Regulation of efflux pump expression and drug resistance by the transcription factors Mrr1, Upc2, and Cap1 in Candida albicans

Abstract

Constitutive overexpression of the Mdr1 efflux pump is an important mechanism of acquired drug resistance in the yeast Candida albicans. The zinc cluster transcription factor Mrr1 is a central regulator of MDR1 expression, but other transcription factors have also been implicated in MDR1 regulation. To better understand how MDR1-mediated drug resistance is achieved in this fungal pathogen, we studied the interdependence of Mrr1 and two other MDR1 regulators, Upc2 and Cap1, in the control of MDR1 expression. A mutated, constitutively active Mrr1 could upregulate MDR1 and confer drug resistance in the absence of Upc2 or Cap1. On the other hand, Upc2 containing a gain-of-function mutation only slightly activated the MDR1 promoter, and this activation depended on the presence of a functional MRR1 gene. In contrast, a C-terminally truncated, activated form of Cap1 could upregulate MDR1 in a partially Mrr1-independent fashion. The induction of MDR1 expression by toxic chemicals occurred independently of Upc2 but required the presence of Mrr1 and also partially depended on Cap1. Transcriptional profiling and in vivo DNA binding studies showed that a constitutively active Mrr1 binds to and upregulates most of its direct target genes in the presence or absence of Cap1. Therefore, Mrr1 and Cap1 cooperate in the environmental induction of MDR1 expression in wild-type C. albicans, but gain-of-function mutations in either of the two transcription factors can independently mediate efflux pump overexpression and drug resistance.

Fig. 1.

Activation of the MDR1 promoter by hyperactive UPC2 and MRR1 alleles. C. albicans strains carrying a P_MDR1-GFP reporter fusion in the indicated genetic backgrounds were grown to log phase in YPD medium, and the mean fluorescence of the cells was determined by flow cytometry. The results obtained with two independently generated reporter strains are shown in each case (with means and standard deviations from three experiments). The following strains were used (see also Table S1 in the supplemental material): SCMPG2A and -B (wild type), SCMRR1M4MPG2A and -B (mrr1Δ), UPC2M4MPG2A and -B (upc2Δ), SCUPC2R12MPG2A and -B (UPC2/UPC2^G648D, wild type), SCUPC2R14MPG2A and -B (UPC2^G648D/UPC2^G648D, wild type), Δmrr1UPC2R12MPG2A and -B (UPC2/UPC2^G648D, mrr1Δ), Δmrr1UPC2R14MPG2A and -B (UPC2^G648D/UPC2^G648D, mrr1Δ), SCMRR1R32MPG2A and -B (MRR1/MRR1^P683S, wild type), SCMRR1R34MPG2A and -B (MRR1^P683S/MRR1^P683S, wild type), Δupc2MRR1R32MPG2A and -B (MRR1/MRR1^P683S, upc2Δ), Δupc2MRR1R34MPG2A and -B (MRR1^P683S/MRR1^P683S, upc2Δ). The background fluorescence of the parental strains, which do not contain the GFP gene, is indicated by the black part in each column (one measurement).

Fig. 2.

Activation of the MDR1 promoter by hyperactive CAP1 and MRR1 alleles. C. albicans strains carrying a P_MDR1-GFP reporter fusion in the indicated genetic backgrounds were grown to log phase in YPD medium, and the mean fluorescence of the cells was determined by flow cytometry. The results obtained with two independently generated reporter strains are shown in each case (means and standard deviations from three experiments). The following strains were used (see also Table S1 in the supplemental material): SCMPG2A and -B (wild type), SCMRR1M4MPG2A and -B (mrr1Δ), SCCAP1M4MPG2A and -B (cap1Δ), SCCAP1R12MPG2A and -B (CAP1/CAP1^ΔC333, wild type), SCCAP1R14MPG2A and -B (CAP1^ΔC333/CAP1^ΔC333, wild type), Δmrr1CAP1R12MPG2A and -B (CAP1/CAP1^ΔC333, mrr1Δ), Δmrr1CAP1R14MPG2A and -B (CAP1^ΔC333/CAP1^ΔC333, mrr1Δ), SCMRR1R32MPG2A and -B (MRR1/MRR1^P683S, wild type), SCMRR1R34MPG2A and -B (MRR1^P683S/MRR1^P683S, wild type), Δcap1MRR1R32MPG2A and -B (MRR1/MRR1^P683S, cap1Δ), Δcap1MRR1R34MPG2A and -B (MRR1^P683S/MRR1^P683S, cap1Δ). The background fluorescence of the parental strains, which do not contain the GFP gene, is indicated by the black portion of each column (one measurement). Data are from the same experiments as in Fig. 1, and the values of the control strains are included for comparison.

Fig. 3.

