The tests that gate release, and the numbers we publish.

Every protein lot is sequence-confirmed against its reference construct and released against two independent purity measurements, SDS-PAGE and RP-HPLC, because a single band on a gel can hide a co-eluting impurity that a reversed-phase separation catches, and vice versa. Release limits run ≥95–98% depending on the SKU; the exact figure is printed on that product’s datasheet. Because purification and QC sit in the same Cambridge building, a sample goes from the polish column to the gel without a courier in between.

Bacterial endotoxin is tested by the Limulus Amebocyte Lysate (LAL) method on the filled liquid and reported per milliliter (EU/mL); for purified proteins we publish the figure per microgram of protein (EU/µg) alongside it, since that is the unit a dose is set in. Endotoxin above a few tenths of an EU/µg is enough to activate TLR4 signaling in immune and stem cells and confound anything downstream that depends on a quiescent baseline, a common, and often uninvestigated, source of irreproducible cell-culture data. Representative release limits: recombinant proteins around <0.1 EU/µg; formulated media around 1 EU/mL.

Purity confirms a protein folded and stayed intact; it doesn’t confirm it still does the job. Growth factors and cytokines are released against a cell-based bioassay, a responsive cell line is dosed across a concentration series and the response (usually proliferation) is fit to a curve to derive an ED50, the concentration giving half-maximal effect. ED50 is lot-tested against a working reference standard so drift shows up before it reaches you, not after.

Media and liquid supplements are sterilized by 0.1–0.2 µm membrane filtration and tested to a USP <71>-style method before release, filled units are held, not shipped, until the sterility test comes back clean. Products intended for live cell culture are additionally tested for mycoplasma, which a standard sterility test doesn’t catch. Osmolality and pH are released against the target window on the datasheet, typically pH 7.0–7.4 and 280–340 mOsm/kg by freezing-point depression, depending on the formulation.

Chromatography media are qualified on capacity, not just chemistry. Dynamic binding capacity (DBC) is measured to 10% breakthrough with a reference protein under defined flow and residence-time conditions, ligand leakage is bounded by ELISA against the coupled ligand, and the clean-in-place tolerance printed on the datasheet is validated against the same functionalized lot.

A unit of enzyme activity only means something with the definition attached, one supplier’s “unit” isn’t necessarily another’s. Every enzyme datasheet spells out its unit definition and the assay used to measure it lot-to-lot, so a titration you ran against a prior lot still means what you think it means.