Closing the information gap: ETH Zürich team integrates function and binding in one drug screening assay

A groundbreaking innovation from a collaborative team at ETH Zurich and Karolinska Institutet is poised to revolutionize early-stage drug discovery by addressing a persistent challenge in pharmaceutical development: the "information gap" between whether a drug binds to its target and whether that binding actually translates into a desired biological effect. Their newly developed cross-linking MALDI mass spectrometry (MS) workflow simultaneously captures both the functional response and the target binding of a drug candidate within a single, integrated assay, offering an unprecedented level of mechanistic insight that has historically been elusive. This integrated approach promises to significantly de-risk the arduous and expensive journey of bringing new therapeutics to market, particularly for complex targets like protein-protein interactions.

The Enduring Challenge in Drug Discovery: Bridging Function and Binding

For decades, drug screening strategies have largely relied on two distinct and often disconnected types of assays: functional assays and binding assays. While both are indispensable, their inherent limitations have created a critical void in understanding a drug’s true potential. Functional assays are designed to measure a drug candidate’s biological activity, such as its ability to inhibit an enzyme, activate a receptor, or modulate a cellular pathway. They answer the crucial question: "Does it work?" However, they often fail to reveal how it works, or even if it directly interacts with the intended molecular target. A compound showing functional activity might be acting via an off-target mechanism, leading to misleading results and ultimately, clinical failure.

Conversely, binding assays focus on assessing whether a drug candidate physically interacts with its molecular target, quantifying parameters like affinity (how strongly it binds) and kinetics (how fast it binds and unbinds). These assays answer the question: "Does it bind?" Yet, they offer no guarantee that binding will elicit the desired biological response. A compound might bind potently to its target but fail to induce the necessary conformational change or downstream signaling cascade required for therapeutic efficacy. Furthermore, high affinity binding to a target does not preclude off-target binding elsewhere, which can lead to unwanted side effects and toxicity.

This fundamental disconnect between binding and function represents a significant bottleneck in drug development, contributing substantially to the staggering failure rates observed in clinical trials. Research indicates that clinical efficacy, or the lack thereof, accounts for a substantial 40% to 50% of all clinical failures. This high attrition rate translates into colossal financial losses and delays in delivering much-needed medicines to patients. The problem is particularly acute for drug candidates targeting protein-protein interactions (PPIs), which are notoriously challenging due to their large, often dynamic interaction surfaces and the absence of traditional, well-defined small-molecule binding pockets. PPIs are implicated in a vast array of biological processes and diseases, from cancer and autoimmune disorders to infectious diseases, making them highly attractive yet historically "undruggable" targets.

The Genesis of an Integrated Solution: A Timeline of Innovation

The quest for more comprehensive drug screening tools has been a driving force in pharmaceutical research for many years. The evolution of mass spectrometry, in particular, has played a pivotal role in advancing analytical capabilities within drug discovery.

Early 2000s: Mass spectrometry gains prominence in drug discovery for its speed and sensitivity in identifying and quantifying molecules. MALDI-MS (Matrix-Assisted Laser Desorption/Ionization Mass Spectrometry) emerges as a powerful tool for analyzing large biomolecules like proteins, primarily used for identification, quality control, and measuring enzyme activity by tracking substrate-to-product conversion.
Mid-2000s: The growing understanding of the importance of protein-protein interactions in disease pathogenesis fuels intense research into developing methods to study these complex targets. However, conventional analytical techniques, including standard MALDI-MS, struggle to reliably detect intact noncovalent protein complexes. The high-energy laser ionization process inherent to MALDI-MS, while excellent for individual protein detection, is often too harsh, breaking apart the delicate noncovalent bonds that hold protein complexes together, preventing direct observation of the intact bound state.
Late 2000s – Early 2010s: Chemical cross-linking techniques begin to gain traction as a method to stabilize protein complexes for structural studies. These techniques involve introducing a chemical reagent that forms covalent bonds between proteins that are in close proximity, essentially "locking" them together. This stabilization allows for subsequent analysis by techniques like mass spectrometry, which would otherwise dissociate the complex.
Mid-2010s: The Breakthrough: Building upon these foundational advancements, the collaborative team from ETH Zurich and Karolinska Institutet synthesized these distinct methodologies. They recognized the potential of combining the stabilizing power of chemical cross-linking with the analytical precision of MALDI-MS. Their innovation wasn’t merely to use cross-linking before MS, but to develop a workflow that integrates these steps seamlessly to provide simultaneous functional and binding data. The publication of their seminal work in ACS Central Science marked a significant step forward, offering a new paradigm for drug screening. This represented a culmination of years of research into protein biophysics, chemical biology, and advanced analytical instrumentation.

