Blood Oxygen Equilibrium Curve Research Articles

Very high blood oxygen affinity and large Bohr shift differentiates the air-breathing siamese fighting fish (Betta splendens) from the closely related anabantoid the blue gourami (Trichopodus trichopterus)

The Siamese fighting fish, Betta splendens, and the blue gourami, Trichopodus trichopterus, are two closely related air-breathing anabantoid fishes. B. splendens is a sedentary facultative air breather frequenting often hypoxic waters, while T. trichopterus is a more active obligatory air-breather inhabiting better oxygenated waters. Despite their close taxonomic relationship, previous studies have shown inter-specific differences in both physiological and morphological plasticity. Consequently, we hypothesized that B. splendens would have the higher blood oxygen affinity characteristics typical of more hypoxia-tolerant fishes. Whole blood oxygen equilibrium curves were determined at 27 °C and pHs of 7.62, 7.44 and 7.25. At a pH of 7.62, the blood O2 affinity (P50) of B. splendens was just 2.9 mmHg, while that of T. trichopterus was ~5 times higher at 14.7 mmHg. There were no significant differences in P50 between males and females in either species. The Bohr coefficient in B. splendens and T. trichopterus was −1.79 and − 0.83, respectively. B. splendens, unlike T. trichopterus, showed a large Root effect. Hills cooperatively coefficient, n, was ~2 in both species, indicating a significant binding cooperative between oxygen and hemoglobin. Collectively, these differences in blood O2 transport characteristics in these two closely related species are likely correlated with the differing habitats in which they breed and inhabit as adults, as well as different activity levels. Finally, the very high blood O2 affinity of B. splendens is not extraordinary among air-breathing fish, as revealed by a review of the literature of blood oxygen affinity in air-breathing fishes.

Effects of anaerobic exercise accompanying catch-and-release fishing on blood-oxygen affinity of the sandbar shark ( Carcharhinus plumbeus, Nardo)

Recovery from anaerobic exercise is thought to be prolonged in elasmobranchs because they lack several mechanisms for maintaining or increasing oxygen delivery that are present in teleosts. For example, teleosts increase hematocrit and maximal blood-oxygen carrying capacity through red cell ejection from the spleen. Teleosts also counteract the reduction in hemoglobin oxygen affinity resulting from metabolic acidosis through an adrenergic-mediated increase in red cell Na +–H + exchanger activity. To begin to assess the consequences of anaerobic exercise accompanying catch-and-release fishing occurring within the estuarine nursery habitats of juvenile sandbar sharks ( Carcharhinus plumbeus, Nardo), we constructed blood-oxygen equilibrium curves using samples from individuals 1 h after capture by hook and line (exercise-stressed) and samples from fully-recovered animals maintained in a shore-side tank (control sharks). We also compared exercise-stressed and control sharks for hemoglobin concentration, hematocrit, red cell count, intracellular pH, and nucleoside triphosphate concentration ([NTP]). In contrast to results from previous studies on elasmobranchs, we found an elevation in both hematocrit (≈ 21%) and blood hemoglobin concentration (≈ 10%) in exercise-stressed sharks. There was also clear evidence of red cell swelling. Mean red cell volume was ≈ 28% higher and mean cell hemoglobin concentration was ≈ 10% lower in exercise-stressed sharks. Most important, in spite of significant metabolic acidosis (0.3 pH units), blood from exercise-stressed sharks had an oxygen affinity equivalent to that of blood from control sharks. This was a direct consequence of intracellular pH being alkalinized by approximately 0.15 pH units relative to plasma pH in exercise-stressed sharks. Our results using isolated hemoglobin solutions showed that the observed reduction (≈ 15%) in intracellular [NTP] also contributed to the leftward shift in the oxygen equilibrium curves. As expected, we found sandbar shark red cells to be unresponsive to exogenous catecholamines. Regardless, sandbar sharks appear able to prevent the decrease in blood-oxygen affinity resulting from anaerobic exercise (and the concomitant decreases in plasma pH), as has been well-documented in teleosts. Our results suggest, therefore, that oxygen delivery following exhaustive exercise is not necessarily compromised in juvenile sandbar sharks, and that hook and line capture and subsequent release do not increase rates of mortality, although both are yet to be directly confirmed.

Blood oxygen binding in double-muscled calves and dairy calves with conventional muscle conformation.

