Haemoglobin-Oxygen Binding Properties in the Blood of Xenopus Laevis, With Special Reference to the Influences of Aestivation and of Temperature and Salinity Acclimation

Xenopus laevis is neither a frog nor a toad, but is one of the ancient Aglossans (tongueless anurans). Xenopus is completely aquatic and lives in ponds and small lakes at physical conditions (i.e. temperature, CO₂ and O₂ tensions and dissolved solutes) that vary widely both diurnally and seasonally. During droughts Xenopus may escape dehydration and death by burrowing into the substrate and entering aestivation until the rainy season returns (Deuchar, 1975).

Xenopus, unlike other amphibians, does not change from arnmonotelism to ureotelism during metamorphosis, which accords with its purely aquatic habitat (Cragg,

Balinsky & Baldwin, 1961). However, adults become increasingly ureotelic under arid conditions or on exposure to saline water (Balinsky, Cragg & Baldwin, 1961; Ireland, 1973; Funkhouser & Goldstein, 1973; Romspert, 1976; Hillman, 1978; McBean & Goldstein, 1970). During aestivation high concentrations of urea accumulate in the blood of Xenopus (Balinsky et al. 1967) as is the case with dormant spadefoot toads, Scaphiopus (Shoemaker, McClannahan & Ruibal, 1969) and with aestivating lungfish (Jansens & Cohen, 1968).

Xenopus is exposed to wide seasonal temperature variations. The effects of temperature acclimation on aerobic metabolism in poikilothermic vertebrates are well known, but only a few investigations on the effect of temperature acclimation on O₂-binding properties of the blood have been carried out on amphibians. Oxygen equilibrium curves of blood from cold-acclimated (15 °C) Rana esculenta are shifted slightly to the right of those of warm-acclimated (25 °C) specimens, when measured at either temperature (Gahlenbeck & Bartels, 1968). In a fish and a reptile, a similar change in 0₂ affinity following temperature acclimation has been ascribed to changes in the concentrations of intra-erythrocytic cofactors (Grigg, 1969, 1972; Wood et al. 1978).

This paper reports on 0₂ affinity of blood from Xenopus laevis, and concerns the possible adaptive changes in the blood respiratory properties during aestivation and acclimation to altered salinity and temperature. Mechanisms responsible for the changes in the 0₂ binding properties of the blood were also studied in experiments on the influences of temperature, pH, urea, ATP, C0₂ and inorganic ions on the oxygenation properties of the stripped (cofactor-free) haemoglobin in solution.

Specimens of Xenopus laevis, of both sexes and of various size (most weighing 50–100 g), were obtained from a South African animal supplier (Thomas Cook, Johannesburg). The animals were kept at 25 °C, in groups of ten, in 60 1 plastic containers with circulating, aerated fresh water for at least one month before experiments. They were fed chopped liver weekly.

Animals were first transferred to 2%o salt water (Tropic Marine Export, Fulda) at 25 °C. Changing the water weekly, the salinity was then increased in 0·5–1 ‰ increments to 12 ‰ over a 3-month period. Animals were acclimated to this salinity for 2 months before measurements were performed.

Animals that were pre-conditioned to 25 °C were used. Their water was changed weekly while the temperature was lowered 2 °C each time until 10 °C was attained. At 10 °C cold acclimation was continued for an additional two month period.

Aestivation was induced by placing animals in glass cylinders (60 cm high, 20 cm diameter) half filled with mud and containing about 10 cm supernatant water. An infra-red lamp placed about 50 cm above the cylinders caused gradual increase in dehydration. The animals burrowed until they were completely covered with mud. In preliminary experiments the animals died when the mud became dry and hard. The mud was therefore kept moist during aestivation. The data reported here are based on results from two specimens that survived for two months in the aestivating condition.

Blood was sampled into heparinized syringes from the ventricle after exposing the heart. Sampling was carried out as rapidly as possible (within 5 min) to minimize acidification due to handling stress. With rapid sampling the lactate concentration was less than 5 mm.

