Effect of the Histone Deacetylase Inhibitor Trichostatin A on the Metabolome of Cultured Primary Hepatocytes James K. Ellis,† Pui Hei Chan,† Tatyana Doktorova,‡ Toby J. Athersuch,† Rachel Cavill,† Tamara Vanhaecke, Vera Rogiers,‡ Mathieu Vinken,‡ Jeremy K. Nicholson,† Timothy M. D. Ebbels,*,† and Hector C. Keun*,† Biomolecular Medicine, Department of Surgery and Cancer, Faculty of Medicine, Imperial College London, Sir Alexander Fleming Building, South Kensington, London, United Kingdom, SW7 2AZ, and Department of Toxicology, Vrije Universiteit Brussel, Laarbeeklaan 103, 1090 Brussels, Belgium Received August 27, 2009
Trichostatin A (TSA) is a histone deacetylase inhibitor that has antiproliferative and differentiationinducing effects on cancer cells, and in cultures of primary hepatocytes has been shown to maintain xenobiotic metabolic capacity. Using an NMR-based metabolic profiling approach, we evaluated if the endogenous metabolome was stabilized and the normal metabolic phenotype retained in this model. Aqueous soluble metabolites were extracted from isolated rat hepatocytes after 44 and 92 h exposure to TSA (25 µM) together with time-matched controls and measured by 1H NMR spectroscopy. Multivariate analysis showed a clear difference in the global metabolic profile over time in control samples, while the TSA treated group was more closely clustered at both time points, suggesting that treatment reduced the time related effect on metabolism that was observed in the control. TSA treatment was associated with decreases in glycerophosphocholine, 3-hydroxybutyric acid, glycine and adenosine, an increase in glycogen, and a reduction in the decrease of inosine, hypoxanthine, and glutathione over time. Collectively, our data suggest that TSA treatment reduces the loss of a normal metabolic phenotype in cultured primary hepatocytes, improving the model as a tool to study endogenous liver metabolism, xenobiotic metabolism, and potentially affecting the accuracy of all biological assays in this system. Keywords: metabolic profiling • trichostatin A • primary hepatocytes • xenobiotic metabolism • HDACi
Introduction The use of primary hepatocytes is useful in pharmacotoxicological testing of xenobiotics and is preferential to dosing regimes conducted in cell lines with the potential to replace at least some current in vivo testing.1 Due to modification with recombinant constructs, immortalized cell lines are metabolically distinct from the primary cells from which they are derived, and as such, the use of primary cells is highly preferable as it represents the in vivo situation far more closely.2 Better characterization of in vitro assays capable of accurately modeling in vivo conditions, specifically hepatotoxicity, will contribute to a reduction of expensive and time-consuming bioassays and ultimately reduce the number of animals needed for such assays. Culture of primary hepatocytes (both human and rodent) has traditionally been hindered by dedifferentiation of the cells once harvesting has occurred, resulting in the loss of liver-specific gene expression and associated metabolic competence (see ref 3 for a review of the molecular mechanisms of hepatocyte dedifferentiation). Many differentiated * To whom correspondence should be addressed. Dr. Hector C. Keun, e-mail:
[email protected]. Dr. Timothy M. D. Ebbels, e-mail: t.ebbels@ imperial.ac.uk. † Imperial College London. ‡ Vrije Universiteit Brussel. 10.1021/pr9007656
2010 American Chemical Society
functions are lost regardless of the culture conditions,4 and dedifferentiation may begin during isolation when the liver architecture is disrupted.3 If this dedifferentiation could be prevented and the metabolic competence of the isolated primary cell could be preserved, then the use of hepatocyte based in vitro assays could be more widely utilized as a real alternative to in vivo assays. Trichostatin A (TSA) is a potent, reversible histone deacetylase inhibitor5 (HDACi) that acts through hyperacetylation of core histones.5 TSA, and the HDACi’s, have been extensively studied due to their chemotherapeutic applications (reviewed in6). Indeed, TSA has antiproliferative and differentiationinducing effects on cancer cells