Investigation of fluconazole susceptibility to Candida albicans by MALDI-TOF MS and real-time PCR for CDR1, CDR2, MDR1 and ERG11

Background C. albicans is a pathogenic yeast that is the most common cause of fungal infections in humans. Unfortunately, the yeast’s resistance to the antifungal medication fluconazole (FLC) is increasing; furthermore, testing its susceptibility to FLC by conventional methods takes time, resulting in treatment failure. The susceptibility of C. albicans to FLC was investigated using MALDI-TOF Mass Spectrometry and Real-time PCR tests for CDR1, CDR2, MDR1 and ERG11. Overall, 32 C. albicans strains made up of four reference strains (three FLC susceptible [S] and one FLC resistant [R], one spontaneous mutant strain [FLC susceptible-dose-dependent (SDD)] and 27 clinical strains obtained from two Thai University Hospitals) were tested for susceptibility to FLC. The following tests were performed: SensititreYeastOne and broth microdilution method, FLC resistant expression mechanism by Real-time PCR, and the major peak determination by MALDI-TOF MS. Results The change of CDR1 and CDR2 mRNA expression was only significantly observed in SDD and R strains. MALDI-TOF MS was performed after incubation for six hours; the change of mass spectral intensity at range 3376–3382 m/z (major peak) was significantly related to FLC susceptibility as SDD (decreased at 4 µg/mL and increased at 8 µg/mL), S (all increased), and R (all slightly decreased or no change). All 27 clinical strains showed FLC minimum inhibitory concentrations (MIC range 0.25-2 µg/mL), no change in CDR1 and CDR2 expression and S major peak type. The FLC resistant C. albicans with CDR1and CDR2 expression may possibly affect the change of mass spectral intensity at range 3376–3382 m/z. Conclusions The MALDI-TOF MS may be used to simultaneously classify and predict FLC resistant C. albicans strains associated with CDR1 and CDR2 expression. Further studies are essential to clarify the methodology and improve the reliability of this assay for routine diagnosis.

(FLC), a triazole antifungal medicine, is the most widely used treatment for this infection. Fluconazole has a fungistatic effect by inhibiting cytochrome P450 enzyme lanosterol demethylase (14α-demethylase), encoded by ERG11, in the ergosterol biosynthesis pathway. Many C. albicans clinical isolates overexpress ERG11, the gene encoding the target of the azoles. However, in many cases the level of overexpression is minimal or else observed in combination with other resistance mutations, making it difficult to assess the direct impact of such overexpression on the resistant phenotype. In addition to ERG11 overexpression, two main classes of efflux proteins --ATP-binding cassette (ABC) transporters encoded by CDR1 and CDR2, and Major facilitator superfamily (MFS) encoded by MDR1-can increase resistance to FLC by reducing the effective intracellular drug concentration [3,4]. The azole resistance associated with ERG11, MDR or CDR overexpression has been reported in many cases [5,6]. Currently, the broth microdilution (BMD) method is the gold standard of antifungal susceptibility testing (AFST) of Candida spp. This method of determining the MIC level of antifungal agents has been standardized by the Clinical and Laboratory Standards Institute (CLSI) and European Committee on Antibiotic Susceptibility Testing (EUCAST). However, this method is time-consuming and tedious, and an alternative method is desirable. The matrix assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) is generally used for microbial identification by referring to validated databases of reference spectra [1]. Several studies have adapted the MALDI-TOF MS for antifungal susceptibility testing of bacteria and fungi because of its simple, rapid, high-throughput, low waste, and low expense method [7][8][9]. The aim of this study is to investigate FLC susceptibility of C. albicans using MALDI-TOF MS and Real-time PCR for CDR1, CDR2, MDR1 and ERG11.

