ABSTRACT
The response of cannulated rainbow trout (Oncorhynchus mykiss) to acute hypoxia was studied in fish acclimated to two temperatures (5 and 15 °C). Blood/water respiratory variables and plasma catecholamine levels were measured before and 15 min after exposure to hypoxic water varying between 4.0 and 10.7 kPa (30–80 mmHg) oxygen partial pressure . Arterial blood and oxygen content fell during hypoxia in a similar manner at both temperatures, although the changes in were often more pronounced in the fish acclimated to 15 °C. Regardless of acclimation temperature, plasma catecholamine levels were consistently elevated at values below 8.0 kPa (60 mmHg); the largest increases in plasma catecholamine levels occurred below =5.3 kPa (40 mmHg). Adrenaline was the predominant catecholamine released into the circulation. Adrenaline was released at values of 8.0 kPa or below, whereas noradrenaline was released at values of 6.7 kPa or below.
The construction of in vivo oxygen dissociation curves demonstrated an obvious effect of acclimation temperature on haemoglobin (Hb) oxygen-affinity; the P50 values at 15 °C and 5 °C were 3.6 kPa (26.7 mmHg) and 1.9 kPa (14.0 mmHg), respectively. At 15 °C, catecholamines were released into the circulation abruptly at a threshold of 4.6 kPa (34.5 mmHg) while at 5 °C the catecholamine release threshold was lowered to 3.3 kPa (24.5 mmHg). The difference in the catecholamine release thresholds was roughly equivalent to the difference in the P50 values at the two distinct temperatures. Catecholamine release thresholds, calculated on the basis of arterial blood oxygen-saturation (expressed as /[Hb]), were similar at both temperatures and were approximately equal to 53–55 % Hb O2-saturation.
The results support the contention that the lowering of blood oxygen content/saturation rather than per se is the proximate stimulus/signal causing catecholamine release in rainbow trout during acute hypoxia.
Introduction
In response to severe environmental hypoxia, fish release the catecholamines adrenaline and noradrenaline into the circulation (Butler et al. 1978; Boutilier et al. 1988; Metcalfe and Butler, 1989; Ristori and Laurent, 1989; Aota et al. 1990; Fievet et al. 1990; Kinkead et al. 1991; Perry et al. 1991; Kakuta and Murachi, 1992; Perry and Reid, 1992; Thomas et al. 1992). The elevation of plasma catecholamine levels during hypoxia initiates a series of compensatory physiological processes directed towards the enhancement of branchial oxygen transfer and blood oxygen transport (see reviews by Perry and Wood, 1989; Randall, 1990; Thomas and Perry, 1992). These include (i) an enhancement of gill O2-diffusing capacity (Pettersson, 1983; Perry et al. 1985), (ii) an increase in blood oxygen-carrying capacity owing to the release of sequestered red blood cells from the spleen (Nilsson and Grove, 1974; Perry and Kinkead, 1989), and (iii) a rise in haemoglobin oxygen-binding affinity/capacity owing to activation of red blood cell Na+/H+ exchange (see reviews by Jensen, 1991; Nikinmaa, 1992; Thomas and Perry, 1992). In addition, elevated levels of circulating catecholamines may help to sustain the hyperventilatory response during severe hypoxia [Aota et al. 1990; Randall and Taylor, 1991 (see also, review by Perry et al. (1992) for an opposing view on the role of catecholamines in the control of breathing in fish].
