Two Types of Macrophages: M1 and M2 Macrophages

2. Molecular Drivers and Signaling Pathways that Affect Macrophage Polarization

3. Research Methods and Technical Advances in Macrophage Phenotypic Identification

4. Macrophages' Functional Roles in Disease Pathogenesis and Resolution

5. Therapeutic Targeting of Macrophage Polarization

1. The M1/M2 phenotype

Macrophages aren't rigidly programmed cells—instead, they have remarkable plasticity, allowing them to adjust their functions in response to local environmental signals. This functional specialization is known as polarization. Macrophages are derived from circulating monocytes or tissue-resident progenitors and are primarily divided into two archetypes: the pro-inflammatory M1 and the reparative M2 macrophages.

The M1/M2 binary model was proposed in the early 2000s, drawing inspiration from the Th1 and Th2 classification of T helper cells. M1 macrophages, often classically activated by signals like IFN-γ and lipopolysaccharide (LPS), are the body's first line of defense against pathogens and tumors. They produce high levels of pro-inflammatory cytokines such as TNF-α, IL-1β, and IL-6, generate reactive nitrogen and oxygen species (ROS), and promote Th1 immune responses ^[1,2]. They primarily orchestrate a robust inflammatory response to eliminate pathogens and damaged cells.

In contrast, M2 macrophages, alternatively activated by cytokines like IL-4 and IL-13, are involved in resolving inflammation, clearing cellular debris, and promoting tissue remodeling and repair, angiogenesis, and immunoregulation ^[1]. They express markers like CD206 and ARG-1 and secrete anti-inflammatory cytokines such as IL-10 and TGF-β ^[3,4]. The M2 phenotype can be further subdivided into: M2a, M2b, M2c, and M2d.

Figure 1. The polarization process of macrophages and its triggering targets

Table 1. Phenotypes, stimulants, biomarkers, secretions, and functions of macrophages ^[12].

Macrophage phenotype	Stimulants	Stimulants	Secretions	Functions
M1 (classically activated macrophages)	IFN-γ, LPS, TNF-α, GM-CSF	CD86, CD40, CD38, NF-κB, STAT1	TNF-α, IL-1α, IL-12, IL-23, IL-1β, IL-6, ROS, and RNS	Promote Th1 immune response, promote inflammatory response, fight pathogens, and inhibit the occurrence and development of tumors
M2a (wound healing macrophages)	IL-4, IL-13, M-CSF	CD206, IL-1R, CCL17, Fizz1, STAT6	TGF-β, IL-10, insulin-like growth factor (IGF), and fibronectin	Promote tissue repair and remodeling, promote fibrosis, and promote type II immune response by enhancing polyamines, collagen synthesis
M2b (regulatory macrophages)	Immune complex, TLR agonist, IL-1R agonist	IL-10, CCL1, LIGHT, CD86, SPHK1, TNF-α, IL-6, ERK, AP-1	Proinflamatory cytokines (IL-1β, IL-6, and TNF-α), anti-inflammatory cytokine (IL-10 and low levels of IL-12)	Involve in proinflammatory and anti-inflammatory responses, immunomodulation, and Th2 activation
M2c (acquired inactivated macrophages)	Glucocorticoids, TGF-β, IL-10	CD163, MerTK, STAT3	IL-10, TGF-β	Immune tolerance and tissue repair, suppress inflammation, promote phagocytosis and cholesterol efflux
M2d (tumor-associated macrophages)	TLR, adenosine A2A receptor γ, IL-6	VEGF-A, HIF-1α	Proteolytic enzymes (MMP-2), growth factors (VEGF), and anti-inflammatory mediators (TGF-β)	Beneficial for angiogenesis and tumor metastasis

2. Molecular Drivers and Signaling Pathways that Affect Macrophage Polarization

Macrophage polarization into classically activated (M1) or alternatively activated (M2) phenotypes is a fundamental process in immunity, inflammation, and tissue repair. This functional switch is tightly regulated by a complex network of intracellular signaling pathways [5,6], which respond to specific microenvironmental cues.

