zyxwvutsrqponm
Copyright 0 1991 by the Genetics Society of America
zyxw
zyxw
Length and Sequence Variationin Evening Bat D-Loop mtDNA
Gerald S. Wilkinson and Alyson M. Chapman
Department of Zoology, University of Maryland, College Park, Maryland 20742
Manuscript received November 12, 1990
Accepted for publication March 2 1, 1991
ABSTRACT
Length variation in D-loop mitochondrial DNA was observed after amplificationwith the polymerase chain reaction (PCR) in 28% of 195 evening bats, Nycticeius humeralis, from seven colonies.
Nucleotide sequences of PCR products show that this heteroplasmy is characterized by an 81-bp
region which is tandemly repeated five to eight times. Southern blots using PCR products as probes
on HaeIII genomicdigests confirm the presence ofheteroplasmy. Furthermore, densitometry of
electrophoresed PCR products from 109 mother-offspring pairs indicate that heteroplasmy
is stably
transmitted from mother to offspring with one exception: a heteroplasmic offspring had a homoplasmic mother and sib. Nucleotide sequencesfrom this family reveal that a repeat duplication and
deletionoccurred. The observed mutation rate pergeneration, p, for lengthpolymorphism is
comparable toan independent estimate, p = 1O-', based on hierarchical diversity statistics.With the
exception of the repeat nearest the prolinetRNA gene, sequencesimilarities between repeats within
bats are consistent with a model of concerted evolution due to unidirectional replication slippage.
Selection is inferred to act on the first repeat because in comparison to other repeats it has the least
sequence divergence among bats, the fewest transversions, and the lowest minimum free energy
associated with folding.
zyxw
zyxwvutsr
M
ITOCHONDRIAL DNA of most animals
ranges in size from 16 to 18 kb and contains
13 proteingenes, 22transferRNAs,2
ribosomal
RNAs, and a regulatory region known as the control
region in invertebrates or the displacement loop (Dloop) in vertebrates. T h e D-loop lies between the
phenylalanine tRNA (tRNAPhe)
and theproline tRNA
(tRNAP'") and is so named because the two parent
strands are displaced by a short, variable length replication product known as 7s mtDNA (CLAYTON
1982). T h e presence of this single-stranded component enables this region of the mtDNA molecule to
bind
hydrophobic
regulatory
proteins
(ALBRING,
GRIFFITHand ATTARDI1977) and to undergoduplication and deletion events when repeats are present
et al. 1990).Comparison of rat, mouse,
(BUROKER
human, cow and Xenopus laevis D-loop sequences reveal that most of the D-loop is A T rich with the
C rich region containing
exception of a central G
an open reading frame that shows substantial similarity across species at the level of amino acid function
(SACCONE,
ATTIMONELLI
and SBISA1987).
Substantial length variationhas recently beenfound
in or near the control region
or D-loop of many animal
mtDNAs. For example, three species of bark weevils
possess mitochondria ranging in size from 30 to 36
kb (BOYCE,ZWICK and AQUADRO1989) because an
0.8-2.0-kb sequence is tandemly repeated adjacentto
the control region. All individuals sampled had more
than one mtDNA form, i.e. were heteroplasmic. Het-
+
+
eroplasmy has also been attributed to variable copy
number of tandem repeats in the control region of
sea scallop (LA ROCHEet al. 1990), Drosophila mauritiana (SOLIGNAC,
MONNEROTand MOUNOLOU1986),
and field crickets (RAND and
HARRISON
1989), andin
the D-loop of shad (BENTZEN,
LEGGETTand BROWN
1988), sturgeon(BUROKER
et al. 1990), whiptail lizards
(MORITZand BROWN 1987), and rabbit
(MIGNOTTEet
al. 1990). In humans, heteroplasmy is rare but has
been detected in association with several disorders as
an 8-kb duplication encompassing the D-loop (POULTON, DEADMAN
and GARDINER 1989)
and as a series
of multiple deletions within the D-loop (ZEVIANIet al.
1989). Length variation between humans, in contrast,
is well-known for D-loop mtDNA (CANNand WILSON
1983; GREENBERG,
NEUBOLD
and SUGINO1983). The
apparent scarcity of heteroplasmy among mammals
has prompted some investigators (e.g. RAND and HARRISON 1989) to speculate that homeotherms, due to
their higher metabolic rates, may experience stronger
selection for smaller and less variable mtDNAs than
poikilotherms.
In addition to length variation, substantial nucleotide sequence variability in the D-loop has been recorded bothwithin (AQUADRO
and GREENBERC 1983)
and between species (FORAN,HIXSONand BROWN
1989). Because the polymerases for both replication
and transcription are nuclear-coded, the promoters
for transcription of both strands lie between the open
reading frame of the D-loop and the tRNAPhe, and
zy
zyxwvutsrqpon
Genetics 128: 607-617 (July, 1991)
zyxwvuts
zyxwvut
zyxwvuts
zyx
zyx
zy
zyxwvutsrqpon
608
G. S . Wilkinson and A. M. Chapman
replication of the heavy (H) strand is primed by Dloop light (L) strand RNA (CHANGand CLAYTON
1985), severalworkers (GREENBERG,
NEUBOLDand
SUGINO1983; BROWN1985; CHANGet ai. 1985;
FORAN,
HIXSONand BROWN1989) have claimed that
the species-specificity of the D-loop sequence is evidence for nuclear-mitochondria1genome coevolution.
This interpretation assumes that mitochondrial Dloop nucleotide sequences have been under selection
for their ability to facilitate nuclear enzyme binding.
An alternative, although not exclusive, hypothesis is
that D-loop sequence variability between speciesmay
be the result of concerted evolution, i.e. the creation
of tandemly repeated sequences through replication
slippage or unequal crossing over (OHNO1970; HILLIS
et al. 1991). Concerted evolution could generate species-specific D-loop sequences independent of functional differences if separate lineages accumulate substitutions independently.
In this paper we use the polymerase chain reaction
(PCR) to characterize length and sequence variation
in the D-loopof evening bat (Nycticeius humeralis)
mitochondrial DNA. By examining PCR product sizes
and sequences between known mother-pup pairs we
verify that length heteroplasmy is transmitted maternally and is caused by a duplication or deletion of an
8 1-bp sequence that is tandemly copied five to eight
times in the D-loop. By convention we use the term
heteroplasmy to refer to multiple mtDNA genomes
within an individual rather than within a cell. But, in
contrast to prior studies in which length variation is
assessed using restriction enzyme digests of DNAextracted from an organ or the entire animal, we can
localize heteroplasmyto relatively few cells within an
individual using PCR. By comparing nucleotide sequence similarity between 81-bp repeats within bats
to sequencesimilarityof
corresponding repeats in
different bats, we evaluate concerted evolution and
nuclear-mitochondria1coevolutionaspossible alternative explanations for creating and maintaining the
substantial nucleotide sequence variation observed between bats both between and within nursery colonies.
MATERIALS AND METHODS
Study sites: Tissue samples were taken from adult female
evening bats, a small (10 g) insectivorous Vespertilionid bat,
at seven attic nursery colonies (Table I), sixin northern
Missouri and southern Iowa and one in North Carolina (G.
