Skip to content
Black Week: 20% off everything, code BLACK

Search

What are you looking for?

Type at least 2 characters to see suggestions.

Quality and Legal Status Marek Wawrzyk

Endotoxin Testing in Peptides: Why HPLC Does Not Tell the Whole Quality Story

Material status information.

Endotoxin Testing in Peptides: Why HPLC Does Not Tell the Whole Quality Story

Material status information. This article concerns the quality control of chemical reagents intended solely for laboratory research (Research Use Only). The peptides discussed are not medicinal products, dietary supplements or foods and are not intended for use in humans or animals. This material is informational and educational and concerns analytical methodology.

A peptide solution can pass sterilising filtration, produce a clean sterility-test result and still contain enough endotoxin to confound an entire cell-culture experiment. The reason is straightforward: an endotoxin is not a bacterium. It is a fragment of the bacterial cell envelope that remains in a sample long after the bacterium itself has ceased to exist. A filter with a pore diameter of 0.22 µm retains a bacterial cell, but a lipopolysaccharide molecule can pass through it.

This is why peptide-material quality is increasingly assessed along two separate axes: chemical purity by HPLC and endotoxin burden by an LAL or rFC test. The first indicates what is in the vial. The second indicates whether what is in the vial also contains something that the chromatograph cannot see.

What are endotoxins?

Endotoxins are lipopolysaccharides (LPS), components of the outer membrane of Gram-negative bacteria. The molecule has three parts: lipid A, which is responsible for biological activity; an oligosaccharide core; and the O antigen. Endotoxins are released in two ways: during normal bacterial growth, when outer-membrane vesicles (OMVs) are shed, and in incomparably larger amounts when a bacterium dies and its envelope breaks down. Paradoxically, this can also occur as a result of effective sterilisation.

Three properties make endotoxins an analytical problem distinct from ordinary microbiological contamination:

  • Heat stability. Autoclaving at 121 °C destroys bacteria but does not break down endotoxins. Removing them from laboratory glassware, a process called depyrogenation, requires dry heat at approximately 250 °C for about 30 minutes. No peptide would withstand those conditions. The validation criterion for such a process is a 3-log, or thousand-fold, reduction in endotoxin burden.
  • Passage through sterilising filters. An LPS monomer has a mass of approximately 10–20 kDa and forms micellar aggregates reaching about 1,000 kDa in aqueous solution. These are still orders of magnitude smaller than 0.22 µm filter pores. Filtration through such a membrane retains bacterial cells, not their dissolved fragments. Endotoxin removal requires separate operations, such as ultrafiltration with a cut-off of approximately 10 kDa or adsorption on anion-exchange media or polymyxin B.
  • Activity at trace concentrations. A cellular response to LPS is observed at concentrations measured in picograms per millilitre. Primary human monocytes and dendritic cells can respond to just a few pg/mL. This is far below the detection threshold of a chromatograph’s UV detector.

This leads to a principle that is often misunderstood: sterility and the absence of endotoxins are two different parameters. A sample can be sterile and strongly pyrogenic at the same time.

HPLC and endotoxins: why they require two independent tests

High-performance liquid chromatography separates the components of a sample and measures their relative proportions, most commonly with UV detection at approximately 214 nm, a wavelength characteristic of the peptide bond. A result of “purity ≥98% by HPLC” means that at least 98% of the total peak area in the analysed profile is attributed to the main compound. A chromatogram is very effective at showing what is relevant in peptide synthesis: truncated sequences, oxidised and deamidated forms, hydrolysis products and isomers.

Endotoxins do not belong to this category. They do not produce a peak that can be compared with the main peak, they are not related compounds formed during synthesis, and their concentrations are several orders of magnitude below the sensitivity of UV detection. The scale of the difference is easy to see numerically: a limit of approximately 1 EU/mg corresponds to roughly 0.1 ng of endotoxin per milligram of peptide, or about 0.00001% by mass, whereas the detection threshold for an impurity in UV chromatography is typically 0.02–0.05% of total peak area. A chromatograph simply does not see them, nor is that its purpose.