Figure 2. Factors affecting macrophage polarization

TLR/NF-κB Pathway: The NF-κB pathway is a master regulator of M1 pro-inflammatory gene expression. Its activation typically occurs through receptors such as TLR4 upon binding ligands such as LPS. This triggers a cascade that leads to the phosphorylation and degradation of inhibitory proteins like IκBα, allowing the NF-κB complex (e.g., p65) to translocate to the nucleus. In the nucleus, NF-κB drives the expression of key M1 markers, including iNOS, TNF-α, IL-6, and IL-1β ^[3,7].
JAK-STAT1 Pathway: M1 polarization is often co-driven by the JAK-STAT1 and NF-κB pathways. IFN-γ binding to its receptor activates receptor-associated JAK1 and JAK2, which phosphorylate STAT1 ^[10]. Phosphorylated STAT1 dimerizes and translocates to the nucleus. Binds to gamma-activated sequence (GAS) elements, driving the expression of a vast array of M1-associated genes such as IRF8, CIITA, and GBP.
Notch Signaling Pathway: The binding of ligands (such as Jagged1, Jagged2, DLL1, DLL4) on neighboring cells to Notch receptors (Notch1-4) on macrophages triggers the canonical Notch signal. Canonical cleavage releases NICD, which complexes with RBP-J DNA-binding protein (CSL or CBF1) to activate transcription. Amplifies TLR signaling, directly induces Nos2 transcription, and cooperates with NF-κB to enhance IL-6/IL-12 expression, thereby promoting M1 polarization ^[8]. NICD also interacts with HIF-1α and enhances glycolytic activity involved in M1 activation.
IL-4/IL-13/JAK-STAT6 Pathway: IL-4 or IL-13 binding to their receptors (IL-4Rα/γc or IL-4Rα/IL-13Rα1) activates receptor-associated JAK1/JAK3 or JAK1/JAK2/TYK2, respectively. This leads to phosphorylation and dimerization of STAT6. Nuclear p-STAT6 drives transcription of classic M2a genes, including Arg1, Fizz1 (Retnla), Ym1, and Mrc1 (CD206) ^[10]. STAT6 cooperates with the transcription factor PPAR-γ and induces IRF4, which further promotes M2a gene expression and represses M1 genes ^[1,7].
IL-10/JAK-STAT3 Pathway: IL-10 binding to its receptor (IL-10R) activates JAK1 and TYK2, leading to phosphorylation and dimerization of STAT3. STAT3 induces the transcription of M2c macrophage-associated genes such as IL-10, TGF-β, Arg1, HIF2α, and SOCS3, and suppresses pro-inflammatory cytokine production ^[9].
TGF-β/SMAD Pathway: TGF-β binding to its receptor leads to phosphorylation of receptor-regulated SMADs (R-SMADs: SMAD2/SMAD3), which complex with SMAD4 and translocate to the nucleus. Drives expression of extracellular matrix (ECM) components, fibronectin, and PAI-1, contributing to the fibrogenic and tissue-remodeling functions of M2 macrophages, particularly in wound healing and fibrosis.
PI3K/Akt/mTOR Pathway: bly activated by IL-4 via engagement of IRS-2. Positively promotes M2 polarization. Akt/mTORC1 signaling enhances the expression of M2 markers (e.g., Arg1, Mrc1) and supports the oxidative metabolic reprogramming (increased fatty acid oxidation and mitochondrial respiration) characteristic of M2 macrophages ^[10]. This contrasts with its inhibitory role in M1.
AMPK Pathway: Activated by metabolic stress (e.g., low ATP) or upstream kinases like LKB1. Promotes M2 polarization by inhibiting mTORC1 and supporting mitochondrial biogenesis and oxidative metabolism. AMPK acts as a metabolic checkpoint favoring the M2 program.
Cross-regulatory Networks: The IL-6-JAK-STAT3 pathway is associated with sustaining the M2 phenotype ^[11], while its inhibition can favor M1.

3. Research Methods and Technical Advances in Macrophage Phenotypic Identification

Research on macrophage phenotypes has moved well beyond simple M1/M2 marker panels and now integrates multi‑omic, imaging, and biophysical approaches for more precise identification.

Table 2. A summary of the macrophage phenotyping methods.

