Acute-on-chronic liver failure (ACLF) and decompensated cirrhosis (DC) are life-threatening syndromes that can develop at the end-stage of chronic hepatitis B virus (HBV) infection. Both ACLF and DC are complicated by hepatic and extrahepatic pathogeneses. To better understand the compartment-specific metabolic modulations related to their pathogenesis, HBV-DC, HBV-ACLF patients, and controls (30 each) were analyzed by metabolomics using portal (Port), hepatic vein (Hep), and peripheral (Peri) serum. Compartment ratios of metabolites (RatioHep/Port, RatioPeri/Hep, and RatioPort/Peri) were calculated. The liver tissues (10 per group) were analyzed using transcriptomics and metabolomics. An additional 75 patients with ACLF, 20 with DC, and 20 with liver cirrhosis (LC) were used to confirm oxlipid dysregulation. Both multi-omics datasets suggest suppressed energy, amino acid, and pyrimidine metabolism in the ACLF/DC liver. The serum metabolomic variations were contributed primarily by disease rather than sampling compartments, as both HBV-ACLF and HBV-DC patients demonstrated abnormal profiles of amino acids and peptides, indoles, purines, steroids, and benzimidazoles. In ACLF/DC patients, impaired hepatic metabolism resulted in a highly correlated hepatic and portal vein serum metabolome and release of inflammatory lipids and heme metabolites from the liver. HBV-ACLF showed higher RatioPeri/Hep of extrahepatic inflammatory oxlipids, while HBV-DC patients showed higher RatioPort/Peri of gut microbial metabolites. An inflammatory oxlipid outburst was confirmed in the early stages of HBV-ACLF. The inflammatory effects of the selected oxlipids were confirmed in monocytes. These findings support a synergy between liver-specific mechanisms and systemic inflammation in ACLF/DC development, and that pro-inflammatory oxlipids are metabolic signatures of early HBV-ACLF.

Bacteroides intestinalis ( B. intestinalis ) is an abundant gastrointestinal commensal bacterium and is able to produce secondary bile acids (BAs) among other important metabolic functions. However, deoxycholic acid (DCA) is known to suppress Bacteroides , suggesting differential molecular impact of different BA species on Bacteroides . Among major human gastrointestinal BA components, we first demonstrated that DCA and chenodeoxycholic acid (CDCA) and their taurine-conjugated species at 1 mM showed significantly higher inhibitory effects on the growth of B. intestinalis than cholic acid (CA) and lithocholic acid (LCA) and their taurine-conjugated species. Then, high-throughput proteomic strategy was used to show that both TCDCA and TDCA caused more proteome-wide modulation than TCA and TLCA. In response to incremental BA toxicity, the main functional changes of B. intestinalis include enhanced protein synthesis, DNA integrity maintenance, and suppressed central metabolic activities. Importantly, key energy and BA metabolism enzymes of B. intestinalis were inhibited by TCDCA and TDCA. These findings provide a basis for future studies to explore how Bacteroides respond to bile stress and how BA composition modulate gut microbiome homeostasis.

Aging is a major risk for developing cardiac and skeletal muscle dysfunction, yet the underlying mechanism remains elusive. Here we demonstrated that the mitochondria-associated endoplasmic reticulum membranes (MAMs) in the rat heart and skeletal muscle were disrupted during aging. Using quantitative morphological analysis, we showed that the mitochondria-endoplasmic reticulum contacts (MERCs) were reduced by half over the lifespan with an early onset of accelerated thickening in the clefts. The ultrastructural changes were further validated by proteomic profiling of the MAM fractions. A combination of subcellular fractionation and quantitative mass spectrometry identified 1306 MAM-enriched proteins in both heart and skeletal muscle, with a catalog of proteins dysregulated with aging. Functional mapping of the MAM proteome suggested several aging signatures to be closely associated with the ER-mitochondria crosstalk, including local metabolic rewiring, calcium homeostasis imbalance, and impaired organelle dynamics and autophagy. Moreover, we identified a subset of highly interconnected proteins in an ER-mitochondria organization network, which were consistently down-regulated with aging. These decreased proteins, including VDAC1, SAMM50, MTX1 and MIC60, were considered as potential contributors to the age-related MAM dysfunction. This study highlights the perturbation in MAM integrity during the striated muscle aging process, and provides a framework for understanding aging biology from the perspective of organelle interactions.

