Macrophages do not occupy a small set of permanent boxes. They integrate developmental origin, tissue-specific signals, metabolic substrates, infection or injury cues, and interactions with neighboring cells. The result is a continuum of durable identities and reversible activation states that extends far beyond the familiar M1/M2 shorthand.

Terms such as M4, Mox, M(Hb), Mhem, M(IL-10), regulatory macrophage (Mreg), resolution-phase macrophage, and tumor-associated macrophage (TAM) arose from different experimental systems. Some describe an inducing signal, some describe a tissue or disease context, and others summarize a proposed function. They are therefore not equivalent ranks in a single taxonomy.

Creative Biolabs provides a research-oriented map of those "other" types and sub-types. It emphasizes operational definitions rather than rigid labels, distinguishes established observations from hypotheses, and connects phenotype with functional measurements.

Beyond the M1/M2 Binary: A More Useful Classification Framework

The M1/M2 model remains useful as a teaching device and as shorthand for tightly controlled in vitro stimulation. Interferon-γ plus microbial products can generate a strongly inflammatory program, whereas IL-4 or IL-13 can generate a distinct alternative activation program. In vivo, however, macrophages are usually exposed to simultaneous and changing inputs. Single-cell and spatial studies repeatedly recover macrophage clusters that co-express genes assigned to both poles or that are dominated by programs—lipid handling, iron recycling, matrix remodeling, interferon response, angiogenesis, or efferocytosis—that the binary model does not describe well.

Fig.1 The hypothesis of the existence of the M3 phenotype: the switching phenotype.^1,2

A robust classification therefore uses at least three intersecting axes. The first is ontogeny: a macrophage may derive from embryonic progenitors and self-renew locally, arise from circulating monocytes, or represent a mixture whose proportions change after injury. The second is niche: local epithelial, stromal, neuronal, vascular, and metabolic cues stabilize tissue-adapted transcriptional programs. The third is state: inflammatory mediators, immune complexes, lipids, heme, hypoxia, dying cells, drugs, or tumor-derived factors can shift function over hours to days. None of these axes alone predicts all behavior.

For reporting, stimulus-based nomenclature such as M(IFN-γ), M(LPS), M(IL-4), M(IL-10), or M(CXCL4) is often clearer than assigning a broad M-number. It records exactly what was done and makes comparisons reproducible. Tissue or disease names—Kupffer cell, microglia, alveolar macrophage, TAM—should be accompanied by origin, location, marker panel, and functional readouts. Cluster names from transcriptomic studies should likewise be treated as dataset-specific until orthogonal evidence confirms that they represent reproducible populations rather than transient states or technical partitions.

At-a-Glance Map of Selected Non-Canonical Macrophage States

The categories below are working descriptors, not mutually exclusive cell lineages. Marker expression varies by species, tissue, time point, and platform, so the table is a starting point for panel design rather than a universal gating manual.

Signal- and Substrate-Defined Macrophage States

M4 or M(CXCL4): A CXCL4-Conditioned Vascular Phenotype

The M4 designation was proposed after primary human monocytes differentiated in the presence of the platelet chemokine CXCL4 developed a transcriptome distinct from macrophages generated with M-CSF and from canonical M1- or M2-like controls. In the defining studies, CXCL4-conditioned cells lost the hemoglobin–haptoglobin scavenger receptor CD163, a change that could not simply be reversed by later exposure to M-CSF. This made M4 a useful model for asking how platelet-derived signals reshape monocyte differentiation in atherosclerotic tissue.

Mox: Adaptation to Oxidized Phospholipids

Mox macrophages were originally defined in murine systems exposed to oxidized phospholipids, especially oxidized PAPC. The state is organized around NRF2, a redox-sensitive transcription factor that induces antioxidant and detoxification genes. In the foundational model, Mox cells differed from M1- and M2-like macrophages and showed reduced phagocytic and chemotactic capacity. Mox-like cells were also identified in advanced atherosclerotic lesions, linking the in vitro stimulus to an oxidative disease environment.

Hemoglobin- and Heme-Associated States: M(Hb), HA-mac, and Mhem

Hemorrhage creates a distinctive macrophage niche. Extracellular hemoglobin binds haptoglobin and can be internalized through CD163, while heme activates detoxification and iron-recycling pathways that include heme oxygenase-1. Studies of human atherosclerotic plaques and monocyte-derived macrophages described CD163-high, HLA-DR-low hemorrhage-associated macrophages and related M(Hb) or Mhem programs. These cells may limit oxidative damage and foam-cell formation under some conditions, but later work also linked CD163-positive macrophages in hemorrhagic plaques with angiogenesis and vascular permeability. The context is therefore not uniformly protective.

