Poster Abstracts

Bioinformatics

1
THE GENOMIC DIVERSITY AND PHENOTYPE CONNECTION (GDPC): MIDDLEWARE FOR
GENOMIC DIVERSITY AND PHENOTYPIC DATA
Casstevens, Terry M {1} Buckler, Edward S {2}
{1} Bioinformatics Research Center, North Carolina State University, Raleigh, NC 27695-7566 {2} USDA-ARS,
Department of Genetics, North Carolina State University, Raleigh, NC 27695-7614
Databases and software tools that integrate information on genomic diversity with phenotypic data provide an
opportunity to help bridge genomics and plant breeding. GDPC provides a middleware interface that gives access
to data on genomic diversity (SNPs, RFLP, AFLP, etc.) and phenotypic data that may be collected in field,
genetic, or physiological experiments.
The goal of this project is to simplify access to our data (i.e. loci, taxi, experiments, phenotypes, etc.) by creating a
middle layer between the software tools and the data itself. This middleware separates the tools from the data
format, allowing developers to concentrate on the tools purpose. Software tools request data from the GDPC
interface in a standardized format. Since the middleware manages the interface between the tools and the data
format, the tools, once designed, are able to work with any number of databases. GDPC also provides a way for
the database developers to write connections for their database, allowing any given database to work
automatically with software tools supporting the GDPC interface. This would allow tools to take advantage of
new databases (with supporting connections) without additional software development. Finally, interactions with
the databases would be separated from the analysis and presentation, while allowing software tools to work with
larger sets of genomic and phenotypic data.
The early release version of GDPC is available at http://www.maizegenetics.net.
2
The Emerson Summer Genetics Program: Carrying on the legacy of maize cooperation through high
school and undergraduate outreach
Costich, Denise E. {1} Ahern, Kevin {1} Conrad, Liza {1} Markelz, Nicole H. {1} Sawers, Ruaridh {1}
Sheehan, Moira J. {1} Singh, Manjit {1} Brutnell, Thomas P. {1}
{1} Boyce Thompson Institute for Plant Research
The Emerson Summer Genetics Program has been developed at BTI and Cornell in honor of Rollins A. Emerson,
who in the early 1930's brought together a team of scientists at Cornell including Barbara McClintock, George
Beadle, Charles Burnham and Marcus Rhoades. Through a constant sharing of ideas and materials, these scientists
helped to lay the foundation of genetics over the course of the century. In the same spirit of community fostered
by Emerson, the Emerson Summer Genetics Program aims to increase public understanding of science through
the direct interaction of students, scientists and educators. An important goal of the program is to encourage
scientific discourse among students, teachers, young scientists and senior researchers from the US and abroad.
During the summers of 2001 and 2002, Brutnell lab members mentored a total of 18 undergraduate and high
school students on the Ac Tagging and Phytochrome Projects. In addition to participating in lab and field
experiments, the students also formed journal clubs and gave presentations at the Colonel's Cup Challenge, an
undergraduate symposium named in honor of the founder of the Institute, Colonel William Boyce Thompson.
Most of the students also gave presentations about their experiences at their home institutions. We are currently
recruiting students for Summer '03.

3
MaizeMeister: Phenotypic Data Collection and Seed Management System
Garcia, Arturo {1} Buckler, Edward IV {1,2}
{1} Department of Genetics, North Carolina State University {2} USDA-ARS
Accurate and complete phenotypic data are essential to genetic analysis. Often analysis of quantitative traits
becomes complicated due to large population sizes and the measurement of multiple traits. The vast amount of
data required can be overwhelming. Technology can be exploited to improve the phenotypic data collection
process, be it simply through better tools or information management systems.
MaizeMeister is a system with the ability to collect phenotypes quickly and accurately with little manual data
entry using information-handling technologies such as bar codes, personal data assistants (PDAs), and computers.
Currently, MaizeMeister uses mostly bar codes with its data entry tools to minimize data entry errors.
MaizeMeister is comprised of four components:
Currently, MaizeMeister uses mostly bar codes with its data entry tools to minimize data entry errors.
MaizeMeister is comprised of four components:

• The desktop application has an easy-to-use interface used to design field seasons, enter data via
synchronization with PDAs or digital-output tools, register harvested ears, and keep a seed inventory
record. All data are stored in an Access database.
• A bar coded plant tag with a unique bar code identification number (PlantID) is attached to each plant.
The plant tag has removable sections that can be used for ear and sample collection. The bar code travels
with tissue samples and harvested ears so you can track where the samples/ears came from.
• The PDA data collection program, PhenoData, is a PalmOS application for collecting phenotypic data
in the field and requires a PDA with a built-in bar code reader. PhenoData works in conjunction with a set
of data entry tools for data collection.
• Specifically designed bar coded data entry tools eliminate as much keyboard entry as possible.

All software and source code is freely available for download at www.maizegenetics.net
4
Discovery and Expression Analysis of Alternatively Spliced Genes in Arabidopsis thaliana
George, Robert {1} Vasquez, Adrian {1} Oakley, Fredrick {2} Brendel, Volker {2} Lal, Shailesh {1}
{1}Department of Biological Sciences, Oakland University, Rochester, MI 48309-4476 {2}Department of
Zoology and Genetics, Iowa State University, Ames, IA 50011-3260
Alternative splicing is a process by which some eukaryotic genes encode multiple transcripts via differential
selection of splice sites during pre-mRNA processing. Gene regulation via alternative splicing is extensively
documented in vertebrates and has been linked to important roles in cell growth and differentiation. Genome wide
expression analysis of human ESTs indicates that 30-60% of human genes are alternatively spliced, suggesting
this process plays an integral role in increasing the functional complexity of the human genome. The availability
of the nearly complete Arabidopsis genome sequence along with a large collection of ESTs have provided a novel
resource to study this process in plants from a genome wide perspective. A database of potentially alternatively
spliced Arabidopsis genes discovered by computational analysis is available at the AtGDB
(http://www.plantgdb.org/AtGDB/prj/ZSB03PP/alternativeSplicing/). To test the efficacy of our computer
prediction, we selected 18 genes from AtGDB for experimental analysis. These genes were selected based on two
criteria: (1) the encoded protein products of the genes shared significant sequence similarity to characterized
genes in the database, (2) the EST predicted alternative splicing utilized an alternative donor or acceptor site in
one intron. RT-PCR analysis of total RNA extracted from Arabidopsis root, shoot and flower tissues using
primers flanking the alternatively spliced regions of these 18 genes provided experimental evidence of alternative
splicing in 12 cases. The results of these studies and the impact of alternative splicing on plant gene expression
are discussed.

5
Conserved Noncoding Sequence Comparisons as a Strategy to Identify Candidate Regulatory Elements in
Cereal Gene Promoters
Guo, Hena {1} Moose, Stephen {1}
{1} University of Illinois at Urbana-Champaign
The identification and analysis of conserved noncoding sequences among genes from related species is one
strategy to identify candidate promoter elements that regulate gene expression. We surveyed for CNS among a
large set of orthologous gene sequences from maize, rice, sorghum, wheat and barley to assess their general
properties and utility in identifying promoter regulatory elements. A number of phylogenetic footprinting tools
were evaluated for their ability to predict 30 known regulatory elements present in 10 maize-rice gene pairs, with
the VISTA tool performing best. Comparisons of 81 maize-rice gene pairs found that maize-rice CNS are
relatively short (mean of 11.9-bp) and represent a small proportion (mean 8.7%) of promoter sequence. Extension
of these comparisons to 18 orthologous gene sets containing sequences from at least three cereal species showed
that the frequency, length, and nucleotide substitution rates of cereal CNS are consistent known phylogenetic
distances among the cereals. Our dataset was used to develop criteria to assess the statistical significance of CNS
identified in cereal gene promoters, based on the frequency of conserved sequence blocks in comparisons of
orthologous gene promoters relative to their occurrence in random promoter sequence comparisons. We have also
initiated tests for CNS function in regulating promoter activity and will present our progress to date for CNS
identified in the maize Shrunken2 gene and its rice ortholog.
6
PROTIC : A database and web-based application to manage, analyse and web-publish plant proteome
expression data.
Joets, Johann {1} Houel, Gwenn {1} Ferry-Dumazet, HÈlËne {2} de Daruvar, Antoine {2} Plomion,
Christophe {3} Zivy, Michel {1}
{1} UMR de GÈnÈtique VÈgÈtale du Moulon, Gif-sur-Yvette, France {2} Centre de Bioinformatique de
Bordeaux, France {3} INRA, GÈnÈtique et AmÈlioration des Arbres Forestiers, Bordeaux, France
Identification of gene involved in plant traits is a first step to understand biological processes underlying these
traits. Transcriptome and proteome studies allow the analysis of gene expression in function of plant tissue,
genotype, environmental condition, etc. Then correlation may be evidenced between traits and genes expression
patterns. Other data about these genes such as functional annotation, co-localization with QTL, etc.. must then be
taken into account to support or to invalidate these gene as candidate.
Large international effort have provided community with databases and softwares dedicated to transcriptome data
mining. However, less effort has been done for proteomics. Although proteomics is far from being as automated
and high-throughput as transcriptomics is, the advantages of proteomics are well recognized. Transcript and
protein levels do not always correlate, and post-translational modifications can not be deduced from cDNA
sequence or transcriptome analysis. To our knowledge, databases developed so far for 2-D PAGE and compliant
with plant proteomics do not deal with quantitative data or relations between protein spots. Then new
developments are needed to store and analyse proteome data and to compare them with transcriptome data.
In a first step, we developed the PROTeome bioinformaTICs Protics; a database and a web-based application to
manage, track, query and web-publish proteomics data.
The scope of the database is to store complete sets of data generated by experiments, from experiment design
description to spot identification and quantitative variations. Schema was designed mainly around 2D-gel
electrophoresis, proteins spot quantification and MS identification of spot but the database may also be fed with
results from other protein identification methods like Edman microsequencing.
In order to ease discovering of correlation of protein expression levels between experiments, we included in the
database schema possibilities to aggregate different types of object like plant, sample, gel or spot. Then data from
those groups can be easily extracted whenever needed, for instance to conduct statistical analysis. One of the
important features of proteomics analysis is relations between spot, such as ìthe two spots are the same protein but
in different post-translational modification stateî. These spot relations can be evidenced only by specific analysis.
As this information may be very important when comparing experiments, we managed to store it in the database.
This system can be extended to any type of relation that could exist between two proteins from the same or
different samples. The Protic application can then automatically build relations network using transitivity. Thus if
a non identified spot is related to an identified one by, for example an allelism relationship, then the first one will
inherit of the identification result of the second one and so on.
Database feeding may be achieved by automated uploading of different formatted files or using web forms in case
of low amount of data. Data files are either files from widely used proteomics softwares like Melanie or tabulated
files from spreadsheets.
Protic interface provides users with a set of web forms to query the database and a graphical tool to query 2D-gel
by clicking.
Protic is based on Oracle or Postgresql DBMS and will be freely available.

7
ActionMap : A bioinformatic package for genetic mapping automation
Albini, Guillaume {1} Falque, Matthieu {1} Joets, Johann {1}
{1} UMR de GÈnÈtique VÈgÈtale du Moulon, Gif-sur-Yvette, France
Genetic linkage map computation may be achieved by using different softwares such as MapMaker (Lander,
Green et al. 1987), JoinMap (Stam, 1993) or GMendel (Holloway and Knapp, 1994). We choose MapMaker
because it is a free software adapted to our mapping populations and widely used by the maize genetics
community. But for a high number of loci, genetic mapping remains time-consuming. Furthermore, manipulating
numerous files to import data and export results may cause mistakes. Thus, automation of genetic linkage map
computation is interesting to increase the throughput as well as the reliability of large mapping projects. However,
building a map de novo requires too much human expertise to be easy to automate. But once a robust framework
map is built, MapMaker command steps required to assign new markers to the framework are similar for each
marker. It becomes then possible to automate these command steps. We thus developed ActionMap, which is
made of a relational database to store raw data and mapping results, and a software that automates the assignation
of markers to framework maps. The Actionmap package is used by the GÈnoplante maize high-throughput
mapping project. Framework raw data of maize community mapping project will be added soon to the database.
ActionMap is packaged as a web server, which may be used on line or locally. The users can then access the
database maize and mapping automation software via usual web browsers. The web interface is used to input data,
to set up the automatic process, and to display and export the results.
ActionMap is a at http://moulon.inra.fr/~bioinfo. Package may be used on line or downloaded for free.

Holloway J. L. and S.J. Knapp, 1994, GMendel 3.0 Users Guide, Department of Crop and Soil Science,Oregon
State University, Corvallis OR 97331, USA.
Lander E. S., Green P., et al . (1987). MAPMAKER : an interactive computer package for constructing primary
database maize and mapping automation software via usual web browsers. The web interface is used to input data,
to set up the automatic process, and to display and export the results.
ActionMap is a at http://moulon.inra.fr/~bioinfo. Package may be used on line or downloaded for free.

Holloway J. L. and S.J. Knapp, 1994, GMendel 3.0 Users Guide, Department of Crop and Soil Science,Oregon
State University, Corvallis OR 97331, USA.
Lander E. S., Green P., et al . (1987). MAPMAKER : an interactive computer package for constructing primary
genetic linkage maps of experimental and natural populations. Genomics 1(2) : 174-81
Stam P., 1993. Construction of integrated genetic linkage maps by means of a new computer package: JoinMap.
The Plant Journal 3: 739-744
8
Maize Genome Database - Maize[G]DB – Inputs and Synthesis.
Polacco, Mary {1,2} Hancock, Denis C {2} Sanchez-Villeda, Hector {1} Coe, Edward H {1.2}
{1} USDA-ARS {2} University of Missouri
Curation of hypothesis-driven research and high-throughput sequencing or other projects share many aspects, but
from a database perspective need to be handled differently. (Inputs 1) Reading the literature and extracting
information requires expertise in genetics, biochemistry, physiology, plant development ; extensive familiarity
with the organism being described; expertise in nomenclature and Ontologies with skill in handling conflicts/grey
areas; skill in systematic integration. Typically databases fund such positions, rather than rely on well-intended,
but time-constrained, researchers. (Inputs 2) In contrast high-throughput data is best handled by the funded
project, performing necessary checks of semantic consistency, missing data and relevant validations in a
collaboration with the integrating genome database, eg MaizeGDB. This takes advantage of considerable
expertise on-hand for the research project while in full-tilt, without unduly burdening either the central genome
database or the research project. Examples of each curation mode will be presented, both as practiced and as
foreseen. (Syntheses)
Systematic integration of data permits ready generation of syntheses otherwise difficult to enable.. The IBM
neighbors maps, a consensus genetic map product , grounded on the IBM high resolution genetic map will be
presented.

9
Plant Genomic Resources at NCBI
Smith-White, Brian {1} Resenchuk, Sergey {1} Chetvernin, Vyachesalv {1} Tatusova, Tatiana {1}
{1} National Center for Biotechnology Information
We wish to describe the plant genomic resources available from the NCBI Web site Plant Genomes Central at
http://www.ncbi.nlm.nih.gov/PMGifs/Genomes/PlantList.html. These resources are 1) classic genetic maps for
corn, rice, wheat, oat, barley and soybean, 2) sequence maps for Arabidopsis thaliana and the International Rice
Genome Sequencing Project, 3) text searching of any map data, 4) BLAST-mediated location of potential
homologs. The goal of the current NCBI effort is to prepare a framework for those plants currently with only
genetic maps but where chromosome sequence or transcriptome sequence is anticipated with the next five years.
10
ElucidateIt: A Bioinformatics Workflow and Analysis System
Stapleton, Ann E. {1} Johnson, Kenisha V. {2} Brown, Jeffrey L. {3} Hudson, Thomas C. {2}
{1} Department of Biological Sciences, University of North Carolina at Wilmington {2} Department of Computer
Science, University of North Carolina at Wilmington {3} Department of Mathematics and Statistics, University of
North Carolina at Wilmington
Biologists now have access to vast amounts of DNA sequence and annotation data, but we do not yet have tool
sets that make it easy for a novice to ask biological questions about these large sequence data sets. We are creating
a visual user interface that will enable users to easily set up and answer questions like ìIs there an amino acid
sequence pattern that is present in significant excess?î. This system will allow users to ask about frequency,
correlation, and significance, queries typical of data mining methods, without writing programs. Analysis modules
will also include clustering methods, pattern detection, and new statistical tools. Our system will be especially
useful for crop plant EST collections and other frequently updated sequence sets, as complex queries and analyses
can be logged and thus easily rerun when the databases are updated.

11
Stalking an A-Maize-ing Plant. Corn In Culture And Science; an Integrated Curriculum; Science
Enhancement for K-5 Teachers
Tracy, William F. {1} Steinbach, Shirley {1}
{1} University of Wisconsin-Madison
The Wisconsin Teacher Enhancement Program, UW-Madison, is a national professional development program for
K-14 science and health educators. Courses in the program are designed to enhance teachersí understanding of
biology and provide curriculum directly applicable to the classroom. A key aspect of this program is the teaming
of professors with master teachers for course development and instruction. In 1993, we developed 'Corn in
Culture and Science; An Integrated Curriculum.' Usually it is taught in the summer as a one credit, one-week
module. Focusing on plant biology, agriculture, and environmental science, we provide an integrated curriculum
in which corn is used to teach all subjects. We focus on inquiry based science activities and provide information
and curriculum for using corn in the class. We introduce 'corny' activities and strategies to challenge students in
designing their own experiments and investigations using the scientific method. We provide information on
growth systems for corn in the classroom. Participants gain knowledge in corn physiology, history, geography,
economics, and genetics. Field trips, art activities, and projects, are a part of the course curriculum.
12
Plant Ontologies and the Plant OntologyTM Consortium (POC)
Vincent, Leszek {1} Coe, Edward {2} Polacco, Mary {2}
{1} Plant Science Unit, Department of Agronomy, Curtis Hall, University of Missouri-Columbia, Columbia, MO
65211-7020, USA {2} USDA/ARS Plant Genetics Research Unit, Curtis Hall, University of Missouri-Columbia,
Columbia, MO 65211-7020, USA
Plant genomic databases need to accurately and consistently document features (e.g. gene products, functions,
phenotypes, traits, developmental stages, anatomy [incl. morphology]), using a syntax that facilitates inter-
database searches. This interoperability will enable comparative genomic strategies to elucidate plant functions.
The Plant OntologyTM Consortium (POC) (www.plantontology.org) is applying and extending the Gene
OntologyTM (GO) (www.geneontology.org) paradigm to knowledge domains pertinent to plant taxa. The POC
aims at providing ontologies and controlled vocabularies for monocot and dicot plant taxa - currently Zea mays,
Oryza sativa and Arabidopsis thaliana, but extending to other important taxa in due course. The POC aims to
facilitate the communications, productivity and collaborations amongst the core participants of the POC involved
in developing ontologies and controlled vocabularies for some monocot and dicot taxa. Further aims: further
representation of other plant taxa, numerical growth of participants, ongoing and extended collaboration with the
research of the GO Consortium. It is anticipated that the POC will impact the bioinformatics research of other
national and international plant-based research groups/researchers (e.g. soybean, sugarcane, cassava, potato,
tomato, trees, grains etc.), via the provision of ontology products, community resources and educational inputs. A
sample of ontology and controlled vocabulary for Zea mays (maize/corn) is presented. Funded by NSF Plant
Genome Grant DBI 9872655.

13
Comparative analysis of splicing related proteins in plants
Wang, Bingbing {1} Brendel, Volker {1}
{1} Iowa State University
Although the splicing machinery is generally conserved among eukaryotic organisms, plant pre-mRNA splicing
has several distinct features that may correlate with plant specific splicing factors. In this study, the maize EST
database was searched using 262 Arabidopsis splicing related proteins and their human homologs. A recursive
BLAST strategy was employed to retrieve maize splicing related proteins. In addition, the spliced alignment
program GeneSeqer was used to map the query proteins onto the rice genome to retrieve rice orthologs. For most
of the query proteins, homologs in maize and rice could be identified. This fact further proves the conservation of
the splicing mechanism between animals and plants. However, compared with animal homologs, plant splicing
related proteins have more duplications. More than half of the splicing related proteins were duplicated. Some
proteins such as SAP90 have high copy number in Arabidopsis but low copy number in maize. Interestingly,
many of the duplications happened after the divergence of monocot and dicot plants, indicating that the ancestor
of plants only have one copy of the duplicated gene. It seems that plants may have developed special ways to
regulate splicing during their adaptation to the environment
14
The maize DEK1 calpain sub-domain functions as a cystein proteinase.
Wang,Cunxi {1} Barry, Jennifer {1} Min, Zhao {1} Tordsen, Gabrielle {1} Rao, Guru {1} Olsen, Odd-Arne
{1}
{1} Pioneer Hi-Bred International, A DuPont company, Johnston, IA 50131
Maintenance of aleurone cell fate depends on the activity of the Defective kernel 1 (dek1) gene, encoding a
predicted 240 kDa membrane anchored protein with a carboxyl-terminus encoding a protein with high similarity
to animal calpain domains II and II. In order to characterize the function of DEK1 in cell signaling, we used
protein 3D modeling to show that DEK1 domain II contains a conserved catalytic site similar to animal calpains.
Bacterially expressed DEK1 domain II display weak casinolytic activity that is enhanced by Ca++ , similar to
domain II of animal m-calpain. DEK1 domain II & III exhibited a high level of the activity toward casein as
compared to domain II alone, indicating that domain III plays important role for activation of DEK1. Mutation of
Cys1796 to serine, part of the calpain cartalytic triad, led to a loss of caseinolytic activity, supporting the
conclusion that the DEK1 calpain domain function as a cysteine protease. In situ hybridization analysis shows that
the Dek1 transcript is present in all cell types in developing maize seeds, suggesting that the activity of the DEK1
cystein proteinase is not regulated at the transcription level. The potential role of DEK1 in aleurone signaling will
be discussed.

15
Comparative Physical Maps of Maize and Rice in Gramene.
Ware, Doreen {1} Teytelman, Leonid {1} Zhao, Wei {1} Clark, Kenneth {1} Stein, Lincoln {1}
{1}Cold Spring Harbor Laboratory
Gramene (www.gramene.org) is a comparative mapping database for rice and other cereals. The public resource
leverages the rice genomic sequence to allow researchers working in maize and other monocots to apply
knowledge derived from the study of the rice genome to corresponding syntenic region in their species of interest.
In this abstract we present preliminary results of our comparative physical maps of maize and rice based upon
finger print contig (FPC) physical maps of maize and rice available from the Arizona Genomic Institute and the
MMP project, and from the public rice sequence draft available from the International Rice Genome Sequencing
project (IRGSP). The analysis and displays are based upon the use of the rice genome as a scaffold to align maize
sequences used for genetic and physical mapping studies. Using our mapping protocol
(http://www.gramene.org/documentation/) 61% of the maize sequences were aligned, with 86% aligning to single
rice BACs or PACs. Of the sequences aligned, 98.8% of the sequences fall within a single rice FPC contig. Of the
cornsensus clones mapped to the rice genome over 5,000 have been hybridized to BACs used in the maize FPC
maps as overgo probes, providing reference points to the rice genome. The sequence-based similarities become
map-to-map correspondences within Gramene's comparative map viewer CMap, an open source application
available at www.gmod.org. CMap dynamically generates maps based upon correspondences, in this case, to the
expressed maize genes. The CMap displays are interactive, and provide links to known maize genes, in addition to
rice sequence annotation available for the region of similarity. We welcome suggestions from the maize
community on innovative displays that will provide maize researchers with the functionality to mine the rice
genome for gene sequences of agronomic importance.

Genome Resources

16
The Maize Genome Sequencing Project at the Donald Danforth Plant Science Center
Barbazuk, Brad {1} Whitelaw, Catherine {2} Quackenbush, John {2} Bennetzen, Jeff {3} Schubert, Karel
{1}
{1} Donald Danforth Plant Science Center. 975 N. Warson Rd. St. Louis MO. 63132 USA {2} The Institute for
Genomic Research. 9712 Medical Center Drive Rockville, MD 20850 {3} Department of Biological Sciences,
Purdue University, West Lafayette, IN 47907 USA
Sequencing the maize genome will greatly influence our understanding of the molecular basis of important
agronomic traits, gene regulation, transposable element function, genome evolution, plant development and
biology. However, the large size of the maize genome and the expectation that upwards of 80 % of the genome is
represented by repetitive elements has prompted the examination of sequencing technologies expected to target
gene rich regions as an alternative to whole genome sequencing. A consortium consisting of the Donald Danforth
Plant Science Center, TIGR, Purdue University and Orion Genomics has been awarded an NSF plant genome
grant to develop and evaluate high-throughput and robust strategies to isolate and sequence maize genes. The
project objectives are to examine the maize gene space by analyzing sequence obtained from a methyl-filtered
subclone library produced by Orion, and one high Cot library produced by Purdue University. The methyl-filtered
libraries are clones with inserts composed of non-methylated maize genomic DNA. Methyl-filtered libraries have
been demonstrated to be gene rich. High Cot selection exploits the relatively low abundance of the gene
sequences, which are present in a small number of copies in the genome. Clone sequencing and sequence
processing is being performed at TIGR, while the Danforth Center will provide an overall analysis of the gene hit
rate/coverage of each method. A detailed overview of the projectís methods and analysis goals, as well as current
subclone library produced by Orion, and one high Cot library produced by Purdue University. The methyl-filtered
libraries are clones with inserts composed of non-methylated maize genomic DNA. Methyl-filtered libraries have
been demonstrated to be gene rich. High Cot selection exploits the relatively low abundance of the gene
sequences, which are present in a small number of copies in the genome. Clone sequencing and sequence
processing is being performed at TIGR, while the Danforth Center will provide an overall analysis of the gene hit
rate/coverage of each method. A detailed overview of the projectís methods and analysis goals, as well as current
progress and results of the analysis to date will be presented.

17
HIGH RESOLUTION PHYSICAL MAPPING OF THE MAIZE GENOME
Bharti, Arvind K. {1} Butler, Ed {2} Yu, Yeisoo {2} Goicoechea, Jose L. {2} Fuks, Galina {1} Nelson, Will
{2} Soderlund, Cari {2} Wing, Rod {2} Messing, Joachim {1}
{1} The Plant Genome Initiative at Rutgers, Waksman Institute, Rutgers, The State University of New Jersey,
Piscataway, NJ 08854, USA (http://pgir.rutgers.edu) {2} Arizona Genomics Institute, 303 Forbes Building,
University of Arizona, Tucson, AZ 85721, USA (http://www.genome.arizona.edu)
Maize is one of the most important crops worldwide. Its genome is composed of 2,500 Mb of DNA spread over
10 chromosomes. An effort has been funded through NSF (The National Science Foundation) to develop
resources and infrastructure for sequencing the maize genome (http://mips.gsf.de/proj/maize), which would
provide invaluable data for refining molecular breeding strategies. A pre-requisite to genome sequencing is a
high-resolution genetic and physical map. The HindIII (ZMMBBb), EcoRI and MboI (ZMMBBc) BAC libraries
(constructed at CUGI and CHORI, respectively) of Zea mays ssp. mays cv B73 form the main basis for the maize
physical map. Together they consist of 449,989 clones (average insert size 155 kb). As of now, >256,000 BACs
have been fingerprinted by HindIII digest and agarose gel electrophoresis at AGI
(http://www.genome.arizona.edu/fpc/maize) into 4,518 contigs. In an effort to further bring down this number, a
high-resolution fluorescent fingerprinting method capable of identifying smaller overlaps is being carried out. To
have a deeper coverage (28X), all 449,989 BACs are being subjected to HICF (High-Information-Content
Fingerprinting) using type IIS restriction enzyme and resolved on ABI 3700 capillary sequencers. The fingerprints
would be assembled using FPC. It is expected that resolution of the HICF-generated contigs would be high
enough to select the entire Minimum Tiling Path (MTP) at once as opposed to the clone-by-clone approach. To
evaluate the MTP generated by HICF, 14 BACs would be sequenced from each chromosome. In addition, end
sequencing would be carried out with more than half of the BACs from both ends. These STCs (Sequence Tagged
Connectors) would serve as an in-silico mapping tool. Collinearity of STCs with the rice genomic sequence can
also be used to order FPCs on the genetic map. Furthermore the library of sequence repeats thereby created would
provide invaluable data on the sequence organization of the maize genome.
18
Activator (Ac) Mutagenesis in Maize
Conrad, Liza {1} Kolkman, Judith {1} Ahern, Kevin {1} Farmer, Phyllis {1} Singh, Manjit {1} Sawers,
Ruairidh {1} Hardeman, Kristine {1} Lebejko, Sara {2} Chomet, Paul {2} Brutnell, Thomas {1}
{1} Boyce Thompson Institute, Ithaca NY, USA; {2} Monsanto Co., Mystic CT, USA.
The utility of Ac as a mutagenesis tool has been hindered by the propensity of Ac to move to linked sites and the
limited number of Ac donor sites in the maize genome. As part of an NSF-funded Plant Genome Program grant,
we have been distributing and mapping the transposable element Ac throughout the maize genome. The goal of
this project is to create approximately 200 near-isogenic W22 lines, each containing a single Ac element at
approximately 20 cM intervals. In the initial phase of this project, approximately 200 unlinked transpositions were
generated from two separate locations in the genome. Several methods are being employed to facilitate the
cloning of sequences flanking these Ac insertions. We are examining the efficiency of both short-range and long-
range inverse-PCR protocols to amplify Ac-flanking sequences from 300bp to 8 kb away from the site of Ac
insertion. To date, 53 anchor Ac insertions have been cloned and verified as active and 41 Ac insertions have been
mapped. Ac elements have been placed on each of the 10 maize chromosomes and on most chromosome arms. In
the second phase of this project, approximately 1800 transpositions have been generated from anchor Ac's
positioned on 8 of the maize chromosomes. We have initiated two-point testcross linkage analysis to map
approximately 1000 of these transpositions relative to the donor Ac's. Insertions that fall within 10 to 30 cM of the
donor Ac will be further characterized. A subset of these insertion sites will be cloned and placed on the IBM94
map. Together, these lines and methodologies to rapidly sequence Ac insertion sites will provide unique and
valuable resources to the maize genetics community.

19
Finishing of the publicly-funded physical map of ëB73í at the Arizona Genomics Institute and Arizona
Genomics Computational Laboratory
Butler, Ed {1} Goicoechea, Jose Luis {1} Engler, Fred {2} Nelson, Will {2} Soderlund, Cari {2} Wing, Rod
{1}
{1}Arizona Genomics Institute, University of Arizona {2}Arizona Genomics Computational Laboratory
As a member of the Maize Mapping Project (www.maizemap.org), we are nearing completion of a physical map
for maize, constructed by FPC assembly of DNA fingerprints from HindIII, EcoRI and MboI BAC libraries of the
inbred line B73. More than 291,000 BAC clones have been fingerprinted by HindIII digestion and agarose gel
electrophoresis. The resulting assembly provides a15x coverage of the maize genome. Over 13,600 molecular
markers have been placed onto the map by hybridization. Manual editing of the map and contig merging are in
progress and expected to be completed in late spring of 2003. The unedited assembly can be viewed via WebFPC
at www.genome.arizona.edu/fpc/maize; the data set and FPC may be downloaded, as well. Approximate
chromosomal locations of genetic markers and contigs may be viewed at this site using the WebChrom tool. In
addition to the production of a high resolution map of B73 based upon agarose fingerprinting, AGI is also
collaborating in a re-analysis of the same BAC libraries using capillary electrophoresis-based high information
content fingerprinting (HICF). This re-analysis is part of an NSF-supported project to increase the resolution of
the maize physical map and create an enhanced sequence-ready platform for maize. The BAC libraries
(ZMMBBb and ZMMBBc) may be obtained from Clemson University Genomics Institute
(http://www.genome.clemson.edu/orders/). This work was supported by NSF award #9872655
20
FTIR and NIR spectroscopy to identify mutants in cell wall biogenesis
Vermerris, Wilfred{1} Bleecker, Anthony{2} Koch, Karen{3} McCann, Maureen C.{1} McCarty,
Donald{3} Patterson, Sara{2} Reiter, Wolf-Dieter{4} Reuhs, Bradley{1} Staiger, Chris{1} Thomas,
Steven{5} Settles, A. M.{3} Hannah, L.C.{3} Carpita, N{1}
{1} Purdue University {2} University of Wisconsin {3} University of Florida {4} University of Connecticut {5}
National Renewable Energy Laboratory
The plant cell wall is of major importance because of its role in providing support to tissues, defense against pests
and pathogens, and as a source of food, fibers, and fuels. Plant cell walls are composed of independent but
interacting networks of carbohydrates, proteins, and aromatic substances. Interacting with this complex matrix are
several hundred enzymes and other proteins that carry out many functions, including wall assembly and
disassembly. As part of a multi-institutional project funded by the NSF Plant Genome Program (No. 0217552) we
are targeting genes involved in cell wall biogenesis in both maize and Arabidopsis, two species that represent very
different types of cell wall. In maize, we are screening the UniformMu transposon-tagging population . The
UniformMu families share a common genetic background (W22) and genetic mechanism for Mu inactivation. We
are screening coleoptiles using Fourier-transform infrared (FTIR) spectroscopy to identify changes in the primary
cell wall. Near-infrared (NIR) spectroscopy is used to screen field-grown plants to identify mutants affected in
primary and secondary cell wall composition. Putative mutants are self-pollinated and backcrossed to W22. After
the heritability of the mutation has been confirmed, the mutated gene will be cloned using the Mu-TAIL protocol.
In parallel, a reverse genetics approach is implemented in order to observe the mutant phenotype resulting from
insertional mutagenesis in genes that impact cell wall biogenesis. Confirmed mutants will be subjected to detailed
chemical and cytological analyses in order to define the function of the targeted genes. Resulting data will be
available on the following web site: http://cellwall.genomics.purdue.edu These combined efforts will eventually
result in a detailed picture of the many aspects of cell wall biogenesis. This in turn is expected to contribute
significantly to the society at large and agriculture in particular.

21
The Zea mays microarray resource
Elumalai, Rangasamy {1} Gchachu, Maya {1} Watson, Tom {1} Chandler, Vicki {1} Galbraith, David {1}
{1} University of Arizona, Tucson, AZ 85721, USA
At the University of Arizona, we have been providing a microarray resource for the academic and not-for-profit
community for the least three years, as part of the NSF Plant Genome project entitled: Maize Gene Discovery,
Sequencing and Phenotypic Analysis (Virginia Walbot, P.I.). The microarrays comprise DNA elements that are
robotically deposited onto glass microscope slides. The elements are produced as amplicons from cDNA clones,
previously sequenced at Stanford University, using universal primers. To date, a total of 128,787 ESTs have been
sequenced, from 20 different libraries. Sequence analysis, done at Iowa State University (Volker Brendel), has
allowed contig identification and assembly, and the definition of a Unigene set. We have provided microarrays for
four of the individual libraries (endosperm, ear tissue, root, and leaf primordia) as well as for the unigene set
(~18,000 unigenes), which has been released in subsets over the last year. To date over 1,100 microarrays have
been distributed to 80 researchers world-wide at a discounted cost relative to production. This project is funded
through September 2003. Immediate future plans are amplification of another 7,000 ESTs generating a total of
25,000 amplified unigenes and consolidation of these onto 1 slide, which will be offered at $100/slide. Details on
the performance of the slides and the project timetable will be presented on the poster.
four of the individual libraries (endosperm, ear tissue, root, and leaf primordia) as well as for the unigene set
(~18,000 unigenes), which has been released in subsets over the last year. To date over 1,100 microarrays have
been distributed to 80 researchers world-wide at a discounted cost relative to production. This project is funded
through September 2003. Immediate future plans are amplification of another 7,000 ESTs generating a total of
25,000 amplified unigenes and consolidation of these onto 1 slide, which will be offered at $100/slide. Details on
the performance of the slides and the project timetable will be presented on the poster.
22
Functional Genomics of Maize Centromeres
Dawe, R. Kelly {1} Jiang, Jiming {3} Birchler, James A. {2} Parrott, Wayne A. {1} Jin, Weiwei {3} Kato,
Akio {2} Marshall, Joshua B. {1} Mroczek, Rebeccaa J. {1} Nagaki, Kiyotaka {3} Phan, Bao H. {1} Theuri,
James {2} Topp, Christopher {1} Zhong, Cathy X. {1}
{1} University of Georgia {2} University of Missouri {3} University of Wisconsin
The poster will describe results from our collaborative work to better understand maize centromeres. It will be
broken down as follows: 1) Objective: To sequence 300 kb of maize centromeric DNA from the A centromeres,
and develop a database of repetitive DNA elements associated with maize centromeres. Status: The sequencing is
complete and provide further support for the idea that centromeres are composed of satellite repeats and
retrotransposons, many of which are centromere-specific. 2) Objective: To fully sequence the centromeres from
two derivatives of the maize B chromosome that have suffered a series of centromere deletions. Status: We have
constructred libraries and carried out about half of the sequencing necessary. Fiber-FISH approaches are helping
in this effort. 3) Objective: Identify putative functional repeats by chromatin immunoprecipitation using
antibodies to maize CENH3 (Centromeric Histone 3). Status: We demonstrated that CentC and the centromeric
retrotranspson CRM interact strongly with CENH3, providing the first evidence in plants that centromere-
localized DNAs are indeed functional components of the centromere/kinetochore complex. 4) Objective:
Transform maize with promising centromeric BACs to determine whether they organize a kinetochore and
generate chromosome movement. We hope to use the information to develop first-generation artificial
chromosomes. Status: Transformation efforts with maize have been slow. We have turned to rice transformation
to speed the process, and have now successfully incorporated rice centromeric DNA into rice with very
encouraging results.

23
Evolutionary genomics of maize
Buckler, Edward S. {1,2} Doebley, John {3} Gaut, Brandon {4} Goodman, Major {1} Kresovich, Stephen
{4} Muse, Spencer {1} Weir, Bruce {1}
{1} North Carolina State University {2} USDA-ARS {3} University of Wisconsin-Madison {4} University of
California-Irvine {5} Cornell University
The complete sequence of crop genomes or EST libraries for a crop species is changing the landscape in which
plant breeders, biotechnologists, and biologists operate. Among the wealth of new opportunities will be one to
exploit genomic sequence data to better utilize naturally occurring variation for important traits in crops. As new
genes of agronomic importance are identified, we need to identify the amount, distribution and nature of
functional variation in these genes that exist in the germplasm pools of crop species. Our project is creating the
necessary infrastructure to do this in maize. We will define where in the maize germplasm pool and where in the
maize genome useful variation is most apt to be found. We are examining nucleotide diversity in a set of
candidate genes for agronomic traits and testing whether specific DNA sequence polymorphisms can be
associated with phenotypic variation for these traits. Our overall goals are to better understand the distribution of
genetic diversity within the maize genome and to facilitate the identification of polymorphisms at the nucleic acid
level in candidate genes that control variation at the phenotypic level in agronomic traits.
24
AN INTEGRATED GENETIC AND PHYSICAL MAP FOR MAIZE
Gardiner, Jack {1} Cone, Karen {2} McMullen, Mike {3} Polacco, Mary {3} Davis, Georgia {1} Yim,
Young-Sun {1} Soderlund, Cari {4} Wing, Rod Wing {4} Butler, Ed {4} Coe, Ed {3}
{1} Dept. of Agronomy, {2} Div. of Biological Sciences, University of Missouri {3} USDA-ARS, Columbia,
Missouri {4} University of Arizona
The goals of the Maize Mapping Project (MMP) are to develop genetic, physical, and database resources for the
maize research community. The centerpiece of this effort is a comprehensively integrated genetic and physical
map. To develop a deep-coverage physical map, 460,000 B73 BACs from HindIII, EcoRI, and MboI libraries
have been fingerprinted to generate approximately 300,000 usable HindIII fingerprints that have been assembled
into contigs using FPC. The final FPC build, producing 4518 contigs, has been completed, and manual editing is
underway. Manual editing is expected to reduce the number of contigs further. Concurrently, a high-resolution
genetic map containing over 1800 RFLP and SSR markers has been generated. Currently, the task at hand is to
integrate these two maps using hybridization, PCR, and in silico approaches. Over 9300 EST unigenes have been
anchored to the physical map via 40 bp overgos, which are gene-specific hybridization probes. Selective SNP
development and genetic mapping is focused on EST unigenes, which will allow unambiguous anchoring of
into contigs using FPC. The final FPC build, producing 4518 contigs, has been completed, and manual editing is
underway. Manual editing is expected to reduce the number of contigs further. Concurrently, a high-resolution
genetic map containing over 1800 RFLP and SSR markers has been generated. Currently, the task at hand is to
integrate these two maps using hybridization, PCR, and in silico approaches. Over 9300 EST unigenes have been
anchored to the physical map via 40 bp overgos, which are gene-specific hybridization probes. Selective SNP
development and genetic mapping is focused on EST unigenes, which will allow unambiguous anchoring of
contigs in a maximally efficient manner. BAC pools are being utilized to allow genetically mapped SSRs to be
anchored to the BAC physical map. In silico approaches have been fruitful in identifying previously mapped SSRs
and RFLPs in the EST unigene set that have been anchored to the BAC physical map via overgo probes. See
http://www.maizemap.org/ and http://genome.arizona.edu/fpc/maize/

25
Full-length cDNA sequencing for the functional genomics of endosperm development
Lai, Jinsheng {1} Dey, Nrisingha {2} Beraft, Philip {2} Kim, Cheol-Soo {3} Larkins, Brian {3} Rudd,
Stephen {4} Meyer, Klaus {4} McCarthy, Don {5} Messing, Joachim {1}
{1} Waksman Institute, Rutgers University, Piscataway, NJ08854 {2} Dept. of Zoology and Genetics, Iowa State
University, Ames, IA50011 {3} Dept. of Plant Science, University of Arizona, Tucson, AS 85721 {4} Munich
Information Center for Protein Sequences, Neuherberg at Munich, Germany {5} Vegetable Crops Department,
University of Florida, Gainesville, FL32611
Maize (Zea mays L.) is one of the most important cereal crops in the world. Central to this productivity is the
endosperm of the maize kernel. Endosperm not only makes up 85% of the seed weight, but also is a unique
triploid tissue that results from double fertilization. To identify specifically the genes involved in endosperm
development, we made cDNA libraries from tissue of a specific genetic background that has been micro-dissected
from different developmental stages, 4 to 23 days after pollination. The tissue is derived from the W22 inbred
used to generate UniformMu population so that tagged genes can be linked to cDNA sequences. The libraries
have been normalized by hybridizing the EST colonies with two probes 1) cDNA prepared from endosperm
mRNA and 2) a pool of already sequenced cDNA clones. In the first batch, both 3' and 5' ends of about 8,000
clones have been sequenced. Assembling the sequences from the 3' end showed a redundancy of roughly 50%.
Out of 3,861 tentative unigenes from the first batch, we identified 1,378 (35%) unique sequences that are not yet
present in the maize EST database. Initial assembly of the 5' and 3' end sequences showed that over 50% of the
clones are longer than 1.5 kb. Comparative analysis is being used to identify full length cDNAís and to estimate
remaining gap sizes.
26
Sequence and phylogenetic analysis of the fie-orp intervals in the two subgenomes of maize with sorghum
and rice as a reference
Lai, Jinsheng {1} Ma, Jianxin {2} Miesak-More, Barbara {1} Ramakrishna, Wusirika {2} Bennetzen,
Jeffrey L. {2} Messing, Joachim {1}
{1} Waksman Institute, Rutgers University, 190 Frelinghuysen Rd., Piscataway, NJ08854 {2} Dept. of Biological
Sciences, Purdue University, W. Lafayette, IN 47907-1392
We have used chromosome walking to connect two closely mapped loci in maize based on deep-coverage BAC
libraries of maize inbred B73. Both loci are duplicate factors and known as fie1/fie2 and orp1/orp2. The two
intervals are on chromosome 4 and 10 and have a length of 358 kb and 285 kb, respectively. Although these
intervals have arisen from tretraploidization, their sequences differ significantly. As a reference we isolated and
sequenced the same interval from sorghum (220 kb); the interval from rice (136kb) was available from the
International Rice Genome Sequencing Project. Changes in these intervals affect gene density, gene content, gene
order, gene orientation, and retrotransposon density. There appears to be greater gene conservation of the same
interval of maize chromosome10, sorghum, and rice than maize chromosome 4. Relative to maize chromosome 4,
the other intervals not only have a dramatically lower retrotransposon density and additional genes but some of
the genes are also amplified. Furthermore, there are also non-collinear genes in the same interval of maize
chromosome 10, sorghum, and rice. An interesting aspect is that the fie homolog is tandemly duplicated in
sorghum and rice, but their upstream copies are conserved on maize chromosome 10, while the downstream
copies are conserved on maize chromosome 4. Phylogenetic analysis of the duplicate factors in maize and their
counterparts in sorghum and rice provide new insights in the evolution of the maize genome.

27
Molecular analysis of Mu-insertions derived from smk mutants.
Latshaw, Susan (1) Wu, Shan (1) Settles, A. Mark (1) Hannah, L. Curtis (1) Koch, Karen E. (1) McCarty,
Donald R. (1)
(1) University of Florida
The Robertsonís Mutator transposon system is a powerful tool for functional genomics. Mu elements
preferentially insert in low-copy regions of the maize genome and transposon activity can be controlled
genetically. The UniformMu transposon-tagging population was created by introgressing Robertsonís Mutator
into color-converted W22. By purging visible seed mutations from the Mu-active parents in each generation, we
ensured independence of the large collection of seed mutants generated by this population. For this report,
fourteen smk mutants from the fourth and fifth backcross generations were selected for molecular analysis.
MuTAIL-PCR products were amplified, and these flanking sequences were cloned to generate 14 microlibraries
for single-pass sequencing. We describe the population of Mu-flanking sequences recovered from these lines and
present progress towards identifying candidate genes for the smk loci. In addition to aiding identification of
mutant genes, our database of sequence-tagged insertion mutations provides an important reverse genetics
resource. These sequences will be available in the microlibrary database at www.endosperm.org.
28
Functional Genomics of Chromatin Genes in Maize
McGinnis, Karen {1} Chandler, Vicki {1} Napoli, Carolyn {1} Kaeppler, Heidi {2} Kaeppler, Shawn {2}
Springer, Nathan {2} Cone, Karen {3}
{1} University of Arizona, Department of Plant Sciences {2} University of Wisconsin, Department of Agronomy
{3} University of Missouri-Columbia, Division of Biological Sciences
Chromatin level control of gene expression is an important regulatory component for growth and development in
all living organisms. To better understand this regulation in plants, we are studying the function of the entire
known complement of chromatin proteins in maize. The goal of the NSF-funded Plant Genome Research Project
on Functional Genomics of Plant Chromatin Genes is to provide information and tools to the research community
to enhance the study of the role of chromatin structure and modification. Using known chromatin genes as queries
we have searched maize ESTs, identifying to date 239 genes that have been assigned to 42 functional classes on
the basis of homology or domain structure. With the ultimate goal of acquiring additional mRNA sequence for
each targeted gene, EST clones were obtained and sequenced, and, in some cases, RACE has been performed.
Additional sequence is being submitted to Genbank and is posted on the project website (http://chromdb.org). For
three functional classes, the additional sequence information has been utilized to analyze the phylogenetic
relationships of these genes in maize, Arabidopsis and other model organisms. The chromosome location of each
gene is being mapped relative to known markers in the maize genome. Constructs have been made to target 128
chromatin genes for silencing using a plasmid that will generate dsRNA, which is expected to interfere with the
expression of the target homologous mRNA. These constructs have been introduced into HiII embryo tissue using
microprojectile bombardment, herbicide selection of callus, and regeneration. The resulting dominant negative
mutants are being assayed to confirm silencing of the targeted gene, and to determine if expression of the dsRNA
targeted to chromatin genes impacts other epigenetic phenomena, such as DNA methylation, epimutations,
paramutation and transgene silencing. Gene sequences, phylogenetic trees, mapping information, assay data and
dominant negative mutants will be made available to the research community through the project website
(http://chromdb.org). This poster presents an update on the progress of this project.