To assess in vivo blood oxygen binding in double-muscled calves and dairy calves with conventional muscle conformation. 58 dairy and 48 double-muscled calves. Calves were classified as neonatal (24 hours old) or older calves (2 to 26 days old). Venous and arterial blood samples were collected, and hemoglobin concentration, pH, PCO2, and PO2 were determined. Blood oxygen equilibrium curves (OEC) under standard conditions were constructed, and the oxygen exchange fraction (OEF) and the amount of oxygen released at the tissue level by 100 ml of blood (OEF Vol%) were calculated. In each breed, partial pressure of oxygen at 50% saturation of hemoglobin (P50) under standard conditions was significantly higher in older than in neonatal calves, indicating a right shift in OEC with age. Venous P50 was significantly lower in neonatal double-muscled calves than in neonatal dairy calves, but arterial and venous P50 were significantly higher in older double-muscled calves than in older dairy calves. In double-muscled, but not in dairy, calves, OEF was significantly higher in older than in neonatal calves. In neonatal calves, OEF Vol% was not significantly different between breeds, but OEF Vol% was significantly higher in older double-muscled calves than in older dairy calves. The lower OEF in neonatal double-muscled calves, compared with dairy calves, could contribute to the higher sensitivity of double-muscled calves to hypoxia. Blood oxygen affinity decreased with age, but OEF and OEF Vol% were unchanged with age in dairy calves, whereas they increased with age in double-muscled calves.

Effects of hyperchloremia on blood oxygen binding in healthy calves.

Three different levels of hyperchloremia were induced in healthy Friesian calves to study the effects of chloride on blood oxygen transport. By infusion, the calves received either 5 ml/kg of 0.9% NaCl (low-level hyperchloremia; group A), 5 ml/kg of 7.5% NaCl (moderate hyperchloremia; group B), or 7.5 ml/kg of 7.5% NaCl (high-level hyperchloremia; group C). Blood was sampled from the jugular vein and the brachial artery. Chloride concentration, hemoglobin content, arterial and venous pH, PCO2, and PO2 were determined. At each time point (0, 15, 30, 60, and 120 min), the whole blood oxygen equilibrium curve (OEC) was measured under standard conditions. In groups B and C, hyperchloremia was accompanied by a sustained rightward shift of the OEC, as indicated by the significant increase in the standard PO2 at 50% hemoglobin saturation. Infusion of hypertonic saline also induced relative acidosis. The arterial and venous OEC were calculated, with body temperature, pH, and PCO2 values in arterial and venous blood taken into account. The degree of blood desaturation between the arterial and the venous compartments [O2 exchange fraction (OEF%)] and the amount of oxygen released at tissue level by 100 ml of bovine blood (OEF vol%) were calculated from the arterial and venous OEC combined with the PO2 and hemoglobin concentration. The chloride-induced rightward shift of the OEC was reinforced by the relative acidosis, but the altered PO2 values combined with the lower hemoglobin concentration explained the absence of any significant difference in OEF (% and vol%). We conclude that infusion of hypertonic saline induces hyperchloremia and acidemia, which can explain the OEC rightward shift observed in arterial and peripheral venous blood.

Influence of age and breed on the binding of oxygen to red blood cells of bovine calves.

The influence of somatic growth and genetic selection on the whole blood oxygen equilibrium curve (OEC) was measured under standard conditions in double-muscled and dairy calves during their first 3 mo of life. Crossbreed animals were also investigated. Hemoglobin, 2,3-diphosphoglycerate (DPG), Cl, and Pi concentrations were also measured. The percentage of fetal hemoglobin (HbF) was determined. The influence of exogenous Cl, Pi, and pH on the OEC was also assessed. The PO2 at 50% hemoglobin saturation (P50) increased during somatic growth, probably because of the increase in DPG recorded in double-muscled neonates and to the progressive disappearance of HbF in both breeds. The oxygen exchange fraction (OEF%) was used to assess the combined influence of the OEC shift and OEC shape changes on blood oxygen desaturation under standard conditions, when the PO2 decreases within a physiological range. The OEF% showed an increase during the first month, then a stabilization. The effects of Cl, Pi, and pH in Friesian calves were similar as in adult cattle. Double-muscled neonates had a lower P50, OEF% values, and DPG concentrations and higher hemoglobin and Cl concentrations than Friesian neonates. The Pi concentration and the percentage of HbF were similar in both breeds. The pH and the Cl concentration had significantly less effect on the OEC in double-muscled than in Friesian calves. Crossbreed animals exhibited intermediate parameter values, between those recorded for double-muscled and Friesian calves. All differences between breeds progressively disappeared during the first month. These data show that blood function changes markedly in calves during the first month of life and that genetic selection can alter blood function.

