Dynamic changes in scope for heart rate and cardiac autonomic control during warm acclimation in rainbow trout

Acute warming in fish typically results in an exponential rise in oxygen consumption rate and cardiac output, driven by increased resting heart rate (f_H,rest), with Q₁₀ values of ∼2–3 (Clark et al., 2008; Gollock et al., 2006). However, with more chronic seasonal temperature changes, compensatory physiological adjustments (i.e. thermal acclimation) can be initiated to counteract the thermal effects, hence reducing Q₁₀ (see Seebacher et al., 2015). Moreover, thermal acclimation capacity may be crucial for the resilience of ectotherms to global warming (see Chevin et al., 2010; Seebacher et al., 2015). While most thermal acclimation studies characterize physiological traits after extended exposure to fixed temperature regimes, the time course and rate of the acclimation response have received much less attention although this information can be essential for revealing the underlying mechanisms and time constraints associated with thermal acclimation (Sandblom et al., 2014; Somero, 2015).

If the increase in f_H,rest during acute warming is greater than the increase in maximum heart rate (f_H,max), the scope for heart rate (f_H,scope) decreases, which may constrain aerobic performance capacity (Farrell et al., 2009). Conversely, if thermal acclimation lowers f_H,rest, the scope may be restored (Franklin et al., 2007). Hence, the thermal plasticity of f_H,rest and f_H,scope is a key component in the overall compensatory acclimation response to elevated temperatures in ectotherms. Under steady-state conditions, f_H,rest is determined by the tonic activity of extrinsic stimulatory adrenergic nerves and circulating catecholamines, inhibitory cholinergic (‘vagal’) nerves, as well as the intrinsic heart rate (f_H,intr) (Nilsson, 1983; Sandblom and Axelsson, 2011). Thermal acclimation of f_H,rest in the intact animal probably involves changes to both f_H,intr (Aho and Vornanen, 2001; Haverinen and Vornanen, 2007) and extrinsic control mechanisms (Graham and Farrell, 1989; Priede, 1974; Sureau et al., 1989; Wood et al., 1979). Yet, there is little information about the time course and dynamic interaction between intrinsic and extrinsic factors determining f_H,rest and f_H,scope during thermal acclimation.

We determined resting, intrinsic and maximum heart rates, as well as adrenergic and cholinergic tone in separate groups of rainbow trout, Oncorhynchus mykiss, acclimated to 9°C and when acutely transferred to 16°C. The subsequent dynamic warm acclimation response was then determined after 2, 4, 7, 11, 18 and 39 days at 16°C. This temperature span is well within the natural thermal tolerance range and has previously been used in thermal acclimation studies on rainbow trout (Gräns et al., 2009; Priede, 1974). We hypothesized that an initial increase in f_H,rest with acute warming would be counteracted by increased cholinergic tone, reducing f_H,rest (Ekström et al., 2014). However, as f_H,intr was expected to gradually decline during the acclimation period, the importance of cholinergic inhibition of f_H,rest was predicted to decrease as warm acclimation progressed. We further hypothesized that f_H,scope would initially be lowered with the acute warming as a result of a greater increase in f_H,rest than f_H,max, but would then recover because of gradually reduced f_H,rest.

Rainbow trout (O. mykiss, N=115, mass 37.4±0.6 g) of mixed sexes were obtained from Antens Laxodling AB (Alingsås, Sweden) and kept in a 500 l holding tank supplied with aerated freshwater from a recirculating system. They were acclimated to 9°C for at least 8 weeks prior to experimentation under a daily 12 h:12 h light:dark cycle and fed commercial feed once weekly. Experimental procedures were covered by ethical permit 65-2012 from the ethical committee of Gothenburg.

Fish were anaesthetized in freshwater containing MS-222 (tricaine methanesulphonate; 150 mg l⁻¹) buffered with NaHCO₃ (300 mg l⁻¹). The fish was placed ventral side up on wet foam on an operating table where anaesthesia was maintained by irrigating the gills with water containing MS-222 (75 mg l⁻¹) and NaHCO₃ (150 mg l⁻¹). For heart rate measurements, two ECG electrodes (AS 631-2, Cooner Wire, Chatsworth, CA, USA) were inserted between the pectoral fins using a 23 gauge needle (Sterican, B. Braun Medical AB, Danderyd, Sweden). One electrode was directed caudally and one was directed cranially with the tip close to the heart. A polyethylene catheter (PE 50) for injection of pharmacological substances was then introduced intraperitoneally. The electrode wire and catheter were secured to the skin using 4.0 silk sutures. Following surgery, fish were placed in the experimental setup (see below) and allowed at least 24 h of post-surgery recovery.

