Evolution of the Isd11-IscS complex reveals a single alpha-proteobacterial endosymbiosis for all eukaryotes.

Abstract

Giardia and Trichomonas are eukaryotes without standard mitochondria but contain mitochondrial-type alpha-proteobacterium-derived iron-sulfur cluster (ISC) assembly proteins, located to mitosomes in Giardia and hydrogenosomes in Trichomonas. Although these data suggest a single common endosymbiotic ancestry for mitochondria, mitosomes, and hydrogenosomes, separate origins are still being proposed. Here, we present a bioinformatic analysis of Isd11, a recently described essential component of the mitochondrial ISC assembly pathway. Isd11 is unique to eukaryotes but functions closely with the alpha-proteobacterium-derived cysteine desulfurase IscS. We demonstrate the presence of homologues of Isd11 in all 5 eukaryotic supergroups sampled, including hydrogenosomal and mitosomal lineages. The eukaryotic invention of Isd11 as a functional partner to IscS directly implies a single shared alpha-proteobacterial endosymbiotic ancestry for all eukaryotes. This pinpoints the alpha-proteobacterial endosymbiosis to before the last common ancestor of all eukaryotes without ambiguity.

Full text

Page 1

Evolution of the Isd11–IscS Complex Reveals a Single a-Proteobacterial
Endosymbiosis for All Eukaryotes
Thomas A. Richards* and Mark van der Giezen?
*School of Biosciences, University of Exeter, Exeter, United Kingdom; and ?School of Biological and Chemical Sciences,
Queen Mary, University of London, London, United Kingdom
GiardiaandTrichomonasareeukaryoteswithoutstandardmitochondriabutcontainmitochondrial-typea-proteobacterium–derived
iron–sulfur cluster (ISC) assembly proteins, located to mitosomes in Giardia and hydrogenosomes in Trichomonas.
Althoughthesedatasuggestasinglecommonendosymbioticancestryformitochondria,mitosomes,andhydrogenosomes,
separate origins are still being proposed. Here, we present a bioinformatic analysis of Isd11, a recently described essential
component of the mitochondrial ISC assembly pathway. Isd11 is unique to eukaryotes but functions closely with
the a-proteobacterium–derived cysteine desulfurase IscS. We demonstrate the presence of homologues of Isd11 in all
5 eukaryotic supergroups sampled, including hydrogenosomal and mitosomal lineages. The eukaryotic invention of
Isd11 as a functional partner to IscS directly implies a single shared a-proteobacterial endosymbiotic ancestry for all
eukaryotes. This pinpoints the a-proteobacterial endosymbiosis to before the last common ancestor of all eukaryotes
without ambiguity.
Endosymbiosis played a key role in the evolution of
eukaryotic cells, but the number and ancestry of endosym-
biotic events remains contentious (Martin and Mu ¨ller 1998;
Martinetal.2001;Dyalletal.2004).Didorganellessuchas
hydrogenosomes, mitosomes, and mitochondria originate
separately from different symbiotic bacteria or from just
one endosymbiosis (Dyall et al. 2004; Embley and Martin
2006)?
In many species, mitochondria are essential for energy
transduction by oxidative phosphorylation. However, sys-
tematic knock-out of mitochondrial proteins has demon-
strated that yeast mitochondria perform only one essential
function, the assembly of iron–sulfur clusters (ISCs) (Lill
and Mu ¨hlenhoff 2005). This suggests that mitochondria
are retained for compartmentalized ISC assembly (Lill
and Mu ¨hlenhoff 2005; van der Giezen and Tovar 2005).
Wiedemann et al. (2006) and Adam et al. (2006)
recently demonstrated that the ISC protein Isd11 is local-
izedwithinmitochondria andisessentialforISCbiogenesis
in yeast and that Isd11 forms a complex with Nfs1. Nfs1
and its orthologue IscS play a role as a vital cysteine desul-
furase of the ISC assembly machinery (Adam et al. 2006;
