In physiological homeostasis, the liver functions as the principal organ for lactate clearance in the human body, sustaining lactate balance through the glucose-pyruvate-lactate metabolic cycle. In the development of HCC, liver cancer cells are often accompanied by the 'Warburg effect'. This metabolic pattern will significantly increase the production of lactic acid and reduce the clearance ability, leading to a large accumulation of lactic acid within the hepatic TME (9). The buildup of lactate furnishes the necessary substrate for the alteration of proteins by lactate, hence substantially enhancing the likelihood of these modifications taking place.

Lactylation is a recently identified important post-translational modification (PTM). In the context of tissue dynamic metabolic balance, some lactate generated by cells is involved in energy metabolism, while another fraction aids in epigenetic changes and non-histone lactylation processes (12). Specifically, lactylation is a chemical process induced by increased lactate concentrations in the TME, during which lactate groups are preferentially appended to lysine residues on proteins. It has been identified as a fundamental method for HCC cells to acclimatize to metabolic stress and modulate the tumor immunological microenvironment (13).

In the process of enzyme-dependent protein acylation modification, there is usually a classic regulatory mode of 'write-recognition-erase': The 'writers' are responsible for transferring the acyl group in acyl-CoA to the side chain of amino acid residues such as lysine, glycine, cysteine or serine in the protein; the 'readers' recognize and bind to the modified amino acid residues through specific structural domains, and initiates the downstream functional effect; the 'erasers' can remove the acyl group on the amino acid residues, so as to realize the dynamic reversible regulation of the modification process (14).

Since both lactic acid and alanine are three-carbon compounds with carboxylic acid groups, their molecular structures are highly similar. The core difference lies only in that alanine contains an amino group (-NH₂), while lactic acid contains a hydroxyl group (-OH). Consequently, it may be deduced that lactylation adheres to the previously indicated 'write-recognize-erase' regulatory framework (Fig. 1). The alteration process is predominantly governed by four principal variables, as indicated by current study advancements (Table I).

Lactate transferases function as the principal enzymes that initiate enzymatic lactylation. The presently recognized members predominantly belong to the histone acetyltransferase (HAT) family and the aminoacyl-tRNA synthetase (AALS) family. P300, as a member of the traditional HAT family, was the first identified lactoyl-transferase capable of dynamically modulating histone lactylation levels. Experimental results indicated that the inhibition of p300 function through gene knockdown or pharmacological inhibitors markedly decreases intracellular histone lactylation levels (7,15,16). Subsequent investigations showed that the inhibition of p300/CREB-binding protein (p300/CBP) acetyltransferase activity or the suppression of p300/CBP gene expression in macrophages significantly reduced the lactate-induced lactylation of the high mobility group box 1, indicating that p300/CBP also plays a role in non-histone lactylation (17). Of note, the evidence supporting p300 as a lactylation writer is still largely based on correlative and mechanistic studies, and its substrate selectivity in HCC requires a deeper validation. AALS1/2, functioning as ATP-dependent lactate transferases, specifically recognize and bind lactate through designated domains. Utilizing ATP as an energy donor, they transform lactate into the high-energy intermediate 'lactyl-CoA', subsequently facilitating the lactylation modification of lysine residues in substrate proteins (18-21). HAT binding to ORC1, a member of the MOZ, Ybf2/Sas3, Sas2, Tip60 HAT family, operates as a lactylating transferase. It aggregates at transcription start sites and facilitates carcinogenesis by modulating histone H3 lysine 9 lactylation and oncogene expression (22).

Deacetylases are essential components that sustain the dynamic balance of histone acetylation modifications. Previous studies have preliminarily identified a variety of enzymes with delactylase activity, mainly including members of the histone deacetylase (HDAC) family and members of the silencing information regulator (SIRT) family (23,24). Among them, studies have proved that HDAC1-3 are able to remove the lactylation of histone H3 lysine 18 (H3K18) (24,25), among which HDAC3 has the strongest delactylase activity (26). SIRT2 and SIRT3 are also delactylases, but their activity is markedly lower than that of HDAC3. Zessin et al (27) found that the delactylase activity of SIRT2 is several order of magnitude lower than that of HDAC3. However, in general, the identification of delactylases is still incomplete, and their specific role and regulatory network in liver cancer have not been fully clarified. This remains a key area for future research.

