Supplementary Materials for -...

www.sciencemag.org/cgi/content/full/339/6124/1207/DC1

Supplementary Materials for

Gene Transfer from Bacteria and Archaea Facilitated Evolution of an Extremophilic Eukaryote

Gerald Schönknecht,* Wei-Hua Chen, Chad M. Ternes, Guillaume G. Barbier, Roshan P. Shrestha, Mario Stanke, Andrea Bräutigam, Brett J. Baker, Jillian F. Banfield, R. Michael Garavito, Kevin Carr, Curtis Wilkerson, Stefan A. Rensing, David Gagneul, Nicholas E.

Dickenson, Christine Oesterhelt, Martin J. Lercher, Andreas P. M. Weber*

*To whom correspondence should be addressed. E-mail: [email protected] (G.S.);

[email protected] (A.P.M.W.)

Published 8 March 2013, Science 339, 1207 (2013) DOI: 10.1126/science.1231707

This PDF file includes:

Materials and Methods Supplementary Text Figs. S1 to S27 Tables S1 to S3 Caption for data table S4 References (21–79)

Other Supplementary Material for this manuscript includes the following: (available at www.sciencemag.org/cgi/content/full/339/6124/1207/DC1)

Data table S4 (Microsoft Excel file)

2

Materials and Methods

Strains and Media Galdieria sulphuraria strain 074W was cultivated axenically in minimal mineral medium

supplemented with 25 mM Gal at 37°C in the dark as described previously (3).

DNA isolation and sequencing G. sulphuraria cells were harvested from heterotrophic cultures and ground to a fine

powder with a mortar and a pestle. Total nucleic acids were extracted by incubating the ground tissue overnight in 50 mM Tris-Cl pH7.5, 5 mM EDTA, and 1% (w/v) SDS, followed by extraction of proteins with phenol:chloroform:isoamylalcohol (24:24:1), and precipitation of DNA from the aqueous phase by ethanol. The pellet was dissolved in 10 mM Tris-Cl pH 7.5, 1 mM EDTA, and RNA was removed by incubation with DNAse-free RNAse, followed by de-proteination with phenol:chloroform:isoamylalcohol, and DNA-precipitation by ethanol. Nuclear DNA was further purified by CsCl-density gradient centrifugation of bis-benzamide-treated total DNA (21).

Three different libraries were generated for sequencing of genomic DNA: (i) a small-insert (approx. 2 kB inserts) shotgun sequencing plasmid library that was constructed in pSMART-HC Kan (Lucigen, Middleton, WI; www.lucigene.com); (ii) a Fosmid library containing 40 kbp inserts was constructed by physically shearing genomic DNA and ligation of end repaired, size fractionated DNA into the Fosmid vector pCC1FOS, followed by packing into Lambda phages. (iii) two BAC libraries (BamHI, HindIII) containing > 100 kbp inserts were constructed by the TAMU GENEfinder Genomics Resource Center (Texas A&M University, College Station, TX).

DNA sequencing was conducted using fluorescence-labeled dye terminators on an ABI 3730xl capillary DNA sequencing systems. Sequence data and chromatograms were stored on a Geospiza Finch server (Geospiza, Seattle, WA). After cleaning the raw sequence data (base quality > Q20, exclusion of contaminating vector or bacterial sequences), 1,769 and 1,699 sequence reads were generated from the BamH1 and HindIII BAC libraries, respectively, 5,817 sequence reads were generated from the Fosmid library, and 190,911 sequence reads were generated from three different small-insert libraries. A total of 147,097,538 Q20-bases were thus sequenced by Sanger technology, yielding approx. 10-fold genome coverage by Sanger-sequenced Q20-bases, based on an approximate genome size of 14 Mb. In addition, we generated 163,727 sequence reads (8.55 x genome coverage) from physically sheared genomic DNA using a GS20 Genomic Sequencer (Roche, Indianapolis, IN). A hybrid (i.e., Sanger and 454) assembly using the ARACHNE genome assembler (22) yielded 433 scaffolds with an N50 of 172,322 bases, giving a total scaffold length of 13712004 bases of which 292650 bases (2.1%) are gaps. Sequencing of mRNA as described previously (23).

Generation of gene models and annotation of protein coding genes

The AUGUSTUS (24) gene prediction program was used for the prediction of genes on the ARACHNE-generated scaffolds. AUGUSTUS was trained with EST-sequences generated by Sanger-technology (25) and by mRNA sequence data generated from two different normalized libraries (cells grown autotrophically and heterotrophically) using a GS20 sequencing system. The genes were then predicted with AUGUSTUS (24) integrating evidence from the mRNA sequencing and from proteins of C. merolae, but also allowing for ab initio predicted models in absence of such data. As a result 6,623 protein coding genes were predicted, 551 with two splice

3

variants, and three with three splice variants. Routine analyses were performed with splice variant 1, which is most confident according to the model of AUGUSTUS. Functions were assigned to the 6,623 predicted major isoforms of proteins based on sequence similarity to annotated genes from NCBI non-redundant (nr) protein database (26), NCBI Conserved Domain Database (CDD) (27), Arabidopsis thaliana genome (28), C. merolae genome (7), KEGG (29), and UniProt (30) databases. In addition, Gene Ontology (GO) (31), Mercator (32), InterProScan (33), and HHpred (34) were used to identify protein functions. Similarity searching was done with the BLAST program with soft masking being turned on (-F “m S”) (35). A BLAST bit-score cutoff of 50 was used, unless stated otherwise. Membrane transporter proteins were annotated with the TransAAP tool from TransportDB (36), with TransportTP (37), and by BLAST against the Transporter Classification Database (TCDB) (38). SignalP 4.0 (39) was used to search for secretory signal peptides. To test for enrichment of certain functional categories in subsets of annotated proteins High-Throughput GoMiner (40) was used. For this analysis G. sulphuraria proteins were substituted by UniProt (30) identifiers of best BLAST hits.

Generation of protein families

G. sulphuraria gene families were constructed to investigate both the size and functions of proteins associated with these families. We performed an all-against-all BLAST (35) using protein sequences in G. sulphuraria, parsed the bit scores from the BLAST output and submitted them to MCL (Markov Clustering; run with default parameters) to build gene families. Each resulting gene family contains either multiple proteins (table S2) that were derived from a common ancestor through duplication events or only a single protein (singlet). InParanoid version 2 (41) with default parameters was used to search for orthologous relationships between proteins in G. sulphuraria and proteins from 208 sequenced genomes (http://dx.doi.org/10.5061/dryad.84r5q) of common model organisms, thermophilic bacteria and archaea, as well as genomes containing top BLAST hits of G. sulphuraria proteins in the NCBI nr database (26). We chose orthologous groups with score 1 for further analysis when testing for horizontal gene transfer (HGT; see below). To investigate the evolutionary histories of G. sulphuraria protein families, we used a method similar to Merchant et al. (42) to expand G. sulphuraria gene families with orthologous proteins (identified using InParanoid, see above) plus paralogous proteins of the identified orthologs from 43 eukaryotic genomes (http://dx.doi.org/10.5061/dryad.84r5q). These expanded protein families were used for phylogenetic analyses and to compare expansion or reduction of a specific protein family among different eukaryotic species.

Identification of horizontal gene transfer (HGT) candidates

All HGT candidates were identified by phylogenetic analyses. Detailed phylogenetic trees were constructed manually (see below) for most large protein families (i.e., those with ten or more paralogs; table S2), for some protein families with best BLAST hits (NCBI, nr) in bacterial or archaeal sequences, and for some protein families with protein domains (PFAM) and/or annotations characteristic for Bacteria or Archaea. This resulted in the identification of several HGT candidates.

In addition, a genome-wide, systematic screen for HGT candidates was performed using bioinformatics methods. To estimate a lower bound for HGT into G. sulphuraria, we aimed to identify only unequivocal cases of HGT. Proteins giving only hits in bacterial or archaeal sequences when blasted against our 208-species database were further analyzed as described

4

below (next paragraph a) to e)). For proteins giving best BLAST hits in bacterial or archaeal sequences in addition to hits in eukaryotic sequences, we performed a phylogenetic analysis using the following procedure. A multiple sequence alignment (MSA) for each of the proteins and their orthologous proteins (identified using InParanoid) from the same 208 genomes was constructed using MUSCLE (43), with the maximum number of iterations set to 100, followed by GBLOCKS (44) with parameters ‘-b3=8 –b4=2 –n=y’ to remove poorly aligned regions. We selected the best protein evolution model for each MSA using ProtTest (45), and used it to reconstruct the phylogenetic relationships for the proteins in the MSA with PhyML (46). To select putative HGT candidates with statistical significance, we performed a RELL analysis implemented in the ‘codeml’ command of the PAML package (47) on each phylogenetic tree indicating HGT. This test compares the protein family tree obtained from our analysis, indicating HGT from a bacterium or archaeon into G. sulphuraria, with the best tree that enforces monophyly of eukaryotic sequences. HGT is only assumed if the tree enforcing monophyly of eukaryotic sequences gets a RELL bootstrap support (pRELL) of 5% or less out of 10,000 replicates in this analysis. Following this systematic bioinformatics screen, the phylogenetic tree for each statistically significant (according to the RELL analysis) HGT candidate was manually checked and only accepted when a clear pattern of horizontal gene transfer was observed.

Very stringent criteria were used for our phylogenetic analyses: a) Sequences that were too short (<150 amino acids) to build reliable MSAs were not accepted; b) phylogenetic trees with less than ten species were excluded: c) genes that were potentially transferred from cyanobacteria were only accepted as HGT candidates when homologs were absent from other photosynthetic eukaryotes, and when the annotation did not indicate a function in photosynthesis, to discriminate against endosymbiotic gene transfer; d) in cases where a corresponding phylogenetic tree did not allow conclusions about the origin of a gene in G. sulphuraria the gene was removed; e) in order to check whether a potential HGT candidate had significant sequence similarity with proteins from eukaryotic species not included in the 208 genomes used for the systematic bioinformatics screen, each potential HGT candidate was submitted to the NCBI Web BLAST service (nr) and a tree of best BLAST hits was generated by the Tree View option; HGT candidates that were not confirmed by a tree of best BLAST hits were not accepted. In cases where an identified HGT candidate was a member of a protein family with two or more paralogs, and the other paralogs had not been identified as HGT candidates, a detailed phylogenetic tree was constructed for the complete protein family to decide which family members were HGT candidates. To facilitate manual inspection of the phylogenetic trees from the systematic bioinformatics screen, we colored tree branches and leaf labels with predefined rules using ColorTree (48).

Data analysis and generation of phylogenetic trees

Non-linear regression analysis was performed with GraphPad Prism 5.01 (GraphPad Software Inc., La Jolla, CA). Two-tail P values for Fisher's exact tests were calculated as defined by Agresti (49) (at http://www.langsrud.com/stat/fisher.htm ). To generate the phylogenetic trees presented as figures, homologous sequences from protein families (see above) and sequences resulting from BLAST (35) searches (NCBI non-redundant (nr) protein database (26) and KEGG (29)) were collected using MEGA (50). At NCBI BLAST runs were carried out separately for different clades to improve broad phylogenomic sampling. Protein sequences from bacteria or archaea that were identified as potential ‘donor’ organisms for HGT into G. sulphuraria were used as BLAST queries to further improve the resolution of phylogenetic trees. Incomplete or

5

highly redundant sequences (>90% amino acid identity) were removed. Multiple sequence alignments were generated with T-Coffee in the ‘mcoffe’ and the ‘accurate’ mode (51). High scoring portions of multiple sequence alignments (T-Coffee score 5 to 9) were extracted (44), followed by comparison of the two resulting alignments (‘mcoffee’ versus ‘accurate’). The better alignment, based on scores (T-Coffee scores from 0 to 100), length, number of invariant positions, and visual inspection, was used to construct a phylogenetic tree. An estimation which model of protein evolution best fits the multiple sequence alignment was generated by ProtTest (45). The best models were used to generate phylogenetic trees with PhyML 3.0 (46) estimating branch support values by non-parametric bootstrap with 200 replicates, and with the MPI version of MrBayes 3.1 (52). MrBayes was run with twelve chains for 5,000,000 generations sampling every 100th generation; the first 25% of samples were ignored when calculating parameters and consensus tree. Phylogenetic trees were visualized in Dendroscope (53), MEGA (50), and EvolView (54). Only support values >50% are given.

Supplementary Text

Early divergence of G. sulphuraria and C. merolae The low degree of colinearity between the genomes of G. sulphuraria and C. merolae (fig.

S2) indicates an early divergence of these two species. To obtain a rough estimate as to when G. sulphuraria and C. merolae diverged, percent amino acid identity of orthologous gene pairs was compared for genomes from different pairs of species, and normalized cumulative frequencies were plotted (fig. S1). At 50% percent cumulative frequency (median value), amino acid identity for annotated proteins from the two red algal genomes was 44.9%, compared to 45.4% for the species pair Homo sapiens / Drosophila melanogaster. Taking percent amino acid identity as a rough measure for divergence time, this means that G. sulphuraria and C. merolae split slightly earlier (lower amino acid identity) than H. sapiens and D. melanogaster did. The divergence between vertebrates and insects is dated to about 910 (±300) million years ago (55), indicating that - assuming comparable mutation rates - G. sulphuraria and C. merolae diverged about 900 million years ago.

