Assay Development Enables Large-Scale Metabolic Profiling

February 1, 2018 (Vol. 38, No. 3)

Examining the Molecular Composition of Samples

To improve prevention, detection, and treatment of diseases, it is crucial to gain understanding of the complex interactions between human genes and the environment. These interactions create our metabolic phenotypes, which are reflective of disease risk.

Metabolic profiling provides critical insight and knowledge into the interrelationships between an individual’s phenome and external influences, and has the potential to transform disease states and permit significant advances in medical treatments.

The Potential of Metabolic Profiling

Funded by the Medical Research Council (MRC) and National Institute for Health Research (NIHR), the MRC-NIHR National Phenome Centre (NPC) is located at Imperial College London, where it provides bioanalytical and metabolomics- and bioinformatics-related client-focused research solutions to scientists and innovators. The NPC offers the United Kingdom’s research and translational medicine community, both academic and industrial, a range of services including high-precision untargeted profiling and bespoke quantitative assays.

The NPC employs quadrupole time-of-flight (Q-tof) mass spectrometry instruments that are used exclusively for discovery, and tandem quadrupole instruments that are used for targeted assays and validation of what is found in the NPC’s discovery platform. This enables high-throughput analysis and a high volume of research, with the focus on understanding the mechanisms of disease, the environmental impacts on metabolism, and the underpinnings of human biochemistry on a population scale. The NPC has developed the technology and expertise to provide metabolite analysis for a single biochemical to broad metabolite profiling using multiple assays to cover the diverse physiochemical properties of metabolites.

Incomplete Metabolome Coverage

A core challenge faced by the NPC was that no one technology platform or assay could cover the metabolome thoroughly enough to provide useful insights into the complexity of human biochemistry.

To address this challenge, the NPC has developed and implemented a complementary mix of two analytical technologies to maximize metabolome coverage and refine the technological platform underpinning its assays. Working with Waters and using its Ultra-Performance Liquid Chromatography^® mass spectrometry (UPLC^®-MS) technology,¹ and working with Bruker BioSpin and its nuclear magnetic-resonance (NMR) spectroscopy instrumentation, the NPC can provide unparalleled insight into the molecular composition of samples.

As part of a two-stage process, after initial NMR screening, UPLC-MS profiling provides deep metabolic coverage, encompassing quantification or relative quantification of thousands of individual metabolites and complex lipids. This is possible because UPLC-MS provides multidimensional high-resolution separations and sensitive detection across a broad range of chemical species.

Early work that pushed the technique with larger batches and large patient cohorts encountered several difficulties: batch-to-batch or study-to-study variation, unacceptably high in-run coefficients of variance (CVs), run order effects, and lack of reproducibility. In addition, the technique needed to be adapted to a research environment in which scientists were focused on extracting the highest density of data and achieving the ultimate in sensitivity.

The original analytical technique was often incompatible with the reproducible collection of very large datasets. It soon became evident that the technique would have to be optimized if it were to help the NPC achieve its goal of building and deriving value from epidemiological-scale metabolic data sets. 

Large-Scale Metabolic Profiling

To achieve this goal, the NPC worked to ensure that the UPLC-MS systems could sustain 24/7 operations with little intervention and without running out of samples. The focus was on reproducible and robust methodology, rather than on ultimate per-sample analytical performance and sensitivity.

The collaboration between Waters and the NPC enabled optimization of the instrument configuration and development of an automatic adjustment capability, allowing for continuous analysis of large sample sets, ultimately delivering unprecedented raw data precision at such a large scale (Figure 1).

Running 80 study samples per instrument per day (excluding quality control samples) with 10 profiling instruments means that the NPC can perform 800 assays in a single day. This may be 800 individual samples by a single assay, or 200 samples multiplied by four different assay types. Running the platform in such a parallel configuration enables the NPC to quickly build comprehensive molecular coverage of sample sets.

In the profiling analyses, typically the NPC looks at relative quantification, scrutinizing the relative differences between people, between groups of people, or among larger sets of groups. The specific, absolute quantification is completed in a targeted validation stage if the analyses are successful in their discovery. The Q-tof platform, which serves as a wide-open screening tool, uses a combination of chromatographic assays that provides as much metabolic coverage as possible. From this comprehensive information, the NPC designs specific targeted assays that can provide greater linearity, sensitivity, and specificity, and that can, in many instances, add absolute quantification.

The platform is managed in such a way that each instrument is dedicated to a specific analysis type on a specific biofluid. Each instrument runs in parallel, which means that when a sample set comes into the NPC, all assays can be performed simultaneously for fast turnaround of data (Figure 2).^1–3

Figure 1. The NPC has incorporated Waters instrumentation into preparative, analytical, and data processing procedures for a stratified medicine application: the metabolic profiling of urine by UPLC-MS.

Looking to the Future

The NPC has made fundamental achievements so far. Standardized high-throughput methodology with unprecedented reproducibility at scale has generated very large datasets, which have already produced new insights into what had been considered well-characterized diseases. Such insight has permitted a composite image of what human health looks like in a normal control population and in people with various diseases.

Moving from data to clinically actionable information continues to be the ultimate goal. A new focus on real-time data processing, as well as harmonization across studies, will rapidly accelerate the advance and value of metabolic profiling in healthcare.

Figure 2. The NPC has developed a fit-for-purpose Ultra-Performance Liquid Chromatography® mass spectrometry (UPLC®-MS) technology, method set, and sample analysis workflow. Working together, these components can sustain analysis on an industrial scale and allow batch-free operation for large studies.1,3

Matthew Lewis, Ph.D. ([email protected]), is chief operating officer at the MRC-NIHR National Phenome Centre (NPC), and director of metabolic profiling at Imperial College London. Dr. Lewis and colleagues at the NPC partnered with Waters to improve the metabolic coverage of HPLC-MS profiling. To find out more about solutions for metabolomics and metabolic profiling, visit www.waters.com/metabolomics.