Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

In 1976, the hydroxamic acid Trichostatin A (TSA³) (Fig. 2) was isolated from Streptomyces hygroscopicus and identified as an antifungal antibiotic, exhibiting anti-Trichophyton activity (Tsuji et al., 1976). About 10 years later, TSA was rediscovered as a potent, specific, and reversible inhibitor of histone deacetylase (HDAC) showing activity both in vivo and in vitro (Yoshida et al., 1990b). The equilibrium between HDAC and histone acetyltransferase activities determines the acetylation level of the N-terminal tails of core histones. Histone acetylation is an important mechanism for the regulation of eukaryotic gene expression. In general, transcriptionally active genes are associated with hyperacetylated histones, whereas hypoacetylation can result in transcriptional repression and gene silencing (Krajewski, 1999; Magnaghi-Jaulin et al., 1999, 2000; Cress and Seto, 2000; Mahlknecht et al., 2000). In a number of mammalian cell lines, nanomolar concentrations of TSA induce histone hyperacetylation, accompanied by various cellular phenotypic changes and a characteristic blockage of the cell cycle at G₁ and G₂ phases (Yoshida and Beppu, 1988; Yoshida et al., 1990b; Hoshikawa et al., 1994). The hydroxamic acid function and the natural R-(+)-configuration of the chiral center at the 6 position of TSA seem to be crucial in this respect, since Trichostatic acid and S-()-TSA are shown to be inactive (Yoshida and Beppu, 1988; Yoshida et al., 1990a).

The promotion of the transition of dedifferentiated, proliferating cells into more differentiated and less proliferative ones by TSA could potentially be employed to treat a variety of tumorous and fibroproliferative diseases. In vitro antitumoral effects of TSA have been shown in a wide variety of human transformed cell lines including neuroblastoma, leukemia, melanoma, and colon carcinoma cells (Li et al., 1996; Lin et al., 1998; Kosugi et al., 1999; Saunders et al., 1999; Marks et al., 2000). In vivo, TSA shows potent antitumor activity against breast cancer without apparent toxicity, as evaluated in a N-methyl-N-nitrosourea carcinogen-induced rat mammary carcinoma model (Vigushin et al., 2001). Promising pharmacological results were also obtained in a Lewis lung carcinoma model (Kim et al., 2001). Niki et al. (1999) found that TSA suppresses myofibroblastic differentiation of rat hepatic stellate cells in primary culture by hyperacetylation of histones (H4). These in vitro antifibrogenic properties have also been seen in vivo, in a CCl₄-induced Balb/C mouse model of hepatic fibrosis with no apparent toxic effects (A. Geerts, personal communication).

Up to now and to the best of our knowledge, no data on the metabolic fate of TSA in humans or animals have been published in the current literature. However, biotransformation is of key importance in the understanding of the pharmacological value of a candidate drug such as TSA. In the early phase of drug development, biotransformation studies are often performed through the use of in vitro technology based on human and animal cells and tissues. Such in vitro studies provide information on the rate of metabolism, the kind of metabolites formed, and about the biotransformation (iso)enzymes involved (Tarbit et al., 1993; Wrighton et al., 1993; Ball et al., 1995; Maurel, 1996). In the present paper, the major in vitro phase I biotransformation pathways of TSA in rat and human liver microsomal suspensions and in freshly isolated rat hepatocytes are described. Separation and detection of the phase I metabolites were carried out using high-performance liquid chromatography (HPLC) and simultaneous electrospray ionization mass spectrometric (ESI-MS)/UV detection, respectively. Structural assignments of the metabolites for which no reference standards were available were based on the interpretation of the spectrum obtained from HPLC/MS and from collision-induced dissociation (CID) MS/MS experiments, both with ESI. A reference standard was available for Trichostatic acid.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals and Reagents. TSA or 7-[4-(dimethylamino)phenyl]-4,6-dimethyl-7-oxo-hepta-2,4-dienoic acid hydroxamide (purity 98%), crude collagenase type I, bovine serum albumin fraction, 7-ethoxyresorufin, resorufin, HEPES, NADP⁺, glucose 6-phosphate and acetonitrile were purchased from Sigma-Aldrich NV/SA (Bornem, Belgium). Glucose 6-phosphate dehydrogenase (Grade I) was obtained from Roche Diagnostics (Mannheim, Germany). Trifluoroacetic acid (TFA) of analytical reagent grade was obtained from Merck Eurolab (Leuven, Belgium). Trichostatic acid was synthesized as described by Mori and Koseki (1988). All other solvents and chemicals were obtained from various commercial sources and were HPLC or analytical grade. Stock solutions of TSA (60 mM) were prepared in methanol, stored at 20°C, and diluted as required for each experiment.

