Marker, kit and medicament for predicting or diagnosing sepsis in children

Abstract

This invention provides biomarkers, kits, and drugs for predicting or diagnosing childhood sepsis. The biomarkers include *Enterococcus raffinis*. Based on statistical analysis, machine learning modeling results, and independent cohort validation, this invention found significant enrichment of *Enterococcus raffinis* and sulfated bile acid derivatives in pediatric sepsis patients. In vitro and in vivo experiments confirmed that symbiotic gut microbiota (especially *Enterococcus raffinis*) produces at least 80% of its total DCA-3S production, highlighting the dominant role of the gut microbiota in DCA-3S biosynthesis.

Description

Technical Field

[0001] This invention relates to the field of biotechnology, and more specifically to biomarkers, reagent kits, and drugs for predicting or diagnosing sepsis in children. Background Technology

[0002] Sepsis is a systemic inflammatory response syndrome triggered by infection and is one of the leading causes of death worldwide. Approximately 20.3 million cases of sepsis occur annually, with 41.5% occurring in children under the age of five, accounting for 20% of global deaths. Childhood sepsis not only poses a direct threat to the patient's life but can also lead to long-term organ dysfunction and neurodevelopmental disorders, placing a heavy burden on families and society. Because children's immune systems are not fully developed, sepsis often progresses more rapidly in children, and early symptoms are often atypical, easily confused with other illnesses such as the common cold or gastroenteritis, leading to delayed diagnosis.

[0003] Currently, the diagnosis of sepsis mainly relies on clinical manifestations, laboratory tests, and imaging examinations. While clinical manifestations such as fever, chills, and hypotension are common, they often lack specificity in the early stages and are easily confused with other infectious or non-infectious diseases. Laboratory tests typically include the detection of inflammatory markers such as white blood cell count, C-reactive protein (CRP), and procalcitonin (PCT). However, while these indicators can reflect the inflammatory state, they cannot accurately distinguish between infectious and non-infectious inflammation, and their assessment of disease severity is not precise enough. Imaging examinations such as X-rays, CT scans, and ultrasound, while helpful in identifying the source of infection, are often difficult to detect abnormalities in the early stages of infection and have a certain lag, failing to reflect the dynamic changes of the disease in a timely manner.

[0004] Mounting evidence suggests that bile acids (BAs) may be potential regulators in the pathological process of sepsis. Recent studies indicate that cholestasis accompanied by BA metabolic disturbances is a very early event in the pathological progression of sepsis, making dynamic changes in BAs a potential indicator for predicting the risk and prognosis of sepsis patients. For example, elevated serum total BA levels are significantly associated with an increased risk of death in pediatric sepsis patients. Currently, research on the impact of BAs on sepsis progression is still limited. It has been reported that chenodeoxycholic acid (CDCA) and deoxycholic acid (DCA) can induce sepsis-related damage, while taurine deoxycholic acid (HDCA) can reduce inflammation and protect the body from sepsis.

[0005] Given the intricate interactions between the gut microbiota and bile acid (BA) metabolism, studying the gut microbiota-bile acid axis is crucial for elucidating the physiological role of bile acids in the progression of childhood sepsis. The gut microbiota participates in a wide range of bile acid modifications through various processes, including debinding, 7α-dehydroxylation, oxidation, and epimerization, as well as the recently discovered amino acid binding mediated by bile salt hydrolases. Furthermore, microbial-derived secondary bile acids are reabsorbed into the liver via the enterohepatic circulation, where liver enzymes further modify these bile acids to produce tertiary bile acids with unique biological properties. Conversely, bile acids can also significantly influence the composition and function of the gut microbiota, thereby indirectly regulating host physiological functions. Integrative multi-omics analyses of bile acids and the gut microbiota provide a powerful tool for elucidating the gut microbiota-bile acid axis in the context of sepsis, thus comprehensively and reliably revealing the physiological effects of bile acids and microorganisms on sepsis.

[0006] Despite these advances, insufficient understanding of the gut microbiota-bile acid (BA) axis in sepsis hinders the development of diagnostic and treatment strategies. The complex and multidimensional characteristics of microbial species and the modifying properties of bile acids present significant challenges to elucidating the relationship between the microbiota-BA axis and sepsis. Currently, only a few multi-omics analyses attempt to elucidate the causal relationships between gut microbiota dynamics, bile acid metabolism, and sepsis outcomes. The diagnostic and therapeutic potential of the gut microbiota-BA axis, particularly in pediatric sepsis, which accounts for nearly half of all sepsis cases, remains largely unexplored. Summary of the Invention

[0007] To address the aforementioned technical problems, the present invention provides biomarkers for predicting or diagnosing sepsis in children, said biomarkers including Enterococcus raffinis.

[0008] In one embodiment, the markers also include Veillonella and Timon Rombutz.

[0009] In one embodiment, the marker further includes deoxycholic acid-3-sulfate.

[0010] In one embodiment, the present invention provides a kit for predicting or diagnosing sepsis in children, the kit comprising reagents for detecting Enterococcus raffinis in a sample to be tested.

[0011] In one embodiment, the kit includes reagents for detecting Veillonella and Timon Rombutz in a sample to be tested.

[0012] In one embodiment, the kit includes a reagent for detecting deoxycholic acid-3-sulfate in a sample to be tested.

[0013] In one embodiment, the use of a reagent for detecting Enterococcus raffinis in the preparation of a kit for predicting or diagnosing sepsis in children is provided.

[0014] In one embodiment, a medicament for treating sepsis in children is provided, the medicament comprising Enterococcus raffinis.

[0015] In one embodiment, the drug further includes deoxycholic acid-3-sulfate.