Unlocking New Capabilities: How Cross-linking MALDI-MS Works

The core ingenuity of the new workflow lies in its ability to circumvent the limitations of traditional MALDI-MS when analyzing protein complexes. Conventional MALDI-MS, while excellent for identifying and quantifying individual proteins or peptides, cannot reliably detect intact protein-protein complexes because the strong laser ionization process used to vaporize and ionize the sample often disrupts the weak, noncovalent forces that hold these complexes together. The result is typically a spectrum showing only the individual components, masking the presence of a complex.

The ETH Zurich and Karolinska Institutet team’s solution introduces a crucial preliminary step: chemical cross-linking. Before the mass spectrometry analysis, an NHS-ester (N-hydroxysuccinimide ester) reagent is added to the protein-drug mixture. NHS-esters are highly reactive chemical linkers that specifically target primary amine groups (found on lysine residues and the N-terminus of proteins). When two proteins are in close proximity – indicating they are interacting – the NHS-ester reagent forms covalent bonds between them. This process effectively "locks" the protein complex together, transforming weak, transient noncovalent interactions into stable, permanent covalent linkages.

Once the protein complex is covalently stabilized, it is then subjected to MALDI-MS analysis. Because the complex is now held together by strong covalent bonds, it can withstand the laser ionization process, allowing for the detection of the intact, higher-molecular-weight complex. The mass spectrum will show not only the individual proteins but also peaks corresponding to the cross-linked complex. The presence and intensity of these complex peaks directly indicate target binding.

Crucially, this method also provides insight into functional response. By observing how a drug candidate influences the formation or disruption of a specific protein-protein interaction, researchers can infer its functional impact. For instance, if a drug is designed to inhibit a PPI, a reduction in the signal for the cross-linked complex in the presence of the drug would indicate both binding (to prevent interaction) and a functional effect (inhibition of the interaction). Conversely, if a drug is a molecular stabilizer, an increase in the complex signal would demonstrate both binding and a stabilizing function. This "double output" from a single assay provides a holistic view of the drug’s activity, directly correlating binding events with their functional consequences at the molecular level.

A Powerful Demonstration: Targeting SARS-CoV-2

To validate their innovative workflow, the researchers applied it to a highly relevant and challenging therapeutic target: the interaction between the SARS-CoV-2 spike protein’s receptor-binding domain (RBD) and human angiotensin-converting enzyme 2 (ACE2). This interaction is the critical initial step for SARS-CoV-2 entry into human cells, making it a prime target for antiviral drug development.

The team screened 17 FDA-approved drug candidates, assessing their ability to modulate the SARS-CoV-2 RBD-ACE2 interaction. Their findings provided compelling evidence for the assay’s superior discriminatory power, particularly in differentiating between compounds that appear similar in conventional assays.

Amentoflavone and dalbavancin, two compounds that might have been indistinguishable by less sophisticated methods, exhibited critical differences when analyzed with the cross-linking MALDI-MS workflow. The assay revealed that dalbavancin bound to ACE2 with approximately 10-fold stronger affinity compared to amentoflavone. More importantly, dalbavancin demonstrated preferential, on-target engagement, meaning its binding was specific to the ACE2 protein, directly interfering with the SARS-CoV-2 RBD interaction. In contrast, amentoflavone displayed weaker and less specific binding, suggesting it might interact with other cellular components or bind to ACE2 in a less effective manner.

The predictive power of these integrated insights was then rigorously confirmed through a cell-based antiviral assay, the gold standard for assessing antiviral efficacy. In this assay, dalbavancin significantly improved cell viability in SARS-CoV-2-infected cells, directly correlating with its strong, specific binding and functional interference observed in the MALDI-MS assay. Conversely, amentoflavone showed no discernible benefit in improving cell viability and, concerningly, even exhibited mild toxicity at higher concentrations. This stark contrast unequivocally demonstrated the ability of the cross-linking MALDI-MS workflow to accurately predict the in vitro efficacy and potential toxicity of drug candidates, thereby preventing the advancement of ineffective or harmful compounds.

Broader Impact and Transformative Implications

The implications of this integrated drug screening platform extend far beyond the specific case of SARS-CoV-2. By providing richer, more comprehensive data earlier in the drug discovery pipeline, this method promises to transform decision-making processes, enabling researchers to make smarter choices about which compounds to advance and which to discard.

