Screening of amide analogues of Trichostatin A in cultures of primary rat hepatocytes: search for potent and safe HDAC inhibitors
- First Online:
- Cite this article as:
- Fraczek, J., Deleu, S., Lukaszuk, A. et al. Invest New Drugs (2009) 27: 338. doi:10.1007/s10637-008-9180-x
- 74 Views
The vast majority of preclinical studies of HDAC inhibitors (HDAC-I) focus on the drug–target (cancer) cell interaction, whereas little attention is paid to the effects on non-target healthy cells, which could provide decisive information to eliminate potential cytotoxic compounds at a very early stage during drug development. In the current study we used cultures of primary rat hepatocytes as a read out system to select for the most potent HDAC-I in the group of structural analogues of an archetypal HDAC-I, namely Trichostatin A. This kind of approach allowed selecting compounds with high biological activity and with no apparent toxicity towards cultured hepatocytes.
KeywordsHistone deacetylase inhibitorsTrichostatin AHepatotoxicityPrimary rat hepatocytes
During the last years HDAC-I have become one of the most promising classes of chemotherapeutics with profound anti-tumour effects . They inhibit proliferation and trigger differentiation and/or apoptosis in neoplastic cells. Recently, also anti-metastatic, anti-angiogenic and immunomodulatory properties of HDAC-I have been described [2–4]. These findings suggest, that in the future HDAC-I could become ultimate anticancer drugs, and stimulate the search for novel, more potent and selective HDAC-inhibiting agents.
In order to select interesting compounds with a favourable pharmacological profile, among a variety of potential HDAC-I, comprehensive analyses are needed. Qualification criteria typically include IC50 values and examination of pharmaco-dynamic and pharmaco-efficacy biomarkers in cancer-derived cell lines. It is remarkable, that at that stage, harvesting of preliminary pharmaco-toxicological data is usually disregarded. This may be a costly mistake. Indeed, HDAC-I similarly to other groups of current chemotherapeutics, do not target cancer-specific molecules/markers but a group of enzymes which regulate essential cellular processes in both diseased and healthy cells . The effectiveness of HDAC-I relies on the higher sensitivity of the former cells towards these agents, due to the general disturbance of their cellular balance . Potential adverse effects in non-target cells, however, may not be excluded. Therefore, here we present a screening method for novel HDAC-I based on the use of healthy primary cells, namely metabolically competent hepatocytes.
Hepatocytes, are by number the largest cell population in the liver . They posses a battery of xenobiotic-metabolizing enzymes and hence are central in drug metabolism. In fact, the majority of drugs is metabolised in the liver. Oral drugs in particular are absorbed by the gut and transported via portal circulation directly to the liver. Consequently, hepatocytes are often the primary cell type within the organism to be affected by deleterious effects of pharmaceuticals . From this point of view, hepatocytes occupy a vital position in in vitro pharmaco-toxicological research . In comparison to currently commercially available hepatic cell lines, such as HepG2, primary hepatocytes demonstrate the closest reflection of in vivo hepatic metabolism [10, 11]. The former cell lines not only lack or have diminished expression of hepatic-specific transcription factors and xenobiotic-metabolising enzymes, but are also characterised by a disturbed regulation of the basic cell cycle and cell death control mechanisms, due to their cancer-related origin . Pioneer research of our group could show that HDAC-I affect a number of cellular processes, namely cell proliferation, differentiation and apoptosis in primary hepatocytes [12–15]. The underlying molecular mechanisms were, however, shown to be different from those involved in cancer cells . Trichostatin A (TSA), a hydroxamic acid-based histone HDAC-I, was shown in cultured adult rat hepatocytes; (1) to inhibit cell cycle re-entry, triggered by the isolation of these cells from the liver, (2) to improve xenobiotic biotransformation capacities and albumin secretion, both indicative of hepatocyte functionality, and (3) to increase cell survival by delaying apoptosis [12–14]. These properties were used here to evaluate the biological activity of several HDAC-I candidates in a primary rat hepatocytes as a read-out system.