Method Category	Specific Method	Key Principles / Measured Parameters	Key Features / Applications
Classical Methods	Flow Cytometry & Immunophenotyping	Panels of surface/intracellular markers (e.g., CD64, CD86 for M1; CD206, Arg1 for M2).	Uses polychromatic panels (e.g., 11-color) with partially redundant markers for robust classification ^[13,14]; species-adapted panels validated via transcriptomics and functional assays ^[15].
	Gene Expression Assays (RT-qPCR & Bulk Transcriptomics)	RT‑qPCR is widely used to quantify hallmark cytokines and polarization markers (IL1B, IL6, IL10, NOS2, ARG1, etc.) and to validate M1/M2 states alongside protein‑level readouts ^[16].	Bulk RNA‑seq and microarray studies enabled early “spectrum” models of macrophage activation, revealing that polarization in response to diverse stimuli occupies a continuum rather than discrete M1/M2 states; network-based analyses of bulk transcriptomes have identified gene modules associated with specific activation cues, providing a framework for more nuanced classification ^[17].
	Secretome & Protein-based Assays (ELISA, Multiplex Bead Arrays, WB)	ELISA and multiplex bead arrays are routinely used to profile cytokines and chemokines (e.g., TNF, IL‑1β, IL‑6, IL‑10, CXCL10) secreted by polarized macrophages; WB and immunoblot-based workflows are used to validate newly identified protein biomarkers, such as polarization-associated membrane proteins discovered by proteomics.	Complementing flow and qPCR data; essential for phenotype verification in human and animal models ^[18].
Advanced Imaging & Biophysical Approaches	Morphology-based High-Content Imaging	Image‑based machine learning analysis of morphometric features (size, roundness, elongation, protrusions) ^[19].	Classifies M0, M1, M2 subtypes (and even M2a vs M2c subtypes) with ~93% accuracy; links morphology to functional output (e.g., IL-10 prediction).
	Membrane-Order-Sensitive Probes (e.g., Di-4-ANEPPDHQ)	A 2025 study compared RT‑qPCR, flow cytometry, and the membrane-order probe Di‑4‑ANEPPDHQ in THP‑1-derived macrophages, showing that the probe distinguishes M0, M1, and M2 phenotypes via distinct red/blue fluorescence shifts corresponding to depolarized (M1) and hyperpolarized (M2) membranes ^[16].	Integrating Di‑4‑ANEPPDHQ imaging with gene and surface marker analysis improved sensitivity and provided real‑time information on dynamic changes in macrophage activation state.
	Live-Cell Imaging & Functional Readouts	Time‑lapse microscopy, often combined with fluorescent reporters, allows direct visualization of phagocytosis, migration, and interaction with other cells ^[19].	Links phenotypic state to dynamic functional behavior; combined with single-cell tracking and machine learning to connect behavioral patterns with molecular signatures.
Single-Cell & Spatial Omics	Single-Cell RNA Sequencing (scRNA-seq)	Unsupervised identification of transcriptionally distinct macrophage subsets within tissues.	Reveals population heterogeneity beyond M1/M2 (e.g., profibrotic subsets) ^[19,20]; defines novel subpopulations.
	Spatial Transcriptomics & Multiplex Imaging	Overlays gene or protein expression data onto tissue architecture to localize macrophage subsets within specific niches, such as fibrotic regions or tumor invasive fronts.	Distinguishes resident vs. infiltrating cells in situ combined with multiplex immunofluorescence or imaging mass cytometry; crucial for understanding how local microenvironments shape macrophage states in fibrosis, cancer, and infection ^[19,20]
	Multi-Omics Integration	Integrates scRNA‑seq with surface proteomics (CITE‑seq), chromatin accessibility (scATAC‑seq), and metabolomic signatures to build multi‑layered maps of macrophage states ^[17,19].	Better capture activation networks and lineage relationships; supports data-driven phenotypic definitions.
Functional Assays	Antimicrobial & Phagocytic Assays	Phagocytosis assays using fluorescent beads, bacteria, or apoptotic cells are used to quantify uptake capacity across phenotypes ^[15]; antimicrobial functional assays (bacterial killing or parasite control readouts) have been combined with transcriptomics and flow cytometry to correlate effector functions with surface marker and gene expression profiles in species such as dog and mouse ^[19].	Correlates effector functions with molecular signatures from omics or flow cytometry data.
Functional Assays	Cytokine Production & Signaling Dynamics	Measures stimulus-induced cytokine release by ELISA or multiplex assays, combined with phospho‑flow cytometry for signaling intermediates (e.g., STAT1, STAT3, NF‑κB) ^[17,19].	Provides a dynamic perspective on responsiveness; helps distinguish transient versus stable activation states.
Emerging & Integrative Approaches	Machine Learning & Computational Phenotyping	Algorithms integrating multi-dimensional data (morphology, flow, transcriptomics) for classification ^[17,19].	Enhances classification accuracy beyond single-modality approaches; can predict functional outputs like cytokine levels ^[14].
Emerging & Integrative Approaches	Biophysical & Microenvironment-Mimicking Platforms	Microfluidic and biophysical platforms are being developed to study macrophages under controlled shear stress, stiffness, and chemokine gradients, enabling more physiologic phenotyping ^[19].	Organoid and organ‑on‑chip systems that co‑culture macrophages with tissue‑specific cells (e.g., epithelium, endothelium, tumor cells) allow to examine how complex microenvironments drive context-specific phenotypes beyond simple M1/M2 states.