29
SEQUENCE COMPARISONS OF MAIZE AND RICE MITOCHONDRIAL GENOMES AND THE
REARRANGEMENTS WITHIN THREE MAIZE MITOCHONDRIAL STRAINS
Minx, Patrick {1} Sandra Clifton {1} Kathleen Newton {2} Melissa Rugen {2} Louis Meyer {2} Christiane
Fauron {3} Mike Gibson {3}
{1} Washington University Genome Sequencing Center {2} University of Missouri {3} University of Utah
To date very few plant mitochondrial genomes have been completely sequenced. Those sequenced include a
liverwort, Marchantia polymorpha, and two dicots, Arabidopsis thaliana and Beta vulgaris. Now the
mitochondrial genomes of two monocots, Zea mays and Oryza sativa have been sequenced. The sequencing of
maize includes three mitochondrial strains: fertile strains NB (B37) and NA (A188), as well as cytoplasmic male
sterile strains CMS-C. The strains provide the first analysis of gene rearrangements in closely-related plant
mitochondria. Sequences of maize strains are compared and illustrated to reveal major rearrangements utilizing
known genes as markers. Gross rearrangements occur between maize strains, but sequences are otherwise highly
conserved. Sequence homologies within maize mitochondrial genomes and rice are also examined. The major
duplications within each genome are presented.
30
A RADIATION HYBRID SYSTEM FOR THE GENETIC AND PHYSICAL MAPPING OF THE MAIZE
GENOME
Phillips, Ronald {1} Rines, Howard {2} Retzel, Ernest {3} Kynast, Ralf {1} Okagaki, Ron {1} Odland, Wade
{1} Grannath, Shannon {1} Jacobs, Morrison {1} Schmidt, Christina {1} Stec, Adrian {1}
{1} Department of Agronomy and Plant Genetics, University of Minnesota {2} USDA-ARS and Department of
Agronomy and Plant Genetics, University of Minnesota {3} AC-Research Computing, University of Minnesota
The goal of our project is to generate and use chromosome-specific radiation hybrid (RH) maps with high
physical resolution for each of the ten different maize chromosomes. We are developing the RH maps from DNA
panels of oat plants that carry single maize chromosome segments in the form of added maize deletion
chromosomes or oat-maize translocations. The maize chromosome aberrations are generated by gamma-
irradiating monosomic oat-maize chromosome addition seeds. Monosomic additions are produced by
backcrossing specific disomic oat-maize chromosome addition plants to their parental oats. From a large number
of oat x maize crosses, fertile disomic oat-maize addition lines have been recovered for all maize chromosomes
with the exception of chromosome 10. A ditelosomic for the short arm of chromosome 10 has recently been
recovered. To date, we are finalizing RH low-resolution panels allowing marker assignment to one of 10 to 20
segments along each of maize chromosomes 2, 4 and 9. Seed and DNA is being produced for distribution. Such
panels serve high-throughput mapping of large marker numbers. We are beginning studies with the 30 and 50
additional chromosome 2 and chromosome 9 RH lines available to determine the eventual number of RH lines to
produce a high resolution, 5 Mbp or less, panel. A panel of five RH lines has been produced that divides
chromosome 6 into five segments for low resolution mapping; additional lines are being isolated. Induced
chromosome breaks in maize chromosome 3 RH lines tend to occur in a small genetic region that represents
approximately half of the short arm. This material is based upon work supported by the National Science
Foundation under Grant No. 0110134. Any opinions, findings, and conclusions or recommendations expressed in
this material are those of the author(s) and do not necessarily reflect the views of the National Science
Foundation.

31
Methylation-Filtration results in gene enrichment in plants but not in animals because plant genes are
hypomethylated and animal genes are often methylated
Rabinowicz, Pablo{1} O'Shaughnessy, Andrew{1} Palmer, Lance{1} Balija, Vivekanand{1} Nascimento,
Lidia{1} Dedhia, Neilay{1} McCombie, W. Richard{1} Martienssen, Rob{1}
{1} Cold Spring Harbor Laboratory, 1 Bungtown Rd., Cold Spring Harbor, New York, 11724.
The maize and other crop genomes are much larger and repetitive than those of rice and Arabidopsis, which have
been fully sequenced. For this reason full sequencing of complex genomes is currently unaffordable and
alternative approaches are being developed to sequence and map their genic fraction or gene space. One of these
approaches, called Methylation-Filtration, has been developed in our lab and due to its preliminary success,
Methylation-Filtration became one of the two technologies selected to carry out a large scale NSF-funded, gene-
targeted maize genomic sequencing project. The technology is based on the fact that the repetitive (intergenic)
DNA is methylated in plants, while low copy DNA (i.e. genes) is hypomethylated. This differential methylation
can be used to select genes using E. coli methylation-restriction systems, in particular McrBC. It has been
reported that genes can also be methylated. Such genes would be excluded from methylation-filtered libraries.
However, although gene methylation is frequent in animals, the methylation pattern of few genes has been studied
in plants and methylation was usually found towards the ends of the genes. Using McrPCR, a new approach to
study DNA methylation, we showed that most exons are unmethylated in plants while most of animal exons are
methylated. These results show that most plant genes can be isolated using Methylation-Filtration. Furthermore,
they support the idea that methylation is involved in the control of transposon activity in plants while in animals
may be related to other biological functions.
32
Regulation of Inflorescence Architecture in Maize
Rocheford, Torbert {1} Brendel, Volker {2} Jackson, David {3} Kellogg, Toby {4} Martienssen, Rob {3}
Schmidt, Bob {6} Vollbrecht, Erik {3} McSteen, Paula {6} Chuck, George {6} Satoh, Namiko {3} Wang,
Bing-Bing {2} Whipple, Clinton {5} Gallavotti, Andrea {5} Stanfeld, Sharon {5} Bandong, Simona {5}
Malcomber, Simon {4} Penly, Anya {4} Kleweis, Shelby {4} Doust, Andrew {4} Wassom, James {1} Schultz,
Jennifer {1} Walbot, Virginia {7} Hake, Sarah {6}
{1} University of Illinois, Urbana {2} Iowa State University {3} Cold Spring Harbor {4} University of Misouri,
St. Louis {5} University of California, San Diego {6} University of California, Berkeley {7} Stanford University
St. Louis {5} University of California, San Diego {6} University of California, Berkeley {7} Stanford University
Our project is focused on identifying genes that regulate tassel and ear infloresecence architecture in maize. We
hope to use these genes to understand the same developmental processes in other grasses. A number of different
approaches are being employed to find these genes and determine their function. EMS mutagenized populations
have been made in two different inbreds and are being screened. Populations made by other NSF funded projects,
such as RescueMu and UniformMu, are being examined for new inflorescence mutations. We are identifying QTL
that regulate inflorescence architecture in mapping populations and relating QTL regions to map position of
known mutants to identify candidate genes. We have initiated collaborative associative genetic studies for some
candidate genes by evaluating allelic variation in a diverse set of inbreds. These inbreds are being studied by
Major Goodman and Ed Buckler as part of the NSF funded Evolutionary Genomics project. cDNA libraries
enriched in genes expressed at early stages of inflorescence development have been prepared and subtracted.
These genes have been arrayed with the Maize Gene Discovery project for expression analysis. We have also
synthesized gene specific oligos for spotting on slides and are in the process of testing this method for expression
analysis. We are identifying genes in other grasses that are orthologous to genes whose function that have already
been characterized in maize.

33
MaizeGDB: A Next Generation Maize Database
Seigfried, Trent {1} Polacco, Mary {2} Brendel, Volker {1} Brekke, Michael {1} Campbell, Darwin {1}
Dong, Qunfeng {1}
{1} Iowa State University {2} University of Missouri - Columbia
MaizeGDB (Maize Genetics/Genomics DataBase) is an on-going project initiated by the USDA-ARS. The goals
of this project are (1) to build a next generation maize database providing curated and integrated data such as
sequences, maps, genetic markers, phenotypes; and (2) to provide a comprehensive online workbench for maize
biologists to analyze the data. Since our development work started on April 1, 2002, we have begun to integrate
the two existing major maize databases, ZmDB (http://www.zmdb.iastate.edu) and MaizeDB
(http://www.agron.missouri.edu). Data sets from ZmDB and MaizeDB have gone through a process of evaluation
and excavation before being added into MaizeGDB's re-designed schema. We have ported MaizeDB and ZmDB
data into the Oracle-based MaizeGDB. A new intuitive web interface has been developed that provides easy
access to the data and related information. Analytical tools are embedded within data display to facilitate in-depth
study. Other developments include protocols and standards for data sharing, i.e., XML specifications for various
data types and text downloads. We are developing web-based curation tools for both designated experts and
general researchers. We work closely with a nation-wide MaizeGDB Steering Committee on both the scientific
and technical aspects of the database. In addition, the Steering Committee members have been serving as beta
testers of MaizeGDB as well as guiding the site development to meet the needs of the maize research community.
MaizeGDB is publicly available for testing at http://www.maizegdb.org.
34
Applications of oligonucleotide microarrays for maize functional genomics
Springer, Nathan M. {1} Kaeppler, Shawn M. {1}
{1} University of Wisconsin-Madison
Two prototype oligonucleotide microarrays have been synthesized using a NimbleGen maskless array synthesizer
(MAS). The first array was designed to test the utility of oligonucleotide microarrays for SNP detection in maize
mRNA or genomic DNA samples. We report the effects of oligonucleotide length, position of the SNP within the
oligonucleotide, and nucleic acid source on signal intensity and on the level of discrimination between SNPs. The
second prototype array was designed to test the utility of oligonucleotide microarrays for expression analysis in
maize. Specifically, we designed this array to gain an understanding of the ability to separate the expression
pattern of closely related genes and to compare the results with those obtained using cDNA spotted arrays.
Preliminary analysis of the results obtained using both prototype arrays will be presented along with a discussion
of potential future applications of oligonucleotide microarrays such as EMS mutation detection and custom-made
expression arrays.

35
Maize Gene Discovery Project -- Update 2003
Walbot, Virginia {1}
{1} Stanford University
Assembly of maize ESTs from the MGDP and other sources at ZmDB (zmdb.iastate.edu) defines ~32,000 genes,
and most are in contigs. There is nearly complete coverage of the maize life cycle of maize. We have started 5'
and 3' end sequencing from a commercially prepared, normalized 'full length' library with diverse tissue sources;
the Gateway compatible vector permits later manipulations without restriction enzymes. Primer walking will be
used to close a subset of the cDNAs. cDNA microarrays using collections from four libraries (immature ear,
seedling root, leaf primordia, endosperm) and Unigenes (eventually totaling ~25,000 cDNAs) can be ordered
through ZmDB. A second aspect of functional genomics is recovery of mutants and RescueMu insertion alleles.
Grids of up to 48 rows x 48 columns are used for Mu tagging and construction of individual row and column E.
coli libraries of RescueMu insertions in library plates. These plates are screened by PCR to identify which plants
have insertions into genes of interest; order the corresponding seed from the Maize Coop. We have sequenced
either the rows or column libraries of Grids G, H, I, K, M, P, and S with AA and Q partway completed,
identifying thousands of likely germinal insertions (project analysis picked 25 'putatives' and all were transmitted
to the next generation). >30 MB of genomic sequence has been coassembled with ESTs showing that 2/3 of likely
germinal RescueMu sites match a maize EST; this confirms the strong preference of Mu insertion into or near
genes. Most mutations in the selfed progeny of tagging grids are caused by standard Mu elements; as a
community resource we are scoring the seed, seedling, and adult segregating phenotypes, which are found in a
searchable PhenotypeDB at ZmDB. By project completion we expect data from ~48,000 ears, ~15,000 seedling,
and 10,000 adult families.
36
Functional Genomics of Chloroplast Biogenesis: The Photosythetic Mutant Library.
Walker, Nigel {1} Belcher, Susan {1} Herschberger, Nicholas {1} Osborne, Erin {1} Barkan, Alice {1} Stern,
David {2} Brenchley, Jamie {2} Harris, Faith {2} Kikis, Elise {2} Hanson, Maureen {3} Peeters, Nemo {3}
{1} Institute of Molecular Biology, University of Oregon, Eugene {2} Boyce Thompson Institute, Ithaca, NY {3}
Cornell University, Ithaca, NY
The study of mutants deficient in chloroplast biogenesis can elucidate chloroplast gene regulation, protein
targeting, membrane assembly, and organellar cross talk, as well as aspects of prokaryotic biology that have been
conserved during chloroplast evolution. To this end, we have generated a large collection of Mutator-induced
maize mutants. The library currently consists of ~2100 mutants with visible chlorophyll deficiencies (from subtle
to severe), phenotypes known to result from mutations that disrupt the expression, targeting, assembly, stability,
or function of components of the photosynthetic apparatus. The collection has been made available to the
scientific community via a searchable database which describes the defects associated with each mutation at the
pigment, chloroplast protein, and chloroplast RNA levels http://chloroplast.uoregon.edu. The collection is also the
basis of a reverse-genetic service for genes potentially involved in chloroplast function; users can request we
screen the library for Mu insertions in their gene of interest. Multiple alleles are generally identified in successful
reverse genetic screens indicating that the collection is nearing saturation.

The major photosynthetic enzyme complexes in each mutant are cataloged via immunoblot, and chloroplast
mRNAs encoding missing proteins are subsequently analyzed on Northern blots. Protein data for 1100 mutants
and RNA data for ~500 mutants are currently available online, providing comprehensive phenotypic information
to help users select mutants of interest. Protein defects include; the specific loss of each single photosynthetic
enzyme complex, global protein losses, protein losses diagnostic for defects in the cpSec and DpH thylakoid
targeting pathways, and the loss of subsets of photosynthetic complexes in virtually every combination. This latter
class of mutants is essentially unexplored and may point to unanticipated circuitry coordinating the synthesis or
assembly of components of the photosynthetic apparatus. Of the 500 mutants thus far examined at the RNA level,
44 show defects in chloroplast RNAs metabolism. Many of the RNA phenotypes have not been described
previously, revealing the benefit of this large scale analysis of nuclear genes involved in chloroplast RNA
metabolism. Seed from mutants with phenotypes of interest can be requested for detailed study.

37
Physical localization of single copy sequences on 2-D maize pachytene chromosomes by fluorescence in situ
hybridization
Wang, CJ. Rachel (1,2) Harper, Lisa (2) Cande, W. Zacheus (2)
(1) Department of Botany, National Taiwan University, Taipei, Taiwan 106, Taiwan. (2) Department of
Molecular and Cell Biology, University of California at Berkeley, Berkeley, CA 94720, USA.
Fluorescence in situ hybridization (FISH) is a powerful technique for physical mapping. However, the sensitivity
of FISH in plant chromosomes has been low because of limited probe penetration and high levels of background
staining in the cytoplasm. Here we demonstrate that single copy sequences can be mapped on ethanol:acetic acid
fixed maize pachytene chromosome squashes by FISH using directly labeled fluorescent DNA probes. A 13.4 kb
genomic clone, pHF10 generously provided by Fu and Dooner, which contains single copy sequences from the
bronze (bz) region is localized on the short arm of chromosome 9, corresponding to its genetic locus. Another 8.5
kb genomic clone, pAMY2, which contains a beta-amylase gene shows 2 signals on chromosome 2 and 7,
respectively, whereas genetic mapping data suggests that there is only one gene each on chromosome 2 and 7.
This is the first report of physical mapping of single copy DNA sequences by FISH in plants using directly
labeled fluorescent probes. Our attempts to localize single copy sequences on 3-D maize pachytene chromosomes
fixed maize pachytene chromosome squashes by FISH using directly labeled fluorescent DNA probes. A 13.4 kb
genomic clone, pHF10 generously provided by Fu and Dooner, which contains single copy sequences from the
bronze (bz) region is localized on the short arm of chromosome 9, corresponding to its genetic locus. Another 8.5
kb genomic clone, pAMY2, which contains a beta-amylase gene shows 2 signals on chromosome 2 and 7,
respectively, whereas genetic mapping data suggests that there is only one gene each on chromosome 2 and 7.
This is the first report of physical mapping of single copy DNA sequences by FISH in plants using directly
labeled fluorescent probes. Our attempts to localize single copy sequences on 3-D maize pachytene chromosomes
fixed to preserve nuclear architecture have been less successful. In order to compare the morphology of 2-D
squashes of ethanol:acetic acid fixed chromosomes related to 3-D formaldehyde fixed chromosomes, we have
measured chromosome lengths, arm ratios, and placement of 5S rDNA on 2-D and 3-D chromosomes in two
different inbred lines using these different FISH techniques. The development of a technology for localizing
single copy sequences will be useful to the maize community for mapping sequences, ordering clones, and for
correlating the genetic and cytological maps.
38
Establishment of Robust Maize Transformation Systems for the Public Sector
Wang, Kan (1) Frame, Bronwyn (1) McMurray, Jennifer (1) Gelvin, Stanton (2) Crane, Yan (2) Davis,
Susan (2) Kaeppler, Heidi (3) McGill, Annie (3) Zeng, Peiyu (3) Thompson, Bill (4) Nguyen, Tuyen (4)
Allen, George (4) Calloway, Anton (4)
(1) Iowa State University (2) Purdue University (3) University of Wisconsin-Madison (4) North Carolina State
University
The long-term goal of this research is to establish a robust maize transformation system to enable the maize
research community for future functional genomic research as well as crop improvement. Our specific objectives
are: 1) Developing a routine Agrobacterium-mediated transformation system. We will conduct experiments to
systematically optimize transformation and regeneration parameters for Agrobacterium-mediated maize
transformation using non-super binary vector systems. 2) Enhancing transgene integration and expression. We
will transfer and express the Arabidopsis and maize rat (resistant to Agrobacterium transformation) genes/proteins
in maize to evaluate their effects in enhancing transgene integration. We will also evaluate the effect of tobacco
and maize MARs (Matrix Attachment Regions) on the efficiency of our transformation protocols and on the
frequency of somatic silencing events in maize. 3) Investigating germline transformation protocols. We will
investigate tissue culture-independent transformation protocols such as meristem transformation and female
gametophyte transformation. 4) Exploring inbred line transformation. We will conduct research to improve inbred
line transformation on B73, H99, Oh43 and W22. In addition, we will facilitate transfer of improved protocols to
the public sector by providing a more efficient transformation service at the Plant Transformation Facility at Iowa
State University and organizing transformation workshops during the course of this program. It is also our
intention that any vector systems, reliable transformation protocols and information generated from this research
program will be made available for the maize community upon request. Deliverables: 1) Both Biolistic-mediated
and Agrobacterium-mediated maize transformation services are available at the Plant Transformation Facility of
Iowa State University(http://www.agron.iastate.edu/ptf/Web/mainframe.htm) 2) A Maize Transformation
Workshop is scheduled at Madison, WI, on March 10-13, 2003

39
Gene Discovery and Mapping in Maize
Wen, Tsui-Jung{1} Guo, Ling{1} Nienaber, Amy{1} Borsuk, Lisa{1} Ashlock, Daniel{1} Schnable, Patrick
S.{1}
{1}Iowa State University
More than 32,000 ESTs have been submitted to Genbank. Most were generated by sequencing the 3í ends of
cDNAs derived from the inbred line B73. Hence, it is possible for these ESTs to be assembled into a set of unique
genes with a very high degree of confidence. 28,500 3í B73 ESTs from two normalized libraries defined 3,794
singletons and 3,679 contigs (7,473 unique genes), resulting in an overall rate of gene discovery of 26%. IDPs
(InDel Polymorphisms) are a class of PCR-based, genetic markers that detect the insertions and deletions (indels)
that exist among maize alleles. Because they can be detected using agarose gel electrophoresis, these markers are
suitable for use by almost all maize genetics laboratories. Over 1,000 IDP markers that distinguish B73 and Mo17
have been developed. On average any given IDP marker has an approximately 50% chance of being polymorphic
in any pair of 22 tested inbred lines. Hence, it is likely that approximately 500 of the IDP markers will prove
informative in any experiment that requires genetic markers. Over 1,000 ESTs have been genetically mapped
using the IBM mapping population and another ~700 have been mapped using other populations involving B73 or
Mo17. By substantially increasing the number of sequence-based connections between the maize genetic map and
the rice physical map, these data will enhance the value of the rice genome sequence to the maize genetics
community. Over half of the genetically mapped ESTs exhibit a high degree of sequence similarity (a stringent
>e-75) to the EST contigs used to design the overgo probes being used by the NSF Project led by Ed Coe to
physically map ESTs to BACs. Hence, in combination these experiments promise to yield hundreds of cross-
linkages between the genetic and physical maps of maize.
community. Over half of the genetically mapped ESTs exhibit a high degree of sequence similarity (a stringent
>e-75) to the EST contigs used to design the overgo probes being used by the NSF Project led by Ed Coe to
physically map ESTs to BACs. Hence, in combination these experiments promise to yield hundreds of cross-
linkages between the genetic and physical maps of maize.
40
CONSORTIUM FOR MAIZE GENOMICS - APPROACH EVALUATION FOR TARGETED
SEQUENCING OF MAIZE GENES
Whitelaw, Catherine {1} Quackenbush, John {1} Schubert, Karel {2} Beachy, Roger {2} Lakey, Nathan {3}
Bennetzen, Jeffrey {4}
{1} The Institute for Genomic Research {2} The Donald Danforth Plant Science Center {3} Orion Genomics {4}
Purdue University
Maize is both a classical genetic model for plant research and an economically important crop; however, the size
and complexity of the maize genome deem it recalcitrant to whole genome sequencing. Current estimates indicate
that genes constitute a mere 15-20% of the maize genome with the remainder consisting of highly repetitive DNA.
The initial objective of the Consortium for Maize Genomics is to evaluate two approaches to sequencing the
maize ëgenespaceí (methylation filtration and high Cot selection) in order to provide the most rapid and cost-
effective alternative to sequencing the whole genome. At TIGR, we are generating paired end sequence reads
from 250,000 methylation filtered clones and 250,000 high Cot clones. The sequences are clustered and
assembled, both independently and in combination, at quarterly intervals. The resulting maize genomic assemblies
and singletons are annotated based on homology searches, with subsequent development of improved methods for
annotation. The results of these analyses, a BLAST-searchable database and Maize Assembly Annotator are
presented in the TIGR Maize Database (http://www.tigr.org/tdb/tgi/maize). The results of the latest assemblies
will be presented and discussed.

41
Gene Enrichment Technologies for Selective Sequence Analysis of the Maize Genome
Yuan, Yinan {1} Emberton, John {1} SanMiguel, Phillip{2} Bennetzen, Jeffrey {1}
{1} Department of Biological Sciences, Purdue University, West Lafayette, INDIANA 47907 {2} Purdue
University Genomics Core, S004 WSLR, Purdue University, West Lafayette, INDIANA 47907
Like other grass genomes, the nuclear DNA of maize is mostly composed of highly repetitive sequences. Over
60% of the maize genome is composed of LTR retrotransposons that are often found as nested blocks of 10-100
kb intermixed with short genic regions. The large size and the similarity of these retrotransposons will make it
very difficult to assemble shotgun sequence data from the maize genome into long contiguous sequences
(contigs). In order to avoid the inefficiency of sequencing the unwanted repetitive DNA, we and others have
developed various gene enrichment technologies based on methylation pattern differences and copy number
differences between repetitive DNAs and genes. These techniques will allow a shotgun approach for relatively
efficient sequencing of gene-rich regions in the maize genome and other higher plant genomes. We will discuss
and compare these technologies, indicating their respective strengths and weaknesses, especially with regard to
the genic components that might be missed and their representation of the boundaries between repetitive DNAs
and genic regions. Several of these techniques appear to be highly complementary. If used in a combined
approach, these gene enrichment techniques should produce almost all genes in sequence contigs, and will also
locate these contigs on genetic and physical maps.

Biochemical Genetics

42
Correlation between SU1 isoamylase expression level and amylopectin structure
Bierwagen, Tracie {1} Mou, Beiquan {1} Myers, Alan {1} James, Martha {1}
{1} Iowa State University
Genetic analysis of maize sugary1 (su1) mutants indicates that isoamylase-type starch debranching enzymes
(DBE) are required for normal starch levels, although whether or not DBEs are directly involved in starch
biosynthesis is unknown. This study utilized a transgenic approach to test the hypothesis that DBEs provide a
molecular editing function that directly influences the structure of the amylopectin (Ap) component of starch.
Expression of Su1, which codes for one of the three isoamylase-type DBEs conserved in maize and other plants,
was modified in transgenic maize in an attempt to correlate increased or reduced expression of SU1 isoamyylase-
type DBE with alterations in the distribution of chain lengths and/or branch linkages in Ap. Thirteen transgenic
maize lines are now in at least their fifth generation and have been introgressed into the B73 inbred background.
Reduced expression of Su1 has been achieved to various degrees in antisense transformants and in two gene-
silenced lines. Chain length profiles of endosperm Ap from these lines by fluorophore assisted carbohydrate
electrophoresis (FACE) revealed that Ap molecules have more short chains relative to wild type, similar to su1-
mutants. Unlike the genetic mutants, however, most of the transgenic plants do not accumulate the water-soluble
type DBE with alterations in the distribution of chain lengths and/or branch linkages in Ap. Thirteen transgenic
maize lines are now in at least their fifth generation and have been introgressed into the B73 inbred background.
Reduced expression of Su1 has been achieved to various degrees in antisense transformants and in two gene-
silenced lines. Chain length profiles of endosperm Ap from these lines by fluorophore assisted carbohydrate
electrophoresis (FACE) revealed that Ap molecules have more short chains relative to wild type, similar to su1-
mutants. Unlike the genetic mutants, however, most of the transgenic plants do not accumulate the water-soluble
glucan phytoglycogen. Thus, phytoglycogen production is not a necessary condition for modification of Ap
structure as a result of SU1 isoamylase deficiency. Three lines have been identified that over-express Su1. FACE
analysis of endosperm Ap shows that increased Su1 expression results in fewer short chains and more
intermediate length and long chains, which is the opposite effect of the su1- mutations and transgenically-induced
deficiency in SU1 isoamylase-type DBE. The reciprocal effects of increased and decreased Su1 activity on Ap
structure suggest that the SU1 isoamylase-type DBE directly catalyzes hydrolysis of branch linkages in a pre-Ap
substrate.

43
Identification of a novel starch hydrolytic enzyme in maize
Colleoni, Christophe {1} Myers, Alan {2} James, Martha {3}
{1} Iowa state University
Starch is mobilized by a combination of phosphorolytic and hydrolytic enzyme activities. Despite the critical
functions of starch degradation in plant metabolism, the molecular details of the process remain to be described.
In plants, it is known that debranching enzymes, disproportionating enzymes, and starch phosphorylases are
potential degradative enzymes, as are the _-1,4 specific hydrolases of the _-amylase and _-amylase classes. In this
study a novel starch degradative enzyme activity was identified, beginning with the observation of a low mobility
band in native activity gels (zymograms) containing starch as the substrate. The novel enzyme (termed SHE for
starch hydrolyzing enzyme) was partially purified from developing maize endosperm by a combination of anion
exchange chromatography and gel permeation chromatography. Proteins from fractions in which SHE activity
was enriched were separated by native PAGE and transferred to a series of gels containing different glucan
polymers. Dye staining of each glucan substrate revealed that SHE hydrolyzes amylopectin, the _-limit dextrin of
amylopectin, amylose, and glycogen, but does not hydrolyze pullulan, a linear polymer of isomaltotriosyl units.
Separation of the hydrolysis products by fluorophore-assisted carbohydrate electrophoresis (FACE) indicated that
maltose was the sole hydrolysis product in all instances. Although _-amylases are known to release maltose from
the non-reducing ends of glucan polymers, acting in an exo-specific manner to cleave _-1,4 linkages, they have
limited activity toward amylose and are unable to further hydrolyze _-limit dextrin. Because SHE readily
hydrolyzes these substrates, we predict that this enzyme acts endo-specifically, cleaving internal _-1,4 linkages to
release maltose. Maltogenic amylases have been reported in bacteria, although these enzymes also have activity
toward pullulan. This is the first description of an endo-acting maltogenic glucan hydrolase in a plant. Further
experiments are underway to define SHE enzymatic activity and to discern its function in maize starch
metabolism.
44
Comparison of cell wall proteins in drought tolerant and susceptible maize lines
Dierking, Ryan {1} Soto, Jennifer {1} Guenther, Eric {1} Davis, Georgia {1}
{1} University of Missouri-Columbia
Drought is a major limitation to global crop production. Many factors affect drought response including root
architecture, root growth and dimension. Expansin expression increases in roots of water-stressed compared to
well-watered, indicating changes in wall structure or chemistry Wu et al. (1996, 2001). Our goal is to identify root
cell-wall specific proteins that play a role in drought tolerance. Our objectives are: 1) to evaluate the Bio-Rad
Sequential Extraction Kit II for its ability to separate cytosolic and cell wall proteins using 2-dimensional
isoelectric focusing gels; 2) to compare the cell wall proteins from root and shoot of seedlings at the 6-leaf stage,
and; 3) to evaluate genotypic differences in root cell wall proteins. For this experiment inbreds Mo17, B73, B97,
and Fr697 were used. Four replicates of five plants were planted in a completely random design in the
greenhouse. All plants were well watered and root and shoot tissue was harvested at the 6-leaf stage. Proteins
were extracted using a Sequential Extraction Kit from Bio-Rad. The proteins were separated on 11 cm pH 3-10
strips and 4-20% Tris HCl gels, and then stained with coomassie blue. Gel images were compared and analyzed
using PD Quest to identify differences in position and relative intensity of spots. Subsequent experiments will
compare cell wall proteins in water stressed vs. well watered roots to identify candidate cell wall proteins
associated with drought tolerance.
45
A Comprehensive Real-time PCR Expression Analysis of the Maize Starch Debranching Enzyme Gene
Family
Dinges, Jason R. {1} James, Martha G. {1} Myers, Alan M. {1}
{1} Iowa State University
This study describes the full complement in maize of starch debranching enzymes (_(1 _ 6) glucan hydrolase, or
DBE), and their expression pattern. Plants contain two conserved types of DBE, the isoamylase- and pullulanase-
type enzymes. Genetic analysis of sugary1 (su1) mutations in maize and other cereals indicates that at least one
isoamylase-type DBE is required for normal starch biosynthesis. Genetic analysis of maize zpu1, coding for the
pullulanase-type DBE, suggests functions both in starch degradation and starch biosynthesis, and that the
biosynthetic function overlaps partially with that of the isoamylase-type DBE coded for by Su1. Here we report
the cloning of two additional isoamylase-type DBE cDNAs, Iso2 and Iso3. Like su1 and zpu1, these sequences
have been conserved throughout the evolution of plants, i.e. Arabidopsis, potato, barley, wheat, sorghum, and rice
all possess three isoamylase-like sequences that have been independently conserved to a high degree. A
comprehensive analysis of the expression of all four DBE transcripts was conducted in metabolically active
tissues. Debranching enzymes are expressed ubiquitously throughout the plant, although to varying levels. Su1 is
the most highly expressed mRNA, and is present during times of net starch synthesis. The pattern of Su1
expression follows closely that of Iso2, consistent with the idea that these isoforms form a heteromultimer. The
expression of Iso3 is most abundant in germinating seed, consistent with a role for this isoform in starch
degradation. Thus, individual DBE isoforms appear to be dedicated to specific catabolic and/or anabolic
functions.
46
Functional Significance of the Evolutionary Steps that Shaped the R2R3 Myb Gene Family
Heine, George F. {1} Dias, Anusha P. {1} Grotewold, Erich {1}
{1} The Ohio State University
Plants express hundreds of R2R3 MYB transcription factors. The expansion of the R2R3 Myb gene family
occurred in plants 250-450 million years ago. R2R3 Myb transcription factors are derived from widely distributed
R1R2R3 Myb genes, represented in animals by the c-myb proto-oncogene and in the plants by the members of the
small pc-myb gene family. Discrete evolutionary steps involving the loss of the R1 MYB repeat, the insertion of a
Leu residue in the R2 MYB repeat and the replacement of the first Trp in R3 by a hydrophobic residue shaped the
R2R3 MYB family that dramatically expanded in plants. In addition, recent duplications of R2R3 Myb genes
have characteristic changes in the MYB domains. One example is provided by the P-to-A clade formed in maize
by at least 10 members and characterized by the change of a very conserved Pro to Ala in the hinge region that
joins R2 and R3. Here, we investigated how each of these changes influenced the activity of R2R3 Myb proteins,
starting from the P1 gene, a member of the P-to-A group of maize R2R3 Myb factors. Our results indicate that,
while DNA-binding has clearly been influenced by the changes that forged the R2R3 Myb gene family, some
changes must have other functions unrelated to DNA binding, and what those functions might be will be
discussed
47
Photosynthetic Mutant Library (PML): A Reverse Genetics Resource for Chloroplast Biogenesis Genes
Herschberger, Nicholas {1} Walker, Nigel {1} Belcher, Susan {1} Osborne, Erin {1} Barkan, Alice {1} Stern,
David {2} Brenchley, Jamie {2} Harris, Faith {2} Kikis, Elise {2} Hanson, Maureen {3} Peeters, Nemo {3}
{1} Institute of Molecular Biology, University of Oregon, Eugene {2} Boyce Thompson Institute, Ithaca, NY {3}
Cornell University, Ithaca, NY
The goal of this project is to develop, use, and disseminate a set of resources for the genetic and biochemical
dissection of chloroplast biogenesis. We have developed a core resource consisting of ~2100 transposon-induced
chloroplast-defective maize Mutants. Mutants were selected from Mutator (Mu) maize lines based upon their
chlorophyll-deficient leaves and/or increased level of chlorophyll fluorescence. Previous studies support the
notion that a disruption of many aspects of chloroplast biogenesis (e.g. import of proteins into the organelle, lipid,
pigment, and prosthetic group synthesis, chloroplast gene expression, intra-chloroplast protein sorting, assembly
of the photosynthetic apparatus) will result in one of these easily scored phenotypes. One application of the
Mutant collection is as a reverse genetics resource that can be used to determine the roles of genes of known
sequence with suspected roles in chloroplast biogenesis. Pooled Mutant DNA samples are screened by PCR for
Mu insertions in genes of interest. Mutants are identified and seed from the Mutant line is provided as a service.
The current collection of ~2100 Mutants appears to be nearing saturation because successful reverse genetic
screens generally yield multiple Mutant alleles. This resource has proven useful for testing the in vivo relevance of
yeast two-hybrid interactions, for identifying new Mutant alleles of previously cloned genes, for testing the
functions of proteins identified through biochemical study of chloroplast import, and for testing functions of plant
homologs of bacterial secretory proteins.
yeast two-hybrid interactions, for identifying new Mutant alleles of previously cloned genes, for testing the
functions of proteins identified through biochemical study of chloroplast import, and for testing functions of plant
homologs of bacterial secretory proteins.
48
The Protein Disulfide Isomerase Multigene Family in Maize
Houston, Norma {1,2} Fan, Chuanzhu {1} Xiang, Qiu-Yun (Jenny) {1} Jung, Rudolf {3} Boston, Rebecca
{1,2}
{1} Department of Botany, North Carolina State University Raleigh, NC {2} Functional Genomics Program,
North Carolina State University Raleigh, NC {3} Pioneer Hi-Bred International, Inc. Johnston, IA
In immature seed, storage protein synthesis and subsequent packaging into protein bodies is a major metabolic
activity. Perturbing protein folding can have dramatic phenotypic effects, such as the lower amounts of storage
proteins and abnormally shaped protein bodies of the maize endosperm mutants floury-2 (fl2), Defective
Endosperm B-30 (De-B30), and Mucronate (Mc). The biological response to accumulation of non-native protein
domains is the induction of a battery of molecular chaperones, including protein disulfide isomerase (PDI). PDI
provides a key role in aiding misfolded protein to achieve their native conformation by catalyzing disulfide bond
redox/isomerization. The enzyme is encoded by a multigene family within a subgroup of the thioredoxin
superfamily. PDIs are characterized by their domain structures that generally contain two active thioredoxin-like
domains minimally composed of a CXXC tetrapeptide and a distinct fold pattern. Although a number of PDIs
have been described, the evolution of plant PDIs is poorly understood. Using BLAST analysis, we performed
iterative searches with published plant sequences for PDIs in the Arabidopsis genome and subsequently removed
putative thioredoxins. We then searched maize and rice genomes to obtain a data matrix of thirty-one putative
plant PDIs for phylogenetic analysis. Both Bayesian and maximum likelihood phylogenetic analyses of PDIs
suggest that at least five duplication events occurred before the divergence of monocots and dicots. The
phylogeny also implies that there were frequent duplications within different lineages of flowering plants. We
have initiated gene and protein expression analyses of PDIs in maize. The major PDI, which has homologues in
most higher eukaryotes, has increased mRNA and protein expression levels in endosperm of the fl2, De-B30 and
Mc mutants compared to the normal inbred. A second putative PDI appears to be plant specific and has a single
thioredoxin-like domain. Immunoblot analysis shows induction of this unusual PDI in fl2 but not in De-B30 or
Mc.
49
Evidence for physical interactions among specific starch metabolizing enzymes
Kim, Seungtaek {1} Marsh, Rebekah {1} Bierwagen, Tracie {1} Mou, Beiquan {2} Myers, Alan {1} James,
Martha {1}
{1} Iowa State University, Ames, IA {2} USDA-ARS, Salinas, CA
Multiple isoforms of starch synthases (SS), starch branching enzymes (BE), and starch debranching enzymes
(DBE) are evolutionarily conserved, suggesting each has a distinct function in starch metabolism. Reduced
expression of a particular isoform can trigger pleiotropic effects on other enzymes in the pathway, indicating that
some enzyme activities are inter-dependent and coordinately regulated. A likely molecular explanation involves
protein-protein interactions between components of the starch biosynthetic system. This study investigated the
potential for protein complex formation among maize SS, BE, and DBE isoforms by three different approaches:
the yeast two-hybrid interaction system, affinity chromatography, and co-immunoprecipitation. Reciprocal pair-
wise tests for protein interactions in yeast indicate that physical associations occur between SU1 (the isoamylase-
type DBE that is the product of the sugary1 gene) and BEIIb (the product of the amylose extender gene), and
between BEIIa and DU1 (the SSIII isoform that is the product of the dull1 gene). Positive unilateral results were
obtained for a number of isoforms when SU1, BEIIa, and DU1 were expressed as GAL4 binding domain fusions,
including evidence for a BEIIa/SSI association. Affinity column chromatography provides further support for this
interaction. Immunoblot analysis revealed that BEIIa was tightly bound to a SSI affinity column, and the
reciprocal experiment showed that SSI was tightly bound to a BEIIa affinity column. These results are supported
by the co-immunoprecipitation of BEIIa and SSI from protein extracts of developing maize endosperm, using
antibodies to both proteins. Similar experiments with anti-DU1 antibodies reveal that both BEIIa and SSI
immunoprecipitate with DU1. The discoveries of these interactions provide strong support for the existence of
protein complexes comprised of specific SS, BE, and DBE isoforms. We hypothesize that such protein complex
formation serves to coordinate and regulate the activities of the enzymes involved, and provide for the
architectural specificity of glucosyl linkages inherent in starch granules.
50
Utility of 2 Dimensional Isoelectric Focusing in Identifying Biochemical Networks.
Kea, Molly {1} Soto, Jennifer {1} Guenther, Eric {1} Dierking, Ryan {1} Gerau, Michael {1} Musket,
Theresa {1} Davis, Georgia {1}
{1} University of Missouri-Columbia
There are two distinct phases in Zea mays L. leaf development, juvenile and adult phase. The first leaves are
juvenile and have a dull green appearance with thick bluish wax rich in fatty alcohols. Adult leaves have a
distinctive glossy appearance. Their epicuticular wax is high in long chain wax esters. The number of juvenile
leaves is correlated to insect damage from the fall armyworm, (Spodoptera frugiperda ( J.E. Smith), and the
southwestern corn borer, (Diatraea grandiosella Dyar). The larvae of both species cause significant whorl stage
damage. Glossy15 (Gl15) is an apetala-2 type transcription factor. Mutations at this locus cause an abbreviated
juvenile phase. This shortened juvenile stage plays a key role in insect resistance. Studies indicate that the
juvenile-to-adult phase change occurs earlier in resistant than in susceptible corn lines. Twenty-two glossy
mutants have been identified to date. Our objective is to use regulatory and structural genes from the epicuticular
wax biosynthesis (Glossy15) pathway to evaluate the potential for identifying biochemical networks using single
gene mutants and 2-dimensional isoelectric focusing. Leaves of seven families segregating for glossy genes in
different genetic backgrounds were collected by phenotype. Glossy phenotype and wild-type for each structural
gene were compared against Gl15 protein profiles to identify putative proteins involved in epicuticular wax
biosynthesis.
51
Characterization of a putative maize ERAD protein
Kirst, Mariana (1) Meyer, David (2) Jung, Rudolf (2) Boston, Rebecca (1)
(1) Department of Botany, North Carolina State University, Raleigh, NC (2) Pioneer Hi-Bred International, Inc.
Johnston, IA
Accumulation of mutant or foreign polypeptides in the endoplasmic reticulum (ER) can lead to channeling of the
polypeptides for degradation through the ER-associated protein degradation (ERAD) pathway. Several of the
polypeptides of this multi-subunit degradation machinery have been described in yeast but little is known about
the phenomenon in plants. We used the amino acid sequence of Der1p, a ~24 kDa component of the yeast ERAD
pathway in a homology search of maize ESTs. We identified two maize sequences that encode putative Der1
proteins and have initiated studies to determine whether they are involved in ERAD in maize. 'Digital northern'
analysis of EST representation among ~480,000 cDNA clones in the Pioneer database showed widespread
representation in different organs although the relative abundance within a given organ differed for the two
clones. Yeast Der1p is highly hydrophobic and is localized to the ER where it is predicted to span the membrane
four times. Both maize proteins show similar in silico transmembrane domains and localization. Subcellular
fractionation followed by immunoblotting showed Der1-1 in both protein bodies and ER. Further fractionation
showed tight association of Der1-1 with the ER membrane where it remained in the membrane fraction following
treatment with TritonX-100, tauricholate, and deoxycholate but was released into the supernatant in the presence
of digitonin. Prior to ERAD, proteins that fail to fold or assemble properly often induce an ER stress response
which is accompanied by increases in accumulation of both molecular chaperones and polypeptides of the ERAD
machinery. We assayed Der1 expression in maize endosperm mutants which we have previously shown to exhibit
an endosperm-specific unfolded protein response (UPR). RNA profiling and immunoblot analyses showed
overexpression of both Der1 genes in endosperm of the mutants relative to the normal inbred.
52
Function of the nitrilases ZmNIT1 and ZmNIT2 of Zea mays in auxin biosynthesis
Park, Woong June (1) Glawischnig, Erich (1) Kriechbaumer, Verena (1) Piotrowski, Markus (2) M¸ller,
Axel (2) Weiler, Elmar W (2) Meeley, Robert B (3) Gierl, Alfons (1)
(1) Lehrstuhl f¸r Genetik, Technische Universit‰t M¸nchen, Germany (2) Lehrstuhl f¸r Pflanzenphysiologie,
Ruhr-Universit‰t-Bochum, Germany (3) Pioneer Hi-Bred International, Johnston, USA
Nitrilases are suggested to convert indole-3-acetonitrile (IAN) to the auxin indole-3-acetic acid (IAA). Maize
seedlings grown in water containing IAN showed typical auxin-like effects, e.g. decreased root length, and
increased nitrilase expression. In maize kernel we detected highest nitrilase expression and activity in protein
extracts from aleurone/pericarp and embryo. Heterologeously expressed ZmNIT2 hydrolyses IAN to IAA 7 to
20fold more efficiently than Arabidopsis thaliana AtNIT1-3 and showed also high activity towards different
nitrile compounds like 3-phenylpropionitrile and 4-phenylbutyronitrile. In spite of a 69% identity in amino acid
sequence between ZmNIT2 and AtNIT4, ZmNIT2 did not accept the AtNIT4 substrate fl-cyanoalanine, which is
discussed as an intermediate in cyanide detoxification. The fact that ZmNIT2 is expressed in maize kernel, the
high turnover rate compared to Arabidopsis thaliana and the finding of the substrate IAN indicate an involvement
of nitrilases in auxin biosynthesis.
53
Color complementation in E. coli for the functional testing of a cDNA required for maize carotenoid
biosynthesis
Licciardello, Nicholas {1} Gallagher, Cynthia E. {1} Cervantes-Cervantes, Miguel {1} Wurtzel, Eleanore T.
{1,2}
{1} Lehman College, The City University of New York, Bronx, NY 10468 {2} Graduate Center of the City
University of New York, New York, NY 10016
Carotenoids represent a class of over 600 compounds with a 40-carbon backbone derived from 5 carbon
isoprenoid units; in plants, they are essential in photosynthesis and serve as precursors to abscisic acid. On the
basis of GenBank homology searching, we identified a cDNA predicted to encode betacarotene hydroxylase
(HYD), which catalyzes the addition of hydroxyl groups at position three on the cyclic rings of betacarotene to
form the xanthophyll zeaxanthin. This maize EST was sequenced and found to encode a protein with 56% identity
to a homolog from the dicot, Arabidopsis thaliana. The cDNA, cloned into EcoRI/XhoI sites of pET23c, encodes
a 359 amino acid polypeptide. Since sequence homology implies, but does not always indicate enzyme specificity
and activity, we used a heterologous bacterial system to demonstrate function of the putative maize HYD. The
cloned cDNA was tested by co-transformation with a second plasmid conferring accumulation, of the HYD
substrate, betacarotene, in Escherichia coli. Expression of a functional cDNA product altered colony pigmentation
of the betacarotene accumulating strain. HPLC analysis of the carotenoid products confirmed the accumulation of
zeaxanthin and the HYD intermediate, betacryptoxanthin. This research is funded in part by the National Institutes
of Health.
54
Enhancing Essential Amino Acids in Crop Plants
Back, Stephanie {1} Fagaly, Tanya {1} Gruys, Ken {1} Luethy, Michael {2} Manjunath, Siva {1}
Oulmassov, Tim {1} Ream, Joel {3} Reznick, Brad {1} Vaduva, Gabriella {1} Varagona, Rita {1} Voyle,
Dale {2} Wang, Joan {1} Weaver, Lisa {1}
{1} Quality Traits, Monsanto Company, St. Louis, MO 63167 {2} Quality Traits, Monsanto Company, Mystic,
CT 06355 {3} Trait Development, Monsanto Company, St. Louis, MO 63167
Several amino acids are essential to animals because they are required for growth and maintenance, but cannot be
synthesized de novo. Both cereals and oil seed crops make up an important part of livestock feed. However, both
types of seed proteins do not contain optimal amounts of several essential amino acids including shikimate
derived (Phenylalanine and tryptophan) and aspartate derived (lysine, threonine, methionine and isoleucine)
amino acids. One way to provide a more nutritionally balanced meal is to produce and accumulate these essential
amino acids in the seed. We are utilizing a transgenic approach to increase the accumulation of free amino acids
by changing the allosteric regulation of key enzymes in essential amino acid biosynthetic pathways. This
presentation focuses on anthranilate synthase for the production of tryptophan.
55
A maize ribosome-associated membrane protein is overexpressed in the floury-2 mutant
Sreenivasa Rao, TJV {1} Jung, Rudolf(2) Boston, Rebecca S (1)
{1} Department of Botany, North Carolina State University Raleigh, NC (2) Pioneer Hi-Bred International, Inc.
Johnston, IA
Synthesis of functional secretory proteins requires the coordinated interaction of several multi-subunit protein
complexes: ribosomes that associate with the ER, the translocation apparatus for transport of the nascent protein
into the ER, and molecular chaperones for protein folding and assembly. Synthesis of both molecular chaperones
(called the ER stress response) and ER associated degradation (ERAD) proteins is induced upon accumulation of
mutant or foreign proteins in the ER. Recently, ribosome-associated membrane protein 4 (RAMP4) was found to
be induced by environmental stress and pharmacological agents that cause an ER stress response in
rats(Yamaguchi et al.,1999). We used the rat RAMP4 amino acid sequence to screen maize ESTs and isolated two
putative homologs, RAMP4-1 and RAMP4-2. Both encode 69 amino acid polypeptides with a single predicted
membrane spanning domain. RNA profiling experiments showed expression of both genes in a variety of tissues.
RAMP4-1 was at least 10-fold more abundant in endosperm than in any other tissue while RAMP4-2 showed
little variation in expression from tissue to tissue. Polyclonal antibodies were raised against synthetic peptides for
both RAMP4 polypeptides. No cross-reacting material was detected in crude endosperm extracts of either normal
or fl2 maize. However, immunoblots of proteins from a ribosome fraction containing associated membrane
proteins showed weak cross-reactivity in normal endosperm and much stronger signal in the fl2 mutant. The
higher levels of RAMP4 in the fl2 mutant are suggestive of cross talk among molecular chaperones, ERAD
associated proteins and the translocation machinery. Yamaguchi, A., O. Hori, D.M. Stern, E. Hartmann, S.
Ogawa, M. Tohyama (1999) 147:1195-1204.
Ogawa, M. Tohyama (1999) 147:1195-1204.
56
From Proplastid to Chloroplast: Understanding Plastid Differentiation in Maize through Microarray and
Proteome Analysis (NSF Award - #0211935).
Van Wijk, Klaas {1} Stern, David {2} Clemente, Thomas (3) Brutnell, Thomas {2} Majeran, Wojciech {1}
Cai, Yang {1} Takacs, Elizabeth {2} Sawers, Ruairidh {2} Markelz, Nicole {2} Anufrikova, Katya {2} Sun,
Qi {4}
{1} Department of Plant Biology, Emerson Hall, Cornell University, Ithaca, NY14853. {2} Boyce Thompson
Institute, Tower Rd, Ithaca, NY1483. {3} Dept. of Agronomy & Horticulture, N308 Beadle Center, University of
Nebraska, Lincoln, NE 68588-0665. {4} Department of Theory & Simulation Sciences and Engineering, Cornell
University, Ithaca, NY14853.
Although best known for their role in photosynthesis, plastids perform many other important functions. These
include the synthesis of hormones, amino acids, fatty acids, lipids, vitamins, nucleotides and secondary
metabolites. Despite their essential role, little is known about the developmental signals governing plastid
differentiation. To elucidate the mechanisms of plastid biogenesis in maize, we will utilize genomic and
proteomic tools in conjunction with genetic analysis. Bioinfomatic approaches have enabled us to identify a set of
approximately 3500 maize EST sequences that are predicted to encode chloroplast targeted proteins. We intend to
use these sequences as the basis for building a DNA microarray. We are currently confirming the identities of
clones corresponding to approximately 1500 EST sequences selected for the Phase I array. This array will also
fully represent the maize chloroplast and mitochondrial genomes. In parallel, we have begun a broad survey of
proteins that accumulate in maize bundle sheath and mesophyll plastids using mass spectromerty. We have
separated and visualized proteins from purified plastids by high resolution 2-Dimensional Electrophoresis (2-DE).
These proteins will then be identified using both Matrix-Assisted-Laser-Desorption Ionization Time-of-Flight
Mass Spectrometry (MALDI-TOF MS) and on-line LC-Electrospray Ionisation Tandem Mass Spectrometry (LC-
ESI-MS/MS). Preliminary data will be presented. We will create reference protein profiles of plastid membrane
proteins using solvent-based fractionations, 1-dimensional electrophoresis (1-DE) gels and mass spectrometry. As
a further tool to aid in the characterization of photosynthetic development, transgenic plants carrying bundle
sheath and mesophyll cell-specific markers will be constructed. Work has begun on constructs to express GFP and
GUS in plastids, the cytosol or in mitochondria of both bundle sheath and mesophyll cells. The bioinformatic
integration of proteomic and expression profiling data will provide a detailed picture of the transcriptional and
translational control mechanisms utilized by maize plastids. The approaches described above will result in an
improved understanding of the biochemical functions of different plastid types and novel insights into the
differentiation process. Chloroplast function is directly related to agronomic performance, and therefore these
studies will greatly improve our understanding and ability to engineer important plant traits.
57
A Functional Genomics Program for the Illinois Long Term Protein Selection Strains
Schneerman, Martha C. {1} Uribelarrea, Martin {1} Seebauer, Juliann R. {1} Below, Fred E. {1} Moose,
Stephen P. {1}
{1} University of Illinois
The Illinois Protein Strains (Illinois High Protein and Illinois Low Protein) represent the known genetic extremes
for endosperm storage product accumulation and nitrogen metabolism in vegetative tissues. They are also the
product of the worldís longest-running breeding and selection experiment. The Illinois Protein Strains are thus an
excellent experimental system to investigate genome responses to selection and the genetic control of whole plant
nitrogen metabolism. We are employing a number of functional genomics strategies to address several important
questions about the Illinois Protein Strains. We are comparing genotypes/haplotypes for candidate genes and
molecular markers from generations 65 through 100 of the experiment to assess changes in allele frequencies
during the last 35 cycles of selection. mRNA expression profiling is being used to survey the global
transcriptional responses of the genome to 100 years of divergent selection for grain protein concentration, with
particular emphasis on genes associated with zein accumulation and nitrogen metabolism. Zein profiles will also
reveal the relative contribution of transcriptional and post-transcriptional regulation to the dramatic differences in
zein accumulation observed among the Illinois Protein Strains. To complement these profiling and gene discovery
efforts, mutations with known effects on grain composition are being introgressed into the Illinois Protein Strains
and EMS mutagenized populations have been generated for Illinois High Protein and Illinois Low Protein. These
populations are being screened for novel mutant phenotypes in these phenotypically extreme genetic backgrounds.
This poster will present our progress to date in each of the above research areas.
58
A genetic approach to dissecting the maize cell wall
Vermerris, Wilfred {1,2} Fontaine, Anne-Sophie {1,3} Bout, Siobhan {1} Boon, Jaap {4} Carpita, Nick {5}
Marita, Jane {6}
{1} Department of Agronomy, Purdue University, West Lafayette, IN 47907, USA {2} Department of
Agricultural & Biological Engineering, Purdue University, West Lafayette, IN 47907, USA {3} Unite de
Genetique et d'Amelioration des Plantes Fourrageres, INRA, Lusignan 86600, France {4} Institute for Atomic and
Molecular Physics, Amsterdam, The Netherlands {5} Department of Botany and Plant Pathology, Purdue
University, West Lafayette, IN 47907, USA {6} US Dairy Forage Research Center, University of Wisconsin,
Madison, WI 53706
The presence of lignin in the plant cell wall has evolved to serve multiple purposes, including water transport,
provision of rigidity, and defense against pests and pathogens. It is likely that the composition of the cell wall as
we know it, is a compromise that provides all functions in an adequate manner as long as the environmental
conditions stay within certain boundaries. With the current agricultural practice and the prospect of global
climatic change, however, a different cell wall composition may be optimal. In order to design the optimal cell
wall, a thorough understanding of the relationship between cell wall structure and function is required. We are
using genetics, in combination with detailed chemical, morphological and physiological analyses to address this
challenge. Three different strategies are employed: (a) we are developing near-isogenic lines in which different
mutations that affect the cell wall are combined, (b) we use forward and reverse genetics approaches to obtain
novel cell wall mutants (as part of an NSF Plant Genome project), and (c) we perform molecular and chemical
analyses on different lines that display significant variation in a trait of interest, such as forage quality. Recent
results include the finding that (1) there is an evolutionarily conserved interaction between cell wall composition
and flowering time, (2) the Brown midrib1 (Bm1) locus affects cell wall composition and plant development in a
dose-dependent manner, (3) the Bm2gene is important for the establishment of tissue-specific patterns of
lignification, (4) the cell wall of the bm3 mutant contains benzodioxanes, representing a novel structural feature of
lignin, (5) combining certain bm mutations has dramatic effects on plant growth and development, and (6) there is
a distinct but complex chemical basis underlying variation in forage quality of maize.
59
Inhibition of Aspergillus flavus growth by genistin and diosmin
Woodruff, Dana {1} Weston, Patrick {1} Musket, Theresa {1} Davis, Georgia {1}
{1} University of Missouri-Columbia
Aflatoxin is produced by Aspergillus flavus and A. paraciticus. Maize, figs, peanuts, cotton, and rice are all hosts
of A. flavus. Aflatoxin is the most potent natural carcinogen identified. The U.S. Food and Drug Administration
prohibits the sale of grain with aflatoxin levels exceeding 20 parts per billion (ppb). Once maize is found to be
contaminated, very few detoxification and utilization options are available. Some flavanoid compounds and
derivatives can alter the growth rate of fungi in vitro including A. flavus or the amount of toxin produced.
Flavanoids from soybeans such as genistein, genistin, daidzein, and daidzin have been shown to reduce plasma
cholesterol levels. Studies have also shown that a healthy diet that includes flavanoids reduces the risk of cancer.
In maize a defect in chalcone synthase (c2), a gene controlling the rate-limiting step in anthocyanin biosynthesis,
results in a 7-fold increase in toxin production. This indicates that flavanoids may also have an in vivo role in
reducing aflatoxin. A. flavus strain NRRL3357 was grown on Czepak's media plus 10g/L NaCl. The media was
supplemented with diosmin, daidzein, daidzin, genistein, and genistin at concentrations of 50, 100, 150, 200, and
250 mM. Fungal growth was measured every 2 days for 14 days after inoculation. Diosmin significantly reduced
growth at all time points compared to both the unsupplemented and DMSO supplemented media. Genistin at
concentrations > 150mM reduced growth from 4 to14 days after inoculation. This information will aid in the
identification of the enzymatic reactions that are critical in reducing aflatoxin production and in the identification
of naturally occurring maize alleles that can be used to produce low toxin accumulating maize lines.
60
Center for Eukaryotic Stuctural Genomics: Facility for Structure Determination of Proteins from
Arabidopsis thaliana and Other Model Eukaryotes.
Wrobel, Russell {1} Aceti, Dave {1} Amasino, Rick {1} Arangarasan, Raj {1} Bingman, Craig {1} Blommel,
Paul {1} Burch, Heather {1} Cao, John {1} Cornelescu, Claudia {1} Cornilescu, Gabriel {1} Doreleijers,
Jurgen {1} Dyer, Dave {1} Eghbalnia, Hamid {1} Fox, Brian {1} Fredrick, Ronnie {1} Hegeman, Adrian {1}
Hruby, Dave {1} Jeon, Won Bae {1} Johnson, Ken {1} Kimball, Todd {1} Kjer, Kelly {1} Lee, Peter {1} Ji,
Jing {1} Livny, Miron {1} Markley, John {1} Narayama, Ramya {1} Newman, Craig {1} Phillips, George
{1} Ramirez, Bryan {1} Rayment, Ivan {1} Runnels, Michael {1} Seder, Kory {1} Smith, David {1} Song,
Jikui {1} Sreenath, Hassan {1} Sussman, Mike {1} Thao, Sandy {1} Ulrich, Eldon {1} Vinaraov, Dmitriy {1}
Wesensberg, Gary {1} Westler, Milo {1} Zhang, Jianhua {1} Zhao, Qin {1} Zolnai, Zsolt {1} Volkman,
Brian {2} Peterson, Francis {2} Lytle, Betsy {2} Dunker, Keith {3} Oldfield, Chris {3} Linial, Michal {4}
Endo, Yaeta {5} Sawasaki, Tatsuya {5} Kainosho, Masatsune {6} Salemme, Ray {7} Carver, Ted {7}
{1} University of Wisconsin, Madison {2} Medical College of Wisconsin, Milwaukee {3} Molecular Kinetics,
Pullman, Washington {4} Hebrew University, Jerusalem, Israel {5} Ehime University, Matsuyama, Japan {6}
Tokyo Metropolitan University, Tokyo, Japan {7} 3D Pharmaceuticals, Exton, PA, USA.
The Center for Eukaryotic Structural Genomics (CESG) was founded as a collaborative effort to develop critical
technologies for determining three-dimensional structures of proteins rapidly and economically. CESG's initial
focus is on the genome of the model plant Arabidopsis thaliana. CESG has developed its own laboratory
information management system ('Sesame') designed to track, evaluate, and eventually direct all steps in the
process leading from gene to published structure. CESG's software periodically analyzes and ranks the entire
Arabidopsis proteome, consisting of ~29,000 open reading frames, to determine the targets to be carried forward
through a defined strategy leading to structure determination by either X-ray crystallography or NMR
spectroscopy. Priority is given to targets likely to open up important regions of conformational space or to
elucidate novel fold-function relationships. CESG also considers proteins of structural interest suggested by the
plant science community. Progress at CESG is tracked at the Structural Genomics web site at the PDB.
Technology and products developed by CESG, including clones, expression vectors, and excess protein, are being
made available to the scientific community. Initially, CESG has concentrated on bottlenecks associated with
cloning eukaryotic targets and the large-scale production of soluble protein. Gene chips produced by maskless
array DNA synthesis are being used to determine the presence of targets in cDNA pools generated by RT-PCR of
RNA isolated from an Arabidopsis callus cell line. CESG's standard protocol utilizes Invitrogens's 'Gateway'
plasmid construction system. Preliminary results on the steps leading to structure determinations will be
presented. Also presented will be specific steps that members of the maize community can use to tap into and help
fill CESG's pipeline.