Immediately following sampling the blood samples were subdivided and subjected to the following procedures. Blood pH was measured in duplicate by means of Radiometer equipment BMS-2 MK-2 connected to a PHM 64 using precision buffers for calibration. Haematocrit values were determined in duplicate by centrifugation in glass capillaries (3 min at 12000 rev/min). Plasma and blood samples were frozen at –18 °C for urea assay. Proteins were precipitated (as described below) in preparation for determination of blood lactate and red cell phosphates and the protein-free preparations were frozen at – 18 °C.

Haemoglobin contents in whole blood (quoted in Table 1) were calculated from O₂-capacity measurements, obtained using a Tucker chamber (below). In experiments on haemolysates, the concentrations were determined spectrophotometrically with a Unicam SP 1800 spectrophotometer after diluting 10 μl samples in 1 ml water. Pigment concentrations were calculated from the millimolar extinction coefficients for oxygenated Hb of 14·6 and 13·8 at 577 and 541 nm, respectively (Antonini & Brunori, 1971).

Nucleoside triphosphate (NTP) concentrations were assayed by thin layer chromatography (Johansen et al. 1976). Prior to measurement 50 μl blood samples were deproteinized by addition of an equal volume of 12% trichloracetic acid (TCA). Concentrations inside the erythrocytes were calculated from the haematocrit values assuming that all NTP is erythrocytic.

Lactate was measured in 100/d blood samples deproteinized with 200μl 0·6 N perchloric acid (PCA) according to the lactate dehydrogenase enzymic method (Boehringer-Mannheim).

The osmotic concentrations in blood, plasma and acclimation water were determined with a Knauer semi-micro osmometer, type M.

Intra-erythrocytic pH was measured in 70 μl blood samples which had been equilibrated for 20 min in a Radiometer BMS-2 MK-2 with mixtures of CO₂ (1–5%) and atmospheric air delivered by Wösthoff gas mixing pumps. After blood pH measurement, the samples were returned from the pH electrode into the tonometers. The blood was then layered below paraffin oil contained in 1·5 ml Eppendorf tubes using a haematocrit suction adapter and capillary tubes that had been rinsed in the equilibration gas mixture. The plasma and red cells were subsequently separated in an Eppendorf 3200 microcentrifuge, and pH values in the plasma (drawn from under the paraffin oil into the BMS pH electrode) were measured. The pH values of blood and of plasma were not significantly different. After removal of the plasma the red blood cells were lysed by three times freezing (immersing the tubes in alcohol at –18 °C) and thawing, and the cellular pH was then measured.

Whole blood oxygen equilibria were measured by equilibrating 50μl samples for 20 min periods to different gas tensions and measuring O₂ content using a Tucker chamber (Tucker, 1967) modified after Bridges et al. (1979). O₂ capacity (i.e. 100% saturation) was measured in samples equilibrated with a gas mixture having a p_o of 650 mmHg. Haemoglobin-bound oxygen was obtained by subtraction of dissolved oxygen calculated from the Bunsen O₂ solubility coefficients (Christoforides & Hedley-Whyte, 1969).

Red cells were washed three times in 0·9 % NaCl and lysed in thrice their volume of 0·01 M Tris HC1 buffer at pH 8. The haemoglobin was stripped of cofactors and other ions, firstly by passage through mixed ion exchange resin (Amberlite MB-3) and subsequently by chromatography on a 30 cm column of Sephadex G-25 superfine gel eluting with 0·05 M Tris buffer pH 7·5 containing 0·1 M-NaCl (Berman, Benesch & Benesch, 1971). Curiously, the haemoglobin denatured when Sephadex stripping preceded the MB-3 treatment, and the absorption peak near 577 nm decreased strongly with CO, so CO treatment was avoided. The haemolysate was dialysed against four changes of 0·01 M Tris buffer pH 7·5, containing 5 × 10⁻⁴ M EDTA for 48 h. The haemoglobin solutions were concentrated on Amicon B15 concentrators (Oosterhout, Holland) to a tetrameric concentration of about 0·4 mm. (Control experiments showed that 0₂ affinity of the stripped haemoglobin was virtually independent of concentration in the range of 0·15–0·7 mm tetramer.)

