Development of 5-hydroxybenzothiophene derivatives as multi-kinase inhibitors with potential anti-cancer activity

Abstract

Aim: Chemoresistance in cancer challenges the classical therapeutic strategy of ‘one molecule-one target’. To combat this, multi-target therapies that inhibit various cancer-relevant targets simultaneously are proposed. Methods & results: We introduce 5-hydroxybenzothiophene derivatives as effective multi-target kinase inhibitors, showing notable growth inhibitory activity across different cancer cell lines. Specifically, compound 16b, featuring a 5-hydroxybenzothiophene hydrazide scaffold, emerged as a potent inhibitor, displaying low IC₅₀ values against key kinases and demonstrating significant anti-cancer effects, particularly against U87MG glioblastoma cells. It induced G2/M cell cycle arrest, apoptosis and inhibited cell migration by modulating apoptotic markers. Conclusion: 16b represents a promising lead for developing new anti-cancer agents targeting multiple kinases with affinity to the hydroxybenzothiophene core.

GRAPHICAL ABSTRACT

Plain language summary

Summary points.

• Several 5-hydroxybenzothiophene derivatives were synthesized and assessed as multi-target kinase inhibitors.

• The 5-hydroxybenzothiophene amide derivatives 4, 6, 7 and 10 displayed selective inhibition toward haspin.

• The hydrazide derivative 16b displayed the most potent multi-target kinase inhibition, with IC₅₀ values of 11, 87, 125.7, 163, 284 and 353.3 nM against Clk4, DRAK1, haspin, Clk1, Dyrk1B and Dyrk1A respectively.

• 16b displayed broad spectrum anticancer activity, as it could inhibit the growth of HCT-116, A549, U87MG and Hela cells, with the highest growth inhibition observed in the U87MG cells (IC₅₀ = 7.2 μM).

• 16b induces G2/M cell cycle arrest, apoptosis and inhibits migration in U87MG cells.

Cancer is a highly complex disease that is characterized by a multifactorial nature and therapeutic resistance development [1]. That is why despite the advancement in anti-cancer chemotherapy, the overall survival rates for cancer patients have not significantly improved over the last three decades [2]. Inhibition of protein kinases is regarded as one of the most successful strategies to treat cancer [3], however, the classic therapeutic strategy of ‘one molecule-one target’ seems to be less effective against some advanced tumors that can find escape routes to the targeted kinase leading to the development of resistance [3,4]. A mechanism through which cancer evades the effects of targeted kinase inhibitors involves overexpressing an alternative kinase with similar functions to the targeted one. For instance, in non-small-cell lung cancer (NSCLC), tumor cells respond to EGFR inhibitors by increasing the expression of genes for MET and HER2. These upregulated receptor tyrosine kinases then activate downstream signaling pathways, which are ordinarily triggered by the inhibited EGFR kinase [5]. Therefore, multi-target therapies could provide a resolution for the problem of cancer chemoresistance, as simultaneous inhibition of different targets would grant synergistic pharmacodynamic activities that enhance therapeutic effectiveness and would delay the development of resistance [6].

The benefits of polypharmacology, where a single drug can be used in the treatment of several diseases, have been validated in several studies. Imatinib, a US FDA-approved drug for treating chronic myeloid leukemia that targets the BCR-ABL tyrosine kinase, was also proven effective in treating other rare cancers, including gastrointestinal tumors. This efficacy was correlated to its ability to inhibit KIT and PDGF protein kinases, that are overexpressed in gastrointestinal tumors [3]. Likewise, silmitasertib (also known as CX-4945) demonstrated efficient anti-cancer activities both in vitro against cancer cell lines and in vivo in xenograft mouse models via inhibiting cell proliferation and angiogenesis [7,8]. This could be in part attributed to its multi-target inhibitory capability, particularly against kinases overexpressed in cancer cells. Specifically, it showed an IC₅₀ value of 14 nM against CK2 [9] and IC₅₀ values of 82, 4 and 90 nM against Clk 1, 2 and 3, respectively [10]. In addition, it exhibited IC₅₀ values of 6.8 and 6.4 nM against Dyrk 1A and 1B, respectively [11]. Furthermore, hybrid molecules in which pharmacophoric moieties of two bioactive substances are combined in a single drug have gained attention as multi-target anti-cancer agents that overcome the dose-limiting toxicities and drug–drug interactions drawbacks of combination therapies as well as chemoresistance [12,13].

Targeting kinases that regulate the pre-mRNA splicing process serves as a potential novel treatment of cancer, as the modulation of pre-mRNA splicing is implicated in all the hallmarks of cancer including metabolism, apoptosis, cell cycle control, metastasis, invasion and angiogenesis [14]. Clk1 and Dyrk1A were reported to be among the most essential regulators of pre-mRNA splicing. Clk1 is a dual specificity protein kinase that can auto-phosphorylate on Ser, Thr and Tyr residues, but phosphorylates exogenous substrates mainly on Ser/Thr residues [15]. Other members of the Clk family include Clk2, 3 and 4 [15], with the highest sequence identity between Clk1 and Clk4 (78.4%) [16]. Similar to Clk1, Dyrk1A is a dual specificity protein kinase that phosphorylates exogenous substrates on Ser/Thr residues, however, it catalyzes its self-activation by autophosphorylation at a conserved tyrosine residue [17]. Other members of the Dyrk family include Dyrk1B, 2, 3 and 4 [18]. Both Clks and Dyrks belong to a larger superfamily of the human kinome known as the CMGC group [18,19]. Clk1 and Dyrk1A are able to phosphorylate and activate serine and arginine-rich (SR) proteins [20], which are then released from nuclear speckles into the nucleoplasm where they govern the process of pre-mRNA splice site recognition [16,21]. Several SR proteins were reported to be substrates of both Clk1 and Dyrk1A, including AF2/ASF, serine/arginine-rich splicing factor-like protein (SC35), and SRp55 [21–23]. Schmitt et al. showed that the dual inhibition of Clk1 and Dyrk1A was more effective in modulating the pre-mRNA splicing of SC35 and Clk1 in the immortal embryonic fibroblasts (STO cells) than the inhibition of Clk1 alone [24]. Moreover, Clk1 could regulate the splice site utilization of splicing factor 45 (SPF45), a non-SR splice factor that is overexpressed in various cancers [25]. In addition, Clk1 inhibition resulted in the downregulation of SRSF2 in gastric cancer cells which led to a decrease in their proliferation, migration and invasion [26]. As for Dyrk1A, Yang et al. highlighted the role of Dyrk1A in various types of cancer, providing evidence that Dyrk1A is a promising therapeutic target for treating triple negative breast cancer, prostate cancer, pancreatic ductal adenocarcinoma, hepatocellular carcinoma, B-cell acute lymphoblastic leukemia, acute myelogenous leukemia as well as EGFR dependent glioblastoma [27]. The study also validated the significant role of Dyrk1A in the proliferation and differentiation of glioblastoma stem cells as well as its crucial role in the migration of glioma cells, thus targeting Dyrk1A could be a successful strategy in treating the highly complex glioblastoma.

One of the most common off targets for Clk1/Dyrk1A inhibitors is haspin. Haspin is an atypical Ser/Thr protein kinase, as its domain sequence is different from other members of the eukaryotic protein kinase super family [28–30]. Haspin's main function is to phosphorylate histone H3 at Thr3. It functions in chromosome alignment, centromeric cohesion and spindle stability during mitosis or meiosis [28,30]. The central role of haspin in chromosomal segregation and cellular proliferation makes it an interesting target for cancer therapy [29]. Haspin was found to be overexpressed in several tumors including chronic lymphocytic leukemia [31], NSCLC [32], breast cancer [33], colorectal cancer [34], pancreatic ductal adenocarcinoma [35], bladder cancer [36], hepatocellular carcinoma [37], prostate cancer [38] and ovarian cancer [39].

Taking into consideration that several types of cancer share an overexpression of Clk1, Dyrk1A and haspin kinases, in addition to their significant roles in cancer development and progression, the aim of this study is to design compounds with multi-target inhibition against Clk1, Dyrk1A and haspin and evaluate their cellular activity against various cancer types.

