Role of HDAC3-miRNA-CAGE Network in Anti-Cancer Drug-Resistance

Histone modification is associated with resistance to anti-cancer drugs. Epigenetic modifications of histones can regulate resistance to anti-cancer drugs. It has been reported that histone deacetylase 3 (HDAC3) regulates responses to anti-cancer drugs, angiogenic potential, and tumorigenic potential of cancer cells in association with cancer-associated genes (CAGE), and in particular, a cancer/testis antigen gene. In this paper, we report the roles of microRNAs that regulate the expression of HDAC3 and CAGE involved in resistance to anti-cancer drugs and associated mechanisms. In this review, roles of HDAC3-miRNAs-CAGE molecular networks in resistance to anti-cancer drugs, and the relevance of HDAC3 as a target for developing anti-cancer drugs are discussed.

HDAC3 can bind to the promoter sequences of Runt-related transcription factor 2 (Runx 2) and suppress the metastatic potential of colorectal cancers [6]. HDAC3 can also bind to cAMP response element-binding protein (CREB) and decrease the migration potential of metastatic breast cancer cells [7]. HDAC3, a downstream target of AKT serine/threonine kinase (Akt) and glycogen synthase kinase 3 beta (GSK3β), can promote neuronal death [8] and beta cell apoptosis induced by inflammatory cytokines [9]. HDAC3 can also decrease expression of Hypoxia-inducible factor-1 α (HIF-1α) by negative regulation of NF-κB in metastatic breast cancer cells [10] and regulates the signal transducer and activator of transported 3 (Stat3) which inhibits Beclin 1, a prognostic marker and autophagy in non-small cell lung cancer cells [11]. Down-regulation of HDAC3 enhances the tumorigenic potential of lung cancer cells [12]. HDAC3 can also decrease the angiogenic potential of melanoma cells [13]. The role of HDAC3 in cancer development is known to be tissue-specific.
MDR to confer anti-apoptotic effects [49]. These results suggest the important role of HDAC3 in anti-cancer drug resistance.
Taxol-resistant ovarian tumors display altered expression of tubulins [50]. Paclitaxel resistance is accompanied by high expression levels of tubulin β3 in cancer cells such as non-small cell lung cancer cells [51,52]. It has been reported that taxol resistance is due to MAPK activation in colon cancer cells [53]. Taxol resistance is induced by the Phosphoinositide 3-kinase (PI3K)-Akt-JNK signaling pathway in osteosarcoma cells [54]. HDAC3 can inhibit JNK signaling [3] and MAPK activation [5].
HDAC3 can bind to promoter sequences of tubulin β3, HDAC6, and MDR1, and negatively regulate expression levels of these proteins [55] ( Figure 1A). Down-regulation of HDAC6 or tubulin β3 enhances sensitivity of Malme3M R cells to anti-cancer drugs [55]. HDAC6 shows an interaction with tubulin β3 that is disrupted by HDAC3 [55] ( Figure 1B). HDAC6 can interact with beta tubulin in yeast two-hybrid assays and deacetylate tubulin [56]. HDAC3 functions upstream of HDAC6 and tubulin β3 ( Figure 1B). Therefore, tubulin β3 is a downstream target of HDAC3 that confers resistance to anti-cancer drugs.
