p53 tumor suppressor encoded by the TP53 gene becomes activated under stress conditions and during uncontrolled cell division. Activated p53 regulates a huge amount of genes involved in various biological processes due to its transcriptional properties. P53 induces cell cycle arrest by p21 activation to stop uncontrolled cell growth, and it triggers pro-apoptotic genes such as BAX to cause cell death when the DNA damage is beyond repair ^[1][35].

This section presents the importance of modulation of p53 by HDACs and HDIs, including biological consequences. HDIs reduce p53 expression at mRNA and protein levels [36,37,38] and increase p53 acetylation enabling its activation and stabilization [37,38]; therefore, HDIs are of great importance in search of targeted therapy.

The control of p53 protein stability is essential to maintain its tumor suppressor functions. The p53 stability can be modulated by deleted in breast cancer 1 (DBC1) protein, which acts as an endogenic inhibitor of SIRT1 ^[2][39]. DBC1 interplays with SIRT1, reducing its deacetylase activity and maintaining the stable acetylated p53 (p53-Ac) level. DBC1 interacts with breast cancer metastasis suppressor 1 (BMRS1), interrupting DBC1-SIRT1 association and giving a conclusion that BRMS1 can downregulate SIRT1-mediated p53 deacetylation upon DNA damage ^[3][40]. DBC1 becomes phosphorylated by ataxia telangiectasia-mutated (ATM), ataxia telangiectasia, and Rad3-related (ATR) kinases in response to DNA damage. Phosphorylated DBC1 (DBC1-P) binds to SIRT1, leading to the dissociation of the SIRT1–p53 complex and the promotion of p53 activation via its acetylation and p53-dependent apoptosis ^[4][41].

In colorectal cancer (CRC) cells, suppression of nicotinamide phosphoribosyl transferase (NAMPT) decreases SIRT1 activity, which results in the upregulation of p53 acetylation. Subsequently, p53-Ac promotes G0/G1 cell arrest by inducing p21 expression and enhancing caspase-3-mediated apoptosis ^[5][42]. The activity of SIRT1 deacetylase can be enhanced by brahma-related gene-1 (BRG1) protein, a subunit of the SWI/SNF chromatin-remodeling complex. BRG1 binds to SIRT1 and increases SIRT1-mediated deacetylation of p53, leading to its destabilization. Knockdown of BRG1 promotes cell senescence and inhibits CRC growth by modulating the SIRT1/p53/p21 signaling axis ^[6][43].

SIRT1-dependent p53 deacetylation can be inhibited by HDIs, which has significant clinical importance. Preclinical studies focused on p53 deacetylation by SIRT1 showed that some HDIs can be considered as potential anti-cancer treatment tools. Tenovin-6, an SIRT1 inhibitor, in combination with metformin, an activator of AMP-activated protein kinase (AMPK), inhibits growth in non-small cell lung cancer (NSCLC) cells ^[7][15]. Both inhibitors synergistically reduce the expression of SIRT1, increase acetylation and stability of p53, and, as a consequence, induce the expression of p53 downstream target proteins, such as p21, and growth arrest, and DNA damage-45 alpha (GADD45α), which subsequently promotes caspase3-dependent apoptosis ^[7][15].

Except for SIRT1-mediated p53 deacetylation through protein–protein interactions, several miRNAs seem to affect SIRT1 and p53 expression and activity. In colon cancer, the miR-34a/SIRT1/p53 feedback loop is repressed by long-noncoding RNA (lncRNA) HNF1A-antisense 1 RNA1 (HNF1A-AS1). HNF1A-AS1 competitively binds miR-34a and increases SIRT1 expression, leading to the upregulation of the set of proteins involved in canonical Wnt signaling pathway and its activation, which ultimately promotes the metastatic progression of CRC ^[8][9][44,45]. In some cancer cells, miR-34a directly inhibits SIRT1 expression and parallelly induces the expression of p53, which leads to p53-mediated apoptosis and the suppression of cancer development^[10][46]. In turn, a decrease in miR-204 expression in prostate cancer (PCa) cell lines and tissues upregulates SIRT1, enhancing the deacetylation of p53 ^[11][47]. miR-204 promotes the doxorubicin (DOX)-induced p53 acetylation through a decrease in SIRT1 expression. Acetylated p53 (p53-Ac) upregulates the expression of pro-apoptotic proteins Noxa and Puma and induces mitochondrial apoptosis ^[11][47]. The overexpression of miR-590-3p in breast cancer (BC) cells decreases the SIRT1 protein level, leading to an increased p53 level and its acetylation. It results in the upregulation of BAX and p21 expression, ultimately inducing apoptosis and suppressing cell survival ^[12][48]. Moreover, in BC cells, the miR34a, miR34c, and S-adenosyl-L-methionine (AdoMet) enhance p53 acetylation by decreasing SIRT1 and HDAC1 protein levels and potentiate apoptosis induced by AdoMet ^[13][49].

