1. Neuroscience

Discovery and characterization of a specific inhibitor of serine-threonine kinase cyclin-dependent kinase-like 5 (CDKL5) demonstrates role in hippocampal CA1 physiology

  1. Anna Castano
  2. Margaux Silvestre
  3. Carrow I Wells
  4. Jennifer L Sanderson
  5. Carla A Ferrer
  6. Han Wee Ong
  7. Yi Lang
  8. William Richardson
  9. Josie A Silvaroli
  10. Frances M Bashore
  11. Jeffery L Smith
  12. Isabelle M Genereux
  13. Kelvin Dempster
  14. David H Drewry
  15. Navlot S Pabla
  16. Alex N Bullock
  17. Tim A Benke  Is a corresponding author
  18. Sila K Ultanir  Is a corresponding author
  19. Alison D Axtman  Is a corresponding author
  1. Department of Pharmacology, University of Colorado School of Medicine, United States
  2. Kinases and Brain Development Laboratory, The Francis Crick Institute, United Kingdom
  3. Structural Genomics Consortium, UNC Eshelman School of Pharmacy, University of North Carolina at Chapel Hill, United States
  4. Centre for Medicines Discovery, Nuffield Department of Medicine, University of Oxford, United Kingdom
  5. Division of Pharmaceutics and Pharmacology, College of Pharmacy and Comprehensive Cancer Center, The Ohio State University, United States
  6. Lineberger Comprehensive Cancer Center, School of Medicine, University of North Carolina at Chapel Hill, United States
  7. Departments of Pediatrics, Pharmacology, Neurology and Otolaryngology, University of Colorado School of Medicine, United States
https://doi.org/10.7554/eLife.88206
Share this article
Cite this article
  1. Anna Castano
  2. Margaux Silvestre
  3. Carrow I Wells
  4. Jennifer L Sanderson
  5. Carla A Ferrer
  6. Han Wee Ong
  7. Yi Lang
  8. William Richardson
  9. Josie A Silvaroli
  10. Frances M Bashore
  11. Jeffery L Smith
  12. Isabelle M Genereux
  13. Kelvin Dempster
  14. David H Drewry
  15. Navlot S Pabla
  16. Alex N Bullock
  17. Tim A Benke
  18. Sila K Ultanir
  19. Alison D Axtman
(2023)
Discovery and characterization of a specific inhibitor of serine-threonine kinase cyclin-dependent kinase-like 5 (CDKL5) demonstrates role in hippocampal CA1 physiology
eLife 12:e88206.
https://doi.org/10.7554/eLife.88206
  2. Download BibTeX
  3. Download .RIS
14 figures, 4 tables and 1 additional file

Figures

Figure 1
Download asset Open asset
Screening of CDKL5 inhibitors in rat primary neurons using western blotting.

Western blots showing expression of total EB2 and levels of Ser222 EB2 phosphorylation in DIV14-16 rat primary neurons upon treatment of 1 hr with 5 nM, 50 nM, 500 nM, and 5000 nM of selected CDKL5 inhibitors.

Figure 1—source data 1

Source data contains raw blots.

https://cdn.elifesciences.org/articles/88206/elife-88206-fig1-data1-v2.zip
Figure 2
Download asset Open asset
Structures of CDKL5 inhibitor leads and corresponding parent compounds.
Figure 3 with 3 supplements
Download asset Open asset
Kinome-wide selectivity data for CDKL5 lead compounds.

Kinome tree diagrams illustrate the selectivity of these compounds when profiled against 403 wild-type (WT) human kinases at 1 µM at Eurofins DiscoverX in their scanMAX panel. Each red circle represents a kinase that binds with high affinity (percent of control [PoC] <10) to the compound being assayed: (A) B1, (B) B4, and (C) B12. These kinases have been labeled for clarity. The percent control legend shows red circles of different sizes corresponding with percent control value each kinase binds the small molecule in this large binding panel. Also included are selectivity scores (S10 [1 µM]), which were calculated using the PoC values for WT human kinases in the scanMAX panel only. The S10 score is a way to express selectivity that corresponds with the percent of the kinases screened that bind with a PoC value <10. In the embedded tables, kinases are listed by their gene names and ranked by their PoC value generated in the scanMAX panel. Rows colored green demonstrate enzymatic IC50 values within a 30-fold window of the CDKL5 binding IC50 value. A binding assay was run only for CDKL5, all other IC50 values in the penultimate column of the nested tables were generated using an enzymatic assay (Figure 3—figure supplements 1 and 2).