4. Macrophages' Functional Roles in Disease Pathogenesis and Resolution

Macrophages are central orchestrators of the immune response, playing a dual and paradoxical role in human health and disease. Their activities span from initiating and sustaining inflammation to actively promoting its resolution and facilitating tissue repair.

4.1 Cancer

Tumor-associated macrophages (TAMs) often exhibit an M2-like, pro-tumorigenic phenotype, supporting immunosuppression, angiogenesis, and metastasis ^[21-23]. However, evidence shows cancers like triple-negative breast cancer can induce a mixed M1/M2 phenotype that still promotes aggressiveness ^[24]. Furthermore, in certain contexts like oral squamous cell carcinoma, M1-like TAMs may also contribute to tumor progression [25]. The therapeutic goal is often to repolarize TAMs from M2 to M1 to reinstate anti-tumor immunity ^[11,23].

4.2 Inflammatory & Autoimmune Diseases

In rheumatoid arthritis (RA), synovitis is characterized by a dominant M1/Th1 response that drives joint destruction via pro-inflammatory cytokines and matrix metalloproteinases ^[2]. Shifting the balance toward M2/Th2 can promote remission. Similarly, in atherosclerosis, M1 macrophages contribute to plaque instability, while M2 macrophages promote resolution ^[26,27]. Agents like ELABELA show therapeutic potential by restoring this balance ^[27].

4.3 Tissue Repair and Fibrosis

The timely transition from M1 to M2 is critical for wound healing. However, persistent M2 activation or the emergence of specific profibrotic subsets, such as CD9+TREM2+ macrophages driven by type 3 inflammation (GM-CSF, IL-17A), can lead to pathological fibrosis in organs like the liver and lung ^[28]. Conversely, in necrotizing enterocolitis (NEC), administering M2-derived nanovesicles to foster M2 polarization can resolve damaging intestinal inflammation ^[29].

4.4 Metabolic and Organ-Specific Diseases

In chronic liver diseases, omega-3 fatty acids and their specialized pro-resolving mediators (e.g., MaR1) protect by promoting a beneficial M2 polarization, reducing oxidative stress and inflammation ^[30]. In retinopathy, the natural progression involves a phase change where early M1 activity drives neovascularization, followed by a late M2 phase that promotes regression, suggesting timed therapeutic interventions ^[7].

5. Therapeutic Targeting of Macrophage Polarization

This is where fundamental research translates into clinical potential. Detail the innovative approaches being developed.

Nanotechnology for Targeted Delivery: A major challenge is selectively delivering agents to specific macrophage subsets in vivo. Nanoparticles, such as mannose-modified liposomes or ferritin nanocages decorated with targeting peptides (e.g., M2pep), can deliver siRNA (e.g., against IKKβ) or TLR agonists (e.g., CpG ODNs) to reprogram M2 TAMs into tumoricidal M1 macrophages ^[23,31].
Biological and Pharmacological Modulators: Agents derived from natural products, like the glycopeptide dCP1 from Codonopsis pilosula, which can reprogram M2-like TAMs to M1 via metabolic and signaling pathways ^[11]. Similarly, compounds like Myricetin can ameliorate atopic dermatitis by inhibiting M1 and promoting M2 polarization through JAK-STAT and NF-κB pathways ^[3].
Cell-Derived Therapeutics: Highlight novel biologics such as exosomes or nanovesicles. For example, M2 macrophage-derived nanovesicles (M2NVs) can be harvested and used as therapeutics to induce M1-to-M2 polarization and resolve inflammation, as shown in NEC ^[29]. Conversely, blocking tumor-derived exosomes carrying factors like HSP90B1 can prevent their M1-to-M2 polarizing effect and inhibit pre-metastatic niche formation ^[21].
Metabolic Reprogramming: Polarization is underpinned by distinct metabolic programs: glycolytic flux for M1 and oxidative phosphorylation for M2 ^[10,11]. Therapeutic strategies that target these metabolic pathways, such as modulating phosphate ion uptake and the AMPK-mTOR axis, are emerging ^[10].

