The cell maintains R-loop homeostasis through dedicated enzymes that remove R-loops once formed. Most of these are ribonuclease H (RNASEH)1 and 2, which specifically recognize RNA:DNA hybrids and cleave RNA strands (2). RNASEH1 is active throughout the cell cycle and serves as the first line of defence against R-loops, degrading RNA components and dismantling the hybrid (15). RNASEH2 operates especially during the S/G₂ phases, not only to resolve R-loops but also to excise ribonucleotides embedded in DNA (16).

Along with RNASEH, several DNA and RNA helicases resolve R-loops by unwinding the RNA:DNA hybrid. Senataxin (SETX) is a key helicase that resolves R-loops during transcription termination, preventing persistent hybrids at gene 3′ ends (17). The Asp-Glu-Ala-Asp box (DEAD-box) RNA helicases DEAD-box helicase (DDX)19 and DDX21 are also important; DDX21 unwinds R-loops at ribosomal RNA (rRNA) genes and collaborates with NAD-dependent protein deacetylase sirtuin-7 to suppress R-loops during replication stress (18,19).

Other factors involved in R-loop suppression include the transcription-export (THO) complex/3′ repair exonuclease complex, which couples transcription to mRNA export and various RNA-processing factors, such as splicing factors. Comprehensive mRNA splicing and 3′ end processing are important to prevent R-loops; if transcription outpaces processing or if splicing is defective, the unprocessed nascent RNA is more prone to hybridize with DNA. This increase in hybridization due to insufficient mRNA processing explains why mutations in spliceosome genes, such as splicing factor 3b subunit 1 and U2 small nuclear RNA auxiliary factor 1, can induce R-loop accumulation and activate DNA damage responses (DDRs) (12,14).

Homeostatic mechanisms may be altered in HCC. Tumor cells often have markedly elevated transcriptional activity, which can favor R-loop formation beyond the capacity of normal regulatory mechanisms. There is evidence that HCC cells upregulate certain R-loop-resolving factors, presumably to cope with the increased R-loop load (1,2). In HCC, dysregulated R-loops contribute to genomic instability by inducing replication stress and DNA double-stranded breaks, thereby activating the ataxia telangiectasia and Rad3-related protein (ATR)-checkpoint kinase (CHK)1 and ataxia telangiectasia mutated (ATM)-CHK2 DNA damage response pathways (2,6). Oncogenic drivers such as metastasis-associated protein 2 (MTA2) promote R-loop formation at the promoters of differentiation genes such as β-hydroxybutyrate dehydrogenase 1 (BDH1), leading to epigenetic silencing and enhanced stemness in HCC cells (9). Furthermore, unresolved R-loops can initiate inflammatory responses through cytosolic DNA-sensing pathways, such as cyclic GMP-AMP synthase and stimulator of interferon genes (cGAS-STING), contributing to immune modulation within the tumor microenvironment (TME) (1). For instance, RNASEH1 expression is elevated in breast cancer, colorectal cancer, leukemia, and HCC and associated with poor prognosis (2). One study noted that a long non-coding RNA (lncRNA) antisense to RNASEH1 (RNASEH1-AS1) is frequently upregulated in HCC and associated with worse survival prognosis (10). Additionally, certain oncogenic pathways in HCC actively induce R-loops to alter gene expression, such as the chromatin modifier MTA2, which increases R-loops at specific gene promoters (9). Thus, the balance between R-loop formation and resolution is maintained in HCC cells.

In addition to their impact on DNA stability, R-loops can influence gene expression and the chromatin state. Physiological R-loops often occur in the promoter or terminator regions of genes and have regulatory functions (2). For example, R-loops formed over 5′-cytosine-phosphate-guanine-3′ island promoters maintain an open chromatin configuration by displacing nucleosomes, thereby facilitating transcription initiation of these genes (19). R-loops at the gene termini have been implicated in pausing RNA polymerase II and promoting transcription termination (20,21).

