FKS1/2 -variant independent mechanisms underlying the emergence of resistance in echinocandin-refractory Candida auris infections

Human fungal pathogens, Candida auris, emerging resistance, antifungal drug 16 tolerance, stress tolerance, echinocandin resistance, population heterogeneity, 17 heteroresistance, cell wall, antimicrobial resistance 18 19 Short Title: Emergence of echinocandin resistance in Candida auris 20 21

22 23 The emerging fungus Candida auris is a drug resistant global public health threat 24 and WHO critical priority pathogen. Recommended first-line invasive candidiasis treatment 25 is echinocandin monotherapy, but C. auris can develop on-treatment resistance via 26 FKS1/2 gene mutations and additional, previously unexplained mechanisms. To better 27 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 2 understand echinocandin failure in C. auris, we sequenced the genomes of echinocandin 28 refractory FKS1/2 wild-type C. auris serial isolates from two critically unwell patients in 29 London, UK. Population analysis profiling revealed echinocandin heteroresistance, and 30 in vitro culture of clinical isolates at supra-MIC concentrations of anidulafungin (8 μg/ml) 31 exhibited morphotypic heterogeneity. Small colony variants (SCVs) and large colony 32 variants (LCVs) showed elevated MICs with polyploidy (to 4n and above) alongside 33 adaptive changes in cell wall β-1,3-glucan content. LCVs contained significantly more 34 mutations in calcineurin-related stress tolerance pathway gene CRZ1 compared to clinical 35 parents and SCVs, associated with further increases in MIC. These findings indicate 36 progressive step-wise accrual of adaptation to echinocandins, including genomic 37 instability, alterations in stress tolerance pathways, and cell wall remodeling, paving the 38 way for resistance emergence. 39 40 41 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 3 Graphical abstract 42 43 44 Graphical Abstract, created with Procreate and BioRender. Illustrations © the authors. 45 46

47 48 Candida auris, also named Candidozyma auris1, is a World Health Organization 49 critical priority human fungal pathogen2. This fungus causes invasive infections 50 (candidiasis) with an associated mortality of up to 45%3, and propensity for resistance to 51 all key antifungal classes: azoles, echinocandins, and polyenes4,5. After detection in Japan 52 in 20086, six clades have emerged near-simultaneously across the world7,8. C. auris 53 colonises the skin, particularly of patients who are critically ill or in long-term ventilation 54 facilities9. Furthermore, C. auris is challenging to eradicate from hospital environments, 55 Bronchial aspirate & blood isolates FKS1/2-variant independent mechanisms underlying the progressive emergence of resistance in echinocandin-refractory Candida auris infections Large colony variants (LCVs) + Anidulafungin Polyploidy Population heterogeneity Escape Small colony variants (SCVs) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 4 resulting in inter-patient spread within and between hospitals10–13. Colonisation and 56 infection rates continue to rise (in Europe14, North America15, South America16, Africa17, 57 and Asia18) with a recent resurgence in UK cases prompting the inclusion of C. auris in the 58 list of notifiable organisms, updates to guidelines for healthcare settings, and the 59 declaration of a national public health incident in 202519,20. 60 61 As the majority of C. auris isolates are resistant to fluconazole (>97% for clades I 62 and III) and a high proportion to amphotericin B (47% for clade I)21, echinocandin 63 monotherapy is currently the initial treatment recommended for invasive infections5,22,23. 64 Initially, echinocandin resistance was uncommon across clades I-IV (0-9%)21, but cases 65 continue to emerge. Echinocandins target the essential cell wall enzyme β-glucan 66 synthase, encoded by FKS1 and FKS2 in C. auris24. Resistance is normally defined as 67 growth above a minimum inhibitory concentration (MIC) using epidemiologically-derived or 68 clinical breakpoints based on correlation with clinical outcomes25,26. Typical resistance is 69 thought to arise primarily through mutations in FKS1 hot-spots27–31 that lead to treatment 70 failure in vivo, as demonstrated in murine infection models32,33. Several clinical case 71 studies have shown resistance arising due to FKS1 mutations: in a five isolate series from 72 a single patient over 1 year, only the terminal isolate contained the F635Y FKS1 variant 73 and was echinocandin resistant34. In a nineteen isolate series from a single patient over 72 74 days, F639Y/F635C variants emerged alongside resistance to four classes of antifungal 75 (pan-resistance) with further mutations in genes related to azole, polyene, and flucytosine 76 resistance35. 77 78 Echinocandin resistance can also be caused by FKS1/2-independent mechanisms 79 that either promote resistance directly, or that lead to FKS mutations. In four cases of 80 urinary infection, FKS1 mutation was necessary for echinocandin resistance, but other 81 mutations related to cell wall stress and DNA repair/chromatin remodeling were present in 82 strains with elevated MICs36. In vitro micro-evolution experiments have shown additional 83 mutations outside FKS1 hot-spots, including in ERG3, associated with echinocandin 84 resistance37,38. Large scale genome-wide association studies have suggested that 85 mutations in other cell wall-related genes (IFF4, FCR1 and GWT1) may promote 86 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 5 echinocandin resistance39. In Candida albicans40 (and other fungal pathogens41,42), 87 aneuploidy and copy number variation can also lead to echinocandin resistance. Such 88 structural variation has not yet been described in C. auris with respect to echinocandin 89 resistance43. Furthermore, the development of secondary echinocandin resistance, and 90 even pan-resistance to azoles, polyenes, echinocandins and flucytosine, has been 91 reported in patients receiving treatment for clinically refractory infection, described in as 92 many as 3% of clade I isolates15,21,35,44,45. The step-wise mechanisms underlying the 93 emergence of echinocandin resistance remain poorly understood and are urgent questions 94 for C. auris therapy in light of such limited options43,46. 95 96 In addition to standard antifungal drug resistance, heteroresistance and tolerance 97 phenomena can contribute to population heterogeneity, stress adaptation, and treatment 98 failure in infections caused by other pathogenic Candida species47, including 99 C. parapsilosis48, C. glabrata49 and C. albicans50. Heteroresistance describes a small 100 intrinsically resistant sub-population of cells (<1%) that is selected for and expands under 101 drug pressure47,48,51. Tolerance represents the ability of a larger sub-population of cells 102 (10-50%) within an isogenic, drug susceptible population (using MIC) to survive and grow 103 slowly (>24 h) at concentrations above the MIC47. In other pathogenic Candida species, 104 echinocandin tolerance has been shown to arise via increased chitin synthesis and cell 105 wall remodeling triggered by protein kinase C (PKC), Hog1, Hsp90 and calcineurin 106 signaling52,53. In C. auris, echinocandin exposure in vitro can lead to the increased 107 expression of genes involved in cell wall synthesis and remodeling including genes 108 encoding chitin synthases, cell wall adhesin Als5 and the drug efflux pump Cdr124,54,55. 109 These transcriptional changes correlate with echinocandin-induced adaptations such as 110 cellular adhesion, aggregation, and biofilm formation54–56. 111 112 Here, we explore underlying mechanisms associated with the emergence of 113 echinocandin refractory infection in two critically ill patients with C. auris bloodstream 114 infections. Our data provide evidence for the sequential accumulation of a series of 115 physiological and genetic adaptations leading to the emergence of resistance and 116 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 6 echinocandin treatment failure. 117 118

