Why enzymes plateau: reading v vs [S] curves

Plot the initial rate of an enzyme-catalysed reaction against substrate concentration and you get a hyperbola that starts linear at low [S], curves over, and asymptotes to a maximum. Every enzyme in every biochemistry textbook makes this shape. This guide unpacks why the curve looks like this, what its two parameters (Vmax and Km) actually mean, and what kcat/Km tells you about how good an enzyme really is at its job.

Why enzymes saturate

The distinguishing feature of an enzyme, compared with an inorganic catalyst, is that it has a specific binding pocket for its substrate. Reaction proceeds through a two-step mechanism:

$\mathrm{E + S} \xrightleftharpoons[k_{-1}]{k_1} \mathrm{ES} \xrightarrow{k_{cat}} \mathrm{E + P}$

The enzyme binds substrate to form an enzyme-substrate complex ES, then the ES complex converts to product and releases free enzyme. When [S] is small, most of the enzyme is unoccupied, and adding more substrate produces proportionally more ES and hence more product — the reaction looks first-order in substrate. When [S] is large, essentially every enzyme molecule is already bound (all E is ES), and adding more substrate cannot make the enzyme go any faster because there is no more free enzyme to bind. That is saturation, and it produces the hyperbolic curve.

A simple homogeneous catalyst (a proton in solution, say) does not have a binding pocket, so there is no equivalent of ES and no saturation. The reaction stays first-order in substrate no matter how high [S] gets. This is one of the two features that separates biological catalysis from ordinary catalysis; the other is specificity, which we will come back to.

The Michaelis-Menten equation

The rate law for the mechanism above, derived under Briggs and Haldane’s 1925 steady-state assumption (d[ES]/dt = 0 when the reaction has been running long enough for ES to reach its quasi-equilibrium concentration), is

$v = \frac{V_{max}\,[\text{S}]}{K_m + [\text{S}]}$

with two parameters:

• Vmax = kcat · [E]_total — the rate when every enzyme molecule is bound as ES.
• Km = (k₋₁ + kcat) / k₁ — a composite constant with units of concentration.

The curve v([S]) has three regions worth naming:

• Low-[S] limit ([S] ≪ Km): v ≈ (Vmax/Km) · [S]. Linear in [S]; slope is Vmax/Km.
• [S] = Km exactly: v = Vmax/2. This is the operational definition of Km — the substrate concentration that gives half the maximum rate.
• High-[S] limit ([S] ≫ Km): v ≈ Vmax. Saturated; adding more substrate is useless.

The transition between the two limits happens over roughly one order of magnitude in [S] around Km. To reach 90% of Vmax you need [S] = 9·Km; for 99%, [S] = 99·Km. This is why some experiments never reach “true” saturation — if your enzyme has Km = 10 mM and your substrate is only soluble to 20 mM, you top out at 67% Vmax.

What Km actually means

Km is not a binding constant, though it is often confused for one. The rapid-equilibrium limit of Michaelis and Menten’s original 1913 derivation gives Km = Kd = k₋₁/k₁ (the ES dissociation constant), but this limit requires kcat ≪ k₋₁ — the catalytic step is much slower than substrate dissociation. In the more general Briggs-Haldane treatment, Km = (k₋₁ + kcat)/k₁ ≥ Kd, and the inequality can be substantial: fast enzymes have Km well above their true binding constant.

What Km reliably does tell you is how much substrate the enzyme “sees” under physiological conditions. If [S]_cell is well below Km, the enzyme is operating in its linear regime and is highly responsive to changes in [S]; if [S]_cell is well above Km, the enzyme is saturated and insensitive to further increases. Regulatory enzymes tend to have Km values close to physiological [S] so they respond to small changes; housekeeping enzymes often have very low Km and are effectively always saturated.

What kcat and kcat/Km mean

kcat (units s⁻¹) is the turnover number — how many substrate molecules each active site processes per second at saturation. It can be computed from Vmax and the total enzyme concentration:

$k_{cat} = V_{max} / [\text{E}]_{total}$

Reasonable values range from 1 s⁻¹ (many regulatory enzymes; Lehninger 7e Table 6-6) to 10⁷ s⁻¹ (catalase decomposing hydrogen peroxide). Higher kcat means the enzyme cycles faster when substrate is abundant.

kcat/Km (units M⁻¹·s⁻¹) is the specificity constant — the effective second-order rate constant for the enzyme working on free substrate under sub-saturating conditions. In the limit [S] ≪ Km, v ≈ (kcat/Km) · [E] · [S], which has the same shape as an ordinary bimolecular rate law. It is called the specificity constant because when two substrates compete for the same enzyme, the ratio of their processing rates equals the ratio of their kcat/Km values (Cornish-Bowden 2012 §4). If an enzyme has kcat/Km = 10⁶ for its natural substrate and 10³ for a near-relative, it discriminates by a factor of 1000 under physiological conditions.

