In mice, pancreatic glucagon-producing cells convert into insulin producers when a single molecular lock is lifted from their RNA, without transplantation or organ injury
In mice, pancreatic glucagon-producing cells convert into insulin producers when a single molecular lock is lifted from their RNA, without transplantation or organ injury
On September 18, 2026, biologists at the Joslin Diabetes Center, Harvard Medical School, reported in Nature Metabolism on the enzyme METTL14, which maintains alpha-cell identity in the pancreas. In mice where this enzyme's gene was knocked out specifically in alpha cells, a subset of those cells spontaneously began producing insulin instead of glucagon. The authors traced the mechanism to a single nucleotide in the YY1 gene's RNA and showed that the same regulatory node operates in human cells as well.
The pancreas maintains blood sugar balance through two cell types within each islet: beta cells secrete insulin, which lowers blood sugar, and alpha cells secrete glucagon, which raises it. In 2010, Pedro Herrera's laboratory in Geneva showed that after the destruction of nearly all beta cells in mice, some alpha cells spontaneously begin producing insulin. Since then it has been known that alpha cells are capable of this switch, but what prevents it from happening in a healthy organ remained unclear.
The new study identified this brake: the enzyme pair METTL3 and METTL14 deposits the chemical mark m6A on alpha-cell RNA, and the level of this mark depends on nutrient status. Amino acids such as L-arginine and low glucose, signals that normally stimulate glucagon release, raise METTL14 and the m6A mark in a dose-dependent manner, while insulin and the fatty acid palmitate lower it. METTL3 barely responded in the same experiments, and subsequent work focused on METTL14. The cell's decision about its own identity is thus embedded in the same nutrient signal that governs hormone release.
When the researchers knocked out Mettl14 specifically in mouse alpha cells, using lineage tracing to mark cell origin, the animals lost glucagon secretion in response to amino acids, alpha-cell mass declined, and beta-cell mass increased. A subset of the new insulin-producing cells carried the alpha-cell lineage marker, meaning they arose not from donor tissue and not after organ injury, but through direct conversion in place. A similar approach has been tested with a different gene: knocking out ALDH3B2 in pancreatic duct cells also activates the insulin gene, but only after such cells are cultured outside the body and transplanted under the kidney capsule of diabetic mice. Single-nucleus sequencing showed that the conversion is incomplete: the cells lose their native genes and acquire genes associated with insulin production, but retain signs of immaturity. The authors describe this as progression toward beta-cell identity, not a finished outcome.
The authors identified the precise m6A target by overlaying the map of m6A-marked RNA sites with the list of factors that change upon loss of METTL14. The intersection yielded a single candidate, the protein YY1. Without the m6A mark, the YY1 transcript is degraded more slowly and persists longer, so the cell accumulates more of the protein. A point mutation at that nucleotide confirmed this directly. Excess YY1 suppresses the alpha-cell identity genes PAX6 and MAFB and activates the same growth signaling pathways triggered by direct loss of METTL14. Both excess and deficiency of YY1 disrupt cell identity: the protein acts as a dosage-sensitive regulator, not a simple switch.
Analysis of data from more than 200,000 human pancreatic cells confirmed this pattern: the more METTL14 an alpha cell contains, the fewer stray traces of insulin and somatostatin it shows (genes that are normally active in neighboring islet cell types). The causal experiment was performed only in mice, but this correlation suggests that the same mechanism likely operates in the human body as well.
"Targeted modulation of m6A-dependent pathways is one possible approach to regulating islet cell plasticity for regenerative purposes in diabetes," the authors write.