- Email: [email protected]
Acute exposure to graded levels of hypoxia in rainbow trout (Salmo gairdneri): metabolic and respiratory adaptations
Respiration Physiology (1988) 71, 69-82
69
Elsevier
RSP 01358
Acute exposure to graded levels of hypoxia in rainbow trout (Salmo gairdneri): metabolic and respiratory adaptations R.G. Boutilier l, G. Dobson 2, U. Hoeger 3 and D.J. RandalP 1Department of Biology, Dalhoasie University, HalO~ax,Nova Scotia, Canada B3H 4Jl, 2NationalInstitutes of Health, Laboratory of Metabolism, NIA,L4, Rm 55, 12501 WashingtonAve., Rockville MD 20852, U.S.A., 3FakultiitJ~r Biologie, Universit~tKonstanz, 7750 Konstunz, F.R.G., and 4Department of Zoology, University of British Columbia, Vancouver,British Columbia, Canada V6T 2,49 (Accepted for publication 4 August 1987) Abstract. We have studied the mechanisms of acute hypoxia tolerance in rainbow trout (Salmo gairdneri). Fish held at 9 °C were exposed to various levels of bypoxia for 24 h. At an environmental Po2 of 30 Tort, the fish showed an initial plasma acidosis probably of metabolic origin which was subsequently offset such that blood pH returned to normal within about 4 h. Over this time period, red cell pH was maintained constant. Comparing the effects of different levels of hypoxia following 24 h exposure, oxygen consumption of the animal remained unchanged over a broad range of inspired oxygen tensions but declined by over 30 % of normoxic values at inspired water Po2 levels of 80 Tort. This appeared to be a true metabolic depression because signs of increased anaerobic metabolism did not occur until there was a further reduction in water oxygen levels. Rainbow trout appear to be able to maintain a relatively high energy status in their white muscle during 24 h exposure to severe hypoxia (water Po2 = 30 Torr). As the level ofhypoxia was intensified, there was a reduction in the oxygen gradient across the gills, probably facilitated in part by the release of catecholamines into the blood. The erythrocytic ATP: Hb4 molar ratios declined with increasing hypoxic stress as did the pH gradient between the erythrocyte and plasma. The overall effect was no change in Hb O2-affinity alter 24 h exposure to severe hypoxia" Abstract. Anaerobiosis; Hypoxia; Muscle metabolism; Rainbow trout; Red cell pH
Exposure of fish to reduced levels of environmental oxygen initiates a whole host of adaptational processes, most of which are aimed at increasing the availability of oxygen at the level of the tissue. Many of these processes act immediately, such as the changes in ventilation and perfusion of the gill (Randall, 1982), while others including changes in erythrocytic adenylates, muscle mitochondrial densities and enzyme activities (Weber, 1982; Hochachka and Somero, 1984) exert their effects over progressively longer time periods. Another potential strategy of water breathing fish is the lowering of metabolic
Correspondence address: Dr. R.G. Boutilier, Department of Biology, Dalhousie University, Halifax, Nova Scotia, Canada B3H 4Jl. 0034-5687/88/$03.50 © 1988 Elsevier Science Publishers B.V. (Biomedical Division)
70
R.G. BOUTILIERet al.
rate which in effect minimizes the impact of the hypoxic insult by conforming to the environmental Po2. Exposure of rainbow trout to water Po2 tensions of 30 Torr leads to a marked extracellular acidosis within the In'st 20 rain and subsequently to an elevation of blood pH to values approaching (Tetens and Lykkeboe, 1985) or exceeding (Thomas and Hughes, 1982) those of the normoxic controls. These initial responses are complex, since the increased gill irrigation leads to a minor respiratory alkalosis whereas the decreased availability of oxygen causes a lactacidemia (Thomas and Hughes, 1982). The time course and severity of the hypoxia greatly influences the degree of the acidosis and therefore H b - O z transport. Recent studies on rainbow trout have shown that red cell pH is regulated at a constant level during and following periods of extracellular acidosis brought on either by anaerobic exercise (Primmett etal., 1986) or direct acid infusion (Boutilier etal., 1986). In both instances catecholamines are released into the blood and would appear, through fl-adrenergic mechanisms at the erythrocytic membrane, to be responsible for turning on cellular ion exchange processes which ultimately lead to the regulation of the pH environment of the haemoglobin (Nikinmaa, 1983, 1986). As an oxygen safeguarding system, the ability to regulate the pH of the red cell during periods of extraceHular acidosis would be of vital importance in facilitating oxygen loading at the gills and in maintaining oxygen delivery to the tissues. Decreased oxygen availability per se, with or without a corresponding acidosis, represents a direct challenge to the oxygen transport and delivery system of the animal. Studies on the marine elasmobranch, Scyliorhinus canicula, have shown that plasma catecholamine concentrations increase in response to hypoxia (Butler et al., 1978). Moreover, it has been shown that hypoxia in rainbow trout causes an increase in red blood cell volume which can be produced in vitro by addition of adrenaline to blood (Nikinmaa, 1986). The present study attempts to clarify the mechanisms of acute hypoxia tolerance in rainbow trout (Salmo gairdneri). Not only will a number of fundamental organismic strategies be explored but emphasis will be placed on integrating these with adrenergic regulation of H b - O 2 carrying capacity and/or affinity as well as muscle metabolism.
Materials and methods
Animals and preparation. Rainbow trout (S. gairdneri) weighing 220-460 g were obtained from the Sun Valley Trout Farm, Mission, B.C. and were kept in outdoor tanks supplied with dechlorinated tap water (5-12 ° C). Feeding was suspended and the animals were transferred to a 9 °C holding tank at least 48 h prior to surgery. The dorsal aorta was chronically cannulated with P.E. 50 tubing after the animals had been anaesthetized in a 0.05Y/o solution of tricainemethanesulfonate (MS-222, Sigma) buffered to pH 7.5 with NaHCO 3. During placement of the cannula, the gills were irrigated with a lighter dose (0.013 ~o MS-222) of the anaesthetic solution which was oxygenated to ambient Po2 levels. Following the operation, the fish were transferred
HYPOXIAIN RAINBOWTROUT
71
to darkened Perspex boxes supplied with flowing aerated water at 9 °C where they remained for at least 72 h before being used for experiments. The cannulae were flushed daily with 0.1 ml Cortland saline containing 10 i.u./ml of heparin (Sigma).
Exposure offish to hypoxia. All experiments were carried out on fish kept in darkened Perspex chambers. Each of the chambers could be individually supplied with flowing water of known oxygen partial pressures produced by oxygen-stripping columns similar to those described by Fry (1951). Thirty fish were exposed to normocapnic conditions of normoxia or various degrees of hypoxia for 24 h before sampling took place. In each instance, a 2 ml blood sample was then drawn from the dorsal aortic catheter and taken into a gas tight syringe for analyses of blood gases, pH and metabolites (see below). Fish were then rapidly transferred from the black holding chambers and decapitated, after which samples of white epaxial muscle were immediately excised from a site posterior to the dorsal fin. Samples were immediately freeze-clamped in aluminum blocks (pre-cooled to - 196 °C in liquid nitrogen) and stored at - 7 0 °C until required. The time between capture of fish and freeze clamping muscle was measured to be within 6-10 see. In a second series of experiments, six rainbow trout were exposed to three levels of environmental hypoxia for determination of oxygen consumption by measuring the O z partial pressure difference and rates of water flow through sealed Perspex chambers. Measurements were taken while the animals were resting in normoxic conditions and after 24 h exposure to environmental hypoxia. Accurate measurement of the Po2 difference became increasingly difficult as the inspired Po2 levels were decreased below 80 Torr, and so were discarded. Reducing the water flow in the chamber to achieve a larger inspired-expired Po2 difference has the effect of increasing water Pco~ and thereby exacerbating steady-state normocapnic measures of Oz consumption. In a final series of experiments six rainbow trout were cannulated as before for dorsal aortic blood sampling. Blood samples were taken before and at specified periods following acute exposure to inspired Po2 levels of 30 Torr. The blood samples were used to measure plasma and red cell pH as outlined below. In vitro determination of Hb O2-affinity. Animals with dorsal aortic catheters were exposed for 24 h to normoxic (Plo2 155 Torr) or severely hypoxic (PIo2 35 Torr) environments, after which blood samples were taken and pooled to the required volumes in two glass tonometer flasks. The blood samples were equilibrated at 10 °C with humidified and thermostatted gas mixtures containing 0.2~o CO2, the balance of which was either air or nitrogen (WOsthoff gas mixing pumps, Bochum, F.R.G.). Oxygen dissociation curves were prepared using the mixing method as detailed by Scheid and Meyer (1978). Analyticalprocedures. Measurements of whole blood pH and red cell lysate were made using Radiometer microdectrode units coupled with PHM 84 pH meters (Boutilier et al., 1985). The freeze-thaw method was used to produce the red cell lysate whose
72
R.G. BOUTILIERet al.
pH is not measurably different than that determined by application of the DMO method in rainbow trout erythrocytes (Milligan and Wood, 1985). The pH electrodes were calibrated before each measurement with Radiometer precision buffer solutions S 1500 and S 1510. Oxygen partial pressures in blood and water were measured using Radiometer electrodes and 02 meters. Oxygen contents of whole blood and CO2 contents of true plasma were measured using a Lex-O2-Con (Lexington Instntments, MA) and a Carle Series 100 gas chromatograph (see Boutilier et al., 1985) respectively. Hemoglobin concentration was determined by the speetrophotometric method of Van Kampen and Zijlstra (1965). Plasma adrenaline and noradrenaline concentrations were measured by. high pressure liquid chromatography (cf.Primmett et al., 1986) using a Bioanalytical Systems HPLC with electrochemical detection. Whole blood and plasma extracts. Samples of either whole blood or true plasma (150 #1) were immediately deproteinized with 300 #1 of ice-cold 10~o trichloroacetic acid (TCA). Following a 15-30 min extraction time on ice, the samples were centrifuged for 30-45 see at 11000 rpm. A known volume of clear supernatant was removed and then neutralized with a pre-determined volume of saturated Tris Base [tris (hydroxymethyl)-aminomethane]. The neutralized extracts were immediately frozen in liquid nitrogen and stored at - 70 °C until required for metabolite analysis.
