Background

Genome refers to the complete set of genetic information in the form of nucleotide sequence inside the DNA, whereas the epigenome refers to complex modifications inside the genomic DNA [1]. In simple terms, epigenetics involves a set of structural modifications within the nucleic acids and histone that do not involve a change in an individual’s genetic code [2,3,4] and can be termed as ‘on top’ or ‘in addition’ to genetics [5].

Epigenetic mechanisms regulate gene transcription and genomic stability and maintain normal cell growth, development, and differentiation [6,7,8,9,10]. As such, epigenetic regulation is a dynamic and reversible process and epigenetic modifications are carried out by writers (DNMTs, HATs, ubiquitin E3 ligases and HMTs) that catalyze the addition of epigenetic marks onto either DNA or histone tails, readers (bromodomains) that recognizes or are recruited to a specific epigenetic mark and erasers (HDACs, KDMs and deubiquitinating enzymes) that removes the epigenetic marks [11,12,13,14,15,16,17,18,19,20,21].

Though epigenetics is a key component of an organism’s normal development, from embryonic development through adulthood, epigenetic dysregulation can significantly contribute to the origin and progression of human diseases such as cancer, cardiovascular diseases, metabolic diseases and neurological diseases. Extensive explorations conducted to enhance the understanding of the epigenome reveals that localised differences existing in epigenetic states of normal and disease tissues can be utilized as disease biomarkers [22,23,24,25,26].

Literature precedents indicate that all the three families of epigenetic proteins—readers, writers, and erasers are druggable targets. This disclosure coupled with the improved understanding of epigenetics in diverse complications dramatically spurred and expedited the translational investigation of the epigenetic inhibitors. In particular, exhaustive investigations predominantly on small-molecule inhibitors were carried out at the clinical level and the subsequent efforts have culminated in the identification of efficacious inhibitors, with some of them being used in the clinic currently. To add on, the preclinical and preliminary studies have also comprehensively explored the epigenetic tools (DNMT/HDAC/LSD1/DOT1L/BET/EZH2 inhibitors) in pursuit of leveraging enhanced antiproliferative effects. Albeit the clinical stage investigations have been appropriately directed towards the evaluation of the epigenetic inhibitors as single agents, a significant proportion of the efforts is also covered by the studies inclined towards the utilization of the epigenetic tools as a part of combination regimens. This information raises a critical question regarding the therapeutic credibility of some of the epigenetic inhibitors as single agents to attain conclusive benefits in cancer. The doubts are further strengthened by the fact that only seven drugs have approved till date despite the epigenetic targets being at the forefront of the strategized explorations. Nevertheless, the medicinal chemist at the preclinical/preliminary level has been quite proficient to employ rational drug design approaches to maximize the benefits of the predefined pharmacophore models of the epigenetic targets. Indeed, the preclinical/preliminary findings (section) bears a relatively higher degree of fascination for the researchers as efforts invested have not just been confined to elucidate the mechanistic insights responsible for exerting antitumor effects via inhibition of the epigenetic targets, rather the chemist has looked beyond this strategy to attain favourable effects via degradation of the proteins also (PROTACS). Sagaciously evidenced on the literature precedential basis, degradation of the target proteins can be achieved at low exposures by PROTACs (protein degraders) owing to their catalytic mode of action and this emerging approach is likely to steer the wheels of the drug discovery field towards the class of degraders bearing appropriately installed epigenetic tools in the near future. For the selective targeting, the concept of antibody–drug conjugates have also attracted the eyeballs of the researchers working in the field of epigenetic inhibitors. This strategy of targeted drug delivery is anticipated to overcome the issue of systemic toxicity and narrow therapeutic window that limits the clinical use of the available epigenetic inhibitors. CRISPR/Cas9-based strategies to target the cancerous epigenetic regulators also represent an emerging potential approach that is being foreseen as a tool to correct genetic mutations. Other than this, the approach of multitarget assemblage construction has continued garnering significant attention to extract enhanced antitumor effects via concomitant inhibition of the biochemically correlated targets and is also conceived to be one of the preferred futuristic strategies as a potential alternative to the combination therapy. To sum up, it is highly likely that the ship of epigenetic inhibitors will sail through the implementation of the aforementioned approaches.

Despite the significant promise demonstrated by the aforementioned strategies, there is no denying the fact that the conventional approaches will continue receiving tantamount attention of the research groups for the development of new inhibitors. Fragment stitching approach on existing drugs coupled with lead modification studies ascertaining the impact of scaffold installation, regiovariation, bioisosteric replacement, structure simplification approach, structure rigidification approach and other subtle structural variations on the activity profile exemplifies some of these potential approaches.

In light of the current scenario and the amount of efforts currently being invested in this field, it is highly likely that this decade might evidence the therapeutic growth of a handful of epigenetic drugs presently undergoing efficacy and safety evaluations at the clinical level and many new agents might enter the clinic. This review article presents an update of FDA approved epigenetic drugs along with the epigenetic inhibitors undergoing clinical stage investigations. The compilation also encompasses a detailed discussion of the rational strategies that can prove to be instrumental in the development of new inhibitors. The covered literature in this review indicates that the future attempts in the epigenetic drug discovery filed needs to headed in the following directions: (i) explorations of natural product based libraries for the development of non-nucleoside based DNMT inhibitors (ii) initiation of parallel programs on non-metal chelating type HDAC inhibitors as well as anilides to transpose the focus from hydroxamic acid type scaffolds owing to the pharmacological liabilities associated with latter class (iii) exhaustive studies needs to be conducted to ascertain the expression level of epigenetic enzymes in diverse malignancies (iv) fabrication of selective isoform inhibitors of HDAC to extract amplified anticancer effects despite of the fact that the clinical success, till date, have only been attained through pan HDAC inhibitors (v) exploration of additional structural templates other than the framework of tranylcypromine to expand the size of LSD1 inhibitors pipeline (vi) design of dual EZH1/EZH2 inhibitors in view of the fact EZH1, complements EZH2 in mediating H3K27 methylation and is also endowed with HMT activity. (vii) Expanding the size of the libraries of DOT1L inhibitors (viii) utilization of the existing chemical architectures of BET and HDAC inhibitors in the PROTAC model and antibody–drug conjugate model (ix) explorations of combination of epigenetic inhibitors with immunotherapy.

Epigenetics and cancer

Epigenetic processes comprises of inherited, somatic and reversible changes in gene expression in cancer cells. DNA methylation, histone modification (acetylation, methylation, phosphorylation, etc.) and noncoding RNAs are the major epigenetic mechanisms that control gene activity leading to a number of complex cancers [4]. In most of the cancers, DNA is hypomethylated along with the hypermethylation at other sites [27]. The two anomalous processes i.e. hypomethylation and hypermethylation activates oncogenes and inhibits the tumor suppressor genes, respectively [28]. Apart from methylation process, histone modification is another process that plays important role in cancer. Histone modifications control the active and inactive state of chromatin which ultimately influences the gene expression within the former region [29]. MicroRNAs are responsible for degradation of mRNA as well as inhibition of target mRNA through respective complementary base pairing and partial base pairing [30]. All these epigenetic changes start taking place a long time ago before the occurrence of cancer and are considered accountable for any genetic changes in cancer, also labelling them as “first hits” for tumorigeneses [27].

Role of DNA methylation in cancer cells

DNA methylation is an epigenetic process that can be described as the covalent transfer of methyl groups to the fifth carbon of cytosine (5-mC) within 5′-CpG-3′ dinucleotides catalysed by DNMTs with SAM as the methyl donor [31, 32]. In mammals, three major types of DNMT enzymes are found, DNMT1, DNMT3a, and DNMT3b. DNA methylation is appointed as an epigenetic marker that manage the time and location of genes expression in both normal and diseased cells [33]. In cancers like breast, colon, esophageal, lung, pancreas, ovary, prostate, and other cancers, altered patterns of DNA methylation have been observed [34]. The hypomethylation results in re-expression of silenced genes and genomic instability leading to demethylation of two elements that consists of long interspread transposable elements and short interspread transposable elements [35, 36]. Besides hypomethylation, the outcome of hypermethylation is the silencing of TSGs, such as P15INK4b, P16INK4a, P14ARF, CDH1 or EXT1 [37]

Acetylation and deacetylation

It is well known that the acetylation and deacetylation of N-terminal of lysine residue of histone is a critical part of gene regulation and the process is controlled by two enzymes HAT or HDAC [38]. The acetylation results in condensed chromatin structure leading to cell transcription promotion while deacetylation leads to relaxed chromatin causing suppression of gene transcription [39]. This balance between HAT and HDAC manages the chromatin structure and gene expression [40]. Any imbalance in the activity of HAT and HDAC results in cancer. HAT enzyme is associated with various transcription factors like GCN5-related Nacetyltransferase, MYST, and cAMP response element binding protein (CREB/p300) families. Dysbalances in histone acetylation has been evidenced in Rubinstein–Taybi syndrome, glioblastomas, lung cancers, and AML [41]. On the other side any alteration in expression of different isoforms of HDACs also causes various cancers like increased levels of HDAC 2 and 3 is observed in colon cancer, rise in levels of HDAC 1 is observed in gastric cancer while in lung cancer reduced expression of HDAC5 and HDAC10 is observed [42]. Furthermore, over expression of HDAC 1 is reported in prostate and esophageal squamous cell carcinoma [43].

