1. science
2. biology
3. zoology
4. endangered species

Genome Biology and Evolution Advance Access published July 22, 2014 doi:10.1093/gbe/evu152
Molecular Phylogeny of Sequenced Saccharomycetes
Reveals Polyphyly of the Alternative Yeast Codon
Usage
Stefanie Mühlhausen1, Martin Kollmar1§
1
Group Systems Biology of Motor Proteins, Department of NMR-based Structural
§
Corresponding author, Tel.: +49-551-2012260, Fax: +49-551-2012202
Email addresses:
SM: [email protected]
MK: [email protected]
© The Author(s) 2014. Published by Oxford University Press on behalf of the Society for Molecular Biology
and Evolution.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License
(http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and
reproduction in any medium, provided the original work is properly cited.
-1-
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
Biology, Max-Planck-Institute for Biophysical Chemistry, Göttingen, Germany
Abstract
The universal genetic code defines the translation of nucleotide triplets, called codons,
into amino acids. In many Saccharomycetes a unique alteration of this code affects the
translation of the CUG codon, which is normally translated as leucine. Most of the
species encoding CUG alternatively as serine belong to the Candida genus and were
grouped into a so-called CTG clade. However, the “Candida genus” is not a
monophyletic group and several Candida species are known to use the standard CUG
ancestor of the Candida, or to several branches independently leading to a
polyphyletic alternative yeast codon usage (AYCU). In order to resolve the
monophyly or polyphyly of the AYCU, we performed a phylogenomics analysis of 26
motor and cytoskeletal proteins from 60 sequenced yeast species. By investigating the
CUG codon positions with respect to sequence conservation at the respective
alignment positions we were able to unambiguously assign the standard code or
AYCU. Quantitative analysis of the highly conserved leucine and serine alignment
positions showed, that 61.1% and 17% of the CUG codons coding for leucine and
serine, respectively, are at highly conserved positions, while only 0.6% and 2.3% of
the CUG codons, respectively, are at positions conserved in the respective other
amino acid. Plotting the codon usage onto the phylogenetic tree revealed the
polyphyly of the AYCU with Pachysolen tannophilus and the CTG clade branching
independently within a time span of 30 to 100 Mya.
Keywords
Genetic code, codon reassignment, codon usage, evolution, Candida
-2-
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
translation. The codon identity could have been changed in a single branch, the
Introduction
The standard genetic code has been altered in many organisms. In eukaryotes, natural
alterations have been identified in mitochondria, in which the universal stop codon
UGA is translated as tryptophane (Yokobori et al. 2001; Watanabe & Yokobori
2011), in ciliates, in which the function of the stop codons UAA and UGA was
changed to code for glutamine (Tourancheau et al. 1995), and in some Candida yeasts
codon CUG translation was changed to serine. In taxonomic terms, the definition of
the genus Candida is not very specific. Candida species comprise a group of about
850 organisms (Robert et al. 2013), which can be distantly related as for example
Candida glabrata, Ca. albicans, and Ca. caseinolytica. Although it has been claimed
that almost all Candida species, with the exceptions of Ca. glabrata and Ca. krusei,
belong to a single clade (Butler et al. 2009), which is commonly referred to as
“Candida clade”, many analyses showed that Candida species are spread over the
entire Saccharomycetes clade (Diezmann et al. 2004; Kurtzman & Suzuki 2010;
Kurtzman et al. 2011; Kurtzman 2011). In addition, different names have been given
to the same yeast species in the past depending on the reproductive stage, in which
they had been identified. For example, the names Meyerozyma guilliermondii
(teleomorph), Ca. guilliermondii (anamorph), Yamadazyma guilliermondii and Pichia
guilliermondii are used for the same holomorph (whole fungus, including anamorph
and teleomorph).
-3-
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
(Ohama et al. 1993; Sugita & Nakase 1999; Santos et al. 2011), in which the leucine
Because Ca. albicans is the most prominent representative of the genus Candida and
was the first shown to use the alternative CUG encoding, the term “Candida clade” is
often used synonymous to “CTG clade” and most studies about CUG encoding
concentrate on Ca. albicans and closely related Candida species. Therefore,
depending on which node is used for defining the “Candida clade”, several species of
the branch containing Ca. albicans might decode CUG as leucine and many species
outside the “Candida clade” are also named Candida. Similarly, there are many nonCandida species decoding CUG as serine. In order to resolve this ambiguity, one of
single branch or to several branches independently. The latter scenario would be in
agreement with the proposed timing of the code alteration (~ 270 Ma) and the split of
the Saccharomyces and Candida clades about 170 Ma (Massey et al. 2003; Santos et
al. 2011) implying that the CUG codon was highly ambiguous for approximately 100
Myr. Species that branched within this time span could have adopted either of the two
decoding possibilities. Most phylogenetic analyses of yeast species used 28s rRNA,
18s rRNA, actin and translation elongation factor-1α for reconstructing single or
combined trees, which is a very successful approach when analyzing hundreds of
species (Kurtzman 2011). However, higher accuracy for tree topologies is obtained in
phylogenomic studies that use dozens to thousands of concatenated homologs in tree
computations (Fitzpatrick et al. 2006). For the latter approach it is important to
distinguish orthologs from paralogs. Only a few yeast species have been included in
phylogenomic studies so far.
In addition to assigning the CUG codon usage by phylogenetic comparisons it has
been suggested to base the translation of codon CUG on the presence of Q9 as the
-4-
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
the major questions is whether the change of identity of the CUG codon happened to a
major ubiquinone (Sugita & Nakase 1999). Ubiquinone (Coenzyme Q, CoQ) is an
electron carrier in the respiratory chain and involved in oxidative stress resistance
(Szkopińska 2000; Turunen et al. 2004). Most enzymes of the ubiquinone
biosynthesis pathway are conserved between Saccharomyces cerevisiae and Ca.
albicans. However, the Ca. albicans ubiquinone has a nonaprenyl side chain (Q9)
whereas the budding yeast has six isoprene units (Q6). Although a great variety of
CoQ types are present in the genus Candida (Schauer & Hanschke 1999), the species
analyzed at that time, which were known to decode CUG as serine, lacked galactose
Based on these ideas, an extensive study revealed 89 yeast strains to possess
coenzyme Q9 (except for Cl. lusitaniae) and to miss galactose in the cell (Sugita &
Nakase 1999). However, in vitro translation assays showed 11 of these species to use
codon CUG as leucine. The incongruence in many instances of the presence of Q9,
absence of galactose, and molecular phylogenetic data has also been reported by
others (Kurtzman & Robnett 1997; Diezmann et al. 2004; Kurtzman & Suzuki 2010).
Sixty yeast species (not counting strains and data under embargo) have already been
sequenced and assembled (http://www.diark.org, visited May 20, 2014). Only half of
these genomes, which include both yeasts of the “CTG clade” and of the
Saccharomyces clade, have been analyzed in detail yet. In addition, sequencing efforts
increasingly focus on genomes of rarely studied species, whose phylogenetic
relationships are not clear. For most of these genomes, gene predictions were even
done and made available using both the Standard Code and the Alternative Yeast
Nuclear Code (AYNC) resulting in mixed datasets. In order to determine whether
there was a time span with a CUG codon usage ambiguity or whether the codon usage
-5-
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
in the cells and possessed Q9, with the exception of Clavispora lusitaniae (Q8).
alteration could be attributed to a single event, we performed a phylogenomic analysis
of highly validated protein sequences of 26 members of the actin, actin-related,
tubulin, myosin, kinesin, dynein, and actin-capping protein families of the sequenced
yeast species. CUG codon translation has been assigned by analyzing the respective
amino acids of the sequences in the context of the sequence alignments.
Materials and Methods
Some fungal actin-related proteins and myosins have been extracted from previously
published datasets (Odronitz & Kollmar 2007; Hammesfahr & Kollmar 2012). The
sequences were updated based on newer genome assemblies if necessary and
corrected for CUG usage. The data for the other proteins and the species not included
in the published datasets have essentially been obtained as described (Odronitz &
Kollmar 2007). Shortly, the corresponding gene regions have been identified in
TBLASTN searches starting with the respective protein sequences of homologs of
Saccharomyces cerevisiae. The respective genomic regions were submitted to
AUGUSTUS (Stanke & Morgenstern 2005) to obtain gene predictions. However,
feature sets are only available for a few species of the Saccharomycetes clade.
Therefore, all hits were subsequently manually analyzed at the genomic DNA level.
When necessary, gene predictions were corrected by comparison with the homologs
already included in the multiple sequence alignments. Especially, the short exons at
the 5’ ends of the tubulin and actin genes were not always identified in automatic
gene predictions. EST data to help in the annotation are only available for a few
species in public databases.
-6-
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
Identification and Annotation of the Proteins
In the last years, genome-sequencing efforts have been extended from
sequencing species from new branches to sequencing closely related organisms. Here,
these species include for example Saccharomyces and Eremothecium species. Protein
sequences from these closely related species have been obtained by using the crossspecies functionality of WebScipio (Odronitz et al. 2008; Hatje et al. 2011).
Nevertheless, for all these genomes TBLASTN searches have been performed. With
this strategy, we sought to ensure that we would not miss more divergent protein
family homologs, which might have been derived by species-specific inventions or
therefore these cross-species gene reconstructions only worked for a few species.
All sequence-related data (protein names, corresponding species,
sequences, and gene structure reconstructions) and references to genome sequencing
centers are available at CyMoBase (http://www.cymobase.org, (Odronitz & Kollmar
2006)). A list of the analyzed species, their abbreviations as used in the alignments
and trees, as well as references and accession numbers is also available as
supplementary file S1, Supplementary Material online. WebScipio (Hatje et al. 2011;
Odronitz et al. 2008) was used to reconstruct the gene structure of each sequence.
Throughout this study we use the teleomorph names of all species.
Generating the Multiple Sequence Alignment
The protein sequences were added to the already existing multiple sequence
alignments (Odronitz & Kollmar 2007; Hammesfahr & Kollmar 2012). In detail, we
first aligned every newly predicted sequence to its supposed closest relative using
ClustalW (Thompson et al. 2002) and added it then to the multiple sequence
alignment. During the subsequent sequence validation process, we manually adjusted
-7-
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
duplications. However, the Saccharomycetes belong to the fast-evolving species and
the obtained alignment by removing wrongly predicted sequence regions and filling
gaps. Still, in those sequences derived from low-coverage genomes many gaps
remained. To maintain the integrity of exons preceded or followed by gaps, gaps
reflecting missing parts of the genomes were added to the multiple sequence
alignment. Sequence homologs of the same protein family (e.g., α-tubulin, β-tubulin,
and γ-tubulin) were kept in one single protein family alignment. The actin/actinrelated protein alignment therefore contains 11 sub-families, the CapZ alignment 2,
the tubulin alignment 3, the myosin alignment 3, and the kinesin alignment 6. Only
sequence alignments can be obtained from CyMoBase.
Computing Sequence Conservation
The residue conservation at alignment positions was calculated with the conservation
code toolbox as implemented by (Capra & Singh 2007). Conservation was estimated
with the property-entropy method, an entropy measurement refined with respect to
chemical properties of amino acids. Scores were calculated with conservation of
adjacent amino acids incorporated (window size 3) and not (window size 0). Except
for window size and scoring method, standard parameters were used. Unexpectedly,
although window sizes were applied, the rest of the alignment still seemed to have an
influence, albeit marginal, on the conservation scores. In addition, amino acids given
as “X” are replaced by hyphens “-“ by the software, which denote gap positions in the
alignment. Gap positions, however, indicate a biological relevant deletion of this
position in the sequence and thus have a strong influence on the conservation score.
Therefore, conservation estimates were performed on small sections of the
concatenated alignment including Schizosaccharomyectes sequences. For each serine
-8-
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
single dynein heavy chain homologs have been found in the analyzed species. The
and leucine position, the alignment was reduced to this position and 15 adjacent
positions in each direction. Sequences with CUG codons at the respective leucine or
serine position were removed from the alignment sections.