Diffusion sets an upper bound on kcat/Km: the substrate has to reach the active site before anything can happen, and simple diffusion-controlled encounters in water top out at ~10⁸–10⁹ M⁻¹·s⁻¹ for a small molecule and a protein. Enzymes at that limit are called catalytically perfect (Lehninger 7e Box 6-1):

These enzymes are so efficient that speeding up the chemistry any further would give no benefit — every substrate that reaches the active site is already being converted. Evolution has selected for the maximum useful catalytic efficiency and stopped there.

Reading a v vs [S] plot

Practical enzymology means fitting real experimental data to the Michaelis-Menten form. The right way, in 2026, is to do a nonlinear least-squares fit of v = Vmax·[S] / (Km + [S]) directly. Every modern kinetics software (GraphPad, Origin, R’s nlme, Python’s scipy.optimize) does this in one call, and the standard errors on Vmax and Km come out with correct weighting.

The linear transformations you may have seen (Lineweaver-Burk plotting 1/v vs 1/[S]; Eadie-Hofstee v vs v/[S]; Hanes-Woolf [S]/v vs [S]) were invented before computers and are now pedagogically useful but numerically unreliable. Lineweaver-Burk is the worst offender: it weights the noisy low-[S] points most heavily, and small experimental errors at low [S] cause large errors in the fitted parameters. Use these transformations to inspect your data (a curved Lineweaver-Burk plot is a sign that MM does not fit), but not to fit it. Cornish-Bowden 2012 §4 has the definitive treatment.

Worked example — hexokinase

Hexokinase (the enzyme that phosphorylates glucose in the first committed step of glycolysis) has Vmax = 6.5 µmol·min⁻¹ per mg enzyme and Km = 1.5 × 10⁻⁴ M for glucose under standard assay conditions (Lehninger 7e §6.3). Physiological [glucose] in blood is about 5 mM.

Where the Michaelis-Menten model fails

MM is a beautifully simple model. Real enzymes push against its assumptions in several ways.

• Cooperative binding. Allosteric enzymes bind multiple substrate molecules at multiple sites and communicate between them. The v vs [S] curve becomes sigmoidal, not hyperbolic, and is described by the Hill equation instead. Phosphofructokinase is a canonical example.
• Multi-substrate mechanisms. For an enzyme with two substrates (E + A + B → E + P + Q), Km values for each substrate depend on the concentration of the other — you have to specify “Km(A) at fixed [B]” or use a full ternary-complex analysis (Berg 9e §8.4).
• Substrate depletion. MM applies to initial rates only. If your experiment lets [S] fall by more than 5–10% during the measurement, you are integrating over changing [S] and the fitted Km will be biased. Use short assay windows or continuous product monitoring.
• Tight-binding regime. MM assumes [S] ≫ [E], so [S]_free ≈ [S]_total. For tight-binding inhibitors and enzymes at high [E], you need the Morrison equation instead (Cornish-Bowden 2012 §5).
• Substrate inhibition. At very high [S], some enzymes bind a second substrate molecule non-productively, and the v vs [S] curve drops past a maximum. The model becomes v = Vmax·[S] / (Km + [S] + [S]²/Ki), a two-parameter extension.

Practice problems

Where to go next

MM is the entry point to biological catalysis; once you have it, the natural extensions are inhibition (competitive, uncompetitive, non-competitive, mixed — each with its own effect on Vmax and Km), allostery (Hill equation, MWC and KNF models), and multi-substrate mechanisms (ping-pong, ordered, random ternary complex). Berg 9e §8.4 and Cornish-Bowden 2012 Ch. 5–6 have thorough treatments.

For hands-on practice, the Michaelis-Menten calculator solves any of v, Vmax, Km, or [S] given the other three, and optionally computes kcat and kcat/Km when you supply the enzyme concentration. Reproducing the hexokinase example above in the tool is a good sanity check that you have the substitution right.