Design and synthesis of a tetrahydroisoquinoline-based hydroxamate derivative (ZYJ-34v), an oral active histone deacetylase inhibitor with potent antitumor activity

Abstract

In our previous study we developed a novel series of tetrahydroisoquinoline-based hydroxamic acid derivatives as histone deacetylase (HDAC) inhibitors (Bioorg. Med. Chem., 2010, 18, 1761–1772., J. Med. Chem., 2011, 54, 2823–2838.), among which compound ZYJ-34c (1) was identified and validated as the most potent one with marked in vitro and in vivo antitumor potency (J. Med. Chem., 2011, 54, 5532–5539.). Herein further modification of 1 afforded another oral active analogue ZYJ-34v (2) with simplified structure and lower molecular weight. Biological evaluation of compound 2 showed efficacious inhibition against HDAC1, 2, 3 and 6, which was confirmed by western blot analysis results. Most importantly, compound 2 exhibited similar even more potent in vitro and in vivo antitumor activities relative to the approved HDAC inhibitor SAHA.

Introduction

Epigenetic covalent modifications of histone and DNA dramatically affect gene expression and cellular activity, and epigenetic disorder could lead to many diseases, especially cancer. Generally, epigenetic modifications are reversible, catalyzed by pairs of enzymes with converse activity. Among these enzymes, histone deacetylases (HDACs) have attracted the most attention over the past decades (1).

HDACs are amidohydrolases with multiple functions. In cell nuclear, HDACs catalyze the deacetylation of lysine residues located at the N-terminal of nucleosomal histones, which results in condensed chromosomal DNA and transcriptional repression (2–4). Over the past few years, kinds of non-histone proteins involved in cell growth and survival pathways have been identified as HDAC substrates, such as transcription factors, cytoskeletal proteins, molecular chaperones and nuclear import factors, etc (5).

Eighteen HDAC isoforms, grouped into 4 classes, are encoded in the human genome. The enzymes of classes I (HDACs 1–3 and 8), II (HDACs 4–7, 9 and 10) and IV (HDAC 11) are Zn²⁺-dependent metallohydrolases, whereas the class III HDACs (sirtuins 1–7) utilize NAD⁺ as a cofactor. It has been revealed that Zn²⁺-dependent isozymes, especially class I and class II HDACs, are involved in tumorigenesis and development, and inhibition of these isoforms could result in proliferation inhibition, cellular differentiation, apoptosis, susceptibility to chemotherapy, anti-angiogenesis and migration inhibition of tumor cells (6). In this context, categories of HDACs inhibitors (HDACI) against Zn²⁺-dependent HDACs have been developed. Currently, over twenty HDACI are in clinical trials as antitumor agents and two of them, SAHA (Figure 1) and FK228 (Figure 1), are already on the market (7).

Figure 1.

Most HDAC inhibitors, including the intracellular active form of FK228 (Figure 1), share a common pharmacophore consisting of three domains: a zinc-binding group (ZBG) that coordinates to the catalytic Zn²⁺ at the bottom of the active site, a cap group able to interacts with the rim of active site entrance, and a linker that occupies the active site tunnel and tethers ZBG to cap (Figure 1).

According to the HDACI pharmacophore, we previously designed and synthesized a novel series of tetrahydroisoquinoline-based hydroxamate derivatives as HDACI (Figure 2) (8–10). Among these analogues, compound ZYJ-34c (1, Figure 2) was identified and validated as the most potent one with marked in vitro and in vivo antitumor potency through several rounds of structural optimization and activity screening (10). In the present study, compound ZYJ-34v (2, Figure 2) was designed and synthesized by replacing the R₂ group of 1 with the valproyl group. There are three reasons for us to attempt this replacement:

Figure 2.

1. Valproic acid (VPA) is a HDACI in clinical trials. Considering its hydroxamic acid analogue valpropylhydroxamic acid exhibited lower class I HDACs inhibitory activity compared with VPA (11), we hypothesized that VPA inhibited HDACs by its valproyl group binding with HDACs rather than by its carboxylic acid group chelating with Zn²⁺. The binding site of VPA in HDACs might be the same binding site of other co-factros. Therefore, VPA could prevent HDACs from interaction with other proteins and the subsequent formation of multi-protein complex, which are crucial for the function of HDACs (6). Moreover, VPA may possibly act as an allosteric inhibitor of HDACs. Based on the above hypothesis, we hoped that the introduction of valproyl group could increase the affinity between 2 and HDACs.

2. There is a possibility that the in vivo or intracellular partial hydrolysis of 2 could release free VPA and compound 3 (Figure 3), which was identified as a moderately potent HDACI in our previous research (8). If so, there will be three active components (unhydrolyzed ZYJ-34v, VPA and compound 3) performing multiple antitumor effects.

