THE EMERGENCE OF HSP90 AS AN ANTIFUNGAL TARGET

The appreciation of Hsp90's role as a potentiator of antifungal drug resistance emerged following seminal studies in plants and animals, which established the profound impact of this chaperone on the translation of genotype to phenotype. Pioneering work in Drosophila demonstrated that Hsp90 functions as a capacitor such that it buffers the expression of genetic variation, maintaining it in a phenotypically silent state until Hsp90 function is compromised, such as by environmental stress (Rutherford and Lindquist, Reference Rutherford and Lindquist1998; Rutherford, Reference Rutherford2003). Stress can overwhelm Hsp90 function by causing an increased burden of misfolded cellular proteins, thereby providing an environmentally contingent mechanism to reveal genetic variation. In this context, inhibition of Hsp90 exposes new traits that are contingent upon cryptic genetic variation. Selection can render these traits independent of Hsp90 function, providing a mechanism by which new traits can be assimilated (Sangster et al. Reference Sangster, Lindquist and Queitsch2004). Subsequent work established that Hsp90 also buffers epigenetic variation (Sollars et al. Reference Sollars, Lu, Xiao, Wang, Garfinkel and Ruden2003; Tariq et al. Reference Tariq, Nussbaumer, Chen, Beisel and Paro2009), and that it functions as a capacitor in plants (Queitsch et al. Reference Queitsch, Sangster and Lindquist2002; Sangster et al. Reference Sangster, Salathia, Lee, Watanabe, Schellenberg, Morneau, Wang, Undurraga, Queitsch and Lindquist2008a , Reference Sangster, Salathia, Undurraga, Milo, Schellenberg, Lindquist and Queitsch b ). Compromising Hsp90 function can also modulate RNA silencing mechanisms leading to transposon activation and the generation of novel variants (Specchia et al. Reference Specchia, Piacentini, Tritto, Fanti, D'Alessandro, Palumbo, Pimpinelli and Bozzetti2010).

The scope and breadth of Hsp90's effects on phenotypic variation is stunning, as it can also have the inverse impact from capacitance in the context of acting as a potentiator. In this case, the impact of Hsp90's role in stabilizing mutant cellular regulators is to enable the emergence of new traits; compromise of Hsp90 function can mask the phenotypic effects of new mutations (Cowen, Reference Cowen2008). Classic precedent for Hsp90's role as a potentiator comes from mammalian cancer cells. For many cellular regulators, the acquisition of mutations that cause promiscuous activity and activate their oncogenic potential also renders them unstable and prone to misfolding. For example, oncogenic mutations in the cellular Src tyrosine kinase (c-Src) often involve truncation of the inhibitory domain, rendering the mutant v-Src protein hyperactive and contingent upon Hsp90 for stability and function (Xu and Lindquist, Reference Xu and Lindquist1993; Xu et al. Reference Xu, Singer and Lindquist1999; Falsone et al. Reference Falsone, Leptihn, Osterauer, Haslbeck and Buchner2004). In this context, inhibition of Hsp90 can reverse the oncogenic phenotypes that result from diverse genetic alterations.

Hsp90's multifaceted roles in modulating the translation of genotype to phenotype in an environmentally contingent manner motivated studies to explore the impact of Hsp90 on evolutionary process in a tractable fungal system, the model yeast Saccharomyces cerevisiae. This work revealed that compromise of Hsp90 function blocks the rapid evolution of resistance to azole antifungal drugs, and abrogates resistance acquired by diverse mechanisms (Cowen and Lindquist, Reference Cowen and Lindquist2005). Hsp90 enables the phenotypic effects of resistance mutations by stabilizing a key regulator of cellular stress responses (Fig. 2). Hsp90 is poised to influence diverse aspects of fungal biology and disease given that Hsp90 modulates the phenotypic effects of approximately 20% of pre-existing genetic variation in S. cerevisiae, where it functions as a capacitor and potentiator in almost equal frequency (Jarosz and Lindquist, Reference Jarosz and Lindquist2010).

Fig. 2.

Hsp90 enables the phenotypic effects of resistance mutations. (A) Under normal conditions, fungal cells contain ergosterol in their cell membranes and stress responses are not required; (B) Treatment with the azole fluconazole blocks ergosterol synthesis and leads to incorporation of a toxic sterol in the membrane, culminating in cell wall stress. Hsp90 stabilizes key regulators of cellular stress response (shown in green), enabling signal transduction pathways required for the emergence and maintenance of drug resistance; (C) Stress response pathways are blocked by Hsp90 inhibitors such as geldanamycin, leading to cell death. This prevents the evolution of drug resistance and abrogates resistance once it has evolved. Adapted by permission from Macmillan Publishers Ltd.: Nature Publishing Group (Cowen, Reference Cowen2008), © 2008.

