Contents
- The short answer
- What endotoxin actually is
- Why laboratories care: contamination changes the result
- How endotoxin is measured
- Sterility is a third question
- Four tests, four questions
- Why research-use-only material usually has neither
- What to look for, and what to conclude
- The honest conclusion
- Where to go next
Key findings
- 1Endotoxin is a fragment of the outer membrane of Gram-negative bacteria. It is a molecule, not a living organism, so it survives sterilisation and filtration that kill bacteria.
- 2Chemical purity and endotoxin content are unrelated measurements. HPLC purity, mass-spectrometry identity and endotoxin testing are three separate assays, and a certificate reporting the first two has not tested the third.
- 3Endotoxin carried in on an ingredient has spoiled real laboratory work. In one published case a contaminated cholera toxin reagent pushed a cell-culture product over its endotoxin limit during quality control.
- 4A 1997 survey of 40 fetal bovine serum lots from 13 manufacturers found a median endotoxin level of 46 ng/ml, and five lots on their own triggered tumour necrosis factor production in cultured macrophage-like cells.
- 5Endotoxin testing is defined by pharmacopeial chapters. USP ⟨85⟩ covers the traditional horseshoe-crab-derived LAL assay; a separate chapter ⟨86⟩ covers testing with recombinant reagents such as recombinant Factor C.
- 6Sterility - whether viable organisms are present - is a third, different question again, tested under USP ⟨71⟩, a chapter harmonised internationally through the Pharmacopeial Discussion Group.
Evidence level
This article describes standard quality-control microbiology and its documented effect on laboratory experiments. The compendial framework for bacterial endotoxin and sterility testing, and published cases where endotoxin carried in on a reagent changed experimental results, are cited to pharmacopeial sources and peer-reviewed papers. Nothing here is a claim about what any peptide does in a living body.
Regulatory status
Bacterial endotoxin testing and sterility testing are defined by pharmacopeial general chapters - USP ⟨85⟩ and the newer ⟨86⟩ for recombinant reagents, and USP ⟨71⟩ for sterility, which is harmonised through the Pharmacopeial Discussion Group. Those chapters apply to products released under a pharmaceutical quality system. Material sold for research use only is outside that system, so no rule requires either test to be run or reported. Regulatory status is separate from the analytical facts described here and should be re-verified before it is relied on.
A peptide certificate of analysis is a chemistry document. It reports what the molecule is and how much of the sample is that molecule. Two other questions sit completely outside that document, are answered by completely different tests, and almost never appear on it at all: is there bacterial endotoxin in this material, and is anything alive in it?
The short answer
Purity and identity are chemistry. Chromatography separates a mixture; mass spectrometry weighs what comes out. Both describe molecules.
Endotoxin and sterility are microbiology. Endotoxin testing asks whether a specific bacterial membrane fragment is present. Sterility testing asks whether viable organisms are present. Neither is visible to a chromatograph.
These are three separate questions with three separate assays. A document answering the first has not touched the other two. That is not a hidden failure — it is a scope limit, and it is worth understanding rather than assuming away.
What endotoxin actually is
Endotoxin is lipopolysaccharide: a large molecule from the outer membrane of Gram-negative bacteria. It is a structural component of the cell wall, not a toxin the bacterium secretes.
Three properties follow from that, and they explain the whole topic.
It is a molecule, not an organism. It cannot be killed, because it was never alive. Sterilisation and filtration that eliminate bacteria do not automatically eliminate the membrane fragments they leave behind.
It is heat-stable. Autoclaving destroys the organism and leaves plenty of the molecule.
It is biologically potent at very low concentrations. Immune cells detect it at trace levels, which is precisely why it is measured in specialised units and why it matters in an experiment.
Put those together and you get the reason endotoxin is tested for separately. A sample can be sterile, chemically pure, and still carry endotoxin. None of the other tests would show it.
Why laboratories care: contamination changes the result
The reason to take this seriously is not hypothetical. Endotoxin arriving on an ingredient has demonstrably corrupted laboratory work.
