Small-Molecule Inhibitors of Acetyltransferase P300 Identified by High-Throughput Screening Are Potent Anticancer Agents

Published OnlineFirst April 26, 2013; DOI: 10.1158/1535-7163.MCT-12-0930

Molecular Cancer Chemical Therapeutics Therapeutics

Small-Molecule Inhibitors of Acetyltransferase p300 Identiﬁed by High-Throughput Screening Are Potent Anticancer Agents

Heng Yang1,6, Christie E. Pinello3, Jian Luo1, Dawei Li1,7, Yunfei Wang1,8, Lisa Y. Zhao1, Stephan C. Jahn2, Sanjay Adrian Saldanha3, Jamie Planck9, Kyla R. Geary9, Haiching Ma9, Brian K. Law2, William R. Roush4, Peter Hodder3,5, and Daiqing Liao1

Abstract Acetyltransferase p300 (KAT3B) plays key roles in signaling cascades that support cancer cell survival and sustained proliferation. Thus, p300 represents a potential anticancer therapeutic target. To discover novel anticancer agents that target p300, we conducted a high-throughput screening campaign. A library of 622,079 compounds was assayed for cytotoxicity to the triple-negative breast cancer (TNBC) cell line MDA-MB-231 but not to the human mammary epithelial cells. The resulting compounds were tested in a biochemical assay for m inhibiting the enzymatic activity of p300. One compound (L002, NSC764414) displayed an IC50 of 1.98 mol/L against p300 in vitro, inhibited acetylation of histones and p53, and suppressed STAT3 activation in cell-based assays. L002 could be docked to the active site of the p300 catalytic domain. Biochemical tests of a series of related compounds revealed functional groups that may impact inhibitory potency of L002 against p300. Interestingly, these analogs showed inhibitory activities against the cellular paralog of p300 (CBP), p300/CBP- associated factor, and GCN5, but not to other acetyltransferases (KAT5, KAT6B, and KAT7), histone deacetylases, and histone methyltransferases. Among the NCI-60 panel of cancer cell lines, leukemia and lymphoma cell lines were extremely sensitive to L002, whereas it is toxic to only a limited number of cell lines derived from solid tumors. Notably, breast cancer cell lines, especially those derived from TNBC, were highly susceptible to L002. In vivo, it potently suppressed tumor growth and histone acetylation of MDA-MB-468 xenografts. Thus, these new acetyltransferase inhibitors are potential anticancer therapeutics. Mol Cancer Ther; 12(5); 1–11. 2013 AACR.

Introduction

1 EP300 Authors' Affiliations: Departments of Anatomy and Cell Biology, Shands p300 (KAT3B) encoded by was originally iden- Cancer Center and Genetics Institute; 2Pharmacology and Therapeutics, tified as a binding protein of adenovirus E1A (1). It is a University of Florida College of Medicine, Gainesville; 3The Scripps Research Institute Molecular Screening Center, Lead Identification Divi- large protein with multiple domains that bind to diverse sion, Translational Research Institute; Departments of 4Chemistry, and proteins including many transcription factors (2). p300 is a 5Molecular Therapeutics, Scripps Florida, Jupiter, Florida; 6Department of lysine acetyltransferase that catalyzes the attachment of Hematology and Oncology, Beijing University of Chinese Medicine, Beijing; 7Department of Urology, Qilu Hospital, Shandong University, Jinan, Shan- an acetyl group to lysine residues of histones and other dong; 8Department of Biochemistry and Molecular Biology, College of Life cellular proteins such as p53 (3). p300-catalyzed acetyla- Sciences, Northwest Agriculture and Forestry University, Yangling, tion of histones and other proteins is pivotal to gene Shaanxi, China; and 9Reaction Biology Corp., Malvern, Pennsylvania activation (2). Heightened p300 expression and related Note: Supplementary data for this article are available at Molecular Cancer activities have been observed in advanced human malig- Therapeutics Online (http://mct.aacrjournals.org/). nancies, such as prostate (4, 5) and liver cancers (6, 7), and Current address for J. Luo: Department of General Surgery, Huazhong seem to be associated with the poor prognosis of these University of Science and Technology, Wuhan, Hubei, China; current address for L.Y. Zhao: Department of Medicine, University of Florida, cancer types. Elevated expression levels of p300 were also Gainesville, Florida; and current address for S.A. Saldanha: Lead Discov- observed in primary human breast cancers and in mouse ery, Bristol-Myers Squibb Company, 5 Research Parkway, Wallingford, mammary carcinomas induced by the polyomavirus Connecticut. middle-T oncogene (8). These observations suggest that H. Yang, C.E. Pinello, and J. Luo contributed equally to this work. p300 might be a potential therapeutic target for treating Corresponding Author: Daiqing Liao, Department of Anatomy and Cell cancer (9). Biology, University of Florida, 1333 Center Drive, Gainesville, Florida Importantly, p300 is a critical coactivator of several fl 100235. Phone: 352-273-8188; Fax: 352-392-3305; E-mail: dliao@u .edu oncogenic transcription factors, such as STAT3, NF-kB, doi: 10.1158/1535-7163.MCT-12-0930 and hypoxia-inducible factor-1a (HIF-1a). Genes regulated 2013 American Association for Cancer Research. by these transcription factors are involved in cytokine- or

www.aacrjournals.org OF1

Yang et al.

