Introduction

Small molecules derived from natural sources play a key role in human welfare, serving as drugs or drug precursors with useful pharmaceutical properties [1]–[2]. Among these natural products, antibiotics are particularly important, abating human suffering and death from infectious disease [1], [3]. The ability of microbes to rapidly evolve resistance compromises the efficacy of antibiotics, necessitating the continuous discovery of structurally novel natural products with antimicrobial properties [e.g., [3]–[5]]. However, traditional sources for discovering novel antibiotics appear to be largely exhausted [3], [6]. For example, contemporary bioprospecting of soil Actinobacteria, the most significant source of new antibiotics in the twentieth century, largely results in the rediscovery of already known compounds [e.g., [3], [6]–[7]]. Current strategies for addressing the urgent need for new antibiotics include metabolic engineering, synthetic chemistry, and genomic or metagenomic approaches [2], [7]–[13]. Another approach is identifying sources of microbes that have not been explored for their potential natural products [5]–[6], [14]–[21].

Symbiotic microbes may represent a particularly promising source because microbial symbioses are widespread [e.g., [22]], largely unexplored for natural products [cf. [5]; but see [16]], and often involve the exchange of small molecules between symbionts and hosts [cf. [16]], including compounds mediating host defense [e.g., [11], [17]–[19]]. Examples of novel small molecules from symbiotic microbes include pederins produced by uncultured endosymbiont bacteria associated with beetles and sponges [8], [11], [23], stilbene-derivatives produced by Photorhabdus luminescens associating with Heterorhabditis nematodes [24]–[25], and a number of compounds with antimicrobial properties derived from bacteria associated with bryozoans [reviewed by [26]]. Among symbiotic associations, insect-microbe symbioses that involve Actinobacteria may be of particular interest in natural product discovery. Recently, a Streptomyces-derived novel compound was discovered, mycangimycin, that appears to be responsible for the selective inhibition of an antagonistic fungus of the mutualistic fungus associated with Dendroctonus frontalis beetles [17], [19]. Similarly, the novel compound dentigerumycin was obtained from a fungus-growing ant-associated actinobacterium (genus Pseudonocardia), which aids in the protection of the ants' mutualistic fungus from a specialized parasite [18], [27], [28].

Here we examine if solitary wasp-associated Actinobacteria can serve as a valuable source of novel natural products. Specifically, we explore whether two species of solitary mud dauber wasps harbor Actinobacteria, and, if so, whether these bacteria produce bioactive natural products of ecological and pharmaceutical interests. We focused on these wasps, in part because of the recently discovered symbiosis between Streptomyces and 27 species of congeneric solitary European beewolf (genus Philanthus) wasps, where the bacteria are believed to produce antibiotics inhibiting fungal pathogens of the wasp larvae [29]–[31]. ‘Mud daubers’, as they are commonly referred to, belong to the families Sphecidae and Crabronidae, with the most common species in the US being the black and yellow mud dauber (Sceliphron caementarium, family Sphecidae) and the blue mud dauber (Chalybion californicum, family Sphecidae) (Fig. 1) [e.g., [32]]. Sceliphron caementarium mud daubers construct nests that are typically tubular and partitioned into multiple cells using mud collected from water puddles (Fig. 1) [e.g., [33]]. Individual cells contain a mud dauber egg and paralyzed prey (typically spiders), which serves as food for the developing larvae [32]. In contrast, C. californicum typically takes over an already established mud dauber nest, empties the contents of individual cells, and places its own paralyzed prey and eggs. Regardless of their life cycle, both mud dauber species frequently interact with the soil and prey insects, exposing both adults and brood to potentially pathogenic microbes.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. The two mud dauber species from which Streptomyces were obtained.

A) shows a black and yellow mud dauber (Sceliphron caementarium) (courtesy of Linda Hendry), B) shows a blue mud dauber (Chalybion californicum) (courtesy of Patrick Belardo), and C) shows a nest of Sceliphron caementarium (courtesy of Jay King).

https://doi.org/10.1371/journal.pone.0016763.g001

Here we demonstrate that enrichment isolations for Streptomyces spp. from two species of mud dauber wasps can provide morphologically, phylogenetically, and chemically diverse Actinobacteria. Chemical characterization of the secondary metabolites produced revealed distinct and structurally diverse secondary metabolites, including a novel polyunsaturated and polyoxygenated macrocyclic lactam. Further, we assess the antibiotic properties of Streptomyces strains in antifungal and antibacterial Petri plate bioassays. We discuss our findings in relation to the potential role of Streptomyces in associations with solitary wasps, and suggest that our findings support that insect-associated Actinobacteria are a valuable source for novel natural products.