Materials and methods
Crude collagenase type I, bovine serum albumin (BSA), bovine insulin, L-glutamine, glucagon, ethylene diamine tetraacetic acid (EDTA), phenylmethanesulphonylfluoride (PMSF), pepstatin A, leupeptin, aprotinin, 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS), piperazine-N,N′-bis(2-ethanesulfonic acid) (PIPES), dithiotreitol (DTT) Trichostatin A (TSA), were purchased from Sigma-Aldrich (Bornem, Belgium); recombinant human epidermal growth factor (EGF) was from Promega (Leiden, The Netherlands). Williams’ Medium E came from Gibco (Belgium), Minimum Essential Medium (MEM) and Medium 199 (M199) were purchased from Sigma-Aldrich (Bornem, Belgium) and foetal bovine serum (FBS) was from Invitrogen Life Technologies (Belgium). Nembutal came from Ceva Santé Animal (Brussels, Belgium) and heparin and glucagon were from Novo Nordisk, Denmark. [methyl3H]-Thymidine (25 Ci/mmol) was purchased from Amersham Pharmacia Biotech (Amersham Bucks, UK). Benzyl penicillin was from Continental Pharma (Brussels, Belgium), streptomycin sulphate, kanamycin monosulphate and sodium ampicillin came from Sigma-Aldrich. All other chemicals and reagents were commercial products of the highest grade available.
Isolation and culture of rat hepatocytes
Hepatocytes were isolated from male outbred Sprague-Dawley rats (200–300 g; Charles River Laboratories, Belgium) and the viability was tested by trypan blue dye exclusion . Experiments were performed in accordance with the regulations of the Animal Experiments Ethical Committee of the Vrije Universiteit Brussel.
Upon isolation, hepatocytes were cultured under proliferation-  or differentiation-  promoting conditions. Briefly, to promote cell proliferation, hepatocytes were plated out at a density of 0.4 × 105 cells/cm2 in MEM-M199 (3:1) medium supplemented with 10% (v/v) FBS, 7 ng/ml glucagon, 292 mg/ml L-glutamine, 5 μg/ml bovine insulin, 1 mg/ml BSA and antibiotics (7.3 U/ml benzyl penicillin, 50 μg/ml streptomycin sulphate, 50 μg/ml kanamycin monosulphate and 10 μg/ml sodium ampicillin). Cells were allowed to attach to the plastic substrate for 4 h. Then, serum-containing medium was removed and fresh, serum-free culture medium supplemented with hydrocortisone sodiumhemisuccinate (Pharmacia, Brussels, Belgium; 500 ng/ml) and, where appropriate, EGF (50 ng/ml), was added to the cultures.
To promote cell differentiation hepatocytes were plated out at a density of 0.57 × 105 cells/cm2 in Williams’ medium E supplemented as described for the proliferation condition. Cells were allowed to attach to the plastic substrate for 4 h. Then, serum-containing medium was removed and fresh, serum-free culture medium supplemented with hydrocortisone sodiumhemisuccinate (25 μg/ml) and bovine insulin (0.5 μg/ml), was added to the cultures. All cultures were placed at 37°C in an atmosphere of 5% CO2 and 95% air and 100% relative humidity and the medium was renewed daily.
25 mM and 50 mM stock solutions of the compounds in ethanol were stored at −20°C. For exposure, intermediate solutions of 10 mM in Mili Q water were added to the cell cultures. Exposure started from the time of plating onwards and continued for 72 h in the case of proliferation-promoting conditions and for 96 h or 7 days in differentiation-promoting cell cultures. The analogues were added daily, during culture medium renewal. No effect of the solvent (ethanol) was seen compared to untreated controls.
Origin and culture of the 5T33MM vt cell line
The 5T33MMvt cell line was derived from C57BL/KalwRij mice (Harlan, Horst, NL) as previously described . The cells were cultured at 37°C in a humidified incubator with 5% CO2, in RPMI 1640 medium (BioWhittaker, Verviers, Belgium) supplemented with 10% fetal calf serum (FCS) (Fetal Clone I; Hyclone, Logan, UT, USA), 1% natriumpyruvate, 100U/ml penicillin, 100μg/ml streptomycin, 2 mM l-glutamine and 1% minimum essential medium (supplements from Bio-Whittaker).
