Discovery Of Biomarkers For Diagnosis Of Colorectal Cancer By LC-MS/MS Based Targeted Amino Acid Metabolomics

Colorectal cancer?CRC?,also known as large intestine carcinoma,is a common malignant tumor of the digestive system.In the past decade,the incidence and the age-standardized incidence of CRC has been continuously increasing in China.CRC has characteristics of a long incubation period,slow growth and progression.The diagnosis and clinical staging has a great influence on the prognosis of patients.Early diagnosis can effectively improve the therapeutic efficacy and the prognosis of patients.Nowadays,methods for CRC diagnosis mainly include biochemical test,imaging examination and pathological examination.However,the early diagnosis rate is still low due to lacking of sensitive and specific biomarkers.Colonoscopy,the diagnostic gold standard for CRC,is not suitable for early screening in large-scale populations.It is of great clinical significance to explore new biomarkers for the effective diagnosis of CRC.These years,many novel biomarkers have been discovered and identified by the different“omics”technologies.As one branch of omics,metabolomics with high sensitivity and specificity could detect small molecular substances involving in cell signal transduction,regulation and biochemical reactions.Because metabolome could monitor any tiny biochemical changes in the human body,it is more sensitive than currently clinical endogenous biomarkers.Metabolome has become a powerful potential tool for the diagnosis and personalized treatment of various cancers,including CRC.It is considered to be non-invasive or minimally invasive,ideal tool for large-scale population screening,by using of easily available body fluids such as plasma and urine with small sample volumes for biomarker analysis.Amino acids play critical roles in metabolism of the body,which are necessary to maintain normal cell function and biological survival.Amino acids that changes with the change of tumor cell metabolism may reflect the formation,growth,infiltration and metastasis of primary tumors to some extent.In the metabolomics research of CRC,it has been reported the difference of amino acids in the blood,urine and feces of CRC patients and healthy volunteers?HVs?.There are also differences between cancerous tissues and normal tissues.Based on the above factors,firstly,this study established a liquid chromatography-tandem mass spectrometry?LC-MS/MS?method for the detection of 34amino acids in cancerous tissue of CRC patients,including glycine?Gly?,alanine?Ala?,serine?Ser?,valine?Val?,threonine?Thr?,leucine?Leu?,isoleucine?Ile?,aspartic acid?Asp?,lysine?Lys?,glutamine?Gln?,glutamic acid?Glu?,phenylalanine?Phe?,arginine?Arg?,tyrosine?Tyr?,proline?Pro?,asparagine?Asn?,methionine?Met?,tryptophan?Trp?,cysteine?Cys?,histidine?His?,citrulline?Cit?,asymmetric dimethylarginine?ADMA?,cystine?Cyss?,sarcosine?Sar?,3-aminopropanoic acid??-alanine,Apa?,3-amino-2-methylpropanoic acid??-aminoisobutyric acid,Amp?,4-aminobutyric acid??-aminobutyric acid,Aba?,5-oxo-proline?Opr?,4-hydroxy-proline?Hpr?,ornithine?Orn?,2-amino-hexanoic diacid??-aminoadipic acid,Ahd?,hippuric acid?Hia?,symmetric dimethylarginine?SDMA?and kynurenine?Kyn?.This method required only 100 mg tissue sample for homogenizing and avoided derivatization of amino acids.The analysis completed within 10 minutes,with good specificity and linearity.Extraction recovery and matrix effect,lower limit of quantification,intra-and inter-day accuracy and precision,stability,and dilution effect for all the analytes conformed to the requirements of biological sample analysis.This method was suitable for high-throughput analysis of tissue samples.This method was successfully applied to determine the contents of 34 amino acids in cancerous tissue,paracancerous tissue and normal tissue samples of 94 CRC patients.Then the targeted amino acids metabolomics study was performed.The results showed that the contents of Gly,Asp,Glu,Pro,Cys,ADMA,Cyss,Kyn,Orn,Aba,Apa,Hpr,Ala,Thr,Val,Leu,Ile,Phe,Tyr,Asn,Met,Trp,His,Opr,SDMA,Sar and Amp in cancerous