Wiedemann et al. 2006). Isd11 is suggested to function
as an adapter and stabilizer of Nfs1 (Adam et al. 2006;
Wiedemann et al. 2006). Homologues of the Isd11 gene
have been detected in plant, fungi, and animal genomes,
which possess mitochondria, but no prokaryote homologue
has been found (Wiedemann et al. 2006). This suggests that
Isd11 is a eukaryotic molecular innovation that arose, at the
earliest, during the primary process of endosymbiosis that
gave rise to mitochondria and which installed IscS, the
functional partner of Isd11 (Adam et al. 2006; Wiedemann
et al. 2006), in the eukaryotes. To test this hypothesis, we
searched for putative homologues from available prokary-
otic genomes using multiple PSI-Blast searches of Gen-
Bank and confirmed, given current sampling, that the
Isd11 gene is exclusive to eukaryotes. Blast searches of
all available eukaryotic genomes revealed the presence
of putative IscS homologues in all eukaryotes surveyed, ex-
cept for incomplete genome projects (fig. 1A). Further,
comparative genome analyses detected putative homo-
logues of Isd11 in numerous eukaryotes including hydro-
genosomal and mitosomal lineages. BlastP analyses
demonstrated 47% amino acid sequence identity with the
yeast Isd11 protein for both the Trichomonas vaginalis
and Nosema locustae (synonym Antonospora locustae) pu-
tative Isd11 proteins. Together with the protein alignment,
this suggests that the amitochondrial sequences are true ho-
mologues (fig. 1B). Isd11 has a high helical propensity (see
fig. 1B). Interestingly, neither PSSM (Kelley et al. 2000)
nor the newer PHYRE algorithm (http://www.sbg.bio.ic.
ac.uk/phyre/) recognize any known fold in Isd11, suggest-
ing that Isd11 might represent a new type of protein fold.
Further structural studies, especially in complex with Nfs1,
shouldproveinteresting.AlltheputativeIsd11homologues
analyzed are predicted to contain 3 a-helices, consistent
with their shared homology. Bioinformatic predictions of
mitochondrial targeting for Isd11 were unconvincing for all
lineages investigated, including the yeast Isd11, known to
be located in mitochondria (Adam et al. 2006; Wiedemann
et al. 2006) but demonstrated a candidate target peptide
intheputativeIsd11ofT.vaginalis(fig.1Band Supplemen-
tary Table 1, Supplementary Material online).
The process of horizontal gene transfers (HGTs) could
theoretically distribute genes of endosymbiotic ancestry in-
to lineages that did not undergo the original endosymbiosis
(Roger et al. 1998). The Isd11 gene is short (encoding ;90
amino acids) and unlikely to be a reliable gene for phylog-
eny; indeed, like many single-gene eukaryote phylogenies,
the terminal branches were unresolved (fig. 2A—see
branches within the gray zone). Comparison tests of alter-
nativeterminalbranchingordersconfirmedthatmanyofthe
pictured terminal node relationships are unresolved (see
Supplementary Methods and Supplementary Figure 1, Sup-
plementary Material online). However, phylogenetic anal-
ysis did not provide strong support for HGT over the more
parsimonious scenario of vertical inheritance for the T. vag-
inalis Isd11 gene (fig. 2A). Furthermore, Nosema Isd11
consistently grouped within a weakly resolved clade of al-
veolates and formed a moderately supported sister group
relationship with the Plasmodium parasites in all analyses.
Key words: mitochondria, mitosome, hydrogenosome, iron–sulfur
cluster, origin of the eukaryotic cell.
E-mail: m.vandergiezen@qmul.ac.uk.
Mol. Biol. Evol. 23(7):1341–1344. 2006
doi:10.1093/molbev/msl001
Advance Access publication April 28, 2006
? The Author 2006. Published by Oxford University Press on behalf of
the Society for Molecular Biology and Evolution. All rights reserved.
For permissions, please e-mail: journals.permissions@oxfordjournals.org

