Strains used in this study are listed in Supplementary Table 1 . Stock cultures were preserved in 25% glycerol and maintained at -80˚C. Cells were routinely grown in YPD media (1% [w/v] yeast extract, 2% [w/v] peptone and 2% [w/v] D-glucose) at 30˚C in a shaking incubator at 150–200 rpm. SDC agar plates contained 0.67% [wt/vol] yeast nitrogen base without amino acids, 2% [wt/vol] d-glucose, 0.2% [wt/vol] complete amino acid mixture, and 2% [wt/vol] agar. For solid medium, 2% [w/v] agar was added. For the selection of gene knockout strains, YPD agar containing 400 μg/mL nourseothricin (Werner BioAgents) medium was used (YPD+NAT). To evict the disruption cassette, yeast nitrogen base (YNB)–bovine serum albumin (BSA) (0.17% [w/v] YNB, 2% [w/v] D-glucose, 0.02% [w/v] BSA, 2% [w/v] agar) plates were used. Drugs were dissolved in dimethyl sulfoxide (DMSO) and stored at -20˚C.

NAT1 flipper gene deletion cassette was amplified from plasmid pJK863 (Shen et al., 2005). Approximately 500 bp upstream and 500 bp downstream regions of the gene to be deleted was amplified from the genomic DNA of SC5314. The upstream region of each gene was then fused by PCR to the 5′ region of the cassette and the downstream region of the gene was fused by PCR to the 3′ region of the cassette, such that the two PCR products had an overlap of 500–1,000 bp within the cassette. All primer sequences used for plasmid and strain construction are listed in Supplementary Table 2 . The upstream and downstream fusion products for each gene were then simultaneously transformed in C. albicans following the lithium acetate method (Wilson et al., 2000). Transformants were selected on YPD+NAT agar plates. The replacement of the gene with the NAT1 flipper cassette was confirmed by diagnostic PCR, using primers that annealed outside the flanking homology regions. The NAT1 flipper was then evicted by streaking the clones on YNB-BSA plates, which induces the Flp recombinase.

The CLSI M44-A2 guidelines (Institute, C.a.L.S., 2009) for antifungal disk diffusion susceptibility testing were followed, with slight modifications. Strains were grown on YPD-agar plates, cell density was adjusted to 1 × 10⁶ cells/mL, and 100 μL of cell suspension was plated on plates. One paper disk (GE Healthcare, USA) supplemented with 50 μg miconazole was placed in the center of each plate. The plates were then incubated for 48 h then photographed. Each strain was evaluated in duplicate, with two independent biological replicates.

Cells were suspended in distilled water and adjusted to 1x10⁶ cells/mL. 3 µL of 10-fold serial dilutions were spotted on YPD with or without drugs. The plates were incubated at 30˚C and photographed after 48 h. The assay was repeated twice, with independent experiments conducted at distinct time points.

Cells were suspended in distilled water and adjusted to 1x10⁷ cells/mL. 100 µL of cell suspension were spread on YPD plates supplemented with MCZ. The plates were incubated at 30˚C for 5 days. 9–30 adaptors were randomly chosen.

Total DNA extraction and genomic DNA library preparation were performed as described previously (Yang et al., 2021). The final libraries were sequenced by BGISEQ-500. Raw fastq files were uploaded to YMAP (version 1.0) (http://lovelace.cs.umn.edu/Ymap/) (Abbey et al., 2014). Read depth was plotted as a function of chromosome position using the Assembly 22 version of the SC5314 reference genome (http://www.candidagenome.org/download/sequence/C_albicans_SC5314/Assembly22/current/C_albicans_SC5314_A22_current_chromosomes.fasta.gz).

Two clinical isolates and one reference strain were tested for MCZ tolerance. SC5314 is the commonly used reference strain. YJB-T1891 and YJB-T490 were clinical isolates collected from Israeli patients (Yang et al., 2023). Two temperature conditions were tested: 30°C and 37°C. Two type of media was utilized: YPD and SDC. YPD is a rich, non-defined media that supports rapid growth. SDC is a more defined and controlled media. Two assays were performed: DDA and spot assay.

