Comparative EST analysis provides insights into the basal aquatic fungus Blastocladiella emersonii

Since the sequencing of the first complete fungal genome, the budding yeast Saccharomyces cerevisiae [1], fungal genomics and the specific area of comparative genome analysis in fungi have experienced a recent but impressive advance. Following sequencing of the genomes of two other ascomycetes, Schyzosaccharomyces pombe and Neurospora crassa [2, 3], efforts have focused on species throughout the fungal kingdom that represent diverse scientific interests. In this sense, genomes from plant and human pathogenic basidiomycetes have been sequenced [4, 5] and there are drafts or genome projects in progress of other fungi. Likewise, the sequencing of one zygomycete genome has been completed and there are two chytrid genome projects underway (see [6] for an overview of fungal genome sequencing projects).

Expressed sequence tag (EST) data from fungi, even though less numerous than genome sequences, are also showing to be useful to specific and diverse aims, such as mapping previously characterized genes [7], investigation of patterns of fungal genome evolution [8], prediction of novel genes [9], prediction of pathogenicity determinants [10], identification of disease-related sequences [11], improvement of functional assignments [12], and identification of alternatively spliced mRNA species [13].

Recently, we reported a sequencing program of nearly 17,000 ESTs corresponding to different developmental stages of Blastocladiella emersonii life cycle, an early diverging fungus that belongs to the Chytridiomycete class [14, 15]. Approximately 52% of the uniseqs presented similarity to sequences deposited in public data banks. Interestingly, several of these ESTs revealed similarity with known genes not previously reported in fungi, and which had been recognized as animal or plant specific proteins. Despite the fact that a consensus phylogenetic tree seems to resolve fungi and animals as sister groups [16], we wondered if some ancestral characteristics of fungi and animals or of fungi and plants could have been retained in this basal fungus and have been lost or become highly divergent in members of the late-branching group of fungi.

Our previous study provided the functional identification of putative unique transcripts based on sequence comparison, and contributed to linking in silico expression profile data with previous information about biological processes occurring throughout the fungal life cycle [14]. In this sense, the survey increased the knowledge about this interesting biological model. However, our approach did not provide a direct link between expressed sequences in B. emersonii and gene expression in major groups, like animals and plants.

In the present work, we carried out a large-scale comparative analysis of B. emersonii ESTs against protein and transcript sequences of fungi, animals and plants, using databases constructed ad hoc. Our goals were to identify putative orthologues in B. emersonii of genes previously classified as specific to animals and/or plants, as well as to find B. emersonii sequences common to fungi but which have evolved at a lower rate in this chytrid than in late-diverging fungi. Based on our results, we discuss possible relationships between expressed sequences and structures and/or biological processes occurring in animals and/or plants and B. emersonii, including a metabolic pathway previously reported only in animals and bacteria.

We carried out a comprehensive comparative EST analysis that identified one hundred and sixty-six expressed sequences from the aquatic fungus B. emersonii encoding putative proteins not previously reported in other fungi.

Among the ESTs with significant similarity to animal sequences, we found assembled sequences encoding enzymes involved in coenzyme B₁₂-dependent propionyl-CoA metabolism. Propionate, the second most abundant fatty acid in soil, is formed by fermentative processes from carbohydrates and several amino acids [37]. Propionate is converted to propionyl-CoA, which is also formed by oxidation of odd-chain fatty acids and several amino acids, and then converted to succinyl-CoA that then enters the central metabolism. This pathway is used in diverse metabolic processes and homologues of intervening enzymes were found within archaeal, bacterial and eukaryal genomes, but not in plants or fungi. In mammals, this route is employed in the catabolism of valine, isoleucine, methionine, threonine, thymine, cholesterol, as well as odd-chain fatty acids [38], and defects in some of the enzymes involved lead to the rare but severe inherited disease methylmalonyl aciduria [17].