[0016] In one embodiment, the use of Enterococcus raffins in the preparation of a medicament for treating sepsis in children is provided.

[0017] This application integrates bile acid-targeted metabolomics and fecal metagenomics to conduct a dual cohort analysis of pediatric sepsis patients. We identified a broad metabolomic profile containing 99 bile acids, significantly expanding the known source library of bile acid metabolites in sepsis. Based on statistical analysis, machine learning modeling results, and independent cohort validation, we found *Enterococcus raffinis* and sulfated bile acid derivatives (especially deoxycholic acid-3-sulfate, DCA-3S, CAS No.: 67030-48-2, molecular formula: C) in pediatric sepsis patients. 24 H 40 O8S (molecular weight: 472.64 g / mol) was significantly enriched. In vitro and in vivo experiments confirmed that symbiotic gut microbiota (especially Enterococcus raffins) produced at least 80% of its total DCA-3S production, highlighting the dominant role of the microbiota in DCA-3S biosynthesis. Further experiments on septic mouse models and intestinal organoid models revealed that DCA-3S exerts its anti-septic protective effect by restoring intestinal barrier function and reducing inflammatory response. This application provides an important reference for the diagnosis and treatment of sepsis targeting the microbiota-bile acid axis. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments recorded in this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 The main coordinate analysis (PCoA) (Brey-Curtis distance) plot of bile acid profiles between children with sepsis and healthy controls was used, and the differences in bile acid profiles were analyzed using adonis (i.e., permutation multivariate analysis of variance, permanova). Figure 2This is a volcano diagram showing the accumulation of bile acids in children with sepsis and healthy controls; Figure 3 This is a ROC curve plot showing the diagnostic potential of bile acid features in childhood sepsis (feature selection, FS). Figure 4 This is a permutation importance score map of bile acid characteristics in the diagnostic model, showing only bile acids for which there are significant differences in levels between healthy controls and sepsis patients. The color indicates the enrichment of bile acids in healthy controls (blue) or sepsis patients (orange). Figure 5 In the validation cohort, the DCA-3S level was significantly higher in patients with sepsis compared with healthy controls. P values ​​were calculated using a two-sided Wilcoxon rank-sum test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001; Figure 6 This is a partial Spearman correlation plot (adjusted for age, sex, and BMI) between bile acid levels and clinical diagnostic indicators of sepsis (including PCT, FIB, hs_CRP, D_dimer, IL6, and IL10). An asterisk (*) indicates a significant association (P.adj < 0.05). The p-values ​​of multiple tests were adjusted using the Benjamini-Hochberg method. Figure 7 This is a graph validating the increase in DCA-3S levels in sepsis patients compared to healthy controls. P values ​​were calculated using a two-sided Wilcoxon rank-sum test. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001; Figure 8 This is a graph comparing DCA levels between healthy controls and children with sepsis in the cohort. The p-value was calculated using a two-sided Wilcoxon rank-sum test. Figure 9 The permutation importance score plot of the microbiome features in the diagnostic model of gut microbiota in children with sepsis only shows the microbiota with significant differences in abundance between the healthy control group and sepsis patients. The color represents the enrichment of microbiota in the healthy control group (blue) or sepsis patients (orange). Figure 10 The effect of DCA-3S-mediated Enterococcus raffinis on childhood sepsis is shown in the figure (P < 0.05) showing the significant correlation between the abundance of Enterococcus raffinis, Enterococcus virionensis, DCA-3S level and 6 clinical diagnostic indicators of sepsis. Red lines indicate positive correlation and blue lines indicate negative correlation. Figure 11Enterococcus raffinosus is the main bacterial producer of DCA-3S. The extracted ion chromatograms of DCA-3S standard (i), fecal sample (ii), and Enterococcus raffinosus (iii) at mass-to-charge ratio (m / z) of 471.3 to 96.9 are shown. Detailed Implementation

[0020] To enable those skilled in the art to better understand the technical solutions in this application, the present invention will be further described below with reference to embodiments. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this application.

[0021] Unless otherwise specified, the experimental methods used in the following specific embodiments are conventional methods. Unless otherwise specified, the materials and reagents used in the following specific embodiments are commercially available.

[0022] I. Queue Collection All human subjects in this study were from Wuhan Children's Hospital, Tongji Medical College, Huazhong University of Science and Technology. Two cohorts were included: a discovery cohort and a validation cohort. The discovery cohort, recruited in 2023, included 117 children with sepsis and 48 age-matched healthy controls; the validation cohort, recruited in 2024, included 49 children with sepsis and 47 age-matched healthy controls. The diagnosis of sepsis in children was based on the Sequential Organ Failure Assessment (SOFA) score. Inclusion criteria were: (i) age not exceeding 10 years, (ii) clinically diagnosed sepsis, and (iii) no history of malignant tumors. Children without sepsis symptoms were also recruited as a control group. Metadata (e.g., age, sex, body mass index) and clinical data (e.g., organ function assessment and inflammatory markers) were collected from participants through on-site assessment and clinical examination. All participants were Chinese citizens. All experimental protocols in this study underwent rigorous evaluation and approval by the Medical Ethics Committee of Wuhan Children's Hospital, Tongji Medical College, Huazhong University of Science and Technology (Approval No.: 2024R050-E01). Each participant (or their guardian) signed an informed consent form before participating in the study. Demographic analysis showed that the sepsis group and the control group were well-matched in terms of age (P=0.28, discovery cohort; P=0.54, validation cohort), sex distribution (P=0.07, discovery cohort; P=0.64, validation cohort), and body mass index (BMI) score (P=0.51, discovery cohort), providing a solid foundation for subsequent comparative analyses.

