Novel ERG11 and TAC1b Mutations Associated with Azole Resistance in Candida auris

Candida auris is a novel Candida species that has spread in all continents, causing nosocomial outbreaks of invasive candidiasis. C. auris has the ability to develop resistance to all antifungal drug classes. Notably, many C. auris isolates are resistant to the azole drug fluconazole, a standard therapy for invasive candidiasis.

ABSTRACT

Candida auris is a novel Candida species that has spread in all continents, causing nosocomial outbreaks of invasive candidiasis. C. auris has the ability to develop resistance to all antifungal drug classes. Notably, many C. auris isolates are resistant to the azole drug fluconazole, a standard therapy for invasive candidiasis. Azole resistance in C. auris can result from mutations in the azole target gene ERG11 and/or overexpression of the efflux pump Cdr1. TAC1 is a transcription factor controlling CDR1 expression in C. albicans. The role of TAC1 homologs in C. auris (TAC1a and TAC1b) remains to be better defined. In this study, we compared sequences of ERG11, TAC1a, and TAC1b between a fluconazole-susceptible and five fluconazole-resistant C. auris isolates of clade IV. Among four of the resistant isolates, we identified similar genotypes with concomitant mutations in ERG11 (F444L) and TAC1b (S611P). The simultaneous deletion of tandemly arranged TAC1a/TAC1b resulted in a decrease of MIC for fluconazole. Introduction of the ERG11 and TAC1b mutations separately and/or combined in the wild-type azole-susceptible isolate resulted in a significant increase of azole resistance with a cumulative effect of the two combined mutations. Interestingly, CDR1 expression was not significantly affected by TAC1a/TAC1b deletion or by the presence of the TAC1b S611P mutation, suggesting the existence of Tac1-dependent and Cdr1-independent azole resistance mechanisms. In conclusion, we demonstrated the role of two previously unreported mutations responsible for azole resistance in C. auris, which were a common signature among four azole-resistant isolates of clade IV.

INTRODUCTION

Candida auris has emerged as a novel pathogenic yeast causing nosocomial outbreaks of candidemia and invasive candidiasis (1–3). Four different genotypic clades have been identified from different parts of the world (South Asia, I; East Asia, II; South Africa, III; and South America, IV) (1, 4). Rapid acquisition of antifungal resistance is a particular feature of C. auris and can affect all three current antifungal drug classes (azoles, polyenes, and echinocandins) (1). Most isolates from clades I and III (>90%) exhibit acquired resistance to the azole drug fluconazole (1, 5), which is a cornerstone in the treatment of invasive candidiasis. However, a lower rate of resistance (14%) has been reported among isolates of clade IV (6).

Azole resistance in Candida spp. can result from different mechanisms, such as mutations in hot spots of the target gene ERG11 or overexpression of efflux pumps (transporters) of the ATP binding cassette (ABC) superfamily or the major facilitator superfamily (MFS) (7). CDR1 and CDR2 (two ABC transporters) and MDR1 (an MFS transporter) were shown to play a significant role in azole resistance (8–11). In Candida albicans, overexpression of CDR1/CDR2 and MDR1 is controlled by gain-of-function mutations in the transcription factors TAC1 and MRR1, respectively (12–15).

Comparable mechanisms of resistance have been suggested for C. auris. Several ERG11 mutations have been characterized for C. auris azole-resistant isolates, some of them being associated with specific geographic clades (5, 16). Rybak et al. have recently observed overexpression of CDR1 and MDR1 in azole-resistant C. auris isolates (17). Deletion of CDR1 in one specific isolate resulted in a significant increase of azole susceptibility, while deletion of MDR1 had only a minor impact on itraconazole resistance (17). In addition, TAC1 orthologs (TAC1a and TAC1b) were annotated in C. auris, and TAC1b was recently found to play a role in azole resistance of C. auris (18, 19). Notably, several TAC1b mutations have been characterized for their potential contribution to azole resistance (18). While these results were derived from analyses performed on isolates of clade I, ERG11 and TAC1b mutations and their impact on azole resistance of isolates from other clades, notably clade IV, remain to be investigated.

