Hidden RNA Molecules Emerge as Master Regulators of Type 2 Diabetes

A sweeping review published in Discover Endocrinology and Metabolism argues that long noncoding RNAs, once dismissed as transcriptional noise, sit surprisingly high in the regulatory hierarchy that drives type 2 diabetes. These RNA molecules, longer than 200 nucleotides and lacking protein-coding capacity, appear to modulate insulin signaling, glucose and lipid metabolism, and the survival of insulin-producing beta cells. The authors, a team led by Axel Tonatiuh Marroquin-Aguilar and Héctor Iván Saldívar-Cerón, synthesized human studies published between 2015 and 2025, supplemented with experimental models where human causal evidence remains thin.

The scale of the problem they address is enormous. Type 2 diabetes currently affects more than 537 million adults worldwide and is projected to exceed 780 million cases by 2045. The disease arises from a dual defect: peripheral insulin resistance and failing insulin secretion from pancreatic beta cells, with prolonged hyperglycemia damaging the kidneys, retina, peripheral nerves and cardiovascular system. More than 98 percent of the human transcriptome does not code for proteins, and within that vast non-coding landscape, lncRNAs modulate chromatin architecture, act as transcriptional scaffolds, interact with RNA-binding proteins and serve as molecular sponges for microRNAs.

In the liver, the review highlights two contrasting players. MEG3, or Maternally Expressed Gene 3, has been associated in human and animal studies with dysregulated FOXO1 signaling and altered glucose output. MALAT1 takes a lipogenic route: in hepatocytes it stabilizes nuclear SREBP-1c, augmenting de novo lipogenesis and triglyceride accumulation. Hepatic steatosis in turn promotes insulin resistance through lipotoxic intermediates and low-grade inflammation. Experimental silencing of MALAT1, using siRNA in palmitate-treated HepG2 and primary mouse hepatocytes and in ob/ob mice, reduced hepatic triglyceride accumulation, lowered nuclear SREBP-1c levels and improved insulin resistance.

Skeletal muscle, which accounts for the majority of insulin-stimulated glucose disposal, hosts its own lncRNA story. The primate-restricted antisense transcript TMEM9B-AS1 is consistently downregulated in skeletal muscle from individuals with type 2 diabetes. Loss-of-function studies in primary human myotubes show that its depletion destabilizes MYC mRNA, impairs ribosome biogenesis and blunts protein synthesis, favoring loss of contractile proteins and myofiber atrophy, a phenotype consistent with the sarcopenia seen in long-standing diabetes. Restoration of TMEM9B-AS1 rescued translational capacity, suggesting its suppression may mechanistically link hyperglycemia to muscle protein loss. Meanwhile, the imprinted lncRNA H19 enhances oxidative metabolism and fatty-acid oxidation through PGC-1α-dependent pathways; reduced H19 in muscle biopsies from diabetic individuals has been associated with insulin resistance, while restoring H19 in db/db mice and muscle cells improved mitochondrial respiration and insulin responsiveness.

Adipose tissue adds an inflammatory dimension. SRA, enriched in adipocytes, co-activates PPARγ and promotes adipocyte differentiation and lipid storage, which under chronic caloric excess favors hypertrophic, dysfunctional depots strongly associated with systemic insulin resistance. HOTAIR, upregulated in adipose tissue from individuals with obesity and insulin resistance, epigenetically represses IRS1, attenuating the AKT–GLUT4 pathway and limiting glucose uptake; HOTAIR knockdown improved insulin responsiveness in obesity-associated insulin resistance in vivo models. NEAT1, induced by inflammatory stimuli, activates the NLRP3 inflammasome and IL-1β secretion in metabolically stressed macrophages, promoting insulin resistance in neighboring adipocytes, though much of this mechanistic detail comes from rodent and ex vivo studies.

In pancreatic islets, the review describes lncRNAs as architects of beta-cell identity. PLUTO, transcribed upstream of PDX1, facilitates long-range enhancer-promoter contacts; its expression is reduced in human islets from donors with diabetes or impaired glucose tolerance, and experimental depletion disrupts chromatin looping, lowers PDX1 and impairs insulin production. The rodent-specific βlinc1 knockdown disrupted beta-cell maturation and caused glucose intolerance in mice. βlinc3 is downregulated in human islets from diabetic donors, and suppressing its murine analogs triggered apoptosis without major changes in insulin content. MEG3 illustrates striking tissue context: deleterious in liver, it is required for normal beta-cell function, where silencing reduced PDX1 and MAFA expression, cut insulin mRNA and increased apoptosis. TUG1 knockdown in rodent and human beta cells likewise induced apoptosis and blunted glucose-stimulated insulin secretion.

