Effect of Preexposure to Triazoles on Susceptibility and Virulence of Rhizopus oryzae

Abstract

Triazole prophylaxis has become the norm in patients with hematological malignancies. Breakthrough infections caused by Mucorales during triazole prophylaxis remain a challenging problem. We found that preexposure of Rhizopus oryzae to antifungal triazoles (fluconazole, voriconazole, posaconazole, and itraconazole) did not modify the in vitro susceptibility of Rhizopus oryzae to posaconazole. In contrast, preexposure of Rhizopus to triazoles was associated with the enhanced in vitro susceptibility of R. oryzae to amphotericin B. Preexposure to posaconazole did not alter the virulence of R. oryzae in the fly model of mucormycosis.

TEXT

Mucormycosis is a severe, frequently fatal invasive mold infection that has emerged in immunocompromised patients with hematological malignancies and in hematopoietic stem cell recipients (1). Rhizopus oryzae is the most common cause of mucormycosis (2). Mucorales are not susceptible to fluconazole (FLU), itraconazole (ITRA), voriconazole (VRC), or the echinocandins (2). These antifungal agents, which are commonly used as prophylaxis and empirical and preemptive treatments in immunocompromised patients, target candidiasis and aspergillosis (3). Breakthrough mucormycosis is a persistent and challenging problem in patients receiving these antifungal agents (1, 3). Among the hypotheses that have been proposed to explain these breakthrough infections is the selective pressure that is imposed by antifungal agents, which creates an ecological vacuum that is filled by Mucorales or creates subtle changes in the susceptibility and virulence of Mucorales that have been exposed to triazoles (4, 5).

Posaconazole (PCZ), a triazole with broad-spectrum antifungal activity, has been increasingly prescribed for prophylaxis against a broad spectrum of fungal pathogens, including Mucorales (6). However, even with posaconazole prophylaxis, breakthrough mucormycosis is observed (7). While the increased virulence of VRC-preexposed Rhizopus was previously reported experimentally (4, 8), no such data are available for PCZ-preexposed Mucorales. To fill this gap in understanding, we investigated the in vitro susceptibilities of R. oryzae to PCZ and amphotericin B (AMB) after serial exposure to triazoles (FLU, ITRA, VRC, or PCZ). We also tested the virulence of PCZ-preexposed R. oryzae using a fly infection model of mucormycosis.

Three clinical isolates of R. oryzae (RO-969, RO-275, and RO-449) were obtained from patients with invasive mucormycosis at the MD Anderson Cancer Center and were used in all in vitro experiments. The strains were exposed to FLU, ITRA, VRC, or PCZ in 4 serial passages on yeast extract agar glucose (YAG) plates (64 μg/ml FLU, 1 μg/ml ITRA, 4 μg/ml VRC, or 0.25 μg/ml and 0.5 μg/ml PCZ, respectively). After each passage and after incubation for 48 h at 37°C sporangiospores were harvested, washed with saline, and immediately plated onto fresh drug-free YAG plates (control) and onto drug-containing YAG plates, as previously described (9).

Broth microdilution was performed as recommended by CLSI guidelines (10). Twofold serial drug dilutions of PCZ and AMB were prepared and inoculated with freshly isolated R. oryzae sporangiospores (10³ spores/ml). After 24 h of incubation at 37°C, the MICs of PCZ and AMB were determined. To determine the minimum fungicidal concentrations (MFCs) of PCZ and AMB, aliquots taken from wells that showed 100% growth inhibition were plated onto yeast extract-peptone-dextrose (YPD) agar plates (10). After 24 h of incubation at 37°C, the MFCs were recorded as the lowest drug concentrations at which no growth was observed. MICs were log-transformed, and calculated geometric MIC/MFC ratios were determined for all combinations. The statistical analysis compared the ratio of the baseline geometric MIC/MFC to the serial passage MIC/MFC (Welch's test for each drug individually, before and after). The experiment was repeated 3 times on different days.

Our MICs and MFCs for AMB and PCZ following serial passages are shown in Table 1. Preexposure to triazoles (FLU, ITRA, VRC, or PCZ) did not affect the in vitro susceptibility of R. oryzae to PCZ in any of the 3 isolates tested. The possibility that preexposure to a triazole can result in the attenuation of the activity of another triazole was previously studied in Aspergillus fumigatus. It was shown that preexposure of A. fumigatus to VRC did not change the MIC of PCZ; additionally, preexposure of A. fumigatus to PCZ did not change the MIC of VRC (9). In addition, sequential exposure of A. fumigatus to FLU resulted in increased MFC (but not an increased MIC) for ITRA and VRC (11). The results of the present study, using R. oryzae, are consistent with the concept that sequential exposure to triazoles with no Mucorales activity (FLU, VRC, or ITRA) or to triazoles with Mucorales activity (PCZ) does not significantly affect the MIC of PCZ.

TABLE 1.

