1H NMR Serum Metabonomics for Understanding ... - ACS Publications

Article pubs.acs.org/jpr

1

H NMR Serum Metabonomics for Understanding Metabolic Dysregulation in Women with Idiopathic Recurrent Spontaneous Miscarriage during Implantation Window

Priyanka Banerjee,† Mainak Dutta,† Sudha Srivastava,‡ Mamata Joshi,‡ Baidyanath Chakravarty,§ and Koel Chaudhury*,† †

School of Medical Science and Technology, Indian Institute of Technology Kharagpur, Kharagpur Pin-721302, India National Facility for High-field NMR, Tata Institute of Fundamental Research, Mumbai Pin-400005, India § Institute of Reproductive Medicine, Kolkata Pin-700020, India ‡

ABSTRACT: In an attempt to find out the association of metabolic dysregulation with poor endometrial receptivity and pregnancy loss, serum metabonomic profiling of women with idiopathic recurrent spontaneous miscarriage (IRSM) is carried out and compared with fertile controls. 1H nuclear magnetic resonance (NMR)-based metabonomics was used to obtain serum metabolic profiles of 36 women with IRSM and 28 proven fertile women during the window of implantation. The acquired data were analyzed using multivariate principal component analysis, partial least-squares-discriminant analysis, and orthogonal projection to latent structure with discriminant analysis. A clear metabolic differentiation was evident between IRSM and control samples. The distinguishing metabolites, Llysine, L-arginine, L-glutamine, L-histidine, L-threonine, L-phenylalanine, and L-tyrosine are significantly up-regulated in IRSM as compared to controls. These altered metabolites may be involved in the molecular mechanism of exaggerated inflammatory response and vascular dysfunction associated with poor endometrial receptivity in women with IRSM. The present work proposes a vital association of metabolic dysfunction with the disease pathogenesis. KEYWORDS: NMR, metabonomics, multivariate analysis, idiopathic recurrent spontaneous miscarriage, implantation window

■

INTRODUCTION

NMR spectroscopy is an extensively used technique for metabolic fingerprinting and metabolite profiling of biofluids and tissues.11 The increasing popularity of this technique is its ability to identify and quantify a broad range of metabolites with nominal or no sample preparation.12 Moreover, the analysis is nondestructive in nature and only a small quantity of the sample is required.13 The technological advantages of NMR is its rapidity (only 1−15 min is required for profiling of metabolites in a biological sample) and sensitivity to discriminate fine biological differences.14 Complex matrices, comprising several hundreds of proton signals originating from various metabolites,15 represent highresolution 1H NMR spectroscopy data sets.15 This complexity can be simplified by the application of two distinct approaches, chemometrics and targeted profiling. In chemometrics, the NMR spectra are separated into different groups with no assumptions about the identity and quantity of metabolites in the spectra.4,16,17 On the contrary, in targeted profiling, each metabolite in every single NMR spectrum is identified and quantified, so that metabolite concentrations are the variables.17 Various multivariate statistical methods such as principal

Recurrent spontaneous miscarriage (RSM) refers to three or more consecutive losses of pregnancy within 24 weeks of gestation. It affects approximately 1−3% of couples attempting to conceive.1 Various metabolic and endocrinologic abnormalities such as diabetes mellitus, polycystic ovary syndrome (PCOS), thyroid disorder, luteal phase defect, and hyperprolactinemia are considered to be major contributory factors to RSM and associated with approximately 17−20% of the cases.2 Consequently, there exists a high probability that serum metabolites in women with history of RSM will be differently expressed as compared to proven fertile women. Metabolic profiling has become a valuable platform for the identification of low-molecular weight metabolites and their intermediates present in the biological system at a specific time point. It provides unanticipated insights into the regulation of the cell metabolism,3 presents clinically useful information related to disease pathogenesis,4 and identification of diagnostic markers.5 The analytical techniques that are generally used to identify and quantify metabolites include nuclear magnetic resonance (NMR),6 liquid chromatography (LC) coupled with mass spectrometry (MS),7,8 and gas chromatography MS (GC/ MS).9,10 © 2014 American Chemical Society

