Peptide-Like Nylon-3 Polymers with Activity against Phylogenetically Diverse, Intrinsically Drug-Resistant Pathogenic Fungi

Fungi reside in all ecosystems on earth and impart both positive and negative effects on human, plant, and animal health. Fungal disease is on the rise worldwide, and there is a critical need for more effective and less toxic antifungal agents. Nylon-3 polymers are short, sequence random, poly-β-amino acid materials that can be designed to manifest antimicrobial properties. Here, we describe three nylon-3 polymers with potent activity against the most phylogenetically diverse set of fungi evaluated thus far in a single study. In contrast to traditional peptides, nylon-3 polymers are highly stable to proteolytic degradation and can be produced efficiently in large quantities at low cost. The ability to modify nylon-3 polymer composition easily creates an opportunity to tailor efficacy and toxicity, which makes these materials attractive as potential broad-spectrum antifungal therapeutics.

T he pervasiveness of fungi and their impact on globally important processes can be seen in the evolution of the human immune system (1), the chemistry of soil (2), the genomes of plants (3,4), and even the progression of climate change (5). Fungi interact extensively with plants, animals, bacteria, and other organisms; these relationships range from mutualists to saprotrophs to pathogens. The influence of fungi across of variety of biological processes has its roots in the massive degree of diversity within the kingdom of fungi. Fungi are characterized by varied morphologies, diverse growth strategies, and assorted nutrient acquisition capabilities. Though only roughly 100,000 species of fungi are accepted in the current taxonomy (6), it is estimated that there exist on the order of 1.5 million fungal species and that up to one-quarter of the world's biomass is of fungal origin (7,8). Our imperfect understanding of fungal diversity arises from the lack of taxonomically relevant morphological features and pleomorphisms of fungi, the difficulty or impossibility of culturing many fungi isolated from natural environments, and the inability to consistently identify diagnostic sexual structures in fungi grown in culture or observed in the environment. The great ecological success of fungi, either as filamentous or unicellular growth forms, is a testament to their diversity and functional importance in the global ecosystem.
Understanding the dimensions of fungal diversity has major implications for disease control, crop management, and decomposition of both recalcitrant organic and synthetic materials. For environmental, agricultural, and biomedical purposes, we need agents that target fungi broadly. The innate immune response to infection by fungi and other microbes includes production of host defense peptides (HDPs), and these molecules offer clues for development of synthetic antifungal agents. HDPs have diverse sequences. Many of these peptides are relatively short (10 to 50 amino acid residues) and adopt globally amphiphilic conformations, although the nature of these conformations is variable (␣-helix versus ␤-sheet versus irregular) (9). HDP amphiphilicity is hypothesized to be critical for disruption of microbial membranes, which results in either microbial killing or inhibition of growth (other antimicrobial mechanisms of action have been proposed as well) (10,11).
Since their discovery in the 1980s, HDPs and their synthetic analogs have been subjects of interest as potential therapeutic agents. The high cost of production and inherent toxicity to the host manifested by HDPs, however, have discouraged clinical development of these peptides thus far (9). The production challenge has encouraged evaluation of random copolymers with hydrophilic and lipophilic subunits for antimicrobial activity (12)(13)(14)(15)(16)(17)(18). Random copolymers can be synthesized in a more economical manner than can sequence-specific peptides.