S. WILKINSON,
unpublished) during the summer between
1987 and 1990. At these sites, females migrate in April
from winter hibernacula and faithfully return to theirnatal
nursery colonies in either hollow trees or attics (WATKINS
1970).
T o document transmission of length and sequence variants, 58 pregnant females were kept in captivity through
parturition until their young were large enough to sample.
In 1988 we kept ten bats and in 1989 nine bats from the
Hutton colony and in 1989 we also kept nine bats from the
Grim colony. In 1990 we kept 30 females captured in an
attic in Edenton, North Carolina. All bats were released
near the site of capture before the young were old enough
to feed independently.
Sample collection and preparation: Bats were captured
with hoop or mist nets as they departed from attics at dusk.
After banding and measuring, each adult bat's chest fur was
clipped, a 3-mm excisionwas made, and approximately 1-3
mg of pectoral muscle was excised while applying ethylene
chloride topically as a local anesthetic. Tissue samples were
stored in liquid nitrogen until returned to the laboratory
where they were kept at -80" until DNA was extracted.
Muscle biopsies were minced in 300 p1 of buffer (0.05 M
Tris/HCl pH 8.0, 0.1 M EDTA, 0.1 M NaCI, and 1% SDS),
incubated overnight at 55" with proteinase K (0.5 pg/ml),
and then kept at 37" for 1 h with RNase (0.1 pg/ml) prior
to
several
phenol/phenol:chloroform:isoamyl/chloroform:isoamyl alcohol extractions (SAMBROOK, FRITSCH
and
MANIATIS1989). The resulting supernatant contained 501000 ng of genomic DNA and was purified and concentrated using Centricon-30 microconcentrators following
manufacturer's instructions (AmiconDivision,Danvers,
Massachusetts).
PCR primers and reactions: Three pairs of 20-bp sequences were usedas primers to amplify and sequence
mtDNA (Figure 1). The location, abbreviation, sequence
and orientation (L or H) of eachstrand is cytochrome B (C):
L, 5'-TGAATTGGAGGACAACCAGT-3', tRNAP'" (P): L,
5'-TCCTACCATCAGCACCCAAAGC-3',initiation of
the
repeat
array
(I):
L, 5"TGAAAAAACTACACACATGTAC-3', termination of the repeat (T): H, 5'-TTGACTGTATGGGGTATG TAC-3', conserved sequence block F (F): 5"GTTGCTGGTTTCAH,
CGGAGGTAG-3', and conserved sequence block E (E): H,
5'-CCTGAAGTAGGAACCAGATG-3'. Conserved sequence blocks(CSB) are highly homologous regions that
have been found in dolphin, cow, human and mouse
mtDNA (SOUTHERN,
SOUTHERN
and DIZON1988). While
the I and T primers are specific for evening bats, the other
two pairs of primers will amplify D-loop mtDNA from at
leastfivefamiliesof
bats (G. S. WILKINSON
and A. M.
CHAPMAN,
unpublished).
Double-stranded amplifications using the PCR were performed following standard protocol (Perkin-Elmer-Cetus,
Norwalk, Connecticut) and included a control with no template DNA. Forty cycles of 95" for 1 min, 55" for 1.5 min
and 72" for 2 min were followed by 7 min at 72". PCR
product size was determined by agarose gel electrophoresis
and ethidium bromide staining (see Figures 2 and 3).
T o quantify the relative amount of each PCR product,
Polaroid photographs of UV illuminated gels were digitized
with a video camera, 16-bit frame-grabber board, and Macintosh computer and then measured using the program
Image 1.24. The average inverse pixel value withinan 8 by
40 pixel rectangle positioned over each band was recorded
for each fragment and for anadjacent background sample.
To compute relative frequencies, the background value was
subtracted before normalizing each fragment value for an
individual to sum to one. Only relative frequencies greater
than 0.05 are recorded. This technique provided a rapid
and highly repeatable measure of fragment frequency.
Southern blot analysis:To corroborate estimates of heteroplasmybased on PCR products we digested purified
genomic DNA with HaeIII. This enzyme cuts the mtDNA
outside each end of the tandem repeat array. Digested DNA
was electrophoresed through 1.O% agarose, blotted to ZetaProbe (Bio-Rad, Richmond, California) nylon and probed
with a PCR product obtained using the C and F primers.
zyxw
Bat D-Loop Sequence Variation
81 tpdrectrepeatr
A
C
I
0
2
1
I
S
6
4
3
609
>
I
1
z
zyxw
z
zyxw
]
zyxwvutsrqponm
zyxwvutsrqp
A
A
P
I
I
I
I
I
I
I
I
I
100
200
300
400
500
600
700
800
A
A
T
F
A
E
I
mbp
FIGURE1.-Locations of PCR primers (arrows) and tandem repeats within evening bat mtDNA. See text for primer sequences.
square value as large or larger than that observed. This
The probe was labeled with"P
using a random primer
technique permits inclusion of categories in which cell values
labeling kit (United States Biochemical, Clevelend, Ohio).
are small by randomizing elements while keeping row and
Prehybridization and hybridization were carried out using
Denhardt's reagent as a blocking agent (SAMBROOK, FRITSCHcolumn totalsconstant.
To assess the difference between nucleotide sequences
and MANIATIS 1989). Mitochondrial DNA fragments that
from the same repeat we calculate the expected number of
hybridized to the PCR probe were visualizedafter 3 days of
substitutionsper site using the Jukes-Cantor distance,
exposure to x-ray film.
DNA sequencing: Sequencing of double-stranded (ds)
d = % In (3/(4q - 1))
and single-stranded (ss) DNA was performed by the dideoxy
chain termination method (SANGER,
NICKLENand C O U L ~ ~ N where q is estimated as the fraction of homologous nucleo1977) using a Sequenase kit (United States Biochemical).
tides which have the same base. We use this distance
metric
Double-stranded DNAwas obtained from PCR products
rather than 1 - q because q overestimates true similarity
that had been cut from an agarose gel and purified using
due to multiplesubstitutionevents and because d scales
glass beads (GENECLEAN,BIO 101 Inc., La Jolla, Califorlinearly with time if substitution rates are equivalent at all
nia). Single-strandedDNA was created either by denaturing
sites (JUKES and CANTOR
1969).
dsDNA or by asymmetrical PCR (GYLLENSTEN
and EHRLICH
We apportion the similarity between sequencesto differ1988). Single-stranded binding protein (SSB, 0.5 rg) was
ent regions analogous to the diversity statistics described
added to the labeling mix to eliminate compression zones in
above
by computing the average number of nucleotide
the gel. The SSB was inactivated by incubating each termisubstitutions either between pairs of bats from the same
nation reaction at 95 O for 15 min with 0.1 Pg of proteinase
colony, from different colonies or from different regions.
K. The sequencing reaction products were separated in an
The proportion of DNA divergence attributable to each
8.0% acrylamide/urea gel for which the top buffer was 0.5
levelis then found by taking the difference in average
X TBE (SAMBROOK,
FRITSCHand MANIATIS1989)while the
substitution
rates and scaling by the substitution rate for bat
bottom buffer was 1 X TBE and 3 M sodium acetate in a
pairs
from
the
highest level, i e . in different regions (NEI
2:l ratio. These procedures allowed us to score approxi1987).
mately 350 bp from each 50-cm lane.