THE IMPACT OF HSP90 ON DRUG RESISTANCE OF FUNGAL PATHOGENS

Hsp90's role in enabling the emergence and maintenance of azole resistance is conserved in fungal pathogens. Hsp90's impact on drug resistance has been studied most extensively in Candida species, which are the fourth most frequent cause of hospital-acquired infections, imposing an economic burden on the healthcare system of over $1 billion annually in the USA alone (Miller et al. Reference Miller, Hajjeh and Edwards2001). Candida albicans is the most prevalent cause of systemic candidiasis, with Candida glabrata emerging as a major threat due at least in part to its intrinsic resistance to azoles; together with Candida parapsilosis, Candida tropicalis and Candida krusei, these species account for over 90% of all cases of candidiasis (Pfaller and Diekema, Reference Pfaller and Diekema2007, Reference Pfaller and Diekema2010). Pharmacological inhibition of Hsp90 blocks the rapid emergence of azole resistance in C. albicans, consistent with findings from S. cerevisiae (Cowen and Lindquist, Reference Cowen and Lindquist2005; Cowen et al. Reference Cowen, Carpenter, Matangkasombut, Fink and Lindquist2006). Genetic or pharmacological compromise of Hsp90 function also transforms azoles from fungistatic to fungicidal (Cowen et al. Reference Cowen, Singh, Kohler, Collins, Zaas, Schell, Aziz, Mylonakis, Perfect, Whitesell and Lindquist2009), and abrogates azole resistance that evolved in a human host by multiple mechanisms (Fig. 3) (Cowen and Lindquist, Reference Cowen and Lindquist2005). Inhibition of Hsp90 function with molecules that are in clinical development for cancer, such as 17-allylamino-17-demethoxygeldanamycin (17-AAG) transforms azoles from ineffective to highly efficacious in rescuing otherwise lethal C. albicans infections in the tractable invertebrate infection model, Galleria mellonella (Cowen et al. Reference Cowen, Singh, Kohler, Collins, Zaas, Schell, Aziz, Mylonakis, Perfect, Whitesell and Lindquist2009). Furthermore, genetic reduction of C. albicans Hsp90 levels enhances the efficacy of azoles in a murine model of systemic candidiasis (Cowen et al. Reference Cowen, Singh, Kohler, Collins, Zaas, Schell, Aziz, Mylonakis, Perfect, Whitesell and Lindquist2009). These findings provide proof-of-principle for Hsp90 as an attractive target for combination therapy to treat fungal infections.

Fig. 3.

Azole resistance of clinical isolates is abrogated by inhibition of Hsp90. Growth in liquid medium with increasing concentrations of the azole fluconazole was measured by absorbance at 600 nm and normalized relative to the no-drug control. Data were quantitatively displayed with colour using Treeview (see colour bar). Clinical isolates (CaCi) recovered from an HIV-infected patient undergoing fluconazole treatment are ordered with those recovered early in treatment at the top and those recovered late at the bottom. All isolates have increased growth compared with the fluconazole susceptible laboratory strain SC5314, with the isolates recovered at the latest stages showing the most robust growth at all concentrations of fluconazole tested. Inhibition of Hsp90 by geldanamycin reduces growth of all clinical isolates in the presence of fluconazole, and affects early isolates to a greater extent than isolates recovered from later stages in treatment. Figure adapted from LaFayette et al. (Reference LaFayette, Collins, Zaas, Schell, Betancourt-Quiroz, Gunatilaka, Perfect and Cowen2010), © LaFayette et al. PLoS Pathogens, 2010.

