INTRODUCTION

Top Abstract Introduction Materials and methods Results Discussion References

THE OCCURRENCE OF HEMOGLOBIN (Hb) in invertebrates is widespread among invertebrate phyla, where an extensive range of adaptive roles has been postulated (16, 31). Among mollusks that commonly use hemocyanin for gas transport, extracellular Hbs occur in aquatic pulmonate gastropods and two families of bivalves (13, 38).

In contrast to the intracellular vertebrate Hbs, which weigh ~68 × 10³ Da, extracellular planorbid Hbs are large, multidomain and multisubunit proteins (28) with molecular mass values of 1.60-1.75 × 10⁶ Da [1.70 × 10⁶ Da for Biomphalaria glabrata (2)]. The molecules contain 3% sugars (1) and 1 heme per 17,700 Da protein (23). Only scant information is available on the quaternary structure of B. glabrata Hb. However, on the basis of electron micrographs, the Hb of the related Planorbis corneus has been interpreted as consisting of hexagonal ring structures (37) and that of the pulmonate Helisoma trivolvis as, variously, a 10-membered ring structure (22), 12 single polypeptide chain subunits that each carry 10 hemes and are grouped in pairs held together by disulfide bonds (10), and a compact two-layer pentameric ring structure of decamers stabilized by disulfide linkages (9). In contrast, the Hbs of the heterodont bivalve families Astartidae and Carditidae exhibit higher molecular masses (8-12 × 10⁶ Da), greater size heterogeneity of the native molecules, and subunits of 240,000-390,000 Da that contain 1 mol heme per 17,000-20,000 g protein (20).

As with vertebrate Hbs, planorbid Hbs exhibit inhibitory, heterotropic interactions between O₂ and proton binding sites (Bohr effects) as well as homotropic heme-heme interactions (cooperativity) that are responsible for the sigmoid shape of O₂ binding curves (23, 26, 38). Invertebrate heme-carrying pigments are insensitive to anionic organic phosphates like glycerate-2,3-bisphosphate and ATP (16, 30, 31) that decrease O₂ affinity of Hb in vertebrate red blood cells by lowering the affinity constant of the tense, deoxygenated form of the Hb molecule (K_T). O₂ affinity of the high-molecular weight, extracellular Hbs and chlorocruorins from annelids are increased by inorganic cations (5, 11, 32), and data of van Aardt and Naude (26) provide evidence for similar effects in Biomphalaria. In the annelid pigments, inorganic cations and decreasing proton concentrations increase O₂ affinity primarily by raising the affinity constant of the relaxed, oxygenated form of the Hb molecule (K_R) (8, 11, 25, 32).

The Hbs of pulmonate snails appear to play an important role in O₂ transport. In P. corneus the presence of Hb correlates with a 20-min delay in the onset of anaerobiosis compared with that in the Hb-free pulmonate Lymnaea stagnalis and increased diving potential, lower postdiving pulmonary O₂ tensions, and a greater exploitation of the pulmonary O₂ store than in L. stagnalis (12).

Biomphalaria typically inhabits swamps that present drastic variations in the physicochemical factors that affect Hb-O₂ binding. Extreme hypoxia (PO₂ commonly falling to 0-7 mmHg) is frequently accompanied by marked hypercapnia [PCO₂ values rising to 35 mmHg (13)] and large temperature variations. As with many invertebrates, planorbid snails do not maintain constancy in the hemolymph ionic concentrations in response to changing osmolality of the medium (14).

We have studied the oxygenation properties of B. glabrata hemolymph and Hb and the effects of CO₂, inorganic ions, pH, and temperature, seeking to identify adaptations to natural environmental conditions and mode of life and the mechanisms that regulate the oxygenation process in pulmonate Hbs.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and methods Results Discussion References

Specimens of the freshwater snail B. glabrata (strain Puerto Rico; albino 770302 l-01) were reared in aquariums with water plants at the location at 20°C. The animals were fed 3-4 times per week on a mixed diet that included trout pellets (Biomar, Ecolife 19).

