INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Azole antifungal agents inhibit sterol 14-demethylation (referred to as 14-dMe) in ergosterol biosynthesis, thus causing the accumulation of 14-methylated sterols in the fungal membranes. Notably, however, the effect of 14-dMe inhibition on cell growth is not simple. While cells of some fungi cannot grow in the presence of 14-dMe deficiency (i.e., they are intolerant to 14-dMe deficiency), those of other fungi can (i.e., they are tolerant to 14-dMe deficiency). Thus far, Saccharomyces cerevisiae has been thought to belong to the former class (3), whereas Candida albicans has been thought to belong to the latter class (2, 3, 7, 9, 10). On the other hand, it is generally assumed that 14-dMe inhibition constitutes the basis of the therapeutic efficacies of azole antifungal agents. In effect, there is evidence that C. albicans cells grown under conditions of 14-dMe deficiency show increased susceptibility to the fungicidal mechanisms of phagocytes (8, 11), suggesting that the clinical efficacies of this class of drugs are brought about by cooperation of 14-dMe inhibition and host defense mechanisms.

In fungi intolerant to 14-dMe deficiency, the minimum concentration of an azole drug required for a complete inhibition of 14-dMe (MDIC [12] for minimum demethylation-inhibitory concentration) is expected to be equal or at least close to its MIC. However, for fungi tolerant to 14-dMe deficiency, for which the MDIC is lower than the MIC by definition, the MIC gives no information about the MDIC, which should be important in clinical settings. Previously, we have described a simple method for estimation of the MDIC of an azole for C. albicans (12). That method uses growth inhibition by acetate in 14-dMe-deficient cells and determines the MIC in an acetate-containing yeast extract-peptone-glucose (YEPG) medium. In the present study, we found, by using this method, that some species of the genus Candida were tolerant to 14-dMe deficiency, like C. albicans, while others were not. Furthermore, we also showed that measurement of the 80% inhibitory concentration (IC₈₀) in a special medium, now widely used as a routine medium for azole susceptibility testing, gave results fairly close to the MDIC estimate.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Fungal strains. The fungal strains used in this study are listed in Table 1.

Culture media. YEPG medium and acetate (0.24 M)-supplemented YEPG (YEPG-Ac) medium were described previously (12). YEPG and YEPG-Ac were solidified with 2% (wt/vol) agar to give YEPG agar and YEPG-Ac agar, respectively. The RPMI 1640 medium (RPMI) was as described previously (5), and it was supplemented with 0.24 M sodium acetate to give RPMI-Ac. Azole drugs were dissolved in dimethyl sulfoxide and were added to sterile media, with the final concentration of the solvent being adjusted to 1% (vol/vol).

Agar plate assays of azole susceptibility. Assays with YEPG or YEPG-Ac agar plates containing a concentration gradient of an azole were done as described previously (12).

MIC measurement and estimation of MDIC by dilution assay. A test cell suspension (10⁷ CFU/ml) was prepared by washing and suspending in distilled water cells grown in YEPG at 35°C overnight with shaking. The conventional MIC of an azole was measured as follows. Twofold dilutions (0.1 ml each) of the test drug made in YEPG or RPMI were mixed in plastic microplate wells with 0.1-ml aliquots of a 100-fold dilution of the cell suspension made in the same medium (containing 10⁴ CFU), and the plates were incubated at 35°C for 2 days without agitation. The lowest drug concentration that gave a visibly clear culture was taken to represent the MIC. It should be noted that the MIC thus obtained is not equivalent to the IC₈₀ (see below) determined by the routine standardized by the National Committee for Clinical Laboratory Standards (5) even when RPMI is used. The MDIC of an azole was estimated by measuring the MIC in YEPG-Ac or RPMI-Ac (signified by MIC_Ac) in a way similar to that described above.

Determination of IC₈₀ of the azole. Determination of the IC₈₀ of the azole was carried out in RPMI by the method standardized by the National Committee for Clinical Laboratory Standards (5). Thus, dilution assays were done with RPMI as described above, and the drug concentration that caused 80% growth inhibition was determined by turbidimetry, followed by interpolation when appropriate.

