Advances in the regulation of macrophage polarization by the tumor microenvironment

Abstract

The tumor microenvironment (Tumor Microenvironment, TME) is a core regulatory factor in the occurrence, development, and treatment resistance of tumors. Macrophages, as key immune cell components in the TME, have a profound impact on the tumor process (Visser and Joyce in Cancer Cell 41:374–403, 2023). This review aims to systematically elucidate the characteristics and functional differences of macrophage polarization into M1 and M2 phenotypes within the TME. Additionally, it endeavors to dissect the regulatory mechanisms by which metabolic products, cytokines, and extracellular matrix components secreted by tumor cells modulate macrophage polarization (Wang et al. in Mol Cancer 23:268, 2024). Moreover, the metabolic reprogramming of tumorassociated macrophage (TAM) is a core mechanism for their functional shift, and intervening in metabolic pathways holds promise for reprogramming TAMs to inhibit tumor progression (Jin et al. in Nat Cancer 6:239–252, 2025). Within the TME, macrophages can be polarized into classically activated M1 and alternatively activated M2 types (Ge and Wu in Zhongguo Fei Ai Za Zhi 26:228–237, 2023). Accumulating evidence indicates that classically activated M1 macrophages orchestrate anti-tumor immunity by secreting pro-inflammatory cytokines, including tumor necrosis factor-α (TNF-α) and interleukin-12 (IL-12), which collectively activate cytotoxic T lymphocyte responses, induce tumor cell apoptosis, and enhance immune surveillance (Luo et al. in Front Immunol 15:1352946, 2024). In contrast, M2 macrophages are induced in the TME and promote tumor angiogenesis, immune evasion, tumor cell proliferation, and metastasis by secreting factors such as vascular endothelial growth factor (VEGF) and transforming growth factor-beta (TGF-β) (Wang et al. in NPJ Precis Oncol 8:31, 2024). Therefore, in-depth research on the mechanisms of macrophage polarization in the tumor microenvironment provides an important basis for developing new tumor immunotherapy strategies and has significant clinical translational value.

Characterization and composition of the tumor microenvironment

The tumor microenvironment is a dynamic ecosystem surrounding tumor cells and consists of a variety of cellular and non-cellular components, whose properties and composition have a significant impact on tumor progression and treatment response [1]. TME has significant immunosuppressive properties and impairs anti-tumor immune responses by recruiting immunosuppressive cells (e.g., myeloid-derived suppressor cells, MDSCs, and regulatory T cells, Tregs) and by modulating metabolic processes [7]. In addition, the TME is a dynamically changing ecosystem whose composition and function change in response to tumor progression and therapeutic interventions [8]. At the same time, TME also exhibits a high degree of heterogeneity, and its composition and properties may vary significantly across tumor types and even different regions within the same tumor, and this heterogeneity increases the complexity of treatment [9].

The composition of TME includes both cellular and non-cellular components [10]. The cellular components mainly include immune cells (e.g., T cells, B cells, macrophages, etc.), stromal cells (e.g., cancer-associated fibroblasts CAFs, endothelial cells, etc.), and tumor cells themselves [11]. In particular, macrophages can polarize to the pro-tumorigenic M2 type TAMs, which promote tumor growth and immune escape by secreting cytokines and chemokines [12]. Non-cellular components mainly include extracellular matrix (ECM) and soluble factors [13]. The ECM consists of collagen, fibronectin, and laminin, which provide physical support for tumor cells and are involved in intercellular signaling [14]. Soluble factors, on the other hand, include cytokines (e.g., IL-6, IL-10, TGF-β, etc.), chemokines (e.g., CXCL8, CCL2, etc.), and metabolites (e.g., lactate, adenosine, etc.) [15]. These factors transmit signals between cells and regulate cellular behavior and function; for example, Wnt5a induces M2 polarization in tumor-associated macrophages, which promotes colorectal cancer progression [16]. These non-cellular components interact with cellular components and together shape the complexity of the TME.

Biology and origin of macrophages

Macrophages are derived from bone marrow stem cells, Myeloid stem cells in bone marrow develop into pre-monocytes by the bone marrow microenvironment [17, 18]. Pre-monocytes develop into monocytes in the bone marrow in response to certain cytokines and enter the bloodstream [19]. Monocytes remain in the bloodstream for only 12–24 h before entering the adult organs and maturing into macrophages [20]. Macrophages that enter and settle in tissues usually do not return to the bloodstream but become free-moving and swim in the interstitial space or become fixed in the tissues. Macrophages belong to the mononuclear phagocytic system, which is composed of a heterogeneous group of cells that play a critical role [21]. Their heterogeneity enables them to play an indispensable role in immune defense, tissue homeostasis maintenance, inflammation regulation, and injury repair [22]. During the immune response, macrophages with different phenotypes can precisely regulate the balance between innate and adaptive immunity by recognizing pathogens, secreting cytokines, and presenting antigens [23]. In terms of tissue homeostasis, alveolar macrophages can remove inhaled particles and maintain the integrity of the respiratory epithelium, while hepatic Kupffer cells are responsible for removing senescent red blood cells and toxins from the blood, jointly maintaining the stability of the organ microenvironment [24]. Also, macrophages are plastic cells that can undergo phenotypic switching [25]. Different macrophage phenotype, such as M1 and M2 macrophages, form in response to environmental changes, but they play different roles in normal immune responses and disease development [26, 27]. Monocytes are recruited by chemokines, cytokines, and growth factors produced by tumor cells or mesenchymal stromal cells and further differentiate to give rise to TAM [28, 29]. TAM, as one of the major immune components, plays a very important role in the TME [30]. Macrophages also play a key role in lymphangiogenesis during development and inflammation [31]. Lymphangiogenesis can lead to tumor metastasis [32]. Therefore, understanding and further researching the regulatory roles of macrophages is of great importance.

