Insights into the Multi-Azole Resistance Profile in Candida haemulonii Species Complex

The Candida haemulonii complex (C. duobushaemulonii, C. haemulonii, and C. haemulonii var. vulnera) is composed of emerging, opportunistic human fungal pathogens able to cause invasive infections with high rates of clinical treatment failure. This fungal complex typically demonstrates resistance to first-line antifungals, including fluconazole. In the present work, we have investigated the azole resistance mechanisms expressed in Brazilian clinical isolates forming the C. haemulonii complex. Initially, 12 isolates were subjected to an antifungal susceptibility test, and azole cross-resistance was detected in almost all isolates (91.7%). In order to understand the azole resistance mechanistic basis, the efflux pump activity was assessed by rhodamine-6G. The C. haemulonii complex exhibited a significantly higher rhodamine-6G efflux than the other non-albicans Candida species tested (C. tropicalis, C. krusei, and C. lusitaneae). Notably, the efflux pump inhibitors (Phe-Arg and FK506) reversed the fluconazole and voricolazole resistance phenotypes in the C. haemulonii species complex. Expression analysis indicated that the efflux pump (ChCDR1, ChCDR2, and ChMDR1) and ERG11 genes were not modulated by either fluconazole or voriconazole treatments. Further, ERG11 gene sequencing revealed several mutations, some of which culminated in amino acid polymorphisms, as previously reported in azole-resistant Candida spp. Collectively, these data point out the relevance of drug efflux pumps in mediating azole resistance in the C. haemulonii complex, and mutations in ERG11p may contribute to this resistance profile.


Introduction
The recently described Candida haemulonii clade has caused deep concerns in hospital environments worldwide, since its members are able to cause life-threatening infections presenting high rates of clinical treatment failures [1]. The C. haemulonii clade is composed of emerging, opportunistic,

Antifungal Susceptibility Assay
Antifungal susceptibility testing was performed according to the standardized broth microdilution technique described by the Clinical and Laboratory Standards Institute (CLSI) in document M27-A3 and interpreted according to document M27-A3 [26]. The tested antifungal drugs include fluconazole (FLC), itraconazole (ITC), ketoconazole (KTC), posaconazole (PSC), and voriconazole (VRC) (Sigma-Aldrich, USA). The minimum inhibitory concentration (MIC), modal MIC, geometric mean (GM)-MIC, MIC 50 , MIC 90 , median, and antifungal concentration range were calculated using Prism version 8 (GraphPad Software, USA). Until now, no species-specific clinical breakpoints are established for rare species regarding azole susceptibility both in CLSI and EUCAST. In this sense, the manuscripts focused on this fungal complex usually base their results on the CLSI breakpoints established for the Candida genus (CLSI document M27-A3) in order to have a minimum (even not precisely) parameter to interpret these data. The clinical breakpoints defined for Candida spp. were used for the interpretation of the MIC data as follows: susceptible (S) ≤8 mg/L, susceptible-dose dependent (S-DD) 16-32 mg/L, resistant (R) ≥64 mg/L for FLC; S ≤0.125 mg/L, S-DD 0.25-0.5 mg/L, R ≥1 mg/L for ITC; S ≤1 mg/L, S-DD 2 mg/L, R ≥4 mg/L for VRC. For PSC, the susceptibility breakpoint was considered 0.06 mg/L, determined by EUCAST for Candida spp. No interpretive criteria for KTC are available in either the CLSI or EUCAST documents.

Efflux Pump Activity
To assess the ABC-type drug transporter activity of fungal cells, we evaluated the glucose-induced efflux of rhodamine 6G (R6G) (Sigma-Aldrich, USA) assay, as previously described [6,27]. Yeasts (10 7 cells/mL) were incubated with R6G (10 µM) to enable uptake under carbon source-depleted conditions for 60 min. Cells were washed in phosphate-buffered saline (PBS), and tubes with glucose-free PBS (control) and PBS supplemented with 100 mM of glucose were prepared to start R6G efflux. Finally, after 1 and 3 h of incubation with glucose at 37 • C, the cells were quantified in a flow cytometer (FACSCalibur, BD Bioscience, USA) and analyzed at 527 nm. Representative data from the 10,000 events analysis were shown and the results were expressed as the percentage of fluorescent cells (%FC). Fungal cells not stained with R6G were used as a negative control. In parallel, a different group was also pretreated for 1.5 h with FLC (64 mg/L) and VRC (16 mg/L) before the addition of R6G and glucose.

