Activities of Temporin Family Peptides against the Chytrid Fungus (Batrachochytrium dendrobatidis) Associated with Global Amphibian Declines

Abstract

Temporin A and structurally related peptides produced in amphibian dermal granular glands and in wasp venom were tested for growth inhibition of Batrachochytrium dendrobatidis, a pathogen associated with global amphibian declines. Two natural amphibian temporins, a wasp temporin, and six synthetic analogs effectively inhibited growth. Differences in potency due to amino acid substitution suggest that ability to penetrate membranes and form an α-helical structure is important for their effectiveness against this pathogen.

Global amphibian declines (reviewed in references 4, 6, and 10) have been associated with a novel chytrid fungus (Batrachochytrium dendrobatidis) that infects the skin (2, 3, 15, 19). Among the innate defenses employed by amphibians to resist skin infections is the production of antimicrobial peptides in dermal granular glands (reviewed in references 17, 24, and 33). Our previous studies have shown that 17 antimicrobial peptides representing 11 families of peptides inhibit growth of B. dendrobatidis. Six are active against a second fungal pathogen (Basidiobolus ranarum) associated with declines of the Wyoming toad, and two can inhibit in vitro plaque formation by iridoviruses pathogenic to amphibians and fish (5, 20-22). One of the peptides with antichytrid activity was a member of the temporin family (temporin 1Ob from Rana ornativentris) (20). For the present study, we have determined the antichytrid activity of several other members of this peptide family. Temporins, originally isolated from the European red frog, Rana temporaria (25), and later isolated from a number of North American and European ranid species (7-9, 11, 12, 23), are linear peptides containing 10 to 14 amino acids. All are α-amidated at their carboxyl-terminal ends. They occur not only in the skin secretions of a number of amphibians but also in wasp venom (reviewed in reference 27). They are most active against gram-positive bacteria, but some show activity against gram-negative bacteria and two have been shown to be active against the fungus Candida albicans (7-9, 11, 12, 25, 28). We show here that three additional natural temporins (two from amphibians and one from a wasp) and six synthetic analogs of temporin A (TA) can inhibit growth of B. dendrobatidis. Differences in potency due to amino acid substitution suggest that the likely mechanism of action against this pathogen is attachment to the membrane, followed by the folding of the temporins into an α-helical structure that facilitates disruption of the membrane. The effectiveness of the peptides may also depend on their ability to resist fungal proteases.

B. dendrobatidis was cultured and peptide inhibition of chytrid growth was assayed as previously described (20-22). MIC is defined as the lowest concentration at which no growth was detectable. The peptides examined in these experiments are listed in Table 1. Shown are the amino acid sequences, species of origin, numbers of amino acids, net charges at pH 7, and percentages of hydrophobic residues. All peptides were synthesized by solid-phase techniques using 9-fluorenylmethoxy carbonyl chemistry as previously described (28). The peptides were purified by reverse-phase high-pressure liquid chromatography and characterized by amino acid analysis and electrospray ionization mass spectrometry. All peptides were dissolved in glass-distilled water, filter sterilized, frozen in small aliquots at high concentration and used at various dilutions for culture.

TABLE 1.

Natural and synthetic temporin-like peptides^a

Peptide	Amino acid sequencef	No. of amino acids	NCg	%Hh
TAb	F-L-P-L-I-G-R-V-L-S-G-I-L-NH2	13	+2	61
T-1Pc	F-L-P-I-V-G-K-L-L-S-G-L-L-NH2	13	+2	61
Rana-6d	F-I-S-A-I-A-S-M-L-G-K-F-L-NH2	13	+2	69
VesCP-Me	F-L-P-I-I-G-K-L-L-S-G-L-L-NH2	13	+2	61
DTA	f-l-p-l-i-g-r-v-l-s-g-i-l-NH2	13	+2	61
LDTA	F-l-P-l-I-G-R-v-L-s-G-i-L-NH2	13	+2	61
W1-TA	W-L-P-L-I-G-R-V-L-S-G-I-L-NH2	13	+2	54
I4G10	F-L-P-I-I-A-S-L-L-G-K-L-L-NH2	13	+2	69
I4S10	F-L-P-I-I-A-S-L-L-S-K-L-L-NH2	13	+2	69
CATA	K-W-K-L-F-K-K-L-P-L-I-G-R-V-L-NH2	15	+6	47

^aAdapted from Table 2 of reference 28.

^bR. temporaria.

^cRana pipiens.

^dRana catesbeiana.

^eVespa mandarinia.

^fNH₂, amide (-CO-NH₂). d amino acids are in lowercase; Gly is nonchiral.

^gNC, net charge at pH 7.

^h%H, percentage of hydrophobic residues (average, 61%) based on the scale of Kyte and Doolittle (14).

The activities of three natural amphibian temporins as inhibitors of growth of mature cells of B. dendrobatidis are shown in Fig. 1a to c, and MICs are shown in Table 2. TA and temporin 1-P (T-1P) significantly inhibited growth at concentrations above 25 μM (35 μg/ml). Ranatuerin-6 (Rana-6) showed little activity against mature chytrid cells but did weakly inhibit growth of zoospores at a concentration above 50 μM (70 μg/ml) (data not shown). It is not yet clear why Rana-6 had significantly reduced activity in comparison with other members of the temporin family. All of the amino acid differences are conservative changes with respect to the hydrophobicity or hydrophilicity of the position.

FIG. 1.

