Identification and Validation of Urinary Metabolite Biomarkers for Major Depressive Disorder

Abstract

Major depressive disorder (MDD) is a widespread and debilitating mental disorder. However, there are no biomarkers available to aid in the diagnosis of this disorder. In this study, a nuclear magnetic resonance spectroscopy–based metabonomic approach was employed to profile urine samples from 82 first-episode drug-naïve depressed subjects and 82 healthy controls (the training set) in order to identify urinary metabolite biomarkers for MDD. Then, 44 unselected depressed subjects and 52 healthy controls (the test set) were used to independently validate the diagnostic generalizability of these biomarkers. A panel of five urinary metabolite biomarkers—malonate, formate, N-methylnicotinamide, m-hydroxyphenylacetate, and alanine—was identified. This panel was capable of distinguishing depressed subjects from healthy controls with an area under the receiver operating characteristic curve (AUC) of 0.81 in the training set. Moreover, this panel could classify blinded samples from the test set with an AUC of 0.89. These findings demonstrate that this urinary metabolite biomarker panel can aid in the future development of a urine-based diagnostic test for MDD. Major depressive disorder (MDD) is a widespread and debilitating mental disorder. However, there are no biomarkers available to aid in the diagnosis of this disorder. In this study, a nuclear magnetic resonance spectroscopy–based metabonomic approach was employed to profile urine samples from 82 first-episode drug-naïve depressed subjects and 82 healthy controls (the training set) in order to identify urinary metabolite biomarkers for MDD. Then, 44 unselected depressed subjects and 52 healthy controls (the test set) were used to independently validate the diagnostic generalizability of these biomarkers. A panel of five urinary metabolite biomarkers—malonate, formate, N-methylnicotinamide, m-hydroxyphenylacetate, and alanine—was identified. This panel was capable of distinguishing depressed subjects from healthy controls with an area under the receiver operating characteristic curve (AUC) of 0.81 in the training set. Moreover, this panel could classify blinded samples from the test set with an AUC of 0.89. These findings demonstrate that this urinary metabolite biomarker panel can aid in the future development of a urine-based diagnostic test for MDD. Major depressive disorder (MDD) 1The abbreviations used are:HChealthy controlMDDmajor depressive disorderNMRnuclear magnetic resonanceOPLS-DAorthogonal partial least-squares discriminant analysisTCAtricarboxylic acid. 1The abbreviations used are:HChealthy controlMDDmajor depressive disorderNMRnuclear magnetic resonanceOPLS-DAorthogonal partial least-squares discriminant analysisTCAtricarboxylic acid. is a debilitating mental disorder affecting up to 15% of the general population and accounting for 12.3% of the global burden of disease (1Kessler R.C. Berglund P. Demler O. Jin R. Koretz D. Merikangas K.R. Rush A.J. Walters E.E. Wang P.S. The epidemiology of major depressive disorder: results from the National Comorbidity Survey Replication (NCS-R).JAMA. 2003; 289: 3095-3105Crossref PubMed Scopus (6308) Google Scholar, 2Reynolds E. Brain and mind: a challenge for WHO.Lancet. 2003; 361: 1924Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar). Currently, the diagnosis of MDD still relies on the subjective identification of symptom clusters rather than empirical laboratory tests. The current diagnostic modality results in a considerable error rate (3Mitchell A.J. Vaze A. Rao S. Clinical diagnosis of depression in primary care: a meta-analysis.Lancet. 2009; 374: 609-619Abstract Full Text Full Text PDF PubMed Scopus (769) Google Scholar), as the clinical presentation of MDD is highly heterogeneous and the current symptom-based method is not capable of adequately characterizing this heterogeneity (4Chen L.S. Eaton W.W. Gallo J.J. Nestadt G. Understanding the heterogeneity of depression through the triad of symptoms, course and risk factors: a longitudinal, population-based study.J. Affect. Disord. 