Anal. Chem. 1998, 70, 425-429
Analysis of Pichia pastoris Components in Recombinant Human Serum Albumin by Immunological Assays and by HPLC with Pulsed Amperometric Detection Wataru Ohtani,* Toyoo Ohda, Akinori Sumi, Kaoru Kobayashi, and Takao Ohmura
Research Division, The Green Cross Corporation, Hirakata, Osaka 573, Japan
We have developed a recombinant human serum albumin (rHSA) from Pichia pastoris which expresses high levels of heterologous proteins. rHSA is used clinically in high concentration (approximately 250 mg/ml in a 50 mL vial). We had to consider not only proteins from host cells as impurities but also mannan, which exhibits harmful effects on humans. However, the analysis of mannan in biopharmaceuticals produced from yeast has not been reported. Contaminating mannans in the final product were one important index to assess the clinical safety of rHSA. We have developed a highly sensitive enzyme immunoassay (EIA), utilizing an avidin-biotin system, for the detection of either the protein or mannan polysaccharide components from P. pastoris components (PPC) in rHSA. In addition, we used anion exchange chromatography with pulsed amperometric detection (AE-PAD) for monosaccharide analysis of glycoconjugates for the detection of mannan from PPC in rHSA. The detection limits of the EIA for PPC (PPC EIA) and the AE-PAD were 1 ng of protein/250 mg of rHSA and 180 ng of mannose/mg of rHSA, respectively. The mannan content in partially purified rHSA as determined by the AE-PAD was about same as the PPC content as determined by the PPC EIA. We showed that the PPC EIA and the AE-PAD are useful methods for the purity analysis of biopharmaceuticals produced from yeast. Human serum albumin (HSA)1 is the major protein component of human plasma. It is a 66.5 kDa single-chain, non-glycosylated polypeptide chain of 585 amino acids.2 HSA is a heart-shaped molecule, comprised of three domains.3,4 HSA is synthesized in the liver. The primary function of HSA is the maintenance of the * Address correspondence to this author at Research Division, The Green Cross Corp., 2-25-1, Shodai-Ohtani, Hirakata, Osaka 573, Japan. (1) Abbreviations used are as follows: EIA, enzyme immunoassay; PPC, Pichia pastoris components; HSA, human serum albumin; rHSA, recombinant human serum albumin; pHSA, plasma-derived HSA; PPC EIA, enzyme immunoassay for Pichia pastoris components; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; HPLC, high-performance liquid chromatography; PBS, phosphate-buffered saline; PVDF, poly(vinylidene difluoride); SPBS, phosphate-buffered saline containing 0.4 % skim milk; AE-PAD, anion exchange chromatography with pulsed amperometric detection; CV, coefficient of variation. (2) Minghetti, P. P.; Ruffner, D. E.; Kuang, W.-J.; Dennison, O. E.; Hawkins, J. W.; Beattie, W. G.; Dugaiczyk, A. J. Biol. Chem. 1986, 261, 6747-6757. S0003-2700(97)00596-9 CCC: $15.00 Published on Web 01/15/1998
© 1998 American Chemical Society
colloid osmotic pressure within the circulatory system. However, HSA also acts as a carrier for several ligands, i.e., bilirubin and fatty acids. HSA is used clinically to treat severe hypoalbuminemia or traumatic shock.5,6 The usual clinical doses of HSA are in excess of 10 g/dose. The current HSA preparation is derived from human plasma. Human plasma is limited in supply and may contain unknown viral contaminants. The plasma-derived HSA (pHSA) products have been shown not to transmit human disease, viral or otherwise, through years of clinical experience. Several groups have attempted to produce HSA (rHSA) by recombinant DNA technology.7-11 We have developed a highlevel HSA expression system using Pichia pastoris and have established a purification method to produce highly purified rHSA.12,13 P. pastoris is a yeast which is particularly useful for expression of high levels of heterologous proteins. We