To determine heterogeneity of the haemoglobin and to isolate the component haemoglobins, preparative isoelectric focussing was carried out at +4 °C in no ml columns using ampholytes of pH ranges 5–8 (0·6%) and 3–10 (0·2%) (LKB, Sweden).

All preparative procedures were carried out at 0–5 °C.

Oxygen equilibria of haemoglobin solutions were measured at 436 nm using a modified gas diffusion chamber technique (Sick & Gersonde, 1969; Weber, Lykkeboe & Johansen, 1976). When measuring the effect of C0₂ on 0₂ affinity, separate samples of the haemoglobin solutions were equilibrated to the same gas mixtures in a BMS-2 for pH measurement. The effects of ATP and urea were investigated by addition of stock solutions assayed with Boehringer test chemicals. The urea solutions used were first deionized with mixed ion-exchange resin. In view of the specific effect of chloride on oxygen affinity and its interactions with organic phosphates, care was taken to keep the chloride concentration constant when studying effects of other parameters. The effect of the anion (Cl⁻) and/or cations (Na⁺, Mg²⁺) were investigated by addition of Cl⁻ assayed stock solutions.

Oxygen equilibrium curves of blood of Xenopus acclimated to 10 °C and 25 °C and measured at both temperatures are given in Fig. 1. For 25 °C acclimated animals the P₅₀ values at pH 7· 6 are 27·0 and 13·7 mmHg (3·60 and 1-83 kPa) at 25 and 10 °C, respectively. Oxygen capacity averaged about 15 vol%, which is unusually high compared to other amphibians (Wood et al. 1975).

It is seen that acclimation to lower temperature (10 °C) decreases O₂ affinity; P₅₀ Values increased to 31·3 and 15·6 mm (4·18 and 2·08 kPa) at 25 and 10 °C. It follows that the effect of temperature is greater in animals subjected to acute temperature changes than in those acclimated to different environmental temperatures. The Bohr factors were –0·39 and – 0·47 when measured at 25 and 10 °C, respectively (Fig. 2).

Interpolation from Fig. 2 at pH 7·6 shows that 0₂ affinity of Xenopus blood has low temperature sensitivity; the ΔH values, calculated from the Van’t Hoff equation as – 2·303.R.Δlog P₅₀ΔT⁻¹, where T is the absolute temperature and R is the gas constant (Wyman, 1964), were –7·8 and – 8·0 kcal.mol^-1 for warm and cold-acclimated animals, respectively. In 10 °C animals, blood pH and red cell urea concentrations were significantly higher than those at 25 °C (Table 1). Intraerythrocytic pH values were less than 0·2 units lower than those in plasma (Fig. 3).

Salt water acclimation as well as aestivation resulted in higher 0₂ affinities than in active, freshwater specimens (Fig. 4, 5). At pH 7·6 P₆₀ values were 21·1 and 25·2 mm Hg, (2·81 and 3·36 kPa) respectively. Also, salt water acclimation and aestivation significantly raised the levels of urea in red cells, blood and plasma (Table 1 and Fig. 4)

Higher salinity and aestivation lowered the haemotocrit values. In salt water, blood chloride increased about 20% (Table 1), while body weights decreased by 20–25% during aestivation and salinity acclimation. As expected, the blood and plasma osmotic concentrations were essentially identical (equivalent to 12·6 and 12·4 g/1 NaCl, respectively), whereas the Cl⁻ levels were about 16% higher in the plasma than in whole blood (equivalent to 6·6 and 5·7 g/1 NaCl, respectively) Non-chloride compounds were therefore correspondingly higher in blood and in the cells than in plasma (equivalent to 6·9 and 5·8 g/1 NaCl, respectively).

There is no evidence for a significant effect of aestivation and salinity acclimation on blood haemoglobin concentrations (range, 1·5-1·9 mm tetramer.1⁻¹ (B)) nor on the ATP levels in blood and red cells. The relationship between 0₂ affinity and pH in stripped haemoglobin in the absence and presence of urea and NaCl, is shown in Fig. 6. At 10 and 25 °C the Bohr factor, ϕ= Δ log P₅₀/ΔpH, was –0·7 to –0·40. The inverse relation between measurement temperature and 0₂ affinity (Fig. 6) reflects a AH value of –5·6 kcal.mol⁻¹ for the stripped pigment. Urea has a strong positive effect on 0₂ affinity at all pH values studied, while NaCl decreases 0₂ affinity. Urea and NaCl, however, exert no significant effect on the cooperativity coefficient, n (Fig. 6).

