Lipid droplets in the tumor microenvironment: Biogenesis, functional diversity, and therapeutic targeting.

Introduction
Metabolic reprogramming, a fundamental hallmark of malignant tumors, enables cancer cells to remodel their energy metabolism to sustain rapid proliferation and adapt to the dynamic tumor microenvironment (TME).[1] Although the Warburg effect has long been considered the cornerstone of tumor metabolism research, recent multiomics-based studies have revealed a complex metabolic landscape that extends beyond glycolysis. This comprehensive network encompasses diverse metabolic pathways, including carbohydrate, lipid, amino acid, and nucleotide metabolism.[2–5] Tumor cells dynamically exploit these metabolic networks to meet biosynthetic demands, maintain redox homeostasis, and modulate epigenetic modifications.[2–5]
Lipid droplet (LD) metabolism has recently been recognized as a critical regulator of metabolic reprogramming in cancer.[6] The biogenesis of LDs may enable tumor cells to withstand various stresses and adapt their energy metabolism in response to changes in nutrient availability. Within the TME, the accumulation of LDs in tumor-associated immune cells has been closely associated with immunosuppressive phenotypes. For instance, tumor-associated immune cells enriched in LDs, such as macrophages and fibroblasts, contribute to tumor progression. Collectively, these findings highlight LDs as promising targets for metabolic intervention in cancer therapy.[7–9]
Once regarded merely as passive lipid storage particles, LDs are now recognized as dynamic metabolic organelles that serve as central hubs for maintaining cellular metabolic homeostasis [Supplementary Table 1, http://links.lww.com/CM9/C705]. Recent research using cellular ultrastructure analysis has revealed extensive membrane connections between LDs and other organelles, including the endoplasmic reticulum (ER), mitochondria, and peroxisomes.[10] These complex interactions suggest intricate mechanisms for coordinated lipid exchange among organelles. For example, ER–LD contact sites are critical for lipid synthesis,[11] whereas mitochondria–LD interactions regulate fatty acid β-oxidation (FAO).[12] In addition to their classical roles in energy storage and transfer, LDs act as essential buffers against cytotoxic free fatty acids (FAs) and regulate the production and release of lipid-derived signaling molecules (e.g., arachidonic acid derivatives).[6] Moreover, LDs dynamically change their size during stress adaptation (e.g., hypoxia or infection) to precisely balance membrane biosynthesis with energy demands.[13] They also actively participate in immune modulation and in determining cell fate.[14] Emerging evidence indicates that LD dysregulation is a pathological feature of diverse diseases, including obesity,[15] nonalcoholic fatty liver disease,[16] neurodegenerative disorders,[17] and cancer,[6] underscoring the therapeutic potential of LDs as target for disease-specific metabolic vulnerabilities.[18]
In this review, we synthesize current knowledge regarding the biogenesis and diverse functions of LDs within the TME. We specifically highlight how LDs regulate critical tumor cell processes, including survival, metabolic adaptation, and invasive migration. Moreover, we emphasize the role of LDs in metabolic remodeling within tumor-associated immune cells to promote an immunosuppressive TME. We further examine the emerging concept of LD-mediated metabolic transfer both at the intracellular and intercellular levels. Finally, we propose future research directions for developing innovative LD-targeted strategies in TME-based cancer therapy.