Page 2

Plasmodium and Microsporidia are intracellular parasites
that infect insects, and in both Plasmodium and Microspor-
idia, there are reported cases of HGT between these para-
sites and cells that they come into close contact with
(Richards et al. 2003). Comparative topology tests could
not reject an alternative topology where the Microsporidia
branched below the alveolate group (e.g., ‘‘approximately
unbiased’’ test, P5 0.49;Shimodaira andHasegawa2001),
which is inconsistent with a Plasmodium-to-Microsporidia
gene transfer. However, taken together, we cannot exclude
a case of HGT between the Apicomplexa (Plasmodium)
andMicrosporidia(Nosema)basedontheIsd11phylogeny,
but this scenario is less parsimonious than vertical inheri-
tance.
FIG. 1.—(A) Eukaryotic genome survey for putative homologues of IscS and Ids11. Yellow indicates putative homologue present. Black indicates
absence of similar sequences. Red indicates evidence of HGT (van der Giezen et al. 2004) possibly explaining absence of the Isd11–IscS complex by
functional replacement with a eubacterial IscS. Triangles indicate amitochondrial lineages. (B) Alignment of a taxonomically diverse representation of
putative Isd11 proteins. Residues with 50% conservation are shaded black. The LYR/K residue block (Pfam PF05347) conserved in Isd11 and B14 and
B22 components of mitochondrial complex I are illustrated, note that all sampled Excavata possess a tyrosine deletion and that sequence similarity
between the Isd11 and B14 and B22 is negligible beyond the LYR/K block. Circles below the alignment represent positions conserved in 50% or more
ofthesequencesandpresentinTrichomonas(green)orNosema(orange).Thepurplepeptidemodelshowsthepredictedconsensusstructuralarrangement
of a-helices. Purple alignment shading indicates conformity with predicted helical regions. Predicted amphipathic a-helix of the putative Trichomonas
hydrogenosomal targeting sequence is shaded purple with a red border (see Supplementary Table 1 for details about putative N-terminal target peptide
detection, Supplementary Material online).
1342Richards and van der Giezen