The experiments were conducted between February and June. The fish were placed in opaque holding tubes (length 250 mm, diameter 63 mm) in a 250 l tank, continuously supplied with flow-through freshwater from the recirculating system. A 9 kW heater (K060, Värmebaronen, Kristianstad, Sweden) was used for temperature control. Eight separate groups of fish were examined at the initial acclimation temperature of 9°C (day 0), immediately after warming to 16°C (day 1) or subsequently after 2, 4, 7, 11, 18 and 39 days of acclimation to 16°C.

On each measurement day, f­_H,rest was first recorded for several hours in all fish to confirm a stable baseline. Individual fish were then randomly allocated to one of two experimental protocols, either to determine f_H,max using a chasing protocol (protocol 1) or to determine cardiac autonomic tone and f_H,intr (protocol 2), as described below.

Protocol 1: the chasing experiments to determine f_H,max were performed in a round tank (diameter 0.6 m, height 0.3 m, ∼85 l) supplied with flow-through aerated water from the recirculating system, maintaining the same temperature as the experimental tank. Fish were individually chased for 5 min until fatigue and quickly returned to the holding tube to record f_H,max.

Protocol 2: the pharmacological treatment group largely followed the protocol of Altimiras et al. (1997) and modified by Sandblom et al. (2010). Briefly, fish were intraperitoneally injected with atropine sulphate (1.2 mg kg⁻¹) to block muscarinic receptors and f_H was recorded for approximately 30 min, allowing the atropine to take full effect. Next, β-adrenergic receptors were blocked by injecting propranolol (3 mg kg⁻¹) and f_H was recorded for 1 h to allow the drug to have full effect. Pharmacological substances were dissolved in saline (0.9% NaCl) and injected as a 1 ml kg⁻¹ bolus followed by 0.5 ml of saline to flush the catheter. All chemicals and pharmacological substances were purchased from Sigma-Aldrich (St Louis, MO, USA). Following the experiments of either protocol, fish were killed in water containing a high dose of MS-222 (500 mg l⁻¹).

In the pharmacological experiments, f_H,intr was obtained after complete autonomic blockade and cholinergic and adrenergic tone were calculated according to Altimiras et al. (1997).

Q₁₀ values for the group means between 9°C and subsequent days at 16°C were calculated for the rate-dependent variables (f_H,rest and f_H,intr) using the Van't Hoff equation (Seebacher et al., 2015).

Values are means±s.e.m. unless otherwise stated. Experimental sample sizes were based on previous studies on these variables in rainbow trout. To evaluate the effects of temperature change, a comparison between the values at 9°C (day 0) and values at subsequent days at 16°C (days 1–39), was conducted using a general linear model with Dunnett's 2-tailed post hoc tests. To further investigate the acclimation response at 16°C, a general linear model including the seven groups measured at 16°C was used. Normality and homogeneity of variances were verified using Shapiro–Wilk and Levene tests, respectively. Maximum heart rate was power transformed to meet these assumptions. Statistical significance was accepted at P≤0.05.

The f_H,rest at 9°C was 39±2 beats min⁻¹, which corresponds well with previous studies on rainbow trout (Ekström et al., 2014; Gamperl et al., 1995; Priede, 1974). Acute warming to 16°C elicited an expected increase in f_H,rest, which was highest in the group acclimated to 16°C for 2 days (60±3 beats min⁻¹, Q₁₀=1.9; Fig. 1A,B). However, the f_H,rest in the groups acclimated to 16°C was significantly affected by acclimation time and decreased from the initial acute response to 45±3 beats min⁻¹ in the group acclimated for 39 days (Q₁₀=1.2; Fig. 1A,B). This is close to full thermal compensation (i.e. Q₁₀=1) and the degree of thermal compensation of f_H,rest in this study is indeed high when compared with other acclimation studies on trout (Q₁₀=1.5–2.6) (Gräns et al., 2009; Priede, 1974; Wood et al., 1979). These discrepancies are probably due to the shorter acclimation times used in these studies (2–4 weeks), resulting in a lower degree of compensation. Nonetheless, the present study shows that following an acute increase from 9 to 16°C, f_H,rest remains significantly elevated for the first week and it takes somewhere in the range of 18–39 days for the compensatory acclimation response to level out (Fig. 1A,B). However, we cannot exclude the possibility that further thermal compensation (e.g. complete) would have occurred had we extended the acclimation time further.

One of the few other investigations of dynamic changes in cardiovascular variables during thermal acclimation in fish is a recent study on f_H,rest in the estuarine longjaw mudsucker, Gillichthys mirabilis (Jayasundara and Somero, 2013). In contrast to the present study, Jayasundara and Somero (2013) found little thermal compensation of f_H,rest and suggested that this may reflect a strategy minimizing the energetic costs associated with acclimation in a highly thermally variable environment. In fact, the pronounced heart rate acclimation capacity observed in rainbow trout may reflect a physiological phenotype adapted to accommodate predictable seasonal temperature changes typical of temperate regions (see Klaiman et al., 2011).