Structural Organization and Dynamic Biogenesis of LDs
LDs are highly dynamic intracellular organelles. They originate from the ER and display unique architecture characterized by a neutral lipid core rich in triglycerides (TAGs) as well as cholesteryl esters (CEs). Surrounding this core is a phospholipid monolayer membrane.[19] This distinctive structure enables LDs to perform both lipid storage and mobilization. The biogenesis of LDs occurs in three main steps: neutral lipid synthesis, nucleation, and budding.[20–22] Notably, this process is coordinated by a complex regulatory network that integrates neutral lipid synthesis with membrane modeling. TAG biosynthesis requires the sequential action of glycerol-3-phosphate acyltransferases, acylglycerol-3-phosphate acyltransferases, phosphatidic acid phosphohydrolases, and diacylglycerol acyltransferases (DGATs).[22] CE production is facilitated by acyl-CoA: cholesterol acyltransferases (ACATs). Regulators such as Seipin, lipid droplet-assembly factor 1 (LDAF1), and perilipins (PLINs) are involved in the subsequent maturation of LDs.[22] Moreover, oncogenic signaling pathways and various stressors within the TME, such as hypoxia and fluctuations in nutrient levels, dysregulate these mechanisms, further resulting in abnormal LD accumulation and metabolic reprogramming in cancer cells.
DGAT enzymes, primarily located in the ER and mitochondria, play a crucial role in catalyzing the terminal and rate-limiting step of TAG synthesis by depositing TAG into the hydrophobic interleaflet space of the ER membrane.[23] Two isoforms, DGAT1 and DGAT2, are involved in this process and share overlapping catalytic function. Overexpression of DGAT1 and DGAT2 enhances TAG synthesis and increases LD accumulation. Preclinical studies have shown that the inhibition of DGAT1 markedly decreases plasma TAG levels and provides protection against diet-induced obesity, highlighting its therapeutic potential in metabolic disorders.[24–26] In various tumor types, such as melanoma and glioblastoma, overexpression of DGAT1 promotes LD accumulation, enabling tumor cells to sequester excessive FAs and mitigate lipotoxicity under metabolic stress.[27,28] DGAT1 inhibition enhances reactive oxygen species (ROS) generation, promoting tumor cell apoptosis, and thereby overcoming chemoresistance and immune evasion.[28,29] DGAT2 inhibition not only blocks TAG synthesis but also suppresses sterol regulatory element-binding protein-1 (SREBP-1), a key transcriptional regulator of hepatic lipogenesis.[30] Recent studies have identified a DGAT-independent pathway mediated by the DGAT1/2-independent enzyme synthesizing storage lipids, which redirects membrane phospholipids into TAG pools to maintain mitochondrial function during nutrient deprivation in thioredoxin-related transmembrane protein 1-deficient conditions.[31] However, the substrate specificity and functional interactions of these pathways under different metabolic states remain to be fully elucidated.
The ACAT family comprises two isoforms, ACAT1 and ACAT2, which exhibit distinct subcellular distribution patterns. ACAT1 primarily resides in the ER, whereas ACAT2 localizes to both the ER and Golgi apparatus. These enzymes play a crucial role in maintaining cellular cholesterol homeostasis by esterifying extra cholesterol to form CEs, which are stored in LDs and support cell survival under stress conditions.[32] In atherosclerosis, ACAT-mediated accumulation of CE in macrophages leads to the formation of a foam cell, a hallmark of early atherosclerotic lesions.[22,33] LD accumulation dependent on ACAT1 is also associated with cancer progression in tumors, such as ovarian, colon, and prostate cancers.[34–36] Inhibition of ACAT1 in ovarian cancer cells induces apoptosis and suppresses tumor cell invasion and migration.[35] Moreover, when combined with cisplatin treatment, ACAT1 inhibition synergistically enhances the chemosensitivity of tumor cells.[35]
The PLIN family (PLIN 1–5) comprises LD-associated proteins that play a critical role in regulating lipid storage and metabolism. Although they exhibit tissue-specific expression patterns, all PLINs modulate lipase activity toward LDs, thereby maintaining the delicate balance between lipogenesis and lipolysis.[37] PLIN1 serves as a marker of adipocyte differentiation and functions as an anti-inflammatory modulator in adipose tissue.[38] Recent CRISPR-Cas9 screening studies have provided mechanistic insights into PLIN2-mediated LD regulation.[39] Under TME stress, PLIN2 and PLIN3 are frequently upregulated in various cancers and correlate with poor clinical outcomes.[40] For instance, oral squamous cell carcinomas with elevated PLIN3 expression demonstrate enhanced proliferative and metastatic potential,[41] whereas PLIN3 knockdown increases radiosensitivity in prostate cancer models.[42] Notably, not all PLIN family members exhibit protumorigenic effects, as evidenced by the ability of PLIN1 to inhibit adipose triglyceride lipase (ATGL) activation and suppress tumor growth in glioblastoma through reduced lipolysis.[40]
LD nucleation is typically initiated once TAG accumulation reaches a critical threshold between the two leaflets of the ER membrane, leading to phase separation and the formation of lens-like structures through demixing processes.[22,34] This phenomenon is primarily driven by elevated TAG levels and is regulated by a complex protein network.[43–45] Seipin, a highly conserved protein localized at LD formation sites, forms oligomeric ring structures that surround the developing lipid lenses. These Seipin complexes function as molecular scaffolds that prevent aberrant LD fusion and maintain structural integrity by stabilizing the phospholipid monolayer at the LD surface.[46–48]
LD budding constitutes a highly coordinated membrane remodeling originating from the ER. This process is initiated by the accumulation of neutral lipid lenses within the ER bilayer, during which the cytoplasmic leaflet of the ER expands to encapsulate the lipid core, thereby forming a phospholipid monolayer consisting of specific LD-associated proteins. Key regulators, including LDAF1, multiple C2 and transmembrane domain-containing protein 1/2, and nuclear envelope morphology protein 1 (Nem1) localize to budding sites and enhance the efficiency of LD budding.[49–51] The peroxisomal protein 30 (Pex30) promotes LD budding by regulating the lipid composition of the ER inner leaflet, whereas Seipin facilitates budding by modulating conformational changes.[52] Following budding, LDs undergo size expansion and functional maturation through LD–LD fusion as well as LD–ER material exchange. The cell death-inducing DNA fragmentation factor alpha-like effector family proteins mediate LD–LD fusion by forming channels for neutral lipid transfer between LDs, consistent with an Ostwald ripening mechanism, in which lipids move from smaller droplets to larger ones.[18,53] Moreover, rapid LD–LD fusion occurs directly under conditions of phosphatidylcholine deficiency or increased phosphatidic acid accumulation.[54,55] LD–ER material exchange is facilitated by membrane bridges activated by the ADP-ribosylation factor-coat protein complex I complex, which mediates the transfer of TAG-synthesizing enzymes and substrates from the ER to LDs[56,57] Although the relationship between neutral lipid nucleation and budding in tumor cells remains unclear, emerging evidence suggests that Seipin-mediated regulation of ER stress may modulate cancer cell metabolism. Notably, Seipin mutation in human colorectal carcinoma cells triggers induced ER stress and apoptosis, thereby positioning Seipin as a potential metabolic checkpoint for tumor therapy.[58]

Multifunctional Roles of LDs in Tumor Cell Biology
Elevated LD accumulation is a metabolic characteristic of diverse tumors, driven by enhanced lipogenesis and FA uptake.[6,59] These dynamic organelles facilitate tumor progression through two primary mechanisms. First, they provide energy through FA mobilization. Second, they promote stress adaptation by buffering lipids.

LDs modulate membrane biosynthesis and energy metabolism in tumor cells
The dynamic regulation of LD biogenesis and membrane synthesis is closely linked to the nutrient availability in tumor cells. Under nutrition-rich conditions, both de novo lipogenesis and lipid uptake pathways are upregulated in a coordinated manner within tumor cells, simultaneously promoting membrane biogenesis and LD formation. In contrast, when extracellular lipids are scarce, tumor cells mobilize FAs from LDs via lipolytic processes to sustain membrane biosynthesis, thereby maintaining cell survival and proliferation.[60,61] AMP-activated protein kinase (AMPK) functions as a critical metabolic sensor under nutrient-deprived conditions, orchestrating a metabolic shift from anabolic biosynthesis to catabolic breakdown.[62,63] Activated AMPK stimulates lipolysis and directs released FAs toward mitochondrial β-oxidation, generating essential ATP and nicotinamide adenine dinucleotide phosphate (NADPH) to maintain redox balance and energy homeostasis.[13,64]
The fluctuating nutrient availability within the TME drives tumor cells to adopt adaptive metabolic strategies, including a “feast-and-famine” approach characterized by anticipatory LD accumulation during nutrient-replete periods.[65,66] In pancreatic ductal adenocarcinoma, this adaptation is mediated by oncogenic KRAS-dependent suppression of hormone-sensitive lipase (HSL), which promotes LD storage. Disruption of this KRAS–HSL axis leads to lipid depletion and markedly impairs metastatic potential in vivo.[67] Metabolic profiling studies revealed that invasive tumor subpopulations preferentially use LD-derived lipids for oxidative phosphorylation.[67] Similarly, in RAS-mutant breast cancer cells, even minimal supplementation with unsaturated FAs induces substantial LD accumulation, which subsequently enhances cell survival during nutrient deprivations.[68,69]