The complications section may be the most clinically consequential. In diabetic kidney disease, MALAT1 is upregulated and drives inflammation, epithelial-to-mesenchymal transition and podocyte injury; silencing it in diabetic mouse models protected against albuminuria and glomerular damage. NEAT1 knockdown alleviated mesangial proliferation and extracellular matrix accumulation. PVT1, whose gene locus carries one of the strongest genetic links to diabetic nephropathy, is upregulated in diabetic kidneys, and elevated circulating PVT1 correlates with faster renal function decline. Conversely, restoring GAS5 in diabetic mice lowered blood glucose and sharply attenuated renal fibrosis, likely via miR-542-3p and ERBB4 signaling.

In the retina, MIAT acts as a sponge for miR-342-3p, derepressing caspase-1 and triggering inflammatory pyroptosis of pericytes; silencing MIAT in diabetic rodents reduced capillary degeneration and IL-1β and VEGF levels. MALAT1 binds Sp1 to increase ICAM-1 and VEGF, and intravitreal siRNA reduced vascular leakage and abnormal angiogenesis in diabetic rats. Elevated serum MIAT is associated not only with retinopathy but also with coronary artery disease and poor glycemic control. In peripheral neuropathy, a transcriptomic analysis of sural nerve biopsies showed the MALAT1/miR-1-3p/CXCR4 axis is dysregulated, and MALAT1 knockdown in diabetic mice improved nerve conduction velocity and reduced neural inflammatory markers.

The biomarker potential is quantified. A recent systematic review and meta-analysis pooling 17 human studies reported a combined area under the ROC curve of approximately 0.84, with sensitivity around 0.79 and specificity around 0.75, for circulating lncRNAs distinguishing type 2 diabetes from non-diabetic controls, dropping to roughly 0.65 for prediabetes. Multi-marker serum panels quantifying nine lncRNAs differentiated patients with and without diabetic retinopathy at AUC values in the 0.8 to 0.9 range. Higher baseline HOTAIR, PVT1 and NR2F1-AS1 have been associated with incident diabetes, retinopathy progression or faster renal decline, though most studies are small, single-center cohorts with heterogeneous effect estimates.

The authors are careful about limits. Most human evidence remains associative or transcriptomic; lncRNA-targeted therapies stay at the preclinical stage, using research-grade siRNA and shRNA tools rather than clinically optimized antisense oligonucleotides. Poor sequence conservation between rodents and humans, the primate restriction of transcripts like TMEM9B-AS1, pleiotropic actions of the same molecule across tissues, and unresolved delivery, toxicity and dosing questions all constrain translation. The review concludes that lncRNAs are best viewed today as candidate biomarkers and experimental therapeutic entry points, complementary to established measures such as HbA1c, albuminuria and retinal imaging, pending validation in large, multi-ethnic, longitudinal cohorts.

Subject of Research: The roles of long noncoding RNAs in insulin resistance, beta-cell dysfunction and diabetic complications of type 2 diabetes

Article Title: Long noncoding RNAs in type 2 diabetes mellitus: emerging regulators of insulin resistance, β-cell dysfunction and diabetic complications

Article References: Marroquin-Aguilar, A. T., Minaya-Pérez, E., Martínez-Islas, D. A., Avelino-Vivas, F., Solis-Galván, D. M., Laguna-González, A. A., García-García, B. M., Barrón-Torres, M. T., Armas-López, L., & Saldívar-Cerón, H. I. (2026). Long noncoding RNAs in type 2 diabetes mellitus: emerging regulators of insulin resistance, β-cell dysfunction and diabetic complications. Discover Endocrinology and Metabolism, 2(1), Article 5. https://doi.org/10.1007/s44417-026-00018-3

Image Credits: AI Generated

DOI: 10.1007/s44417-026-00018-3

Keywords: long noncoding RNAs, type 2 diabetes, insulin resistance, beta-cell dysfunction, diabetic nephropathy, diabetic retinopathy, diabetic neuropathy, MALAT1, biomarkers, noncoding transcriptome, lncRNA therapeutics, metabolic regulation