Resistance to fluconazole in Candida albicans from AIDS patients correlated with reduced intracellular accumulation of drug

ELSEVIER

MICROBIOLOGY LETTERS FEMS Microbiology Letters 131(1995)

337-341

Resistance to fluconazole in Candida albicans from AIDS patients correlated with reduced intracellular accumulation of drug K. Venkateswarlu a, D.W. Denning b, N.J. Manning ‘, S.L. Kelly a7* a Krebs Institute for Biomolecular Research, Department of Molecular Biology and Biotechnology, Sheffield University, Sheffield SlO 2UH, UK b Department of Infectious Disease and Tropical Medicine, Department of Medicine, University of Manchester, North Manchester General Hospital and Hope Hospital, Saljord M6 8HD, UK ’ Neonatal Screening Laboratory, Sheffield Children’s Hospital, Western Bank, Sheffield SIO 2UH, UK Received 13 July 1995; revised 19 July 1995; accepted 19 July 1995

Abstract Mucosal candidosis is an almost inevitable consequence of AIDS. Resistance to fluconazole therapy associated with enhanced tolerance, detectable in microbiological estimation of sensitivity, occurs in up to 10% of cases with late-stage AIDS. We report here our biochemical analysis of the basis of resistance in a study of two susceptible and two resistant isolates. Resistance was not associated with a change in the target enzyme sterol 14a-demethylase, as indicated by equivalent levels of fluconazole inhibition of activity in extracts from all four isolates, or by mutations in sterol A5r6 desaturase as previously observed in Saccharomyces cerevisiae and Ustilago maydis. Reduced cellular content of fluconazole in the resistant isolates of between six to ten-fold was observed which could account for their resistant phenotype. Keywords: Candida albicans; Fhmnazole; Resistance; Candidosis

1. Introduction

Azole antifungals have proved to be valuable agrochemical and pharmaceutical agents, and their mode of action is through the inhibition of a cytochrome P450 present in fungi which undertakes 14a-demethylation of lanosterol in 5’uccharomyces cerevisiae and Candida glabrata, but of 24-methyl-

Corresponding author. Tel: +44 (114) 2824249; (114) 2728697. l

Federation of European Microbiological Societies SSDI 0378-1097(95)00282-O

Fax: +44

enedihydrolanosterol in most other fungi (for review see [l]). Plants and mammals also have an equivalent activity, but the azole antifungals are selective inhibitors of the fungal enzyme. The consequence of inhibition is a reduction in the levels of ergosterol, the main sterol of fungi, and an increase in substrate and other abnormal sterols produced by further metabolism of the substrate, but without removal of the 14a-methyl group. The altered sterol composition is not able to support further growth. Resistance to azole antifungals has been an important problem in agriculture for some years [2], but

338

K. Venkateswarlu

et ai. / FEMS Microbioloa

has also become a large obstacle to successful therapy in late stage AIDS patients as about 10% now develop mucosal candidosis resistant to fluconazole [3]. Also a recent study has identified several fluconazole resistant isolates of C. albicans causing fungaemia in non-AIDS patients [4]. Azole antifungals are not metabolised by fungi and previous resistance studies have indicated a number of mechanisms are possible. The best elucidated mechanism of resistance to date is in Saccharomyces cerevisiae where a change in sterol profile results from resistance caused by mutation in sterol A5,h-desaturase. Instead of accumulating a 6-hydroxylated sterol under treatment, 14a-methylfecosterol is produced instead which fulfils the requirements of the cell for growth [5]. Similar mutations occur in S. cerevisiae strains which are defective in sterol 14cu-demethylase and may account for their azole resistance and viability [6]. This mechanism would appear to have validity for other fungi and has been demonstrated in Ustilago muydis [7] while the association of a sterol 14cY-demethylase defect with a sterol A5T6-desaturase defect has been observed for C. albicans which accumulated 14a-methylfecosterol [8]. Resistance to ketoconazole antifungal has been studied in other strains of C. albicans obtained from a patient suffering chronic mucocutaneous candidosis and sterol biosynthesis was altered, but not in a way which correlated with resistance [9]. Other ketoconazole resistant isolates obtained from patients suffering the same condition were reported to exhibit altered uptake of the drug [lo]. We report here on a comprehensive evaluation of the biochemical basis of resistance in isolates from AIDS patients which indicates altered intracellular accumulation of drug in resistant isolates.