MIC level of control strains, clinical strains and spontaneous mutant strains
Of the FLC susceptibility testing with four reference strains, C. albicans MYA2876, C. albicans ATCC90028 and C. albicans MYA4440 were susceptible (S) to FLC at MIC level of 0.25, 0.25 and 0.5 µg/mL, respectively, whereas C. albicans ATCC96901 was resistant to FLC at MIC level of ≥ 64 µg/mL. With up to 10 consecutive serial passages of spontaneous mutations, ΔMYA4440 --the mutation strain of C. albicans MYA4440 --was susceptible-dose dependent to FLC (MIC = 4 µg/mL), while G950 strain did not change the MIC level. All 27 clinical strains of C. albicans were susceptible to FLC at MIC range 0.5-2.0 µg/mL (Table 1). Twelve strains collected from Thammasat University Hospital using SensititreYeastOne (SYO) were susceptible to fluconazole, voriconazole, anidulafungin, caspofungin and micafungin when applying CLSI CBPs. When applying the epidemiological cutoff values (ECVs), all the isolates had wild type phenotype drug susceptibility to amphotericin, flucytosine, posaconazole and itraconazole ( Table 2). Figure 1a and b indicate CDR efflux pump overexpression in C. albicans strains: the SDD (ΔMYA4440) and R   (ATCC96901) strains significantly overexpressed both CDR1 and CDR2 (p -value < 0.0001) compared to S strains. Compared to C. albicans MYA4440 (wild type) strain (MIC 0.5 µg/mL), its mutant strain ΔMYA4440 (MIC 4 µg/mL) expressed CDR1 and CDR2 at a higher level. After incubation with 8 µg/mL of FLC for 24 h, the SDD strain (ΔMYA4440) and R strain (ATCC96901) significantly overexpressed CDR1 (p -value < 0.0001) but significantly repressed CDR2 (p -value < 0.0001), while S strains (MYA2876, ATCC90028 and MYA4440) and all clinical S strains didn't significantly express CDR1 and CDR2 (p -value > 0.0001).

The protein spectrum related to the fluconazole resistance of C. albicans in control standard strains
This experiment was to observe FLC mass spectral change among S, SDD and R control strains after incubation with 4 and 8 µg/mL for 6 h. Before initiating the experiment, the quality control of reagent, microorganisms, drug, and incubation time period was investigated as shown in Figs. 2 and 3. Figure 2 shows the mass spectral range 0-600 m/z of 0.1%HCCA, FLC S C. albicans strain, FLC SDD C. albicans strain, FLC R C. albicans strain and FLC. The mass spectral intensity of FLC appeared at position 307 m/z in the spectrum, while those of 0.1%HCCA and FLC S, SDD and R strains did not show a corresponding peak at 307 m/z. Testing after FLC incubation for 2, 4 and 6 h, the mass spectral intensity of FLC was consistently present (Fig. 3). This indicates that the incubation period does not affect the FLC mass spectrum. Next, S and R C. albicans strains were treated with 4 and 8 µg/mL of FLC and incubated for 6 h. Testing then showed the mass spectral intensity of 8 µg/mL FLC addition was higher than that of 4 µg/ mL FLC addition. No FLC hydrolysis was detected in this experiment (Fig. 4). A peak at spectrum position 307 m/z can found in every condition of fluconazole addition, although the intensity at 8 µg/mL of fluconazole concentration will be higher than at 4 µg/mL of fluconazole concentration (Fig. 4). This suggests that FLC hydrolysis is not the mechanism of the action of FLC in C. albicans. Therefore, we cannot use FLC mass spectrum position to determine among FLC S, SDD and R strains.
Then, the spectra of C. albicans at mass spectral range 2,000-20,000 m/z were investigated after incubation with 0 (control), 4 and 8 µg/mL of FLC addition for 6 h (Fig. 5). Interestingly, the change of mass spectral intensity at range 3376-3382 m/z (major peak) is significantly related to the FLC susceptibility of S, SDD and R reference strains. Three patterns were observed: compared to 0 µg/mL of FLC, 1)the S strain showed a higher peak with 4 and 8 µg/mL; 2) the SDD strain showed a lower peak at 4 µg/mL and a higher peak at 8 µg/mL; and 3) the R strain showed a slight decrease or no change at 4 and 8 µg/mL. After observing the major peak of all 27 clinical strains, their results showed a similar pattern to the S reference strain (data not shown), that were in line with FLC MIC recognized as susceptible (MIC range 0.25-2 µg/mL) and no change in CDR1 and CDR2 overexpression.

