Genetic framework for GATA factor function in vascular biology

Amelia K. Linnemanna, Henriette O’Geenb, Sunduz Kelesc, Peggy J. Farnhamd, and Emery H. Bresnicka,1

aDepartment of Cell and Regenerative Biology, Wisconsin Institutes for Medical Research, University of Wisconsin Carbone Center, University of Wisconsin School of Medicine and Public Health, Madison, WI 53705; bGenome Center, University of California, Davis, CA 95616; cDepartment of Statistics and Biostatistics and Medical Informatics, University of Wisconsin School of Medicine and Public Health, Madison, WI 53706; and dUniversity of Southern California/Norris Comprehensive Cancer Center, Los Angeles, CA 90089

Edited by Stuart H. Orkin, Children’s Hospital and the Dana Farber Cancer Institute, Harvard Medical School, and Howard Hughes Medical Institute, Boston, MA, and approved June 28, 2011 (received for review May 26, 2011)

Vascular endothelial dysfunction underlies the genesis and pro- We used ChIP sequencing (ChIP-seq) and expression profiling gression of numerous diseases. Although the GATA transcription to describe a GATA-1– and GATA-2–driven genetic network in factor GATA-2 is expressed in endothelial cells and is implicated in human K562 erythroleukemia cells, which resemble primitive coronary disease, it has been studied predominantly as erythroid cells (12). This dataset, which was validated in primary a master regulator of hematopoiesis. Because many questions re- erythroid cells (12), lacked that impart a unique identity to garding GATA-2 function in the vascular biology realm remain endothelial cells. We hypothesize that the GATA-2 target unanswered, we used ChIP sequencing and loss-of-function strat- ensembles in endothelial and hematopoietic cells differ consider- egies to define the GATA-2–instigated genetic network in human ably, but analogous to its hematopoietic functions, GATA-2 exerts endothelial cells. In contrast to erythroid cells, GATA-2 occupied important activities in . Herein we use ChIP-seq and expression profiling in HUVEC to a unique target gene ensemble consisting of genes encoding key fi – determinants of endothelial cell identity and inflammation. GATA- de ne a GATA-2 driven genetic network that differs greatly from 2–occupied sites characteristically contained motifs that bind acti- that of K562 cells. The composition of the network indicated that vator -1 (AP-1), a pivotal regulator of inflammatory genes. GATA-2 functions in concert with activator protein-1 (AP-1), a pivotal regulator of inflammation (13, 14), thus linking GATA-2 GATA-2 frequently occupied the same chromatin sites as c-JUN and with inflammatory processes. Given the dearth of mechanistic in- c-FOS, heterodimeric components of AP-1. Although all three com- formation regarding vascular functions of GATA-2 and the broad ponents were required for maximal AP-1 target , scope of pathophysiologies involving vascular dysfunction with GATA-2 was not required for AP-1 chromatin occupancy. GATA-2 associated inflammation (15), linking endothelial GATA-2 to in- conferred maximal phosphorylation of chromatin-bound c-JUN at fl – ammatory genes establishes an important framework for un- Ser-73, which stimulates AP-1 dependent , in a derstanding the role of GATA-2 in normal and pathological chromosomal context-dependent manner. This work establishes a vascular states. link between a GATA factor and inflammatory genes, mechanistic insights underlying GATA-2–AP-1 cooperativity and a rigorous ge- Results and Discussion netic framework for understanding GATA-2 function in normal Exquisite Cell Type-Specificity of GATA-2 Chromatin Occupancy. To and pathophysiological vascular states. establish a framework for understanding GATA-2 function in endothelium, we used ChIP-seq to define the GATA-2 chromatin endothelium | genomics | genome-wide | transcriptome target sites in HUVEC. The ChIP was validated by quan- titating GATA-2 occupancy at the GATA2 +9.5 kb enhancer (16), n important approach to understanding the biology and which activates a linked LacZ reporter in the vasculature of Apathophysiology of the vascular system involves dissecting transgenic mouse embryos (17). GATA-2 occupancy at the +9.5 transcriptional mechanisms underlying the development and kb site was ∼15-fold greater than the preimmune control, whereas function of endothelial cells. Transcription factors implicated in no significant occupancy was detected at the negative control endothelial transcriptional mechanisms (1) include GATA bind- necdin (Fig. S1). The ChIP-seq analysis was conducted ing protein 2 (GATA-2), which is required to generate and/or via the standard mode of analysis for the ENCODE Consortium. maintain multipotent hematopoietic precursors during embryo- Immunoprecipitated DNA from two biological replicates was genesis (2, 3). GATA transcription factors function as master analyzed by ChIP-seq. Because the replicates met established regulators of development and have important functions in dif- criteria for reproducibility, the datasets were merged, and peaks ferentiated cells (4). Although GATA-2 is expressed in endothelial were called again. Unique sequences were mapped to the genome GENETICS (HG19). Replicates A and B yielded 9.9 million and 12.6 million cells (5, 6), the GATA-2–driven genetic network in endothelium unique sequences, respectively. Using a false-discovery rate of and its role in vascular biology are unclear. 0.0001, replicates A and B yielded 18,233 and 8,830 GATA-2– The pathophysiological importance of GATA-2 in the vascular fi GATA2 occupied loci (peaks), respectively. Comparison of replicates system is supported by the nding that polymorphisms revealed major overlap, with 7,619 (86%) of the B peaks present correlate with coronary artery disease (7). GATA-2 regulates in A. Merging the replicates yielded 15,529 peaks corresponding genes that encode important mediators of endothelial cell func- tion, including angiopoietin-2 (8), matrix metalloproteinase-2 (9), and vascular -1 (10). In vitro and in vivo Author contributions: A.K.L. and E.H.B. designed research; A.K.L. and H.O. performed studies indicate that GATA-2 functions in human umbilical vein research; A.K.L., H.O., S.K., P.J.F., and E.H.B. analyzed data; and A.K.L., S.K., P.J.F., and endothelial cells (HUVEC) and retinal endothelial cells to medi- E.H.B. wrote the paper. ate mechano-signaling–dependent , which involves The authors declare no conflict of interest. GATA-2–dependent induction of VEGF II (11). Despite This article is a PNAS Direct Submission. Gata2 this compelling evidence, -null mice die at embryonic day Data deposition: The data reported in this paper have been deposited in the Gene Ex- 10.5 (E10.5) because of severe , although their vasculature pression Omnibus (GEO) database, www.ncbi.nlm.nih.gov/geo (accession no. GSE29531). appears to be morphologically normal (3). However, the absence 1To whom correspondence should be addressed. E-mail: [email protected]. of an overt vasculogenesis defect does not negate the importance This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. of GATA-2 in adult endothelium. 1073/pnas.1108440108/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1108440108 PNAS | August 16, 2011 | vol. 108 | no. 33 | 13641–13646 Downloaded by guest on September 26, 2021 Endothelial Cell Function Transcription Factors Linking GATA-2 with the Inflammatory Regulator AP-1. We assessed 200 NOS3 Chr. 7 200 ETS1 -27 Chr. 11 whether the GATA-2 cistrome in HUVEC resembles that of +14.4 100 100 10 kb 20 kb –