Histone methylation and demethylation

The extent and location of methylation and demethylation of histones is another important parameter that controls the gene transcription. Both lysine and arginine residues are prone to methylation but lysine residues H3 and H4 of histone tail are more liable to methylation [44]. The known sites for methylation that controls gene activation are H3K4, H3K48 and H3K79 whilst H3K9 and H3K27 are the sites for gene inactivation [45]. A group of proteins containing the SET (enhancer of-zeste, trithorax) called HMT is required by lysine for methylation process [46]. Histone demethylation enzymes known as KDMs are divided into two groups based on their sequence homology and catalytic mechanism. These includes FAD-dependent amine oxidases superfamily called LSDs [47] and (2) the JmjC domain, contains α-ketoglutarate-dependent enzymes, KDMs and Fe(II) [48]. Any irregulatory in epigenetic effects of methyltransferase enzymes can result in a variety of malignancies [49].

Protein phosphorylation

Phosphorylation takes place at side chains of serine, threonine, and tyrosine via phosphate ester linkages in which histidine, lysine and arginine squeeze through the phosphoramidate linkages, and through the mixed anhydride linkages that occur at amino acids, aspartame acid and glutamate [50]. Phosphorylation helps in regulation of a number of biological processes like various signalling pathways, gene expression, cell division, etc. while majority of the cellular functions that includes energy storage, morphological changes, protein synthesis, gene expression, signaling factor release, muscle contraction, and biochemical metabolism are controlled and managed by phosphorylation [51]. A number of signalling pathways are controlled by protein and lipid kinases for regulation of normal cell functions [52,53,54,55,56]. The abnormalities in activity of kinases results in a variety of pathological events, amongst which cancer is the most prominent [52, 53, 56].

Ubiquitination

Ubiquitin system in body consists of three main enzymes ubiquitin-activating enzymes (E1s), binding enzymes (E2s), ligases (E3s), and degrading enzymes [57]. Ubiquitination performs the following functions localization, metabolism, function, regulation and degradation of proteins. The diminished activity of E3 ubiquitin ligase due to some mutations can cause various cancers like renal cell carcinoma, breast cancer, etc. On the other hand, the increase in ubiquitination activity results in cervical cancer. Further total elimination of ubiquitination will lead to colorectal cancer and glioblastoma [58].

SUMOylation

Small ubiquitin-like modified proteins (SUMO) are very similar to ubiquitin proteins as the name signifies. The process of SUMOylation of target proteins results in varied localization and binding partners which ultimately influences the three main parameters: the stability of protein, its transport between cytoplasm and nucleus and regulation of transcription [59]. The promyelocytic leukaemia protein and the oncogenic fusion protein PML–retinoic acid receptor-α are first discovered substrates of SUMO and the occurrence of cancer due to SUMO can be well explained on the basis of the above-mentioned substrates. An infrequent haematological malignancy occurs due to PML-RARα that is called acute promyelocytic leukaemia. The SUMOylation of PML, when distorted, leads to the expression of PML-RARα thus causing APL. SUMOylation is neither tumour promoting nor tumour suppressive rather it is a required process for all cells [60].

Noncoding RNAs in cancer cells

Noncoding RNA is a novel class of genes that control regulatory functions in normal development of cells which get changed in tumor cells. Small nucleolar RNA, PIWI-interacting RNA, small interfering RNA, and microRNA are some of the examples of noncoding RNAs and exhibits functions like transcriptional and posttranscriptional gene silencing via selective base pairing with their targets [4]. Approximately 60% of genes that codes for different proteins and maintains the cellular processes are regulated by miRNAs [61]. Recently, it is reported that miRNAs behave as oncogenes by altering the tumor suppressing proteins or TSGs by modulating the levels proteins that exhibit oncogenic potential [62]. Although, all kinds of ncRNAs exhibits important functions in maintenance of different cellular processes but any irregularity in their function and expression may lead to carcinogenesis [63]. Another ncRNA is small nucleolar RNAs whose dysregulation is reported to be involved in tumorigenesis [64]. For instance, snoRNA42 (H/ACA snoRNA) is a type of snoRNA which is overexpressed in lung cancer [65].

Epigenetics tools for cancer therapy in cancer

Owing to the well-established role of epigenetic dysregulation towards the origin and progression of cancer, lot of efforts have been invested towards the development of epigenetic drugs for the treatment of cancer. The extensive research conducted on small molecule inhibitors as epigenetic tools (DNMT inhibitors, HDAC inhibitors, DOT1L inhibitors, LSD inhibitors, EZH2 inhibitors, BET inhibitors) makes it evident that the epigenetic proteins are druggable targets. At present, seven agents in three epigenetic target classes (DNMT, HDAC and EZH2 inhibitors) have been approved by the US FDA for the treatment of diverse malignancies (Fig. 1) and a wide range of epigenetic-based drugs are undergoing clinical trials. These include 5-azacytidine (1, DNMT inhibitor approved for the treatment of MDS) [66], 5-Aza-2-deoxycytidine (2, DNMT inhibitor approved for the treatment of MDS) [66], FK-228 (3, HDAC inhibitor approved for the treatment of refractory CTCL) [67], SAHA (4, HDAC inhibitor approved for the treatment of refractory CTCL) [68], PXD101 (5, HDAC inhibitor approved for the treatment of refractory PTCL) [69], LBH589 (6, HDAC inhibitor for the treatment of multiple myeloma) [70] and tazemetostat (7, EZH2 inhibitor approved for the treatment of metastatic or locally advanced epithelioid sarcoma) [71]. Other than the aforementioned FDA approved agents, an anilide type HDAC inhibitor, chidamide (8), has also been approved by CFDA to treat patients with R/R PTCL [72].

Fig. 1
figure 1

FDA and CFDA approved inhibitors of the epigenetic targets

DNMT inhibitors

DNMT blockade is considered to be a successful strategy for the prevention of aberrant DNA hypermethylation. DNMT inhibitors reactivate the aberrantly methylated TSG, thereby causing cancer cells reprogramming that ultimately lead to proliferation arrest and cell death [73, 74]. Literature precedents indicate that various compounds have been identified both at the preclinical as well as clinical level that can erase abnormal methylation patterns via irreversible inhibition of DNMTs, causing proteosomal degradation [75, 76]. This degradation then leads to attenuation of the neoplastic cell phenotype by inducing cell differentiation and tumor cell death [73, 74, 77]. Generally, the inhibitors of DNMT are categorized in to two classes: nucleoside analogs and non-nucleoside analogs (Fig. 2).

Fig. 2
figure 2

DNA methyl transferase inhibitors (nucleoside and non-nucleoside based)

Nucleoside analogs

Comprising of a modified cytosine ring (nitrogen in place of a carbon at 5), nucleoside analogs can be converted to nucleotides and get incorporated into newly synthesized DNA or RNA. The enzyme DNMT gets bound with the analogs through the formation of covalent complexes that leads to the DNA methylation inhibition [78]. 5-Aza cytidine and decitabine (5-aza-2′-deoxycytidine) represents the nucleoside analogues that have been approved by FDA for the treatment of AML and MDS [74, 77]

5-Aza cytidine

5-Aza cytidine (Cytidine analog) is a ribonucleoside analog that undergoes phosphorylation to get incorporated in to the RNA. 5-Azacytidine can also get incorporated into DNA via the ribonucleotide reductase pathway. At present, 5-Aza-cytidine is undergoing several clinical stage investigations for diverse malignancies. A phase 3 clinical trial of azacitidine conducted in patients with higher-risk MDS demonstrated that azacitidine (75 mg/m2 per day, 7 days every 28 days) increased the OS in comparison to conventional care [79]. The phase 4 clinical investigation of azacytidine was also conducted in patients with higher-risk MDS. The study design involved the administration of azacitidine 75 mg/m2/day for 7 days/28-day cycle for up to six cycles. The results of the study demonstrated that out of the 44 patients enrolled for the study, response-evaluable patient (n = 33) did not achieve complete remission or partial remission. However, haematological improvement was attained in 50% patients. RBC transfusion independence was attained in 12 of 32 patients and platelet transfusion independence was achieved in 7 of 18 patients. Neutropenia (52%) and leukopenia (39%) was observed as the common grade 3–4 TAAEs [80] (NCT01201811). In a phase 3 study assessing the benefits of azacitidine over the conventional care regimens in old patients with newly diagnosed AML, it was observed that the treatment with 5-azacytidine (N = 129, 8.9 months) led to remarkable prolongment of the median OS versus conventional care regimens (CCR) (n = 133, 4.9 months) [81] (NCT01074047).

