LC-MS Metabolomics and Next-Generation Ionization Technologies

In this episode of ‪ChromatographyTalk‬, David speaks with Joseph Corstvet from the Fernandez Lab at Georgia Tech about the future of mass spectrometry, LC-MS metabolomics, and the development of low-cost ionization technologies.

Joe discusses his work using LC-MS lipidomics to monitor mesenchymal stromal cell (MSC) senescence and explains how metabolomics could eventually become a real-world quality control tool for regenerative medicine and cell therapy manufacturing.

The conversation also dives into triboelectric nanogenerators (TENGs) and how these low-cost, modular ionization systems could help make mass spectrometry more portable, accessible, and affordable. Joe explains how TENG-powered electrospray ionization works, why it differs from traditional ESI, and how his group is exploring applications ranging from mass spectrometry imaging to mobile MS platforms.

Topics include:

• LC-MS metabolomics workflows
• Lipidomics and MSC senescence
• Biomarker discovery in regenerative medicine
• Untargeted metabolomics challenges
• TENG ionization technology
• Portable and lower-cost mass spectrometry
• Mass spectrometry imaging
• Ion mobility and future MS trends
• ASMS and the future of analytical chemistry

Video Transcription

From Cell Health to Portable Mass Spectrometry: Joseph Corstvet on LC–MS, Lipid Biomarkers, and Triboelectric Ionization

Joseph Corstvet’s path into analytical chemistry began with an undergraduate degree in chemistry, continued through biotechnology and public-health research, and ultimately led him to the development of new ionization technologies and metabolomics workflows at Georgia Tech. In an interview for Concentrating on Chromatography, he discussed his introduction to GC–MS at the US Centers for Disease Control and Prevention, the analytical challenges of monitoring mesenchymal stromal cell health, and his interest in making mass spectrometry more affordable, modular, and portable.

Discovering mass spectrometry at the CDC

After completing his bachelor’s degree, Corstvet initially worked for a biotechnology company. He later moved to Atlanta and joined the CDC as an ORISE fellow, where his group used gas chromatography–mass spectrometry to investigate volatile organic compounds in blood.

Most of the work relied on single-quadrupole GC–MS, but the experience gave him extensive access to analytical instrumentation. Beyond assigned projects and formal deliverables, researchers were encouraged to explore their own ideas and conduct small independent experiments.

That freedom proved important. Corstvet became interested not only in collecting data, but also in understanding how the instruments worked. He enjoyed testing different configurations, disassembling and cleaning components, returning systems to vacuum, and investigating how changes affected analytical performance.

When the COVID-19 pandemic reduced the amount of work available to him in his existing research area, he applied to graduate programs. He was accepted at Georgia Tech and joined the Fernández laboratory as a PhD student, where mass spectrometry became the central focus of his research.

Why ionization became a research priority

Every mass spectrometric experiment depends on converting neutral molecules into ions that can be manipulated, separated, and detected. Corstvet was therefore drawn to ionization as one of the most fundamental parts of the analytical workflow.

Modern commercial mass spectrometers have reached a level of sophistication that makes it extremely difficult for academic laboratories to construct a complete instrument capable of competing with commercially available systems. Ion sources, however, still offer considerable room for experimentation.

New ionization technologies can reduce sample-preparation requirements, improve compatibility with chromatography, expand the range of compounds that can be measured, or support applications such as mass spectrometry imaging.

The Fernández laboratory had previously developed plasma-based ionization sources for applications including the analysis of compounds generated during plastic pyrolysis and the measurement of volatile organic compounds. Plasma ionization was also combined with mass spectrometry imaging to spatially resolve the distribution of ibuprofen within a tablet.

For Corstvet, these projects demonstrate the value of combining established analytical platforms with customized components. Building such systems can be frustrating when they fail, he noted, but successfully converting an experimental concept into a functioning analytical device is particularly rewarding.

Monitoring mesenchymal stromal cell health

One major part of Corstvet’s doctoral research concerns mesenchymal stromal cells, commonly abbreviated as MSCs. These naturally occurring mammalian cells are being studied as potential whole-cell therapies because of their roles in immune regulation, tissue repair, and regeneration.

