INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

THE HEMOGLOBIN (Hb) of crocodiles is unique compared with other vertebrate Hbs by showing an oxygenation-linked binding of bicarbonate ions (1). Bicarbonate binds to the deoxy (T) structure of the Hb at the entrance to the central cavity between the two -chains; i.e., the site involved in organic phosphate binding in other vertebrate Hbs (21). A few amino acid substitutions in crocodile Hbs have made the binding site incompatible with organic phosphate and compatible with the binding of two ions (19, 21). The allosteric binding of underlies the large effect of CO₂ on blood-Hb O₂ affinity seen in crocodiles (1, 2, 12, 28). As noted by Perutz and co-workers (21), the oxygenation-linked binding of gives a simple and direct reciprocal action between O₂ and one of the end products of oxidative metabolism. The mechanism has, however, not been adopted by other vertebrates and it is unknown what significance it may have for the crocodilians (21). A link between O₂ and metabolically produced CO₂ (that after hydration forms and H⁺) is also present in other vertebrates through the oxylabile binding of H⁺ to Hb (the fixed acid Bohr-Haldane effect).

Whereas the molecular mechanism of bicarbonate binding and the effects of CO₂ on Hb O₂ affinity are well documented in crocodiles, the consequences of the special functional properties of crocodile Hbs have remained largely unexplored at the red blood cell (RBC) level. This particularly applies to CO₂ transporting properties. Hb is important for both O₂ and CO₂ transport in the blood. The functional significance of binding may therefore equally well be related to its consequences for CO₂ transport as it may be related to its effects on O₂ affinity. Bicarbonate binding to Hb can be predicted to provide crocodile RBCs with unique CO₂ transporting properties, where the fraction of total blood CO₂ contained within the RBCs is larger than in other vertebrates. In the normal vertebrate CO₂ transport pattern, two processes drive the CO₂ hydration reaction in the RBCs toward bicarbonate formation. These are H⁺ binding to the Hb and a shift of produced to the plasma in exchange for Cl via the anion exchanger (known as AE1 or band 3). In crocodiles, bicarbonate binding to Hb provides an additional mechanism of removing free from the RBC cytosol, increasing the amount of CO₂ that is hydrated and thereby the blood CO₂ carrying capacity. Anion exchange is generally considered the rate-limiting step in the uptake of CO₂ in tissue capillaries and the excretion of CO₂ across the respiratory epithelium. This particularly applies if band 3 transport capacity and/or velocity is reduced. On this basis, it is possible that oxygenation-linked binding has evolved in crocodiles to compensate for a slow /Cl exchange across the RBC membrane.

Oxygenation-linked binding of bicarbonate can also be expected to have consequences for RBC acid-base status and the nonpermeable charge carried by the Hb molecule, so that the influence of oxygenation on RBC intracellular pH (pH_i) and on the Donnan-like distribution of permeable ions across the RBC membrane in crocodiles should be different from other vertebrates.

The present study was designed to investigate these ideas in the alligator, Alligator mississippiensis. Specifically, we have 1) traced the unique allosteric binding to the RBC level, 2) analyzed the influences of CO₂ and of oxygenation on extra- and intracellular acid-base status, 3) studied CO₂ and H⁺ binding properties of blood and Hb, 4) made a partitioning of various forms of CO₂ in erythrocytes, and 5) determined the Cl, , and H₂O permeability of the RBC membrane.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Alligators (A. mississippiensis, 1.2-1.7 m in length, n = 6) were kept at 30°C in a 20-m² room with facilities for basking and immersion in water. They were fed chicken once a week and dry dog food (pellets) ad libitum. One to two days before blood sampling, a catheter was inserted occlusively in the femoral artery of one of the hindlegs under local anesthesia (Xylocaine). The animals were then placed individually in a box, which was partly covered with a blanket to allow sampling of blood without disturbing the animal. Approximately 40 ml of blood was drawn from the femoral artery of each animal. The first 10 ml was used for blood tonometry and the remaining 30 ml was used for flux experiments.