CESG (http://uwstructuralgenomics.org/) is supported as a Pilot Project in the NIH Protein Structure Initiative
under Grant P50 GM64598 from the National Institute of General Medical Sciences.
61
The maize gene sugary2 codes for starch synthase IIa
Zhang, Xiaoli {1} Colleoni, Christophe {1} Ratushna, Vlada {1} James, Martha {1} Myers, Alan {1}
{1} Iowa State University
The sugary2 (su2) gene of maize is of industrial interest owing to phenotypic effects of mutations on starch
qualities, including reduced gelatinization temperature. The product of Su2 was not known previously, although
the gene that codes for starch synthase IIa (SSIIa) is tightly linked to the su2 locus. Here we used transposon
tagging in a forward genetic strategy to demonstrate that su2 does in fact code for SSIIa. Starting with the
published cDNA sequence for SSIIa, the sequence of the complete genomic locus that codes for this enzyme was
obtained from the wild type W64A background. The gene contains 9 introns in a transcription unit that codes for
an mRNA of approximately 2.2 Kb. The translational start site is located in exon 2 and is preceded by a large 5'
UTR. Two independent mutations, su2-2279 and su2-5178, were isolated from populations with an active Mutator
transposon system. Each mutation was introgressed into W64A by at least four backcrosses. Both mutations failed
to complement the reference mutation su2-Ref. The SSIIa coding region was examined by PCR amplification for
the presence of transposon insertions in homozygous su2-2279 or su2-5178 plants. Mutant su2-2279 was found to
contain a Mu1 element in the third exon of the SSIIa gene, and mutant su2-5178 contained a small
retrotransposon-like insertion in exon 10. Because two independent su2- mutations are coincident with physical
alterations of the gene that codes for SSIIa, we conclude that su2 and the SSIIa locus are the same genetic
element. Further support for this conclusion comes from the fact that su2-2279 endosperm lacks both SSIIa
protein and enzymatic activity, as shown in native activity gels and immunoblot analyses that definitively
identified SSIIa among all other starch synthase enzymes of maize. Chain length distributions in the amylopectin
component of endosperm starch from five different su2- mutant lines also are reported. We conclude that the
phenotypic effects of su2- mutations on maize starch result from a deficiency in SSIIa activity.

Cytogenetics

62
Single-locus cytogenetic mapping in maize with marker-selected sorghum BACs as FISH probes on
pachytene spreads from maize-chromosome-addition lines of oat.
Koumbaris, George {1} Bass, Hank W. {1}
{1} Florida State University, Tallahassee, FL USA, 32306
We have achieved a major technological advance towards the development of a cytogenetic map for maize (Zea
mays). The use of a maize chromosome-addition line of oat, OMAd9.2 (Kynast et al., 2001; Plant Physiol.
125:1216), provides a distinct advantage for obtaining clear images of optically-isolated chromosomes using a
chromosome spread and painting technique. We designed a direct-labeled oligonucleotide FISH probe, MCCY
(Maize CentC labeled with Cy5), to detect the centromere. The MCCY probe reproducibly hybridizes to a small
discrete region in the centromere of maize-9 ( Koumbaris and Bass, 2002; Maize Newsletter, vol.76). Using 3D
deconvolution microscopy and computer-assisted image analysis, we verified that the maize-9 arm ratio
(Long/Short; 1.6 to 2.1 from prior studies) is preserved in the oat genome background (1.7 in our system). Next
125:1216), provides a distinct advantage for obtaining clear images of optically-isolated chromosomes using a
chromosome spread and painting technique. We designed a direct-labeled oligonucleotide FISH probe, MCCY
(Maize CentC labeled with Cy5), to detect the centromere. The MCCY probe reproducibly hybridizes to a small
discrete region in the centromere of maize-9 ( Koumbaris and Bass, 2002; Maize Newsletter, vol.76). Using 3D
deconvolution microscopy and computer-assisted image analysis, we verified that the maize-9 arm ratio
(Long/Short; 1.6 to 2.1 from prior studies) is preserved in the oat genome background (1.7 in our system). Next
we have determined the cytogenetic map position for three test case core bin marker loci; CBM91 (Core Bin
Marker 9.01, umc109), CBM92 (9.02, bz1), and CBM98 (9.08, csu54b). Marker-selected sorghum BAC clones
were identified by library hybridization screening, purified, and labeled for FISH. We detected FISH signals at
9S.79, 9S.65, and 9L.95+.99 for sCBM91, sCBM92, and sCBM98 (two spots). We also mapped a sorghum GISH
locus (SorGS1) to ~9L.975. The CBM positions are consistent with their distribution on the linkage map and with
previous single-copy FISH-mapped loci on maize-9 from Sadder and Weber (Theor. Appl. Genet., 2002). This
work represents a new approach for efficient FISH-mapping of maize genetic markers. Given the relative ease
with which the first three loci were mapped using standard commercial FISH kits, this method could be extended
to develop a comprehensive cytogenetic map for maize. Extension of this approach could contribute to resource
development for chromosome painting, genome assembly efforts, comparative cytogenetics in grasses, and
placement of translocations and centromeres on the new IBM linkage maps.
63
absence of the first division1 (afd1) is a maize rec8-cohesin
Golubovskaya, Inna (1) Braun, David (2) Pawlowski, Wojtek (1) Cande, Zac (1)
(1) University of California Berkeley (2) Pennsylvania State University
Prominent features of the afd1 mutant phenotype is the bypass of the early prophase I stages, leptotene, zygotene,
and pachytene; and equational instead of reductional segregation of sister chromatids during the first meiotic
division due to sister chromatid cohesion (SSC) failure. Recently we have shown that afd1 mutants fail to
properly reorganize meiotic chromatin during prophase I. After interphase, which is not distinguishable from
normal interphase, afd1 prophase chromosomes fail to form a telomere bouquet. Telomeres do not attach to the
nuclear envelope during prophase I and remain scattered throughout the nucleus. afd1 prophase chromosomes do
not look like normal meiotic chromosomes; the two chromatids of each chromosome can be recognized during the
prophase I stage and even 80 telomeres (one for each chromatid on both chromosome ends) can be counted.
Neither homologous pairing nor synapsis occurs in afd1 meiocytes as monitored by following the behavior of
whole chromosomes or specific (5S rDNA) loci. The synaptonemal complex never forms in afd1 mutants. We
made substantial progress in understanding the nature of the afd1 mutation during the past year, owing to finding
a novel afd1 allele tagged with a Mu1 transposon. Finding this new allele was a result of a collaborative effort to
fine map afd1 and a leaf mutant tiedyed1 (tdy1) using SSR markers and classical genetics. The two genes are very
closely linked on the long arm of chromosome 6. The new allele, afd1 : Mu1, came from a directed tagging
experiment of the tdy1 gene. One tdy1-Reference/tdy : Mu1 mutant found by screening 30,000 plants showed a
unique pattern of inheritance. In a F2 progeny from a cross of the mutant plant to B73, 3/4 plants were fertile and
had normal leaves and 1/4 plants were both tiedyed and sterile. All tiedyed, sterile plants (19 examined) showed a
typical afd1 pattern of chromosome behavior during meiosis. Molecular analyses indicated that a Mu1 insertion
caused a small deletion encompassing both afd1 and tdy1. Cloning the deleted DNA region and sequence analysis
indicated that afd1 is likely a homolog of the yeast meiosis specific cohesin rec8. rec8 mutants in yeast are
defective in SSC and meiotic chromosome morphology and this phenotype is compatible with the mutant
phenotype of afd1.
64
Genetic and Cytogenetic Analysis of Abnormal Chromosome 10 Indicates that at Least Four Loci are
Required for Meiotic Drive
Hiatt, Evelyn N. {1} Dawe, R. Kelly {2}
{1}Kentucky Wesleyan College {2}University of Georgia
In the presence of a rare version of chromosome 10, Abnormal chromosome 10 (Ab10), specific chromosome
regions are preferentially segregated to the megaspore. Neocentromeric activity of knobs (heterochromatic regions
containing specific tandem repeats) is required for meiotic drive. Two regions of Ab10 are indicated in the trans-
activation of neocentromeres. Deletion analysis presented here indicates that only one of the neocentromere-
activating regions is necessary for the activation of many knobs. Recombination in structural heterozygotes is
increased when Ab10 is present in the genome. Deletion analysis has localized the region of Ab10 responsible for
this function. A third function identified by deletion analysis is localized to the tip of the Ab10 chromosome and
results in the complete loss of meiotic drive while maintaining typical Ab10 levels of neocentromeric activity and
recombination. A meiotic drive mutation, suppressor of meiotic drive 3 (smd3) was identified that also abolishes
meiotic drive while maintaining neocentromeric activity and recombination. The mutation, smd3, does not map to
the tip of Ab10, indicating the presence of a fourth function required for meiotic drive. All four functions have
been localized to specific regions of the Ab10 chromosome.
the tip of Ab10, indicating the presence of a fourth function required for meiotic drive. All four functions have
been localized to specific regions of the Ab10 chromosome.
65
Chromosome doubling in maize
Kato, Akio {1} Birchler,James A.{1}
{1}Division of Biological Sciences,University of Missouri - Columbia
Chromosome doubling methods are important for obtaining homozygous lines in a short period or the
establishment of polyploid inbred series to investigate gene dosage effects. During the last several years the
authors have developed three new chromosome doubling techniques in maize. 1) Doubled haploid induction from
parthenogenetically induced haploids. Kernels with a haploid embryo were induced by the stock 6 method (Coe
1963) with the combination of the R1-scm2 embryo color marker. The haploid kernels were germinated in small
pots and treated with nitrous oxide gas (600kPa) at the 6 leaf stage for two days in an air tight iron chamber (Kato
2002). The treated plants were grown in large pots and pollinated in a greenhouse. After self pollination, 44% of
the haploid plants produced doubled haploid kernels. 2) Tetraploid induction from diploid inbred lines. Maize
inbred lines just after pollination (30-36 hrs) were treated with nitrous oxide gas (800-900kPa) for 20 hrs. The ears
were harvested after maturity and chromosome numbers were determined on the germinated seedlings by the
method described by Kato (1999). From the treated ears, kernels with tetraploid embryos were obtained at 2-16%
level. Those tetraploid inbred lines were highly sterile in field conditions but moderately fertile in the winter
greenhouse. Six tetraploid inbred derivatives were established (A188, Oh43, W22, H99, stock 6, and B73). 3)
Triploid induction from diploid inbred lines. Tassels 7-9 days before flowering were treated with trifluralin
solution. By the dual pollination method (Kato 1997) kernels with a triploid embryo were detected as kernels with
white embryo and purple aleurone. Five - ten triploid kernels per ear can be obtained by this method.
66
The maize B chromosome contains multiple centromeric elements located away from the functional
centromere
Lamb, Jonathan C {1} Birchler, James A {1}
{1} Division of Biological Sciences, University of Missouri - Columbia
Various DNA elements have been identified that are located preferentially or specifically at the centromeres of
maize chromosomes. This fact has suggested that these elements play a role in centromere function. FISH analysis
of the maize B chromosome using these various centromeric elements shows that they are distributed outside the
functional centromere. A B-specific repeat shown to reside at the functional B centromere is also located near the
tip of the long arm. The centromere specific Ty3/gypsy-like retroelement CRM hybridizes at the B centromere as
well as near the B-specific repeat on the long arm. The Cent-C repeat, tandemly arrayed only at the centromeres
of the A chromosomes, is distributed in the heterochromatic regions of the long arm of the B-chromosomes in
several clusters. Probing with the Prem1 and Huck retroelements, which are enriched near the centromeres of A
chromosomes, shows that these sequences are enriched in the heterochromatin of the long arm of the B
chromosome. Co-localization of the CENH3 centromeric histone and these elements is being examined. Our
results to date indicate that CENH3 is located only at the functional B centromere. These data contradict a simple
model where centromeric DNA elements alone are sufficient to recruit centromeric proteins and form a functional
centromere. Instead, particular combinations of elements, greater quantities of these elements or an irrelevance of
underlying DNA sequence is involved with centromere formation.
67
Increased Meiotic Recombination in Maize Genotypes After Chronic Water-Deficit Stress
Verde, Luis {1} Woodman-Clikeman, Wendy {1} Long, Mary Jane {1} Nettleton, Daniel {1} Lee, Michael
{1}
{1} Iowa State University
The effect of water stress on meiotic recombination, first reported by our group in 2001, has been investigated in
greater detail and with larger sample sizes for genotypes B73/Mo17 and Mo17/H99. F1 plants were exposed to
either stress or non-stress treatments. For each genotype, 6 populations were created by backcrossing 6 F1 plants
(3 F1 plants from each treatment) as males to B73 or H99, respectively. 93 seedlings of each population were
genotyped at microsatellite loci to create genetic maps for chromosomes 1 and 10. For B73/Mo17, the maps of
chromosomes 1 and 10 were larger for stressed plants (172 vs 149 cM for ch. 1; 83 vs 72 cM for ch. 10). In the 6
B73/Mo17 populations, crossovers were not detected in 19 and 44% of the gametes for chromosomes 1 and 10,
respectively. For Mo17/H99, the maps of chromosomes 1 and 10 were also larger for stressed plants (177 vs 161
cM for ch. 1; 84 vs 65 cM for ch. 10). In the 6 Mo17/H99 populations, crossovers were not detected in 17 and
48% of the gametes for chromosomes 1 and 10, respectively. In both genotypes, 1 interval of chromosome 10
exhibited significantly more recombination.
exhibited significantly more recombination.
68
'Master switches' of meiosis
Pawlowski, Wojtek (1) Golubovskaya, Inna (1) Timofejeva, Ljuda (1) Shi, Liang (2) Li, Jingqiu (2) Kwan,
Waiking (2) Wang, Xun (2) Sheridan, William (3) Cande, W. Zacheus (1)
(1) Department of Molecular and Cell Biology, University of California at Berkeley (2) Torrey Mesa Research
Institute (3) Department of Biology, University of North Dakota
The switch from the somatic to the meiotic cell cycle in maize is regulated by ameiotic1. The am1 gene acts
downstream from genes determining meiocyte identity and upstream from all other meiotic genes. In most mutant
alleles of am1, meiocytes undergo mitosis instead of meiosis or arrest in pre-division interphase. In contrast, in the
am1-praI allele, the meiocytes enter meiosis and arrest in zygotene. It has been previously shown that the am1
gene regulates establishment of a meiosis-specific cytoskeleton. We now demonstrate that am1 also regulates
installation of the meiotic recombination machinery, as meiocytes in all am1 mutants lack chromosomal foci of
the recombination protein RAD51 involved in repairing meiotic double-strand breaks. We cloned the am1 gene
using a new Mutator-tagged allele. It encodes a novel protein with two conserved domains of unknown function
and a coiled-coil domain, frequently associated with protein-protein interactions. Interestingly, the am1 transcript
is constitutively expressed in all plant parts and not limited to meiocytes. These data suggest that am1 acts as a
switch activated by cell cycle-related proteins and triggers a number of downstream events during the meiotic
division. Another key meiotic protein, encoded by the poor homologous synapsis1 gene, is required for the
homology search, which ensures that homologous chromosomes find their proper partners during the meiotic
prophase. phs1 mutants show almost complete chromosome pairing and subsequent synapsis but homologous
pairing is replaced by non-homologous pairing. The mutants are severely defective in meiotic recombination: they
do not install RAD51 foci on chromosomes and show unrepaired double-strand breaks. The phs1 gene encodes a
novel meiosis-specific protein. We hypothesis that the PHS1 protein may be involved in loading recombination
complexes, containing RAD51, onto chromosomes. Unlike most recombination-related proteins, PHS1 shows a
low level of evolutionary conservation but nevertheless has two domains that are conserved among monocots and
dicots.
69
Cytological crossover maps for all maize bivalents using recombination nodules
Anderson, Lorinda {1} Brigham, Brian {1} Carter, Jenna {1} Hooker, Kristina {1} Lai, Ann {1} Rice,
Mindy {1} Stack, Stephen {1}
{1} Colorado State University
Cytological markers such as chiasmata, MLH1 foci, and recombination nodules (RNs) are useful for defining the
frequency and distribution of crossovers along the length of chromosomes. Of these, RNs provide the highest
resolution currently available because they are observed by electron microscopy of synaptonemal complexes
(SCs) in extended pachytene chromosomes. The most useful cytological crossover maps are those in which each
bivalent can be unequivocally identified and related to a specific linkage group. To date, such maps have been
generated only for tomato (using RNs) and mouse (using MLH1 foci). To achieve this goal for maize, we have
prepared an SC karyotype for the maize inbred line KYS, in which each SC can be identified based on its relative
length and arm ratio. Each SC was related to the proper chromosome and linkage group using inversion
heterozygotes. Using this karyotype, we mapped RNs on more than 2000 SCs to produce high resolution maps of
RN frequency and distribution on each bivalent. The average RN frequency per bivalent is closely correlated with
SC length. The crossover frequency using RNs is about 10% higher than estimated using chiasmata. While the
total length of the RN map is about two-fold shorter than current linkage maps, the correlations between the RN
and linkage maps are good (r2 > 0.63) when the cM lengths of the bivalents are compared. Each bivalent has a
unique distribution of crossing over, but all bivalents share certain general characteristics such as a high frequency
of distal RNs and a severe reduction of RNs at and immediately adjacent to kinetochores. The frequency of RNs
at knobs is either similar to or higher than the average frequency of RNs along the SCs. These RN maps represent
an independent measure of crossing over along maize bivalents and provide a means to integrate genetic linkage
maps with chromosome structure.

