Lipidomics in Health & Disease

Translational Bioinformatics 14 Series Editor: Xiangdong Wang, MD, PhD, Prof

Xiangdong Wang · Duojiao Wu  Huali Shen Editors

Lipidomics in Health & Disease Methods & Application

Translational Bioinformatics Volume 14 Series editor Xiangdong Wang, MD, Ph.D. Zhongshan Hospital Institute of Clinical Science, Fudan University Shanghai Medical College, China Shanghai Institute of Clinical Bioinformatics, China

Aims and Scope The Book Series in Translational Bioinformatics is a powerful and integrative resource for understanding and translating discoveries and advances of genomic, transcriptomic, proteomic and bioinformatic technologies into the study of human diseases. The Series represents leading global opinions on the translation of bioinformatics sciences into both the clinical setting and descriptions to medical informatics. It presents the critical evidence to further understand the molecular mechanisms underlying organ or cell dysfunctions in human diseases, the results of genomic, transcriptomic, proteomic and bioinformatic studies from human tissues dedicated to the discovery and validation of diagnostic and prognostic disease biomarkers, essential information on the identification and validation of novel drug targets and the application of tissue genomics, transcriptomics, proteomics and bioinformatics in drug efficacy and toxicity in clinical research. The Book Series in Translational Bioinformatics focuses on outstanding articles/chapters presenting significant recent works in genomic, transcriptomic, proteomic and bioinformatic profiles related to human organ or cell dysfunctions and clinical findings. The Series includes bioinformatics-driven molecular and cellular disease mechanisms, the understanding of human diseases and the improvement of patient prognoses. Additionally, it provides practical and useful study insights into and protocols of design and methodology. Series Description Translational bioinformatics is defined as the development of storage-related, analytic, and interpretive methods to optimize the transformation of increasingly voluminous biomedical data, and genomic data in particular, into proactive, predictive, preventive, and participatory health. Translational bioinformatics includes research on the development of novel techniques for the integration of biological and clinical data and the evolution of clinical informatics methodology to encompass biological observations. The end product of translational bioinformatics is the newly found knowledge from these integrative efforts that can be disseminated to a variety of stakeholders including biomedical scientists, clinicians, and patients. Issues related to database management, administration, or policy will be coordinated through the clinical research informatics domain. Analytic, storage-related, and interpretive methods should be used to improve predictions, early diagnostics, severity monitoring, therapeutic effects, and the prognosis of human diseases. Recently Published and Forthcoming Volumes Computational and Statistical Epigenomics Editor: Andrew E. Teschendorff Volume 7

Allergy Bioinformatics Editors: Ailin Tao, Eyal Raz Volume 8

Transcriptomics and Gene Regulation Editor: Jiaqian Wu Volume 9

Pediatric Biomedical Informatics – Computer Applications in Pediatric Research (Edition 2) Editor: John J. Hutton Volume 10

More information about this series at http://www.springer.com/series/11057

Xiangdong Wang • Duojiao Wu • Huali Shen Editors

Lipidomics in Health & Disease Methods & Application

Editors Xiangdong Wang Zhongshan Hospital Institute of Clinical Science Fudan University Shanghai Medical College Shanghai, China Shanghai Institute of Clinical Bioinformatics Shanghai, China

Duojiao Wu Shanghai Institute of Clinical Bioinformatics Shanghai, China Zhongshan Hospital Institute of Clinical Science Fudan University Shanghai Medical School Shanghai, China

Huali Shen Institutes of Biomedicine Fudan University Shanghai, China

ISSN 2213-2775     ISSN 2213-2783 (electronic) Translational Bioinformatics ISBN 978-981-13-0619-8    ISBN 978-981-13-0620-4 (eBook) https://doi.org/10.1007/978-981-13-0620-4 Library of Congress Control Number: 2018948704 © Springer Nature Singapore Pte Ltd. 2018 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made. Printed on acid-free paper This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd. The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

Contents

1 Clinical Lipidomics: A Critical Approach for Disease Diagnosis and Therapy........................................................ 1 Xiangdong Wang 2 The Role of Lipid Metabolism in the Development of Lung Cancer......................................................................................... 7 Lixin Wang, Weiling Huang, and Xiu-Min Li 3 Bioinformatics of Embryonic Exposures: Lipid Metabolism and Gender as Biomedical Variables...................................................... 21 K. K. Linask 4 An Evaluation of Multivariate Data Analysis Models for Lipidomic Parameters from Patients with Metabolic Syndrome Undergoing Remedial Treatment......................................... 39 D. Farabos, C. Wolf, R. Chapier, A. Lamaziere, and Peter J. Quinn 5 Lipidomics in Carotid Artery Stenosis: Further Understanding of Pathology and Treatment........................... 55 Wei Zhang, Xiushi Zhou, Daqiao Guo, Weiguo Fu, and Lixin Wang 6 Metabolomics of Immunity and Its Clinical Applications.................... 73 Jing Qiu, Fangming Liu, and Duojiao Wu 7 Urinary Lipidomics.................................................................................. 97 Phornpimon Tipthara and Visith Thongboonkerd 8 Breast Cancer and Lipid Metabolism.................................................... 113 Chunfa Huang, Yuntao Li, Yifan Tu, and Carl E. Freter 9 Association of Circulating Oxidized Lipids with Cardiovascular Outcomes.............................................................. 137 Irena Levitan, Ibra S. Fancher, and Evgeny Berdyshev

v

vi

Contents

10 Lipidomics: Mass Spectrometry Based Untargeted Profiling and False Positives.................................................................... 155 Xiaohui Liu, Lina Xu, Xueying Wang, and Yupei Jiao 11 Phospholipid and Phospholipidomics in Health and Diseases.............................................................................................. 177 Tanxi Cai and Fuquan Yang Index.................................................................................................................. 203

Chapter 1

Clinical Lipidomics: A Critical Approach for Disease Diagnosis and Therapy Xiangdong Wang

Abstract  Clinical lipidomics is an important merging discipline to integrate clinical medicine and lipid science for diagnosis and therapy of human disease. The clinical lipidomics is defined as a new integrative biomedicine to discover the correlation and regulation between a large scale of lipid elements measured and analyzed in liquid biopsies from patients with those patient phenomes and clinical phenotypes. One of the important and challenging issues in clinical lipidomics is to define the disease specificity of dyslipidemia and lipid dysregulation. The comparison of lipidomic profile difference between target disease and healthy as well as related diseases is a common approach to perform lipidomics in patients. It is challenging to define the disease specificity of lipids and lipid metabolism, especially for those lipid species and their abundances. The heterogeneity of lipidomic profiles between different diseases is more obvious than that between different stages or severities of one disease. It is a challenge to validate the stage or severity specificity of selected biomarkers and targets. In order to improve the understanding of disease mechanisms in multiple dimensions, clinical lipidomics should/must be merged with clinical phenomes, e.g. patient signs and symptoms, biomedical analyses, pathology, images, and responses to therapies. We believe clinical lipidomics will become one of the most important and helpful approaches during the design and decision-making of therapeutic strategies for individuals. Clinical lipidomics is an important merging discipline to integrate clinical medicine and lipid science for diagnosis and therapy of human disease. The clinical lipidomics is defined as a new integrative biomedicine to discover the correlation and regulation between a large scale of lipid elements measured and analyzed in liquid

X. Wang (*) Zhongshan Hospital Institute of Clinical Science, Fudan University Shanghai Medical College, Shanghai, China Shanghai Institute of Clinical Bioinformatics, Shanghai, China e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2018 X. Wang et al. (eds.), Lipidomics in Health & Disease, Translational Bioinformatics 14, https://doi.org/10.1007/978-981-13-0620-4_1

1

2

X. Wang

biopsies from patients with those patient phenomes and clinical phenotypes, in order to draw a full atlas of human lipids in both physiological and pathophysiological conditions, better understand molecular mechanisms of lipid metabolism and abnormality, and identify diagnostic biomarkers and therapeutic targets. It is not as simple as a measurement of lipids in the patients, a performance of lipidomics in clinic, and a simple correlation between lipid changes and clinical measures. Clinical lipidomics here is to emphasize the word of integration among multiomics which specially include clinical phenomics. As the part of clinical trans-omics, clinical lipidomics is a new approach to map lipidomic profiles of patient liquids, cells, and tissues, provide a big data of lipids with clinical phenomes for the database, and explore new mechanisms by which lipids contribute to disease development. Clinical lipidomics is a powerful tool to discover and validate new diagnoses which can be used to monitor disease severity, during, stage, sensitivity to drugs, and new druggable targets for drug discovery and development. The present chapter with a clear and specific focus on clinical lipidomics aims to define the concept of clinical lipidomics, demonstrate importance of the integration between clinical lipidomics and patient phenomes, and investigate molecular mechanisms of lipid metabolisms.

