Introduction

Candidiasis is a frequent fungal infectious disease, occurring in immunocompromised patients, especially those receiving chemoradiation, bone marrow transplantation or cannula, and AIDS patients as well. It may manifest common ailments, such as thrush, vaginitis, even or fatal invasive infection. Candida albicans is the most important pathogen of candidiasis. Thirty-five to sixty-five percent of candidemia were caused by C. albicans (Krcmery and Barnes 2002). A study showed that C. albicans (44.7 %) were the most frequent species in a total of 1377 isolates causing fungemia in Spain (Pemán et al. 2011). C. albicans constitutes 64 % (163/255) of causative species of candidemia, and the highest degree of resistance to azoles was observed in a university hospital in southwestern Greece (Spiliopoulou et al. 2010). Azoles are the first-line drugs to prevent and treat candidiasis. However, repeated azole therapy for chronic infections has been associated with an increase in azole resistance (Cowen et al. 2000). Our previous research found that 68.6 % of 426 clinical isolates were C. albicans, and 5.1 % of the C. albicans was resistant to fluconazole (Xu et al. 2008).

Azole resistance of C. albicans is a multi-level complex process involving many factors. Several hypotheses have been proposed on the resistance mechanisms: (1) decreased permeability of cell membrane, which may decrease the amount of drugs entering the cell; (2) enhanced ability to degrade the drug; (3) missense mutation in target enzyme coding gene ERG11, leading to target conformational change and reduced affinity to bind azole molecules; (4) overexpression of efflux pump counteracting drug influx; (5) biofilm formation providing a protecting econiche; and (6) vesicular vacuoles in some resistant isolate cells restraining fluconazole molecules (Maebashi et al. 2002). Target conformational change and overexpression of efflux pump were regarded as two major aspects in these resistance mechanisms. Overexpression of ERG11 has been hypothesized to mediate resistance. However, studies have not reached a definite conclusion that this kind of overexpression was related to azole resistance in C. albicans (White 1997; White et al. 2002).

The target enzyme of fluconazole is 14 alpha-demethylase, i.e., Erg11p, encoded by ERG11. Erg11p is one of the key enzymes in the ergosterol synthesis pathway of C. albicans and composed of 528 amino acids that form 12 α-helices of A–L, several β-pleated sheets, and some other helix configurations such as A′, J′, K′, and K″. The active center is located in the deep inside of the protein, near the haemachrome between helix I and L. The substrate interacts with a long access channel and then is demethylated (Ji et al. 2000; Monk et al. 2014; Marichal et al. 1999; Podust et al. 2001; Sanglard et al. 1998). Azole molecules block this process and inhibit cell’s ergosterol synthesis. Mutations in ERG11 change the Erg11p amino acid sequence and possibly the protein conformation. As a consequence, the affinity between azoles and the target enzyme may also be decreased, leading to the resistance to azole.

Overexpression of efflux pumps plays a role in drug resistance of Candida species by decreasing the intracellular drug concentration. Of all the transmembrane transporter coding genes, candida drug resistance genes CDR1, CDR2 and multiple drug resistance gene MDR1 are frequently reported. White (White et al. 2002) and Park (Park and Perlin 2005) reported that overexpression of CDR1, CDR2, and MDR1 along with some certain mutations in ERG11 were related to resistance. CDR1 and CDR2 encode plasma membrane ABC-type transporters Cdr1p and Cdr2p, respectively. Although the two genes share high sequence similarity, Cdr1p makes a much greater functional contribution to fluconazole resistance in C. albicans than Cdr2p (Holmes et al. 2008).

Multiple mechanisms might be involved in a resistant isolate simultaneously. Alternatively, single mechanism may be responsible for the resistance (Goldman et al. 2004; Perea et al. 2001). Studies have shown that even a single base change in ERG11 gene can increase the resistance to azoles (Lamb et al. 1997, 2000).

