Targeted Metabolic Profiling of Post-Mortem Brain from Infants Who

Subscriber access provided by UNIV LAVAL

Article

Targeted Metabolic Profiling of Post-Mortem Brain from Infants Who Died from Sudden Infant Death Syndrome Stewart Francis Graham, Onur Turkoglu, Praveen Kumar, Ali Yilmaz, Trent Corwin Bjorndahl, BeomSoo Han, Rupasri Mandal, David S. Wishart, and Ray O Bahado-Singh J. Proteome Res., Just Accepted Manuscript • Publication Date (Web): 13 Jun 2017 Downloaded from http://pubs.acs.org on June 14, 2017

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 free 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 accessible to all readers and 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.

Journal of Proteome Research 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 36

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

Targeted Metabolic Profiling of Post-Mortem Brain from Infants Who Died from Sudden Infant Death Syndrome Stewart F. Graham1*, Onur Turkoglu1, Praveen Kumar1, Ali Yilmaz1, Trent C. Bjorndahl2, BeomSoo Han2, Rupasri Mandal2, David S. Wishart2 and Ray O. Bahado-Singh1 1.

Beaumont Health, 3811 W. 13 Mile Road, Royal Oak, MI 48073.

2.

Department of Biological and Computing Sciences, University of Alberta, Edmonton,

AB, Canada *Corresponding Author: Email: [email protected]; Phone: +1248-551-2038; Fax: +1248-551-2947

1

ACS Paragon Plus Environment


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

Page 2 of 36

ABSTRACT

Currently little is known about the underlying pathophysiology associated with SIDS and no objective biomarkers exist for the accurate identification of those at greatest risk of dying from SIDS. Using targeted metabolomics, we aim to profile the medulla oblongata of infants who have died from SIDS (n=16) and directly compare their biochemical profile with age matched controls.

Combining data acquired using 1H NMR and targeted DI-LC-MS/MS we have

identified fatty acid oxidation as a pivotal biochemical pathway perturbed in the brains of those infants who have from SIDS (p = 0.0016). Further we have identified a potential central biomarker with an AUC (95% CI) = 0.933 (0.845-1.000) having high sensitivity (0.933) and specificity (0.875) values for discriminating between control and SIDS brains. This is the first reported study to use targeted metabolomics for the study of PM brain from infants who have died from SIDS.

We have identified pathways associated with the disease and central

biomarkers for early screening/diagnosis.

KEYWORDS: Sudden infant death syndrome; targeted metabolomics; biomarkers; post-mortem brain

2


Page 3 of 36

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


Introduction Sudden infant death syndrome (SIDS) is defined as the sudden, unexpected death of a child younger than one year of age. Further, it cannot be explained following a post-mortem investigation, including a complete autopsy, examination of the death scene, and review of the clinical history 1. SIDS is the most common cause of infant mortality between one month and one year of age in the United States 2. Numerous risk factors have been identified for SIDS including young maternal age, maternal smoking, absence of prenatal care, prone sleeping position, low birth weight, short periods between pregnancies, multiple pregnancy, drug intake by pregnant woman, and black and American Indian/Alaskan native race 3-4. The exact cause of death in SIDS remains unclear; however, a triple risk model has been proposed which suggests that it occurs in children with an underlying vulnerability (e.g. brainstem abnormality) who experience a trigger event (e.g., airflow obstruction, maternal smoking), at a vulnerable stage during the development of their central nervous or immune system 5. This risk model has led research efforts to focus on interactions between environmental factors and genetic polymorphisms that may eventually increase the susceptibility to SIDS in critical situations

6-7

.

However, due to overall low rate of SIDS in siblings and lack of concordance in twins, the role of genetic factors that play a significant role in susceptibility to SIDS remains unclear

8-9

.