2000; 59: 1-11Crossref PubMed Scopus (111) Google Scholar). An approach that can be used to circumvent these limitations is to identify disease biomarkers to support objective diagnostic laboratory tests for MDD. healthy control major depressive disorder nuclear magnetic resonance orthogonal partial least-squares discriminant analysis tricarboxylic acid. healthy control major depressive disorder nuclear magnetic resonance orthogonal partial least-squares discriminant analysis tricarboxylic acid. Metabonomics, which can measure the small molecules in given biosamples such as plasma and urine without bias (5Nicholson J.K. Lindon J.C. Holmes E. Metabonomics: understanding the metabolic responses of living systems to pathophysiological stimuli via multivariate statistical analysis of biological NMR spectroscopic data.Xenobiotica. 1999; 29: 1181-1189Crossref PubMed Scopus (3256) Google Scholar), has been extensively used to characterize the metabolic changes of diseases and thus facilitate the identification of novel disease-specific signatures as putative biomarkers (6Nicholson J.K. Lindon J.C. Systems biology: metabonomics.Nature. 2008; 455: 1054-1056Crossref PubMed Scopus (1488) Google Scholar, 7Huang Z. Lin L. Gao Y. Chen Y. Yan X. Xing J. Hang W. Bladder cancer determination via two urinary metabolites: a biomarker pattern approach.Mol. Cell. Proteomics. 2011; 10M111.007922Abstract Full Text Full Text PDF PubMed Scopus (98) Google Scholar, 8Qi Y. Li P. Zhang Y. Cui L. Guo Z. Xie G. Su M. Li X. Zheng X. Qiu Y. Urinary metabolite markers of precocious puberty.Mol. Cell. Proteomics. 2011; 11M111.011072Abstract Full Text Full Text PDF PubMed Scopus (34) Google Scholar, 9Fischer R. Trudgian D.C. Wright C. Thomas G. Bradbury L.A. Brown M.A. Bowness P. Kessler B.M. Discovery of candidate serum proteomic and metabolomic biomarkers in ankylosing spondylitis.Mol. Cell. Proteomics. 2012; 11M111.013904Abstract Full Text Full Text PDF PubMed Scopus (73) Google Scholar, 10Cheng T.M.K. Lu Y.E. Guest P.C. Rahmoune H. Harris L.W. Wang L. Ma D. Stelzhammer V. Umrania Y. Wayland M.T. Identification of targeted analyte clusters for studies of schizophrenia.Mol. Cell. Proteomics. 2010; 9: 510-522Abstract Full Text Full Text PDF PubMed Scopus (23) Google Scholar). Nuclear magnetic resonance (NMR) spectroscopy–based metabonomic approaches characterized by sensitive, high-throughput molecular screening have been employed previously in identifying novel biomarkers for a variety of neuropsychiatric disorders, including stroke, bipolar disorder, and schizophrenia (11Jung J.Y. Lee H.S. Kang D.G. Kim N.S. Cha M.H. Bang O.S. Ryu do H. Hwang G.S. 1H-NMR-based metabolomics study of cerebral infarction.Stroke. 2011; 42: 1282-1288Crossref PubMed Scopus (145) Google Scholar, 12Sussulini A. Prando A. Maretto D.A. Poppi R.J. Tasic L. Banzato C.E. Arruda M.A. Metabolic profiling of human blood serum from treated patients with bipolar disorder employing 1H NMR spectroscopy and chemometrics.Anal. Chem. 2009; 81: 9755-9763Crossref PubMed Scopus (58) Google Scholar, 13Yang J. Chen T. Sun L. Zhao Z. Qi X. Zhou K. Cao Y. Wang X. Qiu Y. Su M. Potential metabolite markers of schizophrenia.Mol. Psychiatry. 2011; 10.1038/mp.2011.131Google Scholar). Specifically with regard to MDD, several animal studies have already characterized the metabolic changes in the blood and urine (14Zhang F. Jia Z. Gao P. Kong H. Li X. Lu X. Wu Y. Xu G. Metabonomics study of urine and plasma in depression and excess fatigue rats by ultra fast liquid chromatography coupled with ion trap-time of flight mass spectrometry.Mol. Biosyst. 2010; 6: 852-861Crossref PubMed Scopus (58) Google Scholar, 15Li Z.Y. Zheng X.Y. Gao X.X. Zhou Y.Z. Sun H.F. Zhang L.Z. Guo X.Q. Du G.H. Qin X.M. Study of plasma metabolic profiling and biomarkers of chronic unpredictable mild stress rats based on gas chromatography/mass spectrometry.Rapid Commun. Mass Spectrom. 2010; 24: 3539-3546Crossref PubMed Scopus (66) Google Scholar, 16Zheng S. Zhang S. Yu M. Tang J. Lu X. Wang F. Yang J. Li F. An 1 H NMR and UPLC–MS-based plasma metabonomic study to investigate the biochemical changes in chronic unpredictable mild stress model of depression.Metabolomics. 