Clotrimazole as a Potent Agent for Treating the Oomycete Fish Pathogen Saprolegnia parasitica through Inhibition of Sterol 14α-Demethylase (CYP51)

Abstract

A candidate CYP51 gene encoding sterol 14α-demethylase from the fish oomycete pathogen Saprolegnia parasitica (SpCYP51) was identified based on conserved CYP51 residues among CYPs in the genome. It was heterologously expressed in Escherichia coli, purified, and characterized. Lanosterol, eburicol, and obtusifoliol bound to purified SpCYP51 with similar binding affinities (K_s, 3 to 5 μM). Eight pharmaceutical and six agricultural azole antifungal agents bound tightly to SpCYP51, with posaconazole displaying the highest apparent affinity (K_d, ≤3 nM) and prothioconazole-desthio the lowest (K_d, ∼51 nM). The efficaciousness of azole antifungals as SpCYP51 inhibitors was confirmed by 50% inhibitory concentrations (IC₅₀s) of 0.17 to 2.27 μM using CYP51 reconstitution assays. However, most azole antifungal agents were less effective at inhibiting S. parasitica, Saprolegnia diclina, and Saprolegnia ferax growth. Epoxiconazole, fluconazole, itraconazole, and posaconazole failed to inhibit Saprolegnia growth (MIC₁₀₀, >256 μg ml⁻¹). The remaining azoles inhibited Saprolegnia growth only at elevated concentrations (MIC₁₀₀ [the lowest antifungal concentration at which growth remained completely inhibited after 72 h at 20°C], 16 to 64 μg ml⁻¹) with the exception of clotrimazole, which was as potent as malachite green (MIC₁₀₀, ∼1 μg ml⁻¹). Sterol profiles of azole-treated Saprolegnia species confirmed that endogenous CYP51 enzymes were being inhibited with the accumulation of lanosterol in the sterol fraction. The effectiveness of clotrimazole against SpCYP51 activity (IC₅₀, ∼1 μM) and the concentration inhibiting the growth of Saprolegnia species in vitro (MIC₁₀₀, ∼1 to 2 μg ml⁻¹) suggest that clotrimazole could be used against Saprolegnia infections, including as a preventative measure by pretreatment of fish eggs, and for freshwater-farmed fish as well as in leisure activities.

INTRODUCTION

Worldwide growth of aquaculture (at ∼6% per annum from 2006 to 2011) is reflective of increased consumption of fish, especially in China (1). Aquaculture now accounts for more than 40% of the total fish production for human consumption, with freshwater aquaculture contributing ∼70% in 2011 compared to ∼30% from marine aquaculture (1). Fish that are diseased as a result of bacterial and oomycete (water mold) infections are the largest cause of economic loss in aquaculture (2). Saprolegnia species are responsible for most oomycete infections in farmed fish (3), with Saprolegnia parasitica being endemic to all fresh water environments around the world. Losses in worldwide salmon aquaculture through disease run into tens of millions of pounds per annum (4). S. parasitica has also been implicated in the worldwide decline in wild salmon populations (5), and Saprolegnia ferax is thought to be partially responsible for declining numbers of amphibians in natural ecosystems (6, 7).

Saprolegnia is an opportunistic pathogen that is saprotrophic and necrotrophic (3), although some S. parasitica strains are very virulent and can cause primary infections (8). Saprolegnia has a fairly broad temperature tolerance (3 to 33°C) (9), with sudden changes in water temperature making fish vulnerable to infection by Saprolegnia (3, 10). Saprolegnia infection can be difficult to eradicate, especially in freshwater hatcheries and fish farms, in part due to the ability of Saprolegnia species to form biofilm communities on its own and with other microorganisms (11), which are more resistant to treatment with antibiotic compounds and act as reservoirs of infection. Infections can affect both eggs and fish. Salmonid eggs are particularly vulnerable to Saprolegnia infection during the several months spent on freshwater riverbeds prior to hatching (12). Initial infections of ova result in dead eggs, from which a mycelial mat spreads over time to engulf neighboring live eggs (12), resulting in further losses and the potential infection of hatchlings. Czeczuga et al. (13) found that Saprolegnia species were present on the external surfaces of all sea trout eggs sampled from the rivers and freshwater hatcheries of Poland. On fish, Saprolegnia invades epidermal tissues, often beginning on the head or fins and eventually spreading over the whole body surface (10), and is able to cause cellular necrosis as well as dermal and epidermal damage (14). However, Saprolegnia infections do not appear to be tissue specific. If untreated, Saprolegnia infection leads to death by osmoregulatory failure (9, 14, 15). Until 2002, Saprolegnia infections were well controlled by using the organic dye malachite green, which is considered the most effective chemical against Saprolegnia (3, 16). However, the use of malachite green in aquaculture was banned by the United States and several other countries (17) due to concerns that the dye is a potential carcinogen (18) and a mutagen (3). Since 2002, infections caused by Saprolegnia species have re-emerged, with Saprolegnia parasitica in particular being an economically important pathogen of farmed fish, such as salmon, trout, and catfish (19). Improving water quality and reducing stress and handling can significantly reduce the incidence and severity of Saprolegnia infection in catfish (20).

Presently there are few licensed chemicals for controlling Saprolegnia infection in salmonid eggs and no chemicals that give sufficient protection against Saprolegnia infection after hatching (21). S. parasitica is inhibited by low concentrations of sodium chloride (22), with the sodium chloride helping to counteract osmotic stress in affected fish with damaged skin. Sodium chloride at high concentrations such as in seawater (∼29 g/liter) and salt water (∼15 g/liter) is lethal to Saprolegnia (17, 23) and is effective for controlling S. parasitica (10). However, use of sodium chloride in freshwater environments is not practical. Ozone treatment of water has also been used to reduce S. parasitica infection (21), although ozone cannot be used to cure infected fish. Formalin has been used to treat eggs and the first stages of larval development (3, 18, 24, 25). However, formaldehyde is harmful to the environment and personal health. Hydrogen peroxide is a promising chemical for the treatment of Saprolegnia (17, 18, 26) with minimal impact on the environment and has been used as a treatment of catfish eggs (20). Cupric sulfate and diquat (0.125 ppm) have been used as prophylactic treatments to inhibit zoospores (20). Recently, alternate chemical approaches have been tried to control S. parasitica infection with varying success, including chemically modified chitosans (27), peracetic acid (28), saprolmycins (29), oridamycins (30), Pyceze (2-bromo-2-nitropropane-1,3-diol) (12), and essential oils and ethanol extracts of medicinal plants (31). In addition, chitin synthases responsible for tip growth in Saprolegnia have been identified as potential targets for anti-oomycete drugs (32). Effective strategies and chemicals still need to be developed to control or eradicate Saprolegnia infections in aquaculture, especially in freshwater environments, with at least the same efficacy as malachite green.

