Tuning epigenetics to enhance cancer virotherapy

Abstract

Cancer remains a major global health challenge, driving intensive research into novel therapeutic strategies. Oncolytic viruses (OVs) including herpesviruses, adenoviruses and echoviruses, have emerged as promising agents for cancer treatment, with several achieving regulatory approval. These viruses selectively lyse tumor cells and stimulate antitumor immunity. Nevertheless, their efficacy is constrained by limitations such as neutralization by serum factors and tumor cell resistance. Concurrently, epigenetic inhibitors modulate gene expression to suppress tumor proliferation and reshape immune responses. Recent advances highlight the strong synergistic potential of combining epigenetic inhibitors with oncolytic viruses. This strategy enhances viral replication and tumor cell killing, reprograms the tumor microenvironment, and ultimately improves therapeutic efficacy. Consequently, the integration of epigenetic modulators with OV therapy represents a promising frontier in oncology. This review comprehensively examines epigenetic inhibitors, with emphasis on their combined application with oncolytic viruses for cancer treatment.

Graphical abstract

Epigenetic regulation modulates oncolytic virotherapy to enhance antitumor efficacy.

1. Introduction

Cancer persists as a critical global health burden, and effective treatment remains particularly challenging for patients with advanced and metastatic disease^1,2. Oncolytic virotherapy emerged as a promising antitumor strategy with early preclinical and clinical studies in the 1950s–1960s3, 4, 5. However, progress stalled during the 1970s–1980s due to limitations in virological understanding, cancer biology knowledge, and technological capabilities3, 4, 5. Advances in these domains, coupled with breakthroughs in genetic engineering and biotechnology, have since driven a remarkable resurgence in oncolytic virus (OV) research over the past two decades3, 4, 5. Numerous OV platforms, including engineered herpesviruses, adenoviruses, and vaccinia viruses are now in clinical development. Key regulatory milestones include China’s 2005 approval of adenovirus H101 for esophageal cancers, the 2015 U.S. Food and Drug Administration (FDA) approval of talimogene laherparepvec (T-VEC) for melanoma, and Japan’s 2021 authorization of G47Δ for malignant gliomas6, 7, 8.

OVs selectively replicate in and lyse tumor cells while sparing normal tissues⁹. Beyond direct oncolysis, they stimulate antitumor immunity. Infection-triggered immunogenic cell death releases tumor associated antigens (TAAs) and danger signals, activating antigen presenting cells and subsequently priming CD4⁺/CD8⁺ T cell mediated tumor clearance^10,11. Importantly, these dual cytolytic and immunostimulatory properties render oncolytic virotherapy particularly attractive for advanced and metastatic cancers, which are often refractory to conventional treatments and characterized by profound immune evasion and therapeutic resistance¹². Nevertheless, therapeutic efficacy faces significant challenges. Systemic delivery is hampered by serum factor neutralization, reticuloendothelial clearance, nonspecific blood cell binding, and elevated tumor interstitial pressure¹³. Furthermore, intrinsic tumor cell resistance, host antiviral defenses, and immunosuppressive stromal components restrict viral replication and spread¹³.

Epigenomic alterations play a key role in cancer metastasis. Processes such as dynamic chromatin remodeling and large-scale regulatory changes determine tumor cell state transitions and therapeutic responses¹⁴. Concurrently, epigenetic inhibitors constitute a distinct therapeutic class. These agents modulate gene expression through interference with DNA methylation and histone modifications, disrupting oncogenic signaling to suppress tumor proliferation^14,15. DNA/histone methyltransferase inhibitors directly impede cancer cell growth¹¹. Notably, many epigenetic modulators exhibit immunomodulatory properties, histone deacetylase (HDAC) inhibitors enhance antigen presentation and immune cell activation, amplifying antitumor immunity¹⁰.

Cancer immunotherapy has evolved rapidly through advances in immune evasion targeting, oncolytic virotherapy, epigenetic regulation, and combination strategies^7,16. These developments highlight the urgent need for innovative approaches to overcome current therapeutic limitations^7,16. The integration of epigenetic inhibitors with OV therapy demonstrates potent synergy (Fig. 1). HDAC inhibitors suppress tumor cell antiviral responses, facilitating enhanced viral entry and replication¹⁰. Preclinical studies confirm that combining HDAC inhibitors with herpesviruses, adenoviruses, or vesicular stomatitis viruses significantly amplifies viral propagation and oncolysis across diverse tumor models. This combination further cooperatively induces tumor cell death: OVs mediate direct cytolysis while epigenetic agents sensitize cells to infection, intensify pro-apoptotic signaling, and promote apoptosis/necrosis¹⁷. Critically, dual therapy reprograms the tumor microenvironment (TME) by modulating cytokine/chemokine networks, enhancing immune cell infiltration/activation, and counteracting immunosuppression^10,11. This synergistic paradigm represents a transformative approach in oncology, offering substantial potential to expand clinical options and improve outcomes for cancer patients.

Figure 1.

Epigenetic regulation modulates oncolytic virotherapy to enhance antitumor efficacy. Key epigenetic mechanisms including DNA methylation, histone modification, chromatin remodeling, and non-coding RNA regulation, affecting viral replication, tumor cell death, immune response, and viral targeting. These processes alter chromatin accessibility, gene expression, and cellular susceptibility, offering synergistic opportunities for combining epigenetic drugs with oncolytic viruses.

In this review, we first introduce the major classes and biological functions of epigenetic inhibitors, and then systematically discuss their synergistic application with OVs in cancer therapy. Finally, we provide an outlook on the current challenges and future directions of this combinatorial approach, with a particular focus on its translational and personalized therapeutic potential.

2. Oncolytic viruses

2.1. Sequential antitumor mechanisms of oncolytic viruses

Oncolytic viruses (OVs) constitute a class of viruses that selectively infect and lyse tumor cells⁹. Their therapeutic potential has established them as a significant focus in cancer immunotherapy and gene therapy research^4,18. OVs are categorized into two primary types: natural strains and genetically modified variants. Natural OVs include enteroviruses, Newcastle disease virus (NDV), poliovirus, and measles virus (MV). Most clinically investigated OVs, however, are genetically engineered to enhance tumor cell tropism, increase selective replication and oncolytic potency, and stimulate host antitumor immune responses¹⁸. Although the concept of oncolytic virotherapy originated decades ago, substantial progress in molecular biology and genetic engineering techniques has enabled rigorous investigation and clinical application of OVs^19,20.

OVs exert antitumor effects through sequential mechanisms. Following administration via routes such as intratumoral or subcutaneous injection, OVs selectively infect tumor cells through inherent tropism and engineered targeting^21,22. Intracellular replication produces viral proteins that disrupt cellular functions, activate oxidative stress responses, and induce autophagic pathways^23,24. Concurrently, viral infection activates immune responses: antigen-presenting cells process viral antigens and present them to activate CD4⁺ and CD8⁺ T cells, thereby enhancing antitumor immunity^22,25,26. Tumor antigens released during cell death further amplify local immune responses. New viral progeny generated within the TME subsequently infect adjacent tumor cells, perpetuating the antitumor cycle²⁷. Through these coordinated mechanisms, direct oncolysis coupled with immune activation, OVs induce durable antitumor effects, underscoring their promise as cancer therapeutics.

2.2. Constraints and optimization directions of oncolytic viruses

Despite their potent cytolytic and immunostimulatory properties, the efficacy of OVs is frequently constrained by intrinsic antiviral defenses, tumor heterogeneity, and an immunosuppressive tumor microenvironment²⁸. These limitations have stimulated growing interest in epigenetic modulation as a strategy to reprogram tumor susceptibility and immune responsiveness, which will be discussed in the following section.

3. Classification and description of epigenetics

Epigenetic regulation, including DNA methylation, histone modifications, chromatin remodeling, and non-coding RNA, governs gene expression without altering the DNA sequence²⁹. These mechanisms shape chromatin structure and cellular behavior, offering therapeutic potential in cancer (Fig. 2). Key epigenetic modulators and their roles in enhancing oncolytic virus-based therapies are summarized (Table 1).

Figure 2.

Major mechanisms of epigenetic regulation. Epigenetic regulation controls gene expression via chromatin remodeling, histone modifications, DNA methylation and non-coding RNA. Target inhibition reverses these changes to enhance transcription and the therapeutic response. DNMTs: DNA methyltransferases; HDACs: histone deacetylases.

Table 1.

Epigenetic types and representative inhibitors.

