Characterization of structurally defined epitopes recognized by monoclonal antibodies produced by chronic lymphocytic leukemia B cells

Abstract

Despite a wealth of information about the structure of surface membrane immunoglobulin (smIg) on chronic lymphocytic leukemia (CLL) cells, little is known about epitopes reacting with their binding sites. Probing phage-displayed peptide libraries, we identified and characterized mimetopes for Igs of 4 patients with IGHV mutated CLL (M-CLL) and 4 with IGHV unmutated CLL (U-CLL). Six of these mAbs were representatives of stereotyped B-cell receptors characteristic of CLL. We found that mimetic epitopes for U- and M-CLL Igs differed significantly. M-CLL-derived peptides exhibited better amino acid motifs, were more similar to each other, aligned more easily, and formed tighter clusters than U-CLL-derived peptides. Mono-, oligo-, and polyreactivity of peptides correlated with structural changes within antigen-binding sites of selecting M-CLL mAbs. Although M-CLL-isolated peptides and certain U-CLL mAbs bound more effectively to the selecting mAb, others were not as specific, reacting with M-CLL and U-CLL mAbs; these data suggest that in vivo structurally diverse epitopes could bind smIgs of distinct CLL clones, thereby altering survival and growth. Finally, an M-CLL-derived peptide inhibited, in a dose-dependent manner, binding of its homologous mAb to human B lymphocytes; therefore peptides that inhibit or alter the consequences of antigen-smIg interactions may represent therapeutic modalities in CLL.

Figure 1

CLL mAbs bind phage-displayed peptides in a dose-dependent manner. Individual phages isolated after 3 rounds of selection with each CLL mAb were amplified and their antibody binding properties were studied by ELISA. A random phage clone derived from the initial phage library before selection was used as a negative control. (A-D) M-CLL mAbs 169, 183, 255, and 412. (E-H) U-CLL mAbs 014, 068, 114, and 270.

Figure 2

Comparisons of pairwise alignment scores and pairwise edit distances between M- and U-CLL mAbs. (A) Pairwise alignment scores. Pooled pairwise alignment scores for M-CLL peptides were significantly higher than those for U-CLL peptides (Student t test, P < .001). (B) Pairwise edit distances. As a group, M-CLL peptides had significantly shorter pairwise distances than did U-CLL peptides (Student t test, P < .001). The error bars extend to the extreme values of the data or 1.5× the height of the box, whichever is less. The height of the box represents the difference between the first and third quartiles. The top error bar in panel B merges with the top of the box plot for U-CLL peptides.

Figure 3

Binding of M-CLL and U-CLL mAbs to synthetic peptides derived by selection with M-CLL mAbs. Binding of M-CLL and U-CLL mAbs to 12mer peptides with sequences derived from phage selection analyses was tested by ELISA. Peptides with a scrambled amino acid order were used as negative controls. (A-D) CLL mAb reactivities with peptides 169-2, 169-8, 183-adapted, and 412-9, respectively. (E-H) Reactivities of mAbs to scrambled peptides.

Figure 4

Binding of M-CLL mAb 412 to peptide 412-9 and its fragments. Fragments of peptide 412-9 were synthesized (peptides 412-9-1, -2, -3, -4; bottom) and tested for binding to CLL 412 mAb in ELISA. Peptide 412-9scr with a scrambled amino acid order was used as a negative control.

Figure 5

Binding of M- and U-CLL mAbs to synthetic peptides derived by selection with U-CLL mAbs. Binding of M- and U-CLL mAbs to 12mer peptides with sequences derived from phage selection analyses with U-CLL mAbs was tested by ELISA. (A-B) Reactivities with peptides 068-4 and 114-1, respectively. (C-D) Reactivities of mAbs with scrambled peptides used as negative controls.

Figure 6

Inhibition of M-CLL mAb 183 binding to an autoantigen on the surface of viable RAMOS B cells by soluble peptide. mAb 183 (12.5 μg/mL) was preincubated with soluble 183-adapted peptide (pep183) or scrambled peptide (pep183scr) at indicated concentrations. Percentage and mean fluorescence intensity (in brackets) of mAb 183 staining of RAMOS cells are presented in each panel. mAb 183 binding without peptide preincubation is shown for comparison.