hypoxia-induced cancer cell survival and sustained further genetic authentication, their characteristic proliferation. Not only does p300 serve as their coactiva- morphologies and certain biochemical features were spe- tor, but STAT3 and NF-kB are also substrates of p300- cifically verified. For example, overexpression of EGF mediated acetylation (10, 11). Recent studies revealed that receptor in MDA-MB-468 was always observed. HTS p300-mediated acetylation of STAT3 at multiple sites is a assays were adapted to the 1,536-well format. In vitro prerequisite for its phosphorylation at Tyr705 by Janus- high-throughput p300 acetyltransferase activity assays activated kinases (JAK; ref. 12). Interestingly, Marotta and were done using a fluorescence assay. Furthermore, colleagues reported that the IL-6/JAK2/STAT3 pathway details of HTS assays are described in the Supplementary þ is preferentially active in CD44 /CD24 breast cancer Material and Methods. Radioactivity-based filter-binding stem cells (CSC) and is required for their growth. Inhibi- (HotSpot) histone acetyltransferase (HAT) assays, in tion of this pathway by a JAK2 inhibitor is effective in which [3H]-acetyl-CoA and histone H3 were used as killing CSCs and causing regression of xenografted substrates, were done by Reaction Biology Corporation. tumors (13). In addition, CD44 has been shown to undergo nuclear translocation. In the nucleus, CD44 mediates the Cell viability, cell cycle, and colony formation assays p300–STAT3 interaction for acetylating STAT3, thus pro- Cell viability and cell-cycle assays were done as moting its activation (14). STAT3 can further strengthen described (22). In colony formation assay, cells were seed- oncogenic signaling through activating NF-kB directly or ed in 6-well plates and exposed to various concentrations indirectly (14, 15). Aside from acetylation of substrates of L002 for 48 hours. The treated cells were trypsinized and that are directly involved in transcription, p300 acetylates 3,500 cells per well were reseeded. Cells were cultured proteins that impact metabolism (16), autophagy, (17) and with a medium containing L002 at the same concentrations motility (18). For example, p300 might contribute to as in the initial treatment. The medium with a proper metastasis through acetylation of the chaperone protein concentration of L002 was changed very 4 days until Hsp90 (18). These observations provide a strong rationale colonies appeared in about 2 weeks. Colonies were stained for targeting p300 as an anticancer therapeutic strategy. with 1% methylene blue. [3H]-thymidine incorporation Targeting p300 may offer additional advantages. Because assays were conducted essentially as described previously multiple growth factor receptors converge to activate (23). Briefly, HCT116 cells were treated for 24 hours with STAT3, NF-kB, HIF-1a, and other transcription factors various concentrations of L002. The cells were then pulse that recruit p300 to their target genes, inhibition of p300 labeled for 2 hours with [3H]-thymidine and the incorpo- may be more effective than suppression of receptor tyro- ration was quantitated and normalized to that of the sine kinases, as inhibition of one kinase often leads to the vehicle dimethyl sulfoxide (DMSO)-treated cells. activation of an alternative pathway that still permits cancer cell survival. Furthermore, chemical inhibition of Histone isolation and Western blotting p300 that possesses intrinsic enzymatic activity is more Cells grown in a complete medium including 10% feasible than blocking transcription factors with small bovine calf serum were exposed to various concentrations molecules, as discovery of chemical inhibitors of tran- of L002 for a specified period of time. One hour before cell scription factors has proven extremely challenging. harvest, cells were treated with DMSO or 0.2 mmol/L of To identify chemical inhibitors of p300 as potential trichostatin A (TSA). Histones were detected in total cell anticancer agents, we conducted a high-throughput lysates or as isolated forms using acid extraction method screening (HTS) campaign. The breast cancer cell line, as described previously (24). Conventional SDS-PAGE or MDA-MB-231, was used in our primary HTS assay cou- acid–urea gel electrophoresis (24) was used to separate pled with counterscreening against normal human mam- histones. Detection antibodies used in this study are mary epithelial cells (HMEC) to discover compounds that described in the Supplementary Materials and Methods. are only toxic to cancer cells, but benign to normal cells. MDA-MB-231 was derived from a breast cancer subtype In silico molecular docking known as triple-negative breast cancer (TNBC) that still L002 was docked to a crystal structure of the p300 has no effective treatments, and consequently carries poor catalytic domain (25). The details of the docking proce- prognosis (19, 20). p300 is expressed in MDA-MB-231 and dure are described in the Supplementary Materials and seems to play a critical role in driving its invasive growth Methods. (21). Here, we report the identification of L002 and its analogs as a new class of lead compounds that inhibit the In vivo efficacy of L002 against TNBC xenograft activity of p300 and related acetyltransferases with in vivo MDA-MB-468 cells (5 106)in50mL of PBS were mixed anticancer effects in a mouse xenograft model. with 50 mL Matrigel, which was injected into the flank of a female NU/NU mouse (5- to 7-week-old, Charles River). Materials and Methods Palpable tumors developed in 2 to 3 weeks. Tumor dimen- HTS assays sions were measured with a digital caliper. Tumor volume All cancer cell lines were obtained from American Type was calculated with the formula (W2 L) 0.5, where W Culture Collection, and HMECs were purchased from is the width and L the length of a tumor (W < L). When Invitrogen. Although cell lines were not subjected to tumor volumes reached approximately 100 mm3, mice

OF2 Mol Cancer Ther; 12(5) May 2013 Molecular Cancer Therapeutics

A New Class of Acetyltransferase Inhibitors

were randomized into 3 treatment cohorts (n ¼ 5): no Of the 4,272 compounds emerging from the primary/ treatment (group 1), DMSO (group 2), or L002 (group 3) at confirmation screens with the MDA-MB-231 cells, 640 0.5 mg in 100 mL DMSO per injection intraperitoneally compounds were chosen based on their selectivity twice weekly for 3 weeks. Tumor-bearing mice were against MDA-MB-231 cells in comparison with activity monitored for 2 additional weeks after treatment termi- versus HMECs for 10-point dose–response assays. The nation. Data are shown as means SEM. The 2-tailed 640 compounds were also tested for inhibiting p300 Student t test was used to compare differences between enzymatic activity in vitro using a fluorescence-based treatment groups, and the differences were considered assay (see Materials and Methods section and Supple- P < statistically significant if 0.05. Tumors were removed mentary Fig. S3). The CC50 cytotoxicity values of the 640 from euthanized mice at the end point of the animal compounds versus MDA-MB-231 in the presence or protocol, and sectioned for hematoxylin and eosin (H&E) absence of SAHA and versus HMEC as well as IC50 of staining and immunohistochemical analysis with an anti- p300 inhibition were determined. Of these 640 com- body against acetylated histone H4 (H4ac, Upstate 06-866, pounds, 10 exhibited selectivity against MDA-MB-231 m m 1:200 dilution). The animal protocol was approved by the (CC50 10 mol/L vs. MDA-MB-231 and 10 mol/ in vitro University of Florida Institutional Animal Care and Use L vs. HMEC), and inhibited p300 with IC50 10 Committee (Gainesville, Florida). mmol/L (Supplementary Fig. S4).

