Definition of a Security Value Determined by Limulus Amebocyte Lysate Assay Targeting the Recombinant Human Epidermal Growth Factor

The limulus amebocyte lysate (LAL) assay is the most widely used test, required by FDA, in quality control for all parenteral drugs in their last-stage manufacturing process previous to the final formulation. However, the rabbit pyrogen test (RPT) is still required in some United States Pharmacopeia (USP) monographs to allow human and animal vaccine batch release. Hence, the correlation between both endotoxin detection tests targeting recombinant human epidermal growth factor (rhEGF) as active molecule was assessed. From the end point chromogenic LAL test initial validation, a correlation coefficient of at least 0.980 in the endotoxin concentration range from 0.5 to 0.06 endotoxin units (EU)/mL was obtained. In the RPT, the specificity of the animal febrile response was previously evaluated. The authors also demonstrated that the LAL enzymatic cascade activation and the induction of the febrile mechanism in rabbits are due to specific bacterial endotoxin contaminants and not linked with rhEGF protein. Finally, a value of 5 EU per 20 µg rhEGF, measured by end point chromogenic LAL, was defined as the security value to be used as the RPT limit.Introduction
Parenteral preparations have to be pyrogen-free because administration of pyrogens may induce fever, shock, or even death. These preparations must therefore be tested for pyrogenic (i.e., fever-inducing) contamination. There are two pharmacopoeial assays to detect pyrogenic contamination: the rabbit pyrogen test (RPT) and the limulus amebocyte lysate (LAL) test. RPT is done in an in-vivo model developing an immune reaction similar to human beings when pyrogens are present. However, this assay is less sensitive than the LAL test and the outcome depends on the rabbit strain, its physiological and housing conditions, and is not suitable for all product categories, besides the use in animals (1, 2).

On the other hand, the LAL test is an in-vitroassay for bacterial endotoxin quantitation. The chromogenic analytical alternative is based on the development of color after cleavage of a synthetic peptide-chromogen
complex in the presence of endotoxin. Significantly, this assay can be run in a 96-well format allowing a higher throughput of samples, although it is a requirement by the United States Pharmacopeia (USP) for vaccine samples.

Current trends prefer the use of such quality control assays, whereby live animal handling and costly procedures can be avoided. The consequence of such trends is the increasing use of the chromogenic LAL test as an alternative to the RPT. However, the factors influencing the correlation between both methods for each API have to be established. In this work, the initial validation of LAL test, the rabbit colony endotoxin specificity, and the correlation between both endotoxin detection tests were evaluated. Finally, the definition of a security value by endpoint chromogenic LAL test for RPT was performed.

MaterialsBiologicals. Human recombinant epidermal growth factor (rhEGF), expressed in Saccharomyces cerevisiae, was obtained from Heber Biotec S.A. (Havana, Cuba). This product is > 95% pure as assessed by high-performance liquid chromatography (HPLC).

The spike experiments with endotoxin to assess the correlation between the RPT and the LAL chromogenic assay were performed using a USP endotoxin standard (10,000 USP endotoxin units per vial, endotoxin LOT G3E069). Endotoxin was reconstituted, diluted, and the concentrate was stored according to USP instructions. The LAL reagent water, which was pyrogen free, was used to dilute the endotoxin. The E. coli O111:B4 endotoxin standard was used in the rabbit colony specificity assay.

Labware. Borosilicate glass vials were rendered endotoxin-free by a validated dry heat cycle, and pipets tips that were free of interfering endotoxins were used to prepare endotoxin and product dilutions. All glassware, needles, and syringes were pyrogen-free.

MethodsLAL test. The endpoint-chromogenic LAL test measures the chromosphere released from a suitable chromogenic peptide by the reaction of endotoxins with the LAL reagent. This method is based on the quantitative
relationship between the concentration of endotoxins and the release of chromophore at the end of an incubation period. The assay is performed as detailed in the current USP and the National Formulary (NF) (3).

