Chapter 3: Mechanisms of O₂ and CO₂ Transport by Erythrocytes

The major function of erythrocytes or red blood cells (RBCs) is transport of the respiratory pigment hemoglobin (Hb). In their mature form, erythrocytes are biconcave discs 7-8 μm in diameter, with thickness of 0.5-2.0 μm, and an average volume of 80-90 μm³ or fL. Whole blood contains large numbers of RBCs, typically 4.5-6.0 × 10⁶/μL. Although living cells, mature erythrocytes are anucleate and contain few organelles. During a typical functional life span of 120 days, each RBC may travel 1,000 km from synthesis in the marrow until destruction in liver, spleen, or other site. Factors that regulate RBC production and longevity are beyond the scope of this book, except as they influence O₂ and CO₂ movements across alveolar septa and into blood vessels that originate in the right ventricle and traverse the lungs before returning to the left ventricle.

As part of a complete blood count (CBC) done on each hospital admission, the following three indices are routinely measured on a patient’s RBCs:

• Hematocrit—the fraction of whole blood that is composed of erythrocytes, most commonly expressed in percent. Historically, blood was placed in glass capillary tubes and centrifuged until the erythrocytes sedimented together at the bottom, with a "buffy coat" of leukocytes and platelets above them, and clear, amber-colored plasma on top. Normal values for hematocrit (Hct) vary with age and sex over predictable ranges (Fig. 3.1).

• Hemoglobin Concentration—the quantity of hemoglobin (Hb) per unit of blood, usually expressed as g/dL or simply g%. Historically, hemoglobin concentration ([Hb]) was determined by diluting whole blood with potassium ferricyanide to lyse the RBCs and oxidize ferrous ions (Fe²⁺) to ferric ions (Fe³⁺), thus converting normal Hb into methemoglobin (metHb) that cannot bind O₂. MetHb combines with cyanide ions to form a quantifiable color product, cyanomethemoglobin. [Hb] also varies with age and sex (Fig. 3.1).

• Red Blood Cell Count—the density of erythrocytes in whole blood, typically reported as ×10 ⁶/μL. Due to their great numbers and small size, whole blood is first diluted with 0.9% NaCl (normal saline). Historically, erythrocytes were counted manually by microscope on a calibrated glass slide (hemacytometer). Contemporary automated channel counters draw dilute red blood cell (RBC) suspensions through a narrow orifice, recording changes in saline conductivity between electrodes caused by passing erythrocytes. RBCs are both counted and sized in this way, yielding a histogram of diameters that is also reported as part of the CBC (Fig. 3.1).

These measured results for Hct, [Hb], and the number of RBCs from a patient’s CBC are then used to calculate three additional derived indices that describe various features of an average erythrocyte and establish useful normal ranges for each:

• Mean Cell Volume—the average volume of an erythrocyte, reported in fL (10^–15 L) or μm³ per RBC. Mean cell value (MCV) is computed by dividing the Hct of whole blood by the number of RBCs in that sample, with appropriate corrections made for differences in units. Among healthy adults, the Normal Range for MCV = 80-90 fL.

• Mean Cell Hemoglobin Content—the average mass of Hb in an erythrocyte, reported in pg. Mean cell hemoglobin content (MCH) is computed by dividing the [Hb] of whole blood by the number of RBCs in the sample, again correcting for differences in units. Among healthy adults, the Normal Range for MCH = 25-30 pg.

• Mean Cell Hemoglobin Concentration—the average [Hb] inside erythrocytes, properly reported as the g Hb/100 mL RBCs, or more conventionally just as a percent. Mean cell hemoglobin concentration (MCHC) is computed by dividing the [Hb] of a blood sample by its Hct, corrected as needed for differences in units. Among healthy adults, the Normal Range for MCHC = 32%-35%.

Based upon the CBC results for millions of presumably healthy adults, the erythrocyte dimensions mentioned above and the Hb content of typical mature RBC populations vary little with postnatal age. In other words, the calculated indices of MCV, MCH, and MCHC are remarkably constant across the broad population. Thus, most observed changes in a person’s Hct or [Hb] represent simple parallel variations in their number of RBC/μL, unless a specific genetic, biochemical, or physiological explanation is found.

Tetrameric Hb contains four ferrous (Fe²⁺) atoms, each of which can reversibly combine with one molecule of O₂ (Fig. 3.2). Under optimal conditions, one mole (M) of Hb can combine with 4 M of O₂. With a molecular weight of 64,458 g, Hb can reversibly bind approximately 1.39 mL of O₂/g Hb [derived from = (4) • (22.4 L O₂/M O₂)/(64,458 g/M Hb)]. This factor of 1.39 is used to compute the maximal O₂ carrying capacity of a sample when its [Hb] is known, assuming 100% oxygenation.

