MONOCLONAL ANTIBODIES IN IMMUNODIAGNOSIS AND IMMUNOTHERAPY Volume 32, Number 4, 2013 ª Mary Ann Liebert, Inc. DOI: 10.1089/mab.2013.0013
Detection of the Single-Chain Precursor in the Production and Purification Process of Recombinant Human Insulin Chunsheng Leng,1,2 Qingwei Li,1,3 Fenfang Wu,1,3 Liyong Chen,1,3 and Peng Su1,3
High quality recombinant insulin requires being free of single-chain precursor (proinsulin), a task that depends on the selectivity and sensitivity of the monitoring process for detecting proinsulin. In this study we developed an enzyme-linked immunosorbent assay (ELISA) system that was specifically tailored to detect recombinant proinsulin. The proinsulin consists of six components: an initiating methionine, 48 amino acids from human growth hormones (HGH, used as the protection peptide), first connecting Arg-residue, B-chain of insulin, and second connecting Arg-peptide and A-chain of insulin. This form of proinsulin is more stable and can be efficiently expressed by E. coli than insulin. Herein, we evaluated the specificity, precision, recovery, sensitivity, and detection range of the proinsulin ELISA kit. The results showed that the ELISA kit is a very useful tool for monitoring the proinsulin yield in early stages of insulin production as well as the residual proinsulin in the final product, insulin.
iological impurities can be divided into two parts: (1) process-related impurities, such as cell matrix (host cell protein, host cell DNA) and cell culture medium (inducing agent, antibiotics, or medium components); and (2) productrelated impurities, such as precursors and degradation products.(1,2) Recombinant human insulin or insulin analogs are usually produced by recombinant DNA technology.(3–7) However, as only a small amount of insulin can be expressed in Escherichia coli and the expressed insulin is easily degraded to a smaller product by host enzymes, direct expression is not an efficient process to produce insulin. Recombinant proinsulin is a fusion protein containing protection peptide and insulin. The increase in size of proinsulin as a result of its fusion with protection peptide leads to improved protein stability and expression efficiency.(8) Proinsulin is also called insulin singlechain precursor, which is part of the impurity associated with the final product. The insulin is synthesized in the form of a peptide chain and then assembled into double-chain in the human body. The N-terminal of the A-chain and the C-terminal of the B-chain of insulin can be joined by the connecting peptide, called the C-peptide.(9) The biological activity of the intact natural proinsulin is only 3% to 5% of insulin, while the C-peptide is inactive. In clinical trials, the presence of natural proinsulin in the insulin preparation was found to have a negative effect on the heart, leading to an increased incidence of myocardial infarction.(10–14) More importantly, recombinant proinsulin is a
heterologous protein. Therefore, the removal of the recombinant proinsulin in the final product of insulin is vital, and the effectiveness of its removal should be analyzed by appropriate methods. The proinsulin present in serum or plasma is generally measured by a sensitive two-site sandwich ELISA (commercialization kit). The assay is based on two monoclonal antibodies, an anti-C-peptide antibody bound to a microtiter plate and a biotin-labeled anti-insulin antibody. However, the recombinant proinsulin prepared in vitro by our laboratory is not detected by a commercially available ELISA kit because the proinsulin only contains protection peptide, insulin, and connected-Arg-peptide, but no connected-C-peptide. In addition, the protection peptide has a negative effect on humans and is also a source of impurity. So, proinsulin specific to the production process must be detectable and removed from the final product, insulin. In this study, we established an ELISA method for the detection of the residual proinsulin during the multi-step purification process of insulin production and evaluated the characteristics of the ELISA kit. Materials and Methods Purification of proinsulin protein The insulin precursor (proinsulin) contained an initiating methionine, 48 amino acids from HGH, first connecting peptide (containing Arg-residues), B-chain of insulin, followed by a second connecting Arg-peptide and A-chain of insulin (Supplementary Fig. 1). The 55 amino acids before the B-chain was added and used as the protection peptide. This
College of Life Science; 3Institute of Marine Genomics & Proteomics, Liaoning Normal University, Dalian, China. Tonghua Dongbao Pharmaceutical Co., Tonghua, China.
