(C) PLOS One
This story was originally published by PLOS One and is unaltered.
 . . . . . . . . . .



Candida albicans selection for human commensalism results in substantial within-host diversity without decreasing fitness for invasive disease [1]

['Faith M. Anderson', 'Department Of Microbiology', 'Immunology', 'University Of Michigan Medical School', 'Ann Arbor', 'Michigan', 'United States Of America', 'Noelle D. Visser', 'Kevin R. Amses', 'Department Of Ecology']

Date: 2023-05 

Candida albicans is a frequent colonizer of human mucosal surfaces as well as an opportunistic pathogen. C. albicans is remarkably versatile in its ability to colonize diverse host sites with differences in oxygen and nutrient availability, pH, immune responses, and resident microbes, among other cues. It is unclear how the genetic background of a commensal colonizing population can influence the shift to pathogenicity. Therefore, we examined 910 commensal isolates from 35 healthy donors to identify host niche-specific adaptations. We demonstrate that healthy people are reservoirs for genotypically and phenotypically diverse C. albicans strains. Using limited diversity exploitation, we identified a single nucleotide change in the uncharacterized ZMS1 transcription factor that was sufficient to drive hyper invasion into agar. We found that SC5314 was significantly different from the majority of both commensal and bloodstream isolates in its ability to induce host cell death. However, our commensal strains retained the capacity to cause disease in the Galleria model of systemic infection, including outcompeting the SC5314 reference strain during systemic competition assays. This study provides a global view of commensal strain variation and within-host strain diversity of C. albicans and suggests that selection for commensalism in humans does not result in a fitness cost for invasive disease.

Funding: Funding for this project included mCubed grant to TRO and TYJ, NIH grant NIH KAI137299 (NIAID) to TRO, NIAID T32 AI007528 to FMA, NIH 1F31HG010569 to AMW, NIH T32GM007544 to KRA, University of Michigan Postdoctoral Pioneer Program to MJM. TYJ is a fellow of the Canadian Institute for Advanced Research program Fungal Kingdom: Threats & Opportunities. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Here, we demonstrate that healthy people are reservoirs for genotypically and phenotypically diverse C. albicans strains that retain their capacity to cause disease. Our results indicate that individuals can be colonized by strains from multiple clades and that these commensal isolates have extensive variation in growth, stress response, biofilm formation, and interaction with macrophages, but this variation was not explainable by sample origin site. We found that although the commensal isolates had a reduced capacity to cause macrophage cell death, the majority of the strains showed increased competitive fitness in invertebrate models of systemic candidiasis compared with the reference SC5314 strain and retained their capacity to cause disease during monotypic infections. Together, these data suggest that the selective pressures experienced by C. albicans during commensal colonization do not necessarily result in decreased pathogenic potential.

Recent work suggests that C. albicans experiences fitness tradeoffs between invasive and colonizing growth, with selection for commensal behavior during colonization [ 36 – 42 ]. In serial passage experiments or competitive fitness experiments in the gut, mutations in key transcription factors controlling hyphal formation, EFG1 and FLO8, resulted in increased fitness in the gut and decreased fitness in systemic models of infection [ 36 – 41 ]. In oral candidiasis, trisomic strains have been identified with a commensal phenotype [ 43 ]. Additionally, the 529L strain of C. albicans causes less damage and inflammation during oropharyngeal candidiasis [ 44 ] and persists at a higher fungal burden in both the mouth and the gut [ 21 ]. These potentially divergent selection pressures imply that commensal C. albicans strains can differ from isolates that cause invasive disease, but the genetic determinants underlying this difference have not been defined.

While host immune status is known to be an important predictor of disease outcomes, whether genetic variation between C. albicans strains also contributes to differences in virulence remains an open question. In the model yeast Saccharomyces cerevisiae, there is extensive variability in genotype and phenotype among different isolates that has been linked to the ability of S. cerevisiae to adapt to a wide range of environmental conditions [ 16 , 17 ]. Recent work has highlighted intraspecies variation in important aspects of C. albicans biology [ 18 – 21 ]. There are currently 17 clades of C. albicans, which were initially defined through multilocus sequence typing (MLST) [ 22 – 24 ]. More recently, genome sequencing has resolved finer population structure in the group [ 25 ]. Genetic epidemiology studies suggest that the major clades of C. albicans may differ in how frequently they are isolated from bloodstream infections or asymptomatic colonization [ 26 ], with 5 clades accounting for the majority of clinical isolates [ 27 – 29 ]. These clinical isolates have demonstrated significant variation in murine models of systemic infection, biofilm formation, cell wall remodeling, secretion of toxins and proteolytic enzymes, and morphological plasticity [ 18 , 26 , 30 – 35 ]. However, our primary understanding of the genotype–phenotype relationship in C. albicans results from analyses of a relatively limited set of pathogenic clinical isolates and their laboratory derivatives, with the majority of the work performed in the SC5314 genetic background. Detailed analysis of the genetic determinants of biofilm regulation between 5 different strains, each representing a different clade of C. albicans, have highlighted that circuit diversification is widespread between strains [ 18 ], adding complexity to our understanding of C. albicans biology.