With the view of investigating their possible roles in the acclimation, the effects of increasing concentrations of urea and chloride were studied (Fig. 7). Raising the urea concentration (Fig. 7 A) from o to 300 mm decreases P₅₀ from about 10 to 4·5 mm Hg (1·33–6·00 kPa) at pH 7·65. Cl⁻ decreases the 0₂ affinity of in vitro haemoglobin preparations most strongly at low concentrations (Fig. 7 B) as found in the red cells (cf. Table 1); at pH 7·6 an increase from 0·05 to 0·5 M raises P₅₀ from about 10·21 mm (1·33–2·80 kPa). Equimolar concentrations of Cl⁻ added as NaCl and MgCl₂ respectively, resulted in similar P₅₀ values, suggesting that the valency of the cation has no distinctive effect on 0₂ affinity.

The organic phosphate, ATP, has a comparatively small effect on 0₂ affinity; at pH 7·55, P₅₀ is raised from 10·7 to only 13·4 mmHg (1·43–1·79 kPa) as the molar ratio of ATP to haemoglobin is increased from 0–3, while n increases slightly from 2·4–2·6 (Fig. 7C). Comparative data are available for few other amphibians. The (fossorial apoda, Boulengerula taitanus, showed a similar low ATP sensitivity; at pH 7·1, 50 times excess ATP to Hb increased P₅₀ from 5·0–7·6 mmHg (0-67-1-10 kPa) (Wood et al. 1975). Hb of the anuran, Rana escalenta, lacks significant ATP sensitivity (R. E. Weber & A. Jokumsen, unpublished data). In the viviparous caecilian Typhlonectes compressicauda there is, however, a large ATP effect. At pH 7·1 a 1 mm concentration changed P₅₀ from 1·9–5·6 mmHg (0·25–0·75 kPa) (Garlick et al. 1979).

The influence of C0₂ and ATP, singly and in combination, on 0₂ affinity and cooperativity in Xenopus haemoglobin is shown in Fig. 8. A specific, pH-independent effect of C0₂ is demonstrated, decreasing 0₂ affinity of the stripped haemoglobin. At levels near 7 and 30 mm (0·93 and 4·00 kPa), these effects were manifest only above pH 7·4 and 7·26 respectively. The specific C0₂ effect is markedly decreased in the presence of ATP. This indicates competition between ATP and C0₂ for common amino groups as in mammals and shows that ATP is bound by the haemoglobin despite its low oxygenation effects. It also suggests that the specific effect of C0₂ will have limited influence on the in vivo P₅₀. A C0₂ effect, however, persists in the presence of ATP when pH is high (above about 7·4) reflecting increased carbamate formation as protonation of the amino groups and competition from the anionic ATP decrease.

Isoelectric focusing resolves Xenopus haemoglobin in one minor and one major component (I and II in Fig. 9), which are isoelectric near pH 9·3 and 7·0, respectively, and account for about 3·9 and 96·% of total haemoglobin. The major component has approximately similar oxygen affinities and n values as the haemolysate (Fig. 10). ATP has the strongest effect on 0₂ affinity at lower pH, as expected from a greater protonation of the binding groups on the molecule under these conditions. Remarkably, Hb I has a significantly higher oxygen affinity, lower cooperativity and a smaller Bohr factor than Hb II; at pH 7·6, P₅₀ and n are 4·3 mmHg (0·57 kPa) and 1·0 respectively, and ϕ =– 0·11. Only a small part of this difference may be attributable to the lower concentration in Hb I (see p. 22).

