Metabolomics Reveals Anaerobic Bacterial Fermentation and

Subscriber access provided by University of Glasgow Library

Article

Metabolomics Reveals Anaerobic Bacterial Fermentation and Hypoxanthine Accumulation for Fibrinous Pleural Effusions in Children with Pneumonia Chih-Yung Chiu, Mei-Ling Cheng, Kin-Sun Wong, ShenHao Lai, Meng Han Chiang, Ming-Han Tsai, and Gigin Lin J. Proteome Res., Just Accepted Manuscript • DOI: 10.1021/acs.jproteome.8b00864 • Publication Date (Web): 13 Feb 2019 Downloaded from http://pubs.acs.org on February 14, 2019

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 28 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

Metabolomics Reveals Anaerobic Bacterial Fermentation and Hypoxanthine Accumulation for Fibrinous Pleural Effusions in Children with Pneumonia Chih-Yung Chiu,†, ‡, * Mei-Ling Cheng,§ Kin-Sun Wong,† Shen-Hao Lai,† Meng-Han Chiang,∥Ming-Han Tsai, ‡ and Gigin Lin*,∥ Author Affiliations: †Division of Pediatric Pulmonology, Chang Gung Memorial Hospital at Linkou, College of Medicine, Chang Gung University, Taoyuan, Taiwan ‡Department of Pediatrics, Chang Gung Memorial Hospital at Keelung, College of Medicine, Chang Gung University, Taoyuan, Taiwan §Department of Medical Biotechnology and Laboratory Science, and Healthy Aging Research Center, Chang Gung University, Taoyuan, Taiwan ∥ Department of Medical Imaging and Intervention, Imaging Core Laboratory, Institute for Radiological Research, and Clinical Metabolomics Core Laboratory, Chang Gung Memorial Hospital at Linkou, College of Medicine, Chang Gung University, Taoyuan, Taiwan

*Corresponding Authors: Chih-Yung Chiu, E-mail: [email protected]. Tel: +886-3-3281200 ext 8966; Fax: 886-2-24313161 & Gigin Lin, E-mail: [email protected]. Tel: +886-3-3281200 ext 2575; Fax: 886-3-3288957

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

Page 2 of 28

ABSTRACT Fibrin formation in infectious parapneumonic effusion (IPE) characterizes complicated parapneumonic effusion and is important for providing guidelines for management of IPEs that require aggressive interventions. We aim to identify metabolic mechanisms associated with bacterial invasion, inflammatory cytokines, and biochemical markers in cases of fibrinous infectious pleural effusions in children with pneumonia. Pleural fluid metabolites were determined by 1H-nuclear magnetic resonance (NMR) spectroscopy. Metabolites contributed for the separation between fibrinous and non-fibrinous IPEs were identified using supervised partial least squares-discriminant analysis (Q2/R2 = 0.84; Ppermutation < 0.01). IL-1β in the inflammatory cytokines, and glucose in the biochemical markers, were significantly correlated with 11 and 9 pleural fluid metabolites, respectively, and exhibited significant

overlaps.

Four

metabolites,

including

glucose,

lactic

acid,

3-hydroxybutyric acid, and hypoxanthine, were significantly correlated with plasminogen activator inhibitor type 1 (PAI-1) in the fibrinolytic system enzymes. Metabolic pathway analysis revealed that anaerobic bacterial fermentation with increased lactic acid and butyric acid via glucose consumption, and adenosine triphosphate (ATP) hydrolysis with increased hypoxanthine appeared to be associated with fibrinous IPE. Our results demonstrate that an increase in lactic acid anaerobic 2 ACS Paragon Plus Environment

Page 3 of 28 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


fermentation and hypoxanthine accumulation under hypoxic conditions are associated with fibrin formation in IPE, representing advanced pleural inflammatory progress in children with pneumonia.

