Cooperative binding: hemoglobin and the Hill equation

Michaelis-Menten kinetics assumes each enzyme active site works independently — the binding of a substrate to one enzyme molecule has no effect on any other. Half the interesting biology violates this assumption. Hemoglobin binds four oxygen molecules, but each successive O₂ binds more easily than the previous one; aspartate transcarbamoylase’s catalytic activity switches on across a narrow window of substrate concentration; allosteric enzymes at metabolic branch points respond to their substrate as a sigmoidal switch, not a hyperbolic drift. Cooperativity is the mechanism. The Hill equation is the arithmetic. Together they explain why hemoglobin can pick up oxygen in the lung and release it in tissue with the same 20-mmHg change in P_O₂ that would move a non-cooperative protein less than 15% of the way through its response curve.

Assume familiarity with Michaelis-Menten from why enzymes plateau and pH buffering from physiological buffers.

What cooperative binding looks like

Plot fractional saturation Y (fraction of binding sites occupied) against ligand concentration [L] for a cooperative protein and you get an S-shaped curve, not a hyperbola:

• At low [L], Y rises slowly — the first ligand binds with difficulty.
• Over a narrow concentration window, Y rises steeply — subsequent ligands bind with rapidly increasing ease.
• At high [L], Y plateaus at 1 — full saturation.

Hemoglobin’s oxygen-binding curve (Lehninger 7e Fig 5-12) shifts from 4% saturation at P_O₂ = 5 mmHg (a low-oxygen tissue) to 98% saturation at P_O₂ = 100 mmHg (the alveoli of the lung) — an 86 percentage-point change over a 20× concentration window. Myoglobin, which is not cooperative, is 74% saturated at P_O₂ = 5 mmHg and 99% at 100 mmHg — it hangs onto oxygen so tightly that it never releases to tissue. Hemoglobin’s cooperativity is precisely what makes it a good oxygen carrier rather than a good oxygen sponge.

The Hill equation

Archibald Hill’s 1910 empirical fit to the hemoglobin curve is:

$Y = \frac{[\text{L}]^n}{K_d^n + [\text{L}]^n}$

with:

• n = Hill coefficient (dimensionless).
• K_d = apparent binding constant, roughly the [L] where Y = 0.5.
• [L] = ligand concentration.

At n = 1, Hill collapses to the standard hyperbolic binding curve (equivalent to Michaelis-Menten shape with K_d ≡ K_m). At n > 1, the curve is positively cooperative — steeper than hyperbolic. Larger n → sharper switch. At n < 1, the protein is negatively cooperative, less common but real.

The Hill coefficient n is a lower bound on the number of binding sites. Hemoglobin has four heme groups, so n_max = 4; the experimental n_Hill for hemoglobin is 2.8-3.0, meaning cooperativity is strong but not perfect (a hypothetically perfect all-or-none tetramer would give exactly n = 4). Berg 9e §7.2 gives the range of measured n for common cooperative proteins:

• Hemoglobin (O₂): n = 2.8-3.0.
• Aspartate transcarbamoylase (substrate): n ≈ 3.
• Phosphofructokinase (ATP inhibition): n ≈ 3.8.
• Calcium-binding proteins (calmodulin): n = 1.5-2 for Ca²⁺.

Deriving n from data: the Hill plot

Rearrange Hill:

$\log\left(\frac{Y}{1-Y}\right) = n \log[\text{L}] - n \log K_d$

A plot of log(Y/(1−Y)) vs log[L] — the Hill plot — is linear with slope n. In practice the plot is only truly linear in the middle of the transition (0.1 < Y < 0.9); at the tails it curves toward the individual K_d values for the first and last binding events. The slope at Y = 0.5 gives the maximum Hill coefficient, which is what “n_Hill” conventionally reports.

The Hill plot is the diagnostic test for cooperativity. If n > 1 at Y = 0.5, cooperativity is positive; n < 1, negative; n = 1, non-cooperative. Voet 5e §10.1 has canonical worked examples using hemoglobin data.

The MWC (Monod-Wyman-Changeux) mechanism

Hill is descriptive but not mechanistic. The mechanism that explains why hemoglobin cooperates is the Monod-Wyman-Changeux (MWC) model (1965): the tetramer exists in equilibrium between two conformational states — T (tense) and R (relaxed) — with all four subunits switching in concert. T has low O₂ affinity; R has high affinity. The equilibrium [R]/[T] shifts toward R as more oxygens bind.