Nylon-3 polymers have been studied as antibacterial agents since 2007, and more recently, examples with both antibacterial and antifungal activities have been reported (19)(20)(21)(22)(23). Nylon-3 polymers are synthesized via the anionic ring-opening polymerization of ␤-lactams. Use of racemic ␤-lactams leads to heterochiral polymer chains. The protein-like polyamide backbone of nylon-3 polymers is believed to confer biocompatibility, while the presence of ␤-amino acid subunits renders these materials resistant to proteolytic degradation (24). Three nylon-3 polymers have previously been shown to have activity against pathogenic fungi (20,22). These polymers, designated MM-TM, DM-TM, and NM ( Fig. 1), inhibit the growth of Candida albicans and Cryptococcus neoformans, pathogenic fungi with yeast morphology, as single agents. In contrast, these polymers were able to halt the growth of the pathogenic, filamentous Aspergillus fumigatus only when applied in combination with established azole-based antifungal agents; the polymers were not active as single agents. The activity of MM-TM, DM-TM, and NM against pathogenic yeasts led us to hypothesize that these polymers might be active against a wide range of unicellular yeasts, but not against filamentous fungi. The work described here presents a test of this hypothesis. We have undertaken, to our knowledge, the broadest phylogenetically susceptibility analysis of any antifungal compound to determine the extent of activity of the MM-TM, DM-TM, and NM polymers. We expanded our evaluation to include a diverse array of both filamentous and nonfilamentous species across the fungal kingdom, including members of the Zygomycetes, Ascomycetes, and Basidiomycetes phyla. Surprisingly, most fungi tested, including those naturally resistant to current antifungal drugs, were sensitive to the nylon-3 polymers, which raises the possibility that nylon-3 polymers could be useful against pathogens for which there are only limited and/or no antifungal agents available at present.

RESULTS
NM, MM-TM, and DM-TM are active at low concentrations against phylogenetically diverse yeasts. To evaluate the antifungal activity of MM-TM, DM-TM, and NM against six species of vegetatively growing yeasts across three different genera, we used the CLSI M27-A3 broth microdilution method (Table 1) (25). Polymers were evaluated by measuring MICs (MIC 100 s). For comparison, we also evaluated the commonly used antifungal drug fluconazole (FLC) in terms of MIC 50 (the MIC to halt 50% of growth per the CLSI M27-A3 standard). All of the yeasts tested were sensitive to all three of the nylon-3 polymers, including those strains resistant to azoles. The polymers were particularly effective against Cryptococcus spp. (Cryptococcus neoformans and Cryptococcus amylolentus) of the Basidiomycota phylum, exhibiting MIC 100 values from 2 to 4 g/ml, which is comparable to or better than concentrations of fluconazole required to halt only 50% of fungal growth. MM-TM, DM-TM, and NM were also active against two genera within the Ascomycota phylum: Candida and Saccharomyces. Saccharomyces cerevisiae was particularly sensitive to all three nylon-3 polymers, with NM  being the most active (MIC ϭ 2 g/ml) and DM-TM the least active (MIC ϭ 8 g/ml). Candida spp. were also susceptible to the polymers, with MIC 100 values for Candida albicans and Candida krusei generally ranging from 4 to 16 g/ml. The polymers were also active against Candida auris. Invasive and drug-resistant C. auris infections have been increasingly reported in health care facilities around the world (26,27). The two strains of C. auris tested both showed sensitivity (MIC 100 s of 4 to 16 g/ml) to all three polymers, with particular sensitivity to NM. Overall, yeasts were highly susceptible to the three nylon-3 polymers tested for efficacy, with NM being the most active polymer against yeasts. The Aspergillus genus is generally insensitive to nylon-3 polymers, NM, MM-TM, and DM-TM. Previous reports show that nylon-3 polymers are largely inactive against the pathogenic fungus Aspergillus fumigatus (22). To determine whether this resistance phenotype was specific to A. fumigatus, we assessed the antifungal activity of MM-TM, DM-TM, and NM against 18 different species within the Aspergillus genus. For 6 of the 18 Aspergillus species examined, none of the three polymers