Plasma microRNA markers of upper limb recovery following human stroke

Abstract

Preclinical investigators have implicated several microRNAs as regulators of gene expression promoting neural plasticity following experimental stroke in rodent models. Our goal was to determine whether similar microRNAs might be identifiable in plasma of humans with variable recovery from stroke. Plasma was collected 19 days post-stroke from 27 participants with mild-moderate upper extremity impairment enrolled in the Critical Periods After Stroke Study (CPASS). MicroRNA expression was assessed using TaqMan microRNA assays. Good clinical recovery was defined as ≥6 point change in the Action Research Arm Test (ARAT) score from baseline to 6 months, with 22 subjects showing good and 5 showing poor recovery. When comparing the good versus poor recovery groups, six microRNAs showed significantly increased expression - miR-371-3p, miR-524, miR-520g, miR-1255A, miR-453, and miR-583, while 3 showed significantly decreased expression - miR-941, miR-449b, and miR-581. MiR-371-3p and miR-941 have previously been associated with neural repair mechanisms; none of the significant microRNAs have previously been associated with stroke. The 9 microRNAs converge on pathways associated with axonal guidance, developmental biology, and cancer. We conclude that plasma microRNAs may be informative regarding human neural repair mechanisms during stroke recovery and probably differ from those seen in experimental stroke models.

Figure 1

Fold-change for microRNAs with significant differential expression between participants with good (ΔARAT ≥ 6) vs. poor (ΔARAT < 6) recovery of the upper limb. Error bars represent standard deviation.

Figure 2

(A) Receiver operating characteristic (ROC) curve for good (ΔARAT ≥ 6) versus poor (ΔARAT < 6) recovery using a combination of five miRNAs - miR-581, miR-519b-3p, miR-941, miR-449b, and miR-616. (B) Predicted class probabilities for the five miRNA predictive panel, demonstrating 25 correctly classified and 2 misclassified participants. The 2 misclassified participants are labeled by their respective ΔARAT scores.