IgG is a tetrameric protein composed of two copies each of the light and heavy chains. The four-chain structure is maintained by strong noncovalent interactions between the amino-terminal half of pairs of heavy-light chains and between the carboxyl-terminal regions of the two heavy chains. In addition, interchain disulfide bonds link each heavy-light chain and also link the paired heavy chains. An engineered human IgG4 specific for human tumor necrosis factor-cy (CDP571) is similar to human myeloma IgG4 in that it is secreted as both disulfide bonded tetramers (approximately 75% of the total amount of IgG) and as tetramers composed of nondisulfide bonded half-IgG4 (heavy chain disulfide bonded to light chain) molecules. However, when CDP571 was genetically engineered with a proline at residue 229 of the core hinge region rather than serine, CDP57 I(S229P), or with an IgGl rather than IgG4 hinge region, CDP571(yl), only trace amounts of nondisulfide bonded half-lgG tetramers were observed. Trypsin digest reversephase HPLC peptide mapping studies of CDP57 1 and CDP57 1 (y I ) with on-line electrospray ionization mass spectroscopy supplemented with Edman sequencing identified the chemical factor preventing inter-heavy chain disulfide bond formation between half-IgG molecules: the two cysteines in the IgG4 and IgGl core hinge region KPSCP and GPPCP, respectively) are capable of forming an intrachain disulfide bond. Conformational modeling studies on cyclic disulfide bonded CPSCP and CPPCP peptides yielded energy ranges for the low-energy conformations of 31-33 kcallmol and 40-42 kcal/mol, respectively. In addition, higher torsion and angle bending energies were observed for the CPPCP peptide due to backbone constraints caused by the extra proline. These modeling results suggest a reason why a larger fraction of intrachain bonds are observed in IgG4 rather than IgGl molecules: the serine in the core hinge region of IgG4 allows more hinge region flexibility than the proline of IgGl and thus may permit formation of a stable intrachain disulfide bond more readily.

Sialylated oligosaccharide structures were determined by the technique of electrospray ionization mass spectroscopy at seven of eight N-linked glycosylation sites of recombinant human ICAM-1des454-532 [tICAM(453)] purified from the tissue culture fluid of Chinese hamster ovary, human embryonic kidney, and mouse myeloma cell lines. The number of structures at each site depended on the cell line and ranged from 8 to 34. N-Glycolyneuraminic acid, a human oncofetal antigen, was found at all sites of all three cell line derived forms of tICAM(453). Tetraantennary complex structures containing one and/or two galactose-beta 1,4 N-acetylglucosamine repeats, characteristic of membrane bound proteins, were found on soluble tICAM(453) primarily at Asn-379. Asn-379, located between the D4 and D5 domains, is believed to be located close to the membrane surface in membrane bound ICAM-1. It has been proposed that the extent of N-linked glycosylation at Asn-240 and Asn-269 in the third domain of ICAM-1 may regulate the binding avidity of ICAM-1 to Mac-1 [Diamond, M. S., Staunton, D. E., Marlin, S. D., & Springer, T. A. (1991) Cell 65, 961-971]. In the present study the tICAM(453) Asn-269 site was found to contain predominantly one oligosaccharide structure that is conserved in all three cell lines. On the other hand, the Asn-240 site was found to contain cell line dependent oligosaccharide structural heterogeneity particularly in the degree of sialylation.

A PEGylated glucagon-like peptide-1 (GLP-1) agonist and glucagon antagonist hybrid peptide was engineered as a potential treatment for type 2 diabetes. To support preclinical development of this PEGylated dual-acting peptide for diabetes (DAPD), we developed a reproducible method for PEGylation, purification, and analysis. Optimal conditions for site-specific PEGylation with 22 and 43 kDa maleimide-polyethylene glycol (maleimide-PEG) polymers were identified by evaluating pH, reaction time, and reactant molar ratio parameters. A 3-step purification process was developed and successfully implemented to purify PEGylated DAPD and remove excess uncoupled PEG and free peptide. Five lots of 43 kDa PEGylated DAPD with starting peptide amounts of 100 mg were produced with overall yields of 53% to 71%. Analytical characterization by N-terminal sequencing, amino acid analysis, matrix-assisted laser desorption/ionization mass spectrometry, and GLP-1 receptor activation assay confirmed site-specific attachment of PEG at the engineered cysteine residue, expected molecular weight, correct amino acid sequence and composition, and consistent functional activity. Purity and safety analysis by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), analytical ion-exchange chromatography, reversed-phase high-performance liquid chromatography, and limulus amebocyte lysate test showed that the final products contained <1% free peptide, <5% uncoupled PEG, and <0.2 endotoxin units per milligram of peptide. These results demonstrate that the PEGylation and purification process we developed was consistent and effective in producing PEGylated DAPD preclinical materials at the 100 mg (peptide weight basis) or 1.2 g (drug substance weight basis) scale.