Long noncoding RNAs (lncRNAs) participate in host antiviral responses; however, how viruses exploit host lncRNAs for immune evasion remains largely unexplored. Functional screening of differentially expressed lncRNA profile in patients infected with influenza A virus (IAV) revealed that lncNSPL (Gene Symbol: LOC105370355) was highly expressed in monocytes. Deregulated lncNSPL expression in infected monocytes significantly increased type I interferon (IFN-I) production and inhibited IAV replication. Moreover, lncNSPL overexpression in mice increased the susceptibility to IAV infection and impaired IFN-I production. LncNSPL directly bound to retinoic acid-inducible gene I (RIG-I) and blocked the interaction between RIG-I and E3 ligase tripartite interaction motif 25 (TRIM25), reducing TRIM25-mediated lysine 63 (K63)-linked RIG-I ubiquitination and limiting the downstream production of antiviral mediators during the late stage of IAV infection. Our findings provide mechanistic insights into the means by which lncNSPL promotes IAV replication and immune escape via restricting the TRIM25-mediated RIG-I K63-linked ubiquitination. Thus, lncNSPL may represent a promising pharmaceutical target for anti-IAV therapy.

Bis (2-ethylhexyl)-2,3,4,5-tetrabromophthalate (TBPH) is an extensively used novel brominated flame retardant that is present ubiquitously in the environment and in biota. However, there is inadequate data on its potential hepatotoxicity to humans. In this study, high-coverage quantitative metabolomics based on 12C-/13C-dansylation labeling LC-MS was performed for the first time to assess the metabolic perturbations and underlying mechanisms of TBPH on human hepatocytes. HepG2 cells were exposed to TBPH at dosages of 0.1,1,10 μM for 24 or 72 h. Overall, 1887 and 1364 amine/phenol-containing metabolites were relatively quantified in cells and culture supernatant. Our results revealed that exposure to 0.1 μM TBPH showed little adverse effects, whereas exposure to 10 μM TBPH for 24 h enhanced intracellular protein catabolism and disrupted energy and lipid homeostasis-related pathways such as histidine metabolism, pantothenate and CoA biosynthesis, alanine, aspartate and glutamate metabolism. Nevertheless, most of these perturbations returned to the same levels as controls after 72 h of exposure. Additionally, prolonged TBPH exposure increased oxidative stress, as reflected by marked disturbances in taurine metabolism. This study sensitively revealed the dysregulations of intracellular and extracellular metabolome induced by TBPH, providing a comprehensive understanding of metabolic responses of cells to novel brominated flame retardants.

SCOPE:Using metabolomics to study the relations of nutrition and health requires stringent control of the experimental conditions used in an animal model. This work investigates the diet effects of autoclaved and irradiated feed on mouse urine and fecal metabolomics.METHODS AND RESULTS:C57BL/6 mice are fed normal-irradiation sterilized diet (n = 9), autoclave sterilized diet (n = 9), and high-irradiation sterilized diet (n = 9) for 4 weeks. Differential chemical isotope labeling liquid chromatography mass spectrometry is used to quantify the metabolome variations of urine and feces collected at five time points. Significant differences are observed in urine or fecal metabolomes of mice fed autoclaved diet versus mice fed high-irradiation diet or fed normal-irradiation diet, while the differences are small between the mice fed normal-irradiation and high-irradiation diet. Correlation studies of metabolite changes of diet- and aging-related biomarkers indicate a large overlap of significantly affected metabolites by the two factors.CONCLUSIONS:Diet can be a confounding factor that needs to be carefully considered when a metabolomics study is designed and metabolomic results of a mouse model of nutritional or other biological study are interpreted. Using the same sterilized diet for a given metabolomics project is essential to control the diet effect.