Results In vitro activity of L002 and its analogs HTS campaign for identifying p300 inhibitors as new One compound (L002, ChemBridge ID 6625948) exhib- anticancer agents ited potent inhibition of the p300 catalytic domain in vitro m Our goal was to identify novel anticancer compounds with an IC50 of 1.98 mol/L. Notably, at high concentra- through inhibiting p300. The primary screen identified tions, L002 completely suppressed the catalytic activity of compounds that are toxic to the TNBC cell line MDA-MB- p300, whereas anacardic acid, a known HAT inhibitor 231 but not to HMECs. These compounds were then tested (27), exerted only approximately 60% inhibition at 52 for their ability to inhibit p300 acetyltransferase activity in mmol/L (Fig. 1A). We have identified 7 additional com- an in vitro biochemical assay. Our original hypothesis was pounds structurally related to L002 (Fig. 1B). All these that inhibition of p300 might potentiate histone deacety- analogs inhibited p300 with varying potencies (Table 1). lase (HDAC) inhibitor (HDACi)-mediated cytotoxicity Independent biochemical experiments using radioisotope based on our initial observation that shRNA-mediated HAT activity assays confirmed the inhibitory effects of knockdown of p300 and MDA-MB-231 cells to HDACi- these compounds on p300 (Table 1). Among these com- induced cell death (data not shown). Thus, the primary pounds, L003, L004, and L007 were least potent, com- screen was conducted in the presence of 1.5 mmol/L of pared with L002 (Table 1), whereas other compounds SAHA (vorinostat), which induced 12% growth inhibition showed less than a 2-fold difference in IC50. Structural for MDA-MB-231 cells (CC12), to discover HDACi "poten- comparison of these analogs suggests that a functional tiators." Each compound in the library was screened as a (methyl, methoxyl, or bromide) group opposite to the single point, single dose at 6.2 mmol/L in the MDA-MB- sulfonyl moiety in the phenyl ring is largely unimportant 231 cytotoxicity assay with the presence of 1.5 mmol/L of or detrimental to p300 inhibition. Regardless, these results SAHA. The primary screen exhibited a Z0 score of 0.82. To reveal a new chemical scaffold that is not shared among identify nominally active compounds (hits), a standard known acetyltransferase inhibitors, which provides a hit-cutoff algorithm was used based on an average plus 3- starting point for further optimization to improve the fold SD calculation method (26). Thus, a compound was inhibitory potency and specificity of these compounds declared a hit if its inhibition exceeded 22.70%. This cut off against p300. yielded a total of 6,471 primary hits at a 1.04% hit-rate We determined the selectivity of these compounds (Supplementary Fig. S1). From these hits, 6,400 com- among various human acetyltransferases. Cellular para- pounds exhibiting the highest inhibition values were log of p300 (CBP), an acetyltransferase closely related to selected for secondary screening. Eight compounds were p300, was inhibited by some of these analogs, and in not available for further testing; therefore, 6,392 were general, they were less potent to CBP than to p300 included in subsequent experiments. These compounds (Table 1). L002 also inhibited p300/CBP-associated factor were retested in the potentiator confirmation assay in (PCAF) and GCN5, 2 proteins of the GNAT (GCN5-relat- the presence of 1.5 mmol/L of SAHA against MDA-MB- ed N-acetyltransferase) family, whereas other com- 231 in triplicate at a single dose (6.2 mmol/L). They were pounds were much less effective (Table 1). In contrast, also tested in counterscreens against MDA-MB-231 and these compounds displayed no inhibition against the HMECs without SAHA. From the set of hits, 4,272 com- MYST family of HATs [KAT5 (Tip60), KAT7 (MYST2)] pounds were confirmed to be active against MDA-MB-231 and KAT6B (MYST4), Table 1). cells, and the vast majority of these compounds exhibited To further assess target specificity of L002, we tested it similar inhibitory activities in assays in the presence or against a panel of HDACs and histone methyltransferas- absence of SAHA, suggesting that few potentiators could es (HMT). L002 did not inhibit HDAC1 (class I), HDAC6 be uncovered in our screen (Supplementary Fig. S2). (class IIb), and HDAC11 (class IV; Supplementary Fig. S5).

www.aacrjournals.org Mol Cancer Ther; 12(5) May 2013 OF3

Yang et al. ) 6 A 6 10 10 100 ND ND ND ND ND 100 IC =1.98 µmol/L ND 50 80 mol/L of IC50=4.15 µmol/L 75 60 4 50 40 (2.67 10 25

20 % Inhibition % Inhibition

0 0 ) 5 –9 –8 –7 –6 –5 –4 –9 –8 –7 –6 –5 –4 log [L002] (mol/L) log [AA] (mol/L) 10 ND ND ND ND ND ND

B H]-acetyl-CoA (3.08 3 (1.77 % activity at 1 – 26%) ND ND 5 5

IC =1.97 µmol/L (1%)

50 (18%) (13%) (24%) ( L001 10 10 L005 4 4 4 4 4 10 10 10 10 10 4.53 6.59 1 1 1 1 > 1 > > > >

IC50=1.98 µmol/L L002 L006 34%) 4 5 (4%) (1%) (17%) (30%) ( 10 10 4 4 uorescence assays are provided in parentheses. 4 4 4 ﬂ 10 10 10 10 10 7.5 1.96 1 1 1 1 > > 1 > >

L007 > L003 IC50=17.7 µmol/L (mol/L) mol/L using the HotSpot HAT assays, in which [ 4 50 curve, percent inhibition [100% (DMSO control)] ) 10 4%) 50 4 5 4 (0%) (28%) (28%) (11%) ( 10 10 10 4 4 4 4 4 values derived from 10 10 10 10 10 50

L004 L008 IC =19.4 µmol/L IC =2.67 µmol/L Compound IC

50 50 2.28 1 (1.94 1 1 1 1 > > > > > tted to an IC ﬁ

Figure 1. Identiﬁcation of L002 and its analogs as p300 inhibitors. A, dose– response curves of p300 inhibition by L002 (black dots; left), and 2%) 38%) 18%)

ﬁ 4 (8%)(9%) 1.2

anacardic acid (AA, right). Puri ed recombinant p300 catalytic domain (22%) ( ( (

m 10 was incubated in a solution containing 50 mol/L of acetyl-CoA and a 4 4 4 4 4 4 10 10

histone H3 N-terminal peptide in the presence of L002 or anacardic acid 10 10 10 10

fi 2.6 at a speci ed concentration. CoASH, a product of the acetylation 1 1 1 1 > > 1 fi fl 1 >

reaction, was detected and quanti ed through a uorescence readout. In > > > separate assays, L002 was added after the completion of the acetylation reaction and the fluorogenic adduct formation to assess whether L002 would quench fluorescence. No quench was observed (gray triangles, ) 28%) left). B, the chemical structures of L002 and its analogs are shown. L001, 4 6 5 5 (2%) (1%) (32%) ( 10 10 10 10 L002, L004, L007, and L008 were distinct hits identified in HTS assays. mode with a 3-fold serial dilution starting at 1 4 4 4 50 L003, L005, and L006 were identified on the basis their structural 4 10 10 10 10 similarity to L002. Their inhibitory activities to acetyltransferases were 1.28 3.47 1 1 (1.98

in vitro 1 > tested (see Table 1). > 1 > >

Similarly, L002 did not display inhibitory effects against a ) 46%) 21%) 6 5 5 5 (4%) (25%) 3.39 panel of 8 diverse HMTs (DOT1, EZH1, G9a, PRMT1, ( ( 10 10 10 10 4 4 SETD2, SET7-9, SMYD2, and SUV39H2; see Supplemen- 4 4 10 10 10 10

tary Fig. S6). Of note, L002 was not ﬂagged as a promis- 9.27 1 (1.97 1 > 1 1

cuous inhibitor in HTS assays. It emerged as a hit only in > > > this HTS campaign among 25 HTS campaigns. Thus, L002 is a speciﬁc inhibitor of acetyltransferases.