1.1  Values of Disease Specificity One of the important and challenging issues in clinical lipidomics is to define disease specificity of dyslipidemia and lipid dysregulation. The comparison of lipidomic profile difference between target disease and healthy as well as related diseases is a common approach to perform lipidomics in patients. For example, Kim et al. investigates altered lipid profiles and dysregulation in patients with behavioral variant frontotemporal dementia, by measuring and analyzing the comprehensive lipidomic profiles of blood plasma, as compared with patients with Alzheimer’s disease and controls, using liquid chromatography-tandem mass spectrometry (Kim et al. 2018). The Alzheimer’s disease was selected as the disease reference in this particular study according to the similarity of the target disease. The disease reference is one of the most important parts of clinical study and designs, which is always ignored or lacked due to the resource of patients. The dyslipidemia and lipid dysregulation can occur in multiple conditions and are closely associated with diets and life styles. Kim’s study measured four major classes of lipids (glycerolipids, phospholipids, sphingolipids, sterols), 17 subclasses of lipids, and 3225 putative individual lipid species in total, as well as a group of dietary lipids (Kim et al. 2018). It still needs to be furthermore clarified about the necessity and importance of “a group of dietary lipids” in clinical lipidomic studies, while such information will definitely benefit for understanding of the disease-associated factors.

1  Clinical Lipidomics: A Critical Approach for Disease Diagnosis and Therapy

3

1.2  Values of Disease Biomarkers There are a large number of preclinical and clinical studies on identification and validation of disease-specific biomarkers, which can have the high specificity of the capacity to monitor one of clinical phenomes, e.g. the severity, stage, duration, response to therapy, and prognosis of patients (Zhu et al. 2016; Shi et al. 2018a, b; Chen et al. 2016; Xu and Wang 2017). Kim selected disease-associated biomarkers in this particular study by comparing the difference between target disease and the same category of diseases and found five lipid molecules-TG (16:0/16:0/16:0), diglyceride (18:1/22:0), phosphatidylcholine (32:0), phosphatidylserine (41:5), and sphingomyelin (36:4) (Kim et al. 2018). The significant difference of lipidomic elements between the similar or same category of diseases is an important criterion for scientists to consider or suggest as potential or developing biomarkers. It should be aware that the disease specificity of biomarkers should have multi-diseases as the reference. It is also challenging to define the disease specificity of lipids and lipid metabolism, especially for those lipid species and their abundances. Lydic and Goo recently provided a comprehensive review on the importance and clinical potentials of lipidomics and described clearly about the complexity of the lipidome in metabolic diseases (Lydic and Goo 2018). Lipidomics can be used to identify and detect the abundances of lipids at the same time, during which methods of samplings, extractions, and measurements can be varied significantly. This particular review addressed a number of challenges in clinical application, including various methods, limited understanding, and stability of measurements, especially global standards of each process to quantify and analyze lipids. There is an urgent and critical need of global databases.

1.3  Monitoring Disease Severity Lipid metabolism can dynamically alter with the development and progression of the disease. Gorden et  al. performed an outstanding and comprehensive clinical study to measure lipid metabolites, aqueous intracellular metabolites, SNPs, and mRNA transcripts in liver biopsies, plasma, and urine samples of patients at different stages of nonalcoholic fatty liver disease, e.g. steatosis, nonalcoholic steatohepatitis, and cirrhosis (Gorden et al. 2015). Although the number of lipid species and clinical cases were relatively small, the study design with a clear focus on different stages of the disease was more logical and impressed by clinicians. The heterogeneity of lipidomic profiles between different diseases is more obvious than that between different stages or severities of one disease. It is a challenge to validate the stage or severity specificity of selected biomarkers and targets, since it is expected that block and knockdown of the biomarkers and targets should prolong or terminate the progression of the disease. Such findings are often detected when the

4

X. Wang

biomarkers and targets are genes or proteins, while the specific biological roles and regulations of lipid species and its abundances should be hardly validated. Ščupáková et al. recently demonstrated that a new concept of “spatial systems lipidomics” was applied for the identification of disease heterogeneity, rather than lipid species per se (Ščupáková et al. 2018). This particular study correlated the distribution of specific lipid species with the severity of nonalcoholic fatty liver diseases by integrating lipidomic profiles of hepatocytes, pathological images, and global databases, to figure out the difference of spatially resolved lipid profiles between non-steatotic and steatotic tissues, severity-specific lipids, their networks, as well as metabolisms. This is an important initiative and a scientific example of clinical trans-omics, which is recently redefined to integrate molecular multiomics, e.g. genomics, proteomics, and lipidomics, with clinical phenomics (Wang 2018).

1.4  Integration with Clinical Phenomes Clinical trans-omics was coined as a new emerging scientific discipline to integrate clinical phenomes with molecular multiomics, in order to further understand molecular mechanisms of disease pathogenesis and progression, patient sensitivity to therapy and prognosis, and therapy design and development (Wang 2018). Like genomics and proteomics, we believe that the concept of “clinical trans-omics” is also applicable and important in the understanding of lipidomics, especially the importance of clinical lipidomics. In order to improve the understanding of disease mechanisms in multiple dimensions, clinical lipidomics should/must be merged with clinical phenomes, e.g. patient signs and symptoms, biomedical analyses, pathology, images, and responses to therapies. We initiated such clinical trans-­ omics studies on the integration of genomics or/and proteomics with clinical phenomes in patients with with acute exacerbation of chronic obstructive pulmonary disease, as compared with healthy individuals, long-term smokers, or individuals at a stable stage of chronic obstructive pulmonary disease or lung cancer (Shi et al. 2018a; Chen et al. 2012a, b, 2016). It is more challenging and difficult to integrate and fuse the information of clinical lipidomics with clinical phenomes, since both results are unstable, less specific, hardly quantified, or unrepeatable. It is also hard to automatically and dynamically generate, collect, integrate, and analyze large-­ scale data of clinical lipidomes and phenomes. Aidoud et al. applied targeted and untargeted lipidomics and positron emission tomography scan imaging to evaluate the effect of adding docosahexaenoic:arachidonic acids (3:2) on brain functional activity and the brain and eye fatty acid and lipid composition (Aidoud et al. 2018). This particular study measured lipidomics and organ function and activity separately, and made the correlation between brain lipidomics and function, which is a milestone step to march the integration of lipidomics with clinical phenomes. Clinical images are the major part of clinical phenomes, including pathological images, ultrasounds, computational tomography, nuclear magnetic resonance images, or face image. It is a challenge to draw the networks of lipidomic species as we did networks of genes and proteins, since it

1  Clinical Lipidomics: A Critical Approach for Disease Diagnosis and Therapy

5

seems the less regulatory roles among lipid species. We recently tried to explore the correlation and network of genomics with lipidomic species and clinical phenotypes, identify disease-specific gene-associated or/and regulated lipidomic profiles, and discovery disease-specific and subtype-specific diagnostic biomarkers and therapeutic targets in patients with lung cancer (Lv et al. 2018). We believe clinical lipidomics will become one of the most important and helpful approaches during the design and decision-making of therapeutic strategies for individuals.

References Aidoud N, Delplanque B, Baudry C, Garcia C, Moyon A, Balasse L, et al. A combination of lipidomics, MS imaging, and PET scan imaging reveals differences in cerebral activity in rat pups according to the lipid quality of infant formulas. FASEB J. 2018:fj201800034R. https://doi. org/10.1096/fj.201800034R. Chen H, Song ZJ, Qian MJ, Bai CX, Wang XD. Selection of disease-specific biomarkers by integrating inflammatory mediators with clinical informatics in AECOPD patients: a preliminary study. J Cell Mol Med. 2012a;16(6):1286–97. Chen H, Wang YL, Bai CX, Wang XD.  Alterations of serum inflammatory biomarkers in the healthy and chronic obstructive pulmonary disease patients with or without acute exacerbation. J Cell Mol Med. 2012b;16(6):1286–97. Chen C, Shi L, Li Y, Wang X, Yang S. Disease-specific dynamic biomarkers selected by integrating inflammatory mediators with clinical informatics in ARDS patients with severe pneumonia. Cell Biol Toxicol. 2016;32(3):169–84. https://doi.org/10.1007/s10565-016-9322-4. Gorden DL, Myers DS, Ivanova PT, Fahy E, Maurya MR, Gupta S, et al. Biomarkers of NAFLD progression: a lipidomics approach to an epidemic. J Lipid Res. 2015;56(3):722–36. https:// doi.org/10.1194/jlr.P056002. Kim WS, Jary E, Pickford R, He Y, Ahmed RM, Piguet O, et al. Lipidomics analysis of behavioral variant frontotemporal dementia: a scope for biomarker development. Front Neurol. 2018;9:104. https://doi.org/10.3389/fneur.2018.00104. eCollection 2018 Lv J, Gao D, Zhang Y, Wu D, Shen L, Wang X. Variations of lipidomic profiles among lung cancer subtypes of patients. J Cell Mol Med. 2018; accepted. Lydic TA, Goo YH. Lipidomics unveils the complexity of the lipidome in metabolic diseases. Clin Transl Med. 2018;7(1):4. https://doi.org/10.1186/s40169-018-0182-9. Ščupáková K, Soons Z, Ertaylan G, Pierzchalski KA, Eijkel GB, Ellis SR, et al. Spatial systems lipidomics reveals nonalcoholic fatty liver disease heterogeneity. Anal Chem. 2018. https://doi. org/10.1021/acs.analchem.7b05215. Shi L, Zhu B, Xu M, Wang X.  Selection of AECOPD-specific immunomodulatory biomarkers by integrating genomics and proteomics with clinical informatics. Cell Biol Toxicol. 2018a;34(2):109–23. https://doi.org/10.1007/s10565-017-9405-x. Epub 2017 Aug 4 Shi L, Dong N, Ji D, Huang X, Ying Z, Wang X, Chen C.  Lipopolysaccharide-induced CCN1 production enhances interleukin-6 secretion in bronchial epithelial cells. Cell Biol Toxicol. 2018b;34(1):39–49. https://doi.org/10.1007/s10565-017-9401-1. Wang X.  Clinical trans-omics: an integration of clinical phenomics with multiomics. Cell Biol Toxicol. 2018 Jun;34(3):163-166. https://doi.org/10.1007/s10565-018-9431-3 Xu M, Wang X. Critical roles of mucin-1 in sensitivity of lung cancer cells to tumor necrosis factor-­ alpha and dexamethasone. Cell Biol Toxicol. 2017;33(4):361–71. https://doi.org/10.1007/ s10565-017-9393-x. Zhu D, Liu Z, Pan Z, Qian M, Wang L, Zhu T, Xue Y, Wu D. A new method for classifying different phenotypes of kidney transplantation. Cell Biol Toxicol. 2016;32(4):323–32. https://doi. org/10.1007/s10565-016-9337-x.

