Regulation of Blood Oxygen Affinity in the Australian Blackfish Gadopsis Marmoratus: I. Correlations between Oxygen-binding Properties, Habitat and Swimming Behaviour

The oxygen equilibrium curve of whole blood is an important component of the oxygen transport system in water-breathing fish. The shape and position of the curve represents a compromise between the need to obtain sufficient oxygen from an environment that, relative to air, is oxygen deficient, and the need to permit satisfactory unloading at the tensions encountered in tissue capillaries. Thus it is not surprising to find that species-specific differences in the oxygen equilibrium curve reflect differences in environmental oxygen availability and in behaviour and physiology (Krogh & Leitch, 1919; Willmer, 1934; Wood & Lenfant, 1979). This’fine’tuning of the oxygen equilibrium curve involves both the intrinsic oxygen-binding properties of haemoglobin, and highly specific interactions between haemoglobin and a number of allosteric modulators operating within the erythrocyte. Among the most important of these modulators of haemoglobin function are organic phosphates, pH, CO₂ and certain inorganic ions (Benesch & Benesch, 1967; Chanutin & Curnish, 1967).

The individual effects of these modulators have been well documented for fish haemoglobins (for review see Riggs, 1979; Wood, 1980), but the aim of the present study was to investigate the relative importance of each modulator in adjusting the oxygen equilibrium curve of whole blood in the freshwater blackfish Gadopsis marmoratus, and seek correlations between the effects of these modulators and the habitat and swimming behaviour of this fish.

The influence of environmental temperature, another important determinant of haemoglobin function in water-breathing fish, is considered in an accompanying paper (Dobson & Baldwin, 1982).

Blackfish (Gadopsis marmoratus) were collected by electrofishing in the Yarra river, 20 km east of Warburton, Victoria, Australia. Mature fish ranging in length from 18 to 25 cm were caught during the summer months when the water temperatures averaged 20 °C. The animals were maintained for at least a month at 20 °C in filtered aerated aquaria housed in a constant-temperature room set on a 12 h light/12 h dark cycle. During this period fish were fed ad libitum on chopped liver.

Fish were stunned by a blow to the head and the tail severed behind the caudal fin. Blood was collected by inserting the needle of a heparinized syringe directly into the caudal vessels, thereby reducing contamination from other body fluids. A limitation of this sampling technique is that measured blood pH may not represent true resting in vivo values; however, to minimize such changes fish were captured, stunned and bled within 45 s.

Haemolysates were prepared by centrifuging fresh whole blood for 5 min at 3000 g and 4 °C. The packed erythrocytes were washed 3 times with ice-cold 150 mm-NaCl. Bis-tris or tris-HCl buffers were added to the packed cells to give a final molarity of 50 mm. The suspension was sonicated at o °C followed by centrifugation for 10 min at 4000 g and 4 °C. Small molecules such as organic phosphates were stripped from the haemolysate solution using the ultrafiltration method of Jelkman & Baeur (1976).

Haematocrits were determined by the micro-method using an IEC model MB macro-capillary centrifuge. Haemoglobin concentration was determined by the spectrophotometric method of Van Kampen & Zijlstra (1965). Methaemoglobin was determined by the method of Evelyn & Malloy (1938). The oxygen-carrying capacity of whole blood was calculated using the method of Dijkhuizen et al. (1977).

The concentrations of total nucleoside triphosphates (NTP), and 2,3-diphospho-glycerate (2,3-DPG) were determined enzymically using test kits (Boehringer Mannheim). Specific nucleotides were separated and quantitated by column chromatography as described by Isaaks et al. (1976) and Bartlett (1978). Column fractions were analysed by absorbance at 260 nm and assayed for total phosphate using the method of Ames & Dubin (1960).

pH measurements were made with a Radiometer blood pH electrode at 20 °C coupled to an acid-base analyser (Radiometer PHM Mk 11). Intraerythrocytic pH values were determined using the method of Enoki et al. (1972). Similar values were obtained using the alternative procedure of Murphy et al. (1977).