In silico docking of L002 to the catalytic domain of Inhibitory potency of L002 and analogs against acetyltransferases p300 To assess whether L002 could dock to the active site of p300, we ﬁtted L002 to a crystal structure of the p300 Table 1. NOTE: The compounds were tested in a 10-dose IC mol/L) and histone H3 were used as substrates. When compound activity could not be MYST4 (KAT6B) GCN5 (KAT2A) TIP60 (KAT5) MYST2 (KAT7) CBP (KAT3A) 7.74 Acetyltransferasep300 (KAT3B) L001 3.7 L002 L003 L004 L005 L006 L007 L008 PCAF (KAT2B) Abbreviation: ND, not determined. a test compound is shown in parentheses. For L001, L002, L004, L007, and L008, the IC catalytic domain (25). In an energy-minimized model of

OF4 Mol Cancer Ther; 12(5) May 2013 Molecular Cancer Therapeutics

A New Class of Acetyltransferase Inhibitors

ABC K1456 K1456 K1456

Q1455 R1462 Q1455 Q1455 R1462 R1462

R1410 R1410 R1410 W1466 W1466 W1466

Figure 2. Docking of L002 to the acetyl-CoA–binding pocket in the p300 catalytic domain. L002 was fitted to the acetyl-CoA–binding pocket of a crystal structure of the p300 catalytic domain. Potential hydrogen bonds between L002 and the indicated residues of p300 are depicted (dotted lines). The residues of interest are labeled in white. A, a cartoon representation of a model of L002–p300 interaction. B, a space-filled representation of the p300 catalytic domain in complex with the inhibitor. C, the L002–p300 complex in a space-filled model. L002 and the p300 catalytic domain are shown in dark and light gray, respectively. the L002–p300 complex, an oxygen atom in the sulfonyl hydrogen bonds with the recently identified p300 inhib- moiety of L002 potentially forms 2 hydrogen bonds with itor C646 (28). In addition, a possible hydrogen bond the side-chain of R1410 in helix a3 of the p300 catalytic might also form between the oxygen linked to the nitrogen domain (Fig. 2). R1410 of p300 was shown to form hydro- atom of the imine bond in L002 and the side-chain of gen bonds with the synthetic bisubstrate p300 inhibitor Q1455 from the L1 loop of the p300 catalytic domain (Fig. Lys–CoA in its cocrystal structure with the p300 catalytic 2). In general, L002 fits nicely into the acetyl-CoA–binding domain (25), and was also proposed to form similar pocket (Fig. 2B and C), making key contacts with residues

L002 A (0.3 µmol/L)

kDa NT DMSO TSA DMSO TSA H3ac B L002 C 17 3 µmol/L 10 µmol/L 30 µmol/L L002 L004 H4ac µmol/L

NT 0.1 1.0 10 30 17 kDa DMSO 0.1 1.0 10 30 80 NT DMSO Etop 2 µmol/L Etop 10 µmol/L DMSO Etop 2 µmol/L Etop 10 µmol/L DMSO Etop 2 µmol/L Etop 10 µmol/L DMSO Etop 2 µmol/L Etop 10 µmol/L p-STAT3 H4ac kDa 46 p53 80 30 p-STAT3 (long) PCNA K382ac 25 80 46 p53 STAT3 p-S392 175 175 80 JAK2 46 p53 58 46 50 Hsp60 25 p21 30 30 25 PCNA 30 PCNA 12345678910 Colloidal blue stain Colloidal blue 17 1 2345678 910111213 12345

Figure 3. Inhibition of p300-dependent functions in cells. A, MDA-MB-468 cells were untreated (NT, lane 1), treated with DMSO (lane 2), TSA (lane 3), L002 (lane 4), and L002 plus TSA (lane 5). The cells cultured in a complete medium with 10% bovine calf serum were exposed to DMSO or L002 for 7 hours. TSA was added to 0.2 mmol/L at 1 hour before lysing cells for Western blotting. A blot was probed with an antibody against acetylated lysine (top) or acetylated histone H4 (H4ac, middle). The blot was reprobed with an anti-PCNA antibody as a loading control. In a separate gel, an equal amount of the samples was loaded and stained with colloidal blue (bottom). B, HCT116 cells were exposed to DMSO or L002 as indicated for 7 hours. Etoposide was added 1 hour after the addition of L002. Cells were lysed for Western blotting with the indicated antibodies. C, MIA PaCa-2 cells were exposed to the indicated concentrations of L002 or L004 for 6 hours before harvesting cells for Western blotting using the indicated antibodies.

www.aacrjournals.org Mol Cancer Ther; 12(5) May 2013 OF5

Yang et al.

in helices a3, a4, and the L1 loop. L002 could also be MIA PaCa-2, in which STAT3 is constitutively active docked to the acetyl-CoA–binding pockets of other acet- (31), was exposed to varying doses of L002 or L004. As yltransferases with the docking scores of 47.3 (PCAF), shown in Fig. 3C, the levels of STAT3 phosphorylated at 43 (p300), 34.5 (KAT5/Tip60), and 17 (MYST1; the Y705 were reduced in a concentration-dependent man- lower the scores, the higher the possible inhibitor–enzyme ner by both L002 and L004, and the former seemed more affinity). These docking scores correlate well with the potent. Collectively, these lines of evidence provide inhibition data shown in Table 1. strong support for our hypothesis that inhibition of p300 is one specific cellular mechanism of action by L002 inhibits p300-mediated acetylation in cell- L002. based assays HTS hit compounds were tested for their effects on Cytotoxicity of L002 to cancer cell lines p300-mediated cellular mechanisms. In TNBC cell line The differential cytotoxic effects of L002 to HMECs and MDA-MB-468, L002 markedly suppressed acetylation of MDA-MD-231 cells, as determined in the HTS assays, histone H3 (H3ac) and H4 (H4ac; Fig. 3A). Inhibition of suggest that it selectively kills cancer cells. To further histone acetylation at various lysine residues by L002 was validate cytotoxicity of L002 to cancer cell lines, we tested also observed in other cancer cell lines including MDA- the sensitivity of a panel of cancer cell lines derived from MB-231 and HCT116 (Supplementary Fig. S7). Together, diverse types of solid cancer to L002. As shown in Table 2 these cell-based assays provide a validation for the inhib- and Fig. 4, these cell lines exhibited varying sensitivities to itory effects of L002 on histone acetylation mediated by L002. Notably, all 4 tested TNBC cell lines were highly p300 and other acetyltransferases. susceptible to treatment with L002 with CC50 at low In addition to histones, p300 acetylates numerous other micromolar concentrations, whereas cell lines of luminal proteins. p53 is specifically acetylated at Lys382 by p300 subtype breast cancer were more resistant. We further (3). Acetylation at this site is elevated in response to DNA examined the impact of L002 treatment on clonal growth damage and other cellular stresses (29). To assess whether L002 could inhibit acetylation of p53 at Lys382 (K382ac), HCT116 cells expressing wt p53 were exposed to L002, the Table 2. CC50 values of L002 against HMEC genotoxic drug etoposide or a combination thereof. As and various cancer cell lines shown in Fig. 3B, K382ac was readily detectable in cells exposed to 10 mmol/L etoposide (lane 4). However, L002 Cell line Tissue/cancer type CC50 (mmol/L) markedly reduced the level of K382ac in a dose-depen- > a dent manner (compare lanes 7, 10, and 13 with 4), suggest- HMEC Normal 15.53 ing again that the L002 specifically targets p300. MDA-MB-231 TNBC 1.03 Genotoxic stress also induces p53 phosphorylation. MDA-MB-468 TNBC 2.57 MDA-MB-436 TNBC 2.85 Indeed, phosphorylation of p53 at Ser392 (p-S392), which b is mediated by several kinases including PKR, p38, and MDA-MB-435S TNBC 1.91 CK2/FACT (30), was markedly elevated in HCT116 cells MCF7 Breast cancer (luminal) 21.67 exposed to etoposide (Fig. 3B). L002 did not exert any ZR-75-1 Breast cancer (luminal) 8.91 obvious inhibition of p53 Ser392 phosphorylation even at SK-BR-3 Breast cancer (luminal) 11.6 30 mmol/L when cells were exposed to 2 mmol/L of LNCaP Prostate cancer 3.54 etoposide (compare lanes 6, 9, and 12 with 3 in the p53 PC-3 Prostate cancer 4.74 p-S392 panel). In cells exposed to 10 mmol/L of etoposide, DU 145 Prostate cancer 3.75 slight reduction of p53 p-S392 levels was noted when cells MIA PaCa-2 Pancreatic cancer 21.7 > a were also treated with 10 or 30 mmol/L of L002 (compare HPAC Pancreatic cancer 30 lanes 10 and 13 with 4). This is likely due to the reduced BxPC3 Pancreatic cancer 18.56 levels of total p53 as dose-dependent suppression of PANC-1 Pancreatic cancer 8.44 etoposide-induced p53 accumulation by L002 was clearly HCT 116 Colon cancer 7.40 observed (see the p53 panel in Fig. 3B). These data thus DLD-1 Colon cancer 8.35 – > a suggest that L002 did not inhibit kinases that phosphor- A549 Non small cell lung 30 ylate p53 S392; however, inhibition of p53 acetylation by cancer L002 seems to compromise p53 protein stability, support- SH-SY5Y Neuroblastoma 0.19 ing the notion that acetylation antagonizes ubiquitin- aCell growth inhibition did not surpass 50% at the highest mediated p53 degradation. Nonetheless, L002 did not concentration tested. affect etoposide-induced p21 expression despite attenu- bAlthough this cell line was isolated from the pleural effusion ated p53 acetylation at Lys382 (Fig. 3B). of a patient with breast cancer and has a gene expression To further understand the cellular mechanism of profile consistent with a breast cancer origin, some studies action for L002, we examined the influence of L002 on showed that it exhibits properties indicative of a melanoma STAT3 activation, which requires p300-mediated acet- origin. ylation of STAT3 (12). The pancreatic cancer cell line