Hsp90's impact on drug resistance extends beyond azoles to echinocandins. Hsp90 has been found to play a key role in echinocandin resistance in C. albicans and C. glabrata (Singh et al. Reference Singh, Robbins, Zaas, Schell, Perfect and Cowen2009; Singh-Babak et al. Reference Singh-Babak, Babak, Diezmann, Hill, Xie, Chen, Poutanen, Rennie, Heitman and Cowen2012), as well as in the mould Aspergillus fumigatus (Cowen et al. Reference Cowen, Singh, Kohler, Collins, Zaas, Schell, Aziz, Mylonakis, Perfect, Whitesell and Lindquist2009), which causes invasive infections associated with mortality rates ranging from 40–90% (Lin et al. Reference Lin, Schranz and Teutsch2001). Compromising Hsp90 function reduces basal tolerance and resistance acquired in patients undergoing echinocandin treatment (Fig. 4) (Cowen and Lindquist, Reference Cowen and Lindquist2005; Cowen et al. Reference Cowen, Singh, Kohler, Collins, Zaas, Schell, Aziz, Mylonakis, Perfect, Whitesell and Lindquist2009; Singh et al. Reference Singh, Robbins, Zaas, Schell, Perfect and Cowen2009; Singh-Babak et al. Reference Singh-Babak, Babak, Diezmann, Hill, Xie, Chen, Poutanen, Rennie, Heitman and Cowen2012). The combination of Hsp90 inhibitors with echinocandins also abrogates azole resistance of A. fumigatus (Lamoth et al. Reference Lamoth, Juvvadi, Gehrke and Steinbach2012). Beyond the effects in vitro, the combination of Hsp90 inhibitors that are well tolerated in humans and echinocandins improves survival of G. mellonella infected with A. fumigatus, relative to those treated with echinocandins alone (Cowen et al. Reference Cowen, Singh, Kohler, Collins, Zaas, Schell, Aziz, Mylonakis, Perfect, Whitesell and Lindquist2009). Further, genetic reduction of C. albicans Hsp90 levels improves the effectiveness of echinocandins in a mouse model of invasive infection (Singh et al. Reference Singh, Robbins, Zaas, Schell, Perfect and Cowen2009). Together, these studies suggest that Hsp90 can be exploited as a target for combination therapy with the two leading classes of antifungal drugs, to combat the most prevalent and threatening fungal pathogens of humans.

Fig. 4.

Genetic or pharmacological compromise of Hsp90 function reduces basal tolerance and resistance to echinocandins. (A) Inhibition of Hsp90 by geldanamycin (GdA) reduces tolerance to the echinocandin micafungin in two laboratory strains of C. albicans, SC5314 and SN95; genetic compromise of C. albicans HSP90 expression by replacing the native promoter with the tetO promoter abrogates tolerance to micafungin, and complementation with a wild-type HSP90 allele restores tolerance. Data were analysed as in Fig. 3. Adapted from Singh et al. (Reference Singh, Robbins, Zaas, Schell, Perfect and Cowen2009). © Singh et al. PLoS Pathogens, 2009; (B) Hsp90 inhibition reduces resistance to the echinocandin caspofungin of C. glabrata clinical isolates. Isolates are arranged in the order they were recovered from a patient undergoing caspofungin treatment. Isolate A was recovered before treatment began, and isolate G was recovered after numerous rounds of caspofungin treatment. Hsp90 inhibition with GdA reduces the resistance of all of the isolates tested, with the exception of isolate F which is a petite mutant lacking mitochondrial function. Data were analysed as in Fig. 3. Adapted from reference Singh-Babak et al. (Reference Singh-Babak, Babak, Diezmann, Hill, Xie, Chen, Poutanen, Rennie, Heitman and Cowen2012). © Singh-Babak et al. PLoS Pathogens, 2012; (C) Pharmacological inhibition of Hsp90 reduces caspofungin resistance of an A. fumigatus clinical isolate. Antifungal test strips produce a gradient of caspofungin with the highest concentration at the top. A reduction in tolerance is seen when Hsp90 is inhibited through addition of GdA to the plates; comparable results are observed with higher concentrations of the geldanamycin analogue 17-AAG. Adapted from Cowen et al. (Reference Cowen, Singh, Kohler, Collins, Zaas, Schell, Aziz, Mylonakis, Perfect, Whitesell and Lindquist2009). Harnessing Hsp90 function as a powerful, broadly effective therapeutic strategy for fungal infectious disease. Proceedings of the National Academy of Sciences USA 106/8, 2818–23; Copyright (2009), with permission from National Academy of Sciences.