tissues were significantly higher than those in paracancerous tissues and normal tissues.This observation had been confirmed that the metabolic profiles of amino acids in cancerous tissues were quite different from those in non-cancerous tissues.With the aid of orthogonal partial least squares-discriminant analysis?OPLS-DA?model,the cancerous tissue was distinguished from the non-cancerous tissue.Eleven amino acids,including Apa,Kyn,Cyss,Cys,Hpr,ADMA,Gly,Asp,Ala,Val and His,had been discovered as potential biomarkers with variable importance for the projection?VIP?values?29?1.0.Among them,the VIP values of Apa,Kyn and Cyss were more than 1.5.And these 3 amino acids could be used as potential biomarkers for diagnosis of colorectal cancerous tissues.Secondly,32 amino acids,including Gly,Ala,Val,Lys,Leu,Ile,Gln,Glu,Met,His,Phe,Arg,Tyr,Trp,Ser,Pro,Thr,Opr,Asn,Orn,Cit,Cyss,Cys,Hpr,Asp,ADMA,SDMA,Kyn,Apa,Sar,Amp and Hia in CRC patient plasma,were analyzed by a modified LC-MS/MS method.These 32 amino acids without derivatization were separated by reverse-phase C₁₈ column,and the total analytical time was within 10 min,with only a small volume of plasma?50?l?.This method was also validated in good specificity,linearity,extraction recovery and matrix effect,lower limit of quantification,intra-and inter-day accuracy and precision,stability,and dilution effect.This method was proved to be reliable for high-throughput analysis of plasma samples.Similaryly,the targeted amino acid metabolomics study was carried out in plasma samples of 246 CRC patients,as well as 411 HVs and 33 precancerosis patients.The results showed that the contents of Gln,Ser,Glu,Asn,Opr,Orn,Lys,Cys,Amp,SDMA and Asp in plasma samples of CRC patients were significantly higher than those in plasma samples of HVs,and the contents of Gly,Ala,Val,Phe,Trp,Pro,Arg,Met,Tyr,Cyss,His,Hia,Hpr,Sar,Apa,Leu and Kyn were all significantly lower than those in plasma samples of HVs.It was of big difference of the plasma amino acid metabolism profile between CRC patients and HVs.The CRC patients and the HVs could be separated in the OPLS-DA model.Fourteen amino acids,including Trp,Sar,Glu,Ser,Met,Ala,Cys,Cyss,Tyr,Opr,Apa,Gln,Hia and Arg,were obtained as promising biomakers with VIP values more than 1.Among them,the VIP values of Trp,Sar and Glu were more than 1.5.A combined diagnostic model was subsequently established based on logistics regression,employing Trp,Sar and Glu.This diagnostic model equation was 0.001×Trp+0.029×Sar-0.002×Glu-9.427?the unit of Trp,Sar,Glu was ng/ml?.The cut off value of the combined factor was 2.433(less than the cut off value,diagnosed as CRC patient.This model showed higher diagnostic efficacy than traditionally used serum indicators,such as carcinoembryonic antigen?CEA?and?-fetoprotein,as well as any other single amino acid.The diagnostic evaluation indicators of this model were all better than those of CEA diagnosis,with the accuracy of 91.9%,the sensitivity of 83.3%,the specificity of 96.6%,the positive predictive value of 93.2%,the negative predictive value of 91.3%for the above-mentioned plasma samples.This diagnostic model might be used for the auxiliary diagnosis of CRC.Lastly,Trp and Kyn contents in 154 plasma samples of CRC patients and 154 plasma samples of HVs,and the contents and enzyme relative activities of indolamine2,3-dioxygenase 1?IDO1?in the same samples were compared in this study.The enzyme relative activities of IDO1 in 94 cancerous tissue samples,94 paracancerous tissue samples and 94 normal tissue samples of CRC patients were also compared.The results showed that the content of IDO1 in plasmas of CRC patients was slightly higher than that of the HVs,but there was no significant difference in two groups.The enzyme relative activity of IDO1 in plasmas of CRC patients was significantly higher than that of the HVs.The enzyme relative activity of IDO1 in cancerous tissues of CRC patients was significantly higher than that in paracancerous tissues and normal tissues.It preliminary confirmed that the disorders of tryptophan metabolism existed in CRC patients.