Discussion
Of all C. albicans strains without FLC addition in this study, only CDR1 and CDR2 genes expression are related to FLC SDD (ΔMYA4440) and R strains (ATCC96901). These results confirm the resistance mechanism via increased efflux of the drug from cells by ABC transporters. The R strain has a higher expression of CDR1 and CDR2 than the SDD strain. Although only two samples --consisting of one spontaneous mutant SDD strain and one R reference strain --are available in this study, the results are in line with previous reports that high expression levels of CDR efflux genes is a major mechanism for fluconazole resistance in C. albicans [6]. MDR1 expression in our study was not significantly different among S, SDD and R strains (p -value > 0.0001). This may support a previous study which revealed that in planktonic cells CDR displays higher expression level after 24 h incubation, while MDR1 expression is even higher after 48 h [10]. Contrary to our study which revealed that only CDR1 overexpressed in SDD and R strains after 24 h incubation   with FLC, a previous study observed that FLC could induce the expression of CDR1, CDR2 and MDR1 [10]. Expression of CDR1 was controlled five cis-acting regulatory elements including one BEE (basal expression element), one DRE (drug-sensitive element), two SRE (sterol regulatory element) and one NRE (negative regulatory element) while CDR2 expression was controlled by only DRE. Among these different elements, only DRE is engaged in essential high expression and up-regulation of CDR1 and CDR2 to fluconazole efflux [11,12]. Contrarily, MDR1 gene does not possess the DRE element, but also MDR1 promoter does not directly react towards fluconazole [13]. For ERG11 expression, all strains showed an increase after incubation with 8 µg/mL FLC for 24 h. This supported the previous report that the expression of ERG11 increased after antifungal treatment in susceptible and nonsusceptible to azole isolates [14]. The most interesting comes to ERG11 gene, the overexpression of which is only partially responsible for the R status. In most cases SNP mutations, which lead to amino acid substitutions that makes the whole protein unable to bind the drug [15]. Additionally, the potential usage of the expression of upregulation transcription factors including TAC1, MRR1, and UPC2 in clinical fluconazole-resistant C. albicans isolates, which is directly responsible for binding with CDR, MDR1 and ERG11 promoters, respectively [16]. Notwithstanding, the susceptibility or resistance of C. albicans towards fluconazole dose not only originate in down-or over-expression of CDR1, CDR2, MDR1 or ERG11 genes. In fact, this might be true to some extent, however, in C. albicans cells, it is not a simple dependence. In case of CDR genes, the higher amount of mRNA potentially leads to higher protein levels and higher azole efflux. However, CDR genes might harbour SNP (single nucleotide polymorphism) mutations, which cannot be sensu stricto interpreted as an "overexpression" [5,15]. On the other hand, the lack of ergosterol or abnormal sterol deposition in plasma membrane (as an effect of azole treatment) might lead to dysfunction of membrane proteins, including ABC transporters, and affected the mislocalization of the Cdr proteins from plasma membrane to the inside of the cell, which in turn results in the protein inactivity (despite their higher levels) [17]. Concerning our attempt to determine FLC susceptibility to C. albicans with MALDI-TOF MS, we found only one spectrum at position 3376-3382 m/z that was linked with FLC resistance of C. albicans after incubation with 4 and 8 µg/mL FLC for 6 h. Indeed, by allowing a peak position's tolerance of ± 3 m/z we were able to overcome the small spectral variations and so could assess as C. albicans housekeeping peaks [18]. The drug concentration and incubation time period seem to be major factors for each Candida species to observe a major peak for differentiating S strain from R strain. In view of the study by Paul, both the resistant and susceptible C. tropicalis isolates showed spectral changes after 4 h when challenged with 128 µg/mL and 1 µg/ mL FLC, respectively [19]. In our study, the major mass spectral intensity change was clearly seen after 6 h incubation with FLC. As reported by Vatanshenassan using MBT ASTRA prototype software, the discrimination between resistant and susceptible strains was accurately detected in 6 h [20]. Furthermore, a previous report explained that prolonged exposure to antifungal drug in some isolates allowed an increasing probability of correctly differentiating between S and R strains [8]. Together with the condition of drug concentration and incubation time period, the method processed in each study could support the identification of FLC resistant C. albicans strains as well as finding different major mass spectra. For example, in the study of Marinach [21], MPCC determination was used to identify FLC resistant C. albicans strains with MALDI-TOF MS. Their study concluded that the most suitable FLC concentrations were 2 and 4 µg/mL and the significant spectral intensity range was 5,800-7,600 m/z to detect FLC resistance. Six spectra in this range could detect FLC resistance of different strains. Many studies have used different procedures including MPCC method [19,21], composite correlation index (CCI) values approach [8,22,23], matrix-assisted laser desorption ionization Biotyper antibiotic susceptibility test rapid assay (MBT-ASTRA) [1,9] and machine-learning algorithms with the most robust pipeline of analysis [18]. Now, we disclose a specific mass spectrum for analyzing FLC S, SDD and R strain for C. albicans with their resistance related to CDR expression efflux pump. Our results, which followed the same protocol, come from more than 27 C. albicans strains from clinical samples at various sites of 27 patients which were investigated to find the major peak type by MALDI-TOF MS and FLC MIC by BMD. All strains presented a novel mass spectral intensity at range 3376-3382 m/z (major peak) and the major peak type was observed as S type and interpreted MIC level as susceptible (data not shown). Unfortunately, we could not find FLC SDD or R C. albicans strain from blood cultures of candidemic patients in the University Hospitals during our study. Moreover, we followed the protocol for determining in 48 C. tropicalis strains including FLC S, SDD and R strains from blood cultures of 48 patients. Unsuccessfully, the major peak type could not be used for differentiating C. tropicalis (data not shown). For this reason, this major peak type is specific for determining FLC susceptibility testing of C. albicans.

