Activity of antimicrobial skin peptides from ranid frogs against Batrachochytrium dendrobatidis, the chytrid fungus associated with global amphibian declines

• Louise A Rollins-Smith^{a, ,} ,
• Cynthia Carey^b,
• Joyce Longcore^c,
• Jennifer K Doersam^a,
• Angela Boutte^a,
• Judsen E Bruzgal^b,
• J.Michael Conlon^d

• ^a Departments of Microbiology and Immunology & Pediatrics, Vanderbilt University, Medical Center North, A-5301, Nashville, TN 37232, USA
• ^b Department of Environmental, Population, and Organismic Biology, University of Colorado, Boulder, CO 80309, USA
• ^c Department of Biological Sciences, University of Maine, Orono, ME 04469, USA
• ^d Department of Biomedical Sciences, Creighton University School of Medicine, Omaha, NE 68178, USA

• Received 10 August 2001. Revised 8 October 2001. Accepted 16 October 2001. Available online 26 November 2001.

• http://dx.doi.org/10.1016/S0145-305X(01)00088-X, How to Cite or Link Using DOI

• Permissions & Reprints

---

Keywords

• Amphibian;
• Antimicrobial peptide;
• Batrachochytrium;
• Brevinin;
• Chytrid;
• Esculentin;
• Ranalexin;
• Ranatuerin

Abbreviations

• MIC, minimal inhibitory concentration;
• PGLa, peptide with amino terminal glycine and carboxyl terminal leucinamide

---

1. Introduction

Amphibian populations in diverse places have been experiencing declines since the 1960s (reviewed in [1], [2] and [3]). Among possible causes for these population declines currently being investigated are introduction of predators [4], exposure to pesticides [5], and exposure to ultraviolet B (UV-B) [6] and [7]. In addition to these possible causes, accumulating evidence supports the hypothesis that a newly discovered fungal skin pathogen, Batrachochytrium dendrobatidis, is responsible for some recent declines [8], [9], [10] and [11]. Ranid frog populations have suffered disproportionate losses compared to other anurans in western North America [12]. Mass mortalities of ranid species were noted first in the mid- to late-1970s in populations of Rana turahumarae in Arizona [13], Rana pipiens in the Colorado Rockies [14] and [15], Alberta [16], and Manitoba [17], and in populations of Rana muscosa in the Sierra Nevada of California [18]. A disease, thought to be of bacterial origin, was suggested as the direct cause of death by Scott [12], Carey [15], and Bradford et al. [18]. Analysis of museum specimens indicates the presence of chytrid infections in R. pipiens collected in Colorado at the time of these die-offs [C. Carey, D. Green, and P.S. Corn, unpublished data]. Presently, chytrid infections resulting in death have been noted in an adult R. pipiens collected in Colorado (A. Pessier, personal communication) and in metamorphosed Rana chiricahuensis, Rana subaquavocalis, and Rana yavapaiensis in Arizona [19]. Chytrids have also been detected in the mouth parts of larval R. muscosa in California [20].

B. dendrobatidis infects and replicates within keratinized skin cells. Motile reproductive spores (zoospores) encyst and insert their contents into epidermal cells, where each unit enlarges and forms a spherical zoosporangium within a skin cell. At maturity, the contents of each zoosporangium cleave into flagellated zoospores, which are released through a papilla that opens to the exterior of the skin. These zoospores frequently infect nearby areas of the skin of the same host, but may infect a new host [8], [9] and [10]. Because the infection involves only the keratinized epithelium, local cell-mediated immune responses do not seem to be evoked [10]. Thus, any deterrent to infection and further spread of this pathogen may depend on the innate defense factors produced by the granular glands of the skin. Among the defensive factors produced at this site are a variety of antimicrobial peptides. Antimicrobial peptides have been isolated from diverse anuran species. These peptides range in size from 12 to 46 amino acid residues. Most are cationic, hydrophobic, and the non-cyclic regions adopt an amphipathic α-helical conformation on binding to the cell membrane. They are thought to bind to charged residues on target cells and disrupt membrane function leading to death of the cell (reviewed in [21], [22] and [23]). The primary structures of the 10 peptides tested in the studies reported here are shown in Table 1. All except temporin-1Ob contain an intramolecular disulfide bridge that forms either a six- or seven-member ring at the carboxyl-terminus of the molecule. Chou–Fasman analysis predicts that all of these peptides would be able to adopt an α-helical conformation (unpublished data). All have been shown to have activity against Escherichia coli, but their activity against amphibian pathogens has not previously been tested. We show here that all can inhibit the growth of zoospores and mature vegetative cells of B. dendrobatidis. Consequently, at effective concentrations, they should be inhibitors of new colonization by the pathogen.

