With the recent shifts in the GLP-1 therapeutics landscape, analytical chemists and bench scientists are increasingly focused on the robust development and quality control of these treatments. A critical frontier in this field involves tracking the neurological impacts of GLP-1 receptor agonists, particularly concerning brain insulin resistance.
In a recent Separation Science interview, Dr. Dimitrios Kapogiannis—a clinician-scientist and Senior Investigator at the National Institute on Aging (NIA) within the National Institute of Health (NIH)—detailed his cutting-edge work translating brain insulin resistance into measurable clinical biomarkers. Having pioneered the use of neuron-enriched extracellular vesicles (EVs) as a "liquid biopsy" for neurodegenerative diagnostics, Dr. Kapogiannis offers a unique bench-to-bedside perspective on the analytical hurdles this field currently faces.
The Promise of Neuronal Extracellular Vesicles (EVs)
Brain insulin resistance refers to the reduced responsiveness of the brain to insulin. Because researchers cannot directly observe insulin's effects on living human brain cells, they must rely on surrogate markers. Drawing from his extensive clinical and translational research, Dr. Kapogiannis's approach achieves this by studying extracellular vesicles of neuronal origin found in peripheral blood. These particles carry the same molecular markers found in the brain, reflecting their cell of origin and providing a vital window into neuronal insulin signaling.
Working in analytical science?
Register for a FREE Separation Science account to subscribe to the Separation Science Newsletter.
Isolating these specific vesicles, however, presents a significant analytical hurdle. Plasma is a complex matrix containing EVs from every tissue in the body. The target population—neuronal EVs—constitutes only 1% to 3% of the total circulating peripheral blood EVs. Navigating this complexity requires acknowledging several key factors in the isolation process.
To capture these rare targets, researchers utilize positive selection targeting specific neuronal antigens. "Technically, we're talking about extracellular vesicles enriched for neuronal origin," Dr. Kapogiannis clarifies. "'Enriched' is the key term. We have an amplification of neuronal cargo, but not an absolute purity."
This reality also dictates the field's terminology. "In reality, you can never isolate pure exosomes," he notes, explaining the preference for the broader term 'extracellular vesicles.' "You can never be ultimately sure about their origin. What you know is that you have particles that are membranous and they have a specific range of sizes."
Understanding these fundamental characteristics is essential for building robust analytical workflows moving forward.
Methodological Evolution and Target Specificity
The baseline workflow for analyzing these biomarkers involves first isolating all extracellular vesicles, followed by enrichment of the neuronally derived subset. Remarkably, this powerful enrichment process can be successfully performed with as little as 0.5 mL of starting plasma. Over time, this methodology has undergone necessary evolution to improve purity and concentration, highlighted by the following shift in techniques:
- Initial iterations of the protocol used particle-precipitation solutions to concentrate EVs, followed by immunoaffinity capture.
- The current, modernized approach uses size-exclusion chromatography (SEC) to produce a purer preparation with fewer soluble contaminants, followed by immunoaffinity capture.
This transition to SEC marks a crucial improvement in minimizing interference from soluble plasma proteins.
A primary focus of this capture process is the neuronal marker L1 CAM. While acknowledging ongoing debates regarding specificity, Dr. Kapogiannis remains confident in the marker's utility. "I stand firmly by L1 CAM as a highly valuable marker," asserts Dr. Kapogiannis. "Evidence clearly demonstrates that L1 CAM-positive extracellular vesicles exist, and a large percentage of these originate directly from neurons." Specifically, these enriched EVs allow researchers to study Insulin Receptor Substrate 1 (IRS1), a key marker of insulin signaling that responds to treatments such as GLP-1 receptor agonists.
To overcome remaining specificity issues and to sample EVs from various parts of neurons, the lab has adopted a multiplexed approach. Current protocols combine L1 CAM with additional neuronal markers, such as GAP-43 and Neurogranin-3, to target three antigens simultaneously.
Despite his lab's success with these specific targets, Dr. Kapogiannis promotes an "open source" mentality regarding capture markers. "Identifying the optimal marker is an empirical challenge rather than an a priori certainty," he stated. "I encourage other investigators to adapt our general methodology and experiment to find the specific markers that work best within their own laboratory environments."