KEYWORDS: childhood pneumonia, fibrin, infectious parapneumonic effusion, metabolomics, nuclear magnetic resonance



INTRODUCTION Pneumonia is a fairly common and potentially serious infection in children. Infectious parapneumonic effusion (IPE) accumulation appears to be due to the imbalance of dynamic drainage of liquid in the pleural cavity resulting from the inflammation in response to pneumonia.1 Pleural inflammation is a continuous process with pleural fluid progression from the exudative effusion to empyema. Failing to adequately control the inflammation, pleural effusion is usually accompanied by fibrin deposition and loculations, leading to the accumulation of pleural fluid, thus requiring aggressive surgical intervention.2 Understanding the complexity of fibrin formation in pleural inflammation may provide guidance for prevention and treatment of IPE in children with pneumonia. Tissue-type plasminogen activators (tPAs) and plasminogen activator inhibitors (PAIs) are key regulators of the fibrinolytic system. The levels of these enzymes have been demonstrated to increase substantially in response to inflammation or infection.3 Increased release of inflammatory cytokines, e.g. interleukin-1β (IL-1β) and tumor necrosis factor-α (TNF-α), in advanced pleural infection is associated with reduced fibrinolytic activity, resulting in fibrin deposition.4 At the same time, an increase in glucose metabolism along with acidosis occurs and requires aggressive pleural drainage.5 However, the complex metabolic networks involved in fibrinous IPE 4 ACS Paragon Plus Environment

Page 4 of 28

Page 5 of 28 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


during pleural inflammatory progress related to pneumonia have not been adequately investigated. For appropriate antibiotic-based treatment of pleural infection, identification and understanding of the responsible pathogens are considered essential. Due to the differences in pathogenic bacteria responsible for pleural cavity infection, various pathogens may influence the course of pleural inflammation and clinical outcomes. Despite this observation, in children with a fibrinous IPE, an early surgical intervention is recommended to re-expand and improve chest dynamics.6 A comprehensive metabolomics-based approach to fibrinous IPE associated with different pathogens may provide insights into the pleural pathogenesis following pneumonia and reveal metabolites that are potentially important for clinical applications in IPE. The purpose of this study was, therefore, to determine the metabolic profiles of IPE using 1H-NMR spectroscopy in children with pneumonia. The metabolic changes in pleural fluid, varying with different pathogens and fibrin depositions were assessed. Their relationships with inflammatory cytokines (TNF- and IL-1), biochemical markers (glucose, LDH, and pH), and fibrinolytic system enzymes (tPA and PAI-1) for fibrin formation were also examined.



Page 6 of 28

MATERIALS AND METHODS Study population Children under 18 years of age who had community-acquired pneumonia complicated by parapneumonic effusions requiring hospitalization were consecutively enrolled in this study. The study population comprised participants with parapneumonic effusions who

underwent

real-time

chest

sonography

performed

by

two

pediatric

pulmonologists and whose sonograms were concurrently interpreted and confirmed by a board-certified radiologist. The selected patients were categorized into two groups based on sonographic findings of pleural fluid:7 (i) non-fibrinous IPE groups defined by simple effusion with a clear anechoic fluid with or without swirling particles and (ii) fibrinous IPE group defined by the presence of septations/loculations or fibrin strands. Pleural effusion or blood was cultured for the identification of pathogens. Streptococcus pneumoniae (S. pneumoniae) pneumonia was defined by a positive latex agglutination test of pleural fluid or a positive S. pneumoniae urine test (Binax, Portland, ME, USA) and necrotic lung parenchyma.8 The criteria of Mycoplasma pneumoniae (M. pneumoniae) pneumonia was a four-fold increase in antibody titers in the absence of other pathogenic bacteria in cultures. For treatment, empirical antibiotics were selected for possible organisms and then adjusted after obtaining 6 ACS Paragon Plus Environment

Page 7 of 28 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


culture results. Additional interventions, such as tube thoracostomy or video-assisted thoracoscopic surgery for pleural fluid drainage were performed for cases with persistent fever above 39°C, respiratory distress, or sepsis which did not improve after antibiotic treatment.9 Data on patient demographics, pleural fluid characteristics, and the organisms identified from blood or pleural effusion were recorded and analyzed. All procedures were approved by the Ethic Committee of Chang Gung Memory Hospital (No. 93-6299). The parents/guardians of the study subjects provided written informed consent prior to participation. Pleural fluid analysis After collection by thoracentesis, pleural fluid samples were immediately analyzed for total cell count, differential cell count, pH, and glucose and lactate dehydrogenase (LDH) concentrations. Another four mL pleural fluid sample was collected in a Greiner Bio-One VACUETTE coagulation tube containing 3.2% sodium citrate anticoagulant solution and immediately centrifuged at 1,500 g for 10 min at 4°C. The supernatant obtained for each sample was then stored at -80°C until further analysis. Commercially available enzyme-linked immunosorbent assay (ELISA) kits were used to assess the pleural inflammatory mediators TNF- and IL-1 (Quantikine; R&D Systems, Minneapolis, MN), and the fibrinolytic system enzymes tPA and PAI-1 (American Diagnostica, Pfungstadt, Germany). All assays were performed in 7 ACS Paragon Plus Environment


