Endotoxin in Recombinant Proteins: Why It Decides Your Cell-Based Assay

Updated 2026-09-28

Endotoxin is invisible in a purity figure and decisive in a live-cell experiment. Here is what it is, where it comes from and when to specify it.

What endotoxin is and where it comes from

Endotoxin is lipopolysaccharide carried over from the host-cell envelope, which is why it is associated mainly with E. coli expression. It is not part of the protein sequence, so it cannot be designed out — it has to be removed by process and confirmed by testing.

Why it matters for some experiments and not others

In an ELISA or a Western blot, trace endotoxin is usually irrelevant. The moment living cells are involved — stimulation experiments, primary cells, signalling work — endotoxin is itself a potent stimulus and will mask the effect you actually set out to measure.

How it is specified and measured

The common grade is < 1.0 EU/µg; cell-based work usually calls for < 0.1 EU/µg, determined by the LAL method. State the endotoxin requirement in your enquiry rather than discovering it after the experiment fails — it is a specification, not an upsell.

How endotoxin is removed, and what removal costs you

Because endotoxin is not encoded in the protein sequence, it has to be removed by process rather than designed out, and every removal method takes something with it. The main options are chromatographic separation exploiting the difference between the lipopolysaccharide and the target's properties, two-phase extraction, and affinity capture with a ligand that binds endotoxin selectively. All of them add at least one purification step, which means added cost, some loss of yield, and an additional opportunity for the protein itself to be perturbed. That is the honest reason low-endotoxin grades cost more, and also the reason it is worth asking whether you need them: paying for a demanding low-endotoxin preparation to run an ELISA wastes money, while trying to stimulate primary cells with an endotoxin-screening-free preparation wastes experiments. State the requirement explicitly rather than assuming that lower is universally better.

Designing the experiment so endotoxin cannot explain the result

When a recombinant protein is applied to living cells, endotoxin is not a trace contaminant but a confounding variable, and the strongest experimental designs control for it directly rather than relying on a specification. Include a vehicle control matched in buffer and, where possible, an endotoxin-spiked control at the level your preparation could plausibly carry; if the cytokine response appears in the spiked control you have learned something about your system rather than about your protein. Heat-inactivation of the preparation is a useful additional check, since endotoxin survives it while most protein activity does not. And record which lot and which grade was used for each experiment, because an endotoxin specification is attached to a lot, not to a product name — the same catalogue number can arrive at two very different endotoxin figures.

Ask a technical question