Synthesis of thiazolyl-based hydroxamic acids as histone deacetylase inhibitors

Herein we report the synthesis of 4-phenyl substituted thiazolyl-based hydroxamates as histone deacetylase inhibitors. The synthesis of the target compounds is comprised of a Hantzsch-thiazole reaction, a HATU-mediated acylation and a hydroxamic acid synthesis. Preliminary docking results indicated an isozyme selectivity by addressing the lower pocket of HDAC4 with 4-phenyl thiazoles. Surprisingly, this new series of thiazolyl based hydroxamates revealed a moderate HDAC6 inhibitory activity in the low micromolar range. Two of the hydroxamic acids did not show an HDAC4 inhibition up to 100 µM, whilst a third with a 4-pyridinyl CAP moiety displayed a moderate HDAC4 inhibitory activity in the micromolar range


Introduction
The term epigenetic refers to inheritable changes in gene expression without altering the underlying DNA sequence. 1Epigenetic mechanisms are subdivided into four different classes: (1) posttranslational modifications of histones, (2) chemical modifications of DNA (3) chromatin-remodeling complexes, and (4) regulatory non-coding RNAs.An essential post-translational modification is protein acetylation/deacetylation, which is regulated by histone acetyltransferases (HATs) and histone deacetylases (HDACs).4][5][6] Their inhibition provides a promising strategy for the development of novel epigenetic anticancer drugs.][9][10][11][12] The common pharmacophore model for HDACi consists of the following four elements: a zinc binding group (ZBG), a linker and a connecting unit (CU), that interact with the substrate binding tunnel and a cap, also known as surface recognition domain. 13,14According to their ZBG, HDACi can be subdivided into 4 main chemical classes: hydroxamic acids, thiols, carboxylic acids and o-amino anilides. 15gure 1.General pharmacophore model of HDACi and chemical structures of FDA approved HDAC inhibitors 1-4 and their indications.9][10][11][12] Thiazoles are privileged structures in medicinal chemistry.Its core scaffold is found in many natural products such as alkaloids, secondary metabolites and cyclopeptides. 16[23] Initial studies for the develoment of thiazolyl-based HDACi were performed by Anandan et al. (Figure 3). 24n the reported HDACi, the thiazole moiety acts as a linker that bridges the ZBG with the cap group.These HDACi displayed in vitro HDAC inhibitory activity in HeLa cell nuclear extracts and antiproliferative effects towards the breast cancer cell line MCF7 in the micromolar range.

Design of 4-phenyl substituted thiazolyl hydroxamates as HDAC inhibitors
6][27][28] So far, only trifluoromethyloxadiazoles (TFMO), cyclopropane hydroxamates as well as benzhydryl based class IIa HDAC inhibitors have been reported.In published crystal structures of class IIa HDACs, the so-called lower pocket was identified as a distinctive structural feature. 29,30 Bürli et al. and  Luckhurst et al. showed that tri-and tetrasubstituted cyclopropanes, providing a scaffold with a suitable 3Dgeometry that occupy the lower pocket with their phenyl moiety (Scheme 1, circled in red). 25,30Here, we report the synthesis of trisubstituted thiazolyl-based hydroxamates as potential class IIa HDACi.
The analysis of the docking conformation of compound 9a in HDAC4 suggests that, as envisioned, the hydroxamate moiety of 9a can coordinate with the zinc cation in a bidentate manner (Figure 4).In addition, the phenyl substituent of 9a occupied the lower pocket in a similar binding mode as the phenyl moiety of compound 14 (Figure 4)..The docking pose of compound 9a is superimposed with the HDAC4 cocrystallised compound 14 (pink).

Chemistry
The starting material of the Hantzsch thiazole synthesis was the α-bromoester 10 that was prepared according to Borowiecki et al.. 35 After the treatment of 10 with thiourea, the thiazole 11 was formed.In the following step, 11 and the respective carboxylic acids were coupled using HATU under basic conditions yielding the corresponding amides 12a-c.][38][39][40] Therefore, the hydroxamic acids 9a-c were synthesized by ester hydrolysis and subsequent conversion of the respective carboxylic acids with HATU, DIPEA and hydroxylamine hydrochloride (Scheme 2).The ester hydrolysis was achieved by the addition of an excess of potassium hydroxide.However, the same protocol was unsuccessful when applied for the cleavage of the corresponding ethyl esters.This finding underscores that the hydrolysis depends on the nature of the alkyl substituent of the ester moiety.Scheme 2. Synthesis of the target structures 9a-c.