In a 2024 report, a group manufacturing cell-based products detected endotoxin during routine quality-control testing. The investigation traced it to a cholera toxin reagent used as a supplement in the culture medium, which itself exceeded acceptable endotoxin levels. Their conclusion was procedural: measure endotoxin in every reagent used, or obtain a certified analysis from the manufacturer[1].
Study details: Lessons learned from contamination with endotoxin originated from the supplement in the cell culture medium
- Study type
- Laboratory investigation of a quality-control failure
- Population
- A cell-based product manufacturing process (laboratory materials, not a study in people)
- Primary result
- During routine quality-control testing, endotoxin was detected in cell culture products. Investigation traced the source to a cholera toxin reagent used as a medium supplement, which exceeded acceptable endotoxin levels. The authors recommend measuring endotoxin in every reagent used to fabricate cell-based products, or obtaining a certified analysis from the manufacturer.
- Limitations
- A single manufacturing investigation. It demonstrates that a supplied reagent can be the contamination source; it does not estimate how often this happens across suppliers.
- Year
- 2024
- Source
- Regenerative Therapy 27:230-233(link not yet independently re-verified)
The more striking demonstration is older and broader. In 1997, Kirikae and colleagues measured lipopolysaccharide in 40 lots of fetal bovine serum from 13 manufacturers — a reagent used in a very large share of all cell culture work. The median endotoxin level was 46 ng/ml. Five of those lots, with nothing else added, induced significant tumour necrosis factor production in cultured macrophage-like cells. Pre-treating the serum with polymyxin B, which binds lipopolysaccharide, abolished the effect, confirming it was the endotoxin doing it[2].
Study details: Endotoxin contamination in fetal bovine serum and its influence on tumor necrosis factor production by macrophage-like cells J774.1 cultured in the presence of the serum
- Study type
- Laboratory survey of reagent lots plus in vitro cell culture experiments
- Population
- 40 lots of fetal bovine serum from 13 manufacturers, tested on cultured J774.1 macrophage-like cells (cell culture, not people)
- Sample size
- 40
- Primary result
- The median endotoxin level across the 40 serum lots was 46 ng/ml. Five lots on their own induced significant tumour necrosis factor production in the cultured cells, and pre-treating the serum with polymyxin B abolished that effect, confirming lipopolysaccharide as the cause.
- Limitations
- A 1997 survey; supplier practice may have changed since. Cell culture only. It shows that a routine reagent can carry enough endotoxin to change an experimental readout, and says nothing about people.
- Year
- 1997
- Source
- International Journal of Immunopharmacology 19(5):255-262(link not yet independently re-verified)
Read that carefully for what it is. The serum was producing an immune response by itself. Any experiment running in that serum was measuring its compound plus an unmeasured background. Two labs using different serum lots could reasonably get different answers about the same compound and never know why.
That is a laboratory reproducibility problem. It is not a claim about what happens in a person, and neither of these studies was done in people.
The contamination does not have to come from an ingredient, either. A 2026 survey of 20 disposable products used in assisted reproduction found measurable pyrogen contamination in seven of them, ranging from 0.014 to 1.110 endotoxin units per product — with gas sterilisation reducing it below the detection limit[3]. Sterile single-use plastic is not automatically endotoxin-free.
Study details: Quality assessment of endotoxin contamination in consumables used for assisted reproductive technology
- Study type
- Laboratory survey of disposable products
- Population
- 20 disposable products used in assisted reproductive technology (laboratory consumables, not people)
- Sample size
- 20
- Primary result
- Seven of the 20 disposable products tested carried measurable pyrogen contamination, in the range 0.014 to 1.110 endotoxin units per product. Ozone and hydrogen peroxide gas sterilisation reduced contamination below the detection limit.
- Limitations
- A survey of one product category in one setting. It illustrates that sterilised single-use plastics can still carry endotoxin; it is not about peptides.
- Year
- 2026
- Source
- PLOS ONE 21(6):e0341246(link not yet independently re-verified)
How endotoxin is measured
The compendial framework is a pharmacopeial general chapter, not a company protocol.
