Gut microbiome and serum metabolome as biomarkers of clinical severity and prognosis in patients with traumatic brain injury

To the Editor: Traumatic brain injury (TBI) is a life-threatening condition and global public health concern in both developed and developing countries.[1] During the last few decades, the morbidity and mortality associated with TBI have decreased. However, the precise mechanisms that contribute to the outcomes remain unclear. Recently, researchers have become aware that owing to the brain–gut axis interaction, the central nervous system is closely related to the gut microbiota.[2] Many studies have correlated the gut microbiota and its metabolites with autism, depression, stroke, and Parkinson’s disease.[3] Whether changes in the gut microbial composition affect TBI remains largely unknown. Metabolomics, a method focuses on deriving a comprehensive analysis of low-molecular-weight endogenous metabolites in biofluids or tissues and has been used to explore metabolic changes in a variety of diseases. However, the relationship between the metabolome and TBI outcomes or progression and the relationship between the gut microbiome and TBI remain unclear. Hence, we compared the gut microbiome and serum metabolome of patients with TBI and healthy individuals to identify specific bacterial and serum metabolome differences. The First Affiliated Hospital of Anhui Medical University Institutional Review Board approved the study (No. PJ 2023-11-15). Patients with TBI were admitted to the Department of Neurosurgery in the First Affiliated Hospital of the Anhui Medical University from October 2020 to June 2022. The enrollment criteria included age ≥18 years and exclusion criteria included pregnant women, and patients transferred from outside hospitals. Consent to participate and continue the study was obtained from the healthy volunteers and patient or a legally authorized representative after admission. Healthy volunteers were enrolled for comparison. Stool samples were collected within 6 h following admission and before the introduction of antibiotics. All fecal samples were stored at −80°C within 15 min of sample collection. The gut microbiota in the stool was analyzed using 16S ribosomal RNA (rRNA). Details of the 16S rRNA are provided in the Supplementary Material, https://links.lww.com/CM9/C155. All serum specimens were thawed at 4°C, and 100 μL was transferred into 2-mL centrifuge tubes. Samples were centrifuged at 4°C for 10 min at 12,000 rpm (centrifuge radius 8.6 cm), and the supernatant was transferred into another 2-mL centrifuge tube. Samples were dissolved with 150 μL of 80% methanol solution containing 2-chlorobenzalanine (4 parts per million), and the supernatant was filtered through a 0.220 μm membrane to obtain the samples for liquid chromatography–mass spectrometry (LC–MS). Details of the LC–MS can be obtained in the Supplementary Material, https://links.lww.com/CM9/C155. The generated FASTQ files were imported, demultiplexed, quality-filtered, and processed using the Quantitative Insights into Microbial Ecology (QIIME) 2 v2020.8 pipeline. Partial 16S rRNA bacterial sequences were qualitatively trimmed, and the reads were truncated into Operational Taxonomic Units (OTUs), referring to the Phred quality scores of the bases of the sequences, using the software package Divisive Amplicon Denoising Algorithm 2 in QIIME 2 (https://curr-protoc-bioinformatics.qiime2.org). This resulted in a median quality score of 36 (interquartile range, 25–37) for the forward reads and 37 (interquartile range, 33–37) for the reverse reads. Taxonomy was assigned based on the SILVA database v138 at the 99% OTU level. A phylogenetic tree was created to generate phylogenetic diversity measures using the q2-phylogeny plugin of QIIME 2, which performed multiple sequence alignments of the OTU sequences and generated a phylogenetic tree of the OTUs from the alignment. Statistical analysis methods are shown in the Supplementary Materials, https://links.lww.com/CM9/C155. A total of 47 patients with TBI were recruited into different Glasgow Coma Scale (GCS) TBI groups (G1, G2, and G3 groups, G1: GCS 13–15; G2: GCS 9–12; G3: GCS 3–8) with similar age, sex, and other factors. Of these, nine were excluded because of inadequate fecal or serum samples. Twelve volunteers were assigned to the control group. In total, samples from four healthy individuals and 34 patients with TBI (eight patients in G1, six patients in G2, and 20 patients in G3) underwent serum metabolome analysis, while gut microbiome samples from 12 healthy individuals and 36 patients with TBI (seven patients in G1, five patients in G2, and 24 patients in G3) underwent 16S rRNA analysis. The microbial diversity and richness indices differed significantly between the two groups. Compared to the healthy control