Descriptive statistics are given as mean f one standard
Sequences wereobtained starting with either the C or P
error.
primer and moving into the D-loop or using the E or T
primers and sequencing toward the tRNAProgene. These
primer pairs allowed us to sequence complimentarystrands
RESULTS
through the region of overlap in the middle of the tandem
repeats and provided an average of 64 bp of single-copy
Repeat inheritanceand mutation: PCR resulted in
sequencebetween the C primer andthe I primer that
amplificationproducts
that varied in length both
included part of the cytochrome 6 gene and the tRNAThr
within and betweenindividuals (Figures 2 and 3).
gene. Substantial sequence similarity between evening bat
Direct sequencing ofeach product revealed that fragand other mammalian sequencesfor these genes confirmed
ments
differ insize d u e to the addition or deletion of
that we were amplifyingand sequencing mtDNA.
Statistics: T o apportion the variability in fragment numan 81-bp sequence that is tandemly repeated five to
ber within individuals, among individuals within colonies,
eight times in the D-loop (Figure 1). Southern blot
and among colonies we used the diversity indices (BIRKY,
analysis of Hue111 digested genomic DNA obtained
MARUYAMA
and FUERST1983),Ki = 1 - S xV2, where xV is
from muscletissue and probed withPCRproduct
the frequency of the jth size class in the ith level. In this
confirmed the same pattern of fragments among and
study we consider three levels for i:individual, colony and
region. T o maintainconsistencywith BIRKY MARUYAMA within individuals as obtained by PCR (Figure 2).
and FUERST(1983)and RAND and HARRISON
(1989)we
PCR band patterns arevery similar among females
denote the diversity index within individuals asKb, among
and pups from the same
family (Figure 3). To quantify
individuals within colonies
as K,, and among colonies within
this
similarity
we
estimated
the repeatability,R , within
regions as Kd. T o quantify relative variation in diversity at
families
of
the
arcsin-square
root transform of thesix
each level we used three C statistics (RAND andHARRISON
1989).The frattion of diversity found within individualsis
repeat frequency obtained by densitometry. With an
given by CI = &/&. The fraction among individuals within
average of 2.9 bats measured in each of 58 families,
colonies is CIC = (Kc - k)/& andthe frac5on among
R = 0.989 & 0.003 (BECKER1975). Densitometric
colonies within regionsis given by CCR= ( K d - Kc)/&.
estimatesof six repeat frequenciesinpups
closely
To test for differences in genotype or size class frequencorrespond
to
those
of
their
mothers
(Figure
4)
with
cies among colonies we used a Monte Carlo randomization
one exception. The boxinFigure
4 indicates a n
procedure (ROFF and BENTZEN
1989) for computing the
significanceassociatedwith obtaining a contingencychioffspring that produced two amplification products
zyxwv
zyxwvuts
610
zyxwvutsrqp
zyxwvuts
zyxwvu
zyxw
G . S. Wilkinson and A. M. Chapman
1 .o
38
:Yo
zyxwvutsr
zyxwvu
0.8 -
B
0.6-
1 2 3 4 5
'-
0.4 -
0.2 -
0.0
1
2
3
4
0.0
5
FIGURE
2.-PCR products using the C-E primer pair (A) and
Southern blots of HuellI-digested genomic DNA probed with a PE PCR product (B)from five bats. Lane I contains a blot from a
bat that was homoplasmic for five repeats, lanes 2 and 3 contain
siblings' blots that were heteroplasmic for five and six repeats, lane
4 contains a blot of a bat heteroplasmic for six and seven repeats,
and lane 5 contains a blot of a bat homoplasmic for seven repeats.
Outside lanes in (A) contain length standards.
1 2 4
3
5 6 7 8 9 1 0 1 1 12
0.2
0.4
0.6
0.8
1.0
Maternal Frequency
FIGURE4.-Frequency of six repeat PCR productsfor 109
mother-offspring pairs. The box indicates an offspring which was
heteroplasmic with a homoplasmic mother. Numbers indicate r e p
licate values for homoplasmic pairs. The least squares regression
equation is y = 0.0029 + 0.986x, r2 = 0.98.
zyxwvu
-1
1 . 2 - 3 . 4 . 5
1
1018 bp b
516 bp b
394 bp b
298 bp b
FIGURE
3.-PCR products from three families resulting from
amplification using the P-E primer pair. Lanes 1 , 5 and 9 are length
standards. Lanes 2, 6 and 10 contain samples from adult females
while the two lanes following them contain samples from their two
pups. Lanes 10-12 display heteroplasmy for fragments with five
and six 8 1-bp repeats.
while its mother and sibling produced just one. Thus,
a length mutation must have occurred in the mother.
Given that we scored 109 meiotic events, the effective
rate at which such mutations can be scored in each
zygote by PCRis 1/109 = 0.0092 -C 0.0091. Note
that this is not the mutation rate per mitochondrion
because we have not assessed the number or distribution of mitochondrial types per gamete. Furthermore, this is a lower estimate of mutation rates per
zygote because PCR cannot detect simultaneous duplication and deletion events. As shown below, these
do occur.
Because each maternal repeat had a characteristic
pattern of base substitutions, we were able to deter-
FIGURE5.-Diagram
of the positions of identical repeat sequences in the one family in which a homoplasmic mother had a
heteroplasmic offspring. Note that repeat 3 duplicated in all offspring arrays while repeat 4 was deleted in the 5 repeat offspring
fragments.
mine how the length change in this family occurred.
T h e six repeat offspring fragmentis one repeat longer
than the maternal fragment because repeat three has
beenduplicated(Figure
5). Furthermore,the five
repeat fragment for the
heteroplasmic pup and its
homoplasmic sib share this duplication but lack maternal repeat four. Thus, the
five repeat fragments of
both pups also show a deletion and concordantduplication of internal maternal repeatswhile the six repeat
fragment of the heteroplasmic pup is due just to a
duplication.
Over 500 bp of nucleotide sequence for five other
families, i.e. all repeats for a mother and her
two pups,
were also obtained. In all mother-offspring comparisons, including afive repeat family and a seven repeat
family, both pup sequences were identical to the maternal sequence.
