Finally, recalcitrant peptides are depleted in cytoplasmic and mitochondrial domains and strongly enrichedby up to 16-foldin calcium-dependent phospholipid-binding C2 domains. C2 domains mediate Ca2+-regulated binding of proteins to anionic phospholipids on membrane surfaces, (58) and their enrichment indicates that peptide persistence is disproportionately associated with proteins stabilized at membranes, rather than within freely soluble intracellular compartments. Loss of ionic homeostasis and rising intracellular Ca2+ during post-mortem decay (6) would be expected to further promote sustained membrane association of such proteins. Membrane-adjacent microenvironments are structurally constrained, concentrating redox-active substrates and catalysts (such as lipids and metals) while restricting diffusion and oxygen availability relative to the cytosol. In this context, radical oxidation is less likely to proceed via chain propagation, which requires continuous access to molecular oxygen, (59) and more likely to locally terminate via cross-linking. Such cross-linking would be expected to reduce molecular mobility and solubility, sterically hinder enzymatic and hydrolytic attack, and promote the formation of insoluble aggregates resistant to degradation. (60)

A further factor likely reinforcing this taphonomic trajectory is the brain’s unusually large and heterogeneous reservoir of redox-active iron. (61) Nervous tissue contains high iron concentrations relative to most other soft tissues, distributed across multiple pools such as heme proteins, ferritin-bound stores, and iron-rich compartments associated with mitochondria, myelin, and oligodendrocytes. (62) In life, these pools are tightly regulated to support oxidative metabolism while limiting collateral damage. (63) After death, however, progressive membrane failure and loss of regulatory control are expected to alter iron speciation and availability, increasing the likelihood of local metal-catalyzed radical generation. Importantly, such chemistry need not produce uniform oxidative destruction: when redox reactions involving iron occur within membrane-adjacent or diffusion-limited microenvironments, they may favor the formation of short-lived aromatic radicals that terminate by covalent cross-linking rather than by chain-propagating oxidation (Table S22). The enrichment of peroxidaseswhich often involve heme iron or metal cofactors (64)among recalcitrant peptides is compatible with localized, metal-associated redox chemistry, in which iron-containing cofactors may contribute to peroxide-driven radical formation without sustaining chain-propagating oxidation.

The brain is particularly predisposed to follow this trajectory. In life, it is among the most oxidatively stressed organs: it consumes a disproportionate share of oxygen, is rich in redox-active metals, and relies heavily on antioxidant and repair systems to maintain protein integrity. (65) Additionally, the brain combines extreme membrane density, (66) an abundance of structurally stable, long-lived proteins that accumulate heterogeneous oxidative modifications during life, (67) and anatomical sequestration within the cranial vault. Together, these features establish a post-mortem environment characterized by pre-existing chemical and structural heterogeneity, limited molecular mobility, and restricted oxygen exchange: conditions that favor local, diffusion-limited radical reactions and termination by cross-linking rather than runaway, chain-propagating oxidation. Notably, the molecular features that define this post-mortem pathway closely parallel those that stabilize aggregation-prone protein assemblies in neurodegenerative disease: enrichment of -sheet and structurally ordered fragments, (68) redox-active residue modifications, (69) and oxidative cross-links (70) are hallmarks of pathological protein aggregation in vivo. While the biological contexts differ fundamentally, these parallels indicate that common chemical processes govern protein persistence across clinical and geological time scales.

I know that 28 was wrong.

However, you seem to be totally ignorant with regards to algorithmic complexity. That means you do not even have the very basics needed to be in this exchange.