Abstract
Background: Histone deacetylase (HDAC) inhibitors are promising antitumoral drugs. Currently there are no data regarding the comparison of different HDAC inhibitors on hepatoma cells. Materials and Methods: Hepatoma cells were incubated with the HDAC inhibitors MS-275, SAHA, FK901228 and trichostatin. Proliferation was assessed via BrdU incorporation and apoptosis rate via flow cytometry. Trichostatin, SAHA and MS-275 were applied in a rat hepatoma model. Results: The agents showed antiproliferative and pro-apoptotic effects time- and dose-dependently. SAHA and MS-275 were moderately effective at 10 μM, while trichostatin A and FK901228 showed higher potency. Caspases 3 and 8 were activated upon treatment with the drugs. The agents increased the acetylation rate. Hyperacetylation did not correlate with antitumoral efficacy. In vivo, SAHA was superior to MS-275 and trichostatin A. Conclusion: The HDAC inhibitors were effective both in vitro and in vivo. The potency of SAHA and MS-275 was similar. In spite of differing affinity to the 11 known HDACs, the agents induced comparable effects. These findings suggest that these agents have further antitumoral effects apart from HDAC inhibition.
Hepatocellular carcinoma (HCC) remains one of the most common tumor entities (1). Systemic cytostatic chemotherapy has shown no life-prolonging effects in the past (1-3). Scientific research has developed substances directed against tumor cell-specific mechanisms, such as neoangiogenesis. Among these substances is sorafenib, a molecular inhibitor, which reduces neoangiogenesis and hyperproliferation (3). Research has, thus, taken a new direction: Abnormal activity of histone-deacetylases (HDAC) has been shown to interact with the regulatory processes of cancer such as tumor suppressor silencing, cell-cycle control and growth signaling, differentiation, angiogenesis, cell adhesion and tissue invasion (2, 3). HDAC inhibitors have been shown to reduce tumor growth and to enhance the effects of other antitumoral agents (4, 5). They influence the post-translational modification of DNA structure, which can be much more pronounced in malignant than in non-malignant cells (5).
DNA is mostly found in a highly condensed form, the chromatin structure. The DNA forms nucleosomes, which consist of a histone-octamer surrounded by the DNA. This structure is also part of the regulatory apparatus of DNA transcription. The N-terminal regions of the histone proteins can be post-translationally modified and lose their positive charges, which is essential for the link with the DNA structure (4-6). This may lead to a less condensed chromatin structure and therefore allow a higher grade of transcription. The mediating enzymes of these epigenetic mechanisms are histone-acetyltransferases (HAT), HDAC and methyltransferases. These enzymes interact with transcription factors and co-factors binding to their specific DNA region. In tumor cells with a low grade of acetylation pro-proliferating and anti-apoptotic factors are strongly expressed (5, 7-9). Therefore, an activation of HAT or an inhibition of HDACs might correct the changed expression profile in tumor cells.
However, HDACs not only interact with histone proteins, they also affect non-histone proteins (6). Among these non-histone proteins are nuclear hormone receptors and other factors affecting some pathways of tumor development and survival (5, 6).
The family of HDACs consists of four different classes. Class I (HDAC 1, 2, 3, 8) is localized in the nucleus. Class II HDACs (6-8) act on histones and non-histone proteins and may be subclassified: class IIa HDACs shuttle from the nucleus to the cytoplasm and vice versa, whereas class IIb HDACs are located primarily in the cytoplasm (6) and interact with the cytoskeleton (3, 8). Class III shows no homology to the others and its function is not fully understood. Currently, the only member of the fourth class is HDAC 11, which shares attributes with both, class I and II HDACs. Several natural and designed HDAC inhibitors exist (4, 5). The known HDAC inhibitors are directed to class I and II HDACs. Beside their direct antitumoral activity, they exhibit chemosensitizing and radiosensitizing ability (9-11). However, antitumoral mechanism has not been fully investigated.
The effects of HDAC inhibitors are induced by low doses, which induce much less harm to non-malignant cells, such as primary hepatocytes (2). Hence they are an attractive treatment option even for chemoresistant malignancies, such as hepatocellular carcinoma (HCC). Some HDAC inhibitors are in Phase II to III trials or were recently approved by the Food and Drug Administration (FDA) (9, 10 ,12). However, the known HDAC inhibitors belong to different chemical subtypes and exhibit different affinities for HDAC 1-11. A direct comparison of several HDAC inhibitors is lacking.
