Clin Exp Metastasis (2012) 29:71–82 DOI 10.1007/s10585-011-9430-8
RESEARCH PAPER
Down-regulation of matrix metalloproteinase-7 inhibits metastasis of human anaplastic thyroid cancer cell line Szu-Tah Chen • Dah-Wel Liu • Jen-Der Lin • Fang-Wu Chen • Yu-Yao Huang • Brend Ray-Sea Hsu
Received: 30 September 2010 / Accepted: 7 October 2011 / Published online: 1 November 2011 Ó Springer Science+Business Media B.V. 2011
Abstract Epigenetic drugs such as histone deacetylase inhibitors (HDACIs) possess anticancer properties due to its ability to regulate genes associated with tumor growth, differentiation, apoptosis and metastasis. In addition to its apoptotic effect, phenylbutyrate (PB), a carboxylic acid HDACI, inhibited an anaplastic (ATC) thyroid cancer cell line ARO from penetrating a matrigel coated transwell with concomitant suppression of a metastasis-associated gene, matrix metalloproteinase-7 (MMP-7) and stimulation of a transformation suppressor protein, reversion-inducing- cysteine-rich protein with Kazal motifs without affecting MMP-2 expression levels. Direct evidence suggesting MMP-7 down-regulated cancer metastasis came from the observation of a
S.-T. Chen and D.-W. Liu contribute equally to this work.
Electronic supplementary material The online version of this article (doi:10.1007/s10585-011-9430-8) contains supplementary material, which is available to authorized users. S.-T. Chen J.-D. Lin Y.-Y. Huang B. R.-S. Hsu Division of Endocrinology and Metabolism, Department of Internal Medicine; College of Medicine, Chang-Gung University, Taoyaun, Taiwan, R.O.C
decreased pulmonary metastasis in SCID mice xeno-transplanted with MMP-7-knocked-down ARO cells. In addition, H-89, a protein kinase A inhibitor, remarkably restored the down-regulaed MMP-7 level treated by PB. Thus, the suppressive effect of PB on MMP-7 was partially carried out through H3 phosphoacetylation. To conclude, our findings suggest PB inhibits MMP-7 expression epigenetically through phosphoacetylation of histone proteins, and thereby, reduced invasive ability of an ATC thyroid cancer cell line. Keywords Thyroid cancer MMP-7 Metastasis Phenylbutyrate siRNA Abbreviations HDACI Histone deacetylase inhibitors PB Phenylbutyrate MMP Matrix metalloproteinase RECK Reversion-inducing-cysteine-rich protein with Kazal motifs TIMP Tissue inhibitor of MMP siRNA Small interference RNA
Introduction D.-W. Liu Division of Thoracic and Cardiovascular Surgery, Chang-Gung Memorial Hospital; College of Medicine, Chang-Gung University, Taoyuan, Taiwan, R.O.C F.-W. Chen Division of Endocrinology, Department of Internal Medicine, National Taiwan University Hospital, Taipei, Taiwan, R.O.C B. R.-S. Hsu (&) Division of Endocrinology and Metabolism, Chang-Gung Memorial Hospital, 5 Fushin st. Kwei-san County, Taoyaun Hsien, Taiwan, R.O.C e-mail:
[email protected]
Anaplastic (ATC) cancer of the thyroid is notorious for its rapid growth, extensive local invasion and distant metastasis [1]. Regardless of the enormous efforts being addressed, little success was achieved to improve patient’s life expectancy partly because the rapid progression of the tumor, and the complexity of the metastatic process as well [2]. Even so, thanks to a rapid-growing branch of epigenetic therapy, remarkable breakthrough has been achieved with a pleiotropic anti-cancer regimen targeting on cell growth, differentiation, apoptosis, and metastasis [3, 4].
123
72
Many studies involving tumor metastasis have focused on a group of protease categorized as the matrix metalloproteinases (MMPs) due to the recognition that metastasis-prone tumors express aberrant amounts of MMPs and inappropriate proportion of their counterpart tissue inhibitor of MMPs (TIMPs) in tumor as well as neighboring stromal cells [5]. These concepts have opened new avenues for treating advanced cancers with either broad-spectrum MMP inhibitors [6] or artificially over-expressed TIMPs [7]. Nonetheless, neither of each achieved satisfactorily clinical outcomes, which could probably be attributed to the redundant effect of MMPs during diverse stages of tumorigenesis/metastasis, and/or the expression of specific, rather than broad-spectrum of MMP subtype(s) in different kind of tumors [8, 9]. For the latter case, Shiomi et al. [10] found that MMP-2 and TIMP-2 correlate well with the invasiveness of cancers such as papillary (PTC) thyroid cancer and breast cancer by applying a sandwich enzyme immunoassay; whereas, MMP-7 was advocated to play important roles in the metastasis of endometrial and gastrointestinal cancers by the others. Researchers have later focused their interest on another endogenous MMP inhibitor, the reversion-inducingcysteine-rich protein with Kazal motifs (RECK) gene, which may suppress tumor invasion and metastasis through its inhibition of MMPs-2, -7 and -9 [11–13]. Although the implication of RECK gene in association with the abundance of certain MMPs have been regarded as valuable prognostic factors in some cancers, the importance of the MMPs and their inhibitors in ATC thyroid cancers awaits clarification. To answer that question, a study analyzing the abundance and correlation of MMPs-2, -7, -9 and their endogenous inhibitors (TIMPs-1, -2 and RECK) in thyroid cancer cell lines was conducted. Through hyperacetylation of histone proteins, a group of histone acetylase inhibitors (HDACIs) regulates transcriptional activity of specific genes by allowing the access of the transcriptional machinery to the loosely histone-tagged chromatins [14–16] to regulate tumor cell growth, apoptosis and metastasis; through such mechanisms, HDACIs were currently applied as anti-cancer drugs with appreciable clinical success. Phenylbutyrate (PB), a small molecular weight carboxylate among one of the HDACIs widely used in clinical phase I/II trials [17], was applied to study its effect on tumor invasion in this study.
