- tamsinmarshall4
- 12 hours ago
- 11 min read
Peer Review Article | Open Access | Published 2nd July 2026
Comparative Endotoxin Determination of Naturally Occurring Endotoxins (NOEs) from Pharmaceutical Facility Isolates
Alessandro Pauletto | EJPPS | 3102 (2026) | https://doi.org/10.37521/ejpps31211
Abstract
Endotoxin detection is essential in pharmaceutical manufacturing due to the strong pyrogenic activity of lipopolysaccharides (LPS) from Gram-negative bacteria. While the Limulus Amebocyte Lysate (LAL) assay has traditionally served as the reference method, sustainability concerns have driven the development of recombinant alternatives such as recombinant Factor C (rFC).
This study compared the performance of an rFC assay with five LAL methods (three KCA and two KTA) using Naturally Occurring Endotoxins (NOEs) obtained from seven Gram-negative bacterial isolates cultivated in nutrient rich and nutrient poor media. All media were confirmed endotoxin-free prior to inoculation, and final samples were standardized to approximately 1 EU/mL.
Across isolates and matrices, rFC consistently produced recovery values within the accepted 50–200% range and closely aligned with the LAL mean reference.
Variability among LAL reagents revealed formulation specific differences: the detected endotoxin varied with bacterial species and growth medium, while nutrient conditions did not meaningfully affect endotoxin detectability.
Overall, results show that rFC performs comparably to LAL in detecting autochthonous endotoxins from real-world Gram-negative contaminants, supporting its suitability as a reliable and sustainable alternative to traditional LAL-based methods.
Key words: endotoxin testing, rFC, NOE, naturally occurring endotoxins, autochthonous isolates
Introduction
Lipopolysaccharide (LPS), commonly referred to as endotoxin, is a structurally complex amphipathic molecule that exhibits a strong tendency to self-aggregate in aqueous environments ¹. It represents a fundamental component of the outer membrane of Gram-negative bacteria and is involved in multiple biological activities, notably its pronounced pyrogenic effect ². This capacity to elicit febrile and inflammatory responses underscores the critical need to prevent endotoxin contamination in parenteral pharmaceuticals and medical devices ³.
The Limulus Amebocyte Lysate (LAL) test has long been considered the reference method for endotoxin detection. This reagent is derived from the amebocytes of the horseshoe crab, where it plays a key role in the animal’s innate immune defense. It consists of a series of interconnected enzymatic factors. In the presence of endotoxins, factor C (one of the components of this cascade) initiates a reaction that leads to the formation of a coagulin clot capable of preventing the spread of invading microorganisms ⁴,⁵. Although the extraction of this reagent from Horseshoe Crabs blood does not typically result in immediate death, its use is associated with mortality rates estimated between 15% and 30% ⁶,⁷. The continuous increase in global Bacterial Endotoxin Testing (BET) demand has raised concerns about sustainability of LAL use, prompting the development of alternative solutions based on recombinant reagents. The first recombinant reagent introduced was rFC, which relies exclusively on factor C, the key element responsible for endotoxin recognition and activation of the enzymatic cascade. The cDNA sequence of factor C was cloned from Carcinoscorpius rotundicauda ⁸ and expressed in heterologous systems by Ding et al. in 1997 ⁹⁻¹¹. In the rFC assay, endotoxin activates the recombinant protein, which in turn cleaves a fluorogenic substrate, generating a measurable fluorescent signal ¹²⁻¹⁴. The second recombinant reagent developed was rCR (recombinant Cascade Reagent), a method based on recombinant Factor C, recombinant Factor B, and a recombinant Pro-Clotting Enzyme. In the rCR assay, a reaction cascade occurs in which recombinant Factor C activates recombinant Factor B, which in turn activates the recombinant clotting enzyme. This cascade ultimately cleaves a chromogenic substrate, producing a measurable color change proportional to the endotoxin concentration ¹⁷.No rCR reagents are included in this comparative study.