In DDAs performed on YPD-agar, at both temperatures, YJB-T1891 had colonies growing inside the ZOI and YJB-T490 had clear ZOI. SC5314 had clear ZOI at 30°C and colonies growing inside the ZOI at 37°C. In addition, SC5314 and YJB-T490 had smaller ZOI at 37°C than at 30°C ( Figure 1A left panel). In DDAs performed on SDC-agar, all strains displayed clear ZOI. YJB-T1891 and YJB-T490 had bigger ZOI at 37°C than at 30°C ( Figure 1A right panel).

Analysis of DDA plates with diskimageR indicated that, on YPD-agar, the RAD₂₀ values of SC5314 at 30°C and 37°C were 13.0 ± 0.0 and 12.5 ± 0.7, respectively. The RAD₂₀ values of YJB-T1891 at 30°C and 37°C were 11.0 ± 0.0 and 10.5 ± 0.7, respectively. And the RAD₂₀ values of YJB-T490 at 30°C and 37°C were 13.0 ± 0.0 and 12.0 ± 0.0, respectively. Thus, when tested with YPD-agar, temperature condition did not alter extent of resistance. On SDC-agar, the RAD₂₀ values of SC5314 at both 30°C and 37°C were 18.0 ± 0.0. The RAD₂₀ values of YJB-T1891 at 30°C and 37°C were 20.0 ± 0.0 and 21.5 ± 0.7, respectively. And the RAD₂₀ values of YJB-T490 at 30°C and 37°C were 20.0 ± 0.0 and 22.5 ± 0.7, respectively ( Figure 1A ). Thus, when tested with SDC-agar, strains had slightly increased resistance at lower temperature.

As compared to the slight change of RAD, the change of FoG was more obvious on YPD-agar. The FoG₂₀ values of SC5314 at 30°C and 37°C were 0.16 ± 0.06 and 0.70 ± 0.08, respectively. The FoG₂₀ values of YJB-T1891 at 30°C and 37°C were 0.83 ± 0.04 and 0.68 ± 0.05, respectively. The FoG₂₀ values of YJB-T1891 at 30°C and 37°C were 0.13 ± 0.03 and 0.17 ± 0.07, respectively ( Figure 1A ).

In spot assays performed on YPD-agar, strains were tested for the ability of growing in the presence of a wide range of MCZ concentrations (0.002–4 μg/mL). At both temperatures, YJB-T1891 could tolerate 1 μg/mL MCZ. SC5314 could tolerate 1 μg/mL MCZ only at 37°C, while it was obviously inhibited by 0.002 μg/mL MCZ at 30°C. YJB-T490 could only tolerate 0.002 and 0.004 μg/mL MCZ at 30°C and 37°C, respectively ( Figure 1B left panel). On SDC-agar plates, at both temperatures, all strains could only tolerate 0.015 μg/mL MCZ. Both YJB-T1891 and YJB-T490 grew better at 30°C than at 37°C on plates supplemented with or without MCZ ( Figure 1B right panel).

Thus, MCZ tolerance was strain background-dependent and was modulated by physiological factors including temperature and medium composition.

In order to investigate how non-tolerant isolate adapts to miconazole, SC5314 was exposed to high concentrations of MCZ (0.004–4 μg/mL) and the plates were incubated at 30°C ( Figure 2A ). After 5 days, randomly 9 colonies were chosen from each of plates supplemented with 0.008–2 μg/mL MCZ. All the 81 adaptors were tested for MCZ tolerance by DDAs ( Supplementary Figure 1 ). Only 3 adaptors derived from 0.008 μg/mL MCZ (#3, #7 and #9), and another 2 adaptors derived from 0.015 μg/mL MCZ (#4 and #5) were not tolerant. All the other 76 adaptors were tolerant. Of note, none of the 81 adaptors displayed altered radius of the inhibition zone, indicating they were not resistant.