Figure 3—figure supplement 1
Download asset Open asset
Curves corresponding with CDKL5 affinity measurements in Figure 3.

(A) Curves generated in CDKL5 split-luciferase assay for B1, B4, and B12. Error bars represent SD. IC50 values with error included are B1 IC50 = 6.7 ± 0.9 nM, B4 IC50 = 6.4 ± 0.4 nM, B12 IC50 = 8.7 ± 0.6 nM. (B) Curves generated in CDKL5 NanoBRET assay for B1, B4, and B12. Assays were run in singlicate (n = 1).

Figure 3—figure supplement 2
Download asset Open asset
Curves corresponding with GSK3 NanoBRET measurements in Figure 3.

(A) Curves generated in GSK3α and GSK3β NanoBRET assays for B1. Assays were run in singlicate (n = 1). (B) Curves generated in GSK3α and GSK3β NanoBRET assays for B4. Assays were run in singlicate (n = 1). (C) Curves generated in GSK3α and GSK3β NanoBRET assays for B12. Assays were run in singlicate (n = 1). (D) Curves generated in CDK9, CDK16, CDK17, and CDK18 NanoBRET assays for B1. Error bars represent SD.

Figure 3—figure supplement 3
Download asset Open asset
CDKL family selectivity evaluation via thermal shift and binding/enzymatic assays.

CDKL1–4 were evaluated in enzymatic assays, while CDKL5 was evaluated using a binding assay. Error bars represent SD.

Figure 4
Download asset Open asset
B1 potently inhibits the kinase activity of CDKL5.

B1 was simultaneously assayed with other CDKL5 antagonists (Ong et al., 2023). (A) Representative graph from a CDKL5 kinase assay demonstrating that purified WT human CDKL5 retains kinase activity, while the kinase dead (KD, CDKL5 K42R) human protein is functionally inactive (n = 3). (B) Representative western blot illustrating that equal expression of WT and KD proteins was observed in the kinase assay experiments. (C, D) Kinase assays using purified WT human CDKL5 in the presence of 10 or 100 nM of the indicated compounds (n = 3). One-way ANOVA with Dunnett’s multiple-comparison test used. ***p<0.0001 and nonsignificant comparisons not shown. Control, DMSO, positive (AST-487), and negative (Lapatinib) controls are as in Ong et al., 2023.

Figure 4—source data 1

Source data contains raw blots.

https://cdn.elifesciences.org/articles/88206/elife-88206-fig4-data1-v2.zip
Figure 5 with 2 supplements
Download asset Open asset
CAF-382 (B1), HW2-013 (B4), and LY-213 (B12) compounds reduce CDKL5 activity in rat primary neurons.

HW2-013 (B4) and LY-213 (B12) also downregulate GSK3 activity. (A) Western blot and quantification showing expression of EB2 phosphorylation and β-catenin phosphorylation in DIV14-15 rat primary neurons after an hour treatment with different concentrations of CAF-382 (B1). (B) Western blot and quantification showing expression of EB2 phosphorylation and β-catenin phosphorylation in DIV14-15 rat primary neurons after an hour treatment with different concentrations of HW2-013 (B4). (C) Western blot and quantification showing expression of EB2 phosphorylation and β-catenin phosphorylation in DIV14-15 rat primary neurons after an hour treatment with different concentrations of LY-213 (B12). Each concentration was compared to the control using a Kruskal–Wallis test. n = 3 biological replicates with two repetitions. *p≤0.05; **p≤0.01; ***p≤0.001. Error bars are SD.

Figure 5—source data 1

Source data contains raw blots.

https://cdn.elifesciences.org/articles/88206/elife-88206-fig5-data1-v2.zip
Figure 5—figure supplement 1
Download asset Open asset
CDKL5, total EB2, and β-catenin expression in rat primary neurons after treatment with CAF-382 (B1), HW2-013 (B4), and LY-213 (B12) compounds.

(A, D, G) Quantification of CDKL5 expression in DIV14-15 rat primary neurons after an hour treatment with different concentrations of CAF-382 (B1), HW2-013 (B4), and LY-213 (B12) compounds, respectively. Each concentration was compared to the control using one-way ANOVA. N = 3 biological replicates with two repetitions. Error bars are SD. (B, E, H) Quantification of total EB2 expression in DIV14-15 rat primary neurons after an hour treatment with different concentrations of CAF-382 (B1), HW2-013 (B4), and LY-213 (B12) compounds, respectively. Each concentration was compared to the control using one-way ANOVA. N = 3 biological replicates with two repetitions. Error bars are SD. (C, F, I) Quantification of total β-catenin expression in DIV14-15 rat primary neurons after an hour treatment with different concentrations of CAF-382 (B1), HW2-013 (B4), and LY-213 (B12) compounds, respectively. Each concentration was compared to the control using one-way ANOVA. N = 2 biological replicates with two repetitions. Error bars are SD.