Occurrence of heteroplasmy: Of the195 adult bats
sampled, 27.7% of the individuals were heteroplasmic
(Table 1). Only three heteroplasmic animals had evidence of three different size classes; the remaining
heteroplasmic individuals contained two sizeclasses
(Table 2). Although the NorthCarolina colony sample
appears to have a lower incidence of heteroplasmy
than most of the Missouri colonies, this difference is
Bat D-Loop Sequence Variation
TABLE 1
zyxwv
zy
61 1
Frequency of mtDNA size classesand heteroplasmy, and diversity indices among sites
Site (n)
P(5)
P(6)
P(7)
P(8)
P( H )
a
Kc
ArcsinG(6)
Total (1 95)
Missouri (1 66)
Busby ( 1 7)
Grim (1 4)
Easton (37)
Smith (23)
Zion (48)
Hutton (27)
North Carolina (29)
0.313
0.279
0.218
0.268
0.264
0.133
0.205
0.601
0.509
0.643
0.677
0.7 14
0.696
0.693
0.705
0.790
0.397
0.446
0.040
0.039
0.068
0.036
0.044
0.130
0.005
0.002
0.045
0.004
0.004
0.000
0.000
0.000
0.03 1
0.000
0.000
0.000
0.277
0.295
0.294
0.357
0.324
0.304
0.292
0.222
0.172
0.125
0.133
0.126
0.186
0.139
0.135
0.135
0.094
0.080
0.487
0.462
0.439
0.443
0.449
0.467
0.333
0.481
0.540
1.005 f 0.046
1.059 k 0.048
1.103 k 0.148
1.064 f 0.149
1.08 1 f 0.098
1.114 f 0.132
1.243 k 0.071
0.627 f 0.134
0.692 f 0.134
~~
zyxwv
zyxwvu
zyxwvutsrq
zyxwvut
zyxwv
Site indicates the location where the sample was obtained, i.e. either state or cdony. n = sample size, p(5) = frequency of the size class
containing 5 tandem repeats, p ( H ) = frequency of heteroplasmic individuals, K b = mean diversity index over all individuals at that
site, K c = diversity index for that site, and arcsinG(6) = mean arcsinG(6) over all individuals.
TABLE 2
Frequencies of mtDNA genotypes at each site
Site (n)
Total ( 1 95)
Missouri ( 1 66)
Busby ( 1 7)
Grim (1 4)
Hutton (27)
Easton (37)
Smith (23)
Zion (48)
North Carolina (29)
f ( 5 ) f(6) f ( 7 ) f(56) f ( 6 f7()5 6 f7()7 8 )
40
27
3
2
1
5
1
3
13
97
87
9
6
3 8
1 9
1 4
3 1
10
3
2
0
20
0
31
21
0
1
38
34
1
4
5
9
3
12
4
12
21 1
3
0
3
3
1
0
2
0
0
1
0
0
0
1
0
0
2
0
0
square root transformed frequencies indicates that
this variation is caused by two colonies, Hutton and
the NorthCarolina colony, which have lower
frequencies of sixand higher frequencies of five repeats than
the remaining Missouricolonies (Table 1). Similar
results are obtained for genotype frequencies upon
posthoc comparison(Table 2). The genotype frequencies at the Hutton and Edenton colonies are independent of colony(?? = 3.266, P = 0.91,500 randomizations) asare those at the five other Missouri colonies
(X2= 29.256, P = 0.17, 500 randomizations). Thus,
although differences in both genotype and repeat
frequency can be demonstrated betweencolonies,
they do not increase with geographic distance.
Sequence similarity among sites: To determine if
there were more nucleotide substitutionsbetween bats
from different colonies or different regions than
within colonies we computed the Jukes-Cantor distance between all possible pairs of the 52 unrelated
bats for which sequence data was obtained for six
repeats. This gave 549 pairs of bats from the same
colony, 630 pairs from two different colonies in the
same region, and 147 pairs from different regionsone from a Missouri colony and the other from the
North Carolina colony. We tested for differences between thesethree sets of pairs by taking 1000 samples
of 549 pairs of distancesat random without regard to
the identity of the bats, ordering the means, and then
counting the number of sampled means less than the
observed mean obtained from bats, in this case, from
the same colony. Identical randomization tests were
conducted with 630 pairs and 147 pairs to determine
if two bats from different colonies in the same state
or two different states, respectively,have more or
fewer nucleotide substitutions than expected by
chance.
Although very little sequence
variation
exists
among batsbetween
the cytochrome b andthe
tRNAP'" gene sequences, there were somedifferences
between individuals in the number of adenosine nu-
zyxwvutsrq
zyxwvu
1
1
Site indicates the location where the sample was obtained, i e .
either state or colony. n = sample sizef(56) = number of individuals
which are heteroplasmic for 5 and 6 tandem repeats,f(5) number
of individuals which are homoplasmic for 5 repeats.
not sufficient to cause heteroplasmy frequency to vary
significantly among colony sites (x'= 1.77, d.f. = 6,
P = 0.939).
These results are consistent both with an analysis of
variance on Kb and with the hierarchical C statistics.
The within individual diversity, Kb, does not differ
significantly among colonies (F6,188 = 0.596, P =
0.733). In terms of relative diversity, the variation
among colonies, CCR= 7.2%, is much less than either
the diversity among individuals within colonies,CIc =
66.8%or the diversity within individuals,CI= 25.6%.
Population subdivision: In contrast to the lack of
variation in the proportion of heteroplasmic individualswithin each colony,significant variation does
occur among colonies in genotype frequencies (x'=
70.431, P < 0.001, 1000 randomizations, Table 2)
and in the arcsin-square root transform of the frequencyofsix repeats (F6,188= 4.523, P = 0.0003,
Table 1). Note that the genotype frequencies only
score the presence or absence of a size class while the
frequency of sixrepeats is calculated from theindividual frequencies obtained from densitometric measurements.ApplicationofFisher's
partial least squares
differences (PLSD) posthoc comparison to the arcsin-
zyxwvut
zyxwvu
612
yj
zyxwvutsrqp
zyxwvutsrqp
zyxwvutsrqponmlkjihgfe
0
zyxwvutsrqpo
G.'S. Wilkinson and A. M. Chapman
5
b
.
I
.
C
td
2
a,
Y
2
0.15
0.02
zyxwv
0.10
0.05
0.01
0.00
-J
n
nn
_."
1
2
3
4
5
6
Repeat Position
FIGURE6.-Jukes-Cantor distances between pairs of bats in the
same colony, in different colonies in Missouri,or between bats from
Missouri and North Carolina colonies. The distances for pairs within
;I colony include sequences for 33 bats at the Zion colony, four bats
from the Hutton colony, and three bats at each of the other five
colonies. Sequences were obtained either from bats homoplasmic
for six repeats or from the six repeat fragment of a heteroplasmic
bat. Asterisks indicate those sets of paired distances that were
significantly different from the overall mean distance according to
r.;lndomi;lation tests as described in the text. Repeat position 1 is
closest to the tRNA"'" gene.
cleotides at the beginningof the D-loop. Most individuals had seven adenosine residues at this position, but
both six and eight were recorded. However, randomization tests indicated that no significant differences
exist between any of the three sets of pairs and the
overall mean. T h e overall average Jukes-Cantor distance between batpairs was 0.0044 f 0.0003. In
contrast, if the sequence data from all six repeats is
pooled, a significant difference is obtained for pairs
of bats within colonies versus pairs between colonies
o r between regions ( P C 0.001).T h e average JukesCantor distance within the repeat region is 0.0143 f
0.0002 indicating greater sequencedivergence between bats within the repeat region than outside the
repeats. To determine if anynucleotidesequence
heterogeneity exists among thesix repeats, we analyze
each repeat separately below.