[0023] II. Extraction of bile acids from fecal samples To extract bile acids from human feces, a 30 mg fecal sample was first weighed and transferred to a 1.5 mL microcentrifuge tube. Next, 1 mL of ice-cold physiological saline was added to the tube, and homogenization was performed using an ULTRA-TURRAX® IKA T10 homogenizer (Wilmington, Delaware, USA) for 3 minutes. Then, 4 mL of ice-cold methanol was added to the tube, and the mixture was vortexed thoroughly. The mixed sample was incubated overnight at 4°C to complete the bile acid extraction process. After incubation, the sample was centrifuged at 6,000 × g at 4°C for 15 minutes, and the supernatant was collected. The supernatant was filtered using a nylon 66 syringe filter (13 mm, 0.22 µm), and the filtrate was dried under nitrogen. Finally, the dried extract was resuspended in an acetonitrile / water (1 / 9, 100 µL) solution containing deuterated cholic acid (cholic acid-2,2,4,4-d4, 100 ng / mL) as an internal standard. 5 µL of this solution was injected into a liquid chromatography-triple quadrupole mass spectrometry (LC-QqQ-MS) system for bile acid analysis. III. Targeted Bile Acid Analysis

[0024] Quantitative analysis of bile acids was performed using a liquid chromatography-tandem mass spectrometry (LC-MS / MS) system coupled with a Shimadzu LC-30AD ultra-high performance liquid chromatography system (Kyoto, Japan) and a Triple Quad™ 4500 mass spectrometer (ABSciex, Framingham, Massachusetts, USA). Chromatographic separation was performed using an ACQUITY UPLC BSH C18 column (50 mm × 2.1 mm, 1.7 μm, Waters, Milford, Massachusetts, USA) at 40°C with a flow rate of 0.4 mL / min. Mobile phase A was 0.1% formic acid aqueous solution, and mobile phase B was acetonitrile. Linear gradient elution was used as follows: 0–1 min, 10% B; 1–11 min, 10–90% B; 11–12 min, 90–100% B; 12–13 min, 100–10% B. Reequilibration was then performed for 5 min at 10% B. All analytes were detected using multiple reaction monitoring (MRM) mode. LC-MS / MS operation control was performed using Analyst® software (version 1.6.3), and quantitative analysis was performed using MultiQuant™ software (version 3.0.2). IV. Statistical Analysis

[0025] All statistical analyses were performed using R (version 4.3.2). The Shapiro-Wilk normality test was used to determine the normality of the data distribution. Principal coordinate analysis based on Bray-Cutis distance (PCoA) was used to analyze the differences in bile acid abundance between healthy controls and children with sepsis. Permuted multivariate ANOVA (also known as ADONIS, R package vegan, version 2.6-4) was used to compare whether there were significant differences in bile acid profiles between the two groups. For statistical comparisons, two-tailed t-tests (for normal probability distributions) or Wilcoxon rank-sum tests (for outlier probability distributions) were used to compare continuous variables. The Benjamini-Hochberg method was used to adjust for p-values ​​in multiple tests. P < 0.05 was considered significant.

[0026] V. Experimental Results This invention employs advanced liquid chromatography-multiple reaction monitoring mass spectrometry (LC-MRM-MS) to analyze the bile acid metabolomic profile in children's fecal samples. Using this method, we successfully identified 92 bile acids, including 44 free bile acids, 42 amino acid-bound bile acids, 5 glucuronidated bile acids, and 1 acetylated bile acid, and accurately determined their levels. This comprehensive bile acid profiling provides rich data support for subsequent biomarker screening.

[0027] Based on the obtained bile acid (BA) profiles, we analyzed the differences in BA between pediatric sepsis patients and healthy individuals. Principal coordinate analysis (PCoA) provided a panoramic overview of the BA profiles, showing a significant separation in BA between the two groups. Figure 1 Permutational multivariate ANOVA, Permanento test, P=1.00e-03. Difference analysis (Wilcoxon-P.adj<0.05, |log2FC|>1) showed significant differences in 19 types of BA between the two groups, of which 17 types of BA remained significant after adjusting for age, sex, and BMI using a generalized linear model (GLM). Figure 2Integrating these findings with partial least squares discriminant analysis (PLS-DA) results, we obtained a sepsis-specific BA profile comprising 13 unique BAs, as detailed in Table 1, which identifies enriched bile acids in pediatric sepsis patients and healthy controls. Among these 13 BAs, sulfated BAs constituted a significant subgroup, accounting for 23.08% (3 / 13) of sepsis-associated BAs. Specifically, deoxycholic acid-3-sulfate (DCA-3S) (log2FC=1.72, P.adj=2.22e-05), 3-sulfoglycocholic acid (Gly-DCA-3S; log2FC=2.70, P.adj=6.95e-04), and 3-sulfoglycodeoxycholic acid (Gly-DCA-3S; log2FC=2.99, P.adj=8.97e-04) consistently showed elevated levels in sepsis patients. These results suggest a potential link between BA sulfation and sepsis pathology.