Materials and Methods

Phylogenetic placement of Streptomyces strains

DNA extraction, PCR amplification and DNA sequencing of bacterial isolates were carried out according to procedures in Poulsen et al. [34]. Near full length sequences of 16S rRNA were obtained from PCR using universal primers [27F and 1492R; [35]]. Positive bands on a 1.5% agarose gel were direct-sequenced at the University of Wisconsin-Madison Biotechnology Center (www.biotech.wisc.edu). Sequences were corrected for mismatches using Sequencher 4.6 for Windows (Gene Codes Corporation, Ann Arbor, MI), and the sequences were submitted to GenBank (Accession numbers: GQ351298-GQ351312) (Fig. S1). The wasp-associated strains were placed phylogenetically by performing a phylogenetic analysis that included the closest matches for each strain obtained in a type-strain search in the Ribosomal Database Project (http://rdp.cme.msu.edu/; [36] (Figs. 2; S1). The phylogeny was generated using PAUP* [37], after automatic and manual alignment using Clustal X [38] and MacClade [39], respectively.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 2. Phylogenetic and chemical diversity of mud dauber-associated Streptomyces.

The phylogenetic placement based on partial 16S rDNA sequences of the 15 Streptomyces strains examined for their secondary metabolites. The tree is rooted with Kribbela (GenBank accession number: GQ351313) and Nocardioides (GenBank accession numbers: AF004988). Mycangimycin-producing Streptomyces strains isolated from Dendroctonus frontalis (SPB071 and SPB074) and three Streptomyces strains (S. philanthi) from European Beewolves (Philanthus venustus, P. gibbosus, and P. triangulum) are included [17], [29]. Branch values indicate bootstrap support (>50 are given) of 1000 pseudoreplicates under Maximum Parsimony. The right panel shows the structures and names of the secondary metabolites identified, including the structurally novel macrocyclic lactam, scheliphrolactam. Strain names indicate the wasp species isolation source, which, as well as the compound names, are color-coded according to the compounds produced. Compounds could not be identified from strains indicated in grey font.

https://doi.org/10.1371/journal.pone.0016763.g002

Antifungal and antibacterial activities of Streptomyces-derived secondary metabolites

Antifungal and antibacterial activities were tested in two Petri plate bioassay experiments. The first experiment tested the antifungal effects of the 15 Streptomyces strains against 16 diverse fungi, including entomopathogens of insects, as well as a set of fungi isolated from mud daubers and Sirex wood wasps (Fig. 3). The identity of all fungi involved was confirmed by partial sequencing of 18S rDNA (not reported). Pure-culture Streptomyces were point inoculated on YMEA plates, left for three weeks until reaching a diameter of ca. 1.5 cm, after which the fungus was point inoculated at the edge of the plate. When a clear zone of inhibition (ZOI) had formed, typically within 2–3 weeks after fungal inoculation, the minimum distance was measured [cf. [28]]. Three replicates were performed for each pairing. The antibacterial assay was performed by pairing the 15 Streptomyces strains against each other in all possible combinations, thereby examining antibacterial properties against bacteria known to be present with the wasps. The inhibiting strain was point inoculated in the center of a Petri plate containing YMEA, and after three weeks the second strain was inoculated to the entire unoccupied plate area by applying 200 µl of autoclaved water containing 800–1200 cells/µl. The plates were checked bi-weekly until the second strain had either grown to fill the Petri plate or a distinct zone of inhibition (ZOI) had established [cf. Fig. 1 in [34]], which was then measured. Three replicates were performed for each pairing.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 3. Bioassay evaluation of the antibiotic activity of the Streptomyces isolates from mud dauber wasps.