LDs buffer tumor cells against TME stress
The interaction between tumor cells and other components of the TME creates a dynamic stress landscape in which LDs act as crucial metabolic buffers. Tumor cells profoundly reshape the TME through their high metabolic demands and proliferative capacity, while simultaneously adapting to the resulting stresses through LD-mediated mechanisms.[70,71] First, LDs function as regulators of lipid homeostasis, mitigating lipotoxicity by sequestering free lipids and controlling their regulated release. This storage-and-release mechanism prevents cytotoxic lipid accumulation while maintaining cellular equilibrium.[72] Second, LDs critically regulate the balance between saturated and unsaturated FAs to alleviate ER stress in tumor cells. Elevated levels of saturated FAs (e.g., palmitate) trigger ER stress and apoptosis,[73–75] whereas hypoxia in the TME inactivates oxygen-dependent stearoyl-CoA desaturase 1, impairing unsaturated FA production.[76,77] Concurrent mechanistic/mammalian target of rapamycin (mTOR) activation increases membrane expansion demands, creating a lipid–protein imbalance that typically induces ER stress.[78] Tumor cells overcome this by enhancing the uptake and LD storage of exogenous unsaturated FAs, thereby maintaining membrane homeostasis.[78–81] In clear cell renal cell carcinoma (ccRCC), the hypoxia-inducible factor (HIF2α)–PLIN2 axis drives characteristic LD accumulation, with PLIN2 both mediating and being required for the oncogenic effects of HIF2α.[81] Notably, PLIN2 overexpression alone can restore lipid storage and ER homeostasis even in HIF2α-deficient cells.[60,81] Nutrient-replete ccRCC cells preferentially incorporate serum-derived unsaturated FAs into LDs,[76,81] which are subsequently mobilized during hypoxia or nutrient stress to maintain membrane integrity.[82] This adaptive mechanism underscores the dual functions of LDs as reservoirs of FAs and regulators of redox metabolism. Third, LDs protect against oxidative stress by sequestering peroxidation-sensitive polyunsaturated FAs (PUFAs), which are highly susceptible to ROS-mediated damage within the TME.[83,84] The TME, enriched in ROS as a result of both internal metabolic dysregulation and external stressors, preferentially oxidizes membrane-incorporated PUFA. This process leads to the accumulation of lethal lipid peroxide that induces ferroptosis, an iron-dependent form of cell death.[85] To counter this threat, tumor cells use LDs as protective reservoirs, where phospholipases relocate PUFAs from the vulnerable membranes into the more oxidation-resistant LD compartment.[86] This sequestration mechanism is amplified during cell cycle arrest, when cells trigger DGAT-dependent LD formation to partition accumulating PUFAs into TAGs, thereby suppressing ferroptosis.[87] Beyond serving as protective storage places, LDs also function as cellular “oxidation sinks”. They accumulate not only peroxidized phospholipids but also oxidatively damaged proteins, thereby reducing cytotoxic damage.[86,88] This multifaceted antioxidant system substantially enhances tumor cell resilience against oxidative challenges, including the iron-rich conditions encountered during hematogenous metastasis and oxidative burst induced by radiotherapy.[89]
Notably, LDs exhibit a context-dependent dual role in ferroptosis, functioning either as protective buffers or as facilitators of cell death. Under cystine deprivation and exposure to dehydroascorbic acid, LDs enhance sensitivity to ferroptosis, whereas inhibition of DGAT1 suppresses cell death by reducing LD accumulation.[90] A similar effect has been observed in ccRCC within a hypoxia-induced LD-associated protein-mediated ferroptosis model.[91] Importantly, ccRCC cells depend on the uptake of exogenous FAs rather than de novo lipogenesis for LD biogenesis, a process regulated by acyl-CoA synthetase 3 (ACSL3).[92] Suppression of ACSL3 not only disrupts LD formation and induces cytotoxicity, but also reduces cellular susceptibility to ferroptosis, thereby modulating ferroptotic susceptibility in a manner dependent on the composition of exogenous FAs.[92] In other contexts, LDs exert potent protective effects against ferroptosis, particularly in therapy-resistant models. This phenomenon is exemplified in radioresistant tumors, where surviving cancer cells exhibit enrichment of oleic acid (OA) and palmitoleic acid within LDs. OA protects cells from ferroptosis in an ACSL3-dependent manner by limiting the accumulation of peroxidation-prone polyunsaturated phospholipids and reducing lipid ROS.[93] This functional dichotomy likely arises from three key factors: compositional variations in LD and exogenous lipid species (e.g., pro-ferroptotic phospholipids containing polyunsaturated fatty acyl chains vs. protective monounsaturated FA such as OA);[93,94] differential interactions with redox-active organelles such as mitochondria and peroxisomes;[10,95,96] and modulation of ferroptotic sensitivity by the TME through lipid metabolic crosstalk involving immune and stromal cells as well as lipid-metabolizing enzymes.[97] The precise molecular mechanisms underlying this switch between protective and destructive roles remain to be fully elucidated and represent an important focus for future investigation in cancer metabolism research.

Roles of LD-related Metabolism in Regulating Immune Cell Function within the TME
The TME, marked by hypoxia and nutrient deprivation, exhibits substantial LD accumulation that extends beyond malignant cells to include various immune cell populations [Table 1]. This widespread LD deposition has emerged as a defining metabolic signature associated with tumor cell–mediated immunosuppression.[6,97]

