Macrophages, a subset of long-lived phagocytes integral to the innate immune system, are widely distributed across most tissues and serve as the first line of defense against pathogens. However, under the specific pathological context of TIME, macrophages undergo phenotypic transformation, lose their protective functions, and become TAMs (14). Within the TIME of gastric cancer, the majority of TAMs arise from the differentiation and polarization of monocytes recruited to the tumor site, with M2-TAMs representing the predominant subset within this compartment (15).

M2-TAMs exert their protumorigenic effects through the secretion of extracellular vesicles (EVs), bioactive proteins, and multiple chemokines, coupled with metabolic reprogramming mechanisms that collectively drive tumor cell proliferation, metastatic progression, and therapeutic resistance. Additionally, these macrophages facilitate gastric cancer pathogenesis through dual-phase modulation of the hepatic microenvironment: priming the premetastatic niche formation during early dissemination stages and subsequently fostering angiogenesis establishment in metastatic lesions (16). A recent investigation utilizing multiplex immunohistochemistry co-localized CD163 and CD206 markers on TAMs in gastric cancer specimens. Notably, the CD206⁺ M2-TAM subset was identified as a phenotypically distinct subpopulation exhibiting PD-L1 positivity (17). Wang et al. demonstrated that exosomes derived from M2-TAMs carry the long non-coding RNA MALAT1 and transfer it to gastric cancer cells. MALAT1 stabilizes δ-catenin and upregulates HIF-1α expression, thereby enhancing glycolysis and promoting the proliferation, metastasis, and chemoresistance of gastric cancer cells (18). Fu et al. demonstrated that transmembrane protein 205 (TMEM205) promotes proliferation and stemness, and enhances epithelial–mesenchymal transition (EMT), migration, and angiogenesis in gastric cancer cells. Moreover, TMEM205-induced polarization of M2-TAMs further accelerates gastric cancer progression (19). Recent studies have shown that SERPINE, a serine protease inhibitor derived from gastric cancer cells, serves as a major driver of gastric cancer growth and promoting M2-TAM polarization. Through autocrine activation of the JAK2/STAT3 signaling pathway to mediate the transfer of exosomal let-7g-5p, thereby facilitating this polarization (20).

MDSCs constitute a heterogeneous population predominantly comprising myeloid progenitor cells and immature myeloid cells (IMCs). Under physiological conditions, they are generated in the bone marrow and serve as precursors to macrophages, dendritic cells, and granulocytes. However, during tumorigenesis, IMCs frequently fail to complete their differentiation, resulting in an accumulation of MDSCs (21). In tumor-bearing mice, two principal MDSC subpopulations have been characterized: polymorphonuclear MDSCs (PMN-MDSCs) and monocytic MDSCs (M-MDSCs) (22). Among them, PMN-MDSCs are defined as CD11b⁺ Ly6G⁺ Ly6C^low cells, whereas M-MDSCs are defined as CD11b⁺ Ly6G^- Ly6C^high cells. In mice, CD49d can be used in place of Gr-1 to identify highly immunosuppressive MDSCs. Because humans lack Gr-1, the equivalent human subsets are defined as CD11b⁺ CD14^- CD15⁺ cells (or CD11b⁺ CD14^- CD66b⁺ cells) and CD11b⁺ CD14^- HLA-DR^-/low CD15^- cells, respectively (23).

