Noncompetitive and Mixed Inhibition

Mixed and noncompetitive inhibition are not textbook curiosities — they describe how many clinically important drugs and toxins actually work. Allosteric HIV protease inhibitors and certain kinase inhibitors bind pockets distinct from the substrate cleft, producing mixed-inhibition kinetics that pharmacologists must fit correctly to compute true inhibition constants for drug potency comparisons (a competitive-only fit to mixed data systematically misestimates $K_i$). Heavy metal ions such as $\text{Hg}^{2+}$ and $\text{Pb}^{2+}$ frequently act as noncompetitive inhibitors by binding cysteine thiols at sites away from the active site, distorting global enzyme conformation rather than directly blocking substrate entry — exactly the $K_i\approx K_i'$ scenario where $V_{max}$ collapses but substrate affinity survives. Allosteric feedback inhibition in metabolic pathways — glycolytic enzymes inhibited by downstream ATP or citrate, amino acid biosynthetic enzymes inhibited by their end products — is almost universally mixed or noncompetitive rather than competitive, because feedback metabolites rarely resemble the enzyme's actual substrate structurally and instead bind dedicated regulatory sites, consistent with the allosteric-site logic developed in the Enzyme Inhibition Overview. In drug discovery pipelines, high-throughput kinetic screens routinely fit dose-response data to the general modifier equation to classify hit compounds by mechanism before committing resources to structural studies, and correctly identifying a mixed inhibitor's $K_i$ versus $K_i'$ can determine whether a compound will remain effective at the high substrate concentrations found in a cell versus only in a dilute test tube assay.

:::quiz[If a mixed inhibitor has K_i = 2 μM and K_i′ = 20 μM, what happens to K_m^app as [I] increases?]

• K_m^app decreases toward zero
• K_m^app stays exactly equal to K_m
• K_m^app increases, approaching a finite limit ✓
• K_m^app becomes negative

:::quiz[In the general modifier rate law, the ratio V_max^app / K_m^app at low [S] depends on:]

• only K_i′, the ES-binding constant
• both K_i and K_i′ equally
• only K_i, the free-enzyme binding constant ✓
• neither constant, only [S]

Ask the tutor: How would you extend this rapid-equilibrium model to a full steady-state treatment where inhibitor binding is not assumed fast compared to $k_{cat}$?

Picture the enzyme as a molecular machine with two conformational "modes" relevant to catalysis: the free form E, and the substrate-loaded form ES poised to react. A competitive inhibitor is a rival for the substrate's own binding site — it can only grab E, never ES, because once S is docked there is no room left for I. A noncompetitive or mixed inhibitor is different: it binds an entirely separate pocket, often an allosteric site, and this pocket may be present on E, on ES, or on both with different affinities. Because the inhibitor site is distinct from the substrate site, S and I do not physically exclude each other — both can be bound simultaneously, giving the dead-end ternary complex EIS. The catalytic consequence is that EIS is unproductive: even though substrate is bound, the inhibitor-bound enzyme cannot complete the chemical step (it may distort the active site, block a conformational change needed for catalysis, or sequester a required cofactor). This is fundamentally different from competitive inhibition, where raising $[S]$ can always outcompete I and restore full velocity. Here, if $K_i'$ is small, high $[S]$ actually drives more enzyme into the inhibited EIS complex rather than away from it, and $V_{max}$ can never be fully recovered. The full reaction network has four enzyme species — E, ES, EI, EIS — connected by four equilibria (two substrate-binding steps with constant $K_m$ in each direction of the square, two inhibitor-binding steps with constants $K_i$ and $K_i'$). Only ES reacts, releasing product with rate constant $k_{cat}$. The elegance of the model is that this four-species square, worked through steady-state or rapid-equilibrium algebra, collapses into a Michaelis-Menten-shaped equation with an effective $V_{max}^{app}$ and $K_m^{app}$, both simple functions of $[I]$.

The governing rate law for a general modifier (mixed) inhibitor, derived under rapid-equilibrium binding of both S and I, is:

where:

• $V_{max} = k_{cat}[E]_T$ is the maximal velocity of the uninhibited enzyme,
• $K_m$ is the uninhibited Michaelis constant,
• $[I]$ is the free inhibitor concentration,
• $K_i$ is the dissociation constant for I binding free E ($\mathrm{E+I \rightleftharpoons EI}$),
• $K_i'$ is the dissociation constant for I binding ES ($\mathrm{ES+I \rightleftharpoons EIS}$).

Rearranged into standard Michaelis-Menten form by dividing through by $(1+[I]/K_i')$, this becomes:

where:

• $V_{max}^{app}$ is the apparent maximal velocity measured at a fixed $[I]$,
• $K_m^{app}$ is the apparent Michaelis constant at that same $[I]$,
• the two correction factors $(1+[I]/K_i)$ and $(1+[I]/K_i')$ act on $K_m$ and $V_{max}$ respectively, but $K_m^{app}$ carries both because it is a ratio of rate constants.

Reading: $V_{max}$ is suppressed exactly by the fraction of ES diverted into EIS — governed purely by $K_i'$. Meanwhile $K_m^{app}$, being an affinity-like ratio, is reshaped by both equilibria: if the inhibitor prefers E over ES ($K_i < K_i'$), it effectively raises $K_m^{app}$ (looks like weaker substrate affinity); if it prefers ES over E ($K_i' < K_i$), it can lower $K_m^{app}$ — the paradoxical case where an inhibitor makes the enzyme appear to bind substrate more tightly, precisely because it is trapping ES in the dead-end complex and pulling the equilibrium of substrate binding forward (Le Chatelier acting on the binding step, not the catalytic step). The special case $K_i = K_i'$ (pure noncompetitive inhibition) makes the two factors identical:

so $K_m$ is completely unchanged while $V_{max}^{app} = V_{max}/(1+[I]/K_i)$ falls uniformly — the textbook "pure noncompetitive" signature. The catalytic efficiency ratio collapses beautifully regardless of mechanism type:

where this shows $k_{cat}/K_m$ (the specificity constant, relevant at low $[S]$) depends only on $K_i$, the inhibitor's affinity for free E — never on $K_i'$ at all.

