Low levels of extracellular glucose limit cardiac anaerobic metabolism in some species of fish

There is a wide interspecific range in the level of blood glucose amongst teleost fishes. Some species exhibit levels in excess of 10 mmol l⁻¹ while other species have extremely low levels by vertebrate standards, often below 1 mmol l⁻¹ (Chavin and Young, 1970; Polakof et al., 2012). How interspecific variability influences the utilization of glucose by different tissues is only now being addressed in a rigorous fashion.

Fish red blood cells (RBCs) have mitochondria and a predominantly aerobic respiration (see Driedzic et al., 2013 for references). The rate of glucose metabolism was determined in RBCs from Atlantic cod (Gadus morhua), Atlantic salmon (Salmo salar), cunner (Tautogolabrus adspersus) and short-horned sculpin (Myoxocephalus scorpius) that had plasma glucose levels of 4, 4, 2, and 0.7 mmol l⁻¹, respectively (Driedzic et al., 2013). At physiological levels of extracellular glucose, the rate of glucose metabolism was significantly higher in RBCs from Atlantic cod than either cunner or short-horned sculpin, with Atlantic salmon RBCs being intermediate between Atlantic cod and the other two species (Driedzic et al., 2013). A follow-up study involved Atlantic cod (plasma glucose, 3 mmol l⁻¹) and short-horned sculpin (plasma glucose, 0.3 mmol l⁻¹) (Driedzic et al., 2014). At physiological levels of extracellular glucose, the rate of oxygen consumption by RBCs was similar between species, but the rate of glucose metabolism by Atlantic cod RBCs was sixfold higher than in short-horned sculpin RBCs. The shortfall in metabolic fuel is likely made up by the catabolism of on-board amino acids (Driedzic et al., 2014). These experiments were extended to isolated myocytes that have rates of oxygen consumption 2.5- to 3-fold higher than RBCs. As with RBCs, at physiological levels of extracellular glucose there was no difference in the rate of oxygen consumption of isolated heart cells, but the rate of glucose metabolism was fourfold higher in Atlantic cod than short-horned sculpin. In isolated heart preparations, oxygen consumption and power development are likely maintained by endogenous triglycerides in the absence of exogenous metabolic fuels (Driedzic and Hart, 1984; Clow et al., 2004, 2016). Collectively, these data suggest that in species comparisons of RBCs and heart cells, operating at basal or sub-basal levels of metabolism under normoxic conditions, the higher the level of extracellular glucose the higher the rate of utilization of glucose.

The present study assesses the above putative relationship in perfused hearts from various species. Contracting, whole hearts were utilized in order to increase the level of energy demand and to provide a more physiological setting than isolated cells. In addition to Atlantic cod (G. morhua Linnaeus 1758) and short-horned sculpin [M. scorpius (Linnaeus 1758)], the experimental matrix included rainbow trout [Oncorhynchus mykiss (Walbaum 1792)] and lumpfish (Cyclopterus lumpus Linnaeus 1758), which have plasma glucose levels of approximately 5 and 1 mmol l⁻¹, respectively. Furthermore, rates of glucose metabolism and lactate production were assessed under both normoxic and hypoxic conditions, as it is well recognised that oxygen limitation increases rates of anaerobic metabolism in fish hearts (Gamperl and Driedzic, 2009). In perfused hearts from rainbow trout receiving only trace amounts of radiolabelled glucose in the medium, cyanide treatment resulted in increased utilization of glucose (West et al., 1993). In addition, hypoxia led to an increase in uptake of the non-metabolizable glucose analog 2-deoxyglucose by isolated heart preparations from American eel (Anguilla rostrata) (Rodnick et al., 1997) and Atlantic cod (Clow et al., 2004). The impact of hypoxia on the rate of metabolism of glucose at physiologically relevant levels of glucose has never been directly assessed. Here we test the hypothesis that under normoxic and hypoxic conditions, species with high physiological levels of glucose have higher rates of glucose metabolism than species with low extracellular glucose levels. Rates of glucose metabolism were assessed by the production of ³H₂0 from [2-³H] glucose (Driedzic et al., 2014; Clow et al., 2016). The most important finding is that glucose metabolism is compromised under hypoxia at low levels of extracellular glucose within the physiological range for some species.