Propionate is generally toxic to fungi and bacteria, and this is the reason why it is widely used as a preservative [39]. Despite its toxicity, many bacteria and fungi are able to use propionate as carbon and energy sources under aerobic conditions, using an alternative pathway to that mentioned above, the "methyl citrate cycle" that catalyses the oxidation of propionate to pyruvate.

Fungi such as S. cerevisiae and Aspergillus nidulans seem to lack cobalamin-dependent functions and therefore cannot use the methyl-malonyl-CoA pathway [18]. Leadley et al. [18] suggested that the maintenance of this pathway in proto-eukaryotes would have meant a high evolutionary cost, due to the need to preserve also the enzymes capable of producing coenzyme B₁₂, and at the same time, the existence of other pathways for propionate utilization may have superceded the selective pressure for preserving this metabolic route.

The question is why B. emersonii would express coenzyme B₁₂-dependent enzymes under the conditions tested. The presence in archaea, eubacteria and animals of coenzyme B₁₂ and coenzyme B₁₂-dependent enzymes seems to indicate that the conservation of these functions is important to diverse processes and this principle can also be applied to fungi. Despite the absence of genes encoding cobalamin-dependent proteins and enzymes of coenzyme B₁₂ biosynthesis in the fungal genomes sequenced, pathways involving this coenzyme could be active under conditions not frequently tested in other fungi whose genomes have not been sequenced yet. In fact, some of B. emersonii ESTs encoding these enzymes were isolated from a cDNA library constructed with mRNA isolated from cells exposed to high concentrations of cadmium. Differently from cadmium, several transition metals, such as cobalt, play a role as catalysts in a variety of enzymatic reactions. These metals, which are normally useful to the cells, can be toxic when in excess. Thus, many molecular mechanisms for cell detoxification have been developed. Some of these mechanisms are promiscuous, being responsible for detoxification of more than one of these heavy metals [40]. In this sense, we cannot rule out the possibility that some B. emersonii genes induced by exposure to cadmium could be involved in cobalt metabolism.

Our analysis has also shown eleven sequences associated to flagella-related proteins expressed in B. emersonii and animals, nine of them also present in green algae. The absence of these sequences from fungi and plant databases was expected, as a consequence of the bias in the most investigated species, which mainly belong to late-diverging fungi and land plants. Thus, B. emersonii could be a good model to study processes related to flagella structure and movement, probably contributing to the characterization of differences between animals and other flagellated cells.

Comparison of B. emersonii ESTs with plant sequences revealed two assembled sequences with high similarity to genes found only in plants, but encoding proteins that belong to families with members also in animals and fungi: a putative Isp4 protein (an oligopeptide transporter) and a putative syntaxin 71 (a member of SYP7 family of protein receptors). Oligopeptides can be used as source of amino acids, nitrogen and carbon, and their transporters have been documented in bacteria, fungi and plants. The identification of multiple OPTs in Arabidopsis, with tissue-specific expression patterns, supports the idea of different functional roles for these transporters, e. g., regulators of hormone activity in hormone-peptide conjugates [22]. There is also evidence indicating that members of other peptide transporter family, the PTR, have a role in plant growth and development. Thus, it is possible that the putative OPT found in B. emersonii has a specific function, different from those described in other fungi, as regulation of growth and differentiation.

In this same context, we can include the matches of B. emersonii ESTs with two plant receptors associated with the control of proliferation and development in plants. Even though the alignments extend over a conserved region in the plant sequences, domains characteristic of the assignments do not overlap. Consequently, B. emersonii proteins could be involved in completely different processes. Further studies will be necessary to clarify this hypothesis.

The syntaxin family of proteins is well represented in B. emersonii transcriptome. We found representatives for all except one (SYP8) of the defined families [24] (Table 3). The group included syntaxin 71, a member of SYP7 family, which seems exclusive of plants. Such broad representation suggests that syntaxin 71 could have specific functions in B. emersonii, perhaps related to functions developed in plants.