1. Materials & methods

1.1. Chemistry

1.1.1. General

Solvents and reagents were obtained from commercial suppliers and used as received. Varian Mercury VX 400 (CA, USA) and Bruker DRX 500 spectrometers (MA, USA) were used to obtain ¹H NMR and ¹³C NMR spectra. Chemical shifts (δ) were reported in parts per million (ppm) downfield from TMS; and all coupling constants (J) are given in Hz. Multiplicities are abbreviated as: s: singlet; d: doublet; t: triplet; q: quartet; m: multiplet; dd: doublet of doublet; dt: doublet of triplet; td: triplet of doublet. At least 95% purity in all the tested compounds was obtained by means of HPLC coupled with mass spectrometry. Mass spectrometric analysis (UPLC-ESI-MS) was performed using Waters ACQUITY Xevo Triple Quadrupole (TQD) system as previously described [19]. Melting points were determined by a Buchi B-540 Melting Point apparatus (Flawil, Switzerland) and are uncorrected.

1.1.2. Synthesis

1.1.2.1. Procedure A: Synthesis of (Z)-5-(3-hydroxybenzylidene)-2-thioxothiazolidin-4-one (A) & (Z)-5-(3-methoxybenzylidene)-2-thioxothiazolidin-4-one (D)

Intermediate A and Intermediate D were prepared by the reaction of rhodanine with 3-hydroxybenzaldehyde, or 3-methoxybenzaldehyde, respectively, in the presence of anhydrous sodium acetate, glacial acetic acid and toluene according to the previously reported procedure in Ref. [40] Intermediate A: melting point (mp) 245–246°C; (ESI-MS) m/z = 238.30. [M + H]⁺ [41]; Intermediate D: mp 230–232°C; (ESI-MS) m/z = 252.2 [M + H]⁺ [42].

1.1.2.2. Procedure B: Synthesis of (Z)-3-(3-hydroxyphenyl)-2-mercaptoacrylic acid (B) & (Z)-2-mercapto-3-(3-methoxyphenyl)acrylic acid (E)

Intermediate B and Intermediate E were prepared by the reaction of the benzylidene rhodanine derivative A or D, respectively, with 15% sodium hydroxide solution according to the previously reported procedure in Ref. [40] Intermediate B: mp 172–173°C; (ESI-MS) m/z = 197.22 [M + H]⁺ [43]; Intermediate E: mp 98–100°C; (ESI-MS) m/z = 209.1 [M - H]^- [42].

1.1.2.3. Procedure C: Synthesis of 6-hydroxybenzo[b]thiophene-2-carboxylic acid (C) & 6-methoxybenzo[b]thiophene-2-carboxylic acid (F)

Compound C and Compound F were prepared by the reaction of the mercaptoacrylic acid derivatives B or E, respectively, with iodine in dry THF according to the previously reported procedure in Ref. [40] Compound C: mp 262–264; (ESI-MS) m/z = 193 [M - H]^- [41]; Compound F: mp 214–216; (ESI-MS) m/z = 207.05 [M - H]^- [40].

1.1.2.4. Procedure D: General procedure for the synthesis of the 5-hydroxybenzo[b]thiophene-2-amide derivatives (1–12)

1 mmol of intermediate C was dissolved in minimal volume of dry DMF, and then 2 equiv. of EDC and DMAP were added. The reaction mixture was stirred for 10 min. This was followed by the addition of 3 equiv. of the appropriate amine or aniline. The reaction mixture was left to stir at room temperature overnight. The solvent was then evaporated under reduced pressure and high temperature. Then the residue was dissolved in 50 mL of dichloromethane (DCM) and 20 mL of water were added for extraction. The aqueous layer was extracted with DCM (50 mL × 3). Organic layers were then combined and dried over anhydrous MgSO₄ and evaporated under vacuum followed by purification using column chromatography (CC).

1.1.2.5. Procedure E: General procedure for the synthesis of 5-methoxybenzo[b]thiophene-2-carbohydrazide derivatives (13a–19a)

Compound F (1 mmol, 1 equiv.) was added to a well stirred mixture of oxalyl chloride (4 mmole, 4 equiv.) and a catalytic amount of DMF (80 μl) in 30 mL dry DCM and heated to reflux for 3 h. The reaction was monitored by quenching an aliquot of the reaction with dry methanol and tracing the formation of the methyl ester of the acid either by TLC or LC-MS. After reaction completion, 20 mL toluene was added to the hot mixture and the solvent was evaporated thoroughly under reduced pressure to remove any excess oxalyl chloride or DMF. The resulting acid chloride residue was then dissolved in 10 mL anhydrous THF. This was followed by the addition of 4 mL DIPEA and the appropriate hydrazine hydrochloride (2 mmole, 2 equiv.). The reaction mixture was stirred at room temperature for 18 h. Afterward, the mixture was diluted with water and extracted with ethyl acetate (3 × 30 mL). The combined organic extracts were then dried over anhydrous MgSO₄ and evaporated under reduced pressure. The crude product was purified by CC to give hydrazide as major product with trace impurities. The pure hydrazide was then obtained by addition of 3 mL DCM followed by addition of 30 mL hexane where the product was precipitated as pure crystals, that were then collected by filtration.

1.1.2.6. Procedure F: General procedure for esterification (synthesis of intermediate G)

To an ice-cooled solution of 5-methoxybenzo[b]thiophene-2-carboxylic acid (F) (1 mmole) in 50 mL of methanol, acetyl chloride (0.4 mL) was added drop wise. Then the mixture was heated to reflux for at least 5 h. After completion of the reaction, the solvent was evaporated under vacuum, then it was dissolved in 50 mL DCM and extracted with 100 mL of 30% NH₃ solution. The aqueous layer was further extracted using another 50 mL of DCM. The organic layers were combined and washed by using a brine solution. Then the organic layers were collected and dried over MgSO₄ followed by evaporation of solvent under reduced pressure to give the methyl ester of intermediate F (intermediate G) that was used directly for the next steps without further purification; (ESI-MS) m/z = 237 [M + H]⁺ [24].

1.1.2.7. Procedure G: General procedure for the synthesis of intermediate (H)

A mixture of compound G (1 mmol) and 99% pure hydrazine hydrate (5.5 mL) were stirred to reflux for 3 h. After completion of the reaction, the reaction mixture was then transferred to a beaker containing ice so that the hydrazide derivative (H) would precipitate. The precipitate (intermediate H) was then filtered, collected and used in the next step without further purification.

1.1.2.8. Procedure H: General procedure for the synthesis of 5-methoxybenzo[b]thiophene-2-hydrazoneamide derivatives (20a & 21a)

Intermediate H was dissolved in 15 mL of absolute ethanol. This is followed by the addition of 2–3 drops of glacial acetic acid and 1.5 equiv. of the appropriate aldehyde. The reaction mixture was left to stir under reflux for 3 h. After completion of the reaction, the solvent is evaporated under reduced pressure followed by purification using CC.

1.1.2.9. Procedure I: General procedure for demethylation

The compound to be demethylated is dissolved in 30 mL of HPLC DCM, the flask is immersed in ice before the addition of 5 mL of BBr₃. The mixture was then left to stir for 18 h. After completion of the reaction, water was added to the mixture and extraction was done using DCM (50 mL × 3). Organic layers were then collected and dried over anhydrous MgSO₄ and evaporated under vacuum followed by purification using CC.

1.1.3. N-hexyl-5-hydroxybenzo[b]thiophene-2-carboxamide (1)

The title compound was synthesized according to procedure D using intermediate C and hexyl amine. The product was purified using CC (DCM/Methanol 100:1) to give a brown solid: yield (28.19%); mp 270–271°C; ¹H-NMR (400 MHz, DMSO-d₆) δ 9.46 (s, 1H), 8.60 (t, J = 5.4 Hz, 1H), 7.87 (s, 1H), 7.76 (d, J = 8.7 Hz, 1H), 7.20 (d, J = 2.0 Hz, 1H), 6.95 (dd, J = 8.7, 2.2 Hz, 1H), 4.46 (d, J = 4.4 Hz, 2H), 1.51 (dd, J = 13.6, 6.7 Hz, 2H), 1.36–1.22 (m, 6H), 0.86 (t, J = 6.4 Hz, 3H).; ¹³C-NMR (101 MHz, DMSO-d₆) δ 161.87, 155.60, 141.35, 140.93, 131.54, 124.35, 123.85, 117.22, 109.44, 72.92, 31.44, 29.46, 26.56, 22.49, 14.36. MS (ESI): m/z = 278.0 (M + H)⁺.

