|Home | About | Journals | Submit | Contact Us | Français|
We report on the in vitro activity of the Hos2 fungal histone deacetylase (HDAC) inhibitor MGCD290 (MethylGene, Inc.) in combination with azoles against azole-resistant yeasts and molds. Susceptibility testing was performed by the CLSI M27-A3 and M38-A2 broth microdilution methods. Testing of the combinations (MGCD290 in combination with fluconazole, posaconazole, or voriconazole) was performed by the checkerboard method. The fractional inhibitory concentrations were determined and were defined as <0.5 for synergy, ≥0.5 but <4 for indifference, and ≥4 for antagonism. Ninety-one isolates were tested, as follows: 30 Candida isolates, 10 Aspergillus isolates, 15 isolates of the Zygomycetes order, 10 Cryptococcus neoformans isolates, 8 Rhodotorula isolates, 8 Fusarium isolates, 5 Trichosporon isolates, and 5 Scedosporium isolates. MGCD290 showed modest activity when it was used alone (MICs, 1 to 8 μg/ml) and was mostly active against azole-resistant yeasts, but the MICs against molds were high (16 to >32 μg/ml). MGCD290 was synergistic with fluconazole against 55 (60%) of the 91 isolates, with posaconazole against 46 (51%) of the 91 isolates, and with voriconazole against 48 (53%) of the 91 isolates. Synergy between fluconazole and MGCD290 was observed against 26/30 (87%) Candida isolates. All 23 of the 91 Candida isolates that were not fluconazole susceptible demonstrated a reduced fluconazole MIC that crossed an interpretive breakpoint (e.g., resistant [MIC, ≥64 μg/ml] to susceptible [MIC, ≤8 μg/ml]) when fluconazole was combined with MGCD290 at 0.12 to 4 μg/ml. The activity of fluconazole plus MGCD290 was also synergistic against 6/10 Aspergillus isolates. Posaconazole plus MGCD290 demonstrated synergy against 14/15 Zygomycetes (9 Rhizopus isolates and 5 Mucor isolates). Voriconazole plus MGCD290 demonstrated synergy against six of eight Fusarium isolates. Thus, MGCD290 demonstrated in vitro synergy with azoles against the majority of clinical isolates tested, including many azole-resistant isolates and genera inherently resistant to azoles (e.g., Mucor and Fusarium). Further evaluation of fungal HDAC inhibitor-azole combinations is indicated.
At present, the azole class of antifungal agents constitutes one of the cornerstones of therapy for opportunistic mycoses due to many yeasts and molds (3, 16, 20, 24, 28, 30, 31, 33). Unfortunately, the clinical efficacy of this class of agents may be compromised by intrinsic or acquired resistance (11, 25, 27, 30). Resistance to azoles has been studied most extensively in Candida spp., in which the upregulation of genes encoding the lanosterol demethylase target enzyme (ERG11) and the Candida drug resistance (CDR) efflux pumps may occur upon exposure of the organism to azole antifungal agents. These alterations in gene regulation can result in increases in MICs and compromised clinical efficacy (25, 27, 30, 34). Given the proven safety, efficacy, and ease of use of these agents, the availability of strategies that may be used to avoid the emergence of resistance is important. Combination antifungal therapy with agents of different mechanistic classes could promote fungal killing and clinical efficacy and provide an alternative to monotherapy for patients with infections caused by multiresistant species and for patients who fail to respond to standard treatments.
Histone deacetylases (HDACs) are a family of enzymes which deacetylate lysines on core histones and other cellular proteins (9, 10, 32). They play an important role in gene regulation and also in the control of other cellular functions, such as proliferation, cell death, and motility (9, 10, 22, 32). Inhibitors of HDACs belong to several chemical classes that act by binding to the catalytic site of the enzyme, causing cell cycle arrest, apoptosis, and terminal differentiation (9, 22). HDAC enzymes have been explored as potential targets in the treatment of cancer cells and infections caused by several eukaryotic microorganisms (1a, 7, 9, 22, 26, 29). Modulation of fungal gene expression through fungal HDAC inhibition may be an alternative approach to the treatment of fungal infections (17, 29). Smith and Edlind (29) have shown in Candida albicans and two other Candida spp. that trichostatin A, a nonselective HDAC inhibitor with cytoxic properties in mammalian cells, markedly decreased the upregulation of the ERG11 and CDR genes following exposure to sterol biosynthesis inhibitors, such as fluconazole and terbinafine.