In teleosts, including the rainbow trout, the predominant source of circulating catecholamines is the chromaffin tissue contained within the walls of the posterior cardinal vein at the level of the anterior (head) kidney (see review by Randall and Perry, 1992). Several physiological stimuli appear to trigger the mobilization of catecholamines from the chromaffin tissue. These include release of the neurotransmitter acetylcholine from preganglionic cholinergic fibres of the sympathetic nervous system (Nilsson et al. 1976; Perry et al. 1991) and direct local effects of altered blood chemistry (Opdyke et al. 1983; Hathaway et al. 1989; Perry et al. 1991). During hypoxia in Atlantic cod (Gadus morhua) and presumably rainbow trout, catecholamines are released in response to both sympathetic (neural) stimulation and the local effects of blood hypoxaemia on the chromaffin cells (Perry et al. 1991). Owing to the obligate relationship between and blood oxygen content, it has proved difficult to distinguish between the effects of these two variables on promoting catecholamine release during hypoxia. Consequently, there is considerable uncertainty concerning the relative contributions of changes in and O2 content in both the neural and the local control of catecholamine release (Fievet et al. 1990; Perry and Reid, 1992). Recently, it was suggested that the lowering of blood oxygen content is the proximate signal causing catecholamine release during hypoxia (Perry and Reid, 1992). This suggestion was based on uniform thresholds for catecholamine release in rainbow trout or American eel (Anguilla rostrata) when these thresholds were expressed on the basis of arterial O2 content rather than . However, the study of Fievet et al. (1990) implies a specific effect of , despite the inherent difficulties in distinguishing the possible specific effects of O2 content versus O2 partial pressure.
In the present study, we have attempted to determine the precise nature of the stimulus causing catecholamine release during hypoxia. The experimental design was to manipulate the intrinsic properties of Hb O2-binding in trout and to assess the impact on the dynamics of catecholamine release during acute hypoxia. The Hb O2-binding affinity was altered by acclimating fish to either 5 or 15 °C. If catecholamines are indeed released into the circulation at a particular and variable threshold corresponding to a critical and uniform reduction in O2 content (Perry and Reid, 1992), then the predicted effect of acclimating fish to lower water temperature would be a lowering of the threshold at which catecholamines are mobilized, in agreement with the increase in Hb O2-affinity induced by low temperature.
Materials and methods
Experimental animals
Rainbow trout [Oncorhynchus mykiss (Walbaum)] weighing between 200 and 300 g (experimental N=116) were obtained from Linwood Acres Trout Farm (Campbellcroft, Ontario) and were transported in hyperoxic water to the University of Ottawa. Fish were maintained on a 12 h:12 h L:D photoperiod in large fibreglass aquaria supplied with flowing, aerated and dechlorinated City of Ottawa tap water ([Na+]=0.10 mmol l−1, [Cl−]=0.15 mmol l−1, [Ca2+]=0.35–0.40 mmol l−1, [K+]=0.03 mmol l−1, pH 7.7–8.0). Fish were maintained at 12 °C for approximately 2 weeks before being separated into two different temperature acclimation groups; one group was acclimated to 5 °C and the other to 15 °C. The temperature was varied by 0.5 °C per day until the desired final temperature was reached. Fish were maintained under the final acclimation conditions for at least 2 months before experimentation. Trout were fed daily to satiation using a diet of commercial trout pellets; food was withheld for 48 h prior to experimentation.
Animal preparation
Fish were anaesthetized in a 0.1 g l−1 solution (5 or 15 °C) of ethyl-m-aminobenzoate (MS 222; Sigma Chemical Company), which was adjusted to pH 7.5 with NaHCO3, and then placed onto an operating table to allow continuous retrograde irrigation of the gills with anaesthetic solution. To permit periodic blood sampling, an indwelling cannula was implanted into the dorsal aorta (Soivio et al. 1975) using flexible polyethylene tubing (Clay-Adams PE 50; internal diameter 0.580 mm, outer diameter 0.965 mm). Trout were revived on the operating table by irrigation of the gills with aerated water, then transferred to individual opaque acrylic experimental chambers (volume 3 l) supplied with aerated flowing water of appropriate temperature, where they were allowed to recover from the effects of anaesthesia and surgery for at least 48 h before experimentation commenced. Cannulae were flushed daily with freshwater teleost saline containing 50 i.u. ml−1 ammonium heparin (Sigma Chemical Company).