The molecular mechanisms of macrophage polarization

Macrophage polarization refers to the process by which macrophages differentiate into different functional subsets under the stimulation of various microenvironments, mainly including the pro-inflammatory M1 type and the anti-inflammatory M2 type [33]. The molecular mechanisms of macrophage polarization are intricate, encompassing various signaling pathways, transcription factors, and metabolic regulation. As depicted in Fig. 1, M1 polarization is predominantly triggered via the Toll-like receptor (TLR) pathway. For instance, exposure to lipopolysaccharide (LPS) and interferon-γ (IFN-γ) activates the Janus kinase (JAK)/signal transducer and activator of transcription (STAT) signaling cascade [34, 35]. In addition, the ATM-Chk2 axis activated by reactive oxygen species (ROS) promotes M1 polarization through metabolic reprogramming [36]. In addition to transcription factors, microRNAs (miRNAs) also play a significant role in regulating macrophage polarization. For example, let-7e targets caspase-3 to inhibit apoptosis in myeloid-derived suppressor cells (MDSCs) and promotes their polarization towards the M1 phenotype under LPS stimulation [37, 38].

Fig. 1.

Mechanism Diagram of the Signaling Pathway

M2 polarization exhibits distinct regulatory mechanisms within the TME. This polarization process is influenced by type 2 T helper (Th2) cell-derived factors including interleukin-4 (IL-4), IL-6, IL-10, and IL-13, as well as other mediators such as transforming growth factor-β (TGF-β), glucocorticoids, immune complexes, Toll-like receptor (TLR) or IL-1R ligands, and leukemia inhibitory factor (LIF) [39]. M2 polarization is induced by cytokines such as IL-4 and IL-10, which activate signaling pathways like STAT6 and PI3K/AKT [33]. For instance, TIPE2 drives M2 polarization through boosting AKT activity and modulating phosphatidylinositol metabolism [40]. M2-type macrophages suppress T cell function by depleting arginine via upregulated Arg1 expression while simultaneously promoting tumor proliferation through enhanced polyamine synthesis [41]. At the transcriptional level, the core regulatory factors STAT6 and IRF4 in M2-type macrophages not only promote the secretion of anti-inflammatory factors (e.g., IL-10 and TGF-β), but also enhance lipid metabolism by activating the PPARγ/LXR signaling network [42, 43].

Metabolic regulation is also crucial for macrophage polarization. M1 macrophages predominantly utilize glycolysis, generating substantial amounts of nitric oxide (NO) and pro-inflammatory cytokines [44]. In contrast, M2 macrophages preferentially rely on oxidative phosphorylation and fatty acid oxidation to produce anti-inflammatory cytokines such as IL-10 [45]. These intricate interplays among signaling pathways, metabolic regulation, and transcription factors collectively determine the functional status of macrophages in the tumor microenvironment, thereby providing potential targets for cancer immunotherapy.

Limitations of the M1/M2 classification

While the M1/M2 dichotomy provides an important framework for studying macrophage functions, its oversimplified nature may fail to fully capture the complexity and dynamic nature of macrophages within the TME [46]. State-of-the-art single-cell transcriptomics and spatial multi-omics technologies have fundamentally transformed our comprehension of macrophage biology. These advanced approaches demonstrate that macrophage phenotypes in vivo exist along a dynamic continuum rather than discrete categories, exhibiting remarkable plasticity shaped by: (1) local microenvironmental cues including metabolic signaling and extracellular matrix composition; (2) developmental origins distinguishing tissue-resident from monocyte-derived populations; and (3) disease-specific pathological contexts. This multifaceted regulation network renders conventional binary classification systems biologically inadequate for capturing the true complexity of macrophage functional states [47]. Recent studies have revealed a paradigm-shifting phenomenon where certain macrophage populations can co-express both classical M1 markers (e.g., iNOS) and alternative M2 markers (e.g., Arg-1), demonstrating an unprecedented capacity for rapid functional switching between antimicrobial defense and tissue repair processes. This remarkable plasticity fundamentally challenges the explanatory power of traditional classification systems, exposing their inability to capture the dynamic functional spectrum of these versatile immune cells [26, 45]. Moreover, macrophages exhibit extraordinary phenotypic plasticity that enables dynamic adaptation to diverse TME cues—including cytokine gradients, metabolic byproducts, and hypoxic conditions. This continuous state of functional reprogramming renders conventional static classification schemes inadequate for characterizing their context-dependent roles in cancer immunobiology [48]. Furthermore, current surface marker-based detection methods (e.g., CD86 for M1-like or CD206 for M2-like phenotypes) may fail to comprehensively characterize macrophage heterogeneity, particularly when analyzing at single-cell resolution. This technical limitation underscores the need for multi-dimensional profiling approaches to fully decipher macrophage functional states [22, 49]. Therefore, future investigations should integrate single-cell multi-omics with spatial profiling technologies to achieve higher-resolution characterization of macrophage subsets, ultimately enabling the establishment of functionally defined classification frameworks that transcend conventional surface marker-based paradigms [50].