Previously, oomycetes, such as Phytophthora species, were considered sterol auxotrophs utilizing sterols from the surrounding environment or the host organism (33) and lacking functional CYP51 genes (34). This explained the ineffectiveness of azole antifungal agents against many oomycete species, as the azole target enzyme CYP51 was absent. However, the recent discovery of a sterol metabolism pathway in the oomycete Aphanomyces euteiches (a legume root pathogen) (35), including a CYP51 gene, led us to investigate whether such a pathway was also present in S. parasitica. Previous investigations into the sterol profile of Saprolegnia were limited to S. ferax (36, 37).

In this study, we identified the CYP51 gene from the genome sequence of S. parasitica and cloned, expressed, and purified the recombinant CYP51 protein in an active state. We characterized the effectiveness of 15 azole antifungal agents used in the clinic (pharmazoles) and in agriculture (agriazoles) at inhibiting S. parasitica CYP51 activity in vitro and growth of Saprolegnia species. Finally, we discuss the potential of using azole antifungal agents to control Saprolegnia infections in freshwater aquaculture.

MATERIALS AND METHODS

Construction of the pCWori⁺:SpCYP51 expression vector.

The S. parasitica CYP51 gene (SpCYP51; BROAD Institute accession number SPRG_09493.2 [http://www.broadinstitute.org/annotation/genome/Saprolegnia_parasitica/MultiHome.html]) was identified by performing a BLASTP search against the S. parasitica genome database using the amino acid sequence of Candida albicans CYP51 (UniProtKB accession number P10613). The SpCYP51 gene was synthesized by Eurofins MWG Operon (Ebersberg, Germany) incorporating an NdeI restriction site at the 5′ end and a HindIII restriction site at the 3′ end of the gene cloned into pUC57. In addition, the first eight amino acids were changed to MALLLAVF (38), and a six-histidine extension (CATCACCATCACCATCAC) was inserted immediately before the stop codon to facilitate protein purification by Ni²⁺-nitriloacetic acid (NTA) agarose affinity chromatography. The SpCYP51 gene was excised by NdeI/HindIII restriction digestion followed by cloning into the pCWori⁺ expression vector using Roche T4 DNA ligase. Gene integrity was confirmed by DNA sequencing.

Heterologous expression in Escherichia coli and isolation of recombinant SpCYP51 protein.

The pCWori⁺:SpCYP51 construct was transformed into competent DH5α E. coli cells, and transformants were selected using 0.1 mg ml⁻¹ ampicillin. Growth and expression conditions were identical to those previously reported (39). Protein isolation was according to the method of Arase et al. (40), except that 2% sodium cholate was used in the sonication buffer. The solubilized SpCYP51 protein was purified by Ni²⁺-NTA agarose affinity chromatography as previously described (41) using 1% l-histidine in 0.1 M Tris-HCl (pH 8.1) and 25% glycerol to elute SpCYP51, followed by dialysis against 5 liters of 25 mM Tris-HCl (pH 8.1) and 10% glycerol. Ni²⁺-NTA agarose-purified SpCYP51 was used for all subsequent spectral and IC₅₀ (50% inhibitory concentration) determinations. Protein purity was assessed by SDS polyacrylamide gel electrophoresis followed by staining with Coomassie brilliant blue R-250.

Determination of cytochrome P450 protein concentrations.

Cytochrome P450 concentration was determined by reduced carbon monoxide difference spectra (42), with carbon monoxide passed through the cytochrome P450 solution prior to addition of sodium dithionite to the sample cuvette, using an extinction coefficient of 91 mM⁻¹ cm⁻¹ (43) for the absorbance difference between 447 and 490 nm (ΔA_447−490). Absolute spectra (700 to 300 nm) were determined using 2 μM SpCYP51 in 0.1 M Tris-HCl (pH 8.1) and 25% glycerol as previously described (41). Confirmation that isolated SpCYP51 was active was demonstrated by measuring the 14α-demethylation of lanosterol, eburicol, and obtusifoliol using the CYP51 reconstitution assay detailed below.

CYP51 reconstitution assay system.

IC₅₀ determinations were performed using the CYP51 reconstitution assay system (500-μl final reaction volume) previously described (44) containing 0.5 μM SpCYP51, 1 μM Aspergillus fumigatus cytochrome P450 reductase (AfCPR1 [UniProtKB accession number Q4WM67]), 50 μM eburicol, 50 μM dilaurylphosphatidylcholine, 4% (wt/vol) (2-hydroxypropyl)-β-cyclodextrin (HPCD), 0.4 mg ml⁻¹ isocitrate dehydrogenase, 25 mM trisodium isocitrate, 50 mM NaCl, 5 mM MgCl₂, and 40 mM MOPS (pH ∼7.2). Azole antifungal agents were added in 2.5 μl dimethylformamide followed by 10 min incubation at 30°C prior to assay initiation with 4 mM β-NADPH-Na₄. Samples were then shaken for 30 min at 30°C. In addition, trial SpCYP51 assays were also performed using 50 μM lanosterol and 50 μM obtusifoliol as substrates. Sterol metabolites were recovered by extraction with ethyl acetate followed by derivatization with N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA) and tetramethylsilane (TMS) in pyridine prior to analysis by gas chromatography-mass spectrometry (GC-MS) (45). Eburicol was chosen as the substrate for the CYP51-azole IC₅₀ determinations, as the TMS-derivatized 14-demethylated product could be readily separated from TMS-derivatized eburicol during gas chromatography. IC₅₀ in this study is defined as the inhibitor concentration required causing 50% inhibition of CYP51 activity under the stated assay conditions.

Sterol binding properties of SpCYP51.

Stock 2.5 mM solutions of lanosterol (the CYP51 substrate in animals and most yeasts), eburicol (the CYP51 substrate in higher fungi), and obtusifoliol (the CYP51 substrate in plants) were prepared in 40% (wt/vol) HPCD. Sterol was progressively titrated against 4 μM SpCYP51 protein in a quartz semimicrocuvette (light path, 4.5 mm) with equivalent amounts of 40% (wt/vol) HPCD added to the reference cuvette also containing 4 μM SpCYP51. The absorbance difference spectrum between 500 and 350 nm was determined after each incremental addition of sterol (up to 75 μM). Sterol saturation curves were constructed from ΔA_385−421, derived from the difference spectra. The substrate binding constants (K_s) were determined by nonlinear regression (Levenberg-Marquardt algorithm) using the Michaelis-Menten equation. Each binding determination was performed in triplicate.