Blood Tonometry

Extracellular blood pH (pH_e) and RBC pH_i were measured with a Radiometer (Copenhagen, Denmark) capillary pH electrode and a Radiometer PHM 73 monitor. Hb was measured spectrophotometrically after conversion of Hb to cyanmethemoglobin using a millimolar extinction coefficient of 11 at 540 nm. Hct was assessed by centrifugation in glass capillaries. The red cell Hb concentration ([Hb]) was calculated from blood [Hb] and Hct. N_RBC counts were performed using a Bürger-Türk counting chamber and a microscope. Mean cellular volume (MCV) was calculated from Hct/N_RBC. Plasma lactate was assessed with the Boehringer-Mannheim lactate dehydrogenase method. C_T values in plasma and in whole blood were measured with the Cameron (8) method. Plasma bicarbonate was calculated as plasma C_T (C^plasma_T)  ()PCO₂, using a plasma CO₂ solubility () at 30°C of 0.0366 mmol · l¹ · mmHg¹ [calculated from the formula of Heisler (14) and the concentrations of ions in alligator plasma (11)]. The apparent RBC bicarbonate concentration ([]_app; in mmol/l RBC) was calculated from blood and plasma C_T and the fractional Hct (F_Hct), according to the formula using an RBC value of 0.032 mmol · l¹ · mmHg¹ (extrapolated from human values; Ref. 23). The calculated quantity is an []_app, as it also includes Hb-bound [] and carbamino compounds. [HCO₃]_app values were converted to concentrations in millimoles per liter (cell water) using the corresponding red cell F_H2O values.

The apparent pK' for the CO₂/ system in alligator plasma at 30°C was calculated from measured plasma C_T and pH values, using the rearranged Henderson-Hasselbalch equation Linear regression on the data gave the following relationship between pK' and pH for alligator plasma: pK' = 0.0817 × pH + 6.7818.

Data from blood tonometry are presented as means ± SE, and the significance of differences between treatments were evaluated by a two-factor (oxygenation degree and CO₂ level) analysis of variance (ANOVA) (repeated-measures design) followed by the Tukey multiple-comparison test.

Flux Measurements

The rate of tracer efflux was determined at 30°C by means of the continuous flow tube technique that has a time resolution in milliseconds and therefore is well-suited to measure RBC anion transport (4, 6). In brief, 0.5 ml of the packed and radioactive loaded RBCs was continuously mixed with 270 ml of isotope-free saline in a mixing chamber. The saline was air equilibrated in experiments with oxygenated RBCs and N₂ equilibrated in experiments with deoxygenated RBCs. The dilute suspension flowed through a pipe where cell-free filtrates were collected at predetermined distances. The filtrates contained increasing amounts of radioactivity with increasing distance from the mixing chamber. With known flow rates, distances could be converted to times. Because tracer efflux follows a monoexponential course, plotting the extracellular radioactivity versus time in a semilogarithmic plot allowed a determination of the rate of tracer efflux by linear regression analysis.

The rate coefficient, k (s¹), of the unidirectional efflux of tracer was used to calculate a permeability coefficient as P = k · (V · A¹), where (V · A¹) is the ratio of the cell water volume to the cell membrane area. Because transport of Cl and saturates, the permeability coefficient is concentration dependent. We therefore use the term "apparent permeability" (P_app) to denote that the value of P refers to a given anion concentration (i.e., the concentration resulting from the physiological saline used). Data from the flux experiments are presented as means ± SE and were statistically evaluated by a two-factor (compound and oxygenation degree) ANOVA followed by the Tukey multiple-comparison test.

Hb H⁺ Titration

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Blood CO₂ Content, Acid-Base Status, and Hematology

View larger version (14K): [in this window] [in a new window]   Fig. 1.   CO2 equilibrium curves for oxygenated () and deoxygenated () true plasma (A), whole blood (B), and red blood cells (RBCs; C) of the alligator (Alligator mississippiensis) at 30°C. Total CO2 is given in mmol/l plasma, mmol/l blood, and mmol/l cell water for the 3 compartments, respectively. Means ± SE (n = 6).

View larger version (33K): [in this window] [in a new window]   Fig. 2.   Whole blood tetrameric Hb concentration ([Hb4]; A), fractional hematocrit (FHct; B), mean cellular Hb concentration (C), and fractional water content (FH2O; D) of RBCs in oxygenated (hatched bars) and deoxygenated (solid bars) alligator blood at different CO2 levels. Means ± SE (n = 6).