Received: April 15, 2014 Published: April 17, 2014 3100

dx.doi.org/10.1021/pr500379n | J. Proteome Res. 2014, 13, 3100−3106

Journal of Proteome Research

Article

echo time of 38 ms. The strong signal from the solvent was suppressed using solvent presaturation pulse sequence. The spectra were recorded over 512 transients, 16 384 data points (TD), spectral width of 14 000 Hz, and an acquisition time of 0.58 s. Prior to Fourier transformation, line broadening of 0.3 Hz was applied to the FIDs. Phase and baseline correction was applied to the resulting spectra using MestReNova version 7.1.0 (Mestrelab Research, Santiago de Compostela, Spain).22,23 TSP served as the chemical shift reference point (δ = 0.00 ppm) and a concentration standard. Metabolites present in the samples were identified using earlier published articles and literature.24−28

component analysis (PCA), partial least-squares-discriminant analysis (PLS-DA), and orthogonal projection to latent structure with discriminant analysis (OPLS-DA) are frequently applied to both these approaches for the analysis of complex data structures.4,14 When the pathophysiology of recurrent miscarriage remains unknown, the condition is referred to as unexplained or idiopathic RSM (IRSM). Approximately 50% of the RSM cases are reported to be idiopathic.18 For understanding the altered cellular metabolism associated with poor endometrial receptivity in women with IRSM, serum metabolic profiles of women with IRSM are compared with those of fertile controls during the implantation window in the present study. 1H NMR spectroscopy (700 MHz) followed by multivariate statistical analysis was used for this purpose.

■

Data Preprocessing

The spectral region of 0.5 to 4.5 ppm was bucketed/binned into 2000 integrated segments of equal frequency window (0.002 ppm) using MestReNova. The region corresponding to water (4.5 to 5.15 ppm) was excluded during binning to avoid possible artifacts due to presaturation of water. Furthermore, the aromatic region (after 5.15 ppm), which had poor signal-tonoise ratio, was analyzed separately. The binned data matrix of 2000 variables were normalized and scaled to the working region of 0.5 to 4.5 ppm. The need to analyze metabonomics data sets by normalization and Pareto scaling is described by Craig et al.29 and van den Berg et al.30 Normalization (by sum) was performed on the integrated data sets to compensate for the variation in concentration between the samples and to represent each data point as a fraction of the total integral value of the spectra. Following normalization, Pareto scaling was performed. This scaling technique divides the mean-centered data by the square root of the standard deviation to provide equal weight to all variables, irrespective of their absolute value.31 The preprocessed data was subjected to multivariate data analysis using SIMCA 13.0.2 (Umetrics, Sweden).32 Univariate analysis was used to integrate and process the peaks beyond 5.15 ppm (aromatic region).

MATERIALS AND METHODS

Subject Selection

Women with IRSM reporting at the Institute of Reproductive Medicine, Kolkata, India for infertility treatment were enrolled in this study. The study was approved by the Institute Ethics Committee, and written informed consent was obtained from all couples. The women were randomly selected from the cohort of subjects described in our earlier studies.19,20 Briefly, two different groups of subjects were included: (i) women with a history of IRSM and (ii) proven fertile women undergoing sterilization (controls) for comparison purposes. The IRSM group was composed of 31 women (age < 35 years; BMI ≤ 28) with a history of three or more consecutive miscarriages of unknown etiology occurring within the first trimester (up to 12 weeks of gestation). It was ensured that these women did not have any other gynecological disorder, had not received any medication over the past three months, and did not have any apparent cause of recurrent miscarriage. The control group consisted of 28 age and BMI-matched women having normal regular menstrual cycle and parity between 2 to 5. It was ensured that these women have had no failed pregnancies and no other clinical abnormalities.

Multivariate Statistical Analysis

Following data preprocessing, unsupervised PCA and supervised classifications including PLS-DA and OPLS-DA were performed using SIMCA 13.0.2 (Umetrics, Sweden). PCA, which reduces complexity of multidimensional data sets in an unbiased manner, was initially applied to separate the IRSM cases from controls. General trends, patterns, and outliers were identified from the principal component (PC) score plots of the reduced dimension data. Following PCA, supervised classifications such as PLS-DA and OPLS-DA were used to further maximize separation between the two groups. Concentration differences in the metabolites between the two groups are shown as coefficient of variation plots. A correlation coefficient of ±0.65 was considered as the threshold to select the variables that best correlated with the discriminative scores of OPLS-DA. R2 and Q2 values are important indicators used to assess the validity of the model. The OPLS-DA model was cross-validated by permutation analysis using 100 different model permutations.