In our collection of C. albicans isolates, we found that 15 isolates were resistant to fluconazole. Sequence analysis detected some interesting mutations, especially G487T (A114S) and T916C (Y257H) in ERG11 gene, which might be related to fluconazole resistance (Xu et al. 2008). In order to analyze the potential mechanisms in these 14 resistant isolates with mutations G487T and T916C, different primers were designed to amplify the ERG11 gene of the related isolates and transcription of ERG11, CDR1, CDR2, MDR1, and FLU1, and expression of CDR1 and CDR2 was determined by real-time PCR and Western blot, respectively. Our results indicated that Erg11p conformational alteration caused by these mutations might be responsible for azole resistance in these isolates.

Materials and methods

Isolates and type strains

Fifteen fluconazole-resistant isolates of Candida albicans with a fluconazole MIC >64 μg ml−1 were obtained from vaginal secreta, or urine, or secreta of raw surface of non-HIV patients in Ji′nan (Nos. 592, 4263, and 4266) or in Guangzhou (No. GZ03, GZ04, GZ09, GZ15, GZ16, GZ17, GZ18, GZ23, GZ29, GZ34, GZ51, and GZ58) of China. The fluconazole-susceptible isolates were from patients in Ji′nan (Nos. C261, C307, and C522) or in Beijing (Nos. 0461, 0920, 2827, 2855, and 2928). Candida krusei ATCC 6258, Candida parapsilosis ATCC 22019, and the following type strains of C. albicans, including fluconazole-susceptible C1b, C1c, C1d, C1e, and fluconazole-resistant ATCC 76615-19, were obtained from Chinese Cultural Collection Commission for Microbiology (CCCCM, Nanjing, China). C. albicans ATCC 10231 was purchased from Anhui Provincial Center for Disease Prevention and Control (Hefei, China). C. albicans Darlington strain was provided by Professor John E. Bennett of National Institutes of Health, Bethesda, Maryland. Saccharomyces cerevisiae AD/CDR1 (MATα PDR1-3 ura3 his1Δyor1::hisGΔsnq2::hisGΔpdr10::hisGΔpdr11::hisGΔycf1::hisGΔpdr3::hisGΔpdr15::hisGΔpdr5::CaCDR1A-URA3) and AD/CDR2 (MATα PDR1-3 ura3 his1Δyor1::hisGΔsnq2::hisGΔpdr10::hisGΔpdr11::hisGΔycf1::hisGΔpdr3::hisGΔpdr15::hisGΔpdr5::CaCDR2A-URA3) were gifts from Professor Ann R. Holmes of Otago University, New Zealand. All isolates and standard strains were reserved on Sabouraud dextrose agar (SDA) at 4 °C.

Agents and media

SDA and yeast extract peptone dextrose (YEPD) broth were prepared according to conventional laboratory methods. The ingredients of SDA include dextrose 20 g, peptone 10 g, agar 20 g, chloromycin 50 mg, and distilled water to 1000 ml. YEPD broth is composed of 10 g yeast extract, 20 g peptone, and 20 g dextrose per liter. The E.Z.N.A.™ Yeast DNA kit and RNA kit were purchased from OMEGA Corp. (USA). The Hi-Fi Primer STAR HS DNA polymerase and the SYBR RT-PCR Kit (Perfect Real Time) were purchased from TaKaRa Corp. (Japan). Bicinchoninic acid (BCA) protein assay kit was purchased from Beyotime Corp. (China). Nitrocellulose membrane was purchased from Amersham Phamacia Biotech Corp. (USA). Proteinase inhibitors were purchased from Sigma-Aldrich Co. LLC. (USA). Anti-Cdr1p antibody was gift kindly provided by Professor Dominique Sanglard of University of Lausanne, Switzerland, and anti-Cdr2p was produced by CWBIO Corp. (China). Enhanced chemiluminescence (ECL) kit was purchased from Santa Cruz Corp. (USA).

Fluconazole susceptibility tests

Sensitivity and resistance of strains were measured using the M27-A3 broth dilution method as recommended by the Clinical and Laboratory Standards Institute (CLSI 2008). Strains ATCC 6258 and ATCC 22019 were used as controls.