Marsupials are one of three major mammalian lineages that include the placental eutherians and the egg-laying monotremes. The marsupial brushtail possum is an important protected species in the Australian forest ecosystem. Molecules encoded by the MHC genes are essential mediators of adaptive immune responses in virus-host interactions. Yet, nothing is known about the peptide presentation features of any marsupial MHC class I (MHC I). This study identified a series of possum MHC I Trvu-UB*01:01 binding peptides derived from wobbly possum disease virus (WPDV), a lethal virus of both captive and feral possum populations, and unveiled the structure of marsupial peptide/MHC I complex. Notably, we found the two brushtail possum-specific insertions, the 3-aa Ile52Glu53Arg54 and 1-aa Arg154 insertions are located in the Trvu-UB*01:01 peptide binding groove (PBG). The 3-aa insertion plays a pivotal role in maintaining the stability of the N terminus of Trvu-UB*01:01 PBG. This aspect of marsupial PBG is unexpectedly similar to the bat MHC I Ptal-N*01:01 and is shared with lower vertebrates from elasmobranch to monotreme, indicating an evolution hotspot that may have emerged from the pathogen-host interactions. Residue Arg154 insertion, located in the α2 helix, is available for TCR recognition, and it has a particular influence on promoting the anchoring of peptide WPDV-12. These findings add significantly to our understanding of adaptive immunity in marsupials and its evolution in vertebrates. Our findings have the potential to impact the conservation of the protected species brushtail possum and other marsupial species.

SummaryMillions of people die from liver diseases annually, and liver failure is one of the three major outcomes of liver disease. The gut microbiota plays a crucial role in liver diseases. This study aimed to explore the effects of Lactobacillus casei strain Shirota (LcS), a probiotics used widely around the world, on acute liver injury (ALI), as well as the underlying mechanism. Sprague Dawley rats were intragastrically administered LcS suspensions or placebo once daily for 7 days before induction of ALI by intraperitoneal injection of D‐galactosamine (D‐GalN). Histopathological examination and assessments of liver biochemical markers, inflammatory cytokines, and the gut microbiota, metabolome and transcriptome were conducted. Our results showed that pretreatment with LcS reduced hepatic and intestinal damage and reduced the elevation of serum gamma‐glutamyltranspeptidase (GGT), total bile acids, IL‐5, IL‐10, G‐CSF and RANTES. The analysis of the gut microbiota, metabolome and transcriptome showed that LcS lowered the ratio of Firmicutes to Bacteroidetes; reduced the enrichment of metabolites such as chenodeoxycholic acid, deoxycholic acid, lithocholic acid, d‐talose and N‐acetyl‐glucosamine, reduce the depletion of d‐glucose and l‐methionine; and alleviated the downregulation of retinol metabolism and PPAR signalling and the upregulation of the pyruvate metabolism pathway in the liver. These results indicate the promising prospect of using LcS for the treatment of liver diseases, particularly ALI.

O-linked N-acetylglucosamine (O-GlcNAcylation) is a ubiquitous post-translational modification of proteins that is essential for cell function. Perturbation of O-GlcNAcylation leads to altered cell-cycle progression and DNA damage response. However, the underlying mechanisms are poorly understood. Here, we develop a highly sensitive one-step enzymatic strategy for capture and profiling O-GlcNAcylated proteins in cells. Using this strategy, we discover that flap endonuclease 1 (FEN1), an essential enzyme in DNA synthesis, is a novel substrate for O-GlcNAcylation. FEN1 O-GlcNAcylation is dynamically regulated during the cell cycle. O-GlcNAcylation at the serine 352 of FEN1 disrupts its interaction with Proliferating Cell Nuclear Antigen (PCNA) at the replication foci, and leads to altered cell cycle, defects in DNA replication, accumulation of DNA damage, and enhanced sensitivity to DNA damage agents. Thus, our study provides a sensitive method for profiling O-GlcNAcylated proteins, and reveals an unknown mechanism of O-GlcNAcylation in regulating cell cycle progression and DNA damage response.