Tissue Oxygenation in Patients with Hemoglobinopathy H

To evaluate, the degree of tissue hypoxia in patients with hemoglobinopathy H disease, whole blood oxygen affinity was estimated and analyzed in 33 patients. Twenty patients with iron deficiency anemia, matched for degree of anemia, served as controls. The results were as follows: Whole blood oxygen equilibrium curves of patients with HbH disease are biphasic because of a combination of the rectangular hyperbolic curve of HbH and the normal sigmoid curve of HbA and are shifted toward the left (P50 3.66 ± 0.33 kPa). Patients with iron deficiency anemia have right-shifted oxygen equilibrium curves (P50 4.02 ± 0.13 kPa) compared with normal. Oxygen release to the tissues in HbH disease is decreased (1.4 ± 0.3 mmol/L) as compared with iron-deficient patients (1.6 ± 0.2 mmol/L) with a similar degree of anemia. Red cell indices vary between the two groups. In patients with HbH disease the mean corpuscular hemoglobin concentration was 268 ± 17 g/L as compared with 294 ± 18 g/L in iron deficiency anemia. These findings indicate that whole blood oxygen affinity is a reliable index of tissue oxygenation in patients with hemoglobinopathy H.

Hb South milwaukee [β105 (G7) leu → Phe]: A newly‐identified hemoglobin variant with high oxygen affinity

Fifteen individuals among four generations of a family of English ancestry demonstrated elevated hemoglobin levels accompanied by leftward-shifted whole blood oxygen equilibrium curves. Five of the affected family members have required phlebotomies for relief of symptoms attributable to erythrocytosis. An abnormal hemoglobin or globin chain could not be isolated, but 43% of the beta chains of the affected individuals contained a Leu----Phe substitution at position 105 (G7). Oxygen equilibrium curves demonstrated a normal Bohr effect but decreased cooperativity.

The haemocyanin of the tarantula Lasiodora erythrocythara—the influence of CO 2, organic cofactors and temperature on oxygen affinity

1. 1. Whole blood oxygen equilibrium curves for the haemocyanin of the Honduran tarantula Lasiodora erythrocythara have for the first time been investigated in vitro. 2. 2. Oxygen affinity was pH dependent with a mean Bohr coefficient of −0.71 ± 0.06 between pH 7.3 and 8.1. At 25°C and pH 7.6 the P 50 value was 23 Torr. 3. 3. The fixed acid Bohr effect was not significantly different from the CO 2 Bohr effect indicating the lack of a specific effect of CO 2 on oxygen affinity. 4. 4. Dialysis had no measurable effect on haemocyanin oxygen affinity, indicating the absence of unidentified dialysable factors. The haemocyanin was also insensitive to changes in the concentration of d and l lactate, succinate, guanosine and urate. 5. 5. Temperature was found to have little influence on oxygen affinity between 20 and 25°C illustrated by the low Δ H value of −5.6 kJ mol −1. At the temperature extremes of 15 and 30°C Δ H values increased to −29kJmol −1. 6. 6. Cooperativity, expressed as n 50, was both temperature and pH dependent.

Biological effects of short-term, high-concentration exposure to methyl isocyanate. IV. Influence on the oxygen-binding properties of guinea pig blood.

