Strain background interacts with chromosome 7 aneuploidy to determine commensal and virulence phenotypes in Candida albicans

Abstract

The human fungal pathobiont Candida albicans displays extensive genomic plasticity, including large-scale chromosomal changes such as aneuploidy. Chromosome trisomy appears frequently in natural and laboratory strains of C. albicans. Trisomy of specific chromosomes has been linked to large phenotypic effects, such as increased murine gut colonization by strains trisomic for chromosome 7 (Chr7). However, studies of whole-chromosome aneuploidy are generally limited to the SC5314 genome reference strain, making it unclear whether the imparted phenotypes are conserved across C. albicans genetic backgrounds. Here, we report the presence of a Chr7 trisomy in the "commensal-like" oral candidiasis strain, 529L, and dissect the contribution of Chr7 trisomy to colonization and virulence in 529L and SC5314. These experiments show that strain background and homolog identity (i.e., AAB vs ABB) interact with Chr7 trisomy to alter commensal and virulence phenotypes in multiple host niches. In vitro filamentation was consistently reduced by Chr7 trisomy in SC5314, but this result was not consistent for 529L. Oral colonization of mice was increased by the presence of a Chr7 trisomy in 529L but not SC5314; conversely, virulence during systemic infection was reduced by Chr7 trisomy in SC5314 but not 529L. Strikingly, the AAB Chr7 trisomy in the SC5314 background rendered this strain avirulent in murine systemic infection. Increased dosage of NRG1 failed to reproduce most of the Chr7 trisomy phenotypes. Our results demonstrate that aneuploidy interacts with background genetic variation to produce complex phenotypic patterns that deviate from our current understanding in the genome reference strain.

Fig 1. Isolation of Candida albicans strains disomic and trisomic for chromosome 7.

Results of whole genome sequencing of five strains from two genetic backgrounds (SC5314 and 529L) were visualized using the YMAP platform [27] . Copy number is indicated on the vertical axis such that Chr7 trisomy is reflected in black bars above the lengthwise 2N midpoint of each chromosome. For SC5314 strains, homologs A and B for each chromosome based on the reference genotype are indicated by cyan and magenta, respectively. Heterozygous positions are marked in grey.

Fig 2. Growth in rich medium and medium-chain fatty acids is minimally affected by Chr7 trisomy.

(A) Each strain was grown in liquid YPD at 30°C for 31 h. The average optical density is plotted as a thick line with the 95% confidence intervals shaded. N = 6. (B) The fastest doubling time and the carrying capacity (max OD) was determined from each growth curve and plotted. Boxplots represent the interquartile ranges with the median marked and whiskers extending to the outermost data points up to 1.5 times the interquartile range. One-way ANOVAs for each strain background, followed by Tukey’s HSD test (*** represents p < 0.001). (C) The optical density of each strain was determined after 48 h of growth across a range of concentrations of medium-chain antifungal fatty acids, undecanoic acid and undec-10-noic acid. Means are plotted with standard error. N = 3 (two-way ANOVAs for each strain background for each day, with strain and concentration as the two fixed factors).

Fig 3. In vitro filamentation is reduced by Chr7 trisomy in both strain backgrounds.

(A) Cellular morphology of the five strains grown under hyphal-inducing conditions, liquid RPMI at 30°C, for three hours. Scale bar = 20 μm. (B) At least 50 cells across 5 + fields of view were categorized as either yeast or filamentous for each strain per replicate and plotted as the average with standard error. Binomial GLMs for each strain background; * for p < 0.05, ** for p < 0.01, and *** for p < 0.001. N = 3. (C) Between 80-120 cells were plated to rich (YPD) and hyphae-inducing (Spider) solid media and allowed to grow for 5 d at 30°C and 37°C. Representative colony morphologies of the five strains are shown. (D) The filamentation score for colonies from each plate was quantified using the area of radial filamentation and center yeast colony as described in [30]. Boxplots represent the interquartile ranges with the median marked and whiskers extending to the outermost data points up to 1.5 times the interquartile range. One-way ANOVAs for each strain background; * for p < 0.05, ** for p < 0.01, and *** for p < 0.001. N = 6.

Fig 4. Only Chr7x3 ABB trisomy in SC5314 altered epithelial cell interactions.

(A) Confocal microscopy of internalized C. albicans by oral epithelial cells after incubation for 2.5 h. Nuclei from all cells are visualized using DAPI, C. albicans (Ca) in green, and extracellular C. albicans (Ca) in red. An example of an internalized C. albicans cell segment is marked with an arrow. Scale bar 50µm. (B) C. albicans yeast from each strain were incubated with OKF6/TERT-2 cells for 2.5 h. Monolayers were washed to remove non-adherent cells and the C. albicans remaining were determined per field of view. HPF = High Power Field. N = 9. (C) Infected monolayers were labeled by rabbit anti-C. albicans antiserum conjugated with Alexa 488, permeabilized, and stained with mouse anti-C. albicans antiserum conjugated to Alexa 568 to determine internalization. 100 cells were assayed across 10-12 fields of view. N = 9. (D) OKF6/TERT-2 cells pre-loaded with Cr⁵¹ were incubated with 2x10⁵ C. albicans cells from each strain and allowed to interact. After 7 h, the supernatant of each culture was collected and the Cr⁵¹ released determined by autoradiography. All boxplots represent the interquartile ranges with the median marked and whiskers extending to the outermost data points up to 1.5 times the interquartile range. One-way ANOVAs within the two strain backgrounds for each assay, followed by Tukey’s HSD test (* for p < 0.05, ** for p < 0.01). N = 9.

