Disruption of STAT3-DNMT1 interaction by SH-I-14 induces re-expression of tumor suppressor genes and inhibits growth of triple-negative breast tumor

Epigenetic regulation of gene expression is an emerging target to treat several human diseases including cancers. In cancers, expressions of many tumor suppressor genes are suppressed by hyper-methylation in their regulatory regions. Herein, we describe a novel carbazole SH-I-14 that decreased the level of the acetyl-STAT3 at the K685 residue. Mutation analysis revealed that SH-I-14 disrupted STAT3-DNMT1 interaction by removing acetyl group from K685 of STAT3. Finally, the inhibition of STAT3-DNMT1 interaction by SH-I-14 resulted in re-expression of tumor suppressor genes such as VHL and PDLIM4 through de-methylation of their promoter regions. In addition, SH-I-14 showed anti-proliferative effect in triple-negative breast cancer (TNBC) cell lines in vitro and anti-tumor effect in a mouse xenograft model of MDA-MB-231 tumor. Taken together, our results suggest that targeting acetyl-STAT3 (K685) provides potential therapeutic opportunity to treat a subset of human cancers.


INTRODUCTION
Signal transducer and activator of transcription 3 (STAT3), a member of STAT family, is a transcriptional regulator that mediates transduction of extracellular signals to the nucleus in response to various cytokines [1]. STAT3 is activated by phosphorylation of Y705 residue that triggered by cytokine receptor associated kinases such as Janus kinases (JAKs) or receptor tyrosine kinases upon ligand binding [1]. STAT3 is widely recognized as a potential drug target because: a) activation and/or overexpression of STAT3 is widely associated with many human cancers [2][3][4]; b) it induces tumor-promoting inflammation and suppresses anti-tumor immunity [2]; and c) it induces various anti-apoptotic and/or pro-proliferative gene expressions [2].
The regulation of STAT3 activity through reduction of phosphorylation [2,4] and various inhibitors that reduce the phosphorylation of STAT3 are under active investigation [3,4]. By reducing the phospho-STAT3, these inhibitors suppress the STAT3-mediated transcriptional activation of anti-apoptotic and/or prosurvival proteins [3,4]. Recent advances in STAT3 biology, however, suggest new aspects of STAT3 function in cancers. First, STAT3 has been reported to have the repressive effects on the expression of tumor suppressor (TS) or pro-apoptotic genes [5][6][7][8][9][10][11]. Second, it has been reported that STAT3 is regulated by multiple acetylation [12] and acetylation of STAT3 at K685 and is an important player in DNA methylation of promoter regions for TS genes through interaction with DNA (cytosine-5)-methyltransferase 1 (DNMT1) [13,14]. It was also found that acetylation of STAT3 is elevated in tumors and contributes to tumor progression by inducing DNA methylation [13]. Therefore, acetyl-STAT3 could be a potent target for tumor treatment and several small-
Since 9H-carbazole was first described in 1872 [17], numerous natural carbazoles have been identified from various plants and microorganisms [17,18]. After discovering antitumor activity of ellipticine in the 1960s, antitumor activity of carbazoles has been widely studied [18]. However, few of them have advanced to clinical investigation [18]. As an attempt to discover novel potent small molecules with anti-tumor activity, recently we synthesized a series of carbazoles with fluorescent moiety [19,20]. One of them, SH-I-14 has been identified as an inhibitor of phospho-STAT3 (Y705) by induction of protein-tyrosine phosphatase non-receptor type 6 (PTPN6/ SHP-1) expression [20]. Here, we further evaluated SH-I-14 as an inhibitor of acetyl-STAT3 (K685). Inhibition of acetyl-STAT3 by SH-I-14 resulted in disruption of STAT3-DNMT1 interaction, de-methylation of TS genes' promoter regions in triple-negative breast cancer (TNBC) cells in vitro, re-expression of TS genes and inhibition of TNBC tumor growth in vivo. Our findings further suggest that acetylation of STAT3 (K685) is a novel therapeutic target to treat human cancers.