2011; 7: 413-423Crossref Scopus (34) Google Scholar, 17Zheng S. Yu M. Lu X. Huo T. Ge L. Yang J. Wu C. Li F. Urinary metabonomic study on biochemical changes in chronic unpredictable mild stress model of depression.Clin. Chim. Acta. 2010; 411: 204-209Crossref PubMed Scopus (124) Google Scholar, 18Liu X.J. Li Z.Y. Li Z.F. Gao X.X. Zhou Y.Z. Sun H.F. Zhang L.Z. Guo X.Q. Du G.H. Qin X.M. Urinary metabonomic study using a CUMS rat model of depression.Magn. Reson. Chem. 2012; 50: 187-192Crossref PubMed Scopus (44) Google Scholar, 19Wang X. Zhao T. Qiu Y. Su M. Jiang T. Zhou M. Zhao A. Jia W. Metabonomics approach to understanding acute and chronic stress in rat models.J. Proteome Res. 2009; 8: 2511-2518Crossref PubMed Scopus (66) Google Scholar). These studies provide valuable clues as to the pathophysiological mechanism of MDD. However, no study has been designed with the aim of diagnosing this disease. Recently, using an NMR-based metabonomic approach, this research group identified a unique plasma metabolic signature that enables the discrimination of MDD from healthy controls with both high sensitivity and specificity (20Zheng P. Gao H.C. Li Q. Shao W.H. Zhang M.L. Cheng K. Yang D.Y. Fan S.H. Chen L. Fang L. Plasma metabonomics as a novel diagnostic approach for major depressive disorder.J. Proteome Res. 2012; 11: 1741-1748Crossref PubMed Scopus (173) Google Scholar). These findings motivated further study on urinary diagnostic metabolite biomarkers for MDD, which would be more valuable from a clinical applicability standpoint, as urine can be more non-invasively collected. Moreover, previous studies have also demonstrated the feasibility of identifying diagnostic metabolite biomarkers of psychiatric disorders in the urine. For example, using an NMR-based metabonomics approach, Yap et al. (21Yap I.K. Angley M. Veselkov K.A. Holmes E. Lindon J.C. Nicholson J.K. Urinary metabolic phenotyping differentiates children with autism from their unaffected siblings and age-matched controls.J. Proteome Res. 2010; 9: 2996-3004Crossref PubMed Scopus (201) Google Scholar) identified a unique urinary metabolite signature that clearly discriminated autism patients from healthy controls. As systemic metabolic disturbances have been observed in the urine of a depressed animal model, it is likely that diagnostic metabolite markers for MDD can be detected in human urine. Therefore, in this study, NMR spectroscopy combined with multivariate pattern recognition techniques were used to profile 82 first-episode drug-naïve MDD subjects and 82 healthy controls (the training set) in order to identify potential metabolite biomarkers for MDD. Furthermore, 44 unselected MDD subjects and 52 healthy controls (the test set) were employed to independently validate the diagnostic performance of these urinary metabolite biomarkers. Prior to the collection of urine samples, written informed consents were obtained from all subjects. The protocol of this study was reviewed and approved by the Ethical Committee of Chongqing Medical University. A total of 126 depressed subjects were recruited from the psychiatric center of the First Affiliated Hospital at Chongqing Medical University. All diagnoses were carried out according to the Structured Psychiatric Interview using DSM-IV-TR criteria (22American Psychiatric Association Diagnostic and Statistical Manual of Mental Disorders. American Psychiatric Association, Washington, D.C2001Google Scholar). The 17-item version of the observer-rated Hamilton Depression Rating Scale (HDRS) was used to assess depression severity (23Williams J.B.W. A structured interview guide for the Hamilton Depression Rating Scale.Arch. Gen. Psychiatry. 