Chapter 2

The Role of Lipid Metabolism in the Development of Lung Cancer Lixin Wang, Weiling Huang, and Xiu-Min Li

Abstract  The increase of lipid synthesis in tumor tissues has been considered as an important component of substance and energy metabolism during cell transformation. In recent years, the role of lipids in the transformation of cells into tumors, tumor growth, invasion and metastasis have attracted much attention. This article reviews the effects of lipid metabolism related enzymes, membrane lipids and extracellular environment on the development of lung cancer. Finally, the application of lipid metabolism related drugs in lung cancer was briefly summarized.

2.1  Introduction Lung cancer is one of the deadliest cancer in all cancers, and about 85% are non-­ small cell lung cancers (NSCLC), which consist of squamous cell carcinoma, large cell carcinomas and adenocarcinoma (Blandin Knight et  al. 2017). Although the abundance of therapeutic methods and drugs (e.g. chemotherapy, radiotherapy, surgery and traditional medicine) has been carried out, the mortality rate of lung cancer still remains at high level. Up to now, many studies have focused on areas of genome, transcriptome, proteome and metabolome, trying to find out the effective methods and molecular characterization to treat lung cancer, but the lipidome still poorly understand. Nevertheless, lipids play a key role in cell biology and lipid metabolism is a critical factor in the development of tumor (Huang and Freter 2015).

L. Wang Department of Integrated Traditional Chinese and Western Medicine, Shanghai Pulmonary Hospital, Tongji University, Shanghai, China W. Huang Graduate school, Shanghai University of Traditional Chinese Medicine, Shanghai, China X.-M. Li (*) Center of Integrative Medicine for Allergies and Wellness, Icahn School of Medicine at Mount Sinai, New York, NY, USA e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2018 X. Wang et al. (eds.), Lipidomics in Health & Disease, Translational Bioinformatics 14, https://doi.org/10.1007/978-981-13-0620-4_2

7

8

L. Wang et al.

Lipids are one of the three major nutrients with two main functions. One is that lipids are the main component of cells (Muro et al. 2014). Phospholipids (glycerol phospholipids and sphingomyelin) and cholesterol are the major components of cell membrane (Schuster and Sleytr 2014). Its metabolism changes directly affect cell membrane synthesis and proliferation, and lipids are the major active molecules in cell activities. Variety of lipid molecules and their metabolic intermediates are involved in the cell signaling, inflammation and vascular regulation, and promote cell proliferation, cell adhesion and movement (Gómez de Cedrón and Ramírez de Molina 2016). Therefore, abnormal of lipid metabolism is not only closely related to cardiovascular disease, but also closely related to tumorigenesis, development, invasion and metastasis. The metabolic activity in tumor cells is usually different from that in normal cells. Lipid metabolism is an important component in the alteration of tumor metabolism, also known as tumor metabolic reprogramming. The reprogramming of metabolic process contributes to increase production of proteins, nucleic acids, lipids, and metabolic intermediates, which is a prerequisite for rapid proliferation of tumor cells. For the tumor cells, the major abnormal of lipid metabolism is the sustain lipid synthesis and altered fatty acid metabolism, in particular the synthesis of de novo fatty acids, which are necessary for the proliferation of tumor cells (Daniëls et al. 2014). However, in tumor hosts, lipid metabolism alteration is opposite, which continuously consume lipid, as well as disorderly utilize exogenous lipid. All of these changes are closed to the alteration of metabolic enzyme, the function of membrane and the changes of tumor environment.

2.2  Natural Lipid Metabolism in Mammalian Cells When citric acid is transported out of mitochondria, pyrolysis by ATP citrate lyase (ACLY) into acetyl CoA and oxaloacetate; then transformed by the acetyl coenzyme A carboxylase (ACC) into malonyl-CoA, acetyl-CoA and malonyl-CoA coupling to acyl carrier generating 16 alkaline carbon saturated fatty acid by the effect of repeated condensation of fatty acid synthase (FASN), palmitic acid by condensation, to produce a variety of effects of saturated and unsaturated fatty acids for use in mammalian cells. The mammalian main desaturase is stearoyl coenzyme A desaturase (SCD), acid palmitic acid and stearic acid in the delta 9 introduced double bond to produce monounsaturated fatty acids (Fig. 2.1) (Qiang et al. 2015). However, the body itself cannot produce ω-3, orω-6 unsaturated fats Acids, these linoleic acids, and α-linoleic acid, must be ingested from food (Simopoulos 2016). Fatty acids can be converted into diacylglycerol (DAG) and triacylglycerol (TAG) by the glycerol phosphate pathway in glycolysis, while TAG is mainly converted to lipid droplets to store energy (Eichmann and Lass 2015; Kopf et al. 2014). At the same time, fatty acids can be converted into a variety of glycerol phosphate such as phosphatidyl choline (PC) and phosphatidylethanolamine (PE), phosphatidyl glycerol (PG) and phosphatidylserine (PS) which are the major components of biological membranes.

2  The Role of Lipid Metabolism in the Development of Lung Cancer

9

Fig. 2.1  Lipid synthesis in mammalian cells

Another important type of lipid for membrane function is sterols, such as cholesterol and cholesterol esters (CE). Because of the fluidity of lipid bilayer structure, cholesterol is an important membrane component. It can also be used to synthetize estrogen, progesterone and other hormone drugs. Fatty acid also can be transformed into sphingolipids, phosphatidylinositol, eicosanoids and others (Montagne et  al. 2014). Recent studies have also shown that tumor cells can provide citrate by the reductive metabolism of glutamine. Under certain conditions, acetate directly added from outside also contributes to the biosynthesis of acetyl coenzyme A (Zaidi and Swinnen 2012).

2.3  Alteration of Lipid Metabolism in Lung Cancer As early as the middle of the last century, it was proved that lipid synthesis was enhanced in tumor cells and tumor tissues, resulting in the accumulation of lipids, including FA and phospholipids (Medes et al. 1953). By comparing tumor tissue and tumor adjacent tissue, some experiments revealed the rate of FA synthesis in tumor tissue is significantly high (Medes et al. 1953). Although FA may be ingested from the cell tissue environment, the de novo synthesis of lipids by the tumor itself provides the vast majority of lipids during tumor growth. Further studies have shown that the specific antigen OA-519, which regulates lipid synthesis in tumor cells, is FASN (Kuhajda et al. 1994). Subsequent studies confirmed that loss of lipid

10

L. Wang et al.

synthesis or inhibition of enzyme synthesis during lipid synthesis could block tumor cell growth (Sounni et al. 2014; Hopperton et al. 2014). Genomic analysis showed that the enhancement of FA biosynthesis is an early feature of tumor progression (Carracedo et al. 2013; Jerby et al. 2012). The progressive tumor showed a decrease in proliferative capacity and an increase in antioxidant capacity. This suggests that anabolism is important in the early expansion of tumors, whereas advanced tumors are more dependent on the antioxidant activity of reactive oxygen species (ROS) (Dayal et al. 2014).

2.4  Lipid Synthetase and Lung Cancer Fatty acids are important fuel for cell growth and are essential components of cell membranes. The substrate for de novo fatty acid synthesis is acetyl coenzyme A, that comes from two main sources: first is citric acid, mainly from the three carboxylic acid cycle. When citric acid is transported out of mitochondria, pyrolysis by ACLY into acetyl CoA and oxaloacetate. ACC carboxylates acetyl-CoA to form malonyl-CoA, which is then used as a substrate for fatty acid synthase to form the 16 carbon saturated fatty acid palmitate. Second, tumor cells directly ingest acetic acid from the extracellular region and catalyze acetyl CoA synthesis by acetyl CoA synthase. In this process, the abnormalities of some key metabolic enzymes will affect the synthesis of fatty acids, thereby affecting the proliferation of tumors (Fig. 2.2).

