Untangling the early diversification of

rspb.royalsocietypublishing.org eukaryotes: a phylogenomic study of the

evolutionary origins of Centrohelida,
Haptophyta and Cryptista
Research
Fabien Burki1, Maia Kaplan1, Denis V. Tikhonenkov1,2, Vasily Zlatogursky3,
Cite this article: Burki F, Kaplan M,
Tikhonenkov DV, Zlatogursky V, Minh BQ,
Bui Quang Minh4, Liudmila V. Radaykina2, Alexey Smirnov3, Alexander
Radaykina LV, Smirnov A, Mylnikov AP, Keeling P. Mylnikov2 and Patrick J. Keeling1,5
PJ. 2016 Untangling the early diversification of 1
Department of Botany, University of British Columbia, Vancouver, British Columbia, Canada
eukaryotes: a phylogenomic study of the 2
Institute for Biology of Inland Waters, Russian Academy of Sciences, Borok, Russia
3
evolutionary origins of Centrohelida, Department of Invertebrate Zoology, St Petersburg State University, St Petersburg, Russia
4
Haptophyta and Cryptista. Proc. R. Soc. B 283: Center for Integrative Bioinformatics, Max F. Perutz Laboratories, University of Vienna, Medical University
of Vienna, Vienna, Austria
20152802. 5
Canadian Institute for Advanced Research, Integrated Microbial Biodiversity Program, Toronto, Ontario, Canada
http://dx.doi.org/10.1098/rspb.2015.2802
Assembling the global eukaryotic tree of life has long been a major effort
of Biology. In recent years, pushed by the new availability of genome-scale
data for microbial eukaryotes, it has become possible to revisit many evolution-
Received: 24 November 2015 ary enigmas. However, some of the most ancient nodes, which are essential for
Accepted: 22 December 2015 inferring a stable tree, have remained highly controversial. Among other
reasons, the lack of adequate genomic datasets for key taxa has prevented
the robust reconstruction of early diversification events. In this context, the cen-
trohelid heliozoans are particularly relevant for reconstructing the tree of
eukaryotes because they represent one of the last substantial groups that was
Subject Areas: missing large and diverse genomic data. Here, we filled this gap by sequencing
evolution, taxonomy and systematics high-quality transcriptomes for four centrohelid lineages, each correspond-
ing to a different family. Combining these new data with a broad eukaryotic
Keywords: sampling, we produced a gene-rich taxon-rich phylogenomic dataset that
phylogenomics, eukaryotes, centrohelids, enabled us to refine the structure of the tree. Specifically, we show that (i) cen-
trohelids relate to haptophytes, confirming Haptista; (ii) Haptista relates
tree of life, plastid evolution
to SAR; (iii) Cryptista share strong affinity with Archaeplastida; and
(iv) Haptista þ SAR is sister to Cryptista þ Archaeplastida. The implications
of this topology are discussed in the broader context of plastid evolution.
Authors for correspondence:
Fabien Burki
e-mail: burkif@mail.ubc.ca
Patrick J. Keeling
1. Introduction
Reconstructing the tree of life is a challenging task, because the long evolutionary
e-mail: pkeeling@mail.ubc.ca
history since the origin of life has often confounded the phylogenetic signal that
can be recovered today. Nevertheless, molecular-based phylogenies have made
possible profound rearrangements in the tree, most recently using phylogenomics
(i.e. the use of genomic-scale datasets with stronger phylogenetic power) [1].
Accordingly, the global tree of eukaryotes has been reshuffled once again, leading
to a better understanding of the relationships between the largest assemblages, or
supergroups, and the origins of some ‘orphan’ lineages [2]. However, contentious
nodes between supergroups remain, as well as a few lingering ‘orphans’. Resol-
ving the positions of these orphans is necessary for understanding their evolution,
but also impacts the tree as a whole, because poorly sampled ‘orphan’ groups
may lead to instability in the tree.
One such group lacking proper genomic data is Centrohelida, a monophyletic
Electronic supplementary material is available group of free-living predatory protists mainly found in freshwater and soil habitats,
but also increasingly recognized to occur widely in marine environments [3]. With
at http://dx.doi.org/10.1098/rspb.2015.2802 or
about 90 described species and a vast diversity of environmental sequences [4], cen-
via http://rspb.royalsocietypublishing.org. trohelids have traditionally constituted the core of the original phylum Heliozoa,

& 2016 The Author(s) Published by the Royal Society. All rights reserved.
which included a subset of microbial eukaryotes characterized Raineriophrys erinaceoides (Pterocystidae), as well as two 2
by a special type of pseudopodia, the axopodia. Heliozoa was undescribed species: Acanthocystis sp. (Acanthocystidae) and

rspb.royalsocietypublishing.org
shown to be a polyphyletic assemblage, and today several Choanocystis sp. (Choanocystidae). Our analyses unambigu-
relatively minor lineages are scattered across the tree [5]. ously confirm that centrohelids share a common origin with
Centrohelids, however, have remained one of the last sub- haptophytes. More generally, we present compelling evidence
stantially diverse groups of eukaryotes that has eluded for the phylogenetic position of the centrohelid–haptophyte
phylogenetic placement in the tree of life. Different analyses group and Cryptista, altogether bringing us one step closer
of the 18S rRNA and a small number of protein-coding to a fully resolved eukaryotic tree of life.
genes (actin, a-tubulin, b-tubulin, EF2, HSP70, HSP90) led
to placement in various regions of the tree, but never with
good statistical support. For example, centrohelids were
weakly inferred to branch close to members of the Viridiplan-
2. Methods