Experimental protocol
Individual groups of fish (N=5–8 in each group) were acutely exposed to levels of hypoxia ranging between 4.0 and 10.7 kPa (30–80 mmHg). Specifically, the levels of hypoxia utilized were 10.7, 9.3, 8.0, 6.7, 6.0, 5.3, 4.7 and 4.0 kPa (80, 70, 60, 50, 45, 40, 35 and 30 mmHg). The lower limits of 4.0 kPa (30 mmHg; 5 °C) and 4.7 kPa (35 mmHg; 15 °C) were selected on the basis of the results of preliminary experiments showing marked mortality at more severe levels of hypoxia (especially at the higher temperature). In the present study, mortalities were observed at 4.7 kPa (1 fish) and 4.0 kPa (2 fish) in the fish acclimated at 15 °C and 5 °C, respectively; the results from these animals have not been incorporated. The upper limit of 10.7 kPa (80 mmHg) was selected on the basis of the preliminary results showing a lack of catecholamine mobilization at this level of hypoxia and the level immediately preceding it (9.3 kPa/70 mmHg).
Hypoxia was initiated by first stopping the air and water flow to the experimental chamber and then quickly re-establishing water flow at the same rate using hypoxic water exiting a water/gas equilibration column previously set to the target . The was adjusted by manipulating the rates of nitrogen (Air Products Inc.) and water flow through the column. The water flow rate into the experimental chamber was always in excess of 5 l min−1 and was sufficient to achieve the desired degree of hypoxia in the chamber within 5 min. Fish were returned to normoxic conditions by re-establishing normal (normoxic) water flow and aeration.
The inflowing water and the water within the experimental chamber were continuously monitored for . Usually, the of the inflow and chamber did not vary by more than 0.2–0.3 kPa and the of the experimental chamber, rather than the column, was used in calculating the mean of the various hypoxic groups.
Blood samples (0.6 ml) were withdrawn from the dorsal aortic cannula pre-hypoxia, at 5 and 15 min after reaching the desired level of hypoxia, and 15 min after return to normoxic conditions. The arterial blood was analyzed immediately after sampling to determine , oxygen content , whole-blood pH (pHa) and haemoglobin concentration ([Hb]). The remaining blood was centrifuged and the plasma (200–250 µl) stored at -80 °C, for no longer than 2 weeks, prior to determination of catecholamine levels. The red cell pellet was resuspended in teleost saline and reinjected into the dorsal aorta.
Analytical procedures
Whole-blood pH was determined with a microcapillary pH electrode (Radiometer G299A). Blood or water was measured using electrodes (Radiometer E5046) housed in thermostatted cuvettes (ambient water temperature 5 or 15 °C). from the equilibration column or the experimental chamber was monitored continuously by allowing water to flow by siphon through the measuring chambers of the electrodes. The pH and electrodes were maintained at the appropriate acclimation water temperature and utilized in conjunction with Radiometer PHM-71 acid–base analyzers and BMS3 Mk2 blood micro-systems. The electrodes were calibrated with water equilibrated to known values using air-saturated water. The pH electrode was calibrated using precision buffers (Radiometer). was measured on 20 µl samples according to an established method (Tucker, 1967) using a Radiometer electrode in a sealed chamber maintained at 37 °C. [Hb] measurements were performed in duplicate on 20 µl blood samples using a commercial spectrophotometric haemoglobin assay kit (Sigma Chemical Company).
Plasma adrenaline and noradrenaline levels were determined on alumina-extracted samples using high performance liquid chromatography (HPLC) with electrochemical detection (Woodward, 1982). 3,4-Dihydroxybenzylalamine hydrobromide (DHBA) was used as an internal standard in all determinations.
In vivo oxygen dissociation curves
Oxygen dissociation curves were constructed using the measured values of , and [Hb] from samples withdrawn pre-hypoxia and at 5 and 15 min of hypoxia; the values obtained after the return to normoxia were not utilized because of the possibility of persistent alterations of Hb O2-affinity initiated by catecholamines. To adjust for differences in [Hb] and variable quantities of physically dissolved O2 in the plasma, the amount of haemoglobin-bound O2 per unit haemoglobin ([O2]/[Hb]), in mol O2 mol−1 Hb, was calculated using O2 solubility coefficients for trout plasma (Boutilier et al. 1984). Oxygen dissociation curves were constructed by plotting [O2]/[Hb] as a function of and fitting the data to a sigmoidal function using an iterative curve-fitting function in a commercial graphics software package (Sigmaplot 5.0; Jandel Scientific). Blood oxygen-affinity (P50; the at half-maximal Hb O2-saturation) and Hill coefficients were derived from the Hill plot.