Here, we compared the antitumoral activity of different types of HDAC inhibitors: suberoxylanilide hydroxamic acid (SAHA or vorinostat) inhibits the activity of all 11 known HDACs. SAHA was shown to induce growth arrest and death in a broad variety of malignant cells (12). MS-275, an aminophenyl benzamide, was developed as an orally available HDAC inhibitor. It is directed against class I and II HDACs with the strongest affinity to HDAC 1 and 3 (9, 10, 13). Trichostatin A (TSA) is the first described natural HDAC inhibitor with activity against class I and II HDACs. TSA inhibits HDAC 6 and 10 more effectively than MS-275 does, while HDAC 4 and 8 are inhibited weakly by both agents (2, 14). FK901228 (FK), a cyclic tetrapeptide, is a strong direct inhibitor of class I HDACs, while the effects on class II are indirect and of minor impact (9, 15). Here we compared the antitumoral efficacy of these related, but different agents. The target parameters were the antitumoral activity in vivo and in vitro as well as the activation of pro-apoptotic proteins/enzymes.
Materials and Methods
Cell culture and reagents. All cell lines were obtained from the German Collection of Microorganisms and Cell Cultures (DSMZ, Braunschweig, Germany). The hepatoma cell lines HepG2 and Hep3B and human foreskin fibroblasts (HF; non-malignant control cell line) were cultured in Dulbecco's modified Eagle's medium (DMEM, Biochrom, Berlin, Germany), containing 10% fetal calf serum (FCS; GibcoBRL, Karlsruhe, Germany), penicillin (107 U/l) and streptomycin (10 mg/l), at 37°C and in 5% CO2. Cells were starved for 24 h in a medium containing 0.125% FCS, seeded at a density of 0.5×106 per well (FACS analysis, protein assays) and incubated in the presence of TSA, FK, SAHA, or MS-275 at 0.1 to 100 μM. All agents were dissolved in dimethylsulfoxide (DMSO; Sigma, Deisenhofen, Germany) and for the in vitro experiments they were further diluted with culture medium. For in vivo experiments, the stock solutions were diluted 20-fold with Aqua injectable (Baxter, Deerfield, IL, USA) to a final maximal DMSO concentration of 5% before injection. SAHA was purchased from Calbiochem-Novabiochem (La Jolla, California, USA), TSA was purchased from Sigma. MS-275 was kindly provided by Schering (Berlin, Germany). FK was a gift from Fujisawa Pharmaceuticals (Japan).
Flow cytometric analysis of apoptosis. For quantification of apoptosis, culture supernatants were collected and cells were washed twice with PBS, trypsinized and lysed in a hypotonic solution containing 0.1% sodium citrate, 0.1% Triton X-100 and 50 μg/ml propidium iodide (Sigma). Analysis of labeled nuclei was performed on a FACSCalibur fluorescence-activated cell sorter (FACS) using CELLQuest software (both from Becton Dickinson, Heidelberg, Germany). The percentage of apoptotic cells was determined by measuring the fraction of nuclei with a sub-diploid DNA content. Ten thousand events were collected for each analyzed sample.
Measurement of DNA synthesis. Cells were seeded at a density of 10,000/ml into 96-well plates and incubated with medium-alone or the HDAC inhibitors in different concentrations. DNA synthesis was determined by bromodeoxyuridine (BrdU) incorporation, using the Cell Proliferation ELISA (Roche Molecular Biochemicals, Mannheim, Germany), which is based on incorporation of BrdU into newly-synthesized DNA and on antibody-mediated detection of incorporated BrdU, as recommended by the manufacturer.
Assessment of caspase-3 and -8. The Caspase Colorimetric Assay (R&D Systems, Minneapolis, MN, USA) was used to determine the enzymatic activity of caspases 3 and 8. The assay was performed according to the manufacturer's instructions, after 6 to 24 h of incubation with the agents. Caspase activation leads to the cleavage of the provided colorimetric substrates conjugated to p-nitroanaline [(DEVD)-pNA] and can be photometrically measured at 405 nm. The experiments were carried out four times. Results are given compared to the activity of untreated cells which was set to a value of 1.
