Seawater acclimation affects cardiac output and adrenergic control of blood pressure in rainbow trout (Oncorhynchus mykiss)—implications for salinity variations now and in the future

Climate change may exacerbate salinity variations in aquatic ecosystems, potentially affecting fish performance. We show that long-term seawater exposure in rainbow trout increases cardiac output and reduces blood pressure and vascular resistance through altered α-adrenergic control. Possible impacts of salinity fluctuations on fish aerobic swimming and digestive capacity are discussed.


Introduction
Global climate change is predicted to affect the salinity and its variability in aquatic environments (Kultz, 2015;Seebacher and Franklin, 2012). For example, exacerbated transient reductions in salinity may be predicted for shallow coastal areas due to increases in precipitation and freshwater (FW) run off (Meier et al., 2012), whereas increases in salinity has been postulated for semi-arid regions due to lower precipitation and increased evaporation (Jeppesen et al., 2015). Such changes in salinity can be challenging and constrain the performance of fish. Euryhaline fishes possess physiological traits that allow them to inhabit a wide range of salinities. Osmoregulatory capacity may, therefore, be of added importance for the ability of fish to cope with future environmental changes (Kultz, 2015), and understanding of the physiological constraints and mechanisms underlying euryhalinity is important to predict future resilience of fish populations and inform management efforts (Cooke et al., 2013;Kultz, 2015;Seebacher and Franklin, 2012).
FW generally has a volume loading effect on fish (i.e. gain of water and loss of ions), whereas seawater (SW) has a volume depleting effect (i.e. loss of water and gain of ions; Olson, 1992;Smith, 1932). These passive effects of water salinity are counteracted by a range of active physiological and behavioural modifications (Marshall and Grosell, 2006). In FW, fish take up ions from the surrounding water via specialized cells in the gills and excrete dilute urine (Evans et al., 2005;Perry et al., 2003;Wood and Bucking, 2010). In contrast, in SW, they actively drink and create an inward directed flow of water through solute-linked water absorption mechanisms in the intestine (Bath and Eddy, 1979;Sundell and Sundh, 2012). Excess ions are then actively excreted across the gills and kidneys (Evans et al., 2005;Grosell et al., 2010), and water is conserved by maintaining low urine volumes (Linhart et al., 1999;Smith, 1930). Collectively, these changes allow euryhaline teleosts to maintain osmotic homeostasis with a constant plasma osmolality of ∼300 mOsm across environmental salinities (Evans, 2008;McCormick and Saunders, 1987;McCormick et al., 1998).
Various cardiovascular adjustments are also important for euryhaline fishes when responding and acclimating to water salinity (Brijs et al., 2015(Brijs et al., , 2016(Brijs et al., , 2017. Gastrointestinal blood flow increases at least 2-fold in chronically SW-acclimated rainbow trout due to a combination of increased cardiac output (mediated by increased stroke volume) and an increased proportion of blood flow directed to the gastrointestinal tract (Brijs et al., 2016). This elevation in gastrointestinal blood flow is believed to be essential for the convection of absorbed ions and water, as well as for supplying oxygen and nutrients to metabolically active gastrointestinal tissues (Brijs et al., 2015(Brijs et al., , 2016. In theory, the elevated gastrointestinal blood flow of trout in SW could be caused either by an elevated arterial blood pressure and/or a reduced gastrointestinal vascular resistance (Olson, 2011). However, the dorsal aortic blood pressure (P DA ) decreased by 11-21% in FW-acclimated trout acutely exposed to SW for 24 hours (Maxime et al., 1991), or short-term acclimated to SW for 2 weeks (Olson and Hoagland, 2008). Nonetheless, knowledge gaps remain as no study has determined the effects of salinity on systemic vascular resistance (R SYS , the sum of the gastrointestinal and the somatic vascular resistances), and it is largely unknown how the hemodynamic status of fish is affected after more chronic SW exposure. Moreover, the only previous study that recorded blood pressure responses to short-term acclimation to SW used rainbow trout with surgically opened pericardia (Olson and Hoagland, 2008), which is known to negatively affect cardiac performance and blood pressure dynamics (Farrell et al., 1988;. Baseline systemic vascular resistance in teleost fishes is to a great extent determined by the α-adrenergic tone on the resistance vasculature, which is primarily mediated by adrenergic neuronal activity (Sandblom and Axelsson, 2011;Sandblom and Gräns, 2017;Smith, 1978;Smith et al., 1985). Thus, it could be hypothesized that the elevated gastrointestinal blood flow in SW-acclimated trout is mediated by a reduced αadrenergic vasomotor tone on the gastrointestinal resistance vasculature. Indeed, the gastrointestinal vasculature is under α-adrenergic control because gastrointestinal blood flow is markedly reduced following injection of α-adrenergic agonists (Axelsson and Fritsche, 1991;Axelsson et al., 1989Axelsson et al., , 2000Sandblom et al., 2012;Seth, 2010), and changes in gastrointestinal blood flow and gastrointestinal vascular resistance with feeding, exercise and hypoxia are at least partially due to changes in α-adrenergic vasomotor tone (Axelsson and Fritsche, 1991;Seth and Axelsson, 2010;Seth et al., 2008). Interestingly, hypertensive trout fed a high salt diet had a decreased dorsal aortic α-adrenoreceptor mRNA expression along with a blunted P DA response to exogenous catecholamines (Chen et al., 2007). While this shows that vascular αadrenoreceptor density and vascular adrenergic sensitivity can be dynamically regulated in trout, it is unknown how the adrenergic control of blood pressure is affected by acclimation to different water salinities. Thus, there is a need for simultaneous measurements of arterial pressure and flow to resolve how R SYS and P DA changes with salinity in euryhaline fishes. This information is of importance to understand how cardiovascular and aerobic performance traits of fishes are affected by transient and chronic salinity changes, which may have implications for dispersal and fitness of estuarine and migratory fish species now and in a future with more pronounced salinity variations.
Here, we used rainbow trout (Oncorhynchus mykiss, Walbaum 1792) as a euryhaline model species that can tolerate a wide range of salinities while being suitable for in vivo cardiovascular recordings. Indeed, some strains of wild rainbow trout (i.e. 'steelhead') are anadromous and naturally migrate between FW and SW, whereas most farmed rainbow trout strains can acclimate to SW (Brijs et al., 2017;Quinn and Myers, 2004 along with P DA to calculate R SYS in vivo in chronically FWand SW-acclimated rainbow trout; with the hypothesis that SW-acclimated trout would exhibit reduced R SYS and P DA . Further, we examined whether differences in R SYS and P DA across salinities could be explained by altered resistance vessel sensitivity to α-adrenergic stimulation or through changes in intrinsic α-adrenergic tone by using specific α-adrenergic pharmacological tools.

