Hypoxic acclimation negatively impacts the contractility of steelhead trout (Oncorhynchus mykiss) spongy myocardium

Abstract

Cardiac stroke volume (S_V) is compromised in Atlantic cod and rainbow trout following acclimation to hypoxia (i.e., 40% air saturation; ~8 kPa O₂) at 10–12°C, and this is not due to changes in heart morphometrics or maximum achievable in vitro end-diastolic volume. To examine if this diminished S_V may be related to compromised myocardial contractility, we used the work-loop method to measure work and power in spongy myocardial strips from normoxic- and hypoxic-acclimated steelhead trout when exposed to decreasing Po₂ levels (21 to 1.5 kPa) at several frequencies (30–90 contractions/min) at 14°C (their acclimation temperature). Work required to lengthen the muscle, as during filling of the heart, was strongly frequency dependent (i.e., increased with contraction rate) but was not affected by hypoxic acclimation or test Po₂. In contrast, although shortening work was less frequency dependent, this parameter and network (and power) 1) were consistently lower (by ~30–50 and ~15%, respectively) in strips from hypoxic-acclimated fish and 2) fell by ~40–50% in both groups from 20 to 1.5 kPa Po₂, despite the already-reduced myocardial performance in the hypoxic-acclimated group. In addition, strips from hypoxic-acclimated trout showed a poorer recovery of net power (by ~15%) when returned to normoxia. These results strongly suggest that hypoxic acclimation reduces myocardial contractility, and in turn, may limit S_V (possibly by increasing end-systolic volume), but that this diminished performance does not improve the capacity to maintain myocardial performance under oxygen limiting conditions.

INTRODUCTION

Hypoxia (low water oxygen levels that compromise physiological function; 44) can occur naturally in aquatic environments due to biological oxygen demand, fluctuations in the tidal cycle, density stratification, changing wind patterns, and the isolation of water bodies for prolonged periods of time (e.g., in tidepools) (10, 44), and these events have become more prevalent in recent years due to accelerated climate change (1, 5). Research on the effects of acute and chronic environmental hypoxia is a current focus in the field of fish physiology, and this research has demonstrated a number of impacts of hypoxia on fish behavior and physiology, and highlighted interspecific differences in hypoxia sensitivity and tolerances (45). For example, sculpins, which inhabit intertidal zones exposed to cyclical and frequent hypoxic episodes, display aerial surface respiration and aerial emergence, along with a suite of physiological and biochemical adaptations (44). The epaulette shark (Hemiscyllum ocellatum) is able to depress its metabolism and increase ventilatory frequency in response to graded hypoxia (36). The freshwater Crucian carp (Carassius carassius), which has the ability to endure anoxia for weeks at a time, can maintain cardiac function and produce ethanol as an end product of anaerobic metabolism instead of lactate to avoid self-acidosis (51). Hypoxia intolerant fish such as Atlantic cod (Gadus morhua) (41) and rainbow trout (Oncorhynchus mykiss) (37) also encounter hypoxic waters (6, 31, 32). However, the effects of prolonged hypoxia and hypoxic-acclimation (i.e., weeks to months of exposure to low O₂ conditions) on hypoxia-intolerant species such as these has only received limited attention (2, 9, 20, 35, 40–42).

Recent studies on Atlantic cod and steelhead trout report that exposure to chronic moderate hypoxia (>6 wk at 8.5 kPa) greatly diminished the capacity of these fish to increase stroke volume (S_V), and thus, cardiac output (Q) when swum to exhaustion or exposed to an acute incremental temperature increase until the fish’s critical thermal maximum (35, 41). Given that this reduced cardiac pumping capacity was also evident in in situ Atlantic cod hearts [i.e., eliminating the possibility that the decrease in S_V was largely due to alterations in neuronal or humoral regulation of cardiac function or venous (filling) pressure (42)], and not related to changes in relative ventricular mass, percent compact myocardium or maximum achievable in vitro stroke volume (35, 42), other aspects of cardiac function must be dysregulated following exposure to chronic hypoxia. Syme et al. (53) investigated the interactive effects of graded (acute) hypoxia and elevated temperature (10 vs. 20°C) on the work and power output of cod ventricular myocardial strips. These authors showed that shortening power decreased across all Po₂ levels (94.5 to 7 kPa) at 10°C, whereas at 20°C there was no change in shortening power but a dramatic increase in lengthening power (i.e., the muscle became stiffer) when Po₂ fell below 21 kPa. These data suggest that the ventricular myocardium of hypoxic-acclimated fish can either not fully eject blood into the circulation due to decreased contractility (i.e., end-systolic volume is no longer zero) or that an increase in myocardial stiffness makes it difficult to fill the ventricle (i.e., end-diastolic volume decreases). Indeed, this latter suggestion is supported by the results of Petersen and Gamperl (42) who showed that the input pressure required to maintain resting in situ cardiac output was consistently higher by ~0.2–0.3 kPa in hypoxic-acclimated cod, and that maximum cardiac function was reduced in this group independent of input pressure.

To examine the hypothesis that the diminished S_V in hypoxic-acclimated fish is associated with alterations in myocardial performance (i.e., shortening and/or lengthening work and power), we acclimated steelhead trout (seawater-acclimated rainbow trout) to a hypoxic Po₂ of ~8.5 kPa or to normoxia (21 kPa) for 8–12 wk at 13–14°C, and used cycling strips from the spongy myocardium to measure the muscle’s contractile performance at Po₂ levels of 21, 13.5, 10.5, 6.9, 4.2, and 1.5 kPa across a range of contraction frequencies (30–90 contractions/min). The Po₂ values of 13.5 and 6.9 kPa approximate the Po₂ values in the arterial and venous blood, respectively, of normoxic fish (50, 56), 1.5 kPa is the blood Po₂ that has been suggested as the lower limit at which trout cardiac function can be sustained (14, 52), and 10.5 and 4.2 kPa provide intermediate (additional) values for the examination of the relationship between Po₂ and myocardial function in hypoxic-acclimated fish.

