Endotoxin and Contamination Control in Hydrogel Preparation

Team Dynamic Matrices | 31 March 2026

A hydrogel can pass every sterility check your lab runs and still carry enough endotoxin to activate a macrophage. That sentence surprises more people than it should. This post is the practical companion to our argument for why endotoxin matters: where it comes from, why standard QC misses it, and the routine we use to keep it out.

Sterile is not endotoxin-free

Sterility means no viable organisms. Endotoxin is the lipopolysaccharide left behind from Gram-negative bacteria that were once present. It is heat-stable: autoclaving kills bacteria and leaves their LPS intact, and standard dry-heat sterilisation at 160 to 170 °C does not destroy it either. Filtration through 0.22 µm removes bacteria and passes LPS aggregates freely. So a filter-sterilised, autoclaved-glassware preparation can be perfectly sterile and heavily pyrogenic.

Mammalian cells respond at picogram-per-millilitre levels through TLR4 (Beutler and Rietschel, Nature Reviews Immunology, 2003). Immune cells and tumour-immune co-cultures are the most sensitive, but epithelial and stromal cells also respond, and the readout is inflammatory signalling that looks like biology.

Where it enters

• Raw materials. Polymers produced by bacterial fermentation (hyaluronic acid, alginate, some PEGs), animal-derived proteins, and peptides synthesised without endotoxin control. Gorbet and Sefton (Biomaterials, 2005) documented how common this is in biomaterials research.
• Water. Deionised or distilled water that is not certified endotoxin-free. Water systems grow biofilms.
• Buffers and salts. Made up in that water, stored in that glassware.
• Labware. Glass and plasticware that has been used, washed and autoclaved. Washing removes bacteria; it does not remove LPS adsorbed to glass.
• Handling. Gloves, benches and pipette tips are minor sources but add up over a long preparation.

None of these require a lapse in sterile technique. Endotoxin contamination is a supply-chain and materials problem, not a hygiene problem.

Why standard QC misses it

Bioburden testing detects viable organisms. Visual inspection detects turbidity. Mycoplasma PCR detects mycoplasma. None of them measure LPS. The only assays that do are the Limulus amebocyte lysate (LAL) family (gel-clot, turbidimetric, kinetic chromogenic) and the recombinant Factor C (rFC) assay, which is animal-free and now accepted by the European Pharmacopoeia and USP. If your QC panel has no LAL or rFC line, you have no endotoxin data.

The prevention and testing routine we use

Sourcing

1. Buy polymers, peptides and proteins with a stated endotoxin specification and a certificate of analysis. "Cell culture grade" without a number is not a specification.
2. Use certified endotoxin-free water (under 0.005 EU/mL) for every solution that touches the gel. Buy it or validate your own system by LAL quarterly.

Preparation

1. Depyrogenate glassware by dry heat at 250 °C for at least 30 minutes (or 180 °C for 3 hours). This destroys LPS; ordinary autoclaving does not.
2. Prefer single-use, certified pyrogen-free plasticware for anything that cannot be dry-heated.
3. Keep preparation time short and solutions cold. Bacterial growth in a buffer at room temperature over a weekend is a classic source.

Testing

1. Run kinetic chromogenic LAL or rFC on raw materials at receipt and on the finished hydrogel as prepared for cells. Test the gel, not just the components, because the final formulation is what cells see.
2. Set an acceptance threshold for your assay. Our recommendation: under 0.1 EU/mL for macrophage, dendritic cell or T cell work; under 0.5 EU/mL for epithelial-only culture.
3. Record the value with the batch.

Controls in immune assays

1. Include a polymyxin B-treated arm. Polymyxin binds LPS; if it changes your baseline, you have contamination.
2. Include an LPS-spiked positive control so you know what an endotoxin response looks like in your system.

How it shows up in data

Unexplained IL-6, TNF or IL-1β release; macrophage activation markers (CD80, CD86) without a stimulus; NF-κB reporter activity in the vehicle arm; and inflammatory gene signatures in RNA-seq that track with a reagent lot rather than a treatment. Any of these, particularly if they appeared after a supplier or lot change, warrant an LAL test before anything else.

What we do for DyNAtrix

Every lot is tested by LAL and the value appears on the certificate of analysis. Our components are synthetic and our water is certified. That removes the biological-extract source entirely, but we still tell customers to test their own final preparations if they are doing immune work, because their buffers and labware are outside our control.