Myeloma paraprotein epitope identification using a phage library of sterically constrained peptides

Abstract

6626 Background: Chronic antigenic stimulation has been hypothesized to be a causative factor in rare cases of multiple myeloma (MM). If true, one might identify relevant antigens by using phage display technology and biopanning to analyze epitopes of m-proteins in sera of MM patients (pts). This method obviates the need to prospectively identify candidate antigens. Methods: The Ig fractions of 17 pts with IgG paraproteins were purified on Protein G beads. The bead-bound m-proteins were then used to biopan a T7 library of random heptapeptides sterically constrained by intramolecular disulfide bridges (Gift of G. Froman). Phage selected via 3–5 rounds of biopanning were cloned by plaque isolation. Binding specificity was confirmed by blotting and chemiluminescent detection. The cloned phage inserts’ DNA was sequenced and the amino acid sequences of each insert deduced. Proteins containing homologies with these peptides were found by short-peptide BLAST search. To test whether the identified epitopes are common among MM pts, the phage clones were screened against IgG m-proteins from sera of 70 additional pts. Results: Phage clones were identified which reacted specifically with 3 of the 17 pt m-proteins in the panel. There were 7 selected clones in total, as multiple clones were found for two pts. None of the peptide sequences displayed strong homology with known human autoantigens. However, the peptide inserts of the isolated clones exhibited strong homologies with proteins of pathogens including Mycoplasma sp., Fusobacterium sp., influenza A, and Candida sp., some of which are antigens ubiquitously present in the environment. None of the Ig from 70 other sera cross-reacted with any of the selected phage. Conclusions: We have shown at least some MM paraproteins bind epitopes expressed by known human pathogens, supporting the possible role of chronic antigenic stimulation in myeloma pathogenesis. These findings are consistent with a previously published report (Scand. J. Immunol. 57:583;2003). In addition, the lack of any cross-reacting Igs from the large panel of sera makes it statistically unlikely that any of these pathogens, even if contributory to individual cases, are commonly causative. No significant financial relationships to disclose.