MATERIALS AND METHODS

Experimental Animals

The female steelhead trout (Oncorhynchus mykiss) (mean 0.61 ± 0.03 kg; range 0.32–0.82 kg) used in this study were initially housed for approximately two months at the Dr. Joe Brown Aquatic Research Building (Ocean Sciences Centre, Memorial University of Newfoundland, St. John’s, Newfoundland, Canada) in a 3-m³ tank supplied with aerated, 10–11°C, seawater. Photoperiod was maintained at 12 h light:12 h dark, and the trout were fed a commercial salmon diet three times a week at a ration of 1.5% of body mass per day. All procedures were approved by the University of Calgary’s and MUN’s (protocol no. 15–89-KG) Institutional Animal Care Committees and followed the guidelines of the Canadian Council on Animal Care.

Holding and Acclimation Conditions

After the initial holding period described above, 80 trout were divided equally between two ~1.2 m³ square tanks equipped with wooden lids to control for external stimuli (noise, human presence, etc.) that were supplied with 13–14°C seawater with a water oxygen partial pressure ($P{w}_{O_2}$) of 19–20 kPa (i.e., >95% air saturation). After a 2-wk acclimation period, the water oxygen level in the tank that was to be made hypoxic was slowly lowered to the desired value (8.4 ± 0.1 kPa; 40% air saturation) over a period of 2 wk. This was accomplished by: 1) gradually reducing the flow to the tank from 10 to 5 L/min (allowing fish metabolism to partially reduce the water O₂ saturation); and 2) using a custom-designed solenoid valve system (Electronic Workshop, Memorial University of Newfoundland). This system monitored the oxygen saturation in the tank continuously by pumping tank water through an external circuit of tubing (Tygon Food, 6–419, Cole Parmer) that contained a galvanic oxygen electrode (CellOx 325, WTW, Weilheim, Germany) placed in a D201 flow cell (WTW). The oxygen probe was connected to an oxygen meter (Oxi 340, WTW), which was interfaced with two solenoid valves: one that bubbled pure N₂ into the tank when $P{w}_{O_2}$ reached the upper limit of 8.8 kPa; the other bubbled air into the tank when oxygen levels reached 8 kPa. This allowed for control of the oxygen levels in the hypoxic tank within a narrow O₂ range (±0.4 kPa). During the acclimation period, fish in both tanks were fed daily by hand with commercial trout pellets at 1% body mass per day. However, if the hypoxic fish failed to eat their full ration, the normoxic fish were only given the same amount of food. Ammonia and nitrite levels were also checked weekly (initially) or biweekly using a Lamotte Smart3 Colorimeter (and appropriate test kits; Lamotte, Chestertown, MD) to ensure that these compounds did not exceed 0.05 and 0.5 mg/L, respectively.

Isolated Myocardial Strips

Fish from each treatment (N = 9) were arbitrarily selected from their holding tank following 8–12 wk of acclimation, swiftly killed by cerebral percussion, measured for length and body mass, and had their hearts removed. The ventricle was separated from the bulbus arteriosus and atrium, weighed, cut in half lengthwise, and rinsed in ice-cold physiological saline for marine teleosts (41), pH 7.6 at 20°C. This saline contained (in g/L) 10.5 NaCl, 0.49 MgSO₄·7H₂O, 0.37 KCl, 0.33 CaCl₂·2H₂O, 0.14 NaH₂PO₄·H₂O, 1.84 sodium TES base (C₆H₁₄NO₆SNa), 0.59 TES acid (C₆H₁₅NO₆S), and 1.0 glucose. The ventricular halves were pinned to the bottom of a temperature-controlled (4°C) dissecting dish containing saline, and a small segment of spongy trabecular muscle (~5–10 mm in length and <1 mm² in cross section) was isolated from the inner (luminal) surface of the ventricle under a dissecting microscope. Muscle segments (3.1 ± 2.3 mg wet mass and 4.9 ± 1.5 mm resting length) were selected so that the majority of fibers ran parallel to the long axis of the preparation, and there was minimal branching along their length. Once dissected free, a short piece of 6–0 silk suture was tied to each end of the strip, and the strips were attached to the arm of a servomotor (300C-LR; Aurora Scientific, Aurora ON, Canada) on one end and a force transducer (404A; Aurora Scientific) on the other end using the methods as described in Syme et al. (53). The muscle segments were bathed in physiological saline, and the temperature was maintained at 14 ± 0.2°C using Peltier thermoelectric modules and a temperature controller (TC-24–12; TE Technology, Traverse City, MI). A custom program using LabView software (National Instruments, Austin, TX) controlled a 12-bit analog/digital converter card (PCI MIO 16E 4; National Instruments) that regulated the stimulator and servomotor (5 kHz D/A output), and collected muscle force, muscle length (servomotor arm position), and stimulus signals (1 kHz A/D input).

Saline Po₂ was maintained at the below levels by bubbling a mixture of O₂ and N₂ gas delivered from a Wöstoff gas mixing pump (DIGAMIX 6KM301, Bochum, Germany) into a reservoir of saline that was then drained via stainless-steel tubing into the muscle chamber, and the Po₂ in the muscle chamber was continuously monitored using a calibrated fiber optic dipping oxygen probe (PSt3; PreSens, Regensburg, Germany) and oxygen meter (Fibox 3; PreSens). Complete turnover of the saline in the 30-mL bath occurred approximately once per minute. The Po₂ of the bath was sequentially set at 21 ± 0.5, 13.5 ± 0.5, 10.5 ± 0.5, 6.9 ± 0.5, 4.2 ± 0.5 and 1.5 ± 0.5 kPa during the experiment, with muscle strips allowed to equilibrate for 10–15 min at each oxygen level. Recovery of muscle work and power was also assessed 10–15 min after the saline in the bath was returned to 21 kPa (normoxic conditions).

