Use of gnotobiotic cod larvae (Gadus morhua L.) to unravel host-microbe interactions

Use of gnotobiotic cod larvae to unravel host-microbe interactions In this project we wanted to investigate the effect of bacteria on growth, survival, gene expression and metabolome of cod larvae. We have previously developed a system where we can hatch and maintain cod larvae in the absence of bacteria. Such germfree cod larvae can then be compared with larvae supplemented with specific bacteria; these are called gnotobiotic as we know exactly which bacteria are present. Cod larvae are dependent on live feed. In previous gnotobiotic experiments we have used germfree rotifers, restricting the experimental time to 17 days after hatching (as larger cod larvae prefer lager feed). One of the part-objectives of this study was to improve our experimental system and allow for longer experimental time (up to 30 days). To meet the increasing feed demands of the larvae, a larger feed organism, Artemia needs to be introduced. In 2014 we performed a start-feeding experiment with germfree, gnotobiotic and conventionalized cod larvae. One gram-positive and one gram-negative bacterial strain were added in equal amounts to the gnotobiotic larvae. Both strains are non-pathogenic and were originally isolated from cod guts. Conventionalized larvae were initially germfree but subsequently reared in seawater with a complex and undefined microbiota. Samples were taken at day 1, 4, 8, 13, 16 and 21 after hatching. On day 17, germfree Artemia was introduced as feed, and after a few days Artemia remains were visible in all cod guts. The final survival in all rearing bottles was relatively high, around 50% for the germfree and gnotobiotic, and 33% for the conventionalized larvae. None of the germfree or gnotobiotic rearing bottles were contaminated by bacteria during the experiment. The experiment lasted until 21 days after hatching, due to accumulation of rapidly growing Artemia in the rearing bottles (which only has a low level of water exchange), which we feared might become detrimental to the larvae over time. We believe however, that after adjustments to the rearing system there will be no problems reaching the 30 day goal. The development of the bacterial communities in the rearing water and cod larvae over time was followed by the use of 16S DGGE PCR. These analyses showed that in the gnotobiotic bottles one of the added bacteria, the gram negative Vibrio strain dominated, and only small amounts of the gram positive Microbacterium could be detected. In the conventionalized bottles, the microbial diversity gradually decreased over time, and the composition of the microbial community in the water and fish larvae became more and more dissimilar. Even though the germfree larvae had the highest survival rates, they weighed significantly less on day 16 after hatching, pointing towards a nutritional effect of the gut microbiota. Microarray analysis showed that 436 genes were significantly differentially regulated between the germfree, gnotobiotic and conventionalized larvae at day 16 (p<0.05, fold change >1.5). If we include more moderate responses (fold changes >1.2) this increases up to more than 1500 genes. Genes involved in growth and development showed higher expression in the cod larvae with bacteria, corresponding with previous observations in zebrafish. Many of the genes with higher expression in the germfree larvae were involved in the immune response, indicating a possible regulatory role of the cod microbiota. In addition, genes involved in digestion (especially proteolytic activity) were upregulated in the germfree larvae. The gnotobiotic larvae only had 22 genes with significantly different expression than the germfree larvae, and were most dissimilar compared to the conventionalized larvae. The high similarity in gene expression between gnotobiotic and germfree larvae could be due to the low diversity of bacteria in the gnotobiotic bottles, as we have previously seen how different compositions of bacteria lead to different host gene expression.

Through the use of gnotobiotic systems (=free of, or containing only known bacteria), the contributions of bacteria on their vertebrate hosts have been confirmed. The overall goal of this proposal is to develop our unique gnotobiotic model for cod (Gadus morhua)larvae to expand our knowledge into the complex interactions that take place between cod larvae and its gut microbiota. The proposals main objectives are system development and host-response detection. Taking advantage of available high-throughput methods for analysis such as a cod specific microarray and non-target metabolomics by UPLC-QTOF MS, we anticipate a significant amount of information from this studies. Increased knowledge into host-microbe interactions in cod larvae may ultimately lead t o improved strategies of microbial management and higher survival of intensively reared larvae. These improvements could positively affect both the financial and resource management aspects of larval rearing.