Candida albicans is a common colonizer of humans, with between 20% and 80% of the world’s population estimated to be asymptomatically colonized at any given time [ 1 , 2 ], although this depends on many factors, including host health status and diet [ 3 – 5 ]. Colonization occurs at multiple body sites including the mouth, skin, GI tract, and vaginal tract [ 6 , 7 ]. These sites present a wide range of physiological stresses to colonizing fungi, including variation in pH, temperature, and oxygen levels, as well as nutrient limitation and host immune responses [ 6 , 8 – 10 ]. C. albicans–host interactions are generally commensal, but C. albicans can also act as an opportunistic pathogen, resulting in an estimated 400,000 serious bloodstream infections per year [ 11 – 13 ]. Additionally, C. albicans can cause more minor mucosal infections, including oral and vaginal thrush and skin infections [ 14 ]. As a consequence, C. albicans represents the second most common human fungal pathogen and the most common source of healthcare-associated fungal infections [ 15 ].

Results

Phenotypic characterization of commensal C. albicans strains We sourced C. albicans from oral and fecal samples from undergraduate student donors, using these as a representative sample of colonizing C. albicans strains from healthy individuals. In this population of students, 29% (29/98) were positive for oral Candida colonization and 10% (10/98) were positive for fecal colonization, including 4 students who exhibited both oral and fecal colonization (Fig 1A). From each host and site, we collected every individual colony present on the BD ChromAgar plates and confirmed species identity through ITS amplicon sequencing. Overall, we obtained 910 C. albicans isolates (fecal: n = 84 colonies, oral: n = 826 colonies) (Fig 1B). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 1. Characterization of isolates from healthy donors reveals extensive phenotypic heterogeneity. (A) The number of healthy donors from whom C. albicans colonies were obtained by isolation site with 4 donors exhibiting both oral and fecal colonization. (B) Number of C. albicans colonies isolated per donor from each sample site. Error bars represent median and interquartile range in number of colonies from each positive individual donor. Significance determined by Wilcoxon rank sum test. (C) Strains varied in maximum growth rate and carrying capacity in response to different carbon sources. Growth curves were performed on each isolate under 3 carbon sources at 30°C for 24 h for dextrose and 48 h for glycerol and galactose in biological duplicate. Rate and carrying capacity were determined using the GrowthcurveR analysis package. Error bars represent median and interquartile range of the growth parameter. (D) All strains retained the capacity to filament in liquid inducing cues, but some demonstrated altered morphology and aggregation. Strains were incubated in the indicated conditions and imaged at 40× magnification. Scale = 10 μM. (E) Strains varied in their capacity to invade into solid agar. Colonies were incubated on the indicated conditions for 5 days before gentle washing and imaging for invasion. (F) UMAP plot for strain phenotypic similarity. All growth conditions and invasion phenotypes were nonlinearly projected into 2D space and colored by donor. https://doi.org/10.1371/journal.pbio.3001822.g001 We then performed growth assays on all 910 isolates and the SC5314 reference strain in rich medium (yeast peptone) with dextrose, galactose, or glycerol as the carbon source. We observed primarily unimodal distributions with a long tail of slow-growing strains for both exponential growth (μ max ) and saturating density (K) in rich media (Fig 1C). Although the 5 slowest-growing isolates were all obtained from oral samples, we did not observe a significant difference in growth rate between oral and fecal samples (S1A Fig), and overall, the growth rates between the different carbon sources were not correlated (S1B Fig). As filamentation has been tightly associated with virulence, and because previous work in murine models has suggested that gut adapted strains may lose their ability to filament [21,36,38], we examined each isolate