The O₂ affinity (P₅₀ = 27 mmHg or 3·60 kPa at 25 °C and pH 7·6) and the Bohr factor (ϕ, –0·39 to –0·47) of Xenopus blood are higher than those found in most other amphibians, where P₅₀ values range from 18–50 mm (2·4–6·67 kPa) and ϕ from –0·13 to –0·29; the O₂ affinity is particularly high compared to values for other anurans (cf. Wood et al. 1975). These properties may be adaptive to the aquatic existence. The high affinity will favour cutaneous O₂ uptake, whereas the high pH sensitivity and the specific CO₂ effect will facilitate O₂ unloading in the tissues as rises during diving (cf. Emilio & Shelton, 1974). A comparison of the Australian and South American lungfish shows a decrease in Bohr factor with increasing dependence on air as a respiratory medium (Lenfant, Johansen & Grigg, 1966/7; Johansen & Lenfant, 1967), whereas a similar relation between O₂ affinity and air dependence was found in a series of amphibians studied by Lenfant & Johansen (1967).

Our data show that the O₂ affinity of blood of Xenopus is subject to environmental modulation; O₂ affinity is increased following acclimation to high temperature and saline water, and during aestivation. ATP, the main allosteric modulator of blood O₂ affinity in lower vertebrates, however, does not seem to play a major role, in view of its small direct effect on the O₂ affinity of the haemoglobin and the small changes in its concentration. The latter, moreover, excludes a significant indirect effect of ATP via changes in intracellular pH coupled with its influence on the Donnan equilibrium of protons across the red cell membranes (Wood & Johansen, 1973).

The low values for the overall heat of oxygenation of the whole blood and the cofactor-free haemoglobin of Xenopus (ΔH, –7·9 and –5·6kcal.mol⁻¹, respectively, compared to –10 to – 14 in most other vertebrates) suggest relatively constant O₂ affinities in the face of varying external temperatures. In acclimated animals the temperature sensitivity is even lower than the acute in vivo effect; measured at the Acclimation temperature, ΔH approximates –6·3 kcal.mol^-1 (Fig. 1 inset; dashed line). These results agree qualitatively with those found for Rana escalenta by Kirberger (1952), Straub (1957) and Gahlenbeck & Bartels (1968).

In air-breathing vertebrates with an efficient pulmonary gas exchange, the arterial blood will be essentially saturated with O₂, so a decrease in O₂ affinity is the only way in which tissue O₂ delivery can be enhanced through affinity modulation. A high temperature sensitivity will normally thus be favourable, enhancing O₂ unloading as O₂ demand increases at high temperature. Although Xenopus primarily relies on lung breathing, its skin plays an important role in gas exchange. Emilio & Shelton (1974) showed that it may account for 30% of O₂ uptake and 80% of CO₂ elimination, and Jones (1967) found that during diving O₂ uptake is 43 ml.kg⁻¹.h⁻¹, compared to pre-and post-dive rates of 63 and 91 ml.kg⁻¹.h⁻¹, respectively. The low temperature sensitivity of Xenopus haemoglobin may thus be adaptive in the sense that it safeguards O₂ loading in ponds where O₂ tension may be low and temperature high. This property may also favour O₂ uptake when similar conditions prevail in the microenvironment of the aestivating animals.

During air breathing, when pulmocutaneous vasodilation accompanies increased pulmonary blood flow (Shelton, 1970) the O₂ tensions approximate 80 mm in the systematic and 60 mmHg in the pulmocutaneous blood of Xenopus (Emilio & Shelton, 1974) providing evidence for the functional separation between blood flows through the single heart ventricle (De Graaf, 1957). However, pulmocutaneous blood is mixed due to a shunt of blood from the left to the right ventricle cavity (De Graff, 1957). During diving, however, when O₂ tensions fall in all parts of the circulation to about 45 mmHg (Emilio & Shelton, 1974), the above cardiovascular adjustments during lung ventilation are reversed (Shelton, 1970). The functional significance of this regulation of blood flow in conjunction with conveying mixed blood in the pulmocutaneous circulation may be that it allows for a gradual depletion of the lung O₂ stores during submersion or whenever Xenopus experiences hypoxic habitats.