White epaxial muscle was powdered under fiquid nitrogen using a pre-cooled mortar and pestle. Tendons and fragments of connective tissue were dissected free and discarded. About 500 nag of powdered tissue was transferred to a pre-cooled, pre-weighed vial containing 1.0 ml of ice-cold 0.6 N perchloric acid (PCA) and then reweighed. A further 1.0 ml PCA was added and the powder homogenized for 15 see at 0°C using an Ultra-Turrax homogenizer. The homogenization procedure was repeated and the side of the vial washed down with a further 0.5 ml PCA. The suspension was stirred under low speed and 100 #1 was removed in duplicate for glycogen determination. The remaining homogenate was then centrifuged for 2 min at 13000 rpm and 4 °C after which approximately 2 ml of supernatant was immediately neutralized (pH 7.0) with Tris Base, frozen in liquid nitrogen and kept at - 70 °C until required. The homogenization and neutralization procedure described takes about 6 min to complete. In a parallel set of validation studies, the highly acid-labile phosphoereatine underwent not more than 5 ~o hydrolysis with this technique. Muscle homogenization, extraction and neutralization.
Plasma and muscle lactate concentrations, and muscle glycogen concentrations were measured in a Pye-Unicam SP8-100 UV-VIS spectrophotometer at 340 nm using the routine NADH- or NAD-coupled enzymatic procedures described in Bergmeyer (1983). Each assay was validated with appropriate standards. Metabolite determinations.
HYPOXIA IN RAINBOW TROUT
73
Determination ofnucleotides andphosphagens. Concentrations of the nucleotides ATP, ADP, AMP and IMP and the phosphagens, phosphocreatine (PCr) and creatine (Cr) were analytically determined by high pressure liquid chromatography (HPLC). The procedure was carded out on a Spectra Physics 8000 B HPLC coupled to a Kratos Spectroflow 773 absorbance detector set at 254 nm. The nucleotides were passed through a 250 x 4.6 n'an AX-300 anion exchange column (Brownlee Laboratories) at a flow rate of 2 ml/min and eluted in 30 rain using a gradient of potassium dihydrogen phosphate. Buffer A consisted of 50 mM KHzPO 4 (pH 2.31 at 55 °C) and buffer B of 600 mM KHzPO4 (pH 2.63 at 55 °C). The elution conditions were as follows: 100~ A for 4 min, 0 to 50% B from 4 to 5 rain, 50 to 55~ B from 5 to 10 rain, 55 to 100% B from 10 to 11 rain, 100~ B from 11 to 30 min. PCR, Cr and AMP were eluted at the same flow rate of 2 ml/min using an isocratic 50 mM KHzPO 4 buffer at pH 3.1, and detected at 210 nm (10 #1 of neutralized sample was required for each set of determinations). Following every set of six chromatographic runs, the column was washed with 600 mM KHzPO 4 and a single 10 #1 injection of known standards was passed through the system to check retention times of each metabolite. Internal standards were used to validate the procedure for muscle analysis. When stated, the significance (P < 0.05) of differences between means was assessed with Student's t-test.
Results
Data for metabolic and respiratory changes in the blood compartments of trout exposed to several stages of hypoxia for 24 h are shown in table 1. Compared with normoxic animals, the blood pH was unchanged after 24 h of hypoxic exposure even though transient changes occur within that time period (fig. 1). After 24 h exposure to deep hypoxia, plasma pH appears to be regulated at normoxic levels by offsetting the pronounced lactacidemia with a decrease in arterial Pco~, presumably through an increase in gill ventilation (tables 1 and 2). Changes in plasma and erythrocytic pH are shown in fig. 1 as a function of time after fish were acutely exposed to an inspired Po~ (PIo~) of 30 Torr. A plasma acidosis was observed after 30 rain exposure, recovery from which took between 2-4 h. Despite the marked reduction of plasma pH, that of the erythrocyte was held at a constant level over the first 4 h of hypoxia and thereafter increased. At the end of 24 h, the erythrocytic pH was within 0.227 units of plasma pH, 0.3 units closer than the normoxic conditions (fig. 1). These latter results are similar to those observed after 24 h exposure to various levels of hypoxia (table 1). Arterial pH after 24 h exposure to all levels of hypoxia was unchanged from that recorded in normoxia. Despite the constancy of arterial pH, the difference in pH between plasma and erythrocytes (ApH) was reduced under extreme hypoxia (table 1). Relative to normoxic fish, plasma adrenaline and noradrenaline concentrations were elevated during hypoxia, with the highest levels being recorded in the animals exposed to the lowest inspired 02 tensions (table 1).
TABLE
1
0.50 f 0.05
[lactate-],
22.5
1.1
10.5
8.5
+ 1.4
f 0.2
+- 0.4
+ 0.4
k 2.5
[lactate-], mmol.kgg’ cell water
6.53 f 0.58
1.22 f 0.09
1.13 * 0.13
ATP: Hb, molar ratio
ml- 1 packed
[ATP], mol. RBC
RBC
+ 0.04
0.54
PH, - PHi (PHI
[GTP], mol. ml - ’ packed
7.517 * 0.037
0.338 f 0.058
Red blood cell: PH
0.5
f
+
0.082
1.27 f
1.24 f
0.14
0.83
0.07
+ 0.09
0.77 *
0.83
0.440 f
1.491 +_ 0.032
0.01
0.6
0.1
+_ 0.3
+
f 10.0
0.35 j:
22.7
8.0
10.8
9.2
79.9
2.08 * 0.41
0.22 +-0.13
116.0
1.78 f 0.17
0.34 f 0.02
blood
f 0.7
1.23 + 0.21
8.5
f 0.2
7.926 f 0.068 2.4
0.42 + 0.09
f 0.7
MCHC, g.rn-t
Hct, %
[Hb], g. 100 ml-’
Maximum 0, concentration vols y0
O2 concentration, vols %
Arterial blood: Po, (To=)
’ [noradrenaline], run01. 1- 1
[adrenaline], mnol 9 1-
mmol~l-’
8.6
[HCO,-],mmol.l-’
+ 0.5
1.855 + 0.084
2.9
Pco, (Torr)
:
Arterial plasma PH
120 (n = 6)
155
(Torr)
(n = 6)
Inspired water P,
f 0.3
+ 0.2
* 0.01
f 2.7
f 0.4
+ 1.2
f 0.9
f 5.3
1.21 f 0.05
6.45 + 0.16
0.77 f 0.03
1.11 + 0.09
0.380 f 0.048
7.483 +_0.030
0.35
23.5
8.2
11.0
1.4
51.1
2.11 * 0.22
1.66 f 0.28
0.72 f 0.17
6.1
2.1
7.883 f 0.044
(n = 5)
90
Temperature = 9°C. n = number of animals.
+ 0.8
+ 0.2
+ 2.1
f 0.02
+ 2.6
+ 0.2
+ 0.4
+ 0.4
5.23 + 0.34 1.14 + 0.12
0.64 f 0.07
1.24 * 0.08
7.510 + 0.027 0.240 + 0.043
0.32
30.3
9.6
12.5
5.2
25.9
0.95 f 0.18 7.10 + 2.68 4.40 + 2.80
6.1
2.6
f 0.3 f 0.8
f 2.1
+ 0.4
* 1.0
f 0.7
+- 0.6
0.92 & 0.07
4.21 + 0.59
0.62 f 0.14
3.53 * 0.45
0.169 + 0.053
7.662 +_0.037
0.29 + 0.03
31.2
9.2
12.1
3.6
13.5
13.79 +_4.11
8.81 k 3.56
6.14 k 1.03
5.5
1.9
7.840 f 0.098
(n = 4)
7.810 * 0.051
30
50 (n = 5)
Means + 1 SEM of measured variables from rainbow trout (S. guirdneri) after 24 h exposure to different levels of oxygen partial pressure (Po,) of inspired water.