Amplification of ERG11 gene and sequencing

Genomic DNA of C. albicans isolates was obtained following the manufacturer’s instructions of the E.Z.N.A.™ yeast DNA kit. Primer pair 5′-TGTAAGCTTGGAATTCAATCGTTATTC-3′ and 5′-AATGGATCCCTGATTGAGTCATCCTAAC-3′ was designed to amplify an expected PCR product of 1852 bp, which includes the complete ERG11 gene sequence. The products were purified and sequenced with an ABI PRISM 3700 DNA analyzer (Applied Biosystems). The ERG11 of all the fifteen resistant isolates, eight susceptible isolates, and the six standard strains were amplified and sequenced. Two strains (ATCC 76615-19 and the Darlington strain) whose ERG11 sequence was known were used as controls (Kakeya et al. 2000; Long et al. 2002).

Real-time PCR: mRNA semi-quantification of ERG11, CDR1, CDR2, MDR1, and FLU1

Fourteen out of fifteen fluconazole-resistant isolates (except for isolate GZ04) had both ERG11 mutations (G487T and T916C). The 14 resistant isolates were cultured overnight in fresh YEPD broth with constant shaking at 30 °C until mid-logarithmic phase was reached. ATCC 10231 was used as control. The freshly harvested cells were washed thrice thoroughly with sterile distilled water. Total cellular RNA was extracted following the manufacturer’s instructions of the E.Z.N.A.™ Yeast RNA kit. All the RNA extracts were treated with RNase-free DNase I. The quality of RNA was examined using a spectrophotometer (DU® 800, Beckman, Palo Alto, CA, USA). A ratio of A260 to A280 of 1.8–2.0 corresponded to 90–100 % pure nucleic acid.

The cDNA was synthesized following the manufacturer’s instructions of SYBR RT-PCR Kit (Perfect Real Time). The volume of reverse transcription system was 20 μl containing 4 μl 5 × PrimerScript buffer, 1 μl Oligo dT primer, 1 μl PrimerScript RT enzyme, 1 μl random mers, 13 μl RNA, and RNase-free distilled water. The system was incubated at 37 °C for 60 min, 95 °C for 5 min and held at 4 °C in a T-gradient 96 thermal cycler (Whatman Biometra, Germany). The cDNA samples were stored at −20 °C.

The real-time PCR primers listed in Table 1 were designed with Primer Premier 5.0 to amplify C. albicans ERG11, CDR1, CDR2, MDR1, FLU1, and housekeeping gene 18SrRNA which is used as a control. The volume of PCR was 20 μl containing 1 μl forward primer, 1 μl reverse primer, 1 μl cDNA, 7 μl distilled water, and last 10 μl SYBR Green I, a specific DNA-binding dye used to detect amplifications. The cycling conditions on LightCycler 2.0 apparatus (Roche Diagnostics, Germany) included 40 cycles of denaturation for 10 s at 95 °C, annealing for 5 s at 57 °C for 18SrRNA, MDR1, CDR2, and ERG11, at 62 °C for FLU1, and at 67 °C for CDR1, and elongation for 10 s at 72 °C, followed by termination in a final cooling step for 30 s at 40 °C. When the amplification was completed, 8-µl product was separated by 2 % agarose gel electrophoresis, visualized, and photographed with Image Master Video Documentation System (Pharmacia Biotech) to verify the specificity of products. The mRNA levels of ERG11, CDR1, CDR2, MDR1, and FLU1 were evaluated using the 2−ΔΔCt method (Livak and Schmittgen 2001). The Ct value was the average of the threshold cycle numbers from three independent experiments. Data were presented as the fold change in gene expression normalized to 18SrRNA as control.

