- Email: [email protected]
Intracellular carbonic anhydrase contributes to the red blood cell adrenergic response in rainbow trout Oncorhynchus mykiss
Respiratory Physiology & Neurobiology 184 (2012) 60–64
Contents lists available at SciVerse ScienceDirect
Respiratory Physiology & Neurobiology journal homepage: www.elsevier.com/locate/resphysiol
Short communication
Intracellular carbonic anhydrase contributes to the red blood cell adrenergic response in rainbow trout Oncorhynchus mykiss D.W. Carrie, K.M. Gilmour ∗ Department of Biology, University of Ottawa, ON, Canada
a r t i c l e
i n f o
Article history: Accepted 26 July 2012 Keywords: Rainbow trout Carbonic anhydrase Red blood cell NHE pH
a b s t r a c t In many teleost fish, catecholamines activate a red blood cell (RBC) Na+ /H+ exchanger (NHE), raising RBC intracellular pH to protect haemoglobin-O2 loading. The present study tested the hypothesis that RBC intracellular carbonic anhydrase (CA) contributes to this adrenergic response. The pH of rainbow trout (Oncorhynchus mykiss) blood was monitored continuously in vitro using blood flowing in a semiclosed loop or in vivo using an extracorporeal circulation. Addition or injection of isoproterenol activated the NHE, causing blood pH to fall (in vitro pH = −0.28 ± 0.03 pH units, N = 16; in vivo, −0.12 ± 0.02 pH units, N = 6). Both in vitro and in vivo, inhibition of RBC CA by acetazolamide significantly decreased the magnitude of the adrenergic response (in vitro, pH = −0.22 ± 0.02 pH units, N = 16; in vivo, −0.02 ± 0.01 pH units, N = 6) as well as the rate of recovery of blood pH following the adrenergic response. These results support the hypothesis that RBC intracellular CA plays an important role in the RBC adrenergic response of rainbow trout, and fuel speculation that interspecific differences in RBC CA activity are associated with the magnitude of the RBC adrenergic response. © 2012 Elsevier B.V. All rights reserved.
1. Introduction An enduring (and under-studied) question in the literature concerns the significance of interspecific differences in red blood cell (RBC) carbonic anhydrase (CA) activity (Gilmour, 2010). Carbonic anhydrase catalyzes the reversible reaction of CO2 and water to bicarbonate ions and protons, and therefore plays a fundamental role in a host of physiological processes including metabolism, gas exchange and acid–base balance. Differences in RBC CA activity spanning an up to 25-fold range have been documented across species within vertebrates in general (see Gilmour, 2010) and teleost fish specifically (Esbaugh et al., 2004). However, the physiological significance of these differences remains unclear. Within teleost fish, one possibility that warrants consideration is that interspecific differences in RBC CA activity are associated with the magnitude of the RBC adrenergic response. In many teleost fish, catecholamines (adrenaline and noradrenaline) that are released into the circulation bind to -adrenoreceptors on the RBC membrane, activating a Na+ /H+ exchanger (NHE) via a cAMP signal transduction pathway. This response alkalinizes the RBC intracellular environment, and
∗ Corresponding author at: Department of Biology, University of Ottawa, 30 Marie Curie, Ottawa, ON K1N 6T2, Canada. Tel.: +1 613 562 5800x6004; fax: +1 613 562 5486. E-mail address: [email protected] (K.M. Gilmour). 1569-9048/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.resp.2012.07.022
is thought to have evolved in parallel with Bohr-Root shift haemoglobins (Hb; i.e. Hb for which O2 binding is pH sensitive) to protect Hb-O2 loading at the gill during systemic acidosis (Berenbrink et al., 2005). Carbonic anhydrase is expected to contribute to the RBC adrenergic response by catalyzing the hydration of CO2 within the RBC to produce the protons that are extruded by the NHE. This reaction also yields HCO3 − that exit the RBC more slowly, in exchange for Cl− entry. In plasma, on the other hand, it is the absence of CA that is important, at least in the gill (Rummer and Brauner, 2011). The reaction of H+ and HCO3 − that exit the RBC is not catalyzed, and the slow rate of this reaction limits the speed of the Jacobs-Stewart cycle, allowing for alkalinization of the RBC interior (Motais et al., 1989; Nikinmaa et al., 1990). In tissues where extracellular CA is present to catalyze the dehydration reaction, the rapid return of protons to the RBC by the Jacobs-Stewart cycle may short-circuit the adrenergic response, favouring O2 unloading from pH-sensitive Hb and hence O2 delivery to the tissue (Rummer and Brauner, 2011). Although it is widely accepted that intracellular CA contributes to the RBC adrenergic response (e.g. Motais et al., 1989), empirical data in support of this hypothesis are sparse. Therefore, in the present study the RBC adrenergic response was evaluated in the absence and presence of the CA inhibitor acetazolamide; CA inhibition was predicted to diminish both the speed and magnitude of the adrenergic response. The rainbow trout, Oncorhynchus mykiss, was used as the study animal because the blood of this species exhibits a well-studied and robust RBC adrenergic response. The RBC
D.W. Carrie, K.M. Gilmour / Respiratory Physiology & Neurobiology 184 (2012) 60–64
adrenergic response was assessed both in vitro, in a simplified system with control over most variables, and in vivo, to place emphasis on the physiological significance of intracellular CA to the RBC adrenergic response. Detection of a relationship between CA activity and the RBC adrenergic response within a single species was viewed as a first step towards exploring the possibility that interspecific variation in RBC CA activity in teleost fish could be related to the magnitude and/or function of the RBC adrenergic response. 2. Materials and methods 2.1. Experimental animals Rainbow trout [O. mykiss Walbaum; mass 422 ± 15 g (mean ± standard deviation of the mean, SEM), N = 32] were obtained from Linwood Acres Trout Farm (Campbellcroft, ON, Canada). Fish were maintained on a 12 h:12 h, L:D photoperiod in large fibreglass aquaria supplied with flowing, aerated and dechloraminated city of Ottawa tap water at 13 ◦ C, and were fed every other day using commercial trout pellets. Trout were allowed to acclimate to the holding conditions for at least 2 weeks prior to any experimentation and were not fed for at least 24 h before experimentation. All experiments complied with institutional guidelines, were approved by the Animal Care Committee of the University of Ottawa (protocol BL-228), and were in accordance with guidelines of the Canadian Council on Animal Care for the use of animals in research and teaching. Fish used for in vitro experiments were fitted with an indwelling dorsal aortic cannula to permit blood sampling, whereas both the dorsal aorta and caudal vein were cannulated in fish used for in vivo experiments to allow an extracorporeal blood circuit to be established (see below). Fish were anaesthetized in an oxygenated solution of 1:10,000 (w/v) ethyl-p-aminobenzoate (benzocaine; Sigma), and placed onto an operating table that permitted continuous irrigation of the gills with the same anaesthetic solution. Blood vessels were cannulated with flexible polyethylene tubing (PE 50; Clay Adams), and cannulae were flushed daily with saline (0.9% NaCl). Trout were revived on the operating table and then placed in individual holding boxes supplied with flowing, aerated water for a 24–48 h recovery period. 