caused any decrease in hyphal growth relative to a no-treatment control ( Table 2, boldface rows, rows 1 to 7). A. fumigatus, A. flavus, and A. terreus are considered to be the most pathogenic species of the Aspergillus genus (28). It seems possible that the high pathogenicity of these three species is related to their relatively high resistance to the nylon-3 polymers tested. The remaining 12 species of Aspergillus evaluated showed low levels of growth inhibition, with MIC 100 s ranging from 8 to Ͼ64 g/ml, with the DM-TM copolymer demonstrating the most promising antifungal activity among the three nylon-3 polymers. While all 18 species of Aspergillus were tested according to CLSI M38-A2 methodology (conidia in liquid culture), four species, selected for their varied response to polymer, were also tested for susceptibility to polymer as hyphal clumps in liquid culture. MIC 100 values against hyphal fragments were determined as the concentration of agent required to prevent hyphal growth, monitored as an increase in optical density at 600 nm (OD 600 ), after incubation at 35°C for 48 h. Three of these species, A. fumigatus, A. terreus, and A. flavus, were highly resistant to polymer as conidia, and their hyphal fragments were also insensitive to polymer. The conidia of the fourth species of Aspergillus selected, A. nidulans, were susceptible to polymer, with MIC 100 values ranging from 8 to 16 g/ml; however, polymer activity was diminished when incubated with A. nidulans hyphal clumps (MIC 100 values of 32 to Ͼ64 g/ml). As expected from precedent, the efficacy of itraconazole against Aspergillus species as hyphal clumps was also greatly diminished relative to treatment of conidia (MIC 100,conidia~1 g/ml; MIC 100,hyphae~8 to Ͼ64 g/ml) (29). Overall, nylon-3 polymers are ineffective against species within the Aspergillus genus, regardless of the life form assayed (conidia or hyphae) for sensitivity. NM, MM-TM, and DM-TM are active against phylogenetically diverse filamentous fungi. On the basis of our observation of low sensitivity of Aspergillus spp. toward the polymers as single agents, we hypothesized that nylon-3 polymers were intrinsically less effective against filamentous fungi relative to yeasts. To test this hypothesis, we assessed MM-TM, DM-TM, and NM activity against 10 genera of filamentous fungi, following the CLSI M38-A2 methodology (conidia in liquid culture) or a modified CLSI method (hyphal fragments in liquid culture). Surprisingly, MM-TM, DM-TM, and NM were very active against phylogenetically diverse filamentous fungi, with MIC 100 values of 4 to 8 g/ml ( Table 3). The polymers were much more effective against medically relevant Rhizopus arrhizus (a causative agent of mucormycosis), the emerging pathogen Paecilomyces variotii (30), and Fusarium oxysporum isolates (which, along with A. flavus, is a serious agent of keratitis [31]) than they were against Aspergillus spp. Notably, all three nylon-3 polymers were active against Scedosporium spp., both Scedosporium apiospermum and Scedosporium prolificans. The latter species is an emerging fungal pathogen of both immunocompetent and immunocompromised individuals that is intrinsically resistant to most antifungal drugs (voriconazole [VOR] MIC 100 Ͼ 16 g/ml); S. prolificans infections are often fatal (32).
For the dermatophytes Trichophyton tonsurans, Trichophyton rubrum, and Microsporum canis, MIC 100 values were determined as the concentration of the agent required to halt hyphal growth, monitored as an increase in OD 600 , after 4 days of incubation at 29°C in RPMI 1640. Mixed efficacy of the nylon-3 polymers was observed across the three species of dermatophytes tested. The highest levels of antifungal activity were observed against T. tonsurans and M. canis (8 to 16 g/ml), and the lowest levels were observed against T. rubrum (16 to 64 g/ml) ( Table 3). The MIC 100 of the dermatophyte Pseudogymnoascus destructans was evaluated after 7 days of incubation of conidia at 12°C in RPMI 1640. The activity against P. destructans, the causative agent of white nose syndrome in bats (33), is encouraging because there are limited options for preventing the spread of this devastating pathogen. The facility with which nylon-3 structure can be varied could provide opportunities for the development of topical agents with high specificity for particular fungi.