SH-I-14 inhibits the proliferation of TNBC cells
First, we determined the cytotoxic effect of SH-I-14 ( Figure 1A) in three TNBC cell lines. The cells were treated with a range of concentrations of SH-I-14 for ~72 hr and the viable cells were measured by MTT assay. As reported recently [20], all these TNBC cells tested were sensitive and more than 60% reduction of viable cells was observed at 1 µM ( Figure 1B). Indeed, SH-I-14 showed the antiproliferative effect on a broad range of TNBC cell lines tested [20]. On the contrary, a JAK1/2 inhibitor CP690550 [21] had no significant effect on the proliferation of TNBC cells up to 10 µM ( Figure 1C).
The cytotoxic effect of SH-I-14 was further analyzed by western blot and cell cycle analysis in HS578T cells. As shown in Figure 1D, SH-I-14 induced cleavage of poly(ADP-ribose) polymerase (PARP), a hallmark of apoptotic cell death [22], in a dose-dependent manner within 16 hr post-treatment. In addition, 0.3 µM of SH-I-14 induced a modest increase of sub-G1 population in cell cycle analysis ( Figure 1E).

SH-I-14 disrupts DNMT1-STAT3 interaction
Given that acetylation of STAT3 at K685 is crucial for binding of STAT3 to DNMT1 [13], we determined the STAT3-DNMT1 interaction in the presence of SH-I-14.
HS578T cells were treated with increasing concentrations of SH-I-14 for 24 hr and STAT3-bound proteins were immunoprecipitated by anti-STAT3 antibody. Western blot analysis demonstrated that STAT3 bound to DNMT1 in HS578T cells. Under this condition, STAT3-DNMT1 interaction was abolished by SH-I-14 as low as 0.3 µM (Figure 2A, left). Although the highest concentration (1 µM) of SH-I-14 reduced the level of DNMT1 protein, the disruption of STAT3-DNMT1 interaction by SH-I-14 occurred at much lower concentration (0.1 µM) that did not affect the level of DNMT1. SH-I-14 also disrupted STAT3-DNMT1 interaction in SUM1315MO2 cells (Figure 2A, right). On the contrary, the JAK2 inhibitor (AZD1480) or JAK1/2 inhibitor (CP690550) did not affect STAT3-DNMT1 interaction ( Figure 2B).
Since the histone acetyltransferase p300 acetylates STAT3 (K685) [25,26], we further tested the effect of p300 overexpression on the SH-I-14-mediated deacetylation of STAT3. HEK293T cells were transfected with STAT3 expression vector in the presence or absence of FLAG-p300 expression vector, further treated with   for 48 hr and subjected to DNA methylation assay A., qRT-PCR for mRNA expression B., or western blot analysis C.. A. Genomic DNAs from the cells treated with SH-I-14 were treated with methylation-specific restriction enzymes followed by qRT-PCR to determine the methylation status of each promoter. B. The cDNAs, synthesized from total RNAs of the cells incubated with SH-I-14, were used for qRT-PCR to measure the level of mRNAs. GAPDH level was used to normalize the level of cDNAs. (A~B) *P < 0.05; **P < 0.01; and ***P < 0.001. C. The lysates from the cells treated with SH-I-14 for 48 hr were analyzed by western blot with indicated antibodies. β-actin was used as a loading control.