Atlantic cod, rainbow trout, lumpfish and short-horned sculpin were maintained in running seawater at a temperature of 8–10°C and kept on a natural photoperiod with fluorescent lights set by an outdoor photocell. Atlantic cod and lumpfish were hatchery reared at the Ocean Sciences Centre (OSC). Rainbow trout were obtained from the Marine Institute of Memorial University of Newfoundland (MUN) and transferred to the OSC. Short-horned sculpin were collected locally as required by the field services unit (OSC) during the summer period.

Atlantic cod, rainbow trout and lumpfish received Optiline Salmonid Feed (Skretting, St Andrews, NB, Canada) every day. Short-horned sculpin were fed to satiation weekly with chopped herring. Body masses and sample sizes were as follows: Atlantic cod (980±68 g; N=23), rainbow trout (565±28 g; N=12), short-horned sculpin (862±78 g; N=20) and lumpfish (293±16 g; N=10). Relative ventricle masses [expressed as % ventricle mass (g)/body mass (g)] were as follows: Atlantic cod (0.08±0.003), rainbow trout (0.07±0.003), short-horned sculpin (0.07±0.004) and lumpfish (0.06±0.004). Blood was collected for glucose analysis from the caudal vessel with a heparinized syringe prior to removal of the heart. Blood was centrifuged at 1500 g for 5 min and plasma was frozen for future analysis. Animal protocols were approved by the Animal Care Committee at the Memorial University of Newfoundland.

The basic perfusion/incubation medium consisted of (in mmol l⁻¹): 182 NaCl, 5 KCl, 1.99 MgSO₄, 2.3 CaCl₂, 7.33 TES base, 2.58 TES acid and various concentrations of glucose (pH 7.8 at 8°C). The heart perfusion apparatus was similar to that described in Clow et al. (2004). The system was designed to recirculate 10 ml of perfusion medium at a temperature of 8°C. Temperature of the input chamber was maintained using a recirculating water bath. A metal cannula attached to the input chamber was placed just above the heart, allowing the perfusate to fill the heart by gravity. The input cannula was inserted into the atrium and the heart was perfused without recirculation for 30 s to allow washout of blood. The aorta was then cannulated with PE tubing in order to effectively collect the fluid into a graduated cylinder for recirculation. The fluid in the cylinder was pumped back to the input chamber. Hearts were electrically paced using a Grass Model SD9 square wave generator set at 5 V and 200 ms duration. One electrode was fused to the input cannula and a second electrode, connected to a small hook, was inserted into the ventricle. The contraction rate was set to 30–35 beats min⁻¹, typical for marine teleosts at the temperature of this study, and only those hearts that showed visible contractions were used in analysis. Hearts were perfused for 5 to 10 min with basic medium in order to allow contraction to stabilize. Thereafter, hearts were perfused with recirculating medium containing [2-³H]-glucose (7.4 kBq; American Radiolabeled Chemicals, Burnaby, Canada).

Preliminary experiments were conducted with hearts isolated from two Atlantic cod to assess whether the preparation was responsive to a change in oxygen level in the medium. In these studies, perfusion was initiated with medium gassed with air and after 45 min of perfusion the gas was switched to 100% nitrogen, resulting in a drop of oxygen level. Oxygen readings were taken using a fibre-optic oxygen meter (model OXY-4 mini) and a precalibrated dipping probe (PreSens). Rates of glucose metabolism and lactate production were monitored for 210 min by collecting 150 µl of medium every 15 min for 60 min and every 30 min thereafter. Following the perfusion, hearts were frozen and later analyzed for lactate and glycogen content.