Conclusion

Macrophages are versatile immune cells whose functional polarization into pro-inflammatory M1 or reparative M2 states is central to health and disease. This balance, driven by specific signaling pathways like JAK-STAT and NF-κB, dictates outcomes in cancer, chronic inflammation, and tissue repair. In pathologies, the equilibrium is often disrupted, for example, in tumors, where M2-like macrophages promote immunosuppression. Consequently, therapeutic strategies now aim to reprogram macrophage polarization using nanotechnology for targeted delivery, biological agents, and metabolic interventions. Looking ahead, the field is moving beyond the simple M1/M2 dichotomy. Future research will leverage single-cell analyses to decipher macrophage heterogeneity and develop context-dependent, precision therapies that dynamically modulate these critical cells to restore immune homeostasis and improve clinical outcomes.

References

[1] Yunna C, Mengru H, Lei W, Weidong C. Macrophage M1/M2 polarization [J]. Eur J Pharmacol. 2020 Jun 15;877:173090.

[2] Cutolo M, Campitiello R, Gotelli E, Soldano S. The Role of M1/M2 Macrophage Polarization in Rheumatoid Arthritis Synovitis [J]. Front Immunol. 2022 May 19;13:867260.

[3] Zhang X, Li J, Liu Y, et al. Myricetin alleviates DNCB-induced atopic dermatitis by modulating macrophage M1/M2 polarization [J]. Int Immunopharmacol. 2025 Oct 10;163:115212.

[4] Hung CH, Hsu HY, et al. Arsenic Induces M2 Macrophage Polarization and Shifts M1/M2 Cytokine Production via Mitophagy [J]. Int J Mol Sci. 2022 Nov 10;23(22):13879.

[5] Strizova, Z., Benesova, I., et al. (2023). M1/M2 macrophages and their overlaps – myth or reality [J]? Clinical Science (London, England: 1979), 137(15), 1067.

[6] Wang, N., Liang, H., & Zen, K. (2014). Molecular Mechanisms That Influence the Macrophage M1–M2 Polarization Balance [J]. Frontiers in Immunology, 5, 614.

[7] Li J, Yu S, et al. The phase changes of M1/M2 phenotype of microglia/macrophage following oxygen-induced retinopathy in mice [J]. Inflamm Res. 2021 Feb;70(2):183-192.

[8] Pagie, S., Gérard, N., & Charreau, B. (2018). Notch signaling triggered via the ligand DLL4 impedes M2 macrophage differentiation and promotes their apoptosis [J]. Cell Communication and Signaling: CCS, 16, 4.

[9] Miki, S., Suzuki, J. I., et al. (2021). S-1-Propenylcysteine promotes IL-10-induced M2c macrophage polarization through prolonged activation of IL-10R/STAT3 signaling [J]. Scientific Reports, 11, 22469.

[10] Sun X, Li Z, Wang X, He J, Wu Y. Inorganic Phosphate as "Bioenergetic Messenger" Triggers M2-Type Macrophage Polarization [J]. Adv Sci (Weinh). 2024 Apr;11(13):e2306062.

[11] Liu H, Yao M, Ren J. Codonopsis pilosula-derived glycopeptide dCP1 promotes the polarization of tumor-associated macrophage from M2-like to M1 phenotype [J]. Cancer Immunol Immunother. 2024 May 14;73(7):128.

[12] Peng, Y., Zhou, M., et al. (2023). Regulatory Mechanism of M1/M2 Macrophage Polarization in the Development of Autoimmune Diseases [J]. Mediators of Inflammation, 2023, 8821610.

[13] Liu L, Stokes JV, Tan W, Pruett SB. An optimized flow cytometry panel for classifying macrophage polarization [J]. J Immunol Methods. 2022 Dec;511:113378.

[14] Selig, M., Poehlman, L., Lang, N. C., Völker, M., Rolauffs, B., & Hart, M. L. (2024). Prediction of six macrophage phenotypes and their IL-10 content based on single-cell morphology using artificial intelligence [J]. Frontiers in Immunology, 14, 1336393.

[15] Chow, L., Soontararak, S., Wheat, W., Ammons, D., & Dow, S. (2022). Canine polarized macrophages express distinct functional and transcriptomic profiles [J]. Frontiers in Veterinary Science, 9, 988981.

