Kinetics of Oxygen Uptake and Release by Red Blood Cells of Chicken and Duck

In recent years, the analysis of O₂ transfer kinetics of red blood cells (RBCs) using data obtained by stopped-flow techniques has been shown to be limited by resistance to O₂ diffusion in the medium surrounding the RBCs (Gad-el-Hak, Morton & Kutchai, 1977; Coin & Olson, 1979; Rice, 1980; Vandegriff & Olson, 1984). This realization has led to a general criticism of the stopped-flow method, but has also stimulated refinements in technology and modelling (Holland, Shibata, Scheid & Piiper, 1985 ; Yamaguchi et al. 1985).

In the present study the O₂ kinetics of avian RBCs has been studied for the first time to find out if the technique is suitable for avian RBCs and if there are major differences between avian and mammalian RBCs. Such differences could be expected on the basis of shape and size differences, the nucleated duck and chicken RBCs being larger than human RBCs and more spherical with a smaller surface area to volume ratio. These differences are expected to yield slower O₂ transfer kinetics for the avian, compared with the human, RBC.

The need for information on the kinetics of avian RBCs in O₂ transfer has arisen with the progress in the analysis of gas exchange in bird lungs (reviewed by Scheid, 1979), in particular in connection with studies aimed at analysis of gas—blood equilibration of O₂ (Scheid & Piiper, 1970; Burger, Meyer, Graf & Scheid, 1979; Geiser, Gratz, Hiramoto & Scheid, 1984) and the morphometry of gas exchange structures in avian lungs (Abdalla et al. 1982; Maina & King, 1982; Maina, 1984).

Blood was obtained by venipuncture from three ducks (domestic form of the muscovy duck, Cairina moschata, average body mass T9 kg) and three domestic hens (Gallus domesticus, average body mass T9 kg), was anticoagulated with heparin (5 i.u. ml^-1 blood Vetren^®, Promonta, Hamburg, FRG) and immediately stored on ice. Experiments were completed within 12 h after withdrawal of the blood.

Since the methods were the same as those described previously (Yamaguchi et al. 1985), only a brief account will be given.

The O₂ capacity of blood was determined from the O₂ content measured with a Lex-O₂-Con (Lexington Instrument Corporation, Waltham, MA, USA) in blood equilibrated with 50% O₂, subtracting physically dissolved O₂. Haematocrit was determined by centrifugation for 5 min in a microcentrifuge (M1100, Compur Electronic GmbH, Munich, FRG). The O₂ capacity of RBCs (mmol O₂l^-1 RBC) was calculated as O₂ capacity of blood divided by haematocrit. The RBC volume (mean corpuscular volume) was calculated from haematocrit and red cell number determined by an electronic cell counter (Coulter Counter^®, Model ZBI, Coulter Electronics, Luton, England).

The measurements of O₂ saturation kinetics of RBC suspensions were performed using a stopped-flow apparatus attached to a dual-wavelength spectrophotometer (Sigma-ZWS-II, Biochem, Munich, FRG) measuring the extinction difference between two wavelengths (560 and 577 nm). Each kinetic curve used for calculation of the specific O₂ conductance of RBCs, G, was based on eight successive measurements, which were averaged by a signal averager (Nicolet 1070, Nicolet Instruments Corp., Offenbach, FRG).

The measurements were performed at 41 °C, the average body temperature of the experimental animals.

Aliquots of RBC suspension obtained by diluting (1:50) whole blood in isotonic saline-phosphate-bicarbonate buffer were equilibrated with water vapour-saturated gas mixtures of a range of O₂ concentrations (0%, 5%, 6% and 8% for measurements in duck blood and 0%, 6%, 8% and 10% for measurements in chicken blood). Aliquots of the same buffer were equilibrated with gas mixtures of O₂ concentrations varied from 50% to 70% in such a manner as to achieve a similar final after mixing and stopped-flow equilibration. For all equilibrations, gas of constant CO₂ concentration of 5 % was used to yield close to 35 Torr and pH close to 7·5, corresponding to the values measured in arterial blood of the unanaesthetized, undisturbed chickens and ducks (Kawashiro & Scheid, 1975). For calibration, the initial and final values of were reproduced by mixing red cell suspension with buffer solution of identical ⁠. For an accurate measurement at ⁠, RBCs suspended in buffer containing sodium dithionite were used. Using the same technique, RBC suspensions and buffer equilibrated at identical ⁠, between 14 and 100 Torr, were mixed and was recorded to construct the actual blood O₂ dissociation curve which is required for the calculation of G.

