Allosteric Modulation of Haemocyanin Oxygen-Affinity By L-Lactate and Urate in the Lobster Homarus Vulgaris: I. Specific and Additive Effects on Haemocyanin Oxygen-Affinity

The oxygen affinity of many arthropod haemocyanins can be modulated by allosteric effectors. It has been known for some time that inorganic ions, such as H⁺, Ca²⁺, Mg²⁺ and Cl⁻, affect haemocyanin oxygen-affinity (for a review, see Ellerton et al. 1983). The organic modulator that was first demonstrated to increase the oxygen affinity of Carcinus maenas and Cancer pagurus haemocyanin was L-lactate (Truchot, 1980). This effect of L-lactate on haemocyanin oxygen-affinity has been confirmed for many other species of crustaceans (see table in Bridges and Morris, 1986). Similarly, urate and other purine bases have a potentiating effect on haemocyanin oxygen-affinity (Morris et al. 1985, 1986a).

Dopamine has also been demonstrated to increase the oxygen affinity of Cancer magister haemocyanin in vitro (Morris and McMahon, 1989a, b).

In contrast to their well-documented effect on oxygen affinity, there is still some uncertainty whether allosteric effectors also influence cooperativity of oxygen binding to haemocyanin. A higher cooperativity was measured in the presence of Ca²⁺ for Limulus polyphemus and Callinectes sapidus (Brouwer et al. 1983), but L-lactate was reported to decrease cooperativity in Cancer magister (Graham et al. 1983) and Callinectes sapidus (Johnson et al. 1984; Bridges and Morris, 1986). Other authors have observed no measurable influence on cooperativity in Callinectes sapidus and Homarus vulgaris (Booth et al. 1982; Taylor and Whiteley, 1989). Urate has been demonstrated to have no effect on the cooperativity of oxygen binding in Austropotamobius pallipes (Morris et al. 1985) and Astacus leptodactylus (Bridges, 1990).

These previous studies were carried out in vitro for each modulator separately, leaving all other factors unchanged. Therefore, less information is available concerning the interactions of two or more allosteric modulators. The experimental approach for investigations on simultaneous effects of several effectors includes variation of the concentrations of all components studied. This was done by Morris et al. (1986,b, 1987) for L-lactate and Ca²⁺ in Austropotamobius pallipes-, they found mutual agonistic effects of both ions (Morris et al. 1986,b). Various combinations of L-lactate and urate concentrations present in the haemolymph of Austropotamobius pallipes were able to enhance haemocyanin oxygen-affinity (Morris et al. 1986a; Morris and Bridges, 1986); however, the effects were not additive.

In the present study, the effect of L-lactate and urate on lobster haemocyanin oxygen-affinity is first considered for each modulator individually and then the simultaneous effect of L-lactate and urate is examined. The results are discussed in the light of our present knowledge of modulation of haemocyanin oxygen-affinity.

Five European lobsters (Homarus vulgaris Milne-Edwards) were purchased in Roscoff, Brittany (France). The animals were transported to Düsseldorf and held there at the Institut für Zoologie, Lehrstuhl für Tierphysiologie, in a recirculating artificial seawater aquarium at 15±1°C, pH8.0±0.1 and a salinity of 38± 1%o. Sea salt was obtained from Wiegandt (Krefeld, Germany). Animals were fed regularly with squid muscle tissue and exposed to a 12h:12h light.’dark regime.

Haemolymph samples were taken by piercing the arthrodial membrane at the base of the walking legs and drawing a haemolymph sample into a syringe. A 5 ml sample was taken every 4 weeks from each individual and samples were pooled. To avoid clotting, the haemolymph was filtered through gauze and cellular material was removed by centrifugation at 8000g (Hermle ZK 400, Kontron, Munich, Germany) at 4°C for 10min. To remove endogenous effectors, the haemolymph was dialysed three times against Ringer’s solution whose volume was 40 times the haemolymph volume. The Ringer’s solution was made up as follows: 600 mmol l⁻¹ NaCl, 10 mmol l⁻¹ CaSO₄, 10 mmol l⁻¹ MgSO₄, 5 mmol P¹ KC1, 5 mmol l⁻¹ NaHCO₃, pH7.99±0.03. This reflected the concentrations of inorganic ions in the artificial sea water and those measured in whole haemolymph. Bacterial growth was prevented by adding 0.02% NaN₃ to the Ringer’s solution. The haemocyanin solutions were stored at 4°C for a maximum of 4 weeks.