Peak Height K562 cells, in which the majority of GATA-2 occupied peaks

NOS3 ABCB8 reside at sites containing the GATA motif (12). Of the 15,529 KCNH2 ATG9B ETS1 GATA-2 peaks, 12,930 (83%) contained at least one WGATAR Chr. 8 200 +139 Chr. 3 200 ANGPT1 FOXP1 +475 motif. Of the 6,976,111 GATA motifs in the , +165 100 50 kb 100 +12.4 100 kb 25,823 (0.37%) resided within the 15,529 peaks. Constrained de Peak Height fi fi FOXP1 novo motif nding with Cosmo (29) identi ed WGATAA that ANGPT1 RSPO2 EIF4E3 we defined previously for GATA-1 occupancy in K562 cells (Fig. +5 +3.3 Chr. 8 200 +149 Chr. 12 A 200 ANGPT2 -25 bp ETV6 +96 +212 3 ) (12). A broader GATA consensus, [CG][AT]GATAA[GAC] +76 +15.4 -36 +70 100 -7.2 20 kb 100 +214 50 kb [GAC], resided at 8,176 (43.3%) of the GATA-2–occupied K562 Peak Height MCPH1 ETV6 BCL2L14 cell peaks and 4,698 (35.4%) of the HUVEC peaks. ANGPT2 PRB2 LRP6 De novo motif finding using MEME (30) to pinpoint cis-ele- – 200 -9.7 Chr. 17 200 Chr. 5 ments at GATA-2 occupied sites in HUVEC revealed canonical ICAM-2 +64 bp MEF2C +56 100 +139 -1.1 fl B 10 kb 100 +3.7 20 kb motifs for the in ammatory regulator AP-1 (14) (Fig. 3 ) and c-

Peak Height C C17orf72 ETS (Fig. 3 ), which has important functions in endothelial cells ICAM-2 ERN1 LOC645323 MEF2C (31). AP-1 motifs (TGA[G/C]TCA) (32) resided in 45.4% of the HUVEC GATA-2 peaks (7,055 of 15,523 peaks). This was highly 200 ANGPTL2 +9.5 Chr. 9 200 KLF2 -8.3 Chr. 19 +73 -41 +14 +27 fi P < 100 20 kb 100 10 kb signi cant ( 0.01) relative to 16.3% of GATA-2 peaks in K562 Peak Height cells (3,451 of 21,167 peaks). Because 69% of GATA-2–occupied RALGPS1 KLF2 ANGPTL2 EPS15L1 peaks overlapped with those occupied by the jun proto-oncogene (c-JUN) component of AP-1 (Fig. 3D and Dataset S3), 29% Fig. 1. Representative GATA-2 target genes in HUVEC. Signal maps of GATA-2 overlapped with FBJ murine osteosarcoma viral oncogene ho- occupancy sites in HUVEC identified by ChIP-seq. Genes important for endo- molog (c-FOS) (Fig. 3E and Dataset S4), and only 1% (137) thelial cell function and transcription factors are depicted. Arrows indicate ChIP- overlapped with peaks of c- occupancy (23,525 total c-Myc seq peak locations relative to the transcription start site of the respective GATA- peaks), this GATA-2–AP-1 link has broad importance. The lo- 2targetgene(kb). cation with respect to neighboring genes of sites containing both GATA-2 and c-JUN did not differ significantly from the location of the complete cohort of GATA-2 occupancy sites. Examples of to 6,081 genes (Dataset S1). The mean peak height was 75 bp AP-1 target genes in which GATA-2 overlapped with c-JUN and (range: 14–229 bp), and the mean width was 417 bp. IL-8 CXCL2 CSF2 – c-FOS include (33), (34), and (35). GATA-2, The GATA-2 occupied genes encode known to be c-JUN, and c-FOS overlapped at RSPO3, which is not known to be important for endothelial cell function, including endothelial an AP-1 target (Fig. 3F). A high fraction of chromatin sites bound NOS3, angiopoietins 1, 2, and 3, and intercellular adhesion by both GATA-2 and c-Jun were enriched in H3K4me1 or molecule-2 (Fig. 1). Certain genes encoded transcription factors, H3K4me3 (76 and 83%, respectively). including Kruppel-like factor-2 (KLF2) (Fig. 1). KLF2 is induced We compared the frequency in which the c-ETS motif resided by shear stress, mediates shear stress signaling (18), determines at GATA-2–occupied sites in HUVEC vs. K562 cells. In HUVEC, vascular tone (19), mediates endothelial cell thrombotic function (20), and promotes vessel maturation (21). KLF2 is repressed by 2% 1% fl β GATA-2 Occupancy the proin ammatory cytokine IL-1 (18) and mediates the anti- A B 3% 1% inflammatory actions of statins (22). Additional transcription 39% 54% factors included the forkhead factor FOXP1 and multiple v-ets 2,283 13,256 18,885 erythroblastosis virus E26 oncogene homolog (ETS)/ETS-variant (85%) (89%) (ETV) factors (Dataset S2), including ETV6, which is required Enhancer Promoter HUVEC K562 for hematopoiesis and maintenance of the vascular network (23, Intron Immediate downstream 24). GATA-2 occupied myocyte-specific enhancer factor 2C Exon 3’ UTR (MEF2C), which is essential for vascular development (25). MEF2C contains an endothelial enhancer, corresponding to the C D fi 14,179 5,505 GATA-2 peak at +139 kb, which directs endothelial cell-speci c 1,344 123,788 10,018 38,791 expression of a reporter in E7.5–E9.5 mice (26). The ChIP-seq (91%) (35%) analysis revealed a rich set of targets, including numerous genes GATA-2 H3K4me1 GATA-2 H3K4me3 not linked previously to endothelial cell biology. The majority of GATA-2–occupied peaks (13,256 peaks, 85%) were not occupied in our prior K562 cell analysis (12) (Fig. 2A). E F 152 411 In HUVEC and K562 cells (12), the majority of peaks resided 15,371 57,205 15,112 60,385 (1%) (3%) distal to the genes in putative enhancer regions (Fig. 2B). The