Azole binding studies.

Binding of azole antifungal agents to 2 μM SpCYP51 was performed as previously described (46) using quartz split cuvettes with a 4.5-mm light path. Stock 1-, 0.5-, 0.2-, and 0.1-mg ml⁻¹ solutions of the pharmaceutical azole antifungals (pharmazoles) clotrimazole, econazole, fluconazole, itraconazole, ketoconazole, miconazole, posaconazole, and voriconazole and the agricultural azole antifungals (agriazoles) epoxiconazole, prochloraz, propiconazole, prothioconazole, prothioconazole-desthio, tebuconazole, and triadimenol were prepared in dimethylformamide. Azole antifungals were progressively titrated against 2 μM SpCYP51 in 0.1 M Tris-HCl (pH 8.1) and 25% glycerol at 22°C, with equivalent volumes of dimethylformamide also being added to the SpCYP51-containing compartment of the reference cuvette. The absorbance difference spectra between 500 and 350 nm were determined after each incremental addition of azole with binding saturation curves constructed from ΔA_428−412 against azole concentration. The dissociation constant of the enzyme-azole complex (K_d) for each azole was determined by nonlinear regression (Levenberg-Marquardt algorithm) using a rearrangement of the Morrison equation for tight ligand binding (47, 48). K_d values determined using the Morrison equation are accurate to ∼1/100 of the protein concentration (49). Each binding determination was performed in triplicate. The chemical structures of the azole antifungals used in this study are shown in Fig. 1.

FIG 1.

Chemical structures of azole antifungal agents. The chemical structures of clotrimazole (molecular weight [MW], 345), econazole (MW, 445 as nitrate salt), fluconazole (MW, 306), itraconazole (MW, 706), ketoconazole (MW, 531), miconazole (MW, 479 as nitrate salt), posaconazole (MW, 701), voriconazole (MW, 349), epoxiconazole (MW, 330), prochloraz (MW, 377), propiconazole (MW, 342), prothioconazole (MW, 344), prothioconazole-desthio (MW, 312), tebuconazole (MW, 308), and triadimenol (MW, 296) are shown. Also shown are the triazole and imidazole ring numbering systems, using the voriconazole and econazole structures as examples.

MIC studies with Saprolegnia species.

Saprolegnia parasitica Coker (CBS 223 63), Saprolegnia diclina Humphrey (CBS 326 35), and Saprolegnia ferax Gruith (CBS 173 42) were retrieved from the CBS-KNAW Fungal Biodiversity Centre (Netherlands) and maintained on 90-mm-diameter yeast mold agar plates (ATCC YM medium 200) containing yeast extract (3%), malt extract (3%), glucose (1%), peptone (0.5%), and agar (2%). Agar plates were routinely inoculated with a single 5-mm plug of mycelium cut from stock cultures exhibiting active radial growth. All strains were grown for 3 days at 20°C (±2°C) prior to use. Azole susceptibility assays were performed in 24-well (flat-bottom, 16-mm-diameter round wells) culture plates (3820-024; Iwaki) using an adaptation of the Clinical and Laboratory Standards Institute (CLSI) M27-A2 broth dilution method. Briefly, all azoles and malachite green were dissolved in dimethyl sulfoxide to a stock concentration of 25.6 mg ml⁻¹. Dilutions were then made with dimethyl sulfoxide to achieve 100× stock solutions (1-ml volumes) of 25.6, 12.8, 6.4, 3.2, 1.6, 0.8, 0.4, 0.2, 0.1, 0.05, and 0.025 mg ml⁻¹. These stocks were initially diluted 10-fold using YM medium prior to addition of 100 μl directly into culture plate wells containing 900 μl YM to achieve final azole concentrations of 256, 128, 64, 32, 16, 8, 4, 2, 1, 0.5, and 0.25 μg ml⁻¹ and control wells containing 1% (vol/vol) dimethyl sulfoxide. Culture wells were inoculated centrally with a single 3-mm plug of mycelium excised from YM agar plates and incubated for 3 days at 20°C (±2°C). MICs were determined in duplicate and scored manually. MIC₁₀₀ is defined here as the lowest antifungal concentration at which growth remained completely inhibited after 72 h at 20°C.

Sterol composition of Saprolegnia species.

Mycelial cakes were removed from the plate wells of treated (compound concentration immediately below MIC₁₀₀) and untreated Saprolegnia species and then washed three times with sterile water. Nonsaponifiable lipids were extracted as previously reported (50) and were derivatization with N,O-bis(trimethylsilyl)trifluoroacetamide and tetramethylsilane in pyridine prior to analysis by gas chromatography-mass spectrometry (45). Sterol composition was calculated using peak areas from the gas chromatograms, and the mass fragmentation patterns were used to confirm sterol identity.

Data analysis.

Curve fitting of ligand binding data was performed using the computer program ProFit 6.1.12 (QuantumSoft, Zurich, Switzerland). Spectral determinations were made using quartz semimicrocuvettes with a Hitachi U-3310 UV/VIS spectrophotometer (Hitachi, San Jose, CA).

The N-terminal membrane anchor region of SpCYP51 was predicted using TargetP version 1.1 (http://www.cbs.dtu.dk/services/TargetP/) software. Subcellular location was predicted using WoLF PSORT (http://www.genscript.com/psort/wolf_psort.html) software. Phylogenetic analyses were performed by comparing the SpCYP51 amino acid sequence (SPRG_09493.2) against selected fungal, plant, and animal CYP51 proteins from the UniProtKB database (http://www.uniprot.org/help/uniprotkb) using NCBI BLAST2 (http://blast.ncbi.nlm.nih.gov/Blast.cgi?PAGE_TYPE=BlastSearch&PROG_DEF=blastn&BLAST_PROG_DEF=megaBlast&BLAST_SPEC=blast2seq) and ClustalX version 1.81 (http://www.clustal.org/) sequence alignment software with a phylogenetic tree produced from the ClustalX generated Phylip-dnd file using TreeViewX (https://code.google.com/p/treeviewx/) software.