Western blot analysis. After incubation, cells were lysed at a density of 5,000 cells/μl. Protein from 75,000 cells was separated by 12 or 14% SDS-PAGE (pre-cast gels; Novex, San Diego, CA, USA) and transferred to nitrocellulose (pore size 0.2 μm, 240 mA, 30 min). The blotting membrane was blocked overnight with PBS/0.1% Tween 20 containing 4% low fat milk powder. The following primary antibodies were used for a 2 h incubation: Bcl-2–associated X protein (BAX)/sc-493, 1:300, and B-cell lymphoma 2 (BCL2)/sc-783, 1:150 (both from Santa Cruz; Santa Cruz, CA, USA); cyclin-dependent kinase inhibitor 1 (p21CIP/WAF1)/C24420, 1:500 (Transduction Laboratories, Lexington, UK). Membranes were washed three times for 10 min in a buffer containing PBS, 0.1% Tween 20 and 4% low fat milk powder and incubated with an anti-rabbit IgG coupled to peroxidase (1:1000; Sigma) for 1 h at room temperature. Reactive bands were detected with the ECL chemiluminescence reagent (Amersham Pharmacia Biotech, Freiburg, Germany).
Assay for assessment of acetylation. The acetylation assay determines the grade of acetylation in a standardizised cell lysate. The assay was performed according to the manufacturer's instructions (Roche Pharmaceuticals, Mannheim, Germany). In brief, the standardized cell lysate (HeLA) was incubated with the HDAC inhibitors at different concentrations (incubation period 30 min). The acetylated sites are detected via antibodies and colorimetric development in an ELISA Reader (405nm). The experiments were repeated at least 3 times. The assay shows maximal acetylation at baseline in an untreated control. If the cell lysate remains untreated, reduced acetylation can be detected. If the cell substrate is incubated with the standardizised control HDAC inhibitor or the experimental agents, the rate of de-acetylation decreases. The results are given as the percentage of the maximal baseline acetylation before incubation with HDAC inhibitor.
In vivo experiments. Animal maintenance and experimental procedures were approved by the Government of Middle Franconia and carried out according to “The 1996 Guide for the Care and Use of Laboratory Animals” as published in ILAR (16). Buffalo rats (male, 200-250 gr) were purchased from Charles River Laboratories (Schweinfurt, Germany). For tumor induction, MH7777A cells (107 cells in 100 μl 0.9% NaCl solution) were injected into the sub-capsular space of the left liver lobe, while rats were anesthetized. On the seventh postoperative day (at an approx. tumor size of 5 mm diameter) treatment was started. The drugs were administered intraperitoneally. The doses used, were those recommended by the manufacturers (MS-275 10 mg/kg bw, SAHA 50 mg/kg bw, TSA 10 μg/kg bw). After 21 days of treatment (day 28 after tumor implantation), the rats were euthanized. The number of animals per group was six. After sacrifice, the liver was removed and the tumor volume calculated using the formula a×b2, where a and b are the perpendicular diameters.
Results
Growth inhibition and apoptosis in vitro. The HDAC inhibitors induced antiproliferative effects as shown in the BrdU assay (Table I). MS-275 had the most prominent effect at a concentration of 100 μM and reduced tumor cell growth by 54% (HepG2) and 62,3% (Hep3B), while that of HF cells was reduced by 27%. At a higher dilution, the antiproliferative effects faded. SAHA gave comparable results at 100 μM and 10 μM. FK and TSA had higher potency and reduced the growth of HepG2 and Hep3 B cells even at 1 μM (Table I). The effects depended not only on tumor cell type and agent concentration, but also on time, with greater effects after 48 h of incubation (data not shown).
FACS analysis (Figure 1A and B) showed pro-apoptotic effects at similar concentrations. The effects were time-dependent. Again, MS-275 induced moderate to high rates of sub-G events at 100 μM and 10 μM, while the lower concentrations were nearly ineffective. SAHA had some efficacy even at concentrations of 1 μM with a 19.9% apoptosis rate after 72 h for HepG2 and 38.6% for Hep3B cells. At 10 μM FK induced maximum cell death after 48 h in all of tested cell lines (including the non-malignant control cells), indicating a cytotoxic effect. At a concentration of 1 μM FK, we observed moderate apoptosis rates of 39% and 49% after 72 h of incubation (HepG2 and Hep3B cells, Figure 1 A and B). The effect of TSA was comparable to that of FK, with moderate apoptosis induction at 1 μM and no efficacy at lower concentrations. We collected the FACS and BrdU data for TSA in this setting and confirmed previously published data (2). Hep3B cells were more sensitive to treatment than HepG2 cells.
BrdU incorporation showing the effect of histone deacetylase inhibitors on cell lines. Mean results were compared to the BrdU incorporation of untreated cells. Results are given as the mean±standard deviation (n=18/experiment).