Proc. R. Soc. B 283: 20152802

Details of experimental procedure for culturing, molecular work,
tae, specifically glaucophytes [6] or red algae [7]. Other
sequencing, assembling and gene preparation are described in
studies showed the centrohelids to share affinities with hap-
the electronic supplementary material.
tophytes [4,8], or were inconclusive [9]. Even a larger-scale
multigene analysis involving 127 genes was unsuccessful at
the task, placing centrohelids with low confidence as sister (a) Phylogenomic datasets construction
to either haptophytes or the enigmatic telonemids [10]. Following the preparation of 263 genes for phylogenomic analysis
More recently, the partial transcriptome sequencing for the (see electronic supplementary material), all taxa were listed with
tiny centrohelid Oxnerella marina was included in a 187 SCaFoS [29], which amounted to 274 taxa. This list was first reduced
genes dataset, which resulted in a less ambiguous monophy- to 234 taxa after removing all taxa with more than or equal to 20%
letic grouping with haptophytes [11], reinforcing the phylum missing genes. A 234-taxa, 263-gene (234/263) supermatrix was
Haptista originally proposed on weaker evidence [4,12]. then constructed to infer an initial maximum-likelihood (ML) tree
with IQ-TREE v. 1.3.0 [30] under the LG þ G model. Based on this
Beyond their intrinsic interest as a large group of eukar-
initial tree, a phylogeny-driven taxon selection approach was
yotes with unknown evolutionary origin, centrohelids also
applied to reduce further the number of taxa by retaining only repre-
hold some of the clues to better understand a larger and sentative sequences within strongly supported monophyletic groups
a priori unrelated evolutionary mystery. Owing to their poss- (100% bootstrap support), discarding the longest branches and/or
ible link to haptophytes [11], centrohelids may help to shed least complete sequences. Chimeric concatenated sequences were
light on one of the most puzzling aspects of plastid evolution: also allowed by pooling highly incomplete taxa of the same genus
the origin and evolution of complex red plastids [13,14]. (see electronic supplementary material, table S1 for details). This
Centrohelids are heterotrophs, and no permanent plastid approach led to a final taxon sampling composed of 150 operational
has ever been observed [15], although kleptoplasty has taxonomic units (OTUs). Because removal of ambiguously aligned
been reported [16]. Haptophytes, on the other hand, are sites is directly influenced by the proportion of gaps, we then re-
phototrophs and possess complex plastids derived from an extracted from the unaligned and untrimmed fasta files the 150
OTUs corresponding to our final selection, which were re-aligned
endosymbiotic event with a red alga [17]. Haptophytes rep-
with MAFFT-LINSI v. 7 and trimmed with BMGE v. 1.1 [31]
resent one of four lineages harbouring such plastids, the
using conservative settings (removal of sites with more than 20%,
others being ochrophytes ( photosynthetic stramenopiles), minimum block size of 8, substitution matrix BLOSUM 75). Finally,
myzozoans (alveolates with plastids: apicomplexans, dinofla- from our starting dataset of 263 seed genes, only 250 were retained to
gellates and chrompodellids) and cryptophytes (belonging to enter the final concatenated alignment (55 554 aa positions), which
Cryptista, which also include goniomonads, katablepharids corresponded to genes with less than 50% missing OTUs. See elec-
and Palpitia). Whereas the origins of stramenopiles and alveo- tronic supplementary material, table S1 for details about missing
lates are better understood [18,19], haptophytes and Cryptista data, and electronic supplementary material, table S2 for complete
have notoriously remained challenging to place in the tree. gene names. These 250 genes, containing up to 150 OTUs, were
They are sometimes grouped together, along with telonemids concatenated into a supermatrix (150/250) with SCaFoS [29].
and centrohelids [10,11,20–23], which resulted in the establish- From the full 150/250 dataset, two reduced datasets were
considered. First, Telonema subtilis and Picomonas sp. were
ment of Hacrobia [24]. However, haptophytes and Cryptista
removed (see Results and Discussion for the justification), lead-
have also been shown to have polyphyletic origins in several
ing to the 148/250 dataset. This dataset was reduced further by
recent multigene analyses [25–27]. Thus, untangling the con- eliminating the 19 047 fastest-evolving positions corresponding
troversial phylogenetic positions of these two groups, along to bin10, according to the tree-independent method described
with their closely related plastid-lacking lineages such as in [32]; this dataset was named 148/250-slow.
centrohelids, is a much-needed step to better explain the
observed distribution of red plastids in the eukaryotic tree.
In this study, we used a phylogenomic approach includ- (b) Phylogenetic analyses
ing a broad sampling of diversity to investigate the deep Our supermatrices were analysed by ML and Bayesian tree recon-
evolutionary relationships among eukaryotes, with particular struction methods. ML analyses were performed with IQ-TREE
focus on centrohelids, haptophytes and Cryptista. For that pur- v. 1.3.0–1.3.10 [30]. Gene-partitioned and unpartitioned alignments
were analysed; in all cases, the model that best fits the data was
pose, we filled an important gap in genome datasets by
determined by IQ-TREE according to the Bayesian information cri-
sequencing high-quality transcriptomes for four centrohelid
terion (BIC). The partitioned analysis was applied to the 150/250
species, and combined those with recent transcriptomes for a and 148/250 datasets, where the best-fit model was chosen accord-
very large diversity of marine microbial eukaryotes (the ing to a greedy strategy that sequentially merges genes from the fully
MMETSP initiative [28]). Cultures for four species were partitioned alignment (250 partitions) until the model fit stops
established, each representing a different centrohelid family, improving. We opted for the new model selection procedure (-m
namely Raphidiophrys heterophryoidea (Raphidiophryidae), TESTNEW), which additionally implements the FreeRate
heterogeneity model inferring the site rates directly from the data In total, four models of evolution were tested on the differ- 3
instead of being drawn from a gamma distribution [33]. Owing to ent datasets: ML analyses employed a partition approach with
the large size of the partition schemes, only the top 20% was checked

rspb.royalsocietypublishing.org
250 gene partitions allowing each gene to have its own model,
using the relaxed clustering algorithm (-rcluster 20), as described in the LG þ Rx þ F model and the LG þ C60 þ F model (elec-
[34]. For both datasets, the best-fit partitioning scheme contained the
tronic supplementary material, table S4); Bayesian analyses
original 250 partitions, i.e. no merging was deemed necessary. This
were run under the CAT þ GTR þ G4 model. To select the
partitioning scheme was then used to specify a model for each par-
best-fitting model in ML, we followed the BIC score selection
tition, allowing each gene to have its own rate (-spp). For the
unpartitioned analyses of both 150/250 and 148/250 supermatrices, criterion, which showed that the LG þ C60 þ F model consist-
the best-fitted model corresponded to the LG matrix with relative ently achieved better scores than the other two models
rates estimated from the data using the non-parametric FreeRate (electronic supplementary material, table S4). In Bayesian fra-
model with 10 categories and empirical amino acid frequencies mework, the CAT þ GTR þ G4 model has been repeatedly
(LG þ R10 þ F). The best-fitted model for the unpartitioned analysis shown to have a better fit than simpler models based on