The devastating coronavirus disease 2019 (COVID-19) pandemic has prompted worldwide efforts to study structural biological traits of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and its viral components. Compared to the Spike protein, which is the primary target for currently available vaccines or antibodies, knowledge about other virion structural components is incomplete. Using high-resolution mass spectrometry, we report a comprehensive post-translational modification (PTM) analysis of nucleocapsid phosphoprotein (NCP), the most abundant structural component of the SARS-CoV-2 virion. In addition to phosphoryl groups, we show that the SARS-CoV-2 NCP is decorated with a variety of PTMs, including N-glycans and ubiquitin. Based on newly identified PTMs, refined protein structural models of SARS-CoV-2 NCP were proposed and potential immune recognition epitopes of NCP were aligned with PTMs. These data can facilitate the design of novel vaccines or therapeutics targeting NCP, as valuable alternatives to the current vaccination and treatment paradigm that is under threat of the ever-mutating SARS-CoV-2 Spike protein.

Comprehensive analysis of the liver metabolome can be very useful for discovering disease biomarkers and studying diseases, especially liver-related diseases. However, the presence of a relatively large amount of blood in liver tissue may have a profound effect on liver tissue metabolome analysis. We designed a study to address this issue in order to develop a liver metabolomics workflow based on high-coverage quantitative metabolome analysis using differential chemical isotope labeling (CIL) LC-MS. In the first set of experiments, we compared the metabolomes of mouse serum, non-perfused liver, and perfused liver without and with varying amounts of blood added. We found that there was a significant metabolome difference between the perfused liver and non-perfused liver. To illustrate the effects of perfusion conditions on tissue metabolome analysis, we analyzed the mouse livers that were subjected to perfusion under two different conditions. We found that ice-cold temperature perfusion led to less change of the liver metabolome, compared to room temperature perfusion; however, there was still a significant metabolome difference between the ice-cold-perfused liver and the non-perfused liver. Finally, we applied the method to a chemical (carbon tetrachloride) exposure liver injury model to examine the effects of blood in liver on the detection of significantly changed metabolites in two comparative groups of mice. Using multivariate and univariate analyses of the serum and liver metabolomes of control and diseased mice, we detected many unique significant metabolites in serum as well as in liver. This work demonstrates that perfusion can alter the liver metabolome significantly. Therefore, we recommend the use of non-perfused liver for high-coverage liver metabolomics.

Chronic hepatitis B (CHB) remains a major public health problem globally. HBeAg seroconversion is a vital hallmark for the improvement of CHB. The plasma metabolic profile has not been clear in CHB patients and searching metabolic candidates to represent HBeAg seroconversion is also difficult currently. In this study, CHB patients were recruited, followed and divided into the HBeAg-positive (HBeAg-pos.) group (n = 29) and the HBeAg-negative (HBeAg-neg.) group (n = 29) based on HBeAg seroconversion or not. The plasma metabolic profiles were measured by gas chromatography-mass spectrometry (GC-MS) at 0 week (0w), 24 weeks (24w) and 48 weeks (48w) after administration. The acquired data was analyzed using orthogonal partial least squares discriminate analysis (OPLS-DA) and the differential metabolites were further assessed by self and group comparison. No differences of age, gender and serological characteristics were observed between two groups at 0w and 48w separately. The OPLS-DA score plots depending on administration time displayed robust metabolic differences no matter HBeAg turned to be negative or not. According to VIP> 1.0, a total of 15 differential metabolites were same in the two groups, 7 differential metabolites (glycolic acid, D-talose, L-proline, L-(-)-arabitol, ethyl-alpha-D-glucopyranoside, L-leucine and dihydroxybutanoic acid) were derived from one group alone and considered as metabolic candidates. At 0w versus (vs.) 24w, only 3 of 7 candidates (L-proline, L-(-)-arabitol, dihydroxybutanoic acid) showed nonuniform in the two groups, while at 0w vs. 48w, all of them varied inconsistently. Conclusively the dynamic metabolic profiles assayed by GC-MS were different between CHB patients with and without HBeAg seroconversion. The 7 metabolic candidates probably had the ability to reflect the CHB progression for HBeAg seroconversion and 3 of them showed strong relationship with HbeAg seroconversion early.