recently found high titers of antibodies to P. pastoris components (PPC) in normal human serum. These antibodies to PPC are directed primarily against the mannan component of the PPC.14 Since the clinical dose of rHSA is in the multigram range, a small amount of PPC contamination in the rHSA could present a threat to the safety of rHSA. The purity of recombinant products may be estimated by methods such as sodium dodecyl sulfate-polyacrylamide gel (3) He, X. M.; Carter, D. C. Nature (London) 1992, 358, 209-215. (4) Carter, D. C.; Chang, B.; Ho, X. J.; Keeling, K.; Krishnasami, Z. Eur. J. Biochem. 1994, 226, 1049-1052. (5) Peters, T., Jr. In The Plasma Proteins, 2nd ed.; Putnam, F. W., Ed.; Academic Press: New York, 1975; Vol. 1, pp 133-181. (6) Foster, J. F. In Albumin Structure, Function and Uses; Rosenoer, V. M., Oratz, M., Rothschild, M. A., Eds.; Pergamon Press: Oxford, 1977; pp 5384. (7) Latta, M.; Knapp, M.; Sarmientos, P.; Brefort, G.; Becquart, J.; Guerrier, L.; Jung, G.; Mayaux, J. F. Bio/Technology 1987, 5, 1309-1314. (8) Saunders, C. W.; Schmidt, B. J.; Mallonee, R. L.; Guyer, M. S. J. Bacteriol. 1987, 169, 2917-2925. (9) Sijmons, P. C.; Dekker, B. M. M.; Schrammeijer, B.; Verwoerd, T. C.; van den Elzen, P. J. M.; Hoekema, A. Bio/Technology 1990, 8, 217-221. (10) Sleep, D.; Belfield, G. P.; Ballance, D. J.; Steven, J.; Jones, S.; Evans, L.R.; Moir, P. D.; Goodey, A. R. Bio/Technology 1991, 9, 183-187. (11) Fleer, R.; Yeh, P.; Amellal, N.; Maury, I.; Fournier, A.; Bacchetta, F.; Baduel, P.; Jung, G.; L’Hote, H.; Becquart, J.; Fukuhara, H.; Mayaux, J. F. Bio/ Technology 1991, 9, 968-975. (12) Sumi, A.; Ohtani, W.; Kobayashi, K.; Ohmura, T.; Yokoyama, K.; Nishida, M.; Suyama, T. Biotechnol. of Blood Proteins 1993, 227, 293-298. (13) Ohi, H.; Miura, M.; Hiramatsu, R.; Ohmura, T. Mol. Gen. Genet. 1994, 243, 489-499. (14) Ohtani, W.; Kobayashi, K.; Ohmura, T. J. Clin. Lab. Anal. 1997, 11, 196201.
Analytical Chemistry, Vol. 70, No. 2, January 15, 1998 425
electrophoresis (SDS-PAGE) or high-performance liquid chromatography (HPLC).15-17 However, conventional analytical methods are not sufficiently sensitive to detect the amounts of PPC which may be immunologically active in the rHSA. Immunological methods, such as immunoblotting and immunoassays, have been employed for other products.18-20 However, there are no reports on the purity analysis of recombinant HSA, from yeast, containing mannan and which had been used in multigram doses (12.5 g of HSA/50 mL vial) by intravenous injection. We have established a highly sensitive enzyme immunoassay (EIA) for the detection of low levels of PPC in the presence of a high concentration of recombinant product. Furthermore, we have used anion exchange chromatography with pulsed amperometric detection (AE-PAD) for monosaccharide analysis of glycoconjugates for the detection of mannan from P. pastoris. We report here on the analysis of protein and mannan impurities from yeast in rHSA. EXPERIMENTAL SECTION Materials. Purified rHSA was prepared according to the rHSA purification method,21 starting with a culture of the P. pastoris UHG42-3 strain carrying the gene for HSA.21 Partially purified rHSA was the in-process rHSA material after the rHSA capturing step (>95% purity as determined by SDS-PAGE and silver staining) according to the rHSA purification method.21 A PPC standard was the in-process material after the rHSA capturing step according to the rHSA purification method,15,21 starting with a culture of the P. pastoris UC43-14 strain lacking the gene for HSA.21 pHSA was from GCC Co. Ltd. (Osaka, Japan). Mannan was purchased from Sigma (St. Louis, MO). Protein Concentration. The rHSA concentration was determined by Kjeldahl’s method.22 The protein concentration of the PPC standard was determined by Lowry’s method using pHSA as the standard.23 Preparation of Rabbit Anti-PPC