Animals with a shell diameter of 1-2 cm, weighing 0.4-1.5 g, were used for experiments. The hemolymph was sampled directly from the pericardial sinus after piercing of the shell and body wall with a needle and imbibing it into thinly drawn-out glass capillaries. Samples from individual animals were stored separately on ice without freezing. Those that showed no met-Hb formation were pooled and used in experiments within at most 3 days.

In vivo hemolymph pH values were measured using a BMS 2 Mk 2 meter coupled to a PHM72 millivolt meter (Radiometer, Copenhagen, Denmark) after fitting the sharp end of a hypodermic needle to the capillary tube of the pH electrode, which was then used to pierce the animals, allowing hemolymph to be drawn directly into the pH electrode without contact with air. Hemolymph osmolality was measured in five individual snails using a Knauer semi-micro osmometer (Berlin, Germany).

Gel filtration of the native hemolymph was carried out using a 40 × 1.6 cm (height × diameter) Sephadex G200 column, eluted with 0.1 M Tris buffer, pH 7.1. Absorbances of eluted fractions were read at 280 and 415 nm (for identifying proteins and heme components, respectively). Isoelectric focusing was performed on hemolymph dialyzed against 0.01 M Tris buffer in a 110-ml column (LKB, Bromma, Sweden), using ampholines of pH 3-6 (0.6%), 5-7 (0.07%), and 3.5-10 (0.07%). After focusing (at 400 V), 1.1-ml fractions were collected for absorbance and pH measurements (at 23°C).

O₂ binding measurements were performed on native hemolymph and on samples that had been dialyzed to remove possible cofactors to Hb-O₂ binding. Dialyses were carried out in preboiled, semipermeable tubing (molecular weight cutoff, 12,000-14,000; Struers) against at least three changes of CO-equilibrated 0.1 M Tris buffer (pH 7.5 at 25°C).

O₂ equilibria were determined on 3-µl Hb samples using a modified gas diffusion chamber (32, 33) linked to cascaded Wösthoff gas mixing pumps (Bochum, Germany), where O₂ tensions were increased stepwise by mixing air or O₂ with highly pure (>99.998%) N₂ while absorption was recorded continuously. O₂ saturations were evaluated from absorbances relative to those values for the fully oxygenated and fully deoxygenated samples, which were obtained after equilibration with pure O₂ and N₂, respectively. Reductions in the absorption difference between the fully oxygenated and deoxygenated Hb during the recordings (that indicate met-Hb formation) were usually absent and always <5%.

The equilibria were recorded at different pH values, which were obtained either by adding Tris buffer to a final concentration of 0.1 M (to assess the "fixed-acid Bohr effect") or by varying the CO₂ tension in the equilibrium gases ("CO₂ Bohr effect"). Half-saturation O₂ tension (P₅₀) values at specific pH values were interpolated from the P₅₀-to-measured pH relationship. pH measurements were carried out in duplicate, at the same temperature as the O₂ equilibrium measurements, on 50-µl subsamples after a minimum of 6 min of equilibration at 90-95% O₂ saturation, using the pH meter described above.

Oxygenation data involving at least four equilibrium steps between 30 and 70% saturation were converted to Hill plots [log (Y/1Y) vs. log P], where Y is the fractional O₂ saturation and P the O₂ tension for estimation of P₅₀ and Hill's cooperativity coefficient at half-saturation (n₅₀). The heat of oxygenation (H) was calculated as H = 2.303 · R(log P₅₀/T), where R is the gas constant and T the absolute temperature. The effects of inorganic ions on O₂ binding were examined by adding accurate volumes of 1-M solutions of CaCl₂, MgCl₂, KCl, or NaCl and was checked by measuring Cl concentrations using a CMT 10 chloride titrator (Radiometer, Copenhagen, Denmark).

Precise equilibria measurements for extended Hill plots that emphasize extreme O₂ saturation values near 1-5 and 95-99% were carried out as previously described (35). Errors resulting from possible incomplete saturation of the Hb after equilibration with pure O₂ at atmospheric pressure were corrected by end-point extrapolation as described (35). The data were analyzed in terms of the parameters of the two-state Monod-Wyman-Changeux (MWC) equation (17) where S denotes saturation, L the allosteric constant, and q the number of interacting O₂ binding sites. Additional analyses were carried out with q fixed at 6 (the approximate value obtained in analyses with "free" q; cf Table 1).