The diskImageR analysis indicates that the parent had RAD₂₀ and FoG₂₀ values of 13.2 ± 0.42 and 0.11 ± 0.05, respectively, while the adaptors had RAD₂₀ and FoG₂₀ values of 13.2 ± 0.41 and 0.53 ± 0.11, respectively ( Figure 2B ). The results of the Student’s t-test show that there was no significant difference in RAD₂₀ between the parent and the adaptors (p>0.05), whereas the adaptors had significantly higher FoG₂₀ values than the parent (p<0.001).

We asked to what extent the adaptors could tolerate MCZ. Spot assay was performed on YPD-agar plates supplemented with 0.002–4 μg/mL MCZ. Growth of the parent SC5314 and 4 non-tolerant adaptors, #3 and #9 derived from 0.008 μg/mL MCZ, as well as #4 and #5 derived from 0.015 μg/mL MCZ, could only grow in the presence of 0.002 μg/mL MCZ. Another non-tolerant adaptor, #7 derived from 0.008 μg/mL MCZ, could only grow in the presence of 0.008 μg/mL MCZ. Two tolerant adaptors, #8 and #9 derived from 0.015 μg/mL MCZ plate, could grow in the presence of 0.25 μg/mL MCZ, and all the other 74 tolerant adaptors could grow on plates containing 2 μg/mL MCZ, but none of them could grow in the presence of 4 μg/mL MCZ ( Figure 3 ). The difference in the levels of tolerance might be due to different genetic or genomic changes happened in different adaptors.

Randomly 4 tolerant adaptors which were derived from each concentration of MCZ (0.008–2 μg/mL) and could tolerate 2 μg/mL MCZ, were sequenced. All the 36 sequenced adaptors had chromosome R trisomy (ChrRx3): 27, and 9 adaptors had duplication of A and B holomog, respectively ( Figure 4 ; Supplementary Figure 2 ). Thus, under different selection force, the same aneuploidy appeared and thereby conferring similar extent of MCZ tolerance.

Two less tolerant adaptors (#8 and #9) derived from 0.015 μg/mL MCZ plate were also sequenced. Both were euploid ( Supplementary Figure 3 ).

Increased efflux is one of the major mechanisms of drug resistance across the kingdoms. But essentiality of efflux in maintenance of tolerance is largely unexplored. In C. albicans genome, CDR1 and CDR2 encode multidrug transporter of the ATP-binding cassette (ABC) superfamily, and MDR1 encodes multidrug resistance protein of the major facilitator superfamily (MFS). TAC1 encodes a Zn(2)-Cys(6) binuclear cluster type transcriptional activator of drug-responsive genes including the ABC drug transporters, CDR1 and CDR2. In this study, both alleles of CDR1, CDR2, MDR1 and TAC1 were deleted from SC5314. The deletion strains as well as the wild type were tested with DDAs and spot assays.

In DDAs, the wild type, cdr2 Δ/Δ, mdr1 Δ/Δ and tac1 Δ/Δ strains had clear ZOI at 30°C and had colonies growing inside the ZOI at 37°C, while cdr1 Δ/Δ strain also had clear ZOI at 30°C and had obviously less colonies growing inside the ZOI at 37°C ( Figure 5A ).

In spot assays, the wild type and the strains with deletions of CDR2, MDR1 and TAC1 could tolerate 1 μg/mL MCZ at 37°C and the growths were obviously inhibited by 0.002 μg/mL MCZ at 30°C. Growth of cdr1 Δ/Δ strain was inhibited by 0.002 μg/mL MCZ at both 30°C and 37°C ( Figure 5B ).

Thus, CDR1, but not CDR2, MDR1 or TAC1, was required for maintaining MCZ tolerance at 37°C.