Developmental Genetics

70
A screen for non-reduction mutants in maize
Barrell, Philippa {1} Grossniklaus, Ueli {1}
{1} University of Zurich
Development of normal kernels in maize strictly requires a 2m:1p ratio of maternal to paternal genomes in the
endosperm. If an unreduced central cell is fertilised by a normal sperm, the kernel aborts due to an imbalance of
this genome dosage ratio in the endosperm (4m:1p). Seed abortion is also observed if a diploid female is crossed
with a tetraploid male resulting in a 2m:2p ratio, and any other deviation from the 2m:1p ratio. This ploidy barrier
can be utilised to screen for mutants that produce functional, non-reduced gametes. Fertilisation of an unreduced
embryo sac with pollen from a tetraploid results in a 4m:2p ratio ensuring normal endosperm development. As
anthocyanin pigmentation can easily be scored visually, a tetraploid stock tetraplex for the R-Navajo (R-nj/R-
nj/R-nj/R-nj) marker that pigments both the crown of the aleurone and the embryo, was developed. The R-nj
tetraploid stock was used as the pollen parent in order to distinguish progeny derived from contaminating pollen.
Families lacking anthocyanin pigmentation (r-g/r-g) derived from individuals with high Mu activity were used as
female parents and scored for segregating plants that produce plump kernels when pollinated with a 4n pollen-
donor. Three putative non-reduction mutants were identified. The mutants were out-crossed to inbred lines. Their
phenotypes were confirmed in reciprocal test-crosses to diploid and tetraploid plants. Flow cytometry revealed
that two of the mutants produced unreduced embryo sacs. In order to analyse the altered female gametophyte
development in the mutants, a protocol for confocal laser scanning microscopy was developed. To determine the
viability of the protocol, embryo sac formation of the maize mutant elongate (which produces both reduced and
unreduced embryo sacs) was analysed. In addition, transposon display will be used to identify the genes disrupted
by Mu.
71
Differential expression of a new HSP70-like gene and of the transposable elements Pong and PREM during
in vitro androgenesis in maize NILs underlines new aspects of induction mechanisms
Barret, Pierre {1} Brinkman, Michele {1} Dufour, Philippe {2} Murigneux, Alain {2} Beckert, Michel {1}
{1} INRA-UBP, UMR AmÈlioration et SantÈ des Plantes, Domaine de Crouel, 234 avenue du BrÈzet, 63 000
Clermont-Ferrand, France {2} Limagrain Genetics, 24 avenue des landais, 63170 AubiËre, France
Extensive studies have been conducted to understand the genetic control of in vitro androgenesis, but little is
know about the genes and the mechanisms involved in the switch that allows an immature pollen grain to start
embryo development. We have developed two maize isogenic lines, named AH5-44 and AH5-49, with high
androgenetic aptitude through backcross and selection from a high responsive DH229 line on the non-responding
A188 line genetic background. The genomic structure of these two lines was precisely described with
microsatellites. Five regions issued from the parent DH229 highly responsive to androgenesis were localised in
both AH5-44 and AH5-49. Sequences expressed on microspores extracted from the four lines were amplified
using a cDNA AFLP protocol. In addition, for each line, eight culture conditions were compared: microspores
extracted after tassel recovery, after 7 or 14 days in cold room, and after 1, 2, 3 and 4 days of in vitro culture. This
genetic and developmental screening allowed us to identify a new HSP 70 like candidate gene, a retrotransposon
PREM-1 analogue strongly expressed after microspore culture only in responsive lines, and a Pong-like
transposon recently described in rice, with expression restricted to the non responsive A188 line. Possible
implication of the HSP70-like sequence and of transposable elements in androgenesis response is discussed.
72
Quantitative analyses and transgenic overexpression support the maize FLORICAULA/LEAFY homologs
as QTL candidates.
Bomblies, Kirsten {1} Helentjaris, Tim {2} Doebley, John {1}
{1} Department of Genetics, University of Wisconsin, Madison {2} Crop Genetics Research, Pioneer- A DuPont
Company, 7300 NW 62nd Ave., Johnston, IA 50131 USA
One of the many morphological differences between maize and its wild ancestor teosinte, is a drastic increase in
kernel row number and length of the ear of maize. QTL for this morphological change have been mapped, and a
large effect QTL maps to the same chromosomal location as one of two maize FLORICAULA/LEAFY homologs,
zfl2. The orthologous genes are known in dicots to control inflorescence architecture by affecting the identity of
lateral structures produced by inflorescence meristems. We have previously supported the candidacy of zfl2 for
this QTL by analyzing its role in development through mutant analysis. We showed that plants carrying mutations
in the two maize FLORICAULA/LEAFY homologs, are indeed associated with lower kernel row number and
shorter ears than wild type siblings. To further test the candidacy of these genes, we analyzed quantitative effects
of mutations in the zfl genes on inflorescence architecture in more detail, and used transgenic plants ectopically
expressing zfl1 to ask whether increasing zfl activity in the inflorescence apex can specifically lead to higher
kernel row number and ear length (more cupules per row) in maize ears. The transgenic method allows us to test
that associated quantitative trends observed are specific to zfl activity, rather than the effects of linked genes.
Preliminary results suggest that transgenic plants ectopically expressing the zfl1 cDNA show an increase in kernel
row number compared to non-transgenic siblings. This suggests that increasing zfl activity in the inflorescence
promotes higher row number, and supports our hypothesis that increased zfl activity in the ear was selected during
the domestication of maize to increase ear size.
row number compared to non-transgenic siblings. This suggests that increasing zfl activity in the inflorescence
promotes higher row number, and supports our hypothesis that increased zfl activity in the ear was selected during
the domestication of maize to increase ear size.
73
tie-dyed1 promotes the sink/source transition in developing leaves
Braun, David {1} Inada, Noriko {2} Freeling, Michael {2}
{1} 208 Mueller Lab, Pennsylvania State University, University Park, PA 16802 {2} 351 Koshland Hall, UC-
Berkeley, Berkeley, CA 94720
We are interested in the signals and genetic programs coordinating leaf cell differentiation and development. To
identify genes controlling these processes we are characterizing mutations with altered pattern formation during
the final stages of leaf development. We focused initially on studying the tie-dyed1 (tdy1) mutant due to its
nonclonal variegation. tdy1 is a recessive mutation that develops yellow and green sectors in leaves. Sectoring
requires high light, is restricted to a narrow developmental time and results in the yellow tissue
hyperaccumulating starch. We hypothesize that tdy1 integrates developmental, environmental and metabolic
signals to promote the switch from a young, carbon importing tissue (sink) to a mature, carbon exporting tissue
(source).
74
An analysis of CNS-lg3-i2 in non-ligule forming plants and plants with altered ligules
Brooks, Lee {1} Lubkowitz, Mark {1}
{1} Saint Michael's College Colchester, VT 05439
Development occurs through the coordinated temporal and spatial regulation of genetic networks. Genetic
networks are integrated through many mechanisms including binding sites for regulatory proteins. Cross species
comparisons have proven a useful tool for identifying conserved elements in orthologous genes. These sequences
have been termed Conserved Non-Coding Sequences (CNSs). Through maize-rice comparisons the Freeling Lab
has identified a Conserved Non-Coding sequence in the second intron of liguleless3 and liguleless4 knotted1-like
homeobox (KNOX) genes. This family of genes is thought to encode DNA binding proteins that are involved in
lateral organ development. The CNS identified in liguleless3 and liguleless4 has been named CNSlg3-i2. An
analysis of lg3 orthologs from grasses representing each subfamily demonstrated that CNSlg3-i2 is most likely
conserved in position, sequence, and polarity in all grasses (Lubkowitz and Freeling, unpublished). This high
degree of conservation in a non-coding sequence suggests functionality. We are attempting to determine if the
absence or alteration of CNSlg3-2 will be observed in plants that fail to make a ligule. To this end, we have
cloned and sequenced the second intron of Lg3-O mutants and are attempting to sequence the second intron from
grass relatives such as rushes, sedges, cattails, and bur reeds.
75
Maize Sprouts Polarity as Factor Defining Exogenetic Phytohormones Influence
Buryak, Alla {1}
{1} Khmelnytskyy Scientific Research Institute of Crop Selection
The general morphogenesis theory, the most important side of which is systemic functions such as polarity,
gradients, correlation, etc., has been actively developed for the last years. These functions correlation with
hormone regulation processes and sex expression have been studied in this work by marking of gibberellic acid
(–ê3 - 2.9 10-1 m/l) and benzylaminopurine (6-BAP - 4.4 10-2 m/l) the various parts of plant sprouts of two maize
sorts (Zea mays Imdurata) early ripe – Podolskaya-98 and late ripe Krasnodarskaya 382. Phytohormones’
treatment has discovered that –ê3 has the greatest effect when marking apical or middle sprout parts, but 6-BAP
does when treating lower or middle ones (phytohormones distribution polarity is changing in 3-4 times). Initial
stages of development and flowering phases proved to be the most sensitive to the above treatment. Exogenetic
–ê3 is able to accelerate late ripe sorts earing for 5-7 days and early ripe ones for 4-6 days. The cytokinin has had
the opposite effect, particularly, by injecting 6-BAP into basal sprout part, which has caused delay of earing for 5-
6 days in early ripe sorts. 6-BAP rather delayed coming next development phases, but later development phases of
corncob went faster. In general, the cytokinin aided the earlier formation of ears and the increase of their amount.
The –ê3 usually caused decrease of the ears amount, particularly in early development phases. It has been
displayed that the accurate correlation between sprout parts, treated by the 6-BAP or –ê3, and the ears location
height. It is obvious that ear placing and development (possibly and sprout node ones) require specific
phytohormone balance in definite sprout part. To describe growth processes the mathematical models have been
developed (approximation accuracy R2 is never less than 0,999). Character of the plants growth speed of the both
sorts is evidence of A3 accelerating plants development. The model, describing 6-BAP influences, proved to be
the mirror reflection of the model describing A3 but with less effect. Furthermore, 6-BAP is removing critical
point of dependences change for 1,5-2 metameter higher, but for all that critical point is located at the same height
as in model using A3.
point of dependences change for 1,5-2 metameter higher, but for all that critical point is located at the same height
as in model using A3.
76
Pangloss1 and Pangloss2 are Required for Polarization of Subsidiary Mother Cells in the Formation of
Maize Stomata.
Cartwright, Heather (1) Gallagher, Kimberly (1) Smith, Laurie (1)
(1) University of California, San Diego
One of the most important events in plant development is the formation of stomata, which control the exchange of
gases and moisture between the plant and its environment. In maize, stomata are composed of four cells--a central
guard cell pair flanked by a pair of triangular subsidiary cells. These subsidiary cells are formed through the
asymmetric divisions of subsidiary mother cells (SMCs). We have isolated mutations in two genes that are
required for the polarized divisions of subsidiary mother cells. In pangloss1 and pangloss2 (pan1 and pan2)
mutants, SMCs often fail to polarize correctly, and frequently invade neighboring cell files. Cytoskeletal analysis
of pan1 and pan2 has revealed abnormalities in both actin and microtubule organization during subsidiary mother
cell division. Preliminary analysis of pan1;pan2 double mutants indicates these genes function independently of
one another to polarize SMCs and specify subsidiary cell fate. In brick1 (brk1) mutants, subsidiary cells often fail
to divide and specify correctly. Preliminary analysis of pan1;brk1 double mutants shows an additive phenotype,
indicating pan1 and brk1 likely act in separate pathways to specify subsidiary cell fate.
77
TRAITMILL TM :AN APPLIED GENOMICS PLATFORM FOR THE IMPROVEMENT OF CEREALS.
REUZEAU, Christophe {1} Frankard, Valerie {1} Hatzfeld, Yves {1} Sanz, Anabel {1} VanCamp, Wim {1}
Lejeune, Pierre {1} HervÈ, Philippe {1} Dillen, Willy {1} Peerbolte, Rindert {1} Broekaert, Willem {1}
{1} CropDesign NV, Technologie Park 3, Gent Zwijnaarde 9052, Belgium
CropDesign's TraitMillTM platform aims at closing the application gap between classical genomics and the
development of crops with improved or novel traits. TraitMill allows high-throughput and high-resolution testing
of the effect of plant-based transgenes on agronomically valuable traits in crop plants. The focus of the TraitMill
platform is on rice, a globally important crop, which is also a good model for other important cereals such as corn
and wheat. The Traitmill involves the following key components: i) Using a high throughput proprietary gene
cloning system, expression levels of transgenes and transgene combinations can be modulated throughout entire
plants or in selected tissues. ii) An industrialized plant transformation system generates the tens of thousands of
transgenic plants required annually to analyse the transgene constructs. iii) Several automated plant evaluation
technologies, including digital image analysis of plants, are used to detect alterations brought about in important
traits such as biomass production, seed yield, thousand kernel weight, and stress tolerance. At the input side of the
TraitMill, CropDesign's research team focuses on genes that modulate the size, shape and growth of specific plant
tissues, as well as the plant as a whole. Plant organ growth, as well as overall plant growth and development, are
also linked to interactions with the environment via multiple signal exchange networks, and by the production,
uptake, translocation and sensing of nutrients. Many of the genes controlling these complex processes remain
functionally undefined and are the target of functional genomics research at CropDesign. The TraitMill has
generated a range interesting product leads over the last year. These leads include transgenic rice lines showing
increased seed yield, increased green biomass, or increased tillering. A phenotypic description of these transgenic
lines will be presented.
78
Maize mutants arrested in early embryogenesis disclose an irregular pattern of cell divisions and altered
programmed cell death
Consonni, Gabriella {1} Dolfini, Silvana {1} Giulini, Anna {1} Colombo, Monica {1} Pilu, Roberto {1}
Gavazzi, Giuseppe {1}
{1}University of Milan
Embryo-specific (emb) mutants exhibit defects in the development of the embryo, while the endosperm is
essentially normal. Thus they may define genes whose products are specifically required for embryo
morphogenesis. The maize emb mutants analyzed in this work are arrested at the proembryo or early transition
stage. The block in embryo development is associated with abnormal suspensor proliferation, and disorganized
tissue proliferation persisting during seed maturation. The program of cell death appears impaired in these
mutants. In wild-type immature embryos, 14 days old, TUNEL positive nuclei are detected in the scutellar region
surrounding the shoot primordium and in the coleoptile whereas in the mutants no positive cells are detectable on
the embryo proper or on the long suspensor. An unexpected result is obtained when mutant embryos rescue is
attempted. Immature mutant embryos transferred to a basal medium germinate and produce single seedlings that
appear morphologically normal although retarded in their growth. The culture of these embryos may cause the
induction of the shoot by promoting the formation of a single SAM in an undifferentiated region of the embryo.
The analysis of emb mutants appears a promising tool to elucidate crucial points of embryo development such as
the coupling of cell division with morphogenesis, the cell to cell interactions, the relationship between embryo
and endosperm development and the interaction between embryo proper and suspensor.
appear morphologically normal although retarded in their growth. The culture of these embryos may cause the
induction of the shoot by promoting the formation of a single SAM in an undifferentiated region of the embryo.
The analysis of emb mutants appears a promising tool to elucidate crucial points of embryo development such as
the coupling of cell division with morphogenesis, the cell to cell interactions, the relationship between embryo
and endosperm development and the interaction between embryo proper and suspensor.
79
Functional analysis of the maize FT/TFL homologs reveals potential players in the floral transition.
Danilevskaya, Olga N. {1} Hermon, Pedro {1} Shirbroun, David {1} Bruggemann, Ed {1} Muszynski,
Michael G. {1} Ananiev, Evguenii V. {1}
{1} Pioneer Hi-Bred International, Inc.
The transition from vegetative to reproductive growth is a critical point in plant development. Despite the
importance of flowering time as a developmental process and agronomic trait, only one maize flowering gene, id1
(indeterminate-1), has been cloned. However, genetic pathways leading to the initiation of flowering are well
defined in Arabidopsis. The Arabidopsis flowering genes FT (flowering locus T) and TFL (terminal flower)
encode 23-kD proteins with similar structures, but antagonistic functions. FT promotes flowering, whereas TFL
acts as a repressor. The FT/TFL proteins are evolutionarily conserved between dicots and monocots. Eight maize
FT/TFL homologs were found in the Pioneer/DuPont maize EST (Expressed Sequencing Tags) database.
Phylogenetic analysis placed 6 maize genes into the FT clade and 2 into the TFL clade. Electronic Northern and
RT-PCR revealed different expression patterns of maize FT/TFL homologs. Both ZmTFL1 and ZmTFL2 are
expressed in the shoot apical meristem during vegetative growth but not expressed in the reproductive meristem.
This result is consistent with their possible roles as repressors of flowering. Conversely, only one maize FT
homolog, ZmFT1, is expressed in the apical meristem just after the floral transition. Five other FT homologs show
expression in different tissues. ZmFT1 transcripts are absent in the id1-m1 mutant, suggesting that ZmFT1
activation in the apical meristem during the floral transition requires Id1 function. ZmFT1 maps to the flowering
QTLs on chromosome 8. Our results suggest that ZmFT1 and ZmTFL1- ZmTFL2 are candidates for maize
flowering genes with possible antagonistic functions like the Arabidopsis genes.
80
Biochemical characterization of the FEA2 protein
Deleu, Wim {1} Wang, Jing {1} Jackson, Dave {1}
{1}Cold Spring Harbor Laboratory
The maize mutant fasciated ear2 (fea2) displays a massive over-proliferation of the ear inflorescence meristem.
The modestly affected floral meristem gives rise to a small increase in floral organ number. This suggests that
FEA2 acts to limit the growth of these meristems. Cloning of the FEA2 gene via transposon tagging showed it
encodes a leucine-rich repeat receptor-like protein, predicted to be plasma membrane targeted. The most closely
related protein is CLAVATA2 (CLV2) from Arabidopsis, which also negatively controls the proliferation of
pluripotent stem cells in shoot apical meristems. This indicates that the CLAVATA pathway has been functionally
preserved beyond the monocot to dicot split in evolution. Genetic and molecular data suggest CLV1and CLV2
form part of a receptor complex, to which a ligand encoded by CLV3 binds. Protein extracts from young ears
were size fractionated to see whether FEA2 is part of a higher molecular weight complex. Preliminary data shows
FEA2 is part of a membrane protein complex eluting between 400-500kDa. The FEA2 profile overlaps with that
of CLV1, which forms a complex of around 450kDa, though CLV1 also forms a 185kDa complex. In addition, it
was found that under denaturing conditions, the FEA2 monomer migrated at a higher molecular weight of 70 kDa
than the predicted 64kDa. Specific removal of oligosaccharides from N-linked glycoproteins accounted for this
size difference.
81
Molecular Genetic Analysis of rgh Endosperm Mutants
Fajardo, Diego {1} Georgelis, Nick {1} Parani, Madasamy {2} Positano, Mark {1} Carson, Chris {3} Coe,
Ed {3} Settles, Mark {1}
{1} University of Florida {2} Univesity of Toledo {3} University of Missouri
Much effort has been directed to maize endosperm improvement for its importance in feed and food making as
well as its use for secondary products such as oils. The maize endosperm shows many similar developmental
processes of other plant organs in addition to being an important storage reserve for the seed. An understanding of
endosperm developmental processes could provide a basis for develop more efficient approaches to seed
improvement as well as generate a model for organ development in maize. The rough endosperm (rgh) class of
seed mutants disrupts normal endosperm and embryo development and is characterized by seeds with a pitted or
etched surface. The development of a mutagenic inbred (UniformMu ) (McCarty et al., NSF award 007676)
provides a unique tool for efficient and rapid characterization of transposon-tagged mutants. UniformMu is a
Robertsonís Mutator transposon-active population that it is introgressed into the W22 color-converted inbred.
Robertonís Mutator elements typically have 50-200 copies/genome which creates challenges for molecular
genetic analysis. The UniformMu population is carefully pedigreed with known non-mutant progenitors and
carries a somatic transposon active marker to select against further transposition once mutants are identified. We
have identified ~80 rgh mutant isolates from the UniformMu population. Because of the high number of isolates
provides a unique tool for efficient and rapid characterization of transposon-tagged mutants. UniformMu is a
Robertsonís Mutator transposon-active population that it is introgressed into the W22 color-converted inbred.
Robertonís Mutator elements typically have 50-200 copies/genome which creates challenges for molecular
genetic analysis. The UniformMu population is carefully pedigreed with known non-mutant progenitors and
carries a somatic transposon active marker to select against further transposition once mutants are identified. We
have identified ~80 rgh mutant isolates from the UniformMu population. Because of the high number of isolates
(~180 estimated) required to uncover all rgh loci (~30 rgh loci) and the fact that most rgh homozygous mutants
are not viable, a traditional genetic analysis is not reasonable. Instead, we are using phenotypic, molecular and
genetic map data to sort rgh mutants into allelic groups. Once loci are identified we will concentrate on detail
genetic and morphological analysis of 3-5 rgh loci that represent the range of seed phenotypes within this class of
mutants.
82
Three Brick Genes Are Required for Polarized Growth and Division in Maize Leaf Epidermal Cells
Frank, Mary J. {1} Smith, Laurie G. {1}
{1} University of California, San Diego
Cells of the maize leaf epidermis assume a wide variety of shapes, and cell polarity plays an important role in
determining these diverse shapes. To gain new insights into the mechanisms governing cell polarity, we isolated
mutations in three Brick genes, which affect maize leaf epidermal cell morphogenesis. In wild type leaves,
epidermal pavement cells have finger-like projections along the lateral margins, which interlock with those of
neighboring cells. In brk mutants, pavement cells expand normally without forming these marginal lobes,
resulting in rectangular cells. Previous studies on the morphogenesis of lobed epidermal and mesophyll cells in a
variety of species have attributed lobe formation to the organization of cortical microtubules into bands that direct
a non-uniform pattern of cellulose deposition. Also, when wheat mesophyll cells were treated with an actin-
disrupting drug, these cells failed to form lobes, and microtubules did not form bands, suggesting a role for F-
actin in microtubule organization. In expanding brk epidermal cells, microtubules rearrange and form bands, but
the bands are less distinct than in wild type cells. However, localized enrichments of cortical F-actin seen at the
tips of emerging and expanding lobes in wild type cells were never observed in expanding brk cells, suggesting an
additional role for actin in lobe formation. In addition to the effect on pavement cell shape, in brk leaves, stomatal
subsidiary cells form abnormally, and epidermal hairs are shorter and blunter than wild type hairs. Double mutant
analysis indicated that all three Brk genes are involved in the same pathway or process affecting epidermal cell
morphogenesis. The Brk1 gene was cloned and shown to encode a small, novel protein, which is highly conserved
in both plants and animals. Recent work on HSPC300, the BRK1 mammalian homolog, has shown that HSPC300
forms a complex with four other proteins including WAVE1, an activator of the Arp2/3 complex. These
observations suggest that BRK proteins are involved in the local regulation of actin polymerization in plant cells.
83
Functional analyses of EMPTY PERICARP2, an essential regulator of the heat shock response that is
required for maize embyogenesis.
Fu, Suneng {1} Rogowsky, Peter {2} Meeley, Robert{3} Scanlon, MJ{1}
{1}Department of Plant Biology, University of Georgia, Athens, Georgia 30602. {2}UMR 5667 INRA-CNRS-
ENSL-UCBL, Lyon, F-69364, FRANCE {3}Pioneer Hi-Bred International, Johnston, Iowa 50131.
The heat shock response (HSR) is an evolutionarily conserved molecular/biochemical reaction to thermal stress
that is essential to the survival of eukaryotic organisms. Previous analyses reveal that recessive, null mutations at
the maize empty pericarp2(emp2) locus lead to dramatically increased expression of heat shock genes, retarded
embryo development, and early-stage abortion of embryogenesis. Cloning and sequence analyses revealed that
emp2 encodes a predicted protein with high similarity to HEAT SHOCK BINDING PROTEIN1, which was first
described in animals as a negative regulator of the HSR. The emp2 locus is duplicated in maize (Zmhsbp2) and
other grass species, although only one hsbp is identified in animals and eudicots. However, the embryo-lethal
emp2 mutant phenotype reveals that the hsbp homologues emp2 and Zmhsbp2 are functionally divergent during
kernel development. Clonal analyses further reveal that emp2 and hsbp2 perform non-redundant functions in adult
shoot tissues. Surprisingly, emp2/-sectors condition phenocopies of maize mutants affecting phyllotaxy, ligule
development and leaf development. Likewise, the differential interaction of EMP2 and ZmHSBP2 with maize
HEAT SHOCK FACTORS as revealed by yeast two-hybrid analyses further supports the model that EMP2 and
HSBP2 are functionally divergent. Moreover, the developmental retardation of emp2 mutant kernels before they
are competent to invoke the HSR suggests an additional role for EMP2 during embryo development, beyond
regulation of the HSR. We are currently investigating this model by yeast two-hybrid screening and
immunoprecipitation analyses of developing maize kernels.
84
The Protein Secretory Pathway is Upregulated in Several opaque Mutants
Gibbon, Bryan C {1} Larkins, Brian A {1} Jung, Rudolf {2}
{1} University of Arizona {2} Pioneer Hi-Bred international
To better understand the mechanisms that lead to an opaque or starchy endosperm phenotype, we performed a
thorough analysis of mRNA transcript profiles in several opaque mutants: opaque1, opaque2 (o2), opaque5,
opaque9, opaque11, floury2, defective endosperm B-30, and mucronate. The mutations were introgressed into the
inbred line W64A, backcrossed six generations, and transcript profiles of mRNAs present at 23 DAP were
obtained using Curagenís Gene Calling technology. Genes that showed changes in expression for a majority of the
mutants revealed information about disrupted metabolic processes. One generally altered metabolic pathway was
protein synthesis. The expression of several ribosomal protein genes was reduced approximately two-fold.
However, there was no evidence for a reduction in the abundance of protein synthesis initiation or elongation
factors. The most surprising change was a substantial increase in expression of genes related to vesicle trafficking
and organelle biogenesis. These included cytoskeletal genes, such as actin and katanin. Additionally, mRNAs
encoding Sec31 and t-SNARE homologs, factors involved in vesicle budding and fusion, were increased.
Furthermore, vacuolar sorting receptors, tonoplast intrinsic proteins, and proteins predicted to localize to vacuolar
compartments were increased in expression. These changes indicate that protein storage may be redirected from
ER-bound zein protein bodies to protein storage vacuoles. Proteomic analysis of CM105, CM105 o2, and
modified CM105 o2 endosperm proteins revealed that globulin 1 and globulin 2 are increased significantly, as is a
Rab GTPase, further supporting the hypothesis that the protein secretory pathway is altered in opaque mutants.
85
GENOMIC CLONING THROUGH APOMIXIS RESULTS IN EXTENSIVE EPIGENETIC VARIATION
Grimanelli, Daniel {1,2} Leblanc, Olivier {1,2} Perotti, Enrico {2}
{1} IRD, Institut de Recherche Pour le DÈveloppement {2}, CIMMYT, APDO 6-641, 06600 Mexico DF Mexico
Apomixis in higher plants is defined as the process of cloning through seeds. Apomictic plants bypass both female
meiosis and fertilization, resulting in progenies that are genetic replica of their mother plant. This method of
genetic cloning, which does not exist as yet in any major crop is regarded as a revolutionary tool in agriculture; it
could allow fixing indefinitely the best allelic combinations, including those of high yielding hybrids. Here, we
studied the genetic consequences of apomictic cloning in hybrid plants between maize and its apomictic wild
relatives, Tripsacum dactyloides. We show that apomixis, while allowing faithful reproduction of the genome of
the maternal plant, also results in extensive phenotypic variation. This variation results from inaccurate
reproduction of the epigenetic state prevailing in the mother genome.
86
Clonal analysis of Wavy auricle in blade (Wab)
Foster, Toshi (1) Johnston, Robyn (2) Hay, Angela (3) Hake, Sarah (1)
(1) Institute of Developmental Phenomenology, New Zealand (2) Massey University, Palmerston, NZ (3) Oxford
University, England
Maize leaves are composed of two distinct parts, the distal blade, which is flat and leans away from the plant, and
the sheath, which wraps around the culm. At the junction of the blade and sheath is the ligule region which
contains the ligule itself, a flap of epidermal fringe, and the wedge-shaped auricles on either side of the leaf. The
auricles function as hinges for the blade. The liguleless mutants, lg1 and lg2, lack auricle and ligule, but still have
a clearly distinct blade and sheath. Wavy auricle in blade (Wab) mutants are characterized by two distinct
features, the leaves are more narrow than wild-type, and proximal tissues, such as sheath and auricle, are found in
the blade. The leaf blade of double mutants between lg1 or lg2 and Wab is very narrow. The sheath extends into
the blade such that a distinction between the two tissues is lost except at the midrib. We have carried out a clonal
analysis, in the presence and in the absence of a functional lg1 gene, to determine how Wab regulates leaf shape.
Our results suggest that the two defects may be separable, white sectors in the blade (indicating the absence of
Wab) lack sheath and auricle characteristics, but are still narrow. The data also allow us to produce a fate map of
the narrow lg; Wab leaves.
87
ragged seedling2 fails to maintain the dorsoventral axes of leaf tissues in maize
Henderson, David {1} Muehlbauer, Gary {2} Heinen, Shane {2} Scanlon, Mike {1}
{1} University of Georgia, Department of Plant Biology {2}University of Minnesota, Department of Agronomy
and Plant Genetics
Unlike previously described dorsoventral mutants in various plant species, radial leaves of ragged seedling2
(rgd2) mutants retain dorsoventral polarity. Less-severe mutants exhibit planar leaves with defects in the
dorsoventral orientation of vascular and epidermal tissue: vascular bundles are rotated out of their normal
positions and the lamina of mildly affected leaves display patches of epidermal tissue that have an inverted
dorsiventral identity. rgd2 is a novel, recessive mutation that maps to chromosome 1S, proximal to the
translocation breakpoint and near roughsheath2. Evaluation of rgd2 mutants provides important insights into
characteristics of adaxial and abaxial leaf tissue, dorsoventral patterning of maize leaves, and leaf development in
general.
dorsoventral orientation of vascular and epidermal tissue: vascular bundles are rotated out of their normal
positions and the lamina of mildly affected leaves display patches of epidermal tissue that have an inverted
dorsiventral identity. rgd2 is a novel, recessive mutation that maps to chromosome 1S, proximal to the
translocation breakpoint and near roughsheath2. Evaluation of rgd2 mutants provides important insights into
characteristics of adaxial and abaxial leaf tissue, dorsoventral patterning of maize leaves, and leaf development in
general.
88
PARAMUTATION AND PLANT DEVELOPMENT REQUIRE rmr12 FUNCTIONS
Hollick, Jay {1}
{1} University of California, Berkeley
Genetic and molecular analyses of paramutation at the purple plant 1(pl1) locus are ongoing to gain insights into
general chromosome processes used in gene control. The PL1 R2R3-Myb protein is a transcriptional activator of
the genes encoding enzymes required for anthocyanin biosynthesis. Heritable epigenetic changes occurring at pl1
are thus reflected by visual differences of anthocyanin pigment levels and tissue-specific pigment patterns. The
Pl1-Rhoades allele (Pl) can confer relatively high levels of anthocyanin pigment production in many sporophytic
tissues. However, this allele is unstable and can spontaneously change to a transcriptionally-repressed,
meiotically-heritable, derivative designated Pl1'-mahogany (Pl'). The Pl allele exclusively changes to Pl' when
carried in heterozygous condition with another Pl' allele; a phenomenological hallmark of paramutation. Stable
maintenance of the silenced Plí state requires undefined interactions with a second Pl-Rh or Plí allele and at least
eight distinct trans-acting factors, mop1 (mediator of paramutation; Dorweiler et al 2000), rmr1 and rmr2
(required to maintain repression; Hollick and Chandler 2001), rmr6, rmr7, rmr8, rmr9, and rmr11 (J. Hollick,
unpublished). Genetic and phenotypic analyses of two mutations isolated from an EMS pollen mutagensis define
one additional trans-acting factor encoded by a novel locus, rmr12. Both mutations behave as stable recessive
alleles that are unlinked to the pl1 locus and fail to genetically complement each other. Homozygous mutant
plants that are genotypically Plí/Plí display a strong Pl-like phenotype while the expression of other pl1 alleles
appears unaffected. Unlike most rmr mutations, homozygous rmr12 mutants have strong defects in plant growth
and development. These defects are distinct from those associated with mop1 and rmr6 mutations, thus indicating
that individual components required for paramutation-based epigenetic repression play selective, rather than
simply general, roles in the control of developmental genetic programs. Hollick, J.B. and Chandler, V.L. (2001)
Genetics 157: 369-378. Dorweiler, J.E. et al. (2000) The Plant Cell 12 (11): 2101-2118.
89
The characterization of pleiotropic shoot phenotype in leafy coleoptile(lco)1-R
Inada, Noriko{1} Akulova-Barlow, Zoya{1} Freeling, Michael{1}
{1}University of California, Berkeley
In the novel maize coleoptile mutants leafy coleoptile(lco), the coleoptile is converted into leafy structure. The
lco1-R coleoptile has multiple veins with blade/sheath and ligule development at the margin. It originally splits in
the embryo and lacks mesocotyl. Besides this coleoptile phenotype, lco1-R exhibits various shoot defects. The
dwarf plant lacks any reproductive organs. The leaves are narrower and shorter than normal leaves and without
midrib. Not only the midrib, but the vasculature is also affected, and epidermal structure is disorganized. The
ligule displacement is often found in the seedling leaves. Immunolocalization has given a pattern of RS1/GN1
knox genes expression in leaves that indicates some extopic expression, which is now being confirmed.
90
Hypomethylation at Pl-blotched is reset by shoot meristem culture-induced rejuvenation
Irish, Erin {1} Doug McMurray {1}
{1} The University of Iowa
Phase change is the transition from one epigenetic state, the non-flowering, juvenile phase, to another, the adult
phase where flowering can occur. Methylation status of the DNA at the blotched allele of Pl isolated from leaves
is correlated with phase of the plant at the time the leaf differentiated (1), and the juvenile pattern of methylation
is restored with each new generation. To test whether methylation status can be modified independently of
passage through a sexual generation, we used shoot meristem culture (2) to rejuvenate Pl-blotched plants. We
found that of 22 plants rejuvenated by culture, all showed initial reductions in methylation, followed by a return to
adult patterns of methylation upon acquisition of the adult phase morphology. Some 13 of these plants had
methylation patterns identical to seedlings in their basal leaves. Thus, methylation status can be reversed upon
rejuvenation of the shoot. This result supports the view that methylation provides a mechanism of maintaining an
epigenetic state. 1. Hoekenga, Muszynski, and Cone (2000), Genetics 155, 1889-1902. 2. Irish and Karlen (1998),
Int. J. Pl. Sci.159, 695-701.
91
Inbred modifier effects on adaxial patterning of the Maize leaf
Juarez, Michelle {1,2} Timmermans, Marja {1}
{1} Cold Spring Harbor Lab, Cold Spring Harbor, NY 11724 {2} Graduate Program in Genetics, State Univ. of
New York at Stony Brook, Stony Brook, NY11794
We have identified three mutant loci that affect adaxial patterning in the developing maize leaf. All mutant loci
have been introgressed five times into several different inbred lines. Dramatic differences in the severity of the
mutant phenotypes were observed in the various inbred backgrounds. Interestingly, each inbred line modifies the
severity of the set of mutations in a similar way. This suggests that all three mutations affect the same process,
which is modified in a consistent manner in all inbreds analyzed. One exception to the general pattern is the
ragged seedling (lbl1-rgd1) allele of leafbladeless1 (lbl1) in the W22 background. In contrast to the extremely
mild phenotypes of the other adaxial patterning mutants, the phenotype of lbl1-rgd1 is most severe in W22. This
suggests the presence of a lbl1-rgd1 allele specific modifier in W22. Intercrosses were made with lbl1-rgd1 in the
W22 and B73 inbred backgrounds to determine the action of a W22 inbred specific modifier. The phenotypes
observed are consistent with the presence of a single dominant modifier in W22.

In the W22 inbred the lbl1-rgd1 mutation is embryo lethal. The lbl1-rgd1 mutant embryo fails to form any apical
structures, however root development appears normal. SEM analysis and knotted1 expression in the early embryo
illustrates a smaller apical region and absence of meristematic cells in the apex. Perhaps the loss of the adaxial
cell identity in the early embryo results in a meristem maintance defect. In wild type vegetative SAM the maize
yabby gene (zyb) is expressed in the adaxial domain of the young leaf primordia. We are currently analyzing the
expression pattern of zyb in the scutellum. Which will provide more insight into the nature of the polarity defect in
the lbl1-rgd1 mutant embryo.
92
Interactions between XCL1 and KNOX genes
Kessler, Sharon {1} Sinha, Neelima {1}
{1} University of California, Davis
An important question in plant developmental biology is how division and differentiation are balanced in shoot
apical meristems (SAMs) in order to maintain meristematic cells and to allow for the production of leaf primordia
from the flanks of the meristem. KNOX genes play an important role in the maintenance of meristem identity. In
simple-leafed model species such as maize and Arabidopsis, class 1 KNOX genes are downregulated in the P0
and remain turned off throughout leaf development. Dominant, neomorphic mutations in maize KNOX genes
such as KNOTTED1 (KN1), lead to the production of knots of tissue over veins and ectopic ligules which require
periclinal cell divisions in epidermal cells (Gelinas et al., 1969). In addition, the severity of kn1 loss-of-function
phenotypes is worsened when the mutation is introgressed into inbred backgrounds with shorter meristems
(Vollbrecht et al., 2000), indicating that KN1 plays a role in maintaining meristem size. Extra cell layers1 (Xcl1)
is a semi-dominant, hypermorphic mutation that causes the overproduction of a normal gene product and affects
cell division and differentiation patterns in developing leaves and kernels. Oblique, periclinal divisions in the
protoderm give rise to multiple epidermal layers in leaves. Leaves are twice as thick but only half as wide as
normal sibling leaves. In addition, Xcl1 meristems are 60% shorter than normal sibling meristems, and recruit
more cells into P0 (Kessler et al., 2002). Thus, Xcl1 shares some phenotypic similarities with KNOX mutants.
The short meristems and increased recruitment of cells into P0 seen in Xcl1 mutants may reflect compromised
KNOX activity. Double mutant analysis between Xcl1 and dominant KNOX mutants indicate that Xcl1 is a
suppressor of the Kn1 phenotype, but may be an enhancer of Gnarly1 and Rough sheath1. The developmental
implications of these regulatory relationships will be discussed.
93
Spatial and temporal control of programmed cell death (PCD) in developing caryopses of maize.
Kladnik, Ales {1} Chamusco, Karen {2} Dermastia, Marina {1} Chourey, Prem {2,3}
{1} University of Ljubljana, Slovenia {2} University of Florida {3} USDA ARS
We show here that a developing caryopsis is chimeric in spatial and temporal control of PCD that occur
throughout the duration of kernel development. PCD was visualized at a cellular level in both maternal (placento-
chalazal layer in pedicel and nucellus) and filial endosperm tissues (regions surrounding the embryo and starchy
cells). PCD in the placento-chalazal layer was fertilization-dependent and was marked by loss of nuclei and all
subcellular membranous organization in all cell layers during 6 ñ 12 days after pollination. These ëemptyí cells,
however, remained intact throughout kernel development. Because enucleation was initiated early, coincident
with endosperm cellularization, and was completed rapidly prior to the beginning of the storage phase in
endosperm, we suggest that PCD may be causal to functional activation of the placento-chalazal layer.
Remarkably, a second wave of PCD, apoptotic in nature as diagnosed by TUNEL stain, was seen in only a subset
of cells underneath the enucleated placento-chalazal cells. Furthermore, such TUNEL staining was seen in the
wild type, Mn1, but not in the invertase-deficient miniature seed1 (mn1) seed mutant that leads to reduced size
and weight of the endosperm. Overall, the two pathways to PCD in these maternal cells appeared to be signaled
by various developmental changes in filial endosperm and its metabolic sink strength. By contrast, the enucleation
endosperm, we suggest that PCD may be causal to functional activation of the placento-chalazal layer.
Remarkably, a second wave of PCD, apoptotic in nature as diagnosed by TUNEL stain, was seen in only a subset
of cells underneath the enucleated placento-chalazal cells. Furthermore, such TUNEL staining was seen in the
wild type, Mn1, but not in the invertase-deficient miniature seed1 (mn1) seed mutant that leads to reduced size
and weight of the endosperm. Overall, the two pathways to PCD in these maternal cells appeared to be signaled
by various developmental changes in filial endosperm and its metabolic sink strength. By contrast, the enucleation
and PCD in autolysing nucellar tissue and endosperm cells surrounding embryo, and in starchy endosperm that
showed only the enucleation in otherwise well-packed cells that remain intact until germination, preceded and/or
coincided with the TUNEL stain.
94
Sucrose synthase isozyme SUS1 in the maize root cap is preferentially localized in the endopolyploid outer
cells
Kladnik, Ales {1} Vilhar, Barbara {1} Chourey, Prem S. {2,3} Dermastia, Marina {1}
{1} University of Ljubljana, Slovenia {2} University of Florida {3} USDA-ARS
Spatial distribution of sucrose synthase isozyme SUS1 was studied in the maize root cap in relation to the level of
endopolyploidy. Median longitudinal root cap sections were analyzed using immunolocalization, quantitative
DNA staining and image cytometry. High immunolocalization signal for the SUS1 protein was detected in the
cells of the outer root cap, while the signal was low in the meristematic and the central root cap region. The
meristematic cells had ploidy levels of 2C and 4C corresponding to the G1 and G2 phases of the cell cycle,
respectively (1C is the DNA content of the unreplicated haploid genome). With respect to spatial distribution,
both the SUS1 signal and the level of endopolyploidy increased from calyptrogen towards the root cap periphery,
and were thus the highest in the outer root cap cells. About 90% of cells with ploidy level 16C or higher exhibited
a high SUS1 signal, compared to only 5% of the 2C cells. The high amount of the SUS1 protein in the outer,
endopolyploid cells suggests an association between endoreduplication and the abundance of this enzyme.
Because the outer root cap cells are involved in mucilage production, sucrose synthase located in these cells may
provide monosaccharide precursors for mucilage synthesis.
95
Mu-Tagged Empty Pericarp Mutants in the UniformMu Maize Population
Xue, Song (1) Settles, A. Mark (1) Huang, Li-Fen (1) Avigne, Wayne T. (1) Larkins, Brian A. (2) Becraft,
Philip (3) Dukes, Will (1) Baier, John W. (1) Hannah, L. Curtis (1) McCarty, Donald R. (1) Koch, Karen E.
(1) University of Florida (2) University of Arizona (3) Iowa State University
A phenotypic analysis was undertaken to further define the extent of variation among the Mu-tagged empty
pericarp (ep) mutants isolated from the UniformMu population. The ep mutants identified thus far (ca 25% of
total Mu-tagged seed mutants) were initially classified on the basis of external visual analysis that indicated little
or no internal seed structure was present. However, examination via longitudinal sections showed significant
endosperm and/or embryo formation in many 'ep' kernels. The range in degree of mutant severity revealed
corresponding differences in timing of developmental termination for embryo and/or endosperm. The ep mutants
of intermediate severity in this population often showed evidence of initial endosperm or nucellar development
resulting in hollow, translucent, paper-like structures within kernels. Further analysis will be directed toward
identifying the sites and timing of developmental cessation in the different ep classes observed in the UniformMu
population, and on molecular isolation of the relevant genes.
96
The indeterminate floral apex1 (ifa1) mutant phenotype is associated with a mutation in the Zea mays
MADS box 14 (ZMM14) gene.
Krolikowski, Katherine {1} Yamawaki, Traci {2} Laudencia-Chingcuanco, Debbie {3} Newman, Lisa {4}
Hake, Sarah {1}
{1} Plant Gene Expression Center, USDA-ARS and University of California, Berkeley {2} University of
California, San Francisco {3} USDA-Agricultural Research Service, Albany, CA {4} Pioneer Hi-Bred
International, Inc. De Moines, IA
Maize inflorescence architecture is controlled by the activity of inflorescence meristems during development.
Both the identity and determinacy of inflorescence meristems are afffected in the indeterminate floral apex1 (ifa1)
mutant. Ifa1 mutants have a loss of determinacy in the Floral Meristem (FM), Spikelet Meristem(SM), and
Spikelet Pair Meristem (SPM) of the tassel, displaying extra floral organs, extra florets and and extra spikelets in
the male inflorescence. In the female inflorescence, the FM?s are affected, with nucellar-like proliferative tissue
growing at the floral apex between rudimentary silks. MADS box genes are involved in floral development in a
wide range of plants, affecting both organ identity and meristem determinacy. Many lines of evidence indicate
that the ifa1 phenotype is caused by a mutation of the MADS box gene, ZMM14. Mapping data and expression
studies show that ZMM14?s map position and expression are linked to the mutant phenotype. A Mu insertion
150bp 5?of ATG shows a weak ifa1-like phenotype when homozygous, and complementation crosses between
this insertion and the ifa1-Reference allele are ongoing. A strong ifa1-like phenotype is observed when the
duplicate gene, ZMM8, is interrupted by a Mu insertion. Taken together, this evidence suggests that the ifa1
phenotype is probably caused by a mutation at the ZMM14 locus.
that the ifa1 phenotype is caused by a mutation of the MADS box gene, ZMM14. Mapping data and expression
studies show that ZMM14?s map position and expression are linked to the mutant phenotype. A Mu insertion
150bp 5?of ATG shows a weak ifa1-like phenotype when homozygous, and complementation crosses between
this insertion and the ifa1-Reference allele are ongoing. A strong ifa1-like phenotype is observed when the
duplicate gene, ZMM8, is interrupted by a Mu insertion. Taken together, this evidence suggests that the ifa1
phenotype is probably caused by a mutation at the ZMM14 locus.
97
The maize Disorgal 1 and Disorgal 2 genes restrict mitotic division plane in the aleurone layer and is
necessary for normal epidermal cell development and plant stature.
Lid, Stein Erik {1} Al, Ronald {1} Li, Bailin {3} Krekling, Trygve {1} Meeley, Liz {2} Utskarpen, Audrun
{1} Meeley, Robert B. {2} Opsahl-Ferstad, Hilde-Gunn {1} Olsen, Odd-Arne {1, 2}
{1} Dept. of Chemistry and Biotechnology, Agricultural Univ. of Norway, P.O. Box 5040, N-1432 ≈s, Norway
{2} Pioneer Hi-Bred International, a DuPont company, Johnston, IA 50131, USA {3} DuPont Crop Genetics,
DTP 200, 1 Innovation Way, Newark, DE 19711, USA
The aleurone layer of maize endosperm consists of a single layer of highly regular epidermal like cells. As part of
our effort to identify genes controlling the cell division plane in the aleurone layer, we are characterizing two
mutant loci, disorgal 1 (dal1) (three alleles) and disorgal 2 (dal2) (one allele) which display a highly irregular
organization of the aleurone layer. Both dal1 and dal2 are inherited as simple mendelian traits. In dal1 and dal2
homozygous endosperms, the control of mitotic division plane is relaxed or missing, resulting in grains with
disorganized aleurone layers. The cells of the aleurone layer are highly irregular, both with respect to cell size and
shape, and often more than one layer of aleurone cells are present. Homozygous dal1 and dal2 grains are
shrunken due to reduced size of the starchy endosperm and developmental arrest of the embryo at premature
stages. Mature mutant grains germinate at a very low rate and fail to develop into plants. However, homozygous
mutant plants can be obtained through embryo rescue. dal1 plants have broad leaves that become necrotic towards
the tip, whereas dal2 plants are extreme dwarfs. Both dal1 and dal2 plants display irregular leaf epidermis as well
as reduced number of root hairs. Taken together, these results indicate that dal1 and dal2 are involved in mitotic
control and cell patterning in epidermal like tissues. Efforts aiming to clone the dal1 and dal2 genes are
underway.

98
Genetic Analysis of the Maize Rop2 Gene
Arthur, K.{1} Vejlupkova, Z.{1} Fowler, J.E. {1}
{1} Molecular and Cellular Biology Program and Department of Botany and Plant Pathology. Oregon State
University, Corvallis, Oregon 97331, USA.
Rop family GTPases have been implicated in the regulation of plant signal transduction and cell morphogenesis.
In collaboration with Pioneer Hi-Bred, we have isolated a collection of mutant rop2 alleles to explore ROP2
function in maize. In genetic crosses, three of five rop2 mutant alleles (rop2-m1, rop2-m2, and rop2-m5) transmit
at a lower frequency through the pollen than the wild-type allele. Transmission through the female gametophyte
was not affected by the rop2 mutation. The male-specific transmission defect associated with the rop2 mutations
strongly supports an important role for rop2 in pollen function. We have been investigating the effects of the rop2
mutation on pollen germination, pollen tube growth and guidance, and fertilization. Measurement of pollen tube
germination and growth rate in vitro indicated that both are unaffected by the rop2 mutation. In vivo pollen tube
growth patterns, as well as pollen tube penetration of the style, did not appear significantly different between
wild-type and rop2 mutant pollen. We are now focusing on the final stages of pollen function: pollen tube growth
patterns near the ovule and fertilization of the egg sac.
99
Comparative Analysis of MADS-Box Sequences in Arabidopsis and Rice
Mao, Long {1} Kang, Hui {1} Wang, Danhui {1} Duvall, Melvin {1}
{1} Department of Biological Sciences, Northern Illinois University, DeKalb, IL 60115
MADS-box genes are important transcription factors involved in various biological processes in plants, especially
in floral development. The availability of the whole genomes of Arabidopsis and rice makes it possible to conduct
comparative analysis of this gene family in the two model plants of dicots and monocots respectively. We
compared 62 rice MADS-box sequences with 72 Arabidopsis MADS-box sequences. Among nine phylogenetic
groups containing both rice and Arabidopsis MADS-box sequences, six were involved in flower development or
flowering time; one group represented by AGL12 works in root development, and the other two groups have not
been functionally studied. There are nine groups containing only Arabidopsis MADS-box sequences with more
than two members in each group, including the MADS-box genes responsible for vernalization represented by
FLOWER LOCUS C. Five groups contain only rice MADS-box sequences, including the groups composed by
OsMADS2 and OsMADS4 that are potential orthologues of Arabidopsis PISTILATA (PI). A consensus sequence
was generated using MADS-box sequences from Arabidopsis, rice and tomato. Comparison of MADS-box
sequences with this consensus sequence showed that functionally known MADS-box sequences were most similar
been functionally studied. There are nine groups containing only Arabidopsis MADS-box sequences with more
than two members in each group, including the MADS-box genes responsible for vernalization represented by
FLOWER LOCUS C. Five groups contain only rice MADS-box sequences, including the groups composed by
OsMADS2 and OsMADS4 that are potential orthologues of Arabidopsis PISTILATA (PI). A consensus sequence
was generated using MADS-box sequences from Arabidopsis, rice and tomato. Comparison of MADS-box
sequences with this consensus sequence showed that functionally known MADS-box sequences were most similar
to the consensus sequence, whereas nearly half of the MADS-box sequences that have not been studied
functionally were highly diversified. Our results showed that, although rice and Arabidopsis have respectively
developed MADS-box genes to meet their special developmental needs such as flowering time, the ABC model
that was firstly established in dicots could almost fully extend to monocotelydonous plants such as rice.
100
Developmental Analysis of Transgenic Maize Lines that Overexpress Glossy15
Moose, Stephen {1} Kampani, Archana {1} Lauter, Nick {1}
{1} University of Illinois at Urbana-Champaign
The glossy15 (gl15) gene activates juvenile and suppresses adult leaf epidermal cell identity in response to
upstream factors that regulate vegetative phase change. Previous genetic and molecular analyses suggested that
the expression of juvenile leaf epidermal traits during shoot development is correlated with gl15 activity, which is
stimulated by the Corngrass1 (Cg1) and Teopod (Tp) mutations but downregulated by factors such as gibberellins
and Viviparous8 (Vp8) that promote the switch to adult vegetative development. To test this hypothesis, we
generated transgenic maize lines that overexpress gl15 in an appropriate developmental context, due to the
integration of additional functional copies of a gl15 genomic fragment that fully complements gl15 mutants. The
phenotypic effects of gl15 overexpression were characterized in a number of different genetic backgrounds,
including inbred lines that differ in their proportion of juvenile leaves as well as the Cg1, Tp1, Tp2, dwarf1 and
vp8 mutations that enhance the expression of juvenile traits during shoot development. The results from these
analyses support a model where gl15 operates via a threshold activity mechanism to regulate leaf epidermal cell
identity, with gl15 activity levels being determined by the antagonistic effects of upstream factors that promote
either juvenile or adult shoot development. High gl15 activity is sufficient to specify juvenile and suppress adult
leaf epidermal cell differentiation, resulting in leaves with only juvenile traits. Intermediate levels of gl15 activity
result in the expression of both juvenile and adult epidermal cell traits (e.g. transition leaves), whereas only adult
traits are expressed once gl15 activity falls below a minimum threshold required to activate juvenile traits.
101
Functional genomics of rice grain development : Establishement of a database of mutant phenotypes and
enhancer trap gus expression in mature seeds.
BARNOLA, Isabelle {1} GAY, CÈline {2} MODAT, Sylvain {3} GUILLEMINOT, Jocelyne {3}
LARMANDE, Pierre {2} DEVIC, Martine {3} THIS, Dominique {2} GUIDERDONI, Emmanuel {2} PAUL,
Wyatt {1} PEREZ, Pascual {1}
{1} BIOGEMMA, Gene Function and Maize Traits Group, 24, Avenue des Landais, 63 170 AUBI»RE,
FRANCE. {2} UMR 1096 GACA, CIRAD Amis, TA 40/03, 34398 MONTPELLIER, FRANCE. {3} Laboratoire
GÈnome et DÈveloppement des Plantes, CNRS-UMR 5096, UniversitÈ de Perpignan, 52 Avenue de Villeneuve,
66860 PERPIGNAN-cedex, FRANCE.
In the framework of the French national initiative GÈnoplante, a T-DNA/Tos17 insertion lines collection in rice is
being developed in Montpellier (See communication 'Creation of a genome-wide library of insertion lines in rice',
Sallaud et al.). This mutants library gives the very interesting opportunity to collect knowledge on grain
development for rice and cereals in general. We are screening the insertion lines for seed mutant phenotypes and
for expression of a reporter gene in mature seeds. In most cases, the collection allows to directly observe some
seed phenotypes on the primary transformantsí progenies without additional sowing. Indeed, we can easily
observe a mendelian segregation where a quarter of the seeds presents a mutant phenotype in a case of a recessive
mutation. We have already evaluated several thousand lines mainly for seed and endosperm morphology. We will
present the different types of mutations found and their frequencies. Using the gusA gene enhancer trap located in
the T-DNA, we have also evaluated the gus expression pattern in dry seeds. Examples and frequencies of gus
expression in seeds will be presented. Our first results in rice and the results obtained in well-characterised
Arabidopsis mutants collections will be compared. Moreover, FST sequences available for a large proportion of
the library will give us access to putative key genes involved in seed development and filling. Cytological and
molecular aspects of key developmental stages of rice seed are also investigated in order to further characterise
more precisely interesting mutants. Finally we will discuss how the development of such rice kernel phenotype
database will allow in silico reverse genetics for any rice and cereals genes and will serve most of the Genoplante
Seed and Grain projects.
102
extended auricle (eta), an Essential Component in the Developmental Network Controlling Maize Leaf
Development
Sadeghian, Nasim {1} Osmont, Karen S. {1} Freeling, Michael {1)
{1} University of California-Berkeley
We have characterized a novel mutant eta, extended auricle, which is involved in proximodisal patterning in the
maize leaf. The most noticeable phenotype is an extension of auricle tissue in to blade and a distal displacement of
the blade/sheath boundary. While the eta phenotype is very similar to the suite of dominant Knotted1-like
homeobox (knox) mutants, it is a recessive mutation and does not ectopically express knox genes. However, eta
acts as an enhancer in combination with proximodistal patterning mutants like Kn1, Lg3, Gn1, Rs1and rs2. There
is a precedent for this type of interaction in Arabidopsis. The Arabidopsis ortholog of rs2, asymmetric1shows a
genetic interaction with the chromatin remodeling factor pickle (pkl). Although pkl itself does not ectopically
express knat1 and knat2, it is thought to act as a general repressor of knox genes in the leaf based on its
enhancement of as1. eta also has a synergistic interaction with lg1 and a dosage effect with lg2. Neither lg1 nor
lg2 are misexpressed in the eta mutant. SEM analysis reveals Eta+ functions early in leaf development to properly
establish the blade/sheath boundary. In addition, we describe results from a mosaic analysis showing Eta+
functions cell autonomously and is required in all three internal cell layers for wildtype phenotype.
103
A transient expression system for maize silks
Sandahl, Jeanne {1} Crane, Virginia {1}
{1} Disease Resistance/Crop Genetics Research and Development DuPont Agriculture and Nutrition
From the moment they emerge from the husks, maize silks provide an entryway into the ear for pollen and various
pathogens. As part of an effort to engineer plants resistant to earmold, we developed a transient expression system
for the analysis of novel gene expression in silks. Silks are bombarded while still attached to cob explants, and are
then cultured on a sucrose/ascorbic acid-containing media for up to several days. We have used this system to
study silk gene expression both before and after pollination. Experiments with the maize ubiquitin promoter
driving GUS or a reef coral fluorescent protein reveal pollination-triggered changes in expression patterns. This
transient system serves as a tool for screening transgenes in silks prior to larger-scale, stable analysis.
104
Developmental analysis of the ramosa3-fasciated ear 1 mutant
Satoh, Namiko {1} Jackson, David {1}
{1} Cold Spring Harbor Lab.
To understand what kind of genetic and molecular mechanisms regulate maize inflorescence architecture, we have
analyzed the mutants that show fasciated phenotypes, in our lab. Interestingly, the fasciated ear 1 (fea1) mutant
which is allelic to ramosa 3 mutant displays a fasciated ear phenotype in some genetic backgrounds (A619 and
W23) and a branched ear phenotype in another background (B73). That is, in A619 and W23, fea1 ears
abnormally flatten with irregular rows of seeds, and in B73 some spikelet pair meristems at the bottom of the ears
go on to form indeterminate branch meristems. It seems the genetic modifier in A619 is dominant to the one in
B73. We will present scanning electron microscopy analysis, histological analysis and in situ hybridization with
the KNOTTED 1 (KN1) gene marker. Preliminary results of double mutant analysis with kn1, and mapping data
using simple sequence repeat polymorphisms will also be presented.
105
Inhibition of polar auxin transport disrupts KNOX protein regulation, founder cell recruitment, and
elaboration of leaf margins in maize shoots.
Scanlon, Mike {1}
{1} University of Georgia
Maize leaves develop basipetally (Poethig and Szymkowiak, 1995); the upper blade emerges first from the shoot
apical meristem (SAM) whereas the sheath expands later in leaf development. Founder cells, maize leaf initials
located in the peripheral zone of the SAM, are distinguished from cells in the SAM proper by the differential
accumulation of KNOX proteins (Smith et al., 1992). KNOX proteins accumulate in the SAM, but are excluded
from maize leaf primordia and leaf founder cells. As in Arabidopsis and tomato (Reinhardt et al., 2000), maize
shoots containing five extant leaf primordia fail to initiate new leaves when cultured in the polar auxin transport
inhibitor NPA. We demonstrate that NPA-induced arrest of leaf initiation in maize is correlated with the failure to
downregulate KNOX accumulation in the peripheral zone. Moreover, inhibition of polar auxin transport in
cultured apices results in the formation of tubular leaf bases, in which the margins fail to emerge and separate in
the lower leaf zone. The tubular leaf bases always formed in the fourth leaf from the meristem (P4). Invariably,
the unseparated margin regions of the tubular leaves accumulate high levels of ectopic KNOX proteins. These
data suggest that the lower, sheath margins emerge from the disc of insertion (Sharman, 1942) late in maize leaf
development, and that the elaboration of leaf margins is correlated with auxin transport and downregulation of
KNOX proteins. In addition, these results suggest that the regulation of KNOX accumulation is not upstream of
cultured apices results in the formation of tubular leaf bases, in which the margins fail to emerge and separate in
the lower leaf zone. The tubular leaf bases always formed in the fourth leaf from the meristem (P4). Invariably,
the unseparated margin regions of the tubular leaves accumulate high levels of ectopic KNOX proteins. These
data suggest that the lower, sheath margins emerge from the disc of insertion (Sharman, 1942) late in maize leaf
development, and that the elaboration of leaf margins is correlated with auxin transport and downregulation of
KNOX proteins. In addition, these results suggest that the regulation of KNOX accumulation is not upstream of
polar auxin transport; a more complex feedback relationship is implied. A model for L1-derived margin
development in maize leaves is presented.
106
RAB2 contributes to orderly cell division and expansion during leaf development
Sylvester, Anne (1) Zhang, Jiaming (1,2) Nielson, Amy (1)
(1) University of Wyoming, Department of Botany, Laramie WY (2) National Key Biotechnology Laboratory of
Tropical Crops, Chinese Academy of Tropical Agricultural Sciences, Hainan 571101, China
The maize leaf is composed of an orderly array of cells that grow by coordinated cell division and expansion. Cell
pattern mutations, however, cause general disorder when cells expand and divide abnormally, disrupting the
balance of cell size and shape. It is likely that similar cellular defects are due to mutations in many different
classes of genes. One of these mutations identifies a gene, warty1 (wty1, also rli1) that encodes RAB2, a small
GTP binding protein involved in vesicle transport in plants and yeast. We speculate that disrupted RAB2 alters
vesicle transport sufficiently to prevent cells from receiving signals to stop expanding. To examine the effect of
various disruptions in Rab2 on cellular phenotype, we isolate cell clusters from leaves and use a modification of in
situ PCR to identify somatic Mu insertions in Rab2. Leaves showing late somatic Mu activity are used. Clusters
composed of 10 to15 abnormal wart-like cells and adjacent normal cells are removed by microdissection,
amplified using Mu and gene-specific primers, and the products sequenced. A majority of isolated warts are due
to insertions of Mu into Rab2 based on statistical analysis of up to 100 wart and non-wart clusters. Furthermore,
Mu inserts throughout the gene, but at statistically higher frequency into intron 2 of Rab2. The remaining warts
not caused by Rab2 insertion could be due to insertions in candidate genes that encode proteins in the RAB
signaling pathway. Similar analyses could help to identify these myriad interacting genes as sequence data
become available from maize, rice and other organisms.
107
Adaxial/Abaxial polarity specification in Maize leaf development
Juarez, Michelle {1,2} Timmermans, Marja {1}
{1}Cold Spring Harbor Lab, Cold Spring Harbor, NY 11724 {2}Graduate Program in Genetics, State Univ. of
New York at Stony Brook, Stony Brook, NY11794
The first steps in lateral organ formation include the recruitment of founder cells from the meristem and the
establishment of new developmental axes relative to the main body axis. Characterization of the
leafbladeless1(lbl1) mutant phenotype suggested that lbl1 is required to establish adaxial cell identity in leaves
and leaf-like lateral organs. In the absence of LBL1 activity, cells obtain an abaxial identity that results in the
formation of radially symmetric, abaxialized leaves.