Derived parameters were calculated from the fitted parameters as follows. P₅₀ was obtained by solving (for PO₂) the equation The median O₂ tension (P_m) was calculated as where c = K_T/K_R.

The maximum slope of log [S/(1  S)] vs. log PO₂ [i.e., maximum cooperativity coefficient (n_max)] was calculated by first solving (for PO₂) the equation and then calculating {log [S/(1  S)]}/log PO₂ at that PO₂.

The free heme-heme energy of interaction (G) was calculated as Heme concentrations were determined from the - and -absorption peak values of oxygenated Hb after 25- to 101-fold dilution in 0.1 M Tris buffer (pH 7.5 at 25°C). Oxygenated Hb was obtained by flushing with O₂, deoxy Hb by flushing with N₂ and adding a trace of solid sodium dithionite, and the carboxy Hb by CO flushing of the deoxygenated sample. Addition of potassium ferricyanide (for obtaining met-Hb spectra) resulted in protein denaturation.

	RESULTS

Top Abstract Introduction Materials and methods Results Discussion References

Whole hemolymph. The in vivo pH value in the hemolymph of B. glabrata measured at 25°C was 7.78 (SD = 0.062, n = 8). Hemolymph osmolality was measured to be 57 mosmol/kg (SD = 29.4, n = 5).

Heme concentrations in the pooled hemolymph samples varied greatly, from 0.31 to 0.87 mM (n = 20). Gel filtration of whole hemolymph samples and isoelectric focusing of dialyzed Hb (Fig. 1) showed that the Hb constitute ~86% of the total hemolymph protein and consists of a single component with an isoelectric point (pI) of 4.7 (Fig. 1).

View larger version (22K): [in this window] [in a new window]   Fig. 1.   Isoelectric focusing and gel filtration (inset) profiles of Biomphalaria glabrata hemolymph carried out at 5°C as described in MATERIALS AND METHODS. Abs, absorbance.

The native hemolymph displayed a moderately high O₂ affinity (P₅₀ at 25°C = 6.1 and 2.8 mmHg at pH 7.7 and 8.4, respectively; Fig. 2A). Although a Bohr effect was observed in the entire pH range measured, the Bohr factor decreased from 0.50 at alkaline pH values (7.4-8.2) to 0.22 at pH 6.8-7.1 (Fig. 2A). n₅₀ increased from 1.1 at pH 7.0 to 2.0 at pH 8.0, exhibiting maximum values at alkaline conditions where the Bohr effect is greatest (Fig. 2A). CO₂ exerted no significant, specific (pH independent) effect on O₂ binding as judged from similar O₂ affinity and cooperativity when the pH was varied by adding either CO₂ or buffers (resulting in the same CO₂ and fixed-acid Bohr effects; Fig. 2B).

View larger version (19K): [in this window] [in a new window]   Fig. 2.   A: CO2 Bohr effect at 25°C of the whole hemolymph () and the dialyzed hemoglobin (Hb) in the absence () and presence () of 0.1 M Ca2+. Heme concentration = 0.44 mM () and 0.41 mM ( and ). B: CO2 Bohr effect () and fixed-acid (FA) Bohr effect (, in 0.1 M Tris) of the hemolymph, measured at 15°C. P50, half-saturation O2 tension; n50, Hill's cooperativity coefficient at half-saturation.

Extended Hill plots at different pH values are shown (Fig. 3). Analysis of the data in terms of the two-state MWC model and the derived parameters (Table 1) showed that in the pH range of 7.8 to 6.9 that spans physiological conditions, changes in proton concentration have a much greater effect on the O₂ affinity of the oxygenated state (K_R), than the deoxygenated state (K_T) (Fig. 3, A and B). Whereas K_R increased from 0.27 mmHg¹ at pH 6.8 to 0.99 mmHg¹ at pH 8.4, reflecting a Bohr factor for binding the last O₂ molecule of 0.41 (Fig. 4B), K_T showed virtually no pH dependence between 6.8 to 7.7, but increased from 0.087 to 0.23 mmHg¹ between pH 7.7 and 8.4. Similar pH-induced variation in K_T and K_R values (not illustrated) was obtained in analyses with q fixed at 6, except for the lowest pH tested (6.86), where fixing q markedly increased K_R and L values.