RBC suspension was equilibrated with a gas mixture containing 50% O₂ and 5 % CO₂ (in N₂). This suspension of completely oxygenated RBCs was mixed with deoxygenated isotonic buffer solution containing 80 mmol l^-1 sodium dithionite (Na₂S₂O₄) the pH of which was titrated to 7·5 using 1moll^-1 NaOH at ⁠.

Table 1 contains mean experimental values for O₂ capacity, haematocrit and RBC concentration in chicken and duck, together with values obtained with the same technique in human blood (Yamaguchi et al. 1985). Derived from these measurements are the RBC O₂ capacity and mean RBC volume.

The mean values for P₅₀ and for Hill’s n in the range 0·3 < < 0·9 are 51·1 Torr and 3·3 for the chicken, and 49·0Torr and 3·3 for the duck.

Average G values for O₂ uptake by chicken and duck RBCs are plotted against time after flow stop in Fig. 1, and against in Fig. 2. In these figures, as in Fig. 3, the standard deviation is omitted for clarity; however, the maximum difference between corresponding measurements in different animals of one species was less than 10%. The following features are evident.

1. The initial G values are not clearly dependent on initial ⁠.

2. The G values decrease markedly with time or with increasing ⁠.

3. There is a tendency for the decline of G with time or with increasing to be steeper with higher initial ⁠.

4. The initial G values are higher for chicken RBCs than for duck RBCs, the ratio for G at corresponding initial values, 1·13 ±0·04 (mean ± S.E.), being significantly different from unity (P< 0·01).

The G values for O₂ release from RBCs into 40 mmol 1^-1 dithionite are plotted against in Fig. 3 (the plots against time are similar).

1. There is a tendency for G initially to increase with decreasing ⁠. This is followed by a plateau and a decrease to about 60 % of the plateau value at the end of desaturation.

2. The G values for chicken are higher than those for duck. In the plateau range (0·4 < < 0·8) G (mmol min^-1 Torr^-1 ml^-1 RBC). for chicken averages 0·33, for duck 0·25, and their ratio over the entire range is 1·29 ±0·01 (mean ± s.E.M.). The mean G values are plotted in Table 2 together with human data.

The deoxygenation kinetics of duck and chicken haemoglobin solution show time courses which are almost identical, and yield the following values for the rate constant k (s^-1): chicken, 172; duck, 186. For human haemoglobin solution (at 37 °C) the time course is very similar, k averaging 185 s^-1.

The values for O₂ capacity, haematocrit, RBC number and mean RBC volume (Table 1) for the chicken are within the range of values reported in the literature (Christensen & Dill, 1935; Morgan & Chichester, 1935; Chiodi & Terman, 1965; Sturkie, 1976; Baumann & Baumann, 1977; Steel, Petersen, Blanks & Smalley, 1977). Our O₂ capacity value for the duck is close to that previously reported by us (Scheid & Kawashiro, 1975; Holle, Meyer & Scheid, 1977; Meyer, Holle & Scheid, 1978). The mean RBC volume of the Pekin duck, about 180μm³ (Sturkie, 1976; Gaehtgens, Schmidt & Will, 1981) is somewhat higher than our value for the muscovy duck.

The P₅₀ value for chicken blood (51·1 Torr) is close to that reported by others (Chiodi & Terman, 1965 ; Baumann & Baumann, 1977; Hirsowitz, Fell & Torrance, 1977). The P50 value for the duck (49·0 Torr) is within the wide range of values reported for Pekin and muscovy ducks (cf. Meyer et al. 1978).

The Hill coefficient values reported for chicken and duck blood range from 2·6 to 3·3 (Christensen & Dill, 1935; Morgan & Chichester, 1935; Bartels, Hiller & Reinhardt, 1966; Scheipers, Kawashiro & Scheid, 1975; Wells, 1976), our values being at the upper margin. There was a definite tendency of Hill’s n to increase with ⁠.