Fig. 2.2  The process of lipid synthetase in lung cancer

2  The Role of Lipid Metabolism in the Development of Lung Cancer

11

2.4.1  ATP Citrate Lyase (ACLY) ACLY is a cytoplasmic enzyme catalytic the pyrolysis of citric acid and oxaloacetate to generate acetyl-CoA (Lin et al. 2013). It has been identified that ACLY is upregulated or activated in many tumors, such as lung cancer (Lin et al. 2013), if the activity of ACLY can obviously block the proliferation of tumor cells. Lei (Lin et al. 2013) found that acetylation occurs at the polypeptide chain of ACLY 540, 546 and 554 lysine residues (K) sites, promoting lipid biosynthesis and tumor growth, and confirmed that the acetylation level of ACLY was significantly higher in lung cancer. However, Csanadi A et  al. (2015) indicated that overexpression of ACLY in NSCLC has different impacts between young and old patients, e.g. overexpression of ACLY prolongs young patients’ overall survival, while opposite in old patients.

2.4.2  Acetyl-CoA Carboxylase (ACC) ACC is the rate-limit enzyme of fatty acid synthesis, is inactivated by phosphorylation by AMP-activated protein kinase (AMPK) and potentially regulated by many other kinases (Zordoky et al. 2014; Yamanaka et al. 2016). ACC has two isoforms in mammals, of which ACC1 is related to FA synthesis occurs in lipogenic tissues such as the liver, whereas ACC2 is highly expressed in oxidative tissues such as heart and muscle, preventing FA degradation (Aragane et al. 2007; O’Neill et al. 2014). Jeon et al. (2012) observed that the silencing of ACC accelerated the growth in lung cancer cells by promoting NADPH-dependent redox balance, while a contradictory result was observed with ACC that ND-646, an allosteric inhibitor of the ACC enzymes, inhibits de novo fatty acid synthesis in  vitro and in  vivo of NSCLC (Svensson et al. 2016).

2.4.3  Fatty Acid Synthase (FASN) FASN is a key rate-determining enzyme for de novo FA biogenesis (Kuhajda et al. 1994). FASN catalyzes successive condensation reactions to form a fatty acid from malonyl-CoA and acetyl-CoA substrates, producing mainly 16-carbon palmitate. In most tissues, the expression of FASN is low, however strongly high in some tumor tissues, which indicated that fatty acid was synthesized in some cancer cells and tumors (Jiang et al. 2014; Menendez and Lupu 2017). FASN is a particularly appealing therapeutic target because most cancer cells depend upon FASN-mediated de novo FA synthesis, whereas most non-cancer cells prefer exogenous FA. FASN is required for adult neuronal stem cell function (Knobloch et al. 2013). Wang et al. (2002) found that FAS expression in the stage I of lung cancer was associated with a poorer prognosis and suggested that FAS overexpression in early lung tumors may be a signal of aggressiveness.

12

L. Wang et al.

2.4.4  A  cyl-Coenzyme A (CoA) Synthetase Long-Chain Family (ACSL) Acyl-coenzyme A (CoA) synthetase long-chain family are critical enzymes, involved in the initial steps of FA metabolism. ACSL preferentially converts fatty acids with C12-C20 into fatty Acyl-CoA esters, which served as the substrates for lipid synthesis and β-oxidation (Prior et al. 2014). ACSL has five different isoforms (i.e., ACSL1, 3, 4, 5, and 6) and each isoform has different effects on lung cancer. Chen et  al. (2016) demonstrated that the different expressions of ACSL family might predict different cancers during carcinogenesis. For instance, the overexpression of ACSL1 suppresses tumor proliferation, migration, and anchorage-­ independent growth in lung cancer, but the high expression of ACSL4 and ACSL5 might predict a better prognosis in lung cancer. Padanad (Padanad et  al. 2016) showed that ACSL3 plays a potential oncogenic role in mutant KRAS lung cancer in  vivo and in  vitro by promoting fatty acid uptake, accumulation of lipids, and β-oxidation. ACSL6 displayed to promote tumor cell proliferation in neuroblastoma cells and pheochromocytoma (Yan et al. 2015), however no studies have confirmed it related to lung cancer.

2.4.5  Stearoyl-CoA Desaturase 1 (SCD1) Stearoyl-CoA desaturase 1 (SCD1) is one of the enzyme of fatty acids desaturases family, catalyzing the saturated fatty acids (SFAs) into monounsaturated fatty acids (MUFAs) (Sampath and Ntambi 2014). In various types of tumor, SCD1 has been shown to be involved in the rapid proliferation of cancer cells, escaping apoptosis, promoting the malignant transformation of tumor cells. Huang et  al. (2016) also identified that SCD1 was generally and highly expressed in lung adenocarcinoma and accelerated the cell proliferation, migration and invasion in vivo and in vitro. Pisanu et al. (2017) demonstrated that SCD1 is a diagnostic and prognostic marker able to predict the outcome for patients with lung ADC and a promising target for therapeutic intervention in combination with chemotherapy. Byagowi et al. (2015) showed that the expression and activity of SCD1 was functioned by the MEK/ ERK1/2 and EGFR-dependent pathway containing with peroxisome proliferative-­ activated receptorδ (PPARδ) protein.

2.5  The Function of Lipid Membrane and Lung Cancer The structural function of lipid is an important factor in phenotypic transformation. Changes in lipid membrane composition can affect membrane fluidity, signaling and gene expression. Cholesterol and other membrane lipids form into special structure of the cell membrane and cell inter membrane, which are involved in membrane

2  The Role of Lipid Metabolism in the Development of Lung Cancer

13

transportation and served as a signal complex for plasma membrane platforms. Epidermal growth factor receptor (EGFR) (Yamashita et al. 2013) and monounsaturated fatty acids (MUFAs) (Mojumdar et al. 2014) are the main component of membranes, the latter is the precursors of phospholipids, tryglicerides, cholesterol esters, diacylglycerols and wax esters. Recently, Zhang found that that EGFR stabilized SCD1 through phosphorylation of Y55 sites, which increased the synthesis of intracellular MUFAs, thus promoted lung cancer growth (Zhang et al. 2017). It has been identified that the alterations of FA can dramatically affect membrane functions and further influence various organelles. The high expression of FA was associated with theslow speed of membrane fluidity (Parsons and Rock 2013).

2.6  Lipid Signaling Molecules and Lung Cancer In addition to be one of the important membrane components, lipids are also important signaling molecules. Inositol phosphate, an important second messenger, transmits activated signals from growth factor receptors to organelles and serves as highly specific binding platforms to recruit target proteins to specific membranes. The most typical is the phosphatidylinositol-3,4,5- trisphosphate (PIP3), which is generated by P13K and responds to growth factor signaling and mediates the recruitment and activation of Akt and mTOR. PIP3 is also the substrate of phosphatase and tensile protein homolog (PTEN), and PTEN is one of the most frequently mutated or deleted genes in tumor cells (Knoch et al. 2014). Other lipid second messengers include lysophosphatidic acid (LPA), phosphatidic acid (PA) and DAG. LPA is produced by extracellular phospholipaseautocrine protein, and couples with G protein coupled receptors, which can enhance proliferation, migration and survival of tumor cells (Riaz et al. 2016; Foster 2013; Eichmann and Lass 2015). Another important type of signaling lipids is sphingolipid. Pro-apoptotic signals, such as ultraviolet radiation and chemotherapy, produce ceramide and sphingosine. Ceramide regulates tumor cell growth inhibitory signaling and participates in inducing tumor cell apoptosis and inhibiting growth (Goldkorn et al. 2013). Sphingosine-­ 1-­Phosphate (SIP) can also promote cell proliferation, migration and angiogenesis (Książek et al. 2015). Prostaglandin E2 (PGE2) can activate the signal transduction pathway in RAS-ERK cells and induce tumor cell proliferation in an autocrine manner (Corti et al. 2013). Arachidonic acid may modulate inflammation and promote tumorigenesis and progression (Brown et al. 2014).

2.7  Lipid and Autophagy in Lung Cancer Lipid metabolism is closely related to autophagy. Autophagy is a self – degradation mechanism that removes defective proteins and organelles under nutritional restriction. Lipid is a component of the whole autophagy process and has a certain influence on autophagy. The initial step of autophagy involves the combination of PE

14

L. Wang et al.

Fig. 2.3  The process of macrolipophagy occurs by the de novo formation of a limiting membrane that sequesters cytosolic lipid droplets and delivers the lipid cargo to the lysosomes for degradation

with autophagy associated protein ATG8 (LC3). Lipid droplets can respond to autophagosomes in nutritional needs to release the lipid droplets (Fig. 2.3). At the same time, the inhibition of autophagy can lead to the retention of lipid droplets. With the importance of autophagy in cancer, the link between autophagy and lipid metabolism has become increasingly clear (Guo et al. 2013). Autophagy can guarantee the energy supply of RAS tumor cells in hunger and cancer (White 2012).