Blood [Hb], Hct, and RBC [Hb] did not change significantly with oxygenation degree (Fig. 2). Mean Hct was 23.5%, and the tetrameric Hb concentration ([Hb₄]) of the blood was ~1.14 mmol/l (corresponding to a mean cellular [Hb₄] of 4.85 mmol/l RBC). The F_H2O was close to 0.64 (Fig. 2). The N_RBC in alligator blood (millions/µl) was 0.698 ± 0.043, and the MCV at 4% CO₂ was 352 ± 11 µm³.

Oxygenated blood equilibrated with a physiological CO₂ tension (4% CO₂ = 29 mmHg) had a pH_e of 7.56 and a plasma [] of 25.7 mmol/l (Fig. 3). This acid-base status was similar to the in vivo arterial acid-base status in alligators at 30°C (11). Plasma lactate concentration was the same in oxygenated (0.9 ± 0.3 mmol/l) and deoxygenated (0.9 ± 0.2 mmol/l) blood. Plasma [] and pH_e changed significantly in both oxygenated and deoxygenated blood when the CO₂ content of the equilibration gas was decreased to 2% CO₂ or increased to 7% CO₂ (Fig. 3). The true plasma nonbicarbonate buffer value (_nb = []/pH_e) was 15.7 mmol · l¹ · pH unit¹ for oxygenated blood and 17.2 mmol · l¹ · pH unit¹ for deoxygenated blood. The corresponding buffer values in whole blood (not illustrated) were 22.7 and 20.9 mmol · l¹ · pH unit¹, respectively.

View larger version (15K): [in this window] [in a new window]   Fig. 3.   True plasma concentration ([]) vs. extracellular pH (pHe) in oxygenated () and deoxygenated () blood of the alligator (Alligator mississippiensis) at 30°C. Dotted lines are iso-PCO2 curves calculated from the Henderson-Hasselbalch equation, using experimentally determined pK' values (cf. MATERIALS AND METHODS). Means ± SE (n = 6).

The pH_e did not differ significantly between oxygenated and deoxygenated blood at any of the CO₂ levels, and plasma [] in deoxygenated blood was only slightly higher than in oxygenated blood (Fig. 3).

RBC CO₂ Content and Acid-Base Status

1

1

View larger version (13K): [in this window] [in a new window]   Fig. 4.   A: apparent concentration (in mmol/l cell water) of bicarbonate (i.e., total CO2 minus dissolved CO2) as function of intracellular pH (pHi) in oxygenated () and deoxygenated () RBCs of the alligator. B: ratio between apparent [] ([]app) and [Hb4] as a function of pHi in oxygenated () and deoxygenated () RBCs from the alligator. Means ± SE (n = 6). Arrow a in A indicates actual change on deoxygenation at 4% CO2. Arrow b is a hypothetical path along a tentative iso-PCO2 curve (dotted curve), which should have been followed if binding to the Hb had been absent and the difference between oxygenated and deoxygenated RBCs had been due to oxygenation-linked H+ binding (see RBC CO2 Content and Acid-Base Status for further explanation).

When the intracellular []_app was referred to the [Hb₄], it was evident that the difference in []_app/[Hb₄] between deoxygenated and oxygenated RBCs was close to two (Fig. 4B), suggesting binding of two equivalents per Hb tetramer on deoxygenation. The change in []_app/[Hb₄] on deoxygenation was, however, pH dependent, being ~2.3 at pH_i 7.4 and ~1.5 at pH_i 7.2 (Fig. 4B). The slopes of the linear relationships between []_app/[Hb₄] and pH_i (Fig. 4B) revealed apparent Hb-specific buffer values of 15.1 and 11.2 mol · mol Hb₄¹ · pH unit¹ in oxygenated and deoxygenated RBCs, respectively.

