Carbon Dioxide Transport

Carbon dioxide diffuses out of metabolizing cells into capillary blood because tissue PCO₂ (roughly 46 mmHg) exceeds arterial PCO₂ (40 mmHg). Once in the blood, CO₂ does not simply stay dissolved. It partitions into three chemical forms, and the proportions differ from what casual intuition suggests: dissolved CO₂ obeys Henry's law and is a small fraction of the total, carbamino compounds form when CO₂ reacts directly with free amino groups on proteins (chiefly hemoglobin's N-terminal residues), and bicarbonate forms when CO₂ hydrates to carbonic acid, which then dissociates into a proton and bicarbonate ion. The equilibrium CO₂ + H₂O ⇌ H₂CO₃ ⇌ H⁺ + HCO₃⁻ is central to acid-base physiology generally, not just to CO₂ transport, because it is the same equation used in the Henderson-Hasselbalch relationship for blood pH. The key catalytic fact is that this hydration reaction is intrinsically slow — with a half-time of tens of seconds — but red blood cells are packed with carbonic anhydrase (CA), a zinc metalloenzyme that accelerates it by roughly three to four orders of magnitude, bringing the reaction to near-completion within the roughly 0.75-second transit time of blood through a systemic or pulmonary capillary. Because plasma has essentially no carbonic anhydrase, almost all of this hydration chemistry happens inside the erythrocyte, and the resulting bicarbonate must then be shuttled back out to plasma, where most of the CO₂ is actually stored in transit. This creates a transport problem the cell solves with a dedicated membrane protein, the anion exchanger AE1 (band 3), which trades outgoing HCO₃⁻ for incoming Cl⁻ — the chloride shift. The other core idea is that hemoglobin's conformational state is not a passive bystander in all this. Deoxygenated hemoglobin has a higher affinity for both protons and CO₂ than oxygenated hemoglobin, a reciprocal relationship to the Bohr effect (in which pH and CO₂ affect O₂ affinity). This reciprocal coupling is the Haldane effect, and it means that at the tissues, where hemoglobin is releasing O₂, it simultaneously becomes a better CO₂ and H⁺ sink — helping load CO₂ exactly where CO₂ production is highest. At the lung, where hemoglobin is picking up O₂, it becomes a worse CO₂ and H⁺ sink, helping unload CO₂ exactly where it needs to leave the blood.

Step 1 — CO₂ diffuses from tissue into plasma and then into the red cell. Tissue PCO₂ (~46 mmHg) exceeds capillary PCO₂ (~40 mmHg), driving diffusion down the gradient. CO₂ is roughly 20 times more soluble in plasma than O₂, so a meaningful fraction (~7% of total blood CO₂ content) simply stays dissolved and is carried as such; this dissolved fraction is what a blood-gas PCO₂ electrode actually measures.

Step 2 — Carbonic anhydrase inside the erythrocyte hydrates CO₂. Roughly 90% of the CO₂ entering blood diffuses into red cells (which have abundant CA, mostly the fast cytosolic isozyme CA II), where it is rapidly converted: CO₂ + H₂O → H₂CO₃ → H⁺ + HCO₃⁻ The reaction reaches near-equilibrium within the capillary transit time only because of this catalysis — uncatalyzed, it would take far longer than blood spends in a capillary.

Step 3 — Hemoglobin buffers the liberated protons. The H⁺ generated would otherwise acidify the cell steeply. Deoxyhemoglobin, however, is an excellent intracellular buffer: its imidazole side chains (histidine residues) pick up protons especially well when hemoglobin is in the deoxy (tense, T-state) conformation. Buffering H⁺ pulls the CA-catalyzed reaction further toward HCO₃⁻ production (Le Chatelier's principle), so oxygen unloading at the tissue directly potentiates bicarbonate formation — one thread of the Haldane effect.

Step 4 — Bicarbonate leaves the red cell via the chloride shift. As HCO₃⁻ concentration rises inside the cell, it is exported to plasma through the band 3 anion exchanger (AE1), a high-capacity electroneutral antiporter that exchanges one HCO₃⁻ out for one Cl⁻ in. This keeps the red cell electrically neutral and osmotically stable, and it means plasma — not the red cell cytoplasm — ends up storing most of the bicarbonate generated. Because Cl⁻ enters the red cell in venous blood, venous red cells contain slightly more Cl⁻ than arterial red cells (the classic Hamburger shift, named for Hartog Jacob Hamburger who described it in the 1890s), with a corresponding small osmotic water shift that makes venous red cells marginally larger than arterial ones.