Example 1 — Foundational: computing $V_{max}^{app}$ and $K_m^{app}$ for a mixed inhibitor. An enzyme has $V_{max} = 120\ \mu\text{mol min}^{-1}\text{mg}^{-1}$ and $K_m = 4.0\ \text{mM}$. A mixed inhibitor is present at $[I] = 6.0\ \mu\text{M}$, with $K_i = 3.0\ \mu\text{M}$ (affinity for free E) and $K_i' = 9.0\ \mu\text{M}$ (affinity for ES, weaker binding, larger $K_i'$). Compute the two correction factors:

Apparent maximal velocity:

Apparent Michaelis constant:

Both $V_{max}$ and $K_m$ have moved — $V_{max}$ dropped 40%, $K_m$ rose 80% — confirming genuine mixed behavior (since $K_i\neq K_i'$, neither pure competitive $K_m$-only nor pure noncompetitive $V_{max}$-only pattern applies). As a limiting-case check, if instead $K_i' \to \infty$ (inhibitor cannot bind ES at all), the factor $1+[I]/K_i' \to 1$, giving $V_{max}^{app}=V_{max}=120$ unchanged and $K_m^{app}=K_m\times3.00=12.0$ mM — exactly the competitive-inhibition result, confirming the general formula degenerates correctly.

Example 2 — Applied: a feedback inhibitor and metabolic flux control. Consider aspartate transcarbamoylase-like feedback regulation (illustrative numbers): the target enzyme has $V_{max}=50\ \mu\text{M s}^{-1}$, $K_m = 2.0\ \text{mM}$ substrate. A pathway end-product I acts as an allosteric inhibitor with $K_i = 15\ \mu\text{M}$ for the enzyme's low-activity (T-like) conformation representing free E, and $K_i' = 45\ \mu\text{M}$ for the ES complex (the inhibitor binds the substrate-loaded form three-fold more weakly). At physiological feedback concentration $[I] = 30\ \mu\text{M}$:

At the physiological substrate concentration $[S]=1.0$ mM (below $K_m^{app}$), velocity is

compared to the uninhibited velocity at the same $[S]$:

a roughly 61% flux reduction — showing how a mixed inhibitor operating physiologically below saturating substrate can throttle pathway flux sharply even though $K_i, K_i'$ are both micromolar and $[S]$ is millimolar. This is exactly the strategy pathways use for feedback control: the inhibitor need not compete head-on with a much higher-affinity, higher-concentration substrate if it instead exploits a separate allosteric pocket.

Example 3 — Exam/competition level: back-calculating $K_i$ and $K_i'$ from two double-reciprocal lines. An enzyme assay gives, in the absence of inhibitor, $V_{max}=200\ \text{nmol min}^{-1}$ and $K_m = 5.0\ \text{mM}$. Adding inhibitor at $[I]=10\ \mu\text{M}$ gives a Lineweaver-Burk line with slope $0.075\ \text{min}\cdot\text{mM}/\text{nmol}$ and y-intercept $0.0125\ \text{min/nmol}$. Find $K_i$ and $K_i'$. The Lineweaver-Burk form is

From the y-intercept:

From the slope:

So $K_i = 5.00\ \mu\text{M}$ (tighter binding to free E) and $K_i'=6.67\ \mu\text{M}$ (weaker binding to ES) — genuine mixed inhibition, with $K_i < K_i'$, meaning the inhibitor slightly prefers the free enzyme. Predict where the Lineweaver-Burk lines cross:

Since $K_i<K_i'$ here, the crossing point should lie below the x-axis (this can be confirmed by plugging $1/[S]=-0.150$ into either line's equation and checking $1/v$ is negative — an algebraic marker, not a physically attainable point, but a valid diagnostic feature of the fitted lines). As a limiting check, if the fitted numbers had instead given $K_i=K_i'$ exactly, the slope-derived and intercept-derived factors would coincide, correctly recovering pure noncompetitive inhibition with the crossing point pinned to the x-axis.

The modern general-modifier framework for enzyme inhibition traces its clearest articulation to William Wallace Cleland, whose systematic nomenclature and rapid-equilibrium/steady-state treatments of multi-substrate and multi-inhibitor enzyme kinetics, developed through the 1960s, became the standard language of the field. Cleland's 1963 papers in Biochimica et Biophysica Acta laid out systematic rules for writing rate equations for arbitrarily complex binding schemes — precisely the machinery used above to derive the four-species square — and introduced the now-universal "Cleland nomenclature" for enzyme mechanisms (ordered, random, ping-pong) alongside rigorous inhibition-pattern classification (competitive, noncompetitive, uncompetitive, and mixed) based on which kinetic constants are affected. Earlier foundations came from Hans Lineweaver and Dean Burk's 1934 linearization, and from J.B.S. Haldane and others who recognized in the 1930s–1950s that inhibitors binding ES complexes produce qualitatively different plots than those binding only free enzyme, but it was Cleland who gave the field the systematic algebraic method — solving conservation equations across an arbitrary binding graph — that makes the general modifier model a routine derivation rather than a special trick for each new inhibitor.