Identification of antibiotics for use in selection of the chytrid fungi Batrachochytrium dendrobatidis and Batrachochytrium salamandrivorans

Global amphibian populations are being decimated by chytridiomycosis, a deadly skin infection caused by the fungal pathogens Batrachochytrium dendrobatidis (Bd) and B. salamandrivorans (Bsal). Although ongoing efforts are attempting to limit the spread of these infections, targeted treatments are necessary to manage the disease. Currently, no tools for genetic manipulation are available to identify and test specific drug targets in these fungi. To facilitate the development of genetic tools in Bd and Bsal, we have tested five commonly used antibiotics with available resistance genes: Hygromycin, Blasticidin, Puromycin, Zeocin, and Neomycin. We have identified effective concentrations of each for selection in both liquid culture and on solid media. These concentrations are within the range of concentrations used for selecting genetically modified cells from a variety of other eukaryotic species.


Introduction
Chytrids are early diverging fungi that are commonly found in aquatic and moist environments [1]. They play key ecological roles, particularly by cycling carbon between trophic levels [2,3]. Chytrids have a biphasic life cycle characterized by motile and sessile stages (Fig 1) [4][5][6]. They begin their life as motile "zoospores," which use a flagellum to swim through water and, for some species, actin-based motility to crawl along surfaces [7,8]. Zoospores then transition to a sessile growth stage by retracting their flagellum and building a cell wall in a process referred to as encystation. Encysted spores of many species develop into sporangia and develop hyphal-like structures called rhizoids and grow rapidly. Each sporangium produces many zoospores that exit via discharge papillae to begin the life cycle anew.
Many chytrids are pathogens that infect protists, plants, algae, fungi, and vertebrates [2]. The most infamous chytrids are the vertebrate pathogens Batrachochytrium dendrobatidis (Bd) and B. salamandrivorans (Bsal). Both pathogens cause chytridiomycosis, a skin disease plaguing amphibians worldwide [4,6]. Recent estimates indicate that Bd has affected several hundred amphibian species and has been recorded on every continent except for Antarctica [9][10][11]. Bsal was more recently discovered in 2013 after a steep decline in fire salamander populations in Belgium [6]. Management strategies for these pathogens have been developed and implemented in limited contexts, but implementation in real world settings remains a challenge. To develop better treatments, we need to understand the biology of chytrids in order to identify targets for drug development. However, studying the molecular mechanisms driving pathogenesis remains challenging due to the lack of genetic tools available for chytrid fungi. Electroporation protocols have been developed for Bd and Bsal, which could be used to deliver molecular payloads for genetics manipulation such as plasmids and/or CRISPR-Cas9 complexes [12]. The recent success in genetic manipulation of a related chytrid species, Spizellomyces punctatus (Sp), is a major breakthrough for our ability to study chytrid biology [8]. We and others are now striving to adapt this technology to Bd and Bsal to further our understanding of chytridiomycosis.
A key step to genetic tool development is the identification of methods for selection of successful transformants. The most commonly used selection method is antibiotic resistance: As illustrated here with images of Bsal, chytrid fungi have a biphasic life cycle characterized by a stationary growth phase called a sporangium (top) and a motile dispersal phase called a zoospore (bottom). Images taken at 100X using differential interference contrast (DIC) microscopy.
https://doi.org/10.1371/journal.pone.0240480.g001 incorporating a gene that provides specific drug resistance allows transformed cells to survive exposure to the antibiotic while all of the other cells are killed [13]. Distinct classes of antibiotics are commonly used for selection, each with their own molecular targets and corresponding organismal specificity. In addition to testing whether a given antibiotic kills cells of interest, it is important to pay attention to the effective concentration of each antibiotic. This is because a low concentration will not apply sufficient selective pressure and a high concentration could produce off-target effects and kill cells indiscriminately [14].
In this paper, we examine five antibiotics used in fungal and animal systems and identify the effective inhibitory concentration(s) necessary to prevent cell growth in liquid and solid media. Hygromycin, Blasticidin, and Puromycin inhibit protein translation in both bacterial and eukaryotic cells. Hygromycin inhibits protein synthesis by binding to the small ribosomal subunit and stabilizing the tRNA in the A site, preventing the progression of translation [15]. Blasticidin inhibits the terminating step of translation while Puromycin causes the ribosome to prematurely detach from mRNA [16,17]. Although neomycin targets the prokaryotic 30S ribosomal subunit and causes codon misreading and mistranslation, it has been used in eukaryotes because of the similarity between mitochondrial and chloroplast ribosomes and bacterial ribosomes [18]. Zeocin intercalates in the DNA of both bacteria and eukaryotes and introduces double-stranded breaks, ultimately causing cell death [19].