The eukaryotic CYP51 sequences used for phylogenetic comparison are detailed in Table S1 in the supplemental material. BLASTP searches of the Broad Institute S. parasitica genome database (http://www.broadinstitute.org/annotation/genome/Saprolegnia_parasitica/MultiHome.html) were performed using sterol biosynthesis protein homologs from A. euteiches (35), Aspergillus fumigatus (Δ⁵ sterol desaturase [Q4WDL3] and Δ³ sterol dehydrogenase [Q4X017]), and Homo sapiens (Δ⁸ sterol isomerase [Q15125]). Several sequences for Δ³ sterol keto reductase (fungal and mammalian) were used. All S. parasitica protein sequences so identified were then used as query sequences at NCBI BLASTP (http://blast.ncbi.nlm.nih.gov/Blast.cgi?PROGRAM=blastp&PAGE_TYPE=BlastSearch&LINK_LOC=blasthome) to elucidate the most likely sterol biosynthesis candidate.

Chemicals.

All chemicals, including azole antifungals except voriconazole, were obtained from Sigma Chemical Company (Poole, United Kingdom). Voriconazole was supplied by Discovery Fine Chemicals (Bournemouth, United Kingdom). Growth media, sodium ampicillin, isopropyl-β-d-thiogalactopyranoside (IPTG), and 5-aminolevulenic acid were obtained from Foremedium Ltd. (Hunstanton, United Kingdom). Ni²⁺-NTA agarose affinity chromatography matrix was obtained from Qiagen (Crawley, United Kingdom).

RESULTS

Analysis of SpCYP51 protein sequence.

The SpCYP51 sequence we identified agreed with that published (SPRG_09493.2) by Jiang et al. (51) during the course of our work. TargetP predicted the SpCYP51 N-terminal membrane anchor to consist of the first 36 amino acid residues, while WoLF PSORT predicted that SpCYP51 would be located in the endoplasmic reticulum. This agrees with the observation that eukaryotic CYP51 proteins are associated with the membranes of the endoplasmic reticulum (52). Alignment of SpCYP51 against other eukaryotic CYP51 proteins confirmed that all 23 conserved amino acid residues previously identified for CYP51 proteins were present (53). The resultant phylogenetic tree (Fig. 2) showed that SpCYP51 clustered with the CYP51 of fellow oomycete A. euteiches, sharing 79% sequence identity. S. parasitica CYP51 was more closely related to plant and animal CYP51 proteins than to fungal CYP51 proteins. For example, SpCYP51 shared 48% sequence identity with CYP51 of the marine diatom Thalassiosira pseudonana and 45% identity with that of the thermoacidophilic unicellular red alga Galdieria sulfuraria. SpCYP51 shared 41 to 44% sequence identity with CYP51 enzymes from higher plants (44% with cucumber CYP51) and 36 to 38% identity with animal CYP51 proteins (37% identity with human CYP51), in contrast to just 30 to 35% sequence identity with the “true” fungal CYP51 proteins (34% identity with C. albicans CYP51 and 31% identity with A. fumigatus CYP51A).

FIG 2.

Phylogenetic tree of eukaryotic CYP51 enzymes. A phylogenetic tree of selected eukaryotic CYP51 proteins, including members from the fungus, plant, and animal kingdoms in addition to trypanosomal and oomycete CYP51 proteins, was constructed. The individual CYP51 sequences used to construct the phylogenetic tree are detailed in Table S1 in the supplemental material. Species names have been left nonitalicized for clarity. A, CYP51A; B, CYP51B.

Expression and purification of SpCYP51.

The yield of SpCYP51 was ∼120 (±20) nmol per liter of E. coli culture, as determined by absolute spectroscopy (41) after purification by Ni²⁺-NTA agarose chromatography. This yield was ∼10-fold greater than that calculated by carbon monoxide difference spectroscopy under reductive conditions (42) of the crude sodium cholate cell extract, due to the reduced CO-SpCYP51 adduct formed being unstable. SDS-polyacrylamide gel electrophoresis confirmed the purity of the Ni²⁺-NTA agarose-eluted SpCYP51 protein to be greater than 95% when assessed by Coomassie brilliant blue R-250 staining intensity, with an apparent molecular weight of ∼52,000 ± 2,000, which was close to the predicted value of 55,487, including the six-histidine C-terminal extension.

Spectral properties of recombinant SpCYP51 protein.

The absolute spectrum of SpCYP51 (Fig. 3A) was typical for a ferric cytochrome P450 enzyme predominantly (∼80%) in the low-spin state (41, 54) with α, β, Soret (γ), and δ spectral bands at 565, 535, 417, and 360 nm, respectively. Reduced carbon monoxide difference spectra (Fig. 3B) produced the characteristic red-shifted Soret peak at 445 to 447 nm typical of ferrous cytochrome P450 enzymes complexed with CO (42, 43). However, the reduced CO adduct at 447 nm was unstable, with the absorbance peak quickly dissipating (half-life [t_1/2] = 3.2 ± 0.2 min), along with an accompanying increase in the “inactive” P-420 complex absorbance at 421 nm.

FIG 3.

Spectral properties of SpCYP51. (A) Absolute oxidized absorption spectrum between 700 and 300 nm for 2 μM SpCYP51. (B) Reduced carbon monoxide difference spectra between 480 and 410 nm using 2 μM SpCYP51 at sequential 45-s intervals. Arrows indicate the progressive fall in A₄₄₇ and rise in A₄₂₁ with time. Matched quartz semimicrocuvettes with a light path of 10 mm were used.

Binding studies with sterols.

All three 14α-methylsterols produced strong type I binding spectra with SpCYP51 (Fig. 4A), which was typical of the interaction of substrates with cytochromes P450 (54), with a peak at ∼385 nm and trough at ∼420 nm. Sterol saturation curves were constructed (Fig. 4B) to derive substrate binding constants (K_s). SpCYP51 had similar affinities for lanosterol, eburicol, and obtusifoliol at 2.8 ± 0.4, 3.3 ± 0.5, and 4.6 ± 0.2 μM, respectively. All three sterols were 14α-demethylated in the CYP51 reconstitution assay using purified SpCYP51 protein and AfCPR1 as the redox partner, with turnover numbers of 0.63, 1.04, and 2.55 min⁻¹ for lanosterol, eburicol, and obtusifoliol, respectively. This confirmed that the putative CYP51 was indeed as predicted and that SpCYP51 had been isolated in a fully functional form.

FIG 4.