256 proinsulin was expressed in E. coli in the form of inclusion bodies. The E. coli cells were lysed with a high pressure homogenizer and the cell extract was centrifuged at 10,000 g for 30 min at 4C to collect the inclusion bodies. The inclusion bodies were washed three times with a buffer containing 50 mM Tris-HCl (pH 8.0), 2 mM EDTA, 100 mM NaCl, and 1% Triton X-100 and then dissolved in 8 M urea containing 4 mM mercaptoethanol (pH 10.4). The concentration of the protein was adjusted to 20–30 mg/mL and the sample was centrifuged at 10,000 g for 30 min at 4C to remove the insoluble material. The supernatant was diluted with 2 M urea. Then HGH-proinsulin fusion protein was refolded for 24 h at 4C in the refolding buffer containing 50 mM Tris-HCl (pH 9.3), 100 mM NaCl, 3.2 M urea, 1% glycine, 5% glycerol, 0.2% PEG, 1 mM oxidized glutathione, and 1 mM reduced glutathione. The correctly refolded and soluble protein was collected by centrifugation at 10,000 g for 30 min at 4C and then concentrated by ultrafiltration using a 10 kDa Ultracel membrane (Millipore, Billerica, MA). The presence of urea in the preparation was finally removed by exchanging the buffer into phosphate buffer (pH 3.5). Preparation of anti-55-peptide monoclonal antibodies
LENG ET AL. recombinant insulin (purchased from Novo Nordisk, Bagsvaerd, Denmark) coupled CNBr-activated Sepharose. The titer of the anti-insulin pAbs was assayed by indirect ELISA. Indirect ELISA ELISA plate (96-well, Corning, NY) was coated with carbonate/bicarbonate buffer (containing 15 mM Na2CO3, 35 mM NaHCO3, and 0.2 M NaCl [pH 9.6]) and 20 mg of antigens. The plate was blocked with a blocking solution containing 0.5% BSA. The affinity-purified anti-55-peptide or anti-insulin antibody was diluted 2000–128,000 fold and then added to the corresponding plate. The plate was washed three times with PBS-T (PBS buffer containing 0.05% Tween-20) buffer, followed by the addition of hydrogen peroxide (HRP)labeled secondary antibody ( Jackson ImmunoResearch Laboratories, Baltimore, MD) in 1:10,000 dilution. After washing three times with PBS-T, 100 mL tetramethylbenzidine were added to the plate and the color development was allowed to proceed for 15 min before it was terminated by the addition of 2 M sulfuric acid. The absorbance of the plate was read at 450/ 620nm using a microtiter plate reader (Molecular Devices, Sunnyvale, CA).
To generate anti-55-peptide monoclonal antibodies (Manti-55-peptide MAbs), BALB/c mice were immunized with the synthesized 55-peptide (MFPTIPLSRLFDNAMLRAH RLHQLAFDTYQEFEEAYIPKEQKYSFLQNPLGTGPR). The peptide contained an initiating methionine, followed by 48 amino acids of the HGH and the first connecting peptide of proinsulin (used as the protection peptide). A brief description is as follows. Three adult female BALB/c mice (6 to 8 weeks of age) were injected intraperitoneally with 55-peptide as antigen. Each mouse was first injected with 100 mg antigen in 250 mL phosphate-buffered saline (PBS, pH 7.2) mixed with an equal volume of complete Freund’s adjuvant (SigmaAldrich, St. Louis, MO) and then boosted with 50 mg antigen in 250 mL PBS mixed with an equal volume of incomplete Freund’s adjuvant at 2-week intervals. Five to 7 days after the third immunization, blood from the tail vein was collected and the antibody’s titer was measured. Before fusion, the mice were boosted again with 100 mg antigen in PBS. After four immunizations, spleen cells of the immunized mice were fused with SP2/0 myeloma cells and the hybridoma cells were selected by growth in HAT (hypoxantin, aminopterin, and thymidin) medium. Monoclonal cell lines were screened by the limiting dilution method. The antibodies secreted by the different clones were then assayed using ELISA. The subclass of MAbs was determined using mouse hybridoma subtyping kit (Boehringer Mannheim, Germany). To obtain a large amount of antibodies, the hybridomas were injected into the peritoneal cavity of mice, and antibody-rich ascites were collected. The antibodies in the ascites were purified by protein-G Sepharose (GE Healthcare, Pittsburgh, PA). The titer of anti-55-peptide MAb was assayed by direct ELISA.