Figure 5—figure supplement 1—source data 1

Source data contains raw blots.

https://cdn.elifesciences.org/articles/88206/elife-88206-fig5-figsupp1-data1-v2.zip
Figure 5—figure supplement 2
Download asset Open asset
Phosphorylation of β-catenin, a substrate of GSK, is reduced upon treatment with a GSK inhibitor CHIR 99021 (Tocris) but EB2 phosphorylation is not changed.

Western blot showing expression of total EB2, pS222 EB2, total β-catenin, phospho-β-catenin, and tubulin in DIV14-15 rat primary neurons after an hour treatment with the GSK inhibitor.

Figure 6 with 1 supplement
Download asset Open asset
CAF-382 (B1) reduced phosphorylation of EB2 in CA1 hippocampal slices.

(A) Example blots demonstrate no alterations in CDKL5 expression across treatments (stats in text) and suggest relative CAF-382 (B1) dose-dependent differences in EB2 phosphorylation. (B) Normalized quantification of EB2-phosphorylation (density of EB2-phosphorylation bands/ density of EB2-total bands). CAF 382 (B1) (control: 1 ± 0.10 vs. 100 nM: 0.49 ± 0.02, n = 8, p=0.002, RM-ANOVA) (control: 1 ± 0.08 vs. 45 nM: 0.16 ± 0.03, n = 10, p<0.001, RM-ANOVA) (control: 1 ± 0.14 vs. 10 nM: 0.62 ± 0.11, n = 7, p=0.003, RM-ANOVA) reduced EB2 phosphorylation at all concentrations. *p≤0.05; **p≤0.01; ***p≤0.001. Error bars are SE.

Figure 6—source data 1

Source data contains raw blots.

https://cdn.elifesciences.org/articles/88206/elife-88206-fig6-data1-v2.zip
Figure 6—source data 2

Source data contains raw blots.

https://cdn.elifesciences.org/articles/88206/elife-88206-fig6-data2-v2.zip
Figure 6—source data 3

Source data contains raw blots.

https://cdn.elifesciences.org/articles/88206/elife-88206-fig6-data3-v2.zip
Figure 6—figure supplement 1
Download asset Open asset
SNS-032 (Selleck Chemicals) inhibits phosphorylation of EB2 in rat hippocampal slices in a time- and concentration-dependent fashion.

Effect of SNS-032 is compared to control slices with similar concentration of DMSO vehicle under three different conditions. Compare to Figure 6.

Figure 7
Download asset Open asset
CAF-382 (B1) reduced postsynaptic fEPSP responsiveness, as measured by input–output curves.

(A) Sample input (fiber volley)/output (fEPSP slope) (I/O) curves at baseline, 30 min and 60 min after CAF-382 (B1) (100 nM) in a hippocampal slice. Slopes of I/O curves were obtained at baseline, and 30 min and 60 min after treatment with CAF-382 (B1). Numbers in (A) indicate sample traces in (B) from treated slice; stimulus artifact (arrow) has been removed for clarity. Initial negative deflection after the artifact is the fiber volley, followed by the fEPSP. (C) RM-ANOVA of I/O slopes of CAF-382 (B1) (100 nM and 45 nM) were significantly decreased after 30 min (100 nM: 1.91 ± 0.23, n = 9, p=0.006; 45 nM: 1.58 ± 0.24, n = 10, p=0.013) and 60 min (100 nM: 2.00 ± 0.30, n = 9, p=0.001; 45 nM: 1.57 ± 0.22, n = 10, p=0.009) compared to baseline (100 nM: 3.24 ± 0.52, n = 9; 45 nM: 2.29 ± 0.23, n = 10). Baseline I/O slope was used to normalize each recording to allow comparisons across all treatments. CAF-382 (B1) (10 nM) did not alter I/O slopes after 30 or 60 min (RM-ANOVA, n = 9, p=0.07). *p≤0.05; **p≤0.01; ***p≤0.001. Error bars are SE. fEPSP, field excitatory postsynaptic potentials.

