NMR-Based Serum Metabolomics of Patients with ... - ACS Publications

Subscriber access provided by University of Sussex Library

Article

NMR-based Serum Metabolomics of patients with Takayasu arteritis (TA) - Relationship with disease activity Avinash Jain, Dinesh Kumar, Anupam Guleria, Durga Prasanna Misra, Abhishek Zanwar, Smriti Chaurasia, Sandeep Kumar, Umesh Kumar, Shravan K. Mishra, Ruchika Goel, Debashish Danda, and Ramnath Misra J. Proteome Res., Just Accepted Manuscript • DOI: 10.1021/acs.jproteome.8b00456 • Publication Date (Web): 10 Aug 2018 Downloaded from http://pubs.acs.org on August 10, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

NMR-based Serum Metabolomics of patients with Takayasu arteritis (TA) Relationship with disease activity Avinash Jain*, Dinesh Kumar+, Anupam Guleria +, Durga Prasanna Misra *, Abhishek Zanwar, Smriti Chaurasia , Sandeep Kumar *, Umesh Kumar+, Shravan K Mishra*, Ruchika Goel **, Debashish Danda**, Ramnath Misra* Department of Clinical Immunology, S.G.P.G.I.M.S, Lucknow, India, +Centre of Biomedical Research, Lucknow India and ** Christian Medical College, Vellore, India

Correspondence: Dr Ramnath Misra, Professor and Head Department of Clinical Immunology, Sanjay Gandhi Postgraduate Institute of Medical Sciences (SGPGIMS), Rae Bareli Road, Lucknow, Uttar Pradesh, India PIN 226014 E-mail: [email protected] Phone +91 522 2494284 Dr. Dinesh Kumar Assistant Professor Centre of Biomedical Research (CBMR), SGPGIMS Campus, Raibareli Road, Lucknow-226014 Uttar Pradesh, India Mobile: +91-8005409932, +91-8953261506 Email: [email protected]

1 ACS Paragon Plus Environment

Journal of Proteome Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ABSTRACT

Takayasu arteritis (TA) is a large vessel vasculitis of unknown pathogenesis. Assessment of disease activity is a challenge and there is an unmet need for relevant biomarker(s). In our previous study, NMR based serum metabolomics had revealed distinctive metabolic signatures in TA patients compared with age/sex matched healthy controls and Systemic Lupus Erythematosus (SLE). In this study we investigate whether the metabolites correlate with disease activity. Patients with TA fulfilling American College of Rheumatology (ACR) criteria were enrolled and disease activity was assessed using Indian Takayasu Clinical Activity Score using Acute Phase Reactant – Erythrocyte Sedimentation Rate [ITAS-A (ESR]. Sera were analysed using 800 MHz NMR spectrometer to identify metabolites [based on Partial Least Squares Discriminant Analysis (PLS-DA) VIP (variable importance in projection) score >1.0 and permutation test, p-value < 0.01]. 45 active and 53 inactive TA patients with median age 27 [(IQR) 22-35 years] and 27 [(IQR) 23-37 years] female to male ratio 3.5:1 and 4.9:1 and median duration of illness 5 [(IQR) 2-9 years] and 3 [(IQR) 1-6 years] years respectively were enrolled. The key metabolites with highest discriminatory potential in active TA (ITAS-A ≥4) were glutamate and N-acetyl glycoprotein (NAG), both elevated, with area under the curve 0.775 and 0.769 (p-value 1)

Figure 2: Correlation of disease activity and metabolic signature (A and B) Principal Component Analysis [PCA] and Orthogonal Partial Least Square – Discriminant Analysis [OPLS-DA] revealed separate clustering of active TA (ITAS-A using ESR ≥ 4) from inactive TA patients (ITAS-A using ESR ≤3). (C) After procurement of this plot, VIP scores were calculated to find out metabolic parameters of significance. Glutamate, NAG, glucose, leucine, isoleucine, alanine, phosphoglyceride were found to be increased in active patients and LDL/VLDL, lactate, PUFA, choline, Glycerophosphocholine [GPC], acetone and glutamine were decreased in active TA patients.