We previously examined the potential chemosensitizing interaction between a novel Hos2 inhibitor, MGCD290, developed by MethylGene, Inc. (Montreal, Quebec, Canada), and three triazole antifungal agents (fluconazole, itraconazole, and voriconazole) against a panel of 45 clinical isolates of Candida spp. (16 of which were fluconazole resistant) and 16 clinical isolates of Aspergillus spp. In the previous study, MGCD290 displayed synergy with fluconazole against 76% of the Candida isolates tested and with voriconazole and itraconazole against 69% of the Aspergillus isolates tested (8a). Our results suggest a potential clinical use for the combination of HDAC inhibitors and azoles in the treatment of fungal infections.
In the present study, we expand upon our initial findings by examining the interaction between MGCD290 and three triazoles (fluconazole, voriconazole, and posaconazole) against a larger and more diverse collection of yeasts and molds, most of which were azole resistant.
Isolates of Candida spp. (11 C. albicans, 14 C. glabrata, and 5 C. krusei isolates), Cryptococcus neoformans (10 isolates), Rhodotorula spp. (4 Rhodotorula glutinis isolates, 1 R. rubra isolate, and 3 Rhodotorula isolates not otherwise identified), Trichosporon spp. (5 isolates), Aspergillus spp. (3 Aspergillus fumigatus, 2 A. flavus, 2 A. niger, and 2 A. terreus isolates), isolates of the Zygomycetes order (5 Mucor isolates and 10 Rhizopus isolates.), Fusarium spp. (8 isolates), and Scedosporium apiospermum (5 isolates) were obtained from the organism collection of the Molecular Epidemiology and Fungus Testing Laboratory (University of Iowa, Iowa City). All isolates had previously been identified by standard mycological methods (14) and were stored as water suspensions or on agar slants until they were used in the study. The collection was selected specifically to maximize the number of isolates expressing resistance to one or more of the triazole antifungal agents.
The HDAC inhibitor MGCD290 was provided by the manufacturer (MethylGene, Inc.). Reference powders of fluconazole (Pfizer), posaconazole (Schering Plough), and voriconazole (Pfizer) were provided by their respective manufacturers. Stock solutions were prepared in RPMI 1640 (Sigma) buffered to a pH of 7.0 with 0.165 M morpholinepropanesulfonic acid.
The MICs of the three azoles and MGCD290 against the various yeasts and molds were determined by the broth microdilution method exactly as described in Clinical and Laboratory Standards Institute (CLSI) documents M27-A3 (4) and M38-A2 (6). The MIC endpoints for all four agents were designated as the first clear well (complete growth inhibition) for filamentous fungi or the first well that demonstrated a prominent reduction in growth (a ≥50% reduction relative to that for the growth control) for yeasts. The range of drug concentrations tested alone and in combination were 0.12 to 256 μg/ml for fluconazole, 0.007 to 8.0 μg/ml for posaconazole and voriconazole, and 0.25 to 32 μg/ml for MGCD290. Interpretive breakpoints for yeasts and fluconazole were those defined by the CLSI (5): susceptibility, ≤8 μg/ml; susceptible dose dependent (SDD), 16 to 32 μg/ml; and resistant, ≥64 μg/ml. Likewise, the breakpoints for voriconazole were those defined by the CLSI (5); susceptible, ≤1 μg/ml; SDD, 2 μg/ml; and resistant, ≥4 μg/ml. In the absence of azole breakpoints for filamentous fungi, we used those listed above. Provisional breakpoints for posaconazole were those defined above for voriconazole.
Testing of the combination of MGCD290 with each of the three triazoles was accomplished by means of a two-dimensional checkerboard procedure, as described in the Clinical Microbiology Procedures Handbook (21). The 96-well plates were prepared, according to CLSI M27-A3 standards (4), by using checkerboard combinations and were stored at −70°C until use. The plates were inoculated as directed in CLSI documents M27-A3 (4) and M38-A2 (6) and were incubated at 35°C for 48 h (for Candida spp., Aspergillus spp., Zygomycetes, Fusarium spp., and Trichosporon spp.) or 72 h (for Clostridium neoformans, Rhodotorula spp., and S. apiospermum). The MIC endpoints for the combinations were read as described above for the individual agents tested alone. Drug combination interactions were calculated algebraically by determining the fractional inhibitory concentration (FIC) as described previously (2). FICA was calculated as the MIC of drug A in the combination divided by the MIC of drug A alone, and FICB equals the MIC of drug B in the combination divided by the MIC of drug B alone. The sum of the FICs (ΣFIC) was calculated as follows: ΣFIC = FICA + FICB. The interpretation of ΣFIC was as follows (2): <0.5, synergistic; ≥0.5 to <4.0 indifferent (no antagonism); and ≥4.0, antagonistic. This method of evaluation of antifungal drug combinations was shown to be highly reproducible in a recent multicenter study (2).
