Enzyme Inhibition

What inhibition actually does to an enzyme

An enzyme is a catalyst: it binds a substrate (S) in its active site, helps it cross the reaction barrier, and releases product (P). An inhibitor (I) is any molecule that makes that cycle run slower. The interesting part is where and when the inhibitor binds, because that single choice determines exactly how the rate equation bends.

The baseline behavior of an uninhibited enzyme is the Michaelis–Menten scheme. The enzyme grabs substrate to form the ES complex, then either lets it go again or turns it into product:

          k1        kcat
   E + S ⇌  ES  ───────→  E + P
          k-1

   v = Vmax · [S] / (Km + [S])
   Vmax = kcat · [E]_total        Km = (k-1 + kcat) / k1

Two numbers summarize everything. Vmax is the top speed when every active site is full of substrate. Km is the substrate concentration that gives exactly half of Vmax — a practical stand-in for how tightly the enzyme grabs substrate (low Km = high apparent affinity). An inhibitor works by changing one or both of these. The three classes are defined entirely by which of Vmax and Km move, and in which direction.

Competitive inhibition: a fight for the active site

A competitive inhibitor looks enough like the substrate to slot into the same active site. While it sits there, no substrate can bind, so that enzyme molecule is temporarily idle. The enzyme is now partitioned three ways:

   E + S ⇌ ES → E + P
   E + I ⇌ EI        (dead-end, no product)
   I binds free E only — never ES

The key word is competition: substrate and inhibitor are fighting for one location. Flood the enzyme with substrate and you statistically swamp the inhibitor — every active site is so often occupied by substrate that the inhibitor rarely gets in. So Vmax is unchanged (you can still reach full speed), but it takes more substrate to get there, which means Km goes up (the enzyme appears to bind substrate more weakly). Quantitatively, the apparent Km scales by the factor α:

   Km(app) = Km · α       where α = 1 + [I]/Ki
   Vmax(app) = Vmax        (unchanged)

   v = Vmax · [S] / (α·Km + [S])

Worked number: methotrexate inhibits dihydrofolate reductase (DHFR) with a Ki near 1 picomolar — among the tightest competitive inhibitors known. At a free drug concentration of just 10 pM, α = 1 + 10/1 = 11, so the enzyme's apparent Km for dihydrofolate jumps eleven-fold while its Vmax sits unchanged. That is why high-dose methotrexate rescue with folinic acid works: pour in enough downstream substrate-analog and you bypass the competition.

Non-competitive inhibition: poisoning from a second site

A non-competitive inhibitor binds somewhere other than the active site — an allosteric pocket — and bends the protein so catalysis falters. Crucially it binds free E and the ES complex equally well, so substrate occupancy is irrelevant to whether the inhibitor lands:

   E + I  ⇌ EI       ES + I ⇌ ESI       (same Ki for both)
   EI and ESI are catalytically dead

Because the inhibitor doesn't care whether the active site is occupied, adding substrate cannot displace it. A fixed fraction of the enzyme population is always knocked out, so the ceiling speed falls: Vmax goes down. But the enzyme molecules that are free still grab substrate exactly as before, so the apparent Km is unchanged:

   Vmax(app) = Vmax / α       α = 1 + [I]/Ki
   Km(app) = Km               (unchanged)

   v = (Vmax/α) · [S] / (Km + [S])

Pure non-competitive inhibition (identical Ki for E and ES) is the textbook ideal; in practice most "second-site" inhibitors bind E and ES with different strengths, giving the more general mixed inhibition, which moves both Vmax and Km. True non-competitive behavior is the special, symmetric case of mixed inhibition where the two Ki values are equal.

Uncompetitive inhibition: binding only after the substrate

The strangest of the three. An uncompetitive inhibitor binds only the ES complex — never the free enzyme. The substrate has to arrive first and reshape the protein before the inhibitor's site even exists:

   E + S ⇌ ES → E + P
   ES + I ⇌ ESI      (dead-end; I cannot bind free E)

This flips the competitive logic on its head. Adding more substrate makes more ES, which gives the inhibitor more to bind — so high substrate makes uncompetitive inhibition worse, not better. Removing ES by Le Chatelier also pulls the E + S ⇌ ES equilibrium to the right, which is why Km drops. Both Vmax and Km fall by the same factor α′:

   Vmax(app) = Vmax / α'      Km(app) = Km / α'
   α' = 1 + [I]/Ki'   ... reported as constant scaling:

   v = (Vmax/α') · [S] / ((Km/α') + [S])

Because Vmax and Km shrink by the same factor, their ratio Vmax/Km — the second-order rate constant at low substrate — is the one quantity left untouched (the α′ cancels). That is the opposite of the competitive case, where Vmax stays put but Km rises, so Vmax/Km falls. Lithium's inhibition of inositol monophosphatase, the leading molecular hypothesis for how lithium treats bipolar disorder, is a classic uncompetitive example.