Within gastric cancer, MDSC subpopulations exhibit considerable complexity. Ding et al. (24) demonstrated that, during Helicobacter pylori infection, bone marrow–derived Schlafen4⁺ (SLFN4⁺) MDSCs migrate to the stomach and activate the NF-κB pathway, leading to robust induction of MiR-130b. MiR-130b not only sustains persistent NF-κB activation but also contributes to TIME formation by suppressing T-cell responses and directly stimulating epithelial cell proliferation, thereby promoting metaplasia and cancer progression. Subsequent studies have demonstrated that Toll-like receptor 9 (TLR9) upregulates IFNα expression in gastric epithelial cells and plasmacytoid dendritic cells (pDCs) during Helicobacter pylori infection, thereby facilitating the polarization of SLFN⁺ MDSCs and suppressing T cell proliferation (25). S100 A8/A9 heterodimer upregulates CXCL1 expression in gastric cancer cells via the TLR4/p38 MAPK/NF-κB pathway, thereby driving PMN-MDSC accumulation in the tumor microenvironment. Furthermore, PMN-MDSCs exploit S100 A8/A9 to inhibit CD8⁺ T-cell glycolysis, proliferation and TNF-α/IFN-γ production through the TLR4/AKT/mTOR pathway, ultimately inducing T-cell exhaustion (26).

Under normal physiological conditions, fibroblasts are classified as mesenchymal cells. However, their phenotypic diversity, functional heterogeneity, and lack of specific markers make it difficult to precisely determine their origin and function. During cancer development, CAFs are defined as non-epithelial, non-cancerous, non-endothelial, non-immune fibroblasts located within or adjacent to the tumor, which may derive from tissue-resident fibroblasts or from pancreatic and hepatic stellate cells (27). Years of investigation have identified two principal CAF subtypes with shared genetic profiles: myofibroblastic CAFs (myCAFs) and inflammatory CAFs (iCAFs). Paracrine factors secreted by epithelial cells—including TGF-β, IL-1α, and PDGF—play a central role in CAF reprogramming. Notably, the TGF-β and IL-1α axes have been shown to preferentially drive the differentiation of myCAFs and iCAFs, respectively (28).

Li et al. (29) identified four CAF subpopulations with distinct properties in gastric cancer. Among these, iCAFs and extracellular matrix CAFs (eCAFs) engage in bidirectional signaling with neighboring immune cell subpopulations in the tumor microenvironment. Specifically, iCAFs modulate T-cell behavior by secreting IL-6 and CXCL12, whereas eCAFs promote M2-TAM recruitment through elevated periostin (POSTN) expression. Another study demonstrated that the fat content and obesity-related gene (FTO) promote M2-TAM polarization by regulating m6A-dependent demethylation of Nicotinamide N-methyltransferase (NNMT) in CAFs, thereby driving gastric cancer progression (30). In addition, Eckert et al. (31) identified NNMT as a key metabolic regulator of CAF differentiation and showed that inhibition of its activity can reverse the CAF phenotype. A recent study identified antigen-presenting CAFs (apCAFs) in gastric cancer that not only enhance T cell activation, cytotoxicity, and proliferation—thereby augmenting T cell–mediated antitumor immunity—but also drive macrophage polarization toward a proinflammatory phenotype. These polarized macrophages, in turn, reinforce the formation of apCAFs, establishing a positive feedback loop that further amplifies antitumor immune responses (32).

Regulatory T cells (Tregs) are an immunosuppressive subset of CD4⁺ T cells that are essential for maintaining immune homeostasis. FoxP3, the master transcription factor governing Treg differentiation, development and functional integrity, is critical to their suppressive activity. Within the TIME, Tregs are typically highly activated and exhibit potent immunosuppressive functions, characterized by upregulated expression levels of FoxP3 and Helios (33, 34). A recent study reported that hemokine (C-C motif) receptor 8 (CCR8) is highly expressed on effector tumor-infiltrating regulatory T cells (TI-Tregs), but its expression remains relatively low on peripheral Tregs and conventional T cells in both mice and humans, rendering CCR8 an ideal candidate for the selective targeting of TI-Tregs (35).

