The development of a method for determining hydrophobicity constants for small, organic molecules by reversed-phase liquid chromatography (RPLC) is presented. The method uses capacity ratios measured at a number of different compositions of methanol to obtain derived values, denoted log k'w, upon which a new scale of hydrophobicity constants can be developed. This scale eliminates potential problems such as peak inversion that hamper RPLC methods using isocratic data to estimate hydrophobicity. The differential hydrogen bond effect observed in most correlations of RPLC data with logarithms of octanol-water partition coefficients (log Po/w) for compounds of opposite net hydrogen bonding capabilities (noncongeners) was minimized by adding trace quantities of n-decylamine and 1-octanol to the eluent and using an octyl-modified silica gel stationary phase. Values of log k'w are shown to be largely column-independent as long as the hydrophobic properties of columns are similar. The correlation of log k'w values with the logarithms of bovine serum albumin binding constants (log 1/C) is shown to be statistically indistinguishable from the correlation of log 1/C with log Po/w, indicating that this data models log 1/C as well as log Po/w for these compounds. Additionally, the chromatographic system is automatable and thus capable of higher sample throughput than measurements of log Po/w by the shake-flask method.

This report describes the effect on antigen binding of an isomerized aspartate residue located in the complementarity-determining regions (CDRs) of a recombinant monoclonal antibody. The antibody, which binds human IgE, contains two Asp-Gly sequences within its CDRs, but only one site was found to be labile to isomerization. Isolation and characterization of antibody fragments differing in the labile sequence were facilitated by using a technique involving hydrophobic interaction chromatography (HIC) that separates aspartyl, isoaspartyl, and cyclic imide variants to the residue located in CDR-L1. The variants were isolated for structural characterization and for determination of their relative antigen binding affinities. Mutants were constructed with altered residues to obviate the effects of isomerization and were evaluated for their ability to bind to IgE. Inspection of published crystal structures of CDRs of antibodies indicated that hydrogen binding of the Asp side chain of the unreactive residue may be the constraint that prevents isomerization. The strategy outlined here may prove to be of general utility in the biochemical and immunochemical characterization of recombinant antibodies.

No abstract

No abstract

The secretory glycoprotein DNase I acquires mannose 6-phosphate moieties on its Asn-linked oligosaccharides, indicating that it is a substrate for UDP-GlcNAc:lysosomal enzyme N-acetylglucosamine-1-phosphotransferase (phosphotransferase) (Cacia, J., Quan, C., and Frenz, J. (1995) Glycobiology 4, 99). Phosphotransferase recognizes a conformation-dependent protein determinant that is present in lysosomal hydrolases, but absent in most secretory glycoproteins. To identify the amino acid residues of DNase I that are required for interaction with phosphotransferase, wild-type and mutant forms of bovine DNase I were expressed in COS-1 cells and the extent of oligosaccharide phosphorylation determined. Phosphorylation of DNase I oligosaccharides decreased from 12.6% to 2.3% when Lys-50, Lys-124, and Arg-27 were mutated to alanines, indicating that these residues are required for the basal level of phosphorylation. Mutation of lysines at other positions did not impair phosphorylation, demonstrating the selectivity of this process. When Arg-27 was replaced with a lysine, phosphorylation increased to 54%, showing that phosphotransferase prefers lysine residues to arginines. Mutation of Asn-74 to a lysine also increased phosphorylation to 50.3%, and the double mutant (R27K/ N74K) was phosphorylated 79%, equivalent to the values obtained with lysosomal hydrolases. Interestingly, Lys-27 and Lys-74 caused selective phosphorylation of the neighboring Asn-linked oligosaccharide. Finally, mutation of Lys-117 to an alanine stimulated phosphorylation, demonstrating that some residues may be negative regulators of this process. We conclude that selected lysine and arginine residues on the surface of DNase I constitute the major elements of the phosphotransferase recognition domain present on this secretory glycoprotein.In many cell types, the sorting of newly synthesized acid hydrolases from secreted proteins is mediated by the phosphomannosyl recognition system (1, 2). The specificity of this pathway is determined by the enzyme UDP-GlcNAc:lysosomal enzyme N-acetylglucosamine-1-phosphotransferase (phosphotransferase) 1 which recognizes a conformation-dependent protein determinant that is present in lysosomal hydrolases but absent in most secretory glycoproteins (3). Using a variety of approaches, evidence has been obtained that the recognition determinant involves a broad surface patch that includes critical lysine residues (4 -10). The interaction of phosphotransferase with its acid hydrolase substrates results in the transfer of GlcNAc-P to mannose residues on the Asn-linked high mannose oligosaccharides of the lysosomal hydrolases. The GlcNAc residues are then removed by N-acetylglucosamine-1-phosphodiester ␣-N-acetylglucosaminidase to generate phosphomannosyl residues which mediate binding of the hydrolases to mannose 6-phosphate (Man-6-P) receptors present in the Golgi. These complexes are subsequently translocated via clathrincoated vesicles to endosomes where the hydrolases are discharged for packaging into lysosomes.While ph...