M(IL-10) and Regulatory Macrophages

IL-10 can suppress inflammatory cytokine production and reconfigure antigen-presentation and metabolic programs, making M(IL-10) a useful stimulus-based description. Regulatory macrophages, often abbreviated Mreg, are a broader operational class generated by distinct protocols that can involve IL-10, immune complexes, apoptotic cells, glucocorticoids, prostaglandins, or combinations of signals. Their defining property is not simply the presence of an "anti-inflammatory" marker but a demonstrated capacity to regulate immune activation or promote tolerance.

Resolution-Phase Macrophages and Mres

Resolution is an active phase of inflammation, not a passive return to baseline. Macrophages clear apoptotic neutrophils, change mediator production, coordinate lymphocyte repopulation, and help restore tissue physiology. Macrophages isolated from resolving murine peritonitis displayed a hybrid program with features outside the conventional M1/M2 scheme. Transcriptomic studies found enrichment of antigen processing and presentation, lymphocyte-directed chemokines, proliferation-related genes, Alox15, Timd4, and Tgfb2. Other work proposed that extensive or "satiating" efferocytosis can drive a distinct Mres state with altered phagocytosis, anti-fibrotic programs, and emigration from the resolving site.

M3: A Hypothesized Switching Phenotype

M3 has been proposed as a "switch" phenotype that would respond to pro-inflammatory conditions by promoting an anti-inflammatory transition and respond to anti-inflammatory conditions by promoting a pro-inflammatory transition. The concept is intended to explain plasticity and feedback behavior that does not fit a one-directional M1-to-M2 spectrum. Unlike M4 or Mox, however, M3 was introduced principally as a hypothesis and does not have a broadly accepted induction protocol, marker set, or stable in vivo counterpart.

Disease-Conditioned Macrophage Populations

Tumor-Associated Macrophages Are an Ecosystem, Not an M2 Synonym

TAMs include macrophages that reside in or are recruited into tumors. Their origins can include pre-existing tissue-resident populations and circulating monocytes, and their programs vary with tumor type, treatment, oxygenation, vasculature, stromal composition, and anatomical territory. Many TAMs support immune suppression, angiogenesis, invasion, matrix remodeling, or resistance to therapy, but others can present antigen, produce inflammatory mediators, or support antitumor responses. Collapsing the compartment into "M2 macrophages" discards precisely the heterogeneity that matters for biomarkers and therapeutic targeting.

Single-cell studies often identify recurrent TAM programs enriched for C1Q genes, SPP1, FCN1, CCL18, interferon-stimulated genes, lipid metabolism, or proliferative signatures. These names are useful coordinates, not universal cell types. A C1QC-high cluster in one cancer may not be functionally identical to a similarly named cluster in another. Spatial data are especially important because perivascular, hypoxic, invasive-margin, necrotic, and tertiary-lymphoid-structure niches can impose different macrophage functions. Protein confirmation and perturbation are needed before a transcriptomic association is treated as a mechanism.

TAM experiments should combine identity, localization, and function. Determine whether cells are resident- or monocyte-associated where the model allows; quantify checkpoint ligands, scavenger receptors, antigen-presentation machinery, and secreted mediators; map interactions with tumor, T, NK, endothelial, and stromal cells; and measure outcomes such as phagocytosis, cytotoxic support, T-cell activation, angiogenesis, invasion, or treatment response. Therapeutic claims should identify the targeted subset and verify that the intervention does not simply deplete beneficial macrophages elsewhere.

Lipid-Associated, Fibrotic, and Interferon-Responsive Programs

Chronic disease frequently generates macrophage states named for a dominant program rather than a classic inducer. Lipid-associated macrophages, including TREM2- and lipid-metabolism-enriched populations, have been described in adipose tissue, atherosclerosis, fatty liver, and tumors. SPP1-high macrophages recur in fibrotic or remodeling environments and are often associated with extracellular-matrix interaction, though their origin and function differ by organ. Interferon-responsive macrophages express interferon-stimulated genes during infection, autoimmunity, cancer, or therapy. These programs can coexist within the same cell or change over time.