Azacytidine has also been evaluated in various combination regimens. The combination of azacytidine (75 mg/m2) and standard induction therapy was found to be feasible in older patients with AML [82] (Phase 2, NCT00915252). The phase 2 trial of 5-azacytidine with lirilumab (BMS-986015) in patients with refractory/relapsed AML was terminated as the response rate did not meet the anticipated minimum 30%. (NCT02399917). In a phase 2 trial evaluating the combination of 5-azacytidine and sorafenib in older patients (n = 27) with untreated FLT3-ITD Mutated AML, 78% ORR, 26% CR, 44% Cri/CRp and 7% PR was observed. The median OS was 8.3 months and 9.2 months in the 19 responders. Overall, the results demonstrated that the combination was well tolerated in the specified population [83] (NCT02196857). The study evaluating the advantages of sequential azacitidine and lenalidomide in subjects with R/R AML demonstrated that this regimen was only effective in a minority of patients (only 11%). Moreover, significant toxicity was evidenced in some of the cases and three treatment-related deaths occurred [84] (NCT01743859). In an investigation evaluating the efficacy of sequential azacitidine and lenalidomide, or azacitidine in old patients with newly diagnosed AML, it was deduced that the regimen (sequential azacitidine and lenalidomide) is not favoured over azacitidine administered in conventional dose and schedule. With sequential azacitidine and lenalidomide, one-year survival was 44% (95% CI: 28, 60%) where as the one year survival with azacitidine only was 52% (95% CI: 35, 70%) [85] (NCT01358734). In a phase II study conducted in elderly population of higher risk MDS or AML that were, as such, considered unfit for intensive chemotherapy, the combination of azacytidine with escalated doses of lenalidomoide was not well tolerated and was discontinued in majority of the patients owing to toxicity issues. However, some positive results were observed in terms of cytogenetic response in the study. [86] (NCT01088373). In a phase 2 study conducted recently, prophylactic low-dose azacytidine and donor lymphocyte infusions following allogeneic hematopoietic stem cell transplantation for high-risk AML (n = 30 patients) and MDS (n = 10 patients) was evaluated. The study results demonstrated that azacytidine was well tolerated but was discontinued in 20 patients owing to graft-versus-host disease and relapse. The overall and disease-free survivals were 65.5% (CI 95% = 48.2–82.8) at 2 years. On the basis of these results, it was concluded that 5-azacytidine demonstrated potential as a prophylactic treatment to reduce the risk of post-transplantation relapse [87] (NCT01541280). A clinical study for assessing the efficacy of the combination of lirilumab and azacitidine in patients with MDS was conducted and 10 patients were enrolled for the investigation. Two patients achieved CR, 5 achieved marrow CR and 3 demonstrated SD. Grade > 3 AEs (infection or neutropenic fever) were observed in five patients. Overall, the combination of azacitidine and lirilumab demonstrated clinical activity [88] (NCT02599649). Azacitidine in combination with midostaurin in subjects (n = 14 in phase 1 and n = 40 in phase 2, enrolled) with AML and high risk MDS was also evaluated. The study design involved the administration of azacytidine 75 mg/m2 on days 1–7 and midostaurin 25 mg bid (in cohort 1 of phase I) or 50 mg bid (in cohort 2 of Phase I and in Phase II) orally. The results of the study demonstrated that the combination is safe as well as effective for patients with FLT3 mutations that were not previously treated with other FLT3 inhibitors [89] (NCT01202877). In a phase 2 evaluation assessing the combination of azacitidine and etanercept for the treatment of MDS, azacitidine (75 mg/m2/day for 7 days) was administered to twenty-three patients in combination with etanercept (25 mg sc twice a week for 2 weeks every 28 days). The results of the study indicated that a total of 14 patients responded, with CR evidenced in five patients and PR in 8 patients. A hematologic improvement of neutrophils was observed in 1 patient. Overall, the combination was deduced to be endowed with favourable trends in comparison to azacytidine alone. [90]

In a phase 1 study of azacytidine (monotherapy, combination with carboplatin or nab-paclitaxel) conducted in patients with R/R solid tumors, RP2D was determined as 300 mg (every day, days 1–14/21). PR (three/eight) and SD (four/eight) in patients with nasopharyngeal cancer were observed with CC-486 (oral azacitidine) monotherapy. Overall, the study demonstrated that the drug is well tolerated in monotherapy as well as in combination with carboplatin or nab-paclitaxel (NCT02269943) [91]. A phase 3 study was conducted for the assessment of the platelet supportive effects of eltrombopag administered concomitantly with azacitidine. In comparison to azacytidine alone, the combination of eltrombopag and azacitidine led to the worsening of platelet recovery, with lower response rates. Moreover, increased progression to AML was evidenced [92] (NCT02158936). In another recently conducted phase 2 study, the combination of ruxolitinib and azacitidine was found to be safe. Improvement in bone marrow fibrosis coupled with significant spleen response rate was attained in patients with MF. The study design enrolled 46 patients and involved the administration of ruxolitinib twice per day continuously in 28-day cycles for the first 3 cycles followed by the addition of azacitidine (25 to 75 mg/m2, days 1–5) starting with cycle 4 [93] (NCT01787487).

Recently, Onureg (azacitidine 300 mg tablets, CC-486) was approved by US FDA for the continued treatment of adult patients in first remission with AML. The promising results of the AML-001 study (Phase 3 clinical trial) laid the foundation of FDA approval as statistically significant improvement in OS (10 months, median OS time 24.7 months, 95% CI: 18.7–30.5) compared to placebo (median OS time 14.8 months, 95% CI: 11.7–17.6) was attained by the use of onureg. [94] It is noteworthy to mention that a chemical stable analog of 5-Azacytidine, dihydro-5-azacytidine (DHAC), is also biologically active and is relatively less toxic [95, 96].

Decitabine

Decitabine, another nucleoside type DNMT inhibitor, is a desoxyribose analog of cytosine which only gets incorporated in DNA. Decitabine also leads to DNMT depletion and genome hypomethylation. Like, 5-Aza cytidine, decitabine has also been approved by FDA for the treatment of AML and MDS [74, 77]. In a phase II clinical investigation conducted to evaluate the efficacy of decitabine (IV, 15 mg/m2, 5 days–2 weeks) in patients with CML resistant to imatinib mesylate, 35 patients were enrolled (12 in chronic phase, 17 in accelerated phase, and six in blastic phase). The results of the study demonstrated complete hematologic responses in 12 patients (34%) and partial hematologic responses in seven patients (20%). Six patients exhibited major cytogenetic responses, and 10 demonstrated minor cytogenetic responses and the overall cytogenetic response rate observed was 46%. Major adverse effect evidenced was myelosuppression. Overall, it was concluded that decitabine is endowed with clinical activity in imatinib refractory CML [97]. A phase 2 clinical trial for the assessment of decitabine as maintenance therapy for younger adults with AML was conducted in anticipation that 1 year of maintenance therapy would lead to an improvement of disease-free survival for AML patients < 60 years, who as such were not responsive to allogeneic stem cell transplantation in first remission. The results of the study were not encouraging as the maintenance with decitabine did not exert any benefits [98] (NCT00416598). The dynamics of neoplastic cell clearance during decitabine treatment using quantitative monitoring of mutant alleles by pyrosequencing was investigated. The study results demonstrated that the drug was endowed with a noncytotoxic mechanism of action that leads to altered biology of the neoplastic clone and/or normal cells [99] (NCT00067808). A retrospective analysis was conducted to evaluate the response to decitabine in patients with advanced stage MDS. In the study, outcome of patients with baseline marrow blasts ≥ 20% and < 30% (refractory anaemia with Excess Blasts in Transformation—RAEB-t group) and < 20% (MDS group) were compared. A better duration of response was demonstrated by the patients with MDS (9.9 vs. 5 months; P = 0.024) and OS (16.6 vs. 9.0 months) in comparison to patients with RAEB-t [100] (NCT00043381, NCT00260065). A gene expression analysis to assess the gene expression patterns associated with response to decitabine was conducted in a multicenter phase II trial in older AML patients deemed unsuitable for induction chemotherapy. The results of the study indicated that the efficacy of decitabine is partly dependent on immunomodulatory effects [101] (NCT00866073).

In a Phase II study conducted with an aim to assess tosedostat in combination with cytarabine or decitabine in patients (newly diagnosed older) with AML or high‐risk MDS, 34 patients ≥ 60 years old were randomized and tosedostat (120 mg on days 1–21 or 180 mg continuously) was administered with decitabine (20 mg/m2/d) every 35 d. The study outcome indicates that combination of tosedostat and decitabine was tolerated well and resulted in a CR/CRi rate of > 50%. (NCT01567059) [102]. Recently, an inqovi (decitabine and cedazuridine) tablet for treatment of adult patients was approved by US FDA for the treatment of MDS and chronic myelomonocytic leukemia. The approval was attributed to the results of clinical trial that demonstrated similar drug concentrations between intravenous decitabine and inqovi. It was also observed that a considerable proportion of patients that were previously dependent on transfusions did not require the transfusions during an 8-week period. Moreover, intravenous decitabine displayed a similar safety profile to inqovi [103].