Proc. R. Soc. B 283: 20152802

of the 148/250-slow dataset was LG þ R6 þ F. A more complex empirical exchangeability matrices such as LG, or even
empirical mixture model not evaluated by the selection strategy in CAT þ G4 alone [40,41]. However, the size of our datasets
IQ-TREE was also tested on all datasets: following recommendation makes comparing the fit of these complex models computation-
in [35], the LG matrix was combined to an amino acid class fre-
ally prohibitive, and thus topologies corresponding to the
quency mixture model with 60 frequency component profiles plus
best-fitting LG þ C60 þ F model (ML) and the CAT þ GTR þ
a class of empirical amino acid frequency of the alignment, and
G4 model (Bayesian) are discussed in the following sections.
four gamma categories to take into account the across-site rate het-
erogeneity (LG þ C60 þ F). To assess branch support, all IQ-TREE
analyses used the ultrafast bootstrap approximation (UFboot) with (b) Evolutionary relationships among major eukaryotic
1000 replicates [36] and the SH-like approximate likelihood ratio
test (SH-aLRT) also with 1000 bootstrap replicates [37]. lineages
Bayesian analyses were performed with PHYLOBAYES MPI The LG þ C60 þ F and CAT þ GTR þ G4 analyses of the
v. 1.5a [38], under a site-heterogeneous mixture model combin- complete dataset (150/250) recovered with maximal support
ing infinite profile mixtures and exchange rates inferred from (100% UFboot and SH-aLRT; 1.0 PP) a monophyletic assem-
the data with the rates across site drawn from a discrete blage including centrohelids and haptophytes (figure 1;
gamma distribution (CAT þ GTR þ G4). Constant sites were
electronic supplementary material, S1). More generally,
removed to decrease computational time (-dc). Three indepen-
these analyses recovered many of the major eukaryotic
dent Markov chain Monte Carlo (MCMC) chains were run, for
groups, namely Obazoa, Amoebozoa, Excavata and the
at least 3000 generations but up to 7000 for the smaller
148/250-slow dataset. The burnin period was determined after SAR assemblage (stramenopiles, alveolates, Rhizaria). The
plotting the evolution of the log-likelihood (Lnl) across the iter- association previously suggested between cryptomonads,
ations, removing the generations anterior to the stabilization of katablepharids and the marine biflagellate Palpitomonas bilix
the Lnl. Convergence between the chains was assessed by exam- into the Cryptista clade was supported with 100% UFboot
ining the difference in frequency between all bipartitions and SH-aLRT and 1.0 PP [25,27,43]. In the LG þ C60 þ F tree,
(maxdiff ). Owing to the large size of our taxon sampling, conver- the Archaeplastida lineages (i.e. green algae and land plants,
gence was generally not globally achieved (maxdiff  0.46), an glaucophytes and red algae) were paraphyletic, with Cryptista
issue that has been reported in other taxon-rich phylogenomic branching with green plants and glaucophytes (96% UFboot;
studies [11,22]. The discrepancies between the chains mostly con-
88% SH-aLRT). In the CAT þ GTR þ G4 analysis, the position
cerned nodes not under active discussion in this study, except for
of Cryptista among the Archaeplastida lineages was unresolved
the monophyly of Archaeplastida, which was accordingly labelled
owing to incongruent nodes in the independent MCMC chains
unsupported; electronic supplementary material, figures S2 and
S5 show the trees inferred from each individual chains to allow (electronic supplementary material, figure S2a–c). Telonemids
visual assessment of the discrepancies. were recovered as sister to SAR (93% UFboot; 99% SH-aLRT;
0.78 PP) and Picozoa as sister to the red algae (93% UFboot;
100% SH-aLRT; 1.0 PP).
Following the inference of a close evolutionary link
3. Results between centrohelids and haptophytes, the next important
question is where Haptista goes in the global tree. The analyses
(a) Improved dataset and model selection of the 150/250 dataset placed Haptista as sister to SAR, a
To place the centrohelids in a broad eukaryotic framework, relationship that received no support under the LG þ C60 þ
we took special care to include a very large diversity for all F model, but 1.0 PP under the CAT þ GTR þ G4 model
known main lineages. Building on previously published (figure 1). To investigate this and the deeper structure of the
datasets [25,39], we more than doubled the taxon sampling, tree in more detail, we reduced our dataset in two successive
mostly using recently released high-quality transcriptomes steps. First, we removed two orphan lineages, T. subtilis and
for marine microbial species [28] (electronic supplementary Picozoa, leading to the 148/250 dataset. These enigmatic taxa
material, table S3). Our carefully curated taxon sampling con- mirror in many ways the problems we sought to solve here
tained 150 OTUs for 250 genes (55 554 aa positions), globally for centrohelids. They are still extremely poorly represented
characterized by only 21% of missing data (electronic sup- in genomic databases, being the sole representatives of a
plementary material, table S1). Importantly, the four new much higher lineage diversity [44,45], which translates into
centrohelid sequences missed only between 7.4% and 12.9% high proportions of missing data (67% for telonemids and
positions, which corresponded to at least 48 399 aa positions 86% for Picomonas sp.; electronic supplementary material,
included, representing many fold improvements compared table S1). Second, we removed from the 148/250 supermatrix
with the 76.3% missing data for the older Polyplacocystis the 19 047 fastest-evolving positions using the similarity
contractilis dataset [10]. between characters as an estimate of the evolutionary rates
 4

rspb.royalsocietypublishing.org
Proc. R. Soc. B 283: 20152802

Figure 1. Phylogenetic tree of eukaryotes inferred from the complete dataset (150/250). The topology shown corresponds to the ML tree under the LG þ C60 þ F
model, with both ML and Bayesian support value reported. Black dots on branches mean maximal support (i.e. 100% UFboot and SH-aLRT, and 1.0 Bayesian PP; the
Bayesian CAT þ GTR þ G4 topology is shown in electronic supplementary material, figure S1). When not maximal, values are indicated only if deemed robust as
follows: UFboot  95%/SH-aLRT  80%/PP  0.9. The tree is drawn rooted between Obazoa, Amoebozoa, Collodictyon, Malawimonas and the rest of eukaryotes
after [42], though we note that the position of the root is under active debate.
 5

rspb.royalsocietypublishing.org
)
zaria
ata, Rhi
Alveol
iles,
enop
tram
R (S
SA

98/91/1
hyta
Haptop

Proc. R. Soc. B 283: 20152802

Centrohelida
phyta)
ntae, Glauco
98/92/1 (V iridipla
plastida
Archae

ta
Cryptis
100/98/1
(Rhodophyta)
Archaeplastida
98/94/1

Excavata
Malawimonas
tyon)
Collodic
Obazoa
(+

Amoebozoa

Figure 2. Schematics of the new backbone for the eukaryotic tree, highlighting the relationships among the main groups. The topology is based on the 148/250-
slow supermatrix, and corresponds to both ML and Bayesian reconstructions under the LG þ C60 þ F and CAT þ GTR þ G4 models, respectively. The complete
tree is presented in electronic supplementary material, figure SX. Black dots on branches mean maximal support (i.e. 100% UFboot and SH-aLRT, and 1.0 Bayesian
PP). When not maximal values are indicated as followed: UFboot/SH-aLRT/PP. All supergroups indicated by the triangles received maximal support, with the excep-
tion of the grouping of Viridiplantae and glaucophytes, which was unsupported (shown by dashed lines). The size of the triangles roughly represents the diversity of
taxa included in our analyses, as well as the length of the longest branch in each group. The root is placed in the same position as in figure 1.