Analytical Roadblocks: Phosphorylation and Mass Spectrometry
Once enriched EVs are isolated, the choice of downstream analytical methodology depends heavily on the specific target and the drug mechanism. While some investigators use these same enriched EVs to extract and analyze microRNAs or messenger RNAs, those focusing on proteins must navigate several critical analytical realities.
"A drug, dietary intervention, or similar treatment does not necessarily change the total abundance of a given protein. Instead, it frequently alters the protein's phosphorylation state," Dr. Kapogiannis explains, emphasizing the importance of measuring the phosphotype of a signaling molecule. "By solely measuring total protein levels, investigators risk missing the underlying mechanism of action entirely, meaning generalized untargeted proteomics may fall short.”
To capture these critical functional changes, researchers require sensitive immunoassays. However, Dr. Kapogiannis points out that the field is rapidly evolving; newer multiplexed and targeted proteomic platforms (such as nucleic acid-linked immunosorbent assay (NULISA)) are beginning to embed antibodies against specific phosphotypes. This integration hints at where analytical technology is heading to bridge the gap between untargeted discovery and functional specificity.
Balancing these considerations is paramount as researchers strive to map the nuanced mechanisms of action for novel therapeutics.
Working with the minuscule sample volumes derived from the rare neuronal EV population naturally limits starting protein material. High-resolution mass spectrometry faces compounding challenges in this matrix. A major issue involves isolating low-abundance peptides against a massive background of highly abundant peptides and soluble contaminants. Furthermore, the protein "corona" that forms around the outside of EVs complicates the precise interpretation of where a signal originates.
For bench scientists grappling with this corona effect, Dr. Kapogiannis suggests a specific, rigorous experimental approach: running parallel experiments in which the corona is chemically stripped prior to proteomics and compared against unstripped samples, or isolating just the vesicular membrane to specifically analyze transmembrane proteins.
Ultimately, dealing with these analytical hurdles forces a philosophical choice between absolute precision and clinical pragmatism. As a researcher actively conducting early-phase interventional clinical trials, Dr. Kapogiannis argues that while exact interpretation is scientifically satisfying, functionality is key. "If a biomarker effectively discriminates between patients and controls, or successfully demonstrates a drug response, we sometimes choose clinical utility over absolute mechanistic precision," he asserts.
The Path to Clinical Validation
Despite impressive internal lab reproducibility—with Dr. Kapogiannis noting that his lab achieves R-squared values greater than 0.9 both between and within operators—EV assays face a steep climb to meet the stringent regulatory standards required for routine clinical trials. EV technologies inherently introduce variance at multiple stages, including total EV isolation, immunoaffinity capture, and the final analytical assay. Because of these compounding variables, Dr. Kapogiannis cautions, "Currently, our analytical validation is not yet robust enough to support using an extracellular vesicle study as a primary clinical outcome," particularly for major regulatory submissions such as an FDA application for a change in patent use.
"From a clinician's perspective, the primary indication must usually be a definitive clinical outcome," he explains. Therefore, "extracellular vesicle data is currently best positioned as secondary supportive evidence to corroborate those primary clinical findings." Ultimately, achieving primary clinical readout status will strictly require reproducible, well-published protocols and universally standardized materials that can be shared to definitively demonstrate interlab reproducibility. Overcoming these validation challenges remains the most significant milestone in transitioning EV assays from the bench to routine clinical diagnostics.
The EV Technology Wishlist
When asked to mandate a single technological leap from instrument manufacturers to solve current bottlenecks, Dr. Kapogiannis points directly to automation. The ultimate objective is a seamless, automated procedure that handles size-exclusion chromatography, followed directly by immunocapture.
By eliminating human pipetting error, this ideal system would allow researchers to load a raw plasma sample and automatically generate a suspension of isolated, neuronally enriched EVs, ready for downstream analysis. Crucially, Dr. Kapogiannis notes the required output of such a system: "The ideal scenario is a fully automated process where you start with a raw plasma sample and end with a purified suspension that is at least as concentrated as the initial volume, preventing further downstream sample losses caused by manual reconcentration steps.”