The naming rule should follow the evidence. If a cluster is defined by RNA, call it a transcriptional program until protein, spatial, and functional data support a discrete population. If a perturbation reduces an SPP1-high cluster, test whether matrix deposition or fibroblast activation also changes. If a lipid-associated state is proposed, measure lipid uptake, storage, oxidation, and efflux rather than relying on TREM2 alone. If an interferon program dominates, specify type I, type II, or mixed signaling and measure its consequences for antigen presentation and pathogen or tumor control. This discipline turns descriptive atlases into mechanistic models.

Tissue-Resident Macrophage Identities and Local Subtypes

Tissue-resident macrophages illustrate why activation labels cannot substitute for anatomical identity. Local niches establish enhancer landscapes, metabolic dependencies, self-renewal behavior, and specialized functions. Embryonic progenitors contribute substantially to several resident populations, but the degree of monocyte replacement varies by organ, age, injury, and experimental model. Within a single tissue, multiple macrophage populations may occupy distinct microanatomical locations and respond differently to disease.

Central Nervous System: Microglia and Border-Associated Macrophages

Microglia are parenchymal macrophages of the central nervous system with a developmentally and transcriptionally specialized identity supported by signals such as TGF-β and the neural environment. Border-associated macrophages occupy meninges, perivascular spaces, and the choroid plexus and should not be automatically merged with microglia. During neuroinflammation, both compartments can acquire disease-associated, interferon-responsive, lipid-handling, or proliferative programs. Markers that distinguish them in homeostasis may blur during injury, so anatomical localization and lineage-aware approaches complement flow cytometry and RNA profiling.

Liver, Spleen, Bone, and Serosal Cavities

Kupffer cells reside in hepatic sinusoids and integrate microbial products arriving from the gut with iron, lipid, and metabolic signals. Injury can recruit monocytes that generate macrophages resembling parts of the resident program without being identical to long-lived Kupffer cells. In the spleen, red-pulp macrophages recycle erythrocytes and iron, while marginal-zone and metallophilic macrophages sample blood-borne material in distinct compartments. Osteoclasts are multinucleated macrophage-lineage cells specialized for bone resorption, with differentiation controlled by M-CSF and RANKL. Peritoneal and pleural cavities contain resident macrophage populations whose abundance, origin, and activation change markedly after inflammation.

Lung: Alveolar and Interstitial Macrophage Heterogeneity

Alveolar macrophages occupy the airspace and are adapted to surfactant handling, particle clearance, and restrained immune surveillance. Interstitial macrophages reside within lung tissue and include spatially and functionally diverse populations associated with vasculature, nerves, airways, or immune aggregates. Inflammation recruits monocytes that can become macrophages and sometimes persist after recovery. Consequently, a lung digest or bronchoalveolar lavage sample captures different compartments and cannot be interpreted without sampling context.

Human alveolar studies demonstrate how rapidly a familiar resident label can subdivide when disease and time are considered. In acute hypoxemic respiratory failure, investigators combined single-cell transcriptomics and flow cytometry to define alveolar monocyte–macrophage subsets, including a matricellular program associated with SPP1, MMP7, CHI3L1, GPNMB, and PLA2G7. Such findings do not establish one permanent catalog for every lung disease; they show why sampling time, clinical context, protein validation, and functional testing must accompany cluster labels.

Designing Experiments to Identify Macrophage Sub-types

A macrophage subtype claim is most credible when it is built from convergent evidence. Begin with the biological question—origin, state transition, disease association, or therapeutic response—then choose the sampling and perturbation scheme. A panel designed only to reproduce a published label may miss the population that matters in a new species or tissue. Conversely, unconstrained clustering can generate names without a functional hypothesis. The workflow below keeps discovery and validation connected.