IC50 determination was performed using HDAC Fluor De Lys Fluorescent activity assay/drug discovery kits from BIOMOL (Tebu-bio, Boechout, Belgium). Briefly, cellular proteins were extracted by sonification of freshly isolated rat hepatocytes in buffer containing 50 mM TrisHCl, 0.25 mM sucrose, 25 mM KCl and 5 mM MgCl2. After centrifugation (5 min, 16000 g) the protein concentration in the supernatant was determined according to Bradford, using BSA as a standard. Stock solutions of TSA analogues in pure ethanol were serially diluted in assay buffer containing 25 mM Tris HCl, 137 mM NaCl, 1 mM MgCl2 × 6H2O. 10μl was added to 96-well microplates, followed by 15 μl of cellular protein extract and 25 μl Fluor de Lys substrate. After 15 min, the reaction was stopped by addition of 50 μl developer. After another 15 min, the results were read in a microplate reading fluorimeter (Victor 3, Perkins Elmer) at 460 nm against appropriate controls.
Assessment of DNA synthesis
in cultured rat hepatocytes; DNA synthesis in cultured hepatocytes was measured as incorporation of radioactively labelled thymidine into newly synthesized DNA. Cells were incubated with [methyl-3H]-thymidine (25 Ci/mmol (2 μCi/ml)) for 24 h and harvested in ice-cold phosphate buffered saline (PBS) at 72 h of culture. DNA was precipitated overnight at 4°C with 15% trichloroacetic acid (TCA; Sigma-Aldrich). After centrifugation, cells were subsequently washed, once with 10% TCA and twice with 5% TCA. The pellet was solubilized in formic acid and [methyl-3H]-thymidine incorporation was measured by liquid scintillation counting (Wallac 1410; counting efficiency: 49%). Cells cultured in the absence of EGF served as a negative control, while cells stimulated with EGF alone served as positive control of hepatocytes’ proliferation. The results are expressed as a % of the DNA synthesis of the positive control.
in the 5T33MM-vt cells; 5T33MM-vt cells were plated in 96-well flat-bottom microtiter plates at a cell density of 1 × 106 cells/ml. TSA analogues were added at different concentrations. After 56h incubation, cells were pulsed with 1°μCi (methyl-3H) thymidine (0.037 MBq; Amersham) for 16h. Cells were harvested by a cell harvester (Inotech, Wohlen, Switserland) on paper filters (Filtermat A; Wallac, Turku, Finland) which were further sealed in bags (Wallac) with 4 ml Optiscint Scintillation Liquid (Wallac). Radioactivity was measured using a 1450 Microbeta Liquid Scintillation Counter (Wallac). Results are expressed as a % of the DNA synthesis of the control (untreated cells).
Cell morphology and cytotoxicity
Light microscopic pictures (×100) were taken using an Optiphot Nikon Phase Contrast microscope (Cetec, Brussels, Belgium). A Merckotest (VWR International, Belgium) was used to test hepatocyte membrane integrity. Lactate dehydrogenase (LDH) index was calculated as follows: [100 × LDH activity in supernatant]/[LDH activity in (supernatant + cells].
At the indicated time points, cultured hepatocytes were harvested in ice-cold PBS. Total cellular proteins were extracted as previously described  and their concentration determined according to Bradford using BSA as a standard.
Proteins (40 μg) were resolved on an 8% SDS-PAGE, transferred onto nitrocellulose membranes and detected using the ECL detection system (Amersham Pharmacia Biotech, UK). Ponceau S staining was used to verify equal sample loading. Afterwards, non-specific binding sites were blocked with 5% skimmed milk/Tris buffered saline for 1 h at room temperature. Incubations with primary antibodies anti-CYP1A1, anti-CYP2B1 and anti-CYP3A2 (all from BD Gentest, Belgium) were performed overnight at 4°C under gentle shaking. Proteins were detected using the ECL detection system (Amersham Pharmacia Biotech, Buckinghamshire, UK) according to the manufacturer’s instructions. Heterologously expressed CYP1A1, CYP2B1 and CYP3A2 (BD Gentest, Belgium) were used as positive controls.