View larger version (36K): [in this window] [in a new window]   Fig. 3.   A: extended Hill plots for CO2 Bohr effect curves in native hemolymph at 25°C and the indicated pH values. The intercepts of upper and lower asymptotes of the Hill plots (dashed lines) with the horizontal axis at log [Y/1  Y] = 0, where Y is fractional O2 saturation, indicate the log and log KR values. Heme concentration = 0.44 mM. B: pH dependence of the parameters KT (), KR (), P501 (), Pm1 (). KT and KR, O2 association equilibrium constants of Hb in tense (deoxygenated) and relaxed (oxygenated) states, respectively; Pm, median O2 tension.

View larger version (37K): [in this window] [in a new window]   Fig. 4.   Temperature (T) dependence of oxygenation parameters. A: CO2 Bohr effect of whole hemolymph at 10, 17.5, and 25°C. B: van't Hoff plot, pH 7.7 for CO2 buffered () and fixed-acid buffered () hemolymph, showing overall heat of oxygenation (H) values at different temperature intervals. C: extended Hill plots of CO2 buffered whole hemolymph at pH 7.7 at 10, 17.5, and 25°C. D: van't Hoff plot of oxygenation parameters derived from C. Heme concentration = 0.57 mM.

The oxygenation process exhibited strong temperature dependence. At pH 7.7, hemolysate P₅₀ values were 6.1, 3.6, and 1.8 mmHg at 25, 17.5, and 10°C (Fig. 4A). The Bohr effect was little affected by temperature (Bohr factor = 0.43, 0.38, and 0.44 at 10, 17.5, and 25°C, respectively). Extended Hill plots at different temperatures (Fig. 4C) and the MWC parameters (Table 1, Fig. 4D) showed that similar increases in K_T and K_R values with decreasing temperature underlie the temperature insensitivity of cooperativity and G values.

H, which includes the heat of solution of O₂ and other oxygenation-linked processes, was not linear but decreased with increasing temperature (Fig. 4B); for the CO₂-buffered hemolymph, it was 63 and 49 kJ/mol below and above 17.5°C, respectively (interpolated at pH = 7.7). Similar nonlinearity of H values with respect to temperature was evident under fixed-acid buffering conditions (Fig. 4B).

Astrup titrations carried out to examine the Haldane effect and differences in the buffer capacities of oxygenated and deoxygenated whole hemolymph (Fig. 5) revealed higher pH values in deoxy than in oxy hemolymph at CO₂ tensions below PCO₂ = 12 mmHg (i.e., above pH 7.3), the difference increasing with increasing pH (as did the Bohr effect; Fig. 2A). At PCO₂ = 3.7 mmHg, pH values were 7.79 and 7.69 in deoxy and oxy hemolymph, respectively.

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Astrup titration of CO2 buffered deoxygenated () and oxygenated () hemolymph at 15°C. Heme concentration = 0.64 mM.

Dialyzed Hb solutions. While displaying similar O₂ affinity and cooperativity values as the whole hemolymph at pH < 7.3, the dialyzed Hb showed distinctly lower affinities at higher pH values (P₅₀ values at 25°C were 7.4 and 6.0 mmHg at pH 7.7 and 8.1, respectively, compared with 6.1 and 3.9 mmHg in native hemolymph; Fig. 2). Dialysis, moreover, decreased the Bohr factor (from 0.50 in the whole hemolymph to 0.32 at pH 8.0), but had no marked effect on cooperativity.