Animals. Male outbred Sprague-Dawley rats (200-300 g) were obtained from Iffa Credo (Brussels, Belgium). They were kept under controlled environmental conditions (12 h light/dark cycle) and fed a standard diet (Animalabo A 04, water ad libitum). Procedures for the housing of rats and isolation and culture of rat hepatocytes were approved by the local ethical committee of the Vrije Universiteit Brussel (Brussels, Belgium).

Liver Microsomal Fractions (Rat, Human). Human liver specimens were acquired from organ donors at the Academic Hospital of the Vrije Universiteit Brussel in accordance with the regulations of the local Medical Ethical Committee for experiments on humans. They were immediately frozen in liquid nitrogen and preserved for subsequent microsome preparation (maximal 5 years storage). To detect possible interindividual differences in the biotransformation of TSA, liver tissues of three patients of variable age and different sexes were used for the preparation of microsomes. Patient characteristics are reported in table 1. Rat livers (n = 3), however, were freshly isolated for immediate microsome preparation. In both cases, liver microsomes were prepared as described (Hales and Neims, 1977), suspended in 0.1 M sodium potassium phosphate buffer either at pH 7.8 (for 7-ethoxyresorufin-O-deethylase (EROD)/7-pentoxyresorufin O-dealkylase activity determinations) or at pH 7.4 (for incubations with TSA), and stored in liquid nitrogen.

Hepatocytes (Rat). The isolation of rat hepatocytes was performed as described by De Smet et al. (1998). Cell integrity was assessed by trypan blue exclusion and only suspensions with a minimum of 84% observed viability were used. Incubations were carried out in HEPES buffer (pH 7.65, 37°C) at a final cytocrit of 2 × 10⁶ cells/ml (3 ml). Concentrations of TSA were 5, 25, and 50 µM, and incubations were conducted up to 6 h. Control samples were obtained by boiling the hepatocyte suspension for 2 min prior to incubation with TSA. Cell-free control incubations contained the substrate in the incubation medium without hepatocytes. Incubations of rat hepatocytes with 50 µM Trichostatic acid were performed under identical conditions, and samples were taken every 1 h up to 3 h.

Microsomal Proteins. Microsomal protein concentration was determined according to the Bradford procedure (Bradford, 1976) using a Bio-Rad protein assay kit (Bio-Rad, Brussels, Belgium) with bovine serum albumin as a standard.

Cytochrome P450-dependent Activities. Microsomes were incubated with 5 µM 7-ethoxyresorufin. The formation of resorufin was measured fluorimetrically according to a modified procedure of Burke and Mayer (1974). EROD activity was expressed versus microsomal protein content.

Sample Preparation. Samples from incubations with hepatocytes were thawed on ice and centrifuged at 120g for 2 min. With the aim of sample clean-up, the supernatants (extracellular medium) were subjected to solid phase extraction (Waters Oasis HLB cartridges; Waters Corporation, Milford, MA). After prewashing with 1 ml of methanol followed by 1 ml of milliQ-water, the cartridges were loaded with 0.8 ml of sample and washed [5% (v/v) methanol, 1 ml]. TSA and its metabolites were eluted with 1.2 ml of methanol. HPLC analysis after centrifugation (2000g, 30 min, 4°C) of the ultrasonicated pellet in 1 ml of methanol revealed the intracellular biotransformation profile.