group, patients with TBI were characterized by decreased Shannon, Chao, and Ace indices and an increased Simpson index (Figure 1A, Shannon P = 0.0006, Simpson P = 0.0007, Chao P = 0.0005, Ace P = 0.0004), suggesting decreased species diversity and richness indices in patients with TBI.Figure 1: (A) Microbial richness significantly differed in patients with TBI based on Shannon, Simpson, Ace, and Chao relative to controls. (B) Microbial richness significantly differed in GCS TBI groups. (C) OPLS-DA shows the serum metabolism of TBI and the control grou. (D) Seventeen differential serum metabolites enrichment for patients with TBI compared to those in control in 17 KEGG pathways. (E) PLS-DA shows serum metabolism of TBI in different GCS TBI groups. (F) Twenty-five differential serum metabolites for different GCS TBI groups. * P <0.001, † P <0.05; ‡ P <0.01. C: Control group; G: GCS TBI group; G1: Patients with GCS 13–15; G2: Patients with GCS 9–12; G3: Patients with GCS 3–8; GCS: Glasgow Coma Scale; KEGG: Kyoto Encyclopedia of Genes and Genomes; OPLS-DA: Orthogonal Partial Least Squares Discrimination Analysis; PLS-DA: Partial Least Squares Discrimination Analysis; TBI: Traumatic brain injury.No differences in baseline characteristics were observed among patients with different GCS scores. Among the different GCS TBI groups, significant differences were observed in alpha diversity (Figure 1B, Shannon G1 vs. G2 P = 0.0499, G1 vs. G3 P = 0.0049; Simpson G1 vs. G2 P = 0.0217, G1 vs. G3 P = 0.0024; Chao G1 vs. G3 P = 0.0014, and Ace G1 vs. G3 P = 0.0087), which included species diversity indices (Shannon and Simpson) and species richness indices (Ace and Chao). The LC–MS-based metabolomics method analyzed the serum metabolome. The serum metabolic phenotype in the TBI group was notably higher than that in the healthy control group [Figure 1C]. To identify the microbial characteristics that distinguished the healthy control and TBI groups, the linear discriminant analysis effect size (LEfSe) method was used to analyze the differential OTUs between these two groups [Supplementary Figure 1, https://links.lww.com/CM9/C155]. Furthermore, to identify the microbial characteristics that distinguish the severity of TBI, LEfSe analysis was used to analyze the differential OTUs among the three groups [Supplementary Figure 2, https://links.lww.com/CM9/C155]. These differential metabolites were further subjected to pathway analysis using MetaboAnalyst 4.0 (https://www.metaboanalyst.ca). Among the 17 identified pathways, the lipid-related metabolic pathway (synthesis and degradation of ketone bodies) was the most significantly enriched [Figure 1D, all P <0.050]. The serum metabolomics phenotypes differed among the different GCS TBI groups [Figure 1E]. Differential metabolites were subjected to pathway analysis using MetaboAnalyst 4.0. Among the 25 identified pathways, the D-glutamine metabolic pathway was the most significantly enriched [Figure 1F, P ≤0.050]. Several studies have focused on the association between serum biomarkers and outcomes in patients with TBI.[4,5] This study demonstrated that multiple metabolic functions are closely related to the gut microbiota and the serum metabolome may be altered by the gut microbiota. LEfSe analysis indicated that lipid-related metabolic pathways were significantly enriched in the TBI group, which may be a novel biomarker for predicting progression in patients with TBI. These findings support the hypothesis that ketone metabolism plays a significant role in the neuroprotective effects of TBI. This study has several strengths and limitations. Differential gut microbiota and serum metabolome data were obtained from sequencing data. However, their roles in the pathogenesis of TBI remain unclear. Whether these analyses can be applied to clinical treatments requires many clinical trials for verification. In the current study, 16S rRNA gene sequencing combined with LC–MS analysis investigated the crosstalk between the gut microbiota and serum metabolome in patients with TBI to improve future therapies. The lipid-related metabolic pathway (synthesis and degradation of ketone bodies) was the most significantly enriched, and we identified a potential ketogenic diet treatment for patients with TBI. Funding This project was supported by grants from National Natural Science Foundation of China (No. 82202438), Clinical and Translational Research Project of Anhui Province (No. 202427b10020130), Beijing Postdoctoral Foundation (No. 2020-ZZ-008), Returned overseas students from Anhui Province in 2020 Innovation and Entrepreneurship Support Plan Project (No. 2020LCX016), Funding of The Anhui Universal Natural Science Foundation (No. KJ2020A0172), and Cultivate Funding of the National Natural Science Foundation of China (No. 2016KJ12). Conflicts of interest None.