In a healthy organism, MSCs can respond to injury by releasing signaling factors that recruit immune cells and other cell populations involved in repair. Their ability to respond dynamically to local biological conditions has made them attractive candidates for regenerative medicine and cell-based therapeutic applications.

At the time of the interview, Corstvet highlighted the December 2024 US approval of an MSC-based therapy for pediatric graft-versus-host disease as an important milestone for the field.

However, manufacturing living-cell therapies creates analytical challenges that are very different from those associated with conventional small-molecule drugs. A pharmaceutical dose can often be described using a defined mass or concentration. For a cell therapy, the number of cells alone does not necessarily indicate therapeutic potency.

The physiological condition of those cells is equally important.

Cellular senescence as a manufacturing challenge

MSCs must be expanded in culture to produce sufficient numbers for therapeutic use. During repeated culture and replication, however, the cells can enter a state of cellular senescence.

Senescent cells remain viable but lose much of their replicative potential and may no longer perform in the same way as younger, healthier cells. A preparation containing a large number of senescent MSCs may therefore be less effective than one containing the same number of biologically active cells.

This makes senescence a critical quality-control issue in cell-therapy manufacturing. Laboratories need a way to determine not only how many cells are present, but also whether those cells retain the desired biological state.

Traditional senescence measurements include assays based on β-galactosidase activity. Corstvet explained that these tests are useful but imperfect because the measured enzyme activity may also be observed in healthy cells. His research therefore explores whether metabolite profiles can provide a more detailed assessment of cellular age and condition.

Why metabolomics can reveal cell phenotype

Genomic information is relatively stable, while metabolism changes rapidly in response to activity, stress, nutrient availability, and environmental stimuli. Even a simple event such as drinking coffee can alter metabolic processes within the body.

For this reason, metabolomics is often described as one of the omics disciplines closest to phenotype. It provides a snapshot of what a cell is doing at a particular moment rather than only describing its underlying genetic potential.

Young and senescent MSCs can remain genomically similar. Their metabolic activity, however, may differ substantially.

Corstvet’s objective is to identify panels of metabolites that reflect these differences. In a future manufacturing workflow, a sample could be collected from a bioreactor or cell culture, analyzed by LC–MS, and evaluated using a statistical model capable of estimating cellular age or predicting the degree of senescence.

Such an approach could provide a quantitative and reproducible method for monitoring cell health during culture expansion.

The role of LC–MS in lipid metabolomics

Liquid chromatography–mass spectrometry is one of the most widely used platforms for untargeted metabolomics. Corstvet’s study focused particularly on lipids, which represent a chemically diverse group of compounds with very different ionization efficiencies.

If a whole-cell extract is introduced directly into a mass spectrometer, compounds compete for a limited amount of charge during ionization. Molecules that ionize efficiently can suppress the response of less readily ionized analytes.

Phosphatidylcholines, for example, generally ionize strongly. Without prior chromatographic separation, they can dominate the signal and reduce the detectability of other lipid classes.

Coupling LC to MS reduces this competition by separating compounds in time. In Corstvet’s method, phosphatidylcholines eluted during one part of the chromatographic run, while triglycerides appeared later. Because the lipid classes reached the ion source at different times, they were not competing for charge to the same degree.

Chromatography therefore increased metabolite coverage and supported the detection of compounds that might otherwise have been suppressed.

Retention time also contributed to identification. If a signal was tentatively assigned as a triglyceride but eluted in the chromatographic region normally associated with phosphatidylcholines, that inconsistency would reduce confidence in the annotation.

From thousands of features to a biomarker panel

The untargeted LC–MS dataset initially contained approximately 4,000 detected features. Through annotation workflows supported by the Georgia Tech mass spectrometry core, the researchers assigned identities to more than 200 lipids.

Annotation involves connecting an observed mass-to-charge ratio with a likely compound identity using multiple forms of evidence, including accurate mass, chromatographic behavior, and fragmentation spectra.

The resulting dataset was evaluated using common metabolomics tools. Principal component analysis was used to determine whether sample groups separated according to cellular condition. Volcano plots and univariate statistical analyses then helped identify compounds that were both significantly altered and strongly associated with either young or senescent cells.

The researchers subsequently used the scientific literature to interpret why particular lipid classes changed. For example, elevated triglyceride levels in senescent cells could provide information about altered lipid storage, energy regulation, or cellular stress.