Experimental animals
Rainbow trout (O. mykiss) were obtained from a local hatchery (Vänneåns fiskodling, Sweden; see Table 1 for mass and length) and kept in a 1000-l tank with aerated recirculating FW (salinity 0-1 ppt) at 10.5 ± 1.0°C for at least 2 weeks. A subset of 30 fish was subsequently randomly assigned for transfer to another identical 1000-l tank with recirculating aerated SW (salinity 30-33 ppt) at the same temperature (10.5 ± 1.0°C). The fish were then acclimated to their respective salinity treatment for a minimum of 6 weeks prior to experimentation. During the holding and acclimation periods, they were fed three times per week with dry commercial trout pellets (9 mm Protec Trout pellets, Skretting, Stavanger, Norway), but fasted for 1 week prior to surgery and experimentation. Animal handling and surgical procedures were performed in accordance with ethical permit #165-2015, approved by the ethical committee in Gothenburg.

Surgery and instrumentation
Individual rainbow trout were anesthetized in FW containing Tricaine methanesulphonate (MS-222, 150 mg l −1 ) buffered with NaHCO 3 (300 mg l −1 ). Length and weight were determined before placing the fish dorsally on water-soaked foam on a surgical table. The gills of the fish were continuously irrigated with recirculating aerated FW (i.e. for both FWand SW-acclimated fish) at 10°C containing MS-222 (75 mg l −1 ) buffered with NaHCO 3 (150 mg l −1 ) throughout the surgery. FW was used as anaesthetic solvent for both acclimation groups since earlier attempts of using SW as solvent for the SWacclimated group resulted in impaired post-surgical recovery, possibly due to impaired drinking during the anesthetized state. The dorsal aorta was cannulated with a custom-made PE-50 catheter using a steel wire guide Smith and Bell, 1964). The cannula was inserted dorsally at ã 45°angle, between the second and third pair of gill arches inside the mouth cavity (Axelsson and Fritsche, 1994). The catheter was filled with heparinized (100 IU ml −1 ) 0.9% saline and exteriorized through the snout and locked in place by a bubble on the catheter (Soivio et al., 1975). The fish was then placed on its side and the operculum and the gill arches were lifted to expose the opercular cavity . The ventral aorta was gently dissected free without damaging nearby nerves and vessels. A Transonic 2.5PSL flow probe (factory calibrated to 10°C; Transonic systems, Inc, Ithaca, NY, USA) was placed around the aorta with the help of a silk suture (size 4-0). Finally, the probe lead and the catheter were attached to the skin with several silk sutures. After surgery, the fish were immediately placed in the experimental setup, which consisted of individual opaque holding tubes with a volume of~3 l, floating in a 120-l tank receiving a continuous supply of aerated FW or SW (11 ± 1°C) depending on the acclimation salinity. All fish were allowed a recovery time of at least 40 h before experiments were initiated.

Experimental protocol
Baseline recordings of cardiac output, heart rate and P DA were first performed for a minimum of 2 h at the start of each experiment. When stable baseline conditions had been confirmed, four dosages of the α-adrenergic agonist phenylephrine (10, 30, 60 and 100 μg kg −1 ) and saline (0.9%) as a control, were injected into the dorsal aortic catheter, followed by 0.3-0.4 ml saline (0.9%) to clear the catheter dead space. The administration of phenylephrine dosages was randomized, and an additional 0.3-0.4 ml saline (0.9%) was injected when the peak P DA response had leveled off to ensure that all traces from the previous injection was cleared from the catheter. Last, one dosage of the α-adrenergic antagonist prazosin (1 mg kg −1 ) was injected in the same way to obtain a complete α-adrenergic blockade. All injections were administered in volumes of 1 ml kg −1 body mass (M b ). While the administration order of phenylephrine and saline injections was randomized for each fish, prazosin was always administered last. Before a new injection was administered, care was taken to allow all cardiovascular variables to return to stable baseline levels. The time for this varied among individuals and injections but was typically never longer than one hour. After the experiments, the fish were killed with a sharp blow to the head.

Data acquisition and analysis
The dorsal aortic catheter was connected to a pressure transducer (pvb Medizintechnik, Kirchseeon, Germany) that was calibrated against a static water column with the water level in the experimental tank serving as the zero reference. The signal from the pressure transducer was amplified using a 4ChAmp pre-amplifier (Somedic, Hörby, Sweden). The Transonic flow probe was connected to a Transonic flow meter (Transonic systems, Inc, Ithaca, NY, USA). The signals from the flow meter and pressure transducer were relayed to a 16SP PowerLab system (ADInstruments, Castle Hill, Australia) that was connected to a computer with LabChart pro data acquisition software (7.3.2, ADIinstruments). Heart rate was determined from the pulsatile blood flow or pressure recordings using the blood pressure module in LabChart pro. R SYS was calculated from P DA and cardiac output as: Stroke volume was calculated as: = stroke volume cardiac output heart rate Mean values for baseline P DA , cardiac output and heart rate for each individual were obtained from representative calm periods, which were taken at the end of the initial 2 h baseline recording period. To analyze the responses to αadrenergic stimulation with the different dosages of phenylephrine, mean values for all cardiovascular variables were taken at the peak blood pressure response after each drug injection. The order of administration for the different dosages of phenylephrine was randomized for each fish. Cardiovascular variables after complete α-adrenergic blockade with prazosin were obtained approximately two hours after the drug had been administered and the blood pressure had reached a new steady state. All mean values were typically calculated as 30 s means.

