Metabolomics and metabolic pathway networks from human colorectal cancers, adjacent mucosa, and stool
© The Author(s). 2016
Received: 26 January 2016
Accepted: 16 May 2016
Published: 6 June 2016
Colorectal cancers (CRC) are associated with perturbations in cellular amino acids, nucleotides, pentose-phosphate pathway carbohydrates, and glycolytic, gluconeogenic, and tricarboxylic acid intermediates. A non-targeted global metabolome approach was utilized for exploring human CRC, adjacent mucosa, and stool. In this pilot study, we identified metabolite profile differences between CRC and adjacent mucosa from patients undergoing colonic resection. Metabolic pathway analyses further revealed relationships between complex networks of metabolites.
Seventeen CRC patients participated in this pilot study and provided CRC, adjacent mucosa ~10 cm proximal to the tumor, and stool. Metabolomes were analyzed by gas chromatography-mass spectrometry (GC/MS) and ultra-performance liquid chromatography-mass spectrometry (UPLC-MS/MS). All of the library standard identifications were confirmed and further analyzed via MetaboLyncTM for metabolic network interactions.
There were a total of 728 distinct metabolites identified from colonic tissue and stool matrices. Nineteen metabolites significantly distinguished CRC from adjacent mucosa in our patient-matched cohort. Glucose-6-phosphate and fructose-6-phosphate demonstrated 0.64-fold and 0.75-fold lower expression in CRC compared to mucosa, respectively, whereas isobar: betaine aldehyde, N-methyldiethanolamine, and adenylosuccinate had 2.68-fold and 1.88-fold higher relative abundance in CRC. Eleven of the 19 metabolites had not previously been reported for CRC relevance. Metabolic pathway analysis revealed significant perturbations of short-chain fatty acid metabolism, fructose, mannose, and galactose metabolism, and glycolytic, gluconeogenic, and pyruvate metabolism. In comparison to the 500 stool metabolites identified from human CRC patients, only 215 of those stool metabolites were also detected in tissue. This CRC and stool metabolome investigation identified novel metabolites that may serve as key small molecules in CRC pathogenesis, confirmed the results from previously reported CRC metabolome studies, and showed networks for metabolic pathway aberrations. In addition, we found differences between the CRC and stool metabolomes.
Stool metabolite profiles were limited for direct associations with CRC and adjacent mucosa, yet metabolic pathways were conserved across both matrices. Larger patient-matched CRC, adjacent non-cancerous colonic mucosa, and stool cohort studies for metabolite profiling are needed to validate these small molecule differences and metabolic pathway aberrations for clinical application to CRC control, treatment, and prevention.
Metabolomics has helped cancer research to elucidate specific biomarkers of disease that facilitate personalized therapies and may improve clinical outcomes non-invasively following primary or secondary diagnoses . Colorectal cancer (CRC) is the third leading cause of cancer-related death in the USA  and while there are many screening techniques (i.e., colonoscopy, double contrast barium enema, immunochemical-based fecal occult blood test, and serum CEA test [3–5]), high throughput and sensitive molecular tools are needed to support identification of novel biomarkers for metabolic pathway aberrations. Metabolomics is an emerging analytical tool that can be described as the systematic study of the entire profile of small molecules in a clinical sample that are detected using mass spectrometry. Metabolites from a multitude of matrices, such as serum, urine, stool, and tissue, represent the downstream functional products of gene expression and protein synthesis and include environmental exposures and microbial metabolism, all of which influence CRC processes . There remains a gap in knowledge for the spectrum of metabolic pathway alterations that co-exist in the CRC tumor microenvironment which merits a non-targeted, global metabolomics approach .
Chan et al. reported 31 metabolites in 2009 that were differentially expressed in paired CRC and adjacent mucosal samples, with glucose levels that were ~67 % higher in the mucosa . In 2014, Qiu et al. identified 15 significantly altered metabolites in matched surgical specimens by comparing CRC and adjacent mucosa from 3 different hospitals in China and 1 in the USA . These metabolite signatures were then used to predict the rate of recurrence and survival in patients after treatment . Most CRCs originate from polyps that exhibit metabolic alterations due to distinct mutations that allow them to replicate uncontrollably, evade host immunity, and become invasive (i.e., Warburg effect, APC/KRAS/BRAF gene mutations, MSI status) [10–12]. Metabolic fingerprints that could distinguish adjacent mucosa from CRC may reveal stages of pathogenesis, inform the frequency of follow-up screening, and response to preventive measures needed for improved prognosis [13, 14].
