Identification of heparin-binding amino acid residues in antibody HS4C3 with the potential to design antibodies against heparan sulfate domains

Abstract

Heparan sulfate (HS) is a linear polysaccharide with high structural and functional diversity. Detection and localization of HS in tissues can be performed using single chain variable fragment (scFv) antibodies. Although several anti-HS antibodies recognizing different sulfation motifs have been identified, little is known about their interaction with HS. In this study the interaction between the scFv antibody HS4C3 and heparin was investigated. Heparin-binding lysine and arginine residues were identified using a protect and label methodology. Site-directed mutagenesis was applied to further identify critical heparin-binding lysine/arginine residues using immunohistochemical and biochemical assays. In addition, computational docking of a heparin tetrasaccharide towards a 3-D homology model of HS4C3 was applied to identify potential heparin-binding sites. Of the 12 lysine and 15 arginine residues within the HS4C3 antibody, 6 and 9, respectively, were identified as heparin-binding. Most of these residues are located within one of the complementarity determining regions (CDR) or in their proximity. All basic amino acid residues in the CDR3 region of the heavy chain were involved in binding. Computational docking showed a heparin tetrasaccharide close to these regions. Mutagenesis of heparin-binding residues reduced or altered reactivity towards HS and heparin. Identification of heparin-binding arginine and lysine residues in HS4C3 allows for better understanding of the interaction with HS and creates a framework to rationally design antibodies targeting specific HS motifs.

Keywords: antibodies; heparan sulfate; heparin-binding residues; protect & label; site-specific mutations.

Fig. 1

Location of heparin-binding lysine and arginine residues. A) Amino acid sequence of HS4C3 with heparin binding lysine and arginine residue highlighted in blue and green, respectively. Non-binding residues are highlighted in magenta. B) Heparin-binding residues were localized in a HS4C3 homology model. Heparin-binding residues in the CDRs seem to cluster, forming a heparin-binding site. Heparin-binding residues located in other regions are scattered across the antibody. FR: Framework region. CDR: Complementarity determining region. V_H: Variable domain heavy chain. V_L: Variable domain light chain.

Fig. 2

Immunofluorescent staining of rat kidney cryosections with mutant HS4C3 antibodies. The HS epitopes of HS4C3 and mutant antibodies were localized in rat kidney sections. A) Typically, HS4C3 staining comprises the glomerulus (orange arrow) and peritubular capillaries (blue arrow). B) Mutant antibody R21S displays a similar staining pattern as HS4C3. C) Mutant antibody R104S antibody stains primarily Bowman’s capsule (white arrow). D) Mutant antibody K107S stains both the glomerulus and Bowman’s capsule. E) Mutant R105D displays a diminished staining. F) Control staining without an antibody. G) After treatment with heparinase I, II and III, the staining typical for the respective antibody is removed, as demonstrated for mutant antibody R104S. Scale bar: 75 μm.

Fig. 3

Immunoprecipitation of HS oligosaccharides by mutated anti-HS antibodies. Chemical structure of HS oligosaccharides GT228 (2S, NS), GT229 (6S, 2S, NS), and GT230 (3S, 6S, 2S, NS); HS oligosaccharides that were immunoprecipitated were subjected to PAGE and silver stained. The bound oligosaccharides in the eluates (A) and the non-bound oligosaccharides in the flow-through (B) were compared to those of the native HS4C3 antibody. The first lanes of all gels contain a mixture of the oligosaccharides; the last three lanes of both gels contain each oligosaccharide separately. 3S: 3-O sulfate. 6S: 6-O sulfate. 2S: 2-O sulfate. NS: N-sulfate.

Fig. 4

Docking of a heparin tetrasaccharide onto antibody HS4C3 using the ClusPro server. A tetrasaccharide was docked onto the homology model for HS4C3 using the ClusPro server. In each panel the heparin-binding arginines and lysines are highlighted in green and blue respectively, whereas the non-binding arginines and lysines are coloured magenta. The heparin tetrasaccharide is shown in red. In the first confirmation (A), heparin interacts with R¹⁰², R¹⁰⁴, R¹⁰⁵, R²²³, K⁵⁴, K¹⁰⁷, and K¹⁶². In the second conformation (B), heparin interacts with R¹⁰², R¹⁰⁴, R¹⁰⁵, R²²³, and K¹⁰⁷. In the third (C), fourth (D) and fifth (E) conformations heparin interacts with R⁵², R¹⁰⁴, R¹⁰⁵, R²²³, K⁵⁴, K¹⁰⁷.

Fig. 5

Schematic overview of the protect and label methodology. In two separate procedures, the arginine and lysine residues were modified. First, the periplasmic fraction, containing the scFv antibodies, was incubated with heparin beads (A) and the non-bound proteins were washed away (B). Subsequently, phenylglyoxal and NHS-acetate were added to chemically modify (=protect) the non-bound arginine and lysine residues respectively, indicated in magenta (C). Then the HS4C3 antibody was dissociated from the beads (D) and the arginine and lysine residues that were previously bound to heparin were labeled with (p-hydroxyphenyl)glyoxal and NHS-biotin, indicated in green and blue respectively (E). Finally, the antibody was digested using chymotrypsin and analyzed by mass spectrometry (F).