The long-term objective is not to retain hundreds of markers. Corstvet envisions narrowing the dataset to a targeted panel of perhaps ten lipids that are highly descriptive of senescence.

A small aliquot could then be collected at different stages of cell culture and analyzed using a targeted LC–MS method. Concentrations or relative abundances could be compared with experimentally established thresholds, allowing manufacturers to classify cells as healthy or increasingly senescent.

Corstvet suggested that such monitoring might eventually be performed using an LC–MS run of approximately 12 minutes.

Optimizing extraction and chromatographic separation

Developing the workflow required optimization of both sample preparation and chromatography.

Corstvet initially tested an extraction containing equal proportions of water and isopropanol. The results were not satisfactory, so the procedure was changed to extraction of whole cells using 100% isopropanol, which produced better lipid recovery for the intended analysis.

The chromatographic method used a C30 reversed-phase column. Because many lipids are highly nonpolar, the longer alkyl stationary phase provided stronger retention and better separation than a conventional C18 phase in the laboratory’s testing.

The choice of column was also supported by an established lipid-analysis workflow already used by the Georgia Tech mass spectrometry core.

Retention-time reproducibility was essential, particularly for the proposed transition from untargeted discovery to targeted quality control. Corstvet recalled that the method generally maintained retention-time shifts within approximately five to ten seconds.

Consistent retention reduces the risk of confusing one compound with another and improves confidence when comparing many samples across extended studies.

Why tandem mass spectrometry is essential

Accurate mass and retention time alone are rarely sufficient for confident lipid identification. Many lipid species have similar elemental compositions, overlapping masses, or comparable chromatographic behavior.

Corstvet’s group therefore avoids assigning definitive lipid identities without MS/MS fragmentation data.

When a precursor ion is fragmented, the resulting product ions can reveal diagnostic structural features. Phosphatidylcholines, for example, generate characteristic fragments that help distinguish them from triglycerides and other lipid classes.

The researchers compared fragmentation spectra with an in-house database maintained in collaboration with the Georgia Tech core facility. When no internal match was available, external resources such as the Human Metabolome Database and LIPID MAPS were consulted.

Without reliable annotations, statistical analysis would only show that unnamed numerical features differed between conditions. Tandem mass spectrometry transforms those features into chemically meaningful biomarkers that can be interpreted biologically.

Generating electrospray with a triboelectric nanogenerator

A second major area of Corstvet’s work involves the triboelectric nanogenerator, or TENG.

The device functions as a compact power source capable of generating the high voltage required for electrospray ionization. Instead of relying on a conventional direct-current power supply, it uses triboelectricity—the same principle responsible for static charge generated by rubbing a balloon against hair or walking across a carpet in wool socks.

In the laboratory’s current design, polytetrafluoroethylene and polyurethane repeatedly move against one another. Copper-foil electrodes collect the resulting charge, which can then be transferred to an electrospray emitter.

The movement generates alternating pulses of electrical potential. Applied to an emitter, these pulses produce short bursts of electrospray and introduce ions into the mass spectrometer.

According to Corstvet, the device can be assembled in approximately one day using around USD 250 worth of materials. Because the system is built within the laboratory, its dimensions and configuration can be adapted for different applications.

The TENG has generated pulses reaching approximately nine kilovolts. Applying the same voltage continuously with a conventional DC supply could damage an electrospray emitter. The short duration of the triboelectric pulse allows very high voltage to be used without maintaining that stress continuously.

A route toward lower-power and portable MS

The triboelectric system also operates with relatively low power consumption. Corstvet estimated the complete setup draws approximately 20–25 watts, making it potentially suitable for battery-powered applications.

Replacing the conventional ionization power supply does not solve every cost associated with mass spectrometry. The mass analyzer, vacuum system, electronics, and data-processing software remain expensive.

Corstvet identified the price of the instruments themselves as the largest barrier to accessibility. He estimated that even relatively inexpensive new mass spectrometers may cost USD 80,000–100,000, while specialized data-processing licenses introduce additional expenses.

Nevertheless, reducing the cost, size, and power demand of individual components can contribute to the broader development of more accessible instruments.

Corstvet is particularly interested in systems that could be transported beyond centralized analytical laboratories. A compact tabletop mass spectrometer could support environmental analysis close to the sampling site, including the measurement of stream water or local air quality.