We investigated the role of ubiquitination in the expression regulation of HDAC3. The loss of Seven in Absentia Homolog 2 (SIAH2) is known to increase chemo-sensitivity [57]. E3 ubiquitin ligase activity of SIAH2 is necessary for resistance to death receptor-mediated apoptosis [58]. Inhibition of SIAH2 ubiquitin ligase blocks melanoma tumorigenesis [59]. Loss of SIAH2 suppresses tumorigenesis in a large tumor suppressor kinase 2 (LATS2)-dependent manner in a xenograft mouse model [60]. SIAH2 expression shows an inverse correlation with the tumor grade, p53, and human epidermal growth factor receptor 2 (HER2) [60]. The inhibition of SIAH2 activity reduces metastatic potential, but not tumorigenic potential, via HIF-1α [61]. SIAH2 targets p53 and regulates the acetylation status of p53 [62]. Ski protein stabilizes HDAC3 by binding to SIAH2 [63]. cancer cells [51,52]. It has been reported that taxol resistance is due to MAPK activation in colon cancer cells [53]. Taxol resistance is induced by the Phosphoinositide 3-kinase (PI3K)-Akt-JNK signaling pathway in osteosarcoma cells [54]. HDAC3 can inhibit JNK signaling [3] and MAPK activation [5]. HDAC3 can bind to promoter sequences of tubulin β3, HDAC6, and MDR1, and negatively regulate expression levels of these proteins [55] ( Figure 1A). Down-regulation of HDAC6 or tubulin β3 enhances sensitivity of Malme3M R cells to anti-cancer drugs [55]. HDAC6 shows an interaction with tubulin β3 that is disrupted by HDAC3 [55] ( Figure 1B). HDAC6 can interact with beta tubulin in yeast two-hybrid assays and deacetylate tubulin [56]. HDAC3 functions upstream of HDAC6 and tubulin β3 ( Figure 1B). Therefore, tubulin β3 is a downstream target of HDAC3 that confers resistance to anti-cancer drugs.
We investigated the role of ubiquitination in the expression regulation of HDAC3. The loss of Seven in Absentia Homolog 2 (SIAH2) is known to increase chemo-sensitivity [57]. E3 ubiquitin ligase activity of SIAH2 is necessary for resistance to death receptor-mediated apoptosis [58]. Inhibition of SIAH2 ubiquitin ligase blocks melanoma tumorigenesis [59]. Loss of SIAH2 suppresses tumorigenesis in a large tumor suppressor kinase 2 (LATS2)-dependent manner in a xenograft mouse model [60]. SIAH2 expression shows an inverse correlation with the tumor grade, p53, and human epidermal growth factor receptor 2 (HER2) [60]. The inhibition of SIAH2 activity reduces metastatic potential, but not tumorigenic potential, via HIF-1α [61]. SIAH2 targets p53 and regulates the acetylation status of p53 [62]. Ski protein stabilizes HDAC3 by binding to SIAH2 [63].
TargetScan analysis has predicted that miR-335 is a negative regulator of SIAH2. MiR-335 binds to 3′-UTR of SIAH2 and decreases expression of SIAH2, which in turn increases the expression of HDAC3 to confer anti-cancer drug sensitivity in anti-cancer drug-sensitive melanoma cells, such as Malme3M cells [64] (Figure 2A). HDAC3 may increase expression of miR-335. In Malme3M R cells (anti-cancer drug-resistant melanoma cells), SIAH2 can binds to HDAC3, resulting in degradation of HDAC3 by ubiquitination [64] ( Figure 2B). Thus, HDAC3 may confer sensitivity to anti-cancer drugs by regulating expression of SIAH2.  TargetScan analysis has predicted that miR-335 is a negative regulator of SIAH2. MiR-335 binds to 3 -UTR of SIAH2 and decreases expression of SIAH2, which in turn increases the expression of HDAC3 to confer anti-cancer drug sensitivity in anti-cancer drug-sensitive melanoma cells, such as Malme3M cells [64] (Figure 2A). HDAC3 may increase expression of miR-335. In Malme3M R cells (anti-cancer drug-resistant melanoma cells), SIAH2 can binds to HDAC3, resulting in degradation of HDAC3 by ubiquitination [64] ( Figure 2B). Thus, HDAC3 may confer sensitivity to anti-cancer drugs by regulating expression of SIAH2.

HDAC3-miRNA Network in Angiogenesis and Anti-Cancer Drug Resistance
Anti-cancer drug-resistant phenotypes are under epigenetic regulation [65,66]. Among various HDACs, HDAC3 can inhibit the invasion, migration, and angiogenic potential of hepatic cancer cell lines and melanoma cell lines [67,68]. HDAC3 also confers sensitivity to anti-cancer drugs [67,69] ( Figure 3A). Many reports have suggested a close relationship between angiogenic potential and anticancer drug resistance [68]. HDAC3 inhibits the angiogenic potential of cancer cells by decreasing expression levels of angiogenic factors, such as VEGF and plasminogen activator inhibitor-1 (PAI-1) [22] ( Figure 3B). VEGF and HDAC3 can form a negative feedback loop and regulate endothelial cell tube formation [13]. Down-regulation of HDAC3 enhances angiogenic potential of Malme3M cells, while overexpression of HDAC3 decreases the angiogenic potential of Malme3M R cells [13] ( Figure 3C).