Guadecitabine

Guadecitabine, a next-generation hypomethylating agent, is a dinucleotide antimetabolite of a decitabine linked via phosphodiester bond to guanosine. Guadecitabine prolongs the exposure of tumor cells to the active metabolite, decitabine, leading to an enhanced uptake of decitabine into the DNA of rapidly dividing cancer cells. Guadecitabine also offers resistance to degradation by cytidine deaminase [104]. A study (Phase I/II) with an aim to determine the genomic and epigenomic predictors of response to guadecitabine in R/R AML was recently carried out. The study results indicated a 17% response rate to guadecitabine (2 CR, 3 CR with CRi or CR with CRp in the phase I component and 23% (14 CR, 9 CRi/CRp) in phase II. Peripheral blood blasts and haemoglobin were identified as predictors of response and cytogenetics, gene expression, RAS mutations, and haemoglobin as predictors of survival. [NCT01261312, [105]. In a phase 2 study evaluating the combination of guadecitabine with carbotaxol in heavily pretreated patients (n = 100 enrolled) with platinum-resistant recurrent ovarian cancer, promising activity was attained. No serious adverse events were observed in the study. Neutropenia (67%), leukopenia (25%) and anemia (14%) were evidenced as grade 3/4 events. The efficacy evaluation results were as follows: ORR (16%), DCR (37%), PFS (4.1 months), OS (11 months) [106]. In a phase 2 study conducted in patients with HCC, guadecitabine (45 mg/m2) administered on a 28-day cycle was well tolerated in subjects with HCC previously progressed on sorafenib. The study outcome revealed that potent global DNA demethylation (LINE-1) was observed in blood and tumor DNA. To add on, demethylation was seen in patients on promoter of TSG MZB1, which as such, is silenced in HCC [107] (NCT01752933). Recently, the efficacy and safety of guadecitabine was evaluated in phase III study (ASTRAL-1 study) in adults with previously untreated AML. The patients selected were ineligible for intensive induction chemotherapy. The study design involved the administration of guadecitabine, delivered (SC, 60 mg/m2/day for 5 days) in combination with either azacitidine (IV or SC 75 mg/m2/day, 7 days), decitabine (IV 20 mg/m2/day, 5 days) or low dose cytarabine (SC 20 mg bid, 10 days), administered in 28-day cycles. The results of this investigation revealed that primary end points of CR rate or OS were not met (NCT02348489) [108]. Recently, Astex and Otsuka announced the evaluation results of guadecitabine in phase 3 ASTRAL-2 and ASTRAL-3 studies in patients with previously treated AML and MDS or CML. It is disappointing to mention that the guadecitabine did not improve the OS and the study was unable to meet the primary end point [109].

5-Fluoro-2′-deoxycytidine (FdCyd)

5-Fluoro-2′-deoxycytidine represents another deoxyribonucleoside analog that undergoes phosphorylation and is capable of getting incorporated into DNA. The combination of FdCyd and the CD inhibitor tetrahydrouridine (THU) was evaluated in phase I study conducted in cynomolgus monkeys. The results of the investigation indicated that THU administration with FdCyd led to increase in the exposure to FdCyd and improved PO FdCyd bioavailability from < 1 to 24%. Moreover, THU and FdCyd concentrations achieved after PO administration were found to be associated with CD inhibition and hypomethylation, respectively [NCT00378807] [110]. In another phase I investigation of oral 5-fluoro-2′-deoxycytidine with oral THU in patients (N = 40) with advanced solid tumors, FdCyd was administered for 3 − 7 days q wk × 2 in 21-day cycles in combination with THU (administered, PO 30 min prior to Foci). The results of the study are as follows: MTD: FdCyd (160 mg) + THU (3000 mg), 1 × daily days 1 − 6 and 8 − 13, grade 4 toxicities: thrombocytopenia (1 pt), neutropenia (3 pts) and lymphopenia (3 pts), SD: 19 pts [111]. A phase I study was conducted to establish the pharmacokinetic and pharmacodynamics profile of FdCyd (IV) administered with THU (fixed dose − 350 mg/m2) in subjects with advanced cancer. The results of the study are as follows: MTD: Fdcyd (134 mg/m2) + THU (350 mg/m2), days 1–5 and 8–12 every 4 weeks, Phase II dose determined − 100 mg/m2/day FdCyd with 350 mg/m2/day THU, good plasma exposures and the sustained PR was observed at 67 mg/m2/day [112] (NCT00378807). Recently, another study was carried out to evaluate the efficacy of 5-FdCyd in patients with advanced solid tumors. In the study, 93 patients were enrolled (29 breast, 21 head and neck cancer, 25 NSCLC, and 18 urothelial). The outcome of the study was not satisfactory as insufficient responses were achieved and only three PRs were attained. It is noteworthy to mention that the results were only promising in patients with urothelial carcinoma as the preliminary 4-month PFS rate of 42% was attained in the urothelial stratum. In 69% of the patients evaluable for clinical and CTC response, p16-expressing cytokeratin-positive CTCs were increased. Overall, the results observed in this study indicate exploration of FdCyd + THU in future is warranted in urothelial carcinoma [113].

Zebularine

Other than these FDA approved DNMT inhibitors, zebularine (4-Deoxyuridine, ribonucleoside analog), an oral DNA-demethylating drug has demonstrated stability in acidic environments as well as aqueous solutions. Despite being a potential DNMT inhibitor, its clinical translation has been hindered by the limited bioavalability in (< 7%) and primates (< 1%) along with high dose requirements in millimolar concentrations. [77, 114].

Non-nucleoside DNMT inhibitors

Risk of mutagenicity and genomic instability associated with the use of nucleoside DNMT inhibitors [75] has led to the initiation of numerous investigations with an aim of develo** nonnucleoside analogs. Most of the non-nucleoside DNMT inhibitors developed so far is small molecule agents that directly target the catalytic sites rather than incorporating into DNA. This section presents a brief account of non-nucleoside inhibitors for natural and synthetic sources.

The sponge Pseudoceratina purpurea yields Psammaplin, a non-nucleoside based dual inhibitor of DNMT and HDAC [115]. A Polyphenol from green tea, EGCG ((-)-epigallocatechin-3-gallate reversibly demethylates methyl-DNA leading to the reactivation of multiple key genes (hMLH1, P16, and RA, in colon, esophageal, and prostate cancer cell lines) [116]. A polyphenolic compound, curcumin, has also been reported to induce global hypomethylation in MV4-11 leukemia cell lines possibly through covalently blocking of the catalytic thiolate of DNMT1, inhibiting DNA methylation [117]. Hydralazine and procainamide have demonstrated tumor suppressor reactivating and antitumor actions in breast cancer [118,119,120]. In a phase II study conducted to combat the issue of chemotherapy resistance in refractory solid tumors, addition of hydralazine and valproate to the same chemotherapy schedule that the patients were receiving, yielded clinical benefits in the selected population. [NCT00404508) [121]. A randomized phase III, epigenetic therapy with hydralazine valproate and chemotherapy in patients with advanced cervical cancer was also carried out. The study design involved the administration of hydralazine (182 mg—rapid acetylators, or 83 mg—slow acetylators along with valproate (30 mg/kg). The study was conducted in 36 patients and four PRs to CT (cisplatin topotecan) + HV (hydralazine valproate) and one in CT + PLA were achieved. SD in five (29%) and six (32%) patients was observed whereas eight (47%) and 12 (63%) showed progression (P = 0.27). Moreover, the study indicated substantial benefits in context of PFS [122] (NCT00532818). Other than these small molecule inhibitors, a second generation phosphorothioate antisense oligodeoxynucleotide, MG98 prevents DNMT1 mRNA translation effects and is under detailed preclinical studies and clinical stage investigations (phase I/II clinical trials) in solid tumors [123,124,125]. Another, small molecule inhibitor, RG-108 is reported to directly inhibit DNMT1 catalytic domain and block DNMTs without causing enzyme degradation [115, 126]. Disulfiram was also identified as a DNMTi as it was found to reduce global 5mC levels, as well as demethylate and reactivate the expression of epigenetically silenced TSGs [127]. SGI-1027, a quinolone based compound, exhibited inhibitory potential towards DNMT1, DNMT3A and DNMT3B, leading to demethylation and reactivation of TSGs [128]. Table 1 presents the clinical update of DNMT inhibitors undergoing clinical stage investigations.