TSA, MS-275 and SAHA induced a significantly lower apoptosis rate in non-malignant control cells (HF cells). However, 10 μM of FK led to nearly complete death of HF cells after 72 h incubation. Even 1 μM of FK induced 41% of cell death after 48 h. These effects are similar to those in malignant cells, confirming the possible cytotoxic effects of FK (Figure 2).
Additional assays. These experiments were performed with the best effective concentrations, as tested in the BrdU assay and FACS analysis. As expected, the evaluated agents induced the expression of p21 in a time-dependent way in HepG2 and Hep3B cells. The marker proteins for apoptosis, caspase-3 and BAX were induced, while BCL-2 was downregulated. These experiments were carried out qualitatively. In parallel, all tested agents doubled the activation of caspases 3 and 8 after 24 h of incubation. FK even induced 3.2-fold caspase-3 activation after 48 h (Table II).
The assay for acetylation (Figure 3) showed dose-dependent hyperacetylation after incubation with the agents, when compared to the de-acetylation rate without HDAC inhibition. Despite the differing grade of antitumoral activity, the evaluated HDAC inhibitors affected the grade of acetylation in a comparable way (Figure 3). MS-275 had the lowest effects, with sufficient HDAC inhibition only at doses ≥10 μM (53.1% acetylation at 10 μM). The other agents showed a higher potency with comparable results even at doses of 1 μM and less (69% for TSA, 65% for FK and 61% for SAHA at 1 μM). TSA and FK induced moderate acetylation even at 0.01 μM (45.7% and 42.8% vs. 13% and 12% for SAHA and MS-275, respectively).
Apoptosis rate in HepG2 cells (A) and Hep3B cells (B) on treatment with MS-275, SAHA, trichostatin A (TSA) or FK901228 (FK) (FACS analysis, n=6/experiment).
Apoptosis rate in non-malignant human foreskin fibroblasts (HF-cells) on treatment with MS-275, SAHA, trichostatin A (TSA) or FK901228 (FK) (FACS analysis, n=6/experiment).
Acetylation rate after incubation of HeLa lysate with MS-275, SAHA, trichostatin I (TSA) or FK901228 (FK). Maximal acetylation= acetylation grade before incubation (set to 100%). Minimal acetylation=acetylation grade after incubation with a placebo (untreated control).
In vivo experiments. The evaluated HDAC inhibitors reduced macroscopic tumor growth in the syngeneic Morris hepatoma model compared to untreated rats. In this setting, SAHA was the most effective (Figure 4). The untreated control animals and the TSA-treated rats gained weight due to tumor burden and ascites, while the most effective compounds reduced the animals' weight. The weight changes ranged for about 10% of the starting weight and did not reduce the overall condition in any detectable way. No animal died or had to be sacrificed before the set end-point.
Tumor growth reduction in a syngenic hepatoma rat model after treatment with MS-275, SAHA or trichostatin II (TSA). The results are given as a percentage to that of the untreated animals. Results are the mean±standard deviation (n=6/experiment).
Discussion
In past decades cytostatic therapy was the only systemic treatment option for malignancies. Its efficacy depends on the grade of tumor cell resistance and the stability of the non-malignant tissue. Chemoresistant tumor types, such as hepatocellular carcinoma, remained untreatable, when the tumor burden was high or metastasis occurred (1). However, there have been new insights regarding tumorigenesis, the abnormal regulation of protein expression and cell-cycle regulation. This has opened the way to new, directly antitumoral compounds. Only recently was epigenetic protein modulation of histone proteins identified as a major regulator of gene transcription. The mediating enzymes, the histone-deacetylases, are overexpressed in a range of tumor types (3-7).
We know that HDACs not only change the histone acetylation grade, but interact with proteins in the nucleus or the cytoplasm. In particular, cell-cycle regulators and proliferation factors are among the proteins which are affected by HDACs, either directly or via modulation of gene expression (6, 7, 11). HDACs differ in their exact function and substrate, although the specific role of each HDAC has not been completely decoded. Some of them may act as proto-oncogenes, if they are overexpressed (8). HDAC1 enhances tumor cell proliferation as a cell-cycle promoter, HDAC2 has antiapoptotic functions in cancer cells (8). Hence, the inhibition of histone deacetylation may disrupt the tumor cell life cycle.
Caspase activation assay. Results are given compared to the activity in untreated cells (n=4/experiment).
