drugs (9, 26-28) in attempts to target these agents to tumors. These latter modifications all involve covalent attachment to tyrosines, e-amino side chains of lysines, carboxyl side chains of aspartic and glutamic acids, or sulfhydryl groups generated by reduction of cystines. Although each of these applications has shown promise, none, as yet, has reached the point of proven clinical utility. Since so many of the diagnostic and therapeutic applications envisioned for monoclonal antibodies require coupling of antibodies to other substances, there is a need for methods of covalent modification that can be used on a broad spectrum of different antibodies with minimal effect on antigen binding properties. This paper will present a strategy for site-specific covalent modification of antibodies based upon attachment through the oligosaccharide moiety of the molecule. Because of the restricted localization of the glycosylation sites on immunoglobulins, such an approach offers the potential advantage of modification of the antibodies at a site distal to the antigen combining site (29). Data will be presented that first compares the effect of various techniques of monoclonal antibody modification on the antigenbinding homogeneity and affinity of the conjugate. Second, the comparative in vivo biodistribution and tumor localization of additional radiolabeled monoclonal antibodies will be presented using nude mice bearing subcutaneous tumor xenografts.
MATERIALS AND METHODSIn Vitro Binding. The mouse monoclonal anti-phosphocholine IgM, HPCM2 (30), was grown in BALB/c mice and purified from ascites by ammonium sulfate precipitation, followed by affinity chromatography (31). The following conjugates of HPCM2 were prepared.(i) For oligosaccharide attachments, the oligosaccharides moieties were oxidized to aldehydes by incubation in the dark with 10-30 mM NaIO4 in phosphate-buffered saline (0.15 M NaCl/0.01 M sodium phosphate) at pH 6.0 on ice for 1 hr. After passage through a Sephadex G-25 column equilibrated with phosphate-buffered saline at pH 6.0, oxidized antibody was incubated with a 270-fold molar excess of 1,6-diaminohexylethylenediaminedi-o-hydroxyphenylacetic acid (1,6-diaminohexyl-EDDHA) § for 1 hr at room temperature. Sodium cyanoborohydride (Aldrich) was added to a final concentration of 10 mM, and the solution was incubated for an additional 4 hr and then dialyzed at 4°C versus several changes of phosphate-buffered saline.(ii) For aspartic/glutamic acid attachments, the antibody was incubated in 20 mM 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide (Sigma) at pH 5.9 for 2 hr at room temperature. The 1,6-diaminohexyl-EDDHA was added to a final concentration of 10 mM, and the solution was incubated at room temperature for an additional 2 hr. Ten microliters of 1 M ethanolamine was added, and, after 1 hr the reaction mixture was dialyzed versus phosphate-buffered saline.(iii) For lysine attachments, the scheme for aspartic/glutamic acid modification was used except that EDDHA (5 mM; Sigma) replaced the 1,6-diamino...