Macrophages
Macrophages, as central components of the innate immune system, maintain tissue homeostasis through potent phagocytic activity while orchestrating host defense and tissue repair mechanisms.[98–100] They exhibit remarkable phenotypic plasticity, classically categorized into pro-inflammatory and anti-inflammatory polarization states based on their responses to different stimuli in vitro.[101] Macrophages acquire lipids through multiple pathways, including endocytosis, scavenger receptor–mediated uptake, and lipid receptor–dependent mechanisms.[14] These distinct activation states display characteristic metabolic profiles: pro-inflammatory macrophages rapidly generate ATP via aerobic glycolysis to sustain phagocytic activity while using a truncated tricarboxylic acid (TCA) cycle to provide precursors for lipid synthesis and inflammatory mediator production,[102–104] whereas anti-inflammatory macrophages depend on a complete TCA cycle with enhanced FAO and oxidative phosphorylation.[102–104]
Lipid metabolism plays a pivotal role in shaping macrophage phenotypes and functions [Figure 1]. Tumor-associated macrophages (TAMs) consistently exhibit marked LD accumulation across multiple cancer types, correlating with an immunosuppressive phenotype.[9,105–108] Furthermore, LD-loaded TAMs directly fuel tumor progression by secreting cytokines such as C-C motif chemokine ligand 6 and vascular endothelial growth factor A, thereby enhancing migration and angiogenesis.[9,108] In gastric cancer, TAMs display enhanced extracellular lipid uptake that activates the γ isoform of phosphoinositide 3-kinase signaling, driving their protumoral polarization.[105]
In vitro experiments have revealed that unsaturated FAs induce CD206+ major histocompatibility complex (MHC) IIlow immunosuppressive TAMs, suggesting LD accumulation as a potential therapeutic target for modulating the macrophage immunosuppressive phenotype.[106] This is supported by the finding that inhibition of LD accumulation with DGAT inhibitors retards tumor growth in colon cancer models.[106] Similar mechanisms are observed in breast cancer, where caspase-1–mediated cleavage of peroxisome proliferator-activated receptor gamma in TAMs attenuates FAO and establishes an LD-rich, tumor-promoting state.[107] These findings indicate that inhibiting LD accumulation restores FAO, thereby suppressing tumor growth. Clinical studies in patients with glioblastoma also demonstrate that LD-rich macrophages exhibit protumorigenic transcriptional signatures and correlate with poorer prognosis.[9]
The mechanisms underlying LD accumulation in macrophage involve multiple pathways: upregulated expression of lipid uptake receptors, such as the macrophage receptor with collagenous structure (MARCO), which are overactivated by tumor-derived cytokines;[108–110] uptake of tumor-derived or adipocyte-derived extracellular vesicles (EVs) and various lipid components in TAMs (FAs, cholesterol, low-density lipoprotein, oxidized low-density lipoprotein [ox-LDL]);[9,111] and phagocytosis of cholesterol-rich myelin debris.[112] Notably, myelin-derived lipids can be transferred to glioma cells via liver X receptor-ATP-binding cassette subfamily A member 1(LXR-ABCA1) dependent mechanisms to promote malignant progression.[112] In addition to exogenous lipid scavenging, TAMs exhibit enhanced de novo lipogenesis, and inhibition of key biosynthetic enzymes has been shown to reverse their tumor-promoting effects.[113]
Collectively, LD accumulation in TAMs, driven by both exogenous uptake and endogenous synthesis, represents a critical metabolic adaptation that fosters an immunosuppressive TME. Therapeutic strategies targeting LD formation–through DGAT inhibition, scavenger receptor blockade, or lipogenesis inhibition–can shift TAMs toward pro-inflammatory, antitumor phenotypes.[106,107,113,114] Although TAM heterogeneity in the TME exceeds the simplicity of in vitro models, disrupting LD metabolism remains a promising approach for cancer immunotherapy. Further studies are warranted to delineate tissue-specific lipid sourcing and to define the temporal dynamics of TAM metabolic reprogramming during tumor progression.

T cells
T cells in the TME exhibit functionally opposing roles. CD8+ T cells act as central effectors of antitumor immune responses through T cell receptor–mediated recognition of tumor antigen–MHC I complexes.[115] Upon activation, they differentiate into cytotoxic T lymphocytes, which eliminate tumor cells via perforin/granzyme-mediated apoptosis and IFN-γ secretion.[115] In contrast, regulatory T cells (Tregs) suppress effector T cell function through inhibitory receptors (cytotoxic T lymphocyte-associated antigen-4 [CTLA-4] and PD-1) and immunosuppressive cytokines (IL-10 and TGF-β), with their infiltration correlating with poor clinical outcomes.[116] Notably, lipid accumulation in the TME differentially affects these subsets – impairing CD8+ T cell function while enhancing Treg-mediated immunosuppression [Figure 2].
CD8+ tumor-infiltrating lymphocytes (TILs) encounter metabolic stresses such as hypoxia, glucose deprivation, and lipid overload, all of which contribute to their functional exhaustion.[117,118] To maintain their cytotoxic capacity under adverse conditions, CD8+ TILs undergo metabolic reprogramming. Specifically, they shift their primary energy generation pathway from glycolysis to FAO.[119] Upregulation of carnitine palmitoyltransferase1 (CPT1), the rate-limiting enzyme for mitochondrial FA import, was observed in CD8+ TILs from both colorectal cancer and melanoma models.[119–121] Notably, PD-1 signaling promotes CPT1A expression and FAO, suggesting a metabolic synergy with checkpoint blockade.[122] Nevertheless, lipid over-accumulation impairs the function of CD8+ TILs through several distinct mechanisms. First, excessive accumulation of long-chain FAs induces mitochondrial dysfunction, and paradoxically, reduces FAO capacity.[123,124] Second, activation of the p38 mitogen-activated protein kinase signaling pathway suppresses the production of key antitumor mediators such as IFN-γ.[125] Finally, ox-LDL-mediated lipid peroxidation triggers the induction of ferroptosis.[123,125] These pathological effects are mediated by the abnormal overexpression of lipid-scavenging receptors such as CD36 on exhausted CD8+ TILs, thus establishing a clear dose-response relationship. Physiological lipid levels support proper metabolic function, whereas supraphysiological accumulation leads to T cell dysfunction.
Conversely, Tregs exploit lipid metabolism to exert their immunosuppressive effects by upregulating three essential metabolic processes: lipid uptake, de novo lipid synthesis, and FAO.[126] Multiple studies have shown that the lipid transporter protein CD36 is highly overexpressed on Tregs in various mouse cancer models, including colorectal carcinoma, melanoma, and glioblastoma, thereby facilitating the uptake of extracellular lipids.[127,128] This CD36-mediated lipid accumulation activates the peroxisome proliferator-activated receptor β (PPARβ) signal pathway, further promoting mitochondrial metabolism. This enables Tregs to adapt to the lactate-rich TME while simultaneously strengthening their immunosuppressive function.[127] Moreover, SREBPs drive increased biosynthesis of both FAs and cholesterol, providing further reinforcement of the Treg phenotype.[129] At the energy metabolism level, Tregs preferentially use free FAs as primary substrates for oxidative phosphorylation. The AKT–mTORC1 signaling axis functions as a master regulator that concurrently enhances both FA and cholesterol metabolism pathways to promote Treg proliferation and upregulates the expression of key immune checkpoint molecules such as CTLA-4 and inducible T cell co-stimulator.[130] Paradoxically, inhibition of fatty acid binding protein 5 (FABP5) induces mitochondrial abnormalities, including cristae disorganization, as well as impairments in lipid metabolism and oxidative phosphorylation that trigger the release of mitochondrial DNA.[131] Mitochondrial distress activates the cyclic GMP-AMP synthase-stimulator of interferon genes-dependent type I IFN signaling cascade, ultimately increasing IL-10 production and enhancing the suppressive activity of Tregs.[131] Although FABP5 inhibition induces compensatory increases in glycolytic flux, prolonged treatment ultimately results in mitochondrial failure and cell death,[131] highlighting the remarkable metabolic plasticity that enables Tregs to maintain their functionality under diverse metabolic conditions within the TME.
Although no studies have yet defined the specific role of LDs in T cell biology, current knowledge regarding LD functions and the observed effects of lipid accumulation on T cells suggests that LDs significantly regulate T cell metabolism and function. LDs act as critical buffers that maintain intracellular lipid concentrations at optimal levels, preventing functional exhaustion of CD8+ T cells owing to lipid overload while preserving the metabolic flexibility required for antitumor activity.[120] In contrast, in Tregs, LDs likely function as metabolic hubs that coordinate the intricate interaction between lipid uptake, synthesis, and oxidative pathways, collectively enhancing their immunosuppressive capacity. These findings highlight the complex, subset-specific ways in which lipid metabolism influences the T cell phenotypes within the TME, suggesting that LD biology represents a potentially critical yet not fully explored aspect of tumor immunometabolism, promising therapeutic implications.

