The clinical problem the review addresses is stark. Although surgery, platinum-taxane chemotherapy, and newer maintenance strategies such as PARP inhibitors have improved outcomes in recent years, most patients are diagnosed at an advanced stage, and recurrent tumors frequently acquire resistance to the very drugs that initially controlled the disease. Classical explanations of chemoresistance have focused on enhanced DNA repair capacity, reduced intracellular drug accumulation, evasion of apoptosis, and the plasticity that allows tumor cells to shift between epithelial and other states. What makes this review distinctive is its argument that each of these classical mechanisms is increasingly inseparable from metabolic adaptation. DNA repair consumes ATP and NAD⁺; apoptosis evasion depends on antioxidant capacity and lipid signaling; cell plasticity is fueled by shifts in substrate preference. Metabolism, in other words, is not a bystander in resistance—it is an active enabler.

At the center of the metabolic argument is the balance between glycolysis and mitochondrial oxidative phosphorylation. The review details how, under therapeutic pressure, subsets of ovarian cancer cells shift toward a glycolysis-dominant state, upregulating key enzymes and transporters such as hexokinase 2 (HK2), phosphoglycerate kinase 1 (PGK1), phosphoglycerate mutase 1 (PGAM1), pyruvate kinase M2 (PKM2), lactate dehydrogenase A (LDHA), and monocarboxylate transporters, particularly MCT4. This Warburg-like configuration allows cells to generate ATP rapidly and to channel glycolytic intermediates into biosynthetic pathways that support survival, while the export of lactate acidifies the tumor microenvironment and can impair the activity and penetration of chemotherapeutic agents. The hypoxia-inducible factor HIF-1α emerges as a central transcriptional driver of this program, linking low oxygen conditions commonly found in advanced ovarian tumors to both glycolytic switch and chemotherapy failure.

Importantly, the authors emphasize that glycolysis is not the whole story. Other resistant tumors instead become dependent on mitochondria, relying on oxidative phosphorylation and the tricarboxylic acid cycle to sustain their energy demands. In these mitochondria-dependent states, glutamine metabolism becomes critical: the enzyme glutaminase (GLS) feeds glutamine-derived carbon into the TCA cycle, while glutamic pyruvate transaminase 2 (GPT2) supports anabolic and redox needs. This bidirectional plasticity—some cells abandoning respiration while others deepen their reliance on it—helps explain why single-agent metabolic inhibitors have often disappointed in the clinic. A drug that blocks glycolysis may spare a mitochondrial subpopulation, and vice versa, allowing residual cells to repopulate the tumor. The review argues that mapping which metabolic state dominates in a given patient’s tumor at a given time could be essential to choosing the right metabolic vulnerability to target.

Perhaps the most vivid section of the review concerns lipid metabolism, an area that has gained traction in ovarian cancer research partly because of the disease’s characteristic pattern of peritoneal and omental spread. Ovarian cancer cells floating in ascites or colonizing fatty omental tissue are surrounded by an environment rich in lipids, and resistant cells appear to exploit this bounty. The review describes upregulation of the fatty acid transporter CD36 and fatty acid binding protein 4 (FABP4), which enhance uptake of exogenous fatty acids, alongside increased expression of fatty acid synthase (FASN) for endogenous lipid production. Downstream, enzymes such as stearoyl-CoA desaturase 1 (SCD1), squalene epoxidase (SQLE), and HMG-CoA reductase (HMGCR)—the latter under the control of the sterol regulatory element-binding protein SREBP2—reshape the lipid composition of cellular membranes. These lipid adaptations do more than supply energy: they maintain membrane integrity against drug-induced damage, alter signaling through lipid-modified proteins, and buffer cells against the oxidative stress that platinum agents generate.

This lipid remodeling connects directly to one of the most discussed topics in modern cancer biology: ferroptosis, an iron-dependent form of cell death driven by the accumulation of lipid peroxides in cellular membranes. Chemotherapy generates reactive oxygen species, and ferroptosis represents a vulnerability that many tumors actively suppress. The review outlines the central defensive axis built around the cystine importer SLC7A11 and glutathione peroxidase 4 (GPX4), which together import cystine, generate glutathione (GSH), and enzymatically repair oxidized membrane lipids. A second, GPX4-independent shield is provided by ferroptosis suppressor protein 1 (FSP1), which reduces coenzyme Q10 at the plasma membrane using NADPH. The acyl-CoA synthetase ACSL4 also features prominently, because it determines which fatty acids are incorporated into membranes and therefore how susceptible a cell is to lipid peroxidation in the first place. Resistant ovarian cancer cells, the authors argue, frequently display a ferroptosis-resistant profile characterized by high antioxidant capacity and altered membrane lipid composition, effectively rendering them invisible to a form of cell death that chemotherapy might otherwise induce.