Results
To establish appropriate selection compounds for use with Bd and Bsal, we first identified antibiotics commonly used for selection with both mammalian and fungal systems. We chose five compounds (Hygromycin, Blasticidin, Puromycin, Zeocin, and Neomycin) to test based on the mechanism of action of each compound, their proven efficacy for use with both animal and fungal cells, and the availability of resistance genes ( Table 1). We next tested the ability of these five compounds to inhibit the growth of Bd and Bsal cells in liquid culture. Although solid agar media is typically used for colony selection in chytrid and other fungi [8,20,21], we chose to use liquid culture to identify initial working concentrations because measuring zoospore release in liquid media is rapid and easily quantified.
To measure the effect of each antibiotic on Bd and Bsal growth, we added a wide range of antibiotic concentrations to cultures of age matched zoospores and allowed them to grow for one full life cycle: three (Bd) or four (Bsal) days. We then measured the concentration of released zoospores in each culture. Initial concentrations were selected based on known inhibitory concentrations for other organisms ( Table 1) and spanned many orders of magnitude. Based on these preliminary experiments (not shown), we then identified possible working concentration ranges for each antibiotic in both species and tested intermediate concentrations using three biological replicates separated in time (Figs 2 and 3). To enable comparison of zoospore release from replicate experiments conducted on different days, we normalized counts for each replicate to its antibiotic-free control.
� the organism had to be made susceptible for the antibiotic to work. ‡ the neo resistance gene is also used for resistance to the drug G418 which was not tested in this study. https://doi.org/10.1371/journal.pone.0240480.t001 Having identified working concentrations of these compounds for use with liquid media, we next tested their efficacy on solid media. Growing cells on solid media allows for colony formation, which is useful for isolating successful and independent genetic transformants by "picking" colonies that grow under selection. To identify useful concentrations for selection on solid media, we inoculated zoospores on nutrient agar plates containing varying antibiotic concentrations. After a full growth cycle on selective media (three days for Bd, four days for Bsal), we compared zoospore release to antibiotic-free control cultures by flooding plates with water and looking for motile zoospores (S1 and S2 Videos). We defined successful concentrations as those which yielded no zoospore release in either replicate. We found at least one concentration for each antibiotic that prevented zoospore release in the timeframe of a typical growth cycle (Figs 4 and 5).
Because detection of colony formation often requires multiple growth cycles, we evaluated the efficiency of growth inhibition by growing plates with no zoospore release for 14 days. We

PLOS ONE
Identification of antibiotics for selection of Batrachochytrium dendrobatidis and B. salamandrivorans found that all the tested antibiotics inhibited Bd growth on solid media, but only Hygromycin, Blasticidin and Zeocin inhibited growth in Bsal. For Bd, Hygromycin has the lowest minimum concentration at 0.1 μg/ml, with Blasticidin and Zeocin both following at 10 μg/ml, Puromycin at 100 μg/ml, and Neomycin at 1 mg/ml (Fig 4). In Bsal, Hygromycin, Blasticidin, and Zeocin all prevented growth for at least 14 days at a concentration of 10 μg/ml, while Puromycin and Neomycin did not prevent growth on solid media (Fig 5). The recommended concentrations for selection are highlighted in orange on the tables in both figures (Figs 4B and 5B).

Discussion
This study identified drug concentrations that reproducibly inhibited Bd and Bsal growth in either liquid culture or on solid media. When a drug worked in both liquid culture and solid

PLOS ONE
Identification of antibiotics for selection of Batrachochytrium dendrobatidis and B. salamandrivorans media, the solid media typically required a higher concentration of antibiotic. This may be because of the additional minerals found in the agar not present in the liquid media [66]. Hygromycin, Zeocin, and Blasticidin worked well for both species and at concentrations within the typical range used for genetic selection in other species ( Table 1). Puromycin and Neomycin were both able to inhibit growth of Bd and Bsal, but required higher concentrations than are used for animal cell lines. Although Hygromycin, Zeocin, and Blasticidin are all effective for preventing growth of Bd and Bsal, we recommend first using Hygromycin for genetic selection because it has been successfully used for selection of transformants in the nonpathogenic chytrid Spizellomyces punctatus, and is widely used for other fungal species [8,[36][37][38].
The ability to select for genetically transformed cells will allow for tractable genetic models to facilitate hypothesis testing in Bd and Bsal. The identification of useful selection agents and appropriate working concentrations is an important first step in developing genetic tools for use with Bd and Bsal. The natural step forward will be the design of selection cassettes, most commonly in the form of transformation plasmids. We look forward to the development of these and related molecular tools that will help us answer questions about the basic cell biology of chytrids, fungal evolution, and amphibian pathology.