accordance with the respective manufacturer’s instructions.4 Pleural fluid sample preparation The preparation of pleural fluid samples has been previously reported.5 In brief, 500 μL pleural effusion was mixed with 250 μL phosphate buffer (0.075 M Na2HPO4, pH 7.4) in deuterium water containing 0.08% TSP [3-(trimethylsilyl)-propionic-2,2,3,3-d4 acid sodium salt] as an internal chemical shift reference standard. Next, the samples were vortexed for 20 s and centrifuged at 12,000 g for 30 min at 4°C. Lastly, 600 μL supernatant was transferred to a 5-mm NMR tube for analysis. 1H–nuclear

1H-NMR

magnetic resonance (NMR) spectroscopy

spectroscopy was performed using a Bruker Avance 600 MHz spectrometer

(Bruker-Biospin GmbH, Karlsruhe, Germany) located at the Chang Gung Healthy Aging Research Center, Taiwan. In total, 64 scans were collected for NMR spectra into 64 K computer data points with a spectral width of 10,000 Hz (10 ppm), relaxation delay of 4.0 s and acquisition time of 3.07 s. Prior to zero-filled Fourier transformation at an exponential line broadening of 0.3 Hz, 1D spectra were applied. Then, the resulting 1H-NMR spectra were manually phased and baseline-corrected. TopSpin 3.2 software (Bruker BioSpin, Rheinstetten, Germany) was further used to reference the chemical shift of TSP (δ 0.0 ppm). NMR data processing and analysis 8 ACS Paragon Plus Environment

Page 8 of 28

Page 9 of 28 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


Raw 1H-NMR spectra were bucketed using AMIX software (Version 3.9.12; Bruker BioSpin GmbH, Rheinstetten, Germany). After spectral region exclusion and TSP spectral normalization, 1H-NMR spectra were subdivided into 0.01 ppm integrated regions corresponding to the region δ 0-10 ppm. To avoid spectral interference, all regions containing residual water (δ 4.825 - 4.725 ppm) were excluded. Further, the spectra were normalized to the integral of the citric acid peak at δ 2.575-2.525 ppm to overcome the variation in pleural fluid volume during sample collection as previously described.5 The Chenomx NMR Suite 8.1 software (Chenomx Inc., Edmonton AB, Canada) using reference spectra from the Human Metabolome Database was used for metabolite identification. Consistent with established NMR data analysis methods,5 normalized 1H-NMR bucket data underwent generalized log transformation (glog) were then uploaded to MetaboAnalyst 4.0 (http://www.metaboanalyst.ca) for the identification of the metabolites used for discrimination between the groups using partial least squares-discriminant analysis (PLS-DA). Spectral variables were mean-centred and scaled using Pareto scaling. To evaluate the resulting statistical models, 10-fold internal cross-validation using the diagnostic measures R2 and Q2 was performed.10 The potential metabolites with a variable importance in projection (VIP) score ≥ 1.0 or P-value < 0.05 were selected. The functional pathway analysis of these potential 9 ACS Paragon Plus Environment


metabolites was based on the Kyoto Encyclopedia of Genes and Genomes database (http://www.genome.jp/kegg/). Statistical analysis The baseline characteristics and pleural variables of children with non-fibrinous or fibrinous IPE were compared using univariate parametric and non-parametric tests as appropriate, including Student’s t-test, Mann-Whitney test, 2, and Fisher’s exact test. Changes in metabolites between the two groups were assessed using non-parametric Mann-Whitney test with the MetaboAnalyst web server. Hierarchical clustering was performed using the Ward clustering algorithm. Relationships between metabolites and inflammatory cytokines, biochemical markers and fibrinolytic system enzymes were assessed using Spearman’s rank correlation coefficient. All tests were two-tailed, and a P-value < 0.05 was considered statistically significant. Analysis was performed using Statistical Program for Social Sciences software (IBM SPSS Statistics for Windows, Version 20.0; IBM, Armonk, NY, USA).