2.2. In vitro experiments The extent of RBC adrenergic stimulation in vitro was assessed by real-time monitoring of the pH of whole blood flowing in a semi-closed loop, essentially as described by Perry and Thomas (1993). The synthetic -adrenergic agonist isoproterenol was used to activate the RBC adrenergic response, and acetazolamide was used to inhibit RBC CA activity. Briefly, ∼5 mL of blood (N = 16) was withdrawn slowly from the dorsal aortic cannula into a preheparinized syringe; blood sampling was terminated at the first sign of agitation or struggling. The blood was mixed with heparin (to achieve a final concentration of 50 IU mL−1 ; ammonium heparin, Sigma), transferred to a 50 mL round-bottomed flask held on ice, gassed with O2 for 2 min, and allowed to rest for 3 h before experimentation, to promote the degradation of any endogenous catecholamines (Perry et al., 1996). Blood haematocrit was measured [duplicate heparinized microcapillary tubes (Fisher), centrifuged at 6000 × g for 5 min] to ensure that it was between 20 and 25% (mean = 20.4 ± 0.4%, N = 16). A 1 mL aliquot of blood was transferred to an Eschweiler tonometry flask (5 mL volume) that was partially immersed in a 13 ◦ C water bath, and the blood was then equilibrated for 20 min with a humidified, hypoxic gas mixture (0.25% CO2 , 1.5% O2 , balance N2 ; GF-3/MP gas mixing flowmeter, Cameron Instruments). Pre-incubation of blood samples to hypoxic
61
conditions was designed to maximize the magnitude of the RBC adrenergic response (e.g. Perry et al., 1996). Acetazolamide (final concentration 5 × 10−4 mol L−1 ) or saline (control) was added to the blood sample for the equilibration period. Following equilibration, blood was pumped (peristaltic pump; 0.4 ml min−1 ) through a temperature-controlled chamber (13 ◦ C) housing a pH electrode (pHC4000 combined pH electrode; Radiometer) linked to a blood gas analyzer (BGM200; Cameron Instruments) and data acquisition system (Biopac Systems with AcqKnowledge software v3.7.3, sampling rate set to 5 Hz; Harvard Apparatus Canada), and then returned to the tonometry flask. The pH electrode was calibrated by pumping precision buffers through the loop, and prior to pumping blood, the loop was pre-treated for 20 min with heparinized (540 IU ml−1 ) saline to prevent blood clotting. Upon achieving a stable baseline pH (usually within 5 min), 20 L of isoproterenol stock solution was added to the tonometer (final concentration 10−7 M) and whole blood pH was monitored for an additional 30 min. The maximal fall in pH (pH, usually achieved within 5 min of agonist addition) was used as an index of the magnitude of the adrenergic response. The time required to reach half the maximum drop in pH was used as an index of the rapidity of the pH fall. The rate of pH recovery was determined as the rate of change of pH from the time of lowest pH to the time at which 50% recovery of pH had occurred. Preliminary experiments confirmed that addition of saline (carrier vehicle) to the tonometry flask was without significant effect on blood pH, either under control conditions (incubation with saline, pH = 0.02 ± 0.01, N = 8, not significantly different from zero, one-sample Student’s t-test, P = 0.164) or following incubation with acetazolamide (pH = 0.02 ± 0.01, N = 8, not significantly different from zero, one-sample Student’s t-test, P = 0.086). Because acetazolamide is a buffer and changes in plasma buffering could affect the magnitude of pH (independently of CA inhibition), the non-bicarbonate buffer capacity of rainbow trout plasma with and without added acetazolamide was measured. Blood samples were collected from a separate group of fish (N = 4) and centrifuged to obtained separated plasma. Plasma non-bicarbonate buffer capacity was determined by measuring pH (using the tonometry set-up described above) and total CO2 concentration (duplicate 50 L samples; Corning 965 Carbon Dioxide Analyser, Olympic Analytical Service) for plasma samples to which acetazolamide or saline had been added (as above) and that had been equilibrated with a range of CO2 tensions (0.1–2.1%; GF-3/MP gas mixing flowmeter, Cameron Instruments). Plasma non-bicarbonate buffer capacity was calculated as the slope of the relationship between total CO2 concentration and pH. 2.3. In vivo experiments The extent of RBC adrenergic stimulation in vivo was assessed by continuous, real-time monitoring of arterial blood pH (pHa) using an extracorporeal blood circulation, essentially as described by Perry and Thomas (1993). The set-up described above also was used for these experiments; blood was pumped from the dorsal aorta of the fish and returned to the caudal vein. In addition to pHa, arterial CO2 tension (PaCO2 ) was monitored by adding a PaCO2 electrode (16-720, Microelectrodes Inc.) to the external loop. The PaCO2 electrode was calibrated using water equilibrated with appropriate gas mixtures (0.5 and 1% CO2 ; GF-3/MP gas mixing flowmeter, Cameron Instruments), and was linked to the blood gas analyzer and data acquisition system. Three treatment groups were utilized, saline-treated (control; N = 6), acetazolamide-treated (N = 6) and hypercapnia-exposed (N = 4). Each fish received two injections of isoproterenol (1 mL kg−1 injection volume, final nominal circulating concentration 10−7 mol L−1 ), spaced approximately 1 h apart, via the dorsal aortic cannula. Preliminary experiments confirmed that injection of
62
D.W. Carrie, K.M. Gilmour / Respiratory Physiology & Neurobiology 184 (2012) 60–64
Fig. 1. The effect of CA inhibition by acetazolamide on the RBC adrenergic response of rainbow trout (Oncorhynchus mykiss) blood in vitro. (A) The magnitude of the fall in pH (pH) following activation of the RBC adrenergic response by isoproterenol addition. (B) The time required after addition of isoproterenol to reach half the maximum drop in pH, an index of the rapidity of the pH fall. (C) The rate of recovery of pH calculated as the rate of change of pH from the time of lowest pH to the time at which 50% recovery of pH had occurred. In all panels, values are means ± SEM (N = 16 for A and B, N = 8 for C). An asterisk indicates a significant difference between treatments (paired Student’s t-tests, P = 0.018, <0.001 and 0.003 for panels A, B and C, respectively). (D) Representative original data recordings depicting the effect of CA inhibition on blood pH in vitro following activation of the RBC adrenergic response by addition of isoproterenol (indicated by the arrow). Values for pH have been normalized by subtracting from each value the initial pH prior to isoproterenol injection.