Although heterogeneity exists among themean distances for all repeats depending on whether the pairs
were from the same or different colonies, the most
instructive difference between these groups occursin
repeat one, that repeat closest to the tRNA"'" gene
(Figure 6). This repeat is the most conserved repeat
in that it shows significantly less divergence between
pairs of bats than any of the other repeats. In comparison to the within repeat nucleotide substitution
rate between regions, 25.5% of the genetic differentiation observed at repeatone is due to variation
within colonies, 30.9% is due to variation between
colonies, and 43.6% is due to sequence variation between regions. Incontrast,repeat
six shows much
greater divergence between bats, althoughbats within
the same colony do not have consistently more similar
sequences at repeat six than bats from different colo-
0
zyxwv
3
4
No. Intervening Repeats
FIGURE7.-Jukes-Cantor distances between the sequences of two
81-bp repeats within each of 52 bats. The number of repeats
separating the two repeats is indicated on the abscissa. Sequences
were obtained either from bats homoplasmic for six repeats or from
the six repeat fragment of a heteroplasmic bat. Error bar indicates
one standard error obtained by bootstrapping.
nies (Figure 6). Internal repeats show distances between individuals which are intermediate tothose for
repeats one andsix, and also fail to show consistently
greater divergence between bats from different colonies.
Sequence similarity among repeats within
bats: If
either unequal crossing over or replication slippage
causes heteroplasmy, then the greatest sequence similarity should occur between those repeats that most
frequently undergo duplication events (OHTA1980).
Under this premise, duplication almost certainly is
restricted to neighboring repeats because of the 52
bats sequenced, adjacent repeatsare most similar and
the distances between repeat sequences diverge at an
exponentially increasing rate as the number of intervening repeats increases (Figure 7).
Because the adjacent repeat category in Figure 7
pools all pairs of adjacent repeats together, any effect
of repeat position on adjacent repeat sequence similarity is obscured. Figure 8 shows that repeat location
does affect the similarity between adjacent repeats.
Adjacent pairsof repeats in the middle of the tandem
repeatregion
show the most sequence similarity
whereas adjacent repeats at each end of the repeat
region show significantly greater distances. However,
this divergence in sequence similarity is asymmetrical
in that the repeat furthest from the tRNAPr", repeat
six in most individuals, shows much greater differentiationfrom its neighboringrepeatthandoes
the
repeat closest to the tRNA"'". Note, however, that
repeat one is much less similar to repeat two than
internal adjacent repeats.Thus, repeat oneis unlikely
to be the result of a recent duplication event.
Substitution bias and location within repeats:
Base
pair substitutionshave not occurredat random among
repeats with respect to base pair identity or position.
Within the consensus repeatsequence, i.e. that sequence obtainedby using the base pair observed most
often in all bats at each of the 81 positions, there are
zyxwvutsrq
00
92
zyxwvut
zyxwvuts
zyxwvutsrqpo
zyxwvuts
zyxwvutsrqpon
zyxwvut
zyxwvutsrqpo
Bat
Variation
D-Loop Sequence
0.15
8
K
a
c.
v)
6
b
c
c
vnr
0.05
a,
Y
3
7
0.00
2-3
1-2
3-4
4-5
5-6
Repeat Pair Position
FIGUREIl.-Jukes-Cantor distances between adjacent 81-bp repeats w i t h i n each of 52 bats. Sequences were obtained either from
Ixlts homophsmic for six repeats or from the six repeat fragment
of a heteroplasmic bat. Repeat positions are the .same as in Figure
6.Error bar indicates one standard error obtained by bootstrapping.
TABLE 3
Substitution bias within each repeat using all pairs of 52
sequences
Repeat
Transversions
Transitions
1
0
2
3
4
6
Total
(Figure 10) with a stem and terminal loop when the
sequence is folded to minimize the free energy of the
structure (ZUKERand STIECLER1981). Comparison of
the minimum free energies associated withall first,
second and end repeats, of which there are 9, 13 and
17 haplotypes, respectively, shows a significant differenceamongthe
mean bindingenergies (F2.155 =
68.14, P = 0.0001). Fisher PLSD tests indicate that
significant differences exist between all three repeats.
Note,however,
thatrepeatone
shows noticeably
lower average free energies than either repeat
t w o or
six (Table 4).
T h e sequence in the terminal loop of all structures
formed is of interest because it includes the first 7 bp,
5’-ACATAAA-3‘, which DODA,WRIGHTand CLAYTON (1981) identified in humans as being 51-53 bp
upstream from the termination of replication of the
7 s daughter strand.In the mouse, there are foursuch
termination-associated sequences (TAS),one of which
is the same as the human, and the other threediffer
by 1 bp, i.e. 5”ACATTAA-3 ’. T h e 7-bp human TAS
also appears along oneside of the stem in the secondary structures. T h e probability of finding these 7 bp
together anywhere in the 81-bprepeat, given the
frequency of each nucleotide in the most common
first repeat, is 0.069. T h e probability of finding these
7bp in the terminalloop is 0.00086. Thus,the
consistent location of this 7-bp sequence in the terminal loop can be considered a nonrandom event.
As with sturgeon mtDNA repeats (BUROKERet al.
1990), lower free energies are obtained when more
than one repeat is allowed to fold into a secondary
structure. For example, folding repeat one and
two
for one bat gave a minimum free energy of -25.4,
repeats one through three gave -40.9, repeats one
through four -51.2, and repeats one through five
-64.4. All of thesestructures were two-branched
stem-loops with 5’-ACATAA-3’ in each terminal
loop.
zyxwv
0.10
v)
5
613
194
378
1253
21 17
Percent
Transversions
497
1718
1411
1130
1050
784
6590
0
5.5
12.0
14.6
26.5
61.5
32.1
~~
~
~
5 (6.1%) guanosine, 9 (1 1.15%) cytosine, 29 (35.8)
thymidine and 38 (46.9) adenosinebases. Given these
base pair frequencies, transversions should represent
78.4% of all substitutions. Instead, only 32.1 % of all
substitutions are transversions. Transversions increase
progressively from 0% in repeatone to 61.5% in
repeat six (Table 3).
Furthermore,thedistribution
of transversions is
not random with respect to nucleotide position within
a repeat (Figure 9). T h e transversions in repeats two
through five are all due toanA-Tsubstitution
at
position 40 in the repeat. Although substitutions have
occurred at least once at 2 1of the 8 1 sites in a repeat
(Figure 9) only 11 sites have altered in repeats two
through five. Nine of those 11 sites have also changed
in repeat six, but only two of the 11 sites have ever
changed in repeat one. Both repeat one and repeat
six have substitutions at an additional five sites each.
Some, but notall, of these substitutions occurin bulge
areas or are paired with complimentary substitutions
in stem regions of secondary structures (Figure 10)
that can be computed, as described below, for each
repeat.