This early divergence is further supported by a comparison of all annotated proteins in the genomes of G. sulphuraria and C. merolae. Only 1,259 annotated proteins (19%) from G. sulphuraria have proteins from C. merolae as best BLAST hits, and only slightly more than 60% (4,017 proteins) give a significant BLAST hit (score >50) with C. merolae. The latter percentage is comparable to that of BLAST hits with proteins from Arabidopsis thaliana (4,032 proteins), emphasizing the early divergence of the two red algae These data are in agreement with phylogenetic studies of the Cyanidiophyceae showing a very early divergence of the lineages leading to G. sulphuraria and C. merolae (56, 57). Based on estimates that the Cyanidiophyceae split from the other red algae (Rhodophyta) approximately one billion years ago (57, 58), this also indicates that G. sulphuraria and C. merolae probably diverged about 900 million years ago.

Horizontal gene transfer (HGT)

While HGT has been observed for unicellular eukaryotes (59), its scope and function in eukaryotic evolution are not well understood. To develop an idea about the extent of horizontal gene transfer in G. sulphuraria, and to statistically confirm HGT candidates, an independent, genome-wide screen combining bioinformatics methods and manual inspection was performed (see above). This search was limited to genes that were acquired specifically by

6

Cyanidiophyceae (G. sulphuraria and C. merolae), and in most cases the genes were only observed in G. sulphuraria and no other eukaryotic organism. To be conservative, we excluded genes that might have been acquired before red alga and green plants split (60) and therefore occur in Cyanidiophyceae as well as green algae and/or land plants. Moreover, we included only genes that most likely have originated from Bacteria or Archaea. As outlined above, we attempted to exclude endosymbiotic gene transfer from cyanobacteria, i.e., the massive gene transfer from the cyanobacterial genome of the evolving plastid into the nuclear genome of the host cell during primary endosymbiosis (61). It seems more likely that a gene of potential cyanobacterial origin only observed in G. sulphuraria originates from HGT, compared to the alternative of endosymbiotic gene transfer followed by loss in all other eukaryotic photosynthetic lineages. Moreover, for genes potentially transferred from cyanobacteria to G. sulphuraria, the phylogenetic position of the descendant of the potential ‘donor’ clade (see below; table S4) was compared with the recently published phylogenetic position of the primary plastid within the cyanobacterial tree (62). HGT candidates for which the potential donor was a close neighbor of the primary plastid were excluded. Similarly, it is unlikely that HGT candidates result from endosymbiotic gene transfer that occurred in the process of mitochondrial evolution after the primary endosymbiosis of an alphaproteobacterium. HGT candidates (only observed in G. sulphuraria) of alphaproteobacterial origin seem more likely to originate from HGT, as the alternative hypothesis of endosymbiotic gene transfer requires subsequent losses in all other eukaryotic lineages. Moreover, no HGT candidate shows similarity to sequences from Rickettsia, which are believed to be similar to the primary endosymbiont (63) (table S4).

178 G. sulphuraria proteins had significant BLAST hits (score > 50) only in bacterial or archaeal sequences, meaning that no RELL analysis (see above) could be performed due to a lack of eukaryotic homologs. Out of those 178 proteins, 110 were accepted as HGT candidates after further inspection using the NCBI Web BLAST service (nr) combined with the Tree View option and criteria as described above. Phylogenetic analyses indicated that those 110 HGT candidates resulted from 25 HGT events. Members of the large Archaeal ATPase families with numerous paralogs (see fig. S11) were included here. 618 proteins had best BLAST hits in bacterial or archaeal sequences. Out of those, RELL analyses for maximum likelihood phylogenetic trees confirmed horizontal gene transfer for 163 proteins with statistical significance (5% significance level). From these 163 proteins, 50 were accepted as HGT candidates after further inspection. Phylogenetic analyses indicated that those 50 HGT candidates resulted from 44 HGT events.

Genes were excluded as HGT candidates during further inspection in case the encoded protein a) was too short (<150 amino acids); b) had too few BLAST hits; c) potentially originated from endosymbiotic gene transfer; d) resulted in a phylogenetic tree that did not allow conclusions; e) had significant sequence similarity with proteins from eukaryotic species not included in the 208 genomes used for the systematic bioinformatics screen, resulting in a phylogenetic tree of best BLAST hits (NCBI nr) that did not confirm HGT. These very stringent criteria for the manual inspection of each HGT candidate were aimed at preventing false positives as far as possible. Out of eight cases where a more detailed manual phylogenetic analysis was performed for a protein family that contained HGT candidates identified by the systematic bioinformatics screen, only one phylogenetic tree indicated a false positive (probably caused by limited phylogenomic sampling).

Together with HGT candidates identified during detailed, manual phylogenetic analyses of large protein families and of protein families showing obvious similarity to bacterial or archaeal

7

proteins, a total of 337 genes resulting from 75 horizontal gene transfers were detected (see table S4). The genome-wide bioinformatics screen detected 69 HGT events, and missed six out of 18 HGT events that were identified during detailed, manual phylogenetic analyses of protein families from G. sulphuraria. The genome-wide bioinformatics screen thus had a non-negligible number of false negatives. This is even more obvious when total numbers of HGT candidates are compared: out of a total of 337, only 160 HGT candidates were identified by the genome-wide bioinformatics screen. The low sensitivity was accepted in favor of high stringency. Some proteins did not give a sufficient number (>10) of significant BLAST hits (score >50) in our 208-species dataset, as for example the Archaeal ATPases of protein family #1, where just one family member out of 133 was detected. In other cases, best BLAST scores in eukaryotic sequences did not correctly predict the position of a G. sulphuraria protein sequence in a detailed, manually constructed phylogenetic tree, also resulting in false negatives in the systematic bioinformatics screen. The rather stringent criteria for the manual inspection of each HGT candidate that aimed at preventing false positives are likely to have increased the number of false negatives as well. Due to the high stringency, the number of 75 HGT events giving rise to 337 genes should be considered a lower bound for HGT in G. sulphuraria.

As a control for contaminations with bacterial sequences during sample preparation and sequencing, all HGT candidate genes were compared by BLASTN against the NCBI nt database. The only significant hit was with a 'Cyanidium caldarium' (a historic name for G. sulphuraria and related Cyanidiophyceae) gene. All HGT candidates were mapped onto the 433 scaffolds, and the percentage of all genes on each scaffold was compared with the percentage of HGT candidates on each scaffold. This comparison did not show any indication for an enrichment of HGT candidates on specific scaffolds, excluding the possibility that HGT candidates were largely located on small scaffolds that cannot be assembled into the overall genome. A comparison of protein sequence lengths showed little difference between HGT candidates and all other proteins.

While bacterial (or archaeal) genes usually do not have introns, 72.4% of G. sulphuraria genes contain introns (table S1). For genes transferred recently from bacteria or archaea into the G. sulphuraria genome one therefore would expect fewer or no introns. Genes inferred to originate from horizontal gene transfer (table S4) on average have 0.80 introns per gene (55% have no intron), less than half of the 2.06 introns per gene for the entire genome (median 1 intron per gene; fig. S6). This significantly smaller number of introns per gene (P = 0.0012, Mann-Whitney test) is expected for genes that are of bacterial or archaeal origin and thus initially did not contain any introns. Similarly, the GC content of genomes from different organisms can differ significantly, and as a result, genes acquired by HGT can deviate in their GC content from the GC content of the entire genome (64). For G. sulphuraria the average GC content of all 6623 annotated genes is 39.9±0.034% compared to 38.5±0.16% for all 337 HGT candidate genes, and 40.6±0.2% for the 120 HGT candidates without Archaeal ATPases (fig. S7). All 217 Archaeal ATPase genes go back to a single HGT event (fig. S9; possibly by a gene with a rather low GC content; fig. S7), but make up almost two thirds of all genes likely to result from HGT. Their inclusion would thus strongly bias comparisons of HGT vs. non-HGT genes. Comparing the GC content of 120 HGT candidates (without Archaeal ATPase genes) to the entire genome indicates a significantly higher GC content for HGT candidates (P = 0.0030, t-test).

Finally, oligonucleotide usage of genomes from different organisms can differ, and as a result, genes acquired by HGT can deviate in their oligonucleotide usage from the rest of the

8

genome. To test this, we calculated ‘genomic signatures’ (zero order Markov distributions) using di-, tri-, tetra-, and hexanucleotides (65). The signature for each gene is defined by a vector containing the relative frequencies of all words of length n (n = 2, 3, 4, or 6), divided by the relative frequencies of the individual nucleotides (65). We first calculated the signatures of the coding sequences of all non-HGT genes, and averaged them to define a genomic reference signature. We then scored the signature of each individual gene by calculating the Pearson correlation coefficient (R) between the signature of the gene and the genomic reference signature. The dinucleotide score distribution of the 120 HGT candidates (without the Archaeal ATPases) showed a significantly stronger deviation from the genomic reference signature compared to non-HGT genes (P = 0.00034) (fig. S7). We obtained similar results using signatures of tri-, tetra-, and hexanucleotides (P = 0.00011, 0.0014, and 0.011, respectively). The lower average intron number of HGT candidates (fig. S6), the significantly different GC content (fig. S7), and the deviating oligonucleotide usage (fig. S7) can be taken as additional support that these genes in G. sulphuraria did indeed originate from bacterial or archaeal genes that had no introns, on average slightly higher GC contents, and different oligonucleotide usages compared to the core genome of G. sulphuraria.

To gain insight into the possible ‘donor’ organisms from which G. sulphuraria might have acquired genes by HGT, we determined for each HGT candidate the organism containing the protein that gave the best BLAST hit, or that was closest in a phylogenetic tree (table S4). Since most HGT events in G. sulphuraria likely occurred a long time ago, the extant strains identified in this way did probably not transfer genetic material into G. sulphuraria, but are likely to be descendants (or close relatives of descendants) of the original ‘donor’ organisms. This analysis indicates that there is no special bacterial or archaeal clade that mainly ‘donated’ genes to G. sulphuraria. Instead, G. sulphuraria acquired genes from a wide variety of different clades of Archaea and Bacteria (fig. S8).

It is becoming increasingly clear that ecological similarity shapes horizontal gene transfer (66), and therefore one would expect that a large fraction of potential gene ‘donors’ lived in the same extreme environment as G. sulphuraria does (and most likely did at the time of the HGTs). Indeed, when looking at the habitat of the bacterial organisms identified as descendants of possible gene donors, almost one third (22 out of 67) are thermophilic or thermoacidophilic, compared to less than 10% of all sequenced bacterial genomes (67 out of 927 as of February 2010). Thus, there is a significant enrichment in thermophilic and thermoacidophilic bacteria among the potential ‘donors’ for HGT into G. sulphuraria (P = 7.8×10-9, Fisher's exact test; Odds Ratio = 6.28 = (22/45)/(67/860)). In addition, we calculated the number of protein families that have orthologs in extremophilic bacteria or archaea. This resulted in 66 out of 76 proteins families containing HGT candidates, compared to 1648 out of 5117 non-HGT protein families (Odds Ratio = 13.89 = (66/10)/(1648/3469)). Thus, there is a significant enrichment of protein families with orthologs in extremophiles among HGT candidates (P = 1.52×10-22, Fisher's exact test).

Most HGTs from Bacteria or Archaea into G. sulphuraria seem to be ancient according to our phylogenetic analyses. The resulting protein sequences from G. sulphuraria in most cases branch off relative ‘deep’ within phylogenetic trees, forming rather long branches, and often subsequent gene duplications into several paralogs are observed. In a few instances, HGT even happened before the G. sulphuraria and C. merolae lineages split (see, e.g., fig. S13). This long evolutionary history within the G. sulphuraria genome probably explains why 45% (152 out of

9

337) of HGT candidates have acquired introns, even though they are of bacterial or archaeal origin.

Excreted proteins

Using a newly designed rRNA-specific oligonucleotide probe (Cya1208, 5’-AGCCCAGGACATCAAAGG-3’), we detected metabolically active Galdieria spp. cells in a subsurface mine, the Richmond Mine in northern California, by FISH (fluorescence in-situ hybridization) (http://dx.doi.org/10.5061/dryad.84r5q). G. sulphuraria cells growing in environments such as the Richmond Mine are dependent on the uptake of organic nutrients from the environment to drive heterotrophic growth, as the absence of light does not permit photosynthesis. It is reasonable to hypothesize that G. sulphuraria is ‘grazing’ on sulfur-oxidizing microbial communities to satisfy its energy needs since these bacteria are the sole net producers in this environment. Likely, G. sulphuraria is able to degrade the microbial biofilm by secretion of hydrolytic enzymes that release monosaccharides and other small metabolites from the organic polymeric film matrix. Supporting this hypothesis is the fact that G. sulphuraria possess a highly cross-linked, protein-rich cell wall (67), frequently incrusted with silica (68), that protects the alga from its own extracellular enzymes. Proteomic analyses have indicated that G. sulphuraria excretes several proteins, which may be related to its heterotrophic life style or may be involved in cell wall metabolism (18, 69). Genome-wide analyses of proteins containing a predicted secretory signal peptide (39, 70) showed more than half (138 from 266, or 52%) annotated as ‘hypothetical protein’, significantly more than in the genome as a whole (36%). Similarly, out of 17 proteins that were identified in a proteomic analysis (18) as being excreted (i.e., at least five-times more spectral counts in the medium compared to the soluble fraction), eight were annotated as ‘hypothetical proteins’ due to a lack of similarity to any protein with known function.

10

Fig. S1.

Molecular divergence between G. sulphuraria and C. merolae. As a measure of molecular divergence, percent amino acid identity of orthologous gene pairs was compared for different species pairs. Sets of orthologous gene pairs were identified by reciprocal best BLAST hits with BLAST scores > 50. Normalized (to 100%) cumulative frequencies are plotted against % amino acid identity for six different species pairs. The two red algal genomes display a slightly lower amino acid identity (median 44.9%) compared to Homo sapiens / Drosophila melanogaster (45.4%). The vertebrate – insect divergence probably occurred about 910 million years ago (55), indicating a similar age for the G. sulphuraria – C. merolae split.