The upstream metabolic regulation maintains the balance between lactate production and clearance, hence ensuring a consistent substrate supply for lactylation and acting as a vital upstream regulator. This process is predominantly regulated by two fundamental metabolic pathways: As tumor cells expand fast, the requirement for cellular energy escalates. At this time, lactate dehydrogenase (LDH) becomes a key regulatory enzyme of metabolic flux.

Under the catalysis of LDH, pyruvate is reduced to lactate, providing a core substrate for lactylation (28-30). In contrast to the role of LDH, pyruvate dehydrogenase (PDH) can catalyze the oxidation and decarboxylation of pyruvate to generate acetyl-CoA, which enters the tricarboxylic acid (TCA) cycle for complete oxidation and energy supply. This procedure diverts pyruvate from the lactate generation pathway, facilitating indirect lactate clearance. The PDH-catalyzed process is irreversibly functioning as a critical 'shunt valve' that preserves intracellular lactate homeostasis (31).

In addition to LDH and PDH, the polyol pathway constitutes a metabolic branch that indirectly modulates the intracellular lactate pool available for lactylation. Under the hyper-glycolytic conditions in HCC, aldose reductase (AR) competes directly with the glycolytic pathway for the glucose (32). This competitive metabolic shunting effectively diverts glucose flux, thereby limiting pyruvate availability and the subsequent generation of lactate (32,33). Through this mechanism, elevated AR activity functions as a metabolic gate that attenuates intracellular lactate accumulation and constrains the substrate supply for downstream Kla. Concurrently, the consumption of NADPH by AR may further perturb the redox environment required for the optimal function of lactylation regulatory enzymes (34). Therefore, the polyol pathway represents an indirect upstream modulator that governs lactylation substrate availability in HCC.

Readers bind lactylated lysine residues and mediate downstream transcriptional regulation. Recent studies have identified several candidate readers. Brahma-related gene 1 (BRG1), a core subunit of the SWItch/sucrose non-fermentable chromatin-remodeling complex, recognizes histone H3 lysine 18 lactylation (H3K18la) and promotes transcriptional activation through chromatin remodeling (35). Double PHD fingers 2 preferentially binds H3K14la and facilitates tumor-associated gene expression, linking lactylation to oncogenic transcriptional programs (36). In addition, the bromodomain-containing protein tripartite motif-containing 33 can directly recognize lactylated lysine residues, suggesting that canonical acetyl-lysine reader domains may also accommodate lactylation (37). However, the number of identified lactylation readers remains limited, and their functional roles in HCC require further investigation.

Lactylation is essential for maintaining the equilibrium between proliferation and death in HCC cells. Research has shown that Kla of cyclin E2 (CCNE2) accelerates cell cycle progression and promotes HCC cell proliferation. This process is governed by SIRT3. Honokiol (HKL), acting as an activator of SIRT3, enhances the delactylase activity of SIRT3 toward CCNE2, which induces apoptosis in HCC cells and suppresses tumor growth in vivo (50). Lactylation at the K124 location of centromeric protein A similarly enhances its transcriptional regulatory function. This modification synergizes with transcription factor Yin Yang 1 to promote mitosis and proliferation in liver cancer cells (71). In addition, certain natural substances can demonstrate anticancer properties by inhibiting lactylation. Demethyl zelasterol and royal jelly acid can reduce the level of histone lactylation, thus inhibit the proliferation and migration of liver cancer cells (13,72).

Lactylation can also affect the invasion and metastasis of HCC. Yang et al (73) found that the lactylation of AK2 at the K28 site will influence the p53 signaling pathway, thus promoting the proliferation and metastasis of tumor cells through lactomics analysis of hepatitis B virus-related HCC samples. Furthermore, particular lactylation sites of ubiquitin-specific protease 14 and ATP-binding cassette subfamily F member 1 in HCC samples may function as prospective molecular markers for HCC metastasis (74). Of note, insufficient microwave ablation is a contributing factor to HCC progression. Studies reveal that sublethal thermal stress induces elevated lactate levels in HCC cells, accompanied by enhanced histone lactylation, which promotes cell migration and drug resistance, thereby accelerating tumor progression (75). While these findings are compelling, most evidence is derived from in vitro or preclinical models, and their physiological relevance in human HCC remains to be fully validated.