[32], leading to the 148/250-slow dataset. Fast-evolving pos-

itions are more likely to concentrate undetected multiple
4. Discussion
substitutions, even by advanced models of evolution such as (a) Towards resolving the eukaryotic tree
the mixture models used here. Removing these positions Over the past decade, several phylogenomic studies have
from large alignments diminishes the amount of undetected attempted to resolve the deep-level relationships among the
multiple substitutions, but maintains enough phylogenetic main lineages of eukaryotes [18,19,25–27,47]. These studies
information to reconstruct even ancient events, so this have greatly improved our model for the tree of eukaryotes,
approach has shown great potential in other phylogenomic but several questions remain unsolved owing to the lack
studies [26,46]. of data from poorly studied groups. Among these unsolved
The resulting topologies were similar to those based on the questions, the relationships between centrohelids, hapto-
full dataset. However, whereas the analyses of the 148/250 phytes, Cryptista and the main Archaeplastida lineages
dataset did not improve the general statistical support of the (green plants, glaucophytes and red algae) have all proved to
tree (electronic supplementary material, figures S3, S4 and be refractory to robust phylogenetic inferences. A combination
S5a–c), the reconstructions based on the 148/250-slow dataset of three important sources of artefact is most likely to explain
led to consistent and more robust topologies (figure 2; elec- the poor resolution for the placement of these lineages:
tronic supplementary material, S6). Here, Haptista received (i) lack of data; (ii) too few representative species with genomic
strong support (98% UFboot; 91% SH-aLRT; 1.0 PP) for its datasets; a (iii) models of evolution that fail to account for
position as sister to SAR, and the Archaeplastida lineages homoplasic positions. In this study, we addressed these
and Cryptista were strongly inferred to share a common ances- possible sources of incongruence by (i) sequencing the
tor (100% UFboot; 98% SH-aLRT; 1.0 PP). Archaeplastida transcriptome of four centrohelid lineages, (ii) using a
remained paraphyletic, but this was still unsupported and considerable amount of newly available taxon diversity, and
should be further tested (69% UFboot; 80% SH-aLRT; 0.89 (iii) reducing non-phylogenetic signal by removing fast-
PP). In these analyses, the Archaeplastida–Cryptista grouping evolving sites and applying site-heterogeneous models of evol-
branched with SAR þ Haptista to the exclusion of all other ution in both ML and Bayesian frameworks.
eukaryotes with maximal support (100% UFboot; 100% Our analyses recovered Haptista with maximal support,
SH-aLRT; 1.0 PP). regardless of the dataset or the model used, strongly
confirming that centrohelids share a direct common ancestry eukaryotes have important implications for how we interpret 6
with haptophytes [4,11]. For the deeper relationships among some major evolutionary and ecological transitions in

rspb.royalsocietypublishing.org
eukaryotic groups, we found that a greater taxon diversity eukaryotic history. The groups investigated here and their
together with the systematic use of site-heterogeneous relationships to the SAR and Archaeplastida supergroups rep-
models, allowing us to take into account site-specific substi- resent a complex mixture of photosynthetic and heterotrophic
tution patterns (C60 mixture and CAT models), improves eukaryotes, as well as lineages for which we have little evi-
the general statistical confidence of the tree. When combined dence as to whether they harbour a plastid or not [13,54].
with a less noisy dataset (removal of the fastest-evolving Many of these lineages possess plastids bounded by three
sites), these models converged towards a similar picture in or four membranes, which are the result of eukaryote-to-
both ML and Bayesian frameworks (figure 2). In this tree, Hap- eukaryote endosymbioses where heterotrophic organisms
tista are closely related to the SAR assemblage with high acquired plastids from red algae [55]. What makes the evol-
support, in agreement with weaker results based on lower ution of complex red plastids so hard to decipher is the

Proc. R. Soc. B 283: 20152802

taxon diversity and different models [25,48]. Another relation- apparent discrepancy between plastid and host phylogenies.
ship to receive strong support for the first time is the grouping Plastid phylogenies have generally been consistent with the
of Archaeplastida with Cryptista. This affinity between notion that all red plastids are the product of a single secondary
Archaeplastida and Cryptista has been noted before in several endosymbiosis [17,56–58]. This idea of a single origin was first
nuclear [25–27,48] and mitochondrial-based [42] phyloge- formalized in the chromalveolate hypothesis, which posited
nomic investigations, as well as in many 18S rDNA that there was a single engulfment of a red alga in a common
molecular studies [6], but unlike here, it never received signifi- ancestor of stramenopiles, haptophytes, cryptophytes and
cant support. Taken together, the affinities of Cryptista to alveolates [59]. Host-derived phylogenies, on the other hand,
Archaeplastida and of Haptista to SAR further diminish the have generally failed to provide any strong evidence that all
support for Hacrobia, which was initially a less controversial red-algal-containing lineages (and their associated plastid-
assemblage when poorer taxon sampling was available lacking relatives) are monophyletic, which is required under
[10,20,21,24]. Even though recent phylogenomic analyses con- the single endosymbiotic origin scenario. However, host phylo-
tinued to show a monophyletic Hacrobia, this was with no genies have thus far not provided any convincing alternative
support [11,22], or with better confidence only when a large topologies either, making it difficult to see how plastid and
part of the diversity was removed [11]. host data can be best reconciled.
One group of Cryptista (the cryptomonads) includes In this context, our work can help us understand the evol-
lineages with plastids of red algal origin (see below), which ution of red plastids. Specifically, the strongly supported
may confound our ability to discriminate vertically inherited grouping of Archaeplastida and Cryptista de facto rules out
genes from endosymbiotically derived ones. Indeed, it is the scenario of a single red plastid origin in a hypothetical
at face value possible that the phylogenetic relationship ancestor of a unified chromalveolate assemblage (figure 3a).
between Cryptista and Archaeplastida observed here and else- As stated above, the lack of support for the monophyletic
where [25–27,48] is due to undetected red algal genes in origin of red plastids from host data is not new, but this is
phylogenomic datasets, rather than common ancestry. This the first time, to the best of our knowledge, that a phylogenetic
is formally possible, because endosymbiotic gene transfer tree strongly argues against it. Indeed, had Cryptista branched
(EGT) is common during endosymbiosis, but there are several elsewhere in the tree, a single origin of chromalveolate plastids
reasons to suggest this is not affecting our results. First, if it was could be explained by positing additional plastid loss events,
the case that large numbers of unrecognized red algal genes however likely that may be. However, because Cryptista
invaded eukaryotic genomes after endosymbiosis, then one branches with the same lineage from which the plastid is
would expect all red algal plastid-containing lineages to con- derived (i.e. Archaeplastida), a single origin of red plastids is
tain many such genes, and accordingly, all be attracted to formally impossible, because those plastids would have
Archaeplastida, not only cryptomonads. Second, large-scale needed to travel backwards in time to result in this topology.
investigations of EGT in various eukaryotes (including the With what are now robust relationships for both plastids
whole genome of the cryptomonad Guillardia theta) have and hosts, how can we best reconcile their apparent conflictual
shown that the endosymbiotic contribution to the host topologies? Two main scenarios exist to explain the origin and
genome, although real, is probably less substantial than orig- present distribution of complex red plastids: (i) indepen-
inally envisioned [49–52]. Third, phylogenomic datasets dent secondary endosymbioses (figure 3b) and (ii) a unique
usually consist of highly expressed housekeeping genes that secondary endosymbiosis followed by additional layers of
show no sign of widespread red algal signal. Careful inspection endosymbioses (i.e. tertiary or quaternary; figure 3c). Even
of our dataset allowed us to detect various contamination in though the first scenario of independent endosymbioses invol-
different lineages, but not specifically from red algae, and sus- ving different red algae could explain the tree topologies, such
picious topologies were not included, as in the case of the a model is unlikely in the light of several other pieces of
translation elongation factor 2 [53]. Overall, we observed no evidence showing that all or substantial subsets of the ‘chro-
genes in our dataset that individually showed a strong affinity malveolate’ plastids trace back to a single secondary red algal
between Cryptista and red algae, suggesting that this relation- endosymbiont [23,60–62]. Lately, the second scenario of
ship is a reflection of vertical inheritance rather than owing to a serial endosymbioses (figure 3c) has received increased atten-
cryptic contamination of endosymbiont genes. tion, being now supported by a growing body of empirical
data [48,63]. Several versions of this serial endosymbiotic fra-
mework for red plastid evolution have been proposed, all
(b) Implications for plastid evolution involving the idea of one secondary endosymbiosis with a
Beyond these taxonomic considerations, the positions of red alga, followed by subsequent eukaryote-to-eukaryote
centrohelids, haptophytes and Cryptista in the tree of endosymbioses [48,62,64,65]. Recently, an explicit model was
 (a) single secondary endosymbiosis 7
diversification