Expression levels of HDAC1 and HDAC2 are higher in melanoma cells resistant to anti-cancer drugs than those in melanoma cells sensitive to anti-cancer drugs [64] ( Figure 4A). It is probable that HDAC1 and HDAC-2 might be able to confer resistance to anti-cancer drugs. Malme3M R cells show lower expression of HDAC3 than parental sensitive Malme3M cells [64] ( Figure 4B). Melanoma cells that are naturally resistant to anti-cancer drugs also showed lower expression of HDAC3 than anticancer drug-sensitive melanoma cells [64]. Overexpression of HDAC3 overcomes resistance of Malme3M R cells to anti-cancer drugs [64].

HDAC3-miRNA Network in Angiogenesis and Anti-Cancer Drug Resistance
Anti-cancer drug-resistant phenotypes are under epigenetic regulation [65,66]. Among various HDACs, HDAC3 can inhibit the invasion, migration, and angiogenic potential of hepatic cancer cell lines and melanoma cell lines [67,68]. HDAC3 also confers sensitivity to anti-cancer drugs [67,69] ( Figure 3A). Many reports have suggested a close relationship between angiogenic potential and anti-cancer drug resistance [68]. HDAC3 inhibits the angiogenic potential of cancer cells by decreasing expression levels of angiogenic factors, such as VEGF and plasminogen activator inhibitor-1 (PAI-1) [22] ( Figure 3B). VEGF and HDAC3 can form a negative feedback loop and regulate endothelial cell tube formation [13]. Down-regulation of HDAC3 enhances angiogenic potential of Malme3M cells, while overexpression of HDAC3 decreases the angiogenic potential of Malme3M R cells [13] ( Figure 3C).
Expression levels of HDAC1 and HDAC2 are higher in melanoma cells resistant to anti-cancer drugs than those in melanoma cells sensitive to anti-cancer drugs [64] ( Figure 4A). It is probable that HDAC1 and HDAC-2 might be able to confer resistance to anti-cancer drugs. Malme3M R cells show lower expression of HDAC3 than parental sensitive Malme3M cells [64] ( Figure 4B). Melanoma cells that are naturally resistant to anti-cancer drugs also showed lower expression of HDAC3 than anti-cancer drug-sensitive melanoma cells [64]. Overexpression of HDAC3 overcomes resistance of Malme3M R cells to anti-cancer drugs [64]. miR-200b, miR-217, and miR-335 in Malme3M cells ( Figure 4A). Increased expressions of miR-200b, miR-217, and miR-335 confer sensitivity to anti-cancer drugs. In Malme3M R cells, miR-326 binds to 3′-UTR of the HDAC3 gene and decreases expression of HDAC3 ( Figure 4B). Decreased expression of HDAC3 decreases expression levels of miR-200b, miR-217, and miR-335 to confer anti-cancer drug resistance ( Figure 4B). These reports suggest the important role of the HDAC3-miRNA network in anti-cancer drug resistance.   miR-200b, miR-217, and miR-335 in Malme3M cells ( Figure 4A). Increased expressions of miR-200b, miR-217, and miR-335 confer sensitivity to anti-cancer drugs. In Malme3M R cells, miR-326 binds to 3′-UTR of the HDAC3 gene and decreases expression of HDAC3 ( Figure 4B). Decreased expression of HDAC3 decreases expression levels of miR-200b, miR-217, and miR-335 to confer anti-cancer drug resistance ( Figure 4B). These reports suggest the important role of the HDAC3-miRNA network in anti-cancer drug resistance.   MicroRNAs (miRNAs) are small, non-coding RNAs (21-23 nucleotides) that functions in post-transcriptional regulation of gene expression. MiRNAs can regulate the expression of various oncogenes and tumor suppressor genes [70]. MicroRNA array analyses show differential expression of miRNAs, such as miR-326, miR-200b, miR-217, and miR-335 between Malme3M cells and Malme3M R cells [64]. MiR-326 negatively regulated by Notch can decrease the tumorigenic potential of glioma cells [71]. MiR-326 enhances sensitivity of etoposide (VP-16)-resistant breast cancer cell lines, MCF-7/VP to VP-16 and doxorubicin [72]. MiR-335 increases the expression of HDAC3 by preventing SIAH2 from inducing ubiquitination of HDAC3 [64]. MiR-326 directly regulated by HDAC3 ( Figure 4A) can regulate responses to anti-cancer drugs [64]. Chromatin immunoprecipitation (ChIP) assays have revealed that HDAC3 can bind to promoter sequences of miR-200b, miR-217, and miR-335 in Malme3M cells ( Figure 4A). Increased expressions of miR-200b, miR-217, and miR-335 confer sensitivity to anti-cancer drugs. In Malme3M R cells, miR-326 binds to 3 -UTR of the HDAC3 gene and decreases expression of HDAC3 ( Figure 4B). Decreased expression of HDAC3 decreases expression levels of miR-200b, miR-217, and miR-335 to confer anti-cancer drug resistance ( Figure 4B). These reports suggest the important role of the HDAC3-miRNA network in anti-cancer drug resistance.

