Microbiota composition of the dorsal patch of reproductive male Leptonycteris yerbabuenae

Bacteria and other types of microbes interact with their hosts in several ways, including metabolic pathways, development, and complex behavioral processes such as mate recognition. During the mating season, adult males of the lesser long-nosed agave pollinator bat Leptonycteris yerbabuenae (Phyllostomidae: Glossophaginae) develop a structure called the dorsal patch, which is located in the interscapular region and may play a role in kin recognition and mate selection. Using high-throughput sequencing of the V4 region of the 16S rRNA gene, we identified a total of 2,847 microbial phylotypes in the dorsal patches of eleven specimens. Twenty-six phylotypes were shared among all the patches, accounting for 30 to 75% of their relative abundance. These shared bacteria are distributed among 13 families, 10 orders, 6 classes and 3 phyla. Two of these common bacterial components of the dorsal patch are Lactococcus and Streptococcus. Some of them—Helcococcus, Aggregatibacter, Enterococcus, and Corynebacteriaceae—include bacteria with pathogenic potential. Half of the shared phylotypes belong to Gallicola, Anaerococcus, Peptoniphilus, Proteus, Staphylococcus, Clostridium, and Peptostreptococcus and specialize in fatty acid production through fermentative processes. This work lays the basis for future symbiotic microbe studies focused on communication and reproduction strategies in wildlife.


Introduction
The microbiome is the assemblage of archaea, bacteria, viruses and other microorganisms associated with any multicellular organism (the host) in a particular environment [1,2]. The microbiome is a vital component in the evolution of the host, and in vertebrates it plays essential roles in almost all of the organism's functions [3][4][5][6][7]. The relationships between the animal host and its microbiome are, for the most part, mutually beneficial [8]. Gut and skin a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 microbiota have coevolved with their hosts, and this mutualism has led to the evolution of intra-and interspecific interactions [9,10]. Microbiota associated with skin and glands in mammals are and diverse, and host species is an important predictor of the microbial community assemblage [11]. It has been demonstrated that microbial communities play a role in the production of odors and chemical signals, which are important in species recognition, kin recognition, and mate choice [12][13][14][15]. The interaction between host and microbiome can thus alter the host's social behavior at various points during its life cycle [14,15].
The role of bacteria in the processes of fermentation and production of volatile fatty acids has been previously reported in studies of canids and bats [16,17]. These short chain fatty acids, produced by bacteria, are the precursors of chemical signals, known as scents, which influence the intraspecific interactions of many mammals. Bacteria associated with the skin contribute to the host's scent through the direct production of odors and the fermentation of organic compounds produced endogenously by the host, which is known as the "fermentative hypothesis" [18,19]. The fermentative hypothesis is based on two main assumptions: first, that volatile odorants are produced by bacteria that colonize mammal scent glands and ephitelial tissues and second, that individual "odor prints" are generated by differences among individuals in the composition of their bacterial communities [20].
Scent production as a consequence of bacterial fermentation has been reported in several bat species [17,21,22], most often among reproductive males [22][23][24]. Some examples of bats that possess odor-producing structures colonized by bacteria are the piscivorous Noctilio leporinus (Noctilionidae), which have a typical strong odor associated with Staphylococcus aureus [17], male sac-winged bats, Saccopteryx bilineata (Emballonuridae), which have a scent pouch that contains an odoriferous liquid used in courtship [17,25], and the big brown bat Eptesicus fuscus (Vespertilionidae), in which roost-mates have a shared odor signature [21]. Adult males of the nectar-feeding bats Leptonycteris curasoae and Leptonycteris yerbabuenae develop an oderiferous "dorsal patch", a temporary structure in the interscapular region that contains fatty acids and signals mating readiness [22,26,27].
Male L. yerbabuenae develop the odoriferous dorsal patch for a short time during the breeding season, using their forelimbs to spread fluids from the anus, penis, and mouth onto their backs in a behavior known as "smearing" [22,28] (Fig 1). The fact that all males with dorsal patches have enlarged testes, and that males with small testes do not develop this trait, suggests a strong association between dorsal patch development and sexual maturity and possibly readiness for reproduction [22][23][24]27]. This dorsal patch has a strong odor and a complex chemical profile [22,27], made of compounds that have been reported to be important during the reproductive season of L.yerbabuenae and Leptonycteris curasoae in female attraction [28,29]. Thus, overall, the available evidence suggests that the dorsal patch is involved in female mate choice mediated through odor [22,24,26,27]. Males with dorsal patches had fewer ectoparasites than those without them, suggesting that this structure might function as a dual signal (odoriferous and visual) of mating readiness and health status [22,[26][27][28]. It has been hypothesized that the odor of the dorsal patch is an honest signal of health status [26].
In this study, we aimed to evaluate the microbial diversity associated with the dorsal patch of reproductive L. yerbabuenae males using a high throughput 16S rRNA amplicon sequencing approach.