Table 1 Clinical update of DNMT inhibitors

EZH2 inhibitors

EZH2, a crux subunit of the PRC2, is a HMT enzyme responsible for methylating lysine 27 (mono-, di- and trimethylation) in histone H3 (H3K27). H3K27me3 is more frequently interlinked with transcriptional repression, and it is a significant epigenetic phenomenon during tissue development and stem cell fate determination. Specifically, functioning of EZH2 in biological processes occurs through 3 types of mechanism viz. PRC2-dependent H3K27 methylation, PRC2-dependent non-histone protein methylation, and PRC2-independent gene transactivation [129,130,131,132,133,134,315]. Yang et al. recently reported HDAC6 degraders via recruitment of VHL (ligand for E3 ligase) instead of CRBN and identified an extremely potent PROTAC 77 as HDAC6 degrader. Overall the results of the study were highly optimistic and indicated that PROTAC can be utilized as a specific chemical probe for HDAC6 degradation to further investigate HDAC6-related biological pathways (Fig. 10) [316].

Small-molecule BET degraders (PROTAC)

Previously conducted structure-guided design of [1,4] oxazepines by Qin et al. led to the identification of QCA276 that was used as a starting point for the design of small-molecule BET degraders (PROTAC). The degraders were synthesized and subsequently evaluated for BET degradation. Owing to the exhaustive pharmacological explorations, one of the compound, QCA570 (78) emerged as an extremely potent BET degrader 78 that demonstrated BET proteins degradation along with cell growth inhibitory activity towards human acute leukemia cell lines at low picomolar concentrations. In leukemia xenograft models, complete and durable tumor regression was attained by the use of QCA570 (78) at well-tolerated dose-schedules [317]. Crew et al. reported ARV-771 (79) as a pan-BET degrader that caused degradation of c-MYC and induced apoptosis of cells through PARP cleavage. The degrader was evaluated in vivo employing the VCaP tumor model and it was observed that the degrader could exert significant tumor growth inhibition. On the whole, ARV-771 (79) demonstrated high potential for treating CRPC compared to enzalutamide [318]. Bardners et al. furnished a BET degrader via conjugation of JQ1 and pomalidomide. The resulting degrader 80 displayed substantial degradation potential towards BRD4 and could also inhibit the growth of tumor as evidenced in an in vivo murine hind-limb xenograft model with human MV4-11 leukemia cells [319]. Wang et al. designed a BET degrader 81 using a previously reported BET inhibitor (BETi-211) for the treatment of TNBC. The degrader targeted BRD2, BRD3 and BRD4 in a dose-dependent manner and inhibited a TNBC cell growth at nanomolar concentration. A time dependent downregulation of MCL1 protein was also exerted by the degrader. Moreover, in the patient-derived xenograft model of TNBC, the degrader demonstrated high efficacy [320]. A BRD9 degrader was generated via conjugation of the VHL ligand and a BRD9 inhibitor by cullin et al. The degrader 82 was found to be high efficacious as it could induce the degradation of BRD9 and BRD7 at nonomolar concentrations. The CRBN based PROTAC also displayed remarkable cytotoxic effects against acute myeloid eosinophilic leukemia and malignant rhabdoid tumor [321]. A BRD9 degrader 83 was designed and synthesized by Bradner et al. that could induce the degradation of the target protein at nanomolar concentration and also exhibited more pronounced antiproliferative effects in the human AML MOLM-13 cell line than the inhibitor (BRD 9 inhibitor) used for the construction of the PROTAC assemblage (Fig. 10) [322].

Challenges associated with PROTAC approach

In light of the aforementioned, PROTAC appears to be a fascinating stratagem that holds tremendous potential to save the sinking ships of epigenetic inhibitors particularly the ones that are considered to be failures as single agents and are being only considered as suitable candidates for combination therapy. Albeit, endowed with significant merits, PROTAC approach also poses some challenges that need to be addressed to extract a cancer therapeutic out of it in the longer run. At present, it is reported that an estimated 600 E3 ligases have unique activity profiles and distribution patterns throughout the body, however only a limited number of ligands have been identified for them. Thus an exhaustive screening program is required to be initiated for identifying the ligands for E3 ligases as picking the right ligase to tag the target protein is extremely imperative for the success of this program. Other than this, no obvious explanation has been provided so far in the majority of the reported studies behind the rational for selection of the linker and it appears a random selection of linkers is usually made. In this context, a more pragmatic inclusion of the linkers is required to ascertain conclusive benefits of this approach. In addition, to unleash the true potential of PROTACs, comprehensive explorations are required to be conducted at the clinical level to gain deeper mechanistic insights of the PROTACs that have passed the preclinical stage.

Multitargeting agents

The concept of multitargeting approach is not a new concept and numerous hybrid scaffolds composed of more than one pharmacophore to exert concomitant modulation of two or more targets have been successfully furnished in the last decade. Indeed, epigenetic inhibitors field should be duly credited for the regained attention towards the scaffolds that are promiscuous and were once considered to be “not the first choice” of the medicinal chemist. Usually a biochemical relation between the two targets or the optimistic results demonstrated by a cocktail of drugs lays the foundation for the construction of such scaffolds. Interestingly, this strategy confers a broad platform to the chemist for the generation of logically constructed scaffolds with enhanced antitumor effects. It is noteworthy to mention that this approach has particularly capitalized on the flexibility of the three component HDAC inhibitory pharmacophore and thus the pipeline of multitargeting epigenetic inhibitors is flooded with the agents that inhibit HDAC isoforms along with the other biochemically correlated target. Moreover, the approach has also been extended towards non-epigenetic target and numerous hybrid scaffolds decorated with one fragment from epigenetic inhibitors and the other fragment form epigenetic or non-epigenetic inhibitor are reported. Decades of extensive research indicates that the chemist has displayed utmost proficiency for rationally designing dual epigenetic inhibitory agents and few representative studies that excellently exemplifies this concept are presented below:

Dual HDAC-HSP90 inhibitor

Synergistic anticancer efficacy attained with HDAC and HSP90 inhibition coupled with the evidenced ability of HDAC inhibitors to induce acetylation and inhibit the ATP binding and chaperone function of HSP90 protein provides a strong rationale for the fabrication of dual HDAC-HSP90 inhibitor [323,324,325,326,327,328,329]. To add on, revelations ascertaining that the beneficial effects that can be attained via targeting of the HDAC6/HSP90 Axis in NSCLC [330] further strengthens the logic behind the design of hybrid scaffolds (85) composed of pharmacophoric features of HDAC as well as HSP90 inhibitors. With this background, a hybrid scaffold was furnished bearing the key structural units of a second generation HSP90 inhibitor, AT-13387 (84) and FDA approved hydroxamic acid, SAHA (Fig. 11). The results of the in-vitro and in-vivo studies demonstrated that the resulting adduct (85) could remarkably inhibit the HDAC6 isoform (IC50 = 4.3 nM) and HSP90 protein (IC50 = 46.8 nM) and could also exert substantial antiproliferative effects against the NSCLC (A549 and H1975). Expressions of signatory biomarkers associated with HDAC6 and HSP90 inhibition were also modulated by the hybrid compound (85). Other than the striking in-vitro antiproliferative profile, the hybrid scaffold was also endowed with tumour growth inhibitory potential as evidenced in human EGFR-resistance NSCLC H1975 xenograft model in vivo [331].

Fig. 11
figure 11

Design of dual HDAC-HSP90 inhibitor

Dual HDAC-DNMT inhibitors

Motivated by the optimistic results attained with a cocktail of DNMT and HDAC inhibitor in context of the anticancer efficacy including the suppression of the tumorigenicity of cancer stem-like cells and enhancing cancer immune therapy [332,333,334,335,336], Yuan et al. designed and synthesized a dual DNMT and HDAC inhibitor C02S (87) (Fig. 12) that demonstrated significant enzymatic inhibitory activities against DNMT1, DNMT3A, DNMT3B and HDAC1. It was quite evident from the results that the hybrid compound could also inhibit DNMT and HDAC at cellular levels via inversion of mutated methylation and acetylation and increased expression of tumor suppressor proteins. Detailed investigation of C02S revealed that it could induce reexpression of p16, p21 and TIMP3 and cause DNA damages, modulate multiple cancer hallmarks simultaneously and exert tumor growth suppression in mouse breast cancer models [337].

Fig. 12
figure 12

Design of dual HDAC-DNMT inhibitor

Dual HDAC-LSD1 inhibitors

Literature precedents underscores the association of gene expression silencing with HDAC1/2 and LSD1 enzymatic activities within the CoREST complex (HDAC complex that includes HDAC1, HDAC2, the scaffolding protein CoREST, and LSD1) that contributes to cancer and other diseases [338, 339]. These notions spurred a group of researchers to design dual LSD1/HDAC inhibitors (Fig. 13) anticipating that such constructs might demonstrate enhanced activity coupled with an improved therapeutic window. The results of the study led to the identification of a dual inhibitor, corin, endowed with magnificent anti-proliferative activity against several melanoma lines and cutaneous squamous cell carcinoma lines. It is noteworthy to mention that the dual LSD1/HDAC inhibitors displayed more pronounced efficacy than its parent monofunctional inhibitors. Detailed investigation of corin (89) revealed that its striking pharmacological profile relied on an intact CoREST complex. In the melanoma mouse xenograft model, treatment with corin (89) resulted in slowing of the tumor growth [340].

