Why an In Vitro Concentration Is Not a Dose

A paper reports an effect at ten micromolar in cultured cells. Somebody multiplies that by an assumed volume, divides by a molecular weight, and arrives at a number of milligrams. Every step of that arithmetic is valid and the result is meaningless, because the quantity being converted was never a property of an organism in the first place.

What an in vitro concentration actually describes

It describes the concentration in the medium surrounding cells in a dish, held approximately constant for the duration of the exposure, with nothing between the compound and the cells.

Each of those conditions is doing work. The concentration is imposed and maintained rather than arrived at. There is no distribution into tissues, no binding to circulating proteins, no metabolism, no clearance, and no barrier to cross. The number is an experimental setting, in the same category as the temperature of the incubator — it is something the experimenter chose, not something the system produced.

The four things that break the conversion

• Exposure profile. A dish holds a constant concentration for hours. Nothing in an organism behaves that way; concentration rises, distributes and falls, and for many peptides falls quickly — the half-life question in half-life and analog modifications. The area under a curve and a flat line at the same peak value are not equivalent exposures.
• Protein binding. In a biological fluid a substantial fraction of many compounds is bound and not free to act. In a dish with minimal protein, nearly all of it is free. The same nominal concentration therefore means different free concentrations.
• Distribution. A compound does not spread evenly through a volume. It reaches some compartments well, others poorly, and some not at all — which is why the “assumed volume” step in the arithmetic is the weakest link in the chain.
• Stability. Peptides are degraded by peptidases that a culture dish largely lacks. A compound intact for six hours in medium may not be intact for six minutes elsewhere, for the reasons in how peptides degrade.

Why the number in the dish may not even be the number in the dish

Before any extrapolation, the nominal concentration is frequently wrong in the experiment itself. The compound adsorbs to the plastic, in the way set out in adsorptive loss to surfaces, so the free concentration is lower than intended and lower still by the end of the incubation.

It may also be partly aggregated, per peptide aggregation in solution, or partly degraded during preparation. Where the stock was made by weighing, the net content of the powder is not the same as its mass, per net peptide content. A stated ten micromolar can comfortably be six or seven micromolar of intact free compound in practice, which is a large error to build a further calculation on.

The concentrations themselves are often chosen, not found

Many in vitro peptide studies work in the micromolar range. Where a receptor-mediated effect has a potency in the nanomolar range — the distinction between affinity and functional potency in Ki, IC50 and EC50 — a micromolar exposure is orders of magnitude above the concentration at which the specific mechanism operates.

At that distance from the potency, effects may be occurring through mechanisms unrelated to the one being studied. This is not a criticism of the experiments; high concentrations are often used deliberately to produce a measurable signal. It does mean the reported concentration is frequently a methodological choice rather than a biologically meaningful quantity, and converting it treats it as the latter.

What the field actually does instead

Where the question genuinely has to be bridged, the work is a discipline in its own right and bears no resemblance to unit conversion. It involves measuring free rather than nominal concentrations, characterising binding, measuring clearance and distribution empirically in a living system, building a model that relates exposure over time to effect, and testing that model against observations it was not fitted to.

Every one of those steps requires data that does not exist for most research peptides. That is the substantive point: the bridge is not difficult to build here, it is unbuildable, because the measurements it would be built from have never been made.

Reading a paper without performing the conversion

An in vitro result supports statements about mechanism: that a compound binds a target, that it produces a measurable response in a particular cell type, that the response depends on concentration in a particular way. Those are the claims the experiment can carry, and they are the useful ones.

What it does not support is any statement about quantity in an organism. The honest reading of “effect at ten micromolar in cells” is that an effect occurs at ten micromolar in those cells under those conditions — a sentence which contains no quantity applicable anywhere else, and which is why the difference between a cell model and an animal model, discussed in the limits of the published literature, matters as much as it does.

The position this leaves

The material on this site is supplied for laboratory research, and this is one of the more concrete reasons that restriction is not a formality. The published record for these compounds consists overwhelmingly of in vitro work and animal studies, and neither category contains a quantity that transfers out of the system it was measured in.

A calculation that produces a number regardless is doing arithmetic on inputs that do not support it. The arithmetic is not the problem; the premise that the starting figure describes an organism is.