Strikingly, nylon-3 polymers were ineffective against Filobasidiella depauperata (MIC 100 Ͼ 64 g/ml [ Table 3]), which is closely related to C. neoformans and grows exclusively as filaments (34). Resistance of F. depauperata to nylon-3 polymers was surprising because of the close phylogenetic relationship between C. neoformans and F. depauperata and the consistent sensitivity of C. neoformans isolates. F. depauperata spores and hyphae were both strongly resistant to polymer in SD medium (synthetically defined, minimal medium) (35); this species of fungi did not grow in the RPMI 1640 growth medium specified by CLSI methods. C. neoformans JEC20x21 spores and yeasts were both sensitive to all three polymers in SD medium (see Tables S1 and S2 in the supplemental material). As this level of inactivity against germinating conidia was reminiscent of the results observed with Aspergillus species, we asked whether the nylon-3 polymer MM-TM could act synergistically with an azole to inhibit germination of F. depauperata conidia, as had previously been observed with various A. fumigatus isolates. The copolymer MM-TM exhibited strong synergy (⌺FIC [fractional inhibitory concentration] index value of 0.04) with itraconazole against the azole-sensitive strain CBS 7855, resulting in Ͼ20-fold decreases in azole and polymer MIC 100 s when the polymers are given in combination ( Table 4). The MM-TM copolymer was previously shown to act synergistically with azole-sensitive strains of A. fumigatus (22). Overall, all three nylon-3 polymers demonstrated strong activity against filamentous fungi spanning the fungal tree of life, with the notable exceptions of F. depauperata and the Aspergillus genus.
NM, MM-TM, and DM-TM are active against dimorphic fungi. Given the sensitivity of yeasts and most filamentous fungi to NM, MM-TM, and DM-TM, we hypothesized that dimorphic fungi would be sensitive to nylon-3 polymers. The CLSI M38-A2 broth macrodilution method was used to test all three polymers against Coccidioides, Blastomyces, and Histoplasma germinating conidia (36). Each of these dimorphic fungi represents a significant health threat for humans, and the dimorphism of these fungi is an important feature for their pathogenicity (37,38). Activity was assessed based on 80% inhibition of growth (MIC 80 per the CLSI M38-A2 standard); the positive control for these studies was voriconazole (Table 5). MM-TM, DM-TM, and NM have MIC 80 values of Ͻ4 g/ml for all strains of Histoplasma capsulatum, Blastomyces dermatitidis, and Coccidioides spp. assayed. NM, MM-TM, and DM-TM have moderate antifungal activity against Pneumocystis spp. Pneumocystis spp. are known to cause lethal pneumonia in immunocompromised patients and are particularly associated with AIDS patients. Relatively little is known about Pneumocystis spp., ranging from its life cycle to survival strategies within mammalian hosts, since there currently exists no way to culture these species in a laboratory setting. Antifungal activity of the three polymers was assessed against cryopreserved and characterized Pneumocystis spp. Fungal viability in the presence of each polymer was measured using an ATP production assay (ATP-liteM assay) after 24, 48, and 72 h of polymer exposure. Calculating percent ATP reduction for all samples allowed us to determine 50% inhibitory concentrations (IC 50 s) ( Table 6 and Tables S3  and S4). MM-TM and DM-TM exhibited moderate activity, with 72 h IC 50 values ranging from 2 to 5 g/ml. The homopolymer NM was less active against Pneumocystis murina compared to Pneumocystis carinii, with 72-h IC 50 values of 15 and 3 g/ml, respectively. These data illustrate that even noncanonical fungi with unusual adaptations and niches (Pneumocystis species were once thought to be protozoan parasites [39]) show sensitivity to nylon-3 polymers, expanding opportunities for the use of nylon-3 materials against diverse fungi.

DISCUSSION
Our data show that the nylon-3 polymers MM-TM, DM-TM, and NM are effective against a surprisingly broad spectrum of fungi (Fig. 2), with only low to moderate toxicity toward mammalian cells (see Text S1 and Tables S5 and S6 in supplemental All three polymers showed roughly equivalent activity against highly diverse fungi across the fungal kingdom. Any strain for which polymer activity was not equivalent across all three polymers is indicated by the pound symbol. Any species assessed in this study in which a different strain was used to compose the phylogeny tree is indicated by an asterisk. The strain tested in a previous publication (11) is indicated by a caret. N/A, not available. material). Here we were able to assess sensitivity of 18 pathogenic genera toward the nylon-3 chemotype, based on measurements with 41 species and 72 isolates. Visualization of our results in the context of fungal phylogeny indicates general sensitivity to the polymers, with notable exceptions in the Aspergillus clade and the basidiomycete Filobasidiella depauperata (Fig. 2). Overall, these data provide support for the idea that nylon-3 polymers could be useful as broad-action therapeutic agents.