saline (carrier vehicle) was without significant effect on blood pH (pH = 0.00 ± 0.00, N = 6) and therefore control (saline) injections were not performed. To examine the role of CA in the adrenergic response, trout were treated with acetazolamide (30 mg kg−1 , final nominal circulating concentration 4.5 × 10−4 mol L−1 ) between the first and second isoproterenol injections. A second group received saline instead of acetazolamide and served as a control for the effect of repeated isoproterenol injection. Because acetazolamide caused a respiratory acidosis by inhibiting RBC CA and hence CO2 excretion, a third treatment group was included in the experimental design in which exposure to hypercapnia was used to alter pHa and PaCO2 to levels comparable to those measured in acetazolamide-treated fish. Preliminary trials revealed that supplying the experimental chamber with water from an equilibration column gassed with 3.5% CO2 in air (GF-3/MP gas mixing flowmeter; Cameron Instruments) achieved appropriate PaCO2 levels. The maximal fall in pHa (pHa; usually achieved within 5 min of agonist addition) was used as an index of the magnitude of the adrenergic response. The rate of pH recovery was calculated as the rate of change of pH from the time of lowest pH to the time at which 50% recovery of pH had occurred.
2.4. Statistical analysis All data are presented as mean values ±1 SEM. The statistical significance of treatment effects was assessed by paired or unpaired Student’s t-tests (or their non-parametric equivalents where the data failed assumptions of normality or equal variance), or by twoway repeated measures analysis of variance (RM ANOVA) followed by Holm-Sidak post hoc tests, as appropriate. The fiducial limit of significance was 5%. Statistical analyses were carried out using commercial software (SigmaStat v3.5 in SigmaPlot v11; SPSS Scientific Inc.). 3. Results Inhibition of RBC CA using acetazolamide caused a significant 13% decrease in the magnitude of the RBC adrenergic response in vitro (Fig. 1A; paired Student’s t-test, P = 0.018). The fall in pH upon addition of isoproterenol was significantly slower with CA inhibition (Fig. 1B; paired Student’s t-test, P < 0.001), as was the rate of recovery of pH (Fig. 1C, paired Student’s t-test, P = 0.003). These differences occurred in the absence of a significant
D.W. Carrie, K.M. Gilmour / Respiratory Physiology & Neurobiology 184 (2012) 60–64
63
Fig. 2. The effect of CA inhibition by acetazolamide on the RBC adrenergic response of rainbow trout (Oncorhynchus mykiss) in vivo. (A) The magnitude of the fall in arterial pH (pHa) following activation of the RBC adrenergic response by injection of isoproterenol in vivo before and after treatment of fish with saline (control; N = 6) or acetazolamide (Az; N = 6), or before and after exposure of fish to hypercapnia (N = 4). An asterisk (*) indicates a statistically significant difference in the magnitude of pH before and after treatment, within a treatment group (paired Student’s t-tests, P = 0.956, 0.006 and 0.388 for control, Az and hypercapnia, respectively). (B) The rate of recovery of pHa calculated as the rate of change of pH from the time of lowest pH to the time at which 50% recovery of pH had occurred. An asterisk (*) indicates a statistically significant difference in the rate of pH recovery before and after treatment, within a treatment group (paired Student’s t-tests, P = 0.629, 0.012 and 0.608 for control, Az and hypercapnia, respectively). Values in both panels are means ± SEM. ISO 1 denotes pHa or the rate of recovery of pHa for the first isoproterenol injection, before treatment with saline, Az or hypercapnia, whereas ISO 2 denotes measurements for the second isoproterenol injection, i.e. after treatment. (C) Representative original data recordings illustrating the changes in arterial pH associated with activation of the RBC adrenergic response by injection of isoproterenol (at the arrow) before (control) and after inhibition of RBC CA by acetazolamide. Values for pHa have been normalized by subtracting from each value the initial pHa prior to isoproterenol injection.
effect of acetazolamide on plasma non-bicarbonate buffer capacity (−4.3 ± 2.0 mmol L−1 pH unit−1 for acetazolamide-treated and −2.8 ± 1.4 mmol L−1 pH unit−1 for control plasma, N = 4 in each case; paired Student’s t-test, P = 0.64). The magnitude of the fall in pH with isoproterenol injection also was significantly reduced, by 81%, when RBC CA was inhibited in vivo (Fig. 2A; paired Student’s t-test, P = 0.006), as was the rate of recovery of pH (Fig. 2B; paired Student’s t-test, P = 0.012). Acetazolamide treatment caused PaCO2 to rise by 10.9 ± 3.0 Torr (N = 4) and pHa to fall by 0.67 ± 0.07 pH units (N = 6); comparable changes in PaCO2 (8.0 ± 0.7 Torr, N = 4; not significantly different from the rise with acetazolamide, rank sum test, P = 0.486) and pH (−0.60 ± 0.1 pH units, N = 4; see Table 1) were achieved by exposure to hypercapnia. However, neither the magnitude of the pH change nor the rate of pH recovery differed between isoproterenol injections in fish exposed to hypercapnia, or in control fish (Fig. 2A and B; paired Student’s t-tests, P > 0.05 in all cases). 4. Discussion In the present study, both the magnitude and speed of the RBC adrenergic response were dependent upon RBC CA activity. These findings support the hypothesis that intracellular CA contributes to the RBC adrenergic response, by catalyzing the hydration of
CO2 to provide H+ for export from the RBC via the NHE, and HCO3 − for (slower) export via the Cl− /HCO3 − exchanger. The role of intracellular CA in the RBC adrenergic response also was examined by Nikinmaa et al. (1990). These authors reported a larger isoproterenol-induced fall in pH following acetazolamide treatment for trout RBCs suspended in physiological saline (Nikinmaa et al., 1990), i.e. a result opposite to that of the present study. An explanation for this difference may lie in the use of RBCs suspended
Table 1 Arterial blood pH prior to isoproterenol injection in vivo in rainbow trout (Oncorhynchus mykiss). Treatment
ISO 1 (pre-treatment)
ISO 2 (post-treatment)
Acetazolamide (N = 6) Hypercapnia (N = 4) Control (N = 6)
7.97 ± 0.12a 8.03 ± 0.08a 8.02 ± 0.06a
7.30 ± 0.16c * 7.46 ± 0.05c * 7.90 ± 0.12d
ISO 1 denotes a measurement preceding the first isoproterenol injection, whereas ISO 2 denotes arterial blood pH measured before the second isoproterenol injection, i.e. after treatment. Values are means ± SEM. An asterisk (*) indicates a significant difference between ISO 1 and ISO 2 within a treatment group, and within a sampling time, groups that share a letter are not significantly different from one another (two-way RM ANOVA, P = 0.115 for the effect of treatment group, P < 0.001 for sample time, P < 0.001 for the interaction of treatment group and sample time).
