Timea Marton1,2, Adeline Feri1,2, Pierre-Henri Commere3, Corinne Maufrais4, Natacha Sertour1, Gavin Sherlock5, Marie-Elisabeth Bougnoux1, Christophe d’Enfert1, Melanie Legrand1

1Institut Pasteur, INRA, Labex IBEID, Unité Biologie et Pathogénicité Fongiques, Paris, France
2Univ. Paris Diderot, Sorbonne Paris Cité, Cellule Pasteur, rue du Docteur Roux, Paris, France
3Institut Pasteur, Imagopole, Plate-Forme de Cytométrie, Paris, France
4Institut Pasteur, Centre d’Informatique pour la Biologie, Paris, France
5Department of Genetics, Stanford University, Stanford, California, USA

Identification of recessive lethal alleles in the Candida albicans genome: a constraint on loss-of-heterozygosity events in this species

Objectives
Candida albicans is a frequent human commensal yeast responsible for both mucosal and the majority of life-threatening nosocomial fungal infections. Its heterozygous diploid genome is highly plastic, with examples of aneuploidies and frequent loss-of-heterozygosity (LOH) events commonly associated with acquired antifungal resistance. Although LOH events can be seen on all 8 chromosome pairs, a chromosome homozygosis bias is observed for certain chromosomes regarding the retained homolog in the homozygous state. This suggests the occurrence of recessive lethal allele(s) (RLA) that would prevent large-scale LOH events from being stably maintained.

Methods
To further explore the impact of LOH on C. albicans biology, we took advantage of the Saccharomyces cerevisiae I-SceI endonuclease and fluorescent markers (GFP/BFP) to establish C. albicans strains whereby 1) a DNA double-strand break (DSB) can be induced at a specific location in the genome and 2) the occurrence of long-tract homozygosis resulting from repair of the induced DNA-DSB can be detected by FACS. 

Results
This approach was applied to identify RLAs on Chr7 of C. albicans strain SC5314, where haplotype B (HapB) is never found in the homozygous state, unlike haplotype A, suggesting the presence of RLA on Chr7B. I-SceI successfully targeted and induced a DNA-DSB on both Chr7 homologs as illustrated by an increase in LOH. However, our inability to recover cells homozygous for the right arm of Chr7B confirmed the existence of RLA(s). Mining of genome sequencing data for 154 C. albicans isolates allowed identifying a nonsense-generating SNP within the HapB allele of C7_03400c/MTR4, encoding a putatively essential RNA helicase. Further experiments confirmed that this nonsense SNP is indeed a RLA responsible for homozygosis bias on Chr7. As well, an RLA and a deleterious recessive allele on another chromosome (Chr4B) have been found explaining why Chr4A cannot be lost in strain SC5314.

Conclusion
This methodology also allowed highlighting that repeated regions, such as major repeat sequences (MRS), are hotspots for homologous recombination between homologs and play a role in LOH dynamics by generating new allele combinations upon haplotype swapping. Overall, our work uncovers the importance of MRSs and the constraint RLAs impose on LOH events in C. albicans.

3 / Identification of recessive lethal alleles in the Candida albicans genome: a constraint on loss-of-heterozygosity events in this species