119 120 FKS1 mutations and aneuploidy are insufficient to account for echinocandin 121 refractory C. auris infection 122 123 We identified C. auris isolates from two patients with bloodstream infections (BSI) 124 refractory to echinocandin therapy treated in the intensive care units of St George’s 125 (isolates StG1-5, Figure 1A), and King’s College Hospitals (isolates K1-2, Figure 1B) in 126 London, UK. The first patient suffered persistent candidaemia for three successive days on 127 treatment, whilst the second patient experienced an initial BSI, followed by breakthrough 128 infection on echinocandin treatment after 33 days. Antifungal susceptibility testing 129 performed by the UK Mycology Reference Lab in Bristol suggested that these fungal 130 isolates were not echinocandin resistant (Figure 1C) based on the tentative CLSI 131 breakpoint (≥4 μg/mL) for anidulafungin in C. auris57. However, anidulafungin monotherapy 132 failed to clear the fungal infection and both patients required a switch to amphotericin B-133 based combination treatment and subsequently died. Therefore, we sequenced the 134 genomes of these C. auris isolates to explore the basis of their lack of response to 135 echinocandin therapy. 136 137 To understand the genetic mechanisms underpinning echinocandin therapy failure 138 in these two patients, we sequenced the genomes of their C. auris isolates. Whole 139 genome sequencing revealed that each of the isolates belonged to clade I (Figure S1) 140 and, surprisingly, no non-synonymous mutations were present in FKS1/2 genes. 141 Therefore, we sought to understand additional pathogen-related mechanisms by which 142 these infections were refractory to treatment. We identified 2,567 single nucleotide 143 polymorphisms (SNPs) compared to the B8441 clade I reference genome, including 2,104 144 in both StG and K series, of which 198 were present only in StG isolates, and 264 only in 145 K isolates. Aside from SNPs in intergenic regions (71.0%), there were a total of 469 non-146 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 7 synonymous variants, 6 nonsense variants, and 66 indels in coding regions (Figure 2A-B). 147 The greatest variation across clinical isolate series was observed in HYR3 148 (B9J08_004100: one deletion, four non-synonymous variants, and nine synonymous 149 variants), which is predicted to encode a GPI-anchored cell wall protein58 (Figure 2A-B). 150 We also identified mutations in genes associated with cell wall stress response (BCK1), 151 cell wall synthesis (KAR2, RBR3, SCF1), secretory pathways and intracellular trafficking 152 (GDI1, SVL3, VPS5, YPT6), RNA synthesis (HAS1, IWR1), and histone acetylation 153 (SPT10), suggesting genetic adaptations involving cell wall remodeling in response to drug 154 pressure (Table S1). SNPs resulting in non-synonymous mutations were identified in 155 genes related to azole resistance: ERG11 (K143R), CDR1 (V704L) and TAC1b (A640V) in 156 isolates StG1-5; and ERG11 (Y132F), CDR1 (E709D) and TAC1b (A583S) in isolates K1-157 2. No non-synonymous mutations were observed in MDR1-2, MRR1a-b, CAS5, or FUR1. 158 159 We examined the genomes of the StG and K isolates (using thresholds of >1.4 and 160 <0.6 normalised depth of coverage, Figure S2A-B) to identify gene or chromosome copy 161 number variation (CNV) associated with echinocandin refractory infection. Though we did 162 not identify aneuploidy, small regions showing increased per-gene CNV included loci such 163 as NTO1 (part of putative histone acetyltransferase machinery) and FGR14 (homologue of 164 a retroviral endonuclease-reverse transcriptase). Areas of low copy number include 165 multiple genes (e.g. CDC13, ECM21, ECM42, HEM14, IWR1, MRF1, MRP49, PRD1, 166 RCO1, TMA17). To our knowledge, none of these CNV-affected genes have been 167 associated with echinocandin resistance (Table S2). 168 169 C. auris isolates display echinocandin heteroresistance and rapidly develop further 170 resistance in vitro 171 172 Based on CLSI breakpoints, both StG and K isolates were echinocandin sensitive 173 and yet obtained from clinically refractory infections. Therefore, we tested whether initial 174 StG1 and K1 isolates displayed anidulafungin heteroresistance using population analysis 175 profiling (PAP, Figure 3A). Heteroresistance was indicated by the growth of small sub-176 populations of cells (<0.01%) on YPD-agar at anidulafungin concentrations 256-fold 177 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 8 greater than those for the susceptible majority (Figure 3A). After extended culture (144 h), 178 two distinct colonial morphologies emerged that were most pronounced on YPD-agar 179 containing 8 μg/ml anidulafungin (Figure 3B): slower growing “small” colony variants 180 (SCVs), and faster growing “large” colony variants (LCVs). 181 182 We compared the phenotypes of SCVs and LCVs: fluorescence microscopy 183 revealed that, in the presence of anidulafungin, both SCVs and LCVs contained 184 moderately larger cells than the parental isolate alongside aggregation and DNA content 185 increase on SYBR green staining (Figure 3C). The SCV and LCV morphotypes persisted 186 when they were re-plated onto drug-free plates: re-plated SCVs (mean colony size 187 0.58 mm3) formed significantly smaller colonies (p < 0.0001) than LCVs (mean colony size 188 3.58 mm3) and parent isolates (mean size 2.45 mm3) following 48 h growth at 37 °C 189 (Figure 3D-E). SCVs and LCVs grew at similar rates to parental isolates in drug-free liquid 190 YPD. However, in YPD containing 8 μg/mL anidulafungin, SCV cells grew significantly 191 faster than parental cells, and LCVs even faster than SCV cells (Figure 3F). Furthermore, 192 both SCVs and LCVs were resistant to anidulafungin according to updated EUCAST 193 criteria for C. auris (sensitive, S: ≤ 0.25 μg/mL; resistant, R: >0.25 g/mL25,26): MIC against 194 anidulafungin was highest for StG1 and K1 LCVs (modal MIC 4-8 μg/ml, range 2-195 16 μg/ml); followed by SCVs (modal MIC 2 μg/mL, range 1-4 μg/mL) and parental isolates 196 (modal MIC 1-2 μg/mL, range 1-4 μg/mL, Table S3). SCVs also exhibited significantly 197 higher tolerance to anidulafungin compared to the parent isolate, measured using both 198 supra-MIC growth (Figure 3G) and confirmatory Etest inhibition strips (Figure 3H). LCVs 199 were resistant according to Etest (Figure 3H). According to the updated EUCAST 200 guidelines, parental isolates were also resistant. 201 202 Taken together, these findings indicate the presence of stable heterogeneous sub-203 populations within treatment-refractory C. auris clinical isolates, which are consistently 204 isolated at high anidulafungin concentrations. These distinctive small and large colonial 205 morphotypes appear to reflect a spectrum of drug-adapted sub-populations, as shown by 206 increased growth rates in the presence of drug, higher tolerance and anidulafungin MICs- 207 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 9 compared to the parent (rather than a difference in actual cell size). 208 209 CRZ1 mutations in LCVs are associated with anidulafungin resistance 210 211 We explored the genomic foundations of the colony morphotypes with associated 212 tolerance and resistance phenotypes by sequencing SCV (n = 18) and LCV (n = 17) 213 colonies. Mutations potentially underlying the observed population heterogeneity were 214 identified by comparing their genomes with those of parental controls: StG1, K1 and K2 215 grown on YPD, with additional SCV/LCV isolates sequenced directly on anidulafungin at 216 8 μg/ml (n = 3 per isolate). No FKS1/2 mutations were observed in any of the SCV or LCV 217 genomes, indicating that anidulafungin resistance had emerged via FKS1/2-independent 218 mechanisms. 219 220 To seek these FKS1/2-independent mechanisms, we examined sequence variants 221 displaying a frequency difference of >25% between groups (e.g. SCV vs LCV, Figure 222 S3A) and compared the number of strains in each group that displayed variants (Figure 223 S3B-C, Supplementary Note 1). Strikingly, variation in the CRZ1 gene was significantly 224 more common in LCVs compared to parent isolates (14/17 vs 0/7, 82.3% vs 0%, adjusted 225 p-value = 0.032) and between LCVs and SCVs (2/18, 11.1%, adjusted p-value = 5.93 x10-226 5). A total of 16 non-synonymous mutations were identified in CRZ1, excluding S237Y 227 which was present in all strains (Figure 4). The majority of CRZ1 mutations were highly 228 likely to alter function, including nonsense (n = 8) and frameshift deletions (n = 3), 229 compared to non-synonymous mutations (n = 5). Thus LCVs, which were anidulafungin 230 resistant, carried the largest number of CRZ1 mutations (Figure 3F), consistent with the 231 recent finding that C. auris crz1 knockout mutants are anidulafungin resistant59. 232 233 The CRZ1 calcineurin-responsive transcription factor gene was one of two genes in 234 the calcineurin pathway that displayed highly significant variation in SCVs and LCVs. 235 Another calcineurin-pathway related gene (RCN2, regulator of calcineurin) was 236 significantly enriched for associated intergenic SNPs that flanked this gene in LCVs 237 (88.2%) vs SCVs (27.8%, adjusted p-value 0.017). Furthermore, four frameshift mutations 238 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 10 in VPS5 (predicted vacuolar sorting protein) were identified in fourteen strains, but these 239 were not significantly enriched in SCVs or LCVs relative to the parental isolates. However, 240 two non-synonymous variants were significantly enriched in daughters compared to clinical 241 parents: KAR2 V410I (p = 0.0077) and HP_4105 R202P (p = 0.033). The change in Kar2 242 (a member of the Hsp70 chaperone family) could conceivably contribute to the tolerance 243 profile of C. auris through stress response adaptations. However, the function of HP_4105, 244 and hence the potential impact of the R202P mutation, remains obscure. 