0

10

20

30

40

50

60

70

80

90

100

20 30 40 50 60 70 80

Cu

mu

lati

ve F

req

uen

cy (%

)

% Amino Acid Identity

G. sulphuraria / T. brucei

G. sulphuraria / H. sapiens

G. sulphuraria / O. sativa

G. sulphuraria / C. merolae

H. sapiens / D. melanogaster

H. sapiens / S. purpuratus

11

Fig. S2

Colinear regions between G. sulphuraria and C. merolae genomes. A colinearity plot of all 20 C. merolae chromosomes (Y axis) against 433 G. sulphuraria scaffolds (X axis) was generated with ColinearScan (http://colinear.cbi.pku.edu.cn/#overview) using a minimum BLAST score of 100. Blue dots indicate orthologous genes identified using InParanoid (41), which are linked by red lines if they are in synteny blocks. Genes in one block are not necessarily next neighbors, but may be separated by other genes, to allow for gene loss, gene creation, and/or minor chromosomal re-arrangements.

0 2 4 6 8 10 15 20 25 30 35 40 45 60 100 22050 432

234

5

6

7

8

1

9

10

11

12

13

14

15

16

17

18

19

20C

ya

nid

iosc

hy

zon

me

rola

e c

hro

mo

som

es

Galdieria sulphuraria scaffolds

2 M

b

2 Mb

12

Fig. S3.

Distribution of distances between coding regions in different unicellular algae, displayed as normalized (to 100%) cumulative frequencies against length (log-scale). Median values in bp are (from smallest to largest): Ostreococcus tauri, 228 (71); G. sulphuraria, 230; Ostreococcus lucimarinus, 284 (72); Thalassiosira pseudonana, 634.5 (73); Phaeodactylum tricornutum, 687 (10); C. merolae, 1404.5 (7); Chlamydomonas reinhardtii, 1534 (42); and Aureococcus anophagefferens, 1563 (74). Color coding: Chlorophyta, Rhodophyta, Stramenopiles).

0

10

20

30

40

50

60

70

80

90

100

1 10 100 1000 10000 100000

Cu

mu

lati

ve F

req

uen

cy (

%)

Distance between coding regions (bp)

G. sulphurariaC. merolaeP. tricornutumT. pseudonanaA. anophagefferensC. reinhardtiiO. lucimarinusO. tauri

13

Fig. S4.

Histogram (log-scale) for the number of introns per gene. The distribution of genes with a certain number of introns is described by an exponential decay function (light blue line),

expy A K x with start amplitude A = 2860±59 and rate constant K = 0.47±0.01.

14

Fig. S5.

Intron and exon size distributions of the G. sulphuraria genome. Size distributions of 13630 introns (top) and 20050 exons (bottom) are displayed as frequency against logarithmic length (log10 bp). (Top) The distribution of logarithmic intron lengths is described by a simple Gaussian distribution with Ampl = 2796±34, Mean = 1.752±0.0005 (56.5 bp), and SD = 0.0363±0.0005. (Bottom) The distribution of logarithmic exon lengths is more complex and was described by the sum of two Gaussian distributions

2 2

1 21 2

1 2

0.5 0.5x Mean x Mean

y Ampl exp Ampl expSD SD

with Ampl1 = 926±13, Mean1 = 2.20±0.01 (158 bp), SD1 = 0.299±0.008, Ampl2 = 422±15, Mean2 = 2.95±0.02 (891 bp), and SD2 = 0.27±0.01. Bin width is 0.02 for the upper and 0.5 for the lower graph; note different X-axis scaling in top and bottom graph.

15

Fig. S6.

Genes originating from horizontal gene transfer have fewer introns. The histograms show the percentage of genes with a given number of introns (compare fig. S4). A comparison of all 6623 annotated genes (blue) with the 337 genes likely to originate from horizontal gene transfer (orange; HGT) shows a significantly lower number of introns in genes originating from HGT (average of 0.8 introns per gene compared to 2.06 for all genes; P = 0.0012, Mann-Whitney test).

0%

10%

20%

30%

40%

50%

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Fre

qu

en

cy

# of Introns

All GenesHGT Candidates

16

Fig. S7.

Genes originating from horizontal gene transfer differ in GC content (top) and dinucleotide usage (bottom) from the rest of the genome. Top: Histogram of GC content in 1% intervals. A comparison of all 6623 annotated genes (blue) with the 120 HGT candidates (*excluding the 217 genes encoding Archaeal ATPases; orange) shows a significantly higher GC content of HGT candidates (P = 0.0030, t-test). The 217 genes encoding Archaeal ATPases (grey) show a significantly (P = 0.0035) lower GC content compared to all genes (blue). Bottom: Histogram of dinucleotide usage scores in 0.05 intervals. A comparison of all non-HGT genes (green) with the 120 HGT candidates (*excluding the 217 genes encoding Archaeal ATPases; orange) shows different dinucleotide usage scores (‘genomic signatures’). Dinucleotide usage score was defined as Pearson’s correlation coefficient R between normalized dinucleotide frequencies in each gene and the corresponding mean normalized frequencies across all G. sulphuraria non-HGT genes (for details see text). As all 217 genes encoding Archaeal ATPases result from the transfer of a single gene (fig. S9), their inclusion would strongly bias both analyses (top and bottom), and consequently these genes were excluded from the statistical comparisons.

0

5

10

15

20

25

30

22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54

Fre

qu

ency

(%

)

% GC

All Genes

HGT Candidates*

Archaeal ATPases

0

10

20

30

40

50

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Fre

qu

ency

(%

)

R (for Dinucleotide Usage)

Non-HGT Genes

HGT Candidates*

Archaeal ATPases

17

Fig. S8.

Number of horizontal gene transfers from different phyla into the genome of G. sulphuraria. For each horizontal gene transfer the phylum of the ‘donor’ organism from which the gene might have originated was determined by phylogenetic analysis or best BLAST hits (see table S4 for details).

18

Fig. S9.

Phylogenetic tree of the MNS clade of the STAND class of P-loop ATPases. The unrooted Bayesian (52) tree calculated with a CpREV+I+G model of protein evolution shows posterior probabilities above the branches and PhyML (75) bootstrap support values (using LG+I+G+F) below the branches. Thick branches indicate 1.0 posterior probability. Constraining all eukaryotic sequences into one monophyletic branch outside the Archaeal ATPases resulted in a tree with pRELL = 0.0. For clarity some sub-branches with sequences from the same clade have been collapsed (elongated triangles) with the height of the triangles reflecting the number of taxa included (3 to 22). The tree was constructed from 136 sequences: 41 from G. sulphuraria, all 56 seed sequences of the ‘Arch_ATPase’ family (PF01637) (76), plus sequences mentioned in Leipe et al. (20) plus BLAST (35) hits of these three groups for sequences outside the Archaeal ATPases. Sequences with >90 % identity were removed, and thus only one sequence from protein family #19 is included. Square brackets indicate proteins of monophyletic origin. Six families as established by Leipe et al. (20) are indicated; MJ-type, SSO-type, and PH-type families of the Archaeal ATPases, plus BL0662 family, SpsJ family, and Npun2340/2341 family outside the Archaeal ATPases. Color coding: Archaea, Cyanobacteria, other Bacteria, Rhodophyta, Amoebozoa. Bar represents 0.2 changes per site.

Dictyostelium discoideum DDBG0277693 Dictyostelium discoideum DDBG0269752

Npun2340/2341 Cyanobacteria( )

Lyngbya majuscula LYNGBM3L_74060 Trichodesmium erythraeum Tery1785

Cyanobacteria Pseudomonas syringae PSYAR_12904

Synechocystis sp. slr1243 Proteobacteria

SpsJ

( )BL0662 Actinobacteria Bacteroides thetaiotaomicron BT0660

Parabacteroides sp. HMPREF0619_03070 Thermococcus barophilus TERMP_02055 Pyrococcus furiosus PF0007

Thermococcus gammatolerans TGAM_0978 Methanosarcina acetivorans MA1854

. GI:1914801Thermococcus sp Methanocaldococcus jannaschii MJECL04

MTH196Methanothermobacter thermautotrophicus Pyrococcus horikoshii PH0977

PAB1582Pyrococcus abyssi PH0436Pyrococcus horikoshii

PAB1598Pyrococcus abyssi PAB1371Pyrococcus abyssi

Fusobacterium nucleatum FN0123 PH0846Pyrococcus horikoshii

PAB2383Pyrococcus abyssi

PH

0846

Arc

h_1

SSO1545Sulfolobus solfataricus CTN 1567Thermotoga naphthophila TM1011Thermotoga maritima

Pyrococcus SS

O1

54

5

Protein family #2

Protein family #19 Protein family #1b

Protein family #1a

Ga

ldie

ria

sulp

hu

rari

a

Methanocaldococcus jannaschii PH0541Pyrococcus horikoshii

PAB0800Pyrococcus abyssi MJ0

074

Euryarchaeota

PAE2855Pyrobaculum aerophilum

Pcal1542Pyrobaculum calidifontis Tpen1344Thermofilum pendens

Hbut0744Hyperthermus butylicus Pcal1338Pyrobaculum calidifontis

Pcal0469Pyrobaculum calidifontis Pcal0902Pyrobaculum calidifontis

Caldivirga maquilingensis Tpen1381Thermofilum pendens

Cmaq1422Caldivirga maquilingensisPyrobaculum

Cre

nar

chae

ota

Arc

ha

eal A

TP

ases

0.9870

100

100

100

100

100

0.94

0.81

0.690.8261

0.93

98

0.82

0.61

0.9986

98

0.94

0.935583

8382

0.530.70 72

0.8

0

0.97

820.77

96

0.90100

100

100100

100

100

100

10095

0.5

1

72

0.96

0.9089

98

0.88

0.5

7

0.5

9

0.97

870.580.56

0.2

19

Fig. S10.

Manual alignment of Archaeal ATPase domains of proteins from different families of the MNS clade of the STAND class of P-loop ATPases (20). Top to bottom: ten seqs. from the MNS clade not belonging to the Archaeal ATPases (six of the Npun2340/2341 family and four of the SpsJ family), 15 seqs. from the Archaeal ATPases (three each of the five families; fig. S9), and ten seqs. from G. sulphuraria (four each of protein family #1 and #2, two of family #19). An asterisk indicates a column of strict conservation. A poorly conserved region between strand 2 and 3 indicated by -X- was omitted, as was the region between strand 3 and 4 separating the two blocks of alignment shown. Secondary structure elements indicated below the alignment. Amino acids are color coded according to their biochemical properties: hydrophobic (A, F, I, L, M, V in yellow), neutral (N, Q, S, T, W in green), negative (D, E in red), positive (K, R in blue), cysteine (C in olive), glycine (G in fuchsia), histidine (H in teal), tyrosine (Y in lime), proline (P in blue).

20

Gasu_29710 Gasu_61330

Gasu_07060 Gasu_63600

Gasu_07070 Gasu_61130 Gasu_15680 Gasu_63480 Gasu_63570 Gasu_64870

Gasu_65290 Gasu_26630 Gasu_65440 Gasu_64910

Gasu_64580 Gasu_51710

Gasu_56180 Gasu_63950

Gasu_24660 Gasu_63640

Gasu_36120 Gasu_65980

Gasu_42890 Gasu_64690

Gasu_34410 Gasu_62190 Gasu_31660 Gasu_19180 Gasu_32560 Gasu_63670

Gasu_38370 Gasu_65590 Gasu_00960

Gasu_39160 Gasu_61070

Gasu_51120 Gasu_62770

Family #1a

Gasu_30810 Gasu_61090

Gasu_36780 Gasu_61670

Gasu_25720 Gasu_41410

Gasu_19170 Gasu_37570

Gasu_63280 Gasu_65430

Gasu_27640 Gasu_56170

Gasu_64600 Gasu_29700

Gasu_15690 Gasu_63200 Gasu_31650 Gasu_03880 Gasu_51100

Gasu_07080 Gasu_51130

Gasu_63590 Gasu_65050

Family #1b

Gasu_63350 Gasu_33390

Gasu_59780 Gasu_31670

Gasu_61280 Gasu_34390

Gasu_25640 Gasu_54000 Gasu_49850

Gasu_61270 Gasu_24690

Gasu_62800 Gasu_64370

Gasu_42170 Gasu_61430 Gasu_42880

Gasu_36830 Gasu_25710


Gasu_64950 Gasu_61230

Gasu_21290 Gasu_51110

Gasu_34430 Gasu_00970

Gasu_61390 Gasu_41370

Gasu_33830 Gasu_24670

Gasu_64220 Gasu_03890 Gasu_46060

Family #2

Gasu_39090 Gasu_48380


Family #19

0.5

8

0.9

0

92

97

100

0.96

0.93

0.750.75

0.75

0.75

0.75

0.750.75

0.74

0.71

0.97

99

100

0.800.820.91

0.54

0.90

0.69

0.540.51

0.50

0.2

Fig. S11.

Phylogenetic tree of G. sulphuraria protein families #1, #2, and #19. The unrooted Bayesian (52) tree calculated with a JTT+G model of protein evolution shows posterior probabilities at the branches. PhyML (75) percent bootstrap support values (using JTT+G+F) below the branches, for clarity, are only given for major branches. Thick branches indicate ≥0.99 posterior probability. The tree was constructed from all 101 complete sequences (including those >90% identical) of protein families #1, #2, and #19. Bar represents 0.2 changes per site.

21

Fig. S12.

