Epigenetic Regulation of Vascular Smooth Muscle Cells by Histone H3 Lysine 9 Dimethylation Attenuates Target Gene-Induction by Inflammatory Signaling

Arteriosclerosis, Thrombosis, and Vascular Biology

BASIC SCIENCES Epigenetic Regulation of Vascular Smooth Muscle Cells by Histone H3 Lysine 9 Dimethylation Attenuates Target Gene-Induction by Inflammatory Signaling

Jennifer L. Harman, Lina Dobnikar, Joel Chappell, Benjamin G. Stokell, Amanda Dalby, Kirsty Foote, Alison Finigan, Paula Freire-Pritchett, Annabel L. Taylor, Matthew D. Worssam, Ralitsa R. Madsen, Elena Loche, Anna Uryga, Martin R. Bennett, Helle F. Jørgensen

OBJECTIVE: Vascular inflammation underlies cardiovascular disease. Vascular smooth muscle cells (VSMCs) upregulate selective genes, including MMPs (matrix metalloproteinases) and proinflammatory cytokines upon local inflammation, which directly contribute to vascular disease and adverse clinical outcome. Identification of factors controlling VSMC responses to inflammation is therefore of considerable therapeutic importance. Here, we determine the role of Histone H3 lysine 9 di-methylation (H3K9me2), a repressive epigenetic mark that is reduced in atherosclerotic lesions, in regulating the VSMC inflammatory response.

APPROACH AND RESULTS: We used VSMC-lineage tracing to reveal reduced H3K9me2 levels in VSMCs of arteries after injury and in atherosclerotic lesions compared with control vessels. Intriguingly, chromatin immunoprecipitation showed H3K9me2 enrichment at a subset of inflammation-responsive gene promoters, including MMP3, MMP9, MMP12, and IL6, in mouse and human VSMCs. Inhibition of G9A/GLP (G9A-like protein), the primary enzymes responsible for H3K9me2, significantly potentiated inflammation-induced gene induction in vitro and in vivo without altering NFκB (nuclear factor kappa-light-chain- enhancer of activated B cell) and MAPK (mitogen-activated protein kinase) signaling. Rather, reduced G9A/GLP activity enhanced inflammation-induced binding of transcription factors NFκB-p65 and cJUN to H3K9me2 target gene promoters MMP3 and IL6. Taken together, these results suggest that promoter-associated H3K9me2 directly attenuates the induction of target genes in response to inflammation in human VSMCs.

CONCLUSIONS: This study implicates H3K9me2 in regulating the proinflammatory VSMC phenotype. Our findings suggest that reduced H3K9me2 in disease enhance binding of NFκB and AP-1 (activator protein-1) transcription factors at specific inflammation-responsive genes to augment proinflammatory stimuli in VSMC. Therefore, H3K9me2-regulation could be targeted clinically to limit expression of MMPs and IL6, which are induced in vascular disease.

VISUAL OVERVIEW: An online visual overview is available for this article.

Key Words: arteries ◼ chromatin immunoprecipitation ◼ cytokines ◼ epigenetics ◼ histone ◼ inflammation

ascular smooth muscle cell (VSMC) accumulation is See accompanying editorial on page 2199 a hallmark of cardiovascular disease (CVD), including Vatherosclerosis and vascular remodeling after injury. In contractile proteins and upregulate selective gene sets, response to proinflammatory stimuli, VSMCs downregulate including matrix metalloproteinases, and proinflammatory

Correspondence to: Helle F. Jørgensen, PhD, ACCI Level 6, Box110, Addenbrookes’ Biomedical Campus, Cambridge, CB2 0QQ, United Kingdom. Email hfj22@cam. ac.uk The online-only Data Supplement is available with this article at https://www.ahajournals.org/doi/suppl/10.1161/ATVBAHA.119.312765. For Sources of Funding and Disclosures, see page 2301. © 2019 The Authors. Arteriosclerosis, Thrombosis, and Vascular Biology is published on behalf of the American Heart Association, Inc., by Wolters Kluwer Health, Inc. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution, and reproduction in any medium, provided that the original work is properly cited. Arterioscler Thromb Vasc Biol is available at www.ahajournals.org/journal/atvb

Arterioscler Thromb Vasc Biol. 2019;39:2289–2302. DOI: 10.1161/ATVBAHA.119.312765 November 2019 2289 Basic Sciences - VB implicate MMP12 inplaquedevelopment and instability. implicate MMP12 compared with placebo inpatients compared withplacebo ing antibody, significantlyloweredcardiovascularevents IL1-β ing inflammationwithCanakinumab,an lesions. in vulnerablecomparedwithstablehumanatherosclerotic activityisobserved and MMP12 MMP9, increased MMP3, 2290 Decreased levels ofhistoneH3lysine9 di-methylation or reducedaccessibilitytothe transcriptionmachinery. that modifyhistoneproteins, resultingineitherincreased to theprogressionofvasculardisease. [interleukin]-6),which directlycontribute cytokines (eg,IL Harman etal progression ofvasculardysfunction. protein kinase)/AP-1signalingislinkedtotheinitiationand κ and dysregulationofNF cells) andAP-1(activatorprotein1)transcriptionfactors, (nuclear factorkappa-light-chain-enhancer ofactivatedB κ responsive geneexpression iscontrolledbytheNF cular disease. and mediate many structural changes associated with vas- proteases thatdegradeextracellular matrixcomponents of considerabletherapeuticimportance. geneexpressionassociated changes istherefore inVSMC Identification ofmechanisms thatregulateinflammation- elevated serumIL-6 isassociatedwithgreaterCVD risk. intima formationaftercarotidligationinmice. migrationand neo- promoteVSMC andMMP9 MMP3 proliferation, andvascularcalcification. migration, contractilitywhilepromotingVSMC ates VSMC important roleinvasculardiseaseprogression.IL-6 attenu- VSMC UNC α TNF- NF mVSMC MMP MAPK LCCA JNK IL hVSMC HFD H3K9me2 GLP CVD ChIP AP-1 Nonstandard AbbreviationsNonstandard andAcronyms Gene expression is regulated by epigenetic pathways MMPs (matrix metalloproteinases) comprise a family of (matrixmetalloproteinases)compriseafamilyof MMPs

κB

11–14

November 2019

The proinflammatory cytokine IL-6 The proinflammatory cytokine alsoplays an 6,7 For example, animalatherosclerosisstudies vascular smoothmusclecell UNC0638 tumor necrosisfactor-alpha enhancer ofactivatedBcell nuclear factorkappa-light-chain- murine VSMC murine VSMC matrix metalloproteinase mitogen-activated proteinkinase left commoncarotidartery N-terminalkinase cJUN interleukin human aorticVSMC high-fat diet histone H3Lysine 9di-methyl G9A-like protein cardiovascular disease chromatin immunoprecipitation activator protein-1 B and MAPK (mitogen-activated (mitogen-activated B andMAPK 4 3 althoughnotinmice. Indeed,directlyreduc- 15–17 1,2 Inflammation- Furthermore, Furthermore, 2019;39:2289–2302. DOI: 10.1161/ATVBAHA.119.312765Arterioscler ThrombVasc DOI: Biol. 2019;39:2289–2302. 10 In addition, In addition, neutraliz- 18–20 8,9 21 B 5

