Endotoxin, Sterility and What “Research Grade” Does Not Mean

Last reviewed 9 August 2026

“Research grade” reads like a specification. It is not one. No pharmacopoeia contains an article of that name, no statute attaches a meaning to the phrase, and no test establishes that a material meets it. It describes an intended use, which carries no information about what was measured.

Three attributes are the usual casualties: sterility, bacterial endotoxin content and microbial count. Each is a statement about what is absent, and an absence cannot be inferred from the two numbers a research certificate normally carries. A purity percentage and a conforming mass are both directed at the peptide fraction, not at a viable organism, a lipopolysaccharide or a mould count.

What the phrase does not name

Contrast a designation that carries content. The two water articles compared in the diluent page are pharmacopoeial articles, each defined by a monograph fixing composition, container designation, labelling and numerical limits: a material either complies or it does not. There is no monograph for “research grade” — nothing to comply with, and nothing to fail.

Manufacturing standards attach the same way, by legal classification and use rather than by molecule. ICH Q7 is explicit about its own trigger: When a material is classified as an API in the region or country in which it is manufactured or used in a drug product, it should be manufactured according to this Guide. For sterile material it applies only up to the point immediately prior to the APIs being rendered sterile: even inside the medicines framework, the standard governing synthesis stops short of the step that makes a substance sterile.

The phrase with the closest thing to a codified meaning is Research Use Only, and it is not about chemical grade at all: 21 CFR 809.10(c)(2)(i) requires an in vitro diagnostic product in the laboratory research phase to bear the statement For Research Use Only. Not for use in diagnostic procedures. That is a labelling rule for diagnostic devices. What governs a laboratory chemical in Great Britain is chemical-safety law — GB CLP classification and the REACH safety-data-sheet duty — which sets no sterility requirement and no endotoxin limit, since neither is a chemical-hazard question. Lawful supply is a separate matter, set out in the page on the UK regulatory position.

Three absences, three methods

None substitutes for another. The commonest error is reading the first row as covering the second.

Endotoxin is a molecule, not an organism

Raetz and Whitfield (2002) describe bacterial lipopolysaccharide as consisting of a hydrophobic domain known as lipid A (or endotoxin), a nonrepeating ‘core’ oligosaccharide, and a distal polysaccharide (or O-antigen), and identify TLR4 as the lipid A signalling receptor of animal cells. Three consequences follow.

Killing the organism does not remove the molecule

Lipid A is a structural constituent of the Gram-negative outer membrane, so a process that kills the cell liberates it rather than destroying it. EU GMP Annex 1, in force since 25 August 2023, keeps the operations apart, defining depyrogenation as A process designed to remove or inactivate pyrogenic material (e.g. endotoxin) to a specified minimum quantity in a glossary entry separate from sterilisation, and requiring raw materials to be controlled for bioburden and for endotoxin as two named things. Lyophilisation is defined there too, as a drying process undertaken primarily to achieve product or material stability — the vocabulary of stabilisation, not of sterilisation, and covered in the page on lyophilisation, mannitol and cake appearance.

It is thermally robust well past sterilising conditions

USP ⟨85⟩ directs that glassware be depyrogenated in a hot-air oven, recording that Commonly used minimum time and temperature settings are 30 minutes at 250°. Li, Wilbur and Mintz (2011) modelled inactivation kinetics for an Escherichia coli O113:H10 endotoxin standard in water at 210–270 °C and 6.2 MPa, in a bench-scale continuous-flow reactor, reporting roughly 5 log of inactivation within about one second at 250 °C followed by a markedly slower phase. Those are hydrothermal conditions rather than the dry heat of a depyrogenation oven, so the two figures describe different processes and are not directly comparable.

A sterilising filter is directed at the organism

Annex 1 defines a sterilising grade filter as one that will remove a defined microbial challenge from a fluid or gas producing a sterile effluent, usually at a pore size of 0.22 µm or less. The challenge validated is microbial. Clearing a lipopolysaccharide from solution is a separate problem, reviewed by Magalhães and colleagues (2007), who concluded that efficient, cost-effective endotoxin removal remains challenging.

How endotoxin is measured, and what the number is

The classical test is biological. USP ⟨85⟩ states that it uses Limulus Amebocyte Lysate (LAL) obtained from the aqueous extracts of circulating amebocytes of horseshoe crab, by gel-clot or photometric techniques, gel-clot being the referee method in dispute. Animal-free reagents now sit beside it: Ph. Eur. 2.6.32, for recombinant Factor C, appeared in Supplement 10.3 in 2020, and USP ⟨86⟩ became official in May 2025. The result is not a mass but an activity against a defined preparation — the USP reference standard has a defined potency of 10,000 USP Endotoxin Units (EU) per vial, one such unit equalling one International Unit.

Nor is a bare figure interpretable without its method. Ph. Eur. 5.1.10 requires a preliminary test for interfering factors, since some materials cannot be adjusted to pH 6.0-8.0 or they inhibit or activate gel formation; it warns that endotoxins may adsorb onto tubes and pipettes of certain plastics and glasses; it notes that cellulose derivatives, as β-D-glucans, can produce false positives; and it requires any interference-removal step to be shown by spike recovery not to strip endotoxin with the interference. An endotoxin row carrying none of that is the same class of assertion as a purity figure with no chromatogram, a pattern set out in the page on reading a certificate of analysis.