1.1.4. N-cyclohexyl-5-hydroxybenzo[b]thiophene-2-carboxamide (2)

The title compound was synthesized according to procedure D using intermediate C and cyclohexyl amine. The product was purified using CC (DCM/Methanol 100:0.5) to give a brown solid: yield (35.22%); mp 284.5–285.5°C; ¹H-NMR (500 MHz, DMSO-d₆) δ 9.56 (s, 1H), 8.40 (d, J = 7.9 Hz, 1H), 7.93 (s, 1H), 7.76 (d, J = 8.7 Hz, 1H), 7.20 (d, J = 2.3 Hz, 1H), 6.96 (dd, J = 8.7, 2.4 Hz, 1H), 3.76–3.69 (m, 1H), 1.84 (d, J = 8.0 Hz, 2H), 1.78–1.70 (m, 2H), 1.62 (d, J = 12.1 Hz, 1H), 1.37–1.25 (m, 4H), 1.20–1.09 (m, 1H); ¹³C-NMR (126 MHz, DMSO-d₆) δ 161.09, 155.62, 141.61, 140.99, 131.60, 124.46, 123.88, 117.21, 109.47, 48.98, 32.87, 25.68, 25.35; MS (ESI): m/z = 278.0 (M + H)⁺.

1.1.5. N-cyclopropyl-5-hydroxybenzo[b]thiophene-2-carboxamide (3)

The title compound was synthesized according to procedure D using intermediate C and cyclopropyl amine. The product was purified using CC (DCM/Methanol 100:1) to give a brown solid: yield (44.5%); mp 201–202°C; ¹H-NMR (400 MHz, DMSO-d₆) δ 9.53 (s, 1H), 8.61 (s, 1H), 7.87–7.70 (m, 2H), 7.17 (s, 1H), 6.94 (s, 1H), 2.81 (s, 1H), 1.46–1.11 (m, 2H), 0.87–0.65 (m, 2H).; ¹³C-NMR (101 MHz, DMSO-d₆) δ 163.21, 155.62, 141.06, 140.91, 131.58, 124.59, 123.86, 117.28, 109.45, 23.43, 6.20; MS (ESI): m/z = 234.0 (M + H)⁺.

1.1.6. 5-hydroxy-N-(3-hydroxyphenyl)benzo[b]thiophene-2-carboxamide (4)

The title compound was synthesized according to procedure D using intermediate C and 3-aminophenol. The product was purified using CC (DCM/Methanol 100:2) to give a brown solid: yield (26.27%); mp 218.7–219.7°C; ¹H-NMR (400 MHz, DMSO-d₆) δ 10.26 (s, 1H), 9.60 (s, 1H), 9.44 (s, 1H), 8.16 (s, 1H), 7.81 (d, J = 8.6 Hz, 1H), 7.32 (s, 1H), 7.27 (s, 1H), 7.18–7.10 (m, 2H), 7.00 (d, J = 8.6 Hz, 1H), 6.52 (d, J = 7.2 Hz, 1H); ¹³C-NMR (101 MHz, DMSO-d₆) δ 160.71, 158.00, 155.74, 141.24, 140.88, 140.11, 131.94, 129.78, 125.64, 123.93, 117.67, 111.44, 111.41, 109.65, 107.78; MS (ESI): m/z = 218.7 (M + H)⁺.

1.1.7. 5-hydroxy-N-(2-hydroxyphenyl)benzo[b]thiophene-2-carboxamide (5)

The title compound was synthesized according to procedure D using intermediate C and 2-aminophenol. The product was purified using CC (DCM/Methanol 100:1) to give a dark brown solid: yield (30.28%); mp 220–221°C; ¹H-NMR (400 MHz, DMSO-d₆) δ 9.79 (s, 1H), 9.61 (s, 2H), 8.15 (s, 1H), 7.81 (d, J = 8.6 Hz, 1H), 7.64 (d, J = 7.4 Hz, 1H), 7.27 (s, 1H), 7.07–6.99 (m, 2H), 6.94 (d, J = 7.7 Hz, 1H), 6.83 (t, J = 7.1 Hz, 1H); ¹³C-NMR (101 MHz, DMSO-d₆) δ 160.71, 158.00, 155.74, 141.24, 140.88, 140.11, 131.94, 129.78, 125.64, 123.93, 117.67, 111.44, 111.41, 109.65, 107.78; MS (ESI): m/z = 285.1 (M + H)⁺.

1.1.8. 5-hydroxy-N-(4-hydroxyphenyl)benzo[b]thiophene-2-carboxamide (6)

The title compound was synthesized according to procedure D using intermediate C and 4-aminophenol. The product was purified using CC (DCM/Methanol 100:1.5) to give a brown solid: yield (14.9%); mp 217.7–219.7°C; ¹H-NMR (400 MHz, DMSO-d₆) δ 10.26 (s, 1H), 9.60 (s, 1H), 9.44 (s, 1H), 8.32 (s, 1H), 7.57 (s, 1H), 7.34 (s, 1H), 7.29 (s, 1H), 6.83(d, J = 7.4 Hz, 2H), 6.81 (d, J = 1.4 Hz, 2H); ¹³C-NMR (101 MHz, DMSO-d₆) δ 158.25, 156.57, 154.90, 139.40, 135.17, 132.35, 129.97, 124.38, 122.93, 122.32, 118.72, 116.05, 114.14; MS (ESI): m/z = 285.1 (M + H)⁺.

1.1.9. 5-hydroxy-N-(4-methoxyphenyl)benzo[b]thiophene-2-carboxamide (7)

The title compound was synthesized according to procedure D using intermediate C and para-anisidine. The product was purified using CC (DCM/Methanol 100:2) followed by the dropwise addition of DCM until a precipitate of the pure final product was formed and collected by filtration to give a dark brown solid: yield (23.6%); mp 194–195°C; ¹H-NMR (400 MHz, DMSO-d₆) δ 10.30 (s, 1H), 9.60 (s, 1H), 8.13 (s, 1H), 7.81 (d, J = 8.6 Hz, 1H), 7.66 (d, J = 7.1 Hz, 2H), 7.27 (s, 1H), 7.00 (d, J = 8.7 Hz, 1H), 6.94 (d, J = 7.1 Hz, 2H), 3.75 (d, J = 1.8 Hz, 3H); ¹³C-NMR (101 MHz, DMSO-d₆) δ 160.45, 156.14, 155.72, 141.30, 140.90, 132.10, 131.86, 125.34, 123.93, 122.33, 117.58, 114.30, 109.60, 55.63; MS (ESI): m/z = 299.1 (M + H)⁺.

1.1.10. 5-hydroxy-N-(3-methoxyphenyl)benzo[b]thiophene-2-carboxamide (8)

The title compound was synthesized according to procedure D using intermediate C and meta-anisidine. The product was purified using CC (DCM/Methanol 100:2) followed by the dropwise addition of DCM until a precipitate of the pure final product was formed and collected by filtration to give a light brown powder: yield (20.81%); mp 171–172.4°C; ¹H-NMR (400 MHz, DMSO-d₆) δ 10.37 (s, 1H), 9.62 (s, 1H), 8.17 (s, 1H), 7.82 (d, J = 8.7 Hz, 1H), 7.45 (t, J = 1.9 Hz, 1H), 7.35 (d, J = 8.1 Hz, 1H), 7.28 (d, J = 2.4 Hz, 1H), 7.26 (d, J = 8.1 Hz, 1H), 7.01 (dd, J = 8.7, 2.3 Hz, 1H), 6.70 (dd, J = 8.1, 2.1 Hz, 1H), 3.76 (s, 3H); ¹³C-NMR (101 MHz, DMSO-d₆) δ 160.85, 159.91, 155.77, 141.05, 140.86, 140.29, 132.00, 129.96, 125.76, 123.97, 117.76, 112.86, 109.87, 109.68, 106.33, 55.47; MS (ESI): m/z = 299.1 (M + H)⁺.