2. Materials and methods 2.1. Strains C. albicans isolates were obtained from AIDS patients. Ca505 and Ca033 were sensitive isolates of C. albicans, but Cal79 and Ca284 exhibited fluconazole resistance. Isolates were identified by the germ tube test and typical colony morphology as described previously [ill. All culture used was RPM1 1640 medium (Sigma), unless specified.

Letters 131 (1995) 337-341

2.2. Chemicals Unless specified, all chemicals were obtained from Sigma Chemical Co., Poole, UK. Fluconazole was purchased from Pfizer and itraconazole and ketoconazole from Janssen Pharmaceutics. [ l4 Clfluconazole (specific activity 22 &i/mg) was a gift from Pfizer and [2-‘4C]mevalonate, dibenzethylenediamine salt (specific activity 53 mCi/mmol) was obtained from Amersham. 2.3. Growth inhibition studies Stationary phase cells were obtained from plate cultures incubated at 37°C on Sabouraud medium (Difco) with 2% (w/v) Difco Bactoagar and were inoculated in 2 ml medium contained in a 60-ml Sterilin container at 10000 cells/ml. Treatment with various doses of antifungal compound occurred over 2 days at 37°C 150 rpm and growth assessed by cell counts and colony forming units/ml on YEPD consisting of 2% (w/v) glucose, 2% (w/v) Difco Bactopeptone, 1% (w/v) Difco yeast extract and 2% (w/v) Difco Bactoagar. Each test was repeated at least three times and minimum inhibitory concentrations were constant. 2.4. identification

of sterols by CC/MS

Samples for GC/MS were prepared from lOO-ml cultures in the exponential phase of growth on RPM1 1650. The cell pellet was saponified in 15% (w/v) KOH in 90% (v/v) ethanol at 80°C for 1 h. Nonsaponifiable lipids (sterols and sterol precursors) were extracted with 3 x 5 ml hexane and dried under nitrogen. Following silylation for 1 h at 60°C with BSTFA (20 ~1) in 100 ~1 of toluene, sterols were analysed by GC/MS (VG 12-250 (VG BIOTECH) using split injections with a split ratio of 2O:l. Sterol identification was by reference to relative retention time and mass spectra, as reported previously [5,6]. 2.5. Inhibition

of Ila-demethylase

activity

Fluconazole inhibition of P450 was investigated by assessing the cell-free biosynthesis of ergosterol, according to methods similar to previous studies [12-141. After growth of l-l culture in 2-l flasks at