Further, there remains a paucity of predictive genetic markers that can accurately classify those infants at greatest risk of dying as a result of SIDS. Metabolomics, an emerging member of “omics” family, globally investigates metabolic pathways in biological systems with the focus on metabolites

10

. It involves high-throughput

characterization and interpretation of the small molecule metabolites that are produced by cells, 3



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

Page 4 of 36

tissues and microorganisms 11. Two dominant metabolomic platforms include Nuclear Magnetic Resonance (1H NMR) and Mass Spectrometry (MS). Due to the vast chemical and physical diversity of metabolites and the complexity of metabolome, no single analytic platform can measure the concentrations of all metabolites. Hence, a combination of 1H NMR and targeted mass spectrometry coupled with liquid chromatography (DI-LC-MS/MS) can provide a more comprehensive understanding of the metabolome of infants who die of SIDS

12

. Additionally,

large-scale quantification of metabolites, in a given biological system, enables researchers to identify an increased number of significantly perturbed biochemical pathways, their interconnectivity with other compounds (proteins, lipids, genes etc.) and environmental perturbations 13. Current efforts aimed at identifying those infants susceptible to SIDS have led to emerging evidence that delayed maturation and dysregulation of the neuro- and cardiorespiratory control in the brainstem maybe the cause of death

14

. The most consistent findings suggest

abnormalities of serotonergic neurotransmission in the medulla and gliosis in caudal brain stem 15-17

. Additional reports on post-mortem (PM) analysis of medulla also suggest cytoarchitectural

anomalies in the arcuate nucleus, including reduced neuronal volume and maturity

18

and

reduced density of neurones in the inferior olivary nucleus 19. Despite these valuable efforts and convincing evidence of brainstem abnormalities in PM infants that died from SIDS, the exact biologic mechanism of pathogenesis and the underlying cause of alterations in neurotransmission are poorly resolved 15. Previously, we employed an untargeted metabolomics platform (high resolution MS), to biochemically profile PM brain tissue (medulla oblongata) from infants who died from SIDS. 4


Page 5 of 36

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


Untargeted metabolomic profiles of brain samples were capable of distinguishing SIDS cases from controls with high predictive power while identifying several features that may be involved in SIDS pathogenesis

20

. However, while being the most powerful technique available to

metabolomists, the greatest drawback of all untargeted analyses is that accurate identification of all the features of interest is not currently possible. Given this limitation, and the importance of identifying the metabolic pathways disrupted as a direct result of SIDS, we present a targeted metabolomics approach. This study describes the combination of data acquired using 1H NMR and targeted MS from the medulla oblongata of infants who died from SIDS. To our knowledge, this is the first targeted metabolomics study for the investigation of SIDS.

5



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

Page 6 of 36

Experimental Procedures Tissue Samples Only a limited number of specimens and tissue volume was available for this pilot study. In brief, tissue samples (medulla) were obtained from postmortem SIDS cases with no co-sleeping (n = 16) and age- and gender-matched control subjects (n = 7) as previously described by Graham et. al., (2017)

20

. This study was approved by the Beaumont Health System’s Human

Investigation Committee (HIC no.: 2014-210). The methods were carried out in accordance with the approved guidelines. Details such as age, gender, race and postmortem delay can be found in Table S1. Samples were prepared as previously described 21-22. Combined Direct Flow Injection and LC-MS/MS compound identification and quantification We have applied a targeted quantitative metabolomics approach to analyze the brain samples using a combination of direct injection mass spectrometry (AbsoluteIDQ™ Kit) with a reversephase LC-MS/MS Kit (AbsoluteIDQ p180, Biocrates, Innsbruck, Austria), as previously described 22-23. In combination with an API 4000 Q-Trap (Applied Biosystems/MDS Sciex) mass spectrometer, the kit was used to accurately identify and quantify up to 180 different endogenous metabolites to include: amino acids, acylcarnitines, biogenic amines, glycerophospholipids, sphingolipids and sugars. Brain specimens were extracted as described by Urban et al., (2010) 24 and Pan et al., (2016)

22

. Briefly, brain samples were lyophilized and milled to a fine powder.

10 mg of powdered PM brain tissue was extracted in 300 µl of solvent (85 % ethanol and 15 % phosphate-buffered saline solution).

