2.1. Subjects
A total of 120 ASD children aged 2-6 years were selected for this study from the Maternal and Child Care Health Hospital of Hainan Province, China, after a comprehensive assessment. Inclusion criteria were a diagnosis of ASD, which was made by developmental pediatrician of our research team through a series of structured interviews according to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) criteria[13]. The Childhood Autism Rating Scale (CARS)[14] was used as an assistant to diagnose by scores above the cut-off point 30. Exclusion criteria included a history of other developmental disorders, neurological or psychiatric diseases, genetic metabolic disease, major physical illness, recent infection, recent use of special diets, recent use of antibiotics or probiotics within one months before sampling.
Symptoms of the children with ASD were assessed with the Autism Behavior Checklist (ABC)[14], Social Responsiveness Scale (SRS) [15]and the CARS test. Higher scores of ABC, CARS, and SRS scores indicate more serious autistic symptoms. Neurodevelopment in ASD children was assessed with the revised Gesell Developmental Scale (GDS) [16],which is extensively used in China to evaluate cognitional and behavioral development, and the development quotient scores (DQ) reflect the levels of intellectual and behavioral development. DQ <75 indicates developmental delay, and the lower DQ score, the more severe developmental delay.
Additionally, a control group of 60 typically developing (TD) children was recruited and matched to the ASD group by age, gender, and region. The TD children received health examinations at the Department of Child Health in Maternal and Child Care Health Hospital of Hainan Province. They are healthy, and they did not have any signs of developmental disorders or psychiatric diseases, and noticeable gastrointestinal symptoms. Other exclusion criteria were the same as for the ASD group.
Participation in this research was voluntary. The study protocol was approved by the institutional review board of Children’s Hospital, Chongqing Medical University. This cross-sectional case-control study was based on a clinical trial which was registered in the Chinese Clinical Trial Registry (ChiCTR; registration number: ChiCTR-ROC-14005442).
2.2. Fecal sample collection and LC-MS metabolomics analysis
2.2.1. Fecal sample collection
Fecal samples were collected from each participant and immediately frozen and stored at −80°C until further analysis. The 100 mg of stool for each sample was preserved in sterile tubes for metabolism analysis.
2.2.2.Metabolites extraction
The 100 mg of stool for each sample was separately ground with liquid nitrogen and the homogenate was resuspended in prechilled 80% methanol and 0.1% formic acid by vortexing thoroughly. Samples were incubated on ice for 5 min then centrifuged at 15,000 rpm at 4°C for 5 min. Some supernatants were diluted with LC-MS-grade water to a final concentration of 60% methanol. Hereafter, samples were transferred into a fresh Eppendorf tube through a 0.22 μm filter then centrifuged at 15,000 g at 4°C for 10 min. Finally, the filtrate was injected into the LC-MS/MS system for analysis.
2.2.3. UHPLC-MS/MS analysis
LC-MS/MS analyses were performed using a Vanquish UHPLC system (Thermo Fisher, USA) and an Orbitrap Q Exactive HF-X mass spectrometer (Thermo Fisher, USA). Briefly, metabolites were first separated and characterized by using a liquid chromatography system and further detected with a mass spectrometry system. Samples were injected onto an Hyperil Gold column (100 × 2.1 mm, 1.9 μm) at a flow rate of 0.2 mL/min and separated using a 16 min linear gradient. The eluents for positive polarity mode were eluent A (0.1% formic acid in water) and eluent B (methanol). The eluents for negative polarity mode were eluent A (5 mM ammonium acetate, pH 9.0) and eluent B (methanol). The solvent gradient was set as follows: 2% B 1.5 min, 2-100% B 12.0 min, 100% B 14.0 min, 100-2% B 14.1 min, 2% B 16 min. The Q Exactive HF-X mass spectrometer was operated in positive/negative polarity mode with a spray voltage of 3.2 kV, a sheath gas flow rate of 35 arb, an aux gas flow rate of 10 arb, and capillary temperature of 320°C.
2.2.4 Metabolite analysis
Compound Discoverer 3.0 (CD 3.0, Thermo Fisher) was used to process and normalize the raw data files generated by UHPLC-MS/MS to perform peak alignment, peak selection, and quantification for each metabolite. The main parameters were set as follows: retention time tolerance 0.2 min, actual mass tolerance 5 ppm, signal intensity tolerance 30%, signal/noise ratio 3, minimum intensity 100,000. Peak intensities were normalized against the total spectral intensity, and normalized data were used to predict the molecular formula based on additive ions, molecular ion peaks, and fragment ions. Peaks were matched with mzCloud (https://www.mzcloud.org/) and ChemSpider (http://www.chemspider.com/) databases to obtain accurate qualitative and relative quantitative results.
The normalized metabolism data were analyzed by the CentOS (CentOS release 6.6), statistical software R (R version R-3.4.3), and SPSS statistical software (version 19.0, SPSS Inc., USA). With individual metabolite dataset, Partial least squares discriminant analysis (PLS-DA) models were built to visualize the metabolic alteration patterns between ASD and TD children. Furthermore, the cross-validation ANOVA (CV-ANOVA) was calculated to assess the reliability of the models. Differential metabolites between the two groups were selected by combined multivariate and univariate analysis methods. Gut metabolites with fold change >1.5, variable important in projection (VIP) >1, and FDR-corrected P values < 0.05 for student’s t-test or Mann-Whitney U test were considered significantly differential metabolites between groups. To further demonstrate the biological functions of the associated differential metabolites, the KEGG pathways enrichment analysis was performed (http://www.genome.jp/kegg/), and a hypergeometric test was used to assess the significance of KEGG pathway.
The metabolomics analysis was carried according to the standard protocols recommended by Novogene Technology Co., Ltd. (Beijing, China). The raw data were deposited into the MetaboLights database (accession number: MTBLS1946, www.ebi.ac.uk/metabolights).
2.3. Statistical Analysis
The demographics and clinical assessment data were analyzed using SPSS statistical software (version 19.0, SPSS Inc., USA). Continuous variables are described as the means with standard deviations or medians (interquartile ranges) when appropriate, and categorical variables are described as percentages. The two-tailed student’s t-test, Mann-Whitney U test, and the chi-square test were used to compare levels between groups. The correlations between metabolites levels with clinical symptoms scores were analyzed by Spearman correlation. P-value < 0.05 was presumed as statistically significance.