HYPOXIA IN RAINBOW TROUT
75
8.1pH
7.9 ~7.7 7..~ 72
0
~
~
~
~
,,/
~'4
Time ( h )
Fig. 1. Changesin medal bloodplasmapH andred bloodcellpH afterexposureto a stepchangeininspired water Po2from150Torrto 30Tort (step changeindicatedby verticallineat time0). C = normoxiecontrol values. Temperature= 10.5°C. The effect of acute exposure to graded levels of hypoxia on the concentration of guanosine 5-triphosphate (GTP) and adenosine 5-triphosphate (ATP) in the red cells of trout is shown in table 1. The concentration of GTP decreased sjt,nificantly by about 30% in response to mild hypoxia (a decrease in PIo2 from 152 to 122 Torr), but stabilized at this level during the more severe hypoxic states. The red cell ATP concentrations followed a different pattern of change with no significant decrease occurring until the environmental Po2 fell to 52 Torr. A further percentage reduction in ATP occurred at the lowest inspired 0 2 tension of 32 Torr (table 1). In general, the reduction in ApH was correlated with a decrease in RBC ATP levels (fig. 2). Haemoglobin concentrations in whole blood increased with haematocrit and decreasing environmental Poz. The increased haematocrit was due to both haemoconcentration and red cell swelling, the latter leading to a corresponding decrease in the mean cellular haemoglobin concentration (MCHC; table 1). The decrease in RBC ATP was somewhat greater than the reduction in RBC haemoglobin concentration due to cellular swelling, the net result being a slight decrease in the ATP : Hb ratio during severe hypoxia (table 1). The effect of decreasing environmental Po2 on the oxygen consumption of rainbow trout is shown in fig. 3. It is clear that after 24 h of mild hypoxia (PIo~ = 120 Torr), oxygen consumption was unchanged, but declined significantly by over 30% at the lower inspired Po2 of 80 Torr. The concentrations of a selected number of metabolites in white muscle of rainbow trout are presented in table 2. Phosphocreatine (PCr), glycogen, lactate, ATP, adenosine 5-monophosphate (AMP) and inosine 5-monophosphate (IMP) underwent no appreciable change in concentration after 24 h of varying degrees of hypoxia down to an inspired Po: of 52 Torr. In contrast, at an inspired Po2 of 32 Torr, glycogen decreased by 50% (from 25.54 to 12.95 pmol/g wet wt), PCr decreased by about 40% (from 16.58 to 19.80 #mol/g wet wt), and lactate increased 2.5 fold (3.33 to 8.33 #mol/g wet wt) compared to the normoxic state. No significant changes occurred in the concentrations of either ATP, AMP, IMP or ammonium ion (NH2-) throughout the various stages of Q hypoxic exposure (table 2).
TABLE 2
19.1
26.5
38.5
41.7
37.0
16.58 (_+ 1.09) 17.74 (_+0.70) 18.41 (_+0.73) 17.90 (_+2.02) 10.80 ( _+ 1.12)
116.0 (_+2.5) 79.9 (_+ 10.0) 51.1 (-+5.3) 25.9 (_+2.7) 13.5 ( _+0.6)
153.0 (_+0.9) 121.6 (_+ 1.3) 89.6 (_+4.2) 52.4 (_+ 1.9) 32.6 ( _+2.8)
39.00 (_+3.00) 39.29 (_+ 1.07) 37.35 (_+1.44) 35.30 (_+2.38) 43.98 ( _+2.25)
Cr
PCr
Arterial Po2
Inspired Po2
APo2
#mol" g - ~ wet weight
Torr
1.18 ( _+0.21)
-
-
1.06 (_+0.15) -
NH4 +
0.86 (_+0.09) 0.95 (_+0.47) 0.82 (_+0.20) 0.6 (_+0.16) 0.77 ( _+0.14)
IMP
0.77 (_+0.03) 0.81 (_+0.05) 0.74 (_+0.22) 0.70 (_+0.04) 0.76 ( _+0.26)
ADP
6.25 (_+0.16) 6.57 (_+0.12) 6.79 (_+0.14) 6.73 (_+0.20) 6.23 ( _+0.07)
ATP
0.047 (_+0.006) 0.064 (_+0.013) 0.065 (_+0.030) 0.029 (_+0.004) 0.043 ( _+0.009)
AMP
3.33 (_+0.60) 4.70 (+_ 1.23) 2.95 (_+0.24) 4.40 (_+ 0.18) 8.33 ( _+2.09)
Lactate
25.54 (_+0.71) 27.14 (_+2.22) 23.48 (_+0.55) 25.09 (_+ 1.14) 12.95 ( _+ 1.44)
Glycogen
4
4
5
6
5
N
Concentration of white muscle metabolites after 24 h ofnormoxia and various levels of hypoxia in rainbow trout (S. gairdneri) at 9°C. Values are means _+ 1 SEM N = number of animals.
rn
¢¢ O
O~
77
HYPOXIA IN RAINBOW TROUT
8 ~-~
<~6
~~..4 2
i
i 0.1
i
i I I 0.2 0.3 ~4
I
~5
]
0'.6 '
01.7
ApH Fig. 2. Relationship between red blood cell ATP concentrations and the pH difference (ApH) between plasma and red cell in fish exposed to various levels of environmental oxygenation as in fig. 3. Linear regression analysis of the relationship gives the line y = 6.15x + 4.12, regression coefficient r = 0.64. T e m p e r a t u r e = 9 °C.
2b
÷
liE "2
155 120 80 Inspired Poz (Torr) Fig. 3. Steady-state oxygen consumption of rainbow trout after 24 h of exposure to three environmental Po2 levels. Number of animals at each inspired Po2 = 6. T e m p e r a t u r e = 9 ° C.
The changes in plasma and muscle ;lactate concentrations as a function of inspired Po2 are found in tables 1 and 2. Over the entire range of hypoxic exposure levels, a significant lactate difference of between 2.0 and 3.5 mM existed between the two compartments. Whole blood oxygen dissociation curves were not significantly different between blood taken from animals exposed to normoxic (inspired Po: - 155 Torr) or hypoxic (inspired Po~ - 35 Torr) environments for 24 h (fig. 4).
Discussion Adaptation at the organismic level. Mechanisms of hypoxic tolerance can be divided into two broad categories; the first, and potentially the most effective strategy, is the
78
R.G. BOUTILIERet al. 100
~
% Soz
/
o
o
o
50
O Normoxia @ Hypoxia
/I
I
I
I
/
I
I
I
I
50
I
I00 PO2 (Torrl
Fig. 4. Oxygendissociationcurvesfor wholebloodtaken fromfishexposedto 24 h ofnormoxic(1~o2- 155 Tort) or hypoxic (Pie2 ___35 Torr) water at 9 *C. Plasma pH at Pso value = 8.050 (normoxia), 8.025 (hypoxia). reduction of the animal's metabolic rate in concert with the degree of hypoxia. By conforming in this way, the ATP requirements primarily can be met by oxidative metabolism, albeit occurring at a reduced rate. The second, and opposing strategy would be to defend the metabolic rate and either singly or in combination, extract more oxygen from the environment or make up the deficit in ATP by activating anaerobiosis. The disadvantage of selecting the latter two mechanisms is that both are to some degree substrate-limited. Increasing gill ventilation and perfusion is energetically expensive while recruiting anaerobic glycogenolysis for extended periods may become glycogen limited. Both processes are thus constrained by time, with survival being dependent upon the nature and duration of the hypoxic excursion. Our study indicates that rainbow trout respond to 24 h of maintained levels of hypoxia by (i) metabolic depression, and (ii) increasing their reliance on anaerobic metabolism. However, the fact that anaerobic glycogenolysis is not activated at an inspired Po2 of 52 Torr despite a 30~o reduction in metabolic rate (PIo2 = 80 Torr) strongly suggests a hierarchical recruitment of responses favouring adjustments in metabolic rate during moderate levels of hypoxia (table2; fig. 3). Activation of anaerobic glycogenolysis, indicated by decreasing glycogen and increasing lactate levels, occurred at some critical Po2 below 52 Torr. At the most extreme hypoxic state (PIo2 = 32 Torr) white muscle glycogen was 50~o depleted (table 2). Further studies are required to quantify the critical Po2 at which anaerobic processes begin to be recruited into the energy budget of the animal. The immediate response of water breathing fish to decreased environmental 02 is an increase in gill ventilation which appears to be mediated through a receptor that monitors blood oxygen content (Smith and Jones, 1982). Increases in blood pressure during hypoxia are also thought to facilitate 02 uptake by promoting lameilar recruitment which leads to a more equitable distribution
Adjustments to the 02 transport system.
HYPOXIA IN RAINBOW TROUT
79
of blood flow within the lameUae (see Randall, 1982). These changes in ventilation and perfusion probably contribute to the apparent increase in giU diffusing capacity during hypoxia (Fisher et al., 1969). The elevation of plasma catecholamine concentrations during hypoxia in rainbow trout (table 1) may play a further role in facilitating the uptake of oxygen at the gills since adrenaline has been shown to increase the water permeability of the fish gill (Isaia et al., 1978) and, presumably, its diffusional conductance to the respiratory gases (Perry et al., 1985). For example, the difference between inspired and arterial Po2 was 40 Ton" after 24 h of normoxia, decreasing progressively as the levels of inspired Po2 were reduced below 100 Ton" (table 2). After 24 h at Ploe levels of 30 Ton', the PIo~-Pao~ difference (--- 17 Ton') was less than half that seen during normoxia. Catechol~mine release in hypoxic fish (table 1) may confer short-term advantages to oxygen uptake, storage and delivery before being supplemented by the more slowly developing cellular and molecular adaptations. At the level of the red cell, polyanionic phosphates (NTP) are well known modulators of haemoglobin oxygen affinity (Weber, 1982) as is erythrocytic pH (Nikinmaa, 1986). Previous work has shown that decreases in the concentrations of erythrocytic NTP occur after hypoxic exposure periods of about one week and lead to increases in Hb O2-affmity (see Weber, 1982 for review). The adaptive advantage of an increased 02 affinity during chronic hypoxia is that it facilitates 02 loading at the gills and presumably reduces the energetic demands for gill ventilation. This process is thought to be comparatively slow in development and thus to be of lesser importance in the early stages of hypoxia. Rather, it appears that the enhanced cardiorespiratory efforts in the early stages act as an interim measure prior to the full development of longer-term molecular events. Even after 24 h, however, there can be appreciable decreases in the erythrocytic ATP levels depending on the severity of the hypoxic condition (Tetens and Lykkeboe, 1985; table 1). Moreover, as the ATP : Hb4 and GTP:Hb4 molar ratios declined (table 1), erythrocytic pH was brought closer to that of the plasma (fig. 2), possibly because of a change in the Don_nan distribution of protons due in part to the decrease in negative charge intracellularly as the organic phosphates decrease (fig. 2). This decrease in the proton gradient across the cell undoubtedly leads to some modulation of Hb O2-carrying capacity (Boutilier et al., 1986) and affinity (Nikinmaa, 1983; Tetens and Lykkeboe, 1985). The decrease in ApH during hypoxia (table 1) will also be effected by deoxygenation of the haemoglobin (Haldane effect) as has been observed to occur when blood samples of tench were equilibrated in vitro (Jensen, 1986). Using the previously determined relationship for the Root effect of rainbow trout blood (ml O2/gHb = 5.005 + 0.8 pHrb~; Boutilier etaL, 1986), the movement of RBC pH to within 0.18 unit of pile (PIo2 = 30 Tort, table 1), would lead to a 20~o increase in the ability of haemoglobin to carry oxygen. Such increases would offer adaptive advantages over and above those already known to be related to haemoconcentration (Weber, 1982). In this study, the potential effect on Hb O2-binding, caused by the decrease in erythrocytic NTP in fish exposed to severe hypoxia, was not simply related to a change in pHi (table 1, fig. 2), the end result being no change in Hb O2-dissociation curves in blood taken from normoxic and hypoxic fish (fig. 4).