Number of Archaeal ATPase genes per 1000 coding genes in genomes of thermophilic and hyperthermophilic organisms as function of optimum growth temperature. Genomes from 57 thermophilic and 29 hyperthermophilic archaea and bacteria were downloaded from NCBI via the “microbial genomes properties” portal (http://www.ncbi.nlm.nih.gov/genomes/lproks.cgi ). Optimum growth temperatures were obtained from NCBI. If temperatures were given as a range (58-60°C for example), the highest number was chosen (60°C in this case). In total, we obtained valid optimum growth temperatures for 38 thermophilic (including 8 archaea, in red, and 30 bacteria, in cyan) and 26 hyperthermophilic (all archaea) organisms. The HMM model of Archaeal ATPases (accession ID: PF01637.10) was downloaded from the Pfam database (76). HMMER version 3 beta (http://hmmer.janelia.org/ ) was used to score each protein sequence in each thermophilic and hyperthermophilic species against this HMM model. A protein was considered an Archaeal ATPase if the e-value of the HMM search was less than 10-5. A trend line (solid) plus correlation coefficient are given for both data sets combined. A trend line for data points from archaea only gave a similar result (dashed line; R = 0.56).

0

5

10

15

20 40 60 80 100 120

AT

Pas

es /

1000

gen

es

Optimum Growth T / °C

R = 0.72

22

Fig. S13.

Phylogenetic tree of the monovalent cation/proton antiporter-1 (CPA1, TC 2.A.36) family, indicating a high phylogenetic diversity in G. sulphuraria. The unrooted Bayesian (52) tree calculated with a Blosum62+G model of protein evolution shows posterior probabilities above branches and the PhyML (75) percent bootstrap support (using WAG+I+G+F) below the branches. Thick branches indicate ≥0.99 posterior probability. For clarity some sub-branches have been collapsed (elongated triangles), with the height of the triangles reflecting the number of taxa included (3 to 12). Curly brackets indicate different cellular locations according to TCDB annotations (38, 77). The dashed line separates the upper branch, mainly containing bacteria, from the lower two branches, mainly containing eukaryotes. Color coding: Bacteria, Streptophyta, Chlorophyta, Rhodophyta, Stramenopiles, Alveolata, Excavata, Amoebozoa, Fungi, Animals & Choanoflagellates. Bar represents 0.5 changes per site.

Gasu_24490Galdieria sulphuraria


CMR382CCyanidioschyzon merolae

CHU 1783Cytophaga hutchinsonii

Noc 0159Nitrosococcus oceani

Bacteria

APJL 1022Actinobacillus pleuropneumoniae

PM0365Pasteurella multocida

GSPATP00033903001Paramecium tetraurelia

Land Plants {Endosomal}

Green Algae

Green Algae

Fungi {Vacuolar}

Bacteria

Eukaryotes

Land Plants {Vacuolar}

CHLREDRAFT 1749Chlamydomonas reinhardtii

Chlre3 206027Chlamydomonas reinhardtii

86.m00157Tetrahymena ttetraurelia

. OSTLU 1871Osterococcus ucimarinus

Ostta 3 1816Ostreococcus tauri

Stramenopiles

152.m00122Entamoeba histolytica


Naegr1 46137Naegleria gruberi

130175Phytophthora sojae

Phyra1_1_72664Phytophthora ramorum

CMN286CCyanidioschyzon merolae


DDB0231789Dictyostelium discoideum

Monbr1 32693Monosiga brevicollis

Animals {Golgi / Intracellular}

LmjF23.0830Leishmania major



CMS152CCyanidioschyzon merolae

CMS154CCyanidioschyzon merolae

Stramenopiles

Alveolata



Alveolata

Auran1 22521Aureococcus anophagefferens

Ctha 1817Chloroherpeton thalassium

GI 1485395Strongylocentrotus purpuratus

ENSP00000356687Homo sapiens Green Algae

Land Plants {SOS1}

0.9

8

0.56

100

100

100 100

100

100

100

100

100

100

100

100

100

100

100

100

100

77

61

96

79

89

9271

99

100

100

82

80

65

100

77

9053

0.67

0.8

1

0.93

0.87

0.8

6

0.9

40.69

0.98

0.8

4

0.940.52

0.93

0.74

0.94

0.87

0.5

Intr

ace

llu

lar

Pla

sma

Mem

bra

ne

23

Fig. S14.

Phylogenetic tree of the S-adenosylmethionine-dependent methyltransferase (SAM) superfamily, indicating horizontal gene transfer of a sarcosine dimethylglycine methyltransferase gene from halophilic cyanobacteria to G. sulphuraria. The unrooted Bayesian (52) tree calculated with a WAG+I+G model of protein evolution shows posterior probabilities above the branches and PhyML (75) percent bootstrap support (using WAG+I+G) below the branches. Thick branches indicate ≥0.99 posterior probability. Square brackets group proteins according to their origin from the two domains of life (Bacteria, Eukaryota). Curly brackets indicate different subfamilies of the SAM superfamily according to NCBI CDD (27, 78). Color coding: Cyanobacteria, other Bacteria, Streptophyta, Chlorophyta, Rhodophyta, Stramenopiles, Alveolata, Animals. Bar represents 0.2 changes per site.

RS9917 03558Synechococcus sp.

WH7805 07371Synechococcus sp.

P9303_17021Prochlorococcus marinus

MIT9313 PMT0551Prochlorococcus marinus




Tgr7 1180Thioalkalivibrio sp.

Ctha 2187Chloroherpeton thalassium

NB-231 10508Nitrococcus mobilis

GI 9392587Halorhodospira halochloris

Hhal 1677Halorhodospira halophila

kuste4122Candidatus Kuenenia stuttgartiensis

GI 28316392Aphanothece halophytica

L8106 04936Lyngbya sp.

M23134 00297Microscilla marina

SvirDRAFT 04Saccharomonospora viridis

SACE 3901Saccharopolyspora erythraea

GI 9392589Actinopolyspora halophila

HochDRAFT 10430Haliangium ochraceum

FP2506 12834Fulvimarina pelagi

Bac

teri

aB

act

eri

a

Phatr2_12280Phaeodactylum tricornutum



SB234M12.18Sorghum bicolor

GI:162461873Zea mays

GI:116785181Picea sitchensis

MAL13P1.214Plasmodium falciparum

CBP02171Caenorhabditis briggsae

ENSXETP00000044432Xenopus tropicalis



Eu

kary

ota

Thaps3_269095Thalassiosira pseudonana

Auran1_59958Aureococcus anophagefferens

MICPUCDRAFT49588Micromonas pusilla

OSTLU16375Ostreococcus lucimarinus

Ostta4_34203 0800Ostreococcus tauri

0.60

100

100

100

98

80

87

56

7899

9198

820.58

0.92

0.6

8

0.83

98

98

0.9255

0.91

0.54

99

94

7051

100

0.83

0.56

0.92

0.88

0.2

Sa

rco

sin

e-D

imet

hyl

gly

cin

e M

eth

ylt

ran

sfe

rase

Ph

osp

ho

eth

ano

lam

ine

N-M

eth

yltr

an

sfe

ras

e

Cyclopropane FattyAcid Synthase

TocopherolO-Methyltransferase

TocopherolO-Methyltransferase

24

Fig. S15. 1H-NMR spectra showing choline-O-sulfate, proline betaine, and glycine betaine accumulations in G. sulphuraria. The spectra shown were obtained from crude extracts of heterotrophic cells grown under control conditions (black line) or treated with 500 mM NaCl for 84 h (red line). Extraction and measurement were performed as described earlier (79). Abbreviations used: GB, glycine betaine; PB, proline betaine; COS, choline-O-sulfate; t-but, ter-butanol. Values in ppm assigned to NMR signals are chemical shifts of either the 9 protons of the -N+(CH3)3 group of glycine betaine or choline-O-sulfate, the 3 protons of each of the methyl group of the -N+(CH3)2 or those of the methylene group of those substances when detectable. Ter-butanol was selected as an internal standard for quantification of betaines. The 9 protons of its -C(CH3)3 group give a sole signal at 1.200 ppm. The signals at 3.181, 3.217, and 3.251 ppm were taken as specific target for quantification of choline-O-sulfate, glycine betaine and proline betaine, respectively. The spectra are presented at the same scale (standardized with ter-butanol).

25

Fig. S16.

Phylogenetic tree of the anion permease ArsB/NhaD superfamily, indicating horizontal gene transfer of an ArsB gene from bacteria to G. sulphuraria. The unrooted Bayesian (52) tree calculated with a CpREV+I+G model of protein evolution shows posterior probabilities above the branches and PhyML (75) percent bootstrap support (using LG+I+G) below the branches. Thick branches indicate 1.0 posterior probability. Square brackets group proteins according to their origin from Eukaryota, Archaea, or different phyla of Bacteria. Curly brackets indicate different subfamilies of the anion permease ArsB/NhaD superfamily (27). The label ‘Thermophilic/Acidophilic’ marks G. sulphuraria and four bacteria being thermophilic and/or acidophilic, living in similar habitats as G. sulphuraria does. Color coding of major phylogenetic groups: Bacteria, Archaea, Streptophyta, Chlorophyta, Rhodophyta, Stramenopiles, Excavata, Fungi, Animals & Choanoflagellates. Bar represents 0.5 changes per site.

FBgn0051693Drosophila melanogaster FBgn0035332melanogasterDrosophila

FBgn0036329melanogasterDrosophila FBgn0041150 melanogasterDrosophila

ENSXETP00000009004Xenopus tropicalis ENSP00000346659Homo sapiens MGI 97454Mus musculus

29712Monosiga brevicollis Ostreococcus tauri 8380

44788Phaeodactylum tricornutum T 262258halassiosira pseudonana

182280Physcomitrella patens 190982Physcomitrella patens

139294Physcomitrella patens Os03g01700Oryza sativa

Os10g39980Oryza sativa Os02g57620Oryza sativa

Os03g05390Oryza sativa GSVIVP00037686001Vitis vinifera GSVIVP00000441001Vitis vinifera GSVIVP00032514001Vitis vinifera

69397Naegleria gruberi 93857 m00324Trypanosoma vaginalis

91127m00285Trypanosoma vaginalis 91660 m00056Trypanosoma vaginalis

174287Chlamydomonas reinhardtii AAW41753Cryptococcus neoformans

CHG09358Chaetomium globosum Afu2g15090Aspergillus fumigatus AN5302Aspergillus nidulans

Eu

kary

ota

TVN0285Thermoplasma volcanium GI 16082492Thermopasma acidophilum

050500004700Ferroplasma acidarmanus PTO0083Picrophilus torridus A

rch

ae

a

ArsBListeria welshimeri SSPP115Staphylococcus saprophyticus

ArsBBacillus halodurans ArsBEscherichia coli

PROPEN_04102Proteus penneri Psyr_1503Pseudomonas syringae

ArsB-1Pseudomonas putida ArsB (Deinococcus-Thermus)Deinococcus geothermalis


Th

erm

op

hili

cA

cid

op

hili

c

Ars

B,

arsen

ical

pu

mp

mem

bra

ne

pro

tein

An

ion

per

mea

seA

rsB

/Nh

aD

Pe

rme

ase

P/

P p

rote

in

Sili

con

eff

lux t

ran

sp

ort

er(a

nio

n p

erm

ease

Yb

iR)

Gasu_31570Galdieria sulphuraria ArsB (Nitrospirae)Leptospirillum ferriphilum

sp. ArsB1Thiomonas

ArsBPsychrobacter arcticus

Lferr_2479Acidithiobacillus ferrooxidans ArsB2Roseomonas cervicalis ArsBMethylobacterium extorquens Mrad2831_1380Methylobacterium radiotolerans

Pro

teo

bac

teri

a

Proteobacteria

Firmicutes

0.51

0.6

9

0.59

61

60

58

0.560.92

68

0.9

0

71

98

92

0.84

84

9396

89

78100

100100

100

100

100

100

100100

100

100

100

100

100

100

0.9

8

80

0.65

0.9

7

96

0.8880

69

0.97

58

0.98

0.9

3

0.56

92

63100

0.5

26

Fig. S17.

Phylogenetic tree of mercuric reductase. The unrooted Bayesian (52) tree calculated with a WAG+G model of protein evolution shows posterior probabilities above the branches and PhyML (75) percent bootstrap support (using WAG+G+F) below the branches. Thick branches indicate 1.0 posterior probability. For bacterial sequences the class for each species is given in abbreviated form for Gammaproteobacteria and Betaproteobacteria. Color coding of major phylogenetic groups: Bacteria, Rhodophyta. Bar represents 0.1 changes per site.

merA, Methylococcus capsulatus Gammaprot’

merA, Pseudomonas aeruginosa Gammaprot’

mer, Stenotrophomonas maltophilia Gammaprot’

Daci0483, Delftia acidovorans Betaprot’

merA, Pseudomonas fluorescens Gammaprot’

merA2, Pseudomonas aeruginosa Gammaprot’

Nhal1676, Nitrosococcus halophilus Gammaprot’

sp. MerA, Thiobacillus Betaprot’

merA, Acidithiobacillus ferrooxidans Gammaprot’

merA, Salmonella enterica Gammaprot’

sp. B1M04319, Burkholderia Betaprot’

merA, Enterobacter cloacae Gammaprot’

Neut0030, Nitrosomonas eutropha Betaprot’

sp. MerA, Bacillus Bacilli

merA, Proteus mirabilis Gammaprot’

Galf1900, Gallionella capsiferriformans Betaprot’

Tint1608, Thiomonas intermedia Betaprot’

CMJ014CCyanidioschyzon merolae

Tbd1341, Thiobacillus denitrificans Betaprot’


sp. BW13 merA, Pseudomonas Gammaprot’

Sfri3488, Shewanella frigidimarina Gammaprot’

sp. Sputw3181_3208, Shewanella Gammaprot’

Sputcn32_0168, Shewanella putrefaciens Gammaprot’

sp. Shewana3_4311, Shewanella Gammaprot’

merA, Escherichia coli Gammaprot’

100

100

98

83

90

99

0.86/550.67

0.95/80

0.87

0.98100

0.64

0.9851

59

65

99

95

94

92

8968

0.90

100

0.1

27

Fig. S18.