64
D.W. Carrie, K.M. Gilmour / Respiratory Physiology & Neurobiology 184 (2012) 60–64
in saline (Nikinmaa et al., 1990) versus plasma (present study). Any CA released from lysed RBCs into the saline would tend to short-circuit the adrenergic response, reducing the magnitude of the pH change (Motais et al., 1989; Nikinmaa et al., 1990). Addition of acetazolamide would then inhibit this extracellular CA, leading to an increase in the extent of the pH fall. This problem is avoided when RBCs are suspended in plasma because trout plasma contains an endogenous CA inhibitor (Haswell and Randall, 1976). The linkage between intracellular CA activity and the magnitude and speed of the RBC adrenergic response, demonstrated in the present study using pharmacological manipulation of CA activity in a single species, suggests that variation in RBC CA activity across species should be investigated with respect to variation in the RBC adrenergic response. Current models propose that the NHE evolved in teleost fish as part of a complex physiological system for effective O2 secretion to the eye (Berenbrink et al., 2005). The evolution of low haemoglobin buffer values together with a pronounced Bohr-Root effect and a choroid rete provided a mechanism for ocular O2 secretion, but at the cost of causing O2 loading into the blood to be vulnerable to systemic acidosis. Evolution of the NHE reduced this risk by providing a mechanism for RBC alkalinization. To add to this hypothesis, recent work suggested that in addition to safeguarding O2 loading at the gill, the NHE could be exploited to promote O2 delivery in tissues possessing extracellular CA activity (Rummer and Brauner, 2011). Addition of CA to adrenergically stimulated trout RBCs in vitro short-circuited NHEmediated pH regulation by catalyzing CO2 formation in plasma. Owing to the Bohr-Root effects, the resulting rapid transfer of protons into the RBC caused PO2 to rise (Rummer and Brauner, 2011). The effectiveness of this mechanism will depend upon intracellular CA to catalyze the hydration of CO2 entering the cell. Indeed, both the rate and magnitude of proton-driven O2 release from haemoglobin were reduced by CA inhibition in suspensions of RBCs from goosefish (Lophius americanus), another teleost fish (Maren and Swenson, 1980). Collectively, these data suggest that species with pronounced Bohr-Root effect haemoglobins and marked RBC adrenergic responses also should possess high levels of RBC CA activity, because high intracellular CA activity permits full expression of both NHE-mediated pH regulation and proton-driven O2 release from haemoglobin. By linking the rate and magnitude of the RBC adrenergic response to intracellular CA activity both in vitro and in vivo in a single species, the present study fuels speculation
that interspecific differences in RBC CA activity may be associated with the expression and/or function of the RBC adrenergic response. Acetazolamide is expected to fully inhibit RBC CA activity (Gilmour et al., 2001), creating a more substantial difference in CA activity in the present experiments than is likely to occur in comparisons of RBC CA activity across species. Thus, comparisons of CA activity and adrenergic responses for RBCs from a range of fish species are now required. Acknowledgements This work was supported by NSERC of Canada Discovery and Research Tools & Instruments grants to K.M.G. D.W.C. received an NSERC Undergraduate Student Research Award. Thanks are extended to Bill Fletcher for his excellent care of the fish in the aquatic facility. References Berenbrink, M., Koldkjær, P., Kepp, O., Cossins, A.R., 2005. Evolution of oxygen secretion in fishes and the emergence of a complex physiological system. Science 307, 1752–1757. Esbaugh, A., Lund, S.G., Tufts, B.L., 2004. Comparative physiology and molecular analysis of carbonic anhydrase from the red blood cells of teleost fish. Journal of Comparative and Physiology B 174, 429–438. Gilmour, K.M., 2010. Perspectives on carbonic anhydrase. Comparative Biochemistry and Physiology A 157, 193–197. Gilmour, K.M., Perry, S.F., Bernier, N.J., Henry, R.P., Wood, C.M., 2001. Extracellular carbonic anhydrase in dogfish, Squalus acanthias A role in CO2 excretion. Physiological and Biochemical Zoology 74, 477–492. Haswell, M.S., Randall, D.J., 1976. Carbonic anhydrase inhibitor in trout plasma. Respiration Physiology 28, 17–27. Maren, T.H., Swenson, E.R., 1980. A comparative study of the kinetics of the Bohr effect in vertebrates. Journal of Physiology 303, 535–547. Motais, R., Fievet, B., Garcia-Romeu, F., Thomas, S., 1989. Na+ –H+ exchange and pH regulation in red blood cells: role of uncatalyzed H2 CO3 dehydration. American Journal of Physiology 256, C728–C735. Nikinmaa, M., Tiihonen, K., Paajaste, M., 1990. Adrenergic control of red cell pH in salmonid fish: roles of the sodium/proton exchange, Jacobs-Stewart cycle and membrane potential. Journal of Experimental Biology 154, 257–271. Perry, S.F., Reid, S.G., Salama, A., 1996. The effects of repeated physical stress on the adrenergic response of the rainbow trout red blood cell. Journal of Experimental Biology 199, 549–562. Perry, S.F., Thomas, S., 1993. Rapid respiratory changes in trout red blood cells during Na+ /H+ exchange activation. Journal of Experimental Biology 180, 27–37. Rummer, J.L., Brauner, C.J., 2011. Plasma-accessible carbonic anhydrase at the tissue of a teleost fish may greatly enhance oxygen delivery: in vitro evidence in rainbow trout, Oncorhynchus mykiss. Journal of Experimental Biology 214, 2319–2328.