80
R.G. BOUTILIERet al.
During acute exposure to severe environmental hypoxia, erythrocytic pH was maintained at a constant level, despite there being significant changes in the pH of the plasma (fig. 1). This form of pH regulation is reminiscent of that seen following acute extracellular acidoses brought on by burst exercise or acid-infusion, where the strategy is thought to be related to preservation of haemoglobin oxygenation (Nikinmaa, 1983; Boutilier et aL, 1986; Primmett et aL, 1986), and subsequent aerobic performance. Such regulation of erythrocytic pH is mediated through the action of catecholamines (Nikinmaa, 1983) which increase in proportion to the magnitude of the plasma acidosis (Boutilier et al., 1986) and whose action can be offset in vivo by the ~-adrenergic blocking agent, propranolol (Primmett et al., 1986). It seems likely that a similar adrenergic effect leads to the regulation of erythrocytic pH observed during acute acidoses associated with hypoxia (fig. 1) and to progressively alkalinizing effects as exposure is prolonged (table 1). Our results show that anaerobic glycogenolysis was not activated in white muscle until the inspired Po2 decreased to some level below 52 Torr (table 2). One biochemical mechanism advanced to explain how a muscle cell mobilizes its glycogen stores, and coordinates pathway flux to make up the ATP deficit during severe hypoxia, is the catecholamine linked cyclic AMP-dependent activation of phosphorylase b to phosphorylase a (Cohen, 1983). Indeed, the plasma catecholamine levels recorded during severe hypoxia, when blood lactate levels increased (table 1), are similar to those that increase lactate release from resting, perfused skeletal muscle of rat (Richter et al., 1982). If we assume that the relative contribution of anaerobic glycogenolysis to energy demand was constant in muscle over the 24 h period at PIo2 of 32 Torr, the ATP generated from glycogen per unit time is estimated to be about 77 #mol/g wet wt/24 h or 0.05 #mol/g wet wt/min. This would be associated with the production of 25 #tool lactate/g wet wt/24 h. However, only a small fraction of the total lactate produced (about 30%) remained in white muscle (table 2). Presumably, the remaining 70% was effluxed into the blood where it either can be oxidized by aerobic tissues directly, or converted slowly back into liver glycogen via the Cod cycle. The latter alternative, however, is considered too expensive since 6 ATP are required to convert 2 lactate to glucose via gluconeogenesis (Hochachka and Somero, 1984). What is clear from these studies is the capacity of the trout to maintain a relatively high energy status (i.e. maintained ATP concentrations) in its white muscle during even the most severe hypoxic states. There is no evidence of anaerobic mobilization of ATP (AMP, IMP and NH~- remaining constant; Portner et al., 1984; Hochachka, 1985) and the increased reliance on anaerobiosis during hypoxia is accomplished by extensive use of white muscle glycogen and PCr reserves. Utilization of white muscle ATP, as observed following exhaustive exercise in trout (Dobson and Hochachka, 1986), may represent the last in a series of metabolic defense mechanisms which become mobilized during periods of functional tissue hypoxia.
Aerobic-anaerobic transitions.
HYPOXIA IN RAINBOW TROUT Acknowledgements.
81
Supported by N.S.E.R.C. of Canada operating grants to R.G.B. and D.J.R.
References Bergmeyer, H. A. (1983). Methods of enzymatic analysis: Metabolites I., Carbohydrates: Vol. 6, 3rd edition, New York, Academic Press. Boutilier, R. G., G. K. Iwama, T.A. Heming and D.J. Randall (1985). The apparent pK of carbonic acid in rainbow trout blood plasma between 5 and 15 °C. Respir. Physiol. 61: 237-254. Boutilier, R. G., G. K. Iwama and D.J. Randall (1986). The promotion of ¢atecbolamine release in rainbow trout, Salmo gairdneri, by acute acidosis: Interactions between red cell pH and haemoglobin oxygencarrying capacity. J. Exp. Biol. 123: 145-157. Butler, P.J., E.W. Taylor, M.F. Capra and W. Davison (1978). The effect of hypoxia on the levels of circulating catecholamines in the dogfish Scyliorhinus canicula. J. Comp. Physiol. 127: 325-330. Cohen, P. (1983). Protein phosphorylation and the control of glycogen metabolism in skeletal muscle. Phil. Trans. R. Soc. Lond. B. 302: 13-25. Dobson, G.P. and P.W. Hochachka (1987). Role of glycolysis in adenylate depletion and repletion during work and recovery in teleost white muscle. J. Exp. Biol. 129: 125-140. Fisher, T.R., R.F. Coburn and R.E. Forster (1969). Carbon monoxide diffusing capacity in the bullhead catfish. J. Appl. Physiol. 26: 161-169. Fry, F. E.J. (1951). A fractionating column to provide water of various dissolved oxygen content. Can. J. Technol. 29: 144-146. Hochachka, P.W. and G.N. Somero (1984). Biochemical Adaptation. Princeton, N J, Princeton University Press, 537 p. Hochachka, P.W. (1985 ). Fuels and pathways as designed systems for support of muscle work. J. Exp. Biol. 115: 149-164. Isaia, J., J. P. Girard and P. Payan (1978). Kinetic study of gill epithelial permeability to water diffusion in the freshwater rainbow trout, Salmo gairdneri: effect of adrenaline. J. Membr. Biol. 41: 337-347. Jensen, R.B. (1986). Pronounced influence of H b - O 2 saturation on red cell pH in tenth blood/n vivo and in vitro. J. Exp. Zool. 238: 119-124. Milligan, C.L. and C.M. Wood (1985). Inttacellular pH transients in rainbow trout tissues measured by dimethadione distribution. Am. J. Physiol. 248: R668-R673. Nikinmaa, M. (1983). Adrenerglc regulation of haemoglobin oxygen affinity in rainbow trout red cells. J. Comp. Physiol. 152: 67-72. Nikinmaa, M. (1986). Control of red cell pH in teleost fish. Ann. Zool. Fennici, 23: 223-235. Perry, S.F., C. Daxboeck and G.P. Dobson (1985). The effect of perfusion flow rate and adrenergic stimulation on oxygen transfer in the isolated saline perfused head of rainbow trout (Salmo gairdneri). J. Exp. Biol. 116: 251-269. Portner, H.O., Grieshaber, M.K. and N. Heisler (1984). Anaerobiosis and acid-base status in marine invertebrates: a theoretical analysis of proton generation by anaerobic metabolism. J. Comp. Physiol. 155B: 1-12. Primmett, D.R.N., D.J. Randall, M. Mazeand and R.G. Boutilier (1986). The role of catecholamines in erythrocyte pH regulation and oxygen transport in rainbow trout (Salmo gairdneri) during exercise. J. Exp. Biol. 122: 139-148. Randall, D.J. (1982). The control of respiration and circulation in fish during exercise and hypoxia. J. Exp. Biol. 100: 275-288. Richter, E.A., N. B. Rudermann and H. Galbo (1982). Alpha and beta adrenergic effect on metabolism in contracting, perfused muscle. Acta Physiol. Scand. 116: 215-222. Scheid, P. and M. Meyer (1978). Mixing technique for study of oxygen-hemoglobin equilibrium: a critical evaluation. J. Appl. Physiol. 45: 818-822.
82
R.G. BOUTILIER et al.
Smith, F., and D.R. Jones (1982). The effect of changes in blood oxygen-carrying capacity on ventilation volume in the rainbow trout - Salmo gairdneri. J. Exp. Biol. 97: 325-334. Tetens, V. and Lykkeboe, G. (1985). Acute exposure of rainbow trout to mild and deep hypoxia: 02 affinity and 02 capacitance of arterial blood. Respir. PhysioL 61: 221-235. Thomas, S. and G.M. Hughes (1982). A study of the effects of hypoxia on acid-base status of rainbow trout blood using an extracorporeal blood circulation. Respir. Physiol. 49: 371-382. VanKampen, E.J. and W.G. Zijlstra (1965). Determination of haemoglobin and its derivatives. Adv. Clb~. Chem. 8: 141-187. Weber, R.E. (1982). Intraspecific adaptation of hemoglobin function in fish to oxygen availability. In: Exogenous and Endogenous Influences on Metabolic and Neural Control, edited by A. D. F. Addink and N. Spronk. Oxford, Pergamon Press, pp. 87-102.