Catecholamine release thresholds
The and [O2]/[Hb] thresholds for catecholamine release were calculated as described by Perry and Reid (1992). The technique is used to estimate the point at which plasma catecholamine levels rise significantly above the baseline values. First, the mean baseline catecholamine levels were calculated by incorporating all values above a critical , the criterion for which was that plasma catecholamine levels were stable for at least 10 mmHg below this value. Next, the highest individual value with a catecholamine level statistically higher than baseline (outside the 95 % confidence interval) was determined and the mean was calculated for all values below that value. The threshold was then calculated as that mean plus its 95 % confidence interval. [O2]/[Hb] thresholds were calculated in a similar fashion.
Statistical analysis
Where appropriate, data are presented as mean values ±1 standard error of the mean. The results have been statistically analyzed by analysis of variance followed by Fisher’s LSD test for multiple comparisons; 95 % was accepted as the level of difference.
Results
Blood acid–base/respiratory variables
The effects of exposure to acute graded hypoxia on arterial blood respiratory/acid–base variables at the two acclimation temperatures are shown in Tables 1 (5 °C) and 2 (15 °C). For clarity, only the pre-hypoxic and 15 min hypoxic data are presented as there were no significant differences in any of the measured variables between 5 and 15 min of hypoxia. After 15 min of recovery from hypoxia, all of the measured blood respiratory/acid–base variables had returned to pre-hypoxic levels except for whole-blood pH (pHa) at the two severest levels of hypoxia (4.7 and 5.3 kPa ) in the 15 °C acclimated fish (data not shown).
Exposure of fish to hypoxia caused reductions in , and O2 bound to haemoglobin ([O2]/[Hb]) that were roughly proportional to the severity of the hypoxia. The changes in and [O2]/[Hb] were more pronounced in the fish acclimated to 15 °C (compare Tables 1 and 2). Blood [Hb] was essentially unaltered during the hypoxia; the decreases in [Hb] that were occasionally observed (see Tables 1 and 2) were probably attributable to the blood sampling rather than being a consequence of hypoxia per se. Whole-blood pH remained constant at values above 5.3 kPa (40 mmHg) and 4.7 kPa (35 mmHg) in the 15 °C and 5 °C fish, respectively. Below these levels of , whole-blood pH declined significantly (Tables 1 and 2); the reduction in pHa was more pronounced in the 15 °C acclimated fish.
Whole-blood pH was consistently (although not always) higher in the 5 °C acclimated fish in accordance with the predicted inverse relationship between blood temperature and pHa (see Heisler, 1984). Occasionally, the pre-hypoxia values of , , [Hb] or [O2]/[Hb] were significantly different between the two acclimation groups, although no obvious pattern was evident.
Blood oxygen dissociation curves
In vivo O2 dissociation curves (Fig. 1) were constructed for the 5 and 15 °C acclimated fish using the blood respiratory data gathered during the acute hypoxia experiments (Tables 1 and 2); the recovery data were not used. In theory, the [O2]/[Hb] of fully saturated haemoglobin is 4 mol O2 mol−1 Hb. In the present study, the [O2]/[Hb] at maximal binding was 3.2 mol mol−1, indicating approximately 83 % saturation. This value is somewhat lower than those reported for trout haemoglobin in previous studies (95–100 %; Holeton and Randall, 1967; Eddy, 1976) and indicates a significant fraction of non-functional haemoglobin of unknown origin. The [O2]/[Hb] at maximal binding was identical at both temperatures, thereby allowing valid comparison between the two experimental groups. The apparent P50 values were determined assuming that maximal binding occurred at a [O2]/[Hb] value of 3.2 mol O2 mol−1 Hb. To permit valid comparisons with previous studies, however, the Hb O2-saturation catecholamine release thresholds were calculated assuming that 100 % saturation corresponded to 4 mol O2 mol−1 Hb.