The concentrations of protein and haemocyanin in the haemolymph were determined by spectrophotometric analysis at wavelengths between 250 and 450nm (Uvikon 810, Kontron, Munich, Germany) according to the method of Nickerson and van Holde (1971). The maximum absorption of haemocyanin was measured at 334nm. Haemocyanin and protein concentrations were calculated using the extinction coefficients reported for Homarus americanas (⁠⁠; Nickerson and van Holde, 1971).

The aggregation state of haemocyanin was determined by gel filtration using a fast protein liquid chromatography (FPLC) system equipped with a Superose-6 column (Pharmacia, Freiburg, Germany). About 2 mg of protein was used for this purpose and eluted with 0.1 ml min⁻¹ of Ringer’s solution (see above). The haemolymph proteins were monitored spectrophotometrically at 280nm. The different aggregation states of haemocyanin were identified by their relative molecular mass using the values reported by Markl et al. (1979) for dodecamers and hexamers of this species. Ovalbumin, aldolase, catalase, ferritin and thyreo-globulin were used for calibration.

L-Lactate concentrations were measured using the method of Gutmann and Wahlefeld (1974), modified by addition of EDTAto the assay (Engel and Jones, 1978). Urate concentration was determined using the HPLC technique according to Wynants et al. (1987). Proteins were removed from the samples by extraction with perchloric acid (0.6 mol l⁻¹), coagulated proteins being sedimented by centrifugation at 8000g at 4°C for 10min (Biofuge A, Heraeus Christ, Osterode, Germany). A 20μl sample of the supernatant was used for chromatography without preceding neutralisation.

Oxygen equilibrium curves were determined spectrophotometrically using a diffusion chamber (Sick and Gersonde, 1969; modified by Bridges et al. 1979). The O2 tensions in a 10μl sample were varied using gas-mixing pumps (303a/f Wösthoff, Bochum, Germany). The pH was altered by changing the CO₂ tension from 0.1 to 0.8kPa. pH values were measured in a corresponding sample equilibrated in a BMS II tonometer (Radiometer, Copenhagen, Denmark) with the same gas mixtures. The pH was determined at the half-saturation point (P₅₀) of each equilibrium curve. The oxygen-dependent changes in haemocyanin absorption were measured at a wavelength of 365 nm using a spectrophotometer (Eppendorf 1101M, Hamburg, Germany). The experimental equipment was thermostatted at 15°C. The P₅₀ and n₅o values were calculated employing regression analysis of the data between 25 and 75% saturation according to the Hill equation (Hill, 1910).

Oxygen binding experiments were carried out with native haemolymph and dialysed haemocyanin solutions. To obtain different L-lactate and urate trations in the haemocyanin samples, 200 μl samples of the dialysed solutions were centrifuged at 100000g for 1 h (Airfuge, Beckman, CA, USA). 100μ1 of the supernatant was replaced by Ringer’s solutions containing defined amounts of L-lactate and/or urate; the haemocyanin pellet was redissolved (replacement technique according to Bridges et al. 1984). The exact L-lactate and urate concentrations of the resulting haemocyanin solutions were measured as described above. The following results are based on 42 oxygen equilibrium curves.

Unless otherwise stated, values given are means±s.D. The data were analysed according to Student’s t-test. When covariance analysis was carried out, the P values are given in the text.