Extended Hill plots of dialyzed Hb showed similar pH dependence as in the whole hemolymph (Fig. 6, A and B; Table 1), K_R varying strongly (0.71 and 0.25 mmHg¹ at pH 8.1 and 7.0, respectively) and K_T changing only slightly with proton concentration. At pH 8.1, where cooperativity is maximal (n₅₀ = 1.96), the K_T and K_R values (0.085 and 0.71 mmHg¹) reflect an 8.4-fold greater O₂ affinity in the oxy than in the deoxy state and a G value of 5.06 kJ/mol. Analyses in terms of the MWC model with q fixed at 6 gave similar values for the K_R-to-K_T affinity ratio (8.1) and G (4.99 kJ/mol).

View larger version (33K): [in this window] [in a new window]   Fig. 6.   A: Hill plots at 25°C of dialyzed Hb at different pH values (fixed-acid Bohr effect measured in 0.1 M Tris); heme concentration = 0.41 mM. B: pH dependence of KR, KT, and P501 parameters derived from A.

The decrease in O₂ affinity on dialysis prompted examination of the effects of inorganic cations on Hb oxygenation. In contrast to the effects of increasing proton and salt concentrations in vertebrate Hbs (3), cations increased Hb-O₂ affinity of B. glabrata Hb. Raising the Na⁺, Mg²⁺, or Ca²⁺ concentration in dialyzed Hb to 25 mM decreased P₅₀ from 8.8 mmHg to 8.1, 6.3, and 4.8 mmHg, respectively, at pH 7.5 (Fig. 7), showing much greater effects of divalent than monovalent cations and showing that Ca²⁺ is a more potent effector than Mg²⁺. The cation effect was pH dependent; addition of 100 mM Ca²⁺ lowered P₅₀ from 9.5 to 4.3 mmHg at pH 7.0 and from 6.0 to 3.7 mmHg at pH 8.1, thus decreasing the Bohr factor (from 0.32 to 0.08 at pH 8.0; Fig. 2A). Significantly, Ca²⁺ almost completely obliterated cooperativity (n₅₀ decreased from 1.96 to 1.12 at pH 8.1).

View larger version (20K): [in this window] [in a new window]   Fig. 7.   Effects of K+, Na+, Mg2+, and Ca2+ on O2 affinity at 25°C. Fixed-acid conditions, 0.1 M Tris, pH 7.5. Heme concentration = 0.52 mM.

The effects of cations on O₂ affinity can be expressed in terms of the basic linkage equation log P_m/log cation concentration = X, where X is the amount of cations bound to the Hb on O₂ binding (3, 33). The overall correspondence between the P₅₀ and P_m values, and between n₅₀ and n_max values (Table 1; see also Figs. 3B and 9B), indicates symmetrical O₂ binding curves, permitting analysis of the P₅₀ data in terms of linkage equations. In the cation concentration range of 20-100 mM, the log P₅₀/log cation concentration relationship indicates oxygen-linked binding of 0.04 Na⁺ or K⁺ ions and 0.17 Ca²⁺ ions per heme group oxygenated.

At constant pH, the shift in K_T and K_R values in the absence (cf Fig. 6A) or presence (cf Fig. 8A) of Ca²⁺ indicates that cations increase O₂ affinity primarily by raising K_T (Fig. 8B). As interpolated at pH 7.8, Ca²⁺ increased K_T 2.7-fold (from 0.083 to 0.22 mmHg¹) and K_R 1.3-fold (from 0.50 to 0.65 mmHg¹). This trend was confirmed in analyses with q fixed at 6. 

View larger version (31K): [in this window] [in a new window]   Fig. 8.   A: Hill plots of dialyzed Hb in 0.1 M Tris at the indicated pH values (fixed-acid conditions) and in the presence of 100 mM Ca2+ at 25°C. Heme concentration = 0.41 mM. B: pH dependence of the association constants KT (squares) and KR (circles) of dialyzed Hb in the absence (open symbols; from Fig. 6A) and the presence (filled symbols; from A) of 100 mM Ca2+.