Paralleled investigations of CRC and adjacent mucosal tissues alongside patient-matched stools that reflect the tumor microenvironment are limited . In the present pilot study, we determined metabolite profile differences between patient-matched CRC and adjacent mucosa and performed metabolic pathway analyses to examine correlations between metabolite expression. We hypothesized that identification of metabolite signatures of colonic tissue that are proximal to and within the CRC represent potential metabolic targets for secondary CRC control, prevention, and treatment. The overlapping metabolites from patient-matched stool and tissue metabolomes as aberrancies in metabolic pathway networks presented herein merit further interrogation for enhanced knowledge of CRC etiology and pathogenesis.
Ethics, consent, and permissions
Seventeen individuals with a scheduled colonic resection after a CRC diagnosis were recruited for this study. All patients provided informed, written consent and knew that data would be used for publication. The patient inclusion criteria for this study included not taking antibiotics prior to surgery (intravenous pre-operative antibiotics were admissible). Tissue and stool samples were collected from Poudre Valley Hospital (PVH; Fort Collins, CO, USA) and de-identified for personal information before processing at Colorado State University (protocol nos. 10-1006 and 10-1670H).
Colorectal cancer patient and tumor characteristics
Age (years) [mean ± SD]
58.8 ± 13.8
13 (76 %)
4 (24 %)
BMI (kg/m2) [mean ± SD]
29.3 ± 4.7
1 (6 %)
1 (6 %)
3 (18 %)
3 (18 %)
8 (46 %)
1 (6 %)
1 (6 %)
5 (31 %)
1 (6 %)
6 (38 %)
3 (19 %)
Tumor size (cm)
0 < 2
6 (35 %)
2 < 4
5 (29 %)
4 < 6
3 (18 %)
6 < 8
3 (18 %)
15 (94 %)
1 (6 %)
Sample accessioning and preparation
Tissue and stool metabolite extractions for GC-MS and UPLC-MS/MS were completed by Metabolon Inc. Both platforms were chosen to provide broad, non-targeted detection of metabolites. Each sample was accessioned into the Metabolon Laboratory Information Management System (LIMS) and assigned a unique identifier associated with the original de-identified study code number. This identifier was used to track all sample handling, tasks, and results. All samples were maintained at −80 °C until analysis .
Samples were extracted using the automated MicroLab STAR® robotics system from Hamilton Company (Reno, NV, USA). A set of recovery standards, tridecanoic acid, chloro-phenylalanine, D6-cholesterol, and fluoro-phenylglycine was added prior to the first step in the extraction process for quality control purposes, with final extraction standard concentrations ranging from 2.5 to 25 μg/mL. Sample preparation was conducted using a methanol extraction to remove the protein fraction while allowing the maximum recovery of small molecules. Tissue samples were homogenized in water (5 μL/mg of sample) using a GenoGrinder 2000 bead grinder (Glen Mills, Clifton, NJ, USA), shaking for 5 min at 1000 strokes per minute at room temperature. Methanol extraction utilized five volumes of methanol (5:1 methanol/water) with vigorous shaking at room temperature for 2 min followed by centrifugation at 680 ×g for 3 min. The resulting methanol extract was divided into four fractions: one for analysis by UPLC-MS/MS with positive ion mode electrospray ionization and one for negative ion mode electrospray ionization, one for analysis by GC-MS, and one sample was reserved for backup. Samples were placed briefly on a TurboVap® from Zymark Corporation (Hopkinton, MA, USA) to remove the organic solvent. The samples were stored overnight under nitrogen before preparation for UPLC-MS/MS and each sample was dried under vacuum for a minimum of 18 h before derivatization for GC-MS analysis. Samples were analyzed in concert with several types of controls to allow instrument performance monitoring and aid chromatographic alignment.
Samples were derivatized under nitrogen using bistrimethyl-silyltrifluoroacetamide and separated on a 5 % diphenyl/95 % dimethyl polysiloxane-fused silica column (20 m × 0.18 mm ID; 0.18-μm film thickness) with helium as carrier gas and a temperature ramp from 60 to 340 °C in a 17.5-min period. Internal standards amylbenzene, 1-phenylhexane, 1-phenyloctane, 1-phenyldecane, 1-phenyldodecane, hexadecylbenzene, octadecylbenzene, tetradecylbenzene, and 2,6-di-tert-butyl-4-methylphenol were added to each sample (250 ng of each standard per sample). Samples were analyzed on a Thermo-Finnigan Trace DSQ fast-scanning single-quadrupole mass spectrometer using electro-impact ionization (EI) and operated at unit mass resolving power. The scan range was from 50–750 m/z.