Activity of the nylon-3 polymers was observed against several species of fungi with limited or ineffective treatment options. For instance, the nylon-3 polymers were active against Rhizopus arrhizus, one of the causative agents of mucormycosis, which is a life-threatening disease in both immunocompetent and immunocompromised patients (40). Depending on the predisease status of the patient and route of infection, mucormycosis may present in pulmonary, rhino-orbital-cerebral, cutaneous, gastrointestinal, or disseminated forms. Treatment of mucormycosis often requires administration of amphotericin B after surgical debridement of necrotic tissues (40). Even with rigorous treatment regimes, mortality rates are high (Ͼ40%), and amphotericin B toxicity is problematic for patients. The sensitivity of R. arrhizus to nylon-3 polymers suggests a new strategy to combat a challenging and deadly fungal disease for which current treatment options are highly limited (40).
All three nylon-3 polymers also showed efficacy against dermatophytes as preestablished hyphae. It is estimated that 20 to 25% of the world's population has skin mycoses (41), and while these infections are often superficial and cosmetic in nature, they can have severe impact on quality of life. Topical treatment of dermatophyte infections relies mostly on two classes of antifungal drugs, the azoles (e.g., fluconazole and ketoconazole) and the allylamines (e.g., terbinafine and naftifine). However, there have been increasing incidences of resistant and refractory dermatophyte infections in the clinic, and daily dosages of terbinafine can reach 1,000 mg (42,43). Nylon-3 polymers, with their ease of synthesis and low anticipated cost of production, could provide a viable option for treating such topical fungal infections.
Comparisons between sensitive and resistant genera may ultimately enable us to identify the origin(s) of sensitivity and to elucidate mechanisms of nylon-3 action. A particular opportunity to understand how nylon-3 polymers confer antifungal activity emerges from the observation that these materials were ineffective against Filobasidiella depauperata, despite the polymers' strong activity against closely related Cryptococcus species. Since the polymers were active against other filamentous fungi, it is likely that properties not associated with filamentous morphology are responsible for this polymer resistance phenotype. Genetic comparisons of C. neoformans and F. depauperata may reveal insights into the mechanisms by which nylon-3 polymers exert their antifungal effects.
While all three nylon-3 polymers studied, NM, MM-TM, and DM-TM, had similar antifungal activities, subtle differences in activity profiles were noted throughout our phylogenetically broad analysis. NM was the most active polymer against the ascomycete yeasts Candida and Saccharomyces, while DM-TM was the most potent polymer against filamentous fungi. Overall hydrophobicity represents one of the main chemical properties that varied among the polymers (NM Ͻ MM-TM Ͻ DM-TM). All subunits in NM bear a short cationic side chain. Therefore, this homopolymer therefore cannot display a large and well-defined hydrophobic surface, which is believed to be an important feature of host defense peptides (HDPs) and their synthetic mimics. The MM-TM and DM-TM copolymers, on the other hand, should manifest greater overall hydrophobicity relative to NM because there are one or two additional nonpolar CH 2 units in the MM and DM cationic subunits relative to NM, and each of the copolymers contains 20 to 30% of the entirely hydrophobic TM subunit. The copolymer expected to be most hydrophobic, DM-TM, was the most active among the three against filamentous fungi; however, this enhanced activity comes at the cost of increased mammalian cell toxicity relative to MM-TM or NM (Text S1 and Tables S5 and S6). The subtle differences in antifungal activity across the three polymers suggest that increased overall hydrophobicity may be required to enhance activity against filamentous fungi, particularly of the Aspergillus clade. Polymers with increased hydrophobicity relative to those discussed here, however, may require additional design features to avoid enhanced toxicity toward eukaryotic cells. Here we present the activities of materials with a novel chemotype, nylon-3 polymers, against a varied array of fungi from across the fungal kingdom. The surprisingly broad spectrum of nylon-3 antifungal activities, including inhibition of difficult-to-treat human pathogens, offers promise for the development of polymeric compounds with therapeutically useful properties. The ease of synthesis and structural diversification of nylon-3 polymers provide a broad scope for future efforts to optimize activity against pathogenic fungi while limiting toxicity toward the host.