These co-transcriptional R-loops can recruit or occlude specific factors. One study showed that promoter-proximal R-loops can attract DNA demethylases, leading to local DNA hypomethylation and robust gene expression (2). Conversely, stable R-loops on gene bodies or regulatory elements may interfere with transcriptional elongation, effectively repressing these loci. Furthermore, R-loops interact with chromatin-modifying complexes and alter histone marks, modulate chromatin accessibility and either promote or repress gene transcription (5). For example, R-loop structures are recognized by the polycomb repressive complex and DNA methyltransferases, linking hybrid formation to heritable epigenetic changes (2).

These R-loops may be co-opted or dysregulated in cancer cells, including HCC cells. A notable HCC-related example is the MTA2 pathway; MTA2, a component of the nucleosome remodeling and deacetylase chromatin-remodeling complex, increases R-loop formation at the promoter of BDH1, which is a gene that influences both differentiation and metabolism (9,22). The R-loop recruits histone deacetylase (HDAC)2/chromodomain-helicase-DNA-binding protein 4 (CHD4) specifically to the promoter region of BDH1, leading to the epigenetic silencing of BDH1 and stemness in HCC cells (9,23). Thus, an oncogenic factor exploits R-loop formation as part of a transcriptional repression mechanism that favors tumor cell dedifferentiation and self-renewal.

Another example involves the THO complex (THOC). THOC1 knockdown impairs the THO complex's role in co-transcriptional RNA processing and export, causing RNA-DNA hybrids to accumulate, which increases R-loop formation and consequent DNA damage, which paradoxically leads to reduced tumor cell proliferation, indicating that THOC1 aids HCC cells by preventing lethal levels of R-loop accumulation (8). This suggests that HCC cells require proper mRNA processing to avoid transcription-associated damage. Therefore, R-loops serve as both outputs and regulators of transcription. With widespread transcriptional reprogramming, HCC cells are likely to exhibit an altered landscape of R-loops across the genome. The influence of R-loops on oncogenic gene expression is bidirectional: Certain R-loops promote tumor progression by activating oncogenes or enhancing stemness, while others suppress tumor progression by inducing transcriptional repression or DNA damage-mediated checkpoints (1,6). The interplay between R-loops and epigenetic machinery is a frontier area; for instance, how R-loop formation might interface with the frequent epigenetic alterations seen in HCC.

One of the most important aspects of R-loops is their ability to induce DNA damage, if not properly mitigated. During the S phase, the coexistence of transcription and replication machinery on the DNA template can lead to transcription-replication conflicts. R-loops are a prominent source of such conflicts. An R-loop can impede the replication fork, especially when it collides with the transcription complex (2). Conflicts are particularly deleterious in a head-on orientation, when transcription and replication move toward each other on opposite strands, as the replication fork encounters the R-loop and stalled RNA polymerase headlong (23). Head-on transcription-replication conflicts markedly increase the likelihood of fork stalling or collapse, resulting in DNA double-strand breaks (DSBs) and chromosome rearrangements (2,24).

ATR kinase is a central responder to replication stress caused by R-loops. Stalled replication forks with stretches of single-stranded DNA coated with replication protein A activate ATR, which, in turn, phosphorylates CHK1 and coordinates fork stabilization (25). ATR signaling is associated with head-on collisions. Notably, ATR not only halts the cell cycle to provide time for repair, but also actively promotes R-loop resolution (3), as ATR signaling recruits SETX to transcription-replication clash sites to help unwind R-loops (2).

Similar to ATR, the ATM kinase is activated when R-loop-associated DSBs are formed. ATM can promote end resection and homologous recombination (HR) or facilitate non-homologous end joining, depending on the cell cycle stage and local chromatin environment (2). ATM-CHK2 and ATR-CHK1 are both engaged in response to R-loops; however, R-loops that result in DSBs preferentially trigger ATM, whereas those that cause stalled forks without immediate breakage primarily trigger ATR (26).

Notably, the ATR and ATM signaling pathways suppress R-loop-driven instability. R-loop accumulation is notably toxic to cells lacking these kinases or their downstream factors. For example, deficiencies in the Fanconi anemia (FA) pathway lead to hypersensitivity to R-loop-induced DNA damage (27). FA proteins such as FA group D2 protein bind R-loops and stalled forks, stabilizing them and recruiting RNASEH2 to remove RNA:DNA hybrids. Breast cancer susceptibility gene (BRCA)1, apart from its role in HR repair, directly contributes to resolving R-loops by recruiting SETX to stalled transcription sites (28). Similarly, BRCA2 knockdown leads to abnormal R-loop accumulation, DNA breaks and chromosomal aberrations (29).

