Turnover Number k_cat and the Specificity Constant

$V_{max}$ tells you how fast a particular tube of enzyme goes. It is a property of that tube, and it doubles if you add twice as much enzyme, which makes it useless for comparing anything. Divide it by the concentration of enzyme actually present and you get $k_{cat}$, the turnover number: substrate molecules processed per enzyme molecule per second. That number belongs to the ENZYME. Dilute the preparation tenfold and $V_{max}$ falls tenfold while $k_{cat}$ does not move at all.

The range is wide and instructive. Lysozyme manages about 0.5 turnovers per second. A typical well-characterised enzyme sits somewhere between 1 and 10⁴ s⁻¹. Carbonic anhydrase runs near 10⁶ s⁻¹ and catalase near 4 × 10⁷ — a single catalase molecule destroying forty million hydrogen peroxide molecules a second, which is a fact worth pausing over.

The specificity constant $k_{cat}/K_m$ is the companion number and arguably the more important one. It has units of M⁻¹s⁻¹, and it is the apparent second-order rate constant for free enzyme meeting free substrate. Because it governs behaviour at LOW substrate — where enzymes mostly work, both inside a cell and inside a dilute process stream — it is the right figure of merit far more often than $k_{cat}$ alone. It is also the correct measure of preference between two competing substrates: given a mixture, an enzyme processes them in the ratio of their $k_{cat}/K_m$ values, and neither constant alone predicts that. An enzyme can have a higher $k_{cat}$ on one substrate and still prefer the other, because it binds it far more tightly.

There is a ceiling on the specificity constant, and it is set by physics rather than by biology. An enzyme cannot turn substrate over faster than substrate diffuses into the active site, which puts an upper bound somewhere around 10⁸ to 10⁹ M⁻¹s⁻¹. A handful of enzymes sit at that bound — triosephosphate isomerase, acetylcholinesterase, carbonic anhydrase, superoxide dismutase — and are described as catalytically perfect: essentially every productive encounter between enzyme and substrate results in catalysis, and no further improvement is physically available. Triosephosphate isomerase is the celebrated case, and the phrase belongs to Albery and Knowles, who argued in 1976 that its evolution had run out of room to improve.

Two cautions on the enzyme concentration, because they are where the number usually goes wrong. $[E]_t$ must be the concentration of ACTIVE SITES. It is not the protein concentration unless the preparation is pure and fully active, and it rarely is — an enzyme preparation that is 60% active gives a $k_{cat}$ 40% too low, which is the commonest reason a measured turnover number falls short of a published one. And a multimeric enzyme carries several active sites per molecule, so the site concentration is the subunit count times the molecular concentration; getting that wrong puts the answer out by a factor of two or four.

Run the relation backwards and it becomes a useful check. Measure $V_{max}$, take $k_{cat}$ from the literature, and the division tells you how much functional enzyme is really present. Compare that with what you weighed out and the difference is the inactive fraction — misfolded, denatured, partly proteolysed, or simply the other proteins in a preparation less pure than its label. The rigorous version of the same measurement is an active-site titration with a stoichiometric reagent that reacts once per site and is counted directly, which needs no assumed $k_{cat}$ at all.

One last note for anyone moving from the bench to a reactor. An industrial enzyme is rarely running at its laboratory $k_{cat}$. It is often immobilised on a support, which imposes diffusion limits that can dominate the observed rate entirely; it is working in a real process stream rather than a clean buffer; and it is losing activity steadily as the run proceeds. Sizing an enzyme charge on a textbook $k_{cat}$ alone is how a reactor ends up half the size it needed to be.