3. Compared with 1, compound 2 possesses simplified structure, less chiral center and lower molecular weight, which are beneficial to its further research and development.

Figure 3.

Compound 2 was evaluated against HDAC1, 2, 3, 6 using Boc-Lys (acetyl)-AMC substrate and against MDA-MB-231 cell lysate using Boc-Lys (triflouroacetyl)-AMC substrate to profile its HDACs inhibitory activity and selectivity. Western blot analysis and in vitro antiproliferative assays were performed to validate its intracellular HDACs inhibitory activity. Moreover, in vivo antitumor activity of 2 was assessed in a human breast carcinoma (MDA-MB-231) xenograft model. SAHA was used as reference drug in all above-mentioned assays.

Materials and Methods

Chemistry

Compound 2 was synthesized following the procedures described in Scheme 1. The starting material compound 4 could be smoothly obtained according to our previously published methods (9). Acylation of 4 with valproic acid and subsequent treatment with NH₂OK led to the target compound 2 (Scheme 1).

Scheme 1. Reagents and conditions.

(a) i). TBTU, Et3N, THF; ii). NH₂OK, CH₃OH, 32% for two steps.

In vitro HDACs inhibition fluorescence assay

The HDACs inhibition fluorescence assays were conducted by the similar method as previously described (9). Boc-Lys (acetyl)-AMC substrate was used in inhibition assays against class I (HDAC1, HDAC2, HDAC3) and class IIb (HDAC6), while Boc-Lys (triflouroacetyl)-AMC substrate for class IIa (MDA-MB-231 cell lysate). In brief, 10 μL of enzyme solution was mixed with various concentrations of tested compound (50 μL). Five minutes later, 40 μL of fluorogenic substrate was added, and the mixture was incubated at 37 °C for 30 min and then stopped by addition of 100 μL of developer containing trypsin and TSA. After incubation at 37 °C for 20 min, fluorescence intensity was measured using a microplate reader at excitation and emission wavelengths of 390 nm and 460 nm, respectively. The inhibition ratios were calculated from the fluorescence intensity readings of tested wells relative to those of control wells, and the IC₅₀ values were calculated using a regression analysis of the concentration/inhibition data.

In vitro antiproliferative assay

In vitro antiproliferative assay was determined by the MTT (3-[4,5-dimethyl-2-thiazolyl]-2,5-diphenyl-2H-tetrazolium bromide) method as previously described (9). Briefly, all cell lines were maintained in RPMI1640 medium containing 10% FBS at 37 °C in 5% CO₂ humidified incubator. Cells were passaged the day before dosing into a 96-well cell plate, allowed to grow for a minimum of 4 h prior to addition of compounds. After compounds addition, the plates were incubated for an additional 48 h, and then 0.5% MTT solution was added. After further incubation for 4 h, formazan formed from MTT was extracted by DMSO for 15 min. Absorbance was then determined using an ELISA reader at 570 nm and the IC₅₀ values were calculated according to the inhibition ratios.

Western blot analysis

For western blot analysis of acetylated tubulin, acetylated histone H3, H4 and p21, total protein extracts were separated on a polyacrylamide gel, transferred onto PVDF membranes and blotted as previously described (12). Protein immunoblots were shown for acetylated tubulin (Sigma), acetylated histone H3 (Sigma), acetylated histone H4 (Sigma) and p21 (Cell Signaling) using β-Actin (Sigma) as a loading control.

In vivo MDA-MB-231 xenograft models

In vivo MDA-MB-231 xenograft model was established as previously described (9). In brief, conventionally cultured MDA-MB-231 cells were inoculated subcutaneously in the right flanks of female athymic nude mice (BALB/c-nu, 5–6 weeks old, Slac Laboratory Aniamal, Shanghai, China). About ten days after injection, tumors were palpable (about 100 mm³) and mice were randomized into treatment and control groups (7 mice per group). The treatment groups received 90 mg/kg of compound 2 or SAHA by oral gavage for 22 consecutive days, and the blank control group received an equal volume of PBS solution containing DMSO. During treatment, subcutaneous tumor volume and body weight were monitored regularly. After treatment, mice were sacrificed and dissected to weigh the tumor tissues, livers and spleens. All the obtained data were used to evaluate the antitumor potency and toxicity of compounds. Data were analyzed by Student’s two-tailed t test. A P level < 0.05 was considered statistically significant.