What is the physiological significance of 1.39 mL O₂/g Hb? Consider that at 37°C the maximal amount of dissolved O₂ = 2.33 mL O₂/dL blood/atm of pure O₂ (1 atm = 760 mm Hg). At sea level and by Dalton’s law of partial pressures (Chap. 1), we know that the ambient Po₂ in dry air = F_Io₂ • (barometric pressure, P_B) = (0.209) • (760) = 160 mm Hg. Thus, one deciliter of blood containing 15 g Hb/dL and equilibrated with air at sea level would contain 0.5 mL dissolved O₂ [= (2.33 mL O₂/dL blood/atm Po₂) • (160/760)], plus 20.8 mL O₂ reversibly bound to Hb.

The total O₂ content of any solution (including blood) depends upon the ambient Po₂ acting as the driving pressure. For saline and plasma, dissolved O₂ is proportional to ambient Po₂: Plasma equilibrated with 2 atm of O₂ contains twice as much dissolved O₂ as plasma equilibrated with 1 atm of O₂ at the same temperature. Respiratory pigments like Hb are different in two important ways. First, Hb binds and releases O₂ in a nonlinear manner. Second, oxygenation of Hb is saturable, so that once maximized, any further elevations in ambient Po₂ increase only the dissolved O₂ fraction, which is both linear and nonsaturable under normal clinical conditions that are encountered.

These phenomena are best studied by first equilibrating a blood sample to a gaseous environment containing no O₂, causing its Hb to become completely deoxygenated (less correctly termed "desaturated" or "unsaturated"). A suitable gas for this purpose is 95% N₂/5% CO₂, which yields by Dalton’s law a clinically meaningful partial pressure of CO₂ (Pco₂) = 40 mm Hg. As we then increase the Po₂ to which the blood is equilibrated (holding Pco₂ at 40 mm Hg), its O₂ content increases both in absolute terms (mL O₂/dL blood) and as a percent of maximal Hb oxygenation. The resulting oxygen dissociation curve (ODC) for whole blood is sigmoidal (Fig. 3.3). When comparing the ODC within or across species, the Po₂ required to half-oxygenate the blood is termed its P₅₀, a bit of physiological shorthand that has gained wide acceptance.

The sigmoidal nature of the ODC is due to changes in the shape of Hb molecules that occur with their binding of sequential O₂ molecules. Binding of the first O₂ (ie, 25% oxygenation) induces conformational changes within heme and globin subunits that facilitate cooperative binding of the second and third O₂ molecules (ie, 50% and 75% oxygenation). The symmetrical arrangement of the four heme sites suggests no preferential binding sequence, while binding of the fourth O₂ exhibits a more gradual approach to full Hb oxygenation. This cooperativity of binding causes the central portion of an ODC to be steeply sloped, so that large fluctuations in percentage of oxygenation and overall blood O₂ content can occur around the P₅₀ with only small changes in ambient Po₂. Myoglobin (Mb), an intracellular pigment with one Fe²⁺ and the mass of one-fourth of a Hb molecule, exhibits a simple saturable approach to 100% oxygenation, but with higher affinity and thus a lower P₅₀ (Fig. 3.4). The higher O₂ affinity of Mb illustrates an important physiological event: Oxygenated pigments with low affinity (high P₅₀) will surrender O₂ to any pigments of higher affinity that are not fully oxygenated. In this instance, oxygenated Hb will lose bound O₂ to deoxygenated Mb.

The shape, slope, and P₅₀ of an ODC are influenced by several physiologically important parameters. The most significant of these are pH and Pco₂, whose effects are related if not identical (Fig. 3.5). Acidification of blood with protons (hydrogen ion, H⁺) from any metabolic source reduces HbO₂ affinity and increases the measured P₅₀, the so-called acid Bohr effect. Indeed, fluctuations in P₅₀ due to Pco₂ are mediated primarily by pH changes caused when dissolved CO₂ interacts with H₂O to form weakly dissociative carbonic acid, H₂CO₃. Such Bohr effects are critical physiologically, since high hydrogen ion concentration ([H⁺]) and CO₂ levels in tissues acidify the arriving arterial blood, increasing its P₅₀ and facilitating dissociation of Hb-bound O₂. This O₂ transfer would occur even if tissue Po₂ equaled arterial Po₂, when in fact it is much lower. The reverse situation occurs as blood returns to the lungs. There, the high tissue CO₂ that caused acidification of venous blood is excreted, causing blood pH to rise. The P₅₀ decreases as a result, signifying higher HbO₂ affinity (alkaline Bohr effect) and a greater ability of Hb to bind alveolar O₂.