At 5 °C, the in vivo P50 value was 1.9 kPa (14.0 mmHg) whereas at 15 °C the P50 value was increased to 3.6 kPa (26.7 mmHg). The Hill coefficients (nH) were 1.29 and 1.76 at 5 and 15 °C, respectively.
Plasma catecholamine levels
Regardless of acclimation temperature, plasma catecholamine levels were elevated at values less than or equal to 8.0 kPa (60 mmHg; Fig. 2). Adrenaline was the sole catecholamine released into the circulation at =8.0 kPa (60 mmHg). Below 8.0 kPa (60 mmHg), both adrenaline and noradrenaline levels were elevated, with adrenaline being the prevalent circulating catecholamine in most instances (Fig. 2). At values of 5.3 kPa (40 mmHg) and 4.7 kPa (35 mmHg), plasma adrenaline levels (Fig. 2A) were significantly greater in the 15 °C acclimated fish. At values of 6.7 kPa and below, plasma noradrenaline levels (Fig. 2B) were always higher (with the exception of 4.7 kPa) in the 15 °C fish.
Fig. 3 illustrates the relationships between and the total plasma catecholamine levels (adrenaline plus noradrenaline). At each acclimation temperature, plasma catecholamine concentrations remained remarkably constant over a wide range of but then increased abruptly when a critical threshold was reached. The thresholds were widely different at the two acclimation temperatures. At 5 °C, the calculated threshold for catecholamine release was 3.3 kPa (24 mmHg; Fig. 3A), whereas at 15 °C the corresponding threshold was 4.6 kPa (34.5 mmHg; Fig. 3B). The difference in these catecholamine release thresholds (1.3 kPa) was approximately equal to the difference in the in vivo P50 values (1.7 kPa) at the two distinct temperatures.
The relationships between [O2]/[Hb] and plasma catecholamine levels are shown in Fig. 4. The calculated catecholamine release thresholds were similar at each acclimation temperature, varying only between 2.2 (5 °C) and 2.1 (15 °C) mol O2 mol−1 Hb, corresponding to 53–55 % Hb O2-saturation.
Water–blood relationships
The relationships between and during normoxia and hypoxia are illustrated in Fig. 5. declined with decreasing in an essentially similar manner in both acclimation groups. During moderate hypoxia (>8.0 kPa), for any given value of , at 5 °C was generally higher than that at 15 °C. These differences were not apparent at the more severe levels of hypoxia.
Discussion
Effects of acclimation temperature on catecholamine release
Recently, Perry and Reid (1992) proposed a mechanism to explain the abrupt release of catecholamines into the circulation during exposure of teleost fish to environmental hypoxia. According to this theory, catecholamines are mobilized from the chromaffin tissue as the blood oxygen content is lowered to a critical catecholamine release threshold. Further, the theory predicts that the blood at which this threshold is reached will vary according to the affinity of Hb O2-binding. Thus, fish possessing blood of high Hb O2-affinity (low P50) would be expected to release catecholamines at considerably lower blood values than fish with low Hb O2-affinity (high P50). In each instance, however, catecholamine levels would rise at a uniform value of Hb O2-saturation corresponding to 45–60 % Hb O2-saturation. This model was based on data gathered from two species possessing widely different Hb O2-affinities, the American eel (Anguilla rostrata) and the rainbow trout (Oncorhynchus mykiss). As appreciated by the authors (Perry and Reid, 1992), it was conceivable that the differences in the catecholamine release profiles in eel and trout simply reflected intrinsic differences in release thresholds and that the uniformity of the O2 content release thresholds may have been coincidental. The present study was designed to test this model by using a single species (rainbow trout) acclimated to different temperatures as a means to modify Hb O2-affinity experimentally (see below).