Page 10 of 28

Page 11 of 28 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


RESULTS Population characteristics Fifty-eight children with community-acquired pneumonia complicated with parapneumonic effusions were consecutively enrolled in this study. There were 32 boys and 26 girls with a median age of 4.3 years (range, 0.3-17 years). S. pneumoniae and M. pneumoniae pneumonia were identified in 38 and 6 children respectively. Non-fibrinous IPE was diagnosed in 30 (52%) children; whereas fibrinous IPE was diagnosed in 28 (48%) children based on ultrasonography. There was a significant decrease in only the pleural pH and glucose levels in children with IPE caused by streptococcal pneumonia, as compared to that caused by Mycoplasmal pneumonia (P < 0.01). In contrast, WBCs in hemogram; IL-1β in inflammatory cytokines; glucose, LDH, and pH value in biochemical markers; and PAI-1, tPA in fibrinolytic system enzymes were significantly different between children with and without fibrin formation in IPE (Table 1). A higher rate of receiving interventional treatments (79% vs. 30%, P < 0.001) and a longer hospital stay (15.5 ± 6.1 vs. 12.8 ± 5.4, P = 0.036) were found in children with fibrinous IPE than in children with non-fibrinous IPE. Age was significantly different with a wide range of distribution; however, the median age was similar for both the groups. Metabolite sets between sexes, pathogens and IPE with and without fibrin 11 ACS Paragon Plus Environment


One hundred and eighty-four of 1,000 buckets were obtained from 1H-NMR data of pleural fluid, corresponding to 26 known metabolites.5 Unsupervised principal components analysis revealed no clear separation between the groups. Metabolites contributed for the discrimination between the groups were then identified using supervised PLS-DA. The PLS-DA cross-validation and permutation test for distinguishing between sexes, S. pneumoniae and M. pneumoniae pathogens, and IPE with and without fibrin are shown in Table S1. Figure S1 shows the PLS-DA score plots from the analysis of pleural fluid 1H-NMR spectra between groups. Significant associations (Q2/R2 > 0.5) were found between the sets with S. pneumoniae and M. pneumoniae pathogens, and between fibrinous and non-fibrinous IPE (Ppermutation < 0.05). Metabolites identified and the extent of change of expression level between different pathogens, and between fibrinous and non-fibrinous IPE are shown in Table 2. A significantly lower pleural glucose level accompanied with higher lactic acid level was found in children with S. pneumoniae IPE than in M. pneumoniae IPE. Compared with non-fibrinous IPE, glucose was found to be significantly lower, whereas thymine, phenylalanine, leucine/isoleucine, tryptophan, glutamic acid, lactic acid, and hypoxanthine were significantly higher in fibrinous IPE. Metabolites associated with inflammatory cytokines, biochemical markers and fibrinolytic system enzymes 12 ACS Paragon Plus Environment

Page 12 of 28

Page 13 of 28 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 1 shows the correlations of pleural fluid metabolites with inflammatory cytokines, biochemical indices, and fibrinolytic system enzymes. IL-1β in inflammatory cytokines; glucose and pH value in biochemical markers; and PAI-1 in fibrinolytic system enzymes were significantly correlated with 11, 9, 6, and 4 pleural fluid metabolites respectively. A Venn diagram showed distribution of the metabolites that correlated with IL-1β, glucose, pH, and PAI-1 (Figure 2A). Majority of the correlated pleural fluid metabolites (82%) overlapped with IL-1β and glucose. Four common metabolites including glucose, lactic acid, 3-hydroxybutyric acid and hypoxanthine were significantly correlated with PAI-1. Metabolic pathway and functional analysis The results of the metabolic pathway analysis demonstrated that the metabolites associated with fibrin formation were important for carbohydrate metabolism (Table S2). Glucose was associated with glycolysis or gluconeogenesis; lactic acid, 3-hydroxybutyric acid, and hypoxanthine were related to propanoate and butanoate metabolism. In fibrinous IPE, glucose dissimilation proceeded through glycolysis, leading to the fermentation of lactic acid and butyric acid, and ATP hydrolysis of hypoxanthine. Metabolic pathways related to fibrin formation in infectious parapneumonic

effusions

are

shown


in

Figure

2B.