unique chromatin-bound sites in HUVEC vs. K562 cells indi- GATA-2 H3K27me3 GATA-2 H3K36me3 cates that GATA-2 occupies a very small subset of genomic GATA motifs in a cell type-specific manner. To explore the re- G lationship between GATA-2 occupancy and the epigenome, we 15,419 104 5,882 evaluated histone marks in HUVEC at occupied sites. This (1%) analysis revealed that 91% of GATA-2–occupied sites overlap with H3K4me1 (Fig. 2C), a diagnostic enhancer mark (27). A GATA-2 H4K20me1 smaller subset of occupied sites (35%) overlaps with H3K4me3 (A) – (Fig. 2D). As expected for a mark associated with transcrip- Fig. 2. Computational mining of ChIP-seq data. Comparison of GATA-2 occupied peaks in HUVEC and K562 cells. (B) Locations of GATA-2–occupied tion (28), 61% of these overlapping peaks reside in introns and peaks relative to nearest-neighbor genes determined with the Cis Element exons. A small cohort of occupied sites overlaps with H3K27me3 Annotation System (http://liulab.dfci.harvard.edu/CEAS/; http://ceas.cbi.pku. (Fig. 2E), H3K36me3 (Fig. 2F), and H4K20me1 (Fig. 2G) (1%, edu.cn/). (C–G) Comparison of GATA-2–occupied peaks and peaks of H3K4me1 3%, and 1%, respectively). (C), H3K4me3 (D), H3K27me3 (E), H3K36me3 (F), and H4K20me1 (G).

13642 | www.pnas.org/cgi/doi/10.1073/pnas.1108440108 Linnemann et al. Downloaded by guest on September 26, 2021 5,446 (35.1%) of the 15,523 GATA-2–occupied sites contained c- Direct GATA-2 Target Gene Ensemble in Endothelium. The frequent ETS motifs. In K562 cells, 2,737 (12.9%) of the 21,167 GATA-2– occurrence of AP-1 motifs at GATA-2–occupied sites and occupied sites contained c-ETS motifs. Based on a binomial test GATA-2 sharing of occupancy sites with a key component of AP-1 for proportions, the differential frequency of c-ETS motifs at suggest that GATA-2 and AP-1 interact functionally. We used GATA-2–occupied sites in HUVEC vs. K562 cells was highly RNA interference to establish GATA-2 target genes in HUVEC. − significant (P < 2.2e 16). Only 2,249 (14.5%) of the 15,523 GATA2 mRNA was considerably lower in HUVEC than in K562 GATA-2–occupied sites in HUVEC contained both c-ETS and cells (Fig. 4A) (12), but endogenous GATA-2 in HUVEC was AP-1 motifs. A GATA-2 peak was more likely to contain an AP-1 detectable by Western blotting (Fig. 4C). GATA2 siRNA signifi- motif if it lacked a c-ETS motif (P < 0.001). Thus, it is likely that cantly reduced GATA2 mRNA and protein in HUVEC vs. control GATA-2 does not function frequently as a ternary complex with siRNA-transfected cells (Fig. 4 B and C) without overtly affecting AP-1 and c-ETS in HUVEC. HUVEC morphology. Expression profiling of two biological replicates was conducted, and genes dysregulated upon GATA-2 knockdown were analyzed further. The direct GATA-2 targets GATA AP-1 c-ETS were identified by comparing the differential expression with A 2.0 B 2.0 C 2.0 1.5 1.5 1.5 GATA-2 occupancy determined by ChIP-seq. This analysis revealed 116 direct targets differentially expressed by twofold or 1.0 1.0 1.0 D 0.5 0.5 0.5 more upon GATA-2 knockdown (Fig. 4 ). Because knockdowns

0 0 0 do not eliminate expression, the analysis underestimates the tar- Information Content Information Content 12345678 1234567891011 Information Content 12345678 Position Position Position get gene ensemble. The direct targets encode regulators of cell migration, in- D HUVEC E HUVEC cluding semaphorin 3A (36), and podocalyxin-like (37). GATA- 2–activated direct targets encode proinflammatory mediators, including interleukins, the chemokine ligand CXCL2, and the 10,761 4,578 4,778 (69%) 39,712 11,354 (29%) 7,017 A B C 1.2 0.20 GATA-2 c-JUN GATA-2 c-FOS 0.16 0.8 -3 * 0.12 52 GATA-2 F IL8 0.08

0.4 r

0.04 M x 10 mRNA Levels mRNA Levels (Relative Units) 150 1 kb (Relative Units) 0 0 Tubulin GATA-2 Control GATA2 -RT -RT K562 siRNA siRNA Control GATA2 siRNA siRNA 100 HUVEC c-JUN D E F SIPA1L3 CPEB4 8 6 PODXL ARHGAP26 60 RNF149 PDE3A Peak Height 6 MALL PRKCE 4 c-FOS FYCO1 BNC1 SLC44A1 CD274 4 PTRF ZC3H12C 2 S100A6 CSF2 2 Fold Change CD47 FAM108C1 Fold Change CXCL2 (Upregulated)

ID1 GATA2 (Downregulated) 0 0 160 GOSR2 FRZB 2 kb MFI2 BDNF FST IL8 STT3A IL1RL1 CD55 AIP3CSF2 KLF2 DDAH1RSPO3PHTF2 RUNX1 PODXL YES1 FAM83D SEMA3A TNF RNF149 HS3ST1 GATA-2 RNF141 DUSP10 SPRED1 LRRFIP1 G TSPAN9 KIF18A 0.08 SRPX DCBLD2 Preimmune CDCA7 KBTBD8 GATA-2 80 TUBGCP3 KIAA1279 0.06 c-JUN PTPRG FBN2 FANCL TSLP SETBP1 SERPINB2 0.04

Peak Height NUCKS1 RGS2 TMCO1 IL6 50 c-FOS WIPI1 TGFB2 DPYSL2 PLEKHB2

(Relative Units) 0.02 AK3L1 ARID5B

CAPN2 KITLG GATA-2 Occupancy SMARCA2 SYDE2 CSF2 TMEM87B LPP 0 CYB5R3 COL21A1 130 2 kb SLC1A1 KLF5 FST ATG16L2 CD200 PHTF2 RSPO3 DDAH1 (-32kb) (-30 kb) -globin (-67 kb) (+28 kb) (+13 kb) β ITGB4 IL7R SEMA3A (-170 kb) (-146 kb) (-268 kb) (-428 kb) (-434 kb) (-477 kb) (-339 kb)