Table 1 Primers used to amplify C. albicans ERG11, CDR1, CDR2, MDR1, and FLU1 using real-time PCR
Full size table

Western blot: semi-quantification of Cdr1 and Cdr2

Cdr1 and Cdr2 protein of the 14 fluconazole-resistant C. albicans isolates with ERG11 mutations G487T and T916C were semi-quantified using Western blot. S. cerevisiae AD/CDR1 and AD/CDR2 were used as controls. After growing overnight at 30 °C in YEPD broth to the mid-exponential phase, the cell density would be 7.5 × 107 /ml or was adjusted to this density if it was tested higher; yeast cells were centrifuged, washed thrice with cold distilled water, and suspended in lysis buffer (10 % proteinase inhibitors, 50 mM Tris (pH 7.5), and 2.5 mM EDTA). Glass beads (spherical, Φ0.5 mm) were added into the suspension and vortexed for 6 times (each time contained 30 s of vortex and 30 s of interval). Cell debris was removed after a 15-min centrifuge at 3000×g. The supernatant was centrifuged at 15,000×g for 1 h. Subsequently, the pellet was resuspended in buffer containing 10 mM Tris (pH 7.5), 0.5 mM EDTA, and 10 % glycerol. Plasma membrane (PM) was obtained by a discontinuous gradient containing an equal volume of 53.5 and 43.5 % sucrose. All the steps above were operated at 4 °C. The concentration of protein was quantified using the BCA protein assay kit.

The extracted proteins (40 µg) were electrophoresed on a 7.5 % sodium dodecyl sulfate polyacrylamide gel (SDS-PAGE). Proteins were transferred to a 0.22-µm nitrocellulose membrane at 200 mA for 3 h. The blots were incubated at 4 °C overnight with a 1:500 dilution of anti-Cdr1p antibody (Micheli et al. 2002) and a 1:1000 dilution of anti-Cdr2p antibody (a polyclonal antibody first raised in rabbits, by Beijing ConWin Biotech Co., Ltd). For anti-Cdr1p antibody, immunoreactivity was detected with horseradish peroxidase-labeled goat anti-rabbit IgG antibodies at a 1:5000 dilution, and for anti-Cdr2p antibody, 1:10,000. The blots were visualized with ECL kit. Densitometric analysis of the band intensity was carried out using Quantity One software (Bio-Rad Laboratories, Hercules, CA).

Results

Mutations in ERG11 gene

PCR products of ERG11 gene from 29 C. albicans isolates or standard strains were obtained as a single band as revealed by ethidium bromide staining. The sequenced ERG11 gene was 1852 bp in length. Nineteen missense mutations were found in 15 resistant and 8 susceptible isolates.

Homozygous T495A, A530C, and C1567A were detected in the control strain ATCC 76615-19, and mutations of T541C and T1559C were observed in Darlington strain as reported previously (Kakeya et al. 2000; Long et al. 2002). Homozygous mutations G487T and T916C were present simultaneously in 14 resistant isolates: 592, 4263, 4266, GZ03, GZ09, GZ15, GZ16, GZ17, GZ18, GZ23, GZ29, GZ34, GZ51, and GZ58. Other mutations of ERG11 were not found in the 14 resistant isolates. In another resistant isolate GZ04, four homozygous: T495A, A530C, T541C, and T1493A were detected. Homozygous T495A, A504C, A530C, G640A, G820C, G820T, and A945C and heterozygous T495A/T, T566G/T, G630T/G, G635C/G, C1271A/C, G1289T/G, and G1609A/G were detected in the susceptible standard strains C1b, C1c, C1d, and C1e.

Semi-quantification of gene transcription in the resistant isolates

The LightCycler relative quantification method was used to measure the mRNA level of ERG11, CDR1, CDR2, MDR1, and FLU1 genes in 14 fluconazole-resistant isolates with G487T and T916C mutations in ERG11 gene. Melting curves and melting peaks of the tested genes gained from systems software of LightCycler apparatus showed no primer dimers and nonspecific amplification. Single band of products of expected length in agarose gel electrophoresis confirmed the specificity of the LightCycler PCR.