The traditional method is the LAL assay. Limulus amebocyte lysate comes from the blood cells of the horseshoe crab, which clot in the presence of endotoxin through an enzyme cascade. This is the basis of USP general chapter ⟨85⟩, the Bacterial Endotoxins Test.
The newer method uses a recombinant protein. Recombinant Factor C reproduces the first enzyme in that cascade without the animal-derived lysate. USP issued a general announcement on 22 August 2023 introducing a separate general chapter, ⟨86⟩ Bacterial Endotoxins Test Using Recombinant Reagents, for publication in Pharmacopeial Forum 49(6) — noting at that point that it was not being proposed for introduction into a monograph or the General Notices[5].
Whether the two approaches give the same answer has been tested directly. Kang and colleagues compared recombinant Factor C against LAL on biopharmaceutical samples. Accuracy against reference standard endotoxin was comparable. Spike recovery was 98.9–116.9% for the recombinant assay versus 67–89% for LAL. Some products interfered more with the recombinant assay — vaccines containing aluminium adjuvants in particular — and diluting the sample resolved it[4].
Study details: A Study on the Application of Recombinant Factor C (rFC) Assay Using Biopharmaceuticals
- Study type
- Analytical method comparison
- Primary result
- Compared the recombinant Factor C assay with the traditional Limulus amebocyte lysate assay on biopharmaceutical samples. Accuracy against reference standard endotoxin was comparable. Spike recovery was 98.9-116.9% for the recombinant reagent versus 67-89% for LAL. Some products, notably vaccines containing aluminium adjuvants, produced greater interference with the recombinant assay, which dilution resolved.
- Limitations
- A method-comparison study on biopharmaceutical products, not peptides sold as research reagents. It establishes that the two assay families can be compared and validated against each other.
- Year
- 2024
- Source
- Microorganisms 12(3):516(link not yet independently re-verified)
The practical point for reading a document: "endotoxin tested" is not a complete statement. Which method, against which standard, at what dilution, and with what limit? Those details are the test.
Sterility is a third question
Sterility testing asks something different again: are there viable microorganisms in this material? The sample, or a filtered portion of it, is incubated in growth media and watched for growth over a defined period.
This sits in USP general chapter ⟨71⟩, which is internationally harmonised — USP records that a harmonised Sterility Test standard has been signed off through the Pharmacopeial Discussion Group, with the European Pharmacopoeia as the coordinating pharmacopeia[6].
Two things are worth knowing about it.
A sterility test is a sampling test. It examines a portion of a batch. A pass is evidence about the units tested, extrapolated to the batch under defined statistical assumptions — not an inspection of every vial.
It does not measure endotoxin. A sterile batch can carry endotoxin from bacteria that were present earlier and are no longer alive. This is the single most common misunderstanding in the area, and it is why the two chapters exist separately.
Four tests, four questions
| Test | Question it answers | What it cannot tell you |
|---|---|---|
| HPLC purity | What share of the detected material was the main peak? | Nothing about endotoxin, organisms, or how many milligrams are present. |
| Mass spectrometry identity | Does the main component weigh what the sequence predicts? | Nothing about how much, and nothing about biological contaminants. |
| Bacterial endotoxin test (USP ⟨85⟩, or ⟨86⟩ with recombinant reagents) | Is bacterial endotoxin present, and at what level? | Whether anything is alive; nothing about chemical purity. |
| Sterility test (USP ⟨71⟩) | Do viable organisms grow from the sample tested? | Whether endotoxin is present; it is also a sample, not a full inspection. |
Why research-use-only material usually has neither
The structural answer is simple, and it is not an accusation.
Endotoxin and sterility testing belong to a pharmaceutical release specification: a defined list of tests, each with an acceptance criterion, applied to every batch under a quality system, with the results reviewed before the batch is released. Products marketed for research use only are not released under that system. No rule sets which tests must be run, and no rule requires the results to be published.
So the absence of endotoxin and sterility results on a certificate does not mean a batch failed them. It almost always means nobody ran them. That is a different conclusion, and a more accurate one.
It also means the question "is this product endotoxin-free?" has no answer available from the document in front of you. The honest reading of a silent certificate is unknown, not fine.