Figure 8
Download asset Open asset
CAF-382 (B1) did not alter presynaptic release as measured by paired-pulse ratio (PPR).

(A) PPRs of fEPSPs (slope of fEPSP2/Slope of fEPSP1), reflective of presynaptic release, were unaltered across a range of stimulus intervals (RM-ANOVA, n = 7) by CAF-382 (B1) (100 nM) for 60 min compared to baseline. (B) Sample trace for 50 ms interval. Reduction of initial fEPSP slopes were consistent with Figure 7. Stimulus artifacts (arrows) have been removed for clarity. Error bars are SE. fEPSP, field excitatory postsynaptic potentials.

Figure 9
Download asset Open asset
CAF-382 (B1) did not alter NMDAR-mediated component of fEPSPs.

Maximal fEPSPs recorded in stratum pyramidale of CA1 with reduced extracellular Mg2+ demonstrate multiple population spikes; previous studies have demonstrated that secondary population spikes in these conditions are NMDAR-dependent (Coan and Collingridge, 1985). CAF-382 (B1) (100 nM) did not alter population spike 1 (control –4.057 ± 0.584 mV, CAF-382 (B1) 3.929 ± 0.736 mV, n = 10, p=0.742, RM-ANOVA) or NMDAR-mediated population spike 2 (control –2.093 ± 0.568 mV, CAF-382 (B1) –1.656 ± 0.4442, n = 10, p=0.867, RM-ANOVA). For comparison, the selective NMDAR antagonist D-APV (50 μM) completely blocked population spike 2, as previously reported (Coan and Collingridge, 1985). Stimulation artifact has been removed (arrow) for clarity. fEPSP, field excitatory postsynaptic potentials.

Figure 10
Download asset Open asset
CAF-382 (B1) acutely reduces the expression of long-term potentiation (LTP) in CA1 hippocampus.

Hippocampal slices were initially incubated in CAF-382 (B1) for at least 30 min and then continuously perfused with CAF-382 (B1) (100 nM). fEPSP slope was normalized to initial baseline, expressed as 100% (dotted line); mean ± SEM shown. Following stable baseline of fEPSP slope (>20 min), LTP was induced by two theta-burst (TBS) trains. Compared to interleaved control slices, CAF-382 (B1) significantly reduced LTP as measured by fEPSP slope at 56–60 min post theta-burst (control: 115.6 ± 2.2 n = 13 slices, 12 rats; CAF-382 (B1): 106.4 ± 3.5, n = 5 slices, four rats; p=0.029, two-way ANOVA, Holm–Sidak post hoc). Sample traces for control (black filled circle) and CAF-382 (B1) (red filled circle) before (1, solid trace) and after LTP (2, dotted trace) are shown to the right. Error bars are SE. fEPSP, field excitatory postsynaptic potentials.

Figure 11
Download asset Open asset
CAF-382 (B1) acutely reduces the expression of long-term potentiation (LTP) mediated by AMPA-type glutamate receptors in CA1 hippocampus.

Hippocampal slices were initially incubated in CAF-382 (B1) (100 nM) for at least 60 min prior to recording. Peak negative current of AMPA-type glutamate receptor-mediated synaptic responses were normalized to initial baseline post break-in (time = 0), expressed as 100% (dotted line); mean ± SEM shown. Following baseline, LTP was induced by a pairing protocol. Compared to interleaved control slices, CAF-382 (B1) significantly reduced LTP at 27–31 min post break-in (control: 197.5 ± 8.4 n = 14 slices, 14 rats; CAF-382 (B1): 123.14 ± 9.0, n = 12–13 slices, 13 rats; p<0.001, two-way ANOVA, Holm–Sidak post hoc). Sample traces for control (black filled circle) and CAF-382 (B1, red filled circle) before (1, solid trace) and after LTP (2, dotted trace) are shown to the right. Error bars are SE.

Figure 12
Download asset Open asset
Snapshot of pharmacokinetic (PK) results for B1.

(A) Plasma concentration measurements at timepoints post-administration. (B) Plot of mean plasma concentration over the time course of the PK study. (C) Summary of B1 PK properties (n = 2). Values are expressed as the mean calculated after two animals were dosed. T1/2, half-life; Tmax, time of maximum observed concentration; Cmax, maximum observed concentration, occurring at Tmax (if not unique, then the first maximum is used); AUClast, area under the curve from the time of dosing to the last measurable positive concentration; AUCinf, area under the curve from dosing time extrapolated to infinity, based on the last observed concentration; AUC_%Extrap_obs, percentage of AUCinf due to extrapolation from last time point to infinity; MRTinf_obs, mean residence time extrapolated to infinity for a substance administered by intravascular dosing using observed last concentration; AUClast/D, area under the curve from the time of dosing to the last measurable concentration divided by the dose; F, bioavailability.

