Lukiw and colleagues performed Western blotting (anti-E. coli LPS antibody ab35654, normalized to β-actin, densitometry in ImageQuantTL) and confocal immunohistochemistry on postmortem human brain with PMI ≤3.8 h, reporting LPS at ~37 kDa: >7-fold enrichment in AD temporal-lobe neocortex and >21-fold in AD hippocampus versus age- and gender-matched controls (ANOVA, p<0.01) . Immunohistochemistry with DAPI counterstain showed perinuclear LPS–nuclear-envelope association in AD, with >75% of LPS signals associated with nuclei versus random association in controls . The work builds on a prior companion study detecting LPS in AD hippocampus and on an independent Neurology report of Gram-negative bacterial molecules co-localizing with AD pathology .
Reported observations: antibody-detected LPS signal enrichment and perinuclear clustering. Author interpretations: that the LPS originates from GI-tract Gram-negative bacilli, "leaks" across GI and blood–brain barriers, and may block mRNA export through nuclear pores . BGPT inference: the gut-origin and nuclear-pore-export claims are not directly tested here — they are conjectures layered on a correlative staining result. Key blind spots the paper does not fully resolve: (i) the anti-E. coli LPS antibody may cross-react with host or other bacterial epitopes; no mass spectrometry or structural (lipid A) confirmation is provided; (ii) all-female samples limit generalizability and the Western-blot N per group is 3–4; (iii) postmortem contamination or PMI-related microbial translocation (the paper itself invokes the thanatomicrobiome) is not ruled out by culture-independent controls; (iv) it is a Perspectives article — much of the mechanistic edifice cites the authors' own work and "unpublished observations," a self-citation and transparency weakness; (v) no data-availability statement and no independent-lab replication within this paper; (vi) 16S-based bacterial sequences in AD brain similarly cannot exclude contamination. The paper does acknowledge several of these (small N, unknown significance of perinuclear clustering) but not the antibody-specificity or contamination problems.
The core enrichment finding would be falsified or weakened if blinded, structurally confirmed (MALDI/MS lipid A) assays in larger mixed-sex cohorts found no AD-control difference, or if nuclear-envelope LPS localization disappeared with rigorous contamination controls. The causality claim (gut-derived LPS driving neurodegeneration) would require interventional data — e.g., germ-free versus colonized amyloid-model mice — which this paper does not supply. Confidence in the descriptive finding: low–moderate; in the mechanistic narrative: low.
Know what changed, what holds up, and what remains uncertain. Every Friday. No ads.