Sample Collection and Preparation

Following confirmation of ovulation, venous blood samples were collected from women of both the groups during day 18− 22 of their menstrual cycle. Samples were incubated at room temperature for 45 min to allow clotting followed by centrifugation at 1500 × g for 10 min. Two hundred microliter aliquots of the supernatant (serum) were transferred into sterile cryovials, frozen, and stored at −80 °C. The samples were prepared for NMR analysis, as described previously.4,5 Briefly, serum samples were thawed and homogenized before performing the NMR experiments. Two hundred microliters of thawed serum sample was mixed with 400 μL of deuterium oxide (D2O) containing 1 mM sodium salt of 3-(trimethylsilyl) propionic-2,2,3,3,d4 acid (TSP) and centrifuged at 8000 rpm for 5 min. From each sample, 600 μL of supernantant was transferred to 5 mm NMR tubes and subjected to analysis.

Quantitative Analysis of Variations in Selected Metabolites

NMR Measurements

The relaxation property of a particular chemical group may influence the signal intensity of a CPMG spectra. It is, however, presumed that any such influence in signal intensity would be consistent across samples and thus not affect the relative changes in metabolite concentration.33 NMR peaks of

NMR experiments were performed as discussed earlier.4,5 Briefly, Carr−Purcell−Meiboom−Gill (CPMG)21 spin−echo spectra were recorded on a 700 MHz Bruker Avance AV III spectrometer at 298 K with 16 K data points each and a total 3101



Article

Figure 1. 1H NMR (700 MHz) CPMG spectra (δ10.0−0.0) of serum obtained from (A) a woman with history of IRSM and (B) control.

PLS-DA method, filters out variability that is not directly related to the class separation and thereby facilitates easy interpretation.34−36 OPLS-DA, in the present study, optimized separation between the two groups (Figure 3C). In the permutation test, R2 and Q2 values were significantly higher than the permutated models, indicating a good predictive ability (Figure 4). Several procedures were used for the identification of the main metabolites responsible for separation between IRSM and controls. First, the key metabolites were identified using the correlation-based approach and then extracted from the S-line plot using SIMCA (Figure 5). The variables with higher values of correlation coefficient were the major metabolites contributing maximum toward cluster formation between the two groups. These metabolites were further analyzed based on their VIP values. Variables with VIP scores > 1 were considered to be influential for effective class separation. Variables with VIP scores < 1 were not considered further. Ten metabolites met the above screening procedure. Next, to validate the binning technique, a relative quantitative metabolomic approach was applied. In this approach, expected chemical-shift and couplingconstant values were used to identify the metabolites.37−40 Baseline corrections were carried out, wherever necessary. The distinguishing metabolites, summarized according to p < 0.05 in Table 1, indicate that L-lysine, L-arginine, L-glutamine, Lhistidine, and L-threonine are significantly up-regulated in IRSM as compared to controls. Since relative quantification

significantly altered metabolites, identified by multivariate statistical analysis, were integrated for quantitative measurements. MestReNova version 7.1.0 (Mestrelab Research) was used to integrate individual peaks. Normalization to total intensity of the spectra was performed by dividing the identified peak integrals within a spectrum by the total integral of that spectrum. Signal integrals representing significantly altered metabolites were then compared between women with IRSM and controls. All integrals were compared using Student’s t test (Ky Plot version 2.0 beta 13 software, Koichi Yoshioka). Statistical significance was considered to be p < 0.05. The aromatic peaks were separately quantified and analyzed using nonparametric Mann−Whitney U test (Ky Plot version 2.0 beta 13 software). A p value < 0.05 was considered to be significant.

■

RESULTS Typical serum 1H NMR CPMG spectra of women with IRSM and controls are shown in Figure 1A,B. Figure 2 depicts the signals of common metabolites, which include organic acids, amino acids, and carbohydrates. The peaks are assigned according to previous publications and literatures.24−28 PCA of binned spectra showed unbiased clustering among diseased and control groups. The trend toward unsupervised separation between IRSM and controls is evident in the t1 vs t2 scores scatter plot (Figure 3A). PLS-DA further increased the class separation (Figure 3B). OPLS-DA, an extension of the 3102



Article

Figure 2. 1H NMR (700 MHz) CPMG serum spectrum of a patient with history of IRSM. Numbers indicate the following metabolites: 1, TSP; 2, lipoproteins (LDL and VLDL); 3, unsaturated lipid; 4, creatinine; 5, L-arginine; 6, glycerophosphatidylcholine; 7, D-glucose; 8, ornithine; 9, citrate; 10, L-lysine; 11, L-tyrosine; 12, L-histidine; 13, L-phenylalanine; 14, formate; 15, choline; 16, L-threonine; 17, acetate; 18, L-glutamine; 19, succinate; 20, acetone; 21, adipic acid; 22, L-isoleucine; 23, alanine; 24, L-aspartate; 25, 3-hydroxybutyric acid; 26, propylene glycol; 27, valine; 28, leucine; 29, creatine; 30, pyruvate; 31, lactate; 32, proline.