Parameter HPLC purity Endotoxin test (LAL or rFC)
What it measures The percentage of the main compound and the related-substance profile The presence and quantity of bacterial lipopolysaccharides
Result Percentage of total peak area, for example ≥98% Endotoxin units per mass or volume, such as EU/mg or EU/mL
Detects synthesis-related impurities Yes: truncated peptides, oxidised forms and isomers No
Detects endotoxins No Yes
Question it answers What is in the vial and in what proportion? Does the material carry a pyrogenic burden?

The practical conclusion is that a certificate reporting 99.2% HPLC purity contains no information about the endotoxin level. The same applies to identity confirmation by mass spectrometry. These are three different questions and three different tests.

Where endotoxins in peptide materials come from

Solid-phase peptide synthesis (SPPS) is itself a chemical process. Bacteria do not participate in it, so the synthesis does not generate endotoxins. Risk arises at later stages wherever an aqueous phase or contact surface enters the process:

  • Process water is the most common source. Purified water is not equivalent to water for injection (WFI) with respect to endotoxin limits. The WFI monograph requires a result of ≤0.25 EU/mL as a mandatory parameter, while the purified-water monograph does not impose an equivalent routine limit.
  • Buffers and aqueous solutions are used during cleavage, purification and lyophilisation.
  • Contact surfaces, including glassware, vials, stoppers and equipment, may introduce contamination if they have not undergone depyrogenation.
  • The environment and handling matter, including air quality in the filling area and aseptic technique.
  • Recombinant expression is a special case. Proteins and longer peptides produced in Escherichia coli are formed inside a Gram-negative bacterium, so endotoxin burden is inherent to the production technology and requires a deliberate removal step.

The last point explains why endotoxin control is especially important for protein materials, including recombinant growth factors and hormones, rather than only for short synthetic peptides.

How endotoxins are tested

The pharmacopeial standard is the bacterial endotoxins test described in USP <85> and Ph. Eur. 2.6.14. The rabbit pyrogen test was historically used for this purpose, but the European Pharmacopoeia has withdrawn it completely: references were removed from its texts in 2025, and chapter 2.6.8, Pyrogens, ceased to apply on 1 January 2026. Current testing is based on in vitro methods that use the enzyme cascade in horseshoe-crab amoebocyte lysate from Limulus polyphemus, hence the name LAL, or Limulus Amebocyte Lysate.

Four approaches are encountered in laboratory practice: three LAL formats and one recombinant method.

  • Gel-clot method. This is the simplest format: the sample either forms a clot or it does not. It produces a threshold result, such as “below the sensitivity limit”, rather than a numerical value. A semi-quantitative dilution series can estimate concentration to the accuracy of the dilution step.
  • Turbidimetric method. This measures the kinetics of increasing turbidity and produces a quantitative result.
  • Chromogenic method. This measures a coloured product of the enzymatic reaction. Its kinetic format is often selected for highly sensitive quantitative testing.
  • rFC, or recombinant factor C. This animal-free format uses a recombinant version of the first enzyme in the cascade. It is described in Ph. Eur. 2.6.32, effective since July 2021, and since 2025 has also had its own USP chapter, <86> Bacterial Endotoxins Tests Using Recombinant Reagents. It is therefore no longer a marginal alternative but a fully recognised pharmacopeial approach.

Results are reported in endotoxin units (EU) relative to mass or volume. For a peptide lyophilisate, EU/mg is typical; for a solution, EU/mL is typical. The unit is defined against an international reference standard as a measure of activity rather than mass. As a general approximation, 1 EU corresponds to about 0.1 ng of standard endotoxin, although the exact conversion depends on the standard used. The monocyte activation test (MAT, Ph. Eur. 2.6.30) is a separate method that also detects pyrogens other than endotoxins and is used less often when a broader assessment is required.