Antimicrobial properties of Streptomyces in Petri plate boioassay tests against diverse fungi (left) and Streptomyces bacteria (right). Boxes represent average ZOI (n = 3) of a given pairing and different colors indicate the degree of inhibition. For Streptomyces-fungus bioassay: White: ZOI = 0 mm, Light yellow: ZOI = 0.01–1.0 mm, Yellow: ZOI = 1.01–2.0 mm, yellow-orange: ZOI = 2.01–3.0 mm, light orange: ZOI = 3.01–4.0, and orange: ZOI >4.01 mm. For Streptomyces-Streptomyces bioassay: White: ZOI = 0 mm, Light yellow: ZOI = 0.1–2.0 mm, Yellow: ZOI = 2.1–4.0 mm, yellow-orange: ZOI = 4.1–6.0 mm, light orange: ZOI = 6.1–8.0, and orange: ZOI >8.1 mm.

https://doi.org/10.1371/journal.pone.0016763.g003

Results

Discussion

Our findings show that diverse Actinobacteria in the genus Streptomyces can be readily isolated from a few individuals of a single insect group. Isolations from eight C. californicum and 25 S. caementarium yielded more than 200 strains of Streptomyces representing 24 distinct morphotypes. Fourteen of these morphotypes were isolated from individuals belonging to both wasp species, and five were unique to each wasp species. On average, individual S. caementarium wasps yielded more morphotypes than C. californicum; however, this is likely because three sub-cultures were performed per C. californicum isolation plate versus eight per S. caementarium. Our phylogenetic analysis of the 15 representative strains chosen for chemical analyses indicated that these strains represent a diverse collection of bacteria distributed across the genus Streptomyces (Figs. 2; S1), despite 16S rDNA providing only limited phylogenetic resolution [e.g., [49]].

Our chemical analyses of the 15 Streptomyces strains revealed the production of a diverse collection of compounds. Ten of the 15 strains produce diffusible secondary metabolites from six structural classes: antimycins, bafilomycin A1 and B1, daunomycin, mycangimycin, streptazolin and streptazon B, and the previously unknown macrocyclic lactam, sceliphrolactam. As expected, different Streptomyces strains belonging to the same ’species’ can produce different secondary metabolites, while taxonomically diverse Streptomyces strains can produce identical metabolites. Embedded within the finding of diverse chemical compounds from the isolated Streptomyces bacteria was the discovery of sceliphrolactam: a structurally novel macrocyclic lactam produced by two genetically distinct Streptomyces strains (e113 and e122) (Fig. 2). Sceliphrolactam bears polyene and polyol moieties and could act as an antifungal by destabilizing the fungal cell membrane functions [cf., [50]]. Interestingly, none of the Streptomyces secreted compounds from more than one class in vitro, suggesting that individual strains produce a limited set of compounds within a single compound group under these conditions (Fig. 2).

Our Petri plate bioassay experiments against fungi and bacteria confirmed that different secondary metabolites are secreted by the Streptomyces strains, with strains varying in their antibacterial and antifungal activity (Fig. 3). Some of these differences were expected based on the known properties of previously reported compounds. For example, antimycins are potent inducers of cellular apoptosis in hepatocytes and bind to the hydrophobic groove of Bcl-2/Bcl-x proteins on the surface of mitochondria [51], which likely explains the antifungal activity (Fig. 3). Bafilomycins have both antifungal and cytotoxic properties [48], and they are potent inhibitors of vacuolar H⁺-ATPase [52]. Daunomycin inhibits DNA topoisomerase II and it thereby induces a cytotoxic antibiotic effect [53]. Mycangimycin is a selective antifungal agent, only recently obtained from a southern pine beetle-associated Streptomyces (Fig. 2) [17], [19]. While streptazolin itself showed limited antimicrobial activities [54], some of its derivatives, such as 3,9-dihydrostreptazolin, show enhanced antimicrobial and cytotoxic activities [43], which is consistent with the results of our assays. The strains producing these compounds mainly displayed weak or no inhibition of other Actinobacteria in bioassay, while their antifungal activity appeared restricted to a few Aspergillus strains and one Trichoderma strain (Fig. 3), suggesting narrow antimicrobial activity. Our antifungal assay revealed that the Streptomyces strains producing sceliphrolactam have antifungal properties (e.g., inhibiting certain strains of Beauvaria); however, they generally displayed stronger levels of inhibition of other Streptomyces bacteria than fungi (Fig. 3). Differences between strains known to produce the same compounds were observed, likely due to differences in concentrations of the compounds produced and/or due to differences in the relative composition of secretions in strains producing multiple compounds. Contrary to our expectations, we did observe antimicrobial properties of Streptomyces strains for which we did not detect secondary metabolites, suggesting that compounds indeed are produced and secreted by these strains (Fig. 3). Possible reasons for not detecting these compounds include secretion in concentrations below our detection threshold, or that compounds are not produced in pure culture, but elicited by the presence of another microbe.