Sterol binding properties of SpCYP51. (A) Type I difference spectra were obtained by progressive titration of lanosterol, eburicol, and obtusifoliol against 4 μM SpCYP51 using 4.5-mm-light-path quartz semimicrocuvettes. (B) Sterol saturation curves were constructed for lanosterol (filled circles), eburicol (open circles), and obtusifoliol (bullets). All spectral determinations were performed in triplicate, although values for only one replicate are shown.

Azole antifungal agent binding studies.

Tight type II binding spectra (peak at 428 nm and trough at 412 nm) were observed between 2 µM SpCYP51 and all 15 azole antifungal agents (Fig. 5 and 6) with the exception of prothioconazole. Type II binding spectra are caused by the triazole N-4 (Fig. 1, voriconazole) or the imidazole N-3 (Fig. 1, econazole), coordinating as the sixth ligand with the heme iron (55) to form the low-spin CYP51-azole complex, resulting in a red shift of the heme Soret peak. No reproducible binding spectra between prothioconazole and 2 μM SpCYP51 could be obtained (data not shown), probably because of steric hindrance from the sulfur atom preventing the direct coordination of the triazole N-4 atom with the heme ferric ion (Fig. 1). Tight binding is normally observed for azole antifungals against fungal CYP51 proteins, where the K_d for a ligand is similar to or lower than the concentration of the enzyme present (49). The imidazole-based pharmazoles gave approximately 30% more intense (greater ΔA_max) difference spectra with SpCYP51 than the triazole-based pharmazoles (Fig. 5), whereas the agriazoles (all triazoles with the exception of the imidazole prochloraz) gave type II difference spectra (Fig. 6) of similar intensity to the triazole-based pharmazoles. Azole-SpCYP51 saturation curves (Fig. 7) indicated that azole binding affinities were similar, with K_d values typically between 3 and 16 nM (Table 1), with the exception of clotrimazole, fluconazole, and prothioconazole-desthio, which bound slightly less tightly, with K_d values of 25, 40, and 51 nM, respectively. These marginally higher K_d values may result from the relative compactness of these three azoles preventing additional stabilizing interactions with amino acid side chains that line the SpCYP51 heme binding pocket. This would need to be verified by in silico ligand docking experiments. Therefore, SpCYP51-catalyzed sterol demethylation should be strongly inhibited by azole antifungal agents.

FIG 5.

Pharmazole binding to SpCYP51. Type II difference spectra were obtained by progressive titration of 2 μM SpCYP51 with individual pharmazoles. Matched quartz semimicrocuvettes with a light path of 4.5 mm were used, and all spectral determinations were performed in triplicate, although data from only one replicate are shown.

FIG 6.

Agriazole binding to SpCYP51. Type II difference spectra were obtained by progressive titration of 2 μM SpCYP51 with individual agriazoles. No reproducible difference spectra were obtained between prothioconazole and 2 μM SpCYP51 (data not shown). Matched quartz semimicrocuvettes with a light path of 4.5 mm were used, and all spectral determinations were performed in triplicate, although data from only one replicate are shown.

FIG 7.

Azole saturation curves for SpCYP51. Saturation curves were constructed for the imidazole-based pharmazoles (A), triazole-based pharmazoles (B), and agriazoles (C) from the change in absorbance (ΔA_428−412) against azole concentration using a rearrangement of the Morrison equation (47) for the tight ligand binding observed with SpCYP51 (Fig. 5 and 6). (A) The imidazole-based pharmazoles used were clotrimazole (filled circles), econazole (open circles), ketoconazole (bullets), and miconazole (crosses). (B) The triazole-based pharmazoles used were fluconazole (filled circles), itraconazole (open circles), posaconazole (bullets), and voriconazole (crosses). (C) The agriazoles used were epoxiconazole (filled circles), prochloraz (open circles), propiconazole (bullets), prothioconazole-desthio (crosses), tebuconazole (asterisks), and triadimenol (triangles). All determinations were performed in triplicate, although data for only one replicate are shown.

TABLE 1.

Affinity of SpCYP51 for azole antifungal agents

Class and azole	Kd (nM)a	IC50 (μM)b	Residual CYP51 activity (%)c
Imidazole pharmazoles
Clotrimazole	24.6 ± 5.2	0.95	27
Econazole	≤7.8 ± 1.5	ND
Ketoconazole	≤12.7 ± 5.2	2.27	38
Miconazole	≤8.8 ± 3.1	ND
Triazole pharmazoles
Fluconazole	39.5 ± 5.0	1.35	30
Itraconazole	≤8.9 ± 3.7	0.84	19
Posaconazole	≤3.2 ± 1.4	ND
Voriconazole	≤14.6 ± 6.3	ND
Imidazole agriazole
Prochloraz	≤5.8 ± 1.2	0.15	0
Triazole agriazoles
Epoxiconazole	≤3.5 ± 1.1	0.17	0
Propiconazole	≤5.7 ± 2.1	0.20	0
Prothioconazole-desthio	50.8 ± 16.8	ND
Tebuconazole	≤14.6 ± 5.7	0.47	27
Triadimenol	≤16.4 ± 11.8	ND

^aValues are means and standard deviations for three replicates.

^bND, not determined.

^cResidual CYP51 activity observed in the presence of 4 μM azole antifungal agent, expressed as a percentage of the CYP51 activity observed in the absence of azole. Prothioconazole did not bind to SpCYP51.

IC₅₀ determinations for azole antifungal agents on SpCYP51 activity.

IC₅₀ determinations using 0.5 μM SpCYP51 with the pharmazoles fluconazole, itraconazole, ketoconazole, and clotrimazole (Fig. 8A) and the agriazoles epoxiconazole, propiconazole, prochloraz, and tebuconazole (Fig. 8B) confirmed that all eight azoles severely inhibited SpCYP51 activity using eburicol as the substrate and AfCPR1 as the redox partner. The binding of epoxiconazole, propiconazole, and prochloraz to SpCYP51 was tight, with similar IC₅₀s of 0.17, 0.20, and 0.15 μM, respectively, being observed, along with no residual CYP51 activity in the presence of 2 μM azole. The IC₅₀ for a tightly binding azole that cannot be displaced by substrate would be half the CYP51 concentration (∼0.25 μM), as was observed for epoxiconazole, propiconazole, and prochloraz. The binding of tebuconazole, itraconazole, clotrimazole, fluconazole, and ketoconazole appeared to be slightly less tight, with IC₅₀s of 0.47, 0.84, 0.95, 1.35, and 2.27 μM, respectively, with residual SpCYP51 activities of 27%, 19%, 27%, 30%, and 38% observed at 4 μM azole, suggesting that these five azoles were displaced from the SpCYP51 heme by the substrate during catalysis.