Our data do not permit inferences on the mechanisms underlying temperature-induced changes in O₂ affinity. It is, however, likely that changes in the erythrocytic concentrations of inorganic ions are implicated. Straub (1957) and Gahlenbeck & Bartels (1968) report that whereas the erythrocytic concentration of K⁺ is higher in warm-than in cold-adapted Rana esculenta, the opposite applies to the concentrations of Na⁺, K⁺ and Cl⁻ in the plasma. Thermoacclimatory responses in inorganic erythrocyte ions are also known in fish. Recently (1979) Houston & Smeda showed that with increasing acclimation temperature, the K⁺: haemoglobin ratio in rainbow trout and the Na⁺: haemoglobin and Cl⁻: haemoglobin ratios in carp increased whereas the Mg²⁺: haemoglobin in carp decreased. The inverse effect of NaCl on O₂ affinity of Xenopus haemoglobin corresponds to that generally encountered in vertebrates (Antonini, Amiconi & Brunori, 1972). For intact human erythrocytes Sommer-kamp et al. (1961) have shown an inverse relation between cation content and O₂ affinity.

The increases in O₂ affinity induced by osmotic stress and aestivation were accompanied by significant increases in the urea levels in blood, plasma and red cells. In conjunction with the specific effect of urea on the affinity of Xenopus haemoglobin (Fig. 7 A), this suggests that the changes in blood O₂ affinity result from the increased urea levels. The fact that salts and ATP decrease O₂ affinity (Fig. 7B, C) shows that the higher affinities cannot be attributed to increases in the concentrations of these factors. On the contrary, changes in their concentrations may have reduced the urea effect. The concentrations of urea measured in the aestivating specimens (∼11 mm (P)-Table 1) were, moreover, much lower than those (∼ 30 m-mol.kg⁻¹(P)) Balinsky et al. (1967) found in naturally aestivating specimens. It follows that in nature the possible effects of urea on the functional properties of the blood will be greater than those we could document here.

Seymour (1973) found a similar increase in the O₂ affinity of the blood of the spadefoot toad during dormancy, and related the change (measured under unspecified pH conditions) to an O₂ storage function of the haemoglobin. As with Xenopus, the affinity increase may have been caused by raised blood urea levels.

What may be the biological significance of an increased O₂ affinity during aestivation ? A high O₂ affinity implies that the haemoglobin will only unload when O₂ tension is low. In the absence of compensatory increases in the circulation rate (cf. Coulson, Hernandez & Herbert, 1977), a high affinity could result in a decrease in the rate of O₂ delivery and thus lower O₂ tensions at the mitochondria, which will favour the lower metabolic rates that characterize dormancy. In lungfish, a significant increase in blood O₂ affinity during aestivation is ascribable to a marked fall in the erythrocytic concentration of the cofactor guanosine triphosphate (Johansen et al. 1976). Studies on squirrels and hamsters have correspondingly showed 29–48% decreases in erythrocytic diphosphoglycerates during hibernation (Burlington & Whitten, 1971; Tempel & Musacchia, 1975). The present contention that the increased O₂ affinity is adaptive to lower tissue O₂ tensions rather than to high O₂ saturation of the pigment at the respiratory surfaces is analogous to the suggestion of Manwell (1959) that the high affinities of some invertebrate haemoglobins may protect their tissues from injurious effects of high O₂ tension (O₂ poisoning). A similar mechanism may be basic to the findings (Wood et al. 1975) that blood P₅₀ and ATP values increase when plaice from hypoxic water is subjected to atmospheric O₂ tensions, but then decrease again as increases to hyperoxic levels.

The high sensitivity of O₂ affinity of Xenopus haemoglobin to urea is similar to that of Human Hb (Rossi-Fanelli, Antonini & Caputo, 1961, 1964), but contrasts with the low sensitivity seen in Hb from aestivating lungfish (Weber et al. 1977) and the skate (Bonaventura, Bonaventura & Sullivan, 1974), which in life similarly experience high concentrations of urea. Urea causes dimerization of human haemoglobin (Bonaventura et al. 1974) whereas the quaternary structure of the elasmobranch Hb appears only slightly affected by this compound. This correlates with the different effects of Cl⁻ on O₂ affinity which indicates that the urea insensitivity in the skate results from stronger electrostatic interactions between the Hb subunits. However, Xenopus Hb shows Cl⁻ sensitivity similar to that of human haemoglobin indicating that different molecular adaptations are basic to the tolerance of Xenopus and skate haemoglobins to urea.