What to look for, and what to conclude
- Is any microbiological test reported at all? Endotoxin and sterility are separate line items. Chemistry results do not imply them.
- If endotoxin is reported, by which method? LAL under USP ⟨85⟩, or a recombinant assay under ⟨86⟩. Different methods have different known interferences.
- Against what limit, and in what units? An endotoxin figure needs units — typically endotoxin units per milligram or per vial — and a stated limit to be interpretable.
- If sterility is reported, by what method and on how many units? A sterility statement without a method and a sample size is an assertion, not a result.
- Does the document say what was not tested? The most useful certificates state their scope explicitly. Silence should be read as absence of information.
The honest conclusion
Chemical purity is the easiest quality attribute to measure and the easiest to advertise, so it is the one that gets printed. Endotoxin and sterility are harder, cost more, and require a different laboratory — so on research-grade material they are usually simply not done.
The point of knowing this is not to be alarmed by every vial. It is to stop treating one number as though it covered everything. A purity figure is a real measurement of a real thing, and it is silent about biology. Recognising the silence is the whole skill.
Where to go next
- HPLC purity and LC-MS identity answer two different questions — the same problem one level up, inside the chemistry itself.
- Why a 10 mg peptide vial does not contain 10 mg of peptide — purity is a ratio, content is a mass.
- Peptide purity, contamination, and the limits of a certificate of analysis — what a certificate cannot tell you at all.
- Why peptides degrade: the chemistry behind stability and storage — what changes after the batch was released.
Frequently asked questions
- If a peptide is 99% pure, can it still contain endotoxin?
- Yes, and the two facts are unrelated. Purity is measured by chromatography and describes the chemical composition of the material. Endotoxin is measured by a completely separate biological assay. A sample can be chemically clean and still carry endotoxin picked up from water, glassware, a reagent or the environment, because nothing in the purity test looks for it.
- Does sterilising something remove endotoxin?
- Not necessarily. Endotoxin is a molecule from the outer membrane of Gram-negative bacteria, not a living organism. Killing or filtering out the bacteria does not automatically remove the molecule they leave behind, which is why sterility and endotoxin are tested separately rather than one standing in for the other.
- What is the difference between the LAL test and recombinant Factor C?
- Both detect bacterial endotoxin. The traditional LAL assay uses lysate from horseshoe crab blood cells. The recombinant Factor C assay uses a single recombinant protein instead. USP describes the traditional test in general chapter ⟨85⟩ and has introduced a separate chapter ⟨86⟩ for testing with recombinant reagents. A published comparison found comparable accuracy against reference endotoxin, with different interference behaviour on some products.
- Why would endotoxin matter in a laboratory experiment?
- Because it is biologically active at very low concentrations and can produce an effect that gets attributed to whatever was being tested. In one published survey, five of 40 fetal bovine serum lots induced tumour necrosis factor production in cultured cells on their own. That is a reproducibility problem: two labs testing the same compound can get different answers because their reagents differed.
- Should I expect endotoxin and sterility results on a research-use-only certificate?
- Usually they are absent. Those tests are part of a pharmaceutical release specification, and research-use-only material is not released under one. Their absence is not evidence that a batch failed them - it is evidence that nobody ran them, which is a different and more accurate thing to conclude.
References
Numbered in order of first use. Study type is shown for every source; see our methodology for how we rank evidence.
- 1.
Uehara K, Oshiro E, Ochiai A, Takagi R, Yamato M, Kato A. Lessons learned from contamination with endotoxin originated from the supplement in the cell culture medium Regenerative Therapy 27:230-233, 2024.
Case reportA cell-based product manufacturing process (laboratory materials, not a study in people)
Result: During routine quality-control testing, endotoxin was detected in cell culture products. Investigation traced the source to a cholera toxin reagent used as a medium supplement, which exceeded acceptable endotoxin levels. The authors recommend measuring endotoxin in every reagent used to fabricate cell-based products, or obtaining a certified analysis from the manufacturer.
Limitations: A single manufacturing investigation. It demonstrates that a supplied reagent can be the contamination source; it does not estimate how often this happens across suppliers.