Fig. 13
figure 13

Design of dual HDAC-DNMT inhibitor

Dual HDAC-BET protein inhibitors

In light of role of BET and HDAC proteins as central regulators of chromatin structure and transcription coupled with the evidenced efficacy attained with the combined BET and HDAC inhibition in pancreatic ductal adenocarcinoma at the preclinical level [341], a hybrid scaffold composed of the structural features of a BET inhibitor ( +)-JQ1 and class I HDAC inhibitor CI994 was generated (Fig. 14). Remarkable tumor cell proliferation was achieved with the hybrid compound in comparison to ( +)-JQ1, CI994 alone or combined treatment of both inhibitors. The hybrid scaffold (91) demonstrated more pronounced inhibition of HDAC1 isoform and retained a similar inhibitory potency against BRD4 bromodomains as that of ( +)-JQ1. [342]

Fig. 14
figure 14

Design of dual HDAC-BET inhibitor

Dual HDAC-EZH2 inhibitors

Previous reports reveal that EZH2 works in tandem with HDACs in the same protein complex and mediates gene transcription repression by increasing histone H3 Lys [343,344,345,346,347,348]. Owing to this functional link, design of dual EZH2/HDAC inhibitor has been conceived as a rational approach to control a number of epigenetic-dependent carcinogenic pathways. In an attempt to capitalize on this useful information, a group of researchers led by Romanelli et al. designed a first-in-class dual EZH2/HDAC inhibitor (93) (Fig. 15) that displayed a balanced inhibitory potential towards both the targets and also inhibited the proliferation of U937, THP1 (hematological malignancies) RH4 (rhabdomyosarcoma), SH-N-SK (neuroblastoma) and U87 (glioblastoma) cancer cell lines. Moreover, in U937 and RH4 cells, the dual inhibitor caused cell cycle arrest in the subG1 phase, induced apoptosis and increased the expression of cell differentiation markers [349].

Fig. 15
figure 15

Design of dual HDAC-EZH2 inhibitor

Dual HDAC-PI3K inhibitor

In pursuit of attaining synergistic effects from simultaneous inhibition of PI3K and HDAC, Thakur et al. designed quinazolin-4-one based hydroxamic acids using a rational approach for the tetheration of the pharmacophores of both the inhibitors (Fig. 16). Resultantly, some of the hybrid adducts were found to be potent as well as selective against selective against PI3Kγ, δ and HDAC6 enzymes. The dual inhibitors also exhibited cell growth inhibitory effects exhibited against multiple cancer cell lines. One of the most promising dual inhibitor induced necrosis in several mutant and FLT3-resistant AML cell lines and primary blasts from AML patients and was not found to be toxic. The hybrid compound (95) was also endowed with a good pharmacokinetic profile when evaluated in mice via imp administration. [350]

Fig. 16
figure 16

Design of dual HDAC-PI3K inhibitor

Dual HDAC-tubulin

An investigation conducted for evaluating the efficacy of vincristine (microtubule destabilizing agent) and vorinostat (HDAC inhibitor) reported that the cocktail of the aforementioned drugs demonstrates synergistic antitumor effects in vitro and in vivo. These promising results were basically attributed to the alteration of the microtubules dynamics via vorinostat exerted HDAC inhibition [351]. Motivated by these findings, dual inhibitors of tubulin polymerization and HDAC were designed by Lamaa et al.employing the key pharmacophoric features of 1,1-diarylethylenes (isoCA-4) and belinostat (Fig. 17). Subsequent evaluation of the dual inhibitors revealed that two of the inhibitors (98 and 99) were endowed with striking antiproliferative activity mediated through substantial inhibition of tubulin polymerization as well as HDAC8. One of the compounds, 99, could induce cell cycle arrest of cancer cells at the G2/M phase via disruption of microtubule organization. Furthermore, docking study also rationalized the binding of these hybrid molecules with both tubulin and HDAC active sites. The compound 99 also exhibited cell growth inhibitory effects against tumoral cell lines such as K562, PC3, U87, and BXPC3 and was also active against the CA-4 refractory human colon adenocarcinoma cell line HT-29 and possessed acceptable physicochemical properties [352].

Fig. 17
figure 17

Design of dual HDAC-tubulin inhibitor

The aforementioned studies perspicuously highlight that this approach holds substantial promise and optimism and its implementation is expected to be continued to shoulder the progress of new scaffolds as inhibitors of the epigenetic targets. It is evident that single molecule multiple targets”, “multiple ligands” or “hybrid” anticancer agents can enhance efficacy and lower drug resistance as multiple cross talks between the signaling networks are involved in cancer. The dual inhibitors also score over the strategy of using the cocktail of drugs (combination therapy) as they eliminate the need of extensive investigations such as dose limiting toxicity, drug-drug interactions, pharmacokinetics, bioavailability, establishment of the dosage regimen that, as such, are required to be conducted for combination therapy [353]. Endowed with the benefits of lowering the risk of possible drug interactions, simplified drug metabolism, improved drug transport and reduced drug R&D costs, it is anticipated that the multitargeting agents might outshine the candidature of combination therapy for the treatment of cancer. The main challenge in front of the medicinal chemist is to furnish assemblage that can exert balanced modulation of both the targets to produce synergistic antiproliferative effects and this can only be accomplished via careful selection of targets, pharmacophores as well as the site of tetheration.

Isoform selective inhibitors: Employing a structural template to furnish selective isoform inhibitors of the enzymes

Selecting an epigenetic target for exploration of chemical entities only leads to partial accomplishment of the task as there is still a galactic challenge in front of the medicinal chemist to design isoform selective inhibitors of the enzyme. In general, selective isoform inhibitors are conceived to be less toxic and more efficacious than the non-selective ones. In this context, the ongoing wave in the field of epigenetic drug discovery filed is heavily inclined towards the construction of selective isoform inhibitors to attain anti-tumor effects. This scenario can be best explained by the consideration of the recent trends in the HDAC inhibitors field. As such, HDAC inhibitors garnered limelight with the discovery of pan-HDAC inhibitors namely vorinostat, romidepsin, belinostat, and panobinostat that were later approved by FDA for treatment of diverse malignancies [67,68,69,70]. Despite the clinical success evidenced with the pan HDAC inhibitors, their use has been associated with some side effects, such as fatigue, diarrhea, nausea, QTc-interval prolongation and thrombocytopenia [354,355,356,357]. These disappointing revelations spurred the researchers to draw their attention towards selective inhibitors of HDAC isoforms. Attaining isoform selectivity via structural engineering approaches appears to be extremely feasible for the HDAC inhibitors in light of their flexible and modular structural template composed of three parts: CAP-Linker-Zinc binding motif. The efforts invested in the past in this context have been delved into the below mentioned categories and some selected examples are discussed.

CAP modification The modular nature of the HDAC inhibitory pharmacophore allows the chemist to fine tune each component to attain an inhibitor with a selective isoform inhibitory profile. For instance, a cap construct tolerates a wide variety of modifications such a placement of aryl/heteroaryl ring (bicyclic/tricyclic as well as fused rings), cycloalkanes, bridgehead adducts, spirocyclic rings, steroidal/terpenoidal framework and so on. Other than the placement of diverse scaffolds as CAP component, literature survey indicates that the chemist has also employed several strategies to modify the cap construct such as structural simplification approach, site translocation and CAP rigidification or confernment of some flexibility to the surface recognition part (CAP). Recently, one of the study conducted in our laboratory attempted lead modification of MS-275 employing a CAP rigidification approach that culminated in the identification of a potent compound that was not only a more potent inhibitor of the class I HDACs but was also endowed with substantial antiproliferative effects against TNBC (Fig. 18) [358].

Fig. 18
figure 18

CAP rigidification approach

It is noteworthy to mention that the CAP component remains to be the most comprehensively explored part of the HDAC inhibitory pharmacophore not just in pursuit of attaining isoform selectivity but also for induction or amplification of antitumor effects. The HDAC inhibitory pipeline, at present, is also endowed with some candidates that are potent as well as selective isoform inhibitors of the enzyme but are unable to produce anti proliferative effects. Tubastatin, a highly selective HDAC6 inhibitor, exemplifies one of such case where the selective HDAC6 inhibition is just able to induce neurodegenerative effects irrespective of that fact that HDAC6 isoform is overexpressed in various malignancies. In this context, a study was conducted by our research group to modify the cap construct of tubastatin (structure simplification approach) and the resulting compounds displayed excellent activity profile against multiple myeloma coupled with a striking HDAC6 inhibitory profile (Fig. 19) [359]. Thus, the role of CAP construct also appears to the crucial for activating an enzyme inhibitor to demonstrate cellular potency.