HCC cells depend on effective glycolysis to satisfy their accelerated energy requirements for proliferation (76). Lactylation forms a positive feedback loop in this process, promoting HCC initiation and progression. On the one hand, lactate produced by glycolysis provides ample substrate for lactylation. On the other hand, lactylation, particularly histone lactylation, activates gene expression of multiple key glycolytic enzymes, including hexokinase 2 (77,78), further enhancing lactate production and sustaining high glycolysis.

Meanwhile, lipid metabolism in HCC cells is also significantly reprogrammed to support cell membrane synthesis and energy storage (79). It has been found that in non-alcoholic fatty liver disease, the lactylation of fatty acid synthase at the K673 site inhibits its enzymatic activity, thereby mediating the downregulation of hepatic lipid accumulation by mitochondrial pyruvate carrier (MPC) 1 and reducing lipid accumulation (80). However, in certain cases, lactylation can also promote lipid synthesis. Meng et al (81) demonstrated that lactate dehydrogenase A (LDHA)-induced mutations in the H3K18lac gene can increase total cholesterol and triglyceride levels, as well as the expression of fatty acid synthesis-related proteins stearoyl-CoA desaturase 1, fatty acid synthase and sterol regulatory element-binding protein 1, providing material and energy reserves for the rapid growth of liver cancer cells. However, the conclusions are mainly based on specific cell models, and their universality in the complex metabolic environment of the human body still needs to be verified.

Furthermore, lactylation is intricately linked to resistance against ferroptosis. Ferroptosis depends on iron molecules, which is a type of programmed cell death characterized by lipid peroxidation (82). An increasing number of studies have shown that ferroptosis can closely influence the proliferation, metastasis and drug resistance of HCC. It has been demonstrated that following sublethal thermal stress, such as microwave ablation, lactate levels and histone lactylation (particularly H3K18la) are significantly upregulated in HCC cells. Histone lactylation promotes the expression of cysteine desulfurase NFS1, thus enhancing the synthesis of iron-sulphur clusters. Through this step, the resistance of HCC cells to iron death is improved, which ultimately promotes their survival and metastasis (75). This elucidates a novel mechanism through which lactylation modifications impart therapeutic resistance to HCC cells by modulating the ferroptosis pathway.

In liver cancer cells, enhanced glycolysis leads to a large accumulation of lactates, which acts as a substrate to drive lactylation modification of histones, thereby triggering changes in gene transcription programs related to tumor malignancy. Numerous histone lactylation sites have been shown to be intricately associated with malignant progression in HCC. H3K18la is among the most thoroughly investigated lactylated modifications. It primarily accumulates in the promoter regions of active genes and is considered an activating epigenetic marker (7). Research demonstrates that O-linked β-N-acetyl-glucosamin glycosylation of Y-box binding protein 1 (YBX1), a crucial regulator of glycolysis, stabilizes pyruvate kinase M2 mRNA and increases glycolytic activity, consequently leading to increased H3K18la levels. H3K18la enhances YBX1 expression, creating a positive feedback loop that stimulates metabolic activation and malignant advancement in HCC (83). Furthermore, an elevated histone H3 lysine 56 lactylation (H3K56la) has been demonstrated to be closely associated with the tumorigenicity and stemness maintenance of liver cancer stem cells (LCSCs) (84). In LCSCs, H3K56la levels are significantly upregulated, and its enrichment directly activates the transcription of multiple stemness-related genes, including the oncogene MYC and the pro-angiogenic factor vascular endothelial growth factor A (VEGFA), promoting the self-renewal and tumorigenicity of LCSCs (85). Targeting lactate production or interfering with lactylation can effectively reduce H3K56la levels, thereby inhibiting the malignant phenotype of LCSCs and providing a new target for the treatment of HCC.