Geometric mean MIC and MFC for AMB and PCZ obtained after four passages on YAG plates containing FLU (64 μg/ml), ITRA (1 μg/ml), VRC (4 μg/ml), or PCZ (0.5 μg/ml) for 3 R. oryzae strains (RO 969, RO 275, and RO 749)^a

Drug	GMIC, μg/ml (95% CIb)	Ratio of GMIC versus baseline (95% CI)	P valuec	GMFC, μg/ml (95% CI)	Ratio of GMFC versus control (95% CI)	P valuec
AMB baseline	0.11 (0.05–0.26)			0.68 (0.46–1.0)
Serial passage
FLU	0.10 (0.06–0.5)	0.92 (0.33–2.52)	0.87	0.34 (0.21–0.54)	0.5 (0.2–0.88)	0.02
ITRA	0.09 (0.04–0.19)	0.79 (0.27–2.24)	0.64	0.34 (0.23–0.50)	0.5 (0.30–0.82)	0.01
VRC	0.11 (0.07–0.18)	1.01 (0.42–2.45)	0.97	0.32 (0.22–0.46)	0.46 (0.28–0.76)	0.005
PCZ	0.09 (0.06–0.15)	0.79 (0.32–1.96)	0.58	0.23 (0.15–0.35)	0.34 (0.20–0.57)	0.005
PCZ baseline	3.17 (2.17–4.62)			4.32 (3.61–5.16)
Serial passage
FLU	2.16 (1.31–3.54)	0.68 (0.38–1.21)	0.19	3.43 (2.71–4.33)	0.79 (0.60–1.04)	0.09
ITRA	2.33 (1.63–3.33)	0.73 (0.45–1.18)	0.19	3.70 (3.10–4.42)	0.86 (0.68–1.08)	0.18
VRC	2.72 (1.70–4.35)	0.85 (0.49–1.50)	0.56	4.32 (3.61–5.16)	1.00 (0.79–1.26)	1.0
PCZ	2.33 (1.63–3.32)	0.73 (0.45–1.18)	0.19	4.32 (3.61–5.16)	1.00 (0.79–1.26)	1.0

^aThe baseline corresponds to the MIC and MFC of AMB and PCZ obtained after four passages on drug-free YAG plates. Geometric mean MICs (GMICs) and geometric mean MFCs (GMFCs) were stable for PCZ (P > 0.05). While GMICs were stable for AMB, a consistent decrease of GMFCs was observed after exposure to each azole.

^bCI, confidence interval.

^cWelch's t test versus control.

While preexposure to triazoles (FLU, ITRA, VRC, or PCZ) did not affect the geometric mean MIC of R. oryzae to AMB, a consistent decrease of the geometric MFC to AMB was observed for each azole (Table 1). However, this result was modest (one dilution of difference). This observation may be the consequence of modified ergosterol content in the fungal cell membrane, which is the target of AMB action (12). AMB activity was attenuated in Aspergillus fumigatus following exposure to ergosterol-depleting agents (13–17). The ergosterol content in fungal cell membranes seems to be the consequence of a complex regulation, as enhanced in vitro activity toward ITRA or caspofungin (CAS) was observed after ITRA or CAS preexposure in A. fumigatus (18). The results of the present study, using R. oryzae, are consistent with the concept that sequential exposure to triazoles may also result in enhanced in vitro activity toward AMB.

In order to study the virulence of R. oryzae (RO-969), we used Oregon^R wild-type (WT) flies as previously described (19). Flies were infected in the afternoon (3 p.m.) with 5 × 10⁷ spores/ml of Rhizopus serially passaged on the following media: (i) drug-free YAG plates, (ii) VRC-containing YAG plates (4 μg/ml), and (iii) PCZ-containing YAG plates (0.25 μg/ml and 0.5 μg/ml). Survival was assessed until day 7 after infection. Each experiment was performed in triplicate on different days. Differences in survival rates were analyzed using the log-rank test in the GraphPad Prism software (version 5.0; GraphPad Software, Inc., La Jolla, CA, USA). A P value of ≤0.05 was considered statistically significant (Mantel-Cox test).

No difference in survival rate was observed in flies infected with R. oryzae that was exposed or not exposed to PCZ, even when the strain was exposed to the higher PCZ concentration (P > 0.05, Mantel-Cox test) (Fig. 1). In contrast, increased virulence was observed in flies infected with R. oryzae that was exposed to VRC (P < 0.05, Mantel-Cox test) (Fig. 1B) as previously described (4, 8). Additionally, a previous study showed that serial passages of the same strain of R. oryzae (RO 969) on YAG plates containing either ITRA, AMB, or CAS did not influence its virulence in the fly model of mucormycosis (4). Virulence changes in triazole-preexposed A. fumigatus were addressed in a previous study that showed that preexposure to VRC or PCZ in vitro did not affect A. fumigatus virulence in flies (9). In contrast, preexposure to VRC seems to affect the virulence of R. oryzae (4, 8).

FIG 1.

(A) Survival rate of female WT flies infected with RO-969 grown serially on drug-free YAG plates (Rhizopus, 4 passages) and on PCZ-containing YAG plates (Rhizopus serially passaged on 0.25 μg/ml PCZ and Rhizopus serially passaged on 0.5 μg/ml PCZ). (B) Survival rate of female WT flies infected with RO-969 grown serially on drug-free YAG plates (Rhizopus, 4 passages) and on VRC-containing YAG plates (4 μg/ml; Rhizopus serially passaged on VRC).

Our study has several limitations. Specifically, we used only one strain of R. oryzae for the virulence assay. Therefore, the generalizability of our findings would be strengthened by testing multiple R. oryzae strains and other Mucorales species. In addition, it is unclear whether the experimental conditions of the inoculum, media, and O₂/CO₂ atmosphere that were used in the in vitro susceptibility study accurately simulated in vivo conditions. We did not test the effect of echinocandin preexposure on R. oryzae susceptibility and virulence. Finally, the mechanisms of the enhanced virulence of VRZ-pretreated R. oryzae remain to be further elucidated.

In summary, our findings support the hypothesis that breakthrough mucormycosis in patients receiving posaconazole is due mainly to poor immunity, high inoculum exposures, and/or low posaconazole levels, and is not due to the selection of resistance or of enhanced virulence of the infecting strain of R. oryzae following preexposure to triazoles.