cells. contractility of pulmonary and airway smooth muscle mark associatedwithfacultativeheterochromatin sion areunclear. Interestingly, H3K9me2,arepressive gene inflammation-associated expres- changes in VSMC other cells, directly regulatingtheexpression ofinduciblegenesin spurious inductionofaproinflammatorystatein VSMCs. suggest H3K9me2 as an important mechanism to prevent consequences oflossH3K9me2invasculardiseaseand scription factorbinding.These findingsidentifyfunctional κ genes byinhibitingNF attenuates inflammation-associatedupregulationoftarget arteries. human atheroscleroticplaquescomparedwithcontrol derivedfrom (H3K27me3) areobservedinVSMCs (H3K9me2) andhistoneH3lysine27tri-methylation committee. Wild-type andApolipoproteinE-null (Apoe Office (PPL70/7565 andP452C9545)thelocalethics All experiments wereapprovedbytheUnitedKingdomHome Animal Experiments available withinthearticleandin online-onlyDataSupplement. All other supportingdataare accession numberGSE131212. made publicly available at the Gene Expression Omnibus with the datahavebeen The chromatin immunoprecipitation(ChIP)-seq Disclosure Statement METHODS ated with CVD. MMP3, genes,including set ofinflammation-responsive VSMC increased inflammation.H3K9me2isdepositedatasub- undergoing injury-inducedremodeling,concomitantwith arereducedinatherosclerosisandarteries VSMCs remainlargelyunknown. of H3K9me2inVSMCs • • • We demonstratethatgloballevelsofH3K9me2in loproteinase) andcytokinegenepromoters. (matrixmetal- their cognatesequenceswithinMMP AP-1 (activator protein-1) transcriptionfactorsto kappa-light-chain-enhancer ofactivatedBcell)and κB (nuclearfactor H3K9me2 inhibitsbindingofNF matrix degradationandphenotypicswitching. in vitroandvivo,withfunctionalconsequencesfor induced upregulationofdisease-associatedgenes smooth muscle cells exacerbates inflammation- Loss ofH3K9me2inhumanandmousevascular cular diseasemodels. smooth musclecellsoninflammationincardiovas- 9 Dimethylation(H3K9me2)isreducedinvascular The repressiveepigeneticmarkHistoneH3Lysine 29,30 MMP9, 22 Yet, thefunctionalimportanceandtargetgenes However, howthesemechanisms control H3K9me2 Regulates the VSMC InflammatoryResponse H3K9me2 RegulatestheVSMC 24–28 MMP12, andIL6,which arestronglyassoci- 11–14 affects thegrowthrate,migration,and Importantly, we show that H3K9me2 Highlights B/p65 and/or AP-1/cJUN tran- and/orAP-1/cJUN B/p65 −/− 23 ) mice and Basic Sciences - VB is 22 2291 - ). Tamoxi 33 transgene, ERt2 and 40% to 60% 34 November 2019 /Rosa26-Confetti mice ERt2 cells within the media, core, + Importantly, the fluorescent Importantly, with recombination reporter or Rosa26-EYFP 31 32 33,35 mice Immunostaining revealed significantly Immunostaining revealed ERt2+ test, Kruskal-Wallis 1-way ANOVA with Dunn 1-way ANOVA U test, Kruskal-Wallis ), suggesting Supplement), Data online-only - mice to assess H3K9me2 levels in athero 34 41,42 H3K9me2 Regulates the VSMC Regulates H3K9me2 Response Inflammatory −/− Apoe Myh11-Cre 1B). This indicates that the previ- κB p65 (Figure 1B). This To assess whether H3K9me2 regulates early events To We first crossed Myh11-Cre We decreased H3K9me2 levels in lineage-labeled VSMClineage-labeled in levels H3K9me2 decreased arter- nuclei in ligated compared with nonligated control and 28 days post-ligation (Figure 1C and Fig- ies at 5, 7, the in I ure that H3K9me2 may regulate the inflammatory response of VSMCs. because of downregulation in VSMCs. VSMCthe in liga- performed we response, inflammatory elicits repro- which tion of the left common carotid artery, ducible vessel remodeling concomitant with changes in VSMC including induction of IL6 gene expression, and MMPs. with and fibrous cap in atherosclerotic arteries compared with and fibrous cap in atherosclerotic arteries blot analysis cells in control vessels (Figure 1A). Western in aortas also indicated reduced H3K9me2 expression a high-fat diet for 6 months compared with of mice fed control aortas that had little or no visible plaque, and this was associated with increased levels of phosphorylated NF levels in human ath- reduced H3K9me2 ously reported erosclerotic plaques compared with healthy controls fen treatment of these animals induces VSMC-specific treatment of these fen expression protein recombination and stable fluorescent to 95% of VSMCs in Confetti in 70% in medial VSMCsmedial in VSMC-derivedin and and neo-intimal plaque cells. Mann-Whitney Mann-Whitney Wilcoxon exact pairs, 1-sided specific sample test to compare (described in the online-only or a linear model rank-sum tests - threshold for consider ) with P<0.05 as the Data Supplement to be statistically significant. ing results RESULTS VSMCs Within H3K9me2 Is Reduced in Arterial Remodeling, and Atherosclerosis Concomitant Inflammation With VSMCs lineage markers in atheroscle- downregulate specific quantification of making rosis and after injury, We therefore used genetic H3K9me2 in VSMCs difficult. VSMCsdefinitively identify lineage tracing to by cross- ing lineage reporters are stably expressed, independent of lineage reporters are stably expressed, status of the Myh11-Cre the expression in EYFP reporter mice. sclerosis. Male mice received tamoxifen and were then sclerosis. Male mice received tamoxifen high-fat diet (21% fat, an atherosclerosis-inducing fed diet. Immunohis- standard chow or a 0.2% cholesterol) artery sections of animals after 3 in carotid tochemistry significantly lower to 4 months of high-fat diet revealed H3K9me2 levels in Confetti alleles (Rosa26-Confetti

40 Surgery 34 transgene transgene mice have mice −/− ERt2 Rosa26-EYFP /Rosa26-EYFP+ Apoe ERt2 Experimental animals (all Experimental animals (https://rgd.mcw.edu/rgdCura and reads per gene promoter aligned to the mouse GRCm38mouse aligned to the 39 - a marker of differen MYH11 is , Rosa26-Confetti 31,33–36 38 37

transgene is Y-linked) received 10 transgene is Y-linked) 31,32 31 ERt2 (2 ng/mL, Peprotech), and human recom- (2 ng/mL, Peprotech), (Y-linked), , 90 ng/mL, Peprotech (tumor necrosis factor-α), 90 ng/mL, Peprotech ERt2 tion/?module=portal&func=show&name=cardio). males as the Myh11-Cre genome using Bowtie2 v2.2, from patients undergoing aortic valve replacement with ethics mouse committee approval and used at passage 6 to 15. Primary aortic VSMCs were derived from 8- to 12-week-old wild-type males and used at passage 4. VSMCs were treated C57Bl/6 with UNC0638 RNA small interfering tar- (UNC; 1 μM Tocris), RNA Santa Cruz), control small interfering geting G9A (sc-43777, Santa Cruz), SP600125 (10 μM, Abcam), human (sc-37007, α recombinant IL-1 binant TNF-α as indicated and analyzed as described in the Methods in the online-only Data Supplement. been described previously. been described males were treated with A366males were treated with osmotic pumps (30 delivered by mg/kg per day in 98:2 PEG 400/polysorbate 80), or vehicle directly or subjected to ligation of the alone, and either analyzed artery as described previously. left common carotid intraperitoneal injections of 1 mg/mL tamoxifen between 4 and of 1 mg/mL tamoxifen intraperitoneal injections G9A/GLP inhibit (G9A- labeling. To 8 weeks of age for lineage Myh11-Cre like protein), lineage-labeled Statistics Data are shown as mean±SEM, unless otherwise indicated. The number of animals per group and number of biologically inde- least (at legends Figure in indicated are experiments pendent were done using isolates 3). Independent in vitro experiment individuals for hVSMCsderived from different and independent primary cultures for murine VSMCs. Data were analyzed using VSMC Culture Human aortic VSMCs (hVSMCs) were isolated as described ChIP and Analysis ChIP was performed as described in the Methods in the online- genome-wide analysis, the DNA only Data Supplement. For SMART ChIP-seq 634865) kit (Clontech, was used to gen- from 100 pg erate Illumina-compatible sequencing libraries to 2 ng of DNA from 2 independent H3K9me2 ChIP- experi sequenced (Illumina ments and associated input. Libraries were bp reads. ChIP-seq using paired-end 75 reads were NextSeq) v1.9, using Cutadapt trimmed on a C57Bl/6 background were purchased from Charles River. River. from Charles were purchased background on a C57Bl/6 Myh11-Cre (within ±1 kb of the transcription start sites) quantified using (within ±1 kb of the transcription start sites) SeqMonk v1.4 (http://www.bioinformatics.babraham.ac.uk/ counts than 20 read projects/seqmonk). Genes showing fewer further analysis and in the input samples were removed from computed. Of genes the ratio of H3K9me2/input signal was 25th percentile, 63 with H3K9me2/input ratios in the top according to the genes were associated with arteriosclerosis Cardiovascular Disease Portal used here has been extensively tested for VSMC-specific tested has been extensively used here in major arteries. expression and tissue processing are described in the Methods in the are described in the Methods in the and tissue processing . online-only Data Supplement tiated smooth muscle cells and the Myh11-Cre muscle cells and tiated smooth Biol. 2019;39:2289–2302. DOI:Arterioscler Thromb Vasc 10.1161/ATVBAHA.119.312765 (enhanced yellow fluorescent protein), and protein), fluorescent yellow (enhanced Harman et al et Harman Basic Sciences - VB control Myh11-Cre (of Confetti (red, blue,yellow, (4′,6-diamidino-2-phenylindole;white) areshown.The green)andDAPI dotplots showH3K9me2intensityinindividualnuclei ). A indicates adventitia; L, lumen; M, media; NI, neointima;andP, Supplement).Aindicatesadventitia;L,lumen;M, media;NI, Data plaque. (errorbars)areindicated.*P<0.05(linearmodel,see online-only batch (batchesthe same areindicated withsymbols).Mean(line)andSEM 2292 Harman etal , Representative immunofluorescence imagesandquantificationofH3K9me2signalintensityinnucleiConfetti A, Representative and vascularremodeling. inatherosclerosis Figure 1.HistoneH3lysine9dimethylation(H3K9me2)isreducedinvascular smoothmusclecells(VSMCs) carotid arteries(LCCAs) fromlineage-labeledMyh11-Cre quantification ofH3K9me2signal intensity innucleiofConfetti immunofluorescenceimagesand Supplement.C,Representative N.D., notdetected.Pwerecalculatedasdescribedinthe online-onlyData animals,normalizedtoβ-Actin;n=3animalspergroup. HFD SD and inaorticmediafrom and quantificationofH3K9me2p-p65(Ser536) mice.B,WesternHFD blot in n=5animals).H3K9me2quantificationisdisplayedseparatelyformedial,plaquecore,andcap VSMCs (HFD, November 2019 + , nonligated control samples inC)analyzed inA,nonligatedcontrolsamples cells)relativetotheaverageH3K9me2 signalintensityincontrolanimals(SD ERt2+ /Rosa26-Confetti + , Signals for H3K9me2 (magenta), Confettireporterproteins mice.n=5animalspergroup.Aand C, SignalsforH3K9me2(magenta), 2019;39:2289–2302. DOI: 10.1161/ATVBAHA.119.312765Arterioscler ThrombVasc DOI: Biol. 2019;39:2289–2302. ERt2+ /Rosa26-Confetti + cellswithinLCCAs from ligated(7dayspost-ligation)relativetonosurgery + /Apoe −/− males fed a standard (SD, n=3animals)orhigh-fatdiet (SD, malesfedastandard H3K9me2 Regulates the VSMC InflammatoryResponse H3K9me2 RegulatestheVSMC + cellswithinleftcommon Basic Sciences - VB

