AUTHOR=Wang Dan , Song Juan , Cheng Ye , Xu Yiran , Song Lili , Qiao Yimeng , Li Bingbing , Xia Lei , Li Ming , Zhang Jin , Su Yu , Wang Ting , Ding Jian , Wang Xiaoyang , Wang Sujuan , Zhu Changlian , Xing Qinghe TITLE=Targeting the metabolic profile of amino acids to identify the key metabolic characteristics in cerebral palsy JOURNAL=Frontiers in Molecular Neuroscience VOLUME=16 YEAR=2023 URL=https://www.frontiersin.org/journals/molecular-neuroscience/articles/10.3389/fnmol.2023.1237745 DOI=10.3389/fnmol.2023.1237745 ISSN=1662-5099 ABSTRACT=Background

Cerebral palsy (CP) is a neurodevelopmental disorder characterized by motor impairment. In this study, we aimed to describe the characteristics of amino acids (AA) in the plasma of children with CP and identify AA that could play a potential role in the auxiliary diagnosis and treatment of CP.

Methods

Using high performance liquid chromatography, we performed metabolomics analysis of AA in plasma from 62 CP children and 60 healthy controls. Univariate and multivariate analyses were then applied to characterize different AA. AA markers associated with CP were then identified by machine learning based on the Lasso regression model for the validation of intra-sample interactions. Next, we calculated a discriminant formula and generated a receiver operating characteristic (ROC) curve based on the marker combination in the discriminant diagnostic model.

Results

A total of 33 AA were detected in the plasma of CP children and controls. Compared with controls, 5, 7, and 10 different AA were identified in total participants, premature infants, and full-term infants, respectively. Of these, β-amino-isobutyric acid [p = 2.9*10(−4), Fold change (FC) = 0.76, Variable importance of protection (VIP) = 1.75], tryptophan [p = 5.4*10(−4), FC = 0.87, VIP = 2.22], and asparagine [p = 3.6*10(−3), FC = 0.82, VIP = 1.64], were significantly lower in the three groups of CP patients than that in controls. The combination of β-amino-isobutyric acid, tryptophan, and taurine, provided high levels of diagnostic classification and risk prediction efficacy for preterm children with an area under the curve (AUC) value of 0.8741 [95% confidence interval (CI): 0.7322–1.000]. The discriminant diagnostic formula for preterm infant with CP based on the potential marker combination was defined by p = 1/(1 + e−(8.295–0.3848* BAIBA-0.1120*Trp + 0.0108*Tau)).

Conclusion

Full-spectrum analysis of amino acid metabolomics revealed a distinct profile in CP, including reductions in the levels of β-amino-isobutyric acid, tryptophan, and taurine. Our findings shed new light on the pathogenesis and diagnosis of premature infants with CP.