Type	Target	Epigenetic drug	Mechanism	Effect	Model
DNA methylation modification	DNMT	Azacytidine	Cytosine analog → incorporated into DNA/RNA → DNMT activity↓, degradation↑	Tumor cell presentation ↑, immunogenicity ↑, apoptosis↑, cell cycle arrest	MDS30,31, hematologic malignancies32, CIC33
		Decitabine	Cytosine analog → incorporated into DNA → DNMT activity↓, degradation↑	T-neopeptide↑, endogenous retrovirus activation↑, DNA damage↑, anti-tumor immunity↑, apoptosis↑	MM34, AML35,
		Guadecitabine	Cytosine analog, precursor of DAC	MHC-I↑, CD8+ T cell recruitment↑, ICB-mediated T cell regeneration↑, anti-tumor immunity↑	Breast cancer36
Histone modification	HDAC	SAHA	Binds HDAC via hydroxamic acid →chelates zinc ion → inhibits deacetylation	Tumor antigen presentation ↑, immune cell function ↑, apoptosis ↑, angiogenesis ↓	Retinoblastoma37
		LBH589	Binds HDAC via hydroxamic acid → chelates zinc ion → inhibits deacetylation	PD-L1 expression ↑, tumor growth↓	Melanoma38
		TSA	Binds HDAC via hydroxamic acid → chelates zinc ion → inhibits deacetylation	Immune activation↑,apoptosis↑	Retinoblastoma37
	EZH2	CPI-1205	SAM-competitive inhibitor → EZH2 activity↓, H3K27me3↓	CD8+/CD4+ T cell function↑, TME remodeling↑, tumor growth↓	Colorectal cancer39
		GSK126	Binds EZH2 catalytic site → blocks SAM binding → H3K27me3↓	Tumor proliferation↓, macrophage polarization toward M1↑	Colorectal cancer40
		EPZ6438	Inhibits EZH2 catalytic activity → H3K27me3 deposition↓	Tumor proliferation↓, macrophage polarization toward M1↑	Colorectal cancer40
		EPZ005687	Covalent binding to EZH2 amino acid residues → EZH2/PRC2 inhibition	H3K27 methylation↓, selective killing of EZH2-mutant lymphoma cells	Mutant lymphoma cell41
	PRMT	PT1001B	Inhibits PRMT via competitive/non-competitive substrate binding	PD-L1↓, CD8+ T cell infiltration↑, apoptosis↑, tumor growth↓	PDAC42
		JNJ-64619178	Prolonged binding to SAM and substrate pocket → PRMT5 activity↓	Cancer cell proliferation↓, tumor growth ↓	Lung, breast, pancreatic and hematologic malignancies43
		PF-06939999	Competitive binding to SAM site → PRMT5 activity↓, SDMA↓	Cancer cell proliferation↓, altered splicing of precursor mRNAs	NSCLC44
	HAT	C646	Binds to acetyl-CoA pocket → blocks acetyl transfer to histones	Oncogene expression↓, cell cycle arrest↑, apoptosis↑	AML45
		CPI703, CPI644	Binds to CBP/p300 bromodomain → transcription of H3K27ac↓	FOXP3 and other Treg genes↓, Treg differentiation↓, Th17 cytokine production↓	Colorectal cancer46
		I-CBP112	Competitively binds CBP/p300 substrate site	Leukemia colony formation↓, leukemia cell differentiation↑	AML47
	LSD	Tranylcypromine	Binds to flavin adenine dinucleotide site of LSD1 → inhibits demethylation	HSV-1 replication↓, antiviral effect↑	HSV-1 infection-induced lethal encephalitis and corneal blindness 48
		OG-L002	Blocks H3K9 demethylation → maintains epigenetic repression	Viral load↓, gene expression↓, prevents EHV-1 reactivation	EFKCs/PB leukocytes49
Chromatin remodeling	SWI/SNF	BRM-014	Inhibits ATPase activity of BRM/BRG1 → SMARCA2 function↓	Tumor cell survival↓	Leukaemia50
	ISWI	NVS-CECR2-1	Binds CECR2 subunit → disrupts ISWI complex interaction	Cytotoxicity↑, apoptosis↑, CECR2-independent mechanisms	Colorectal cancer51
Non-coding RNA	lncRNA/miRNA	ASO	Hybridize target RNA → RNase H cleavage/gene silencing	lncRNA/miRNA↓, tumor migration↓, target cleavage↑	Breast cancer52
	miRNA	miRNA mimics	Restore downregulated miRNA expression → repress epigenetic enzymes	DNMT3A/B↓, SIRT1↓, tumor growth↓, metastasis↓	Breast cancer53
Others	BETi	PLX51107	Inhibits BET binding to acetylated histones↓	T Cell activity↑, PD-L1/IDO in TME↓, tumor growth↓	Melanoma54
		OTX015	Occupies acetyl pocket of BRD2/3/4 → downstream signaling↓	c-Myc↓, cell cycle arrest↑, tumor cell proliferation↓	Disseminated human myeloma55, MPM56
	IDH mutant inhibitor	AGI-5198	Inhibits IDH1 → D-2-HG↓	Metabolic reprogramming↑, tumor cell proliferation↓	Glioma236
		IDH305	Binds mutant IDH1 active site → D-2-HG↓	Restores α-KG-dependent pathways↑, epigenetic regulation↑, tumor growth↓	AML and MDS57

3.1. DNA methylation

DNA methyltransferases (DNMTs) function as central epigenetic regulators, critically implicated in aging-related disorders such as Alzheimer’s disease, myelodysplastic syndromes (MDS), and diverse cancers including colon, breast, and bladder malignancies^10,93, 94, 95. DNMTs catalyze methylation at the 5′-position of cytosine within CpG dinucleotides to form 5-methylcytosine (5 mC), leading to transcriptional repression^96,97. Mammals express six DNMT isoforms: DNMT1, DNMT3A, DNMT3B, DNMT3C, DNMT3L, and DNMT2. DNMT1 maintains methylation patterns⁹⁸, while DNMT3A and DNMT3B establish de novo methylation⁹⁹. DNMT3C’s function remains less defined but may relate to reproduction¹⁰⁰. DNMT3L lacks catalytic activity but enhances DNMT3A/3B function allosterically¹⁰¹, and DNMT2 methylates tRNA rather than genomic DNA¹⁰².

DNMT inhibitors (DNMTis) suppress tumor suppressor gene hypermethylation by inhibiting DNMT enzymatic function¹⁰³. These compounds comprise nucleoside analogs and non-nucleoside analogs. First generation FDA approved DNMTis, the cytidine analogs azacitidine (AZA) and decitabine (5-aza-2′-deoxycytidine, DAC) are clinically utilized for MDS, hematologic malignancies, and solid tumors^30,32,104. Both agents incorporate into nucleic acids to sequester DNMTs, with AZA targeting both DNA and RNA versus DAC’s DNA specificity, conferring distinct therapeutic applications¹⁰⁵.

In addition to direct gene reactivation, DNMTis enhance tumor immunogenicity through neoantigen presentation. Guadecitabine upregulates antigen presentation machinery and immune co-stimulatory molecules such as CD80/CD86/CD40^36,106. DNMTi-mediated induction of C-X-C chemokine receptor three ligands promotes T cell recruitment, while enhanced major histocompatibility complex class 1 (MHC-I) antigen presentation activates CD8⁺ T cell responses³⁶. Additionally, DNMTis induce immunogenic t-neopeptides from endogenous retroviral elements. Goyal et al.³⁵ demonstrated DAC induced t-neopeptide-specific T cell responses in acute myeloid leukemia (AML) patients via comprehensive transcriptomic analysis.

Besides immune modulation, DNMTis further promote tumor cell apoptosis and cell cycle arrest¹⁰⁷. These inhibitors activate DNA damage response pathways, evidenced by decitabine-induced γ-H2AX foci formation, G0/G1 or G2/M arrest, and caspase-dependent apoptosis in multiple myeloma models³⁴.

3.2. Histone modifications

Histone modifications represent a versatile epigenetic code that orchestrates chromatin organization and gene expression, thereby influencing tumor development, immune dynamics, and viral susceptibility. These reversible marks—principally acetylation, deacetylation, methylation, and demethylation—are installed or removed by specialized “writers” and “erasers”, whose dysregulation underlies oncogenesis and immune escape¹⁰⁸. Beyond classical modifications, emerging marks such as lactylation further bridge metabolic states to immune regulation¹⁰⁹. Histone lactylation, a recently identified post-translational modification derived from intracellular lactate, has been shown to directly link altered tumor metabolism to transcriptional reprogramming of immune-related genes, thereby influencing macrophage polarization and inflammatory responses^110,111. Therapeutically, targeting histone modifications offers opportunities to reprogram tumor immunity and enhance sensitivity to oncolytic virotherapy.

3.2.1. Acetylation

Histone acetylation loosens chromatin and promotes gene transcription, immune responses, and antiviral defense¹¹². This process is catalyzed by histone acetyltransferases (HATs) and reversed by histone deacetylases (HDACs)¹¹³. HATs include GNAT (GCN5, PCAF), MYST (MOF, Tip60), CBP/p300, and Rtt109 families, each with distinct roles in transcription and repair^46,114, 115, 116. Aberrant HAT activity promotes oncogenesis and immune evasion through chromatin remodeling¹¹⁷. HAT dysregulation in cancer promotes oncogene expression, cell cycle progression, and stemness maintenance¹¹⁸. HAT inhibitors (HATis) such as C646 block CBP/p300 acetyl-CoA binding to suppress oncogene transcription and induce apoptosis⁴⁵. I-CBP112 inhibits CBP/p300 bromodomain, reducing leukemia colony formation and promoting differentiation⁴⁷ Bromodomain inhibitors (CPI703, CPI644) modulate regulatory T cell (Treg) and Th17 cells via FOXP3 and H3K27ac suppression, reshaping the immune microenvironment^46,119. HATis also show promise in inflammatory diseases like asthma and arthritis^120,121.

3.2.2. Deacetylation

HDACs remove acetyl groups to condense chromatin and silence gene expression^122,123. Dysregulation contributes to cancers and immune dysfunctions^124,125. HDACs are categorized into four classes based on sequence homology and cofactor dependence¹¹². HDACs are classified into Class I (HDAC1/2/3/8)¹²⁶. Class IIa (HDAC4/5/7/9) and IIb (HDAC6/10)¹²⁷. Class III (SIRT1–7)^128,129, and Class IV (HDAC11)¹³⁰. HDACis, such as suberoylanilide hydroxamic acid (SAHA) and LBH589, induce histone hyperacetylation, reactivating silenced genes and enhancing immune surveillance^131,132. HDACis upregulate MHC-I and promote T and NK cell cytotoxicity while suppressing Treg activity^64,133,134. Mechanistically, HDACis chelate zinc ions at HDAC catalytic sites, blocking deacetylation^135,136. SAHA enhances tumor antigen presentation, immune activation, and apoptosis in retinoblastoma models³⁷. Trichostatin A (TSA) activates the Rb‒E2F1 axis to promote immune responses and apoptosis³⁷. Transcriptomic data show HDAC is upregulate TRAIL, DR5, Bax, APAF1 and downregulate BCL2, XIAP, sensitizing tumors to apoptosis¹³⁷. HDACis also inhibit angiogenesis and metastasis by suppressing vascular endothelial growth factor (VEGF), basic fibroblast growth factor, hypoxia inducible factor 1α, and endothelial nitric oxide synthase 138, 139, 140. LBH589 enhances programmed cell death protein ligand 1 (PD-L1) expression and synergizes with PD-1 blockade in melanoma models³⁸.

3.2.3. Methylation

Histone methylation regulates chromatin structure, gene expression, and immune responses through methyltransferases (“writers”) and demethylases (“erasers”).

EZH2: H3K27 methylation and oncogenic activity. Enhancer of Zeste Homolog 2 (EZH2), the catalytic subunit of polycomb repressive complex 2 (PRC2), mediates H3K27 trimethylation (H3K27me3) to silence transcription^141,142. EZH2 also methylates non-histone proteins such as GATA4 and RORα, expanding its regulatory roles¹⁴³, it modulates oncogenic transcription through lncRNAs and mRNAs interactions¹⁴⁴. EZH2 overexpression drives tumor growth and metastasis in prostate, breast, lung, and colorectal cancers^141,145. In ovarian cancer, EZH2 suppresses type I chemokines to promote immune evasion¹⁴⁶. Additionally, EZH2 regulates CD8⁺ T cell memory differentiation and chemotherapy responsiveness^147,148. EZH2 inhibitors (EZH2is), such as CPI-1205, block H3K27me3 and enhance CD4⁺/CD8⁺ T cell responses in colon cancer³⁹. GSK126 and EPZ6438 inhibit EZH2’s SAM-binding site, reducing tumor growth and polarizing macrophages toward M1 via STAT3 demethylation⁴⁰. EPZ005687 covalently inhibits EZH2 in lymphoma with Y641/A677 mutations⁴¹. Other strategies disrupt PRC2 via SAH-EZH2 peptides or EZH2‒EED interaction inhibitors^149,150.