Another important concept is matrix interference. Some substances disrupt the enzyme cascade and increase or decrease the reading. Endotoxin masking in matrices containing chelators and surfactants, known as low endotoxin recovery, is a separately described phenomenon. A reliable report therefore confirms recovery from a spiked sample through a positive product control. This demonstrates that the sample did not conceal endotoxin that was present. The pharmacopeially accepted recovery range is 50–200%.

Why this matters in laboratory research

This is the central issue, and in the context of research reagents it may matter more than the quality claim itself. Endotoxin is not an inert contaminant. It is a highly biologically active ligand.

LPS binds to the TLR4/MD-2 receptor complex with the involvement of LPS-binding protein (LBP) and CD14. This activates the MyD88- and TRIF-dependent adaptor pathways and the NF-κB transcription factor. The result is expression of pro-inflammatory cytokines, primarily TNF-α and IL-6, as well as the IL-1β precursor. The mechanism is so well established that LPS is the canonical positive stimulation control in inflammation research: when researchers want to induce an inflammatory response in a macrophage culture, they add lipopolysaccharide.

The consequence for an experiment is unavoidable. If the tested peptide carries an endotoxin burden, the cellular system receives a signal that does not originate from the molecule under investigation. The observed “activity” may be entirely attributable to contamination. The problem has the greatest impact on:

  • immunological studies and inflammation models, including cytokine measurements, macrophage and microglial activation, and neuroinflammation models;
  • cell-proliferation and viability assays, because LPS modifies cell survival and the culture’s metabolic profile;
  • endothelial and barrier models, because endotoxin changes permeability and adhesion-molecule expression;
  • in vivo models, where a febrile response and systemic immune activation can distort endpoints.

Peptides investigated for anti-inflammatory activity present a particularly ironic case. Material contaminated with pro-inflammatory LPS acts in the opposite direction to the hypothesis under study. It can mask a real effect or generate a result that another laboratory cannot reproduce. This is not merely theoretical: in a published analysis of commercial recombinant proteins, residual endotoxins were sufficient to activate human dendritic cells, and the effect attributed to the protein disappeared after the endotoxin was neutralised (Schwarz et al., PLoS ONE, 2014). The same mechanism applies to peptide preparations, although it is described less often in that context.

In other words, endotoxin control in research material is not a cosmetic addition to a certificate. It is a condition for ensuring that the experimental result reflects the compound under study rather than the quality of the water used during purification.

How to read an endotoxin test report

A trustworthy document contains several elements that should be reviewed in the following order:

  1. Batch number. It must match the number on the vial. A report for a different batch describes different material.
  2. Method. This may be a gel-clot, turbidimetric or kinetic chromogenic LAL test, or an rFC test. The gel-clot method provides a threshold answer; kinetic methods produce a numerical value.
  3. Result with a unit. This should be EU/mg or EU/mL. A percentage or the word “negative” without a unit conveys little information.
  4. Assay sensitivity. A result such as “<0.5 EU/mg” means “below the detection limit of the method used”, not “zero”. The threshold depends on reagent sensitivity, designated λ and expressed in EU/mL, together with the dilution and concentration of the tested sample.
  5. Recovery control. This confirms that the composition of the tested sample did not interfere with the assay.
  6. Test date and testing laboratory. A document without a date and an identified entity is a declaration, not a test result.

The numerical limits used in pharmacopoeias apply to medicinal products. They are calculated using the K/M formula, where K is the permitted endotoxin exposure, 5 EU/kg of body weight per hour for parenteral administration and 0.2 EU/kg for intrathecal administration, and M is the maximum dose of the product. Research Use Only materials are not medicines and are not subject to these requirements. The limits nevertheless remain a useful reference when assessing whether a batch of reagent is suitable for a planned research system. The researcher must choose the acceptance threshold according to the protocol methodology.

