3.1 Characterization of exosomes obtained from moyamoya disease patient plasma
The clinical information of the sample is shown in Table 2. There is no significant difference between the age and gender of the patient. According to whether the patient will see smoke-like changes in the brain during the digital subtraction cerebrovascular angiography examination, we can distinguish patients with moyamoya disease from patients with non-moyamoya disease. For example, Figs. 1A and B show a patient with the same hemorrhagic moyamoya disease (bilateral), characterized by cerebral hemorrhage, occlusion of the end of the internal carotid artery on the right side, and smoke vessel formation). Figures 1C and D are the same healthy person, showing normal cerebral blood vessels. To characterize exosomes derived from moyamoya disease patient plasma, transmission electron microscopy, nanoparticle tracking analysis and western blotting were used to characterize exosome diameters and protein markers. CD63 and CD81 are protein markers of exosomes, and CD63 and CD81 are protein markers of exosomes, and were detected in NC exosomes, MMD exosomes and plasma by Western blotting analysis, it confirmed that we isolated exosomes (Fig. 1E). Transmission electron microscopy showed a typical rounded morphology with a sagged double membrane (Fig. 1F). The nanoparticle tracking analysis further confirmed that the exosome isolated from moyamoya disease patients and health people were 30–150 nm in diameter (Fig. 1G). These results indicated that exosomes were successfully purified from all plasma samples.
Table 2
Clinical information of the samples
|
NC
|
MMD
|
P-value
|
Male sex
|
5(50)
|
5(56%)
|
0.967
|
Age
|
46.1 ± 6.4
|
48.9 ± 12.1
|
0.603
|
WBC(×109)
|
6.08 ± 1.24
|
7.23 ± 1.37
|
0.169
|
RBC(×109)
|
4.45 ± 0.32
|
4.45 ± 0.41
|
0.967
|
Hb(mmol/L)
|
133 ± 7.1
|
132 ± 13
|
0.791
|
Glu(mmol/L)
|
5.31 ± 0.62
|
5.21 ± 0.65
|
0.729
|
K+(mmol/L)
|
4.02 ± 0.21
|
3.95 ± 0.21
|
0.456
|
Na+(mmol/L)
|
139.7 ± 3.4
|
141.1 ± 2.5
|
0.306
|
ALT(U/L)
|
19.5 ± 6.6
|
24.45 ± 15.18
|
0.364
|
TB(umol/L)
|
9.38 ± 3.03
|
8.58 ± 3.01
|
0.561
|
DB(umol/L)
|
1.92 ± 0.19
|
1.75 ± 0.39
|
0.232
|
TC(mg/DL)
|
3.63 ± 1.03
|
3.88 ± 1.57
|
0.678
|
TG(mg/DL)
|
1.41 ± 0.69
|
1.54 ± 0.87
|
0.737
|
HDL(mg/DL)
|
0.99 ± 0.07
|
0.97 ± 0.27
|
0.85
|
LDL(mg/DL)
|
2.88 ± 0.63
|
3.02 ± 0.77
|
0.662
|
Table 3
Sequencing data output statistics and quality control
Sample Name
|
Raw Reads
|
Clean Reads
|
Q20
|
Q30
|
GC%
|
MMD1
|
13394601
|
11470036
|
0.994038
|
0.979452
|
0.541002
|
MMD2
|
16789838
|
15597279
|
0.993253
|
0.977721
|
0.543984
|
MMD3
|
19767133
|
14471822
|
0.993802
|
0.979251
|
0.541744
|
MMD4
|
10192610
|
8096998
|
0.994197
|
0.979526
|
0.552548
|
MMD5
|
13430381
|
11318666
|
0.993618
|
0.978735
|
0.542509
|
MMD6
|
19573465
|
17755102
|
0.994607
|
0.98019
|
0.538964
|
MMD7
|
20702651
|
17529882
|
0.99429
|
0.980239
|
0.527753
|
MMD8
|
18527733
|
17109638
|
0.99283
|
0.977045
|
0.502724
|
MMD9
|
21319758
|
20473411
|
0.993715
|
0.978213
|
0.487527
|
NC1
|
12058148
|
11396788
|
0.984477
|
0.949508
|
0.552691
|
NC2
|
37649858
|
36311886
|
0.990424
|
0.966974
|
0.548631
|
NC3
|
21111513
|
18796751
|
0.989343
|
0.965364
|
0.566005
|
NC4
|
35556828
|
29895610
|
0.986185
|
0.954861
|
0.548253
|
NC5
|
11792452
|
11097666
|
0.985199
|
0.951597
|
0.559581
|
NC6
|
15052720
|
4073973
|
0.992414
|
0.974378
|
0.585476
|
NC7
|
11093372
|
5542318
|
0.992088
|
0.973208
|
0.580912
|
NC8
|
10618213
|
7722455
|
0.991523
|
0.972106
|
0.588273
|
NC9
|
11282238
|
7778548
|
0.992302
|
0.974028
|
0.582235
|
NC10
|
30780243
|
20582756
|
0.992305
|
0.974105
|
0.583265
|
3.2 GO and KEGG analyses of the target genes of differentially expressed miRNAs