The limit is a property of a product, not of a substance

Here the question becomes structural. Ph. Eur. 5.1.10 sets the endotoxin limit for the active substance of a medicinal product as K / M, where K is a threshold constant per kilogram of body mass and M is the maximum recommended quantity of that product per kilogram of body mass. K is not a single number: Table 5.1.10.-1 suggests a different value for each administration route it covers, spanning a 25-fold range, and for routes it does not cover the chapter fixes no constant at all, leaving the criterion to be established from development data. The limit therefore depends on the product and its route of administration and is stated in the monograph.

Both terms are properties of a product with a defined, authorised use. A substance with no monograph and no defined use has no M, so no limit can be computed for it at all. That is why an endotoxin row on a research certificate is awkward to read even where the measurement is sound: ICH Q7 asks a certificate to list each test including the acceptance limits, and the numerical results obtained, and here the second half can be supplied while the first cannot exist. A result without a criterion is a measurement, not a judgement.

Sterility is a manufacturing outcome, not a test result

The compendial test says so about itself, in its opening lines. USP ⟨71⟩: These Pharmacopeial procedures are not by themselves designed to ensure that a batch of product is sterile or has been sterilized. This is accomplished primarily by validation of the sterilization process or of the aseptic processing procedures. A pass, it adds, indicates only that no contaminating micro-organism was found in the sample examined, under the conditions of the test.

The mechanics explain that modesty. The test runs by membrane filtration at a pore size not greater than 0.45 µm, or by direct inoculation, in two media incubated for not less than 14 days. It is destructive, so the containers examined are precisely those that can never be supplied, and it is sampled: for a batch of more than 500 items the chapter requires only 2 per cent or 20 items, whichever is less, per medium. Ph. Eur. 5.1.9 states the consequence plainly: interpretation rests on the assumption that the contents of every container in the batch, had they been tested, would have given the same result.

Annex 1 states the conclusion outright: Monitoring or testing alone does not give assurance of sterility, and Sole reliance for sterility or other quality aspects should not be placed on any terminal process or finished product test. What carries a sterility claim is the rest of that document — facility design, validated sterilisation, environmental monitoring, a documented contamination control strategy. Where a product is terminally sterilised, Annex 1 defines that process as achieving a predetermined sterility assurance level (SAL) of 10⁻⁶ or better; aseptically processed material is instead subject to periodic aseptic process simulation, for which the Annex sets a target of zero growth. A line on a certificate is the smallest component of a claim of either shape.

Between sterile and unspecified lies a middle ground: under USP ⟨61⟩/⟨62⟩ and ⟨1111⟩ a substance can carry a microbial limit without being sterile, which is a genuine specification.

Pyrogens are a wider category than endotoxins

Endotoxins are the most common cause of pyrogenic contamination in pharmaceutical products, which is why the endotoxin test came to stand in for the pyrogen test. Ph. Eur. 5.1.10 states the inference with its condition attached: the absence of bacterial endotoxins generally justifies concluding that pyrogenic components are absent, provided the presence of non-endotoxin pyrogenic substances can be ruled out. Elsewhere it notes flatly that the terms ‘pyrogens’ and ‘endotoxins’ denote groups of entities that do not coincide completely. That conditional is now formalised: at its 179th session in June 2024 the Ph. Eur. Commission adopted 57 revised texts with the rabbit pyrogen test deleted, alongside a new general chapter, 5.1.13 Pyrogenicity, in Supplement 11.8 for implementation on 1 July 2025 — under which the endotoxin test is selected where non-endotoxin pyrogens can be excluded, and the monocyte-activation test where they cannot.

What a silent certificate establishes

The two numbers on a research certificate cannot carry these attributes even in principle. Chromatographic purity is a ratio of detector responses among species that both elute under one gradient and absorb at one wavelength, examined in the page on area percent and net peptide content; a conforming mass establishes elemental composition against a known sequence. Neither is directed at a lipopolysaccharide, a bacterial spore or a mould. Nor is the omission exotic for this class of material: USP ⟨1503⟩, on the quality attributes of synthetic peptide drug substances, lists microbiological contamination and bacterial endotoxins among the attributes a specification should cover.

Stated and measured figures diverge, as an adjacent material class shows. Schwarz and colleagues (2014) assayed five commercially obtained recombinant proteins by amoebocyte lysate and reported that three carried more endotoxin than the supplier had stated; one carried approximately 140 pg per microgram of protein against a stated ceiling equivalent to 100 pg, while the same product from a second supplier carried approximately 2 pg per microgram. They further reported that human CD1c+ dendritic cells in culture responded to quantities in that range. Those were bacterially expressed proteins, a route with an endotoxin source built into it, so the finding does not transfer to solid-phase synthesis; what it establishes is that a stated level and a measured level are different quantities. An attribute absent from a certificate was not tested, which is a different statement from tested and found absent.