1.1.11. N-(4-fluorophenyl)-5-hydroxybenzo[b]thiophene-2-carboxamide (9)

The title compound was synthesized according to procedure D using intermediate C and 4-fluoro aniline. The product was purified using CC (DCM/Methanol 100:0.5) to give a light brown powder: yield (20.7%); mp 220.2–221.2°C; ¹H-NMR (400 MHz, DMSO-d₆) δ 10.47 (s, 1H), 9.62 (s, 1H), 8.15 (s, 1H), 7.81 (d, J = 8.8 Hz, 1H), 7.79–7.75 (m, 2H), 7.28 (d, J = 2.3 Hz, 1H), 7.24–7.18 (m, 2H), 7.01 (dd, J = 8.7, 2.3 Hz, 1H); ¹³C-NMR (101 MHz, DMSO-d₆) δ 160.78, 158.83 (d, ¹J_C-F = 240.6 Hz), 155.78, 140.84, 135.45 (d, ⁴J_C-F = 2.6 Hz), 131.96, 129.09, 125.78, 123.97, 122.56 (d, ³J_C-F = 7.9 Hz), 117.76, 115.77 (d, ²J_C-F = 22.3 Hz), 109.67; MS (ESI): m/z = 287.0 (M + H)⁺.

1.1.12. N-(3-fluorophenyl)-5-hydroxybenzo[b]thiophene-2-carboxamide (10)

The title compound was synthesized according to procedure D using intermediate C and 3-fluoro aniline. The product was purified using CC (DCM/Methanol 100:0.5) to give a brown solid: yield (22.10%); mp 206.3–207.3°C; ¹H-NMR (500 MHz, DMSO-d₆) δ 10.62 (s, 1H), 9.72 (s, 1H), 8.21 (s, 1H), 7.82 (d, J = 8.7 Hz, 1H), 7.77 (dt, J = 11.7, 2.1 Hz, 1H), 7.58 (dd, J = 8.2, 1.0 Hz, 1H), 7.40 (dd, J = 15.2, 8.1 Hz, 1H), 7.34 (d, J = 2.3 Hz, 1H), 7.05 (dd, J = 8.7, 2.3 Hz, 1H), 6.94 (td, J = 8.5, 2.4 Hz, 1H); ¹³C-NMR (126 MHz, DMSO-d₆) δ 162.57 (d, ¹J_C-F = 241.3 Hz), 161.17, 155.89, 140.96, 140.86, 140.64, 132.17, 130.77 (d, ³J_C-F = 9.6 Hz), 126.18, 123.99, 117.97, 116.37 (d, ³J_C-F = 2.4 Hz), 110.76 (d, ²J_C-F = 21.1 Hz), 109.82, 107.45 (d, ²J_C-F = 26.5 Hz); MS (ESI): m/z = 287.0 (M + H)⁺.

1.1.13. N-(3-chlorophenyl)-5-hydroxybenzo[b]thiophene-2-carboxamide (11)

The title compound was synthesized according to procedure D using intermediate C and 3-chloro aniline. The product was purified using CC (DCM/Methanol 100:1.5) to give a brown solid: yield (20.9%); mp 200.5–201.5°C; ¹H-NMR (400 MHz, DMSO- d₆) δ 10.56 (s, 1H), 8.18 (s, 1H), 7.92 (s, 1H), 7.83 (d, J = 8.7 Hz, 1H), 7.70 (d, J = 8.1 Hz, 1H), 7.40 (t, J = 8.1 Hz, 1H), 7.30 (d, J = 1.2 Hz, 1H), 7.18 (d, J = 7.6 Hz, 1H), 7.02 (dd, J = 8.7, 1.6 Hz, 1H); ¹³C-NMR (101 MHz, DMSO- d₆) δ 160.43, 155.16, 140.13, 139.93, 139.83, 132.81, 131.44, 130.23, 125.54, 123.35, 123.33, 119.41, 118.31, 117.28, 109.09; MS (ESI): m/z = 303.0 (M + H)⁺.

1.1.14. N-(4-chlorophenyl)-5-hydroxybenzo[b]thiophene-2-carboxamide (12)

The title compound was synthesized according to procedure D using intermediate C and 4-chloro aniline. The product was purified using CC (DCM/Methanol 100:0.5) followed by the dropwise addition of DCM until a precipitate of the pure final product was formed and collected by filtration to give a brown solid: yield (20%); mp 205–206°C; ¹H-NMR (400 MHz, DMSO-d₆) δ 10.51 (s, 1H), 9.62(s, 1H), 8.15 (s, 1H), 7.81 (d, J = 1.1 Hz, 1H), 7.78 (d, J = 8.6 Hz, 2H), 7.44–7.37 (m, 2H), 7.26 (s, 1H), 6.99 (d, J = 8.6 Hz, 1H); ¹³C-NMR (101 MHz, DMSO-d₆) δ 160.93, 155.79, 140.80, 140.66, 138.07, 132.03, 129.09, 127.94, 126.01, 123.99, 122.20, 117.86, 109.70; MS (ESI): m/z = 303.0 (M + H)⁺.

1.1.15. N′-(3-chlorophenyl)-5-hydroxybenzo[b]thiophene-2-carbohydrazide (13b)

The title compound was prepared by demethylation of compound 13a according to the general procedure for demethylation. The product was purified using CC (DCM/Methanol 100:1.5) to give a brown solid: yield (10.7%); mp 205–206°C; ¹H-NMR (400 MHz, DMSO-d₆) δ 10.58 (s, 1H), 9.62 (s, 1H), 8.29 (s, 1H), 8.04 (s, 1H), 7.81 (d, J = 8.7 Hz, 1H), 7.26 (s, 1H), 7.18 (t, J = 8.2 Hz, 1H), 7.00 (d, J = 8.7 Hz, 1H), 6.75 (m, J = 12.2 Hz, 3H); ¹³C-NMR (101 MHz, DMSO-d₆) δ 161.68, 155.14, 150.66, 140.19, 137.72, 133.31, 131.05, 130.32, 124.92, 123.31, 118.01, 117.10, 111.27, 110.71, 108.98; MS (ESI): m/z = 319.0 (M + H)⁺.

1.1.16. N′-(4-chlorophenyl)-5-hydroxybenzo[b]thiophene-2-carbohydrazide (14b)

The title compound was prepared by demethylation of compound 14a according to the general procedure for demethylation. The product was purified using CC (DCM/Methanol 100:1) to give a brown solid: yield (10.7%); mp 190–191°C; ¹H-NMR (400 MHz, DMSO-d₆) δ 10.58 (s, 1H), 9.62 (s, 1H), 8.29 (s, 1H), 8.04 (s, 1H), 7.72 (s, 1H), 7.28 (s, 1H), 7.20 (d, J = 7.5 Hz, 2H), 6.81 (s, 1H), 6.60 (d, J = 7.5 Hz, 2H); ¹³C-NMR (101 MHz, DMSO-d₆) δ 158.29, 154.90, 147.46, 139.40, 135.17, 132.72, 128.69, 124.23, 122.32, 122.04, 118.72, 114.14, 113.09; MS (ESI): m/z = 319.0 (M + H)⁺.

1.1.17. N′-(4-fluorophenyl)-5-hydroxybenzo[b]thiophene-2-carbohydrazide (15b)

The title compound was prepared by demethylation of compound 15a according to the general procedure for demethylation. The product was purified using CC (DCM/Methanol 100:1) to give a brown solid: yield (11.3%); mp 207.7–208.7°C; ¹H-NMR (400 MHz, DMSO- d₆) δ 10.53 (d, J = 2.7 Hz, 1H), 9.59 (s, 1H), 8.01 (s, 1H), 7.95 (d, J = 2.4 Hz, 1H), 7.78 (d, J = 8.7 Hz, 1H), 7.23 (d, J = 2.1 Hz, 1H), 6.99–6.94 (m, 3H), 6.80–6.75 (m, 2H); ¹³C-NMR (101 MHz, DMSO-d₆) δ 162.33, 156.40 (d, ¹J_C-F = 233.6 Hz), 155.76, 146.23 (d, ⁴J_C-F = 1.6 Hz), 140.85, 138.71, 131.65, 125.32, 123.93, 117.65, 115.68 (d, ²J_C-F = 22.4 Hz), 113.96 (d, ³J_C-F = 7.6 Hz), 109.57; MS (ESI): m/z = 303.1 (M + H)⁺.

