1H-NMR-based metabolomic analysis of cerebrospinal fluid from adult bilateral moyamoya disease: comparison with unilateral moyamoya disease and atherosclerotic stenosis

Abstract

Although metabolomics has been increasingly used to observe metabolic pattern and disease-specific metabolic markers, metabolite profiling for moyamoya disease (MMD) has not yet been done in adults. This study investigated cerebrospinal fluid (CSF) metabolites specific to bilateral MMD (B-MMD) and compared them to those of unilateral MMD (U-MMD) or atherosclerotic stenosis with hydrogen-1 nuclear magnetic resonance spectroscopy to identify metabolic biomarkers associated with MMD in adults.CSF samples of B-MMD (n = 29), U-MMD (n = 11), and atherosclerotic cerebrovascular disease (ACVD) (n = 8) were recruited. Principal component analysis, partial least square discriminant analysis, and orthogonal projections to latent structure discriminant analysis (OPLS-DA) were done for the comparisons. Diagnostic performance was acquired by prediction of 1 left-out sample from the distinction model constructed with the rest of the samples.B-MMD showed an increase in glutamine (P < 0.001) and taurine (P = 0.004), and a decrease in glucose (P < 0.001), citrate (P = 0.002), and myo-inositol (P = 0.006) than those in ACVD. U-MMD showed a higher level of glutamine (P = 0.005) and taurine (P = 0.034), and a lower level of glutamate (P < 0.004) than those in ACVD. No difference at the metabolite level was observed between B-MMD and U-MMD. Cross-validation with the OPLS-DA model showed a high accuracy for the prediction of MMD.The results of the study suggest that a metabolomics approach may be helpful in confirming MMD and providing a better understanding of MMD pathogenesis. Elevated glutamine in the CSF may be associated with MMD pathogenesis, which was different from ACVD.

FIGURE 1

Representative 600-MHz ¹H-NMR spectra of CSF. Key for CSF is as follows: 1, Isopropanol; 2, lactate; 3, acetate; 4, glutamine; 5, citrate; 6, 2-aminoadipate; 7, myo-inositol; 8, glucose; 9, taurine; 10, creatine; 11, glutamate. CSF = cerebrospinal fluid, ¹H-NMR = hydrogen-1 nuclear magnetic resonance, PPM = ??.

FIGURE 2

OPLS-DA and S-plot analysis between B-MMD and ACVD (A and B), and U-MMD and ACVD (C and D). Potential markers that were significantly different between the 2 groups are enclosed in dotted boxes. ▪, B-MMD; ●, U-MMD; ♦, ACVD. ACVD = atherosclerotic cerebrovascular disease, B-MMD = bilateral moyamoya disease, OPLS-DA = Orthogonal projections to latent structure discriminant analysis, U-MMD = unilateral MMD.

FIGURE 3

Y-predicted scatter plot of an OPLS-DA model between B-MMD and ACVD (A), and U-MMD and ACVD (B). The OPLS-DA model of B-MMD and ACVD patients had a slightly higher goodness of fit (R2X = 0.424, Q2 = 0.531) than that of U-MMD and ACVD (R2X = 0.647, Q2 = 0.252). ▪, B-MMD; ●, U-MMD; ♦, ACVD; ▴; test set; ▵; mispredicted sample; Y predicted cutoff of 0.5. ACVD = atherosclerotic cerebrovascular disease, B-MMD = bilateral moyamoya disease, MMD = moyamoya disease, OPLS-DA = Orthogonal projections to latent structure discriminant analysis, U-MMD = unilateral MMD.

FIGURE 4

Summary of the metabolic pathway in patients with ACVD. The red characters represent the metabolites identified in this study. The metabolites with arrows are significantly changed. A red or blue arrow shows the trend of the metabolites in ACVD compared with B-MMD or U-MMD, respectively. ACVD = atherosclerotic cerebrovascular disease, B-MMD = bilateral moyamoya disease, CoA = coenzyme A, U-MMD = unilateral MMD.