245 246 Anidulafungin induces cell wall remodeling in both SCVs and LCVs 247 248 In C. albicans, the inhibition of β-1,3-glucan synthesis by echinocandins induces 249 compensatory increases in chitin synthesis and cell wall remodeling60, in part via Crz1 250 signaling61. In C. auris, caspofungin induces chitin synthase gene expression in a CRZ1-251 dependent fashion59,62. Therefore, we examined the impact of anidulafungin on the cell 252 walls of LCVs and SCVs. Flow cytometry of Calcofluor-White (CFW) stained cells revealed 253 that chitin levels increased in response to anidulafungin in both LCVs and SCVs 254 (Figure 5A), with associated increases in β-1,3-glucan exposure (Figure 5B). 255 Transmission electron microscopy (TEM) showed corresponding changes in cell wall 256 architecture; exposure to anidulafungin led to significant thickening of the inner (chitin-rich) 257 cell walls of both LCVs and SCVs (Figure 5C-E). Anidulafungin also induced a slight 258 increase in the outer mannan layer of the cell wall, but this was minor compared to the 259 dramatic changes to the inner layer. High pressure ion chromatography (HPIC) of cell wall 260 carbohydrates provided further evidence of anidulafungin-induced increases in 261 chitin/mannan content with corresponding decreases in glucan both across both the StG 262 clinical isolates (Figure 5F-H) and in SCVs, but not in LCVs (Figure 5I-K). 263 264 Clinical and drug-tolerant/drug-resistant sub-populations of C. auris are 265 characterised by alterations in ploidy 266 267 Stress-induced changes in ploidy have been proposed to precede the emergence in 268 drug resistance in C. albicans63,64, and tetraploidy has been described in C. albicans 269 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 11 clinical isolates from human hosts65–68. Therefore, we compared the ploidies of C. auris 270 StG1 and its SCV and LCV daughters, analysing cells taken directly from colonies on 271 plates using flow cytometry: StG1 cells from YPD plates, and SCV and LCV cells from 272 were taken directly from YPD plates containing 8 μg/ml anidulafungin. Saccharomyces 273 cerevisiae haploid and diploid strains were used as controls. The S. cerevisiae controls 274 showed clean 1n-2n and 2n-4n ploidies, as expected for mixed populations containing 275 cells pre- and post- S-phase (Figure 6A-B). Although C. auris is purportedly haploid69,70, 276 StG1 cells displayed heterogeneous ploidies ranging from 1n to >4n (Figure 6A-B). Even 277 higher ploidy ranges, with a higher proportion of cells with ploidies of >4n, were observed 278 for the SCV and LCV morphotypes isolates from YPD+anidulafungin. Our bioinformatic 279 analyses, which included allele frequency tallies (Figure S4A), were consistent these 280 increased ploidies, though identification of heterozygous sites was not sensitive to inter-281 strain differences (Figure S4B). 282 283 The heterogeneous ploidy of StG1 cells was unexpected (Figure 6A-B). Therefore, 284 to test whether other C. auris isolates display this phenotype, we examined isolates from 285 clades I to V. Ten epidemiologically divergent isolates all displayed heterogeneous 286 ploidies when grown on YPD agar (Figure S4C) with strict single-cell gating strategies 287 (Figure S4D). These findings suggest that genomic instability and polyploidy are features 288 of C. auris that might underlie this species’ phenotypic heterogeneity and propensity 289 towards drug tolerance/resistance, promoting treatment-refractory infection and 290 emergence of secondary echinocandin resistance. 291 292

293 294 In this study we investigated serial invasive C. auris isolates from critically ill 295 patients, which did not respond to anidulafungin despite the absence of FKS mutations. 296 We have identified the presence of phenotypically heterogeneous sub-populations in 297 C. auris, which display distinct colony morphotypes that show increased ploidy and 298 enhanced adaptation to drug stress, relative to the parent clinical isolate. The larger colony 299 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 12 variant (LCV) grows faster in the presence of anidulafungin, has an elevated anidulafungin 300 MIC, and carries significantly more CRZ1 mutations than the small colony variant (SCV) 301 and parental isolate. Whilst C. albicans and C. glabrata crz1 mutants display echinocandin 302 sensitivity71–73, a recent study has suggested a C. auris crz1 null mutant is anidulafungin-303 resistant59. Consistent with this, CRZ1 mutations observed in LCVs were nonsense/stop or 304 deletion/frameshift mutations, which likely result in Crz1 disruption/dysfunction, and these 305 were associated with increased anidulafungin resistance. CRZ1 is likely to promote the 306 fitness of C. auris under drug pressure, but C. auris crz1 knockout strains do not display 307 attenuated virulence in a murine model of systemic candidiasis59. 308 309 Based on our findings, we propose that the following mechanisms contributed to the 310 failure of anidulafungin therapy in the patients in our study. A small sub-population of drug-311 tolerant C. auris was able to survive high supra-MIC concentrations of anidulafungin 312 (giving rise to SCVs). This sub-population of cells was able to adapt to anidulafungin by 313 elevating chitin synthesis and thickening their inner cell walls, which may have contributed 314 to the drug tolerance/intermediate resistance phenotype of SCVs. We reason that the 315 survival of these cells under drug pressure then enabled the emergence of genetic 316 resistance, in part via CRZ1 escape mutations, to yield the faster growing, anidulafungin 317 resistant LCVs. The Crz1 transcription factor regulates cell wall genes in many fungi73, and 318 therefore the loss of Crz1 functionality likely contributed to the observed blunting of the cell 319 wall remodeling response of LCVs to anidulafungin. 320 321 Previous reports have suggested that echinocandin resistance in C. auris primarily 322 arises via mutations in FKS1/2 hot-spots in vivo27–31. We did not identify FKS1/2 mutations 323 in clinical isolates or sub-populations of SCVs/LCVs. Instead, we observed a 324 predominance of HYR3 variation within the clinically evolved case series. HYR3 encodes a 325 predicted GPI-anchored cell wall protein that is under selection in BSI-causing C. auris 326 clades I, III and IV58. Also, HYR3 was highly upregulated in a murine catheter infection 327 biofilm model in vivo74, and in mature biofilms with coincident echinocandin resistance75. 328 Our findings reinforce the view that this locus is significant during human infection. 329 330 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 13 Variable ploidy is likely to be an important aspect of generating phenotypic 331 heterogeneity in C. auris, and is perhaps an overlooked feature of genomic plasticity in this 332 pathogen. Reports of polyploidy in C. auris are rare70, though diploid isolates have been 333 reported from clades I and III, which are associated with higher virulence in a murine 334 systemic infection model76. Polyploidy can accelerate genomic evolution and is viewed as 335 a common and reversible fungal stress response that can increase the availability of 336 beneficial mutations41,42,65,77–79 and even potentially increase virulence63,80. We provide 337 evidence for drug-induced increases in ploidy in treatment-refractory C. auris isolates, 338 strongly suggesting that changes in ploidy promote therapeutic escape by this important 339 fungal pathogen69,70. 340 341 In light of our findings, we suggest updates to the three-stage model for the 342 development of echinocandin resistance in Candida species36,81. We propose that the 343 emergence of resistance begins with echinocandin tolerance driven by reversible 344 physiological changes, for example via protein kinase C, calcineurin-Crz1, HOG and 345 Hsp90 signalling82. Then, the exposure of sub-populations of physiologically relatively 346 tolerant cells to sub-MIC echinocandin concentrations, especially in difficult-to-penetrate 347 sites (e.g. catheter-related biofilms or intra-abdominal compartment, as was the case for 348 our two critically ill patients), selects for the progressive emergence of resistance. Here, 349 drug-induced increases in C. auris ploidy promote the accumulation of escape mutations. 350 Additional drug pressure would then select for further mutations, for example in FKS1 351 and/or CRZ1, that promote increased resistance and faster growth, thereby yielding the 352 LCV phenotype. This updated model can account for the observed population 353 heterogeneity of clinical isolates from patients that have undergone protracted drug 354 treatments, where parental cells coexist with SCVs and LCVs displaying varying degrees 355 of tolerance and resistance and may represent parallel pathways towards overcoming 356 antifungal pressure. This update may also inform the developing conceptualisation of bet-357 hedging in fungal pathogens83. 