TARGETING HSP90 TO ABROGATE DRUG RESISTANCE OF FUNGAL BIOFILMS

For many fungal pathogens, drug resistance can emerge not only as a result of the acquisition of specific resistance mutations, but also as a consequence of a change in cellular state. Biofilms represent one such cellular state that is associated with intrinsically high levels of antifungal drug resistance (Blankenship and Mitchell, Reference Blankenship and Mitchell2006; d'Enfert, Reference d'Enfert2006). Fungal biofilms are complex communities that form upon adherence to specific surfaces such as medical devices and indwelling catheters. Both C. albicans and A. fumigatus cause prevalent biofilm infections, with C. albicans being the third most frequent cause of intravascular catheter infection and A. fumigatus implicated in biofilm infections on medical devices and bronchial epithelial cells (Nett and Andes, Reference Nett and Andes2006; Ramage et al. Reference Ramage, Mowat, Jones, Williams and Lopez-Ribot2009; Finkel and Mitchell, Reference Finkel and Mitchell2011). For C. albicans, inhibition of Hsp90 function in vitro reduces biofilm formation, abrogates azole resistance, and further blocks the dispersal of biofilm cells, which could otherwise serve as a reservoir for further infection (Robbins et al. Reference Robbins, Uppuluri, Nett, Rajendran, Ramage, Lopez-Ribot, Andes and Cowen2011). Inhibition of Hsp90 also reduces echinocandin resistance of A. fumigatus biofilms in vitro. In a rat central venous catheter model of biofilm infection, azoles alone are ineffective while the combination of genetic or pharmacological compromise of Hsp90 function with azoles sterilizes the catheter without host toxicity (Fig. 5) (Robbins et al. Reference Robbins, Uppuluri, Nett, Rajendran, Ramage, Lopez-Ribot, Andes and Cowen2011). The powerful and broad-spectrum efficacy of Hsp90 inhibitors combined with antifungals suggests that targeting Hsp90 may provide a much-needed strategy to cripple fungal pathogens.

Fig. 5.

Compromise of C. albicans Hsp90 function enhances the efficacy of fluconazole against azole-resistant infections in a rat central venous catheter model of biofilm infection. (A) Genetic reduction of HSP90 levels was achieved using a tetO-HSP90/hsp90Δ strain, in which the only allele of HSP90 is under control of a doxycycline repressible promoter. Scanning electron microscopy images of tetO-HSP90/hsp90Δ biofilms after 24 h of growth in rat venous catheters with or without 20 μg mL⁻¹ doxycycline (DOX) for transcriptional repression of HSP90 expression, followed by treatment with the azole fluconazole (FL) for 24 h. The combination of fluconazole treatment with the genetic depletion of Hsp90 abrogates biofilm growth; (B) Pharmacological inhibition of Hsp90 with 17-allylamino-17-demethoxygeldanamycin (17-AGG) combined with azole treatment sterilizes the rat catheter. 17-AAG and FL were administered after the biofilm had formed; catheter fluid was serially diluted and plated to calculate colony forming units. Asterisk indicates P < 0·001. Adapted from Robbins et al. (Reference Robbins, Uppuluri, Nett, Rajendran, Ramage, Lopez-Ribot, Andes and Cowen2011). © Robbins et al. PLoS Pathogens, 2011.

HSP90 REGULATES KEY FUNGAL VIRULENCE TRAITS

Beyond its impact on fungal drug resistance, Hsp90 is also a key regulator of traits of central importance for fungal virulence. For example, Hsp90 has a profound impact on the capacity of C. albicans to transition between yeast and filamentous growth, which is correlated with virulence (Noble et al. Reference Noble, French, Kohn, Chen and Johnson2010). Filaments are crucial for tissue invasion, escape from macrophages, and expression of a multitude of virulence factors such as adhesins and proteases; yeasts are thought to enable early infection as well as dissemination (Gow et al. Reference Gow, van de Veerdonk, Brown and Netea2012). Filamentation is induced by numerous environmental cues, such as exposure to serum and elevated carbon dioxide, in a manner that is contingent upon elevated temperature (Shapiro et al. Reference Shapiro, Robbins and Cowen2011). The molecular basis for the impact of temperature on morphogenesis remained enigmatic for decades, until Hsp90 was implicated as the key temperature sensor governing this developmental programme. Compromising Hsp90 function induces filamentation under conditions that otherwise favour growth of the yeast form (Fig. 6) (Shapiro et al. Reference Shapiro, Uppuluri, Zaas, Collins, Senn, Perfect, Heitman and Cowen2009, Reference Shapiro, Sellam, Tebbji, Whiteway, Nantel and Cowen2012a ). Elevated temperature compromises Hsp90, thereby relieving the repressive effect on morphogenesis. Consistent with the importance of morphological flexibility for virulence, depletion of C. albicans Hsp90 attenuates virulence in a murine model of systemic infection (Fig. 6) (Shapiro et al. Reference Shapiro, Uppuluri, Zaas, Collins, Senn, Perfect, Heitman and Cowen2009). Genetic studies have also implicated Hsp90 in virulence of C. glabrata in a murine model of disseminated disease (Singh-Babak et al. Reference Singh-Babak, Babak, Diezmann, Hill, Xie, Chen, Poutanen, Rennie, Heitman and Cowen2012), further validating Hsp90 as an attractive therapeutic target to treat fungal infectious disease.

