Fig. 3: Proline-rich motifs interact and downregulate kinase activity.

a Immunoblotting analysis of signaling activity of FGFR2IIIb isoforms. FGFR2^C1; FGFR2^C1Δ34; FGFR2^C2 and FGFR2^C3 were transfected into HEK293T cells starved overnight. The levels of receptor phosphorylation and downstream ERK phosphorylation on each isoform were probed with the indicated antibodies. Tail sequences of FGFR2^C1; FGFR2^C1Δ34; FGFR2^C2 and FGFR2^C3 are shown below. Densitometric bar graph represents 3 independent experiments. The error bars are presented as the standard deviation. b Proline-rich CT inhibits in vitro kinase activity. Recombinant KD-CT^C1, KD-CT^C2, KD-CT^C3 and control clone: KD-CT^C1Δ34 were incubated with ATP/Mg²⁺ at 25 °C and quenched with 100 mM EDTA at different time points as indicated. The activation level was measured using an anti-pY657/658 antibody. Bottom: The densitometric analysis of kinase activity (KD-CT^C1 – blue; KD-CT^C2 – green; KD-CT^C3 – purple and KD-CT^C1Δ34 – red). c The affinities of CT^C1 to JM-KD^pY1 (red) and KD^pY1 (blue) determined using MST. CT^C1 was labeled with Atto488 dye then titrated with unlabeled JM-KD^pY1 and KD^pY1. The error bars represent the standard deviation of 3 technical replications. d HSQC spectra of unbound ¹H-¹⁵N-labeled CT^C1 overlaid with KD^pY1-bound CT^C1 at different ratio (black (0:1) to red (12:1)). Examples of peaks with high chemical-shift perturbations (CSPs) are shown by labels indicating the assignment of given peaks. CSPs chart of ¹⁵N-KD^pY1 titrated by CT^C1 was derived from ¹H-¹⁵N HSQC spectra. Large changes occur on both N-terminal and C-terminal residues of CT^C1. e Wild type GST-CT^C1 and individual P to A mutants, the first 24 residues of CT (CT^C1Δ34), the last 34 residues (CT^L34), CT^C2 and CT^C1 D802F or D802W (to explore the importance of the charged acid group in binding) were used for a GST pulldown experiment with KD^pY1. The symmetric dimerization of KD^pY1 at the concentration range used in this experiment (1 µM) was assumed to have negligible impact on binding of the various CT variants. Densitometric bar graph represents 2 independent experiments.