Clin Exp Metastasis (2012) 29:71–82
lines were surveyed in this study. Thyroid cancer cell lines WRO and ARO were gifts from Dr. GJF Juillard (UCLA, Los Angeles, CA, USA). A poorly differentiated thyroid cancer cell line SW-579 was obtained from the American Type Culture Collection (ATCC); two other ATC cancer cell lines 8305C and 8505C derived from Japanese Collection of Research Bioresources (JCRB) were kindly provided by Dr. SK Liao (Chang Gung University). Both BHP 18-21 and KAT-4C cell lines were generously offered by Dr. YY Huang (Chang-Gung Memorial Hospital). All thyroid cancer cell lines were cultured in RPMI 1640 medium (Invitrogen, Carlsbad, CA, USA) supplemented with 5% (for all the solutions used in this study is defined as v/v unless otherwise specified) fetal calf serum (FCS), 2 mM L-glutamine, 100 l/ml penicillin and 0.1 mg/ml streptomycin (Sigma, St. Louis, MO, USA) at 37°C in a CO2 incubator. An average of about 5 9 105 thyroid cancer cells were allowed to grow to *80% confluence in a 60 mm dish before treatment. For different experimental purposes, cells were treated with PB (5 mM), PB with cyclohexamide (CHX, 10 lg/ml added 1 h before PB) and PB with H-89 (N-[2-(pbromocinnamylamino)ethyl]-5-isoquinolinesulfonamide, 25 lM/ml) or vehicle (DMSO in equivalent volume); respectively, for 24 h before gene expression levels were determined. Chemicals including PB, CHX, and H-89 were purchased from Sigma Company. For comparison, thyroid tissues obtained from six benign lesions (nodular goiter, NG) and 17 cancers [14 PTC and three anapalstic (ATC)] were collected with the disease-free counterparts of glands excised served as normal control. All specimens were examined by expert pathologists and freeze immediately for RNA extraction. Slide sections of paraffin-embedded tissues characterized with ATC cancers were subjected for immunohistochemical (IHC) study. The application for the surgically excised tissues used in this study was approved by the local Institutional Review Board (registration no. 92-568). RNA preparation Total RNA was extracted with TRIZOL reagent (Invitrogen) as previously described [18]. Briefly, cells were lysed in an adequate volume of TRIZOL reagent according to manufacturer instructions. After chloroform extraction and high speed centrifugation (12,0009g, 15 min), total RNA were purified and stored at -70°C until use.
Materials and methods Thyroid cancer cell lines and human thyroid specimens
Reverse transcription/polymerase chain reaction (RT-PCR)
Seven human thyroid cancer-derived cell lines including follicular (WRO), PTC (BHP 18-21) and ATC (ARO, KAT-4C, SW-579, 8305C and 8505C) type of cancer cell
Total RNA extracted from thyroid tissues and cancer cell lines were reversely transcribed as described elsewhere [18]. Table 1 displays primer sequences corresponding to
123
Clin Exp Metastasis (2012) 29:71–82
73
Table 1 Primer sequences and thermocycling conditions of MMPs-2, -7, -9, TIMPs-1, -2 and RECK genes Gene
Sense
MMP2
50 -CAGGCTCTTCTCCTTTCACAAC-30 0
Antisense
0
50 -AAGCCACGGCTTGGTTTTCCTC-30 0
0
Product size (bp)
Rx temp (oC)
Cycles
387
60
30*
MMP7
5 -ACCAAATCAACCATAGGT-3
492
60
30*
MMP9
50 -TGGGCTACGTGACCTATGACAT-30
50 -GCCCAGCCCACCTCCACTCCTC-30
171
60
30*
TIMP1
50 -GGGGCTTCACCAAGACCTACAC-30
50 -AAGAAAGATGGGAGTGGGAACA-30
278
55
35
TIMP2
50 -CCACCCAGAAGAAGAGCCTGAG-30
50 -TGGACCCATGGGATGACTGTTT-30
369
55
35
RECK
50 -CCTCAGTGAGCACAGTTCAGA-30
50 -GCAGCACACACACTGCTGTA-30
477
60
30*
5 -TAAATTTACCATGGAATA-3
30* denotes 10 min extension of the reaction at 72°C after 30 cycles of fixed reaction time
the MMP family (MMPs-2, -7 and -9) and their tissue inhibitors (RECK, TIMPs-1 and -2). The PCR reaction was performed in a thermocycler (PTC-200 DNA Engine, San Francisco, CA, USA) with the cycle parameter as listed in Table 1. Relative quantities of all tested genes in various samples were evaluated after adjustment to a simultaneously amplified internal control of b-actin gene unless otherwise stated [18]. Quantitative RT-PCR (QRT-PCR) The ABI PRISM 7000 Sequence Detection System (Applied Biosystems, CA, USA) was used to perform RTPCR using the TaqMan One-Step RT-PCR Master Mix Reagents Kit according to manufacturer instructions. The expression levels of MMP-7 gene in ARO cells with or without treatment were determined after completion of a preset cycling condition consisting of a denaturing cycle at 95°C for 10 min followed by 40 cycles of denaturing at 95°C for 15 s and annealing and elongation at 60°C for 1 min. Relative abundance of conditioned MMP-7 transcripts was normalized against the glyceraldehyde-3phosphate dehydrogenase (GAPDH) gene under corresponding situations. Both primers and probes of MMP-7 (Assay ID: Hs00159163_m1) and GAPDH (Assay ID: Hs99999905_m1) were premade Assays-on-Demand Gene Expression probes purchased from Applied Biosystems. In vitro assay for invasion activity The invasion activity of tumor cells was assessed by the transwell method. In brief, transwell chambers (Costar, Cambridge, MA, USA) with a 6.5 mm diameter polyvinyl/ pyrrolidone (PVP)-free polycarbonate filter of 8 lm pore size were used for the invasion assays. Each filter was coated with 100 ll of a 1:20 diluted matrigel (Sigma) in cold RPMI to form a thin continuous film on the top of the filter. Cells were preconditioned with or without PB treatment for 24 h. A 100 ll aliquot of serum-free medium containing 2 9 105 of cells was then added to each of
triplicate wells after trypan blue (Sigma) selection. Cells were further incubated for 24 h to observe the number of tumor cells transversing the filter under a BX50 light microscope (Olympus Co., Shinjuku-ku, Tokyo, Japan) and a detector (SPOT RT-KE Mono, Diagnostic instrument inc., Sterling Heights, MI, USA) image system. Cell numbers migrating through the bottom of the membrane were documented under 9400 high power field. By taking means of the cell numbers counted from three randomly selected views served as an index for invasive activity. Western blotting and IHC studies To measure changing protein levels before and after PB treatment, antibody (AB8118, Chemicon, Temecula, CA, USA) that can specifically recognize both the latent (28 kD, pro-MMP-7) and the active (18 kD) forms of MMP-7 protein was used for western blot analysis as previously described [18]. A b-actin specific monoclone antibody (Sigma) was concomitantly used to assure equivalent loading of each aliquot. The expression level of each protein was then determined by measuring the densities of bands representing TM each target protein documented by using an Immobilon western chemiluminescent HRP substrate (Millipore, Billerica, MA, USA). IHC follows the following procedures: after dewaxed and fixed in two xylene baths (5 min each) and a series of alcohol diluents (100, 95, 70%), slide sections were rehydrated in distilled water and then submitted to heatinduced antigen retrieval. After cooling for 20 min, endogenous peroxidase activity were inhibited with 3% hydrogen peroxide (Fisher Scientific, ON, Canada), followed by a phosphate buffered saline (PBS) wash. Sections were incubated with primary antibody specific for MMP-7 in a humidified chamber at 4°C overnight. Following incubation with the appropriate secondary antibody (Jackson ImmunoResearch Laboratories, PA, USA) for 60 min, the brown colored cytoplasmic target protein was visualized with a standard light microscope after Vectastain ABC kit (Vector Laboratories, CA, USA) and 3,30 -diaminobenzidine (DAB, Sigma-Aldrich, ON, Canada) treatment.
