TUDCA and 4-PBA in Preclinical Models of Beta-Cell Secretory Failure: A Systematic Review and Bayesian Meta-Analysis.

The progressive failure of pancreatic beta-cells under chronic glucolipotoxicity drives the pathogenesis of type 2 diabetes mellitus (T2DM). This metabolic stress overwhelms the folding capacity of the endoplasmic reticulum (ER), hyperactivates the unfolded protein response (UPR), engages terminal pro-apoptotic signaling through C/EBP-homologous protein (CHOP), and promotes beta-cell dedifferentiation. In this systematic review and meta-analysis, registered with PROSPERO (CRD420261370436), we evaluated the preclinical efficacy of the low-molecular-weight chemical chaperones tauroursodeoxycholic acid (TUDCA) and 4-phenylbutyrate (4-PBA) in preserving beta-cell exocytotic identity and mitigating ER stress. Following PRISMA 2020 guidelines, a systematic search of PubMed, Scopus, and Web of Science (January 2016-May 2026) identified four eligible experimental studies. Preclinical models (INS-1 and βTC-6 cell lines, Wistar rats, and C57BL/6 mice) exposed to a high-fat diet (HFD), a high-fat/high-fructose diet (HFHFD), cholesterol loading, or protein restriction followed by high-fat feeding showed impaired or dysregulated glucose-stimulated insulin secretion (GSIS) and upregulated ER-stress markers. Co-administration of TUDCA or 4-PBA moved secretory output toward the healthy-control phenotype in every model and reduced pro-apoptotic markers in the three models in which they were measured. A hierarchical Bayesian random-effects meta-analysis of the between-arm GSIS restoration ratio at stimulatory glucose yielded a pooled ratio of 1.85 (95% credible interval [CrI] 1.38 to 2.43), with the entire credible mass above the null (posterior probability of benefit 0.996). This estimate was stable across nine prior specifications for the between-study standard deviation and in every leave-one-out analysis, including exclusion of the single hypersecretion model (1.98, 95% CrI 1.09 to 3.18). Between-study variance was small but weakly identified from only four studies and is reported as exploratory. Pooling instead on the registered within-arm stimulation-index scale, a change in metric declared as a protocol deviation, gave 1.46 (95% CrI 0.73 to 2.58) with substantial heterogeneity (I2 = 89.3%), so the evidence supports restoration of absolute glucose-stimulated insulin output rather than of fold glucose responsiveness. Because no source report documents blinding of outcome assessment, the pooled estimate should be read as an upper bound. In conclusion, TUDCA and 4-PBA act as chemical chaperones that alleviate ER stress and may prevent terminal UPR activation and preserve the beta-cell exocytotic machinery, positioning them as candidate disease-modifying agents that merit confirmatory clinical evaluation.