In this work, we describe the respective and cumulative role in azole resistance of two previously unreported mutations in ERG11 (F444L) and TAC1b (S611P), which were a common feature of four C. auris azole-resistant isolates of clade IV.

RESULTS

Gene sequencing revealed novel single nucleotide polymorphisms (SNPs) in ERG11 and TAC1b.

We obtained six C. auris clinical isolates from Colombia (clade IV) (20), including a fluconazole-susceptible isolate (fluconazole MIC: 4 μg/ml), and five fluconazole-resistant isolates (fluconazole MIC: 128 μg/ml) (Fig. 1). The ERG11, TAC1a, and TAC1b genes were sequenced for all these strains. Compared to the fluconazole-susceptible wild-type isolate (IV.1), four fluconazole-resistant isolates (IV.3, IV.4, IV.5, and IV.6) exhibited a nonsynonymous mutation in ERG11 (F444L) and another one in TAC1b (S611P) (Fig. 1). These mutations have not been previously reported. Another fluconazole-resistant isolate (IV.2) exhibited a single mutation in TAC1b (F214L). A mutation at this position (F214S) has been already described by Rybak et al. for C. auris isolates from clades II and IV and for three fluconazole-evolved strains, suggesting that it may be a gain-of-function (GOF) mutation for azole resistance (18). We did not observe any genotypic variations in TAC1a among our isolates (Fig. 1).

FIG 1.

Amino acid sequences corresponding to ERG11, TAC1a, and TAC1b for six C. auris isolates of clade IV. Bars in different colors indicate different fluconazole susceptibility levels: green, fluconazole susceptible (MIC 4 μg/ml); red, fluconazole resistant (MIC 128 μg/ml). In the absence of defined clinical breakpoints for C. auris, the cutoffs of ≤32 mg/liter and >32 mg/liter were used for defining susceptibility and resistance, respectively, according to the recommendations of the Centers for Disease Control and Prevention (https://www.cdc.gov/fungal/candida-auris/c-auris-antifungal.html). The PCR sequencing results were translated into amino acid sequences and compared to those for the fluconazole-susceptible isolate IV.1. The nonsynonymous mutations are marked in yellow.

Distinct impact of TAC1 deletion in different C. auris isolates.

To further assess the role of TAC1 in C. auris, we performed double deletion of TAC1a and TAC1b by CRISPR-Cas9 in two fluconazole-resistant C. auris isolates harboring different TAC1 genotypes: isolate IV.2 (TAC1b^F214L) and isolate IV.3 (TAC1b^S611P). Double TAC1 deletion in these two isolates (to generate the IV.2 tac1aΔ/tac1bΔ and IV.3 tac1aΔ/tac1bΔ strains, respectively) resulted in different impact on azole susceptibility (Table 1 and Fig. 2). The IV.2 tac1aΔ/tac1bΔ strain exhibited a modest (2-fold) decrease of fluconazole and voriconazole MIC compared to that of the parental strain. The effect of TAC1 deletion was more pronounced in the IV.3 tac1aΔ/tac1bΔ strain with respect to fluconazole (4-fold MIC decrease) and similar with respect to voriconazole (2-fold MIC decrease).

TABLE 1.

MIC values of the parental strains and their derived mutant strains

Isolate	Construction	MICa (μg/ml) of:
		Fluconazole	Voriconazole
IV.2	Wild type	128	1
tac1aΔ/tac1bΔ	64	0.5
IV.3	Wild type	128	1
tac1aΔ/tac1bΔ	32	0.5
IV.1	Wild-type	4	0.06
TAC1bS611P	16	0.25
ERG11F444L	16	0.25
TAC1bS611P ERG11F444L	64	1

^aMIC values were determined according to CLSI protocol (27).