The concentration of haemocyanin in native Homarus vulgaris haemolymph was 37.8±10.3mgml⁻¹, and total protein concentration was 39.6±10.7mgml⁻¹ (N=21). In these measurements, individual haemolymph samples of four animals were taken at different times. L-Lactate concentrations averaged 0.5±0.2 mmol l⁻¹ (N=7) and those of urate 0.08±0.02mmoll⁻¹ (N=7) in undisturbed control animals. In the native haemolymph more than 96% of the haemocyanin molecules were dodecamers and hexamers; at least 80% were dodecamers. The concentration of monomers was negligible. Only solutions with more than 80% dodecameric haemocyanin were used for the experiments.

Oxygen binding curves constructed in the presence of different L-lactate concentrations at pH7.99±0.02 ⁠, are shown in Fig. 1. The P₅₀ in the dialysed haemolymph (0.3 mmol l⁻¹ L-lactate) was 0.89 kPa (6.7 mmHg). Increasing the concentration of L-lactate to 1.5 mmol l⁻¹ resulted in a left shift of the curve, the P₅₀ being 0.79kPa (5.9mmHg). Half-saturation was further lowered to 0.65 kPa (4.9 mmHg) by raising the L-lactate concentration to 4.5 mmol l⁻¹. Augmentation of the L-lactate concentration to 11 mmol l1⁻¹ only resulted in a small further increase in the oxygen affinity, the P₅₀ reaching 0.63 kPa (4.7mmHg). The decrease in P₅₀ as an exponential function of the L-lactate concentration is illustrated in Fig. 2. At low L-lactate concentrations, up to 4.5 mmol l⁻¹, P₅₀is proportional to the added L-lactate concentration. Raising the

L-lactate concentration beyond 4.5 mmol l⁻¹ resulted only in a small increase in oxygen affinity.

A quantitative estimate of the L-lactate effect can be obtained from the quotient (ΔlogP₅₀)(Δlog[L-lactate])^−l, which was −0.11 at pH7.99±0.03 (Table 1B). Whole haemolymph showed a higher oxygen affinity than did dialysed haemolymph at the same L-lactate concentration (Fig. 2): AP50 was 0.16 kPa (1.2 mmHg).

The Bohr effect can be estimated by the ratio (ΔlogP₅₀)(ΔpH)⁻¹, which is derived from the regression lines given in Table 1A for the L-lactate concentrations examined. For L-lactate concentrations between 0.3 mmol l⁻¹ and 4.5 mmol l⁻¹ no significant change in the Bohr coefficient was observed. At 11 mmol l⁻¹ L-lactate, however, the slope of the regression line (ΔlogP₅₀)(ΔpH)⁻¹ was significantly lowered (P<0.01).

Similar investigations were carried out with haemocyanin solutions containing different urate concentrations (Fig. 3) at constant pH. The oxygen affinity of haemocyanin was raised by increasing urate concentrations. The P₅₀ in the dialysed solution was 0.89 kPa (6.7 mmHg), with 0.17 mmol l⁻¹ urate it decreased to 0.72 kPa (5.4 mmHg), with 0.56 mmol l⁻¹ urate to 0.53 kPa (4.0 mmHg). A further increase in the urate concentration to 0.93 mmol l⁻¹ only led to a slight decrease in P50 to 0.52kPa (3.9mmHg). As for L-lactate, P₅₀ is an exponential function of the urate concentration (Fig. 4). The measured urate effect at pH7.99±0.03 according to the quotient (ΔlogP₅₀)(Δlog[urate])^−l was −0.18 (Table 1D). Whole haemolymph showed a higher oxygen affinity than did dialysed haemolymph with the same urate concentrations (Fig. 4): AP_5O=0.10kPa (0.8 mmHg).

The dependence of the oxygen affinity on pH was calculated for a constant urate concentration by linear regression, as in the case of L-lactate, giving the Bohr coefficients listed in Table 1C. Urate concentrations exceeding 0.56 mmol l⁻¹ led to a significant decrease in the Bohr coefficient (P<0.01; Table 1C).