Bat sampling
This study was conducted during the mating season in June 2015. Bats were captured using the methods described by Kunz et al. [35], using 12 m mist nets (Avinet, Dryden, New York, USA) placed at the entrance of the cave from 18:00h in the evening to 06:00h the next day, a sampling effort of 12 hours.
Interscapular dorsal patch samples (n = 11, named with the prefix Oax) were obtained following the guidelines of the American Society of Mammalogists for capture, handling and care of mammals [36,37]. To minimize animal suffering and distress, each animal was processed by experts only, ensuring effective and harmless handling. No anesthetics were administered for chemical immobilization, as they were deemed unnecessary and would only increase the mortality risk. Samples of the crust and fur of the interscapular patch were taken using gloves and sterile surgical calipers and scissors. A 0.25 cm 2 area of the formed crust and hair was cut without touching the skin, and placed into sterile Eppendorf tubes (1.5 ml). Samples were frozen until processing using a field liquid nitrogen dewar.

Ethics statement
Samples were taken from wild bats that were released in the same area as capture, causing no apparent harm to individuals. Leptonycteris yerbabuenae is not under federal protection by Mexican law (NOM-059-SEMARNAT-2010). Scientific collection activities were carried out under a scientific collection permit number granted by the Mexican Subsecretary of the Environment and Natural Resources (SEMARNAT), number FAUT-0231, SGPA/DGVS/05780/ 15. SEMARNAT specifically approved and authorized the tissue sampling methods under this collection permit. Laboratory activities were carried out in the Ecology Institute of the Universidad Nacional Autonoma de Mexico (UNAM), under the authorization of the Biosafety Commission of the Ecology Institute, UNAM; no specific permit was needed because only tissue and skin samples were used (no in vivo studies were included). All biosafety standard requirements from the Ecology Institute were satisfied.

DNA extraction
DNA was extracted using the DNeasy Blood & Tissue kit (Qiagen, Valencia, CA) with some modifications. Dorsal patch samples were diluted with 180 μl of Animal Tissue Lysis (ATL) extraction buffer and incubated with lysozyme A (30 mg/ml) and proteinase K (10 mg/ml). After enzymatic digestion, the manufacturer's protocol was followed. DNA was precipitated with 1 volume of chilled absolute ethanol and 0.1 volume of 3 M sodium acetate, then washed with 70% ethanol. Finally, DNA was eluted in 30 μl of molecular grade water and stored at -20˚C prior to PCR amplification.
Samples were amplified in triplicate using a PCR protocol that included an initial denaturation step at 95˚C (3 min), followed by 35 cycles of 95˚C (30 s), 52˚C (40 s) and 72˚C (90 s), followed by a final extension 72˚C (12 min). Triplicates were then pooled and purified using the SPRI magnetic bead, AgencourtAMPure XP PCR purification system (Beckman Coulter, Brea, CA, USA). The purified 16S rRNA fragments (~20 ng per sample) were sequenced on an Illumina MiSeq platform (Yale Center for Genome Analysis, CT, USA), generating~250 bp paired end reads. The sequence data are available from the NCBI Bioproject PRJNA496019; SUB4625080.
OTUs that were shared among samples were searched for their closest OTU using a BLAST search against the RefSeq-NCBI database. The three best hits for each OTU were aligned with MUSCLE [47], and a phylogenetic tree was built with PHYML software [48], using the GTR substitution model and 1000 bootstraps.

Results
A total of 17 male L. yerbabuenae individuals were captured, but only 11 samples met the minimal requirements for high-quality DNA extraction. Patches were found on mature males only and covered between 2 and 4 cm 2 of the interscapular area.
A subset of 26 OTUs was shared among all samples, and they collectively accounted for 30-75% of the microbial abundance of the dorsal patch per individual (Figs 7 and 8). Since these were all unassigned OTUs, we generated a phylogenetic tree was generated to associate these 26 shared bacteria to their closest relatives (RefSeq-NCBI Database) (Fig 9). These bacteria are distributed among 13 families; the four with the highest abundance were Tissierellaceae, Enterobacteriaceae, Pasteurellaceae and Streptococcaceae. Three Enterobacteriaceae genera were unassigned, along with one genus each from Pasteurellaceae, Nisseriaceae, Gemellaceae (see supporting information), and Planococcaceae. There was also one unassigned family.
The distribution pattern of 20 of these 26 bacteria was generally similar among samples. Only six showed differential abundance patterns: Pasteurellales OTUs 4466150 and 4353757, Clostridiales OTUs 97301, 768514, 3804335, 4349519 and Bacillales 4454737, 630141 and 4446058. OTUs for which we did not find a species-level relative were: Clostridum     Taxonomic affiliation of each OTU to its three best references (RefSeq-NCBI Database). The code names of related species sequence, shows "NCBI accession id"_"species name".