The epigenetic regulation of lactylation also transpires through non-histone mechanisms. The chromatin remodeling protein BRG1 has been recognized as a 'reader' for histone lactylation (86). It recognizes and binds to H3K18la-enriched promoter regions, utilizing its ATPase activity to remodel chromatin structure and alter its accessibility. This mechanism enables BRG1 to mediate lactylation-driven transcriptional reprogramming, profoundly influencing cellular pluripotency and tumor progression (Fig. 2) (35). Beyond the aforementioned examples, an expanding repertoire of lactylation targets has been implicated in HCC pathogenesis and therapeutic resistance, which are systematically summarized in Table II.

Tregs are a crucial cell type that sustains immunological tolerance and inhibits antitumor immune responses, with their differentiation and function significantly influenced by the tumor metabolic environment (105). In the high-lactic acid, low-glucose microenvironment of HCC, Tregs actively absorb lactate through the pronounced expression of monocarboxylate transporter 1 (MCT1). They convert lactate to pyruvate, which then enters the TCA cycle to support mitochondrial oxidative phosphorylation (OXPHOS) energy production. This allows Tregs to sustain functional stability amid metabolic stress and assume a predominant role within the tumor immunosuppressive network (106,107). Watson et al (106) conducted research indicating that the Treg-specific deletion of MCT1 markedly diminished lactate absorption and immunosuppressive function, hence reinforcing the essential role of lactate metabolism in Treg homeostasis. However, the complete mechanism by which it regulates immunosuppressive molecules in the complex TME still needs further investigation.

Gu et al (108) found that in the HCC model, the lactate produced by tumor cells, when absorbed by Tregs, causes the lactylation of lysine 72 in the intracellular membrane-organizing extension spike protein (MOESIN) protein (MOESIN lysine 72 lactylation). This alteration increases the binding affinity of MOESIN to the transforming growth factor beta receptor type I, consequently augmenting the transforming growth factor-β/SMAD family member 3 signaling cascade. This facilitates the consistent production of the transcription factor forkhead box P3 (FOXP3) and amplifies the immunosuppressive capabilities of Tregs. The lactic acid accumulated in the TME can be absorbed by Tregs and converted into lactyl-CoA in the cell. As a substrate, it can promote p300/CBP-mediated histone H3 lactate modification (109). These modifications accumulate in promoter regions of genes closely associated with Treg function, particularly the FOXP3 gene, activating the transcription of key immune regulatory genes such as FOXP3, and thereby reinforcing the immunosuppressive phenotype of Tregs (110).

Tumor-associated macrophages (TAMs) constitute a significant population of immune cells within the HCC TME, serving a pivotal function in tumor initiation, development and immune evasion. TAMs can be traditionally classified into pro-inflammatory, antitumor M1-type and anti-inflammatory, pro-tumor M2-type based on their functional and phenotypic properties (111). In the TME of HCC, macrophages exhibit significant plasticity. After being regulated by signals such as tumor cell metabolites, hypoxia and cytokines, most are induced to polarize into the pro-tumor M2 TAMs (112). Clinicopathological analysis has shown that, in HCC tissues, the high infiltration of M2 TAMs (cluster of differentiation CD206⁺ or CD163⁺) is closely associated with tumor invasiveness, angiogenesis and poor prognosis (113,114). Numerous evidence has confirmed that high lactic acid accumulation in the TME can induce the polarization of TAMs into the immunosuppressive M2 phenotypes (115). The study by Zhang et al (7) first revealed that lactate from tumor cells can be taken up by macrophages through MCTs and mediated by histone Kla under the catalysis of acetyltransferase p300/CBP. This process exhibits a 'lactate clock'-like dynamic amplification, progressively enhancing M2-associated gene expression as lactate accumulates (7). Chromatin immunoprecipitation sequencing analysis showed that H3K18la enriched in the promoter regions of M2-associated genes such as arginase 1 and VEGFA, directly promoting their transcriptional activation and TAM from M1 phenotype to M2 phenotype, enhancing the immune escape of tumors. In addition, lactylation changes enhanced the stability of the M2 phenotype through the regulation of metabolic reprogramming. Lactate increases tumor-associated macrophage reliance on OXPHOS and the TCA cycle (116), further stabilizing the M2-like metabolic phenotype and reinforcing immunosuppressive functions. A recent study highlighted the significant functional heterogeneity of TAMs in HCC. Emerging evidence has suggested that specific subsets, such as CD48⁺ macrophages, play an irreplaceable role in shaping the immunosuppressive landscape and are priority targets for next-generation immunotherapies (111).