Tonic adrenergic stimulation (i.e., ~5 nM epinephrine) was omitted from these experiments given that it is not possible to maintain constant levels over the course of these lengthy and complex experiments (i.e., it would have increased data variability), and so direct comparisons could be made with Carnevale et al. (unpublished observations; Carnevale C, Syme DA, and Gamperl AK); these experiments examined the effects of hypoxic-acclimation on NO-mediated myocardial function and epinephrine is known to interfere with these effects (39). Furthermore, Roberts and Syme (46) recently showed that net work was not significantly different in cycling muscle strips from the spongy myocardium of rainbow trout at 10% air saturation and 14°C when tested with 0 versus 5 nM epinephrine in the bath.

Measuring Muscle Work and Power

Initially, the length of each preparation was increased systematically until developed twitch force, elicited by a 1-ms supramaximal shock (~10 V), was maximal. Stimulus shocks were applied using a stimulator (Isostim A320, World Precision Instruments, Sarasota, FL) and platinum plates placed on either side of the preparation. Muscle length was then decreased by 5% to reduce passive stress and to better mimic the normal function of cardiac muscle, which operates on the ascending limb of the force-length relationship. Furthermore, Harwood et al. (21) showed no significant difference in work output between muscles operating at 95% versus 100% of the length giving maximal force. This resting length (L_opt) was then used for the remainder of experiments. In a few instances, where passive force rose steeply when strain was applied to the muscle during work loop experiments, suggesting that muscle length was too long, muscle length was decreased slightly to maximize net work output.

Muscle work and power were measured using the work-loop method (26, 54). To measure work, muscle length was oscillated about L_opt in a sinusoidal trajectory by the servomotor, mimicking muscle strain experienced during the cardiac cycle. The amplitude of strain (related to S_V) was set at ±5% of L_opt because this has been shown to elicit maximum work output in spongy ventricular muscle preparations (53); an observation confirmed in our study. The proportion of the period of the sinusoidal strain cycle that comprised shortening versus lengthening was adjusted at each contraction frequency to maximize net work output. This was done to reflect changes in the proportion of the cardiac cycle occupied by systole versus diastole as heart rate changes: i.e., approximately 30% of the cycle period was muscle shortening (the remaining 70% being lengthening) at 30 contractions/min, 50% shortening at 50 and 70 contractions/min, and 80% shortening at 90 contractions/min. The timing of muscle stimulation relative to the imposed length changes (i.e., stimulation phase) was also adjusted at each frequency used (30–90 contractions/min, see below) to achieve maximum net work output. The optimal phase values varied between preparations and frequency and ranged from −10% to 25% across the range of frequencies used.

Work (in J) provides a measure of the mechanical energy produced during each beat of the heart, and was calculated as the integral of muscle force with respect to length change. The work done by the servomotor to lengthen the muscle in each cycle (lengthening work; analogous to filling work in an intact heart), the work done by the muscle when it contracts during each cycle (shortening work; analogous to stroke work), and net work (the difference between shortening work and lengthening work) were all measured. Power (in W) provides a measure of the sustained rate of mechanical energy produced and was calculated as the product of the work done per cycle (shortening, lengthening, or net) and contraction frequency (Hz). Mass-specific work (J/kg) and power (W/kg) were calculated for each muscle strip based on the mass of each preparation.

Contractile performance was measured at 4 randomized contraction frequencies (30, 50, 70, 90 contractions/min) at each of the Po₂ values noted above, and again once the strips were returned to 21 kPa Po₂ (normoxia). These frequencies span the physiological range of heart rates reported in trout at similar temperatures (0.5–1.5 Hz) (43), with 1.5 Hz (90 contractions/min) also being the maximum contraction frequency that the strips could be subjected to before they were unable to maintain consistent work output at every cycle (suggesting that full mechanical restitution was not maintained at higher frequencies). At every combination of frequency and Po₂, 30 isometric contractions were performed before the measurements of work output to ensure the preparation was stable, and then a series of 30 consecutive work loops were recorded. Furthermore, pre-experimental trials determined that there was no decrement in function when the 2-h protocol was performed under normoxic conditions. This is consistent with Roberts and Syme (46), who performed similar studies, and several studies using in situ trout hearts (e.g., 16).

Data and Statistical Analyses

At the conclusion of each experiment, the muscle strips were removed from the bath and trimmed of any tissue past the silk ties and obviously dead tissue, blotted on filter paper to remove surface moisture and weighed on a microbalance (Mettler UMT2; Mettler-Toledo, Columbus, OH). Vital staining was not performed to more accurately determine the amount of viable tissue. Thus, muscle mass is likely an overestimate of viable tissue mass and mass-specific work values are likely to be an underestimate of muscle performance.

A 20-point median filter was applied to every record of muscle force and length before analyses to eliminate background noise, and checks were made to ensure that the filter did not distort the traces. Performance of the working preparations was measured as the average work done during the final 10 contractions of every series of 30. The number of replicates (N) reported for this experiment refers to the number of trabeculae (strips) tested, each originating from a different fish. In these experiments, parameters measured in each muscle strip at 21 kPa served as a reference point for the calculation of relative changes.

A split-split plot mixed general linear model was performed using R to examine the effects of the three controlled variables (acclimation condition, contraction frequency, and saline Po₂,) and one random variable (strip) on muscle performance parameters. Ventricular trabeculae were not controlled for length. This contributed to variability between strips in work done, but this was accounted for in the main model by including strip as a random factor. These analyses were followed by Holm-Sidak pairwise-comparisons. When significant interactions were present, a two-way analysis of variance, followed by Holm-Sidak pairwise-comparisons, was performed using SPSS (version 11.0). Main factors of acclimation and Po₂ (at each frequency), or acclimation and frequency (at each Po₂ level) were used. Fish mass, ventricular mass, relative ventricular mass (RVM in % = (ventricle mass/body mass) × 100), fish length and condition factor (K = (mass/length³) × 100) were compared between acclimation groups using two-tailed t tests. All statistical analyses were performed with the level of statistical significance set at P < 0.05. Graphs and figures were created using GraphPad Prism 5 (GraphPad Software, La Jolla, CA), and all data in the text and figures are expressed as means ± SE.