In the main series of experiments, one group of hearts, hereafter referred to as normoxic, were perfused under conditions of constant aeration (oxygen level, 95±0.7% air saturation, N=28). In a parallel group, hearts were initially perfused with aerated medium and, once vigorous contraction was observed, switched to medium that was gassed continuously with nitrogen. This group is considered to be hypoxic (oxygen level, 12±0.6% air saturation, N=30). All experiments lasted for 180 min. Samples for glucose metabolism and lactate production were collected as described in the preliminary experiment. Experiments were conducted with all four species with medium containing glucose at a level in the physiological range. Levels were 5.0 mmol l⁻¹ for Atlantic cod and rainbow trout, 1 mmol l⁻¹ for lumpfish and 0.5 mmol l⁻¹ for short-horned sculpin. A further series of experiments was conducted with Atlantic cod at a non-physiologically low level of 0.5 mmol l⁻¹ glucose and with short-horned sculpin at a non-physiologically high level of 5.0 mmol l⁻¹ glucose.

The metabolism of glucose was determined by measuring the rate of ³H₂O production that is formed from [2-³H]-glucose via water exchange in the reaction catalyzed by phosphoglucose isomerase (Hutton, 1972; Driedzic et al., 2013). A volume of 100 µl of sample medium at time 0, 15, 30, 45, 60, 90, 120 or 180 min was added to 1 ml of resin in polystyrene columns and four 1 ml water fractions containing the ³H₂O were collected in scintillation vials and counted.

Lactate was measured in the remaining perfusate at each time point. The medium was deproteinized 1:1 with 6% PCA and 25–50 μl of extract was added to 200 μl assay medium containing glycine buffer (Sigma-Aldrich, G5418) and 2.5 mmol l^–1 NAD⁺, pH 9.0 at room temperature. Samples were read after 10 min before adding 30 IU ml^–1 lactate dehydrogenase. Absorbances were read for another 60 min or until stable. All absorbances were determined with a DTX 880 microplate reader (Beckman Coulter, Mississauga, ON, Canada). Lactate was measured in the heart by homogenizing the muscle in 6% PCA and centrifuging at 10,000 g. The supernatant was measured as described above.

Glycogen was measured in hearts based upon a procedure modified from Keppler and Decker (1974) using amyloglucosidase (EC 3.21.3) to hydrolyze glycogen. Briefly, hearts were homogenized and deproteinized in 6% PCA, centrifuged and neutralized with 2 mol l⁻¹ potassium bicarbonate. The neutralized homogenate was added to 56 IU ml⁻¹ amyloglucosidase in 0.4 mol l⁻¹ acetate buffer, pH 4.8 and heated at 40°C. Hydrolysis was stopped after 2 h by adding 12% PCA. Glucose from these samples was measured as described by Clow et al. (2004). Free glucose from these samples without hydrolysis was also measured and subtracted from the glucose produced from glycogen. Plasma glucose was also assayed as described by Clow et al. (2004).

Values are expressed as means±s.e.m. Statistical analyses applied are stated in the legends to figures. In all cases, P≤0.05 was considered to be significant. In the main experiment, rates of glucose metabolism and lactate production were calculated from the initial linear rates. Of the 52 data sets presented, all but two (short-horned sculpin, heart glycogen, normoxia; rainbow trout, glucose metabolism, normoxia) passed a normality test. Given the robustness of ANOVA tests, we elected to accept these anomalies in the analyses.

As anticipated, plasma glucose levels were substantially and significantly higher in Atlantic cod (5.1±0.4 mmol l⁻¹) and rainbow trout (4.8±0.4 mmol l⁻¹) than in lumpfish (0.89± 0.07 mmol l⁻¹) or short-horned sculpin (0.52±0.09 mmol l⁻¹). There was no significant difference in plasma glucose levels between Atlantic cod and rainbow trout, or between lumpfish and short-horned sculpin (Fig. 1).