Role of CAGE-miRNA Network in Anti-Cancer Drug Resistance
Cancer-associated genes (CAGE) have been found in the sera of patients with various cancers [73][74][75]. DNMT1 increases methylation at promoter sequences of CAGE, leading to decreased expression of CAGE [76] ( Figure 5A). CAGE can bind to HDAC2 and confers resistance to various anti-cancer drugs [77] (Figure 5B). CAGE can also inactivate tumor suppressor retinoblastoma (Rb) protein by phosphorylation, increase the expression of cyclins, and act as an oncogene [78] (Figure 6A).
CAGE also binds to GSK3β and increases expression of cyclinD, thus conferring resistance to various anti-cancer drugs in melanoma cells [79] (Figure 6B). TargetScan has predicted binding of miR-200b, miR-217, and miR-335 to 3 untranslated regions (UTR) of CAGE. MiR-200b, miR-217, and miR-335 form a negative feedback loop with CAGE ( Figure 6B). CAGE-derived peptides can bind to CAGE and confer sensitivity to anti-cancer drugs in melanoma cells [79] (Figure 6B). CAGE also binds to the sex-determining region y-box 2 (SOX2) and induces cancer stem cell-like properties in melanoma cells [80] (Figure 6B). In addition, CAGE can bind to Beclin1 and induce autophagic flux (manuscript in preparation).
MiR-200b can form a negative feedback loop with CAGE, a cancer/testis antigen, and regulate invasion of a cancer cell line. It also regulates tumorigenic and angiogenic responses to microtubule-targeting drugs [81]. CAGE binds to miR-200b and directly regulates expression of miR-200b [81]. Recombinant CAGE proteins display angiogenic potential [81]. MiR-217 can form a negative feedback loop with CAGE and negatively regulates the tumorigenic and metastatic potential of melanoma cells [82]. CAGE binds to the epidermal growth factor receptor (EGFR) and human epidermal growth factor receptor 2 (HER2) and confers in vivo resistance to trastuzumab, an inhibitor of HER2 [82] ( Figure 7A). Figure 7A). Down-regulation of EGFR or HER2 can enhance sensitivity to anti-cancer drugs, such as EGFR inhibitors [82]. CAGE directly regulates expression of HER2. It is necessary for the interaction of CAGE with EGFR and HER2 in Malme3M R cells [82] ( Figure 7B). These results indicate that HDAC3 and CAGE can exert an opposite regulation on the expression of miR-200b and miR-217. This has led us to investigate relationships between HDAC3 and CAGE in anti-cancer drug resistance.     (E2F). Activated E2F binds to promoter sequences of cyclins. (B) MiR-200b, miR-217, and miR-335 form negative feedback loops with CAGE. CAGE binds to GSK3β, which increases expression of cyclin D1 to confer resistance to anti-cancer drugs. AQTGTGKT, as a CAGE-derived peptide can disrupt the interaction between CAGE and GSK3β and confer sensitivity to anti-cancer drugs. The Tbar arrows denote negative regulation and both side T-bar arrows denote cross inhibition. The hollow arrows denote positive regulation.

