An A643V Amino Acid Substitution in Upc2p Contributes to Azole Resistance in Well-Characterized Clinical Isolates of Candida albicans

Abstract

The Candida albicans Upc2p transcription factor regulates ERG11, encoding the target of azole drugs. Gain-of-function mutations that contribute to resistance were recently identified in a series of sequential clinical isolates (N. Dunkel, T. T. Liu, K. S. Barker, R. Homayouni, J. Morschhauser, and P. D. Rogers, Eukaryot. Cell 7:1180-1190, 2008). In the present study, UPC2 was sequenced from a matched set of 17 isolates. An A643V substitution was present in all of the isolates in the series that overexpressed ERG11. Azole susceptibility, ergosterol levels, and expression of ERG genes were elevated in the A643V clinical isolates and in reconstructed strains.

The ergosterol biosynthesis pathway in budding yeast is tightly regulated by a number of transcriptional activators and repressors (3, 5, 10, 11). In Saccharomyces cerevisiae, the paralogous transcription factors ScUpc2p and ScEcm22p activate the expression of genes required for ergosterol biosynthesis (ERG genes) in response to sterol depletion during aerobic growth (6, 17). An early study on ScUPC2 identified a dominant gain-of-function mutation in ScUPC2 that resulted in a glycine-to-aspartic acid substitution at position 888 (G888D) within the protein sequence (5). This mutation, termed UPC2-1, conferred constitutive activity on the transcription factor, resulting in increased aerobic expression of genes required for sterol uptake, as well as increased expression of ERG genes.

In Candida albicans, gain-of-function mutations in several transcription factors, such as TAC1 and MRR1, in drug-resistant clinical isolates have been previously documented (4, 7). In fact, these trans-acting mutations have been central to regulation of many of the molecular mechanisms of antifungal resistance, such as CDR1 and MDR1 overexpression. Gain-of-function mutations in UPC2 that result in G648D or A643T substitutions in drug-resistant isolates of C. albicans were recently identified (8, 9). In these studies, the UPC2 point mutations were found to confer a decrease in azole susceptibility and increase expression of Upc2p-regulated genes. The two mutations are in close proximity in the C-terminal portion of the protein, which may indicate that this is a hot spot for mutations or that this region of the protein is important for regulation of its functional activity.

Previous studies of C. albicans azole-resistant clinical isolates revealed a role for overexpression of ERG11, which encodes the enzyme target of the azole drugs (18, 19). In these strains, it is thought that increased expression of the Erg11p target enzyme contributes to a decrease in the susceptibility of the strain to azoles. Genome-wide expression profiling has expanded this view to include overexpression of other ERG genes in association with reduced azole susceptibility (15). Increased ERG gene expression in these strains may result from cis-acting mutations within the promoters of the ERG genes themselves or from alterations in trans-acting factors, such as Upc2p. Because some strains that overexpress ERG11 also show increased expression of other ERG genes, it is likely that alterations in trans-acting factors are responsible for the coordinated upregulation of these genes. Because Upc2p directly activates the transcription of many ERG genes, alterations of the Upc2p activity may be responsible for overexpression of these genes in azole-resistant clinical isolates of C. albicans. The goal of this work was to investigate a well-characterized matched set of C. albicans clinical isolates for point mutations in UPC2 that result in ERG11 overexpression. This analysis should not only contribute to our understanding of how CaUPC2 functions but should also expand our knowledge of how azole resistance develops clinically in C. albicans.

C. albicans azole-resistant clinical isolates in a matched set contain an alanine-to-valine amino acid substitution at the Upc2p C terminus.

A well-characterized set of 17 C. albicans matched clinical isolates in which resistance developed over the course of azole therapy was analyzed. In this set of isolates, a number of mechanisms contribute to resistance, including overexpression of ERG11 (18). The 17 isolates were taken from an AIDS patient who had received a 2-year azole prophylactic treatment regimen. The overexpression of ERG11 began in isolate 13 and may have been due to alterations in the ERG11 promoter sequence and/or in trans-acting factors such as Upc2p. To identify UPC2 point mutations that might contribute to resistance, the UPC2 coding sequence was PCR amplified using oligonucleotides as previously described (17) and sequenced from each isolate in the series. Interestingly, there was a cytosine-to-thymidine transition mutation at position 1928 that occurred in strain 13 and was maintained as a heterozygous mutation through the rest of the series. This mutation within UPC2 results in an alanine-to-valine substitution at amino acid position 643 (A643V).

A643V causes increases in FLC MIC, ergosterol level, and ERG gene overexpression.

Previous work has shown that an increase in fluconazole (FLC) MIC occurs in isolate 13 and that this increase is accompanied by an increase in expression of ERG11 (18), correlating the A643V mutation with azole resistance in this series. Azole susceptibilities of the 17 isolates in the series were determined using the Etest method. As with previous data, there was a 2-fold increase in FLC MIC from 2 μg/ml in strain 12 to 4 μg/ml in strain 13, associating the acquisition of the A6443V mutation with an increase in FLC MIC. To confirm that this increase in MIC was due to the A643V mutation, laboratory mutants expressing the wild type (WT) (A643) and the A643V allele amplified from strain 17 were created. Both alleles were PCR amplified and cloned into pGEM-HIS vector (22) according to the strategy used previously in this laboratory (Seattle Biomedical Research Institute) (16). The WT and A643V constructs were introduced into a upc2Δ upc2Δ deletion strain (16). The strain expressing the A643V allele showed a 2-fold increase in MIC by both Etest and broth microdilution compared to the strain constructed to express the WT allele from strain 17.