This biotechnological approach ensures high purity, endotoxin specificity, and minimal lot-to-lot variability ¹⁵,¹⁶. Furthermore, the absence of factor G eliminates β-glucan interference and false positives ¹⁷. Recombinant production enables unlimited, sustainable availability without the use of animals ¹⁸.
Since the introduction of the first recombinant reagent, an intense debate has arisen regarding the actual detection capability of these methods compared to the traditional reference method, the LAL test. Several peer-reviewed studies have demonstrated that rFC is comparable or even superior to LAL, including its lack of sensitivity to β-glucans ¹⁹. More recently, a number of publications have supported similar claims for rCR reagents as well ²⁰. Part of this discussion has focused on the ability of recombinant methods to detect samples that represent real-world contamination scenarios in pharmaceutical products ²¹. To address this point, in the present study we compared the endotoxin detection performance of one rFC reagent against five different LAL reagents: three KCA (Kinetic Chromogenic Assays) and two KTA (Kinetic Turbidimetric Assays). The samples tested were NOEs (Naturally Occurring Endotoxins) derived from seven different Gram-negative bacterial species isolated during routine bioburden monitoring in pharmaceutical facilities. NOEs refer to endotoxins that, unlike CSE or RSE, are not subjected to significant purification steps. Although the procedure for obtaining NOEs is not fully standardized, it generally involves bacterial cultures followed by sterile filtration ²².
Materials and Methods
Reagents:
For this study, five different LAL reagents from two different suppliers were used. The study does not specify which reagent was used in each individual test. Reagents used were both KCA and KTA reagents. Only one reagent for rFC was used
The reagents included:
Charles River Laboratories (Endochrome®, R19000 and Endosafe® PTS Cartdriges)
Lonza (Kinetic-QCLTM and PyrogentTM-5000)
bioMérieux: ENDOZYME II® kit,
LAL and rFC assays were performed according to manufacturers’ instructions and European Pharmacopoeia requirement. Results were all valid. The experiments were performed at an independent 3rd party laboratory. All calibration curves generated with the different methods consistently used CSE, duly qualified against the primary reference standard as stated in the supplier’s certificate. The standard curve used was always from 5 to 0.005 EU/ml with the exception of PTS cartridges (5-0.05 EU/ml) and PyrogentTM-5000 (10 - 0.01 EU/ml)
Media:
Four different growth media have been used:
Salt solution: 7,5 g/L Disodium hydrogen phosphate, 3 g/L Potassium dihydrogen phosphate, 0,5 g/L Sodium chloride, 1 g/L Ammonium chloride, dissolved in water endotoxin-free (pH 7).
M9 medium: 0,5% Glucose, 2 mM Magnesium sulfate, dissolved in the Salt solution
R2A medium: 0,05% yeast extract, 0,05% Proteose Peptone no. 3, 0,05% Casein hydrolysate, 0,05% Starch soluble, 0,03% Sodium pyruvate, 0,03% dipotassium hydrogen phosphate, 0,005% Magnesiumsulfate, dissolved in water endotoxin-free (pH 7,2).
LB medium: 10 g/L Tryptone, 10 g/L Sodium chloride, 5 g/L yeast extract, dissolved in water endotoxin-free.
All the media were tested to be below the LOD by rFC method.
Sample preparation:
Seven different facility isolates were used to produce specific BIOBALLS® (bioMérieux) according to the supplier’s instructions. These BIOBALLS® were then used to inoculate the corresponding test media. The inoculation level was approximately 550 CFU. After 5 days of incubation at 37 °C, the samples were sterilized by filtration through a 0.2 µm filter (Sartorius) to remove all viable bacteria.
The samples obtained were subsequently diluted in water for BET to reach a final endotoxin concentration of approximately 1 EU/mL
The isolates included:
romobacter xylosoxidans
Acinetobacter johnsonii
Methylobacterium fujisawaense
Moraxella osloensis
Pseudomonas stutzeri
Ralstonia insidiosa
Stenotrophomonas maltophilia
It is noted that not all isolates were able to grow in all media.