Quality control was ensured by testing the following strains: C. parapsilosis ATCC 22019, C. krusei ATCC6258, and A. flavus ATCC 204304.
Table Table11 lists the MICs of the three triazoles and MGCD290 tested alone against the 91 clinical isolates of fungi. Notably, 71% of the 91 isolates were resistant to one or more of the azoles tested. MGCD290 exhibited activity against the various yeasts included in the study: the MIC50s ranged from 1.0 μg/ml for C. neoformans, to 2.0 μg/ml for Candida spp., 4.0 μg/ml for Trichosporon spp., and 8 μg/ml for Rhodotorula spp.; but they were 16 to >32 μg/ml for the filamentous fungi.
The antifungal activities of various combinations of the azoles with MGCD290 are shown in Table Table2.2. Overall, synergy was observed for 60.4% of the isolates tested with MGCD290 plus fluconazole, 50.5% of the isolates tested with MGCD290 plus posaconazole, and 52.7% of the isolates tested with MGCD290 plus voriconazole. Antagonism was observed with only one isolate of Trichosporon tested with the combination of MGCD290 and either fluconazole or posaconazole and three isolates of Trichosporon tested with MGCD290 plus voriconazole. An FIC could not be calculated for two isolates of Fusarium due to off-scale MICs for fluconazole, posaconazole, and MGCD290.
Our findings are consistent with those of our previous studies in which we demonstrated synergy between MGCD290 and the azoles against Candida spp. and Aspergillus spp. MGCD290, in contrast to nonselective HDAC inhibitors, has specific activity against fungal HDACs (MethylGene, Inc., data on file) and has some intrinsic antifungal activity (Table (Table1).1). In the current study, MGCD290 demonstrated synergy against Candida spp. when it was combined with each of the triazoles (Table (Table2).2). Although Smith and Edlind (29) observed minimal or no effects of the HDAC inhibitor trichostatin A on the activities azoles against C. glabrata or C. krusei, we observed synergy between MGCD290 and all three azoles, including fluconazole, which is typically less effective against these two species. In addition, the activity of MGCD290 with the azoles was synergistic against azole-resistant strains of C. albicans. Similar favorable interactions against the other yeasts and yeast-like fungi, C. neoformans, Rhodotorula spp., and Trichosporon spp., were observed with the MGCD290 combinations Notably, these favorable interactions were generally achieved with relatively low concentrations of MGCD290 (0.25 to 8.0 μg/ml) (Table (Table33).
Synergy or indifference against the filamentous fungi was also noted for the various combinations of MGCD290 with the azoles. The most striking observation was the enhanced activity of fluconazole against the various molds (Table (Table2).2). Despite a complete lack of activity of fluconazole against all of the filamentous fungal isolates when it was tested alone, synergy was observed with MGCD290 for 60% of Aspergillus spp., 40% of Zygomycetes, 50% of Fusarium spp., and 60% of S. apiospermum isolates tested. Although voriconazole alone was inactive against all of the Zygomycetes tested, synergistic activity was observed with 73.3% of the isolates when voriconazole was tested in combination with MGCD290. As expected, posaconazole alone was active against some of the Zygomycetes. However, a synergistic potentiation of the activity of posaconazole against 93% of the isolates was found when it was combined with MGCD290.
Although it is encouraging to find that the interaction between MGCD290 and the triazole antifungal agents is primarily one of synergy, such an interaction is most clinically relevant if it results in a shift to a more susceptible category. In Table Table33 we provide a complete listing of the individual isolates for which a categorical change that increased the susceptibility to a triazole was observed when the triazole was tested in combination with MGCD290. Among the 91 isolates examined, a change to a more susceptible category with one or more of the triazoles in combination with MGCD290 was observed with 23 of 30 (76.7%) Candida isolates, 8 of 8 (100%) Rhodotorula isolates, 7 of 10 (70%) Aspergillus isolates, 13 of 15 (87%) Zygomycetes isolates, 8 of 8 (100%) Fusarium isolates, and 4 of 5 (80%) S. apiospermum isolates. Given that all isolates of C. neoformans and Trichosporon spp. were susceptible to all three triazoles, a change to a more susceptible category was not possible; however, a decrease in the azole MIC against the majority of cryptococcal isolates was observed. Importantly, among the isolates of Trichosporon for which antagonism was observed with the MGCD290-azole combinations, a shift to a more resistant category was not observed.
Given the important role of fluconazole in the treatment of candidiasis, it is notable that all 19 fluconazole-resistant isolates of Candida spp. showed a shift to either SDD (6 isolates) or susceptible (13 isolates) when fluconazole was tested in combination with MGCD290 (Table (Table3).3). Likewise, four of eight fluconazole-resistant isolates of Rhodotorula showed a shift to either SDD (two isolates) or susceptible (two isolates) when they were tested with the combination.