Computing and Visualising Phylogenetic Trees
For calculating phylogenetic trees, the alignments of the CapZ, actin and actin-related,
tubulin, kinesin, and the myosin protein families were split into sub-family alignments
and the respective alignments were concatenated (supplementary file S2,
Schizosaccharomycetes (Sc. pombe, Sc. octosporus, Sc. japonicus, and Sc. cryophilus)
were taken. Gblocks v.0.91b (Talavera & Castresana 2007) was run with standard
parameters, and half gap positions allowed to reduce the concatenated alignment from
28,202 amino acids to 9,788 amino acids in 230 blocks. The phylogenetic trees were
generated using four different methods: Neighbor-Joining (NJ), maximum-likelihood
(ML), Bayesian inference and split networks. 1) ClustalW v.2.0.10 (Thompson et al.
2002) was used to calculate unrooted trees with the NJ method. For each dataset,
bootstrapping with 1,000 replicates was performed. 2) ML analysis with estimated
proportion of invariable sites and bootstrapping (1.000 replicates) was performed
using RAxML v.7.3.1 (Stamatakis et al. 2008). To this end, ProtTest v.3.2 was used
first to determine the most appropriate of the available 112 amino acid substitution
models (Darriba et al. 2011). Within ProtTest, the tree topology was calculated with
the BioNJ algorithm and both the branch lengths and the model of protein evolution
were optimized simultaneously. The Akaike Information Criterion with a
modification to control for small sample size (AICc, with alignment length
representing sample size) identified the WAG model with gamma model of rate
-9-
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
Supplementary Material online). As outgroup, sequences from
heterogeneity and site-specific evolutionary rates to be the best. 3) The ML analysis
was repeated on an additional dataset generated from the concatenated alignment. In
this dataset, amino acids encoded by CUG codons were substituted by “X”. Gblocks
was invoked as described above, with standard parameters and half gap positions
allowed, to reduce the alignment to 8,903 positions in 297 blocks. The phylogenetic
tree was generated with RAxML under the WAG+Γ+F model and 1,000 replicates
were performed. 4) Posterior probabilities were generated using MrBayes v3.2.1.
(Ronquist & Huelsenbeck 2003). Using the mixed amino-acid option, two
performed. MrBayes used the WAG model for all protein alignments. The trees were
calculated based on a reduced dataset generated by Gblocks using standard
parameters allowing no gap positions. This dataset includes 5,038 positions in 131
blocks. Trees were sampled every 1,000th generation and the first 25% of the trees
were discarded as “burn-in” before generating a consensus tree. 5) An unrooted
phylogenetic split network was generated with SplitsTree v. 4.1.3.1 (Huson & Bryant
2006). The Neighbor-Net method as implemented in SplitsTree was used to identify
alternative splits. Phylogenetic trees and networks were visualized with FigTree
v.1.3.1 (Rambaut and Drummond) and SplitsTree v.4.13.1, respectively, and are
available as supplementary fig. S1, Supplementary Material online.
Estimating the divergence times
The divergence times of species using AYCU were estimated with a penalizedlikelihood approach as implemented in treePL (Smith & O’Meara 2012). The
underlying tree topology was generated with RAxML under the WAG+Γ+F model on
the alignment with CUG codons assigned. The splits between Sa. cerevisiae and Ca.
- 10 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
independent runs with 100,000 generations, 4 chains, and a random starting tree were
albicans, and Sa. cerevisiae and Sc. pombe (Heckman et al. 2001; Douzery et al.
2004) were constrained both individually and combined.
Results
Phylogeny of Sequenced Yeast Species
In order to obtain a reliable phylogeny of the sequenced yeast species we choose a
small-scale phylogenomics approach. The more data (i.e. orthologs) included in the
important parameter for the reliability of the computed trees is the quality of the
underlying sequence data. Here, we manually determined 26 motor and cytoskeletal
proteins for each of the 60 sequenced yeast species. These 26 proteins belong to six
major protein families, the actin/actin-related proteins, the tubulins, the myosins, the
kinesins, the dynein heavy chain proteins, and the actin capping proteins of the CapZ
complex. We included all proteins of these six families in the analysis except some
kinesins that are unique to certain species. Many of the actin and tubulin genes
contain very short exons of one to three residues that are not included in datasets of
automatic gene predictions. By manually inspecting the genomic DNA sequences we
could completely reconstruct all genes as long as the genomic sequences did not
contain gaps. The concatenated sequences consist of on average 19,300 residues per
species amounting to 28,202 alignment positions. Gblocks (Talavera & Castresana
2007) was used with less stringent and more stringent parameters to reduce the
alignment to 9,788 and 5,038 positions, respectively. NJ and ML methods were used
to construct trees from the extended alignment, and the Bayesian approach to compute
a tree from the shorter alignment (supplementary fig. S1, Supplementary Material
- 11 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
alignment the better and more robust the phylogenetic trees should become. Another
online). To evaluate the influence of misassigned CUG codons on the tree topology
all CUG codons were translated with “X” and the tree reconstruction repeated. The
topology of this unbiased tree is identical to that based on the alignment with assigned
CUG codons. As outgroup we included four Schizosaccharomyces species. The
constructed trees have almost the same topology with the major exceptions being the
placing of Kluyveromyces in the NJ tree, in which they group to Lachancea and
Eremothecium. To highlight alternative phylogenetic relationships between the
species we also constructed a phylogenetic network using the Neighbour-Net method
trees show that Yarrowia lipolytica is the earliest diverging Saccharomycetes. The
Saccharomycetaceae form a highly supported group together with the early separating
Pfaffomycetaceae (supplementary fig. S1, Supplementary Material online). In our
analysis, Pichia kudriavzevii and Ogataea parapolymorpha branch together with
Dekkera bruxellensis, Kuraishia capsulata, and Komagataella pastoris. This branch is
more closely related to the Ca. albicans species group than to the
Saccharomycetaceae (supplementary fig. S1, Supplementary Material online), but the
species encode CUG as leucine and not as serine. The species, for which the AYCU
has been determined previously (e.g., Ca. albicans, Ca. dubliniensis and Ca.
parapsilosis), are part of a highly supported group in all phylogenetic trees (fig. 1 and
supplementary fig. S1, Supplementary Material online). Therefore, we refer to this
group from here on as CTG clade. The CTG clade consists of several “Candida”
species including Ca. albicans and Ca. tenuis, and also many others such as species
from Metschnikowiaceae and Debaryomycetaceae. In all trees, Pachysolen groups
outside the CTG clade as sister to the Pichiaceae/Ogataea (supplementary fig. S1,
Supplementary Material online). In the phylogenetic network, Pachysolen is placed at
- 12 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
(fig. 1). In agreement with other studies (Kurtzman & Robnett 2013b, 2013a), the
the origin of the CTG clade but there are many connections supporting alternative
topologies. Interestingly, species of the Candida genus are spread all over the tree
grouping to the Saccharomycetes and to various subgroups of the CTG clade (fig. 1).
Conservation of Amino Acids at CUG Codon Positions in Class I Myosins
In order to determine, whether CUG codons are coding for conserved amino acids
(leucine or serine) we performed a thorough analysis of the alignments of all protein
subfamilies. As representative example a part of the alignment of the class I myosins
supplementary fig. S2, Supplementary Material online). This analysis provided
several surprising findings: 1) The class I myosin alignment contains 33 positions
with 100% leucine conservation but only 11 positions that show 100% serine
conservation. 2) Except for Wickerhamomyces anomalus and Candida tenuis (no
CUG codons in their class I myosin genes), and Cl. lusitaniae, Ca. tropicalis, and
Pachysolen (two to three CUG codons in their class I myosin genes but not at
conserved positions), all Saccharomycetaceae, Pfaffomycetaceae, Pichiaceae, and
Ogataea species, and D. bruxellensis, K. capsulata, K. pastoris, and Y. lipolytica,
have at least one CUG codon at one of the 100% conserved leucine positions. 3)
Except for the sequences of the Debaryomyces and Spathaspora species investigated
here, whose CUG positions are in almost all cases identical, none of the other closely
related species (e.g., Saccharomyces or Pichiaceae) have conserved CUG positions
(fig. 2). This is especially apparent for the six analyzed Saccharomyces species that
have duplicated class I myosins but do not have even a single conserved CUG
position between the two paralogs. In addition, CUG positions are not conserved
within any of the orthologs. There is one position in the alignment (position 322), at
- 13 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
is presented in figure 2 (the full class I myosin alignment is available in
which ten of the species of the CTG clade have a CUG-encoded serine whereas
serines, alanines and glycines are found at that position in the other species (fig. 2).
Conservation of Amino Acids at CUG Codon Positions in General
The analysis of the complete data with respect to amino acid similarity at alignment
positions, at which at least one leucine/serine is present, showed that positions with
leucines are stronger conserved than positions with serines although in total serines
have been observed at as many alignment positions as leucines (fig. 3A and
conservation of each position alone (window size = 0) and within its closest
environment (window size = 3). The analysis of CUG codons present at leucine
positions showed that CUG codons are not evenly distributed to all positions with
leucines but enriched at highly conserved alignment positions (fig. 3A). In contrast,
there are far less CUG codons at serine positions, and these are in general present at
less conserved positions compared with the leucine encoding CUG codons. However,
although small there are a considerable number of CUG codons at highly conserved
and thus discriminative serine positions (fig. 3A). In order to determine the amino
acid conservation, it is extremely important to resolve protein sub-family
relationships. For example, the kinesin-1 sub-family members have a highly
conserved serine at the same position where the kinesin-5 proteins have a highly
conserved leucine (fig. 3B).
Assignment of the Codon Usage Scheme by Amino Acid Conservation Patterns
In order to determine the CUG codon translation we investigated whether the CUG
codons within the analyzed sequences of each species are at highly conserved leucine
- 14 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
supplementary fig. S3, Supplementary Material online). We determined the
or serine positions in the alignment. Therefore, we first determined the percentage of
CUG codons at extremely conserved leucine and serine positions (conservation score
of ≥90%; fig. 4). The analysis resulted in three groups of species: The first group,
including Sa. cerevisiae, had 2.8-26.4% (on average 14.3%) of the CUG codons at
highly conserved leucine positions. The second group, including Ca. albicans, had
CUG codons at conserved serine positions although the total numbers were
considerably lower than those for the leucine positions (0.9 to 7.5%; on average
3.4%), Pachysolen tannophilus did not have any CUG codons at either conserved
a conservation score of more than or equal to 80% (fig. 4). In the same group of
species as at the conservation score of more than or equal to 90%, 8.4-44.4% (on
average 24.5%) of the CUG codons are at leucine positions. At that conservation
level, one of the CUG codons of Yarrowia is found at a conserved serine position. At
the same conservation score (≥80%), all other species have CUG codons at highly
conserved serine positions (2.2-11.8%; on average 6.1%). In five species, one to two
of the CUG codons are at conserved leucine positions (Candida metapsilosis, Ca.
tenuis, Cl. lusitaniae, Spathaspora arborariae and Millerozyma). Thus, the species
clearly separate into two groups, one with many CUG codons almost exclusively at
highly conserved leucine positions, and the other with CUG codons at serine positions
(fig. 4). At a modest sequence conservation level (conservation score ≥50%), 5591
(61.1%) of the CUG codons of the first group are at leucine positions (the maximum
is 70.2% of the CUG codons in Saccharomyces kudriavzevii). At this conservation
level 54 CUG codons (0.6%) are at modestly conserved serine positions (conservation
score ≥50%). At the same conservation level, 17% of the CUG codons of the second
group (in total 332 codons) are at serine positions, whereas 44 CUG codons (2.4 %)
- 15 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
position. A similar situation was found when analyzing the CUG codon distribution at
are at leucine positions with the same conservation level. The same pattern can be
observed when using scores calculated column wise instead of window wise (window
size 0; supplementary fig. S4, Supplementary Material online).
CUG Codon Positions Shared between Species
It was proposed that all CUG codons had been removed during the time of codon
ambiguity and later reintroduced within the Saccharomycetaceae and Candida
branches (Massey et al. 2003). If this model was true, species of the
positions within but not across these two branches. In order to evaluate CUG codon
position conservation we determined the number of CUG codons shared between
every two species (fig. 5). This CUG codon position analysis is independent from
CUG codon assignment. For comparison, we analyzed the conservation of the CUG
codon positions at all alignment positions (fig. 5, lower triangle) and at positions with
a modest conservation level of more than or equal to 50% (fig 5, upper triangle).