In gastric cancer, Tregs exert immunosuppressive effects via multiple mechanisms that promote tumor progression. Studies have demonstrated that TNFR2⁺ Tregs accumulate in the tumor microenvironment as the disease advances, and their level of infiltration can serve as a prognostic marker. Moreover, in vitro experiments reveal that TNF-α/TNFR2 signaling upregulates Foxp3 expression in CD4⁺CD25⁺ T cells and increases TGF-β secretion by Tregs, further enhancing their immunosuppressive capacity (36). Lv et al. (37) demonstrated that gastric cancer–derived IL-33 induces activation of the p38 MAPK pathway in mast cells, leading to IL-2 secretion and consequent expansion of ICOS⁺ Tregs. This expansion enhances the immunosuppressive capacity of Tregs while attenuating anti-tumor CD8⁺ T-cell activity, thereby driving gastric cancer progression. Wang et al. (38) identified the CLDN18-ARHGAP fusion gene as a primary source of immunogenic neoepitopes. As a pivotal regulator of the tumor immune microenvironment, this fusion enhances Treg survival by activating the PI3K-AKT-mTOR-FAS signaling cascade and augmenting free fatty acid (FFA) production, thereby promoting the establishment of the gastric cancer TIME.

Since Single-Cell RNA Sequencing (scRNA-seq) was designated Technology of the Year by Nature Methods in 2013 (39), scRNA-seq has been widely used to study TIME heterogeneity in gastric cancer. Sathe et al. (40) performed single-cell transcriptome sequencing in patients with gastric cancer and intestinal metaplasia, revealing significant enrichment of stromal cells, macrophages, dendritic cells (DC) and T cells in the gastric cancer TIME, accompanied by an extensive cell-reprogramming phenomenon. By sequencing the single-cell transcriptomes of gastric cancer and adjacent mucosa (AM) samples, the researchers revealed pronounced intra- and intertumoral heterogeneity among tumor epithelial cells, whereas CAFs exhibited predominantly intratumoral heterogeneity. In addition, four CAF subgroups with distinct characteristics were identified; although they resemble resident fibroblasts in the AM, they display enhanced pro-tumor activity (29). Bian et al. (41) utilized the optimized single-cell multi-genomic sequencing method (scTrio-seq3) to map the DNA methylation landscape of gastric cancer at single-cell resolution and to identify candidate DNA methylation biomarkers. Moreover, they systematically delineated the relationships among genetic lineage, DNA methylation patterns, and transcriptional clusters at the single-cell level, providing a more detailed analysis of the molecular features underlying intratumoral heterogeneity and differentiation status in human gastric cancer than traditional approaches. Some researchers performed scRNA-seq and single-cell TCR sequencing (scTCR-seq) on patients with newly treated gastric adenocarcinoma across various Lauren subtypes. Their findings showed that intratumoral heterogeneity (ITH) serves as a prognostic marker for recurrence. Compared with ITH-L tumors, ITH-H tumors exhibited pronounced immunosuppressive features, including a reduced number of activated CD8+ T cells, an increased number of depleted CD8+ T cells, and marked polarization of M2-TAMs (42). Through scRNA-seq analysis of malignant cells in gastric cancer ascites, we identified a gastric-dominant subtype—predominantly composed of gastric cell lines—and a GI-mixed subtype—characterized by a combination of gastric- and colon-like cells. When integrated with immune-infiltration data from public databases, the superior prognosis of GI-mixed tumors appears linked to a more effective antitumor immune response, hallmarked by high levels of polarized B cells and M1-TAMs, low levels of polarized fibroblasts and M2-TAMs, and elevated cytolytic activity (43).

Spatial transcriptomics (ST) sequencing technology was designated Technology of the Year in 2020. Its principal advantage lies in its ability to profile gene expression while preserving spatial context within tissues, offering significant potential to elucidate complex, heterogeneous microenvironments (44, 45). Several recent pioneering studies have further broadened the applications of this technology. For instance, Joakim Lundeberg’s group (46) introduced a depth‐generation model based on spatial expression data that fuses low-resolution ISC datasets with high-resolution histological images to infer super-resolution expression maps, enabling analysis of gene expression within fine-grained anatomical structures. Likewise, the Mingyao Li team (47) developed iStar, a hierarchical image-feature extraction method that integrates ST data with high-resolution histological images to predict spatial gene expression at super-resolution.