More testing means more data available for review

At Pro-Body, we are expanding the quality documentation made available with research materials. Alongside HPLC results, we publish endotoxin test reports tied to a specific batch number. Documentation is being added to successive product pages, starting with:

Further catalogue entries are being supplemented with additional laboratory documentation. The principle is simple: a quality claim without access to a result for the specific batch remains only a claim. A certificate of analysis assigned to a batch number, together with an endotoxin test report, allows the material to be evaluated independently rather than solely on the seller’s assurance.

Frequently asked questions

Does high HPLC purity mean that a material is free of endotoxins?

No. HPLC measures the percentage of the main compound and the profile of synthesis-related substances. Endotoxins do not produce a peak comparable with the peptide peak, and their concentrations are below the UV-detection threshold. A limit of approximately 1 EU/mg corresponds to about 0.00001% of the material by mass, whereas UV chromatography detects impurities from approximately 0.02–0.05% of total peak area. A material with 99% purity can still carry a significant endotoxin burden. These are two independent parameters that require separate tests.

Does filtration through a 0.22 µm membrane remove endotoxins?

No. A membrane with this pore diameter retains bacterial cells but not dissolved lipopolysaccharides. An LPS monomer has a mass of approximately a dozen kDa, and even its micellar aggregates, at approximately 1,000 kDa, remain orders of magnitude smaller than the filter pores. Sterilising filtration addresses the presence of microorganisms, not pyrogens. Removing endotoxins requires other operations, such as ultrafiltration or adsorption.

Is sterility the same as the absence of endotoxins?

No. These are the two concepts most often confused. A sterility test confirms the absence of growth by living microorganisms. Endotoxin is a fragment of the cell envelope of bacteria that are already dead. It remains in a sample after their breakdown and withstands conditions that kill the bacteria themselves. A sterile sample can therefore be strongly pyrogenic.

What exactly does a result of “<0.5 EU/mg” mean?

It means that the endotoxin concentration in the tested sample was below the detection limit of the method used, in this case 0.5 endotoxin units per milligram of material. The threshold depends on the sensitivity of the reagent and the sample dilution. The result does not mean that endotoxins are entirely absent, which is why assay sensitivity is an important part of the report.

Can a chemically synthesised peptide contain endotoxins?

Yes. SPPS itself is a chemical process and does not generate endotoxins, but contamination can arise during purification, lyophilisation and filling through process water, buffers, contact surfaces or the environment. Materials produced through recombinant expression in E. coli are a separate case in which the risk follows directly from the production technology.

Why is this especially important in cell-based research?

LPS is a potent TLR4 ligand that activates the NF-κB pathway and triggers the release of pro-inflammatory cytokines. It is a standard positive stimulation control in inflammation research, and primary cells can respond to concentrations of only a few picograms per millilitre. Endotoxin in a peptide preparation therefore generates a signal that does not originate from the molecule under study, and the observed effect may actually be a response to contamination.

Summary

Peptide-material quality depends on several independent parameters that cannot replace one another. HPLC answers questions about purity and the related-substance profile, mass spectrometry confirms identity, and an endotoxin test, whether gel-clot, turbidimetric or chromogenic LAL, or recombinant rFC, measures a pyrogenic burden that chromatography cannot detect. Endotoxins are lipopolysaccharides from the envelope of Gram-negative bacteria. They are heat-stable, pass through sterilising filters and remain biologically active at trace concentrations. In laboratory research, their control matters beyond the certificate itself: LPS activates the TLR4/NF-κB pathway and can produce a cell-culture response that is easily mistaken for an effect of the peptide under investigation. A reliable report states the batch number, method, result with a unit such as EU/mg or EU/mL, assay sensitivity and recovery control. Publishing this documentation alongside HPLC results allows a specific batch to be assessed from data rather than from a declaration.