Hydrophobic Interaction Chromatography of Soluble Interleukin I

Aug 16, 2008 - HIC resolved the degraded molecules into three peaks. A combination of several analytical techniques, including cyanogen bromide cleava...
0 downloads 0 Views 361KB Size
Anal. Chem. 2008, 80, 7022–7028

Hydrophobic Interaction Chromatography of Soluble Interleukin I Receptor Type II To Reveal Chemical Degradations Resulting in Loss of Potency Yuling Zhang,* Theresa Martinez, Brian Woodruff, Andy Goetze, Robert Bailey, Dean Pettit, and Alain Balland* Department of Analytical and Formulation Sciences, Amgen Inc. 1201 Amgen Court West, Seattle, Washington 98119 A hydrophobic interaction chromatography method was developed to analyze recombinant soluble Interleukin 1 receptor type II (sIL-1R type II) drug substance and assess the stability of the drug under accelerated degradation studies. HIC resolved the degraded molecules into three peaks. A combination of several analytical techniques, including cyanogen bromide cleavage, reversedphase chromatography, mass spectrometry, and N-terminal sequencing, were used to identify the origins of these peaks. We found that accelerated degradation resulted from three different events, deamidation and isomerization at asparagine 317 (Asn317), C-terminal cleavage, and aggregation. The iso-aspartate 317 (iso-Asp317)containing species were shown to elute in HIC peak I and the Asp317-containing species in HIC peak II, respectively. Deamidation-isomerization to iso-Asp317, but not deamidation to Asp317, resulted in altered retention time on HIC companied by loss of potency, presumably by introducing a significant conformational change. CNBr C-terminal analysis showed that the inactive HIC peak I consisted of sIL-1R type II with “large” C-terminal truncations of 13 or 14 amino acids, whereas the active HIC peak II contained C-terminally full length and “small” C-terminal clips of two amino acids. Molecular modeling indicates that the short loop D317-S320, in the third domain of IL-1R type II, has a crucial impact on the stability of the molecule. Hydrophobic interaction chromatography (HIC) separates proteins including monoclonal antibodies (mAbs) based on their weak hydrophobic interactions with ligands on HIC adsorbents. The strength of these interactions depends on hydrophobic moieties exposed on proteins surfaces which govern the elution order in HIC separation. HIC had been widely used as preparative purification procedure for proteins.1 Since HIC operates under native conditions, it is nondestructive to proteins in terms of chemical or conformational structure. For these reasons, HIC is an ideal tool for isolation of protein fractions for further and in * To whom correspondence should be addressed. E-mail: [email protected] (Y.Z); [email protected] (A.B.). (1) Shukla, A. A.; Peterson, J.; Sorge, L.; Lewis, P.; Thomas, S.; Waugh, S. Biotechnol. Prog. 2002, 18, 556–64.