OF6 Mol Cancer Ther; 12(5) May 2013 Molecular Cancer Therapeutics

A New Class of Acetyltransferase Inhibitors

AB 120 [L002] (µmol/L) DMSO 1 µmol/L 5 µmol/L 10 µmol/L 100

80 MDA-MB-231 60

40 MCF7 Cell survival (% of control) 20

0 1.4 4.7 14 28 0.05 0.14 0.47 Control [L002] (µmol/L) MCF7 MDA-MB-231

C 120 MDA-MB-231 HCT116 ([L002]: 30 µmol/L) 100 100

80 No washout 80 60 60

Cell survival 24-hour (% of control) 40 washout 40 Cell survival 20 (% of control) 20 0 1 3 10 30 0 Control [L002] (µmol/L) No washout 24 h washout

Figure 4. In vitro cell growth suppression by L002. A, MDA-MB-231 and MCF7 cells were cultured and exposed to DMSO (control) or the indicated concentrations of L002. The relative abundance of viable cells was determined. Shown are the average values of 3 assays along with SD. B, L002 suppressed clonogenic growth of breast cancer cell lines. The cells were exposed to DMSO or the indicated concentrations of L002. Surviving colonies were stained with methylene blue. The images shown are representative of 3 experiments. C, reversibility of L002-mediated cytotoxicity. MDA-MB-231 or HCT116 cells were exposed to the indicated concentrations of L002. At 24 hours of its addition to a cell culture, L002 was washed out in one set of the experiments as indicated. Cell viability was determined at 96 hours after L002 addition. of cancer cell lines. As shown in Fig. 4B, in the presence of percentage of cells in the S-phase. The proportion of cells m 1 mol/L of L002, the clonal growth of MDA-MB-231 cells with sub-G1 DNA content, indicative of cells undergoing was completely suppressed, whereas colonies of MCF7 apoptosis, was higher in MDA-MB-231 cells exposed to cells were largely unaffected, compared with the control L002 compared with those treated with DMSO. In con- (DMSO-treated cells). For MCF7 cells, a significant num- trast, anacardic acid had no effects on cell-cycle progres- ber of colonies survived at 5 mmol/L of L002, although no sion of MDA-MB-231 cells (Fig. 5A). In HCT116 cells, L002 colonies were detected at 10 mmol/L (Fig. 4B). Inhibition potently inhibited DNA replication, as measured in thy- of cell growth seemed reversible, as drug washout per- midine incorporation assays (Fig. 5B). Taken together, mitted recovery of treated cells (Fig. 4C). Nonetheless, a these results indicate that L002 can induce growth arrest significant fraction of MDA-MB-231 cells was already and apoptosis in cancer cells to exert antiproliferative killed within 24 hours of drug exposure (Fig. 4C). effects. Among these cell lines from solid tumors, SH-SY5Y (neuroblastoma) was most sensitive, whereas A549 In vivo anticancer efficacy of L002 (non–small cell lung cancer) and MIA PaCa-2 (pancre- To test whether L002 could suppress tumor growth in atic cancer) were highly resistant to L002 (Table 2). L002 vivo, L002 was administrated intraperitoneally to mice was further tested against the NCI-60 panel of human bearing tumor xenografts of the TNBC cell line MDA- cancer cell lines. Strikingly, diverse leukemia and lym- MB-468. Figure 6A shows that L002 effectively sup- phoma cell lines were all extremely sensitive to L002 pressed tumor growth during systemic treatment and (Supplementary Table S1). Among the solid tumor cell importantly, tumors did not grow back after treatment lines, most of the breast cancer cell lines including termination. The twice-weekly dosing regimen was well MDA-MB-231, MCF7 and MDA-MB-435 were sensitive tolerated, as the change of body weight was within 10% to L002. In contrast, cell lines of non–small cell lung and (Fig.6B).Tumorsectionsfrom mice treated with DMSO central nervous system cancer were largely insensitive or L002 were subjected to immunohistochemical stain- to L002. ing with an antibody against H4ac. As shown in Fig. 6C, To explore the potential mechanism by which L002 kills the levels of H4ac were markedly lower in tumors from cancer cells, we assessed effects of L002 treatment on cell- micetreatedwithL002incomparisonwiththosefrom cycle profiles of MDA-MB-231 cells. As shown in Fig. 5A, mice treated with DMSO. Altogether, these data repre- compared with treatment with DMSO, exposure of MDA- sent the first in vivo proof-of-principle regarding the MB-231 cells to L002 increased cell numbers in the G1 feasibility of acetyltransferase inhibition for treating phase of the cell cycle with a concomitant reduction of the solid malignancy.

www.aacrjournals.org Mol Cancer Ther; 12(5) May 2013 OF7

Yang et al.