FIG 2.

Drug susceptibility testing by spotting assay of C. auris tac1aΔ/tac1bΔ mutants and their respective background strains. Spotting assays were performed with serial dilutions of yeast cells spotted onto YEPD agar plates containing a gradient of concentrations of fluconazole (A) or voriconazole (B) and in the absence of azoles (C). Concentrations of drugs were chosen according to the MICs of the tested strains and are indicated. Plates were incubated for 24 h at 37°C.

These results suggest that the impact of TAC1 deletion may differ across different strains, which may be related to different TAC1 amino acid substitutions representing GOF mutations for azole resistance. Notably, the impact of TAC1 deletion on fluconazole susceptibility was significant in the IV.3 strain harboring the S611P mutation in TAC1b (previously unreported).

Respective roles of F444L (ERG11) and S611P (TAC1b) mutations in C. auris azole resistance.

In order to investigate the respective roles of the F444L (ERG11) and S611P (TAC1b) mutations that were present in four fluconazole-resistant isolates, we introduced these mutations separately and in combination in the wild-type azole-susceptible isolate IV.1. The presence of the S611P mutation in TAC1b (i.e., strain IV.1 TAC1b^S611P) resulted in a 4-fold increase of fluconazole and voriconazole MICs compared to those for the parental strain (Table 1 and Fig. 3). The presence of the F444L mutation in ERG11 (i.e., strain IV.1 ERG11^F444L) also resulted in a 4-fold increase of fluconazole and voriconazole MICs compared to those for the parental strain (Table 1 and Fig. 3). Introduction of the double mutations S611P (TAC1b) and F444L (ERG11) (i.e., strain IV.1 TAC1b^S611P ERG11^F444L) resulted in a 16-fold increase of fluconazole and voriconazole MICs compared to those for the parental strain (Table 1 and Fig. 3).

FIG 3.

Drug susceptibility testing by spotting assay of C. auris IV.1 and its derived mutant strains harboring the S611P (TAC1b) and/or F444L (ERG11) mutations. Spotting assays were performed with serial dilutions of yeast cells spotted onto YEPD agar plates containing a gradient of concentrations of fluconazole (A) or voriconazole (B) and in the absence of azoles (C). Concentrations of drugs were chosen according to the MICs of the tested strains and are indicated. Plates were incubated for 24 h at 37°C.

These results confirm the respective roles of the S611P (TAC1b) and F444L (ERG11) mutations in the azole resistance of C. auris and their cumulative effects.

Link between TAC1b and CDR transporter expression.

CDR1 and CDR2 expression was measured by real-time reverse transcription-PCR (RT-PCR) in our wild-type and recombinant C. auris isolates. We observed a modest (about 2-fold) increase of CDR1 and CDR2 expression in the TAC1 deletion strains (IV.2 tac1aΔ/tac1bΔ and IV.3 tac1aΔ/tac1bΔ) compared to that of their respective parental strains, but this difference was not statistically significant, with the exception of CDR2 overexpression in IV.3 tac1aΔ/tac1bΔ (P = 0.001) (Fig. 4). In the IV.1 TAC1b^S611P strain, CDR1 expression was similar to that of its parental strain, while we observed an increase (about 2-fold) of CDR2 expression, which was just beyond the cutoff of statistical significance (P = 0.058) (Fig. 4).

FIG 4.

Relative expression of C. auris CDR1 (A) and CDR2 (B) in different C. auris isolates and their respective TAC1 mutant strains. Results are expressed as fold change compared to the azole-susceptible isolate IV.1. Bars represent means with standard deviations of three biological replicates. Numbers for comparisons (double black arrows) are P values calculated by t test. **, P value ≤ 0.05 (i.e., statistically significant).

In agreement with previous observations (19), these results further support the existence of CDR-independent pathways of azole resistance that are under the control of Tac1 in C. auris.