Conclusions
Overall, we discovered a novel mass spectral intensity at range 3376-3382 m/z (major peak) related to the FLC susceptibility of S, SDD and R strains specific for C. albicans. The FLC resistance mechanisms of C. albicans associated with CDR1 and CDR2 expression (via increased efflux of the drug from cells by ABC transporters) may possibly effect the change of mass spectral intensity at range 3376-3382 m/z. The results demonstrate that the MALDI-TOF MS may be used to simultaneously classify Candida species and predict FLC resistant C. albicans strains associated with CDR1 and CDR2 overexpression. However, the limitations of our study include the small number of clinical C. albicans strains, the shortage of SDD and R clinical C. albicans strains, and the unproven protein spectrum at position 3376-3382 m/z. In addition, several molecular mechanisms of fluconazole resistance except the overexpression have not been considered such as insertions, deletions and point mutations. Therefore, further studies are essential to clarify the methodology and improve the reliability of this assay for routine diagnosis.

Yeast samples used in this study
The C. albicans samples in this study were 15 strains isolated from abdomen (1 strains), blood (2 strains), wound/ pus (4 strains), sputum (2 strains), stool (1 strain), and urine (5 strains) in the Microbiology Laboratory Unit, Department of Central Laboratory and Blood Bank, Faculty of Medicine, Vajira Hospital, Navamindradhiraj University, Bangkok, Thailand between April 2015 and December 2016, and 12 strains obtained from positive blood culture in the Microbiology Laboratory, Thammasat University Hospital from January 2012 to April 2015. Four ATCC standard strains were used as the control strains including two susceptible strains (C. albicans MYA2876 and C. albicans ATCC90028), one strain for spontaneous mutation in SDD MIC level (C. albicans MYA4440), and one resistant strain (C. albicans ATCC96901). C. parapsilosisATCC22019 and C. krusei-ATCC6258 were used as quality control strains for susceptibility testing.
For this study, we obtained all yeast strains from routine laboratory collection where their data were done namelessly. Therefore, we did not use any additional data other than those gained from laboratory. Human ethical committees decided to exempt, constituted in accordance with the International Conference on Harmonization-Good Clinical Practice (ICH-GCP).