Fig 5. Oral colonization and cytokine responses were altered by Chr7 trisomy.

(A) Oral infections of male C57BL/6 mice were initiated using cotton balls soaked with 2x10⁷ C. albicans cells from each strain. Tongues were collected on Days 1 and 5 post-infection from independent sets of mice and used to measure the cytokine levels or C. albicans abundance as indicated. (B) Extracted tongue weights of mice infected with each strain at Day 1 and Day 5. Boxplots represent the interquartile ranges with the median marked and whiskers extending to the outermost data points up to 1.5 times the interquartile range. N = 10. (C) Tongues were homogenized and 70 μL of homogenate was plated for colony forming units on YPD. Colonies were counted after 2 d growth at 30°C. Boxplots represent the interquartile ranges with the median marked and whiskers extending to the outermost data points up to 1.5 times the interquartile range. Two-way ANOVAs within each strain background with strain and day as the fixed factors, followed by Tukey’s HSD test (*** represents p < 0.001). N = 10. (D) The abundance of eight cytokines and the CXCL1 chemokine was determined from tongue homogenates collected on Day 1. Scores were normalized within each cytokine by z-score and the average plotted on a colorimetric scale. Cytokines and CXCL1 are arranged in the order of increasing concentrations for the diploid SC5314 strain.

Fig 6. Chr7 trisomy dramatically reduced systemic virulence and promoted organ colonization in SC5314 but not in 529L.

(A) C57BL/6 mice were infected with the five C. albicans strains via tail vein injection. One set of mice was monitored for mortality up to two weeks, while the other set of mice was sacrificed 2 days post-injection to recover their organs and obtain CFU counts. (B) Ten C57BL/6 mice (5 male, 5 female) were injected with 5x10⁵ C. albicans cells for each of the five Chr7 disomic or trisomic strains. Survival of the mice is indicated in the Kaplan-Meier curve and significance of pairwise comparisons was determined by Log-Rank test followed by Benjamini-Hochberg correction (p < 0.01 for all three pairwise comparisons in SC5314 background). (C) The brain, kidney, and spleen were harvested from ten C57BL/6 mice (5 male, 5 female) from each strain at 2 d post-infection and organ homogenates plated to determine colony forming units (CFUs) normalized to organ weight. One-way ANOVAs for each organ within a given strain background, followed by Tukey’s HSD test (* for p < 0.05, ** for p < 0.01, ** for p < 0.01). N = 10.

Fig 7. Addition of NRG1 does not phenocopy Chr7 trisomy.

(A) Diploid strains with an extra copy of NRG1 were assayed for filamentation under hyphal-inducing conditions, liquid RPMI at 30°C, for three hours. At least 50 cells across 5 + fields of view were categorized as either yeast or filamentous for each strain per replicate and plotted as the average with standard error. Binomial GLMs for each strain background; *** for p < 0.001. N = 3. (B) Fungal cell adherence and uptake after co-incubation with OKF6/TERT-2 cells for 2.5 h and epithelial cell damage after 7 h were assayed. All boxplots represent the interquartile ranges with the median marked and whiskers extending to the outermost data points up to 1.5 times the interquartile range. One-way ANOVAs within the two strain backgrounds for each assay (** for p < 0.01, *** for p < 0.001). HPF = High Power Field. N = 9. (C) Ten C57BL/6 mice (5 male, 5 female) were injected via the tail vein with 5x10⁵ C. albicans cells. Survival of the mice is indicated in the Kaplan-Meier curve and significance of pairwise comparisons was determined by Log-Rank test followed by Benjamini-Hochberg correction (p < 0.05 for Chr7x2 v. Chr7x3 (ABB) or Chr7x2 + NRG1).

Fig 8. Phenotypic changes due to Chr7 trisomy reveal strain-specific effects and correlated changes.

(A) Scaled phenotypic values of the Chr7 trisomic strains against the corresponding diploid from the same genetic background. The phenotypes of the two diploids were set to 1 and the aneuploid phenotypic values were linearly scaled against it, except for the CFU count assays, which were scaled on a log scale. Phenotypes are color-coded as indicated. (B) Spearman correlations were conducted between the measured phenotypes, based on all five strains used in this study. The shade of blue or red denotes the strength of positive of negative correlation coefficients, respectively, and the statistical significance is denoted with asterisks (* for p < 0.05, ** for p < 0.01, and *** for p < 0.001). The correlation matrix is ordered using the angular order of the eigenvectors (corrplot package in R).