Oxygen binding curves were then constructed for dialysed haemocyanin solutions containing different concentrations of L-lactate and urate at the same time (Fig. 5). At any given concentration of one effector, the oxygen affinity could still be enhanced by increasing the concentration of the other effector. The lowest half-saturation value of 0.35 kPa (2.6 mmHg) was achieved at the highest L-lactate (11 mmol l^−L) and urate (0.39 mmol l⁻¹) concentrations studied.

To check whether these experimentally determined P₅₀ values agree with the hypothesis that the effects of L-lactate and urate are additive, half-saturation values were calculated from the equations given in Table IB and Table ID for the combinations of L-lactate and urate concentrations studied. For this calculation the shift in oxygen affinity of the dialysed haemolymph expected for the investigated L-lactate concentration was added to the expected decrease in P₅₀ for the studied urate concentration, giving a predicted value of half-saturation (P_50(calc)) which could be compared to the experimental data (P_50(det)) of Fig. 5. These differences between the expected and measured values, P_50(calc) –P_50(det), are given in Table 2. The deviation between the calculated and the determined values averaged −0.02±0.06kPa (−0.15±0.45 mmHg). Statistical analysis revealed that the P₅₀ values of Fig. 5 did not differ significantly from the calculated values based on the hypothesis that the effects of both modulators are additive.

A similar calculation can be carried out for the oxygen affinity of native haemolymph containing 0.7 mmol l⁻¹ L-lactate and 0.08 mmol l⁻¹ urate, showing a P₅₀ value of 0.67 kPa (5.0mmHg) at pH7.99. Dialysis raised the P₅₀ to 0.89 kPa (6.7 mmHg). Assuming their effects to be additive, the addition of 0.7 mmol l⁻¹ L-lactate and 0.08 mmol l⁻¹ urate to the dialysed haemolymph should result in a P₅₀ of 0.71 kPa (5.3 mmHg). This value differs by 0.04 kPa (0.3 mmHg) from the P₅₀ of native haemolymph. Thus, the combined effect of L-lactate and urate accounts for 82.4% of the shift in oxygen affinity.

Oxygen binds to the haemocyanin of Homarus vulgaris in a cooperative way. The 7750 values derived by analysis of the binding data in the Hill equation are summarized in Table 3. The maximal cooperativity is achieved in dialysed haemolymph (n₅₀=4.1). Addition of either L-lactate or urate or the presence of both factors reduces cooperativity. The smallest n₅₀ value (1.7) was measured in the presence of 0.39 mmol l⁻¹ urate and 11 mmol l⁻¹ L-lactate.

The in vivo concentrations of L-lactate shown in the present study are within the range reported by other authors for Homarus vulgaris (Phillips et al. 1977; Bridges and Brand, 1980; Bouchet and Truchot, 1985). Little information is available concerning in vivo concentrations of urate in the haemolymph of this species. The average value of 0.08±0.02mmoll⁻¹ urate determined in this study is low in comparison with 0.31 mmol l⁻¹ measured by Morris and Bridges (1986) but agrees well with data for Astacus leptodactylus (Czytrich et al. 1987; Bridges, 1990), Carcinus maenas (Lallier et al. 1987) and Penaeus japonicus (Lallier and Truchot, 1989b).

The effect of L-lactate in increasing the oxygen affinity of crustacean haemocyanin has been established for many species (for a review see Bridges and Morris, 1986). In the present study, the potentiating effect of L-lactate on the oxygen affinity of lobster haemocyanin (ΔlogP₅₀)(Δlog[L-lactate])⁻¹ was −0.11 at pH7.99±0.03. This value is similar to that cited for Homarus vulgaris (−0.11 at pH7.8, −0.19 at pH7.4) by Bridges et al. (1984), whereas Bouchet and Truchot (1985) measured a higher value (−0.16 at pH 7.9) and Taylor and Whiteley (1989) measured a value of −0.18. The influence of L-lactate calculated by different authors implies a moderate potentiating effect on oxygen affinity in Homarus vulgaris compared to its effect in other species. The highest reported value of (ΔlogP₅₀)(Δlog[L-lactate])⁻¹ is −0.56 at pH7.8 and 15°C for Palaemon elegans (Bridges et al. 1984). The L-lactate effect may, however, be absent in a number of crustaceans (Bridges, 1988).