Although it is well-known that fluconazole has no meaningful activity against filamentous fungi, the combination of fluconazole with relatively high concentrations of MGCD290 (8 to 32 μg/ml) resulted in a decrease in the fluconazole MICs from ≥256 μg/ml to 16 to 32 μg/ml for 6 of 10 isolates of Aspergillus spp., 1 of 8 isolates of Fusarium spp., 1 of 10 isolates of Rhizopus spp., 1 of 5 isolates of Mucor spp., and 2 of 5 isolates of S. apiospermum (Table (Table3).3). Even more striking was the observed decrease in the fluconazole MICs from ≥256 μg/ml to 4.0 μg/ml for one isolate of a Mucor sp., four isolates of Rhizopus spp., one isolate of a Fusarium sp., and one isolate of S. apiospermum. The mechanism for the apparent sensitization of filamentous fungi to inhibition by fluconazole is not known but clearly warrants further study in vivo.
Due to the more potent activities of both posaconazole and voriconazole against this collection of pathogenic fungi, fewer categorical shifts were noted (Table (Table3).3). A shift from posaconazole resistance (MIC, ≥4 μg/ml) to susceptibility (MIC, ≤1 μg/ml) was observed with four of five isolates of C. glabrata, two of five isolates of Fusarium spp., two of two isolates of Zygomycetes, and one of one isolate of a Rhodotorula sp. (Table (Table3).3). Likewise, a categorical shift from voriconazole resistance (MIC, ≥4 μg/ml) to susceptibility (MIC, ≤1 μg/ml) was observed with 2 of 2 isolates of C. glabrata, 1 of 4 isolates of Fusarium spp., 11 of 15 isolates of Zygomycetes, and 1 of 3 isolates of Rhodotorula spp. These major shifts in the susceptibility profiles obtained with the combination of azoles and MGCD290 are such that infections with these isolates could potentially be treatable by agents that would be considered inactive when they are used alone.
The results of this study confirm and extend our earlier findings regarding the potential for HDAC inhibitors to enhance the potencies of azoles against a collection of clinical isolates of both yeasts and molds. The surprising enhancement of the activity of fluconazole against filamentous fungi in general and of voriconazole against the Zygomycetes suggests a similar mode of action HDAC inhibitors against these fungi that begs further investigation. The effect of MGCD290 on azole activity may be due to the modulation of Hos2 transcriptional complexes, which results in the suppression of the azole-induced upregulation of genes involved in azole activity and resistance (1).
The use of combinations of antifungal agents to augment efficacy, expand the spectrum of activity, decrease toxicity, and reduce the emergence of resistance is not a novel concept (8, 12, 30). This generally entails the use of recognized antifungal agents for which complementary mechanisms of action have been proven (8, 12). In recent years, however, the concept of chemosensitization of fungal organisms to standard antifungal agents (e.g., azoles) by exposure to biological agents or nontraditional partner drugs, such as nitric oxide, cyclosporine, benzo analogues, or recombinant antibodies, has emerged (13, 15, 18, 19, 23). Each of these agents may alter the ability of fungal cells to respond or adapt to the insult of a specific antifungal exposure. In this regard, HDAC inhibition may be considered a novel approach to the sensitization of fungal cells to the azole class of antifungal agents at the level of gene transcription. By interfering with the organism's ability to upregulate genes important in resistance to sterol biosynthesis inhibitors, such as the azoles, these agents may both sensitize the organism to the action of the azoles and block the emergence of resistance during therapy. Furthermore, as observed with fluconazole activity in combination with MGCD290 against the intrinsically resistant organism C. krusei and filamentous fungi, the activity of HDAC inhibitors may be sufficient to overcome the inherent resistance of such fungi to azoles. HDAC inhibition may therefore minimize the risk of breakthrough infections that are becoming more prominent with the expanded use of limited-spectrum agents such as fluconazole (25).
In the present study, we provide evidence of the ability of a specific fungal HDAC inhibitor, MGCD290, to enhance the potency of both limited- and expanded-spectrum azoles against a broad collection of resistant yeasts and filamentous fungi. Not only have we demonstrated a general synergistic interaction between these agents, but we have also shown a distinctly favorable influence of MGCD290 on the MICs of the various azoles, resulting in a conversion from resistance to susceptibility. At this point, our observations are limited to in vitro data generated by checkerboard broth microdilution testing. These preliminary results will be confirmed and expanded by using in vivo models of infection. Thus, this novel approach to antifungal therapy warrants further investigation.
We thank Tara Schroder for excellent secretarial support.
This work was supported in part by a research grant from MethylGene, Inc.
Published ahead of print on 30 September 2009.