PRMT: Arginine methylation and immune escape. Protein arginine methyltransferases (PRMTs) catalyze mono- and di-methylation of arginine residues, regulating protein interactions, stability, and gene expression¹⁵¹. PRMTs are categorized into type I (PRMT1, 4, 6), type II (PRMT5, 9), and type III (PRMT7)^152,153. PRMT1 and PRMT5 interact with epstein–barr virus (EBV) antigens EBNA1/2 to promote viral latency and immune evasion^154,155. Aberrant PRMT1/5 expression drives oncogenesis and immune escape through transcriptional repression¹⁵⁶. PRMT5 also suppresses HBV transcription through chromatin and RNA packaging regulation^157,158.

PRMT inhibitors (PRMTis) like PT1001B enhance CD8⁺ T cell infiltration, reduce PD-L1 expression, and improve tumor control in PDAC models⁴². JNJ-64619178 binds SAM/substrate pockets to sustain PRMT5 inhibition and reduce proliferation across various cancers⁴³. PF-06939999 competitively inhibits PRMT5, altering NSCLC cell splicing and reducing SDMA⁴⁴. PRMT5 knockdown enhances antigen presentation by upregulating type-I interferon (IFN-I), chemokines, and MHC-I¹⁵⁹. PRMT5 inhibition blocks EBV-driven B cell transformation and selectively kills HTLV-1-infected T cells160, 161, 162.

3.2.4. Demethylation

Histone lysine demethylases (KDMs) remove methyl groups from histone lysines, altering chromatin and regulating gene expression, differentiation, and DNA damage responses^163,164. KDMs are classified into KDM1 (lysine-specific demethylase 1/2, LSD1/2), KDM2-7, and JmjC-domain proteins¹⁶⁵. LSD1 (KDM1A), the first discovered demethylase, functions independently of JmjC domain 166, 167, 168. JmjC KDMs demethylate di- and trimethyl lysines via Fe(II)/α-KG-dependent dioxygenase activity¹⁶⁹. Notable jumonji C domain-containing histone demethylase (JMJD) proteins include JMJD2B (KDM4B) and JMJD2D (KDM4D), involved in gene silencing, chromatin remodeling, and DNA repair^164,170. Aberrant KDM activity is linked to cancer, neurodegeneration, and immune dysfunction¹⁶⁹.

LSD1 removes H3K4 and H3K9 methylation, facilitating viral genome reactivation and chromatin accessibility^166,167. LSD1 inhibition blocks herpes simplex virus type 1 (HSV-1) and cytomegalovirus replication via Tranylcypromine, which targets the flavin adenine dinucleotide catalytic domain⁴⁸. OG-L002 inhibits H3K9 demethylation to prevent EHV-1 reactivation in equine cells⁴⁹. LSD1 also regulates RNA virus infection and host immunity; its inhibition exacerbates influenza A infection¹¹. LSD1 knockdown triggers dsRNA stress and IFN-I signaling, enhancing T cell infiltration in melanoma¹⁷¹.

3.3. Chromatin remodeling

Chromatin remodeling complexes use ATP hydrolysis to reposition nucleosomes and regulate chromatin accessibility^172,173. Among them, switch/sucrose non-fermentable (SWI/SNF) and Imitation SWI/SNF (ISWI) are the most relevant to cancer and immunity, and are highlighted here for their roles in tumor progression and therapeutic response¹⁷⁴.

3.3.1. SWI/SNF

The SWI/SNF complex, also known as the Brahma-related gene 1 (BRG1)/Brahma (BRM)-associated factor (BAF) complex, is a multi-subunit ATP-dependent chromatin remodeler regulating nucleosome positioning, DNA accessibility, and transcription. It consists of three subfamilies: canonical BAF (cBAF), polybromo-associated BAF (PBAF), and non-canonical BAF (ncBAF)^175,176. SWI/SNF localizes to H3K27ac-marked enhancers and cooperates with transcription factors to maintain open chromatin at active genes^177,178. ATPase subunits SMARCA2 (BRM) and SMARCA4 (BRG1) use ATP hydrolysis to slide or evict nucleosomes, regulating transcription¹⁷⁹. SWI/SNF mutations are common in ovarian, pancreatic, and rhabdoid tumors, often linked to impaired DNA repair and epigenetic dysregulation^180,181.

Targeting SWI/SNF dysfunction is an emerging strategy, via direct inhibition or synthetic lethality. BRM-014 selectively inhibits SMARCA2/4 ATPase, suppressing SMARCA2 in SMARCA4-deficient leukemia, impairing tumor survival⁵⁰. Degraders such as dBRD9, a proteolysis-targeting chimera directed against bromodomain-containing protein 9 (BRD9), disrupt ncBAF bromodomain function and suppress tumor growth in SMARCB1-deficient cancers, including rhabdomyosarcoma^182,183. In these tumors, BRD9 is a synthetic dependency; its depletion reduces chromatin accessibility and oncogenic transcription¹⁸⁴.

SWI/SNF-deficient tumors show vulnerabilities to epigenetic inhibitors. Tazemetostat, an EZH2 inhibitor, reduces proliferation in SMARCA4-deficient ovarian cancer by antagonizing PRC2 silencing¹⁸⁵. Similarly, the HDAC inhibitor Panobinostat suppresses tumor growth and induces multilineage differentiation in rhabdomyosarcoma with SMARCB1 (BAF47) loss¹⁸⁶. Loss of SWI/SNF subunits enhances immune infiltration and antigen presentation, suggesting increased immunotherapy sensitivity¹⁸⁷. These findings highlight the dual value of SWI/SNF inhibitors in chromatin remodeling and immune modulation.

3.3.2. ISWI

The ISWI family of ATP-dependent remodelers regulates nucleosome spacing and chromatin accessibility via ATP-driven histone sliding and repositioning¹⁸⁸. In humans, SMARCA5 (SNF2H) and SMARCA1 (SNF2L) form complexes with distinct regulatory subunits across cell types¹⁸⁹. ISWI remodelers regulate transcription, DNA repair, replication, and chromatin structure by maintaining nucleosome uniformity and gene expression^190,191.

ISWI dysregulation is implicated in breast, ovarian, melanoma, and hepatocellular cancers. SMARCA5 overexpression promotes β-catenin nuclear localization and proliferation in hepatocellular cancer¹⁹², whereas SMARCA1 suppresses melanoma growth via Wnt/β-catenin downregulation¹⁹³. These findings highlight ISWI’s dual, context-specific roles in cancer and its therapeutic potential.

ISWI-targeted therapy remains early-stage, but studies support its relevance in cancer. Bromodomain-containing ISWI subunits are a focus for small-molecule development. NVS-CECR2-1, a selective inhibitor of cat eye syndrome chromosome region, candidate 2 (CECR2), disrupts ISWI assembly and induces cytotoxicity and apoptosis in colon cancer via CECR2-dependent and -independent pathways⁵¹. ISWI subunit crystal structures enable structure-based inhibitor design targeting ATPase or reader domains^194,195.

ISWI inhibitors may enhance antitumor immunity. Preclinical studies show ISWI dysfunction enhances tumor immunogenicity and may synergize with checkpoint blockade¹⁹⁶. In triple-negative breast cancer, combining CECR2 inhibitor GSK2801 with BET inhibitor JQ1 enhances efficacy over monotherapy, suggesting dual epigenetic blockade may overcome resistance¹⁹⁷. Given ISWI’s role in chromatin, its inhibition offers a strategy to reprogram tumor epigenetics and sensitize cancers to existing therapies.

3.4. Non-coding RNA

3.4.1. Classification and mechanisms of epigenetic regulation

Non-coding RNAs (ncRNAs), including microRNAs (miRNAs), long non-coding RNAs (lncRNAs), and small interfering RNAs (siRNAs), modulate gene expression via epigenetic pathways. miRNAs and siRNAs repress mRNA translation or promote degradation via RNA interference198, 199, 200. LncRNAs regulate chromatin modifiers, transcription factors, or RNA molecules201, 202, 203. LncRNAs guide chromatin-modifying complexes to specific genomic loci. HOTAIR recruits PRC2 to catalyze H3K27me3 and transcriptional silencing²⁰¹, while Kcnq1 recruits DNMT1 for promoter hypermethylation²⁰⁴. Additional mechanisms include RNA–DNA triplex formation to recruit DNMT3 or PRC2^205,206, regulating transcription via RNA polymerase II²⁰⁷, and modulation of splicing and translation via RNA–RNA duplexes²⁰⁸. NcRNA dysregulation contributes to aberrant DNA methylation, histone modification, and transcriptional repression in cancers, neurodegeneration, and fibrosis²⁰⁹. Notably, miRNAs and lncRNAs also participate in the epigenetic and epitranscriptomic regulation of innate immune pathways by modulating pattern recognition receptors (PRRs) and their downstream effectors210, 211, 212. Through controlling the expression or RNA fate of key components in RIG-I/MDA5-and cGAS-STING-associated signaling, ncRNAs can influence IFN1 responses and antiviral immunity^213,214.

3.4.2. NcRNA-targeted inhibitors

Therapeutic strategies targeting ncRNAs aim to restore epigenetic balance by silencing oncogenic ncRNAs or restoring tumor-suppressive ncRNAs. Antisense oligonucleotides (ASOs), such as Gapmers, bind lncRNAs and trigger RNase H-mediated degradation. For example, ASOs against MALAT1 inhibit metastasis by reversing chromatin-mediated gene repression^215,216. Folic acid–PEG-modified ASOs enhance nuclease resistance and tumor targeting, significantly improving cellular uptake and antitumor efficacy⁵².

RNA interference via siRNAs selectively depletes oncogenic ncRNAs, while CRISPR/Cas9 enables genomic disruption of lncRNA loci (e.g., GMAN in gastric cancer)^217,218. miRNA mimics or inhibitors modulate epigenetic enzyme expression—miR-29 mimics reduce DNMT3A/B levels, demethylating tumor suppressor genes²¹⁹. Similarly, miRNA mimics can restore the function of downregulated tumor-suppressive miRNAs in cancer cells, thereby inhibiting tumor growth and metastasis⁵³. Small molecules targeting ncRNA interactions (e.g., eRNAs or enhancer–promoter looping regulators) represent emerging strategies to modify chromatin structure indirectly220, 221, 222, 223. These approaches may complement conventional epigenetic inhibitors in correcting transcriptional dysregulation.