FIG 8.

Azole IC₅₀ determinations for SpCYP51. IC₅₀s were determined for the pharmazoles clotrimazole (filled circles), fluconazole (open circles), itraconazole (bullets), and ketoconazole (crosses) (A) and the agriazoles epoxiconazole (open circles), propiconazole (bullets), prochloraz (crosses), and tebuconazole (asterisks), with clotrimazole (filled circles) as a reference (B). The CYP51 reconstitution assay mixtures contained 0.5 μM SpCYP51 and 1 μM AfCPR1, with 50 μM eburicol as the substrate. Relative velocities of 1.00 corresponded to actual velocities of 0.87 ± 0.21 min⁻¹ for SpCYP51. The means and standard errors from two replicates are shown.

MIC studies with Saprolegnia species.

The MIC results obtained with most of the azole antifungals (Fig. 9 and Table 2) were surprising, as all three Saprolegnia species were insensitive to epoxiconazole, fluconazole, itraconazole, and posaconazole (MIC₁₀₀ > 256 μg ml⁻¹), while the remaining azoles, except clotrimazole, gave MIC₁₀₀s between 8 and 64 μg ml⁻¹ in comparison to 1 μg ml⁻¹ for malachite green. Hu et al. (56) also determined the MIC for malachite green to be 1 μg ml⁻¹ for Saprolegnia, supporting the validity of our MIC method. Clotrimazole, though, proved to be as effective as malachite green against Saprolegnia species, with an MIC₁₀₀ of 1 to 2 μg ml⁻¹. Clotrimazole proved to be effective despite having an apparent 8-fold-lower affinity for SpCYP51 than posaconazole (K_d, ∼25 nM compared to ≤3 nM) and an IC₅₀ 6-fold higher than that of prochloraz. The next most effective azoles were econazole and miconazole (MIC₁₀₀, 8 μg ml⁻¹), suggesting that imidazoles were generally more effective at controlling Saprolegnia infection than triazoles.

FIG 9.

MIC determinations for Saprolegnia species. MIC₁₀₀s were determined in duplicate after 72 h at 20°C using serial dilutions of 256, 128, 64, 32, 16, 8, 4, 2, 1, 0.5, and 0.25 μg ml⁻¹ compound along with a dimethyl sulfoxide control. The MIC plates for epoxiconazole, prochloraz, clotrimazole, and malachite green are shown for S. parasitica; identical results were obtained with S. diclina and S. ferax.

TABLE 2.

MIC₁₀₀s for azole antifungals against Saprolgenia species

Inhibitor	MIC100 (t = 72 h)a
	μg ml−1	μM
Clotrimazole	1 to 2	3 to 6
Econazole	8	18
Epoxiconazole	>256	>776
Fluconazole	>256	>836
Itraconazole	>256	>363
Ketoconazole	64	>120
Miconazole	8	17
Posaconazole	>256	>365
Prochloraz	32	85
Propiconazole	64	187
Prothioconazole	16	46
Prothioconazole-desthio	32–64	103–205
Tebuconazole	64	208
Triadimenol	64	216
Malachite green	1	3

^aMIC₁₀₀ experiments were performed in duplicate using S. parasitica, S. diclina, and S. ferax. All three Saprolegnia species displayed identical azole susceptibilities to the azoles tested and to the malachite green positive control at 20°C. We have previously shown that prothioconazole is progressively converted to prothioconazole-desthio in the medium over time (45, 62), explaining why a compound that cannot bind to SpCYP51 can inhibit SpCYP51 activity and sterol biosynthesis in Saprolegnia.

Sterol composition of Saprolegnia species.

The sterol compositions of untreated and prochloraz-treated (16 μg ml⁻¹) S. diclina, S. ferax, and S. parasitica were determined after 72 h at 20°C (Fig. 10 and Table 3). Untreated S. diclina, S. ferax, and S. parasitica contained four main sterols—desmosterol, 24-methylene cholesterol, fucosterol (the Saprolegnia equivalent of ergosterol in fungi), and cholesterol—with traces of lanosterol also being detectable. S. diclina and S. ferax had similar sterol profiles, while S. parasitica had a 10-fold-lower abundance of fucosterol and a 2- to 3-fold-higher abundance of desmosterol.

FIG 10.

Gas chromatograms of sterol content of Saprolegnia species. The nonsaponifiable lipid fraction was isolated (50) from untreated and prochloraz-treated (16 μg ml⁻¹) S. diclina, S. ferax, and S. parasitica after 72 h growth at 20°C. Samples were derivatized with TMS prior to analysis by GC-MS (45). The gas chromatograms are shown for untreated (solid lines) and prochloraz-treated (dashed lines) samples. Gas chromatogram peaks were confirmed to be cholesterol (peak 1), desmosterol (peak 2), 24-methylene cholesterol (peak 3), fucosterol (peak 4), and lanosterol (peak 5) by the mass fragmentation patterns obtained (see Fig. S1 in the supplemental material).

TABLE 3.

Sterol compositions of untreated and azole-treated Saprolegnia species

Sterol	Sterol composition (%)
	Untreated			Prochloraz treated (16 μg ml−1)			Clotrimazole treated (0.5 μg ml−1)
	S. ferax	S. parasitica	S. diclina	S. ferax	S. parasitica	S. diclina	S. parasitica
Cholesterol	2.6	2.0	4.2	0.6			0.8
Desmosterol	17.3	54.7	19.8	0.6	2.5	9.3	1.9
24-Methylene cholesterol	46.3	39.4	49.1	5.9	0.6		4.9
Fucosterol	33.8	3.9	26.9	34.5
Lanosterol	Tr	Tr	Tr	58.4	96.9	90.7	92.4

The sterol compositions of untreated and prochloraz-treated S. ferax, S. parasitica, and S. diclina, in addition to clotrimazole-treated S. parasitica, were determined by extraction of the nonsaponifiable lipid fraction (50) followed by derivatization with BSTFA-TMS-pyridine and analysis by GC-MS (45). All analyses were performed in duplicate. Tr, trace.

