Matthew Robers, an Associate Research Director at Promega Corporation, has developed novel biophysical technologies to assess intracellular target engagement, residence time, and drug cooperativity, including the creation of the NanoBRET Target Engagement platform.
Credit: PROMEGA
Studying how drugs interact with their targets inside living cells has long been a challenge in drug discovery. Traditional biochemical assays often rely on purified proteins in test tubes, giving only a limited picture of what happens in the complex environment of a cell. Bioluminescence resonance energy transfer (BRET) offers a powerful alternative. By harnessing naturally occurring energy transfer between a luminescent donor and a fluorescent acceptor, BRET allows researchers to monitor protein-protein interactions and small molecule-target engagement in real time, in the correct cellular location, and under native conditions.
BRET-based assays are fast, sensitive, cost-effective, and adaptable to high-throughput workflows. They have been used not only to track protein function, phosphorylation, and protease activity in cultured cells, but also in small animal studies to screen genetic and chemical modulators. Unlike conventional biochemical approaches, BRET provides spatio-temporal information, detects transient interactions, and avoids disruptive cell lysis, making it well suited for understanding complex cellular processes and drug mechanisms in living systems.
To explore the opportunities and challenges of these live-cell approaches, DDN spoke with Matthew Robers, Associate Research Director at Promega Corporation.
Why have cellular target engagement assays traditionally lagged behind biochemical assays in early discovery workflows?
Traditional drug discovery has focused on targets that are intrinsically easy to drug, and in many cases, targets that are easy to drug are also often amenable to more biochemical types of analysis using purified protein. For example, if you want to target an enzymatic protein, there are usually high-throughput compatible enzymatic assays to do that, and oftentimes you can identify ligands without a lot of labor in that process. So, simplicity, scalability, and off-the-shelf availability have made biochemical assays part of a lot of workflows.
Because of this traditional focus on proteins that were intrinsically easier to study, biochemical systems weren’t as much of a limitation until more recently. But as pharmacology gets increasingly nuanced, applying an off-the-shelf biochemical tool is no longer a simple option. When you get into proteins that are large, multi-complex proteins, or proteins that come together to create a druggable surface, assay development also becomes more complex. It’s becoming much harder to rely on the same traditional methods.
Are there specific classes of proteins that have been essentially “invisible” to traditional cellular assays, and why?
Definitely. I think the targets that are most difficult to study in cells are the ones that aren’t coupled to any sort of immediate proximal downstream cellular response.
For example, if you're trying to target a protein that's directly coupled to the release of a cytokine or the release of some proximal cell health biomarker, in general, these kinds of targets can often be measured indirectly using that cellular response as a proxy for the target activity. But as you get into more challenging targets, the connection between the target and the phenotype is more difficult to pinpoint. Multiple levels of signaling have to happen. And the more distal you get between the target itself and the biological response – the assay endpoint – the harder it is to rely upon a proxy. A lot of those targets are super valuable, but using a proxy-based assay just does not give you enough specific information. There are too many divergent paths from that target that lead to ambiguity in using proxy-based assays.
What are luminescence- or BRET-based target engagement assays?
BRET-based target engagement assays are essentially biophysical tools that measure molecular proximity in living cells. BRET is based on a pretty elegant principle: When two proteins or a protein and a small molecule get close together, energy can transfer from a luminescent donor to a fluorescent acceptor, and we can detect that as a signal. We’re using this energy transfer as a molecular ruler to tell us when a drug molecule is engaging with its target protein inside a live cell.
Over the past decade, we’ve used BRET to develop robust probe-based assays that can measure target engagement across a wide range of protein classes. In our NanoBRET system, small, cell-permeable fluorescent tracers compete with test compounds for binding to a target protein. When the tracer is bound, you get a BRET signal, and when your compound displaces it, the signal goes away. It’s a competition-based assay that happens in living cells.
More recently, we combined BRET with a target-agnostic type of protein denaturation assay similar to thermal shift. Instead of using a tracer, our BRET shift assay exploits a conformational difference in the target protein that occurs in the presence of a bound ligand relative to the unbound target. We’re essentially using BRET to report on that structural change, which means we can measure engagement without needing to develop a competitive tracer for every target. Both approaches give us that critical biophysical readout of engagement in cells, but they have different advantages depending on the target and the pharmacology you’re trying to interrogate.
How do these assays differ from traditional biochemical or functional assays?
The big difference between BRET-based assays and traditional biochemical assays is the environment where you’re making the measurement. With biochemical assays, you’re typically working with purified protein in a test tube. It’s a reductionist approach where you’ve stripped away all the cellular complexity to get at that one protein target. But with BRET, we’re measuring that same interaction inside a living cell.
That’s a fundamentally different environment. You’ve got all the cellular machinery present, you’ve got the right cofactors and metabolites, the protein is in its native conformation, potentially in complexes with other proteins. You’re capturing pharmacology that might only exist in that cellular context. So biochemical assays are asking, “Can this molecule bind to my purified target” whereas live-cell BRET is asking, “Can this molecule engage my target in the environment where it actually needs to work?” Those are related questions, but they’re not the same question.
Compared to functional assays, I think what makes BRET particularly interesting is that while it’s only measuring molecular proximity, we can get pretty creative about how we deploy that readout. You can use it for looking at protein-protein and protein-small molecule interactions. BRET can also function as a biophysical readout, using optical probes and fluorescent reagents to monitor protein behavior in real time. We can exploit BRET in cellular biophysics in a way that gives us nuanced information about drug-target interactions at proteins and complexes.