Metabolomics study of a biological system often involves the analysis of many comparative samples over a period of several days or weeks. This process of long-term sample runs can encounter unexpected instrument drifts such as small leaks in liquid chromatography-mass spectrometry (LC-MS), degradation of column performance, and MS signal intensity change. A robust analytical method should ideally tolerate these instrumental drifts as much as possible. In this work, we report a case study to demonstrate the high tolerance of differential chemical isotope labeling (CIL) LC-MS method for quantitative metabolome analysis. In a study of using a rat model to examine the metabolome changes during rheumatoid arthritis (RA) disease development and treatment, over 468 samples were analyzed over a period of 15 days in three batches. During the sample runs, a small leak in LC was discovered after a batch of analyses was completed. Reanalysis of these samples was not an option as sample amounts were limited. To overcome the problem caused by the small leak, we applied a method of retention time correction to the LC-MS data to align peak pairs from different runs with different degrees of leak, followed by peak ratio calculation and analysis. Herein, we illustrate that using(12)C-/C-13-peak pair intensity values in CIL LC-MS as a measurement of concentration changes in different samples could tolerate the signal drifts, while using the absolute intensity values (ie,C-12-peak as in conventional LC-MS) was not as reliable. We hope that the case study illustrated and the method of overcoming the small-leak-caused signal drifts can be helpful to others who may encounter this kind of situation in long-term sample runs.

Compared to conventional MS and NMR techniques, high-performance chemical isotope labeling (CIL) LC-MS provides accurate relative quantification of many more metabolites in biological samples. However, to apply this technique for urine and fecal metabolomics studies of animal models, the entire workflow, including the preanalytical process, needs to be strictly controlled to avoid or minimize quantitative errors. In this study, we report our investigation of the effects of mouse urine and fecal sample collection methods on CIL LC-MS metabolome analysis. Metabolic-cage collection and spot-sample collection of urine and feces were compared in a mouse model of CCl4-induced liver disease. 13C-/12C-dansylation LC-MS was used for quantitative profiling of the amine-/phenol-submetabolome changes. A total of 5026, 4963, 4238, and 4600 peak pairs or metabolites were detected from spot urine, spot feces, cage-collected urine, and cage-collected feces, respectively. It was found that samples collected using metabolic cages, widely used in low coverage metabolomics, could be contaminated with food as well as cross-specimen (urine in feces or feces in urine) to the extent that metabolomic comparison of different groups of mice could be seriously compromised in high-coverage metabolomics. In contrast, spot urine and spot feces could be collected without contamination and should be used in CIL LC-MS metabolomics.

We report a chemical isotope labeling (CIL) liquid chromatography mass spectrometry (LC-MS) method generally applicable for tracking metabolomic changes from samples collected in an animal model for studying disease development and treatment. A rat model of surgically induced osteoarthritis (OA) was used as an example to illustrate the workflow and technical performance. Experimental duplicate analyses of 234 plasma samples were carried out using dansylation labeling LC-MS targeting the amine/phenol submetabolome. These samples composed of 39 groups (6 rats per group) were collected at multiple time points with sham operation, OA control group, and OA rats with treatment, separately, using glucosamine/Celecoxib and three traditional Chinese medicines (Epimedii folium, Chuanxiong Rhizoma and Bushen-Huoxue). In total, 3893 metabolites could be detected and 2923 of them were consistently detected in more than 50% of the runs. This high-coverage submetabolome dataset could be used to track OA progression and treatment. Many differentiating metabolites were found and 11 metabolites including 2-aminoadipic acid, saccharopine and GABA were selected as potential biomarkers of OA progression and OA treatment. This study illustrates that CIL LC-MS is a very useful technique for monitoring incremental metabolomic changes with high coverage and accuracy for studying disease progression and treatment in animal models.