Figure 3. Scores scatter plot of (A) PCA showing discrimination between IRSM and controls, (B) PLS-DA showing improved discrimination with R2 = 0.843 and Q2 = 0.709, and (C) OPLS-DA showing optimized discrimination with R2 = 0.882 and Q2 = 0.751.

metabolites in serum of women with IRSM for an improved understanding of the disease mechanism. Multivariate analysis of the global metabolite profiling clearly distinguishes women with IRSM from controls. A total of seven metabolites were found to be up-regulated in women with IRSM (Table 1). The PCA scores plot shows a clear separation between IRSM and controls on the first two components (Figure 3A). An improved class separation was observed in PLS-DA (Figure 3B), whereas OPLS-DA showed an optimized class separation (Figure 3C). The robustness of the OPLS-DA model is assessed by two parameters: R2 (goodness of fit) and Q2 (predictive ability). The values of R2 (0.882) and Q2 (0.751) in the present OPLS-DA model indicate that the model has good fit and can satisfactorily predict IRSM.

could not be carried out for the lipoprotein regions due to spectral overlapping, only the first two validation approaches were relied on to assess its significance. Further, on applying Mann−Whitney U test to the aromatic region, L-phenylalanine and L-tyrosine were found to be significantly increased in IRSM compared with controls.

■

DISCUSSION

Metabonomics is the study of individual metabolic profiles under normal and altered or disease-related conditions.41,14 It is an effective tool for understanding potential biological alterations,42,11 monitoring disease progression,43 and distinguishing between diseased and nondiseased states.44 This study, for the first time, attempts to identify differently expressed 3103



Article

Figure 4. Response permutation test (n = 100) to estimate the statistical significance of the OPLS-DA model. The R2 and Q2 values on the extreme right-hand side of the plot are of the true model, whereas the permutated model parameters are represented on the left-hand side of the plot. The correlation coefficients between true and permutated models represent the X axis. The true class has a correlation of 1.0 with itself. The true model parameters in the validation test exhibited higher values than those of the permutated models. Y-axis intercepts: R2 = (0.0, 0.308) and Q2 = (0.0, −0.236).

Figure 5. Significance of variation in metabolites between IRSM and controls is represented by the color map plotted in MATLAB R2009a (The MathWorks, Inc., USA). A color bar on the right-hand side of the plot represents the modulus of correlation for the metabolites responsible for discriminating IRSM and controls. The blue color represents least correlation, whereas red color represents highest correlation. Peaks with positive loading signify increased metabolites in IRSM (2, L-lysine; 4, L-arginine; 6, L-glutamine; 8, L-valine; 9, L-histidine; 10, L-threonine) in comparison with controls. Decreased metabolites in IRSM (1, lipoproteins; 3, adipic acid; 5, proline; 7, acetone) are shown as peaks with negative loading.

Permutation test statistics was further used to validate the predictive capability of the OPLS-DA model. In this test the R2 and Q2 values of the true model are compared with that of the permutated model. The test is performed by randomly assigning samples to the two different groups, following which OPLS-DA models are fitted to each permutated class variable. R2 and Q2 values are then generated for the permutated models and compared to that of the true model. Results of the permutation test indicate that the true model has much higher R2 and Q2 values, and thus, the true model is far better than the one hundred permutated models (Figure 4). It is well accepted that nitric oxide (NO) is produced by the oxidation of L-arginine in a reaction catalyzed by nitric oxide

Table 1. Major Metabolites Contributing to the Discrimination between Women with History of IRSM and Controls metabolites L-lysine L-arginine L-glutamine L-histidine L-threonine phenylalanine tyrosine

δ 1H (ppm)

fold change (relative to controls)