Members of the syntaxin family are known to play an important role in the fusion of transport vesicles with specific organelles [23], and specifically SYP7 proteins seem to be involved in transport between the ER and the Golgi apparatus [41]. Interestingly, membrane transport and vesicle rearrangement have critical importance during the sporulation stage of B. emersonii life cycle [42], and several ESTs related to this function were exclusively isolated from sporulating cells (see [43] GO:0006886 intracellular protein transport), which includes the EST encoding the possible syntaxin 71.

B. emersonii-animal-plant common sequences included an urocanase, an enzyme with no records in sequenced fungal genomes and which could be indicative of B. emersonii's ability to use histidine as a carbon source. The presence of sequences encoding enzymes possibly involved in the catabolism of valine, isoleucine, methionine and threonine (such as the enzymes that are active in coenzyme B12-dependent propionyl-CoA metabolism), and enzymes possibly involved in the catabolism of histidine, suggest that B. emersonii metabolism might be directed towards amino acid catabolism, as a source of carbon. Early studies in chytrids indicated distinct roles for some amino acids, other than serving as nitrogen source or protein building blocks. For instance, certain amino acids have been shown to be effective in initiating growth on sugars different from glucose, such as mannose and fructose, in Allomyces macrogynus cultures, presumably supplying both carbon and nitrogen sources [44].

An assembled sequence encoding a FBA type C, a member of class I FBA, was also found in our analyses. FBAs are divided into two non-related protein classes: Class I FBA, not found in fungi but with widespread distribution in other eukaryotes and also found in prokaryotes, and Class II FBA, identified mainly in eubacteria and also in eukaryotes, including fungi [29, 45]. Although the scattered taxonomic distribution of FBA classes does not have a consensual evolutionary explanation yet, gene duplication events and replacement of one paralog by the other are events that could have occurred. For instance, there is some evidence for the existence of an ancestral class II aldolase, from the endosymbiosis with a cyanobacterium, which could have been replaced by a class I aldolase in red and green algae, as well as in higher plants [29, 46]. Class II FBA genes of ascomycetes are also of eubacterial origin, and probably consequence of endosymbiosis with mitochondria ancestors [47]. Thus, a gene replacement event such as the proposed for red and green algae and land plants could similarly be proposed for the origin of B. emersonii FBA gene.

Even more interesting than the presence of a member of class I FBA in B. emersonii, could be the type observed, the C type, which is supposed to have evolved after divergence of the B type [48]. In fact, no B type FBA sequences were observed among B. emersonii ESTs. One possible explanation would be that the C type FBA could have replaced the B type. Another explanation, perhaps the simplest one, is that the B type sequence was not found among B. emersonii sequenced ESTs, but the gene is present in the genome. However, why both types of FBAs would be expressed in B. emersonii is not clear yet. In vertebrates, Class I comprises three types of isozymes expressed in different tissues: aldolase A (muscle type, also expressed in brain), B (liver type) and C (brain type). As described for other enzymes of the glycolytic pathway, aldolases A and C display activities different from that observed during glucose metabolism, as they regulate the stability of the light neurofilament mRNA through their ribonuclease activity [49]. A specialized function for the C type aldolase in B. emersonii should not be ruled out.

In addition, two putative genes encoding the alpha and gamma-SNAPs were observed in B. emersonii. Until now, no gamma-SNAPs have been described in fungi, B. emersonii being the first fungus in which this gene has been identified. However, the presence of both alpha and gamma-SNAPs in eukaryotic cells seems to be the rule, with fungi being the exception, considering that five phylogenetically distant species are known to possess both alpha and gamma-SNAP: D. melanogaster, B. taurus, H. sapiens, A. thaliana and D. discoideum [30]. The protein alpha-SNAP is essential for membrane traffic because it allows efficient NSF/SNARE interaction. Instead of this direct function, gamma-SNAP could have a regulatory role in membrane fusion. It was also suggested a role for gamma-SNAP in mitochondrial dynamics, contributing as an adaptor in the attachment of mitochondria to the cytoskeleton [50]. Our results in B. emersonii indicate that D. discoideum is not the only simple eukaryote containing both alpha and gamma-SNAPs.

