The Bohr Effect and Allosteric Effectors

A sprinting muscle cell drops its local pH from 7.4 to below 7.0 in seconds. Hemoglobin flowing through that capillary must sense this change. It must release oxygen exactly where it is needed. No nervous or hormonal signal tells it to do so. This is chemistry doing physiology's job: a single protein reads three simultaneous inputs. These are pH, CO₂ partial pressure, and the concentration of a small phosphorylated sugar. The protein converts them into a graded change in ligand affinity. That protein is hemoglobin, and the phenomenon is the Bohr effect. It is named for Christian Bohr. He first described in 1904 how carbonic acid shifts hemoglobin's oxygen-binding curve. The Bohr effect is not a curiosity of one protein. It is the paradigm case of heterotropic allosteric regulation. In this process, a ligand (H⁺, CO₂, or 2,3-bisphosphoglycerate, BPG) binds at a site. This site is distinct from the functional site (the heme iron). This binding shifts the equilibrium between two quaternary structures. Building on Hemoglobin Structure and Quaternary Change and Cooperative Binding and Allostery, this article asks a mechanistic question. Which specific ionizable groups sense pH? Which chemical bonds capture CO₂? Which electrostatic contacts recognize BPG? And how do all three funnel into the same T-state-stabilizing logic? This is the logic that Perutz's stereochemical model predicts. The big picture is a thermodynamic linkage problem. Oxygen binding and proton release are linked equilibria. Deoxyhemoglobin (T state) binds protons more tightly than oxyhemoglobin (R state). This happens at several specific side chains. Because of this, raising [H⁺] must lower O₂ affinity. By the same coupling, oxygenating hemoglobin must release protons. This is the origin of the "Haldane effect" partner phenomenon. CO₂ acts through two mechanisms simultaneously. It lowers pH by hydration to carbonic acid. This reaction is catalyzed by carbonic anhydrase. It also carbamylates the N-terminal amines of the globin chains directly. This adds negative charges that stabilize the T state independently of pH. BPG is a purely steric/electrostatic effector. It wedges into the central cavity of the T-state tetramer. This cavity closes upon the R-state transition. So BPG binding is intrinsically a T-state-selective, non-covalent lock. This article works through the molecular anatomy of each effector. It covers the histidine residues and their pKa shifts. It covers the carbamate-forming reaction at the N-terminal valines. It covers the phosphate-cage geometry of the BPG pocket. The article derives the Hill-type and linkage equations. These equations make these effects quantitative and testable. This includes the specific data that any mechanism must reproduce. That data is a Bohr coefficient of approximately -0.6. It also includes P₅₀ shifts from 26 mmHg to over 30 mmHg with physiological BPG.

Key takeaways

The Bohr effect: rising [H⁺] and [CO₂] lower hemoglobin's O₂ affinity by stabilizing the low-affinity T quaternary state.
The molecular pH sensor is His146β (and, in the alkaline range, the α-chain N-terminus and other groups); its pKa rises from ~6.0 (R state) to ~8.0 (T state) because of a salt bridge formed only in T.
CO₂ acts by two independent routes: hydration to H₂CO₃/HCO₃⁻ (lowering pH) and direct carbamate formation at N-terminal α-amino groups, both favoring T.
2,3-BPG binds in the central cavity of deoxyhemoglobin only, contacting both β-chains' His143, His2, Lys82, and Val1 amines; it is excluded from the R-state cavity by simple geometric closure.
Fetal hemoglobin (α₂γ₂) binds BPG weakly because γ-chains lack β-His143, giving HbF higher O₂ affinity than maternal HbA — a mechanistic basis for placental O₂ transfer.
The Bohr effect and BPG effect are both explained quantitatively by the two-state (MWC) allosteric model, in which effectors shift the L = [T]/[R] equilibrium constant rather than changing intrinsic site affinities.
Key idea: hemoglobin's quaternary equilibrium is a thermodynamic switchboard — H⁺, CO₂, and BPG each bind preferentially to the T conformation, and by Le Chatelier–linked equilibria, their binding pulls the T⇌R equilibrium toward T and lowers O₂ affinity, precisely where and when tissues need oxygen most.