MATERIALS AND METHODS
Polymer synthesis and characterization. All polymers and monomers were prepared using previously reported methods (20)(21)(22)(23). Please see the supplemental material for more information about polymer synthesis and characterization methods (Text S2 and Table S7).
CLSI M27-A3 protocol. The MIC 100 endpoint of each antifungal agent was determined as the lowest concentration to inhibit 100% of fungal growth compared to the no-drug control. MIC 100 values for all yeasts assayed (Table 7) were determined by the broth microdilution method according to the CLSI M27-A3 guidelines, with slight modifications (25). Briefly, fungal cells at a density of 0.5 ϫ 10 3 to 2.5 ϫ 10 3 cells/ml were incubated at 30 to 35°C in RPMI 1640 plus 0.145 M 3-(N-morpholino)propanesulfonic acid (MOPS) (pH 7.0) in 96-well plates with twofold serial dilutions of nylon-3 polymer or fluconazole (FLC) from 1 to 64 g/ml. After 24 to 72 h, the optical density at 600 nm (OD 600 ) of each well was measured using a microplate reader. Wells containing fungal cells with no drug and wells containing only RPMI 1640 were used as positive and blank controls, respectively. Percent cell growth was determined as [(sample absorbance Ϫ blank absorbance)/(control absorbance Ϫ blank absorbance)] ϫ 100%. All values reported represent the average MIC 100 value for more than two biological replicates and two or more technical replicates each. The average MIC 100 value consistently fell within a twofold serial dilution of the concentration of each experimental replicate. Any modifications to this protocol for a specific species is listed in Table 7.
Revised CLSI M38-A protocol for hyphae of filamentous fungi and dermatophytes. The MIC 100 endpoint of each antifungal agent was determined as the lowest concentration to inhibit 100% of hyphal growth. MIC 100 values for all filamentous fungi assayed were determined by the broth microdilution method according to CLSI M38-A guidelines, with modifications (44).
For the dermatophytes, crushed hyphal fragments of Trichophyton tonsurans, Trichophyton rubrum, and Microsporum canis at a density corresponding to an OD 600 reading of 0.05 were incubated at 29°C in RPMI 1640 plus 0.145 M MOPS, pH 7.0, in 96-well plates with twofold serial dilutions of nylon-3 polymer or with itraconazole (ITRA) from 1 to 64 g/ml. After 96 h, the MIC of each well was measured by monitoring the changes in OD 600 compared to the value at the 0-h time point. From the OD measurements, a difference in OD greater than 0.03 from the values at the 96-and 0-h time points was indicative of growth. All values reported represent the average MIC 100 value for two biological replicates and six technical replicates each. The average MICs consistently fell within a twofold serial dilution of the concentration of each experimental replicate.
Pseudogymnoascus destructans at a density of 1 ϫ 10 4 spores/ml was incubated at 12°C in RPMI 1640 plus 0.145 M MOPS, pH 7.0, in 96-well plates with twofold serial dilutions of nylon-3 polymer or with ITRA from 1 to 64 g/ml. After 96 h, the MIC of each well was measured by monitoring changes in OD 600 compared to the value at the 0-h time point. From the OD measurements, a difference in OD greater than 0.01 from the values at the 120-and 0-h time points was indicative of growth. All values reported represent the average MIC 100 value for two biological replicates and six technical replicates each. The average MICs consistently fell within a twofold serial dilution of the concentration of each experimental replicate.