The dual role of macrophage polarization in tumor progression

Antitumor inhibitory effect

Macrophage polarization exhibits a dual and context-dependent role in tumor progression, demonstrating both tumor-suppressive and tumor-promoting capacities [49]. The activation of M1-type macrophages requires two distinct signals: (1) Th1-derived cytokines (e.g., interferon-γ) and (2) exogenous tumor necrosis factor-α or endogenous TNF inducers (e.g., bacterial components such as lipopolysaccharide) [51]. M1-polarized macrophages are characterized by the production of multiple pro-inflammatory mediators, including TNF-α, IL-1, IL-6, reactive nitrogen species, and oxygen intermediates, which confer potent bactericidal and tumoricidal activities [5]. These effector functions are mediated through the activation of the NADPH oxidase system, generating reactive oxygen species (ROS) that contribute to pathogen clearance during infections [26]. The anti-tumor mechanisms of M1 macrophages involve three principal pathways: (1) direct cytotoxic effects through pro-inflammatory cytokine production and inflammatory cell recruitment, facilitating the clearance of apoptotic cells and cellular debris [18]; (2) enhancement of antigen presentation by dendritic cells (DCs) to initiate adaptive immune responses [52]; and (3) ROS-mediated tumor cell killing. Given these multifaceted anti-tumor properties, pharmacological induction of M1 polarization has emerged as a promising therapeutic strategy in oncology. The M1-polarized state correlates strongly with favorable anti-tumor outcomes, providing a rationale for macrophage-based cancer immunotherapy approaches [53].

Pro-tumorigenic effects

However, various factors in the tumor microenvironment can induce macrophage polarization toward the M2 phenotype, also known as tumor-associated macrophages (TAMs). As presented in Table 1, M2 macrophage polarization leads to four functionally distinct subpopulations (M2a-d), each characterized by unique stimulation responses and transcriptional signatures [54]. Within the tumor microenvironment, resident macrophages acquire TAM phenotypes in response to local cues, exhibiting M2-like characteristics that support tumor progression through the secretion of growth factors and matrix-remodeling enzymes [39]. M2-type macrophages promote malignant progression in the tumor microenvironment through a multiple mechanism.M2 macrophages secrete immunosuppressive molecules such as IL-10 and TGF-β, which inhibit the antitumor activity of T cells and NK cells, thereby facilitating immune escape of pharyngeal carcinoma cells [55]. M2-polarized TAMs stimulate IL-6 production in tumor cells, which functions as a critical growth and survival factor. This IL-6 subsequently promotes tumorigenesis through activation of the JAK-STAT3 signaling pathway [56]. With the assistance of M2 (TAMs), tumor cells invade adjacent tissues and suppress immune responses. Meanwhile, IL-6-mediated chronic inflammation creates a favorable microenvironment for tumor growth [57]. M2-polarized tumor-associated macrophages (TAMs) further facilitate angiogenesis and pre-metastatic niche formation through the secretion of vascular endothelial growth factor (VEGF) and angiopoietin-2 (ANGPT2) [58]. Consequently, the predominant M2-polarized TAM population exhibits tumor-promoting functions that drive cancer progression.

Table 1.

Types ofmacrophage M2 polarization

M2 phenotype	Stimulus	Secreted cytokines	Function
M2a	IL-4, IL-13	TNF-α,TGF-β.IL-10	Anti-inflammatory, repalr, and immunomodulatory
M2b	TLR ligands	IL-1β,TNF-α,IL-6,I-10,CCL1	Weakens immune and inflammatory responses
M2c	IL-10, TGF-β	IL-10, TGF-β,CCL16, CCL18, CXCL13	Phagocytosis of apoptotic cells
M2d	TLR antagonists, IL-6, LIF [59]	IL-10, VEGF	Promote angiogenesis and tumor growth

Multidimensional regulation of macrophage polarization by the tumor microenvironment

TME has a dual role, acting both as a promoter of tumorigenesis and a coordinator of immunosuppression, thereby facilitating cancer progression and enabling immune escape [60]. The TME can influence the recruitment and polarization of macrophages, thereby driving these pro-tumorigenic outcomes [4]. Studying TME-induced macrophage polarization is crucial for further understanding TAM-related pro-tumorigenic outcomes and the potential development of novel therapeutic approaches. Therefore, a comprehensive understanding of the TME and its intricate involvement in tumor initiation, progression, and metastasis is also critical for developing effective anticancer drugs. Macrophage polarization is driven by cues within the TME, which can include signaling pathways, metabolic reprogramming, cell–cell interactions, and physical microenvironmental factors.