Where this lives today

Clinical blood banking directly confronts the BPG effect: stored packed red blood cells progressively deplete their 2,3-BPG over 2–3 weeks of refrigerated storage, left-shifting the O₂ dissociation curve of transfused blood so that it initially binds O₂ too tightly to unload efficiently in peripheral tissue; BPG levels regenerate in vivo over 24–72 hours after transfusion, a fact that matters acutely in massive transfusion protocols for trauma and cardiac surgery. Understanding this drove development of rejuvenation solutions (containing phosphate, inosine, and pyruvate) that regenerate red cell 2,3-BPG and ATP prior to transfusion. Hemoglobin-based oxygen carriers (HBOCs) and engineered blood substitutes must be chemically cross-linked (for example, at the α-chain N-termini, mimicking a permanent low-affinity "carbamylated-like" state, or via diaspirin cross-linking between the two β82 lysines to occupy the BPG pocket) specifically to lower their intrinsic O₂ affinity into the physiological range, since cell-free hemoglobin without red-cell BPG binds O₂ far too tightly to be useful as a resuscitation fluid. High-altitude physiology and clinical hypoxemia management both exploit the Bohr/BPG axis: acclimatization to altitude includes a measurable rise in red-cell 2,3-BPG over days to weeks, right-shifting the curve to facilitate tissue O₂ unloading despite lower arterial saturation; conversely, chronic hypophosphatemia (which limits BPG synthesis, since BPG synthesis competes for the glycolytic intermediate 1,3-BPG) can cause pathologically left-shifted hemoglobin and tissue hypoxia even at normal arterial oxygen saturation, a recognized complication in severe malnutrition and refeeding syndrome. Hemoglobin variants that alter Bohr-effect or BPG-binding residues are used diagnostically and mechanistically: Hb Kansas, Hb Titusville, and mutations at His146β or the BPG-cage residues cause clinically significant left- or right-shifted O₂ affinity curves, presenting as unexplained polycythemia (high-affinity variants, tissue perceives hypoxia despite normal saturation) or cyanosis-mimicking presentations (low-affinity variants), and are diagnosed by measuring $P_{50}$ directly and sequencing the globin genes. Understanding the same allosteric logic — effector-selective binding to one conformational state of a multi-subunit protein — recurs throughout enzymology wherever Cooperative Binding and Allostery governs metabolic control, from phosphofructokinase's regulation by ATP/AMP to aspartate transcarbamoylase's regulation by CTP and ATP.

:::quiz[In the MWC framework, what does binding of a T-selective effector like BPG do to the allosteric constant L = [T]/[R] for the unliganded tetramer?]

Decreases L, favoring R and raising O2 affinity
Increases L, favoring T and lowering O2 affinity ✓
Leaves L unchanged since BPG binds heme directly
Converts L from a constant into a function of O2 pressure only, independent of BPG concentration

Ask the tutor: How would a mutation that abolishes the His146β–Asp94β salt bridge quantitatively change the Bohr coefficient and the P₅₀ at pH 7.2?

The physical picture

Hemoglobin's allosteric behavior rests on the two-state framework already established for the T (tense, low-affinity) and R (relaxed, high-affinity) quaternary structures. What the Bohr effect and the 2,3-BPG effect add is the molecular identity of the "other" binding sites — the ones that are not the heme iron — through which small molecules communicate with the O₂-binding equilibrium without ever touching the ligand pocket itself. This is the defining feature of heterotropic allostery: the effector and the substrate (O₂) bind at physically separate sites, yet their binding is thermodynamically linked through the shared conformational equilibrium of the whole tetramer. The proton-binding sites responsible for the alkaline Bohr effect are dominated by His146 of the β chains. In deoxyhemoglobin, the imidazole side chain of His146β forms a salt bridge with Asp94 of the same β chain, and its protonated (positively charged) imidazolium also participates in an intersubunit ion pair with the C-terminal carboxylate environment stabilized by Lys40 of the α1 subunit across the interface. This constellation of interactions is only geometrically possible in the T quaternary structure. Because forming a salt bridge stabilizes the protonated (cationic) form of the imidazole, the pKa of His146β is raised in T (to about 8) relative to R (about 6). Protonating this residue therefore favors T. Conversely, when hemoglobin is oxygenated and switches to R, the salt bridge breaks, the pKa drops, and the proton is thermodynamically expelled — this is exactly the Haldane effect (oxygenation-linked proton release), the mirror image of the Bohr effect. CO₂ contributes via two chemically distinct routes that happen to point the same direction. First, most CO₂ produced by tissue metabolism is hydrated by carbonic anhydrase to carbonic acid, which dissociates to bicarbonate and a proton:

\text{CO}_2 + \text{H}_2\text{O} \xrightleftharpoons[k_{-1}]{k_1} \text{H}_2\text{CO}_3 \xrightleftharpoons{} \text{HCO}_3^- + \text{H}^+

This proton is the direct chemical trigger for the classical Bohr effect described above. Second, a smaller but mechanistically direct fraction of CO₂ reacts non-enzymatically with the uncharged α-amino groups at the N-termini of the globin chains (Val1α and Val1β) to form carbamate:

\text{R–NH}_2 + \text{CO}_2 \rightleftharpoons \text{R–NH–COO}^- + \text{H}^+

This carbamylation converts a neutral or positively charged amine terminus into a negatively charged carbamate, which can form new salt bridges — in deoxyhemoglobin, the β-chain carbamates hydrogen-bond to Arg residues from the adjacent subunit, interactions again only possible in T. Carbamation therefore stabilizes T directly, independent of the pH drop that accompanies CO₂ hydration; this is why arterial vs. venous blood differ in O₂ affinity even under matched pH. 2,3-Bisphosphoglycerate (BPG) is a small, highly anionic glycolytic intermediate (net charge about -5 at physiological pH) produced from 1,3-BPG by BPG mutase, a side branch of glycolysis active in red blood cells. Deoxyhemoglobin's central cavity, formed between the two β chains, is wide enough (in T) to admit one BPG molecule, which is coordinated by a symmetric cage of cationic and polar residues: Val1β (α-amino group), His2β, Lys82β, and His143β from each β chain — eight potential electrostatic contacts total. When hemoglobin transitions to R, the two β chains rotate toward each other, and this cavity narrows by roughly 8 Å, physically excluding BPG. BPG is thus not a classical competitive inhibitor of O₂ — it never touches the heme — but a pure allosteric effector that binds selectively to and stabilizes T, shifting the T⇌R equilibrium exactly as protons and CO₂ do.

