Direct Differentiation of Human Pancreatic Islet-Derived Mesenchymal Stromal Cells into Glucose-Responsive Insulin-Producing Cells

Technical Summary: Direct Differentiation of Human Pancreatic Islet-Derived Mesenchymal Stromal Cells into Glucose-Responsive Insulin-Producing Cells

Problem Statement
Type 1 diabetes mellitus (T1D) is characterized by the autoimmune destruction of pancreatic β-cells, leading to absolute insulin deficiency. While β-cell replacement via donor islet transplantation or stem cell-derived therapies (e.g., Vertex's VX-880) represents a therapeutic goal, these approaches face significant hurdles, including limited donor availability, the need for lifelong immunosuppression, and challenges regarding the complete maturation and glucose responsiveness of generated cells. Direct lineage reprogramming of somatic cells offers a potential alternative to pluripotent stem cell differentiation, bypassing complex stepwise protocols. However, current reprogramming strategies often yield cells with incomplete β-cell maturation and limited functional glucose responsiveness. This study addresses the need for an efficient, rapid reprogramming strategy using a cell source with intrinsic pancreatic lineage predisposition: human pancreatic islet-derived mesenchymal stromal cells (hPD-MSCs).

Methodology
The researchers developed a sequential viral delivery protocol to reprogram hPD-MSCs into insulin-producing cells. The process involved two distinct phases of transduction:

1. Cell Source: hPD-MSCs were isolated from human pancreatic tail biopsies, immortalized with hTERT, and characterized for MSC markers and multipotency.
2. Phase 1 (ISP Factors): Cells were first transduced with lentiviral vectors encoding PAX6, ISL1, and a short hairpin RNA targeting REST1/2 (shREST). These factors were selected to enhance endocrine lineage specification and identity stabilization.
3. Phase 2 (PMN Factors): Twenty-four hours later, cells were transduced with an adenoviral vector encoding PDX1, NGN3, and MAFA (collectively termed PMN factors), which are critical for β-cell specialization.
4. Differentiation: Following transduction, cells were cultured for 7 days in a specific differentiation medium (NPM) containing NeuroMax, bFGF, and EGF.
5. Evaluation: Reprogramming efficiency was assessed through:
   • RT-qPCR: To measure mRNA expression of insulin (INS), NKX2.2, GCK, and other lineage markers.
   • ELISA: To quantify C-peptide secretion under basal and glucose-stimulated (5.5 mM vs. 25 mM) conditions.
   • Calcium Imaging: Using Fluo-4 AM to assess intracellular calcium dynamics in response to low glucose, high glucose, and KCl depolarization.
   • Bulk RNA Sequencing: To analyze transcriptomic shifts, gene ontology enrichment, and comparison with native human β-cells and other reprogramming datasets.

Key Results

• PMN Overexpression: Adenoviral delivery of PMN factors alone successfully induced INS and NKX2.2 expression and resulted in measurable C-peptide secretion. However, the glucose-stimulated C-peptide response was not statistically significant, and calcium imaging showed only a modest response to high glucose, indicating partial functional maturation.
• Enhancement by ISP Factors: The addition of the ISP factors (PAX6, ISL1, shREST) significantly improved reprogramming outcomes. The ISP–PMN combination resulted in:
  • Higher INS mRNA expression compared to PMN alone.
  • Significantly increased basal and glucose-stimulated C-peptide secretion.
  • Improved intracellular calcium dynamics, with a statistically significant increase in calcium influx upon high-glucose stimulation.
  • Upregulation of the glucose sensor GCK.
• Transcriptomic Profiling: RNA-seq analysis revealed that reprogrammed cells activated endocrine and neuroendocrine gene programs, including genes involved in secretory granule function, exocytosis, and stimulus–secretion coupling (e.g., VAMP1, SCG2). While mesenchymal markers were downregulated, the cells retained expression of certain stellate and progenitor markers (e.g., COL1A1, SOX9), indicating they did not fully mature into primary β-cells but rather adopted a "β-like" intermediate state.
• Comparison: The transcriptional profile of the ISP–PMN reprogrammed cells showed broad similarity to other reported β-like cell reprogramming strategies, though they remained distinct from fully mature primary human β-cells.

Key Contributions

• Cell Source Validation: The study establishes hPD-MSCs as a favorable source for endocrine reprogramming, leveraging their native pancreatic niche-related transcriptional profile and stellate cell identity to facilitate differentiation.
• Optimized Protocol: The authors demonstrate that a sequential combination of ISP factors (PAX6, ISL1, shREST) and PMN factors (PDX1, NGN3, MAFA) is superior to PMN factors alone, significantly enhancing both molecular and functional markers of β-cell identity.
• Functional Evidence: The work provides functional evidence of glucose-responsive insulin production (via C-peptide) and stimulus–secretion coupling (via calcium imaging) in reprogrammed hPD-MSCs, a critical step beyond mere gene expression.

Significance and Claims
The paper claims that this study demonstrates the feasibility of rapidly directing hPD-MSCs toward an insulin-producing, glucose-responsive phenotype using a combinatorial transcription factor approach. The authors posit that hPD-MSCs are a promising cell source for endocrine reprogramming due to their intrinsic predisposition toward endocrine differentiation.

The study concludes that while the resulting cells acquire key molecular and functional features of normal β-cells, they retain a partially intermediate transcriptional state and do not represent fully mature β-cells. The authors explicitly note that the use of viral vectors presents safety concerns for clinical translation and that future work must focus on enhancing maturation, functional stability, and safety. The findings provide a foundational basis for developing cell-based diabetes therapies but do not claim immediate clinical applicability.