Whole blood oxygen equilibrium curves (O2 ECs), blood buffer lines, and several hematologic properties were determined for adult guinea pigs exposed to 700 ppm methyl isocyanate (MIC) for 15 min. MIC inhalation effected a significant reduction of blood O2 affinity; the half-saturation pressure (P50) at 38 degrees C increased from the control (untreated) level of 22.8 +/- 0.1 mm Hg to values ranging from 28.5 to 43.7 mm Hg for experimental animals. MIC exposure had no apparent influence on O2 EC shape or CO2 Bohr effect. Erythrocyte volume, [metHb], O2 binding capacity, and combined red cell organic phosphate concentration (DPG + ATP) were not affected by MIC treatment. However, experimental animals experienced a severe metabolic acid-base disturbance; blood lactate concentration ranged from 8.6 to 24.0 mmole/L. Results indicate that lactic acidosis was solely responsible for increased blood P50 of MIC-treated animals. No direct effects of MIC on hemoglobin function were observed. Reduced Hb-O2 affinity, in conjunction with severe hypoxemia, compromised the guinea pigs' capacity for pulmonary O2 loading; at PaO2 of 30 mm Hg, Hb-O2 saturation (S) decreased from 66% S for controls to 42% S for MIC-treated animals.

Blood oxygen transport in the free-swimming hagfish, Eptatretus cirrhatus.

Arterial and mixed venous blood were sampled through chronically implanted cannulae from rested and swimming hagfish. PaO2 remained high when hagfish were swum for 15 min at a velocity of 20 cm s-1. PvO2 fell from 17.2 mmHg at rest to 3.5 mmHg after swimming, and the arteriovenous pH difference increased from 0.15 to 0.25 pH units. Whole blood oxygen equilibrium curves were essentially hyperbolic (Hill's n value = 1.38) and gave a half-saturation PO2 (P50) value of 12.3 mmHg at pH 7.8 and 16 degrees C. A CO2-Bohr factor (phi = delta logP50/delta pH) of -0.43 and a limited buffering capacity of the blood, amounting to approx. 4 slykes, were observed. The role of the blood in transporting oxygen and carbon dioxide both at rest and after swimming is established by in vivo blood gas measurements and in vitro oxygen-binding data. The low internal PvO2 at rest is close to the P50 measured under similar conditions and the hyperbolic equilibrium curve permits further oxygen unloading when PvO2 falls during swimming.

Effect of chronic cold and submergence on blood oxygen transport in the turtle, chrysemys picta

Whole blood oxygen equilibrium curves (O 2EC's) and related hematologic properties are reported for the turtle Chrysemys picta exposed to two experimental conditions. Summer turtles were maintained at 24°C with free access to air; winter turtles were submerged for 4–12 wk in N 2-bubbled water at 3°C. Half-saturation P O 2 's at 3°C for blood from summer and winter animals were 4.1 and 4.5 Torr, respectively. At 24°C, summerand winter P 50's were 20.2 and 22.7 Torr, respectively. The winter turtle P 50 values were lower than predicted since prolonged submergence effected a severe metabolic acidos; blood pH's for winter turtles were 0.65 pH unit lower than for summer animals at both temperatures. Cold submergence also had a profound influence on O 2EC shape. Winter turtle curves exhibited high O 2 affinity below P 50 while they were distinctly right-shifted above 50% S. Winter animals also exhibited reduced CO 2-Bohr coefficients (Δlog P O 2 /ΔpH) at 3 and 24°C. Prolonged submergence did not affect the animal's isohemoglobin profile (demontrated by isoelectric focusing) or [metHb]. The [ATP] and [DPG] in winter turtle red cells, however, decreased significantly; the ratio of organic phosphate ([ATP]+[DPG]) to Hb tetramer fell from 1.4 in summer animals to 0.5 in winter turtles. These findings suggest that the effect of chronic cold and prolonged submergence on turtle O 2EC position and shape may result from reduction in RBC organic phosphates. Furthermore, these observed changes in blood oxygen transport may facilitate O 2 loading during winter submergence via extrapulmonary gas exchange.

Oxygen affinity and equilibrium curve shape in blood of chicken embryos

Oxygen equilibrium curves of blood from 4- to 18-day chicken embryos were investigated at 38°C. Curves were recorded at the P CO 2 measured in the air cell of each egg. Since arterial P CO 2 is known to closely approximate that of the air cell, these curves reflect in vitro arterial pH. Curves were also recorded at a second, higher P CO 2 , allowig calculation of he curve at estimated venous P CO 2 and of physiological curves by interpolation. Half saturation P O 2 of physiological curves increased from 38 Torr at 4 days to 52 Torr at 8 days, and then decreased to 31 Torr at 198 days. Increasing blood oxygen affinity late in development favors oxygen loading at the falling P O 2 that exists at the chorioallantoic surface, while higher affinity before 8 days may be related to diffussion resistance of the inner shell membrane early in incubation. In blood from 4- to 6-day embryos, the Hill coefficient, n H, of curves at air cell P CO 2 increased from 1.5 at oxygen saturation 0.1 to 6.5 at saturation 0.85. After 6 days, n H steadily increased at low saturation and decreased at high saturation. By 18 days, n H varied only from 2 to 3.4. A biphasic equilibrium curve shape (hump at the low end of the oxygen equilibrium curve) developed in 4- and 5-day embryo blood after a period of storage on ice, and was accentuated if some of the cells were intentionally lysed.