The role of Streptomyces in the associations with mud dauber wasps remains enigmatic; they may be symbionts associated with the wasps or transient microbes picked up by wasps during soil interactions. Nevertheless, the tight association of solitary wasps with the soil, harboring an abundance of parasitic microbes, implies a potential benefit for the mud daubers of engaging in associations with antibiotic-producing bacteria. If so, Streptomyces-derived compounds with antimicrobial properties could serve to protect the wasps, the wasp brood, and/or help preserve the prey. Streptomyces presence on all body parts of all wasps sampled is a first indication that such a symbiotic association may be present; however, documenting the costs and benefits to the partners involved in this potential mutualism will be required [cf. [55]]. The only well-established Streptomyces-solitary wasp association involves the maintenance of the bacteria in antennal structures of European beewolf species, which benefit from the association by deriving antibiotic compounds that protect the wasp brood [29]-[31]. The mud daubers included in our study do not appear to have analogous antennal Streptomyces, and all wasp body parts yielded isolates, suggesting that Streptomyces are present at different, or less specialized, locations on the wasps.

If mud daubers benefit from Actinobacteria-derived compounds with antibiotic properties, they could either associate loosely with phylogenetically and chemically diverse Streptomyces, or more specifically with potentially coevolving Streptomyces. In fungus-growing ants, southern pine beetle insect-Actinobacteria, and European beewolves, one or a few specific Actinobacteria appear to associate with the host [17]–[19], [30]. In the European beewolf-Streptomyces association, the insect achieves protection through individual bacteria species secreting a mixture of antibiotic compounds [31]. In contrast, the diversity of Streptomyces found in this study suggests that the insect potentially takes advantage of multiple Streptomyces strains secreting diverse antibiotic compounds. Irrespective of whether mud daubers engage in specific defensive mutualisms with Streptomyces or not, the omnipresence of Streptomyces on all body parts of all wasps examined, coupled with the antibiotic properties of the bacteria-derived compounds, suggest that the presence of these Streptomyces sp. nevertheless affects the microbial communities associated with the wasps. This may include that compounds produced aid in competition with other bacteria, which is known to be frequent and important in shaping microbial communities [e.g., [56]].

Our findings of actinobacterial associations and identification of diverse bacterial secondary metabolites, especially the discovery of the novel macrocyclic lactam, sceliphrolactam, from mud dauber wasps demonstrate that actinobacterial communities from an insect host can aid in the search for novel natural products with antibiotic properties. Our analyses included only a subset of the Streptomyces we obtained from this insect group, suggesting that more compounds of natural product interest await discovery. Our search yielded a novel compound from chemical characterization of only 15 Streptomyces isolates, whereas searches for novel antibiotics among hundreds of soil Actinobacteria typically result in mainly rediscoveries of already known compounds [cf., [4]–[5], [7]]. Moreover, recent findings suggest that defensive mutualisms between insects and Actinobacteria are far more common than previously assumed [17]–[19], [29]–[30], [57], suggesting that insect-Actinobacteria associations are a particularly promising group from which novel natural products can be elicited. Future research should continue to identify novel insect-Actinobacteria associations, in addition to symbioses with bacteria beyond Actinobacteria [cf. [8], [16], [21]], including establishing the roles of the microbes and their secondary metabolites in the symbiotic systems. Such identification will both provide novel insight into defensive mutualisms and host-symbiont interactions in general, but also lead to the discovery of novel associations and novel secondary metabolites of natural product interest.

Supporting Information

Acknowledgments

We thank S. Adams and R.A. Steffensen for assistance with DNA extraction and PCR amplification, N. Vitt Meyling and J. Eilenberg for M. anisopliae and B. bassiana strains, A. Adams for the Trichoderma strain, B. Pahnke and H. Horn for bioassay assistance, A. Adams, S. Adams, F. Aylward, E. Caldera, J. Gelderman, K. Grubbs, S. Marsh, R.A. Steffensen, G. Suen, and L.A. Schwab for comments on a draft of the manuscript.