[0028] Table 1 bile acids GLM P value Wilcox uncorrected p-value Wilcox correction P-value VIP Points log2FC norCA 0.054899899 0.023035576 0.056474315 1.046000281 0.57410066 CA-3G 0.00239774 3.72287E-05 0.000282938 1.56623685 2.483262948 Gly-CA 0.249089148 0.321846274 0.452968831 0.755480951 1.038878001 UDCA-3S 0.918304048 0.013708762 0.037209496 0.144205564 0.11930319 Ala-CA 0.174053542 0.017943012 0.047023066 0.954308416 -1.51048555 CA-7S 0.296782241 0.878805808 0.908339525 0.302284247 0.209023963 TUDCA / THDCA 0.291083718 0.526617016 0.678354123 0.58846685 1.639545578 T-α-MCA / T-β-MCA 0.816926167 0.688178253 0.804639188 0.292680597 0.306667425 TCA 0.139730621 0.15225965 0.251559422 1.098922949 2.931026965 DHCA 0.533797848 0.488563233 0.640186306 0.301546563 -0.28336216 N6-Lys-CDCA 0.025139801 0.001686493 0.00712075 0.723336442 2.311993485 N6-Lys-DCA 0.28465638 0.83058433 0.876727903 0.536457316 1.025943124 Gly-CDCA-3S 0.042596675 0.000141641 0.000897062 1.488328418 2.989594827 Gly-DCA-3S 0.045960074 0.000100655 0.000695434 1.369204518 2.696888755 Gln-β-MCA 0.555694039 0.068880148 0.137760296 0.079966654 0.116377694 Gln-CA 0.222788918 0.803845491 0.876727903 0.385578947 0.356405863 Glu-DCA 0.333141572 0.124849564 0.220664345 0.473911195 -0.96023151 7,12-KLCA 0.086616256 0.004860533 0.017590501 0.776456841 -0.72040517 isoalloLCA 0.179295599 0.786486667 0.876727903 0.698622257 0.270010994 6-KLCA 0.004233582 7.09391E-07 1.34784E-05 1.277161358 -1.26728163 7-KLCA 0.03034665 0.003576759 0.013591683 1.095050152 -0.69477844 MDCA 0.001389116 0.001656073 0.00712075 1.769956193 -1.36684185 β-UDCA 0.001588223 0.008358954 0.025411219 1.835566557 -1.56572929 β-HDCA 0.019777489 0.120102973 0.220664345 0.912766966 -0.48852548 UDCA / HDCA 0.001203503 0.012584416 0.035422801 1.465970017 -0.76757766 3β-DCA / 12β-DCA 1.30667E-05 4.32251E-06 5.47518E-05 1.947332526 -1.12395154 3-oxoCA 0.411026151 0.179072909 0.283532105 0.623823661 -0.3257498 HCA 0.700421126 0.274625678 0.393802859 0.112645261 -0.05652307 7-oxoCA / 3-oxoACA / 12-oxoCA 0.465994418 0.231494875 0.345729568 0.453470635 -0.21275109 ACA 0.025711167 0.151769565 0.251559422 0.979069329 0.445308106 CA 0.008663364 0.064906579 0.133321622 1.098811076 0.384553806 Gly-CDCA 0.483354501 0.884435853 0.908339525 0.544968139 0.655804713 Gly-DCA 0.960497356 0.337174951 0.465914478 0.107818873 0.262745353 coproCA 0.548184982 0.000858828 0.004079431 0.158239471 -0.20727218 Ala-DCA 0.979209671 0.12471471 0.220664345 0.055187163 0.094270842 CDCA-3S 0.003021338 0.025288527 0.059160882 1.280920578 0.796035308 DCA-3S 0.00111635 1.46051E-06 2.21998E-05 1.61196443 1.719989506 TCDCA 0.085324694 0.05558447 0.120697707 1.282293772 2.95452066 12β-CA / UCA 0.006649375 0.020760637 0.052593614 1.374925159 -0.69984758 TDCA 0.760352857 0.628035693 0.753509927 0.280555099 0.858036155 Val-CA 0.509093079 0.608253437 0.753509927 0.346829365 0.49760128 Gly-LCA-3S 0.563841369 0.006124793 0.020238445 0.402952282 0.523601109 Gln-DCA 0.613823615 0.92405348 0.936374193 0.31125122 1.612875483 Leu-CA 0.236929374 0.176988142 0.283532105 0.688619516 1.243150765 Phe-CA 0.050905631 0.009552137 0.027921631 1.079957469 1.189900515 CDCA-3G / DCA-3G / CDCA-24G 0.001409075 7.02805E-07 1.34784E-05 1.621099344 2.380659712 ω-MCA / 3β-CA 0.050688503 0.007242743 0.022935354 1.031044468 -0.53261299 △5-isoLCA 0.119691187 0.62015925 0.753509927 0.744924829 0.913374976 △11-LCA 0.214424895 0.000572309 0.003106818 0.668510176 -0.74256467 3-dehydroLCA 0.000919974 9.57396E-06 9.09526E-05 1.920584185 -1.9650267 isoLCA 0.092782285 0.000769936 0.00390101 0.677434411 -0.98002029 alloLCA 0.000465273 0.002436175 0.009744702 1.51544242 -1.51640855 LCA 6.02387E-05 8.2778E-06 8.98733E-05 1.790624224 -1.52997295 norDCA 0.524544844 0.349945138 0.474925544 0.074898017 0.048055428 dioxo-HDCA / dioxo-CDCA 0.01983429 6.41786E-07 1.34784E-05 1.131495408 -1.27804468 12-KLCA 0.000387875 0.005670933 0.019590495 1.885252202 -1.48024715 apoCA 0.982152232 0.556266462 0.704604185 0.110525197 -0.11104611 α-MCA 0.772935558 0.789244752 0.876727903 0.049543186 0.02595519 3-oxoCDCA 0.333763903 0.826763279 0.876727903 0.472349095 0.536028225 3-oxoDCA 0.000616405 2.43222E-05 0.000205387 1.638830254 -0.87584092 CDCA 0.011389889 0.092094359 0.17946593 1.095196614 0.420682665 DCA 0.416878852 0.231494875 0.345729568 0.726770683 0.397871797 isoDCA 0.017515384 4.20627E-08 3.19677E-06 1.391683185 -1.37821059 LCA-3S 0.161570882 0.034431414 0.076964337 0.713470684 0.77315298 TLCA 0.000199928 0.000451365 0.002638746 2.195063018 -2.74783632 β-MCA 0.90239617 0.058415458 0.123321522 0.164986506 -0.11500328 Val-CDCA 0.814104637 0.829245234 0.876727903 0.073992912 0.21980773 Val-DCA 0.209622662 0.114614875 0.217768263 0.149754328 -0.20455447 Leu-CDCA 0.552691966 0.826262455 0.876727903 0.347548927 0.989966551 Leu-DCA 0.829824567 0.634534676 0.753509927 0.137029007 0.208495332 Phe-CDCA 0.252672894 0.025688278 0.059160882 0.634859924 1.026182333 Phe-DCA 0.978700973 0.937907946 0.937907946 0.496812348 0.804869746 Gly-HDCA / Gly-UDCA 0.915715583 0.236551809 0.345729568 0.207084013 0.593852913 Tyr-DCA 0.005944127 0.149231092 0.251559422 1.037530877 -1.20483285 Trp-CDCA 0.168448157 0.426782547 0.569043397 0.758489022 1.266092882 Trp-DCA 0.127026402 0.235911643 0.345729568 0.172777169 -0.22850712 .