Fig. 6.

Hsp90 regulates morphogenesis and virulence of C. albicans. (A) Pharmacological inhibition of Hsp90 with geldanamycin (GdA) induces filamentous growth under conditions that otherwise favour the yeast growth state. Scale bar represents 10 μm; (B) Genetic depletion of Hsp90 induces filamentation. Genetic repression was achieved by treating the tetO-HSP90/hsp90Δ strain with doxycycline or growing the MAL2p-HSP90/hsp90Δ strain in glucose; (C) Depletion of Hsp90 results in clearance of kidney fungal burden in a murine model of systemic infection. Genetic compromise of C. albicans HSP90 expression simply by replacement of the native promoter with the tetO promoter, reduces kidney fungal burden. Further depletion of Hsp90 with tetracycline sterilizes the kidney. Asterisk indicates P < 0·001. Adapted from Shapiro et al. (Reference Shapiro, Uppuluri, Zaas, Collins, Senn, Perfect, Heitman and Cowen2009). Hsp90 orchestrates temperature-dependent Candida albicans morphogenesis via Ras1-PKA signaling. Current Biology 19/8, 621–629; Copyright (2009), with permission from Elsevier.

CHALLENGES FOR THE DEVELOPMENT OF HSP90 INHIBITORS TO TREAT FUNGAL INFECTIONS

Despite the compelling evidence in support of targeting Hsp90 in fungal pathogens, there are numerous challenges for the development of Hsp90 inhibitors to treat fungal infections. In the context of a mammalian model of fungal biofilm infections, an Hsp90 inhibitor in clinical development for cancer, 17-AAG, was highly effective in combination with an azole without host toxicity problems (Robbins et al. Reference Robbins, Uppuluri, Nett, Rajendran, Ramage, Lopez-Ribot, Andes and Cowen2011). In this case, the drug delivery and fungal infection are both localized in the catheter environment as the result of the blood flow creating a fluid “lock”. In contrast, when the same Hsp90 inhibitor was deployed systemically to treat a disseminated C. albicans infection in a murine model, there was considerable toxicity that precluded observing any therapeutic benefit (Cowen et al. Reference Cowen, Singh, Kohler, Collins, Zaas, Schell, Aziz, Mylonakis, Perfect, Whitesell and Lindquist2009). One would expect that fungal selective Hsp90 inhibitors would ameliorate host toxicity problems (Cowen, Reference Cowen2013).

The development of fungal selective Hsp90 inhibitors may be quite challenging given the extensive conservation of Hsp90. The feasibility of selectively targeting Hsp90 in fungal pathogens based on structural differences compared with the human orthologues can be better assessed once structures are solved for Hsp90 from fungal pathogens. Further, selective targeting of Hsp90 in fungal pathogens may be facilitated by the divergence in ATPase activity or equilibrium of conformational states between human and fungal Hsp90 orthologues (Southworth and Agard, Reference Southworth and Agard2008). In support of this, conformational differences have been exploited in the development of paralogue selective inhibitors of human Hsp90 proteins (Chan et al. Reference Chan, Reeves, Geller, Yaghoubi, Hoehne, Solow-Cordero, Chiosis, Massoud, Paulmurugan and Gambhir2012; Patel et al. Reference Patel, Yan, Seidler, Patel, Sun, Yang, Que, Taldone, Finotti, Stephani, Gewirth and Chiosis2013). Such differences also underpin parasite selective Hsp90 inhibition with molecules that have increased affinity for Trypanosoma brucei Hsp90 compared with the human counterpart (Pizarro et al. Reference Pizarro, Hills, Senisterra, Wernimont, Mackenzie, Norcross, Ferguson, Wyatt, Gilbert and Hui2013). Further analysis of Hsp90 conformational states and biochemical properties in fungal pathogens may illuminate distinct features that can be exploited for antifungal drug development. An appreciation of the upstream regulators of Hsp90 function and effectors that control drug resistance and virulence in fungi is poised to reveal alternative therapeutic strategies to inhibit the Hsp90 chaperone machine.