HDAC3 Functions Upstream of CAGE and Targets CAGE
Anti-cancer drug-resistant cancer cell lines show an increased expression of phosphorylated epidermal growth factor receptor (pEGFR Y845 ) and an interaction between CAGE and EGFR [82]. CAGE-derived GTGKT and AQTGTGKT peptides inhibit activation of EGFR, enhance sensitivity to gefitinib and trastuzumab, and prevent interactions of CAGE with EGFR and HER2 [82]. Vorinostat, an inhibitor of HDACs, decreases expression of EGFR and inhibits renal growth [83]. HDAC3 directly regulates the expression of EGFR by binding to promoter sequences of EGFR, while HDAC inhibitors can reverse the expression of EGFR in colorectal cancer cells [84]. HIF1β activates EGFR and extracellular regulated kinase (ERK) by HDACs [85]. EGFR signaling can increase expression levels of HDACs and inhibit osteoblast differentiation by down-regulating Runt-Related Transcription Factor 2 (Runx2) [86]. These reports imply that HDAC3 might be able to regulate EGFR signaling.
AG1478, an inhibitor of EGFR tyrosine kinase, inhibits functions of ATP-binding cassette (ABC) transporters, such as ABCB1 and ABCG2 [87]. The targeting of EGFR in cancers is quite limited due to the status of the Kirsten rat sarcoma 2 viral oncogene homolog (KRAS) mutation [88]. KRAS mutants can bypass EGFR to activate Ras/Raf/MEK/ERK signals and significantly weaken the therapeutic effect of cetuximab [89]. C-Met inhibition can enhance sensitivity to anti-cancer drugs, such as 5-Fluorouracil and taxol [90]. Combination therapy employing the inhibition of c-Met and EGFR suppresses tumorigenic potential of hepatocellular carcinoma cells [91]. These reports suggest that HDAC3 may regulate the expression of CAGE and the activation of EGFR in anti-cancer drug-resistance.
The MiR-326 inhibitor increases the expression of HDAC3, which results in the binding of HDAC3 to promoter sequences of CAGE in anti-cancer drug-resistant Malme3M R cells (Figure 8). Decreased expression of CAGE leads to enhanced sensitivity to anti-cancer drugs [69] (Figure 8). HDAC3 negatively regulates the expression of pEGFR Y845 (Figure 8). Therefore, HDAC3 can target

HDAC3 Functions Upstream of CAGE and Targets CAGE
Anti-cancer drug-resistant cancer cell lines show an increased expression of phosphorylated epidermal growth factor receptor (pEGFR Y845 ) and an interaction between CAGE and EGFR [82]. CAGE-derived GTGKT and AQTGTGKT peptides inhibit activation of EGFR, enhance sensitivity to gefitinib and trastuzumab, and prevent interactions of CAGE with EGFR and HER2 [82]. Vorinostat, an inhibitor of HDACs, decreases expression of EGFR and inhibits renal growth [83]. HDAC3 directly regulates the expression of EGFR by binding to promoter sequences of EGFR, while HDAC inhibitors can reverse the expression of EGFR in colorectal cancer cells [84]. HIF1β activates EGFR and extracellular regulated kinase (ERK) by HDACs [85]. EGFR signaling can increase expression levels of HDACs and inhibit osteoblast differentiation by down-regulating Runt-Related Transcription Factor 2 (Runx2) [86]. These reports imply that HDAC3 might be able to regulate EGFR signaling.
AG1478, an inhibitor of EGFR tyrosine kinase, inhibits functions of ATP-binding cassette (ABC) transporters, such as ABCB1 and ABCG2 [87]. The targeting of EGFR in cancers is quite limited due to the status of the Kirsten rat sarcoma 2 viral oncogene homolog (KRAS) mutation [88]. KRAS mutants can bypass EGFR to activate Ras/Raf/MEK/ERK signals and significantly weaken the therapeutic effect of cetuximab [89]. C-Met inhibition can enhance sensitivity to anti-cancer drugs, such as 5-Fluorouracil and taxol [90]. Combination therapy employing the inhibition of c-Met and EGFR suppresses tumorigenic potential of hepatocellular carcinoma cells [91]. These reports suggest that HDAC3 may regulate the expression of CAGE and the activation of EGFR in anti-cancer drug-resistance.