Anti-55-peptide MAb was labeled with NaIO4 and HRP. In brief, 5 mg HRP were dissolved in 1 mL of triple-distilled water and 0.2 mL of 0.1 M NaIO4 was then added and the mixture was stirred at room temperature in the dark for 20 min. The solution was placed in a dialysis bag and dialyzed against 1 mM sodium acetate buffer (pH 4.4) at 4C overnight. Then, 20 mL of 0.2 M sodium carbonate buffer (pH 9.5) were added, and the resulting hydroformylated HRP was immediately added to a tube containing 1 mL of the desired antibody (containing 5 mg anti-55-peptide MAbs in 0.01 M carbonate buffer) and mixed by inverting the tube several times. The sample was then placed on a rotary shaker and shaken for 2 h at room temperature in the dark. Thereafter, 0.1 mL of NaBH4 (4 mg/mL) was added and the sample was mixed by inverting the tube several times, followed by 2 h of shaking on a rotary shaker at 4C. The sample was then transferred into a dialysis bag and dialyzed against 0.02 M PBS (pH 7.4) at 4C overnight, followed by the addition of an equal volume of saturated ammonium sulfate solution and standing at 4C for 1 h, then centrifuged at 9000 g for 10 min at 4C. The precipitate was washed twice with half-saturated ammonium sulfate solution, and dissolved in 1 mL 0.02 M PBS (pH 7.4). The solution was transferred to a new dialysis bag and dialyzed against 0.02 M PB buffer (pH 7.4) to remove the ammonium sulfate, followed by centrifugation at 11,000 g for 30 min to remove the precipitated material. Finally, the supernatant (labeled antibody) was mixed with an equal volume of glycerol and dispensed into small aliquots for storage at - 20C.
Preparation of anti-insulin polyclonal antibodies
Assembly of ELISA kit and detection of proinsulin
New Zealand white rabbits were immunized with the purified recombinant proinsulin fusion protein. Seven days after the last immunization, serum of the immunized animals was collected. The anti-insulin polyclonal antibodies (pAbs) in the antiserum were purified by affinity chromatography using
Microtiter plates (96-well, Corning) were coated with the capture antibody (anti-insulin pAbs) at a concentration of 10 mg/mL in carbonate/bicarbonate buffer (pH 9.6) and incubated overnight at 4C. The coating solution was removed and the remaining protein-binding sites of the plate were
DETECTION OF THE INSULIN SINGLE-CHAIN PRECURSOR
FIG. 1. Purification and identification of anti-insulin pAbs. (A) Purification of anti-insulin pAbs by insulin-coupled CNBractivated Sepharose chromatography. M, protein marker; lane 1, rabbit anti-proinsulin sera; lane 2, flow-through sample from affinity columns; lane 3, balance fluid from affinity columns; lane 4, elution samples containing anti-insulin pAbs. (B) Titer of anti-insulin pAbs by ELISA assay. Irrelated rabbit IgG used as control. Error bars indicate standard error of the mean, n = 3. OD, optical density.
blocked with 1% BSA in PBS overnight at 4C. The blocking solution was removed and the plate was washed three times with PBS. After washing, 100 mL of the appropriately diluted samples were added to each well. For accurate quantitative results, signals of unknown samples were always compared against those of a standard curve. Standards (triplicates) and blanks were carried out simultaneously for each plate to ensure accuracy. Standards were prepared using purified recombinant proinsulin diluted in TBS buffer (containing 0.05% Tween-20 and 1% BSA). The plate was incubated for 90 min at 37C. After incubation, the samples were removed and the plate was washed three times with PBS, and then incubated with the appropriately diluted HRP-labeled anti-55-peptide MAbs for 90 min at room temperature. The plate was washed four times with PBS, followed by the addition of a tetramethylbenzidine solution and a 10– to 15–min incubation before the reaction was terminated by the addition of a stopping solution (2 M H2SO4).