in the collection for their ability to form hyphae using 10% serum and febrile temperatures as 2 inducing cues. Additionally, we examined the morphology of each strain under the non-inducing condition of yeast extract peptone dextrose (YPD) at 30°C to identify constitutive filamentation, as has been observed from isolates collected from sputum samples from cystic fibrosis patients [45]. We observed that while none of the strains were constitutively filamentous, several strains aggregated at YPD 30°C (Fig 1D; represented by 882–46). These aggregative isolates were more likely to reach a lower carrying capacity; however, it is possible that the aggregation interfered with the accuracy of the OD reads. All strains were able to filament in response to the standard laboratory inducing conditions of serum and high temperature, albeit with some variation in the number of hyphal cells, shape, and aggregation of the cells (Fig 1D). Under hyphal-inducing conditions, the aggregating strains formed filaments while the mother yeast cells remained connected, resulting in a star-like pattern. Some isolates showed different filamentation patterns under the different inducing conditions. For example, isolate 813–61 formed filaments that resembled those of SC5314 under 37°C + 10% serum, but under the 42°C inducing cue, this isolate formed fewer filaments with many yeast cells aggregating around the filament. The ability of all our commensal isolates to form filaments may be consistent with the hypothesis that interaction with bacteria in the gut maintains selection for the hyphal program [36,46]. Although there were isolates, such as 813–61, that showed more of a mixture of yeast and hyphae than the SC5314 parent that are potentially consistent with the yeast-locked phenotypes observed in Tso and colleagues (2018) [36], the majority of our commensal strains retained full capacity to filament in our in vitro assays. This may suggest that the antibiotic or germ-free mouse models are not fully recapitulating important features of C. albicans–human interactions. Filamentation programs under solid and liquid growth can involve distinct genetic programs [47], and the human host can present a variety of substrates for C. albicans to utilize. Therefore, we also tested each isolate for its capacity to invade into solid agar media under multiple conditions: YPD agar at 37°C under anaerobic conditions as a minimal inducing cue, YPD agar at 30°C as the baseline condition, YPD agar at 37°C as an intermediate inducing condition, and Spider agar at 37°C as a strong inducing condition (Fig 1D). As expected for a strong inducing cue, the highest degree of invasive growth was observed on Spider agar at 37°C [47]. Interestingly, although all strains showed the ability to form hyphae under liquid culture conditions, many strains failed to invade into solid agar, giving scores of 0 or 1 under multiple conditions (Fig 1E). We also observed substantial phenotypic heterogeneity among strains isolated from the same host, including in strains isolated from the same site, consistent with each individual being colonized by multiple, phenotypically distinct, strains of C. albicans. To examine the relationship between the strains, observed phenotypes, and donors, we generated a uniform manifold approximation projection (UMAP) embedding, which will plot strains with similar phenotypes closer together and strains with dissimilar phenotypes farther apart based on the Euclidean metric (Fig 1F). Using this projection, we saw significant overlap in strain phenotypes from many donors, although there were some donors that had strains that clustered away from others, such as strain 946. Additionally, some donors had 2 separate clusters of strains, such as those from donor 814, 872, and 882. We also observed that some strains from multiple donors clustered together in this projection, including the aggregating strains from donors 811 and 882. However, we did not observe clustering based on sample origin site (S1C Fig). Together, these experiments demonstrate a large variation in phenotypes between commensal isolates, even among traits, such as filamentation and invasion, that are often correlated with virulence.