7022

Analytical Chemistry, Vol. 80, No. 18, September 15, 2008

depth characterization by high-resolution mass spectrometry and potency assays to study post-translational modifications in therapeutic recombinant proteins. A few published articles on analytical HIC separation and characterization included separation of succinimides and fragments of antibodies.2,3 Deamidation of asparagine residues, one of the most common protein post-translational modifications, can impact the stability profiles of therapeutic molecules. Following deamidation of asparagine, proteins can undergo racemization and isomerization4-9 sometimes associated with a reduction of biological activity.9-12 Isolation of iso-aspartate (iso-Asp) species and its succinimide variant from aspartate (Asp) is still not easily achieved due to molecular similarity of these species and rapid hydrolysis of succinimide.13-17 Diverse effects of these modifications on potency have been described in the literature. Recombinant human DNase11 and recombinant soluble CD412 were reported examples of reduction or loss of biological activities, whereas deamidated Asp 10-containing rhSCF showed 150% increase in potency (2) Kwong, M.; Harris, R. Protein Sci. 1994, 3, 147–149. (3) Valliere-Douglass, J.; Jones, L.; Shpektor, D.; Kodama, P.; Wallace, A.; Balland, A.; Bailey, R.; Zhang, Y. Anal. Chem. 2008, 80, 3168–3174. (4) McFadden, P. N.; Clarke, S. J. Biol. Chem. 1986, 261, 11503–11. (5) Stephenson, R.; Clarke, S. J. Biol. Chem. 1989, 264, 6164–6170. (6) Huang, L.; Lu, J.; Wroblewski, V.; Beals, J.; Riggin, R. Anal. Chem. 2005, 77, 1432–1439. (7) Zhang, W.; Czupryn, M. J. Pharm. Biomed. Anal. 2003, 30, 1479–1490. (8) Lyubarskaya, Y.; Houde, D.; Woodard, J.; Murphy, D.; Mhatre, R. Anal. Biochem. 2006, 348, 24–39. (9) Orpiszewski, J.; Schormann, N.; Kluve-Beckerman, B.; Liepnteks, J.; Benson, M. FASEB J. 2000, 14, 1255–1263. (10) Napper, S.; Delbaere, T.; Waygood, E. J. Biol. Chem. 1999, 274, 21776– 21782. (11) Cacia, J.; Quan, C.; Vasser, M.; Sliwkowski, M.; Frenz, J. J. Chromatogr., A 1993, 634, 229–239. (12) Teshima, G.; Porter, J.; Yim, K.; Ling, V.; Guzzetta, A. Biochemistry 1991, 30, 3916–3922. (13) Violand, B.; Schlittler, M.; Kolodiej, E.; Toren, P.; Cabonce, M.; Siegel, M.; Duffin, K.; Zobel, J.; Smith, C.; Tou, J. Protein Sci. 1992, 1, 1634–1641. (14) Bischoff, R.; Lepage, P.; Jaquinod, M.; Cauet, G.; Acker-Klein, M.; Clesse, D.; Laporte, M.; Bayol, A.; Van Dorsselaer, A.; Roitsch, C. Biochemistry 1993, 32, 725–734. (15) Teshima, G.; Stults, J.; Ling, V.; Canova-Davis, E. J. Biol. Chem. 1991, 266, 13544–13547. (16) Chu, G.; Chelius, D.; Gang, X.; Khor, H.; Coulibaly, S.; Bondarenko, P. Pharm. Res. 2007, 24, 1145–1156. (17) Athmer, L.; Kindrachuk, J.; Georges, F.; Napper, S. J. Biol. Chem. 2002, 277, 30502–30507. 10.1021/ac800928z CCC: $40.75  2008 American Chemical Society Published on Web 08/16/2008