Portable systems could also have clinical applications, although instruments must be designed to tolerate movement and changing conditions without compromising mass accuracy or analytical stability.

Where time is spent in a metabolomics workflow

For Corstvet’s research, sample preparation represented only about 5–10% of the total workflow. Instrument operation accounted for approximately 40–50%, while the remaining time was largely devoted to data processing.

This distribution highlights a common feature of untargeted metabolomics: generating the raw data is only one stage of the project. Aligning peaks, removing noise, normalizing signal intensities, annotating compounds, and interpreting multivariate results can require more time than the extraction and chromatographic analysis combined.

Access to suitable instrumentation remained his greatest practical limitation. The necessary chemistry and sample-preparation equipment were generally available, but obtaining time on an instrument with the required capabilities could be more difficult.

Reproducibility in untargeted metabolomics

Corstvet acknowledged that untargeted metabolomics presents reproducibility challenges. Changes to the gradient, stationary phase, column condition, instrument performance, or sample-preparation procedure can influence which features are detected and how they are measured.

Analytical consistency is therefore essential. Samples must be prepared and analyzed using the same procedure, while software tools are used to align chromatographic peaks and normalize the resulting datasets.

Even with these controls, some features may appear in most samples but remain absent from a smaller fraction. If a signal cannot be reproduced consistently, the researchers may exclude it rather than risk basing a biological conclusion on noise or an analytical artifact.

More reproducible workflows would allow researchers to retain a greater number of potentially useful metabolites. Standardized procedures and rigorous quality-control samples will therefore remain important as untargeted methods move toward targeted assays and eventual manufacturing applications.

The future of accessible mass spectrometry

Ion mobility is among the areas of mass spectrometry currently advancing rapidly, but Corstvet is especially interested in the movement toward smaller, less expensive, and more mobile systems.

High-performance Orbitrap and time-of-flight instruments continue to provide exceptional analytical capabilities. Portable systems are not yet able to match all aspects of their performance, but improvements in miniaturization, electronics, ionization, and power consumption are gradually expanding what can be achieved outside a conventional laboratory.

Corstvet hopes that triboelectric nanogenerators can contribute to this transition. A newer TENG design under development in the Fernández laboratory is intended to be even less expensive and more modular while improving the charge delivered to the emitter.

The laboratory is also applying metabolomics to additional biomedical questions, including the search for markers associated with ovarian cancer, a disease for which earlier detection remains an important analytical and clinical objective.

Connecting instrumentation development with biological questions

Corstvet’s research combines two areas that can appear very different: the development of low-cost ionization technology and the use of LC–MS metabolomics to monitor living cells.

Both, however, address the same larger objective—making mass spectrometry more useful in real analytical settings.

In cell-therapy manufacturing, this means moving from thousands of untargeted features toward a small, reproducible biomarker panel that can support quality-control decisions. In instrumentation, it means reducing the cost, size, and power requirements that limit where mass spectrometry can be deployed.

The interview illustrates that analytical innovation does not depend only on developing a more powerful mass analyzer. Progress can also come from improving extraction, selecting a more suitable chromatographic column, strengthening compound annotation, developing robust statistical workflows, or replacing an expensive high-voltage supply with a modular triboelectric device.

By working across these different parts of the analytical process, Corstvet is helping connect fundamental mass spectrometry research with regenerative medicine, metabolomics, environmental testing, and the longer-term goal of portable chemical analysis.

This text has been automatically transcribed from a video presentation using AI technology. It may contain inaccuracies and is not guaranteed to be 100% correct.

Concentrating on Chromatography Podcast

Dive into the frontiers of chromatography, mass spectrometry, and sample preparation with host David Oliva. Each episode features candid conversations with leading researchers, industry innovators, and passionate scientists who are shaping the future of analytical chemistry. From decoding PFAS detection challenges to exploring the latest in AI-assisted liquid chromatography, this show uncovers practical workflows, sustainability breakthroughs, and the real-world impact of separation science. Whether you’re a chromatographer, lab professional, or researcher you'll discover inspiring content!

You can find Concentrating on Chromatography Podcast in podcast apps:

• Spotify
• Apple podcast
• podscan

and on YouTube channel