• Define the coordinate system. State species, tissue or cell source, anatomical compartment, developmental or disease stage, treatment, and collection time. Record whether macrophages are primary, monocyte-derived, immortalized, or engineered.
• Preserve the relevant niche where possible. Compare isolated monoculture with co-culture, matrix, organoid, or ex vivo conditions when stromal, epithelial, vascular, neuronal, or tumor signals are central to the hypothesis.
• Use a multi-marker identity panel. Combine lineage exclusion, pan-myeloid markers, tissue-associated markers, activation markers, and viability. Do not infer a subtype from CD163, CD206, HMOX1, TREM2, SPP1, or any other single marker.
• Measure function directly. Select phagocytosis, efferocytosis, antigen presentation, cytokine secretion, migration, lipid or iron handling, redox state, metabolism, matrix remodeling, microbial control, or cell–cell interaction according to the proposed biology.
• Resolve mixtures and transitions. Use single-cell, imaging, spatial, trajectory, barcoding, or time-course approaches when population averages could hide rare states or opposing responses. Validate computational clusters with an independent method.
• Build in perturbation and reversibility. Remove or add the proposed cue, block its receptor or pathway, and test whether phenotype and function change together. Rechallenge experiments reveal whether the state is stable, trained, tolerant, or readily reversible.
• Report uncertainty. Separate observations from interpretation, provide full gating and analysis criteria, and avoid presenting dataset-specific cluster names as universal cell types without cross-cohort replication.

Research Support for Complex Macrophage Phenotypes

Projects that extend beyond M1/M2 often begin with the right biological source. Creative Biolabs offers macrophage isolation and culture services for multiple human and murine sources and supports tissue-specific macrophage development when organ context is central. Model selection can be aligned with the proposed subtype, disease niche, species, sample availability, and downstream assay.

Phenotype definition can combine a customized macrophage marker development strategy with broader macrophage characterization services. Rather than relying on one surface marker, a project can integrate morphology, multi-marker flow cytometry, targeted gene expression, cytokine or chemokine output, and functional confirmation. This is particularly useful for signal-defined states such as M(CXCL4), Mox, M(IL-10), or heme-associated macrophages.

Functional validation can be selected around the biological claim. Options include quantitative phagocytosis capacity analysis, antigen-presenting capacity assessment, and cytokine expression profiling. For disease-conditioned states, co-culture and perturbation can connect macrophage phenotype with effects on tumor cells, T cells, fibroblasts, epithelium, endothelium, or extracellular-matrix remodeling.

A tailored project should start with the decision the data must support: confirming a literature-defined state, discovering heterogeneous populations, testing a pathway, comparing donors or treatments, or qualifying an assay. From that point, cell source, controls, stimulation schedule, panel composition, functional endpoints, and analysis strategy can be designed as one coherent study. Contact Creative Biolabs to discuss a macrophage phenotype workflow that fits the biological question and translational stage.

Related Products

Q & A

Q: Are M4, Mox, M(Hb), and Mreg macrophages fixed lineages?

A: Usually not. These terms generally describe macrophage states generated or enriched by particular signals or environments. The strength of the label depends on the model, induction history, marker program, and functional evidence. Tissue-resident identities may be developmentally durable, while activation states can be reversible or transitional.

Q: Can one marker identify a non-canonical macrophage subtype?

A: No single marker is sufficient. CD163, HMOX1, TREM2, SPP1, CD206, or PD-L1 can be informative, but each is shared across contexts. Use a multi-marker panel together with source, stimulus, timing, and a functional assay relevant to the proposed state.

Q: How are tumor-associated macrophages different from M2 macrophages?

A: TAM is an anatomical and disease-context term, whereas M2 refers to a simplified activation framework. Tumors contain resident- and monocyte-derived macrophages with diverse inflammatory, immunoregulatory, angiogenic, lipid-handling, interferon-responsive, and matrix-remodeling programs. Some resemble parts of an M2-like program, but TAMs should not be treated as a synonym for M2.

Q: What is the best way to study macrophage state transitions?

A: Use time-resolved sampling with a defined perturbation, then combine single-cell or high-dimensional phenotyping with protein and functional validation. Washout, rechallenge, receptor blockade, or pathway perturbation can test reversibility and causality. Lineage tracing or barcoding is valuable when origin and state change must be separated.

Q: Which assays are most useful for validating a proposed subtype?

A: Choose assays that test the defining biology. Examples include phagocytosis or efferocytosis for clearance states, T-cell co-culture for regulatory or antigen-presenting states, redox and lipid measurements for Mox-like states, heme and iron handling for M(Hb)/Mhem states, and spatial co-localization plus tumor interaction assays for TAM programs.

References

1. Malyshev, Igor, and Yuri Malyshev. "Current concept and update of the macrophage plasticity concept: intracellular mechanisms of reprogramming and M3 macrophage "switch" phenotype." BioMed research international 2015.1 (2015): 341308. https://doi.org/10.1155/2015/341308
2. Distributed under Open Access license CC BY 3.0, without modification.