Page 3

In conclusion, we suggest that Isd11 originated during
or shortly after the single endosymbiosis that gave rise to
mitochondria, hydrogenosomes, and mitosomes. Isd11
evolved as an exclusively eukaryotic addition to the a-
proteobacterium–derived ISC biogenesis pathway. Like
the ISC biogenesis pathway itself (Tachezy et al. 2001),
Isd11 has been conserved in hydrogenosomal and mitoso-
mallineages.WedemonstratethatIsd11representsaunique
shared derived character of all sampled eukaryote super-
groups(fig.2B),includingmitochondrial,hydrogenosomal,
and mitosomal lineages. Isd11 therefore pinpoints the an-
cestry of the eukaryotic ISC biogenesis pathway to a single
endosymbiotic event. The alternative assumption of sepa-
rate origins of the IscS–Isd11 complex requires Isd11 to
have evolved convergently in separate lineages and IscS
to be acquired separately from 2 closely related proteobac-
teria, whereas separate endosymbiotic origins for any com-
bination of the 3 organelles must include a process of
endosymbiotic replacement (Dyall et al. 2004); these alter-
native scenarios may be possible but are very unparsimo-
nious. Similarities in mitochondrial, hydrogenosomal, and
mitosomal N-terminal targeting peptides also hint at
a shared derived import machinery in all 3 organelle types,
suggesting a common origin of these organelles (van der
Giezen and Tovar 2005). However, although genes encod-
ing components of a mitochondrial import machinery have
been identified in mitosomal lineages (Henriquez et al.
2005), only one putative component of the mitochondrial
import machinery has been identified in hydrogenosomes
(Dolezal et al. 2005). In conclusion, the IscS–Isd11 com-
plex was present in the last common ancestor of all the eu-
karyotes prior to the division of all eukaryotic supergroups
FIG. 2.—(A) Phylogeny of Isd11 calculated from an alignment of 23 taxa and 58 conserved amino acid alignment characters. Topology is shown
unrooted. Topology support values are shown when PHYML bootstrap values are in excess of 50%. Support values are shown in the order 1) Bayesian
posterior probability, 2)percentagebootstrapsupport from 1000PHYMLbootstrap replicates, and 3)percentagebootstrap supportfrom 1000 PROTPARS
bootstrapreplicates.Eukaryotesupergroupclassificationsarenotatednexttospeciesnames.Analternative branchingorder,demonstratingalveolatemono-
phyly and tested using CONSEL (see Supplementary Methods, Supplementary Material online), is illustrated with a gray branch. The gray zone indicates
tree nodes unresolved in all analyses (i.e., all topology support values below 0.5 posterior probability and 50% bootstrap support). GenBank accession
numbers for sequences used in phylogeny: At, AAM66015; Os, NP_921405; Cm, CMN136C (http://merolae.biol.s.u-tokyo.ac.jp/); Lm, CAJ06332; Tc,
XP_807608; Tb, XP_823406; Sc, NP_010968; Kl, XP_454147; Um, EAK83440; Dp, EAL31501; Rn, XP_574009; Dr, XP_701063; Dd, XP_635573; Pf,
CAD52245; Pb, XP_680329. All other sequences were predicted open reading frames from genome projects as discussed in the Supplementary Methods
(SupplementaryMaterialonline).(B)PinpointingtheevolutionoftheIsd11–IscScomplexwithintheeukaryotes.Notethatthebikont/unikontrootmodelis
used(StechmannandCavalier-Smith2002;RichardsandCavalier-Smith2005).TheIsd11tyrosinedeletion(fig.1B),strongresultsofmultigenephylogeny
(Rodriguez-Ezpeleta et al. 2005), and morphological characters suggest Excavata monophyly (Cavalier-Smith 2003a; Simpson 2003), whereas bikont
monophyly is supported by the dhfr-ts fusion character (Stechmann and Cavalier-Smith 2002), inferred ancestral morphological characters (Stechmann
andCavalier-Smith 2002; Cavalier-Smith 2003b), andresults of unrooted multigene phylogenies (Rodriguez-Ezpeleta et al. 2005). This data, coupled with
shared derived HGTs present in Trichomonas and Giardia (Andersson et al. 2005), favor the tree topology shown. The alternative Giardia/Trichomonas
root requires fission events in the ancestral unikont dhfr-ts and separate origins or separate evolutionary reductions of highly complex morphological
characters and the presence of such characters in the last common eukaryotic ancestor (Cavalier-Smith 2003a; Simpson 2003). However, this alternative
rootwouldstillpinpointtheIsd11/IscScharactertothebaseoftheeukaryoteradiation,demonstratingthatthea-proteobacterialendosymbiosisgivingriseto
mitochondria, hydrogenosomes, and mitosomes occurred in the last common eukaryotic ancestor.
Isd11 and the Mitochondrial Endosymbiosis 1343