Red cell pH_i did not change significantly with oxygenation degree, whereby the relationship between pH_i and pH_e was the same for oxygenated and deoxygenated blood (Fig. 5A). The linear relationship between pH_i and pH_e had a slope, pH_i/pH_e, of 0.60 (Fig. 5A). The distribution ratio of H⁺ across the red cell membrane (rH⁺ = [H⁺]_e/[H⁺]_i = 1) was linearly related to pH_e, with a slope of 0.50 (Fig. 5B). The apparent distribution ratio of bicarbonate ([]_{i app}/ []_{e cor}, where []_{e cor} is plasma [] corrected to mmol/l H₂O, assuming a plasma water content of 94%) was much higher than rH⁺ and r values for bicarbonate in deoxygenated blood were significantly above those in oxygenated blood (Fig. 5B). Due to the very high intracellular []_app (Fig. 4A), all apparent r values for bicarbonate, except for oxygenated blood at high pH_e, were >1, reflecting the unusual feature of larger intra- than extracellular values of []_app. The apparent r values for were linearly related to pH_e (Fig. 5B). In deoxygenated blood the slope was 0.52 (i.e., similar to that for H⁺), whereas it was 1.7 in oxygenated blood (Fig. 5B).

View larger version (11K): [in this window] [in a new window]   Fig. 5.   A: relationship between RBC pHi and pHe in oxygenated () and deoxygenated () alligator blood. B: pHe dependencies of the distribution ratio of H+ and the apparent distribution ratio of bicarbonate across the RBC membrane in oxygenated (open symbols) and deoxygenated (closed symbols) alligator blood. Means ± SE (n = 6).

Partitioning of Total CO₂ in Deoxygenated RBCs

View larger version (20K): [in this window] [in a new window]   Fig. 6.   Partitioning of total CO2 into various forms of carbon dioxide in deoxygenated alligator RBCs. Concentrations are given in mmol/l cell water. See Partitioning of Total CO2 in Deoxygenated RBCs for further explanation.

RBC , Cl, and H₂O Permeabilities

1

View larger version (12K): [in this window] [in a new window]   Fig. 7.   Typical examples of 36Cl, [14C], and 3H2O efflux from deoxygenated alligator RBCs under self-exchange conditions (30°C, pH 7.6). Ordinate represents fraction of tracer that remains in the cells at time t. at, a0, and a denote extracellular radioactivity at time t, time 0, and at equilibrium. Linearity of the curves in the semilogarithmic plot indicates that the efflux follows a monoexponential course, and the slope of the curves hence equals the negative value of the rate coefficient k (s1).

View larger version (13K): [in this window] [in a new window]   Fig. 8.   Transport parameters in oxygenated (hatched bars) and deoxygenated (solid bars) alligator RBCs. A: rate coefficients (k) for the unidirectional efflux of Cl, , and H2O under self-exchange conditions. B: times required for 63% equilibration (1/k). C: apparent permeabilities (P) for Cl at [Cl]e = 110 mM and at [] = 25 mM, and the diffusional H2O permeability. Means ± SE (no. oxygenated/no. deoxygenated: Cl, 14/10; , 11/10; H2O, 4/4).

The presence of DNDS at 2.5 mmol/l reduced k for unidirectional ³⁶Cl efflux by 94% and 30 µmol/l phloretin inhibited the transport by >99%.

The time required for 63% equilibration of the fluxes (i.e., 1/k) varied from 192 ms for Cl in oxygenated RBCs to 24 ms for water in oxygenated RBCs (cf. Fig 8B).

The apparent membrane permeabilities for Cl and and the diffusive H₂O permeability were evaluated from the k values and the measured MCV of 352 µm³, a fractional RBC water content of 0.64 (Fig. 2D), and an estimated membrane area (A_m) of 367 µm². A_m was evaluated by assuming that the elliptical alligator RBCs were geometrically similar to fish RBCs and then calculating A_m as the average of the two values resulting from considering the RBC as being either an ellipsoid or two elliptical surfaces separated by a marginal band (17). The apparent Cl permeability was ~4 µm/s in both oxygenated and deoxygenated RBCs (Fig. 8C). The P_app for bicarbonate was significantly higher, being approximately the double of P_Cl (Fig. 8C). The diffusive water permeability was significantly higher than the values for Cl and and the water permeability was significantly higher in oxygenated than in deoxygenated RBCs (Fig. 8C).