Antiserum. Three female white rabbits were given subcutaneous injections of 1 mL of the PPC standard (1.0 mg of protein) emulsified in an equal volume of complete Freund’s adjuvant (Difco Laboratories, Detroit, MI). Booster injections were given three times at biweekly intervals, using the same dose as the first immunization. The rabbits were bled three weeks after the final booster injection. The anti-PPC antiserum was collected and kept frozen (-80 °C) until use. Preparation of Rabbit Anti-PPC IgG. The rabbit anti-PPC antiserum (22.5 mL), comprised of an equal volume of each rabbit serum, was dialyzed against 17.5 mM phosphate buffer, pH 6.3. The serum was chromatographed on a DEAE-Sepharose fast flow column (1.6 cm × 20 cm, Pharmacia Biotech, Tokyo, Japan) that (15) Garnick, R. L. J. Pharm. Biomed. Anal. 1989, 7, 255-266. (16) Anicetti, V. In Analytical Biotechnology Capillary Electrophoresis and Chromatography; Horvath, C.; Nikelly, J. G., Eds.; ACS Symposium Series 434; American Chemical Society:Washington, DC, 1990; pp 127-140. (17) O’Keefe, D. O.; Will, M. L. In Chromatography of pharamaceuticals; Ahuja, S., Ed.; ACS Symposium Series 512; American Chemical Society: Washington, DC, 1992; pp 121-134. (18) Anicetti, V. R.; Fehskens, E. F.; Reed, B. R.; Chen, A. B.; Moore, P.; Geier, M.D.; Jones, A. J. S. J. Immunol. Methods 1986, 91, 213-224. (19) Briggs, J.; Panfili, P. R. Anal. Chem. 1991, 63, 850-859. (20) O’Keefe, D. O.; DePhillips, P.; Will, M. L. Pharm. Res. 1993, 10, 975-979. (21) Yokoyama, K.; Ohmura, T. Jpn. J. Apheresis 1995, 14, 19-20. (22) Conway, E. J. Microdiffusion Analysis and Volumetric Error, 5th ed.; Croby Lockwood and Son Ltd.: London, 1962. (23) Lowry, O. H.; Rosebrough, N. J.; Farr, A. L.; Randall, R. J. J. Biol. Chem. 1951, 193, 265-275.
426
Analytical Chemistry, Vol. 70, No. 2, January 15, 1998
had been previously equilibrated with the same buffer. Passthrough fractions were pooled, concentrated using a Centriprep10 (Amicon, Inc., Beverly, MA), and dialyzed against 0.15 M phosphate-buffered saline (PBS). Rabbit anti-PPC IgG (10 mL, A280 ) 15.4) was prepared. Preparation of Biotinylated Rabbit Anti-PPC IgG. Rabbit anti-PPC IgG was biotinylated using Sulfo-NHS-Biotin (Pierce, Rockford, IL) according to the general methods of Hnatowich et al.24 SDS-PAGE. SDS-PAGE was performed using 1.0-mm-thick 10-20% acrylamide gradient slab gels (Daiichi Pure Chemicals Co. Ltd., Tokyo, Japan) according to the method of Laemmli.25 Silver staining was performed using the Wako silver stain kit (Wako Co. Ltd., Osaka, Japan). Immunoblotting and Lectin Blotting. Following SDSPAGE, proteins were transferred to a poly(vinylidene difluoride) (PVDF) membrane (Tefco Co. Ltd., Tokyo, Japan) using a Tefco Model TB-8 transfer blotter (Tefco Co. Ltd.). Each PVDF membrane for immunoblotting and lectin blotting was immersed in PBS containing 0.4% skim milk (SPBS, pH7.2, Yukijirushi Co. Ltd., Sapporo, Japan) for 16 h at room temperature and 2 h at room temperature, respectively. The membranes were then incubated with gentle shaking in a solution of rabbit anti-PPC antiserum (100 µL in 10 mL of SPBS) for 2 h at room temperature or in a solution of biotinylated concanavalin A (100 µL in 10 mL of SPBS, Genzyme Co. Ltd., Cambridge, MA) for 16 h at room temperature. After the membranes were washed with saline containing 0.05% Tween 20, they were then incubated with gentle shaking in a solution of goat anti-rabbit IgG peroxidase conjugate (diluted to 1:500 with SPBS, KPL Inc., Gaitherburg, MD) for 2 h at room temperature or avidin peroxidase conjugate (diluted to 1:103 with SPBS, Vector Laboratories, Inc., Burlingame, CA) for 5 h at room temperature. After washing, substrate solution containing 4-chloro-1-naphthol (Wako Co. Ltd.) and H2O2 (Wako Co. Ltd.) was incubated with the membranes. The reaction was quenched with distilled water. EIA