Interaction of Cell Subtypes in Tumor Microenvironment and Its Implications for Immunotherapy: Summary, Strengths, Weaknesses, and Future

Hong Jiang¹Lingfeng Fu²Takatsugu Ishimoto^2*Tao Zhang^1*Jun Zhang^1,3*

• ¹Liaoning Cancer Hospital, China Medical University, Shenyang, Liaoning Province, China
• ²Japanese Foundation For Cancer Research, Tokyo, Japan
• ³Kumamoto University, Kumamoto, Japan

The tumor microenvironment (TME) is a highly dynamic and heterogeneous network where immune, stromal, and malignant cells interact to modulate cancer progression and immunotherapy outcomes. transporters to sustain CAF energy production (7). This metabolic coupling enhances immunosuppression and induces drug resistance through epigenetic modifications (e.g., histone acetylation via LDHA) (8). For instance, LDHAcatalyzed lactate dehydrogenase promotes histone acetylation of genes involved in drug resistance (e.g., ABC transporters). Additionally, senescent pancreatic cancer cells secrete CCL20, recruiting M2 macrophages and exacerbating immune suppression (7).Additionally, metabolic reprogramming within the TME, including lactate accumulation and hypoxia, further suppresses immune responses by inducing T cell exhaustion and promoting regulatory immune cell populations. A particularly novel finding is the role of tumor cell senescence in shaping the TME. In pancreatic cancer, senescent tumor cells secrete CCL20, which recruits M2 macrophages and enhances immune suppression (7). Hypoxia and lactic acid accumulation could drive T cell exhaustion and M2 TAM polarization via the mTOR-HIF-1α axis. HIF-1α upregulates PD-1 expression on T cells and inhibits their mitochondrial function, while lactate activates NLRP3 inflammasomes to promote Treg expansion. ECM stiffness, mediated by PI3K/Akt signaling, further induces M2 TAM polarization and metastasisassociated gene expression (5). In breast cancer models, hard collagen matrices increased MMP-9 expression and liver metastasis, whereas soft matrices suppressed these effects (9). These insights suggest that combining metabolic (e.g., LDHA/MCT4 inhibitors) and mechanical (e.g., PI3K inhibitors) interventions could synergistically reverse immune suppression.To counteract these challenges, several emerging immunotherapeutic strategies have been proposed. One promising approach is targeting the Hippo signaling pathway, which influences macrophage polarization and immune cell differentiation. Hippo pathway inhibition enhances M1 TAM polarization by downregulating YAP/TAZ transcription factors, thereby augmenting antitumor immunity (10). For example, YAP/TAZ knockout macrophages secreted elevated levels of IL-12 and IFN-γ, leading to enhanced CD8+ T cell infiltration and tumor regression (11). Another strategy involves integrating spatial transcriptomics with multi-omics profiling to define patient-specific TME signatures, which could guide the development of personalized immunotherapies (12). In melanoma patients, spatial metabolic analysis revealed that lactate accumulation correlated with reduced CD8+ T cell infiltration, suggesting MCT4-targeted interventions could improve treatment responses. Clinical trials demonstrated that triple checkpoint blockade (CD47-SIRPα+PD-1) achieved 70% complete remission in melanoma patients, while anti-IL-8 nanobodies increased CTLA-4 efficacy from 18% to 54% by blocking neutrophil infiltration (13). These approaches hold promise for improving immune checkpoint blockade (ICB) responses and overcoming resistance mechanisms.Despite these valuable contributions, several limitations remain. Current studies predominantly rely on xenograft models (>70%), which lack humanspecific features such as neuro-immune interactions (fimmu-15-1434030). B cell dynamics, CD8+ T cell clonality, and immune temporal evolution remain underexplored, while static analyses limit understanding of TME dynamics.Another challenge is the variation in experimental methodologies across studies, making it difficult to standardize findings and draw definitive conclusions. Experimental variability across systems (cell lines/PDX/organoids) and technologies (scRNA-seq/spatial metabolomics) hinders cross-validation.Future efforts should integrate 4D spatiotemporal omics (live imaging + singlecell temporal analysis) with AI-driven predictive models (e.g., TME-DynaPredict) to monitor TME evolution in real time. Patient stratification strategies, such as proteinase inhibitor therapy for high-senescence-risk pancreatic cancer patients (7), aim to reverse chemoresistance. Targeted delivery systems (e.g., nanobodies) could enable precise TME reprogramming, advancing immunotherapy toward precision medicine (14,15).