The effect of varying Ca²⁺ concentrations on dialyzed Hb was examined at pH 8.1, where cooperativity is maximal (Fig. 9A). With increasing Ca²⁺ concentration, n₅₀ decreased (from 2.0 to 1.1 at 100 mM Ca²⁺), and affinity increased to a relatively stable value at Ca²⁺ concentrations above 10 mM. These effects are attributable to a marked increase in K_T (that also was evident when the MWC model was fitted with q fixed at 6) with increasing Ca²⁺ levels (Fig. 9B; Table 1) and a transient increase in K_R. Significantly, addition of Ca²⁺ to concentrations up to 0.1 M did not significantly affect the pH of the Hb solution at pH 8.1 and only slightly decreased pH at lower pH values. This was confirmed in separate experiments that showed 1) pH = 0.026 [Ca²⁺] + 0.0017 (r = 0.121) at pH = 8.1 and heme concentration = 0.37 mM, where [Ca²⁺] is the molar calcium concentration and 2) pH⁽⁰¹⁰⁰⁾ = 0.026 pH  0.216 (r = 0.906), where pH⁽⁰¹⁰⁰⁾ is the pH induced by 100 mM Ca²⁺.

View larger version (27K): [in this window] [in a new window]   Fig. 9.   A: Hill plots of dialyzed Hb in 0.1 M Tris at 25°C and pH 8.1 in the presence of the indicated concentrations of Ca2+ (fixed-acid buffering). Heme concentration = 0.41 mM. B: Ca2+ dependence of KT (), KR (), P501 (), Pm1 (), and n50 ().

Spectrophotometric characteristics. Spectrophotometric analysis (Fig. 10) showed - and -absorption maxima at 574 and 539 nm, respectively (compared with 577 and 542 nm for human Hb A), and an -to--absorption ratio of 1.01 [compared with 1.07 and 1.04 in human Hb A (27) and extracellular annelid Hb from the giant earthworm Megascolides australis (34), respectively]. The lack of a distinctive absorption peak at 630 nm reflected absence of Met-Hb. With this in view, the absorptions near 630 nm and the low -to--absorption ratio (~0.94) earlier recorded for P. corneus (37) likely represent met-Hb formation rather than a pulmonate character. The carboxy compound showed - and -absorption peaks at 568 and 539 nm, thus differing only at the -peak compared with oxy Hb, and an -to--absorption ratio of 0.94. The span between the -peaks for oxy and carboxy Hb (that reflect the affinity difference of Hb for the two ligands in vertebrate Hbs) is 6 nm. Deoxy Hb showed a single broad band with a peak at 555 nm. The Soret band maxima for oxy, deoxy, and carboxy Hb were at 413, 430, and 419 nm, and exhibited an absorption ratio of 1:1.35:0.94.

View larger version (23K): [in this window] [in a new window]   Fig. 10.   Spectral characteristics of derivatives of B. glabrata oxy Hb (solid line), deoxy Hb (dashed line), and carboxy Hb (dotted line) in 0.1 M Tris, pH 7.5. Heme concentration = 0.0044 mM (370-450 nm; left) and 0.018 mM (500-600 nm; right).

	DISCUSSION

Top Abstract Introduction Materials and methods Results Discussion References

Invertebrate Hbs display an intriguing diversity in structure and function, correlating with extraordinary variations in the physicochemical conditions under which they operate in vivo, and the diversity may compensate for a lesser-developed organization at the organ level and a shift of the regulatory burden toward the molecular level compared with vertebrates (31). Compared with vertebrates, little is known about the relation between physiological function and molecular characteristics in invertebrate Hbs, particularly in planorbid Hbs that form a distinctive class as regards quaternary structure. These Hbs may be compared with the larger (3-4 × 10⁶ Da) extracellular annelid pigments that exhibit highly characteristic, two-tiered hexagonal structures (hexagonal bilayers), may be dissociated into 12 submultiples, and may contain nonheme "linkers" in addition to heme-carrying chains (28).

O₂ affinity and cooperativity and their pH dependence. The relatively high O₂ affinities (P₅₀ = 6.1 mmHg at pH 7.7; Fig. 2A) and low cooperativities (n₅₀ = 2.0 and 1.1 at 25°C and pH 8.4 and 6.9, respectively) are in general agreement with previous data for planorbid snails (Table 2).