The relevance of these mechanisms in HCC is supported by several observations. HCC tumors often harbor mutations or dysregulation of DNA repair genes; for example, tumor protein P53 is commonly mutated, whilst BRCA2 or ATM mutations are less frequent but occur in subsets. Even without specific DDR mutations, chronic replication stress in HCC likely indicates that HCC cells rely heavily on ATR-CHK1 signaling to survive (30,31).

If R-loops are prevalent in HCC cells, ATR would be expected to remain continuously engaged in managing replication-transcription conflicts. Recent research has shown that HCC cells with high R-loop scores exhibit an activated defined set of genes involved in the DDR pathway and are more reliant on stress response pathways (1). When R-loop levels become excessive, cells may enter senescence or apoptosis because of irreparable DNA damage (1,6). This could act as a barrier to tumor proliferation, which HCC cells must overcome by enhancing R-loop tolerance. The aforementioned THOC1 example illustrates that reducing the ability of HCC cells to resolve R-loops leads to notable DNA damage and slow proliferation in HCC cells (8).

Furthermore, etiological factors such as HBV can exacerbate genomic instability in HCC via R-loops. HBV genome integration and HBV X protein (HBx) expression interfere with host DNA repair. Although not yet supported by evidence, it is conceivable that HBV infection facilitates R-loop accumulation, for instance, by causing replication stress or by HBx-mediated downregulation of the RNASEH or ATR pathways (32).

In summary, R-loop-induced DNA damage is likely associated with myriad sources of genomic instability in HCC. The ability of HCC cells to survive and proliferate may depend on how effectively they activate ATR/ATM and associated repair factors to manage R-loop conflicts.

Aberrant R-loop accumulation can provoke DDRs, contributing to genomic instability in liver cancer (6). THOC1 knockdown triggers R-loop accumulation and DNA damage and curbs tumor proliferation (8). Excess R-loops can also initiate innate immune signaling; cytosolic RNA:DNA hybrids excised from nuclear R-loops activate the cGAS-STING pathway and related receptors, leading to interferon regulatory factor 3-driven inflammatory cascades and tumor cell apoptosis (13). Furthermore, the dynamics of R-loops intersect with cancer metabolism. High R-loop levels are associated with metabolic reprogramming, for instance, the HCC R-loop score correlates with elevated lipid synthesis and an immunosuppressive microenvironment (1). Elevated lipid synthesis provides energy and membrane components to support rapid HCC cell proliferation, while an immunosuppressive microenvironment inhibits anti-tumor immune responses, together promoting tumor growth and progression. Upregulation or increased activity of specific R-loop regulators such as clathrin heavy chain 1 (CLTC) enhances fatty acid metabolism, providing energy and biosynthetic substrates that support tumor growth and survival (1). Finally, R-loops interact with epigenetic mechanisms and result in both tumor promotion and suppression; the chromatin modifier MTA2 creates R-loops to repress BDH1. The increase in histone β-hydroxybutyrylation is a direct consequence of BDH1 repression. BDH1 normally metabolizes β-hydroxybutyrate (BHB); when BDH1 is repressed, BHB accumulates, serving as a substrate for histone β-hydroxybutyrylation, which promotes pro-tumorigenic gene expression and stemness (9).

Similarly, RNA-binding proteins such as G patch domain-containing protein 4 (GPATCH4) guard ribosomal DNA loci by unwinding nucleolar R-loops. This activity of GPATCH4 primarily drives HCC progression. By unwinding nucleolar R-loops, it maintains rRNA transcription and protein synthesis, supporting rapid proliferation and tumor growth. However, GPATCH4 could also represent a therapeutic vulnerability, as its inhibition may increase R-loop accumulation, impair ribosome biogenesis, and selectively stress HCC cells (33). Collectively, these findings indicate that R-loops engage multiple signaling pathways, including DNA damage repair, immune surveillance, metabolic adaptation and epigenetic regulation, all of which can drive HCC progression or present therapeutic vulnerabilities (34).