The accumulation of lactate markedly diminishes the anticancer efficacy of cytotoxic T lymphocytes (CTLs). It has been shown that lactate can inhibit the activity of the CTL effect by activating the c-Jun N-terminal kinase/c-Jun signaling pathway. Simultaneously, lactate reduces the release of inflammatory cytokines such as interferon (IFN)-γ and TNF-α, and downgrades the expression of perforin and granzyme B, thus significantly weakening the cytotoxic response mediated by CTL and weakening the antitumor immune response (117).

Natural killer (NK) cells are also strongly inhibited by lactate. Increased lactate concentration in tumor tissues can weaken NK cell function through multiple pathways, including by downregulating the expression of surface-activated receptor NKp46, inhibiting mechanistic target of rapamycin signaling and preventing the nuclear translocation of promyelocytic leukemia zinc finger, thereby reducing metabolic activity and cytotoxicity (76,118-120). In addition, the lactate-mediated acidic environment can also directly induce the apoptosis of liver-resident NK cells, further weakening innate immune defense (76). However, the aforementioned study mainly focuses on the CRLM model, and its universality in HCC still needs further exploration.

Dendritic cells (DCs) are key antigen-presenting cells that bridge innate and adaptive immunity. A recent study has shown that lactate can activate the sterol regulatory element-binding protein 2 pathway (121). Through the action of Toll-like receptor (TLR), lactate can upregulate IL-10 and inhibit the expression of IL-12, thus hindering the maturation and antigen presentation function of DC. In melanoma and prostate cancer, there are tumor-associated DCs, which are characterized by reduced IL-12 secretion and reduced CD1a expression. Gottfried et al (122) found that lactate accumulation can induce this DC phenotype (122). Plasmacytoid DCs (pDCs) can produce type I interferon, which promotes tumors to avoid immune attacks. On the contrary, lactate can activate the G-protein-coupled receptor 81 (also known as HCAR1) receptor, induce cytoplasmic Ca²⁺ mobilization and then downgrade the IFN-I, thus weakening its antitumor immune activity (123). In addition, glycolysis functions as the principal energy source for pDC IFN-γ generation upon TLR stimulation; thus, lactate accumulation hinders pDC energy metabolism by obstructing glycolysis, therefore further reducing their antitumor efficacy (124).

Lactate is a principal metabolic byproduct in the TME, which functions both as an indicator of active glycolysis and an essential signaling molecule. It promotes angiogenesis through multiple mechanisms, hence endorsing persistent tumor cell proliferation and invasion in hypoxic environments. Depending on whether hypoxia-inducible factor (HIF) is involved, the pro-angiogenic effect mainly involves the following two mechanisms.

In a hypoxic environment, the accumulated lactate can stabilize HIF-1α, thereby enhancing multiple angiogenesis-related pathways and promoting neovascularization. Lactate specifically competes with 2-ketoglutarate for binding and decreases the action of pyruvate dehydrogenase E1 component subunit beta (PDH2) (125). Decreased PDH2 activity inhibits its degradation and then impairs HIF-1α hydroxylation through the PHD2/von Hippel-Lindau pathway, thereby increasing HIF-1α stability (126). Stable HIF-1α can increase the expression of downstream related factors (including VEGFA and basic fibroblast growth factor), which can promote the proliferation of tumor-associated endothelial cells and angiogenesis (127).