Oxygen equilibrium curves of whole blood and concentrated haemoglobin solutions were determined at 20 °C by the mixing technique described by Edwards & Martin (1966) after Haab, Piiper & Rahn (1960). This technique involved mixing known volumes of fully oxygenated and fully deoxygenated blood in varying proportions using a specially calibrated 1 ·0 ml gas-tight syringe, then measuring the partial pressure of oxygen. Partial pressures of oxygen were determined with a Radiometer oxygen electrode (E 5046) housed in a Radiometer thermostated cell (D 616). The electrode was coupled to a Radiometer acid-base analyser (PHM 71 Mk n). The method was verified using human blood as a standard at a of 40 mmHg and 38 °C. The whole curve was determined and the P₅₀ value of 26 · 5 ± 0·5 and n value of 2·98 are in good agreement with the published data of Dill & Forbes (1941) and Astrup et al. (1965). Blood and haemoglobin solutions were equilibrated for 10–15 min in 25 ml temperature-controlled tonometers with humidified gas mixtures comprising 3·95 mmHg in air and in nitrogen. Gases were supplied and analysed by C.I.G., Victoria.

The Root effect of blackfish haemoglobin was determined by the spectrophotometric method described by Weber & De Wilde (1975).

The relative proportions of red and white muscle along the body of the blackfish were measured by the method of Mosse & Hudson (1977).

Maximum activities of the enzymes hexokinase, phosphorylase, phosphofructokinase and lactate dehydrogenase in blackfish white muscle were determined as described by Baldwin & Seymour (1977). The lactate dehydrogenase isoenzymes in muscle homogenates were separated by electrophoresis and quantified as described by Muller & Baldwin (1978). The subunit compositions of these isoenzymes were determined immunochemically using antisera prepared against the M₄ and H₄ lactate dehydrogenase enzymes purified from Sardinops neopilchardus (Holmes & Scopes, 1974).

The concentrations of lactate in blood and muscle were measured enzymically using test kits (Boehringer Mannheim). Muscle samples were prepared for analysis using standard procedures described in Bergmeyer (1974). Blood lactates were determined during 20% trichloroacetic acid extracts of whole blood (5:1 vol/vol).

The oxygen equilibrium curve for blackfish whole blood is plotted in Fig. 1, and the oxygen-binding properties are summarized in Table 1 together with published data for other freshwater fishes. The P₅₀ values listed in Table 1 show that the blackfish whole blood curve is positioned to the right of the oxygen equilibrium curves reported for other fishes.

Oxygen equilibrium curves for blackfish whole blood, sonicated whole blood (cells lysed and haemoglobin released into plasma) and a solution of stripped haemoglobin adjusted to the intraerythrocytic pH of 7 ·0, are shown in Fig. 1. Haemolysis has a marked effect on the position and shape of the whole blood curve, with the P₅₀ decreasing from 37·5 to 7·0mmHg, and Hill coefficient, n, decreasing from 1·58 to 1 · 20. The position of the whole blood oxygen equilibrium curve is not restored for a solution of stripped haemoglobin adjusted to intraerythrocytic pH value of 7·0(P₅₀ is 18 · 0 mmHg, n is 1·60).

Enzymic analysis of red cell extracts from blackfish showed high concentrations of nucleoside triphosphates (Table 3). The mean concentration was 6 · 43 μmole NTP/ml packed erythrocytes, giving a molar ratio of total NTP:Hb₄ of 1·68. The major organic phosphate in erythrocytes of most mammals, 2,3-DPG, was not detected. Ion-exchange chromatography revealed that the total NTP comprised ATP and GTP in the ratio 5:1. Total phosphate analysis showed that the ATP and GTP recovered from the column accounted for the total NTP measured in whole blood (Fig. 2, Table 3).

The effects of ATP and GTP on the oxygen affinity (P₅₀) of stripped haemoglobin solutions are shown in Fig. 3. Increasing concentrations of both ATP and GTP markedly decrease oxygen affinity (increase P₅₀), with the maximum decrease occurring at molar ratios between 1·0 and 2·0. For a given molar ratio, GTP is more effective in reducing the oxygen affinity of blackfish haemoglobin.

An in vitro system designed to mimic the whole blood curve at 20 °C was constructed by adding physiological levels of ATP to a solution of stripped haemoglobin at intraerythrocytic pH (Fig. 4). Following the addition of ATP, the oxygen affinity decreased from a P₅₀ value of 18·0 to 34·0 mmHg, a value which is similar to that obtained for whole blood. The Hill coefficient, n, remained essentially unchanged at 1·60.

The mean partial pressure of carbon dioxide in blood taken from the caudal vessels was 3·5 ± 0·5 (Table 2).

