Hypothesis

We propose that aging is driven not by a global increase or decrease in O‑GlcNAcylation but by a loss of tissue‑specific O‑GlcNAc set points, where the optimal ratio of O‑GlcNAc‑modified to phosphorylated sites on shared serine/threonine residues diverges between metabolic and neuronal compartments. Restoring the set point in each tissue—by boosting HBP flux in neurons (via GFAT2 activation) while attenuating it in liver, adipose, and heart (via GFAT1 inhibition)—will uncouple protective O‑GlcNAc signaling from pathological HBP‑driven modifications, thereby extending healthspan independent of glucose levels.

Mechanistic Basis

• O‑GlcNAc/phosphate switch: Many signaling nodes (e.g., ERK1/2, FOXO, REST) contain serine/threonine residues that can be either phosphorylated by kinases or O‑GlcNAcylated by OGT. The outcome depends on the local flux through HBP versus the activity of kinases/phosphatases.
• Isoform‑specific GFAT regulation: GFAT1 predominates in liver, adipose, and cardiac tissue, whereas GFAT2 is enriched in brain and pancreatic β‑cells. Differential expression creates a built‑in bias for HBP flux that can be tipped pharmacologically.
• Feedback via sirtuins: NAD⁺‑dependent SIRT2 deacetylates OGT, reducing its activity; SIRT3 modulates OGA mitochondrial localization. Age‑related NAD⁺ decline therefore skews the O‑GlcNAc/phosphate balance toward phosphorylation in metabolically active tissues and toward O‑GlcNAc deficiency in neurons.
• Evidence: HBP flux drives O‑GlcNAcylation of ERK1/2, REST, and FOXO in diabetic hearts, promoting hypertrophy and fibrosis [[https://www.frontiersin.org/journals/endocrinology/articles/10.3389/fendo.2022.984342/full]]. Conversely, OGA inhibition raises O‑GlcNAc and protects against α‑synuclein aggregation [[https://pubmed.ncbi.nlm.nih.gov/41146299/]]. Healthy aging shows an early anabolic HBP shift that remodels ECM [[https://doi.org/10.1101/2023.11.17.567640]].

Predictions and Experimental Design

1. Biomarker prediction: In human blood monocytes, the ratio of O‑GlcNAc‑ERK1/2 to p‑ERK1/2 will be low in individuals with metabolic syndrome and high in those with early neurodegeneration; the ratio will predict multimorbidity better than HbA1c alone.
2. Intervention test: Aged mice will receive (a) a liver‑targeted GFAT1 antisense oligonucleotide (to lower HBP flux) and (b) a brain‑penetrant GFAT2 activator (e.g., a small‑molecule allosteric enhancer). We expect:
   • Reduced cardiac ERK1/2 O‑GlcNAcylation, lower fibrosis, and preserved ejection fraction.
   • Increased neuronal O‑GlcNAcylation of synaptic proteins, enhanced autophagic flux, and reduced neuroinflammatory markers.
   • Improved grip strength, glucose tolerance, and cognitive performance versus controls receiving scrambled oligonucleotides or vehicle.
3. Falsifiability: If dual tissue‑specific modulation fails to improve at least two of the three healthspan metrics (cardiovascular function, metabolic homeostasis, cognition) relative to single‑tissue or global HBP inhibition, the hypothesis is refuted.

Potential Implications

This model shifts the focus from "more or less O‑GlcNAc" to "right O‑GlcNAc in the right place". It suggests that combination therapies—tissue‑selective HBP modulators plus NAD⁺ boosters—could uncouple the protective and deleterious arms of the O‑GlcNAc cycle, offering a strategy to treat diabetic complications and neurodegenerative aging without relying solely on glucose control.