DISCUSSION

The ability to rapidly develop fluconazole resistance is a major characteristic of the emerging yeast pathogen C. auris, and associated mechanisms seem to be multiple and complex. Several mutations were previously identified in ERG11 of C. auris (ERG11 V125A, F126L, Y132F, and K143R) (1, 5, 16), which correspond to well-known hot spot mutations of fluconazole resistance in C. albicans (21). In this study, we identified a novel (previously unreported) ERG11 mutation (F444L), which was present in four C. auris clinical isolates of clade IV, and we demonstrated its role in azole resistance. The position of this mutation corresponds to a heme-binding region and may therefore affect the positioning and affinity of azole drugs to the modified protein structure (22). Besides mutations in the azole target gene, overexpression of the multidrug transporters (CDR1/CDR2) under the control of the transcription factor Tac1 has also been recognized as a relevant mechanism of azole resistance in C. albicans (12). Mayr et al. have performed separate deletions of TAC1a and TAC1b in two isolates (clades III and IV) with high levels of fluconazole resistance (MIC ≥256 μg/ml). They observed no impact of TAC1a deletion on azole susceptibility and a modest effect of TAC1b deletion (about 2-fold MIC decrease) (19). Both strains harbored an ERG11 mutation (Y132F and F126L) with a known role in azole resistance, which may have limited the impact of TAC1 deletions. In our present work, we performed double TAC1a and TAC1b deletion in two isolates of clade IV (one with a wild-type ERG11 genotype and another one with a previously unreported mutation at position F444). The impact of these deletions on azole susceptibility was relatively modest, with fluconazole MICs that remained relatively high (≥32 μg/ml).

The TAC1 effect on azole resistance could be mainly dependent on GOF mutations. Rybak et al. demonstrated the role of the TAC1b A640V mutation, which induced an 8-fold increase of fluconazole resistance when introduced in a susceptible isolate (18). They also identified other TAC1b mutations (R495G and F214S) among in vitro-evolved fluconazole-resistant strains, which were also observed in several clinical isolates (18). A mutation at position 214 was also detected in one of the fluconazole-resistant clinical isolates in the present study (IV.2), but it was associated with another amino acid substitution (F214L instead of F214S). In addition, we identified a novel TAC1b mutation (S611P) in several isolates of clade IV. We further demonstrated the GOF effect of this mutation, which induced a significant MIC increase (4-fold) for both fluconazole and voriconazole when it was introduced in an azole-susceptible strain with a TAC1b wild-type allele. Interestingly, this TAC1b mutation was repeatedly observed in four clinical isolates of clade IV harboring a concomitant F444L mutation in ERG11. We further demonstrated the cumulative effect of these two mutations (16-fold MIC increase for both fluconazole and voriconazole). Of note, we did not observe this particular genotype of combined S611P (TAC1b) and F444L (ERG11) in several clinical isolates from other clades (I, II, and III), for which we have performed complete TAC1b and ERG11 sequencing (data not shown). Whether these mutations are clade IV specific should be further investigated.

The link between Tac1 and Cdr transporters remains an open question. Rybak et al. showed the role of Cdr1 in azole resistance of C. auris (17). The GOF mutations that they identified in TAC1b (R495G and F214S) were associated with CDR1 overexpression (18). We did not identify any significant change of CDR1 expression in the strain in which the S611P mutation was introduced (IV.1 TAC1b^S611P). However, this strain exhibited some increase of CDR2 expression (albeit not reaching the criteria of statistically significance). The clinical fluconazole-resistant isolate harboring this S611P mutation (IV.3) had CDR1 and CDR2 expression similar to that of the TAC1b wild-type fluconazole-susceptible strain (IV.1). These findings suggest that the S611P mutation in TAC1b governs azole resistance by downstream effectors that are Cdr independent and still need to be discovered. The existence of such alternative pathways is also suggested by the work of Mayr et al., in which no significant CDR1 expression change was measured following TAC1b deletion in two isolates from clades III and IV (19). Unexpectedly, we observed increased expression (about 2-fold) of both CDR1 and CDR2 in our two TAC1a/TAC1b deletion strains of clade IV, which was, however, statistically significant only for CDR2 in the IV.3 tac1aΔ/tac1bΔ mutant.