Contents lists available at SciVerse ScienceDirect
Respiratory Physiology & Neurobiology journal homepage: www.elsevier.com/locate/resphysiol
Short communication
Intracellular carbonic anhydrase contributes to the red blood cell adrenergic response in rainbow trout Oncorhynchus mykiss D.W. Carrie, K.M. Gilmour ∗ Department of Biology, University of Ottawa, ON, Canada
a r t i c l e
i n f o
Article history: Accepted 26 July 2012 Keywords: Rainbow trout Carbonic anhydrase Red blood cell NHE pH
a b s t r a c t In many teleost fish, catecholamines activate a red blood cell (RBC) Na+ /H+ exchanger (NHE), raising RBC intracellular pH to protect haemoglobin-O2 loading. The present study tested the hypothesis that RBC intracellular carbonic anhydrase (CA) contributes to this adrenergic response. The pH of rainbow trout (Oncorhynchus mykiss) blood was monitored continuously in vitro using blood flowing in a semiclosed loop or in vivo using an extracorporeal circulation. Addition or injection of isoproterenol activated the NHE, causing blood pH to fall (in vitro pH = −0.28 ± 0.03 pH units, N = 16; in vivo, −0.12 ± 0.02 pH units, N = 6). Both in vitro and in vivo, inhibition of RBC CA by acetazolamide significantly decreased the magnitude of the adrenergic response (in vitro, pH = −0.22 ± 0.02 pH units, N = 16; in vivo, −0.02 ± 0.01 pH units, N = 6) as well as the rate of recovery of blood pH following the adrenergic response. These results support the hypothesis that RBC intracellular CA plays an important role in the RBC adrenergic response of rainbow trout, and fuel speculation that interspecific differences in RBC CA activity are associated with the magnitude of the RBC adrenergic response. © 2012 Elsevier B.V. All rights reserved.
1. Introduction An enduring (and under-studied) question in the literature concerns the significance of interspecific differences in red blood cell (RBC) carbonic anhydrase (CA) activity (Gilmour, 2010). Carbonic anhydrase catalyzes the reversible reaction of CO2 and water to bicarbonate ions and protons, and therefore plays a fundamental role in a host of physiological processes including metabolism, gas exchange and acid–base balance. Differences in RBC CA activity spanning an up to 25-fold range have been documented across species within vertebrates in general (see Gilmour, 2010) and teleost fish specifically (Esbaugh et al., 2004). However, the physiological significance of these differences remains unclear. Within teleost fish, one possibility that warrants consideration is that interspecific differences in RBC CA activity are associated with the magnitude of the RBC adrenergic response. In many teleost fish, catecholamines (adrenaline and noradrenaline) that are released into the circulation bind to -adrenoreceptors on the RBC membrane, activating a Na+ /H+ exchanger (NHE) via a cAMP signal transduction pathway. This response alkalinizes the RBC intracellular environment, and
∗ Corresponding author at: Department of Biology, University of Ottawa, 30 Marie Curie, Ottawa, ON K1N 6T2, Canada. Tel.: +1 613 562 5800x6004; fax: +1 613 562 5486. E-mail address: [email protected] (K.M. Gilmour). 1569-9048/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.resp.2012.07.022
is thought to have evolved in parallel with Bohr-Root shift haemoglobins (Hb; i.e. Hb for which O2 binding is pH sensitive) to protect Hb-O2 loading at the gill during systemic acidosis (Berenbrink et al., 2005). Carbonic anhydrase is expected to contribute to the RBC adrenergic response by catalyzing the hydration of CO2 within the RBC to produce the protons that are extruded by the NHE. This reaction also yields HCO3 − that exit the RBC more slowly, in exchange for Cl− entry. In plasma, on the other hand, it is the absence of CA that is important, at least in the gill (Rummer and Brauner, 2011). The reaction of H+ and HCO3 − that exit the RBC is not catalyzed, and the slow rate of this reaction limits the speed of the Jacobs-Stewart cycle, allowing for alkalinization of the RBC interior (Motais et al., 1989; Nikinmaa et al., 1990). In tissues where extracellular CA is present to catalyze the dehydration reaction, the rapid return of protons to the RBC by the Jacobs-Stewart cycle may short-circuit the adrenergic response, favouring O2 unloading from pH-sensitive Hb and hence O2 delivery to the tissue (Rummer and Brauner, 2011). Although it is widely accepted that intracellular CA contributes to the RBC adrenergic response (e.g. Motais et al., 1989), empirical data in support of this hypothesis are sparse. Therefore, in the present study the RBC adrenergic response was evaluated in the absence and presence of the CA inhibitor acetazolamide; CA inhibition was predicted to diminish both the speed and magnitude of the adrenergic response. The rainbow trout, Oncorhynchus mykiss, was used as the study animal because the blood of this species exhibits a well-studied and robust RBC adrenergic response. The RBC
D.W. Carrie, K.M. Gilmour / Respiratory Physiology & Neurobiology 184 (2012) 60–64
adrenergic response was assessed both in vitro, in a simplified system with control over most variables, and in vivo, to place emphasis on the physiological significance of intracellular CA to the RBC adrenergic response. Detection of a relationship between CA activity and the RBC adrenergic response within a single species was viewed as a first step towards exploring the possibility that interspecific variation in RBC CA activity in teleost fish could be related to the magnitude and/or function of the RBC adrenergic response. 2. Materials and methods 2.1. Experimental animals Rainbow trout [O. mykiss Walbaum; mass 422 ± 15 g (mean ± standard deviation of the mean, SEM), N = 32] were obtained from Linwood Acres Trout Farm (Campbellcroft, ON, Canada). Fish were maintained on a 12 h:12 h, L:D photoperiod in large fibreglass aquaria supplied with flowing, aerated and dechloraminated city of Ottawa tap water at 13 ◦ C, and were fed every other day using commercial trout pellets. Trout were allowed to acclimate to the holding conditions for at least 2 weeks prior to any experimentation and were not fed for at least 24 h before experimentation. All experiments complied with institutional guidelines, were approved by the Animal Care Committee of the University of Ottawa (protocol BL-228), and were in accordance with guidelines of the Canadian Council on Animal Care for the use of animals in research and teaching. Fish used for in vitro experiments were fitted with an indwelling dorsal aortic cannula to permit blood sampling, whereas both the dorsal aorta and caudal vein were cannulated in fish used for in vivo experiments to allow an extracorporeal blood circuit to be established (see below). Fish were anaesthetized in an oxygenated solution of 1:10,000 (w/v) ethyl-p-aminobenzoate (benzocaine; Sigma), and placed onto an operating table that permitted continuous irrigation of the gills with the same anaesthetic solution. Blood vessels were cannulated with flexible polyethylene tubing (PE 50; Clay Adams), and cannulae were flushed daily with saline (0.9% NaCl). Trout were revived on the operating table and then placed in individual holding boxes supplied with flowing, aerated water for a 24–48 h recovery period. 