Repeat secondary structures: T h e most common
sequence of the L strand first repeat, as well as all
other repeat sequences, forms a secondary structure
zy
DISCUSSION
T h e resultspresented in this paperdemonstrate
that PCR can be used to quantify length variation in
mtDNA among populations. T h e advantages of this
technique over traditional restriction fragment length
polymorphism analysis are several: PCR is quicker,
does not require radioactivity, can be conducted on
minute samples obviating the need to sacrifice small
animals and can be used for direct sequencing. Furthermore, estimates of sequence differences between
individuals based on restriction maps may be misleading if there is duplication and deletion of repeats in a
tandem array. Frequent duplication/deletion
events
will cause substitutionrates to appearhigher than
they really are. On the other hand, the presence of
substantial length and nucleotide variation means that
614
zyxwvutsrqpon
zyxwvuts
zyxwv
zyxwvutsrqponm
zyxw
G . S. Wilkinson and A. M.Chapman
ATTAAACTATATTCCACATGAATATTAAACATGTACATAAATATATTAATATTACATAAGACATATAATGTATAATTGTAC
1:.C....T
o
w
. . . . . ..T...G.A
. .
w
w
N O D
2:.....
..
0
Wol
N
NN
N
P
W
w
0
m
zyxwv
4
........G. .........T . . . . . . . . . . . . . . . . . . .A
P
N
w
0
4
03
P O
VIN
..........
W
P
.C .......A
........G.
.....................
0
.....................
0
N
......T... .CC
. . . . . . . . . . . . . . . . .G. ..........T C. . . . . . . . . . . . . . . . . . . A
0
N
5:
................................
........G. .........T 1 . . . . . . . . . . . . . . . . . . .A
.CC......A
3:......T...
..
.CG.......
00
Nh)
..... .CC......A
..
Lnw
No
4:
...................
o w
0
0
m
N
. . . . . . . . . . . . . . .C. ....
0
N
......... T ................... A
P
0
W
P
0
0
h)
W
03
N
0
N
.....C....
......c.. .
0
0
zyxwvutsrqponm
6: ...... T...
2
..
.CC......A
3
P
ow
0
0
9 4
4
.....C..G.
b
,?
0 0
.A
Io
4
....... .T .C......G..
0 4
W
0
w
w
w
N
N
.........T .....G.... CG....C
P
0
w
0
.
P
w.
JO
....
'
w
4
FIGURE9.-Location and frequency of substitutions within the 3' to 5' consensus sequence of the light strand of the 81-bp repeat. The
proportion of the 52 bats carrying a particular substitution in each repeat is indicated below the respective base. Each substitution involved
only one base pair change as indicated with four exceptions: ') one of three changes was A to G , ') one of two changes was T to A, ') one of
two changes was G to C, and ') one of 39 changes was A to G .
Repeat #I
Consensus
A
T
FIGURE10.-Secondary structures for the most common first
repeat andthe consensus sequence obtained by minimizing the free
energy according to the Zuker-Stiegler method. The location and
number of substitutions observed among repeats two through six
for 52 bat sequences are indicated on the consensus repeat.
TABLE 4
Minimum free energies associated withstem-loop secondary
structures of each repeat
~
Repeat
1
2
6
Mean
(kcal/mole)
-9.78
-8.03
-7.44
SE
No. of different
structures
0.09
0.13
0.22
13
17
9
PCR of mtDNAcan be used in parentage studies
when maternal relationships are unknown.
Utility for matriline assignment:A high mutation
rate for length variation reduces genetic differentiation of colonies. In the evening bat D-loop, mutation
rates for length polymorphism are too high to allow
fixation of different length variants in isolated populations, but not high enough for most of the variation
to be present within an individual. Thus,the frequency of mtDNA size classes within
an individual can
be used as a first approximation to assign individuals
to matrilines because the mostlikely source fora
length variant is through acommon maternal ancestor
rather than from a mutational event. However, a more
accurate classification can be made by comparing the
sequence patterns within repeats because there are
many more unique repeat genotypes than length variants andtheper
generation mutation rate for a
nucleotide substitution in each repeat is much smaller
thanthe mutation ratefor length polymorphism.
Thus, by compiling a catalog ofrepeat haplotypes and
then comparing the sequences of the repeats, rather
than the nucleotides, it is not only possible to identify
individuals to matrilines (G. S. WILKINSONunpublished), but also to track repeat evolution.
Estimationof length mutation rates: The existence
of up to eight copies of a direct repeat suggests that
mammal mtDNA maynot experience strong selection
for reduced mtDNA size as hasbeen proposed (HARRISON 1989). Because PCR hasonlyrecentlybeen
used to assess mtDNA length variation, we predict
that more examples of heteroplasmy in mammalswill
be found. For example, in addition to evening bats,
we have found tandem repeats and similar frequencies
of length heteroplasmy intwo other Vespertilionid
zyxwvu
zyxwvutsrqp
zyxwvutsrqpo
zyxwvuts
Bat D-Loop Sequence Variation
bats, Myotis lucijigus and Eptesicus fuscus (G. S. WILKINSON and A. M. CHAPMAN,
in preparation). Rabbits
exhibit similar length variation except that the tandem
repeats lie between the CSBs and the tRNAPhe(MIGNOTTE et al. 1990). Although the frequency of heteroplasmyin evening bats, 28%, is lower than that
estimated in Gyllusfirmus, 60%, or Gryllus pennsyluanicus, 45%, (RANDand HARRISON
1989), thediversity indices of allhybrid crickets, K c = 0.471 and Kc =
0.149, are very similar to the diversity indices for all
evening bats, Kc = 0.487 and E b = 0.125. These
indices show that most of the length variation, 66.8%
in evening bats and between 61.3and 67.4% for
crickets, occurs within a colony, or in the caseof
crickets, lineage. T o maintainthislevelof
length
variation in the absence ofselection, the mutation rate
for length polymorphism mustbe relatively highand/
or the population size must be large to counteract the
effects of random assortment of mitochondria into
gameteswhich will rapidly remove variation unless
the poolof mitochondria per gamete is very large
(CLARK1988).
If the effective number of females is known and
transmission is strictly maternal, then K , = 2Np/
(2Np 1) where N is the number of breeding females
in the population and p is the mutation rate per
individual (BIRKY,MARUYAMA
and FUERST1983). If
we substitute the observed mutation rate of 0.0092
and the observed K , of 0.487 into this equation then
N = 52. Because this estimate of population size is
very close to the average number of adult females in
the seven colonies we sampled, we conclude that the
per generation mutation rate is near9 X
This
mutation rate is about 20 times greater than that
estimated for field crickets (RANDand HARRISON
1989).
The length mutation event described in this paper
required both a duplication anda deletion event.
Initially, a duplication must have occurred in repeat
three which presumably affected at least some of the
maternal mtDNA molecules in the ovaries. In other
words, the mother must have had different mtDNA
forms in different tissues. This would explainwhy the
maternal sequence amplified from a chest muscle biopsy could differ from both offspring sequences. The
duplication event was apparently followed by a deletion of repeat four in some mtDNA molecules. If an
ovum received one mtDNA type, a homoplasmic offspring would result; if it received both types, the
offspring would be heteroplasmic.
Mechanisms for length variation: Four different
mechanisms have beenproposed to account for length
variation in mitochondrial DNA: intra- and intermolecular recombination (RANDand HARRISON
19$9),
slipped mispairing (STREISINGER
et al. 1966; EFSTRATIADIS et al. 1980), illegitimate elongation (BUROKER
615
et al. 1990), and transposition (RANDand HARRISON
1989). The recent documentation of DNAbeing
transferred into mitochondria by a protein (VESTWEBER and SCHATZ
1989)opens the possibility that transposition could occur among mitochondria. Transposition wouldnot, however, produce both a duplication
and deletion nor would it result in adjacent internal
repeats being more similar than external repeats.