Growth inhibition of B. dendrobatidis at 4 days of culture by TA (a), T-1P (b), Rana-6 (c), DTA (d), LDTA (e), and VesCP-M (f). Each data point represents the mean ± standard error (SE) of three or more replicate wells. Where no error bar is shown, the SE was less than the diameter of the symbol. ∗, significantly less growth than positive controls (one-tailed Student t test; P ≤ 0.05). PF, paraformaldehyde. The results are representative of two assays for each peptide except LDTA and VesCP-M, which were assayed once. Each peptide was also assayed at least once for inhibition of zoospore-enriched cultures (see Table 2). MIC is the lowest concentration at which no growth was detected. O.D.₄₉₂, optical density at 492 nm.

TABLE 2.

MICs necessary to completely inhibit growth of B. dendrobatidis (mature cells or zoospores)

Peptide	Species of origin	MIC against mature cells at day 4		MIC against zoospores at day 5
		μM	μg/ml	μM	μg/ml
TA	Rana temporaria	100	140	66	92
T-1P	Rana pipiens	50	68	50	68
Rana-6	Rana catesbeiana	>100	>140	>100	>140
DTA	Synthetic	6.25	8.7	6.25	8.7
LDTA	Synthetic	138	193	>25	>35
VesCP-M	Vespa mandarinia	25	34.6	6.25	8.6
W1-TA	Synthetic	36	52	25	36
I4G10	Synthetic	57	80	29	40
I4S10	Synthetic	100	143	32	46
CATA	Synthetic	100	184	100	184

TA is composed entirely of l amino acids. In comparison, an all-d isomer, designated DTA, had significantly greater potency in the inhibition of growth of B. dendrobatidis (Fig. 1d and Table 2). Antimicrobial peptides are thought to act independently of specific membrane receptors, and all-d isomers are predicted to have activities very similar to those of the naturally occurring all-l-isomer forms (29). We speculate that the enhanced activity of DTA may be due to its stability against proteolytic enzymes produced by B. dendrobatidis. Although specific proteolytic enzymes have not been identified or characterized for this species, preliminary studies suggest that secreted products of growing B. dendrobatidis can degrade casein and gelatin (J. Piotrowski, S. Annis, and J. E. Longcore, unpublished observations).

The TA isomer designated LDTA has the same amino acid sequence as TA and DTA, but alternate amino acids (amino acids 2, 4, 8, 10, and 12) are of the d configuration. Natural TA has been shown to adopt an α-helical conformation in aqueous solutions of trifluoroethanol, a model system for the hydrophobic membrane environment (31). LDTA is predicted to be incapable of forming an α-helix (28), although it might form a larger-diameter helix such as that found in gramicidin A (13). An α-helical conformation has been shown to be required for the antimicrobial activities of many antimicrobial peptides. Therefore, the activity of LDTA is predicted to be less than that of TA, and this prediction was confirmed experimentally (Fig. 1e and Table 2). This is persuasive evidence that the ability of temporins to assume an α-helical conformation is important for their activity against B. dendrobatidis.

Another natural analog of TA, which is found in the venom of the wasp Vespa mandarinia (VesCP-M) (27, 28), and a second, synthetic analog of TA in which the amino-terminal phenylalanine is replaced by tryptophan (W1-TA) had significantly better activity for inhibiting the growth of B. dendrobatidis than TA (Fig. 1f and Table 2). Similar results were obtained with bacteria as targets (30). Substitution of the amino-terminal tryptophan for phenylalanine in the analog W1-TA may enable the peptide to insert itself into membranes more freely.

The peptide designated CATA is a hybrid composed of the amino-terminal sequence (residues 1 to 7) of cecropin A, an antimicrobial peptide originally isolated from the hemolymph of pupae of the silk moth Hyalophora cecropia (26), and residues 2 to 9 of TA. Its antimicrobial activity profile and MIC against B. dendrobatidis are similar to those of TA and other natural amphibian temporins (Table 2).

The analogs I4G10 and I4S10 were synthesized to match two possible consensus amino acid sequences derived by comparison of the amino acid sequences of 30 temporin-like peptides found in the skin secretions of various amphibian species (32). Their antimicrobial activity profiles and MICs against B. dendrobatidis are very similar to those of the natural temporins (Table 2).

Studies of replication and transmission of B. dendrobatidis suggest that spread of infection from one area of the skin to another or to a new host occurs by means of motile zoospores. Thus, the effectiveness of each antimicrobial peptide against the zoospore stage was tested to determine whether the peptide could inhibit infections. A comparison of the MICs for purified zoospores versus those for mature cells shows that, for most peptides tested, zoospores were completely inhibited at a concentration that was an average of 66% of that necessary to completely inhibit mature cells (Table 2).

The mechanism(s) of action of these peptides is unknown. Possible mechanisms include formation of pores in microbial bilayer membranes and membrane solubilization by a “carpet-like” mechanism that leads to a disruption in the internal homeostasis of the cell and death (1, 18). The first step in such a process would be the binding of the peptide to the cell membrane. For B. dendrobatidis, the predominant mode of binding does not appear to be via electrostatic interactions between the positively charged peptides and negatively charged membrane phospholipids because CATA, the peptide with the greatest net positive charge (+6), was not as active as many of the other peptides of Table 1 that are less positively charged (+2). It may be that the major mode of binding to the membrane of this organism is through hydrophobic interactions, as has been suggested for bacteria (16).

Because amphibian species are threatened on a global scale, further research is urgently needed to understand the role of temporins and other antimicrobial peptides in the innate defense capacity of amphibians.