1988; 45: 742Crossref PubMed Scopus (1693) Google Scholar). The depressed subjects with HDRS scores of greater than 17 were recruited. The majority of these MDD subjects (n = 95) were first episode and drug naïve, and the remaining MDD subjects (n = 31) were being treated with various anti-depressants. The detailed individual demographic and clinical data of the recruited subjects are presented in supplemental Table S1. Exclusion criteria for the MDD subjects included any pre-existing physical or other mental disorders and/or illicit drug use. During the same healthy control subjects were recruited from the center of First Affiliated Hospital at Chongqing Medical University. controls were to have no previous of or systemic The recruited MDD subjects and healthy controls were a training set and a test set. The training including 82 first-episode drug-naïve MDD subjects and 82 healthy was used to identify potential urinary metabolite markers for the remaining subjects were used to the test set to independently validate the diagnostic generalizability of these urinary metabolite The of samples in the is an to with the of identified biomarkers in clinical studies C. S. G. of and the of National Washington, Google Scholar). The clinical of the recruited MDD subjects and healthy controls are in Table All MDD subjects on the HDRS than healthy controls in both training and test The MDD group and group not in of and mass in set. As for the MDD subjects and were in the training set not in the test and clinical of recruited test for and HDRS test for and HDRS as the as the as the healthy MDD, major depressive mass Hamilton Depression Rating for test for and HDRS as the in a healthy MDD, major depressive mass Hamilton Depression Rating for of the urine samples were in and All urine samples were at for The was and at For NMR urine samples were and at for to Then, of urine was with of 1 and at for samples of were NMR The were on a operating at 1H A was used and data were with a of an of and a of The was and an of was to the to resonance were according to from and and NMR (20Zheng P. Gao H.C. Li Q. Shao W.H. Zhang M.L. Cheng K. Yang D.Y. Fan S.H. Chen L. Fang L. Plasma metabonomics as a novel diagnostic approach for major depressive disorder.J. Proteome Res. 2012; 11: 1741-1748Crossref PubMed Scopus (173) Google Scholar, Angley M. Veselkov K.A. Holmes E. Lindon J.C. Nicholson J.K. Urinary metabolic phenotyping differentiates children with autism from their unaffected siblings and age-matched controls.J. Proteome Res. 2010; 9: 2996-3004Crossref PubMed Scopus Google Scholar, Nicholson J. A. P. Lindon J. S. M. J. Holmes E. Nuclear magnetic resonance spectroscopic and analysis biochemical of model Res. 11: PubMed Scopus Google Scholar). All were and to resonance at The NMR were using the of the and were in order to of The remaining in NMR were to the total of the to for in the The were as A multivariate approach, orthogonal partial least-squares discriminant analysis was on the data to discrimination and MDD subjects J. S. to Scopus Google Scholar, M. M. O. Nicholson J.K. Holmes E. J. discriminant the of and Scopus Google Scholar). The of the was in of and which were by the and were used to the was employed to assess the of the model S. of metabolomic data using support Chem. 2008; PubMed Scopus Google Scholar). out the of a test was Y. Lee J. J. Lee Ryu Hwang G.S. of the of by 1H NMR-based Chem. 2010; PubMed Scopus Google Scholar). the of and from the model were than the from the the model was S. of metabolomic data using support Chem. 2008; PubMed Scopus Google Scholar). The of the model were used to identify the for on the scores O. J. A. Lindon J.C. Nicholson J.K. Holmes E. of the orthogonal on model limitations by and of biomarker changes in 1H NMR spectroscopic metabonomic Chem. PubMed Scopus Google Scholar). on the of samples used to the a of was as a for statistical based on the discrimination at the of = The of the are to the of the to the discrimination based on a high and no Moreover, of the of NMR the of in the NMR was to the in the samples (the of the were to that of which was the samples as a the identified as to the discrimination MDD subjects and using multivariate analysis were by of The test was used to the two The in identifying a set of urinary metabolite biomarkers for MDD is in diagnosis based on the of a small of would be more and in clinical