K. Venkateswarlu et al. /FEMS Microbiology Letters 131 (1995) 337-341

150 rpm, 37°C to late logarithmic phase, cells were homogenised using a Braun disintegrator (Braun GmbH, Mesungen, Germany) operating at 4000 rpm with 4 X 30 s bursts with liquid carbon dioxide cooling. 20 g of glass beads (0.45-0.5 mm) were mixed with the cells and made up to 50 ml using 20% (w/v) glycerol, 100 mM phosphate buffer pH 7.4. Cell-free extracts were obtained following centrifugation at 1500 X g and protein concentration estimated using the bicinchoninic acid method (Sigma). The reaction mixture consisted of cell-free extract (924 pi/ml, lo-15 mg/ml protein concentration), cofactor solution (50 ~1; containing 1 pm01 NADP, 1 pmol NADPH, 1 pmol NAB, 3 pmol glucose 6-phosphate, 5 pmol ATP and 3 pmol reduced glutathione in distilled water), divalent cation solution (10 yl of 0.5 M MgCl, and 5 ~1 of 0.4 M MnCl,), solution of azole antifungal compound dissolved in dimethylsulphoxide (1 ~0, [2-‘4]-mevalonate (10 ~1; 0.25 PCi) and adjusted to pH 7.4 (by addition of 10 M KOH). The mixture was incubated at 37°C for 2 h with shaking (110 rpm) after which the reaction was stopped by adding 1 ml of freshly prepared saponification reagent (15% (w/v> KOH in 90% (v/v> ethanol). Non-saponifiable lipids (sterols and sterol precursors) were extracted with 2 X 3 ml petroleum ether (bp 40-60°C) and dried under nitrogen. The non-saponifiable lipid was applied to silica gel thin layer chromatography plates (ART 573, Merck) and developed using toluene: diethyl ether 9:l (v/v). Radioactive sterols were located by autoradiography and excised for scintillation counting. The production of 4-desmethyl sterol was assessed for inhibition as described previously [12-141 and comprised more than 30% of the sterol produced. Experiments were performed in triplicate and IC,, values for inhibition of ergosterol biosynthesis calculated. Table 1 Minimum Isolate

ca505 Ca033 Cal79 Ca284

inhibitory

concentration

of different antifungal

Minimum inhibitory

concentration

339

2.6. Comparison of azole content in cells Cellular content of fluconazole was investigated using lo9 cells incubated in 10V5 M 14Cfluconazole in 0.1 M potassium phosphate buffer pH 7 at 37°C 150 ‘pm. The cells were harvested by centrifugation, washed three times in 10 ml lop4 M unlabelled fluconazole at room temperature prior to collection on Whatman GFC filters. This procedure was used to establish a favourable fluconazole concentration gradient to reduce loss of radiolabelled fluconazole from the cells and to wash off non-specifically bound fluconazole. Cellular fluconazole content, a balance of uptake and efflux, reached a plateau after 30 min. Autoclaved cells indicated a background of nonspecific binding at less than 10% of the values shown. The samples were assayed for radioactivity on a Philips 4700 scintillation counter and efficiency was examined by the external standard method.

3. Results and discussion Table 1 shows the minimum inhibitory concentrations observed for the clinical isolates of C. albicans. They exhibited a range of sensitivities with isolates Cal79 and Ca284 being resistant to fluconazole in comparison to 0505 and Ca033. Cross-resistance was also seen to extend to ketoconazole and itraconazole. Susceptibility to amphotericin B varied with a maximum six-fold difference between strain Ca505 and both Cal79 and Ca284. The fluconazole resistant isolates also showed increased resistance to cycloheximide. Although comparison of sensitivities (MICs) between non-isogenic strains makes definitive conclusions as to identifying the cause of resistance difficult, biochemical changes likely to cause the resistance were investigated. These included in-

drugs for four C. albicans isolates

( FM)

Fluconazole

Itraconazole

Ketoconazole

Amphotericin

12.8 25.6 205.1 205.1

3.0 6.0 48.0 48.0

0.9 1.9 15.1 15.1

0.5 1.1 3.3 3.3

B

Cyclohexamide 10.7 17.8 71.1 71.1

340

K. Venkateswarlu et al. / FEMS Microbiology Letters 131 (1995) 337-341

vestigation of changes in sterol composition, in the target enzyme and in the cellular content of the drug. Analysis of the sterol composition of the resistant isolates showed that they exhibited no differences, which would be indicative of a defect in ergosterol biosynthesis corresponding to sterol 14cr-demethylase or sterol A5,6-desaturase (Table 2). All isolates accumulated ergosterol as their predominant sterol and smaller quantities of intermediates. Cal79 and Ca284 were observed to maintain their ergosterol levels under treatment with 12.8 PM fluconazole (the MIC observed for Ca505) at 65% and 61% of total sterol respectively. Previously the observation of altered amphotericin B resistance associated with reduced ergosterol levels in mutants defective in sterol biosynthesis has been central to the model of amphotericin B action through binding to ergosterol [El. However, the changes in resistance seen here between the isolates were not correlated with altered sterols. Investigation of the sensitivity of the target enzyme to inhibition was assessed through study of in vitro sterol biosynthesis. The assay indicates a reduction of ergosterol biosynthesis, together with accumulation of lanosterol which are separated on TLC and assessment of the IC,, (dose required for a 50% reduction) of ergosterol biosynthesis can be made. The possible inhibition of sterol A22-desaturase, also a P450 [16], does not change the mobility of sterol from that of ergosterol, hence the wide application of the assay in antifungal discovery studies. No difference between the isolates was observed in the sensitivity of the cell-free extracts in their IC,, for ergos-