The MiR-326 inhibitor increases the expression of HDAC3, which results in the binding of HDAC3 to promoter sequences of CAGE in anti-cancer drug-resistant Malme3M R cells (Figure 8). Decreased expression of CAGE leads to enhanced sensitivity to anti-cancer drugs [69] (Figure 8). HDAC3 negatively regulates the expression of pEGFR Y845 (Figure 8). Therefore, HDAC3 can target CAGE to regulate the activation of EGFR signaling, responses to various anti-cancer drugs, and the tumorigenic potential of cancer cells. CAGE to regulate the activation of EGFR signaling, responses to various anti-cancer drugs, and the tumorigenic potential of cancer cells.

Discussion and Conclusions
Reports from our laboratory and others indicate that HDAC3 regulates responses to anti-cancer drugs, tumorigenic potential, and angiogenesis. MicroRNAs that regulate the expression of HDAC3 can also regulate responses to anti-cancer drugs, and the tumorigenic potential and angiogenic potential of cancer cells. Our studies provide the mechanism of anti-cancer drug resistance regulated by HDAC3. We discovered the role of CAGE, a cancer/testis antigen, in anti-cancer drug resistance and the associated mechanisms.
MicroRNAs that form negative feedback loops with CAGE can regulate responses to anti-cancer drugs and form positive feedback loops with HDAC3. HDAC3 functions upstream of CAGE and negatively regulates expression of CAGE. CAGE-promoted anti-cancer drug resistance involves direct regulation of EGFR and HER2 by CAGE and interactions of CAGE with EGFR and HER2.
To overcome anti-cancer drug-resistance, it might be necessary to employ miRNA-inhibitors or miRNA-mimics that can increase expression of HDAC3 while decreasing expression of CAGE. It would be also necessary to identify small molecules that could increase expression of HDAC3. For this, it would be necessary to develop screening tools based on the structure of HDAC3. For better understanding of anti-cancer drug resistance regulated by HDAC3, it would be necessary to employ cancer patient-derived tissues and cell lines that show low level expression levels of HDAC3 but high expression of CAGE. This approach will give better understanding of anti-cancer drug resistance and strategy to develop anti-cancer drugs. Overall, HDAC3 can serve as a target for developing anticancer drugs and/or drugs that can enhance sensitivity to anti-cancer drugs in clinical use.

Discussion and Conclusions
Reports from our laboratory and others indicate that HDAC3 regulates responses to anti-cancer drugs, tumorigenic potential, and angiogenesis. MicroRNAs that regulate the expression of HDAC3 can also regulate responses to anti-cancer drugs, and the tumorigenic potential and angiogenic potential of cancer cells. Our studies provide the mechanism of anti-cancer drug resistance regulated by HDAC3. We discovered the role of CAGE, a cancer/testis antigen, in anti-cancer drug resistance and the associated mechanisms.
MicroRNAs that form negative feedback loops with CAGE can regulate responses to anti-cancer drugs and form positive feedback loops with HDAC3. HDAC3 functions upstream of CAGE and negatively regulates expression of CAGE. CAGE-promoted anti-cancer drug resistance involves direct regulation of EGFR and HER2 by CAGE and interactions of CAGE with EGFR and HER2.
To overcome anti-cancer drug-resistance, it might be necessary to employ miRNA-inhibitors or miRNA-mimics that can increase expression of HDAC3 while decreasing expression of CAGE. It would be also necessary to identify small molecules that could increase expression of HDAC3. For this, it would be necessary to develop screening tools based on the structure of HDAC3. For better understanding of anti-cancer drug resistance regulated by HDAC3, it would be necessary to employ cancer patient-derived tissues and cell lines that show low level expression levels of HDAC3 but high expression of CAGE. This approach will give better understanding of anti-cancer drug resistance and strategy to develop anti-cancer drugs. Overall, HDAC3 can serve as a target for developing anti-cancer drugs and/or drugs that can enhance sensitivity to anti-cancer drugs in clinical use.
Funding: This work was supported by National Research Foundation Grants (2017R1A2A2A05001029 and 2017M3A9G7072417), a grant from the BK21 plus Program. This work was supported by a grant from Kangwon National University (520170495). This research was also supported by (2018R1D1A1B07043498 to Y. Kim).

Conflicts of Interest:
The authors declare no conflict of interest.