3.5. Others

Beyond classical epigenetic mechanisms, additional processes contribute to the regulation of chromatin dynamics and gene expression. These include epigenetic readers, such as the bromodomain and extra-terminal domain (BET) protein family, and metabolism-epigenetics crosstalk mediated by mutant isocitrate dehydrogenase (IDH). Both mechanisms represent important targets in the context of epigenetic therapies for cancer.

3.5.1. BET proteins as epigenetic readers

BET proteins act as epigenetic readers²²⁴, comprising BRD2, BRD3, BRD4 and testis-restricted BRDT²²⁵. BET protein recognizes ε-N-acetylated lysine in histone tails²²⁴. BET proteins recruit repair factors and initiate cell cycle checkpoints by recruiting transcription factors or remodeling chromatin to regulate DNA repair genes²²⁶. BET is aberrantly activated in cancer, driving excessive transcription; thus, BET-targeted therapies are promising.

BET inhibitors (BETis) block BET chromatin binding, suppressing oncogene and immune evasion gene transcription²²⁷. BETis are classified as pan-BETis, BD1-or BD2-selective agents, BRD4-specific inhibitors, or BET degraders²²⁷. PLX51107, a non-selective BETi, blocks BRD histone binding, enhances T cell infiltration, and downregulates PD-L1 and IDO in the TME, promoting tumor regression in melanoma⁵⁴. However, BETi-induced MYC suppression may paradoxically increase PD-L1, highlighting a dual regulatory axis²²⁸. OTX015 (MK-8628), an oral BETi, binds the acetyl-lysine pocket of BRD2/3/4, displacing them from chromatin and suppressing RNA polymerase II signaling. This downregulates c-Myc, induces cell cycle arrest, and reduces proliferation in myeloma and mesothelioma^55,56. Despite promise, BETis may cause resistance via compensatory pathways and on-target toxicities like thrombocytopenia and immunosuppression²²⁶. Nonetheless, BETis suppress oncogenic transcription and reprogram immunity, supporting their use in cancer epigenetic therapy.

3.5.2. Metabolism-epigenetics crosstalk: IDH mutations

Emerging evidence highlights the critical role of metabolic-epigenetic crosstalk in cancer, with isocitrate dehydrogenase (IDH) mutations as a key example²²⁹. IDH enzymes (IDH1‒3) convert isocitrate to α-ketoglutarate (α-KG), an essential co-factor for α-KG-dependent dioxygenases such as TET DNA demethylases and JmjC-domain histone demethylases, thereby linking metabolism to epigenetic regulation^230,231. IDH1/2 contribute to redox homeostasis via NADPH, while IDH3 drives tricarboxylic acid cycle flux. Cancer-associated IDH1/2 mutations produce the oncometabolite D-2-hydroxyglutarate (D-2-HG) by reducing α-KG^232,233. D-2-HG competitively inhibits α-KG-dependent dioxygenases, leading to global DNA and histone hypermethylation and contributing to tumorigenesis in glioma and AML²³⁴.

Targeting mutant IDH restores α-KG dependent enzymatic activity, lowers D-2-HG, and reverses epigenetic abnormalities²³⁵. Approved agents include ivosidenib (IDH1) and enasidenib (IDH2), which promote differentiation and suppress tumor growth in IDH-mutant AML^57,234. In glioma, IDH1 inhibitors like AGI-5198 reduce H3K9me3 and promote glial differentiation despite persistent DNA hypermethylation²³⁶. Combination strategies, such as IDH inhibitors with DNMT inhibitors like 5-azacytidine, further enhance therapeutic effects by reactivating lineage markers and sensitizing tumors to chemotherapy (e.g., temozolomide) in IDH1 R132H mutant glioma²³³. Additionally, IDH inhibition depletes NADPH, increasing oxidative stress and contributing to anti-tumor effects²³⁷.

3.6. Toxicities and resistance mechanisms of epigenetic inhibitors

Although epigenetic inhibitors exhibit considerable therapeutic potential, their clinical application is frequently constrained by dose-dependent toxicity and the development of acquired resistance²³⁸.

Toxicity profiles mainly arise from non-specific disruption of normal epigenetic regulation. DNMT inhibitors, frequently cause hematologic toxicities (thrombocytopenia, neutropenia, anemia, bone marrow suppression), gastrointestinal symptoms (nausea, vomiting, diarrhea), and fatigue239, 240, 241, 242. HDAC inhibitors are linked to thrombocytopenia, fatigue, gastrointestinal issues, and cardiac effects such as QT prolongation243, 244, 245, 246. BET and EZH2 inhibitors share similar hematologic and gastrointestinal toxicities, with occasional neurocognitive side effects reported in prolonged use^247,248. Fortunately, most adverse events are reversible and can be managed by dose adjustment, supportive care, or optimized scheduling²⁴⁹.

Resistance mechanisms stem largely from the plasticity of the cancer epigenome^250,251. Intrinsic resistance often results from pre-existing tumor heterogeneity and baseline epigenetic states that evade target inhibition²⁵². Acquired resistance commonly involves adaptive transcriptional reprogramming, metabolic rewiring, and compensatory upregulation of alternative epigenetic enzymes²⁵³. Well-known examples include Wnt/β-catenin pathway activation that bypasses BET protein dependency²⁵⁴, as well as clonal selection and microenvironmental protection that drive relapse^255,256.

To overcome these limitations, rational combination strategies are both conceptually and mechanistically justified¹⁰⁴. In particular, integrating epigenetic inhibitors with oncolytic viruses exploits complementary vulnerabilities by enhancing viral replication and immune activation while potentially enabling dose de-escalation to reduce cumulative toxicity^92,257,258.

4. Epigenetic inhibitors and OVs as combination therapy

As an emerging cancer treatment modality, oncolytic viruses (OVs) have demonstrated significant clinical translation in recent years⁴. Talimogene laherparepvec (T-VEC) is indicated for melanoma and exhibits oncolytic activity in multiple tumor types, including breast and prostate cancers²⁵⁹.

Notably, OV-based combination therapies represent a promising strategic advancement in oncology²⁶⁰. OVs synergize with chemotherapy, radiotherapy, targeted therapy, and immunotherapy owing to their unique mechanisms261, 262, 263. In particular, they enhance immune checkpoint inhibitors (ICIs) efficacy by increasing tumor immunogenicity and immune infiltration, thereby overcoming resistance in immunologically “cold” tumors²⁶⁴, as demonstrated by improved responses to T-VEC plus ipilimumab in melanoma²⁶⁵. Furthermore, OVs potentiate chimeric antigen receptor T-cell therapy by augmenting T-cell recruitment to tumor sites²⁶⁶. These properties have further stimulated interest in combining OVs with epigenetic inhibitors to modulate antiviral signaling and tumor immunity, and the following sections will introduce the synergistic application of epigenetic inhibitors and oncolytic virotherapy in detail (Table 2).

Table 2.

Summary of epigenetic modulators combined with OVs.

Target	Epigenetic modulator	Oncolytic virus	Effect	Tumor type	Ref.
HDACis	TSA	γ1 34.5-deleted- oHSV-1	Cell cycle protein D1↓	Colon cancer, lung cancer, brain cancer	59
		γ1 34.5-deleted- oHSV-1	NF-κB activation↑, viral replication↑	Oral squamous cell carcinoma	58
		BoHV-1	Viral replication↑, tumor lysis↑	Lung adenocarcinoma	60
		VV	Virus replication↑	Melanoma, bowel cancer	61
		Ad5-TRAIL	CAR expression↑, tumor lysis↑, caspase-2 activity↑	Cervical cancer	62
	VPA	rQNestin34.5	IFN-responsive antiviral genes↓, viral gene replication and transcription↑, viral transmission↑	Glioma	63
		HSV	Viral replication↑, GM-CSF production↑, TAA-specific CTL production ↑, NKG2D ligand expression on cancer cells↑	Melanoma	64
		H1PV	Apoptosis↑	Pancreatic cancer, cervical cancer	65
	Scriptaid	P/V-CPI- mutant	Tumor lysis↑, apoptosis↑	Small cell lung cancer, laryngeal cancer cells	66
	Entinostat	RV	Viral replication↑, apoptosis and inflammatory response↑, tumor lysis↑	Lymphoma, head and neck squamous cell carcinoma, multiple myeloma	67
	Res	MeV	Cytotoxicity↑, tumor lysis↑	Hepatocellular carcinoma, pancreatic cancer	68
	SB	Ad5-TRAIL	CAR expression↑, tumor lysis↑, caspase-2 activity↑	Cervical cancer	62
	Romidepsin, MS-275	AdTRAIL	CAR expression↑, caspase-2 activity↑	Prostate cancer	62
	Vorinostat	AdTRAIL	CAR expression↑, TRAIL transcription↑, tumor lysis↑, NF-κB activation↓, decomposition of Bcl-2↑	Lung cancer	69
		VSVΔ51	Viral replication ↑, tumor lysis↑, antiviral response↓	Prostate cancer	70
	SAHA	VSVΔ51	Viral replication ↑, tumor lysis↑, antiviral response↓	Prostate cancer	70
	Butyrate	Telomerase-specific OAd	Apoptosis and inflammatory response↑, tumor lysis↑	Colon carcinoma	71
	AR42	Telomerase-specific OAd	Tumor lysis↑, Akt signaling↓	HCC	72
DNMTis	5-AZA	BHV-1	Viral replication ↑, tumor lysis↑	Lymphoma	73
		HSV rQNestin34.5	Viral replication↑, tumor lysis↑	Glioma	74
		P/V	Tumor cell killing↑, caspase activation↑	Neuroblastoma	75
Chromatin remodeling	SWI/SNF	VVDD	Cell cycle protein D1↓	AT/RT	76
miRNAi	P19	Ad1, -2, -5, -6	Tumor cell killing↑	Lung cancer	77
	miR-222 inhibitor	AdNuPAR-E-miR222-S	Viral replication↑, cytotoxicity↑	Pancreatic cancer	78
miRNA	miR-99b, miR-485	ICOVIR15	Viral replication↑, tumor lysis↑	PDAC	79
	amiR-4	VSVd51-amiR-4	Tumor cell killing↑	PDAC Melanoma	80
	miR-26b	OAd	Viral transmission↑	Prostate cancer	81
	miR-34a	AdCN205	Viral cytotoxicity↑, sensitized cancer cells to IL-24↑, angiogenesis↓	HCC	82
lncRNAi	lncRNAi	AdSVPE1a-lncR	Tumor cell proliferation and migrations↓, tumor lysis↑	HCC	83
	lncRNAi9	AdSVP-lncRNAi9	Tumor cell proliferation and migrations↓	TNBC	84
lncRNA	lncRNA 8244	SVA	IFN-β and TLR3 expression↑	NA	85
siRNA	ERCC1-siRNA	Ad-siERCC1	Tumor cell killing↑, tumor cell proliferation and migrations↓	Ovarian cancer	86
	siGNAQ	H101	Tumor cell proliferation↓, tumor lysis↑	UM	87
	siPDL1	siPDL1/HVJ-E	PD-L1 expression↓, tumor cell proliferation↓, Tregs↓	GBM	88
	Ki67-ZXC2 siRNA, Ki67-ZXC2-Ki67 siRNA, Ki67-ZXC2-hTERT siRNA, Ki67-ZXC2-double siRNA	OAd	Viral replication↑, tumor lysis↑	Renal cancer	89
	1504-siRNA	Ad-TERTp-E1A-1504	Tumor cell killing↑, tumor lysis↑	Prostate cancer, gastric carcinoma, hepatic carcinoma	90
	Eg5 siRNA	HVJ-E	Tumor lysis↑	Glioblastoma	91
BETi	OTX-015, iBET-762, Neo-2734	AdΔΔ、Ad-3Δ-A20T	c-Myc and Myb expression↓, tumor cell killing↑, viral replication ↑	Pancreatic ductal adenocarcinoma	92