69
Elsevier
RSP 01358
Acute exposure to graded levels of hypoxia in rainbow trout (Salmo gairdneri): metabolic and respiratory adaptations R.G. Boutilier l, G. Dobson 2, U. Hoeger 3 and D.J. RandalP 1Department of Biology, Dalhoasie University, HalO~ax,Nova Scotia, Canada B3H 4Jl, 2NationalInstitutes of Health, Laboratory of Metabolism, NIA,L4, Rm 55, 12501 WashingtonAve., Rockville MD 20852, U.S.A., 3FakultiitJ~r Biologie, Universit~tKonstanz, 7750 Konstunz, F.R.G., and 4Department of Zoology, University of British Columbia, Vancouver,British Columbia, Canada V6T 2,49 (Accepted for publication 4 August 1987) Abstract. We have studied the mechanisms of acute hypoxia tolerance in rainbow trout (Salmo gairdneri). Fish held at 9 °C were exposed to various levels of bypoxia for 24 h. At an environmental Po2 of 30 Tort, the fish showed an initial plasma acidosis probably of metabolic origin which was subsequently offset such that blood pH returned to normal within about 4 h. Over this time period, red cell pH was maintained constant. Comparing the effects of different levels of hypoxia following 24 h exposure, oxygen consumption of the animal remained unchanged over a broad range of inspired oxygen tensions but declined by over 30 % of normoxic values at inspired water Po2 levels of 80 Tort. This appeared to be a true metabolic depression because signs of increased anaerobic metabolism did not occur until there was a further reduction in water oxygen levels. Rainbow trout appear to be able to maintain a relatively high energy status in their white muscle during 24 h exposure to severe hypoxia (water Po2 = 30 Torr). As the level ofhypoxia was intensified, there was a reduction in the oxygen gradient across the gills, probably facilitated in part by the release of catecholamines into the blood. The erythrocytic ATP: Hb4 molar ratios declined with increasing hypoxic stress as did the pH gradient between the erythrocyte and plasma. The overall effect was no change in Hb O2-affinity alter 24 h exposure to severe hypoxia" Abstract. Anaerobiosis; Hypoxia; Muscle metabolism; Rainbow trout; Red cell pH
Exposure of fish to reduced levels of environmental oxygen initiates a whole host of adaptational processes, most of which are aimed at increasing the availability of oxygen at the level of the tissue. Many of these processes act immediately, such as the changes in ventilation and perfusion of the gill (Randall, 1982), while others including changes in erythrocytic adenylates, muscle mitochondrial densities and enzyme activities (Weber, 1982; Hochachka and Somero, 1984) exert their effects over progressively longer time periods. Another potential strategy of water breathing fish is the lowering of metabolic
Correspondence address: Dr. R.G. Boutilier, Department of Biology, Dalhousie University, Halifax, Nova Scotia, Canada B3H 4Jl. 0034-5687/88/$03.50 © 1988 Elsevier Science Publishers B.V. (Biomedical Division)
70
R.G. BOUTILIERet al.
rate which in effect minimizes the impact of the hypoxic insult by conforming to the environmental Po2. Exposure of rainbow trout to water Po2 tensions of 30 Torr leads to a marked extracellular acidosis within the In'st 20 rain and subsequently to an elevation of blood pH to values approaching (Tetens and Lykkeboe, 1985) or exceeding (Thomas and Hughes, 1982) those of the normoxic controls. These initial responses are complex, since the increased gill irrigation leads to a minor respiratory alkalosis whereas the decreased availability of oxygen causes a lactacidemia (Thomas and Hughes, 1982). The time course and severity of the hypoxia greatly influences the degree of the acidosis and therefore H b - O z transport. Recent studies on rainbow trout have shown that red cell pH is regulated at a constant level during and following periods of extracellular acidosis brought on either by anaerobic exercise (Primmett etal., 1986) or direct acid infusion (Boutilier etal., 1986). In both instances catecholamines are released into the blood and would appear, through fl-adrenergic mechanisms at the erythrocytic membrane, to be responsible for turning on cellular ion exchange processes which ultimately lead to the regulation of the pH environment of the haemoglobin (Nikinmaa, 1983, 1986). As an oxygen safeguarding system, the ability to regulate the pH of the red cell during periods of extraceHular acidosis would be of vital importance in facilitating oxygen loading at the gills and in maintaining oxygen delivery to the tissues. Decreased oxygen availability per se, with or without a corresponding acidosis, represents a direct challenge to the oxygen transport and delivery system of the animal. Studies on the marine elasmobranch, Scyliorhinus canicula, have shown that plasma catecholamine concentrations increase in response to hypoxia (Butler et al., 1978). Moreover, it has been shown that hypoxia in rainbow trout causes an increase in red blood cell volume which can be produced in vitro by addition of adrenaline to blood (Nikinmaa, 1986). The present study attempts to clarify the mechanisms of acute hypoxia tolerance in rainbow trout (Salmo gairdneri). Not only will a number of fundamental organismic strategies be explored but emphasis will be placed on integrating these with adrenergic regulation of H b - O 2 carrying capacity and/or affinity as well as muscle metabolism.
Materials and methods
Animals and preparation. Rainbow trout (S. gairdneri) weighing 220-460 g were obtained from the Sun Valley Trout Farm, Mission, B.C. and were kept in outdoor tanks supplied with dechlorinated tap water (5-12 ° C). Feeding was suspended and the animals were transferred to a 9 °C holding tank at least 48 h prior to surgery. The dorsal aorta was chronically cannulated with P.E. 50 tubing after the animals had been anaesthetized in a 0.05Y/o solution of tricainemethanesulfonate (MS-222, Sigma) buffered to pH 7.5 with NaHCO 3. During placement of the cannula, the gills were irrigated with a lighter dose (0.013 ~o MS-222) of the anaesthetic solution which was oxygenated to ambient Po2 levels. Following the operation, the fish were transferred
HYPOXIAIN RAINBOWTROUT
71
to darkened Perspex boxes supplied with flowing aerated water at 9 °C where they remained for at least 72 h before being used for experiments. The cannulae were flushed daily with 0.1 ml Cortland saline containing 10 i.u./ml of heparin (Sigma).
Exposure offish to hypoxia. All experiments were carried out on fish kept in darkened Perspex chambers. Each of the chambers could be individually supplied with flowing water of known oxygen partial pressures produced by oxygen-stripping columns similar to those described by Fry (1951). Thirty fish were exposed to normocapnic conditions of normoxia or various degrees of hypoxia for 24 h before sampling took place. In each instance, a 2 ml blood sample was then drawn from the dorsal aortic catheter and taken into a gas tight syringe for analyses of blood gases, pH and metabolites (see below). Fish were then rapidly transferred from the black holding chambers and decapitated, after which samples of white epaxial muscle were immediately excised from a site posterior to the dorsal fin. Samples were immediately freeze-clamped in aluminum blocks (pre-cooled to - 196 °C in liquid nitrogen) and stored at - 7 0 °C until required. The time between capture of fish and freeze clamping muscle was measured to be within 6-10 see. In a second series of experiments, six rainbow trout were exposed to three levels of environmental hypoxia for determination of oxygen consumption by measuring the O z partial pressure difference and rates of water flow through sealed Perspex chambers. Measurements were taken while the animals were resting in normoxic conditions and after 24 h exposure to environmental hypoxia. Accurate measurement of the Po2 difference became increasingly difficult as the inspired Po2 levels were decreased below 80 Torr, and so were discarded. Reducing the water flow in the chamber to achieve a larger inspired-expired Po2 difference has the effect of increasing water Pco~ and thereby exacerbating steady-state normocapnic measures of Oz consumption. In a final series of experiments six rainbow trout were cannulated as before for dorsal aortic blood sampling. Blood samples were taken before and at specified periods following acute exposure to inspired Po2 levels of 30 Torr. The blood samples were used to measure plasma and red cell pH as outlined below. In vitro determination of Hb O2-affinity. Animals with dorsal aortic catheters were exposed for 24 h to normoxic (Plo2 155 Torr) or severely hypoxic (PIo2 35 Torr) environments, after which blood samples were taken and pooled to the required volumes in two glass tonometer flasks. The blood samples were equilibrated at 10 °C with humidified and thermostatted gas mixtures containing 0.2~o CO2, the balance of which was either air or nitrogen (WOsthoff gas mixing pumps, Bochum, F.R.G.). Oxygen dissociation curves were prepared using the mixing method as detailed by Scheid and Meyer (1978). Analyticalprocedures. Measurements of whole blood pH and red cell lysate were made using Radiometer microdectrode units coupled with PHM 84 pH meters (Boutilier et al., 1985). The freeze-thaw method was used to produce the red cell lysate whose
72
R.G. BOUTILIERet al.
pH is not measurably different than that determined by application of the DMO method in rainbow trout erythrocytes (Milligan and Wood, 1985). The pH electrodes were calibrated before each measurement with Radiometer precision buffer solutions S 1500 and S 1510. Oxygen partial pressures in blood and water were measured using Radiometer electrodes and 02 meters. Oxygen contents of whole blood and CO2 contents of true plasma were measured using a Lex-O2-Con (Lexington Instntments, MA) and a Carle Series 100 gas chromatograph (see Boutilier et al., 1985) respectively. Hemoglobin concentration was determined by the speetrophotometric method of Van Kampen and Zijlstra (1965). Plasma adrenaline and noradrenaline concentrations were measured by. high pressure liquid chromatography (cf.Primmett et al., 1986) using a Bioanalytical Systems HPLC with electrochemical detection. Whole blood and plasma extracts. Samples of either whole blood or true plasma (150 #1) were immediately deproteinized with 300 #1 of ice-cold 10~o trichloroacetic acid (TCA). Following a 15-30 min extraction time on ice, the samples were centrifuged for 30-45 see at 11000 rpm. A known volume of clear supernatant was removed and then neutralized with a pre-determined volume of saturated Tris Base [tris (hydroxymethyl)-aminomethane]. The neutralized extracts were immediately frozen in liquid nitrogen and stored at - 70 °C until required for metabolite analysis.