The UPLC-MS/MS portion of the platform was based on a Waters ACQUITY UPLC and a Thermo-Finnigan LTQ MS operated at nominal mass resolution, which consisted of an electrospray ionization (ESI) source and linear ion-trap (LIT) mass analyzer. The dried sample extract was reconstituted in acidic or basic UPLC-compatible solvents, each of which contained 11 to 13 injection standards at fixed concentrations . One aliquot was analyzed using acidic positive ion-optimized conditions and the other using basic, negative ion-optimized conditions in two independent injections using separate dedicated columns (Waters UPLC BEH C18-2.1 × 100 mm, 1.7 μm). Extracts reconstituted in acidic conditions were gradient eluted using water and methanol containing 0.1 % formic acid, while the basic extracts, which also used water/methanol, contained 6.5 mM ammonium bicarbonate. The MS analysis alternated between MS and data-dependent MS/MS scans using dynamic exclusion and the scan range was from 80–1000 m/z.
Data extraction and compound identification
Raw data was extracted and peak-identified as previously described . Biochemical identifications were based on (1) retention index within a narrow RI window of the proposed identification, (2) accurate mass match to the library +/− 0.005 amu, and (3) the MS/MS forward and reverse scores between the experimental data and authentic standards. The MS/MS scores were based on a comparison of the ions present in the experimental spectrum to the ions present in the library spectrum. More than 3300 commercially available purified standard compounds have been acquired and registered into LIMS for distribution to both the UPLC-MS/MS and GC-MS platforms for determination of their analytical characteristics . All detected metabolites were identified with level 1 or 2 (noted by an asterisk symbol) confidence of identification .
Metabolomics statistical analysis
Metabolite profiles in CRC patients were quantified in terms of relative abundance and median scaled to 1. Following log transformation and imputation of missing values, if any, with the minimum observed value for each compound imputed, statistical analyses were performed to identify significant differences between experimental groups. Statistical analyses were performed in ArrayStudio (Omicsoft, Cary, NC, USA), R version 2.14.2, and/or SAS v9.4 . Metabolite profile distinctions between CRC and adjacent mucosa were evaluated by matched pair t tests. An estimate of the false discovery rate (q value) was calculated to take into account the multiple comparisons that normally occur in metabolomic-based studies. A q value threshold of ≤0.10 was used to correct for false discovery of statistically significant compounds due to multiple hypothesis testing. Metabolites with q values that exceeded this threshold were discarded from further analysis. Fold difference (FD) was determined by dividing the relative abundance of the metabolite in the CRC by the relative abundance of the metabolite in the adjacent mucosa. Metabolites with p values of ≤0.05 with q values below the threshold of ≤0.10 were considered statistically significant in this study.
Metabolic pathway networks and analysis
To visualize and analyze small molecules within relevant networks of metabolic pathways, the detected metabolites in CRC and adjacent mucosa were subjected to MetaboLync pathway analysis (MPA) software (portal.metabolon.com). Significantly altered pathways were determined by completing pathway set enrichment analysis within MPA software which was determined by the following equation:
# of significant metabolites in pathway (k)/total # of detected metabolites in pathway (m)/total # of significant metabolites (n)/total # of detected metabolites (N) or (k/m)/(n/N).
A pathway impact score greater than one indicates that the pathway contains a higher number of experimentally regulated compounds relative to the overall study in CRC than adjacent mucosa, suggesting that the pathway may be of interest to the metabolite perturbations observed. Finally, the significantly altered pathways and metabolites were visualized via a Pathway Visualizations tool using Cytoscape v 2.8.3 software .