7 2293 3D), indicating November 2019 . VSMC-lineage-labeled4A). ani- /Rosa26-EYFP) were adminis- 3B), demonstrating that the effect the effect that demonstrating 3B), ERt2 H3K9me2 Regulates the VSMC Regulates H3K9me2 Response Inflammatory nor UNCnor α IL-1 neither contrast, In 3C). - treat α and UNC before incubating with dye-quenched α-induced gelatin digestion (1.6-fold) compared with To assess whether MMP regulation by H3K9me2 To , which is constitutively expressed by VSMCs. is constitutively expressed Mmp2, which tered A366, a G9A/GLP inhibitor that displays similar specificity and efficacy but improved pharmacokinetics We next assessed the effect of reducing H3K9me2 in H3K9me2 reducing of effect the assessed next We vivo on injury-induced MMP following carotid expression (Figure ligation artery mals (Myh11-Cre VSMC-Specific of Inflammation- Upregulation Genes After Responsive H3K9me2-Target Ligation Is Enhanced by G9A/GLPCarotid Inhibition did not affect expression of UNC+IL-1 expression α did not affect Importantly, H3K9me2 showed no Il6 and Ccl2 (Figure 3C), which (Figure enrichment UNCof MMPon of a not because is expression gene This response. general upregulation of the inflammatory to specifically analysis suggests that H3K9me2 acts in response to repress the induction of target genes inflammatory signaling. VSMC matrix remodeling, we could affect extracellular whether G9A/GLP matrix examined inhibition affects treated cultured murine VSMCs with degradation. We IL-1 is converted into fluorescent peptides gelatin, which upon digestion. Inhibition of G9A/GLP further increased IL-1 α alone (Figure VSMCs treated with IL-1 that H3K9me2-regulation of MMP in VSMCs expression vascular remodeling in disease. might affect α-mediated upregulation of significantly potentiated IL-1 and Mmp12 (7.1-fold) Mmp3 (3.2-fold), Mmp9 (1.7-fold), (Figure expression the affected combination, in or alone, ment of (Table (Table online- Mmp9, 39 2A). The The 2A). promoters Mmp12 treatment α IL-1 after Mmp12 and Mmp9, Mmp3, and Mmp9, ), including several including Supplement), Data online-only Mmp3, ChIP-qPCRMmp12. murine primary cultured using To directly test the importance of H3K9me2-marking To The reduction of H3K9me2 in VSMCs in reduction of H3K9me2 The concomitant with examine local to increased inflammation prompted us arteriosclerosis- levels of H3K9me2 at promoters of by ChIP-seqassociated genes. Genome-wide mapping promoters gene at levels H3K9me2 that demonstrated the associated of expression with negatively correlated VSMCsmurine vivo ex IIin (Figure gene the in H3K9me2 Attenuates Inflammation-Induced H3K9me2 Attenuates Inflammation-Induced MMP Gene Induction for a repres- is expected only Data Supplement), which the genes found that among sive epigenetic mark. We transcrip- their of ±1kb within levels H3K9me2 high with arteriosclerosis- as listed 63 genes were site start tion Portal associated on the Cardiovascular Disease Histone H3 Lysine 9 Dimethylation (H3K9me2) is enriched at a subset of IL (H3K9me2) is enriched 9 Dimethylation (interleukin)-1α−responsive gene Figure 2. Histone H3 Lysine promoters. A, Chromatin immunoprecipitation-qPCRIL-1 analysis for H3K9me2 in untreated or smooth muscle cells (mVSMCs;α-treated murine vascular and Mmp12, compared with the promoters of Mmp3, Mmp9, (relative to Magea2) at Actb (negative control) and 6 h). H3K9me2 enrichment show mean±SEM control IgG (immunoglobulin). Graph signal observed using negative cultures. n.s., not significant of 3 independent primary (reverse transcription with quantitativeMmp3, Mmp9, Mmp12, and Hprt1 reaction) analysis of polymerase chain B, RT-qPCR (Kruskal-Wallis). in control and IL-1α-treated mVSMCs. Expression (mean±SEM is shown relative to IL-1 in 6 independent primary cultures) α-treated mVSMCs, U test). ). *P<0.05 (2-tailednormalized to housekeeping genes (Hmbs and Hprt1 Mann-Whitney (Figure 2B), the levels of the repressive H3K9me2 the at modification VSMCs levels at Mmp3, confirmed abundant H3K9me2 Mmp9, and Mmp12 as well as the positive control locus Magea2 compared with the negative control, Actb (Fig- despite increased mRNA- expres ure 2A). Surprisingly, of sion I in the in I inflammation-responsive MMPs, Mmp3, as such and Biol. 2019;39:2289–2302. DOI:Arterioscler Thromb Vasc 10.1161/ATVBAHA.119.312765 treatment (Figure treatment α IL-1 after reduced not were continued presence of this repressive epigenetic mark at target genes suggests that their activation in response to inflammation could be obstructed. at MMP cultured murine gene promoters, we pretreated VSMCs with UNC, dimeth- an inhibitor of the main H3K9 α stimulation. UNC before IL-1 G9A/GLP, yltransferases treatment significantly reduced H3K9me2 levels glob- ally and at target gene promoters (Figure 3A and 3B). UNC basal MMP did not affect but gene expression, Harman et al et Harman Basic Sciences - VB 2294 compared withUNC0638 normalizedtoβ h)relativetocontrols(–UNC), 48 A α. metalloproteinase) genepromotersandpotentiatestheirupregulation inresponsetoIL-1 (matrix (G9A-like protein)reducesHistoneH3LysineFigure 3.InhibitionofG9A/GLP 9Dimethylation(H3K9me2)atMMP Harman etal sequent to A366 administration,the leftcommoncarotid sequent toA366 Figure 4C), comparedwithvehicle-treated controls.Sub- taining (45%,Figure 4B)andWestern blotting(40%, as revealed byimmunos- H3K9me2 levelsinVSMCs treatment resultedinreproducible reductionofglobal analysis ofMmp2, Ccl2 comparedwithnegativecontrolIgG.*P 4 independentprimarycultures)atcontrolloci(Magea2,positiveandActb , negative)andthepromotersofMmp2 U housekeeping genes(HmbsandYwhaz).*P α of4–5independentprimarycultures)isshownrelativetocellstreatedwithUNC+IL-1 Expression (mean±SEM α (top right),IL-1 digestionincontrol(topleft),UNC quenched (DQ)-gelatin corrected total fluorescence (CTF), relative to untreated cells (mean±SEM of6independentprimarycultures).*P fluorescence(CTF), relativetountreatedcells(mean±SEM corrected total (4′ (green)andDAPI for digestedDQ-gelatin test).B (UNC0638)-treated murine vascular smooth muscle cells (mVSMCs; murinevascularsmoothmusclecells (mVSMCs; (UNC0638)-treated UNC westernblotandquantification ofH3K9me2in , Representative November 2019 , Chromatin immunoprecipitation-qPCR analysis in control and UNC-treated mVSMCs showing H3K9me2 enrichment (mean±SEM of of showingH3K9me2enrichment (mean±SEM mVSMCs analysisincontrolandUNC-treated , Chromatinimmunoprecipitation-qPCR Mmp3, Mmp9, Mmp12, 43, 44 . Two weeksofA366 in control (white bars), UNC (black),α IL-1 Il6, andCcl2incontrol(whitebars),UNC <0.05, n.s.,notsignificant(Kruskal-Wallis). D <0.05 (Kruskal-Wallis). C ,6-diamidino-2-phenylindole)-stained nuclei (blue) are shown. Scale bars, 25 μm. Bar plot shows the plotshowsthe nuclei(blue)areshown.Scalebars,25μm.Bar ,6-diamidino-2-phenylindole)-stained -actin. Graph shows mean±SEM of4independentprimarycultures.*P -actin. Graphshowsmean±SEM 2019;39:2289–2302. DOI: 10.1161/ATVBAHA.119.312765Arterioscler ThrombVasc DOI: Biol. 2019;39:2289–2302. , RT-qPCR polymerase chain reaction) (reverse transcriptionwithquantitative α (lowerleft),andUNC+IL-1 promoter in VSMCs withintissue(Figure 4D, at theIl6promoterinVSMCs trol animals revealed that H3K9me2 was also enriched the protocol(Figure 4D).Interestingly, analysisofcon- were reducedattargetgene promotersattheendof (Figure 4A). We thatH3K9me2 levels confirmedbyChIP artery wasligatedandvessels wereanalyzed7dayslater , Representative imagesandquantification ofdye- , Representative H3K9me2 Regulates the VSMC InflammatoryResponse H3K9me2 RegulatestheVSMC α (gray),andIL-1 -treated mVSMCs (lower right). Signals (lowerright).Signals -treated mVSMCs (red bars) and normalized to (redbars)andnormalizedto <0.05 (Kruskal-Wallis). Mmp3, +UNC-treated (red) mVSMCs. (red)mVSMCs. +UNC-treated <0.05 (Mann-Whitney Mmp9, Mmp12, and Il6, and Basic Sciences - VB 2295 November 2019 <0.05 (linear model, see online-only Data , Representative immunofluorescence staining and H3K9me2 Regulates the VSMC Regulates H3K9me2 Response Inflammatory

