| dc.description.abstract | Atlantic cod (Gadus morhua L.) larvae hatch at an early developmental stage, and poor growth, high mortality and low reproducibility in juvenile production is considered the main bottleneck of cod farming. This high mortality has been attributed to opportunistic bacteria in the rearing water. In order to improve the survival and growth of cod larvae we need a better understanding of the interactions that take place between larvae and bacteria, however these interactions are incredibly complex. To study host-microflora interactions this inherent complexity needs to be reduced, this can be achieved by the use of gnotobiotic systems. Gnotobiotic (“gnotos”=known) systems are either free of, or contain only known bacteria.
We have developed a protocol for bacteria-free cod larvae, which is based on surface disinfection of cod eggs with glutaraldehyde, and subsequent hatching in sterile seawater containing rifampicin and ampicillin. This protocol is highly reproducible and has no adverse effects on the hatchability of cod eggs. After hatching no additional antibiotics are used, which facilitates gnotobiotic studies with bacterial strains that are not resistant to antibiotics.
To identify genes whose expression is altered by the presence of bacteria (host-response genes) two different approaches was chosen, Suppression Subtractive Hybridization PCR (SSH-PCR) as an unbiased, and Quantitative PCR of selected genes as a biased approach.
While some putative host-response genes were identified from the subtracted libraries, over 50% of the transcripts could not be identified through Blastn/Blastx searches. In addition, when we tried to verify the results by qPCR on larvae from a second experiment, the expression levels varied significantly between larvae from replicate conventional (with undefined microflora) rearing bottles, reflecting a large difference in microbial composition. Expression analysis of host-response genes selected based on findings in zebrafish also identified putative host-response genes. However, the conventional replicates analyzed were the same as for the SSH-PCR part, and also for these genes the expression varied between the two “replicates”.
As the differences in the microbial content of conventional rearing bottles most likely caused the large differences in gene-expression, mono -and poly-gnotobiotic conditions were used to attempt to clarify the effect bacteria have on expression of host-response genes. Consequently, in 2 different experiments (1 short-term without feeding and 1 longterm with feeding) the effects of various gnotobiotic conditions on the expression of genes involved in immunological, nutritional and morphological aspects were investigated. In the short-term experiment, microbial specificity of 4 host-response genes was observed both with respect to the composition and the activity (i.e. dead or alive) of the bacteria present. Live bacteria generally resulted in higher transcript levels of all the genes investigated, indicating that these host-responses are partly dependent on microbial activity, and not just due to interactions with bacteria as particles. When comparing live and dead gnotobiotic conditions containing the same microbes, survival of cod larvae was higher in the conditions with dead bacteria
In the long-term experiment, the effects of live vs. dead probiotic candidate Microbacterium ND 2-7 on the expression of 15 genes was investigated. No difference with regards to survival of the larvae was observed, and expression levels were similar for ¾ of the genes investigated. However, for larvae exposed to live Microbacterium, expression of 4 genes involved in cell growth/differentiation and nutrient uptake were slightly, but statistically significantly up-regulated. This may suggest that live Microbacterium promote growth and digestive capacity of the cod larvae through processes other than simple particle recognition. | nb_NO |