Chemosensitization Assay Combining Efflux Pump Inhibitors and Azoles
The efflux pump inhibitors (EPIs), Phe-L-Arg-β-naphthylamine dihydrochloride (Phe-Arg) and tacrolimus (FK506) (Sigma-Aldrich, USA), were used in combination with FLC or VRC in order to determine if the antifungal activity could be enhanced [28]. MICs were developed in the presence of azole (64 to 0.125 mg/L) with and without the presence of each EPI at a concentration of 64 mg/L. All the systems were incubated for 48 h at 37 • C and the results were expressed as the lowest azole concentration that resulted in the total inhibition of visible fungal growth.

RNA Extraction
After initial culture conditions (item 2.1), C. haemulonii isolate LIPCh4 (a pan-azole C. haemulonii sensu strictu isolate) was diluted to an initial inoculum of 10 7 cells/mL in fresh Sabouraud (20 mL) and then exposed to FLC at 64 mg/L or VRC at 16 mg/L for 1.5 h with shaking. Following drug exposure, the fungal cells were harvested for RNA isolation as previously described [29]. Complementary DNA (cDNA) was synthesized with the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems™, Thermo Scientific, USA).

Quantitative Reverse Transcription-Polymerase Chain Reaction (RT-qPCR)
RT-qPCR was undertaken to estimate the expression of the ChCDR1, ChCDR2, ChMDR1, and ChERG11 ortholog genes by C. haemulonii complex strains during FLC or VRC exposure (after 1.5 h of exposure). cDNA was analyzed by RT-qPCR with the StepOne™ Real-Time PCR System (Applied Biosystems™, Thermo Scientific, USA) and specific primers (Table 1). Universal SYBR green Supermix was used for PCRs according to the manufacturer's recommendations. The 2 −∆∆CT method was used for the relative quantification of gene expression [30], and the data were normalized to the ACT1 (actin) gene expression.
The letters F and R in the primer names describe the 5 -to-3 orientations of the primers as follows: F, forward (sense); R, reverse (antisense). b NCBI reference sequence.

Amplification and Sequencing of the ERG11 Gene
The ERG11 gene encoding lanosterol 14α-demethylase was amplified and sequenced with specific primers (Table S1). The BigDye ® Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, USA) was used for sequencing the reaction, precipitated with ethanol/EDTA/sodium acetate according to the protocol suggested by the manufacturer and sequenced on the ABI3730xl DNA Analyzer platform (Applied Biosystems, USA). DNA sequences and the corresponding amino acid sequences were analyzed with the SeqMan II and Bioedit software packages (Lasergene; DNAStar, USA). Consensus sequences were aligned with different reference ERG11 sequences from Candida spp. For the mutation analysis, the C. albicans strain SC5314 ERG11 gene assembly sequence from Candida Genome Database (http://www.candidagenome.org/) was used [31][32][33].

Sequences Accessions
The ERG11 gene sequences obtained from clinical isolates belonging to the C. haemulonii species complex studied herein were deposited in GenBank with the following accession numbers: MT860384 to MT860395.

Statistics
All the experiments were performed in triplicate, in three independent experimental sets. The results were analyzed statistically by the Analysis of Variance One-Way ANOVA (comparisons between three or more groups). All the analyses were performed using the program GraphPad Prism8. In all analyses, p values of 0.05 or less were considered statistically significant.

