Scientists Built a Renewable Cancer-Fighting Cell Factory From Immune Progenitors

For decades, the cell therapy field has chased a deceptively simple goal: build an immune cell you can grow in bulk, engineer to fight cancer, freeze, ship, and use in any patient. T-cell therapies have come close for blood cancers, but solid tumors have mostly resisted them. A new approach published in Cell on June 19, 2026 by a USC Stem Cell-led team may open a different path, one that works upstream of mature immune cells entirely. The findings are early and mouse-based, but the underlying concept could reshape how researchers think about immunotherapy manufacturing.

Granulocyte-monocyte progenitors self-renew outside the body

The USC team, led by Qi-Long Ying, MD, PhD, focused on granulocyte-monocyte progenitors, or GMPs, a class of progenitor cells that normally give rise to macrophages and closely related immune cells inside the body. The conventional assumption in stem cell biology has been that long-term self-renewal is essentially the exclusive property of hematopoietic stem cells; GMPs were thought to be too far along the developmental ladder to divide indefinitely without losing their identity. The new study challenges that view directly. By developing a defined chemical cocktail, the team found they could maintain and expand mouse and human GMPs in the laboratory over the long term, preventing them from differentiating prematurely. First author Shi Yue, MD, from the Ying Lab led the bench work. Critically, collaborators from Ravi Majeti's lab at Stanford University independently reproduced both the long-term maintenance and the genetic engineering of the cells, a replication step that adds meaningful weight to the core claim.

Expanded GMPs retained their cellular and molecular identity and their ability to generate functional immune cells, meaning the expansion process did not simply produce a generic, undifferentiated mass. That preservation of identity is what makes the platform potentially useful: you still get macrophages and related cells at the end, not a developmental dead-end.

CAR-GMPs delay leukemia and HER2-positive solid tumor progression in mice

The team then pushed further, engineering the expanded GMPs with Chimeric Antigen Receptors, or CARs, the same targeting technology that powers approved CAR-T therapies. They programmed GMPs to recognize two distinct cancer markers: CD19, found on certain leukemia cells, and HER2, present on a subset of solid tumors including some breast and gastric cancers. When these CAR-GMPs were injected into mice, they engrafted into the bone marrow and generated a continuous, renewable supply of engineered immune cells rather than delivering a single wave of mature cells that would be rapidly cleared. In mouse models of both blood cancer and solid tumors, CAR-engineered GMPs delayed disease progression.

The researchers also added an additional immune-activating signal to some GMP constructs. Specifically, a CAR incorporating an immunoglobulin G Fc domain was introduced to recruit host immune cells expressing Fc receptors and to enable T-cell priming even across MHC mismatch, the immunological incompatibility that normally makes donor cells a poor fit for a recipient. GMPs engineered with both CARs and this activating signal provided an even greater benefit in cancer models. That Fc-domain addition is what enables the "off-the-shelf" concept: donor cells, even from a mismatched donor, could theoretically be used because the engineered signal helps recruit the patient's own immune system to amplify the response.

The team also tested GMPs in a different disease context entirely: mice with chronic granulomatous disease, an inherited immune deficiency that leaves patients unable to fight certain bacterial infections. Engineered GMPs restored infection-fighting ability in these mice, suggesting the platform is not narrowly limited to oncology.

How a chemical cocktail replaced stem-cell dependency

The methodological pivot here is the choice of developmental stage. Rather than starting from hematopoietic stem cells, which are scarce, difficult to engineer, and require intensive expansion protocols, the USC team worked with GMPs, which sit further downstream but, under the right conditions, can apparently be coaxed into sustained self-renewal. The defined chemical cocktail blocks premature differentiation while allowing the cells to proliferate. The study also identified myeloperoxidase as a regulator of GMP proliferation, a mechanistic detail that could inform how researchers fine-tune the expansion protocol.

This upstream strategy addresses a real bottleneck. Mature macrophages are notoriously hard to expand outside the body, genetically engineer with high efficiency, and protect from damage during the freezing and distribution process. Lab-grown human immune organoid research at Georgia Tech has similarly highlighted the difficulty of recapitulating functional immune environments outside the body. Working with progenitors rather than mature cells sidesteps several of those fragility problems at once.

Open questions about human translation and solid tumor microenvironments

The study is, for now, mouse work. The central question is whether the same expansion and engraftment dynamics will hold in human clinical settings. Challenges in solid tumor immunotherapy run deeper than cell manufacturing: antigen heterogeneity and the immunosuppressive tumor microenvironment can neutralize even well-engineered immune cells before they accomplish much. CAR-GMPs will face those same barriers in human tumors. How well the Fc-domain immune-activating signal works across the range of MHC mismatches seen in real patient populations also remains to be demonstrated outside a controlled mouse model.

Myeloperoxidase's role as a proliferation regulator is intriguing but preliminary; understanding whether it can be safely modulated without unintended effects will require follow-up work. The team also has not yet shown long-term safety data for engrafted GMPs in any model, a necessary step before human trials could begin.

GMP platform opens a path toward scalable off-the-shelf immunotherapy

"This method for the expansion and engineering of GMPs opens the door to numerous translational applications, much like T cell expansion and engineering," said Ravi Majeti, MD, PhD, of Stanford. "We have already demonstrated engineering of these cells to drive multiple potent functions, and there is a lot more to be explored."

The broader ambition is to position GMPs as a manufacturing platform analogous to what T cells became after CAR-T technology matured: a cell type you can grow at scale, engineer with precision, and eventually distribute as a standardized therapy. The future of engineered immune cell therapies is widely understood to depend on solving the off-the-shelf problem, and a progenitor-based approach that engrafts and self-replenishes represents a genuinely different angle on that challenge compared to current allogeneic T-cell strategies. Whether GMPs can clear the long runway from mouse model to approved therapy is an open question, but the platform's ability to address blood cancers, solid tumors, and infectious disease in a single framework gives researchers several directions to pursue simultaneously.

Evidence rating

INVESTIGATIONAL — 71/100

Caveats:

• All efficacy data are from mouse models; no human clinical trials have been conducted.
• Funding sources were not disclosed in the available reporting.
• Results have not yet been replicated by fully independent groups — the Stanford validation was a collaborative replication within the same study.

This is a preclinical study published in Cell demonstrating that mouse and human granulocyte-monocyte progenitors (GMPs) can be expanded long-term, genetically engineered with CAR constructs, and tested in mouse models of leukemia, solid tumors, and chronic granulomatous disease. All efficacy outcomes derive from animal experiments; no human clinical data are presented. An independent laboratory at Stanford reproduced key culture and engineering steps within the same publication, which strengthens platform validity, but fully independent replication in separate studies has not yet been reported.

References

1. New method generates renewable supply of progenitor immune cells. https://www.news-medical.net/news/20260619/New-method-generates-renewable-supply-of-progenitor-immune-cells.aspx
2. Lab-Grown Human Immune System Model Uncovers Weakened Response in Cancer Patients – Georgia Tech Research. research.gatech.edu. https://research.gatech.edu/news/lab-grown-human-immune-system-model-uncovers-weakened-response-cancer-patients
3. Ziyad M. Althafar (2025). Engineered immune cell therapies for solid tumors: pharmacological advances, clinical outcomes, and future directions. Frontiers in Pharmacology. https://doi.org/10.3389/fphar.2025.1614325
4. The future of engineered immune cell therapies. Science. https://www.science.org/doi/10.1126/scitranslmed.add2791