Several micro(mi)RNA are deregulated in brain, cerebrospinal fluid (CSF), and serum/plasma from patients with Alzheimer's disease (AD). The aim of the study was to profile circulating miRNAs in serum as non-invasive biomarkers for AD, correlating them with those identified in CSF, the biological fluid which better reflects biochemical changes occurring during pathological processes in the brain and may provide a robust indicator of AD-related disease pathogenesis thanks to the evidence of low amyloid and high levels of tau and hyperphosphorylated tau. Using a two-step analysis (array and validation through real-time PCR), a down-regulation (mean fold change ± SEM) of miR-125b (0.415 ± 0.11 versus 1.381 ± 0.36, p = 0.009), miR-23a (0.111 ± 0.03 versus 0.732 ± 0.14, p < 0.001), and miR-26b (0.414 ± 0.11 versus 1.353 ± 0.39, p < 0.01), out of 84 tested, was shown in serum from 22 AD patients compared with 18 non-inflammatory and 8 inflammatory neurological controls (NINDCs and INDCs) and 10 patients with frontotemporal dementia. Significant down-regulation of miR-125b and miR-26b was also confirmed in CSF from AD patients versus NINDCs (miR-125b: 0.089 ± 0.03 versus 0.230 ± 0.08, p < 0.001; miR-26b: 0.217 ± 0.06 versus 1.255 ± 0.29, p < 0.001, mean fold change ± SEM, respectively), whereas data were not replicated for miR-23a. In serum, miR-125b had an AUC of 0.82 to distinguish AD from NINDCs (95% CI: 0.65-0.98, p = 0.005). In conclusion, we demonstrated that cell-free miR-125b serum levels are decreased in serum from patients with AD as compared with NINDC and distinguish between AD and NINDCs with an accuracy of 82%.
Keywords: Alzheimer's disease; biomarker; miRNA; serum.