A total of 1 ϫ 10 4 conidia/ml of Aspergillus fumigatus, Aspergillus terreus, Aspergillus flavus, or Aspergillus nidulans were inoculated into RPMI 1640 plus 0.145 M MOPS, pH 7.0 in 96-well plates and incubated at 37°C for 18 to 24 h to allow for growth of the extensive hyphal network. Hyphae were then treated with twofold serial dilutions of nylon-3 polymer or with ITRA from 1 to 64 g/ml and allowed to incubate at 37°C. OD 600 readings were obtained at 0, 24, and 48 h. From the OD measurements, a difference in OD greater than 0.05 from the values at the 48-and 0-h time points was indicative of growth. All values reported represent the average MIC 100 value for two biological replicates and six technical replicates each. The average MICs consistently fell within a twofold serial dilution of the concentration of each experimental replicate.
A total of 1 ϫ 10 4 spores/ml of Penicillium expansum conidia were inoculated into RPMI 1640 plus 0.145 M MOPS, pH 7.0, in 96-well plates and incubated at room temperature for 24 to 48 h to allow for germination into established hyphae. Hyphae were then treated with twofold serial dilutions of nylon-3 polymer or with ITRA from 1 to 64 g/ml and allowed to incubate further at room temperature. OD 600 readings were obtained at 0, 24, and 48 h.    (Table 7) were determined by the broth macrodilution method according to the CLSI M38-A2 guidelines, with slight modifications (36). Briefly, fungal cells at a density of 0.5 ϫ 10 4 to 5 ϫ 10 4 conidia/ml were incubated at 35°C in RPMI 1640 plus 0.145 M MOPS, pH 7.0, in 96-well plates with twofold serial dilutions of the nylon-3 polymer or with fluconazole (FLC), voriconazole (VOR), or itraconazole (ITRA) from 1 to 64 g/ml. After 24 to 72 h, the MIC of each well was measured visually using a dissecting microscope. All values reported represent the average MIC 100 values for more than two biological replicates and two or more technical replicates each. The average MICs consistently fell within a twofold serial dilution of the concentration of each experimental replicate. Any modifications to this protocol for a specific species are listed in Table 7. ATP-liteM assay for Pneumocystis spp. Cryopreserved and characterized Pneumocystis carinii isolated from rat lung tissue and Pneumocystis murina isolated from mouse lung tissue were distributed into triplicate wells of 48-well plates with a final volume of 500 l at a final concentration of 5 ϫ 10 7 nuclei/ml P. carinii and 5 ϫ 10 6 nuclei/ml P. murina. Controls and compounds were added and incubated at 36°C and 5% CO 2 . At 24, 48, and 72 h, 10% of the well volume was removed, and the ATP content was measured using PerkinElmer ATP-liteM luciferin-luciferase assay. The luminescence generated by the ATP content of the samples was measured by a BMG PolarStar optima spectrophotometer. A sample of each group was examined microscopically on the final assay day to rule out the presence of bacteria.
For 50% inhibitory concentration (IC 50 ) calculations, background luminescence was subtracted, and triplicate well readings of duplicate assays were averaged. For each day's reading, percent reduction in ATP for all groups was calculated as follows: [(medium control Ϫ experimental value)/medium control] ϫ 100. IC 50 values were calculated using GraphPad Prism 6 linear regression program (Tables S3 and S4).
Evolutionary analysis. Multigene-based phylogeny between all fungal species depicted in Fig. 2 was constructed based on amino acid sequences of 14 genes (Table S8) identified in Fusarium spp. to be conserved for various energetic processes (45). Orthologous proteins for genes listed in Table S8 were identified in each species using HMMer (Ensembl Fungi). Ten genes from Wiemann et al. (45) were omitted from the analysis due to lack of orthologous protein identity in one or more of the species tested. Resulting protein sequences for each of the gene were aligned to well-conserved regions using MAFFT (46) and trimmed using Gblocks (47) using default parameters. All positions containing gaps and missing data were eliminated. Trimmed alignments were then concatenated for each species, and a maximum likelihood (ML) phylogeny was inferred based on the JTT model (48) using MEGA7 (49) and viewed in FigTree (http://tree.bio.ed.ac.uk/software/figtree/). Data availability. All data are included in the article and supplemental files.