The unequal distribution of urea between erythrocytes and the plasma (Fig. 4, inset) in Xenopus is in agreement with that found in human and skate blood (Ralls, 1943; Browning, 1978). The higher red cell urea concentration may be due to binding of urea to the haemoglobin molecule. The cationic properties and the amide structure of urea particularly favour hydrogen binding of urea to the normally dissociated and hence negative terminal carboxyl groups of the subunits of haemoglobin (Krichevskaya, Lukash & Kartasheva, 1973). This suggests that urea may act as a positively charged cofactor undergoing O₂-linked binding of the Xenopus haemoglobin molecule, in analogous fashion to anionic phosphate cofactors from most vertebrate erythrocytes.

We demonstrate a direct pH independence effect of CO₂ on O₂ binding of Xenopus Hb. As HCO₃⁻ is always present in experiments with CO₂, this anion could be contributing to the effect attributed to CO₂. However, the absence of CO₂ and substitution of HCO₃⁻ by acetate, an ion of similar size and charge distribution, resulted in unchanged O₂ affinity in lungfish haemoglobin (Farmer, 1979) indicating that the specific CO₂ effect is due to carbamino CO₂. In human haemoglobin, CO₂ binds at the α and β chain amino termini competing with DGP for the latter sites. Competition between CO₂ and the organic phosphate in Xenopus haemoglobin suggests qualitatively similar interaction to that which occurs in human haemoglobin A, and that carbamino formation at the free a amino terminal groups mediates the CO₂ effect in Xenopus, In conjunction with the low levels of ATP in Xenopus blood, and its small influence on oxygen affinity, this competition also indicates that oxylabile carbamates may contribute significantly to CO₂ transport in Xenopus blood. Farmer (1979) suggested that a CO₂-induced decrease in O₂ affinity in three fishes and one amphibian of only half of that found in human Hb A may be due to blocked a chain amino termini. Acetylation of the a chain amino terminal groups has in fact been reported in some fish and amphibians (Riggs, 1970; De Witt & Ingram, 1967; Sullivan, 1974)

Though most CO₂ in Xenopus is eliminated through the skin, of blood fluctuates (7–13 mmHg) in relation to the breathing rhythm, with lower values during breathing periods and higher ones during immersion (Emilio & Shelton, 1974). During diving CO₂ may thus increase the O₂ unloading pressure, resulting in larger oxygen gradients from the blood capillaries to the cellular metabolic sites.

The higher O₂ affinity found in the haemolysate compared to that in the whole blood may be attributed to various interacting factors, including lower intracellular pH, different concentrations of interacting ions, a specific effect of CO₂ and the lower haemoglobin concentration. Although the haemoglobin concentration effect was insignificant in the stripped haemolysate (p. 22) it may increase drastically at higher pigment concentrations in the presence of cofactor as was recently demonstrated for carp haemoglobin (Lykkeboe & Weber, 1978).

The presence of one minor component, Hb I, and one major component, Hb II, (3·9 and 96·1%, respectively) agrees with what has been found by Maclean & Jurd (1971). Gel-electrophoretic studies of Just, Schwager & Weber (1977) revealed two Hb components of different molecular weights, evidencing dimerization of the minor Hb component. The distinctly different O₂ binding properties of the two fractions is reminiscent of many fish where an electrophoretically cathodal Hb (cf. Hb I) has high O₂ affinity and lacks cooperativity and pH sensitivity, while anodal Hbs (cf. Hb II) have low O₂ affinity and distinct cooperativity and pH sensitivity. In fish (Hashimoto, Yamaguchi & Matsuura, 1960; Powers, 1972; Weber & De Wilde, 1976), these differences may represent a division of Laboure, the cathodal component ensuring O₂ uptake and transport when O₂ tensions or pH is low. Despite its low relative concentration the properties of component I of Xenopus foreshadows a similar role.

Advice criticism and help from Dr Helmut Schrock, Karlsruhe, West Germany and Professor Kjell Johansen, Gunnar Lykkeboe and Winnie Heidemann, Aarhus, Denmark, are gratefully acknowledged. The work was supported by the Danish Natural Science Council.