NFE2L3 TNFAIP3 IL8

GATA-2 (int3 peak1) KLF2 (-8 kb) CSF2 (-3 kb) COL13A1 PLAU (int3 peak2) KLF2 (+27 kb) CSF2

ROR1 PMCH PODXL PODXL RNF149 (-21 kb) HS3ST1 (-77 kb) RUNX1 RUNX1 RUNX1 RUNX1 RUNX1 PODXL HS3ST1 (-120 kb) HS3ST1 (-475 kb) HS3ST1 (-476 kb) C9orf6 CD55 TNFAIP3 TNFAIP3 RUNX1 110 ZFP36L1 FGF18 RUNX1 CGNL1 GSTO2 RALB SEMA3A 0.05 c-JUN PAG1 CXCL3 H GENETICS ZNF175 DLX2 GRAMD1B DKK1 0.04

Peak Height 75 RUNX1T1 INHBA 2 CENPE CXCL2 0.03 R = 0.68 c-FOS IGF2BP1 IL8 ABL2 GEM RFX3 RELN 0.02 VIP PHTF2 RUNX1 RSPO3 5 RSPO3 BIRC3 DDAH1 3 0.01 PCDH17 ATF3 2 NFKBIB ARHGAP29 0 0 70 10 kb PTGS2 FST -2 Quantitative ChIP Signal 50 100 150 200 250 GATA-2 DEPDC1 HIST1H4H -3 MAP1LC3B IFIT2 -5 ChIP-seq Peak Height

70 Fig. 4. Direct GATA-2 target gene ensemble. (A) Comparison of GATA2 c-JUN levels in HUVEC and K562 cells. (B) Quantitative RT-PCR analysis of GATA2 transcript levels after treatment with nontargeting control siRNA or GATA2 Peak Height 50 c-FOS siRNA. (C) Western blotting of GATA-2 in HUVEC transfected with control or GATA2 siRNA. Whole-cell samples were analyzed. *, nonspecific band. (D) Heat map depicting the mean fold change of expression at GATA-2–occu- Fig. 3. Linking GATA-2 and AP-1 function. (A–C) Logos of overrepresented pied loci resulting from GATA2 knockdown (n = 2). (E and F) Quantitative motifs from GATA-2–occupied ChIP-seq peaks: (A) GATA motif from con- RT-PCR validation of array results in GATA-2–activated (down-regulated strained motif analysis and (B) AP-1 and (C) c-ETS motifs from de novo motif with knockdown) (E) and repressed (up-regulated) genes (F). (G) Quantita- finding. (D and E) Comparison of GATA-2–occupied peaks and peaks of (D) tive ChIP validation of GATA-2 occupancy at loci dysregulated by the c-JUN and (E) c-FOS occupancy in HUVEC. (F) Representative profiles dem- knockdown. Data shown are mean ± SE; n =3.(H) Correlation between onstrating overlap among GATA-2, c-JUN, and c-FOS occupancy peaks. height of ChIP-seq peak and quantitative ChIP signal.

Linnemann et al. PNAS | August 16, 2011 | vol. 108 | no. 33 | 13643 Downloaded by guest on September 26, 2021 cytokine CSF2 (GM-CSF). GATA-2 directly repressed KLF2, phorylated c-JUN occupancy (P < 0.05), but not total c-JUN and c- the key mediator of shear stress signaling in endothelium that FOS occupancy, at the IL-8 promoter and distal sites at RSPO3 and opposes inflammation. GATA-2 activated RSPO3, which con- ATF3 genes (Fig. 6D). Loss of chromatin-bound phosphorylated trols placental vascular plexus development and regulates an- c-JUN was not associated with a change in total c-JUN (Fig. 6C). giogenesis during embryogenesis (38). Secreted RSPO3 activates To assess the specificity of the GATA-2–dependent phosphor- wingless-type/β-catenin signaling, which exerts a proinflammatory ylated c-JUN occupancy, occupancy was analyzed at the GATA-2– function (38). GATA-2 induced expression of activating tran- insensitive IL-1B and IL-17D genes (Fig. 6D). GATA-2 knock- scription factor 3 (ATF3), a stress-regulated down did not affect phosphorylated c-JUN occupancy at these that protects endothelial cells from TNF-α–induced cell death genes. Occupancy of all factors was low at the HBB promoter and (39) and negatively regulates allergic airway inflammation (40). an unoccupied site 1 kb upstream of the IL-8 promoter (Fig. 6D). A subset of GATA-2–activated and –repressed targets was sub- The context-dependent GATA-2 requirement for phosphorylated jected to additional validation by real-time RT-PCR (Fig. 4 E and c-JUN occupancy may explain why certain AP-1 target genes that F) and quantitative ChIP analysis using the human β-globin (HBB) are occupied by GATA-2 and AP-1 (Datasets S3 and S4) are in- promoter as a negative control (Fig. 4G). ChIP-seq peak height sensitive to reducing GATA-2. In principle, GATA-2 may promote correlated (R2 = 0.68) with the quantitative ChIP signal (Fig. 4H). c-JUN phosphorylation indirectly or may promote phosphorylated Our results indicate that GATA-2, c-JUN, and c-FOS com- c-JUN chromatin occupancy, and it has been reported that GATA- monly occupy the same chromatin sites containing GATA and 2 and c-JUN can interact (43). Coimmunoprecipitation analysis AP-1 motifs. Taken together with the fact that AP-1 is a key with HUVEC whole-cell lysates did not provide evidence for regulator of inflammatory genes such as IL-8 and CSF2 that are a stable complex containing GATA-2 and c-JUN. GATA-2 targets (Fig. 4 D and E), this finding suggests the ex- We also tested whether c-JUN promotes GATA-2 chromatin istence of a GATA factor–AP-1 regulatory network to control occupancy, but knocking down c-JUN did not affect GATA-2 inflammatory genes. Ontological analysis of the direct GATA-2 chromatin occupancy (Fig. S2). target-gene ensemble revealed an enrichment of inflammatory To test whether GATA-2 facilitation of AP-1–mediated tran- genes, the majority of which are GATA-2–activated (Fig. 5A). scription has biological implications, we asked whether GATA-2 regulates secretion of the critical inflammatory mediator IL-8. GATA-2 Requirement for Assembly of Functional AP-1 Complexes on GATA-2–knockdown HUVECs were treated with the proin- Chromatin. In our loss-of-function analysis, HUVEC were cul- flammatory factor TNF-α, which induced IL-8 secretion several tured in medium containing FBS, EGF, and FGF2, which induce fold as measured by ELISA (Fig. 6E). Knocking down GATA-2 signaling that can activate AP-1 (13). Our results suggested that significantly reduced IL-8 secretion (P = 0.0015), consistent with GATA-2 and AP-1 confer maximal AP-1 target gene transcrip- the mRNA expression data. tion. We tested whether activity of presumptive AP-1 targets requires AP-1 in HUVEC. with dominant-negative Linking GATA-2, AP-1, and Inflammatory Genes. The results de- c-FOS (A-FOS) (41, 42) significantly reduced IL-8 (P < 0.01), scribed herein, representing the genome-wide analysis of a GATA CSF2 (P < 0.01), and IL-6 (P < 0.01) expression (Fig. 5C). We factor in endothelium, provide a rigorous genetic framework for tested genes that emerged from our direct GATA-2 target screen understanding GATA-2 function in vascular biology. GATA-2 that were not established AP-1 targets. Expression of RSPO3 (P < controls expression of genes involved in establishing endothelial 0.01), CXCL1, and RUNX1 was reduced by A-FOS. c-JUN, and c- cell and inflammation. Analysis of 116 GATA-2– FOS knockdowns confirmed AP-1 regulation of select GATA-2 occupied genes differentially regulated upon knocking down targets; CSF2 and CXCL1 expression were not affected signifi- GATA-2 using Ingenuity Pathways Analysis (http://www.ingenuity. cantly (Fig. 5 B and C). Thus, GATA-2 and AP-1 ensure maximal com/products/pathways_analysis.html) revealed a complex net- activity of a cohort of the target genes. work (Fig. S3) in which factors were knitted together based on We considered potential mechanisms underlying the GATA-2– established protein–protein interactions and abilities to regulate AP-1 cooperation (Fig. 6A). At target genes (Fig. 6AI), reduced expression of other factors. The network highlights a GATA-2 GATA-2 occupancy might decrease AP-1 occupancy (Fig. 6AII), function in regulating genes mediating inflammation (Fig. 6F). decrease AP-1 phosphorylation (Fig. 6AIII), or induce AP-1– GATA-2 targets also included genes encoding anti-inflammatory independent molecular alterations (Fig. 6AIV). To distinguish factors, including TNFAIP3, which is induced by proinflammatory among these possibilities, occupancy of c-FOS, c-JUN, and c-JUN factors and suppresses proinflammatory signaling (44, 45) and phosphorylated at serine 73 was analyzed in control siRNA- and ATF3 (40), as described above. Certain AP-1 targets are regulated GATA2 siRNA-transfected cells. The analysis was conducted un- by the proinflammatory factor NF-κB (46). Scanning the GATA-2 der conditions in which GATA2 expression was greatly reduced peaks with the NF-κB position/weight matrix from the JASPAR (Fig. 6B). Knocking down GATA-2 significantly reduced phos- database using the FIMO tool (47) revealed that only 9.1% of the