White epaxial muscle was powdered under fiquid nitrogen using a pre-cooled mortar and pestle. Tendons and fragments of connective tissue were dissected free and discarded. About 500 nag of powdered tissue was transferred to a pre-cooled, pre-weighed vial containing 1.0 ml of ice-cold 0.6 N perchloric acid (PCA) and then reweighed. A further 1.0 ml PCA was added and the powder homogenized for 15 see at 0°C using an Ultra-Turrax homogenizer. The homogenization procedure was repeated and the side of the vial washed down with a further 0.5 ml PCA. The suspension was stirred under low speed and 100 #1 was removed in duplicate for glycogen determination. The remaining homogenate was then centrifuged for 2 min at 13000 rpm and 4 °C after which approximately 2 ml of supernatant was immediately neutralized (pH 7.0) with Tris Base, frozen in liquid nitrogen and kept at - 70 °C until required. The homogenization and neutralization procedure described takes about 6 min to complete. In a parallel set of validation studies, the highly acid-labile phosphoereatine underwent not more than 5 ~o hydrolysis with this technique. Muscle homogenization, extraction and neutralization.
Plasma and muscle lactate concentrations, and muscle glycogen concentrations were measured in a Pye-Unicam SP8-100 UV-VIS spectrophotometer at 340 nm using the routine NADH- or NAD-coupled enzymatic procedures described in Bergmeyer (1983). Each assay was validated with appropriate standards. Metabolite determinations.
HYPOXIA IN RAINBOW TROUT
73
Determination ofnucleotides andphosphagens. Concentrations of the nucleotides ATP, ADP, AMP and IMP and the phosphagens, phosphocreatine (PCr) and creatine (Cr) were analytically determined by high pressure liquid chromatography (HPLC). The procedure was carded out on a Spectra Physics 8000 B HPLC coupled to a Kratos Spectroflow 773 absorbance detector set at 254 nm. The nucleotides were passed through a 250 x 4.6 n'an AX-300 anion exchange column (Brownlee Laboratories) at a flow rate of 2 ml/min and eluted in 30 rain using a gradient of potassium dihydrogen phosphate. Buffer A consisted of 50 mM KHzPO 4 (pH 2.31 at 55 °C) and buffer B of 600 mM KHzPO4 (pH 2.63 at 55 °C). The elution conditions were as follows: 100~ A for 4 min, 0 to 50% B from 4 to 5 rain, 50 to 55~ B from 5 to 10 rain, 55 to 100% B from 10 to 11 rain, 100~ B from 11 to 30 min. PCR, Cr and AMP were eluted at the same flow rate of 2 ml/min using an isocratic 50 mM KHzPO 4 buffer at pH 3.1, and detected at 210 nm (10 #1 of neutralized sample was required for each set of determinations). Following every set of six chromatographic runs, the column was washed with 600 mM KHzPO 4 and a single 10 #1 injection of known standards was passed through the system to check retention times of each metabolite. Internal standards were used to validate the procedure for muscle analysis. When stated, the significance (P < 0.05) of differences between means was assessed with Student's t-test.
Results
Data for metabolic and respiratory changes in the blood compartments of trout exposed to several stages of hypoxia for 24 h are shown in table 1. Compared with normoxic animals, the blood pH was unchanged after 24 h of hypoxic exposure even though transient changes occur within that time period (fig. 1). After 24 h exposure to deep hypoxia, plasma pH appears to be regulated at normoxic levels by offsetting the pronounced lactacidemia with a decrease in arterial Pco~, presumably through an increase in gill ventilation (tables 1 and 2). Changes in plasma and erythrocytic pH are shown in fig. 1 as a function of time after fish were acutely exposed to an inspired Po~ (PIo~) of 30 Torr. A plasma acidosis was observed after 30 rain exposure, recovery from which took between 2-4 h. Despite the marked reduction of plasma pH, that of the erythrocyte was held at a constant level over the first 4 h of hypoxia and thereafter increased. At the end of 24 h, the erythrocytic pH was within 0.227 units of plasma pH, 0.3 units closer than the normoxic conditions (fig. 1). These latter results are similar to those observed after 24 h exposure to various levels of hypoxia (table 1). Arterial pH after 24 h exposure to all levels of hypoxia was unchanged from that recorded in normoxia. Despite the constancy of arterial pH, the difference in pH between plasma and erythrocytes (ApH) was reduced under extreme hypoxia (table 1). Relative to normoxic fish, plasma adrenaline and noradrenaline concentrations were elevated during hypoxia, with the highest levels being recorded in the animals exposed to the lowest inspired 02 tensions (table 1).
TABLE
1
0.50 f 0.05
[lactate-],
22.5
1.1
10.5
8.5
+ 1.4
f 0.2
+- 0.4
+ 0.4
k 2.5
[lactate-], mmol.kgg’ cell water
6.53 f 0.58
1.22 f 0.09
1.13 * 0.13
ATP: Hb, molar ratio
ml- 1 packed
[ATP], mol. RBC
RBC
+ 0.04
0.54
PH, - PHi (PHI
[GTP], mol. ml - ’ packed
7.517 * 0.037
0.338 f 0.058
Red blood cell: PH
0.5
f
+
0.082
1.27 f
1.24 f
0.14
0.83
0.07
+ 0.09
0.77 *
0.83
0.440 f
1.491 +_ 0.032
0.01
0.6
0.1
+_ 0.3
+
f 10.0
0.35 j:
22.7
8.0
10.8
9.2
79.9
2.08 * 0.41
0.22 +-0.13
116.0
1.78 f 0.17
0.34 f 0.02
blood
f 0.7
1.23 + 0.21
8.5
f 0.2
7.926 f 0.068 2.4
0.42 + 0.09
f 0.7
MCHC, g.rn-t
Hct, %
[Hb], g. 100 ml-’
Maximum 0, concentration vols y0
O2 concentration, vols %
Arterial blood: Po, (To=)
’ [noradrenaline], run01. 1- 1
[adrenaline], mnol 9 1-
mmol~l-’
8.6
[HCO,-],mmol.l-’
+ 0.5
1.855 + 0.084
2.9
Pco, (Torr)
:
Arterial plasma PH
120 (n = 6)
155
(Torr)
(n = 6)
Inspired water P,
f 0.3
+ 0.2
* 0.01
f 2.7
f 0.4
+ 1.2
f 0.9
f 5.3
1.21 f 0.05
6.45 + 0.16
0.77 f 0.03
1.11 + 0.09
0.380 f 0.048
7.483 +_0.030
0.35
23.5
8.2
11.0
1.4
51.1
2.11 * 0.22
1.66 f 0.28
0.72 f 0.17
6.1
2.1
7.883 f 0.044
(n = 5)
90
Temperature = 9°C. n = number of animals.
+ 0.8
+ 0.2
+ 2.1
f 0.02
+ 2.6
+ 0.2
+ 0.4
+ 0.4
5.23 + 0.34 1.14 + 0.12
0.64 f 0.07
1.24 * 0.08
7.510 + 0.027 0.240 + 0.043
0.32
30.3
9.6
12.5
5.2
25.9
0.95 f 0.18 7.10 + 2.68 4.40 + 2.80
6.1
2.6
f 0.3 f 0.8
f 2.1
+ 0.4
* 1.0
f 0.7
+- 0.6
0.92 & 0.07
4.21 + 0.59
0.62 f 0.14
3.53 * 0.45
0.169 + 0.053
7.662 +_0.037
0.29 + 0.03
31.2
9.2
12.1
3.6
13.5
13.79 +_4.11
8.81 k 3.56
6.14 k 1.03
5.5
1.9
7.840 f 0.098
(n = 4)
7.810 * 0.051
30
50 (n = 5)
Means + 1 SEM of measured variables from rainbow trout (S. guirdneri) after 24 h exposure to different levels of oxygen partial pressure (Po,) of inspired water.