The role of TME soluble factors

The diverse cytokines secreted by tumor cells and immune cells within the TME form a core regulatory network governing polarization [61]. As summarized in Table 2, Cytokines such as IL-4, IL-10, and TGF-β promote M2-type polarization by activating the STAT6 and SMAD signaling pathways [62]; In contrast, IFN-γ and LPS induce M1-type polarization through the STAT1 and NF-κB signaling pathways [63]. IL-6 can promote M2-like polarization via STAT3 signaling, enhances PD-L1 expression, and contributes to immune suppression [64, 65]. Chronic TNF-α exposure can drive M2-like polarization via NF-κB and p38 MAPK pathways or induce secretion of IL-10 and VEGF, aiding immune evasion and angiogenesis [6, 66]. Notably, these cytokines exhibit spatiotemporal dynamics within the TME, with pro-inflammatory factors dominating in early stages and immunosuppressive cytokines becoming predominant during advanced progression [67]. Moreover, tumor cell-derived colony-stimulating factor 1 (CSF-1) binds to CSF-1R on macrophages, activating the Ras/MEK1/2/ERK1/2 signaling pathway. This activation upregulates Ets family transcription factors (e.g., Ets2) and cell cycle regulators (e.g., Cyclin D), thereby promoting TAM proliferation. Concurrently, ERK1/2 pathway activation suppresses differentiation-related factors (including c-Maf/MafB) while enhancing the secretion of anti-inflammatory cytokines (such as IL-10), collectively driving TAM polarization toward an M2-like phenotype [68, 69].

Table 2.

Cytokines and Their Functions in the TME

Cytokine type	Representative factor	Major sources	Core functions in TME	Pro-/Anti-tumor role	Mechanism/target
Pro-inflammatory Cytokines	IL-1β	Macrophages, TAMs, Cancer cells	Activates NF-κB pathway to enhance IL-6 secretion; promotes tumor invasiveness and chronic inflammation-driven progression [70]	Pro-tumor	Binds IL-1R (CD121a/b) to activate downstream signaling
	IL-6	Macrophages, Tumor cells, Fibroblasts	Promotes tumor proliferation, inhibits apoptosis; induces MDSC accumulation and T-cell suppression [71]	Dual role (concentration-dependent)	Binds IL-6R/gp130 complex to activate JAK-STAT3 pathway
Anti-inflammatory/Immunosuppressive	TGF-β	Tregs, CAFs, Tumor cells	Suppresses CD8⁺ T cells; promotes Treg differentiation; induces EMT to drive metastasis [72]	Pro-tumor (advanced stage)	Smad-dependent and -independent pathways
Th1-type Immune Activators	IFN-γ	CD8⁺ T cells, NK cells	Enhances antigen presentation (MHC-I/II upregulation); induces tumor apoptosis; upregulates PD-L1 to drive immune evasion [73]	Dual role (time-dependent)	Binds IFN-γR to activate JAK-STAT1 pathway
Th2-type/Pro-fibrotic	IL-4/IL-13	Th2 cells, TAMs	Drives M2 macrophage polarization; activates CAFs to promote fibrosis and immunosuppression [74]	Pro-tumor	IL-4Rα-mediated STAT6 activation
Chemokines	CXCL12 (SDF-1)	CAFs, Endothelial cells	Recruits CXCR4⁺ immunosuppressive cells (e.g., MDSCs, Tregs); promotes angiogenesis and metastasis	Pro-tumor	CXCL12-CXCR4 axis activates MAPK/ERK pathways
Growth Factors	VEGF	Tumor cells, TAMs	Promotes angiogenesis; induces immunosuppressive TME by suppressing DC maturation [75]	Pro-tumor	Binds VEGFR to activate PI3K-Akt and MAPK pathways
Metabolic Regulators	Arg1 (Arginase 1)	MDSCs, TAMs	Depletes arginine to inhibit T-cell proliferation; drives tumor metabolic reprogramming [75]	Immunosuppressive (indirectly pro-tumor)	Converts arginine to urea and ornithine

