MMP28 gene expression is regulated by Sp1 ...

Biochemical Journal Immediate Publication. Published 09 Feb 2010 as manuscript BJ20091798 Authoron manuscript, published in "Biochemical Journal 427, 3 (2010) 391-400" DOI : 10.1042/BJ20091798

us cri pt

MMP28 GENE EXPRESSION IS REGULATED BY SP1 TRANSCRIPTION FACTOR ACETYLATION T. E. Swingler*, L. Kevorkian*1, K.L. Culley*, S.A. Illman†, D.A. Young‡, A. E. Parker§2, J.Lohi†, I.M. Clark* From *School of Biological Sciences, University of East Anglia, Norwich UK Dept. of Pathology, University of Helsinki, Finland ‡ Dept. of Rheumatology, University of Newcastle-upon-Tyne, UK § Respiratory & Inflammation Dept., AstraZeneca Pharmaceuticals, Cheshire, UK †

current address: UCB Celltech, Slough, Berkshire, UK current address: OPSONA Therapeutics Ltd, Dublin, Ireland.

Short title: MMP28 and acetylation

an

2

dM

Address correspondence to: Prof. Ian Clark, Biomedical Research Centre, School of Biological Sciences, University of East Anglia, Norwich, NR4 7TJ , Tel. No: 01603 592760, Fax No: 01603 592250, Email: [email protected]

ce

pte

Abbreviations: CBP, CREB-binding protein; HAT, histone acetyltransferase; HDAC, histone deacetylase; MMP, matrix metalloproteinase; P/CAF, p300/CBP-associated factor; Sp, specificity protein; STRAP, serine threonine receptor-associated protein; TSA, trichostatin A

Ac

peer-00479282, version 1 - 30 Apr 2010 THIS IS NOT THE VERSION OF RECORD - see doi:10.1042/BJ20091798

1

1

Licenced copy. Copying is not permitted, except with prior permission and as allowed by law. ' 2010 The Authors Journal compilation ' 2010 Portland Press Limited

an

us cri pt

SYNOPSIS MMP-28 (epilysin) is a recently cloned member of the MMP family. It is highly expressed in the skin by keratinocytes, the developing and regenerating nervous system and a number of other normal human tissues, as well as a number of carcinomas. The MMP28 promoter has previously been cloned and characterized identifying a conserved GT-box that binds Sp1/Sp3 proteins and is essential for the basal expression of the gene. This study demonstrates that MMP28 expression is induced by histone deacetylase (HDAC) inhibitors and that this effect is mediated through the GT-box. Transient transfection assays have shown that the induction of MMP28 expression by the HDAC inhibitior trichostatin A (TSA) is mediated via Sp1 at the GT-box. Immunoprecipitation experiments have shown that the acetylation of Sp1 and Sp3 is increased by TSA treatment, however no effect on DNA binding was observed. Histone acetyltransferases such as p300 and P/CAF increased induction of the MMP28 promoter by Sp1. Knockdown of HDAC1 using siRNA also induces the MMP28 promoter. Oligonucleotide pulldown identified serine threonine receptor-associated protein (STRAP) as a further protein recruited to the MMP28 promoter and acting functionally with Sp1. KEY WORDS MMP, protease, acetylation, HDAC, Sp1, GT-box

pte

dM

INTRODUCTION Matrix metalloproteinases (MMPs) are a family of 23 structurally related enzymes in man that are able to remodel and degrade the extracellular-matrix (ECM). MMPs are also able to act on non-ECM components to mediate release and activation of soluble factors such as growth factors and cytokines from the ECM [1]. A range of evidence strongly implicates MMPs in physiological (e.g. embryonic development, wound healing) and pathological events (e.g. tumour growth and metastasis, cartilage destruction in arthritis, cardiovascular disease).

ce

MMP-28, also known as epilysin, is a recently cloned member of the MMP family [2]. The primary sequence of MMP-28 has homology to the other members of the MMP family, containing a signal sequence, propeptide, zinc-binding catalytic domain and haemopexin-like C-terminal domain. Within the propeptide is a furin activation sequence, with both mouse [3] and human [4] MMP-28 shown to be furin activated. The exon-intron structure of the MMP28 gene is unusual amongst MMPs, and there is evidence that it undergoes alternative splicing [3]. Northern blot analysis suggests that the MMP28 gene is strongly expressed in testis, and also in lung, heart, colon, intestine and brain [3, 5]. It is also expressed in the developing and regenerating nervous system prior to myelination, as well as in several carcinomas [6, 7]. We have recently reported expression of MMP28 in human cartilage and synovium with significantly increased expression in osteoarthritis [8, 9]. The only substrate reported for MMP-28 is casein, a non-specific substrate for many proteinases [2].

Ac


Biochemical Journal Immediate Publication. Published on 09 Feb 2010 as manuscript BJ20091798

Since MMP-28 is expressed strongly in keratinocytes, its expression in wound healing has been examined. The MMP-28 protein is expressed in the basal and suprabasal

2



epidermis in intact skin, whilst in injured skin, expression is seen in basal keratinocytes both at and some distance from the wound edge; this is distinct from other MMPs [2]. Injury to the epidermis was required for MMP-28 induction [10].

us cri pt

dM

an

In primary keratinocytes, expression of MMP28 was induced by tumour necrosis factor α (TNFα) [11]. The proximal promoters of both human and mouse MMP28 genes have been isolated, and partially characterised in immortalised keratinocytes (HaCaT) [3]. The MMP28 promoter has no TATA box, nor AP-1 motif often found in the genes of inducible MMPs [12]. Conserved motifs were identified that could control the expression of MMP28 including a GT-box situated 300 bp upstream from the translation start site that binds both Sp1 and Sp3 proteins and is essential for the basal expression of MMP28 in HaCaT cells. The Sp family of transcription factors recognise and bind specifically to GC and GT DNA motifs [13]. Several isoforms of Sp1 and Sp3 have been described [14, 15]. These GC- and GT-boxes are required for the expression of numerous genes and are involved in processes e.g. regulating cell growth, cell cycle and developmental patterning [16].

ce

pte

Work in our laboratory and elsewhere has shown that histone deacetylase (HDAC) inhibitors are chondroprotective in a number of models [17-19]. The induced expression of many MMPs is decreased by HDAC inhibitors in chondrocytes, but interestingly, the basal expression of a subset of genes is increased, and these include MMP28. MMP expression is regulated primarily at the level of transcription, however little is known about the role of chromatin modification in this process [20]. Eukaryotic chromatin is formed from DNA wound round a histone octamer consisting of two molecules each of histones H2A, H2B, H3 and H4. Chromatin structure and function can be modified by post-translational modifications such as acetylation. Acetylation of positively charged lysine residues in the N-terminal tails of histones is mediated by histone actyltransferases (HATs). Simplistically, this is thought to regulate expression of a subset of genes by loosening the histone:DNA complex and allowing access of the transcriptional machinery. Conversely, histone deacetylation is catalysed by HDACs. HATs and HDACs are recruited to their target promoters through a physical interaction with sequence specific transcription factors. In addition, HATs and HDACs can modify many other signalling proteins and transcription factors including both Sp1 and Sp3 [21-23].