Sterol analysis. Extraction of cellular lipids and thin-layer chromatography on silica gel plates (Merck) were carried out as described previously (9). Identification of sterols was achieved by comparing the R_fs with the reference values obtained in our previous work (9).

Chemicals. Fluconazole (FLCZ) was donated by Pfizer Pharmaceuticals (Tokyo, Japan); ketoconazole (KCZ) and itraconazole (ITZ) were gifts from the Janssen Research Foundation (Beerse, Belgium).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Relationship between 14-dMe deficiency and acetate-mediated growth inhibition in Candida species. We have previously shown that azole-treated C. albicans cells are subject to growth inhibition in YEPG-Ac and that the extent of this inhibition increases as the degree of 14-dMe (the level of cellular ergosterol) diminishes, finally becoming complete when the azole concentration is high enough to make the ergosterol undetectable (12). (In a previous study [12], some cell growth can be seen in YEPG-Ac even in the virtual absence of 14-dMe [Fig. 5A therein], giving the impression that the MIC_Ac is higher than the MDIC. However, this is due to the high initial cell density (5 × 10⁵ CFU/ml; A₅₄₀  0.04) needed for sterol analysis, combined with gradual growth cessation intrinsic to azole-treated cells [see Fig. 4A of reference 12]. In MIC assays, an initial cell density of 5 × 10⁴ CFU/ml is used [see Materials and Methods], with which the residual growth alone barely yields visible turbidity. Hence, the MIC_Ac roughly coincides with MDIC.) On the basis of these observations, we have proposed that the MDICs of azoles can be estimated by determining the MIC_Ac, i.e., the MIC measured in YEPG-Ac (12).

View larger version (92K): [in this window] [in a new window]   FIG. 1.   Growth of Candida strains on a concentration gradient of FLCZ. (A) YEPG agar; (B) YEPG-Ac agar. Top, photographic images; bottom, schematic representations with gradations depicting growth. FLCZ concentrations, 0 to 10 µg/ml from left to right (I to IV). Rows: 1, C. tropicalis 820567; 2, C. kefyr B2455; 3, C. parapsilosis MCC499; 4, C. guilliermondii IFO0454; 5, C. glabrata CBS138; 6, C. krusei B2399.

View larger version (46K): [in this window] [in a new window]   FIG. 2.   Thin-layer chromatography analysis of sterols from cells grown on a YEPG agar plate containing the FLCZ concentration gradient (plate A in Fig. 1). Lanes: 1, C. tropicalis 820567; 2, C. kefyr B2455; 3, C. parapsilosis MCC499; 4, C. guilliermondii IFO0454; 5, C. glabrata CBS138; 6, C. krusei B2399; I to IV, sampling zones (see Fig. 1). Identification of sterols: a, ergosterol; b, 4,14-methylated sterols; c, 4,4',14-methylated sterols.

Relationship between MIC and MIC_Ac (MDIC) of the azole. On the basis of the results presented above, we went on to determine the MIC_Ac and MIC using YEPG-Ac and YEPG, respectively, with three different azoles and a wider array of strains of the genus Candida (Table 2). A strain of S. cerevisiae was included as a reference. It has been reported that cells of S. cerevisiae are incapable of growth in the absence of 14-dMe (3). In accordance with this notion, the MIC_Ac and MIC of each of the azoles used were shown to be identical for the test strain of this yeast. On the other hand, for all C. albicans strains except B59630 the MIC_Ac was much lower than the MIC. Strain B59630 is an azole-resistant mutant (6) which overexpresses the drug efflux pump Cdr1 (1). While the MIC did not distinguish B59630 from the other C. albicans test strains, the MIC_Ac did: compared with the other strains, the MICs of all three azoles for strain B59630 were elevated, with the increase being most marked with FLCZ. Notably, however, this strain was found to be similar to the others in terms of the relationship between the MIC_Ac and the MIC: the MIC_Ac and MIC were sharply different from each other for KCZ and ITZ, although they were fairly close to each other for FLCZ due to the marked increase in the MIC_Ac. The test strains of C. glabrata, C. krusei, and C. parapsilosis were similar to S. cerevisiae in this respect; differences between the MIC_Ac and the MIC were absent or only slight, with the factor being not more than 2. The behaviors of the strains of C. guilliermondii, C. kefyr, and C. tropicalis were similar to those of the C. albicans strains, with the MIC_Ac being lower than the MIC at least by an order of magnitude.