Pyrogen test. New Zeland white rabbits, weighing between 1.5 to 2.5 kg, were used. The animals were housed in individual cages in an animal room at 20 ºC and had free access to food and water. The animals were fasted for 24 h before the experiments. The rabbit pyrogen test was performed at the National Center of Bio-Pharmaceutical Preparation (NCB, Mayabeque, Cuba) according to USP standards (4). The results are expressed as the difference between the initial and highest temperatures recorded within 3 h of the injection.

Rabbit colony specificity assay. The rabbits were tested using the specificity assay. This test was performed using a reference standard endotoxin (RSE) from E. coli strain 0111:B4 with a potency of 22 EU/mL (0.1 ng/mL) at 1 ng/kg of body weight (kg bw). A comparison between two groups of rabbits was studied; one group received 1 ng/kg bw of the RSE and the second group received pyrogen-free saline solution. The rest of the test was performed as detailed in USP35-NF30 (4). The rabbit colony is suitable to be used in the pyrogen test if the animals receiving the pyrogen-free saline solution do not experience an increase in initial temperature and if those receiving 1 ng/kg bw of lipopolysaccharide (LPS) display a temperature increase (∆T) more than 0.5 °C.

rhEGF sample characterization. The rhEGF was separated in a discontinuous sodium dodecyl sulfate-polyacrylamide gel electrophoresis according to methods reported by Schägger and von Jagow in 1987 (5). The specific composition was as follows: 16.5%T-3% separating gel, spacer gel 10%T-3%C and stacking gel 4%T-3%C. Ten µg rhEGF was loaded into each lane and detection of resolved protein bands was made by Coomassie Blue staining (Colloidal Coomassie, Invitrogen, Paisley, UK) (6).

HPLC was performed using a Merck-Hitachi system and a reverse-phase C18 column (Vydac, Hesperia, CA). The samples were injected into the column and the analysis was performed using a water-acetonitrile (AcN)-trifluoroacetic acid (TFA) buffer system (7) and a linear gradient of 0-40% AcN from 0-60 min.

Statistics. A Tukey’s post-hoc test for multiple comparisons was performed after ordinary one-way analysis of variance (ANOVA) statistical test. The statistical analysis was performed using the GraphPad Prism software (Prism 5 for Windows, v. 5.00).

Results and DiscussionLAL and pyrogen tests performance
The endpoint-chromogenic LAL and rabbit pyrogen test, used in this work, were performed according to USP35-NF30 specifications (3, 4). Using these methods does not require validation of accuracy and reliability but users had to verify their suitability under actual conditions of use (8). For LAL test, an initial validation assessing the linearity and specificity parameters was performed according to USP (3, 9). The main results are summarized in Table I. The results obtained are in agreement with USP specifications (3) for the end point chromogenic LAL test. The absolute value of the correlation coefficient is greater than or equal to 0.980 in the endotoxin range of 0.5 to 0.06 EU/mL. Endotoxin recovery values, obtained in all validation runs, ranged from 85.9 up to 103%. Thus, a limit range from 80% to 125% for the specificity requirement was herein established.

Table I: Summary of chromogenic limulus amebocyte lysate (LAL) test validation results.

Validation test

Correlation coefficient

Linearity (Endotoxin standard curve ranging from 0.5 to 0.06 EU/ mL)           

Assay 1

0.9915

Assay 2

0.9935

Assay 3

0.9931

Assay 4

0.9944

Assay 5

0.9917

Assay 6

0.9931

Validation test

Control endotoxin recovery (%)

Specificity (inhibition/enhancement)

Assay 1

99,3

Assay 2

96,6

Assay 3

86,4

Assay 4

94,3

Assay 5

103

Assay 6

85,9

This range allows more stringency than the USP reported limit values (50% to 200%) in releasing sample results.

In the rabbit pyrogen test, the animal colony endotoxin specificity was preliminarily assessed using a reference standard endotoxin (RSE) as previously described. The results are shown in Table II. All the rabbits inoculated with the RSE (group 1) showed a temperature rise from 0.5 °C to 1.2 °C. On the other hand, no temperature rise was detected in rabbits used as control group (group 2), which received pyrogen-free saline solution. These results indicate the specificity of the animal febrile response to endotoxin, suggesting its physiologic adequacy to be used in the pyrogen test.