Page 4

(Simpson and Roger 2004). Therefore, the a-proteobacte-
rial endosymbiosis is placed firmly before the radiation of
all eukaryotes (fig. 2B).
Supplementary Material
Additionaldetailed methods (Supplementary Methods),
results of alternative topology comparison tests (Sup-
plementary Figure 1), and comparisons of putative Isd11
mitochondria targeting peptides (Supplementary Table 1)
are available at Molecular Biology and Evolution online
(http://www.mbe.oxfordjournals.org/).
Acknowledgments
We are grateful to The Institute for Genomic Research
(http://www.tigr.org), The Department of Energy (http://
www.jgi.doe.gov/), Genoscope (http://www.genoscope.
cns.fr/), and M. B. L. Woods Hole (http://jbpc.mbl.
edu/Nosema/) for making their genome data available
(01/06). T.A.R. thanks the Department for Environment,
Food and Rural Affairs for fellowship support. We thank
N. J. Talbot and J. F. Allen for comments and P. Foster with
assistance with topology comparison tests.
Literature Cited
Adam AC, Bornho ¨vd C, Prokisch H, Neupert W, Hell K. 2006.
The Nfs1 interacting protein Isd11 has an essential role in Fe/S
cluster biogenesis in mitochondria. EMBO J 25:174–83.
Andersson JO, Sarchfield SW, Roger AJ. 2005. Gene transfers
from nanoarchaeota to an ancestor of diplomonads and para-
basalids. Mol Biol Evol 22:85–90.
Cavalier-Smith T. 2003a. The excavate protozoan phyla Metamo-
nada Grasse ´ emend. (Anaeromonadea, Parabasalia, Carpedie-
monas, Eopharyngia) and Loukozoa emend. (Jakobea,
Malawimonas): their evolutionary affinities and new higher
taxa. Int J Syst Evol Microbiol 53:1741–58.
Cavalier-Smith T. 2003b. Protist phylogeny and the high-level
classification of protozoa. Eur J Protistol 39:338–48.
Dolezal P, Smid O, Rada P, Zubacova Z, Bursac D, Sutak R,
NebesarovaJ,LithgowT,TachezyJ.2005.Giardiamitosomes
and trichomonad hydrogenosomes share a common mode of
protein targeting. Proc Natl Acad Sci USA 102:10924–29.
Dyall SD, Brown MT, Johnson PJ. 2004. Ancient invasions: from
endosymbionts to organelles. Science 304:253–7.
Embley TM, Martin W. 2006. Eukaryotic evolution, changes and
challenges. Nature 440:623–30.
Henriquez FL, Richards TA, Roberts F, McLeod R, Roberts CW.
2005. The unusual mitochondrial compartment of Cryptospo-
ridium parvum. Trends Parasitol 21:68–74.
Kelley LA, MacCallum RM, Sternberg MJ. 2000. Enhanced
genome annotation using structural profiles in the program
3D-PSSM. J Mol Biol 299:499–520.
Lill R, Mu ¨hlenhoff U. 2005. Iron-sulfur-protein biogenesis in
eukaryotes. Trends Biochem Sci 30:133–41.
MartinW,HoffmeisterM,RotteC,HenzeK.2001.Anoverviewof
endosymbiotic models for the origins of eukaryotes, their ATP-
producing organelles (mitochondria and hydrogenosomes), and
their heterotrophic lifestyle. Biol Chem 382:1521–39.
Martin W, Mu ¨ller M. 1998. The hydrogen hypothesis for the first
eukaryote. Nature 392:37–41.
Richards TA, Cavalier-Smith T. 2005. Myosin domain evolu-
tion and the primary divergence of eukaryotes. Nature 436:
1113–8.
Richards TA, Hirt RP, Williams BA, Embley TM. 2003. Horizon-
tal gene transfer and the evolution of parasitic protozoa. Protist
154:17–32.
Rodriguez-Ezpeleta N, Brinkmann H, Burey SC, Roure B, Burger
G, Loffelhardt W, Bohnert HJ, Philippe H, Lang BF. 2005.
Monophyly of primary photosynthetic eukaryotes: green
plants, red algae, and glaucophytes. Curr Biol 15:1325–30.
Roger AJ, Svard SG, Tovar J, Clark CG, Smith MW, Gillin FD,
Sogin ML. 1998. A mitochondrial-like chaperonin 60 gene in
Giardia lamblia: evidence that diplomonads once harbored an
endosymbiont related to the progenitor of mitochondria. Proc
Natl Acad Sci USA 95:229–34.
Shimodaira H, Hasegawa M. 2001. CONSEL: for assessing the
confidence of phylogenetic tree selection. Bioinformatics
17:1246–7.
Simpson AG.2003.Cytoskeletalorganization, phylogenetic affin-
ities and systematics in the contentious taxon Excavata (Eukar-
yota). Int J Syst Evol Microbiol 53:1759–77.
Simpson AG, Roger AJ. 2004. The real ?kingdoms? of eukaryotes.
Curr Biol 14:R693–6.
Stechmann A, Cavalier-Smith T. 2002. Rooting the eukaryote tree
by using a derived gene fusion. Science 297:89–91.
Tachezy J, Sa ´nchez LB, Mu ¨ller M. 2001 Mitochondrial type iron-
sulfur cluster assembly in the amitochondriate eukaryotes Tri-
chomonas vaginalis and Giardia intestinalis, as indicated by
the phylogeny of IscS. Mol Biol Evol 18:1919–28.
van der Giezen M, Cox S, Tovar J. 2004. The iron-sulfur cluster
assembly genes iscS and iscU of Entamoeba histolytica were
acquired by horizontal gene transfer. BMC Evol Biol 4:7.
van der Giezen M, Tovar J. 2005. Degenerate mitochondria.
EMBO Rep 6:525–30.
Wiedemann N, Urzica E, Guiard B et al. (11 co-authors). 2006.
Essential role of Isd11 in mitochondrial iron-sulfur cluster syn-
thesis on Isu scaffold proteins. EMBO J 25:184–95.
William Martin, Associate Editor
Accepted April 19, 2006
1344Richards and van der Giezen