The results of this study support the mechanism proposed by Perry and Reid (1992). At each acclimation temperature, catecholamines were released into the circulation as the blood oxygen status traversed a critical threshold corresponding to a uniform value of approximately 53–55 % Hb O2-saturation. This catecholamine release threshold was similar to the release threshold (45–60 % Hb O2-saturation) reported in a previous study (Perry and Reid, 1992). Owing to the decrease in P50 in the 5 °C acclimated fish, this threshold was reached at a lower than in the 15 °C acclimated fish. In other words, the difference in the thresholds at the two temperatures was essentially equal to the difference in the P50 values. The simplest explanation for these data is that the lowering of Hb O2-saturation (or a closely related variable such as blood O2 content) is the proximate signal causing the release of catecholamines rather than a lowering of blood per se. Regardless of the mechanism underlying the relationship between blood O2 content and catecholamine release, the obvious and important consequence of this relationship is that catecholamines are released into the circulation only upon marked impairment of blood O2 transport. Unlike the ventilatory and cardiovascular adjustments to hypoxia (see reviews by Perry et al. 1992; Fritsche and Nilsson, 1993), which often begin with only slight reductions in , the release of catecholamines is not initiated until is lowered to very low levels [usually below 8 kPa (60 mmHg)] (see review by Randall and Perry, 1992). Given that the predominant effect of elevated catecholamine levels is to enhance branchial O2 transfer and blood O2 transport, the physiological significance of their delayed release into the circulation after only a severe reduction of blood O2 content is evident. From a design viewpoint, it would be impractical to link changes in per se to catecholamine mobilization because is not a reliable indicator of blood O2 content. This reflects the non-linear relationship between and Hb O2-saturation as well as the interactive effects of numerous allosteric modifiers of Hb O2-binding, such as CO2, H+ and organic phosphates (Weber and Jensen, 1988; Jensen, 1991). Thus, the reliance of catecholamine release on a critical lowering of blood O2 content may have evolved so as to improve O2 delivery when it is compromised by a developing hypoxaemia.
A linkage between blood O2 content and Hb O2-saturation and plasma catecholamine levels has been suggested by the results of previous studies in addition to the aforementioned comparison of trout and eel (see above). First, anaemic fish release catecholamines into the circulation (Iwama et al. 1987; Perry et al. 1989) even under hyperoxic conditions (Perry et al. 1989). In anaemic fish, Hb O2-saturation is not lowered, which suggests that there is a specific role for the lowering of blood O2 content in causing release. Second, it was shown that the cause of catecholamine release during hypercapnic acidosis in trout is the associated hypoxaemia (owing to the Root effect) rather than the acidosis itself (Perry et al. 1989). Third, Fievet et al. (1990) reported that the threshold for catecholamine release in trout was substantially lowered after repeated episodes of acute hypoxia. Our own interpretation of these data is that Hb O2-affinity was raised after the initial episode of hypoxia [a result, at least in part, of catecholamine release (Nikinmaa, 1983)] and thus led to a lowering of the threshold. Although Fievet et al. (1990) alluded to a controlling role of blood O2 tension in catecholamine release, the underlying reasons for this assumption are unclear. Indeed, to our knowledge, there are no data in the literature supporting a specific/direct role of blood in catecholamine release in fish.
Catecholamine release is not the only physiological process that appears to be associated with blood O2 content levels in fish. Numerous studies have implicated the reduction of blood O2 content as a key variable controlling ventilation (see reviews by Randall, 1982; Shelton et al. 1986), whereas there is considerably more uncertainty as to the specific role of blood .
Effects of acclimation temperature on Hb O2-affinity
The advantages and disadvantages of determining P50 values from in vivo O2 dissociation curves have been discussed in detail previously (Perry and Reid, 1992). Briefly, the merit of the in vivo O2 dissociation curve is that it yields functional P50 values that encompass the net effects of potential curve modifiers, including changes in blood acid–base status and elevated catecholamine levels. The P50 values reported in the present study are similar to those previously reported in vitro (Milligan and Wood, 1987; Soivio et al. 1980; Vorger, 1986) or in vivo (Tetens and Christensen, 1987) values. The reasons for the significant amount of non-functional haemoglobin (Hb O2-saturation was only 83 % at maximal binding) in the fish used in the present are unknown, but it presumably reflects an unusually large fraction of methaemoglobin.