Effects of combining hyperactive MRR1 and CAP1 alleles on MDR1 promoter activity and drug resistance. (A) C. albicans strains that are homozygous for the indicated hyperactive MRR1 and CAP1 alleles and contain a P_MDR1-GFP reporter fusion were grown to log phase in YPD medium, and the mean fluorescence of the cells was determined by flow cytometry. The results obtained with two independently generated reporter strains are shown in each case (means and standard deviations from three experiments). The following strains were used (see also Table S1 in the supplemental material): SCMPG2A and -B (wild type), SCMRR1R34MPG2A and -B (MRR1^P683S), SCCAP1R14MPG2A and -B (CAP1^ΔC333), SCMRR1R34CAP1R14MPG2A and -B (MRR1^P683S + CAP1^ΔC333). Data are from the same experiments as in Fig. 1 and 2, and the values of the control strains are included for comparison. (B) Susceptibilities to fluconazole and cerulenin of the wild-type parental strain SC5314 and mutant derivatives in which both resident MRR1 and/or CAP1 alleles were replaced by the hyperactive MRR1^P683S and CAP1^ΔC333 alleles, respectively. The results obtained with two independently generated strains are shown in each case. The following strains were used: SC5314 (wild type), SCMRR1R34A and -B (MRR1^P683S), SCCAP1R14A and -B (CAP1^ΔC333), SCMRR1R34CAP1R14A and -B (MRR1^P683S + CAP1^ΔC333).

Fig. 4.

Activation of the MDR1 promoter by benomyl and H₂O₂ in wild-type, mrr1Δ, cap1Δ, and upc2Δ strains. Overnight cultures of C. albicans strains carrying a P_MDR1-GFP reporter fusion in the indicated genetic backgrounds were diluted 10⁻² in three tubes with fresh YPD medium and grown to log phase. One culture was left untreated, and 50 μg/ml benomyl or 0.005% H₂O₂ was added to the other cultures to induce MDR1 expression. The cultures were incubated for 80 min, and the mean fluorescence of the cells was determined by flow cytometry. The results obtained with two independently generated reporter strains are shown in each case (means and standard deviations from three experiments). The following strains were used (see also Table S1 in the supplemental material): SCMPG2A and -B (wild type), SCMRR1M4MPG2A and -B (mrr1Δ), SCCAP1M4MPG2A and -B (cap1Δ), UPC2M4MPG2A and -B (upc2Δ).

Fig. 5.

(A) MDR1 promoter activity in C. albicans mrr1Δ mutants expressing a wild-type MRR1 allele or the MRR1^P683S allele, without or with a C-terminal 3×HA tag, from the endogenous MRR1 promoter or the strong ADH1 promoter. The strains were grown to log phase in YPD medium, and the mean fluorescence of the cells was determined by flow cytometry. The results obtained with two independent transformants are shown in each case (means and standard deviations from three experiments). The following strains were used: CAG48MRR1M4B (mrr1Δ), CAG48MRR1M4K2B1 and -2 (P_MRR1-MRR1), CAG48MRR1M4H2B1 and -2 (P_MRR1-MRR1-HA), CAG48MRR1M4K3B1 and -2 (P_MRR1-MRR1^P683S), CAG48MRR1M4H3B1 and -2 (P_MRR1-MRR1^P683S-HA), CAG48MRR1M4E2B1 and -2 (P_ADH1-MRR1), CAG48MRR1M4EH2B1 and -2 (P_ADH1-MRR1-HA), CAG48MRR1M4E3B1 and -2 (P_ADH1-MRR1^P683S), CAG48MRR1M4EH3B1 and -2 (P_ADH1-MRR1^P683S-HA). (B) Expression of the HA-tagged Mrr1^P683S protein in mrr1Δ single and mrr1Δ cap1Δ double mutants. Whole-cell protein extracts of the strains were analyzed by Western immunoblotting with an anti-HA antibody. The following strains were used: 1, SC5314; 2, SCΔmrr1MEH3A; 3, SCΔmrr1Δcap1MEH3A; 4, SCΔmrr1MEH3B; 5, SCΔmrr1Δcap1MEH3B; 6, SCΔmrr1ME3A; 7, SCΔmrr1ME3B. The HA-tagged Mrr1^P683S was expressed at similar levels in independent transformants of the mrr1Δ (lanes 2 and 4) and mrr1Δ cap1Δ (lanes 3 and 5) mutants. No cross-reacting proteins were detected in the parental strain SC5314 (lane 1) or in mrr1Δ mutants expressing untagged Mrr1^P683S (lanes 6 and 7).

Fig. 6.

Identification of putative Mrr1 DNA binding motifs. (A) Weblogo representation of the DCSGHD motif identified by using SCOPE. (B) Enrichment of the DCSGHD motif in the 710 sequence data set (filled circles), compared to what would be expected randomly in the whole genome (open circles). Each of the 710 1-kb sequences was divided into 20 intervals of 50 bp, and the number of DCSGHD motif occurrences was compiled and plotted for each interval. The maximum fold enrichment value (3.5) is observed in the center of the analyzed 1-kb sequences. (C) Enrichment of the DCSGHD motif in the sequences corresponding to 40 genes bound and regulated by Mrr1 (filled circles) compared to all genes (open circles). The P value (3.E−06) represents the probability of observing this motif distribution in random data sets.