Preliminary experiments were conducted to assess whether hearts responded to a change in oxygen level in the perfusion medium (Fig. 2). In the case of the experiment with Atlantic cod specimen 1, the initial level of oxygen was 92% air saturation. The medium was gassed with N₂ after 45 min of perfusion, and at the end of the experiment (210 min) the oxygen level was 7% of air saturation. The slopes of arbitrarily selected linear portions of the data sets were calculated. The initial rate of glucose metabolism was 8.33 nmol g⁻¹ min⁻¹, which increased to 79.1 nmol g⁻¹ min⁻¹ for the perfusion period between 120 and 210 min. Lactate production, determined from lactate accumulation in the medium, followed a similar pattern: the initial rate was 4.4 nmol g⁻¹ min⁻¹ and this increased to 372 nmol g⁻¹ min⁻¹. The experiment with Atlantic cod specimen 2 also showed rate increases, although the linear portions of the curves were shifted. In this case, the initial oxygen level was 82% and the final level was 14% of air saturation. The rate of glucose metabolism increased from an initial level of 23.3 nmol g⁻¹ min⁻¹ to 108 nmol g⁻¹ min⁻¹ for the period between 60 and 120 min. The rate of lactate production increased from 3.2 to 397 nmol g⁻¹ min⁻¹. Although limited, the data clearly show that hearts are responsive to a decrease in oxygen availability. The transition to a hypoxic environment led to an average sevenfold increase in glucose metabolism that was accompanied by an average increase in lactate production by approximately 100-fold. These preliminary studies also illustrate, at least in the case of Atlantic cod hearts perfused with 5 mmol l⁻¹ glucose in the medium, that the rate of increase in the use of exogenous glucose under hypoxia is not sufficient to support lactate production. The reason for the variability between preparations is not known but could be related to the time course of the decrease in oxygen level in the medium. In order to control for this, all further experiments were conducted with individual hearts exposed to either normoxic or hypoxic conditions from time zero.

In the main series of experiments, all heart preparations were perfused for 180 min. Lactate and glycogen levels following perfusion were determined in all hearts. Given that we were only interested in assessing whether hypoxia resulted in a change in lactate or glycogen level, within a species at a specific level of extracellular glucose, the data were analyzed with a t-test. There was no significant difference in heart lactate between hearts perfused under normoxic or hypoxic conditions for any species under conditions of 5, 1 or 0.5 mmol l⁻¹ glucose in the medium (Fig. 3A). Because lactate levels in the heart do not change, this means that any lactate produced is released to the medium. Glycogen levels were significantly lower in hearts from Atlantic cod perfused with 5 mmol l⁻¹ in the medium under hypoxic than normoxic conditions (Fig. 3B). A similar decrease was evident for hearts from rainbow trout. There was no significant difference between normoxic and hypoxic treatments for either lumpfish or short-horned sculpin. We caution, though, that given the high individual variability within short-horned sculpin hearts, it would be difficult to reveal any change owing to treatment with the small sample size.

The relationship between glucose metabolism or lactate production and time for the four species studied is presented in Figs 4–7. The raw data alone suggest a number of qualitative findings. Foremost, rates of glucose metabolism appear to be higher under hypoxic than normoxic conditions for Atlantic cod (Fig. 4A,C) and rainbow trout (Fig. 5A), but not lumpfish (Fig. 6A) or short-horned sculpin (Fig. 7A,C). Rates of lactate production appear to be higher for Atlantic cod (Fig. 4B,D) and rainbow trout (Fig. 5B) under hypoxia than normoxia. The raw data also reveal that in some cases glucose metabolism (Fig. 4C) and in many cases lactate production (Figs 4B,D, 5B, 7B,D) appear to lose linearity with respect to time and plateaus. On the basis of this finding, rates of glucose metabolism and lactate production were calculated for the initial linear portion of the curves for each individual experiment.