Metabolite Separation and Identification. Separations were performed by reversed-phase HPLC using a Kontron chromatographic system, which consisted of a low-pressure gradient pump type 325, a UV detector type 332, and an autosampler type 465 (Kontron Interments, Milan, Italy). The HPLC system was coupled to a VG Quattro II triple quadrupole mass spectrometer with an ESI interface (Micromass, Manchester, UK). Data collection and processing was done by Masslynx software version 2.22 (Micromass). Separations (20 µl of samples) were achieved at room temperature on a Discovery C₁₈, 5 µm, 250 × 4.6 mm column (Supelco; Sigma-Aldrich NV/SA). The mobile phase consisted of component A [0.1% (v/v) aqueous TFA] and component B [0.1% (v/v) TFA in acetonitrile]. A linear gradient from 3 to 80% (v/v) B was used over 30 min at a flow rate of 1.0 ml/min. The HPLC effluent was split (ACURATE by LC Packings, Amsterdam, The Netherlands) 9:1 to direct 100 µl/min into the mass spectrometer. UV detection was at 266 nm, whereas ESI was performed in the positive mode. The operating parameters of the mass spectrometer were set as follows: capillary voltage 3kV, cone potential 50V, source temperature 70°C. Nitrogen was used as nebulizer and drying gas at a flow rate of 20 and 250 l/h, respectively. The data from the HPLC/MS analysis were collected over a mass range from m/z 110 to 850 at a scan rate of 1 s/scan. The full scan ESI-MS spectrum from the HPLC/MS analysis allowed us to determine the protonated molecular ions [M + H]⁺ of the metabolites. However, for structural identification of the metabolites, the mass spectrometer was operated off-line in the MS/MS mode (product-ion scanning mode). Samples were directly introduced to the mass spectrometer by continuous-flow injection using a syringe pump (Harvard Apparatus model 55-2222; Harvard Apparatus, Holliston, MA). Water/acetonitrile (50:50 v/v) was used as a carrying solvent, and the sample flow rate was set at 10 µl/min. Argon was used as collision gas at a pressure of 2.5 × 10³ mbar and collision energy was set to 35eV. The first quadrupole of the mass spectrometer, operated in the static mode, was set to monitor the selected molecular ions of the metabolites, and the third quadrupole was used in the scan mode, over a mass range of m/z 75 to 350 at a scan rate of 1 s/scan, to detect all the fragments obtained from the CID of the selected molecular ions. Other parameters of the mass spectrometer were set as described above. Structural assignments of the metabolites were based on the interpretation of the spectra obtained from HPLC/MS and from the CID MS/MS experiments, both with ESI.

Statistical Analysis. Results of LDH leakage in suspensions of freshly isolated rat hepatocytes and microsomal EROD activity measurements were subjected to a paired Student's t test.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Phase I Biotransformation of TSA in Rat and Human Liver Microsomes.

Enzymatic activities in rat and human microsomes EROD (rat and human CYP 1A1/2-dependent) activity was measured as a marker enzyme for quality control of the microsomes used. Based on values before and after storage (results not shown), we assumed that the microsomal phase I biotransformation activity had remained intact.

TSA metabolism in rat liver microsomes. Microsomal samples (n = 3) containing 0.0, 0.5, 1.0, 1.5, and 2.0 mg of microsomal protein/ml were exposed to 140 µM TSA during 150 min. In all cases, the breakdown of TSA was incomplete. No degradation of TSA occurred when incubated without microsomes or with boiled microsomes. An increase in protein concentration resulted in an increased breakdown of TSA during the first 30 min, an earlier "reaction stop", and a reduced percentage of unmetabolized TSA (Fig. 1). Addition of higher concentrations of NADPH-generating system or MgCl₂ to the incubation mixtures had no effect on the metabolic rate (results not shown).

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Kinetics of TSA metabolism by rat liver microsomes. Each point represents the mean (± S.D.) of three experiments. The initial concentration of TSA was 140 µM. At the indicated time points, samples were removed from the incubation mixture, diluted with acetonitrile (1:2.5, v/v), and centrifuged prior to HPLC analysis.

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Proposal for phase I biotransformation pathways of TSA and structures of the identified phase I metabolites formed in suspensions of rat and human liver microsomes and in freshly isolated rat hepatocytes. Phase I metabolites of R-(+)-TSA or 7-[4-(dimethylamino)phenyl]-4,6-dimethyl-7-oxo-hepta-2,4-dienoic acid hydroxamide were identified by ESI CID MS/MS as N-monodemethylated TSA (1), N-didemethylated TSA (2), Trichostatic acid (3), N-monodemethylated Trichostatic acid (4), TSA amide (5), N-monodemethylated TSA amide (6), and N-didemethylated TSA amide (7).

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Representative CID MS/MS-product ion spectra for the structural elucidation of TSA metabolites. Spectra of TSA (A), N-didemethylated TSA (2, B), Trichostatic acid (3, C), and N-monodemethylated TSA amide (6, D) are shown.