1.1.18. N′-(3-fluorophenyl)-5-hydroxybenzo[b]thiophene-2-carbohydrazide (16b)

The title compound was prepared by demethylation of compound 16a according to the general procedure for demethylation. The product was purified using CC (DCM/Methanol 100:0.25) to give a brown solid: yield (6.19%); mp 102–103°C; ¹H-NMR (400 MHz, DMSO-d₆) δ 10.58 (s, 1H), 9.63 (s, 1H), 8.30 (s, 1H), 8.05 (s, 1H), 7.81 (d, J = 8.6 Hz, 1H), 7.26 (d, J = 2.2 Hz, 1H), 7.18 (dd, J = 15.4, 8.2 Hz, 1H), 7.01 (dd, J = 8.7, 2.3 Hz, 1H), 6.62 (d, J = 7.7 Hz, 1H), 6.55–6.48 (m, 2H); ¹³C-NMR (101 MHz, DMSO-d₆) δ 163.57 (d, ¹J_C-F = 250.1 Hz), 162.42, 155.78, 151.93 (d, ³J_C-F = 10.3 Hz), 140.84, 138.45, 131.68, 130.89 (d, ³J_C-F = 9.8 Hz), 125.53, 123.94, 117.72, 109.61, 108.70, 105.26 (d, ²J_C-F = 21.4 Hz), 99.18 (d, ²J_C-F = 25.6 Hz); MS (ESI): m/z = 303.1 (M + H)⁺.

1.1.19. N′-(2-fluorophenyl)-5-hydroxybenzo[b]thiophene-2-carbohydrazide (17b)

The title compound was prepared by demethylation of compound 17a according to the general procedure for demethylation. The product was purified using CC (DCM/Methanol 100:1.5) to give a brown solid: yield (20.34%); mp 145–150°C; ¹H-NMR (500 MHz, DMSO-d₆) δ 10.59 (d, J = 1.9 Hz, 1H), 9.63 (s, 1H), 8.05 (s, 1H), 7.94 (s, 1H), 7.81 (d, J = 8.7 Hz, 1H), 7.26 (d, J = 2.2 Hz, 1H), 7.11 (dd, J = 11.2, 8.1 Hz, 1H), 7.01–6.98 (m, 2H), 6.85 (t, J = 7.8 Hz, 1H), 6.78–6.72 (m, 1H); ¹³C-NMR (126 MHz, DMSO-d₆) δ 162.40, 155.82, 150.68 (d, ¹J_C-F = 239.0 Hz), 140.89, 138.62, 137.28 (d, ²J_C-F = 10.5 Hz), 131.72, 125.50, 125.09, 124.00, 119.43 (d, ³J_C-F = 6.6 Hz), 117.74, 115.43 (d, ²J_C-F = 17.7 Hz), 114.14, 109.64; MS (ESI): m/z = 303.1 (M + H)⁺.

1.1.20. N′-(3,5-difluorophenyl)-5-hydroxybenzo[b]thiophene-2-carbohydrazide (18b)

The title compound was prepared by demethylation of compound 18a according to the general procedure for demethylation. The product was purified using CC (DCM/Methanol 100:1) to give a brown solid: yield (13%); mp 164.4–165.4°C; ¹H-NMR (500 MHz, DMSO-d₆) δ 10.64 (s, 1H), 9.64 (s, 1H), 8.60 (s, 1H), 8.06 (s, 1H), 7.82 (d, J = 8.7 Hz, 1H), 7.27 (d, J = 2.1 Hz, 1H), 7.01 (dd, J = 8.7, 2.2 Hz, 1H), 6.48 (t, J = 9.3 Hz, 1H), 6.40 (d, J = 8.0 Hz, 2H); ¹³C-NMR (126 MHz, DMSO-d₆) δ 163.79 (dd, ¹J ³J_C-F = 242.5, 15.8 Hz), 162.38, 155.84, 152.88 (t, ³J ³J_C-F = 12.9 Hz), 140.85, 138.13, 131.76, 125.85, 123.99, 117.83, 109.67, 95.36 (d, ²J_C-F = 29.0 Hz), 93.72 (t, ²J ²J_C-F = 26.5 Hz); MS (ESI): m/z = 321.0 (M + H)⁺.

1.1.21. N′-(3,5-dichlorophenyl)-5-hydroxybenzo[b]thiophene-2-carbohydrazide (19b)

The title compound was prepared by demethylation of compound 19a according to the general procedure for demethylation. The product was purified using CC (DCM/Methanol 100:1) to give a brown solid: yield (15.0%); mp 166–167°C; ¹H-NMR (500 MHz, DMSO-d₆) δ 10.67 (d, J = 1.6 Hz, 1H), 9.66 (s, 1H), 8.60 (d, J = 1.6 Hz, 1H), 8.06 (s, 1H), 7.83 (d, J = 8.7 Hz, 1H), 7.28 (d, J = 2.3 Hz, 1H), 7.02 (dd, J = 8.8, 2.3 Hz, 1H), 6.87 (t, J = 1.6 Hz, 1H), 6.76 (d, J = 1.7 Hz, 2H); ¹³C-NMR (126 MHz, DMSO-d₆) δ 160.27, 153.72, 150.08, 138.73, 135.88, 132.85, 129.67, 123.80, 121.88, 115.87, 115.77, 108.67, 107.59; MS (ESI): m/z = 321.0 (M + H)⁺.

1.1.22. N′-(3-fluorobenzylidene)-5-methoxybenzo[b]thiophene-2-carbohydrazide (20a)

The title compound was synthesized according to procedure H using intermediate H and 3-fluorobenzaldehyde. The product was purified using CC (DCM/Methanol 100:0.25) followed by the dropwise addition of petroleum ether until a precipitate of the pure final product was formed and collected by filtration to give a white solid: yield (67.7%); mp 225.1–226.1°C; ¹H-NMR (400 MHz, DMSO-d₆) δ 12.17 (s, 2H), 8.48 (s, 1H), 8.34 (s, 1H), 8.16 (s, 2H), 7.95 (d, J = 5.7 Hz, 2H), 7.59 (m, 8H), 7.30 (s, 2H), 7.15 (d, J = 8.7 Hz, 2H), 3.85 (s, 6H).¹³C-NMR (101 MHz, DMSO-d₆) δ164.08, 161.66, 160.48 (d, ¹J_C-F = 244.0 Hz), 157.88 (d, ³J_C-F = 12.6 Hz), 147.08, 143.87, 140.58, 138.87, 133.39, 132.17, 131.49 (d, ³J_C-F = 9.9 Hz), 126.09, 124.08, 117.34 (d, ²J_C-F = 21.5 Hz), 113.98 (d, ²J_C-F = 20.3 Hz), 107.34, 55.87; MS (ESI): m/z = 329.0 (M + H)⁺.

1.1.23. N′-(4-fluorobenzylidene)-5-methoxybenzo[b]thiophene-2-carbohydrazide (21a)

The title compound was synthesized according to procedure H using intermediate H and 4-fluorobenzaldehyde. The product was purified using CC (DCM/Methanol 100:0.25) followed by the dropwise addition of petroleum ether until a precipitate of the pure final product was formed and collected by filtration to give a white solid: yield (67.7%); mp 194.7–195.7°C; ¹H-NMR (400 MHz, DMSO-d₆) δ 12.13 (s, 1H), 11.99 (s, 1H), 8.46 (s, 1H), 8.32 (s, 1H), 8.13 (s, 2H), 7.92 (d, J = 8.9 Hz, 4H), 7.80 (s, 2H), 7.55 (s, 1H), 7.48 (s, 1H), 7.37–7.26 (m, 4H), 7.14 (d, J = 2.2 Hz, 1H), 7.12 (d, J = 2.2 Hz, 1H), 3.83 (s, 6H); ¹³C-NMR (101 MHz, DMSO-d6) δ 163.63 (d, ¹J_C-F = 246.6 Hz), 163.55 (d, ¹J_C-F = 252.4 Hz), 162.07, 158.66, 157.96, 157.75, 147.37, 143.99, 140.60, 139.67, 138.91, 136.21, 134.62, 133.31, 132.01, 131.23, 131.10, 130.06 (d, ³J_C-F = 6.0 Hz), 129.83 (d, ³J_C-F = 7.8 Hz), 125.92, 124.07, 123.83, 118.02, 117.54, 116.58 (d, ²J_C-F = 12.6 Hz), 116.36 (d, ²J_C-F = 12.8 Hz), 107.33, 107.22, 55.88, 55.83; MS (ESI): m/z = 329.0 (M + H)⁺.