Results and Discussions

Chemistry

All reactions were monitored by TLC with 0.25 mm silica gel plates (60GF-254). UV light, iodine stain and ferric chloride were used to visualize the spots. Silica gel or C18 silica gel was used for column chromatography purification. All commercially available starting materials, reagents and solvents were used without further purification unless otherwise stated. Melting points were determined uncorrected on an electrothermal melting point apparatus. ¹H NMR spectra were recorded on a Bruker DRX spectrometer at 600 MHz, δ in parts per million and J in hertz, using TMS as an internal standard. High-resolution mass spectra were conducted by Shandong Analysis and Test Center in Ji’nan, China. ESI-MS spectra were recorded on an API 4000 spectrometer. Compound 2 are >95% pure by HPLC analysis, performed on a Agilent 1100 HPLC instrument using a Phenomenex Synergi 4 μ Polar-RP 80A column (250 mm × 4.6 mm), eluted with 50% acetonitrile/50% water (containing 0.1% formic acid) over 30 min, with detection at 254 nm and a flow rate of 1.0 mL/min.

(S)-Methyl 2-((3-((4-methoxyphenyl)carbamoyl)-1,2,3,4-tetrahydroisoquinolin-7-yl)oxy)acetate (4) was synthesized according to the methods in our previous work (9).

(S)-7-(2-(hydroxyamino)-2-oxoethoxy)-N-(4-methoxyphenyl)-2-(2-propylpentanoyl)-1,2,3,4-t etrahydroisoquinoline-3-carboxamide (2)

At room temperature, to a solution of VPA (5) (1.32 g, 9.2 mmol) in anhydrous THF (40 mL), was added Et₃N (1.02 g, 10 mmol) followed by 2-(1H-benzotriazole-1-yl)-1,1,3,3-tetramethyluronium tetrafluoroborate (TBTU, 3.24 g, 10 mmol). After 15 min, the amine compound 4 (3.33 g, 9.0 mmol) was added. Stirring was continued until compound 4 disappeared by TLC, then THF was evaporated with the residue being taken up in EtOAc (50 mL). The EtOAc solution was washed with saturated Na₂CO₃ (3 × 10 mL), 1 N HCl (3 × 10 mL) and brine (3 × 10 mL), dried over MgSO₄, and evaporated under vacuum. The obtained crude product was treated with a solution of NH₂OK in anhydrous methanol for 1 h, and then the solvent was evaporated under vacuum. The residue was acidified with 2 N HCl until pH 5–6 then extracted with EtOAc (3 × 30 mL). The organic layers were combined, washed with brine (3 × 20 mL), dried over MgSO₄ and evaporated with the residue being purified by C18 reversed-phase column chromatography (H₂O/MeOH 3:7) to give desired compound 2 (1.75 g, 39% yield) as a white powder. Mp: 96–98 °C. ¹H-NMR (DMSO-d₆, 600 MHz) δ 0.67–0.89 (m, 6H, CH₃CH₂CH₂CHCH₂CH₂CH₃), 1.06–1.55 (m, 8H, CH₃CH₂CH₂CHCH₂CH₂CH₃), 2.96–3.17 (m, 3H, CH₂CHCH₂ and PhCH₂CH), 3.70 (s, 3H, OCH₃), 4.45 (s, 2H, OCH₂CO), 4.49–4.94 (m, 3H, NCHCO and PhCH₂N), 6.74–6.92 (m, 4H, benzene protons), 7.11–7.15 (m, 1H, benzene proton), 7.36–7.42 (m, 2H, benzene protons), 8.96 (s, 1H, NHOH), 9.88 (s, 1H, PhNH), 10.82 (s, 1H, NHOH). HRMS (AP-ESI) m/z calcd for C₂₇H₃₆N₃O₆ [M+H]+ 498.2604, found 498.2611. Retention time: 6.3 min.

HDACs inhibition assay

To compare their HDACs inhibitory activity and isoform selectivity, compound 2 and SAHA were tested against HDAC1, HDAC2, HDAC3 and HDAC6 using acetylated substrate. Besides, the class IIa inhibitory activity was evaluated against MDA-MB-231 cell lysate using class IIa-specific triflouroacetylated substrate (13). Results listed in Table 1 showed that compared to SAHA, compound 2 exhibited superior inhibitory capacities against HDAC1 and HDAC2, while inferior inhibitory capacities against HDAC3 and HDAC6. Neither compound 2 nor SAHA exhibited obvious inhibition against class IIa HDACs up to 10 μM, which was in line with literature information that SAHA was not active against class IIa HDACs (14). Overall, compound 2 was equipotent to SAHA.

Table 1.

HDACs inhibitory activity and isoform selectivity of compound 2 and SAHA^a

Cpd	class I			class IIb	class IIa
	HDAC1	HDAC2	HDAC3	HDAC6	cell lysate
2	37.6±12.1	174.8±6.2	87.7±11.7	675±125	NAβ
SAHA	75.6±11.1	256.4±2.1	28.4±11.2	118±11.9	NAβ

^aAssays were performed in replicate (n≥2); IC₅₀ (nM) values are shown as mean ± SD.