358 359 Our investigation of fungal population heterogeneity in patient isolates has revealed 360 progressive pathways towards the emergence of echinocandin resistance in C. auris that 361 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 14 may otherwise have been overlooked. The causes of treatment failure are complex, 362 involving also drug delivery to the site of infection, as well as host immunosuppression 363 associated with critical illness. Nevertheless, defining the mechanisms that underlie the 364 emergence of drug tolerance and resistance in fungal pathogens, together with the 365 temporal dynamics of their impacts in vivo, may ultimately lead to improved patient 366 outcomes. Therapeutic strategies could be tuned in real time, to minimise the development 367 of drug tolerance or resistance in an individual patient47,84. Combination therapy is one 368 such powerful option that could enhance efficacy, addresses intrinsic fungal population 369 heterogeneity, and reduce the inherent risk of resistance selection by currently 370 recommended echinocandin monotherapy85,86: for example, the ongoing Wellcome-Trust 371 funded COMBAT Candida clinical trial will compare micafungin alone to combination with 372 micafungin and flucytosine for treatment of candidaemia in a setting of high C. auris 373 prevalence in South Africa, incorporating resistance emergence as an endpoint. 374 375 In summary, we have uncovered pathogen-based mechanisms whereby C. auris 376 sub-populations adapt to first-line therapy, contributing to persistence and evolution. The 377 identification and association of CRZ1 mutation with resistance provides a valuable 378 candidate locus for future functional investigation, potentially serving as an early marker of 379 resistance evolution. Future research should focus on validating CRZ1 mutations by using 380 structural biology, gene editing, and population genetics approaches. The clinical 381 implications of our findings are that treatment refractory C. auris infections require a more 382 nuanced approach to resistance beyond MICs, and novel therapeutic approaches are 383 needed to address the array of adaptive mechanisms that C. auris displays in the face of 384 echinocandin pressure. 385 386

387 388 DNA Extraction: Clinical isolates were confirmed as C. auris using MALDI-TOF 389 and tested for MIC by CLSI standards, stored at -80 °C in YPD containing 20% glycerol, 390 and sub-cultured at 30 °C on YPD agar (Sigma Aldrich, UK). Yeasts were grown overnight 391 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 15 in 10 ml YPD at 30 °C before centrifugation at 13,000 rpm, resuspension in 3 ml water and 392 transferring into a 2 ml tube. Cells were centrifuged at 13,000 rpm for 5 min. prior to the 393 supernatant being removed and cells resuspended in the residual liquid. Cells were snap-394 frozen by immersion in liquid nitrogen for 1 min. and then transferred into a 65 °C water 395 bath for 3 min. Disruption buffer included 400 mg of glass beads, 200 μl of DNA extraction 396 buffer (2% Triton X-100, 1% SDS, 100 mM NaCl, 1 mM EDTA, 10 mM Tris HCl (pH 8)), 397 and 200 μl of phenol:chloroform:isoamyl alcohol (25:24:1), which was added to yeast 398 samples, which were vortexed for four rounds of 20 s with 1 min. breaks on ice to cool 399 samples. A sample of 200 μl 1x TE was added to the mixture before vortexing again for a 400 few seconds and centrifugation at 13,000 rpm for 10 min. The aqueous layer was 401 transferred into a fresh tube and mixed with 1 ml of 100% ethanol and centrifuged at 402 13,000 rpm for 5 min. The pellet was resuspended in 0.4 ml TE and 3 μl of 10 mg/ml 403 RNAase A and incubated for 15 min. at 37 °C. Samples of 10 μl of 4 mM ammonium 404 acetate and 1 ml of 100% chilled ethanol were added. The mixture was centrifuged at 405 13,000 rpm for 10 min. at 4 °C. The supernatant was discarded, and pellet resuspended in 406 100% chilled ethanol. The mixture was spun at 13,000 rpm at 4 °C for 10 min. The 407 supernatant was discarded, and the pellet was dried in a heat block at 65 °C for 10 min. 408 DNA was resuspended in 50 μl water. DNA concentration was checked using Nanodrop 409 and DNA quality by gel electrophoresis. 410 411 Sequencing and variant calling: Sequencing was performed by the University of 412 Exeter Sequencing Service (ESS) in the on the NovaSeq 6000 with SP flow cell (Illumina, 413 San Diego, USA). Quality control was performed with MultiQC v1.10.187. Variant Calling 414 was performed using the Genome Analysis Toolkit (GATK) v4.1.2.088 with alignment of 415 raw sequences to the B8441 v2 reference genome (GCA_002759435.2)7 and the 416 mitochondrial genome (NC_053321.1)89 using BWA-MEM v0.7.1790. HaplotypeCaller was 417 executed in GVCF mode with the haploid ploidy flag. Hard filters were used to remove 418 spurious variants, including the filters quality by depth (QD) 60.0 419 and root mean square mapping quality (MQ) <40.0. 420 421 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 16 Population analysis profiling (PAP): Overnight cultures of C. auris isolates were 422 serially diluted 10-fold 6 times in sterile distilled water. 5 μl aliquots of each serial dilution 423 were spotted in triplicate onto YPD agar plates infused with range of anidulafungin 424 concentrations (0-8 μg/ml). Plates were incubated at 30 °C for 24-48 h and CFU counting 425 performed to calculate CFU/ml able to grow on each drug concentration, which was then 426 normalised to CFU/ml on drug free plate to calculate proportion of the population able to 427 grow at each anidulafungin concentration. Due to noticeable heterogeneity in colony size 428 at the maximum anidulafungin concentration (8 μg/ml), extended incubation of individual 429 8 μg/ml PAP plates inoculated with 107 cells was performed. Plates were incubated at 430 30 °C for 7 days, and images taken after 48 h, 96 h and 144 h. Colony sizes were 431 measured manually using ImageJ at each timepoint. Briefly, images were converted to 432 black and white, sharpened, and then threshold changed to remove parts of the plate with 433 no colony growing. To assess stability of SCV/LCV phenotypes, replica plating of isolates 434 was performed by preparing inoculum of SCV/LCV in distilled water and then plating 102 435 cells on drug free and 8 μg/ml anidulafungin-containing YPD agar. Plates were incubated 436 at 30 °C for 48 h and colony sizes were calculated again using ImageJ as described 437 previously. 438 439 Growth curves: Microdilution plates were set up to contain 0, 0.25, or 8 μg/ml 440 anidulafungin. Wells were inoculated with 1x105 log phase cells such that the final volume 441 was 200 μL/ well. Plates were incubated at 30 °C for 50 h and constantly shaken at 442 100 rpm. OD600 was measured using a Spectrostar Nano every 10 min. for 50 h. Prior to 443 OD600 reading, the plates were shaken at 600 rpm. 444 445 Susceptibility and tolerance assays: Clinical isolates were tested at the point of 446 isolation using CLSI methodology confirmed at St. George’s and King’s College Hospitals 447 with confirmation by the National Reference Laboratory in Bristol. Additionally, the 448 susceptibility of isolates to anidulafungin was determined per EUCAST guidelines with 449 slight modifications. Briefly, 96 well plates were prepared to contain two-fold dilutions of 450 anidulafungin (Gereon B1224) with the first and last well containing no drug. Isolates were 451 defrosted onto YPD agar and grown at 37 °C for 24 h. Five colonies were selected and 452 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 17 resuspended in PBS and diluted such that 105 cells were added to each well in a total 453 volume of 200 μL/well except the last. Susceptibility in RPMI 1640 media supplemented 454 with 2% dextrose was tested at 37 °C. OD450 was read using a Tecan plate reader. The 455 tentative clinical breakpoints were used to determine whether an isolate was susceptible or 456 resistant. Plates were also read at 48 and 72 h to determine supra-MIC growth (SMG). 457 SMG was determined as average 72 h growth in the wells above MIC divided by the 458 growth in the no-drug well. For Etest strips, isolates were grown for 48 h on YPD plates at 459 37 °C. Three to four colonies were taken and used to prepare 0.5 Mcfarland suspension. 460 Sterile cotton swabs were used to streak each suspension across YPD plates in 3 461 separate directions to cover the whole plate in cells, and plates were allowed to dry. Sterile 462 forceps were then used to place an Etest strip onto each plate, ensuring no air bubbles 463 were present. Inoculated plates were incubated at 37 °C and were read and photos taken 464 at 24 h and 48 h. Readings were made according to CDC guidelines for interpretation of 465 Etest antifungal susceptibility testing results91. 