Deep phenotyping of commensal isolates The set of isolates for sequencing were initially chosen based on variation in growth rate in rich medium and alterations in invasion into agar. However, we hypothesized that we may identify site-specific adaptations, as host sites commonly colonized by C. albicans vary dramatically in environmental cues. Additionally, we hypothesized we may identify phenotypes associated with specific C. albicans clades, as we were able to identify structural variants shared between closely related isolates. To test this, we performed a set of growth analyses under multiple environmental conditions, including pH stresses, nutrient limitation, cell wall stressors, and antifungal drugs (Fig 4A). These analyses produced a dense array of quantitative phenotypic information for each strain. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 4. Deep phenotyping reveals heterogeneity in in vitro and host response phenotypes. (A) Growth curve analysis under multiple environmental conditions. Carrying capacity (K) was normalized to SC5314 and the fold-change plotted by heatmap. Aggregating strains (882–60, 882–46, and 811–7) demonstrate a consistently lower carrying capacity. (B) Relative macrophage phagocytosis rates of commensal isolates to reference strain SC5314. Black data points indicate an oral isolate and pink data points indicate a fecal isolate. Triangles represent bloodstream isolates. (C) Macrophage cell death rates. (D) Representative images of isolates following 4 h macrophage infection. Representative filamentation score of 0 (top). Representative filamentation score of 5 (bottom) and 20× magnification. Scale = 50 μM. For phagocytosis and cell death rates, significant differences from the SC5314 reference strain were determined by one-way ANOVA, with Dunnett’s multiple correction testing. Asterisks indicate P < 0.05 (*), P < 0.005 (**), P < 0.001 (***), and P < 0.0001 (****). https://doi.org/10.1371/journal.pbio.3001822.g004 From these data, we identified 3 strains, 882–60, 882–46, and 811–7, that consistently reached a lower carrying capacity than the wild type under multiple conditions; this is consistent with the aggregative phenotype exhibited by these isolates at 30°C (Fig 1D). These strains all belonged to Clade C and were closely related, despite arising from 2 donors (Fig 4A). Growth rates in the nutrient limitation conditions were generally correlated with each other. However, we did not observe a correlation between body site and growth rate, even in response to cues that would appear to be specific for a particular body site, such as anaerobic growth. Across the commensal isolates, we noted the most variation in growth in response to caffeine and the antifungals fluconazole and rapamycin. In addition to growth, we measured each of the strains for their ability to form biofilms on plastic surfaces [58]. Although we observed variation between the strains, there was no correlation between isolation site or clade with the propensity of isolates to form biofilms (S4 Fig). Our dense array of phenotypic data across 45 C. albicans isolates and 8 clades reveal that commensal isolates largely retain the plasticity to grow efficiently under diverse environmental cues, even those not immediately relevant to their colonizing site, as we did not observe growth enrichment in cues specific to isolation sites. A major stress condition and environmental factor impacting C. albicans in the host is the immune response. Therefore, we moved from pure growth assays to measuring host–microbe interactions, using macrophages as representative phagocytes. We first hypothesized that the oral strains may show decreased recognition by macrophages, as persistent oral isolates were recently shown to result in reduced immune recognition and inflammation in both an OPC model of infection and in cell culture [59]. We tested this by measuring phagocytosis of each strain by immortalized bone marrow-derived macrophages and determining the ratio of internalized to external cells by differential staining and microscopy (Fig 4B) [60]. Although most isolates were not significantly different from the SC5314 reference, the isolates generally had a lower phagocytic rate than SC5314. Additionally, there was no correlation between sample origin site or clade with phagocytosis rate. As phagocytosis was not a major differentiating factor between strains, we then wanted to examine whether the strains would induce different levels of macrophage cell death. We primed bone marrow-derived macrophage for 2 h with LPS before infecting with each of our isolates for 4 h. Following infection, we stained the cells with propidium iodide (PI) as a measure of cell death (Fig 4C). Strikingly, many of the commensal isolates induced less cell death than SC5314. Our results are consistent with previous reports that commensal oral isolates of C. albicans are characterized by a reduced capacity to damage host cells [61]. To test whether this increased killing capacity of SC5314 was shared with other bloodstream isolates, we measured the phagocytosis and cell death rates for 12 bloodstream isolates originally described in Hirakawa and colleagues (2015) [19]. These results indicated a higher rate of phagocytosis for the bloodstream isolates (Fig 4B). However, for most of the bloodstream isolates, we also observed decreased capacity for macrophage killing (Fig 4C). The bloodstream isolate P94015, which was identified by Hirakawa and colleagues as having a loss of function of the transcription factor EFG1, behaved similarly to the other bloodstream isolates in these assays, despite having a decrease in virulence in their animal model [19]. These results indicate that the ability to kill host cells is not conserved across bloodstream isolates. Recently, we showed that C. albicans mutants that filament in serum are not always filamentous within macrophages [62]. As filamentation is linked, but not required, for inflammasome activation within host phagocytes [62–64], and clinical isolates show variability in induction of host inflammatory responses [65,66], we examined the morphology of the commensal isolates after incubation for 4 h with macrophages. We observed considerable variation in the extent of filamentation among the natural isolates (Fig 4D), including strains that completely failed to filament and those that filamented more than the SC5314 reference strain. The aggregative strains, 882–60, 882–46, and 811–7 failed to filament during incubation with macrophages. Notably, the extent of filamentation did not correlate with colony morphology or invasion on agar, with many strains showing invasion into agar but no filamentation inside the macrophage, and vice versa (S5 Fig). Additionally, oral and fecal isolates both demonstrated defects in filamentation in macrophages, and filamentation in macrophages was not predictive of the phagocytic rate or cell death rate. Using individual phenotypic measures, we were unable to identify associations between strains based on body site or donor. However, it is possible that the combined phenotypic and genotypic profile would identify clusters of strains with similar distinct phenotypes or reveal connections between isolates. Therefore, we turned to UMAP embedding, which will plot strains with similar phenotypes closer together and strains with dissimilar phenotypes farther apart based on the cosign metric. We did not observe any obvious clusters that segregated by isolation site, clade, or participant (S6 Fig). In sum, all of the commensal isolates showed extensive phenotypic variation, but this was not dependent on the body site or participant from which they were collected.