Table 2 Relative percentage isolates Percentage

sterol compositions

of the four C. albicans

of total sterols

Sterol

ca505

Ca033

Cal79

Ca284

Ergosta-tetraenol Ergosterol Fecosterol Ergosta-5,7-dienol Episterol Obtusifoliol Eburicol Unidentified sterols

3.5 64.4 5.7 10.7 6.5 5.1 2.0 2.1

ND 65.1 2.4 ND 8.0 18.0 ND 6.5

ND 70.4 12.4 2.7 4.3 9.0 1.2 ND

ND 64.7 10.6 4.4 0.8 12.9 0.9 5.7

ND = not detected.

Table 3 (concentration of drug inhibiting incorporation of [2l&l “C]mevalonate into ergosterol by 50%) of fluconazole for in vitro ergosterol biosynthesis and intracellular accumulation of [‘4Cjfluconazole in C. albicans isolates Isolate

lC,, (nM)

[ l4 C]fluconazole (pmol/lO’ cells)

Ca505 Ca033 Cal79 Ca284

42.5 + 7.9 55.0+4.2 58.4k5.6 56.7k3.7

20.9kO.6 12.4kO.4 1.7kO.7 2.0 f 0.5

terol biosynthesis on treatment with fluconazole (Table 3). This assay would indicate alterations in the specific content or changes in the apoprotein of the target both of which could be envisaged to reduce the inhibitory effect of fluconazole. In contrast to studies on sterol content and enzyme sensitivity, a difference was observed between the isolates when the intracellular content of radiolabelled fluconazole was assessed after treatment. Substantially reduced content of fluconazole was detected in the resistant isolates of between six- to ten-fold when compared to the susceptible isolates. The levels detected were correlated with the rank order of the MICs, i.e. Ca505 < Ca033 < Ca179;Ca284. This paper concerns the initial investigation of the basis of fluconazole resistance in C. albicans isolates following fluconazole therapy of AIDS patients. Resistance was not associated with any alteration in genes of sterol biosynthesis of the type which alter the sterol profile under treatment or with an alteration in the target site which alters azole antifungal binding. Instead resistance was correlated with decreased content of the drug in cells. A similar change was observed for two ketoconazole isolates from patients suffering chronic mucocutaneous candidosis [lo] and this can now be anticipated as a common mechanism in AIDS patients. In the isolates studied here, cross-resistance to the other orally administered antifungal azoles was found suggesting a common mechanism may also be operative for other drugs. Recently, an example of multi-drug resistance have been associated with laboratory studies on an ABC membrane protein of Candida albicans [17] and this may be the basis of the changes observed here or

K. Venkateswarlu et al. / FEMS Microbiology Letters 131 (1995) 337-341

through another membrane transporter. Understanding the molecular basis of resistance in a range of clinical isolates is important as azoles are well established and extremely useful antifungal compounds. This may allow the development of improved therapy to overcome the problem which now exists.

Acknowledgements K.V. is supported by a Commonwealth Scholarship.

References [II

El

Kelly, S.L. and Kelly, D.E. (1993) Molecular studies on azole sensitivity in fungi. In: Molecular Biology and Applications to Medical Mycology (Maresca, B., Kobayashi, G.S. and Yamaguchi, H., Eds.), pp~ 199-213. Springer-Verlag, Berlin. Holloman, D. (1993) Resistance to azole fungicides in the field. B&hem. Sot. Trans. 21, 1047-1051.