2. Methods

2.1. Culture and maintenance of B. dendrobatidis

B. dendrobatidis was isolated from a diseased blue poison dart frog (Dendrobates auratus) by J.E.L. [9]. It was grown on TGhL agar (16 g tryptone, 4 g gelatin hydrolysate, 2 g lactose, and 10 g agar per 1 l of glass distilled water) or in H broth (10 g tryptone and 3.2 g glucose per 1 l of glass distilled water) at 22–23 °C. Broth cultures were passaged twice weekly to assure that the cells were in an active phase of growth. Zoospores were harvested by flooding the agar surface of 10-day old established cultures with approximately 3.0 ml of sterile broth twice. The broth containing swimming zoospores was passed over sterile nylon spectra/mesh filters (20 μm mesh opening, Fisher Scientific, Pittsburgh, PA) to remove more mature cells. After separation of mature cells, the zoospore population was >99% pure.

2.2. Growth inhibition assay

Our growth inhibition assay [24] was adapted from an assay described by Mor and Nicolas [25]. Briefly, 5×10⁴ mature cells or 5×10⁵ zoospores in a volume of 50 μl of broth were plated in replicates of three or more in a 96-well flat bottom microtiter plate (Costar 3596, Corning, Corning NY, USA) with or without addition of 50 μl serial dilutions of each peptide in broth. The plates were covered not sealed, wrapped in a plastic wrap to limit moisture loss, and incubated on a laboratory bench at 23 °C. Positive control wells received 50 μl of broth without peptide, and negative control wells (on a separate plate) received 50 μl of broth containing 0.4% paraformaldehyde [24]. Growth at 24, 48, 72, and 96 h (23 °C) was measured as increased optical density at 492 nm (OD₄₉₂) with an ELISA plate reader. For growth inhibition assays using zoospores at 23 °C or mature cells at lower temperatures, the incubation period was extended up to eight days. Minimal inhibitory concentration (MIC) is defined as the lowest concentration at which no growth was detectable. That is, the OD₄₉₂ was not significantly greater than that observed for negative control wells containing paraformaldehyde.

2.3. Peptides

Some peptides were isolated from the pooled homogenized skins or skin secretions of adult frogs by J.M.C as previously described [26] and [27]. Ranatuerin-1 was isolated from the American bullfrog, Rana catesbeiana[26]. Ranatuerin-2La and esculentin-2L were isolated from the spotted frog, Rana luteiventris[27]. Esculentin-1A and palustrin-3A were isolated from the crayfish frog, Rana areolata (unpublished data). Brevinin-2Ob and Temporin-1Ob were isolated from the Japanese mountain brown frog, Rana ornativentris[28]. Ranatuerin-2P and esculentin-2P were custom-synthesized (Sigma Genosys, Houston, TX) from the sequence described for these purified peptides obtained from R. pipiens[27]. Ranalexin, originally isolated from young postmetamorphic R. catesbeiana[29], was purchased as a synthetic product from Sigma Chemical (St Louis, MO, USA). All peptides were dissolved in glass distilled water or 0.01% trifluoracetic acid, filter sterilized, and frozen in small aliquots at high concentration (usually 1 mM) at −70 °C and used at various dilutions in broth for culture [24]. Because the activity of many antimicrobial peptides can be greatly affected by ionic strength and presence of divalent cations, we measured the concentrations of sodium, potassium, calcium, and magnesium ions in the assay broth. The broth had no measurable sodium or potassium ions. The concentration of calcium ions in broth was <0.5 mg/dl, and the concentration of magnesium ions was <0.2 mg/dl.