Figure 13 with 1 supplement
Download asset Open asset
Scheme 1: preparation of compound B1.

(a) EtOH, NaBH4, acetone, 80°C, 1 hr, 57% yield; (b) HATU, DIPEA, 1-(tert-butoxycarbonyl)piperidine-4-carboxylic acid, DMF; (c) TFA, CH2Cl2, 36% yield over two steps. 5-(((5-isopropyloxazol-2-yl)methyl)thio)thiazol-2-amine (Int 1). To a flask was added 5-thiocyanatothiazol-2-amine (400 mg, 1.0 equiv, 2.5 mmol) in EtOH (24 mL), followed by addition of NaBH4 (183 mg, 1.9 equiv, 4.8 mmol) portion-wise at room temperature. The mixture was stirred for 1 hr at room temperature then acetone (12 mL) was added. After 1 hr, a solution of 2-(chloromethyl)–5-isopropyloxazole (412 mg, 1.0 equiv, 2.6 mmol) in EtOH (4 mL) was added, and the reaction was heated at 80°C for 1 hr. The resulting mixture was cooled, concentrated in vacuo, and then partitioned between EtOAc and brine. The organic phase was separated, dried with MgSO4, and concentrated in vacuo to give a crude solid. The crude material was triturated with diethyl ether/hexane to provide the desired product 4-(((5-isopropyloxazol-2-yl)methyl)thio)thiazol-2-amine as a red oil (368 mg, 57 %). N-(5-(((5-isopropyloxazol-2-yl)methyl)thio)thiazol-2-yl)piperidine-4-carboxamide 2,2,2-trifluoroacetate (B1). To a flask was added 1-(tert-butoxycarbonyl)piperidine-4-carboxylic acid (99 mg, 1.1 equiv, 431 μmol) and DIPEA (205 μL, 3 equiv, 1.2 mmol) and HATU (194 mg, 1.3 equiv, 509 μmol) and the flask was stirred at room temperature for approximately 15 min. Next, 5-(((5-isopropyloxazol-2-yl)methyl)thio)thiazol-2-amine (100 mg, 1.0 equiv, 392 μmol) was added and the reaction was stirred at room temperature for 16 hr. The reaction mixture was concentrated and purified by column chromatography (SiO2, MeOH 0–10% in CH2Cl2) to yield the desired product as a white solid to yield the intermediate tert-butyl 4-((5-(((5-isopropyloxazol-2-yl)methyl)thio)thiazol-2-yl)carbamoyl)piperidine-1-carboxylate. To this intermediate was added 20% TFA and CH2Cl2 (5 mL) and the reaction stirred for 1 hr. The reaction was concentrated in vacuo and after addition of MeOH a white precipitate crashed out. The solid was filtered under vacuum to yield the desired product N-(5-(((5-isopropyloxazol-2-yl)methyl)thio)thiazol-2-yl)piperidine-4-carboxamide 2,2,2-trifluoroacetate as a white solid (20 mg). The remaining crude material was purified by preparative HPLC (MeOH 10–100% in H2O [+0.05% TFA]) to yield the desired product N-(5-(((5-isopropyloxazol-2-yl)methyl)thio)thiazol-2-yl)piperidine-4-carboxamide 2,2,2-trifluoroacetate (B1) as a white solid (32 mg). Yield over two steps N-(5-(((5-isopropyloxazol-2-yl)methyl)thio)thiazol-2-yl)piperidine-4-carboxamide 2,2,2-trifluoroacetate (52 mg, 36%).

Figure 13—figure supplement 1
Download asset Open asset
Characterization of N-(5-(((5-isopropyloxazol-2-yl)methyl)thio)thiazol-2-yl)piperidine-4-carboxamide 2,2,2-trifluoroacetate (B1).