466 467 Phylogenomics: FastTree v2.1.11 with Jukes-Cantor modelling92 was used to 468 construct phylogenetic trees based on multiple sequence FASTA alignments produced by 469 ECATools93 using default parameters; midpoint rooted trees used 1,015 (StG clinical 470 series plus laboratory evolved daughters) or 1,053 (K clinical series plus laboratory 471 evolved daughters) phylogenetically informative sites. Additional sequences used to 472 contextualise clinical samples within clades were obtained from the largest global genomic 473 epidemiology study to date21 and the first five isolates were called and used for 474 phylogenetic reconstruction, for which 192,032 sites were entirely covered in all. 475 476 Statistical genomics and annotation: Copy number variation estimation and 477 depth-of-coverage plots were calculated in comparison to normalised depth of coverage 478 across the whole genome compared to the depth of coverage across either all positions 479 for each gene locus or each 10 kb sliding window, from pileups created with Samtools 480 mpileup94. Significance testing for variants/loci/functional annotations was performed using 481 Fisher’s exact test (two-sided) with Benjamini-Hochberg (BH) testing for multiple correction 482 with a cut-off of 0.0595. Significance testing for copy number variation was performed with 483 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 18 Student’s T-test (two-sided), also with BH testing, but with a cut-off FDR of 0.005. 484 Functional annotations and gene names were imported from prior analyses as described 485 previously96. To identify the promoter site for Crz1 binding upstream of FKS1, we used 486 Yeastract+97 to ascertain the presence of GNGGCKCA98 at contig PEKT02000002.1 487 position 1009439-1009439 (893-900 bases upstream from FKS1 start) as the putative 488 binding site for Crz1 (GAGGCCGCA). Contig edges were inferred using the B8441 v3 489 assembly, which demonstrates seven contigs99. To test for polyploidy, variants were re-490 called using GATK with the diploid flag to count heterozygous positions per 10 kb. Allele 491 frequency counts were calculated from mpileups derived as above. 492 493 Quantification of total chitin and β-glucan exposure: The chitin content in the 494 cell walls of C. auris isolates and their segregants was compared using previously 495 described methods100. Cells were grown in YPD at 30 °C for 5 h with or without 8 μg/ml 496 anidulafungin, fixed with 50 mM thimerosal, and stained with 10 μg/mL Calcofluor-White 497 (CFW) in the dark for 60 min. Stained cells were washed twice with PBS, and their 498 fluorescence quantified using an Attune NxT flow cytometer. The plots represent three 499 biological replicate experiments, in each of which 10,000 events were acquired. As a 500 negative control, cells were treated as above but without the addition of CFW. Median 501 Fluorescence Intensities (MFI) were determined using FlowJo v.10 software. The 502 exposure of β-glucan at the C. auris cell surface was quantified by flow cytometry as 503 described previously101. Cells were fixed overnight with thimerosal and stained with Fc-504 Dectin-1 and anti-human IgG linked to Alexafluor 488 (Jackson ImmunoResearch, Ely, 505 UK). The fluorescence of 10,000 cells per condition was assayed using an Attune NxT flow 506 cytometer. Median Fluorescence Intensity (MFI) was quantified using FlowJo v.10 507 software, and fold changes in β-1,3-glucan exposure were calculated relative to the 508 control. 509 510 Transmission electron microscopy: Transmission electron microscopy (TEM) 511 was performed as described previously101. Cells were subjected to high pressure freezing 512 and freeze substitution102, and fixed and stained with 1% osmium tetroxide and 0.5% 513 glutaraldehyde. Ultrathin sections (60 nm) were prepared using lead citrate for contrast 514 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 19 and imaged (JEOL 1400 JEM transmission electron microscope with ES1000W Gatan 515 CCD camera). Cell wall sections were imaged at a nominal magnification of x100k. For 516 each condition, approx. 30 cells were imaged and 10 measurements of inner and outer 517 cell wall diameter taken for each cell using the line tool in ImageJ103. 518 519 Cell wall carbohydrate analysis: Determination of cell wall mannan, chitin, and b-520 glucan content was achieved by acid hydrolysing the polymers, and quantifying mannose, 521 glucosamine, and glucose content, respectively, by high-performance anion-exchange ion 522 chromatography with pulsed amperometric detection (HPIC) as previously described104. 523 524 Ploidy testing: The ploidy of cells in C. auris colonies growing on YPD plates 525 containing 8 μg/ml anidulafungin was assayed by flow cytometry using previously 526 described procedures69. Using a toothpick, cells were harvested directly from individual 527 colonies and fixed overnight in 70% ethanol. The next day, cells were harvested by 528 centrifugation (8,000 rpm, 5 min.) and resuspended in 50 mM sodium citrate, pH 7.5 and 529 incubated with RNase A (250 µg per 107 cells) and proteinase K (1000 µg per 107 cells) for 530 4 h at 37 °C. Cells were then washed in PBS, resuspended in 0.25% Triton-X 100 (Sigma-531 Aldrich) and stained with SYBR Green I (1:500; Sigma-Aldrich) overnight at 4 °C 532 overnight. Before flow cytometry, samples were sonicated and washed with PBS. Flow 533 cytometry was performed on an Attune NxT flow cytometer using an excitation wavelength 534 of 488 nm. SYBR Green I fluorescence was detected with a 530/30 band pass filter and 535 50,000 events, gated for single cells, were recorded for each sample. The proportions of 536 individual cells that displayed ploidies of 1n, 2n, 3n, 4n or >5n were quantified using 537 FlowJo v.10 software, using isogenic haploid and diploid colonies of S. cerevisiae grown 538 on YPD plates as controls: W303-1B a (MATa, ade2, his3, leu2, trp1, ura3); W303-1B 2M 539 (MATa/MATα, ade2/ade2, his3/his3, leu2/leu2, trp1/trp1, ura3/ura3). The cells were also 540 examined by fluorescence microscopy, staining them as described above with the addition 541 of CFW (25 µg/ml) for 5 min. Cells were imaged using a DeltaVision Elite fluorescence 542 microscope with a 60 x objective. Fluorescence excitation was generated by a 543 LumencorLED light source and 10 μm Z-stacks of 50 images were captured by a pco.edge 544 sCMOScamera. The 3D stacks were then deconvolved to remove out of focus light and 545 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 20 maximum intensity projections used to create a 2D image by Image-J v1.54. 546 547 Data availability: Raw reads have been deposited via SRI via BioProject accession 548 PRJNA1373000. 549

550 551 We acknowledge funding from the MRC Centre for Medical Mycology at the 552 University of Exeter (MR/N006364/2, MR/V033417/1), MRC Doctoral Training Grants 553 (MR/P501955/2, MR/W502649/1), Wellcome Trust Career Development Award 554 (215239/Z/19/Z), Wellcome Trust Fellowship (219551/Z/19/Z), and the NIHR Exeter 555 Biomedical Research Centre. The views expressed are those of the authors and not 556 necessarily those of the NIHR or the Department of Health and Social Care. We thank the 557 Exeter Sequencing Service facility and support from Wellcome Trust Institutional Strategic 558 Support Fund (WT097835MF) to TB, NG, AB, and RAF, Wellcome Trust Multi User 559 Equipment Awards (WT101650MA and 218247/Z/19/Z), Medical Research Council Clinical 560 Infrastructure Funding (MR/M008924/1) and BBSRC LOLA award (BB/K003240/1). We 561 thank the University of Exeter High-Performance Computing (HPC) facility, funded by the 562 UK MRC Clinical Research Infrastructure Initiative (award number MR/M008924/1). BC 563 and SM were funded via the St George’s Hospital Charity grant to TB (19-20-001). Medical 564 Research Foundation Emerging Leaders award in AMR to TB (MRF-160-0009-ELP-BICA-565 0802), and the NIHR Exeter BRC. NG and AB also acknowledge the support of Wellcome 566 Trust Investigator, Collaborative, Equipment, Strategic and Biomedical Resource awards 567 (101873, 200208, 215599, 224323), and the MRC (MR/M026663/2, MR/Y002164/1). TB 568 acknowledges salary support from MR/Y002164/1. We are grateful for comments on the 569 manuscript from Dr Johanna Rhodes. 570 571 Author contributions 572 573 All authors contributed to conceptualisation, writing and editing of the paper. TB 574 originally conceived the study alongside AB, NG, and HG. TB and ST provided the clinical 575 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 21 isolates and metadata for StG and K isolate series. All bioinformatic work was performed 576 by HG with support and supervision from RF. BC and SM obtained comprehensive 577 population profiling, colony morphotype analysis and EUCAST MIC testing, AP performed 578 cell wall analysis, IL performed DNA extractions and microscopy, ME performed HPIC, TC 579 performed additional experiments with DW. 580 581 Figures 582 583 Figure 1: Clinical case series for patients suffering from C. auris candidaemia during antifungal therapy, both pre- and post-clinical isolate collection. (A) St George’s Hospital case series (isolates StG1-5), beginning at day -67 from first blood culture isolation (day 0), and ending at death on day 17. (B) King’s College Hospital case series (isolates K1- preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 22 2), beginning at day -53 before first blood culture and ending at their death on day 81. This case was previously included (case no. 10) in an outbreak report11. (C) Clinical isolates included in this study by hospital site, day of isolation and CLSI MIC values against anidulafungin as reported by NHS Microbiology laboratories. Figure 2: Common sequence variants identified in the genomes of the clinical series. (A) Genomic variants between clinical isolates in the StG1-5 series, excluding intergenic variants. (B) Genomic variants between clinical isolates in the K1-2 isolate series, excluding intergenic variants. (C) Counts of all genomic variants present across both clinical series. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 23 Figure 3: C. auris isolates display heteroresistance with sub-populations of cells displaying contrasting colonial morphotypes. (A) Population analysis profiling (PAP) of StG1 and K1 clinical isolates demonstrates anidulafungin heteroresistance. Points represent mean of 3 biological replicates. Error bars represent standard error of the mean (SEM). (B) Emergence of small and large morphotypes after protracted growth under drug pressure. StG1 and K1 isolates were plated onto YPD containing 8 μg/mL anidulafungin, and colony sizes measured using ImageJ after 96, 120 and 144 h 0 2 4 6 0 0.03 0.06 0.12 0.25 0.5 1 2 4 8 0.00001 0.0001 0.001 0.01 0.1 1 10 100 StG1 % population 0 0.03 0.06 0.12 0.25 0.5 1 2 4 8 0.00001 0.0001 0.001 0.01 0.1 1 10 100 K1 K1StG1 % population Anidulafungin (μg/ml)Anidulafungin (μg/ml) YPD YPD + anidulafungin 8 μg/ml Colonysize(mm2) 1 2 4 6 8 10 24 36 48 72 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 Hours 1 2 4 6 8 10 24 36 48 72 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 Hours OD600 OD600 LCV SCV Parent LCV SCV Parent A FE HG LCVSCVParent 96 120 144 96 120 144 0 1 2 3 0 1 2 34 Hours Hours Colony size (mm2) B C Parent **** **** ****** **** **** LCV SCV Parent LCV SCV Parent LCV SCV D StG1 Parent StG1 SCV K1 Parent K1 LCV K1 SCV 0.0 0.2 0.4 0.6 SMG ns 5 μm preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 24 incubation (biological triplicates). (C) Fluorescence microscopy of individual cells, their cell walls stained with CFW and their DNA content with SYBR green. (D) Colony morphotypes of StG1 are stable following re-plating. StG1 large colony variants (LCVs) and small colony variants (SCVs) were taken from a YPD+anidulafungin 8 μg/mL plate, and the StG1 parental isolate taken from a YPD plate. These cells were replated onto YPD at 102 CFU and imaged after 48 h at 30 °C. (E) Colony sizes were then measured on these plates (F) by ImageJ. Plots represent distribution of all colony sizes from 3 biological replicates for each cell type. One-way ANOVA: *, p <0.05; **, p ≤0.01; ***, p ≤0.001; ****, p ≤0.0001. (F) Growth (OD600) of the StG1 parent, LCV and SCV colonies in YPD containing 0 or 8 μg/ml anidulafungin. Points represent mean of 3 biological replicates, error bars represent SEM. (G) Supra-MIC growth for StG1 SCVs vs parent isolate. Each point represents mean of technical triplicate results. One-way ANOVA, p- values as above. (H) Etest strips reveal the zones of inhibition for the parental isolate StG1, LCVs and SCVs. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 25 Figure 4: Significant within-gene variation in clinical isolates and their large and small morphotypes. FastTree phylogenies for StG and K clinical series isolates are shown above. The colours in the legend (top right) and below the phylogenies indicate the group: clinical isolate versus large or small colony variant; genome sequenced following growth overnight on- or off- anidulafungin (8 μg/mL). Black boxes indicate the presence of the variant specified to the right of the figure: NSY: non-synonymous, SYN: synonymous, NON: nonsense, INS: insertion, DEL: deletion, AFG: anidulafungin. Differences between isolates K1 and K2 were driven by 237 intergenic variant differences (95 present only in isolate K1, 142 present only in isolate K2) in addition to those outlined in Figure 2B. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 26 Figure 5: Impact of anidulafungin on chitin content and β-1,3-glucan exposure of cell walls of StG parents and large and small morphotypes: (A) The chitin content of the cell wall of StG parents and large and small morphotypes during growth in the presence (+, red) or absence (-, dark blue) of 8 μg/ml anidulafungin (ANF) was quantified by Calcofluor-White (CFW) staining and flow cytometry. Data represent means and standard deviations from three replicate experiments, one-way ANOVA: ns, not CA B D No drug Anidulafungin (ANF) 200 nm LCV SCV Parent E ANF LCVSCV WT Parent SCV LCV Parent Chitin Glucan Mannan Chitin Glucan Mannan SCV LCV Parent SCV LCV Parent SCV LCV ANF ANF ANF F I G J H K - + - + - 0 100 200 300 400 500 Inner cell wall diameter (nm) ns Parent SCV LCV *** **** **** **** **** **** **** - + - + - 0 50 100 150Outer cell wall diameter (nm) Parent SCV LCV ANF ANF ANF preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 27 significant; *, p <0.05; **, p ≤0.01; ***, p ≤0.001; ****, p ≤0.0001. (B) The fold change in β-1,3-glucan exposure in large and small morphotypes, compared with parental isolates, in the presence (+, red) or absence (-, dark blue) of anidulafungin was quantified by dectin-1 staining and flow cytometry. Statistics as per (A). (C) Transmission electron microscopy (TEM) reveals cell wall remodeling in response to anidulafungin. The diameters of the (D) inner and (E) outer cell wall layers were quantified from TEM images using ImageJ. Data represent means and standard deviations from n >30 cells (10 measurements per cell) and were analysed using Brown-Forsythe and Welch ANOVA, p-values as above. (F-K) High pressure ion chromatography (HPIC) was used to measure the chitin (F), glucan (G) and mannan (H) contents of the cell walls of the StG1-5 series following growth on YPD containing 0 or 8 μg/mL anidulafungin. The chitin (I), glucan (J) and mannan (K) contents of the cell walls of the parental isolate StG1 and its SCV and LCV isolates and their corresponding changes in cell wall content inc. chitin, glucan, and mannan after growth without or with (8 μg/mL) anidulafungin. One-way ANOVA, p-values as above. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 28 Figure 6: Exposure to anidulafungin increases the ploidy of C. auris. (A) The ploidy of C. auris cells in colonies taken directly from plates was measured by staining their DNA content with SYBR green and quantifying the fluorescence by flow cytometry. Haploid and diploid S. cerevisiae strains and a C. auris clade I isolate were used as controls. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 29 Colonies for the C. auris clade I and S. cerevisiae controls were taken from YPD plates. Colonies for the C. auris StG1 parent and its small and large colony variants were harvested from YPD plates containing 8 μg/mL anidulafungin. Peak volumes were quantified using were quantified using FlowJo v.10 software. (B) The proportions of 1, 2n, 3n, 4n and >4n cells in each sample were calculated from the peak volumes obtained by flow cytometry in (A) with unpaired t-test comparison between groups (C): p <0.05; **, p ≤0.01; ***, p ≤0.001; ****, p ≤0.0001.

584 585 1. Liu, F. et al. Phylogenomic analysis of the Candida auris-Candida haemuli 586 clade and related taxa in the Metschnikowiaceae, and proposal of thirteen new genera, 587 fifty-five new combinations and nine new species. Persoonia : Molecular Phylogeny and 588 Evolution of Fungi 52, 22–43 (2024). 589 2. WHO. World Health Organization (WHO) fungal priority pathogens list to 590 guide research, development and public health action. (2022) 591 doi:https://www.who.int/publications/i/item/9789240060241. 592 3. Chen, J. et al. Is the superbug fungus really so scary? A systematic review 593 and meta-analysis of global epidemiology and mortality of Candida auris. BMC Infectious 594 Diseases 20, 827 (2020). 595 4. Fisher, M. C. & Denning, D. W. The WHO fungal priority pathogens list as a 596 game-changer. Nature Reviews Microbiology 21, 211–212 (2023). 597 5. Lionakis, M. S. & Chowdhary, A. Candida auris infections. New England 598 Journal of Medicine 391, 1924–1935 (2024). 599 6. Satoh, K. et al. Candida auris sp. Nov., a novel ascomycetous yeast isolated 600 from the external ear canal of an inpatient in a Japanese hospital. Microbiology and 601 Immunology 53, 41–44 (2009). 602 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 30 7. Lockhart, S. R. et al. Simultaneous emergence of multidrug-resistant 603 Candida auris on 3 continents confirmed by whole-genome sequencing and 604 epidemiological analyses. Clinical Infectious Diseases 64, 134–140 (2017). 605 8. Gifford, H., Rhodes, J. & Farrer, R. A. The diverse genomes of Candida 606 auris. The Lancet Microbe 0, (2024). 607 9. Sexton, D. J. et al. Positive correlation between Candida auris skin-608 colonization burden and environmental contamination at a ventilator-capable skilled 609 nursing facility in Chicago. Clinical Infectious Diseases 73, 1142–1148 (2021). 610 10. Eyre, D. W. et al. A Candida auris outbreak and its control in an intensive 611 care setting. New England Journal of Medicine 379, 1322–1331 (2018). 612 11. Taori, S. K. et al. Candida auris outbreak: Mortality, interventions and cost of 613 sustaining control. Journal of Infection 79, 601–611 (2019). 614 12. Zhu, Y. et al. Laboratory analysis of an outbreak of Candida auris in New 615 York from 2016 to 2018: Impact and lessons learned. Journal of Clinical Microbiology 58, 616 e01503–19 (2020). 