Interestingly, fluctuations in CO₂ levels also affect P₅₀ even if blood pH is held constant, causing what some authors have termed a pure CO₂ Bohr effect. In any case, the basis for the Bohr effects of both molecular CO₂ and protons is allosteric competition: [H⁺] and CO₂ both can bind reversibly to specific residues on Hb proteins that are remote from heme-coordinated Fe²⁺ binding sites for O₂. Association of [H⁺] or molecular CO₂ at those sites induces conformational changes in Hb that reduce its affinity of Fe²⁺ sites for O₂. As might be expected, the reverse of this allosteric competition is also true: Oxygenation of hemoglobin induces conformational shifts in the protein subunits that decrease their affinity for [H⁺]. Consequently, Hb becomes a stronger acid as it is oxygenated, a protein that more readily dissociates its bound H⁺. This Haldane effect is important in the lungs, where Hb oxygenation causes localized release of [H⁺] that are then available to combine with plasma and erythrocytic HCO^–₃ to reform H₂CO₃ and then gaseous CO₂.

Another important regulator of HbO₂ affinity in vivo is the erythrocytic intracellular concentration of the glycolytic intermediate, 2,3-diphosphoglycerate ([2,3-DPG]). At a normal concentration within RBCs of about 0.8 mol 2,3-DPG/mol Hb, the [2,3-DPG] reversibly binds with α-amino groups on β-globulin chains, having an allosteric effect analogous to those of CO₂ and [H⁺] by reducing HbO₂ affinity and raising the P₅₀ (Fig. 3.6). However, increased levels of intracellular [2,3-DPG] also makes O₂ uptake more difficult in the lungs, since [2,3-DPG] levels persist breath-to-breath, unlike CO₂ and [H⁺]. Thus, when [2,3-DPG] increases as sea-level inhabitants ascend rapidly to altitude, it does not facilitate O₂ uptake to the extent postulated when first described in 1968. More likely, acute altitude probably evokes the normal response to tissue hypoxia evident at sea level to increase erythrocyte [2,3-DPG] and improve tissue O₂ delivery without presuming a decline in inspired Po₂ as does occur at altitude. In some mammals lacking a distinct fetal Hb, the P₅₀ of fetal blood is kept below that of maternal blood by maintaining fetal erythrocytes at a low [2,3-DPG]. Thus, O₂ moves readily into the fetal circulation with its higher-affinity blood from the lower-affinity maternal circulation.

Core temperature is a final variable that influences HbO₂ affinity. As blood temperature increases, so does its P₅₀, analogous to the effect that thermal energy has on other chemical processes like the dissociation of weak acids. Although temperature gradients exist in the human body at rest and during exercise, it may be erroneous to attach much adaptive significance to this effect. In true poikilothermic (cold-blooded) vertebrates, however, such temperature effects have striking implications on seasonal patterns of ventilation and acid-base regulation.

A number of permanent changes in P₅₀ have been discovered that are due to minor changes in the globin chains of Hb. Most notable among these is that between fetal and maternal hemoglobins. In humans, the P₅₀ of whole blood composed of Hb-A is 26-27 mm Hg, while the P₅₀ of Hb-F in the fetus is about 20 mm Hg. Recent evidence suggests that most of this 6-7 mm Hg differential occurs because Hb-F is not modulated allosterically by [2,3-DPG]. In addition to such normal physiological shifts in P₅₀, there are scores of described hemoglobin variants in humans with altered P₅₀’s attributable to amino acid substitutions in the globin chains. Certain variants, including sickle cell hemoglobin (Hb-S) and hemoglobin Kansas, have abnormally low O₂ affinity and high P₅₀’s. Other mutations like those causing hemoglobin Rainier and hemoglobin Capetown show exceptionally high oxygen affinity and low P₅₀’s. Except where an amino acid substitution produces a specific functional abnormality in the tetramer, for example, the precipitative tendency of homozygous Hb-S at low Po₂, it remains unclear what impact such altered P₅₀’s have on systemic oxygen delivery.

Several final comments should be made concerning Hb and its O₂ transport capacity. While 1 g of Hb can theoretically combine with 1.39 mL O₂, the observed O₂ content of well-oxygenated blood seldom exceeds 97% of that predicted by the [Hb]. This is due to the occurrence of circulating Hb that cannot bind O₂ due to the presence of a competitive molecule at the O₂ binding site, or a modification of the site itself. An example of the former is carbon monoxide (CO) that reversibly binds to the normal HbO₂ site as carboxyhemoglobin (HbCO), but with very high affinity (P₅₀ ~1 mm Hg). Even low concentrations of inspired CO can effectively block O₂ transport by Hb. A different type of interference with Hb function is caused by strong oxidants like cyanide that convert Fe²⁺ to Fe³⁺ (see earlier discussion of [Hb] assay). The oxidized product methemoglobin (metHb) remains within erythrocytes but cannot reversibly bind O₂. Indeed, the conversion of Hb to metHb by common environmental oxidants like ozone is normally maintained at <2% of total [Hb] by the erythrocytic enzyme methemoglobin reductase. In poisoning cases, the production rate of metHb obviously exceeds the capacity of this reducing enzyme to reverse an ongoing oxidative process.