Fig. 19
figure 19

CAP modification

The surface recognition part has also demonstrated scope for structural alteration to afford a transposition in the pattern of inhibitory effects from pan-HDAC inhibition to selective isoform inhibitors of the enzyme. For example, a study focusing on the structural alteration of PXD-101 reported that conferring few degrees of rigidity to the surface recognition part (CAP) via placement of bicyclic ring (Azaindoles) in the chemical architecture of PXD-101 instilled a transposition in inhibitory effects from pan-HDAC inhibition to selective HDAC6 inhibition. (Fig. 20) [360]

Fig. 20
figure 20

CAP modification

CAP and Linker modification On similar lines, the linker part has also demonstrated significant accommodative ability and several functionalities have been employed for the design of HDAC inhibitor. Digging the structural architecture of FDA approved inhibitors viz. Vorinostat, belinostat and panabinostat reveals that each of the structure comprises of different type of a linker (long chain alkyl in SAHA, benzyl acrylamide in LBH-589 and benzenesulfonyl acrylamide in PXD-101) yet exerts pan-HDAC inhibition. However, utilization of the same linker of SAHA with an altered CAP construct (chemically bulky) leads to two selective HDAC6 inhibitors namely ACY-1215 [361] and Tubacin [362]. The identification of ACY-1215 and tubacin presents only one example but the fact that diverse combinations of CAP and linkers can be exploited to extract selective isoform inhibition is supported by a plethora of studies. A relative comparison of the template of SAHA with tubacin and ACY-1215 makes it evident that expanding the size of the CAP construct can induce a transposed pattern of HDAC enzyme inhibition from pan HDAC inhibition to selective HDAC6 inhibition. Likewise, taking cognizance of the structural features of another highly potent and selective HDAC6 inhibitor, Tubastatin [363], it can be conceived that a relatively rigid and fused CAP construct requires a change in the chemical nature of linker to exert selective HDAC 6 inhibition (a relative comparison of tubastatin with tubacin and ACY-1215) (Fig. 21).

Fig. 21
figure 21

Clinically and preclinically active HDAC inhibitors

Several structure alteration programs on a lead compound (already reported selective HDAC6 inhibitors) have also been conducted in the past where modifications of cap and linker part were concomitantly attempted. A recent lead modification study on tubastatin excellently exemplifies this case as the rigid CAP construct of tubastatin was modified into ring opened indole ring bearing a flexible dimethyl amino substituent via a structural simplification approach and acrylamide unit as present in FDA approved agents PXD-101 and LBH-529 was installed in the linker region. The results of the study were overwhelmingly positive as the structurally modified tubastatin analogs exhibited significant cellular growth inhibitory effects and maintained their tendency to exert preferential HDAC6 inhibition (Fig. 22).

Fig. 22
figure 22

Ring opening strategy

C. Modification at the zinc binding motifs Like the CAP construct and the linker part, selection of the zinc binding motif is also an area that needs tantamount attention to attain selective inhibition of HDAC isoforms, thereby extracting the anticancer effects in malignancies having overexpressed pattern of those isoforms. Practically, the competition is majorly between two classes of zinc binding motif viz. hydroxamic acid and aminoanilides. As such, the former class clearly outshines the latter in terms of clinical success with 3 hydroxamic acid type HDAC inhibitors receiving FDA approval for use in cancer [68,69,70]. However, chidamide represents the only aminoanilide that has been approved by CFDA to treat patients with recurrent or refractory PTCL [72]. Regardless of the clinical success, the present trend in this field is equally aligned towards the development of both the classes of inhibitors. An obvious explanation to this is the evidenced susceptibility of hydroxamic acids to glucuronide conjugation leading to inactivation and off-targeting whereas some explorations reported that aminoanilides are less prone to glucuronide metabolism and are endowed with high efficacy possibly due to their slow, tight binding and slow disassembling with HDACs [190, 191, 363,364,365,366,367,368]. Another notable variation between the two classes can be observed in their enzymatic inhibitory profile as pan HDAC inhibition and selective HDAC 6 inhibition is mostly exerted via use of hydroxamic acid based HDAC inhibitors while class I HDAC selectivity is usually attained via the fabrication of aminoanilides. Other than these two types of zinc binding motifs, a trifluoromethyloxadiazolyl moiety (TFMO) as a non-metal chelating group has further led to a categorical division of HDAC inhibitors as the metal chelators and the non-metal chelators. TFMO interacts by one of the fluorine atoms and its oxygen with the active Zn2 + atom in the catalytic center and represents another class of zinc binding motif [369, 370] (Fig. 23). In light of the debated candidature of hydroxamic acids and to some extent, amino anilides also, due to the susceptibility to glucuronidation based inactivation, explorations have been accelerated for non-metal chelating class of HDAC inhibitors in recent times. As a result, TMP-269 has emerged as a prototype inhibitor that bears a trifluoromethyloxadiazolyl moiety (TFMO) as a nonchelating MBG and acts as a class IIa HDAC selective inhibitor [369]. The attributes of TFMO bearing adducts act as boon for the medicinal chemist working in this area as utilization of this zinc binding motif can accomplish HDAC inhibitors that are non-susceptible to glucoronidation and can generate selective/preferential class IIa inhibitory adducts (class IIa HDAC bias). To add on, leveraging the TFMO group for furnishing the HDAC inhibitory assemblage presents a rational approach that might challenge the strategy of targeting a metalloenzyme only with a metal chelators. It is anticipated that conclusive therapeutic benefits can be achieved by shifting the inclination towards weak binders of the metal that are duly supported by other structural features to garner favourable interactions with the amino acid residues of the active site. This aforementioned information clearly provides a platform to pragmatically design new chemical entities that can target the malignancies having an overexpressed pattern of class IIa HDACs.

Fig. 23
figure 23

Classification on the basis of zinc binding motif

Overall, the approach to construct HDAC inhibitors will require an in-depth literature survey to extract the information regarding the over-expression of various HDAC isoforms in diverse malignancies and accordingly adducts can be designed to bear carefully recruited components of the HDAC inhibitory model. Consideration of these notions can certainly enable us to create a depository of isoform selective HDAC inhibitors that upon further evaluation such as evaluation of ADME properties, high dose pharmacology, toxicity studies, evaluation in animal models that mimics the human condition can advance to clinical stage investigation.

Fragment stitching approach: Fragment stitching on an existing drug to produce more potent derivatives

Unlike the approach of using a structural template to create a compendium of compounds (via diverse permutations and combinations of key components) to target various isoforms of the enzyme, this strategy simply employs a FDA approved drug to stich diverse fragments that either leads to the induction or potentiation of anticancer effects. Usually a transposition in the isoform inhibitory effects is not desired rather inhibition of the same isoform is capitalized to derive antitumour effects in diverse malignancies. This approach can be best explained by the numerous investigations conducted on a LSD1 inhibitor, tranylcypromine, which has been exhaustively utilized in the past to fill the library of rationally designed analogues with anticancer efficacy (Fig. 24). To exemplify this, some selected structures have been presented in Fig. 24 that were generated via stitching of diverse fragments on the tranylcypromine core. Compound 108 was recently reported by Vianello et al. as potent LSD 1inhibitor that displayed substantial activity in the in vivo model (in the murine promyelocytic leukemia model) on oral administration. The results of the study also demonstrated that compound 108 was well tolerated and led to remarkable prolongment of the survival time of the treated mice (35% and 62% at the doses of 11.25 and 22.50 mg/kg, respectively). [371] Rotili et al. conducted an investigation that led to the identification of LSD1 inhibitors (109 and 110). The inhibitors induced an increase in H3K4 and H3K9 methylation levels in cells, caused growth arrest and apoptosis in LNCaP prostate and HCT116 colon cancer cells. [372] A study by gehling et al. led to the identification of a highly potent and selective LSD 1 inhibitor, compound 111, (< 4 nM biochemical, 2 nM cell, and 1 nM GI50). Compound 111 exhibited cell growth inhibitory effects in a panel of AML cell lines along with notable antitumor potential in a Kasumi-1 xenograft model of AML at the dose of 1.5 mg/kg once daily (administered orally) [373].

Fig. 24
figure 24

Fragment stitching approach

The structures shown in Fig. 24 only represents some of the selected examples that have been furnished using the aforementioned approach, however, the preclinical pipeline of LSD1 inhibitors is endowed with numerous candidates synthesized via fragment stitching approach. It is noteworthy to mention that this approach has been implemented to all classes of epigenetic inhibitors and the main advantage of employing this approach is that substantial amount of information in terms of protocol for synthesis, structure activity relationship as well as the toxicity profile is available and accessible. This information plays a key role in expediting the library generation of compounds and it is conceived that this startegem is likely to be continually employed in the near future to create antitumor assemblages.