Metabolite differences between CRC and adjacent mucosa
Metabolites with statistically significant differences between colorectal cancer tissue and adjacent mucosa
Fold difference (CRC/mucosa)
Glycine, serine, and threonine metabolism
Isobar: betaine aldehyde, N-methyldiethanolamine
Purine metabolism, adenine containing
Leucine, isoleucine, and valine metabolism
Short-chain fatty acid
Nicotinate and nicotinamide metabolism
Phenylalanine and tyrosine metabolism
Fructose, mannose, and galactose metabolism
Leucine, isoleucine, and valine metabolism
Secondary bile acid metabolism
Alanine and aspartate metabolism
Glycolysis, gluconeogenesis, and pyruvate metabolism
Metabolic network differences between CRC and adjacent mucosa
Stool metabolome of CRC patients
CRC stool metabolite numbers within each of the assigned metabolic pathways
Number of metabolites
Alanine and aspartate
Glycine, serine, and threonine
Leucine, isoleucine, and valine
Methionine, cysteine, SAM, and taurine
Phenylalanine and tyrosine
Urea cycle; arginine and proline
Disaccharides and oligosaccharides
Fructose, mannose, and galactose
Glycolysis, gluconeogenesis, and pyruvate
Cofactors and vitamins
Ascorbate and aldarate
Hemoglobin and porphyrin
Nicotinate and nicotinamide
Pantothenate and CoA
Fatty alcohol, long chain
Primary bile acid
Secondary bile acid
Purine and pyrimidine
(Hypo) xanthine/inosine containing
Purine metabolism, adenine containing
Purine metabolism, guanine containing
Pyrimidine metabolism, cytidine containing
Pyrimidine metabolism, orotate containing
Pyrimidine metabolism, thymine containing
Pyrimidine metabolism, uracil containing
Gamma-glutamyl amino acid
Metabolome overlap and comparison between CRC and stool
This multi-platform CRC and adjacent mucosa metabolome analysis led to the identification of 11 novel metabolites that have not previously been detected for CRC relevance in animals or humans. Adenylosuccinate, isovalerate, valerate, N1-methyl-2-pyridone-5-carboxamide, stearoyl sphingomyelin, 4-hydroxyphenylpyruvate, aspartylvaline, aspartyltryptophan, tryptophyglycine, glucose-6-phosphate, and fructose-6-phosphate merit further investigation as key small molecules in CRC pathogenesis. It is noteworthy that tryptophylglycine was detected in a fecal metabolomics study comparing CRC patients and matched healthy controls , whereas our report is the first to our knowledge to show tryptophylglycine metabolite levels in human tissue. One limitation of this study is the small sample size and subsequent low representation of each tumor stage and location. A strength of this pilot investigation was the capacity to obtain CRC and adjacent mucosa with stool samples from the same patient and utilize a multi-platform analytical approach that is comparable to other human datasets (GC/LC-MS) [8, 9, 23].
Our findings supported the identification of 8 metabolites that were discriminatory between CRC and adjacent mucosa in previous reports and included 2-aminoadipate, isobar: betaine aldehyde, N-methyldiethanolamine, alpha-hydroxyisovalerate, cys-gly, oxidized, deoxycholate, 7-ketodeoxycholate, sorbitol, and asparagine [9, 14, 23]. However, these results were not consistent by the relative abundance, such that the asparagine, 2-aminoadipate, alpha-hydroxyisovalerate, deoxycholate, 7-ketodeoxycholate, sorbitol and cys-gly, oxidized had higher expression in adjacent mucosa compared to CRC, whereas isobar: betaine aldehyde, N-methyldiethanolamine had higher expression in CRC than in adjacent mucosa in this study . Intraindividual differences in tumor heterogeneity and analytical platforms may, in part, explain these discrepancies .
Two secondary bile acid metabolites (7-ketodeoxycholate and deoxycholate) showed higher expression in adjacent mucosa compared with CRC and were detectable across sample matrices (Additional file 1: Table S1). These metabolites are associated with generation of reactive oxygen and nitrogen species [24–27], altering the stability of the cell and mitochondrial membranes [25, 26], and inducing oxidative DNA damage [25, 26], mutation [25, 26], and apoptosis in CRC [25, 26]. Deoxycholic acid was shown to be diet-modifiable in the plasma metabolome , and had higher expression in adjacent mucosa in this study. Alternatively, the variable levels we observed between tissue matrices could be due to gut microbial metabolism. The gut mircobiota are known to convert primary bile acids to secondary bile acids, a process that may be pH-dependent . Variable gut microbial populations throughout the colon and from person to person may also contribute to the differential expression of secondary bile acids observed in this study (Fig. 3a).