1.1.24. N′-(3-fluorobenzylidene)-5-hydroxybenzo[b]thiophene-2-carbohydrazide (20b)

The title compound was prepared by demethylation of compound 20a according to the general procedure for demethylation. The product was purified using CC (DCM/Methanol 100:1.5) followed by dissolving in DCM, then dropwise addition of petroleum ether till the precipitation of the pure final product as a white solid: yield (62.23%); mp 237.4–238.4°C; ¹H-NMR (400 MHz, DMSO-d₆) δ 12.15 (s, 1H), 12.05 (s, 1H), 9.61 (s, 2H), 8.45 (s, 1H), 8.24 (s, 1H), 8.13 (s, 1H), 8.07 (s, 1H), 7.83 (s, 2H), 7.70–7.46 (m, 6H), 7.28 (s, 4H), 7.01 (d, J = 8.3 Hz, 2H); ¹³C-NMR (101 MHz, DMSO-d₆) δ 164.08, 161.66, 160.59 (d, ¹J_C-F = 243.9 Hz), 155.77 (d, ³J_C-F = 7.8 Hz), 146.95, 143.68, 143.64, 140.77, 139.06, 137.14, 131.96, 131.42 (d, ³J_C-F = 8.8 Hz), 125.89, 123.97, 117.92, 117.34 (d, ²J_C-F = 16.9 Hz), 113.58 (d, ²J_C-F = 20.9 Hz), 109.71; MS (ESI): m/z = 315.0 (M + H)⁺.

1.1.25. N′-(4-fluorobenzylidene)-5-hydroxybenzo[b]thiophene-2-carbohydrazide (21b)

The title compound was prepared by demethylation of compound 21a according to the general procedure for demethylation The product was purified using CC (DCM/Methanol 100:1) followed by dissolving in DCM, then dropwise addition of petroleum ether till the precipitation of the pure final product as a white solid: yield (60.11%); mp 229.5–230.5°C; ¹H-NMR (400 MHz, DMSO-d6) δ 12.07 (s, 1H), 11.97 (s, 1H), 9.62 (s, 2H), 8.46 (s, 1H), 8.25 (s, 1H), 8.15 (s, 1H), 8.07 (s, 1H), 7.90 (s, 2H), 7.83 (d, J = 8.1 Hz, 4H), 7.33 (s, 6H), 7.02 (d, J = 7.7 Hz, 2H); ¹³C-NMR (101 MHz, DMSO-d6) δ 163.37 (d, ¹J_C-F = 246.1 Hz), 163.24 (d, ¹J_C-F = 254.9 Hz), 158.49, 155.58, 155.38, 155.33, 147.01, 146.98, 143.57, 140.55, 139.03 (d, ⁴J_C-F = 2.3 Hz), 138.83 (d, ⁴J_C-F = 3.5 Hz), 131.62, 131.57, 130.99, 130.97, 129.77 (d, ³J_C-F = 13.8 Hz), 129.59 (d, ³J_C-F = 12.0 Hz), 125.46, 125.44, 123.74, 123.71, 123.48, 117.96, 117.59, 116.25 (d, ²J_C-F = 18.7 Hz), 116.17 (d, ²J_C-F = 21.4 Hz), 109.45; MS (ESI): m/z = 315.0 (M + H)⁺.

1.2. Biological activity

Experimental details for all biological assays are provided in Supplementary materials.

2. Results & discussion

2.1. Compound design

Schmitt et al. reported the discovery of several 5-hydroxybenzothiophene ketones as dual Clk1/Dyrk1A inhibitors. Compounds (Cpds.) I and II were the two most potent dual inhibitors (Figure 1), both exhibiting a relatively lower activity against haspin (83 and 69% inhibition at a screening dose of 5 μM, respectively) [24]. On the other hand, the Weinreb amide (III) having a 5-methoxy group instead of the 5-hydroxy at the benzothiophene scaffold retained the inhibitory activity against Clk1 (IC₅₀ = 100 nM) but had decreased potency against Dyrk1A (IC₅₀ = 900 nM) (Figure 1). Therefore, the 5-hydroxy substitution in this class of compounds was identified as a key feature for targeting multiple kinases.

Figure 1.

Rationale for compound design along with the structure of target scaffolds.

Later, Engel and co-workers introduced different structural extensions to the amide function of compound III leading to a huge boost in potency and selectivity against Clk1; (cpd. IV) [44] having a methylated benzyl amide, cpd. V [19] having an imide linker and cpd. VI [40] having a hydrazide linker (Figure 1).

In this work we based our modifications on the 5-hydroxybenzothiophene scaffold to target Clk1, Dyrk1A and haspin through extending the carbonyl linker in the previously reported 5-hydroxybenzothiophene ketones (compounds I and II) [24] to an amide, a hydrazide, or a hydrazone linker, along with various aliphatic, mono-substituted or di-substituted aryl extensions (Figure 1, right panel).

2.2. Chemistry

Amides (1–12) were synthesized from 5-hydroxybenzothiophene-2-carboxylic acid (C). Synthesis of C involved a tripartite synthetic process that was adapted from the methodology described by Misra et al [41]. Rhodanine was first reacted with 3-hydroxybenzaldehyde in presence of glacial acetic acid and sodium acetate to obtain the 3-hydroxybenzylidenerhodanine (A), which was then subjected to alkaline hydrolysis yielding the β-substituted-α-mercaptoacrylic acid (B). This was followed by the oxidative cyclization of (B) using iodine, affording the 5-hydroxybenzothiophene-2-carboxylic acid (C) in a good yield. Intermediate C was then coupled to different aliphatic amines and mono-substituted anilines in the presence of EDC and DMAP to yield the final amides (1–12) in different yields (Figure 2).

Figure 2.

Synthesis of cpds. (1–12). Reagents and conditions: (i) 0.17 equiv. of anhydrous sodium acetate, 1 mL glacial acetic acid in toluene, reflux, 4 h; (ii) 15% aqueous NaOH, 60–70°C, 1 h; (iii) 1.5 equiv. of iodine in dry THF, reflux, 42 h; (iv) 2 equiv. of EDC, 2 equiv. of DMAP and 3 equiv. of the appropriate amine or aniline in minimal volume of dry DMF, RT, overnight.

RT: Room temperature.

As for the 5-hydroxybenzothiophene-2-hydrazide derivatives, they were obtained from the 5-methoxybenzothiophene-2-carboxylic acid (F). Intermediate F was synthesized using the same procedure as intermediate (C), but 3-methoxybenzaldehyde was used in the first step. Intermediate F was then converted to the corresponding acid chloride by reacting it with oxalyl chloride in presence of catalytic amounts of DMF. The acid chloride was then coupled with different phenyl hydrazine hydrochloride derivatives in presence of DIPEA to yield hydrazides (13a–19a), which were then demethylated using BBr₃ to yield the desired 5-hydroxybenzothiophene-2-hydrazides (13b–19b) (Figure 3).

Figure 3.

Synthesis of compounds 13b–19b. Reagents and conditions: (i) 0.17 equiv. of anhydrous sodium acetate, 1 mL glacial acetic acid in toluene, reflux, 4 h; (ii) 15% aqueous NaOH, 60–70°C, 1 h; (iii) 1.5 equiv. of iodine in dry THF, reflux, 42 h; (iv) 4 equiv. of oxalyl chloride, catalytic DMF in dry DCM, reflux, 3 h; (v) 1 mL DIPEA, 2 equiv. of the appropriate phenyl hydrazine hydrochloride in dry THF, RT, 18 h; (vi) 5 mL BBr₃ in DCM, RT, overnight.

Cpd: Compound; RT: Room temperature.

To synthesize the benzylidene hydrazide derivatives, intermediate F was converted to the corresponding 5-methoxybenzothiophene-2-hydrazide (G) through esterification using methanol, followed by hydrazinolysis. Intermediate G was then reacted with 3-fluoro or 4-fluorobenzaldehyde in the presence of a catalytic amount of glacial acetic acid to give the benzylidene hydrazides (20a and 21a), which were then demethylated using BBr₃ to produce compounds 20b and 21b (Figure 4).

Figure 4.

Synthesis of compounds 20b and 21b. Reagents and conditions: (i) 0.4 mL of acetyl chloride in 50 mL methanol, reflux, 5 h; (ii) 5.5 mL hydrazine hydrate, reflux, 3 h; (iii) 2–3 drops of glacial acetic acid, 1.5 equiv. of the appropriate aldehyde in 15 mL of absolute ethanol, reflux, 3 h; (iv) 5 mL BBr₃ in HPLC DCM, RT, overnight.

Cpds: Compounds; HPLC: High-performance liquid chromatography; RT: Room temperature.