The effects of increasing the partial pressure of carbon dioxide on the oxygen affinity of whole blood, and haemoglobin solutions, are shown in Fig. 5. Increasing ⁠, over the range 3·95 to 22·4 mmHg decreases the oxygen affinity of whole blood, with P₅₀ values increasing from 17·0 to 65·0 mmHg. The CO₂-Bohr coefficient (Δ log P₅₀/Δ pH) over this range was —0·625 (Fig. 6).In marked contrast, the oxygen affinity of strongly pH-buffered haemoglobin solutions is insensitive to changes in carbon dioxide tension, both in the presence and in the absence of 100 mm sodium chloride and ATP (Fig. 5, Table 4).

In 20 °C-acclimated blackfish, the mean intraerythrocytic pH of 7·05 is considerably lower than the value of 7·52 obtained for whole blood (Table 2). The effects of pH on oxygen affinity and n values for haemoglobin solutions in the presence and absence of ATP are shown in Figs. 6 and 7. Blackfish haemoglobin is very sensitive to pH over the range 6·5–7·0, with decreasing pH leading to a marked increase in P₅₀. The alkaline Bohr coefficient in this case is —1·05 between pH 6·5 and 7·0, a range that probably represents intraerythrocytic values (see Table 2, Fig. 10). Accompanying this large decrease in oxygen affinity at low pH is an increase in Hill coefficient, n, which changes from 1·6 at pH 7·0 to 2·5 at pH 6·5, as the oxygen equilibrium curve becomes increasingly sigmoidal.

When blackfish were exercised to exhaustion for 1 min, whole blood pH fell, reaching a minimum of 7·17 after 30 min recovery (Fig. 10). The CO₂-Bohr effect for whole blood (Fig. 5) shows that this decrease in blood pH would result in a 1·5-fold reduction in oxygen affinity.

The effect of pH on the total amount of oxygen bound by solutions of blackfish haemoglobin (Root effect) is shown in Fig. 8. The Root effect is small, as haemoglobin equilibrated with air unloads only about 10% of its total bound oxygen when the pH falls from 7·5 to 6·0. Analysis of swimbladder gas revealed a partial pressure of oxygen of 143 mmHg, a value similar to environmental tensions.

The relative proportions of red and white muscle masses in the body musculature of the blackfish, and a number of other fishes of differing swimming behaviour, are shown in Fig. 9. In blackfish, the red muscle is only a minor component accounting for less than 1 % of the total propulsive body musculature.

The maximum activities of the glycolytic enzymes hexokinase, phosphorylase, phosphofructokinase and lactate dehydrogenase in blackfish white muscle are shown in Table 5. The lactate dehydrogenase activity was not significantly inhibited by high pyruvate concentrations up to 10 mm. Electrophoretic and immunochemical analysis revealed that the M₄ isoenzyme accounted for 92 % of the total lactate dehydrogenase activity present in the tissue.

The concentrations of lactate in whole blood and in skeletal muscle of blackfish at rest, and at various times following exercise, are shown in Fig. 11. Muscle lactate increased in a linear fashion throughout the 3 h recovery period, with values rising from 14 to 40 μmole lactate/g wet weight muscle. Whole blood lactate showed quite a different response to that observed for muscle. During the first 30 min of recovery

blood lactate increased from 0·7 to a maximum of 4·6 μmole lactate/ml whole blood, then decreased to a value of about 3·0 μmole/ml and remained at this level during the remainder of the 3 h recovery period, despite the steady rise in muscle lactate.

The oxygen equilibrium curves in Fig. 1 clearly demonstrate the importance of intraerythrocytic factors in reducing the high intrinsic oxygen affinity of blackfish haemoglobin into the physiologically relevant range observed for whole blood. The major factors known to regulate haemoglobin function in vertebrates are organic phosphates, pH and CO₂, and the relative importance of these factors in modulating haemoglobin function in blackfish whole blood, are considered in the following discussion.

A number of studies have demonstrated the importance of organic phosphates as regulators of haemoglobin oxygen affinity in fish erythrocytes. Evidence for their role is threefold: the presence within the erythrocyte of high concentrations of organic phosphate compounds, often exceeding the molar concentration of haemoglobin (Bartlett, 1976, 1978, 1980); the ability of physiological concentrations of these naturally occurring phosphates to lower haemoglobin oxygen affinity in vitro, in a manner similar to that displayed by 2,3-DPG with human blood (Gillen & Riggs, 1971; Weber & De Wilde, 1975); and finally, correlations between changes in the levels of these organic phosphates and changes in whole blood oxygen affinity in fish subjected to hypoxia (Wood & Johansen, 1972; Weber, Lykkeboe & Johansen, 1975, 1976; Weber & Lykkeboe, 1978; Greaney & Powers, 1978; Wood, 1980), and thermal acclimation (Powers, 1974; Wood, 1980; Dobson & Baldwin, 1982).