It is known that SIRT3 activity affects p53 protein status in cancer cells. However, several studies reported contradictory results that require further investigation. SIRT3 deacetylates and promotes lipid phosphatase activity of PTEN that reduces mouse double minute 2 homolog (MDM2) transcription and protects p53 from MDM2-mediated ubiquitin-proteasome degradation in BC and CRC cells ^[14][50]. In human hepatocellular carcinoma (HCC) tissues, the overexpression of SIRT3 increases the p53 protein level through the downregulation of MDM2 that reduces MDM2-dependent p53 degradation, indicating that SIRT3 acts as a tumor suppressor ^[15][51]. In contrast, some reports suggest that SIRT3 can decrease p53 protein expression by its direct deacetylation. It promotes p53 binding to MDM2 and increases p53 ubiquitination and degradation, contributing to malignancy in PTEN-deficient cancers (Figure 2G) ^[16][52]. Interestingly, it seems that p53 regulates the SIRT3 level affecting its post-transcriptional modification through proteasome-dependent mechanism ^[17][53]. A significantly higher SIRT3 protein level was observed in p53 wild-type (p53-WT) lung and CRC cancer cell lines in comparison to p53-deficient cell lines. In p53-depleted cells, increased level of S-phase kinase-associated protein 2 (SKP2) E3 ligase and enhanced turnover of SIRT3 were observed, suggesting that p53 regulates SIRT3 protein level through proteasome-pathway that involves SKP2 ^[17][53].

SIRT6 deacetylates p53 at lysine 382, which may lead to ubiquitin-dependent p53 degradation, suggesting SIRT6 potential role in regulating stress resistance and apoptosis ^[18][54]. In HCC cells, SIRT7 interacts and deacetylates p53 at lysine 320 and lysine 373, by which it regulates doxorubicin (DOX)-induced apoptosis. The deacetylation of p53 decreases the Noxa transcription, thus blocking the 53-dependent apoptosis machinery in HHC in vitro model. In the mouse xenograft model, suppression of SIRT7 increases p53 activation induced by DOX, leading to apoptosis and inhibition of HCC growth ^[19][55]. Additionally, upregulation of SIRT7 expression is observed in HCC cell lines and patients’ tissues, correlating with shorter overall survival^[19][55].

In addition to SIRTs, HDAC1 and HDAC8 influence the function of p53 by its deacetylation. In cutaneous T-cell lymphomas (CTLC) treatment with HDIs, such as valproic acid (VPA), trichostatin A (TSA), and PCI-34051, enhances p53 acetylation and mRNA levels of p53-regulated genes Bcl-xl, NOXA, and PUMA that are involved in p53-mediated apoptosis, emphasizing the importance of p53 as a respond target to HDIs ^[20][56]. Consequently, in glioma cells with AT-Rich Interaction Domain 4B (ARID4B) (oncoprotein involved in tumor progression) silencing, HDAC1 is upregulated, which leads to p53-Ac deacetylation suppressing cell apoptosis ^[21][57].