The experiment designed to assess rates of glucose metabolism and lactate production at physiological levels of extracellular glucose involved eight conditions with different species, oxygen levels or extracellular glucose levels (Fig. 8). The data were analyzed with two-way ANOVA including all conditions, but the figures are partitioned out to better visualize the relevant physiological comparisons consistent with the hypotheses being tested. Under conditions of normoxia and physiological levels of glucose, there was no difference in rates of either glucose metabolism or lactate production across species (Fig. 8A,B). The pattern changed under hypoxia, where rates of glucose metabolism and lactate production by hearts from Atlantic cod and rainbow trout were significantly higher than those by hearts from lumpfish and short-horned sculpin (Fig. 8C,D). Although of similar magnitude, the rate of glucose metabolism was significantly higher in Atlantic cod than rainbow trout hearts. There was no difference in rates of lactate production by hearts from these two species. Lumpfish and short-horned sculpin hearts exhibited similar rates of both glucose metabolism and lactate production under hypoxia. Rates of glucose metabolism and lactate production between normoxic and hypoxic conditions may be assessed by comparing Fig. 8A with C and B with D. These rates were significantly higher under hypoxic than normoxic conditions for hearts from both Atlantic cod and rainbow trout. There was no significant difference for either glucose metabolism or lactate production for lumpfish and short-horned sculpin hearts between normoxic and hypoxic conditions. However, the average values for lactate production were approximately twofold higher for lumpfish and 17-fold higher for short-horned sculpin under hypoxia than normoxia, respectively.

The above data suggest that hearts from Atlantic cod and rainbow trout respond to hypoxia with an increase in glucose metabolism and lactate production, but lumpfish and short-horned sculpin hearts have no capacity to increase glucose metabolism under hypoxia and at the very best a limited capacity to increase lactate production. This could be because of the low extracellular glucose or other properties inherent to the heart. In an attempt to tease out these possibilities, hearts were perfused with medium containing non-physiological levels of extracellular glucose. More specifically, Atlantic cod hearts received 0.5 mmol l⁻¹ glucose and short-horned sculpin hearts 5.0 mmol l⁻¹ glucose in the medium. Rates of glucose metabolism and lactate production are presented in Fig. 9 and for comparative purposes values at physiological levels of extracellular glucose are also included. For Atlantic cod, there was no significant difference in rates of glucose metabolism under normoxic conditions between conditions of 0.5 and 5.0 mmol l⁻¹ glucose in the medium, nor between normoxia and hypoxia at 0.5 mmol l⁻¹ glucose (Fig. 9A). Lactate production was significantly higher in Atlantic cod hearts perfused with hypoxic than normoxic medium with 0.5 mmol l⁻¹ glucose in the medium (Fig. 9B). For short-horned sculpin hearts there was no difference in rate of glucose metabolism between normoxia and hypoxia with either 5.0 or 0.5 mmol l⁻¹ glucose in the medium (Fig. 9C). The rate of glucose metabolism was significantly higher (approximately threefold) at 5.0 than at 0.5 mmol l⁻¹ glucose under normoxic conditions. Lactate production was significantly higher under hypoxia than normoxia at both levels of extracellular glucose.

The range of plasma glucose for the four species under study was as anticipated, with Atlantic cod and rainbow trout at approximately 5 mmol l⁻¹, which is in the window for most teleosts, typically ranging from 2 to 8 mmol l⁻¹. Lumpfish and short-horned sculpin had similar levels of plasma glucose at less than 1 mmol l⁻¹. This is extremely low even by fish standards (Chavin and Young, 1970; Polakof et al., 2012). As such, the selection of species was suitable to assess whether low plasma glucose compromised glucose utilization. Isolated, perfused hearts were used as the experimental preparation. Hearts received oxygen and glucose directly from the medium being pumped by the atrium and ventricle, as is the normal physiological condition for these hearts that have either no or only poorly developed coronary arteries (Driedzic and Gesser, 1994). Hearts were filled from a pressure head and emptied against atmospheric pressure. Thus the level of mechanical work, which is a primary determinant of energy demand, was similar for all species and considered to be sub-basal. Hearts contracted during the trial period, although it appeared that the strength of contraction decreased in some preparations, especially under hypoxic conditions, as evidenced by the force at which the perfusate was expelled. The input oxygen level was 12% of air saturation. Based on measurements in myoglobin-containing hearts, oxygen consumption by Atlantic cod, rainbow trout and short-horned sculpin hearts should be approximately 20% of normoxia, whereas myoglobin-devoid lumpfish hearts should be anaerobic (Bailey et al., 1990). Glucose metabolism was determined by the rate of ³H₂O production from [2-³H]-glucose. The virtue of this technique is that the specific activity of extracellular glucose is close to that of [2-³H]-glucose 6-phosphate, the substrate that produces ³H₂O (Neely et al., 1972). In preliminary experiments, hearts from Atlantic cod responded to a decrease in oxygen level with an increase in rate of ³H₂O production. Thereafter, in experiments that applied normoxic or hypoxic conditions from time zero there was a linear rate of production of ³H₂O that lasted for 180 min under normoxia and at least 120 min under hypoxia in most preparations. Rates of glucose metabolism were calculated from the initial linear period. Rates of glucose metabolism in Atlantic cod and short-horned sculpin were approximately twofold higher than in isolated myocytes (Clow et al., 2016). In summary, the experimental approach used here of providing [2-³H]-glucose to isolated hearts appears to be quite suitable to address the issue of species-specific rates of glucose metabolism under normoxic and hypoxic conditions given the viability of the preparation, the similarity of mechanical work for all species, the responsiveness to a change in oxygen level, the linearity of ³H₂O production, and higher rates of metabolism compared with isolated myocytes.