123
74
Gelatin zymography Aliquots of conditioned serum free media and diluted fetal bovine serum equivalent to 20 lg of cell proteins were loaded onto 0.1% gelatin containing sodium dodecyl sulfate (SDS) polyacrylamide gels, which were prepared by 0.1% SDS, 8% polyacrylamide, 0.3% bisacrylamide, 375 mM Tris/HCl and 0.05% ammonium persulfate. After electrophoresis, gels were washed twice with 2.5% Triton X-100 followed by overnight incubation in 40 mM Tris/ HCl, 10 mM CaCl2 and 0.01% NaN3. Finally, gels were stained with 0.1% Coomassie brilliant blue, 20% methanol and 5% glacial acetic acid to detect clear zones indicating the presence of proteolytic activities of MMPs-2 and -9. All the chemicals were purchased from the Sigma Company.
Clin Exp Metastasis (2012) 29:71–82
Crutz, CA, USA) as the manufacture’s instruction or permanently by transfecting a computer-designed (sense: 50 GATCCCGCAGGTCA GGACTATCTCATTCAAGAGA TGAGATAGTCCTGAGCCTGTTTTTTGGAAA-30 ; antisense: 50 -AGCTTTTCCAAAAAACAGGCTCAGGACTA TCTCATCTCTTG AATGAGATAGTCCTGAGCCTGCG G-30 ) siRNA specifically targeting on the nucleotides 147–165 of the MMP-7 gene. After ligating into a mammalian expression vector, pSilencer 2.0-U6 (Applied Biosystems), the reconstructed plasmid was stably transfected into ARO cells by using lipofectamine (Invitrogen) and selected with 1 mg/ml geneticin (G418) as previously described [20]. A self-ligated pSilencer 2.0-U6 served as a negative control in this study. Assay of spontaneous metastasis
Histone acetylation and phosphorylation Histone proteins were extracted from cultured cells in accordance with the procedure described by Yoshida et al. [19] with some modifications. Briefly, *75% confluent cells were treated with or without PB for 24 h before harvest. Shortly after submergence in a hypotonic lysis buffer (20 mM Hepes/pH 7.4, 10 mM KCl, 1 mM MgCl2, 0.5% NP-40, and 0.5 mM DTT), cells were scratched with a rubber policeman and lyzed by repetitive freeze and thaw. The nuclei fraction was collected by centrifugation at 1,0009g for 10 min and washed with PBS. The pellet was then resuspended in 0.1 ml of ice-cold H2O and acidified to a final concentration of 0.4 N by concentrated H2SO4. After incubation for at least 1 h at 4°C followed by centrifugation for 5 min at 15,000 rpm, histone proteins were extracted from the supernatant by mixing with 1 ml of acetone. After overnight incubation at -20°C, the acid soluble histone fraction was collected by microcentrifugation. Concentration of the extract was quantitated by protein assay kit (Bio-Rad). After loading onto a 12% polyacrylamide gel, aliquots of 10 lg of the acid soluble fraction extracted from various tested conditions were subjected to electrophoresis. The phospho/acetylated core histones were detected by using polyclonal antibodies against the acetylated (H3 and H4), phosphorylateled (H3) and the phosphoacetylated (H3) histones (Upstate Biotechnology, Lake Placid, NY, USA).
Female severe combined immuno-deficient (SCID, C.B17/ Icr-scid) mice of 4–6 weeks old were purchased from the Laboratory Animal Center of the National Taiwan University (Taiwan) and housed under pathogen-free conditions. All of the animal experiments followed the local animal protection law with approval from the Institutional Animal Care and Use Committee of Chung Gung Memorial Hospital described herein. For the experimental metastatic assays, animals were subcutaneously injected with parental or transfectant ARO cells (1 9 106 cells resuspended in 100 lL of PBS) into the flank region. The mice were sacrificed six weeks after tumor cell transplant. The tumors (xenografts), livers and lungs were then removed, weighted and fixed in neutral-buffer formalin and embedded in paraffin, sections were then cut and stained with hematoxylin-eosin. The existence of metastatic foci was examined under a microscope in sections prepared from tissue blocks obtained from organs at 1-mm intervals. Statistics Raw data were accumulated in more than three consecutive measurements and presented as mean ± standard error (SE) to evaluate the reproducibility of the measurements. A P value less than 0.05 by paired Student t and Pearson’s v2 tests was considered statistically significant.
