.
Consumption of fresh fruit and vegetables (F&V) contributes to a healthy and balanced diet. F&V provide many compounds of nutritional interest such as fiber, minerals (potassium), vitamins C, B9, E and K, provitamin A carotenoids (including β-carotene), polyphenols and glucosinolates. After harvest, F&V undergo physiological changes. These are slowed down by storage technologies such as refrigeration, possibly combined with a modified storage atmosphere. F&V shelf life remains extremely variable, ranging from a few days for fragile species to a few months for species with long-term storage potential. Nutritional compounds fulfill essential physiological or structural functions in plants. They are used, and therefore their concentration change, during storage in ways that vary depending on their type (leading to significant changes in vitamin C, for example, but far fewer in minerals or fiber). Post-harvest changes in levels of compounds of nutritional importance in F&V are just one of the factors contributing to their variability, along with environmental factors (soil, climate…) and production methods, although species and genotypes have the greatest impact on composition. Despite relative losses of certain nutrients during storage, overall levels remain sufficiently high to meet nutritional needs and provide the health benefits associated with regular consumption of F&V.
White matter hyperintensities (WMH), a key MRI-marker of cerebral small vessel disease (cSVD), are common in older adults and associated with an increased risk of stroke and dementia. The latest WMH genome-wide association study (GWAS) identified 27 loci involving genes enriched for extracellular matrix, myelination, and membrane transport. Genetically predicted WMH correlates with white matter microstructural alterations in young adults and shows causal effects on stroke and dementia. However, biological pathways underlying WMH and their contribution to clinical outcomes remain unclear, and validated polygenic risk scores (PRS) for WMH are lacking. We applied global and pathway-specific PRS (ps-PRS) to generate robust WMH-PRS and identify biological pathways contributing to WMH across the lifespan. We leveraged the largest European-ancestry WMH-GWAS (N=46,944) and data from 15,320 UK Biobank participants with MRI (UKB-MRI) (mean age=67 (7.7) years) to optimize a WMH global-PRS and construct 3,794 ps-PRS based on canonical pathways from the Molecular Signatures Database (v2023.2). The best WMH global-PRS in UKB-MRI (P=1.78 x e-193; delta-R-squared=+4.1%) predicted WMH in independent cohorts: young adults (i-Share, N=1,578, age=22(2.3); P=0.0024; delta-R-squared=+0.5%), older community-dwellers (Three City-Dijon, N=1,443, age=73 (4.1); P=1.38xe-9; delta-R-squared=+2.2%), and memory-clinic patients (Memento, N=1,831, age=71 (8.5); P=1.11xe-19; delta-R-squared=+3.3%). In the UK Biobank (N up to 355,180), WMH global-PRS was associated with incident stroke (HR=1.06 [1.038, 1.082], P=2.6xe-8), including both ischemic stroke (HR=1.061[1.037, 1.086], P=1.6xe-8) and intracerebral hemorrhage (HR=1.083[1.026, 1.143], P=0.004). Higher global-PRS was also associated with incident all-cause dementia (HR=1.051 [1.030, 1.080], P=1xe-5) and its vascular or mixed dementia sub-type (HR=1.168[1.103, 1.237], P=1.1xe-7), but showed no association with Alzheimer's disease. Permutation-based pathway enrichment, performed in UKB-MRI, identified 127 ps-PRS consistently enriched for WMH and clustering into ten biological domains. Of these, 61 ps-PRS (six clusters and 55 individual pathways) were enriched in at least one follow-up cohort: 14 in older community-persons, 17 in young adults, and 37 in memory-clinic patients. Secondary analyses highlighted four ps-PRS involved in lipid metabolism, ciliogenesis, and signal transduction enriched in both young and older adults and associated with stroke and dementia. In the memory-clinic some ps-PRS, notably involved in sphingolipid metabolism, were also associated with dementia. Seven ps-PRS, mostly lipid-related, showed evidence of modulation by hypertension. In summary, we generated a validated WMH global-PRS showing robust association with cSVD clinical complications and introduce a multi-cohort WMH ps-PRS framework that reveals candidate biological pathways with differential associations across the lifespan and clinical outcomes. These findings may inform precision prevention and drug development for cSVD.
In the living world, copper is both toxic in excess and necessary for the activity of specific oxidoreductases and electron transfer chains and as such is involved in the host-pathogen interface. Mammalian hosts deploy anti-microbial strategies of copper intoxication or starvation of invading microorganisms, collectively called nutritional immunity, and bacteria have developed both protection and acquisition systems in response. We recently described a TonB-dependent copper importer, CrtABp in the whooping cough agent Bordetella pertussis. Here we characterized another protein encoded in the same operon and similarly upregulated by copper starvation, CrpH. By combining in vitro and in vivo experiments with transcriptomics, we showed that CrpH contributes to bacterial fitness and enhances respiration by the heme-copper oxidoreductases (HCO) of B. pertussis. CrpH belongs to the PepSY_TM superfamily of membrane-associated bacterial enzymes, whose known members catalyze heme-mediated ferrisiderophore reduction in the periplasm. The corresponding heme-binding motifs of CrpH are similarly required for function. Furthermore, we uncovered a synthetic growth phenotype of a double crpH-ccoG mutant, the latter gene encoding a putative copper reductase involved in HCO assembly. In silico analyses identified thousands of CrpH orthologs, leading us to define a new subfamily of PepSY_TM proteins found in diverse bacterial species all harboring HCO genes. Collectively, our results indicate that CrpH of B. pertussis is a prototype of a family of proteins supporting HCO function.
Current evidence suggests that LLM assistance could augment the diagnostic accuracy of clinicians. However, these systems are black boxes, susceptible to hallucinations, and project a potentially misleading level of confidence. It is currently unknown whether physicians are susceptible to accepting fabricated LLM suggestions, and whether this susceptibility varies with experience. We poisoned the system prompt of an LLM-based diagnostic assistant, forcing it to suggest a fictitious disease (neurocadmiumatosis) within an otherwise legitimate differential diagnosis. Across two independent phases, 18 of 41 participants (44