In conclusion, this work identified two novel mutations in ERG11 and TAC1b that were associated with azole resistance. Interestingly, these mutations were observed concomitantly in four clinical isolates of clade IV and displayed a cumulative effect on azole resistance. The actual incidence of the tandem mutation among isolates of clade IV, as well as its possible occurrence in other clades, should be further investigated. Our results also further support the existence of Tac1-dependent pathways of azole resistance that are distinct from the expression of multidrug transporters (CDR1/CDR2). Identification of these TAC1 downstream effectors would deserve further investigations in order to identify potential novel therapeutic targets against C. auris.

MATERIALS AND METHODS

Isolates, plasmids, and media.

The C. auris isolates (20) used in this study are listed in Table 2. Plasmids pJK795 (containing the nourseothricin resistance cassette) and pYM70 (containing the hygromycin resistance cassette) were used in this study (23, 24). Isolates were grown in yeast extract-peptone-dextrose (YEPD) medium. Cultures were incubated for 16 to 20 h at 37°C on solid YEPD agar plates or in liquid medium under constant agitation (220 rpm).

TABLE 2.

Description of the strains used in this study

Name	Description	Reference
IV.1 (LMDM 1219)	Clinical isolate	20
IV.2 (LMDM 1231)	Clinical isolate	20
IV.3 (LMDM 1218)	Clinical isolate	20
IV.4 (LMDM 1232)	Clinical isolate	20
IV.5 (LMDM 1242)	Clinical isolate	20
IV.6 (LMDM 1245)	Clinical isolate	20
IV.2 tac1aΔ/tac1bΔ (JLY0001)	IV.2 tac1aΔ/tac1bΔ::NatR	This study
IV.3 tac1aΔ/tac1bΔ (JLY0002)	IV.3 tac1aΔ/tac1bΔ::NatR	This study
IV.1 TAC1bS611P (JLY0003)	IV.1 TAC1bS611P::NatR	This study
JLY0004	IV.1 ERG11F444L::HygR/thr1Δ	This study
JLY0005	IV.1 TAC1bS611P::NatR/ERG11F444L::HygR/thr1Δ	This study
IV.1 ERG11F444L (JLY0006)	IV.1 ERG11F444L/thr1Δ::THR1	This study
IV.1TAC1bS611PERG11F444L (JLY0007)	IV.1 TAC1bS611P::NatR/ERG11F444L/thr1Δ::THR1	This study

The antifungal drugs fluconazole and voriconazole (Sigma-Aldrich, St. Louis, MO) used in this study were obtained as powder suspensions and dissolved in dimethyl sulfoxide (DMSO).

PCR and sequencing.

The TAC1 orthologs of C. auris were identified using the Basic Local Alignment Search Tool (BLAST) of the National Center for Biotechnology Information (NCBI). The sequence of the TAC1 (orf19.3188) gene of the C. albicans SC5314 isolate was BLAST searched against the sequence of an Indian C. auris genome (6684), which resulted in the identification of two tandemly arranged TAC1 orthologs (TAC1a/TAC1b). Gene amplification was performed with Phusion high-fidelity DNA polymerase (Thermo Fisher Scientific Inc.), and DNA was purified with the QIAquick PCR purification kit (Qiagen Inc.). Sequencing was performed by the Sanger method by Microsynth (Balgach, Switzerland). All primers for PCR and sequencing are listed in Table S1 in the supplemental material.

Construction of C. auris mutants.