Adding a genuinely novel dimension, the review devotes substantial attention to cuproptosis, a recently described form of regulated cell death triggered by copper-dependent toxicity. Unlike ferroptosis, cuproptosis does not depend on lipid peroxidation; instead, excess copper promotes the aggregation of lipoylated mitochondrial enzymes, principally dihydrolipoamide S-acetyltransferase (DLAT), disrupting respiration and causing proteotoxic stress. The machinery of protein lipoylation—lipoic acid synthase (LIAS) and lipoyltransferase 1 (LIPT1)—and the copper-transporting ATPases ATP7A and ATP7B, along with the copper transporter CTR1 (SLC31A1) and the mitochondrial protein ferredoxin 1 (FDX1), all modulate sensitivity to this pathway. The review suggests that certain ovarian cancer subtypes, particularly those with high mitochondrial lipoylation, may be inherently cuproptosis-sensitive, raising the possibility of copper ionophores or copper-mobilizing strategies as a way to kill tumors that have survived conventional therapy. Intriguingly, CTR1 is also implicated in cisplatin uptake, linking copper biology directly to platinum drug transport.

A recurring theme throughout the review is dynamism. Resistant tumors are portrayed not as uniformly glycolytic or uniformly mitochondrial, but as ecosystems in which distinct metabolic and cell-death states—glycolysis-dominant, mitochondria-dependent, lipid-adapted, ferroptosis-resistant, or cuproptosis-sensitive—coexist and shift over time in response to chemotherapy, recurrence, and microenvironmental selection pressures such as hypoxia and lipid availability. A tumor sampled at diagnosis may present a very different metabolic face from the same tumor after six cycles of carboplatin and paclitaxel. This temporal evolution explains both why initial biopsies have limited predictive power and why static biomarker studies of metabolic genes have produced inconsistent results. The authors advocate for serial metabolic characterization of tumors, potentially using non-invasive imaging or liquid biopsy approaches, as a foundation for treatment selection.

The therapeutic implications are considerable, and the review is careful to frame them as biologically informed strategies rather than ready-made protocols. Combination approaches emerge as the logical consequence of the model: pairing a metabolic inhibitor matched to the dominant resistance state with chemotherapy or with agents that collapse specific antioxidant defenses. For glycolysis-dominant tumors, targeting HK2, LDHA, or lactate export might resensitize cells to platinum; for mitochondria-dependent tumors, inhibition of GLS or respiratory complexes could be preferable; for lipid-adapted tumors, blocking CD36, FASN, or SCD1 might strip away a critical survival layer. In parallel, inducing ferroptosis by inhibiting SLC7A11 or GPX4 could be combined with chemotherapy to convert a hidden stress into lethal damage, while cuproptosis-sensitizing strategies could open an entirely orthogonal cell-death route untouchable by existing resistance mechanisms. The unifying principle is that the choice of metabolic target should be dictated by the tumor’s measured metabolic state, moving treatment design away from broadly applied combinations toward rational, individualized regimens.

As a review, the article does not present new experimental data, and the authors acknowledge that much of the evidence comes from cell lines, xenografts, and retrospective biomarker analyses; translating metabolic profiling into validated clinical decision tools will require prospective studies and, ultimately, biomarker-guided clinical trials. Funding was provided by the Proof of Concept Project of Fudan University Huashan Hospital and the Yangfan Plan of the Shanghai Science and Technology Commission. Nevertheless, the synthesis arrives at a timely moment, as ferroptosis-targeting compounds and several metabolic inhibitors progress toward and through early-phase clinical testing across oncology. For a disease in which the median survival for advanced-stage patients has improved only incrementally over decades, the message that chemoresistance is a tractable metabolic problem—one with concrete molecular nodes, measurable states, and emerging pharmacological tools—offers a coherent roadmap for the next generation of ovarian cancer research. The challenge now, the authors imply, is not to find one more drug, but to learn to read the tumor’s metabolic handwriting and strike where it is most vulnerable.

The tumor microenvironment is a chemically hostile and highly dynamic ecosystem. Rapidly dividing cancer cells consume large quantities of glucose, amino acids and oxygen, while releasing lactate, carbon dioxide, extracellular nucleotides and other metabolites. These changes create regions of hypoxia, acidity and nutrient deprivation. Innate immune cells entering the tumor, including macrophages, neutrophils, dendritic cells and myeloid-derived suppressor cells, must adapt to these conditions. Rather than remaining fixed in a single functional state, they can be pushed toward inflammatory, tissue-remodeling or immunosuppressive programs that may ultimately support tumor survival.

