SUNY at Albany
June 19-23, 2001
Phylogenetic analysis based on Hsp90 heat-shock protein sequences
Ubiquitous, conserved heat-shock proteins (chaperones) such as most well characterized Hsp70 and chaperonin Cpn60 are widely used in evolutionary studies as reliable tracers of endosymbiotic origin of energy-producing organelles. In agreement with other data, phylogenetic analyses based on Cpn60 and Hsp70 protein sequences point to the origin of mitochondria and chloroplasts from a-Proteobacteria and cyanobacteria, respectively. Furthermore, mitochondria were shown by using both Cpn60 [1-4] and Hsp70  to trace their descent to the order Rickettsiales which comprises exceptionally obligate intracellular endosymbiontes of eukaryotes . The genes encoding the above proteins reside in the nuclear genomes with their products being imported into organelles. Besides those encoding organellar chaperones, nuclear genomes contain the genes for nucleocytoplasmic forms. Cytosolic chaperonin called CCT, or Tric, is known to be of archaebacterial origin  while two related Hsp70, cytosolic and endoplasmic reticulum (ER) chaperones, which have arisen via ancient duplication in eukaryotes, are apparently of proteobacterial origin .
Like Hsp70, eukaryotic Hsp90 also exists in two (cytosolic and ER) forms [3, 8]. In addition to Hsp90 of chloroplasts, Hsp90 chaperones were recently reported in mitochondria . In this phylogenetic study, we used the Hsp90 protein sequences available in databases to see whether the genes for organellar proteins have originated from aforementioned eubacterial phyla. Fig. 1 shows a dendrogram constructed by the maximum-likelihood (ML) method of PUZZLE 5.0 . Quite surprisingly, Hsp90 of chloroplasts exhibited an affiliation to ER forms. This relationship was also supported by the presence of two indels, specific insertions or deletions within conserved protein regions , one of which encompasses a charged domain . Another surprise, Hsp90 sequences of human and fruit fly, the only two mitochondrial sequences known to date, appeared to group with their cyanobacterial counterparts. Unfortunately, no htpG gene encoding Hsp90 was reported in a-Proteobacteria except for Rickettsia prowazekii Ð the nearest known relative of mitochondria [4, 6]. It is of interest in this regard that the latter occupies a right position on the phylogenetic trees within Proteobacteria and spirochaetes (Fig.1). Phylogenetic placement of mitochondria, chloroplasts, and R. prowazekii was supported by confidence limits (bootstrap values) to be above 50 % and did not change upon species sampling and eventual elimination from multiple alignments of alignable sequence regions of poor homology. Moreover, similar data were obtained by using ML in MOLPHY 2.2 as well as maximum parsimony and distance matrix-based analysis implemented in PAUP 4.0 and PHYLIP 3.6.
Legend to Fig. 1
Abbreviations of the cluster names are: lowGC and highGC - gram-positive bacteria with low and high G+C content respectively; a, g, and e are subdivisions of Proteobacteriae; spi - spirochaetes; cya - cyanobacteria; MT - mitochondria; CHL - chloroplasts; CYT and ER - eukaryotic cytoplasmic and microsomal forms.
The obtained results may be explained in the following way. Mitochondrial progenitor, which was suggested to have derived from yet highly reduced rickettsia-like endosymbiont , has lost htpG. A proto-eukaryote has acquired htpG from either cyanobacterium or chloroplast with Hsp90 being subsequently recruited to serve in mitochondria. If this is true, the common ancestor of all eukaryotes has once undergone cryptic cyanobacterial symbiosis or even harbored chloroplasts. Hsp90 of chloroplasts has in turn originated via duplication of the gene encoding Hsp90 of ER. It remains to be seen whether Hsp90 of free-living a-Proteobacteria which diverged well before the Rickettsiaceae family and mitochondria sister clades will group with rickettsial or mitochondrial homologs.Acknowledgements.
We are especially grateful to the staff of the Interregional Supercomputer Center of Russian Academy of Sciences (Moscow) for providing a possibility to use their powerful computers.References and Footnotes
Olga K. Mamayeva (1) and Victor V. Emelyanov (2)
Engelhardt Institute of Molecular Biology (1), Vavilov str. 32, 119991 Moscow, Russia,