Page 5

1
Supplementary Methods
Alignment and sequence analysis. The yeast Isd11 protein sequence was used as a
seed for BLASTp and tBLASTn searches of all the eukaryote genome projects available
at NCBI genome BLAST server
(http://www.ncbi.nlm.nih.gov/sutils/genom_tree.cgi?organism=euk), additional BLAST
searches were performed on the Phanerochaete, Phytophthora, Chlamydomonas
(http://genome.jgi-psf.org/),
Nosema
(http://jbpc.mbl.edu/Nosema/),
Paramecium
(http://www.genoscope.cns.fr/) and
Cyanidioschyzon
(http://merolae.biol.s.u-
tokyo.ac.jp/) genome projects. PSI-BLAST (3 iterations) and BLASTp analyses were
conducted against GenBank nr database to identify additional eukaryotic and putative
prokaryotic homologues. The BLAST process was repeated using the Trichomonas
putative Isd11 as a seed. A diverse taxonomic selection of putative homologues was
aligned using T-COFFEE (http://www.igs.cnrs-mrs.fr/Tcoffee/tcoffee_cgi/index.cgi)
and then manually refined using the alignment program Se-Al
(http://evolve.zoo.ox.ac.uk/software.html?id=seal). Highly variable positions that could
not be aligned with confidence were removed from the alignment prior to phylogenetic
reconstruction. MITOPROT (Claros and Vincens 1996) and PREDOTAR (Small et al
2004) was used to crudely assess evidence of mitochondrial N-terminal target
sequences. Primary and secondary sequence features were analysed using SMART
(Letunic et al. 2004), 3D-PSSM protein fold recognition software (Kelley, MacCallum,
and Sternberg 2000) and the new protein homology/analogy recognition engine,
PHYRE (http://www.sbg.bio.ic.ac.uk/phyre/) using standard parameters.

Phylogeny. Trees of Isd11 were calculated from an alignment of 23 taxa and 58-
conserved amino acid alignment characters. Prior to tree reconstruction the alignment
was analysed using MODELGENERATOR