potent p300 inhibitor reported thus far, C646, inhibits A p300 at submicromolar potency (28). Inhibition of the MDA-MB-231 80 * enzymatic activity of p300 by C646 has been shown in diverse applications (36, 37). Nonetheless, C646 seems to 60 be inactivated in the presence of serum (38). Furthermore, DMSO the requirement of the nitroaromatic moiety for C646 to in vivo 40 L002 inhibit p300 (28) might limit its application due to % Cells AA hepatotoxicity associated with nitroaromatic compounds 20 * * (39). Thus, novel and potent chemical inhibitors of p300 ns could broaden the appeal of such compounds for further 0 preclinical and clinical evaluation. G SG–M Apoptosis 1 2 L002 is structurally distinct from any known HAT *,P ≤ 0.05 ns: Not significant inhibitors. The 2 ring moieties (quinone imine and methyoxyphenyl group) are connected by the sulfonyl B HCT116 120 group. The 2 rings are not in the same plane, and this 3- dimensional arrangement seems to ﬁt well in the acetyl- 100 CoA pocket of the p300 catalytic domain (Fig. 2). Initial 80 SAR assessment suggests that the absence of the oxygen atom in the methyoxyl group enhances the inhibitory 60 potency (Table 1). Furthermore, replacement of the

(% control) 40 methyoxy group with a bromine atom in L002 also

20 markedly reduced the inhibitory potency against p300 Thymidine incorporation Thymidine (Table 1). 0 Interestingly, L002 and its analogs displayed differ- 0 0.156 0.313 0.625 1.25 2.5 5 10 ential inhibition to p300/CBP and other acetyltransfera- [L002] (µmol/L) ses. Most of the 8 analogs showed potent inhibition to p300, whereas only L001 and L002 detectably inhibited Figure 5. L002 induces cell-cycle arrest and apoptosis. A, MDA-MB-231 PCAF and GCN5 (Table 1). In contrast, these com- cells were exposed to DMSO, 10 mmol/L of L002, or anacardic acid for 24 pounds did not inhibit the MYST family of acetyltrans- fl hours. Cells were then processed for ow cytometry analysis. Percentage ferases (Table 1). Furthermore, L002 did not inhibit of cells in different phases of the cell cycle as well as cells with sub-G1 DNA content (apoptotic cells) is plotted. The P value was calculated on HDACs and a panel of diverse HMTs (Supplementary the basis of a Student t test. B, HCT116 cells were exposed to DMSO or Figs. S5 and S6). Although not extensively tested against the indicated concentrations of L002 for 24 hours. Percent thymidine kinases, L002 did not seem to impact p53 phosphoryla- incorporation is plotted against L002 concentration. tion (Fig. 3B). The demonstrated selectivity of these compounds to different classes of acetyltransferases per- haps reflects their structural differences. Although these Discussion enzymes have divergent amino acid sequences, their Here, we reported the discovery of a new class of catalytic domains share structural similarity in the cen- inhibitors of acetyltransferases. These compounds share tral core associated with acetyl-CoA binding, but they the same core chemical scaffold and potently inhibited also have pronounced differences (25). Structure-guided p300 and related acetyltransferases in vitro and in cells. A modifications of these analogs will likely improve selec- number of chemical inhibitors of p300 and CBP have been tivity and potency against acetyltransferases. identified. These compounds include natural products Targeting p300 might have therapeutic implications in a curcumin (32), garcinol (33), and anacardic acid (27). range of different diseases including heart malfunction These natural products exhibited only moderate inhibi- (40), diabetes mellitus (41), and HIV infection (42), prob- tory potency and structure—activity relationship (SAR)- ably due to the fact that p300 is a pleiotropic protein that is based approaches to improve selectivity and potency involved in diverse biologic mechanisms ranging from are quite challenging due to their complex chemical cell survival, proliferation, metabolism, and viral infec- structures. A HTS of 69,000 compounds identified iso- tion. More than 400 protein-binding partners have been thiazolones as inhibitors of PCAF and other acetyltrans- described for p300/CBP, making them among the most ferases (34). These compounds seem to form covalent connected interaction "hubs" in cells (2). p300 and CBP are adducts with their targets and seem to induce irreversible found in mammals and Drosophila, and their placement in inhibition and toxicity (34), thus limiting their potential the interaction hubs indicates that they play diverse roles applications. A series of analogs of the synthetic bisub- in a multicellular organism, some of which may be essen- strate HAT inhibitor Lys–CoA have been described. These tial for cell and organism viability (2). Indeed, homozy- synthetic compounds are highly potent in inhibiting acet- gous knockout of either p300 or CBP as well as their yltransferases (35), but they have relatively large molec- compound heterozygous knockout causes embryonic ular weights and are poorly cell permeable. The most lethality in mice (2, 43, 44). These observations indicate

OF8 Mol Cancer Ther; 12(5) May 2013 Molecular Cancer Therapeutics

A New Class of Acetyltransferase Inhibitors

AB12 600

500 8 ) 3

400 4 No treatment No treatment 300 DMSO DMSO 0 L002 200 L002

Tumor volume (mm volume Tumor -4 100 ** ** ** ** ** **,P < 0.01 Body weight change (% ± SEM) 0 -8 70 14 21 28 31 70 14 21 3128 Days Days