MMP-7 knocked-down
Results
To determine its role played in the theme of cancer metastasis, the endogenous MMP-7 of ARO cells was knocked-down either transiently by transfecting a commercially available MMP-7 small interference RNA (siRNA, sc-41553, a pool of three target-specific 20–25 nucleotide siRNAs, Santa Crutz Biotechnology, Santa
The expression of metastasis-associated genes in thyroid tissues and cancer cell lines
123
The expression of MMPs-2, -7, -9, TIMPs-1, -2 and RECK genes were semi-quantitatively determined in seven thyroid cancer cell lines with various pathologic
Clin Exp Metastasis (2012) 29:71–82
75
differentiation and normal thyroid tissues by RT-PCR. All of the MMPs, TIMPS and RECK, except MMP-7 were detectable in the normal thyroid tissues. While TIMPs-1 and -2 expressed abundantly in all cancer cell lines with equal abundance to the normal tissues, trends of decreased MMPs-2, -9, RECK and increased MMP-7 gene expression was observed in several lineages of the cancer cells lines. Among which, RECK was suppressed in KAT-4C, BHP 18-21, and absent in ARO cell lines; MMP-2 was downregulated in KAT-4C, 8305C, 8505C and lost in ARO cell lines; and MMP-9 was dramatically down-regulated in two ATC (ARO and SW579) cancer cell lines. To the contrary, MMP-7, was up-regulated in three (ARO, KAT-4C and SW579) out of the five ATC cancer cell lines (Fig. 1a). Enzyme activity of MMPs-2, -9 and the abundance of MMP-7 protein level of ARO, 8305C, 8505C and WRO cells were examined by gelatin zymography and western blot analysis, respectively. As shown in Fig. 1b, while none of the enzyme activity was detectable in ARO cells, clear bands stood for proteolytic activities of MMP-9 (92 kD)
and MMP-2 (66 kD) were documented in 8305C and 8505C cells; whereas, only MMP-2 activity was preserved in WRO cells. In accordance with the presence of MMP-7 transcript, the 28 kD pro-MMP-7 protein was documented in ARO cells (Fig. 1c). The expression of MMP-7 in 23 surgically excised thyroid tissues (14 PTC, 3 ATC and 6 NG) were examined by QRT-PCR and/or IHC (Supplement). When evaluated by relative quantitation, the abundance of MMP-7 was found significantly higher in the PTC group than in the NG group; while less significantly, the ATC group also expressed appreciable more abundant MMP-7 than the NG group (Supplementary material a). The expression of MMP-7 protein was then documented by IHC in two paraffin-embedded ATC, which showed either tumorwide (Supplementary material b1, 2) or local invasive front (Supplementary material b5–8) expression of MMP-7, this may signify the role of MMP-7 in direct tumor invasion in ATC thyroid cancer as it did in colorectal cancers [21].
Fig. 1 Documentation of MMPs and their tissue inhibitors (TIMPs and RECK) in normal thyroid tissues and thyroid-derived cancer cell lines. a Relative abundance of MMPs-2, -7, -9, TIMPs-1, -2 and RECK genes in seven thyroid cancer cell lines (ATC ARO, KAT-4C, SW-579, 8305C, 8505C; PTC BHP 18-21 and follicular WRO) and two normal thyroid tissues (NT1 and NT2) were semi-quantitatively documented by RT-PCR. The housekeeping gene b-actin was coamplified to assure equal loading among samples and served as an internal control. b Proteolytic activities of MMPs-2 and -9 were
documented by gelatin zymography by loading 20 lg of proteins derived from supernatants obtained from ARO, WRO, 8305C and 8505C cells. Clear bands standing for MMP-9 (92 kD) and MMP-2 (66 kD) enzyme activities were demonstrated. c After incubation with the MMP-7 specific antibody (AB8118, Chemicon), the expression of MMP-7 protein in ARO, 8305C, 8505C and WRO cells was shown; the 28 kD pro-MMP-7 could be identified from whole cell lysates of ARO cells. The abundance of b-actin protein was concomitantly determined in this study to serve as internal control
123
76
PB suppressed in vitro ARO cell migrating When evaluating the invasion ability of different thyroid cancer cell lines by counting cells migrated through a matrigel-coated membrane, the ATC cancer cells showed overwhelming invasion ability than that of the well-differentiated one by a decreasing order of 96.7 ± 16.5, 89 ± 3.0, 56 ± 6.1 and 34 ± 9.5 [(mean ± SE)/9400 HPF] for ARO, 8305C, 8505C and WRO cell lines (Fig. 2a). Treatment effect of PB on cell migration was carried out by treating ARO and WRO cells with 5 mM of PB for 24 h before seeding onto the transwell; as a consequence, a trend of reduced cell migration was observed
Clin Exp Metastasis (2012) 29:71–82
in ARO, but not in WRO cells (-34.5% vs. ?33.8%, Fig. 2b), in a reproducible manner. PB affects metastasis-associated genes by downregulating MMP-7 in ARO cells in an CHX unsuppressible manner As shown in Fig. 3a, PB suppressed MMP-7 in ARO cells, up-regulated RECK and MMP-9 without affecting MMP-2 (also TIMP-1 and TIMP-2, data not shown) transcripts in both ARO and WRO cells. When the expression levels of the metastasis-associated genes were correlated to the invasion outcome, our result strongly suggests that the migration-inhibitory effect of PB was associated with the down-regulated MMP-7. Although Liu et al. [22] demonstrated that RECK overexpression suppressed tumor invasion , the absence of negative MMPs-2 and -9 responses in this study made the influence of RECK less important. Thus, MMP-7 was considered the main factor controlling tumor cell migration in ARO cells. To realize whether PB regulates MMP-7 gene through de novo protein synthesis, cells were then co-treated with a protein biosynthesis inhibitor, CHX. As shown through Fig. 3b–d, the regulation of MMP-7 by PB was not affected by CHX neither at transcriptional, nor at translational levels. Similarly, CHX did not alter the cellular response induced by PB to enhance RECK gene expression (Fig. 3b). Establishment of MMP-7 knocked-down ARO cells Stable transfectants presenting the *500 bp PCR fragment of the pSilencer-specific sequence were selected after PCR screening with T7 (50 -TAATACGACTCACTATAGGG30 ) and 2.0rev (50 -AGGCGATTAAGTTGGGTA -30 ) primers, among which, two (A721 and A726) and one (A706) clones of ARO tranfectants expressing 40 (±14)%, 66 (±7)% and 124 (±17)% of the parental level of MMP-7 (Fig. 4a, b) were chosen for positive and negative control for in vivo metastasis assay, respectively. Meanwhile, one clone (Asc704) of the siRNA sc-41553 transfectants expressing knockdown MMP-7 abundance (34 ± 8% of the parental level of MMP-7 at 1 week after transfection) was also selected for metastasis assay.