4.1. DNMTis and OVs as combinatorial therapy

Combining oncolytic viruses with epigenetic inhibitors represents a growing therapeutic paradigm. Rational combination regimens that integrate signaling pathway modulators, immunomodulators, and epigenetic agents can significantly enhance antitumor immunity and therapeutic efficacy^104,267. Specifically, DNMT inhibitors (DNMTi) enhance OV efficacy by promoting viral replication, reactivating endogenous retroviruses (ERVs), and increasing tumor immunogenicity. These effects collectively enhance tumor cell lysis and antitumor immunity (Fig. 3).

Figure 3.

DNMT inhibitors enhance oncolytic virus efficacy through epigenetic regulation. (A) DNMTis facilitate viral replication by demethylating the nestin promoter/enhancer. (B) DNMTis lift ERV silencing, activating dsRNA via LTR demethylation to trigger cytokine release, promoting tumor cell apoptosis and lysis. (C) DNMTis mediate immune cell reprogramming by upregulating MHC-I to enhance CD8⁺ T cell cytotoxicity, suppressing Treg activity, and activating NKG2D-dependent NK cell killing.

4.1.1. Herpes simplex viruses

Oncolytic herpes simplex viruses (oHSVs), engineered from their wild-type counterparts, epitomize a rationally designed viral platform that achieves both selective replication and precise tumor tropism⁴. For instance, T-VEC is an engineered oHSV-1 designed to preferentially replicate in tumor cells and induce robust anti-tumor immunity. Currently, T-VEC is primarily utilized in treating patients with recurrent melanoma post-surgery²⁶⁸.

Accumulating evidence indicates that DNMTis substantially augment herpes simplex viruses (HSV) replication across both in vitro and in vivo models, thereby reinforcing their translational relevance. Specifically, the DNMTi 5-azacytidine (5-aza) significantly enhances the replication and cytotoxicity of bovine herpesvirus 1 (BHV-1), indicating potential synergy between DNMTis and herpesvirus-mediated oncolysis⁷³. Moreover, DNMTis can demethylate the nestin promoter/enhancer, which the engineered HSV variant rQNestin34.5 exploits to enhance selective viral replication. This enhancement is evidenced by increased viral gene expression and a higher number and larger size of the green fluorescent protein (GFP)-positive infected glioma cells⁷⁴.

Building on these mechanistic insights, recent investigations have advanced toward combinatorial strategies integrating DNMTis with oHSVs, unveiling synergistic antitumor effects¹⁰. For instance, dual-reactive DNMT inhibitor nanoprecursors (ACNPs), when combined with oHSV, were shown to significantly enhance Gasdermin E-mediated pyroptosis. The combination therapy using ACNPs and oHSV effectively suppressed tumor growth, remodeled the immunosuppressive tumor microenvironment (TME), and improved responses to immune checkpoint blockade (ICB) therapy²⁵⁷. These findings highlight DNMTis’ capability to potentiate oHSV infection efficacy, thereby improving therapeutic outcomes in cancer immunotherapy.

4.1.2. Parainfluenza virus

Oncolytic parainfluenza virus (OPV) is a modified member of the parainfluenza virus family, characterized by an enhanced ability to selectively infect and kill cancer cells²⁶⁹. This modification of the wild-type parainfluenza virus (P/V) significantly increases its specificity and cytotoxic effects toward malignant cells²⁶⁹. For instance, a genetically altered variant, parainfluenza virus 5 (PIV5), demonstrates substantial cytopathic effects and robust induction of antiviral cytokines^270,271.

By reversing epigenetic silencing, DNMTis upregulate apoptosis-associated proteins, thereby rendering malignant cells more susceptible to parainfluenza virus-mediated cytolysis and improving therapeutic efficacy. Cells infected by P/V activate apoptotic caspases including caspase-8, -9, and -3/7. However, the antitumor efficacy of P/V infection alone remains limited, as 5-azacytidine treatment and P/V infection individually induce caspase-dependent and caspase-independent cell death pathways, respectively. Kritika Kedarinath et al.⁷⁵ demonstrated that combining these two treatments significantly improves antitumor effects, primarily through the caspase-dependent pathway.

Beyond apoptosis, the combinatorial regimen amplifies innate immune signaling, as increased double-stranded RNA (dsRNA) accumulation provokes a potent type-I interferon (IFN-I) response and robust induction of interferon-stimulated genes (ISGs), the notably oligoadenylate synthase (OAS) family members. Activated OAS proteins further stimulate ribonuclease L, an enzyme that cleaves cellular and viral RNAs, thereby disrupting viral replication and promoting apoptosis^272,273. Gainey et al.²⁷⁴ also reported increased levels of dsRNA produced during P/V viral replication. Consequently, Kedarinath et al.⁷⁵ suggest that the enhanced virus-mediated death observed in neuroblastoma cells pretreated with 5-azacytidine arises from the accumulated dsRNA and downstream signaling pathways triggered by combined drug treatment and viral infection.

Furthermore, DNMTi-P/V co-treatment disrupts viral protein synthesis, culminating in markedly attenuated viral gene expression and replication. For example, 5-azacytidine treatment alone upregulates the expression of crucial antiviral genes, such as IFN-β and OAS2. This antiviral effect is further amplified upon subsequent P/V infection, effectively suppressing viral gene expression and proliferation⁷⁵.

4.1.3. Vesicular stomatitis virus

Owing to the inherent deficiencies in IFN signaling within malignant cells, vesicular stomatitis virus (VSV) intrinsically demonstrates tumor selectivity, enabling preferential oncolysis while preserving normal tissues. VSV selectively infects and kills cancer cells by activating apoptotic pathways^70,275. Furthermore, studies have demonstrated that deleting the 51st amino acid methionine on the VSV-M protein (VSVΔM51) or substituting it with arginine (rM51R-M) can alleviate the inhibitory effect of the M protein on IFN expression in infected cells^276,277. These modifications enhance viral safety in normal tissues^276,277. Additionally, treatment of EL-4 cells from mice with acute T-cell lymphoblastic leukemia using VSVΔM51 followed by DNMTis has shown significant therapeutic efficacy¹⁰.

4.1.4. Adenovirus

Recent studies have revealed that the therapeutic synergy between DNMTis and oncolytic adenoviruses (OAd) extends beyond pharmacological co-administration, as epigenetic modulation can also be directly engineered into viral platforms. An innovative oncolytic adenovirus encoding DNMT1 targeting short hairpin RNA (OAd.shDNMT1) has been developed, enabling virus-mediated epigenetic silencing within the TME²⁷⁸. Preclinical studies in murine 4T1 and CT26 tumor models demonstrated that OAd.shDNMT1 infection generated shDNMT1-loaded small extracellular vesicles capable of reprogramming myeloid-derived suppressor cells (MDSCs), reducing their suppressive function, and restoring antitumor immunity. This strategy markedly suppressed tumor growth, prolonged survival, and synergized with immune checkpoint blockade, leading to complete tumor eradication in moderately immunogenic models. These findings indicate that integrating DNMT inhibition directly into oncolytic adenoviruses represents a novel paradigm in epigenetic-virotherapy, offering a targeted approach to remodel the immunosuppressive TME and enhance the efficacy of cancer immunotherapy.

4.2. HDACis and OVs as combinatorial therapy

Histone modifications regulate viral replication, tumor cell death, and antitumor immunity^267,279. HDAC inhibitors (HDACis) enhance OV efficacy by targeting these processes through multiple mechanisms (Fig. 4).

Figure 4.

Roles of histone modifications in regulating OVs therapy efficacy. (A) HDACis enhance viral replication by attenuating IFN responses through IFNR signaling inhibition. (B) HDACis promote virus-induced tumor cell death via NF-κB activation, facilitating autophagy and apoptosis. (C) HDACis facilitate immune cell remodeling by reducing regulatory T cell (Treg)-mediated suppression, enhancing CD8⁺ T cell cytotoxicity via major histocompatibility complex class 1 (MHC-I) upregulation and PD-1/PD-L1 blockade, and boosting NK cell-mediated killing through granzyme and perforin release. (D) HDACis augment viral attachment and entry through upregulation of viral receptors.

4.2.1. Herpes simplex viruses

HDACis markedly potentiate HSV replication and transcriptional activity within malignant cells, thereby amplifying their oncolytic potential. For instance, in glioma cells, valproic acid promotes gene replication and transcription of an infected cell protein 6 (ICP6)-deleted HSV mutant expressing ICP34.5⁶³. Transcriptomic analyses by Kawamura et al.²⁸⁰ identified selective alterations in HSV mRNA processing and splicing induced by HDACis, favoring viral mRNA expression. Submicromolar concentrations of pan-HDACis, trichostatin A (TSA), and panobinostat significantly increased the infectivity and spread of the oncolytic HSV variant G47Δ in multiple myeloma cells in vitro, enhancing viral-mediated cell killing even at low multiplicity-of-infection (MOI). Cody et al.⁵⁹ demonstrated that TSA activates nuclear factor κB (NF-κB), enhancing the replication of γ(1)34.5-deficient oHSV in oral squamous cell carcinoma (OSCC) models. Similarly, HDACis improved HSV replication in breast cancer cells⁵⁹.