Independent of whether sequence conservation at alignment positions is required the
yeast species group into two distinct classes. Species of the first group containing the
Saccharomycetaceae, Pfaffomycetaceae, Pichiaceae, and Ogataea species, and
Dekkera bruxellensis, Komagataella pastoris, Kuraishia capsulate, and Yarrowia
lipolytica share considerably more CUG positions than those from the second group
formed by the CTG-clade species. When all CUG positions are considered, there are a
few positions shared between the two groups. However, if only CUG positions at a
modest conservation level of more than or equal to 50% are evaluated, the two groups
do not share any CUG positions. The few CUG positions shared at all alignment
positions between the two species groups reflect CUG positions in disordered and
- 16 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
Saccharomycetaceae and Candida would be expected to have shared CUG codon
non-conserved protein regions (see, e.g., the alignment of the class-1 myosin tail
regions in supplementary fig. S2, Supplementary Material online). Interestingly,
Pachysolen, does not share any CUG positions with any other species (fig. 5).
Mapping the Assigned Codon Usage onto the Species Tree
In order to determine the mono- or polyphyly of the AYCU assignment, we mapped
the assigned codon usage onto the reconstructed species tree (fig. 6). This mapping
supports the polyphyly of the AYCU assignment because the CTG clade and
Saccharomycetales. The CUG codon position conservation is also in accordance with
the species tree and the codon usage assignment. The CTG clade species group
together in the species tree and share CUG codon positions implying that at least
some of them have been introduced in the last common ancestor of the CTG clade.
Pachysolen diverged separate from the CTG clade and therefore does not share any
CUG codon positions with species from the CTG clade.
Timeline for the Separation of the CTG Clade
In order to estimate the time span, in which the CUG codon usage might have been
ambiguous, we applied published branching time estimations to our species trees.
However, the divergence time estimations for the splits between Archiascomycetes
and other Ascomycota and within the Saccharomycetes differ by more than 700 Myr
although extensive molecular and fossil record data have been used in respective
studies ((Heckman et al. 2001; Hedges et al. 2004; Douzery et al. 2004; Berbee &
Taylor 2007), for more references see (Hedges et al. 2006)). Therefore, we choose
some of the most extreme estimates for the calibration (table 1). The divergence time
- 17 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
Pachysolen branched at different points during the evolution of the
of the CTG clade was calibrated on the basis of the topology and branch lengths of the
ML tree under penalized-likelihood constraints as implemented in treePL (Smith &
O’Meara 2012). The splits of Sc. pombe and Sa. cerevisiae, and of Sa. cerevisiae and
Ca. albicans were set to 1,144 and 841 Ma (Heckman et al. 2001), and 587 and 235
Ma (Douzery et al. 2004), respectively. Under these two sets of constraints, the
separation of the CTG clade was estimated to 440 and 135 Ma, respectively (table 1),
implying that the reassignment of the CUG codon in the CTG clade happened at most
400 and 100 Myr, respectively, after the split of the Saccharomycetaceae and the
obtained for the emergence of Pachysolen indicating that ancient yeast species
emerging in the time of 220 to 190 Ma (740 to 640 Ma, respectively) could have
independently reassigned the CUG codon translation. This time span is the minimum
time of CUG codon ambiguity. The maximum time span depends on the last split
before the first appearance of the AYCU (split of the Saccharomycetaceae and the
branch containing the CTG clade, Pichiaceae, and Ogataea; fig. 6) and both the first
split after the last assignment of the AYCU (split of the branch containing Pichiaceae
and Ogataea from Pachysolen) and the divergence of the CTG clade (135 and 440
Ma, respectively). The maximum time span of codon ambiguity is thus estimated to
be 100 Myr (400 Myr, respectively).
Discussion
The AYCU has already been assigned to ten of the sequenced yeast genomes, Ca.
albicans (Jones et al. 2004; Butler et al. 2009), Ca. dubliniensis (Jackson et al. 2009),
Ca. parapsilosis (Butler et al. 2009), Ca. tropicalis (Butler et al. 2009), Cl. lusitaniae
- 18 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
CTG clade (fig. 6). Similarly, a date of at least 190 Ma (640 Ma, respectively) was
(Butler et al. 2009), Debaryomyces hansenii (Dujon et al. 2004; Butler et al. 2009),
Lodderomyces elongisporus (Butler et al. 2009), Mey. guilliermondii (Butler et al.
2009), Scheffersomyces stipites (Jeffries et al. 2007), and Spathaspora passalidarum
(Wohlbach et al. 2011) and the standard codon usage was shown for several of the
Saccharomyces species. Mass spectrometry and Edman sequencing of SMKT
demonstrated that Millerozyma farinosa also uses the AYCU (Suzuki et al. 2002). An
in vitro translation study showed 78 Candida species to use the AYCU including the
above listed Candida strains (Sugita & Nakase 1999), and in addition Ca.
recently (Wohlbach et al. 2011). Nevertheless, the gene prediction datasets for Ca.
tanzawaensis and Ca. tenuis available at the respective sequencing centers include
predictions both with AYCU and with standard codon usage. However, for highly
resolved phylogenetic studies, protein ortholog assignments and functional studies,
naming only a few applications, the accuracy of the protein sequences is important.
The best and most accurate way to determine a species codon usage would be the
experimental analysis of some proteins known to contain CUG codons by, for
example, mass spectrometry. However, as long as proteins cannot be enriched this is
extremely difficult to conduct, and the most prevalent endogenous proteins might not
contain CUG codons. For example, our analysis showed that actin genes from only a
few yeasts (11 out of 60 analyzed) contain CUG codons. Other approaches only
provide indirect evidence. As discussed, the presence or absence of cellular
components such as galactose or Q9 does not correlate with CUG codon usage. Here,
we hypothesized, that the codon usage can unambiguously be assigned by analyzing
three types of data: A highly resolved phylogenetic tree, sequence conservation at
CUG coding positions, and conservation of CUG codon positions across species. To
- 19 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
tanzawaensis and Ca. tenuis, for which the genome sequence had been assembled
obtain significant numbers of sequences we were restricted to analyse sequenced
genomes. Up to May 2014, genomes of 60 yeast species have been sequenced,
assembled and released. We did not include genomes in the analysis, which were
sequenced with low-coverage, and only included one strain per species (e.g., 73
different Sa. cerevisiae strains have been sequenced, of which only the reference
strain was included in our analysis). To avoid bias by misassigning CUG codons we
excluded all CUG positions from the alignment and the sequence conservation
calculations.
proteins, which are key-components of eukaryotic cells. We decided to only use
proteins in the analysis that are at least present in all unikonts (the eukaryotic superkingdom including Fungi, Metazoa, Apusozoa and Amoebozoa) to ensure that we do
not have extensive missing data. The topology of the species tree was almost identical
independent of the reconstruction method used (NJ, ML and Bayesian; fig. 6).
Although this particular set of species has never been analyzed before, the branching
of most sub-branches and individual species is comparable to those found in
previously published yeast trees (see references in Introduction section). The species
known to use either the standard code (Saccharomycetaceae) or the AYCU (see
above) grouped into two distinct clades. It seemed obvious to assume, but had to be
shown, that all other species within these two groups, for which the codon usage has
not yet been confirmed, use the same codon usage as the known species. In terms of a
mono- or polyphyletic origin of the AYCU it would be necessary to resolve the codon
usage of the species of the other major branches (Phaffomycetaceae clade and the
- 20 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
For the reconstruction of the yeast tree we concatenated 26 cytoskeletal and motor
branch containing the Pichiaceae and Ogataea) and the species that diverged before
the combined Saccharomycetaceae/CTG clades such as Y. lipolytica.
To determine the most probable CUG codon usage for each species we analyzed the
amino acid conservation at the respective CUG codon positions in the protein
sequence alignments. Amino acid positions highly conserved over hundreds of
million years (divergence time of Schizosaccharomyces and Saccharomycetes) are
expected to represent structurally and functionally important sites in proteins. A small
residues in the other species, seems very unlikely. Similarly, large hydrophobic
residues at positions conserved in small and polar amino acids should be strongly
disfavoured. Our data show that the assignment of the standard codon usage is
facilitated by the enrichment of CUGs at highly conserved leucine positions. In this
respect only determining the class I myosin of a species with unknown CUG codon
usage and comparing this sequence to the available data (fig. 2 and supplementary fig.
S2, Supplementary Material online) would in most cases be enough to find out
whether the respective species uses the standard or the alternative yeast CUG codon
translation. If CUG codons can be mapped to conserved leucine positions, it will be
highly probable that the respective species translates CUG as leucine. The AYCU will
be assigned to a species if CUG codons are mapped to highly conserved serine
positions or if CUG codons are not present at highly conserved leucine positions
(proof by contradiction). In addition to the class I myosins, several proteins could be
identified and compared until a conserved serine position was found. The latter
approach would lead to an unambiguous assignment and was used in our analysis
here. It has been proposed that CUG codons were not reintroduced randomly in CTG
- 21 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
polar residue at a position, which is characterized by conserved large hydrophobic
clade genomes but by avoiding structurally sensitive sites (Rocha et al. 2011). This is
in agreement with the different biochemical properties of leucine and serine and the
associated consequences for protein folding and solubility. Leucine positions are in
general stronger conserved than serine positions, leucines are usually located in the
core of protein domains, and mutations at leucine positions are therefore restricted.
Compared with serine, the introduction of leucine encoding CUG codons is possible
by synonymous mutations from CUN codons or by transition from thymine from the
UUG codon. Transitions are more favourable and generated at higher frequencies than
conservation of the CUG codons within the yeast species using the Standard Code. In
contrast, mutating an existing serine codon to a CUG codon would require at least two
mutations. Mutations from other codons than leucine or serine codons would either
require the less likely transversions (e.g., GUG to CUG) or very unusual amino acid
mutations (e.g., from proline to serine). Therefore, it is not surprising to find less
CUG codons in CTG clade species, and those codons at less conserved positions.
Overall, we could demonstrate, that especially the CUG codons coding for leucine,
and also a substantial number of the CUG codons in species using the AYCU, are at
conserved alignment positions (fig. 4). Although seven of the CUG codons coding for
leucine and serine, respectively, are at positions supporting the respective opposite
codon translation based on an amino acid conservation level of more than or equal to
80%, the majority of the CUG codon positions supports either the standard or the
AYCU. This is in accordance with protein structure and in vivo studies showing that
single point mutations and even partial reversion of the CUG identity are tolerated in
Ca. albicans although large numbers of mutations cause profound cellular changes
such as morphological variation and altered gene expression (Cutfield et al. 2000;
- 22 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
transversions. Such silent substitutions seem to be very common given the low
Gomes et al. 2007; Miranda et al. 2007; Rocha et al. 2011; Miranda et al. 2013;
Bezerra et al. 2013). Because we excluded all CUG codon positions from the analysis,
our assignments are independent from previous knowledge. Our data unambiguously
supports the assignment of the standard codon usage to the Saccharomycetaceae, and
the AYCU to the eight species, which were known to translate CUG as serine. From
these data we conclude, that the CUG codon assignment can unambiguously be
resolved by analyzing a considerable amount of sequences and that our CUG codon
translation assignment is robust. Therefore, we assigned the AYCU to the CTG clade
assignment of the CUG codon usage, we acknowledge that the final judgement should
come from experiments.
When plotting the assigned AYCU onto the species in the phylogenetic trees it
became evident that the CUG codon usage must have been ambiguous for a long time
in the ancestry of the Saccharomycetes leading to polyphyly of the AYCU (fig. 6).
This means that ancestral species that emerged during this time span could have
adopted either the standard or the AYC usage. In the case of a long time span of
codon ambiguity, the original CUG codons in the ancestor of the Saccharomycetes
were most likely erased and later reintroduced in the emerging branches either as
leucine codons or as serine codons. In the less likely and mutually exclusive scenario,
the CUG ambiguity appeared only once in the ancestor of the CTG clade. The
combination of the reconstructed species trees with our CUG codon assignment
strongly supports the first scenario of a polyphyletic origin of the AYCU.
- 23 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
species and Pachysolen. Although our data strongly support an unambiguous
Based on the phylogeny of the analyzed species Pachysolen does not group to the
CTG clade. In addition, Pachysolen does not have any CUG codons at highly
conserved leucine positions, but at conserved serine positions. The phylogenetic
grouping of Pachysolen has not been analyzed in detail so far, and not at all in the
context of a taxonomic sampling as used in our study. The evolutionary grouping and
CUG codon usage were also not analyzed in the study of the Pachysolen draft genome
sequence (Liu et al. 2012). A previous study based on 28s rDNA and 18s rDNA only
resolved the Saccharomycetes but not the grouping of Pachysolen and none of the
addition, support for the branchings was low (Suh et al. 2006). In contrast, we used a
phylogenomics approach for the tree reconstruction resulting in strongly supported
branching of Pachysolen.