UP-STREAM REGULATORS OF HSP90 FUNCTION IN FUNGAL PATHOGENS

Our understanding of regulation of Hsp90 function in fungal pathogens has been largely informed by studies in mammalian systems and in the model yeast S. cerevisiae. Hsp90 function is modulated by ATP binding and hydrolysis, by co-chaperones, and by post-translational modifications including phosphorylation, acetylation, methylation and even nitrosylation (Kovacs et al. Reference Kovacs, Murphy, Gaillard, Zhao, Wu, Nicchitta, Yoshida, Toft, Pratt and Yao2005; Martinez-Ruiz et al. Reference Martinez-Ruiz, Villanueva, Gonzalez de Orduna, Lopez-Ferrer, Higueras, Tarin, Rodriguez-Crespo, Vazquez and Lamas2005; Murphy et al. Reference Murphy, Morishima, Kovacs, Yao and Pratt2005; Scroggins et al. Reference Scroggins, Robzyk, Wang, Marcu, Tsutsumi, Beebe, Cotter, Felts, Toft, Karnitz, Rosen and Neckers2007; Mollapour et al. Reference Mollapour, Tsutsumi, Donnelly, Beebe, Tokita, Lee, Lee, Morra, Bourboulia, Scroggins, Colombo, Blagg, Panaretou, Stetler-Stevenson, Trepel, Piper, Prodromou, Pearl and Neckers2010, Reference Mollapour, Tsutsumi, Truman, Xu, Vaughan, Beebe, Konstantinova, Vourganti, Panaretou, Piper, Trepel, Prodromou, Pearl and Neckers2011; Donlin et al. Reference Donlin, Andresen, Just, Rudensky, Pappas, Kruger, Jacobs, Unger, Zieseniss, Dobenecker, Voelkel, Chait, Gregorio, Rottbauer, Tarakhovsky and Linke2012; Xu et al. Reference Xu, Mollapour, Prodromou, Wang, Scroggins, Palchick, Beebe, Siderius, Lee, Couvillon, Trepel, Miyata, Matts and Neckers2012). These post-translational modifications and interactions with co-chaperones can alter Hsp90 conformational states, thereby influencing client protein recognition and chaperone function. Despite conservation of Hsp90, it is notable that the composition of the co-chaperone machinery varies considerably across the eukaryotic kingdom (Johnson and Brown, Reference Johnson and Brown2009). Those regulators of Hsp90 function that have been identified in fungal pathogens are far more divergent between pathogen and host than Hsp90, providing a broader window for the development of fungal selective therapeutic agents.

Regulators of Hsp90 function in fungal pathogens have been explored in most detail in C. albicans. The first analysis of a co-chaperone demonstrated that Sgt1 physically interacts with Hsp90, and that depletion of Sgt1 phenocopies depletion of Hsp90, abrogating drug resistance and inducing filamentation (Shapiro et al. Reference Shapiro, Zaas, Betancourt-Quiroz, Perfect and Cowen2012b ). Sgt1 is far more divergent between fungi and humans than Hsp90, providing an alternative therapeutic target and motivating the study of additional co-chaperones.

Beyond co-chaperones, post-translational modifications have recently been implicated as key for modulating Hsp90 function in C. albicans. As with S. cerevisiae and mammalian cells (Mollapour et al. Reference Mollapour, Tsutsumi, Truman, Xu, Vaughan, Beebe, Konstantinova, Vourganti, Panaretou, Piper, Trepel, Prodromou, Pearl and Neckers2011), Hsp90 function is regulated by phosphorylation mediated by protein kinase CK2 in C. albicans (Fig. 7) (Diezmann et al. Reference Diezmann, Michaut, Shapiro, Bader and Cowen2012). Deletion of CK2 regulatory subunits reduces phosphorylation of Hsp90, as well as the co-chaperone Cdc37, thereby compromising Hsp90 client protein stability and function (Diezmann et al. Reference Diezmann, Michaut, Shapiro, Bader and Cowen2012). Consistent with an impact on Hsp90 function, CK2 has been implicated in regulation of azole resistance (Bruno and Mitchell, Reference Bruno and Mitchell2005). Beyond phosphorylation, acetylation provides an additional post-translation modification with a profound impact on Hsp90 function. Inhibition of lysine deacetylases (KDACs) with the broad-spectrum agent trichostatin A impairs the stability and function of multiple Hsp90 client proteins, and abrogates azole resistance (Fig. 7) (Robbins et al. Reference Robbins, Leach and Cowen2012). The KDACs of functional importance for drug resistance in S. cerevisiae are Hda1 and Rpd3 (Robbins et al. Reference Robbins, Leach and Cowen2012), while the KDACs in C. albicans remain to be identified. Notably, the regulatory subunits of the Hda1 complex are not conserved in metazoans providing opportunities for development of fungal specific inhibitors of the Hsp90 chaperone machinery.

Fig. 7.