The large build-up of lanosterol in the prochloraz-treated samples (Fig. 10 and Table 3) indicated prochloraz uptake in Saprolegnia and inhibition of CYP51 activity, preventing the 14α-demethylation of lanosterol, which leads to the depletion of the other major Saprolegnia sterols. Prochloraz-treated S. parasitica and S. diclina had similar sterol profiles, with lanosterol constituting 90 to 97% of the total sterol and the depletion of cholesterol, desmosterol, 24-methylene cholesterol, and fucosterol. However, the sterol profile of prochloraz-treated S. ferax differed in that fucosterol content remained constant while cholesterol, desmosterol, and 24-methylene cholesterol contents were depleted and lanosterol accumulation was lower, at only 58% of the sterol content. Further investigations are required to explain the differences in sterol composition among Saprolegnia species in response to azole treatment. Clotrimazole treatment (0.5 μg ml⁻¹) of S. parasitica also resulted in a large accumulation of lanosterol (92% of total sterol) and the depletion of the other sterols (Table 3). Therefore, clotrimazole was effective at inhibiting CYP51 activity in vivo at a 32-fold-lower concentration than prochloraz. In contrast, the sterol profiles of malachite green-treated Saprolegnia were unaltered, indicating that the mode of action of malachite green was different from that of azole antifungals and did not inhibit CYP51.

Searching the Broad Institute S. parasitica genome database for sterol biosynthesis enzymes identified 12 enzymes (Table 4). For lanosterol synthase and Δ⁵ sterol desaturase, two potential candidates were identified for each gene, in agreement with the work of Jiang et al. (51). For lanosterol synthase, SPRG_117832.2 (805 amino acids) and SPRG_17895.2 (472 amino acids) shared 99% identity. SPRG_17895.2 was a truncated version of SPRG_117832.2 starting at residue 334 and contained five point mutations. For Δ⁵ sterol desaturase, SPRG_11773.2 (270 amino acids) and SPRG_18544.2 (272 amino acids) also shared 99% identity, with just three point mutations in the first 263 amino acid residues; however, the C termini (7 to 9 residues) of the two proteins differed. Functional analysis of the lanosterol synthase and Δ⁵ sterol desaturase candidates (expression, purification, and reconstitution assays) will be required to demonstrate the catalytic function of each protein. The S. parasitica sterol biosynthesis enzymes identified here (Table 4) agree with those identified by Jiang et al. (51). However, the sterol pathway proposed by Jiang et al. (51) did not relate to the sterols identified in this study. Our pathway (Fig. 11) shows the four main sterols identified in Saprolegnia (desmosterol, 24-methylene cholesterol, fucosterol, and cholesterol), although the exact catalytic sequence of the enzymes is not known. We could not identify a Δ³ sterol keto reductase candidate in S. parasitica, although the presence of desmosterol and 24-methylene cholesterol in the cell membranes indicates that this chemical reaction occurs in vivo. This may be because the S. parasitica Δ³ sterol keto reductase has low sequence identity to mammalian and fungal Δ³ sterol keto reductases or because this chemical reaction is performed by a phylogenetically unrelated enzyme in S. parasitica. Jiang et al. (51) were unable to identify a Δ³ sterol keto reductase candidate in S. parasitica, as were Madoui et al. (35) in A. euteiches.

TABLE 4.

Sterol biosynthetic genes in Saprolegnia parasitica^a

Enzyme	Gene locus ID	Query accession no.
Squalene synthase	SPRG_06233.2	B2ZWF5
Squalene epoxidase	SPRG_11641.2	B2ZWF6
Lanosterol synthase	SPRG_11783.2 and SPRG_17895.2	B2ZWF7
Δ14 sterol demethylase (CYP51)	SPRG_09493.2	B2ZWE1
Δ14 sterol reductase	SPRG_00418.2	B2ZWF8
Δ4 sterol methyl oxidase	SPRG_01623.2	B2ZWF9
Δ3 sterol dehydrogenase	SPRG_01499.2	Q4X017
Δ3 sterol keto reductase	Not identified
Δ24 sterol methyltransferase	SPRG_05001.2	B2ZWE2
Δ8 sterol isomerase	SPRG_13330.2	Q15125
Δ5 sterol desaturase	SPRG_11773.2 and SPRG_18544.2	Q4WDL3
Δ7 sterol reductase	SPRG_01085.2	B2ZWG0
Δ24 sterol reductase	SPRG_04988.2	B2ZWG1

^aThe gene locus IDs of the most likely candidates are shown along with the UniProtKB accession number of the sequence used as the query in NCBI BLASTP searches. Sterol biosynthetic gene homologs from A. euteiches (35), A. fumigatus (Q4WDL3 and Q4X017), and H. sapiens (Q15125) were used for the BLASTP searches.

FIG 11.

Postlanosterol sterol biosynthetic pathway in Saprolegnia parasitica. The postlanosterol biosynthetic pathway was elucidated based on the sterol metabolites observed (Fig. 10) and by searching the S. parasitica genome database using homologs from A. euteiches, A. fumigatus, and H. sapiens. Dashed lines indicate reactions that involve several enzymes.

DISCUSSION

Our phylogenetic analysis of the SpCYP51 protein suggested that S. parasitica was more closely related to plants and animals than “true” fungi, in agreement with previous findings that true fungi and oomycetes were phylogenetically distinct (35, 57). However, the low sequence identity between SpCYP51 and the fungal CYP51 proteins suggests that azole antifungal agents commonly used to fight fungal infection in the clinic and in agriculture may not be so efficacious against oomycetes containing CYP51 enzymes.

SpCYP51 sterol binding affinities were similar to those reported for the trypanosomal CYP51 enzymes of Trypanosoma cruzi and Leishmania infantum (58, 59). However, purified SpCYP51 displayed 3- to 8-fold-higher affinity for sterol than the CYP51 enzymes of C. albicans, A. fumigatus, Mycosphaerella graminicola, and H. sapiens (60–62). The ability of SpCYP51 to 14α-demethylate lanosterol, eburicol, and obtusifoliol confirmed that the putative CYP51 was indeed as predicted and that SpCYP51 had been isolated in a fully functional form.

SpCYP51 binding affinities for pharmazoles were similar to those reported for C. albicans CYP51 (K_d, 10 to 56 nM) (60). SpCYP51 binding affinities for the agriazoles epoxiconazole, prochloraz, propiconazole, and triadimenol were 5-, 8-, 6- and 4-fold tighter (K_d, 4 to 16 nM) than those observed with C. albicans CYP51 (K_d, 22 to 68 nM) (60). Therefore, SpCYP51 should be strongly inhibited by azole antifungal agents, as no inherent resistance (high K_d values) to azole antifungals was evident, unlike with A. fumigatus CYP51A and CYP51B, which both bound fluconazole weakly (K_ds, 11,930 and 4,030 nM) (61). This was further supported by the IC₅₀ results, which showed strong inhibition of CYP51 activity by azole antifungal agents (Fig. 7), especially epoxiconazole, propiconazole, and prochloraz (IC₅₀s, 0.17, 0.20, and 0.15 μM). The SpCYP51 IC₅₀s with the agriazoles were 2- to 3-fold lower than the equivalent IC₅₀s with C. albicans CYP51 (60), suggesting that agriazoles would prove effective at inhibiting Saprolegnia growth. However, the relatively poor performance of most azoles against Saprolegnia, with the exception of clotrimazole, in terms of MIC₁₀₀s was in agreement with the limited previous azole antifungal studies with Saprolegnia species.