TSA metabolism in human liver microsomes. The biotransformation patterns of TSA in rat and human microsomes were similar and the breakdown of TSA was incomplete in both species. Interindividually, the three human samples showed no qualitative differences.

Biotransformation of TSA in Suspensions of Freshly Isolated Rat Hepatocytes.

Cell viability No significant toxic effects (LDH leakage) were seen in cell suspensions incubated with 5, 25, and 50 µM TSA for 6 h and with 50 µM Trichostatic acid for 3 h as compared with control suspensions without TSA/Trichostatic acid and suspensions containing only the solvent (results not shown).

Biotransformation of TSA. Preliminary experiments showed that incubation of 5, 25, and 50 µM TSA with 2 × 10⁶ rat hepatocytes/ml led to a complete degradation within the first hour. Since control suspensions showed no changes in the extracellular level of TSA, the degradation of TSA was completely catalyzed by liver biotransformation enzymes. In Fig. 4, an HPLC chromatogram of the extracellular medium of rat hepatocytes incubated with 50 µM TSA for 1 h is shown. Seven metabolites (1-7) could be chromatographically resolved and structurally identified.

View larger version (11K): [in this window] [in a new window]   Fig. 4.   High-performance liquid chromatogram of the extracellular medium of freshly isolated rat hepatocytes after 1 h incubation with 50 µM TSA. 2 × 106 freshly isolated rat hepatocytes/ml were incubated with 50 µM TSA as described under Materials and Methods during 1 h. 2, 7, 1, 6, 5, 4, and 3 represent the metabolites of TSA present in the extracellular medium as indicated in Fig. 2, eluting at 17.42, 18.20, 20.23, 21.09, 23.46, 24.03, and 26.53 min, respectively.

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Effect of the initial TSA concentration on metabolite formation in suspensions of freshly isolated rat hepatocytes. 2 × 106 rat hepatocytes/ml were exposed to 5 (A), 25 (B), and 50 µM (C) TSA during 1 h. The degradation of TSA and the formation of its identified metabolites were followed by comparing peak areas detected at 266 nm. Samples were taken after 0, 10, 20, 30, and 60 min of incubation. 2, 7, 1, 6, 5, 4, and 3 represent the metabolites eluting at 17.42, 18.20, 20.23, 21.09, 23.46, 24.03, and 26.53 min, respectively, as indicated in Fig. 2. Since the extinction coefficients of the metabolites at the detection wavelength are not known, the results are semiquantitative.

View larger version (19K): [in this window] [in a new window]   Fig. 6.   Comparison of the extracellular (A) and intracellular (B) formation of the identified phase I metabolites of TSA in rat hepatocytes as a function of time. The initial concentration of TSA was 50 µM. 2, 7, 1, 6, 5, 4, and 3 represent the metabolites eluting at 17.42, 18.20, 20.23, 21.09, 23.46, 24.03, and 26.53 min, respectively. Their structures are shown in Fig. 2. Their formation was evaluated semiquantitatively by comparing the peak areas detected at 266 nm in samples taken at the indicated time points.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Based on its chemical structure and knowledge of the biotransformation pathways of molecules with comparable functional groups, one could assume that the most likely phase I biotransformation pathways of TSA (Fig. 2) are N-demethylation of the dimethylaminogroup, reduction, hydrolysis, and carbonyl reduction of the hydroxamic acid function, together with oxidation of the conjugated double bonds (Gibson and Skett, 1994; Meyer, 1996; Parkinson, 1996). Thus, TSA can be expected to undergo extensive and rapid biotransformation. To the best of our knowledge, no specific data are available on the biotransformation of TSA.

Incubation of TSA with both human and rat liver microsomes led to the formation of two major metabolites, N-monodemethylated TSA (1) and N-didemethylated TSA (2) and several minor metabolites, as identified by CID MS/MS. Although relatively high concentrations of microsomal protein were used (up to 2.0 mg/ml), the biotransformation of TSA remained slow and incomplete. This cannot be explained by a decrease in biotransformation activity during storage (up to 5 years for human liver samples and 6 months for rat liver microsomes), as the microsomal phase I marker enzyme activity EROD was intact at the beginning of each experiment. Minor changes in the cytochrome P450 enzyme pattern, however, cannot be excluded.