How to diagnose the class: the Lineweaver–Burk fingerprint

Take the reciprocal of the Michaelis–Menten equation and it becomes a straight line in 1/v versus 1/[S]:

   1/v = (Km/Vmax)·(1/[S]) + 1/Vmax
        └ slope ┘            └ y-intercept ┘
   x-intercept = −1/Km

Run the assay at several substrate concentrations with and without inhibitor, plot both lines, and the geometry tells you the class at a glance:

	Competitive	Non-competitive (pure)	Uncompetitive
Inhibitor binds	Free E (active site)	E and ES (allosteric)	ES only
Apparent Km	↑ increases (×α)	unchanged	↓ decreases (÷α′)
Apparent Vmax	unchanged	↓ decreases (÷α)	↓ decreases (÷α′)
Vmax/Km	↓ decreases	↓ decreases	unchanged
Beat it with more [S]?	Yes — fully reversible	No	No — gets worse
Lineweaver–Burk lines	Cross on the y-axis	Cross on the x-axis	Parallel (same slope)
Drug example	Statins, methotrexate	Some metal-chelating poisons	Lithium on IMPase

The mnemonic that survives every exam: same y-intercept = competitive (Vmax shared), same x-intercept = non-competitive (Km shared), parallel lines = uncompetitive (slope Km/Vmax shared because both scale together).

Worked example: how much does a competitive inhibitor slow you?

An enzyme has Vmax = 100 µM/min and Km = 2.0 mM. A competitive inhibitor with Ki = 0.5 mM is present at [I] = 1.5 mM, and the assay runs at [S] = 2.0 mM (i.e. at the uninhibited Km). What is the velocity?

α = 1 + [I]/Ki = 1 + 1.5/0.5 = 4.0
Km(app) = α·Km = 4.0 × 2.0 mM = 8.0 mM

v = Vmax·[S] / (α·Km + [S])
  = 100 × 2.0 / (8.0 + 2.0)
  = 200 / 10
  = 20 µM/min

Without inhibitor, [S] = Km gives exactly half of Vmax = 50 µM/min. The inhibitor has cut the rate to 20 µM/min — a 2.5-fold slowdown — purely by quadrupling the apparent Km. Now raise substrate to a flooding [S] = 100 mM: v = 100 × 100 / (8 + 100) = 92.6 µM/min, almost back to Vmax. That recovery at high substrate is the experimental signature of competition.

Reversible vs irreversible: when the enzyme never comes back

Everything above is reversible inhibition — the inhibitor binds and releases by non-covalent forces governed by an equilibrium Ki, so dialysis or dilution restores activity. Irreversible inhibitors instead form a covalent bond and the enzyme stays dead until the cell synthesizes a fresh copy. The chemistry is real arrow-pushing:

• Aspirin → cyclooxygenase (COX). The acetyl group of aspirin is transferred to Ser530 of COX-1, blocking the channel that arachidonic acid uses. Platelets can't remake COX (no nucleus), so a single low-dose aspirin suppresses platelet thromboxane for the platelet's entire ~8–10 day lifespan.
• Penicillin → transpeptidase. The strained β-lactam ring acylates the active-site serine of the bacterial cell-wall transpeptidase (a penicillin-binding protein), permanently capping it. The bacterium can't cross-link its peptidoglycan and bursts under osmotic pressure.
• Organophosphates → acetylcholinesterase. Nerve agents and many insecticides phosphorylate the catalytic serine; the adduct "ages" into a bond that even the antidote pralidoxime can't break. This is why sarin is lethal at sub-milligram doses.

Irreversible inhibitors don't have a true Ki. They are described by a second-order inactivation rate kinact/KI — how fast the covalent kill happens — which is the modern design metric for "targeted covalent inhibitors" like the kinase drug ibrutinib (covalently captures Cys481 of Bruton's tyrosine kinase).

Where enzyme inhibition runs the world of medicine

• Statins (competitive). Atorvastatin binds HMG-CoA reductase with a Ki in the low nanomolar range, out-competing the natural HMG-CoA substrate to throttle cholesterol synthesis. Because it's competitive, the body responds by upregulating the enzyme — which is why statins are dosed continuously.
• ACE inhibitors (competitive). Drugs like lisinopril mimic the C-terminal dipeptide of angiotensin I and occupy the active-site zinc of angiotensin-converting enzyme, lowering blood pressure.
• Allopurinol (competitive → suicide). It competes with hypoxanthine at xanthine oxidase, but the enzyme then converts it to oxypurinol, which binds the reduced molybdenum cofactor near-irreversibly — a "suicide" inhibition that drops uric acid in gout.
• Protease inhibitors (competitive/transition-state mimics). HIV protease inhibitors such as saquinavir are designed to resemble the tetrahedral transition state of peptide-bond hydrolysis, binding far tighter than substrate (Ki in the sub-nanomolar range).
• Negative feedback in metabolism (non-competitive/allosteric). The end product of a pathway often binds an allosteric site on the first committed enzyme — ATP and CTP inhibiting phosphofructokinase and aspartate transcarbamoylase — shutting the line down when supply is adequate.

Common misconceptions and pitfalls

• "Competitive inhibition lowers Vmax." No — at infinite substrate you always reach the same Vmax. It's Km that changes. If your data show Vmax dropping, it is not purely competitive.
• "Non-competitive means the inhibitor binds outside the active site, so it must be weak." Binding site has nothing to do with potency. Ki sets potency; a nanomolar non-competitive inhibitor is far more potent than a millimolar competitive one.
• "Adding substrate always relieves inhibition." Only for competitive. For uncompetitive, more substrate makes it worse because it generates more ES for the inhibitor to attack.
• "Km is the inhibitor's binding constant." Km describes the substrate; Ki describes the inhibitor. They are independent. Conflating them is the single most common kinetics error.
• "Mixed and non-competitive are the same thing." Pure non-competitive (equal Ki for E and ES) is the symmetric special case of mixed inhibition. Real allosteric inhibitors are usually mixed, shifting both Vmax and Km.
• "You can read the class off a single velocity." You need a substrate series ± inhibitor. One v value can be reproduced by all three classes at the right concentrations; only the shape of the curve discriminates.