Dendritic cells (DCs)
DCs, as professional antigen-presenting cells, play a pivotal role in initiating antitumor immune responses by capturing, processing, and presenting tumor antigens to T cells in lymph nodes. However, within the TME, abnormal lipid accumulation in DCs results in functional impairment, characterized by diminished antigen-presenting capacity and reduced production of pro-inflammatory cytokines, thereby facilitating tumor immune escape [Figure 3].[132] Pathological lipid accumulation in tumor-associated DCs arises from dysregulation of both lipid uptake and synthesis pathways. In the TME, autophagy deficiency (particularly autophagy-related protein 5 deletion) and tumor-derived factors synergistically upregulate the scavenger receptors CD36 and macrophage scavenger receptor 1, significantly enhancing lipid uptake by DCs.[133,134]. Concurrently, constitutive activation of FA synthase drives excessive de novo lipogenesis in DCs.[135] These metabolic alterations impair DC function through multiple mechanisms, including defective antigen cross-presentation in which electrophilic oxidatively truncated lipids within LDs covalently modify heat shock protein 70, trapping peptide–MHC class I complexes in late endosomes/lysosomes and preventing their surface expression for T cell activation.[7] In addition, X-box binding protein 1-mediated lipid peroxide accumulation induces ER stress, promotes TAG biosynthesis and LD formation, and further compromises antigen cross-presentation.[136]
In addition to lipid uptake and synthesis, FAO is enhanced in tumor-associated DCs. Tumor cells can drive FAO in DCs by upregulating the expression of CPT1A through the paracrine Wnt5a–β-catenin–PPAR-γ signaling pathway, which suppresses the expression of pro-inflammatory cytokines IL-6 and IL-12.[137] This metabolic reprogramming ultimately promotes the aggregation of Tregs in the TME. Blocking lipid accumulation and FAO in tumor-associated DCs may restore their antitumor efficacy. However, research has shown that immunity-related GTPase family M protein 3 (IRGM3), an ER-resident immune-related GTPase expressed in DCs, localizes to the ER membrane and associates with LD membranes through binding to the LD-associated protein adipose differentiation-related protein (ADRP).[138] Genetic knockout of either IRGM3 or ADRP leads to defective LD formation in DCs, which severely impairs MHC class I antigen presentation.[138] Similarly, Toll-like receptor activation–induced de novo FA synthesis has been identified as essential for the rapid activation of DCs.[139]
Overall, LDs and lipid metabolism play dual roles in tumor-associated DCs. This functional duality appears to depend on the types and ratios of lipids stored within LDs, as well as the timing and extent of FAO activation. Certain lipid accumulation enhances the immunosuppressive functions of DCs, whereas other lipid configurations promote their antigen-presenting capacity. Dysregulated lipid metabolism in DCs profoundly influences their phenotype and, in turn, affects downstream T cell activation and the overall antitumor immune response.

Myeloid-derived suppressor cells (MDSCs)
MDSCs represent another immunosuppressive population within the TME. They originate from myeloid progenitor cells and are generally categorized into two subpopulations: monocytic MDSCs and polymorphonuclear MDSCs (PMN-MDSCs).[140] MDSCs suppress antitumor immunity by inhibiting the function of T and natural killer (NK) cells through multiple mechanisms, including the production of immunosuppressive cytokines that regulate tumor-associated inflammation.[140] Within the TME, MDSCs undergo substantial metabolic reprogramming, characterized by enhanced lipid uptake and FAO, which fuels their immunosuppressive effects [Figure 3].[141,142] This process is driven by tumor-derived cytokines, such as granulocyte colony-stimulating factor and granulocyte-macrophage colony-stimulating factor, which activate the signal transducer and activator of transcription 3/5 signaling pathways, leading to upregulation of lipid transport receptors and subsequent lipid uptake from the TME.[143] Notably, tumor-activated PMN-MDSCs exhibit pronounced overexpression of CD36, FA transport proteins, and very low-density lipoprotein receptors. CD36 inhibition attenuates MDSC-mediated immunosuppression.[143] Furthermore, in multiple mouse models, tumor-infiltrating MDSCs consistently display elevated FAO and upregulated CPT1 expression. Pharmacological inhibition of FAO in MDSCs effectively disrupts their immunosuppressive functions, diminishing the production of inhibitory cytokines.[144] Notably, these findings have clinical relevance, as similar increases in FA uptake and expression of FAO-related enzyme have been observed in MDSCs isolated from both tumor tissues and the peripheral blood of patients with cancer.[144]
The composition of accumulated lipids has a significant impact on the function of MDSCs. For instance, unsaturated FAs, such as linoleic acid, enhance the immunosuppressive activity of MDSCs compared with saturated FAs such as palmitic acid. Treatment of mesenchymal stem cells 2 (MSC-2) cells with sodium oleate induces a regulatory phenotype characterized by an increase in LD formation and NO production, thereby enhancing the suppressive capacity of MSC-2 cells.[145] Notably, this effect can be reversed by inhibiting LD formation, but remains unaffected by blockade of de novo synthesis of free FAs. These findings highlight the critical role of exogenous FAs stored in LDs in regulating MDSC function.
Collectively, these findings demonstrate that the uptake of exogenous lipid is essential for activating the immunosuppressive functions of MDSCs. The observation that elevated levels of FAO are present in peripheral blood MDSCs suggests that lipid metabolic reprogramming may begin even before these cells enter the TME. Within MDSCs, LDs serve several critical roles: they act as regulatory hubs that link increased lipid uptake to enhanced FAO, and they function as preloaded stores of exogenous lipids. This unique functional plasticity enables MDSCs to rapidly activate FAO and exert potent immunosuppressive effects upon infiltrating the TME, thereby supporting their survival and activity during tumor progression.

