Exosomal MicroRNAs Contribute to Cognitive Impairment in Hypertensive Patients by Decreasing Frontal Cerebrovascular Reactivity

Abstract

Mechanisms underlying cognitive impairment (CI) in hypertensive patients remain relatively unclear. The present study aimed to explore the relationship among serum exosomal microRNAs (miRNAs), cerebrovascular reactivity (CVR), and cognitive function in hypertensive patients. Seventy-three hypertensive patients with CI (HT-CI), 67 hypertensive patients with normal cognition (HT-NC), and 37 healthy controls underwent identification of exosomal miRNA, multimodal magnetic resonance imaging (MRI) scans, and neuropsychological tests. CVR mapping was investigated based on resting-state functional MRI data. Compared with healthy subjects and HT-NC subjects, HT-CI subjects displayed decreased serum exosomal miRNA-330-3p. The group difference of CVR was mainly found in the left frontal lobe and demonstrated that HT-CI group had a lower CVR than both HT-NC group and control group. Furthermore, both the CVR in the left medial superior frontal gyrus and the miRNA-330-3p level were significantly correlated with executive function (r = -0.275, P = 0.021, and r = -0.246, P = 0.04, respectively) in HT-CI subjects, and the CVR was significantly correlated with the miRNA-330-3p level (r = 0.246, P = 0.040). Notably, path analysis showed that the CVR mediated the association between miRNA-330-3p and executive function. In conclusion, decreased miRNA-330-3p might contribute to CI in hypertensive patients by decreasing frontal CVR and could be a biomarker of early diagnosis.

Keywords: cerebrovascular reactivity; cognitive impairment; exosomal microRNA; hypertension; mediation.

FIGURE 1

Characteristics of the obtained exosomes from human serum in this study. (A) The transmission electron microscopy of the exosomes from the serum. Scale bar = 100 nm. (B) Identification of exosomal markers with indicated antibodies by western blot. Measurement of size distribution (C) and morphology (D) of exosomes purified from serum by nanoparticle-tracking analyzer.

FIGURE 2

Screening on differentially expressed miRNA. (A) Flow chart of study design. (B) Volcano plot of differentially expressed miRNAs between HT-CI and control groups. (C) Volcano plot of differentially expressed miRNAs between HT-CI and HT-NC groups. HT-NC, hypertensive patients with normal cognition; HT-CI, hypertensive patients with cognitive impairment.

FIGURE 3

MicroRNA difference among the three groups. (A) Subjects in the HT-CI group demonstrated significantly lower concentration of miRNA-330-3p than the other two groups. (B) Subjects in the HT-CI group displayed significantly lower concentration of miRNA-432-5p than the HT-NC group. (C) Subjects in the HT-CI group and HT-NC group showed significantly lower miRNA-625-3p than the control group. (D) Subjects in the HT-CI group and HT-NC group showed significantly higher miRNA-6852-3p than the control group. Error bar indicates standard deviation. *P < 0.05, **P < 0.01, ***P < 0.001. HT-NC, hypertensive patients with normal cognition; HT-CI, hypertensive patients with cognitive impairment.

FIGURE 4

Bioinformatics analysis of identified miRNAs. (A) The results of Gene Ontology function analysis on differentially expressed miRNA target genes. (B) The entries of Kyoto Encyclopedia of Genes and Genomes pathway enrichment analysis on differentially expressed miRNA target genes.

FIGURE 5

The CVR difference among the three groups. (A) Significant CVR regions on the brain. The four regions with significant difference were all located in the left frontal lobe. (B) Subjects in the HT-NC group demonstrated significantly higher CVR in the left prefrontal white matter than the other two groups. (C) Subjects in the HT-CI group demonstrated significantly lower CVR in the left triangular inferior frontal gyrus. (D) Subjects in the HT-NC group demonstrated significantly higher CVR in the left medial superior frontal gyrus than the other two groups. (E) Subjects in the HT-CI group and HT-NC group showed significantly lower CVR in the left precentral gyrus than the control group. The threshold was set at corrected P-value < 0.005, determined by Gaussian random field for multiple comparisons. Error bar indicates standard deviation. *P < 0.05, **P < 0.01, ***P < 0.001. Prefrontal white matter_L, left prefrontal white matter; Frontal_Inf_Tri_L, left triangular inferior frontal gyrus; Frontal_Sup_Medial_L, left medial superior frontal gyrus; Precentral_L, left precentral gyrus; r.u., relative unit.

FIGURE 6

Correlations between CVR, miRNA, and cognitive performance in the HT-CI group. (A) The CVR value in the left triangular inferior frontal gyrus was negatively correlated with visuospatial function. (B,C) The CVR of the left medial superior frontal lobe had a negative correlation with executive function and language. (D) A significant inverse correlation between serum miRNA-330-3p and executive function was discovered in the HT-CI group. (E,F) The CVR of the left medial superior frontal lobe had a positive correlation with serum miRNA-330-3p and miRNA-452-5p. Frontal_Inf_Tri_L, left triangular inferior frontal gyrus; Frontal_Sup_Medial_L, left medial superior frontal gyrus; r.u., relative unit.

FIGURE 7

The mediation effect analysis of CVR in the left medial superior frontal lobe on the association between miRNA-330-3p and executive function in HT-CI patients. The standard coefficient and P-value were displayed on the pathway controlling for age, years of education, and history of lacunar stroke. The indirect mediation effect of β = –0.23 and a 95% CI as (–0.583, –0.001) was present in the center of the diagram. Red block indicated the significant CVR region. Frontal_Sup_Medial_L, left medial superior frontal gyrus.