Double mutants of both C. auris TAC1 orthologs TAC1a and TAC1b were constructed in two different parental isolates, IV.2 and IV.3 (Table 2), on the basis of a CRISPR-Cas9 approach with corresponding DNA repair fragments. The gene disruption fragments containing homolog regions and selection cassette NatR (nourseothricin resistance) were obtained by fusion PCR of three PCR fragments with overlapping sequences. The first PCR fragment amplified the 5′-flanking region (520 bp) of TAC1a with primers TAC1del_fusion_PF1 and TAC1del_fusion_PR1. The second PCR fragment amplified the marker NatR from plasmid pJK795 with primers TAC1del_fusion_PF2 and TAC1del_fusion_PR2. The third PCR fragment amplified the 3′-flanking region (520 bp) of TAC1b with primers TAC1del_fusion_PF3 and TAC1del_fusion_PR3. The final PCR was performed with the three purified fragments using the QIAquick PCR purification kit (Qiagen Inc.) and nested primers TAC1del_fusion_PF4 and TAC1del_fusion_PR4 in the presence of 1.3 M betaine (Fig. S1A).

The Cas9-CRISPR-based approach used RNA-protein complexes (RNPs) reconstituted with purified Cas9 protein combined with scaffold- and gene-specific guide RNAs for transformation as previously described (25). Designs of the constructs are shown in Fig. S1B. CRISPR gene-specific RNA guides were designed to contain 20-bp homologous sequences of the upstream region of TAC1a and the downstream region of TAC1b. The mix of the guide RNAs, the Cas9 nuclease 3NLS (Integrated DNA Technologies [IDT]), and tracrRNA (a universal transactivating CRISPR RNA) were prepared according to the protocol previously described (26). Transformation of C. auris cells was performed by electroporation with about 1 μg of gene disruption construct as previously described (26). Transformants were selected at 35°C on YEPD agar containing 200 μg/ml of nourseothricin (Werner BioAgents). The colonies were verified by PCR screening for integration of the marker NatR as well as by PCR in the gene TAC1b to ensure the deletion of the genes (Fig. S2).

For construction of the IV.1 TAC1b^S611P strain, the gene repair fragments containing homologous regions comprising S611P in TAC1b of strain IV.3 and selection cassette NatR (nourseothricin resistance) were obtained by fusion PCR of three PCR fragments with overlapping sequences. The first PCR fragment amplified the region containing the mutation S611P of TAC1b (1,482 bp) from isolate IV.3 and with primers S611Pinduit_PF1 and S611Pinduit_PR1. The second PCR fragment amplified the marker NatR from plasmid pJK795 with primers S611Pinduit_PF2 and S611Pinduit_PR2. The third PCR fragment amplified the 3′-flanking region of the terminator of TAC1b (538 bp) with primers S611Pinduit_PF3 and S611Pinduit_PR3. The final PCR was performed with the three purified fragments and QIAquick PCR purification kit (Qiagen Inc.) and nested primers S611Pinduit_PF4 and S611Pinduit_PR4 in the presence of 1.3 M betaine (Fig. S3A).

One CRISPR gene-specific RNA guide containing 20 bp was designed between the two homologous regions, while another CRISPR gene-specific RNA guide recognized the downstream sequence of the 3′-homologous region (Fig. S3B). The preparation of CAS9 mix and the transformation via electroporation were as described above. The colonies were verified by PCR screening for integration of the marker NatR (Fig. S4) as well as by TAC1b sequencing to ensure S611P introduction.

