**3. Conclusion**

306 Gene Duplication

integrated by CgYps12 and ScBar1 of *S. cerevisiae*, a putative orthologous pair with low similitude synteny, but with a clear ancestor-descendant relationship (Fig. 6G). Finally, family E was formed by a representative of each *Candida* spp. Yps, CgYps7 and ScYps7. This family forms a sub tree with the same topology as those phylogenies constructed with ribosomal and other protein sequences (Diezman *et al.,* 2004). The CgYPS7 and ScYPS7 genes exhibited an extensive synteny (Fig. 6C), but no synteny with *CaYPS* (orf19.6481) and *CdYPS* (Cd36\_72090) was observed (Fig. 6D). In *C. albicans* and *C. dubliniensis* genome databases these *YPS* are described as *ScYPS7* orthologous genes (Schaefer *et al*., 2007). Nevertheless, both *YPS* exhibited low similarity with *ScYPS7* (37.2-38.7%) and no-synteny. The final decision to consider family E as an orthologous family will depend on comparative

Families C, F, G and H have not any *C. glabrata* or *S. cerevisiae* Yps representative protein*.*  Families C and H were formed only by one ClYps gene of *C. lusitaniae* and seven CpYps genes of *C. parapsilosis*, respectively (Fig. 5). Curiously, *C. lusitaniae* is the species that harbours the fewest Cl*YPS* (n=1) and *SAP* (n=3) genes, and its isolation frequency from clinical samples ,as well as its virulence, are lower than the other *Candida* species (Abi-Said et al. 1997)*.* This evidence supports a hypothesis of relevance of aspartyl proteases in virulence. That is, species with numerous aspartyl proteases in virulence; species with broad aspartyl proteases are more virulent than those with a limited number of these proteins. Family F harboured *C. albicans, C. dubliniensis* and *C. tropicalis* yapsins organized congruently according to the ribosomal phylogenetic tree. The *C. albicans* CaBar1 (orf19.2082) and *C. dubliniensis* CdBar1 (Cd36\_15430) gene, found in family F, has been described as orthologous to *S. cerevisiae BAR1* (Schaefer *et al.,* 2007) found in family C. In both species, *C. albicans* and *S. cerevisiae*, the protein is involved in alpha pheromone degradation and secreted to the periplasmic space of mating alpha-type cells. These proteins help cells find mating partners by cleaving and inactivating the alpha factor, which allows cells to recover from alpha-factor-induced cell cycle arrest (Mackay *et al.,* 1988). The *in silico* analysis performed in this work established that these proteins and the Bar1 from *C. dubliniensis* are extracellular, but anchored to the cell wall or cell membrane. Also, phylogenetic analysis shows that Bar1 from *C. albicans* and *C. dubliniensis* belongs to the Yps superfamily, with a similarity of 40%, and are not grouped with CgYps12 of *C. glabrata*  (CgYps12 or CgBar1) and Bar1 of *S. cerevisiae*. The reason for which an aspartyl protease, that apparently is secreted, is groupedwith the yapsines superfamily could be a mistake in the cell location method because almost all software use the signal peptide, transmembranal regions, and the GPI site in the C-terminal, to predict the cell location. In *C. albicans* it has been detected that aspartyl proteases are associated with the plasmatic membrane, or to both the plasmatic membrane and cell wall. This makes the experimental corroboration of the cell location necessary. The Bar1 protein of *C. albicans* has been described as a protein with three domains: 2 aspartyl protease domains and another unidentified. Apparently, this GPI-membrane anchored domain determines that Bar proteins are not secreted, but anchored to cellular membranes, and their two actives sites are oriented to cellular membranes, and their two actives sites are oriented to the exterior to inactivate alpha pheromone, which is secreted by Mat-alpha cells. In *C. albicans*, the degradation of secreted alpha pheromone is not exclusive to Bar1. CaYPS7 (orf19.6481) of family E also encodes for this function with lesser efficiency (Schaefer *et al*., 2007). This physiological redundancy has not been demonstrated in *S. cerevisiae* ScYps7. *C. albicans* can mate under some *in vitro* and *in* 

analyses of functional features not yet performed.

Why have *C. albicans/C. dubliniensis* and *C. glabrata*/*S. cerevisiae* been suffering some genetic duplication events in their Sap and Yps superfamilies? This is something that has not been resolved, but it is clear that the decrease in virulence in null mutants, in both *CaSAP* and *CgYPS,* endorse the idea that the presence and expansion of *SAP* and *YPS* families is necessary for adaptation to the host, and therefore for survival and virulence. Also, species with broad aspartyl protease families are more virulent than those with a limited number of these proteins. *C. glabrata* belongs to a phylogenetic group with no pathogenic yeast, and its virulence attributes could be evolving independently from the CTG clade, where *C. albicans* is the main opportunistic pathogenic species. The expansion of the *CgYPS* gene superfamily of *C. glabrata* maintains a parallelism with the expansion of the *SAP* gene superfamily of *C. albicans*, and constitutes a possible example of convergent evolution. The transition from a commensally life style to a successful opportunistic pathogen could be related to gene expansion that encodes for each kind of aspartyl protease. A lot of experimental methodologies must be performed to recognize the orthologous gene families, as well as the virulence, participation and transition commensal-pathogen roles of aspartyl proteases, including Sap and Yps.