SH-I-14 induces the re-expression of TS genes via demethylation of DNA
The STAT3-DNMT1 interaction has been reported to be important for DNA methylation in the promoter region of TS genes [7,13,14]. Since SH-I-14 disrupts STAT3-DNMT1 interactions by decreasing the level of acetyl-STAT3 (K685), we further determine the effect of SH-I-14 on the methylation of DNA. The DNA methylation status of a panel of specific primers for 22 TS gene promoters was analyzed in HS578T cells treated with 0.3µM of SH-I-14 for 48 hr. As results, we identified four TS genes, that were de-methylated by SH-I-14 in HS578T cells, including retinoic acid receptor beta (RARB), neurogenin 1 (NEUROG1), PDZ and LIM domain 4 (PDLIM4), and Von Hippel-Lindau tumor suppressor (VHL) (data not shown). We further analyzed the methylation status of these promoters in two TNBC cells. HS578T and MDA-MB-231 cell lines were treated with 0.3µM of SH-I-14 for 48 hr and promoter methylations were analyzed by specific primers for each gene's promoter. In both cells, the promoter regions of VHL and PDLIM4 genes were highly methylated and SH-I-14 near commonly demethylated these promoters ( Figure 4A).
To further verify the consequence of promoter demethylation, mRNA and protein expression of VHL and PDLIM4 was analyzed by qRT-PCR and western blot analyses. First, the cells were treated with an increasing amount of SH-I-14 for 48 hr and RNAs from these cells were subjected to qRT-PCR analysis. De-methylation of VHL and PDLIM4 gene promoters by SH-I-14 resulted in reactivation of mRNA expression of these genes in both HS578T and MDA-MB-231 in a dose-dependent manner ( Figure 4B). Another TS gene, a Ras homologue member I (ARHI), that regulated by STAT3-DNMT1-dependent methylation in ovarian cancer cells [14] was also induced by SH-I-14 ( Figure 4B). Next, lysates from the cells treated with different concentration of SH-I-14 for 48 hr were used to determine the expression of these proteins. Consistent with mRNA expression, the levels of VHL and PDLIM4 proteins were increased by SH-I-14 in a dosedependent manner ( Figure 4C).

SH-I-14 reduces tumor growth in vivo
To address whether SH-I-14 affects tumor growth in vivo, SH-I-14 (10 mg/kg) was administrated three times per week to athymic nude mice bearing MDA-MB-231 human TNBC tumors. As shown in Figure 5A, SH-I-14 markedly reduced the tumor growth. No apparent loss of body weights was observed during the treatment ( Supplementary Information, Figure S2).
To determine the level of markers in tumor samples, we administered SH-I-14 to MDA-MB-231 xenograft mice daily for 5-days and western blot analysis was performed with lysates from tumor samples. As shown in Figure 5B, the level of acetyl-STAT3 (K685) was reduced in tumor samples treated with SH-I-14 compared to tumor samples treated with DMSO control. In addition, the level of two TS proteins, VHL and PDLIM4, was increased in SH-I-14-treated tumor samples.