The semi-dominant Rolled1-O (Rld1-O) mutant of maize also affects adaxial/abaxial patterning in that the polarity
of the leaf is inverted. Double mutants between Rld1-O and lbl1 resulted in a mutual suppression of both
phenotypes. These results suggest that lbl1 and rld1 act in an opposing fashion on the same pathway, or that lbl1
and rld1 negatively regulate each other.

In order to further characterize the lbl1, Rld1 and lbl1;Rld1 double mutant phenotypes, we isolated the maize
homologs of the Arabidopsis YABBY genes which are expressed in the incipient primordium and in the abaxial
domain of developing lateral organs. The maize yabby genes (zyb) are also expressed in the incipient and young
leaf primordia, but interestingly expression is restricted to the adaxial side. In situ hybridization and RT-PCR data
both suggest that zyb is down-regulated in lbl1 mutants and up-regulated in Rld1 mutants. The zyb expression
pattern further suggests they function during lateral outgrowth. Our genetic and zyb expression data suggest lbl1
and Rld1 act in the same pathway and upstream of the zyb genes. We have preliminary data on the cloning of rld1.
Experiments with further analysis will be presented.
108
Vegetative Phase Change and Response to Puccinia sorghi in Sweet Corn
Basso, C. Federico {1} Hurkman, Meg {1} Tracy, William F. {1}
{1} University of Wisconsin-Madison
We performed three cycles of divergent full sib recurrent selection to develop early (adult earlier in development)
and late (adult later in development) phase change populations. Divergent selection for vegetative phase change
was successful with last leaf with juvenile wax exhibiting a strong linear response to selection. The average last
leaf with juvenile wax in C0 was 9.7. Average leaf number for C3L (late) was 11.7 while that for C3E (early) was
8.2. We evaluated the populations for resistance to common rust (Puccinia sorghi) by inoculating them at three
different developmental stages, v5, v10, and v15. Rust damage was evaluated on leaves seven through thirteen,
individually. The differences among cycles were greater in v10 than in v5 or v15. In the late direction, in v5 there
were significant linear trends over cycles for leaves 7, 8, 9, and 10; in v10 there were significant linear trends on
leaves 7, 8, and 9; and in v15 there were significant linear trends in leaves 9 and 10. In all these cases C3L had
more rust than either C0 or C3E. There were no linear trends in the early direction with C3E not different from
C0. Close examination, revealed that within leaves, rust pustules were more abundant on juvenile tissues, as
indicated by epicuticular wax. Fewer pustules per leaf and fewer leaves with pustules below the ear should reduce
the amount of inoculum for allo-infection and could reduce the total amount of damage.
109
Role of Pollen Allergens and Beta-Expansin Zea m 1 in Pollen Development and Fertilization
Valdivia,Elene R.{1} Durachko, Daniel M.{1} Wagner,Edward{1} Stephenson, Andrew G.{1} Cosgrove,
Daniel J.{1}
{1}Department of Biology, Penn State University
Group-1 grass pollen allergens are glycoproteins 28-35 kDa in size and are specifically expressed in grass pollen.
Group-1 pollen allergens make up a sub class of beta-expansins, which are cell wall loosening proteins that loosen
plant cell walls in characteristic ways. Group-1 pollen allergens are proposed to loosen the cell walls of the silks
to aid in the penetration and growth of the pollen tube through the silks. In maize, the group-1 pollen allergens are
called Zea m 1 and there are two divergent forms of Zea m 1. Each of the two forms is encoded by two genes,
making a total of four Zea m 1 genes in maize. To assess the function of Zea m 1, maize plants were transformed
with an RNAi construct for RNA silencing, driven by a pollen specific promoter. The first generation of
transgenic plants have the following unusual traits: low pollen production, shriveled anthers, and continued silk
growth post pollination. These phenotypes are more extensive than that found previously for a maize line carrying
an Mu insertion in one of the Zea m 1 genes. The first generation RNAi plants are being analyzed by northern blot
and immunoblot blot assays. Immunohistostaining, using monoclonal antibodies specific to group-1 pollen
allergens, is also being performed on male and female flowers at various developmental stages to examine the
localization of allergens during development.
110
Transgenic analyses of a duplicate pair of maize MADS-box genes, Zag1 and Zmm2, suggest protein
subfunctionalization of the C class in the traditional ABC model of flower development.
Whipple, Clinton {1} Ambrose, Barbara A. {1} Mena, MontaÒa {1} Schmidt, Robert J. {1}
{1} University of California San Diego
In the ABC model of flower development as defined by Arabidospsis, the C class is controlled by a single
MADS-box gene, AGAMOUS (AG). In ag mutants, stamens are converted to petals, and in the place of carpels a
new flower arises, such that ag flowers reiterate a patten of speal, petal, petal. In a search for maize genes similar
to AG, two genes were originally identified as potential AG orthologs: Zea AGamous1 (ZAG1) and Zea Mays
Mads2 (ZMM2). ZAG1 and ZMM2 have overlapping but non-identical expression patterns, such that ZAG1 is
more highly expressed in carpels, and ZMM2 more highly expressed in stamens. Maize plants homozygous for a
null zag1-mum1 allele show a phenotype consistent with a partial loss of C-function. However, attempts to isolate
ZMM2 loss of function have shown no phenotype. These results are consistent with the hypothesis that gene
duplication was followed by subfunctionalization via different expression domains. In order to test the possibility
that during evolution the ZAG1 and ZMM2 proteins themselves have subfunctionalized C class gene activity, we
have tested their ability specify stamen and carpel development in an Arabidopsis ag mutant. Our results indicate
that ZAG1 is better than ZMM2 at promoting carpel identity, while ZMM2 is better than ZAG1at promoting
stamen identity. This suggests that subfunctionalization has occurred at the protein level in addition to that seen in
the expression pattern. Recently we have cloned a duplicate locus of ZMM2, which we call ZMM2b. The
sequence and expression of ZMM2b suggest it also contributes to the maize C class activity, which could explain
previous difficulties in identifying a ZMM2 knockout. Work is underway to isolate a ZMM2/ZMM2b loss of
function, as well as to determine the region(s) of ZAG1 and ZMM2 that have evolved to result in differential
organ-identity function.
111
Functional analysis of the maize root specific gene ZmGrp3
Woll, Katrin (1) Hochholdinger, Frank (1)
(1) Center for Plant Molecular Biology (ZMBP), Eberhard-Karls-University, Tuebingen, Germany
Plant roots play an important role in water and nutrition uptake and anchoring the plant in the soil. In maize there
are four different types of roots: the embryonic primary- and seminal-roots and the postembryonic crown- and
brace-roots. In addition, lateral roots originate from all root types. So far, only little is known about the molecular
and genetic mechanisms underlying root formation in monocotyledonous plants. In an effort to isolate root-
specific genes, a differential screening of a coleoptilar node cDNA-library of maize was conducted in Dr. Feix lab
(University of Freiburg, Germany) which revealed a full-length cDNA clone designated ZmGrp3. ZmGrp3
represents a glycine rich, putative cell wall protein which is specifically expressed in roots. No homologues of this
gene are found in the completely sequenced rice and Arabidopsis-genomes. A functional characterization of the
ZmGrp3 gene expression and protein accumulation was now initiated. Gene expression was studied in RNA-gel-
blot and in situ -hybridization experiments. Northern-blot experiments with expression data from various
aboveground organs of the plants supported the notion that ZmGrp3 is exclusively expressed in roots. Further,
experiments including the root initiation-mutant lrt1, that fails to initiate lateral roots on the embryonic primary-
and seminal-roots (Hochholdinger et al., 1998, Plant J., 848- 855) and the mutant rtcs, that lacks the formation of
crown- and seminal- roots (Hetz et al., 1996, Plant J., 247-255) led to the hypothesis that the gene ZmGrp3 is
exclusively expressed in root-tips. In situ- hybridization experiments were performed to analyze the cell-type
specific expression of ZmGrp3 in root-tips. Sections of embryos, crown roots and lateral roots of different
developmental stages of the maize inbred-line B73 were hybridized using a DIG-labeled ZmGrp3-RNA probe.
Hybridization signals showed that ZmGrp3 is exclusively expressed in root-tips, confining the expression to the
epidermis and the columella. Further experiments on the protein-level including western-blot analyses and
immunolocalization experiments as well as screening for a knock out mutant with a reverse genetic gene machine
are in progress.
112
The Ac-tagged aberrant pollen transmission 1 (apt1) is a homologue of the Arabidopsis gene SABRE
required for root cell expansion
Xu, Zhennan {1} Dooner, Hugo {1}
{1} Waksman Institute, Rutgers University, Piscataway, NJ 08854
The Ac-linked apt1 mutation was detected by the aberrant pollen transmission of Ac. A cross between an apt1/+
heterozygous female and a wild-type male gives, as expected, a 1:1 ratio for Ac, but the reciprocal cross gives a
1:12 ratio for Ac. Thus, Ac transmission is normal in the female germline, but aberrant in the pollen.
Heterozygous apt1/+ plants segregate pollen with normal tubes and shorter tubes. The DNA adjacent to Ac was
isolated and shown by Southern blot analysis to cosegregate with the apt1 mutation. A revertant with normal
pollen transmission was recovered. It left an 8-bp excision footprint at the prior Ac insertion site. Southern blot
analysis of Ac-carrying progeny from the cross between apt1(Ac)/+ males and wild-type females shows that
apt1(Ac) is rarely, if ever, male-transmitted . All the bz-m spotted progeny from the cross carry a transposed Ac
(trAc). trAcs are recovered at a frequency of 8%. The APT1 protein is homologous to SABRE, a novel protein
involved in root cortex cell elongation in Arabidopsis . Northern blots show that apt1 is expressed in pollen. The
apt1 transcript is about 4 kb in size, unusually large for plants. We are currently working on isolation and
characterization of the apt1 gene.
113
NAC genes in maize: highly cell type specific expression patterns in embryogenesis
Zimmermann, Roman (1)
(1) University of Cologne
In monocot plants like Zea mays little is known about pattern formation processes organizing the developing
embryo. In contrast to the early stereotypic cell division pattern in Arabidopsis, cell division planes in the early
maize embryo are unpredictable. Developmental studies may be improved by molecular markers which monitor
cellular differentiation. An essential aspect of early maize embryogenesis is the specification of the root/shoot axis
in the transition stage. This decision results in a separation of the globular embryo proper into the scutellum, the
single maize cotyledon, and cells that will contribute to the future plant body. A similar developmental decision is
taken in the globular stage Arabidopsis embryo where the prospective shoot apical meristem (SAM) forms
between the two cotyledons. Arabidopsis seedlings which are mutant for CUC1 and CUC2 (CUP-SHAPED
COTYLEDON) lack the anlage of a SAM which and show a characteristic fusion of cotyledons to a cup-shaped
structure (Aida et al.,1997). Recently, a closely related gene, CUC3, was cloned and characterized (Casper
Vroemen, pers. communication). The embryonic expression patterns for CUC1, CUC2 and CUC3 suggest that all
three genes are redundantly required for meristem initiation and organ separation. The CUC-genes are members
of the plant-specific NAC gene family which is thought to encode transcription factors. Based on the high
sequence conservation of the NAC domain, we screened for CUC-homologous genes in maize to establish
developmental markers with crucial functions in SAM-initiation. Several candidate genes of the maize NAC gene
family were isolated. The expression pattern of a putative CUC3 homologue suggests that it may set a boundary
between the SAM and leaf primordia and between the SAM and the coleoptile. All isolated NAC members show
cell-type specific expression patterns during embryogenesis e.g. in the vascular system, the coleorhiza or the
developing endosperm.
sequence conservation of the NAC domain, we screened for CUC-homologous genes in maize to establish
developmental markers with crucial functions in SAM-initiation. Several candidate genes of the maize NAC gene
family were isolated. The expression pattern of a putative CUC3 homologue suggests that it may set a boundary
between the SAM and leaf primordia and between the SAM and the coleoptile. All isolated NAC members show
cell-type specific expression patterns during embryogenesis e.g. in the vascular system, the coleorhiza or the
developing endosperm.

Genome Structure/Synteny

114
MAPPING MAIZE MUTANTS WITH SSR MARKERS
Carson, Chris {1} Robertson, Jarrod {1} Mayginnes, John {1} Schlueter, Jessica {1} Bennett, Jennifer {1}
Lewis, Nora {1} Hampton, Ryan {1} Bradley, Lucas {1} Dauernheim, Alison {1} Anyanwu, Nkemdi {1}
Melia-Hancock, Susan {1} Neuffer, Gerry {1} Coe, Ed {2}
{1} University of Missouri, Columbia, MO, 65211 {2} USDA-ARS University of Missouri, Columbia, MO,
65211
Mutants that can be identified visually are a powerful source of biological information. They can be utilized for
both gene discovery and gene function studies. We have been working on a project designed to create and
evaluate a system to gain map information for a large set of mutants. We have analyzed over 476 mutants and
have mapped more than 384 using SSR markers. Further, we have produced F2 families for 596 placed kernel,
seedling and adult plant mutants from B-A translocation tested EMS stocks, and 234 seedling, viviparous, and
adult plant mutants from the Maize Genetics Cooperation-Stock Center. The mutants analyzed by this project
represent nearly every class and phenotype found in maize. F2 families are derived from crosses between mutant
stocks and the inbred parents A619, A632, B73, and Mo17. Predicted mutants appear in more than 80% of our F2
families. Some placed mutants (16%) were not located on the chromosome arm predicted. In turn, we are
characterizing many new mutants that we identified in some of these families. Our mapping strategy begins with
bulked-segregant analysis using SSR markers that span either the whole genome for unplaced mutants or specific
genomic segments for placed mutants. The phenotypes of the individuals in mutant families allow for the
identification of homozygous and segregating samples. These are pooled separately and subjected to PCR with
SSR primers. Polymorphic band patterns are compared between the two pools for reciprocal differences that
indicate linkage between a SSR marker allele and the homozygous allele of the mutant. Most BSA results were
tested by examining the individual homozygous samples with the putative linked SSR markers. New loci have
been identified, and many more have been accurately targeted for future experiments between tightly linked loci
having the same phenotype. NSF Plant Genome Grant DBI 9872655.
115
Duplication and Divergence of the R2R3 Myb Gene Family in the Grasses
Dias, Anusha P {1} Grotewold, Erich {1}
{1} The Ohio State University
R2R3 Myb genes are widely distributed in higher plants and comprise one of the largest known families of
regulatory proteins. The dramatic expansion of this family of transcription factors provides a striking example of
how gene amplifications followed by divergence may have impacted plant evolution. A group of recently
duplicated maize R2R3 Myb genes characterized by a change Pro 63 to Ala (MybPtoA) provides a unique
opportunity to understand the role of amplification and divergence of regulatory genes in the development of
novel cellular functions. MybPtoA comprises at least 10 members in maize, including the well characterized P gene,
which controls a subset of flavonoid biosynthetic genes, a regulatory function acquired recently in the evolution of
grasses. Members of the MybPtoA clade are also found in sorghum and rice, further supporting the ancient origin,
yet recent expansion of this group. To identify the patterns of evolutionary divergence after duplication of R2R3
Myb genes, we have cloned and characterized several additional members of the MybPtoA clade from maize and
other grasses. The patterns of sequence divergence of the Myb domains and the corresponding C-termini, the
evolution and size expansion of the introns and their possible contributions to the evolution of function of these
recently amplified R2R3 Myb genes in the grasses will be discussed.
116
Large-scale Maize cDNA mapping for candidate gene approach
Falque, Matthieu {1} DÈcousset, Laurent {2} Murigneux, Alain {2} Dautrevaux, Nolwen {1} Dervins,
Delphine {1} Jacob, Anne-Marie {1} RibiËre, Nicolas {2} Ridel, CÈline {1} Albini, Guillaume {1} Joets,
Johann {1} Charcosset, Alain {1}
{1} INRA UMR Genetique Vegetale du Moulon {2} BIOGEMMA
One of the ways to look for genes controlling traits of agronomical importance is to pre-select in silico EST
sequences with a putative interesting function, and to genetically map the corresponding cDNAs to look for co-
localization with QTLs. This poster presents the maize cDNA mapping project that is being carried out within the
Genoplante consortium. In this project, we use two Intermated Recombinant Inbred Lines mapping populations:
the Intermated B73xMo17 (IBM) population developed by Mike Lee (ISU) and available from Missouri
University, and the Intermated F2xF252 (LHRF) population developed by Alain Charcosset at INRA Le Moulon,
France. Both populations were developed following the same scheme. Genetic mapping of the cDNAs is
performed by RFLP, on a subset of 81 to 85 lines, whereas the framework maps were constructed with mostly
SSRs on 94 lines. The IBM framework map has 237 RFLP and SSR markers, including 84 Coremarkers. The
LHRF framework map has 250 RFLP and SSR markers, including 51 Coremarkers. 96 markers are common to
both framework genetic maps. Linkage analysis to place the cDNAs on the framework maps is automated by
using the ActionMap software (Albini, Falque & Joets, publication in prep.). 652 cDNAs, corresponding mostly
to genes with no publicly available mapping information, revealed 1160 loci with an average of 1.8 locus per
cDNA. Monomorphic genes on both populations represented 4% of the probes, and 6% of the genes could not be
genotyped. Among the cDNAs successfully mapped, 18% were monomorphic in the IBM population but could be
mapped with LHRF. This population provided also additional loci for some genes already mapped with IBM. The
LHRF population will be made available publicly (after 2003 summer multiplication, request to Alain Charcosset:
charcos@moulon.inra.fr). In addition to the framework map and the cDNAs, 1829 RFLP and SSR markers
(including 1316 raw mapping data obtained from MaizeDB) have been placed on our IBM framework map for
anchoring purposes with the QTL maps.
117
MAIZE MITOCHONDRIAL GENOMICS
Fauron, Christiane (1) Gibson, Mike (1) Clifton, Sandy (2) Minx, Patrick (2) Spieth, John (2) Newton,
Kathy (3) Rugen, Melissa (3)
(1) University of Utah, Eccles Institute of Genetics, 15 N. 2030 E., Salt Lake City, UT 84112 (2) Washington
University School of Medicine, Genome Sequencing Center, 4444 Forest Park Boulevard, St. Louis, MO 63108
(3) University of Missouri, Tucker Hall, Columbia, MO 65211
Plant mitochondrial genomes are unusual in their diversity of structure and rapidity of change. A comparison of
the complete sequences of mitochondrial DNA from two dicot species showed that, although most of the known
coding regions are very conserved, large intergenic regions show no sequence similarities. The regions between
genes have important roles in the rapid structural evolution seen for plant mitochondrial genomes. Moreover,
these 'non-coding' regions can become functional as components of chimeric genes that have been described as
being responsible for several cases of cytoplasmic male sterility (CMS). In order to understand how non-coding
regions of mitochondrial genomes contribute to their structural and functional diversity, it is necessary to examine
closely related species. Our NSF-funded Plant Genome Project is focused on the organization, evolution and
expression of the mitochondrial genomes of maize and its relatives, Currently, six mitochondrial genomes within
the genus Zea are being sequenced, including fertile (NB and NA genotypes) and male sterile (CMS-T and CMS-
C genotypes) maize, Zea parviglumis (the probable progenitor of maize) and Zea perennis. In addition, the
mtDNA sequences of related grasses, Tripsacum dactyloides and Sorghum bicolor, are being determined. The
sequencing of the Zea mays mays NB mitochondrial genome is now complete. The final assembly generated a
circular map of 569,630 base pairs with an average G+C content of 43.9%. The size compares well with the
previous restriction enzyme analyses (Fauron and Havlik, NAR 16:10395, 1988), which generated a map of
570kb. Annotation data and comparative analysis with other plant mitochondrial genomes will be presented.
118
Conserved Noncoding Sequences (CNSs) in Grasses
Choffnes-Inada, Dan (1) Bashir, Ali (1) Lee, Chunghau (1) Thomas, Brian C. (1) Ko, Cynthia (2) Goff,
Stephan A. (2) Freeling, Michael(1)
(1) University of California, Berkeley (2) Syngenta Biotechnology
Space surrounding and between exons carry discreet sequences that bind proteins or RNAs that wire any gene into
a regulated network of expression. One way to identify such cis-acting sites is as conserved noncoding sequences
(CNSs), these being phylogenetic footprints so large and exact that just two orthologous gene sequences are
required. CNSs have been found in higher plant genes (Kaplinsky et al., 2002. PNAS 99: 6147) as well as
mammalian genes, and presumably become discernable when orthologs are diverged for enough time to
randomize functionless sequence (approximately 50 MY), but not for so long as to realize mutations in all but
absolutely essential nucleotides. As with mice and man, maize and rice have diverged for a useful amount of time.
We add data on 52 new maize-rice orthologs where exon identities are supported experimentally. The average
grass gene has about three CNSs (greater than or equal to a 15/15 exact bp match within the gene space), with a
mode of 0 at 28%. The mean CNS size is about 25 bp, these being located about evenly in 5í, 3íand intron space. -
-CNSs make excellent pan-grass PCR primer sites. --Grasses have about 10X fewer and much shorter CNSs as
compared to mammals. --There is a poor correlation between CNS and known transcription factor binding sites. --
Some grass genes are CNS-rich, from 7-34. These tend to be upstream regulatory genes. --One such CNS-rich
gene is homeobox gene knotted1. Here, an intron region shown by Greene and coworkers (1994. Genetics 138:
grass gene has about three CNSs (greater than or equal to a 15/15 exact bp match within the gene space), with a
mode of 0 at 28%. The mean CNS size is about 25 bp, these being located about evenly in 5í, 3íand intron space. -
-CNSs make excellent pan-grass PCR primer sites. --Grasses have about 10X fewer and much shorter CNSs as
compared to mammals. --There is a poor correlation between CNS and known transcription factor binding sites. --
Some grass genes are CNS-rich, from 7-34. These tend to be upstream regulatory genes. --One such CNS-rich
gene is homeobox gene knotted1. Here, an intron region shown by Greene and coworkers (1994. Genetics 138:
1275) to bind factors turning KN off in leaves is extraordinarily CNS-rich. By comparing higher plants and
animals for gene space structure, morphological and regulatory complexity and the illusive ìplasticity,î useful
inferences will be made as to what CNSs bind and to what end.
119
Sequence diversity at the y1 locus of maize.
Hahn, Elizabeth A. {1} Krueger, Katherine C. {1} Massman, Kerry L. {1} Schnable, Patrick S. {2}
Buckner, Brent {1}
{1} Truman State University {2} Iowa State University
Previous studies of various alleles of the y1 gene of maize have demonstrated that single nucleotide
polymorphisms (SNP), insertion/deletion polymorphisms (InDel) and microsatellite length variability can be used
to distinguish alleles. In order to further characterize the degree of sequence diversity at this locus, additional
regions of the gene, including exons and introns, were cloned and sequenced from seventeen inbred lines, six
open-pollinated landraces and four accessions of teosinte. The inbred lines investigated in this study were selected
because they were used to establish the agronomically important BSSS and BSCB1 breeding populations.
Multisequence alignments and phylogenetic trees of these y1 alleles established that 1 SNP occurred every 101 bp
of exonic sequence and every 44 bp of intronic sequence. No InDels were found in exonic sequences while, on
average, there was 1 InDel per 220 bp of intronic sequences. The inbred lines exhibit relatively little sequence
diversity as compared to the open-pollinated landraces and teosinte. The teosinte alleles analyzed in this study had
a degree of greater sequence similarity to the recessive than the dominant maize alleles.
120
INTERGRATING A MAIZE PHYSICAL AND GENETIC MAP USING PCR BASED MARKERS
CONVERTED FROM SINGLE COPY RFLP MARKERS ON BAC POOL DNAs
January,Eboni {1} Yim,Young-Sun {1} Moak,Patricia {1} Robertson,Jarrod {1} Schroeder,Steve {1}
Sanchez-Villeda,Hector{1} Musket,Theresa {1} Soderlund,Carol {2} Coe Jr.,Edward {1,3} Davis,Georgia
{1}
{1} Department of Agronomy, University of Missouri-Columbia, 1-87 Agriculture, Columbia, MO 65211 {2}
Plant Science Department, University of Arizona, Tucson, AZ 85721 {3} USDA-ARS, PGRU, 210 Curtis Hall,
Columbia, MO 65211
The long-term goal of this project is to provide an anchored physical and genetic map of the maize genome. In
order to meet this goal, PCR-based physical mapping of genetically mapped single copy RFLPs was performed on
6-dimensional pools from the HindIII BAC library. Single copy genes were chosen for anchoring because of the
high amount of repetitive DNA and large size of the maize genome. The BAC pooling strategy is efficient and
cost effective. Exactly 110,592 maize BAC clones were pooled into 6 different dimensions to create 288 pools of
BAC DNA, 48 pools per dimension. Individual BAC clones containing the gene of interest can be determined
from actual and predicted positives. The quality of the BAC DNA pools and their role in identifying BACs
containing certain genomic sequences were initially tested using SSR primers. Single copy RFLP markers were
converted into PCR-based markers to provide more anchoring points quickly. Primers were first tested for
amplification on B73 genomic DNA, then amplified on B73 pool DNAs. Among the 176 primers designed for
PstI RFLP clones, 32 primers (18%) did not amplify bands on genomic DNA. The remaining primers were
analyzed on BAC pool DNAs. Amplification products from the pools were deconvoluted into individual BAC
address using a program known as Resolve Script. The marker data will be integrated with the FPC fingerprint
assembly generated by AGI to anchor the BAC contigs and contribute to the manual editing process.
121
Hormonal, Light and Organospecific Regulation of Ribosomal Protein S14 Gene Expression
Kurylo, Vasyl {1}
{1} Khmelnytskyy Scientific Research Institute of Crop Selection / Ukraine
Phytohormones, cytokinin and abscisic acid (AA) are most important endogenic regulators of protein synthesis in
plants. It is shown participation in regulation of individual matrices accumulation, rRNA synthesis, ribosomes
new formation, polysomes forming and degradation and also ribosomal protein phosphorylation. However, the
effect of phytohormones on ribosomal protein gene expression, the most important components of translation
system, is not studied yet. The purpose of this work consisted in analysis of cytokinin and abscisic acid, white and
red light effect on accumulation of ribosomal protein S14 and its matrix in isolated cotyledons of two maize sorts
(Zea mays Imdurata) early ripe ñ Voronezhskaya-76 and late ripe Odesskaya-10. Based on the differential
screening it was isolated full-size cDNA, corresponding to transcript of the gene rps14, which expression is being
regulated by cytokinin (benzylaminopurine), abscisic acid and light in isolated cotyledons of maize. To examine
hormone and light regulation of the gene rps14 expression, maize seeds have been germinated in the dark in
system, is not studied yet. The purpose of this work consisted in analysis of cytokinin and abscisic acid, white and
red light effect on accumulation of ribosomal protein S14 and its matrix in isolated cotyledons of two maize sorts
(Zea mays Imdurata) early ripe ñ Voronezhskaya-76 and late ripe Odesskaya-10. Based on the differential
screening it was isolated full-size cDNA, corresponding to transcript of the gene rps14, which expression is being
regulated by cytokinin (benzylaminopurine), abscisic acid and light in isolated cotyledons of maize. To examine
hormone and light regulation of the gene rps14 expression, maize seeds have been germinated in the dark in
humid filter paper for three days. Then cotyledons have been separated from the sprouts and have been incubated
for 24 hours on the water to exhaust endogenic phytohormones content in it. Incubation has been continued in
benzylaminopurine solution (2,2 x 10-5 M), abscisic acid (7,6 ı 10-5 Ã) and in the water in the dark and in the
light for 1 ñ 120 hours. Analysis of the protein S14 distribution in various organs of 8-days sprouts, grown in the
light, has shown that organs with mitotic active tissue prevalence (hypocotyl and new leaves) are characterised by
the most protein S14 content. The minimum protein S14 content is available in exhausting cotyledons and
finished growth root sections. Level comparison of mRNA and protein S14 displays important role of post-
transcriptional regulation of this gene expression in various maize organs. S14 level reflects ribosomesí content,
which is much more in cells of actively growing tissues. Therefore for the first time it is shown participation of
cytokinin and abscisic acid on mRNA and protein level in regulation of expression of rps14 gene, coding one of
ribosomal proteins. It is suggested that many ribosomal proteins genes are under co-ordinated hormone control,
however it is not to except differential regulation of these genes expression. The full-size cDNA isolation and also
obtaining of antibody to protein S14 significantly extend opportunities to study mechanisms of hormone
regulation of transcription and translation processes in plants.
122
An integrated physical map in maize
Fengler, Kevin (1) Faller, Marianna (1) Dam, Thao (1) Tingey, Scott (1) Morgante, Michele (1) Li, Bailin
(1)
(1)DuPont Crop Genetics P.O.Box 6104 Newark, DE 19714-6104
An integrated genetic and physic map in maize serves as the foundation for positional cloning of genetically
mapped genes. A whole-genome physical map was constructed by fluorescent fingerprinting of a corn BAC
library. Manual curation has greatly improved the quality of the map. More than 140,000 BAC end sequences
were obtained from the BAC clones. These sequences provide a rich source for genetic markers and automatically
place several thousands genes on the map. Moreover, about 17,000 ESTs and genetic markers have been placed
on the BAC contigs through overgo hybridization. A map-based cloning project will be given as an example to
demonstrate the utility of the physical map.
123
Genomic regions comprising the entire alpha zein gene family of Zea mays in a single inbred line
Song, Rentao {1} Messing, Joachim {1}
{1} Waksman Institute, Rutgers, The State University of New Jersey
Zeins are the major storage proteins in maize seeds. Based on sequence homology, the alpha zein gene family can
be divided into four subfamilies: z1A, z1B, z1C and z1D. We have isolated BAC clones from each subfamily by
specific probes and sequenced the genomic regions comprising all copies of alpha zein genes of maize inbred line
B73. This yielded a total of more than 1.4 Mb genomic sequences, representing the most extensive genomic
sequence analysis in maize to date. A total of 41 alpha zein genes were discovered from this analysis, with 12
genes of the z1A subfamily, 8 genes of the z1B subfamily, 16 genes of the z1C subfamily, and 5 genes of the z1D
subfamily, respectively. Except for the z1D subfamily, which only occupies one genomic location, all other
subfamilies occupy two genomic locations. Distance analysis using alpha zein gene coding regions indicated that
alpha zein genes arose from a single ancestor gene around 35 mya. Gene duplication, followed by translocation
and further amplification formed todayís genomic organization of this gene family. Less than half of the alpha
zein genes still remain active in the B73 inbred line, and their expression level also varies greatly. Having now the
positional information of all alpha zein genes, we can study in parallel the expression of each gene copy, the
different haplotypes of this gene family in maize, and the combinatorial aspects of different haplotypes in gene
expression. Such combinatorial studies could provide important insights into phenomena like heterosis.
124
Haplotypes of a genomic region containing z1C gene family in maize
Song, Rentao {1} Messing, Joachim {1}
{1}Waksman Institute, Rutgers, The State University of New Jersey
Genomic regions of nearly every species diverged into different haplotypes, mostly based on point mutations,
small deletions and insertions that do not affect the collinearity of genes within a species. However, when an
orthologous genomic regions of two inbred lines of Zea mays were compared, not only genes were lost from the
entire genome but also the regulation of genes were different. As a consequence, the expression potential of
inbreds and hybrids differed significantly. The orthologous regions also differed in length, 360 kb versus 263 kb,
and a survey of 43 common maize inbred lines illustrated similar structural differences. However, the same
orthologous region even between two subspecies rather than cultivars of rice was conserved in length and gene
order. Changes between two maize inbreds were as drastic as inter-species changes of maize and sorghum. Intra-
orthologous genomic regions of two inbred lines of Zea mays were compared, not only genes were lost from the
entire genome but also the regulation of genes were different. As a consequence, the expression potential of
inbreds and hybrids differed significantly. The orthologous regions also differed in length, 360 kb versus 263 kb,
and a survey of 43 common maize inbred lines illustrated similar structural differences. However, the same
orthologous region even between two subspecies rather than cultivars of rice was conserved in length and gene
order. Changes between two maize inbreds were as drastic as inter-species changes of maize and sorghum. Intra-
species changes were also enhanced by the insertion of different transposable elements. Changes in gene order
affect not only the cluster of z1C storage protein genes but also three unrelated genes. Analysis of the expression
of individual members of the z1C genes exhibited unexpected shifts of gene expression patterns in hybrid crosses,
exemplifying allelic and non-allelic interactions of the two haplotypes. Given that chromosomes could
conceivably consist of intervals of ìbinaryî haplotypes that are highly diverged, one could envision endless
breeding opportunities because of their linear arrangement along a chromosome and their expression potential in
hybrid combinations.
125
Validation of in silico connection between maize physical map and genetic map by PCR-based screening on
BAC pools.
Yim, Young-Sun Yim {1} Moak, Patricia {1} Robertson, Jarrod {1} Sanchez-Villeda, Hector {1} Musket,
Theresa {1} Soderlund, Carol {2} Coe Jr., Edward {1, 3} Davis, Georgia {1}
{1} Department of Agronomy, University of Missouri-Columbia, 1-87 Agriculture, Columbia, MO 65211 {2}
Plant Science Department, University of Arizona, Tucson, AZ 85721, www.genome.arizona.edu {3} USDA-
ARS, PGRU, 210 Curtis Hall, Columbia, MO65211
The construction of an integrated genetic and physical map of the maize genome (2500 Mbp) is the primary goal
of our ongoing maize genome project. To accomplish this goal, several strategies were used simultaneously. An
overgo hybridization approach provided a large number of anchoring points to the maize BAC contig map. Our
objective was to validate the in silico connections between the maize physical and genetic maps using a BAC
pooling strategy combined with a high-throughput PCR-based screening method. The BAC DNAs were pooled
from 6X haploid genome equivalents (110,592 maize BAC clones) of a maize HindIII BAC library in six different
dimensions, producing 288 pools of BAC DNA (48 pools per dimension). A total of 95 PCR-based markers
associated with overgo probe by in silico methods were screened on the pools. Amplified PCR products from the
pools were deconvoluted into individual BAC addresses using Resolve script. On average, PCR screening on
BAC pool and overgo hybridization identified 4.72 and 5.26 BAC clones per each probe analyzed. BAC: probe
data shows 88% (eighty-four out of ninety-five probes) containing BAC clones identified by both approaches,
strongly supporting the in silico associations. This data was integrated with the fingerprinting data generated by
Arizona Genome Institute (AGI) to assemble the BAC contigs using FPC and contribute to the process anchoring
physical map to the genetic map.

