Background

Candida bloodstream infections are an important healthcare issue known to be associated with high morbidity and mortality. There have been increasing reports of antifungal resistance. We have previously reported decreasing azole susceptibilities in our hospital, particular in Candida tropicalis. More than 20% of C. tropicalis were non-susceptible to fluconazole [1]. There are various mechanisms mediating azole resistance. It has been suggested that molecular mechanisms such as presence of mutations may be a predictive marker of clinical failure in Candida infections [2]. Whilst this has been established for echinocandin resistance, azole resistance mechanisms are not as well studied, particularly for non-albicans species. Elucidation of these mechanisms is crucial to make progress in understanding and treating invasive Candida infections.

Methods

In this study, we characterised the molecular mechanisms of azole resistance in 26 fluconazole non-susceptible Candida bloodstream isolates. These isolates were identified from a retrospective surveillance study conducted at a major regional tertiary referral hospital between 2012 and 2015 [1]. In brief, non-duplicate Candida bloodstream isolates from all adult inpatients (at least 21 years old) with temporally-related clinical signs and symptoms of infection admitted to the hospital during the study period were included. Antifungal susceptibility testing was performed using Sensititre YeastOne® YO10 panel (Trek Diagnostics System, West Sussex, England) according to manufacturer’s recommendations. Minimum inhibitory concentrations were interpreted according to the current species-specific clinical breakpoints provided by the Clinical and Laboratory Standards Institute (CLSI) M27-S4 document or epidemiological cut-off values (ECV), where CLSI breakpoints were unavailable [3, 4]. For Candida albicans and C. tropicalis, isolates meeting the susceptible-dose-dependent (SDD) and resistant criteria were included, whereas only resistant Candida glabrata were included in this study. A total of 26 fluconazole non-susceptible isolates [C. albicans - 4/62 (6%); C. glabrata - 5/82 (6%); C. tropicalis - 17/78 (22%); C. parapsilosis - 0/35 (0%)] were identified from 257 Candida spp. isolates included in the surveillance study.

ERG11, CDR1, CDR2 and MDR1 gene expression were quantified in triplicates using real-time reverse transcription-PCR (RTPCR) with total RNA extracted from exponential-phase yeast peptone dextrose broth cultures on a CFX96 Real-Time PCR Detection System (Bio-Rad, USA). The primers used were adopted from previous publications [5,6,7,8,9,10,11], except for C. glabrata CDR1 gene [F – TGGTGTTGCTAATGTCGCCA, R – GTCCCAAGTACTCGCCACAA] and C. glabrata ERG11 gene [F – CCACCCATTGCACTCTTTGT, R – AGAACGTGGTAGTCCCTTGG]. Quantification of target genes was normalised to the level of ACT1, an endogenous reference gene. Relative gene expression was calculated as the fold change in expression of the isolates compared to the respective ATCC reference strains (C. albicans ATCC 90028, C. glabrata ATCC 2950, C. tropicalis ATCC 750). A fold increase of 3 times was considered to be an overexpression of the target gene. The ERG11 gene was amplified and sequenced to identify amino acid mutations by comparing with reference wild-type GenBank sequences (C. albicans – X13296; C. tropicalis – M23673; C. glabrata – L40389).

Results

The susceptibility profiles of the isolates are displayed in Table 1. Cross-resistance to all azoles was observed in all isolates except for one C. albicans (CW138) and two C. glabrata (CW262 and CW378) isolates. All isolates retained susceptibility to other anti-fungals including anidulafungin, caspofungin, micafungin and amphotericin B (data not shown). In C. albicans, all isolates showed non-synonymous homozygous ERG11 substitutions which included three distinct substitutions (A114S, Y257H and E266D). I166S substitutions were detected in two of the six C. glabrata isolates. Of the 17 C. tropicalis isolates, eight (47%) had ERG11 substitutions. The most common substitutions were the concurrent observation of Y132F and S154F, which occurred primarily in resistant isolates with fluconazole MICs ≥8 μg/mL. Only two of the eight ERG11 substitutions were homozygous, and there does not appear to be any correlation of the type of substitutions with MICs. The ERG11 substitutions observed in all of the Candida spp. have been previously reported in literature except for I166S.

Table 1 Molecular characteristics of clinical fluconazole non-susceptible Candida spp. blood isolates
Full size table

Among the different gene targets, it appeared that ERG11 expression levels were mostly similar compared to the respective wild-type reference strains. CDR2 expression was consistently elevated in fluconazole non-susceptible C. albicans. In the two resistant isolates with MIC 128 μg/mL, MDR1 was also up-regulated. CDR1 and CDR2 co-expression was observed in all C. glabrata isolates. Gene overexpression was not consistent among C. tropicalis isolates – there were five isolates with CDR1 overexpression and six isolates with MDR1 overexpression. All C. tropicalis isolates only had overexpression of a single gene target. Interestingly, there were three C. tropicalis isolates with no ERG11 mutations or any gene up-regulation.

Discussion

In this study, we evaluated the molecular mechanisms associated with azole resistance in various Candida species in our institution. Identification of antifungal susceptibilities through phenotypic methods such as MIC testing is often limited by the length of time required. Furthermore, current fungal MIC breakpoint interpretations are not supported by robust clinical data and are not predictive of clinical success/failure. Therefore, there is interest in identifying genotypic markers which could be rapidly identified for use in clinical prediction. Various previous studies have investigated different mechanisms of azole resistance in Candida species [5, 12,13,14]. Some of these studies have identified key ERG11 substitutions which are associated with azole resistance e.g. Y132F, S154F [8, 15] and suggested that these mutations could be potential predictive markers of azole resistance.

In our context, it appeared that there was an interplay of various different mechanisms, including mechanisms which were not studied here, responsible for azole resistance in Candida spp. ERG11 mutations were commonly detected in C. albicans, whereas the role of overexpression of azoles efflux pumps appeared to be more prominent in C. albicans (CDR1) and C. glabrata (CDR1, CDR2). In C. tropicalis, presence of Y132F and S154F substitutions was unable to explain the mechanisms of majority of our isolates. Less than half of the azole-resistant C. tropicalis harboured these amino acid substitutions. This was in contrast to the high frequency identified in another local study where > 90% of the isolates had Y132F and S154F substitutions [15]. Likewise, in another study, these mutations accounted for 95% of the fluconazole-resistant C. tropicalis [16].

Our study was limited by the small sample size although we had included all azole-resistant bloodstream isolates between 2012 and 2015. In addition, we did not perform further functional verification of the ERG11 mutations and homology modelling experiments, therefore the clinical significance of I166S amino acid substitution in C. glabrata remains to be validated.

Conclusions

In conclusion, our results indicated that the mechanisms mediating azole resistance in our isolates are heterogeneous. There were isolates with unidentified resistance mechanisms warranting further exploration. Moving ahead, the use of more advanced molecular technologies such as next-generation sequencing might be considered for an in-depth molecular characterisation of azole-resistant Candida spp to aid the identification of potential resistance markers.