DISCUSSION
Here we report a small molecule compound, SH-I-14, that inhibits in vivo tumor growth in a human TNBC xenograft model. SH-I-14 inhibits STAT3-DNMT1 interaction by reducing the acetylation of STAT3 (K685). In addition, SH-I-14 induces re-expression of TS genes, VHL and PDLIM4, through de-methylation of DNA in their promoter regions. Administration of SH-I-14 markedly reduced the in vivo tumor growth in a mouse xenograft model-bearing tumor of human MDA-MB-231 cells. It was also observed that de-acetylation of STAT3 and expression of VHL and PDLIM4 was induced by SH-I-14 in tumors from xenograft mice.
STAT3-DNMT1 interaction implicates new functional roles of STAT3 in epigenetic gene silencing in human cancer [7,13]. DNMT1 is the major DNMT that expressed ubiquitously [31] and has important role in tumorigenesis by silencing tumor suppressor genes by hyper-methylation [32,33]. Since DNMT1 has been reported to be up-regulated in many human cancers [34], compounds inhibiting DNMT1 are currently under www.impactjournals.com/oncotarget active development [32,33]. High levels of acetyl-STAT3 (K685) was also found in many types of cancers including melanomas, colon cancers and TNBCs [13]. In addition, acetyl-STAT3 (K685) was suggested to crucial for promoter methylation of tumor suppressor genes such as STAT1, p53, SOCS3, SHP-1, and p16 (CDKN2A) in mouse embryonic fibroblasts, A2058 melanoma cell, and HCT116 colon cancer cell [7,13]. In the present study, the functional consequence of disrupting the STAT3-DNMT1 interaction by SH-I-14 was re-expression of TS genes including VHL and PDLIM4 through de-methylation of their promoter DNAs. SH-I-14 induced de-methylation of these genes' promoters near completely within a 48 hr post-treatment. The expression of VHL and PDLIM4 protein by SH-I-14 was also confirmed in both cell culture and mouse xenograft model. Re-expression of these TS genes has been reported to suppress various cancer cells' growth both in vitro and in vivo [35][36][37][38][39][40][41]. Taken together, our data suggest that blocking the STAT3-DNMT1 interaction through inhibition of acetyl-STAT3 (K685) Oncotarget 83464 www.impactjournals.com/oncotarget by SH-I-14 is sufficient to re-express TS genes by demethylation of their promoter DNAs ( Figure 6).
Our present study raises several questions to be addressed: 1) How does SH-I-14 facilitate the removal of acetyl group from K685 of STAT3? Although overexpression of p300 could overcome SH-I-14-mediated de-acetylation of STAT3 (K685), even 10 µM of SH-I-14 failed to inhibit acetyltransferase activity of p300 in vitro (data not shown). In fact, we performed a series of in vitro enzyme assays with purified HATs, HDACs, and SIRTs (from Reaction Biology Corp, Malvern, PA). However, SH-I-14 had no apparent activity toward these enzymes up to 10 µM concentration (data not shown). Future studies should be focused on the identification of molecular target(s) of SH-I-14. Identification of SH-I-14-binding protein(s) may provide additional target(s) to further develop potential therapeutics. 2) What is the fate of DNMT1 after dissociation from STAT3 by SH-I-14? In our present study, high concentration of SH-I-14 reduced the level of DNMT1 protein. Inhibition of STAT3-binding by SH-I-14 may affect the localization and/or stability of DNMT1. Several lines of evidences suggest that DNMT1 might be regulated by localization and/or stability. Nuclear localization signals in the N-terminal region of DNMT1 implicates that nucleocytoplasmic shuttling may be an important regulatory mechanism of DNA methylation [42]. In addition, stability of DNMT1 was increased in human cancers [43,45] and DNMT1 is degraded by ubiquitin-dependent proteasomal degradation [45]. Interestingly, a HDAC inhibitor trichostatin A (TSA) has been reported to reduce DNMT1 protein in urothelial carcinoma cell lines [46] and reduce nuclear DNMT1 while increase cytoplasmic DNMT1 in HepG2 cell [47]. However, the effect of STAT3-interaction on the stability of DNMT1 has not been revealed yet. 3) What kinds of genes are re-repressed by SH-I-14 through de-methylation? Since DNMT1 is not a DNA-binding protein, DNA binding proteins recruit DNMT1 by protein-protein interaction to specific DNA sequences and determine the set of genes to be regulated by these protein complexes. Although recent reports demonstrated that STAT3 facilitate DNMT1 binding to several genes' promoter [7,13], currently there is no detailed study on the specific DNA sequences for STAT3-DNMT1 complex. As a probe to disrupt STAT3-DNMT1 interaction, SH-I-14 may provide valuable tool to determine the specific sequence for STAT3-DNMT1-DNA binding and the set of methylated genes by STAT3-DNMT1 complex. Since the method used to measure DNA methylation in this study is based upon restriction enzyme cleavage and the number of targets tested is limited, further study is needed to measure the global methylation pattern in the absence or presence of SH-I-14. 4) We cannot exclude additional negative regulation of STAT3 function by SH-I-14 on the suppression of TS genes. In the present study, we found that SH-I-14 could not disrupt STAT3 K685Q -DNMT1 interaction while Oncotarget 83465 www.impactjournals.com/oncotarget reduced the phospho-STAT3 K685Q (Y705) in PC-3 cells. We also found that SH-I-14 induced the expression of PTPN6/SHP-1 [20] and SOCS3 (data not shown), which negatively regulated phospho-STAT3 [3,48] in TNBC cells. These results suggest possible alternative repression of PTPN6/SHP-1 expression by STAT3. Since DNMT1 functions to maintain DNA methylation patterns by recognizing hemimethylated CpG sequences after DNA replication [49], it might require additional time for DNA replication after inhibition of DNMT1 to sufficiently demethylate DNA for re-expression of suppressed TS genes. As an example, it has been reported that maximal DNA de-methylation was achievable after 48 hr treatment with DNMT inhibitors, 5-azacytidine or 2'-deoxy-5-azacytidine (decitabine) [50,51] and the mRNA expression of p16 TS gene was detected beginning 36 hr after decitabinetreatment in T24 bladder carcinoma cells [51].
In conclusion, in this research work we described a potent inhibitor of acetyl-STAT3 (K685). We confirmed that our novel inhibitor SH-I-14 could reduce the acetylation of STAT3 at K685 residue. We further found that SH-I-14 disrupted STAT3-DNMT1 interaction because of its de-acetylation of STAT3, which resulted in re-expression of tumor suppressor genes such as VHL and PDLIM4 through de-methylation of their promoter regions. Our data suggest that targeting acetylation of STAT3 (K685) by small molecule inhibitors is a plausible new approach to treat human cancers. Since both acetylation of STAT3 and the level of DNMT1 are known to be elevated in a broad range of human cancers, it is likely that the STAT3-DNMT1 complex may participate the repression of TS genes by DNA methylation in various cancers. Future profiling of acetyl-STAT3 and DNMT1 status in human cancers may provide an alternative therapeutic opportunity with a distinct mechanism of targeting the STAT3-DNMT1 interaction. Automated Cell Counter (Logos Biosystems, Gyunggi-do, Korea). AZD1480 was purchased from Selleck Chemicals (Houston, TX) and CP690550 was obtained from LC Labs (Woburn, MA). Stock solutions of compounds were made at 10 mM concentration in dimethyl sulfoxide (DMSO) and stored at -20°C in small aliquots.