In basal cell carcinoma (BCC) progenitors, the deletion of HDAC1 and HDAC2 significantly inhibits proliferation and enhances cell apoptosis, which is slightly restored after PT53 or P16 knockout (KO). A similar effect has been observed after non-selective HDI (romidepsin) treatment (normal epidermis and lesions cells) ^[22][58]. Considering that deletion of PT53 partly salvages cell proliferation and apoptosis, the moderate effect of another pan-inhibitor, SAHA, towards tumor progression prevention, in a mice xenograft model with p53-depleted BCC cells is well-founded ^[23][59]. All these observations conclude that the efficiency in HDAC1 and HDAC2 inhibition is partly dependent on p53 and p16 status in BCC.

Since p53 mutations frequently appear in pancreatic cancer, they can be considered a potential anti-cancer therapeutic target. The last data showed that HDAC1 and HDAC2 contribute to p53 mutant (p53^R273H) expression in vivo. In the murine pancreatic ductal adenocarcinoma (PDAC) model, the expression of p53 mRNA is reduced in HDAC1/2-deficient cells compared to normal cells, suggesting that HDACs upregulate p53 and maintain the mutant TP53 transcription. Moreover, after SAHA treatment, which targets HDAC1 and HDAC2, p53^R273H expression is downregulated; therefore, SAHA-based therapy could have the potential for further solid tumors research ^[24][60]. In turn, in glioma cells, HDAC1 activity seems to be irrelevant in terms of p53-mutated cells. Unlike p53-WT cells, in p53-mutant cells, cell growth arrest and apoptosis are not observed after HDAC1 silencing. Moreover, in p53-WT cells, HDAC1-KD affects cell phenotype changes in a p53-dependent manner ^[25][61]. These studies show the importance of HDAC1 in glioblastoma, indicating the high need to develop HDAC-specific isoform inhibitors.

Another research confirms the previous observations that HDAC2 could be a proper pharmacological target in tumors bearing p53 mutations ^[26][62]. HDAC2 is downregulated by RCY1, an E3 ligase involved in ubiquitin-mediated proteins degradation, in different cancer cells, including p53-WT, mutant, or p-53 depleted cells. Moreover, HDAC2 expression is enhanced by RCHY1 knockdown, and an inverse correlation between RCHY1 and HDAC2 levels was found in patients’ samples ^[26][62]. In turn, silencing of HDAC2 reduces the ATM/p53-mediated cell death in osteosarcoma cells after DOX treatment, which suggests that HDAC2 takes part in DNA early damage response and could be a coactivator for p53. In detail, DOX-induced cell apoptosis in p53-WT cells is attended by significant p53 and H2A histone family member X (H2AX) accumulation. Silencing of HDAC2 meaningfully deteriorates sensitivity to DOX and reduces p53-mediated DNA damage responses through downregulation of ATM and p53 (at Ser-15) phosphorylation. Taking into account that ATM-mediated H2AX phosphorylation occurs during early DNA damage response, and p53 is a substrate for ATM-mediated propagation of DNA damage signaling, HDAC2 may be involved in ATM activation. Interestingly, silencing HDAC2 in p53-null lung cancer cells shows a weak effect of DOX-mediated phosphorylation of ATM, proving that HDAC2 influence on ATM depends on p53 ^[27][63]. The HDAC2 impact on DNA-damaging agents resistance has also been established for colorectal adenocarcinoma CRC cell lines with different statuses of TP53 (TP53-mutated, TP53-WT, and TP53-deficient cells) by determining the pharmacological effect between them and SAHA or VPA. In untreated p53-mutated cells, the HDAC2 expression is lower in comparison to p53-WT cells. Upregulated HDAC2 expression is associated with drug resistance, and HDAC2 depletion sensitizes multidrug-resistant cells (HT-29) to chemotherapeutic agents (5-FU or oxaliplatin) commonly used in CRC treatment. Combined treatment with SAHA and 5-FU or oxaliplatin downregulates HDAC2 expression, inducing cell apoptosis. The synergetic effect of combined treatment has been validated in vivo, where xenograft tumor growth is reduced by half after drugs treatment. These results speculate that CRC outcomes after combined treatment with HDIs and DNA-damage agents depend more on HDAC2 expression than p53 mutation status ^[28][64]. The effect of VPA has been also investigated against HCC cell lines, showing that VPA downregulates HDAC1/2/3 and upregulated P21 and PT53 gene expression simultaneously, which resulted in cell apoptosis ^[29][65].