We asked if CDR1 was essential for gain of miconazole tolerance. The cdr1 Δ/Δ strain was spread on YPD plates supplemented with MCZ. On plates containing 0.002 – 0.5 μg/mL MCZ, colonies appeared within 5 days ( Figure 6A ). Randomly 9 adaptors from each drug plate were tested with DDAs. Among the 81 adaptors, only 7, 2, 3 and 2 adaptors derived from 0.002, 0.004, 0.008 and 0.25 μg/mL MCZ, respectively, were not tolerant to MCZ. Of note, like SC5314 derived adaptors, none of the 81 adaptors derived from cdr1 Δ/Δ strain had altered size of radius of inhibition zone ( Supplementary Figure 4 ). The diskImageR analysis indicates that the parent cdr1 Δ/Δ strain had RAD₂₀ and FoG₂₀ values of 14.4 ± 0.52 and 0.18 ± 0.03, respectively, while the adaptors had RAD₂₀ and FoG₂₀ values of 14.6 ± 0.60 and 0.41 ± 0.13, respectively ( Figure 6B ). The results of the Student’s t-test show that there was no significant difference in RAD₂₀ between the parent and the adaptors (p>0.05), whereas the adaptors had significantly higher FoG₂₀ values than the parent (p<0.001). Thus, exposure of cdr1 Δ/Δ strain to MCZ mainly induced development of tolerance.

We asked to what extent the cdr1 Δ/Δ derived adaptors could tolerate MCZ. The 14 non-tolerant adaptors as well as the wild type were similarly inhibited by 0.001 μg/mL MCZ, while most of the 67 tolerant adaptors could grow in the presence of 1 μg/mL MCZ. There was no obvious difference in the extent of tolerance among the tolerant adaptors ( Figure 7 ). We sequenced 2 tolerant adaptors (#4 and #5) derived from 0.002 μg/mL MCZ and 4 tolerant adaptors (#1-#4) derived from 0.5 μg/mL MCZ. All 6 adaptors had trisomy of ChrR. Notably, there was no evidence of biased amplification of either ChrR homolog. Three adaptors exhibited amplification of the A homolog (AAB), while the remaining three adaptors showed amplification of the B homolog (ABB) ( Supplementary Figure 3 ).

Heat shock protein 90 (Hsp90) is a highly conserved vital chaperone protein required for the stability and/or folding of hundreds of client proteins. In yeasts, Hsp90 potentiates development of antifungal resistance (Cowen et al., 2006) and is required for the maintenance of tolerance to azoles (Rosenberg et al., 2018; Xu et al., 2021). In order to investigate essentiality of Hsp90 for MCZ tolerance, DDAs were performed using YPD-agar plates supplemented with and without NVP-HSP990. NVP-HSP990 is a potent and selective synthetic small-molecule Hsp90 inhibitor (Menezes et al., 2012).

At both 30°C and 37°C, addition of NVP-HSP990 resulted in bigger ZOI. At 37°C, addition of NVP-HSP990 resulted in disappearance of growth inside the ZOI ( Figure 8A ). Thus, inhibition of Hsp990 caused decreased tolerance and resistance to MCZ.

We asked if cells could still develop resistance or tolerance when Hsp90 was inhibited. Approximately one million cells of SC5314 were spread on YPD-agar plate supplemented with NVP-HSP990 and MCZ ( Figure 8B ). 5 days later, randomly 30 adaptors were chosen. DDAs indicated all the adaptors had clear ZOI, and all adaptors had similar size of ZOI as compared to the parent SC5314 ( Figure 8C ). Thus, cells with suppressed Hsp90 could not develop resistance or tolerance to MCZ.

