Search

Folate is an important B vitamin required for methylation reactions, nucleotide and neurotransmitter synthesis, and maintenance of homocysteine at nontoxic levels. Its metabolism is tightly linked to that of choline, a precursor to acetylcholine and membrane phospholipids. Low folate intake and genetic variants in folate metabolism, such as the methylenetetrahydrofolate reductase (MTHFR) 677 C>T polymorphism, have been suggested to impact brain function and increase the risk for cognitive decline and late-onset Alzheimer's disease. Our study aimed to assess the impact of genetic and nutritional disturbances in folate metabolism, and their potential interaction, on features of cognitive decline and brain biochemistry in a mouse model. Wild-type and Mthfr +/- mice, a model for the MTHFR 677 C>T polymorphism, were fed control or folate-deficient diets from weaning until 8 and 10 months of age. We observed short-term memory impairment measured by the novel object paradigm, altered transcriptional levels of synaptic markers and epigenetic enzymes, as well as impaired choline metabolism due to the Mthfr +/- genotype in cortex or hippocampus. We also detected changes in mRNA levels of Presenillin-1, neurotrophic factors, one-carbon metabolic and epigenetic enzymes, as well as reduced levels of S-adenosylmethionine and acetylcholine, due to the folate-deficient diet. These findings shed further insights into the mechanisms by which genetic and dietary folate metabolic disturbances increase the risk for cognitive decline and suggest that these mechanisms are distinct.

Plasma microRNAs (miRNAs) have been proposed as potential biomarkers in Alzheimer's disease (AD). Here, we explored their use as early sensors of the preclinical phase of the disease, when brain pathology is being developed and no cognitive loss is detected. For this purpose, we analyzed a set of ten mature plasma miRNAs in symptomatic patients with AD from a cohort that also included healthy controls (HC) and patients with preclinical Alzheimer's disease (PAD) (cohort 1). Plasmas from subjects with Parkinson's disease (PD) were used to control for disease specificity. We found that miR-15b-5p, miR-34a-5p, miR-142-3p, and miR-545-3p levels significantly distinguished AD from PD and HC subjects. We next examined the expression of these four miRNAs in plasma from subjects with PAD. Among these, miR-34a-5p and miR-545-3p presented good diagnostic accuracy to distinguish both AD and PAD from HC subjects, according to the receiver operating characteristic (ROC) curve analysis. Both miRNAs also demonstrated a significant positive correlation with Aβ1-42 levels in cerebrospinal fluid (CSF). Taking into account the clinical potential of these findings, we decided to validate the diagnostic accuracy of miR-34a-5p and miR-545-3p in plasma samples from an independent cohort (cohort 2), in which we did not observe the alterations described by us and others in AD and PAD samples. Although miR-34a-5p and miR-545-3p might be promising early biomarker candidates for AD, our study highlights possible sources of variability in miRNA analysis across hospitals, which currently prevents their use as reliable clinical tools.