↑ back to text - 2.
Kirikae T, Tamura H, Hashizume M, Kirikae F, Uemura Y, Tanaka S, Yokochi T, Nakano M. Endotoxin contamination in fetal bovine serum and its influence on tumor necrosis factor production by macrophage-like cells J774.1 cultured in the presence of the serum International Journal of Immunopharmacology 19(5):255-262, 1997.
Preclinical40 lots of fetal bovine serum from 13 manufacturers, tested on cultured J774.1 macrophage-like cells (cell culture, not people)n = 40
Result: The median endotoxin level across the 40 serum lots was 46 ng/ml. Five lots on their own induced significant tumour necrosis factor production in the cultured cells, and pre-treating the serum with polymyxin B abolished that effect, confirming lipopolysaccharide as the cause.
Limitations: A 1997 survey; supplier practice may have changed since. Cell culture only. It shows that a routine reagent can carry enough endotoxin to change an experimental readout, and says nothing about people.
↑ back to text - 3.
Tomari H, Sugizaki E, Ibrahim S, Hashiguchi Y, Koyama G, Nakamura Y, Nagata M, Nagata Y, Haishima Y. Quality assessment of endotoxin contamination in consumables used for assisted reproductive technology PLOS ONE 21(6):e0341246, 2026.
Preclinical20 disposable products used in assisted reproductive technology (laboratory consumables, not people)n = 20
Result: Seven of the 20 disposable products tested carried measurable pyrogen contamination, in the range 0.014 to 1.110 endotoxin units per product. Ozone and hydrogen peroxide gas sterilisation reduced contamination below the detection limit.
Limitations: A survey of one product category in one setting. It illustrates that sterilised single-use plastics can still carry endotoxin; it is not about peptides.
↑ back to text - 4.
Kang DH, Yun SY, Eum S, Yoon KE, Ryu SR, Lee C, Heo HR, Lee KM. A Study on the Application of Recombinant Factor C (rFC) Assay Using Biopharmaceuticals Microorganisms 12(3):516, 2024.
Expert analysis
Result: Compared the recombinant Factor C assay with the traditional Limulus amebocyte lysate assay on biopharmaceutical samples. Accuracy against reference standard endotoxin was comparable. Spike recovery was 98.9-116.9% for the recombinant reagent versus 67-89% for LAL. Some products, notably vaccines containing aluminium adjuvants, produced greater interference with the recombinant assay, which dilution resolved.
Limitations: A method-comparison study on biopharmaceutical products, not peptides sold as research reagents. It establishes that the two assay families can be compared and validated against each other.
↑ back to text - 5.
⟨86⟩ Bacterial Endotoxins Test Using Recombinant Reagents - General Announcement United States Pharmacopeia (USP-NF), 2023.
Guideline
Result: USP general announcement, dated 22 August 2023, introducing a new general chapter ⟨86⟩ Bacterial Endotoxins Test Using Recombinant Reagents, to be published in Pharmacopeial Forum 49(6) for public comment. The announcement states the chapter was not at that time proposed for introduction into a monograph or the General Notices.
↑ back to text - 6.
Sterility Test - PDG harmonised general chapter United States Pharmacopeia.
Guideline
Result: USP page recording that a harmonised standard for the Sterility Test has been signed off by the Pharmacopeial Discussion Group, with the European Pharmacopoeia as coordinating pharmacopeia, under USP general chapter ⟨71⟩. The page lists sign-off history back to 2002, most recently 10 June 2009.
↑ back to text
Advertising disclosure: ads on this page are for NuNu Research, a research-use-only chemical storefront that shares an owner with PeptidesInfo. Ads never influence the evidence grades, regulatory statuses, or conclusions above, which come solely from the cited primary sources.
Review status: Editorially reviewed against primary sources. This article was fact-checked against the primary sources listed in the references by our editorial team, and it has not been reviewed by a licensed clinician. It is educational content, not medical advice. Read our editorial policy and methodology. Spotted an error? Tell us.
Comments
Have a question, or a study to add? Join the discussion — no account needed. Comments are moderated before they appear.