Perutz’s 1970 X-ray crystallography confirmed the atomic-scale reality: T and R are two distinct quaternary structures of the tetramer that differ in the alignment of subunit-subunit contacts. Oxygen binding to one heme triggers a conformational change transmitted through those contacts, pulling nearby subunits into R and raising their O₂ affinity. The mechanism is concerted — the whole tetramer flips at once, not one subunit at a time.

The alternative model (Koshland-Némethy-Filmer, KNF, or “sequential” model) allows individual subunits to switch independently. Real hemoglobin sits between the two — the MWC picture is a good first approximation, with small deviations that KNF captures better.

Regulators of cooperativity: allosteric effectors

Cooperativity is not fixed — it responds to other molecules. Hemoglobin’s O₂ affinity is regulated by four heterotropic effectors:

• CO₂ / H⁺ (Bohr effect): high CO₂ and low pH (as in exercising muscle) stabilize T. Hemoglobin’s O₂ affinity drops; oxygen is released more readily where it is most needed. The Bohr effect is why exercising tissue gets O₂ efficiently despite ambient P_O₂ being similar to resting tissue.
• 2,3-BPG (bisphosphoglycerate): binds to the central cavity of the T state and stabilizes it. High-altitude adaptation involves upregulating [2,3-BPG], shifting hemoglobin’s O₂ affinity lower so more O₂ is released to tissue at reduced arterial P_O₂.
• Temperature: higher T stabilizes R; hyperthermia decreases O₂ affinity, so oxygen is easier to release into hot exercising tissue.
• Fetal hemoglobin (HbF): contains γ-chains instead of β-chains, binds 2,3-BPG less tightly, and thus has higher O₂ affinity than adult HbA. That difference is what allows the fetus to extract O₂ from the mother’s blood at the placenta.

Each of these effectors changes the position of the sigmoidal curve on the P_O₂ axis without necessarily changing its shape or n_Hill. The clinical shorthand is “P₅₀” — the P_O₂ at which Y = 0.5. Normal adult Hb has P₅₀ = 26 mmHg. HbF has P₅₀ = 19 mmHg. High-altitude adaptation raises P₅₀ to ~30 mmHg through 2,3-BPG upregulation.

Worked example — hemoglobin O2 saturation at rest vs exercise

When to reach for Hill vs Michaelis-Menten

Practical decision tree:

• Plotting rate or binding vs [L]: does the curve look hyperbolic or sigmoidal? If sigmoidal, use Hill.
• Number of binding sites known: if 1, use Michaelis-Menten. If ≥ 2, consider Hill (the protein might still be non-cooperative).
• Hill plot slope at Y = 0.5: n = 1 means non-cooperative (use MM); n > 1 means positively cooperative (use Hill); n < 1 means negatively cooperative (use Hill, but interpret cautiously).

Hill treats cooperativity phenomenologically; MWC and KNF give the mechanism. For teaching or curve-fitting, Hill is enough. For enzyme kinetics near a regulatory setpoint or for structure-function studies, upgrade to MWC.

Practice problems

Where to go next

• The Michaelis-Menten calculator is not directly usable for cooperative enzymes but is a good comparison — plot v/Vmax for the same K and see how much more gradual the MM curve is than the Hill curve at n = 3.
• The buffer calculator models the pH environment that drives the Bohr effect; simulating pH shifts under exercise loads gives an intuition for how physiological pH couples to O₂ delivery.
• Beyond hemoglobin, cooperative binding underlies gene regulation (transcription factors on multiple binding sites), calcium signaling (calmodulin’s four EF hands), and drug-receptor pharmacology (agonists with n > 1). The Hill equation shows up throughout biology.

Cooperative binding is what lets a protein act like a switch instead of a dimmer. Michaelis-Menten describes proteins whose response is monotonic and gradual; Hill describes proteins whose response is sigmoidal and sharp. Every regulatory decision point in biology — muscle contraction, gene expression, hemoglobin loading and unloading, allosteric enzyme control — reaches for cooperativity to convert a gradual chemical signal into a decisive functional output.