Hb H⁺ Binding Properties

1

1

View larger version (11K): [in this window] [in a new window]   Fig. 9.   Hydrogen ion equilibria of stripped alligator Hb in 0.1 mol/l KCl. A: net H+ charge (ZH, mol H+/mol tetramer) as function of pH in oxygenated and deoxygenated Hb solutions. B: buffer values (dZH/dpH, mol H+ · mol1 tetramer · pH unit1) for oxygenated and deoxygenated Hb as function of pH. Hb was kept oxygenated by equilibration with pure O2 and was deoxygenated with pure N2. Tetrameric Hb concentration = 0.21 mmol/l; temperature = 30°C.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Binding to Hb in Intact RBCs

The difference in the ratio []_app/[Hb₄] between deoxygenated and oxygenated RBCs was close to two but changed slightly with pH (Fig. 4B). Minor differences in the absolute value may have resulted from experimental uncertainties, because []_app/[Hb₄] was assessed from measurement of several different parameters. The value slightly higher than two (i.e., 2.3) at high RBC pH_i could, however, also reflect a minor oxygenation-linked carbamate formation (i.e., preferential reaction of CO₂ with uncharged -NH₂ groups in deoxygenated Hb). The value of []_app/[Hb₄] between deoxygenated and oxygenated RBCs decreased to 1.5 at pH_i 7.2 (Fig. 4B). This finding suggests that binds not only to deoxygenated Hb but also to oxygenated Hb when pH is decreased. Binding of occurs to the quaternary T-structure of the Hb. At the highest pH (lowest PCO₂) here examined, O₂ tension at half saturation (P₅₀) of alligator blood at 30°C is ~16 mmHg (extrapolated from data in Ref. 28). With this relatively high O₂ affinity, RBCs equilibrated with 98% air (PO₂ = 149.5 mmHg) primarily contain Hb with the R (oxy)-conformation, and oxygenation-linked binding is absent. Lowering of pH, however, gradually shifts the RT equilibrium toward the T-state. P₅₀ is ~40 mmHg (28) at the lowest pH here examined, and blood equilibrated with 93% air (PO₂ = 141.9 mmHg) therefore contains a significant quantity of T-structure Hb, leading to some oxygenation-linked binding in "oxygenated" RBCs.

The apparent distribution ratio of bicarbonate across the RBC membrane was higher than that for protons (Fig. 5B) because []_{i app} includes not only free but also carbamate and bound to the Hb. If, however, in deoxygenated RBCs, two molecules are bound to each Hb₄ molecule throughout the pH interval examined, then the slope of the relationship between r and pH_e should be similar for bicarbonate and protons. This indeed was the case (Fig. 5B). The steeper slope in oxygenated RBCs (Fig. 5B), can be attributed to a gradual binding to oxygenated Hb as pH is lowered. Binding of to Hb in oxygenated RBCs at low pH also influences the relationship between []_app and pH_i (Fig. 4A). Thus in oxygenated RBCs the value of d[]_app/dpH_i will be higher than that caused by nonbicarbonate buffering of H⁺ per se. In deoxygenated RBCs, d[]_app/dpH_i can be directly attributed to nonbicarbonate buffering.

H⁺ Binding Properties and Red Cell pH

1

1

1

1

The proton uptake on deoxygenation reached a maximum of 1.6 mol H⁺/mol Hb tetramer in alligator Hb. This value is higher than that of elasmobranch Hb, comparable to that of some mammalian Hbs, and lower than that of teleost fish (16). The Z_{H max} of 1.6 at pH 7.2 in alligator Hb is similar to the Z_H value of 1.51 in caiman Hb (1).

Normally, the uptake/release of H⁺ on deoxygenation/oxygenation causes pH_i changes, the magnitudes of which depend on the amount of H⁺ exchanged and the intracellular buffer capacity. In human blood pH_i between deoxygenated and oxygenated RBCs is 0.03-0.05 pH units (23) and in teleost (that have high Z_H values and low buffer values) it can be up to 0.35 pH units (15). In alligator RBCs, pH_i was not affected by changes in oxygenation degree (Fig. 5A). The important feature that distinguishes crocodilian RBCs from other vertebrates is the oxygenation-linked binding of both H⁺ and to the Hb. The amounts of H⁺ and taken up on deoxygenation (or released on oxygenation) are about equal in the alligator, which is also the case in the caiman (1). The oxylabile protons therefore balance the H⁺ formed in the CO₂ hydration reaction as result of binding to the Hb, and pH_i is not changed. Furthermore, if minor differences are present in the amounts of H⁺ and bound, the high intracellular buffer capacity will strongly limit pH_i changes.