Table II: Result of the rabbit colony sensitivity assay.

Group

Initial temperature
(ºC)

Final temperature
(ºC)

Temperature difference
(ºC)

1^a

39.6

40.7

1.1

1^a

39.3

39.8

0.5

1^a

39.2

39.9

0.7

1^a

38.9

40.1

1.2

1^a

39.4

40.3

0.9

1^a

39.5

40.4

0.9

1^a

39.3

40.1

0.8

1^a

39.2

39.7

0.5

2^b

39.4

39.3

-0.1

2^b

39.3

39.2

-0.1

2^b

39.6

39.6

0.0

2^b

39.5

39.5

0.0

2^b

39.6

39.5

-0.1

^a rabbit that received 1ng LPS/kg body weight. LPS is lipopolysaccharide.

^b rabbit that received pyrogen-free saline solution.

Assessment of rhEGF protein action in the LAL and pyrogen tests
The ability of analytical assays to unequivocally assess a molecule of interest should be evaluated prior to routine testing. Equally important is the effect of deactivation of the API, which is essential for the specificity assessment of endotoxin, while studying the febrile rise in rabbits and/or LAL enzymatic cascade activation.

For this purpose, two purified rhEGF samples were selected. The first one (rhEGF-1) was purified without full assurance of the aseptic environment. When the pyrogen test for this sample was initially performed, the rabbit temperature rise of more than the USP regulated limit values was induced. According to the USP criteria, “…if not more than three of the eight rabbits show individual rises in temperature of 0.5 °C or more and if the sum of the eight individual maximum temperature rises does not exceed 3.3°C, the material under examination meets the requirements for the absence of pyrogens” (4). The second sample (rhEGF-2) was purified under GMP standards. In this case, when the pyrogen test was initially performed, there was no increase over that USP regulated limits in the rabbits’ temperature, indicating that rhEGF-2 was pyrogen free.

Both API samples, rhEGF-1 (pyrogenic sample) and rhEGF-2 (pyrogen-free sample), were fully heat degraded at 95 °C for 12 h. The efficiency of this degradation procedure was verified by Tricine sodium dodecyl sulfate
polyacrylamide gel electrophoresis (SDS-PAGE) (5) and RP-HPLC (Figures 1 A and B, respectively). The results showed that rhEGF was fully degraded in both heated samples. All these samples, either intact or heat-degraded, were analyzed by LAL and rabbit pyrogen tests. The results are shown in Table III. The LAL values of rhEGF-1 intact and degraded samples were more than 1000 EU/mL and, consequently, rejected in RPT. In contrast, when both rhEGF-2 intact and degraded samples were tested, no detectable LAL reagent activation signals were obtained and both passed the pyrogen test.

Table III: Chromogenic limulus amebocyte lysate (LAL) and rabbit pyrogen test results of recombinant human epidermal growth factor (rhEGF) samples; rhEGF-1 = pyrogenic sample; rhEGF-2 = pyrogen-free sample.

Sample

LAL test (EU/mL)

Rabbit pyrogen test*

rhEGF-1

1005

fail

rhEGF-1 heat degraded

1102

fail

rhEGF-2

<0.012

pass

rhEGF-2 heat degraded

<0.012

pass

*Rabbit pyrogen test according to the regulations described in USP (USP32-NF28 Chapters 85 and 151).

The EGF protein sequence integrity determines its biological activity. In this sense, any change in the protein sequence, from losing a few amino acids to full degradation, has a marked effect on its total biological activity,
which could be partially or completely inactivated (10).

The results suggest that the LAL enzymatic cascade activation and the induction of the febrile mechanism in rabbits are specific for bacterial endotoxin contaminants, and also independent of the rhEGF API.

LAL and RPT after reference standard endotoxin spiking
To correlate the rabbit pyrogen test with the LAL chromogenic assay, 20 µg rhEGF-2 samples were spiked with several USP standard endotoxin (reference standard endotoxin; RSE) concentrations. The RSE concentrations ranged from 2.5 to 20 EU/mL. Two controls were used in the experiment, one containing 20 µg rhEGF-2 without endotoxin and the other of 20 EU/mL RSE without rhEGF. The RSE
recovery results obtained by LAL chromogenic assay are shown in Table IV. No specific enzymatic cascade activation was detected in the first sample without RSE. The obtained RSE recovery values ranged from 95 to 113%, which agree with those expected for each challenged sample. Thus, there is no evidence of rhEGF interference (inhibition/enhancement) with the LAL reagent.