Metabolic reprogramming

The unique metabolic features of the TME significantly influence macrophage polarization. Hypoxia-induced HIF-1α activation promotes glycolysis, driving M2-like polarization. Additionally, the hypoxic tumor microenvironment can stimulate macrophages to secrete extracellular vesicles (EVs), especially exosomes containing hypoxia-inducible factors and metabolites [76]. These exosomes mediate metabolic reprogramming of TAMs through three core mechanisms: (1) delivering key glycolytic enzymes (such as HK2 and PKM2) to enhance aerobic glycolysis [77]; (2) transporting miRNAs regulating mitochondrial oxidative phosphorylation (such as miR-21-5p and miR-155) [78]; (3) transferring lipid metabolites to alter membrane receptor signaling [79]. This metabolic interaction promotes tumor-promoting functions such as angiogenesis, metastasis, and immune suppression [76].As critical "messengers" mediating intercellular communication within the TME, exosomes play a pivotal role in driving TAM polarization [80]. As key intercellular messengers in the TME, tumor-derived exosomes coordinate TAM polarization through metabolite exchange. Their cargoes (such as lactate and succinate) can directly inhibit AMPK while activating the HIF-1α pathway, thereby maintaining the M2-like metabolic state [81]. However, current exosome inhibition strategies lack specificity—broad-spectrum inhibitors like GW4869 simultaneously block both tumor-promoting and immune-stimulatory vesicle transport, failing to selectively reverse M2 polarization [82, 83]. In addition to promoting M2 polarization of macrophages in hypoxic conditions, exosomes can also exert this effect in normal oxygen environments by delivering a variety of molecules to macrophages [84]. Therefore, we may potentially modulate the levels of miRNAs or lncRNAs within exosomes and deliver them to TAMs via exosomal transport to exert specific functional effects.

In recent years, lactic acid, as a vital metabolite in the TME, has increasingly attracted research attention. Cancer cells undergo aerobic glycolysis through the "Warburg effect," generating substantial amounts of lactic acid, which significantly raises the concentration of lactic acid in the TME [85]. Lactic acid serves not only as an energy source for tumor cells but also influences tumor progression by modulating the functions of immune cells. Research has shown that lactic acid can promote tumor growth, invasion, and immune evasion by inducing the polarization of macrophages toward the M2 phenotype [86]. Moreover, the accumulation of lactic acid enhances Arg-1 expression via the GPR132 receptor and inhibits the functions of other immune cells. This process further facilitates the polarization of M1-type TAMs towards the M2 phenotype, thereby creating an immunosuppressive environment conducive to tumor growth [87]. Abnormal lipid metabolism (such as the endocytosis of oxLDL via CD36) and tryptophan metabolism (the kynurenine pathway mediated by IDO) both contribute to M2 polarization [88]. Recent studies have found that metabolic enzymes carried by tumor cell exosomes (such as PKM2) can be internalized by macrophages, directly altering their metabolic patterns [89].

Epigenetic modifications

The TME precisely regulates macrophage polarization through multi-layered epigenetic mechanisms [90]. At the level of histone modifications, metabolites in TME, such as lactic acid, can induce lactylation modifications like H3K18la, thereby promoting the expression of M2-related genes. Meanwhile, hypoxic conditions can increase the enrichment of repressive marks such as H3K27me3 in the promoter regions of M1-related genes in a HIF-1α-dependent manner [91, 92]. In terms of DNA methylation, factors in the TME such as TGF-β upregulate the expression of DNMTs, leading to hypermethylation in the promoter regions of M1-related genes. Meanwhile, M2 signature genes exhibit a hypomethylated state. Additionally, miR-29b carried by tumor cell exosomes can influence the DNA demethylation process by targeting TET enzymes [93, 94]. In the regulation by non-coding RNA networks, TME-specific miRNAs (such as pro-M2 miR-21/miR-146a and pro-M1 miR-155) function by regulating target genes like SOCS1/STAT3. Long non-coding RNAs (such as HOTAIR) can recruit modifying enzymes like EZH2 to form repressive chromatin structures. Circular RNAs (such as circRNA_002178) act as ceRNAs to adsorb miRNAs, thereby indirectly regulating mRNA stability [95, 96]. These mechanisms are closely intertwined. Metabolites supply substrates for modifications, signaling pathways activate epigenetic regulatory enzymes, and non-coding RNAs provide targeting specificity. Together, they form a unique epigenetic regulatory network in the TME, enabling macrophages to dynamically respond to changes in the microenvironment [97]. This multi-layered epigenetic regulation not only reveals new mechanisms by which the TME shapes an immunosuppressive niche, but also provides a theoretical basis for the development of epigenetic drugs, such as histone deacetylase inhibitors. It also highlights the need to develop personalized therapeutic strategies targeting the specific epigenetic features of different tumor types [98].

