Anti-GRAP2 monoclonal antibody

The ribosomal acidic P0 protein, an essential component of the eukaryotic ribosomal stalk, was found to interact with the helix-loop-helix protein human Grap2 and cyclin D interacting protein (GCIP)/D-type cyclin-interacting protein 1/human homolog of MAID protein. Using in vivo and in vitro binding assays, we show that P0 can interact with the N and C termini of GCIP via its N-terminal 39-114 amino-acid residues. Although the P0-GCIP complex was detected mainly in cytoplasmic fraction, polysome pro. le analysis indicated that the P0-GCIP complex did not coelute with either polysomes or 60S ribosomes, suggesting that GCIP associates with the free form of P0 in the cytoplasm. Transfection of GCIP into MCF-7 cells resulted in decreased levels of pRb phosphorylation. Cotransfection of P0 with GCIP, however, resulted in GCIP-mediated reduction of pRb phosphorylation level which was repressed by P0. Furthermore, overexpression of P0 in breast cancer and hepatocellular cancer cell lines promoted cell growth and colony formation compared to control transfectants. Overexpression of P0 also increased cyclin D1 expression and phosphorylation of pRb at Ser780. Interestingly, P0 mRNA was overexpressed in 12 of 20 pairs of breast cancer/ normal breast specimens (60%). Together, these data indicate that P0 overexpression may cause tumorigenesis in breast and liver tissues at least in part by inhibiting GCIP-mediated tumor suppression.

To demonstrate the utility of phage display in generating highly specific antibodies, affinity selections were conducted on 20 related Src Homology 2 (SH2) domains (ABL1, ABL2, BTK, BCAR3, CRK, FYN, GRB2, GRAP2, LYN, LCK, NCK1, PTPN11 C, PIK3R1 C, PLC gamma 1 C, RASA1 C, SHC1, SH2D1A, SYK N, VAV1 and the tandem domains of ZAP70). The domains were expressed in Escherichia coli, purified and used in affinity selection experiments. In total, 1292/3800 of the resultant antibodies were shown to bind the target antigen. Of the 695 further evaluated in specificity ELISAs against all 20 SH2 domains, 379 antibodies were identified with unique specificity (i.e. monospecific). Sequence analysis revealed that there were at least 150 different clones with 1-19 different antibodies/antigen. This includes antibodies that distinguish between ABL1 and ABL2, despite their 89% sequence identity. Specificity was confirmed for many on protein arrays fabricated with 432 different proteins. Thus, even though the SH2 domains share a common three-dimensional structure and 20-89% identity at the primary structure level, we were able to isolate antibodies with exquisite specificity within this family of structurally related domains.