Page 6

2
(http://bioinf.nuim.ie/software/modelgenerator) to find the best model for phylogeny
(WAG+Γ, 8 categories). MRBAYES 3.1.2 (Ronquist and Huelsenbeck 2003) was run
with two separate MCMCMC analyses for 500,000 generations at a sampling frequency
of 100 generations, Each MCMCMC run had four MCMC chains (3 heated 1 cold –
“temperature” parameter = 0.2). The covarion option was used. Comparisons of
likelihood scores within and between the two independent runs confirmed that the tree
log-likelihood scores and parameters had reached a plateau and converged by 50,000
generations. Consequently 500 samples were excluded as burnin. However, the terminal
branching order of the two topologies from the two separate Bayesian runs had different
topologies, each differing relationship having very weak posterior probability support in
either result. 1,000 PROTPARS (PHYLIP - Felsenstein 1989) bootstrap replicates were
analysed (random input order – 3x jumbling) and 1,000 PHYML (Guindon et al 2005)
bootstrap replicates were run using the WAG+Γ (8 categories) model. TREE-PUZZLE
(Schmidt et al. 2002) ML-distance analyses (WAG+Γ, 8 category model) demonstrated
that the Plasmodium falciparum and Trypanosoma brucei sequence failed the chi-square
test of amino acid composition. Even though the chi-square test is probably not relevant
here due to the short sequences analysed, we removed these two branches and repeated
the analyses in PHYML. Comparisons of the quartet puzzling values (PUZZLE),
posterior probabilities of the two Bayesian phylogenies (MRBAYES), PROTPARS-
bootstrap values (1,000 replicates) and the two PHYML bootstrap analyses (1,000
replicates each) demonstrate that the terminal relationships are weakly resolved in all
the analyses i.e. below 0.5 posterior probability or 50% bootstrap support value (see
grey zone Figure 2A). Topological comparisons tests showed that the two Bayesian
trees, the PHYML tree (all 23 taxa included) and an adjusted tree topology with
monophyletic alveolates (i.e. no Plasmodium-Microsporidia HGT - see grey branch
Figure 2A) could not be rejected by AU or SH tests (P < 0.05) (Supplementary Figure
1). Prior to topology comparisons branch lengths for the four alternative topologies

Page 7

3
were optimized using maximum likelihood (model = WAG+Γ, 8 categories) using P4
(http://www.nhm.ac.uk/zoology/external/p4.html) and compared using CONSEL
(Shimodaira and Hasegawa 2001). The best scoring phylogeny, Bayesian search 1, (not
significantly better) is shown on Figure 2A.

Literature Cited

Claros, M. G., and P. Vincens. 1996. Computational method to predict mitochondrially
imported proteins and their targeting sequences. Eur. J. Biochem. 241:779-786.
Felsenstein, J. 1989. PHYLIP - Phylogeny Inference Package (Version 3.2). Cladistics
5:164-166.
Guindon, S., F. Lethiec, P. Duroux, and O. Gascuel. 2005. PHYML Online-a web
server for fast maximum likelihood-based phylogenetic inference. Nucleic Acids
Res. 33:W557-559.
Kelley, L. A., R. M. MacCallum, M. J. Sternberg. 2000 Enhanced genome annotation
using structural profiles in the program 3D-PSSM. J. Mol. Biol. 299:499-520.
Letunic, I., R. R. Copley, F. D. Ciccarelli, T. Doerks et al. 2004. SMART 4.0: towards
genomic data integration. Nucleic Acids Res. 32:D142-144.
Ronquist, F. and J, P, Huelsenbeck. 2003. MrBayes 3: Bayesian phylogenetic inference
under mixed models. Bioinformatics 19:1572-1574.
Schmidt, H. A., K. Strimmer, M. Vingron, and A. von Haeseler. 2002. TREE-PUZZLE:
maximum likelihood phylogenetic analysis using quartets and parallel
computing. Bioinformatics 18:502-504.
Shimodaira, H., and M, Hasegawa. 2001. CONSEL: for assessing the confidence of
phylogenetic tree selection. Bioinformatics 17:1246-1247.
Small, I., N. Peeters, F. Legeai, and C. Lurin. 2004. Predotar: A tool for rapidly
screening proteomes for N-terminal targeting sequences. Proteomics 4:1581-
1590.