Mechanical signaling

In the TME, the increased stiffness of the ECM activates the YAP/TAZ signaling pathway through the integrin-focal adhesion kinase (FAK) pathway, thereby promoting M2 polarization [99]. Notably, studies have reported that YAP/TAZ activation not only drives macrophage polarization but may also further alter ECM composition by promoting TGF-β secretion, thereby establishing a positive feedback loop of "matrix stiffening-mechanical signal amplification [100]". Additionally, physical factors such as fluid shear stress also influence the polarization and function of macrophages [101]. Experimental studies have demonstrated that laminar shear stress suppresses the anti-tumor capacity of macrophages by activating PIEZO1 mechanosensitive ion channels on the cell membrane, which subsequently inhibits IFN-γ-mediated STAT1 phosphorylation. In contrast, turbulent shear stress activates NF-κB signaling and induces pro-inflammatory cytokine release. These findings highlight how the complexity of macrophage phenotypic outputs is determined by both the type of mechanical force (laminar vs. turbulent) and its duration of application [102]. This mechanical regulation, together with chemical signaling, forms a unique “mechano-chemical” coupling regulatory pattern in the TME [103]. The coupling between mechanical and chemical signaling extends beyond simple synergy to include competitive regulation. For instance, YAP activation driven by ECM stiffness can inhibit hypoxia-inducible factor HIF-1α degradation, leading to aberrant HIF-1α accumulation under normoxic conditions - thereby transducing mechanical stimuli into metabolic reprogramming signals (e.g., enhanced glycolysis) [103]. Conversely, HIF-1α may reciprocally augment cellular sensitivity to substrate stiffness by modulating integrin β1 expression, establishing a bidirectional mechano-metabolic crosstalk [104].

Other factors influencing macrophage polarization.

Firstly, nanocarriers, as tools for drug delivery, can also influence the polarization direction of macrophages, thereby modulating the levels of inflammation [105]. For example, drug-loaded nanoparticles can reprogram macrophages, repolarizing them from the tumor-promoting M2 state to the anti-tumor M1 state, thereby supporting other types of immunotherapy [106]. Secondly, peroxisome proliferator-activated receptor gamma (PPARγ) also plays a crucial role in the polarization of macrophages. As a key transcription factor for M2 macrophages, PPARγ influences the function of macrophages by regulating gene expression [107]. The mTOR signaling protein 6D (Sema 6D)-PPARγ axis plays a key role in macrophage polarization [108]. mTOR or Sema 6D can lead to upregulation of PPARγ expression, thereby promoting the polarization of macrophages towards the anti-inflammatory M2 phenotype. PPARγ can control the direction of macrophage polarization by promoting polarization towards the M2 phenotype and inhibiting polarization towards the M1 phenotype [109]. For example, the use of decitabine can lead to hypomethylation of RIPK3, thereby relieving its silencing effect. This results in an upregulation of the kinase's expression level while reducing the expression of PPARα and PPARγ. Consequently, the polarization state of TAMs is reversed from the M2 state to the M1 state [110].

Therapeutic strategies for macrophage polarization

Macrophage-centered immunotherapeutic strategies for cancer primarily focus on regulating the functional phenotypes of TAMs. As shown in Table 3, Current approaches mainly employ several treatment methods, including: inhibition of M2-type macrophage polarization and elimination of TAMs, metabolic reprogramming therapy, genetically engineered therapies, and other innovative strategies.

Table 3.

Therapeutic Approaches Targeting Macrophages in Tumors

Therapy Type	Representative Strategies	Mechanism of Action	Target/Signaling Pathway
Genetically Engineered Therapies	CAR-Macrophage (CAR-M) Cell Therapy	CAR-engineered macrophages target tumor antigens via phagocytosis and activate adaptive immunity [111]	HER2/CD19
Metabolic Reprogramming Therapy	Arginase-1 (Arg1) Inhibition	Targeting TAM arginine metabolism to restore T-cell function and reverse M2 polarization [3]	Arg1/urea cycle
	Pyruvate kinase M2 (PKM2) activation	Forcing macrophages into oxidative phosphorylation (OXPHOS) mode to drive M1 polarization and IL-12 secretion [3]	PKM2/HIF-1α
	siRNA delivery via nanocarriers	Selective targeting and silencing of M2-associated genes (e.g., Arg1 and CD206) to reprogram M2-polarized macrophages
Inhibiting M2 polarization or eliminating TAMs	CSF-1R inhibitor	Significant reduction in TAMs infiltration with enhanced antitumor immune response [68]	Ras/MEK1/2/ERK1/2
Innovative therapeutic strategies	Natural compounds -calciobitrin B	Specific binding to MMP12 inhibits M2-polarized TAM differentiation and tumor infiltration, effectively reducing the M2/M1 ratio within the TME [112]
	Tyrosine kinase inhibitor (TKI)	Blocking the SPP1-CD44 signaling axis to restore T cell-mediated antitumor activity [113]	SPP1-CD44 signaling axis

Inhibiting M2 polarization or eliminating TAMs

Targeting the inhibition of M2 macrophage polarization and the elimination of TAMs have become important strategies in cancer immunotherapy [53]. Currently, this goal is primarily achieved through the following approaches: First, CSF-1R inhibitors (e.g., Pexidartinib) can effectively block macrophage recruitment and survival. Preclinical studies demonstrate their ability to significantly reduce TAM infiltration and enhance anti-tumor immune responses [114]. Secondly, blocking antibodies targeting the CD47-SIRPα "don't eat me" signal (e.g., Magrolimab) can relieve tumor cell-mediated suppression of macrophage phagocytosis, enhancing macrophage-dependent tumor cell clearance [114]. Furthermore, drugs targeting the CCL2-CCR2 axis can disrupt monocyte recruitment to tumor sites, thereby reducing the replenishment of TAMs at their source [115].