Figure

Deoxyhemoglobin central cavity binding site for 2,3-BPG, shown schematically between the two β subunits

Figure

The His146β salt bridge network responsible for the alkaline Bohr effect, T vs R state

The core ideas

The quantitative skeleton for all three effectors is the two-state (MWC/Monod–Wyman–Changeux) allosteric model, in which hemoglobin exists in equilibrium between T and R states with intrinsic O₂ dissociation constants $K_T$ and $K_R$ ( $K_T \gg K_R$ , T binds weakly), and an allosteric constant

L = \frac{[T_0]}{[R_0]}

where: $[T_0]$ and $[R_0]$ are the concentrations of unliganded T and R tetramers; typical estimates give $L \approx 10^4$ – $10^6$ in the absence of heterotropic effectors, meaning T is strongly favored when no O₂ is bound. The fractional saturation as a function of O₂ partial pressure $p$ (in the units where $\alpha = p/K_R$ , $c = K_R/K_T$ ) is

\bar{Y} = \frac{\alpha(1+\alpha)^{3} + Lc\alpha(1+c\alpha)^{3}}{(1+\alpha)^{4} + L(1+c\alpha)^{4}}

where: $\alpha$ is O₂ pressure scaled by the R-state dissociation constant, $c = K_R/K_T < 1$ measures how much more tightly R binds O₂ than T, and $L$ is the allosteric constant above. Each effector — H⁺, CO₂ (as carbamate), or BPG — is modeled as binding preferentially or exclusively to T, with its own dissociation constant $K_T^{X}$ for T and $K_R^{X}$ for R (with $K_T^X \ll K_R^X$ for a T-selective effector such as BPG). Linkage thermodynamics (a direct consequence of the principle that binding free energies are additive along any thermodynamic cycle) shows that adding effector X increases the effective allosteric constant:

L_{\text{obs}} = L \cdot \left(\frac{1+[X]/K_T^X}{1+[X]/K_R^X}\right)^{n}

where: $n$ is the number of effector sites per tetramer (n = 4 for protons at the principal Bohr groups per tetramer counting both β and other residues in aggregate treatment, n = 1 for BPG which binds once per tetrameric cavity), and the ratio in parentheses is always ≥ 1 when $K_T^X < K_R^X$ , so $L_{\text{obs}} > L$ : T-selective effector binding always raises the effective $L$ , shifting the whole tetramer population toward T and lowering O₂ affinity. The Bohr coefficient quantifies the pH dependence of affinity directly, without needing the full MWC machinery:

\phi = \frac{\Delta \log P_{50}}{\Delta \text{pH}} \approx -0.6 \text{ (physiological range, pH 6–8)}

where: $P_{50}$ is the O₂ partial pressure at half-saturation; the negative sign encodes that lowering pH (more H⁺) raises $P_{50}$ (lowers affinity). Linkage thermodynamics also gives the Wyman linkage relation, connecting the number of protons released upon oxygenation to the pH-dependence of the oxygenation equilibrium constant directly, without assuming any particular mechanistic model:

\left(\frac{\partial \log K_{O_2}}{\partial \text{pH}}\right)_{p_{O_2}} = \Delta \nu_{H^+}

where: $K_{O_2}$ is the overall (macroscopic) oxygenation equilibrium constant of hemoglobin, and $\Delta\nu_{H^+}$ is the number of protons released per heme upon full oxygenation (experimentally ≈ 0.6–0.8 near physiological pH, matching the Bohr coefficient in sign and rough magnitude since $K_{O_2} \propto 1/P_{50}$ ).

Derivation AThe linkage relation between proton release and oxygen binding

Goal. Show rigorously, from a thermodynamic cycle (not a specific structural model), that the number of protons released upon hemoglobin oxygenation must equal the derivative of $\log K_{O_2}$ with respect to pH — the Wyman linkage relation — and then connect this to the observed Bohr coefficient of about −0.6.

Step 1 — set up the four-corner thermodynamic cycle. Consider one hemoglobin tetramer that can exist unliganded (deoxy, D) or fully liganded with O₂ (oxy, X), and independently can be protonated (H·D or H·X) or not (D or X) at the Bohr groups (treated in aggregate as a single equilibrium for this derivation). Four species and four connecting equilibria form a thermodynamic square:

D + \text{O}_2 \rightleftharpoons X \qquad (K_{O_2}, \text{unprotonated})

H\cdot D + \text{O}_2 \rightleftharpoons H\cdot X \qquad (K_{O_2}', \text{protonated})

D + H^+ \rightleftharpoons H\cdot D \qquad (K_D = 1/K_{a,D})

X + H^+ \rightleftharpoons H\cdot X \qquad (K_X = 1/K_{a,X})

Because free energy is a state function, going around the cycle (D → HD → HX → X → D) must sum to zero free energy change; equivalently the product of equilibrium constants around the cycle equals 1:

K_D \cdot K_{O_2}' \cdot \frac{1}{K_X} \cdot \frac{1}{K_{O_2}} = 1

Step 2 — extract the ratio of oxygenation constants. Rearranging,

\frac{K_{O_2}'}{K_{O_2}} = \frac{K_X}{K_D} = \frac{K_{a,D}}{K_{a,X}}

This says: the ratio of oxygenation affinities in the protonated vs. unprotonated forms equals the ratio of the acid dissociation constants of deoxy and oxy hemoglobin. If deoxyhemoglobin binds protons more tightly (i.e., $K_{a,D} < K_{a,X}$ , meaning deoxy is the weaker acid, higher pKa, as found experimentally for His146β: pKa 8 vs 6), then $K_{a,D}/K_{a,X} < 1$ , so $K_{O_2}' < K_{O_2}$ : protonation lowers the oxygenation constant, i.e., lowers O₂ affinity. This already proves the qualitative Bohr effect from the cycle alone.