The characteristics of the kinetics of O₂ uptake and release by RBCs as well as the deoxygenation kinetics of oxyhaemoglobin are in every respect similar to those reported for human RBCs by Yamaguchi et al. (1985). These authors concluded from the analysis of their results that O₂ transfer of RBCs was limited by diffusion both within RBCs and in the medium around the RBCs. The increase of the effective O₂ diffusion pathway by progressive depletion of the medium surrounding the RBCs was held to be responsible for the decrease of G with time during O₂ uptake. But even the initial value, determined for the time interval 0—5 ms after flow stop, was considered to be influenced by the extracellular O₂ gradient.

The important finding in human RBCs that with increasing dithionite concentration the O₂ release kinetics approached an upper limit (Yamaguchi et al. 1985) indicates that O₂ release into a medium with a sufficiently high dithionite concentration is not limited by diffusion or reaction kinetics (of Na₂S₂O₄ with O₂) in the medium. Thus in these conditions, diffusion inside the RBC is most probably limiting O₂ release.

C. Hook, K. Yamaguchi, J. Piiper & P. Scheid (in preparation) use the deoxygenation kinetics of human haemoglobin solution in model calculations and conclude that diffusion plays a more significant role in O₂ transfer in the RBC interior than in the reaction. This result is further supported in experiments in which the temperature and pH dependence of G are measured (K. Yamaguchi, J. Glahn, P. Scheid & J. Piiper, in preparation).

It appears justifiable to apply these conclusions to the avian RBC as well. The dependence of G on ⁠, particularly when using 40 mmol I^-1 dithionite, is very similar to that of human RBCs (Yamaguchi et al. 1985). Second, the deoxygenation kinetics of avian haemoglobin solution is not much slower than for human haemoglobin solution, even if the human k value is extrapolated to that expected at 41 °C using the Q₁₀ of K. Yamaguchi, J. Glahn, P. Scheid & J. Piiper (in preparation) whereby a value of k = 300 s^-1 is estimated, compared with the avian estimate of about 180 s^-1.

Thus, the plateau values for G for O₂ release may be considered as the best approximations to diffusive O₂ conductance of the RBC and to be valid also for O₂ uptake in a medium without external O₂ diffusion limitation, i.e. in a sufficiently stirred medium. These values are listed in Table 2 together with the corresponding value for human RBCs (Yamaguchi et al. 1985).

In the mammalian RBC, the cell membrane appears to play no important limiting role in diffusive O₂ transfer (Kreuzer & Yahr, 1960; Kutchai & Staub, 1969; Rotman et al. 1980; Huxley & Kutchai, 1985). Although experimental evidence is lacking, there is no reason to assume that this is not also true for the avian RBCs.

Table 2 shows a sequence in G values: G(duck) < G(chicken) < G(man). In this comparison, differences in the cell volume, and thus the number of RBCs per ml, must be considered. It is hence of interest to calculate and compare O₂ transfer conductance for a single RBC, g, which is obtained as G × V_RBC. The values of g in Table 2 show less variation among species than the G estimates. Can the remaining differences be attributed to differences in cell shape and size?

On the other hand, the experimental ratio g(man)/g(chicken) = 1·09, whereas the predicted ratio, using equation 3 and V_RBC from Table 1, is 0·94. This shows that the O₂ exchange conductance of human RBCs is about 15 % larger than would be predicted from the avian data on the assumption of isomorphy between avian and human RBCs.

On the other hand, differences in experimental O₂ exchange kinetics among mammalian species appear to be related to size differences of RBCs. Thus, Holland & Forster (1966) found the velocity constants of the initial rate of O₂ uptake in mammalian RBCs of widely differing size, to decrease as cell volume increases, and Jones (1979) obtained the best correlation of his data with the surface area to volume ratio, i.e. (V_RBC)^1//3. Thus, both for a comparison between mammalian species and between avian species the difference in g appears to be predictable on the basis of size differences in isomorphous bodies. This prediction appears, on the other hand, to be invalid for a comparison between mammalian and avian RBCs.

There are some differences between avian and mammalian RBCs which may have relevance to their O₂ exchange conductance. First, in comparison with the discoid mammalian cell, the shape of the avian RBC is more spherical, so that the surface area for a given volume, and hence the conductance g, is smaller.

