Capillary electrophoresis mass spectrometry based metabolomics

Since its first formal definition more than a decade ago, metabolomics, or, the comprehensive analysis of all metabolites present within a biological system, has attracted growing interest in clinical research by academia, industry and government labs. This is most prevalent in biomarker and drug development applications where a considerable change has been witnessed in how new diagnoses, prognoses, and therapeutic options are being discovered and developed using omic technologies. Moreover, many chronic diseases suggest a strong metabolic involvement or even a clear metabolic cause, including cancer. Together with the other omic disciplines, including genomics and proteomics, metabolomics plays a key role in the implementation of personalized medicine; evidence-based medicine designed for individually designed healthcare strategies. In turn, biomarker discovery and the understanding of biochemical pathways typically rely on a multimodal approach. Among these modalities, there continues to be a growing interest in CE-MS based development and implementation in clinical development.


Introduction
Since its first formal definition more than a decade ago, metabolomics, or, the comprehensive analysis of all metabolites present within a biological system, has attracted growing interest in clinical research by academia, industry and government labs.This is most prevalent in biomarker and drug development applications where a considerable change has been witnessed in how new diagnoses, prognoses, and therapeutic options are being discovered and developed using omic technologies.Moreover, many chronic diseases suggest a strong metabolic involvement or even a clear metabolic cause, including cancer.Together with the other omic disciplines, including genomics and proteomics, metabolomics plays a key role in the implementation of personalized medicine; evidence-based medicine designed for individually designed healthcare strategies.In turn, biomarker discovery and the understanding of biochemical pathways typically rely on a multimodal approach.Among these modalities, there continues to be a growing interest in CE-MS based development and implementation in clinical development.

Formulating the problem
Depending on how you classify the metabolome, there is a complex chemical space separated by hydrophilicity, polarity and size, excluding a broad range of metabolites classified as lipids, separately referred to as lipidomics.This class of metabolites also covers a large range of physical properties for which specifically designed platforms work the best.For example, for years liquid chromatography-mass spectrometry (LC-MS) has been used to capture a host of hydrophilic and hydrophobic metabolites, while gas chromatography-mass spectrometry (GC-MS) has been used to capture small molecular weight metabolites.For this review, the metabolome refers to endogenous molecules with a molecular weight typically less than 1000 Kd.To measure, catalogue, and compare the entirety of the metabolic space, the implementation of comprehensive mass spectral databases was needed (e.g., Human Metab-olome Database (HMDB), METLIN, MassBank, LIP-ID MAPS, LipidBlast, NIST 14).These databases have helped drive the field and enable untargeted discovery by CE-MS.The HMDB, for example, lists thousands of metabolites, their physical properties and, for many common metabolites, metabolite concentrations in various biological matrices (blood, urine, saliva, CSF).While METLIN reports structure and mass for thousands of metabolites, small peptides, and xenobiotic metabolites found across the plant and animal kingdom.Because of this diversity in physical properties and size of the metabolome in biological systems, there has been a need for the development of several analytical protocols based on different chromatographic methods to select specific subgroups of metabolites based on these different chemical properties beyond the technical capabilities of LC-MS and GC-MS.These new methods have their own advantages for selecting specific types of molecules: lipids, nucleotides, aminoacids or steroids.Nonmass spectrometric methods such as nuclear magnetic resonance (NMR) and non-chromatographic methods such as matrix-assisted laser desorption-time of flight (MALDI-TOF) imaging are successfully being used for selected metabolomic analyses.The focus of this review is recent publications that use CE-MS to extend the polar metabolome beyond what is observed by LC-MS and GC-MS.

Metabolomics -current methods
Using the appropriate technology can be a critical decision when starting a discovery program.Different methods or technical platforms have advantages within certain chemical spaces, and the choice of platform can effect linear dynamic range, lower level of quantitation, resolution of isomers, baseline biological noise and insource ion suppression.No universal methods exist so care should me made when making conclusions in evaluating unbiased and untargeted metabolomic data.With the diversity of metabolites captured by any one method, one can expect a range of different responses effecting dynamic range and linearity of response depending upon the specific metabolite properties.The choice of technology can affect the number of false positives, how to assess quality control and the type of statistical analysis in any metabolomics study.The most commonly used separation method uses high-pressure liquid chromatography (HPLC).HPLC methods vary to accommodate a broader range of metabolites.Ion pairing, ion exchange, hydrophilic interaction chromatography (HILIC) and reverse phase columns can all be used to select specific subtypes of metabolites.Capillary electrophoresis (CE) offers a novel approach with distinct advantages as seen in a growing list of publications.