Table IV: Reference standard endotoxin (RSE) recovery from spiked recombinant human epidermal growth factor pyrogen-free samples (rhEGF-2).

Sample

rhEGF:RSE spike ratio (µg : EU)^a

Recovered
RSE (EU/mL)

Recovery
values (%)

1

20 : 0

nd

-

2

20 : 2.5

2.4

96.00

3

20 : 5.0

5.25

105.50

4

20 : 10

11.21

112.10

5

20 : 20

19.11

95.58

6

0 : 20

20.77

103.85

^a mixtures were prepared in 1 mL volume and injected by kg of rabbit weight. nd = not detected.

The pyrogenic activity of those samples was assessed in vivo according to USP (4). The temperature rise of each rabbit was measured and the results are shown in Table V. In the first group, several rabbits that received 10 and 20 EU/mL RSE showed a temperature rise (ΔT) greater than 0.5º C. The sum of ΔT from each of these rabbits (4-6R1) was over 3.3 ºC, the limit value according to USP specifications (4). Thus, these samples were rejected in the first RPT.

Table V: Rabbit pyrogen test results of recombinant human epidermal growth factor (rhEGF) spiked with reference standard endotoxin (RSE).

Sample

rhEGF:RSE spike ratio

Temperature raise in rabbits (°C)

Total sum

 (µg : EU) ^c

I

II

III

IV

V

VI

VII

VIII

(°C)

Group 1

1R1^a

20:0.0

0.1

0.0

0.2

nu

nu

nu

nu

nu

0.3

2R1^a

20:2.5

0.3

0.4

0.4

nu

nu

nu

nu

nu

1.1

3R1^a

20:5.0

0.1

0.4

0.0

nu

nu

nu

nu

nu

0.5

4R1^a

20:10

1.0

0.5

0.1

0.4

0.4

0.8

1.1

0.1

4.4*

5R1^a

20:20

0.7

0.2

0.0

0.4

1.0

0.4

0.5

0.2

3.4*

6R1^a

0:20

0.2

1.1

0.0

1.0

0.6

0.4

0.8

0.5

4.6*

Group 2

1R2^b

20:0.0

0.6

0.2

0.3

0.4

0.1

0.2

0.3

0.2

2.3

2R2^b

20:2.5

0.2

0.5

0.0

0.3

0.4

0.2

0.1

0.4

2.1

3R2^b

20:5.0

0.1

0.3

0.6

0.6

0.2

0.5

0.5

0.3

3.1*

4R2^b

20:10

0.8

0.4

0.3

0.8

0.5

1.0

0.6

0.5

4.9*

5R2^b

20:20

0.7

0.9

0.9

1.1

0.7

1.1

0.7

0.9

7.0*

6R2^b

0:20

0.8

1.2

0.9

0.6

0.6

1.0

0.5

0.7

6.3*

^a samples of the first replicates.
^b samples of the second replicates.
^c sample ratio/kg of rabbit weight.
nu = rabbit not used according to USP specifications (USP35-NF30 chapter <151>).
* = rejected sample.

When the second group of animals was analyzed, the rhEGF samples spiked with 10 and 20 UE/mL were also rejected. These results agree with the first experiment replicates. In the 3R2 case, the fact that four rabbits showed a temperatures rise greater than or equal to 0.5 ºC was the criteria used to reject this sample, although the sum of ΔT for this group was 3.1 ºC. The USP specifications indicate that more than three out of eight rabbits employed for this kind of assay with a temperature rise over 0.5 ºC is also used as a reject criteria. This result does not coincide with those in the first experiment replicate (samples 3R1 vs 3R2).

The influence of physiological factors (independent of the pyrogenic substances action of sample tested) could alter the results in the RPT. The rabbit thermoregulation mechanism is labile and produces false positive results (11, 12). Hence, the 3R2 could be rejected due to the rabbit physiological factors and not because of the sample endotoxin content.