rspb.royalsocietypublishing.org
secondary
endosymbiosis

Proc. R. Soc. B 283: 20152802

(b) multiple secondary endosymbioses
multiple secondary diversification
endosymbioses

(c) cryptic serial endosymbioses

diversification

secondary
endosymbiosis

additional serial
endosymbioses

Figure 3. Scenarios for the origin and evolution of complex red plastids. These scenarios do not refer to any specific taxa, but rather illustrate the various pos-
sibilities discussed in the text, and show that the same diversity of plastid types can be generated by different combinations of events. (a) A single secondary
endosymbiosis in the ancestor of all red plastid-bearing eukaryotes was followed only by descent with modification, as formalized in the chromalveolate hypothesis;
this scenario is not supported by the current analyses. (b) Multiple independent secondary endosymbioses take place with different red algal symbionts, followed by
descent with modification; this is compatible with current phylogenetic evidence from hosts, but not with evidence from plastids. (c) A single secondary endo-
symbiosis takes place, but is followed by serial eukaryote-to-eukaryote endosymbioses; several versions of this scenario have been proposed (see text for references),
and they are consistent with current phylogenetic data.

devised using regression analyses to measure the expected strongly supported trees can be shown to be misleading with
similarity between genomes of various ‘chromalveolate’ additional data. Moreover, the breadth for plastid genome
lineages [63]. This approach resulted in a model where crypto- data has now been far exceeded by nuclear data, so that it is
phytes first engulfed a red alga, which was then transferred likely that changes to the plastid tree will occur after the
to the ochrophytes by tertiary endosymbiosis, and to the addition of new sequences, as recently demonstrated [58]. All
haptophytes by quaternary endosymbiosis [63]. of this could ultimately affect our interpretation, but more
In this context, our results are compatible with such a importantly various kinds of data, not only phylogenetics,
‘cryptophyte-first’ model, although we note that phylogenetic will be needed to validate a particular model. Plastids are cel-
lines of evidence are not compelling by themselves. More gen- lular structures of great complexity that have integrated with
erally, our results will need to stand the test of time, as even their hosts in many ways [54,66]. Serial endosymbiosis is
currently known for certain only in a few dinoflagellate Assembled transcriptomes: Dryad data depository (http://data- 8
lineages, whose endosymbionts display peculiar ways of inte- dryad.org) accession http://dx.doi.org/10.5061/dryad.rj87v.
Untrimmed sequences, trimmed alignments and single-gene

rspb.royalsocietypublishing.org
grating that are very different from what we observe in lineages
trees: Dryad data depository (http://datadryad.org) accession
like haptophytes, ochrophytes or most alveolates [67–69]. http://dx.doi.org/10.5061/dryad.rj87v.
Thus, an integrative model of plastid evolution will need to Authors’ contributions. F.B. designed the study, participated in the dataset
explain many aspects to be comprehensive, from phylogeny construction, carried out the phylogenetic analyses and drafted the
to genetics to fine cellular processes. manuscript; M.K. participated in the dataset construction; D.V.T.
and V.Z. collected field samples, established cultures, carried out
molecular laboratory work and drafted the manuscript; L.V.R. col-
lected field samples and established cultures; B.Q.M. carried out
5. Concluding remarks phylogenetic analyses and drafted the manuscript; A.S. and A.P.M.
participated in the design of the study and critically revised the
Our centrohelid transcriptomes fill an important diversity manuscript; P.J.K. designed the study and drafted the manuscript.