The presence of high concentrations of ATP and lower concentrations of GTP in blackfish erythrocytes (Table 3, Fig. 2), together with their pronounced effect on the oxygen affinity of haemoglobin solutions at physiological molar ratios and intraerythrocytic pH (Fig. 4), indicates that these organic phosphates are the major regulators of oxygen affinity in blackfish blood.

The ability of increased ⁠, to lower blood oxygen affinity may be elicited through two separate mechanisms; firstly, the effect of CO₂ to reduce intraerythrocytic pH via the carbonic acid system, giving rise to the Bohr effect, and secondly, the direct combination of CO₂ with haemoglobin as bound carbamate, which lowers the affinity of haemoglobin for oxygen (Henriques, 1928; Kilmartin & Rossi-Bernardi, 1973).

While increasing ⁠, markedly lowers the oxygen affinity of blackfish whole blood, it does so only through the effect of pH, as no such change in oxygen affinity was observed with strongly pH-buffered haemoglobin solutions. To our knowledge, this lack of oxygen-linked CO₂ binding has not been reported for any other naturally occurring haemoglobins, that is haemoglobins that have not been chemically modified. In mammals, CO₂ binds preferentially with the terminal a-amino groups of the β-chain of deoxyhaemoglobin at physiological pH (Rossi-Bernardi & Roughton, 1967; Kilmartin & Rossi-Bernardi, 1969, 1973; Arnone, 1972; Perella, Bresciani & Rossi-Bernardi, 1975; Bauer & Kurtz, 1977). Since the a-terminal amino groups of the β-chains also bind organic phosphates (Bauer, 1969, 1970; Perutz, 1970; Arnone, 1972), competition for a common binding site explains the antagonistic effect of CO₂ on the binding of 2,3-DPG (Bunn & Briehl, 1970). Similar competition between CO₂ and organic phosphates has been reported for carp haemoglobin (Weber & Lykkeboe, 1978). However, competition studies with blackfish haemoglobin clearly demonstrate that ATP binding is insensitive to CO₂ (Table 4). The lack of oxygen-linked CO₂ binding site(s) on blackfish haemoglobin indicates not only that the terminal a-amino groups of the α-chains are blocked, as occurs with a number of other teleost fish haemoglobins (Riggs, 1970; Weber & Lykkeboe, 1978), but also that the terminal α-amino groups of the β-chain must be blocked or modified to prevent CO₂ binding.

The physiological significance of direct oxygen-linked CO₂ binding to most vertebrate haemoglobins is not well understood (Kilmartin, 1976; Farmer, 1979). As waterbreathing fish have very low circulating CO₂ tensions compared with air-breathers (Rahn, 1966), carbamate formation can make little contribution to oxygen or CO₂ transport. The absence of oxygen-linked CO₂ binding with blackfish haemoglobin at physiological pH supports this view.

Blackfish whole blood and haemoglobin solutions are very sensitive to changes in pH, with a decreasing pH leading to a reduction in oxygen affinity. The functional significance of the alkaline Bohr effect is thought to lie in enhancing oxygen delivery to the tissues at times of greatest oxygen demand. During muscle work, acidic metabolites will wash out into the blood and lower pH, and the subsequent uptake of hydrogen ions by haemoglobin causes the release of more bound oxygen than would occur merely in response to the lowering of tissue PO₂ during activity.

The freshwater blackfish is largely a bottom-dwelling animal frequenting snags and fallen timber in the slower sections of well-oxygenated streams in south-eastern Australia. General body shape, and muscle morphology and biochemistry are consistent with the observed swimming behaviour in nature and laboratory aquaria. The elongated rounded body tapering gradually to the caudal peduncle (Fig. 9) is characteristic of short-burst performers (Webb, 1978) and this, together with the high percentage of white muscle (99%), suggests predator-prey interactions involving shortterm bursts of high-speed swimming, rather than prolonged periods of steady-state cruising. In the laboratory, fish can be easily exhausted in about 1 min of continuous stimulation, and this lack of endurance is reflected in their poor fighting abilities when captured by rod and line.