Beyond direct viral replication, HDACis augment HSV-mediated oncolysis by upregulating granulocyte-macrophage colony stimulating factor (GM-CSF), a pleiotropic cytokine that orchestrates myeloid proliferation, inflammation, and cytotoxic effector functions. Specifically, HDACis boosted HSV replication, cell killing, and GM-CSF secretion in melanoma cell models⁶⁴. Moreover, the combinatorial application of HDACis with HSV robustly amplifies antitumor immunity, reshaping cytokine secretion profiles and augmenting immune effector recruitment. Jennings et al.⁶⁴ showed that melanoma cells pretreated with valproic acid before HSV infection exhibited reduced IL-10 and increased IFNγ secretion in co-cultures with immature dendritic cells (iDCs), indicating enhanced production of tumor associated antigen (TAA)-specific cytotoxic T lymphocytes (CTLs). Additionally, TSA and HSV synergistically suppress angiogenesis and tumor growth, linked to increased degradation of cell cycle protein D1 and vascular endothelial growth factor (VEGF) inhibition⁵⁸. TSA also promotes leukocyte infiltration into infected tissues by enhancing viral activation of glycosyltransferase genes (FUT3, FUT5, FUT6), leading to increased Lewis X ligand expression, critical for leukocyte‒endothelial interactions^281,282. Lastly, HDACis upregulate natural killer group 2, member D (NKG2D) ligand expression in melanoma cells, enhancing NK cell-mediated cytotoxicity by activated peripheral blood mononuclear cells (PBMC)⁶⁴.

4.2.2. Parvoviridae

HDACis effectively attenuate IFN induction and secretion, thereby dismantling intrinsic antiviral defenses that otherwise constrain viral replication. IFNs exert antiviral effects by stimulating the expression of numerous ISGs, which inhibit viral replication, transcription, and other critical cellular processes²⁸³. Specifically, HDACis directly diminish the synthesis of exogenous human IFN-β, subsequently lowering ISG expression and enhancing P/V-induced viral spread and cancer cell death⁶⁶. Following viral infection, the induction of IFN-β requires phosphorylation and nuclear translocation of interferon regulatory factor-3 (IRF-3), processes that are disrupted by HDACis, thereby inhibiting IFN-β gene transcription⁶⁶.

In addition, HDACis synergistically heighten the oncolytic potency of parvoviruses, driving robust caspase-mediated apoptosis and amplifying tumor cytotoxicity. Caspases, a family of cysteine-dependent aspartate-specific proteases, are crucial mediators of programmed cell death, including apoptosis and necroptosis, as well as inflammatory responses⁶⁶. HDACis amplify cell death and caspase-dependent apoptotic signaling in cells infected with P/V-CPI mutants, partly through pathways involving the recognition of dsRNA⁶⁶. Furthermore, HDACis significantly enhance the activation of caspase-3/7, -8, and -9, thereby increasing overall cytotoxicity following P/V-CPI infection⁶⁶. Bretscher et al. demonstrated that combined treatment with HDACis (VPA) and H1 parvovirus (H1PV) induced oxidative stress and apoptosis in cancer cells⁶⁵.

4.2.3. Bovine herpesvirus type 1

HDACis markedly facilitates the replication dynamics of bovine herpesvirus type 1 (BoHV-1), thereby intensifying its lytic activity in tumor contexts. In an A549 tumor xenograft mouse model, treatment with the HDAC inhibitor TSA significantly enhanced BoHV-1 replication, increased productive viral infection, and mitigated cytopathic effects (CPE) associated with viral infection⁶⁰.

Beyond replication, HDACis significantly magnify BoHV-1-driven tumor lysis, as evidenced by pronounced DNA damage and apoptotic signaling in malignant cells. Qiu et al.⁶⁰ employed the comet assay—a sensitive method for evaluating DNA strand breaks at the single-cell level. They quantified DNA damage using the tail DNA percentage (tailDNA%), defined as the ratio of DNA fluorescence in the comet tail to the total cellular DNA⁶⁰. Their findings revealed that BoHV-1 infection alone resulted in a tailDNA% of 34.28%, whereas the addition of TSA increased this value to 64.14%, confirming that HDACis (specifically TSA) significantly augments BoHV-1-induced DNA damage⁶⁰.

Interestingly, BoHV-1 infection reciprocally modulates host HDAC expression, revealing a bidirectional interplay between viral infection and epigenetic regulation. During infection of the A549 adenocarcinoma cell line, the protein levels of multiple HDAC isoforms progressively declined at different post-infection time points, indicating that BoHV-1 infection downregulates histone deacetylase expression⁶⁰.

4.2.4. Vesicular stomatitis virus

Silent information regulator sirtuin 1 (SIRT1), tightly regulated by the oncogenic microRNA miR-34a, is a pivotal determinant of VSV susceptibility in tumor cells²⁸⁴. Inhibition of SIRT1 suppresses cancer cell proliferation and induces cell cycle arrest or apoptosis²⁸⁵. In many cancers, genetic or epigenetic inactivation of the tumor suppressor gene P53 leads to the upregulation of SIRT1 transcription²⁸⁶. Notably, reduced SIRT1 expression has been directly linked to enhanced VSVΔM51-mediated oncolysis following HDACis treatment, suggesting that SIRT1 and/or miR-34a levels may serve as predictive biomarkers for the efficacy of oncolytic viral therapies.

HDACis further heighten the permissiveness of prostate cancer PC-3 cells to VSVΔM51 infection, accelerating viral replication and promoting extensive cytolysis. This effect is mediated through HDACis-induced upregulation of miR-34a, which downregulates SIRT1 and the anti-apoptotic protein Bcl-2. The resulting inhibition of SIRT1 activates NF-κB signaling, which further promotes VSVΔM51 infection and cancer cell death⁷⁰. NF-κB signaling also negatively regulates SIRT1 and binds to the miR-34a promoter, inducing miR-34a expression in a positive feedback loop^287,288.

Additionally, HDACis subvert innate antiviral defenses, rendering even drug-resistant cancer cells highly vulnerable to VSVΔM51 replication and spread. For instance, the HDACis SAHA modulates NF-κB target gene expression and inhibits IFN production, enhancing virus-induced apoptosis, autophagy, and cell lysis²⁵⁸.

Notably, HDACis exacerbate VSV-associated lymphopenia, profoundly impairing immune cell recovery and thereby augmenting viral oncolytic efficacy. VSV infection is known to cause rapid and severe depletion of lymphocytes following intravenous administration²⁸⁹. Bridle et al.²⁹⁰ observed that combined treatment with MS-275 and VSV led to near-complete elimination of pre-B and immature B cells in the bone marrow as early as Day 3 post-infection. This combination resulted in prolonged lymphopenia and delayed reconstitution of peripheral T and B cells, possibly due to heightened IFN responses²⁹⁰. These findings support a synergistic effect between HDACis and VSV in impairing immune cell recovery and enhancing oncolytic efficacy.

4.2.5. Reolysin

Reolysin is a naturally occurring, non-pathogenic, dsRNA virus with inherent oncolytic activity. It is currently under investigation in phase I–III clinical trials across a range of tumor types. Reolysin belongs to the type three dearing strain of the mammalian orthoreovirus family.

HDACis markedly augment Reolysin-mediated tumor infection and cytotoxicity in vivo by upregulating viral entry receptors and sensitizing malignant cells. In both human (SCC74A) and murine (mouse tonsil epithelial, MTE) squamous cell carcinoma models, HDACis treatment led to the upregulation of junctional adhesion molecule 1 (JAM-1), the primary cellular receptor for reovirus entry. This increase in JAM-1 expression significantly enhanced the susceptibility of head and neck cancer cells to Reolysin infection⁶⁷. Additionally, Islam et al. demonstrated that HDACis targeting of the signal transducer and activator of transcription 1 (STAT1) pathway reduced the expression of IRF1 and STAT1, thereby augmenting the anti-lymphoma effects of oncolytic reovirus²⁹¹.

In addition, HDACis significantly potentiate Reolysin replication, driving elevated capsid protein synthesis and viral propagation within tumor cells. Ramirez et al.⁶⁷ observed that treatment with AR-42 or SAHA significantly increased expression of reovirus capsid proteins (σ-NS), as indicated by enhanced green fluorescence in SCC74A cells expressing red fluorescent protein. This finding confirms that HDAC inhibition boosts reovirus replication by promoting capsid protein production following viral infection.

Furthermore, combinatorial treatment with HDACis and Reolysin exerts profound synergistic cytotoxicity while simultaneously reshaping the TME toward a pro-inflammatory, immunostimulatory state. Monocyte chemotactic protein-1 (MCP-1), also known as chemokine (C‒C motif) ligand 2 (CCL2), plays a critical role in orchestrating inflammatory responses by recruiting immune cells and enhancing the expression of other inflammatory mediators⁶⁷. In both SCC74A and MTE models, combined treatment with HDACis and Reolysin led to upregulated expression of IFN-α and MCP-1, suggesting that this combination induces a pro-inflammatory tumor response that may contribute to enhanced anti-tumor immunity⁶⁷.

4.2.6. Measles virus

HDACis substantially amplify the cytotoxic and tumor-lytic efficacy of measles virus (MeV), underscoring their value as sensitizers in oncolytic virotherapy. In a study by Ruf et al.⁶⁸, tumor cell viability was assessed based on metabolic activity. The results showed that combination therapy with resminostat and MeV reduced HepG2 cell viability to 49.2%, compared to 62.5% with MeV-SCD alone and 80.4% with resminostat alone, indicating that the combined treatment significantly increased cytotoxicity and apoptosis induction⁶⁸. Similarly, in pancreatic cancer models, resminostat treatment resulted in a notable reduction in tumor mass across all tested pancreatic cancer cell lines²⁹².

HDACis also enhance the early infection dynamics of MeV in hepatocellular carcinoma cells, facilitating superior viral entry and replication. Resminostat treatment increased the proportion of MeV-GFP-infected HepG2 cells from 13.2 to 21.9%, and Hep3B cells from 32.5 to 45.0% at 24 h post-infection⁶⁸. It suggested improved viral entry or early replication efficiency in the presence of HDACis⁶⁸.

Beyond direct viral facilitation, HDACis remodel innate immune signaling pathways, thereby mitigating antiviral resistance and enabling more efficient MeV-driven oncolysis. MeV infection normally induces IFN signaling through phosphorylation of STAT1 and upregulation of IFN-stimulated genes such as IFIT1. However, Ruf et al.⁶⁸ demonstrated that the addition of resminostat (5 μmol/L) to hepatocellular carcinoma cells pretreated with human IFN-β (1000 U/mL) significantly suppressed IFIT1 expression in HepG2, Hep3B, and PLC/PRF/5 cells⁶⁸. This finding indicates that resminostat can attenuate the innate antiviral response, thereby helping to overcome resistance to MeV-mediated oncolysis.