2.2. In vitro experiments The extent of RBC adrenergic stimulation in vitro was assessed by real-time monitoring of the pH of whole blood flowing in a semi-closed loop, essentially as described by Perry and Thomas (1993). The synthetic -adrenergic agonist isoproterenol was used to activate the RBC adrenergic response, and acetazolamide was used to inhibit RBC CA activity. Briefly, ∼5 mL of blood (N = 16) was withdrawn slowly from the dorsal aortic cannula into a preheparinized syringe; blood sampling was terminated at the first sign of agitation or struggling. The blood was mixed with heparin (to achieve a final concentration of 50 IU mL−1 ; ammonium heparin, Sigma), transferred to a 50 mL round-bottomed flask held on ice, gassed with O2 for 2 min, and allowed to rest for 3 h before experimentation, to promote the degradation of any endogenous catecholamines (Perry et al., 1996). Blood haematocrit was measured [duplicate heparinized microcapillary tubes (Fisher), centrifuged at 6000 × g for 5 min] to ensure that it was between 20 and 25% (mean = 20.4 ± 0.4%, N = 16). A 1 mL aliquot of blood was transferred to an Eschweiler tonometry flask (5 mL volume) that was partially immersed in a 13 ◦ C water bath, and the blood was then equilibrated for 20 min with a humidified, hypoxic gas mixture (0.25% CO2 , 1.5% O2 , balance N2 ; GF-3/MP gas mixing flowmeter, Cameron Instruments). Pre-incubation of blood samples to hypoxic
61
conditions was designed to maximize the magnitude of the RBC adrenergic response (e.g. Perry et al., 1996). Acetazolamide (final concentration 5 × 10−4 mol L−1 ) or saline (control) was added to the blood sample for the equilibration period. Following equilibration, blood was pumped (peristaltic pump; 0.4 ml min−1 ) through a temperature-controlled chamber (13 ◦ C) housing a pH electrode (pHC4000 combined pH electrode; Radiometer) linked to a blood gas analyzer (BGM200; Cameron Instruments) and data acquisition system (Biopac Systems with AcqKnowledge software v3.7.3, sampling rate set to 5 Hz; Harvard Apparatus Canada), and then returned to the tonometry flask. The pH electrode was calibrated by pumping precision buffers through the loop, and prior to pumping blood, the loop was pre-treated for 20 min with heparinized (540 IU ml−1 ) saline to prevent blood clotting. Upon achieving a stable baseline pH (usually within 5 min), 20 L of isoproterenol stock solution was added to the tonometer (final concentration 10−7 M) and whole blood pH was monitored for an additional 30 min. The maximal fall in pH (pH, usually achieved within 5 min of agonist addition) was used as an index of the magnitude of the adrenergic response. The time required to reach half the maximum drop in pH was used as an index of the rapidity of the pH fall. The rate of pH recovery was determined as the rate of change of pH from the time of lowest pH to the time at which 50% recovery of pH had occurred. Preliminary experiments confirmed that addition of saline (carrier vehicle) to the tonometry flask was without significant effect on blood pH, either under control conditions (incubation with saline, pH = 0.02 ± 0.01, N = 8, not significantly different from zero, one-sample Student’s t-test, P = 0.164) or following incubation with acetazolamide (pH = 0.02 ± 0.01, N = 8, not significantly different from zero, one-sample Student’s t-test, P = 0.086). Because acetazolamide is a buffer and changes in plasma buffering could affect the magnitude of pH (independently of CA inhibition), the non-bicarbonate buffer capacity of rainbow trout plasma with and without added acetazolamide was measured. Blood samples were collected from a separate group of fish (N = 4) and centrifuged to obtained separated plasma. Plasma non-bicarbonate buffer capacity was determined by measuring pH (using the tonometry set-up described above) and total CO2 concentration (duplicate 50 L samples; Corning 965 Carbon Dioxide Analyser, Olympic Analytical Service) for plasma samples to which acetazolamide or saline had been added (as above) and that had been equilibrated with a range of CO2 tensions (0.1–2.1%; GF-3/MP gas mixing flowmeter, Cameron Instruments). Plasma non-bicarbonate buffer capacity was calculated as the slope of the relationship between total CO2 concentration and pH. 2.3. In vivo experiments The extent of RBC adrenergic stimulation in vivo was assessed by continuous, real-time monitoring of arterial blood pH (pHa) using an extracorporeal blood circulation, essentially as described by Perry and Thomas (1993). The set-up described above also was used for these experiments; blood was pumped from the dorsal aorta of the fish and returned to the caudal vein. In addition to pHa, arterial CO2 tension (PaCO2 ) was monitored by adding a PaCO2 electrode (16-720, Microelectrodes Inc.) to the external loop. The PaCO2 electrode was calibrated using water equilibrated with appropriate gas mixtures (0.5 and 1% CO2 ; GF-3/MP gas mixing flowmeter, Cameron Instruments), and was linked to the blood gas analyzer and data acquisition system. Three treatment groups were utilized, saline-treated (control; N = 6), acetazolamide-treated (N = 6) and hypercapnia-exposed (N = 4). Each fish received two injections of isoproterenol (1 mL kg−1 injection volume, final nominal circulating concentration 10−7 mol L−1 ), spaced approximately 1 h apart, via the dorsal aortic cannula. Preliminary experiments confirmed that injection of
62
D.W. Carrie, K.M. Gilmour / Respiratory Physiology & Neurobiology 184 (2012) 60–64
Fig. 1. The effect of CA inhibition by acetazolamide on the RBC adrenergic response of rainbow trout (Oncorhynchus mykiss) blood in vitro. (A) The magnitude of the fall in pH (pH) following activation of the RBC adrenergic response by isoproterenol addition. (B) The time required after addition of isoproterenol to reach half the maximum drop in pH, an index of the rapidity of the pH fall. (C) The rate of recovery of pH calculated as the rate of change of pH from the time of lowest pH to the time at which 50% recovery of pH had occurred. In all panels, values are means ± SEM (N = 16 for A and B, N = 8 for C). An asterisk indicates a significant difference between treatments (paired Student’s t-tests, P = 0.018, <0.001 and 0.003 for panels A, B and C, respectively). (D) Representative original data recordings depicting the effect of CA inhibition on blood pH in vitro following activation of the RBC adrenergic response by addition of isoproterenol (indicated by the arrow). Values for pH have been normalized by subtracting from each value the initial pH prior to isoproterenol injection.