[0029] Given the significant differences in bile acid (BA) profiles between sepsis patients and healthy individuals, we investigated the diagnostic potential of BA profiles for childhood sepsis. Based on individual BA levels, we constructed a random forest model to distinguish between childhood sepsis patients and age-matched healthy controls. Figure 3 A greedy feature selection method was employed to remove irrelevant features, thereby improving the model's diagnostic performance. Notably, DCA-3S became the most significant predictor in the final diagnostic model, with an area under the curve (AUC) of 0.91 in the discovery cohort. When applied to the independent validation cohort, the model maintained robust diagnostic accuracy, with an AUC of 0.85. Figure 3 Table 2 shows the diagnostic efficacy of the model based on microbial and bile acid biomarkers in the discovery and validation cohorts. Furthermore, BA profile analysis revealed that 8 BAs in the validation cohort exhibited the same significant alterations as the 13 significantly altered BAs in the discovery cohort (Table 3 shows that 8 of the 13 differentially abundant bile acids (including DCA-3S) identified in the discovery cohort were validated in the validation cohort). Importantly, deoxycholic acid-3-sulfate (DCA-3S) showed persistent and significant enrichment in sepsis patients in both cohorts. Figure 5 ).

[0030] Table 2 Model queue accuracy Sensitivity Specificity Accuracy Recall rate F1 value AUC AUPR bile acid model Discovery queue 0.7455 0.7009 0.8542 0.9213 0.7009 0.7961 0.8560 0.9261 Validation queue / / / / / / / / Bile acid model (feature screening) Discovery queue 0.8545 0.8889 0.7708 0.9043 0.8889 0.8966 0.9088 0.9607 Verification queue 0.8125 0.7959 0.8298 0.8298 0.7959 0.8125 0.8506 0.8772 Microbiome model Discovery queue 0.7576 0.7094 0.8750 0.9326 0.7094 0.8058 0.8567 0.9341 Verification queue / / / / / / / / Microbiome Model (Feature Screening) Discovery queue 0.8242 0.7778 0.9375 0.9681 0.7778 0.8626 0.9045 0.9620 Verification queue 0.7917 0.7143 0.8723 0.8537 0.7143 0.7778 0.8545 0.8827 Microbiome + Bile Acid Combined Model Discovery queue 0.7939 0.7607 0.8750 0.9368 0.7607 0.8396 0.8780 0.9406 Verification queue / / / / / / / / Microbiome + Bile Acid Combined Model (Feature Screening) Discovery queue 0.8545 0.8376 0.8958 0.9515 0.8376 0.8909 0.9257 0.9681 Verification queue 0.8438 0.8163 0.8723 0.8696 0.8163 0.8421 0.8934 0.9169 .

[0031] Table 3 .

[0032] Furthermore, we explored the association between plasma protein A (BA) profiles and clinical indicators of sepsis. Random forest regression analysis showed that the BA profile had reliable predictive power for multiple clinical indicators, especially high-sensitivity C-reactive protein (hs-CRP), interleukin-6 (IL-6), and D-dimer levels. Correlation analysis further confirmed that each BA was significantly correlated with clinical diagnostic indicators of sepsis. Figure 6 Notably, DCA-3S was significantly positively correlated with all six clinical endpoints, including procalcitonin (PCT, adjusted P=3.30e-03), fibrinogen (FIB, adjusted P=2.71e-03), hs-CRP (adjusted P=5.68e-03), D-dimer (adjusted P=1.64e-02), IL-6 (adjusted P=9.63e-04), and interleukin-10 (IL-10, adjusted P=2.61e-04). In contrast, the precursor of DCA-3S, DCA, was not significantly correlated with any of these clinical endpoints. Figure 7 ).

[0033] The altered bile acid (BA) profile (primarily DCA-3S) observed in pediatric sepsis patients suggests that BA may be involved in the progression of sepsis. Given the lack of a significant difference in DCA levels between sepsis patients and healthy controls (P = 0.23), the elevated DCA-3S levels may be attributed to the upregulation of sulfation modification. Figure 8 ).