The results of the present study demonstrated an increase in Hb O2-affinity as temperature was lowered from 15 to 5 °C [P50 decreased from 3.6 kPa (26.7 mmHg) to 1.9 kPa (14.0 mmHg)]. Although the inverse relationship between temperature and Hb O2-affinity is well documented following acute changes in blood temperature (e.g. Vorger, 1986; see also reviews by Weber and Jensen, 1988; Jensen, 1991), considerably less is known of the chronic effects of acclimation to different temperatures. It is generally believed, however, that chronic temperature changes elicit smaller effects on Hb O2-affinity than do acute changes (see review by Wood, 1980). For example, Weber et al. (1976) demonstrated that acclimation of rainbow trout to temperatures varying between 5 and 22 °C for as long as 4 months was without effect on the Hb O2-affinity of whole blood (as assessed in vitro). Clearly, the results of the present study showing an effect of chronic acclimation on Hb O2-affinity are in marked contrast to the study of Weber et al. (1976). Further, the changes in P50 were essentially similar to the changes that accompany acute temperature changes in vitro (e.g. Vorger, 1986). Red blood cell intracellular pH (red blood cell pHi) or levels of red blood cell organic phosphates were not measured in the present study, making it difficult to compare these results with those of previous studies. Although not always statistically significant, whole-blood pH (pHa) was generally elevated in the fish acclimated to 5 °C (compare Tables 1 and 2) in accordance with the usual inverse relationship between pHa and temperature (Heisler, 1984). Thus, it is also likely that red blood cell pHi was elevated in the 5 °C acclimated fish given that hydrogen ions are passively distributed across the red blood cell membrane (see Nikinmaa, 1992).
Other potential causes/correlates of catecholamine release
Exposure of fish to the more severe levels of hypoxia (4.0–5.3 kPa; 30–40 mmHg) elicited marked acidosis of the blood (Tables 1 and 2). It is tempting to speculate, therefore, that the blood acidosis is a cause (or one of the causes) of the greatly elevated plasma catecholamine levels during severe hypoxia. Although several studies have reported significant correlations (Tang and Boutilier, 1988; Perry and Reid, 1992) or relationships (Boutilier et al. 1986; this study) between the extent of blood acidosis and circulating catecholamine levels, it is difficult to ascribe a direct role to blood acidosis in causing catecholamine release because of the hypoxaemia that normally accompanies acidosis in teleost fish. Furthermore, it has been demonstrated (Perry et al. 1989; Aota et al. 1990) that acidosis itself does not initiate catecholamine release in trout unless it is associated with blood hypoxaemia. An additional problem in assigning a role for blood pH changes in the control of catecholamine release is the inherent difficulty in separating the cause of release from the consequences of release. Catecholamines, when released into the blood, cause acidification of the plasma as a result of stimulation of red blood cell Na+/H+ exchange (see reviews by Nikinmaa, 1992; Thomas and Perry, 1992) and thus high levels of catecholamines would normally accompany blood acidification even if acidosis itself were not a cause of catecholamine release. Finally, at moderate levels of hypoxia (>5.3 kPa; 40 mmHg), blood acidosis is clearly not a factor in triggering release because blood pH is either unaltered at such times (this study) or even elevated owing to hyperventilation. Variations in the versus relationship also cannot explain the differing patterns of release at the different temperatures, as these relationships were essentially indistinguishable at the levels at which catecholamines are released (Fig. 5).
The results of this and other studies (Perry et al. 1989; Thomas et al. 1992; Perry and Reid, 1992) provide compelling evidence that lowering of blood Hb O2-saturation and/or blood oxygen content is the factor that signals catecholamine release during hypoxia. However, owing to the nature of this study in which acclimation to different temperatures was used as a tool to modify Hb O2-affinity, we cannot exclude the involvement of other temperature-dependent factors such as thermal modulation of a receptor. The important theme that is emerging from these studies is that catecholamines are released into the circulation of fish only when a critical threshold of blood O2 content is reached. Further experiments are required to elucidate fully the mechanisms linking depression of blood O2 content to the mobilization of catecholamines from chromaffin tissue.
ACKNOWLEDGEMENTS
This study was supported by NSERC of Canada operating and equipment grants to S.F.P. S.G.R was the recipient of a NSERC Postgraduate Scholarship.