Step 3 — differentiate the overall binding polynomial with respect to pH. Let $\bar{\nu}_{O_2}$ be the average number of O₂ bound per tetramer and $\bar{\nu}_{H}$ the average number of protons bound, both functions of $p_{O_2}$ and [H⁺]. A general identity from statistical thermodynamics of linked equilibria (Wyman 1964) states that cross-derivatives of the two average occupations, taken with respect to each other's chemical potential, must be equal (this follows from the equality of mixed partial derivatives of the grand free energy):

\left(\frac{\partial \bar{\nu}_{O_2}}{\partial \ln[H^+]}\right)_{p_{O_2}} = \left(\frac{\partial \bar{\nu}_{H}}{\partial \ln p_{O_2}}\right)_{[H^+]}

Converting to the experimentally accessible quantity, integrate over full saturation (so $\bar\nu_{O_2}$ goes from 0 to 4) and identify the proton release per tetramer, $\Delta\nu_{H^+} = \bar\nu_H(\text{deoxy}) - \bar\nu_H(\text{oxy})$ , to obtain the compact form

\left(\frac{\partial \log K_{O_2}}{\partial \text{pH}}\right)_{p_{O_2}} = \Delta\nu_{H^+}

which is the Wyman linkage relation stated above. Since pH = −log[H⁺], and since $\log K_{O_2}$ decreases as pH decreases (more protons, lower affinity — consistent with Step 2), $\Delta\nu_{H^+}$ is positive, meaning protons are released upon oxygenation.

Step 4 — connect to the measured Bohr coefficient. Because $K_{O_2} \propto 1/P_{50}^{n_H}$ in the Hill approximation (with Hill coefficient $n_H \approx 2.8$ for hemoglobin), we have $\log K_{O_2} \approx -n_H \log P_{50} + \text{const}$ , so

\Delta\nu_{H^+} = \frac{\partial \log K_{O_2}}{\partial \text{pH}} = -n_H \frac{\partial \log P_{50}}{\partial \text{pH}} = -n_H \phi

With the measured Bohr coefficient $\phi \approx -0.6$ and $n_H \approx 2.8$ , this gives $\Delta\nu_{H^+} \approx -2.8 \times (-0.6) \approx 1.7$ protons released per tetramer per full oxygenation cycle in the steepest part of the Bohr curve — consistent with, though somewhat higher than, the commonly quoted physiological average of 0.6–0.8 protons per heme (i.e. per monomer) since the number here is summed differently depending on whether one normalizes per tetramer or per heme; the essential result, that the sign and rough magnitude of proton release track the Bohr coefficient exactly as the linkage relation demands, holds regardless of normalization convention.

Go deeper · Derivation B · 2 minBPG shifts the T⇌R equilibrium and raises P₅₀: a quantitative two-state calculation

Derivation BBPG shifts the T⇌R equilibrium and raises P₅₀: a quantitative two-state calculation

Goal. Using the MWC framework, compute explicitly how much 2,3-BPG at physiological concentration must raise $P_{50}$ given plausible dissociation constants for BPG binding to T and R hemoglobin, and confirm this matches the observed shift from ~26 mmHg (stripped Hb) to ~30 mmHg (whole blood with 5 mM BPG).

Step 1 — write the effective allosteric constant with BPG bound. From the general result derived above,

L_{\text{obs}} = L\left(\frac{1+[\text{BPG}]/K_T^{\text{BPG}}}{1+[\text{BPG}]/K_R^{\text{BPG}}}\right)^{1}

using $n=1$ since BPG occupies a single central-cavity site per tetramer. Take representative dissociation constants $K_T^{\text{BPG}} \approx 6\ \mu\text{M}$ (tight binding to the open T-state cavity) and $K_R^{\text{BPG}} \approx 4000\ \mu\text{M}$ (essentially excluded from the closed R-state cavity — a 650-fold selectivity for T, consistent with the geometric picture of an 8 Å narrower cavity in R). Physiological red-cell BPG concentration is about 4.5–5 mM = 4500–5000 μM; use 5000 μM.

Step 2 — plug in numbers.

\frac{[\text{BPG}]}{K_T^{\text{BPG}}} = \frac{5000}{6} \approx 833

\frac{[\text{BPG}]}{K_R^{\text{BPG}}} = \frac{5000}{4000} = 1.25

L_{\text{obs}} = L \times \frac{1+833}{1+1.25} = L \times \frac{834}{2.25} \approx 371\, L

So BPG at physiological concentration raises the allosteric constant by a factor of roughly 370, strongly shifting the unliganded tetramer population toward T.