for PPC (PPC EIA) Procedure. Microtiter plates (Costar, Cambridge, MA) were coated with 100 µL/well of 10.0 µg/mL rabbit anti-PPC IgG in 0.05 M sodium carbonate buffer, pH 9.6, overnight at room temperature. The antibody solution in the wells was aspirated and the excess binding sites were blocked with 300 µL/well of SPBS for 2 h at room temperature. The SPBS in the wells was aspirated and 50 µL of either the PPC standard or samples was added to appropriate wells. After incubation for 3 h at 37 °C, the plates were washed with saline containing 0.05% Tween 20, and 50 µL per well of biotinylated rabbit anti-PPC IgG (diluted to 1:3000 with SPBS) was added. The plates were incubated for 2 h at 37 °C and again washed with saline containing 0.05% Tween 20. Fifty microliters per well of avidin peroxidase conjugate (diluted to 1:104 with SPBS) was then added. The plates were incubated for 1 h at 37 °C and again washed with saline containing 0.05% Tween 20. Fifty microliters of o-phenylenediamine substrate solution supplied with the ELISA color reagent kit (Sumitomo Bakelite Co. Ltd., Tokyo, Japan) was then added to each well and allowed to react for 30 min at room temperature. The reaction was stopped with 50 µL/well of 2 N H2SO4. The (24) Hnatowich, D. J.; Virzi, F.; Rusckowski, M. J. Nucl. Med. 1987, 28, 12941302. (25) Laemmli, U. K. Nature (London) 1970, 227, 680-685.
Figure 1. Immunoblotting of PPC standard with rabbit anti-PPC antiserum and lectin blotting patterns of PPC standard with concanavalin A. Lanes 1, 3, and 5, molecular weight standards: (A) Lane 2, silver staining following SDS-PAGE (10 µg of protein/lane); (B) lane 4, immunoblotting (10 µg of protein/lane); (C) lane 6, lectin blotting (50 µg of protein/lane).
resulting absorbance in each well was read at 490 nm using a Tmax Microplate reader (Molecular Devices, Menlo Park, CA). Analysis of P. pastoris Mannan by AE-PAD. Samples were mixed with an equal volume of 10% trichloroacetic acid, incubated for 15 min at room temperature and clarified by centrifugation. The clarified supernatant was evaporated to dryness, dissolved in water, and analyzed by the method of Hardy et al.26 Statistical Analysis. Statistical comparisons were analyzed using Student’s nonpaired t-test or a one-way ANOVA using the software package Microsoft EXCEL (Microsoft Co. Ltd., Tokyo, Japan).27 RESULTS AND DISCUSSION SDS-PAGE, Immunoblotting, and Lectin Blotting. The results of silver staining for the PPC standard and the immunoblotting pattern using rabbit anti-PPC antiserum are shown in Figure 1. Numerous bands were observed in the 10 -100 kDa molecular mass range in the silver-stained SDS-polyacrylamide gel. The immunoblotting pattern did not exactly match the PPC standard pattern obtained by silver staining. Smeared bands not evident following silver staining of the SDS-polyacrylamide gel were observed above 100 kDa by immunoblotting (Figure 1). A sharp band and a smeared band were observed at molecular masses of approximately 55 and above 100 kDa by lectin blotting, respectively (Figure 1). Lectin (concanavalin A) possesses a high binding specificity for the D-mannopyranose unit.28 Anti-PPC protein and polysaccharide antibodies were present in the rabbit anti-PPC IgG preparation. It is well known that P. pastoris possesses mannans, which share a common structure with some yeast and Candida albicans mannans, and that humans have a high titer of antibody to mannan.29-34 The mannans possess biological effects which induce an anaphylaxis-like reaction and release histamine.35 (26) Hardy, M. R.; Townsend, R. R.; Lee, Y. C. Anal. Biochem. 1988, 170, 5462. (27) Tango, T. In Igaku Toukeigaku Handbook; Miyahara, H., Tango, T., Eds.; Asakura Shoten Co. Ltd.: Tokyo, 1995; pp 514-541. (28) Tojo, M.; Shibata, N.; Osanai, T.; Mikami, T.; Suzuki, M.; Suzuki, S. Carbohydr. Res. 1991, 213, 325-330.