617 13. Wasylyshyn, A. & Stoneman, E. K. Management of Candida auris. JAMA E1, 618 (2024). 619 14. Kohlenberg, A., Monnet, D. L., Plachouras, D. & Group, C. auris survey 620 collaborative. Increasing number of cases and outbreaks caused by Candida auris in the 621 EU/EEA, 2020 to 2021. Eurosurveillance 27, 2200846 (2022). 622 15. Lyman, M. et al. Worsening spread of Candida auris in the United States, 623 2019 to 2021. Annals of Internal Medicine (2023) doi:10.7326/M22-3469. 624 16. Davi Marena, G., Nosanchuk, J. D. & Taborda, C. P. Candidozyma auris 625 alert in South America: An epidemiological and therapeutic update. Current Tropical 626 Medicine Reports 12, 1 (2025). 627 17. Ibe, C. & Pohl, C. H. Epidemiology and drug resistance among Candida 628 pathogens in Africa: Candida auris could now be leading the pack. The Lancet Microbe 629 100996 (2024) doi:10.1016/j.lanmic.2024.100996. 630 18. Prayag, P. S., Patwardhan, S., Panchakshari, S., Rajhans, P. A. & Prayag, 631 A. The dominance of Candida auris: A single-center experience of 79 episodes of 632 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 31 candidemia from western India. Indian Journal of Critical Care Medicine 26, 560–563 633 (2022). 634 19. Jesudason, T. UKHSA publishes updated guidance for C. auris. The Lancet 635 Microbe 6, (2025). 636 20. UK Health Security Agency. Candidozyma auris: Guidance for acute 637 healthcare settings. GOV.UK (2025). 638 21. Chow, N. A. et al. Tracing the evolutionary history and global expansion of 639 Candida auris using population genomic analyses. mBio 11, 15 (2020). 640 22. Lass-Flörl, C. et al. Invasive candidiasis. Nature Reviews Disease Primers 641 10, 20 (2024). 642 23. Cornely, O. A. et al. Global guideline for the diagnosis and management of 643 candidiasis: An initiative of the ECMM in cooperation with ISHAM and ASM. The Lancet 644 Infectious Diseases 0, (2025). 645 24. Lara-Aguilar, V. et al. Adaptation of the emerging pathogenic yeast Candida 646 auris to high caspofungin concentrations correlates with cell wall changes. Virulence 12, 647 1400–1417 (2021). 648 25. Meletiadis, J. et al. A multicentre study for determination of epidemiological 649 cut-off values for Candida auris with EUCAST broth microdilution reference methodology. 650 Journal of Antimicrobial Chemotherapy 80, 2459–2465 (2025). 651 26. Arendrup, M. C. et al. How to interpret MICs of amphotericin B, 652 echinocandins and flucytosine against Candida auris (Candidozyma auris) according to 653 the newly established European Committee for Antimicrobial Susceptibility Testing 654 (EUCAST) breakpoints. Clinical Microbiology and Infection S1198743X25003404 (2025) 655 doi:10.1016/j.cmi.2025.07.002. 656 27. Kordalewska, M. et al. Understanding echinocandin resistance in the 657 emerging pathogen Candida auris. Antimicrobial Agents and Chemotherapy 62, e00238–658 18 (2018). 659 28. Chowdhary, A. et al. A multicentre study of antifungal susceptibility patterns 660 among 350 Candida auris isolates (2009-17) in India: Role of the ERG11 and FKS1 genes 661 in azole and echinocandin resistance. The Journal of Antimicrobial Chemotherapy 73, 662 891–899 (2018). 663 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 32 29. Asadzadeh, M. et al. Molecular characterisation of Candida auris isolates 664 from immunocompromised patients in a tertiary-care hospital in Kuwait reveals a novel 665 mutation in FKS1 conferring reduced susceptibility to echinocandins. Mycoses 65, 331–666 343 (2022). 667 30. Al-Obaid, I. et al. Fatal breakthrough candidemia in an immunocompromised 668 patient in Kuwait due to Candida auris exhibiting reduced susceptibility to echinocandins 669 and carrying a novel mutation in hotspot-1 of FKS1. Journal of Fungi 8, 267 (2022). 670 31. Kordalewska, M. et al. Novel non-hot spot modification in Fks1 of Candida 671 auris confers echinocandin resistance. Antimicrobial Agents and Chemotherapy 67, 672 e00423–23 (2023). 673 32. Sharma, D. et al. Impact of FKS1 genotype on echinocandin in-vitro 674 susceptibility in Candida auris and in-vivo response in a murine model of infection. 675 Antimicrobial Agents and Chemotherapy AAC0165221 (2021) doi:10.1128/AAC.01652-21. 676 33. Hirayama, T. et al. Echinocandin resistance in Candida auris occurs in the 677 murine gastrointestinal tract due to FKS1 mutations. Antimicrobial Agents and 678 Chemotherapy 67, e01243–22 (2023). 679 34. Wang, T. W., Putnam, N., Johnson, J. K. & Jabra-Rizk, M. A. In vivo 680 evolution of Candida auris multidrug resistance in a patient receiving antifungal treatment. 681 The Journal of Infectious Diseases jiaf394 (2025) doi:10.1093/infdis/jiaf394. 682 35. Jacobs, S. E. et al. Candida auris pan-drug-resistant to four classes of 683 antifungal agents. Antimicrobial Agents and Chemotherapy 66, e0005322 (2022). 684 36. Tian, S. et al. Evolutionary accumulation of FKS1 mutations from clinical 685 echinocandin-resistant Candida auris. Emerging Microbes & Infections 13, 2377584 686 (2024). 687 37. Carolus, H. et al. Genome-wide analysis of experimentally evolved Candida 688 auris reveals multiple novel mechanisms of multidrug resistance. mBio 12, e03333–20 689 (2021). 690 38. Rybak, J. M., Cuomo, C. A. & David Rogers, P. The molecular and genetic 691 basis of antifungal resistance in the emerging fungal pathogen Candida auris. Current 692 Opinion in Microbiology 70, 102208 (2022). 693 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 33 39. Wang, Y. & Xu, J. Associations between genomic variants and antifungal 694 susceptibilities in the archived global Candida auris population. Journal of Fungi 10, 86 695 (2024). 696 40. Sah, S. K., Hayes, J. J. & Rustchenko, E. The role of aneuploidy in the 697 emergence of echinocandin resistance in human fungal pathogen Candida albicans. 698 PLOS Pathogens 17, e1009564 (2021). 699 41. Selmecki, A. M. et al. Polyploidy can drive rapid adaptation in yeast. Nature 700 519, 349–352 (2015). 701 42. Vande Zande, P., Zhou, X. & Selmecki, A. The dynamic fungal genome: 702 Polyploidy, aneuploidy and copy number variation in response to stress. Annual Review of 703 Microbiology 77, 341–361 (2023). 704 43. Shor, E., Perlin, D. S. & Kontoyiannis, D. P. Tolerance and heteroresistance 705 to echinocandins in Candida auris: Conceptual issues, clinical implications, and 706 outstanding questions. mSphere e00161–25 (2025) doi:10.1128/msphere.00161-25. 707 44. Ostrowsky, B. et al. Candida auris isolates resistant to three classes of 708 antifungal medications - New York, 2019. Morbidity and Mortality Weekly Report 69, 6–9 709 (2020). 710 45. Lyman, M. Notes from the field: Transmission of pan-resistant and 711 echinocandin-resistant Candida auris in health care facilities - Texas and the District of 712 Columbia, January-April 2021. Morbidity and Mortality Weekly Report 70, (2021). 713 46. Lockhart, S. R., Chowdhary, A. & Gold, J. A. W. The rapid emergence of 714 antifungal-resistant human-pathogenic fungi. Nature Reviews Microbiology 1–15 (2023) 715 doi:10.1038/s41579-023-00960-9. 716 47. Berman, J. & Krysan, D. J. Drug resistance and tolerance in fungi. Nature 717 Reviews Microbiology 18, 319–331 (2020). 718 48. Zhai, B. et al. Antifungal heteroresistance causes prophylaxis failure and 719 facilitates breakthrough Candida parapsilosis infections. Nature Medicine (2024) 720 doi:10.1038/s41591-024-03183-4. 721 49. Ben-Ami, R. et al. Heteroresistance to fluconazole Is a continuously 722 distributed phenotype among Candida glabrata clinical strains associated with in vivo 723 persistence. mBio 7, e00655–16 (2016). 724 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 34 50. Rosenberg, A. et al. Antifungal tolerance is a subpopulation effect distinct 725 from resistance and is associated with persistent candidemia. Nature Communications 9, 726 2470 (2018). 727 51. Gautier, C., Maciel, E. I. & Ene, I. V. Approaches for identifying and 728 measuring heteroresistance in azole-susceptible Candida isolates. Microbiology Spectrum 729 12, e04041–23 (2024). 730 52. Munro, C. A. et al. The PKC, HOG and Ca2+ signalling pathways co-731 ordinately regulate chitin synthesis in Candida albicans. Molecular microbiology 63, 1399–732 1413 (2007). 733 53. Walker, L. A., Gow, N. A. R. & Munro, C. A. Elevated chitin content reduces 734 the susceptibility of Candida species to caspofungin. Antimicrobial Agents and 735 Chemotherapy 57, 146–154 (2013). 736 54. Fayed, B., Jayakumar, M. N. & Soliman, S. S. M. Caspofungin-resistance in 737 Candida auris is cell wall-dependent phenotype and potential prevention by zinc oxide 738 nanoparticles. Medical Mycology (2021) doi:10.1093/mmy/myab059. 739 55. Zamith-Miranda, D. et al. Transcriptional and translational landscape of 740 Candida auris in response to caspofungin. Computational and Structural Biotechnology 741 Journal 19, 5264–5277 (2021). 742 56. Szekely, A., Borman, A. M. & Johnson, E. M. Candida auris isolates of the 743 southern Asian and South African lineages exhibit different phenotypic and antifungal 744 susceptibility profiles in vitro. Journal of Clinical Microbiology 57, e02055–18 (2019). 