TEXT

Top Abstract Text References

Amphotericin B lipid complex (ABLC) and liposomal amphotericin B (L-AmB) are formulations of amphotericin B that have been shown to have decreased toxicities compared to amphotericin B deoxycholate (AmB). It is unclear how active amphotericin B is released from these lipid formulations in vivo. Phospholipases can hydrolyze ester linkages in glycerophospholipids (3), and it has been postulated that phospholipases of both host and fungal origin can release active amphotericin B from ABLC (5). A previous report demonstrated that phospholipase-deficient Candida albicans mutants created using random chemical mutagenesis were resistant to ABLC in vitro (5). Interestingly, these mutants remained sensitive to AmB in vitro and were found to be deficient in extracellular phospholipase activity. It was concluded that fungal phospholipases were important for the in vitro susceptibility testing of ABLC and that the decreased toxicities of ABLC relative to AmB may be due to selective release of active amphotericin B at the sites of infection through the actions of fungal and/or host-derived phospholipases (5).

The predominant secreted phospholipase in the human pathogenic fungi C. albicans and Cryptococcus neoformans is phospholipase B (1-4). Molecular pathogenesis studies have demonstrated that phospholipase B is an important virulence factor for both of these fungi (2, 4). In these studies, phospholipase-deficient mutants were created using targeted gene disruption, and we wanted to take advantage of these strains for in vitro susceptibility testing against various formulations of amphotericin B.

The cloning of genes encoding for extracellular phospholipase B activities from C. albicans and C. neoformans, and the creation of phospholipase-deficient mutants using targeted gene disruption has been described (2, 4). The plb1 strain is a phospholipase-deficient mutant derived from the C. neoformans wild-type strain H99. The caplb1 strain is a phospholipase-deficient mutant derived from the C. albicans wild-type strain SC5314. All strains were maintained in glycerol stocks at 80°C and grown on either potato dextrose or Sabouraud dextrose agar (Difco Laboratories, Detroit, Mich.).

Susceptibility testing for C. neoformans was performed according to National Committee for Clinical Laboratory Standards (NCCLS) protocol M27A with the modification that antibiotic medium 3 was substituted for RPMI 1640. Strains were tested against AmB (Fungizone; Bristol-Myers Squibb, Princeton, N.J.), ABLC (Abelcet; The Liposome Company, Princeton, N.J., and L-AmB (Ambisome; Fujisawa Healthcare, Deerfield, Ill.). MICs at which 80% of isolates were inhibited (MIC₈₀s), MIC₁₀₀s, and minimum fungicidal concentrations (MFCs) were determined. MFCs were determined by plating 100 µl from each tube that showed no growth onto Sabouraud dextrose agar. The MFC was considered the lowest concentration of drug that yielded three colonies or fewer (i.e., 97% killing) after incubation at 37°C for 3 days.

Susceptibility testing for C. albicans was performed according to the NCCLS M27-A broth microdilution methodol in RPMI 1640 (American Biorganics Inc., Niagara Falls, N.Y.). The antifungal activities these drugs were also tested using two additional media: Sabouraud dextrose broth (SAB) and yeast nitrogen base (YNB: Difco Laboratories) supplemented with 0.5% glucose. An inoculum suspension of 2 × 10³ to 5 × 10³ CFU/ml was prepared in RPMI 1640, SAB, or YNB. The effect of other inocula on MICs was examined using three different inoculum sizes (0.5 × 10³, 2.0 × 10³, and 5.0 × 10³ CFU/ml). Additionally, the susceptibilities of both strains were evaluated at both 30 and 35°C. In accordance with the NCCLS M27-A document, Candida krusei (ATCC 6258) was used as a control strain and tested with each assay. MICs obtained for this quality control strain were within the expected range. All assays were done in duplicate.

Susceptibility data for the C. neoformans strains are shown in Table 1. There were no significant differences in MIC₈₀, MIC₁₀₀, and MFC between the phospholipase mutant (plb1 strain) and the isogenic wild-type strain (H99) for the three formulations of amphotericin B tested. Thus, extracellular phospholipases appear to play no role in in vitro susceptibility testing of C. neoformans against ABLC, L-AmB, and AmB.

Susceptibility testing for the C. albicans strains was performed using a variety of conditions in order to control for any effects of temperature and media on phospholipase activity. The data are presented in Tables 2 and 3 and demonstrate that there were no significant differences in the ABLC MICs for the two strains under any of the conditions tested. The MICs of ABLC for both strains were 1 to 3 dilutions lower than those of L-AmB, and AmB, and all of the results were consistent at 48 and 72 h (Table 2). It also appears that larger inoculum, increased temperature, and longer incubations may increase ABLC MICs by 1 to 2 dilutions for both the wild-type and phospholipase-deficient mutant strains (Table 3). Presumably, the higher MICs are due to increased growth of yeast under these conditions.

These data demonstrate that mutants of C. neoformans and C. albicans specifically lacking extracellular phospholipase activities do not have any significant changes in in vitro susceptibilities to ABLC or L-AmB compared to the corresponding parent strains having wild-type phospholipase activities. Previous data have shown that these two yeasts have extracellular phospholipase activity when grown under the conditions used in in vitro susceptibility testing (2, 4; unpublished data). We conclude that extracellular phospholipase B activities of yeast origin do not influence susceptibility testing of these agents, and phospholipase activities do not have to be controlled for when ABLC or L-AmB MIC data for different strains are being compared.

Our results are quite different from what was found in an excellent study by Swenson et al. (5). These investigators selected for mutants that were resistant to ABLC and found that these mutants retained in vitro sensitivity to AmB. Furthermore, they found that these mutants were also deficient in extracellular phospholipase activity. They attributed the selective resistance of these mutants to ABLC to the decreased levels of phospholipases, and they were able to restore sensitivity to ABLC by adding exogenous phospholipases to the strains. There are several explanations for the discrepancies between our results and those of Swenson et al. The strains used by Swenson et al. were generated through random chemical mutagenesis and likely have other undefined mutations besides the phospholipase deficiency that was detected. Those mutants were selected for ABLC resistance, rather than phospholipase deficiency, and the concern is that the three independent mutants share some mutation in a gene (or genes) other than phospholipase that is the actual cause of the ABLC resistance. On the other hand, the phospholipase B mutant strains used in the present study resulted from targeted mutagenesis of the phospholipase B-encoding genes, and sensitive radiometric assays confirmed the absence of the vast majority of extracellular phospholipase activities (2, 4). There was no preselection for ABLC resistance; thus, the use of these strains appears to be more valid in determining whether extracellular phospholipases affect ABLC MIC values. The data from Swenson et al. do support the possibility of host phospholipases having a role in the release of active amphotericin B from ABLC in vivo, but that is beyond the scope of our study.