^βNA: not active at 10 μM.

Western blot analysis

We also confirmed that compound 2 was cell permeable and able to inhibit intracellular even nuclear HDACs by monitoring the acetylation levels of tubulin, histones H3 and H4 in MDA-MB-231 cell line. Acetylated tubulin is a known target of HDAC6 and histones are the common targets of HDAC1 and HDAC2. Moreover, the effect on the expression level of the cyclin-dependent kinase (CDK) inhibitor p21 was also investigated. Silencing of the tumor suppresor gene p21 through hypoacetylation is a hallmark of many cancers, and HDACs inhibition in the nucleus could induce apoptosis via re-establishing expression of p21 (15). As indicated in Figure 4, both compound 2 and SAHA effectively inhibited deacetylation of tubulin, histone H3 and H4, and significantly induced increase in the protein level of p21 at 1 μM after 24 h of treatment.

Figure 4.

In vitro antiproliferative activity assay

The potent HDACs inhibitory activities of compound 2 promoted us to evaluate its in vitro antiproliferative activity against several tumor cell lines (Table 2). Overall, the antiproliferative activity of compound 2 was similar even superior to that of SAHA, which correlated well with aforementioned biological test results.

Table 2.

Antiproliferative activity of compound 2 with SAHA as positive control

Compd	IC50 (μM) a
	MDA-MB-231	HCT116	PC-3	HepG2	MCF-7
2	1.41	0.63	2.62	4.23	4.76
SAHA	1.52	0.60	3.97	6.29	3.44

^aValues are the mean of three experiments. The standard derivations are <20% of the mean.

In vivo antitumor activity assay

Encouraged by its promising in vitro activity, the in vivo antitumor potency of compound 2 was evaluated in a subcutaneous MDA-MB-231 xenograft model. Tumor growth inhibition (TGI) and relative increment ratio (T/C) were used as the evaluation indicators to reveal the antitumor effects in tumor weight and tumor volume, respectively.

TGI = (the mean tumor weight of control group − the mean tumor weight of treated group)/the mean tumor weight of control group.

Tumor volume (V) was estimated using the equation (V = ab²/2, where a and b stand for the longest and shortest diameter, respectively). T/C was calculated according to the following formula:

$$\text{T}/\text{C}=\text{the}\text{mean}\text{RTV}\text{of}\text{treated}\text{group}/\text{the}\text{mean}\text{RTV}\text{of}\text{control}\text{group}.$$

RTV, namely relative tumor volume = V_t/V₀, (V_t: the tumor volume measured at the end of treatment; V₀: the tumor volume measured at the beginning of treatment)

The data present in Table 3, tumor growth curve (Figure 5a) and the final tumor volume (Figure 5b) showed that compound 2 (TGI = 59%, T/C = 49%) possessed comparable in vivo antitumor activity to the approved drug SAHA (TGI = 56%, T/C = 52%). According to the final body weight and the dissected liver and spleen weight of treated mice (Table 3), there was no evidence that compound 2 was more toxic than SAHA. In fact, during experiment no obvious body weight loss was detected in mice treated by compound 2.

Table 3.

Effects of compounds on a mouse MDA-MB-231 xenograft model

Compoud	Body weight (g)	Liver weight (g)	Spleen weight (g)	TGIa	T/Ca
Blank Control	25.37±1.27	2.1±0.43	0.2±0.07	-	-
SAHA (90 mg/kg, po)	23.23±2.08	1.69±0.2	0.17±0.06	56%	52%
2 (90 mg/kg, po)	25.61±1.41	1.67±0.58	0.16±0.06	59%	49%

^aCompared with the control group, all treated groups showed statistically significant (P < 0.05) TGI and T/C by Student’s two-tailed t test.

Figure 5.

Conclusion and Future Directions

Structural modification of our previously discovered potent HDACI compound 1 led to another oral active HDACI compound 2 with efficacious inhibitory activity against HDAC1, HDAC2, HDAC3 and HDAC6. Compared with the approved HDACI SAHA, compound 2 exhibited similar even superior in vitro antiproliferative activity and in vivo antitumor potency in a MDA-MB-231xenograft model. Compared with its parent compound 1, compound 2 possesses more convenient synthetic route and less chiral center, which facilitates its further research and development as a new drug candidate. Currently, related pharmacokinetic research of compound 2 is underway to find out if VPA could be released by in vivo metabolism of compound 2. On the other hand, the simplified structure and lower molecular weight of compound 2 leave a margin of subsequent structural modification to further increase its activity. We are endeavouring to search more potent HDACI using compound 2 as the lead.

Abbreviations

HDAC

histone deacetylase

HDACIs

histone deacetylase inhibitors

ZBG

zinc binding group

CDK

cyclin-dependent kinase

VPA

valproic acid