Establishing causality for protein lactylation in Alzheimer's disease: site validation, evidence standards, and...

Protein lysine lactylation has emerged as a candidate interface between metabolic stress and Alzheimer's disease (AD), but the strength of individual claims varies. This focused review evaluates whether reported histone and non-histone lactylation events meet analytical and experimental requirements for causal interpretation. Direct AD evidence is more robust for plaque-associated microglial histone H4 lysine 12 lactylation (H4K12la)-pyruvate kinase M2 (PKM2) feedback in an amyloid model. Histone H3 lysine 18 lactylation (H3K18la)-nuclear factor-κB (NF-κB) signaling is supported principally in senescent microglia and ageing or AD-related models. Human tau K331 lactylation demonstrates disease-associated occurrence but not quantitative occupancy, temporal precedence or residue-specific necessity. Amyloid precursor protein (APP) K612 lactylation provides a potentially protective mechanism supported by site-resolved identification, cellular perturbation and in vivo rescue. Histone H3 lysine 9 lactylation (H3K9la) and additional substrates remain hypothesis-generating. We distinguish putative enzymatic L-lactylation from methylglyoxal-derived D-lactoylation and define minimum standards for antibody specificity, stereochemistry, site localization, occupancy, protein abundance, cell composition, post-mortem quality, isotope tracing, genetic substitution and rescue. A four-tier framework separates extrapolated evidence, human occurrence or relevant in vivo association, site-specific AD-model mechanisms supported by intervention and rescue, and quantified, cell-resolved, residue-causal evidence independently replicated in human-relevant systems. Six experimental stages are proposed for advancing a candidate site from detection to human relevance. Current evidence supports protein lactylation as a context-dependent disease-modifying layer rather than a single upstream driver. Therapeutic development should prioritize validated sites and biomarker-defined cell states rather than indiscriminate suppression of lactate metabolism or global lactylation.