Second, the presence of the nucleus in the avian red cell is expected to affect g in at least two ways. On the one hand, the haemoglobin concentration in the cytoplasm is higher than that calculated on the basis of the RBC volume, and therefore the O₂ diffusivity may be expected to be reduced. In fact, the values for O₂ capacity of avian RBCs must be corrected for the nucleus volume, 22 % in the chicken (Abdalla et al. 1982) and 19 % in the duck (Maina & King, 1982), which gives cytoplasmic haemoglobin concentrations of 23·5 and 23·8 mmol I^-1 cytoplasm, which are thus about 20 % larger than the mammalian value.

On the other hand, for a given cytoplasm volume, the cell surface area, and thus the area available for O₂ diffusion, is increased by the presence of the nucleus. In fact, the avian RBC may be modelled as a spherical shell containing haemoglobin, with a haemoglobin-free nucleus. C. Hook, K. Yamaguchi, J. Piiper & P. Scheid (in preparation) have shown that the g value for the spherical shell model is superior to that of the sphere (of equal volume). These differences show that in fact the avian RBC cannot simply be considered as isomorphous with the mammalian RBC, and that quantitative predictions of differences in g are difficult to obtain.

Use of G instead of θ, and of equation 6 instead of equation 4, has important advantages, particularly in comparative physiology. Whereas θ refers to a normal, standardized haematocrit, G accounts for the fact that the haematocrit is highly variable in an individual animal, according to the physiological state, and between individuals and species. Therefore, it is useful to single out the haematocrit, h, as a separate variable in the report of blood O₂ kinetics. The same objective would be attained by considering G for a single cell, g (see above), with the RBC count as a measured parameter.

Abdalla et al. (1982) have estimated the total volume of RBCs in chicken pulmonary capillaries as Q_c × h = 2·2 cm³. Using the estimate of the maximum D_tot for the chicken lung of Scheid & Piiper (1970), 67μmolmin^-1 Torr^-1, in equation 6 yields an estimate for D_m of about 74μmolmin^-1 Torr^-1 compared with the blood conductance, Q_c × h × G = 730μmol min^-1 Torr^-1. These estimates suggest that the gas/blood membrane of the parabronchial air capillaries offers a sizeably larger resistance to O₂ transfer than does the blood. For the duck, Q_c × h = 2·2 cm³ has been measured by Maina & King (1982), and D_tot about 100μmolmin^-1 Torr^-1 by Burger et al. (1979), yielding an estimate of 550 for Q_c × h × G and of 120 μmol min^-1 Torr^-1 for D_m. Again, the major resistance appears to reside in the membrane. These results are in contrast to the conclusion of the morphometrists (Abdalla et al. 1982; Maina, 1984). It should, however, be appreciated that morphometric and physiological techniques do not necessarily measure the same functional parameters. The morphometric estimates of diffusing capacity are based on functional parameters, like Krogh’s diffusion constant for O₂ in lung tissue and blood O₂ transfer kinetics which are unknown or were unknown to Maina & King (1984) and to Maina (1982) at the time of their study. On the other hand, measurements of D_tot are influenced by ‘functional inhomogeneities’ in the lungs (e.g. unequal distribution of ventilation to perfusion). If not taken into account, the inhomogeneities lead to an apparently decreased D_tot, as shown for avian lung models by Burger et al. (1979). Thus the calculated high Q_c × h × G/D_m ratios may be overestimated due to underestimation of D_tot.

The most important result of the present study is that the O₂ uptake and release kinetics of chicken and duck RBCs is similar to that of human RBCs. As in mammals the O₂ kinetics of avian RBCs appear to be size-dependent. For the single cell, the O₂ transfer conductance is smaller for the chicken than for the duck red cell, and this difference is expected since the ratio surface area/radius, which is important for O₂ diffusion in the cell, is smaller for the chicken RBC. For whole blood, the specific O₂ transfer conductance (i.e. O₂ conductance per unit RBC volume) is larger in the chicken than in the duck, and this is because the larger number of small cells in blood of a given haematocrit more than compensates for the smaller conductance of the single cell.