Molecular Genetics

126
Keeping the babies healthy. Defence mechanisms operate at the base of the endosperm in developing seeds.
BalandÌn, Teresa {1} MuÒiz, Luis Miguel {2} Lopez, Isabel {3} Sanz, Yolanda {4} Hueros, Gregorio {5}
Department of Cell Biology and Genetics. University of Alcala
Endosperm transfer cells, located at the basal part of the seed, are designed to facilitate nutrient import into the
growing maize kernel. Specific genetic systems devised to improve transport capacity are therefore thought to be
operative in this tissue. Nevertheless, in a series of screening experiments conducted to identify transfer cell
specific genes, we could not find any gene putatively involved in the transport of metabolites. Many genes have
been identified, however, apparently involved in defence mechanisms and in one case, the BAP family of
proteins, antimicrobial activity has been previously demonstrated in vitro. We report here on the characterisation
of another antimicrobial activity, in this case isolated from a different endosperm tissue also located at the base of
the seed, the embryo surrounding region. We propose that, in addition to their role in nutrient transport, cells at
the base of the endosperm are actively involved in the formation of a barrier to impede the entry of pathogens into
the seed.
127
Expression and inheritance patterns of a modified porcine a-lactalbumin transgene in maize kernels
Bicar, Earl (1) Scott, Paul (1)(2) Lee, Michael (1)
(1) Iowa State University, Department of Agronomy (2) USDA-ARS Collaborator
A codon-adjusted version of a gene encoding the porcine milk protein a-lactalbumin was synthesized to create
maize with improved nutritional quality. Three expression vectors containing the synthetic gene were constructed
and transformed into maize callus by particle bombardment. The constructs (P45, P64, P57) differ in their
promoters and subcellular targeting domains. The objective of this study was to characterize the expression and
inheritance patterns of these transgenes. a-La DNA sequences were detected by PCR analysis of F1 plants in 6 of
8 events of P64, 6 events of P45 and 4 of 5 events of P57, indicating that the a-La gene was integrated into the
maize genome and transmitted to the F1 generation. Western blot analysis indicated that the a-La protein
accumulated in the endosperm, showing that the transgenes were functional. Expression and inheritance of the a-
La transgenes were assessed. In 5 of 6 events of P64, the a-La transgene was inherited as a single locus through
the F3 generation. In P45, irregular expression and inheritance patterns were observed. In 3 events, the a-La
transgene was inherited as a single locus in the BC1F1 and F2 generations. In the F3 generation, segregation
distortion manifested by an increase in frequency of kernels with no detectable a-La protein was observed. PCR
analysis of these kernels confirmed presence of the a-La transgene, indicating that the transgene was not
functional. In 2 events that expressed a-La in the F1, the a-La protein was not detected in the BC1F1 and F2
generations. However, in F3 kernels derived from self-pollination of these F2 null plants, a-La was detected,
suggesting that the transgene was reactivated. This phenomenon was also observed in the F3 generation in 2
events of P57 that manifested transgene inactivation in the BC1F1 and F2 generations. Possible causes of these
irregular inheritance patterns are discussed.
128
Class III Aldehyde Dehydrogenase Genes of Maize
Cao, Jun {1} Simmons, Carl {2} Meeley, Robert B {2} Schnable, Patrick S {1}
{1} Iowa State University, Ames, Iowa 50011, USA {2} Pioneer Hi-Bred Intl, Johnston, Iowa 50131, USA
The maize rf2a gene, which is a restorer of T-cytoplasm male sterility, encodes a Class II aldehyde dehydrogenase
(ALDH). The maize genome contains another gene that encodes a Class II ALDH (rf2b) and two that encode
Class I ALDHs (rf2c and rf2d). Four maize genes that encode Class III ALDHs have been identified (zmfaldh1,
zmfaldh2. zmfaldh3 and zmfaldh4). The deduced amino acid sequences of the corresponding proteins exhibit
~60% similarity to human fatty ALDH (ALDH10), the deficiency of which causes the genetic disorder Sjogren-
Larsson syndrome. These proteins are also >70% similar to Class III ALDHs from Craterostigma plantagineum,
which are thought to be involved in ABA-dependent responses to dehydration stress. Each of the maize Class III
ALDHs contains a putative ER targeting sequence at its C-terminal. Three of the four Class III ALDHs genes
have been genetically mapped. Mutant alleles have been isolated for each gene. Mu transposon insertion alleles of
zmfaldh1, zmfaldh2 and zmfaldh4 were obtained using the TUSC system. An EMS-induced mutant of zmfaldh3,
identified via TILLING, contains a single nucleotide transition that causes a G to E change in the protein.
129
Use of DNA pooling to assess diversity within and among maize populations. Application to the
investigation of maize introduction into Europe.
Dubreuil, Pierre {1} Rebourg, CÈcile {1} Warburton, Marilyn {2} Chastanet, Monique {3} Gouesnard,
Brigitte {4} Hoisington, David {2} Charcosset, Alain {1}
{1} INRA, CNRS, UPS, INAPG, Station de GÈnÈtique vÈgÈtale du Moulon, 91190 Gif sur Yvette {2}
CIMMYT, Apdo. Postal 6-641, 06600 Mexico, D.F., Mexico {3} CNRS, UMR 8054, Centre de Recherches
Historiques et Juridiques ñ UniversitÈ Paris I, 9 rue Malher, 75181 Paris cedex 04, France. {4} INRA, domaine
de Melgueil, 34130 Mauguio, Franc
Large collections of traditional maize open pollinated varieties (OPVs) are maintained by the breeding programs
and genetic resource centers of many institutions (e.g., CIMMYT, USDA, INRA). It is widely acknowledged that
these collections contain as yet unexploited genetic diversity with a high potential value for genetic investigations
and breeding. Their practical use in research programs is presently limited by a low level of information regarding
the structure of genetic diversity, such as the level of redundancy among accessions, presence of groups of related
accessions, and presence of unique diversity in a few, key populations. The use of neutral molecular markers has
the potential to provide interesting new information regarding these topics. However, high levels of within
population genetic diversity typical in maize OPVs call for the analysis of a representative sample of individuals
for each accession, which makes analyses difficult and time-consuming. We have tested a method first using
RFLP markers and then SSRs, which consists in analyzing bulks of leaves from different individuals, and
estimating allelic frequencies using signal intensities of alleles present within a pool. The use of controlled pools
of inbred lines shows that this approach is accurate for RFLPs (overall correlation between estimated and
expected frequencies of 0.93). The same approach can be used for SSRs provided (i) loci with low amplification
competition between alleles are selected and (ii) frequency estimation takes into account possible ìstutteringî
phenomena (presence of additional fragments of predictable sizes due to errors of the Taq polymerase). Overall
correlation between estimated and expected frequencies was 0.96 in this cases. Data were obtained for 29 RFLP
loci and 24 SSR loci on a set of 174 representative American OPVs (from South, Central, and North America) and
a set of 131 representative European OPVs. Results obtained on American material show strong structure patterns
associated with geographical origin, consistent with recent results of Matsuoka et al. (2002, PNAS 99:6080-6084).
Comparison of European and American diversity (i) confirms the contribution of Caribbean material introduced
by Colombus in Spain to the establishment of European diversity and (ii) shows a major contribution of Northern
Flint populations cultivated by Native Americans in the North-Eastern United States. A detailed historical
loci and 24 SSR loci on a set of 174 representative American OPVs (from South, Central, and North America) and
a set of 131 representative European OPVs. Results obtained on American material show strong structure patterns
associated with geographical origin, consistent with recent results of Matsuoka et al. (2002, PNAS 99:6080-6084).
Comparison of European and American diversity (i) confirms the contribution of Caribbean material introduced
by Colombus in Spain to the establishment of European diversity and (ii) shows a major contribution of Northern
Flint populations cultivated by Native Americans in the North-Eastern United States. A detailed historical
investigation shows that this introduction occurred very shortly after new world discovery (before 1539). This
research program was supported by PROMAIS members: ASGROW France (Monsanto SAS), CARGILL
SEMENCES (Monsanto SAS), CAUSSADE SEMENCES, CEBECO SEMENCES, EURALIS GENETIQUE,
MAœSADOUR SEMENCES, NOVARTIS SEEDS (Syngenta), LIMAGRAIN GENETICS, PIONEER
GENETIQUE, S.D.M.E. K.W.S. France, VERNEUIL RECHERCHE
130
Certain Mutations Preventing Paramutation Heritably Reactivate a Transcriptionally Silent Transgene
Clay, Catherine {1} Lin, Yan {1} Carey, Charles C {1} Chandler, Vicki L {1}
{1}University of Arizona, Tucson, AZ 85721, USA
Paramutation is a natural gene silencing phenomenon involving an interaction between alleles that leads to a
heritable decrease in the expression of one of the alleles. Previous studies isolated several trans-acting mutants
affecting paramutation in maize. To address whether the paramutation mutants would affect transgene silencing,
the mutants were crossed with a silent transgenic line to obtain transgenic stocks segregating each paramutation
mutant. Our initial results demonstrate that mutations in 5 genes that prevent paramutation can activate a
previously silent transgene. The silent transgene contains a 35S promoter driving the transcribed region of B-I,
which when active results in dark purple plants. The b1 locus encodes a transcription factor that activates the
anthocyanin pathway. Sibling plants with the transgene, and heterozygous or homozygous for each of the
recessive mutations, were compared. The B-I genomic transgene was activated in homozygous paramutation
mutant plants (darkly purple), while it remained silent in heterozygous plants (green). Northern blots and nuclear
run-on assays demonstrate that the transgene is transcriptionally silenced and that the 'absence' of the wild type
gene products in each of the recessive mutants relieves this silencing. In the paramutation mutants, another
transcriptionally silenced transgene (35S Bar) within the same transgene array is also activated. We have been
investigating whether or not the reactivated transgene remains active when the paramutation mutants are
segregated away. We have identified two classes of mutants. With the first class, represented by the rmr1-1
mutation, the transgene is efficiently resilenced in plants heterozygous for the wild type allele and the mutant. In
contrast, when segregated away from the other mutations, the transgene can remain active for a number of
generations in the absence of the mutation responsible for its initial reactivation.
131
Comparative mapping of drought responsive genes in Zea mays L. and Arabidopsis thaliana
Cordia, John {1} Musket, Theresa {1} Yim, Young-Sun {1} Davis, Georgia {1}
{1} University of Missouri-Columbia
Drought is a major limitation to productivity of Zea mays world wide. The search for drought responsive genes in
maize is therefore a very important and valuable undertaking. Studies have shown that abscisic acid (ABA)
concentration increases in response to drought stress. ABA application can also affect the expression of drought
responsive genes suggesting that ABA acts as a signal for water deficit response (Leung and Giraudat, 1998). In
Arabidopsis thaliana, numerous genes that respond to dehydration stress have been identified. These genes have
been categorized as rd (responsive to dehydration) and erd (early response to drought). Several of these genes
have been found to be induced in the ABA pathways. The fact that maize and Arabidopsis have many common
genes leads to the possibility that the erd and rd genes in Arabidopsis could be found in maize. The objectives of
this experiment are to determine if the rd and erd gene sequences in Arabidopsis can be found in maize, to map
these genes in maize and to sequence the maize PCR products to determine if they are the selected rd or erd gene
from Arabidopsis. Six genes from Arabidopsis were selected and primers were designed from their sequence.
DNA from maize inbreds, Mo17 and B73 was amplified to determine if the corresponding sequences were present
in the maize genome. Primer products were obtained for all six genes. Five of the six primer pairs had mappable
polymorphisms. Further research is needed to determine the function of these genes in maize.
132
Screening phenotypic diversity in light-responsiveness to identify potential phytochrome mutants in maize
Costich, Denise E. {1} Sheehan, Moira J. {1} Brutnell, Thomas P. {1}
{1} Boyce Thompson Institute for Plant Research
Mesocotyl length is a phenotypically plastic trait in different light environments and provides a quantitative
measure of light responsiveness during early seedling development. Mesocotyl elongation is sensitive to several
wavelengths of light, implicative of red/blue photoreceptor systems. Low light levels in growth chamber screens
have revealed significant underlying genetic variation in this trait among 100 maize inbreds. We compared mean
mesocotyl length in non-stiff and stiff stalk lines originating from the North American Corn Belt with those from
semi-tropical lines and found that semitropical lines had mesocotyls that were on average 41-49 % shorter than
the corn belt inbreds. This suggests that a significant loss of light responsiveness at the seedling emergence stage
measure of light responsiveness during early seedling development. Mesocotyl elongation is sensitive to several
wavelengths of light, implicative of red/blue photoreceptor systems. Low light levels in growth chamber screens
have revealed significant underlying genetic variation in this trait among 100 maize inbreds. We compared mean
mesocotyl length in non-stiff and stiff stalk lines originating from the North American Corn Belt with those from
semi-tropical lines and found that semitropical lines had mesocotyls that were on average 41-49 % shorter than
the corn belt inbreds. This suggests that a significant loss of light responsiveness at the seedling emergence stage
accompanied selection by breeders in northern temperate regions. Has diversity in phytochrome gene sequence or
gene expression played a role in this process? To begin to answer this question, we have designed a molecular-
genetic screen to identify large-scale changes in the highly conserved region of the duplicated phytochrome B
genes, spanning Exons I and II. In our assay, no allelic diversity in phyB1 has been detected, but seven of the 88
inbreds screened appear to contain an approximately 6 Kb insertion in phyB2. A subset of these mutants will be
targeted for sequence and expression analysis, as well as phenotypic characterization of other phytochrome-
mediated traits.
133
Is the maize C2-Idf mutation controlled by an RNA-based silencing mechanism?
Della Vedova, Chris {1} Cone, Karen {1} Birchler, Jim {1}
{1} University of Missouri
The maize Inhibitor-diffuse (C2-Idf) mutation is a dominant loss-of-function allele of c2. The c2 gene product is
chalcone synthase, the enzyme catalyzing the first dedicated step in the anthocyanin and phlobaphene pigment
pathways. Without a functional c2 product, no pigments are made. Plants that are heterozygous for wild-type
functional C2 and a recessive non-functional c2 allele make half the normal amount of pigment. By contrast,
plants that are heterozygous for wild-type C2 and C2-Idf make virtually no pigment. We are interested in
understanding how C2-Idf inhibits the function of a normal C2 allele. Experimental observations suggest that C2-
Idf inhibition involves regulation at the epigenetic level, perhaps through a RNA-based silencing mechanism. C2-
Idf/C2-Idf homozygotes accumulate very little c2 mRNA. In C2-Idf/C2 heterozygotes, c2 mRNA is reduced to
roughly 10% of wild-type. This implies both that there is a negative interaction between the inhibitor and wild-
type alleles and that the loss of function associated with C2-Idf inhibition is controlled at the level of RNA
accumulation. The promoter of the C2-Idf allele is hypermethylated relative to the wild type, further suggesting an
epigenetic effect on regulation of this allele. As a means to test these ideas, we infected C2-Idf plants with maize
dwarf mosaic virus, a potyvirus that carries an inhibitor of the RNA-based silencing mechanism. In contrast to the
unpigmented phenotype of uninfected plants, infected plants were highly pigmented. The increased pigmentation
was accompanied by increased c2 mRNA accumulation. These results suggest an RNA-based silencing
mechanism as the means for C2-Idf inhibition, although it remains to be determined whether changes in RNA
accumulation are modulated post-transcriptionally by RNA degradation or transcriptionally by chromatin-level
regulation.
134
Characterization of a Peptide Transport Gene from Oryza sativa
Donahue, Kerry {1} Lubkowitz, Mark {1}
{1} Saint Michael's College 1 Winooski Park Colchester, VT 05439 USA
Peptide transport is a common phenomenon in both prokaryotic and eukaryotic organisms. The obvious role of
peptide transport is to translocate amino acids across cellular membranes for protein synthesis and for nitrogen
acquisition. However, the abundance of peptide transport genes, particularly in plants, suggests that transport
proteins may play other important roles. We have begun characterizing the PTR family (Peptide TRansporter) of
peptide transport genes in both maize and rice. The recent discovery of 51 putative PTR transporters in
Arabidopsis suggests this family may be playing a significant biological role in plants (Stacey et al., 2002). A
peptide transport gene (OsPTR1) from Oryza sativa was identified through homology searches with an existing
peptide transport gene from Arabidopsis. We are currently in the process of cloning the OsPTR1 open reading
frame into a yeast expression vector for heterologous expression in a Saccharomyces PTR mutant. In a second
experiment, we are investigating a possible connection between plant hormones and peptide transport by
determining if OsPTR1 expression is affected by hormone levels in germinating embryos.
135
One gene family that may influence maize endosperm betacarotene content.
Gallagher, Cynthia E. {1} Licciardello, Nicholas {1} Murillo, Christina I. {1} Wurtzel, Eleanore T. {1, 2}
{1} Lehman College, The City University of New York, Bronx, NY 10468 {2} The Graduate Center of the City
University of New York, New York, NY 10016
Carotenoids, synthesized by plants and other organisms, include over 600 different compounds, many of which
are essential in human health. Endosperms of major food crops, including maize, are low in carotenoid content
and therefore potential targets for improvement via marker-assisted selection or transgenic metabolic engineering
approaches. To provide the foundation needed to develop rational engineering strategies, we are investigating
regulation of the endosperm biosynthetic pathway through identification of the genes and gene families that
encode the pathway enzymes. We wish to determine the contribution of each family member to tissue specificity
of carotenoid accumulation and more specifically to assembly of pathway enzymes with regard to plastid
are essential in human health. Endosperms of major food crops, including maize, are low in carotenoid content
and therefore potential targets for improvement via marker-assisted selection or transgenic metabolic engineering
approaches. To provide the foundation needed to develop rational engineering strategies, we are investigating
regulation of the endosperm biosynthetic pathway through identification of the genes and gene families that
encode the pathway enzymes. We wish to determine the contribution of each family member to tissue specificity
of carotenoid accumulation and more specifically to assembly of pathway enzymes with regard to plastid
membrane localization. Here we focused on betacarotene hydroxylase (HYD), the enzyme that converts
provitamin A carotenoids (betacarotene) to nonprovitamin A xanthophylls (zeaxanthin). A maize cDNA,
demonstrated to encode a HYD enzyme, was used as a probe to screen a maize B73 BAC genomic DNA library.
Characterization of the isolated clones revealed three gene groups. On the basis of sequence analysis, one family
member (HYD3) appeared to be complete, while the other two (HYD1 and HYD2), though homologous, appeared
to share deletions that would likely interfere with expression of a functional enzyme. From database searching, it
appeared that other varieties had normal HYD1/2 sequences suggesting that allelic variations may be responsible
for variations in endosperm carotene content. This research is funded in part by the National Institutes of Health.
136
THE LETHAL LEAF-SPOT 1 (LLS1) PROTEIN IS LOCALIZED TO CHLOROPLAST MEMBRANES
AND IS HIGHLY CONSERVED IN OXYGENIC PHOTOSYNTHETIC ORGANISMS.
Yang, Manli (1) Wardzala, Ellen (1), Haller, Steve (1) Greenberg, Jean (2) Johal, Gurmukh S. (3)
Reinbothe, Steffen (4), Gray, John (1)
(1) Dept. of Biology, University of Toledo, OH 43606. (2) Mol. Genetics and Cell Biology Dept, University of
Chicago, IL 60637. (3) Plant Pathology Dept., Purdue University, IN 47906. (4) Laboratoire de Genetique
Moleculaire des Plantes, Universite Joseph Fourier et CNRS, Grenoble, France.
The LLS1 protein provides an important protection against light-dependent cell death in higher plants but its
molecular function remains undetermined. The cell death that occurs in lls1 plants is mediated by mature
chloroplasts leading us to hypothesize that LLS1 provides a protective function in that cell compartment (Gray et
al, 2002, Plant Physiology, 130:1894-1907). We provide evidence to show that the LLS1 protein is localized to
the inner membrane of the chloroplast in both monocots and dicots. Firstly, the acd-1 (accelerated cell death 1)
mutation of Arabidopsis was complemented using the maize Lls1 cDNA. Secondly, subcellular fractions of
isolated pea chloroplasts and mitochondria were analyzed by western blot analysis using an anti-LLS1
monoclonal antibody. The LLS1 protein was detected only in fractions enriched for the inner and outer plastid
membrane. Thirdly, in vitro -labeled LLS1 protein was imported into isolated chloroplasts and found to be
specifically localized to the inner chloroplast membrane. Fourthly, transgenic Arabidopsis plants expressing an
LLS1::GFP fusion protein were examined by confocal microscopy. The LLS1::GFP fusion proteins were
localized exclusively to the chloroplast and the predicted LLS1 transit peptide was sufficient to enable this
targeting. The LLS1 protein was also detectable by western blot in many plant species including lower plants such
as moss and ferns. A bioinformatics survey revealed that Lls1 homologues exist in algae (Chlamydomonas) and
cyanobacteria (Synechocystis, Anabaena and Trichodesmium) but not in the anoxygenic bacterium Chlorobium.
Although the LLS1 protein is always present in photosynthetic tissues the protein is also found in many non-
photosynthetic tissues and expression of the Lls1 gene increases following wounding. Our findings suggest that
the Lls1 gene evolved to help protect oxygenic photosynthesizers from free radical associated damage. This role is
of particular importance in the chloroplast compartment where chlorophyll and other pigments can readily
become phototoxic.
137
Identification of proteins associated with aflatoxin accumulation levels by 2-dimensional gel electrophoresis
Guenther, Eric {1} Soto, Jennifer {1} Musket, Theresa {1} Woodruff, Dana {1} Song, Qisheng {1} Gerau,
Michael {1} Sharp, Lindsey {1} Davis, Georgia {1}
{1} University of Missouri-Columbia
Food and grain contamination by aflatoxin causes worldwide crop loss. Aflatoxin is produced by the fungus
Aspergillus flavus and is a potent carcinogen. The FDA requires that aflatoxin may not exceed 20 parts per billion
for corn going to market. Host plant resistance to A. flavus is the best means of solving the problems caused by
aflatoxin. One tropically derived maize inbred, Mp313E, has lower aflatoxin accumulation levels when compared
to other lines. In contrast,Va35 accumulates high levels of toxin. Our goal is to identify proteins associated with
low toxin accumulation in Zea mays L. using 2-dimensional isoelectric focusing. In this experiment protein
patterns from infected or uninfected ear tissue Mp313E and Va35 were analyzed by means of 2-dimensional gel
electrophoresis. All material was self-pollinated 17 days before inoculation with A. flavus strain NRRL 3557.
Three replicates of inoculated and two replicates of uninoculated material were evaluated for each inbred. Plant
material was sampled at 5 days after inoculation. Proteins were prepared according to Hurkman and Tanaka
(1986) with modification and analyzed using 11cm, pH 5-8 strips in 12.5% Tris HCl gels stained with Coomassie
Blue. Differences in position and intensity of various protein spots on the gels were determined using PDQuest
software. These proteins are candidates for mass spectrometry to determine protein identity.
Blue. Differences in position and intensity of various protein spots on the gels were determined using PDQuest
software. These proteins are candidates for mass spectrometry to determine protein identity.
138
Allelic Variation of Gene Expression in Maize Hybrids
Guo, Mei {1} Rupe, Mary {1} Smith, Howie {1}
{1} Pioneer Hi-Bred International, Inc.
For nearly a century, geneticists and breeders have known that the hybrids between genetically unrelated
individuals often exhibit hybrid vigor or heterosis. The genetic factors contributing to hybrid vigor have mainly
been described by DNA sequence differences between the parental alleles. However, little is known at the RNA
level how the ìhetero-allelesî are expressed and their relevance to hybrid performance. In this study we measured
allele-specific transcript accumulation of a random set of 15 genes in an old and new maize hybrid grown together
in different environments. We discovered that alleles of 14 genes expressed differentially in at least one hybrid or
environment, ranging from expressing both alleles (bi-allelic) to expressing a single allele (mono-allelic). The
allelic expression varied during plant development and responded to drought and density stresses. The older
hybrid also tended to express one allele whereas the newer hybrid expressed both alleles. The results suggest an
unequal contribution of parental alleles in the hybrid and the advantage of bi-allelic expression in hybrids over
mono-allelic expression in inbreds, which may contribute to heterosis. The yield gains of the newer hybrids are
primarily due to the genetic improvement in tolerance to stresses, which could be the result of breeding selection
for alleles that respond to stresses. The large extent of allelic expression variation suggests that a substantial
fraction of quantitative trait variation is the result of sequence polymorphisms in the regulatory regions of the loci.
139
Cell Death in the lethal leafspot1 mutant of maize and its ortholog in sorghum, dropdead1
Hallar, Brittan {1} Hardee, Matthew {1} Janick-Buckner, Diane {1}
{1} Truman State University
The lethal leafspot1 (lls1) mutant of maize and its orthologous mutant in sorghum, dropdead1 (dd1), are not able
to suppress cell death in response to wounding. Although the function of the LLS1 protein is not known, it
appears to be localized in chloroplasts. We examined nuclear fragmentation and cytochrome c release from
mitochondria, characteristics of programmed cell death (PCD), in wounded leaves of lls1 and wild-type (WT)
plants. Fragmentation of DNA was examined 2 and 20 hours following wounding. DNA fragmentation in WT and
lls1 plants was evident at 2 hours, however, only lls1 plants displayed DNA fragmentation 20 hours post-
wounding. A limited amount of cytochrome c appeared to be released into the cytosol of wounded WT leaves.
However, a large amount of cytochrome c was released from the mitochondria into the cytosol, even in
unwounded lls1 leaves. This may indicate that these organelles in the lls1 plants are particularly fragile and
susceptible to breakage during the mitochondrial isolation procedure. Previously, we have observed that lls1 and
dd1 chloroplasts exhibit a decreased number of starch granules. This decreased starch accumulation indicates that
lls1 and dd1 plantsí photosynthetic apparatus may be compromised. By using western blotting, we analyzed the
levels of eight different light harvesting complex (LHC) proteins in both dd1 and WT leaves and found particular
LHC proteins to be increased in the dd1 leaves. We also examined the level of select photosynthetic enzymes
(Rubisco and PEP carboxylase) and mitochondrial proteins (alpha and beta subunits of ATP synthase, porin and
cytochrome c) in these plants. Rubisco was significantly increased in the dd1 leaves compared to WT, but none of
the other proteins showed significant differences. These findings suggest that there is a selective increase in the
expression of some proteins involved in photosynthesis, perhaps in response to a decreased overall photosynthetic
capability.
140
DISSECTION OF PLANT PROMOTER FUNCTION IN VIVO
Hernandez, J. Marcela {1} Lockwood, Robert D. {2} Matthews, Nicole {4} Szalma, Stephen J. {5}
McMullen, Michael D.{6} Grotewold, Erich {1, 2, 3, 4}
{1} Ohio State Biochemistry Program, Ohio State University {2} Molecular, Cellular and Developmental
Program, Ohio State University {3} Dept. of Plant Biology, Ohio State University {4} Plant Biotechnology
Center, Ohio State University {5} Genetics Area Program, University of Missouri {6} Plant Genetics Research
Unit, USDA-ARS, Columbia, MO 65211.
Combinatorial interactions between MYB and HLH transcription factors are required for the regulation of several
important processes in plants. Protein-DNA binding and transient expression experiments established a modular
structure for several maize flavonoid biosynthetic gene promoters, in which high- and low-affinity MYB binding
sites alternate. While these cis-regulatory elements are sufficient to drive expression of naÔve minimal promoters,
we have determined that insertions and mutations caused by transposons have effects not predicted from in vitro
protein-DNA binding or transient expression experiments. These results suggest that other cis-elements are also
important in vivo. The behavior of transposon insertions in the promoter of the anthocyaninless1 (a1) gene
(encoding dihydroflavonol reductase) provides an ideal system to investigate the role of transposons as control
elements. To further define which cis-elements are involved in the in vivo regulation, we analyzed the
corresponding promoter sequences of the a1 and colorless2 (encoding chalcone synthase) genes in a large number
of maize inbred lines, in various Zea species, and other grasses. This 'natural variation footprint' method provided
protein-DNA binding or transient expression experiments. These results suggest that other cis-elements are also
important in vivo. The behavior of transposon insertions in the promoter of the anthocyaninless1 (a1) gene
(encoding dihydroflavonol reductase) provides an ideal system to investigate the role of transposons as control
elements. To further define which cis-elements are involved in the in vivo regulation, we analyzed the
corresponding promoter sequences of the a1 and colorless2 (encoding chalcone synthase) genes in a large number
of maize inbred lines, in various Zea species, and other grasses. This 'natural variation footprint' method provided
information on the degree of variation allowed for experimentally-verified cis-regulatory elements, and identified
several conserved non-coding sequences (CNS) that may serve as additional binding sites for transcription factors.
141
Molecular dissection of the interaction between ZmMRP-1 and the promoter of BETL-1.
Gomez, Elisa {1} Barrero, Cristina {1} Royo, Joaquin {1} Grijota, Carmen {1} Hueros, Gregorio {1}
{1} Department of Cell Biology and Genetics. University of Alcala
ZmMRP-1 is a transfer cell specific transcriptional activator containing a single MYB-like DNA binding domain.
We have previously shown that ZmMRP-1 is necessary and sufficient to trans-activate various transfer cell
specific promoters in homologous and heterologous systems. We present here new results on the detailed
characterisation of the interaction ZmMRP-1/BETL-1 promoter. We also report on the functional characterisation
of ZmMRP-1 through the generation of a collection of 1000 mutated forms of the protein, by microbial
transposon insertion, and evaluation of their capacity to transactivate the promoter of BETL-1.
142
Expression Analysis of Glu-1Dx5 Transgenic Corn
Jia, Hongwu (1) Sangtong, Varaporn (1) Nettleton, Dan (1) Scott, Paul (2)
(1) Iowa State University (2) USDA-ARS
Maize endosperm is one of the most significant agronomic tissues because it is the major component of feed made
from maize grain. An important trait of endosperm is therefore its content of essential amino acids. Because the
majority of endosperm protein is in the form of seed storage proteins, the seed storage proteins have a large
impact on the amino acid content of the seed. The main storage proteins in maize endosperm are the zein proteins,
which are encoded by large gene families. We set out to manipulate seed protein content by introducing a gene
from wheat called Glu-1Dx5 into transgenic corn plants. This gene encodes a seed storage protein called high
molecular weight glutenin. We observed differences in nitrogen content between transgenic kernels and non-
transgenic kernels from the same ear. To characterize this difference in greater detail, the levels of the Glu-1Dx5
protein and the zeins in transgenic endosperms were quantified by HPLC. A peak of Glu-1Dx5 that reacts with a
Glu1Dx5 antibody can be observed in transgenic plants. Through this quantification, we found that the levels of
several zeins were altered in the seeds expressing the Glu-1Dx5 transgene. In order to identify differences in
transcript levels, a cDNA microarray was used to study the differences in expression patterns of 8064 selected
genes between transgenic and non-transgenic endosperms generated on the same ear. Two zein genes with
statistically significant differences in transcript levels were identified. Interestingly, the Glu-1Dx5 promoter
region contains a G-box like motif that could possibly serve as the binding site for the maize transcription factor
Opaque2. We have produced F2 seed of a cross between o2 and Glu1-Dx5 lines. It will be interesting to see the
effect of o2 on transgene expression.
143
Microarray Analysis of Maize Opaque2 In Developing Endosperms of Eight Inbred Lines
Jia, Hongwu (1) Nettleton, Dan (1) Vazquez, Gricelda (1) Scott, Paul (2)
(1) Iowa State University (2) USDA-ARS
The Opaque2 (O2) mutation can significantly improve the protein nutritional quality of corn kernels by increasing
the lysine content of maize endosperm. Since this discovery, it has been the subject of intense research. Studies
conducted over the past 40 years have revealed that the o2 mutation reduces the synthesis of zein storage proteins
and increases the level of free amino acids, including lysine. The biochemical and molecular mechanisms leading
to these phenomena are only partially understood. We analyzed the grain from eight inbred lines and their
backcross o2 conversions and found that the extent of change in amino acid content of an o2 mutant is dependent
on its genetic background. In recent years, high throughput microarray studies have been conducted to study the
pattern of expression in normal and o2 mutant with a single genetic background. However, background-specific
effects of o2 can obscure the main action of o2. In order to eliminate such marginal effects, we sought to identify
genome-wide gene expression changes that are consistently maintained in the o2 mutant lines in several genetic
backgrounds. Fourteen days after pollination, developing endosperms were harvested from eight pairs of near-
isogenic lines. The pooled mRNA was labeled and hybridized to Unigene arrays from ZmDB. A duplicate dye-
swap design was used for the microarray experiment and an ANOVA method was used to analyze the data. Our
results showed that more than 100 genes are significantly down-regulated, including expected genes such as zeins,
a pyruvate orthophospate dikinase, and ribosome-inactivating protein b32 and a few are up-regulated significantly
in o2 mutants. Further analysis will be done on these genes. In the future work, we hope to understand factors
influencing amino acid content by studying global expression patterns of o2 at different developmental stages and
o2 mutants from specific genetic backgrounds.
results showed that more than 100 genes are significantly down-regulated, including expected genes such as zeins,
a pyruvate orthophospate dikinase, and ribosome-inactivating protein b32 and a few are up-regulated significantly
in o2 mutants. Further analysis will be done on these genes. In the future work, we hope to understand factors
influencing amino acid content by studying global expression patterns of o2 at different developmental stages and
o2 mutants from specific genetic backgrounds.
144
Evaluation of Near-Isogenic Lines for Kernel Oil Concentration QTL in Maize
Johnson, Jeremy {1} Rocheford, Torbert {1}
{1} University of Illinois, Department of Crop Sciences
Value-added specialized corn hybrids, which may provide efficiencies and economic benefits, have the potential
to be grown much more extensively in the Midwest. One such example is high-oil corn (HOC) grown for its
livestock feeding value. However, negative associations with grain yield have hindered the success of HOC. We
evaluated sets of near-isogenic lines (NILs) for a chromosome region with a major oil concentration quantitative
trait locus (QTL) or linked QTL derived from a (Illinois High Oil x B73) B73 BC1S1 population. NILs may
facilitate confirmation of QTL, fine mapping, and dissecting the basis of the QTL. Three SSR markers linked to
the major oil QTL on chromosome 6 were used in the development of BC1 S5:7 and S7:9 NILs. Testcross
progeny were made from BC1S8 NIL sublines to Mo17 for evaluation of NILs in a hybrid context. Grain
collected from the lines were evaluated for their kernel oil, protein and starch concentration using near-infrared
(NIR) analysis. To better dissect the basis of the QTL, kernel weight, embryo weight and embryo oil
concentration, using nuclear magnetic resonance (NMR), of selected sub lines from the S5:7, S7:9, and testcrosses
will be reported. Single factor analysis of variance confirmed the kernel oil QTL in the S5:7 NILs evaluated in
2001. Evaluation of a second year of S5:7 NILs and two environments of S7:9 NILs in both per-se and testcross
evaluations in 2002 show significant differences in kernel oil concentration between marker class means,
confirming presence of QTL. Testcross evaluations for yield detected a negative association with the kernel oil
concentration QTL and final grain yield. However, the 2002 growing year was a stressful drought environment
and yields were poor in general.
145
A DNA methyltransferase mutation increases expression of the epigenetically-regulated maize gene, Pl-
Blotched
Brewer, Nathan {1} Cone, Karen {1}
{1} Division of Biological Sciences, University of Missouri-Columbia, Columbia, MO 65211
Changes in DNA methylation and chromatin structure have been associated with altered gene expression in a
variety of different organisms, including maize. In maize, an example of such an epigenetic effect has been
observed at Pl-Blotched, an allele of the anthocyanin regulatory gene purple plant1 (pl1). Pl-Blotched leads to a
variegated pattern of anthocyanin pigmentation compared to the uniformly pigmented phenotype of plants with
the Pl-Rhoades allele. The phenotypic difference of Pl-Blotched is accompanied by a different pattern of DNA
methylation and by a more condensed chromatin state. To help us understand the relationship of DNA
methylation to chromatin structure and expression of Pl-Blotched, we introduced a mutant allele of the DNA
methyltransferase Zmt2. This DNA methyltransferase gene is one of two duplicates in the maize genome that
encode enzymes with methylation specificity for CNG sequences. In Pl-Blotched plants with the Zmt2 mutant,
anthocyanin pigmentation was higher than in plants with the wild-type DNA methyltransferase gene. The increase
in pigmentation coincided with a modest decrease in cytosine methylation at diagnostic restriction sites in the
gene, opening of chromatin structure in the 3' region of the gene, and an increase in pl1 mRNA accumulation.
These results indicate that disrupting DNA methylation leads to a more open chromatin state and higher levels of
gene expression. These observations point to the importance of DNA methylation of the Pl-Blotched gene in
maintaining the chromatin structure of this epiallele.
146
Genomics analysis of lipoxygenase gene family
Kolomiets, Michael (1) Navarro, Pedro (2) Zhang, Jinglan (1) Nemchenko, Andriy (1) Simmons, Carl (2)
Yalpani, Nasser (2) Meeley, Robert (2) Duvick, Jon (2)
(1) Texas A&M University (2) Pioneer Hi-Bred Intl.
Lipoxygenases (LOX) are dioxygenases that catalyze the hydroperoxidation of polyunsaturated fatty acids such as
linolenic and linoleic acids. Several lines of correlative evidence strongly implicate lipoxygenases and their
metabolic products in the resistance mechanisms to diverse plant pathogens and in accumulation of mycotoxins
such as aflatoxin. As the first step in our attempt to clearly define LOX functions, we have undertaken a genomic
scale analysis and identification of the entire LOX multigene family of maize, rice and Arabidopsis using wealth
of Pioneer and DuPont EST collection as well as public EST and genomic databases. Using bioinformatics
analysis of ESTs representative of more then 130 different cDNA libraries, full insert sequencing and 5íRACE
cloning technique we have identified that maize genome contains at least 11 different LOX genes. In comparison,
rice contains at least 14 different genes and Arabidopsis genome encodes only 6 LOX genes. This number has
been independently confirmed by using gene-specific probes in Southern blotting analysis and by identification of
chromosome location of each individual gene by using gene-specific PCR primers and genomic DNA from oat-
of Pioneer and DuPont EST collection as well as public EST and genomic databases. Using bioinformatics
analysis of ESTs representative of more then 130 different cDNA libraries, full insert sequencing and 5íRACE
cloning technique we have identified that maize genome contains at least 11 different LOX genes. In comparison,
rice contains at least 14 different genes and Arabidopsis genome encodes only 6 LOX genes. This number has
been independently confirmed by using gene-specific probes in Southern blotting analysis and by identification of
chromosome location of each individual gene by using gene-specific PCR primers and genomic DNA from oat-
maize addition lines. Homologs of two of these LOX genes have never been reported in any other plant species
and can be grouped into two novel classes of plant LOX genes. RNA profiling and Northern blotting results
suggested that many of the LOX genes are up-or down-regulated by pathogens, their elicitors or defense-
associated signaling molecules such as JA, ethylene and SA suggesting their involvement in pathogen-induced
defense responses. By using reverse genetics strategy, we have identified maize mutants in which function of 8
LOX genes is interrupted by insertions of Mutator transposable elements in their coding sequences. These mutants
will undergo thorough testing for disease resistance and mycotoxin accumulation levels to unambiguously define
LOX function in defense responses and interaction between pathogens and the plant host.
147
A maize Y1 homolog encodes a functional phytoene synthase.
Li, Faqiang (1) Gallagher, Cynthia E. (1) Matthews, Paul D. (2) Wurtzel, Eleanore T. (1,2)
(1) Lehman College, The City University of New York, Bronx, NY 10468 (2) The Graduate Center of the City
University of New York, New York, NY 10016
Plant-derived carotenoids are important in human health and include a class of over 600 structures. In plants,
carotenoids function in light harvesting, protect against photo-oxidative damage and act as precursors to abscisic
acid (ABA). The carotenoid biosynthetic pathway takes place on plastid membranes by nuclear-encoded enzymes
that require appropriate plastid-targeting domains. The marked difference in plastid membrane architecture
between endosperm and photosynthetic tissue suggests possible differences for pathway assembly depending on
tissue/plastid type. As part of an ongoing effort to investigate regulation of the pathway, we are characterizing
gene families and enzymes for the entire pathway, including isoprenoid precursors. In this study, we investigated
the first step of the carotenoid biosynthetic pathway, synthesis of phytoene from geranylgeranyl pyrophosphate
(GGPP), a key compound which lies at a metabolic crossroad. This step is catalyzed by phytoene synthase (PSY)
and is rate-controlling for carotenoid accumulation. We identified a maize cDNA (PSY2) showing homology to
maize Y1, the previously identified structural gene for PSY. When the PSY2 cDNA was used as a hybridization
probe of maize genomic DNA, a different hybridization pattern was produced as compared when Y1 was used as a
probe, suggesting that PSY2 represented a second gene. To test whether this second gene encoded a functional
phytoene synthase or was a pseudogene, we used a heterologous complementation system to confirm function of
PSY2. A gene cluster encoding the carotenoid biosynthetic pathway enzymes, with the exception of PSY, was
introduced into E. coli. Only when a cDNA encoding a functional PSY was introduced, would zeaxanthin
accumulate. When the Y1 homolog cDNA (PSY2) was introduced, zeaxanthin accumulated. These results indicate
that there is a second phytoene synthase gene in maize which encodes a functional phytoene synthase. We are
currently investigating the roles of both phytoene synthase enzymes and their corresponding genes in conditioning
carotenoid accumulation in maize. This research was funded in part by the National Institutes of Health.
148
RAD51 is required for chromosome segregation but not for chromosome pairing or cell viability in maize
Li, Jin {1} Weber, David {2} Meeley, Robert B. {3} Bowen, Ben {4} Schnable, Patrick S. {1}
{1} Iowa State University, Ames, IA 50011 {2} Illinois State University, Normal, IL 61790 {3} Pioneer Hi-Bred
Intl. Inc., Johnston, Iowa 50131 {4} Previously, Pioneer Hi-Bred Intl. Inc., Johnston, Iowa 50131; Currently,
Lynx Therapeutics, Inc., Hayward CA 94545
RAD51, a RECA-homolog of yeast, plays a central role in homologous recombination and DNA repair. To define
the functions of RAD51 in plants, a reverse genetic method was used to isolate Mu insertion alleles of the two
maize rad51 homologs. A PCR-based screen was used to isolate deletions of exonic DNA adjacent to Mu
insertions. Flanking deletion derivatives were obtained at a rate close to 1% (4/500). Deletions ranged from 69 to
177 bp. rad51 double mutants are male sterile and have seed sets of only 5-10% of wild type. Cytogenetic analysis
of these double mutants revealed that pachytene cells were normal in that homologous chromosomes were
completely synapsed. In contrast, bivalents prematurely dissociated during diakinesis and more than 25% of
quartets carried cells that were either missing an organized nucleolus or had two nucleoli. These observations
indicate that nondisjuction occurs at both meiosis I and II in cells that lack RAD51. Hence, although RAD51 not
required for synapsis, it is required for proper interhomolog and sister chromosome segregation. This study
demonstrates a significant difference between the roles of RAD51 in plants and vertebrates. Vertebrate cells that
lack RAD51 can not propagate due to the accumulation of chromosome breaks. In contrast, somatic cells of maize
do not require RAD51 for normal division.
149
Color complementation in E. coli for functional testing of a cDNA required for maize carotenoid
biosynthesis
Licciardello, Nicholas (1) Gallagher, Cynthia E. (1) Cervantes-Cervantes, Miguel (1) Wurtzel, Eleanore T.
(1,2)
(1) Lehman College, The City University of New York, Bronx, NY 10468 (2) The Graduate Center of the City
University of New York, New York, NY 10016
Carotenoids represent a class of over 600 compounds with a 40-carbon backbone derived from 5 carbon
isoprenoid units; in plants, they are essential in photosynthesis and serve as precursors to abscisic acid. The
biosynthetic pathway takes place on plastid membranes by nuclear-encoded enzymes that require appropriate
plastid-targeting domains. The marked difference in plastid membrane architecture between endosperm and
photosynthetic tissue suggests possible differences for pathway assembly depending on tissue/plastid type. As part
of an ongoing effort to investigate regulation of the pathway, we are characterizing gene families and enzymes for
the entire pathway including isoprenoid precursors. From limited sequence in GenBank, we identified a cDNA
predicted to encode a betacarotene hydroxylase (HYD), an enzyme that catalyzes the addition of hydroxyl groups
at position three on the cyclic rings of betacarotene to form the xanthophyll, zeaxanthin. This maize (Zea mays)
cDNA was sequenced and found to encode a 359 amino acid polypeptide with 56% identity to a homolog from
the dicot, Arabidopsis thaliana. Since sequence homology implies, but does not always indicate enzyme
specificity and activity, we used a heterologous bacterial system to demonstrate function of the putative maize
betacarotene hydroxylase. The cDNA was cloned in-frame into the pET23c vector and then tested by co-
transformation with a second plasmid conferring accumulation in Escherichia coli of the HYD substrate,
betacarotene. Presence of the maize HYD cDNA was associated with altered colony pigmentation as compared to
the control strain accumulating betacarotene but lacking the maize cDNA. HPLC analysis confirmed
accumulation of the HYD enzymatic product, zeaxanthin and the intermediate, betacryptoxanthin, thereby proving
the cDNA encoded a functional betacarotene hydroxylase. This undergraduate research project was funded in part
by the National Institutes of Health.
150
Isolation and characterization of a maize cDNA encoding a telomere repeat DNA oligonucleotide-binding
protein
Marian, Calin O {1} Goltz, Marion {1} Santarella, Rachel A {1} Danilevskaya, Olga {2} Meeley, Robert {2}
Bass, Hank W {1}
{1} Florida State University, Tallahassee, FL USA, 32306 {2} Pioneer Hi-Bred Int., Johnston IA, USA, 50131
To determine the molecular basis of meiotic telomere behavior, we have set out to identify genes that encode
telomeric complex proteins using a conserved myb-like telomere-binding domain of human TRF1 to query EST
databases. From these searches and by cDNA library screens, we have discovered and sequenced five full-length
maize cDNAs that represent a small gene family with a novel domain organization that is apparently unique to
plants. Here we describe the preliminary characterization of one member (called MHCP5) of this family. The
ORF of MHCP5 was expressed in E. coli, and the recombinant protein was found to bind double-stranded
telomere repeat oligonucleotides in vitro. Gel mobility shift assays with 21 different mutant variant oligos
established that MHCP5 binds most strongly to the wild-type telomere repeat. Southern blots revealed that the
MHCP5 cDNA hybridizes to several restriction fragments at moderate stringency, consistent with the idea that
MHCP5 belongs to a small gene family. RNA gel-blot analysis with a 3í UTR probe indicated that this gene is
expressed at low levels in a few plant tissues. A pair of primers for MHCP5 produced two PCR products with