both in vivo7 and in vitro.8 In cultures of primary hepatocytes TSA has been shown to induce cell cycle arrest,8 reduce apoptosis9 and maintain the metabolic competence of primary cultured rat hepatocytes,10 in particular, elevated albumin secretion and phase I biotransformation by preservation of CYP450 expression.10 The current study utilizes primary hepatocyte cultures that are treated with TSA from the time of plating the cells, thus addressing the causes of dedifferentiation instead of reducing its consequences. NMR-based metabolic profiling allows the untargeted analysis of the small molecule composition of a biological sample and is a powerful and flexible tool in biomarker discovery.11 Journal of Proteome Research 2010, 9, 413–419 413 Published on Web 11/09/2009
research articles Characterization of biofluids and tissues can reflect perturbations in metabolism related to disease or toxicant exposure.12 NMR-based metabolic profiling has been previously used to explore liver cell metabolism, specifically in vivo13-16 toxicology studies, and partial hepatectomy in the rodent.17 It has also been demonstrated that metabolic profiling can play a role in validation of in vitro assays; for example Bollard et al. have shown the effect of the toxin D-galactosamine on liver spheroid cultures.18 Characterization of the primary cultured hepatocyte metabolic phenotype would be highly complementary to existing molecular data and would allow a more comprehensive comparison of the TSA treated primary hepatocyte model to the in vivo situation. While it is clear that the HDACi TSA preserves some xenobiotic metabolism, its effects on intermediary endogenous metabolism are unknown. Here we utilized 1H NMR spectroscopy to investigate the effect of TSA treatment on isolated hepatocytes (over a 92 h period) and test the hypothesis that this treatment preserves the postisolation endogenous metabolome in a state that is more like the in vivo hepatocyte than the untreated culture.
Methods Chemicals and Reagents. TSA was obtained by Errant Gene Therapeutics LLC (Chicago, IL) and was kept as 30 mM stock solutions in dimethylsulfoxide (DMSO). All other chemicals and reagents were commercially available products of analytical grade and were supplied by Sigma-Aldrich (Bornem, Belgium or Dorset, U.K.) unless specified otherwise. Hepatocyte Isolation and Culture with TSA. Procedures for the housing of rats, and isolation and cultivation of hepatocytes were approved by the local ethical committee of the Vrije Universiteit Brussel (Belgium). Male outbred Sprague-Dawley rats weighing approximately 250 g (Charles River Laboratories, Brussels, Belgium) were kept under controlled environmental conditions (12 h light/dark cycles) with free access to food and water. Hepatocytes were isolated by use of a 2-step collagenase method and cell viability was assessed by trypan blue exclusion.19 Viable (>80%) hepatocytes were plated at a density of 0.56 × 105 cells/cm2 in William’s medium E (Invitrogen, Belgium) containing 7 ng/mL glucagon, 292 µg/mL L-glutamine, antibiotics (7.33 IU/mL sodium benzyl penicillin, 50 µg/mL kanamycin monosulfate, 10 µg/mL sodium ampicillin, 50 µg/mL streptomycin sulfate), and 10% v/v fetal bovine serum. All cell cultures were kept at 37 °C in an atmosphere of 5% CO2 and 95% air. After 4, 16, 44, and 68 h, the culture medium was removed and replaced by serum-free medium supplemented with 25 µg/mL hydrocortisone sodium hemisuccinate and 0.5 µg/mL insulin. Starting at the time of cell plating, two experimental conditions were used; vehicle control (condition i; cultures treated with DMSO at either 0.0833% or 0.283% v/v)) and TSA treated (condition ii; cultures treated with final concentration of 25 mM TSA and DMSO at 0.0833 or 0.283% v/v)). Three replicates of the experiment were conducted; including isolation, culture and dosing of hepatocytes. Principal component analysis (PCA) observed no discrimination between the groups that contained 0.0833 or 0.283% v/v of DMSO (within both condition i and ii); so the two were combined to increase the sample number (total sample number in study: n ) 24) for the control (n ) 12) and TSA treated (n ) 12) groups. 414