Fig. 5. GATA-2/AP-1–dependent regulation of in- -3 52 * -3 A Gene Biological Function B 52 fl A fl x 10 x 10 FOS r c-JUN r ammatory genes. ( ) Genes involved in in am- CD274 Programmed cell death ligand 38 M 38 M CD47 Receptor for C-terminal cell binding domain of thrombospondin Tubulin mation and their expression changes with GATA- CD55 Decay accelerating factor for complement Tubulin CSF2 Cytokine; colony stimulating factor 2 (granulocyte-macrophage) Control c-JUN Control c-FOS 2 knockdown. Blue, down-regulated with GATA-2 CXCL2 Chemokine siRNA siRNA siRNA siRNA CXCL3 Chemokine C knockdown; Yellow, up-regulated with GATA-2 ** p<0.01 A-FOS DCBLD2 Modulates PDGF signaling; upregulated after vascular injury 100 * p<0.05 ** HIST1H4H Histone H4 family member * knockdown. (B) Western blotting of endogenous c- 75 ** IL1RL1 Interleukin receptor; induced by proinflammatory stimuli 3.00 IL6 Inflammation and B cell maturation 50 ** JUN and c-FOS in HUVEC transfected with c-JUN or c- 2.00 ** 25

IL7R V(D)J recombination mRNA Levels 1.00 (% of Control) FOS siRNA, respectively. Transfected cells were boiled IL8 Chemoattractant, angiogenic factor, and inflammatory response mediator 0.00 INHBA Erythroid growth/differentiation factor; regulates osteotrophic factors IL8 IL6 CSF2 RSPO3 CXCL1 RUNX1 -1.00 in SDS-sample buffer, and whole-cell samples were ITGB4 Reduces human endothelial inflammation -2.00 ** p<0.001 c-JUN siRNA KITLG Functions pleiotropically; cell migration 100 fi C -3.00 * p<0.005 analyzed by Western blotting. *, nonspeci cband.( ) LRRFIP1 Represses alpha 75 NFKBIB I-kappa-B-beta; inhibits NF-kappa B complex * Real-time RT-PCR analysis of gene expression in 50 PAG1 Regulation of activation ** ** ** 25 ** mRNA Levels PLAU Extracellular matrix degradation; possibly tumor cell migration and proliferation (% of Control) HUVEC transfected with a dominant-negative AP-1 PTGS2 Inducible cyclooxygenase RUNX1 Transcription factor c-JUN IL8 IL6 CSF2 RSPO3 CXCL1 RUNX1 antagonist (A-FOS), c-JUN siRNA, or c-FOS siRNA. The SEMA3A Modulates cell motililty; chemoattractant ** p<0.01 c-FOS siRNA 100 expression of empty vector-transfected cells (A-FOS) SERPINB2 Placental plasminogen activator inhibitor * p<0.05 TGFB2 Cytokine; suppression of IL-2-dependent T-cell growth 75 or nontargeting control siRNA-transfected cells (c-JUN TNFAIP3 Induced by TNF; Inhibits NF-kappa-B activation and TNF-mediated * 50 * TSLP Hemopoietic cytokine ** * 25 ** and c-FOS) was designated as 100%, and expression mRNA Levels ZFP36L1 Putative transcription factor; TPA-responsive gene (% of Control) for each gene is represented as a percentage of con- c-FOS IL8 IL6 CSF2 RSPO3 CXCL1 RUNX1 trol expression. Data shown are mean ± SE; n ≥ 3.