HYPOXIA IN RAINBOW TROUT
75
8.1pH
7.9 ~7.7 7..~ 72
0
~
~
~
~
,,/
~'4
Time ( h )
Fig. 1. Changesin medal bloodplasmapH andred bloodcellpH afterexposureto a stepchangeininspired water Po2from150Torrto 30Tort (step changeindicatedby verticallineat time0). C = normoxiecontrol values. Temperature= 10.5°C. The effect of acute exposure to graded levels of hypoxia on the concentration of guanosine 5-triphosphate (GTP) and adenosine 5-triphosphate (ATP) in the red cells of trout is shown in table 1. The concentration of GTP decreased sjt,nificantly by about 30% in response to mild hypoxia (a decrease in PIo2 from 152 to 122 Torr), but stabilized at this level during the more severe hypoxic states. The red cell ATP concentrations followed a different pattern of change with no significant decrease occurring until the environmental Po2 fell to 52 Torr. A further percentage reduction in ATP occurred at the lowest inspired 0 2 tension of 32 Torr (table 1). In general, the reduction in ApH was correlated with a decrease in RBC ATP levels (fig. 2). Haemoglobin concentrations in whole blood increased with haematocrit and decreasing environmental Poz. The increased haematocrit was due to both haemoconcentration and red cell swelling, the latter leading to a corresponding decrease in the mean cellular haemoglobin concentration (MCHC; table 1). The decrease in RBC ATP was somewhat greater than the reduction in RBC haemoglobin concentration due to cellular swelling, the net result being a slight decrease in the ATP : Hb ratio during severe hypoxia (table 1). The effect of decreasing environmental Po2 on the oxygen consumption of rainbow trout is shown in fig. 3. It is clear that after 24 h of mild hypoxia (PIo~ = 120 Torr), oxygen consumption was unchanged, but declined significantly by over 30% at the lower inspired Po2 of 80 Torr. The concentrations of a selected number of metabolites in white muscle of rainbow trout are presented in table 2. Phosphocreatine (PCr), glycogen, lactate, ATP, adenosine 5-monophosphate (AMP) and inosine 5-monophosphate (IMP) underwent no appreciable change in concentration after 24 h of varying degrees of hypoxia down to an inspired Po: of 52 Torr. In contrast, at an inspired Po2 of 32 Torr, glycogen decreased by 50% (from 25.54 to 12.95 pmol/g wet wt), PCr decreased by about 40% (from 16.58 to 19.80 #mol/g wet wt), and lactate increased 2.5 fold (3.33 to 8.33 #mol/g wet wt) compared to the normoxic state. No significant changes occurred in the concentrations of either ATP, AMP, IMP or ammonium ion (NH2-) throughout the various stages of Q hypoxic exposure (table 2).
TABLE 2
19.1
26.5
38.5
41.7
37.0
16.58 (_+ 1.09) 17.74 (_+0.70) 18.41 (_+0.73) 17.90 (_+2.02) 10.80 ( _+ 1.12)
116.0 (_+2.5) 79.9 (_+ 10.0) 51.1 (-+5.3) 25.9 (_+2.7) 13.5 ( _+0.6)
153.0 (_+0.9) 121.6 (_+ 1.3) 89.6 (_+4.2) 52.4 (_+ 1.9) 32.6 ( _+2.8)
39.00 (_+3.00) 39.29 (_+ 1.07) 37.35 (_+1.44) 35.30 (_+2.38) 43.98 ( _+2.25)
Cr
PCr
Arterial Po2
Inspired Po2
APo2
#mol" g - ~ wet weight
Torr
1.18 ( _+0.21)
-
-
1.06 (_+0.15) -
NH4 +
0.86 (_+0.09) 0.95 (_+0.47) 0.82 (_+0.20) 0.6 (_+0.16) 0.77 ( _+0.14)
IMP
0.77 (_+0.03) 0.81 (_+0.05) 0.74 (_+0.22) 0.70 (_+0.04) 0.76 ( _+0.26)
ADP
6.25 (_+0.16) 6.57 (_+0.12) 6.79 (_+0.14) 6.73 (_+0.20) 6.23 ( _+0.07)
ATP
0.047 (_+0.006) 0.064 (_+0.013) 0.065 (_+0.030) 0.029 (_+0.004) 0.043 ( _+0.009)
AMP
3.33 (_+0.60) 4.70 (+_ 1.23) 2.95 (_+0.24) 4.40 (_+ 0.18) 8.33 ( _+2.09)
Lactate
25.54 (_+0.71) 27.14 (_+2.22) 23.48 (_+0.55) 25.09 (_+ 1.14) 12.95 ( _+ 1.44)
Glycogen
4
4
5
6
5
N
Concentration of white muscle metabolites after 24 h ofnormoxia and various levels of hypoxia in rainbow trout (S. gairdneri) at 9°C. Values are means _+ 1 SEM N = number of animals.
rn
¢¢ O
O~
77
HYPOXIA IN RAINBOW TROUT
8 ~-~
<~6
~~..4 2
i
i 0.1
i
i I I 0.2 0.3 ~4
I
~5
]
0'.6 '
01.7
ApH Fig. 2. Relationship between red blood cell ATP concentrations and the pH difference (ApH) between plasma and red cell in fish exposed to various levels of environmental oxygenation as in fig. 3. Linear regression analysis of the relationship gives the line y = 6.15x + 4.12, regression coefficient r = 0.64. T e m p e r a t u r e = 9 °C.
2b
÷
liE "2
155 120 80 Inspired Poz (Torr) Fig. 3. Steady-state oxygen consumption of rainbow trout after 24 h of exposure to three environmental Po2 levels. Number of animals at each inspired Po2 = 6. T e m p e r a t u r e = 9 ° C.
The changes in plasma and muscle ;lactate concentrations as a function of inspired Po2 are found in tables 1 and 2. Over the entire range of hypoxic exposure levels, a significant lactate difference of between 2.0 and 3.5 mM existed between the two compartments. Whole blood oxygen dissociation curves were not significantly different between blood taken from animals exposed to normoxic (inspired Po: - 155 Torr) or hypoxic (inspired Po~ - 35 Torr) environments for 24 h (fig. 4).
Discussion Adaptation at the organismic level. Mechanisms of hypoxic tolerance can be divided into two broad categories; the first, and potentially the most effective strategy, is the
78
R.G. BOUTILIERet al. 100
~
% Soz
/
o
o
o
50
O Normoxia @ Hypoxia
/I
I
I
I
/
I
I
I
I
50
I
I00 PO2 (Torrl
Fig. 4. Oxygendissociationcurvesfor wholebloodtaken fromfishexposedto 24 h ofnormoxic(1~o2- 155 Tort) or hypoxic (Pie2 ___35 Torr) water at 9 *C. Plasma pH at Pso value = 8.050 (normoxia), 8.025 (hypoxia). reduction of the animal's metabolic rate in concert with the degree of hypoxia. By conforming in this way, the ATP requirements primarily can be met by oxidative metabolism, albeit occurring at a reduced rate. The second, and opposing strategy would be to defend the metabolic rate and either singly or in combination, extract more oxygen from the environment or make up the deficit in ATP by activating anaerobiosis. The disadvantage of selecting the latter two mechanisms is that both are to some degree substrate-limited. Increasing gill ventilation and perfusion is energetically expensive while recruiting anaerobic glycogenolysis for extended periods may become glycogen limited. Both processes are thus constrained by time, with survival being dependent upon the nature and duration of the hypoxic excursion. Our study indicates that rainbow trout respond to 24 h of maintained levels of hypoxia by (i) metabolic depression, and (ii) increasing their reliance on anaerobic metabolism. However, the fact that anaerobic glycogenolysis is not activated at an inspired Po2 of 52 Torr despite a 30~o reduction in metabolic rate (PIo2 = 80 Torr) strongly suggests a hierarchical recruitment of responses favouring adjustments in metabolic rate during moderate levels of hypoxia (table2; fig. 3). Activation of anaerobic glycogenolysis, indicated by decreasing glycogen and increasing lactate levels, occurred at some critical Po2 below 52 Torr. At the most extreme hypoxic state (PIo2 = 32 Torr) white muscle glycogen was 50~o depleted (table 2). Further studies are required to quantify the critical Po2 at which anaerobic processes begin to be recruited into the energy budget of the animal. The immediate response of water breathing fish to decreased environmental 02 is an increase in gill ventilation which appears to be mediated through a receptor that monitors blood oxygen content (Smith and Jones, 1982). Increases in blood pressure during hypoxia are also thought to facilitate 02 uptake by promoting lameilar recruitment which leads to a more equitable distribution
Adjustments to the 02 transport system.
HYPOXIA IN RAINBOW TROUT
79
of blood flow within the lameUae (see Randall, 1982). These changes in ventilation and perfusion probably contribute to the apparent increase in giU diffusing capacity during hypoxia (Fisher et al., 1969). The elevation of plasma catecholamine concentrations during hypoxia in rainbow trout (table 1) may play a further role in facilitating the uptake of oxygen at the gills since adrenaline has been shown to increase the water permeability of the fish gill (Isaia et al., 1978) and, presumably, its diffusional conductance to the respiratory gases (Perry et al., 1985). For example, the difference between inspired and arterial Po2 was 40 Ton" after 24 h of normoxia, decreasing progressively as the levels of inspired Po2 were reduced below 100 Ton" (table 2). After 24 h at Ploe levels of 30 Ton', the PIo~-Pao~ difference (--- 17 Ton') was less than half that seen during normoxia. Catechol~mine release in hypoxic fish (table 1) may confer short-term advantages to oxygen uptake, storage and delivery before being supplemented by the more slowly developing cellular and molecular adaptations. At the level of the red cell, polyanionic phosphates (NTP) are well known modulators of haemoglobin oxygen affinity (Weber, 1982) as is erythrocytic pH (Nikinmaa, 1986). Previous work has shown that decreases in the concentrations of erythrocytic NTP occur after hypoxic exposure periods of about one week and lead to increases in Hb O2-affmity (see Weber, 1982 for review). The adaptive advantage of an increased 02 affinity during chronic hypoxia is that it facilitates 02 loading at the gills and presumably reduces the energetic demands for gill ventilation. This process is thought to be comparatively slow in development and thus to be of lesser importance in the early stages of hypoxia. Rather, it appears that the enhanced cardiorespiratory efforts in the early stages act as an interim measure prior to the full development of longer-term molecular events. Even after 24 h, however, there can be appreciable decreases in the erythrocytic ATP levels depending on the severity of the hypoxic condition (Tetens and Lykkeboe, 1985; table 1). Moreover, as the ATP : Hb4 and GTP:Hb4 molar ratios declined (table 1), erythrocytic pH was brought closer to that of the plasma (fig. 2), possibly because of a change in the Don_nan distribution of protons due in part to the decrease in negative charge intracellularly as the organic phosphates decrease (fig. 2). This decrease in the proton gradient across the cell undoubtedly leads to some modulation of Hb O2-carrying capacity (Boutilier et al., 1986) and affinity (Nikinmaa, 1983; Tetens and Lykkeboe, 1985). The decrease in ApH during hypoxia (table 1) will also be effected by deoxygenation of the haemoglobin (Haldane effect) as has been observed to occur when blood samples of tench were equilibrated in vitro (Jensen, 1986). Using the previously determined relationship for the Root effect of rainbow trout blood (ml O2/gHb = 5.005 + 0.8 pHrb~; Boutilier etaL, 1986), the movement of RBC pH to within 0.18 unit of pile (PIo2 = 30 Tort, table 1), would lead to a 20~o increase in the ability of haemoglobin to carry oxygen. Such increases would offer adaptive advantages over and above those already known to be related to haemoconcentration (Weber, 1982). In this study, the potential effect on Hb O2-binding, caused by the decrease in erythrocytic NTP in fish exposed to severe hypoxia, was not simply related to a change in pHi (table 1, fig. 2), the end result being no change in Hb O2-dissociation curves in blood taken from normoxic and hypoxic fish (fig. 4).
