Symbiotic bacteria can produce secondary metabolites and volatile compounds that contribute to amphibian skin defense. Some of these symbionts have been used as probiotics to treat or prevent the emerging disease chytridiomycosis. We examined 20 amphibian cutaneous bacteria for the production of prodigiosin or violacein, brightly colored defense compounds that pigment the bacteria and have characteristic spectroscopic properties making them readily detectable, and evaluated the antifungal activity of these compounds. We detected violacein from all six isolates of Janthinobacterium lividum on frogs from the USA, Switzerland, and on captive frogs originally from Panama. We detected prodigiosin from five isolates of Serratia plymuthica or S. marcescens, but not from four isolates of S. fonticola or S. liquefaciens. All J. lividum isolates produced violacein when visibly purple, while prodigiosin was only detected on visibly red Serratia isolates. When applied to cultures of chytrid fungi Batrachochytrium dendrobatidis (Bd) and B. salamandrivorans (Bsal), prodigiosin caused significant growth inhibition, with minimal inhibitory concentrations (MIC) of 10 and 50 μM, respectively. Violacein showed a MIC of 15 μM against both fungi and was slightly more active against Bsal than Bd at lower concentrations. Although neither violacein nor prodigiosin showed aerosol activity and is not considered a volatile organic compound (VOC), J. lividum and several Serratia isolates did produce antifungal VOCs. White Serratia isolates with undetectable prodigiosin levels could still inhibit Bd growth indicating additional antifungal compounds in their chemical arsenals. Similarly, J. lividum can produce antifungal compounds such as indole-3-carboxaldehyde in addition to violacein, and isolates are not always purple, or turn purple under certain growth conditions. When Serratia isolates were grown in the presence of cell-free supernatant (CFS) from the fungi, CFS from Bd inhibited growth of the prodigiosin-producing isolates, perhaps indicative of an evolutionary arms race; Bsal CFS did not inhibit bacterial growth. In contrast, growth of one J. lividum isolate was facilitated by CFS from both fungi. Isolates that grow and continue to produce antifungal compounds in the presence of pathogens may represent promising probiotics for amphibians infected or at risk of chytridiomycosis. In a global analysis, 89% of tested Serratia isolates and 82% of J. lividum isolates were capable of inhibiting Bd and these have been reported from anurans and caudates from five continents, indicating their widespread distribution and potential for host benefit.

Management of hyper-virulent generalist pathogens is an emergent global challenge, yet for most disease systems we lack a basic understanding as to why some host species suffer mass mortalities, while others resist epizootics. We studied two sympatric species of frogs from the Colombian Andes, which coexist with the amphibian pathogen Batrachochytrium dendrobatidis (Bd), to understand why some species did not succumb to the infection. We found high Bd prevalence in juveniles for both species, yet infection intensities remained low. We also found that bacterial community composition and host defense peptides are specific to amphibian life stages. We detected abundant Bd-inhibitory skin bacteria across life stages and Bd-inhibitory defense peptides post-metamorphosis in both species. Bd-inhibitory bacteria were proportionally more abundant in adults of both species than in earlier developmental stages. We tested for activity of peptides against the skin microbiota and found that in general peptides did not negatively affect bacterial growth and in some instances facilitated growth. Our results suggest that symbiotic bacteria and antimicrobial peptides may be co-selected for, and that together they contribute to the ability of Andean amphibian species to coexist with the global pandemic lineage of Bd.