As a second approach to discover non-typical fungal genes in B. emersonii, we carried out a comparative analysis with the expressed sequences of this aquatic fungus and other fungal sequences. We intended to be conservative at the time of selection and very few sequences were identified. Likewise, several difficulties arose due to the complexity of dealing with large multigene families. Indeed, two of the four selected sequences initially collected were not orthologues, and the relationship between the other two is not evident, but we decided to include these sequences in our analysis because the information extracted was also relevant. In fact, even though the sequences encoding the Rjl protein were not reported in fungi or plants, we did not detect them among the 105 B. emersonii-animal-shared sequences selected in our first approach. The divergence found for the other three cases (EF1α, ADP/ATP translocase and aminocarboxymuconate semialdehyde decarboxylase) is also noteworthy, because it could reflect high evolutionary rates, gene duplication and replacement (as suggested for EF1α in [34]), gene conversion, or horizontal gene transfer from prokaryotes to eukaryotes or among eukaryotes, which seems to be more common that previously thought (see [31]).

This collection of selected B. emersonii assembled sequences represents the result of approaches that use comparative EST analyses to address differences and similarities between chytrids and other eukaryotes (other fungi, animals and plants). The results of such analyses will probably suffer modifications when more fungal sequences are available. Specifically, other chytrid and zygomycete sequences will contribute to define retained, lost and divergent genes. Moreover, at least part of the borderline sequences (with an Evalue between 10^-3 and 10^-5 against ad hoc fungal database), which could be true divergent homologues, could constitute a group of interest to help understand phylogenetic relationships among fungi.

Through two different approaches involving comparative sequence analyses, and using computational tools and manual revision, we identified 162 protein-coding sequences from B. emersonii previously described in animals (such as coenzyme B₁₂-dependent propionyl-CoA pathway members, and proteins related to flagella structure or movement), in plants (such as protein receptors, a putative member of small olipeptide transporter, and a SYP7 family member of syntaxins), and in animals and plants (such as an urocanase, a fructose-bisphosphate aldolase (FBA) type C, members of the MtN3 family and a gamma-SNAP representative). We also found 4 sequences from B. emersonii, which were identified in a fungal sequence comparison as not found or highly divergent from other fungal species: a Rbj-like protein (similar to animal proteins), an EF-like protein (dispersely distributed in taxa, already described in [14]), an ADP/ATP translocase (similar neither to plant nor to animal sequences) and a 2-amino-3-carboxymuconate-6-semialdehyde decarboxylase (different from fungal sequences, poorly characterized). When the selected ESTs were classified according to the biological processes in which they could be involved, cell growth and maintenance, signal transduction and metabolism resulted as the biological processes most represented. Some sequences selected were expected, based on the knowledge about chytrids, like those associated to specific structures not found in other fungi (e. g., flagellar-associated ESTs). Thus, B. emersonii seems to be an interesting model to study flagella-associated structures or functions.

Among the ESTs exclusively isolated from sporulating cells, we collected sequences associated to membrane transport, such as syntaxin 71. Membrane fusion and vesicle rearrangement are crucial events in B. emersonii sporulation, when cytokinesis occurs. A set of core SNAREs is apparently sufficient to mediate most intracellular vesicle fusion events, although multicellular organisms would express additional SNARE proteins for specific functions associated with the body complexity [51]. Thus, proteins like syntaxin 71 are good candidates to function as additional SNARE proteins in the transition of unicellular multinucleated zoosporangia to zoospores during B. emersonii life cycle.

Other collected ESTs were unexpected, like those involved in specific metabolic pathways, such as sequences involved in conversion of propionate and histidine to glutamate. We hypothesize that alternative pathways leading to the use of amino acids and other substrates as carbon and nitrogen sources could have been lost in late-diverging fungi and retained in basal fungi.

Finally, a large number of sequences selected by the first approach were not classified in a known process, which suggests that other structures or biological processes not identified yet can be shared by B. emersonii, animals and plants.