ST technology plays a critical role in gastric cancer research. Joseph J. Zhao’s team utilized it to perform spatial transcriptomic analysis on 67 samples. Their findings revealed significant differences in the molecular characteristics and immune composition between peritoneal metastasis and other types of metastasis (such as liver metastasis), suggesting that peritoneal metastasis may involve a distinct biological mechanism. These results further underscore the pivotal role of TIME in transluminal metastasis (48). In addition, Patrick Tan’s research team integrated spatial transcriptomic and single-cell sequencing data, revealing that gastric cancer exhibits intra-tumor heterogeneity (ITH) based on RNA expression, characterized by two distinct evolutionary trajectories. These trajectories are closely associated with molecular subtypes, clinical prognosis, and the stromal microenvironment, where the tumor-stroma interface demonstrates a unique ecological state mediated by TGF-β signaling. Furthermore, the study identified SOX9 as a potential intrinsically dispersed evolutionary driver (48). These findings not only deepen the understanding of gastric cancer heterogeneity but also offer new perspectives for the development of precise therapeutic strategies and spatial biomarkers.

Warburg’s seminal report, published 100 years ago, demonstrated that cancer tissue sections convert glucose to lactic acid despite sufficient oxygen availability; this phenomenon is now called the Warburg effect (49). Although glycolysis yields far less energy than oxidative phosphorylation, cancer cells nevertheless depend on it by consuming large quantities of glucose. To meet the metabolic demands of rapid proliferation, cancer cells frequently alter receptor-mediated signaling pathways through specific genetic mutations, thereby activating and upregulating nutrient-uptake mechanisms. As a result, their capacity for glucose uptake and utilization is markedly enhanced (50). In gastric cancer cells, cytoplasmic β-Arrestin1 (ARRB1) binds to pyruvate kinase M2 (PKM2), inhibiting PKM2 tetramerization, decreasing its enzymatic activity, and modulating metabolic flux, thereby shifting cellular metabolism from oxidative phosphorylation to aerobic glycolysis (51). In addition, studies have demonstrated that Glucose Transporter 3(GLUT3) expression is markedly up-regulated in gastric cancer cells and is inversely correlated with patient prognosis. GLUT3 may facilitate tumor growth and metastasis by modulating Lactate dehydrogenase A(LDHA) activity (52). Driven by the Warburg effect, lactic acid—the end product of glycolysis—accumulates at high levels within gastric cancer cells and is exported extracellularly, where it plays a pivotal role in promoting immune escape and reshaping the gastric cancer TIME.

Lactatic-driven lactoylation of NBS1K388 promotes the formation of the MRE11-RAD50-NBS1(MRN) complex, thereby enhancing DNA-repair capacity and mediating chemotherapy resistance. This modification is regulated by TIP60-mediated lactoylation and HDAC3-catalyzed delactylation, and elevated NBS1 lactoylation correlates with poor prognosis in patients undergoing neoadjuvant chemotherapy (53). In tumors characterized by high glycolytic flux—including gastric cancer—excessive glucose uptake by tumor cells generates a microenvironment of low glucose and high lactate. Regulatory T cells (Tregs) actively import lactate via monocarboxylate transporter 1 (MCT1), inducing nuclear translocation of activated nuclear factor of activated T cells 1 (NFAT1) and consequent upregulation of PD-1; while PD-1 expression in effector T cells is inhibited, that is, the PD-1 expression and its inhibitory activity of Tregs are enhanced, which leads to poor PD-1 blocking therapy (54).