The p73 protein, a homolog of p53, induces cell apoptosis and cell cycle arrest ^[30][66]. p73 is rarely mutated in tumors ^[31][67], and in vivo research indicates that p73-deficiency does not increase the tumor incidence. Nonetheless, a few studies reported reduced expression or depletion of p73 in specific tumors, suggesting that p73 acts as a tumor suppressor ^[32][68].

SIRTs participate in this regulation through their activity against p73 and transcription factor E2F1 (E2F1) ^[33][69]. They are able to regulate tumor suppressors directly: as a result of deacetylation of p73 by SIRT1, or indirectly as a result of deacetylation of E2F1, and thus the inhibition of p73 activation ^[33][69]. Thereby, the modulation of the p73 network may help overcome chemoresistance in human cancers, as the inactivation of the TP73 gene is associated with an increased chemoresistance ^[33][34][69,70]. SIRT1 binding to p73 suppresses its transcriptional activity and partly inhibits p73-dependent apoptosis in HeLa cell lines ^[34][70]. Deacetylation of p73-Ac by SIRT1 inhibits p73-dependent BAX transcription suggesting that abnormal expression of SIRT1 promotes tumorigenesis  ^[34][70]. SIRT2 also deacetylates p73 at C-terminal lysines, which suppresses its transcriptional activity and increases tumorigenicity and the proliferation of glioblastoma cells^[35][71]. The overexpression of p73 protein and suppression of SIRT2 activity with one of the specific inhibitors (AGK2 or AK7) results in the induction of apoptosis ^[35][71]. Except for SIRTs, HDAC1 is also involved in p73 deacetylation. A simultaneous ectopic expression of p73 and silencing of the HDAC1 gene in metastatic melanoma cells results in enhancement of apoptosis and autophagy ^[36][72].

The family of forkhead (FOXO) proteins is a set of transcription factors that possess a highly conserved DNA-binding domain called the “forkhead box” (FOX). FOXO family regulates a wide range of cellular processes such as stress resistance, apoptosis, and cell cycle arrest. Among FOXO proteins, FOXO3a is a tumor suppressor regulating cell survival or death in response to chemotherapy and metabolic stress ^[37][38][73,74].

Both SIRT1 and SIRT7 can deacetylate FOXO3a, which prevents its phosphorylation at serine 574 (S574) and blocks lipopolysaccharide (LPS)-induced apoptosis in leukemic monocytes ^[38][74]. The formation of phosphorylated FOXO3a (FOXO3a-P) is regulated by the state of its acetylation, wherein acetylated FOXO3a favorably interacts with c-Jun N-terminal kinase (JNK1), resulting in FOXO3a phosphorylation. Deacetylated FOXO3a is maintained mainly due to SIRT1 and SIRT7 deacetylase activity. SIRT1 and SIRT7 stability is reduced by LPS-induced signaling that involves a mitogen-activated protein kinase (MAPK) pathway, leading to an increase in FOXO3a-P level and induction of cell apoptosis ^[38][74].

FOXO3a can increase Bim expression and stimulate lung cancer cells’ apoptosis, which is partly mediated by early growth response protein 1 (EGR1), which binds to the Bim promoter region. SIRT1 stimulates this pro-apoptotic effect through deacetylation of FOXO3a-Ac, which induces EGR1 binding to the Bim promoter and Bim expression ^[39][75]. Conversely, in glioma cells, a reduction in SIRT1 expression leads to increased acetylation of FOXO3a and a higher expression of Bim and PUMA, resulting in decreased proliferation, viability, and an induction of apoptosis ^[40][76].

The role of the SIRT6 and FOXO3a/bromodomain-containing protein 4 (BRD4) axis is underlined in the progression and drug resistance of luminal BC ^[41][77]. BRD4 co-creates transcriptional machinery controlling the expression of genes critical for tumor progression. Furthermore, SIRT6/FOXO3a/BRD4/ cyclin-dependent kinase 6 (CDK6) axis participates in the development of resistance to Akt inhibitors. Akt inhibitors are able to induce dephosphorylation of FOXO3a and disrupt its interaction with SIRT6, leading to acetylation of FOXO3a (FOXO3a-Ac). FOXO3a-Ac recruits the BRD4/RNAPII complex to the CDK6 gene promoter region, inducing its transcription. The inhibition of either CDK6 activity or BRD4/FOXO3a association overcomes the resistance of luminal BC cells to Akt inhibitors in vitro and in vivo^[41][77].