Caspase-3-like activity assay
At the indicated time points, cultured hepatocytes were harvested and washed twice with ice-cold PBS. Cell pellets were lysed in homogenization buffer pH 7.0 (10 mM HEPES, 2 mM EDTA, 1 mM PMSF, 10 μg/ml Pepstatin A, 10 μg/ml Aprotinin, 20 μg/ml Leupeptin, 2 mM CHAPS, 5 mM DTT) by three freeze-thaw cycles. Upon centrifugation (30 min, 16,2 × g, 4°C) the protein concentration was determined with a Bradford method, using BSA as a standard. The volume of cell lysate, containing 50 μg of total protein, was adjusted to 230 μl with pure water and 10 μl reaction buffer pH 7.4 (10 mM PIPES, 2 mM EDTA, 1.6 mM CHAPS, 5 mM DTT) was added to each sample. After 2 min pre-incubation at 37°C, 10 μl of 1,375 mM N-acetyl-Asp-Glu-Val-Asp-7-amino-4-trifluoromethylcoumarin (Ac-DEVD-AFC) substrate was added. After 1h incubation at 37°C, the fluorescence was measured in a microplate reading fluorimeter (Victor 3, Perkins Elmer) at an excitation wavelength of 400 nm and an emission wavelength of 505 nm. The background readings from the buffer were subtracted from the readings of the samples and the caspase-3-like enzymatic activity was expressed as pmol of AFC/minute × mg total protein.
Results expressed are mean ± SD of three independent experiments. Statistical analyses were performed using Student’s t-test. The significance level was set at 0.05.
Synthesis of benzamide analogues of TSA
A series of hydroxamic acid analogues of TSA was described by Jung et al., in which the 4-dimethyloaminobenzoyl group of TSA was coupled to an alkanic acid hydroxamate [18, 19]. Since we have demonstrated that one of the important metabolic pathways of TSA involves N-demethylation , we prepared several benzamide analogues with two different spacer lengths. 4-dimethyloamino substituent was replaced by the cyclic pyrolidyno- (3), piperidino- (5,9), or morfolino- (6,10) substituent, or by a 4-chloro substituent (2). The previously reported reference compounds (1) and (7) were included for comparison. Additionally, the indole amide (4), which was reported to be over tenfold more potent than SAHA  was included as well.
IC50 values of TSA-analogues
In order to select the most potent HDAC-I within the group of newly synthesised TSA-analogues we determined the IC50 of each compound. TSA, an archetypal molecule and potent HDAC-I served as a reference. The results shown in Table 1 confirm a previously made observation namely that the length of the aliphatic linker is a critical factor determining the potency of HDAC-I which could be related to the fact that compounds with a C5 spacer demonstrate more resemblance to the natural HDAC substrate, namely acetylated lysine [18, 24]. Indeed, the molecules with C5 spacer were displaying a higher inhibitory effect than those with a C4 linker, although they shared the same chemical entities in the selectivity region (Table 1). Compounds 1, 3 and in particular 4 seem the most promising molecules with respect to their HDAC inhibitory activity as their IC50 values are in the same order of magnitude as TSA. These three compounds were further evaluated for their biological properties in hepatocytes.
Antiproliferative potency of the selected TSA-analogues
Effects of the selected TSA-analogues on the expression of phase I biotransformation enzymes (CYPs)
In the next step, we examined the effect of the three compounds on the expression of CYP2B1, CYP3A2 and CYP1A1 in a differentiation-promoting culture set up. The latter group of proteins is crucial to the detoxifying capacity of rat hepatocytes, however their expression in cell cultures is irretrievably lost in the course of time .