Dysregulation of R-loops in HCC is associated with context-dependent pro- and antitumor effects (6). Abnormal R-loop accumulation causes genomic instability and DNA breaks that, beyond repair thresholds, trigger replicative senescence or cell death, thereby suppressing cancer proliferation (6). For example, unresolved R-loops arising from THOC1 depletion or GPATCH4 loss in HCC cells can lead to DNA damage, impaired proliferation and heightened chemosensitivity (8). Thus, cancer cells often upregulate R-loop-resolving factors such as THOC1, GPATCH4 and tonicity-responsive enhancer-binding protein to avoid lethal R-loop stress.

By contrast, dysregulation of R-loop formation can drive HCC progression. R-loop formation can activate oncogenic pathways or silence tumor suppressors (1,6). As aforementioned, MTA2 induces R-loops at the BDH1 locus, repressing differentiation-linked metabolism and sustaining cancer stemness (9). Similarly, elevated R-loop levels (via CLTC or other R-loop regulators) reprogram lipid metabolism and correlate with an immunosuppressive microenvironment, ultimately promoting HCC proliferation (1). Excessive R-loops activate cGAS-STING signaling, inducing type I interferons and cytokines that recruit cytotoxic immune cells, which directly kill tumor cells and enhance immune surveillance, thereby constraining HCC growth and progression (1).

R-loop dysregulation contributes to the malignant behavior of HCC cells and is closely associated with changes in the TME. As aforementioned, MTA2 promote R-loop accumulation at the BDH1 locus, silencing this metabolic gene via the recruitment of HDAC2/CHD4 and enhancing cancer stemness and invasiveness (9). Additionally, elevated R-loop levels in HCC have been shown to upregulate lipid metabolic pathways via regulators such as CLTC, promoting tumor cell proliferation and tumor aggressiveness (1). These metabolic shifts reshape the TME by increasing immunosuppressive signaling and nutrient competition with paracancerous cells. Furthermore, unresolved R-loops generate cytosolic DNA fragments that activate cGAS-STING signaling, drive type I interferon responses and alter immune cell infiltration (6,35). Hypoxic or inflammatory TMEs may also exacerbate R-loop formation by impairing RNA processing or TOP activity, further aggravating replication stress (3). Thus, R-loops and HCC TMEs exist in a bidirectional relationship that fuels tumor progression.

As described, ATR kinase is important for cancer cells to survive R-loop-driven replication stress. Therefore, inhibition of ATR can experimentally reveal the lethal consequences of R-loops on tumor cells. Preclinical studies have shown strong synergy between ATR inhibitors and R-loop accumulation (24,38). For example, cancer cells with RNASEH2 dysfunction accumulate R-loops and rely on ATR to mitigate the harmful effects of accumulated R-loops; treating these cells with an ATR inhibitor causes notable DNA breaks and apoptosis (1). The same study found that cells harboring spliceosome mutations accumulate R-loops and show selective sensitivity to ATR inhibitors (1). Translating this to HCC, ATR inhibitors may be effective therapeutic agents in tumors with a high R-loop load or impaired R-loop resolution. ATR inhibitors are currently in clinical trials for solid tumors and hold potential for use in combination with an agent that elevates R-loops for the treatment of HCC (28).

PARP1 inhibitors are the preferred treatment option for tumors with HR deficiencies. There is an emerging connection between R-loop incidence and sensitivity to PARP inhibition (39). R-loops can cause replication fork collapse and one-ended DSBs, which normally require HR for repair (40,41). If HR is compromised, cells become markedly reliant on PARP-mediated repair of replication-associated breaks, making them vulnerable to PARP inhibition; in cancer and HCC, PARP inhibitors exacerbate DNA damage, promote R-loop-induced genomic instability and trigger apoptosis, selectively suppressing tumor proliferation (39). Furthermore, R-loops can sequester BRCA1 and prevent it from localizing to DNA damage sites, impairing homologous recombination repair (12). Although BRCA1 is wild-type, its functional unavailability creates an HR defect, increasing genomic instability (12). In Ewing sarcoma cells, transcription of the EWS RNA binding protein 1-friend leukemia integration 1 transcription factor fusion oncogene leads to R-loops that sequester BRCA1, rendering the cells sensitive to PARP inhibitors, similar to BRCA1 mutants (12). Similarly, conditions in HCC that compromise HR repair can potentially be exploited using PARP inhibitors.

