Early life microbial colonizers shape and support the immature vertebrate immune system. Microbial colonization relies on the vertical route via parental provisioning and the horizontal route via environmental contribution. Vertical transmission is mostly a maternal trait making it hard to determine the source of microbial colonization in order to gain insight in the establishment of the microbial community during crucial development stages. The evolution of unique male pregnancy in pipefishes and seahorses enables the disentanglement of both horizontal and vertical transmission, but also facilitates the differentiation of maternal vs. paternal provisioning ranging from egg development, to male pregnancy and early juvenile development. Using 16s rRNA amplicon sequencing and source-tracker analyses, we revealed how the distinct origins of transmission (maternal, paternal & horizontal) shaped the juvenile internal and external microbiome establishment in the broad-nosed pipefish Syngnathus typhle. Paternal provisioning mainly shaped the juvenile external microbiome, whereas maternal microbes were the main source of the internal juvenile microbiome, later developing into the gut microbiome. This suggests that stability of niche microbiomes may vary depending on the route and time point of colonization, the strength of environmental influences (i.e., horizontal transmission), and potentially the homeostatic function of the niche microbiome.

Sample collection: 

Adult Syngnathus typhle were caught in Orth on Fehmarn (54°26'N 11°02'E) and brought to our aquaria facilities at GEOMAR Kiel, for breeding. Fish were kept in a flow-through aquaria system at 18°C with 18h day/6h night light regime and fed with live Artemia salina and frozen Mysidae spp. twice a day. In each aquarium three males and three females were kept together to allow mating. After the onset of breeding, fish were randomly sampled, regarding their sex and gravity stages, on five days between the end of May 2019 and end of June 2019. We sampled 89 mature S. typhle (18 females, 16 early pregnant males, 19 mid pregnant males, 18 late pregnant males and 19 non-pregnant males). Pregnancy stages (early, mid and late pregnancy) were defined according to (56). In order to detect microbial transfer from parental gonads and pouch tissue to the juveniles, we sampled testes and endometrial inner pouch lining tissue as well as fertilized larvae from the three pregnancy stages in male fish. In female fish, we sampled unfertilized eggs to assess potentially deposited microbes into the eggs or on their surface. The hindgut was sampled irrespective of sex. To investigate maternal microbial transfer through the cytoplasm, we surface-sterilized half of the unfertilized eggs from each female in a bath of 0.5% Polyvinylpyrrolidone-iodine (PVP-I, Solution in sterile-filtered Phosphate buffered Saline (PBS)) for 5 min with subsequent washing three times with 500µl sterile-filtered PBS (adapted from (57)). To sample the cytoplasm of surface-sterilized eggs (sterilized eggs) without contamination through the chorion, the egg was squished in the collection tube. The same sterilization treatment was applied to larvae (sterilized juveniles) of different pregnancy stages to discriminate between external and the internal microbiome. Non-sterilized eggs (untreated eggs) and larvae (untreated juveniles) were directly placed in the collection tubes. We pooled three juvenile and eggs sample from each of the pregnancy stages and sterilization treatment. All sampled organs were collected into collection microtubes from the DNeasy96 Blood and Tissue Kit from Qiagen (Hilden, Germany) and stored immediately at -80 °C.

Juvenile microbiota development:         
Pregnant male S. typhle were caught in Orth on Fehmarn (54°26'N 11°02'E) in late spring 2019, transferred to the aquaria system, kept individually in a flow-through system with 18h day/6h night cycle and fed twice a day with live Mysidae spp. After parturition, free-swimming juveniles of each male were kept in a distinct aquarium and fed ad libitum twice per day with live Artemia salina. First sampling took place after release from the brood pouch, sampling was then continued in 2-3 day intervals. During each sampling three juveniles per family tank were collected individually in a collection tube for the analysis of whole-body microbiome (whole juveniles) development. To test for development of internal vs. external microbiome another three juveniles from the same family tank were euthanized by immersion in MS222 and the gut was removed in a sterile manner (juvenile gut) and collected individually. At each sampling day, controls of water and food samples were taken. 

RNA extraction, library preparation and amplicon sequencing:

Both datasets have been treated with the same DNA extraction and 16S rRNA sequencing protocols. DNA extraction was done with DNeasy Blood & Tissue Kit (QIAGEN, Germany) following the manufacturers protocol including a pre-treatment for Gram-positive bacteria with ameliorations from (58). See Supplementary Material 1 for further details. Library preparation was done by the institute for experimental medicine (UKSH, Campus Kiel) with 20µl of sample DNA from each sample. Amplicons of the V3-V4 hypervariable region (341f/806r) were sequenced using the Illumina MiSeq platform (Illumina, USA) with 2x 300-bp paired-end read settings at the IKMB, Kiel University.