NK cells
NK cells act as crucial effectors of innate immunity. They play significant roles in tumor immune surveillance and clearance through their cytotoxic activity and cytokine production. These immune cells identify tumor cells by a sophisticated recognition system. This system involves two components: (1) activating receptors such as natural killer group 2 member D (NKG2D), which bind to stress-induced ligands including MHC class I chain–related protein A/B and UL16 binding protein on malignant cells; and (2) inhibitory receptors such as killer cell immunoglobulin-like receptors and NKG2A, which detect the absence of MHC class I molecules, thereby enabling selective tumor target.[146] In addition to their direct cytotoxic effects, NK cells also contribute to antitumor immunity by secreting IFN-γ, which promotes the maturation of antigen-presenting cells and also stimulates the anti-tumoral responses of TILs. Additionally, they mediate antibody-dependent cellular cytotoxicity via CD16, thereby enhancing the efficacy of targeted therapeutic strategies.[146]
In the TME, metabolic abnormalities severely impair the antitumor functions of NK cells through multiple mechanisms. Nutrient deprivation and lactate accumulation directly induce NK cell apoptosis.[147] In addition, lipid overload results in intracellular lipid accumulation and functional exhaustion [Figure 3].[148] Studies have shown that the upregulated expression of CD36, along with lipid accumulation in splenic NK cells from tumor-bearing mice, directly impairs their cytotoxic capacity.[134] A similar dysfunction was observed in obesity models, where metabolic alterations diminish NK cell activity and decrease IFN-γ production.[149] These findings highlight the mechanisms by which metabolic disturbances impair the antitumor activity of NK cells across different pathological conditions.
The obesity-associated TME exhibits distinct immune regulatory characteristics. In esophageal adenocarcinoma, omental and hepatic factors from patients with obesity induce NK cell death while polarizing surviving NK cells toward an anti-inflammatory phenotype, characterized by increased IL-10–producing NK cells and decreased TNF-α+ and NKP46+ subsets.[150] In KRAS-mutant pancreatic cancer models, peripancreatic adipose tissue in obese mice releases IL-6, leading to NK cell depletion during preneoplasia and establishing an immune-evasive niche.[151] Lipid accumulation in NK cells may involve exosomal transfer, as demonstrated in breast cancer models, in which lung interstitial cells shuttle neutral lipids to NK cells via EVs, causing dysfunction that promotes metastasis.[152] Mechanistically, lipid-loaded NK cells exhibit enhanced PPARα/δ signaling, reducing cytotoxic granule production and IFN-γ secretion.[149] Inhibiting PPARα/δ signaling or blocking lipid transport can reverse the metabolic paralysis of NK cells and restore their antitumor activity.[149]
Notably, hepatocellular carcinoma models have demonstrated a paradoxical scenario in which cholesterol accumulation enhances NK cell antitumor activity by promoting membrane lipid raft formation.[153] This observation, in contrast to most reports of lipid-induced NK cell inhibition, suggests that different lipid species (e.g., free FAs versus cholesterol) may exert opposing effects on NK cell function. Moreover, LD heterogeneity may strongly influence NK cell phenotypes by regulating intracellular lipid composition. Collectively, these findings highlight that both the species and compartmentalization of accumulated lipids determine functional outcomes.
In summary, NK cell function is differentially modulated by diverse intracellular species and tissue-specific microenvironmental changes, acting through distinct regulatory processes. Further investigation is required to delineate the precise pathways involved and to evaluate their potential clinical applications.

Intercellular Transfer of Lipids and LDs in the TME
In addition to de novo synthesis and uptake of lipids and LDs, recent studies have highlighted the importance of direct intercellular trafficking of lipids and LDs as a critical mechanism of metabolic reprogramming and functional modulation within the TME. This transfer establishes a form of metabolic symbiosis that enables crosstalk between tumor and neighboring stromal cells. One key mechanism involves tumor-derived EVs, such as exosomes and microvesicles, which carry lipids, lipid-metabolizing enzymes, and even preformed LD precursors. For example, in breast cancer models, lung interstitial cells transfer neutral lipids to NK cells through EVs, leading to LD accumulation and functional paralysis of NK cells, thereby facilitating the formation of a metastatic niche.[152] Similarly, tumor cells release EVs that deliver lipids to TAMs and MDSCs, thereby enhancing their protumoral and immunosuppressive phenotypes.[9] In addition, direct lipid exchange through membrane contact sites or scavenger receptors represents another important pathway. In glioblastoma, TAMs engulf myelin-rich debris and subsequently transfer processed cholesterol esters to tumor cells in an LXR-ABCA1–dependent manner, thereby fueling malignant progression.[112] Collectively, these findings indicate that LD-loaded immune cells function as metabolic intermediaries that take up, process, and transfer lipids to tumor cells, ultimately supporting tumor progression.
Intercellular lipid transfer has considerable functional implications in tumor biology. It enables tumor cells to export excess lipids, thereby avoiding lipotoxicity and associated cellular damage. Concurrently, this process accelerates the establishment of an immunosuppressive TME by transferring LDs to immune cells, thereby altering their function. Furthermore, immune cells provide essential fuels and remodel the TME to support tumor cell survival and growth. Therefore, targeting intercellular lipid transfer—whether through inhibition of EV formation, blockade of specific lipid transporters, or disruption of intercellular signaling pathways that regulate this process—represents a promising therapeutic strategy for disrupting the metabolic interdependence among cells within the TME.[154,155]

LD Metabolism as a Promising Therapeutic Target
LDs accumulate ubiquitously in both tumor and immune cells within the TME, where they promote malignant progression, support cell survival, and contribute to metastatic dissemination. They also foster an immunosuppressive TME, generating interest in the development of therapeutic interventions that target LDs in the TME. Current strategies can be broadly categorized into three principal approaches, based on our understanding of LD biogenesis and its diverse roles in cancer biology.