Azoles' Susceptibility Profiles
The MIC values determined for 12 clinical isolates forming the C. haemulonii species complex (C. haemulonii, n = 5 isolates; C. duobushaemulonii, n = 4; and C. haemulonii var. vulnera, n = 3) against five different azoles (FLC, ITC, KTC, PSC, and VRC) were summarized in Table 2 and Table S1. Among the tested azoles, PSC (GM-MIC = 0.62 mg/L) and KTC (GM-MIC = 1.49 mg/L) exhibited potent antifungal activity against all the isolates studied. Notably, all the fungal isolates had MICs of ≥64 mg/L for FLC. For ITC, a modal MIC of ≥16 mg/L was noted, suggesting that all the isolates were resistant to this antifungal. Despite some isolates remaining susceptible to VRC, the MIC distribution exhibited a peak at 16 mg/L, with 66% (n = 8) of isolates being resistant to this azole. Nine isolates (75%) were resistant to at least two different azoles. Six isolates (50%) were considered pan-azole resistant, of which four belong to the species C. haemulonii senso strictu, one belongs to C. duobushaemulonii, and one belongs to C. haemulonii var. vulnera.

Efflux Pumps Play a Primary Role in Azole Resistance in C. haemulonii Complex
In this set of experiments, we firstly assessed the efflux pump activity by the extrusion of RG6. When compared with other non-albicans Candida species (C. tropicalis, C. lusitaniae, and C. krusei), C. haemulonii exhibited a significantly higher R6G efflux after 1 h ( Figure 1) (Table S2). Analyzing the glucose-induced R6G efflux in the three different species forming the C. haemulonii complex, we could not observe statistical differences between them over 3 h ( Figure 2). Interestingly, the efflux pump activities have been shown to be constitutively expressed within fungal cells, since the treatment with either FLC or VRC did not modulate the glucose-induced efflux of RG6 ( Figure 2). J. Fungi 2020, 6, x FOR PEER REVIEW 7 of 16