Journal of Proteome Research • Vol. 9, No. 1, 2010
Ellis et al. For metabolite analysis, cultured hepatocytes were harvested (at 44 and 92 h) by scraping, washed twice with ice-cold PBS, snap frozen, and stored at -80 °C as a cell pellet for subsequent extraction. Sample Preparation for 1H NMR Spectroscopy. Aqueous soluble metabolites were extracted from primary cultured rat hepatocytes using a chloroform/methanol extraction method. Briefly, 300 µL of chloroform/methanol (2:1) solution was added to each Eppendorf tube containing the frozen cell pellet. The sample was then vortex mixed for 30 s, 300 µL of ultrapure water added and vortex mixed again. The sample was centrifuged (16 000g, 10 min) and the aqueous and organic layers removed to separate sample tubes (the organic phase was not analyzed in the present study). The extraction process was repeated for each sample and the aqueous sample pooled and left to evaporate at room temperature for ∼12 h to remove any organic solvent. The aqueous extract was then freeze-dried, reconstituted in 600 µL of phosphate buffer (0.2 M Na2HPO4, 0.043 M NaH2PO4, 100 µM TSP, 3 mM NaN3 in 100% D2O) and centrifuged (16 000g, 5 min); of which 550 µL was transferred to a standard 5 mm glass NMR tube for spectroscopic analysis. All reagents were checked prior to sample preparation by obtaining a 1D 1H NMR spectrum to ensure that they contain no contaminants that may interfere with the downstream spectroscopic analysis. A BCA protein assay (Thermo Fisher Scientific, Cramlington, U.K.) was used to determine the protein concentration of the hepatocyte cell pellet (sample protein) after chloroform/ methanol extraction of the aqueous and organic metabolites. Where necessary, this sample protein measurement was used to scale the NMR data to adjust for the effect of hepatocyte number on metabolite concentration. 1 H NMR Spectroscopy. All acquisitions were made using a 5 mm broadband-inverse tube probehead (Bruker Biospin, Rheinstetten, Germany) at 300 K. Carr-Purcell-Meiboom-Gill (CPMG) 1H NMR spectra were acquired for aqueous cell extract samples using a Bruker AVANCE 600 spectrometer operating at 14.1 T (600.13 MHz 1H frequency) and the pulse sequence (RD-90°-(t-180°-t)n-AQ). The fixed echo time, t, was set to 400 µs, giving a total spin-echo time of 64 ms. During the acquisition period (AQ, 2.73 s), the free induction decay (FID) was recorded into 64k data-points in the time domain, with a spectral width of 20 ppm. Typically, spectra were recorded as the sum of 128 transients following 16 dummy scans. Suppression of the water resonance was achieved by the application of a presaturation pulse during the relaxation delay (RD, 3 s). Samples were introduced to the instrument using a BACS 60 automated sample changer, and acquisitions controlled using Xwin-NMR and Icon-NMR (Bruker Biospin). Gradient shimming was used to improve the magnetic field homogeneity prior to all acquisitions. Prior to Fourier transformation, all FIDs were multiplied by an exponential function equivalent to a line broadening of 0.3 Hz. Assignment of peaks to specific metabolites was based on the addition of known standards to the biological samples, published literature,20 in-house assignment databases and statistical total correlation spectroscopy21 (STOCSY). NMR Data and Statistical Analysis. Data were imported and manipulated in Matlab (Mathworks) using in-house software written and compiled by Dr. T. M. D. Ebbels, Dr. H. C. Keun, Mr. J. T. Pearce, and Dr. R. Cavill. For all analyses, the samples
research articles
Effect of the Histone Deacetylase Inhibitor Trichostatin A were separated into two classes: vehicle control (DMSO treated) and TSA-treated groups (TSA dissolved in DMSO dosing vehicle). 1 H NMR spectra were automatically phased, baseline corrected, and referenced and normalized to the TSP resonance at δ 0. The spectra were digitized using Matlab script developed in-house. Two levels of data resolution were analyzed: full resolution spectra at 32 697 data points and reduced resolution spectra that comprised of 1100 data points of a 0.01 ppm width.
For multivariate analysis the reduced resolution data were exported to SIMCA-P (Umetrics, San Jose, CA) and the TSP (δ