Figure 3.

Analysis of the interaction between the SH3 domain of the SFKs Src and Lyn and the proline-rich motif of recombinant MBNL1_40–41 and MBNL1_42–43 isoforms. (a) The interaction between SH3 domain of two SFK (Src and Lyn) and the proline-rich motifs of recombinant MBNL1_40–41 and MBNL1_42–43 isoforms. The 0.2 μg of recombinant MBNL1_40–41 or MBNL1_42–43 (as such or heated at 55 °C for 10 min, denat. MBNL1) was incubated with 0.1 μg of Src or Lyn in the absence or presence, respectively, of a proline-rich peptide or the GST-Lyn SH3 domain, for 10 min at 30 °C. Three-fourths of the sample underwent immunoprecipitation with anti-Src or anti-Lyn antibodies, respectively. Immunocomplexes were then probed with anti-P11 or anti-P9 antibodies, and subsequently with anti-Src or anti-Lyn antibodies. The presence of MBNL₄₂ or MBNL₄₃ (right panel), but not of MBNL₄₀ or MBNL₄₁ (left), indicates its association with the two SFKs and suggests that the stretch encoded by exon 5 might have a role in this interaction, possibly through a conformational change that renders the PRMs available to binding. Moreover, competition assays carried out in the absence or presence of a synthetic proline-rich peptide (5′-KGGRSRLPLPPLPPPG-3′) known to interact with the Lyn SH3 domain, or the GST-Lyn/SH3 domain, confirmed that this interaction was mediated by PRMs and the SH3 domain. The remainder of the sample underwent WB analysis (lanes 7 and 8) (b) As this interaction was also a prerequisite for enhancement of SFK activity relative to the basal state, we tested Tyr kinase activity of Src (top panel) and Lyn (bottom) on the Src-specific peptide substrate cdc2 in the absence (lane 1) or presence of increasing concentrations of MBNL1₄₁ (left panels, lanes 2–5) and denat. MBNL1₄₁ (left panels, lanes 6–9) or MBNL1₄₃ (right panels, lanes 2–5) and denat. MBNL1₄₃ (right panels, lanes 6–9), alternatively supplemented with the GST-Lyn SH3 domain (lanes 4 and 5, and 8–9) as described in the Materials and Methods section. (c) As this interaction was also a prerequisite for recognition of MBNL1₄₃ as a substrate by SFKs, we tested whether MBNL1₄₁ or MBNL1₄₃ (each 1 μM) were phosphorylated by Lyn and Src in the absence (solid triangles and solid squares, respectively) or presence (open triangles and open squares, respectively) of the GST-Lyn SH3 domain. At various time points, reactions were stopped and the phosphate incorporated was analyzed after SDS-PAGE on the Cyclone Plus. The amount is expressed as mol P incorporated/mol prot. (values represent the mean of three separate experiments±S.D.). As expected, only MBNL1₄₃ is phosphorylated by Src. The same set of experiments conducted with the MBNL1₄₀ and MBNL1₄₂ isoforms demonstrated that these two proteins behave similarly to MBNL1₄₁ and MBNL1₄₃, respectively (data not shown). Only MBNL1_42–43 isoforms were able to significantly increase Src and Lyn activity, suggesting a novel function of these long isoforms in kinase enzyme regulation. Data are expressed as mean±S.D. from three separate experiments. *P<0.05; ***P<0.001 by Kruskal–Wallis analysis