A kinase inhibitor with a 2.4 µM Ki - real or promiscuous?
A student screens a small molecule against a kinase with KM = 0.20 mM for ATP and Vmax = 100 U mg^−1. Initial rates are measured at [ATP] = 0.05–0.8 mM. The compound is kept as a 10 mM DMSO stock and assays contain 1 % DMSO. Its aqueous solubility limit is 10 µM. Dynamic light scattering shows 50–80 nm particles above 3 µM, but the cuvette is clean at 1 µM. With 5 µM compound the fitted parameters are KM(app) = 0.62 mM and Vmax(app) = 101 U mg^−1 (the student quotes Ki = 2.4 µM). The compound also inhibits two unrelated enzymes with similar potency, and its IC50 shifts when enzyme concentration is doubled. Triton X-100 is available.
1. From the fitted parameters, confirm the inhibition mode and calculate Ki, showing the arithmetic.
Read the worked answer
Vmax is unchanged (100 → 101 U mg^−1, within error) while KM rises 0.20 → 0.62 mM. That is the signature of competitive inhibition: only the apparent KM changes, because inhibitor and substrate compete for the same free-enzyme pool.
For competitive inhibition, KM(app) = KM(1 + [I]/Ki): 0.62 = 0.20 (1 + 5/Ki) 3.10 = 1 + 5/Ki 5/Ki = 2.10 → Ki = 2.38 µM ≈ 2.4 µM
Consistency checks that make the conclusion credible:
- The highest [ATP] (0.8 mM) is 4× KM(app), so the fit is not starved of the saturating region.
- All inhibitor concentrations used are below the 10 µM solubility limit, so no free precipitate is present in this particular run.
- Note that the classical Cheng–Prusoff relation IC50 = Ki(1 + [S]/KM) must not be applied blindly here; the KM shift was measured directly, which is the safer route.
2. The potency is called a "lead". Should the Ki be trusted? Give the two most informative experiments and what result would overturn the interpretation.
Read the worked answer
Several red flags, none individually decisive:
- Aggregation. Particles at >3 µM, and the working concentration is 5 µM. The measured Ki (2.4 µM) lies below the aggregation onset, but aggregates at 5 µM will sequester and partly denature enzyme, which mimics non-competitive or mixed inhibition. A 50 nm particle suspension also scatters light and can adsorb the substrate.
- Non-specific behaviour. Inhibition of unrelated enzymes at similar potency and dependence of IC50 on enzyme concentration are hallmarks of promiscuous inhibition, not of a genuine ATP-site binder. A true competitive inhibitor's IC50 is independent of [E] when [E] ≪ Ki.
Experiments:
- Detergent counter-screen: repeat with 0.01 % Triton X-100. If the inhibition vanishes or Ki rises by ≫5-fold, it is aggregation-based.
- Mechanism check with [E] variation and a washout/dilution experiment: a reversible, site-specific inhibitor gives IC50 independent of [E] and recovers activity immediately on dilution; a colloidal aggregate does not.
Overturning result: any of (i) loss of potency in detergent, (ii) IC50 falling with decreasing [E], (iii) a Hill slope ≫1, (iv) time-dependent, non-equilibrium inhibition - all of these mean the 2.4 µM Ki is an artefact and must not be reported as a lead parameter.
3. Assuming the competitive model holds, predict KM(app) at 10 µM inhibitor and comment on whether that measurement is worth making.
Read the worked answer
KM(app) = 0.20 (1 + 10/2.38) = 0.20 × 5.20 = 1.04 mM, i.e. a ~5-fold shift, with Vmax still 100 U mg^−1.
Practical judgement:
- To measure KM(app) = 1.04 mM properly, the substrate range must extend to ~4 mM ATP, far above the 0.8 mM used. Otherwise the curve is truncated below saturation, KM and Vmax become strongly correlated in the fit, and the 5-fold shift will be underestimated.
- 10 µM is the solubility ceiling and above the 3 µM aggregation threshold, so the experiment is being run at exactly the concentration where the interpretation is least reliable. The 5-fold shift, if real, is only convincing in the presence of detergent.
- A better design: hold [I] at 2 µM (below the aggregation onset), extend [ATP] to 4 mM, and confirm that Vmax is unchanged. A constant Vmax across the series is the strongest single piece of evidence that the compound is truly competitive.