Ac


MMP-28 is also expressed in a variety of carcinomas [6]. In oral squamous cell carcinoma and oesophageal carcinoma cells, antisense knockdown of MMP-28 causes a reduction in anchorage-independent growth. In epithelial cells, recent evidence also suggests that MMP-28 can induce an epithelial to mesenchymal transition (EMT) via activation of latent transforming growth factor β (TGFβ) [11].

Here, we show that HDAC inhibitors induce the MMP28 promoter and that this effect is mediated through Sp1 at a promoter proximal GT-box. Sp factors are acetylated in response to HDAC inhibitors, bind to the conserved GT-box of the MMP28 promoter, potentially working in conjunction with HDAC1, p300, P/CAF and serine threonine receptor-associated protein (STRAP).

3



us cri pt

an

Cell culture HeLa cells were cultured in DMEM (Invitrogen) supplemented with 10% fetal bovine serum (FBS) (Sigma) 2mM glutamine, 100 lU/ml penicillin and 100 µg/ml streptomycin. Serum free conditions omitted the FBS. For assays, cells were grown to confluency and starved of serum for 24 hours prior to treatment for 6 hours with trichostatin A (TSA) (Calbiochem) (10-500 ng/ml), valproic acid (VPA) (Calbiochem) (0.5-8 mM), sodium butyrate (NaB) (Calbiochem) (0.5-10 mM), okadaic acid (R & D systems) (100 nM), UO126 (R & D systems) (10 µM), SP610025 (R & D systems) (20 µM), SB203580 (R & D systems) (1 µM).

dM

RNA isolation and cDNA synthesis Cells were grown to confluency and RNA extracted using Trizol reagent (Invitrogen). cDNA was synthesised from 1 µg total RNA using Superscript II reverse transcriptase (Invitrogen) and random hexamers in a total volume of 20 µl according to the manufacturer instructions. cDNA was stored at -20oC.

pte

Quantitative RT-PCR MMP primers and probes were previously described and validated [7]. The 18S ribosomal RNA gene was used as an endogenous control to normalise for RNA loading (18S rRNA primers and probes were purchased from Applied Bioscience).

ce

Relative quantification of genes by the standard curve method was performed using the ABI Prism 7700 sequence detection system in accordance with the manufacturer’s protocol. PCR reactions contained 5ng of reverse transcribed RNA (1ng for 18S), 50% Taqman 2X Master mix (Applied Biosystems), 100 nM of each primer and 200 nM probe in a total volume of 25 µl. Conditions for the PCR reaction were 2 min at 50oC, 10 min at 95 oC and 40 cycles of 15 sec at 95oC and 1 min at 60oC.

Ac


EXPERIMENTAL Plasmid constructs The pGL3-344MMP28wt (wild-type), pGL3-344MMP28mt (mutant) and MMP28 promoter deletion constructs were as previously described [24]. The Sp1 and Sp3 cDNAs were a kind gift from Professor G. Suske (Marburg, Germany) whilst Sp2 cDNA was purchased from MRC Geneservice. These were subcloned into a modified pcDNA4 vector to include a C-terminal Flag tag to give pcDNA4flagSp1, pcDNA4flagSp2, and pcDNA4flagSp3. pCDNA3GAL4-p300, pRSV-CBP, pcDNA-PCAF were donated by Prof Tony Kouzarides (University of Cambridge). The STRAP expression construct pcDNA3.1-STRAP was a gift from Dr Pran Datta (Vanderbilt University Medical Centre, Nashville, USA).

Reporter assays HeLa cells were plated at 2x104 cells per well in a 24 well plate and grown overnight to approximately 80% confluency. Cells were transiently transfected with Fugene6 (Roche) at a 3:1 ratio of Fugene:DNA according to manufacturers instructions and incubated for 24 hours. Unless otherwise stated, promoter-reporter plasmids were used at 200ng per well, with co-transfected expression plasmids (or empty vector equivalents) used at

4



us cri pt

pte

dM

an

Immunoprecipitation 2x106 HeLa cells in 10cm dishes were transiently transfected with 10µg pcDNA4FlagSp1 or pcDNA4Flag-Sp3 expression plasmids using Fugene6 (as above) and incubated for 24 hours. Cells were washed with HBSS three times and incubated overnight with serum free media. Cells were treated with or without 50 ng/ml trichostatin A (TSA) for 6 hours. Cells were washed with cold PBS and harvested in 500 µl RIPA buffer (50 mM Tris-HCl pH7.4/ 1% NP40/ 0.25% deoxycholate/ 150 mM NaCl/ 1 mM EDTA/ 1 mM PMSF/ 1 mM NaF/ 1 EDTA-free protease inhibitor tablet (Roche)) and incubated at 4oC for 15 minutes with rotation. Protein concentration in the supernatant was measured and equalised. The supernatant was pre-cleared by adding 30 µl 50% (v/v) slurry of protein A agarose beads:PBS for 15 minutes whilst rotating at 4oC. Beads were pelleted at 13000rpm for 2 minutes. The supernatant was incubated with 4µl rabbit anti-Sp1 IgG or rabbit anti-Sp3 IgG (Santa Cruz sc-59x, sc-644x) for 2 hours at 4oC with rotation. A 50% slurry of protein A agarose beads:PBS was added and incubated at 4oC for a further hour whilst rotating. The beads were washed 3 times in RIPA buffer and then resuspended in SDS- sample buffer and heated to 100oC for 5 minutes; proteins were then separated on a 10% SDS-PAG. The proteins were transferred onto PVDF membrane, blocked with 5% non-fat milk powder in PBS-Tween20 (0.1%) and probed with antibodies against panacetyl lysine residues (Cell Signalling Technologies #9441S). The membranes were stripped with Strong stripping buffer (Chemicon) and reprobed with anti-Sp1 or anti-Sp3 antibodies and visualised with chemiluminesence (Pierce).

ce

Electromobility shift assay (EMSA) HeLa cells were grown to confluency in a 15cm plate and washed with ice-cold PBS. Cells were pelleted at 500 x g and resuspended in 1ml of PBS containing 0.1% (v/v) Nonidet P-40 for approximately 30 seconds. After centrifugation at 12,000 x g for 10 seconds, supernatant was removed and the nuclear pellet was resuspended in 3 volumes of high salt buffer (25 mM Hepes pH7.9/ 500 mM KCl/ 0.5 mM MgSO4/ 1 mM DTT/ 1x EDTA-free protease inhibitors (Roche)) and incubated on ice for 20 minutes. The lysate was centrifuged for 2 minutes at 12,000 x g and the supernatant collected. The protein concentration was determined by Bradford reagent (Biorad). Nuclear extracts were aliquoted and stored at -80oC.