[0034] In summary, principal coordinate analysis (PCoA) based on Bray-Cutis distance revealed significant differences in bile acid abundance between healthy controls and children with sepsis, with a significant difference in PCOA2 between the two groups. This finding suggests that bile acid metabolism may play an important role in the development and progression of sepsis. Further two-sided Wilcoxon rank-sum test analysis revealed significantly higher DCA-3S abundance in children with sepsis compared to healthy controls, thus establishing DCA-3S as a biomarker.

[0035] VI. Gut microbiota dysbiosis was observed in pediatric sepsis patients. Given the extensive involvement of the gut microbiota in bile acid alterations, changes in the gut microbiota of sepsis patients were subsequently examined. Metagenomic analysis revealed a significant dysregulation in the gut microbiota composition of pediatric sepsis patients. Fecal metagenomic sequencing revealed a high-dimensional microbial community composition in sepsis patients, comprising 496 genera and 1,165 species. Macroscopic analysis showed that α-diversity in sepsis patients was significantly lower than in healthy controls, manifested as decreased Shannon diversity (P=6.20e-08), Simpson diversity (P=4.50e-08), and species richness (P=1.90e-04), along with decreased evenness (P=8.20e-08, Pielou evenness). Principal coordinate analysis of β-diversity (PCoA) further revealed differential microbial composition between sepsis patients and controls (P=1.00e-03, Permanova test), manifested as significant separation along the PC2 axis (P=1.71e-06). At the genus level, sepsis patients were characterized by enrichment of *Enterococcus* and *Rothia*, while the healthy control group was dominated by *Blautia*, *Ruminococcus*, *Mediterraneibacter*, and *Erysipelatoclostridium*. Notably, in the comparison between sepsis patients and the control group, *Enterococcus* showed the highest linear discriminant analysis (LDA) score (LDA = 4.90) (Table 4). Table 4 shows the LEfSe analysis results at the genus level in the cohort; this is consistent with previous studies on *Enterococcus* as a key biomarker for hospital-acquired sepsis.

[0036] Table 4 Taxonomic name enrichment group LDA score Uncorrected P-value Correction P value g__Enterococcus sepsis 4.89767778 6.945E-05 0.0011165 g__Blautia Comparison 4.33477266 2.786E-06 0.0001096 g__Ruminococcus Comparison 4.09633691 0.0127802 0.0350713 g__Mediterraneibacter Comparison 4.07734292 0.0003167 0.0024153 g__Erysipelatoclostridium Comparison 3.99676901 0.0046311 0.0176281 g__Escherichia Comparison 3.95026204 0.0100078 0.0298067 g__Anaerostipes Comparison 3.93369979 1.749E-07 2.064E-05 g__Hungatella Comparison 3.93117242 0.0151062 0.0396117 g__Prevotella Comparison 3.87254441 0.005628 0.0201243 g__Roseburia Comparison 3.84673107 0.0069976 0.0235918 g__Citrobacter Comparison 3.82819435 0.0039794 0.0161922 g__Tyzzerella Comparison 3.75826954 0.0002415 0.0021974 g__Fusicatenibacter Comparison 3.75494939 0.001579 0.0083531 g__Clostriaceae_unclassified Comparison 3.7174935 0.0021743 0.0106902 g__Klebsiella Comparison 3.65075702 0.0010197 0.0066844 g__Ruthenibacterium Comparison 3.61839301 9.462E-05 0.0011165 g__Lacrimispora Comparison 3.60909044 0.0026556 0.0124795 g__Anaerobutyricum Comparison 3.52648044 0.0002614 0.0022033 g__Collinsella Comparison 3.47494554 0.0147597 0.0395828 g__Clostridium Comparison 3.34130693 3.231E-05 0.0006354 g__Veillonella Comparison 3.33994535 0.0005852 0.004062 g__Enterocloster Comparison 3.2804472 0.0073663 0.024145 g__Lachnospira Comparison 3.26965613 0.0012359 0.007292 g__Lachnoclostridium Comparison 3.2357387 0.0042557 0.0167392 g__Rothia sepsis 3.10423787 0.0048 0.0177001 g__Megasphaera Comparison 2.9098384 0.0002421 0.0021974 g__Flavonifractor Comparison 2.8420724 2.16E-06 0.0001096 g__Dialister Comparison 2.69465891 0.0010956 0.0068045 g__Cryobacterium Comparison 2.68372478 0.0094405 0.0293152 g__Eubacteriales_unclassified Comparison 2.61679282 2.287E-05 0.0005397 g__Agathobaculum Comparison 2.6154902 0.0163342 0.0412996 g__Intestinibacter Comparison 2.44031113 2.188E-05 0.0005397 g__Clostridia_unclassified Comparison 2.31148135 0.0106715 0.0302952 g__Acinetobacter Comparison 2.30612366 8.924E-05 0.0011165 g__Faecalicatena Comparison 2.2903673 0.0035375 0.014908 g__GGB79916 Comparison 2.01043607 0.0087531 0.0279154 .

[0037] We further explored species-specific microbial alterations in childhood sepsis. Using linear discriminant analysis of effect size (LEfSe) (LDA value > 2.0, adjusted P < 0.05), 42 bacterial species with differential abundance between the sepsis group and the healthy group were identified from 1,165 annotated microbial species. After adjusting for potential confounding factors (age, sex, and BMI) using a multivariate linear model (MaAsLin2), 3 sepsis-enriched species and 35 control-enriched species remained statistically significant (adjusted P < 0.05). Specifically, in children with sepsis, the abundance of *Enterococcus raffinosus* (LDA = 3.95) and *Enterococcus faecalis* (LDA = 4.43) was significantly increased, while the abundance of various β-lactamase-metabolizing bacteria (including *Anaerostipes hadrus*, *Ruminococcus gnavus*, and *Ruminococcus torques*) was significantly decreased. A similar pattern of altered gut microbiota was observed in the validation cohort, with a significant increase in the abundance of *Enterococcus* species (especially *Enterococcus raffinosus*) in sepsis patients (LDA = 3.33).