The high activities of phosphorylase relative to hexokinase indicate that the white muscle utilizes anaerobic glycolysis with glycogen as the major fuel. The predominance of the M₄ lactate dehydrogenase isoenzyme, which is not significantly inhibited at high pyruvate concentrations, also suggests that the muscle is set up to channel glycolytically derived pyruvate to lactate during bursts of anaerobic work. The maximum activities of phosphorylase and phosphofructokinase are similar and provide for a maximum glycolytic flow of about 26 μmole lactate/min g wet weight muscle. This value is fourfold lower than those calculated from phosphofructokinase activities in white muscle of rainbow trout and dogfish (Crabtree & Newsholme, 1975), and suggests that although blackfish rely upon anaerobic glycolysis during rapid swimming, their anaerobic capabilities are rather limited.

In general, the oxygen-binding properties of fish blood reflect oxygen availability, with high oxygen affinity found in species that encounter hypoxic environments, and lower oxygen affinities in species inhabiting well-oxygenated waters (Powers et al. 1979; Riggs, 1979; Powers, 1980). The low-affinity blood of blackfish compared with other fish bloods (Table 1), suggests possible disadvantages during oxygen loading at the gills. However, this condition is presumably circumvented by the preferred habitat of the animal, a highly oxygenated environment with oxygen tensions between 130 and 140 mmHg (G. P. Dobson, unpublished data). Indeed, this low-affinity blood probably places severe limitations on the ability of blackfish to frequent and survive in waters low in oxygen without major physiological adjustments.

A second important feature of the whole blood oxygen equilibrium curve is the shape, which determines the functional range of oxygen tensions over which blood can unload oxygen to the tissues. Active fish utilizing sustained aerobic muscle work, like the salmonid species, generally display a distinctly sigmoidal curve with n values approaching 2·5–3·0, while less active species like the cyprinids are characterized by a more hyperbolic-shaped curve with lower n values. The hyperbolic curve with an n value of 1·58 obtained for blackfish at 20 °C further justifies placing this animal among the less active species of fish.

Muscle activity associated with swimming can have a profound effect on the oxygenbinding properties of blood by increasing the concentration of hydrogen ions inside the red cell as a result of the hydration of CO₂ and the influx of hydrogen ions from the plasma following washout from those tissues undergoing anaerobiosis. Exercising blackfish to complete exhaustion has the immediate effect of lowering blood pH (Fig. 10), with the maximum decrease occurring after 30 min recovery. This decrease can be correlated with increased muscle anaerobiosis (Fig. 11), but it should be stressed that the initial decrease in blood pH following activity is probably due to the rapid rise in ⁠, (Piiper, Meyer & Drees, 1972; Wood, McMahon & McDonald, 1977).

An interesting feature of the exercise experiment is that changes in blood lactate do not reflect muscle lactate concentrations. The accumulation and retention of high concentrations of muscle lactate, relative to blood, following exercise (Fig. 11), appears to be a general property of fish white muscle (Black et al. 1962; Driedzic & Hochachka, 1978; Wardle, 1978). This phenomenon is thought to reflect the unfavourable transfer conditions resulting from the low capillary density and low perfusion rates of white muscle. The continued increase in muscle lactate concentrations for at least 3 h after exercise suggests that blackfish continues to utilize anaerobic glycolysis, rather than aerobic metabolism, to meet the energy demands of the muscle during the recovery phase. The possibility exists that the continuation of anaerobic glycolysis after exercise may be important to maintain cellular transport processes and assist in the regeneration of creatine phosphate stores in white muscle. The time it takes to switch from predominantly anaerobic to predominantly aerobic metabolism in blackfish in recovery is not known. Similarly, the retention of high lactate concentrations and the eventual fate of lactate in fish white muscle following activity is not well understood.

We conclude that the oxygen-binding properties of blackfish whole blood are not adapted to supply large amounts of oxygen to sustain high levels of aerobic metabolism but, rather, that they are adjusted to maintain adequate tissue oxygenation for slow manoeuvring aerobic swimming in waters of high oxygen tension. The large Bohr effect may be an important adaptation to maximize oxygen delivery to oxygendependent organs and tissues following short-term bursts of high-speed swimming associated with the capture of prey and avoiding predators.

The authors thank Dr Peter Jackson for assisting in obtaining blackfish, and the Victoria Fisheries and Wildlife Division for providing financial support during this study. We also thank Drs P. W. Hochachka, D. J. Randall and S. Perry for their valuable comments and criticism on the manuscript.