CE mechanism
Understanding that HPLC and GC methods are not all inclusive, Smith et al. [1] demonstrated the potential of interfacing CE with mass spectrometry (MS) in 1987, although it was not until Soga et al. [2] published in 2002 that CE-MS could be performed with high reproducibility and sensitivity for biological applications.CE works using electrophoretic movement or electroosmotic flow (EOF) to govern transport and separation of metabolites.EOF is a phenomenon where the electrolyte or running buffer solution itself flows inside the capillary.The flow of the electrolyte solution is the main driving force that pushes samples into the mass spectrometer side of the capillary.A fused-silica capillary contains surface charges of silanol groups present on the inner walls.The silanol groups on the capillary inner wall are ionized presenting an overall negative surface.Opposing ions in the electrolyte solution are attracted to the inner wall surface to achieve a balance of electric charges, resulting in the formation of a double-layer with ionized silanol groups.Under these conditions, a potential difference is created very close to the inner wall.The application of a voltage to both ends of the capillary attracts the positively charged ions of the diffuse double-layer to an anode.In contrast, the silanol groups cannot move due to the fixation on the wall surface and the entire electrolyte solution in the capillary is directed toward the anode with the migration of the positively charged ions, thereby generating a flow.The degree of mobility of any compound relative to others is due to variations in their ionic radius and size and charge of the electrolyte filling the capillary.A compound or metabolite with a larger ionic radius and smaller charge would have limited mobility compared to small, more polar species.Compounds or metabolites with a smaller ionic radius and higher charge would have high mobility.Hence controlling the electrical gradient across the capillary and pH of the electrolyte solution are two of the most important parameters in controlling metabolite separation into the mass spectrometer.Hence CE-MS offers a totally different approach to metabolite separation prior to mass spectrometric detection.Earlier reviews go into more detail on how CE works, the theory of electroosmotic flow (EOF) and CE-MS interfaces [3,15,16,17,18].

Early problems
Despite the strong need to broaden the ability to measure the hydrophilic metabolic space, the usage of CE-MS has been relatively slow.Sensitivity, mass spectrometric stability, migration time reproducibility, and correction software for migration time shift have all contributed to slow uptake.In addition, limited commercial CE-MS solutions and the lack of CE-MS availability in core academic, government or industrial laboratories has limited development in the field.Since CE-MS may measure many new molecules, not in LC libraries, the creation of new small molecule libraries and the peak picking and warping software needed for CE alignment has also slowed progress.Lastly, the ability to provide a stable, sensitive, and reliable interface has been a critical issue to overcome.

Early developments
In 2002, Tomoyoshi Soga first developed a metabolome analysis method based on CE-MS, which enabled the simultaneous analysis of several thousand charged metabolites by cationic and anionic methods [1,7,11,14].Since then, CE-MS began to grow as one of the standard methods in bioscience research.See Soga's recent review in 2015 [18].Soga and other labs have clearly demonstrated that most of these early issues and concerns have been overcome by technological improvements, hence the growing list of publications in this field since 2002.CE-MS protocols have been developed to support metabolic measurements, both relative and quantitative, in a large variety of sample types including urine [4,5,8], bacteria [6], CSF [9], neurons [10] and brain [12].
Next-generation capillary electrophoresis-mass spectrometry approaches in metabolomics.
Recent applications of CE-and HPLC-MS in the analysis of human fluids.
Capillary electrophoresis mass spectrometry as a tool for untargeted metabolomics 2017 [116] Maier TV, Schmitt-Kopplin P. Capillary Electrophoresis in Metabolomics 2016 [117] BUKO AM J. APPL.BIOANAL ods.CE has also been favourably compared to HILIC separations offering higher flexibility, lower complexity, better RT reproducibility, higher peak capacity and better peak shape [17].Another challenge to the popularity of CE-MS has been the effect of high salt concentrations on migration time reproducibility.High salt concentrations in the sample can lead to stacking and peak spreading affecting metabolite resolution, identification and mass accuracy.However, this issue can be avoided by determination of an appropriate dilution ratio for general types of samples (blood, tissue, urine, saliva etc.).Alternatively, some sample types may require a pre-test utilizing a small amount of sample.Lastly, one could use some measurement, such as electric conductivity or creatinine concentration in urine, to determine appropriate dilution ratio.Capillary electrophoresis-tandem mass spectrometry combined with molecularly imprinted solid phase extraction as useful tool for the monitoring of 5-nitroimidazoles and their metabolites in urine samples.SPE on-line [118] Yamamoto M, Ly R, Gill B, Zhu Y, Moran-Mirabal J, Britz-McKibbin P.
Robust and High-Throughput Method for Anionic Metabolite Profiling: Preventing Polyimide Aminolysis and Capillary Breakages under Alkaline Conditions in Capillary Electrophoresis-Mass Spectrometry.
A capillary electrophoresis coupled to mass spectrometry pipeline for long term comparable assessment of the urinary metabolome.
Capillary electrochromatography and nano-liquid chromatography coupled to nano-electrospray ionization interface for the separation and identification of estrogenic compounds.
Metabolic profiling for the identification of Huntington biomarkers by on-line solidphase extraction capillary electrophoresis mass spectrometry combined with advanced data analysis tools.
Multiplatform metabolomic fingerprinting as a tool for understanding hypercholesterolemia in Wistar rats.
Insulin resistance in prepubertal obese children correlates with sexdependent early onset metabolomic alterations.
New insight on obesity and adipose-derived stem cells using comprehensive metabolomics.
Multiplatform [127] chromatography to capillary electrophoresis with advantages and challenges for both.In Methods in Molecular Biology (2016) a whole chapter is provided [117] on CE-MS in metabolomics with detailed protocols and technical information for those getting into this field.