The separate branching of Pachysolen is in agreement with the previously proposed
model of a long time span of codon usage ambiguity (Sugita & Nakase 1999). If the
CUG codon usage was ambiguous for a long time, it would be natural to assume that
species diverged within this time span, and that those species could have adapted
either of the usages. Supposed, that the codon usage was ambiguous for a long time, it
seems highly non-parsimonious, that either species did not branch within this time
span or that all species except the ancestor of the CTG clade adapted the standard
codon usage. A monophyletic group of the species using the AYCU is also highly
unlikely given the missing conservation of CUG codon positions between the CTG
clade species and Pachysolen. To date the codon reassignments we dated the ML tree
based on various separation time estimates for the splits of Sc. pombe and Sa.
cerevisiae, and Sa. cerevisiae and Ca. albicans (fig. 6 and table 1). Accordingly, there
- 24 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
other clades shown to be monophyletic in our trees (including the CTG clade). In
is a maximum time span of codon ambiguity of 100-400 Myr ranging from the first
possible appearance of the serine tRNASerCAG (separation of the Saccharomycetes and
the branch containing the CTG clade, the Pichiacea and Ogataea) to the split of the
most basal species of the CTG clade (Ca. tenuis). Assuming an unambiguous usage of
the CUG codon in Pachysolen and the CTG clade, the minimal time span of CUG
codon ambiguity was 30 to 100 million years (fig. 6 and table 1). This is well in
agreement with a previous study, which suggested an ambiguous usage of the CUG
codon for 100 million years (Sugita & Nakase 1999). However, the presented time
assume that further sequenced species will change the time estimates in two directions
(this might already happen by analyzing the yeast species whose data is still under
embargo). The inclusion of more basal CTG clade species will decrease the current
codon ambiguity time. The identification of further species using the AYCU, which
do not group to the CTG clade or to Pachysolen, instead might increase both the
codon ambiguity and codon reassignment time ranges. Such species include yeasts,
which might even branch before the split of the Saccharomycetaceae and the CTG
clade/Pichiacea/Ogataea, as well as additional species diverging in the sister branch of
the CTG clade, and species separating early in the Saccharomycetaceae branch.
Acknowledgments
We would like to thank Prof. Christian Griesinger for his continuous generous
support, and Prof. Toni Gabaldón and Dr. Qinhong Wang for their permission to use
the Ca.metapsilosis CANME and Ca.maltosa Xu316 genome assembly data prior to
publication. This project has been funded by the Deutsche Forschungsgemeinschaft
- 25 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
estimates are based on the species analyzed in this study. It is well reasonable to
DFG Grant KO 2251/13-1 to M.K. and a Synaptic Systems fellowship to S.M., and
was partly supported by the Göttingen Graduate School of Neurosciences and
Molecular Biosciences (DFG Grants GSC 226/1 and GSC 226/2).
Authors contribution
SM and MK assembled and annotated genes, analyzed the data and wrote the
manuscript.
Berbee ML, Taylor JW. 2007. Rhynie chert: a window into a lost world of complex
plant-fungus interactions. New Phytol. 174:475–479. doi: 10.1111/j.14698137.2007.02080.x.
Bezerra AR et al. 2013. Reversion of a fungal genetic code alteration links proteome
instability with genomic and phenotypic diversification. Proc. Natl. Acad. Sci. U.S.A.
110:11079–11084. doi: 10.1073/pnas.1302094110.
Butler G et al. 2009. Evolution of pathogenicity and sexual reproduction in eight
Candida genomes. Nature. 459:657–662. doi: 10.1038/nature08064.
Capra JA, Singh M. 2007. Predicting functionally important residues from sequence
conservation. Bioinformatics. 23:1875–1882. doi: 10.1093/bioinformatics/btm270.
Crooks GE, Hon G, Chandonia JM, Brenner SE. 2004. WebLogo: a sequence logo
generator. Genome Res. 14:1188 – 1190.
Cutfield JF, Sullivan PA, Cutfield SM. 2000. Minor structural consequences of
alternative CUG codon usage (Ser for Leu) in Candida albicans exoglucanase. Protein
Eng. 13:735–738.
Darriba D, Taboada GL, Doallo R, Posada D. 2011. ProtTest 3: fast selection of bestfit models of protein evolution. Bioinformatics. 27:1164–1165. doi:
10.1093/bioinformatics/btr088.
Diezmann S, Cox CJ, Schönian G, Vilgalys RJ, Mitchell TG. 2004. Phylogeny and
evolution of medical species of Candida and related taxa: a multigenic analysis. J.
Clin. Microbiol. 42:5624–5635. doi: 10.1128/JCM.42.12.5624-5635.2004.
- 26 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
References
Douzery EJP, Snell EA, Bapteste E, Delsuc F, Philippe H. 2004. The timing of
eukaryotic evolution: Does a relaxed molecular clock reconcile proteins and fossils?
Proc Natl Acad Sci U S A. 101:15386–15391. doi: 10.1073/pnas.0403984101.
Dujon B et al. 2004. Genome evolution in yeasts. Nature. 430:35–44. doi:
10.1038/nature02579.
Fitzpatrick DA, Logue ME, Stajich JE, Butler G. 2006. A fungal phylogeny based on
42 complete genomes derived from supertree and combined gene analysis. BMC
Evol. Biol. 6:99. doi: 10.1186/1471-2148-6-99.
Gomes AC et al. 2007. A genetic code alteration generates a proteome of high
diversity in the human pathogen Candida albicans. Genome Biol. 8:R206. doi:
10.1186/gb-2007-8-10-r206.
Hatje K et al. 2011. Cross-species protein sequence and gene structure prediction with
fine-tuned Webscipio 2.0 and Scipio. BMC Res. Notes. 4:265. doi: 10.1186/17560500-4-265.
Heckman DS et al. 2001. Molecular Evidence for the Early Colonization of Land by
Fungi and Plants. Science. 293:1129–1133. doi: 10.1126/science.1061457.
Hedges SB, Blair JE, Venturi ML, Shoe JL. 2004. A molecular timescale of eukaryote
evolution and the rise of complex multicellular life. BMC Evolutionary Biology. 4:2.
doi: 10.1186/1471-2148-4-2.
Hedges SB, Dudley J, Kumar S. 2006. TimeTree: a public knowledge-base of
divergence times among organisms. Bioinformatics. 22:2971–2972. doi:
10.1093/bioinformatics/btl505.
Huson DH, Bryant D. 2006. Application of phylogenetic networks in evolutionary
studies. Mol. Biol. Evol. 23:254–267. doi: 10.1093/molbev/msj030.
Jackson AP et al. 2009. Comparative genomics of the fungal pathogens Candida
dubliniensis and Candida albicans. Genome Res. 19:2231–2244. doi:
10.1101/gr.097501.109.
Jeffries TW et al. 2007. Genome sequence of the lignocellulose-bioconverting and
xylose-fermenting yeast Pichia stipitis. Nat. Biotechnol. 25:319–326. doi:
10.1038/nbt1290.
Jones T et al. 2004. The diploid genome sequence of Candida albicans. Proc. Natl.
Acad. Sci. U.S.A. 101:7329–7334. doi: 10.1073/pnas.0401648101.
Kurtzman C, Fell JW, Boekhout T. 2011. The Yeasts: A Taxonomic Study. 5th edition.
London, Burlington, San Diego: Elsevier.
- 27 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
Hammesfahr B, Kollmar M. 2012. Evolution of the eukaryotic dynactin complex, the
activator of cytoplasmic dynein. BMC Evol. Biol. 12:95. doi: 10.1186/1471-2148-1295.
Kurtzman CP. 2011. Phylogeny of the ascomycetous yeasts and the renaming of
Pichia anomala to Wickerhamomyces anomalus. Antonie Van Leeuwenhoek. 99:13–
23. doi: 10.1007/s10482-010-9505-6.
Kurtzman CP, Robnett CJ. 2013a. Alloascoidea hylecoeti gen. nov., comb. nov.,
Alloascoidea africana comb. nov., Ascoidea tarda sp. nov., and Nadsonia starkeyihenricii comb. nov., new members of the Saccharomycotina (Ascomycota). FEMS
Yeast Res. 13:423–432. doi: 10.1111/1567-1364.12044.
Kurtzman CP, Robnett CJ. 1997. Identification of clinically important ascomycetous
yeasts based on nucleotide divergence in the 5’ end of the large-subunit (26S)
ribosomal DNA gene. J. Clin. Microbiol. 35:1216–1223.
Kurtzman CP, Robnett CJ. 2013b. Relationships among genera of the
Saccharomycotina (Ascomycota) from multigene phylogenetic analysis of type
species. FEMS Yeast Res. 13:23–33. doi: 10.1111/1567-1364.12006.
Liu X, Kaas RS, Jensen PR, Workman M. 2012. Draft genome sequence of the yeast
Pachysolen tannophilus CBS 4044/NRRL Y-2460. Eukaryotic Cell. 11:827. doi:
10.1128/EC.00114-12.
Massey SE et al. 2003. Comparative evolutionary genomics unveils the molecular
mechanism of reassignment of the CTG codon in Candida spp. Genome Res. 13:544–
557. doi: 10.1101/gr.811003.
Miranda I et al. 2007. A genetic code alteration is a phenotype diversity generator in
the human pathogen Candida albicans. PLoS ONE. 2:e996. doi:
10.1371/journal.pone.0000996.
Miranda I et al. 2013. Candida albicans CUG mistranslation is a mechanism to create
cell surface variation. MBio. 4: pii e00285-13. doi: 10.1128/mBio.00285-13.
Odronitz F, Kollmar M. 2007. Drawing the tree of eukaryotic life based on the
analysis of 2,269 manually annotated myosins from 328 species. Genome Biol.
8:R196. doi: 10.1186/gb-2007-8-9-r196.
Odronitz F, Kollmar M. 2006. Pfarao: a web application for protein family analysis
customized for cytoskeletal and motor proteins (CyMoBase). BMC Genomics. 7:300.
doi: 10.1186/1471-2164-7-300.
Odronitz F, Pillmann H, Keller O, Waack S, Kollmar M. 2008. WebScipio: an online
tool for the determination of gene structures using protein sequences. BMC
Genomics. 9:422. doi: 10.1186/1471-2164-9-422.
Ohama T et al. 1993. Non-universal decoding of the leucine codon CUG in several
Candida species. Nucleic Acids Res. 21:4039–4045.
- 28 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
Kurtzman CP, Suzuki M. 2010. Phylogenetic analysis of ascomycete yeasts that form
coenzyme Q-9 and the proposal of the new genera Babjeviella, Meyerozyma,
Millerozyma, Priceomyces, and Scheffersomyces. Mycoscience. 51:2–14. doi:
10.1007/s10267-009-0011-5.
Rambaut A, Drummond A. 2009. FigTree v1.3.1.
http://tree.bio.ed.ac.uk/software/figtree/, visited May 9, 2014.
Robert V et al. 2013. MycoBank gearing up for new horizons. IMA Fungus. 4:371–
379. doi: 10.5598/imafungus.2013.04.02.16.
Rocha R, Pereira PJB, Santos MAS, Macedo-Ribeiro S. 2011. Unveiling the structural
basis for translational ambiguity tolerance in a human fungal pathogen. PNAS.
108:14091–14096. doi: 10.1073/pnas.1102835108.
Ronquist F, Huelsenbeck JP. 2003. MrBayes 3: Bayesian phylogenetic inference
under mixed models. Bioinformatics. 19:1572–1574.
Santos MAS, Gomes AC, Santos MC, Carreto LC, Moura GR. 2011. The genetic
code of the fungal CTG clade. C. R. Biol. 334:607–611. doi:
10.1016/j.crvi.2011.05.008.
Smith SA, O’Meara BC. 2012. treePL: divergence time estimation using penalized
likelihood for large phylogenies. Bioinformatics. 28:2689–2690. doi:
10.1093/bioinformatics/bts492.
Stamatakis A, Hoover P, Rougemont J. 2008. A rapid bootstrap algorithm for the
RAxML Web servers. Syst. Biol. 57:758–771. doi: 10.1080/10635150802429642.