Noh and colleagues emphasize that metabolism is not simply a source of energy for immune cells. Metabolic pathways also generate molecules that directly influence gene regulation. Acetyl-coenzyme A, for example, supplies acetyl groups for histone acetylation, a chromatin modification generally associated with more accessible DNA and active transcription. S-adenosylmethionine provides methyl groups for DNA and histone methylation, while α-ketoglutarate supports enzymes that remove methyl marks from chromatin. In this way, the nutrients and metabolites available to a cell can determine which genes are switched on or silenced.

Other metabolites can exert an opposing influence. Succinate and fumarate, which accumulate when mitochondrial metabolism is altered, can inhibit α-ketoglutarate-dependent dioxygenases, including enzymes involved in DNA and histone demethylation. Their accumulation may therefore stabilize particular epigenetic states. Lactate, long regarded mainly as a waste product of aerobic glycolysis, can also function as a signaling and regulatory molecule. It can affect transcription, chromatin-associated processes and the behavior of neighboring immune cells, helping establish an environment in which immune responses become less effective against malignant tissue.

The review describes hypoxia as another major force connecting metabolism to epigenetic remodeling. Low oxygen activates hypoxia-inducible factors, transcriptional regulators that alter glucose utilization, angiogenesis, survival and inflammatory signaling. Hypoxia can also change the activity of chromatin-modifying enzymes whose reactions depend on oxygen. As a result, oxygen limitation does not merely force immune cells to use alternative fuels; it can leave a lasting molecular imprint on their identity and function. Cells that encounter these signals repeatedly may retain altered transcriptional programs even after local conditions change.

Macrophages provide one of the clearest examples of this plasticity. In tumors, they may acquire features associated with tumor-supportive macrophages, including promotion of blood-vessel formation, extracellular-matrix remodeling and suppression of cytotoxic lymphocytes. These changes are not controlled by a single “on” or “off” switch. Instead, nutrients, oxygen, cytokines and metabolites converge on transcription factors and chromatin regulators. Fatty-acid oxidation, mitochondrial activity, glycolytic flux and amino-acid availability can all influence the epigenetic landscape that determines how macrophages respond to the tumor.

The same principle applies to other innate immune populations. Neutrophils exposed to tumor-derived signals can develop phenotypes that assist invasion, angiogenesis or immune suppression. Dendritic cells may lose efficiency in processing and presenting tumor antigens, weakening the activation of T cells. Myeloid-derived suppressor cells can expand under chronic inflammatory and metabolic stress, consuming nutrients and producing mediators that inhibit antitumor immunity. The review frames these outcomes as interconnected rather than isolated: metabolic competition and epigenetic memory can reinforce one another across several immune-cell types.

This framework has important implications for cancer therapy. Drugs that inhibit histone deacetylases, DNA methyltransferases or specific metabolic enzymes could potentially reprogram immune cells within tumors. Blocking lactate production or transport, altering glutamine metabolism, targeting hypoxia pathways or restoring mitochondrial function may also change the immune landscape. However, the authors’ discussion highlights a major challenge: the same metabolic pathway can have different effects depending on cell type, tumor region and disease stage. A treatment that strengthens antitumor immunity in one setting could impair immune function or damage healthy tissue in another.

Future therapies may therefore need to combine metabolic and epigenetic interventions with immunotherapy rather than targeting either system alone. Careful mapping of metabolites, chromatin states and immune-cell activity at single-cell and spatial resolution could help identify which populations are suppressive, which remain therapeutically recoverable and which metabolic dependencies are unique to the tumor. The review’s central message is that innate immune plasticity is not an accidental consequence of cancer metabolism. It is a dynamic, potentially reversible process in which the tumor’s chemical environment is translated into long-lasting gene-regulatory programs—offering both an explanation for immune failure and a possible route to restore immune attack.

Subject of Research: Metabolic–epigenetic regulation of innate immune cell plasticity in the tumor microenvironment

Article Title: Metabolic–epigenetic crosstalk in innate immune cell plasticity within the tumor microenvironment

Article References: Noh, J., Lee, J., You, C. et al. Metabolic–epigenetic crosstalk in innate immune cell plasticity within the tumor microenvironment. Exp Mol Med (2026). https://doi.org/10.1038/s12276-026-01802-3

Image Credits: AI Generated

DOI: 10.1038/s12276-026-01802-3

Keywords: tumor microenvironment, innate immunity, immune-cell plasticity, cancer metabolism, epigenetics, macrophages, hypoxia, lactate, chromatin remodeling, immunotherapy