Ac


100ng per well (Figure 4 and 6), or 10ng/well (Figure 9), following initial titration . Hence, within an experiment, the total amount of plasmid DNA transfected into each well was constant. The cells were washed in Hank’s Balanced Salt Solution (HBSS, Invitrogen) three times and incubated in serum free media for 24 hours. When required, the cells were either treated with TSA (50 ng/ml) for 6 hours or mock treated with DMSO. Cells were washed with ice cold PBS and lysed in 1X passive lysis buffer (Roche) according to manufacturers instructions. Luciferase assays were performed following manufacturers instructions (Roche) and presented as relative light units or fold change compared to control.

MMP28 oligonucleotides (MMP28 wild-type, Wt 5’-GCG GGG TGG GTG GGG CGG GAG/ MMP28 mutant, Mt 5’-GCG GGG TGG TTT GGG CGG GAG) or Sp1 consensus oligonucleotides (5’ ATT CGA TCG GGG CGG GGC GAG C) were resuspended in

5



us cri pt

an

Oligonucleotide pulldown Assay MMP28 oligonucleotides (wild-type and mutant, see above) were biotinylated on the 5’ end of the top strand (Sigma Genosys). Oligonucleotides (0.1 µg/µl) were resuspended in annealing buffer (20 mM Tris-HCl pH7.4/ 2 mM MgCl2/ 50mM NaCl), heated to 95oC for 5 minutes and left to cool to room temperature.

ce

pte

dM

HeLa cells were grown to confluency in a 15 cm dish. Cells were washed in PBS and lysed in 500 µl Schindler’s buffer (50 mM Tris pH7.9/ 0.1 mM EDTA/ 10% glycerol/ 10 mM NaF/ 150 mM NaCl/ 1% NP40/ 1 mM DTT/ 1x EDTA-free protease inhibitors (Roche)) and incubated for 30 minutes at 4oC with rotation. The lysate was centrifuged at 12,000 x g for 5 minutes. The supernatant was quantified for protein concentration using Bradford reagent. The lysate was pre-cleared with 50% (v/v) slurry of streptavidin coated resin (Pierce): Schindler’s buffer for 15 minutes at 4oC whilst rotating. The resin was removed by centrifugation at 12,000 x g for 1 minute. The supernatant was incubated with 1 µg annealed oligonucleotide and 50 µl 50% streptavidin resin and incubated at 4oC for 2 hours whilst rotating. The beads were washed 3 times in Schindler’s buffer. The protein was eluted in elution buffer (7 M urea, 4% CHAPS, 2 M thiourea, 50 mM DTT, 1x protease inhibitor tablet Roche) and separated by SDS-PAGE. An approximately 37kD band was excised from the wild-type lane and analysed by tryptic digest and mass spectroscopy. STRAP was identified using Mascot software (www.matrixscience.com). siRNA knockdown 2x104 HeLa cells were plated in each well of a 24 well plate and incubated overnight. The cells were tranfected with siRNA targeted to HDAC1 (Dharmafect SmartPool) using Dharmafect1 (Dharmacon) and incubated for 24 hours at 37oC. The media was changed and the cells were further transfected with 0.2 µg pGL3-344 MMP-28wt promoter construct using Fugene6 (Roche) and incubated for 24 hours. The cells were washed with PBS and lysed in 1X lysis buffer (Roche). Luciferase assays were performed following manufacturers instructions (Roche). In parallel, wells were harvested at the end of the experiment into TRIzol and RNA purified. This was reverse transcribed as above and 50ng cDNA used in RT-PCR for HDAC1 using forward primer 5’-

Ac


annealing buffer (20 mM Tris-HCl pH7.4/ 2 mM MgCl2/ 50mM NaCl) to a concentration of 100 µM. Oligonucleotides were heated to 95oC for 5 minutes and left to cool to room temperature. Annealed oligonucleotides were end-labelled with 32P-γATP (Amersham) using polynucleotide kinase (Promega) and unincorporated radioactivity removed on a G25 spin column (GE Healthcare). 1 µl probe was incubated with 5 µg nuclear extract, 1 µg dI-dC, 1 µl Binding buffer (100 mM Tris-HCl 7.5/ 5 mM MgCl2/ 5 mM DTT/ 500 mM NaCl/ 50% glycerol), 1 µl competitor in 10 µl final volume with H20. The binding reaction was incubated on ice for 30 minutes. For super-shifts the nuclear extract was pre-incubated with 1 µl antibody (anti-Sp1 antibody and anti-Sp3 antibody were Santa Cruz sc-59x, sc-644x respectively) on ice for 20 minutes prior to the addition of remaining components of the binding reaction. Loading buffer (1 µl) without dye (30% glycerol) was added and complexes separated on a 5% acrylamide gel at 150V. The gel was dried and visualised by autoradiography.

6



CAGCCTAGTGCGGTGGTCTTACAGTG-3’ and reverse primer CTCCCAGCATCAGCATAGGCAGGTTA-3’ and annealing temperature of 58°C.

5’-

us cri pt

an

RESULTS MMP28 expression is induced by HDAC inhibitors Our previous studies demonstrated that HDAC inhibitors have no effect on the basal expression of the majority of MMP genes, but our early data in SW1353 chondrosarcoma cells showed an induction in MMP17, MMP23 and MMP28 expression [19]. Figure 1 shows that in HeLa cells MMP28 expression is induced in a dose-dependent manner by trichostatin A (TSA) (10 – 100ng/ml) (Figure 1a), sodium butyrate (NaB) (0.5-10mM) (Figure1b) and valproic acid (VPA) (0.5 – 8mM) (Figure1c), with maximum induction seen at 50ng/ml TSA, 5mM NaB and ≥ 8mM VPA. The dose-dependent induction of MMP28 expression by these different chemical classes of HDAC inhibitor demonstrates the specificity of the effect since the non-specific activities of each class are unlikely to overlap. A similar pattern of induction is seen in HeLa, SW1353, C28/I2 and SV40MRC5 cells and further experiments were performed with the former line.

dM

Figure 2a shows that expression from human MMP28 promoter-reporter constructs, transiently transfected into HeLa cells, is also induced by TSA. A deletion set from 3031 to -344 bp of the human MMP28 promoter shows strong (between ~16-fold and ~27-fold) induction of reporter gene expression. Deletion between -344 to -298 bp leads to a drop in TSA induction to ~7-fold with the empty pGL3-basic vector retaining ~2.5fold induction.

pte

The -344 / -298 region contains a previously described GT-box which contributes to basal expression of the gene in other cell types. Figure 2b shows that functional mutation of this site (pGL3-344MMP28mt) abrogates TSA induction of the -344 bp MMP28 wildtype (pGL3-344MMP28wt) promoter construct.

ce

The MMP28 promoter sequence binds Sp1 and Sp3 transcription factors Electrophoretic mobility shift assay shows the same pattern of binding to the MMP28 GT-box and to a consensus Sp1-binding oligonucleotide (Figure 3a and data not shown). Of the specific protein complexes observed, supershift analysis identifies the slowest migrating complex as containing Sp1, whilst two more rapidly migrating complexes contain Sp3 (Figure 3b). No evidence of Sp2 binding was observed. Recombinant Sp1 alone binds poorly to the oligonucleotide but binding is greatly enhanced with the addition of nuclear extract suggesting synergy with other nuclear proteins.