NovoVita’s position

NovoVita’s published specification claim is a floor rather than a point value: third-party tested at greater than 99% purity. Third-party certificates are not published, for the reasons set out on the page explaining that position.

Purity is a statement about composition. It is not a statement about sterility, endotoxin or pyrogenicity — separate attributes, established by separate methods, against criteria that exist only where a monograph and a defined use exist. Conflating them is the commonest misreading in this category. Purity is also not a claim that a compound does anything, and nothing described in the compound library is authorised for human or veterinary use. Storage of the solid is covered in the page on stability and shelf life, and the diluents sit in the reconstitution consumables section.

References

1. United States Pharmacopeia. General Chapter ⟨85⟩ Bacterial Endotoxins Test (PDG-harmonised text). Definition and techniques; Apparatus and glassware; Preparation of the standard endotoxin stock solution; Establishment of endotoxin limits. Regulatory document.
2. United States Pharmacopeia. General Chapter ⟨86⟩ Bacterial Endotoxins Test Using Recombinant Reagents. Approved by the USP Microbiology Expert Committee 26 July 2024; published for early adoption November 2024; official May 2025. Regulatory document.
3. United States Pharmacopeia. General Chapter ⟨71⟩ Sterility Tests (PDG-harmonised text). Culture media and incubation temperatures; Number of articles to be tested; Membrane filtration. Regulatory document.
4. United States Pharmacopeia. General Chapters ⟨61⟩ Microbiological Examination of Nonsterile Products: Microbial Enumeration Tests; ⟨62⟩ Tests for Specified Microorganisms; ⟨1111⟩ Microbiological Examination of Nonsterile Products: Acceptance Criteria for Pharmaceutical Preparations and Substances for Pharmaceutical Use. Regulatory documents.
5. United States Pharmacopeia. General Chapter ⟨1503⟩ Quality Attributes of Synthetic Peptide Drug Substances. Regulatory document.
6. European Pharmacopoeia. General text 5.1.10, Guidelines for using the test for bacterial endotoxins (01/2010:50110), sections 1, 2, 4, 7, 8 and 10, and Table 5.1.10.-1. Regulatory document.
7. European Pharmacopoeia. General text 5.1.9, Guidelines for using the test for sterility: guidance to manufacturers; observation and interpretation of results. Regulatory document.
8. European Pharmacopoeia. General chapters 2.6.14, Bacterial endotoxins; 2.6.32, Test for bacterial endotoxins using recombinant factor C (Supplement 10.3, available from 1 July 2020); 2.6.30, Monocyte-activation test; 5.1.13, Pyrogenicity. Regulatory documents.
9. EDQM, Council of Europe. Ph. Eur. Commission, 179th session, June 2024: adoption of 57 revised texts with the rabbit pyrogen test deleted, and of general chapter 5.1.13 Pyrogenicity, published in Supplement 11.8 with implementation 1 July 2025. Regulatory publication.
10. European Commission. EudraLex Volume 4, Annex 1: Manufacture of Sterile Medicinal Products, C(2022) 5938 final, 22 August 2022; in force 25 August 2023. Sections 1, 2.1, 2.2, 2.7, 9.32, 9.46 and Glossary. Regulatory document.
11. International Council for Harmonisation. Good Manufacturing Practice Guide for Active Pharmaceutical Ingredients, Q7, Step 4 version dated 10 November 2000. Sections 1.2, 1.3 and 11.4. Regulatory document.
12. 21 CFR 809.10(c)(2)(i), labelling for in vitro diagnostic products in the laboratory research phase of development; and US Food and Drug Administration, Distribution of In Vitro Diagnostic Products Labeled for Research Use Only or Investigational Use Only, guidance for industry and FDA staff, 25 November 2013. Regulatory documents.
13. Regulation (EC) No 1907/2006 (REACH) as assimilated in Great Britain, Article 31 and Annex II; Regulation (EC) No 1272/2008 (CLP) as assimilated in Great Britain. Regulatory documents.
14. Raetz CRH, Whitfield C. Lipopolysaccharide endotoxins. Annu Rev Biochem. 2002;71:635–700. PMID 12045108. DOI 10.1146/annurev.biochem.71.110601.135414. Review.
15. Li L, Wilbur CL, Mintz KL. Kinetics of hydrothermal inactivation of endotoxins. Appl Environ Microbiol. 2011;77(8):2640–7. PMID 21193667. DOI 10.1128/AEM.01460-10. In vitro, physicochemical kinetics.
16. Magalhães PO, Lopes AM, Mazzola PG, Rangel-Yagui C, Penna TCV, Pessoa A Jr. Methods of endotoxin removal from biological preparations: a review. J Pharm Pharm Sci. 2007;10(3):388–404. PMID 17727802. Review.
17. Schwarz H, Schmittner M, Duschl A, Horejs-Hoeck J. Residual endotoxin contaminations in recombinant proteins are sufficient to activate human CD1c+ dendritic cells. PLoS One. 2014;9(12):e113840. PMID 25478795. DOI 10.1371/journal.pone.0113840. In vitro, human primary cells.