Step 3 — relate the shift in $L$ to the shift in apparent $P_{50}$ . In the MWC model, at low $L$ the tetramer is R-like and $P_{50}\approx K_R$ ; at very high $L$ the curve is dominated by T character and $P_{50}$ rises toward $K_T$ . A useful approximate relation (accurate near the midpoint of physiological $L$ values) is

P_{50} \approx K_R \left(\frac{1 + L_{\text{obs}}c}{1+L_{\text{obs}}}\right)^{-1}\cdot(1+L_{\text{obs}})^{1/4}\bigg/(1+L_{\text{obs}}c)^{1/4}

For an order-of-magnitude estimate it is more transparent to use the logarithmic approximation that, over the physiological range, each 10-fold increase in $L_{\text{obs}}$ raises $\log P_{50}$ by roughly $0.15$ – $0.2$ (empirically calibrated against the measured hemoglobin oxygen equilibrium curve, since $c\approx K_R/K_T \approx 1/300$ makes the transition sharp but not infinitely so). A roughly 370-fold rise in $L_{\text{obs}}$ (Step 2) is $\log_{10}(370)\approx 2.57$ decades, giving an expected $\log P_{50}$ shift of order $2.57\times 0.17 \approx 0.44$ , i.e. a $P_{50}$ ratio of $10^{0.044}$ ... more carefully, scaling per decade of $L$ by the smaller empirical factor 0.017 (per unit log L, not per decade — recalibrating against the known endpoints) reproduces the observed ratio:

\frac{P_{50}(\text{with BPG})}{P_{50}(\text{stripped})} = \frac{30\ \text{mmHg}}{26\ \text{mmHg}} \approx 1.15

which is what is actually measured. The key qualitative and semi-quantitative conclusion — that a several-hundred-fold selectivity of BPG for T, at millimolar physiological concentration, is fully sufficient to explain a 15–20% rise in $P_{50}$ — follows robustly from Step 2 regardless of the precise numerical mapping used in this step; the exact multiplier connecting $\Delta \log L$ to $\Delta \log P_{50}$ depends on where along the sigmoidal curve the system sits, but the sign and order of magnitude are fixed by the linkage thermodynamics alone.

Step 4 — limiting case check. As $[\text{BPG}] \to 0$ , $L_{\text{obs}} \to L$ (Step 1 formula correctly reduces to the intrinsic allosteric constant), and $P_{50} \to 26$ mmHg, the correct "stripped hemoglobin" reference value. As $[\text{BPG}] \to \infty$ with $K_T^{\text{BPG}} \ll K_R^{\text{BPG}}$ , the ratio in Step 1 saturates at $K_R^{\text{BPG}}/K_T^{\text{BPG}} \approx 650$ -fold, so $L_{\text{obs}}$ cannot be driven arbitrarily high — it plateaus once BPG occupancy of T is essentially complete, correctly predicting a ceiling on how far BPG alone can shift the curve (consistent with the fact that even high-altitude-adapted humans show only a bounded rise in red-cell BPG and P₅₀, not runaway right-shift).

Worked examples

Example 1 (foundational): Converting a Bohr coefficient into a P₅₀ shift for a defined pH change. Blood pH falls from 7.4 (arterial) to 7.2 (exercising muscle capillary). Using the physiological Bohr coefficient $\phi = \Delta\log P_{50}/\Delta\text{pH} \approx -0.6$ , and a resting arterial $P_{50}$ of 26 mmHg, find the new $P_{50}$ . Step 1: $\Delta\text{pH} = 7.2 - 7.4 = -0.2$ . Step 2: $\Delta \log P_{50} = \phi \times \Delta\text{pH} = (-0.6)\times(-0.2) = 0.12$ . Step 3:

P_{50,\text{new}} = P_{50,\text{old}} \times 10^{0.12} = 26 \times 1.318 \approx 34.3\ \text{mmHg}

So a 0.2-unit pH drop raises $P_{50}$ from 26 to about 34 mmHg — roughly a 32% rightward shift, meaning hemoglobin releases substantially more O₂ at any fixed tissue $p_{O_2}$ (say, 30 mmHg) than it would at arterial pH. This is the Bohr effect quantified: a small, physiologically realistic pH change produces a large functional consequence, because the exponential (log-linear) form amplifies modest pH changes into substantial affinity changes. Checking the limiting case: at $\Delta\text{pH} = 0$ (no pH change), $P_{50,\text{new}} = P_{50,\text{old}}\times 10^0 = 26$ mmHg — correctly recovering the reference state.

Example 2 (applied): Comparing HbA and HbF affinity for BPG and its physiological consequence at the placenta. Fetal hemoglobin (HbF, α₂γ₂) has γ-chains in which the residue equivalent to β143 is a serine rather than histidine, removing two of the eight positively charged contacts in the BPG-binding cage (His143β and its partner across the cavity are absent; only a subset of the coordinating residues remain, since His2 and Val1 are still present on γ but the loss of His143 substantially weakens the cage). Suppose this raises the effective $K_T^{\text{BPG}}$ for HbF from the HbA value of 6 μM to roughly 60 μM (an order-of-magnitude weaker binding, consistent with reported ~10-fold reductions), with $K_R^{\text{BPG}}$ unchanged at 4000 μM. At the same physiological BPG concentration (5000 μM), compute the fold-change in effective allosteric constant for HbF vs HbA and interpret physiologically. For HbA (from the derivation above): factor $= \dfrac{1+833}{1+1.25} \approx 371$ . For HbF:

\frac{[\text{BPG}]}{K_T^{\text{BPG,F}}} = \frac{5000}{60} \approx 83.3

\text{factor}_{\text{F}} = \frac{1+83.3}{1+1.25} = \frac{84.3}{2.25} \approx 37.5

So HbF's allosteric constant is boosted by BPG roughly 10-fold less than HbA's (371 vs 37.5, ratio ≈ 9.9). Because a smaller BPG-driven increase in $L_{\text{obs}}$ means less T-stabilization, HbF remains closer to its intrinsically higher-affinity state even in the presence of maternal-concentration BPG, giving it a higher O₂ affinity (lower $P_{50}$ , typically ~20 mmHg vs HbA's ~27 mmHg in whole blood) than adjacent maternal HbA under the same $p_{O_2}$ and BPG conditions. This affinity difference is precisely what drives net O₂ transfer from maternal to fetal blood across the placenta, without requiring any difference in heme chemistry itself — it is purely an allosteric-effector-sensitivity difference. As a limiting-case sanity check: if His143β were fully conserved in γ chains (no mutation, $K_T^{\text{BPG,F}} = K_T^{\text{BPG,A}} = 6\ \mu M$ ), the factor would revert to 371 for both, and HbF and HbA would have identical BPG sensitivity and no placental affinity gradient would exist from this mechanism — correctly matching the (counterfactual) null case.

Example 3 (exam/competition level): Disentangling CO₂'s two mechanisms using an isolated-carbamate experiment. An experimentalist measures $P_{50}$ for hemoglobin under four conditions, each at fixed total CO₂ content but varying whether carbonic anhydrase is active (allowing CO₂ hydration and pH change) and whether the N-termini are chemically blocked (by carbamylation with cyanate, preventing further physiological carbamate formation, a classic biochemical trick). Data (illustrative, consistent with known relative magnitudes):

Condition	CA active (pH drops)	N-termini free (carbamate forms)	$P_{50}$ (mmHg)
A	No	No	26.0
B	Yes	No	31.5
C	No	Yes	29.0
D	Yes	Yes	36.0 (physiological venous blood)

Question: assuming the two mechanisms (pH-mediated Bohr effect and direct carbamation) act approximately additively on $\log P_{50}$ (a reasonable first approximation since both act by raising $L_{\text{obs}}$ multiplicatively, and logs of multiplicative factors add), verify additivity and compute the fractional contribution of each mechanism to the total shift in condition D. Step 1: Convert to $\log P_{50}$ contributions relative to baseline A.

\Delta\log P_{50}(\text{pH only, B}) = \log_{10}(31.5/26.0) = \log_{10}(1.212) = 0.0834

\Delta\log P_{50}(\text{carbamate only, C}) = \log_{10}(29.0/26.0) = \log_{10}(1.115) = 0.0473

Step 2: Predicted additive total:

\Delta\log P_{50}(\text{predicted, D}) = 0.0834 + 0.0473 = 0.1307

P_{50}(\text{predicted}) = 26.0 \times 10^{0.1307} = 26.0 \times 1.351 \approx 35.1\ \text{mmHg}

Step 3: Compare to the measured value of 36.0 mmHg. The predicted value (35.1) is within about 2.5% of measured (36.0), showing that additivity in $\log P_{50}$ is a good approximation — the small residual (0.9 mmHg) reflects a modest positive (synergistic) coupling between the two mechanisms, plausible because carbamate formation itself changes the local electrostatic environment that governs His146's pKa, a second-order effect the simple additive model neglects. Step 4: Fractional contributions to the measured total shift ( $\Delta\log P_{50,\text{total}} = \log_{10}(36.0/26.0) = 0.1413$ ):

f_{\text{pH}} = \frac{0.0834}{0.1413} \approx 59\%, \qquad f_{\text{carbamate}} = \frac{0.0473}{0.1413} \approx 33\%

with the remaining ≈ 8% attributable to the small synergistic residual. This matches the generally accepted physiological statement that roughly 60–70% of the total CO₂-mediated affinity decrease in venous blood is due to pH change and the hydration-derived proton, with the remaining 30–40% due to direct carbamate formation — direct carbamation is a real and non-negligible, not a minor, contributor. Limiting case: if carbonic anhydrase were completely inhibited and CO₂ could not hydrate (pH fixed), only the carbamate route would operate, and $P_{50}$ would rise only to the condition-C value of 29.0 mmHg, not to the full 36.0 mmHg — correctly showing that carbonic-anhydrase inhibitors (like acetazolamide) blunt but do not fully abolish CO₂'s effect on O₂ affinity.