Heeres et al. (63) established that the MIC₅₀ for terconazole was 10 μg ml⁻¹ and the MIC₁₀₀ was 100 μg ml⁻¹, whereas the MIC₁₀₀ for clotrimazole was >100 μg ml⁻¹ in Sabouraud broth (63), which was ∼100-fold higher than that determined in this study. Heeres et al. (63) also determined the MIC₁₀₀ for clotrimazole with C. albicans to be >100 μg ml⁻¹ in Sabouraud broth, which was surprisingly high considering that clotrimazole is used for the treatment of Candida infections in humans and calls the methodology into question. More recent MIC determinations for clotrimazole against C. albicans range from 0.03 to 1 μg ml⁻¹ (64–66), suggesting that the MIC determinations reported by Heeres et al. (63) were over 100-fold higher than present determinations. In contrast, Marking et al. (67) found the MIC for clotrimazole to be 10 to 30 μg ml⁻¹ after a 15-min exposure, which fell to 10 μg ml⁻¹ after 60 min with Saprolegnia hypogyna. Our MIC determinations exposed Saprolegnia to azoles for 72 h and included a number of different species, with results similar to those of Marking et al. (67). Hu et al. (56) determined the Saprolegnia MIC for propiconazole to be 100 μg ml⁻¹, compared to 64 μg ml⁻¹ in this study, and Saprolegnia MICs for thiabendazole, difenoconazole, and diniconazole to be >100, 50, and 100 μg ml⁻¹. High azole MIC₁₀₀s were probably due to either poor uptake into the cells or the efficient efflux of the azoles out of the cells. Clotrimazole is a relatively compact molecule, unlike many of the ineffective azoles, suggesting that size is important in efficient uptake of azole antifungals in Saprolegnia.

For azole antifungal agents to be effective at combating Saprolegnia infection, they need to display high potency against the intended target CYP51 (SpCYP51) enzyme, with minimal interaction with the host CYP51 (and other off-target host CYP enzymes). Morrison et al. (68) determined the K_d values for ketoconazole and propiconazole to be 260 and 640 nM with recombinant Danio rerio (zebrafish) CYP51 by type II difference ligand binding spectroscopy and 76 and 3,700 nM by in silico azole ligand docking experiments. Therefore, the apparent selectivities for SpCYP51 over the zebrafish homolog based on azole K_d values were 20- and 112-fold for ketoconazole and propiconazole (azole binding studies) or 6- and 649-fold (in silico studies). This is encouraging, and it would be of interest to determine the selectivity for the more potent Saprolegnia inhibitor clotrimazole.

The change in sterol composition of Saprolegnia in response to exposure to prochloraz and clotrimazole confirmed that CYP51 was being inhibited in vivo by the accumulation of lanosterol in the cell membranes. Similar lanosterol accumulation was observed in S. ferax that had been treated with triarimol, a CYP51-inhibiting pyrimidine fungicide (36), and accumulation of eburicol was observed in fungi treated with azole antifungals (69–71). Malachite green-treated Saprolegnia had unaltered sterol profiles, indicating that the mode of action of malachite green was different from that of azole antifungals, with no inhibition of CYP51 function. This is in agreement with previous findings that malachite green acts independently of known antifungal targets, such as sterol biosynthesis and drug efflux pump proteins in C. albicans (MIC₅₀, 0.1 μg ml⁻¹) (72). Malachite green was found to affect 207 genes in C. albicans, with the upregulation of 167 genes involved in oxidative stress, virulence, carbohydrate metabolism, heat shock, and amino acid metabolism and the downregulation of 37 genes involved in iron acquisition, filamentous growth, and mitochondrial respiration (72). Malachite green affects multiple mechanisms to exert its antifungal effect and leads to a shift in metabolism toward fermentation, increased generation of reactive oxygen species, iron depletion, and cell necrosis in C. albicans (72) and is likely to behave similarly in Saprolegnia species.

There appear to be no published studies investigating the effectiveness of clotrimazole at combating Saprolegnia infection of salmonid eggs and fry in vivo. Bailey and Jeffrey (73) screened 215 candidate fungicides for suitability as aquatic fungicides, 120 of which, including clotrimazole, they deemed unsuitable, but did not elaborate why clotrimazole was deemed unsuitable. The OSPAR commission (74) established that clotrimazole was not acutely or chronically toxic to fish (Brachydanio rerio and Oncorhynchus mykiss). The report, however, concludes that at present there is no risk to the environment from the current usage of clotrimazole. Azole antifungal agents are not presently used to treat Saprolegnia infections in aquaculture. However, azole antifungal agents are widely used in medicine to combat infections by human fungal pathogens such as Candida and Aspergillus species and in agriculture to combat infections by plant fungal pathogens such as Mycosphaerella graminicola and Fusarium species. Clotrimazole proved to be as effective at inhibiting Saprolegnia growth in vitro as malachite green (MIC, ∼1 μg ml⁻¹). It is our contention that clotrimazole would prove an effective control agent of Saprolegnia infection, especially if used to pretreat salmonid eggs in commercial hatcheries. Fungal diseases have been detected in many commercially important fish species besides salmonids, including catfish, pike, bass, tilapia, roach, carp, mullet, and sturgeon (3, 75–78), such that treatment with clotrimazole analogs could potentially reduce financial losses significantly by effective control of oomycete and fungal infections. Therefore, studies need to be performed on salmonid eggs in vivo to determine optimum clotrimazole doses and pretreatment lengths for the prevention of Saprolegnia infections as well as for other farmed species and for infected fish. Hu et al. (56) found that the strobilurin antifungals kresoxim-methyl and azoxystrobin were most effective against Saprolegnia, with MICs of 1 and 0.5 μg ml⁻¹; however, both of these compounds are toxic to fish, highlighting the problems of translating successful in vitro experiments into effective in vivo treatments.