Figure 2. Calibration curve of PPC EIA. Curve fit: four-parameter equation, Y ) (A - D)/(1 + (X/C)B) + D, where A ) 0.009 06, B ) 1.04, C ) 146, D ) 4.37, and the correlation coefficient ) 0.999.
Validation of the PPC EIA and Purity Analysis of rHSA. The calibration curve of a solid phase EIA is often calculated as a nonlinear curve. A four-parameter logistic fit was used for the calibration curve for the PPC EIA.36 The four-parameter equation is y ) (A - D)/(1 + (X/C)B) + D. The coefficient of correlation was 0.999 (Figure 2). X and Y axes are expressed as the log of PPC protein concentration and as optical density, respectively. The sugar content in the PPC standard, which was determined by the phenol-sulfuric acid reaction using mannose as the standard,37 was 1/6.35 of the protein content. Equivalent concentrations (250 mg/mL) of purified rHSA and pHSA were measured in the PPC EIA. The resulting OD between purified rHSA and pHSA was not significantly different, as determined by Student’s two-tailed t-test (degrees of freedom, 4; p > 0.05). PPC recoveries were within 5% over the range of PPC from 4.0 to 64.0 ng of protein/mL, and the coefficients of variation (CV) were 1.6-2.3% (Table 1). The EIA was sensitive to 1.0 ng of protein/mL of PPC. The absorbance at 1.0 ng of protein/mL of PPC was significantly different from that of the 0 ng of protein/mL of PPC blank sample (degrees of freedom, 2; p < 0.01). We validated the PPC EIA using partially purified rHSA containing purification process-specific PPC. The partially purified rHSA diluent was assayed for 6 days (three times per day). Intraassay, interassay, and total assay precisions were analyzed by one-way ANOVA (Tables 2 and 3). Intraassay precision was not significantly different from interassay precision (degrees of (29) Kumar, B. V.; Medoff, G.; Kobayashi, G. S.; Sieling, W. L. Infect. Immun. 1985, 48, 806-812. (30) Kobayashi, H.; Shibata, N.; Suzuki, S. Arch. Biochem. Biophys. 1986, 245, 494-503. (31) Ohtani, W.; Ohmizu, A.; Sumi, A.; Ohmura, T.; Uemura, Y.; Yokoyama, K. Yakugaku Zasshi 1988, 108, 339-344. (32) Mbawala, A.; Mahmood, S. A.; Loppinet, V.; Bonaly, R. J. Gen. Microbiol. 1990, 136, 1279-1284. (33) Namba, M.; Hanafusa, T.; Kono, N.; Tarui, S. (The Gl-G Hypoglycemia Study Group). Diabetes Res. Clin. Pract. 1993, 19, 133-138. (34) Suzuki, S. Yakugaku Zasshi 1995, 115, 280-294. (35) Nagase, T.; Mikami, T.; Suzuki, S.; Schuerch, C.; Suzuki, M. Microbiol. Immunol. 1984, 28, 997-1007. (36) Tijssen, P. In Laboratory Techniques in Biochemistry and Molecular Biology; Burdon, R. H., van Knippenberg, P. H., Eds.; Elsevier Science Publishers B. V.: Amsterdam, 1985; pp 407-411. (37) Dubois, M.; Gilles, K. A.; Hamilton, J. K.; Rebers, P. A.; Smith, F. Anal. Chem. 1956, 28, 350-356.