DISCUSSION Infectious parapneumonic effusion (IPE) is not uncommon in childhood pneumonia and an aggressive surgical intervention is needed in children with fibrinous septations within the pleural fluid.2 In children with pneumonia, pleural inflammation followed by fibrin deposition in pleural fluid usually occurs in advanced pneumonia and pleurisy. In this study, a significant increase in inflammatory cytokines accompanied with a decrease in pleural pH and glucose was found in fibrinous IPE. Furthermore, children with fibrinous IPE appeared to have a higher rate of advanced interventional procedures and longer hospital stays. These features support the idea that severe pleural infection may drive the production of inflammatory cytokines leading to fibrin deposition and subsequent surgical intervention. Several studies have shown that glucose consumption associated with lactate and CO2 excretion leads to a low pleural fluid pH level with progress of pleural inflammation.4,

5, 11

In this study, pleural fluid metabolites related to glucose were

mostly correlated with inflammatory cytokine IL-1β. Furthermore, glucose and lactic acid were the most useful metabolites for differentiating non-fibrinous IPE from fibrinous IPE. In addition, a shift in pH to a more acidic value appears to correlate with the polymerization of isolated fibrin monomers.12, 13 These findings indicate that glucose consumption with pleural acidosis related to advanced pleural infection plays 14 ACS Paragon Plus Environment

Page 14 of 28

Page 15 of 28 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


a crucial role in the development of fibrinous IPE. A shift of serum proteins including albumin from blood to pleural fluid is known as a feature in severely infected children with massive pleural effusion.14 A dynamic change of proteins in pleural fluid provides potential diagnostic and therapeutic values in guiding the management of IPE.15 In this study, metabolism of amino acid synthesis appeared to be significantly active in the fibrinous IPE. However, among them, glutamate is important for bacterial virulence and has been shown to be significantly higher in the plasma in severe childhood pneumonia.16, 17 Tryptophan, an essential amino acid for protein synthesis, is synthesized from molecules such as phosphoenolpyruvate in bacteria.18 In our study, very importantly, these amino acids were significantly correlated with inflammatory cytokines but not with the expression of PAI-1, a key enzyme in fibrinolysis. These findings thus indicate that amino acids associated with fibrinous IPE may mainly represent the increased invasion of bacteria rather than the formation of fibrin deposition. S. pneumoniae, a facultative anaerobe, causes most childhood bacterial parapneumonic effusions in Taiwan as seen in this study.19 In contrast to M. pneumoniae, S. pneumoniae is more capable of switching to anaerobic fermentation in the absence of oxygen.20 Both propionic and butyric acid are the early products observed as a result of bacterial anaerobic respiration.21, 22 In this study, the levels of 15 ACS Paragon Plus Environment


lactic acid (related to propanoate metabolism) and 3-hydroxybutyric acid (related to butanoate metabolism) were not only significantly correlated with the pleural pH value but also PAI-1 levels contributing to fibrin formation. These observations emphasize the importance of microorganisms undergoing anaerobic bacterial fermentation in the absence of oxygen for the formation of fibrin in IPE. Hypoxia is a common insult in children with advanced pneumonia. The degree of lung perfusion impairment is strongly positively correlated with the severity of lung necrosis and pneumonia.23 Uncontrolled infection or sepsis induces substantial alterations in the homeostasis of energy metabolism. Hypoxanthine is a breakdown product of ATP that maintains a high energy charge.24 An increased plasma level of hypoxanthine has been reported in bacterial meningitis, septic shock and severe childhood pneumonia.17, 25 In this study, there was a significantly positive correlation of hypoxanthine level in pleural fluid with pleural inflammatory status as well as the activity of anti-fibrinolysis, indicating that the pulmonary ischemia and hypoxia resulting from severe pneumonia most likely participates in fibrin formation in IPE. A relatively low sensitivity due to low concentration levels in the samples for NMR for estimating biological matrices and the small sample size are the major limitations of this study. A wide range in the age distribution of children with pneumonia could result in an increase in the variability of metabolic profiles. 16 ACS Paragon Plus Environment

Page 16 of 28

Page 17 of 28 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


However, similar medians of the age distribution between groups minimize the possible influence of individual variation. Furthermore, a high cross-validation reproducibility (Q2/R2 > 0.5) with a significant permutation test indicates that the current PLS-DA model is reliable for predicting metabolic changes. Further correlation analysis between metabolites and clinical biochemical indices provides solid evidence for the entire theoretical framework of the IPE metabolism. CONCLUSION NMR-based metabolomic analysis provides a comprehensive insight into the molecular pathogenesis of fibrinous IPE in children with pneumonia. With the progress of pleural infection, increased inflammatory response appears to be strongly associated with increased bacterial fermentation via glucose metabolism along with pleural acidosis. In advanced pneumonia, lactic acid and butyric acid (related to anaerobic bacterial metabolism) and hypoxanthine (through ATP hydrolysis) under hypoxic conditions are significantly positively correlated with fibrin formation in IPE, indicating that these metabolites could serve as potential biomarkers for the management of IPE. However, further research with a larger sample size is warranted to ascertain the clinical applications.