Percent of proteome annotated as membrane transporter for 49 different eukaryotes. The red line indicates the 5% level. Total numbers of transporter proteins were taken from TransportDB (36). The TransAAP tool from TransportDB annotated 368 membrane transport proteins for the G. sulphuraria genome. After manual curration this number was corrected to 247 membrane transport proteins; for consistency, only those were considered here.

0 1 2 3 4 5 6 7

Physcomitrella patensArabidopsis thaliana

Oryza sativaOstreococcus sp.

Ostreococcus tauriMicromonas pusilla CCMP1545

Micromonas pusilla RCC299Chlamydomonas reinhardtii

Cyanidioschyzon merolaeGaldieria sulphuraria

Phytophthora infestansPhytophthora ramorum

Phytophthora sojaeEctocarpus siliculosus

Aureococcus anophagefferensPhaeodactylum tricornutumThalassiosira pseudonana

Babesia bovisBabesia equi

Cryptosporidium parvumPlasmodium falciparum 3D7

Plasmodium vivaxTheileria parva

Toxoplasma gondii B7Paramecium tetraurelia

Tetrahymena thermophilaTrichomonas vaginalis

Leishmania majorTrypanosoma brucei

Trypanosoma cruziDictyostelium discoideum

Entamoeba disparEntamoeba histolyticaEntamoeba invadensAspergillus fumigatus

Aspergillus nidulansAspergillus oryzaeNeurospora crassa

Saccharomyces cerevisiaeSchizosaccharomyces pombe

Cryptococcus neoformansEncephalitozoon cuniculi

Brugia malayiCaenorhabditis elegans

Aedes aegyptiAnopheles gambiae

Drosophila melanogasterMus musculusHomo sapiens

% of proteins

28

Animals

Monosiga brevicollis Monbr1_10586

Land Plants

Trichomonas vaginalis


Chlre3_196325Chlamydomonas reinhardtii

Ostta4_29494Ostreococcus tauri

SS1G_01070Sclerotinia sclerotiorum

Animals

Animals

. 136367Physcomitrella p

. 113769Physcomitrella p

Naegr1_65611Naegleria gruberi

Alveolata

Alveolata

Saccharomyces cerevisiae

Ustilago maydis UM05023

Phytophthora

Cyanidioschyzon merolae CMK066

Oryza sativa Os07g0559700

Bacteria

Bacteria

Bacteria

Haloarcula marismortui GI:55376617

sp. RHA1_ro01361Rhodococcus

TcaBSaccharopolyspora erythraea

GbCGDNIH1_1264Granulibacter bethesdensis

sp. GI:169189897Paenibacillus

Animals

Galdieria sulphuraria Gasu_04230

Ostreococcus lucimarinus OSTLU34203

Aureococcus anophagefferens Auran1_1914


Land Plants

Land Plants

Monbr1_8633Monosiga brevicollis

MONBRDRAFT_16319Monosiga brevicollis

Stramenopiles

Fungi

Land Plants

Excavata

Phytophthora

Aureococcus anophagefferens Auran1_71194

Phaeodactylum tricornutum Phatr2_27361

Animals

Land Plants

Fungi

Galdieria sulphuraria


SPT1

VGT1

SPT2

Frt1








Gasu_57830.Galdieria sulphuraria

Ascomycota (Fungi)

100

100

100

81

100

52

0.98

0.68

0.83

0.54

0.5

8

0.84

0.5

40.

810

.95

0.62

0.94

0.9

10.94

0.8

7

0.76

0.57

0.94

0.83

57

73

0.99

100

100

100

100

9999

99

70

98

99

58

100

0.95

0.90

0.62

100

10068

86

100100

100

78

99

100

0.97

100

100

100

0.57

540.74

0.89

0.6

30.81

100

0.91

0.66

100

75

100

97

5480

100

0.75

100

82

0.70

0.72100

1000.62

99

100

0.2

TC 2.A.1.1.69

TC 2.A.1.1.45

TC 2.A.1.1.33

TC 2.A.1.1.70

29

Fig. S19.

Phylogenetic tree of the sugar porter (SP) family of the major facilitator superfamily (MFS). The unrooted Bayesian (52) tree calculated with a CpREV+I+G model of protein evolution shows posterior probabilities above the branches and PhyML (75) bootstrap support values (using LG+I+G+F) below the branches. Thick branches indicate 1.0 posterior probability. For clarity most sub-branches have been collapsed (elongated triangles) with the height of the triangles reflecting the number of taxa included (3 to 29). Curly brackets indicate different subfamilies of the sugar porter family according to the Transporter Classification Database (38). The founding member of each subfamily is indicated in white on the collapsed branch where it is included. Color coding of major phylogenetic groups: Archaea, Bacteria, Streptophyta, Chlorophyta, Rhodophyta, Stramenopiles, Alveolata, Excavata, Amoebozoa, Fungi, Choanoflagellates & Animals. Bar represents 0.2 changes per site.

30

Fig. S20.

Phylogenetic tree of the amino acid / auxin permease (AAAP) family. The unrooted Bayesian (52) tree calculated with a CpREV+I+G model of protein evolution shows posterior probabilities above the branches and PhyML (75) percent bootstrap support (using LG+G+F) below the branches. Thick branches indicate 1.0 posterior probability. For clarity most sub-branches have been collapsed (elongated triangles) with the height of the triangles reflecting the number of taxa included (3 to 46). Curly bracket indicates a subfamily of fungal amino acid permeases (TC 2.A.18.4), which includes all G. sulphuraria sequences, according to best BLAST hits in the Transporter Classification Database (38). Color coding of major phylogenetic groups: Streptophyta, Chlorophyta, Rhodophyta, Stramenopiles, Excavata, Amoebozoa, Fungi, Choanoflagellates & Animals. Bar represents 0.2 changes per site.

Fungi (Ascomycota)



Fungi

Pyrenophora tritici-repentis GI:189189672

Fungi






MONBRDRAFT_38331Monosiga brevicollis


Ot03g02520Ostreococcus tauri

UM03049.1Ustilago maydis

Land Plants

Phytophthora

Animals

Land Plants



88788.m00034Trichomonas vaginalis

93204.m00168Trichomonas vaginalis

Animals

Fungi

Entamoeba histolytica


Excavata

0.57500.92

100

100

100

100

100

100

100

100

100

100

100

100

98

52

73

95

94

56

58

68

99

98

90

94

63

0.53

0.95

0.53

0.99

0.93

0.51

0.94

0.53

0.51

0.66

0.59

0.57

0.83

0.5

TC

2.A

.18.

4 (

fun

gal

per

me

ases

)

31

Fig. S21.

Phylogenetic tree of acetate permeases of the YaaH family. The unrooted Bayesian (52) tree calculated with a Blosum62+G model of protein evolution shows posterior probabilities above the branches and PhyML (75) percent bootstrap support (using LG+G+F) below the branches. Thick branches indicate 1.0 posterior probability. Three subfamilies of acetate permeases according to the Transporter Classification Database (38) are indicated, with the founding member of each subfamily indicated in white on the collapsed branch where it is included. Color coding of major phylogenetic groups: Archaea, Bacteria, Streptophyta, Chlorophyta, Rhodophyta, Stramenopiles, Alveolata, Excavata, Fungi). Bar represents 0.2 changes per site.

Leishmania major LmjF03.0400


Ichthyophthirius multifiliis IMG5_002930

Calditerrivibrio nitroreducens Calni_1246

Dbac_3478Desulfomicrobium baculatum

MCON_2287Methanosaeta concilii

CfE428DRAFT_126Chthoniobacter flavus

Gmet_1126Geobacter metallireducens

Mhun_0634Methanospirillum hungatei

Mboo_0065Methanoregula boonei

sp. CAT7_04262Carnobacterium

Cphy_0364Clostridium perfringens

MCP_2738Methanocella paludicola

archaeon RC-I RCIX138

TC 2.A.96.1.1

sp. PCNPT3_10676Psychromonas

Desal_0840Desulfovibrio salexigens

sp. GM18DRAFT 1497Geobacter

Gammaproteobacteria

TPY_0697Sulfobacillus acidophilus

sp. BH160DRAFT_3638Burkholderia

BuboB_01010000444Burkholderia ubonensis

GI:218289874Alicyclobacillus acidocaldarius

Euryarchaeota

Acel_1067Acidothermus cellulolyticus

Namu_0029Nakamurella multipartita

Gobs_1305Geodermatophilus obscurus

Firmicutes











Eukaryota

Fungi

Bacterial origin Actinobacteria

Green Algae


Cyanidioschyzon merolae CMM079C

Leishmania major LmjF03.0370

Physcomitrella

Aspergillus oryzae Ao090701000842

GI:57227606Cryptococcus neoformans

Ustilago maydis UM00196.1

SPAC5D6.09cSchizosaccharomyces pombe

TC 2.A.96.1.3

TC 2.A.96.1.4

AcpA

YaaH

Ady2

0.83

0.87

0.690.56

0.95

0.83

0.56

0.79100

0.72

0.76

0.78

0.61

0.72

0.79

0.79

0.9493

74

59

100

10088

74

75

75

70

92

62

100

100

0.66

0.96

0.99

0.96

0.96

0.60

0.85

0.60

0.99

0.8

5

0.9292

97

78

78

75

8060

71

99

0.98

0.98

0.850.940.2

32

Fig. S22.

Phylogenetic tree of amino acid-polyamine-organocation (APC) superfamily (TC 2.A.3). The unrooted Bayesian (52) tree calculated with a Blosum62+G model of protein evolution shows posterior probabilities above the branches and PhyML (75) percent bootstrap support (using LG+G+F) below the branches. Thick branches indicate 1.0 posterior probability. Curly brackets indicate different families of the amino acid-polyamine-organocation superfamily, according to best BLAST hits in the Transporter Classification Database (38). Color coding of major phylogenetic groups: Archaea, Bacteria, Streptophyta, Chlorophyta, Rhodophyta, Stramenopiles, Alveolata, Excavata, Amoebozoa, Fungi, Animals. Bar represents 0.2 changes per site.

Land Plants

Stramenopiles

Phytophthora ramorum Phyra1_1_95019



Cyanidioschyzon merolae CMR055C




EHI 020320Entamoeba histolytica

Excavata The

Po

lya

min

e:H

Sym

port

er

(PH

S)

fam

il y (

TC

)

+

2.A

.3.1

2

Animals

Fungi

Fungi

Fungi

CHG02138.1Chaetomium globosum

Trichomonas vaginalis

AAW41080.1Cryptococcus neoformans

SSO1009Sulfolobus solfataricus





Ta0877Thermoplasma acidophilum

TVN0718Thermoplasma volcanium

Archaea

CME062CCyanidioschyzon merolae




Tb927.6.4660 [2.A.3.3.1]Trypanosoma brucei

Tc00.104705350683Trypanosoma cruzi

Land Plants

Bacteria

Green Algae

3386891 ctrADictyostelium discoideum

3386917 ctrCDictyostelium discoideum

Phytophthora ramorum Phyra1_1_72585

Phytophthora sojae 130075

3392553 ctrBDictyostelium discoideum

Animals

Land Plants

Ca

tion

ic A

min

o a

cid

Tra

nsp

orte

r (C

AT

)fa

mily

(T

C

)2.

A.3

.3

0.99

0.6762

9592

79

9899

99

99

53

53

92

9356

59

99

75

79

76

98

98

93

8992

97

100

100

100

100

100

100

100

100

100

100

100

100

0.50

0.92

0.59

0.87

0.610.88

0.96

0.97

0.95

0.84

0.84

0.88

0.81

0.5

70.

50

0.5

8

0.7

9

0.2

33

Fig. S23.

Phylogenetic tree of transporters from the major intrinsic protein (MIP; TC 1.A.8) family. The unrooted Bayesian (52) tree calculated with a Blosum62+I+G model of protein evolution shows posterior probabilities above the branches and PhyML (75) percent bootstrap support (using LG+I+G+F) below the branches. Thick branches indicate ≥0.99 posterior probability. The glycerol uptake facilitator (TC 1.A.8.1.1) subfamily of the MIP family according to best BLAST hits at the Transporter Classification Database (38) is indicated by a curly bracket. All other sequences shown belong to different MIP subfamilies. Color coding of major phylogenetic groups: Bacteria, Streptophyta, Chlorophyta, Rhodophyta, Stramenopiles, Alveolata, Excavata, Fungi, Animals & Choanoflagellates. Bar represents 0.2 changes per site.

Land Plants

Drosophila melanogaster FBgn0015872

ENSAPMP00000029623Apis mellifera






NCU08052Neurospora crassa

Cryptococcus neoformans gi:57225766


Bacteria






Proteobacteria

Bacteria

Bacteria

gi:348669471Phytophthora sojae

gi:348669470Phytophthora sojae

gi:325188770Albugo laibachii



PF11_0338Plasmodium falciparum

1963.m00065Plasmodium yoelii

Animals

Tb10.61.2650Trypanosoma brucei




Fungi

0.96

0.66

0.68

0.7

9

0.94

0.98

0.93

64

97

72

67

82

86

80

99

75

6390

97

55

92

100

100

100

100

100

100

100

100

100

0.65

0.58

0.2

Gly

cer

ol u

pta

ke f

acil

itat

or

(TC

1.A

.8.1

.1)

34

Fig. S24.

