Comparative genomics of white and opaque cell states supports an epigenetic mechanism of phenotypic switching in Candida albicans

Abstract

Several Candida species can undergo a heritable and reversible transition from a 'white' state to a mating proficient 'opaque' state. This ability relies on highly interconnected transcriptional networks that control cell-type-specific gene expression programs over multiple generations. Candida albicans, the most prominent pathogenic Candida species, provides a well-studied paradigm for the white-opaque transition. In this species, a network of at least eight transcriptional regulators controls the balance between white and opaque states that have distinct morphologies, transcriptional profiles, and physiological properties. Given the reversible nature and the high frequency of white-opaque transitions, it is widely assumed that this switch is governed by epigenetic mechanisms that occur independently of any changes in DNA sequence. However, a direct genomic comparison between white and opaque cells has yet to be performed. Here, we present a whole-genome comparative analysis of C. albicans white and opaque cells. This analysis revealed rare genetic changes between cell states, none of which are linked to white-opaque switching. This result is consistent with epigenetic mechanisms controlling cell state differentiation in C. albicans and provides direct evidence against a role for genetic variation in mediating the switch.

Keywords: Candida albicans; comparative genomics; epigenetic switch.

Figure 1

Cell morphologies of Candida albicans white and opaque cells. Cells were obtained from SC5314 derivatives that are MTLa/- (lineage A1) or MTLa/a (lineage B1). Scale bars, 10 μm.

Figure 2

Chromosome copy numbers and heterozygosity levels for white and opaque isolates. (A) Coverage was normalized to disomic levels and relative read depth was plotted across all eight Candida albicans chromosomes to determine their relative copy numbers, plotted in blue. The number of heterozygous positions was examined for each chromosome and average heterozygosity per 10 kbp window is plotted in red. Lines and tick marks denote chromosomes and centromere positions, respectively, on each chromosome. Shaded boxes indicate LOH regions unique to the B-1 and B-2 lineages. The parental strain SC5314 is included for reference. WH, white state; OP, opaque state. (B) Allele frequencies were determined for white/opaque cells and parental strain SC5314. Frequencies were compared across chromosomes as indicated for homolog A (blue) and homolog B (green) for each isolate. Note that allele frequencies have a median value of ∼50% for heterozygous disomic chromosomes, whereas median values are skewed toward 0% or 100% for chromosomes that are homozygous. Center lines represent median frequencies, upper and lower lines indicate 25th and 75th percentiles, shaded boxes indicate LOH regions unique to the B-1 and B-2 lineages.

Figure 3

Differences in DNA sequence identified between paired white and opaque cells of each lineage. Mutations were identified with either GATK or Pilon and were inspected in IGV. For mutations identified in close proximity to ORFs, the distance relative to the respective gene is specified. LOH, loss of heterozygosity; GOH, gain of heterozygosity; SNP, single nucleotide polymorphism; nonsyn, nonsynonymous mutation. Shaded rows indicate nonsynonymous mutations.

Figure 4

Copy number variation (CNV) between white and opaque cells. CNVs were identified using read coverage depth across 1 kbp windows (A) and different chromosomal elements (C) for the four A and B lineages (using two standard deviations above or below coverage levels as cutoffs for CNV). Comparisons were made between white/opaque states within a lineage and between A and B lineages for cells of the same type within the same sequencing run. Numbers indicate the total number of CNV regions identified in each comparison (above or below two standard deviations), lines indicate means and standard deviations. OP, opaque; WH, white. Venn diagrams indicate the number of shared CNVs identified between cell states for 1 kbp windows (B) and for chromosomal elements (D). Control comparisons between lineages as well as a full list of the common regions along with corresponding Candida albicans genome annotations are included in Supplementary Table S1.