Stanke M, Morgenstern B. 2005. AUGUSTUS: a web server for gene prediction in
eukaryotes that allows user-defined constraints. Nucleic Acids Res. 33:W465–467.
doi: 10.1093/nar/gki458.
Sugita T, Nakase T. 1999. Non-universal usage of the leucine CUG codon and the
molecular phylogeny of the genus Candida. Syst. Appl. Microbiol. 22:79–86. doi:
10.1016/S0723-2020(99)80030-7.
Suh S-O, Blackwell M, Kurtzman CP, Lachance M-A. 2006. Phylogenetics of
Saccharomycetales, the ascomycete yeasts. Mycologia. 98:1006–1017.
Suzuki C, Kashiwagi T, Hirayama K. 2002. Alternative CUG codon usage (Ser for
Leu) in Pichia farinosa and the effect of a mutated killer gene in Saccharomyces
cerevisiae. Protein Eng. 15:251–255.
Szkopińska A. 2000. Ubiquinone. Biosynthesis of quinone ring and its isoprenoid side
chain. Intracellular localization. Acta Biochim. Pol. 47:469–480.
Talavera G, Castresana J. 2007. Improvement of phylogenies after removing
divergent and ambiguously aligned blocks from protein sequence alignments. Syst
Biol. 56:564–577. doi: 10.1080/10635150701472164.
Thompson JD, Gibson TJ, Higgins DG. 2002. Multiple sequence alignment using
ClustalW and ClustalX. Curr Protoc Bioinformatics. Chapter 2:Unit 2.3. doi:
10.1002/0471250953.bi0203s00.
- 29 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
Schauer F, Hanschke R. 1999. [Taxonomy and ecology of the genus Candida].
Mycoses. 42 Suppl 1:12–21.
Tourancheau AB, Tsao N, Klobutcher LA, Pearlman RE, Adoutte A. 1995. Genetic
code deviations in the ciliates: evidence for multiple and independent events. EMBO
J. 14:3262–3267.
Turunen M, Olsson J, Dallner G. 2004. Metabolism and function of coenzyme Q.
Biochim. Biophys. Acta. 1660:171–199.
Watanabe K, Yokobori S-I. 2011. tRNA Modification and Genetic Code Variations in
Animal Mitochondria. J Nucleic Acids. 2011:623095. doi: 10.4061/2011/623095.
Wohlbach DJ et al. 2011. Comparative genomics of xylose-fermenting fungi for
enhanced biofuel production. Proc. Natl. Acad. Sci. U.S.A. 108:13212–13217. doi:
10.1073/pnas.1103039108.
Figure Legends
Figure 1 Phylogenetic relationship between Saccharomycetes
The unrooted phylogenetic network was generated using the Neighbor-Net method as
implemented in SplitsTree 4.1.3.1. The Schizosaccharomycetes species were included
as outgroup. The network strongly supports the Saccharomycetaceae and the CTG
clade (highlighted in orange). The grouping of Pachysolen tannophilus is not
unambiguously resolved. Species of the genus "Candida" are highlighted in blue
(teleomorph names) and green (if anamorphs of the species are called "Candida")
showing the paraphyly (or mis-assingment) of this genus. Orange and purple dots
mark species, for which alternative or standard codon usage has already been shown
elsewhere.
- 30 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
Yokobori S, Suzuki T, Watanabe K. 2001. Genetic code variations in mitochondria:
tRNA as a major determinant of genetic code plasticity. J. Mol. Evol. 53:314–326.
doi: 10.1007/s002390010221.
Figure 2 Sequence alignment of the yeast class I myosins highlighting leucines
and serines encoded by CUG
The protein sequence alignment represents part of the class I myosin alignment (for
the complete alignment see fig. S2, Supplementary Material online). Numbers on the
left denote the residue numbers of the first amino acids of the sequences in this
section of the alignment. All CUG positions occurring in the aligned class I myosin
genes are highlighted. We assigned the most probable translation scheme to each
species and translated the CUG codons accordingly. Blue and green boxes indicate
Figure 3 Conservation of serine and leucine positions
A) The charts show the amino acid conservation at all alignment positions of the
Gblocks reduced concatenated alignment of the 26 cytoskeletal and motor proteins, at
which at least one leucine (upper charts) or one serine (lower charts) is present. The
sequence conservation has been determined based on the Property-Entropy
divergence, as described in (Capra & Singh 2007). With a window size of 0, each
column is scored independently (left row), while the surrounding 3 columns are also
taken into account with a window size of 3 (right row). Blue bars represent the
number of alignment positions with a conservation score for leucine and serine
residues, respectively, within the given half-bounded intervals. Red bars denote the
number of alignment positions with respect to conservation, at which at least one
CUG codon is present independent of its translation. Green bars give the total
numbers of CUG codons at the respective alignment positions. B) The weblogos
(Crooks et al. 2004) show the sequence conservation of two kinesin sub-families,
kinesin-1 and kinesin-5, within the family-defining motor domain around the highly
- 31 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
CUG codons coding for leucine and serine, respectively.
conserved switch II and α-helix α4 motifs. At the position within α4 marked by a
grey bar kinesin-1 sequences contain a highly conserved serine while kinesin-5
sequences contain a highly conserved leucine indicating the need to resolve subfamily relationships when determining CUG codon usage by sequence conservation.
Figure 4 CUG codon usage at conserved leucine and serine positions
The graph presents the CUG codon usage with respect to alignment position
alignment positions with conservation scores of ≥90%, ≥80%, and ≥50% and a
window size of 3, respectively, in the concatenated alignment of the 26 motor and
cytoskeletal proteins. Red and blue colours denote the percentages of CUG codons
present at alignment positions enriched in serines and leucines, respectively. For
comparison, we plotted the percentages of CUG codons at positions of the assigned
codon translation to the left (% CUG codons at leucine positions for species using the
standard code and % CUG codons at serine positions for species using the AYCU)
and the percentages of CUG codons present at alignment positions enriched in the
respective other amino acid to the right. When considering only highly conserved
alignment positions (≥90% conserved) the CUG codon translation assignment is
unambiguous. Species using the AYCU are highlighted in bold.
Figure 5 CUG codon positions conserved between species
The heatmap represents the number of CUG codon positions shared between every
two species. The upper triangle shows the number of shared CUG codon positions at
those positions in the concatenated alignment of the full-length sequences, which have
- 32 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
conservation. For each species we determined the percentage of CUG codons at
a conservation score of at least 50%, the lower triangle the number of shared CUG
codons at all alignment positions. The diagonal represents the total number of CUG
codons in the respective species. The number of CUG codons is coloured on a
logarithmic scale. Species encoding CUG as serine are typed in bold.
Figure 6 Timeline of the CUG codon reassignment
The tree presents the maximum-likelihood topology generated under the Γ + WAGF
model in RAxML showing branch lengths for the concatenated alignments of 26
bootstrap replicates), MrBayes (posterior probabilities) and ClustalW trees (1,000
replicates) is indicated (for more details see fig. S1, Supplementary Material online).
Species using the AYCU are highlighted in bold. With the splits of Sc. pombe and Sa.
cerevisiae and Sa. cerevisiae and Ca. albicans set to 587 mya and 235 mya,
respectively, the divergence of the CTG clade was estimated to 190 mya using treePL
(Smith & O’Meara 2012). The scale bar denotes amino acid substitutions per site. The
width and colour of the branches to extant species represent the total number of CUG
codons in the respective concatenated sequences.
Tables
Table 1 Estimation of the time span of the CUG codon reassignment
Sc. pombe-Sa.
cerevisiae
Sa. cerevisiae-Ca.
albicans
CUG codon
ambiguity
CUG codon
reassignment
( 1,850
841a
520-841
740-855)
1,144a
730
415-730
605-695
- 33 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
cytoskeletal and motor proteins. Support for major branches of the RAxML (1,000
1,144a
841a
440-841
640-740
485
235b
135-235
195-225
587b
360
205-360
295-340
587b
235b
135-235
190-220
Estimated times in million years for the split between Sa. cerevisiae and Ca. albicans,
Sa. cerevisiae and Sc. pombe and the reassignment of the CUG codon were calculated
with a Penalized-likelihood program for the phylogenetic tree generated with RAxML
a
Time constrains for the splits between Sa. cerevisiae and Ca. albicans, and Sa.
cerevisiae and Sc. pombe as reported by (Heckman et al. 2001) and b(Douzery et al.
2004), respectively, were included both individually and combined. For each CUG
codon reassignment time computation, constrained time estimates are typed in bold.
The CUG codon reassignment time estimates are given as differences between the
time estimates for the divergence of P. tannophilus from the Pichiaceae and Ogataea
species (first number), and the first occurrence of the tRNASerCAG (second number;
separation of the CTG clade from the branch containing P. tannophilus, the
Pichiaceae and Ogataea species). The CUG codon ambiguity time estimates, however,
are given as differences between the time estimates for the latest possible fixation of
the tRNASerCAG (first number; separation of Ca. tenuis, the most basal species in the
CTG clade) and the earliest possible invention of the tRNASerCAG (second number;
separation of the Saccharomycetaceae and the CTG clade). The time estimate for the
divergence of Sc. pombe and Sa. cerevisiae (1,850 mya) based on the constraint of
841 mya for the split of Sa. cerevisiae and Ca. albicans predates the emergence of
eukaryotes and should be ignored.
- 34 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
using the WAG+Γ+F model.
Supporting Information
Figure S1 Phylogenetic trees
This file contains the phylogenetic trees generated with ClustalW, RAxML and
MrBayes based on the concatenated alignments of 26 motor and cytoskeletal proteins.
Bootstrap support values and posterior probabilities are reported in absolute values
dated by constraining the time for the splits between Sa. cerevisiae and Ca. albicans,
Sa. cerevisiae and Sc. pombe is included. The scale bar denotes million years.
Moreover, a RAxML tree was generated on a modified alignment, in which CTG
encoded amino acids were replaced by “X”. In all trees, major branches as well as the
“CTG-clade” are highlighted with colours.
Figure S2 Alignment of the yeast class I myosins with all CUG positions
highlighted.
All CUG positions occurring in the aligned class I myosin genes are highlighted. Blue
and green boxes indicate CUG codons coding for leucine and serine, respectively.
- 35 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
(ClustalW) and relative values (RAxML and MrBayes). In addition, the RAxML tree
Figure S3 Conservation of serine and leucine positions
Opposed to fig. 3, alignment positions were scored on all sequences of our dataset,
including inferred CUG translations. Scores were calculated with window sizes of 0
(charts on the left) and 3 (charts on the right).
Figure S4 CUG codon usage at conserved leucine and serine positions
Alignment positions were scored without sequences containing CUG codons.
Opposed to fig. 4, the window size for score computations was set to 0.
This table lists the teleomorph name (used as reference) of every analyzed species,
names of anamorphs, alternatively used scientific names and the species abbreviations
as used throughout this analysis. For every species the number of CapZ, actin and
actin-related, tubulin, kinesin, myosin, and dynein heavy chain genes is given.
Genome analyses, if published, accession numbers and sequencing centres are
referenced.
File S2 Concatenated alignment
This zip-compressed file contains the concatenated multiple sequence alignment of
the 26 cytoskeletal and motor proteins.
File S3 Conservation of serine and leucine positions
This file contains the raw data to fig. 3 and fig. S3.
- 36 -
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
File S1 List of analyzed species
File S4 CUG codon usage at conserved leucine and serine positions
This file contains the raw data to fig. 4 and fig. S4.