Compared with susceptible strain of C. albicans ATCC10231, all of the 14 resistant isolates showed a 1.6-fold to eightfold up-regulation of mRNA level for CDR1 gene, and 3.7- to 52-fold up-regulation of mRNA level for CDR2 gene. The mRNA of MDR1 was up-regulated by 1.6- to 2.1-fold in GZ03, GZ17, 592, and 4266 and down-regulated in the others (except for GZ23, which was similar to the control). The mRNA of ERG11 was up-regulated by 1.1- to 2.8-fold in GZ03, GZ17, GZ23, GZ51, 592, and 4266 and down-regulated in the other 9 isolates. The mRNA of FLU1 was up-regulated by 1.3- to 5.6-fold in GZ03, GZ17, GZ23, GZ34, 592, and 4266, but not in other isolates (Table 2).

Table 2 N-fold differential mRNA transcription of CDR1, CDR2, MDR1, ERG11, and FLU1 in the 14 fluconazole-resistant isolates compared with sensitive C. albicans ATCC10231
Full size table

Protein level of Cdr1 and Cdr2 in the resistant isolates

Cdr1p expression was determined in the 14 resistant isolates as well as in S. cerevisiae AD/CDR1. Comparison of the blots intensity showed Cdr1 semi-quantification ratios of the 14 isolates were all <100 %, 4 of them were <25 %, and another 4 were <50 %. Cdr1p expression in the resistant isolates was lower than that in S. cerevisiae AD/CDR1. Cdr2p expression was variable in different isolates. The Cdr2p blot intensity ratios of GZ17, GZ18, GZ23, GZ29, and 592, compared with S. cerevisiae AD/CDR2, were more than 100 %, and not in the others (Table 3).

Table 3 Western blot intensity ratio of expression for Cdr1p and Cdr2p from the 14 fluconazole-resistant isolates compared with S. cerevisiae AD/CDR1 and AD/CDR2, respectively
Full size table

Discussion

The ERG11 mutations existing in the susceptible isolates or standard strains were most likely unrelated to C. albicans resistance to fluconazole. We are most interested in those mutations that occur in the resistant isolates but not in the susceptible isolates. Particularly, homozygous G487T and T916C mutations resulting in two amino acid changes (A114S and Y257H, respectively) in the target enzyme Erg11p were exclusively present in the resistant isolates. In 14 resistant isolates derived from different patients in two different cities about 1800 km away from each other, identical ERG11 mutations G487T and T916C were detected, and no other mutations were observed in this gene. G487T and T916C along with T395A had been found in an induced fluconazole-resistant strain of C. albicans, and neither of them was present in the fluconazole-susceptible parent strain or any other susceptible strains as reported in 2006 (Jiang et al. 2006). Ge et al. 2010 reported in that these two homozygous ERG11 mutations were exclusively present in 23 vaginal isolates of C. albicans, which were characterized as microsatellite CAI 32-46 genotype and showed much lower susceptibility to fluconazole than the other isolates. It was hypothesized that C. albicans isolates with genotypes CAI 32-46 might be more resistant to azoles based on their previous studies (Ge et al. 2010). G487T and/or T916C might be related to fluconazole resistance in C. albicans, although it is possible that these 14 isolates are derived from the same resistant strain in the population. A114 locus in Erg11p is near the substrate channel, and mutations in this position may decrease the binding affinity between azole molecule and the target. Y257H is located in G helix, which is far away from the active center or substrate access channel of the protein. A multi-drug resistant clinical C. albicans isolate was reported harboring a single A114S substitution in Erg11p as well as a sequence insertion in Erg5p (Martel et al. 2010). Although there was no experimental evidence, this report still might be regarded as a support to the linkage between A114S and azole resistance.

T495A, A530C, and T541C found in another resistant isolate GZ04 have been studied extensively. T541C was shown to contribute to azole resistance (Kakeya et al. 2000). Mutations in T1493 locus have been reported as T1492C and T1493C in resistant isolates leading to the amino acids changes of F449L and F449S in Erg11p (Marichal et al. 1999; Perea et al. 2001). F449 is close to the terminal of I helix and is located in front of the heme group. Substitutions in this position have the chance to influence the function of the active center. The role of T1493A (F449Y) in GZ04 is remained to be further investigated.