13644 | www.pnas.org/cgi/doi/10.1073/pnas.1108440108 Linnemann et al. Downloaded by guest on September 26, 2021 − peaks match significantly (P ≤ 1e 4) to the matrix, a finding that – κ A I is inconsistent with a frequent GATA-2 NF- B association at + P N GATA-2

HUVEC chromatin sites. FOS JUN C Because certain GATA-2 target genes with established roles in inflammation were co-occupied by GATA-2, c-JUN, and c-FOS, GATA-2 knockdown reduced phosphorylated c-JUN occupancy III fl II IV and IL-8 secretion, and AP-1 is a pivotal mediator of in ammation, + + P + – X X X it is attractive to propose that GATA-2 AP-1 cooperation is a FOS JUN FOS JUN fl common mode of orchestrating in ammatory processes. Vascular Loss of AP-1 Loss of Phosphorylation Independent endothelial dysfunction involving inflammation underlies diseases including , rheumatoid arthritis, and type II di- GATA2

B 1.0 C -3 abetes (15). 52 p-c-JUN x 10 Given the biological importance of genes constituting the r 38 (Ser73) M GATA-2 genetic network in HUVEC, why does the Gata2- 0.5 knockout mouse appear to have morphologically normal vascu- lature, despite dying at E10.5 because of severe anemia (3)? Since Tubulin 0.0 Control GATA2 Control GATA2 multiple GATA factors are expressed in endothelium (48), a rig- Relative mRNA Levels siRNA siRNA orous loss-of-function approach to ascertain GATA-2 function in siRNA siRNA 0.12 the adult vasculature might require multigenic disruptions to re- D IL8 Promoter IL8 (-1 kb) veal critical GATA factor-dependent pathways; alternatively, an 0.08 important vascular activity of GATA-2 might not be apparent * given the early embryonic lethality. 0.04 Occupancy

The direct regulation of inflammatory genes by GATA-2 has (Relative Units) 0.00 considerable biological and pathophysiological implications. It 0.12 will be important to implement additional functional analyses RSPO3 (+28 kb, intron 1) IL1B 0.08 and to determine whether endothelial cell subtypes differ in this * mechanism. During the review of our manuscript, a genome-wide 0.04 analysis of GATA-2 occupancy in human microvascular endo- Occupancy (Relative Units) thelial cells (HMVEC) was reported (49). Comparison of the 0.00 0.20 HUVEC and HMVEC datasets will be informative. ATF3 (-12 kb) IL17D Given the oncogenic activity of AP-1 subunits (50, 51), the role 0.16 for GATA-3 in (52), and links between GATA-2 0.12 dysregulation and leukemia (53), it will be important to determine 0.08 Occupancy 0.04 * whether GATA factors cooperate with AP-1 in diverse biological (Relative Units) 0.00 0.12 PI c-FOS c-JUN p-c-JUN contexts and to what extent disruption of this mechanism under- HBB Promoter lies human pathologies. Because strong evidence implicates in- (Ser73) 0.08 flammation in a plethora of disease processes, including cancer, Control siRNA GATA2 siRNA studies to evaluate GATA-2 function in endothelium in diverse 0.04 Occupancy

disease models will be particularly instructive. (Relative Units) 0.00 PI c-FOS c-JUN p-c-JUN Materials and Methods (Ser73) Cell Culture. HUVEC were maintained in Medium 200 (Cascade Biologics) 1000 F supplementedwithLowSerum GrowthSupplement(Cascade Biologics)and1% E Control siRNA *

800 N antibiotic/antimycotic (Invitrogen). Cells were passaged at ∼70–80% conflu- GATA2 siRNA P GATA-2

FOS JUN C ence with 0.05% trypsin, and cells from passages four and six were used. 600

400 Antibodies. Rabbit anti–GATA-2 polyclonal antibody was described pre- IL-8 (pg/mL) Proinflammatory 200 viously (54). Anti–c-FOS (SC-253) and anti–c-JUN (SC-45) were from Santa Anti-inflammatory – 0 Cruz Biotech. Anti p-c-JUN (Ser73) was from Cell Signaling Technology Vehicle TNFα (#9164). Mouse monoclonal α-tubulin (#CP06) was from Calbiochem. Fig. 6. GATA-2 requirement for assembly of functional AP-1 complexes on – chromatin. (A) Models as described in Results and Discussion.(B) Real-time

Transfection and RNA Interference. HUVEC were seeded 3 5 d before siRNA GENETICS GATA2 C or plasmid transfection and grown to ∼70% confluence. After trypsiniza- RT-PCR quantitation of mRNA. ( ) Western blotting of c-JUN phos- tion, ∼106 cells per experiment were resuspended in HUVEC Nucleofector phorylated at serine 73 [p-c-JUN (Ser73)] in HUVEC transfected with control GATA2 D Solution (Lonza), and 0.45–0.60 nmol of siRNA, 5 μg A-FOS expression vector, or siRNA. Whole-cell samples were analyzed by Western blotting. ( ) or empty vector was transfected using the Nucleofector system (Lonza). Quantitative ChIP of preimmune control (PI), c-FOS, c-JUN, and p-c-JUN (Ser- IL-8 RSPO3 ATF3 siGENOME SMARTpool GATA2 siRNA, c-JUN siRNA, c-FOS siRNA, and control 73) occupancy at the promoter, , and , AP-1 targets that are GATA-2–independent (IL-1B, IL-17D), and negative controls (HBB and a site 1 siRNA (SMARTpool #1) were from Dharmacon (Thermo Scientific). CMV-500- kb upstream of the IL-8 promoter). Data shown are mean ± SE; n ≥ 3. (E) A-Fos and CMV-500 empty vector were provided by Charles Vinson (National ELISA quantitation of IL-8 levels in medium from control siRNA- or GATA-2 Cancer Institute, Bethesda, MD). Cells were analyzed 24 or 40 h after siRNA siRNA-transfected cells treated with 10 ng/mL TNF-α or vehicle for 6 h. Data or A-Fos transfection, respectively. shown are mean ± SE; n =6.(F) GATA-2–AP-1 coregulation of inflammatory genes in endothelium. Generation of ChIP-Seq Data. HUVEC histone ChIP-seq data were generated at the Broad Institute and by the B. Bernstein group (Massachusetts General Hospital, Boston, MA). HUVEC c-JUN ChIP-seq data were generated by D. Raha and Mike Snyder (Stanford University, Stanford, CA). HUVEC c-Myc Quantitative ChIP Analysis. ChIP was conducted as described (55) using 1 × 108 ChIP-seq data were generated by Vishy Iyer (University of Texas at Austin, TX). HUVEC cells per condition for ChIP-seq, or 1 × 107 HUVEC cells per condition HUVEC GATA-2, HUVEC c-FOS, and K562 GATA-2 ChIP-seq data were gener- for quantitative ChIP. ated by the P. Farnham group. These data were collected as part of the EN- CODE Consortium (http://www.genome.gov/10005107) and are available on ChIP-Seq Analysis. HUVEC were crosslinked for 10 min with 1% formaldehyde, the University of California, Santa Cruz Genome Browser. snap frozen, and stored at −80 °C. ChIP was conducted as described (http://