Ac


Statistical analysis Comparison between groups was by Student’s t-test. In order to combine data from replicate experiments, data are plotted as fold against control.

TSA induces MMP28 promoter via Sp1 at the GT box The pGL3-344MMP28wt or pGL3-344MMP28mt promoter constructs were cotransfected with pcDNA4FlagSp1, pcDNA4FlagSp2 or pcDNA4FlagSp3. Figure 4a

7



us cri pt

shows that over-expression of Sp1 induces expression of the -344 MMP28wt construct, and this is abrogated by the GT-box mutation. Sp3 over-expression does not induce expression from the -344 MMP28wt promoter construct. Similarly Sp2 has no effect (data not shown).

dM

an

TSA induces acetylation but not expression of both Sp1 and Sp3 Western blot analysis (either before or after immunoprecipitation) demonstrates that treatment of cells with TSA does not increase expression of Sp1 or Sp3 (Figure 5a, c, d and f). The literature suggests that both of these factors can be acetylated [23, 25-27]. In order to demonstrate this in our system, cells were treated with or without TSA and whole cell extracts harvested. Extracts were immunoprecipitated with anti-Sp1 or antiSp3 antibodies, and these immunoprecipitates subjected to western blot and probed first with an anti-(pan acetyl lysine) antibody, then stripped and reprobed with the precipitating antibody. Figure 5b and e show that both Sp1 and Sp3 are acetylated in response to TSA treatment, but interestingly, only one isoform of each protein identified by western with each cognate antibody appears to be acetylated The IP-westerns do show additional bands, not seen in the westerns before IP. A band of approximately 50kD is a non-specific cross-reaction with the heavy chain of IgG and othe bands are likely due to an increase in sensitivity since the IP effectively concentrates the starting sample.

ce

pte

Histone acetyl transferases mediate Sp1 induction of the MMP28 promoter Since TSA mediates its induction of MMP28 via Sp1, and Sp1 is acetylated in response to TSA, a number of HATs were assessed as potential Sp1 acetyl transferases in this system. Expression constructs for p300, CBP, P/CAF and equivalent mutants lacking the HAT domain, were co-transfected with the pGL3-344MMP28wt construct +/- an Sp1 expression construct. Figure 6a shows that basal expression from the pGL3344MMP28wt promoter is induced by p300 but not by CBP or P/CAF, compared to the empty vector controls. This is largely dependent on an active HAT domain (compare delta HAT construct with wild-type). In the additional presence of Sp1, p300 and P/CAF (but again, not CBP) further induce the pGL3-344MMP28wt promoter. Whilst Sp1 does induce the pGL3-344MMP28mt construct to some extent, p300 cannot further induce the expression though P/CAF does, suggesting an additional site of action for this HAT (Figure 6b). siRNA-mediated knock down of HDAC1 induces the MMP28 promoter TSA inhibits all of the 11 classical HDACs to some extent [28]. HDAC1 has been implicated in Sp1-mediated repression of a number of genes [26, 29]. Cells were transfected with HDAC1 siRNA or a scrambled version, then 24 hours later, further transfected with the pGL3-344MMP2wt promoter construct. Figure 7a demonstrates that knock down of HDAC1 expression induces the MMP28 promoter. Figure 7b shows that

Ac


Figure 4b shows that treatment with TSA superinduces the Sp1-induced pGL3344MMP28wt reporter expression, demonstrating that TSA mediates its induction of the pGL3-344MMP28wt construct via Sp1 at the GT-box.

8



siRNA targeted to HDAC1 successfully knocks down expression of HDAC1 at the mRNA level.

us cri pt

an

Since the MEK-ERK pathway can lead to phosphorylation of Sp1, we repeated the above experiment treating cells with either ERK, p38 or JNK inhibitors prior to treatment with TSA. Interestingly, Figure 8b shows that the JNK inhibitor SP610025 decreases the induction of the MMP28 promoter by TSA. The impact of either pharmacological inhibitor could be direct or indirect and the effect of the JNK inhibitor may suggest a role for c-Jun in MMP28 expression.

pte

dM

STRAP binds to the MMP28 promoter In order to identify other proteins which might form complexes at the MMP28 promoter GT-box we performed oligonucleotide-pulldown assays with a 37mer biotinylated oligonucleotide which spans the conserved GT-box. Protein complexes formed on the wild-type oligonucleotide were compared with those on a mutant oligonucleotide with a two base pair mutation in the GT-box. Pulldowns were performed and proteins were separated on a SDS-PAG and silver stained (Figure 9a). A 37 kDa protein was identified which bound the wild-type oligonucleotide but not the mutant oligonucleotide. This band was excised and identified by mass spectrometry as serine threonine receptor associated protein (STRAP). Of 31 peptides in the tryptic digests, 9 were used by Mascot to identify STRAP with a probability-based Mowse score of 94 (p=0.0012). Sp1 was not identified as a differentially bound protein in this experiment, likely reflecting its abundance or its ability to bind to sites outside of the GT-box in the mutant sequence.

ce

STRAP increases the Sp1 induction of the MMP28 promoter In order to test whether STRAP has any effect on the MMP28 promoter we co-transfected HeLa cells with the pGL3-344MMP28wt luciferase reporter, pcDNA4Flag-Sp1 and pcDNA3.1-STRAP. STRAP alone had no effect on the promoter above background. However, when STRAP was co-transfected with Sp1, STRAP was able to increase the Sp1 induction of the MMP28 promoter (Figure 9b).

Ac


Phosphorylation events impact upon TSA induction of the MMP28 promoter Since Sp1 is known to undergo phosphorylation either increasing or decreasing its activity [30] and in some signalling pathways phosphorylation and acetylation can be competing modifications, cells transfected with the pGL3-344MMP28wt promoterreporter construct were treated with a protein phosphatase inhibitor, okadaic acid. This compound has previously been shown to increase the level of Sp1 phosphorylation [30]. Figure 8a clearly shows that okadaic acid reduces the ability of TSA to induce the MMP28 promoter from approximately 8-fold to 2-fold.