Blood is widely used for discovery metabolomics to search for disease biomarkers. However, blood sample matrix can have a profound effect on metabolome analysis, which can impose an undesirable restriction on the type of blood collection tubes that can be used for blood metabolomics. We investigated the effect of blood sample matrix 'on metabolome analysis using a high-coverage and quantitative metabolome profiling technique based on differential chemical isotope labeling (CIL) LC-MS. We used C-12-/C-13-dansylation LC:MS to perform relative quantification of the amine/phenol submetabolomes of four types of samples (i.e., serum, EDTA plasma, heparin plasma, and citrate plasma) collected from healthy individuals and compare their metabolomic results. From the analysis of 80 plasma and serum samples in experimental triplicate, we detected a total of 3651 metabolites with an average of 1818 metabolites per run (n = 240). The number of metabolites detected and the precision and accuracy of relative quantification were found to be independent of the sample type. Within each sample type, the metabolome data set could reveal biological variation (e.g., sex separation). Although the relative concentrations of some individual metabolites might be different in the four types of samples, for sex separation, all 66 significant metabolites with larger fold-changes (FC >= 2 and p < 0.05) found in at least one sample type could be found in the other types of samples with similar or somewhat reduced, but still significant, fold-changes. Our results indicate that CIL LC-MS could overcome the sample matrix effect, thereby greatly broadening the scope of blood metabolomics; any blood samples properly collected in routine clinical settings, including those in biobanks originally used for other purposes, can potentially be used for discovery metabolomics.

Chronic HBV (CHB) infected patients with intermediate necroinflammation and fibrosis are recommended to receive antiviral treatment. However, other than liver biopsy, there is a lack of sensitive and specific objective method to determine the necroinflammation and fibrosis stages in CHB patients. This study aims to identify unique serum metabolomic profile associated with histological progression in CHB patients and to develop novel metabolite biomarker panels for early CHB detection and stratification. A comprehensive metabolomic profiling method was established to compare serum samples collected from health donor (n = 67), patients with mild (G < 2 and S < 2, CHB1, n = 52) or intermediate (G ≥ 2 or S ≥ 2, CHB2, n = 36) necroinflammation and fibrosis. Multivariate models were developed to differentiate CHB1 and CHB2 from controls. A set of CHB-associated biomarkers was identified, including lysophosphatidylcholines, phosphatidylcholines, phosphatidylinositol, phosphatidylserine, and bile acid metabolism products. Stratification of CHB1 and CHB2 patients by a simple logistic index, the PIPSindex, based on phosphatidylinositol (PI) and phosphatidylserine (PS), was achieved with an AUC of 0.961, which outperformed all currently available markers. A panel of serum metabolites that differentiate health control, CHB1 and CHB2 patients has been identified. The proposed metabolomic biosignature has the potential to be used as indicator for antiviral treatment for CHB management.

Human urine and feces can be non-invasively collected for metabolomics-based disease biomarker discovery research. Because urinary and fecal metabolomes are thought to be different, analysis of both biospecimens may generate a more comprehensive metabolomic profile that can be better related to the health state of an individual. Herein we describe a method of using differential chemical isotope labeling (CIL) liquid chromatography mass spectrometry (LC-MS) for parallel metabolomic profiling of urine and feces. Dansylation labeling was used to quantify the amine/phenol submetabolome changes among different samples based on (12)C-labeling of individual samples and (13)C-labeling of a pooled urine or pooled feces and subsequent analysis of the (13)C-/(12)C-labeled mixture by LC-MS. The pooled urine and pooled feces are further differentially labeled, mixed and then analyzed by LC-MS in order to relate the metabolite concentrations of the common metabolites found in both biospecimens. This method offers a means of direct comparison of urinary and fecal submetabolomes. We evaluated the analytical performance and demonstrated the utility of this method in the analysis of urine and feces collected daily from three healthy individuals for 7 days. On average, 2534 ± 113 (n = 126) peak pairs or metabolites could be detected from a urine sample, while 2507 ± 77 (n = 63) peak pairs were detected from a fecal sample. In total, 5372 unique peak pairs were detected from all the samples combined; 3089 and 3012 pairs were found in urine and feces, respectively. These results reveal that the urine and fecal metabolomes are very different, thereby justifying the consideration of using both biospecimens to increase the probability of finding specific biomarkers of diseases. Furthermore, the CIL LC-MS method described can be used to perform parallel quantitative analysis of urine and feces, resulting in more complete coverage of the human metabolome.