4.2.7. Adenovirus

HDACis heighten tumor cell susceptibility to adenoviral entry by upregulating key viral receptors, thereby enhancing the efficiency of initial infection events. Studies have shown that HDACis treatment upregulates the expression of the coxsackievirus and adenovirus receptor (CAR), which facilitates viral entry. In C4-2b prostate cancer cells, HDACis treatment significantly elevated CAR levels, enhancing adenoviral uptake and resulting in increased infection efficiency during initial viral delivery²⁹³. Similarly, in lung cancer models, vorinostat treatment upregulated CAR expression and led to increased luciferase expression in Ad-luc-transduced cells, as well as enhanced expression of the pro-apoptotic ligand TRAIL in Ad-TRAIL-transduced cells, demonstrating synergistic anti-tumor effects⁶⁹. Recent studies have demonstrated a significant synergistic effect between the P53-expressing telomerase-specific OAd OBP-702 and butyrate, a gut microbiota-derived metabolite that functions as an HDACi, in colorectal cancer models⁷¹. Butyrate directly enhances OBP-702 infectivity by upregulating CAR and integrins. In addition, it activates the cGAS-STING pathway to increase MHC-I expression, and also stimulates chemokine (C-X-C motif) ligand 10 secretion to recruit CD8⁺ T cells. These findings highlight the therapeutic potential of integrating gut microbial metabolites with oncolytic virotherapy as a novel immunotherapeutic strategy for colorectal cancer.

In addition, HDACis potentiate the tumor-lytic activity of recombinant adenoviral vectors, synergizing with pro-apoptotic payloads such as TRAIL to overcome resistance in malignant cells. In prostate cancer models, sodium butyrate and TSA, two HDAC inhibitors, enhanced CAR expression and adenoviral infectivity, thereby increasing TRAIL-mediated cytotoxicity⁶². Notably, in TRAIL-resistant DU-145 cells, HDACis treatment also suppressed protein kinase CK2 activity, which in turn activated caspase-2, suggesting an additional mechanism by which HDACis sensitize resistant prostate tumor cells to Ad5-TRAIL-induced apoptosis. More recently, the combination of HDACis with telomerase-specific OAd has been explored in hepatocellular carcinoma (HCC)⁷². Preclinical trial results indicate that the HDACi AR42 attenuated Telomelysin-induced phospho-Akt activation and enhanced Telomelysin-induced apoptosis, resulting in synergistic antitumor efficacy in HCC models. Similarly, the HDACi panobinostat was shown to markedly enhance the anti-myeloma efficacy of the SLAMF7-targeted oncolytic adenovirus Ad (CE1A)²⁹⁴.

4.2.8. Vaccinia virus

HDACis enhance the replication and dissemination of fusogenic oncolytic vaccinia virus (FUVAC) in tumor cells. In mouse melanoma B16-F10 cells, TSA modulated the expression of constitutive cellular genes and influenced cell structure, motility, and tight junction integrity⁶¹. These changes promoted intercellular fusion, thereby facilitating FUVAC replication and spread within tumor tissues. Similarly, in CT26 colorectal cancer cells, TSA activated viral replication through an IFN-independent mechanism⁶¹.

Moreover, TSA promoted cowpox virus dissemination even in the presence of IFN signaling in tumor cells. However, TSA was unable to overcome IFN-induced antiviral defenses in normal cells, highlighting the selective enhancement of viral replication in malignant tissues while preserving antiviral protection in non-cancerous cells²⁹⁵.

4.3. Chromatin remodeling and OVs as combinatorial therapy

Chromatin remodeling dysfunction is a key driver in certain tumors, such as atypical teratoid/rhabdoid tumor (AT/RT), which is characterized by loss of hSNF5, a subunit of the SWI/SNF complex. A modified double-deleted vaccinia virus (VVDD) engineered to express hSNF5 demonstrated enhanced antitumor activity by restoring cell cycle regulation, reducing S-phase entry, and downregulating cyclin D1. This strategy highlights the potential of combining oncolytic viruses with chromatin remodeling gene restoration to target epigenetically driven cancers⁷⁶.

Emerging evidence underscores the pivotal role of chromatin remodeling in orchestrating host-virus interactions, with profound implications for immune evasion and viral pathogenesis. OVs, like vaccinia virus, have been shown to actively modulate host chromatin architecture to suppress antiviral gene expression. For example, vaccinia virus infection induces long-range intra-chromosomal interactions while reducing overall chromatin accessibility, leading to the downregulation of immune-related genes and facilitating viral immune evasion²⁹⁶. Similarly, studies on SARS-CoV-2 have revealed its capacity to disrupt chromatin organization, resulting in the loss of structural specificity at loci governing antiviral responses, thereby weakening host immune surveillance mechanisms. These findings suggest that OVs may leverage chromatin remodeling not only to promote viral replication but also to evade innate immune defenses, providing a novel angle for enhancing virotherapy efficacy through epigenetic targeting²⁹⁷. These insights open new avenues for designing OVs that strategically modulate chromatin architecture to suppress tumor immune resistance and enhance therapeutic efficacy.

In a recent study, researchers identified BRD9²⁹⁸, a subunit of the ncBAF chromatin remodeling complex, as a key resistance factor to oHSV1 in glioblastoma (GBM). CRISPR-Cas9 screening revealed that BRD9 loss enhanced viral replication and antitumor activity by suppressing RELA-mediated antiviral gene expression. Clinically, low BRD9 expression correlated with significantly improved OV treatment outcomes²⁹⁸. The BRD9 inhibitor IBRD9 combined with oHSV1 achieved complete tumor regression in multiple models, highlighting a promising virus-epigenetic inhibitor strategy for overcoming OV resistance.

4.4. ncRNA and OVs as combinatorial therapy

In recent years, the combination of ncRNAs and OVs has emerged as a promising strategy in cancer therapy, offering enhanced antitumor efficacy through synergistic mechanisms. Small RNAs such as miRNAs⁷⁷, lncRNAs⁸⁵, and siRNAs⁸⁶ can modulate tumor-related gene expression, thereby improving viral replication, promoting tumor cell lysis, reshaping the TME, and enhancing immune responses. The following sections provide an overview of recent advances in combining OVs with miRNAs, lncRNAs, and siRNAs, highlighting their therapeutic mechanisms and potential applications across various solid tumors.

4.4.1. miRNA and OVs as combinatorial therapy

The strategic incorporation of miRNA inhibitors into OV platforms markedly amplifies tumor lysis, establishing a potent combinatorial approach. Doerner et al.⁷⁷ developed novel OAds derived from species C Ad1, Ad2, Ad5, and Ad6, incorporating a 24-bp deletion in the E1A gene for tumor-selective replication and expressing the RNAi suppressor P19. OAds based on Ad1, Ad2, and Ad6 exhibited stronger cell lysis activity than Ad5, indicating enhanced oncolytic potential. Their improved antitumor efficacy was linked to P19 expression in a cell type-dependent manner. In vivo, these P19-expressing OAds significantly delayed tumor growth compared to H101. Furthermore, an oncolytic adenovirus engineered to express a miR-222 sponge (AdNuPAR-E-miR222-S) demonstrated significant tumor growth reduction in an in vivo model of pancreatic ductal adenocarcinoma (PDAC). This construct effectively suppressed miR-222 expression while upregulating its downstream target genes, thereby enhancing both adenoviral replication and cytotoxicity in preclinical PDAC models. Notably, the elevated expression of miR-222 in PDAC and its inhibitory effect on adenoviral propagation underscore the therapeutic potential of the miR-222 sponge⁷⁸.

Engineering OVs to express miRNAs or artificial miRNAs (amiRs) significantly enhances their oncolytic potency and antitumor efficacy, either by directly boosting viral replication or through synergistic combinations with other therapies. Rigau et al.⁷⁹ demonstrated that miR-99b and miR-485 enhance adenoviral oncolysis by suppressing ELF4, MDM2, and KLF8—genes that negatively regulate E1A and late viral protein expression—resulting in improved virion production; ICOVIR15 adenoviruses expressing these miRNAs showed superior replication and tumor control in PDAC models, significantly reducing tumor growth in mice compared to the parental virus. Similarly, oncolytic VSV expressing amiR-4 exhibited enhanced efficacy against pancreatic tumors when combined with the EZH2 inhibitor GSK126 or CTLA-4 blockade, with amiR-4 transferred via extracellular vesicles to uninfected tumor cells, inducing bystander killing⁸⁰. VSVΔ51-shPD-L1, which uses hairpin RNA to silence PD-L1, also effectively suppressed tumor growth in syngeneic mice without toxicity, highlighting the capacity of miRNA-expressing OVs to remodel the TME⁸⁰.

Beyond enhancing oncolysis, OVs can serve as delivery vehicles for therapeutic miRNAs, enabling precise gene regulation and immune modulation. Brachtlova et al.⁸¹ engineered adenoviruses to express miR-1 (tumor-suppressive) and miR-26b (pro-viral replication). They showed that constructs expressing pri-miRNA with flanking sequences produced higher levels of mature miRNAs and effectively silenced target genes, establishing a robust platform for miRNA-based therapy⁸¹.

Furthermore, Lou et al.²⁹⁹ developed AdCN205, an oncolytic adenovirus co-expressing miR-34a and IL-24, which synergistically enhanced antitumor effects: miR-34a inhibited Bcl2 and Sirt1 to induce apoptosis and block metastasis, while also sensitizing cells to IL-24. This construct also reduced PD-L1 expressing tumor-associated macrophages and promoted Th1 and CD8⁺ naïve T cell infiltration, demonstrating potent immune remodeling alongside tumor-selective cytotoxicity⁸².

4.4.2. lncRNA and OVs as combinatorial therapy

LncRNAs serve as key regulators in both antiviral responses and oncolytic virus-based cancer therapies. During seneca valley virus (SVA) infection, lncRNA 8244 is downregulated and modulates the miR-320–CCR7 axis to activate Toll-like receptor (TLR) signaling and promote IFN-β production, shedding light on lncRNA-mediated antiviral mechanisms⁸⁵. Additionally, engineered adenoviruses expressing artificial lncRNAs targeting oncogenic miRNAs have demonstrated potent therapeutic effects. One design targeting 12 HCC-related miRNAs induced apoptosis and suppressed tumor growth in HCC models⁸³ , while another targeting nine epithelial–mesenchymal transition (EMT)-associated miRNAs in triple-negative breast cancer reduced proliferation, reversed EMT, and enhanced tumor suppressor expression in xenografts⁸⁴.