Results
Control experiments of the pure media:
All culture media used for bacterial growth in this study were previously tested with an rFC reagent (ENDOZYME® II). As shown in Fig.1, the rFC reagent did not reveal any detectable endotoxin contamination in any of the media tested. The media were analyzed by diluting the sample in BET water at a 1:10 dilution for LB and M9, and at 1:100 for R2A and Salt media. No interference was observed, and all spike recoveries fell within the acceptable range 50-200%.
Result analysis
All seven bacterial isolates were analyzed using the different methods described in the Materials and Methods section. Sample dilutions were selected to ensure that endotoxin concentrations fell within the quantifiable range of the standard curve associated with each analytical method.
To meet this requirement, the dilutions applied ranged from 1:1,000 to 1:1,000,000. All dilutions were prepared in water for BET to avoid potential interferences and to maintain full methodological consistency. For standardization purposes, the 100% baseline was defined as the mean value derived from the five LAL assays, facilitating a comparison between methods. Individual results were subsequently benchmarked against this mean and expressed as a percentage recovery relative to the reference. The results presented in Fig. 2 show the values obtained with each individual method, and are expressed, as noted earlier, as percentages relative to the mean value observed with the LAL methods used as the reference.
Methodological note. In all instances, data interpretation considers the commonly applied acceptance range (50–200%) for spike recovery in BET methods, as well as the inherent variability associated with the matrix (medium) and the isolate.
Achromobacter xylosoxidans. For this isolate, all four media were evaluated. The rFC reagent consistently returned values within the 50–200% accuracy range generally expected for BET assays. The only exception was observed with the Salt medium, which showed overquantification, with values slightly above 200%. Overall, the LAL methods produced results within the 50–200% range, with a few exceptions. Notably, the KTA2 reagent yielded a marked underestimation of the endotoxin generated in the LB medium; this behavior was not observed in the other media tested.
Acinetobacter johnsonii. For this isolate, only LB medium was assessed. Here again, KTA2 showed atypical (high) values that remained within the 50–200% range, but close to the upper limit.
Methylobacterium fujisawaense. Testing was conducted only in LB medium. This isolate exhibited a particularly interesting profile: three LAL reagents (KCA1, KTA1, and KTA2) produced values below the reference mean, whereas one LAL reagent returned values well above 300%. These elevated values could be explained by a potential interaction with β-glucans possibly present in the culture medium (not demonstrated), to which this specific reagent is more sensitive. However, the same reagent did not show the same issue in other bacterial strains grown in the same medium. It is worth noting that this bacterial species has also shown pronounced discrepancies between KTA and KCA methods in previous studies ²².
Moraxella osloensis. In this case, no substantial differences were observed across methods. However, one LAL reagent still produced values below the presumed mean; specifically, the KCA3 method in the LB media only. Both LB and R2A were used for this strain.
Pseudomonas stutzeri: the isolate was grown on LB, M9, and R2A media. For the LB medium, none of the BET methods produced values outside the general acceptance range (50–200%), indicating good consistency across analytical reagents. In contrast, for the M9 medium, two LAL reagents (KCA1 and KCA3) yielded values slightly below 50%, suggesting a potential underestimation of endotoxin concentration. The reagent KTA2, however, behaved in the opposite way, producing overestimated values relative to the expected contamination level. In the R2A medium, only the KCA method resulted in values slightly below the acceptable threshold, a pattern that mirrors what was observed in M9. This suggests that the KCA reagent may have a specific underdetection tendency for this bacterial species.
Ralstonia insidiosa: the growth was performed on LB and R2A media. No significant issues were observed with any method, with the sole exception of KTA2, which produced a marked underestimation of endotoxin levels only in R2A medium.
Stenotrophomonas maltophilia: it was assessed in LB and R2A media. No major inconsistencies were detected; the only deviation involved the KCA3 reagent, which showed slightly lower values in LB medium. However, the results remained within the generally accepted range.