1H NMR (400 MHz, DMSO-d6) δ 7.38 (s, 1H), 6.72 (d, J = 1.1 Hz, 1H), 4.03 (s, 2H), 3.21 (d, J = 12.8 Hz, 2H), 2.93–2.85 (m, 1H), 2.84–2.75 (m, 2H), 2.70 (s, 1H), 1.89 (d, J = 13.6 Hz, 2H), 1.77–1.63 (m, 2H), 1.12 (d, J = 6.9 Hz, 6H). 13C NMR (214 MHz, DMSO-d6) δ 172.45, 161.00, 158.76, 158.31, 145.12, 120.91, 119.04, 42.48, 38.65, 34.03, 25.29, 24.83, 20.48. HPLC purity: >95%. HRMS calculated for [M+H]+ C16H23N4O2S2: 367.1184; observed [M+H]+: 367.1248. (A) 1H spectra and chemical structure (inset). (B) 13C spectra; (C) HPLC-UV Trace; (D) HRMS Trace.

Figure 14 with 2 supplements
Download asset Open asset
Scheme 2: preparation of compounds B4 and B12.

(a) Propylphosphonic anhydride (T3P), DIPEA, THF, 0°C, 1 hr, 93% yield; (b) palladium on carbon (Pd/C), MeOH, THF, 25°C, 16 hr, 99% yield; (c) 4-methoxybenzoic acid, HATU, DIPEA, THF, 25°C, 17 hr, 72% yield; (d) HCl, dioxane, 25°C, 1 hr, 58% yield; (e) 3-cyanobenzoic acid, T3P, DIPEA, THF, 25°C, 16 hr; (f) HCl, dioxane, 25°C, 1 hr, 39% yield over two steps. General procedure for preparation of B4 and B12: To a flask was added 4-nitro-1H-pyrazole-3-carboxylic acid (1.00 g, 1.0 equiv, 6.4 mmol), 1-Boc-4-aminopiperidine (1.28 g, 1.0 equiv, 6.4 mmol), and DIPEA (4.21 mL, 4.0 equiv, 25.5 mmol) in THF (20 mL). The reaction mixture was cooled to 0°C, and then T3P (6.08 mL, 1.5 equiv, 9.5 mmol) was slowly added. After stirring for 1 hr at 0°C, the reaction mixture was concentrated in vacuo, and then partitioned between EtOAc and brine. The organic phase was separated, dried with MgSO4, and concentrated in vacuo to give a crude solid. The crude material was purified using an automated purification system (SiO2) 90%/10% hexanes/EtOAc to 100% EtOAc to give tert-butyl 4-(4-nitro-1H-pyrazole-3-carboxamido)piperidine-1-carboxylate (Int 2) as an amorphous solid (2.00 g, 93% yield). Palladium on carbon (5%, 31.3 mg, 0.1 equiv, 0.295 mmol) was added to tert-butyl 4-(4-nitro-1H-pyrazole-3-carboxamido)piperidine-1-carboxylate (Int 2, 1.0 g, 1.0 equiv, 2.95 mmol) in a mixture of methanol (10 mL) and THF (3.0 mL). The solution was placed under an atmosphere of H2 at room temperature. After 16 hr, the solution was filtered through SiO2 and the eluent was concentrated to afford tert-butyl 4-(4-amino-1H-pyrazole-3-carboxamido)piperidine-1-carboxylate (Int 3) as an amorphous solid (900 mg, 99% yield). To a flask was added tert-butyl 4-(4-amino-1H-pyrazole-3-carboxamido)piperidine-1-carboxylate (Int 3, 1 equiv) and the corresponding carboxylic acid (1.2 equiv) and DIPEA (4.0 equiv) in THF (0.5 M) at room temperature. The reaction mixture was then treated slowly with T3P (3.0 equiv). After stirring for 16–17 hr, the reaction mixture was concentrated in vacuo, and then partitioned between EtOAc and brine. The organic phase was separated, dried with MgSO4, and concentrated in vacuo to give a crude solid. The crude material was purified using an automated purification system (SiO2) 90%/100% hexanes/EtOAc to 100% EtOAc to give an intermediate, which was stirred in HCl/dioxane (3.3 mL) for 1 hr at room temperature. Solvent was removed and the crude material was purified using an automated purification system (SiO2) 90%/100% hexanes/EtOAc to 100% EtOAc to give the desired product (B4 or B12) as an amorphous solid (39–42% yield over two steps).

Figure 14—figure supplement 1
Download asset Open asset
Characterization of 4-(4-methoxybenzamido)-N-(piperidin-4-yl)–1H-pyrazole-3-carboxamide (B4).