The plate was read at dual filter 450/620nm. A standard curve was prepared from known concentrations of proinsulin, and the concentrations of proinsulin in the unknown samples were determined from the standard curve. Identification of proinsulin and insulin by SDS-PAGE The proinsulin and insulin obtained from different purification steps were analyzed by SDS-PAGE using 12% gel. The proteins in the gel were visualized by staining the gel with Coomassie brilliant blue G-250. Statistical analysis Statistical comparisons were analyzed by Student’s nonpaired t-test or one-factor analysis of variance using SAS software (v 8.02).
FIG. 2. Purification and identification of anti-55-peptide MAb. (A) Purification of M-anti-55-peptide MAbs by protein-G Sepharose chromatography. Lane 1, M-anti-55-peptide ascites; lane 2, flow-through sample from affinity columns; lane 3, balance fluid from affinity columns; lane 4, elution samples containing M-anti-55-peptide MAbs; M, protein markers. (B) Titer of M-anti-55peptide MAbs by ELISA assay. Error bars indicate standard error of the mean, n = 3. Unrelated mouse IgG was used as control.
LENG ET AL. Table 1. Precision Assay of Proinsunlin Elisa Kit CV Sample
Mean value (ng/L)
1 2 3 4 Mean value
921 623 305 107
Total assay %
5.6 3.4 2.8 1.8 3.4
8.9 6.2 12.1 6.8 8.5
6.9 5.1 7.5 4.1 5.9
Hook effect in ELISA
FIG. 3. Detection of insulin, synthetic 55-peptide, and proinsulin by proinsulin ELISA kit. Error bars indicate standard error of the mean, n = 3.
Results Purification and identification of antibodies The anti-insulin pAbs generated using the proinsulin fusion protein as antigen could be efficiently purified by protein G affinity chromatography. The heavy chains and light chains of the affinity-purified anti-insulin pAbs were detected by SDS-PAGE (Fig. 1A). The highest titer for the antiinsulin pAb preparation was 1:64,000, as detected by ELISA (Fig. 1B). The anti-55-peptide MAb (1E10) was identified as IgG1a isotype with kappa type. The anti-55-peptide MAb in the ascites was also purified by protein-G Sepharose affinity chromatography, and both the heavy and light chains of the antibody were detected by SDS-PAGE (Fig. 2A). The highest titer achieved by this antibody preparation was 1:32,000, as detected by ELISA (Fig. 2B).
Hook effect is observed when an ELISA system is overwhelmed with the target antigen, resulting in lower than expected OD readings for samples with lower dilutions. Proinsulins with a concentration of up to 1.6 g/L could be detected by our ELISA kits without giving falsely low results. When the concentration of proinsulins is too high (over 1.6 g/ L), the sample should be appropriately diluted before assay. Sensitivity of ELISA kit The lower limit of detection (LOD) was estimated from the mean of the blank and the standard deviation (SD) of the blank, which was defined as 3 · standard deviation of the blank (X blank + 3 SD). LOD obtained from 30 replicates of the blank was 12 ng/ L. Table 2. Recovery Assay of Proinsunlin Elisa Kit Original proinsulin (ng/L)
Proinsulin antigen added (ng/L)
Observed proinsulin assay (ng/L)
Percent recoverya (%)
30 60 180 360 30 60 180 360 30 60 180 360 30 60 180 360 30 60 180 360
56 80 220 370 96 101 254 495 220 214 331 531 472 402 575 799 702 725 816 995
93 89 105 95 107 84 106 118 105 89 92 98 121 96 106 111 94 93 91 92 99
Specificity of ELISA kit Proinsulin could be detected at concentrations ranging from 15.6 to 1000 ng/L using the proinsulin ELISA kit. However, insulin and synthetic 55-peptide could not be detected at this range of concentrations (Fig. 3), which indicated that insulin and synthetic 55-peptide did not interfere with the test and only proinsulin could be specifically detected by the ELISA kit. Precision of ELISA kit
Each sample was analyzed in three replications on 20 different occasions. Intraassay, interassay, and total assay precisions were analyzed by one-way analysis of variance. The mean values of intra- and inter-CV (coefficient of variation) were 3.4% and 8.5%, respectively. The total assay precision was 5.9% (Table 1).
a Recovery: A study was performed where dilutions of a sample with one known concentration of proinsulin was added to another known concentration sample. The concentration of proinsulin was determined using the proinsulin ELISA assay, and the resulting percent recovery was calculated. Percent recovery (%) = observed proinsulin concentration / (original proinsulin concentration + proinsulin concentration added).
