The observed tendency for higher plasma lactate levels
The observed tendency for higher plasma lactate levels in the aerated thermal exposure suggests increased reliance on Vacquinol-1 metabolism to maintain function. This finding is supported by a recent study that observed an increased ṀO2 in European perch (Perca fluviatilis) exposed to a hyperoxic rather than normoxic thermal ramp (Brijs et al., 2015), suggesting that some species can benefit from supplemental oxygen in an environmentally-relevant acute thermal episode. Additionally, the maintenance of plasma lactate in the hyperoxic treatment, coupled with the lack of change in ventricular MDA levels, suggests that trout were largely able to avoid both harmful oxidative damage and anaerobism for the duration of the experiment. Hyperoxia is known to increase ROS production, which is implicated in a wide range of physiological effects, including arrythmogenesis (Tse et al., 2016). Our results suggest there is background signaling during an acute hyperoxic thermal exposure coordinating protein localization, expression, and turnover, even before oxidative damage has accumulated. Specifically, increased mtHKI density during combined hyperoxia and hyperthermia should limit both oxidative stress and apoptosis to protect cardiac muscle. These processes would then lower the risk of arrhythmogenesis and the associated disruptions in oxygen uptake and delivery that limit survival under physiological stress. Furthermore, the gradation in the increase of cardiac protein ubiquitinylation between the aerated and hyperoxic exposures suggests that the degree of response induced can be fine-tuned based on the signal; the amount of ROS produced would likely be compounded by both the hyperthermic and hyperoxic aspects of the exposure and could have induced redox-sensitive signaling in the CM. Our finding that mtHKI density only increases during a hyperthermic event when hyperoxia is also present suggests that either a certain threshold of ROS is needed to trigger this cardioprotective effect, or else that separate hyperoxic and hyperthermic signaling may synergize to increase mtHKI density. We were unable to detect a change in ROS between treatments using malondialdehyde as a proxy for oxidative damage, but this should not be taken to suggest that there is no difference in ROS production between treatments. The contributions of high temperature (Banh et al., 2016) and hyperoxia (Lushchak and Bagnyukova, 2006) to ROS production in fish are well-recognized. In this case, an increase in ROS production may have been masked by adequate antioxidant reserves such as glutathione or thioredoxins over the short time course of the experiment. Sustained or repeated insults may have resulted in higher levels of MDA or other markers of oxidative stress. Together, these results support previous findings of improved performance in terms of higher CTmax (Brijs et al., 2015, Ekström et al., 2016) or decreased plasma lactate (Devor et al., 2015) when hyperthermic events are compounded with hyperoxia. As mtHKI levels seemed highly dynamic when considering the dogma of non-actively regulated mammalian HKI, an in vitro experiment was undertaken to confirm the cardioprotective nature of mtHK. We hypothesized that LND-mediated dissociation of mtHK would increase mortality in isolated CMs by leaving the mPTP open, allowing dissipation of the mitochondrial membrane potential as well as cytochrome C release to the cytosol, which initiates apoptotic processes. Moreover, we hypothesized that LND-mediated dissipation of the proton motive force would effectively short-circuit the electron transport chain and result in increased ṀO2 as metabolism continues at full capacity with little benefit in the form of ATP production. As the ionophore FCCP allows for dissipation of the mitochondrial membrane potential without opening of the mPTP (Pulselli et al., 1996), we finally hypothesized that the administration of FCCP would have a similar and noncumulative effect on ṀO2 as LND, but that mortality would be cumulative between the two treatments. Force production in the ventricular strip preparations was not affected by LND treatment, suggesting that inhibition of HK was not sufficient to depress metabolic flux enough to affect contractility, a highly costly activity in terms of energy consumption. Furthermore, the increase in oxygen utilization upon administration of LND or FCCP supported our hypothesis that metabolism was inhibited at the stage of oxidative phosphorylation, as substantial inhibition of hexokinase would prevent downstream use of O2 to catabolize glycolytic end products. Our results closely matched our predictions, supporting the cardioprotective role of mtHK in salmonids. It is important to note that this experiment was not specific to HKI and may also have influenced HKII localization. The dynamic nature of salmonid ventricular HKI localization strongly suggests some degree of active regulation, specifically in a situation where cardioprotective measures are expected to be undertaken.