4.4.3. siRNA and OVs as combinatorial therapy

OVs combined with siRNA exert antitumor effects against different types of cancer. Zhao et al.⁸⁶ developed a novel oncolytic adenovirus, Ad-siERCC1, which carries a siRNA targeting excision repair cross-complementation group 1 (ERCC1) driven by hTERT and HIF promoters. Ad-siERCC1 effectively inhibited the proliferation, migration, and invasion of ovarian cancer cells, and enhanced sensitivity to cisplatin (DDP) by silencing ERCC1, a gene associated with chemoresistance⁸⁶. Mechanistically, Ad-siERCC1 induced G1-phase cell cycle arrest and promoted apoptosis via the PI3K/AKT–caspase-3 pathway in SKOV3 cells⁸⁶. In vivo studies confirmed its potent antitumor efficacy and its ability to overcome DDP resistance in ovarian cancer. The combination of the oncolytic adenovirus H101 and siRNA targeting GNAQ enhanced antitumor effects in GNAQ-mutated uveal melanoma (UM) cells. In OMM2.3 and 92.1 cell lines, this strategy inhibited proliferation and induced apoptosis by suppressing MEK1/2 phosphorylation and promoting YAP phosphorylation, suggesting a synergistic impact through disruption of the GNAQ–MEK–YAP signaling axis⁸⁷.

An earlier study showed that Ki67 promoter-driven AdV carrying siRNAs against Ki67 and hTERT effectively silenced both targets, enhanced viral replication, and induced apoptosis in renal cancer models, highlighting the potential of dual-siRNA virotherapy⁸⁹. A conditionally replicating adenovirus (Ad-TERTp-E1A-1504) driven by the TERT promoter and carrying siRNA targeting EphA3 (1504-siRNA) showed potent, selective cytotoxicity against EphA3-and TERT-positive tumor cells, with minimal effects on normal cells. Compared to controls, Ad-TERTp-E1A-1504 markedly enhanced viral replication and induced autophagy via suppression of the PI3K/AKT/mTOR pathway⁹⁰.

Study found that HVJ-E delivering Eg5 siRNA induces cell-cycle arrest and apoptosis in glioblastoma, achieving complete tumor regression and prolonged survival in mouse models. This synergistic approach holds promise as a novel glioblastoma therapy⁹¹. A non-replicating viral vector, HVJ-E, was used to deliver PD-L1-targeting siRNA (siPDL1/HVJ-E) into glioblastoma cells, aiming to combine immune checkpoint inhibition with viral immunostimulation. In murine glioma models, intratumoral administration of siPDL1/HVJ-E reduced PD-L1 expression, suppressed tumor growth, and prolonged survival⁸⁸. This treatment enhanced brain infiltration of CD8⁺ T cells and NK cells, while reducing Treg/CD4⁺ ratios⁸⁸. CD8⁺ T-cell depletion abolished the therapeutic effect, confirming a CD8⁺ T cell-dependent antitumor response⁸⁸. These findings suggest that HVJ-E-mediated siRNA delivery represents a promising immunovirotherapy strategy for glioblastoma.

Furthermore, extending their epigenetic regulation of innate immune pathways, ncRNAs may serve as a critical molecular bridge linking epigenetic modulators to OV-induced antitumor immunity. Tumor-associated miRNAs and lncRNAs are frequently dysregulated, often attenuating pattern recognition receptor (PRR)-mediated antiviral signaling (e.g., RIG-I/MDA5 and cGAS-STING pathways) and facilitating immune evasion in “cold” tumors²¹². Upon OV infection, viral nucleic acids trigger these PRRs, promoting type I interferon responses, immunogenic cell death, and adaptive immune priming^300,301. Epigenetic inhibitors such as HDACi and DNMTi may reprogram ncRNA expression profiles or disrupt oncogenic ncRNA networks, thereby partially restoring PRR/STING activation, enhancing viral replication/oncolysis, and amplifying OV-mediated cytokine release and T cell infiltration^302,303.

4.5. BETi and AdV as combinatorial therapy

In a PDAC model, RNA-seq analysis revealed that treatment with epigenetic inhibitors OTX-015, iBET-762, and Neo-2734 enhanced viral E1A gene expression⁹². It modulated the expression of cell cycle regulators and inflammatory mediators, and suppressed oncogenes such as c-Myc and Myb in tumor cells⁹². These findings suggest that combining tumor-selective adenoviruses (AdΔΔ and Ad-3Δ-A20T) with epigenetic modulators represents a promising therapeutic strategy for PDAC225.

4.6. Clinical translation challenges and considerations

Preclinical studies indicate promising synergy between epigenetic inhibitors and OVs^304,305. However, clinical translation remains challenging due to several interconnected hurdles. Safety concerns are prominent. As epigenetic drugs cause myelosuppression, gastrointestinal, and cardiovascular toxicities^305,306, which may be worsened by OV-induced immune activation²⁷. Although rare, severe inflammation has been noted; early trials with HDAC inhibitors plus T-VEC show manageable profiles but require careful dose optimization^258,307.

Antiviral immunity and tumor evasion limit efficacy. Neutralizing antibodies and innate responses restrict OV spread, while heterogeneous “cold” tumors reduce benefit from epigenetic suppression of antiviral pathways^18,308. Epigenetic modulation may also inadvertently enhance immunosuppressive cells (e.g., MDSCs)³⁰⁹. Systemic OV clearance and poor tumor penetration, combined with the need for sustained epigenetic drug exposure, call for advanced strategies such as nanoparticle delivery or intratumoral administration^310,311. Resistance mechanisms include compensatory epigenetic pathways, restored antiviral defenses, and selection of resistant subclones, potentially limiting response durability311, 312, 313. Multi-modal approaches are under exploration^54,314.

Clinical trials of oncolytic virotherapy as monotherapy or in combination with immune checkpoint inhibitors or chemotherapy have reached a relatively mature stage. Epigenetic inhibitors have also undergone extensive clinical evaluation. However, the combination of epigenetic inhibitors and oncolytic viruses remains largely at the preclinical and early translational stage, and no large-scale phase II/III trials have yet reported definitive results for this strategy³¹⁵. Addressing these challenges through biomarkers, AI-optimized combinations, and adaptive designs will be critical to translate this synergy into improved outcomes for refractory cancers^12,316.

5. Perspective

The combined application of epigenetic modulation and oncolytic virotherapy demonstrates considerable promise for enhancing antitumor immunity and overcoming drug resistance. This potential is exemplified by the 62% objective response rate achieved with talimogene laherparepvec (T-VEC) and PD-1 inhibitor combination therapy in advanced melanoma (NCT02263508)³¹⁷. Nevertheless, clinical translation encounters three fundamental obstacles: insufficient mechanistic understanding restricts precise interventions, inadequate delivery efficiency compromises intratumoral drug concentrations, and the absence of personalized approaches fails to address highly heterogeneous tumor microenvironments (TMEs)318, 319, 320, 321. Resolving these bottlenecks demands interdisciplinary technological integration (Fig. 5).

Figure 5.

A conceptual framework illustrating the integration of epigenetics and oncolytic virotherapy. Key strategies include: (1) deciphering gene regulatory mechanisms and enhancing delivery systems to improve specificity and reduce off-target effects; (2) leveraging synthetic biology and organoid models to design intelligent oncolytic viruses and enable high-throughput screening; (3) applying AI-driven diagnostics to stratify patients and guide personalized treatment decisions.

5.1. Molecular mechanism deciphering and delivery innovation

Future investigations must comprehensively elucidate molecular interactions across multiple biological scales. At the epigenetic level, this entails revealing how oncolytic viruses restructure CCCTC-binding factor (CTCF)-mediated chromatin loops to activate endogenous retroviral pathways while implementing novel R-loop single-base resolution techniques to decipher emerging gene regulatory mechanisms³²². Recent studies have further highlighted how metastasis-specific epigenetic alterations, including enhancer reprogramming^323,324, CTCF-dependent insulation loss³²⁵, and ncRNA-mediated chromatin rewiring^218,326, 327, 328, 329, 330, 331, promote immune escape, sustain oncogenic transcriptional programs, and drive resistance to current therapies. These insights not only illuminate novel vulnerabilities for oncolytic virotherapy but also provide a foundation for identifying regulatory elements that could be exploited to achieve tumor-selective viral replication and immune activation. Regarding delivery systems, progress requires engineering bacterial outer membrane vesicles (OMVs) for CRISPR-Cas9^332,333 mediated site-specific DNA demethylation alongside developing soluble ligand-modified lipid nanoparticles^334,335 that optimize pharmacokinetic and pharmacodynamic profiles to enhance tumor penetration while mitigating off-target effects^4,336. Together, these mechanistic insights and delivery innovations pave the way for next-generation viro-immunotherapies with enhanced specificity, efficacy, and translational potential.

5.2. AI-driven precision diagnostics, drug design, and personalized treatment

Diagnostic transformation necessitates integrating multimodal technologies including mapping immune-excluded three-dimensional genomic regions through single-cell epigenetic trajectory tracing combined with chromatin conformation sequencing^337,338, monitoring real-time oncolytic virus replication dynamics and drug distribution via molecular imaging^339,340, and deploying artificial intelligence algorithms to identify therapy-responsive patients, thereby establishing molecular subtyping foundations for precision immunotherapy.

Furthermore, AI-assisted models also support dynamic treatment optimization for personalized epigenetic-oncolytic virotherapy. Machine learning and deep learning frameworks have been shown to integrate high-dimensional genomic, transcriptomic, and radiomic data to predict individual treatment responses and recurrence risk, thereby guiding tailored therapeutic decisions and improving clinical outcomes^341,342. In parallel, AI-driven drug design platforms enable the rational discovery and optimization of epigenetic modulators and viral engineering strategies by predicting drug-target interactions, screening candidate compounds, and modeling virus-host interactions in silico³⁴³. AI-driven decision-support systems enable predictive simulations of tumor-virus-immune dynamics and identifying patient-specific biomarkers for susceptibility or resistance. It holds the potential to refine combination regimens, optimize dosing and sequencing, and ultimately realize precision medicine in the context of epigenetic modulation and oncolytic virotherapy.

5.3. Targeted therapeutics and synthetic biology interventions

Therapeutic advancement requires optimizing metabolic-epigenetic synergistic nanodrugs such as DNMTi-loaded acid-catalyzed nanoparticles²⁵⁷ to amplify antitumor immunity³⁶. Concurrently, constructing AI-organoid high-throughput screening platforms^344,345 will facilitate large-scale identification of oncolytic virus-sensitizing small-molecule epigenetic compounds for precision medicine guidance. Ultimately, synthetic biology approaches may enable engineering intelligent oncolytic viruses capable of dually targeting tumor plasticity and immune evasion to achieve precise therapeutic control^309,346. These integrated strategies hold great promise for ushering in a new era of highly personalized and adaptive cancer therapies.

Author contributions

Conception/design: Zhi-Jun Sun, Fuhua Yang; Manuscript preparing and writing: Yuanyuan Wang, Lanyi Zhang and Yunting Che; Preparation of figures: Yuanyuan Wang; Preparation of tables: Lanyi Zhang and Yunting Che. All authors read and approved the final manuscript.

Conflicts of interest

The authors declare no competing interests.