RESULTS

Cardiac Morphometrics and Fish Size

Acclimation to hypoxia did not significantly (P > 0.05) affect the trout’s length, ventricular mass, or condition factor but had a significant effect on fish mass and relative ventricular mass (RVM) (Table 1). Hypoxic fish weighed ~100 g (20%) less than the normoxic fish, whereas RVM was ~0.02% (27%) higher in hypoxic fish (Table 1). Furthermore, ~30% of the hypoxic-acclimated fish did not survive the first 3–4 wk of exposure to water of 40% air saturation, whereas >95% of the normoxic-acclimated group survived this period.

Table 1.

Body and cardiac morphometrics for steelhead trout

	Animal Mass, g	Fork Length, cm	Condition Factor, K	Ventricular Mass, g	Relative Ventricular Mass, %
Normoxia	678.6 ± 34.5	35.2 ± 0.7	1.55 ± 0.04	0.53 ± 0.04	0.079 ± 0.003
Hypoxia	548.9 ± 45.7*	33.2 ± 0.9	1.48 ± 0.07	0.54 ± 0.04	0.100 ± 0.006*

Values are means ± SE (n = 9) for fish acclimated to either normoxia (water oxygen partial pressure of ~21 kPa) or hypoxia (8–9 kPa) for 8–12 wk. Statistical differences between groups were identified using unpaired t tests.

^*Significant difference between acclimation groups (P < 0.05).

Effects of Acclimation, Contraction Frequency, and Po₂ on Myocardial Performance

Mass-specific work.

At a saline Po₂ of ~21 kPa (normoxia), the shortening work of strips from hypoxic-acclimated fish was comparable to that measured in the normoxic-acclimated group at the lowest frequency (30 contractions/min) (Fig. 1, A and B). However, there was a distinct negative effect of frequency on shortening work in the hypoxic- but not normoxic-acclimated group at 21 kPa (e.g., see inset Fig. 1B) so that at 90 contractions/min shortening work in the hypoxia group was only 73% of that at 30 contractions/min, and only ~60% of that measured in the normoxic-acclimated group. Shortening work declined with decreasing Po₂ in both acclimation groups (Fig. 1, A and B), but interestingly, the fall in shortening work between 21 and 13.5 kPa was inversely related to the contraction frequency in the hypoxic-acclimated group, and this resulted in shortening work being: 1) the same at all contraction frequencies at 13.5 kPa: and 2) ~35% lower than measured in strips from normoxic-acclimated individuals. From 13.5 to 1.5 kPa, there was no effect of frequency on shortening work, and the depressive effect of Po₂ on shortening work was similar between acclimation groups.

Fig. 1.

Shortening (A and B), lengthening (C and D) and net work (E and F) done by ventricular trabeculae from normoxic-acclimated (solid symbols; left) and hypoxic-acclimated (open symbols; right) steelhead trout when exposed to declining O₂ levels and four different contraction frequencies at 14°C. Lengthening work is shown as a negative value to reflect the work that was done by the servomotor to stretch the muscle. The trabeculae were cycled at 30 (black ● and ○), 50 (blue ■ and □), 70 (green ▲ and △), and 90 (red ◆ and ◇) contractions/min. Values are means ± SE; n = 9. Shortening work done by ventricular trabeculae from hypoxic-acclimated steelhead trout at 21 kPa Po₂ at the four contraction frequencies (inset). Differences between the main factors (acclimation, frequency and oxygen level) were analyzed using a split-split plot, mixed, general linear model followed by two-way ANOVAs when interactions were identified (see Table 2). A one-way repeated measure ANOVA followed by Tukey’s post hoc test were used to specifically test whether shortening work at 21 kPa was frequency dependent in the hypoxic-acclimated group (see inset in B). Dissimilar letters indicate a significant difference at P < 0.05.

Neither acclimation condition (normoxia vs. hypoxia), or Po₂, had a significant effect on lengthening work. However, there was a distinct frequency effect (P < 0.05) (Fig. 1, C and D), where work done by the servomotor to lengthen the muscle (i.e., negative work) increased with contraction frequency, and was ~150% greater at 90 contractions/min versus 30 contractions/min in both groups.

Because of the additive effects of shortening work and lengthening work on net work, and the number of interactions between parameters (see Table 2), interpretation of the effects of acclimation, Po₂, and frequency on net work was complicated. Nonetheless, there are a number of apparent (and significant) trends in the data. First, the effect of frequency on net work was qualitatively similar in strips from normoxic- and hypoxic-acclimated fish. Net work decreased by ~30–60% as contraction frequency increased from 30 to 90 contractions/min. Second, exposure to acute reductions in Po₂ had comparable effects on net work in both groups, where a decrease from 21 to 1.5 kPa resulted in an ~40% decrease in net work. Finally, strips from hypoxic-acclimated fish produced ~30–40% less mass-specific net work when compared with those from normoxic-acclimated fish at similar Po₂ values and frequencies.

Table 2.

Results of the statistical analyses that examined the effect of hypoxic acclimation on myocardial contractility