80
R.G. BOUTILIERet al.
During acute exposure to severe environmental hypoxia, erythrocytic pH was maintained at a constant level, despite there being significant changes in the pH of the plasma (fig. 1). This form of pH regulation is reminiscent of that seen following acute extracellular acidoses brought on by burst exercise or acid-infusion, where the strategy is thought to be related to preservation of haemoglobin oxygenation (Nikinmaa, 1983; Boutilier et aL, 1986; Primmett et aL, 1986), and subsequent aerobic performance. Such regulation of erythrocytic pH is mediated through the action of catecholamines (Nikinmaa, 1983) which increase in proportion to the magnitude of the plasma acidosis (Boutilier et al., 1986) and whose action can be offset in vivo by the ~-adrenergic blocking agent, propranolol (Primmett et al., 1986). It seems likely that a similar adrenergic effect leads to the regulation of erythrocytic pH observed during acute acidoses associated with hypoxia (fig. 1) and to progressively alkalinizing effects as exposure is prolonged (table 1). Our results show that anaerobic glycogenolysis was not activated in white muscle until the inspired Po2 decreased to some level below 52 Torr (table 2). One biochemical mechanism advanced to explain how a muscle cell mobilizes its glycogen stores, and coordinates pathway flux to make up the ATP deficit during severe hypoxia, is the catecholamine linked cyclic AMP-dependent activation of phosphorylase b to phosphorylase a (Cohen, 1983). Indeed, the plasma catecholamine levels recorded during severe hypoxia, when blood lactate levels increased (table 1), are similar to those that increase lactate release from resting, perfused skeletal muscle of rat (Richter et al., 1982). If we assume that the relative contribution of anaerobic glycogenolysis to energy demand was constant in muscle over the 24 h period at PIo2 of 32 Torr, the ATP generated from glycogen per unit time is estimated to be about 77 #mol/g wet wt/24 h or 0.05 #mol/g wet wt/min. This would be associated with the production of 25 #tool lactate/g wet wt/24 h. However, only a small fraction of the total lactate produced (about 30%) remained in white muscle (table 2). Presumably, the remaining 70% was effluxed into the blood where it either can be oxidized by aerobic tissues directly, or converted slowly back into liver glycogen via the Cod cycle. The latter alternative, however, is considered too expensive since 6 ATP are required to convert 2 lactate to glucose via gluconeogenesis (Hochachka and Somero, 1984). What is clear from these studies is the capacity of the trout to maintain a relatively high energy status (i.e. maintained ATP concentrations) in its white muscle during even the most severe hypoxic states. There is no evidence of anaerobic mobilization of ATP (AMP, IMP and NH~- remaining constant; Portner et al., 1984; Hochachka, 1985) and the increased reliance on anaerobiosis during hypoxia is accomplished by extensive use of white muscle glycogen and PCr reserves. Utilization of white muscle ATP, as observed following exhaustive exercise in trout (Dobson and Hochachka, 1986), may represent the last in a series of metabolic defense mechanisms which become mobilized during periods of functional tissue hypoxia.
Aerobic-anaerobic transitions.
HYPOXIA IN RAINBOW TROUT Acknowledgements.
81
Supported by N.S.E.R.C. of Canada operating grants to R.G.B. and D.J.R.
References Bergmeyer, H. A. (1983). Methods of enzymatic analysis: Metabolites I., Carbohydrates: Vol. 6, 3rd edition, New York, Academic Press. Boutilier, R. G., G. K. Iwama, T.A. Heming and D.J. Randall (1985). The apparent pK of carbonic acid in rainbow trout blood plasma between 5 and 15 °C. Respir. Physiol. 61: 237-254. Boutilier, R. G., G. K. Iwama and D.J. Randall (1986). The promotion of ¢atecbolamine release in rainbow trout, Salmo gairdneri, by acute acidosis: Interactions between red cell pH and haemoglobin oxygencarrying capacity. J. Exp. Biol. 123: 145-157. Butler, P.J., E.W. Taylor, M.F. Capra and W. Davison (1978). The effect of hypoxia on the levels of circulating catecholamines in the dogfish Scyliorhinus canicula. J. Comp. Physiol. 127: 325-330. Cohen, P. (1983). Protein phosphorylation and the control of glycogen metabolism in skeletal muscle. Phil. Trans. R. Soc. Lond. B. 302: 13-25. Dobson, G.P. and P.W. Hochachka (1987). Role of glycolysis in adenylate depletion and repletion during work and recovery in teleost white muscle. J. Exp. Biol. 129: 125-140. Fisher, T.R., R.F. Coburn and R.E. Forster (1969). Carbon monoxide diffusing capacity in the bullhead catfish. J. Appl. Physiol. 26: 161-169. Fry, F. E.J. (1951). A fractionating column to provide water of various dissolved oxygen content. Can. J. Technol. 29: 144-146. Hochachka, P.W. and G.N. Somero (1984). Biochemical Adaptation. Princeton, N J, Princeton University Press, 537 p. Hochachka, P.W. (1985 ). Fuels and pathways as designed systems for support of muscle work. J. Exp. Biol. 115: 149-164. Isaia, J., J. P. Girard and P. Payan (1978). Kinetic study of gill epithelial permeability to water diffusion in the freshwater rainbow trout, Salmo gairdneri: effect of adrenaline. J. Membr. Biol. 41: 337-347. Jensen, R.B. (1986). Pronounced influence of H b - O 2 saturation on red cell pH in tenth blood/n vivo and in vitro. J. Exp. Zool. 238: 119-124. Milligan, C.L. and C.M. Wood (1985). Inttacellular pH transients in rainbow trout tissues measured by dimethadione distribution. Am. J. Physiol. 248: R668-R673. Nikinmaa, M. (1983). Adrenerglc regulation of haemoglobin oxygen affinity in rainbow trout red cells. J. Comp. Physiol. 152: 67-72. Nikinmaa, M. (1986). Control of red cell pH in teleost fish. Ann. Zool. Fennici, 23: 223-235. Perry, S.F., C. Daxboeck and G.P. Dobson (1985). The effect of perfusion flow rate and adrenergic stimulation on oxygen transfer in the isolated saline perfused head of rainbow trout (Salmo gairdneri). J. Exp. Biol. 116: 251-269. Portner, H.O., Grieshaber, M.K. and N. Heisler (1984). Anaerobiosis and acid-base status in marine invertebrates: a theoretical analysis of proton generation by anaerobic metabolism. J. Comp. Physiol. 155B: 1-12. Primmett, D.R.N., D.J. Randall, M. Mazeand and R.G. Boutilier (1986). The role of catecholamines in erythrocyte pH regulation and oxygen transport in rainbow trout (Salmo gairdneri) during exercise. J. Exp. Biol. 122: 139-148. Randall, D.J. (1982). The control of respiration and circulation in fish during exercise and hypoxia. J. Exp. Biol. 100: 275-288. Richter, E.A., N. B. Rudermann and H. Galbo (1982). Alpha and beta adrenergic effect on metabolism in contracting, perfused muscle. Acta Physiol. Scand. 116: 215-222. Scheid, P. and M. Meyer (1978). Mixing technique for study of oxygen-hemoglobin equilibrium: a critical evaluation. J. Appl. Physiol. 45: 818-822.
82
R.G. BOUTILIER et al.
Smith, F., and D.R. Jones (1982). The effect of changes in blood oxygen-carrying capacity on ventilation volume in the rainbow trout - Salmo gairdneri. J. Exp. Biol. 97: 325-334. Tetens, V. and Lykkeboe, G. (1985). Acute exposure of rainbow trout to mild and deep hypoxia: 02 affinity and 02 capacitance of arterial blood. Respir. PhysioL 61: 221-235. Thomas, S. and G.M. Hughes (1982). A study of the effects of hypoxia on acid-base status of rainbow trout blood using an extracorporeal blood circulation. Respir. Physiol. 49: 371-382. VanKampen, E.J. and W.G. Zijlstra (1965). Determination of haemoglobin and its derivatives. Adv. Clb~. Chem. 8: 141-187. Weber, R.E. (1982). Intraspecific adaptation of hemoglobin function in fish to oxygen availability. In: Exogenous and Endogenous Influences on Metabolic and Neural Control, edited by A. D. F. Addink and N. Spronk. Oxford, Pergamon Press, pp. 87-102.