Small molecules

Measure kon, koff and KD of a small molecule without a label: the target sits on the chip and the compound flows over it.

The question

How fast does my compound bind and release its target, and how strongly?

Coherent mass density (CMD) reports bound mass, so a small molecule gives a small signal. Here the target is on the chip and the compound is the analyte. If the compound carries DNA, as a DEL hit does, put the compound on the chip and use the protein as the analyte. DEL hit validation shows that format, with selectivity and off-target profiling.

The setup

ChipA 1-plex chip: carbonic anhydrase II (CA-II) on all 64 molograms
AnalyteDorzolamide (324 Da), a sulfonamide CA inhibitor. Eight 2-fold steps from 78 pM to 10 nM, 150 s each
AssaySingle-cycle kinetics (SCK), 1:1 fit on the 6 molograms with clean traces

Measured result

Top: a sensorgram of coherent mass density over 68 minutes. Eight dorzolamide injections from 78 pM to 10 nM raise the signal in steps to about 3 pg/mm², with partial dissociation between steps; a dashed 1:1 fit follows the data. Bottom: a log-log map of k_on against k_off with six mologram points close to the 0.1 nM line of equal K_D, and the median marked.
Dorzolamide on CA-II, single-cycle kinetics, 78 pM to 10 nM. Top: one mologram with its 1:1 fit. Bottom: rate constants from the 6 molograms fitted, out of 64 on the chip.

How to read it

How large is the signal?

Each bound analyte adds its own mass. If every ligand binds one analyte, the maximum response is the ligand level times the mass ratio of analyte to ligand. A 324 Da compound on a protein ligand gives a response hundreds of times smaller than an antibody on the same ligand. Compare that response with the noise of the instrument.

Signal size

Set the analyte mass, the ligand mass and the ligand level. Analytes land on the ligands of the ridges; the grooves carry backfill and bind nothing. The scale on the right compares the maximum response with the specified noise. It starts at the measured pair above: dorzolamide on CA-II (29 kDa) in a dense ligand layer. The measured 3 pg/mm² is about two thirds of that maximum. The buttons show the documentation example level.

Rmax = (MWanalyte / MWligand) × ligand level × stoichiometry, from the experiment design documentation. The lab uses a stoichiometry of 1 and a fully active ligand. The ligand level is the immobilization level in pg/mm²; a protein monolayer gives about 400 pg/mm² of CMD. The noise line is the specified 0.05 pg/mm² over 20 minutes at 5 pg/mm². At 50 pg/mm², the specified noise is 0.2 pg/mm². The orange line marks the measured dorzolamide result above, about 3 pg/mm².

Why it works

The ridges and the grooves of a mologram sit about 200 nm apart and see the same solution. A DMSO or buffer step changes both equally and cancels; a 10% DMSO step adds only 1 to 2 pg/mm². Specified noise and drift of 0.05 pg/mm² over 20 minutes keep a response of a few pg/mm² readable. When the compound sits on the chip, the protein is the analyte and gives the larger signal.

Read more: why DMSO and bulk steps cancel, what limits small analytes, and how bound mass becomes coherent mass density.

More