Two additional key small molecules of CRC pathogenesis were elucidated in this global, metabolite profiling: N1-methyl-2-pyridone-5-carboxamide and sorbitol. These small molecules are important because they were discriminatory between CRC and adjacent mucosa in this study and they have been reported in disease contexts including chronic renal failure and diabetes [30, 31]. N1-methyl 2-pyridone-5-carboxamide (N125) has been increased in high fat diet-induced obesity, which is an elevated CRC risk factor . N125 was also shown as diet-modifiable in urinary metabolomes , supporting its possible importance as a diet-modifiable mediator of CRC pathogenesis. Additionally, excess sorbitol has been associated with diabetic-related microvascular complications and retinopathy, and its accumulation was associated with osmotic and oxidative stress damage to the endothelium .
Beyond specific metabolite differences, we also evaluated entire metabolic networks to interrogate their contributions to both CRC and mucosal tissue microenvironments. The perturbation of short-chain fatty acid metabolism between CRC and adjacent mucosa is noteworthy because of the breadth of short-chain fatty acid metabolites emphasized in CRC prevention and control research. Short-chain fatty acids such as butyrate, propionate, and valerate have been shown to cause growth arrest and differentiation in human colorectal cancer cells . In our study, short-chain fatty acid metabolism had a pathway impact score of 23.32 and contained the discriminatory metabolite valerate. These results taken together merit continued metabolomics investigation of short-chain fatty acids in the context of CRC pathogenesis and prevention, specifically valerate [34–36]. In addition to short-chain fatty acid metabolism, our pathway analysis revealed linkages to the glycolytic/gluconeogenic pathway and CRC tumorigenesis (pathway impact score of 5.8). This metabolic pathway contained 2 discriminatory metabolites: fructose-6-phosphate and glucose-6-phosphate. Glucose-6-phosphate and fructose-6-phosphate represent metabolites/intermediates of cellular respiration that are established for enhancing cancer cell energy required for rapid proliferation . In this study, we observed depleted glucose-6-phosphate and fructose-6-phosphate levels in tumor tissues relative to adjacent mucosa which implies a Warburg effect or a switch from oxidative to substrate-level phosphorylation for energy production . Larger cohort investigations may be needed to establish the signatures associated with short-chain fatty acid metabolism, fructose, mannose, and galactose metabolism, glycolysis, gluconeogenesis, and pyruvate metabolism, alanine and aspartate metabolism, leucine, isoleucine, and valine metabolism, and secondary bile acid metabolism (Fig. 2). Given that the CRC metabotype has not yet been expanded to provide metabolite information specific to the colonic location affected (i.e., proximal, distal, rectum), future investigations should consider this to be another experimental parameter by which to discriminate samples from each other and for enhanced precision in biomarker discovery .
Of the 703 metabolites detected in this study, 29.2 % of metabolites overlap across tissue and stool matrices (Fig. 4). Colorectal tissue and stool metabolite profiles derive from several inputs including altered gene expression, oxidative stress responses, xenobiotic metabolism, and the utilization of alternative carbon sources from the gut microbiota. Findings from this pilot investigation of CRC and stool raises awareness regarding the utility of stool metabolites as relevant indicators of CRC tissue microenvironment [39, 40].
This pilot, global, non-targeted metabolome study identified various metabolic alterations in both individual metabolites and metabolic pathway networks between CRC and adjacent mucosa from CRC patients. In addition, we observed a low percentage of conserved detection across tissue and stool matrices (213 shared metabolites of 728 total metabolites; 29.2 % conserved metabolites). Findings from this study reiterate the complexity of CRC biology with regard to individual phenotypes and the utility for a broad-spectrum metabolite detection platform to guide our approaches to treatment, control, and prevention of this complex malignancy.
ANOVA, analysis of variance; APC, adenomatous polyposis coli; BRAF, B-Raf proto-oncogene, serine/threonine kinase; CEA, carcinoembryonic antigen; COX, cyclooxygenase; CRC, colorectal cancer; FD, fold difference; GC/MS, gas chromatography mass spectrometry; HMDB, human metabolome database; KEGG, Kyoto Encyclopedia Genes and Genomes; KRAS, Kirsten rat sarcoma oncogene homolog; LC-MS/MS, liquid chromatography mass spectrometry; m/z, mass-to-charge ratio; MPA, MetaboLync pathway analysis; MSI, microsatellite instability; NSAID, non-steroidal anti-inflammatory drugs; PVH, Poudre Valley Hospital; RI/RT, retention index/retention time.