Statistical analysis
Statistical analyses were conducted using SPSS Statistics 24 (IBM Corp., Armonk, NY, USA). Independent t-tests were used for all comparisons between acclimation groups containing one dependent factor, including all baseline and prazosin treatment analyses, as well as all analyses of the absolute changes induced by prazosin. To assess the general effect of the different dosages of phenylephrine within each acclimation group, as well as the general effect between acclimation groups, a repeated measures ANOVA was used with individuals as subject variables and the dose of phenylephrine as the repeated variable. In the model, we included dose of phenylephrine (0, 10, 30, 60, 100 μg kg −1 ), acclimation group and their interactions as fixed effects. To meet the assumptions of statistical tests, a logarithmic transformation was applied for cardiac output, R SYS and P PULSE , and a square root transformation was applied for P DA . When the assumption of sphericity in the general linear model analyses was not met, we used Greenhouse-Geiser corrections to interpret if the results were significant. Values are presented as means ± SEM and statistical significance was accepted at P ≤ 0.05.

Results
There were no obvious behavioural differences between acclimation groups as fish from both groups generally remained calm in the experimental setup throughout the recording period. There were no differences in body mass, length or condition factor between acclimation groups (P < 0.05; Table 1).

Cardiovascular effects of α-adrenergic drugs in FW-and SW-acclimated trout
Intra-arterial injection of the α-adrenergic agonist phenylephrine caused dose-dependent increases in R SYS and P DA in both acclimation groups, but both variables were consistently lower in SW-acclimated trout (Fig. 3A, B). However, when analyzing the absolute changes in R SYS and P DA from baseline values with each dosage of phenylephrine (data not shown), there were no significant differences between acclimation groups indicating that the responsiveness to α-adrenergic stimulation was unchanged across salinity acclimation groups (F 1 = 2.184, P = 0.159 and F 1 = 0.219, P = 0.645, respectively).
Phenylephrine injections significantly affected cardiac output (F 4 = 2.974, P = 0.026) and stroke volume (F 3 = 16.144, P < 0.001) in both acclimation groups (Table 2). Further, administration of phenylephrine typically reduced heart rate in both SW and FW, indicating a barostatic reflex (F 2.544 = 13.059, P < 0.001; Table 2). Although not statistically tested, cardiac output appeared to increase at the lower dosages of phenylephrine due to an increased stroke volume, whereas it decreased at higher dosages of phenylephrine as stroke volume reached an upper limit while heart rate continued to decrease (Table 2) After complete α-adrenoceptor blockade with prazosin, R SYS remained consistently lower in SW (T 14 = 2.442 and P = 0.028; Fig. 1E), but the absolute reduction in R SYS with prazosin was significantly greater in FW-compared to SW-acclimated rainbow trout (T 8.312 = 2.895, P = 0.019; Fig. 4). P DA was also consistently lower in SW-acclimated trout after prazosin treatment (T 20 = 2.689, P = 0.014; Fig. 1D), although the absolute change in P DA from baseline with prazosin was not statistically different between acclimation groups. All other blood pressure variables also showed a similar magnitude in the absolute change after prazosin across acclimation groups, but again were consistently lower in the SW-acclimated trout (Fig. 2).
While prazosin induced a significantly greater heart rate increase in FW-acclimated trout compared to SW-acclimated trout (data not shown), there was still no significant difference in heart rate between acclimation groups after prazosin (T 20 = 1.579, P = 0.130; Fig. 1B). The magnitude of the absolute changes in stroke volume and cardiac output with prazosin were not significantly different between acclimation groups. However, the clear and significant differences in baseline values for these variables that were observed in untreated trout disappeared with prazosin treatment (cardiac output: T 14 = 0.840, P = 0.415 and stroke volume: T 14 = 1.317, P = 0.209; Fig. 1A, C).