Effects of the selected TSA-analogues on apoptosis
Effects of the selected TSA-analogues on the proliferation of 5T33MM-vt cells
Effects of selected TSA analogues on DNA synthesis in the 5T33MM-vt cell line. DNA synthesis was determined by a [methyl-H3]-thymidine incorporation assay
% DNA synthesis compared to control
101.7 ± 8.2
94.6 ± 5.8
76.8 ± 23.2
71.5 ± 14.6*
55.2 ± 12.0**
34.9 ± 20.2*
1.3 ± 1.2*
0.1 ± 0.1*
0.3 ± 0.2*
With malignant neoplasms positioned in the top ten of the leading death causes in industrialised countries, the pressure to discover effective anti-neoplastic therapeutics is high . Finding an anti-tumour drug is, however, not the most challenging issue. The real challenge lies in finding a molecule which would be equally potent and tumour-selective, and thus safe for non-target healthy cells. Among the variety of chemotherapeutics, HDAC-I are believed to fulfil the latter condition . However, in the majority of the available studies the therapeutic efficacy of these agents in cancer models is emphasized, while the safety for non-target cells is not thoroughly investigated. Among different cell types, hepatocytes constitute a valuable cell system to investigate the effects of candidate pharmaceuticals on non-target cells since the biological activity and potential toxicity of xenobiotics is largely shaped by the phase I and phase II biotransformation reactions taking place in these cells.
Based on our extensive experience on the biological activities of HDAC-I in cultures of primary rat hepatocytes, we hereby present a screen for potent HDAC-I using these cells.
Upon IC50 determination, compounds 1, 3 and 4 showed the best HDAC-inhibiting activity. The molecules with a C5 spacer had lower IC50 values than those with C4-atoms, an observation done earlier and related to the fact that compounds with a C5 spacer demonstrate more resemblance to acetylated lysine, the natural HDAC substrate [18, 26]. In addition, compounds 1, 3 and 4 actively inhibited the proliferation of mitogen-stimulated hepatocytes (decreased DNA synthesis and cell number). In the differentiation-stimulating cultures, compounds 1 and 4 effectively preserved the expression of CYP2B1 and CYP1A1 enzymes respectively, at a higher level than observed for non-treated control cultures. Additionally, caspase-3-like activity was dramatically reduced (over tenfold) upon exposure to any of the 3 selected compounds, indicating a delayed activation of the apoptotic programmes in primary hepatocytes.
These data demonstrate that healthy cells such as primary hepatocytes, may be used in an analogous manner as tumour-derived cell lines in order to discriminate between compounds with HDAC-inhibiting activity. This kind of approach is double-edged. Indeed, it not only allows a selection of the most potent compounds, but at the same time also provides early information about their safety. Accordingly, the selected TSA-analogues, exhibited a high inhibitory potency towards HDAC enzymes, and did not display direct cytotoxic effects in cultures of primary rat hepatocytes. Additionally, the same compounds proved to be effective inhibitors of cancer cell proliferation, as shown in the 5T33MM-vt cell line. The fact, that a considerably lower dose was sufficient to markedly decrease the proliferation of 5T33MM-vt cells is of utmost importance. Indeed, it proves that the HDAC-inhibiting compounds, as selected in our system have a therapeutic potential at the concentration which does not affect the homeostasis between proliferation and apoptosis of non-target cells. This is a critical aspect, particularly in view of the fact that the use of HDAC-I is considered far beyond cancer, namely also in neuroprotective, immunomodulatory and antiprotozoal therapies .
Therefore, to minimalize potential risks of HDAC-I therapy, pan-inhibitors should be replaced by isoenzyme-specific agents, which would be particularly advantageous in pathologies in which specific HDAC isoenzymes are overexpressed or aberrantly recruited. As straightforward as it seems, the latter concept is difficult to execute as our current knowledge of physiological and pathological functions of specific HDACs is limited. Additionally, the high similarity of the amino acid sequences in the catalytic domains of HDACs implies that future isoform-selective inhibitors must be based on the interactions with other regions of the proteins .
This study was financially supported by Vrije Universiteit Brussel (OZR-VUB), Belgium through the GOA project. Part of the project is also supported by the EU FP6 project, Liintop. Thanks go to Mrs. Pauwels M., Mr. Degreef B., Mr. Branson S. and Ms. Bastiaensen E. for their technical assistance.