Parameter	Method	Factor(s)	F	d.f.	P
sWork	1	Acclimation	2.987	1	0.103
		Freq.	7.636	3	5.83×10−5
		Po2	105.094	5	<2×10−16
		Acclimation*Freq.	1.292	3	0.276
		Acclimation* Po2	3.311	5	0.006
		Freq* Po2	1.203	15	0.267
		Acclimation*Freq.* Po2	0.947	15	0.512
lWork	1	Acclimation	0.002	1	0.921
		Freq.	45.602	3	<2×10−16
		Po2	2.501	5	0.030
		Acclimation*Freq.	1.333	3	0.263
		Acclimation* Po2	0.708	5	0.618
		Freq* Po2	0.185	15	0.999
		Acclimation*Freq.* Po2	0.082	15	1
nWork	1	Acclimation	3.465	1	0.081
		Freq.	86.991	3	<2×10−16
		Po2	87.929	5	<2×10−16
		Acclimation*Freq.	4.831	3	0.0026
		Acclimation* Po2	5.092	5	0.00016
		Freq* Po2	1.73	15	0.043
		Acclimation*Freq.* Po2	0.57	15	0.897
sPower	1	Acclimation	2.721	1	0.118
		Freq.	357.164	3	<2×10−16
		Po2	33.651	5	<2×10−16
		Acclimation*Freq.	12.77	3	6.05×10−8
		Acclimation* Po2	1.862	5	0.1
		Freq* Po2	0.454	15	0.961
		Acclimation*Freq.* Po2	0.306	15	0.995
lPower	1	Acclimation	0	1	0.985
		Freq.	110.755	3	<2×10−16
		Po2	0.609	5	0.693
		Acclimation*Freq.	0.167	3	0.919
		Acclimation* Po2	0.08	5	0.995
		Freq* Po2	0.094	15	1
		Acclimation*Freq.* Po2	0.018	15	1
nPower	1	Acclimation	3.396	1	0.083
		Freq.	72.914	3	<2×10−16
		Po2	43.662	6	<2×10−16
		Acclimation*Freq.	14.923	3	3.02×10−9
		Acclimation* Po2	3.41	6	0.003
		Freq* Po2	0.437	18	0.979
		Acclimation*Freq.* Po2	0.419	18	0.983
RelsPower	1	Acclimation	0.004	1	0.842
		Freq.	36.092	3	<2×10−16
		Po2	118.977	5	<2×10−16
		Acclimation*Freq	9.033	3	9.87×10−6
		Acclimation* Po2	0.342	5	0.887
		Freq* Po2	1.484	15	0.109
		Acclimation*Freq.* Po2	0.556	15	0.907
RellPower	1	Acclimation	0.288	1	0.599
		Freq.	0.045	3	0.987
		Po2	5.823	5	3.52×10−5
		Acclimation*Freq.	3.883	3	0.009
		Acclimation* Po2	0.647	5	0.664
		Freq* Po2	0.671	15	0.813
		Acclimation*Freq.* Po2	0.282	15	0.996
RelnPower	1	Acclimation	0.627	1	0.439
		Freq.	7.102	3	0.0001
		Po2	136.856	6	<2×10−16
		Acclimation*Freq.	0.291	3	0.831
		Acclimation* Po2	1.398	6	0.224
		Freq* Po2	0.517	18	0.93
		Acclimation*Freq.* Po2	0.144	18	0.999
sPowerRec	2	Acclimation	0.643	1	0.426
		Freq.	1.55	3	0.211
		Acclimation*Freq.	0.295	3	0.62
lPowerRec	2	Acclimation	0.132	1	0.718
		Freq.	0.199	3	0.897
		Acclimation*Freq.	0.295	3	0.829
nPowerRec	2	Acclimation	6.84	1	0.011
		Freq.	0.65	3	0.586
		Acclimation*Freq.	0.096	3	0.962

Results of the split-split plot, mixed general linear model (method 1) analyses that were used to examine the effects of acclimation group, contraction frequency and Po₂ on mass-specific shortening (s), lengthening (l) and net (n) work and power, as well as relative (Rel) shortening, lengthening and net power. Results of the split plot, mixed, general linear model (method 2) analyses that examined the effects of acclimation group and frequency on the recovery (Rec) of shortening, lengthening and net power are also included.

Mass-specific power.

As expected, power was highly frequency dependent (Fig. 2). This resulted in a positive (not negative, as seen for work) relationship between contraction frequency and shortening power in both acclimation groups (Fig. 2, A and B). Strips from hypoxic-acclimated fish produced, on average, ~30% less shortening power than normoxic-acclimated strips. In both acclimation groups there was an ~25% decrease in shortening power with a decrease in Po₂ across the range studied. Acclimation condition and Po₂ had no significant effect on lengthening power (Fig. 2, C and D), but the negative relationship between frequency and lengthening power was even more pronounced in comparison to lengthening work (see Fig. 1, C and D). Net power increased with contraction frequency in both acclimation groups but to a much lower extent in strips from hypoxic-acclimated fish (Fig. 2, E and F). For example, net power increased by ~120% between 30 and 90 contractions/min in strips from normoxic-acclimated fish, whereas this parameter only increased by ~20% over this frequency range in strips from hypoxic-acclimated trout (Fig. 2). Despite this, the depressive effect of decreasing Po₂ on net power was relatively similar between the two acclimation groups (Fig. 2, E and F).

Fig. 2.

Shortening (A and B), lengthening (C and D) and net power (E and F) done by ventricular trabeculae from normoxic-acclimated (solid symbols; left) and hypoxic-acclimated (open symbols; right) steelhead trout when exposed to declining O₂ levels and four different contraction frequencies at 14°C. Lengthening power is shown as a negative value to reflect the power that was done by the servomotor to stretch the muscle. The trabeculae were cycled at 30 (black ● and ○), 50 (blue ■ and □), 70 (green ▲ and △) and 90 (red ◆ and ◇) contractions/min. Differences between the main factors (acclimation, frequency and oxygen level) were analyzed using a split-split plot, mixed, general linear model followed by two-way ANOVAs when interactions were identified (see Table 2). Values are means ± SE; n = 9.

Relative myocardial power.

To further emphasize some of the relationships between contractile performance, Po₂, and contraction frequency, as well as to remove variability in the results due to errors in estimating viable mass of the preparations, relative power is also shown (Fig. 3). Relative shortening power declined by ~30% over the range of Po₂ values studied at every frequency in the normoxic-acclimated strips. However, in the hypoxic-acclimated group, the relative reduction in shortening power with lowered Po₂ was frequency dependent, decreasing by ~35% at 30 contractions/min but only by 15% at 90 contractions/min. Again, this response was largely due to the frequency-dependent response of the strips to lowering the saline Po₂ from 21 to 13.5 kPa. In both acclimation groups there was no significant effect of frequency or Po₂ on relative lengthening power (Fig. 3, C and D). Similarly, there was no effect of acclimation condition or frequency on relative net power, but relative net power decreased by 40–50% in both groups as oxygen levels fell (Fig. 3, E and F).