Calcineurin is a conserved Ca²⁺-calmodulin activated protein phosphatase and is involved in calcium- dependent signaling and regulation of several important cellular processes including evolution of drug resistance in both yeasts and filamentous fungi (Juvvadi et al., 2017). Calcineurin is a heterodimer comprised of a catalytic subunit and a regulatory subunit. In C. albicans genome, these two subunits are encoded by CMP1 and CNB1, respectively. CRZ1 encodes a C2H2 zinc-finger domain transcription factor which is the downstream target of calcineurin (Onyewu et al., 2004). Both alleles of these three genes were deleted from SC5314 and the deletion strains were evaluated for MCZ tolerance.

In DDAs, cmp1 Δ/Δ and cnb1 Δ/Δ strains had clear ZOI on YPD at 37°C, while crz1 Δ/Δ strain still had colonies growing inside the ZOI, albeit less obvious than wild type ( Figure 9A , top panel). In spot assays, growth of cmp1 Δ/Δ and cnb1 Δ/Δ strains was completely inhibited by 0.002 μg/mL MCZ at both 30°C and 37°C while crz1 Δ/Δ strain could tolerate 1 μg/mL MCZ at 37°C. Surprisingly, crz1 Δ/Δ strain was more tolerant than wild type at 30°C ( Figure 9B bottom panel). Thus, calcineurin was required for maintaining MCZ tolerance at 37°C, but Crz1 was not required.

We asked if cmp1 Δ/Δ and cnb1 Δ/Δ strains could develop MCZ tolerance or resistance. Approximately one million cells of the deletion strains were spread on YPD-agar plates supplemented with 0.004 μg/mL MCZ. 5 days later, randomly 30 colonies were chosen. DDAs indicated all adaptors derived from cmp1 Δ/Δ and cnb1 Δ/Δ had clear ZOI, and none had obvious altered size of ZOI as compared to the parent ( Figure 9C ). Thus, strains with deletion of CMP1 and CNB1 failed to develop MCZ tolerance or resistance.

We investigated whether the non-tolerant clinical isolate YJB-T490 adapted to MCZ in a similar manner to the reference strain SC5314. Approximately one million cells of YJB-T490 were spread on YPD-agar plates supplemented with 0.004–4 μg/mL MCZ ( Figure 10A ). From each of the plates with 0.008–2 μg/mL MCZ, randomly 9 adaptors were tested. In total, 81 adaptors were tested. Among them, 62 adaptors had obvious growth inside the ZOI, and 19 adaptors had clear ZOI. None of them had reduced ZOI ( Supplementary Figure 5A ). The diskImageR analysis indicates that the parent YJB-T490 strain had RAD₂₀ and FoG₂₀ values of 13.1 ± 0.31 and 0.18 ± 0.04, respectively, while the adaptors had RAD₂₀ and FoG₂₀ values of 13.3 ± 0.45 and 0.49 ± 0.19, respectively ( Supplementary Figure 5B ). The results of the Student’s t-test show that there was no significant difference in RAD₂₀ between the parent and the adaptors (p>0.05), whereas the adaptors had significantly higher FoG₂₀ values than the parent (p<0.001). Spot assay verified that all the 62 tolerant adaptors could grow in the presence of 2 μg/mL MCZ. The parent YJB-T490 and the 19 non-tolerant adaptors could only grow in the presence of 0.004 μg/mL MCZ ( Supplementary Figure 6 ).

We randomly sequenced 26 tolerant adaptors. Two adaptors (#2 and #8) were derived from 0.008 μg/mL MCZ, and three adaptors were derived from each of the 0.015–2 μg/mL MCZ plates. All the 26 adaptors were aneuploid of ChrR. 24 adaptors had the same segmental trisomy of ChrR. The segment ranged from the left telomere to the MRS (Major Repeat Sequence) region. 2 adaptors had whole chromosomal trisomy of ChrR ( Figure 10B ; Supplementary Figure 7 ).

Thus, similar to the reference strain SC5314, the clinical isolate YJB-T490 also adapted to MCZ via gain of tolerance. However, although whole chromosomal trisomy of ChrR happened, the major genomic change was segmental amplification of ChrR.