Novel antifungal drugs are badly needed to combat the increasing clinical challenge posed by fungal pathogens. We utilized a S. cerevisiae reporter to create a high throughput bioassay to identify novel antifungal compounds small molecule libraries. Disk diffusion assays of initial hits revealed compounds 13 and 33 as two antifungal drug leads with potent and potentially broad spectrum activity against two major fungal pathogens A. fumigatus and C. albicans. In further testing using broth microdilution assays, compounds 13 and 33 yielded MICs against these two fungi, other molds and C. neoformans with similar potency to currently used antifungals such as fluconazole and amphotericin B. Importantly, both compounds showed fungicidal activity against mold and yeast.
While compounds 13 and 33 exert potent, broad-spectrum activity, they do not appear to act directly against the group III HHKs – the mode of action that was sought in this screen. Why did our screen not identify compounds that directly act on group III HHKs? There may be several explanations. First, we screened only a limited number of small molecules (~20,000). Though sufficient for a preliminary screen, large-scale screening efforts often include hundreds of thousands of compounds. Second, S. cerevisiae has
only one HHK, Sln1, while A. nidulans
encodes 11 HHKs 
. As a result, HHKs redundant in function to NikA could complement its deletion explaining retained compound activity against the ΔNikA strain. Third, it is possible that Hik1 overexpression “reprograms” S. cerevisiae
to become sensitive to compounds 13 and 33. The activity of compounds 13 and 33 therefore requires Hik1 in S. cerevisiae
, but the compounds do not directly target Hik1.
Microarray analysis revealed that compound 13 induced the up-regulation of genes associated with an oxidative stress response. The GSEA network analysis also showed a concordant increase in gene families associated with an oxidative stress response. Oxidative stress is known to damage DNA, and we found that DNA repair-deficient fungi were much more sensitive to compound 13 than were wild type strains. Together, these data argue that compound 13 acts in a manner dependent on the induction of an oxidative stress response.
exposed to compound 33 induced transcripts associated with the stress response to heavy metals. We also found that pre-exposing C. albicans
to heavy metal stress, by initial growth in cadmium, increased its sensitivity to compound 33. Although this result supports possible involvement of heavy metal stress in the mode of action of compound 33, the finding was unanticipated since prior work showed that pre-exposure to a given stress engenders resistance 
. How do we explain this discrepancy? Vylokova et al.
examined osmotic stress while we studied heavy metal stress 
. The different stressors could explain the discrepancy. It is also possible that residual cadmium associated with C. albicans
after pretreatment resulted in the exposure of yeast simultaneously to heavy metal and compound 33. Similarly, Vylokova et al.
observed that concurrent incubation of C. albicans
with osmotic stress together with the osmotic stress-inducing peptide histatin 5 significantly enhanced its sensitivity.
The enhanced sensitivity of DNA repair-deficient fungi to compound 13 is notable from a drug development standpoint, since it suggests the compound may damage DNA. Nevertheless, there are FDA approved antibiotics that damage DNA as part their mode of action. Nitrofurantoin, a commonly used antimicrobial, damages DNA as part of its action 
. Compound 13 contains a nitro group structurally, like in nitrofurantoin, where it is reduced to generate oxidative stress 
. Although compound 33 also contains nitro groups, it did not induce oxidative stress response genes, nor did compound 33 have increased activity against DNA repair-deficient fungi. Thus, the nitro group of compound 13, but not 33, is likely reduced leading to oxidative stress and DNA damage, a possible toxicity concern.
Toxicity is a concern with any potential therapeutic. Highly toxic compounds should have been eliminated during the small molecule secondary screen, when compounds that inhibited the growth of wild type S. cerevisiae were removed. We also used a standard cell membrane lysis assay to assess compound toxicity. Amphotericin B lysed RBCs at low concentrations, whereas compounds 13 or 33 had little effect even at the high concentrations of 300 µg/ml. There are many different types of cell and animal toxicity assays and we limited our initial analysis in vitro to this standard assay, but other cell-based testing may be desirable. We also found that while compounds 13 and 33 bound serum protein substantially, making systemic administration challenging, mice tolerated substantial doses with no overt toxicity (data not shown). Thus, from the in vitro studies in wild-type S. cerevisiae, and the studies with human RBCs, and in mice, we conclude that compounds 13 and 33 are not general cytotoxins.
Fungal biofilms represent one of the most challenging types of fungal infections facing patients and physicians. Compounds 13 and 33 had robust activity against C. albicans biofilm. Compounds 13 and 33 also were highly active in vivo in a rat denture model of C. albicans biofilm infection. Whereas fluconazole alone had little activity against in vitro biofilm, compounds 13 and 33 synergized significantly with fluconazole. Fluconazole-induced oxidative stress may have enhanced its synergy with compounds 13 and 33.
The azole antifungals inhibit lanosterol 14 α–demethylase, a crucial enzyme in the biosynthesis of ergosterol 
, but recent studies have suggested that the generation of oxidative stress may also be involved in their mechanism of action. For example, the addition of free radical scavengers rendered C. albicans
resistant to the azole antifungal miconazole 
. Furthermore, incubation with sub-inhibitory concentrations of fluconazole caused C. albicans
to upregulate oxidative stress response genes and made the organism more resistant to killing by phagocytes 
. Exposure to sub-inhibitory concentrations of fluconazole had a similar affect on C. neoformans
. Oxidative stress has also been postulated to be involved in the synergy of compounds with the azoles. For example, the synergistic activity against C. albicans
of fluconazole with polyphenol curcumin I, a plant-derived antifungal, was abolished by the addition of an antioxidant 
. Thus, the mechanism of synergy between fluconazole and compounds 13 and 33 likely involved oxidative stress.
In summary, by employing a simple high-throughput S. cerevisiae bioassay, we discovered two compounds - 13 and 33 - with potent fungicidal activity against yeast and molds across multiple genera that frequently infect human patients. The compounds also work against fungal biofilms and act in synergy with conventional antifungal drugs. Compounds 13 and 33 therefore represent potentially valuable antifungal drug leads.