4-(4-methoxybenzamido)-N-(piperidin-4-yl)–1H-pyrazole-3-carboxamide (B4) 1H NMR (850 MHz, methanol-d4) δ 8.30 (s, 1H), 7.89 (d, J = 8.8 Hz, 2H), 7.07 (d, J = 8.8 Hz, 2H), 4.22 (tt, J = 10.8, 4.1 Hz, 1H), 3.89 (s, 3H), 3.48 (dt, J = 13.4, 3.9 Hz, 2H), 3.18 (td, J = 12.7, 3.1 Hz, 2H), 2.22 (dd, J = 14.3, 3.9 Hz, 2H), 1.96–1.85 (m, 2H). 13C NMR (214 MHz, methanol-d4) δ 165.66, 165.58, 164.53, 134.00, 130.01, 126.71, 124.57, 121.57, 115.24, 56.06, 45.26, 44.25, 29.54. 1H NMR purity: >95%. HRMS calculated for [M+H]+ C17H22N5O3: 344.1723; observed [M+H]+: 344.1717. (A) 1H spectra and chemical structure (inset). (B) 13C spectra. (C) HRMS Trace.

Figure 14—figure supplement 2
Download asset Open asset
Characterization of 4-(3-cyanobenzamido)-N-(piperidin-4-yl)–1H-pyrazole-3-carboxamide (B12).

4-(3-cyanobenzamido)-N-(piperidin-4-yl)–1H-pyrazole-3-carboxamide (B12) 1H NMR (850 MHz, methanol-d4) δ 8.33 (s, 1H), 8.26 (t, J = 1.7 Hz, 1H), 8.22 (dt, J = 7.9, 1.5 Hz, 1H), 7.97 (dt, J = 7.7, 1.4 Hz, 1H), 7.75 (t, J = 7.8 Hz, 1H), 4.22 (tt, J = 10.9, 4.1 Hz, 1H), 3.49 (dt, J = 14.1, 3.5 Hz, 2H), 3.18 (td, J = 12.9, 3.1 Hz, 2H), 2.22 (dd, J = 14.2, 3.7 Hz, 2H), 1.95–1.88 (m, 2H). 13C NMR (214 MHz, methanol-d4) δ 165.57, 163.51, 136.50, 136.14, 134.35, 132.50, 131.84, 131.32, 124.01, 122.04, 118.88, 114.39, 45.33, 44.29, 29.53. 1H NMR purity: >95%. HRMS calculated for [M+H]+ C17H19N6O2: 339.1569; observed [M+H]+: 339.1563. (A) 1H spectra and chemical structure (inset). (B) 13C spectra. (C) HRMS Trace.

Tables

Table 1
NanoBRET data corresponding to compounds selected for initial study.
CompoundParentCDKL5 NB IC50 (nM)GSK3β NB IC50 (nM)Fold
A01AT-7519363>10,00027.5
A02AT-7519178413523.2
A03AT-751919404.821.3
A04SNS-032108568852.7
A05AT-751912308.225.7
A06SNS-032221778735.2
A07AT-7519619232.0
A08AT-751956137424.5
A09AT-75192148122.9
A10SNS-032424>10,00023.6
A11AT-751993699411.0
A12AT-751920524.826.2
B01SNS-0328>10,0001250.0
B02SNS-032190910047.9
B03SNS-03210>10,0001000.0
B04AT-751914553.939.6
B05SNS-032388800020.6
B06AT-7519155819352.9
B07AT-751936325690.4
B08AT-7519161>10,00062.1
B09AT-751937362798.0
B10AT-751990494554.9
B11AT-7519141310122.0
B12AT-751969363152.6
C01SNS-032323>10,00031.0
Table 2
Solubility and microsomal stability data for B1.
CompoundKinetic solubilityMouse liver microsomal stability (%)
µMµg/mLT = 0 minT = 30 minT = 30 min (-NADPH)
B1196.8*72.1*10086.187.5
  1. *

    Measured solubility is >75% dose concentration, actual solubility may be higher.

Table 3
Brain exposure results for B1.

(A) Brain/plasma concentration–time data after 2.29 mg/kg dose. (B) Brain/plasma concentration–time data after 7.63 mg/kg dose.

AIP dose: 2.29 mg/kgBIP dose: 7.63 mg/kg
Time (hr)Brain concentration (ng/g)Plasma (ng/mL)Ratio (brain/plasma)Time (hr)Brain conc entration (ng/g)Plasma (ng/mL)Ratio (brain/plasma)
115.63400.045911587500.210*
  1. *

    One animal artificially drove up blood/plasma ratio. N = 3 per dose.