Step 5 — Some CO₂ reacts directly with hemoglobin to form carbamino compounds. CO₂ can combine (without an enzyme) with uncharged terminal amino groups (mostly the N-terminal valines of the globin chains) to form carbaminohemoglobin: Hb–NH₂ + CO₂ ⇌ Hb–NH–COOH ⇌ Hb–NH–COO⁻ + H⁺ Deoxyhemoglobin binds roughly 3.5 times more CO₂ as carbamino compound than oxyhemoglobin does at the same PCO₂, because the T-state conformation exposes these amino groups more favorably — the second thread of the Haldane effect. Although only a minority of total blood CO₂ (~23% of the arterial-to-venous CO₂ difference) is carried this way, it is disproportionately important for the change in CO₂ content between tissue and lung, because carbamino binding is so sensitive to hemoglobin's oxygenation state.

Step 6 — At the lung, the entire sequence runs in reverse. Alveolar PCO₂ (~40 mmHg) is lower than mixed venous PCO₂ (~46 mmHg), so dissolved CO₂ diffuses out into alveolar gas. As hemoglobin binds O₂ and flips to the R-state (relaxed) conformation, it releases H⁺ and CO₂ from carbamino sites (Haldane effect in reverse), the freed H⁺ combines with HCO₃⁻ re-entering the cell (in exchange for Cl⁻ leaving — the reverse chloride shift) to re-form H₂CO₃, and carbonic anhydrase dehydrates this back to CO₂ and H₂O, which then diffuses into the alveolus to be exhaled. Quantitatively, in arterial blood, the approximate distribution of total CO₂ content (about 48 mL CO₂ per 100 mL blood) is:

Venous blood carries roughly 4 mL CO₂/100 mL more than arterial blood (about 52 vs 48 mL/100 mL) — a much smaller arteriovenous difference, proportionally, than the 4–5 mL O₂/100 mL swing seen in oxygen transport, but sufficient to carry the ~200 mL CO₂/min produced by resting metabolism given a cardiac output around 5 L/min.

Establishing that CO₂ is carried mainly as bicarbonate, not simply dissolved. In the late 19th and early 20th centuries, physiological chemists including Christian Bohr, John Scott Haldane, and later Joseph Barcroft used gas manometric techniques — extracting all the gas from blood samples under vacuum and measuring its volume and composition — to show that blood held vastly more CO₂ than physical solubility (Henry's law) could account for at physiological PCO₂. Plasma alone, treated with acid, released far more CO₂ than expected from dissolved gas, showing that most of it had been chemically bound as bicarbonate. This directly motivated the recognition that a chemical equilibrium, not simple physical dissolution, dominates CO₂ carriage.

Discovering carbonic anhydrase. In 1932-1933, Nathaniel Roughton and Charles Ashworth Roughton working with Karl Meldrum demonstrated that red blood cell hemolysates catalyzed CO₂ hydration dramatically faster than the uncatalyzed rate — the discovery of what came to be called carbonic anhydrase. Roughton's group used rapid-reaction techniques (flow methods capable of resolving reactions on sub-second timescales) to show that without the enzyme, CO₂ hydration in plasma was far too slow to reach equilibrium within a normal capillary transit time of well under a second, whereas red cell cytoplasm with intact enzyme reached equilibrium fast enough. This explained why nearly all bicarbonate formation must occur inside the erythrocyte rather than in plasma, and it directly predicted that carbonic anhydrase inhibitors (later exploited pharmacologically, e.g., acetazolamide) should slow CO₂ excretion and shift the CO₂ dissociation curve.

The chloride shift. Hartog Jacob Hamburger, in the 1890s, observed that when whole blood was equilibrated with CO₂, red blood cells swelled slightly and plasma chloride concentration fell while red cell chloride rose — the reverse occurring when CO₂ was removed. He correctly interpreted this as an exchange of bicarbonate leaving the cell for chloride entering, maintaining electrical neutrality (the Hamburger shift, still commonly called the chloride shift). This was one of the earliest demonstrations of a specific, reversible, coupled ion-exchange process across the red cell membrane, decades before the anion exchanger protein (band 3/AE1) was molecularly characterized in the 1970s–1980s using SDS-PAGE identification of red cell membrane proteins and subsequent transport reconstitution studies, and later cloned and sequenced, confirming Hamburger's ionic hypothesis at the molecular level.

The Haldane effect. John Scott Haldane and John Gillies Priestley, building on earlier Bohr-effect work, showed in the early 1900s that blood equilibrated with a given PCO₂ held substantially more total CO₂ when hemoglobin was deoxygenated than when it was oxygenated. Concretely, Haldane's manometric measurements demonstrated that at the same PCO₂, fully reduced (deoxygenated) blood carried measurably more CO₂ than fully oxygenated blood — establishing that oxygenation state, not just PCO₂, determines CO₂ content. This reciprocal relationship to the Bohr effect (in which CO₂/pH shifts the oxygen dissociation curve) revealed a two-way molecular coupling between the two gases carried on the same protein, later explained mechanistically by the differential proton and carbamino affinities of the T-state versus R-state hemoglobin conformations once the allosteric MWC/Perutz structural models of hemoglobin were developed in the 1960s–1970s.