Carbamate Formation

The number of carbamate groups in caiman Hb (at pH 7.45, PCO₂ = 49 mmHg, 37°C, and ionic strength 0.1 M) was reported to be 0.9/tetramer (1). The estimated number of carbamate groups in the alligator is approximately double that in the caiman. This difference between the two crocodile species cannot be explained by the total number of amino groups potentially available for carbamate formation. Caiman Hb has 46 Lys residues compared with 44 in alligator Hb, and the -amino groups of the -chains are not acetylated in the caiman (19). The NH₂-terminus of -chains is, however, involved in binding in both crocodilian Hbs, making carbamate formation unlikely at this site in both Hbs (21). A different carbamate formation in alligator and caiman Hb could result from differences in the molecular microenvironment of specific amino groups in the two Hbs, altering their pK_a values and thereby carbamate formation at a given pH. Alternatively, carbamate formation measured in intact RBCs (as with the alligator) may differ from that measured on purified Hb (as with the caiman).

The formation of carbamate does not appear to be oxygen-linked in caiman Hb (1). In the alligator, the major part of the carbamate was similarly not sensitive to oxygenation, although a minor oxygenation-linked component was indicated from the pH dependence of []_app/[Hb₄] between deoxygenated and oxygenated RBCs (cf. above).

Haldane Effect

The Haldane effect in alligator whole blood amounted to 0.72 mmol CO₂/mmol Hb monomer at constant physiological PCO₂, which is about double the value in human blood, but slightly lower than the value of 0.9 in Crocodylus porosus (12). The value in both crocodilian species is close to the respiratory quotient (RQ). On the assumption that the relationship between binding and Hb-O₂ saturation is linear, this implies that for every mole of O₂ released from Hb and consumed in tissue cells, the corresponding amount of CO₂ produced can be taken up by the blood without a change in PCO₂. The O₂ and CO₂ transport at rest can therefore be predicted to proceed with minimal changes in PCO₂ between arterial and venous blood. This also suggests the absence of significant pH changes (Fig. 5A) between arterial and venous blood. In many vertebrates, the CO₂-induced pH decrease in tissue capillaries helps drive off O₂ via the Bohr effect. In the crocodiles, this effect is taken over by allosteric binding.

RBC membrane permeability to Cl, , and H₂O

1

1

Information on band 3-mediated Cl and transport is available for another reptile, the sea turtle (24). The capacity for anion exchange in the RBC of this species appears as high as in alligators. The high anion permeability found in alligator RBCs (Fig. 8) leads to the conclusion that the unique feature of binding to Hb, which has evolved in the alligator, is not related to a low anion exchange capacity via band 3.

Human deoxygenated Hb binds to the cytoplasmic fragment of band 3 with a higher affinity than oxygenated Hb (27). This could potentially lead to an influence of oxygenation degree on band 3 anion transport kinetics (22). The kinetics of Cl and transport were, however, similar in oxygenated and deoxygenated alligator RBCs under equilibrium conditions (Fig. 8). This is also the case in human RBCs and in RBCs from four different teleost species (17).

Water. The diffusive water permeability, P_d, was 27 µm/s in oxygenated alligator RBCs, but significantly reduced to 17 µm/s by deoxygenation (Fig. 8C). The value of P_d of oxygenated cells is similar to P_d of intact human RBCs or resealed red cell ghosts, while P_d of deoxygenated cells is close to the P_d of 10-12 µm/s obtained in human RBCs after maximum inhibition of the water transporting proteins with p-chloromercuribenzenesulfonate (5). To our knowledge this is the first study showing that oxygenation/deoxygenation has an effect on P_d. The origin of the effect is not clear. Colombo et al. (9) showed an allosteric binding of some 60 extra water molecules to Hb when it switches from the fully deoxygenated (T) state to the fully oxygenated (R) state. The reduction of free intercellular water content is, however, small and also goes in the wrong direction to account for the change in P_d. Because the phenomenon apparently is not a general change of the solute permeability characteristics, it may be related to the transport pathways to water. By diffusion, water permeates the human RBC membrane through either the lipid phase or water transporting proteins in almost equal portions (5). If this also applies to alligator RBCs, deoxygenation may affect either the lipids or the proteinaceous water channels, closing either of the two pathways. A possible effect on the lipid pathway may be revealed by studying other RBCs such as chicken RBCs that lack the proteinaceous water channels (7).