Hemodynamic status in FW and SW and possible implications for aerobic performance traits
The present findings demonstrate that P DA and R SYS are significantly reduced in chronically SW-acclimated rainbow trout. Moreover, our results confirm previous observations of significantly elevated cardiac output in SW-acclimated trout (Brijs et al., 2016(Brijs et al., , 2017. These fundamental cardiovascular changes in response to salinity open up a range of important questions of how aerobic performance traits such as digestion and swimming capacity are affected by salinity in euryhaline fishes. For example, it is unknown if the elevated cardiac output in SW affects the scope for cardiac output, or whether the cardiovascular system possesses sufficient phenotypic plasticity to compensate across salinities; e.g. by increasing the maximal cardiac output in SW to maintain cardiac scope. While acclimation to different salinities generally has negligible impacts on the maximum swimming capacity of euryhaline fishes (Beamish, 1978;Christensen et al., 2018;Nelson et al., 1996;Wagner et al., 2006), we are not aware of any study comparing maximum cardiac performance during sustained swimming at different acclimation salinities. Thus, while our findings require consideration in conservation management of fish populations that are exposed to varying environmental salinities in their natural habitats, they also highlight the need for further experiments on cardiorespiratory responses to exercise in euryhaline fishes across salinities.
Our data strongly suggest that the elevated gastrointestinal blood flow previously observed in SW-acclimated trout (Brijs et al., 2015(Brijs et al., , 2016 is caused by a reduced gastrointestinal vascular resistance, since the driving pressure for gastrointestinal blood flow (i.e. P DA ) was markedly reduced in SW and therefore cannot explain the elevated blood flow. However, it is not possible to conclude if a dilation of somatic vascular beds also contributed to the overall reduction in R SYS , or if somatic vascular resistance increased to aid blood flow distribution to the gastrointestinal tract in SW. Forced feeding of FW and SWacclimated rainbow trout increased gastrointestinal blood flow with the same absolute amount, which shows that the baseline difference in gastrointestinal blood flow persists after feeding (Brijs et al., 2016). However, whether the elevated gastrointestinal blood flow and decreased α-adrenergic tone on the gastrointestinal resistance vasculature of unfed SW fish constrains the ability to redistribute blood flow away from the gastrointestinal tract to supply swimming muscles during exercise represents another interesting topic to explore in the future.

6
Seawater acclimation alters the αadrenergic control of cardiovascular function Reduced R SYS can result either from elevated vasodilatory and/or reduced vasoconstrictory stimulation of the resistance vasculature (Nilsson, 1994;Olson and Farrell, 2005). The present study examined the α-adrenergic vasomotor tone and found that SW-acclimation of trout leads to reduced α-adrenergic constriction of the systemic resistance vasculature. This decreased α-adrenergic vasoconstriction can either be due to a reduced vascular α-adrenergic sensitivity (e.g. via downregulation of vascular α-adrenoreceptors), or a reduced intrinsic α-adrenergic neurohumoral tone (Chen et al., 2007;   Cardiovascular variables in freshwater-and seawater-acclimated rainbow trout during baseline conditions and after intra-arterial injection with different dosages of phenylephrine (10, 30, 60, 100 μg kg −1 body mass). The variables are cardiac output (CO, n = 9), heart rate (HR, n = 11-13), stroke volume (SV, n = 9), dorsal aortic diastolic blood pressure (P DIA , n = 11-13), dorsal aortic systolic blood pressure (P SYS , n = 11-13) and dorsal aortic pulse pressure (P PULSE , n = 11-13). Data are presented as means ± SEM. Asterisk (*) denote significant effects of salinity acclimation on baseline variables. Statistical outcomes for treatment effects