3. Results

3.1. Growth inhibition

To understand the possible role of antimicrobial skin peptides in protection from chytridiomycosis, it is necessary to determine what concentrations of individual skin peptides are sufficient to inhibit growth. All tested peptides significantly inhibited growth at 96 h in mixed cultures of B. dendrobatidis containing both mature cells and zoospores. Significant inhibition was noted at concentrations above 25 μM, and esculentin-1A and palustrin-3A had significant inhibitory activity at concentrations as low as 2 μM ( Fig. 1 and Fig. 2). MIC, defined as the concentration at which no growth was detectable (see Section 2), ranged from about 18 to 100 μM (Table 2). Subculture of material from wells treated with peptides at the MIC showed very limited additional growth. Examination of the contents of these wells with an inverted microscope revealed a much-reduced number of intact cells in comparison with positive control wells. We interpret this as an evidence that the peptides are fungicidal not fungistatic at concentrations at or above the MIC. A comparison of the growth inhibitory activities of all the peptides suggests a ranking of relative effectiveness of these peptides against B. dendrobatidis into three groups as follows: brevinin-2Ob, esculentin-1A, palustrin-3A, and ranalexin>esculentin-2L, esculentin-2P, and temporin-1Ob>ranatuerin-1, ranatuerin-2La, and ranatuerin-2P. Preliminary experiments suggest that five other peptides isolated from R. pipiens (brevinin-1Pa, brevinin-1Pb, brevinin-1Pc, brevinin-1Pd, and temporin-1P) [27] are also effective inhibitors of growth of B. dendrobatidis, but their MIC values were not determined (data not shown).

3.2. Increased sensitivity of zoospores

We hypothesized that zoospores, which lack cell walls, might be more sensitive to growth inhibition by antimicrobial peptides than more mature cells that have developed cell walls [9] and [10]. To compare the sensitivity of zoospores with germlings (maturing fungal cells with rhizoids), parallel growth-inhibition experiments were done with purified zoospores or mixed cultures containing both zoospores and mature stages. The results of a number of replicate experiments comparing the activity of ranatuerin-2P and esculentin-2P against zoospores or germlings were plotted as mean percent inhibition. Zoospores were significantly more sensitive over a range of concentrations with each peptide (Fig. 3). If one compares the amount of each peptide required for 50% growth inhibition, zoospores were inhibited by a concentration that was about one third of that necessary to inhibit mature cells. For esculentin-2P, about 8 μM was sufficient to induce 50% growth inhibition of zoospores whereas about 25 μM was necessary to achieve the same degree of growth inhibition of mature cells. For ranatuerin-2P, the concentration necessary for 50% growth inhibition of zoospores was about 25 μM whereas approximately 75 μM was necessary to inhibit 50% of the growth of mature cells (Fig. 3). A comparison of the MIC of each peptide tested on zoospores in comparison with mature cells is shown in Table 2. For every peptide tested, zoospores were more sensitive.

3.3. Temperature effects

To determine whether cold temperature may interfere with the inhibitory effects of antimicrobial skin peptides, we compared the effectiveness of esculentin-2P and rantuerin-2P at 22 and 10 °C. These two peptides were originally isolated from R. pipiens[27], a species that historically has thrived in cold climates and tolerates cold temperature during overwintering. Both peptides were effective at the warm and cold temperature when used at relatively high concentrations (Fig. 4a–d). At lower concentrations, they both appeared to be somewhat more effective at the colder temperature (compare Fig. 4a with b, and c with d).

4. Discussion

Antimicrobial peptides have been isolated from the skin of a number of frogs belonging to the genera Xenopus, Rana, Bombina, Phyllomedusa, and Litoria (reviewed in Ref. [22]). They have been assayed against human pathogens, such as E. coli, Staphylococcus aureus, and Candida albicans in a search for novel antibiotics. However, few have been assayed for their activity against amphibian pathogens. Previously, we tested six peptides in growth inhibition assays against two fungal pathogens of amphibians (B. dendrobatidis and Basidiobolus ranarum) and one bacterial species that is pathogenic to amphibians, Aeromonas hydrophila. The peptides inhibited growth of the fungal species, but none were active against A. hydrophila[24]. In this report, we have focused on peptides isolated from ranid species because R. pipiens and other species of this genus have suffered population declines that are linked to infection by B. dendrobatidis (discussed in Ref. [1], 12–20).