[0038] Alterations in the microbial community were significantly correlated with clinical indicators in sepsis patients. A random forest model based on quantitative microbial signatures demonstrated good predictive performance on multiple clinical indicators. The identified microbial species were highly correlated with six clinical diagnostic indicators of sepsis. We further evaluated the potential of these species as biomarkers for childhood sepsis. The microbial model achieved an AUC of 0.90 in the discovery cohort and 0.85 in the validation cohort. Veillonella parvula and Romboutsia timonensis species enriched in the healthy control group, and E. raffinosus species enriched in sepsis patients, showed importance in the diagnostic model. Figure 9 Furthermore, the integrated model combining bile acid analysis (BA) with microbial biomarkers achieved optimal diagnostic performance, reaching AUC values ​​of 0.93 and 0.89 in the discovery and validation cohorts, respectively.

[0039] VII. The role of DCA-3S-mediated Enterococcus raffinis in childhood sepsis Notably, sepsis-associated gut microbiota was significantly associated with bile acids (BAs), particularly DCA-3S, further confirming their crucial role in sepsis progression. Procrustes analysis showed a significant correlation between bile acid profiles and microbial profiles (P = 1.00e-03). Furthermore, Spearman analysis revealed significant correlations between bile acid abundance and microbial abundance (P = 5.86e-11) or microbial diversity (P = 1.44e-08). We further analyzed the correlation between sepsis-associated microbial species and individual bile acid levels, finding that the abundance of several species, including *V. parvula* and *Enterococcus raffinosus*, was significantly correlated with DCA-3S levels. Based on the significant Spearman correlation results (P < 0.05), we constructed a species correlation network between *V. parvula*, *Enterococcus raffinosus*, DCA-3S, and six clinical diagnostic indicators. The network showed that DCA-3S levels were associated with both the abundance of the two microorganisms and sepsis marker values. Figure 10 This study explored the complex interactions between gut microbiota, bile acids, and sepsis through mediation analysis. The results showed that bile acids mediated most of the computer-simulated associations between the gut microbiota and sepsis. Notably, DCA-3S mediated the effects of *Enterococcus raffinis* on sepsis, particularly on fibrinogen (FIB) levels. This mediating effect was also confirmed in the validation cohort, suggesting that *Enterococcus raffinis* participates in the biosynthesis of DCA-3S, thereby influencing the progression of sepsis.

[0040] 8. *Enterococcus raffinis* is the main bacterial producer of DCA-3S. To determine whether *Enterococcus raffinis* or other bacterial species are involved in the formation of DCA-3S, we conducted a series of in vitro and in vivo bacterial culture experiments. Bacterial species were isolated from a fecal sample of a pediatric sepsis patient with the highest DCA-3S level and compared with… 13 C-DCA was co-cultured in the culture medium. After 24 hours of anaerobic culture, we detected DCA-3S in the culture medium, indicating that the gut microbiota successfully sulfated DCA to DCA-3S. Figure 11 Through 16S rRNA gene sequencing, we identified 27 bacterial species belonging to 4 different bacterial phyla from the cultured isolates. Subsequent experiments showed that 15 of the 27 bacterial species, alone or in combination with... 13C-DCA co-culture demonstrated the ability to sulfate DCA to DCA-3S. Notably, *Enterococcus raffinis* strains exhibited the highest sulfation efficiency, producing more than twice the amount of DCA-3S compared to other strains. Temporal analysis of the *Enterococcus raffinis* culture process further revealed a continuous and gradual accumulation of DCA-3S, eventually reaching a fixed concentration. This temporal accumulation pattern of DCA-3S confirms its dynamic synthesis during culture, thus eliminating potential interference from trace amounts of bile acids (BAs) in the culture medium. In addition to DCA, in vitro co-culture experiments also showed that *Enterococcus raffinis* can sulfate CDCA to form CDCA-3S, while no sulfation was detected in other bile acids tested under the same conditions. Furthermore, analysis of the *Enterococcus raffinis* reference genome identified a gene encoding an aryl sulfate sulfatase, which may be involved in the DCA sulfation process. Functional annotation showed that this gene is present in the genomes of 11 species among 27 bacterial isolates. Of these 11 species, 8 showed strains that converted sulfated DCA into DCA-3S, including Bifidobacterium pseudocatenulatum, which showed the second-highest DCA-3S production.

[0041] Given previous studies reporting that BA sulfation depends on hepatic enzymes, we investigated the contributions of the liver and gut microbiota to DCA-3S biosynthesis. In vitro experiments showed that no DCA-3S was generated in mouse liver homogenate at physiological DCA concentrations (approximately 0.45 nmol / g) (<0.01 nmol / g; limit of detection, LOD). Co-incubation of liver homogenate with DCA-3S ruled out the possibility of hepatic degradation of DCA-3S (P=0.88). Only when the DCA concentration was artificially increased to 50 times above physiological levels (approximately 30 nmol / g) could hepatic enzymes produce measurable amounts of DCA-3S (approximately 0.05 nmol / g). In vivo metabolomics analysis indicated that although DCA is widely distributed in multiple organs, under normal physiological conditions, DCA-3S is only present in the ileum, cecum, and colon (microbially colonized intestinal sites). The absence of DCA-3S in the liver and gallbladder challenges the long-held view that the liver is the primary site of DCA-3S biosynthesis.