Limited diversity exploitation Genome-wide association studies have been a powerful tool for identifying the genetic basis of variation in phenotypes of interest in humans and other recombining species. However, the generally clonal and asexual reproduction of C. albicans creates a population structure that confounds traditional GWAS methods. By sampling multiple isolates from each individual, we were able to obtain phenotypically diverse strains with a limited set of unique SNPs between isolates, allowing us to identify causative variants associated with a particular phenotype. We focused on the 6 strains from donor 814 included in the condensed set, which clustered tightly in Clade 1; we hypothesized this would allow us to identify causative variants associated with particular phenotypes that were divergent between strains. Our agar invasion analysis revealed that isolate 814–168 demonstrated hyper invasion into Spider agar at 37°C, whereas the other 5 isolates from the condensed set were less invasive (Fig 5A). Moreover, from this donor’s 163 total isolates (144 oral isolates and 19 fecal isolates), only this single isolate exhibited the hyper invasive phenotype into Spider agar at 37°C (Fig 5B); this phenotype was the motivation for initially including this strain in the condensed set. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 5. SNV limited diversity exploitation analysis of donor 814 commensal isolates reveals role for Zms1 in regulating agar invasion. (A) Agar invasion images for isolates from donor 814 included in the condensed set. Colonies were grown on YPD or Spider agar for 5 days. Invasion was determined after gentle washing. Isolates in black text are oral samples and isolates in pink text are fecal samples. (B) Agar invasion scores for all isolates from donor 814 under YPD and Spider conditions. (C) Co-expression analysis of Zms1. Width of the lines represents strength of the co-expression score. Gene names and predicted functions from Candida Genome Database. (D) Agar invasion images for allele swap and deletion strains. Colonies were grown on YPD or Spider agar for 5 days before imaging. Invasion was imaged after gentle washing. SNV, single-nucleotide variant; YPD, yeast extract peptone dextrose. https://doi.org/10.1371/journal.pbio.3001822.g005 Variant analysis identified 12 genes with unique SNVs in the 814–168 strain compared with the other 5 sequenced 814 strains, including a heterozygous adenine to thymine SNV in the transcription factor Zms1, resulting in a change in amino acid 681 from a threonine to a serine. Moreover, co-expression analysis [67] of ZMS1 revealed that it is highly correlated with genes involved in regulating the yeast-to-hyphal morphogenic transition (Fig 5C). To test whether this SNV can drive invasion, we generated complementation plasmids encoding each of the alleles from the 2 strain backgrounds and performed allele swap experiments between the high and low invasion strains. In the minimally invasive 814–183 background, replacing 1 copy of the endogenous ZMS1 allele with the ZMS1-T681S allele was sufficient to drive hyper invasion into both YPD and Spider agar (Fig 5D). Similarly, replacing the invasive ZMS1 allele with the ZMS1-S681T allele was sufficient to reduce invasion into both YPD and Spider agar (Fig 5D). Previous work on the function of Zms1 via deletion mutant analysis had not revealed a phenotype [68]; however, this was in the SC5314 genetic background and the impact of a specific transcription factor on a given phenotype can vary depending on the strain [18]. Therefore, we deleted ZMS1 from both 814 backgrounds and tested the strains for invasion and filamentation. On YPD agar, as before, deletion of ZMS1 in both genetic backgrounds had minimal effects, with the mutant strains behaving similarly to their parent strains (Fig 5D). However, on Spider agar, ZMS1 deletion changed the colony morphology from wrinkly to smooth in the 814–168 background, although it did not decrease overall invasion. In contrast, deletion of ZMS1 increased invasion in the otherwise minimally invasive 814–183 background, highlighting the differential impact of ZMS1 mutation in the different genetic backgrounds (Fig 5D). Our results demonstrate that a single SNV changing amino acid 681 to a serine is a dominant negative allele that is sufficient to drive a hyphal invasion program into Spider agar. We also identified natural variation that was distinct from deletion phenotypes. This approach, which we have termed “limited diversity exploitation,” highlights how deep phenotypic analysis of a limited set of natural isolates from a single host can be exploited to identify causative variants and identify new functions for under-characterized genes.

[END]
---
[1] Url: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3001822

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/