SensititreYeastOne antifungal susceptibility testing
Twelve strains collected from Thammasat Hospital were subculturedon Sabouraud Dextrose Agar (SDA) and incubated at 35 °C overnight. The procedure used Sen-sititreYeastOne (SYO) according to the manufacturer's instructions (Thermo Fisher Scientific, Cleveland, OH, USA)for measuring MIC of nine drugs -namely, fluconazole, anidulafungin, amphotericin B, micafungin, caspofungin, 5-Flucytosine, posaconazole, voriconazole and itraconazole. The SYO plate was incubated at 35 °C overnight and MIC endpoints were read after 24 h of incubation by the color change from blue (negative, indicating no growth) to magenta (positive, indicating growth) [24].

CLSI broth microdilution method (CLSI BMD)
Reference BMD testing was performed exactly as outlined in CLSI document M27-4th ed [25], with a final inoculum concentration of 0.5 × 10 3 to 2.5 × 10 3 cells/mL and RPMI 1640 medium with 0.2% glucose, and incubation at 35 °C. MIC values were determined by visually observing the presence or absence of growth after 24 h (Candida and other yeast) and 72 h (C. neoformans) as the lowest concentration of drug that produced a prominent decrease in turbidity (ca. 50% growth reduction) relative to that of the drug-free control.

Selection of spontaneous mutants
C. albicans MYA4440 isolates (MIC = 0.5 µg/mL) and G950 isolates (MIC = 2 µg/mL) were subcultured and incubated at 35 °C overnight. The individual colony from each strain was used to start cultures in RPMI and grown overnight. One hundred microliters of overnight cultures (5 × 10 8 to 1 × 10 9 CFU/mL) were streaked onto SDA plates with (1, 1.5, 2, and 4 µg/mL) or without fluconazole and incubated at 35 °C for 2 days. MIC was checked and resubcultured onto a fresh-passage until the MIC was 4 µg/mL. The spontaneous mutation frequency rate was calculated as the ratio of viable colonies growing on drugcontaining plates over the starting inoculum. Mutant resistance phenotypes were confirmed by subculturing on SDA plates containing an amount of drug equivalent to that used for initial selection. Mutant strains were selected and evaluated by MIC and qRT-PCR.

RNA extraction, cDNA synthesis and quantitative real-time PCR (qRT-PCR)
Each C. albicans isolate was suspended in 5 mL of yeast nitrogen base (YNB) medium supplemented with 50 mM glucose and incubated at 37 °C in a shaker at 75 rpm overnight to prepare each of the starters. After that, 500 µL of the starter were transferred into a flask of 50 mL containing fresh medium and incubated at 37 °C in a shaker at 75 rpm until OD 600 = 1.0. Total RNA was extracted from the growth medium (at 0 h without fluconazole and after 24 h with 8 µg/mL of fluconazole addition) at midexponential (log) phase using the preparation of yeast RNA by extraction with hot acidic phenol. The concentration of RNA was measured using a NanoDrop2000C spectrophotometer (Thermo Scientific). The RNA purity and integrity was evaluated by the ratio of absorbance 260 and 280 nm. Run gel electrophoresis was performed to verify that the RNA was intact. Total RNA was treated with TURBO DNA-free ™ Kit (Invitrogen) according to the manufacturer's instructions (RNA sample > 200 µg nucleic acid per mL). cDNA was synthesized by reverse transcription from 1 µg of total RNA using the iScript ™ Reverse Transcription Supermix for RT-qPCR (Biorad) according to the manufacturer's instructions. The reaction protocols for cDNA synthesis were composed of a priming step at 25 °C for 5 min, a reverse transcription step at 46 °C for 20 min, and an RT inactivation step at 95 °C for 1 min. The mRNA expression level was measured using quantitative real-time RT-PCR (qRT-PCR) following Watamoto's protocol [28]. The sequence primers (macrogen) of the genes CDR1, CDR2, MDR1, ERG11 [28], and PMA1 [29] in this study are listed in Table 3. qRT-PCR was performed in duplicate using iTaq ™ Universal SYBR ® Green Supermix (Biorad). Ten microlitres of PCR mix (5 µLiTaq ™ Universal SYBR ® Green Supermix, 1 µL primer mix, 0.5 µL cDNA and 3.5 µL DEPC water) was used for each gene and qRT-PCR was performed using the following cycling conditions: 95 °C for 5 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 30 s. Fluorescence intensities were quantified using Bio-Rad ® CFX96 ™ (Biorad). The relative quantities of the target genes (CDR1, CDR2, MDR1 and ERG11) were normalized against PMA1 housekeeping gene expression (plasma membrane ATPase pump). The analyses of genes expressions were performed using the comparative 2 (−ΔΔCT) method of relative quantification. The mRNA expression level of each target gene was analyzed. All results were presented by mean ± SD. Statistical analyses were performed using GraphPad Prism 5.0 software. Results were compared using one way ANOVA followed by Bonferroni's post test; the results were considered statistically significant when the p -value was < 0.0001.