A range of natural and synthetic HDAC inhibitors have been identified and evaluated. They are known to reduce tumor proliferation and to re-induce apoptosis in a wide range of tumor types, among them hepatoma, colorectal cancer and lymphoma. There are even substances approved by the FDA or in phase III trials (5, 9, 10). However, the known HDAC inhibitors are not equally effective in inhibition.
For the first time to our knowledge, we compared the effects of four different HDAC inhibitors on hepatoma cells in vitro and in a rat model of hepatoma. These HDAC inhibitors exhibited antitumoral efficacy in vitro and in vivo. We saw differences in their potency: While MS-275 and SAHA were effective only at 100 and 10 μM, TSA and FK had best efficacy at a 10-fold lower concentration. As expected, we detected high apoptosis rates. All of the tested HDAC inhibitors activated caspases 3 and 8. In western blot analysis we saw a time-dependent induction of BAX and a down-regulation of BCL-2. Thus apoptosis was not only initiated by cell membrane ligands (11), but also triggered by the mitochondrial apoptosis pathway. As described before, p21CIP/WAF, a cell cycle regulator, was induced (2). HDACI-deficient cells overexpress p21 and have a reduced proliferation rate (8). HDAC inhibitors have significant less effect on proliferation of p21-deficient HCT-116 cells, compared to wild type HCT cells (experiments by our group, unpublished data).
Surprisingly FK, the most specific HDAC inhibitor, induced complete cell death not only in tumor cells, but also in non-malignant control cells, fact that strongly indicates a possible cytotoxic effect. The other HDAC inhibitors presented antitumoral effects even at low concentration, which did not harm HF cells. Unfortunately, FK could not be tested in vivo due to these probable toxic effects in vitro and due to the amount recommended for use in rats. However, evaluation of tumor growth inhibition in vivo and the induced side-effects would have been interesting. The remaining HDAC inhibitors were effective in the rat model, indicating that sufficient concentrations can be reached in vivo. Here, MS-275 and TSA were inferior to SAHA, which is probably related to drug metabolism in vivo. No animal showed intolerable side-effects or had to be sacrificed before the intended end-point. Therefore, an in vivo therapeutic option, even for the difficult-to-treat hepatoma, seems feasible.
For the proof of the antitumoral pathway via HDAC inhibition, we performed a commercial acetylation assay. We saw reduced de-acetylation (hyperacetylation) after incubation with the HDAC inhibitors in a dose-dependent manner as expected. However, maximum acetylation was neither maintained by the moderate effective compounds (MS-275 and SAHA) nor by the more potent ones, FK and TSA. It is remarkable that the grades of acetylation do not correlate with the antitumoral efficacy: SAHA and MS-275 showed 55-65% acetylation at 10 μM, the concentration with greatest antitumoral activity. TSA and FK maintained even higher rates of acetylation, but were mostly ineffective for tumor suppression in concentrations inducing about 60% acetylation. This may indicate that the antitumoral effect is not only induced by hyperacetylation of the histone proteins and transcription regulation, but via direct antitumoral efficacy. We know that some HDACs, such as HDAC1 and 6 also interact with extranuclear proteins (6, 17, 18). HDAC6 interacts with the cytoskeleton, leading to improved cell mobility and probably enhancing metastasis (17, 18). Additionally, HDAC inhibition leads to significant suppressions of matrix metalloproteinases, which are mediators of metastasis (19).
In conclusion, the tested HDAC inhibitors showed antitumoral effects both in vitro and in vivo. The effects on acetylation do not go parallel with the antitumoral efficacy, indicating relevant extranuclear pathways. Any of the tested HDAC inhibitors may offer a new treatment option for HCC. ‘Pan HDAC inhibition' (TSA and SAHA) seems a reasonable approach, since the exact role of HDACs' in tumor cells is not fully elucidated. Furthermore, pan-HDAC inhibition might overcome tumor cell resistance (8). However, specific HDAC inhibition could reduce the amount of toxicity (5, 9).
The data confirm the antitumoral potential of some HDAC inhibitors and justify further investigation: The effects of specific vs. non-specific HDAC inhibitors in a model for metastasis should be evaluated. Furthermore, the chemosensitizing effects of different HDAC inhibitors should be compared and standardized for identification of the best combination partners in clinical trials.
Acknowledgements
We thank Schering, Berlin (Germany) and Fujisawa Pharmaceuticals (Japan) for MS-275 and FK901228.
- Received September 13, 2012.
- Revision received October 22, 2012.
- Accepted October 23, 2012.
- Copyright© 2012 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