DISCUSSION This study presents evidence that the expression of MMP28 is regulated by acetylation. Previous studies have identified conserved motifs that could control the expression of MMP28 including a GT-box situated 310 bp upstream from the translation start site that binds both Sp1 and Sp3 proteins and is essential for MMP28 basal expression in keratinocytes [24]. We have shown that broad spectrum HDAC inhibitors such as TSA

9



dM

an

us cri pt

Gel mobility shift and supershift assays have shown that Sp1 and Sp3 factors can bind to the MMP28 promoter GT box sequence. However, transient transfection assays with the wild type and mutant promoters show that Sp1, but not Sp3 can induce the MMP28 promoter via this site. Some minimal transactivation of the pGL3-344MMP28mt construct by Sp1 suggests sequences outside the GT box within -344 can mediate this effect and other GC-rich sequences can be identified. Sp1 and Sp3 both bind GT/GCrich sequences, these interactions can be cooperative or Sp3 can decrease Sp1-dependent transactivation with the ratio of each protein being a determining factor. When these factors were added together, no synergy or competition was observed at the MMP28 promoter (data not shown). This could be due however to the delicate balance of each factor required. Studies have shown that the DNA-binding affinity of Sp factors and their ability to transactivate are influenced at the post-translational level by both phosphorylation and acetylation [16, 23, 27, 31, 32]. Western blot and immunopreciptitaion experiments showed that TSA does not affect the expression of Sp1 and Sp3 but does induce their acetylation. Acetylation of Sp3 in its inhibitory domain can turn it from a transcriptional repressor to an activator [22]. This does not occur on the MMP28 gene since even in the presence of TSA, Sp3 still does not transactivate the MMP28 promoter (data not shown). Studies have shown that Sp1 can be acetylated at Lys 703 [23, 26, 27]. This acetylation can alter protein-protein interactions and modify the Sp1-containing protein complexes which act at GC/GT-boxes.

pte

The phosphatase inhibitor okadaic acid, previously shown to increase phosphorylation of Sp1, abrogates both basal expression of the MMP28 promoter and its induction by TSA. Phosphorylation of Sp1 has been shown to increase or decrease its transactivation ability dependent on phosphorylation site and biological context [30] The specific JNK inhibitor is the only MAPK inhibitor to impact upon TSA induction of the MMP28 promoter. Since it reduces the TSA induction of the MMP28 promoter (rather than enhancing it, which would be expected if JNK were the Sp1 kinase), this is not likely to be via phosphorylation of Sp1.

ce

Transcriptional co-activators such as p300/ CBP and P/CAF are recruited to sequence specific transcription factors and have intrinsic HAT activity. Recruitment of P/CAF and p300 has previously been shown to be correlated with acetylation of Sp1 after treatment with TSA [23] We showed that both P/CAF and p300 were able to induce the MMP28 promoter in the presence of Sp1 and at least for p300 this is HAT dependent.

Ac


and NaB can induce MMP28 expression in a number of cell types and that this effect is predominantly mediated through the conserved GT-box with functional mutation abrogating induction of the MMP28 promoter by TSA

Since it had previously been shown that Sp1 could recruit HDAC1 [26, 33], we investigated the function of this HDAC on the MMP28 gene. Knockdown of HDAC1 with siRNA induced the MMP28 promoter in the same manner as treatment with TSA, suggesting that inhibition of HDAC1 is in part responsible for the action of broad spectrum HDAC inhibitors on MMP28 expression. Sp1 has previously been reported to regulate the 12(S)-lipoxygenase gene by recruiting both c-Jun and HDAC1 [26, 34]. A

10



an

us cri pt

Recombinant Sp1 alone does not bind strongly to the MMP28 promoter GT-box, but binding is enhanced in the additional presence of nuclear extracts. As discussed above, this again suggests that Sp1 binds and acts in complex with other protein factors. Oligonucleotide pulldown assays were performed to identify novel proteins that bind to the MMP28 promoter GT-box. We identified a 37kDa protein that bound to the wild-type GT-box sequence, but not to the non-functional mutant. This was identified as serine threonine receptor associated protein STRAP. STRAP is a transforming growth factor (TGF)-beta receptor-interacting protein that associates with both type I and type II TGFbeta receptors and is involved in TGF-beta signaling [35, 36]. STRAP synergizes with Smad7 and inhibits TGF-beta-induced transcriptional responses. The expression of STRAP is also increased in a number of human cancers [37] and it can act via TGF-beta independent mechanisms, e.g. via the ERK pathway [38]. STRAP can localize to both cytoplasm and nucleus and has been shown to inhibit the co-activator function of the Ewing sarcoma (EWS) protein by displacing p300 from the activation complex [39]. We have shown that transfected STRAP had little or no effect alone at the MMP28 promoter but was able to increase Sp1-mediated activation further when co-transfected.

pte

dM

In summary, we present evidence that the expression of MMP28 is regulated by acetylation. The MMP28 gene is induced by HDAC inhibitors in a number of cell lines. This is dependent on a promoter proximal GT-box which binds a protein complex containing Sp1. A number of proteins are involved in transactivation of the gene including p300, P/CAF, HDAC1, STRAP and potentially c-Jun. A number of HDAC inhibitors are being developed as cancer chemotherapeutics and it has been mooted that the ability of such compounds to repress expression of some MMPs may enhance their efficacy [40, 41]. Our study underlines the need to examine the effects of such compounds across the MMP family, since MMP-28 is induced by HDAC inhibitors and may also promote tumourigenesis and/or metastasis.

ce

Acknowledgements: Emily Hutchinson (UEA) is acknowledged for the preliminary promoter assays. We are grateful to Dr. Pran Datta for the STRAP constructs and Professor Tony Kouzarides for the HAT constructs. Funding: TES was funded by Wellcome Trust [073379/Z/03/Z] and Dunhill Medical Trust. LK was funded by BBSRC and Astra Zeneca (CASE).

Ac


role for c-Jun in the Sp1-site dependent regulation of MMP28 expression is also consistent with the ability of the JNK inhibitor SP610025 to abrogate TSA induction of the MMP28 promoter.