Comparison of YEPG and RPMI. As a routine of the clinical laboratory, azole susceptibility testing is done by measuring the IC₈₀ in RPMI by the standard procedure (5). For this reason, we wanted to see what would happen if YEPG was replaced by RPMI in our assay. We found that the RPMI-based MIC_Ac and MIC of FLCZ were similar, if not identical, to the YEPG-based counterparts (Table 3).

Coincidence between IC₈₀ and MIC_Ac (MDIC). We also compared the MIC_Ac and the MIC with the IC₈₀ and found that the IC₈₀ was close to the MIC_Ac, the estimated MDIC, for FLCZ (Table 3). It is this feature of IC₈₀ that probably made possible the identification of the FLCZ-specific azole resistance in B59630 (6).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

The utility of acetate-mediated growth inhibition for estimation of the MDICs of azoles has previously been demonstrated with strains of C. albicans (12). In the present study, it was extended to other Candida species (Fig. 1 and 2), which justified comparison of the MDIC, as estimated from the MIC_Ac, and the MIC with a larger set of strains of the genus Candida (Table 2). We found that the test strains were clearly divided into two distinct classes irrespective of the azole used. For one class, the MDICs and MICs were practically identical, indicating that cells were intolerant to 14-dMe deficiency; for the other class, the MDICs were much lower than MICs, indicating that the cells were tolerant to 14-dMe deficiency. It is unclear whether the tolerance or intolerance to 14-dMe deficiency is a species-dependent phenotype. While C. albicans may possibly be homogeneous with respect to this trait, there is no reason to believe that other species may also be so. More strains must be tested before a conclusion can be drawn.

The dichotomy between tolerance and intolerance to 14-dMe deficiency has two practical aspects that are worth mentioning. First, it has nothing to do with the MIC_Ac-based estimation of the MDIC, the information that should be needed for clinical purposes (8, 11), and hence will cause no problem in the use of MIC_Ac as the key parameter in azole susceptibility testing. Second, the possibility exists that azoles might be clinically more effective against fungi intolerant to 14-dMe deficiency than those tolerant to 14-dMe deficiency. This is because fungal cells of the latter class are still capable of growth under 14-dMe-deficient conditions and their control within the human body is perhaps more dependent on the host's defense mechanisms.

The test results with RPMI are also of note. First, the MIC_Ac (as well as the MIC) measured by use of RPMI and YEPG were practically equivalent. Since RPMI is nutritionally much poorer than YEPG for Candida yeasts, this finding seems to suggest that the choice of the basal medium is not of critical importance in the determination of the MIC_Ac (and, hence, in the estimation of the MDIC). Second, the finding that the IC₈₀ (RPMI-based) is in effect close to the MDIC should be significant because it gives some basis to the totally empirical parameter IC₈₀. However, as a tool for estimation of the MDIC, the YEPG-based MIC_Ac has the obvious advantage of being theoretically well based, simple to measure, accurate, and inexpensive.

The underlying mechanism(s) for the dichotomy with regard to the effect of 14-dMe deficiency on cell growth is an interesting question that remains to be answered. We consider the isolation of mutants with alterations in this trait to be the approach of choice. Studies along this line are under way in the authors' laboratory. Also worth study is the applicability of the present method to azole susceptibility testing of filamentous fungi. This possibility is intriguing in view of the lack of a convenient procedure suitable for this purpose.