There was no significant difference across species in rates of glucose metabolism when considered at physiological levels of extracellular glucose. Under normoxia and physiological levels of glucose, the level of extracellular glucose is not a primary determinant of glucose utilization, at least in hearts working at sub-basal levels of energy demand. That is, lumpfish and short-horned sculpin operating with 1.0 and 0.5 mmol l⁻¹ extracellular glucose, respectively, are as effective in taking up and metabolizing glucose as Atlantic cod and rainbow trout with an availability of 5 mmol l⁻¹ extracellular glucose. Within Atlantic cod, the rate of glucose metabolism was the same at both 5 and 0.5 mmol l⁻¹ extracellular glucose, consistent with the interpretation that extracellular glucose is not an important factor in setting the rate of glucose metabolism under normoxic sub-basal working conditions. In short-horned sculpin hearts there was a threefold increase in rate of glucose metabolism between 0.5 and 5 mmol l⁻¹ extracellular glucose. Although the tissue has some capacity to increase glucose metabolism, this would not be achieved in vivo given that extracellular glucose is never any higher than approximately 0.5 mmol l⁻¹ (MacCormack and Driedzic, 2007; Driedzic et al., 2013, 2014; Clow et al., 2016).

The current finding of statistically similar rates of glucose metabolism in hearts from Atlantic cod and short-horned sculpin differs from results with isolated myocytes, where glucose metabolism was fourfold higher in Atlantic cod than in short-horned sculpin (Clow et al., 2016). Contractility apparently elevated glucose metabolism relatively more in short-horned sculpin than in Atlantic cod, thus eliminating the difference. In an earlier study with rainbow trout hearts, West et al. (1993), using the same experimental approach as here, reported the same rate of glucose metabolism as currently observed. In that study, an increase in work from sub-basal to resting levels of power output (1.75 mW g⁻¹) resulted in a fourfold increase in the rate of glucose metabolism. It would be of interest to determine whether increases in power output to physiological levels are matched with increases in glucose metabolism in species with extremely low plasma glucose concentrations.

Under normoxic conditions, glycolysis leading to lactate production makes only a minor contribution, less than 5%, to ATP turnover in perfused fish hearts and isolated myocytes (Driedzic et al., 1983; Arthur et al., 1992; West et al., 1993; Lague et al., 2012; Clow et al., 2016); however, low rates of cytosolic generation of ATP via lactate production may be important in maintaining resting tension (Bailey et al., 2000; Gesser, 2002). There was no significant difference in rates of lactate production by hearts from the four species in the present study, but the high variability should be noted. Again, increased cardiac energy demand might change this relationship, as increased work led to increased rates of lactate production in perfused hearts from rainbow trout (West et al., 1993) and tilapia (Lague et al., 2012) but not in sea raven (Driedzic et al., 1983), a scorpaeniforme closely related to short-horned sculpin.

