Development of a novel AAK1 inhibitor via Kinobeads-based screening

Abstract

A chemical proteomics approach using Ca²⁺/calmodulin-dependent protein kinase kinase (CaMKK) inhibitor-immobilized sepharose (TIM-063-Kinobeads) identified main targets such as CaMKKα/1 and β/2, and potential off-target kinases, including AP2-associated protein kinase 1 (AAK1), as TIM-063 interactants. Because TIM-063 interacted with the AAK1 catalytic domain and inhibited its enzymatic activity moderately (IC₅₀ = 8.51 µM), we attempted to identify potential AAK1 inhibitors from TIM-063-derivatives and found a novel AAK1 inhibitor, TIM-098a (11-amino-2-hydroxy-7H-benzo[de]benzo[4,5]imidazo[2,1-a]isoquinolin-7-one) which is more potent (IC₅₀ = 0.24 µM) than TIM-063 without any inhibitory activity against CaMKK isoforms and a relative AAK1-selectivity among the Numb-associated kinases family. TIM-098a could inhibit AAK1 activity in transfected cultured cells (IC₅₀ = 0.87 µM), indicating cell-membrane permeability of the compound. Overexpression of AAK1 in HeLa cells significantly reduced the number of early endosomes, which was blocked by treatment with 10 µM TIM-098a. These results indicate TIM-063-Kinobeads-based chemical proteomics is efficient for identifying off-target kinases and re-evaluating the kinase inhibitor (TIM-063), leading to the successful development of a novel inhibitory compound (TIM-098a) for AAK1, which could be a molecular probe for AAK1. TIM-098a may be a promising lead compound for a more potent, selective and therapeutically useful AAK1 inhibitor.

Figure 1

Identification and characterization of AAK1 as a TIM-063–interactant. (a) Protocol for identification of TIM-063–interactants using TIM-063–immobilized sepharose (TIM-127-sepharose) combined with mass spectrometry analysis. Mouse cerebrum extracts were applied onto TIM-127-sepharose, and after extensive washing of resin, TIM-063–interactants were eluted by the addition of 100 µM TIM-063 containing buffer, followed by identification via mass spectrometry analysis as described in the “Methods”. (b) Eluates from TIM-127-sepharose (TIM-127–S) or a control sepharose without the inhibitor (Control–S) as described in (a) were analyzed by immunoblotting with either anti-CaMKKβ/2 antibody (upper panel), anti-AAK1 antibody (middle panel), or anti-ERK1/2 antibody (bottom panel) together with mouse cerebrum extracts (Input). (c) Extracts from COS-7 cells expressing His-tagged AAK1 wild type (1–863) or His-tagged AAK1 catalytic domain (25–396) were incubated with either TIM-127-sepharose or control sepharose, and after extensive washing of resin, TIM-063–interactants were eluted by the addition of 100 µM TIM-063–containing buffer, followed by immunoblot analysis with an anti-AAK1 antibody as described in the “Methods”. An arrow and asterisk indicate His-tagged AAK1 wild type (1–863) and His-tagged AAK1 catalytic domain (25–396), respectively. (d) Phosphorylation of GST-AP2µ2 (145–162) at Thr156 by the His-tagged AAK1 catalytic domain (25–396) for various time periods (0–120 min) was detected by immunoblot analysis with an anti-phospho-AP2µ2 (pThr156) antibody (insert), quantitated by densitometric scanning of the immunoreactive bands, and expressed as a percentage of the value at 120 min of the reaction. An asterisk indicates phospho-Thr156 in GST-AP2µ2 (145–162). (e) Inhibition of AAK1 activity by TIM-063 in a dose-dependent manner. Phosphorylation of GST-AP2µ2 (145–162) by His-tagged AAK1 catalytic domain (25–396) was measured in the presence or absence of various concentrations of TIM-063 at 30 °C for 20 min using 100 µM [γ-³²P]ATP as described in the “Methods”. AAK1 activities are expressed as a percentage of the average value in the absence of the compound. Results represent duplicate experiments. Chemical structure of TIM-063 is indicated (insert). Molecular mass markers (kDa) are indicated in the left lanes of immunoblot panels.

