The Don’t Eat Me Signal: How CD47 Shapes Cell Clearance in Cancer, Heart Disease and the Brain

Every second, the human body quietly destroys and rebuilds itself. Billions of cells die each day through programmed cell death, and if their corpses were left to accumulate, tissues would choke on their own debris. Instead, a specialized disposal operation runs continuously: phagocytes, chiefly macrophages, patrol tissues, recognize dying cells and swallow them whole. This process, known as efferocytosis, is far more than garbage collection. It is an active immunological program that suppresses inflammation, recycles cellular components and preserves the delicate architecture of organs. A new comprehensive review published in Molecular Biology Reports by Congcong Wang, Wei Zhu and colleagues at Mudanjiang Medical University examines how a single transmembrane glycoprotein, CD47, governs this process and how its misregulation feeds into cancer, atherosclerosis, autoimmune disease and neurodegeneration.

Efferocytosis unfolds in coordinated stages. Dying cells first release find-me signals, soluble messengers such as nucleotides and lipids that lure phagocytes toward the site of death. Once in contact, the phagocyte probes the surface of the apoptotic cell for eat-me signals, most prominently phosphatidylserine, a phospholipid that flips from the inner leaflet to the outer leaflet of the membrane as a cell dies. Receptors on the macrophage surface engage these markers and trigger cytoskeletal remodeling, engulfment and degradation of the corpse inside a phagolysosome. Crucially, efferocytosis is immunologically silent or even anti-inflammatory: it drives the production of anti-inflammatory mediators such as interleukin-10 and transforming growth factor-beta, and it promotes the resolution phase that follows any inflammatory insult. When the process falters, apoptotic cells progress to secondary necrosis, spilling their contents and sustaining a chronic inflammatory fire that damages surrounding tissue.

CD47 sits at the heart of the regulatory machinery that decides whether engulfment proceeds. This widely expressed cell-surface glycoprotein binds to signal regulatory protein alpha, SIRPα, on macrophages and delivers what immunologists call the classic don’t eat me signal. When CD47 engages SIRPα, the receptor’s cytoplasmic tail recruits tyrosine phosphatases SHP-1 and SHP-2, which dephosphorylate the signaling intermediates needed for phagocytic cup formation. In effect, the macrophage’s engulfment machinery is switched off before it can engage. The system evolved as a self-preservation mechanism: healthy cells display abundant CD47 and are spared from accidental destruction. The landmark work of Oldenborg and colleagues showed that CD47 acts as a marker of self on red blood cells, and that erythrocytes lacking the protein are rapidly cleared from circulation. In this sense, CD47 functions as a passport presented at the macrophage checkpoint.

The trouble begins when pathological cells hijack the passport. Under disease conditions, abnormally high CD47 expression allows abnormal cells to evade immune clearance, and the CD47/SIRPα axis has consequently emerged as a major immune checkpoint in the regulation of efferocytosis. In tumors, the evidence is striking. Majeti and colleagues demonstrated in 2009 that CD47 is upregulated on acute myeloid leukemia stem cells and serves as an adverse prognostic factor, allowing malignant cells to slip past macrophage surveillance. Subsequent studies have traced the mechanisms of this overexpression in remarkable detail. The oncogene MYC directly regulates CD47 transcription alongside PD-L1, coupling proliferative drive with immune evasion. In breast cancer, a CD47-associated super-enhancer links pro-inflammatory signaling to CD47 upregulation, while NF-kappaB signaling drives CD47 overexpression in radioresistant HER2-positive cells. In glioblastoma, EGFR-induced phosphorylation of CD47 blocks its ubiquitination and degradation, stabilizing the protein on the tumor cell surface.

Therapeutically, the don’t eat me signal has become one of the most actively pursued targets in cancer immunology. Anti-CD47 antibodies such as magrolimab and evorpacept block the CD47-SIRPα interaction, releasing the brakes on macrophage phagocytosis and enabling tumor cell engulfment. Clinical studies in hematologic malignancies have combined CD47 blockade with azacitidine and venetoclax in acute myeloid leukemia, and with rituximab and lenalidomide in non-Hodgkin lymphoma, with phase 1 trials reporting encouraging activity. In solid tumors, the strategy is more nuanced. Glioblastoma research showed that microglia, the resident macrophages of the brain, act as effector cells when the CD47-SIRPα axis is disrupted, and that irradiation or temozolomide chemotherapy can sensitize tumors to anti-CD47 treatment. Recent work has explored dual checkpoint blockade, oncolytic viruses that downregulate CD47, HDAC inhibitors that enhance phagocytosis, and histone demethylase inhibitors that suppress CD47 expression in breast cancer, all aimed at amplifying the macrophage attack on malignant cells.