of phenylephrine (P) and salinity (S) across phenylephrine concentrations between acclimation groups is denoted in the 7 Sandblom and Axelsson, 2011). Both acclimation groups displayed similar dose-dependent increases in P DA and R SYS with phenylephrine injections, indicating that the vascular sensitivity to α-adrenoreceptor stimulation was similar across acclimation groups. A reduced vascular α-adrenergic sensitivity can therefore not explain the lower baseline R SYS and P DA in the SW-acclimated trout. However, the greater reduction in R SYS after α-adrenoreceptor blockade with prazosin in FW-acclimated trout revealed a reduced intrinsic α-adrenergic tone on the resistance vasculature in SW. Consequently, a lower intrinsic α-adrenergic tone at least partly explains the lower baseline P DA and R SYS in SW. This is likely due to modulation of neural α-adrenergic vascular tone exclusively, as a previous study did not find any significant differences in resting levels of circulating catecholamines between FW-and SW-acclimated trout (Tang and Boutilier, 1988).
Still, the reduced α-adrenergic vasomotor tone cannot alone explain the reduction in R SYS in SW, as P DA and R SYS remained significantly lower in the SW-acclimated trout after complete α-adrenergic blockade. This implies that other local or neurohumoral vasodilatory factors, possibly acting on the gastrointestinal resistance vasculature, are also involved. Likely vasodilating candidates include natriuretic peptides such as atrial natriuretic peptide, which has a strong vasodilating effect on the celiacomesenteric artery in rainbow trout (Cousins and Farrell, 1996;Olson and Meisheri, 1989;Smith et al., 2000), and is released in response to elevated cardiac preload as occurs with SW-acclimation in trout (Brijs et al., 2017;Farrell and Olson, 2000). Indeed, natriuretic peptide secretion increased upon SW transfer in the eel Anguilla japonica (Kaiya and Takei, 1996a). This secretion was predominantly triggered by osmotic stimuli from salt loading and cellular dehydration, but also to a lesser extent modified by volume loading Takei, 1996b, 1997). Other possible vasodilator candidates include nitric oxide and nitric oxide derivatives that act as general vasodilators in the vasculature of teleost fish (Olson and Donald, 2009;Sandblom and Gräns, 2017). In mammals, intestinal hyperosmolarity, as would be expected with SW drinking in fish, have both direct vasodilatory effects on the gastrointestinal resistance vasculature, as well as indirect vasodilatory effects by stimulating nitric oxide production (Bohlen, 1998;Levine et al., 1978;Steenbergen and Bohlen, 1993;Zani and Bohlen, 2005). Interestingly, an increased activity of neuronal nitric oxide synthase was found in the anterior intestine of rainbow trout after 7 days of exposure to SW (25 ppt; Gerber et al., 2018). This indicates a role of nitric oxide in SW-acclimation and as a possible mediator of gastrointestinal vasodilation.
The 68% increase in cardiac output in SW-acclimated trout was primarily mediated via an increased stroke volume, as heart rate remained unchanged across salinities (Brijs et al., 2015(Brijs et al., , 2016(Brijs et al., , 2017. Nonetheless, the reduction in heart rate induced by phenylephrine in both acclimation groups reveals a functional cardiac baroreflex response at both salinities (Sandblom and Axelsson, 2005). While the increased stroke volume has previously been attributed to a reduced venous capacitance and an increased central venous pressure with SW-acclimation (Brijs et al., 2017), the present study also indicates that the elevated venous pressure may be due to the reduced R SYS . Interestingly, the significant elevation in baseline cardiac output and stroke volume in SW-acclimated trout was abolished following the prazosin treatment. This could be due to an increased central venous pressure following prazosin treatment in FW-acclimated trout due to altered trans-vascular fluid shifts, as observed in previous in vivo studies (see Sandblom and Gräns, 2017).