Key resources table
Reagent type (species) or resourceDesignationSource or referenceIdentifiersAdditional information
Cell line (human)HEK293ATCCCRL-1573Routinely subjected to extensive quality control and physiological and molecular characterization (STR profiling). Mycoplasma negativity is regularly confirmed.
AntibodyAnti-CDKL5 (sheep polyclonal)University of DundeeAnti-CDKL5 350-6501:2000
AntibodyAnti-pS222 EB2 (rabbit polyclonal)Covalab, from Baltussen et al., 20181:1000-1:2000
AntibodyAnti-EB2 (rat polyclonal)Abcamab457671:1000
AntibodyAnti-FLAG (rabbit polyclonal)Cell Signaling147931:1000
AntibodyAnti-rabbit IgG HRP (goat polyclonal)Jackson ImmunoResearch111-035-1441:5000
AntibodyAnti-tubulin (mouse monoclonal)SigmaT90261:100,000
AntibodyAnti-CDKL5 (rabbit polyclonal)AtlasHPA0028471:1000
AntibodyAnti-β-catenin (rabbit polyclonal)Cell Signaling95621:1000
AntibodyAnti-phospho-β-catenin (Ser33/37 Thr41) (rabbit polyclonal)Cell Signaling95611:250
Commercial assay or kitCDKL5 TE AssayPromegaNLuc-CDKL5 (NV2911); Transfection Carrier DNA; NanoBRET TE Intracellular Kinase Assay, K-11 (N2650)
Commercial assay or kitGSK3α TE AssayPromegaNLuc-GSK3A (NV3191); Transfection Carrier DNA; NanoBRET TE Intracellular Kinase Assay, K-8 (N2620)
Commercial assay or kitGSK3β TE AssayPromegaNLuc-GSK3B (NV3201); Transfection Carrier DNA; NanoBRET TE Intracellular Kinase Assay, K-8 (N2620)
Commercial assay or kitCDK9/Cyclin K TE AssayPromegaNLuc-CDK9 (NV2871); CCNK Expression Vector (NV2661); NanoBRET TE Intracellular Kinase Assay, K-8 (N2620)
Commercial assay or kitCDK16/Cyclin Y TE AssayPromegaNLuc-CDKL16 (NV2741); CCNY Expression Vector (NV2691); NanoBRET TE Intracellular Kinase Assay, K-12 (NF1001)
Commercial assay or kitCDK17/Cyclin Y TE AssayPromegaNLuc-CDKL17 (NV2751); CCNY Expression Vector (NV2691); NanoBRET TE Intracellular Kinase Assay, K-12 (NF1001)
Commercial assay or kitCDK18/Cyclin Y TE AssayPromegaNLuc-CDKL18 (NV2761); CCNY Expression Vector (NV2691); NanoBRET TE Intracellular Kinase Assay, K-12 (NF1001)
Commercial assay or kitscanMAXEurofins DiscoverX Corporation
Commercial assay or kitKinaseSeekerLuceome BiotechnologiesCDKL5 only
Commercial assay or kitRadiometric kinase assaysEurofins DiscoverX Corporation
Chemical compound, drugB1 (CAF-382)This paper, Figures 1—5,7Available from senior authors (alison.axtman@unc.edu)
Chemical compound, drugB4 (HW2-013)This paper, Figures 1—3, Figures 7 and 14Available from senior authors (alison.axtman@unc.edu)
Chemical compound, drugB12 (LY-213)This paper, Figures 1—3, Figures 7 and 14Available from senior authors (alison.axtman@unc.edu)

Additional files

MDAR checklist
https://cdn.elifesciences.org/articles/88206/elife-88206-mdarchecklist1-v2.pdf

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Anna Castano
  2. Margaux Silvestre
  3. Carrow I Wells
  4. Jennifer L Sanderson
  5. Carla A Ferrer
  6. Han Wee Ong
  7. Yi Lang
  8. William Richardson
  9. Josie A Silvaroli
  10. Frances M Bashore
  11. Jeffery L Smith
  12. Isabelle M Genereux
  13. Kelvin Dempster
  14. David H Drewry
  15. Navlot S Pabla
  16. Alex N Bullock
  17. Tim A Benke
  18. Sila K Ultanir
  19. Alison D Axtman
(2023)
Discovery and characterization of a specific inhibitor of serine-threonine kinase cyclin-dependent kinase-like 5 (CDKL5) demonstrates role in hippocampal CA1 physiology
eLife 12:e88206.
https://doi.org/10.7554/eLife.88206