ASSOCIATED CONTENT Supporting information The Supporting Information is available free of charge on the ACS Publications website at DOI: Table S1. PLS-DA cross-validation and permutation test for distinguishing between sexes, pathogens and IPE with and without fibrin. Figure S1. PLS-DA score plots obtained from pleural fluid 1H-NMR spectra analysis between sexes, S. pneumoniae and M. pneumoniae pathogens, and IPE with and without fibrin. Table S2. Functional pathway analysis of four metabolites associated with fibrin formation in infectious parapneumonic effusions (PDF)

AUTHOR INFORMATION Corresponding Author * Chih-Yung Chiu, E-mail: [email protected]. Tel: 886-2-24313131 ext 2606 *Gigin Lin, E-mail: [email protected]. Tel: 886-3-3281200 ext 2575 ORCID Chih-Yung Chiu: 0000-0002-6454-968X Meng-Han Chiang: 0000-0002-2697-4757 Gigin Lin: 0000-0001-7246-1058 18 ACS Paragon Plus Environment

Page 18 of 28

Page 19 of 28 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


Notes The authors declare no competing financial interest. NMR data and quantified peak lists utilized in this manuscript have been deposited to the Figshare data repository (https://figshare.com/s/ea508089bc09c3004531) with the digital object identifier 10.6084/m9.figshare.7599296. ACKNOWLEDGMENTS We acknowledge financial support from the Chang Gung Medical foundation, Chang Gung Memorial Hospital, Taiwan (CMRPG2G0651-3 and CMRPG2E0301) received by Chih-Yung Chiu. The authors are enormously thankful for all support of the metabolomics analysis using 1H-NMR spectroscopy performed at the Metabolomics Core Laboratory, Healthy Aging Research Center (HARC), Chang Gung University and Clinical Metabolomics Core Laboratory, Chang Gung Memorial Hospital.



REFERENCES (1) Marchi, E.; Broaddus, V. C. Mechanisms of pleural liquid formation in pleural inflammation. Curr Opin Pulm Med 1997, 3 (4), 305-309. (2) Chiu, C. Y.; Wong, K. S.; Huang, Y. C.; Lai, S. H.; Lin, T. Y. Echo-guided management of complicated parapneumonic effusion in children. Pediatr Pulmonol 2006, 41 (12), 1226-1232. (3) Philip-Joet, F.; Alessi, M. C.; Philip-Joet, C.; Aillaud, M.; Barriere, J. R.; Arnaud, A.; Juhan-Vague, I. Fibrinolytic and inflammatory processes in pleural effusions. Eur Respir J 1995, 8 (8), 1352-1356. (4) Chiu, C. Y.; Wong, K. S.; Huang, J. L.; Tasi, M. H.; Lin, T. Y.; Hsieh, S. Y. Proinflammatory cytokines, fibrinolytic system enzymes, and biochemical indices in children with infectious para-pneumonic effusions. Pediatr Infect Dis J 2008, 27 (8), 699-703. (5) Chiu, C. Y.; Lin, G.; Cheng, M. L.; Chiang, M. H.; Tsai, M. H.; Lai, S. H.; Wong, K. S.; Hsieh, S. Y. Metabolomic Profiling of Infectious Parapneumonic Effusions Reveals Biomarkers for Guiding Management of Children with Streptococcus pneumoniae Pneumonia. Sci Rep 2016, 6, 24930. (6) Hendaus, M. A.; Janahi, I. A. Parapneumonic Effusion in Children. Clinical Pediatrics 2015, 55 (1), 10-18. (7) Koh, D. M.; Burke, S.; Davies, N.; Padley, S. P. Transthoracic US of the chest: clinical uses and applications. Radiographics 2002, 22 (1), e1. (8) Song, J. Y.; Eun, B. W.; Nahm, M. H. Diagnosis of pneumococcal pneumonia: current pitfalls and the way forward. Infect Chemother 2013, 45 (4), 351-366. (9) Bradley, J. S.; Byington, C. L.; Shah, S. S.; Alverson, B.; Carter, E. R.; Harrison, C.; Kaplan, S. L.; Mace, S. E.; McCracken, G. H., Jr.; Moore, M. R.; St Peter, S. D.; Stockwell, J. A.; Swanson, J. T.; Pediatric Infectious Diseases, S.; the Infectious Diseases Society of, A. The management of community-acquired pneumonia in infants and children older than 3 months of age: clinical practice guidelines by the Pediatric Infectious Diseases Society and the Infectious Diseases Society of America. Clin Infect Dis 2011, 53 (7), e25-76. (10) Westerhuis, J. A.; Hoefsloot, H. C. J.; Smit, S.; Vis, D. J.; Smilde, A. K.; Velzen, E. J. J. v.; Duijnhoven, J. P. M. v.; Dorsten, F. A. v. Assessment of PLSDA cross validation. Metabolomics 2008, 4 (1), 81-89. (11) Utine, G. E.; Ozcelik, U.; Yalcin, E.; Dogru, D.; Kiper, N.; Aslan, A.; Kanra, G. Childhood parapneumonic effusions: biochemical and inflammatory markers. Chest 2005, 128 (3), 1436-1441. (12) Mihalyi, E.; Tercero, J. C.; Diaz-Maurino, T. Changes in pH associated with 20 ACS Paragon Plus Environment