/ROSA26-EYFP (enhanced yellow fluorescent protein) mice were administered A366 ERt2 , RepresentativeVSMCs western blot and quantification of H3K9me2 levels in aortic medial of mice administered (Continued ) C ,6-diamidino-2-phenylindole; white) are shown. Scale bars, 50by signal in individual medial cell nuclei, defined μm. Dot plot shows H3K9me2 , Schematic of experimental, Schematic Lineage-labeled Myh11-Cre model. their location between elastic lamina, relative to the average H3K9me2 signal in vehicle controls analyzed in the same batch (batches are indicated with in the sametheir location between elastic lamina, relative to the average H3K9me2 signal in vehicle controls analyzed (batches batch symbols). Mean (line) and SEM (error bars) are indicated. n=4 animals for vehicle, n=3 animals for A366. *P (30 (CCA). mg/kg per d) or vehicle control and subjected to ligation of the left common carotid artery B quantification of H3K9me2 in left (L) cryosections from mice administered vehicle control or A366CCA for 14 d. Signals for H3K9me2 (magenta) and DAPI (4′ Supplement). Vascular smooth muscle cell (VSMC)-specificFigure 4. Vascular upregulation of histone H3 lysine 9 dimethylation (H3K9me2)-target genes after carotid ligation is enhanced by G9A/GLP protein) inhibition. (G9A-like A Biol. 2019;39:2289–2302. DOI:Arterioscler Thromb Vasc 10.1161/ATVBAHA.119.312765 Harman et al et Harman Basic Sciences - VB normalized tohousekeeping genes(Hprt1andHmbs ). *P<0.05(Kruskal-Wallis). Lindicatesleft,and R,rightCCA. of4experiments) areshownrelativetoligated LCCA fromvehiclecontrol(striped, greenbars)and (mean±SEM Data detected inanysample. 2296 Mmp12, Il6,Ccl2,andSca1(Ly6a) inEYFP replicates. *P<0.05(Kruskal-Wallis). E,RT-qPCR polymerase chain reaction) analysisofMmp2,Mmp3 (reversetranscription with quantitative of5to6biological and Actb,negative)Mmp3Mmp12 , Il6,andCcl2promotersisshownalongsidenegative controlIgG.Mean±SEM orvehicle control.H3K9me2enrichment atcontrolloci(Magea2,positive from mice21daysafterinsertionof anosmoticpumpdeliveringA366 analysisoftheaorticmedial layer(adventitialandendothelialcellsremoved) normalized toβ-actin.D,Chromatinimmunoprecipitation-qPCR of3biologicalreplicates)areshown relative tovehiclecontrol, (mean±SEM for 14days.Data Figure 4Continued.vehiclecontrolorA366 was mostpronounced6hoursafterIL-1α stimulationfor thetranscriptand protein level (Figure 5C, D). Thisboth ated theIL-1α responseofH3K9me2targetgenesat global H3K9me2 levels 10-fold (Figure treatmentreduced MMP2 promoters(Figure 5A).UNC repressive epigeneticmarkwasnotpresentatCCL2and whereasthis was alsoaH3K9me2targetinhVSMCs, IL6 Data Supplement).SimilartomurinetissueVSMCs, served inhuman(Figure 5A;IVtheonline-only control locuswithandwithoutIL-1α treatmentwascon- MMP12 genepromoterscomparedwiththeactiveACTB enrichment ofH3K9me2attheMMP3, culturesrevealedthat analysisinprimaryhVSMC ChIP Expression ChangesIsConservedinHumans Gene Inflammation-Associated VSMC H3K9me2-Mediated Regulationof chain reactiononlineage-labeled,EYFP reverse transcription withquantitativepolymerase weperformed induced generegulationinVSMCs, cells (Figure 3B). green bars),contrastingwithobservationsincultured Harman etal VSMCs, sion ofSca1/Ly6a, amarkerofphenotypicallyswitched Interestingly, alsosignificantlyincreased expres- A366 reduced H3K9me2 at their promoters (Figure 4D). (Figure 4E),correlatingwith those treatedwithA366 Mmp3, treated controlmice,injury-inducedexpression of arteries 7daysaftersurgery. Comparedwithvehicle- lated byflowcytometryfromligatedandcontrolcarotid more proinflammatorystate. phenotypicswitching toain vivoandinhibitsVSMC mation-responsive H3K9me2targetgenesinVSMCs suggest thatH3K9me2attenuatesinductionofinflam- online-only DataSupplement).Together, theseresults inthe expression afterligation(Figure 4E;FigureIII injury, gene orthereductionofcontractileVSMC in uninjuredvessels,theresponseofCcl2toligation- did notaffect expression ofH3K9me2targetgenes the in derived cells (Cd45), beyond background (Figure III ( tions didnotexpress markersofadventitialfibroblasts VSMC cellpopula- firmed thatflowcytometry-purified injury comparedwithcontrols(Figure 4E).We con- ), or bone marrow- Pdgfra), endothelialcells (Cd31), or bone To assessthespecificeffect ofH3K9me2oninjury- online-only Data Supplement). Importantly, A366 November 2019 Mmp12, andIl6wassignificantlyincreasedin 35 suggesting an increased VSMC response to suggesting an increased VSMC + carotid VSMCs isolatedbyflowcytometry7daftercarotidligation.Mmp9transcriptswerenot carotidVSMCs 5B) and potenti- + VSMCs iso- VSMCs 2019;39:2289–2302. DOI: 10.1161/ATVBAHA.119.312765Arterioscler ThrombVasc DOI: Biol. 2019;39:2289–2302. MMP9, and inflammatory cytokinepromoters,includingMMP3 andpro- AP-1 transcriptionfactorstoanumberofMMP tion domainnecessaryforgeneinduction. of thep65subunit,which containsthetransactiva- κ mation-induced NF essary toinducetheirexpression. protein stability, Induced SignalTransduction H3K9me2 DoesNotAffectInflammation- G9A/GLP havenonhistonetargets, G9A/GLP In additiontotheirroleashistonemethyltransferases, (Figure 5F). expression ofMMP2ortheresponseCCL2toIL-1α G9Aknockdown didnotaffect thesiRNA-mediated sion (Figure 5F).Consistentwiththeeffect ofUNC, IL-1 did notinfluencebasaltranscriptlevelsbutenhanced transfected cells(Figure 5E).G9Aknockdown alone H3K9me2 levelscomparedwithscrambledsiRNA- tein levels,which leadtoa2.7-fold reductioninglobal knockdown resultedin>7.5-fold reductioninG9Apro- H3K9-methyltransferase complex.G9A/GLP G9A to silence the G9A subunit of the heterodimeric siRNA weused UNC, sion werebecauseofspecificeffects of also observed24hoursafterstimulation(Figure 5C). MMP3, IL6, whenexpression peaked,whereaspotentiationof MMP9 the activation oftheAP-1transcriptionfactor(FigureVA in and IL-1α α binding andaffects afactorthatiscommontoTNF- actsdownstreamofreceptor This suggeststhatUNC affected (FigureVC in the online-onlyDataSupplement). whereasMMP2orCCL2expression wasnot by UNC, MMP3, α TNF- VB intheonline-onlyDataSupplement).SimilartoIL-1 α inhibition(Figure G9A/GLP affected byUNC-mediated modestly increasedinIL-1 α Supplement). ExpressionlevelsoftheIL-1 receptorwas downstream ofIL-1α activityaffects signalingcascades reducing G9A/GLP lular signalingpathways. To regulatesinflam- investigatewhetherG9A/GLP To ensure that the observed changes in gene expres- κ online-only DataSupplement).BindingofNF α -induced 51 -mediated induction of H3K9me2 target genes -mediated induction of H3K9me2 target genes transcript levels by UNC was MMP9, andMMP12transcriptlevelsbyUNC MMP9, , MMP12, κ signalingsuch asNF H3K9me2 Regulates the VSMC InflammatoryResponse H3K9me2 RegulatestheVSMC MMP3, MMP9,MMP12andIL6expres- 46 was also potentiated MMP12, andIL6wasalsopotentiated protein-proteininteractions, 51 andIL6 (FigureVA intheonline-onlyData B activity, weanalyzedregulation 45,49 -treated hVSMCs but was not butwasnot -treated hVSMCs 53 We thereforetestedwhether , hasbeenreportedasnec- B andMAPK-mediated 45 which mayinfluence 54 7 8 4 47, Compared andcel- B and B and 50-52 , ,

Basic Sciences - VB 2297 November 2019 H3K9me2 Regulates the VSMC Regulates H3K9me2 Response Inflammatory