Figure 2

Characterization of TIM-063 derivatives as AAK1 inhibitors. (a) Protein kinase activities of His-tagged AAK1 catalytic domain (25–396) were measured at 30 °C for 20 min using 100 μM [γ-³²P]ATP in the presence of 26 µM TIM-063 or 2.6 µM of its derivatives (TIM-055 − TIM-106) or absence as described in the “Methods”. AAK1 activities are quantitated and expressed as a percentage of the average value in the absence of the compound (–). Results represent duplicate experiments. (b) Protein kinase activities of CaMKKα/1 (left panel) and β/2 (right panel) were measured at 30 °C for 20 min using 100 µM [γ-³²P]ATP, and 2 mM CaCl₂ with 6 µM CaM in the presence of 2.6 µM TIM-063 or its derivatives (TIM-088 − TIM-106) or absence as described in the “Methods”. CaMKK activities are quantitated and expressed as a percentage of the average value in the absence of the compound (−). Results represent duplicate experiments.

Figure 3

Synthesis of TIM-063 analogs (a) and TIM-098a (b).

Figure 4

Effect of TIM-098a on AAK1, NAKs, and CaMKK activities. (a) Protein kinase activities of His-tagged AAK1 catalytic domain (25–396) were measured at 30 °C for 20 min using 100 μM [γ-³²P]ATP in the presence of various concentrations of TIM-098a or absence as described in the “Methods”. AAK1 activities are quantitated and expressed as a percentage of the average value in the absence of the compound. Results represent duplicate experiments. Chemical structure of TIM-098a is indicated (insert). (b) IC₅₀ value of TIM-098a against NAKs activities. Activities of NAKs, including AAK1, GAK, BIKE and STK19, were measured in the presence of various concentrations of TIM-098a in a solution described in (a) and the “Methods”. IC₅₀ values were estimated from duplicate experiments (see Supplemental Information S3). (c) Activities of CaMKKα/1 (left panels) and β/2 (right panels) were measured at 30 °C for 10 min in the presence of 2.6 µM TIM-063 or TIM-098a or absence (−) via immunoblot analysis using anti-phospho-CaMKIα (pThr177) antibody (upper panels) as described in the “Methods”. CaMKK activities are quantitated by densitometric scanning of the immunoreactive band and expressed as a percentage of the average value in the absence of the compound (−). Results are represented as the mean ± the standard deviation indicated by error bars from triplicate experiments. Molecular mass markers (kDa) are indicated in the left lanes of immunoblot panels. Statistical differences are marked as **p < 0.05; n.s., not significant vs. CaMKK activities in the absence of compounds (−). (d) Energy minimized (Macromodel) Glide top scored docking poses of (A) TIM-063 and (B) TIM-098a with AAK1. (C) Overlay of (A) and (B).

Figure 5

Inhibition of AAK1 activity by TIM-098a in cells. (a) COS-7 cells were transfected with an expression vector of GST-AP2µ2 (145–162) in the presence (+) or absence (−) of the expression vector of His-tagged AAK1 catalytic domain (25–396). After 42 h culture, the cells were cultured with or without (−) the indicated concentrations of TIM-098a for 6 h. Cell extracts were analyzed by immunoblotting using an anti-phospho-AP2µ2 (pThr156) antibody (upper panel), anti-GST antibody (middle panel), or anti-AAK1 antibody (lower panel). The molecular mass markers (kDa) are indicated in the left lanes of the immunoblot panels. (b) AAK1 activity was quantified by densitometric scanning of the immunoreactive band of pThr156 (a, upper panel) and expressed as a percentage of the average value in the absence of the compound (−). The results are representative of duplicate experiments.

Figure 6

Effect of TIM-098a on AAK1 overexpression-reduced early endosome formation. HeLa cells were transfected with an expression vector containing His-tagged AAK1 (1–863). After 42 h of culture, cells were incubated with 10 µM TIM-098a-containing medium or without (DMSO) for 6 h. (a) Cells were fixed and immunostained with anti-AAK1 antibody (red) and anti-EEA1 antibody (green), followed by nuclear staining with Hoechst 33342 (blue). The right panels show the merged images. Scale bars = 10 µm (b) The number of EEA1 vesicles were quantified in 54–169 cells per condition using the ImageJ software as described in the “Methods”. The data are represented as the mean ± the standard deviation indicated by error bars. Statistical differences are marked as **p < 0.05; n.s., not significant vs. untransfected cells without TIM-098a treatment.