Consequently, these results suggest that any sample with at least 10 endotoxin units (measured by the LAL test) per 20 µg rhEGF will produce an increase in temperature on the rabbit pyrogen test over the USP specified limit.

Correlation between LAL and RPT tests
Despite the shortcomings of RPT, it is still recommended by pharmacopeias for the quality control of final parenteral preparations (13). Nevertheless, a worldwide trend is to substitute with equivalent tests to reduce animal handling (14). The LAL tests are accepted by pharmacopeias as in-vitro assays for detection and quantification of bacterial endotoxins (15). Therefore, it provides as an analytical alternative to take into account as a parenteral biopharmaceutical quality control tool.

With this purpose, it is necessary to find a correlation between both methods. The definition of a security value is also a key point. In that sense, a statistical analysis of data shown in Table V was performed. All the obtained values from rhEGF:RSE spike samples from both groups were taken into account. The Tukey’s post-hoc test for multiple comparisons after ordinary one-way ANOVA statistical test was performed. The results are shown in Figure 2.

No statistically significant difference were observed when 20:2.5 and 20:5 rhEGF:RSE ratios were compared with the endotoxin-free control sample. The corresponding p values for both comparisons were 0.991 and 0.971, respectively. Nevertheless, when samples spiked with 10 and 20 RSE units were compared with the same reference control, statistically significant differences were detected, with p values of 0.023 and 0.003, respectively. The results show that samples with either 10 or more endotoxin units measured by LAL will produce a febrile response in the rabbit pyrogen test. Therefore, measuring 5 RSE units by LAL per 20 µg rhEGF, which is the immediate lowest value tested, is defined as the security value in the rabbit pyrogen test, as previously stated. Altogether, a rhEGF sample with an endotoxin value measured by LAL of 5 EU or higher should be considered as pyrogenic.

On the other hand, no statistically significant differences (p = 1.000) were observed when samples containing 20 RSE units with/without 20 µg rhEGF were analyzed. This result suggests, taking into consideration the number of animals used in the test, that the pyrogenicity is rhEGF independent and RSE specific and that they are in agreement with those obtained with heat-degraded rhEGF samples. The results obtained herein are in agreement with those previously reported. USP defines 5 EU/kg body weight as the threshold pyrogenic dose for individual preclinical species (16).

Similar studies with other proteins have been previously published (17). Park et al. reported a marginal response corresponding to a rise in temperature of 0.5 ºC in each rabbit when both Hepatitis B antigen (HB) protein and HB vaccine were spiked with 5 EU/mL of LPS (17). Additionally, a batch of influenza haemagglutinin vaccine having 100 EU/ml, as demonstrated by LAL activity, showed no pyrogenicity in RPT (18). The sensitivity of the pyrogen test should, therefore, be assessed for each APl and established according to the test conditions.

Assessment of the proposed security value in the pyrogen test using several purified rhEGF batches
The evaluation of the proposed security value, 5 endotoxin units per 20 µg rhEGF, was done using 28 purified rhEGF batches. The results are shown in Table VI. All rhEGF samples with a value lower than 5 EU/20ug of rhEGF, detected by LAL, were approved by RPT. On the contrary, samples with LAL values over this security value were rejected by RPT. In summary, the 5 EU/20 µg rhEGF security value can be used as the upper limit in the endpoint chromogenic LAL test for passing the rabbit pyrogen test when a rhEGF API is analyzed.

Table VI: Assessment of limulus amebocyte lysate (LAL) security limit of recombinant human epidermal growth factor (rhEGF) samples in the rabbit pyrogen test.