The authors would like to thank the Poudre Valley Hospital/Oncology Clinical Research team, Erica C. Borresen, and Brittany Barnett for their technical and logistical support of this work. We would also like to thank Dr. Stephen Pettine, and the surgeons of the Northern Colorado Surgical Associates, who participated in recruitment and sample collection. The authors would like to acknowledge pilot project funding from the University of Colorado Cancer Center, Division of Cancer Control and Prevention, an NIH RO3CA150070-2 award, and internal department funds (Ryan) from Environmental and Radiological Health Sciences at Colorado State University for supporting this project.
Availability of data and materials
The data sets supporting the results of this article are included within the article and its additional files.
DB participated in result interpretations, metabolome data analysis, and wrote the manuscript. SR performed the statistical analyses for the study. TW participated in study design, collection of samples for the study, and provided written comments for manuscript. JO and MB participated in patient recruitment, consent, and coordination of sample collection. RB helped coordinate the patient recruitment and helped with the study design. ER conceived the study design and coordinated the regulatory approvals, result interpretations, and writing of the manuscript. All authors read and approved the final manuscript.
The authors declare that they have no competing interests.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Lech G, Slotwinski R, Krasnodebski IW. The role of tumor markers and biomarkers in colorectal cancer. Neoplasma. 2013. doi:10.4149/neo_2014_003.Google Scholar
- Siegel R, Desantis C, Jemal A. Colorectal cancer statistics. CA Cancer J Clin. 2014;64(2):104–17. doi:10.3322/caac.21220.View ArticlePubMedGoogle Scholar
- Phua LC, Chue XP, Koh PK, Cheah PY, Ho HK, Chan EC. Non-invasive fecal metabonomic detection of colorectal cancer. Cancer Biol Ther. 2014;15(4):389–97. doi:10.4161/cbt.27625.View ArticlePubMedPubMed CentralGoogle Scholar
- Ned RM, Melillo S, Marrone M. Fecal DNA testing for colorectal cancer screening: the ColoSure test. PLoS currents. 2011;3, RRN1220. doi:10.1371/currents.RRN1220.PubMedPubMed CentralGoogle Scholar
- Zou M, Zhang PJ, Wen XY, Chen L, Tian YP, Wang Y. A novel mixed integer programming for multi-biomarker panel identification by distinguishing malignant from benign colorectal tumors. Methods. 2015. doi:10.1016/j.ymeth.2015.05.011.Google Scholar
- Schmidt CW. Metabolomics: what’s happening downstream of DNA. Environ Health Perspect. 2004;112(7):A410–5.View ArticlePubMedPubMed CentralGoogle Scholar
- Shirakami Y, Shimizu M, Kubota M, Araki H, Tanaka T, Moriwaki H, et al. Chemoprevention of colorectal cancer by targeting obesity-related metabolic abnormalities. World J Gastroenterol. 2014;20(27):8939–46. doi:10.3748/wjg.v20.i27.8939.PubMedPubMed CentralGoogle Scholar
- Chan EC, Koh PK, Mal M, Cheah PY, Eu KW, Backshall A, et al. Metabolic profiling of human colorectal cancer using high-resolution magic angle spinning nuclear magnetic resonance (HR-MAS NMR) spectroscopy and gas chromatography mass spectrometry (GC/MS). J Proteome Res. 2009;8(1):352–61. doi:10.1021/pr8006232.View ArticlePubMedGoogle Scholar
- Qiu Y, Cai G, Zhou B, Li D, Zhao A, Xie G, et al. A distinct metabolic signature of human colorectal cancer with prognostic potential. Clini Cancer Res. 2014;20(8):2136–46. doi:10.1158/1078-0432.CCR-13-1939.View ArticleGoogle Scholar
- Yamane LS, Scapulatempo-Neto C, Alvarenga L, Oliveira CZ, Berardinelli GN, Almodova E et al. KRAS and BRAF mutations and MSI status in precursor lesions of colorectal cancer detected by colonoscopy. Oncology reports. 2014. doi:10.3892/or.2014.3338.Google Scholar