2.3. Biological activity

2.3.1. Multi-target kinase activity

All the final 5-hydroxybenzothiophene-2-carboxamide, hydrazide, and benzylidene hydrazide derivatives (cpds. 1–12 and 13b–21b) were tested for their ability to inhibit Clk1, Dyrk1A and haspin in vitro at an initial screening dose of 1 μM ([ATP]=Km). IC₅₀ values were determined for compounds showing more than 60% inhibition against the tested kinases (Table 1).

Table 1. In vitro kinase activity against Clk1, Dyrk1A and haspin.

Cpd. no.	R	% inh at 1 μM Clk1† (IC50)‡ (nM)	% inh at 1 μM Dyrk1A† (IC50)‡ (nM)	% inh at 1 μM haspin† (IC50)‡ (nM)	Cpd. no.	R	% inh at 1 μM Clk1† (IC50)‡ (nM)	% inh at 1 μM Dyrk1A† (IC50)‡ (nM)	% inh at 1 μM haspin† (IC50)‡ (nM)
1	n-hexyl	31	16	24	12	4-chlorophenyl	11	1	43
2	cyclohexyl	8	6	14	13b	3-chlorophenyl	72 (620 ± 11)	45	32
3	cyclopropyl	29	12	40	14b	4-chlorophenyl	38	24	n.d.
4	3-hydroxyphenyl	56	24	64 (793 ± 45)	15b	4-fluorophenyl	73 (634 ± 33)	45	82 (460 ± 26)
5	2-hydroxyphenyl	18	4	32	16b	3-fluorophenyl	89 (163 ± 5)	68 (353 ± 11)	85 (125 ± 3)
6	4-hydroxyphenyl	19	4	64 (615 ± 21)	17b	2-fluorophenyl	23	7	n.d.
7	4-methoxyphenyl	37	13	70 (673 ± 32)	18b	3,5-difluorophenyl	92 (181 ± 3)	72 (237 ± 8.4)	18
8	3-methoxyphenyl	15	10	34	19b	3,5-dichlorophenyl	25	9	8
9	4-fluorophenyl	30	4	49	20b	3-fluorophenyl	42	45	30
10	3-fluorophenyl	35	14	64 (685 ± 11)	21b	4-fluorophenyl	23	32	36
11	3-chlorophenyl	7	ni	7

^†Data shown are the mean of at least two independent experiments, S.D. ≤10%; ni, no inhibition.

^‡IC₅₀ value ± SD.

Cpd: Compound; inh: Inhibition; n.d.: not determined.

The amide derivatives (1–3) having an n-hexyl, cyclohexyl or a cyclopropyl extension were not active against any of the tested kinases. Expanding the amide function by adding various mono-substituted phenyl rings (cpds. 4–12) resulted in derivatives that remained inactive against Clk1 and Dyrk1A, however, some of these derivatives showed selective inhibition toward haspin. This included mainly meta and/or para substituted analogues namely cpds. 4, 6, 7 and 10 (IC₅₀ ≈ 600–800 nM), which carry either hydroxy, methoxy or fluoro substituents. On the other hand, having the bulkier electron withdrawing m-chloro or p-chloro groups abolished the haspin inhibitory activity.

Our goal of achieving multi-target inhibitors for Clk1, Dyrk1A and haspin was not reached by modifying the carbonyl linker in cpds. I and II to alkyl and aryl amide, therefore we explored the effect of extending the amide function to a hydrazide linker with different mono- or di-halogen substituted phenyl rings (cpds. 13b–19b). Generally, this modification could boost the inhibitory activity against the three tested kinases. Cpd. 16b, the m-fluorophenyl hydrazide emerged as the most potent/balanced multi-target kinase inhibitor (Table 1), followed by 18b (3,5-difluorophenyl hydrazide) which could potently inhibit Clk1 and Dyrk1A but lacked activity against haspin. When the fluoro substituent(s) in compounds 16b and 18b was replaced with chloro, the activity was greatly reduced against the three tested kinases (compounds 13b and 19b, respectively).

Finally, further extension of the hydrazide to a hydrazone diminished the inhibitory activity against the three tested kinases (compare 20b to 16b and 21b to 15b).

2.3.2. Kinase selectivity profiling

The most potent multi-target inhibitor 16b was further tested against a panel of kinases that have similar binding pockets to Clk1, Dyrk1A and haspin (Supplementary Table S1) [17,18,45–47]. Interestingly, 16b turned out to be a highly potent Clk4 inhibitor (IC₅₀ = 11 nM) with unique selectivity vs other Clk isoforms. Cpd. 16b showed 14-fold selectivity for Clk4 over Clk1, the closest isoform to Clk4 in sequence similarity, >50-fold selectivity over Clk2 and >>85-fold selectivity over Clk3.

Furthermore, 16b showed nearly an equipotent inhibition toward Dyrk1B (IC₅₀ = 284 nM), in comparison to that of Dyrk1A. This was an expected outcome as the catalytic domains of Dyrk1A and Dyrk1B share an identity of 85%, and only one difference is present in their ATP binding pockets, where Met240 in Dyrk1A is replaced by Leu192 in Dyrk1B [48]. Co-inhibition of Dyrk1B was in accordance with our strategy as this Dyrk isoform was previously shown to be an anti-cancer target in rhabdomyosarcoma, ovarian cancer, pancreatic cancer and non-small-cell lung cancer [49–52].

Moreover, among the tested kinases, DRAK1 was strongly inhibited by 16b with an IC₅₀ of 87 nM. Overall, the selectivity screen releveled 16b as a novel selective Clk4 inhibitor with significantly lower inhibition for other Clk and Dyrk isoforms. DRAK1 is considered the 2nd best inhibited kinase by 16b with about eightfold lower potency than Clk4.

2.3.3. Inhibition of cancer cell viability

Next, we tested the effects of the most potent multi-target kinase inhibitor 16b and the dual Clk1/Dyrk1A inhibitor 18b on the inhibition of cell viability in various cancer cell lines: HCT-116 (colorectal adenocarcinoma), A549 (lung carcinoma), U87MG (glioblastoma), Hela (cervical carcinoma) and T24 (urinary bladder carcinoma) (Table 2). Cpd. 16b showed a broad-spectrum anti-cancer activity, where the glioblastoma cell line U87MG was the most sensitive with a IC₅₀ of 7.2 μM. This could be attributed to the kinase inhibition profile of 16b, as the previously reported 7-azaindole based compounds that co-inhibited Clk1, Dyrk1A and Dyrk1B efficiently blocked cell proliferation and migration of glioblastoma cells [53]. Furthermore, compound 16b is a highly potent inhibitor of DRAK1, a kinase known for its elevated expression in glioblastoma [54]. On the other hand, the dual Clk1/Dyrk1A inhibitor 18b displayed a different profile with lower growth inhibitory activity against A549 and U87MG cells (Table 2).

Table 2. Growth inhibitory activity of compounds 16b and 18b against cell lines of different origin.

Cell line	HCT-116	A549	U87MG	Hela	T24
Cpd. no.	% Inhibition @ 20 μM†	% Inhibition @ 20 μM† (IC50 (μM)†)	% Inhibition @ 20 μM† (IC50 (μM)†)	% Inhibition @ 20 μM†	% Inhibition @ 20 μM†
16b	51.6 ± 2.24	51 ± 6.81	66.1 ± 1.70 (7.2 ± 0.26)	50.8 ± 2.01	35.5 ± 6.44
18b	54.8 ± 3.38	31.1 ± 6.16	10.5 ± 4.60	45.5 ± 3.65	49.7 ± 3.50

^†Data are expressed as mean ± S.E.M. (n = 3). p < 0.001 compared with DMSO alone.

Cpd: Compound; S.E.M: Standard error of mean.

2.3.4. Cytotoxicity in nontumor cells

Compound 16b was selected to be further tested for its cytotoxic effect on two nontumor cell lines: IEC-6 (rat intestine epithelial cells) and HaCaT (human keratinocytes) (Supplementary Table S2). 16b exhibited low cytotoxicity in both nontumor cells at a screening dose of 20 μM, indicating that compound 16b selectively targets and inhibits the growth of cancer cells with minimal effect on nontumor cells.