Using univariate analysis, box and whisker plots were structured underscoring the differences in metabolites between active and inactive samples (Fig. 3). These metabolites were further scrutinised using AUROC analysis. Glutamate, NAG, LDL, glucose, leucine, and proline could serve as potential biomarker in distinguishing the two groups. The area under the ROC curve gives an accurate assessment of discriminatory ability (0.5 = no discrimination; 1 = perfect discrimination)

Figure 3 : Area under the receiver operating characteristic curve (AUROC) calculated for selected differential biomarkers in active and inactive TA patients: Representative box-cum-whisker plots and ROC curve showing quantitative variations of relative signal integrals for serum metabolites in Takayasu Arteritis comparing active and inactive patients. For presented metabolite entities, the VIP score >1 and statistical significance was defined as p ≤ 0.05. The centre line refers to median whereas the lines above and below refer to 25th and 75th percentiles. Lower and upper whiskers are 5th and 95th percentiles, respectively. AUROC plot for glutamate and NAG revealed an area of 0.775 and 0.769 with 95% confidence interval of 0.685 to 0.859 and 0.675 to 0.847 respectively. At 2.03 ppm contribution is seen from GP (Glutamate and Proline) improving the AUC to 0.816 [0.721-0.891]. Similarly, glucose, glycerol, phosphoglyceride [PG] and phenylalanine were found to be raised and LDL decreased in active TA patients as shown in Fig. (D-H) respectively. Dotted lines in the box and whisker plot in all figures and arrow represented in Fig. (A and B) defines the cut off at which active and inactive TA patient can be classified with varying sensitivity and specificity.


Page 14 of 26

Page 15 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Glutamate and NAG emerged as an important biomarkers showing a good correlation with disease activity as well (Fig. 3). The strength of correlation of ITAS-A using ESR was maximum with glutamate followed by NAG level. This can be further studied and substantiated using ELISA kits which could be more practical. Using combination of different metabolites, the difference between active and inactive got further accentuated with significant contribution from sum total of Glutamate and proline at 2.03 ppm (p < 0.001) with AUC 0.816 (0.721 to 0.891). Among the other metabolites contributing significantly in separation were glucose, glycerol, phosphoglyceride, phenylalanine which were elevated with AUC 0.76 [0.683- 0.851], 0.746 [0.645-0.85], 0.743 (0.643-0.834), 0.72 [0.611-0.816] respectively and LDL which was decreased in active TA patients with AUC 0.710 [0.6160.807]. Sensitivity and specificity of various metabolites was also calculated using cut off points obtained from top left corner of AUROC It is the point which is farthest vertical distance from an imaginary diagonal line connecting the starting and end point of ROC curve.8 Table S2 highlights the cut off and sensitivity, specificity and p value of these metabolites. A combination of various metabolites of significance was also assessed using Metaboanalyst 3.0 to see if there was improvement in ROC plot. Combining the six metabolites which were increased improved the AUROC to 0.913 [0.837-0.971] which was marginally better than AUROC of glutamate and NAG when assessed separately. This was based on linear Support vector machine model, a multivariate data analysis method [Table S3] Three groups namely active TA, inactive TA and healthy control were also compared by constructing OPLS DA plot (Fig. S3) to assess whether the pattern of metabolites shifts



towards healthy control when disease is inactive. This might help in shedding more light on the pathogenesis of disease and might serve as a potential to monitor therapy and develop newer therapeutic potentials. OPLS –DA plot could differentiate the three groups with good accuracy and significant R2 of 0.79 underlining the fitness of model. Visual impression of the graph also depicted the shift in metabolic profile of the inactive TA patients towards healthy control away from active TA patients. Longitudinal assessment of disease activity and its relationship with metabolic profile 57 patients were monitored with a median follow up duration of 3.1 months with reassessment of disease activity and metabolic profile for evaluating the trend and variation in the profile with disease activity on serial assessments. 32 out of these 57 patients were defined as active on basis of ITAS-A using ESR and rest 25 were inactive. 28 out of 32 active TA patients had become inactive on follow up and OPLS DA 2D score plot revealed separation of patients indicating change in the metabolic signature pattern. (Fig. 4A). Though the accuracy was 0.66 but r2 of 0.34 was not satisfactory. 23/25 inactive patients who remained inactive clustered together in OPLS DA plot with R2 value of 0.12 and showed no differences in the metabolic pattern (Fig 4B). There was no significant separation when other groups were analysed (Table S4).