C DMSO L002 DMSO L002 DMSO L002 DMSO L002

100 µm 100 µm 100 µm 100 µm 100 µm 100 µm 50 µm 50 µm H&E H4ac

Figure 6. In vivo anticancer efficacy of L002. Mice bearing MDA-MB-468 xenografts were untreated or treated with the vehicle (DMSO) or L002. Each treatment group consisted of 5 mice (n ¼ 5). The vehicle or L002 (0.5 mg/injection) was injected intraperitoneally twice weekly for 3 weeks. Tumor volumes (A) and the percentage change of body weights (B) of tumor-bearing mice along with the SEM are plotted. The arrow in the left graph denotes the treatment end point. C, tumor sections of mice treated with DMSO or L002 were subjected to H&E staining or immunohistochemical analysis with an anti-H4ac antibody. Three different magnifications of the tumor sections stained with the anti-H4ac antibody are shown. that pharmacologic inhibition of p300/CBP might cause lymphoma. Significantly, agents targeting p300 acetyl- unintended toxicity. Thus, therapeutic strategies based on transferase activity were effective in suppressing acute targeting p300/CBP might be more applicable against myelogenous leukemia in an ex vivo treatment approach diseases that rely on excessive p300/CBP activity. Our using a mouse model (36). Because systemic administra- data showed that TNBC cell lines are highly susceptible tion of L002 efficiently suppressed tumor growth in vivo to L002 (Table 2). In vivo, L002 potently suppressed (Fig. 6), it is feasible to vigorously test the efficacy of L002 tumor growth of TNBC cell line MDA-MB-468 xenografts for treating hematologic malignancies using various ani- (Fig. 6). This represents the first demonstration of in vivo mal models. anticancer efficacy of an acetyltransferase inhibitor using systemic drug administration. Precisely why L002 Disclosure of Potential Conflicts of Interest is more potent in killing TNBC cells requires further No potential conflicts of interest were disclosed. investigation; however, the importance of p300 in pathways that sustain CSCs in TNBC provides a plausible Authors' Contributions Conception and design: H. Yang, C.E. Pinello, J. Luo, P. Hodder, D. Liao explanation. In particular, because of the critical role of Development of methodology: H. Yang, C.E. Pinello, J. Luo, L.Y. Zhao, p300 in the JAK/STAT (45) and Wnt/b-catenin pathways P. Hodder, D. Liao (46–48) that underpin CSC self-renewal, survival, and Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): H. Yang, C.E. Pinello, J. Luo, S.C. Jahn, S.A. proliferation (15), chemical inhibitors of p300 might be Saldanha, J. Planck, K.R. Geary, H. Ma, B.K. Law, P. Hodder, D. Liao more effective against TNBC and other cancers that are Analysis and interpretation of data (e.g., statistical analysis, biostatis- still refractory to currently available chemotherapies and tics, computational analysis): H. Yang, C.E. Pinello, J. Luo, S.C. Jahn, W.R. Roush, P. Hodder, D. Liao targeted therapies. Writing, review, and/or revision of the manuscript: H. Yang, C.E. Pinello, Notably, among the NCI-60 panel of cancer cell lines, J. Luo, S.C. Jahn, B.K. Law, W.R. Roush, P. Hodder, D. Liao Administrative, technical, or material support (i.e., reporting or orga- those derived from hematologic malignancies were nizing data, constructing databases): H. Yang, C.E. Pinello, J. Luo, D. Li, extremely sensitive to L002 (Supplementary Table S1). Y. Wang, W.R. Roush, P. Hodder, D. Liao This seems to mirror the more pronounced vulnerability Study supervision: J. Luo, P. Hodder, D. Liao of leukemias and lymphomas to agents targeting epige- netic pathways than solid cancers (49). Indeed, HDAC Acknowledgments The authors thank Pierre Baillargeon, Lina DeLuca, and Louis Scam- inhibitors SAHA (vorinostat) and romidepsin (ISTODAX) pavia for compound management and quality control, Katharine Emery are already in clinical use for treating certain types of (Scripps Florida) for secretarial assistance, Jennifer Liao for reading the

www.aacrjournals.org Mol Cancer Ther; 12(5) May 2013 OF9

Yang et al.

article, and the NCI Developmental Therapeutics Program for the NCI-60 The costs of publication of this article were defrayed in part by the cell line testing. payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate Grant Support this fact. The work was supported by a grant from Bankhead-Coley Cancer Research Program, Florida Department of Health (09BW-05 to D. Liao). H. Yang, D. Li, and Y. Wang received scholarships from the China Received September 21, 2012; revised March 6, 2013; accepted March 6, Scholarship Council. 2013; published OnlineFirst April 26, 2013.

References 1. Eckner R, Ewen ME, Newsome D, Gerdes M, DeCaprio JA, Lawrence 20. Lehmann BD, Bauer JA, Chen X, Sanders ME, Chakravarthy AB, Shyr JB, et al. Molecular cloning and functional analysis of the adenovirus Y, et al. Identification of human triple-negative breast cancer subtypes E1A-associated 300-kD protein (p300) reveals a protein with proper- and preclinical models for selection of targeted therapies. J Clin Invest ties of a transcriptional adaptor. Genes Dev 1994;8:869–84. 2011;121:2750–67. 2. Bedford DC, Kasper LH, Fukuyama T, Brindle PK. Target gene context 21. Chan CH, Lee SW, Li CF, Wang J, Yang WL, Wu CY, et al. Deciphering influences the transcriptional requirement for the KAT3 family of CBP the transcriptional complex critical for RhoA gene expression and and p300 histone acetyltransferases. Epigenetics 2010;5:9–15. cancer metastasis. Nat Cell Biol 2010;12:457–67. 3. Gu W, Roeder RG. Activation of p53 sequence-specific DNA binding 22. Li Q, Zhao LY, Zheng Z, Yang H, Santiago A, Liao D. Inhibition of by acetylation of the p53 C-terminal domain. Cell 1997;90:595–606. p53 by adenovirus type 12 E1B-55K deregulates cell cycle control 4. Heemers HV, Debes JD, Tindall DJ. The role of the transcriptional and sensitizes tumor cells to genotoxic agents. J Virol 2011;85: coactivator p300 in prostate cancer progression. Adv Exp Med Biol 7976–88. 2008;617:535–40. 23. Corsino P, Horenstein N, Ostrov D, Rowe T, Law M, Barrett A, et al. A 5. Isharwal S, Miller MC, Marlow C, Makarov DV, Partin AW, Veltri RW. novel class of cyclin-dependent kinase inhibitors identified by molec- p300 (histone acetyltransferase) biomarker predicts prostate cancer ular docking act through a unique mechanism. J Biol Chem 2009;284: biochemical recurrence and correlates with changes in epithelia nucle- 29945–55. ar size and shape. Prostate 2008;68:1097–104. 24. Shechter D, Dormann HL, Allis CD, Hake SB. Extraction, purification 6. Yokomizo C, Yamaguchi K, Itoh Y, Nishimura T, Umemura A, Minami and analysis of histones. Nat Protoc 2007;2:1445–57. M, et al. High expression of p300 in HCC predicts shortened overall 25. Liu X, Wang L, Zhao K, Thompson PR, Hwang Y, Marmorstein R, et al. survival in association with enhanced epithelial mesenchymal transi- The structural basis of protein acetylation by the p300/CBP transcription of HCC cells. Cancer Lett 2011;310:140–7. tional coactivator. Nature 2008;451:846–50. 7. Li M, Luo RZ, Chen JW, Cao Y, Lu JB, He JH, et al. High expression of 26. Hodder P, Cassaday J, Peltier R, Berry K, Inglese J, Feuston B, et al. transcriptional coactivator p300 correlates with aggressive features Identification of metabotropic glutamate receptor antagonists using an and poor prognosis of hepatocellular carcinoma. J Transl Med automated high-throughput screening system. Anal Biochem 2003; 2011;9:5. 313:246–54. 8. Fermento ME, Gandini NA, Lang CA, Perez JE, Maturi HV, Curino AC, 27. Balasubramanyam K, Swaminathan V, Ranganathan A, Kundu TK. et al. Intracellular distribution of p300 and its differential recruitment to Small molecule modulators of histone acetyltransferase p300. J Biol aggresomes in breast cancer. Exp Mol Pathol 2010;88:256–64. Chem 2003;278:19134–40. 9. Dekker FJ, Haisma HJ. Histone acetyl transferases as emerging drug 28. Bowers EM, Yan G, Mukherjee C, Orry A, Wang L, Holbert MA, et al. targets. Drug Discov Today 2009;14:942–8. Virtual ligand screening of the p300/CBP histone acetyltransferase: 10. Yuan ZL, Guan YJ, Chatterjee D, Chin YE. Stat3 dimerization regulated identification of a selective small molecule inhibitor. Chem Biol 2010; by reversible acetylation of a single lysine residue. Science 2005; 17:471–82. 307:269–73. 29. Ito A, Lai CH, Zhao X, Saito S, Hamilton MH, Appella E, et al. p300/ 11. Chen LF, Mu Y, Greene WC. Acetylation of RelA at discrete sites CBP-mediated p53 acetylation is commonly induced by p53- regulates distinct nuclear functions of NF-kappaB. EMBO J 2002;21: activating agents and inhibited by MDM2. EMBO J 2001;20: 6539–48. 1331–40. 12. Nie Y, Erion DM, Yuan Z, Dietrich M, Shulman GI, Horvath TL, et al. 30. Bode AM, Dong Z. Post-translational modification of p53 in tumori- STAT3 inhibition of gluconeogenesis is downregulated by SirT1. Nat genesis. Nat Rev Cancer 2004;4:793–805. Cell Biol 2009;11:492–500. 31. Corcoran RB, Contino G, Deshpande V, Tzatsos A, Conrad C, Benes 13. Marotta LL, Almendro V, Marusyk A, Shipitsin M, Schemme J, Walker CH, et al. STAT3 plays a critical role in KRAS-induced pancreatic SR, et al. The JAK2/STAT3 signaling pathway is required for growth of tumorigenesis. Cancer Res 2011;71:5020–9. CD44CD24 stem cell-like breast cancer cells in human tumors. J Clin 32. Balasubramanyam K, Varier RA, Altaf M, Swaminathan V, Siddappa Invest 2011;121:2723–35. NB, Ranga U, et al. Curcumin, a novel p300/CREB-binding protein- 14. Lee JL, Wang MJ, Chen JY. Acetylation and activation of STAT3 specific inhibitor of acetyltransferase, represses the acetylation of mediated by nuclear translocation of CD44. J Cell Biol 2009;185: histone/nonhistone proteins and histone acetyltransferase-dependent 949–57. chromatin transcription. J Biol Chem 2004;279:51163–71. 15. Korkaya H, Liu S, Wicha MS. Regulation of cancer stem cells by 33. Balasubramanyam K, Altaf M, Varier RA, Swaminathan V, Ravindran A, cytokine networks: attacking cancer's inflammatory roots. Clin Cancer Sadhale PP, et al. Polyisoprenylated benzophenone, garcinol, a Res 2011;17:6125–9. natural histone acetyltransferase inhibitor, represses chromatin tran- 16. Lin YY, Kiihl S, Suhail Y, Liu SY, Chou YH, Kuang Z, et al. Functional scription and alters global gene expression. J Biol Chem 2004;279: dissection of lysine deacetylases reveals that HDAC1 and p300 reg- 33716–26. ulate AMPK. Nature 2012;482:251–5. 34. Stimson L, Rowlands MG, Newbatt YM, Smith NF, Raynaud FI, Rogers 17. Lee IH, Finkel T. Regulation of autophagy by the p300 acetyltransfer- P, et al. Isothiazolones as inhibitors of PCAF and p300 histone ase. J Biol Chem 2009;284:6322–8. acetyltransferase activity. Mol Cancer Ther 2005;4:1521–32. 18. Yang Y, Rao R, Shen J, Tang Y, Fiskus W, Nechtman J, et al. Role of 35. Cole PA. Chemical probes for histone-modifying enzymes. Nat Chem acetylation and extracellular location of heat shock protein 90alpha in Biol 2008;4:590–7. tumor cell invasion. Cancer Res 2008;68:4833–42. 36. Wang L, Gural A, Sun XJ, Zhao X, Perna F, Huang G, et al. The 19. Santana-Davila R, Perez EA. Treatment options for patients with triple- leukemogenicity of AML1-ETO is dependent on site-specific lysine negative breast cancer. J Hematol Oncol 2010;3:42. acetylation. Science 2011;333:765–9.