3.05 1.68 2.09 3.98 4.24 7.33 7.17

1.71 2.1 1.63 2.78 10.56 1.72 1.56

p-value p p p p p p p

≤ ≤ ≤ ≤ ≤ ≤ ≤

0.01 0.05 0.05 0.001 0.01 0.01 0.05

3104



■

synthase (NOS).45 The present findings indicate an elevated level of L-arginine in serum of women with IRSM, which may be attributed to impairment in eNOS activity in these women. This hypothesis is based on the results of our previous work,20 where a significant decrease in eNOS and NO expression in IRSM women due to eNOS polymorphism were observed. We also found the level of L-lysine in serum of IRSM women to be significantly higher as compared to controls. Since L-lysine uses the same intracellular transporter system as L-arginine and therefore competes with L-arginine for its transport within the cells,46 it is logical to presume that accumulation of L-lysine in blood occurs due to increased availability of L-arginine. Histidine is a precursor molecule for histamine, which is involved in inflammatory response. 47 L-Glutamine has immunomodulatory activity48 and is associated with the production of T lymphocyte-derived cytokines including IL1β and TNF-α.49 Also, it is proven that a high threonine concentration is required for inflammation.50 The elevated levels of L-histidine, L-glutamine, and L-threonine in women with IRSM (Table 1) underline the fact that these metabolites are associated with the elevated immune response in these women. These findings are strengthened by the fact that there is evidence of exaggerated pro-inflammatory response in IRSM, as discussed in our earlier work.19 Further, altered expression of lipids corresponding to specific bins was also seen in women with IRSM. This may be attributed to the apolipoprotein E (Apo E) gene polymorphisms associated with recurrent pregnancy loss.51 Other metabolites of potential interest due to their altered expression in IRSM are L-phenylalanine and Ltyrosine. We found elevated levels of L-phenylalanine and Ltyrosine in women with IRSM as compared with controls. Phenylalanine is irreversibly converted to tyrosine by the ratelimiting enzyme, phenylalanine hydroxylase.52 An association between IRSM and enzymatic defects in catabolism of phenylalanine and tyrosine seems likely. The tyrosine catabolic pathway in women with IRSM warrants further investigation.

REFERENCES

(1) Toth, B.; Jeschke, U.; Rogenhofer, N.; Scholz, C.; Würfel, W.; Thaler, C. J.; Makrigiannakis, A. Recurrent miscarriage: current concepts in diagnosis and treatment. J. Reprod. Immunol. 2010, 85 (1), 25−32. (2) Ford, H. B.; Schust, D. J. Recurrent pregnancy loss: etiology, diagnosis, and therapy. Rev. Obstet. Gynecol. 2009, 2 (2), 76−83. (3) Lafaye, A.; Junot, C.; Pereira, Y.; Lagniel, G.; Tabet, J. C.; Ezan, E.; Labarre, J. Combined proteome and metabolite-profiling analyses reveal surprising insights into yeast sulfur metabolism. J. Biol. Chem. 2005, 280 (26), 24723−24730. (4) Jana, S. K.; Dutta, M.; Joshi, M.; Srivastava, S.; Chakravarty, B.; Chaudhury, K. 1H NMR based targeted metabolite profiling for understanding the complex relationship connecting oxidative stress with endometriosis. Biomed. Res. Int. 2013, 2013, 329058. (5) Dutta, M.; Joshi, M.; Srivastava, S.; Lodh, I.; Chakravarty, B.; Chaudhury, K. A metabonomics approach as a means for identification of potential biomarkers for early diagnosis of endometriosis. Mol. Biosyst. 