Endotoxin testing for peptides: LAL methods, EU per mg and how to read the result

What an endotoxin test measures

Bacterial endotoxin is lipopolysaccharide (LPS), a large molecule in the outer membrane of Gram-negative bacteria. It has three parts: a lipid anchor called lipid A, a core sugar chain and an outer polysaccharide called the O-antigen. Lipid A is the part animal cells recognise, through a receptor called TLR4. Pharmacopoeial endotoxin limits start from its threshold pyrogenic dose: the amount per kg of body mass that causes fever. An endotoxin test measures how much of the molecule is in a sample, in endotoxin units (EU).

Endotoxins are not proteins, so peptide tests such as HPLC purity do not see them. The peptide in the sample can still enhance or inhibit the test reaction, which is why the method has controls for interference.

Endotoxin can be present when no bacteria are alive. A vial can grow nothing in a sterility test and still contain endotoxin, so the two tests answer different questions. Our guide to endotoxin vs sterility sets them side by side.

Regulators treat endotoxin as a standard quality attribute for peptide medicines. The EMA guideline on synthetic peptides, in force since 1 June 2026, lists bacterial endotoxins in the specification of a peptide active substance, next to identity, purity, assay, counter-ion and water. Research peptides have no such specification, so the buyer decides whether to test.

LAL testing: what it means

LAL stands for Limulus amebocyte lysate. It is an extract of blood cells from the horseshoe crab (Limulus polyphemus or Tachypleus tridentatus). Endotoxin sets off a clotting reaction in the lysate, and the test reads that reaction. In the LAL test procedure, the lab dissolves the peptide in endotoxin-free water, dilutes it, mixes it with the lysate and incubates the mixture, usually at 37 ± 1 °C. USP <85> Bacterial Endotoxins Test is harmonised with the European Pharmacopoeia (2.6.14) and the Japanese Pharmacopoeia (4.01), so the method is the same in all three.

The chapter describes three techniques. The two photometric ones can run as a single end-point reading or as a kinetic reading over time.

Technique	What it reads	What the result looks like
Gel-clot, limit test	Whether a firm gel forms at one set concentration	Complies or does not comply with a limit
Gel-clot, semi-quantitative	The last dilution that still forms a gel	A value from the end-point dilution, or “less than” a value
Turbidimetric (end-point or kinetic)	Cloudiness as the gel forms	A number read from a standard curve
Chromogenic (end-point or kinetic)	Colour released from a synthetic substrate	A number read from a standard curve

The photometric techniques need a standard curve of at least three endotoxin concentrations, each run at least three times, with a correlation coefficient of at least 0.980. If results from different methods disagree, both the pharmacopoeia and the FDA say the gel-clot limit test decides.

One EU equals one international unit of endotoxin. Labs calibrate against a reference standard traced to the WHO International Standard for endotoxin, which is what makes EU values from different labs comparable. USP also has a separate chapter, <86>, for tests that use recombinant reagents in place of crab lysate, such as recombinant factor C (rFC).

The LAL test and pyrogens

A pyrogen is any substance that causes fever, and bacterial endotoxin is one of them. The LAL test detects endotoxin only. In 1977 FDA set out conditions for using LAL in place of the rabbit pyrogen test. Its March 2026 guidance still asks makers of medicines and devices to assess the risk of non-endotoxin pyrogens. Where that risk exists, it says the rabbit pyrogen test (USP <151>) may be more appropriate, and it names the monocyte activation test as an alternative method. A low LAL result therefore speaks to endotoxin alone.

How a reading becomes EU per mg and EU per vial

The instrument measures EU per mL of the test solution. Two more numbers turn that into a result about the product:

• EU per mg = EU per mL ÷ the peptide concentration of the solution (mg per mL).
• EU per vial = EU per mg × the mg of peptide measured in the vial.