Next, the raw data were obtained by high-throughput sequencing. The connectors at both ends of the reads were cut off by cutadapt software, and the reads with lengths greater than 17 nt were retained. The basic quality information on the reads showed that they can be used for subsequent data analyses (Table 3). A total of 1002 differentially expressed miRNAs were identified based on an false discovery rate < 0.01 and fold change > 2, and they included 585 upregulated and 417 downregulated miRNAs (Fig. 2A). Target genes of differential miRNAs were predicted via the TargetScan database. A gene ontology (GO) analysis and Kyoto Encyclopaedia of Genes and Genomes (KEGG) pathway analysis were conducted to explore the function of the target genes of differentially expressed miRNAs. According to the GO analysis, the results indicated that these decreased expressed miRNAs were primarily associated with biological process terms, including regulation of cellular metabolic process, regulation of cellular process, cellular macromolecule metabolic process(Fig. 2B). They were also were enriched in cellular component terms, including nucleoplasm, intracellular part, intracellular(Fig. 2C), and molecular function terms, including transcription regulatory region sequence-specific DNA binding and regulatory region nucleic acid binding(Fig. 2D). These increased expressed miRNAs were primarily associated with biological process terms, including nervous system development, positive regulation of cellular process, positive regulation of biological process(Fig. 3A); cellular component terms, including cytoplasm, intracellular, and intracellular part(Fig. 3B), and molecular function terms, including protein binding and regulatory region nucleic acid binding (Fig. 3C).The KEGG pathway analysis for these target genes showed that they were mainly enriched in axon guidance, regulation of the actin cytoskeleton and the MAPK signalling pathway (Fig. 3D). There were 63 exosomal miRNAs involved in the axon guidance, regulation of actin cytoskeleton and MAPK signaling pathway (Fig. 4A). Among them, the different target genes of 59 exosomal miRNAs were associated with the three pathways at the same time, especially with regulation of actin cytoskeleton (Fig. 4B). Function predictions and pathway analyses of target genes of differentially expressed miRNAs could provide insights to help regulate the actin cytoskeleton in the pathogenesis of moyamoya disease.
3.3 Prognosis of potential biomarkers by the receiver operating characteristic curve and area under curve
To further screen potential biomarkers of moyamoya disease, the area under the curve of the receiver operating characteristic (ROC) curve was used to evaluate the sensitivity and specificity of biomarkers for predicting events. The sensitivity and specificity of each miRNA were determined by the optimal threshold of the area under the curve (AUC). AUC of ROC were carried out determining the diagnostic values of these 63 miRNAs involved in the axon guidance, regulation of actin cytoskeleton and MAPK signaling pathway. Ten miRNAs (miR-1306-5p, miR-196b-5p, miR-19a-3p, miR-22-3p, miR-320b, miR-34a-5p, miR-485-3p, miR-489-3p, miR-501-3p, and miR-487b-3p) had significantly differentiated moyamoya disease patients from healthy controls based on an AUC value above 0.9 and sensitivity and specificity above 0.8 (Fig. 5). To further explore the expression of these exosomal miRNAs as potential biomarkers of moyamoya disease, RT-QPCR was performed to detect randomly the levels of these exosomal miRNAs extracted from the plasma samples of 9 moyamoya disease patients and 10 healthy individuals. These ten microRNAs showed the same expression patterns obtained from the high-throughput sequencing analysis. Hsa-miR-34a-5p, Hsa-miR-19a-3p, miR-22-3p, miR-196a-5p, miR-320b, miR-485-3p, miR-487b-3p, and miR-501-3p were up-regulated, and miR-489-3p and hsa-miR-1306-5p were downregulated (Fig. 6). These results indicated that these exosomal miRNAs are potential biomarkers that may participate in the regulation of the actin cytoskeleton in moyamoya disease. In summary, the analysis of the differentially expressed exosomal miRNAs in plasma revealed that they may have a functional role in the pathogenesis of moyamoya disease.