Phylogenetic tree of glycerol dehydrogenases (EC 1.1.1.6). The unrooted Bayesian (52) tree calculated with a WAG+I+G model of protein evolution shows posterior probabilities above the branches and PhyML (75) percent bootstrap support (using LG+I+G) below the branches. Thick branches indicate 1.0 posterior probability. The collapsed branch labeled ‘Bacteria (mostly Firmicutes)’ contains sequences from two Actinobacteria, two Synergistetes, and 15 Firmicutes. Color coding of major phylogenetic groups: Cyanobacteria, other Bacteria, Rhodophyta, Excavata, Amoebozoa, Fungi. Bar represents 0.2 changes per site.

Bacteria (mostly Firmicutes)

gi:257125507 (Fusobacteria)Leptotrichia buccalis

Dictyostelium discoideum DDB0217165

gi:330844838Dictyostelium purpureum

sp. gi:312111935 (Firmicutes)Geobacillus

gi:237808641 (Proteobacteria)Tolumonas auensis

gi:310778873 (Fusobacteria)Ilyobacter polytropus

gi:51892405 (Firmicutes)Symbiobacterium thermophilum

gi:51892405 (Firmicutes)Symbiobacterium thermophilum

sp. gi:254229761 (Proteobacteria)VibrioBacteria (Proteobacteria)

gi:289523705 (Synergistetes)Anaerobaculum hydrogeniformans

gi:357419061 (Synergistetes)Thermovirga lienii

gi:296132160 (Firmicutes)Thermincola potens

sp. gi:242133534Crithidia

gi:302391592 (Firmicutes)Acetohalobium arabaticum

gi:188587303 (Firmicutes)Natranaerobius thermophilus




gi:81300462 Synechococcus elongatus

gi:88703938 (Proteobacteria)Congregibacter litoralis

gamma proteobacterium NOR5-3 gi:254515406

Schizophyllum commune gi:302691994

gi:302696107Schizophyllum communeFungi (Basidiomycota)

mGDHSchizosaccharomyces pombe

gi:21340879Schizosaccharomyces japonicusFungi (Ascomycota)

gi:228912388 (Firmicutes)Bacillus thuringiensisBacteria (Firmicutes)

0.98

0.6

3

0.6

4

0.96

0.83

0.98

0.9

4

0.78 75

54

6265

52

7699

88

92

62

100

100

100

100

100

100

0.9

0

0.9

0

0.9

0

0.2

35

Fig. S25.

Phylogenetic tree of the Pi:H+ symporter (PHS) family and related families of the MFS superfamily. The unrooted Bayesian (52) tree calculated with a Blosum62+I+G model of protein evolution shows posterior probabilities above the branches and PhyML (75) percent bootstrap support (using LG+I+G+F) below the branches. Curly brackets indicate different families of the MFS (major facilitator superfamily) according to the Transporter Classification Database (38). Color coding of major phylogenetic groups: Archaea, Bacteria, Streptophyta, Chlorophyta, Rhodophyta, Stramenopiles, Alveolata, Amoebozoa, Fungi, Animals. Bar represents 0.5 changes per site.

GI:154707566Coxiella burnetii

GI:153207273Coxiella burnetii

sp. GI:195953069Hydrogenobaculum

Archaea

GI 14324338Thermoplasma volcanium

0.7

2

57 56

63100

Animals

Ph

osp

ha

te:H

+ S

ymp

ort

er (

TC

2.A

.1.9

)

Oth

er

MF

S T

ran

sp

ort

er

Animals

392.m00008Tetrahymena thermophila



PHUM188680Pediculus humanus

CPIJ011499Culex quinquefasciatus


Ostta4_19850Ostreococcus tauri

OSTLU_38010Ostreococcus lucimarinus

PHYPADRAFT136Physcomitrella

LOC100125647Zea mays

AT3G13050Arabidopsis thaliana

Os09g0559800Oryza sativa

0.98

0.9

1

0.97

0.83

61

66

98

97

97

9688

100

100

100100

100

100

Fungi






CMT284CCyanidioschyzon merolae










Fungi


76690Phytophthora ramorum Phyra1_1_

76686Phytophthora ramorum Phyra1_1_

GI:167393077Entamoeba dispar


Fungi



Land Plants

0.66

0.84

0.57

74

97

96

6185

84

83

97

95

51

54

57

57

98

96

95

9585

100

100

100

100

100

100

100

100

100

100

0.950

.90

0.86

0.5

36

Fig. S26.

Phylogenetic tree of -galactosidases (EC 3.2.1.23; glycosyl hydrolases family 15), indicating two different subfamilies in G. sulphuraria. The unrooted Bayesian (52) tree calculated with a WAG+I+G model of protein evolution shows posterior probabilities above the branches and PhyML (75) percent bootstrap support (using LG+I+G) below the branches. Thick branches indicate 1.0 posterior probability. Numbers in brackets give number of introns for different G. sulphuraria genes; ‘SP’ indicates a secretory signal peptide according to SignalP 4.0 (39). Color coding of major phylogenetic groups: Bacteria, Streptophyta, Rhodophyta, Stramenopiles, Alveolata, Amoebozoa, Fungi, Animals & Choanoflagellates. Bar represents 0.2 changes per site.

Land Plants

DDBDRAFT_0186630Dictyostelium discoideum

PPL 03214Polysphondylium pallidum

GI:188501572Adineta vaga

Stramenopiles

CMP078CCyanidioschyzon merolae

Gasu_40850.1 (10)Galdieria sulphuraria




Bacteria


Gasu_27490.1 (1) ( )Galdieria sulphuraria SP

Gasu_27500.1 (0) ( )Galdieria sulphuraria SP

Fungi




Bacteria

Animals

sp. GI:192814064Geobacillus

s sp. GI:197330841Streptomyce

GI:197932226Streptomyces scabies

Land Plants




Animals

0.9774

73

9784

100

100

99

100

100

100

100

100

100

100

100

100

100

100

0.960.83

0.76

96

0.6

0

0.75

0.66

0.5

8 0.99550.99

0.6

8

0.68 89

0.6

8

560.99

0.2

37

Fig. S27.

Phylogenetic tree of acid phosphatases (EC 3.1.3.2). The unrooted Bayesian (52) tree calculated with a WAG+I+G model of protein evolution shows posterior probabilities above the branches and PhyML (75) percent bootstrap support (using LG+I+G) below the branches. Thick branches indicate 1.0 posterior probability. ‘SP’ indicates a secretory signal peptide according to SignalP 4.0 (64). Curly bracket marks a group of organisms that all are acidophilic. Color coding of major phylogenetic groups: Cyanobacteria, other Bacteria, Streptophyta, Rhodophyta, Excavata, Fungi. Bar represents 0.2 changes per site.

Fungi

gi:328772443Batrachochytrium dendrobatidis


Phypa1_1_51671Physcomitrella

sp. gi:209517346 (Betaproteobacteria)Burkholderia

sp. gi:325521954 (Betaproteobacteria)Burkholderia

gi:225874284 (Acidobacteria)Acidobacterium capsulatum

gi:167567471 (Betaproteobacteria)Burkholderia oklahomensis

sp. gi:309779586 (Betaproteobacteria)Ralstonia

sp. gi:119717424 (Actinobacteria)Nocardioides

gi:359774813 (Actinobacteria)Arthrobacter globiformis

(Acidobacteria)Candidatus Koribacter versatilis

gi:117927441 (Actinobacteria)Acidothermus cellulolyticus

gi:388442093 (Gammaproteobacteria)Rhodanobacter thiooxydans

gi:182678150 (Alphaproteobacteria)Beijerinckia indica

gi:186682052 (Cyanobacteria)Nostoc punctiforme

gi:269839860Thermobaculum terrenum

(Nitrospirae)Leptospirillum

gi:296164862 (Actinobacteria)Mycobacterium parascrofulaceum

gi:342861237 (Actinobacteria)Mycobacterium colombiense

gi:339629066 (Firmicutes)Sulfobacillus acidophilus

gi:339626434 (Firmicutes)Sulfobacillus acidophilus

sp. gi:294341316 (Betaproteobacteria)Thiomonas

Gasu_56160 ( )Galdieria sulphuraria SP

Gasu_24940Galdieria sulphuraria ( )SP




0.97 97

97

86

62

63

55

64

99

94

97

80

68

100

100

100

100

100

100

100

100

100

0.95

0.76

0.7

5

0.6

40.

55

0.7

40

.85

0.5

9

0.2

acid

op

hili

c

38

Table S1.

Orthologous relationships between G. sulphuraria and C. merolae genomes. For intron and exon size distribution in G. sulphuraria see fig. S5, for a comparison of intergenic distances see fig. S3.

G. sulphuraria C. merolae

Genome size (Mb) 13.7 16.5

GC (%) 37.7 55.0

GC (%) coding sequence (excluding introns) 38.6 56.7

CpG occurrence (Obs./Exp.) 0.715 1.151

CpG islands 27 2

Repeat content 713 313

Number of rRNA units (18S/5.8S/28S + 5S) 4 + 10 3 + 3

Number of tRNAs 85 30

Predicted proteins 6623 5771

Gene density (kb per gene) 2.07 2.86

Average gene length (bp) 1601 1553

Average transcript length (bp) 1388 1552

Average number of amino acids per polypeptide 421 518

Average number of exons per gene 3.16 1.005

Average exon length (bp) 417.3 1527.6

Introns 13630 26

Genes with introns (%) 72.4 0.5

Average intron length (bp) 56.5 248

Median intergenic distance (bp) 20.0 1404.5

Coding sequence (%) 77.5 44.9

39

Table S2.

Large protein families in G. sulphuraria. Families of paralogous proteins were determined by BLASTing the G. sulphuraria genome against itself, followed by MCL clustering (see Materials and Methods). Protein families were ranked by size. When annotations differed for members of one family, the annotation for the majority is given. Representative protein domains are indicated by PF numbers (76) or IPR numbers (33); brackets indicate low significance or hits in only some family members.

Rank Size Annotation Protein domains

1 131 Archaeal ATPase PF01637

2 76 Archaeal ATPase PF01637 & PF00536

3 48 Sugar Porter (SP) family of MFS superfamily PF00083

4 29 serine/threonine protein kinase IPR002290

5 26 AAA ATPase PF00004

6 25 DEAD-box RNA helicase PF00270 & PF00271

7 22 Pi:H+ symporter (PHS) family of MFS superfamily PF07690

8 15 Mitochondrial Carrier (MC) family PF00153

9 13 Amino Acid / Auxin Permease (AAAP) family PF01490

10 13 aldehyde dehydrogenase (NAD+) PF00171

11 12 cyclin-dependent serine/threonine protein kinase IPR002290

12 12 ubiquitin-conjugating enzyme E2 PF00179

13 12 kinesin family member PF00225

14 11 ABC transporter, multidrug resistance PF01061

15 11 ADP-ribosylation factor PF00025

16 11 putative acetate transporter PF01184

17 11 hypothetical protein (PF01113)

18 11 hypothetical protein (IPR008238)

19 10 Archaeal ATPase PF01637

20 10 hypothetical protein (IPR015598)

40

Table S3.

Comparison of membrane transport proteins in G. sulphuraria and C. merolae. Annotated membrane transport proteins classified as ‘ATP-Dependent’, ‘Ion Channel’, or ‘Secondary Transporter’ are listed, and the number of transporters per family is given. The classification of transporters into families follows the Transporter Classification Database (38, 77). Species and numbers of transporters per family are as in fig. S18. Transporter numbers were normalized to total number of proteins for each genome before calculating z-scores (z-scores >2.0 in bold).

z-score ATP-Dependent 60 102 0.58 2.12

The ATP-binding Cassette (ABC) Superfamily 28 49 0.24 1.42 The Arsenite-Antimonite (ArsA) Efflux Family ATP-hydrolizing Subunit 3 3 3.44 2.85

The H+- or Na+-translocating F-type, V-type and A-type ATPase (F-ATPase) 3 3 1.72 1.33 The H+-translocating Pyrophosphatase (H+-PPase) Family 1 3 0.27 1.97

The Type II (General) Secretory Pathway (IISP) Family 4 11 -0.02 1.82 The Mitochondrial Protein Translocase (MPT) Family 11 18 1.01 1.89

The P-type ATPase (P-ATPase) Superfamily 10 15 -0.24 0.52 Ion Channel 13 19 -0.61 -0.46

The Ammonia Transporter Channel (Amt) Family 2 2 0.38 0.19 The Anion Channel-forming Bestrophin (Bestrophin) Family 4 0 3.17 -0.36

The Intracellular Chloride Channel (CLIC) Family 0 4 -0.42 5.11 The Major Intrinsic Protein (MIP) Family 1 5 -0.46 0.84

The CorA Metal Ion Transporter (MIT) Family 0 3 -1.12 0.82 The Small Conductance Mechanosensitive Ion Channel (MscS) Family 2 2 0.21 0.10

The Chloroplast Envelope Anion Channel-forming Tic110 (Tic110) Family 1 1 3.03 2.57 The Voltage-gated Ion Channel (VIC) Superfamily 3 1 -0.49 -0.64

Secondary Transporter 108 225 -0.24 1.08 The ATP:ADP Antiporter (AAA) Family 1 1 0.32 0.24

The Amino Acid / Auxin Permease (AAAP) Family 1 17 -0.76 0.89 The Anion Exchanger (AE) Family 3 2 1.98 0.81