In human CRC cell lines, FOXO3a positively regulates SIRT6 expression, leading to BAX-induced apoptosis ^[42][78]. The FOXO3a activity is increased by inactivation of Akt, augmenting its affinity and binding to the SIRT6 promoter that in turn enhances SIRT6 expression. SIRT6 induces apoptosis probably through deacetylation of lysine residues on histone 3 (H3K9) in the survivin gene promoter that leads to inhibition of anti-apoptotic protein survivin expression. The SIRT6 knockdown (SIRT6-KD) abrogates apoptotic responses and grants resistance against PI3K inhibitor (BKM120), indicating that inactivation of Akt and induction of SIRT6 may become a novel combined CRC therapy ^[42][78].

In the BC LM2 cell line, the geminin, a DNA replication inhibitor, selectively tethers FOXO3a to HDAC3, which results in the deacetylation and inactivation of FOXO3a transcriptional activity leading to downregulation of the FOXO3a target Dicer, which is an RNase that suppresses metastasis ^[43][79]. The silencing of HDAC3 or depletion of geminin decreases the migration/invasion potential of LM2 cells as well as their metastatic capacity in vivo. These findings indicate the importance of the FOXO3a/Dicer axis as a downstream effector of geminin/HDAC3-dependent BC metastasis^[43][79].

The regulation of FOXO1 activity by SIRT1 yields opposite biological effects in different cancer cells, emphasizing an individual role of HDACs in tumorigenesis and the response to treatment. In BC cells, FOXO1 nuclear localization is regulated by SIRT1 deacetylase activity, where FOXO1 upregulates multidrug resistance protein 2 (MRP2) gene expression. The overexpression of both SIRT1 and FOXO1 enhances transcription of the MRP2 gene, while the inhibition of SIRT1 decreased both the MRP2 expression and FOXO1 nuclear levels that increases the cytotoxic effect of chemotherapeutic agents, such as DOX and paclitaxel (PAX) ^[44][80]. Contrary, in progestin-resistant endometrial cancer (EC) cells, SIRT1 knockout (SIRT1-KO) results in the upregulation of progesterone receptor (PR) and FOXO1, as well as the downregulation of sterol regulatory element-binding protein-1 (SREBP-1) that increases sensitivity to progestin therapy. Therefore, targeting the SIRT1/FOXO1/SREBP-1 pathway that regulates PR may help overcome progestin resistance in EC cells^[45]81].

In gastric cancer (GC), SIRT1 silencing or inhibiting its activity by EX527 enhances expression of FOXO1, pro-apoptotic BAX, and E-cadherin, whereas the expression of Ki67, Cyclin D1, anti-apoptotic Bcl-2, Vimentin, MMP-2 and MMP-9 are downregulated, suggesting SIRT1 involvement in GC progression ^[46][82]. In turn, treatment of HCC cells (HepG2 and Huh7) with cambinol (SIRT 1/2 Inhibitor) or EX-527 (a selective SIRT1 inhibitor) increases the levels of acetylated FOXO1 and p53, reduces cellular viability and migration ^[47][83].

Among HDACs, HATs are also involved in non-histone protein modification. CBP/p300 acetylates FOXM1 at several lysine residues (63, 422, 440, 603, and 614), which is essential for transcriptional activation of its target genes. The FOXM1 acetylation increases the activity of this protein by increasing its stability, DNA binding affinity, and sensitivity to phosphorylation. Additionally, the acetylation of FOXM1 promotes the proliferation of cervical cancer (HeLa) cells and tumor growth in vivo. On the other hand, SIRT1 acts as a negative cofactor for FOXM1 as it can deacetylate FOXM1, which decreases its stability and transcriptional activity. Therefore, activating SIRT1 to deacetylate FOXM1 may become an efficient strategy for treating cancers with overexpressed FOXM1 ^[48][84].