Efflux Pumps Play a Primary Role in Azole Resistance in C. haemulonii Complex
In this set of experiments, we firstly assessed the efflux pump activity by the extrusion of RG6. When compared with other non-albicans Candida species (C. tropicalis, C. lusitaniae, and C. krusei), C. haemulonii exhibited a significantly higher R6G efflux after 1 h ( Figure 1) (Table S2). Analyzing the glucose-induced R6G efflux in the three different species forming the C. haemulonii complex, we could not observe statistical differences between them over 3 h ( Figure 2). Interestingly, the efflux pump activities have been shown to be constitutively expressed within fungal cells, since the treatment with either FLC or VRC did not modulate the glucose-induced efflux of RG6 ( Figure 2).   Since it seems that efflux pump activities are constitutively present in the C. haemulonii complex, as the FLC and VRC treatment did not modify the R6G extrusion process, we next examined the expression of the genes coding for known azole transporters. In an effort to identify the efflux pump-encoding genes which could participate in the clinical azole resistance in the C. haemulonii species complex, CDR1-like, CDR2-like, and MDR1-like genes were identified by the BLAST approach using the recently updated genome of C. haemulonii (strain B11899), which presented the highest degree of homology genes to C. albicans. The C. haemulonii genes with the highest degree of homology to the C. albicans CDR1 and CDR2 genes were XM_025483931.1 and XM_025488660.1, here referred to as ChCDR1 and ChCDR2 (Table S2), respectively; the C. haemulonii gene with a greater degree of homology with the C. albicans MDR1 gene was XM_025488295.1, here referred to as ChMDR1 (Table S3). Our data showed that the ChCDR1, ChCDR2, and ChMDR1 genes were not modulated by the azole treatment ( Figure 3). A similar result was observed considering the expression of the ERG11 gene ( Figure 3).  Since it seems that efflux pump activities are constitutively present in the C. haemulonii complex, as the FLC and VRC treatment did not modify the R6G extrusion process, we next examined the expression of the genes coding for known azole transporters. In an effort to identify the efflux pump-encoding genes which could participate in the clinical azole resistance in the C. haemulonii species complex, CDR1-like, CDR2-like, and MDR1-like genes were identified by the BLAST approach using the recently updated genome of C. haemulonii (strain B11899), which presented the highest degree of homology genes to C. albicans. The C. haemulonii genes with the highest degree of homology to the C. albicans CDR1 and CDR2 genes were XM_025483931.1 and XM_025488660.1, here referred to as ChCDR1 and ChCDR2 (Table S2), respectively; the C. haemulonii gene with a greater degree of homology with the C. albicans MDR1 gene was XM_025488295.1, here referred to as ChMDR1 (Table S3). Our data showed that the ChCDR1, ChCDR2, and ChMDR1 genes were not modulated by the azole treatment ( Figure 3). A similar result was observed considering the expression of the ERG11 gene ( Figure 3). Since it seems that efflux pump activities are constitutively present in the C. haemulonii complex, as the FLC and VRC treatment did not modify the R6G extrusion process, we next examined the expression of the genes coding for known azole transporters. In an effort to identify the efflux pump-encoding genes which could participate in the clinical azole resistance in the C. haemulonii species complex, CDR1-like, CDR2-like, and MDR1-like genes were identified by the BLAST approach using the recently updated genome of C. haemulonii (strain B11899), which presented the highest degree of homology genes to C. albicans. The C. haemulonii genes with the highest degree of homology to the C. albicans CDR1 and CDR2 genes were XM_025483931.1 and XM_025488660.1, here referred to as ChCDR1 and ChCDR2 (Table S2), respectively; the C. haemulonii gene with a greater degree of homology with the C. albicans MDR1 gene was XM_025488295.1, here referred to as ChMDR1 (Table S3). Our data showed that the ChCDR1, ChCDR2, and ChMDR1 genes were not modulated by the azole treatment ( Figure 3). A similar result was observed considering the expression of the ERG11 gene ( Figure 3).  We then assessed the contribution of these transporters to azole sensitivity by blocking them with classical EPIs (Table 3). When the isolates were grown in the presence of a combination of an azole (FLC or VRC) plus an EPI (Phe-Arg or FK506), we can observe that the MIC decreased in a 4-to more than 64-fold range. Remarkably, all the tested fungal isolates reversed their resistance phenotypes with the addition of EPIs.

ERG11 Mutations and Gene Expression Analysis
A phylogenetic tree of ERG11 genes and their putative orthologs in Candida spp. was built by the MEGAX program using the "Maximum likelihood" method, with the intent to identify their phylogenetic relationships (Figure 4). The tree based on a concatenated alignment of ERG11 genes places C. auris, C. haemulonii, C. duobushaemulonii, and C. pseudohaemulonii as a single clade, confirming the close relationship among these fungal species. Our phylogenetic analysis strongly supported that C. duobushaemulonii and C. pseudohaemulonii were more closely related and form a sister group of C. haemulonii, which appeared as the most branched species [23].  By comparing the amino acid sequences of ERG11p, 12 amino acid substitutions were identified in our isolates, all of them have been previously listed in the hot spot (HS) regions I, II, and III of ERG11p of C. albicans and C. auris ( Figure 5) [23,35,36]. Of the 12 found substitutions in ERG11p, modifications at positions K119S were only present in C. haemulonii sensu stricto and C. haemulonii var. vulnera, whereas changes at positions R267T and A432S were only detected in C. duobushaemulonii isolates.

