The Handbook of Metabolic Phenotyping

John C. Lindon, Jeremy K. Nicholson, Elaine Holmes

• Publisher: Elsevier
• Year: 2018
• ISBN:
  9780128122945
• Topic: Sciences physiques
• Subtopic: Chimie analytique

The Handbook of Metabolic Phenotyping is the definitive work on the rapidly developing subject of metabolic phenotyping. It explores in detail the wide array of analytical chemistry and statistical modeling techniques used in the field, coupled with surveys of the various application areas in human development, nutrition, disease, therapy, and epidemiology to create a comprehensive exploration of the area of study. It covers recent studies that integrate the various -omics data sets to derive a systems biology view. It also addresses current issues on standardization, assay and statistics validation, and data storage and sharing. Written by experts with many years of practice in the field who pioneered many of the approaches widely used today, The Handbook of Metabolic Phenotyping is a valuable resource for postgrads and research scientists studying and furthering the field of metabolomics.

• Contains theoretical and practical explanations of all the main analytical chemistry techniques used in metabolic phenotyping
• Explores, in detail, the many diverse statistical approaches used in the field
• Offers practical tips for successfully conducting metabolic phenotyping studies
• Features reviews of all of the various fields of activity relating to human studies

Chapter 1

An Overview of Metabolic Phenotyping and Its Role in Systems Biology

Elaine Holmes; Ian D. Wilson; John C. Lindon Department of Surgery and Cancer, Imperial College London, London, United Kingdom

Abstract

Understanding the functional relationships between genetic and gene expression variation and real-world biochemical and physiological end points in human, animal, or plant populations poses one of the key challenges in the postgenomic era. The advent of these postgenomic technologies has been facilitated by step changes in analytical technologies and advances in bioinformatics used to analyze and interpret the data generated. In this chapter, we describe the historical context of the development of the core analytical technologies and bioinformatics tools used in metabolic phenotyping and summarize the application fields enhanced by the use of such technologies.

Keywords

Metabonomics; MetabolomicsMetabolic phenotyping; NMR spectroscopy; Mass spectrometry; Biomarker; Disease; Drug toxicity; Personalized medicine; Pattern recognition

Chapter Outline

• 1 The History and Evolution of Metabolic Profiling
  • 1.1 The Early Years
  • 1.2 Modern Metabolic Profiling Using NMR Spectroscopy
  • 1.3 Modern Metabolic Profiling Using Mass Spectrometry
  • 1.4 Definition of the Field and Efforts in Standardization
  • 1.5 Development of Computational Technologies
• 2 Application Areas
  • 2.1 Drug Metabolism and Toxicology
  • 2.2 Plant-Based Metabolomics
  • 2.3 Human Disease
  • 2.4 Mechanistic Studies of Disease Models
  • 2.5 Predictive Metabolic Phenotyping (Pharmacometabonomics)
  • 2.6 Emerging Metabolic Profiling Technologies
• 3 Concluding Remarks
• References

1 The History and Evolution of Metabolic Profiling

1.1 The Early Years

The notion of using spectroscopic analysis to chemically characterize multicomponent signatures of biological samples has undergone several reincarnations, but the use of a multivariate-based urine diagnostic is a product of much older medical systems. In Ayurvedic medicine, the color, turbidity, and behavior of oil dropped into urine was used diagnostically to indicate the type and curability of the disease [1], while Sanskrit documents dating back to 100 BC describe 20 different kinds of urine in relation to disease [2]. In medieval times, the practice of examining a patient's urine and forming a diagnosis based on color, smell, and even taste was popular and the physicians who specialized in this practice were termed “pisse prophets.” Early documented examples of this are the two “urine wheels” shown in Fig. 1 [3]. In Sir Henry Wellcome's 1911 overview of the history of uroscopy, “The Evolution and Development of Urine Analysis,” he draws on Sanskrit medical texts describing urine that “is astringent, sweet, white and sharp.” He referred to urine from individuals with Diabetes mellitus and an “elephant” urine wherein the patient “continuously passes turbid urine like a mad elephant.” [4]. Today, urine color is still useful in the pathology laboratory. A recent photo of a rainbow spectrum of urine was published by Heather West highlighting the effect of factors such as dehydration (dark yellow), liver cancer (orange/brown), porphyria (blue), and kidney stones (red) [5], although several drugs and foods can also cause strange colors in urine.

Fig. 1 (Left) Image of a urine wheel taken from the book, identified as NKS 84 b folio, in the collection of the Royal Library Copenhagen Denmark; (right) an image of a urine wheel published in Ullrich Pinder's Epyphanie Medicorum, 1506. We acknowledge the Royal Library Copenhagen, Denmark, for providing these images.

The 20th century heralded a step change in the development of analytical techniques with use of more sophisticated analytical chemistry methods, in particular, the use of various forms of chromatography and electrophoresis. The field thus transitioned to a foundation in molecular-based identification of specific substances in body fluids with diagnostic properties for a range of diseases. Moreover, the realization that multiple parameters were more specific and held more diagnostic power than single clinical chemical measurements was beginning to dawn. One of the first examples of the application of multianalyte analysis to characterize biofluids originates from the work of scientists such as Dent and Dalgliesh who applied two-dimensional paper chromatography to separate a series of metabolites in urine, which were visualized using a variety of reagents, which they referred to as a “map of the spots” [6, 7] and which was used to differentiate inborn errors of metabolism such as Fanconi syndrome and cystinuria.