Example 1 — Quantifying the Haldane effect's contribution to CO₂ unloading at the lung. Suppose arterial blood (PCO₂ 40 mmHg, hemoglobin ~97% saturated with O₂) carries 48.0 mL CO₂/100 mL blood, and mixed venous blood (PCO₂ 46 mmHg, hemoglobin ~75% saturated) carries 52.0 mL CO₂/100 mL blood. The total arteriovenous CO₂ difference is: 52.0 − 48.0 = 4.0 mL CO₂/100 mL blood Experimentally, if you compared the CO₂ content of blood at a fixed PCO₂ of 46 mmHg but forced full oxygenation (Haldane effect abolished), the content drops by roughly 1.5–2 mL CO₂/100 mL relative to the same PCO₂ at low saturation — meaning that somewhere around a third to a half of the physiological arteriovenous CO₂ difference is attributable not to the change in PCO₂ itself but to the change in hemoglobin's O₂ saturation acting through the Haldane effect. This is why measuring "CO₂ content vs PCO₂" curves separately for oxygenated and deoxygenated blood (as in the SVG above) is essential — a single dissociation curve, unlike the O₂ curve, is not enough; you must specify the saturation state. Clinically, this also explains a subtlety in mechanical ventilation and altitude physiology: giving supplemental O₂ to a patient with chronic CO₂ retention can, independent of any effect on respiratory drive, slightly reduce venous blood's CO₂-carrying capacity (by raising hemoglobin saturation), which is one contributing mechanism — along with reduced hypoxic drive and V/Q mismatch — behind the well-known rise in PCO₂ sometimes seen after oxygen therapy in severe COPD.

Example 2 — Reasoning through a carbonic anhydrase inhibitor problem. A patient is given acetazolamide, a carbonic anhydrase inhibitor, sometimes used clinically as a diuretic and in altitude sickness prophylaxis. Predict what happens to red blood cell CO₂ hydration kinetics and to renal bicarbonate handling. Reasoning: Acetazolamide blocks carbonic anhydrase both in red blood cells and in renal tubular epithelium (where CA normally supports H⁺ secretion coupled to HCO₃⁻ reabsorption in the proximal tubule). In red cells, inhibiting CA slows the CO₂ + H₂O ⇌ H⁺ + HCO₃⁻ reaction toward the uncatalyzed rate (a roughly 1000-fold to 5000-fold slowdown), which means less bicarbonate forms per unit time during a normal ~0.75-second capillary transit — so a larger fraction of CO₂ would have to be carried as dissolved gas or carbamino compound instead, and overall CO₂ unloading in the lung becomes less efficient (though the body compensates over time). In the kidney, blocking CA impairs H⁺ secretion into the tubular lumen, so filtered bicarbonate is not efficiently reclaimed and is lost in the urine, producing a mild metabolic acidosis alongside increased urinary Na⁺, K⁺, and bicarbonate excretion (a diuretic effect). This dual action — respiratory CO₂-handling slowdown plus renal bicarbonate wasting — is precisely why acetazolamide is used prophylactically in acute mountain sickness: the induced mild metabolic acidosis stimulates ventilation (via chemoreceptors sensitive to pH/PCO₂), helping counteract the respiratory alkalosis of hyperventilation at altitude and improving acclimatization.

Arterial and venous blood-gas measurements of PCO₂, HCO₃⁻, and pH are among the most common laboratory tests in medicine, used to diagnose respiratory versus metabolic acid-base disturbances in emergency departments, intensive care units, and anesthesiology. Carbonic anhydrase inhibitors (acetazolamide, dorzolamide) are used clinically as diuretics, in glaucoma to reduce aqueous humor formation, and as prophylaxis against acute mountain sickness, all exploiting the enzyme's dual roles in red cells, kidney, and eye. Understanding the Haldane and Bohr effects together is essential for interpreting oxygen and CO₂ exchange in patients on mechanical ventilators, in extracorporeal membrane oxygenation (ECMO), and in blood-gas-based management of chronic obstructive pulmonary disease, where impaired CO₂ elimination and altered bicarbonate buffering drive chronic respiratory acidosis with renal compensation. In exercise physiology and altitude medicine, the coupling between CO₂ transport chemistry and ventilatory drive (foreshadowing Control of Ventilation) explains hyperventilation responses to both exercise-generated CO₂ and hypoxia-triggered acclimatization. Comparative physiologists also use variations in this system — for instance, the differing carbonic anhydrase distribution and hemoglobin buffering in diving mammals, or the CO₂-handling adaptations of fish gills, discussed alongside Gas Exchange & Diffusion — to understand how different vertebrate lineages solve the same fundamental acid-load problem CO₂ transport poses.

Ask the tutor: How does the Haldane effect quantitatively change the shape and position of the CO₂ dissociation curve compared to the oxygen dissociation curve, and why does that difference matter for interpreting a blood gas?