REFERENCES

1 Mott, J. D. and Werb, Z. (2004) Regulation of matrix biology by matrix metalloproteinases. Curr. Opin. Cell Biol. 16, 558-564 2 Lohi, J., Wilson, C. L., Roby, J. D. and Parks, W. C. (2001) Epilysin, a novel human matrix metalloproteinase (MMP-28) expressed in testis and keratinocytes and in response to injury. J. Biol. Chem. 276, 10134-10144

11


ce

pte

dM

an

us cri pt

3 Illman, S. A., Keski-Oja, J., Parks, W. C. and Lohi, J. (2003) The mouse matrix metalloproteinase, epilysin (MMP-28), is alternatively spliced and processed by a furinlike proprotein convertase. Biochem. J. 375, 191-197 4 Rodgers, U. R., Kevorkian, L., Surridge, A. K., Waters, J. G., Swingler, T. E., Culley, K., Illman, S., Lohi, J., Parker, A. E. and Clark, I. M. (2009) Expression and function of matrix metalloproteinase (MMP)-28. Matrix Biol. 28, 263-272 5 Bernal, F., Hartung, H. P. and Kieseier, B. C. (2005) Tissue mRNA expression in rat of newly described matrix metalloproteinases. Biol. Res. 38, 267-271 6 Marchenko, G. N. and Strongin, A. Y. (2001) MMP-28, a new human matrix metalloproteinase with an unusual cysteine-switch sequence is widely expressed in tumors. Gene. 265, 87-93 7 Nuttall, R. K., Pennington, C. J., Taplin, J., Wheal, A., Yong, V. W., Forsyth, P. A. and Edwards, D. R. (2003) Elevated membrane-type matrix metalloproteinases in gliomas revealed by profiling proteases and inhibitors in human cancer cells. Mol. Cancer Res. 1, 333-345 8 Kevorkian, L., Young, D. A., Darrah, C., Donell, S. T., Shepstone, L., Porter, S., Brockbank, S. M., Edwards, D. R., Parker, A. E. and Clark, I. M. (2004) Expression profiling of metalloproteinases and their inhibitors in cartilage. Arthritis Rheum. 50, 131141 9 Davidson, R. K., Waters, J. G., Kevorkian, L., Darrah, C., Cooper, A., Donell, S. T. and Clark, I. M. (2006) Expression profiling of metalloproteinases and their inhibitors in synovium and cartilage. Arthritis Res. Ther. 8, R124 10 Saarialho-Kere, U., Kerkela, E., Jahkola, T., Suomela, S., Keski-Oja, J. and Lohi, J. (2002) Epilysin (MMP-28) expression is associated with cell proliferation during epithelial repair. J. Invest. Dermatol. 119, 14-21 11 Illman, S. A., Lehti, K., Keski-Oja, J. and Lohi, J. (2006) Epilysin (MMP-28) induces TGF-beta mediated epithelial to mesenchymal transition in lung carcinoma cells. J. Cell Sci. 119, 3856-3865 12 Yan, C. and Boyd, D. D. (2007) Regulation of matrix metalloproteinase gene expression. J. Cell Physiol. 211, 19-26 13 Suske, G. (1999) The Sp-family of transcription factors. Gene. 238, 291-300 14 Sapetschnig, A., Koch, F., Rischitor, G., Mennenga, T. and Suske, G. (2004) Complexity of translationally controlled transcription factor Sp3 isoform expression. J. Biol. Chem. 279, 42095-42105 15 Thomas, K., Wu, J., Sung, D. Y., Thompson, W., Powell, M., McCarrey, J., Gibbs, R. and Walker, W. (2007) SP1 transcription factors in male germ cell development and differentiation. Mol. Cell. Endocrinol. 270, 1-7 16 Black, A. R., Jensen, D., Lin, S. Y. and Azizkhan, J. C. (1999) Growth/cell cycle regulation of Sp1 phosphorylation. J. Biol. Chem. 274, 1207-1215 17 Chung, Y. L., Lee, M. Y., Wang, A. J. and Yao, L. F. (2003) A therapeutic strategy uses histone deacetylase inhibitors to modulate the expression of genes involved in the pathogenesis of rheumatoid arthritis. Mol. Ther. 8, 707-717 18 Lin, H. S., Hu, C. Y., Chan, H. Y., Liew, Y. Y., Huang, H. P., Lepescheux, L., Bastianelli, E., Baron, R., Rawadi, G. and Clement-Lacroix, P. (2007) Anti-rheumatic activities of histone deacetylase (HDAC) inhibitors in vivo in collagen-induced arthritis in rodents. Br. J. Pharmacol. 150, 862-872

Ac



12


ce

pte

dM

an

us cri pt

19 Young, D. A., Lakey, R. L., Pennington, C. J., Jones, D., Kevorkian, L., Edwards, D. R., Cawston, T. E. and Clark, I. M. (2005) Histone deacetylase inhibitors modulate metalloproteinase gene expression in chondrocytes and block cartilage resorption. Arthritis Res. Ther. 7, R503-512 20 Clark, I. M., Swingler, T. E. and Young, D. A. (2007) Acetylation in the regulation of metalloproteinase and tissue inhibitor of metalloproteinases gene expression. Front. Biosci. 12, 528-535 21 Ammanamanchi, S., Freeman, J. W. and Brattain, M. G. (2003) Acetylated sp3 is a transcriptional activator. J. Biol. Chem. 278, 35775-35780 22 Braun, H., Koop, R., Ertmer, A., Nacht, S. and Suske, G. (2001) Transcription factor Sp3 is regulated by acetylation. Nucleic Acids Res. 29, 4994-5000 23 Huang, W., Zhao, S., Ammanamanchi, S., Brattain, M., Venkatasubbarao, K. and Freeman, J. W. (2005) Trichostatin A induces transforming growth factor beta type II receptor promoter activity and acetylation of Sp1 by recruitment of PCAF/p300 to a Sp1.NF-Y complex. J. Biol. Chem. 280, 10047-10054 24 Illman, S. A., Keski-Oja, J. and Lohi, J. (2001) Promoter characterization of the human and mouse epilysin (MMP-28) genes. Gene. 275, 185-194 25 Brasse-Lagnel, C., Fairand, A., Lavoinne, A. and Husson, A. (2003) Glutamine stimulates argininosuccinate synthetase gene expression through cytosolic Oglycosylation of Sp1 in Caco-2 cells. J. Biol. Chem. 278, 52504-52510 26 Hung, J. J., Wang, Y. T. and Chang, W. C. (2006) Sp1 deacetylation induced by phorbol ester recruits p300 to activate 12(S)-lipoxygenase gene transcription. Mol. Cell. Biol. 26, 1770-1785 27 Ryu, H., Lee, J., Olofsson, B. A., Mwidau, A., Dedeoglu, A., Escudero, M., Flemington, E., Azizkhan-Clifford, J., Ferrante, R. J. and Ratan, R. R. (2003) Histone deacetylase inhibitors prevent oxidative neuronal death independent of expanded polyglutamine repeats via an Sp1-dependent pathway. Proc. Natl. Acad. Sci. U S A. 100, 4281-4286 28 Yoshida, M., Kijima, M., Akita, M. and Beppu, T. (1990) Potent and specific inhibition of mammalian histone deacetylase both in vivo and in vitro by trichostatin A. J. Biol. Chem. 265, 17174-17179 29 Doetzlhofer, A., Rotheneder, H., Lagger, G., Koranda, M., Kurtev, V., Brosch, G., Wintersberger, E. and Seiser, C. (1999) Histone deacetylase 1 can repress transcription by binding to Sp1. Mol. Cell. Biol. 19, 5504-5511 30 Chu, S. and Ferro, T. J. (2005) Sp1: regulation of gene expression by phosphorylation. Gene. 348, 1-11 31 Bouwman, P. and Philipsen, S. (2002) Regulation of the activity of Sp1-related transcription factors. Mol. Cell. Endocrinol. 195, 27-38 32 Duan, H., Heckman, C. A. and Boxer, L. M. (2005) Histone deacetylase inhibitors down-regulate bcl-2 expression and induce apoptosis in t(14;18) lymphomas. Mol. Cell. Biol. 25, 1608-1619 33 Kang, J. E., Kim, M. H., Lee, J. A., Park, H., Min-Nyung, L., Auh, C. K. and Hur, M. W. (2005) Histone deacetylase-1 represses transcription by interacting with zincfingers and interfering with the DNA binding activity of Sp1. Cell. Physiol. Biochem. 16, 23-30

