Fig. 4: CT binding to KD^pY1 disrupts the formation of asymmetric dimer.

a Dimerization status of FGFR2 kinases in the presence or absence of CT. KD^pY1 (red curve) includes a dominant population of dimers at 60 µM in SEC (HiPrep Sephacryl 26/60 S-100 HR), whereas the unphosphorylated KD (black curve) prevails as a monomer. The mono-phosphorylated KD^pY1-CT construct (green curve) also exists as a monomer in solution. b Phosphorylation of JM-KD affects CT binding. GST-CT^C1 was used to pulldown the progressively increasing phosphorylation of JM-KD constructs. The pulldown indicates that JM-KD^pY1 was the highest affinity binding partner for GST-CT^C1. This interaction is verified using SPR. See Supplementary Table 5 for the affinity measurements. c Dimerization of FGFR2 by FLIM analysis of the FRET between the FGFR2-GFP and FGFR2-RFP. First panel: Reference lifetime measurements between FGFR2-GFP and RFP-alone, control for no interaction. The mean lifetime is centered around ∼2.1 ns (dark blue), which corresponds to the mean lifetime for isolated CFP alone. Second panel: Dimerization of C1 showing a measurable left shift with of the molecule showing FRET above the control. Note that most interactions are seen in the intracellular vesicles. Third panel: Dimerization of C2. 16% of molecules on plasma membrane showing dimerization above the control threshold (orange). Fourth panel: Dimerization of C3, as with C1, 26% of the molecules are showing interaction (orange) however unlike C1, almost all of the interactions are on the plasma membrane. Inserts with arrows showing exquisite separation of dimeric and non-dimeric FGFR2-C3 on the plasma membrane. Representative of 3 independent experiments. 10–15 cell images were taken for each isoform in each independent experiment. Scale bar = 25 µm. d GST-CT^C1 was used to pull down five mono-phosphorylated constructs of JM-KD^pY1 with progressively truncated JM (JM-KD^pY1, JM^Δ15-KD^pY1, JM^Δ25-KD^pY1, JM^Δ35-KD^pY1, and KD^pY1). The presence of the intact JM enhances the interaction with GST-CT^C1. Densitometric bar graph represents 2 independent experiments. e A GST-CT^C1 pulldown of different FGFR2IIIb mono-phosphorylated proteins that include the presence or absence of JM and/or CT (KD^pY1; JM-KD^pY1; KD^pY1-CT, and JM-KD^pY1-CT), shows that the presence of JM, but not CT, promotes the interaction between kinase ___domain and GST-CT^C1. The presence of CT inhibits the GST-CT^C1 interaction, indicating CT binds through an intramolecular interaction. f A MBP-JM pulldown of different FGFR2IIIb mono-phosphorylated proteins (as described in Fig. 4e), shows that the presence of JM does not block JM binding suggesting that JM of one protomer binds to the other in the mono-phosphorylated dimers (previously identified for KD and JM-KD). The latch to the protomer in the asymmetric dimer leaves an available JM binding site. The presence of CT (in KD-CT^C1 and JM-KD^pY1-CT^C1) reduces JM binding. Densitometric bar graph represents 2 independent experiments. g MST measurement of JM binding to CT. A two-fold serial dilution of CT was titrated into JM which was labeled with Atto 488. The error bars represent standard error of the mean from 3 technical replications. h NMR titration of JM titrated into ¹⁵N-labeled CT using a red-to-white gradient, where white represents the weakest CSP and red depicts the strongest CSP. Proline residues are not visible in this experiment (shown in gray).