2.8  Lipid Metabolism and Hypoxia in Lung Cancer Hypoxia contributes to tumor angiogenesis. When the oxygen requirement of the tumor exceeds the oxygen supply capacity of the tissue, the reduced oxygen concentration activates the hypoxia inducible factor (HIF) (Muz et al. 2015). HIF is the major mediator of cell response to hypoxia. The activation of HIF can induce many metabolic changes under hypoxic conditions, of which the most important is the induction of glycolysis and inhibition of oxidative phosphorylation of mitochondria (Huang et al. 2014; Fan et al. 2013). Hypoxia dependent lipid metabolism in lung cacner has not been studied yet. Cell hypoxia is the substrate that uses glutamine as a source of lipid synthesis. When HIF2 is activated in liver cells, it organizes lipid production and beta oxidation of FA, so that lipids are stored and thus fatty liver is produced. In order to survive in a hypoxic environment, tumor cells retain part of beta oxidation to produce CPTIC (a isoform of carnitine transferase that is required for long chain A transport to degrade energy supply in mitochondria) (Zaugg et al. 2011).

2  The Role of Lipid Metabolism in the Development of Lung Cancer

15

In addition to regulating FA biosynthesis, hypoxia also changes the lipid composition of tumor cells by participating in lipid synthesis and modification pathways and lipid uptake. HIF1αregulates lipid uptake only with peroxisome proliferator activated receptor γ (PPARγ), and affects cardiac hypertrophy via the glycerol phosphate pathway. Hypoxic tumor cells can ingest lysophosphatidic acid to meet its needs for unsaturated FA (Kamphorst et al. 2013). Lipid modification is also affected by hypoxia. Intermittent hypoxia can induce the expression of SCDl and SCD2 in mice. In renal cancer, SCDl (the major isoform of SCD in the body) can be induced by hypoxia and regulated by the positive feedback regulation of the AKT pathway, regulating the expression of HIF2α (Zhang et al. 2013). Changes in lipid expression through SCD eventually affect cell membrane fluidity and induce tumor cell migration. The migration of tumor cells is also regulated by hypoxia. FA desaturation also affects lipid signaling between tumor cells and the vascular system.

2.9  Lipid Metabolism Drugs for Lung Cancer Pharmacological inhibition of lipid synthesis is focused on inhibition of FASN. FASN inhibitors have been shown to be effective in transgenic mice with breast tumors in HER2/neu and can effectively reduce the incidence of chemically induced lung cancer.FASN inhibitors, such as TVB-3166, C75, C93 (Table 2.1) (Ventura et al. 2015; Fu et al. 2015; Relat et al. 2012). TVB-3166 can effectively and selectively induce lung cancer cell death by disrupting lipid raft architecture, inhibiting signaling pathways and oncogenic expression, but has no significant effect on normal cells (Ventura et  al. 2015). Other enzymes in lipid synthesis have also been shown to inhibit tumor cell growth and proliferation, including ACC and SCD (Nashed et al. 2012; Hess et al. 2010). The knockdown of ACLY could block the growth of tumor cells both in vivo and in vitro. ACLY inhibitors (Shi et al. 2016), i.e. SB-204990, could reduce hepatic cholesterol and FA synthesis rates and reduce the formation of xenograft lung cancer and prostate tumors. ACLY is not only essential for the synthesis of FA, but also an essential regulator of histone acetylation. Therefore, it plays a bridging role in cell metabolism and gene expression. There is evidence that increased cholesterol biosynthesis contributes to the development of prostate tumors. The dys-regulation of the alpha hydroxy acid pathway promotes the Table 2.1  Agents modifying lipid metabolism with therapeutic potential in lung carcinoma

Drug TVB-3166 C93 C75 SB-204990 CVT-11127 CP-640186

Action FAS inhibitor FAS inhibitor FAS inhibitor ACLY inhibitor SCD inhibitor ACC inhibitor

References Richard Ventura et al. (2015) Fu et al. (2015) Relat et al. (2012) Shi et al. (2016) Nashed et al. (2012) Hess et al. (2010)

16

L. Wang et al.

transformation of primary mouse embryonic cells into fibroblasts. Statins are an effective inhibitor of hydroxymethyl two coenzyme A reductase (HMGCR) and are routinely used to lower cholesterol levels to prevent cardiovascular disease.

2.10  Conclusion The biosynthesis, uptake and modification of lipids not only affect the proliferation and survival of tumor cells, but also affect the migration, invasion and tumor angiogenesis of tumor cells through more complex pathways. Lipid metabolism and lung cancer is an interesting situation, and potential biomarkers are involved in the occurrence, prognosis, prevention and treatment of lung cancer. Understanding the biological basis of lung cancer and lipid metabolism helps to develop new biomarkers and effective therapeutic strategies.

Reference Aragane K, Tamai Y, Kusunoki J. Hepatic de novo lipogenesis is present in liver-specific ACC1-­ deficient mice. Mol Cell Biol. 2007;27(5):1881–8. Blandin Knight S, Crosbie PA, Balata H, Chudziak J, Hussell T, Dive C. Progress and prospects of early detection in lung cancer. Open Biol. 2017;7(9):170070. Brown M, Roulson J-A, Hart CA, Tawadros T, Clarke NW. Arachidonic acid induction of Rho-­ mediated transendothelial migration in prostate cancer. Br J Cancer. 2014;110(8):2099–108. Byagowi S, NaserpourFarivar T, Najafipour R. Effect of PPARd agonist on stearoyl-CoA desaturase 1  in human pancreatic cancer cells: role of MEK/ERK1/2 pathway. Can J  Diabetes. 2015;39(2):123–7. Carracedo A, Cantley LC, Pandolfi PP. Cancer metabolism: fatty acid oxidation in the limelight. Nat Rev Cancer. 2013;13(4):227–32. Chen W-C, Wang C-Y, Hung Y-H, Weng T-Y, Yen M-C, Lai M-D. Systematic analysis of gene expression alterations and clinical outcomes for long-chain acyl-coenzyme A synthetase family in cancer. Shridhar V, ed. PLoS One. 2016;11(5):e0155660. Corti F, Finetti F, Ziche M, Simons M. The syndecan-4/protein kinase Cα pathway mediates prostaglandin E2-induced extracellular regulated kinase (ERK) activation in endothelial cells and angiogenesis in vivo. J Biol Chem. 2013;288(18):12712–21. Csanadi A, Kayser C, Donauer M. Prognostic value of malic enzyme and ATP-citrate lyase in non-­ small cell lung cancer of the young and the elderly. PLoS One. 2015;10(5):e0126357. Daniëls VW, Smans K, Royaux I, Chypre M, Swinnen JV, Zaidi N.  Cancer cells differentially activate and thrive on de novo lipid synthesis pathways in a low-lipid environment. Lobaccaro J-MA, ed. PLoS One. 2014; 9(9):e106913. Dayal R, Singh A, Pandey A, Mishra KP. Reactive oxygen species as mediator of tumor radiosensitivity. J Cancer Res Ther. 2014;10(4):811–8. Eichmann TO, Lass A. DAG tales: the multiple faces of diacylglycerol—stereochemistry, metabolism, and signaling. Cell Mol Life Sci, CMLS. 2015;72(20):3931–52. Fan J, Kamphorst JJ, Mathew R, et  al. Glutamine-driven oxidative phosphorylation is a major ATP source in transformed mammalian cells in both normoxia and hypoxia. Mol Syst Biol. 2013;9:712.