saline (carrier vehicle) was without significant effect on blood pH (pH = 0.00 ± 0.00, N = 6) and therefore control (saline) injections were not performed. To examine the role of CA in the adrenergic response, trout were treated with acetazolamide (30 mg kg−1 , final nominal circulating concentration 4.5 × 10−4 mol L−1 ) between the first and second isoproterenol injections. A second group received saline instead of acetazolamide and served as a control for the effect of repeated isoproterenol injection. Because acetazolamide caused a respiratory acidosis by inhibiting RBC CA and hence CO2 excretion, a third treatment group was included in the experimental design in which exposure to hypercapnia was used to alter pHa and PaCO2 to levels comparable to those measured in acetazolamide-treated fish. Preliminary trials revealed that supplying the experimental chamber with water from an equilibration column gassed with 3.5% CO2 in air (GF-3/MP gas mixing flowmeter; Cameron Instruments) achieved appropriate PaCO2 levels. The maximal fall in pHa (pHa; usually achieved within 5 min of agonist addition) was used as an index of the magnitude of the adrenergic response. The rate of pH recovery was calculated as the rate of change of pH from the time of lowest pH to the time at which 50% recovery of pH had occurred.
2.4. Statistical analysis All data are presented as mean values ±1 SEM. The statistical significance of treatment effects was assessed by paired or unpaired Student’s t-tests (or their non-parametric equivalents where the data failed assumptions of normality or equal variance), or by twoway repeated measures analysis of variance (RM ANOVA) followed by Holm-Sidak post hoc tests, as appropriate. The fiducial limit of significance was 5%. Statistical analyses were carried out using commercial software (SigmaStat v3.5 in SigmaPlot v11; SPSS Scientific Inc.). 3. Results Inhibition of RBC CA using acetazolamide caused a significant 13% decrease in the magnitude of the RBC adrenergic response in vitro (Fig. 1A; paired Student’s t-test, P = 0.018). The fall in pH upon addition of isoproterenol was significantly slower with CA inhibition (Fig. 1B; paired Student’s t-test, P < 0.001), as was the rate of recovery of pH (Fig. 1C, paired Student’s t-test, P = 0.003). These differences occurred in the absence of a significant
D.W. Carrie, K.M. Gilmour / Respiratory Physiology & Neurobiology 184 (2012) 60–64
63
Fig. 2. The effect of CA inhibition by acetazolamide on the RBC adrenergic response of rainbow trout (Oncorhynchus mykiss) in vivo. (A) The magnitude of the fall in arterial pH (pHa) following activation of the RBC adrenergic response by injection of isoproterenol in vivo before and after treatment of fish with saline (control; N = 6) or acetazolamide (Az; N = 6), or before and after exposure of fish to hypercapnia (N = 4). An asterisk (*) indicates a statistically significant difference in the magnitude of pH before and after treatment, within a treatment group (paired Student’s t-tests, P = 0.956, 0.006 and 0.388 for control, Az and hypercapnia, respectively). (B) The rate of recovery of pHa calculated as the rate of change of pH from the time of lowest pH to the time at which 50% recovery of pH had occurred. An asterisk (*) indicates a statistically significant difference in the rate of pH recovery before and after treatment, within a treatment group (paired Student’s t-tests, P = 0.629, 0.012 and 0.608 for control, Az and hypercapnia, respectively). Values in both panels are means ± SEM. ISO 1 denotes pHa or the rate of recovery of pHa for the first isoproterenol injection, before treatment with saline, Az or hypercapnia, whereas ISO 2 denotes measurements for the second isoproterenol injection, i.e. after treatment. (C) Representative original data recordings illustrating the changes in arterial pH associated with activation of the RBC adrenergic response by injection of isoproterenol (at the arrow) before (control) and after inhibition of RBC CA by acetazolamide. Values for pHa have been normalized by subtracting from each value the initial pHa prior to isoproterenol injection.
effect of acetazolamide on plasma non-bicarbonate buffer capacity (−4.3 ± 2.0 mmol L−1 pH unit−1 for acetazolamide-treated and −2.8 ± 1.4 mmol L−1 pH unit−1 for control plasma, N = 4 in each case; paired Student’s t-test, P = 0.64). The magnitude of the fall in pH with isoproterenol injection also was significantly reduced, by 81%, when RBC CA was inhibited in vivo (Fig. 2A; paired Student’s t-test, P = 0.006), as was the rate of recovery of pH (Fig. 2B; paired Student’s t-test, P = 0.012). Acetazolamide treatment caused PaCO2 to rise by 10.9 ± 3.0 Torr (N = 4) and pHa to fall by 0.67 ± 0.07 pH units (N = 6); comparable changes in PaCO2 (8.0 ± 0.7 Torr, N = 4; not significantly different from the rise with acetazolamide, rank sum test, P = 0.486) and pH (−0.60 ± 0.1 pH units, N = 4; see Table 1) were achieved by exposure to hypercapnia. However, neither the magnitude of the pH change nor the rate of pH recovery differed between isoproterenol injections in fish exposed to hypercapnia, or in control fish (Fig. 2A and B; paired Student’s t-tests, P > 0.05 in all cases). 4. Discussion In the present study, both the magnitude and speed of the RBC adrenergic response were dependent upon RBC CA activity. These findings support the hypothesis that intracellular CA contributes to the RBC adrenergic response, by catalyzing the hydration of
CO2 to provide H+ for export from the RBC via the NHE, and HCO3 − for (slower) export via the Cl− /HCO3 − exchanger. The role of intracellular CA in the RBC adrenergic response also was examined by Nikinmaa et al. (1990). These authors reported a larger isoproterenol-induced fall in pH following acetazolamide treatment for trout RBCs suspended in physiological saline (Nikinmaa et al., 1990), i.e. a result opposite to that of the present study. An explanation for this difference may lie in the use of RBCs suspended
Table 1 Arterial blood pH prior to isoproterenol injection in vivo in rainbow trout (Oncorhynchus mykiss). Treatment
ISO 1 (pre-treatment)
ISO 2 (post-treatment)
Acetazolamide (N = 6) Hypercapnia (N = 4) Control (N = 6)
7.97 ± 0.12a 8.03 ± 0.08a 8.02 ± 0.06a
7.30 ± 0.16c * 7.46 ± 0.05c * 7.90 ± 0.12d
ISO 1 denotes a measurement preceding the first isoproterenol injection, whereas ISO 2 denotes arterial blood pH measured before the second isoproterenol injection, i.e. after treatment. Values are means ± SEM. An asterisk (*) indicates a significant difference between ISO 1 and ISO 2 within a treatment group, and within a sampling time, groups that share a letter are not significantly different from one another (two-way RM ANOVA, P = 0.115 for the effect of treatment group, P < 0.001 for sample time, P < 0.001 for the interaction of treatment group and sample time).