DNA from either B73 or Mo17 inbred lines. One PCR product was from MHCP5 and the other from a putative
duplicate gene termed MHCBOP5. All four PCR products were cloned, sequenced, and then genetically mapped
with recombinant inbred lines using the 96 well PCR-based IBM DNA Mapping Kit. MHCP5 and MHCBOP5
were mapped in duplicated regions on two different chromosomes. Possible functions of MHCP5 and this novel
gene family will be discussed.
151
Characterization of an oligopeptide transporter
Moody, David {1} Lubkowitz, Mark {1}
{1} Saint Michael's College 1 Winooski Park Colchester, VT 05439
Peptide transporters are a well-studied physiological phenomenon in prokaryotes and in fungi. They function in
transporting small peptides across the plasma membrane in an energy-dependant manner. In lower eukaryotes
internalized peptides serve primarily as a source of amino acids for protein synthesis or as a nitrogen source.
Peptide transporters have been classified into different groups based on the energetics and the length of the
substrate. The PTR (Peptide TRansporter) family uses the proton motive force to translocate peptides two to three
amino acids in length. The OPT family (OligoPeptideTransporters) also uses the proton motive force for
translocation but the substrates are 3-6 amino acids in length. OPT and PTR transporters have been studied in
Arabidopsis but none have been characterized in grasses. We have identified several putative OPT type
transporter in Zea mays using orthologous sequences from yeast and Arabidopsis. We are in the process of
characterizing one of these genes named ZmOPT3. We are using heterologous expression of ZmOPT3 in yeast to
substrate. The PTR (Peptide TRansporter) family uses the proton motive force to translocate peptides two to three
amino acids in length. The OPT family (OligoPeptideTransporters) also uses the proton motive force for
translocation but the substrates are 3-6 amino acids in length. OPT and PTR transporters have been studied in
Arabidopsis but none have been characterized in grasses. We have identified several putative OPT type
transporter in Zea mays using orthologous sequences from yeast and Arabidopsis. We are in the process of
characterizing one of these genes named ZmOPT3. We are using heterologous expression of ZmOPT3 in yeast to
determine if ZmOPT3 does indeed encode an oligopeptide transporter. Furthermore, we are attempting to discern
if phytosulfokine, a pentapeptide plant hormone, is a substrate for ZmOPT3.
152
Isolation of Genes Controlling Agronomic Traits in Maize
Murigneux, Alain (1) Lessard, Philippe (1) Martinant, Jean-Pierre (1) Huttner, Eric (1) Debrand, Michel
(1) DeRose, Rick (2) Sarda, Xavier (2) James, Evelyne (2) Chataigner, Virgine (2) Sajot, Nicolas (2)
Gigonzac, Olivier (2) Freyssinet, Georges (2) Joets, Johann (3) Legeai, Fabrice (4) Samson, Delphine (4)
Barillot, Emmanuel (4)
(1) Biogemma (2) RhoBio (3) INRA (4) GÈnoplante-Info
GENOPLANTE, the French Genomic initiative, has initiated in 1999 a Maize Programme devoted to the
identification of genes playing a key role in the control of drought or cold tolerance, earliness, grain filling and
maturation, and whole plant digestibility. Among the overall objectives, one project focused on two main goals:
(1) production of 100,000 ESTs (Expressed Sequence Tags) and (2) identification of sequences differentially
expressed under specific conditions. This work has been performed in collaboration (and under contract) with the
Celera AgGen Company (Davis, California). Plant tissues were sampled from various sources of material
differing in genotype, organ, stage of development, conditions of plant growth (temperature, water availabilityÖ).
Libraries were constructed from 76 mRNA samples. Three kinds of libraries were produced: (1) standard libraries
without any normalisation or subtraction procedure; (2) normalised libraries, with SSH protocol procedure (the
mRNAs were subtracted against themselves), and (3) subtraction libraries (tissue 1 ñ tissue2 Ö). Sequences were
produced for 15 standard libraries, 12 normalised libraries and 30 subtraction libraries. The EST sequencing,
completed by May 2001, produced 107,559 sequenced > 99bp, merging into 24,775 contigs. The combined
assembly of these sequences with publicly available maize ESTs (70,807 in May 2001) produced 31,664 contigs.
A modified cDNA AFLP display developed by Celera AgGen and called AFLP-Transcript Imaging (TI) was used
to obtain the transcript profile of 40 mRNA samples (partially redundant with the 76 mentionned above). In each
sample, about 20,000 tags (Genetags) are characterised for their size and expression level. The comparison among
the various samples allows the identification of differentially expressed genes (TI bands). The sequence
determination of these bands is made possible by the Genetag database that contains about 150,000 cDNA AFLP
tags (Genetags). These were produced through the sequencing of all bands produced with the same TI protocol on
reference samples. Identification is based on comparing the size of the TI bands scored on capillary
electrophoresis runs and the length of the Genetag sequences for corresponding primer pairs. Sequences
corresponding to EST sequencing and the Genetag sequences, in total about 250,000, will be publicly released by
July 1st, 2003. This represents so far the largest release of maize sequences in the public domain. Comparisons
were made between samples relevant for earliness, kernel filling, drought and cold tolerances. 4,562 non
redundant differentially expressed bands (TI bands) were observed and a sequence identification is proposed for
3,352 of them using the Genetag database. Verification of these differentially expressed bands, currently under
process, is necessary as ambiguity in sequence identification may remain when several Genetags sequences
overlap the confidence interval of the size of the TI bands. This verification process made by RT-PCR using the
same samples as used above (for differential expression discovery) allows in a single experiment the confirmation
of sequence identity and confirms the differential expression between the two conditions. The presentation will
illustrate the diversity/redundancy and the specificity of sequences obtained within and across different libraries in
relation with (1) the nature and the complexity of dissected organs, (2) the library construction methodologies.
The presentation of the contig assembly merging EST from this project, public EST and Genetag sequences will
document the complementarities of EST Sequencing versus Transcript Imaging approaches. In silico northern
analysis of these sequences will also be presented and finally the use of some sequences in a functional genomic
programme will be exemplified.
153
Ectopic knotted1 expression does not perturb gibberellin biosynthesis in maize.
Muszynski, Michael {1} Hedden, Peter {2} Archibald, Rayeann {1}
{1} Agronomic Traits, Pioneer Hi-Bred Intl., Inc., Johnston, IA 50131, USA {2} Long Ashton Research Station,
University of Bristol, Long Ashton, Bristol, BS41 9AF, UK
The maize knotted1 homeodomain transcription factor functions to maintain the indeterminate properties of
meristem cells. In maize, ectopic expression of kn1 and other class 1 knox genes through dominant mutations
results in displacement of the ligule into the blade, ectopic ligule formation and the production of tissue
outgrowths or ìknotsî. Ectopic expression of class 1 knox genes in dicots results in altered leaf shape, as well as
formation of ectopic meristems. Recent evidence from dicots indicates that knox misexpression phenotypes
maybe mediated by plant hormones. Ectopic knox expression can result in increased cytokinin (CK) accumulation
and reduced gibberellic acid (GA) levels. In tobacco, the KNOX protein NTH15 directly inhibits transcription of
the GA 20-oxidase biosynthetic gene Ntc12. In order to elucidate targets of maize knotted1, the lipid-transfer
results in displacement of the ligule into the blade, ectopic ligule formation and the production of tissue
outgrowths or ìknotsî. Ectopic expression of class 1 knox genes in dicots results in altered leaf shape, as well as
formation of ectopic meristems. Recent evidence from dicots indicates that knox misexpression phenotypes
maybe mediated by plant hormones. Ectopic knox expression can result in increased cytokinin (CK) accumulation
and reduced gibberellic acid (GA) levels. In tobacco, the KNOX protein NTH15 directly inhibits transcription of
the GA 20-oxidase biosynthetic gene Ntc12. In order to elucidate targets of maize knotted1, the lipid-transfer
protein2 promoter (Ltp2) from barley was used to drive ectopic expression of kn1 in transgenic maize plants.
Ltp2:Kn1 transgenic plants display phenotypes resembling class 1 knox dominant mutations (e.g. Knotted1-O,
Rough Sheath1-O and Gnarley1-R). Transgenic plants have displaced blade-sheath boundaries, ectopic ligules in
the blade and tissue outgrowths localized along the mid-rib. Leaf tissue from transgenic plants shows abundant
accumulation of kn1 mRNA. We tested if genes known to modulate GA biosynthesis in maize, including GA 20-
oxidase, were regulated by kn1. None of the genes tested showed differential expression between the transgenic
and non-transgenic sibs. Endogenous GA levels were measured in 1-week and 2-week seedlings. No differences
in endogenous GA levels were detected in 2-week seedlings and a modest (1.9-fold) reduction in GA1 was
detected in 1-week seedlings. The 1-week seedling experiment was not replicated; therefore the slight reduction in
GA1 levels cannot be distinguished from experimental error. These preliminary data suggest that the ectopic kn1
phenotype in maize is not mediated through altered GA levels.
154
Chromosomal Distribution of Maize Repetitive Sequences
Odland, Wade {1} Kynast, Ralf {1} Okagaki, Ron {1} Rines, Howard {2} Phillips, Ronald {1}
{1} Department of Agronomy and Plant Genetics, University of Minnesota {2} USDA ARS and Department of
Agronomy and Plant Genetics, University of Minnesota
The maize genome is composed largely (85%) of repetitive sequences, most of which are retrotransposons. Many
of these elements are present in high copy numbers and widely dispersed across the genome. A series of oat-maize
chromosome addition (OMA) lines, each with an individual maize chromosome added to a hexaploid oat genome,
provides the opportunity to assess the relative copy number of specific repetitive elements of each maize
chromosome. Microarray technology measures the relative abundance difference of a sequence between two
labeled DNAs. Labeling two OMA lines and hybridizing them to an array with immobilized maize repetitive
sequences can directly compare the amount of an element present between the two maize chromosomes. The
probes in this research were designed to detect whole elements or unique motifs within specific repetitive
elements. Preliminary trials have shown that copy number differences can be detected between some of the
chromosomes. Trials have been done comparing chromosome 4 with 6 and 3 with 9. Maize chromosomes 3 and 9
have distinctive compositions with respect to certain repetitive sequences. Probes for Zeon1 and Opie2a show the
greatest differences between these two chromosomes. The repetitive sequences of chromosomes 4 and 6 are
similar in copy number (within 8%). Such results will improve the clarity of the repetitive sequence organization
within maize. This research is based upon work supported by the National Science Foundation under Grant No.
0110134.
155
The Effects of Host Genetic Background on Aboveground Microsymbionts: Endophytes on Smut-resistant
and Smut-susceptible Corn
Pan, Jean {1} May, Georgiana {2}
{1} Dept. of Ecology, Evolution, and Behavior, University of Minnesota {2} Dept. of Ecology, Evolution, and
Behavior and Dept. of Plant Biology, University of Minnesota
In agricultural systems, selection for resistance is usually to a particular pathogen, normally the one that is most
economically detrimental at the time. However, the ecological and evolutionary effects of this change in plant
environment for other members of the microbial community is not well understood. We are using the Zea mays-
Ustilago maydis (corn-corn smut) system to address this issue. Recombinant inbred lines of corn resistant or
susceptible to corn smut have been developed and genetic differences associated with resistance or susceptibility
have been documented (R. Phillips, see Baumgarten et al. poster). Here, we present preliminary data on endophyte
diversity from a field survey conducted in 2002. Endophytes were cultured from surface-sterilized corn tissue
samples on media selective for fungi. Endophytes were categorized morphologically, based on colony
characteristics. Over 120 unique colony types were identified and over 80 unique colony types were present in the
most fully sampled line. We found that over 300 samples were needed to characterize the endophyte diversity of
each line. A trend toward lower endophyte diversity in smut-susceptible lines is noted, suggesting that
susceptibility to one microbial symbiont, smut, may not increase susceptibility to other symbionts. Next, we plan
to use molecular approaches for endophyte identification in order to more fully characterize the endophyte
diversity on these corn lines.
156
Rmr7 is necessary for silencing at paramutable maize loci.
Parkinson, Susan {1} Hollick, Jay {1}
{1} University of California, Berkeley
The required to maintain repression 7 (rmr7) locus encodes a trans-acting factor necessary for maintaining the
silencing that occurs at the maize purple plant 1(pl1) locus as a result of paramutation. The pl1 locus encodes an
R2R3 Myb-type transcription factor required for anthocyanin production in specific plant tissues. The Pl1-
Rhoades (Pl) allele confers high levels of anthocyanin production, but it occasionally changes to a low expressing
state designated Pl1í-mahogany (Plí). In plants heterozygous for Pl and Plí, the Pl allele is changed to the Plí state,
and is therefore an example of paramutation. In an EMS mutagenesis screen for factors affecting pl1
paramutation, we isolated two mutant alleles of rmr7 (rmr7-1 and rmr7-2). Genetic analyses show that both
mutations represent single locus recessive alleles of a novel genetic locus, located on chromosome 2S. The rmr7
mutant plants can be identified in a Plí background because anthocyanin pigmentation is increased to levels seen
in Pl plants. Molecular data reflects this increase, as rmr7-1 plants have significantly higher levels of pl1 mRNA
than do rmr7-1/+ plants. Furthermore, tests with other recessive pl1 alleles show that the increased anthocyanin
production in rmr7 mutants occurs only in Plí plants, confirming that rmr7 affects Plí silencing specifically and
not anthocyanin biosynthesis in general. The rmr7 locus is not required for normal plant development, as rmr7
mutants are morphologically similar to their wild-type siblings, even after five generations of inbreeding. In
addition to its role in maintaining Plí silencing, genetic tests indicate that rmr7 function is required to establish
paramutation at the booster1 (b1) locus, which encodes another transcription factor involved in anthocyanin
pigmentation. The role of rmr7 in epigenetic repression at both pl1 and b1provides evidence of a mechanistic link
between paramutation at these loci. Furthermore, the developmental stability of rmr7 mutants suggests a specific
rather than global role in gene silencing.
157
EFFECTS OF SELECTION ON SEQUENCE DIVERSITY AND LINKAGE DISEQUILIBRIUM
AROUND THE MAIZE Y1 LOCUS
Palaisa, Kelly {1} Bhatramakki, Dinakar {2} Williams, Mark E {1} Smith, Oscar S.(Howie) {2} Tingey,
Scott {1} Rafalski, J.Antoni {1}
{1} DuPont Crop Genetics Molecular Genetics Group, Newark DE 19711 {2} Pioneer Hi-Bred International,
Johnston, IA 50131-1004
The extent of linkage disequilibrium (LD) in the maize population has important consequences for the usefulness
of genetic association mapping. Recent reports in maize have described a rapid decline in LD, ranging from 100-
200 bp to approximately 1500-2000bp [Tenaillon et al., 2001; Remington et al., 2001; Thornsberry et al., 2001].
To further evaluate the patterns of LD and sequence diversity in maize, the sequences within and surrounding the
Y1 yellow endosperm locus were analyzed. The Y1 gene determines the endosperm color of maize kernels and is
thought to have undergone repeated selection for the yellow endosperm phenotype due to its higher nutritional
content. Taking advantage of the availability of a 1Mb BAC contig with this locus at a central position [Morgante,
2001], various portions of the Y1 gene and its surrounding genomic regions were amplified and sequenced from a
set of 75 white and yellow corn inbreds. All polymorphic sites were identified. Significantly higher nucleotide
diversity was found in the white endosperm inbreds at the Y1 locus, while the yellow endosperm inbreds shared a
common haplotype throughout the 6kb region, thereby showing significant LD. The low copy regions surrounding
Y1 evinced similar patterns of diversity up to 500kb+ away from the Y1 locus on one side only. LD appears to
extend much farther than was observed previously, presumably due to the region being subject to selective forces.
158
Methylation analysis of tissue culture-induced P white cob maize mutants
Rhee, Yong {1} Kaeppler, M. Shawn {1}
{1} The University of Wisconsin-Madison
Plant tissue culture is a mutagenic process producing a high frequency of genetic changes among regenerated
plants and their progenies. Studies linking mutant phenotypes with molecular changes indicate that transposons,
base changes, and small insertions/deletions contribute to tissue culture-induced mutation. The objective of this
study was to characterize a potential case of tissue culture-induced epigenetic silencing of P-wr. Five independent
p-ww alleles were identified based on segregation for a white cob phenotype among regenerant-derived progeny,
and confirmed by allelism tests. The explant source, LH51, contains a P-wr allele. DNA gel-blot analysis
indicated that no major rearrangement of the repeated locus had occurred, but revealed hypermethylation in all
five new p-ww alleles. DNA gel-blot analysis and PCR have been used to determine the pattern of
hypermethylation across the mutant alleles. Results of the methylation analysis will be reported and interpreted in
the context of previous studies on the P locus.
the context of previous studies on the P locus.
159
Transgenic Luciferase Constructs That Respond to Abiotic Stress in Maize
Rich, Patrick J. {1} Bennetzen, Jeffrey L. {1}
{1} Purdue University
We have transformed maize with two stress-responsive reporter constructs. The first construct contains the maize
rab17 promoter linked to firefly luciferase structural gene. The second contains the maize Adh1-S promoter and
luciferase. Both constructs were introduced in separate events into maize HiII embryogenic callus at the Iowa
State Plant Transformation Facility by particle bombardment. Selected plants containing either construct were
regenerated. Events containing rab17-LUC are being assayed for luciferase expression after applying osmotic
stresses (limited irrigation and irrigation with sodium chloride or polyethylene glycol solutions) and abscisic acid.
Response of the transgene with each treatment is measured by CCD imaging. Various tissues at several
developmental stages can be screened non-destructively with this method. Likewise, the Adh1-LUC plants are
assayed for luciferase expression after exposure to hypoxic conditions and cold temperatures. We now have seed
from both types. Those showing patterns of luciferase expression similar to that expected for the native rab17 or
Adh1 genes are being advanced to generate maize lines with reporters that specifically respond to various
environmental stresses. These should be useful tools for studies in plant environmental stress physiology and
genetics.
160
Gene Expression in the Maize Endosperm Revealed by Genome-wide mRNA Profiling
Rupe, Mary {1} Hu, Zihua {2} Guo, Mei {1}
{1} Pioneer Hi-Bred International, Inc. {2) State University of New York at Buffalo.
We have taken a genomic approach to examine global gene expression in the maize endosperm in relation to
dosage and parental effects. Endosperm of eight hybrids generated by reciprocal crosses and their seven inbred
parents were sampled at three developmental stages; 10, 14 and 21 days after pollination. These samples were
subjected to GeneCalling, an open ended mRNA profiling technology, that simultaneously analyzes thousands of
genes. Results indicated that the overall level of gene expression in the maize endosperm was dosage dependent;
that is, the gene expression was proportional to the parental genome contribution of 2n maternal : 1n paternal.
However, approximately 8% of the genes deviated from such allelic dosage expression and exhibited differential
expression in hybrids of reciprocal crosses, resembling either maternally or paternally expressed genes. There
were more genes with maternal-like expression than with paternal-like expression. Allele-specific expression
analysis of four selected genes using the WAVE denaturing HPLC (dHPLC) system revealed several mechanisms
responsible for the deviation from the allelic additive expression in the hybrid endosperm: heterochronic allelic
variation, allelic variation in the level of expression and genomic imprinting. A novel imprinted gene, the maize
homologue of Non Apical Meristem (NAM) was identified.
161
Expression Profiling Across Diverse Maize Germplasm
Saft, Beth {1} Rauh, Brad {1} Goodman, Major {2} Buckler, Edward {1}
{1} USDA/ARS, North Carolina State University, Raleigh NC {2} Department of Genetics, North Carolina State
University, Raleigh, NC
The quality and nutritional value of maize kernels varies greatly in diverse germplasm. Many genes are involved
in the biosynthesis of starch, protein, and oil in the kernel. Upwards of 100 genes are involved in the regulation
and accumulation of zein protein storage bodies alone. To identify additional candidate genes involved in energy
biosynthesis in the maize kernel we completed expression profiles from microarrays produced by the Maize Gene
Discovery group that are specific to the genes expressed during maize endosperm development (series 605).
Expression profiles as well as numerous other phenotypes were collected for analysis in our standard 102 inbred
maize lines that represent much of the diversity of modern maize inbreds. Ears were successfully harvested 14
days after pollination from 45 and 89 diverse inbred lines and mutant stocks throughout the 2000 and 2001
summers, respectively, in North Carolina. Total RNA was isolated, labeled and hybridized to arrays in a loop
design. Through a range of analysis approaches, we show the variation in expression across diverse germplasm,
we examine the heritability of the expression profiles, we relate these expression profiles to kernel quality, and we
relate variation in expression to genetic relatedness based on SSR data. This study demonstrates that expression
profiling is a powerful way to characterize germplasm diversity.
162
Functional analysis of a sorghum myb orthogolous transcription factor promoter in transgenic maize
Sangar,Vineet{1} Peterson,Thomas{2} Chopra,Surinder{1}
{1} The Pennsylvania State University, University Park, PA {2} Iowas State Univeristy, Ames, IA
The central flavonoid pathway is highly conserved in the plant kingdom. However, functions of the end products
and expression mechanisms of the pathways have diverged during the course of evolution. The comparison of
phlobaphenes (flavan 4-ols) accumulation in maize and sorghum tissues exhibit conservation as well as
divergence. The genes p1 in maize and y1 in sorghum regulate the accumulation of phlobaphenes. The p1 and y1
genes are MYB transcription factors and are highly similar in their coding region. However the 5í and 3í regions
as well as genomic organization of these genes show divergence. In this study we tested the hypothesis that the
difference in the expression pattern of the phlobaphenes accumulation in two related grass species is due to the
divergence of their 5í regulatory sequences. Maize line was transformed with the Y1::GUS construct and
expression of the promoter analyzed. Expression results demonstrating the similarity and differences for tissue
preferred function of the orthologous gene promoter in maize will be presented.
163
A novel screen for single-locus kernel composition mutants.
Settles, A. Mark {1} Pearson, Tom {2}
{1} University of Florida {2} USDA-Grain Marketing and Production Research Center
Visible mutant analysis of maize kernel composition has identified ~25 genes involved in starch biosynthesis and
seed storage protein accumulation. The mutant alleles of these genes have been used to partially decipher these
pathways and develop specialized corn varieties with dramatically increased value. All of these genes are
sensitive to genetic background modifiers, and only a fraction of the biochemically predicted genes that should
influence starch and protein composition have identified mutant phenotypes. These observations suggest that there
are kernel composition loci that have phenotypes undetectable to the human eye. We have used single-kernel near
infrared (NIR) spectroscopy to screen for mutants that affect kernel composition but do not affect the visible
appearance of the kernel. Initially, kernels from visible mutants with composition defects were characterized by
NIR to develop a strategy to identify novel mutants with visibly normal kernels. In a screen of M2 families from
the UniformMu population, we identified putative mutants based solely on NIR spectra. A subset of the mutants
were identified as putatively allelic through tight linkage to the bz1 locus. These mutant isolates share a novel
polymorphism in the sh1 locus suggesting that they have a subtle defect in sucrose synthase. These results suggest
a strategy for identifying weak alleles of known visible mutant loci and strong alleles of unknown kernel
composition modifiers.
164
Assessment of transgenic maize events produced by particle bombardment or Agrobacterium-mediated
transformation
Shou, Huixia {1} Frame,Bronwyn {1} Whitham, Steve {1} Wang, Kan {1}
{1} Department of Agronomy, Iowa State University, Ames, IA 50011
We have previously published a protocol to produce transgenic maize using Agrobacterium-mediated
transformation. In this study, we compared transgenic events produced by particle bombardment or
Agrobacterium-mediate transformation with regard to transgene copy number and RNA expression level using
real-time PCR or real time RT-PCR technology. Our results using this new analysis method correlated well with
those obtained using traditional Southern and Northern analyses. Results showed that the majority of
Agrobacterium-derived transgenic lines contained 1 to 3 copies of the transgene, whereas most of the
bombardment-derived lines contained 10 to 277 copies of the transgene. The average amount of transgene
transcript in Agrobacterium-derived lines was four times more than that in bombardment-derived lines. Transgene
expression in the R2 generation was greater than the R1 progenitor for 77% of the Agrobacterium-derived lines.
In contrast, 89% of the bombardment-derived lines had lower transgene expression in the R2 than in the R1
generation. We conclude that transgene expression is more stable in Agrobacterium-derived transformants than in
bombardment-derived transformants. Interestingly, we observed that more than 70% of Agrobacterium-derived
transgenic lines contained various lengths of the bacterial plasmid backbone DNA sequence, indicating that
Agrobacterium transformation was not as precise as previously perceived using the current binary vector system.
In conclusion, although the transformation efficiency of Agrobacterium-mediated transformation using a standard
binary vector is one-half of particle bombardment in our laboratory, the resulting transformants have fewer
transgene copies, and higher and more stable gene expression than their bombardment-derived counterparts.
165
Determination of the physiologically significant substrate of RF2A in fertility restoration
Skibbe, David {1} Liu, Feng {1} Schnable, Patrick {1}
{1} Iowa State University, Ames, IA 50011
Cytoplasmic male sterility is a maternally inherited inability to produce viable pollen. Texas (T)-cytoplasmic
male sterility can be overcome by dominant alleles of two genes, rf1 and rf2a. The rf2a gene encodes a
mitochondrial aldehyde dehydrogenase (ALDH) whose enzyme activity is required for fertility restoration in T-
cytoplasm. Since RF2A can oxidize many aldehydes, the physiologically significant substrate is not known.
Three additional ALDH genes (rf2b, rf2c, and rf2d) have been identified in maize and are being used to reduce the
number of candidate substrates. A two-pronged approach is being used. First, in vitro kinetic enzyme analyses for
over 25 commercially available aldehydes have been performed on RF2A, RF2B, RF2C, and RF2D. Second,
complementation constructs containing the rf2a, rf2b, rf2c, or rf2d coding region, driven by the rf2a promoter and
targeted to mitochondria with the RF2A mitochondrial targeting sequence, were transformed into maize. If the
transformed complementation construct is able to oxidize the physiologically significant substrate in a (T)
rf2a/rf2a background, then viable pollen will be produced. Fertility observations have revealed that both the Rf2a
(control) and Rf2c constructs, but not the Rf2b and Rf2d constructs, complement the male sterile phenotype. Since
Rf2c can complement the male sterile phenotype, the RF2A and RF2C proteins can both oxidize the
physiologically significant substrate. Comparison of the kinetic data between RF2A and RF2C has identified four
aldehydes that have common kinetic characteristics for both proteins. To determine which of these substrates is
physiologically significant, microarray analyses will be performed on tapetal cells of RF2A and RF2C
complementing plants to identify pathways that are coordinately regulated between these two constructs and that
are therefore likely to be involved in the generation of these aldehydes.
166
Microarray Analysis of Trichostatin A and 5-aza-2-deoxycytidine Treatments in Maize Tissue Culture
Smith, Alan {1} Haun, William {1) Kaeppler, Shawn {1}
{1} Department of Agronomy, University of Wisconsin
Microarrays allow for the simultaneous screening of multiple genes for changes in transcription in response to
experimental conditions. The goal of this experiment was to detect genes whose expression is sensitive to
alterations in DNA methylation or histone acetylation. Maize suspension cultures were treated with either the
DNA methyltransferase inhibitor, 5-aza-2-deoxycytidine, or the histone deacetylase inhibitor, trichostatin A. RNA
was isolated and used to interrogate maize 1-1-03 Unigene arrays using untreated cultures as a control. Data were
analyzed to determine which genes were altered in one or both treatments. The results of this experiment will be
presented.
167
Molecular and genetic characterization of a 'gain of function' Mu insertion P1-wr allele
Smith, Brady {1} Meeley, Robert {2} Chopra, Surinder {1}
{1} Department of Crop & Soil Sciences, Pennsylvania State University, University Park, PA 16802 {2} Pioneer
Hi-Bred International Inc., Johnston, IA 50310
The maize p1 locus encodes a MYB-homologous regulator of red phlobaphene pigment biosynthesis in pericarp,
cob glumes, and other plant tissues. The P1-wr allele of the p1 gene specifies the accumulation of pigment in the
cob, but not the pericarp (Styles and Ceska, Maydica 34: 227-237, 1989). In order to gain a greater understanding
of the tissue specificity associated with this allele, induced mutations were identified using the Trait Utility
System for Corn (TUSC) developed by Pioneer Hi-Bred International, Inc. (Meeley and Briggs, MNL 69: 67-82).
Using several P1-wr specific primers, we have detected 13 unique Mu insertion alleles. One of the insertion
alleles shows ëgain of functioní in pericarp. We tested if the gain of function is the result of Mu suppression, as
has been observed in hcf106 mutation (Barkan and Martienssen, Proc. Natl. Acad. Sci. USA 88: 3502-3506,
1991). Genetic and molecular analysis of this ectopic P1-wr-mum1 expression and pericarp suppression in the w-t
allele will be presented
168
Proteins associated with Spodoptera frugiperda resistance in Zea mays L.
Soto, Jennifer {1} Song, Qisheng {1} Mooney, Brian {1} Thelen, Jay {1} Davis, Georgia {1}
{1} University of Missouri-Columbia
Spodoptera frugiperda, fall armyworm (FAW), can cause significant economic loss in maize. Commercial non-
transgenic maize lines are susceptible to feeding damage. We have identified several quantitative trait loci (QTL)
conferring resistance to whorl-stage feeding damage in maize. Tropical maize provides the source for the
resistance genes. One QTL corresponds to the Glossy15 gene, a regulatory gene that causes abbreviated
expression of the juvenile phase in maize leaves. The remainder of the QTL are unknown. An abbreviated
juvenile phase has been linked to resistance. Protein patterns from juvenile and adult leaf tissue of corn (Zea mays
L.) inbreds that are resistant or susceptible to fall armyworm were analyzed through 2-dimensional
electrophoresis. Tissue was collected from F3 families of a cross between the resistant parent (Mp705) and the
susceptible parent (Oh28). Bulks were made from the 13 most resistant and 13 most susceptible F3 families. Adult
and juvenile tissues were analyzed separately. Both presence/absence and protein intensity differences were
identified. Differently expressed proteins were excised from the two-dimensional gels and analyzed through mass
spectrometry to determine protein identity. The goal of this comparative protein research is to identify putative S.
L.) inbreds that are resistant or susceptible to fall armyworm were analyzed through 2-dimensional
electrophoresis. Tissue was collected from F3 families of a cross between the resistant parent (Mp705) and the
susceptible parent (Oh28). Bulks were made from the 13 most resistant and 13 most susceptible F3 families. Adult
and juvenile tissues were analyzed separately. Both presence/absence and protein intensity differences were
identified. Differently expressed proteins were excised from the two-dimensional gels and analyzed through mass
spectrometry to determine protein identity. The goal of this comparative protein research is to identify putative S.
frugiperda resistance genes in maize corresponding to our QTL. These resistance genes can then be integrated
into elite maize lines in order to reduce; the pest population, cost of pesticide application, and hazards to the
environment.
169
The molecular characterization of the sequences required for b1 paramutation and high expression
Stam, Maike {1} Belele, Christiane {2} Rechien Bader {1} Chandler, Vicki L.{2}
{1} SILS, University of Amsterdam, Kruislaan 318, 1098 SM Amsterdam, The Netherlands {2} University of
Arizona, Tucson, Arizona 85721, USA.
Paramutation is a mitotically and meiotically heritable change in the expression of one allele caused by interaction
with another allele. At b1, a regulatory gene of the maize pigmentation pathway, the high expression of the B-I
allele is downregulated to the lower expression level of the Bí allele, in B'/B-I heterozygous plants. New B' alleles
behave like Bí, they paramutate B-I alleles with a 100% efficiency. The sequences required for Bí paramutation
and B-I expression locate in a 6 kb region ~100 kbp upstream of the b1 coding region, suggesting the involvement
of long range communication between sequences far upstream and the promoter proximal region (Stam et al.,
2002, Genes & Dev. 16:1906; Stam et al., 2002, Genetics 162:197). Within the 6 kb region, Bí and B-I have 7
tandemly repeated copies of a 853 bp sequence which are involved in paramutation. There are no sequence
differences between Bí and B-I in this region, establishing they are epialleles. The Bí and B-I repeats are currently
being compared at the molecular level making use of trans-acting mutations affecting paramutation. Analyses of
the repeat DNA methylation levels using methylation sensitive restriction enzymes revealed a complex pattern of
correlations. Some sites are methylated only in Bí and other sites only in B-I. The methylation of the former sites
is partially relieved in a mop1-1 mutant background, but not in a Mop2-1 background. We speculate that the
differential methylated regions may have different functions. Some sites might be involved in paramutation,
others in enhancing expression. Nuclease accessibility assays revealed that the B-I repeats have a more open
chromatin structure than the Bí repeats. New Bí alleles had an intermediate chromatin state relative to B-I and Bí.
The question which chromatin structure changes are correlated with paramutation is currently being studied in
more detail.
170
Identification and Expression of Flavonoid 3'-Hydroxylase in Maize and Sorghum
Svabek, Catherine {1} Chopra, Surinder {2}
{1} Pennsylvania State University {2} Pennsylvania State University
Flavonoids are secondary metabolites that function in plant growth and development through pollen tube
germination, responses to stress, pigmentation, protection from UV light, and as antimicrobial phytoalexins such
as anthocyanidins. Here we attempt to further elucidate the role of flavonoid-3í-hydroxylase, a cytochrome P450,
and its possible function in the flavan-4-ol pathway which leads to the synthesis of phlobaphene pigments.
Previous research has shown that the pr gene may encode for the F3íH activity required for the conversion of
dihydrokaempherol to dihydroquercetin in the 3-hydroxyanthocyanidin pathway (Larson et al., 1986, Biochem.
Genet. 24:615-624), however, the f3íh gene has not been cloned. Our preliminary attempts indicate that f3íh may
also play a role in the flavonoid 3-deoxyanthocyanidin biosynthetic pathway. Comparisons of available genomic
and proposed amino acid sequences were used as a starting point in isolating f3íh cDNA and genomic clones.
Results from expression studies of f3íh will be presented.
171
MAR-induced variegation
Torney, FranÁois {1} Partier, Anne {1} Says-Lesage, VÈronique {1} Barret, Pierre {1} Beckert, Michel {1}
{1} INRA, UMR Amelioration et SantÈ des plantes, Clermont-Ferrand, France
Chromatin loop organization is based on interactions between DNA matrix attachment regions (MARs) and the
nuclear protein matrix. In this study, we recreated an in vivo pre-existing chromatin loop based on MARs. We
used the 5í and the 3í MARs of maize Adh-1 gene to bracket a Gus expression cassette. Unexpected quantitative
results, obtained in various tissues of fully regenerated plants, led us to check the transgene expression pattern.
Surprisingly, maize Adh-1 MARs induced a systematic variegated expression pattern in maize roots. To our
knowledge this constitute the first report of intraspecific MAR induced variegation. Furthermore, we found that
tissue and developmental specificity were superimposed to variegation. Our results reveal new aspects of MAR
influence gene regulation.
tissue and developmental specificity were superimposed to variegation. Our results reveal new aspects of MAR
influence gene regulation.
172
LOCALIZATION OF A BACTERIAL PROTEIN IN STARCH GRANULES OF TRANSGENIC MAIZE
KERNELS
Chikwamba, Rachel {1} Scott, Paul {2} Mason, Hugh {3} Wang, Kan {1}
{1} Iowa State University {2} USDA-ARS {3} Arizona State University
The B subunit of Escherichia coli heat labile enterotoxin (LT-B) is a potent oral immunogen, with potential for
use as a vaccine and an adjuvant to co-administered vaccine. LT-B has been used as a model antigen to
demonstrate the feasibility of producing effective oral vaccines in transgenic plants. We expressed a synthetic
gene encoding the LT-B subunit in transgenic maize kernels. Using immunogold labeling /electronmicroscopy
and cell fractionation/western analysis, we demonstrate that the LT-B protein, when produced with either its
native bacterial signal peptide (BSP) or the signal peptide from the maize 27 kDa gamma-zein, accumulates in
starch granules of transgenic maize kernels. These observations suggest that the BSP in is not required for
localization of LT-B to starch granules, and that starch localization is a feature of the mature protein. Analysis of
ground kernel endosperm by ELISA and Western analysis also indicates the presence of LT-B in the soluble
(extra-granular) fraction. The functional properties of starch encapsulated LT-B are yet to be determined. The
unexpected localization of LT-B in starch granules suggests the existence of a novel amyloplast targeting
mechanism. This targeting property is a potentially valuable cell biology tool for further elucidation of protein
translocation into the amyloplast. Furthermore, LT-B could be used to direct other proteins to the amylopast,
conferring stability to the conjugated proteins, a feature that is desirable in edible vaccines.
173
Targeting Induced Limited Lesions IN Genomes (TILLING) for Maize: Reverse Genetic Analysis of Point
Mutations
Weil, Cliff {1} Till, Bradley {2} Comai, Luca {3} Henikoff, Steve {2}
{1} Purdue University {2} Fred Hutchison Cancer Research Center {3} University of Washington
TILLING is a reverse genetic method that combines chemical mutagenesis, typically EMS, and PCR-based
screening in gene regions of interest. The approach complements other reverse genetic methods and provides a
range of allele types, including missense and knock-out mutations, potentially useful in gene function and
interaction studies. The Arabidopsis TILLING Project (ATP) in Seattle has developed into a, low-cost, high-
throughput operation for plant functional genomics and we are establishing a maize TILLING venue modeled
after it. ATP has developed, interactive, web-based bioinformatics tools that allow users to identify gene regions
most suitable for TILLING, design optimal PCR primers for mutational screening, and to map, analyze and
evaluate the mutations that are discovered. Preliminary data indicate that standard EMS pollen mutagenesis
protocols provide sufficient mutation density to carry out efficient TILLING in maize. Mutagenizing both the
W22 and B73 inbreds, we have set a target of developing ~10,000 mutant lines for each inbred. M1 leaf tissue is
used for DNA analysis and M2 seed will then be distributed to users. Gene segments will be screened, mutations
identified and sequenced and seed stocks for identified mutants made available through the Maize Genetics Stock
Center. TILLING technology has potential benefits for agriculture, where methods are needed for crop
improvements that avoid the expensive regulatory process required for transgenics and that are widely acceptable
to consumers. We provide an overview of the methodology and bioinformatics tools, and we present preliminary
data and our progress to date.
174
The origins of sweet corn
Whitt, Sherry {1} Tracy, William {2} Buckler, Edward IV {1}
{1} USDA-ARS, North Carlina State University {2} University of Wisconsin-Madison
The first mention of US sweet corn, referred to as ìshriveled cornî in the literature is from Thomas Jefferson in
1810. By 1909, Correns had characterized the sugary1 locus, and much later is was discovered that this locus
produces a starch debranching enzyme of the isoamylase type (James et al, 1995), responsible for building starch
molecules in concert with the starch branching enzymes. A mutation at this locus is responsible for producing
kernels that contain a high concentration of phytoglycogen and taste sweet. The phenotype appears as a wrinkled,
translucent kernel that we recognize as normal commercial sweet corn. The data presented here clarify the origin
of sweet corn, for which there were two competing hypotheses. One theory argues a single origin from a Peruvian
race (Chullpi) (Manglesdorf, 1974), whereas others propose independent origins from recurring mutations (Erwin,
1951; Tracy, 2001). Morphological and molecular analyses have confirmed 4 and possibly 5 unique, independent
mutations responsible for the sugary phenotype in 64 sweet corn accessions from North America and South
America. Three of the mutations are single amino acid changes at residues that are highly conserved across plants,
Archaebacteria and Eubacteria. The fourth mutation is the result of a transposable element in exon 1, which
disrupts translation. The fifth putative mutation, present in accessions from Peru, is most likely a modification of
the 5í end of the gene, but further investigation is required. There is strong geographic patterning to these
mutations. The abundance of single amino acid changes responsible for several of the sweet phenotype mutations
is quite surprising and may reflect constraints on the su1 locus for germination.
America. Three of the mutations are single amino acid changes at residues that are highly conserved across plants,
Archaebacteria and Eubacteria. The fourth mutation is the result of a transposable element in exon 1, which
disrupts translation. The fifth putative mutation, present in accessions from Peru, is most likely a modification of
the 5í end of the gene, but further investigation is required. There is strong geographic patterning to these
mutations. The abundance of single amino acid changes responsible for several of the sweet phenotype mutations
is quite surprising and may reflect constraints on the su1 locus for germination.
175
Genotype-specific trans-acting Factors Influence Meiotic Recombination in the 140-kb a1-sh2 Interval and
Elsewhere in the Genome
Yandeau, Marna D. {1} Dunham, Timothy {1} Tenhundfeld, John {1} Toomer, Lakeysha {1} Smith,
Heather {1} Nikolau, Basil J. {1} Schnable, Patrick S. {1}
{1} Iowa State University
Rates of meiotic recombination can be affected by trans-acting factors: genotype-specific modifiers not residing
in the interval of interest. Such trans factors were previously shown to affect rates of recombination in an
identical 140-kb a1-sh2 interval of chromosome 3 that was introgressed into the A632, W64A and Oh43 inbred
backgrounds. Rates of recombination were increased in the A632 and W64A inbred backgrounds two-fold and
high-resolution recombination break-point mapping showed that trans-modifiers in these backgrounds increased
rates of recombination at existing hotspots. To test the hypothesis that patterns of breakpoint distribution differ
within a hotspot, fine-structure mapping of the distribution of breakpoints within the yz1 gene was conducted.
These analyses revealed different patterns of breakpoint resolution among the inbred backgrounds. These
findings demonstrate that trans-factors can affect both the rates and distributions of meiotic recombination in the
a1 sh2 interval. To test the hypothesis that trans-acting factors in these inbred backgrounds can also affect rates
of recombination in other regions of the genome, the genetic distances associated with a genomic interval that is
not located on chromosome 3 were determined in each of the three inbred backgrounds. This genomic interval
has a significantly greater genetic distance in the A632 background as compared to the other inbred backgrounds.
176
Molecular analysis of the maize P1-rw allele
Zhang, Feng {1} Peterson, Thomas {1}
{1} Iowa State University
The maize p1 gene encodes a Myb-like transcription factor which regulates the synthesis of a phlobaphene-like
red pigment in kernel pericarp, cob and other floral organs. Some p1 alleles elicit differential pigmentation of
pericarp and cob: for example, P1-rr, P1-wr, and P1-rw specify red pericarp/red cob, white pericarp/red cod, and
red pericarp/white cob, respectively. These alleles provide excellent resources to study the mechanism of organ-
specific gene expression. The P1-rr and P1-wr alleles have been previously characterized; we are now
investigating the gene structure and molecular expression pattern of the P1-rw allele. Genomic Southern analysis
indicates that, similar to P1-rr, P1-rw contains a single copy gene flanked by long direct repeats. RT-PCR results
showed that the P1-rw 5°Ø UTR and 5°Ø coding sequences are identical to P1-rr and P1-wr. However, the P1-rw
3°Ø coding sequences and 3°Ø UTR are identical to that of the p2 gene, a tightly linked, paralogous gene. To
reconcile these results, we isolated and sequenced two overlapping P1-rw genomic clones. Interestingly, although
P1-rw is structurally similar to that of P1-rr (both contain a coding sequence flanked by two long direct repeats),
P1-rw differs from P1-rr in two major characteristics: First, the P1-rw coding sequence is chimeric, and consists
of a p1-like 5°Ø UTR, followed by p2-like exons and introns, followed by a truncated P1-wr-like exon 3. This
chimeric structure appears to have been generated by recombination between the p1 and p2 genes. Second, P1-rw
has a major polymorphism in the 1.2 kb distal enhancer region which is located 5 kbp upstream of the p1
transcription start site. Our current hypothesis is that polymorphisms in the coding sequences and/or the distal
enhancer region are responsible for the distinct gene expression pattern of P1-rw. Nuclear run-on assay is
underway to test whether P1-rw expression is regulated at the transcriptional or post-transcriptional level, or both.
177
Nutritionally Improved Transgenic Sorghum
Zhao, Zuo-yu {1} Glassman, Kimberly {1} Sewalt, Vincent {1} Wang, Ning {1} Miller, Mike {1} Chang,
Shawn {1} Thompson, Teresa {1} Catron, Sally {1} Wu, Emily {1} Bidney, Dennis{1} Jung, Rudolf {1}
{1} Trait and Technology Development, Pioneer Hi-Bred International, Inc., DuPont Agriculture and Nutrition,
Johnston, IA 50131
Sorghum is the dietary staple food to over half a billion people in developing countries. However, sorghum grain
is low in protein quality due to its low contents of essential amino acids, such as lysine. The reliance on sorghum
results therefore in malnutrition, especially of children, in many countries in Africa and Asia. Advancements in
sorghum tissue culture and transformation research in Pioneer have led to the development of the first efficient
method to genetically transform sorghum by Agrobacterium. With the objective to use this novel technology to
improve the nutritional quality of its grain, we introduced into sorghum a high-lysine gene (HT12) that we
previously found to be efficacious when transgenically expressed in corn. Sorghum was transformed using a
ìsuper-binaryî Agrobacterium vector containing two unlinked T-DNA cassettes. One of these cassettes contained
the HT12 gene and the other contained a herbicide-resistance gene as a selectable marker. In progeny of the
sorghum tissue culture and transformation research in Pioneer have led to the development of the first efficient
method to genetically transform sorghum by Agrobacterium. With the objective to use this novel technology to
improve the nutritional quality of its grain, we introduced into sorghum a high-lysine gene (HT12) that we
previously found to be efficacious when transgenically expressed in corn. Sorghum was transformed using a
ìsuper-binaryî Agrobacterium vector containing two unlinked T-DNA cassettes. One of these cassettes contained
the HT12 gene and the other contained a herbicide-resistance gene as a selectable marker. In progeny of the
primary transformants, this potentially permits the segregation of the marker gene from the trait gene. The
elimination of the marker gene is important, because sorghum can cross-breed with wild relatives and herbicide-
resistant plants can not be released into the environment. Of four independent transgenic events obtained, three
expressed high levels of the HT12 protein in the grain. Several plants from these events were selfed and the
progeny was analyzed for segregation of the herbicide marker and the HT12 gene. Preliminary data demonstrate
the successful elimination of the herbicide-resistance gene from one event. Heterozygous seed from this event
showed a 40-60% increase in lysine. A transgenic, lysine-enriched sorghum variety has the potential to directly
benefit large populations in developing countries. Assistance in the development and in the release of such a
variety would therefore demonstrate to a wide public the possibility of biotechnological solutions to important
agricultural and nutritional challenges.