Lipid metabolic reprogramming is increasingly recognized as a hallmark of tumor cells (55). It plays an important role in the progression of gastric cancer. Kumagai et al. (56) found that gastric cancers harboring the RHOA Y42C mutation exhibited significant enrichment of fatty acid metabolism–related gene sets and upregulation of fatty acid synthase (FASN) expression. Further studies revealed that this mutation not only impaired chemokine recruitment but also enhanced free fatty acid (FFA) production in effector T cells via the PI3K–AKT signaling pathway. The resulting elevated FFA levels promoted Treg cell accumulation, thereby facilitating the formation of TIME; elevated FFA levels, in turn, promote regulatory T-cell (Treg) accumulation and thereby facilitate tumor immune microenvironment (TIME) formation. Furthermore, studies have demonstrated that elevated intracellular lipid levels are a hallmark of lymph node–positive (N+) gastric cancer. In N+ gastric cancer, leptin-induced phosphorylation of angiopoietin-like protein 4 (ANGPTL4) enhances lipid uptake via overexpressed lipoprotein lipase (LPL), thereby stimulating prostaglandin E2 (PGE2) production and ultimately facilitating lymph node metastasis (57). Oleic acid (EA), produced via stearoyl-CoA desaturase (SCD)-dependent fatty acid desaturation, promotes the proliferation and survival of dysplastic gastric epithelial cells, thereby establishing an energy-supply chain through carcinogenic fatty acid metabolism during gastric cancer development. This metabolic reprogramming contributes to the reorganization of the gastric cancer tumor immune microenvironment (TIME) (58). Recent studies have demonstrated that in gastric cancer patients the intestinal microbiota is disrupted, leading to aberrant lipid metabolism that facilitates tumor progression. A significant reduction in the abundance of short chain fatty acid (SCFA)–producing bacteria results in decreased butyrate levels. SCFAs bind specifically to Hydroxycarboxylic acid receptor 2(HCAR2, also called GPR109A), which selectively recognizes butyrate, inhibiting gastric cancer progression by enhancing CD8⁺ T-cell–mediated cytotoxicity, including that of CAR-Claudin 18.2⁺ CD8⁺ T cells (59).

In pan-cancer, the hyperphysiological abundance and dysregulated metabolism of nucleotides in cancer cells not only support rapid proliferation and DNA repair, but also promote immune evasion and drug resistance through multiple mechanisms. For instance, cancer cells modulate MHC-I expression, antigen presentation, and immune-related signaling pathways (such as purine metabolism-mediated immune regulation), thereby attenuating the anti-cancer activity of various cells within the tumor TIME (60).

Multiple studies have demonstrated the critical role of nucleotide metabolism in gastric cancer development. Through comprehensive genomic analysis of Helicobacter pylori, Tang et al. (61) discovered the 171S/L HtrA mutation as a unique bacterial cancer-associated SNP demonstrating significant association with gastric cancer development and progression. In addition, the study found that coactivator associated arginine methyltransferase 1 Gene(CARM1) is overexpressed in gastric cancer. Under the low-glucose conditions in the tumor microenvironment, Nuclear factor erythroid 2-related factor 2 (NRF2) becomes activated and binds to the CARM1 promoter, thereby upregulating its expression, and then triggering CARM1-mediated hypermethylation of histone H3 methylated at R arginine 17 (H3R17me2) in the glucose-6-phosphate dehydrogenase gene body. This epigenetic modification redirects glucose carbon flux toward the pentose phosphate pathway, facilitating nucleotide synthesis (e.g., production of nucleotide precursors such as ribose-5-phosphate) and maintaining redox homeostasis, ultimately promoting the survival and growth of gastric cancer cells (62). Furthermore, pyrimidine biosynthesis not only potentiates Notch signaling but also upregulates c-Myc transcriptional activity, consequently elevating key glycolytic enzymes. Importantly, the enhanced expression of critical pyrimidine synthesis enzymes CAD and DHODH confers chemotherapy resistance in gastric malignancies through glycolytic pathway activation (63). In HER2-positive gastric cancer, tumor cells activate CDC42 signaling to induce phosphorylation of NF-κB p65, thereby promoting the secretion of GLS1-enriched microvesicles. This molecular mechanism coordinates macrophage glutamine metabolic remodeling, facilitates M2 polarization of tumor-associated macrophages, and enhances pro-angiogenic signaling pathways, collectively contributing to the development of trastuzumab resistance (64).