Ergosterol quantitation (as done previously [2]) was performed on representative strain 12, which does not contain A643V, and isolate 13, which does contain the A643V allele, as well as on the reconstructed laboratory strains. Samples were collected after 48 h of growth at 30°C. Ergosterol quantitation showed that strain 13 had a higher total cellular ergosterol content than strain 12 (Fig. 1 A). The increase in ergosterol content occurred in the series with the strain in which the A643V mutation appears, associating the mutation in UPC2 with altered activity of the sterol biosynthetic pathway. The increase in ergosterol content was confirmed to be due to the presence of the A643V copy of UPC2, as the laboratory strains created that express this allele showed an increase in ergosterol content relative to the strains expressing the nonmutated allele (Fig. 1B).

FIG. 1.

Ergosterol quantitation of C. albicans clinical and laboratory strains expressing UPC2 A643V. (A) Increased ergosterol content in a clinical isolate expressing UPC2 A643V (strain 13) compared to the content seen with a matched isolate with nonmutated UPC2 (strain 12). Strains were grown for 48 h, and 100 optical density (OD) units of cells were harvested for heptane extraction of total ergosterol. (B) Introduction of the UPC2 A643V allele confers an increase in ergosterol content compared to the WT (A643) allele level. EC-2 is the previously published UPC2 reconstruction strain. All three strains have a single UPC2 allele. Strains were grown for 48 h, and the graph represents the ergosterol scan results of heptane extraction of 100 OD units of cells. Experiments were repeated three times with similar results. A representative experiment is shown; error bars are derived from the results determined from triplicate samples within that experiment. Where error bars are not visible, they are smaller than the symbols shown.

The MIC and ergosterol scan data suggest that A643V acts as a hyperactivating mutation. This idea was addressed by performing quantitative reverse transcription-PCR (qRT-PCR) to assess ERG gene expression levels in isolates containing the A643V allele (isolates 13, 14, and 17) compared to the levels seen with those without A643V (isolate 12) as well as the levels seen with the reconstructed strains. RNA samples were extracted from 6-h cultures, and QRT-PCR was performed as described previously (14). ERG gene transcript levels were analyzed by determination of the levels of induction of the A643V strains relative to strain 12. The majority of the ERG genes exhibited increased expression in the A643V strains compared to strain 12 (Fig. 2 A). In A643V clinical isolates 13, 14, and 17, expression was increased for each of the genes tested by QRT-PCR, with the exception of ERG3 expression in strain 13, which was not increased in this particular experiment. In the A643V reconstructed strain, expression was increased for each of the genes tested, except ERG1, compared to that seen with the WT UPC2 allele (Fig. 2B). UPC2 expression is thought to be partially controlled by a transcriptional autoregulatory mechanism (1, 12, 21, 23), and all strains expressing the A643V allele also showed increased UPC2 transcript levels, except for strain 13 in the experiment, whose results are shown in Fig. 2. The coordinate upregulation of multiple ERG genes suggests that a trans-acting mutation in UCP2 is responsible for this increased expression rather than simply being the result of cis-acting mutations. However, cis-acting mutations within ERG gene promoters may also contribute to overexpression in the clinical isolates.

FIG. 2.

Expression levels of Upc2p-regulated genes are increased due to the A643V mutation. (A) Expression of ERG genes and UPC2 is increased in clinical isolates containing the A643V allele relative to the expression of a matched isolate not containing the UPC2 mutation. QRT-PCR was performed using RNA extracted from cells grown for 6 h. All data are normalized to an internal ACT1 control and are expressed as fold induction relative to the expression level in strain 12. (B) Expression of ERG genes and UPC2 is increased in the laboratory-reconstructed strain expressing A643V compared to the expression seen with the WT (A643) allele of UPC2. QRT-PCR was performed using RNA extracted from cells grown for 6 h. All data are normalized to an internal ACT1 control and are expressed as fold induction relative to the expression level in the BWP17 WT parental strain.

The A643V mutation and other gain-of-function mutations in Upc2p (8, 9) may result in hyperactivity due to release from a repressor. Work done in studies of S. cerevisiae has suggested that an important component of Upc2p regulation involves a posttranslational control mechanism that may be due to binding by a repressor protein (6, 17). An alternative model in which Upc2p is membrane bound is based on the mammalian sterol regulator SREBP and was proposed in a recent review (20). This model is supported by S. cerevisiae localization experiments that show that the C-terminally tagged Upc2p is not localized in the nucleus (13). In either model, the UPC2 A643V mutation could interfere with normal functioning of the protein by affecting the C-terminal regulatory domain.

These results suggest that a steric or conformational constraint is involved in the regulation of Upc2p, which is consistent with either of the models presented above. Most importantly, these mutations can be found in drug-resistant isolates of C. albicans, stressing the relevance of UPC2 to our understanding of how azole resistance develops in the clinical setting.