Discussion
Drug resistance has become a major public health issue in terms of managing mainly invasive infections, particularly those resulting from pathogenic fungi that present scarce available effective drugs for treatment. Despite still being considered rare, infections by the C. haemulonii species complex have been highlighted by the high capability of developing multidrug-resistance against all clinically available antifungal classes, representing a challenge to the treatment of patients affected by these fungi. The majority of the studies has shown that isolates comprising the C. haemulonii complex demonstrated high MIC values for AMB and also frequently to azoles, all directly associated with clinical treatment failure [18,20]. In general, susceptibility to echinocandins, such as caspofungin and micafungin, in both in vitro and in vivo approaches have been observed [6,37,38]. However, there are already reports of resistance to this "novel" antifungal class in clinical isolates belonging to the C. haemulonii species complex [11,12]. Notably, we remain at the very beginning in understanding the biology of these "rare" fungal opportunistic human pathogens.
Among the diversity of molecular mechanisms related to the antifungal resistance machinery orchestrated by fungal cells, the upregulation of the efflux pumps MDR1 and/or CDR1/CDR2 contributes to azole resistance in most clinical isolates of Candida spp. [39]. Our results align with Zhang et al. [24], who observed that the presence or absence of FLC treatment repeatedly demonstrated the upregulation of the multidrug transporter gene CDR1. Additionally, in that study, treatment of the azole-resistant C. haemulonii strains with EPIs partially restored their susceptibility to FLC, confirming the significant contribution of Cdr1p to azole resistance [24]. It is well known that EPIs, such as Phe-Arg and FK506, play a vital role in azole resistance reversal against C. albicans [40], C. glabrata [41], C. tropicalis [42], and C. auris [28]. Consistent with these findings, in the present study, the combination of Phe-Arg or FK506 with FLC or VRC showed synergism against all the resistant isolates forming the C. haemulonii complex. Hence, here we have provided important evidence of efflux pumps ruling the azole resistance phenotype in the C. haemulonii species complex.
For many years, the imidazole KTC was the only available oral agent for the treatment of systemic fungal infections. Nevertheless, the side effects associated with KTC therapy as well as the  [13]. Colored letters are mutations encountered in the sequence alignment. Red boxes are mutation previously described in resistant strains of C. albicans.