Compounds that perturb RNA processing can induce R-loops by leaving nascent transcripts unprocessed. For example, spliceosome inhibitors such as sudemycin and H3B-8800 cause widespread splicing disruption, leading to R-loop accumulation and DNA damage in cancer cells (14). These inhibitors are proposed for splice-mutant cancers because such cells already have impaired RNA splicing, which increases baseline R-loop formation. Inhibiting relevant pathways exacerbates R-loop accumulation beyond a tolerable threshold, overwhelming DNA repair and promoting cancer cell death (14). In HCC, direct spliceosome mutations are rare; however, the use of a splicing modulator can artificially create an R-loop overload.

In addition to conventional drugs, gene editing and molecular biology techniques offer ways to directly manipulate R-loops and their regulators in cancer cells. One innovative approach involves the use of engineered RNASEH to eliminate R-loops at specific genomic sites. Researchers have developed a CRISPR/dCas9-RNASEH1 fusion protein, which can be guided by single guide RNAs to genomic loci and locally resolve R-loops (42,43). This tool has been used experimentally to target tumorigenic R-loops; for instance, site-specific R-loop removal improved cellular reprogramming efficiency in cultured cells (42,43).

In principle, a similar strategy could be used to target a pathogenic R-loop in HCC; for example, if a particular R-loop silences a tumor suppressor gene, a targeted RNASEH1 might reactivate that gene. Conversely, gene editing can be used to create synthetic vulnerabilities in tumor cells. The CRISPR knockout of R-loop in HCC cells removes backup mechanisms that normally compensate for each other, exacerbating DNA damage (11). This heightened stress increases sensitivity to DNA-damaging agents, such as TOP inhibitors, ATR inhibitors and PARP inhibitors, which exploit defects in replication and repair pathways. Another gene-focused strategy targets non-coding RNAs that modulate R-loops. RNASEH1-AS1, an antisense lncRNA upregulated in HCC, may attenuate RNASEH1 function. Using antisense oligonucleotides to knockdown RNASEH1-AS1 could free RNASEH1 to better resolve R-loops, potentially reducing genomic instability in tumor cells.

The concept of synthetic lethality, in which two non-lethal perturbations kill the cell when combined, is notably relevant to R-loops (11). The inherent DNA damage in cancer cells, arising from defects in HR repair or RNASEH function, activates compensatory pathways like ATR signaling. ATR responds to replication stress and R-loop-induced DNA breaks, stabilizing replication forks and coordinating repair to maintain cell survival. The present review has already discussed some synthetic lethal pairs, such as RNASEH2 loss and ATR inhibition (11) or splicing mutation and ATR inhibition (14).

One combined strategy for therapeutic application to HCC could be coupling ATR inhibitors with agents that increase R-loop formation, as aforementioned. Another notable possibility is the combination of R-loop targeting and immunotherapy. HCC treatment has included immune checkpoint inhibitors (ICIs) in recent years (44–46). There is evidence that tumors with increased DNA damage can be more immunogenic due to the activation of cGAS-STING and type I interferon responses (35). Excess R-loops can activate the cGAS-STING pathway, as displaced single-stranded DNA or R-loop-driven DNA breaks generate ligands for this pathway (1).

A transient burst of R-loop-induced damage may acutely increase the immunostimulatory signals via the activated cGAS-STING and type I interferon pathways. Therefore, a short course of R-loop-inducing therapy can enhance tumor immunogenicity by increasing DNA damage and immune signaling, making individual tumors, including HCC, more susceptible to immune checkpoint inhibitor-mediated targeting (13,47).