Thus, transposition cannot account for the patterns
of sequence variation we observe in N. humeralis, but
it may account for the origin of the tandem repeat
unit. Although both intra- and intermolecular recombination is frequent in plant mtDNA (SEDEROFF
1987), no direct evidence of recombination has yet
been found for animal mtDNA. As RANDand HARRISON (1 989) point out, intermolecular recombination
results in molecules of differing sizes which have not
yetbeen found. However, intramolecular recombination between, for example, 7 s and parent mtDNA,
need not alter molecule size and would be very difficult to distinguish from replication slippage.
Although the slippedmispairingmodel was presented to explain deletions which are flanked by short
direct repeats (EFSTRATIADIS
et al. 1980), a similar
process could also produce duplications. Instead of
the parental strand forming a single-strandedloop
which gets excised before replication occurs, formation of a single-stranded loop in the daughter strand
will result in a repeat duplication after DNA replication and resolution of the heteroduplex molecule.
This is, in fact, the essence of the illegitimate elongation model (BUROKERet al. 1990). Both processes are
clearly a form of replication slippage. BUROKERet al.
(1990) claim that illegitimate elongation is unique to
mitochondrial replication because of triplex
its
nature.
An alternative and perhaps more revealing distinction
is that mtDNA, unlike nuclear DNA, replicates each
strand independently. In the D-loopthisprocess is
unidirectional for 7s mtDNAbecause the L strand
does not begin replicating until H strand replication
reaches the L strand origin of replication, a conserved
noncoding region between the tRNAAsn and the
tRNACysgenes (CLAYTON
1982). Unidirectional replication is significant because repeat duplication can
only occur in one direction. Thus, the most recently
duplicated repeat should be at the end of an array
unless the last repeat in the array is protected in some
way from undergoing a deletion event.
The increased nucleotide divergence between repeats one and two, as compared to adjacent internal
repeats (Figure 8), indicates that repeat one has not
duplicated as recently as have the internal repeats.
The even greater divergence between the last repeat
(number five, six, seven or eight depending on the
size of the repeat array) and the adjacent internal
repeat suggests that the last repeat also has not dupli-
zyx
zyxwvu
+
zyxwvu
616
zyxwvutsrqp
zyxwvut
zyxwvu
G . S.Wilkinson and A. M.Chapman
cated recently. Presumably, the first and last repeats
in the array undergo duplication and deletion events
at much lowerrates, if at all, than theinternal repeats.
Evidence for selection: If a duplication event is not
always accompanied by
a corresponding deletion, then
copy number will change in one strand and heteroplasmy will result. Thus, partial independence between duplicationsand deletions can account for heteroplasmy, but cannot explain why we usually found
six but never less than five or more than eight repeats.
If our estimate of p is correct and the occurrence of
tandem repeats predates the origin of the genus, as
their presence in both Myotis and Eptesicus indicates,
then there has been ample time for more variation in
repeat copy number to occur unless selection also acts
to maintain an optimal number of copies.One possible
reason for an intermediate number of repeats is that
a large secondary structure composed of five, six or
more repeats may form when a 7s strand binds to a
protein. If such protein binding occurs, then an optimal number of repeats probably exists and more or
fewer repeats should decrease binding efficiency. Under this scenario we might expect to find repeat
number, as well as nucleotide sequence, differing between related species.
The explanation just provided is insufficient by
itself, however, to explain the patterns of nucleotide
sequencesimilarity we found between bats at each
repeat. If substitution rates are equal along each repeat and selection is absent, then all repeats, even
those that duplicate, should diverge between bats at
equal rates. Thus, in the absenceofselection
the
nucleotide differences between bats from different
states should be the same for each repeat. However,
the repeat nearest the tRNAP'" is much more conserved between bats than the other five repeats indicating that substitutions are either less frequent or
less tolerated in this repeat. The hypothesis that repeat
one is under purifying selection is consistent with the
absence of transversions in this
repeat but not in other
repeats and with repeat one having the lowest average
free energy associated with a folded stem-loop structure. Therefore,an alternative interpretation of these
findings is that only one repeat is necessary for protein
binding. Some other explanation must then be provided to explainthe apparentlimitation on copy number within a tandem array. If true, this latter scenario
implicatesconcerted evolution as the mechanism causing sequence similarity among internal repeats and
within, but not necessarily between, batswhile nuclear-mitochondrialcoevolution is likely to be responsible for sequence similarity between bats at the first
repeat. Comparison
of
repeat sequences among
closely related species should help to resolve where
and how selection acts on length and sequence variation in this region.
We thank W. STEPHAN,
A. CLARK
and two anonymous reviewers
for helpful comments on the manuscript, F. MAYER,D. O'REILLY,
A. SCHERRER,
S. STEELE,P. TRAIL,R. WILLIAMS
and L. WOLFENBARGER for assistance in the field, and the Smith, Grim, Hutton,
Busby, Easton, and McLain families for access to bat colonies. We
are grateful to T. KOCHER for providing the C and E primers and
to W. RASBAND
for providing the image analysis program. Supported by a Searle Scholar Award/Chicago Community Trust and
by a biomedical research support grant from the University of
Maryland.
zyxwvuts
zyxw
LITERATURECITED
ALBRING,
M., J. GRIFFITHand G. ATTARDI,1977 Association of
aprotein structure of probable membrane derivation with
HeLa cell mitochondrial DNA near its origin of replication.
Proc. Natl. Acad. Sci. USA 74: 1348-1352.
AQUADRO,
C.F., and B.D. GREENBERG,
1983 Human mitochondrial DNA variation and evolution: analysis of nucleotide sequences from seven individuals. Genetics 103: 287-312.
BECKER,
W. A., 1975 Manual of Quantitative Genetics, Ed. 3. Washington State University, Pullman.
BENTZEN,P., W. C.LEGGETT and G. G. BROWN,1988 Length
and restriction site heteroplasmy in the mitochondrial DNA of
American shad (Alosa sapidissima). Genetics 118: 509-518.
BIRKY,
C. W.J., T. MARUYAMAand P. FUERST,1983 An approach
to population and evolutionary genetic theory for genes in
mitochondria and chloroplasts, and some results. Genetics 103:
513-527.
BOYCE, T. M., M. E. ZWICK and C. F. AQUADRO,
1989 Mitochondrial DNA in the bark weevils: size, structure
and heteroplasmy. Genetics 123: 825-836.
BROWN,
W. M., 1985 The mitochondrial genome of animals, pp.
edited by R. J.
95-130 in MolecularEvolutionaryGenetics,
MACINTYRE.Plenum, New York.
BUROKER,
N. E., J.R. BROWN,T. A. GILBERT,
P. J. O'HARA,A. T.
BECKENBACH,
W. K. THOMAS
and M. J. SMITH,1990 Length
heteroplasmy of sturgeon mitochondrial DNA: an illegitimate
elongation model. Genetics 124: 157-163.
CANN,R. L., and A. C. WILSON,1983 Length mutations in human
mitochondrial DNA. Genetics 104: 699-7 11.