Fig. 3.

Relative shortening (A and B), lengthening (C and D) and net power (E and F) done by ventricular trabeculae from normoxic-acclimated (filled in symbols; left) and hypoxic-acclimated (open symbols; right) steelhead trout when exposed to declining O₂ levels and four different contraction frequencies at 14°C. Values of power are expressed relative to the values initially obtained at 21 kPa Po₂ for each preparation. Lengthening power is shown as a negative value to reflect the power that was used by the servomotorto stretch the muscle. The trabeculae were cycled at 30 (black ● and ○), 50 (blue ■ and □), 70 (green ▲ and △) and 90 (red ◆ and ◇) contractions/min. Differences between the main factors (acclimation, frequency, and oxygen level) were analyzed using a split-split plot, mixed, general linear model followed by two-way ANOVAs when interactions were identified (see Table 2). Values are expressed as % change relative to the power output measured at normoxia for each strip. Values are means ± SE; n = 9.

Effects of Acclimation Condition and Contraction Frequency on Recovery

The recovery of shortening, lengthening and net power upon returning to 21 kPa saline was compared between the acclimation groups at the four contraction frequencies (Fig. 4). Neither frequency or acclimation condition had a significant effect on the recovery of lengthening or shortening power (P > 0.05) (Table 2, Fig. 4). In contrast, the recovery of net power was consistently ~20% lower in trabeculae from hypoxic-acclimated fish (P < 0.05), although this difference only reached statistical significance at 30 contractions/min (Table 2, Fig. 4).

Fig. 4.

Recovery of shortening (A), lengthening (B), and net power (C) in myocardial trabeculae from normoxic-acclimated (●) and hypoxic-acclimated (□) steelhead trout at 14°C and four contraction frequencies. Values are expressed as power measured at 21 kPa Po₂ after exposure to reduced levels of oxygen relative to the power output measured initially at 21 kPa. A split plot, mixed, general linear model analysis was used to examine the effects of acclimation and frequency on the recovery of shortening, lengthening and net power. *Significant difference (P < 0.05) between normoxic- and hypoxic-acclimated groups at a particular frequency. Values are means ± SE; n = 9.

DISCUSSION

Effects of Chronic Hypoxia on Fish Size and Cardiac Morphometrics

Acclimation to chronic hypoxia had no effect on fish mass, length, condition factor, ventricular mass or RVM in prior studies on Atlantic cod (40–42) and rainbow trout (35) in our laboratory. However, in the current study, hypoxic-acclimated steelhead trout weighed significantly less, and their RVM was significantly greater (Table 1) despite our attempt to provide equal amounts of food to both acclimation groups. Hypoxia has been shown in numerous studies to decrease appetite and growth rate (6, 7, 58), and a previous study on rainbow trout acclimated to hypoxia (10.5 kPa for 28 days) reported ~50% reductions in food intake and growth rate as compared with normoxic individuals when fed to satiety (19). The one difference between this study and Petersen and Gamperl (40–42) and Motyka et al. (35) was the temperature at which the fish were held (13–14 versus 10–11°C, respectively), and it is possible that the higher temperature used in this study contributed to the lower growth rate and mortalities reported. Basal metabolic rate increases and water oxygen content decreases with increased temperatures, and so the hypoxic challenge was actually more severe in the current experiment. Both of these effects could have significantly constrained the trout’s aerobic scope, and this parameter has a large influence on the growth rate of fish under hypoxic conditions (8). Furthermore, it is very likely that this temperature difference explains the higher RVM in hypoxic-acclimated fish in this experiment. Ventricle size has been linked to cardiac work in prior studies on rainbow trout (15, 24), and the temperature-dependent increase in cardiac workload associated with rearing at 13–14°C (i.e., increased metabolic demand and decreased oxygen availability) might have required the ventricle to be larger relative to body mass to maintain oxygen delivery under hypoxic conditions. Indeed, the data of Anttila et al. (3) suggest that there is an interactive effect of temperature and hypoxia on cardiac morphology. These authors exposed Atlantic salmon (Salmo salar) and Arctic char (Salvelinus alpinus) to daily (overnight) bouts of hypoxia (65% oxygen saturation) for 4 wk at 14.9 versus 7.7°C and showed that although RVM was not different at either temperature, the percentage of compact myocardium was significantly higher in hypoxic- versus normoxic-acclimated fish at 14.9°C. Alternatively, it is conceivable that the heart is protected during periods of impaired growth associated with reduced food intake, which would result in increased RVM through reduced body mass rather than changes in heart mass. Gamperl et al. (unpublished observations; Genge G, Rodnick KJ, and Gamperl AK) deprived Atlantic cod of food for 8–10 wk at 10°C and showed that although the fish weighed significantly less, they had a greater RVM and maintained similar amounts of ventricular protein as compared with fed cod, as appeared to be the case in the present study with steelhead trout (Table 1).

Frequency Dependence of Myocardial Performance

The net work and power produced by the ventricular trabeculae of normoxic-acclimated trout working in normoxic saline (21 kPa Po₂) averaged 0.300 J/kg and 0.280 W/kg, respectively, across the four contraction frequencies employed (Figs. 1E and 2E). These values fall within the range reported for teleosts (21, 22, 49, 53) and are very similar to those recently reported by Roberts and Syme (46) for rainbow trout. However, they are considerably higher than values reported by Shiels et al. (49) for rainbow trout (0.13 J/kg and 0.12 W/kg at 12°C) and by Syme et al. (53) for trabeculae from the spongy myocardium of Atlantic cod (0.15 J/kg and 0.09 W/kg at 10°C). Our protocol was specifically designed to maximize work and power (including the adjustment of the proportion of the strain cycle that comprised muscle shortening and lengthening) and conducted at warmer temperatures, and these differences may explain the higher values reported in our study. In contrast, our values for work and power were less than those reported by Harwood et al. (21, 22) (0.9–1.4 J/kg and 0.8–1.2 W/kg at 15°C). This may be due to where the strips were taken from the ventricle. Based on the description of where the myocardial strips were dissected from (ventral apex of the heart), it is probable that the preparations in the studies by Harwood et al. (21) were from the compact layer versus the spongy layer in the present study, or had a different fiber orientation. Although, Roberts and Syme (46) recently reported that the net power of strips from these two tissue types was not statistically different.