Another scenario using synthetic lethality could involve targeting metabolic pathways that become lethal when the number of R-loops is high. Single-cell analysis by Chen et al (1) highlighted a link between R-loops and lipid metabolism in HCC. The study identified an 8-gene signature of fatty acid metabolism-related R-loop regulator genes; inhibiting CLTC reduces fatty acid metabolism, decreasing lipid synthesis and accumulation in HCC cells. This metabolic disruption impairs tumor cell proliferation and overall tumor growth (1).

The accurate detection of R-loops in cells and tissues is important. The primary method, DNA-RNA immunoprecipitation (DRIP) using the S9.6 monoclonal antibody, has known limitations in terms of specificity (48). Thus, a key challenge is the identification of patients with HCC who have high R-loop levels or dysfunctional R-loop regulation. Without reliable diagnostic assays, patient stratification for R-loop-targeted therapy is difficult.

Targeting R-loops can affect normal cells because they are not unique to cancer cells. All rapidly dividing cells generate R-loops (49). Therapies such as ATR inhibitors or TOP poisons also affect healthy proliferating tissues, causing side effects, such as bone marrow suppression (anemia, neutropenia), gastrointestinal toxicity (nausea, diarrhea) and hair loss (12). However, cancer selectivity remains a major challenge. Cancer selectivity is observed in cells with ~2–3 times higher R-loop levels than normal tissues, providing a therapeutic window for targeting R-loop-dependent vulnerabilities.

As with any cancer therapy, tumors may develop resistance to R-loop-targeting strategies. There are several plausible resistance routes: i) A tumor under ATR inhibitor pressure may upregulate compensatory pathways such as ATM or DNA-dependent protein kinase catalytic subunit (50); ii) a HCC cell may slow its proliferation or transcription rate to manage R-loop stress (51); iii) if a small molecule such as an ATR inhibitor is used, classical drug-specific resistance mechanisms can occur (52,53); and iv) the liver TME might shield HCC cells from treatments.

A number of R-loop-targeting approaches run the risk of causing global DNA damage if not controlled. It is important to avoid therapies that could induce notable genomic instability in healthy tissues, resulting in secondary malignancies or organ failure.

The mechanisms of R-loops in HCC remain yet to be fully elucidated. Most mechanistic insights have been obtained from cell-line studies or other cancer types such as breast cancer, colorectal cancer and leukemia. Care must be taken when suggesting that HCC has unique aspects that could influence R-loop dynamics in ways that have not yet been fully understood.

A more comprehensive understanding of R-loop dynamics in HCC may be achieved by integrating multiple layers of data: Genomics, transcriptomics, epigenomics and proteomics. At single-cell resolution, it is possible to profile transcription and chromatin states, while R-loop mapping remains largely bulk; emerging methods like scDRIP-seq combined with single-cell ATAC-seq or RNA-seq enable integrated analyses. These investigative methods are capable of refining the current understanding of where R-loops form in HCC genomes, as well as their causes and consequences.

The present study has focused on known molecules that interact with R-loops, such as ATR and RNASEH; however, future studies may unveil HCC-specific interactions. The list of R-loop regulatory genes used in recent HCC studies (1) provides a starting point for identifying novel HCC-specific interactions that includes unexpected genes that are not previously linked to canonical R-loop regulation or cancer progression pathways.

As therapies that exploit R-loops are introduced to clinical settings, the parallel development of biomarkers is needed. Potential biomarkers could be the expression of an R-loop resolver, mutations in an R-loop-related gene or the direct measurement of RNA:DNA hybrids in tumor samples, if feasible.

Future research should aim to overcome the challenges outlined in the present review. Nanoparticle- or virus-based delivery of gene therapies may be refined for drug delivery. With small-molecule approaches, prodrugs that are activated in the TME can localize the R-loop-targeting effect (52,53). This refers to a subset of drugs, specifically those designed to target R-loops or exploit R-loop-associated vulnerabilities.

The relationship between R-loops and the immune system is an area of study with direct implications for HCC (1,13). Future studies could clarify this by manipulating R-loops in vivo and observing immune responses.

Finally, the culmination of these future directions of study will be clinical trials that test R-loop-targeted therapies in patients with HCC. For instance, a trial of an ATR inhibitor in combination with an approved HCC therapy could be designed for patients whose tumors exhibit a predefined R-loop signature.