Quantitative Traits

178
Identification of QTLs controlling Ustilago maydis resistance in two populations of recombinant inbred
lines
Baumgarten, Andrew {1} Suresh, Jayanti {2} May, Georgiana {3} Phillips, Ronald L. {2}
{1} Department of Plant Biology, University of Minnesota {2} Department of Agronomy and Plant Genetics,
University of Minnesota {3} Department of Ecology, Evolution, and Animal Behavior, University of Minnesota
We report the identification of QTLs contributing to Ustilago maydis (corn smut) resistance in two populations of
recombinant inbred lines of maize. The two RI populations were generated by crossing a susceptible inbred
(CMV3) to two inbred lines (A188 and W23). During the summers of 2000-2002, the frequency and severity of
U. maydis infection on these RI lines was evaluated in St. Paul, MN and Waseca, MN. We found that resistance to
U. maydis infection was highly heritable (h2 = 0.65-0.84) in both populations of RI lines and several lines
segregated for U. maydis resistance in specific tissues. 100 previously mapped SSR markers were surveyed across
RI lines from both populations and examined for association with U. maydis infection. Several QTLs for U.
maydis resistance were discovered, but significant associations were found in all replicate blocks between markers
mapped to chromosome 2, 4, and 6. Some QTLs were observed only in specific environments, indicating that
genetic by environment interactions may be influencing the U. maydis resistance of several RI lines. Similar to
other studies mapping QTLs for pathogen resistance, QTLs for U. maydis resistance mapped to chromosome
regions containing resistance gene homologs, defense-related genes, and QTLs for pathogen resistance.
179
Association mapping of starch candidate genes with kernel composition and starch viscosity traits
Wilson, LM {1} Whitt, SR {2} Ibanez, AM {3} Rocheford, TR {4} Goodman, MM {1} Doebley, J {5}
Buckler, ES {1,2}
{1} North Carolina State University {2} USDA/ARS {3} University of California, Davis {4} University of
Illinois at Urbana-Champaign {5} University of Wisconsin-Madison
Starch production in the maize kernel is a complex trait controlled by multiple genes. Depending on the specific
needs of production, maize breeders select for kernel composition traits, such as the amount of starch produced in
the kernel. An equally important aspect of maize kernel quality and of great interest for food processing
applications is the pasting and viscosity profile of the starch produced. This study used 97 diverse maize lines in
an association mapping approach to locate DNA polymorphisms within six candidate genes that had an effect on
kernel composition traits (starch, oil, protein) or starch pasting/viscosity traits. Starch candidate genes Ae1, Bt2,
Sh1, Sh2, Su1, and Wx1 were amplified, sequenced, and tested for associations using the software package,
TASSEL. Principle component factor analysis was used to reduce multiple testing and to examine underlying
factors contributing to the multidimensional nature of kernel physicochemical properties. Overall gene analyses
indicated significant results for kernel composition traits from Sh1, Sh2, and Bt2. Significant associations for
starch pasting/viscosity traits were seen in Ae1 and Sh2. In order to assess the effects of Sh2 on starch traits,
haplotype analysis of a 500 bp region was performed. Although a significant effect on amylose was seen in the
winter field season (P < 0.04), it was not significant in the summer field season (P = 0.20). However, Sh2 showed
a significant genotype x environment effect on kernel composition, therefore we propose the hypothesis that with
higher ambient temperatures, the heat labile nature of ADP-glucose pyrophosphorylase (Green and Hannah,
1998), in which Sh2 encodes a subunit, varies between Sh2 haplotypes and has an epistatic effect on amylose
production.
higher ambient temperatures, the heat labile nature of ADP-glucose pyrophosphorylase (Green and Hannah,
1998), in which Sh2 encodes a subunit, varies between Sh2 haplotypes and has an epistatic effect on amylose
production.
180
A large region upstream to the teosinte branched 1 (tb1) locus was selected during maize evolution
Clark, Richard {1} Linton, Eric {2} Messing, Joachim {2} Doebley, John {1}
{1} University of Wisconsin, Madison, Wisconsin 53706, USA {2} Waksman Institute, Rutgers University,
Piscataway, New Jersey 08854, USA
Quantitative trait locus (QTL) mapping studies have suggested that a major locus, teosinte branched 1 (tb1), is
responsible for many of the morphological differences between maize and its wild ancestor (teosinte). The gene
for tb1 was cloned by transposon tagging and is predicted to encode a transcriptional regulator. Levels of tb1
RNA are two fold higher in maize as compared to teosinte, and sequence evolution studies have demonstrated that
the tb1 promoter region, but not the coding region, shows a significant signature of selection. These results
suggest that differences in tb1 regulation were the target of selection during maize domestication from teosinte.
Nonetheless, the molecular changes that were selected at tb1 are not known. To better understand the nature of
selection at tb1, we have generated 220 kb of contiguous genomic sequence around the tb1 locus, and assessed
sequence diversity at sites 5' to the tb1 coding sequence. Sequence analysis has revealed that a large 162 kb region
between tb1 and the next 5' gene contains only transposable elements and non-genic unique sequences. Nucleotide
polymorphism analysis at sites throughout this intergenic region has revealed low sequence diversity (a sign of
selection) among maize lines from the tb1 promoter region extending to at least 58 kb upstream. In contrast, at 93
kb upstream to tb1, as well as at the next gene 5' to tb1, sequence diversity in maize is nearly that observed at
neutral sites in the maize genome. Sequence diversity throughout the region in teosinte is high. These results
suggest that a large region of 58 to 93 kb was selected during maize domestication and that this region is likely to
harbor sequences responsible for the derived maize morphology. The large size of the selected region may reflect
the fact that multiple polymorphisms in the region are required for maize morphology. Alternatively, low
recombination rates upstream to tb1 may have contributed to maintenance of an extensive region of low sequence
diversity following selection at a given site during domestication. The pattern and extent of linkage disequilibrium
in the region will also be discussed.
181
Genetic Control of the Number of Ears per Plant and Related Morphological Traits in the Golden Glow
Maize Population
de Leon, Natalia {1,2} Coors, James G. {1} Kaeppler, Shawn M. {1} Rosa, Guilherme J. M. {2}
{1} University of Wisconsin {2} Michigan State University
Prolificacy is a natural characteristic in maize. Humans, however, have tended to select for the concentration of
resources in a single large ear borne at the apex of the primary lateral branches. Since 1971, bi-parental mass
selection for increased number of ears per plant has been carried out in the open-pollinated maize population
Golden Glow at the University of Wisconsin - Madison. To better characterize the changes in morphology
accompanying selection for prolificacy in Golden Glow, a mapping population has been developed from the cross
of inbred A679 with a highly-prolific plant from cycle 23 of Golden Glow. The objective of this study was to
determine the genetic basis of prolificacy and related traits and to determine whether the genetic control of
prolificacy was related to genetic events associated with the domestication of maize. Significant phenotypic
variation was observed for most traits analyzed in this experiment. Twenty-four QTL were found for the nine
traits described here. In a number of cases, regions of the genome controlled more than one morphological trait at
a time. The number of significant associations between molecular markers and a particular trait, as well as the
total percentage of phenotypic variance accounted for by each QTL, varied significantly from trait to trait. The
largest QTL found in this study explained 36.5% of the phenotypic variance for the frequency of tassels at the end
of tillers. Eleven of the 24 associations described here involved linkage group 1. A particularly interesting region
was located on bin 1.09 of the mapping population. This region shared many similarities with the pattern of
development suggested for the teosinte branched1 mutant. This indicates that the genetic control of prolificacy
and associated traits in Golden Glow resembles the process of maize domestication at least in regard to its
branching morphology.
182
Selection for Methionine and Tryptophan Content in Maize
Darrigues, Audrey {1} Scott, M. Paul {2} Lamkey, Kendall R. {1}
{1} Iowa State University, Ames, Iowa {2} USDA-ARS, Ames, Iowa
Some essential amino acids in maize are nutritionally limiting for animal diets. In plant breeding, selection to
improve a population while conserving genetic variation within the population is referred to as recurrent selection.
The potential of this method to improve maize populations for their composition in two amino acids such as
Methionine (Met) and Tryptophan (Trp) was tested. The objectives of this study are (1) to determine the effect of
selection on the amino acid composition in maize, (2) to conduct divergent selection for both Trp and Met, and (3)
to determine the effect of selection on the nutritional value of the populations. Initially two random-mated maize
populations, BS11 and BS31, were chosen for this study on the basis of their protein content and variability. The
improve a population while conserving genetic variation within the population is referred to as recurrent selection.
The potential of this method to improve maize populations for their composition in two amino acids such as
Methionine (Met) and Tryptophan (Trp) was tested. The objectives of this study are (1) to determine the effect of
selection on the amino acid composition in maize, (2) to conduct divergent selection for both Trp and Met, and (3)
to determine the effect of selection on the nutritional value of the populations. Initially two random-mated maize
populations, BS11 and BS31, were chosen for this study on the basis of their protein content and variability. The
two populations were treated independently and selections were categorized for high Met, low Met, high Trp, and
low Trp within each population in Cycle 0 in 2001. The first cycle of half-sib family selection for Met and Trp
was completed in the summer of 2002. For both populations in Cycle 1, the values for the high Met and Trp
selections were significantly higher than the values for the low Met and Trp selections. These data suggest that a
divergent selection program is a useful method for producing populations with improved Met and Trp content.
These populations will be a valuable resource for researchers interested in the genetics and biochemistry of kernel
amino acid content.
183
Candidate Gene Association of Kernel Composition in Diverse Maize Inbreds
Flint-Garcia, Sherry {1} Buckler, Edward, IV {2}
{1} Department of Genetics, North Carolina State University {2} USDA-ARS
Kernel quality in maize is an important quantitative trait that has been studied extensively and many QTL
have been identified for starch, protein, and oil content. Candidate gene association mapping, also known as
linkage disequilibrium mapping, can be used to dissect QTL on a finer scale than traditional QTL mapping.
Previously, association studies have identified that starch content and quality associate with polymorphisms in
genes late in the starch pathway.
Our lab is also interested in earlier events that lead up to kernel development and starch biosynthesis.
Lachrima1 (lac1 ) is a defective kernel mutant (dek34 ) with an unknown function, brittle1 (bt1 ) is an amyloplast
adenylate translocator, and dull endosperm1 (du1 ) is a soluble starch synthase. Regions of these genes were
sequenced in 32 inbred lines chosen for diversity in kernel composition and analyzed for association with kernel
composition traits, i.e. kernel protein, oil, starch, and amylose.
Several polymorphisms in each gene have been identified that significantly associate with kernel composition.
Efforts are currently underway to sample these polymorphisms in a larger set of 102 inbred lines in the hope of
identifying the causative site. We have also initiated a study of the invertase gene family, which may be involved
in assimilate translocation as their gene products cleave sucrose into glucose and fructose.
184
The genetic and molecular basis of cell wall digestibility in silage maize
Fontaine, Anne-Sophie {1,2} Vermerris, Wilfred {1,3} BarriËre, Yves {2}
{1} Department of Agronomy, Purdue University, West Lafayette, Indiana 47907, USA {2} UnitÈ de GÈnÈtique
et d'AmÈlioration des Plantes FourragËres, INRA, Lusignan 86600, France {3} Department of Agricultural &
Biological Engineering, Purdue University, West Lafayette, IN 47907, USA
Maize feeding value is mainly related to the digestibility of the vegetative parts, and especially of cell wall
components. Lignin is the cell wall biopolymer hindering forage digestibility the most. Several lignification
parameters influence silage maize digestibility: lignin content and composition, and chemical cross-linking with
other cell wall components. In order to understand relationships between cell wall lignification and digestibility,
we have performed QTL analyses of cell wall lignification traits, and investigated allelic polymorphisms of a
candidate gene. The analysis of genetic variation in cell wall traits was performed on 25 lines covering a wide
range of origins and digestibility. Traits related to cell wall digestibility were measured. QTL for some of these
traits were identified in a Recombinant Inbred Line population derived from F288 x F271, and compared with
QTL for digestibility. QTL for p-coumaric acid and ferulic acid were different. Digestibility QTL co-localized
with QTL of lignin and phenolic acid content. Analysis of allelic variation of the gene coding for caffeic acid O-
methyltransferase (COMT) was performed. This gene was selected because mutations in this gene result in the
brown midrib3 phenotype, which is associated with changes in cell wall composition and a major improvement in
digestibility. Seven haplotypes of this gene were characterized. The high degree of polymorphism was partly due
to several recombination events. It supports the idea of screening for favorable alleles to achieve higher
digestibility. Screening of allelic variation of COMT should be performed in a larger germplasm collection. This
will provide a database for the definition of relationships between allele polymorphism and digestibility, which is
helpful for the identification of favorable alleles. They can later be introgressed in breeding material. The various
parts of the genome where QTL were detected will lead to the identification of other candidate genes suitable for
allelic variation study.
185
Whole-Genome Mapping of Maize UV Responses
Fu, Yibing {1} Blanding, Carletha R. {1} Stapleton, Ann E. {1}
{1} Department of Biological Sciences, University of North Carolina at Wilmington
Understanding of the mechanisms plants use for adaption to ultraviolet radiation stress is key in predicting plant
responses to our changing environment. Study of UV perception has been complicated by uncertainty about
whether a particular response is due to damage/inhibition or due to signal transduction. Many different UV
responses have been measured (growth, shape change, photosynthetic parameters, etc.) but these studies generally
suffer from difficulty in determining the underlying mechanism. We are analyzing UV responses at multiple
levels, from gene expression to morphological change. We have carried out the first round of genome wide-scans
to identify and map the most important UV response control loci in the IBM mapping population. We have begun
comparisons of the QTL pattern of UV responses; there are some QTL which are shared among a subset of
responses and others which are response-specific. Statistical analysis of the number and location of the QTL will
allow us to determine if there is a ìmaster switchî that controls all UV responses, or if there is a set of separate
switches for each different kind of response. Mapping of control loci, and examination of map patterns, will allow
integration of a great deal of data that would otherwise exist in isolation.
186
Identification of QTL Responsible for Root Architecture
Gerau, Michael {1} Musket, Theresa {1} Dierking, Ryan {1} Kea, Molly {1} Cordia, John {1} Weston,
Patrick {1} Davis, Georgia {1}
{1} University of Missouri-Columbia
Drought resistance is the single most limiting factor to crop productivity worldwide. Many factors are associated
with drought response including anthesis-silking interval, leaf surface area and root architecture. Our objective is
to identify the QTL for root architecture in Zea mays and investigate their role in drought response. Two
experiments were conducted. Ten lines with different drought tolerance levels were evaluated in two reps in a
randomized complete block design in the greenhouse. After two weeks of growth taproot length, overall root
branching, number of primary branches, root mass, shoot mass and number of leaves were measured. A subset of
94 mapping lines from the IBM population were planted in five reps and the same traits measured. Genotypic data
for 251 markers , evenly distributed throughout the genome, were used for QTL analysis. The genetic map was
constructed with Mapmaker Exp version 3.0 for Unix. QTL analysis was performed using QTL Cartographer
Version 1.16. QTL positions were compared to QTL identified in prior drought tolerance experiments.
187
Identification and characterization of Al tolerance genes in the Intermated B73 x Mo17 (IBM) population
by quantitative trait locus mapping
Hoekenga, Owen {1} Mason, Paul {1} Buckler, Ed {2} Shaff, Jon {1} Kochian, Leon {3}
{1} Cornell University, Department of Plant Biology {2} USDA-ARS/North Carolina State University,
Department of Genetics {3} USDA-ARS/Cornell University, Department of Plant Biology
Aluminum (Al) is the third most abundant element in the earthís crust and toxic to plant roots if solubilized at
acidic pH. Aluminum toxicity, which results from inhibition of root growth and nutrient/water acquisition, is a
serious limitation to crop production, as up to one-half of the worldís potentially arable land is acidic. Aluminum
toxicity is one of the primary soil constraints for Latin America and Sub-Saharan Africa, where maize is the
primary staple crop. Increasing plant aluminum tolerance has been an objective of plant breeders worldwide for
many years, yet no tolerance genes have yet been cloned. The physiological mechanisms that underlie aluminum
tolerance are somewhat better understood. The best-characterized mechanism involves an aluminum-activated
release of low molecular weight di- and tri-carboxylic organic acids (e.g. malate, oxalate, citrate) that can bind
soluble Al in the rhizosphere and prevent its uptake into the root. Other mechanisms (e.g. alteration in rhizosphere
pH, release of phenolics, ameliorating Al-induced free radicals) quite possibly also can be contributing to Al
tolerance in some plant species, but their presence have yet to be confirmed. We initiated a quantitative trait loci
analysis of aluminum tolerance in maize, using the IBM population as our subject, with the intention of
identifying the genes responsible for tolerance to aluminum stress and the underlying physiological mechanisms.
The IBM population is especially attractive, as B73 and Mo17 appear to utilize different mechanisms to create
similar levels of tolerance. Transgressive segregation has generated varieties with tolerance similar to that seen in
elite aluminum tolerant materials developed by EMBRAPA Maize and Sorghum, far more tolerant than either
parent. Using a preliminary dataset, we identified five QTL regions important for aluminum tolerance and explain
48% of the variance observed; Mo17 contributes the superior allele for four of the five loci. Interestingly, these
four loci appear to act epistatically with one another suggesting they function in a common pathway. Comparative
mapping of QTLs between our study and one from EMBRAPA Maize and Sorghum will also be discussed.
188
Searching for Virus Resistance Genes-is mdm1 a Universal Virus Resistance Gene?
Jones, Mark {1} Louie, Ray {1} Redinbaugh, Margaret {1}{2} Todd, Jane {1}
{1}USDA-ARS {2} The Ohio State University
Previously, a gene controlling resistance to Maize dwarf mosaic virus strain A (MDMV) was identified in the
inbred line Pa405. The gene was closely linked to the RFLP marker umc85 on the short arm of chromosome 6.
This gene also conferred resistance to two other viruses in the family Potyviridae: Sugarcane mosaic virus
(SCMV/MDM-B) and Wheat streak mosaic virus (WSMV). To determine whether resistance to these viruses was
linked to the gene on chromosome 6S in a diverse array of maize germplasm, we identified 42 MDMV-resistant
inbred lines, then crossed these lines to three MDMV susceptible inbreds and developed F2 populations. The F2
populations were screened with MDMV and leaf tissue from susceptible and resistant pools was screened with
RFLP and SSR markers. In 40 of the 42 inbreds, chromosome 6S markers associated with the resistant parent
were found in the resistant, but not the susceptible pooled tissue. In addition, markers associated with
SCMV/MDM-B and WSMV resistance on chromosome 3 and WSMV resistance on chromosome10 were found
only in the pooled resistant tissue for 12 and 7 inbreds, respectively. These data suggest that mdm1 is associated
with MDMV resistance in most germplasm, but that other loci may also affect resistance. In addition to resistance
genes, rating date and environmental factors influenced the development of symptoms and phenotypic
classifications. Consequently it was not possible to determine the type of gene action.
189
Molecular evolution of genes encoding AGPase in endosperm of Zea mays
Manicacci, Domenica {1} Falque, Matthieu {1} Le Guillou, Sandrine {2} Arar, Chantal {3} Fourmann,
Marie {1} Rousselet, AgnËs {1}
{1} UMR INRA-UPS-CNRS-INAPG du Moulon, Gif sur Yvette, France {2} INRA, Jouy en Josas, France {3}
Rhobio, Evry, France
Sh2 and Bt2 genes encode for subunits of a major enzyme in endosperm starch biosynthesis pathway in Zea mays,
ADP Glucose Pyrophosphorylase (AGPase). We study the molecular evolution of both genes in wild and
cultivated forms, including different subspecies of teosinte, landraces and inbred modern lines. For Sh2, natural
selection, acting from well before domestication, has strongly reduced nucleotide diversity among both wild and
cultivated forms. Similar haplotypes are found among maize and teosinte. For Bt2, the severe reduction of
nucleotide diversity during maize domestication and the phylogenetic reconstruction suggest that Bt2 is involved
in the domestication syndrome. Although these two genes encode for interacting subunits of the same enzyme and
are both under strong selective pressure, they show independent evolutionary histories, leading to reduced genetic
diversity and no evidence of coevolution at the molecular level.
190
Marker Assisted Selection for Resistance to Aspergillus flavus and Aflatoxin production in Maize.
Paul, Chandra (1) White, Don (1) Rocheford, Torbert (1)
(1) University of Illinois
Genetic studies were performed on Aspergillus ear rot and aflatoxin production in (Tex6xB73)xFRB73 BC1S1
and Tex6xFRB73 F2:3 mapping populations derived from maize inbred Tex6, associated with relatively low
levels of aflatoxin production. Two major QTL regions for low aflatoxin were identified on chromosomes 5 and
10 in Tex6, explaining 41% of total phenotypic variation. Two largest QTLs for resistance were identified in
Mp313E chromosome 4L in Mississippi (Davis et al., 1999). Our objective is to pyramid the two different
chromosomal regions of resistance genes in Tex6 and 4L chromosome regions in Mp313E into a commercially
used susceptible inbred line FR1064 background through marker assisted selection. Different stages of back
crossing onto FRB73 with the various combinations of significant genomic regions for resistance from Tex6, and
FR1064 with the significant fraction of resistance from Mp313E were genotyped. The homozygous families for
the significant chromosomal regions associated with the resistance in Tex6 and Mp313E, were identified. The
initial preliminary analyses of some homozygous lines for Mp313E 4L segment in FR1064 showed that the level
of aflatoxin is in the range of 150-200ppb as opposed to FR1064, in which the range is 700-800ppb. Selected
homozygous lines for 4L from Mp313E and chromosome regions 5 and 10 from Tex6 were crossed; (Mp313E x
FR1064)FR1064 BC2S1 with (Tex6 x FRB73)FRB73 BC2S1 to produce F1s. Segregating progenies for these
regions in F2 were genotyped in summer, 2002. Plants were selected for either homozygous or heterozygous for
the significant regions from both resistant parents and then selfed. Six families were identified and individual
plants were genotyped in the winter green house. A few plants were identified that were homozygous for all
chromosome segments, while few more plants were identified with different combinations of major regions.
These advanced lines will be selfed, genotyped further to pyramid homozygous resistance genes from different
genetic sources in multiple lineages. Lines will be evaluated for resistance recovery and yield performance in the
subsequent years.
genetic sources in multiple lineages. Lines will be evaluated for resistance recovery and yield performance in the
subsequent years.
191
Xenia effect on phosphorus concentration in outcrossed maize seed
Pletsch-Rivera, Laura A. {1} Kaeppler, Shawn M. {1}
{1} University of Wisconsin
Xenia effects have been reported for seed weight and oil concentration in maize seed. However, the effect of
foreign pollen on seed mineral nutrient concentration have not been studied. The objective of this research was to
determine if foreign pollen alters seed phosphorus (P) concentration. Fifteen maize inbreds were crossed onto two
hybrids, B73xMo17 and Oh43xW64A, over three years. Seed weight and P concentration were measured and
used to calculate seed P uptake. A xenia effect was found for P uptake with some inbred crosses resulting in
increased P uptake and some in decreased P uptake compared to the selfed hybrid control. A significant xenia
effect was also found for seed size. Although a difference in P uptake was found, the difference is small enough
that open pollinated maize populations can be used for studies measuring P content of seed.
192
The impact of farmer management practices on maize landrace genetic diversity
Pressoir, Gael {1,2} Berthaud, Julien {1,2}
{1} International Maize an Wheat Improvement Center, CIMMYT {2} Institut de Recherche pour le
Developpement, IRD
We report on the impact of traditional farmer management of maize populations in structuring molecular and
phenotypic diversity and on the population dynamics of maize landraces. These populations, from a sample of
local landraces cultivated by farmers in six villages in the Central Valleys of Oaxaca, Mexico, show small among-
population differentiation (Fst=0.011). Most surprisingly, there is no isolation by distance and small among-
village differentiation (Fst=0.003). One would expect populations to fit Hardy-Weinberg equilibrium, but
significant homozygote excess (Fis=0.13) was found. This homozygote excess shows remarkable interpopulation
and interloci differences. We show that this pattern is related to the variation for flowering traits of a given
population. A short anthesis-silking interval and high level of heterogeneity in flowering precocity will favor
assortative mating. This leads to a loci dependent population substructure giving an unusual case of Wahlund
effect and inbreeding while high levels of seed exchange among farmers prevent population differentiation at both
village and regional levels. In spite of considerable gene-flow, high levels of population differentiation for
quantitative traits were observed (i.e., an among-village Qst value of 0.535 for kernel weight). We show that
phenotypic diversification is the result of farmer divergent selection in spite of predominant, but not strict,
conservative selection. Furthermore, we characterized nonproportional changes in the G matrix structure both
within and among villages that are consequences of farmer selection. As a consequence of these differences in the
G matrix structure, the response to multivariate selection will be different from one population to another. Large
changes in the G matrix structure could indicate that farmers select for genes of major and pleiotropic effect.
Farmersí decision and selection strategies have a great impact on phenotypic diversification in maize landraces.
193
Allele Frequency Changes of Oil Candidate Genes in the High Oil Lines ASK Cycle0 ñ Cycle 28: Selection
or Drift?
Ching, Ada {1} Williams, Mark {1} Jensen, Nicole {2} Winkler, Chris {2} Cooper, Mark {2} Smith, Howie
{2} Rafalski, Antoni {1}
{1} DuPont Crop Genetics, Molecular Genetics, Newark, DE19711 {2} PioneerHi-Bred International,
Associative Genetics, Johnston, IA50131-1004
In this study, we attempt to validate candidate loci for high oil content in maize kernels by following the allele
frequency at these loci in a high oil long term selection population, ASK (Alexho Single Kernel Synthetic). This
population was bred for high oil over a total of 28 cycles of selection (Misevic, D & Alexander, DE. 1989. Crop
Sci. 29:320-324). The hypothesis is that certain allele(s) of a gene involved in the high oil trait will be selected
for, and therefore will either increase in frequency or become fixed after several rounds of selection. Candidate
loci identified in a microarray study (J.-M. Lee, unpublished data; see also J.-M. Lee et al., 2002 Funct. Integr.
Genomics 2 :13-27) were selected for analysis. For each of the candidate genes, SNP haplotypes were identified
and their frequencies in the ASKc0, c5, c20 and c28 populations were determined. The same analysis was also
performed for a set of control loci. Pronounced changes in haplotype frequency at most of the loci examined were
observed, including some of the control loci. To help distinguish between selection and random drift, computer
simulation of the selection process has been conducted. The simulation allowed estimation of the likelihood that
the observed changes in allele frequency could be explained by the effects of genetic drift due to the structure and
operation of the breeding program. Thus, allele frequency changes that has a low estimated likelihood under the
drift model were identified as selected alleles for further investigation In addition to the candidate loci, we have
also examined the genetic diversity at 100 SSR loci for 43 individuals from selection cycles 0,5,10,20, 25 and 28.
These results allowed us to evaluate oil QTLs and candidate loci identified by gene expression analysis, and to
identify new oil-associated loci in the ASK population. Examples of the allele frequency study with SNP, SSR
markers and computer modeling on Chromosome 9 will be presented.
operation of the breeding program. Thus, allele frequency changes that has a low estimated likelihood under the
drift model were identified as selected alleles for further investigation In addition to the candidate loci, we have
also examined the genetic diversity at 100 SSR loci for 43 individuals from selection cycles 0,5,10,20, 25 and 28.
These results allowed us to evaluate oil QTLs and candidate loci identified by gene expression analysis, and to
identify new oil-associated loci in the ASK population. Examples of the allele frequency study with SNP, SSR
markers and computer modeling on Chromosome 9 will be presented.
194
A test for a heritable epigenetic component of heterosis
Ream, Thomas {1} Auger, Donald {1} Birchler, James {1}
{1} University of Missouri-Columbia
Our lab is interested in developing a molecular explanation of heterosis. Toward this end, we conducted a test of
whether there is a heritable epigenetic component to hybrid vigor/inbreeding depression. Classical studies of the
effect of homozygosis of the r locus on its expression served as the impetus for these studies. Styles (1967)
showed that maintaining R mottling alleles in the homozygous state would produce an accumulating decrease,
over several generations, of their paternally imprinted expression. This effect is reversed by R allele
heterozygosity. If this behavior were characteristic of many regulatory genes, then such a phenomenon could
contribute to the basis of heterosis/inbreeding depression. To examine this question, inbreds of Mo17 and B73 and
the two reciprocally produced hybrids were crossed by stock 6 to generate four classes of maternal haploids. This
protocol produces the hybrid condition for one generation and then eliminates it for all genes in the subsequent
generation to assay for any heritable effect. The haploid kernels were selected using anthocyanin markers and
embryo morphology. The mature haploid plants were measured for numerous quantitative traits including plant
height, culm circumference, leaf length, leaf width, number of tassel branches, number of nodes, position of the
ear, ear length and number of tillers among others. If inbreeding depression results from an accumulating heritable
effect that is reversed by the hybrid state, one would expect the haploids derived from the hybrids to perform
better than those derived from the inbred lines. Previous work by Kato (2002) on doubled haploids indicates that
inducing the haploid state itself does not invoke any obvious change in the quantitative characteristics of an inbred
line. The haploids derived from the hybrids showed greater variance as expected from the fact that each individual
has a unique genotype that is different from the inbred lines but overall they did not exhibit greater performance
than the mean of the inbreds. These data do not support the idea that inbreeding depression has a heritable
accumulating component.
195
Molecular Dissection of a Stalk Quality Quantitative Trait Locus on Maize Chromosome 3
Reinders, Jon {1} Aldorfer, Micheal {2} Chang, Hur-Song {3} Zhu, Tong {3} Warner, Todd {1} Arbuckle,
John {1}
{1} Syngenta {2} Novartis {3} Torrey Mesa Research Institute
Maize stalk quality is a quantitative trait associated with yield stability. This study was conducted to validate a
QTL effect on maize chromosome 3 and characterize candidate genes using structural and functional genomics
analyses. The QTL was validated using a marker-assisted backcrossing strategy. Marker-assisted introgression of
the donor segment created a population of NIL contrasts. Phenotypic assessments measuring agronomic traits and
disease inoculations in 2000 and 2001 across 5 locations were completed. The allele effect in the BC2 NIL
contrasts across 2000 and 2001 were a 15, 15, and 12% reduction in stalk rot, stalk lodging, and crown rot means,
respectively. Allele effects in the BC3 NIL contrasts in 2001 were a 6, 6, and 4% reduction in stalk rot, stalk
lodging, and crown rot means, respectively. In 2001, BC2 and BC3 NIL contrasts indicated a 5 and 4% increase in
rind strength, respectively. Candidate genes were characterized using expression profiling of two NIL-derived
testcross entries grown under greenhouse conditions in 2002. The results of total lignin content and the ratio of
lignin monomers assays of whorl tissues, sampled at 4 vegetative stages, indicated heritable differences under
greenhouse conditions. Messenger RNA was extracted from whorl tissues per entry for each stage. Expression
profiles were obtained using an Affymetrix RiceChip array. The results indicated 982 oligos were differentially
regulated. Comparative genomic analyses of mapped maize loci linked to the QTL region aligned to the rice
genome identified a syntenic region on rice chromosome 1. Structural queries within this region have aligned 12
of the 982 RiceChip oligos with gene predictions at this syntenic interval. Efforts to validate these candidate genes
are in progress.
196
Diverse Maize Germplasm Phenotypes: Population Structure and Heterosis
Romero, Susan {1} Flint-Garcia, Sherry {1} Liu, Jack {2} Doebley, John {3} Goodman, Major {4} Buckler,
Edward IV {1,5}
{1} Department of Genetics, North Carolina State University {2} Department of Statistics, North Carolina State
University {3} Genetics Department, University of Wisconsin-Madison {4} Department of Crop Science, North
Carolina State University {5} USDA-ARS
Germplasm diversity is the raw material needed for maize improvement and dissection of complex genetic
traits. Although there is a tremendous amount of variability for many traits in diverse inbred lines, most maize
research is done on a small number of inbred lines, e.g. B73 and Mo17. We have now extensively characterized
300 lines, a germplasm set that captures about 85% of all the SNPs in maize, and their B73 hybrids for up to 50
phenotypes. Phenotypes included tassel measurements, leaf measurements, plant and ear height, flowering dates,
node counts, ear measurements, and kernel characteristics.
Unlike most germplasm surveys, we have accurate estimates of relatedness and population structure based on
100 SSR loci. In this study, we estimate that heritabilities ranged from 0.15 to 0.65 across all traits. We also
discuss which traits are highly correlated with population structure. These correlations with population structure
have important implications for association studies and for allele mining projects. We will also present
preliminary data on how heterotic effects correlate with genetic diversity for a range of non-yield traits.
Characterization of this germplasm at phenotypic and molecular levels will aid in the dissection of complex traits
and the planning of projects to exploit the rich diversity of maize.
197
Determination of QTL associated with root lodging in maize
Scott, Trisha {1} Johns, Mitrick {1}
{1} Northern Illinois University
Root lodging continues to be a serious agronomic issue in maize. In an effort to determine QTL associated with
root lodging, an F2 mapping population of 47 individuals was created from two siblings derived from two lines
exhibiting extreme phenotypes averaging 0.7 and 7.7 percent root lodging. These F2s, the F1 cross of the siblings,
and the originating lines were crossed with a tester historically experiencing no root lodging. Percent root lodging
was evaluated for 20 reps of each testcross in 5 environments and averaged over two years. Sibling parents of the
F2 population were screened with 174 SSR markers for polymorphisms, and testcross progeny were analyzed for
QTL using the resulting 74 SSRs. Three areas of significant loci were found on chromosomes 3, 9, and 10 with
single marker analysis, which correlate to QTL regions found in previous studies of specific root system
morphological traits. Interval mapping showed two of these QTL, near marker bnlg1160 on chromosome 3 and
bnlg1185 on chromosome 10, explained 21.7 and 24.3 percent of the phenotypic variation respectively. The LOD
score of bnlg1185 was above the calculated threshold of 2.6; bnlg1160 was slightly lower with a LOD score of
2.5. Both markers were also found to have significant allele effects, potentially allowing prediction of root lodging
in certain genotypes.
198
Association Mapping of Phenotypic Variation in Maize Flowering Time
Thornsberry, Jeffry {1} Young, Gregory {2} Whitt, Sherry {1} Buckler, Edward {1}
{1} USDA/ARS North Carolina State University {2} North Carolina State University
Flowering time is an agronomically important quantitative trait, which is potentially influenced by a large number
of maize candidate genes. Association testing, using diverse inbred lines, provides a rapid means of dissecting the
effects of these putative flowering time genes at very high resolution, provided that the effects of population
structure are controlled in the statistical analysis. Previous association mapping experiments had concentrated
upon SNP discovery in a set of 100 inbred lines and subsequent statistical tests to identify polymorphisms that
associated with phenotypic variation observed in the field. We will discuss current efforts at confirming
associations by scoring specific polymorphisms in an expanded set of 200 additional inbred lines. We will also
discuss our efforts to streamline the process of SNP discovery for use in association mapping. This approach is
based upon our knowledge of linkage disequilibrium and genotypic diversity, which allows us to perform initial
SNP discovery in a much smaller core set of inbred lines. Understanding the effects of such nucleotide
polymorphisms and their presence in the germplasm will improve selective breeding techniques for these
quantitative traits.

Transposable Elements

199
DNA Repair Genes and Ac/Ds Transposition
Altun, Cagla {1} Mahajan, Pramod {2} Saballos, Ana {1} Friesner, Joanna {3} Marshall, Kelly {1} Zundo,
Andrew {1} Britt, Anne {3} Haber, James {4} Weil, Clifford F. {1}
{1} Purdue University {2} Pioneer Hi-Bred {3} UC Davis {4} Brandeis University
Host cell genes are likely to be involved in repairing the DNA damage created when transposons excise. To
investigate this further, we have been exploiting a system engineered to allow Ac/Ds transposition in yeast. Data
from yeast suggest that the repair of Ac excision sites occurs through a single-strand annealing mechanism and
that it involves, among other host factors, the Mre11/Rad50 and Ku complexes. Data are presented describing
unexpected results from yeast Mre11 separation-of-function mutations that begin to address the DNA processing
roles of this protein. We have also begun to examine the roles in transposition for these complexes in plants. For
example, Ku80-deficient mutants of Arabidopsis have a radiation hypersensitive phenotype and fail to produce
transposon footprints following Ac excision, when compared with wild-type sibs. Current data on mre11, rad50,
ku70, and ku80 mutants of Arabidopsis and mre11 and rad50 mutants of maize are presented.
200
Ac insertional mutagenesis of the vp7/ps1 locus of maize
Brutnell, Thomas {1} Singh, Manjit {1} Lewis, Paul {1} Hardeman,
Kristine {2} Bai, Ling {1} Rose, Jocelyn {3} Mazourek, Michael {3}
Chomet, Paul {2}
{1} Boyce Thompson Institute {2} Monsanto/Mystic Research {3} Cornell
University
We are currently developing a series of near isogenic maize inbred lines
homozygous for a single active Ac transposon. The Ac elements are precisely
positioned in each line and mapped using publicly available recombinant
inbred populations (IBM94 and BNL96).
http://bti.cornell.edu/Brutnell_lab2/Projects/Tagging/BMGG_pro_tagging.html

Ultimately, approximately 200 lines will be generated with Ac insertions

placed at roughly 20 cM intervals throughout the genome. To demonstrate the
utility of this system we have performed a targeted mutagenesis of the vp7/ps1
locus. We selected 400 transposition events from an Ac located approximately
4 cM from the ps1 locus and self-pollinated the plants. Phenotypic screens of
386 ears revealed 7 putative ps1-insertion alleles. Using an inverse-PCR
protocol, we cloned ps1 sequences flanking one of the Ac insertions. To clone
the entire ps1 gene, we developed a novel PCR-based protocol termed 'Ac
casting'. This method exploits the tendency of Ac to insert locally in somatic
tissues of the plant to create a library of insertion alleles. These somatic
insertions can be recovered through two rounds of PCR using gene-specific
and Ac-specific primer sets. In theory, this technique could be utilized to clone
genomic sequences flanking any Ac insertion in the maize genome. Finally,
sequence analysis and HPLC data will be presented that strongly suggests that
the ps1/vp7 locus encodes for lycopene b-cyclase. This enzyme is necessary
for the production of b-carotene derivatives including ABA and a-carotene
derivatives including lutein.
201
Assembly and Analysis of RescueMu Transposon-tagged Maize Genome Survey Sequences
Dong, Qunfeng {1} Brendel, Volker {1, 2} Fernandes, John {3} Walbot, Virginia {3}
{1} Department of Zoology and Genetics, Iowa State University, Ames, IA 50011-3260, U.S.A. {2} Department
of Statistics, Iowa State University, Ames, IA 50011-3260, U.S.A. {3} Department of Biological Sciences,
Stanford University, Stanford, CA 94305-5020, U.S.A.
The Maize Gene Discovery Project (MGDP) is using a genetically engineered Mu transposon (RescueMu) for
insertional mutagenesis and gene tagging in maize. Plants are grown in grids of up to 48 rows x 48 columns with
an individual E. coli library prepared of RescueMu insertions from each row and column. Maize genomic inserts
captured by RescueMu cloning are PCR-amplified and sequenced from the transposon ends. So far, MGDP has
deposited over 100,000 genome survey sequences (GSSs) derived from Grids G, H, I, K, and M; grids P, S, Q and
AA are in progress. The RescueMu GSS assembly is more challenging than the usual genomic fragment assembly.
First, RescueMu transposons are recovered for use as sequencing templates after digestion of DNA with BamH1 +
BglII (http://zmdb.iastate.edu/zmdb/library-plate/GridGprep.html). The resulting right and left sequences flanking
RescueMu are trimmed when either restriction site(s) or a ligation site is encountered. Consequently, a RescueMu
clone can produce up to four genomic DNA sequence fragments. Second, somatic RescueMu insertions should be
sequenced only once, while heritable insertions are likely to be recovered several times. We developed a
RescueMuGSSAssembler program to remove the redundancy in the RescueMu sequences and to derive a GSS-
based maize gene index. Presently, over 8,400 GSS clusters, each representing a unique maize genomic region,
have been assembled from somatic plus heritable sequences. About 56% of those regions match either plant ESTs
or proteins, indicating RescueMu inserts into genic regions. Likely germinal insertions show an even higher
bias for genes. Based on DNA hybridization blot analysis and RescueMu sequences, heritable transpositions range
from 15% to ~100% per grid. To compare the RescueMu tagging approach with other gene-enrichment methods,
RescueMuGSSAssembler program to remove the redundancy in the RescueMu sequences and to derive a GSS-
based maize gene index. Presently, over 8,400 GSS clusters, each representing a unique maize genomic region,
have been assembled from somatic plus heritable sequences. About 56% of those regions match either plant ESTs
or proteins, indicating RescueMu inserts into genic regions. Likely germinal insertions show an even higher
bias for genes. Based on DNA hybridization blot analysis and RescueMu sequences, heritable transpositions range
from 15% to ~100% per grid. To compare the RescueMu tagging approach with other gene-enrichment methods,
we assembled >55,000 GSSs from methylation filtration methods. Our preliminary results show only 7% of the
RescueMu insertions overlapping with the regions sequenced after methylation filtration. Research is supported by
the NSF.
202
Genome-wide analysis of mariner-like transposable elements in rice reveals complex relationships with
Stowaway MITEs.
Feschotte, Cedric {1} Swamy, Lakshmi {1} Wessler, Susan R {1}
{1} Department of Plant Biology, University of Georgia, Athens GA 30602
Stowaway is a superfamily of miniature inverted repeat transposable elements (MITEs) that is widespread and
abundant in plant genomes. Like other MITEs, however, its origin and mode of amplification are poorly
understood. Several lines of evidence point to plant mariner-like elements (MLEs) as the autonomous partners of
the nonautonomous Stowaway MITEs. To better understand this relationship, we have taken advantage of the
nearly complete genome sequences of two rice subspecies to generate the first inventory of virtually all MLEs and
Stowaway families co-existing in a single plant species. Thirty four different MLEs were found to group into
three major clades and 26 families. More than 22,000 Stowaway MITEs were identified and classified into 36
families. Based on detailed sequence comparisons, MLEs were confirmed to be the best candidate autonomous
elements for Stowaway MITEs. Surprisingly, however, sequence similarity between MLE and Stowaway families
was restricted to the terminal inverted repeats (TIRs) and, in a few cases, to adjacent subterminal sequences.
These data suggest a model whereby most of the Stowaway MITEs in rice were co-mobilized by MLE
transposases encoded by distantly related elements.
203
Discovery of Helitron Type Transposable Elements in Maize Genome
Gupta, Smriti {1} Catanzaro, Nina {1} Hannah, Curt {2} Lal, Shailesh {1}
{1} Department of Biological Sciences, Oakland University, Rochester, MI 48309-4476 {2} Department of
Horticultural Sciences and Plant Molecular and Cellular Biology Program, University of Florida, Gainesville, FL
32611-0690
Despite the diversity of transposable elements, virtually all make a duplication of host sequences upon entry. Also
the 3í terminus is either a copy of a poly (A) tail or a duplication of their 5í end. An exception to these rules is a
recently described family of transposable elements called Helitrons. Sophisticated computer searches revealed the
presence of elements in some eukaryotic genomes that do not bear terminal repeats, do not duplicate host
sequences upon insertion and apparently transpose through replication and strand replacement. Until recently, the
lack of large invariant sequence motifs precluded their detection. Despite their abundance, there is no genetic
evidence of an autonomous Helitron. We recently described a maize mutant sh2-7527 that contains an insertion
within the Sh2 gene. The insertion bears striking similarity to Helitrons. Like Helitrons, the insertion lacks
terminal repeats, does not duplicate host sequences and was inserted between the host dinucleotide AT. Also like
Helitrons, the element contains the sequences TC and CTRR at its 5í and 3í termini, respectively. This mutant was
isolated from a conventional breeding program in the 1970s, strongly suggesting that Helitron-type transposable
elements are present and active in the present day maize genome. A search of maize sequences deposited in the
public databases identified three other Helitrons that bear sequence similarity to the sh2-7527 element only at both
the termini. These elements ranging from approximately 9kb to 35kb in length contain randomly distributed
copies of retroelements or pseudogenes. Based on the comparison of the insertion sites with maize ESTs and PCR
amplified genomic sequences, these elements have apparently inserted within a host AT dinucleotide sequence.
The possible impact of these elements on genome evolution is discussed.
204
Horizontal transfer of a Mu-like element (MULE).
Xinmin, Dao {1) Freeling, Michael {1} Lisch, Damon {1}
{1} University of California at Berkeley
Although many instances of horizontal transfer have been inferred from analysis of a wide variety of animal
transposons, to our knowledge there have been no documented instances of similar transfers of plant transposable
elements. This is surprising given the fact that hybridization is known to have occurred throughout plant
evolution. Here we present a case for the horizontal transfer of a MULE element between two subtribes of maize:
Panicoideae and Oryzoideae. The two species examined (Oryza sativa and Setaria italica) are separated by more
than 50 million years, and yet they carry MULE elements that are roughly 90% identical at the nucleotide level
over more than 3.8 kilobases. This sequence similarity, which is as high as that of the well conserved Adh1 gene,
includes both coding and non-coding nucleotides, suggesting that it is not the result of conservation for protein
function. These data is consistent with a horizontal transfer between an ancestor of S. italica and an ancestor of O.
evolution. Here we present a case for the horizontal transfer of a MULE element between two subtribes of maize:
Panicoideae and Oryzoideae. The two species examined (Oryza sativa and Setaria italica) are separated by more
than 50 million years, and yet they carry MULE elements that are roughly 90% identical at the nucleotide level
over more than 3.8 kilobases. This sequence similarity, which is as high as that of the well conserved Adh1 gene,
includes both coding and non-coding nucleotides, suggesting that it is not the result of conservation for protein
function. These data is consistent with a horizontal transfer between an ancestor of S. italica and an ancestor of O.
sativa. Neither element contains an intact mudrA transposase gene, suggesting that subsequent to transfer both
elements became inactive. Although well distributed and present in multiple copies in several species of Setaria,
we only detect a single hybridizing fragment in rice, suggesting that the transfer was from Setaria to rice. These
observations match previous results that indicate that the distribution of many subfamilies of MULE elements are
remarkably patchy, suggesting that horizontal transfer of MULE elements may have been a common feature of the
evolution of these elements
205
STRUCTURES, AGES AND CHROMOSOMAL DISTRIBUTIONS OF LTR RETROELEMENTS IN
THE RICE GENOME
Ma, Jianxin {1} Devos, Katrien {2} Bennetzen, Jeffrey {1}
{1} Department of Biological Sciences, Purdue University, West Lafayette, IN 47907, USA {2} John Innes
Centre, Norwich Research Park, Colney, Norwich NR4 7UH, UK
LTR retrotransposons constitute a large part of the repetitive DNA fraction in the nuclear genomes of higher
plants. LTR retrotransposons and other transposable elements play major roles in the structural and functional
evolution of higher plant genomes. Moreover, LTR retrotransposons are 'relatively neutral' sequence components
within the genome. Hence, study of the relationship between recombination, age and distribution of retroelements
in a relatively complete genome can provide insights into genome-wide mechanisms of sequence evolution. We
analyzed structures, ages and distribution of over 1000 LTR retroelements belonging to 11 families in the genome
of rice variety Nipponbare. Our preliminary analyses indicate that, though the majority of the intact LTR
retroelements are younger than 5 mys, over 75% of the elements have undergone sequence deletion and/or
recombination. Illegitimate recombination and unequal recombination are primarily responsible for LTR
retrotransposon removal. In general, the retroelements investigated are distributed over all the rice chromosomes
and show the typical pericentromeric clustering distribution, but the retroelements belonging to some families
seem to be dispersed over all the chromosomes. Our data suggest that the degree and nature of deletion of LTR
retroelements is related to their ages and chromosomal distribution in the rice genome.
206
Nested Deletions: A new tool for plant genomics research
Peterson, Thomas {1} Zhang, Jianbo {1} Lee, Michael {1} Wang, Kan {1}
{1} Iowa State University
Plant genomes are laden with local sequence duplications and clusters of homologous genes. To simplify the
analysis of these duplications and gene clusters, this project will develop a new genetic technology to generate
large deletions quickly and efficiently. This approach is based on the finding that transposable elements can
participate in alternative transposition pathways that generate novel recombination products, including large
deletions and duplications. The system described here utilizes a transgene construct containing maize Ac/Ds
transposon ends in tandem orientation within a I/dSpm transposon (Ned1; Nested deletions 1). The Ned1
construct will be transformed into maize; subsequent crosses will introduce the En/Spm transposase to mobilize
Ned1 to various genomic locations, and the Ac transposase to activate the deletion process. The action of Ac
transposase on the Ac termini within Ned1 generates an unlimited set of nested deletions with one end anchored at
the transgene locus. The Ned1 construct contains marker genes for detection of both Ned1 transpositions and Ac-
induced deletions, as well as sequences for easy cloning of deletion endpoints via plasmid rescue. If this
demonstration project is successful, this approach could, in the future, be extended to the production of a set of
maize lines containing Ned1 elements at dispersed sites throughout the genome. Researchers could then use these
lines to isolate deletions and other rearrangements in specific regions of the genome for a variety of research
applications, including 1) to map molecular and genetic markers to defined intervals; 2) to assign functions to
individual gene copies within complex loci; 3) to test the effects of gene copy number on expression and
silencing. For further information, contact Tom Peterson, 2206 Molecular Biology, Iowa State University, Ames,
IA 50011. thomasp@iastate.edu. Supported by National Science Foundation-Plant Genome Program Award
0110170.
207
Mechanisms of Mu inactivation in the UniformMu population
Robin, Kevin {1} Latshaw, Sue {1} Suzuki, Masa {1} Tan, Bao-Cai {1} McCarty, Donald R {1}
{1} University of Florida
Efficient genetic control of transposon activity is essential for functional genomics applications of the Robertsonís
Mutator system. The autonomous MuDR element is required for transposition activity of more numerous non-
autonomous Mu transposons in the maize genome. In the UniformMu transposon tagging population, somatic
transposition activity is monitored using the spotted aleurone phenotype of the bz1-mum9 mutation. In advanced
generations of UniformMu, selection of stable Mu-off derivative lines is facilitated by the frequent segregation of
Mu-inactive kernels on selfed ears. Because MuDR elements present in the population are made heterozygous
through continual backcrossing to W22, we hypothesized that loss of MuDR by segregation is a major mechanism
of Mu-inactivation in F2 families. However, epigenetic mechanisms of Mu-inactivation are also known. To
address this question we performed a genetic analysis of Mu-activity in 25 UniformMu BC4 families. Spotted and
non-spotted (stable bronze) kernels were counted on each F2 ear and the goodness of fit to Mendelian ratios
evaluated by chi square. Individuals were grown from spotted and non-spotted classes and analyzed by Southern
blotting using MuDR and Mu1 probes. The Mu1 probed DNAís were digested with HinfI in order to correlate
methylation of Mu1 TIRís with spotting activity. Families that segregate 3:1 and 15:1 spotted:non-spotted kernels
showed a good correlation with segregation of MuDR elements. Other families that fit 1:3, 1:1 or intermediate
ratios did not show a correlation with MuDR detected by Southern suggesting an epigenetic or dominant
mechanism of Mu inactivation. We conclude that both MuDR segregation and epigenetic mechanisms contribute
to Mu-inactivation in UniformMu lines.
208
Identification of the Transposase Controlling the Insertional Activity of MuDR/Mu Elements
Rudenko, George {1} Walbot, Virginia {1}
{1} Stanford University
Transposition activity of MuDR/Mu elements is directed by the autonomous element MuDR. This element
contains two genes mudrA and mudrB. The mudrA encodes several forms of MURA transposase resulting from
alternative splicing. The gene has two transcription start sites at positions +169 and +252 and four exons. It was
assumed that the first exon was a non-coding 5' untranslated region and that translation started with the ATG at
position +450, within the second exon; two large proteins are predicted from initiation at +450, MURA823 from
completely spliced transcripts and truncated MURA736 when the last intron is retained. Transgenic maize
containing 35S:MURA823 (Raizada and Walbot, 2000) or MURA823 and MURA736 transcribed from the native
promoter program Mu somatic excision from anthocyanin reporter alleles. No Mu insertion events have been
observed, even when these MURAs are crossed into lines that express all forms of MURB (35S:mudrB or
deletion derivative d112, Lisch et al., 1999). Additional analysis of mudrA RNA indicated that transcripts initiated
at +169 have an alternatively spliced first intron that eliminates the ATG at +450. An ATG at position +224 opens
a short reading frame of just 48 aa. Surprisingly, antibody against this short polypeptide detects a large ~120 kD
polypeptide in maize tissues; this is the same size as the protein recognized by antibodies to the main ORF in
MURA823. We hypothesized that a translational frameshift could incorporate the novel amino terminus into the
main ORF, resulting in the predicted protein MURA854. To determine if translational frameshifting occurs in
vivo, we constructed transgenic maize plants expressing either the native cDNA (frameshift required) or a
frameshift corrected form. MURA854 protein is produced in vivo from both constructs and programs somatic
excision of Mu elements. MURA854 lines were crossed with [1] mudrB lines to look for germinal insertion and
[2] RescueMu lines to monitor for somatic insertion. We have now identified first Mu insertion events suggesting
that MURA854 is the functional transposase of MuDR/Mu family. In addition, we provide evidence that MURB
might be essential for Mu insertion.
209
Mu killer (Muk) epigenetically silences the Mutator family of transposons
Slotkin, R. Keith (1) Freeling, Mike (1) Lisch, Damon (1)
(1) University of California Berkeley
Mu killer (Muk) is a single dominant locus that epigenetically silences the Mutator family of transposable
elements in maize. Muk completely silences both simple minimal Mutator lines as well as highly active multiple-
MuDR transposon tagging lines. Silencing by Muk is not dependent on either the position of the MuDR element to
be silenced, or on the gender of parent from which Muk was inherited. The silencing of MuDR by Muk occurs
gradually through plant development and is complete by the development of floral structures. Once silenced by
Muk, MuDR elements remain inactive for multiple generations even after Muk is segregated away, suggesting that
Muk is involved in the initiation but not the maintenance of the silenced state. Active MuDR elements express two
transcripts, the putative transposase, mudrA, and the helper gene, mudrB. When active MuDR elements are
crossed to Muk, transcript levels of mudrA decrease substantially. In contrast, mudrB transcript remains in the
total RNA fraction but fails to accumulate in the poly-A RNA fraction. In progeny plants that carry silenced
MuDR but that lack Muk, neither mudrA nor mudrB transcript is present in either the total or poly-A fraction.
From our studies with Muk, we hope to understand how the maize genome has evolved to epigenetically regulate
transposable elements.
total RNA fraction but fails to accumulate in the poly-A RNA fraction. In progeny plants that carry silenced
MuDR but that lack Muk, neither mudrA nor mudrB transcript is present in either the total or poly-A fraction.
From our studies with Muk, we hope to understand how the maize genome has evolved to epigenetically regulate
transposable elements.
210
Distribution and Evolution of the PIF/IS5 Transposon Superfamily and Its Association with Tourist-like
MITEs
Zhang, Xiaoyu {1} Jiang, Ning {1} Feschotte, Cedric {1} Wessler, Susan {1}
{1} University of Georgia
Two recently discovered active DNA transposons, the maize P Instability Factor (PIF) and the rice Pong element,
define a new eukaryotic transposon superfamily named ìPIF/IS5î. In order to examine the distribution and
evolution of this superfamily, a systematic survey combining database searches and a degenerate PCR assay was
performed. Hundreds of related TPase sequences were identified (or isolated) from a variety of plant, animal and
fungal species, indicating that PIF/IS5 is one of the most widespread and abundant superfamilies in eukaryotes.
Subsequent phylogenetic analyses distinguished three major groups (plant PIF-like, plant Pong-like and an animal
group), each represented by multiple distinct lineages. Several lineages of Pong-like elements recently amplified
to a surprisingly high copy number (~1,000) in B. oleracea after its divergence from A. thaliana. PIF and Pong are
responsible for the amplification of two Tourist-like miniature inverted-repeat transposable element (MITE)
families (mPIF in maize and mPing in rice, respectively). Relationships between PIF/Pong-like elements and
Tourist-like MITEs were further examined in rice, where it was found that most rice Tourist-like MITEs are
related to either PIF- or Pong-like. Furthermore, clear-cut associations between PIF- or Pong-like elements and
Tourist-like MITEs were readily identified. Taken together, these results suggest that members of the PIF/IS5
superfamily are responsible for the origin and amplification of Tourist-like MITEs. `
x