Cell growth and synchronization
Batrachochytrium dendrobatidis (Bd) isolate JEL 423 was grown in 1% (w/v) tryptone (Apex Cat. 20-251) in tissue culture treated flasks (Cell Treat 229340) at 24˚C for three days. B. salamandrivorans (Bsal) isolate AMFP 1 was grown in half-strength TGhL liquid media (0.8% Tryptone, 0.2% gelatin hydrolysate, 0.1% lactose (w/v) in tissue culture treated flasks at 15˚C for four days [67]. For both species, we synchronized the release of motile zoospores by gently washing the flask three times with fresh growth media and then incubating with 10 mL of media for 2 hours. Age matched zoospores were then collected by centrifugation at 2000 rcf for 5 mins, resuspended in media, counted, and used for experiments as outlined below.

Drug treatments and quantitation for cells grown in liquid media
Neomycin (Fisher Cat. AAJ67011AE), Hygromycin B (Fisher Cat. AAJ60681MC), Blasticidin (Fisher Cat. BP2647100), Puromycin (Fisher Cat. BP2956100), and Zeocin (Fisher Cat. AAJ671408EQ), were screened for growth inhibition of Bd and Bsal. Cells were diluted to a starting concentration of 5x10 5 cells/mL and 250 uL of cells were added to each well of a sterile tissue culture treated 24-well plate (Cell Treat 229123). 250 μl of appropriately diluted antibiotics and matched carrier controls were added to each well and mixed thoroughly. Plates were sealed with parafilm and grown at either 24˚C for three days (Bd), or 15˚C for four days (Bsal). For each of three biological replicates spaced in time, the concentration of released zoospores was estimated using the average of two independent hemocytometer counts. Zoospore concentrations were normalized to the no drug control and data plotted using Prism (GraphPad v8). The '+' demonstrates the relative zoospore activity of each plate compared to an antibiotic-free control plate. The box highlights zoospores, which appear as small dots while the bracket highlights sporangia. The zoospores in the '0' image are immotile (see S1 Video). Scale bar 50 μm. (B) Bd growth on antibiotic selection plates. Concentrations highlighted in bold and orange are the lowest concentrations that prevent growth for at least 14 days post zoospore plating. https://doi.org/10.1371/journal.pone.0240480.g004

PLOS ONE
Identification of antibiotics for selection of Batrachochytrium dendrobatidis and B. salamandrivorans

Drug treatments and quantitation for cells grown on solid media
We added 1% agar to 50 mL batches of 1% tryptone (w/v) and half-strength TGhL then autoclaved. Each antibiotic was added to a separate, pre-cooled, 50 mL batch of media, and 10 mL of the solution added to one of five 15 mm 2 plates (VWR 25384-090) and allowed to solidify. Equal volume of appropriate carrier liquid was added to the pre-cooled 50 mL batch of agarmedia to create control plates. Plates were wrapped in parafilm and aluminum foil, and stored at 4˚C. Plates were inoculated by evenly spreading 5.0 x 10 6 zoospores across the agar and incubated at 24˚C for three days (Bd) or 15˚C for four days (Bsal). Three control plates were used per replicate to ensure a point of comparison if one were to be contaminated. Zoospore release was evaluated by imaging each plate for 20 seconds at one second intervals using a Nikon Ti2-E inverted microscope equipped with 10x PlanApo objective and sCMOS 4mp camera (PCO Panda) using white LED transmitted light. Approximate zoospore activity was assessed as: 0 (no visible zoospores), + (< 25% zoospore activity of control plates lacking antibiotic), ++ (~50% zoospore activity of control plates), or +++ (equivalent zoospore activity to control plates). To determine the lowest antibiotic concentration that could completely inhibit growth, plates that yielded "0" growth were allowed to grow for 14 days at the appropriate incubation temperature and reassessed as above.