[0042] To investigate the dynamics of DCA-3S generation in vivo, we conducted a DCA gavage experiment in a mouse model. Although the hepatic DCA concentration increased 16-fold after gavage, the DCA-3S level in the mouse liver remained undetectable (<0.01 nmol / g). In contrast, DCA-3S showed significant accumulation in the mouse intestine, particularly in the cecum (0.19 ± 0.07 nmol / g) and colon (0.12 ± 0.04 nmol / g), where the content increased by 1.47-fold and 2.42-fold, respectively, compared to baseline (control group). Since the hepatic DCA-3S concentration remained undetectable even after manually increasing DCA levels, we conservatively set the limit of detection (LOD) at the hepatic DCA-3S concentration (0.01 nmol / g), estimating the upper limit of hepatic DCA-3S content to be 9.64e-03 nmol. Even so, the hepatic DCA-3S content was still significantly lower than the intestinal DCA-3S content (24.05e-03 nmol was detected in the control group), with the latter accounting for 71.39% of the total DCA-3S content. In the DCA gavage group, the intestinal DCA-3S content further increased to 43.45e-03 nmol, accounting for 81.84% of the total DCA-3S content.

[0043] We also used germ-free mice in our experiments to determine the microbial origin of DCA-3S in a more rigorous manner. The germ-free mice were divided into four groups: a germ-free control group, a control group treated with... 13 Germ-free mouse groups with C-DCA, germ-free mouse groups colonized with Enterococcus syringae, and mouse groups colonized with Enterococcus syringae before being given 13 Germ-free mice group receiving C-DCA. As expected, administered alone... 13 Neither C-DCA nor isolated Enterococcus raffinis was detected. 13 The generation of C-DCA-3S was observed. In contrast, germ-free mice colonized with *Enterococcus syringae* and administered 13C-DCA showed significantly reduced [the production of C-DCA-3S]. 13 The C-DCA-3S peak. These results collectively indicate that DCA-3S is produced by the gut microbiota, confirming that the gut microbiota (rather than the liver) is the main contributor to DCA-3S biosynthesis.

[0044] 9. DCA-3S supplementation can reduce the progression of sepsis in mice. To elucidate the impact of elevated DCA-3S levels on the progression of sepsis in children, we conducted a series of experiments in a mouse model. A sepsis model was established in 4-week-old mice using cecal ligation and perforation (CLP). Compared to the sham-operated control group, sepsis-affected mice exhibited significantly elevated DCA-3S levels, accompanied by significant gut microbiota dysbiosis. The elevated DCA-3S levels in both human and mouse sepsis provide a basis for further mechanistic studies using mouse models.

[0045] We further investigated whether DCA-3S administration could alleviate or exacerbate the progression of sepsis in children. In a treatment experiment, CLP-induced sepsis mice were randomly assigned to two groups: a sepsis group treated with PBS (phosphate-buffered saline) and a DCA-3S treatment group treated with DCA-3S for 4 days, with healthy mice serving as controls. Notably, CRP levels in the DCA-3S treatment group were significantly lower than those in the PBS-treated sepsis group (P=2.90e-02). Similarly, PCT levels, another sepsis marker, also decreased after DCA-3S treatment, although they remained higher in the DCA-3S treatment group than in the healthy control group. Hematoxylin and eosin (H&E) staining of intestinal tissues showed that DCA-3S treatment effectively improved CLP-induced intestinal barrier disruption, as evidenced by preserved villous structure integrity and reduced inflammatory cell infiltration. Furthermore, gut microbiota analysis indicated that DCA-3S treatment partially restored microbial diversity and homogeneity. Notably, DCA-3S treatment significantly increased the levels of *Akermansia myxophilus* (…). Akkermansia muciniphila) Gordon's parabacterium ( Parabacteroides gordonii) The abundance of bacteria in the family Muribauculaceae, which are widely recognized as important or potential probiotics, was also observed. Previous studies have highlighted the important role of the gut microbiota in regulating intestinal barrier function. The observed restoration of microbial homeostasis may also contribute to maintaining barrier integrity and may help alleviate sepsis.

[0046] It should be understood that the disclosed invention is not limited to the specific methods, schemes, and substances described, as these are all subject to variation. It should also be understood that the terminology used herein is for the purpose of describing specific embodiments only and is not intended to limit the scope of the invention, which is limited only by the appended claims.

[0047] Those skilled in the art will also recognize, or be able to identify, many equivalents of the specific embodiments of the invention described herein using no more than conventional experiments. These equivalents are also included in the appended claims.

Claims

1. A biomarker for predicting or diagnosing sepsis in children, characterized in that, The markers include Enterococcus raffinis.

2. The marker according to claim 1, characterized in that, The markers also include Veillonella and Timon Rombutz.

3. The marker according to claim 1, characterized in that, The markers also include deoxycholic acid-3-sulfate.

4. A kit for predicting or diagnosing sepsis in children, characterized in that, The kit contains reagents for detecting Enterococcus raffinis in the sample to be tested.

5. The reagent kit according to claim 4, characterized in that, The kit contains reagents for detecting Veillonella and Timon Rombutz in the sample to be tested.

6. The reagent kit according to claim 4, characterized in that, The kit contains reagents for detecting deoxycholic acid-3-sulfate in the sample to be tested.

7. The use of a reagent for detecting Enterococcus raffinis in the preparation of a kit for predicting or diagnosing sepsis in children.

8. A drug for treating childhood sepsis, characterized in that, The drug contains Enterococcus raffinis.

9. The medicament according to claim 8, characterized in that, The drug also includes deoxycholic acid-3-sulfate.

10. Application of Enterococcus raffinis in the preparation of drugs for treating sepsis in children.

Images