Contractile force (and thus work) was expected to decrease substantially with increasing contraction frequency due to the well-known negative force-frequency relationship shown for isometrically contracting muscles (11, 12, 25, 29, 34, 48) and cycling muscle strips (21, 22). However, a strong negative relationship between shortening work and contraction frequency was only evident in strips from hypoxic-acclimated trout at a test Po₂ of 21 kPa (normoxia) (see below) and not at any Po₂ in normoxic-acclimated trout. This is an interesting observation. The symmetry of the sinusoidal trajectory of muscle length change was adjusted in the current study to account for changes in the duration of cardiac contraction (and in the duration of systole and diastole as heart rate increased), as would occur in a beating heart, and this likely reduced the impacts of heart rate on (stroke) work.

In this experiment: 1) shortening power produced by strips from hypoxic-acclimated trout increased approximately twofold, whereas net power changed little or not at all as frequency was increased from 30 to 90 contractions/min (Fig. 2F); and 2) shortening power increased in strips from normoxic-acclimated trout by almost fourfold when frequency was increased from 30 to 90 contractions/min, whereas net power only doubled across the same frequency range and did not change between 70 and 90 contractions/min (Fig. 2E). The latter results are consistent with those of Shiels et al. (49) and Harwood et al. (22) for rainbow trout, where net power of ventricular strips tested at 12–15°C peaked at ~60 contractions/min. The difference in the effect of frequency on shortening versus net power, and the apparent limit in net power at higher frequencies, is explained by the fact that lengthening work and power also increased with contraction frequency, and the effect became relatively greater at higher frequencies (Figs. 1 and 2C). This disproportionate increase in lengthening power at high contraction frequencies would eventually offset the increase in shortening power, resulting in net power not changing or attaining a maximum at high contraction rates. The increase in lengthening power with frequency is in agreement with observations for ventricular muscle from trout (22) and cod (53) and likely reflects a failure to achieve complete mechanical restitution (relaxation, lusitropy) between beats at high heart rates (21).

Impact of Acute and Chronic Hypoxia on Myocardial Work and Power Output

Reduced contractility under normoxic conditions.

When tested at 21 kPa Po₂ (normoxic conditions), net work and power were 30–50% and ~15% lower, respectively, in ventricular trabeculae from hypoxic-acclimated trout as compared with those from normoxic-acclimated conspecifics (Figs. 1 and 2E versus F). This effect was not due to increased stiffness of the myocardium, as lengthening work was similar between the two acclimation groups (Fig. 1C versus D), but to a decrease in shortening work (Fig. 1A versus B). The ~30% decrease in shortening work is similar to the 28% decrease in in situ maximum S_V in hearts from hypoxic- versus normoxic-acclimated Atlantic cod when tested under oxygenated conditions at 10°C (42). As such, this study sheds light on the possible mechanism(s) underlying the reduction in S_V and Q reported in hypoxic-acclimated rainbow trout (35) and Atlantic cod (40–42) hearts. There are at least two potential explanations for the poor shortening work and power output of the ventricular trabeculae from hypoxic-acclimated fish when working under normoxic conditions. First, it is possible that our hypoxic acclimation conditions were severe enough to induce myocardial remodeling and/or myocardial necrosis. Nonetheless, although 3 wk of severe hypoxia (5 kPa O₂ at 21°C) did result in myocardial necrosis in flounder (30), no changes in myocardial ultrastructure were evident in snapper exposed to a 6-wk period of moderate hypoxia (Po₂, 10.2–12.1 kPa at 21°C) (9). Furthermore, there was no myocardial damage or limitations in myocardial energetic or enzymatic function after the exposure of trout hearts to acute severe hypoxia (<30 min, perfusate Po₂ ≤ 1 kPa) (16, 37, 38), which suggests that this explanation is unlikely. Second, the ventricular trabeculae of hypoxic-acclimated trout may have been “stunned.” Mechanical dysfunction that persists after reoxygenation and/or reperfusion, termed “myocardial stunning,” can occur without the presence of irreversible damage to the myofibrils (4) and has been reported in several studies on acute hypoxic effects on cardiac function in trout (16, 17, 37, 38).

Myocardial performance during acute hypoxia.

When bath Po₂ was reduced from 21 to 13.5 kPa, the decrease in myocardial performance (i.e., shortening work and relative power) in strips from normoxic-acclimated animals was similar at all contraction frequencies (Figs. 1 and 3B). In contrast, there was a strong inverse relationship between contraction frequency and the decrease in work and power in the hypoxic-acclimated group between 21 and 13.5 kPa (Figs. 1 and 3B). This effect was not likely due to differences in the accumulation of anaerobic byproducts as 1) the strips only performed 30 contractions during each measurement and 13.5 kPa is well above typical venous Po₂ levels (~4–5 kPa; 14, 52, 56); 2) lactate production is generally higher in muscles when exercised at higher intensities for shorter durations (33, 57), which is at odds with the pattern of changes in power observed; and 3) mitochondria from hypoxic-acclimated fish are also less affected by decreasing oxygen concentration and are able to maintain ATP synthesis better at lower oxygen saturations (9). Although we do not have an explanation for this observation, it may be related to the observations of Syme et al. (53) and Gesser and Rodnick (18). These authors reported Po₂-related decreases in myocardial work and power even when oxygen levels were well above physiological and suggested that this result was due to a mechanism other than reduced energy supply. Several potential oxygen sensing mechanisms have been identified in cardiac tissue [including H₂S, NO, ROS, and ion (Ca²⁺) channels (13, 23, 47, 55)], and it is possible that these mechanisms were more sensitive to changes in Po₂ in hypoxic-acclimated trout.