DETECTION OF THE INSULIN SINGLE-CHAIN PRECURSOR
Table 3. Dilution Linearity Assay of Proinsunlin Elisa Kit Sample
Final dilution Obtained Expected Percent factor value (ng/L) value (ng/L) recoverya (%)
Undiluted 1:2 1:4 1:8 1:16 1:32 1:64 2 Undiluted 1:2 1:4 1:8 1:16 1:32 1:64 3 Undiluted 1:2 1:4 1:8 1:16 1:32 1:64 Average recovery
981 470 234 141 74 24 13 960 495 275 135 71 34 14 889 410 254 133 65 29 16
981 491 245 123 61 31 15 960 480 240 120 60 30 15 889 445 222 111 56 28 14
100 96 95 115 121 78 85 100 103 115 113 118 113 93 100 92 114 120 117 104 115 105
a High-concentration proinsulin were diluted with TBS (Tris buffered saline) buffer. The proinsulin concentration was determined for each dilution and the percent (%) recovery was calculated. Percent recovery (%) = proinsulin concentration obtained/proinsulin concentration expected.
FIG. 4. Detection of proinsulin and insulin in various stages of insulin production by SDS-PAGE. M, protein markers; lane 1, lysed cells transformed by proinsulin expression vector; lane 2, inclusion bodies containing proinsulin; lane 3, Arg (B31)-insulin from proinsulin (12,461 Da) cleavaged by trypsin, then purified by cation exchange chromatography and concentrated by 3 kDa ultrafiltration; lane 4, insulin (containing A chain, 2,383 Da and B chain, 3,429 Da) was from Arg (B31)-insulin cleavaged by carboxypeptidase, then purified by C8 reverse phase HPLC column and concentrated by 3 kDa ultrafiltration; lane 5, insulin crystals dissolved in 0.1 M PBS.
L) was from 78% to 121%, and the average recovery was 105% (Table 3).
The lower limit of quantitation (LOQ) was defined as the lowest concentration of an assay that can be measured with an interassay CV of 20%. LOQ of the proinsulin ELISA assay was < 2 ng/mL. The detection limit concentration of the samples was tested over 20 days spanning a period of 6 months, with each test carried out in three replications, and two tests were performed for each day.
Proinsulin was the major product in the earlier three stages of insulin preparation process. However, in the later three stages, proinsulin was converted into insulin and the level of residual proinsulin in the preparation was too low to be detected by SDS-PAGE (Fig. 4).
Recovery of ELISA kit
Detection of proinsulin by ELISA
The recovery of the proinsulin ELISA across the five separate spiked concentrations ranged from 84% to 121%, and the average recovery was 99% (Table 2). The recovery across the three fold-diluted samples (981 ng/L, 960 ng/L, and 889 ng/
In order to calculate the amount of proinsulin contained in our samples, the standard curve for proinsulin was plotted using the logarithmic regression [y = a + bLn(x)]. The logarithmic regression analysis of the curve (Supplementary Fig. 2)
Detection of proinsulin and insulin by SDS-PAGE
Table 4. Content of Proinsulin In Purification Process Content of proinsulin (ppm)a
Purification step 1 2 3
Lysis cells transformed by proinsulin expression vector. Inclusion bodies containing proinsulin. Proinsulin was cleavaged to become Arg (B31)-insulin by trypsin SP Sepharose Fast Flow cation exchange chromatography 3 kDa ultrafiltration of Arg (B31)-insulin. Arg (B31)-insulin was cleavaged to become insulin by carboxypeptidase. Purification of insulin by C8 reverse phase HPLC column 3 kDa ultrafiltration of insulin. Insulin crystals.
4 5 a
ppm (parts per million) values were calculated by the amount proinsulin corrected for total protein content.
340,000 (34%) 690,000 (69%) < 200 < 50