Potential CD34 interactions with CrkII, Grb2, Shc, C3G, Cbl, paxillin, or Abl.

(A) GST-CD34i_full fusion protein pull-down studies. In an effort to detect associations between CD34 and other signaling intermediates in addition to CrkL, GST-CD34i_fullprecipitation studies were performed, and precipitated proteins were immunoblotted. KG1a cells were stimulated briefly (30 minutes) with 9C5 anti-CD34 (“CD34”) or MOPC 21 isotype control (“Ig”) in adhesion assays, and cells then were washed and lysed. Lysates from 4 × 10⁶ cells per point were precleared with GST-Sepharose, and proteins were precipitated with 20 μg GST-CD34i_full (“CD34i”) or 20 μg GST (“GST”) protein, as indicated. Proteins were separated by SDS-PAGE, transferred to nitrocellulose membrane, and filters were cut according to protein size. Membranes were then immunoblotted with antibodies, as indicated, to determine whether CrkII, Grb2, paxillin, C3G, Cbl, or Shc proteins were present in the protein complex precipitated by GST-CD34i_full. As a positive control for immunoblotting, KG1a lysate was loaded in neighboring lanes, shown at left. Results demonstrate that CrkL protein was selectively precipitated with the GST-CD34i_full fusion protein and that CrkII, Grb2, Shc C3G, Cbl, and paxillin proteins could not be detected in precipitated material. Thus, GST-CD34i_full apparently does not tightly associate with these proteins under conditions employed in our assays. (B) GST-CD34i_full and GST-CrkL pull-down studies. In experiments similar to those described in Figure 5, GST-fusion proteins were used to precipitate c-Abl from KG1a lysates. Abl was identified by immunoblotting with anti-Abl antibody. GST-CD34i_full did not detectably precipitate c-Abl protein; in contrast, Abl was precipitated with GST-CrkL5′SH2-SH3, an important positive control for our studies.