Histone H3 lysine 9 dimethylation (H3K9me2)-mediated regulation of inflammation-associated genes is conserved inFigure 5. Histone H3 lysine 9 dimethylation (H3K9me2)-mediated regulation of inflammation-associated human aortic VSMCs (hVSMCs). A, Chromatin immunoprecipitation-qPCR analysis in cultured hVSMCs, for H3K9me2 relative to MAGEA2 (positive showing enrichment Continuedcontrol) at ACTB (negative control) and MMP2, MMP3, MMP9, IL6, and CCL2 promoters compared with signals for ( ) Biol. 2019;39:2289–2302. DOI:Arterioscler Thromb Vasc 10.1161/ATVBAHA.119.312765 Harman et al et Harman Basic Sciences - VB GLP (Figure 6B). GLP target (Figure 5A),wasunaffected byinhibitionofG9A/ enrichment ofp65atCCL2,which isnot aH3K9me2 NF suggesting thatH3K9me2attheIL6locusprevents cellscorrelatedwithreducedH3K9me2, UNC-treated pretreatment (Figure 6B).Increasedp65bindingin UNC stimulation, andthiswassignificantlyenhancedby At theIL6locus,p65bindingwasincreasedbyIL-1 α the the Comparedwith treated with IL-1hVSMCs α andUNC. (AP-1)in κB) andcJUN forp65(NF performed ChIP transcription factorbindingattargetgenepromoters,we 2298 Figure 5Continued.negativecontrolIgG ure 6A). scriptional startsiteofMMP3, elements havebeenreportedwithin2kbofthetran- κB and AP-1 and IL-1-responsiveBinding sites for NF Promoters to Inflammation-Responsive Target Gene κB andAP-1Binding H3K9me2 AttenuatesNF activation. influence MAPK inhibitiondoesnot ment), suggestingthatG9A/GLP intheonline-onlyDataSupple - stimulation (FigureVII treatmentbeforeIL-1α wereaffected byUNC (p38) kinase (extracellular regulatingkinase)1/2,andp38 N-terminalkinase),ERK (cJUN of phosphorylatedJNK Western blotting showed that neither timing nor levels phosphorylation of AP-1 subunits, such as cJUN. atspecific residues,whichMAPKs ultimatelyleadsto signaling. κ is not because of increased upstream NF by UNC to IL-1α basal conditions nor its nuclear translocation in response ther affected thenuclear-to-cytoplasmicratioofp65at α treatedwithUNC±IL-1 hVSMCs the with untreatedcontrolcells,p65levels(FigureVIAin Harman etal actin. of3experiments)areshown relativetountreatedcellsandnormalizedβ (mean±SEM Data culturedhVSMCs. (+UNC) and UNC-treated enrichment comparedwithACTB). tion of IL-1α Data Supplement).These resultssuggestthatpotentia- (mean±SEM of4experiments).*P (mean±SEM andnormalizedtoβ are shownrelativetomock-transfected hVSMCs Data transfectedhVSMCs. andG9A siRNA scrambled (Scr)siRNA, normalized toβ α IL-1 UNC, of4to10experiments.*P mean±SEM α arerelativetocellstreated withUNC+IL-1 white bars).Data α andafterIL-1 control hVSMCs α treatedwithG9AsiRNA+IL-1 relative tohVSMCs α withandwithoutIL-1 hVSMCs andG9A siRNA-transfected CCL2 inmock, ScrsiRNA α compared withG9A siRNA+IL-1 Inflammatory cytokines induce phosphorylation of κB bindingtoitscognatesequence.Incontrast,the ACTB negativecontrollocus,p65wasenriched at online-only DataSupplement)wereunchanged in IL6 andCCL2genepromotersinuntreatedcells. C , RT-qPCR polymerasechain reaction) analysisof (reverse transcriptionwithquantitative November 2019 α orTNF- 50–53,58 α , andUNC+IL-1 -actin. *P α /TNF- To investigatewhetherH3K9me2affects stimulation(FigureVIintheonline-only <0.05 (Kruskal-Wallis). E -induced α of3experimentsisshownrelative tocellstreatedwithUNC+IL-1 Mean±SEM -treated culturedhVSMCs. treatment for 10, 30 min, 1.5, 3, 6, or 24 h, without (−UNC, black bars) and with prior UNC treatment (+UNC, treatment(+UNC, black bars) andwithpriorUNC min,1.5,3,6,or 24 h,without(−UNC, treatmentfor10,30 ). Each experiment used hVSMC isolatesderivedfromdifferent individuals. ). Each experimentusedhVSMC <0.05 (Kruskal-Wallis, compared withScr).F B [UNC0638]) [UNC0638]) westernblotanddensitometric analysisofH3K9me2levelsincontrol(−UNC , Representative IL6 and <0.05 (Kruskal-Wallis). D IL6, andCCL2(Fig- . UNC treatmentnei- . UNC (immunoglobulin). MMP expression , Representative westernblotanddensitometricanalysisofG9AH3K9me2 levelsinmock, , Representative andnormalizedto housekeepinggenes(HPRT1and YWHAZ). *P 2019;39:2289–2302. DOI: 10.1161/ATVBAHA.119.312765Arterioscler ThrombVasc DOI: Biol. 2019;39:2289–2302. ). Graph show for6handnormalizedtohousekeepinggenes(HMBSYWHAZ).Graphshow Graph show mean±SEM of6experiments.*P Graph showmean±SEM , Representative western blot and densitomeric analysis of MMP3 in control, incontrol, westernblotanddensitomericanalysisofMMP3 , Representative 55–57 B

IL6, andCCL2areregulatedviadifferent mechanisms. strates thatinflammation-induced expression ofMMP3, transcriptionfactorbindinganddemon- AP-1 (cJUN) IL-1 acts directlyatthe κB targetIL6. activation oftheNF IL-1 phosphorylationisnecessaryfor cJUN JNK-mediated inhibition(Figure 6D).These resultssuggestthat JNK induced ing inanattenuatedinductionbyIL-1 α, whereasIL-1α- cells selectivelyaffected MMP3geneexpression result - inSP600125-treated ment). Reducedlevelsofp-cJUN intheonline-onlyDataSupple- (FigureVII or totalcJUN intheonline-onlyDataSupplement ) ure 6C; FigureVII not affecting (Fig- thelevelsofphosphorylatedMAPKs stimulatedwithIL-1 inhVSMCs α, despite in p-cJUN inhibitioncausedamodest(2-fold)increase G9A/GLP levels (3.9-foldto4.7-fold; Figure 6C).Interestingly, IL-1 significantlyreduced α-induced p-cJUN SP600125 of JNK (SP600125 of JNK the sensitivityofIL-1α-induced transcriptiontoinhibition factor bindingattheMMP3andIL6loci,wemeasured the CCL2promoterunderanycondition(Figure 6B). bindingwasobservedat depletion (Figure 6B).NocJUN H3K9me2 UNC-mediated was significantlyincreasedby enrichment promoters onIL-1α stimulation,andcJUN IL6andMMP3 wasdetectedatboth AP-1 subunitcJUN of cellstimulation(Figure 6B).Incontrast,bindingthe MMP3 promoter, reported IL-1-responsive element intheproximal a subsetofinflammation-responsivegenes,including phenotype associatedwithvasculardiseaseandidentify of H3K9me2 in regulating the proinflammatory VSMC We here demonstrate an evolutionary conserved role DISCUSSION for cJUN phosphorylationatSer63. for cJUN upstream ofthetranscriptionalstartsite, Taken together, thesefindings suggestthatH3K9me2 To explore theimpactofdifferences intranscription Notably, p65enrichment wasnotdetectedatthe , RT-qPCR analysisofMMP2, κB (p65) and α-induced upregulation by inhibiting NF α-induced expression ofMMP3butnotrequiredfor IL6 andCCL2expression wasnotaffected by stimulation. Data (mean±SEM of 4 experiments) are of4experiments)are (mean±SEM stimulation.Data H3K9me2 Regulates the VSMC InflammatoryResponse H3K9me2 RegulatestheVSMC 51,52 MMP2, MMP3 and κB site 1.5 kb nor at a reported NF 59 ), the primary MAPK responsible ), theprimaryMAPK MMP3, <0.05 (Kruskal-Wallis, H3K9me2 MMP3, MMP9, IL6 promoter to restrain <0.05 (Kruskal-Wallis, MMP9, MMP12, 60 Treatment with MMP12, 50 IL6, and irrespective and IL6, and and and in CCL2 in -actin -actin - Basic Sciences - VB 2299 November 2019 <0.05 (Kruskal-Wallis). promoters. Red lines indicate chromatin promoters. Red lines indicate chromatin H3K9me2 Regulates the VSMC Regulates H3K9me2 Response Inflammatory 49 −SP600125 and normalized to the average of 2 housekeeping -treated hVSMCs, without and with prior UNC and/or SP600125 (reverse transcription with quantitative polymerase chain (reverse transcription with quantitative polymerase chain , RT-qPCR -treated human aortic VSMCs (hVSMCs; levels at the red) showing enrichment and CCL2 (Sutcliffe et al) 44 , ChIP-qPCR analysis of p65 and cJUN binding relative to input in untreated control (white -TUBULIN levels. D negative control locus compared with negative control IL6, and CCL2 promoters and the ACTB negative control locus compared with negative control IL6 (Gomard et al), (gray), and UNC+IL-1α