EGF preparations

LAL value (UE/mg of EGF)

Pyrogen test results

15

4.58

Pass

20

5.36

Pass

16

5.59

Pass

1

6.18

Pass

26

6.88

Pass

2

6.88

Pass

28

9.74

Pass

11

13.79

Pass

25

17.27

Pass

3

21.90

Pass

18

23.97

Pass

10

45.03

Pass

4

62.80

Pass

13

91.54

Pass

19

101.06

Pass

5

133.41

Pass

23

145.32

Pass

9

150.87

Pass

27

151.35

Pass

24

152.00

Pass

7

160.06

Pass

22

246.70

Pass

6

258.00

Fail

12

275.52

Fail

8

297.91

Fail

17

1016.48

Fail

14

1023.61

Fail

21

3198.78

Fail

Conclusions
The results in this article suggest that the LAL enzymatic cascade activation and the febrile mechanism induction in rabbits are bacterial endotoxin contaminants specific and rhEGF protein independent. A value of 5 endotoxin units, measured by LAL, per 20 µg rhEGF is defined as the security value in rabbit pyrogen test targeting rhEGF as API. A general strategy to define a security value by LAL test to be used in the rabbit
pyrogen test was proposed using as model protein the recombinant human epidermal growth factor.

Acknowledgements
The authors want to thank Magalys García Blanco for proofreading this manuscript, Maria de los Angeles Denis, from the CIGB, for supplying the protein samples used in this work, and the epidermal growth factor
manufacturing and development departments. The authors also thank Professor Elisabeth Díaz for providing language help. Julio César Sánchez García and Neyda Hernández Caso contributed equally to the manuscript.

References
1. P. Castle, Dev. Biol. Stand. 86, 21-29 (1996).
2. T. Hartung et al., ATLA 29, 99-123 (2001).
3. USP35-NF30 General Chapter <85> “Bacterial Endotoxins Test” 1-12.
4. USP35-NF30 General Chapter <151> “Pyrogen Test” 1-3.
5. H. Schägger and G. von Jagow, Anal Biochem. 166 (2) 368-79 (1987).
6. U.K. Laemmli, Nature 227, 680-685 (1970).
7. L.R. Snyder and J.J. Kirkland, “Bonded-Phase Chromatography” in Introduction to Modern Liquid Chromatography (John Willey and Sons, New York, 1979), pp. 269-322.
8. USP35-NF30 General Chapter <1226> “Verification of Compendial Procedures” 1-3.
9. USP35-NF30 General Chapter <1225> “Validation of Compedial Procedures” 2748-2749.
10. D.P. Calnan et al, Gut 47, 622-627 (2000).
11. L. Silveira Rosimar et al., Brazilian J Microbiol. 35, 48-53(2004).
12. K.J. Roberts, “The Pyrogen test” in Endotoxins Pyrogens, LAL Testing and Depyrogenation, K.L. Williams (Eli Lilly & Company Indianapolis, Indiana, U.S.A., Third Edition, New York, 2007), pp. 261-271.
13. S. Poole, “Pyrogen testing of polupeptide and protein drugs” in Polypeptide and Protein Drugs: Production, Characterization and Formulation, R.C. Hider eds. (Ellis Horwood, 1991), pp. 146-153.
14. M. Bernard, et al., Vaccine 20, 2411-2430 (2002).
15. FDA, Guidance for Industry; “Pyrogen and Endotoxins Testing: Questions and Answers” (US Department of Health and Human Services Food and Drug Administration, June 2012 compliance).
16. P. Malyala and M. Singh, J Pharm Sci 97 (6) 2041-2044 (2008).
17. P. C. Yong et al, Biologicals 33, 145-151 (2005).
18. M. Ochia et al, Microbiol Immunol. 46, 527e33 (2002)

About the Authors
Julio Cesar Sánchez García (julio.sanchez@cigb.edu.cu) is head of the Process Control Department; Neyda Hernández Caso is head of the Biochemical Section; Roxana Hernández González is a technician; Alexis Musacchio Lasa, Alexey Llopiz Arzuaga, Hector Santana Millan, and Vivian Pujol García are senior research scientists, all at Center for Genetic Engineering and Biotechnology, Ave 31 e/158 y 190, P.O. Box 6162,C.P. 10 600, La Habana, Cuba; and Virgilio Bourg is a senior researcher at National Center of Bio-Pharmaceutical Preparation, BIOCEN, Mayabeque, Cuba.

Peer-reviewed
Article submitted: July 8, 2013.
Article accepted: August 12, 2013.