The relative quantification PCR was used to investigate ERG11 mRNA in the 14 fluconazole-resistant isolates with G487T and T916C. Five isolates (GZ03, GZ17, GZ23, 592, and 4266) had up-regulated ERG11 transcription, whereas in the others (GZ09, GZ15, GZ16, GZ18, GZ29, GZ34, GZ51, GZ58, and 4263), ERG11 transcription was down-regulated in comparison with that in the standard strain ATCC 10231. Based on previous studies (Marr et al. 1998; Park and Perlin 2005; White et al. 2002), the expression level of ERG11 was not correlated with fluconazole resistance. Therefore, the up- or down-regulation of ERG11 transcription was not regarded as the key point concerning resistance mechanism in the 14 isolates, especially in those with down-regulated ERG11 transcription.

The mRNA level of CDR1 and CDR2 in all the resistant isolates with G487T and T916C in ERG11 was higher than in the control. However, Western blot showed that Cdr1 expression in these isolates was lower than that in S. cerevisiae AD/CDR1. And the data analyzed with t test suggest that all of the differences are significant (P < 0.0005) when the resistant isolates are considered as a whole. It cannot be asserted yet overexpression of CDR1 is not contributed to the azole resistance of these clinical isolates since there is consistent and equivalent hyperexpression of Cdr1p in S. cerevisiae AD/CDR1. But Cdr1p levels of different isolates are distinct from each other, and moreover, 8 of them (GZ09, GZ15, GZ16, GZ29, GZ34, GZ51, GZ58, and 4263) are lower than 50 %, and GZ15, GZ16, GZ58, and 4263 are even lower than 25 %, in which overexpression of CDR1 may not play an important role. Cdr2p overexpression plays a much less significant role than Cdr1p in fluconazole resistance (Holmes et al. 2008). Although Cdr2p expression level in some isolates was higher than that in the control strain S. cerevisiae AD/CDR2, CDR2 might not play a role in the resistance formation of the 14 resistant isolates, especially in those strains (GZ03, GZ09, GZ15, GZ16, GZ34, GZ51, GZ58, 4263, and 4266) in which both CDR1 and CDR2 expression were down-regulated. The bands for Cdr2p of some clinical isolates in Western blot were fragmentary (GZ16 and GZ51) or blank (GZ34), and semi-quantity value of them was relatively lower, and even reached zero (GZ34), while the CDR2 mRNA transcription of these isolates was much higher. Cdr2p translation after transcription of the series of clinical isolates might be instable and weaker than Cdr1p. In addition, there was a possibility that a CDR2 nonsense mutation existed in GZ34 leading to zero production of Cdr2p.

MDR1 overexpression has been regarded as another important factor in fluconazole resistance because of its efflux pump function. MDR1 was observed to be up-regulated in the resistant C. albicans isolates. However, up-regulation of MDR1 has also been reported in some susceptible isolates (Goldman et al. 2004; White et al. 2002). MDR1 mRNA level in four (GZ03, GZ17, 592, and 4266) of our resistant isolates was up-regulated, but in the other isolates, it is not up-regulated. It is possible that MDR1 might be involved in conferring resistance in the four isolates with up-regulated MDR1 mRNA. In contrast, it is less possible that MDR1 is involved in the resistance in the other ten isolates that have reduced level of MDR1 mRNA.

FLU1, which has been extensively studied in azole resistance, was not significantly up-regulated in the resistance isolates (except the four isolates GZ03, GZ17, 592, and 4266) compared to the susceptible control strain ATCC 10231. Therefore, FLU1 might be unrelated to the formation of fluconazole resistance in these isolates.

It should be noticed that the five resistant isolates (GZ09, GZ15, GZ16, GZ58, and 4263) are interesting. Neither up-regulation of mRNA transcription of MDR1, ERG11, or FLU1 nor higher protein expression of Cdr1p or Cdr2p was detected in them. ERG11 mutations G487T and/or T916C might play a most important role in the formation of fluconazole resistance of the five isolates.