Discussion and Conclusions
Stressing microorganisms—and the endotoxin they produce—under both nutrient rich (LB medium) and nutrient poor (R2A medium, M9 medium, and Salt solution-) growth conditions did not influence clearly the detectability or quantification of endotoxin by any of the evaluated methods with some exceptions. It indicates that environmental or nutritional stress does not alter the structural or functional integrity of the endotoxin in a way that affects analytical recovery, reinforcing the robustness of the tested detection systems. This observation applies to strains tested in both types of media.
It is also interesting to note that relevant differences can be observed not only within LAL methods (KTA and KCA), but also among different formulations or brands of reagents belonging to the same methodology. This is likely related to the influence exerted not only by the reagent itself, but also by its specific formulation, which may affect the quantification process. Different suppliers use distinct manufacturing approaches that introduce variability in LAL reagents. These differences may stem from the extraction process used to obtain the reagent or from variations in formulation, such as the composition of salts and buffers. Such factors clearly influence the reactivity of the reagent ²³,²⁴ and, according to the results of this study, may also affect endotoxin quantification in certain cases.
The findings of this study demonstrate that recombinant Factor C (rFC) assays provide a level of performance fully comparable to traditional LAL methods when quantifying autochthonous endotoxin derived from Gram-negative- bacteria naturally occurring in pharmaceutical environments. Across all tested isolates, rFC consistently produced results aligned with those obtained using LAL-based assays, confirming its suitability for endotoxin detection in real-world industrial conditions.
Given that both rFC and LAL assays rely on the same endotoxin sensing enzyme, Factor C, it is not surprising that their analytical behavior remains largely comparable- across conditions and matrices. The collective data supports the conclusion that rFC represents a reliable and consistent method for endotoxin detection, capable of matching LAL performance while offering the additional benefits associated with recombinant technologies.
References
1. Gorman A, Golovanov A. Lipopolysaccharide Structure and the Phenomenon of Low Endotoxin Recovery. European Journal of Pharmaceutics and Biopharmaceutics, Volume 180, November 2022, Pages 289-307
2. Schneier, M.; Razdan, S.; Miller, A.M.; Briceno, M.E.; Barua, S. Current technologies to endotoxin detection and removal for biopharmaceutical purification. Biotechnol. Bioeng. 2020, 117, 2588–2609.
3. Franco, E.; Gracia-Recio, V. Endotoxins from a Pharmacopoeial Point of View. Toxins 2018, 10, 331.
4. Muta, T.; Oda, T.; Iwanaga, S. Horseshoe crab coagulation factor B. A unique serine protease zymogen activated by cleavage of an Ile-Ile bond. J. Biol. Chem. 1993, 268, 21384–21388.
5. Ding, J.L.; Ho, B. Endotoxin detection—From Limulus amebocyte lysate to recombinant factor C. In Endotoxins: Structure, Functionand Recognition; Springer: Dordrecht, The Netherlands, 2010; pp. 187–208.
6. Krisfalusi-Gannon, J.; Ali, W.; Dellinger, K.; Robertson, L.; Brady, T.E.; Goddard, M.K.M.; Tinker-Kulberg, R.; Kepley, C.L.; Dellinger, A.L. The role of horseshoe crabs in the biomedical industry and recent trends impacting species sustainability. Front.Mar. Sci. 2018, 5, 185.
7. Atlantic States Marine Fisheries Commission 2024 Horseshoe Crab Stock Assessment Update
8. Ding, Navas & Ho. Molecular cloning and sequence analysis of factor C cDNA from the Singapore horseshoe crab, Carcinoscorpius rotundicauda Mol Mar Biol Biotechnol 4(1):90–103
9. Ding, J.L.; Chai, C.; Pui, A.W.M.; Ho, B. Expression of full length and deletion homologues of Carcinoscorpius rotundicauda factor C in Saccharomyces cerevisiae: Immunoreactivity and endotoxin binding. J. Endotoxin Res. 1997, 4, 33–43.