Upstream regulators of fungal Hsp90 function. (A) Hsp90 function is regulated by phosphorylation. The stability and activation of kinase clients such as Hog1 are dependent on the phosphorylation of Hsp90 by kinases such as CK2. This is important for key cellular stress responses; (B) Hsp90 function is regulated by acetylation. The acetylation level of Hsp90, balanced by lysine deacetylases (KDACs) and lysine acetyl transferases (KATs), influences its interactions and ability to stabilize client proteins. Inhibition of lysine deacetylases (KDACs) blocks the interaction between Hsp90 and its client protein calcineurin, and alters the stability and function of numerous other clients, thereby blocking key responses to drug-induced cellular stress. Adapted from Cowen (Reference Cowen2013). The fungal Achilles' heel: targeting Hsp90 to cripple fungal pathogens. Current Opinion in Microbiology 16/4, 377–384; Copyright (2013), with permission from Elsevier.

DOWN-STREAM EFFECTORS OF HSP90 IMPORTANT FOR DRUG RESISTANCE AND VIRULENCE

Given Hsp90's pleiotropic effects on cellular signalling, it is not surprising that Hsp90 modulates drug resistance and virulence though multiple downstream effectors. In C. albicans, the first Hsp90 client protein identified was calcineurin (Cowen and Lindquist, Reference Cowen and Lindquist2005; Cowen et al. Reference Cowen, Carpenter, Matangkasombut, Fink and Lindquist2006; Singh et al. Reference Singh, Robbins, Zaas, Schell, Perfect and Cowen2009), a serine threonine protein phosphatase and key regulator of resistance to azoles and echinocandins (Cruz et al. Reference Cruz, Goldstein, Blankenship, Del Poeta, Davis, Cardenas, Perfect, McCusker and Heitman2002; Sanglard et al. Reference Sanglard, Ischer, Marchetti, Entenza and Bille2003; Singh et al. Reference Singh, Robbins, Zaas, Schell, Perfect and Cowen2009). Hsp90 physically interacts with the catalytic subunit of calcineurin in both S. cerevisiae and C. albicans, and depletion of Hsp90 leads to reduction of calcineurin levels (Imai and Yahara, Reference Imai and Yahara2000; Singh et al. Reference Singh, Robbins, Zaas, Schell, Perfect and Cowen2009). Inhibiting calcineurin function with the structurally unrelated immunosuppressants cyclosporin A and FK506 provides a powerful strategy to abrogate drug resistance of diverse fungal pathogens (Steinbach et al. Reference Steinbach, Reedy, Cramer, Perfect and Heitman2007; Lamoth et al. Reference Lamoth, Juvvadi, Gehrke and Steinbach2012). Calcineurin is also required for virulence in species of Candida, Cryptococcus and Aspergillus (Bader et al. Reference Bader, Bodendorfer, Schroppel and Morschhauser2003; Sanglard et al. Reference Sanglard, Ischer, Marchetti, Entenza and Bille2003; Steinbach et al. Reference Steinbach, Cramer, Perfect, Asfaw, Sauer, Najvar, Kirkpatrick, Patterson, Benjamin, Heitman and Perfect2006; Miyazaki et al. Reference Miyazaki, Yamauchi, Inamine, Nagayoshi, Saijo, Izumikawa, Seki, Kakeya, Yamamoto, Yanagihara, Miyazaki and Kohno2010b ; Reedy et al. Reference Reedy, Filler and Heitman2010; Chen et al. Reference Chen, Brand, Morrison, Silao, Bigol, Malbas, Nett, Andes, Solis, Filler, Averette and Heitman2011, Reference Chen, Konieczka, Springer, Bowen, Zhang, Silao, Bungay, Bigol, Nicolas, Abraham, Thompson, Regev and Heitman2012, Reference Chen, Lehman, Lewit, Averette and Heitman2013; Singh-Babak et al. Reference Singh-Babak, Babak, Diezmann, Hill, Xie, Chen, Poutanen, Rennie, Heitman and Cowen2012; Zhang et al. Reference Zhang, Silao, Bigol, Bungay, Nicolas, Heitman and Chen2012; Juvvadi et al. Reference Juvvadi, Gehrke, Fortwendel, Lamoth, Soderblom, Cook, Hast, Asfaw, Moseley, Creamer and Steinbach2013). A central challenge for the development of calcineurin inhibitors for the treatment of fungal infections is the development of analogues that retain antifungal activity in the absence of immunosuppressive effects due to inhibition of calcineurin in the host (Blankenship et al. Reference Blankenship, Steinbach, Perfect and Heitman2003).

