Age‑dependent uncoupling of ATF6α and XBP1 drives proIAPP misprocessing and amyloid nucleation via O‑GlcNAcylation of β‑cell convertases

Hypothesis

Aged human β‑cells lose identity transcription factors (PDX1, MAFA, INS) and exhibit chronic ER stress that uncouples ATF6α signaling from XBP1‑mediated transcription, while simultaneously increasing O‑GlcNAcylation of the prohormone convertases PC1/3 and PC2 due to heightened flux through the arginine biosynthesis and pentose phosphate pathways. This post‑translational modification impairs proIAPP processing, leading to accumulation of incompletely processed proIAPP that binds heparan sulfate proteoglycans (HSPGs) on the ER lumen, seeding amyloid nucleation. The resulting IAPP oligomers exacerbate ER stress, suppress autophagy via ATF4‑dependent transcriptional repression, and create a vicious cycle that drives β‑cell death and T2D progression.

Mechanistic Rationale

• ATF6α–XBP1 uncoupling: In young β‑cells, ATF6α cleavage and XBP1 splicing act cooperatively to expand the ERAD and lysosomal biogenesis programs ([3],[4]). Aging shifts ATF6α to smaller complexes and diminishes IRE1α‑dependent XBP1 splicing, breaking this coordination.
• Metabolic rewiring: Proteomics of aged islets shows up‑regulated arginine biosynthesis and pentose phosphate pathway metabolites ([2]), providing substrates for the hexosamine pathway and increased O‑GlcNAc transferase (OGT) activity.
• O‑GlcNAcylation of convertases: O‑GlcNAc modification of PC1/3 and PC2 reduces their catalytic efficiency, a mechanism previously observed in neuronal secretory pathways and here proposed for β‑cells.
• ProIAPP‑HSPG nucleation: Incompletely processed proIAPP exposes hydrophobic regions that bind HSPG heparan sulfate chains, a step accelerated by heparanase inhibition and blocked by heparanase overexpression ([6],[7]).
• Feedback to UPR: IAPP oligomers provoke PERK‑eIF2α‑ATF4 signaling, which up‑regulates CHOP and induces transcription of autophagy repressors (e.g., Rubicon), further suppressing autophagic clearance of amyloid.

Testable Predictions

1. In islets from donors >60 yr, O‑GlcNAc levels on PC1/3 and PC2 will be significantly higher than in <40 yr islets, correlating with reduced C‑peptide/IAPP ratio.
2. Pharmacological inhibition of OGT (e.g., with OSMI‑1) will restore proIAPP processing, decrease intracellular amyloid load, and increase PDX1/MAFA expression without affecting ATF6α cleavage.
3. Forced expression of a constitutively active XBP1s (spliced) in aged β‑cells will rescue autophagy flux (LC3‑II turnover) and reduce IAPP oligomer formation, even when ATF6α signaling remains uncoupled.
4. Heparanase overexpression will attenuate amyloid nucleation only when OGT activity is normalized, indicating that HSPG binding is downstream of proIAPP misprocessing.

Experimental Approach

• Obtain human pancreatic islets from organ donors stratified by age (20‑40, 40‑60, >60 yr). Measure O‑GlcNAc on PC1/3/PC2 via immunoprecipitation followed by Western blot with RL2 antibody; quantify proIAPP/IAPP ratios by ELISA.
• Treat cultured islets with OSMI‑1 (OGT inhibitor) or vehicle for 48 h; assess autophagy flux (bafilomycin A1 chase, LC3‑II/p62), ER stress markers (HSPA5, phospho‑eIF2α, ATF4), and identity genes (PDX1, MAFA, INS) by qPCR and immunofluorescence.
• Use adenoviral vectors to express XBP1s or a cleavage‑resistant ATF6α construct; evaluate amyloid deposition with Thioflavin T staining and confocal microscopy.
• Modulate heparanase levels (overexpression or siRNA) and examine amyloid formation in the presence/absence of OGT inhibition.
• Statistical analysis: ANOVA with post‑hoc Tukey; n≥6 donors per age group.

If OGT inhibition restores proIAPP processing and reduces amyloid despite persistent ATF6α–XBP1 uncoupling, the hypothesis is supported; failure to rescue processing would falsify the proposed mechanistic link.