Fig. 2 Invasion ability of thyroid cancer cells was determined by documenting the number of cells migrating though a matrigel-coated transwell membrane. a Photographs represent four different kinds of migrated thyroid cancer cells (ARO, WRO, 8305C and 8505C) viewed under high power field (HPF, 9400). Average numbers [(mean ± SE)/9400 HPF] of the migrated cells were 96.7 ± 16.5, 34 ± 9.5, 89 ± 3 and 56 ± 6.1 for ARO, WRO, 8305C and 8505C cell lines, respectively. b Numbers of migrated ARO and WRO cells pre-treated with or without 5 mM of PB were documented under high power field (HPF, 9400) and means of the migrated cell numbers were transformed into columns. The significance of difference was evaluated by a paired Student t test
123
Knock-down of MMP-7 attenuated pulmonary metastasis in SCID mice Three MMP-7 knocked-down ARO transfectants (A721, A726 and Asc704), ARO cells and an empty-vector transfectant (A706) were subcutaneously injected into the flank region of the SCID mice. The average tumor weight after 6 weeks breeding was 5.21 ± 2.96 gm for ARO cells, which was insignificantly different from A706 (5.7 ± 2.29 gm),
Clin Exp Metastasis (2012) 29:71–82
77
Fig. 3 Regulatory effect of PB on the expression of MMPs and RECK genes. a The abundance of MMPs-2, -7, -9, RECK genes and their change after 5 mM PB treatment were detected by RT-PCR. bc Relative abundance of MMP-7 with or without CHX (10 lg/ml added 1 h before PB) treatment were evaluated by RT-PCR (b) and western blot (c). Only the 28 kD pro-MMP-7 protein could be identified in this study by a polyclonal antibody (chemicon). The influence of RECK gene expression by PB with or without CHX cotreatment was also shown in b. Relative quantities of all tested samples were evaluated after adjustment to a simultaneously
amplified b-actin in both RT-PCR and western blot analysis. d After treatment with PB, CHX or both, relative quantitations (CT, n = 5) of MMP-7 (filled circle) and GAPDH (filled triangle) were compared to the levels detected in cells left untreated. By standardizing to the untreated group, CT values of MMP-7 derived from each treatment group were normalized by using the comparative CT method. The characters A and W stand for ARO and WRO cells; N, P, C and PC represent no treatment, treatment with PB, CHX, and both, respectively. *P \ 0.05, NS non-significance
A721 (4.8 ± 2.38 gm), A726 (5.57 ± 2.69 gm) and Asc704 (4.02 ± 2.41 gm) cells. Spontaneous metastasis was mostly found in the lung. Overall, there were 12 out of 30 (12/30) mice transplanted with ARO/A706 cells showed metastasis to the lung with one concomitant metastasis to the intraabdominal cavity. In the other group, significantly fewer mice (5/31, P = 0.038) transplanted with MMP-7 knocked-down transfectants (A721/A726/Asc704 cells) showed metastatic lesions to the lung (Table 2; Fig. 4c). There was no tumor cells metastasize to the liver. In conclusion, our study suggests a significantly higher incidence of pulmonary metastasis in SCID mice xeno-transplanted with MMP-7 overexpressing (ARO/A706) cells, which brings a 3.472-folds of odds ratio (95% confidence interval, 1.040–11.494) for pulmonary metastasis.
H-89 partially restore the PB-suppressed MMP-7 level by antagonizing histone phosphoacetylation On analyzing the treatment effect of PB on histone modification, we found that PB not only hyperacetylated histones H3 and H4, but also hyperphosphorylated H3 by using an anti-phosphoacetylated (Fig. 5a), but not by an anti-phosphorylated H3 antibody (data not shown). Since that histone phosphorylation synergistically modulates target gene transcription through interaction with the neighboring acetylation sites [23], we tried to verify whether histone phosphorylation interfered MMP-7 expression in ARO cells by applying a potent protein kinase A (PKA) inhibitor, H-89. As our expectation, the suppressed MMP-7 transcript and protein were partially, but significantly
123
78
Clin Exp Metastasis (2012) 29:71–82
Fig. 4 Selection of MMP-7 down-regulated transfectants. a Lanes 1, 2 ARO, lane 3 WRO and lanes 4–14 geneticin-resistant ARO transfectants were screened for MMP-7 expression by RT-PCR. Asterisks represent three ARO transfectants including two clones (lane 12 A721, lane 14 A726) of decreased, and one clone (lane 9 A706) of equivalent level of MMP-7 transcript as compared to the wild-type. b The expression of MMP-7 transcript and protein in four clones of ARO cells transfected with either empty vector (A706) or
MMP-7-specific siRNA (A721, A726 and Asc704, see text) was shown. Equivalent amount of cDNA/lysate obtained from ARO cells served as control. c Exhibition of hematoxylin-eosin stained metastatic pulmonary lesions derived from SCID mice xeno-transplanted with ARO a–c and its MMP-7-knocked-down transfectant A721 d–f, where more aggregated lesions were found in ARO than in A721 xeno-transplanted mice. Area indicated by an arrow in e was amplified and shown in f
Table 2 Effect of MMP-7 knock-down on tumor metastasis in an in vivo mouse model by xeno-transplanting ARO cells and its siRNA trasnfectants MMP-7 expression level
Parental cell equivalent
Xeno-transplanted cell line
ARO
A706
A721
A726
Asc704
Mean weight of xenograft tumor (gm)
5.21 ± 2.96
5.7 ± 2.29
4.8 ± 2.38
5.57 ± 2.69
4.02 ± 2.41
No. of mice with lung metastasis
7
5
2
3
0
No. of mice xeno-transplanted
16
14
13
9
9
Sum-up
12/30
versus
5/31
restored by H-89 (Fig. 5 b–d) when the fraction of the PBinduced histone acetylation and phosphoacetylation were remarkably suppressed (Fig. 5a) as previously reported [24]. After standardization, when taking the 2-DDCT values of the MMP-7/GAPDH derived from each treated group normalized to the untreated one, a raise of the value from 0.06 (P) to 0.26 (PH, Fig. 5c) fold (P \ 0.05) suggested that the PB-suppressed MMP-7 was partially restored by H-89.