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
- 37 -
ru
ab
ck
erh
Wi
om
rha
rap
isisp
ora
Tetra
nd
am
cke
Tet
rli
pisis
pora
Vande
ne
yc
mo
es
my
ces
blat
tae
ra
in
an
phaffi
i CBS
i
far
S
B
C
67
ii C
7
BS
C
09
88
sN
rrii
R
NR RL
RL Y-3
Y-1 66
031
628
0.1
4
4417
Yarrowia lipolytica CLIB122
olyspo
ra DSM
delbrueckii C
ii
n
e
s
an
en
r. h
s
a
n
m
v
ii
zy s ha
n
e
o
s 4
r
e
98
4
li le myc s hanC 23
1
YL
M ryo yce TC
R
NR
a om M
s
i
b
20
u
7
y
e
2
n
r
4
e
a
D eb
TCC
da t
A
i
e
D
d
a
Can
itani
s
u
l
ra
77
2
o
a
p
l
o
s
i
c
i
fruct
Clav
a
i
w
o
k
ni
Metsch
6260
C
C
T
A
ii
d
n
o
guillierm
Meyerozyma
a
BR
alu
zyma p
Torulaspora
ii N
om
cife
CBS
rwalto
jad
a
s
no
4
6
70
BS 1146
70294
Schizosaccharomyces
osporus
Schizoct
osacch
cryopharomyces
ilus
Schi
zosa
ccha
p
omb romyc
Sch
e
es
izos
japaocchar
nicuomy
s ces
Zygosaccharomyces bailii ISA1307
ii CBS732
x
u
o
r
s
e
c
y
m
o
Zygosacchar
ns
is
N
ia
R
CB
nia
a
R
f
L
ric
nag
Y- S 4
a
Nak
a
n
12 21
nis
aC
ase
63
h
om
ii C
BS
yce
BS
25 0
s ba
8
17
797
cilli
spo
rus
Cand
CBS
ida g
772
labra
0
ta CB
Nakase
S138
omyce
s delph
ens
Candida niva
riensis CBS 9 is CBS 2170
983
Candida bracarensis CBS 10154
ne
re
te
llii
ai
hs
za
c
Kaz
ach
sta
Ka
um
Na
tan
ym
ac
as
ad
ym
oz
ov
oz
um
ov
Na
nia
spo
Klu
ra v
Klu yv
ine
ae
yv erom
T02
er
om yce
/19
AF
yc s m
arx
es
ian
lac
us
tis
va
NR
r. m
RL
arx
Yian
11
us
40
K
10
-2
54
CD
2
ii U
86
m
18
C
ha
TC
er
ii A
ick
ar
w
95
stu
es
08
15
ae
yc
C1
#72
m
es
TC
PG
ro
yc
ii A
BV
ve om
yp
eD
s
s
uy er
aria
go
Kl yv
bal
m
Klu
ciu
cym
651
the ium
-12
mo hec
RL Y
t
R
N
i
Ere mo
6340
ver
CBS
Ere
kluy
rans
cea
otole
han
therm
a
2644
Lac
ance
i NCYC
Lach ancea walti
Lach
ellii CBS 4332
cc
ha
rom
yce
nse
Candida cast
Sa
Ha
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
H-6
a
l
o
c
i
r
arbo -17217c
s
e
c
y
rom
RL Y e S288815 02
R
N
Saccha
oxus evisia IFO 1 O 18 3-6C
d
a
r
er tae i IF 62
c
es pa
s
c
y
e
a evi us
c
m
k
y
o
i
r
a
h
m avz an
rom
s
a
Sacc
e
h
c
yc
dri bay
Sac
u
m
ro
s k ces
a
e
h
c
c
y
y
m
Sac
m
o
ar
aro
h
h
cc
cc
a
Sa
S
CT
C1
ccha
romyces
Pha
ffo
ia
dr
my
ce
tac
ea
e
ku
Wi
be
e
Sc
hiz
osa
ia
ch
Pi
Cy
Candida orthopsilosis Co 90-125
Candida met
apsilosis CAN
ME
Candid
Cand a albican
s WO-1
Can ida dub
linie
Ca dida tr
nsis
nd
CD3
o
pica
i
6
d
a
l
S
i
s
p
m
Sp at
alto MYAh
a
340
Sc
t a
sa
4
Xu
he has spor
3
p
ffe
16
or a ar
a p bo
rs
om
rar
as
s
yc
ali iae U
es
da
FM
ru
st
Gm
ip
19
i ti
NR
.1A
sC
RL
BS
Y27
60
90
54
7
er
kk
De
Pic
hia
ce
ae
psilosis
Candida para
us
gispor
es el4o2n39
romyc
Lodde NRRL YB-
60
Y-24
RRL
ilus N
noph
93
n ta n
S 19
ysole
a CB
Pach
5
ulat
ap s
S11
-1
ia c
sG
ori
DL
aish
Kur
ast
ha
la p
orp
ael
lym
gat
o
9
ma
ap
49
r
Ko
pa
S2
ea
CB
is
ata
12
ns
Og
ii M
lle
ev
xe
vz
O
a
tae
a
g
CT
Gc
la d
75
55
Reported standard codon usage
tac
eae
Reported alternative codon usage
CnbMyo1A
CdnMyo1A
CacaMyo1A
NabMyo1A
NdMyo1A
CnbMyo1B
CglMyo1A
Sab_aMyo1B
Sc_cMyo1B
SaaMyo1B
SakMyo1B
SmiMyo1B
Sap_aMyo1B
NacMyo1
KaaMyo1
VpMyo1
TtpMyo1A
TtpMyo1B
TtbMyo1A
TtbMyo1B
ErgMyo1
ErcMyo1
HsvMyo1
KlMyo1
KmmMyo1
KlwMyo1
KaMyo1
KnMyo1
LwMyo1
LatMyo1
Lak_aMyo1
Sab_aMyo1A
Sc_cMyo1A
SaaMyo1A
SakMyo1A
SmiMyo1A
Sap_aMyo1A
ZrMyo1
TodMyo1
CglMyo1B
WaMyo1Alpha
WicMyo1
CyjMyo1
YlMyo1
CdnMyo1B
Kop_bMyo1
NdMyo1B
KcMyo1
OgpMyo1
DebMyo1
PiuMyo1
NabMyo1B
ZbMyo1
Ca_bMyo1
StaMyo1
ShpMyo1
CllMyo1
MefMyo1
CnmMyo1
DhhMyo1
DehMyo1
CatMyo1
LoeMyo1
MrgMyo1
ShsMyo1
CameMyo1
MifMyo1Alpha
CadMyo1
CaoMyo1
CapMyo1
Ct_aMyo1
PtaMyo1
CacaMyo1B
NadMyo1
278
278
279
278
278
278
279
278
278
278
278
278
278
278
278
278
278
278
282
280
277
278
277
276
276
276
277
278
278
278
282
278
277
278
268
278
278
278
278
278
278
278
276
279
278
279
277
280
284
283
284
278
278
287
285
284
286
285
285
285
285
285
292
283
281
286
284
288
286
286
284
282
281
278
310
320
330
340
350
360
370
380
390
400
....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|
KAMQIIGLSQDEQDQIFRMLAAILWIGNISFIENEEGNAQVRDTSVTDFVAYLLQID-SPSLTKSLVERIVETNHGMKRGSVYHVPLNIVQATAVRDALA
KAMQVIGLSQEEQDQIFRMLASILWIGNISFIENEEGNAQVRDTSVTDFVAYLLQID-SPSLTKALVERIVETNHGMKRGSVYHVPLNIVQATAVRDALA
KAARTIGLSQEEQDQIFRQLAAILWIGNLTFVENEEGNAQVRDTSVSDFIAYLLQVE-SELLIKCIVERTVETNHGMKRGSVYHVPLNIVQANAVRDALA
KAMQTIGLAQEEQDNIFRMLAAILWIGNISFIENEEGNAQIRDTSVTDFVAYLLQIN-SDILITAITERTVETNHGMKRGSVYHIPMNIVQATAVRDALA
KAMQIIGLSQDEQDQIFRMLASILWIGNISFIENEEGNAQVRDTSVTDFVAYLLQVD-SPSLTKALVERIVETSHGMKRGSVYHVPLNIVQATAVRDALA
KAMEIIGLSKEEQDQIFRMLAVILWIGNISFIENEEGNAQIRDTSVTAFVAYLLQVQ-EELLIKSLIERIIETGHGAKRGSTYHSPLNIVQATAVRDALA
KAMQVIGLAQEEQDQIFRMLAAILWIGNISFIENEEGNAQVRDTSVTDFVAYLLQVD-SQSLIKALVERIVETNHGSRRGSVYHVPLNIVQATAVRDALA
KAMKVIGLGQEEQDQIFRMLAAILWIGNISFIENEEGNAQVRDTSVTDFVAYLLQID-GQLLVKSLVERIMETNHGMKRGSVYHVPLNIVQADAVRDALA
KAMRVIGLGQEEQDQIFRMLAAILWIGNVSFIENEEGNAQVRDTSVTDFVAYLLQID-SQLLIKSLVERIMETNHGMKRGSVYHVPLNIVQADAVRDALA
KAMRVIGLEQDEQDQIFRMLAAILWIGNVSFIENEEGNAQVRDTSVTDFVAYLLQID-SQLLLKSLVERIMETSHGMKRGSVYHVPLNVVQADAVRDALA
KAMRVIGLGQEEQDQIFRMLAAILWIGNVSFTENEEGNSQVRDTSVTDFVAYLLQID-SQLLIKSLVERIMETNHGMKRGSVYHVPLNIVQADAVRDALA
KAMRVIGLGQEEQDQIFRMLAAILWIGNVSFIENEEGNAQVRDTSVTDFVAYLLQID-SQLLIKSLVERIMETNHGMKRGSVYHVPLNVVQADAVRDALA
KAMRVIGLGQEEQDQIFRMLAAILWIGNVSFIENEEGNAQVRDTSVTDFVAYLLQID-SQLLTKSLVERIMETNHGMKRGSVYHVPLNIVQADAVRDALA
KAMEVIGLHQEEQDQIFRMLAAVLWIGNVSFVENEEGNAQVRDTSVTDFVAYLLQID-APLLIKSLVERVMETNHGMRRGSVYHVPLNIVQATAVRDALA
KAMQVIGLTQQEQDEIFRMLSAILWIGNITFIENEEGNAQVRDTSVTDFVAYLLQVD-SQLLIKSIVERTMETNHGMRRGSIYHVPLNIVQATAVKDALA
KAMQVIGLSQDEQDQIFRMLSSILWVGNISFVENEEGNAQVRDTSVTDFVAYLLQID-SQVLMKALVERTMETSHGMRRGSVYHVPLNIVQATAVRDALA
KAMQIIGLSQDEQDQIFRMLASILWIGNISFVENEEGNAQVRDTSVTDFVAYLLQID-ASILIKGLVERTMETSHGMRRGSVYHVPLNIVQANAAKDALA
KAMQIIGLSQDEQDQIFRMLAFILWIGNISFVENEEGNAQVRDTSVTDFVAYLLQID-ASILIKGLVERTMETSHGMRRGSVYHVPLNIVQANAAEDALA
KAMEIIGLDQNEQDQIFRMLAAILWIGNITFEENDEGNAQVRDTSVTDFVAYLLEVD-SPLLIKSLVERIMETSHGSQRGSVYHVPLNITQATAVRDALA
KAMQVIGLSQDEQDQIFRMLSAILWIGNITFVENDEGNAQVADSSVTDFVAYLLQVD-AGVLVKSLVERIMETNHGMKRGSVYHVPLNRVQATAVRDALA
EAMNVIGLSQAEQDEIFRLLSAILWIGNVTFMEDDEGNAKIADTSITDFVAYLLQVD-AGLLVKSLVERTIETTHGMRRGSIYNVPLNIVQATAVRDALA
KAMGVIGLSQAEQDEIFRMLAAILWIGNITFAENDEGNAQVGDTTVTDFLAYLLQVD-PDLLLKSLVERVIETNHGMKRGSIYNVPLNIVQATAVRDALA
KAMNTIGVTQQEQDEIFRMMAAILWIGNVTFVENDEGNAQVKDTSVTDFVAYLLQVD-SQILIKCLVERIMETNHGMKRGSVYHVPLNRVQATAVRDALA
KAMQVIGLSQEEQDQIFRMLSAILWIGNVTFVENNEGNAEVRDTSVTDFVAYLMQVD-SGLLIKCLVERVMETGHGSRRGSVYHVPLNVVQATAVRDALA
KAMDVIGLSQDEQDQIFRMLSAILWIGNISFIENDEGNSQVRDTSVTDFVAYLLQVD-ASLLIKCLVERVMETGHGARRGSVYHVPLNVVQANAVRDALA
KAMEIIGLSQDEQDQIFRMLSAILWIGNISFVENDEGNSEVRDTSVTDFVAYLMQVD-SSLLVKCLVERVMETGHGSRRGSVYHVPLNVVQATAVKDALA
KAMQVIGLSQDEQDQIFRMLAAILWIGNISFVENEEGNAQIRDTSVTDFVAYLLQVD-SNLLIKSLVERVMETGHGSKRGSIYHVPLNIVQASAVKDALA
KAMQIIGLSQEEQDQIFRVLSSILWIGNITFAENEEGSAAVRDTSVTDFVAYLLQID-AALLTQSLVERTMETNHGMRRGSIYHVPLNVVQATAVRDALA
NAMSVIGITQHEQDEVFRFLAAILWIGNISFTENEEGNAQVRDTSVTDFVAYLLQVD-SNLLVQALVERVMETNHGMKRGSIYHVPLNVVQATAVKDALA
KAMQVIGITQQEQDELFRLLSAILWIGNISFTENEEGNAQVCDTSVTDFVAYLLQVD-AHFLTQALVERVMETNHGMKRGSVYHVPLNVVQATAVKDALA
KAMQVIGLSQEEQDQIFRLLATILWIGNISFTEDEEGNAQVRDTSVTDFVAYLLQVD-SQLLIKCLVERVMETSHGMRRGSVYHVPLNIVQATAVKDALA
EAMRTIGLVQEEQDQIFRMLAAILWIGNISFIENEEGNAQVRDTSVTDFVAYLLQVD-APLLIKCLVERIMQTSHGMKRGSVYHVPLNAVQATAARDALA
EAMRTIGLVQEEQDQIFRMLAAILWIGNISFIENEEGNAQVGDTSVTDFVAYLLQVD-ASLLVKCLVERIMQTSHGMKRGSVYHVPLNPVQATAVRDALA