Page 8

Pc
Um
Sc
Kl
Ro
Nl
Tv
TpDd
Plantae
Metazoa
Kinetoplastids
Plasmodium
Ciliates
Dd
Tp
Kl
Sc
Ro
Tv
Nl
Plantae
Metazoa
Plasmodium
Ciliates
Kinetoplastids
Pc
Um
Dd
Tp
Kl
Sc
Ro
Tv
Nl
Plantae
Metazoa
Plasmodium
Ciliates
Kinetoplastids
Pc
Um
Dd
Tp
Kl
Sc
RoTv
Nl
Metazoa
Plasmodium
Ciliates
Kinetoplastids
Pc
Um
Plantae
Trees -ln[L] ∆ln[L] AU SH Possible?
-2295.069 best 0.633 0.893 Y
-2295.178 0.109 0.617 0.822 Y
-2295.422 0.353 0.467 0.684 Y
-2301.073 6.004 0.115 0.19 Y

Page 9

1
Supplementary Figure 1. Comparisons of alternative topologies using likelihood-
based methods. Results are computed with a WAG+Γ (8 categories) model and listed
for the four topologies tested (in the order MRBAYES result 1 [best tree], PHYML
result [paraphyletic fungi], altered MRBAYES result 1 [monophyletic alveolates] and
MRBAYES result 2). Species initials are given – full names are to be found on Figure
2A. AU, Approximately Unbiased test; SH, Shimodaira–Hasegawa test. All alternative
topologies could not be rejected at the 5% level suggesting that there is no phylogenetic
resolution in this gene amongst these terminal nodes.

Page 10

Supplementary Table 1. Summary of MITOPROT/PREDOTAR analyses for putative
Isd11 proteins of representative taxa (Claros and Vincens 1996; Small et al. 2004). The T.
vaginalis putative Isd11 has a predicted N-terminally cleaved peptide (MLSSFLSRT –
Figure 1B) and a higher score for probability of mitochondrial localization than the yeast
Isd11 according to MITOPROT analyses. Similar bioinformatic predictions of
mitosomal/hydrogenosomal localization for candidate proteins have proved unreliable,
with many proteins demonstrating mitosome localization while not possessing
recognizable N-terminal target sequences (Tovar et al. 2003; Regoes et al. 2005).
However, a comparison of the T. vaginalis putative Isd11 N-terminal sequence with other
hydrogenosomal proteins demonstrates four peptide features consistent with
hydrogenosomal targeting: 1) a short serine-rich N-terminal sequence; 2) a leucine
adjacent to the start-codon; 3) an arginine close to the predicted cleavage point; and 4) a
putative amphipathic α-helix within the predicted N-terminal pre-sequence (Figure 1B).

% positives to
S. cerevisiae ISD11
using BLASTp
MITOPROT
probability of export to
mitochondria
PREDOTAR
probability of export
to mitochondria
S. cerevisiae
- 0.60 0.39
R. norvegicus
65 0.67 0.61
A. thaliana
53 0.44 0.18
D. discoideum
65 0.64 0.45
N. locustae
47 0.45 0.03
T. vaginalis
47 0.69 0.02

Page 11

Literature Cited
Claros, M. G., and P. Vincens. 1996. Computational method to predict
mitochondrially imported proteins and their targeting sequences. Eur. J.
Biochem. 241:779-786.
Regoes, A., D. Zourmpanou, G. León-Avila, M. van der Giezen, J. Tovar, and A.
B. Hehl. 2005. Protein import, replication, and inheritance of a vestigial
mitochondrion. J. Biol. Chem. 280:30557-30563.
Small, I., N. Peeters, F. Legeai, and C. Lurin. (2004) Predotar: A tool for rapidly
screening proteomes for N-terminal targeting sequences. Proteomics, 4:
1581-1590.
Tovar, J., G. León-Avila, L. B. Sánchez, R. Sutak, J. Tachezy, M. van der
Giezen, M. Hernandez, M. Müller, and J. M. Lucocq. 2003. Mitochondrial
remnant organelles of Giardia function in iron-sulphur protein maturation.
Nature 426:172-176.