DNA methylation assay
Genomic DNA was isolated from HS578T and MDA-MB-231 treated with 0.3 µM of SH-I-14 for 48 hr using DNeasy Blood & Tissue kit (QIAGEN, Valencia, CA) according to manufacturer's protocol. CpG island DNA methylation was analyzed by EpiTect Methyl II Oncotarget 83466 www.impactjournals.com/oncotarget PCR array system (QIAGEN) following manufacturer's instruction. In brief, genomic DNA was digested by methylation-dependent or/and methylation-sensitive enzymes at 37°C for overnight. Following digestion, DNA was analyzed by quantitative realtime-PCR withEpiTect Methyl II PCR array system or predesigned VHL or PDLIM4 primers (QIAGEN). Percentage of methylation and un-methylation were calculated by data analysis tool provided by QIAGEN.

In Vivo xenograft tumor models
Animal use procedures were approved by the Institutional Animal Care and Use Committees of Georgetown University Medical Center. MDA-MB-231 cells (2.5 × 10 6 cells/mouse) mixed with Matrigel (BD Biosciences, San Jose, CA) were injected subcutaneously into the flank of male athymic nude (Foxn1 nu ) mice of 6-week age (Harlan Laboratories, Frederick, MD). SH-I-14 was dissolved in DMSO at concentration of 30 mg/ ml. For injection of compound, DMSO or SH-I-14 was diluted in a 1:1 mixture of PBS and injection solution. Injection solution was made as described previously [20]. Diluted SH-I-14 (10 mg/kg) or DMSO was administered into peritoneal cavity of mouse with 3 times per week. The body weights and tumor sizes were measured before every injection. Tumor sizes were measured using a digital caliper and tumor volumes were calculated using the formula, [volume (mm 3 ) = width (mm) × length (mm) × height (mm) / 2].
For western blot analysis of protein markers, xenograft mice were established as described above. SH-I-14 (10 mg/kg) or DMSO was administered into peritoneal cavity of mouse daily for 5 days. Mice were euthanized at 1 hr after last administration of drug. Tumors from each mouse were lysed and subjected to western blot analysis.

Statistical analysis
To compare two groups of interest, the Turkey's multiple comparison test (ANOVA) was applied for statistical analysis. * indicates P < 0.05; ** indicates P < 0.01; and *** indicates P < 0.001.