Activity and selectivity of KH-CB19
(A) Binding of KH-CB19 to kinase catalytic domains as assessed in thermal shift assays against 129 human protein kinases (see Table S1 for screened targets). Targets that showed significant temperature shifts are highlighted by blue spheres.
(B) Chemical structures for KH-CB19 and KH-CB20.
(C) Correlation between binding constants determined by AMBIT and temperature shift data. The correlation of the linear least-squares fit was 0.95. One outlier is highlighted (red circle) which corresponds to BIRB-796, a potent p38 inhibitor. This compound showed only weak association in the AMBIT assay possibly due to slow binding kinetics, but gave a high T_m shift as expected for this potent inhibitor. The kinome phylogenetic tree has been used and modified with permission of Cell Signaling Technology.

Induced Fit of the P Loop upon Inhibitor Binding
Shown are superimpositions of CLK1 (A) and CLK3 (B) inhibitor complexes. The different cocrystallized ligands are indicated in the figure. Inhibitor molecules and the P loop phenylalanine (F172 and F167 in CLK1 and CLK3, respectively) are shown in stick representation. The induced structural changes are indicated by an arrow.

Synthetic Route
(A) Lead alkaloid Bauerine C.
(B) Synthetic route for the preparation of the studied inhibitors.

Binding Mode of CLK Inhibitors
(A) Overview of the CLK cocrystal structure. The CLK1 catalytic domain is shown as a ribbon diagram and the ATP binding site has been highlighted in surface representation. Details of the interaction of KH-CB19 with the kinase active site are shown in the detailed view on the right.
(B) Superimposition of the CLK3 cocrystal structure with KH-CB19 and the triazole diamine K00546.
(C) Superimposition of CLK1 complexes with KH-CB19 and hymenialdisine (K0010).
(D) Electron density (2Fc-2Fo) map of KH-CB19 in the CLK1 complex. The map has been contoured using a 2 sigma cutoff.

Influence of CLK Inhibitors on the Phosphorylation State of SR Proteins and mRNA Expression of flTF and asHTF in HMEC-1 Cells
(A) SR protein phosphorylation is shown in resting or TNF-α-induced HMEC-1 cells 2 min post TNF-α stimulation. SR protein family members SRp75 (75 kDa), SRp55 (55 kDa), SRp40 (40 kDa), SC35 (35 kDa), SF2/ASF (30 kDa), and SRp20 (20 kDa) were detected using phosphorylation-dependent anti-SR protein antibodies. Displayed are nonstimulated cells with or without pretreatment with KH-CB19 (10 μM) or TG003 (10 μM) and TNF-α-stimulated cells (10 ng/ml) pretreated with the same CLK inhibitors. The results are representative of at least three independent experiments.
(B) Expression of mRNA encoding flTF (931 bp) and asHTF (771 bp) in untreated cells (control), TNF-α−induced HMEC-1 (TNF-α), and cells pretreated with 10 μM compound KH-CB19 or TG003 with or without TNF-α 1 hr post stimulation. GAPDH was used as a loading control. A 100 bp DNA ladder was used as a marker. The panel shown is representative of at least three independent experiments.
(C) Dose-dependent reduction of phosphorylation of SRp75 and SRp55 and quantitation of the detected phosphorylation (lower panels). GAPDH was used as a loading control. Data were expressed as mean ± SEM using three independent experiments.