The calculated value for (ΔlogP₅₀)(Δlog[urate])⁻¹ of −0.18 at pH7.99±0.03 and 15°C agrees well with the value of −0.16 at pH 7.8 and 15°C calculated from the data of Morris and Bridges (1986) for Homarus vulgaris. In other species, slightly higher values are reported, indicating a more pronounced effect of urate on oxygen affinity. In Carcinus maenas it amounts to −0.23 at pH 7.6 (Lallier and Truchot, 1989a) and in Austropotamobiuspallipes to −0.39 at pH 7.8 (Morris et al. 1985), both at 15°C. A value of −0.48 is reported by Bridges (1990) for Astacus leptodactylus at pH7.8 and 15°C. Spicer and McMahon (1991) report values of −0.38 and −0.23, respectively, for the amphipods Apohyale pugettensis and Megalorchestia californiana at 10°C. In Penaeus japonicus, however, urate causes only a small increase in haemocyanin oxygen-affinity, (ΔlogP₅₀)(Δlog[urate])⁻¹ being −0.03 at pH7.6 and 25°C (Lallier and Truchot, 1989b).

For the lobster, haemocyanin oxygen-affinity showed a pronounced dependence on pH, with a Bohr coefficient of ψ=−1.20. This value corresponds well with the results of other authors for this species ψ= − 1.00, Taylor and Whiteley, 1989; ψ= −1.17, Bouchet and Truchot, 1985). Increasing the concentrations of either L-lactate or urate slightly decreases the Bohr coefficient at high modulator concentrations (Table 1A,C). It is not clear, however, whether the Bohr effect changes as a function of the effector concentration. This question requires a more detailed investigation. In Ocypode saratan the Bohr effect is dependent on L-lactate concentration (Morris and Bridges, 1985), whereas in Carcinus maenas and Callinectes sapidus the Bohr effect is independent of the L-lactate concentration (Truchot, 1980; Johnson et al. 1984). Previous studies on the urate effect did not show a dependency of the Bohr effect on urate concentration (Morris et al. 1985; Lallier et al. 1987; Lallier and Truchot, 1989b; Bridges, 1990).

The use of dialysed haemolymph with added L-lactate and urate guarantees constant concentrations of other factors which may have influenced haemocyanin oxygen-affinity. Observed changes in oxygen affinity must thus result from the combined effects of L-lactate and urate. Comparison of predicted and measured P₅₀ values for various combinations of L-lactate and urate concentrations showed that the values determined were slightly higher than the expected values by 5.2% (Fig. 5, Table 2). Since the statistical analysis revealed no significant difference between calculated and experimental data, this deviation may be due to experimental errors. Therefore, in the concentration range investigated, the effects of L-lactate and urate on oxygen affinity of dialysed haemolymph of Homarus vulgaris are additive.

The decrease in oxygen affinity brought about by dialysis amounted to 0.22 kPa (1.7 mmHg). The effect on P₅₀ due to the native concentrations of both effectors, 0.7 mmol l^−L L-lactate and 0.08 mmol l⁻¹ urate, is calculated to be 0.19 kPa (1.4mmHg). Several reasons for the difference of 0.04kPa (0.3mmHg) between the experimental and calculated values must be considered. It is possible that low molecular weight factors present in the whole haemolymph other than urate and L-lactate are removed by dialysis. Morris et al. (1986,a) have clearly shown that other purines such as hypoxanthine, adenine and inosine increase oxygen affinity in Austropotamobius pallipes. Likewise, dopamine has been shown to increase haemocyanin oxygen-affinity in vitro in Cancer magister (Morris and McMahon, 1989a, b).