Yet the same pathway that protects tumors also drives cardiovascular disease, and here the therapeutic logic inverts. In atherosclerosis, efferocytosis is essential for clearing apoptotic foam cells within arterial plaques. When macrophages fail to engulf these dying cells, the corpses undergo secondary necrosis and form the necrotic cores that destabilize plaques and precipitate heart attacks. CD47 is upregulated in advanced lesions, partly through noncoding RNA circuits such as the lncRNA MIAT sponge that sequesters miR-149-5p and lifts CD47 expression. Kojima and colleagues reported in Nature that CD47-blocking antibodies restore phagocytosis and prevent atherosclerosis progression in mice, and subsequent studies showed that smooth muscle cell-specific CD47 deletion suppresses the disease. Age-related impairments in efferocytosis and autophagy further hinder plaque regression, and recent preclinical work suggests CD47 blockade can reprogram the monocyte-macrophage axis to promote inflammation resolution in atherosclerotic vessels.

Researchers are now engineering delivery systems to exploit this biology with precision. Biomimetic nanoparticles that boost efferocytosis through dual-pathway regulation are being developed for atherosclerosis therapy and imaging, while ROS-responsive liposomes combine CD47 blockade with lipid metabolism regulation to target inflamed plaques. Similar nanotechnology strategies are being applied to the heart: after myocardial infarction, enhanced efferocytosis of dying cardiomyocytes is linked to tissue repair, and acute CD47 blockade during ischemic reperfusion has been shown to enhance phagocytosis-associated cardiac repair in experimental models. Biomimetic nano-degraders targeting the CD47-SIRPα checkpoint and cardiac-homing nanoparticles that suppress SHP-1 are being tested as ways to promote efferocytosis and dampen the inflammation that otherwise remodels and weakens the infarcted heart.

In the nervous system, the story becomes more complicated, because CD47 plays opposing roles depending on context. Microglia, the brain’s phagocytes, must clear myelin debris, dead neurons and protein aggregates, and CD47 on myelin interacts with SIRPα to downregulate this phagocytosis, potentially impeding repair in multiple sclerosis. In experimental autoimmune encephalomyelitis, a model of multiple sclerosis, CD47 antibody blockade suppressed microglia-dependent phagocytosis and impaired recovery, illustrating the pathway’s Janus-like character in autoimmune brain inflammation. In intracerebral hemorrhage, however, CD47-blocking antibodies accelerated hematoma clearance and alleviated brain injury, including in aged animals with intraventricular hemorrhage. Emerging studies also link CD47 to microglial phagocytosis defects in autism models and to synaptic pruning in Alzheimer’s disease, where microglial clearance of amyloid deposits is a central therapeutic goal, and compounds such as luteolin have been reported to promote phagocytosis by inhibiting CD47 pyroglutamation.

The review by Wang and colleagues makes clear that CD47-mediated efferocytosis is not a simple on-off switch but a finely balanced checkpoint whose manipulation must be tailored to each disease. Blocking CD47 unleashes macrophages against tumors and atherosclerotic debris, yet the same intervention can complicate recovery in some autoimmune neuroinflammatory settings, and anemia caused by CD47’s role on red blood cells remains a safety consideration in clinical trials. The next generation of approaches, including engineered antibodies, nanoparticle delivery, combination regimens with chemotherapy, radiotherapy and other immunotherapies, and drugs that modulate CD47 expression epigenetically, aims to sharpen this therapeutic window. As the molecular circuitry of the don’t eat me signal continues to be mapped, the prospect of deliberately tuning cell clearance to resolve inflammation, dissolve plaques and expose tumors to immune destruction is moving steadily from the laboratory toward the clinic.

Subject of Research: CD47-mediated regulation of efferocytosis and its role in cancer, atherosclerosis and neurological disease

Article Title: CD47-mediated efferocytosis in diseases: A comprehensive review

Article References: Wang, C., Yu, S., Kang, J., Yang, S., Pan, Z., Zhao, Y., Li, C., & Zhu, W. (2026). CD47-mediated efferocytosis in diseases: A comprehensive review. Molecular Biology Reports, 53(1), Article 1656. https://doi.org/10.1007/s11033-026-12854-z

Image Credits: AI Generated

DOI: 10.1007/s11033-026-12854-z

Keywords: CD47, efferocytosis, SIRPα, macrophages, cancer immunotherapy, atherosclerosis, phagocytosis, immune checkpoint, neurodegeneration, glioblastoma, myocardial infarction, inflammation resolution