It has been argued that the birth of modern metabolic profiling was by Linus Pauling and colleagues who in 1971 deployed GC for the analysis the volatile components of urine and breath with a view to using the technique to detect signatures of disease [8]. They stated that “sickness in human beings, whether it be from malfunctions of organs and tissues or from invasions of viruses or microorganisms, would be reflected in qualitative and quantitative changes in the metabolites emitted.” In subsequent years, the continuation of this work led to the identification of around 250 and 280 components for breath and urine volatiles, respectively [9], and he and his colleagues went on to show quantitative sex-related differences in amino acid patterns in urine [10]. At around the same time as Pauling's GC profiling work, American scientists at the Oak Ridge National Laboratory also developed methods for untargeted metabolic phenotyping using ion-exchange liquid chromatography with capacity to separate several hundred metabolites [11, 12].

In 1978, Gates and Sweeley published an article on quantitative metabolic profiling based on gas chromatography [13] describing the GC-based quantitative metabolic profiling of volatile components of human biological fluids such as urinary organic acids. Their paper contains subsections entitled “Development of the concept of metabolic profiling” and “Disease diagnosis by metabolic profiling” citing Roger and Williams work in the late 1940s and early 1950s on the use of paper chromatography to differentiate the urine from individuals with different taste preferences. They noted that there was a high degree of interindividual variation but a low degree of intraindividual variation in the biochemical composition of urine and saliva [14]. This analysis used a multivariate star plot to represent each individual. Williams later applied this profiling method to show characteristic metabolic patterns for schizophrenia, alcoholics, and other neurobehavioral conditions. The article by Gates and Sweeley further expands on the need to consider the effect of outliers on biofluid reference ranges and the fact that metabolite distributions are often not normal. They summarize the development of a new field of statistics for simultaneously handling multiple measurements and refer to the work of Winkel who was the first to define a framework for applying multivariate statistics to metabolic profiling data [15].

Despite these substantial breakthroughs between the early 1950s and 1970s, there was very little uptake of the approach for a decade, with the exception of the study of inborn errors of metabolism where sporadic articles appeared using infrared spectroscopy, NMR spectroscopy, and GC-MS [16, 17]. One reason for the limited uptake of the technology was the fact that while the GC or LC methods provided peak fingerprints, relating peaks to specific potential “biomarkers” was difficult due to the lack of robust interfaces with instruments capable of providing structural information, such as mass spectrometers (particularly for LC-based methods).

After a decade or so, newer spectroscopic technologies began emerging as metabolic profiling tools and there was a resurgence of interest in the application of the technology to biomedical challenges as well as in the microbiology and plant science fields.

1.2 Modern Metabolic Profiling Using NMR Spectroscopy

The development of pulse-Fourier transform proton nuclear magnetic resonance (NMR) spectroscopy by the early 1980s opened the doors to robust metabolic screening of small molecules in biological fluids as well as revolutionizing the three-dimensional molecular structural characterization of proteins. Nicholson and Sadler published a series of studies applying the technology to chemically characterize the composition of biofluids such as urine and plasma in healthy humans and demonstrating the systematic alteration in metabolic composition in disease states [18, 19]. They went on to describe the metabolic fate and toxicity of drugs including acetaminophen using two-dimensional pulse sequences to improve metabolite identification [20].

In parallel, both GC-MS and NMR platforms were developed by Iles [21], Wevers [22], and Chalmers [23] for spectroscopic profiling of inborn errors of metabolism and led t...

Table of contents

• Cover image
• Title page
• Table of Contents
• Copyright
• Contributors
• Foreword
• Preface
• Chapter 1: An Overview of Metabolic Phenotyping and Its Role in Systems Biology
• Chapter 2: NMR Spectroscopy Methods in Metabolic Phenotyping
• Chapter 3: The Role of Ultra Performance Liquid Chromatography-Mass Spectrometry in Metabolic Phenotyping
• Chapter 4: GC-MS-Based Metabolic Phenotyping
• Chapter 5: Metabolic Phenotyping Using Capillary Electrophoresis Mass Spectrometry
• Chapter 6: Supercritical Fluid Chromatography for Metabolic Phenotyping: Potential and Applications
• Chapter 7: The iKnife: Development and Clinical Applications of Rapid Evaporative Ionization Mass Spectrometry
• Chapter 8: Univariate Statistical Modeling, Multiple Testing Correction, and Visualization in Metabolome-Wide Association Studies
• Chapter 9: Multivariate Statistical Methods for Metabolic Phenotyping
• Chapter 10: Data-Driven Visualizations in Metabolic Phenotyping
• Chapter 11: Big Data and Databases for Metabolic Phenotyping
• Chapter 12: Progress in Standardization of Metabolic Phenotyping Data
• Chapter 13: Conception, Implementation and Operation of Large-Scale Metabolic Phenotyping Centres: Phenome Centres
• Chapter 14: Applications of Metabolic Phenotyping in Pharmaceutical Research and Development
• Chapter 15: Metabolic Phenotyping in Nutrition Research
• Chapter 16: Metabolic Phenotyping in Clinical Practice
• Chapter 17: Applications of Metabolic Phenotyping in Epidemiology
• Chapter 18: Influence of the Human Gut Microbiome on the Metabolic Phenotype
• Chapter 19: Linking Metabolic Phenotyping and Genomic Information
• Chapter 20: Metabolic Phenotyping: History, Status, and Prospects
• Index