Discussion Even as rapid accumulating lines of evidence established the crucial roles played by MMPs in cancer biology, little consents can be drawn to correlate specific MMPs or their
123
siRNA down-regulated
inhibitors to the development and/or progression of certain kind of tumors because the complex interplay between MMPs and their local milieu. Several types of MMPs and TIMPs reported in the literature were inconsistently associated with clinical disease severity in thyroid follicular lesions [25–30]. To eliminate the complexity instigated by neighboring normal follicular and stromal cells, the expression of MMPs and their inhibitors were analyzed in thyroid-derived cell lines [27, 31, 32], which demonstrated that MMP-2 was constitutively expressed in the normal thyrocytes, follicular adenoma and most of the PTC, follicular and ATC cancers, except MRO and ARO cell lines. The expression of MMP-9 was reported heterogeneously in the normal thyrocytes and was absent or barely detectable in two of the ATC (SW579 and CAL-62) cancer cell lines; whereas TIMPs-1 and -2 were well-conserved in all thyroid
Clin Exp Metastasis (2012) 29:71–82
79
Fig. 5 H-89 counteracts PB by decreasing histone phosphoacetylation (a) and partially restores MMP-7 expression in ARO cells (b–d). a Fraction of the acetylated H3 (AcH3, 17 kD), H4 (AcH4, 10 kD) and the phosophoacetylated H3 (AcPH3, 17 kD) were documented in lysates derived from cells left untreated and 24 h after PB, H-89, or both by using polyclonal antibodies (Upstate). b Semi-quantitative amplification of MMP-7 and b-actin genes by RT-PCR. c Relative quantitations (CT, n = 4) of MMP-7 (filled circle) and GAPDH (filled triangle) were documented by real time RT-PCR; while taking the pretreatment level of MMP-7 as 1.0, comparative CT value (MMP-7/
GAPDH, n = 4) of various treatment was normalized as shown. d The abundance of MMP-7 and b-actin proteins was detected. Quantities of the bands stood for the 28 kD pro-MMP-7 protein were measured by a GS-710 imaging densitometer (Bio-Rad, Hercules, CA, USA). A relative quantity of MMP-7 protein in each condition was normalized by taking the pre-treatment level as standard after adjusting to a simultaneously detected internal control of b-actin. The uppercase characters N, P, PD and PH represent no treatment; treatments with PB, PB ? DMSO and PB ? H-89, respectively. *P \ 0.05
cancer-derived cell lines. Here in this study we demonstrates the expression of MMPs-2, -7, -9; TIMPs-1, -2 and RECK genes by RT-PCR in five ATC (ARO, KAT-4C, SW-579, 8305C and 8505C) and two well differentiated (WRO, follicular and BHP 18-21, PTC) thyroid cancer cell lines. In agreement with previous reports, our results confirmed the constitutive expression of both the MMPs-2, -9 and all the MMP inhibitors (TIMPs-1, -2 and RECK) in normal thyroid tissues, absence of MMP-2 expressions in ARO cells, and absence or barely detectable MMP-9 transcript in ARO and SW579 cell lines. When the proteolytic activities of the enzymes were determined, it showed absence of MMP-9 enzyme activity in WRO cells even with its transcript documented by RT-PCR. The finding of over-expressed MMP-7, the smallest member of the MMP family [33] that may enhance matrix protein degradation and growth factor shedding [34], in three (ARO, KAT-4C and SW579) out of the five ATC cancer cell lines was quite intriguing because it was contradictory to a favorable
outcome found in well-differentiated thyroid cancers [28– 30]. Even so, our results from QRT-PCR and IHC (Supplement) suggest the pivotal roles of MMP-7 played in thyroid cancer although the clinical significance of MMP-7 in thyroid cancer awaits further clarification because the limited cases studied. Furthermore, isoform of MMP-7 expressed in the well-differentiated thyroid cancers was not specified, the prevailing abundance of the premature proMMP-7 protein (28 kD) in colorectal [35] as well as many other cancers [10, 34] suggested that the over-expressed pro-MMP-7 in ATC thyroid cancer cells may implicate its role in tumor aggressiveness. Among the three endogenous MMP inhibitors, the influence of TIMPs-1 and -2 on cancer metastasis was considered trivial because their equal abundance of expression among various thyroid cancer cell lines and normal tissues. Similarly, although RECK (a membrane-anchored protease regulator) was proved to prevent metastasis in a series of studies through its inhibition of MMP mediated breakdown of
123
80
extracellular matrix (ECM) [13, 36], the significance of RECK in thyroid cancers was undetermined because the lack of clinical correlation and only three (ARO, KAT-4C and BHP 18-21) out of the seven cancer cell lines expressed decreased levels of RECK transcript. The invasion ability of four thyroid cancer cell lines expressing different abundance of MMPs-2, -7, -9 and RECK [ARO, MMP-2-MMP-7?MMP-9-RECK-; 8305C, MMP-2?MMP-7-MMP-9?RECK?; 8505C, MMP-2?MMP7-MMP-9?RECK? and WRO, MMP-2(?)MMP-7MMP-9?RECK?, where parenthesis indicates incoherent transcript-protein levels] was then studied by applying an artificial environment imitating physiologic basement membrane. As shown in Fig. 2a, ARO cells overexpressing MMP-7 but absence of MMPs-2, -9 and RECK genes appeared to show the most invasive capability among the four (three ATC and one follicular) thyroid cancer cell lines; this observation, in agreement with Ii [34] and Shiomi et al. [10], suggests that MMP-7, rather than MMPs-2, -9 [27, 28, 37] or RECK [38], determines the aggressiveness of ATC thyroid cancers. ARO cell line that behaved the most aggressiveness was then assessed for possible response to HDACI treatment. A short chain aliphatic fatty acid of the HDACI family, PB, was selected according to its capability of inhibiting cancer cell invasion in sea urchin embryo basement membrane [14]. The in vitro matrigel invasion ability of ARO cells finally came out with a *35% reduction (P = 0.065) after PB treatment. Whether PB regulated MMPs and RECK gene expression was simultaneously studied, which showed a MMP-7 suppression, MMP-9 and RECK up-regulation, and a neutral effect on MMP-2 gene, respectively (Fig. 3a). By applying PB analogues, Joseph and Liu [15, 22] similarly reported that HDACI down-regulated MMP-7 and upregulated RECK genes. Although the invasion inhibitory effect was suggested to be a RECK-mediated consequence [22], the effect of PB in inhibiting ARO cells from invasion in this study was challenged because the absence of MMP-2, the key modulator of RECK [5, 39], in ARO cells. Since the use of PB as an anti-tumor agent involves chromatin modification and clusters of gene reprogramming, whether MMP-7 distinctively determines tumor cell invasiveness among the symphonic groups of genes regulated by PB remains speculated. To explore that, we strategically knocked-down MMP-7 expression by the siRNA technology in ARO cells to study its effect in an in vivo metastasis model (Table 2). By subcutaneously injecting cancer cells into the flank region of the SCID mice, the risk of pulmonary metastasis was remarkably reduced in the group injected with ARO cells knocked-down with MMP-7 than the group xeno-transplanted with ARO or its equivalent A706 cells. The significance of reduced pulmonary metastasis in this study may implicate the role of MMP-7