745 57. CDC. Antifungal Susceptibility Testing for C. auris. CDC (2024). 746 58. Muñoz, J. F. et al. Clade-specific chromosomal rearrangements and loss of 747 subtelomeric adhesins in Candida auris. Genetics 218, iyab029 (2021). 748 59. Cha, H. et al. The calcineurin pathway regulates extreme thermotolerance, 749 cell membrane and wall integrity, antifungal resistance, and virulence in Candida auris. 750 PLOS Pathogens 21, e1013363 (2025). 751 60. Wagner, A. S., Lumsdaine, S. W., Mangrum, M. M. & Reynolds, T. B. 752 Caspofungin-induced 𝛽(1,3)-glucan exposure in Candida albicans is driven by increased 753 chitin levels. mBio 14, e00074–23 (2023). 754 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 35 61. da Silva Dantas, A. et al. Crosstalk between the calcineurin and cell wall 755 integrity pathways prevents chitin overexpression in Candida albicans. Journal of Cell 756 Science 134, jcs258889 (2021). 757 62. Cha, H., Won, D. & Bahn, Y.-S. Signaling pathways governing the 758 pathobiological features and antifungal drug resistance of Candida auris. mBio 16, 759 e02475–23 (2025). 760 63. Smith, A. C., Rizvi, H., Hickman, M. A. & Morran, L. T. Increased virulence 761 and large-scale reduction in genome size of tetraploid Candida albicans evolved in 762 nematode hosts. Frontiers in Fungal Biology 3, 903135 (2022). 763 64. Avramovska, O., Smith, A. C., Rego, E. & Hickman, M. A. Tetraploidy 764 accelerates adaptation under drug selection in a fungal pathogen. Frontiers in Fungal 765 Biology 3, 984377 (2022). 766 65. Gerstein, A. C., Lim, H., Berman, J. & Hickman, M. A. Ploidy tug-of-war: 767 Evolutionary and genetic environments influence the rate of ploidy drive in a human fungal 768 pathogen. Evolution 71, 1025–1038 (2017). 769 66. Suzuki, T., Nishibayashi, S., Kuroiwa, T., Kanbe, T. & Tanaka, K. Variance of 770 ploidy in Candida albicans. Journal of Bacteriology 152, 893–896 (1982). 771 67. Marr, K. A., White, T. C., van Burik, J.-A. H. & Bowden, R. A. Development of 772 fluconazole resistance in Candida albicans causing disseminated infection in a patient 773 undergoing marrow transplantation. Clinical Infectious Diseases 25, 908–910 (1997). 774 68. Abbey, D. A. et al. YMAP: A pipeline for visualization of copy number 775 variation and loss of heterozygosity in eukaryotic pathogens. Genome Medicine 6, 100 776 (2014). 777 69. Bravo Ruiz, G. et al. Rapid and extensive karyotype diversification in haploid 778 clinical Candida auris isolates. Current Genetics 65, 1217–1228 (2019). 779 70. Du, H., Zheng, Q., Bennett, R. J. & Huang, G. Ploidy changes in human 780 fungal pathogens: Going beyond sexual reproduction. PLOS Pathogens 18, e1010954 781 (2022). 782 71. Singh, S. D. et al. Hsp90 governs echinocandin resistance in the pathogenic 783 yeast Candida albicans via calcineurin. PLOS pathogens 5, e1000532 (2009). 784 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 36 72. Walker, L. A., Gow, N. A. R. & Munro, C. A. Fungal echinocandin resistance. 785 Fungal Genetics and Biology 47, 117–126 (2010). 786 73. Ceballos-Garzon, A. et al. CRISPR-Cas9 approach confirms Calcineurin-787 responsive zinc finger 1 (Crz1) transcription factor as a promising therapeutic target in 788 echinocandin-resistant Candida glabrata. PLoS One 17, e0265777 (2022). 789 74. Wang, T. W. et al. Functional redundancy in Candida auris cell surface 790 adhesins crucial for cell-cell interaction and aggregation. Nature Communications 15, 9212 791 (2024). 792 75. Kean, R. et al. Transcriptome assembly and profiling of Candida auris 793 reveals novel insights into biofilm-mediated resistance. mSphere 3, e00334–18 (2018). 794 76. Fan, S., Li, C., Bing, J., Huang, G. & Du, H. Discovery of the diploid form of 795 the emerging fungal pathogen Candida auris. ACS Infectious Diseases 6, 2641–2646 796 (2020). 797 77. Anderson, J. B., Sirjusingh, C. & Ricker, N. Haploidy, diploidy and evolution 798 of antifungal drug resistance in Saccharomyces cerevisiae. Genetics 168, 1915–1923 799 (2004). 800 78. Thomson, G. J. et al. Metabolism-induced oxidative stress and DNA damage 801 selectively trigger genome instability in polyploid fungal cells. The EMBO Journal 38, 802 e101597 (2019). 803 79. Stone, N. R. H. et al. Dynamic ploidy changes drive fluconazole resistance in 804 human cryptococcal meningitis. Journal of Clinical Investigation 129, 999–1014 (2019). 805 80. Smith, A. C. & Hickman, M. A. Host-induced genome instability rapidly 806 generates phenotypic variation across Candida albicans strains and ploidy states. 807 mSphere 5, e00433–20 (2020). 808 81. Healey, K. R. & Perlin, D. S. Fungal resistance to echinocandins and the 809 MDR phenomenon in Candida glabrata. Journal of Fungi 4, 105 (2018). 810 82. Arastehfar, A. et al. Drug-resistant fungi: An emerging challenge threatening 811 our limited antifungal armamentarium. Antibiotics 9, 877 (2020). 812 83. Morawska, L. P., Hernandez-Valdes, J. A. & Kuipers, O. P. Diversity of bet-813 hedging strategies in microbial communities—recent cases and insights. WIREs 814 Mechanisms of Disease 14, e1544 (2022). 815 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 37 84. Rosenberg, A. et al. Antifungal tolerance is a subpopulation effect distinct 816 from resistance and is associated with persistent candidemia. Nature Communications 9, 817 2470 (2018). 818 85. Jacobs, S. E. & Chaturvedi, V. CAF to the rescue! Potential and challenges 819 of combination antifungal therapy for reducing morbidity and mortality in hospitalized 820 patients with serious fungal infections. Open Forum Infectious Diseases 11, ofae646 821 (2024). 822 86. Wake, R. M. et al. Optimizing the treatment of invasive candidiasis — a case 823 for combination therapy. Open Forum Infectious Diseases 11, ofae072 (2024). 824 87. Ewels, P., Magnusson, M., Lundin, S. & Käller, M. MultiQC: Summarize 825 analysis results for multiple tools and samples in a single report. Bioinformatics 32, 3047–826 3048 (2016). 827 88. McKenna, A. et al. The Genome Analysis Toolkit: A MapReduce framework 828 for analyzing next-generation DNA sequencing data. Genome Research 20, 1297–1303 829 (2010). 830 89. Misas, E. et al. Mitochondrial genome sequences of the emerging fungal 831 pathogen Candida auris. Frontiers in Microbiology 11, 560332 (2020). 832 90. Li, H. Aligning sequence reads, clone sequences and assembly contigs with 833 BWA-MEM. (2013) doi:10.48550/arXiv.1303.3997. 834 91. CDC. Antifungal Susceptibility Testing of Yeasts using Gradient Diffusion 835 Strips. Fungal Diseases (2025). 836 92. Price, M. N., Dehal, P. S. & Arkin, A. P. FastTree 2 – approximately 837 maximum-likelihood trees for large alignments. PLOS ONE 5, e9490 (2010). 838 93. Farrer, R. ECATools: Scripts to work with ’Entirely Covered in All’ (ECA) 839 positions from VCFs (i.e. Sites most useful for phylogenetics). (2019) 840 doi:https://github.com/rhysf/ECATools. 841 94. Li, H. et al. The Sequence Alignment/Map format and SAMtools. 842 Bioinformatics 25, 2078–2079 (2009). 843 95. Benjamini, Y. & Hochberg, Y. Controlling the false discovery rate: A practical 844 and powerful approach to multiple testing. Journal of the Royal Statistical Society Series B: 845 Statistical Methodology 57, 289–300 (1995). 846 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint 38 96. Gifford, H. et al. Xenosiderophore transporter gene expression and clade-847 specific filamentation in Candida auris killifish infection. (2024) 848 doi:10.1101/2024.10.22.619744. 849 97. Teixeira, M. C. et al. YEASTRACT+: A portal for the exploitation of global 850 transcription regulation and metabolic model data in yeast biotechnology and 851 pathogenesis. Nucleic Acids Research 51, D785–D791 (2023). 852 98. Yoshimoto, H. et al. Genome-wide analysis of gene expression regulated by 853 the calcineurin/Crz1p signaling pathway in Saccharomyces cerevisiae. The Journal of 854 Biological Chemistry 277, 31079–31088 (2002). 855 99. Cauldron, N. C., Shea, T. & Cuomo, C. A. Improved genome assembly of 856 Candida auris strain B8441 and annotation of B11205. Microbiology Resource 857 Announcements 0, e00512–24 (2024). 858 100. Pradhan, A. et al. Non-canonical signalling mediates changes in fungal cell 859 wall PAMPs that drive immune evasion. Nature Communications 10, 5315 (2019). 860 101. Ma, Q. et al. Impact of secreted glucanases upon the cell surface and fitness 861 of Candida albicans during colonisation and infection. The Cell Surface 11, 100128 (2024). 862 102. McDonald, K. L. & Webb, R. I. Freeze substitution in 3 hours or less. Journal 863 of Microscopy 243, 227–233 (2011). 864 103. Rueden, C. T. et al. ImageJ2: ImageJ for the next generation of scientific 865 image data. BMC Bioinformatics 18, 529 (2017). 866 104. Mora-Montes, H. M. et al. Endoplasmic reticulum alpha-glycosidases of 867 Candida albicans are required for N glycosylation, cell wall integrity, and normal host-868 fungus interaction. Eukaryotic Cell 6, 2184–2193 (2007). 869 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 18, 2026. ; https://doi.org/10.64898/2026.01.17.700071doi: bioRxiv preprint