a based on was employed to the metabolite biomarker K. Li L. Zheng Y. R. D. M. M. T. E. of cancer and to by a biomarker J. 2008; PubMed Scopus Google Scholar). further the diagnostic performance of this set of MDD a characteristic curve analysis was carried out to the of this metabolite biomarker panel to MDD subjects and in both training and test The of the area under the curve in the of Scopus Google Scholar). As of demographic were using the the or the test A of than was In the training analysis was carried out to the metabolic MDD subjects and 1H NMR of urine obtained from an MDD and an are in supplemental S1. The to the are The of the model that the MDD subjects were from with mild = = = The of the and the model were a metabolic MDD subjects and Furthermore, a test was employed to validate the The demonstrated that the model was as the and to the were than all and to the independently validate the diagnostic performance of the model, the model was used to in the test set. The from the model demonstrated that of the 44 MDD subjects and of the 52 were by the model, a of These results that this model by urinary metabolite profiling as an empirical diagnostic for MDD. of the in the identification of with a of The of these for MDD subjects and is presented in Table to MDD subjects were characterized by of formate, and and of m-hydroxyphenylacetate, N-methylnicotinamide, and statistical analysis was to validate the metabolic changes identified through multivariate statistical the majority of In the other that were not as the MDD and are in supplemental Table The of these are presented in Table urinary for the discrimination MDD subjects and of in MDD subjects to in MDD subjects to were from the of in MDD subjects to in MDD subjects to were from the in a In order to identify a metabolite biomarker panel for MDD a based on was A analysis demonstrated that the MDD subjects and could be by five metabolites: formate, N-methylnicotinamide, m-hydroxyphenylacetate, and Therefore, these five biomarkers the for future diagnostic An analysis was further to the diagnostic performance of this panel in both training and test The area under the curve (AUC) of this panel was 0.81 in the training samples MDD subjects and 82 and in the test samples MDD subjects and 52 and The diagnostic performance of this panel is to that of the model with all the the of this urinary metabolite panel in MDD MDD is a widespread and debilitating mental disorder. Currently, no biomarkers are available to aid in diagnosing this disorder. In this study, an NMR-based metabonomic approach was employed to identify potential urinary metabolite biomarkers for MDD. A panel of five urinary metabolite biomarkers—malonate, formate, N-methylnicotinamide, m-hydroxyphenylacetate, and alanine—was identified. This panel the discrimination of MDD subjects from with of 0.81 and in the training set and test These findings demonstrate that urinary metabolite biomarkers can facilitate MDD and could aid the development of objective diagnostic for MDD. In order to the urinary metabolite biomarkers that the changes in the MDD disease first-episode drug-naïve MDD subjects were recruited the training set. However, given that is in MDD unselected and the test set were used to independently validate the diagnostic generalizability of the biomarkers. Furthermore, the MDD subjects and were not age-matched in this test set. these the panel still blinded MDD subjects from with an AUC of in the test the diagnostic of the biomarker In this study, were identified that MDD subjects from The of these discrimination MDD subjects and in the test set with an of This a diagnostic performance of these However, in clinical it is not or to measure a of in order to a disease Therefore, using a based on a biomarker panel of five was to MDD subjects from high of this the biomarker panel is likely to be of more clinical than from previous metabolomic studies on MDD (20Zheng P. Gao H.C. Li Q. Shao W.H. Zhang M.L. Cheng K. Yang D.Y. Fan S.H. Chen L. Fang L. Plasma metabonomics as a novel diagnostic approach for major depressive disorder.J. Proteome