EAMKTIGLVQEEQDQIFRMLAAILWIGNISFVENEEGNAQVRDTSVTDFVAYLLQVD-ASLLIKCLVERIMQTSHGMKRGSVYHVPLNAVQATAVKDALA
------GLGQEDQYQFFRMLAAIMWMANISFIENEEANAQVRDTSVTDFAAYILQVD-APFLIKCMVDRIMQTILGMKRGSVYHVPLNTVQATAVRDALA
EAMRTIGLAQEEQDQIFRMLAAILWVGNISFIENEEGNAQVRDTSVTDFVAYLLQVD-ASLLIKCLVERIMQTSHGMKRGSVYHVPLNPVQATAVRDALA
EAMRTIGLAQEEQDQIFRMLAAILWIGNISFIENEEGNAQIRDTSVTDFVAYLLQVD-ASLLMKCLVERIMQTSHGMKRGSVYHVPLNPVQATAVRDALA
AAMRVIGLSQEEQDEIFRVLAAILWTGNITFMENDEGNAQVRDTSVTDFVAYLLQVD-SQLLVKSLVERIMETNHGMRRGSVYHVPLTTVHATAVKDALA
KAMQVIGLSQEEQDQIFRLLAAILWIGNITFVEDEEGNAKIRDTSVTDFVAYLLQVE-SELLIKSLVERVLETNHGMRRGSVYHVPLNAVQATAVKDALA
NAMRTIGLTKSEQDQIFRALAAILWIGNISFVENEAGNAEIRDKSVTTFVAYLLEVQ-EELLIKALIERIIETTHGAKRGSTYHSPLNIIQATAVRDALA
NAMQTVGITQPEQDQIFRVLAAILWIGNITFVENEEGNAQVRDTSVTDFVAYLLQVD-SETLNKAVTERVMETSHGMRRGSVYNVPLNITQATAVRDALA
SAMQTIGVTQDEQDQIFRILAAILWIGNITFMENDEGNSQVKDTSVTDFVAYLLQVD-SEILIKSITQRVMETSHGMKRGSVYEVPLNITQAIAVKDALA
KAMQIIGIDQEQQDQIFRILAAILWIGNITFVEDEQGNSTVRDSSVTKFVAYLISVD-EQTLINSLITRVMTTMN-----ETYDIPLNPTQATAVKNALA
AAMNLIGLTQAEQDNLFKLLAAILWIGNMSFVEDKDGNAAIADVSVPNFVAYLLEVD-AESVVKAVTQRIMETSRGGRRGSVYEVALNIAQATSVRDALA
NAMNIIGLSKEEQDQIFRMLAAILWIGNISFIEDEQGNAQIRDTSVTTFVAYLLQVQ-EELLIKSLVERIIETGHGAKRGSTYHSPLNIVQATAVRDALA
KAMQTIGLTQDEQDQVFRILAAILWIGNISFVENEDGNAQIRDSSVTDFVAYLLQVN-AQILTNSLIERIVETSHGSKRGSIYHVPLNITQATAVRDALS
RAMDIIGLAKEEQDQIFRLLAAILWIGNISFVENEQGNAQIRDTSVTTFVAYLLQVQ-EDMLIKSLVERVIETSHGAKRGSTYHSPLNIVQATAVRDALA
KAMRTIGMTQEEQDELFSLLAAVLWIGNVSFAEDAEGNSTIRDTSVTDFVAYLLQVD-SQILCKSLTERTLETNHGMKRGSIYHVPLNMTQATAVRDALA
ASMTTIGLTQEEQDQVFRVLAAILWIGNITFVEDAEGNAQVRDTGVTDFVAYLLQVN-SEVLVKSIIERTMETSHGMRRGSVYHVPLNPVQATASRDALA
KAMDVIGITDEERDQIFRILAGILWIGNITFTEDEEGNAAIADTSVTDFVAYLLQVD-AQTLCKSVVERTIQTFHGMRRGSIYHSPLNPVQATASRDALA
KAMDTIGLSSEERDHVFRLLSAILWIGNVSFIEDEEGNSKIRDESVTNFVAYLLQVD-ATLLCKSLTERTMETSHGMKRGSIYHVPLNTTQATAVKDALA
RAMETIGISQEEQDNIFRLLAAILWTGNISFVENEEGNAQIRDTSVTDFVAYLLQVD-SQMLIGSIIERIMETSHGSRRGSIYNVPLNIVQATAVKDALA
KAMQVIGLTQEEQDEIFRVLAAILWIGNITFGENDEGNAEVRDTSVTDFVAYLLQVD-SQLLIKSLVERIMETNHGMRRGSVYHVPLNTVQATAVRDALA
NAMKIIGLTQQEQDNIFRMLASILWIGNISFVEDENGNAAIRDDSVTNFAAYLLDVN-PEILKKAIIEKTIETSHGMRRGSTYHSPLNIVQATAVRDALA
NAMNIIGLTQDEQDNIFKMLAAILWIGNISFVEDESGNAAIRDETVTQFVAYLLDVPSAEIMKKAIIERTIETTHGSRRGSTYHVPLNIVQATSVRDALA
NAMNIIGLTQEEQDNIFRMLAAILWVGNISFVEDESGNAAIRDETVTQFVAYLLDVS-PEIMKKAIIERTIETTHGSRRGSTYHVPLNIVQATSVRDALA
RAMEIIGLSQAEQDNIFRMLASILWIGNISFVEDESGNATIRDAGVTNFVAYLLEVS-PEILQKAIVERIIETSHGMRRGSTYHVPLNIVQATAVRDALA
NAMNVIGLAQVEQDNIFRMLAAILWIGNISFVEDENGNAAVRDEGVTNFVAYLLEVN-AEILKKSIIERIIETSHGMRRGSTYHVPLNITQATAVRDALA
NAMQVIGLTQQEQDSIFRMLASILWIGNISFVEDEHGNAAIRDESVTNFVGYLLDVS-GEIVKKSIIERTIETSFGSRRGSTYHSPLNIVQATAVRDALA
RAMQVIGLSQEEQDNIFRMLASILWIGNISFVEDEHGNAAIRDESVTAFVAYLLDVN-AETLKTSLIQRVMQTSHGMRRGSTYHVPLNIVQATSVRDALA
RAMQVIGLSQDEQDNIFRMLASILWVGNVSFVEDDNGNAAVRDESVTAFIAYLLDVD-AETLKTSLIQRVMQTSHGMRRGSTYHVPLNIVQATSVRDALA
NAMKVIGITPQEQDHIFRMLAAILWVGNISYVEDEHGNAAIRDESVINFVAYLLETD-AESVKKSITEKIVQTSHGMRRGSTYHSPLNIVQATAVRDALA
NAMNIIGLSQAEQDNIFRILASILWTGNISFVEDESGNAAIRDDSVTNFVAYLLDVN-AEILKKAITERTIETSHGMKRGSTYHVPLNIVQATAVRDALA
QAMNTIGLSKAEQDNIFRSLASILWIGNISFVENEDGNAAIRDDTVTTFVAYLLEVD-ANVLKKSILERVIETSHGMRRGSTYHVPLNIVQATASRDALA
SAMKIIGLTELEQNNIFRMIASILWIGNVSFVEDESGNAAIRDDSVTQFVAYLLEVN-PEILKKAIVERVIETTHGMRRGSTYHVPLNIVQATSVRDALA
NAMNVIGLTQDEQDNIFRMLASILWIGNISFVEDESGNAAIRDDSVTNFVAYLLDVN-PEILKKAIIERTIETSHGMKRGSTYHVPLNIVQATAVRDALA
KAMQIIGLTQDEQDNIFRVLASILWIGNISFVEDENGNAIPRDESVTNFVAYLLNVE-PVILKTAFVQRVMQTSHGMRRGSTYHVPLNIVQATAVRDALA
NAMKIIGLTQQEQDNIFRMLAVILWIGNISFIEDENGNATIRDDSVTNFVAYLLDVN-SEILKKAIIERTIETSHGMRRGSTYHSPLNIVQATAVRDALA
NAMNVIGLSQEEQDNIFRMLASILWIGNISFVEDESGNAAIRDDSVTNFVAYLLDVN-PEILKKAIIERTIETSHGMKRGSTYHVPLNIVQATAVRDALA
NAMNVIGLSQDEQDNIFRMLASILWIGNISFIEDESGNAAIRDDSVTNFVAYLLDAN-PDILKKAIIERTIETSHGMKRGSTYHVPLNIVQATAVRDALA
NAMQIIGLSQDEQDQIFRMLASILWVGNISFVEDENGNAAIRDESVTNFVAYLLGVN-TEILKKAIIERTIETSHGMRRGSTYHSPLNIVQATAVRDALA
NAMSIIGLSQEEQDQIFRILSAILWIGNVLFVENEDANSAIGDQGVINYIAYLLQVD-AEQLAKSLTERIMETSHGMKRGSIYHVPLNLTQAYAVRDALA
NAMKTIGLSQTEQDHIFRLLAAILWIGNISFIENEEGSAQIRDTSVTDFVAYLLEVD-SSLLITSIVERILETSHGTKRGSIYHTPLNIVQATAVRDGLA
KAMQTIGLAQEEQDQIFRMLAAILWIGNISFIENEEGNAQVRDTSVTDFVAYLLQVD-AAVLIKALVERIMETSHGMRRGSVYHVPLNIVQATAVRDALA
A
window size 0
window size 3
900
2500
800
2000
leucine residues
700
occurence
occurence
600
1500
1000
500
400
300
200
500
0
0
0
5
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
5
0
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
conservation score [%]
conservation score [%]
450
350
400
300
350
occurence
300
200
150
250
200
150
100
100
50
0
5
0
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
5
0
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
conservation score [%]
B
4
kinesin-1
bits
3
2
I
K
L I N
A
S
R
M
E
N
A
QV
R
H
Y
E
Q
V
Y
4
kinesin-5
bits
3
2
1
T
K
R
Q
R
V
I
DG
R
A
S
F
N
D
A
S
P
KN
L
C
D
H
I S
MYF I
F
V
M
VG L
FVR I
L
Y
I
V
K
Y
F
E
0
N
N
G
R
I SQ
TVTT
I
0
N
LVDLAGSEK KTG A G LE AK IN SLS LG VIN LTD S
S
Q
D
N
A
R
R
I
R
T
A
N
K NLVDLAGSE SG A E G IN SL LG VIN L
K S LF
1 S
VT
I
K
R
T
SN
A I
DR
SQ
T
S
V
M
K
L
V
L
Q
I
G
KL
L
ST
L
Q
T
Q
N
K AS
TST YD
I
R
M
RA
V
N
A N SA GK
S
K
AS
H
I
L
N
F
H
A
C
E
QT
FAR
VMDV
P
R
conservation score [%]
S
L
M
SS
PE
NC
AQ RK
K
LQ
A
A
G
N
N
D
SS
GV
Y
E
H
N
L
I
K
N
NTMVST
Q
V
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
0
50
T
A
F
LQL
S
M
T
I
H
S
I
K
Y
T
R
I
N
FA
S
W
I
A
G
N
F
S
V
D
H
L
H
G
K
T
G
S
A
K
E
Q
S
Y
A
N
I
E
S
T
N
K
Q
K
SIV
V
HMK
T
E
R
DA
V
F
Q
M
E
L
G
Q
S
Q
I
I
S
S
T
I
D
N
L
M
G
I
K
A
A
F
G
H
R
V
KS
N
V
K
N
S
N
X
AS V KS
S
Q
N
I
K
N
V
F
S
L
T
MV
G
T
E KN
S
C
I
I
N
M
H
G
R
C
N
Q
G
L
NAP
S
QT T
AS
N
A
I
Q
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
occurence
250
switch II
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
100
α4
C
serine residues
# alignment positions with a score
in the respective range
# alignment positions with a score
in the respective range and with at
least 1 CUG codon
# CUG codons at alignment
positions with a score in the
respective range
[%] CUG codons
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
Yarrowia lipolytica CLIB122
Wickerhamomyces anomalus NRRL Y-366
Wickerhamomyces ciferrii NRRL Y-1031
Cyberlindnera jadinii NBRC 0988
Kluyveromyces aestuarii ATCC 18862
Kluyveromyces wickerhamii UCD 54-210
Kluyveromyces marxianus var. marxianus KCTC 17555
Kluyveromyces lactis NRRL Y-1140
Lachancea thermotolerans CBS 6340
Lachancea waltii NCYC 2644
Eremothecium gossypii ATCC 10895
Eremothecium cymbalariae DBVPG#7215
Lachancea kluyveri NRRL Y-12651
Tetrapisispora phaffii CBS 4417
Vanderwaltozyma polyspora DSM 70294
Tetrapisispora blattae CBS 6284
Torulaspora delbrueckii CBS 1146
Zygosaccharomyces bailii ISA1307
Zygosaccharomyces rouxii CBS732
Candida castellii CBS 4332
Nakaseomyces bacillisporus CBS 7720
Candida glabrata CBS138
Nakaseomyces delphensis CBS 2170
Candida nivariensis CBS 9983
Candida bracarensis CBS 10154
Saccharomyces bayanus 623-6C
Saccharomyces kudriavzevii IFO 1802
Saccharomyces mikatae IFO 1815
Saccharomyces cerevisiae S288c
Saccharomyces paradoxus NRRL Y-17217
Saccharomyces arboricola H-6
Kazachstania africana CBS 2517
Kazachstania naganishii CBS 8797
Naumovozyma dairenensis CBS 421
Naumovozyma castellii NRRL Y-12630
Hanseniaspora vineae T02/19AF
Komagataella pastoris GS115
Pichia kudriavzevii M12
Dekkera bruxellensis CBS 2499