B

A


Page 16 of 26

Page 17 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4 (A) shows OPLS-DA plot for patients who were active on visit 1 and became inactive on visit 2. There was definite separation but discriminatory potential was not satisfactory. (B) Similarly 23 patients who were inactive in both visits did not show any significant separation on OPLS-DA plot

Effect of Drug Therapy To evaluate the metabolic effect of immunosuppressive agents, a PLS-DA score plot was generated from 1H NMR spectra of serum samples obtained from the inactive TA patients with and without medications. 37 patients were on immunosuppression as against 14 patients who were not taking any immunosuppressive agents. (2 patients’ details were not available) There was no significant impact of medications on the PLS DA score plot (Fig. S4) irrespective of disease activity status. Discussion Takayasu arteritis (TA), a systemic vasculitis, like any other inflammatory disorder is accompanied by systemic alterations in circulating metabolites. Similar to previous study11 we used quantitative NMR and a multivariate PLS-DA statistical method to assess the changes in serum metabolites simultaneously to build metabolite profiles in the disease and to assess their variation with disease activity. This data links the so called “metabolic signature” to clinical status of patients for the first time. It may help in establishing a clinical relevance for this field which will be pursued in more detail in ongoing studies. The inclusion of samples from a new patient cohort allowed confirmation of the distinct metabolic profile of TA seen in our preliminary study11 disclosing similar results with increased N-Acetyl Glycoprotein (NAG), glucose, glutamate, phosphoglyceride, glycerol, glycerophophocholine and decreased glucogenic amino acids, lactate and creatine.

The

combined samples from the two series enhanced the power to discriminate the TA profile from both normal and disease controls. Increased Glucose, Glycerol, PUFA with decrease 17 ACS Paragon Plus Environment


Page 18 of 26

lactate could imply dampened glycolysis and increased utilization of other metabolites like stored fat as an energy source. Whether this decrease in lactate is an end result in an attempt to halt widespread angiogenesis in a setting of hypoxia is unclear, as lactate is known to stabilize hypoxia inducible factor-1α (HIF-1α) and promote angiogenesis.18 This is in contrast to Warburg effect expected in hypoxic environment

19

Higher serum LDL/VLDL in TA

patients is also consistent with previously reported dyslipidemia in TA patients

20

NAG, an

anti-inflammatory metabolite , too has been found to be increased in TA patients , similar to other acute and chronic inflammatory states. 21,22 Lower amino acids like arginine, glutamine, proline, alanine designates increase in inflammatory and autoimmune responses. However lower level of serum histidine, an anti-oxidant, seems to be an effect of increased inflammation and oxidative stress in TA as reported previously 11,23 We had previously also compared TA with SLE as disease control which allows us to state that autoimmune inflammation does not produce a uniform pattern across diseases, confirming the specific disease-related profile seen in TA11. This data is yet to be published (Fig S5). SLE was chosen as a control as gender and age distribution was similar and is treated with similar drugs. The lack of any major difference between the TA samples studied when the patients were not on any therapy confirms this. Future studies of therapy-naive patients followed by serial assessments while on therapy will be important to establish the contribution of drug therapy to the observed TA profile. The relationship of the TA metabolic profile to disease activity indicated by the relationship of the activity score to the metabolic profile in this study is a very important area. Here we used the validated index ITAS-A to score disease activity and compared active TA to inactive group. Glutamate, NAG, glucose, glycerol, phosphoglyceride, phenylalanine were found to be significantly elevated and LDL/VLDL significantly decreased in active TA patients (p

NMR-Based Serum Metabolomics of Patients with ... - ACS Publications

Recommend Documents