2  The Role of Lipid Metabolism in the Development of Lung Cancer

17

Foster DA.  Phosphatidic acid and lipid sensing by mTOR.  Trends Endocrinol Metab. 2013;24(6):272–8. Fu QF, Liu Y, Fan Y, Hua SN, Qu HY, Dong SW, et al. Alpha-enolase promotes cell glycolysis, growth, migration, and invasion in non-small cell lung cancer through FAK-mediated PI3K/ AKT pathway. J Hematol Oncol. 2015;8:22. Goldkorn T, Chung S, Filosto S. Lung cancer and lung injury: the dual role of ceramide. Handb Exp Pharmacol. 2013;216:93–113. Gómez de Cedrón M, Ramírez de Molina A.  Microtargeting cancer metabolism: opening new therapeutic windows based on lipid metabolism. J Lipid Res. 2016;57(2):193–206. Guo JY, Karsli-Uzunbas G, Mathew R. Autophagy suppresses progression of K-ras-induced lung tumors to oncocytomas and maintains lipid homeostasis. Genes Dev. 2013;27(13):1447–61. Hess D, Chisholm JW, Igal RA. Inhibition of stearoylCoA desaturase activity blocks cell cycle progression and induces programmed cell death in lung cancer cells. PLoS One. 2010;5(6):e11394. Hopperton KE, Duncan RE, Bazinet RP, Archer MC. Fatty acid synthase plays a role in cancer metabolism beyond providing fatty acids for phospholipid synthesis or sustaining elevations in glycolytic activity. Exp Cell Res. 2014;320(2):302–10. Huang C, Freter C. Lipid metabolism, apoptosis and cancer therapy. Lemarié A, ed. Int J Mol Sci 2015; 16(1):924–949. Huang S-W, Kao J-K, Wu C-Y, et al. Targeting aerobic glycolysis and HIF-1α expression enhance Imiquimod-induced apoptosis in cancer cells. Oncotarget. 2014;5(5):1363–81. Huang J, Fan X-X, He J, et al. SCD1 is associated with tumor promotion, late stage and poor survival in lung adenocarcinoma. Oncotarget. 2016;7(26):39970–9. Jeon S-M, Chandel NS, Hay N. AMPK regulates NADPH homeostasis to promote tumour cell survival during energy stress. Nature. 2012;485(7400):661–5. Jerby L, Wolf L, Denkert C, et al. Metabolic associations of reduced proliferation and oxidative stress in advanced breast cancer. Cancer Res. 2012;72(22):5712–20. Jiang L, Wang H, Li J, et  al. Up-regulated FASN expression promotes Transcoelomic metastasis of ovarian cancer cell through epithelial-mesenchymal transition. Int J  Mol Sci. 2014;15(7):11539–54. Kamphorst JJ, Cross JR, Fan J. Hypoxic and Ras-transformed cells support growth by scavenging unsaturated fatty acids from lysophospholipids. Proc Natl Acad Sci U S A. 2013;110(22):8882–7. Knobloch M, Braun SM, Zurkirchen L, et al. Metabolic control of adult neural stem cell activity by Fasn-dependent lipogenesis. Nature. 2013;493(7431):226–30. Knoch F, Tarantola M, Bodenschatz E, Rappel W-J. Modeling self-organized spatio-temporal patterns of PIP3 and PTEN during spontaneous cell polarization. Phys Biol. 2014;11(4):046002. Kopf T, Schaefer H-L, Troetzmueller M, et al. Influence of fenofibrate treatment on triacylglycerides, diacylglycerides and fatty acids in fructose fed rats. Oresic M, ed. PLoS One. 2014; 9(9):e106849. Książek M, Chacińska M, Chabowski A, Baranowski M. Sources, metabolism, and regulation of circulating sphingosine-1-phosphate. J Lipid Res. 2015;56(7):1271–81. Kuhajda FP, Jenner K, Wood FD, et al. Fatty acid synthesis: a potential selective target for antineoplastic therapy. Proc Natl Acad Sci U S A. 1994;91(14):6379–83. Lin R, Tao R, Gao X, et al. Acetylation stabilizes ATP-citrate lyase to promote lipid biosynthesis and tumor growth. Mol Cell. 2013;51(4):506–18. Medes G, Thomas A, Weinhouse S. A study of lipid synthesis in neoplastic tissue slices in vitro. Cancer Res. 1953;13:27–9. Menendez JA, Lupu R. Fatty acid synthase regulates estrogen receptor-α signaling in breast cancer cells. Oncogene. 2017;6(2):e299. Mojumdar EH, Helder RW, Gooris GS, Bouwstra JA. Monounsaturated fatty acids reduce the barrier of stratum corneum lipid membranes by enhancing the formation of a hexagonal lateral packing. Langmuir. 2014;30(22):6534–43. Montagne K, Uchiyama H, Furukawa KS, Ushida T.  Hydrostatic pressure decreases membrane fluidity and lipid desaturase expression in chondrocyte progenitor cells. J  Biomech. 2014;47(2):354–9.

18

L. Wang et al.

Muro E, Atilla-Gokcumen GE, Eggert US. Lipids in cell biology: how can we understand them better? Bement W, ed. Mol Biol Cell 2014; 25(12):1819–1823. Muz B, de la Puente P, Azab F, Azab AK. The role of hypoxia in cancer progression, angiogenesis, metastasis, and resistance to therapy. Hypoxia. 2015;3:83–92. Nashed M, Chisholm JW, Igal RA.  Stearoyl-CoA desaturase activity modulates the activation of epidermal growth factor receptor in human lung cancer cells. Exp Biol Med (Maywood). 2012;237(9):1007–17. O’Neill HM, Lally JS, Galic S, et al. AMPK phosphorylation of ACC2 is required for skeletal muscle fatty acid oxidation and insulin sensitivity in mice. Diabetologia. 2014;57(8):1693–702. Padanad MS, Konstantinidou G, Venkateswaran N, et al. Fatty acid oxidation mediated by Acyl-­ CoA synthetase long chain 3 is required for mutant KRAS lung tumorigenesis. Cell Rep. 2016;16(6):1614–28. Parsons JB, Rock CO. Bacterial lipids: metabolism and membrane homeostasis. Prog Lipid Res. 2013;52(3):249–76. Pisanu ME, Noto A, De Vitis C, et al. Blockade of Stearoyl-CoA-desaturase 1 activity reverts resistance to cisplatin in lung cancer stem cells. Cancer Lett. 2017;406:93–104. Prior AM, Zhang M, Blakeman N, et  al. Inhibition of long chain fatty Acyl-CoA Synthetase (ACSL) and ischemia reperfusion injury. Bioorg Med Chem Lett. 2014;24(4):1057–61. Qiang L, Kon N, Zhao W, et al. Hepatic SirT1-dependent gain-of-function of stearoyl-CoA desaturase-­1 conveys dysmetabolic and tumor progression functions. Cell Rep. 2015;11(11):1797–808. Relat J, Blancafort A, Oliveras G. Different fatty acid metabolism effects of (−)-epigallocatechin-­ 3-­gallate and C75 in adenocarcinoma lung cancer. BMC Cancer. 2012;12:280. Riaz A, Huang Y, Johansson S. G-protein-coupled lysophosphatidic acid receptors and their regulation of AKT signaling. Van Craenenbroeck K, Tikkanen R, eds. Int J Mol Sci. 2016;17(2):215. Sampath H, Ntambi JM.  Role of stearoyl-CoA desaturase-1  in skin integrity and whole body energy balance. J Biol Chem. 2014;289(5):2482–8. Schuster B, Sleytr UB.  Biomimetic interfaces based on S-layer proteins, lipid membranes and functional biomolecules. J R Soc Interface. 2014;11(96):20140232. Shi J, Yang H, Duan X, Li L, Sun L, Li Q, et al. Apolipoproteins as differentiating and predictive markers for assessing clinical outcomes in patients with small cell lung Cancer. Yonsei Med J. 2016;57(3):549–56. Simopoulos AP. An increase in the Omega-6/Omega-3 fatty acid ratio increases the risk for obesity. Nutrients. 2016;8(3):128. Sounni NE, Cimino J, Blacher S, et al. Blocking lipid synthesis overcomes tumor regrowth and metastasis after antiangiogenic therapy withdrawal. Cell Metab. 2014;20(2):280–94. Svensson RU, Parker SJ, Eichner LJ, et al. Inhibition of acetyl-CoA carboxylase suppresses fatty acid synthesis and tumor growth of non-small cell lung cancer in preclinical models. Nat Med. 2016;22(10):1108–19. Ventura R, Mordec K, Waszczuk J. Inhibition of de novo palmitate synthesis by fatty acid synthase induces apoptosis in tumor cells by remodeling cell membranes, inhibiting signaling pathways, and reprogramming gene expression. EBioMedicine. 2015;2(8):808–24. Wang Y, Zhang X, Tan W, Fu J, Zhang W. Significance of fatty acid synthase expression in non-­ small cell lung cancer. Zhonghuazhongliuzazhi [Chin J Oncol]. 2002;24(3):271–3. White E.  Deconvoluting the context-dependent role for autophagy in cancer. Nat Rev Cancer. 2012;12(6):401–10. Yamanaka M, Tian Z, Darvish-Ghane S, Zhuo M. Pre-LTP requires extracellular signal-regulated kinase in the ACC. Mol Pain. 2016;12:174480691664737. Yamashita T, Kamada H, Kanasaki S, et al. Epidermal growth factor receptor localized to exosome membranes as a possible biomarker for lung cancer diagnosis. Pharmazie. 2013;68(12):969–73. Yan S, Yang X-F, Liu H-L, Fu N, Ouyang Y, Qing K. Long-chain acyl-CoA synthetase in fatty acid metabolism involved in liver and other diseases: an update. World J Gastroenterol, WJG. 2015;21(12):3492–8.

2  The Role of Lipid Metabolism in the Development of Lung Cancer

19

Zaidi N, Swinnen JV, Smans K. ATP-citrate lyase: a key player in cancer metabolism. Cancer Res. 2012;72:3709–14. Zaugg K, Yao Y, Reilly PT, et  al. Carnitinepalmitoyltransferase 1C promotes cell survival and tumor growth under conditions of metabolic stress. Genes Dev. 2011;25(10):1041–51. Zhang Y, Wang H, Zhang J, et al. Positive feedback loop and synergistic effects between hypoxia-­ inducible factor-2α and stearoyl-CoA desaturase-1 promote tumorigenesis in clear cell renal cell carcinoma. Cancer Sci. 2013;104(4):416–22. Zhang J, Song F, Zhao X, et al. EGFR modulates monounsaturated fatty acid synthesis through phosphorylation of SCD1 in lung cancer. Mol Cancer. 2017;16:127. Zordoky BN, Nagendran J, Pulinilkunnil T. AMPK-dependent inhibitory phosphorylation of ACC is not essential for maintaining myocardial fatty acid oxidation. Circ Res. 2014;115(5):518–24.