The Auxin Efflux Carrier (AEC) Family 0 1 -0.75 0.02 The Amino Acid-Polyamine-Organocation (APC) Family 2 8 -0.39 0.24

The Arsenite-Antimonite (ArsB) Efflux Family Transmembrane Subunit 0 2 -0.56 3.55 The Bile Acid:Na+ Symporter (BASS) Family 3 3 1.56 1.27

The Ca2+:Cation Antiporter (CaCA) Family 1 1 -0.70 -0.77 The Cation Diffusion Facilitator (CDF) Family 3 1 0.50 -1.23

The Chloride Carrier/Channel (ClC) Family 2 5 0.34 2.39 The Monovalent Cation:Proton Antiporter-1 (CPA1) Family 4 5 1.40 1.68 The Monovalent Cation:Proton Antiporter-2 (CPA2) Family 2 2 0.62 0.44

The Divalent Anion:Na+ Symporter (DASS) Family 2 1 0.75 -0.21 The Drug/Metabolite Transporter (DMT) Superfamily 18 20 0.79 0.72

The Folate-Biopterin Transporter (FBT) Family 1 0 -0.30 -0.72 The Glycoside-Pentoside-Hexuronide (GPH):Cation Symporter Family 1 5 -0.03 2.71

The Hydroxy/Aromatic Amino Acid Permease (HAAAP) Family 0 1 -0.51 0.97 The K+ Uptake Permease (KUP) Family 1 2 0.62 1.34

The Lysosomal Cystine Transporter (LCT) Family 2 2 1.86 1.48 The Mitochondrial Carrier (MC) Family 31 34 1.68 1.52

The Major Facilitator Superfamily (MFS) 13 83 -0.62 0.76 The Multidrug/Oligosaccharidyl-lipid/Polysaccharide (MOP) Flippase Superfamily 2 3 -0.25 -0.06

The Nucleobase:Cation Symporter-1 (NCS1) Family 0 2 -0.46 0.46 The Nucleobase:Cation Symporter-2 (NCS2) Family 0 2 -0.67 1.47

The NhaD Na+:H+ Antiporter (NhaD) Family 1 1 3.72 3.18 The Metal Ion (Mn2+-iron) Transporter (Nramp) Family 3 4 2.70 3.28

The Cytochrome Oxidase Biogenesis (Oxa1) Family 2 5 1.12 3.63 The Inorganic Phosphate Transporter (PiT) Family 2 1 1.18 -0.01

The Proton-dependent Oligopeptide Transporter (POT) Family 0 1 -0.52 -0.22 The Resistance-Nodulation-Cell Division (RND) Superfamily 0 1 -0.93 -0.54

The Solute:Sodium Symporter (SSS) Family 1 1 -0.08 -0.17 The Sulfate Permease (SulP) Family 2 2 -0.11 -0.26

The Twin Arginine Targeting (Tat) Family 0 2 -0.40 2.93 The Telurite-resistance/Dicarboxylate Transporter (TDT) Family 1 1 0.67 0.52

The K+ Transporter (Trk) Family 1 0 0.80 -0.63 The Zinc (Zn2+)-Iron (Fe2+) Permease (ZIP) Family 2 3 -0.66 -0.28

41

Additional data table S4.

Genes in G. sulphuraria probably originating from horizontal gene transfer (HGT) and possible ‘donors’ from which they descended. From left to right: gene identifiers; protein family (ranked by size); number of genes/paralogs resulting from the gene transfer (a number in brackets indicates that not all genes in this family result from HGT and gives the total number of family members); annotation of the encoded protein; for those cases where a detailed phylogenetic analysis was performed, the figure displaying the phylogenetic tree; organism that is the closest sequenced relative (or descendant) of the ‘donor’ organism from which the gene might have originated, according to best BLAST hits or phylogenetic tree; typical habitat of this ‘donor’ organism, with ‘M’ standing for mesophilic (optimum growth temperature between 15°C and 45°C), ‘T’ for thermophilic (OGT 45-80°C), ‘HT’ for hyperthermophilic (OGT > 80°C), and ‘TA’ for thermoacidophilic; systematic position of this ‘donor’ organism with domain, phylum, and order.

42

References and Notes 1. W. N. Doemel, T. D. Brock, The physiological ecology of Cyanidium caldarium. J. Gen.

Microbiol. 67, 17 (1971). doi:10.1099/00221287-67-1-17

2. L. J. Rothschild, R. L. Mancinelli, Life in extreme environments. Nature 409, 1092 (2001). doi:10.1038/35059215 Medline

3. W. Gross, C. Schnarrenberger, Heterotrophic growth of two strains of the acido-thermophilic red alga Galdieria sulphuraria. Plant Cell Physiol. 36, 633 (1995).

4. V. Reeb, D. Bhattacharya, in The thermo-acidophilic Cyanidiophyceae (Cyanidiales) in Red Algae in the Genomic Age, J. Seckbach, D. J. Chapman, Eds. (Springer, Netherlands, 2010), pp. 409–426.

5. G. Pinto, C. Ciniglia, C. Cascone, A. Pollio, Species composition of Cyanidiales assemblages in Pisciarelli (Campi Flegrei, Italy) and description of Galdieria phlegrea SP. NOV in Algae and Cyanobacteria in Extreme Environments, J. Seckbach, Ed. (Springer Netherlands, 2007), vol. 11, pp. 487-502.

6. Materials and methods are available as supplementary materials on Science Online.

7. M. Matsuzaki et al., Genome sequence of the ultrasmall unicellular red alga Cyanidioschyzon merolae 10D. Nature 428, 653 (2004). doi:10.1038/nature02398 Medline

8. H. Innan, F. Kondrashov, The evolution of gene duplications: Classifying and distinguishing between models. Nat. Rev. Genet. 11, 97 (2010). doi:10.1038/nrg2689 Medline

9. T. J. Treangen, E. P. C. Rocha, Horizontal transfer, not duplication, drives the expansion of protein families in prokaryotes. PLoS Genet. 7, e1001284 (2011). doi:10.1371/journal.pgen.1001284 Medline

10. C. Bowler et al., The Phaeodactylum genome reveals the evolutionary history of diatom genomes. Nature 456, 239 (2008). doi:10.1038/nature07410 Medline

11. P. J. Keeling, Functional and ecological impacts of horizontal gene transfer in eukaryotes. Curr. Opin. Genet. Dev. 19, 613 (2009). doi:10.1016/j.gde.2009.10.001 Medline

12. R. Overbeek et al., The subsystems approach to genome annotation and its use in the project to annotate 1000 genomes. Nucleic Acids Res. 33, 5691 (2005). doi:10.1093/nar/gki866 Medline

13. E. V. Koonin, Evidence for a family of archaeal ATPases. Science 275, 1489 (1997). doi:10.1126/science.275.5305.1489 Medline

14. J. G. McCoy et al., Discovery of sarcosine dimethylglycine methyltransferase from Galdieria sulphuraria. Proteins Struct. Funct. Bioinformatics 74, 368 (2009). doi:10.1002/prot.22147 Medline

15. I. Enami, H. Akutsu, Y. Kyogoku, Intracellular pH regulation in an acidophilic unicellular alga, Cyanidium caldarium: 31P-NMR determination of intracellular pH. Plant Cell Physiol. 27, 1351 (1986).

16. W. Gross, Ecophysiology of algae living in highly acidic environments. Hydrobiologia 433, 31 (2000). doi:10.1023/A:1004054317446

http://dx.doi.org/10.1099/00221287-67-1-17

http://dx.doi.org/10.1038/35059215

http://dx.doi.org/10.1038/35059215

http://dx.doi.org/10.1038/nature02398


http://dx.doi.org/10.1038/nrg2689


http://dx.doi.org/10.1371/journal.pgen.1001284

http://dx.doi.org/10.1371/journal.pgen.1001284



http://dx.doi.org/10.1016/j.gde.2009.10.001

http://dx.doi.org/10.1016/j.gde.2009.10.001

http://dx.doi.org/10.1093/nar/gki866


http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=16214803&dopt=Abstract

http://dx.doi.org/10.1126/science.275.5305.1489

http://dx.doi.org/10.1126/science.275.5305.1489

http://dx.doi.org/10.1002/prot.22147

http://dx.doi.org/10.1002/prot.22147

http://dx.doi.org/10.1023/A:1004054317446

43

17. J. Ye, C. Rensing, B. P. Rosen, Y.-G. Zhu, Arsenic biomethylation by photosynthetic organisms. Trends Plant Sci. 17, 155 (2012). doi:10.1016/j.tplants.2011.12.003 Medline

18. C. Oesterhelt, S. Vogelbein, R. P. Shrestha, M. Stanke, A. P. M. Weber, The genome of the thermoacidophilic red microalga Galdieria sulphuraria encodes a small family of secreted class III peroxidases that might be involved in cell wall modification. Planta 227, 353 (2008). doi:10.1007/s00425-007-0622-z Medline

19. C. Pál, B. Papp, M. J. Lercher, Adaptive evolution of bacterial metabolic networks by horizontal gene transfer. Nat. Genet. 37, 1372 (2005). doi:10.1038/ng1686 Medline

20. D. D. Leipe, E. V. Koonin, L. Aravind, STAND, a class of P-loop NTPases including animal and plant regulators of programmed cell death: Multiple, complex domain architectures, unusual phyletic patterns, and evolution by horizontal gene transfer. J. Mol. Biol. 343, 1 (2004). doi:10.1016/j.jmb.2004.08.023 Medline

21. W. L. Chiu et al., Oenothera chloroplast DNA polymorphisms associated with plastome mutator activity. Mol. Gen. Genet. 221, 59 (1990). doi:10.1007/BF00280368

22. S. Batzoglou et al., ARACHNE: A whole-genome shotgun assembler. Genome Res. 12, 177 (2002). doi:10.1101/gr.208902 Medline

23. A. P. M. Weber, K. L. Weber, K. Carr, C. Wilkerson, J. B. Ohlrogge, Sampling the Arabidopsis transcriptome with massively parallel pyrosequencing. Plant Physiol. 144, 32 (2007). doi:10.1104/pp.107.096677 Medline

24. M. Stanke, O. Schöffmann, B. Morgenstern, S. Waack, Gene prediction in eukaryotes with a generalized hidden Markov model that uses hints from external sources. BMC Bioinformatics 7, 62 (2006). doi:10.1186/1471-2105-7-62 Medline

25. A. P. M. Weber et al., EST-analysis of the thermo-acidophilic red microalga Galdieria sulphuraria reveals potential for lipid A biosynthesis and unveils the pathway of carbon export from rhodoplasts. Plant Mol. Biol. 55, 17 (2004). doi:10.1007/s11103-004-0376-y Medline

26. K. D. Pruitt, T. Tatusova, D. R. Maglott, NCBI reference sequences (RefSeq): A curated non-redundant sequence database of genomes, transcripts and proteins. Nucleic Acids Res. 35, (Database issue), D61 (2007). doi:10.1093/nar/gkl842 Medline

27. A. Marchler-Bauer et al., CDD: A Conserved Domain Database for the functional annotation of proteins. Nucleic Acids Res. 39 (Database issue), D225 (2011). doi:10.1093/nar/gkq1189 Medline

28. The Arabidopsis Genome Initiative, Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature 408, 796 (2000). doi:10.1038/35048692 Medline

29. M. Kanehisa, S. Goto, M. Furumichi, M. Tanabe, M. Hirakawa, KEGG for representation and analysis of molecular networks involving diseases and drugs. Nucleic Acids Res. 38 (Database issue), D355 (2010). doi:10.1093/nar/gkp896 Medline

30. UniProt Consortium, The Universal Protein Resource (UniProt) in 2010. Nucleic Acids Res. 38 (Database issue), D142 (2010). doi:10.1093/nar/gkp846 Medline

http://dx.doi.org/10.1016/j.tplants.2011.12.003

http://dx.doi.org/10.1016/j.tplants.2011.12.003

http://dx.doi.org/10.1007/s00425-007-0622-z

http://dx.doi.org/10.1007/s00425-007-0622-z

http://dx.doi.org/10.1038/ng1686

http://dx.doi.org/10.1038/ng1686

http://dx.doi.org/10.1016/j.jmb.2004.08.023

http://dx.doi.org/10.1016/j.jmb.2004.08.023

http://dx.doi.org/10.1007/BF00280368

http://dx.doi.org/10.1101/gr.208902

http://dx.doi.org/10.1101/gr.208902

http://dx.doi.org/10.1104/pp.107.096677

http://dx.doi.org/10.1104/pp.107.096677

http://dx.doi.org/10.1186/1471-2105-7-62

http://dx.doi.org/10.1186/1471-2105-7-62

http://dx.doi.org/10.1007/s11103-004-0376-y

http://dx.doi.org/10.1007/s11103-004-0376-y


http://dx.doi.org/10.1093/nar/gkl842


http://dx.doi.org/10.1093/nar/gkq1189

http://dx.doi.org/10.1093/nar/gkq1189

http://dx.doi.org/10.1038/35048692

http://dx.doi.org/10.1038/35048692

http://dx.doi.org/10.1093/nar/gkp896




44

31. M. Ashburner et al., Gene ontology: Tool for the unification of biology. Nat. Genet. 25, 25 (2000). doi:10.1038/75556 Medline

32. http://mapman.gabipd.org/web/guest/app/mercator

33. S. Hunter et al., InterPro: The integrative protein signature database. Nucleic Acids Res. 37 (Database issue), D211 (2009). doi:10.1093/nar/gkn785 Medline

34. J. Söding, A. Biegert, A. N. Lupas, The HHpred interactive server for protein homology detection and structure prediction. Nucleic Acids Res. 33, (Web Server issue), W244 (2005). doi:10.1093/nar/gki408 Medline