Proc. R. Soc. B 283: 20152802

gap in genomic sequencing. In the near future, effort All authors gave final approval for publication.
should be made to provide better-quality datasets for taxa Competing interests. The authors declare no competing interests.
that are still evolutionary mystery but are essential to further Funding. This work was supported by a grant from the Natural
resolve the tree; telonemids and Picozoa represent obvious Sciences and Engineering Research Council of Canada, and by a
targets near to the organisms studied here, but many other grant from the Tula Foundation to the Centre for Microbial Diversity
and Evolution. This work was also partially supported by the
enigmatic microbial eukaryotes probably affect other parts
Russian Foundation for Basic Research (no. 14-04-00554, 15-34-20065,
of the tree in similar ways. More work is also necessary to 15-29-02518, 15-04-18101_a) and by a grant from the President of Rus-
determine the relative position of Cryptista to the Archae- sian Federation MK-7436.2015.4. The work of D.V.T. was supported by
plastida lineages in order to assess the monophyletic origin the Russian Science Foundation (no 14-14-00515). B.Q.M. acknowl-
of the primary plastids. edges financial support to Arndt von Haeseler from the University
of Vienna and the Medical University Vienna.
Data accessibility. Raw reads are available through GenBank sequence Acknowledgements. We thank Compute/Calcul Canada for computing
read archive: SRR2170621, SRR2170625, SRR2170626, SRR2170627, resources and assistance, in particular WestGrid’s Orcinus and
SRR2170634. Calcul Quebec’s Guillimin and Colosse facilities.

References
1. Delsuc F, Brinkmann H, Philippe H. 2005 9. Sakaguchi M, Inagaki Y, Hashimoto T. 2007 16. Patterson DJ, Dürrschmidt M. 1987 Selective
Phylogenomics and the reconstruction of the tree of Centrohelida is still searching for a phylogenetic home: retention of chloroplasts by algivorous heliozoa:
life. Nat. Rev. Genet. 6, 361– 375. (doi:10.1038/ analyses of seven Raphidiophrys contractilis genes. fortuitous chloroplast symbiosis? Eur. J. Protistol. 23,
nrg1603) Gene 405, 47–54. (doi:10.1016/j.gene.2007.09.003) 51– 55. (doi:10.1016/S0932-4739(87)80007-X)
2. Burki F. 2014 The eukaryotic tree of life from a 10. Burki F et al. 2009 Large-scale phylogenomic 17. Yoon HS, Hackett JD, Pinto G, Bhattacharya D. 2002
global phylogenomic perspective. Cold Spring Harb. analyses reveal that two enigmatic protist The single, ancient origin of chromist plastids. Proc.
Perspect. Biol. 6, a016147. (doi:10.1101/cshperspect. lineages, telonemia and centroheliozoa, are Natl Acad. Sci. USA 99, 15 507 –15 512. (doi:10.
a016147) related to photosynthetic chromalveolates. 1073/pnas.242379899)
3. Mikrjukov KA. 2000 System and phylogeny of Genome Biol. Evol. 1, 231–238. (doi:10.1093/ 18. Burki F, Shalchian-Tabrizi K, Minge M, Skjaeveland
Heliozoa: should this taxon exist in modern systems gbe/evp022) A, Nikolaev SI, Jakobsen KS, Pawlowski J. 2007
of protists? Zool Z. 79, 883 –897. 11. Cavalier-Smith T, Chao EE, Lewis R. 2015 Multiple Phylogenomics reshuffles the eukaryotic
4. Cavalier-Smith T, Heyden von der S. 2007 Molecular origins of Heliozoa from flagellate ancestors: new supergroups. PLoS ONE 2, e790. (doi:10.1371/
phylogeny, scale evolution and taxonomy of cryptist subphylum Corbihelia, superclass journal.pone.0000790)
centrohelid heliozoa. Mol. Phylogenet. Evol. 44, Corbistoma, and monophyly of Haptista, Cryptista, 19. Rodriguez-Ezpeleta N, Brinkmann H, Burger G,
1186–1203. (doi:10.1016/j.ympev.2007.04.019) Hacrobia and Chromista. Mol. Phylogenet. Evol. 93, Roger AJ, Gray MW, Philippe H, Lang BF. 2007
5. Cavalier-Smith T. 2003 Protist phylogeny and the 331 –362. (doi:10.1016/j.ympev.2015.07.004) Toward resolving the eukaryotic tree: the
high-level classification of Protozoa. Eur. J. Protistol. 12. Cavalier-Smith T, Chao EE-Y. 2003 Molecular phylogenetic positions of jakobids and cercozoans.
39, 338–348. (doi:10.1078/0932-4739-00002) phylogeny of centrohelid heliozoa, a novel lineage Curr. Biol. 17, 1420–1425. (doi:10.1016/j.cub.2007.
6. Nikolaev SI, Berney C, Fahrni JF, Bolivar I, Polet S, of bikont eukaryotes that arose by ciliary loss. 07.036)
Mylnikov AP, Aleshin VV, Petrov NB, Pawlowski J. J. Mol. Biol. 56, 387–396. (doi:10.1007/s00239- 20. Hackett JD, Yoon HS, Li S, Reyes-Prieto A, Rümmele
2004 The twilight of Heliozoa and rise of Rhizaria, 002-2409-y) SE, Bhattacharya D. 2007 Phylogenomic analysis
an emerging supergroup of amoeboid eukaryotes. 13. Archibald JM. 2015 Genomic perspectives on the supports the monophyly of cryptophytes and
Proc. Natl Acad. Sci. USA 101, 8066– 8071. (doi:10. birth and spread of plastids. Proc. Natl Acad. Sci. haptophytes and the association of rhizaria with
1073/pnas.0308602101) USA 112, 10 147 –10 153. (doi:10.1073/pnas. chromalveolates. Mol. Biol. Evol. 24, 1702–1713.
7. Sakaguchi M, Nakayama T, Hashimoto T, Inouye I. 1421374112) (doi:10.1093/molbev/msm089)
2005 Phylogeny of the Centrohelida inferred from 14. Keeling PJ. 2013 The number, speed, and impact of 21. Patron NJ, Inagaki Y, Keeling PJ. 2007 Multiple gene
SSU rRNA, tubulins, and actin genes. J. Mol. Biol. plastid endosymbioses in eukaryotic evolution. phylogenies support the monophyly of
61, 765–775. (doi:10.1007/s00239-005-0006-6) Annu. Rev. Plant Biol. 64, 27.1–27.25. (doi:10. cryptomonad and haptophyte host lineages. Curr.
8. Nikolaev SI, Berney C, Petrov NB, Mylnikov AP, 1146/annurev-arplant-050312-120144) Biol. 17, 887–891. (doi:10.1016/j.cub.2007.03.069)
Fahrni JF, Pawlowski J. 2006 Phylogenetic position 15. Mikrjukov KA, Siemensma FJ, Patterson DJ. 2000 22. Katz LA, Grant JR. 2014 Taxon-rich phylogenomic
of Multicilia marina and the evolution of Phylum Heliozoa. In The Illustrated guide to the analyses resolve the eukaryotic tree of life and
Amoebozoa. Int. J. Syst. Evol. Microbiol. 56, protozoa (eds JJ Lee, GF Leedale, P Bradbury), reveal the power of subsampling by sites. Syst. Biol.
1449–1458. (doi:10.1099/ijs.0.63763-0) pp. 860–871. Lawrence, KS: Society of Protozoologists. 64, 406 –415. (doi:10.1093/sysbio/syu126)
23. Rice DW, Palmer JD. 2006 An exceptional horizontal 35. Wang H-C, Susko E, Roger AJ. 2014 An amino acid ‘supergroups’. Proc. Natl Acad. Sci. USA 106, 9
gene transfer in plastids: gene replacement by a substitution–selection model adjusts residue fitness 3859– 3864. (doi:10.1073/pnas.0807880106)