- Williams MD, Reeves R, Resar LS, Hill Jr HH. Metabolomics of colorectal cancer: past and current analytical platforms. Anal Bioanal Chem. 2013;405(15):5013–30. doi:10.1007/s00216-013-6777-5.View ArticlePubMedGoogle Scholar
- Donohoe DR, Collins LB, Wali A, Bigler R, Sun W, Bultman SJ. The Warburg effect dictates the mechanism of butyrate-mediated histone acetylation and cell proliferation. Mol Cell. 2012;48(4):612–26. doi:10.1016/j.molcel.2012.08.033.View ArticlePubMedPubMed CentralGoogle Scholar
- Assfalg M, Bertini I, Colangiuli D, Luchinat C, Schafer H, Schutz B, et al. Evidence of different metabolic phenotypes in humans. Proc Natl Acad Sci U S A. 2008;105(5):1420–4. doi:10.1073/pnas.0705685105.View ArticlePubMedPubMed CentralGoogle Scholar
- Montrose DC, Zhou XK, Kopelovich L, Yantiss RK, Karoly ED, Subbaramaiah K, et al. Metabolic profiling, a noninvasive approach for the detection of experimental colorectal neoplasia. Cancer Prev Res. 2012;5(12):1358–67. doi:10.1158/1940-6207.CAPR-12-0160.View ArticleGoogle Scholar
- Park JS, Choi GS, Hasegawa S, Sakai Y, Huh JW, Kim HR, et al. Validation of the seventh edition of the American Joint Committee on cancer tumor node-staging system in patients with colorectal carcinoma in comparison with sixth classification. J Surg Oncol. 2012;106(6):674–9. doi:10.1002/jso.23117.View ArticlePubMedGoogle Scholar
- DeHaven CDE, Anne M, Dai H, Lawton KA. Software techniques for enabling high-throughput analysis of metabolomic datasets. InTech. 2012.Google Scholar
- Evans AM, DeHaven CD, Barrett T, Mitchell M, Milgram E. Integrated, nontargeted ultrahigh performance liquid chromatography/electrospray ionization tandem mass spectrometry platform for the identification and relative quantification of the small-molecule complement of biological systems. Anal Chem. 2009;81(16):6656–67. doi:10.1021/ac901536h.View ArticlePubMedGoogle Scholar
- Sumner LW, Amberg A, Barrett D, Beale MH, Beger R, Daykin CA, et al. Proposed minimum reporting standards for chemical analysis Chemical Analysis Working Group (CAWG) Metabolomics Standards Initiative (MSI). Metabolomics: Official journal of the Metabolomic Society. 2007;3(3):211–21. doi:10.1007/s11306-007-0082-2.View ArticleGoogle Scholar
- Team RC. R: A language and environment for statistical computing. Vienna Austria: R Foundation for Statistical Computing; 2013.Google Scholar
- Smoot ME, Ono K, Ruscheinski J, Wang PL, Ideker T. Cytoscape 2.8: new features for data integration and network visualization. Bioinformatics. 2011;27(3):431–2. doi:10.1093/bioinformatics/btq675.View ArticlePubMedPubMed CentralGoogle Scholar
- Guo L, Milburn MV, Ryals JA, Lonergan SC, Mitchell MW, Wulff JE, et al. Plasma metabolomic profiles enhance precision medicine for volunteers of normal health. Proc Natl Acad Sci U S A. 2015;112(35):E4901–10. doi:10.1073/pnas.1508425112.View ArticlePubMedPubMed CentralGoogle Scholar
- Goedert JJ, Sampson JN, Moore SC, Xiao Q, Xiong X, Hayes RB, et al. Fecal metabolomics: assay performance and association with colorectal cancer. Carcinogenesis. 2014;35(9):2089–96. doi:10.1093/carcin/bgu131.View ArticlePubMedPubMed CentralGoogle Scholar
- Williams MD, Zhang X, Park JJ, Siems WF, Gang DR, Resar LM et al. Characterizing metabolic changes in human colorectal cancer. Analytical and bioanalytical chemistry. 2015. doi:10.1007/s00216-015-8662-x.Google Scholar
- Qin P, Tang X, Elloso MM, Harnish DC. Bile acids induce adhesion molecule expression in endothelial cells through activation of reactive oxygen species, NF-kappaB, and p38. Am J Physiol Heart Circ Physiol. 2006;291(2):H741–7. doi:10.1152/ajpheart.01182.2005.View ArticlePubMedGoogle Scholar