Macrophage reprogramming

Macrophage reprogramming plays a dual role in maintaining health. On one hand, it is involved in regulating the development of the body, maintaining health balance, and resisting the invasion of pathogens. On the other hand, it is closely related to a variety of inflammatory diseases, including cancer, sepsis, diabetes, and others, which involve both infectious and non-infectious inflammation [116]. Reprogramming of macrophages is considered a new hope for treating most clinical diseases. Reactivating M2-type TAMs into M1-type can enhance the tumor-killing function of macrophages and effectively eliminate tumors [117]. Regarding reprogramming approaches, TLR agonists (e.g., CpG ODN) and inhibitors of STAT3/STAT6 can facilitate the transition of M2 macrophages to the M1 phenotype, thereby reinstating their anti-tumor capabilities [118, 119]. Recent studies have found that nanocarriers can deliver siRNA to specifically target and silence M2-associated genes such as Arg1 and CD206, thereby selectively reversing M2 polarization [120]. For instance, studies have shown that USP7 can reprogram macrophages into the M1 phenotype by deubiquitinating and stabilizing TRIM24, thereby enhancing the expression of SPLUNC1 and ultimately inhibiting the growth and metastasis of nasopharyngeal carcinoma [121]. For example, the β-site amyloid precursor protein-cleaving enzyme 1 (BACE1) inhibitor MK-8931 can reprogram TAMs from the M2 phenotype to the M1 phenotype by inhibiting IL-6-STAT3 signaling and enhancing the phagocytic activity of macrophages [122]. Thus, reprogramming macrophages to the M1 phenotype may be an effective immunotherapeutic strategy for cancer.

Future clinical therapeutic strategies for macrophages in cancer.

For instance, M1 polarization can enhance the antitumor activity of chimeric antigen receptor (CAR) macrophages in tumors [123]. CAR-T cell immunotherapy is a clinical treatment approach that utilizes chimeric antigen receptor-engineered T cells to specifically recognize tumor-associated antigen targets. Upon antigen engagement, intracellular signaling cascades trigger T cell activation and proliferation, ultimately leading to potent tumor cell elimination [124]. According to literature studies, chimeric antigen receptor (CAR) T-cell therapy has demonstrated potential therapeutic efficacy in patients. Firstly, CAR-T cell immunotherapy can directly eliminate tumor cells in an MHC-independent manner, thereby circumventing immune evasion during antigen presentation [125]. Secondly, the scope of application is broader. Once a CAR targeting a specific tumor antigen is constructed, it can be used to treat multiple types of tumors. Moreover, most CAR-T cells currently used in clinical trials are engineered with gene sequences that promote T cell proliferation and signal activation. This enables CAR-T cells to proliferate and acquire immunological memory after being introduced into the body, allowing them to survive for extended periods. This makes CAR-T therapy highly cost-effective when applied clinically [126]. However, the efficacy of CAR-T therapy is limited for other types of cancer, especially solid tumors [127]. Compared to blood cancers, solid tumors face more challenges, including the lack of robustly expressed tumor-exclusive antigen targets and a highly immunosuppressive and metabolically challenging tumor microenvironment. These factors limit the safety and effectiveness of the immunotherapy [128]. Poor transport and infiltration, an immunosuppressive tumor microenvironment, and the heterogeneity of tumor antigens are key factors that limit the function of CAR-T cells in solid tumors [123, 129].

CAR-M therapy is an immunotherapy that is currently being researched to determine the ideal CAR cell type for targeting solid tumors [130]. Macrophages, with their phagocytic properties, antigen-presenting capabilities, and natural ability to infiltrate the tumor microenvironment, have emerged as a viable direction for treating solid tumors [131]. CAR-M therapy can enhance antitumor efficacy by manipulating the M1-polarized state of macrophages [132]. Additionally, CAR-M therapy can be tailored to individual patients and has the potential to treat various types of cancer, including both hematological malignancies and solid tumors [133]. Significantly, macrophages are abundant in most solid tumors. They can effectively infiltrate and persist within solid tumor tissues, including those of the breast and lung [134]. These features all indicate that macrophages are more suitable as CAR carrier cells. However, the clinical translation of CAR-M therapy is currently constrained by several critical technical hurdles that must be addressed. Primary challenges include: (1) inherently low transfection efficiency and poor expansion capacity of primary macrophages during ex vivo culture, posing significant obstacles to large-scale manufacturing [135, 136]; (2) phenotypic plasticity leading to potential functional reversal in immunosuppressive tumor microenvironments, which may abrogate therapeutic effects; and (3) insufficient mechanistic understanding of CAR-M biology compared to well-characterized CAR-T cells, particularly regarding long-term safety profiles and sustained antitumor efficacy [137]. Additionally, the dense extracellular matrix in solid tumors may hinder the infiltration and distribution of CAR-M cells, affecting therapeutic efficacy. Despite these challenges, CAR-M therapy is still regarded as one of the most promising directions to break through treatment bottlenecks in solid tumors, owing to its unique biological characteris.