rspb.royalsocietypublishing.org
distant bacterial paralog and evidence that to improve phylogenetic estimation. Mol. Biol. Evol. 48. Baurain D et al. 2010 Phylogenomic evidence for
haptophyte and cryptophyte plastids are sisters. 31, 779– 792. (doi:10.1093/molbev/msu044) separate acquisition of plastids in cryptophytes,
BMC Biol. 4, 31. (doi:10.1186/1741-7007-4-31) 36. Minh BQ, Nguyen MAT, von Haeseler A. 2013 haptophytes, and stramenopiles. Mol. Biol. Evol. 27,
24. Okamoto N, Chantangsi C, Horak A, Leander BS, Ultrafast approximation for phylogenetic bootstrap. 1698– 1709. (doi:10.1093/molbev/msq059)
Keeling PJ. 2009 Molecular phylogeny and Mol. Biol. Evol. 30, 1188–1195. (doi:10.1093/ 49. Burki F, Imanian B, Hehenberger E, Hirakawa Y,
description of the novel katablepharid Roombia molbev/mst024) Maruyama S, Keeling PJ. 2014 Endosymbiotic gene
truncata gen. et sp. Nov., and establishment of the 37. Guindon S, Dufayard J-F, Lefort V, Anisimova M, transfer in tertiary plastid-containing
Hacrobia taxon nov. PLoS ONE 4, e7080. (doi:10. Hordijk W, Gascuel O. 2010 New algorithms and dinoflagellates. Eukaryot. Cell 13, 246–255. (doi:10.
1371/journal.pone.0007080) methods to estimate maximum-likelihood 1128/EC.00299-13)
25. Burki F, Okamoto N, Pombert J-F, Keeling PJ. 2012 phylogenies: assessing the performance of PhyML 50. Burki F, Flegontov P, Obornik M, Cihlar J, Pain A,