The samples were sonicated (5 minutes), vortexed (1 6


Page 7 of 36

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


minute), and centrifuged (13,000 X g for 15 minutes) at 4ºC. Mass spectrometric analysis was performed on an API 4000 Qtrap® tandem mass spectrometry instrument (Applied Biosystems/MDS Analytical Technologies, Foster City, CA) equipped with a solvent delivery system per the manufacturer’s instructions. The Biocrates MetIDQ software was used to control the entire assay workflow, from sample registration to automated calculation of metabolite concentrations and export of data into other data analysis programs. 1

H NMR Analysis

All 1H NMR spectra were acquired as previously described by Ravanbakhsh et al., (2015) 25 and Graham et al., (2016)

21

. Briefly, 50 mg of milled PM brain were extracted using 50 %

MeOH:H2O as previously described by Graham et al., (2016)

21

. All spectra were acquired at

300.0 K on a Bruker Avance III HD 600 MHz spectrometer (Bruker-Biospin, MA, USA) equipped with a 5 mm cryo-probe. Two hundred and fifty-six transients were acquired and chemical shifts (δ) are reported in parts per million (ppm). The singlet at 0.00 ppm produced by the methyl groups of the internal standard 4,4-dimethyl-4-silapentane-1-sulfonic acid (DSS d6) was used for spectral referencing and quantification. All spectra were processed and analyzed using Chenomx NMR Suite (v8.0, Chenomx, Canada). Statistical Analysis Standard metabolomics statistical analyses were utilized 26-27. Metabolite concentrations in SIDS and controls were compared using Wilcoxon Mann-Whitney test. The false discovery rate (FDR, q-values) was also calculated to take into consideration the corrections needed for multiple comparisons. Generalized log-transformation and Pareto-scaling were performed to normalize the metabolite concentration data for multivariate analysis. Principal component analysis (PCA) 7



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

Page 8 of 36

and Partial Least Squares Discriminant Analysis (PLS-DA) were performed to identify distinct metabolite patterns. To assess the significance of the separations achieved by PLS-DA, permutation testing was performed (using 2,000 resampling) to determine the corresponding pvalues. Permutation testing allows one to determine the likelihood that the observed separation between cases and controls on a PLS-DA plot is due to chance. In addition, Variable Importance in Projection (VIP) plots were generated from the PLS-DA data. The VIP plot ranks metabolites based on their importance in discriminating SIDS from the control group. The greater the VIP score the more discriminating the metabolite. Logistic regression analysis was utilized to generate the optimal predictive models for SIDS classification. Two models were developed, one was based on metabolites only, and the other was based on metabolites with age (days). Generalized log-transformed and metabolites with a pvalue < 0.3 (using univariate analysis) were selected for developing logistic regression models. We employed a k-fold cross-validation (CV) technique to ensure that the logistic regression models were robust

26

. Stepwise variable selection was utilized to optimize the SIDS prediction

model components via 10-fold cross-validation. Areas under the Receiver Operating Characteristic curve (AUROC or AUC)

26

along with sensitivity and specificity values were

calculated for all generated models. MetaboAnalyst was used to perform PCA and PLS-DA analyses

28

. Custom programs written using the R statistical software (http://www.r-project.org)

was used to perform all other statistical analyses. Pathway Analysis All metabolites were uploaded to Metaboanalyst (v3.0) and analyzed using the Pathway Topology tool. All data were log transformed and Pareto scaled. The pathway library chosen 8


Page 9 of 36

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


was Homo Sapiens and all compounds in the selected pathways were used when referencing the specific metabolome. Fisher's exact test was applied to perform overrepresentation analysis and “relative betweenness centrality” was chosen for the pathway topology testing. Pathways that had both a Holm adjusted p-value < 0.05 and FDR p-value < 0.05 were considered to be altered due to SIDS 21.

9



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

Page 10 of 36

Results In total, 199 metabolites were accurately identified and quantified in the medulla of infants who died from SIDS. 156 metabolites were identified and quantified using the DI-LC-MS/MS of which 19 were found to statistically significantly (p

Targeted Metabolic Profiling of Post-Mortem Brain from Infants Who

Recommend Documents