Linnemann et al. PNAS | August 16, 2011 | vol. 108 | no. 33 | 13645 Downloaded by guest on September 26, 2021 www.genomecenter.ucdavis.edu/farnham/pdf/FarnhamLabChIP%20Protocol. Expression Profiling. Control siRNA- and GATA2 siRNA-treated samples col- pdf). Chromatin from 108 cells was diluted with three volumes of immuno- lected 24 h posttransfection were analyzed by hybridization to human 244k precipitation (IP) dilution buffer and incubated at 4 °C overnight with 60 μLof expression microarrays (Agilent Technologies). For each sample, 1.2 μg RNA rabbit anti–GATA-2 antibody. StaphA cells were blocked with BSA. StaphA was labeled with Cy-5 (control siRNA) or Cy-3 (GATA2 siRNA), and labeled and Staph-Seq cells (SIGMA) were used. Blocked StaphA/Staph-Seq cells (100 cRNAs were combined and hybridized. Data were collected with an Agilent μL) were added to the antibody/chromatin mixture and rotated for 15 min at scanner. Two biological replicates for control siRNA/GATA-2 siRNA pairs room temperature. StaphA/Staph-Seq cells were washed twice with dialysis were each hybridized in duplicate, yielding similar results, and average sig- buffer and four times with polyclonal IP wash buffer. After reversal of cross- nals between the four replicates were used to establish fold-change cutoffs. links and RNase treatment, DNA was purified and analyzed as described with the Illumina GA2 platform (56). Short-sequence reads were aligned to the PCR Primers. Primer sequences are indicated in Dataset S5. genome, and peaks were called using Sole-search (false-discovery rate, 0.0001; α value 0.001) with sequenced HUVEC input DNA as background (57). ChIP-seq ACKNOWLEDGMENTS. We thank Dr. Charles Vinson (National Institutes of Health) for providing the A-Fos vector. ChIP-seq was supported by National data produced by the ENCODE Project Consortium were downloaded from the Human Genome Research Institute funds. We acknowledge funding from University of California, Santa Cruz browser and analyzed with Sole-search. National Institutes of Health Grants R21 HL091520 and DK068634 (to E.H.B.), Peak overlap was analyzed with Sole-search, allowing a distance of 200 bp HG003747 (to S.K.), and 1U54HG004558 (to P.J.F.). A.K.L. is an American between peaks. Heart Association postdoctoral fellow.