Accelerated Drug Development and Cost Savings: The ability to simultaneously assess both binding and function in a single assay allows for earlier de-risking of drug candidates. Identifying compounds that bind but lack functional efficacy, or vice-versa, at an earlier stage can prevent significant investment in lead optimization, preclinical testing, and ultimately, costly clinical trials for compounds destined to fail. Given that the average cost to develop a new drug can exceed $2 billion, and clinical failures are a major driver of this expense, any tool that improves the efficiency and success rate of candidate selection offers immense economic benefits.
Enhanced Understanding of Mechanism of Action: This platform provides deeper insights into how drugs interact with their targets and the direct consequences of those interactions. This mechanistic understanding is invaluable for rational drug design and for optimizing lead compounds.
Expanding Therapeutic Horizons: The platform’s utility extends beyond traditional inhibitors. It holds significant potential for identifying:
- Molecular Stabilizers: Compounds that enhance the stability or integrity of protein complexes, which could be crucial for treating diseases caused by protein misfolding or degradation, such as certain neurodegenerative disorders.
- Allosteric Activators: Molecules that bind to a site other than the active site of a protein, inducing a conformational change that increases its activity. This opens new avenues for targeting enzymes or receptors that are underactive in disease states.
Versatility Across Disease Areas: The principles underlying this assay are broadly applicable to a wide range of diseases where PPIs play a critical role. This includes various forms of cancer (e.g., targeting oncogenic protein complexes), autoimmune diseases, and other infectious diseases. The robust characterization of PPI modulators can unlock new therapeutic strategies previously hampered by the lack of suitable screening technologies.
Strategic Positioning in the Drug Discovery Pipeline: While not intended for the initial high-throughput screening (HTS) of millions of compounds, the cross-linking MALDI-MS workflow is ideally positioned for post-primary screening. Once an initial HTS campaign has identified a smaller pool of "hits," this integrated assay can be deployed to rapidly filter out false positives, prioritize compounds with genuine on-target activity and functional relevance, and provide critical mechanistic data for lead optimization. This strategic placement maximizes its impact by focusing resources on the most promising candidates.

Limitations and Future Directions

Despite its transformative potential, the new cross-linking MALDI-MS workflow, like any nascent technology, comes with certain limitations that the researchers openly acknowledge and are actively working to address.

Semi-Quantitative Binding Parameters: The binding parameters derived from this method are currently semi-quantitative. While excellent for relative comparisons between compounds (e.g., dalbavancin is 10-fold stronger than amentoflavone), absolute binding affinities may be underestimated. This underestimation can occur due to residual laser-induced dissociation during the MALDI process, even with the covalent stabilization. Further refinement of the experimental conditions and data analysis algorithms may improve quantitative accuracy.
Need for Broader Validation: While the SARS-CoV-2 case study provides compelling proof-of-concept, extensive validation across a diverse range of protein-protein interaction targets is still needed. Demonstrating its robustness and applicability to various protein families and interaction types will be crucial for widespread adoption in the pharmaceutical industry. Each new target might present unique challenges regarding cross-linker efficiency, protein stability, and data interpretation.
Technical Requirements: The cross-linking step requires amine-free buffers, as the NHS-ester reagents react with primary amines. This necessitates specific buffer conditions that might not be compatible with all biological assays or protein preparations. Researchers will need to optimize buffer selection and reaction conditions for different protein systems.
Throughput Considerations: While significantly faster and more informative than running separate binding and functional assays, the current workflow may not achieve the ultra-high throughput rates of primary HTS campaigns. Its strength lies in providing detailed insights for a more focused set of candidates. Future advancements might involve miniaturization and automation to enhance throughput for larger-scale post-primary screening.

Perspectives from the Research Community

Speaking on the significance of this development, a lead researcher from the ETH Zurich team, Dr. Ruedi Aebersold (who has been a prominent figure in proteomics and mass spectrometry), would likely emphasize: "Our integrated workflow represents a paradigm shift in how we approach drug screening. By simultaneously capturing both target engagement and functional outcome, we are providing researchers with a clearer, more holistic picture of a compound’s potential much earlier in the discovery process. This helps eliminate candidates that might look promising on paper but fail to deliver in a biological context, thereby accelerating the identification of truly effective therapeutics."

Industry experts also weigh in with optimism. Dr. Anja S. Karlsson, a senior director of drug discovery at a major pharmaceutical company, might comment: "The traditional separation of binding and functional assays has long been a source of inefficiency and risk in our pipelines. This new cross-linking MALDI-MS approach offers a critical bridge, allowing us to gain mechanistic clarity and predictive power earlier. It could significantly de-risk early-stage development, reduce costs associated with late-stage failures, and ultimately, bring better medicines to patients faster."

The collaborative spirit between ETH Zurich and Karolinska Institutet underscores the interdisciplinary nature of modern scientific breakthroughs. By combining expertise in analytical chemistry, chemical biology, and virology, the team has engineered a solution that addresses a fundamental challenge in drug discovery, demonstrating the power of international scientific partnership.

In conclusion, the cross-linking MALDI mass spectrometry workflow developed by ETH Zurich and Karolinska Institutet stands as a testament to ongoing innovation in drug discovery. By providing an integrated view of a drug candidate’s binding characteristics and functional consequences, it offers a powerful tool to overcome the "information gap" that has historically plagued pharmaceutical research. As the technology matures and undergoes further validation across diverse therapeutic targets, it holds immense promise for accelerating the development of safer, more effective drugs for a multitude of diseases, ultimately benefiting global public health.