IL6, and CCL2 in control and IL-1α 41–43 MMP3, , without and with prior treatment with UNC and SP600125 for 48 h. Data (mean±SEM of 4 experiments) are relative , Representative western blot and densitometric analysis p-cJUN control hVSMCs (Ser63) in untreated and cells -SP600125 and normalized to β , Schematic indicating positions of reported p65 and cJUN indicating positions of reported , Schematic binding sites and an IL-1-responsive element at the MMP3 (Borghaei Figure 6. IL (interleukin)-1α-induced binding of p65 and cJUN at histone H3 lysine 9 dimethylation (H3K9me2)-target genes is enhanced by G9A/GLP protein) inhibition. (G9A-like A et al, Clark et al, Quinones et al), treatment. Data (mean±SEM of 4 experiments) are relative to +UNC+IL-1α experiment used hVSMCsgenes (HPRT1 and YWHAZ). Each derived from different individuals; *P immunoprecipitation (ChIP)-qPCR amplicons. B bars), UNC (UNC0638; IL-1α black), reported binding sites within the MMP3, at ACTB. Graph shows mean±SEMIgG (immunoglobulin). Background levels of p65 binding, indicated by a line, are based on enrichment of 4 experiments. C stimulated with IL-1α to +UNC+IL-1α reaction) analysis of MMP2, Biol. 2019;39:2289–2302. DOI:Arterioscler Thromb Vasc 10.1161/ATVBAHA.119.312765 Harman et al et Harman Basic Sciences - VB induced byhighinsulinlevels. is compared withnondiabeticcontrols,andKDM3A betic patientsalsodisplayreducedlevelsofH3K9me2 H3K9me2 levels plaquecellshavelower vation thatαSMA-positive supportstheobser- analysis oflineagetracedVSMCs other celltypes. ers andtheirexpression hasalsobeenreportedin pancy atMMPandotherinflammatorygenepromot- ated vascularcomplications. tion ofH3K9me2mightcontributetodiabetes-associ- 2300 type controls. virus infection andsuccumbfastertoinfection than wild- RNA Moreover, G9A-deficient fliesarehypersensitiveto patients comparedwithhealthyhumanarteries. H3K9me2 reductioninatheroscleroticplaquesfrom ing aftervascularinjuryisconsistentwiththereported sclerotic lesionsandinarteriesundergoingremodel- tic targetinginCVD. mightthereforebeastrategyfortherapeu- (eg, KDM3A) activity orinhibitingH3K9me2-specificdemethylases Increasing H3K9me2levelsbystimulatingG9A/GLP inflammatoryactivationandvascularremodeling. VSMC in atherosclerosisandafterinjuryisrequiredtoelicit ulate thatthereducedH3K9me2levelsweobserved Itisthereforetemptingtospec- expression inVSMCs. gene restrain inflammation-induced, CVD-associated H3K9me2 suggeststhatthisepigeneticmarkactsto invitroandvivo.This functionalroleof stimuli both genes inresponsetoinflammatory inflammatory VSMC factors andincreasesexpression oftheseMMPandpro- increases binding of inflammation-induced transcription Wetargets in VSMCs. show that loss of H3K9me2 MMP3, Harman etal poly(I:C)-induced upregulation ofinflammatorygenes. reduction ofH3K9me2levels infibroblastspotentiates embryonic fibroblasts. ulates theexpression ofinflammatorygenesinmouse stim- ablation orpharmacologicalinhibitionofG9A/GLP H3K9me2-specific demethylase. a levels werereducedbyoverexpression ofKDM3A, previously detectedindiabeticratswhenH3K9me2 erated neointimaformationfollowingvascularinjury mightalsoexplain theaccel- induced genesinVSMCs the lesioncore,thatdonotexpress contractilemarkers. cellswithin etic markisalsoreducedinVSMC-derived compared withcontrols. expression in retinal endothelial cells from diabetic rats at theMmp9promotercorrelateswithelevated mRNA levels. with enhancedMMP3mRNA moter inprimaryhumandermalfibroblasts,correlating radiation induceslossofH3K9me2attheMMP3pro- An inversecorrelationbetweenH3K9me2occu- The from athero- global loss of H3K9me2 in VSMCs H3K9me2-mediated regulationofinflammation- November 2019 MMP9, 28 These observationssuggest thatG9A/ MMP12, andIL6,asdirectH3K9me2 24–27 61 and further indicates that this epigen- For example, H3K9me2depletion 24 Similar to our findings in VSMCs, Similartoourfindingsin VSMCs, 27 Similarly, solar-stimulatedUV 62,64 62,63 Therefore, dysregula- 25 VSMCs fromdia- VSMCs Inaddition,genetic 2019;39:2289–2302. DOI: 10.1161/ATVBAHA.119.312765Arterioscler ThrombVasc DOI: Biol. 2019;39:2289–2302. 22 Our 24

phatase. phos- might regulateanas-yet-unidentifiedcJUN to chromatin. canphosphorylatesubstratesoncebound MAPKs andERK1/2. JNK, tion andphosphorylationofp38, activation ofupstreamsignaling,includingp65activa- didnotaffect IL-1tion ofG9A/GLP α inhibition (seeGraphicalAbstract).Incontrast,inhibi- reduction ofH3K9me2levelsfollowingG9A/GLP NF response—in avarietyofcelltypes. tight anddynamicregulation—such astheinflammatory nism to control genes involved in processes that require representsanevolutionarilyconservedmecha- GLP the CambridgeNational InstituteforHealthResearch Biomedical Research Cen- core (Wellcome Trust fortechnical assistance, MajorAward [208363/Z/17/Z]) Imaging Research Laboratories, Science,Metabolic Council, InstituteofMetabolic seq dataanalysis,GregoryStrachan andthe Wellcome Trust-Medical Research statistical analysis,MikhailSpivakovforadvise onchromatin immunoprecipitation- We thanktheCambridge Statistics Clinic(DepartmentofStatistics)foradviseon Acknowledgments Research Institute,UniversityofEdinburgh,UnitedKingdom. Kingdom. RalitsaR.Madsen,CentreforCardiovascularScience,Queen’sMedical ofMolecularBiology, Laboratory Cambridge,United Cell BiologyDivision,MRC tal Sciences,UniversityofBirmingham,UnitedKingdom.Paula Freire-Pritchett: Amanda Dalby:InstituteofCardiovascularSciences,CollegeMedicalandDen- Current Addresses Mathematical Sciences,UniversityofCambridge,UnitedKingdom(B.G.S.). Cambridge, UnitedKingdom(L.D.,P.F.-P.); andStatisticalLaboratory, Centrefor R.R.M., E.L.,A.U.,M.R.B.,H.F.J.); BabrahamInstitute,Research Campus, sity ofCambridge,UnitedKingdom(J.L.H.,L.D.,J.C.,A.D.,K.F., A.F., A.L.T., M.D.W., From theDivisionofCardiovascularMedicine,DepartmentUniver- Affiliations Received April9,2019;acceptedAugust7 ARTICLE INFORMATION potentiates transactivation. which lation efficiencyof chromatin-associated cJUN, inhibition couldbebecauseofincreasedphosphory- in IL-1 α in p-cJUN more specificdrugsforfutureclinicaluse. tion-induced genescouldbeleveragedfordeveloping tional regulationofselectiveMMPsandotherinflamma- genes, includingMMPsandIL6.Differential transcrip- proinflammatory spurious induction of CVD-associated response. We propose that H3K9me2 acts to prevent inflammatory tional epigeneticregulatoroftheVSMC chromatin regulation. specific transcriptionfactorusageandlocus-specific a context-specific manner,both which coulddependon contrast, H3K9me2targetgeneinductionisdelayedin levels resultinginrapidinductionuponstimulation.In as Non-H3K9me2targetssuchmechanisms inVSMCs. inflammation-induced geneshavedistinctregulatory We demonstrateincreasedbindingofAP-1and In conclusion,ourdataimplicateH3K9me2asafunc- κ CCL2 displaytranscriptionfactorbindingatbasal B specificallyatH3K9me2targetgeneson 66 Regardless,ourdatademonstratethatkey H3K9me2 Regulates the VSMC InflammatoryResponse H3K9me2 RegulatestheVSMC 65 We proposethatthe2-foldincrease -stimulated cells after G9A/GLP -stimulated cells after G9A/GLP 66,67 , 2019. Alternatively, G9A/GLP α /TNF- -induced Basic Sciences - VB