CO₂ Transport in the Alligator

View larger version (15K): [in this window] [in a new window]   Fig. 10.   CO2 transport pattern in the alligator, showing the main processes and chemical reactions when CO2 is added to blood in tissue capillaries. Thick arrows, major pathways; thin arrows, minor pathways. See CO2 Transport in the Alligator for further explanation.

As blood passes through tissue capillaries, metabolically produced CO₂ diffuses into the RBCs, where most of the CO₂ is rapidly hydrated to carbonic acid that dissociates to and H⁺. The formed and H⁺ are next bound in approximately equal amounts to the Hb in parallel with the release of oxygen. This is evidenced by the similar magnitudes of oxygenation-linked (Fig. 4B) and H⁺ (Fig. 9A) binding and by the constancy in pH_i on desaturation of the Hb (Figs. 4 and 5A). Only a minor fraction of the is shifted to the plasma in exchange with Cl, as indicated by the insignificant rise in plasma C_T (Fig. 1) and [] (Fig. 3) on deoxygenation. This contrasts with the "mammalian strategy," where the contribution of anion exchange to elevate the CO₂-carrying capacity of blood is about equal to that of oxylabile H⁺ binding and where most of the formed in the RBCs is transported to the lungs as plasma (29).

An almost stoichiometric binding of both and H⁺ when alligator Hb is deoxygenated means that the net charge on the Hb molecule changes little with oxygenation degree. This also contrasts with the situation in other vertebrates, where oxylabile H⁺ binding decreases the nonpermeable negative charge carried by the Hb, giving rise to an increase in the Donnan-like distribution ratio of permeable ions and an associated swelling of the RBCs. Minimal oxygenation-linked changes in Hb charge in the alligator correlate with the absence of any influence of oxygenation degree on the distribution ratio of H⁺ (Fig. 5B) and on Hct and red cell [Hb] (Fig. 2), the latter reflecting the minimal water fluxes and cell volume changes. Relative constancy in the Donnan-like distribution of permeable ions means that the concentration of free changes little in both RBCs and plasma when CO₂ is taken up in tissue capillaries.

Dissolved CO₂ also changes little, because the magnitude of the Haldane effect is close to the RQ, allowing the produced CO₂ to be taken up with no significant change in PCO₂.

Carbamate formation can be predicted to contribute little to arteriovenous CO₂ transport in crocodiles, because oxylabile carbamate formation is small and because the PCO₂ change between arterial and venous blood is minor. In humans, oxylabile carbamate accounts for some 13% of the CO₂ exchange (18).

The bulk of CO₂ taken up in tissue capillaries is accordingly carried by the blood within the RBCs, and primarily as bound to Hb, until the blood enters the pulmonary capillaries where CO₂ is reformed and excreted (i.e., the reactions outlined in Fig. 10 proceed in the opposite direction). One other ancient vertebrate group, the lampreys, is also known to carry CO₂ primarily within the RBCs. In the lampreys, however, the mechanism is different from that in the crocodiles. Lamprey RBCs have a high pH_i, and the H⁺ uptake on deoxygenation is large. The resulting high amount of inside the RBCs exists as free and is not shifted to plasma, because the RBC membrane is devoid of an anion exchange system (20, 26).

Although anion exchange is fast in most vertebrate RBCs, it is normally considered rate limiting for CO₂ exchange during exercise, where capillary transit time goes down and where disequilibria may develop as a result of an insufficient time to reach equilibrium between intracellular and extracellular C_T before the blood leaves the capillaries. In this regard the "crocodile strategy" of carrying CO₂ inside the RBCs may be superior. The binding and release of