[31 Baily, G.G., Perry, F.M., Denning, D.W. and Mandal, B.K. (19941 Fluconazole-resistant candidosis in an HIV cohort. AIDS 8,787-792. [41 Goff, D.A., Koletar, S.L., Buesching, W.J., Barnisham, J. and Fass, R.J. (1995) Isolation of fluconazole-resistant Candida albicans from Human Immunodeficiency Virus-negative patients never treated with azoles. Clin. Infect. Dis. 20, 77-83. [51 Watson, P.F., Rose, M.E., Ellis, SW., England, H. and Kelly, S.L. (1989) Defective sterol C5-6 desaturation and azole resistance: a new hypothesis for the mode of action of azole antitimgals. Biochem. Biophys. Res. Commun. 164, 1170-1175. b1 Kelly, S.L., Lamb, D.C., Corran, A.J., Baldwin, B.C. and Kelly, D.E. (19951 Mode of action and resistance to azole antifungals associated with the formation of 14a-methylergosta-8,24(28)-dien-3@,6a-diol. Biochem. Biophys. Res. Commun. 207, 910-915.

[71 Joseph-Horne,

341

T., Manning, N.J., Hollomon, D. and Kelly, S.L. (1995) Defective sterol Ay6’ desaturase as a cause of azole resistance in U&ago maydis. FBMS Microbial. Len. 127, 29-34. Bl Shimokawa, O., Kato, Y., Kawano, K. and Nakayama, H. (1989) Accumulation of 14-alpha-methylergosta-8,24(281dien-3-beta,6-alpha-diol in 14-alpha-demethylation mutants of Candida albicans-genetic evidence for the involvement of 5-desaturase. Biochim. Biophys. Acta 1003, 15-19. [91 Howell, S.A., Mallet, AI. and Noble, W.C. (1990) A comparision of the sterol content of multiple isolates of the Candida albicans Darlington strain with other clinically azole-sensitive and -resistant strains. J. Appl. Bacterial, 69, 692-696. DOI Ryley, J.F., Wilson, R.G. and Barrett-Bee, K.J. (1984) Azole resistance in Candida albicans. J. Med. Vet. Mycol. 22, 53-63. [ill Law, D., Moore, C.B., Wardle, H.M., Ganguli, L.A., Keaney, M.G.L. and Denning, D.W. (1994) High prevalence of antifungal resistance in Candida in AIDS. J. Antimicrob. Chemother. 34, 659-668. [121Marriott, M.S. (1980) Inhibition of sterol biosynthesis in Can&da albicam by imidazole-containing antifungals. J. Gen. Microb. 117, 253-255. t131 Barrett-Bee, K.J., Lane, A.C. and Turner, R.W. (19861 The mode of action of tolnaftate. J. Med. Vet. Mycol. 24, 155160. [141 Ballard, S.A., Ellis, S.W., Kelly, S.L. and Troke, P.F. (1990) A novel method for studying ergosterol biosynthesis by cell-free preparation of Aspergillus fumigatus and its inhibition by azole antifungal agents. J. Med. Vet. Mycol. 28, 335-344. [151 Kelly, S.L., Lamb, D.C., Taylor, M., Corran, A.I., Baldwin, B.C. and Powderly, W.G. (19941 Resistance to amphotericin B associated with defective sterol A8’ ’ isomerase in a Cryptococcus neofomums strain isolated from an AIDS patient. FEMS Microbial. Lett. 122, 39-42. [161Hata, S., Nishino, T., Katsuki, A., Aoyama, Y. and Yoshida, Y. (1987) Characterization of A ** -desaturation in ergosterol biosynthesis of yeast. Agric. Biol. Chem. 51, 1349-1354. P.D., Goffeau, A. and Balzi, E. 1171 Prasad, R., Wergibosse, (19951 Molecular cloning and characterization of a novel gene of Candida albicans, CDRl, conferring multiple resistance to drugs and antifungals. Curr. Genet. 27, 320-329.