Antibody drug conjugates: Accomplished via conjugation of a small molecule inhibitor and a humanized antibody though a chemical linker, ADCs selectively bind to the receptors of tumor cells. Internalization of the receptor–ADC complex usually occurs through the endocytosis pathway. Forth the internalization, the cytotoxic drug is released via cleavage of the linker. This strategy of targeted drug delivery presents enough promise to overcome the issue of systemic toxicity and narrow therapeutic window that limits the clinical use of the available chemotherapeutic agents [374,375,376]. In this context, a study was conducted recently and two antibody-drug conjugates (112 and 113) comprising of a HDAC inhibitor ST7612AA1 and cetuximab employing cleavable and non-cleavable linkers were synthesized. The results of the study were extremely promising and the HDAC inhibitor-antibody conjugate demonstrated efficient internalization in tumour cells. In the in vivo studies, the conjugates exhibited striking antitumor activity (animal models of human solid tumors) without any toxicity that is generally observed with traditional ADCs delivering highly potent cytotoxic drugs. (Fig. 25) [377]

Fig. 25
figure 25

Antibody drug conjugates

Another research group recently furnished an ADC endowed with the potential to deliver a HDAC inhibitor to ErbB2 + solid tumors. In the study, partial reduction of trastuzumab with tris [2-carboxyethyl] phosphine was performed followed by conjugation to ST7464AA1, the active form of the prodrug HDAC inhibitor ST7612AA1 via a maleimide-thiol linker to furnish the target ADC ST8176AA1. The detailed biological evaluation of ST8176AA1 revealed a similar receptor binding and internalization of ST8176AA1 to trastuzumab. However, the conjugate demonstrated higher anti-tumor activity compared to trastuzumab (in vitro assay) that correlated with increased acetylation of histones and α-tubulin. Moreover, similar to trastuzumab, ST8176AA1 also increased the expression of ErbB2 and estrogen receptor in TNBC cells. (Fig. 25) [378]

These aforementioned studies provide a strong rational to extend this concept to the targeted delivery of epigenetic inhibitors that, in turn, can demonstrate fruitful results in cancer by exerting epigenetic modulation at a much safer dose in comparison to standard epigenetic inhibitors based therapeutic tools. In this context, it appears logical to dig the preclinical pipeline of the epigenetic inhibitors and select some candidates that could not advance to the clinical stage investigation owing to the toxicity related hindering factors. Once selected, their accommodation in ADC model should be attempted.

(CRISPR) Cas9

Designing the CRISPR/Cas9-based strategies to target the cancerous epigenetic regulators in a more specific manner is an emerging potential approach that has garnered tremendous interest in the recent past as a tool to correct genetic mutations. CRISPR-based epigenome editors (CRISPR epi-editors) consist of dCas9 and epigenetic effector (fused or non-covalently) [379,380,381,382] and are being given serious consideration as a practical approach in cancer gene therapy as they can activate the tumor suppressor genes and also inhibit the tumor driving genes. On the literature precedential basis, it is well known that epigenetic mechanisms sometimes inactivate tumor suppressors in some cancers. To add on, the epigenetic factors, such as LSD1, EZH2 and NSD2 (tumor drivers,) are overexpressed by either epigenetic or genetic mechanisms in diverse malignancies. In light of these revelations, development of CRISPR/Cas9 based transcriptional regulators appears to be a pragmatic way to: (1) suppress the expression of the aforementioned enzymes in some cancers (2) target the driver genes of cancer as well as the genes essential for cancer maintenance or drug resistance (3) target the “undruggable” epigenetically silenced tumor suppressors [379,380,381,382,383]. In a nutshell, dCas9-fused epigenetic regulators holds enough promise for cancer treatment as they can reversibly manipulate the epigenetic patterns and also regulate the oncogenes and tumor suppressors expressions.

Conclusion and future perspective

Significant advancement has been made in the field of the epigenetic drug discovery in the last decade. All the major classes of the epigenetic tools have received considerable attention. In context of the DNMT inhibitors, the FDA approved DNMT inhibitors 5-azacytidine and decitabine were exhaustively explored in AML and MDS at the clinical level and optimistic results were attained particularly in combination therapy. Other than the FDA approved nucleoside based DNMT inhibitors, guadecitabine, a dinucleotide antimetabolite of a decitabine linked via phosphodiester bond to guanosine and 5-fluoro-2′-deoxycytidine (FdCyd) also demonstrated promise and are likely to be the subjects of number of investigations in diverse cancer types in the near future. In addition, non-nucleoside prototypes namely curcumin, hydralazine, procainamide, disulfiram, SGI-1027 and ((-)-epigallocatechin-3-gal were evidenced to be endowed with interesting antitumor profiles. In light of the current trends in medicinal chemistry, it appears that the balance in context of the future research will be tilted towards the natural product based non-nucleoside inhibitors owing to the beneficial trends evidenced with curcumin coupled with the fact that the chemical architecture of this phytoconstituent comprises of several sites accessible to generate an appropriate sized structure pool. For the EZH2 inhibitory field, several drug candidates are undergoing clinical stage evaluation with GSK126, GSK343, CPI-205, ZLD1039, PF-06821497, UNC1999 representing a few of them. It is noteworthy to mention that the clinical success of tazemetosat has brought the spotlight on EZH2 as a potential target for the design of cancer therapeutics. The studies conducted also indicates that significant benefits can be attained through the simultaneous inhibition of EZH1/EZH2 and thus the dual targeting of the aforementioned is likely to be explored more comprehensively. Moreover, the synergistic effects anticipated to be attained through the combination of EZH2 inhibitors with immunotherapy or chemotherapy are also likely to be evaluated. In context of the preliminary/preclinical exploration, the chemical architectures of tazemetostat offer several sites for chemical alterations that can be exploited to create library of new inhibitors. The area of development of DOT1L inhibitors have relatively received less attention in comparison to the other epigenetic tools and pinometostat, a potent and selective small molecule DOT1L inhibitor, stands as the only DOT1L inhibitor that has been a subject of significant clinical attention. In this context, there is an urgent need to load the armoury of the DOT1L inhibitors with new entries that can maximize the therapeutic benefits of inhibiting this epigenetic target in cancer. HDAC inhibitors as small molecule inhibitors have always been at the forefront of investigations and a number of inhibitors belonging to this class of hydroxamic acids and anilides were tested in diverse malignancies both as single agents as well as the part of combination regimens. While these two classes are expected to continue being evaluated clinically as well as preclinically, it is also anticipated that some initiatives will be taken to probe the candidature of the non-metal chelating class of HDAC inhibitors as a potential therapeutic weapon in cancer. A complete profiling of TMP-269, prototype non-metal chelating HDAC inhibitor, might open an avenue for the construction of similar scaffolds. As the non-metal chelating class is expected to be relatively free of the pharmacological liabilities of the metal chelating class of HDAC inhibitors, safer therapeutics can be established via furnishment of agents that binds weakly with the metals but relies on interactions from other parts of the pharmacophore to afford acceptable levels of isoform inhibition. Progress in the field of LSD1 inhibitors has basically relied on the fragment stitching approach on the chemical architecture of tranylcypromine to access new frameworks capable of producing antiproliferative effects. OPY-1001, IMG-7289, GSK-2879552 and ORY-2001 are the tranylcypromine based LSD1 inhibitors that are undergoing clinical stage investigation in different cancer types. Futuristic attempts are likely to be directed towards the hydrazide type LSD1 inhibitors owing to the striking antitumor potential demonstrated by SP-2509. OTX-015, CPI-0610, BMS-986158, GSK525762, INCB054329, INCB057643, ODM-207, RO-6870810, BAY1238097, CC-90010, AZD-5153, PLX-51107, SF-1126 represents the BET inhibitors in various phases of clinical evaluation. It is quite hopeful that a comprehensive profiling of these BET inhibitors might yield a cancer therapeutic in the longer run. In addition, some preclinical studies have excellently utilized the BET inhibitory scaffolds for installation in the PROTACS model. The optimistic activity profile of the BET inhibitor based degraders can also spur numerous investigations in this direction. To sum up, the armoury of clinical and preclinical epigenetic inhibitors is filled with numerous candidates, however, conclusive benefits can only be ascertained after the completion of all the phases of the clinical trials.

Overall, the findings covered in this perspective present recent advances in the field of epigenetic drug discovery. It is apparent that future research will include parallel programs aimed at the completion of the ongoing clinical studies, initiation of new clinical trials and development of new inhibitors. In addition, numerous brain storming sessions involving expertise of interdisciplinary teams composed of chemist, biologist, and formulation chemist along with researchers well versed with computational aspects of drug design needs to be conducted. Such programs can identify the hurdles and the hindering factors that have halted the clinical growth of several inhibitors of the epigenetic targets and present logical ways to strengthen the candidature of epigenetic targets in context of capacity to generate single target agents.