[16] Siniyeh, A. A., Alshaer, W., et al. (2025). Comparative analysis of RT-qPCR, flow cytometry, and Di-4-ANEPPDHQ fluorescence for distinguishing macrophage phenotypes [J]. Biochemistry and Biophysics Reports, 44, 102225.

[17] Xue, J., Schmidt, S. V., et al. (2014). Transcriptome-Based Network Analysis Reveals a Spectrum Model of Human Macrophage Activation [J]. Immunity, 40(2), 274.

[18] Liu, B., Nguyen, P. L., Li, P., Naldrett, M. J., Alvarez, S., & Yu, J. (2025). Identification of novel protein biomarkers of macrophage polarization using comparative proteomic analyses of murine primary macrophages [J]. The Journal of Immunology Author Choice, 214(10), 2730.

[19] Vishnyakova, P., Elchaninov, A., Fatkhudinov, T., & Kolesov, D. (2025). Unravelling approaches to study macrophages: From classical to novel biophysical methodologies [J]. PeerJ, 13, e19039.

[20] Coulis, G., Jaime, D., et al. (2023). Single-cell and spatial transcriptomics identify a macrophage population associated with skeletal muscle fibrosis [J]. BioRxiv, 2023.04.18.537253.

[21] Li S, Fu X, Ning D, et al. Colon cancer exosome-associated HSP90B1 initiates pre-metastatic niche formation in the liver by polarizing M1 macrophage into M2 phenotype [J]. Biol Direct. 2025 Apr 15;20(1):52.

[22] Wang Q, Ma W. Revisiting TAM polarization: beyond M1- and M2-type TAM toward clinical precision in macrophage-targeted therapy [J]. Exp Mol Pathol. 2025 Sep;143:104982.

[23] Shan H, Dou W, Zhang Y, Qi M. Targeted ferritin nanoparticle encapsulating CpG oligodeoxynucleotides induces tumor-associated macrophage M2 phenotype polarization into M1 phenotype and inhibits tumor growth [J]. Nanoscale. 2020 Nov 12;12(43):22268-22280.

[24] Pe KCS, Saetung R, et al. Triple-negative breast cancer influences a mixed M1/M2 macrophage phenotype associated with tumor aggressiveness [J]. PLoS One. 2022 Aug 12;17(8):e0273044.

[25] Li CX, Gong ZC, Yu JW. Deliberation concerning the role of M1-type macrophage subset in oral carcinogenesis [J]. J Exp Clin Cancer Res. 2024 Aug 8;43(1):220.

[26] Shou X, Wang Y, Jiang Q, Chen J, Liu Q. miR-126 promotes M1 to M2 macrophage phenotype switching via VEGFA and KLF4 [J]. PeerJ. 2023 Mar 31;11:e15180.

[27] Tang L, Yi X, Tan W, et al. ELABELA Ameliorates Atherosclerosis Through Restoring the M1/M2 Macrophage Balance in ApoE(-/-) Mice [J]. J Am Heart Assoc. 2025 Jul 15;14(14):e041261.

[28] Fabre T, Barron AMS, et al. Identification of a broadly fibrogenic macrophage subset induced by type 3 inflammation [J]. Sci Immunol. 2023 Apr 14;8(82):eadd8945.

[29] Liu, C., Gong, J., et al. (2025). M2-type macrophage nanovesicles regulate the inflammatory response after necrotizing enterocolitis by inducing M1 to M2-like macrophage polarization [J]. Frontiers in Cellular and Infection Microbiology, 15, 1664897.

[30] Videla LA, Valenzuela R, Del Campo A, Zúñiga-Hernández J. Omega-3 Lipid Mediators: Modulation of the M1/M2 Macrophage Phenotype and Its Protective Role in Chronic Liver Diseases [J]. Int J Mol Sci. 2023 Oct 24;24(21):15528.

[31] Zhang Y, Chu B, Fan Q, Song X, Xu Q, Qu Y. M2-type macrophage-targeted delivery of IKKβ siRNA induces M2-to-M1 repolarization for CNV gene therapy [J]. Nanomedicine. 2024 Apr;57:102740.

[32] Wang, P., Li, Z., Song, Y., Zhang, B., & Fan, C. (2025). Resveratrol-driven macrophage polarization: Unveiling mechanisms and therapeutic potential [J]. Frontiers in Pharmacology, 15, 1516609.