Biological Evaluation
The synthesised compounds 9a-c were assessed for their antiproliferative activity and for their HDAC inhibitory activity in the human monocytic cell line THP-1 using class-distinguishing substrates (Boc-Lys(Ac)-AMC: class I and IIb HDACs; Boc-Lys(TFAc)-AMC: class IIa, HDAC8).The results are depicted in Table 1 with vorinostat as reference compound.Among the tested compounds, 9b exhibited the highest antiproliferative activity with 61.5% of growth inhibition at 100 µM concentration.The introduction of a 3,5-dimethoxyphenyl (9a) or biphenyl (9c) CAP moiety led to a decreased cytotoxicity with antiproliferative effects below 50% at 100 µM.In a whole-cell HDAC inhibition assay using the Boc-Lys(Ac)-AMC-HDAC substrate, all screened hydroxamic acids 9a-c displayed inhibitory activities below 40% at 100 µM.However, the cellular HDAC inhibition assay using the Boc-Lys(TFAc)-AMC-HDAC substrate revealed a slightly increased inhibition with over 45% at 100 µM for 9a and 9b.
Overall there is a good correlation between the antiproliferative activity and the cellular HDAC inhibition.All tested compounds 9a-c showed a moderate cytotoxicity and cellular HDAC inhibition.
For further evaluation, the synthesized compounds 9a-c were screened against recombinant human HDAC2, HDAC4, HDAC6 and HDAC8 (Table 2).Vorinostat was included as a pan-HDACi reference compound.Interestingly, the new thiazolyl-based hydroxamates 9a-c were identified as moderate HDAC6 inhibitors in the low micromolar range.In contrast to our preliminary in-silico data, the hydroxamic acids 9a and 9c displayed no HDAC4 inhibition in the tested concentration range (≤ 100 µM).However, the hydroxamic acid 9b exhibiting a 4-pyridinyl CAP moiety demonstrated a moderate micromolar activity against HDAC4 (IC 50 (HDAC4) = 48.8µM).

Conclusions
Histone deacetylases are zinc-dependent metalloproteases that catalyse the removal of acetyl functional groups from lysine residues of both histone and nonhistone proteins.HDACs are involved in a multitude of biological processes e.g. in cell cycle progression, cell survival, apoptosis and differentiation.They are clinically validated targets for the treatment of cancer.Selective HDAC inhibitors may serve as important tools for elucidating the role of specific HDACs in certain diseases.Contrary to our preliminary qualitative in-silico screening, all synthesized compounds 9a-c exhibited a moderate HDAC6 inhibitory activity in the low micromolar range.Only the hydroxamic acid 9b exhibiting a 4-pyridinyl CAP moiety displayed a moderate micromolar activity against HDAC4 (IC 50 (HDAC4) = 48.8µM).

Experimental Section
General.All chemicals and solvents were purchased from commercial suppliers (Sigma Aldrich, Alfa Aesar, Fluorochem, TCI, abcr and Acros Organics) and used without further purification.All anhydrous reactions were carried out in flame-dried Schlenk-flasks and under argon atmosphere.Dry solvents were used directly from Seal® bottles from Acros Organics.Analytic Thin Layer Chromatography (TLC) was carried out with Macherey Nagel precoated silica gel plates (ALUGRAM® Xtra SIL G/UV 254 ).Detection was achieved with ultraviolet irradiation (254 nm) and/or staining with potassium permanganate solution (9 g KMnO 4 , 60 g K 2 CO 3 , 15 mL of a 5% aqueous NaOH-solution, and 900 mL demineralised water).Flash column chromatography was performed with CombiFlashRf200 (TeleDyneIsco) with the solvent mixtures specified in the corresponding procedure.

General procedure 1. HATU-coupling
For the synthesis of the amides 12a-c, 1.00 eq of the respective carboxylic acid was dissolved in dry DMF (0.1 mmol/mL) and 1.00 eq HATU and 2.00 eq DIPEA were added.After stirring for 15 min at rt, methyl 2-amino-4phenylthiazole-5-carboxylate (11) (1.00 eq) was added and the resulting mixture was stirred for 16h at 60 °C.

General procedure 2: Methyl ester hydrolysis
For the synthesis of the carboxylic acids 13a-c, 1.00 eq of the methyl ester 12a-c was dissolved in THF (0.1 mmol/mL) and 30.0 eq potassium hydroxide was added.The resulting mixture was refluxed for 16 h.The solvent was removed in vacuo and the residue was resuspended with EtOAc.The organic layer was washed with saturated NaHCO 3 (3x 50 mL).The combined aqueous phases were acidified (pH = 2-3) and the resulting precipitate (product) was filtered off and washed with water.

General procedure 3. Hydroxamic acid formation
For the synthesis of the hydroxamic acids 9a-c, 1.00 eq of the respective carboxylic acid 13a-c was dissolved in dry DMF (0.1 mmol/mL), 1.00 eq HATU and 3.20 eq DIPEA were added.After stirring for 15 min at rt, 1.20 eq hydroxylamine hydrochloride was added and the resulting mixture was stirred for further 16 h at 60 °C.The reaction mixture was then poured into water and the resulting participate was then filtered and washed with aqueous HCl-solution (pH = 4) and water.The crude product was purified by flash column chromatography (prepacked silica cartridge, DCM/30% methanol in DCM) to provide the corresponding hydroxamic acids.

Table 1 .
Cell viability (MTT assay) and whole-cell HDAC inhibition assay in the human monocytic cell line THP-1 Presented data are calculated from at least two experiments each performed in duplicates.IC 50 values were calculated for compounds with an inhibition of more than 50%.Standard deviation of percent inhibition values is less than 14%.Vehicle control was defined as 0% inhibition.

Table 2 .
Inhibitory activity of 9a-c and Vorinostat on recombinant human HDAC2, HDAC4, HDAC6 and HDAC8 Data shown are from at least two experiments each performed at least as duplicates and the IC 50 value of pooled data is reported when IC 50 < 100 µM.