At L-lactate and urate concentrations exceeding the values measured in undisturbed animals, the oxygen affinity of haemocyanin can be raised in vitro beyond the affinity determined in whole haemolymph. At the highest modulator concentrations, a half-saturation value of 0.35 kPa (2.6 mmHg) was achieved, which is 0.32 kPa (2.4 mmHg) lower than the P₅Q determined in whole haemolymph. Even at high L-lactate concentrations, the addition of urate further increases haemocyanin oxygen-affinity and vice versa. Yet the increasing effect of L-lactate and urate present in the haemolymph on haemocyanin oxygen-affinity must be limited, since the effect of each single modulator is progressively reduced at high concentrations (Figs 2, 4, 5).

In the presence of L-lactate and urate, the n₅₀ values decrease markedly at the same time as the oxygen affinity increases. Morris and Bridges (1989) suggest that this is the case when K_Tass is affected to a greater extent than K_R,ass by the studied modulators. This is true in the present investigation. Cooperativity has been explained by the MWC model (Monod et al. 1965) as a symmetrical national change of the entire protein, with the binding of the first ligand facilitating the binding of further ligand molecules. According to this model, the maximal cooperativity is determined by the number of binding sites for a ligand on the protein molecule. So, in the case of the dodecameric haemocyanin, the maximal n₅₀ possible for oxygen binding is 12. The highest value achieved here, however, was 4.1, which is far below the maximum. Moreover, the presence of both L-lactate and urate in a haemocyanin solution decreased cooperativity to a greater extent than each effector was able to do alone (Table 3). These results suggest that the binding of one ligand does not affect the entire haemocyanin molecule, but only a limited environment adjacent to the ligand binding site. This asymmetry is not compatible with the MWC model. For this reason, several authors have extended this model for large protein molecules composed of many subunits; the nesting model (Decker et al. 1986) and the model of interacting cooperative units (Brouwer and Serigstad, 1989) are perhaps better suited to explain the binding properties and cooperativity characteristics of haemocyanin.

While the potentiating effects of L-lactate and urate on haemocyanin oxygen affinity have been confirmed by several in vitro investigations, little information is available on the influence of these factors on the animal’s physiology. To serve as true modulators in vivo, their concentration in the haemolymph must vary with different physiological conditions.

Exercise increases L-lactate concentrations in the haemolymph, as was shown for Callinectes sapidus (Booth et al. 1982) and for Cancer magister (Graham et al. 1983). L-Lactate concentrations are also reported to increase in hypoxic situations in Homarus vulgaris at the critical (Bridges and Brand, 1980; Bouchet and Truchot, 1985; Taylor and Whiteley, 1989). With this increase in L-lactate concentration, a concomitant reduction occurs in P₅₀. Hypoxia also leads to elevated urate concentrations in Astacus leptodactylus (Czytrich et al. 1987), Carcinus maenas (Lallier et al. 1987) and Penaeus japonicus (Lallier, 1988) because the urate concentration in the haemolymph is dependent upon the available oxygen, since the urate oxidising enzyme (uricase) requires molecular oxygen (Mahler, 1963). In contrast to L-lactate formation, urate metabolism is limited by moderate hypoxia, resulting in an accumulation of urate in the haemolymph. L-Lactate, which is the anaerobic end product in Crustacea (Chang and O’Connor, 1983; Gade and Grieshaber, 1986), accumulates during severe hypoxic situations (Bouchet and Truchot, 1985) and exercise (Booth et al. 1982; Graham et al. 1983), when the energy metabolism is switched to anaerobic pathways. Thus, a shortage of oxygen leads to an initial increase in urate concentration followed by an increase in L-lactate concentration as hypoxia becomes more severe. Both factors, in turn, elevate the oxygen affinity of haemocyanin. This feedback may increase the ability to take up oxygen in the gills and enhance the amount of oxygen transported in the haemolymph.

This work was supported by the Deutsche Forschungsgemeinschaft (Gr456/ 12-1) and by the Fond der Chemischen Industrie (MKG).