The plasma miRNAome in ADNI: Signatures to aid the detection of at-risk individuals

Abstract

Introduction: MicroRNAs are short non-coding RNAs that control proteostasis at the systems level and are emerging as potential prognostic and diagnostic biomarkers for Alzheimer's disease (AD).

Methods: We performed small RNA sequencing on plasma samples from 847 Alzheimer's Disease Neuroimaging Initiative (ADNI) participants.

Results: We identified microRNA signatures that correlate with AD diagnoses and help predict the conversion from mild cognitive impairment (MCI) to AD.

Discussion: Our data demonstrate that plasma microRNA signatures can be used to not only diagnose MCI, but also, critically, predict the conversion from MCI to AD. Moreover, combined with neuropsychological testing, plasma microRNAome evaluation helps predict MCI to AD conversion. These findings are of considerable public interest because they provide a path toward reducing indiscriminate utilization of costly and invasive testing by defining the at-risk segment of the aging population.

Highlights: We provide the first analysis of the plasma microRNAome for the ADNI study. The levels of several microRNAs can be used as biomarkers for the prediction of conversion from MCI to AD. Adding the evaluation of plasma microRNA levels to neuropsychological testing in a clinical setting increases the accuracy of MCI to AD conversion prediction.

Keywords: Alzheimer's disease; blood biomarker; cognitive decline; microRNA; mild cognitive impairment; plasma; small non‐coding RNA.

FIGURE 1

Experimental approach. Schematic overview of the experimental approach to analyze the plasma microRNAome of the Alzheimer's Disease Neuroimaging Initiative (ADNI) study and the overall aim. For further details see text.

FIGURE 2

miRNA distribution is largely unaffected by library preparation and independent of sequencing depth. (A) Left panel: experimental design. Right panel: Heatmap showing the cross correlation of the 7 batch control samples across the 21 sequencing batches. (B) Left panel: experimental design. Right panel: Heatmap showing the cross‐correlation of the 15 replicated samples that were randomly distributed across the 21 sequencing batches. (C) Heatmaps showing the spike‐in cross correlation within each of the 21 sequencing batches. 91% of the samples showed a spike‐in cross‐correlations of r > 0.9. Samples with high unknown variance (r < 0.6) were detected in batch 1 and 7 and were removed from further analysis resulting in 803 samples in total for the final analysis. miRNA, microRNA.

FIGURE 3

Selecting miRNAs for ML analysis. (A) Heat map showing miRs significantly correlated with the diagnostic groups (p < 0.05, linear regression analysis). (B) Heat map showing miRs significantly correlated with EMCI (EMCI‐AD) and LMCI (LMCI‐AD) patients converting to AD (p < 0.05, linear regression analysis). (C) Venn diagram comparing the miRNAs significantly correlated to diagnosis, EMCI‐AD, LMCI‐AD and the top5 miRNAs identified by one round of ML for control vs. EMCI, LMCI, MCI, EMCI‐AD, and LMCI‐AD revealing 73 unique miRNAs that were subsequently used for all ML approaches. AD, Alzheimer's disease; EMCI, early mild cognitive impairment; LMCI, late mild cognitive impairment; MCI, mild cognitive impairment; miRNA, microRNA; ML, machine learning.

FIGURE 4

ML identifies miRNA signatures for  EMCI, LMCI, and AD. The panels show the results from ML displayed as ROC curve analysis. Numbers in parenthesis indicate the respective AUCs for miRNA predicting AD (A), EMCI (B), or LMCI (C) and for the CSF biomarkers together with MMSE for AD (D), EMCI (E), and LMCI (F). (G) ROC plot showing the results when combining the data for miR‐4429, miR‐22‐5p, and miR‐1306 with MMSE. AD, Alzheimer's disease; EMCI, early mild cognitive impairment; LMCI, late mild cognitive impairment; MCI, mild cognitive impairment; miRNA, microRNA; ML, machine learning; MMSE, Mini‐Mental State Examination; ROC, receiver operating characteristics.

FIGURE 5

ML identifies miRNA signatures to predict EMCI‐AD and LMCI‐AD converters. (A) ROC curve displaying the results from ML using miR.125b.5p, miR.18a.3p and miR.25b.5p predicting EMCI‐AD converters with an AUC of 0.70. (B) ROC curve showing the performance of MMSE and CSF biomarkers to identify EMCI‐AD converters. (C) ROC curve showing that a signature of miR.338.3p, miR.584.5p and miR.142.3p can predict LMCI‐AD converters with an AUC of 0.75. (D) ROC curve showing the performance of MMSE and CSF biomarkers to identify LMCI‐AD converters. Numbers in parenthesis indicate the respective AUC values. AD, Alzheimer's disease; AUC, area under the curve; CSF, cerebrospinal fluid; EMCI, early mild cognitive impairment; LMCI, late mild cognitive impairment; MCI, mild cognitive impairment; miRNA, microRNA; ML, machine learning; MMSE, Mini‐Mental State Examination; ROC, receiver operating characteristics.

FIGURE 6

ML‐identified miRNA signatures to predict EMCI‐AD and LMCI‐AD converters in combination with the ADAScog13 test. (A) ROC curve displaying the results from ML using miR.151a.5p and miR.652.3p in combination with the ADAScog13 test predicting EMCI‐AD converters. (B) ROC curve showing the performance of miR.1306.3p and miR.532.3p in combination with the ADAscog13 test to predict LMCI‐AD converters Numbers in parenthesis indicate the respective AUC values. AD, Alzheimer's disease; EMCI, early mild cognitive impairment; LMCI, late mild cognitive impairment; MCI, mild cognitive impairment; miRNA, microRNA; ML, machine learning; MMSE, Mini‐Mental State Examination; ROC, receiver operating characteristics.

FIGURE 7

Pathways controlled by miRNAs associated with EMCI, LMCI, and AD. Dot plot showing the top 15 pathways affected in AD, as well as the pathways specific to the LMCI, AD across diagnoses. AD, Alzheimer's disease; EMCI, early mild cognitive impairment; LMCI, late mild cognitive impairment; miRNA, microRNA.

FIGURE 8

Pathways controlled by miRNAs associated with EMCI‐AD and LMCI‐AD converters. Dot plot showing the top common pathways affected in EMCI‐AD and LMCI‐AD, as well as the pathways specific to each condition. AD, Alzheimer's disease; EMCI, early mild cognitive impairment; LMCI, late mild cognitive impairment; miRNA, microRNA.