In the normoxic group, there was a 40–50% reduction in net work and power across the four tested frequencies as the saline Po₂ was lowered from 21 to 1.5 kPa (Figs. 1E and 3E), and myocardial performance (net power) recovered to ~75–80% of prehypoxic levels upon return to 21 kPa (Fig. 4). Comparisons of the effects of reduced Po₂ between the current study and the literature are difficult due to the different methodologies and species used; however, this loss of myocardial/cardiac performance upon acute hypoxic exposure and the degree of recovery are similar to those reported in other studies (17, 28, 42, 46, 53). In contrast, the lack of an effect of hypoxic acclimation on the loss of myocardial force and power when the strips were exposed to acute hypoxia (Figs. 1–3), and the poorer recovery (by ~20%) despite the lower absolute values of work and power produced by strips from hypoxic-acclimated fish (Fig. 4), were not anticipated or predicted. All previous studies that have used hypoxic-acclimated fish or populations that were inherently hypoxia-tolerant report improved myocardial/heart performance during hypoxia and/or recovery once normoxia was re-established. For example, 1) Driedzic et al. (11) showed that hearts from hypoxic-acclimated (4–6 wk, $P{w}_{O_2}$ 4–4.7 kPa) eelpout (Zoarces viviparous) were better able to sustain peak tension during anoxia when bath Ca²⁺ levels were elevated; 2) Petersen and Gamperl (42) showed that although the Q_max of hearts from Atlantic cod that were acclimated to hypoxia (~8–9 kPa) for 6–12 wk at 10°C was lower than normoxic-acclimated individuals, it fell less when exposed to severe hypoxia and recovered better (by 10%) when returned to normoxic conditions; 3) Joyce et al. (27) showed that myocardial muscle from hypoxia-tolerant sea bass (Dicentrarchus labrax) developed 3–5 times more isometric force as compared with strips from hypoxia-sensitive fish when exposed to acute hypoxia; and 4) Faust et al. (16) showed that in situ hearts from a hypoxia-tolerant strain of rainbow required twice the duration of severe hypoxia (~1 kPa), and an elevated workload during hypoxia, before they experienced a “typical” degree of hypoxia-induced loss of function. The reason(s) for the discrepancy between these studies and the present results are not known, but collectively they 1) point to the severity of hypoxia during acclimation as a key determinant of myocardial dysfunction; and 2) suggest that hypoxic-acclimation cannot realize the improvements in myocardial function under oxygen-limiting situations that are achieved through adaptation to hypoxia (i.e., where certain populations or individuals are inherently hypoxia tolerant).

Perspectives and Significance

This is the first study to compare contractility in isolated ventricular trabeculae from normoxic- and hypoxic-acclimated rainbow (steelhead) trout and strongly suggests that the diminished S_V reported in chronically hypoxic trout and cod (35, 40–42) results from an increase in end-systolic volume, i.e., a decrease in ejection fraction. This is proposed because although lengthening work (i.e., related to ventricular stiffness) was not affected by hypoxic acclimation, shortening work and power, and thus the ability to eject blood from the heart, were substantially reduced. The mechanism(s) mediating this decrease in shortening work and power are still unidentified but are unlikely to involve alterations in NO signaling. Hypoxic acclimation does not modify the concentration dependent response of steelhead trout hearts to sodium nitroprusside (10⁻⁹ to 10⁻⁴ M; a NO donor) or the maximum loss (5–30% depending on frequency) of contractility (Carnevale et al., unpublished observations). This study also showed that the response of the myocardium of hypoxic-acclimated trout to a decrease in Po₂ is highly frequency dependent. This is an interesting/novel finding, and we hypothesize that this phenomenon may involve some form of “oxygen sensing.” This would be in line with the findings of Syme et al. (53) and Gesser and Rodnick (18), who showed that in the Atlantic cod and rainbow trout, respectively, myocardial function was very Po₂ dependent at partial pressures well above those that would be encountered by the heart in vivo. This strongly suggests that the response to reduced Po₂ is not simply caused by a failure in the face of reduced oxygen availability.

In this study, we report that acclimation of trout to a level of hypoxia close to their limit of tolerance (i.e., as suggested by the ~30% mortality in the hypoxic-acclimated group) results in an increase in heart size (relative ventricular mass), no improvement in the capacity of the spongy myocardium to perform during hypoxic conditions (i.e., shortening and net work/power) and a reduced capacity to recover function following hypoxic exposure. These results are in clear contrast to previous studies on cod and trout exposed to a less severe hypoxic insult (35, 40–42) and suggest that 1) the impacts of chronic hypoxia on myocardial performance are highly dependent on the severity of the hypoxic challenge; and 2) that cardiac remodeling (i.e., an increase RVM) appears to be required under severe, prolonged, hypoxic conditions to compensate for a loss of myocardial contractility. Furthermore, when all our previous studies on hypoxic acclimation and fish cardiac function (present study; 35, 40–42) are compared with the results of experiments on normally “hypoxia-tolerant” species where individuals (27) or populations (16) show a degree of inherent hypoxia tolerance, the data suggest that alterations in genotype are a prerequisite for substantial improvements in myocardial hypoxia tolerance (i.e., acclimation to hypoxia does not produce this phenotype). Understanding these aspects of the relationship between water O₂ levels and myocardial function, and the mechanisms involved, is key to understanding how fish will respond to climate change.

GRANTS

This research was supported by NSERC Discovery grants to A. K. Gamperl (249926-2011) and D. A. Syme (201190-2012), and an NSERC Accelerator grant to A. K. Gamperl (412325-2011). During these studies, C. Carnevale was supported by a Memorial University School of Graduate Studies fellowship.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

A.K.G, C.C., and D.A.S. conceived and designed research; C.C. and J.C.R. performed experiments; C.C. analyzed data; C.C., A.K.G., and D.A.S. interpreted results of experimental; C.C., D.A.S., and A.K.G. prepared figures.