OF10 Mol Cancer Ther; 12(5) May 2013 Molecular Cancer Therapeutics

A New Class of Acetyltransferase Inhibitors

37. Xu CR, Cole PA, Meyers DJ, Kormish J, Dent S, Zaret KS. Chromatin 43. Tanaka Y, Naruse I, Hongo T, Xu M, Nakahata T, Maekawa T, et al. "prepattern" and histone modifiers in a fate choice for liver and Extensive brain hemorrhage and embryonic lethality in a mouse null pancreas. Science 2011;332:963–6. mutant of CREB-binding protein. Mech Dev 2000;95:133–45. 38. Santer FR, Hoschele PP, Oh SJ, Erb HH, Bouchal J, Cavarretta IT, et al. 44. Yao TP, Oh SP, Fuchs M, Zhou ND, Ch'ng LE, Newsome D, et al. Gene Inhibition of the acetyltransferases p300 and CBP reveals a targetable dosage-dependent embryonic development and proliferation defects function for p300 in the survival and invasion pathways of prostate in mice lacking the transcriptional integrator p300. Cell 1998;93: cancer cell lines. Mol Cancer Ther 2011;10:1644–55. 361–72. 39. Boelsterli UA, Ho HK, Zhou S, Leow KY. Bioactivation and hepato- 45. Yu H, Pardoll D, Jove R. STATs in cancer inflammation and immunity: a toxicity of nitroaromatic drugs. Curr Drug Metab 2006;7:715–27. leading role for STAT3. Nat Rev Cancer 2009;9:798–809. 40. Morimoto T, Sunagawa Y, Kawamura T, Takaya T, Wada H, Nagasawa 46. Hecht A, Vleminckx K, Stemmler MP, van Roy F, Kemler R. The p300/ A, et al. The dietary compound curcumin inhibits p300 histone acet- CBP acetyltransferases function as transcriptional coactivators of yltransferase activity and prevents heart failure in rats. J Clin Invest beta-catenin in vertebrates. EMBO J 2000;19:1839–50. 2008;118:868–78. 47. Morris JPt, Wang SC, Hebrok M. KRAS, Hedgehog, Wnt and the 41. He L, Sabet A, Djedjos S, Miller R, Sun X, Hussain MA, et al. Metformin twisted developmental biology of pancreatic ductal adenocarcinoma. and insulin suppress hepatic gluconeogenesis through phosphoryla- Nat Rev Cancer 2010;10:683–95. tion of CREB binding protein. Cell 2009;137:635–46. 48. Takemaru KI, Moon RT. The transcriptional coactivator CBP interacts 42. Mantelingu K, Reddy BA, Swaminathan V, Kishore AH, Siddappa with beta-catenin to activate gene expression. J Cell Biol 2000;149: NB, Kumar GV, et al. Specific inhibition of p300-HAT alters global 249–54. gene expression and represses HIV replication. Chem Biol 2007;14: 49. Dawson MA, Kouzarides T. Cancer epigenetics: from mechanism to 645–57. therapy. Cell 2012;150:12–27.

www.aacrjournals.org Mol Cancer Ther; 12(5) May 2013 OF11

Small-Molecule Inhibitors of Acetyltransferase p300 Identified by High-Throughput Screening Are Potent Anticancer Agents

Heng Yang, Christie E. Pinello, Jian Luo, et al.

Mol Cancer Ther Published OnlineFirst April 26, 2013.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-12-0930

Supplementary Access the most recent supplemental material at: Material http://mct.aacrjournals.org/content/suppl/2013/03/12/1535-7163.MCT-12-0930.DC1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/early/2013/04/26/1535-7163.MCT-12-0930. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.