64
D.W. Carrie, K.M. Gilmour / Respiratory Physiology & Neurobiology 184 (2012) 60–64
in saline (Nikinmaa et al., 1990) versus plasma (present study). Any CA released from lysed RBCs into the saline would tend to short-circuit the adrenergic response, reducing the magnitude of the pH change (Motais et al., 1989; Nikinmaa et al., 1990). Addition of acetazolamide would then inhibit this extracellular CA, leading to an increase in the extent of the pH fall. This problem is avoided when RBCs are suspended in plasma because trout plasma contains an endogenous CA inhibitor (Haswell and Randall, 1976). The linkage between intracellular CA activity and the magnitude and speed of the RBC adrenergic response, demonstrated in the present study using pharmacological manipulation of CA activity in a single species, suggests that variation in RBC CA activity across species should be investigated with respect to variation in the RBC adrenergic response. Current models propose that the NHE evolved in teleost fish as part of a complex physiological system for effective O2 secretion to the eye (Berenbrink et al., 2005). The evolution of low haemoglobin buffer values together with a pronounced Bohr-Root effect and a choroid rete provided a mechanism for ocular O2 secretion, but at the cost of causing O2 loading into the blood to be vulnerable to systemic acidosis. Evolution of the NHE reduced this risk by providing a mechanism for RBC alkalinization. To add to this hypothesis, recent work suggested that in addition to safeguarding O2 loading at the gill, the NHE could be exploited to promote O2 delivery in tissues possessing extracellular CA activity (Rummer and Brauner, 2011). Addition of CA to adrenergically stimulated trout RBCs in vitro short-circuited NHEmediated pH regulation by catalyzing CO2 formation in plasma. Owing to the Bohr-Root effects, the resulting rapid transfer of protons into the RBC caused PO2 to rise (Rummer and Brauner, 2011). The effectiveness of this mechanism will depend upon intracellular CA to catalyze the hydration of CO2 entering the cell. Indeed, both the rate and magnitude of proton-driven O2 release from haemoglobin were reduced by CA inhibition in suspensions of RBCs from goosefish (Lophius americanus), another teleost fish (Maren and Swenson, 1980). Collectively, these data suggest that species with pronounced Bohr-Root effect haemoglobins and marked RBC adrenergic responses also should possess high levels of RBC CA activity, because high intracellular CA activity permits full expression of both NHE-mediated pH regulation and proton-driven O2 release from haemoglobin. By linking the rate and magnitude of the RBC adrenergic response to intracellular CA activity both in vitro and in vivo in a single species, the present study fuels speculation
that interspecific differences in RBC CA activity may be associated with the expression and/or function of the RBC adrenergic response. Acetazolamide is expected to fully inhibit RBC CA activity (Gilmour et al., 2001), creating a more substantial difference in CA activity in the present experiments than is likely to occur in comparisons of RBC CA activity across species. Thus, comparisons of CA activity and adrenergic responses for RBCs from a range of fish species are now required. Acknowledgements This work was supported by NSERC of Canada Discovery and Research Tools & Instruments grants to K.M.G. D.W.C. received an NSERC Undergraduate Student Research Award. Thanks are extended to Bill Fletcher for his excellent care of the fish in the aquatic facility. References Berenbrink, M., Koldkjær, P., Kepp, O., Cossins, A.R., 2005. Evolution of oxygen secretion in fishes and the emergence of a complex physiological system. Science 307, 1752–1757. Esbaugh, A., Lund, S.G., Tufts, B.L., 2004. Comparative physiology and molecular analysis of carbonic anhydrase from the red blood cells of teleost fish. Journal of Comparative and Physiology B 174, 429–438. Gilmour, K.M., 2010. Perspectives on carbonic anhydrase. Comparative Biochemistry and Physiology A 157, 193–197. Gilmour, K.M., Perry, S.F., Bernier, N.J., Henry, R.P., Wood, C.M., 2001. Extracellular carbonic anhydrase in dogfish, Squalus acanthias A role in CO2 excretion. Physiological and Biochemical Zoology 74, 477–492. Haswell, M.S., Randall, D.J., 1976. Carbonic anhydrase inhibitor in trout plasma. Respiration Physiology 28, 17–27. Maren, T.H., Swenson, E.R., 1980. A comparative study of the kinetics of the Bohr effect in vertebrates. Journal of Physiology 303, 535–547. Motais, R., Fievet, B., Garcia-Romeu, F., Thomas, S., 1989. Na+ –H+ exchange and pH regulation in red blood cells: role of uncatalyzed H2 CO3 dehydration. American Journal of Physiology 256, C728–C735. Nikinmaa, M., Tiihonen, K., Paajaste, M., 1990. Adrenergic control of red cell pH in salmonid fish: roles of the sodium/proton exchange, Jacobs-Stewart cycle and membrane potential. Journal of Experimental Biology 154, 257–271. Perry, S.F., Reid, S.G., Salama, A., 1996. The effects of repeated physical stress on the adrenergic response of the rainbow trout red blood cell. Journal of Experimental Biology 199, 549–562. Perry, S.F., Thomas, S., 1993. Rapid respiratory changes in trout red blood cells during Na+ /H+ exchange activation. Journal of Experimental Biology 180, 27–37. Rummer, J.L., Brauner, C.J., 2011. Plasma-accessible carbonic anhydrase at the tissue of a teleost fish may greatly enhance oxygen delivery: in vitro evidence in rainbow trout, Oncorhynchus mykiss. Journal of Experimental Biology 214, 2319–2328.