The cholinergic tone at 9°C (36±16%) did not differ significantly from the group acutely warmed to 16°C (46±15%), but increased significantly with further warm acclimation to reach a maximum of 69±18% at day 18 (Fig. 1C). The adrenergic tone was 28±10% at 9°C and there was no significant change with acclimation to 16°C (Fig. 1C). The f_H,intr was 41±3 beats min⁻¹ at 9°C and peaked at 68±3 beats min⁻¹ on day 2 at 16°C (Q₁₀=2.1; Fig. 1A,B). While this acute effect resembles the acute effect observed for f_H,rest, no significant acclimation effect was evident for f_H,intr in the groups acclimated to 16°C and the Q₁₀ was still 1.8 for this variable after 39 days at 16°C (Fig. 1A,B). Consequently, the decrease in f_H,rest during warm acclimation was mainly due to an increase in cholinergic tone, which was significant from day 18 (Fig. 1A,C). This is consistent with a comparison of in vivo data with studies on in situ perfused hearts from thermally acclimated trout. In the perfused heart preparation, all extrinsic cardiac control is absent and the thermal compensation of heart rate is often relatively low (Q₁₀=2.0–2.2; Graham and Farrell, 1985; Graham and Farrell, 1989), whereas the thermal compensation of f_H,rest in vivo can be considerably greater (Q₁₀=1.2–1.5, present study and Priede, 1974). Again, this highlights the importance of changes in autonomic tone for the thermal resetting of f_H,rest in the intact animal.

f_H,max increased from 72±2 beats min⁻¹ in the 9°C acclimated group to a peak value of 111±2 beats min⁻¹ in the group acutely warmed to 16°C (day 1), and then remained significantly elevated relative to the 9°C group throughout the warm acclimation period (Fig. 1A). These patterns are qualitatively consistent with previous studies on heart rate acclimation in various Arctic sculpin species, albeit at lower temperatures (Franklin et al., 2007,, 2013). Even so, a significant general acclimation effect on f_H,max was found in trout, suggesting a somewhat decreased f_H,max with warm acclimation (Fig. 1A).

As a result of the concurrent changes in f_H,rest and f_H,max, f_H,scope initially increased significantly from 33±4 beats min⁻¹ at 9°C to 58±3 beats min⁻¹ in the group acutely warmed to 16°C, but then drastically declined to 40±7 beats min⁻¹ after 2 days acclimation to 16°C, primarily due to the elevated f_H,rest in this group (Fig. 1A,D). However, f_H,scope subsequently increased and was again significantly elevated relative to the 9°C group after 39 days of acclimation to 16°C (Fig. 1D). Thus, while the increased f_H,max with acute warming to 16°C explained the initial increase in f_H,scope, our data reveal that increased cholinergic inhibition of f_H,rest was crucial for the subsequent increase in f_H,scope during warm acclimation because f_H,max did not increase further with acclimation. Interestingly, the current findings on the thermal acclimation of heart rate in trout are qualitatively consistent with a similar study on metabolic acclimation in the shorthorn sculpin, Myoxocephalus scorpius, where the recovery and increase in aerobic scope with warm acclimation were mainly the result of reduced standard metabolic rate, whereas maximum metabolic rate was little affected (Sandblom et al., 2014). Thus, the elevated cholinergic tone on the heart with warm acclimation may reflect an adaptive mechanism allowing warm-acclimated fish to quickly modulate heart rate and cardiac output through vagal release during transient periods of additionally elevated oxygen demand (e.g. exercise and digestion), and may therefore be fundamental to maintain a high metabolic scope following warm acclimation, as observed previously in various temperate and Arctic fish species (Gräns et al., 2014; Sandblom et al., 2014; Seth et al., 2013).

The mechanistic underpinnings of the increase in cholinergic tone with warm acclimation are presently not clear, but may reflect a greater expression of vagal activity as the metabolic oxygen demand is gradually reduced during warm acclimation. Another possible mechanism is an up-regulation of cardiac muscarinic receptor density, which would increase the cardiac sensitivity to cholinergic stimulation. While an up-regulation of cardiac β-adrenergic receptors has been demonstrated with cold acclimation in rainbow trout (Keen et al., 1993), the chronic temperature effects on cardiac muscarinic receptor density have to our knowledge not been examined.

The current study on rainbow trout provided several novel insights into the dynamic changes in cardiovascular function and its neuro-hormonal control during warm acclimation. The changes in f_H,scope with warming exhibited a complex pattern with an initial increase due to a greater increase in f_H,max than f_H,rest, followed by a decline due to a further increase in f_H,rest and then a subsequent increase due to elevated cholinergic tone suppressing f_H,rest during warm acclimation. Ultimately, increased knowledge about the dynamics and the temporal and physiological constraints to thermal acclimation will advance our understanding of how species biogeography and ecological interactions change seasonally, as well as how organisms may be affected by global change altering thermal conditions.

The authors thank Michael Axelsson, Catharina Olsson, Fredrik Jutfelt and Jeroen Brijs for valuable input during this study.