In addition to stabilizing HIF-1α, lactate can also promote angiogenesis independently of HIF signaling. After entering tumor-associated endothelial cells through MCT1, lactate can activate the NF-κB pathway by inhibiting the hydroxylation of IκB kinase by PDH2, thereby enhancing IL-8 transcriptional expression and promoting angiogenesis maturation and stabilization (128). Furthermore, lactate can also inhibit the PHD2/VHL-dependent degradation of N-Myc downstream regulatory gene 3 (NDRG3) by binding to it, further prolonging the half-life of NDRG3. Under hypoxic conditions, stable NDRG3 and cellular rapidly accelerated fibrosarcoma (C-Raf) jointly activate the Raf-extracellular signal-regulated kinase pathway and promote angiogenesis (129). This constitutes the HIF-independent lactate-NDRG3-C-Raf axis, which is an important compensatory pathway for tumors to maintain blood supply in a hypoxic environment.

In conclusion, lactylation and lactate signaling jointly participate in the multifaceted regulation of HCC angiogenesis, providing a new metabolic target for anti-angiogenic therapy.

Early diagnosis and recurrence prediction of HCC still face significant challenges. Currently, commonly used clinical biomarkers, such as α-fetoprotein, lack sufficient specificity, and therefore new molecular indicators are urgently needed. Zhang et al (130) established a dual-purpose model based on lactate-associated genes based on multi-cohort data from The Cancer Genome Atlas (TCGA; https://www.cancer.gov/ccg/research/genome-sequencing/tcga) and Gene Expression Omnibus (GEO; https://www.ncbi.nlm.nih.gov/geo/). Their analysis showed that specific markers, including basigin, lysine acetyltransferase 2A and zinc finger E-box binding homeobox 1, not only had a strong diagnostic performance (area under the curve of >0.8) in distinguishing between HCC and normal tissue, but also effectively predict patients' response to immunotherapy. Similarly, Luan (131) also conducted related studies. Using single-cell and spatial transcriptomics techniques, they found that multiple histone lactylation-related genes, such as EP300, SIRT2 and LDHA were significantly upregulated in HCC tissues and were associated with CD8⁺ T cell infiltration, providing a new direction for lactylation-modified radiomics and liquid biopsy. Of note, lactylation modification can also activate gene transcription by remodeling the epigenetic state, promoting cell proliferation, invasion and immune escape (132). In HCC, elevated lactylation-related risk scores are associated with significantly poorer OS and disease-free survival (131). Therefore, detecting serum or tissue lactate levels, LDHA/MCT4 (SLC16A3) expression, and lactylation modification markers may become a new method for predicting the probability of HCC recurrence and progression. Despite the promising diagnostic performance of these gene signatures, the current evidence is predominantly limited by its retrospective nature and heavy reliance on public databases, such as TCGA and GEO. To transition into clinical practice, these markers must undergo rigorous testing in large-scale, multi-center prospective cohorts to account for the high inter-patient heterogeneity of HCC.

However, it is noteworthy that the interpretation of lactylation-related biomarkers in clinical studies can be significantly interfered with by acetylation, as these two PTMs share significant biochemical and regulatory similarities. Structurally, the highly similar physicochemical properties and small mass difference between lactyl and acetyl groups make accurate differentiation using conventional mass spectrometry and antibody-based detection methods difficult (14,26). Both lactylation and acetylation can occur on lysine residues and are catalyzed by overlapping enzyme systems, particularly HATs, such as p300/CBP (7,17,40). Furthermore, members of the HDAC and SIRT families can also function as deacetylases and delactylases, further increasing the overlap in their regulatory networks (24,50,133,134). This leads to signals attributed to lactylation in clinical settings potentially reflecting acetylation or mixed acylation states, reducing the specificity of biomarkers. Therefore, improving detection specificity and developing reliable strategies to differentiate these modifications are crucial steps in advancing the clinical translation of lactylation in HCC.

In HCC, lactylation can regulate metabolic reprogramming and epigenetic regulation, providing a new theoretical basis and breakthrough point for treatment. There are two types of mainstream treatment methods. The first targets lactate production, efflux and signaling pathways, reducing immunosuppression and invasiveness by inhibiting tumor metabolism. The second regulates the activity of lactylation-related enzymes to intervene in lactate-mediated transcriptional reprogramming and immune escape processes (Fig. 4).