Technology and protocols
Innovative technologies continue to advance through the ingenuity of researchers, providing reliable improvements in the technologies and protocols.Such advancements are needed in order to provide solutions to increasingly difficult analytical problems seen in the medical and pharmaceutical fields.Over the past year, several publications (Table 2) are noted demonstrating new ideas and concepts designed to expand and improve CE-MS.Untargeted profiling, when including multiple analytical methods, provides the greatest opportunity to discover biomarkers, understand biological pathways and identify new drug targets.The University of San Pablo published 3 papers in 2016 [125][126][127] using multiple mass spectrometry based untargeted technologies [LC-MS, CE-MS and GC-MS] to create large data rich metabolomics profiles in plasma [125], serum [126] and from cell culture [127] to identify biomarkers in obesity and hypercholes- One of the long-standing issues with CE-MS is the ease and stability of the CE to MS interface.Ramautar [123] and Boizard [120] [137], CSF [130] and serum [136] to both understand the biology of BD and seek biomarkers for diagnosis.Hashimoto [130] aptly used the specificity of CE-MS to understand the role of mitochondrial dysfunction in BD, finding isocitrate as a diagnostic biomarker in CSF.Likewise, the same group looked at serum TCA cycle metabolites using CE-MS to resolve and identify specific energy metabolites in BD sera.While Fuji [138] used the uniqueness of CE-MS to identify metabolites in brain tissue specific to Schizophrenia.

Future
The successful implementation and integration of metabolomics in personalized medicine and drug development will rely on the combination of high-throughput analysis, large metabolite coverage, accurate quantitation, high-value data and low-cost analysis.This paradigm can only be achieved using state-of-the-art analytical technologies and computational techniques combining many modalities and data input with growth in technology and implementation.CE-MS is poised to make significant contributions to these fields as the technology and availability to CE-MS continues to increase.Multi-platform, multi-omics approaches with large studies will develop high-powered algorithms for diagnostic and prognostic studies.Increasing CE databases, improvements in sheathless interfaces and increases in MS detection capability will further push the field forward helping to find solutions to critical questions in drug development and biotechnology.
biomarkers based on proteome profiling of mouse models.Br J Cancer 113 (11)

Table 2 .
Technical and methods papers on CE-MS based metabolomics (year 2016) [109][112][113][114]cal challenges overcome, the advantages of CE-MS are clearly understood.With the availability of commercial CE-MS instrumentation, development in academic labs and contract labs and with a growing list of publications and reviews, we saw in 2016 a balance of papers reflecting the diversity and growth of CE-MS in biotechnology and drug development.Over 200 CEMS publications in 2016In this paper recent CE-MS applications developed for metabolomics covering the literature from Jan 2016 to Dec 2016 are outlined in Tables 1, 2 and 3. Attention will be paid to CE-MS approaches for the profiling of metabolites in the fields of biomedical, clinical and microbial metabolomics.er'slaboratoryatLeidenUniversity provided 4 separate reviews[111][112][113][114]covering different aspects of CE-MS from growing applications to technical developments.Tang et al.[109]provides a discussion comparing HILIC