CO₂ produced by metabolically active cells must enter the vascular space as dissolved gas, with the majority entering the erythrocyte (Fig. 3.7). Once inside an RBC, CO₂ may: (1) remain dissolved as a gas for transport to the lungs (2) combine with intracellular water to form carbonic acid, H₂CO₃ that reversibly dissociates into bicarbonate anion, HCO₃⁻ and free protons, H⁺ or (3) combine reversibly with Hb to form carbaminohemoglobin, HbCO₂. During these potential excretion pathways for CO₂ created in active tissues, the erythrocyte provides: (1) the enzyme carbonic anhydrase (CA) to accelerate H₂CO₃ formation; (2) an exchange medium by which formed HCO^–₃ is transported in plasma after intracellular HCO^–₃ exchanges for extracellular Cl⁻, termed the chloride shift, favoring continued intracellular HCO^–₃ formation while maintaining electrochemical neutrality; (3) Hb as a proton acceptor to buffer H⁺ produced with HCO^–₃; and (4) Hb to act as a reversible sink for molecular CO₂ bound as HbCO₂. The last reaction simultaneously decreases HbO₂ affinity by the CO₂ Bohr effect. These several processes are reversed in the lungs as CO₂ is expired.

In absolute terms, about 91% of the CO₂ in mixed venous blood exists as HCO^–₃, with an additional 5% as HbCO₂ and 4% as dissolved gaseous CO₂ (Table 3.1). To determine how each of these pools contributes to the CO₂ excreted from the lungs, consider the data on differences in CO₂ contents between arterial and venous blood presented in this table. One critical point is quickly apparent: Although the total CO₂ content of venous blood is large, the amount actually excreted during a single passage through the lungs is small, here 1.83 mM or about 7% of the total. As will become apparent in Chaps. 9 and 17, the residual 93% of the blood’s total CO₂ constitutes an important source of labile (volatile) blood buffer. The plasma loss of HCO^–₃ as venous blood traverses pulmonary capillaries to become oxygenated arterial blood is 0.97 mM, with the HCO^–₃ loss from erythrocytes equal to an additional 0.49 mM.

Thus, the total amount of CO₂ excreted as HCO^–₃ from erythrocytes and plasma during alveolar perfusion is 1.46 mM, or roughly 80% of all CO₂ exchanged during ventilation. The remaining CO₂ excreted is about 11% as HbCO₂, and 9% via dissolved CO₂ from erythrocytes and plasma. Since Cl^– ions move between the erythrocyte interior and plasma in the opposite directions to these shifts in HCO^–₃ (Fig. 3.7), the fundamental contribution of RBCs to CO₂ transport is to accelerate HCO^–₃ production via intra-erythrocytic carbonic anhydrase, with Hb molecules acting as proton donors or acceptors. The contribution of HbCO₂ formation to overall CO₂ excretion is minor, but that benefit is magnified due to the advantageous shift in P₅₀ that occurs simultaneously whenever HbCO₂ is formed or dissociated (Haldane effect).

Two additional points are worth emphasizing from the data in Table 3.1. First, the amount of O₂ extracted from a liter of arterial blood, 2.20 mM, is of the same magnitude as the amount of CO₂ added to a liter of venous blood, 1.83 mM. In fact, these two numbers reflect the whole-body rates of O₂ consumption, V̇o₂ and CO₂ production, V̇co₂, respectively. (As mentioned in Chap. 1, placing a dot over a letter like V indicates a rate function, such as mL O₂ consumed per minute here.) The ratio V̇co₂/V̇o₂ is the respiratory quotient (R), here equal to (1.83/2.20) = 0.83. This topic will recur in Chap. 9. Second, the partial pressure gradient between arterial and venous bloods that drives this volume of O₂ into the body by diffusion is 60 mm Hg, or 10 times higher than the partial pressure gradient needed for a nearly equivalent amount of CO₂ to diffuse out of the body, here 6 mm Hg. This impressive difference in diffusional gradients is due to the 20- to 30-fold greater solubility of CO₂ in fluid and tissues versus O₂. That greater solubility derives from the reaction, CO₂ + H₂O ↔ H₂CO₃, that is propelled forward as long as H₂CO₃ continues to dissociate into H⁺ and HCO^–₃.