Key takeaways

  • MUSE cells (Multilineage-differentiating Stress-Enduring cells) are a naturally occurring, pluripotent-like subset of stem cells first described in adult human mesenchymal populations in 2010.
  • They carry a surface marker called SSEA-3, can give rise to cell types from all three germ layers, and in the foundational work did not form tumors, which has historically been a safety concern with more primitive pluripotent cells.
  • Once injected, MUSE cells home to damaged tissue through a signaling pathway called the S1P-S1PR2 axis, then support repair by differentiating into local cell types and by releasing protective signaling factors.
  • Most of the evidence is preclinical, from animal models of heart attack and stroke. The early human work includes a small randomized placebo-controlled stroke trial and a Phase 1 trial in newborns.

What are MUSE cells?

MUSE stands for Multilineage-differentiating Stress-Enduring cells. They are a distinct population of stem cells that occur naturally in the body, first characterized in adult human mesenchymal populations and reported in PNAS in 2010. They can be isolated from skin fibroblasts, bone marrow, or fat tissue, and are identified by a surface marker called SSEA-3. MUSE cells are described as pluripotent-like, meaning a single cell can give rise to cell types from all three germ layers, the building blocks of skin and nerve, muscle and bone, and internal organs.

One feature has drawn particular attention. In that foundational work, MUSE cells were stress-tolerant and, unlike embryonic stem cells, did not form teratomas when injected into immunodeficient mice. That absence of tumor formation matters because tumorigenicity has historically been a safety concern with more primitive pluripotent cells.

How they are thought to work

A 2018 review in Advances in Experimental Medicine and Biology framed MUSE cells as endogenous reparative stem cells, meaning the body holds its own reservoir it can mobilize after injury. In a healthy steady state, low numbers circulate from the bone marrow into the blood and out to the connective tissue of nearly every organ. After serious damage, such as a heart attack or stroke, their numbers in the bloodstream rise sharply within about 24 hours.

The homing mechanism centers on a signaling pathway known as the S1P-S1PR2 axis. Damaged tissue releases sphingosine-1-phosphate, and MUSE cells carry the matching receptor, so they are drawn preferentially toward the injured site. Once there, the same review describes two ways they contribute: they differentiate into the local cell types the tissue needs, and they exert paracrine effects, releasing anti-inflammatory, anti-fibrotic, and anti-apoptotic signals that protect and stabilize the surrounding tissue. These mechanistic findings come largely from laboratory and animal research, and the human work is still building.

Evidence from heart and stroke models

Most of the evidence base is preclinical and mechanistic. In a 2018 study published in Circulation Research, the S1P-S1PR2 homing axis was examined in a rabbit model of acute myocardial infarction (heart attack). MUSE cell treatment reduced infarct size by roughly 52 percent and raised ejection fraction by about 38 percent compared with vehicle at two months, and the allografts kept working for up to six months without immunosuppression. These findings come from a rabbit model, not from people.

The neurological work builds across several studies. A 2016 paper in the Journal of Stroke and Cerebrovascular Diseases documented MUSE cells mobilizing from the bone marrow into the blood in 29 stroke patients, evidence that the homing response described in the lab also occurs in people. A 2017 study in Stroke reported that human MUSE cells engrafted and differentiated into neurons in a subacute stroke model, reconstructing elements of neuronal circuitry. A 2020 follow-up in Stroke showed that intravenously delivered clinical-grade MUSE cells (the product CL2020) migrated to the injured area and improved recovery in a mouse stroke model. In both animal studies the cells produced no tumors over follow-up periods of up to ten months.

The human treatment data are early and limited, but they exist. A 2023 randomized, double-blind, placebo-controlled trial in the Journal of Cerebral Blood Flow and Metabolism gave a single intravenous dose of the allogeneic MUSE cell product CL2020 to patients 14 to 28 days after an ischemic stroke, without immunosuppression. At 12 weeks, 40 percent of treated patients had reached a good functional outcome (modified Rankin Scale of 2 or less) versus 10 percent on placebo, a signal the authors describe as a possible effective treatment that warrants larger study. Separately, a completed Phase 1 trial in Japan (the SHIELD study, NCT04261335) tested CL2020 in newborns with hypoxic ischemic encephalopathy to establish safety and dosing.

Other injury models under study

Heart attack and stroke are the most developed lines of MUSE cell research, but they are not the only ones. Reviews note that the same reparative behavior has been explored in other injury models, including intracerebral hemorrhage, lung injury, liver fibrosis, kidney disease, and skin and cartilage damage. The combination that drives this interest is the one described above: preferential homing to injured tissue, the ability to form cell types from all three germ layers, and a non-tumorigenic profile.

Across all of these areas the evidence is at an early stage, with consistent animal and mechanism data and the first small human trials now underway rather than settled treatments. Whether a stem cell approach makes sense for any individual is a question for a qualified physician after a full assessment.

The evidence

Selected references, each verified against primary sources (PubMed and ClinicalTrials.gov). Explore the full, filterable research library on our Science page.

MECHANISMUnique multipotent cells in adult human mesenchymal cell populations. PNAS (2010). PubMed 20421459
MECHANISMMUSE cells are a primary source of induced pluripotent stem cells in human fibroblasts. PNAS (2011). PubMed 21628574
REVIEWMuse Cells Are Endogenous Reparative Stem Cells. Adv Exp Med Biol (2018). PubMed 30484223
PRECLINICALS1P-S1PR2 axis mediates homing of MUSE cells into damaged heart after AMI. Circ Res (2018). PubMed 29475983
MECHANISMMobilization of pluripotent MUSE cells in ischemic stroke. J Stroke Cerebrovasc Dis (2016). PubMed 27019988
PRECLINICALHuman MUSE cells reconstruct neuronal circuitry in subacute lacunar stroke model. Stroke (2017). PubMed 27999136
PRECLINICALIV transplanted human MUSE cells afford brain repair in mouse lacunar stroke. Stroke (2020). PubMed 31826733
RCTRCT of CL2020 allogenic MUSE cell-based product in subacute ischemic stroke. J Cereb Blood Flow Metab (2023). PubMed 37756573
REVIEWRegenerative potential of pluripotent nontumorgenetic stem cells: MUSE cells. Regen Ther (2020). PubMed 33426206
REVIEWMUSE Cells: A New Era of Stem Cell-Based Therapy. Cells (2023). PubMed 37443710
REVIEWMUSE cells: a powerful tool for tissue damage repair. Front Cell Dev Biol (2024). PubMed 38872932
Phase 1 Trial RegistrySHIELD: phase 1 dose-escalation trial of the CL2020 MUSE cell product in newborns with hypoxic ischemic encephalopathy receiving therapeutic hypothermia (Nagoya University). Completed. ClinicalTrials.gov. NCT04261335

This article is for educational purposes only and is not medical advice, a diagnosis, or a treatment recommendation. MUSE Cells is discussed in the context of the published research; inclusion of a study does not imply a guaranteed outcome. Many of these compounds are investigational and not approved for the uses described in all jurisdictions. Any treatment decision should be made with a qualified physician. Individual results vary.