The Canalised Cleavage of Glucosepane

We report two mechanistically distinct catalytic chemistries for cleavage of glucosepane, the most abundant protein–protein advanced glycation end-product (AGE) crosslink of aged human extracellular matrix. Glucosepane exceeds every other AGE crosslink in collagen and elastin by ~10–100-fold and accumulates progressively throughout life. Beyond the commonly cited tissue-mechanical consequences — arterial stiffening (the strongest independent predictor of cardiovascular mortality in the elderly), renal basement-membrane thickening, reduced pulmonary and joint compliance, and the accelerated vascular aging of diabetes [6–9] — AGE crosslinking of basement-membrane type IV collagen and laminin disrupts integrin-mediated cell–matrix engagement, impairs endothelial adhesion, and drives a premature senescence-like endothelial phenotype (Haitoglou et al., J Biol Chem 1992 [78]; Chen et al., Circ Res 2002 [79]); AGE accumulation and RAGE-mediated signalling at the neurovascular unit are associated with blood–brain-barrier breakdown, pericyte injury, and the cognitive decline of aging and Alzheimer’s disease (Deane et al., Nat Med 2003 [80]; Montagne et al., Neuron 2015 [81]; Sweeney et al., Nat Rev Neurol 2018 [82]). Yet no peer-reviewed chemical cleavage method has been reported in the two decades since the crosslink’s structural identification (Sell et al., J Biol Chem 2005 [4]) nor in the decade since its total synthesis (Draghici, Wang, and Spiegel, Science 2015 [10]); every clinical AGE-breaker (alagebrium, aminoguanidine, pyridoxamine, carnosine) is mechanistically incompatible with glucosepane’s bicyclic aminal–amidine core. The two chemistries converge on a shared ring-opened iminium intermediate. Strategy I is thermal: Lewis acid and mild-acid catalysis raise the ring-opening equilibrium of the bicyclic aminal, and an α-effect nucleophile traps the transient intermediate irreversibly into a thermodynamic sink. Strategy II is photoredox: visible-light single-electron oxidation generates the lysine-side N1 radical cation, and β-scission of the adjacent strained C–N bond produces an irreversible ring-opened fragment. Because each strategy's principal parameter weakness is offset by the other, and because the radical cation required for Strategy II is also accessible by electrochemistry, sonodynamics, and near-infrared activation (Section 6), the architecture tolerates parameter failure in any single channel; kinetic selectivity follows from strained-heterocycle-specific radical fragmentation (Section 4). Aged collagen matrix itself contributes four independent factors that further favor the intervention at the lesion site (Section 5). Each chemical element is independently falsifiable within weeks of bench work at reagent costs of thousands of dollars, with tiered tests and explicit kill criteria specified (Section 9; SI Table S4). Intrafibrillar delivery is addressable through the anti-glucosepane Glu3 DNA aptamer (Li et al., Nanoscale Horizons 2025 [74]). Should the chemistry withstand the proposed tests, the result will be the first chemistry-based cleavage of the dominant human AGE crosslink and the first demonstration that a canalised chemistry can address a class of accumulated protein lesions for which no repair pathway — endogenous or synthetic — has previously been established.