In contrast to microsomes, freshly isolated rat hepatocytes in suspension exhibited a rapid (complete in 40 min) and extensive phase I metabolism of TSA. Seven metabolites could be identified. N-demethylation and reduction of the hydroxamic acid function were shown to be the major phase I biotransformation pathways, whereas TSA hydrolysis to its corresponding acid was a minor pathway. Although N-mono- (6) and N-didemethylated TSA amide (7) were found to be the major metabolites formed in suspensions incubated with 25 and 50 µM TSA, 6 was present in relatively small amounts in experiments with 5 µM TSA, whereas N-didemethylated TSA (2) became one of the main metabolites together with 7. This leads us to conclude that, in the presence of 25 and 50 µM TSA, the cytochrome P450 enzymes responsible for the N-demethylation reaction are saturated, which has to be taken into consideration when in vitro results are extrapolated to the in vivo situation.

Since TSA reduction is a major phase I biotransformation pathway in hepatocytes, but not in microsomes, the reaction is catalyzed by nonmicrosomal enzymes. This is in agreement with the findings by other authors that hydroxamic acids undergo enzymatic reduction to their corresponding amides by cytosolic or mitochondrial liver enzymes such as aldehyde oxidase (Sugihara and Kiyoshi, 1986; Katsura et al., 1993; Kitamura et al., 1994; Beedham et al., 1995). The possibility of interspecies differences between rat and human in phase I biotransformation of TSA cannot be excluded based on our results with microsomal suspensions. Short-term cultures of hepatocytes or liver slices are the preferable model to study human biotransformation of TSA. Human in vivo experiments to further elucidate TSA biotransformation are quite difficult to perform because of the low pharmacologically active dosage. Analyses of blood, urine, and liver samples of treated patients for traces of TSA metabolites are extremely difficult, and administration of higher dosages of TSA is unethical at this stage of the development of TSA as a drug candidate because limited data with respect to its general toxicological profile are available.

In this manuscript, we have shown that TSA undergoes rapid biotransformation in suspensions of freshly isolated rat hepatocytes. In vivo, this results in low plasma levels of the active mother compound for a limited time span and in a mixture of metabolites that can be either pharmacologically active or inactive. Furthermore, the existence of polymorphic phase I pathways can lay at the basis of inter- and intraspecies differences and unwanted drug-drug interactions and should therefore be further investigated (Tarbit et al., 1993; Wrighton et al., 1993; Ball et al., 1995; Maurel, 1996). Whether TSA biotransformation results in pharmacological and/or toxicological inactivation is not yet known. We do know that TSA acid is not biologically active (Yoshida et al., 1990a). The activities of the other metabolites of TSA identified here remain to be further investigated. X-ray crystallographic studies of an archaebacterial HDAC homolog have revealed the structure of the catalytic core of HDACs and the mode by which hydroxamic-acid based HDAC inhibitors like TSA bind the enzyme. Through its hydroxamic acid group, TSA chelates zinc at the bottom of the tube-like pocket and simultaneously forms multiple hydrophobic interactions with the walls of the pocket. The hydrophobic group on the other end of the TSA molecule (the dimethylaminobenzoyl fragment) is present in an open hydrophobic area where the histone core normally sits (Finnin et al., 1999). Jung et al. (1999) synthesized different structural analogs of TSA and compared their HDAC-inhibiting activity as well as their ability to induce terminal cell differentiation in Friend leukemic cells. The pharmacological activity was clearly dependent on aliphatic chain length, whereas modification of the para-substituent in the benzoyl moiety had little influence. Based on these findings, we can predict that the detected N-demethylated TSA metabolites, 1 and 2, are probably capable of inhibiting HDAC at potencies comparable with TSA. On the contrary, reduced metabolites (5, 6, 7) are probably not active because they lack the hydroxamic acid group capable of chelating zinc. Since 2 is a major metabolite in rat hepatocyte suspensions incubated with 5 µM TSA, in contrast to results at higher concentrations of TSA, it is quite possible that the in vivo effects of TSA are at least partially due to the action of 2.

Further experiments will focus on the design of structurally related analogs of TSA with comparable pharmacological activities but which are more metabolically stable, and the use of human hepatocytes to obtain more information on the biotransformation pathways of TSA and these analogs.