At least 11 families of antimicrobial peptides isolated from ranid species can be recognized by their amino-acid sequence similarity [27] and [30]. In this study, we tested the ability of representative members of eight of these families of peptides (ranatuerin-1, ranatuerin-2, esculentin-1, esculentin-2, brevinin-2, temporin, palustrin-3, and ranalexin) to inhibit growth of B. dendrobatidis. All the tested peptides were highly effective inhibitory agents against mature cells of B. dendrobatidis, and four of them (brevinin-2Ob, esculentin-1A, palustrin-3A, and ranalexin) were more effective than the others (Table 2). Infections of new hosts are thought to be due to colonization by swimming zoospores [8], [9] and [10]. Therefore, it was especially important to determine the effective concentrations that inhibit zoospore growth. We observed that the zoospores were about two- to four-fold more sensitive than more mature fungal cells of B. dendrobatidis (Fig. 3 and Table 2). Little is known about the natural triggers of peptide secretion. An alarm response to a predator or an injury are stimuli that may activate the sympathetic nervous system to release peptides to the skin (discussed in [21], [22] and [23]). How frequently this would occur in nature is unknown. There is very little information defining the concentrations of antimicrobial peptides present on the skin of resting frogs, but concentrations on the skin following stimulation with noradrenergic agents can exceed 1 mg/ml ([31] and M. Zasloff, personal communication). Furthermore, some of the peptides (e.g. magainin II and PGLa) act synergistically [24] and [32]. Thus, if the appropriate peptides are present at the effective concentrations defined by our studies, spread of existing infections and new colonization by B. dendrobatidis zoospores should be inhibited.

Four of the families of peptides tested in this study are found in skin secretions of R. pipiens (brevinin, esculentin-2, ranatuerin-2, and temporin) [27]. Yet, R. pipiens has suffered population declines thought to be due to chytridiomycosis [C. Carey, D. Green, and P.S. Corn, unpublished data]. In contrast to R. pipiens, the bullfrog (R. catesbeiana) appears to be able to survive low levels of infection with B. dendrobatidis. Experimental exposure of bullfrogs to B. dendrobatidis failed to induce overwhelming infection (A. Strieby, P. Daszak, and D. Porter, unpublished data) whereas experimental infection of some other amphibian species leads to rapid death [9], [33] and [34]. R. catesbeiana and R. pipiens each possess peptides belonging to two of the families of peptides tested in this study (ranatuerin-2 and temporin). In addition, R. catesbeiana expresses ranalexin and ranatuerin-1 that are absent in R. pipiens. Further detailed studies are necessary to elucidate the differences in the skin peptide defenses between ranid species that are resistant to chytridiomycosis and those that are susceptible. These studies may provide an explanation for the relative resistance of one ranid species in comparison with another.

Populations of R. pipiens and other anuran species appear to be susceptible to chytridiomycosis either at high altitude (Rocky Mountains and Sierra Nevada) [14] and [15], in cold climates [16] and [17] or in warm climates during cold snaps in winter [13], [19] and [34]. R. pipiens hibernate at temperatures near freezing [35] and [36]. We hypothesized that antimicrobial activity of the peptides might also be reduced by cold temperature, because other immune defense parameters (numbers of circulating lymphocytes, proliferation of lymphocytes in response to mitogens, and serum complement levels) are reduced during exposure to cold [37] and [38]. However, we observed that the effective inhibition of growth of zoospores by esculentin-2P and ranatuerin-2P occurred at both 10 and 22 °C, and both peptides were more effective at the lower temperature. We interpret the increased effectiveness of peptides at 10 °C to mean that the peptides can bind and disrupt cell membranes at 10 °C as effectively as at 22 °C, and we speculate that at the higher temperature, the pathogen grows more rapidly and may produce factors such as proteases that interfere with peptide function. Even though cold does not affect antimicrobial activity in vitro, it is very likely to decrease the rate of synthesis and secretion of the antimicrobial peptides by the skin. In nature, there is some evidence that the cold temperature can inhibit synthesis of the peptides, limiting their effectiveness [39]. We speculate that the species with antimicrobial peptides effective against the chytrid do have a natural defensive capability that may hold the infection in check. However, environmental factors, such as cold temperatures, UV-B or toxic chemicals may inhibit synthesis and release. Stress that elevates glucocorticoids may inhibit synthesis [40], allowing the chytrid to survive and eventually kill the amphibians. Further studies are urgently needed to determine whether environmental factors inhibit natural peptide defenses.