Res. 2012; 11: 1741-1748Crossref PubMed Scopus (173) Google Scholar, L.A. K.R. R. D.C. A metabolomic analysis of with and without J. PubMed Scopus Google Scholar). the five biomarkers in the of and were not in the statistical However, these were included in the diagnostic as were identified by multivariate This was the analysis that the of these two in the This the of a multivariate statistical approach in the potential of metabolic to an analysis D.A. H. M. J. A. analysis by chronic molecular 2010; 24: PubMed Scopus (124) Google Scholar). the of MDD, the were in of in metabolic These were to be in and which are in Urinary of tricarboxylic and in MDD subjects to and are the two metabolic the that The and in the urine of MDD subjects in with the of plasma and in MDD subjects from this previous study (20Zheng P. Gao H.C. Li Q. Shao W.H. Zhang M.L. Cheng K. Yang D.Y. Fan S.H. Chen L. Fang L. Plasma metabonomics as a novel diagnostic approach for major depressive disorder.J. Proteome Res. 2012; 11: 1741-1748Crossref PubMed Scopus (173) Google Scholar), likely greater through the by of which were in MDD subjects to This in be a mechanism to for an in Moreover, the of the metabolite also observed in MDD subjects is from a metabolite in with This likely an in MDD which is with the in MDD subjects given that is the primary for this be with the in previously observed in MDD subjects. coupled with the and for the the human a of total this of total A. Brain and and Google Scholar). Therefore, the in to chronic in the of MDD subjects. previous studies have a in in several of MDD patients J.C. C.E. of to of Gen. Psychiatry. PubMed Scopus Google Scholar, P. of and blood in major depressive disorder: a 2000; PubMed Scopus Google Scholar). the was in these the depressive were S.H. K.R. S. H.S. S. S. in with of major J. Psychiatry. PubMed Scopus Google Scholar). Moreover, the urinary of formate, an from MDD subjects in the current study a for These findings that in MDD subjects be by metabolic Urinary of five and in MDD subjects to These are by in the that MDD be with in previous urinary metabonomic analysis in a depressed animal model has that depressed is with changes in S. Yu M. Lu X. Huo T. Ge L. Yang J. Wu C. Li F. Urinary metabonomic study on biochemical changes in chronic unpredictable mild stress model of depression.Clin. Chim. Acta. 2010; 411: 204-209Crossref PubMed Scopus (124) Google Scholar). several clinical studies have demonstrated that MDD patients a high of Depression in patients with in J. PubMed Scopus Google Scholar, and severity of the of the disorders and Disord. 2009; PubMed Scopus Google Scholar), a disorder The in patients of and of and J. J. A of the of the in and the of J. PubMed Google Scholar). These combined findings the potential of in the development of MDD. an of was in MDD subjects to that the is in the G. of the in PubMed Google Scholar), the of urinary observed an of in MDD subjects. is the biochemical of both and (21Yap I.K. Angley M. Veselkov K.A. Holmes E. Lindon J.C. Nicholson J.K. Urinary metabolic phenotyping differentiates children with autism from their unaffected siblings and age-matched controls.J. Proteome Res. 2010; 9: 2996-3004Crossref PubMed Scopus (201) Google Scholar). Therefore, in the of This is in with the that to the of MDD R. G. Major depressive J. 2008; PubMed Scopus Google Scholar). The results and of this study be on of several The diagnostic performance of the urinary metabolite biomarker panel was by MDD subjects from on or not these biomarkers can be to MDD from other psychiatric Moreover, all subjects were of the same and were recruited from the same and be studies heterogeneous from clinical are In with the of a 1H NMR-based metabonomic a panel of urinary metabolite biomarkers for MDD was identified using a set. This panel was independently in a set. metabolite biomarkers—malonate, formate, N-methylnicotinamide, m-hydroxyphenylacetate, and be used to MDD subjects from in both and test These findings the for the future development of a urine-based diagnostic test for MDD. is to Yang and for their in with