CHANG,D. D., and D.A. CLAYTON,1985 Priming of human
mitochondrial DNA replication occurs at the light-strand promoter. Proc. Natl. Acad. Sci. USA 82: 351-355.
CHANG,D. D., T. W. WONG,J. E. HIXSONand D.A. CLAYTON,
1985 Regulatory sequences for mammalian mitochondrial
transcription and replication, pp. 135-144 in Achievements and
Perspectives of MitochondrialResearch, edited byE. QUAGLIARIELLO, E. C. SLATER,F. PALMIERI,
C. SACCONE
and A. M.
KROON. Elsevier, New York.
CLARK,
A. G., 1988 Deterministic theory of heteroplasmy. Evolution 42: 621-626.
CLAYTON,
D. A., 1982 Replication of animal mitochondrial DNA.
Cell 28: 693-705.
DODA,J. N., C. T. WRIGHTand D. A. CLAYTON, 1981Elongation
of displacement-loop strands in human and mouse mitochondrial DNA is arrested near specific template sequences. Proc.
Natl. Acad. Sci. USA 78: 61 16-6170.
EFSTRATIADIS, A.,
J. W. POSAKONY,
T. MANIATIS,
R. M. LAWN,c.
O'CONNELL,
R. A. SPRITZ,J. K. DERIEL,B. G. FORGET,S. M.
WEISSMAN,
J. L. SLIGHTOM, A.
E. BLECHL,
0. SMITHIES,
F. E.
1980 The
BARALLE,
C. C. SHOULDERS
and N. J. PROUDFOOT,
structure and evolution of the human &globin gene family.
Cell 21: 653-668.
FORAN,L. A., J. E. HIXSONand W. M. BROWN,1989 Similarities
of ape and human sequences that regulatemitochondrial DNA
transcription and their DNA synthesis. Nucleic Acids Res. 17:
5841-5861.
zyxw
zyxwvuts
zyxwv
zyx
Bat D-Loop Sequence Variation
GREENBERG,B. D., J. E. NEUBOLD and A. SUGINO,
1983 Intraspecific nucleotide sequence variability surrounding the origin of replication in human mitochonrial DNA. Gene
21: 33-49.
GYLLENSTEN,
U. B., and H. A. EHRLICH,1988 Generation of
single-stranded DNA by the polymerase chain reaction and its
application to direct sequencing of the HLA-DQa locus. Proc.
Natl. Acad. Sci. USA 85:7652-7655.
HARRISON,
R. G., 1989 Animal mitochondrial DNAas a genetic
marker in population and evolutionary biology. TREE 4: 611.
HILLIS, D. M., C. MORITZ,C.A.
PORTERand R. J. BAKER,
199 1 Evidence for biased gene conversion in concerted evolution of ribosomal DNA. Science 251: 308-310.
LA ROCHE,J., M. SNYDER,
D. I. COOK,K. FULLER
and E. ZOUROS,
1990 Molecular characterization of a repeat element causing
large-scale size variation in the mitochondrial DNA of the sea
scallop Placopecten magellanicus. Mol. Biol. Evol. 7: 45-64.
JUKES, T. H., and C. R. CANTOR,1969 Evolution of protein
molecules., pp. 21-132 in Mammalian Protein Metabolism, edited by H. N. MUNRO.Academic Press, New York.
MICNOTTE,
F., M. GUERIDE,
A.-M. CHAMPAGNE
and J.-C. MOUNOLOU,1990 Direct repeats in the noncoding region of rabbit
mitochondrial DNA: involvement in the generation of intra
and inter-individual heterogeneity. Eur. J. Biochem. 194: 561571.
MORITZ,C., and W. M. BROWN,1987 Tandem duplications in
animal mitochondrial DNAs: variation in incidence and gene
content among lizards. Proc. Natl. Acad. Sci. USA 8 4 71837187.
NEI, M., 1987 MolecularEvolutionaryGenetics. Columbia University, New York.
OHNO,S., 1970 Evolution by GeneDuplication. Springer Verlag,
Berlin.
OHTA,T., 1980 EvolutionandVariationinMultigeneFamilies.
Springer Verlag, Berlin.
POULTON,J., M. E. DEADMANand R. M. GARDINER,
1989 Tandem
duplications of mitochondrial DNA in mitochondrial myopathy: analysis of nucleotide sequence and tissue distribution.
Nucleic Acids Res. 17: 10223-10229.
RAND,D. M., and R. G. HARRISON,
1989 Molecular population
617
zyxw
genetics of mtDNA size variation in crickets. Genetics 121:
551-569.
ROFF,D. A., and P. BENTZEN,1989 The statistical analysis of
mitochondrial DNA polymorphisms: 'x and the problem of
small samples. Mol. Biol. Evol. 6 539-545.
SACCONE,
C., M. ATTIMONELLI
and E.SBISA, 1987 Structural
elements highly preserved during the evolution of the D-loopcontaining region in vertebrate mitochondrial DNA. J. Mol.
EvoI. 2 6 205-21 1 .
SAMBROOK,
J., E. F. FRITSCHand T. MANIATIS, 1989 Molecular
Cloning: A LaboratoryManual, Ed. 2 (3 vols.). Cold Spring
Harbor Laboratory, Cold Spring Harbor, N.Y.
SANGER,
F., S. NICKLENand A. R. COULSON,
1977 DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad.Sci.
USA 74: 5463-5467.
SEDEROFF,
R. R., 1987 Molecular mechanism of mitochondrial
genome evolution in higher plants. Am. Nat. 130: S30-S45.
SOLIGNAC, M., MONNEROT
M.
and J.-C. MOUNOLOU,
1986 Concerted evolution of sequence repeats in Drosophila
mitochondrial DNA. J. Mol. Evol. 2 4 53-60.
SOUTHERN,S. O., P. J. SOUTHERNand A. DIZON,
E.
1988 Molecular characterization of a cloned dolphin mitochondrial genome. J. Mol. Evol. 28: 32-42.
STREISINGER,
G., Y. OKADA,
J. EMRICH,
J. NEWTON,A. TSUGITA,
E. TERZAGHI
and M. INOUYE,
1966 Frameshift mutations and
the genetic code. Cold Spring Harbor Symp. Quant. Biol.
31:77-84.
VESTWEBER,
D., and G. SCHATZ,1989 DNA-protein conjugates
can enter mitochondria via the protein import pathway. Nature
338: 170-172.
L. C., 1970 Observations on the distribution and natWATKINS,
ural history of the evening bat (Nycticeius humeralis) in northwestern Missouri and adjacent Iowa. Trans. Kans. Acad. Sci.
72: 330-336.
ZEVIANI, M., S. SERVIDEI,
C. GELLERA,
E. BERTINI,s. DIMAUROand
S. DIWNATO,1989 An autosomal dominantdisorder with
multiple deletions of mitochondrial DNA starting atthe D-loop
region. Nature 3 3 9 309-3 1 1 .
ZUKER,M., and P. STIEGLER,
1981 Optimal computer folding of
large RNA sequences using thermodynamics and auxiliary
information. Nucleic Acids Res. 9: 133-148.
zyxwvutsrqpon
Communicating editor: A. G. CLARK