Chapter 3

Bioinformatics of Embryonic Exposures: Lipid Metabolism and Gender as  Biomedical Variables K. K. Linask

Abstract  Early pregnancy in the first month is a highly vulnerable window for adverse effects of environmental exposures on the developing embryo. We have demonstrated that three seemingly disparate factors, lithium, homocysteine or alcohol, induce cardiac outflow tract defects, when a single exposure occurs during gastrulation stages. Severity of defects relates to dose, timing of exposure during gastrulation, and gender of the embryo. We sought to define what common process in the developing heart may be altered by maternal lithium and homocysteine exposure and is protected by folic acid dietary supplementation. Using microarray studies and bioinformatics analyses, lipid metabolism was predominantly altered with male embryos displaying greater misexpression of genes than the female. Both placental and cardiac lipid metabolism was altered in a sex-dependent manner.

3.1  Introduction 3.1.1  Development Heart development during embryogenesis begins during the period of gastrulation. Gastrulation encompasses the formation of three germ layers, the ectoderm, mesoderm, and endoderm. These germ layers give rise to all of the tissues and organs of the body. The heart is the first organ to develop from the mesoderm layer beginning within the second week after conception. By 3 weeks after fertilization, the heart is a tubular, beating structure. By 5–6  weeks after fertilization, the heart is a four-­ chambered structure that is circulating blood with the necessary nutrients and oxygen from the placenta to the embryo and removing metabolic waste products from the embryo back to the placenta. The coordination of cardiogenesis and the differentiation of trophoblasts extraembryonically that contributes to placental K. K. Linask (*) Department of Pediatrics, USF Morsani College of Medicine, Tampa and St. Petersburg, FL, USA e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2018 X. Wang et al. (eds.), Lipidomics in Health & Disease, Translational Bioinformatics 14, https://doi.org/10.1007/978-981-13-0620-4_3

21

22

K. K. Linask

development take place during this same time frame. This early period of pregnancy encompasses important developmental processes in heart and placental development that occur before a woman usually recognizes her pregnancy at approximately 5–6 weeks of gestation. Therefore, usually no precautionary measures to decrease embryonic exposures to adverse factors are yet being taken by the mother. In the context that 49% of pregnancies are unplanned (Walker et al. 2005), the early pregnancy time frame before 5–6 weeks is a period of high risk for the embryo to be exposed to environmental factors that can induce birth defects. Defining the early processes and pathways affected by embryonic exposures is of especial importance, because cardiovascular malformations are the most common congenital anomalies, involving 4–6 per 1000 livebirths (Ferencz 1990), and those involving valve and associated structures are the most common subtype, accounting for 25–30% of all cardiovascular malformations (Loffredo 2000). These percentages seemingly reflect the early time period that environmental agents can adversely affect cardiovascular development.

3.1.2  Environmental Exposures and Cardiovascular Defects My research using mouse and chick embryos has demonstrated the sensitivity of the vertebrate embryo to exposures during gastrulation to environmental factors and the induction of a high incidence of cardiac defects. Specifically, we addressed the effects on heart development of lithium, a drug used therapeutically for bipolar disorder, of an elevation of the natural metabolite homocysteine that occurs with nutrient deficiency and is part of the folic acid cycle, and of excessive alcohol ingestion known to induce Fetal Alcohol Syndrome. All three factors in the human population have been shown to increase the incidence of cardiac defects above normal. Since 1970, lithium (Li) has been therapeutically used as a mood stabilizer for bipolar disorder, a disease occurring during child-bearing years. Li was shown to mimic Wnt/ß-catenin signaling (Klein and Melton 1996; Zhang et al. 2003) by decreasing the activity of glycogen synthase kinase-3 (GSK3) and augmenting canonical Wnt signaling (Manisastry et al. 2006). Early adverse effects produced embryonic lethality. During human pregnancy, Li is associated with an increase in tricuspid valve defects such as Ebstein’s anomaly. Ebstein’s anomaly in the human is among the most diverse in presentation and severity (Gurvitz and Stout 2007). In its most severe form, it is lethal. An elevation of homocysteine (HCy) in the maternal plasma in the human population results from low dietary folate (folic acid, FA) or from mutations in methylene- tetrahydrofolate reductase (MTHFR). Elevated HCy increases the risk for neural tube, neural crest, craniofacial and congenital heart defects in the offspring (Boot et al. 2003; Huhta et al. 2006; Rosenquist et al. 1996; Tang et al. 2004). The association with elevated HCy is particularly strong for specific outflow tract defects of the heart (Huhta et al. 2006). In the folic acid cycle

3  Bioinformatics of Embryonic Exposures

23

HCy is a substrate for the synthesis of the primary methyl donor S-adenosylmethionine (SAM). Changes in SAM synthesis can have important effects on epigenetic regulation of gene expression. Fetal alcohol syndrome (FAS) is characterized by cardiac defects, fetal growth restriction, neurodevelopmental delays, and craniofacial malformations. In a recent epidemiologic study, periconceptional alcohol use was associated with cardiac birth defects, specifically outflow tract defects and transposition of the great arteries (Grewal et al. 2008). The FAS-related outflow tract congenital birth defects with alcohol exposure are similar to those that we reported in the mouse where embryos in utero were exposed acutely to lithium or to an elevated dose of homocysteine (Han et al. 2009; Serrano et al. 2010). Using the mouse model and noninvasive Doppler ultrasound monitoring of blood flow at mid-gestation on embryonic day (ED)15.5, it became particularly noteworthy that when the early mouse embryo was exposed acutely to the environmental factors on E6.75, similar cardiac defects in a high percentage of the embryos were induced. We found using echocardiography that by the acute intraperitoneal injection of the pregnant dam in this specific narrow window during gastrulation, defects of the outflow tract were primarily observed (Chen et  al. 2008; Han et  al. 2009; Serrano et al. 2010). Normal placental blood flow was also disrupted as detected by ultrasound. Additionally when analyzing the behavior of human trophoblast cell lines in culture exposed to the same three factors, the trophoblast cells displayed abnormal cell migration and protein misexpression (Han et  al. 2012). Thus, we detected commonalities in the abnormal molecular, biochemical, and blood flow biophysical changes that were being induced in the mouse embryos and placentas just by a single exposure on E6.75 to three different environmental factors.

3.1.3  Prevention by Folic Acid Importantly, with all three exposures of the embryo to Li, HCy or to alcohol, the adverse effects could be prevented with maternal dietary 10 mg folic acid supplementation initiated on the morning after conception (Han et al. 2009; Serrano et al. 2010). This prevention suggested that a similar common pathway or process was impacted, the effects of which could be prevented by the high dose folic acid. Certain common perturbations that were prevented in previous studies by folic acid supplementation were the delay in expression of key genes that are activated during heart development, such as Hex and Islet-1 and the misexpression of the growth factor canonical Wnt activity in the heart and valve regions, as well as in the developing placenta (Chen et al. 2008; Han et al. 2012). These genes and Wnt growth factor activity are key components of early heart development that on their own, when misexpressed, can lead to cardiac defects. Subsequently, we asked are these the predominant common signals that are altered with the environmental factor

24

K. K. Linask

exposures or is there still another predominant key aspect of developmental pathways that we should be considering and which can be prevented by folic acid? To answer this question, microarray studies were carried out using six experimental groups done in duplicate followed by bioinformatic analyses of the microarray data.

3.2  E  xperimental Background Leading to Bioinformatic Analyses 3.2.1  Mice The inbred C57BL/6J pregnant mice were used in our studies. It is recognized that mouse strains differ in their response to external factors (Andrikopoulos et al. 2005; Barnabei et al. 2010; Funkat et al. 2004). We used a popular strain that we had also used in our previous studies. The presence of a vaginal plug in the morning following conception is taken as evidence of mating and is designated in my lab as embryonic day 0.5 (E0.5) of gestation.

3.2.2  Control Folic Acid and High Folic Acid Diets The mouse diets were prepared by Harlan Laboratories (Madison, WI). The control diet contains 3.3 mg/kg as the baseline folic acid concentration to maintain health of the mouse and embryos. This is an amount that does not prevent cardiac defects. The “high” folic acid-supplemented mouse diet contains 10.5 mg/kg folic acid and prevented the noted defects. Folic acid, 7.2 mg/kg, was added to the base of 3.3 mg/kg in the baseline diet for a total of 10.5 mg/kg. This level of folic acid represents a supplement of 460 ng/g BW (body weight) taking into consideration metabolic body size (Han et al. 2009). This 10.5 mg/kg folic acid concentration was based on a comparable amount used in human population clinical studies in Hungary (Czeizel et al. 2013).