2012, 8 (12), 3281−3287. (6) Bingol, K.; Brüschweiler, R. Multidimensional approaches to NMR-based metabolomics. Anal. Chem. 2014, 86 (1), 47−57. (7) Zhou, B.; Xiao, J. F.; Tuli, L.; Ressom, H. W. LC-MS-based metabolomics. Mol. Biosyst. 2012, 8 (2), 470−481. (8) Becker, S.; Kortz, L.; Helmschrodt, C.; Thiery, J.; Ceglarek, U. LC−MS-based metabolomics in the clinical laboratory. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci. 2012, 883−884, 68−75. (9) Koek, M. M.; Jellema, R. H.; van der Greef, J.; Tas, A. C.; Hankemeier, T. Quantitative metabolomics based on gas chromatography mass spectrometry: status and perspectives. Metabolomics 2011, 7 (3), 307−328. (10) Tsugawa, H.; Tsujimoto, Y.; Arita, M.; Bamba, T.; Fukusaki, E. GC/MS based metabolomics: development of a data mining system for metabolite identification by using soft independent modeling of class analogy (SIMCA). BMC Bioinf. 2011, 12, 131. (11) Shulaev, V. Metabolomics technology and bioinformatics. Briefings Bioinf. 2006, 7 (2), 128−139. (12) Beckonert, O.; Keun, H. C.; Ebbels, T. M.; Bundy, J.; Holmes, E.; Lindon, J. C.; Nicholson, J. K. Metabolic profiling, metabolomic andmetabonomic procedures for NMR spectroscopy of urine, plasma, serum and tissue extracts. Nat. Protoc. 2007, 2 (11), 2692−2703. (13) Price, W. S. Applications of Pulsed Gradient Spin-Echo NMR Diffusion Measurements to Solution Dynamics and Organization. In Diffusion Fundamentals; Karger, J., Grinberg, F., Heitjans, P., Eds; Leipzig University Press: Leipzig, Germany, 2005. (14) Zhang, S.; Nagana Gowda, G. A.; Ye, T.; Raftery, D. Advances in NMR-based biofluid analysis and metabolite profiling. Analyst 2010, 135 (7), 1490−1498. (15) Ramadan, Z.; Jacobs, D.; Grigorov, M.; Kochhar, S. Metabolic profiling using principal component analysis, discriminant partial least squares, and genetic algorithms. Talanta 2006, 68 (5), 1683−1691. (16) Trygg, J.; Holmes, E.; Lundstedt, T. Chemometrics in metabonomics. J. Proteome Res. 2007, 6 (2), 469−479. (17) Liu, J.; Litt, L.; Segal, M. R.; Kelly, M. J. S.; Pelton, J. G.; Kim, M. Metabolomics of oxidative stress in recent studies of endogenous and exogenously administered intermediate metabolites. Int. J. Mol. Sci. 2011, 12 (10), 6469−6501. (18) Magdoud, K.; Dendana, M.; Herbepin, V.; Hizem, S.; Ben Jazia, K.; Messaoudi, S.; Almawi, W. Y.; Touraine, R.; Mahjoub, T. Identification of specific vascular endothelial growth factor susceptible and protective haplotypes associated with recurrent spontaneous miscarriages. Hum. Reprod. 2012, 27 (5), 1536−1541. (19) Banerjee, P.; Jana, S. K.; Pasricha, P.; Ghosh, S.; Chakravarty, B.; Chaudhury, K. Proinflammatory cytokines induced altered expression of cyclooxygenase-2 gene results in unreceptive endometrium in women with idiopathic recurrent spontaneous miscarriage. Fertil. Steril. 2013, 99 (1), 179−187. (20) Banerjee, P.; Ghosh, S.; Dutta, M.; Subramani, E.; Khalpada, J.; Roychoudhury, S.; Chakravarty, B.; Chaudhury, K. Identification of key contributory factors responsible for vascular dysfunction in