compared to reference with Asn 10.18 These variable responses illustrate the importance of developing a nondestructive analytical method for separation and fractionation to maintain the protein integrity for characterization of deamidation as well as for evaluation of its biological activity or antigenicity or both. Interleukin-l receptor type II (IL-1R type II) is an IL-1 binding protein that cannot mediate any biological signals and naturally regulates IL-1 activity by acting as a decoy receptor.19 Implication of IL-1 in diseases such as arthritis,20 type II diabetes,21 and Alzheimer’s disease22 has been studied extensively, and this cytokine presents an interesting medical target. Consequently, recombinant IL-1R type II has been expressed as a soluble form in CHO cells to evaluate its potential usage as an inhibitor of IL-1 activity for therapeutic applications.23-26 Here we present hydrophobic interaction chromatography at analytical and semipreparative scales as a method to monitor product quality and characterize recombinant soluble IL-1R type II. The HIC assay is shown to have stability-indicating properties and be well-suited as a purification tool to isolate various degradation products for further characterization and assessment of their binding activities. MATERIAL AND METHODS Materials. Commercially available reagents were purchased from suppliers as follows. Guanidine HCl (ultra pure) from ICN Biomedicals (catalog no. 820540); tris (base) from JT Baker (catalog no. 4109-02); dithiothreitol (DTT) from Roche (catalog no. 100034, desiccated); trifluoroacetic acid (TFA, sequanal grade, catalog no. 28902) from Pierce; isopropanol (catalog no. AH3234) from Burdick & Jackson; acetonitrile, HPLC grade (catalog no. AX0142-1) from EM Sciences; CNBr (catalog no. U46619), phosphate buffered saline (10XPBS, catalog no. P3621), tween20 (polyoxyethylene-sorbitan monolaurate, catalog no. P1379), bovine serum albumin (BSA) (catalog no. A7906), biotinamidocaproate N-hydroxysuccinimide ester (“biotin ester”, catalog no. B-2643), dimethyl sulfoxide (DMSO, catalog no. D-5879), and sodium azide (NaN3, catalog no. D-5879) from Sigma; streptavidin: HRPO from Jackson ImmunoResearch Laboratories (catalog no. 016-030-084, 1 mg, lyophilized powder lot no. 35459); HuIL-1β, carrier-free, (18.4 µL at 1.37 mg/mL) from R&D Systems (catalog no. 201-LB-025/CF, 25 µg). sIL-1R Type II. Recombinant human sIL-1R type II was produced as a soluble receptor in Chinese hamster ovary (CHO) cells.23-26 The specific amino acid residues of 1-333 (Figure 1) in the soluble constructs were from genbank NM-004633.25 The (18) Hsu, Y.; Chang, W.; Mendiaz, E.; Hara, S.; Chow, D.; Mann, M.; Langley, K.; Lu, H. Biochemistry 1998, 37, 2251–2262. (19) Colotta, F.; Bertini, M.; Polentarutti, N.; Sironi, M.; Giri, J.; Dower, S.; Sims, J.; Mantovani, A. Science 1993, 261, 472–475. (20) Sims, J.; Smith, D. Eur. Cytokine Network 2003, 14, 77–81. (21) O’Connor, J.; Satpathy, A.; Hartman, M.; Horvath, E.; Kelley, K.; Dantzer, R.; Johnson, R.; Freund, G. J. Immunol. 2005, 174, 4991–4997. (22) Garlind, A.; Brauner, A.; Hojeberg, B.; Basun, H.; Schultzberg, M. Brain Res. 1999, 826, 112–116. (23) Smith, D.; Ketchem, R.; Moore, H.; Anderson, Z.; Renshaw, B.; Friend, D.; Sims, J. J. Biol. Chem. 2002, 277, 47619–47625. (24) Sims, J. Curr. Opin. Immunol. 2002, 14, 117–122. (25) Morris, A. E.; Lee, C. C.; Hodges, K.; Aldrich, T. L.; Krantz, C.; Smidt, P. S.; Thomas, J. N. Animal Cell Technology; Kluwer Academic Publishers: Norwell, MA, 1997; pp 529-534. (26) Smith, D.; Hanna, R.; Friend, D.; Moore, H.; Chen, H.; Farese, A. M.; MacVittie, T. J.; Virca, G. D.; Sims, J. Immunity 2003, 18, 87–96.

Figure 1. Amino acid sequence of soluble IL-1R type II.

amino acid sequence was cloned in the 2a5ib plasmid26 for a stable expression in the CHO cells. The process used a serum-free medium with CHO cells in suspension that was scaled up using three sequential suspension bioreactors to a final production volume of 1 600 L. The recombinant molecule was purified to homogeneity from the clarified supernatant using four chromatographic steps, anion exchange (AEX), hydrophobic interaction (HIC), cation exchange (CEX), and hydroxyapatite (HA). The stability samples at 50 mg/mL formulated in 20 mM phosphate and 120 mM NaCl pH 7 (a formulation buffer) and incubated at various temperatures for various amounts of time include: (1) 2-8 °C bulk drug substance (BDS) as a control for comparison; (2) 37 °C for 3 days (BDS-37C-3D); and (3) 30 °C for 3 months (BDS-30C-3M). The degraded samples were then diluted in the formulation buffer to reach a concentration of 1 mg/ mL and stored at