10. Pui, A.W.M.; Ho, B.; Ding, J.L. Yeast recombinant factor C from horseshoe crab binds endotoxin and causes bacteriostasis. J. Endotoxin Res. 1997, 4, 391–400.
11. Roopashree, S.D.; Ho, B.; Ding, J.L. Expression of Carcinoscorpius rotundicauda factor C in Pichia pastoris. Mol. Mar. Biol. Biotechnol. 1996, 5, 334–343.
12. Maloney, T.; Phelan, R.; Simmons, N. Saving the horseshoe crab: A synthetic alternative to horseshoe crab blood for endotoxin detection. PLoS Biol. 2018, 16, e2006607.
13. Ding, J.L.; Ho, B. A New era in pyrogen testing. Trends Biotechnol. 2001, 19, 277–281. [CrossRef] [PubMed]
14. Loverock, B.; Simon, B.; Burgenson, A.; Baines, A. A recombinant factor C procedure for the detection of gram-negative bacterial endotoxin. Pharmacopeial Forum 2010, 36, 321–329.
15. Abate,W.; Sattar, A.A.; Liu, J.; Conway, M.E.; Jackson, S.K. Evaluation of recombinant factor C assay for the detection of divergent
lipopolysaccharide structural species and comparison with Limulus amebocyte lysate-based assays and a human monocyte activity assay. J. Med. Microbiol. 2017, 66, 888–897.
16. Bolden, J.; Smith, K. Application of recombinant factor C reagent for the detection of bacterial endotoxins in pharmaceutical products. PDA J. Pharm. Sci. Technol. 2017, 71, 405–412.
17. Muroi, M.; Ogura, N.; Mizumura, H.; Aketagawa, J.; Oda, T.; Tanamoto, K.I. Application of a recombinant three-factor chromogenic reagent, PyroSmart, for bacterial endotoxins test filed in the pharmacopeias. Biol. Pharm. Bull. 2019, 42, 2024–2037.
18. Piehler, M.; Roeder, R.; Blessing, S.; Reich, J. Comparison of LAL and rFC assays-Participation in a proficiency test program between 2014 and 2019. Microorganisms 2020, 8, 418.
19. Jay Bolden, Chris Knutsen, Jack Levin, Catherine Milne, Tina Morris, Ned Mozier, Ingo Spreitzer, and Friedrich Von Wintzingerode, Currently Available Recombinant Alternatives to Horseshoe Crab Blood Lysates: Are They Comparable for the Detection of Environmental Bacterial Endotoxins? A Review. PDA J Pharm Sci and Tech 2020, 74 602-611
20. Madeline Kelley, Ingrid Stevens, Toshio Oda, Junko Akiyoshi, Edwin G. Jahngen, A Demonstration of the Validation Process for Alternative Endotoxin Testing Methods Using PyroSmart NextGen® Recombinant Cascade Reagent. BPB Reports Volume 6 (2023) Issue 2
21. Yutaka Kikuchi, et al., Collaborative Study on the Bacterial Endotoxins Test Using Recombinant Factor C-based Procedure for Detection of Lipopolysaccharides. Pharmaceutical and Medical Device Regulatory Science, 48(4), 252-260 (2017)
22. Bowers K, Tran L. Creation of an in-house naturally occurring endotoxin preparation for use in endotoxin spiking studies and LAL sample hold time analysis. Am Pharmaceut Rev. 2011;14(6):92–97.
23. Joël Gysi, Comparison of different recombinant assays. USP Virtual WWorkshop May 14 (2024)
24. P. F. Roslansky and T. J. Novitsky, Sensitivity of Limulus Amebocyte Lysate (LAL) to LAL-Reactive Glucans. Journal of Clinical Microbiology, Nov. 1991, p. 2477-2483
25. R E.Wachtel' and K.Tsuji, Comparison of Limulus Amebocyte Lysates and Correlation with the United States Pharmacopeial Pyrogen Test. Applied and Environmental Microbiology, June 1977, p. 1265-1269 Vol. 33, No. 6
Author
Corresponding Author: Alessandro Pauletto,
bioMérieux
Comments