Glucose enhances the hypoxic performance of isolated heart preparations and lactate production is enhanced under hypoxia in both heart preparations and at the whole-animal level (Gamperl and Driedzic, 2009; Lague et al., 2012; Becker et al., 2013). However, the contribution of extracellular glucose to lactate production is poorly documented. In perfused rainbow trout hearts treated with cyanide, the rate of glucose metabolism and lactate production increased 4- and 20-fold, respectively, relative to control preparations, implying that the use of extracellular glucose was insufficient to meet lactate production and most likely the glycogen pool was mobilized (West et al., 1993). Experiments with rainbow trout ventricle strips present a different scenario in that severe hypoxia led to a sixfold increase in lactate efflux but there was no change in the rate of uptake of 2-deoxyglucose (Becker et al., 2013). Analyzing differences in glucose trafficking between these two experiments is beyond the scope of the present study, but both reports suggest mobilization of the glycogen pool as was shown in the current experiment. Cardiac glycogen mobilization, under oxygen limitation, has also been noted in preparations from American eel (Bailey et al., 2000), Atlantic cod (Clow et al., 2004) and Pacific hagfish (Eptatretus stoutii) (Gillis et al., 2015). The paradigm that emerges is that under oxygen limitation, the fish heart metabolizes on-board glycogen and extracellular glucose, which are utilized to produce lactate with the associated generation of ATP that, in turn, supports contractility. However, none of the studies conducted to date involve species with chronically low (<1 mol l⁻¹) plasma glucose levels, such as lumpfish or short-horned sculpin. The current experiments are the first to assess glucose metabolism under hypoxia in species with exceptionally low plasma glucose levels by vertebrate standards. When considered at physiological levels of extracellular glucose, hypoxia resulted in increases in glucose metabolism in hearts from Atlantic cod (4.5-fold) and trout (7.4-fold) but no change in hearts from lumpfish or short-horned sculpin. As a consequence, under hypoxic challenge the rate of glucose metabolism was higher in Atlantic cod and rainbow trout than in lumpfish and short-horned sculpin hearts. That is, the species with high physiological levels of extracellular glucose have a greater capacity to increase rates of cardiac glucose metabolism under hypoxia. Low levels of extracellular glucose would result in a low diffusion gradient from the extracellular to the intracellular space and therefore could limit glucose utilization. In Atlantic cod heart perfused with medium at the low level of extracellular glucose (0.5 mmol l⁻¹), hypoxia did not result in an increase in glucose metabolism, a finding that supports the contention that low extracellular glucose compromises the ability to increase glucose utilization. In hearts from short-horned sculpin, non-physiologically high levels of glucose did not result in an enhanced capacity to increase glucose metabolism under hypoxia. For unknown reasons, possibly owing to reaching the ceiling of transport or enzymatic capacity, these hearts simply do not have the ability to attain high rates of glucose metabolism, even when provided with extracellular glucose 10 times the normal physiological level.

Rates of lactate production responded in a similar fashion to glucose metabolism. At physiological levels of extracellular glucose, significant increases in lactate production occurred under hypoxia in Atlantic cod (51-fold) and rainbow trout (18-fold) hearts, but not in hearts from lumpfish and short-horned sculpin. As with glucose metabolism, under hypoxic challenge the rate of lactate production was higher in Atlantic cod and rainbow trout than in lumpfish and short-horned sculpin hearts. Again, species with high physiological levels of extracellular glucose have a greater capacity to increase rates of cardiac lactate production under hypoxia. It should be noted that although not significantly different when assessed using two-way ANOVA, involving only data obtained at physiological levels of extracellular glucose, the average rate of lactate production by hearts from short-horned sculpin was 17-fold higher under hypoxia than normoxia, so it may be that this species has some capacity to increase lactate production, but rates would be less than 10% of that observed in Atlantic cod or rainbow trout. Enhanced rates of lactate production occurred under hypoxia at both 5 and 0.5 mmol l⁻¹ in Atlantic cod and short-horned sculpin hearts. Given that glucose metabolism was not increased under three of the four conditions, this finding suggests that glycogen was being called upon. Analysis of data for short-horned sculpin by a one-way ANOVA revealed a significant increase in lactate production under hypoxia, consistent with the suggestion above that hearts from this species have at least a limited capacity for anaerobic glycolysis.