Discussion
Drug resistance has become a major public health issue in terms of managing mainly invasive infections, particularly those resulting from pathogenic fungi that present scarce available effective drugs for treatment. Despite still being considered rare, infections by the C. haemulonii species complex have been highlighted by the high capability of developing multidrug-resistance against all clinically available antifungal classes, representing a challenge to the treatment of patients affected by these fungi. The majority of the studies has shown that isolates comprising the C. haemulonii complex demonstrated high MIC values for AMB and also frequently to azoles, all directly associated with clinical treatment failure [18,20]. In general, susceptibility to echinocandins, such as caspofungin and micafungin, in both in vitro and in vivo approaches have been observed [6,37,38]. However, there are already reports of resistance to this "novel" antifungal class in clinical isolates belonging to the C. haemulonii species complex [11,12]. Notably, we remain at the very beginning in understanding the biology of these "rare" fungal opportunistic human pathogens.
Among the diversity of molecular mechanisms related to the antifungal resistance machinery orchestrated by fungal cells, the upregulation of the efflux pumps MDR1 and/or CDR1/CDR2 contributes to azole resistance in most clinical isolates of Candida spp. [39]. Our results align with Zhang et al. [24], who observed that the presence or absence of FLC treatment repeatedly demonstrated the upregulation of the multidrug transporter gene CDR1. Additionally, in that study, treatment of the azole-resistant C. haemulonii strains with EPIs partially restored their susceptibility to FLC, confirming the significant contribution of Cdr1p to azole resistance [24]. It is well known that EPIs, such as Phe-Arg and FK506, play a vital role in azole resistance reversal against C. albicans [40], C. glabrata [41], C. tropicalis [42], and C. auris [28]. Consistent with these findings, in the present study, the combination of Phe-Arg or FK506 with FLC or VRC showed synergism against all the resistant isolates forming the C. haemulonii complex. Hence, here we have provided important evidence of efflux pumps ruling the azole resistance phenotype in the C. haemulonii species complex. For many years, the imidazole KTC was the only available oral agent for the treatment of systemic fungal infections. Nevertheless, the side effects associated with KTC therapy as well as the the relatively poor response rates, led to the search for a novel chemical group of azole derivatives, namely the triazoles [43]. Overall, the triazoles demonstrate a broader spectrum of antifungal activity and reduced toxicity when compared with the imidazole derivatives [44]. PSC is a second-generation triazole agent that has been developed mainly to combat the appearance of azoles resistance in yeasts, in particular to FLC. Several in vitro and in vivo studies confirmed that PSC has a broad spectrum of activity against the majority of yeasts, filamentous fungi and azole-resistant Candida. Thus, PSC has a superior activity profile when compared with FLC and ITC [44]. For instance, one way of developing azole resistance is by acquiring mutations on the ERG11 gene. Depending on the position and the nature of the alteration, the reduced susceptibility can cause cross-resistance or just resistance to a subset of derivatives [44].
Although the role of ERG11 gene point mutations in azole resistance is a well-recognized phenomenon in Candida spp. [38], only recently a few studies have characterized the polymorphisms of the ERG11 gene in azole-resistant isolates of the C. hameulonii complex [2,13]. In the study of Gade et al. [2], an increasing azole resistance in isolates of C. haemulonii and C. duobushaemulonii was evidenced in Colombia, Panama, and Venezuela, associated with substitutions in the ERG11 gene, suggesting the existence of an evolutionary pressure for the development of azoles resistance in Latin America. In the present study, eight of the 12 ERG11 gene mutations have been reported to be associated with reduced azole susceptibility in C. albicans-namely, F105L, S110A, D116A, K119S, D153E, R267T, A432S, and F487Y. However, Erg11p substitutions F126L, Y132F, and K143R, which were implicated in reduced susceptibilities to azoles upon heterologous expression in S. cerevisiae [45,46], were not identified in our isolates. Corroborating our data, a recent study by Rodrigues et al. [13] again found that Brazilian resistant isolates of C. haemulonii var. vulnera also lacked these common mutations. Thus, the ERG11 gene mutations by themselves cannot explain the differences between the distinct multi-azole profile in our isolates. ERG11, ERG3, and ERG5 combinatorial alterations in Candida have shown to circumvent this pathway and led to the development of cross-resistance to azoles and AMB [47][48][49][50]. Using multi-omics approaches, Zamith-Miranda et al. [51] showed that a resistant isolate of C. auris displays a significant higher abundance of Erg2p and a lower abundance of Erg3p. As a matter of fact, we have demonstrated by GC-MS that the C. haemulonii species sterol profiles were comparable to those of different Candida species with ERG2, ERG3, ERG6m and ERG11 gene mutations, presenting the same cross-resistance profile [21]. Along with our results, recent literature reports have suggested that azole resistance in C. haemulonii clade can be related to multiple-steps modifications in the ergosterol biosynthesis pathway [45,51].
Azole resistance has also been attributed to fungal cells that have lost mitochondrial function resulting in respiratory deficiency [52][53][54][55]. Precisely, in S. cerevisiae, C. glabrata, and A. fumigatus, mitochondrial dysfunction has displayed multiple effects in fungal cells, and this can enlighten us on the enhanced efflux pumps' activities [53,56,57]. These fungal cells usually exhibit an altered sterol profile with a unequal amount of ergosterol; still, no changes in the sequence of ERG11 gene or its expression have been detected [57,58]. Our recently published work demonstrated primary evidence that impariment of the mitochondrial function in C. haemulonii species complex intensifies the tolerance to stress and could be related with a rearrangement in the cellular ergosterol content [21]. In this context, growing evidence from recent elegant reports has revealed that there are more complicated and unknown drug resistance mechanisms apart from the mutational modification of the azole target ERG11p in fungal pathogens [59][60][61][62].

Conclusions
Indeed, the widespread use of azoles has been considered to be an important risk factor for FLC resistance development primarily in non-albicans Candida species. In this context, an assessment of all possible resistance mechanisms is important for a proper patient treatment outcome. Our data suggest an emergence of azole-resistant clinical isolates belonging to the C. haemulonii species complex recovered from Brazilian patients. This study provides significant evidence that efflux pump activity is implicated in high-level FLC resistance and pan-azole resistance in the C. haemulonii complex. Once there is an increasing azole resistance among C. haemulonii complex isolates, further surveillance is warranted.