Early detection is vital to improve the overall survival rate of bladder cancer (BCa) patients, yet there is a lack of a reliable urine-based assay for early detection of BCa. Urine metabolites represented a potential rich source of biomarkers for BCa. This study aimed to develop a metabolomics approach for high coverage discovery and identification of metabolites in urine samples. Urine samples from 23 early stage BCa patients and 21 healthy volunteers with minimum sample preparations were analyzed by a short 30 min UPLC-HRMS method. We detected and quantified over 9000 unique UPLC-HRMS features, which is more than four times than about 2000 features detected in previous urine metabolomic studies. Furthermore, multivariate OPLS-DA classification models were established to differentiate urine samples from bladder cancer cohort and normal health cohort. We identified three BCa-upregulated metabolites: nicotinuric acid, trehalose, AspAspGlyTrp, and three BCa-downregulated metabolites: inosinic acid, ureidosuccinic acid, GlyCysAlaLys. Finally, analysis of six post-surgery BCa urine samples showed that these BCa-metabolomic features reverted to normal state after tumor removal, suggesting that they reflected metabolomic features associated with BCa. ROC analyses using two linear regression models to combine the identified markers showed a high diagnostic performance for detecting BCa with AUC (area under the ROC curve) values of 0.919 to 0.934. In summary, we developed a high coverage metabolomic approach that has potential for biomarker discovery in cancers.

Human fecal samples contain endogenous human metabolites, gut microbiota metabolites, and other compounds. Profiling the fecal metabolome can produce metabolic information that may be used not only for disease biomarker discovery, but also for providing an insight about the relationship of the gut microbiome and human health. In this work, we report a chemical isotope labeling liquid chromatography-mass spectrometry (LC-MS) method for comprehensive and quantitative analysis of the amine- and phenol-containing metabolites in fecal samples. Differential (13)C2/(12)C2-dansyl labeling of the amines and phenols was used to improve LC separation efficiency and MS detection sensitivity. Water, methanol, and acetonitrile were examined as an extraction solvent, and a sequential water-acetonitrile extraction method was found to be optimal. A step-gradient LC-UV setup and a fast LC-MS method were evaluated for measuring the total concentration of dansyl labeled metabolites that could be used for normalizing the sample amounts of individual samples for quantitative metabolomics. Knowing the total concentration was also useful for optimizing the sample injection amount into LC-MS to maximize the number of metabolites detectable while avoiding sample overloading. For the first time, dansylation isotope labeling LC-MS was performed in a simple time-of-flight mass spectrometer, instead of high-end equipment, demonstrating the feasibility of using a low-cost instrument for chemical isotope labeling metabolomics. The developed method was applied for profiling the amine/phenol submetabolome of fecal samples collected from three families. An average of 1785 peak pairs or putative metabolites were found from a 30 min LC-MS run. From 243 LC-MS runs of all the fecal samples, a total of 6200 peak pairs were detected. Among them, 67 could be positively identified based on the mass and retention time match to a dansyl standard library, while 581 and 3197 peak pairs could be putatively identified based on mass match using MyCompoundID against a Human Metabolome Database and an Evidence-based Metabolome Library, respectively. This represents the most comprehensive profile of the amine/phenol submetabolome ever detected in human fecal samples. The quantitative metabolome profiles of individual samples were shown to be useful to separate different groups of samples, illustrating the possibility of using this method for fecal metabolomics studies.