■

CONCLUSIONS In the present work, with the combined use of 1H NMR spectroscopy and multivariate analysis, a clear metabolic discrimination is evident between IRSM and controls. It is envisaged that these distinguishing altered metabolites may be involved in the molecular mechanism associated with exaggerated inflammatory response and vascular dysfunction leading to poor endometrial receptivity in women with IRSM. The findings are expected to provide an improved understanding of the disease pathogenesis and hence facilitate development of suitable therapeutic management.

■

Article

AUTHOR INFORMATION

Corresponding Author

*(K.C.) Tel: 03222-283572. Fax: 03222-282221. E-mail: koel@ smst.iitkgp.ernet.in or [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors would like to acknowledge Francois Abram, ArthroLab Inc., ArthroVision, Imaging Research & Development, Canada and Hamid Abdollahi, Kerman Graduate University of Technology, Kerma n, ̅ Iran, members of Researchgate, for their help with MATLAB scripts. 3105



Article

idiopathic recurrent spontaneous miscarriage. PLoS One 2013, 8 (11), e80940. (21) Meiboom, S.; Gill, D. Modified SpinEcho method for measuring nuclear relaxation times. Rev. Sci. Instrum. 1958, 29, 688. (22) Qin, X. Y.; Wei, F.; Tanokura, M.; Ishibashi, N.; Shimizu, M.; Moriwaki, H.; Kojima, S. The effect of acyclic retinoid on the metabolomic profiles of hepatocytes and hepatocellular carcinoma cells. PLoS One 2013, 8 (12), e82860. (23) Lourenço, A. B.; Roque, F. C.; Teixeira, M. C.; Ascenso, J. R.; Sá-Correia, I. Quantitative 1H-NMR-metabolomics reveals extensive metabolic reprogramming and the effect of the aquaglyceroporin FPS1 in ethanol-stressed yeast cells. PLoS One 2013, 8 (2), e55439. (24) Psychogios, N.; Hau, D. D.; Peng, J.; Guo, A. C.; Mandal, R.; Bouatra, S.; Sinelnikov, I.; Krishnamurthy, R.; Eisner, R.; Gautam, B.; Young, N.; Xia, J.; Knox, C.; Dong, E.; Huang, P.; Hollander, Z.; Pedersen, T. L.; Smith, S. R.; Bamforth, F.; Greiner, R.; McManus, B.; Newman, J. W.; Goodfriend, T.; Wishart, D. S. The human serum metabolome. PLoS One 2011, 6 (2), e16957. (25) Schicho, R.; Shaykhutdinov, R.; Ngo, J.; Nazyrova, A.; Schneider, C.; Panaccione, R.; Kaplan, G. G.; Vogel, H. J.; Storr, M. Quantitative metabolomic profiling of serum, plasma, and urine by (1) H NMR spectroscopy discriminates between patients with inflammatory bowel disease and healthy individuals. J. Proteome Res. 2012, 11 (6), 3344−3357. (26) Cao, M.; Zhao, L.; Chen, H.; Xue, W.; Lin, D. NMR-based metabolomic analysis of human bladder cancer. Anal. Sci. 2012, 28 (5), 451−456. (27) Bertini, I.; Cacciatore, S.; Jensen, B. V.; Schou, J. V.; Johansen, J. S.; Kruhøffer, M.; Luchinat, C.; Nielsen, D. L.; Turano, P. Metabolomic NMR fingerprinting to identify and predict survival of patients with metastatic colorectal cancer. Cancer Res. 2012, 72 (1), 356−364. (28) Wishart, D. S.; Knox, C.; Guo, A. C.; Eisner, R.; Young, N.; Gautam, B.; Hau, D. D.; Psychogios, N.; Dong, E.; Bouatra, S.; Mandal, R.; Sinelnikov, I.; Xia, J.; Jia, L.; Cruz, J. A.; Lim, E.; Sobsey, C. A.; Shrivastava, S.; Huang, P.; Liu, P.; Fang, L.; Peng, J.; Fradette, R.; Cheng, D.; Tzur, D.; Clements, M.; Lewis, A.; De Souza, A.; Zuniga, A.; Dawe, M.; Xiong, Y.; Clive, D.; Greiner, R.; Nazyrova, A.; Shaykhutdinov, R.; Li, L.; Vogel, H. J.; Forsythe, I. HMDB: a knowledgebase for the human metabolome. Nucleic Acids Res. 2009, 37 (Database issue), D603−610. (29) Craig, A.; Cloarec, O.; Holmes, E.; Nicholson, J. K.; Lindon, J. C. Scaling and normalization effects in NMR spectroscopic metabonomic data sets. Anal. Chem. 2006, 78 (7), 2262−2267. (30) van den Berg, R. A.; Hoefsloot, H. C.; Westerhuis, J. A.; Smilde, A. K.; van der Werf, M. J. Centering, scaling, and transformations: improving the biological information content of metabolomics data. BMC Genomics 2006, 7, 142. (31) Eriksson, L.; Johansson, E.; Kettaneh-Wold, N.; Wold, S. Introduction to Multi- and Megavariate Data Analysis Using Projection Methods (PCA & PLS); Umetrics: Umea, Sweden, 1999; pp 213−225. (32) Umetrics. SIMCA-P + The standard multivariate technology; Umetrics: Umea, Sweden, 2009. (33) MacIntyre, D. A.; Melguizo Sanchís, D.; Jiménez, B.; Moreno, R.; Stojkovic, M.; Pineda-Lucena, A. Characterisation of human embryonic stem cells conditioning media by 1H-nuclear magnetic resonance spectroscopy. PLoS One 2011, 6 (2), e16732. (34) Eriksson, L.; Johansson, E.; Wold, N.; Trygg, J.; Wikstrom, C.; Wold, S. Multi- and Megavariate Data Analysis: Advanced Applications and Method Extensions; Umetrics: Umea, Sweden, 2006. (35) Trygg, J.; Wold, S. Orthogonal projections to latent structures (O-PLS). J. Chemom. 2002, 16, 119−128. (36) Cloarec, O.; Dumas, M. E.; Trygg, J.; Craig, A.; Barton, R. H.; Lindon, J. C.; Nicholson, J. K.; Holmes, E. Evaluation of the orthogonal projection on latent structure model limitations caused by chemical shift variability and improved visualization of biomarker changes in 1H NMR spectroscopic metabonomic studies. Anal. Chem. 2005, 77 (2), 517−526.