1. De Val S, Black BL (2009) Transcriptional control of endothelial cell development. Dev 30. Bailey TL, Williams N, Misleh C, Li WW (2006) MEME: Discovering and analyzing DNA Cell 16:180–195. and protein sequence motifs. Nucleic Acids Res 34(Web Server issue):W369–W373. 2. Tsai F-Y, Orkin SH (1997) Transcription factor GATA-2 is required for proliferation/ 31. Zhan Y, et al. (2005) Ets-1 is a critical regulator of Ang II-mediated vascular in- survival of early hematopoietic cells and mast cell formation, but not for erythroid flammation and remodeling. J Clin Invest 115:2508–2516. and myeloid terminal differentiation. Blood 89:3636–3643. 32. Angel P, et al. (1987) Phorbol ester-inducible genes contain a common cis element 3. Tsai FY, et al. (1994) An early haematopoietic defect in mice lacking the transcription recognized by a TPA-modulated trans-acting factor. Cell 49:729–739. factor GATA-2. Nature 371:221–226. 33. Yasumoto K, et al. (1992) Tumor necrosis factor alpha and interferon gamma syner- 4. Bresnick EH, Lee HY, Fujiwara T, Johnson KD, Keles S (2010) GATA switches as de- gistically induce interleukin 8 production in a human gastric cancer cell line through velopmental drivers. J Biol Chem 285:31087–31093. acting concurrently on AP-1 and NF-kB-like binding sites of the interleukin 8 gene. J 5. Dorfman DM, Wilson DB, Bruns GA, Orkin SH (1992) Human transcription factor Biol Chem 267:22506–22511. GATA-2. Evidence for regulation of preproendothelin-1 gene expression in endo- 34. Faouzi S, et al. (2001) Anti-Fas induces hepatic chemokines and promotes in- thelial cells. J Biol Chem 267:1279–1285. flammation by an NF-kappa B-independent, caspase-3-dependent pathway. J Biol 6. Lee ME, Temizer DH, Clifford JA, Quertermous T (1991) Cloning of the GATA-binding Chem 276:49077–49082. protein that regulates endothelin-1 gene expression in endothelial cells. J Biol Chem 35. Wang CY, et al. (1994) Activation of the granulocyte-macrophage colony-stimulating 266:16188–16192. factor promoter in T cells requires cooperative binding of Elf-1 and AP-1 transcription 7. Connelly JJ, et al. (2006) GATA2 is associated with familial early-onset coronary artery factors. Mol Cell Biol 14:1153–1159. disease. PLoS Genet 2:e139. 36. Serini G, Valdembri D, Bussolino F (2006) Integrins and angiogenesis: A sticky busi- 8. Simon MP, Tournaire R, Pouyssegur J (2008) The angiopoietin-2 gene of endothelial ness. Exp Cell Res 312:651–658. cells is up-regulated in hypoxia by a HIF binding site located in its first intron and by 37. Larrucea S, et al. (2008) Expression of podocalyxin enhances the adherence, migra- the central factors GATA-2 and Ets-1. J Cell Physiol 217:809–818. tion, and intercellular communication of cells. Exp Cell Res 314:2004–2015. 9. Han X, Boyd PJ, Colgan S, Madri JA, Haas TL (2003) Transcriptional up-regulation of 38. Kazanskaya O, et al. (2008) The Wnt signaling regulator R-spondin 3 promotes an- endothelial cell matrix metalloproteinase-2 in response to extracellular cues involves gioblast and vascular development. Development 135:3655–3664. GATA-2. J Biol Chem 278:47785–47791. 39. Kawauchi J, et al. (2002) Transcriptional repressor activating transcription factor 3 10. Minami T, Aird WC (2001) Thrombin stimulation of the vascular cell adhesion mole- protects human umbilical vein endothelial cells from tumor necrosis factor-alpha- cule-1 promoter in endothelial cells is mediated by tandem nuclear factor-kappa B induced apoptosis through down-regulation of transcription. J Biol Chem 277: and GATA motifs. J Biol Chem 276:47632–47641. 39025–39034. 11. Mammoto A, et al. (2009) A mechanosensitive transcriptional mechanism that con- 40. Gilchrist M, et al. (2008) Activating transcription factor 3 is a negative regulator of trols angiogenesis. Nature 457:1103–1108. allergic pulmonary inflammation. J Exp Med 205:2349–2357. 12. Fujiwara T, et al. (2009) Discovering hematopoietic mechanisms through genome- 41. Olive M, et al. (1997) A dominant negative to activation protein-1 (AP1) that abolishes wide analysis of GATA factor chromatin occupancy. Mol Cell 36:667–681. DNA binding and inhibits oncogenesis. J Biol Chem 272:18586–18594. 13. Karin M, Liu Z, Zandi E (1997) AP-1 function and regulation. Curr Opin Cell Biol 9:240–246. 42. Gerdes MJ, et al. (2006) Activator protein-1 activity regulates epithelial tumor cell 14. Wagner EF, Eferl R (2005) Fos/AP-1 proteins in bone and the immune system. Immunol identity. Cancer Res 66:7578–7588. Rev 208:126–140. 43. Kawana M, Lee ME, Quertermous EE, Quertermous T (1995) Cooperative interaction 15. Pober JS, Sessa WC (2007) Evolving functions of endothelial cells in inflammation. Nat of GATA-2 and AP1 regulates transcription of the endothelin-1 gene. Mol Cell Biol 15: Rev Immunol 7:803–815. 4225–4231. 16. Grass JA, et al. (2006) Distinct functions of dispersed GATA factor complexes at an 44. Lee EG, et al. (2000) Failure to regulate TNF-induced NF-kappaB and cell death re- endogenous gene . Mol Cell Biol 26:7056–7067. sponses in A20-deficient mice. Science 289:2350–2354. 17. Wozniak RJ, Boyer ME, Grass JA, Lee Y-S, Bresnick EH (2007) Context-dependent 45. Cooper JT, et al. (1996) A20 blocks endothelial cell activation through a NF-kappaB- GATA factor function: Combinatorial requirements for transcriptional control in he- dependent mechanism. J Biol Chem 271:18068–18073. matopoietic and endothelial cells. J Biol Chem 282:14665–14674. 46. Natoli G (2010) NF-kappaB: No longer an island, but a piece of a continent. EMBO Rep 18. SenBanerjee S, et al. (2004) KLF2 Is a novel transcriptional regulator of endothelial 11:246–248. proinflammatory activation. J Exp Med 199:1305–1315. 47. Bailey TL, et al. (2009) MEME SUITE: Tools for motif discovery and searching. Nucleic 19. Lee JS, et al. (2006) Klf2 is an essential regulator of vascular hemodynamic forces Acids Res 37(Web Server issue):W202-8. in vivo. Dev Cell 11:845–857. 48. Song H, et al. (2009) Critical role for GATA3 in mediating Tie2 expression and function 20. Lin Z, et al. (2005) Kruppel-like factor 2 (KLF2) regulates endothelial thrombotic in large vessel endothelial cells. J Biol Chem 284:29109–29124. function. Circ Res 96:e48–e57. 49. Kanki Y, et al. (2011) Epigenetically coordinated GATA2 binding is necessary for 21. Wu J, Bohanan CS, Neumann JC, Lingrel JB (2008) KLF2 transcription factor modulates endothelium-specific endomucin expression. EMBO J 30:2582–2595. blood vessel maturation through smooth muscle cell migration. J Biol Chem 283:3942–3950. 50. Eferl R, Wagner EF (2003) AP-1: A double-edged sword in tumorigenesis. Nat Rev 22. Parmar KM, et al. (2005) Statins exert endothelial atheroprotective effects via the Cancer 3:859–868. KLF2 transcription factor. J Biol Chem 280:26714–26719. 51. Shaulian E, Karin M (2002) AP-1 as a regulator of cell life and death. Nat Cell Biol 4: 23. Wang LC, et al. (1997) Yolk sac angiogenic defect and intra-embryonic apoptosis in E131–E136. mice lacking the Ets-related factor TEL. EMBO J 16:4374–4383. 52. Kouros-Mehr H, et al. (2008) GATA-3 links tumor differentiation and dissemination in 24. Wang LC, et al. (1998) The TEL/ETV6 gene is required specifically for hematopoiesis in a luminal breast cancer model. Cancer Cell 13:141–152. the bone marrow. Genes Dev 12:2392–2402. 53. Zhang SJ, et al. (2008) Gain-of-function of GATA-2 in acute myeloid 25. Lin Q, et al. (1998) Requirement of the MADS-box transcription factor MEF2C for transformation of chronic myeloid leukemia. Proc Natl Acad Sci USA 105:2076–2081. vascular development. Development 125:4565–4574. 54. Im H, et al. (2005) Chromatin domain activation via GATA-1 utilization of a small 26. De Val S, et al. (2004) Mef2c is activated directly by Ets transcription factors through subset of dispersed GATA motifs within a broad chromosomal region. Proc Natl Acad an evolutionarily conserved endothelial cell-specific enhancer. Dev Biol 275:424–434. Sci USA 102:17065–17070. 27. Barski A, et al. (2007) High-resolution profiling of histone methylations in the human 55. Im H, et al. (2004) Measurement of protein-DNA interactions in vivo by chromatin genome. Cell 129:823–837. immunoprecipitation. Methods Mol Biol 284:129–146. 28. Santos-Rosa H, et al. (2002) Active genes are tri-methylated at K4 of histone H3. 56. O’Geen H, Frietze S, Farnham PJ (2010) Using ChIP-seq technology to identify targets Nature 419:407–411. of finger transcription factors. Methods Mol Biol 649:437–455. 29. Bembom O, Keles S, van der Laan MJ (2007) Supervised detection of conserved motifs 57. Blahnik KR, et al. (2010) Sole-Search: An integrated analysis program for peak de- in DNA sequences with cosmo. Stat Appl Genet Mol Biol 6(1): Article 8. tection and functional annotation using ChIP-seq data. Nucleic Acids Res 38:e13.

13646 | www.pnas.org/cgi/doi/10.1073/pnas.1108440108 Linnemann et al. Downloaded by guest on September 26, 2021