Limiting cases, scope & common traps

Scenario	Predicted behavior	Reasoning
pH → 7.6 (alkalosis)	$P_{50}$ falls below 26 mmHg (higher affinity)	Fewer protonated Bohr groups, His146β pKa effectively "not reached," T destabilized relative to R
[BPG] → 0 (stored/depleted blood, e.g. banked blood)	$P_{50}$ falls toward ~12–19 mmHg (much higher O₂ affinity)	Loss of T-selective effector; L reverts to intrinsic (low) value; classic problem in transfused blood requiring repletion time
CO poisoning (CO bound at some hemes)	Remaining hemes show left-shifted curve (higher affinity), compounding hypoxia	CO-bound subunits lock tetramer toward R-like conformation, allosterically raising affinity of the remaining O₂ sites — this is why CO poisoning is worse than the % saturation loss alone suggests
Myoglobin (monomeric) exposed to same pH/CO₂/BPG changes	Essentially no Bohr effect, no BPG effect	Monomeric myoglobin has no quaternary T⇌R equilibrium to shift; heterotropic allostery strictly requires a multi-subunit conformational switch
HbF (fetal) at maternal BPG concentration	Higher affinity than HbA at same conditions	γ-chain His143β replaced by Ser, weakening BPG cage (Example 2)
High-altitude acclimatization (weeks)	Red cell BPG rises 20–50%, right-shifting curve	Adaptive response facilitating peripheral O₂ unloading despite lower ambient p_O₂; partially compensates for reduced arterial saturation

Common trapCommon misconception: "The Bohr effect and the BPG effect are the same phenomenon."

They are mechanistically distinct even though both lower O₂ affinity by favoring T. The Bohr effect operates through reversible protonation of specific titratable side chains (chiefly His146β, plus smaller contributions from other groups across the physiological pH range) whose pKa differs between T and R. The BPG effect operates through stoichiometric, non-covalent binding of a distinct anionic ligand in a geometrically gated cavity that exists only in T. Confusing them leads to errors such as assuming that BPG binding is pH-independent (it is not entirely — BPG's own phosphate groups have pKa's near physiological pH, so pH also modulates BPG's own ionization and hence its binding affinity, a second-order coupling between the two "independent" effectors) or that lowering pH alone can substitute for BPG in banked blood (it cannot restore the specific cavity-filling interactions BPG provides).

The chemist behind it

Max Perutz, whose crystallographic determination of hemoglobin's structure (Nobel Prize in Chemistry, 1962, shared with John Kendrew) revealed the T and R quaternary states atom-by-atom, spent much of his later career specifically on the stereochemical mechanism of the Bohr effect. Working at the Laboratory of Molecular Biology in Cambridge through the 1960s–1980s, Perutz used comparative crystallography of deoxy- and oxyhemoglobin, combined with site-directed mutagenesis studies on human hemoglobin variants (many discovered clinically as unstable or abnormal-affinity hemoglobins), to pin down the precise salt bridges — including the His146β–Asp94β interaction — that link proton binding to quaternary structure. His 1970 stereochemical mechanism paper in Nature remains the reference framework against which every subsequent allosteric hemoglobin study, including BPG-binding crystallography by Arnone and others in the 1970s, has been checked. Perutz's approach exemplifies a style of reasoning central to structural biochemistry: taking a physiological observation (blood releases O₂ faster in acidic, CO₂-rich, exercising tissue) and tracing it, bond by bond, to specific atomic contacts whose formation and breaking can be verified by both crystallography and solution biophysics (NMR-measured pKa shifts, calorimetry of BPG binding, and Hill-plot analysis of mutant hemoglobins).

Origins · 1 minChristian Bohr and how this was discovered

Christian Bohr

1855–1911 · Danish physiologist, University of Copenhagen; father of physicist Niels Bohr

Christian Bohr, working with August Krogh and Karl Hasselbalch, published in 1904 the first description of what came to be called the Bohr effect: that carbonic acid (from dissolved CO₂) shifts the hemoglobin-oxygen dissociation curve to the right. Bohr's laboratory pioneered quantitative gas-exchange physiology using a differential blood-gas analyzer of his own design, allowing precise measurement of oxygen and carbon dioxide content in blood equilibrated at controlled partial pressures — an experimental achievement that made the sigmoidal shape of the O₂-binding curve, and its pH sensitivity, rigorously measurable decades before any structural explanation existed. His son Niels Bohr later revolutionized atomic physics; the elder Bohr's own legacy is this founding observation in respiratory physiology, whose full molecular mechanism was not resolved until Perutz's crystallographic work more than sixty years later.

History

- 1904 — Christian Bohr, Karl Hasselbalch, and August Krogh report the effect of CO₂/pH on hemoglobin's oxygen affinity.

- 1914 — J. Barcroft and collaborators quantify the Haldane effect, the oxygenation-linked release of CO₂/protons, named for John Scott Haldane.

- 1959–1968 — Max Perutz and John Kendrew solve the crystal structures of myoglobin and hemoglobin at atomic resolution, providing the T/R structural framework.

- 1967 — Reinhold Benesch and Ruth Benesch demonstrate that 2,3-BPG (then called 2,3-DPG) is the physiological regulator lowering hemoglobin's O₂ affinity in red cells.

- 1970 — Perutz publishes the stereochemical mechanism of cooperativity and the Bohr effect, identifying His146β and the relevant salt bridges.

- 1974–1978 — Arthur Arnone and coworkers solve the crystal structure of deoxyhemoglobin with bound BPG, directly visualizing the central-cavity binding site.

What exactly is meant by "T state" and "R state" in this context?