Ogataea parapolymorpha DL-1
Kuraishia capsulata CBS 1993
Pachysolen tannophilus NRRL Y-2460
Candida tenuis NRRL Y-1498
Spathaspora passalidarum NRRL Y-27907
Spathaspora arborariae UFMG-19.1A
Candida metapsilosis CANME
Candida orthopsilosis Co 90-125
Candida parapsilosis
Lodderomyces elongisporus NRRL YB-4239
Candida tropicalis MYA-3404
Candida maltosa Xu316
Candida albicans WO-1
Candida dubliniensis CD36
Scheffersomyces stipitis CBS 6054
Millerozyma farinosa CBS7064
Debaryomyces hansenii var. hansenii MTCC 234
Debaryomyces hansenii CBS767
Metschnikowia fructicola 277
Clavispora lusitaniae ATCC 42720
Meyerozyma guilliermondii ATCC 6260
[%] CUG codons
Serine, min. 90% conservation score
Leucine, min. 90% conservation score
Serine, min. 80% conservation score
Leucine, min. 80% conservation score
Serine, min. 50% conservation sore
Leucine, min. 50% conservation sore
Yarrowia lipolytica CLIB122
Wickerhamomyces anomalus NRRL Y-366
Wickerhamomyces ciferrii NRRL Y-1031
Cyberlindnera jadinii NBRC 0988
Kluyveromyces aestuarii ATCC 18862
Kluyveromyces wickerhamii UCD 54-210
Kluyveromyces marxianus var. marxianus KCTC 17555
Kluyveromyces lactis NRRL Y-1140
Lachancea thermotolerans CBS 6340
Lachancea waltii NCYC 2644
Eremothecium gossypii ATCC 10895
Eremothecium cymbalariae DBVPG#7215
Lachancea kluyveri NRRL Y-12651
Tetrapisispora phaffii CBS 4417
Vanderwaltozyma polyspora DSM 70294
Tetrapisispora blattae CBS 6284
Torulaspora delbrueckii CBS 1146
Zygosaccharomyces bailii ISA1307
Zygosaccharomyces rouxii CBS732
Candida castellii CBS 4332
Nakaseomyces bacillisporus CBS 7720
Candida glabrata CBS138
Nakaseomyces delphensis CBS 2170
Candida nivariensis CBS 9983
Candida bracarensis CBS 10154
Saccharomyces bayanus 623-6C
Saccharomyces kudriavzevii IFO 1802
Saccharomyces mikatae IFO 1815
Saccharomyces cerevisiae S288c
Saccharomyces paradoxus NRRL Y-17217
Saccharomyces arboricola H-6
Kazachstania africana CBS 2517
Kazachstania naganishii CBS 8797
Naumovozyma dairenensis CBS 421
Naumovozyma castellii NRRL Y-12630
Hanseniaspora vineae T02/19AF
Komagataella pastoris GS115
Pichia kudriavzevii M12
Dekkera bruxellensis CBS 2499
Ogataea parapolymorpha DL-1
Kuraishia capsulata CBS 1993
Pachysolen tannophilus NRRL Y-2460
Candida tenuis NRRL Y-1498
Spathaspora passalidarum NRRL Y-27907
Spathaspora arborariae UFMG-19.1A
Candida metapsilosis CANME
Candida orthopsilosis Co 90-125
Candida parapsilosis
Lodderomyces elongisporus NRRL YB-4239
Candida tropicalis MYA-3404
Candida maltosa Xu316
Candida albicans WO-1
Candida dubliniensis CD36
Scheffersomyces stipitis CBS 6054
Millerozyma farinosa CBS7064
Debaryomyces hansenii var. hansenii MTCC 234
Debaryomyces hansenii CBS767
Metschnikowia fructicola 277
Clavispora lusitaniae ATCC 42720
Meyerozyma guilliermondii ATCC 6260
# CUGs
770
# CUGs at positions with a conservation score of ≥50%
Yarrowia lipolytica CLIB122
Wickerhamomyces anomalus NRRL Y-366
Wickerhamomyces ciferrii NRRL Y-1031
Cyberlindnera jadinii NBRC 0988
Kluyveromyces aestuarii ATCC 18862
Kluyveromyces wickerhamii UCD 54-210
Kluyveromyces marxianus var. marxianus KCTC 17555
Kluyveromyces lactis NRRL Y-1140
Lachancea thermotolerans CBS 6340
Lachancea waltii NCYC 2644
Eremothecium gossypii ATCC 10895
Eremothecium cymbalariae DBVPG#7215
Lachancea kluyveri NRRL Y-12651
Tetrapisispora phaffii CBS 4417
Vanderwaltozyma polyspora DSM 70294
Tetrapisispora blattae CBS 6284
Torulaspora delbrueckii CBS 1146
Zygosaccharomyces bailii ISA1307
Zygosaccharomyces rouxii CBS732
Candida castellii CBS 4332
Nakaseomyces bacillisporus CBS 7720
Candida glabrata CBS138
Nakaseomyces delphensis CBS 2170
Candida nivariensis CBS 9983
Candida bracarensis CBS 10154
Saccharomyces bayanus 623-6C
Saccharomyces kudriavzevii IFO 1802
Saccharomyces mikatae IFO 1815
Saccharomyces cerevisiae S288c
Saccharomyces paradoxus NRRL Y-17217
Saccharomyces arboricola H-6
Kazachstania africana CBS 2517
Kazachstania naganishii CBS 8797
Naumovozyma dairenensis CBS 421
Naumovozyma castellii NRRL Y-12630
Hanseniaspora vineae T02/19AF
Komagataella pastoris GS115
Pichia kudriavzevii M12
Dekkera bruxellensis CBS 2499
Ogataea parapolymorpha DL-1
Kuraishia capsulata CBS 1993
Pachysolen tannophilus NRRL Y-2460
Candida tenuis NRRL Y-1498
Spathaspora passalidarum NRRL Y-27907
Spathaspora arborariae UFMG-19.1A
Candida metapsilosis CANME
Candida orthopsilosis Co 90-125
Candida parapsilosis
Lodderomyces elongisporus NRRL YB-4239
Candida tropicalis MYA-3404
Candida maltosa Xu316
Candida albicans WO-1
Candida dubliniensis CD36
Scheffersomyces stipitis CBS 6054
Millerozyma farinosa CBS7064
Debaryomyces hansenii var. hansenii MTCC 234
Debaryomyces hansenii CBS767
Metschnikowia fructicola 277
Clavispora lusitaniae ATCC 42720
Meyerozyma guilliermondii ATCC 6260
#CUGs at identical alignment positions
316
100
31
10
3
1
Schizosaccharomyces pombe 972hSchizosaccharomyces cryophilus NRRL Y-48691
Schizosaccharomyces octosporus yFS286
Schizosaccharomyces japonicus yFS275
Yarrowia lipolytica CLIB122
Wickerhamomyces anomalus NRRL Y-366
Wickerhamomyces ciferrii NRRL Y-1031
Cyberlindnera jadinii NBRC 0988
Kluyveromyces aestuarii ATCC 18862
Kluyveromyces wickerhamii UCD 54-210
Kluyveromyces marxianus var. marxianus KCTC 17555
Kluyveromyces lactis NRRL Y-1140
Lachancea thermotolerans CBS 6340
Lachancea waltii NCYC 2644
Eremothecium gossypii ATCC 10895
Eremothecium cymbalariae DBVPG#7215
Lachancea kluyveri NRRL Y-12651
Tetrapisispora phaffii CBS 4417
Vanderwaltozyma polyspora DSM 70294
Tetrapisispora blattae CBS 6284
Torulaspora delbrueckii CBS 1146
Zygosaccharomyces bailii ISA1307
Zygosaccharomyces rouxii CBS732
Candida castellii CBS 4332
Nakaseomyces bacillisporus CBS 7720
Candida glabrata CBS138
Nakaseomyces delphensis CBS 2170
Candida nivariensis CBS 9983
Candida bracarensis CBS 10154
Saccharomyces bayanus 623-6C
Saccharomyces kudriavzevii IFO 1802
Saccharomyces mikatae IFO 1815
Saccharomyces cerevisiae S288c
Saccharomyces paradoxus NRRL Y-17217
Saccharomyces arboricola H-6
Kazachstania africana CBS 2517
Kazachstania naganishii CBS 8797
Naumovozyma dairenensis CBS 421
Naumovozyma castellii NRRL Y-12630
Hanseniaspora vineae T02/19AF
Komagataella pastoris GS115
Pichia kudriavzevii M12
Dekkera bruxellensis CBS 2499
Ogataea parapolymorpha DL-1
Kuraishia capsulata CBS 1993
Pachysolen tannophilus NRRL Y-2460
Candida tenuis NRRL Y-1498
Spathaspora passalidarum NRRL Y-27907
Spathaspora arborariae UFMG-19.1A
Candida metapsilosis CANME
Candida orthopsilosis Co 90-125
Candida parapsilosis
Lodderomyces elongisporus NRRL YB-4239
Candida tropicalis MYA-3404
Candida maltosa Xu316
Candida albicans WO-1
Candida dubliniensis CD36
Scheffersomyces stipitis CBS 6054
Millerozyma farinosa CBS7064
Debaryomyces hansenii var. hansenii MTCC 234
Debaryomyces hansenii CBS767
Metschnikowia fructicola 277
Clavispora lusitaniae ATCC 42720
Meyerozyma guilliermondii ATCC 6260
100/100/100
100/100/100
587 Ma
100/100/100
100/100/88
100/100/100
100/100/100
235 Ma
100/100/100
100/100/52
100/100/100
Downloaded from http://gbe.oxfordjournals.org/ by guest on October 6, 2014
87/100/100
100/100/100
56/100/-
235 220
100
175
135 Ma
amino acid substitutions per site
0.2
0
190
200
300
400
500
600
700
800
# CTG codons