123
Clin Exp Metastasis (2012) 29:71–82
played in the course of thyroid cancer metastasis, especially to the lung (Table 2; Fig. 4). Several mechanisms may explain why MMP-7 promotes cancer invasion; first of all, MMP-7, like the other MMPs, promotes cancer invasion by proteolytic cleavage of the ECM substrate. MMP-7 also activates other MMPs, such as pro-MMP-2 and proMMP-9 to facilitate tumor invasion [40]. Secondly, MMP-7 is also involved in tumor invasion through regulation of the activities of the non-ECM proteins. For example, E-cadherin, a transmembrane protein involved in the positive regulation of cell adhesion by its interaction with the cytoplasmic tail of catenins is ectodomainly shed by MMP-7 [41]. Resulting soluble form inhibits E-cadherin function in a paracrine fashion, thus promotes the migration and invasion of tumor cells. Finally, another possible mechanism that MMP-7 promotes prostate cancer metastasis in bone tissue is that the MMP-7 produced by osteoclasts was able to process the receptor activator of nuclear factor-kappa B ligand to a soluble form that promotes osteoclast activation [42]. Since that non-specific trans-repression induced by de novo protein synthesis was excluded by CHX treatment, our interest was drawn to epigenetic mechanism possibly induced by PB. Albeit thorough understanding remains elucidative, comprehensive studies documenting direct or indirect mechanisms by trans-repressing specific response elements or by epigenetic modifications of DNA [43–46] have provided clues to realize how HDACIs regulate target gene expression. Given that the influence of transcription factors modified by PB [43–45] could not be excluded, the globally increased phosphoacetylated H3, acetylated H3 and H4 (Fig 5d) suggested the roles of chromatin remodeling. Based on these findings, H-89, a selective PKA inhibitor that counteracts HDACI-induced histone modification [24] was applied to determine its effect on MMP-7 expression. As expected, H-89 counteracted the PB effect not only by inhibiting histone phosphoacetylation but also by partially restoring the down-regulated MMP-7 transcript and protein (Fig. 5). Therefore, our study suggests that, in addition to mechanisms such as decreased binding of transcription factors to specific promoter architecture [47] and decreased RNA stability [48, 49], PB down-regulates MMP-7, at least partially, through phosphoacetylation of histones or other non-histone proteins [50, 51] that may hinder specific transcription factors/enhancers to interact with chromatin structure around the MMP-7 promoter. To summarize, interruption of the MMP-TIMP system could be documented in thyroid cancer cell lines, among which, over-expression of MMP-7 was a common finding in ATC thyroid cancers and cell lines (ARO, KAT-4C and SW579). Treatment with a member of the HDACI, PB, suppresses in vitro invasion of ARO cells with concomitant down-regulation of MMP-7 gene. Thus, MMP-7 can be
Clin Exp Metastasis (2012) 29:71–82
applied as one of a genetic panel predicting tumor invasion/ metastasis of ATC thyroid cancer. Furthermore, base on their pleiotropic effect, HDACIs deserve even better appraisal as an adjuvant treatment for ATC thyroid cancer not only for their differentiation- and apoptosis-inducing capabilities, but also because their invasion and metastasis inhibitory effects. Acknowledgments The authors wish to thank Dr. K. H. Lin and Dr. S. K. Liao (Chang-Gung University) for their kind supports and valuable suggestions. The authors also like to express their appreciation to Dr. C. Hsueh (Division of Pathology, Chang-Gung Memorial Hospital) for her expert review of the surgical samples. The authors would like to thank the Chang-Gung Memorial Hospital (CMRPG32046) for financially supporting this research.
References 1. Giuffrida D, Gharib H (2000) Anaplastic thyroid carcinoma: current diagnosis and treatment. Ann Oncol 11(9):1083–1089 2. Mareel M, Leroy A (2003) Clinical, cellular, and molecular aspects of cancer invasion. Physiol Rev 83(2):337–376 3. Sigalotti L, Fratta E, Coral S et al (2007) Epigenetic drugs as pleiotropic agents in cancer treatment: biomolecular aspects and clinical applications. J Cell Physiol 212(2):330–344 4. Peedicayil J (2006) Epigenetic therapy–a new development in pharmacology. Indian J Med Res 123(1):17–24 5. John A, Tuszynski G (2001) The role of matrix metalloproteinases in tumor angiogenesis and tumor metastasis. Pathol Oncol Res 7(1):14–23 6. Coussens LM, Fingleton B, Matrisian LM (2002) Matrix metalloproteinase inhibitors, cancer trials and tribulations. Sci (New York) NY 2002 295(5564):2387–2392 7. Ramer R, Hinz B (2008) Inhibition of cancer cell invasion by cannabinoids via increased expression of tissue inhibitor of matrix metalloproteinases-1. J Natl Cancer Inst 100(1):59–69 8. Folgueras AR, Pendas AM, Sanchez LM et al (2004) Matrix metalloproteinases in cancer: from new functions to improved inhibition strategies. Int J Dev Biol 48(5–6):411–424 9. Overall CM, Kleifeld O (2006) Tumour microenvironmentopinion: validating matrix metalloproteinases as drug targets and anti-targets for cancer therapy. Nature Rev 6(3):227–239 10. Shiomi T, Okada Y (2003) MT1-MMP and MMP-7 in invasion and metastasis of human cancers. Cancer Metastasis Rev 22(2–3):145–152 11. Oh J, Takahashi R, Kondo S et al (2001) The membraneanchored MMP inhibitor RECK is a key regulator of extracellular matrix integrity and angiogenesis. Cell 107(6):789–800 12. Rhee JS, Coussens LM (2002) RECKing MMP function: implications for cancer development. Trends Cell Biol 12(5):209–211 13. Omura A, Matsuzaki T, Mio K et al (2009) RECK forms cowbellshaped dimers and inhibits matrix metalloproteinase-catalyzed cleavage of fibronectin. J Biol Chem 284(6):3461–3469 14. Dyer ES, Paulsen MT, Markwart SM et al (2002) Phenylbutyrate inhibits the invasive properties of prostate and breast cancer cell lines in the sea urchin embryo basement membrane invasion assay. Int J Cancer 101(5):496–499 15. Joseph J, Mudduluru G, Antony S et al (2004) Expression profiling of sodium butyrate (NaB)-treated cells: identification of regulation of genes related to cytokine signaling and cancer metastasis by NaB. Oncogene 23(37):6304–6315