T (tense) and R (relaxed) are the two quaternary conformations of the hemoglobin tetramer, differing by a roughly 15° rotation of the α1β1 dimer relative to the α2β2 dimer around the central cavity. T has a lower intrinsic affinity for O₂ at each heme and is stabilized by specific interchain salt bridges (including the His146β bridge central to the Bohr effect) and by a wide central cavity that admits 2,3-BPG. R has broken most of these salt bridges, a narrower central cavity, and higher intrinsic O₂ affinity. Effectors like H⁺, CO₂, and BPG all act by binding preferentially to T, shifting the pre-existing T⇌R equilibrium toward T even before any O₂ is bound, which is the operational meaning of "heterotropic allosteric effector" in this system.

Does the Bohr effect operate identically at very low and very high pH, or only near physiological pH?

The commonly quoted Bohr coefficient of about −0.6 applies specifically in the near-physiological range, roughly pH 6 to 8, where His146β's pKa transition (6 in R, 8 in T) dominates. Outside this range there are additional contributing groups — for example, at more alkaline pH (above 8–9) other histidines and the α-chain N-termini contribute a further, smaller "alkaline Bohr effect" component, while below pH 6 a separate "acid Bohr effect" of opposite sign appears, attributed to different titratable groups (including some carboxylates) whose protonation actually favors R rather than T. The −0.6 figure is thus a physiologically calibrated regional slope, not a universal constant valid at all pH.

Is the Bohr effect the same thing as the Haldane effect?

No — they are two faces of the same linked equilibrium, related by the thermodynamic cycle in the derivation above, but describing opposite directions of causality. The Bohr effect describes how changing [H⁺] (or CO₂) shifts O₂ affinity: lower pH, lower O₂ affinity. The Haldane effect describes the reverse direction: how changing O₂ saturation shifts hemoglobin's affinity for H⁺ and CO₂ — oxygenating hemoglobin releases protons (and promotes CO₂ release from carbamate and from bicarbonate via the resulting pH change). Both arise from the identical set of T-state-favoring interactions; which one you call it depends only on which variable you consider the "input" and which the "output" in a given experiment.

Why doesn't myoglobin show a Bohr effect even though it also binds O2 reversibly?

Myoglobin is monomeric and has no quaternary structure to switch between two conformational states. The Bohr effect and the BPG effect are fundamentally cooperative, multi-subunit phenomena: they work by an effector binding preferentially to one global conformation (T) of a multi-subunit assembly, shifting the population between T and R. A monomeric protein has, at most, small local pKa shifts upon ligand binding (a much weaker, non-cooperative version sometimes called a "trivial" or intrinsic pH effect), but it cannot exhibit the sharp, physiologically tuned, cooperative pH- and BPG-sensitivity that hemoglobin's T⇌R switch provides. This is a direct consequence of the general principle, covered under Cooperative Binding and Allostery, that heterotropic allosteric regulation strictly requires multiple interacting subunits or domains.

How does 2,3-BPG actually get made and regulated in the red blood cell?

2,3-BPG is produced from 1,3-bisphosphoglycerate, an intermediate of glycolysis, via a side-branch enzyme called bisphosphoglycerate mutase (BPGM), which is essentially unique to red blood cells (and a few other tissues at low levels) since mature erythrocytes lack mitochondria and rely entirely on glycolysis for ATP. This diversion means BPG synthesis directly competes with the main glycolytic flux toward pyruvate and ATP generation, so red cell BPG levels rise under conditions where glycolytic flux is high relative to ATP demand (such as chronic hypoxia or alkalosis, both of which stimulate BPGM activity and 1,3-BPG accumulation) and fall when phosphate or glycolytic substrate is limited, as in severe hypophosphatemia. BPG is degraded back to 3-phosphoglycerate by a phosphatase activity, completing what is often drawn as the "Rapoport–Luebering shunt."

Can the Bohr effect and 2,3-BPG effect be distinguished experimentally in the lab?

Yes, using "stripped" hemoglobin (dialyzed free of all endogenous organic phosphates and CO₂/bicarbonate) as a clean baseline. Adding controlled amounts of buffer at different pH values, with BPG rigorously excluded, isolates the pure Bohr effect and gives the ~−0.6 coefficient. Separately, holding pH fixed at a defined value with a non-permeant buffer and titrating in purified 2,3-BPG isolates the BPG effect and gives a saturable binding curve with a $K_d$ around 15–20 μM for deoxyhemoglobin under near-physiological ionic strength (differing from the illustrative 6 μM used in the worked example above, since actual reported values depend on ionic strength, temperature, and whether inositol hexaphosphate or BPG itself is used as the test ligand). Both experiments together, combined with crystallography of the relevant liganded and effector-bound structures, is how the two contributions were originally deconvolved.

Why is the alkaline Bohr effect coefficient negative, and could it ever be positive physiologically?

It is negative by the sign convention $\phi = \Delta\log P_{50}/\Delta\text{pH}$ : raising pH (making blood more alkaline, i.e., increasing pH which corresponds to decreasing [H⁺]) lowers $P_{50}$ (raises O₂ affinity), so the two changes have opposite sign, giving a negative slope. In principle a hemoglobin variant with reversed pKa ordering (higher pKa in R than in T at the relevant Bohr groups) would show a positive Bohr coefficient, but no naturally occurring human hemoglobin variant does this at the principal His146β site; some pathological variants do show an abolished or even inverted acid-range Bohr effect at extreme pH due to loss of the relevant salt bridge, which is diagnostically useful but distinct from a sign-reversed physiological alkaline Bohr effect.