Conclusions and perspectives
The present study emphasizes that profound cardiovascular changes occur during acclimation to different salinities in a euryhaline teleost fish. These changes are likely important for maintaining osmotic homeostasis but may impact on aerobic performance traits, which requires future research attention and consideration in conservation management. While the present findings suggest that previous observations of increased gastrointestinal blood flow in SW are due to a reduced α-adrenergic tonus on the gastrointestinal resistance vasculature, our data also indicate that other vasoactive factors are important for mediating these responses. Thus, deciphering the apparently complex interplay between the various neural and hormonal cardiovascular control systems at play during FW to SW transition represents another challenging avenue for further research. From an evolutionary perspective, it could be speculated that the evolution of more complex control systems involving both neural and hormonal vascular control systems in teleost fishes, that are not present in elasmobranchs and cyclostomes (Nilsson, 1983,   1994;Sandblom and Axelsson, 2011;Sandblom and Gräns, 2017), has been an important prerequisite for the evolution of euryhalinity in this diverse group of vertebrates. This evolutionary transition has undoubtedly equipped many teleost species with the physiological machinery necessary to tolerate large acute and chronic salinity changes, as well as the ability to exploit and undertake long-distance migrations across environments with highly contrasting salinities. In fact, this capacity may provide euryhaline teleosts with competitive advantages allowing them to better cope with greater salinity variations in the future resulting from climate change and other anthropogenic perturbations. Author contributions E.Sa. conceived and designed the study; E.Su. and D.M. performed the experiments with technical assistance from A.E. and J.B.; E.Su., E.Sa. and A.G. analyzed the data and performed the statistical analysis; E.Su. and E.Sa. wrote the manuscript, with all co-authors providing input on the written text.