- 2301 November 2019 . 2010;143:134–144. doi: 2010;143:134–144. Cell. H3K9me2 Regulates the VSMC Regulates H3K9me2 Response Inflammatory Van Rij RP. The epigenetic regulator G9a mediates tolerance to RNA The Rij RP. Van Drosophila. PLoS Pathog. 2015;11:e1004692. in virus infection doi: 10.1371/journal.ppat.1004692 Z, Singh D, Raj JU. Q, Lu pro- cell BIX-01294 treatment blocks Yang migration and contractility in ovine foetal pulmonary arte- liferation, Prolif. 2012;45:335–344. doi: 10.1111/j. rial smooth muscle cells. Cell 1365-2184.2012.00828.x Clifford RL, AJ. Abnormal histone methylation CE, John AE, Brightling Knox secre- 165a factor growth endothelial vascular increased for responsible is tion from airway smooth muscle cells in asthma. J Immunol. 2012;189:819– 831. doi: 10.4049/jimmunol.1103641 N, Gorbey S, B, Wettschureck M, Leutgeb A, Benyó Z, Lukasova Wirth Horváth B, Maser-Gluth C, Greiner E, et al. G12-G13-LARG-mediOrsy P, - ated signaling in vascular smooth muscle is required for salt-induced hyper- tension. Nat Med. 2008;14:64–68.10.1038/nm1666 doi: van Es JH, van den Born M, Sato T, Snippert HJ, van der Flier LG, C, Barker N, Klein AM, van Rheenen J, Simons BD, et Kroon-Veenboer al. Intestinal crypt homeostasis results from neutral competition between cells. stem Lgr5 dividing symmetrically 10.1016/j.cell.2010.09.016 Owens GK. Detection of histone Nguyen AT, Gomez D, Shankman LS, modifications at specific gene loci in single cells in histological sections. Nat doi: 10.1038/nmeth.2332 Methods. 2013;10:171–177. K, H, Foote Chappell J, Harman JL, Narasimhan VM, Yu Simons BD, of a sub- Extensive proliferation Bennett MR, Jørgensen HF. yet plastic, medial vascular smooth muscle cells set of differentiated, contributes to neointimal formation in mouse injury and atherosclerosis Kurozumi A, Nakano K, Yamagata K, Okada Y, Nakayamada S, Tanaka Y. Y. S, Tanaka Nakayamada K, Okada Y, K, Yamagata A, Nakano Kurozumi histone vascular calcification via is critically involved in Il-6/stat3 pathway Bone vascular smooth muscle cells. of the runx2 promoter in modification Abstracts. 2016;5:425 S-F, Yet Lin C-S, Y-F, Wang Y-T, Tsai C-S, Lin C-Y, Tsai J-Y, G-L, Wu Lee C-C.Kuo vascular smooth regulates axis mapk-il-6 Hsu Y-J, Tlr4-activated Int JMo Sci. 2016;17:1394muscle cell function. T, Nakagawa JH,Li N, Jing A, Kahn HY, Lan W, Mu S, Watanabe activa- pathway in IL-6-induced of JAK/STAT Ohashi R, Johnson RJ. Role cells. Am J Nephrol. 2004;24:387–392.tion of vascular smooth muscle doi: 10.1159/000079706 Ginnetti F, G, Caligiuri G, Rebuzzi AG, Liuzzo G, Fantuzzi Biasucci LM, (IL)-1Rainterleukin of levels Increasing A. Maseri CA, Dinarello IL-6 and - asso are angina unstable in hospitalization of days 2 first the during of in-hospital coronary events. Circulation. ciated with increased risk doi: 10.1161/01.cir.99.16.2079 1999;99:2079–2084. JC. Kaski Inflammatory E, Economou EK, D, Oikonomou F, Crea Tousoulis Eur Heart J. current therapeutic approaches. cytokines in atherosclerosis: doi: 10.1093/eurheartj/ehv759 2016;37:1723–1732. Klassen LW, Mikuls TR, Duryee MJ, Garvin RP, JT, Anderson DR, Poterucha GM. artery dis- and its receptors in coronary Thiele IL-6 Shurmur SW, ease and acute myocardial infarction. Cytokine. 2013;62:395–400. doi: 10.1016/j.cyto.2013.03.020 Rastegar M, Davie JR. Epigenetic control. J Cell Physiol. Delcuve GP, 2009;219:243–250. doi: 10.1002/jcp.21678 M, E, Gebhard H, Schmitt Greißel A, Culmes M, Napieralski R, Wagner J. Alternation of histone HH,Zimmermann A, Eckstein A, Pelisek Zernecke and DNA methylation in human atherosclerotic carotid plaques. Thromb Haemost. 2015;114:390–402. doi: 10.1160/TH14-10-0852 mol- related the and G9a methyltransferase H3K9 M. Tachibana Y, Shinkai doi: 10.1101/gad.2027411 Genes Dev. 2011;25:781–788. ecule GLP. I, Stienen A, Dewell S, Chen MS, U, Mecklenbrauker Schaefer TC, Fang RK, Prinjha H3 lysine 9 Chandwani R, et al. Histone V, Rioja I, Parravicini response. J Exp di-methylation as an epigenetic signature of the interferon Med. 2012;209:661–669. doi: 10.1084/jem.20112343 Solar-simulatedCosta M. radia- ultraviolet MS,Matsui L, Gesumaria T, Kluz in the gene promoters of matrix tion induces histone 3 methylation changes fibroblasts. Exp Der metalloproteinases 1 and 3 in primary human dermal matol. 2015;24:384–385. doi: 10.1111/exd.12675 Zhou L, Jin J, Chen X. Gunawardena HP, JA, X, Wrobel Wei Liu F, Y, Liu C, Yu - reveals a transcrip approach chemoproteomic activity-based A chromatin . 2014;5:5733. tional repressome for gene-specific silencing. Nat Commun doi: 10.1038/ncomms6733 RA. by epigen- metalloproteinase-9 Kowluru of matrix Zhong Q, Regulation Diabetes. retinopathy. etic modifications and the development of diabetic 2013;62:2559–2568. doi: 10.2337/db12-1141 JM, A, Kramer GJ, Schenck Overheul SH,Merkling AW, Bronkhorst

17. 27. 29. 31. 32. 33. 34. 15. 16. 18. 19. 20. 21. 22. 23. 24. 25. 26. 28. 30.

Nucl USA. Sci Dove Medical Press Natl Acad Proc . 2016;118:692–702. doi: 10.1161/CIRCRESAHA.Circ Res. 2016;118:692–702. doi: 10.1161/01.ATV. Biol. 2007;27:571–577. Arterioscler Thromb Vasc Bennett MR, smooth muscle cells in ath- Sinha S, Owens GK. Vascular erosclerosis. 115.306361 smooth muscle cell in the inflammation of ath- S. Role of vascular Lim S, Park doi: 10.5483/bmbrep.2014.47.1.285 erosclerosis. BMB Rep. 2014;47:1–7. Orekhov AN. VA, Myasoedova AV, AA, Grechko Melnichenko Chistiakov DA, anti-inflammatory agents for treatment of atherosclerosis. Exp of Potential Mol Pathol. 2018;104:114–124. doi: 10.1016/j.yexmp.2018.01.008 JG, MacFadyen Chang WH, Ballantyne C, Ridker PM, Everett BM, T, Thuren Anti- Group. Trial Anker SD, et al; CANTOS W, J, Koenig Nicolau F, Fonseca disease. N Engl inflammatory therapy with canakinumab for atherosclerotic doi: 10.1056/NEJMoa1707914 J Med. 2017;377:1119–1131. Mahan S, Gomez D, Baylis RA, Durgin BG,GF, Alencar Newman AAC, SE, ath- has Interleukin-1β al. et Francis A, Waisman W, Müller C, Hilaire St atherosclerotic lesions of mice. Nat Med. in advanced eroprotective effects 2018;24:1418–1429. doi: 10.1038/s41591-018-0124-5 Johnson JL. Matrix metalloproteinases and their inhibitors in cardiovascular potential. clinical and knowledge current pathologies: Ltd. 2014;2014:21–36. ZS,Galis remodeling vascular in metalloproteinases Matrix JJ. Khatri Circ Res. 2002;90: and atherogenesis: the good, the bad, and the ugly. 251–262. Johnson JL,CL. George SJ, Jackson effects Divergent Newby AC, 9, and 12 on atherosclerotic plaque of matrix metalloproteinases 3, 7, arteries. brachiocephalic mouse in stability 2005;102:15575–15580. Morimoto M, Hatakeyama K, Jin Y, T, Kitajima S, Koike Y, Liang J, Liu E, Yu accelerates the pro- et al. Macrophage metalloelastase T, Watanabe Asada Y, gression of atherosclerosis in transgenic rabbits. Circulation. 2006;113:1993– 2001. doi: 10.1161/CIRCULATIONAHA.105.596031 Johnson JL, Dwivedi A, Somerville M, George SJ, Matrix metal- Newby AC. (MMP)-3loproteinase MMP-9 activates muscle vascular smooth mediated cell migration and neointima formation in mice. Arterioscler Thromb Vasc Biol. 2011;31:e35–e44. doi: 10.1161/ATVBAHA.111.225623 TN, Shapiro SD, Halpert I, Sires UI, S, Wight Roby JD, Potter-Perigo by lipid-laden WC. Matrilysin is expressed SA, Parks HG, Wickline Welgus macrophages at sites of potential rupture in atherosclerotic lesions and for the deposition, a proteoglycan substrate localizes to areas of versican enzyme. Proc Natl Acad SciUSA. 1996;93:9748–9753 Johnson JL, G, Sala-Newby GB, Pasterkamp AC, Thomas Ismail Y, Genomics of foam cells and nonfoamy macrophages from rab- Newby AC. produc- oxide nitric of regulator differential as a arginase-I identifies bits tion. 0000256470.23842.94 Rancic Z, SD, K, Selivanova SV, Müller A, Krämer Meletta R, Beck levels of et al. Gene expression T, B, Meding J, Stellfeld Vos PA, Kaufmann matrix metalloproteinases in human atherosclerotic plaques and evaluation vulnerability. plaque for agents imaging as inhibitors radiolabeled of Med Biol. 2014;41:562–569. doi: 10.1016/j.nucmedbio.2014.04.085 of matrix metallo- Increased expression Libby P. Galis ZS, GK, Sukhova Lark MW, proteinases and matrix degrading activity in vulnerable regions of human atherosclerotic plaques. J Clin Invest. 1994;94:2493–2503. doi: 10.1172/JCI117619 7. 2. 3. 4. 5. 6. 1. 8. 9. 14. 10. 11. 12. 13.

REFERENCES Disclosures None. Sources of Funding Sources (BHF) Heart Foundation was funded by a British J.L. Harman studentship (FS/15/38/31516); was funded by a BBSRC L. Dobnikar studentship; DTP funded by the BHFJ. Chappell was - Excel Research Centre for Cardiovascular lence (RE/13/6/30180);by a studentship from the Can- B.G. Stokell was funded was funded by Mathematics of Information. K. Foote tab Capital Institute for the BHFthe PG/16/63/32307; BHFthe by funded were Uryga A. and Finigan A. was funded by BHFRG/13/14/30314. studentship (FS/15/62/32032). A.L. Taylor supported by the BHF Jørgensen were (CH/2000003),M.R. Bennett and H.F. the Ox- bridge BHF (RM/13/3/30159, Centre for Regenerative Medicine RM/17/2/33380), the BHF Excellence (RE/13/6/30180), Centre for Cardiovascular Research and the Center. Research Cambridge Biomedical National Institute for Health Research tre Cell Phenotyping Hub for cell sorting and all members of the Jørgensen group and all members of the Jørgensen Hub for cell sorting tre Cell Phenotyping for helpful discussions. Biol. 2019;39:2289–2302. DOI:Arterioscler Thromb Vasc 10.1161/ATVBAHA.119.312765