Another Hsp90 effector that modulates resistance to both azoles and echinocandins in C. albicans is the terminal mitogen activated protein kinase of the Pkc1 cell wall integrity pathway, Mkc1. Hsp90 stabilizes Mkc1, enabling crucial responses to drug-induced stress (LaFayette et al. Reference LaFayette, Collins, Zaas, Schell, Betancourt-Quiroz, Gunatilaka, Perfect and Cowen2010). While Hsp90 stabilizes both the phosphorylated and unphosphorylated form of Mkc1 in C. albicans (LaFayette et al. Reference LaFayette, Collins, Zaas, Schell, Betancourt-Quiroz, Gunatilaka, Perfect and Cowen2010), Hsp90 interacts exclusively with the phosphorylated form of the orthologue Slt2 in S. cerevisiae (Millson et al. Reference Millson, Truman, King, Prodromou, Pearl and Piper2005). Compromising signalling through this cell wall integrity pathway reduces drug resistance, and attenuates virulence of Candida species (LaFayette et al. Reference LaFayette, Collins, Zaas, Schell, Betancourt-Quiroz, Gunatilaka, Perfect and Cowen2010; Miyazaki et al. Reference Miyazaki, Inamine, Yamauchi, Nagayoshi, Saijo, Izumikawa, Seki, Kakeya, Yamamoto, Yanagihara, Miyazaki and Kohno2010a ). Signalling through Pkc1 also regulates virulence in Cryptococcus neoformans (Gerik et al. Reference Gerik, Donlin, Soto, Banks, Banks, Maligie, Selitrennikoff and Lodge2005). Chemical genomic approaches have recently been successful in the identification of Hsp90 interactors on a global scale in C. albicans, revealing a multitude of additional interactors that include novel regulators of drug resistance and virulence, as well as new therapeutic targets (Diezmann et al. Reference Diezmann, Michaut, Shapiro, Bader and Cowen2012).

CONCLUSIONS AND OUTLOOK

Hsp90 has emerged as a fungal Achilles' heel given that it is a hub of cellular circuitry required for fungal drug resistance, stress response, morphogenesis and virulence (Cowen, Reference Cowen2013). The development of combination therapies with an Hsp90 inhibitor deployed with an antifungal drug may be facilitated by the multitude of Hsp90 inhibitors, including 17 agents currently in clinical development for cancer (Trepel et al. Reference Trepel, Mollapour, Giaccone and Neckers2010; Neckers and Workman, Reference Neckers and Workman2012). Hsp90 inhibitors in clinical development show promise for combination therapy with antifungals in the context of mammalian models where the fungal infection and drug delivery is localized (Robbins et al. Reference Robbins, Uppuluri, Nett, Rajendran, Ramage, Lopez-Ribot, Andes and Cowen2011). To overcome host toxicity issues in the context of disseminated fungal infections and systemic drug delivery, it is likely that fungal selective Hsp90 inhibitors will be required (Cowen et al. Reference Cowen, Singh, Kohler, Collins, Zaas, Schell, Aziz, Mylonakis, Perfect, Whitesell and Lindquist2009). Alternatively, targeting regulators of Hsp90 function that are more divergent between pathogen and host, such as co-chaperones or KDACs, or down-stream effectors required for drug resistance and virulence, such as calcineurin, could provide the foundation of effective combination therapies (Cowen, Reference Cowen2013). Further development of inhibitors of the Hsp90 chaperone network as combination therapeutic agents to treat fungal infections will be facilitated through collaborative efforts between industry and academia. This may provide a broader paradigm for treatment of infectious disease caused by other eukaryotic pathogens, such as protozoan parasites that are the causal agents of malaria and trypanosomiasis (Pallavi et al. Reference Pallavi, Roy, Nageshan, Talukdar, Pavithra, Reddy, Venketesh, Kumar, Gupta, Singh, Yadav and Tatu2010; Shahinas et al. Reference Shahinas, Liang, Datti and Pillai2010, Reference Shahinas, Macmullin, Benedict, Crandall and Pillai2012, Reference Shahinas, Folefoc, Taldone, Chiosis, Crandall and Pillai2013). An appreciation of the importance of combination therapeutic strategies, which are the foundation for the treatment of HIV (Bock and Lengauer, Reference Bock and Lengauer2012), tuberculosis (Zumla et al. Reference Zumla, Hafner, Lienhardt, Hoelscher and Nunn2012) and malaria (Eastman and Fidock, Reference Eastman and Fidock2009), promises to dramatically accelerate the discovery of novel treatment strategies to minimize the global health burden of fungal infectious disease.