81 16. Drummond DC, Noble CO, Kirpotin DB et al (2005) Clinical development of histone deacetylase inhibitors as anticancer agents. Annu Rev Pharmacol Toxicol 45:495–528 17. Acharya MR, Sparreboom A, Venitz J et al (2005) Rational development of histone deacetylase inhibitors as anticancer agents: a review. Mol Pharmacol 68(4):917–932 18. Chen ST, Shieh HY, Lin JD et al (2000) Overexpression of thyroid hormone receptor beta1 is associated with thyrotropin receptor gene expression and proliferation in a human thyroid carcinoma cell line. J Endocrinol 165(2):379–389 19. Yoshida M, Kijima M, Akita M et al (1990) Potent and specific inhibition of mammalian histone deacetylase both in vivo and in vitro by trichostatin A. J Biol Chem 265(28):17174–17179 20. Chen ST, Lin JD, Lin KH (2002) Characterization of a thyroid hormone-mediated short-loop feedback control of TSH receptor gene in an anaplastic human thyroid cancer cell line. J Endocrinol 175(2):459–465 21. Ougolkov AV, Yamashita K, Mai M et al (2002) Oncogenic betacatenin and MMP-7 (matrilysin) cosegregate in late-stage clinical colon cancer. Gastroenterology 122(1):60–71 22. Liu LT, Chang HC, Chiang LC et al (2003) Histone deacetylase inhibitor up-regulates RECK to inhibit MMP-2 activation and cancer cell invasion. Cancer Res 63(12):3069–3072 23. Nowak SJ, Corces VG (2004) Phosphorylation of histone H3: a balancing act between chromosome condensation and transcriptional activation. Trends Genet 20(4):214–220 24. Hauser C, Schuettengruber B, Bartl S et al (2002) Activation of the mouse histone deacetylase 1 gene by cooperative histone phosphorylation and acetylation. Mol Cell Biol 22(22):7820–7830 25. Maeta H, Ohgi S, Terada T (2001) Protein expression of matrix metalloproteinases 2 and 9 and tissue inhibitors of metalloproteinase 1 and 2 in papillary thyroid carcinomas. Virchows Arch 438(2):121–128 26. Frigugliett IC, Mello ES, Castro IV et al (2000) Metalloproteinase-9 immunoexpression and angiogenesis in thyroid follicular neoplasms: relation to clinical and histopathologic features. Head Neck 22(4):373–379 27. Korem S, Kraiem Z, Shiloni E et al (2002) Increased expression of matrix metalloproteinase-2: a diagnostic marker but not prognostic marker of papillary thyroid carcinoma. Isr Med Assoc J 4(4):247–251 28. Cho Mar K, Eimoto T, Tateyama H et al (2006) Expression of matrix metalloproteinases in benign and malignant follicular thyroid lesions. Histopathology 48(3):286–294 29. Buergy D, Weber T, Maurer GD et al (2009) Urokinase receptor, MMP-1 and MMP-9 are markers to differentiate prognosis, adenoma and carcinoma in thyroid malignancies. Int J Cancer 125(4):894–901 30. Ito Y, Yoshida H, Kakudo K et al (2006) Inverse relationships between the expression of MMP-7 and MMP-11 and predictors of poor prognosis of papillary thyroid carcinoma. Pathology 38(5):421–425 31. Baldini E, Toller M, Graziano FM et al (2004) Expression of matrix metalloproteinases and their specific inhibitors in normal and different human thyroid tumor cell lines. Thyroid 14(11): 881–888 32. Roomi MW, Monterrey JC, Kalinovsky T et al (2009) Patterns of MMP-2 and MMP-9 expression in human cancer cell lines. Oncol Rep 21(5):1323–1333 33. Gaire M, Magbanua Z, McDonnell S et al (1994) Structure and expression of the human gene for the matrix metalloproteinase matrilysin. J Biol Chem 269(3):2032–2040 34. Ii M, Yamamoto H, Adachi Y et al (2006) Role of matrix metalloproteinase-7 (matrilysin) in human cancer invasion, apoptosis, growth, and angiogenesis. Exp Biol Med (Maywood) 231(1):20–27
123
82 35. Adachi Y, Yamamoto H, Itoh F et al (1999) Contribution of matrilysin (MMP-7) to the metastatic pathway of human colorectal cancers. Gut 45(2):252–258 36. Clark JC, Thomas DM, Choong PF et al (2007) RECK–a newly discovered inhibitor of metastasis with prognostic significance in multiple forms of cancer. Cancer Metastasis Rev 26(3–4): 675–683 37. Maruyama S, Kawata R, Shimada T et al (2000) Study of matrix metalloproteinase-2 and -9 in thyroid papillary cancer. Nippon Jibiinkoka Gakkai kaiho 103(5):499–505 38. Takagi S, Simizu S, Osada H (2009) RECK negatively regulates matrix metalloproteinase-9 transcription. Cancer Res 69(4):1502– 1508 39. Takahashi C, Sheng Z, Horan TP et al (1998) Regulation of matrix metalloproteinase-9 and inhibition of tumor invasion by the membrane-anchored glycoprotein RECK. Proc Natl Acad Sci USA 95(22):13221–13226 40. Wang F, Reierstad S, Fishman DA (2006) Matrilysin overexpression in MCF-7 cells enhances cellular invasiveness and pro-gelatinase activation. Cancer Lett 236(2):292–301 41. Noe V, Fingleton B, Jacobs K et al (2001) Release of an invasion promoter E-cadherin fragment by matrilysin and stromelysin-1. J Cell Sci 114(Pt 1):111–118 42. Lynch CC, Hikosaka A, Acuff HB et al (2005) MMP-7 promotes prostate cancer-induced osteolysis via the solubilization of RANKL. Cancer Cell 7(5):485–496 43. Rada-Iglesias A, Enroth S, Ameur A et al (2007) Butyrate mediates decrease of histone acetylation centered on transcription
123
Clin Exp Metastasis (2012) 29:71–82
44. 45.
46. 47.
48.
49.
50.
51.
start sites and down-regulation of associated genes. Genome Res 17(6):708–719 Davie JR (2003) Inhibition of histone deacetylase activity by butyrate. J Nutr 133(7 Suppl):2485S–2493S Patel P, Nankova BB, LaGamma EF (2005) Butyrate, a gutderived environmental signal, regulates tyrosine hydroxylase gene expression via a novel promoter element. Brain Res 160(1): 53–62 Struhl K (1998) Histone acetylation and transcriptional regulatory mechanisms. Genes Dev 12(5):599–606 Dehm SM, Hilton TL, Wang EH et al (2004) SRC proximal and core promoter elements dictate TAF1 dependence and transcriptional repression by histone deacetylase inhibitors. Mol Cell Biol 24(6):2296–2307 Rubenstein RC, Lyons BM (2001) Sodium 4-phenylbutyrate downregulates HSC70 expression by facilitating mRNA degradation. Am J Physiol 281(1):L43–L51 Scott GK, Marden C, Xu F et al (2002) Transcriptional repression of ErbB2 by histone deacetylase inhibitors detected by a genomically integrated ErbB2 promoter-reporting cell screen. Mol Cancer Ther 1(6):385–392 Duan H, Heckman CA, Boxer LM (2005) Histone deacetylase inhibitors down-regulate bcl-2 expression and induce apoptosis in t(14;18) lymphomas. Mol Cell Biol 25(5):1608–1619 Glaser KB, Staver MJ, Waring JF et al (2003) Gene expression profiling of multiple histone deacetylase (HDAC) inhibitors: defining a common gene set produced by HDAC inhibition in T24 and MDA carcinoma cell lines. Mol Cancer Ther 2(2):151–163