Ac



13


ce

pte

dM

an

us cri pt

34 Chen, B. K., Tsai, T. Y., Huang, H. S., Chen, L. C., Chang, W. C., Tsai, S. B. and Chang, W. C. (2002) Functional role of extracellular signal-regulated kinase activation and c-Jun induction in phorbol ester-induced promoter activation of human 12(S)lipoxygenase gene. J. Biomed. Sci. 9, 156-165 35 Datta, P. K., Chytil, A., Gorska, A. E. and Moses, H. L. (1998) Identification of STRAP, a novel WD domain protein in transforming growth factor-beta signaling. J. Biol. Chem. 273, 34671-34674 36 Seong, H. A., Jung, H. and Ha, H. (2007) NM23-H1 tumor suppressor physically interacts with serine-threonine kinase receptor-associated protein, a transforming growth factor-beta (TGF-beta) receptor-interacting protein, and negatively regulates TGF-beta signaling. J. Biol. Chem. 282, 12075-12096 37 Kim, C. J., Choi, B. J., Song, J. H., Park, Y. K., Cho, Y. G., Nam, S. W., Yoo, N. J., Lee, J. Y. and Park, W. S. (2007) Overexpression of serine-threonine receptor kinaseassociated protein in colorectal cancers. Pathol. Int. 57, 178-182 38 Halder, S. K., Anumanthan, G., Maddula, R., Mann, J., Chytil, A., Gonzalez, A. L., Washington, M. K., Moses, H. L., Beauchamp, R. D. and Datta, P. K. (2006) Oncogenic function of a novel WD-domain protein, STRAP, in human carcinogenesis. Cancer Res. 66, 6156-6166 39 Anumanthan, G., Halder, S. K., Friedman, D. B. and Datta, P. K. (2006) Oncogenic serine-threonine kinase receptor-associated protein modulates the function of Ewing sarcoma protein through a novel mechanism. Cancer Res. 66, 10824-10832 40 Ailenberg, M. and Silverman, M. (2002) Trichostatin A-histone deacetylase inhibitor with clinical therapeutic potential-is also a selective and potent inhibitor of gelatinase A expression. Biochem. Biophys. Res. Commun. 298, 110-115 41 Ailenberg, M. and Silverman, M. (2003) Differential effects of trichostatin A on gelatinase A expression in 3T3 fibroblasts and HT-1080 fibrosarcoma cells: implications for use of TSA in cancer therapy. Biochem. Biophys. Res. Commun. 302, 181-185

Ac



14



Figure legends

us cri pt

an

Figure 2. TSA induces MMP28 promoter via the GT-box. MMP28 promoter constructs were cloned into pGL3 upstream of the luciferase reporter gene. HeLa cells were transiently transfected with these constructs overnight, washed and incubated in serum-free medium for a further 24 hours, then treated with or without TSA (50ng/ml) for 6 hours. Cells were harvested and luciferase assay performed. (a) A deletion set from -3031 to -298 of the MMP28 promoter (b) A two base mutation was introduced into the GT box contained within the pGL3-344MMP28wt construct to give pGL3-344MMP28mt. Data are presented as fold relative light units in the presence of TSA compared to untreated and are plotted as the mean of three independent experiments each in triplicate (n=9) +/s.e.m.

pte

dM

Figure 3. Sp1 and Sp3 bind to the MMP28 GT-box sequence. (a) Electrophoretic mobility shift assays were performed using 32P-labelled 37mer MMP28 wild-type (wt) probe. Lane 1 shows the free probe. Nuclear extracts were prepared from HeLa cells and incubated with the labelled probe (lane 2). Recombinant Sp1 (rec Sp1) protein was incubated with the probe (lane 3) and in the presence of nuclear extract (lane 4). Unlabeled MMP28wt / MMP28 mutant (mt) / Sp1 consensus oligonucleotide cold competitor was added (lanes 5, 6, 7). (b) Supershift assay was performed using antibodies (ab) against Sp1, Sp2 or Sp3.

ce

Figure 4. TSA induces MMP28 promoter via Sp1 at the GT box. (a) HeLa cells were cotransfected overnight with pGL3-344MMP28wt or pGL3-344MMP28mt and either pcDNA4Flag-Sp1, pcDNA4Flag-Sp3 or empty vector. Cells were incubated overnight in serum free medium and luciferase assay performed. (b) HeLa cells were transiently transfected overnight with pGL3-344MMP28wt and either pcDNA4Flag-Sp1 or empty vector. Cells were incubated overnight in serum free medium and treated with or without TSA (50ng/ml) for 6 hours. Cells were harvested and luciferase assay performed. Data are presented as relative light units plotted as mean +/- s.e.m. from a representative experiment in triplicate (n=3).

Ac


Figure 1. MMP28 expression is induced by HDAC inhibitors. HeLa cells were incubated with (a) TSA 0-100ng/ml, (b) NaB 0-10mM or (c) VPA 0-8mM for 6 hours. Cells were harvested and total RNA was isolated and subjected to quantitative RT-PCR for expression of MMP28. Data were normalised to 18S. Results are plotted as mean +/s.e.m., n=3.