The ratio of lactate production to glucose metabolism provides further insight into the relationship between extracellular glucose metabolism and lactate production. At physiological levels of extracellular glucose, the lactate production:glucose metabolism ratio for Atlantic cod, rainbow trout, lumpfish and short-horned sculpin was 5.1, 7.5, 15 and 3.4, respectively. A lactate production:glucose metabolism ratio greater than 2 implies that glycogen is being mobilized in addition to extracellular glucose. The decrease in tissue glycogen in Atlantic cod and rainbow trout hearts is consistent with this contention. It is likely that lactate production by lumpfish and short-horned sculpin hearts is being supported by glycogenolysis but that this is being missed in the analysis because of the variability in glycogen levels and/or the low rate of lactate production relative to the amount of glycogen.

It is generally viewed that in most fish species, hypoxia results in an increase in cardiac anaerobic metabolism involving both glycogen and glucose as metabolic fuels to support lactate production (Gamperl and Driedzic, 2009). This scenario is pulled together from a range of studies each contributing parts of the story but without one single experiment actually measuring rates of glucose metabolism under hypoxia. The current investigation with isolated heart preparations from Atlantic cod and rainbow trout supports the accepted paradigm. Here it is shown that in hearts from these species receiving perfusate containing a physiological level of extracellular glucose (5 mmol l⁻¹), hypoxia leads to an increase in glucose metabolism as assessed by the rate of release of ³H₂O from [2-³H]-glucose. As well, there was a concomitant increase in the rate of lactate production and glycogen utilization. Studies were also conducted with lumpfish and short-horned sculpin, which have normal plasma glucose levels of less than 1 mmol l⁻¹. Hearts from these species perfused with medium containing physiological levels of glucose, 1 mmol l⁻¹ for lumpfish and 0.5 mmol l⁻¹ for short-horned sculpin, had rates of glucose metabolism similar to Atlantic cod and rainbow trout under normoxia but failed to show any increase in glucose metabolism when subjected to hypoxia. Similarly, although the rate of lactate production appears to be enhanced under hypoxia, rates are still many fold lower than observed in Atlantic cod and rainbow trout. These findings lead to the conclusion that cardiac anaerobic metabolism from plasma glucose is compromised in species with especially low levels of plasma glucose. This is possibly due to a reduced glucose gradient from the extracellular to the intracellular space. This conclusion is supported by the observation that hearts from Atlantic cod perfused with medium containing a low concentration of glucose (0.5 mmol l⁻¹) did not show an increased in glucose metabolism. The position of limited rates of glucose metabolism is based upon isolated hearts performing similar levels of cardiac work both across species and under both oxygen levels. As contractility is the major determinant of energy demand, the next stage in the progression of this experimental program should be to assess rates of glucose metabolism at species-specific levels of pressure development and frequency under normoxia and hypoxic bradycardia.

The response of the isolated heart from short-horned sculpin matches the whole-animal response. Short-horned sculpin subjected to hypoxia, at the limit of their tolerance, had rates of 2-deoxyglucose uptake and lactate levels similar to those of control animals. Anaerobic metabolism was not activated; instead, heart rate and cardiac output were reduced, a response that would contribute to a decrease in energy demand (MacCormack and Driedzic, 2007). The main message of the present study is that in some fish, a low level of extracellular glucose compromises heart anaerobic energy production; therefore, hypoxic survival of this tissue must depend on other factors, such as decreases in contractility, ion transport or protein turnover. In European sea bass (Dicentrarchus labrax), whole-animal hypoxia tolerance is associated with the hypoxic performance of ventricle strips (Joyce et al., 2016). If heart hypoxia tolerance proves to be an important determinant of whole-animal sensitivity to oxygen deprivation, then the current findings of the consequences of low blood glucose on heart energy metabolism become especially important as many marine ecosystems are facing oxygen reduction.

The authors thank Dr K. Gamperl for the use of instrumentation to measure oxygen consumption. We also thank the field services unit (Department of Ocean Sciences) for the collection of short-horned sculpin, D. Boyce (Department of Ocean Sciences) for the supply of Atlantic cod and lumpfish, and Chris Dawe (Marine Institute, Memorial University) for rainbow trout.