Analytical Chemistry, Vol. 70, No. 2, January 15, 1998
427
Table 1. Recoveries of PPC by the PPC EIA PPC (ng of protein/mL) added measd 256 128 64 32 16 8 4 2 1 a
225.3 134.6 65.6 31.5 15.7 8.18 4.10 2.28 1.08
recovery (%)
CVa (%)
88.0 105.2 102.5 98.4 98.3 102.3 102.5 114.2 107.6
19.3 2.3 2.1 1.6 2.3 1.6 2.0 3.4 5.5
Coefficient of variation, n ) 3.
Table 2. Reproducibility of Detection of PPC in Partially Purified rHSA Dilutions:a Determination by the PPC EIA days
mean PPC (ng of protein/mL)
SD
CVb (%)
1 2 3 4 5 6
12.05 11.79 11.84 12.35 11.75 12.34
0.434 0.189 0.210 0.288 0.269 0.252
3.60 1.60 1.78 2.33 2.29 2.04
Figure 3. Silver staining and immunoblotting patterns of rHSA. Lanes 1 and 8, molecular weight standards. (A) Silver staining following SDS-PAGE: lane 2, PPC standard (10 µg of protein/lane); lane 3, partially purified rHSA (0.1 µg/lane); lanes 4, 5, and 6, purified rHSA (Lot RA001, Lot RA027, and Lot RA028, 0.1 µg/lane); lane 7, pHSA (0.1 µg/lane). (B) Immunoblotting: lane 9, PPC standard (10 µg of protein /lane); lane 10, partially purified rHSA (10 µg/lane); lanes 11, 12, and 13, purified rHSA (Lot RA001, Lot RA027, and Lot RA028, 10 µg/lane); lane 14, pHSA (10 µg/lane).
a rHSA concentration was 377.3 ng/mL. b Coefficient of variation, n ) 3.
Table 3. Precision of the PPC EIA
intraassay interassay total assaya a
SD
CV (%)
σA ) 0.285 σE ) 0.218 σ ) 0.358
2.37 1.81 2.98
σ ) (σA2+σE2)1/2.
Table 4. Validation of the PPC EIA Using Partially Purified rHSA Diluent dilution (×104)
mean PPCa (mg of protein/mL)
SD
CVb (%)
6.4 12.8 25.6
1.55 1.54 1.55
0.0436 0.0555 0.0463
2.82 3.60 2.99
a These values were calculated as PPC in partially purified rHSA. rHSA concentration in partially purified rHSA was 48.3 mg/mL. b Coefficient of variation, n ) 3.
freedom: f1 ) 5, f2 ) 12, F ) 2.75; p > 0.05). The inter- and intraassay CVs were 2.37% and 1.81%, respectively. The total assay precision, calculated from intra- and interassay precision, was 2.98%. The PPC concentrations determined in each partially purified rHSA diluent were not significantly different, as determined by one-way ANOVA analysis (degrees of freedom: f1 ) 2, f2 ) 6, F ) 0.0045; p > 0.05, Table 4). The purity of partially purified rHSA was estimated to be 96.8% by the PPC EIA (Table 4). Purity Analysis by Immunoblotting. For partially purified rHSA, three bands were detected by SDS-PAGE followed by 428 Analytical Chemistry, Vol. 70, No. 2, January 15, 1998
Figure 4. Specificity of AE-PAD for the detection of mannan in purified rHSA. Arrows indicate the elution position of mannose. (A) Mannose, (B) purified rHSA containing mannan; (C) purified rHSA.
silver staining, whereas smeared bands, for P. pastoris mannan, with a molecular mass above 100 kDa were observed by immunoblotting (Figure 3). It was shown by immunoblotting using the anti-HSA antibody that the three bands for partially purified rHSA resulted from rHSA with a molecular mass of about 66.5
Table 5. Mannan in rHSA As Determined by the AE-PAD mannan (µg of mannose/mg of rHSA) partially purified rHSA purified rHSA a
4.5