Figure 5. TSA induces acetylation of both Sp1 and Sp3. HeLa cells were treated with or without TSA (50ng/ml) for 6 hours. Cells were harvested and whole cells extracts prepared. Cell extracts were immunoprecipitated (IP) with antibodies (α) against Sp1 or Sp3 and protein

15



us cri pt

an

Figure 6. Histone acetyl transferases mediate Sp1 induction of the MMP28 promoter. (a) HeLa cells were co-transfected with pGL3-344MMP28wt construct and HAT expression constructs expressing pcDNAGAL4-p300/ pRSV-CBP/ pcDNA3-P/CAF or empty vectors. Constructs expressing mutant co-activators with no HAT (dHAT) activity were also co-transfected. Following an overnight incubation the cells harvested and luciferase assay performed. (b) HeLa cells were co-transfected with pGL3-344 MMP28wt or pGL3344 MMP28mt constructs and the co-activators and incubated overnight. Luciferase assay was performed as above. Data are presented as relative light units plotted as mean +/- s.e.m. from a representative experiment in triplicate (n=3).

pte

dM

Figure 7. siRNA-mediated knock down of HDAC1 induces the MMP28 promoter. (a) HeLa cells were transfected with siRNA against HDAC1 (30-100nM) or a non-targeting siRNA control (NT) and the cells incubated for 24 hours. The media was changed and the cells were transiently transfected with pGL3-344MMP28wt promoter construct and incubated for a further 24 hours. The cells were harvested and luciferase assay performed. Data are presented as fold luciferase activity compared to empty vector and plotted as the mean of three independent experiments each in triplicate (n=9) +/- s.e.m. (b) HDAC1 knockdown shown by siRNA against HDAC1, non-targeting control (NT), or no siRNA control (C).

ce

Figure 8. Phosphorylation events impact upon TSA induction of the MMP28 promoter. (a) HeLa cells were transfected overnight with pGL3-344MMP28wt promoter construct. Cells were incubated overnight in serum free medium then treated +/- TSA (50ng/ml) and okadaic acid (100nM) for 6 hours prior to luciferase assay. (b) HeLa cells were transfected with pGL3-344MMP28wt promoter construct. Cells were incubated overnight in serum free medium then treated +/- TSA and MEK, p38-MAPK and JNK inhibitors U0126 (10µM), SB203580 (1µM) and SP610025 (20µM) respectively for 6 hours prior to luciferase assay. Data are presented as relative light units plotted as mean +/- s.e.m. from a representative experiment in triplicate (n=3).

Ac


G-Sepharose beads. The beads were washed and proteins eluted by heating in SDSsample buffer. Proteins were separated on a 10% SDS-PAG and transferred onto PVDF membrane for Western blot (WB). (a) Whole cell extracts probed using antibodies against Sp1. (b) Immunoprecitated proteins were probed using a pan-acetyl lysine antibody. (c) The membrane was stripped and reprobed using antibodies against Sp1. (d) Whole cell extracts probed using antibodies against Sp3. (e) Immunoprecitated proteins were probed using a pan-acetyl lysine antibody. (f) The membrane was stripped and reprobed using antibodies against Sp3.

Figure 9.

16



us cri pt

an dM pte ce Ac


STRAP binds to the MMP28 promoter. (a) Whole cell extracts were prepared from HeLa cells. The extracts were incubated with a biotinylated MMP28wt or mutant 37mer oligonucleotide containing the GT-box and streptavidin coated beads. The beads were washed and eluted proteins separated on a 10% SDS-PAG. The gel was silver stained. A 37kDa band that was pulled down with the wt oligonucleotide but absent in the mutant pulldown was excised from the gel and analysed using mass spectrometry. (b) HeLa cells were co-transfected with pGL3-344MMP28wt and a construct expressing STRAP (pcDNA3-STRAP) and Sp1 (pcDNA4Flag-Sp1). The cells were incubated overnight and luciferase assay performed. Data are presented as relative light units plotted as mean +/s.e.m. from a representative experiment in triplicate (n=3).

17



0.5

(a) P< 0.01

0.3 0.2 0.1 0

50

100

P< 0.01

an (c)

0.5

ce

pte

0.1 0.09 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0

dM

Arbitrary units MMP28/18S

10 25 TSA (ng/ml)

P< 0.05

0


5

(b)

1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

Ac


0

P< 0.05

us cri pt


0.4

0

0.5

1 NaB (mM)

5

10

P< 0.001 ***

*** P< 0.001 ** P< 0.05

1 2 VPA (mM)

4

8

Figure 1a.b.c 18



us cri pt

(a)

TSA - 3031

+ - 1486

+ -

- 1197

+ +

- 344

+ +

- 298

Sox-site GT-box

pGL3basic

+

an

-

dM

0

(b)

5

10

15

20

25

30

35

Fold luciferase activity

pte

30

ce

Fold luciferase activity

25

Ac


- 402

20

-TSA +TSA

15

10

5

0

pGL3-344 MMP28wt

pGL3-344 MMP28mt

Figure 2.a.b.

19


(b)

+

-

+

+

+

+ Nuclear extract

-

-

+

+

-

-

-

-

-

-

+

-

- rec Sp1 protein - MMP28 wt cold oligo

-

-

-

-

-

+

- MMP28 mt cold oligo

-

-

-

-

-

-

+ Sp1 cold oligo

-

+

+

+

+ Nuclear extract

-

-

+

-

-

Sp1 ab

-

-

-

+

-

Sp2 ab

-

-

-

-

+ Sp3 ab

dM

an

-

Sp3 Sp 3

ce

pte

Figure 3a.b

Sp1

Ac


(a)

us cri pt


20



(a) 6000

us cri pt

Relative light units

5000 4000 3000 2000

pcDNA4 pcDNA4 pcDNA4 pcDNA4 pcDNA4 pcDNA4 pcDNA4 pcDNA4 Flag Flag Sp1 Flag Flag Sp1 Flag Flag Sp3 Flag Flag Sp3

pGL3 -344 MMP28wt

40000

30000 25000 20000

pte

Relative light units

35000

pGL3 -344 MMP28mt

pGL3 -344 MMP28wt

pGL3 -344 MMP28mt

dM

(b)

an

0

15000 10000 5000

ce

0 pcDNA4 pcDNA4

Flag

Flag

pcDNA4 FlagSp1

-

+

-

pcDNA4 FlagSp1 +

TSA

Ac


1000

Figure 4a.b.

21



(a)

-

(b)

-

+

-

+

250 150 100 75

250 150 100 75

50

50

50

IP αSp1 WB αAc Lys

-

250 150 100

(e)

+

75

50

50

(f)

+

250 150 100

-

+ TSA

75

dM

75

IP αSp3 WB αAc Lys

50

IP αSp3 WB αSp3

pte

WB αSp3

-

250 150 100

TSA

IP αSp1 WB αSp1

an

(d)

+

us cri pt

250 150 100 75

WB αSp1

ce

Figure 5.a. b. c. d. e. f

Ac


(c)

22



(a) 30000

p