Furthermore, multiple studies have unveiled innovative therapeutic strategies targeting macrophages: in ovarian cancer, SPP1⁺ TAMs drive the formation of an immunosuppressive microenvironment and recruit Treg cells through the SPP1-CD44 signaling axis. The tyrosine kinase inhibitor nilotinib has been shown to disrupt this pathway by specifically inhibiting SPP1 function, thereby restoring T cell-mediated antitumor activity and opening new avenues for combination immunotherapy [113]. For hepatocellular carcinoma (HCC), the natural compound calciobitrin B specifically binds to matrix metallopeptidase MMP12, inhibiting its mediated M2-type TAM polarization and tumor infiltration. This intervention significantly reduces the pro-tumorigenic M2/M1 ratio in the TME (from 4.2:1 to 1.3:1). Clinical data analyses further reveal that high MMP12 expression correlates with TAM enrichment in HCC metastatic lesions and poor patient prognosis, underscoring its potential as a therapeutic target [112]. Furthermore, to address post-RFA(standard medical abbreviation) recurrence in HCC, researchers developed lenvatinib-loaded engineered macrophage-membrane nanovesicles (PML@Len). This innovative delivery system achieves dual functionality through surface-overexpressed PD-1 proteins: (1) evading immune clearance while blocking the PD-1/PD-L1 checkpoint, and (2) precisely releasing lenvatinib to suppress tumor angiogenesis. In animal models, PML@Len demonstrated a 68% reduction in recurrence rates while concurrently activating durable anti-tumor immune memory [138].

Conclusion and future perspectives

This study has unveiled the “three-dimensional complexity” of macrophage polarization regulation in the TME: phenotypic plasticity, metabolic plasticity, and epigenetic plasticity are all interwoven. Specifically, M1/M2 polarization is not a simple binary opposition but rather a dynamic equilibrium that exists on a continuous spectrum. This equilibrium is shaped by a combination of factors in the TME, including metabolic substrates (such as lactic acid and adenosine), hypoxic gradients, and matrix stiffness [139]. It is worth noting that mitochondrial metabolic remodeling, by regulating the ROS/HIF-1α axis, not only affects the direction of macrophage polarization but also directly participates in the regulation of tumor angiogenesis. This provides a key node for the development of "metabolism-immunity" combined intervention strategies [140]. Nowadays, research is no longer confined to the traditional cytokine regulation paradigm but is beginning to construct a more comprehensive multi-omics regulatory network model. However, further in-depth studies are still needed for the critical points of cell polarization state transitions (i.e., threshold mechanisms) and the intercellular sensing phenomena. Future research needs to break through in three dimensions: firstly, establishing a multidimensional analysis platform of "single-cell sequencing + metabolomics" to decipher the spatiotemporal dynamic atlas of TAMs subpopulations at different stages of tumor evolution. Single-cell sequencing technology can reveal the heterogeneity of macrophages in the tumor microenvironment and provides a powerful tool for in-depth research on macrophage polarization mechanisms [50]. Secondly, develop precise regulatory technologies that combine “gene editing and metabolic reprogramming.” By using CRISPR to target and modify key metabolic enzymes (such as IDH and PKM2) or epigenetic modification enzymes (such as HDAC and DNMT), it is possible to achieve targeted reshaping of the functional phenotypes of TAMs [141]; Thirdly, establish organoid-immune cell co-culture systems to simulate the metabolic characteristics of the tumor microenvironment in vitro and screen for immunometabolic regulators that can specifically reverse M2 polarization [142]. On the clinical translation front, it is recommended to conduct closed-loop studies based on the polarization status of TAMs, integrating “liquid biopsy and treatment decision-making.” By monitoring the dynamic phenotypes of peripheral blood macrophages, individualized treatment response prediction models can be established [4]. At the same time, it is necessary to integrate cutting-edge technologies such as single-cell multi-omics, organoid models, and artificial intelligence to accelerate the translation from basic research to clinical application. Through interdisciplinary innovation, precise regulatory strategies targeting the TME-macrophage axis are expected to open up new therapeutic paradigms for cancer immunotherapy and ultimately improve patient outcomes.

Author contributions

Authors' contributions Weiguo Huang：Ideas for article ideas Minzhi Peng：Writing—Original Draft Yuan Zhu、Jianping Wen：Writing—Editing Weiguo Huang：Funding acquisition Yi Hu：Revised Manuscript.

Funding

This work was supported by National Natural Science Foundation of China (81172210 and 82360528), Natural Science Foundation of Guangxi Zhuang Autonomous Region (2023GXNSFAA026021), Guangxi Degree and Graduate Education Reform Project (JGY2023188). Chinese National College Student Innovation Program (202210601021).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

The work in question has never been published before it is not under consideration for publication elsewhere; and Its publication has been approved by all co-authors (if any).

Competing interests

The authors declare no competing interests.