Proc. R. Soc. B 283: 20152802

The evolutionary history of haptophytes and 3.0. Syst. Biol. 59, 307–321. (doi:10.1093/sysbio/ Lukes J, Keeling PJ. 2012 Reevaluating the green
cryptophytes: phylogenomic evidence for separate syq010) versus red signal in eukaryotes with secondary
origins. Proc. R Soc. B 279, 2246 –2254. (doi:10. 38. Lartillot N, Lepage T, Blanquart S. 2009 PhyloBayes plastid of red algal origin. Genome Biol. Evol. 4,
1098/rspb.2011.2301) 3: a Bayesian software package for phylogenetic 626–635. (doi:10.1093/gbe/evs049)
26. Brown MW, Sharpe SC, Silberman JD, Heiss AA, reconstruction and molecular dating. Bioinformatics 51. Deschamps P, Moreira D. 2012 Re-evaluating the
Lang BF, Simpson AGB, Roger AJ. 2013 25, 2286–2288. (doi:10.1093/bioinformatics/ green contribution to diatom genomes. Genome
Phylogenomics demonstrates that breviate btp368) Biol. Evol. 4, 795– 800. (doi:10.1093/gbe/evs053)
flagellates are related to opisthokonts and 39. Burki F, Corradi N, Sierra R, Pawlowski J, Meyer GR, 52. Curtis BA et al. 2012 Algal genomes reveal
apusomonads. Proc. R. Soc. B 280, 20131755. Abbott CL, Keeling PJ. 2013 Phylogenomics of the evolutionary mosaicism and the fate of
(doi:10.1016/j.protis.2010.06.004) intracellular parasite Mikrocytos mackini reveals nucleomorphs. Nature 492, 59 –65. (doi:10.1038/
27. Yabuki A, Kamikawa R, Ishikawa SA, Kolisko M, Kim evidence for a mitosome in rhizaria. Curr. Biol. 23, nature11681)
E, Tanabe AS, Kume K, Ishida K-I, Inagki Y. 2014 1541 –1547. (doi:10.1016/j.cub.2013.06.033) 53. Kim EE, Graham LE. 2008 EEF2 analysis challenges
Palpitomonas bilix represents a basal cryptist 40. Philippe H, Brinkmann H, Copley RR, Moroz LL, the monophyly of Archaeplastida and
lineage: insight into the character evolution in Nakano H, Poustka AJ, Wallberg A, Peterson KJ, Chromalveolata. PLoS ONE 3, e2621. (doi:10.1371/
Cryptista. Sci. Rep. 4, 4641. (doi:10.1038/srep04641) Telford MJ. 2011 Acoelomorph flatworms are journal.pone.0002621)
28. Keeling PJ et al. 2014 The marine microbial deuterostomes related to Xenoturbella. Nature 470, 54. Archibald JM. 2009 The puzzle of plastid evolution.
eukaryote transcriptome sequencing project 255 –258. (doi:10.1038/nature09676) Curr. Biol. 19, R81–R88. (doi:10.1016/j.cub.2008.
(MMETSP): illuminating the functional diversity of 41. Lartillot N, Philippe H. 2008 Improvement of 11.067)
eukaryotic life in the oceans through transcriptome molecular phylogenetic inference and the 55. Bhattacharya D, Archibald JM, Weber APM, Reyes-
sequencing. PLoS Biol. 12, e1001889. (doi:10.1371/ phylogeny of Bilateria. Proc. R. Soc. B 363, Prieto A. 2007 How do endosymbionts become
journal.pbio.1001889) 1463 –1472. (doi:10.1098/rstb.2007.2236) organelles? Understanding early events in plastid
29. Roure B, Rodriguez-Ezpeleta N, Philippe H. 2007 42. Derelle R, Torruella G, Klimeš V, Brinkmann H, Kim evolution. Bioessays 29, 1239–1246. (doi:10.1002/
SCaFoS: a tool for selection, concatenation and E, Vlček Č, Lang BF, Eliáš M. 2015 Bacterial proteins bies.20671)
fusion of sequences for phylogenomics. BMC pinpoint a single eukaryotic root. Proc. Natl Acad. 56. Qiu H, Yang EC, Bhattacharya D, Yoon HS. 2012
Evol. Biol. 7(Suppl. 1), S2. (doi:10.1186/1471-2148- Sci. USA 112, E693 –E699. (doi:10.1073/pnas. Ancient gene paralogy may mislead inference of
7-S1-S2) 1420657112) plastid phylogeny. Mol. Biol. Evol. 29, 3333–3343.
30. Nguyen L-T, Schmidt HA, von Haeseler A, Minh BQ. 43. Cavalier-Smith T. 2013 Symbiogenesis: mechanisms, (doi:10.1093/molbev/mss137)
2015 IQ-TREE: a fast and effective stochastic evolutionary consequences, and systematic 57. Janouskovec J, Horak A, Obornik M, Lukes J, Keeling
algorithm for estimating maximum-likelihood implications. Annu. Rev. Ecol. Evol. Syst. 44, 145 – PJ. 2010 A common red algal origin of the
phylogenies. Mol. Biol. Evol. 32, 268–274. (doi:10. 172. (doi:10.1146/annurev-ecolsys-110411-160320) apicomplexan, dinoflagellate, and heterokont
1093/molbev/msu300) 44. Bråte J, Klaveness D, Rygh T, Jakobsen KS, plastids. Proc. Natl Acad. Sci. USA 107, 10 949–
31. Criscuolo A, Gribaldo S. 2010 BMGE (block mapping Shalchian-Tabrizi K. 2010 Telonemia-specific 10 954. (doi:10.1073/pnas.1003335107)
and gathering with entropy): a new software for environmental 18S rDNA PCR reveals unknown 58. Ševčı́ková T et al. 2015 Updating algal evolutionary
selection of phylogenetic informative regions from diversity and multiple marine-freshwater relationships through plastid genome sequencing:
multiple sequence alignments. BMC Evol. Biol. 10, colonizations. BMC Microbiol. 10, 168. (doi:10.1186/ did alveolate plastids emerge through
210. (doi:10.1186/1471-2148-10-210) 1471-2180-10-168) endosymbiosis of an ochrophyte? Sci. Rep. 5, 10134.
32. Cummins CA, McInerney JO. 2011 A method for 45. Seenivasan R, Sausen N, Medlin LK, Melkonian M. (doi:10.1038/srep10134)
inferring the rate of evolution of homologous 2013 Picomonas judraskeda gen. et sp. nov.: the 59. Cavalier-Smith T. 1999 Principles of protein and
characters that can potentially improve phylogenetic first identified member of the Picozoa phylum nov., lipid targeting in secondary symbiogenesis:
inference, resolve deep divergence and correct a widespread group of picoeukaryotes, formerly euglenoid, dinoflagellate, and sporozoan plastid
systematic biases. Syst. Biol. 60, 833– 844. (doi:10. known as ‘picobiliphytes’. PLoS ONE 8, e59565. origins and the eukaryote family tree. J. Eukaryot.
1093/sysbio/syr064) (doi:10.1371/journal.pone.0059565.s006) Microbiol. 46, 347– 366. (doi:10.1111/j.1550-7408.
33. Soubrier J, Steel M, Lee MSY, Der Sarkissian C, 46. Rodriguez-Ezpeleta N, Brinkmann H, Roure B, 1999.tb04614.x)
Guindon S, Ho SYW, Cooper A. 2012 The influence Lartillot N, Lang BF, Philippe H. 2007 Detecting and 60. Zimorski V, Ku C, Martin WF, Gould SB. 2014
of rate heterogeneity among sites on the time overcoming systematic errors in genome-scale Endosymbiotic theory for organelle origins. Curr.
dependence of molecular rates. Mol. Biol. Evol. 29, phylogenies. Syst. Biol. 56, 389–399. (doi:10.1080/ Opin. Microbiol. 22, 38 –48. (doi:10.1016/j.mib.
3345–3358. (doi:10.1093/molbev/mss140) 10635150701397643) 2014.09.008)
34. Lanfear R, Calcott B, Kainer D, Mayer C, Stamatakis 47. Hampl V, Hug L, Leigh JW, Dacks JB, Lang BF, 61. Stork S, Moog D, Przyborski JM, Wilhelmi I, Zauner
A. 2014 Selecting optimal partitioning schemes for Simpson AGB, Roger AJ. 2009 Phylogenomic S, Maier UG. 2012 Distribution of the SELMA
phylogenomic datasets. BMC Evol. Biol. 14, 82. analyses support the monophyly of Excavata and translocon in secondary plastids of red algal origin
(doi:10.1186/1471-2148-14-82) resolve relationships among eukaryotic and predicted uncoupling of ubiquitin-dependent
 translocation from degradation. Eukaryot Cell 11, 64. Bodyl A, Stiller JW, Mackiewicz P. 2009 Chromalveolate 67. Keeling PJ. 2010 The endosymbiotic origin, 10
1472–1482. (doi:10.1128/EC.00183-12) plastids: direct descent or multiple endosymbioses? diversification and fate of plastids. Phil.

rspb.royalsocietypublishing.org
62. Petersen J, Ludewig A-K, Michael V, Bunk B, Jarek M, Trends Ecol. Evol. 24, 119–121; author reply 121–122. Trans. R. Soc. B 365, 729– 748. (doi:10.1098/rstb.
Baurain D, Brinkmann H. 2014 Chromera velia, (doi:10.1016/j.tree.2008.11.003) 2009.0103)
endosymbioses and the rhodoplex hypothesis: plastid 65. Sanchez Puerta MV, Delwiche CF. 2008 A hypothesis 68. Wisecaver JH, Hackett JD. 2011 Dinoflagellate
evolution in cryptophytes, alveolates, stramenopiles, for plastid evolution in chromalveolates. J. Phycol. genome evolution. Annu. Rev. Microbiol. 65,
and haptophytes (CASH lineages). Genome Biol. Evol. 44, 1097–1107. (doi:10.1111/j.1529-8817.2008. 369–387. (doi:10.1146/annurev-micro-090110-
6, 666–684. (doi:10.1093/gbe/evu043) 00559.x) 102841)
63. Stiller JW, Schreiber J, Yue J, Guo H, Ding Q, Huang 66. Bolte K, Bullmann L, Hempel F, Bozarth A, Zauner 69. Patron NJN, Waller RF. 2007 Transit peptide diversity
J. 2014 The evolution of photosynthesis in chromist S, Maier UG. 2009 Protein targeting into secondary and divergence: a global analysis of plastid
algae through serial endosymbioses. Nat. Commun. plastids. J. Eukaryot. Microbiol. 56, 9 –15. (doi:10. targeting signals. Bioessays 29, 1048– 1058.
5, 5764. (doi:10.1038/ncomms6764) 1111/j.1550-7408.2008.00370.x) (doi:10.1002/bies.20638)

Proc. R. Soc. B 283: 20152802