Stress responses and virulence in human fungal pathogens: a comparative systems biology approach of regulatory networks in Candida species – CANDIHUB

For the two yeasts under investigation, Candida albicans and Candida glabrata, knock-out and overexpression mutants for a set of genes encoding transcription factors were obtained. The phenotypic analysis (culture in the presence of different stresses) of these mutants makes it possible to propose a biological function for these transcription factors. Transcriptomic analysis (RNA-Seq) identifies genes regulated directly or indirectly by these transcription factors. In parallel, mutants expressing tagged versions of the transcription factors have been produced which allow TF binding sites to be identified at the genomic scale (ChIp-Seq). The combination of transcriptomic and binding data makes it possible to reveal target genes for each transcription factor. The combination of the information obtained for several transcription factors makes it possible to establish a regulatory network and to identify regulatory hubs whose function can then be explored in detail using phenotypic tests, omics approaches including proteomics and tests to assess the virulence of yeasts ex vivo and in vivo.
The analysis of regulatory networks has been facilitated by the development of new bioinformatic tools: bPeaks for the identification of binding sites of transcription factors; Pixel for the analysis of large multi-omic datasets; iHKGviewer for exploring regulatory networks.

For each of the two yeasts under investigation, more than 12 transcription factors were studied in detail by phenotypic, transcriptomic and chromatin immunoprecipitation approaches.
In the yeast C. glabrata, a regulatory network comprising 12 of the 17 transcription factors studied has been established, showing, among other things, an original mechanism for regulating the response to excess iron, involving the cooperation of factors of Hap5 and Yap5 transcription. In addition, an unexpected mechanism of adaptation to iron deficiency has also been demonstrated involving the proteins Dom34 and Hbs1.
For the yeast C. albicans, a regulatory network comprising 12 transcription factors has been established, the exploration of which is in progress. Four hubs have already been identified - Rme1, Rfg1, Orf19.217 and Crz2 - which appear to play a central and hitherto unknown role in the stress adaptation of C. albicans. Dedicated studies have shown the involvement of Crz2 in adaptation to acidic pH and hypoxia and hence the ability of C. albicans to colonize the gastrointestinal tract. Further work has shown the importance of Skn7 in the homeostasis of reactive oxygen species during filament formation by C. albicans.

The transcriptomic data produced in the framework of this project are massive and have allowed the establishment of regulatory networks for C. glabrata and C. albicans, the topology and composition of which have only been partially studied. In particular, the identification of hubs is still partial and its continuation should open new research avenues. The first works carried out on a few, Dom34 and Hbs1 in C. glabrata and Rme1, Rfg1, Orf19.217 and Crz2 in C. albicans are promising.
One of the objectives of CANDIHUB was to define regulatory networks involving orthologous transcription factors in C. albicans and C. glabrata in order to compare the structure of the networks obtained and to understand how they had been modified during evolution. This endeavor remains relevant.

Taken as a whole, the CANDIHUB project has so far resulted in 12 publications in international journals. Several publications focus on the characterization of regulatory networks in C. glabrata (PMID: 27242683, 28615656, 30505294). Further publications focus on the detailed characterization of transcription factors in C. albicans (PMID: 28752552, 29998470). Two publications relate to tool developments (PMID: 30286789, 30944779). Finally, three reviews and a methodological article were published (PMID: 30368597, 31050635, 31172012, 31370875).

Candida albicans and Candida glabrata are the two major yeast species responsible for invasive fungal infections in humans. Systemic infections due to C. albicans and C. glabrata are associated with high mortality rates despite the existing antifungal arsenal. A better knowledge of the mechanisms involved in their pathogenesis is likely to reveal new strategies to fight these infections. Several features contribute to C. albicans and C. glabrata virulence among which an ability to adapt to a wide range of variations in environmental parameters and, in the case of C. albicans, the ability to undergo a reversible yeast-to-hypha transition. In this respect, it has been proposed that deciphering the transcriptional regulatory networks that control these different properties and their interplay has the potential to uncover so-called regulatory hubs – transcription factors that show high connectivity within networks and can act as integrators of diverse cellular processes – and that these hubs could represent interesting targets for antifungal development. Hence, the CANDIHUB project aims at deciphering the C. albicans and C. glabrata transcriptional regulatory networks involved in the adaptation of these two species to varying environments and subsequently testing the contribution of the regulatory hubs identified in these networks to C. albicans and C. glabrata fitness and virulence.
In order to achieve this objective, we will take advantage of our previous identification of homologous C. albicans and C. glabrata transcription factor genes that show changes in expression in cells exposed to a pleiotropic stress. For each of these transcription factors, we will define its transcription module (the set of its direct target genes) using genome-wide transcript profiling (by RNA-Seq) and chromatin immunoprecipitation (by ChIP-Seq). The combination of the transcription modules obtained for the 16 C. glabrata and 13 C. albicans transcription factors to be investigated in CANDIHUB will enable us to infer a global stress regulation network for each of these two species, which will be validated in particular through advanced proteomics approaches. Additionally, regulatory hubs will be identified within these networks and their functional characterization using loss-of-function and gain-of-function mutants in vitro or in ex vivo or in vivo models of host-pathogen interaction will reveal whether they constitute relevant targets for antifungal therapies. Furthermore, comparison of the global regulatory networks, inferred using homologous transcription factors, will help estimating the divergence and convergence between the strategies used by these two distantly related yeasts to infect humans. Dissemination of CANDIHUB results will be promoted through a web-based database compiling and mapping transcriptional regulatory interactions in C. albicans and C. glabrata, thus providing an important resource to the Candida community, which is missing at the moment. Overall, our study will be unique by the amount of direct experimental data produced and used for regulatory network reconstruction in Candida species, by the coherence of the data produced and by the parallel analyses of two distantly related yeast pathogens.
The complementary expertise of the three CANDIHUB Partners will ensure the success of this project. Indeed, these Partners are internationally recognized for their expertise in C. albicans post-genomics and biology (Partner 1: Christophe d’Enfert, Institut Pasteur, Paris, France), the study of regulatory networks in hemiascomycetous yeasts and, specifically, C. glabrata (Partner 2: Frédéric Devaux, Université Pierre et Marie Curie, Paris, France), the development of bioinformatics tools and databases and in silico study of transcriptional regulatory networks (Partner 3: Gaëlle Lelandais, Institut Jacques Monod, Paris, France) and proteomics (Partner 3: Jean-Michel Camadro, Institut Jacques Monod, Paris, France).