Harman et al et Harman Basic Sciences - VB

2302 Harman etal 40. 39. 38. 36. 49. 48. 46. 35. 41. 45. 44. 43. 42. 50. 37. 47.

116.309799 models. 10.1161/CIRCULATIONAHA.117.033249 doi: cells promotesatherosclerosis.Circulation.2018;138:1446–1462. excision repair of oxidativedamage in vascular smooth muscle DNA Shah A,GrayK,FiggN,FiniganStarksL,BennettM.Defective base nar/gku1026 Nucleic AcidsRes.2015;43(Databaseissue):D743–D750. doi:10.1093/ 2015: genomic,phenotypicandenvironmentalvariationsdisease. Petri V, Tutaj Smith JR, M, Wang et al. The SJ, ratgenome database Shimoyama M,DePons J,HaymanGT, LiuW, LaulederkindSJ, NigamR, Methods. 2012;9:357–359. doi:10.1038/nmeth.1923 2.Nat Fast gapped-readalignmentwithbowtie Langmead B,SalzbergSL. ej.17.1.200 sequencing reads.EMBnet.journal[Online].2011;17:1.doi:10.14806/ Martin M.Cutadaptremovesadaptersequencesfromhigh-throughput doi:10.1161/ATVBAHA.119.3124492019;39:603–612. Promoters tostudyvascularsmoothmuscle.ArteriosclerThrombVasc Biol. MartinKA. CarraoAC, GreifDM, Krause R,SaddoukFZ, DS, Chakraborty s41467-018-06891-x from healthymousevessels.Nat Commun. 2018;9:4567. doi:10.1038/ transcriptional signaturesidentifiedinindividualsmoothmusclecells SpivakovM,JørgensenHF.Dzierzak E,BennettMR, Disease-relevant Dobnikar L,Taylor AL,ChappellJ,Oldach P, OertonE, HarmanJL, bility. PLoSOne.2015;10:e0141118.doi:10.1371/journal.pone.0141118 sta function oflysinemethyltransferase G9aintheregulationofsox2 protein Lee JY, Ko KimKS, JJ,Park KimC,DK, HeoSH, Lee KS. SH, Novel 10.1038/nrm3915 lular signallingandfunction.NatRevMolCellBiol.2015;16:5–17. doi: Non-histone protein methylation as a regulator of cel- Li SS. Biggar KK, doi:10.1021/ml400496h 2014;5:205–209. tive inhibitorsofhistonemethyltransferase g9a.ACSMedChemLett . Tse C,Vedadi M, etal.Discovery and developmentof potent andselec- Sweis RF, Pliushchev M,BrownPJ, GuoJ,LiF, MaagD,Petros NB, AM, Soni doi:10.1172/JCI88353 Invest. 2017;127:335–348. protectagainstpathologicalcardiachypertrophy. JClin ferases EHMT1/2 Brien P, AlkassK, Tomasso A,Agrawaletal. The H3K9 dimethyltrans- OkkenhaugH,Loche E,Ferrini RobinsonEL, Thienpont A, B,AronsenJM, 2013;33:1564–1573. doi:10.1038/jcbfm.2013.109 implications forcerebralaneurysmpathology. J Cereb Blood Flow Metab. induces phenotypicmodulationincerebralvascularsmoothmusclecells: Koch WJ, OwensGK, α DumontAS.Rosenwasser RH, TNF- GreigNH, GonzalezLF, Tjoumakaris SI, PM, Jabbour StarkeRM, Ali MS, 10.1152/physiolgenomics.00095.2016 doi: ing carotidarteryligation.PhysiolGenomics.2017;49:115–126. Inflammation and vascular smooth muscle cell dedifferentiation follow- Herring BP,JN, Gallagher SL, McClintick HoggattAM,Griffith PJ. 10.1074/jbc.M802132200 enhancer-binding protein-beta. Leutz A. G9a-mediated lysine methylation alters the function of CCAAT/ Pless O,Kowenz-Leutz E,Knoblich M,LausenJ,BeyermannWalsh MJ, methylation. MolCell.2010;39:71–85. doi:10.1016/j.molcel.2010.06.008 etal.Negativeregulationofhypoxic responses viainducedreptin Seo SB, KimJY, KimY,Lee JS, ShinHJ,KimJH, KimB,ChoiHJ,Lee JM, KimIS, differentiation. ProcNatlAcadSciUSA.2012;109:841–846 g9a methylatesthe transcription factor myod and regulates skeletal muscle Gopinadhan S,SartorelliV, Walsh MJ,Taneja R.Lysine methyltransferase Ling BMT, BharathyN,ChungT-K, Kok WK,LiS,Tan YH,RaoVK, . 2004;316:182–188. doi:10.1016/j.bbrc.2004.02.030 Res Commun.2004;316:182–188. promoter. Biochem Biophys polymorphic repressorelementintheMMP-3 Jr, Rawlings PL Borghaei RC, Javadi M, Woloshin binds to a J. NF-kappaB November 2019 Circ Res.2016;119:1313–1323.doi:10.1161/CIRCRESAHA. . 2008;283:26357–26363. doi: J BiolChem.2008;283:26357–26363. 2019;39:2289–2302. DOI: 10.1161/ATVBAHA.119.312765Arterioscler ThrombVasc DOI: Biol. 2019;39:2289–2302. -

60. 66. 55. 64. 63. 62. 61. 59. 58. 56. 54. 53. 52. 51. 65. 67. 57. tory phenotype of vascular smooth muscle cells in diabetes. memoryandinflamma- genetic histoneh3lysine9methylationinmetabolic Villeneuve LM, ReddyMA,LantingLL, Wang M,MengL, NatarajanR.Epi- 1274. doi:10.3892/ijmm.2017.3351 ‑κB pathways.IntJMolMed.2018;41:1265– by suppressingMAPK/NF ates highconcentrationinsulin‑inducedvascularsmoothmusclecellinjury Zhang BF, JiangH,ChenJ,GuoX,HuQ,Yang inhibitionattenu- S.KDM3A 2017;257:152–163. doi:10.1016/j.atherosclerosis.2016.12.007 controls vascularneointimalhyperplasiaindiabeticrats.Atherosclerosis. anovelregulatorofvascularsmoothmusclecells, demethylase KDM3a, Chen J, Zhang J, Yang J,Xu L, HuQ, Xu C, Yang S,Jiang H. Histone 2016;25:79–86. doi:10.1016/j.carpath.2015.11.001 severity ofatherosclerosisinhumancarotidplaques.CardiovascPathol. Pelisek J.Histoneacetylationand methylationsignificantly change with Zernecke A, Greißel A,CulmesM,BurgkartR,ZimmermannEckstein HH, doi:10.1093/emboj/cdg388 3886. phosphorylation ofthetranscriptionfactorc-Jun.EMBOJ.2003;22:3876– Morton S,DavisRJ,McLarenA,CohenP. Areinvestigationofthemultisite 2001;108:73–81. doi:10.1172/JCI12466 expression andjointdestructionininflammatoryarthritis.J Clin Invest . c-Jun N-terminal kinase is required for metalloproteinase Firestein GS. Chang L, Bennett B, KarinHan Z, Boyle DL, M, Yang L, Manning AM, 2009.00143.x AP-1. and thelin-1 inhumanairwaysmoothmusclecellsinvolvesNF-kappaB scriptional regulationofmonocytechemotactic protein-1releasebyendo- BradburyDA,Sutcliffe CorbettLM, AM,ClarkeDL, Patel JA, Knox AJ.Tran- gene.2012.10.033 and transcriptionalregulation.Gene.2013;513:1–13.doi:10.1016/j. Yang Sharrocks AD,Whitmarsh SH, AJ.MAPkinasesignallingcascades 10.3109/10799893.2015.1030412 and apoptosis.JReceptSignalTransduct doi: Res.2015;35:600–604. incellproliferation, differentiation, migration,senescence MAPK/ERK Sun Y, LiuWZ,T, Feng X,Yang N,ZhouHF. Signalingpathwayof 015-1472-2 an update.ArchToxicol doi:10.1007/s00204-. 2015;89:867–882. signalinginhumandiseases: ChoiEJ.CompromisedMAPK Kim EK, Invest. 2001;107:7–11. doi:10.1172/JCI11830 Tak PP, akey role ininflammatorydiseases. NF-kappaB: Firestein GS. 2010;5:e9585. doi:10.1371/journal.pone.0009585 marrow-deriveddendriticcells.PLoSOne . bone cytokines inLPS-activated roleforJunBintheinductionofproinflammatory NF-kappaB-dependent Gomard T, Michaud HA,Tempé D,Thiolon K,Pelegrin M,Piechaczyk M.An 10.1042/bj3020471 2):471–477.tory cytokine,interleukin1.BiochemJ.1994;302(pt doi: tional activation of the human stromelysin-1 gene bythe inflamma- Quinones S, Buttice G, Kurkinen M. Promoter elements in the transcrip- doi:10.1016/j.biocel.2007.12.0062008;40:1362–1378. matrix metalloproteinases and their inhibitors. The regulationof EdwardsDR. SwinglerTE,SampieriCL, Clark IM, 10.1038/353670a0 lation ofc-junmediatedbyMAPkinases.Nature.1991;353:670–674. doi: Pulverer BJ,Kyriakis Avruch Phosphory - J,NikolakakiE,Woodgett JM, JR. 10.1158/1940-6207.CAPR-14-0233 cell transformation.Cancer Prev Res (Phila).2014;7:1270–1281. doi: proteindegradation inthecheliensisin ainhibition of ated byPHLPP Zhao Q,etal.Crucialroleofc-JunphosphorylationatSer63/73 medi- Zhu J, Zhang J, Huang H, Li J, Yu Y, Jin H, Li Y, Deng X, Gao J, Biol. 2013;25:272–277. doi:10.1016/j.ceb.2013.01.002 CurrOpin Cell Chromatin-tetheredMAPKs. Klein AM,ZaganjorE,CobbMH. Sci USA.2008;105:9047–9052 . 2009;157:436–450. Br doi: J Pharmacol . 2009;157:436–450. 10.1111/j.1476-5381. H3K9me2 Regulates the VSMC InflammatoryResponse H3K9me2 RegulatestheVSMC Int JBiochemCellBiol. Proc Natl Acad J Clin