For the construction of IV.1 ERG^F444L and IV.1 TAC1b^S611PERG1^F444L strains, the gene repair fragments containing homologous regions comprising F444L in ERG11 of strain IV.3 and the selection cassette HygR (hygromycin B resistance) were also obtained by fusion PCR of three PCR fragments with overlapping sequences. The first PCR fragment amplified the region containing the mutation F444L of ERG11 (943 bp) from isolate IV.3 and primers F444Linduit_PF1 and F444Linduit_PR1′. The second PCR fragment amplified the marker HygR from plasmid pYM70 with primers F444Linduit_PF2′ and F444Linduit_PR2. The third PCR fragment amplified the 3′-flanking region of the terminator of ERG11 (426 bp) with primers F444Linduit_PF3 and F444Linduit_PR3. The final PCR was performed with the three purified fragments and the QIAquick PCR purification kit (Qiagen Inc.) and nested primers F444Lnduit_PF4′ and F444Linduit_PR4′ in the presence of 1.3 M betaine (Fig. S5A).

Designs of the CRISPR-Cas9 constructs are shown in Fig. S5B. One CRISPR gene-specific RNA guide containing 20 bp was designed between the two homologous regions, while another CRISPR gene-specific RNA guide recognized the downstream sequence of the 3′-homologous region. Transformation via electroporation was performed as described above. Transformants were selected at 35°C on YEPD agar containing 600 μg/ml of hygromycin B (Corning). The colonies were verified by PCR screening for integration of the marker HygR (Fig. S6), followed by ERG11 sequencing.

Because the HygR cassette truncated the opening reading frame (ORF) of THR1 (CJI97_001157, homoserine kinase involved in threonine biosynthesis) downstream of ERG11, we complemented this gene in recombinant strains. The complementation cassette was constructed by PCR with primers Cauris_ERG11_PrimerF1 and F444Linduit2_verif_PR, which includes ORFs of ERG11 and THR1 from IV.3 (Fig. S7). Transformation was performed via electroporation as described above. The colonies were selected by minimum medium YNB (BIO 101, Inc.) with 2% glucose.

All primers used for these constructs and PCRs are described in Table S1.

Antifungal susceptibility testing.

MIC of fluconazole and voriconazole was determined for all C. auris isolates according to the procedure of the Clinical and Laboratory Standards Institute (CLSI) (27). Antifungal susceptibility was also assessed by a spotting assay using 10-fold dilutions of yeast suspension, from 10⁷ cells/ml to 10⁴ cells/ml, spotted (5 μl) onto YEPD plates with or without the test drugs (fluconazole and voriconazole). Plates were incubated for 24 h at 37°C.

Real-time RT-PCR.

Each C. auris isolate was grown overnight in 5 ml of liquid YEPD under constant agitation at 37°C. Overnight cultures were diluted to a density of 0.75 × 10⁷ cells/ml in 15 ml of fresh YEPD and were grown at 37°C under constant agitation for 2 to 3 h to reach a density of 1.5 × 10⁷ cells/ml. Total RNA was extracted with the Quick-RNA fungal/bacterial miniprep kit (Zymo Research), and total RNA extracts were treated with DNase using the DNA-free kit (Thermo Fisher Inc.). The concentration of purified RNA was measured with a NanoDrop 1000 instrument (Witec AG, Switzerland). RNA was stored at −80°C until use. Each isolate was prepared in biological triplicates.

One microgram of RNA of each isolate was converted into cDNA using the Transcriptor high-fidelity cDNA synthesis kit (Roche, Switzerland). RT-PCR was performed in 96-well plates using a master mix (PowerUp SYBR green master mix; Applied Biosystems, USA) with primers of the targeted genes ACT1, CDR1, and CDR2 (Table S1) and supplemented with nuclease-free water up to 20 μl for each reaction. The primers used for ACT1, CDR1, and CDR2 mRNA amplification were the same as previously described (17). Each experiment was performed in biological and technical triplicates. The StepONE software program including a melt curve stage was used with activation at 95°C for 10 min, 40 cycles of denaturation at 95°C for 15 s, and annealing/extension at 60°C for 1 min. Gene expression was calculated with the threshold cycle (2^−ΔΔCT) method (28). Results were analyzed by the t test method (GraphPad).

Data availability.

The complete sequencing data set has been deposited in GenBank under accession numbers MW368389 to MW368409.