Targeting lipid uptake receptors
LD accumulation in both malignant and immune cells promotes tumor progression and immunosuppression, garnering interest in targeting receptors that mediate exogenous lipid uptake as a therapeutic strategy. Among these receptors, CD36 has emerged as a particularly promising candidate, given its dual role in regulating FA uptake and driving metabolic reprogramming across multiple cell types within the TME. This scavenger receptor contributes substantially to tumor initiation, progression, and metastatic dissemination.[156–158] In prostate cancer models, genetic ablation of CD36 in Pten−/− mice markedly reduces FA uptake, decreases levels of oncogenic lipid signaling molecules, and attenuates tumor progression.[159] Within the immune compartment, CD36 deletion in Tregs confers dual benefits by reducing intratumoral Treg accumulation and enhancing TIL antitumor activity while preserving overall immune homeostasis. Similarly, in CD8+ T cells, CD36-mediated lipid uptake impairs cytotoxic cytokine production and antitumor function, effects that can be reversed via CD36 blockade.[127] In addition to CD36, MARCO represents another viable target. In prostate cancers, tumor-derived IL-1β upregulates MARCO expression on macrophages, and therapeutic targeting of MARCO-mediated lipid uptake yields a triad of benefits: inhibition of tumor growth, reduction of invasive potential, and enhanced chemotherapy efficacy in preclinical models.[108] Collectively, these findings highlight receptor-mediated lipid uptake as a therapeutically actionable vulnerability that influences both tumor cell metabolism and immune cell function within the TME. The varied effects observed across cell types—including tumor cells, Tregs, CD8+ T cells, and macrophages—suggest cell type–specific approaches may be required for optimal therapeutic outcomes.

Targeting enzymes in LD biogenesis
Disrupting the enzymatic machinery that regulates LD formation has emerged as a promising therapeutic strategy, particularly because LD accumulation enables tumor cells to bypass dependence on de novo FA synthesis pathways. Among these targets, DGAT enzymes–key catalysts of TAG synthesis and LD biogenesis–are consistently upregulated across multiple malignancies and therefore represent attractive therapeutic candidates. In glioblastoma, DGAT1 overexpression facilitates FA storage in LDs, protecting tumor cells from oxidative damage. Pharmacological inhibition of DGAT1 disrupts this protective mechanism, redirecting excess FAs into mitochondrial β-oxidation pathways that generate lethal ROS levels, ultimately inducing mitochondrial dysfunction and apoptosis.[28] In prostate cancer models, loss of the ephrin B2 receptor function triggers a metabolic shift characterized by DGAT1 upregulation and ATGL suppression, leading to pathological LD accumulation. In this context, DGAT1 inhibition restores lipid homeostasis and impairs tumor progression.[160] Studies on ccRCC have identified the jumonji domain-containing protein 6–DGAT1 axis as a critical regulator of LD formation, with DGAT1 inhibitors effectively reducing LD content in vitro and suppressing tumor growth in vivo.[161] The therapeutic potential of DGAT inhibition extends beyond tumor cells to immune modulation, as demonstrated in colon cancer models where DGAT inhibitors reduce LD accumulation in TAMs, promoting their differentiation into anti-tumoral phenotypes and mediating tumor suppression.[106] Mesothelioma cells in acidic microenvironments secrete TGF-β2 to promote LD formation in DCs, impairing their immunostimulatory capacity—an effect reversible through DGAT inhibition.[162] Collectively, these findings position DGAT inhibition as a multifaceted therapeutic approach capable of simultaneously targeting tumor cell survival mechanisms, metabolic adaptations to oxidative stress, and immune cell dysfunction within the TME. The consistent antitumor effects observed across diverse cancer types and cellular compartments underscore the central role of LD biogenesis in maintaining the pathological state of both malignant and stromal cells.

Targeting LD functional proteins and organelle interactions
Disruption of interactions between LD and mitochondria, along with their associated functional proteins, represents a promising therapeutic approach as these interactions play a critical role in accelerating FAO in tumor cells. Phosphofructokinase, liver type, phosphorylates PLIN2 to strengthen its binding with CPT1A. This interaction not only reinforces the LD-mitochondria contact sites but also increases the efficiency of FAO, thereby driving tumor cell proliferation.[95] In this context, CPT1A inhibition effectively reduces mitochondrial FAO activity and promotes tumor cell apoptosis.[95] Beyond its role in forming membrane contact sites with CPT1, PLIN2 also participates in the HIF2α–PLIN2 regulatory axis, coordinating lipid storage and preventing ER stress. These functions establish PLIN2 as a multifaceted tumor therapy target.[81] LD-targeted strategies also demonstrate notable synergy with existing immunotherapies. Preclinical studies have shown enhanced antitumor efficacy when CD36 blockade is combined with anti-PD-1 antibodies, and pharmacological inhibition of FAO potentiates PD-1 checkpoint inhibition in autologous melanoma models.[123,137]
In summary, LD-targeted therapies represent a promising strategy for disrupting the protumorigenic functions of LDs in the TME. To advance this field, future studies should primarily focus on elucidating the precise molecular mechanisms that govern LD functionality across diverse tumor contexts and on developing additional inhibitors or antibodies against specific LD targets. In addition, efforts should also be directed toward characterizing LD heterogeneity across different cancer types, as well as optimizing targeted delivery systems to enhance therapeutic efficacy and improve targeting accuracy while minimizing systemic toxicity. Ultimately, the successful translation of these strategies into clinical practice will depend on a deeper understanding of LD biology within the TME and the development of rational combination approaches with existing therapies.