(37) Lachenbruch, P. A. An almost unbiased method of obtaining confidence intervals for the probability of misclassification in discriminant analysis. Biometrics 1967, 23, 639−645. (38) Merrifield, C. A.; Lewis, M.; Claus, S. P.; Beckonert, O. P.; Dumas, M. E.; Duncker, S.; et al. A metabolic system-wide characterization of the pig: A model for human physiology. Mol. BioSyst. 2011, 7, 2577−2588. (39) Ranjan, P.; Gupta, A.; Kumar, S.; Nagana, G. G. A.; Ranjan, A.; Sonkar, A. A. Detection of new amino acid marker of liver injury by proton nuclear magnetic resonance spectroscopy. Liver Int. 2006, 26, 703−707. (40) Saxena, V.; Gupta, A.; Nagana Gowda, G. A.; Saxena, R.; Yachha, S. K.; Khetrapal, C. L. 1H NMR spectroscopy for the prediction of therapeutic outcome in patients with fulminant hepatic failure. NMR Biomed. 2006, 19, 521−526. (41) Graça, G.; Duarte, I. F.; Barros, A. S.; Goodfellow, B. J.; Diaz, S.; Carreira, I. M.; Couceiro, A. B.; Galhano, E.; Gil, A. M. (1)H NMR based metabonomics of human amniotic fluid for the metabolic characterization of fetus malformations. J. Proteome Res. 2009, 8 (8), 4144−4150. (42) Wang, H.; Wang, L.; Zhang, H.; Deng, P.; Chen, J.; Zhou, B.; Hu, J.; Zou, J.; Lu, W.; Xiang, P.; Wu, T.; Shao, X.; Li, Y.; Zhou, Z.; Zhao, Y. L. 1H NMR-based metabolic profiling of human rectal cancer tissue. Mol. Cancer 2013, 12 (1), 121. (43) Nicholson, J. K.; Lindon, J. C. Systems biology: metabonomics. Nature 2008, 455 (7216), 1054−1056. (44) Stoop, M. P.; Coulier, L.; Rosenling, T.; Shi, S.; Smolinska, A. M.; Buydens, L.; Ampt, K.; Stingl, C.; Dane, A.; Muilwijk, B.; Luitwieler, R. L.; Sillevis Smitt, P. A.; Hintzen, R. Q.; Bischoff, R.; Wijmenga, S. S.; Hankemeier, T.; van Gool, A. J.; Luider, T. M. Quantitative proteomics and metabolomics analysis of normal human cerebrospinal fluid samples. Mol. Cell. Proteomics 2010, 9 (9), 2063− 2075. (45) Andrew, P. J.; Mayer, B. Enzymatic function of nitric oxide synthases. Cardiovasc. Res. 1999, 43 (3), 521−531. (46) Luiking, Y. C.; Deutz, N. E. Biomarkers of arginine and lysine excess. J. Nutr. 2007, 137 (6 Suppl), 1662S−1668S. (47) Dy, M.; Schneider, E. Histamine-cytokine connection in immunity and hematopoiesis. Cytokine Growth Factor Rev. 2004, 15 (5), 393−410. (48) Newsholme, P. Why is L-glutamine metabolism important to cells of the immune system in health, postinjury, surgery or infection? J. Nutr. 2001, 131 (9 Suppl), 2515S−2522S discussion 2523S−2524S.. (49) Yaqoob, P.; Calder, P. C. Cytokine production by human peripheral blood mononuclear cells: differential senstivity to glutamine availability. Cytokine 1998, 10 (10), 790−794. (50) van der Sluis, M.; Schaart, M. W.; de Koning, B. A.; Schierbeek, H.; Velcich, A.; Renes, I. B.; van Goudoever, J. B. Threonine metabolism in the intestine of mice: loss of mucin 2 induces the threonine catabolic pathway. J. Pediatr. Gastroenterol. Nutr. 2009, 49 (1), 99−107. (51) Korkmazer, E.; Ustunyurt, E.; Tekin, B.; Cilingir, O. Study on potential role of apolipoprotein E in recurrent pregnancy loss. Exp. Ther. Med. 2013, 5 (5), 1408−1410. (52) Matthews, D. E. An overview of phenylalanine and tyrosine kinetics in humans. J. Nutr. 2007, 137 (6), 1549S−1555S.

3106


1H NMR Serum Metabonomics for Understanding ... - ACS Publications

Recommend Documents