3.2.3  Experimental Groups Because much of our previous published data related to effects of embryonic exposure to elevated HCy or to Li+, the microarray studies addressed these two factors, with and without folate supplementation, in comparison to control mouse embryos. Treated pregnant mice on E6.75 were randomly assigned to receive an intraperitoneal (i.p.) injection of either a single dose of lithium chloride (Li+), as previously determined (Chen et al. 2008), or of HCy [(Han et al. 2009). Control mice received

3  Bioinformatics of Embryonic Exposures

25

i.p. injections of physiological saline (0.9% NaCl). An i.p. injection at 17:30 h on E6 was determined to be optimal for inducing outflow tract defects and was designated E6.75. On E0.5, all pregnant mice were placed either on the baseline folic acid 3.3 mg/kg chow or the high dietary folic acid chow of 10.5 mg/kg. All experimental and control animals were then maintained on the defined diets throughout the study. On E15.5, the heart and utero-placental circulations of the embryos were examined noninvasively in utero using Doppler ultrasonography (echocardiography) (Gui et al. 1996). As previously published, the Li+/HCy i.p. injections of dams on maintenance folic acid diet induced similar cardiac valve defects and altered myocardial and umbilical artery blood flow (Gui et al. 1996; Han et al. 2009). E15.5 embryos with abnormal cardiac echo patterns were chosen for the microarray analysis to compare with the control embryos displaying normal blood flow (Han et al. 2009). Detailed analyses and discussion of the abnormal ultrasound patterns can be found in our previous publications (Han et al. 2009; Linask et al. 2014). Six pregnant C57BL/6J mice were used for the microarray analysis. Control embryos and folic acid-protected embryos displaying normal heart function and placental blood flow, and the Li+/HCy exposed embryos displaying semilunar valve regurgitation and abnormal blood flow patterns, were isolated on E15.5. Two different embryos from the same dam’s litter served as duplicate samples for each control and experimental treatment groups; therefore 6 different groups (litters), i.e., a total of 12 embryos were analyzed. Genotyping of embryos by RT-PCR for gender was not done for the microarray analysis at this point, because initially we were not considering gender differences in response to exposures. Because the outflow and right ventricular regions were most affected with our timing of acute exposure, only these cardiac regions were microdissected as one piece and the total RNA was extracted. The left ventricles were not used. The samples were analyzed by microarrays for changes in gene expression on E15.5, after the acute exposures were administered on E6.75. In summary, the treatment groups included embryos that had received: (1) high dietary folic acid (FA) only (10.5 mg/kg body weight), with 0.9% saline by i.p., and had normal echos; (2) Li+ in saline by i.p. injection, maintenance diet FA (3.3 mg/kg body weight), and had abnormal echos; (3) Li+ in saline by i.p. injection, with high dietary FA supplementation, normal echos; (4) HCy in saline by i.p., maintenance FA diet, abnormal echos; (5) HCy in saline by i.p with high dietary FA supplementation, normal echos; or (6) control 0.9% saline by i.p., maintenance FA diet, with normal echos. The samples were processed in one batch. The detailed protocol has been published (Han et al. 2016).

3.2.4  Microarrays For hybridization we used the Affymetrix GeneChip Mouse Genome 430 2.0 arrays (Affymetrix, Inc., Santa Clara, CA). With this microarray an estimated 39,000 distinct transcripts are detected, including over 34,000 well-substantiated mouse genes.

26

K. K. Linask

Each gene is represented by a series of oligonucleotides that are identical to sequence in the gene and oligonucleotides that contain a homomeric (base transversion) mismatch at the central base position of the oligomer used for measuring cross hybridization. Scanned output files were visually inspected for hybridization artifacts and then analyzed using Affymetrix GeneChip Operating Software (GCOS). Signal intensity was scaled to an average intensity of 500 prior to comparison analysis. Using the default settings, GCOS software identified the increased and decreased genes between any two samples with a statistical algorithm that assesses the behavior of 11 different oligonucleotide probes designed to detect the same gene (Liu et al. 2002). Probe sets yielding a change in p-value less than 0.002 were identified as increased or decreased and those that yielded a p-value between 0.002 and 0.002667 were identified as marginally changed. A gene was identified as consistently changed if it was identified as changed in all replicate experiments by the software. Affymetrix CEL files were analyzed using MAS5 algorithm (Hubbell et al. 2002; Liu et al. 2002) of Expression Console Suite (Affymetrix).

3.3  Bioinformatic Analysis The bioinformatics analysis we performed was not conventional, but it has been used previously in the analysis, or for reanalysis, of clinical data (Ptitsyn 2009). Our data suggested that embryonic gender may be involved in the observed results. This meant, however, that when taking gender into account, we had only a single replicate per condition. Because of extreme noise between individual cancer patient samples, the approach described by Pitsyn in 2009, focused on Gene Ontology (GO) terms (or pathways, or other types of gene function groupings) rather than individual genes. This was found to clarify the noise. Our interest was to do a similar analysis with our individual mouse heart samples. Using this approach we were able to take into account gender and to compare the gene expression in male and female embryos in response to HCy and Li exposure, with and without folate supplementation, in comparison to control embryos. A flow chart of the bioinformatics analysis is provided in Fig. 3.1 and is described below. We had six treatment groups, including the control groups. We started our study not addressing gender, but only asking which dominant biological processes were being perturbed by our exposures and were protected with folic acid supplementation. Because developing embryos have been noted to have gender biases in gene expression (Gabory et al. 2009), the initial quality control of the data included gender determination using expression values for X-inactivation transcript Xist (probesets 1427262_at,1427263_at and 1436936_s_at; present in females, absent in males) and Y-chromosome linked gene Ddx3y (1426438_at, 1426439_at and 1452077_at; present in males; absent in females). This we later validated by genotyping the posterior parts of the embryos that had been picked for the analysis and had been kept frozen. Gender was confirmed as had been determined by the specified probesets.

27

3  Bioinformatics of Embryonic Exposures

A. SIX TREATMENT GROUPS 1 CONTROL Embryo #s/Group 2

2 HiFA 2

3 Li

4 HCy

5 LiFA

2

2

2

6HCyFA 2

B. INITIAL QUALITY CONTROL OF MICROARRAY DATA AND GENDER DETERMINATION

Used X-inactivation probesets present only in females (F) Used Y-linked Ddx3y gene probeset present only in males (M) M

F

M F

M

F

M F

F F

M M

C. PAIRWISE ANALYSIS OF THE 12 INDIVIDUAL SAMPLES

D. NON-PARAMETRIC ANALYSIS USING VLAD

E. LISTS OF OVER-REPRESENTED GENE ONTOLOGY (GO) NODES

F. DETERMINATION OF COMMON CATEGORIES LIST G. COMMON CATEGORIES SORTED ACCORDING TO GENDER Fig. 3.1  Flow Chart of the Bioinformatic Analysis

Using the probesets, the randomly chosen embryos represented 6 male and 6 female embryos: Control: 1 male 1 female; High Folate diet: 1 male 1 female; Li+ exposure: 1 male 1 female; HCy exposure: 1 male 1 female; Li+ exposure + High Folate diet: 0 male 2 female; and HCy exposure + High Folate diet: 2 male 0 female. Because gender differences in gene expression between the arrays were contributing the largest variance among the data, we planned a strategy for data analysis with a low number of replicates that was adapted from the analysis done in cancer

28

K. K. Linask

research (Ptitsyn 2009). By focusing on Gene Ontology terms rather than individual genes, this clarified the noise. To increase the sampling power of our microarray data, we tested dysregulation of expression in the groups of genes identified by the GO categories. For this, we performed pairwise analysis for all 12 samples individually in which each probeset was assigned as “present”, “downregulated”, “upregulated” or “inconsistent” depending on the ratio of the signals between samples, and consistency of these ratios among probesets representing the same gene. Next, non-parametric analysis of these data was carried out using Visual Annotation Display (VLAD) software that is available as a web-based application from the Mouse Genome Informatics (MGI) website. VLAD analysis tests for the enrichment of ontology terms in the set of genes submitted by a researcher and the software supports the analysis of multiple gene sets at once. It performs hypergeometric distribution of GO annotations for a given set of genes (i.e., up- or downregulated) in comparison to a “universal set of genes (i.e., all genes “present” in the samples) (Richardson and Bult 2015). A p value p 2 are indicative of the highly collinear variations of two major n3 fatty acids, EPA (20:5 n3) and DHEA (22:6 n3)

distribution, e.g. for vaccenic acid at D21. P-P plots compare the empirical distribution function of a sample with that of a sample distributed according to a normal distribution of the same mean and variance. If the sample follows a normal distribution, the points lie along the first bisector of the plan (Thode et al. 1983).

4.2.2  M  ultivariate Statistics and Modeling of the Serum Lipid Fatty Acid Composition From a close examination of the raw data a number of limitations for the statistical processing of data have emerged using parametric tests and multilinear modelling. A particular difficulty is that the number of observations is not carried through the 5 successive measurement protocol (D0 and D21, n = 96; at 3 M and 6 M, n = 94 and 12 M, n = 58) due to dropout patients. This probably results in a bias for later observations being overweighed for a beneficial outcome in the most compliant patients. As judged by the student t-test of D0-D21 paired values the mean values of FA marking DNL (Fig. 4.1a) vary significantly during residence. However the variations show opposite trends; shorter myristic (14:0, p = 0,007) and palmitoleic (16:1, p