Preface

Although William Harvey in 1628 used a truly comparative approach in his study on ‘The Movement of the Heart and Blood’, mentioning crustaceans, fishes, amphibians, reptiles and birds as well as mammals, physiologists now tend to be classified as either mammalian or comparative. One objective of the discussion meeting and this associated review volume was to bring together experienced scientists from either side of this arbitrary division who between them covered a large proportion of the animal kingdom. The major objective, however, was to view the separate central and peripheral mechanisms controlling the respiratory and circulatory systems, and then to see how these systems are controlled and co-ordinated during various types of behaviour and activities in order that an adequate supply of oxygen to the metabolizing tissues may be maintained, presumably with the least expenditure of energy.Such ‘goal-directed’ control and co-ordination may seem an obvious necessity during exercise, particularly for those air-breathing animals that spend much of their time underwater or for those that live at high altitude. However, a similar objective may be achieved during that most energy-conserving of all activities, sleep. The rhythmic ventilation of most mammals, which is generated by medullary neurones and which modulates output to the cardiovascular system, becomes irregular during desynchronized sleep. This irregular pattern may be controlled by a different central generator from that functioning in the awake animal, but its irregular nature is similar to that seen in a number of resting poikilothermic animals and in the marine mammals. In these animals wide fluctuations in blood gases are experienced, so the concept of internal respiratory homeostasis is not appropriate. Interestingly, the reflex control of circulation during sleep is mainly by way of the peripheral chemoreceptors, which is similar to the condition in the foetus where the baroreceptors do not function. However, the presence of a baroreflex in aquatic vertebrates indicates that it has not evolved solely in response to the influence of gravity on the circulatory system. The general level of arterial blood pressure may, in fact, be set centrally, with the peripheral receptors merely ‘fine-tuning’ the system.Respiratory and cardiovascular adjustments associated with some patterns of behaviour can be studied in traditional ways with anaesthetized or restrained animals. The defence-arousal response and a pattern of sympathetic activity similar to that seen during desynchronized sleep can be obtained by stimulating specific regions of the brain. Other types of activity cannot be so usefully studied under such conditions. It is now clear that the physiological responses to submersion (diving) are different when the animal dives spontaneously compared with forced submersion.At this meeting we heard from scientists studying topics as widely different as the connexions and interactions of central neurones, peripheral sense organs and whole animal activities in crustaceans, fishes, reptiles, birds and mammals (including man). At first sight this may appear an impossibly broad area to cover at a three-day meeting and in a relatively slim volume. In the event it was clear that important mechanisms for compensation are widespread throughout the animal kingdom (e.g. both the cray-fish, Austropotamobius, in hypoxic water and man at high altitude benefit from the leftward shift of the blood oxygen equilibrium curve that results from a respiratory alkalosis). It is hoped that the truly comparative approach to physiology will be fostered by this volume. As Harvey found, by studying a wide range of animals we are better able to understand the mechanisms which determine our own responses to the environment and, maybe, our aspirations within it.

Effect of piracetam on sickle erythrocytes and sickle hemoglobin

Piracetam, 2-oxo-1-pyrrolidine acetamide, inhibits sickling of red cells containing sickle hemoglobin (Hb S). The concentration required for 50% inhibition is about 300 mM. Addition of piracetam into the supersaturated Hb S solution in concentrated phosphate buffer prolongs the delay time prior to gelation. Piracetam shifts the oxygen equilibrium curves of blood toward the right, with a stronger effect at higher piracetam concentrations. Piracetam increases the viscosity of oxygenated cells but reduces the relative viscosity of deoxygenated sickle cells. The mechanism for the antisickling effect of piracetam will be discussed.