Dorsal patch core diversity
Although the dorsal patch of L. yerbabuenae males harbors a large microbial diversity, A subset of 26 OTUs was shared among all samples, and they collectively accounted for 30-75% of the microbial abundance of the dorsal patch per individual. Within this shared microbiome, 16 phylotypes are fermentative bacteria. These anaerobic bacteria are reported to be relevant in the production of short-chain fatty acids (SCFAs), which in several mammals have been shown to contribute to scent production that influences behavior, for example in bats (Saccopteryx bilineata [20] and Leptonycteris yerbabuenae [28]), hyenas (Crocuta crocuta and Hyaena hyaena [49]), and lions (Panthera leo [50]). Most of the shared bacteria from the dorsal patch belong to genera and families found and described in humans as part of the skin microbiome (Finegoldia, Pasteurellaceae), associated with wounds or infections (Helcococcus, Enterococcus) [49], or with the production of fermented products or volatile fatty acids (Peptostreptococcus, Anaerococcus, Gallicola, Peptostreptococcus, Lactococcus, Planococcaceae).
The shared OTUs are shown in the phylogenetic reconstruction, positioned nearest to their closest known relative (Fig 9). Two of the bacteria from the genus Lactococcus and Streptococcus are associated with wounds [49]. OTUs from the genera Helcococcus, Aggregatibacter, Enterococcus, and the family Corynebacteriaceae and one unassigned OTU include bacteria with pathogenic potential [50]. OTUs from the genus Finegoldia and one unassigned OTU were classified as double-function bacteria with fermentative functions [51]. One OTU was unassigned with unknown functions, and OTUs belonging to the genera Gallicola, Anaerococcus, Peptoniphilus, Proteus, unassigned Staphylococcus, Clostridium, Peptostreptococcus specialize in fatty acid production through fermentative processes [52]. More specific information on known functions of these groups is summarized in the supporting information (S1 Table).

Diversity and potential implications of common bacteria in the dorsal patch of L. yerbabuenae
The difference observed in the Alpha and Beta diversity analyses of the bacteria (Table 1 and Fig 3) suggest individual-level variation in the assembly of the microbiota in the dorsal patch of each individual. Sixty-one percent of the shared OTUs in the dorsal patch have a fermentative function (Figs 7 and 8). These fermentative bacteria could metabolize short-chain fatty acids (SCFAs), likely contributing to scent production and possibly to mate attraction [17,[53][54][55].
These results are consistent with the first premise of the fermentative hypothesis that volatile odorants are produced by bacteria that colonize mammal scent structures. This premise has also been supported by several other studies [17,20,54]. The second premise, that "individual "odor prints" result from the differences among individuals in the composition of the bacterial communities of these structures" was not tested in this study [20]. To do that, experimental studies are needed to detect the presence of SCFAs and confirm their role in individual recognition.
The recognition of skin microbiota as a main component of odor production is key to further understanding animal behavior. The olfactory receptors, the main olfactory epithelium, and the vomeronasal organ receive the odor signals produced by bacteria, which are then sent to the brain, a chemical signal pathway known as the microbiome-skin-brain axis [55]. The presence of fermentative bacteria in the dorsal patch of L.yerbabuenae could be important in generating odor, a potential reproductive signal.

Conclusion
This study contributes to the establishment of baseline knowledge of the bacteria associated with the dorsal patch of L. yerbabuenae using molecular methods (16S rRNA sequencing). High-throughput sequencing techniques, along with bioinformatic analysis, allowed us to describe the biodiversity of the bacterial consortium of the dorsal patch of L. yerbabuenae. Current microbiome analyses are based on genetic diversity and composition estimations, coupled with information found in the literature. However, a major remaining challenge is that the large majority of the sequences found in this study were unassigned OTUs, making it necessary to rely on assertive association techniques using phylogenetic trees and other informatic comparisons to understand the system. This information could be substantially deepened by using a metagenomic approach to infer functional and taxonomic information as well as metabolic properties present in the bacterial communities associated with this and other systems, as well as their repercussions in the host and its life cycle. There is a long journey ahead to understand how microbiomes intervene in diverse biological systems, and how they contribute to regulating host behavior and reproduction through chemical signaling.
Supporting information S1