Page 20 of 28

Page 21 of 28 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


clotting of fibrinogen. Kinetic studies of the pH shift and correlation of the pH change with the release of fibrinopeptides and the ensuing polymerization. Biochemistry 1991, 30 (19), 4753-4762. (13) Chernysh, I. N.; Nagaswami, C.; Purohit, P. K.; Weisel, J. W. Fibrin Clots Are Equilibrium Polymers That Can Be Remodeled Without Proteolytic Digestion. Sci Rep 2012, 2, 879. (14) Prais, D.; Kuzmenko, E.; Amir, J.; Harel, L. Association of hypoalbuminemia with the presence and size of pleural effusion in children with pneumonia. Pediatrics 2008, 121 (3), e533-538. (15) Chiu, C. Y.; Hsieh, S. Y.; Wong, K. S.; Lai, S. H.; Chen, J. K.; Huang, J. L. The value of total protein in guiding management of infectious parapneumonic effusion by using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. J Microbiol Immunol Infect 2015, 48 (5), 483-489. (16) Hendriksen, W. T.; Kloosterman, T. G.; Bootsma, H. J.; Estevao, S.; de Groot, R.; Kuipers, O. P.; Hermans, P. W. Site-specific contributions of glutamine-dependent regulator GlnR and GlnR-regulated genes to virulence of Streptococcus pneumoniae. Infect Immun 2008, 76 (3), 1230-1238. (17) Laiakis, E. C.; Morris, G. A. J.; Fornace, A. J.; Howie, S. R. C. Metabolomic analysis in severe childhood pneumonia in the Gambia, West Africa: findings from a pilot study. PLoS One 2010, 5 (9), e12655. (18) Moffett, J. R.; Namboodiri, M. A. Tryptophan and the immune response. Immunol Cell Biol 2003, 81 (4), 247-265. (19) Cashen, K.; Petersen, T. L. Pleural Effusions and Pneumothoraces. Pediatrics in Review 2017, 38 (4), 170. (20) Werno, A. M.; Murdoch, D. R. Medical microbiology: laboratory diagnosis of invasive pneumococcal disease. Clin Infect Dis 2008, 46 (6), 926-932. (21) Border, P. M.; Kierstan, M. P. J.; Plastow, G. S. Production of propionic acid by mixed bacterial fermentation. Biotechnology Letters 1987, 9 (12), 843-848. (22) Van Immerseel, F.; Ducatelle, R.; De Vos, M.; Boon, N.; Van De Wiele, T.; Verbeke, K.; Rutgeerts, P.; Sas, B.; Louis, P.; Flint, H. J. Butyric acid-producing anaerobic bacteria as a novel probiotic treatment approach for inflammatory bowel disease. J Med Microbiol 2010, 59 (Pt 2), 141-143. (23) Lai, S. H.; Wong, K. S.; Liao, S. L. Value of Lung Ultrasonography in the Diagnosis and Outcome Prediction of Pediatric Community-Acquired Pneumonia with Necrotizing Change. PLoS One 2015, 10 (6), e0130082. (24) Saugstad, O. D. Hypoxanthine as an indicator of hypoxia: its role in health and disease through free radical production. Pediatr Res 1988, 23 (2), 143-150. (25) Jabs, C. M.; Neglen, P.; Eklof, B. Breakdown of adenine nucleotides, formation 21 ACS Paragon Plus Environment


of oxygen free radicals, and early markers of cellular injury in endotoxic shock. Eur J Surg 1995, 161 (3), 147-155.


