|Home | About | Journals | Submit | Contact Us | Français|
Multilocus microsatellite polymorphisms in 27 clinical Candida albicans isolates were found to be clearly unrelated to in vitro paradoxical growth or trailing effect with caspofungin. These findings suggest that such in vitro phenotypes are either gained or lost too rapidly to be predicted by more stable genomic markers.
A majority of Candida albicans isolates exposed in vitro to caspofungin display a clear MIC endpoint, but some of them display an altered susceptibility phenotype referred to as either trailing effect (TE) or paradoxical growth (PG) (25, 27). The TE in Candida spp. is well known with azoles and has recently been described with echinocandins. In vitro TE is not linked to clinical outcome with azoles (16, 20), but its clinical significance with echinocandins remains unsolved (14). PG is reminiscent of the paradoxical effect observed for other cell wall-active antimicrobial agents, including beta-lactams, which is also known as the Eagle effect (10). These two in vitro susceptibility testing phenotypes are now well described (1, 16, 24, 25). About 15% of C. albicans isolates display PG in vitro (6, 11, 24). PG has been observed, but not reproducibly, with some isolates in vivo (7). The molecular mechanisms underlying PG have been elucidated recently. First, a compensatory increase in cell wall chitin synthesis was shown in one isolate exhibiting paradoxical growth (26). The roles of cell wall integrity and calcineurin pathways have been suggested by others (31), and recently these pathways have been found to enable C. albicans' survival in the presence of elevated caspofungin concentrations through the stimulation of chitin synthesis (29).
Genetic diversity in C. albicans has been occasionally reported to be associated with antifungal susceptibility, but its association with TE or PG has not yet been investigated. Typing systems for C. albicans, such as DNA fingerprinting with the moderately repetitive probe Ca3 (23) or multilocus sequence typing (MLST) (17), could associate clades with resistance to antifungal agents such as flucytosine (9, 28) (19) or amphotericin B (4, 15). Two studies produced conflicting results regarding the association of C. albicans mating type locus (MTL) homozygosity with azole resistance (18, 21). In one study (8), multilocus genotypes were not predictive of fluconazole resistance. Microsatellites have been used increasingly in recent years as molecular markers for population genetics and rapid genotyping of different organisms, including C. albicans (5, 22). This study aimed to determine whether multilocus microsatellite polymorphisms could be associated with PG or TE of C. albicans when exposed in vitro to caspofungin.
A set of 27 C. albicans isolates, identified by their carbohydrate assimilation profiles on Auxacolor (Sanofi Diagnostic Pasteur, France) and internal transcribed spacer (ITS) sequence analysis results (20), were analyzed in this study. Their characteristics are summarized in Table Table1.1. Briefly, they were isolated over a 1-year period in four teaching hospitals located in Marseille (France), each from different patients, wards, collection sites, and times. Caspofungin susceptibility testing was performed in duplicate with two commercially available assays, the Etest (AB Biodisk, Sweden), an agar diffusion test, and Sensititre Yeast One (Trek Diagnostic Systems Inc.), a liquid broth microdilution assay which differs from the CLSI method only by the use of alamarBlue and RPMI 1640 (2% glucose). Etest strips for caspofungin and Sensititre microdilution plates were used as recommended by the manufacturers. The assays were read after 48 h of incubation at 30°C. Quality control was performed using Candida krusei ATCC 6258 and Candida parapsilosis ATCC 22019. A clear endpoint was defined as the absence of growth of C. albicans at caspofungin concentrations above the MIC. PG was defined as growth of C. albicans in the presence of low caspofungin concentrations, no growth at intermediate concentrations, and growth again at high caspofungin concentrations (Fig. (Fig.1A).1A). TE was defined as a reduced but persistent growth of C. albicans at caspofungin concentrations above the MIC (Fig. (Fig.1B).1B). Eight isolates displayed either PG (n = 4) or TE (n = 4) when exposed to caspofungin. The Etest and Sensititre assays gave identical TE and PG results. C. albicans ATCC 90028 and the 18 remaining isolates displayed a clear MIC endpoint.
Candida albicans typing was carried out using CDC3, EF3, and HIS3 markers as described by Botterel et al. (5). Statistical analysis was performed using GENETIX (3) and FSTAT version 2.9.4 (12) software programs. The genetic diversities over the microsatellite loci were similar for the clear endpoint and in vitro TE or PG caspofungin susceptibility populations (Table (Table2).2). Nei's unbiased expected heterozygosity was 0.797 overall. The Hardy-Weinberg hypothesis of panmixia within populations was rejected for all three loci (P = 0.0008). The EF3 locus was in statistically significant linkage disequilibrium with both the CDC3 and HIS3 loci. No population structure was associated with altered in vitro caspofungin susceptibility (Fst [coefficient of genetic differentiation] = 0.003, P = 0.10), as illustrated in the factorial analysis correspondence graph, where isolate clusters were clearly independent from their phenotypes (Fig. (Fig.2).2). With respect to within-population genotype structure, the log-likelihood G test (13) confirmed the absence of significant genetic differentiation of C. albicans isolates (P = 0.21) associated with the phenotypes.
In this study, PG or TE upon exposure to caspofungin was easily visualized using the Etest (Fig. (Fig.1),1), with results that were fully concordant with those of a broth microdilution assay, while others (30, 31), using 2,3-bis(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide (XTT)-based in vitro viability studies, have reported PG upon exposure to caspofungin for C. albicans ATCC 90028, and this isolate repeatedly displayed a clear endpoint with both Etest and Sensititre assays. The multilocus microsatellite genotyping system revealed a high genetic diversity in the C. albicans sample but could not predict caspofungin in vitro PG and TE phenotypes as would be expected if these phenotypes were the result of inheritable differences among strains. In contrast to what has been suggested with respect to the CDC3 marker and reduced susceptibility to caspofungin (2), our findings indicate that loss of heterozygosity is not associated with in vitro TE or PG. Furthermore, there was no loss of genetic diversity within a population exhibiting distinct in vitro caspofungin susceptibility phenotypes. These findings suggest that PG or TE phenotypes in C. albicans upon exposure to caspofungin are either gained or lost too rapidly to be predicted by more stable genomic markers.
Published ahead of print on 11 January 2010.