35. S. F. Altschul et al., Gapped BLAST and PSI-BLAST: A new generation of protein database search programs. Nucleic Acids Res. 25, 3389 (1997). doi:10.1093/nar/25.17.3389 Medline

36. Q. Ren, K. Chen, I. T. Paulsen, TransportDB: A comprehensive database resource for cytoplasmic membrane transport systems and outer membrane channels. Nucleic Acids Res. 35 (Database issue), D274 (2007). doi:10.1093/nar/gkl925 Medline

37. H. Li, V. A. Benedito, M. K. Udvardi, P. X. Zhao, TransportTP: A two-phase classification approach for membrane transporter prediction and characterization. BMC Bioinformatics 10, 418 (2009). doi:10.1186/1471-2105-10-418 Medline

38. M. H. Saier, Jr., M. R. Yen, K. Noto, D. G. Tamang, C. Elkan, The Transporter Classification Database: Recent advances. Nucleic Acids Res. 37 (Database issue), D274 (2009). doi:10.1093/nar/gkn862 Medline

39. T. N. Petersen, S. Brunak, G. von Heijne, H. Nielsen, SignalP 4.0: Discriminating signal peptides from transmembrane regions. Nat. Methods 8, 785 (2011). doi:10.1038/nmeth.1701 Medline

40. B. R. Zeeberg et al., High-Throughput GoMiner, an ‘industrial-strength’ integrative gene ontology tool for interpretation of multiple-microarray experiments, with application to studies of Common Variable Immune Deficiency (CVID). BMC Bioinformatics 6, 168 (2005). doi:10.1186/1471-2105-6-168 Medline

41. M. Remm, C. E. V. Storm, E. L. L. Sonnhammer, Automatic clustering of orthologs and in-paralogs from pairwise species comparisons. J. Mol. Biol. 314, 1041 (2001). doi:10.1006/jmbi.2000.5197 Medline

42. S. S. Merchant et al., The Chlamydomonas genome reveals the evolution of key animal and plant functions. Science 318, 245 (2007). doi:10.1126/science.1143609 Medline

43. R. C. Edgar, MUSCLE: A multiple sequence alignment method with reduced time and space complexity. BMC Bioinformatics 5, 113 (2004). doi:10.1186/1471-2105-5-113 Medline

44. G. Talavera, J. Castresana, Improvement of phylogenies after removing divergent and ambiguously aligned blocks from protein sequence alignments. Syst. Biol. 56, 564 (2007). doi:10.1080/10635150701472164 Medline

45. F. Abascal, R. Zardoya, D. Posada, ProtTest: Selection of best-fit models of protein evolution. Bioinformatics 21, 2104 (2005). doi:10.1093/bioinformatics/bti263 Medline

http://dx.doi.org/10.1038/75556

http://dx.doi.org/10.1038/75556

http://dx.doi.org/10.1093/nar/gkn785




http://dx.doi.org/10.1093/nar/25.17.3389

http://dx.doi.org/10.1093/nar/25.17.3389




http://dx.doi.org/10.1186/1471-2105-10-418

http://dx.doi.org/10.1186/1471-2105-10-418



http://dx.doi.org/10.1038/nmeth.1701

http://dx.doi.org/10.1038/nmeth.1701

http://dx.doi.org/10.1186/1471-2105-6-168

http://dx.doi.org/10.1186/1471-2105-6-168

http://dx.doi.org/10.1006/jmbi.2000.5197

http://dx.doi.org/10.1006/jmbi.2000.5197

http://dx.doi.org/10.1126/science.1143609


http://dx.doi.org/10.1186/1471-2105-5-113

http://dx.doi.org/10.1186/1471-2105-5-113

http://dx.doi.org/10.1080/10635150701472164

http://dx.doi.org/10.1080/10635150701472164

http://dx.doi.org/10.1093/bioinformatics/bti263

http://dx.doi.org/10.1093/bioinformatics/bti263

45

46. S. Guindon et al., New algorithms and methods to estimate maximum-likelihood phylogenies: Assessing the performance of PhyML 3.0. Syst. Biol. 59, 307 (2010). doi:10.1093/sysbio/syq010 Medline

47. Z. Yang, PAML 4: Phylogenetic analysis by maximum likelihood. Mol. Biol. Evol. 24, 1586 (2007). doi:10.1093/molbev/msm088 Medline

48. W.-H. Chen, M. J. Lercher, ColorTree: A batch customization tool for phylogenic trees. BMC Res. Notes 2, 155 (2009). doi:10.1186/1756-0500-2-155 Medline

49. A. Agresti, A survey of exact inference for contingency tables. Stat. Sci. 7, 131 (1992). doi:10.1214/ss/1177011454

50. K. Tamura, J. Dudley, M. Nei, S. Kumar, MEGA4: Molecular Evolutionary Genetics Analysis (MEGA) software version 4.0. Mol. Biol. Evol. 24, 1596 (2007). doi:10.1093/molbev/msm092 Medline

51. J.-F. Taly et al., Using the T-Coffee package to build multiple sequence alignments of protein, RNA, DNA sequences and 3D structures. Nat. Protoc. 6, 1669 (2011). doi:10.1038/nprot.2011.393 Medline

52. G. Altekar, S. Dwarkadas, J. P. Huelsenbeck, F. Ronquist, Parallel Metropolis coupled Markov chain Monte Carlo for Bayesian phylogenetic inference. Bioinformatics 20, 407 (2004). doi:10.1093/bioinformatics/btg427 Medline

53. D. H. Huson et al., Dendroscope: An interactive viewer for large phylogenetic trees. BMC Bioinformatics 8, 460 (2007). doi:10.1186/1471-2105-8-460 Medline

54. H. Zhang, S. Gao, M. J. Lercher, S. Hu, W.-H. Chen, EvolView, an online tool for visualizing, annotating and managing phylogenetic trees. Nucleic Acids Res. 40, (Web Server issue), W569 (2012). doi:10.1093/nar/gks576 Medline

55. J. E. Blair, Animals (Metazoa) in The Timetree of Life, S. B. Hedges, S. Kumar, Eds. (Oxford Univ. Press, New York, 2009), pp. 223-230.

56. C. Ciniglia, H. S. Yoon, A. Pollio, G. Pinto, D. Bhattacharya, Hidden biodiversity of the extremophilic Cyanidiales red algae. Mol. Ecol. 13, 1827 (2004). doi:10.1111/j.1365-294X.2004.02180.x Medline

57. H. S. Yoon, J. D. Hackett, C. Ciniglia, G. Pinto, D. Bhattacharya, A molecular timeline for the origin of photosynthetic eukaryotes. Mol. Biol. Evol. 21, 809 (2004). doi:10.1093/molbev/msh075 Medline

58. D. Bhattacharya, H. S. Yoon, S. B. Hedges, J. D. Hackett, in Eukaryotes (Eukaryota) in The Timetree of Life, S. B. Hedges, S. Kumar, Eds. (Oxford Univ. Press, New York, 2009), pp. 116–120.

59. P. J. Keeling, J. D. Palmer, Horizontal gene transfer in eukaryotic evolution. Nat. Rev. Genet. 9, 605 (2008). doi:10.1038/nrg2386 Medline

60. J. Huang, J. P. Gogarten, Concerted gene recruitment in early plant evolution. Genome Biol. 9, R109 (2008). doi:10.1186/gb-2008-9-7-r109 Medline

http://dx.doi.org/10.1093/sysbio/syq010

http://dx.doi.org/10.1093/sysbio/syq010

http://dx.doi.org/10.1093/molbev/msm088


http://dx.doi.org/10.1186/1756-0500-2-155

http://dx.doi.org/10.1186/1756-0500-2-155

http://dx.doi.org/10.1214/ss/1177011454



http://dx.doi.org/10.1038/nprot.2011.393


http://dx.doi.org/10.1093/bioinformatics/btg427

http://dx.doi.org/10.1093/bioinformatics/btg427

http://dx.doi.org/10.1186/1471-2105-8-460

http://dx.doi.org/10.1186/1471-2105-8-460

http://dx.doi.org/10.1093/nar/gks576

http://dx.doi.org/10.1093/nar/gks576

http://dx.doi.org/10.1111/j.1365-294X.2004.02180.x

http://dx.doi.org/10.1111/j.1365-294X.2004.02180.x


http://dx.doi.org/10.1093/molbev/msh075

http://dx.doi.org/10.1093/molbev/msh075



http://dx.doi.org/10.1186/gb-2008-9-7-r109

http://dx.doi.org/10.1186/gb-2008-9-7-r109

46

61. J. N. Timmis, M. A. Ayliffe, C. Y. Huang, W. Martin, Endosymbiotic gene transfer: Organelle genomes forge eukaryotic chromosomes. Nat. Rev. Genet. 5, 123 (2004). doi:10.1038/nrg1271 Medline

62. A. Criscuolo, S. Gribaldo, Large-scale phylogenomic analyses indicate a deep origin of primary plastids within cyanobacteria. Mol. Biol. Evol. 28, 3019 (2011). doi:10.1093/molbev/msr108 Medline

63. S. G. E. Andersson et al., The genome sequence of Rickettsia prowazekii and the origin of mitochondria. Nature 396, 133 (1998). doi:10.1038/24094 Medline

64. J. G. Lawrence, H. Ochman, Molecular archaeology of the Escherichia coli genome. Proc. Natl. Acad. Sci. U.S.A. 95, 9413 (1998). doi:10.1073/pnas.95.16.9413 Medline

65. J. Bohlin, E. Skjerve, D. W. Ussery, Reliability and applications of statistical methods based on oligonucleotide frequencies in bacterial and archaeal genomes. BMC Genomics 9, 104 (2008). doi:10.1186/1471-2164-9-104 Medline

66. C. S. Smillie et al., Ecology drives a global network of gene exchange connecting the human microbiome. Nature 480, 241 (2011). doi:10.1038/nature10571 Medline

67. R. W. Bailey, L. A. Staehelin, The chemical composition of isolated cell walls of Cyanidium caldarium. J. Gen. Microbiol. 54, 269 (1968). doi:10.1099/00221287-54-2-269 Medline

68. R. Asada, K. Tazaki, Silica biomineralization of unicellular microbes under strongly acidic conditions. Can. Mineral. 39, 1 (2001). doi:10.2113/gscanmin.39.1.1

69. R. P. Shrestha, A. P. Weber, Acidothermophilic red microalga Galdieria sulphuraria: From genome to an extracellular glucoamylase active at extreme low pH and high temperature. J. Phycol. 43 (suppl. 1), 23 (2007).

70. O. Emanuelsson, S. Brunak, G. von Heijne, H. Nielsen, Locating proteins in the cell using TargetP, SignalP and related tools. Nat. Protoc. 2, 953 (2007). doi:10.1038/nprot.2007.131 Medline

71. E. Derelle et al., Genome analysis of the smallest free-living eukaryote Ostreococcus tauri unveils many unique features. Proc. Natl. Acad. Sci. U.S.A. 103, 11647 (2006). doi:10.1073/pnas.0604795103 Medline

72. B. Palenik et al., The tiny eukaryote Ostreococcus provides genomic insights into the paradox of plankton speciation. Proc. Natl. Acad. Sci. U.S.A. 104, 7705 (2007). doi:10.1073/pnas.0611046104 Medline

73. E. V. Armbrust et al., The genome of the diatom Thalassiosira pseudonana: Ecology, evolution, and metabolism. Science 306, 79 (2004). doi:10.1126/science.1101156 Medline

74. http://genome.jgi-psf.org/Auran1/Auran1.home.html.

75. S. Guindon, O. Gascuel, A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Syst. Biol. 52, 696 (2003). doi:10.1080/10635150390235520 Medline

76. R. D. Finn et al., The Pfam protein families database. Nucleic Acids Res. 38 (Database issue), D211 (2010). doi:10.1093/nar/gkp985 Medline



http://dx.doi.org/10.1093/molbev/msr108

http://dx.doi.org/10.1093/molbev/msr108

http://dx.doi.org/10.1038/24094

http://dx.doi.org/10.1038/24094

http://dx.doi.org/10.1073/pnas.95.16.9413

http://dx.doi.org/10.1073/pnas.95.16.9413

http://dx.doi.org/10.1186/1471-2164-9-104

http://dx.doi.org/10.1186/1471-2164-9-104



http://dx.doi.org/10.1099/00221287-54-2-269

http://dx.doi.org/10.1099/00221287-54-2-269

http://dx.doi.org/10.2113/gscanmin.39.1.1



http://dx.doi.org/10.1073/pnas.0604795103







http://dx.doi.org/10.1080/10635150390235520

http://dx.doi.org/10.1080/10635150390235520




47

77. M. H. Saier, Jr., C. V. Tran, R. D. Barabote, TCDB: The Transporter Classification Database for membrane transport protein analyses and information. Nucleic Acids Res. 34 (Database issue), D181 (2006). doi:10.1093/nar/gkj001 Medline

78. A. Marchler-Bauer et al., CDD: A conserved domain database for interactive domain family analysis. Nucleic Acids Res. 35 (Database issue), D237 (2007). doi:10.1093/nar/gkl951 Medline

79. D. Gagneul et al., A reassessment of the function of the so-called compatible solutes in the halophytic plumbaginaceae Limonium latifolium. Plant Physiol. 144, 1598 (2007). doi:10.1104/pp.107.099820 Medline

http://dx.doi.org/10.1093/nar/gkj001

http://dx.doi.org/10.1093/nar/gkj001




http://dx.doi.org/10.1104/pp.107.099820

http://dx.doi.org/10.1104/pp.107.099820

Supplementary Materials for -...

Documents

Transcript of Supplementary Materials for -...