Page 22 of 28

Page 23 of 28 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 LEGENDS Figure 1. Heatmap of correlations between pleural fluid metabolites with inflammatory cytokines, biochemical indices and fibrinolytic system enzymes. Color intensity represents the magnitude of correlation. Red color = positive correlations; blue color = negative correlations. + symbol means a P-value < 0.05; ++ symbol means a P-value < 0.01.

Figure 2. Venn diagram of the metabolites correlated with pleural variables and metabolic pathways of metabolites significantly related to fibrin formation in infectious parapneumonic effusions. (A) The total metabolite number correlated with IL-1β, glucose, pH and PAI-1 in each set and in the overlapping areas is indicated. Glucose, lactic acid, 3-hydroxybutyric acid and hypoxanthine correlating with PAI-1 are in the common area. (B) Schematic overview of metabolic pathways in infectious parapneumonic effusions with fibrin formation. Metabolites significantly correlated with PAI-1 are shown in blue color.



Page 24 of 28

TABLES Table 1. Baseline characteristics and pleural effusion variables of children between non-fibrinous and fibrinous infectious parapneumonic effusions. Characteristics

Non-fibrin (n = 30)

Fibrin (n = 28)

P-value

4.8 (1.8-17.0)

3.5 (0.3-16.5)

0.003

17 (57%)

15 (54%)

0.813

S. pneumoniae

17 (57%)

21 (75%)

0.142

Mycoplasma pneumoniae

5 (17%)

1 (4%)

0.102

Unknown

8 (27%)

6 (21%)

0.641

WBC, 109/L

11.4 ± 5.7

18.9 ± 9.2

0.001

Hb, g/dL

10.8 ± 1.3

10.1 ± 2.0

0.091

Platelet, 109/L

307.5 ± 168.3

384.3 ± 257.7

0.195

CRP, mg/L

178.5 ± 118.0

188.6 ± 108.8

0.742

TNF-α, pg/mL

13.6 (0.3-799.3)

16.8 (0.3-2007.2)

0.188

IL-1β, pg/mL

13.7 (0.5-820.2)

84.6 (0.5-853.9)

< 0.001

85 (11-133)

38.5 (1-170)

0.001

LDH, IU/L

471 (10-23,243)

1126 (135-25,164)

0.005

pH value

7.40 (7.27-7.93)

7.18 (6.5-7.5)

< 0.001

Age, yr Sex, male Pathogen

Hemogram

Pleural effusion variables Inflammatory cytokines

Biochemical markers Glucose, mg/dL

Fibrinolytic system enzymes PAI-1, ng/mL tPA, ng/mL Intervention procedures Hospital stay, d

744.4 (58.4-15,020.7) 8696.8 (35.3-21,138.1)

0.005

14.9 (3.6-49.5)

5.5 (0.5-51.8)

0.016

9 (30%)

22 (79%)

< 0.001

12.8 ± 5.4

15.5 ± 6.1

0.036

Data shown are mean  SD, median (range) or number (%) of patients as appropriate. yr, year; WBC, white blood cell; Hb, hemoglobin; CRP, C-reactive protein; TNF, tumor necrosis factor; IL, interleukin; LDH, lactate dehydrogenase; PAI-1, plasminogen activator inhibitor-1; tPA, tissue plasminogen activator; d, day.


Page 25 of 28 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


Table 2. Assignment results of pleural fluid 1H-NMR spectra of the identified metabolites and chemical shifts, and VIP scores and fold changes differentially expressed between different pathogens and IPE with and without fibrin formation. Pathogens Compound name

Chemical shift,

VIP

Fold

ppm (multiplicity) scorea changeb

IPE P-value

VIP

Fold

scorea changeb

P-value

Glucose

5.255-5.245 (d)

4.79

0.44

Metabolomics Reveals Anaerobic Bacterial Fermentation and

Recommend Documents