Detection of mass spectral analysis using MALDI-TOF MS
Fresh-isolated colonies on SDA were cultured in RPMI 1640 medium, adjusted turbidity with 0.5 McFarland Standard, and incubated at 37 °C overnight. Two milliliters (mL) of suspension were added into RPMI 1640 medium with different fluconazole concentrations: 4 and8 µg/mL, and without fluconazole as a negative control. The suspension was incubated at 37 °C and harvested at 6 h intervals, centrifuged and the pellets and the supernatant separated. The pellets were extracted with formic acid following Bruker Daltonics' recommended protocol for analysis of C. albicans protein spectrum. Briefly, 300 µL of HPLC grade water and 700 µL of ethanol were added, mixed with a vortex for 1 min, and centrifuged at 13,000 rpm for3 min. The solution was discarded after centrifugation. After drying the pellets, 70% formic acid was added to cover the pellets, then silica beads were added into the solution before mixing with vortex for 2 min. Acetonitrile was added in equal volume of 70% formic acid (ratio 1:1). The mixer was centrifuged at 13,000 rpm for 3 min. One microliter of supernatant was spotted on the MALDI target plate and dried at room temperature (RT). One microliter of 2.5% TFA of α-Cyano-4-hydroxycinnamic acid (HCCA) matrix was covered and dried at RT again. The sample was analyzed by MALDI-TOF MS (Bruker Daltonics, Bremen, Germany). A drop of the supernatant used for fluconazole spectrum analysis was placed onto a plate and dried, covered with0.1% TFA of HCCA matrix, dried at RT, and analyzed by Bruker Autoflex-Speed MALDI-TOF MS. The bacterial test standard (BTS), the ribosomal proteins of Escherichia coli strain DH5alpha, was used as a calibrator. The spectrum was measured by flexControl 3.3 software [19,30] using linear mode (2,000-20,000 m/z) for C. albicans spectrum and using reflector mode (0-3,500 m/z) for fluconazole spectrum. Finally, the data was processed and compared using flexAnalysis. BTS: Bacterial test standard. Chanika Maenchantrarath carried out the study design, sample preparation, experiments, and data analysis and wrote the initial draft of the manuscript. Pradchama Khumdee helped with some parts of MALDI-TOF MS experiment. Seksun Samosornsuk gave guidance and support. Narissara Mungkornkaew identified and collected some yeast samples. Worada Samosornsuk gave advice, contributed to the analysis of the evaluation of the study, and was a major contributor in writing the manuscript. All authors read and approved the final manuscript.

Funding
This work was supported by grants from the National Research Council of Thailand, 2019. The authors gratefully acknowledge the financial support Table 3 Gene-specific primers for qRT-PCR [22] F Forward, R Reverse