30] Continuous determination of the oxygen dissociation curve of whole blood

This chapter deals with continuous determination of the oxygen-dissociation curve of whole blood. Methods for the measurement of the hemoglobin oxygen-equilibrium curve (OEC) have evolved recently allowing the systematic investigation of properties of purified hemoglobin. The determination of the OEC under near-physiological conditions is of critical importance in the study of respiratory functions of blood. Although there are several new methods for continuously recording the OEC, few of them give proper attention to the careful control of the various factors that are known to affect the position and the shape of the curve. The chapter describes a method for the continuous recording of the whole blood OEC, which uses small samples. This method has the inherent advantage that fresh whole blood can be used; the entire curve can be measured in as little as 30 min from venipuncture; measurements of carboxyhemoglobin, methemoglobin (MetHb), and 2,3-diphosphoglycerate (2,3-DPG) can be made on the sample before and after the run; and carbon dioxide partial pressure (PCO 2 ) and pH are constant over the entire oxygenation range.

The effect of temperature on the oxygen equilibrium curve of human blood

The oxygen affinity of human blood was measured with the micro blood film technique over six-degree temperature intervals from 13 to 43°C. Results are expressed both as standard curves at pH 7.4 and for conditions of constant carbon dioxide content. P 50 (7.4) values found were 5.8, 8.3, 13.2, 19.7, 26.9 and 33.7 Torr at 13, 19, 25, 31, 37 and 43°C, respectively. The shape of the blood oxygen equilibrium curve was invariant with temperature. The measured CO 2-Bohr coefficient (Δlog P 50/ΔpH) ranged from −0.46 to −0.51 and was not temperature dependent. P 50 data for 7.4 followed a van't Hoff isochore with a slope, Δlog P 50/Δ(1/T K), of −2350 degrees and an enthalpy per mol of oxygen bound of −8.2 kcal/mol. Temperature dependence of blood oxygen affinity expressed for conditions of constant carbon dioxide content exceeds that of the standard curve. New data are used to calculate the temperature dependence of blood oxygen tension at constant total gas content as a functional of initial P O 2 and saturation.

The separate effects of H+ and 2,3-DPG on the oxygen equilibrium curve of human blood.

Addition of non-saturating amounts of 2,3-DPG (2,3-diphosphoglycerate) within the red cell (2,3-DPG/haemoglobin less than 1) initially reduces Hill's parameter, n. With increasing 2,3-DPG/haemoglobin, n increases until a maximum is reached at 2,3-DPG/haemoglobin greater than 1. Thus, 2-3-DPG influences the shape as well as the position of the whole blood oxygen equilibrium curve (OEC). The importance of this effect on the oxygen carrying capacity of the blood is considered. The effect of 2,3-DPG on the position of the OEC (p50, the pO2 at one-half maximal O2 saturation) is via its allosteric effect on haemoglobin at 2,3-DPG/haemoglobin less than 1. Above the ratio, its effect is to reduce intracellular relative to the extracellular pH.

Improved oxygen release: an adaptation of mature red cells to hypoxia.

Blood from patients with erythrocytosis secondary to arterial hypoxemia due either to congenital heart disease or to chronic obstructive pulmonary disease was shown to have a decreased affinity for oxygen; the average oxygen pressure required to produce 50% saturation of hemoglobin with oxygen was 29.8 mm Hg (average normal, 26.3 mm Hg). Such a displacement of the blood oxygen equilibrium curve promotes the release of oxygen from blood to the tissues. Studies were also performed upon blood from a man with complete erythrocyte aplasia who received all of his red cells by transfusion from presumably normal persons. With mild anemia (hematocrit, 28%), the affinity of his blood for oxygen was slightly diminished (an oxygen pressure of 27.0 mm Hg was required to produce 50% saturation of hemoglobin with oxygen). With severe anemia (hematocrit, 13.5%), however, his blood had a markedly decreased oxygen affinity (an oxygen pressure of 29.6 mm Hg was required to produce 50% saturation of hemoglobin with oxygen). We conclude that patients with various conditions characterized by an impairment in the oxygen supply system to tissues respond with a diminished affinity of their blood for oxygen. Although the mechanism which brings about this adaptation is not known, the displacement of the oxygen equilibrium curve is associated with an increase in heme-heme interaction. The decrease in blood oxygen affinity need not occur during erythropoiesis, but may be imposed upon mature circulating red cells.