2.3.5. Effect on cell cycle

Propidium iodide (PI) staining and flow cytometry were used to determine the cell cycle distribution of U87MG cells after exposure to cpd. 16b for 24 h at 30 and 50 μM doses (Figure 5). Cpd. 16b led to sequestering of the cells in the G2 phase preventing them from entering the mitotic phase (G2/M arrest), as indicated by the doubled number of cells in pre-mitosis upon treatment with 30 μM of 16b in comparison to the control. The result might be explained by the inhibition profile of 16b as a multi-target inhibitor for Clk1/4 and DRAK1 kinases. It has been reported that Clk1 shows the highest expression levels at the G2/M phase [55]. Moreover, genetic knockdown of Clk1 in the Hela, H157 and A549 cell lines led to accumulation of cells with 4N DNA content. In addition, multinucleation was observed upon treating the Hela, H157 and A549 cells with TG003 (a potent dual Clk1/4 inhibitor) [56], which is a sign of defective chromosome segregation [55]. As for DRAK1, it was reported that this kinase and its substrates Sqh/MRLC and Anillin/ANLN regulate mitosis and cytokinesis in gliomas [54].

Figure 5.

Compound 16b causes G2/M arrest in the U87MG cells. U87MG cells were treated with DMSO (0.1%) or 16b (30–50 μM) for 24 h. Cell cycle was examined by propidium iodide (PI) staining and flow cytometry. All data are expressed as the mean ± S.E.M. (n = 3). ##p < 0.01 compared with the DMSO group in G0/G1 phase. ***p < 0.001 compared with the DMSO group in G2/M phase.

2.3.6. 16b induced apoptosis in U87MG cells

Inhibitors of Clk kinases have established roles in induction of apoptosis in cancer cells by depleting their proteosomes [57]. Accordingly, Annexin-V/propidium iodide staining and flow cytometric analysis were performed to determine whether the growth inhibitory activity of 16b on the U87MG cells was accompanied by enhanced cellular apoptosis or not. U87MG cells were incubated with different concentrations of 16b for 24 h, afterward the percentage of apoptotic cells were determined. U87MG cells showed a dose-dependent increase of apoptosis compared with the solvent control, reaching a maximum of approximately 30% at 50 μM (Supplementary Figure S1).

2.3.7. Effect of 16b on the proapoptotic & antiapoptotic markers in U87MG cells

Based on its apoptosis-inducing properties, we investigated how 16b influences the levels of the proapoptotic proteins caspase 3 and Bcl-2-associated X protein (BAX), as well as the antiapoptotic protein Bcl-2 in U87MG cells (Supplementary Figure S2). After treatment of the cells with different concentrations of 16b, a concentration-dependent increase in the levels of the caspase 3 and BAX proapoptotic proteins was observed, as well as a concentration-dependent decrease in the level of Bcl-2 antiapoptotic protein. It has been reported that Clk1 and Dyrk1A overexpression in tumors increased the expression of the antiapoptotic proteins B-cell lymphoma-extra large (Bcl-X_L) and myeloid cell leukemia-1 (Mcl-1; members of the Bcl-2 family) [53], In addition, treatment of PC3 and DU145 cancer cells with TG003 resulted in augmented levels of activated caspase 3 and 7 after 48 h of exposure to the dual Clk1/4 inhibitor [14,58]. The aforementioned previously reported findings might thus correlate the observed effect of 16b on the proapoptotic and antiapoptotic markers in U87MG cells to its multi-target inhibition for Clk1, Clk4 and Dyrk1A kinases.

2.3.8. Wound healing assay

We eventually evaluated the impact of 16b on the wound healing capacity of U87MG cells, utilizing a wound healing assay that is widely used for assessing cell migration in vitro. In this assay, a ‘wound’ is created on a cell monolayer, and the ability of the cells to migrate and close this gap is measured [59]. The cells were treated with doses of 10, 30 and 50 μM of 16b for a period of 24 h. As depicted in Figure 6, there was a notable concentration-dependent reduction in the wound healing ability of the U87MG cells. The assay revealed that wound closure was significantly diminished at 30 and 50 μM of 16b, thus indicating an inhibitory effect of 16b on the migratory capabilities of these cells. It was observed that the overexpression of DRAK1 in head and neck tumors prompted the migration and invasion of tumor cells, while the knockdown of DRAK1 led to reduced migration of glioblastoma cells [60,61]. Likewise, the inhibition of Clk4 might be linked to thwarting cell migration. At least for triple negative breast cancer, it was shown that the pharmacological inhibition of Clk4 significantly decreased the invasion and proliferation of breast cancer cell lines and patient-derived cells [62].

Figure 6.

Effect of 16b on the wound healing ability of U87MG cells. U87MG cells with 90% confluence were scraped with a 200 μl tip followed by incubation with DMSO (0.1%) or 16b (30–50 μM) for 24 h. The wound closure was monitored using light microscope. All data are expressed as the mean ± S.E.M (n = 3). **p < 0.01, ***p < 0.001 compared with the DMSO group.

3. Conclusion

In the present study, we explored the effect of modifying the carbonyl linker in the previously reported 5-hydroxybenzothiophene ketones into an amide, hydrazide or a hydrazone linker on the inhibitory activity toward Clk1, Dyrk1A and haspin in an attempt to develop a multi-target kinase inhibitor with a potential broad-spectrum anti-tumor activity. Multi-target kinase inhibition was not feasible by the hydrazone nor the amide derivatives, but rather selective haspin inhibition was observed for the latter (cpds. 4, 6, 7 and 10). The hydrazide derivative 16b emerged as a potent multi-target kinase inhibitor with IC₅₀s of 163, 353.3 and 125.7 nM against Clk1, Dyrk1A and haspin respectively, with additional inhibitory activities against Clk4 (IC₅₀ = 11 nM), Dyrk1B (IC₅₀ = 284 nM) and DRAK1 (IC₅₀ = 87 nM). 16b displayed growth inhibitory activity against HCT-116, A549, U87MG and Hela cells, showing the highest efficacy against the U87MG cells with a IC₅₀ value of 7.2 μM. 16b could induce G2/M cell cycle arrest and apoptosis in the U87MG cells, in addition, the cell migration capacity of the U87MG cells was inhibited, in line with the simultaneous inhibition of multiple kinases. Thereby, the inhibitory profile of 16b was not promiscuous, as most of the kinases usually co-inhibited by non-selective Clk1/Dyrk inhibitors were not affected. Probably because ‘untouchable’ kinases were not inhibited, the cell viability of nontumor cells was not diminished by 16b.

However, it is important to highlight a few limitations in our work. First, it should be noted that the tumor-promoting role of Clk4, the main target of 16b, has not fully elucidated yet, one reason for that being the lack of inhibitors selectively targeting Clk4 but not other Clk isoforms. Thus, it cannot be assessed whether some of the observed effects on tumor cells in our study were at least in part due to Clk4 inhibition. Second, the compounds displayed in the study harbor a phenolic OH group, which is known to be metabolized extensively via methylation, sulfation and glucuronidation [63], therefore the metabolic stability of 16b might be of concern if it was further tested in in vivo studies.

However, despite these limitations, the ability of 16b to inhibit growth of several cell lines indicates that inhibition of multiple kinases might potentially combat several types of cancers. Future studies will show whether 16b may have higher therapeutic effectiveness than more selective kinase inhibitors targeting only one or two kinases.

Funding Statement

The work was partially supported by E-Da Hospital and the National Science and Technology Council. P-J Chen acknowledges the funding by National Science and Technology Council (112-2320-B-650-001 and 112-2321-B-255-001) E-Da Hospital (EDAHJ112016, EDAHS112009, and NCKUEDA11205).

Supplemental material

Supplemental data for this article can be accessed at https://doi.org/10.1080/17568919.2024.2342708

Author contributions

Writing – review, editing, Writing – original draft, methodology, investigation, formal analysis: All Authors. Funding acquisition: P-J Chen. Project administration, resources, supervision, conceptualization: M Abdel-Halim and M Engel.

Financial disclosure

The work was partially supported by E-Da Hospital and the National Science and Technology Council. P-J Chen acknowledges the funding by National Science and Technology Council (112-2320-B-650-001 and 112-2321-B-255-001) E-Da Hospital (EDAHJ112016, EDAHS112009, and NCKUEDA11205). The authors have no other competing interests or relevant affiliations with any organization or entity with the subject matter or materials discussed in the manuscript apart from those disclosed.

Competing interests disclosure

The authors have no competing interests or relevant affiliations with any organization or entity with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Writing disclosure

No writing assistance was utilized in the production of this manuscript.