- Bernstein H, Bernstein C, Payne CM, Dvorak K. Bile acids as endogenous etiologic agents in gastrointestinal cancer. World J Gastroenterol. 2009;15(27):3329–40.View ArticlePubMedPubMed CentralGoogle Scholar
- Ajouz H, Mukherji D, Shamseddine A. Secondary bile acids: an underrecognized cause of colon cancer. World J Surg Oncol. 2014;12:164. doi:10.1186/1477-7819-12-164.View ArticlePubMedPubMed CentralGoogle Scholar
- Song S, Guha S, Liu K, Buttar NS, Bresalier RS. COX-2 induction by unconjugated bile acids involves reactive oxygen species-mediated signalling pathways in Barrett’s oesophagus and oesophageal adenocarcinoma. Gut. 2007;56(11):1512–21. doi:10.1136/gut.2007.121244.View ArticlePubMedPubMed CentralGoogle Scholar
- Ryan EP, Heuberger AL, Broeckling CD, Borresen EC, Tillotson C, Prenni JE. Advances in nutritional metabolomics. Current Metabolomics. 2013;I:109–20.View ArticleGoogle Scholar
- Christl SU, Bartram HP, Ruckert A, Scheppach W, Kasper H. Influence of starch fermentation on bile acid metabolism by colonic bacteria. Nutr Cancer. 1995;24(1):67–75. doi:10.1080/01635589509514394.View ArticlePubMedGoogle Scholar
- Kinoshita JH, Nishimura C. The involvement of aldose reductase in diabetic complications. Diabetes Metab Rev. 1988;4(4):323–37.View ArticlePubMedGoogle Scholar
- Rutkowski B, Slominska E, Szolkiewicz M, Smolenski RT, Striley C, Rutkowski P, et al. N-methyl-2-pyridone-5-carboxamide: a novel uremic toxin? Kidney Int Suppl. 2003;84:S19–21. doi:10.1046/j.1523-1755.63.s84.36.x.View ArticlePubMedGoogle Scholar
- Boulange CL, Claus SP, Chou CJ, Collino S, Montoliu I, Kochhar S, et al. Early metabolic adaptation in C57BL/6 mice resistant to high fat diet induced weight gain involves an activation of mitochondrial oxidative pathways. J Proteome Res. 2013;12(4):1956–68. doi:10.1021/pr400051s.View ArticlePubMedGoogle Scholar
- Hinnebusch BF, Meng SF, Wu JT, Archer SY, Hodin RA. The effects of short-chain fatty acids on human colon cancer cell phenotype are associated with histone hyperacetylation. J Nutr. 2002;132(5):1012–7.PubMedGoogle Scholar
- Louis P, Hold GL, Flint HJ. The gut microbiota, bacterial metabolites and colorectal cancer. Nat Rev Microbiol. 2014;12(10):661–72. doi:10.1038/nrmicro3344.View ArticlePubMedGoogle Scholar
- Weir TL, Manter DK, Sheflin AM, Barnett BA, Heuberger AL, Ryan EP. Stool microbiome and metabolome differences between colorectal cancer patients and healthy adults. PLoS One. 2013;8(8):e70803. doi:10.1371/journal.pone.0070803.View ArticlePubMedPubMed CentralGoogle Scholar
- Hester CM, Jala VR, Langille MG, Umar S, Greiner KA, Haribabu B. Fecal microbes, short chain fatty acids, and colorectal cancer across racial/ethnic groups. World J Gastroenterol. 2015;21(9):2759–69. doi:10.3748/wjg.v21.i9.2759.View ArticlePubMedPubMed CentralGoogle Scholar
- Hamanaka RB, Chandel NS. Targeting glucose metabolism for cancer therapy. J Exp Med. 2012;209(2):211–5. doi:10.1084/jem.20120162.View ArticlePubMedPubMed CentralGoogle Scholar
- Nam SO, Yotsumoto F, Miyata K, Fukagawa S, Yamada H, Kuroki M, et al. Warburg effect regulated by amphiregulin in the development of colorectal cancer. Cancer Med. 2015;4(4):575–87. doi:10.1002/cam4.416.View ArticlePubMedPubMed CentralGoogle Scholar
- Wang H, Tso V, Wong C, Sadowski D, Fedorak RN. Development and validation of a highly sensitive urine-based test to identify patients with colonic adenomatous polyps. Clini Transl Gastroenterol. 2014;5:e54. doi:10.1038/ctg.2014.2.View ArticleGoogle Scholar
- Zamani Z, Arjmand M, Vahabi F, Eshaq Hosseini SM, Fazeli SM, Iravani A, et al. A metabolic study on colon cancer using (1) h nuclear magnetic resonance spectroscopy. Biochem Res Int. 2014;2014:348712. doi:10.1155/2014/348712.View ArticlePubMedPubMed CentralGoogle Scholar