A result printed only as EU/mL, with no concentration or volume, cannot be converted. In our September 2026 review of public certificates from 25 labs and one rating site’s certificate format (26 in all), none showed endotoxin as a number with a clear basis (EU/mg or EU/vial) in full, and 11 did so in part, for example EU/mL with no volume. The other 15 showed a pass with no number, no endotoxin result, or could not be checked.

Labs dilute the sample to get past interference from the peptide. The pharmacopoeia caps this at the maximum valid dilution (MVD): the endotoxin limit × the sample concentration ÷ the lysate sensitivity, or ÷ the lowest point on the curve. Beyond the MVD the test can no longer detect endotoxin at the limit, so the dilution belongs on the report.

How to read an endotoxin result

Check these items on any endotoxin report. Our guide to reading a COA covers the rest of the document.

Item on the report	What to look for	Why it matters
Method	USP <85> with the technique (gel-clot, kinetic turbidimetric or kinetic chromogenic), or USP <86> with rFC	Tells you whether the result is a number or a limit test
Result and unit	EU per mg and EU per vial	Lets you compare vials and batches
Limit of quantitation	“Less than x EU/mg”, with x stated	A “less than” value is still a number; a bare “not detected” is not
Replicates	At least two readings of the sample	The chapter runs the sample and the spiked sample at least in duplicate
Spike recovery	Recovery of added endotoxin between 50% and 200%	Shows the peptide did not hide or inflate the reading
Curve and dilution	Correlation coefficient, dilution factor, MVD	Shows the run was valid and the dilution was allowed
Limit and its source	Who set the limit, and on what basis	A research peptide has no official limit
Vial tested	Which vial was used, ideally one set aside unopened	Ties the result to a sample you can identify

The spike recovery deserves the most attention. The lab adds a known amount of endotoxin to a portion of your sample and measures how much it gets back. The harmonised chapter accepts the run only when the recovery is 50% to 200%. Outside that range the peptide is interfering, and the sample result is not valid at that dilution. In our September 2026 review, no lab’s main certificate showed a passed spike-recovery control in full, and three showed one in part.

Why a numeric endotoxin result beats pass or fail

A pass says the result was under some limit. It does not say how far under, or which limit. Two vials that both pass can differ tenfold. A number lets you:

• compare batches from one seller over time;
• compare two labs’ results on the same batch;
• apply a limit of your own choosing later, in place of the lab’s;
• see the method’s sensitivity, even when the answer is “less than”.

Buyers raise this often. In forum threads they report pass-only endotoxin results, ask why replicate data is not shown, and some say they distrust labs that never report a measurable endotoxin number. Rating sites reward the number too. Disclosed Labs gives 40% of a seller’s grade to safety-panel coverage, which includes endotoxin, and tracks how often each lab’s certificates report it. PeptideTrust scores sellers on whether they publish sterility and endotoxin results. See which labs report a real endotoxin number.

Maximum endotoxin levels and the USP <85> limit

For licensed injectable medicines, the endotoxin limit is calculated per product as K/M. K is a threshold dose of endotoxin per kg of body mass, and M is the largest single dose of the product per kg, or the largest dose per kg in one hour when it is given often or by infusion. The limit therefore depends on a labelled dose. A research peptide has no labelled dose, so no official limit exists for it.

That has two consequences. First, any limit printed on a research peptide report was chosen by someone: the client, the lab or a seller. The report should say who, and on what basis. Second, a lab result cannot tell you whether a level is safe for any use. It tells you how much endotoxin was in the vial tested.

What HK Analytical reports

Our endotoxin test uses LAL and reports EU per mg and EU per vial, with the limit of quantitation, replicates and spike recovery. Below the limit we print “less than” the limit as a number. Endotoxin cannot show whether live microbes are present, so it pairs with Sterility (USP <71>, 14 days). Every number comes with its raw data, and the report is a signed record you can check on /verify. Results within days of receipt. Sterility (USP <71>, 14 days) is the one test that takes 14 days, because the method incubates for 14 days. See the example report or request a quote.