Modulation of innate immune responses in Atlantic salmon by chronic hypoxia-induced stress.

Atlantic salmon post-smolts were exposed to either chronic hypoxic (Hy) or normal oxygen (No) conditions in seawater tanks for 58 days, mimicking conditions typical of sea cages for farmed salmon at some periods of the year. By day 29 head kidney macrophages were isolated and subjected to in vitro poly I:C stimulation to simulate viral infection, and samples were collected over 48 h. By day 58 fish were subjected to in vivo stimulation using poly I:C or a Vibrio water-based vaccine to simulate viral or bacterial infection, respectively. The fish were monitored for stress responses and expression of several pro-inflammatory genes in head kidney and intestinal tissue up to five days post-injection. Stress load was monitored by plasma cortisol estimation at days 29 and 58, and on days 1, 2, 3 and 5 post-injection in the in vivo trial. Hy exposure resulted in elevated plasma cortisol levels on day 29 compared to No, while on day 58 cortisol levels were higher in the control group. Additionally, both poly I:C and the Vibrio vaccine gave significantly increased cortisol levels one day post-injection compared to PBS treated controls, irrespective of previous oxygen exposure. In vitro stimulation of macrophages with poly I:C revealed higher IFNα mRNA levels at 6, 12 and 24 h and for Mx at 12 and 24 h post-stimulation, for both No and Hy individuals. Moreover, IFNα levels were higher in No than in Hy individuals at all time points, and a similar difference was seen in Mx at 48 h. In vivo stimulation with poly I:C elicited strong elevation of the IL-1β, IFNγ, Mx and IP10 mRNA transcripts in head kidney, while TNFα1 and IFNα were found unaffected. The Vibrio vaccine elicited a strong up regulation of IL-1β, IFNγ and IP10 mRNA, whereas Mx, TNFα1 and IFNα appeared unchanged. Significant differences in expression between different oxygen exposure groups were found for all genes and both stimuli. The overall trend suggests that long-term hypoxia either reduces or delays the expression of these genes in head kidney. Expression of IFNγ and Mx in intestinal tissues also showed a strong up regulation of the genes following poly I:C stimulation, and also here the overall trend suggests that chronic hypoxia results in a lower or delayed expression of the measured genes. In summary, our results indicate that chronic hypoxia modulates the expression of important immune related genes putatively altering the immune response. As the effect is present in isolated macrophages as well as head kidney and intestinal tissue the modulation appears to be affecting local as well as systemic responses.