Challenges and Future Directions in LD Research
Despite significant progress in understanding LDs, several fundamental questions and unresolved discrepancies remain. The precise molecular mechanisms governing neutral lipid accumulation and LD formation within the ER membrane are incompletely characterized, with particular uncertainty surrounding the budding process. Current evidence presents conflicting views on whether LDs fully separate from the ER or maintain connections after budding.[20] Similarly, the regulatory mechanisms underlying LD fusion events—particularly the proteins and signaling pathways mediating mitochondrial-LD crosstalk during fusion—require further elucidation.[20,163–165] The traditional paradigm of LD degradation through lipolysis or autophagy is being challenged by emerging evidence suggesting alternative routes, such as endocytic uptake of LDs by lysosomes. However, the relative contributions and regulation of different degradation pathways across various cellular contexts remain unclear.[18,166–168] Furthermore, although LD interactions with mitochondria and lysosomes are known to regulate cellular metabolism, the molecular basis of these contacts and their functional consequences in distinct physiological states warrant systematic investigation.[12,169–171]
In the context of TME, critical knowledge gaps remain regarding the roles of LD-associated proteins in tumor progression[172,173] and the precise mechanisms through which LD accumulation contributes to tumor initiation, metastasis, and therapy resistance, although it is recognized as a hallmark of metabolic reprogramming.[174–176] The immunological consequences of LD accumulation present another area of uncertainty; although correlative studies demonstrate LD-associated dysfunction across multiple immune cell types in various tumor models,[105,177] mechanistic studies on MDSCs, DCs, and NK cells remain scarce.[109,112] Most notably, the intracellular signaling pathways linking LD accumulation to immune cell dysfunction constitute a significant unresolved question in the field. Addressing these challenges will require the development and application of advanced imaging techniques, systematic omics approaches, and genetically engineered models to explore LD biology with molecular precision, ultimately facilitating the translation of fundamental discoveries into therapeutic strategies for cancer and other diseases characterized by lipid metabolism dysregulation.
Although the therapeutic targeting of LD metabolism holds considerable promise, translating these strategies into clinical practice faces significant hurdles, primarily related to on-target, off-tumor toxicity. Lipid metabolism is essential to the homeostasis of almost all healthy tissues. For instance, systemic inhibition of DGAT1 is known to cause dose-limiting gastrointestinal side effects, such as diarrhea, owing to its critical role in lipid absorption and chylomicron production in the intestine.[178,179] In the immune compartment, nonspecific targeting may impair the function of lipid-dependent immune cells in peripheral tissues, such as CD8+ T cells or Tregs, where LDs play a bidirectional role.[119,125,127] Therefore, the ubiquitous expression of LD regulatory machinery necessitates highly selective delivery carriers to minimize systemic toxicity. To overcome these challenges, future efforts should focus on developing sophisticated TME-selective drug delivery systems. For instance, nanoparticles can be designed that release their payload specifically in response to TME hallmarks, such as low pH (e.g., using pH-sensitive linkers or polymers),[180,181] hypoxia (e.g., using hypoxia-sensitive bonds),[182,183] or enzyme overexpression (e.g., matrix metalloproteinase–cleavable peptides).[184,185] In addition, conjugating inhibitors or antibodies that target receptors that are highly enriched on specific cells within the TME represents a promising strategy.[186,187] Ultimately, the successful clinical translation of LD-targeted therapies will depend not only on target discovery but also on the development of delivery platforms that achieve a favorable therapeutic window.

Conclusions
LDs have evolved from being regarded merely as lipid storage organelles to being recognized as central regulators of tumor metabolic reprogramming and immune microenvironment modulation. Their formations—including neutral lipid synthesis, nucleation, and budding, regulated by key proteins such as DGAT and Seipin—not only provide energy reserves for tumor cells but also enable adaptation to microenvironmental stresses through dynamic membrane synthesis and the maintenance of lipid homeostasis. LDs display notable metabolic plasticity that confers protective functions across diverse tumor types, while simultaneously shaping immunosuppressive phenotypes within the TME. For instance, they promote protumoral polarization of TAMs by enhancing lipid uptake, modulate CD8+ T cell exhaustion through lipid overload, coordinate Treg-mediated immunosuppression, and impair DC and MDSC functions via disrupted antigen presentation or excessive FAO. These multifaceted roles position LDs as critical metabolic–immune hubs, thereby prompting diverse therapeutic strategies. Such strategies include inhibiting lipid uptake receptors such as CD36 and MARCO, disrupting biogenesis enzymes including DGAT1/2 and ACAT1, and targeting LD-mitochondria contacts involving CPT1A and PLIN2. Notably, preclinical studies demonstrate promising synergy when LD-targeted approaches are combined with immunotherapy.
Despite these advances, several fundamental questions remain unresolved. For instance, how does the heterogeneity of LDs across diverse tumor types and immune subsets impact their pro-tumor or antitumor functions? What are the precise molecular mechanisms that determine whether LDs protect against or promote ferroptosis? How do LDs interact with other organelles, such as mitochondria, the ER, and lysosomes, in a spatiotemporally specific manner within the TME? To what extent does LD accumulation in immune cells represent a reversible metabolic state versus a terminally differentiated phenotype? Addressing these questions will necessitate the development of more refined models, such as organelle-specific biosensors, spatial multiomics, and genetically engineered systems, that enable real-time tracking of LD dynamics in vivo. From a therapeutic perspective, targeting LD metabolism holds substantial potential but faces considerable challenges, particularly regarding cell-type specificity and on-target/off-tumor effects. Future studies should focus on developing TME-selective delivery systems, such as nanoparticles responsive to hypoxia, low pH, or overexpressed enzymes, to minimize systemic toxicity while maximizing efficacy. Combination strategies integrating LD-targeted agents with immunotherapies (e.g., anti-PD-1/CTLA-4), chemotherapy, radiotherapy, or ferroptosis inducers represent particularly promising avenues. In addition, dietary interventions that modulate lipid availability may synergize with pharmacological approaches to disrupt LD-driven tumor progression and immune evasion.
In summary, LDs represent a dynamic and context-dependent metabolic–immune interface in cancer. A deeper understanding of their biology, coupled with innovative therapeutic targeting strategies, could open new avenues for LD-based antitumor treatments. Such approaches may enhance patient survival and restore antitumor immunity.

Funding
This research was supported by the Noncommunicable Chronic Diseases-National Science and Technology Major Project (No. 2024ZD0525301) and the National Natural Science Foundation of China (Nos. 82172808 and 82303334).

Conflicts of interest
None.

Supplementary Material