Engineered Stem Cell Therapy Reverses New-Onset T1D in Mice by Reprogramming T Cells

According to authors of a study, a single infusion of mesenchymal stem/stromal cells (MSCs) engineered to overexpress alpha-1 antitrypsin (AAT) reversed new-onset type 1 diabetes (T1D) in more than half of treated female nonobese diabetic (NOD) mice. The findings were published in Molecular Therapy by researchers at the Medical University of South Carolina (MUSC). Using single-cell RNA sequencing, flow cytometry, and functional assays, the investigators found that the therapy works mainly by strengthening Tregs and driving autoreactive CD8+ T cells into an exhausted state.¹

Current T1D management relies on insulin replacement, which controls glucose but does not halt autoimmune destruction of pancreatic β cells and carries risks such as hypoglycemia.¹

“To impact or cure T1D, the stem cells themselves don't need to be there,” lead investigator Hongjun Wang, PhD, coscientific director of the MUSC Center for Cellular Therapy, said in a news release. “This means that when you administer the cell therapy to patients, the effect can last 6 months to 2 years, as seen in clinical trials using MSCs for a range of diseases.”²

He noted that the MSCs were cleared from the body within days, but the immune effects persisted. Wang states his belief is that this is because AAT-MSCs secrete microscopic factors that continue to protect the body’s organs long after the cells disappear.

Why Engineer MSCs to Express AAT?

MSCs have established immunomodulatory and tissue-repair properties. In 2 clinical trials, they showed preliminary safety and efficacy in new-onset T1D, and the FDA approved remestemcel-L (Ryoncil; Mesoblast), an allogeneic bone marrow–derived MSC therapy, for steroid-refractory graft-vs-host disease (GvHD) in pediatric patients.¹ However, the inflammatory environment of established T1D can overwhelm unmodified MSCs before they reverse the disease.²

AAT is a serine protease inhibitor with anti-inflammatory and antiapoptotic effects. In earlier NOD mouse studies, AAT reduced insulitis and protected β cells. The investigators transduced human bone marrow–derived MSCs with a lentiviral vector to overexpress AAT. The authors wrote that, in their earlier work, these AAT-MSCs outperformed unmodified MSCs in preventing T1D onset and protecting against GvHD in mice.¹

Diabetes Reversal and Reduced Insulitis

Female NOD mice with new-onset diabetes received a single intravenous dose of 1 × 10⁶ AAT-MSCs (n = 18) or no treatment (n = 11). By week 5, approximately 61.1% (n = 11) of the treated mice were diabetes-free, compared with none of the controls. At week 10, approximately 54% (n = 7) of the remaining treated mice were still diabetes-free, after 5 had been sacrificed for analysis. None of the control mice achieved remission.¹

Treated mice were also observed to have less islet infiltration. Among controls, approximately 28.2% of islets had less than 5% immune cell infiltration and 50.6% had more than 75%. In the treated group, those proportions were 51.1% and 26.2%, respectively. Pancreatic lymph node (PLN) cells from treated mice produced lower levels of proinflammatory cytokines, including interferon-γ (IFN-γ), interleukin-6, and tumor necrosis factor α.¹

Stronger Tregs, Fewer Th1 Cells

AAT-MSC treatment did not change the overall frequency of Tregs in the PLNs. Instead, it enriched subpopulations associated with stronger suppressive function, increasing Helios+ Tregs from approximately 25.27% in controls to 33.43% (P = .02) and CTLA4+ Tregs from 30.50% to 49.17% (P = .047).¹

In a functional assay, stimulated T cells proliferated at a rate of 85.50%. Adding Tregs from untreated mice reduced proliferation to 35.50%, and Tregs from AAT-MSC–treated mice reduced it further, to 25.70%. The therapy also reduced Th1 responses. CD4+IFN-γ+ cells fell from 5.52% to 4.30% (P = .04), with no significant change in Th17 cells. Single-cell analysis demonstrated that IFN-responsive CD4+ T cells in the PLNs came mostly from the control mice (80%) rather than ones that were treated (20%).¹

The effect extended to human cells. The study demonstrated that when peripheral blood mononuclear cells from healthy donors were cultured with AAT-MSCs, the CD4+CD25+CD127low Treg population rose from 4.09% to as high as 6.72%.¹

Driving CD8+ T Cells Toward Exhaustion

CD8+ T cells are the dominant immune cells infiltrating islets in T1D and directly kill β cells. Exhausted CD8+ T cell signatures have been linked to slower disease progression and to clinical response to immunotherapies such as teplizumab (Tzield; Sanofi). Islets from treated mice contained a higher proportion of Tox+ exhausted CD8+ T cells. In coculture, AAT-MSCs increased both progenitor-exhausted and terminally exhausted CD8+ T cells, identified by markers including PD-1, Tim-3, Tox, and TIGIT.¹

The investigators then tested whether Tregs mediated this effect. Tregs that were “educated” by AAT-MSCs induced the most CD8+ T cell exhaustion, whereas non-Tregs had little effect on their own. Islets that were cocultured with educated Tregs and CD8+ T cells showed significantly less cell death than islets exposed to CD8+ T cells alone. Cell–cell communication analysis supported these findings: Predicted interactions increased from 75 to 104 in the PLNs and from 2211 to 2593 in the islets, with the largest gains in signaling originating from Tregs.¹

Limitations and Next Steps

The authors noted several limitations, including the following: only female NOD mice were studied, meaning that sex-specific differences in response were not evaluated; ATT levels were not measured in vivo; the study’s design, which did not isolate the contribution of the AAT modification from that of the MSCs themselves; and the analysis’s focus on T cells, which excluded other immune populations such as antigen-presenting cells.¹

The authors emphasized that as an engineered cell therapy, AAT-MSCs would need to meet FDA requirements for chemistry, manufacturing, and controls, as well as potency and long-term safety assessment, before this approach can be used clinically. More research is needed before AAT-MSCs can enter clinical trials.¹

Separately, Wang’s research team has reported testing unmodified MSCs in a clinical trial of patients with new-onset T1D (NCT04061746). The group is also exploring the approach in other autoimmune and inflammatory conditions, such as lupus and chronic pancreatitis.²

REFERENCES

1. Wei H, Gou W, Kim J, et al. Taming autoimmunity: alpha-1 antitrypsin overexpressing mesenchymal stromal cells promote regulatory T cell crosstalk to reverse diabetes. Mol Ther. 2026;34(7):4181-4199. doi:10.1016/j.ymthe.2026.03.032

2. Scheldrup G. MUSC researchers develop stem cell therapy that shows promise in one day reversing type 1 diabetes. News release. Medical University of South Carolina. June 2, 2026. Accessed September 25, 2026. https://www.musc.edu/content-hub/News/2026/06/02/promising-stem-cell-therapy-for-diabetes