Importantly, CD5 was one of only two proteins identified with at least Palbociclib mw 2 peptides specifically in the VLR32-containing IP compared to the ‘minus-VLR32’ negative control.

The other identified protein, myosin-9 was only identified with 2 spectra from 2 unique peptides, corresponding to a sequence coverage of only 1%. Using this approach we determined that VLR32 immunoprecipitations contained the CD5 antigen (Table 2). We verified these results in parallel experiments testing the reactivity of VLR32 with cells transfected with CD5–GFP fusion constructs, by western blot analysis and immunoprecipitation experiments. VLR32, but not the negative control VLR4, was able to detect CD5–GFP fusion proteins in cell lysates from transiently transfected HEK293T cells (Fig. 3A). This find more reactivity was limited to lysates separated under non-reducing conditions as separation of cell lysates under reducing conditions abolished VLR32 binding (data not shown). In additional experiments, we demonstrated that VLR32 but not the negative control VLR4

precipitated CD5–GFP fusion proteins from cell lysates of transiently transfected HEK293T cells (Fig. 3B) and that VLR32 but not VLR4 stained HEK293T cells transfected with CD5 expression constructs in flow cytometry experiments (Fig. 3C). These experiments demonstrate the CD5-specificity of VLR32. Prior studies of VLR antibodies suggest that binding of the antibody to the antigen is avidity-based and that the affinity of

the individual antigen-binding unit to the antigen is often comparatively low (Herrin et al., 2008 and Kirchdoerfer et al., 2012). To investigate the affinity versus avidity-based binding of VLR32 to the CD5 antigen, we generated monomeric VLR32 antibodies by deleting the C-terminal 42 residues of the VLR antibody. As triclocarban expected, the resulting individual VLR units displayed a slightly faster migration pattern compared to full-length VLR proteins (Fig. 3B). Only the multimeric VLR32 was able to bind to CD5 efficiently as shown for immunoprecipitation (Fig. 3B) and flow cytometry analyses (Fig. 3D). VLR binding was not detected by flow cytometry using the monomeric VLR32 (Fig. 3D) and only a weak signal was obtained for immunoprecipitated CD5–GFP using the monomeric VLR32 (Fig. 3B). These data indicate an avidity-based contribution to the binding of the VLR32 lamprey antibody to human CD5. In this study, we demonstrate for the first time that monoclonal lamprey VLR antibodies can be used for purification and mass spectrometry-based identification of cell surface expressed protein antigens. Unlike conventional immunoglobulin-based antibodies, VLR antibodies utilize their leucine-rich repeats as basic structural units, resulting in a fundamentally different protein architecture of antigen receptors.