This is a very important complement to functional assays. When you have a functional assay and want to know if that binding influences a functional pharmacology or a cellular response, leveraging a functional assay in combination with BRET is very powerful. The problems arise when you lack a target-proximal functional assay to tell you whether engagement has a functional outcome. That’s often a gap. But my research group focuses on developing biophysical tools that measure small molecule-target interactions in a cellular environment. That’s always going to be a critical question in drug discovery workflows.
How does measuring weak or early-binding interactions in cells influence the way researchers prioritize chemical matter?
There are two ways of looking at that. First, when you're trying to rapidly generate a lead compound, identifying molecules that are bioactive in cells early in the discovery process can massively accelerate a project. Without this information, early-stage activity is often ambiguous. For example, let's say you're generating molecules that might be on the fringe of Lipinski's rule of five and have poor cell permeability or other properties that limit their bioactivity in a cellular environment. Detecting compounds that are truly active in cells as early as possible de-risks the project and reduces the effort required to develop them into bioactive agents later.
On the other hand, there might be scenarios where the cell enhances the pharmacology or enhances the binding characteristics between the drug and the target protein. Going directly into cells early in discovery can therefore increase your chances of identifying true binders. This is especially important for targets whose activity depends on protein complexes or specific conformational states that aren’t captured in biochemical or purified systems. By starting in cells, you may be working in a privileged environment where pharmacology is evident that would otherwise be missed in a cell-free assay.
To summarize, there can be two distinct benefits. You could be getting to legitimate molecules faster, or you could be seeing things that are invisible in other assay formats. I think there’s a huge opportunity to exploit both, helping you rule in and rule out, casting a wider net for certain types of pharmacology, or accelerating the path to bioactive agents.
In what ways does the complexity of different cellular compartments complicate target engagement measurements?
When I first started working in this field, I thought of cells as just a bag of enzymes. And if you can get past the plastic bag, you’re basically doing a biochemical assay inside of it. In reality, cells are nothing like that. They’re way more complicated than we often envision.
The inside of a cell is a crowded, super dense environment where the gradient behavior of metabolites and proteins definitely influences target engagement. We know now that a lot of metabolites aren’t at fixed concentrations across the cell. If there is a critical metabolite that either drives or interferes with engagement at your target, your ability to access the binding pocket can be affected by where the interaction occurs within that gradient. It has to compete with or be enhanced by the presence of that metabolite, and you can’t simulate that in a test tube.
The detection chemistry we’ve recently debuted is compatible with as many cellular compartments as we’ve been able to study, including targets in the nucleus, chromatin, mitochondria, Golgi, plasma membrane, and more. We have seen that there are no compartments off limits with our method. But the million-dollar question, in my opinion, is how those compartments influence or enhance drug pharmacology. For example, do you have an environment in the cell that’s privileged because some particular metabolite or complex there is providing a benefit to target engagement? We’ll be learning more about that over time.
The good news is that there is no cellular compartment that's fundamentally off limits with our technology. That’s an area where we feel particularly strong, and we’re excited to see how this enables exploration of how organelles and other factors influence target engagement with small molecules.
How critical is it to directly measure target engagement when assessing compound selectivity and mechanism of action?
It’s absolutely critical. When you’re trying to understand selectivity or mechanism of action, you need to know with certainty that your compound is actually binding to the target you think it’s binding to. Functional readouts can tell you that something is happening, but they can’t tell you what is happening or where it’s happening. You may see a cellular phenotype that you think is coming from on-target engagement, but it could actually be an off-target effect or some indirect consequence of hitting multiple proteins.
The selectivity question is extremely important because just looking at direct engagement to one protein doesn’t always give you the full picture with things like polypharmacology. Analyzing selectivity in a cell is critical. There are very few drugs that work through single targets; they typically work through secondary effects or polypharmacology, so understanding these interactions in cells is pivotal to defining a drug’s mode of action. Being able to directly measure engagement lets you map out which proteins your compound is actually hitting, and at what concentrations. That’s information you can’t reliably get from functional assays alone.
Establishing mode of action in cells is also critical, especially when dealing with the types of complexes and metabolites mentioned earlier. Measuring target engagement inside a cell is important for defining whether a metabolite or other biomolecule will impact potency, either by enhancing or interfering. That’s a huge opportunity to work with cells to ask those questions. Overall, you need that direct measurement of engagement to nail down the mechanism.
How do the technical requirements differ when measuring small molecules versus larger biologics in live-cell target engagement assays?
There are a lot of biophysical and functional assays out there that are incredibly enabling for large molecule research. Oftentimes, when you're studying biologics, you can couple the activity of that biologic to a very clear-cut biological outcome. It's not to say that the assay development is easy in that space — it could actually be incredibly complex — but the biomarkers that you analyze as a consequence of that biologic’s activity are often very well defined. This can really inform the design and development of large molecules in the biologic space.
Small molecules are a bit different. In many cases, a functional assay might not be reliable or even available to dissect the mode of action. A binding assay is often the only way to establish whether your drug is working against its target. With biologics, you can ask, “Is my drug working?” with fairly clear-cut outcomes, often guided by existing precedent. With small molecules, functional assays are less clear, which is why direct measurements of binding become so critical.
How might earlier access to live-cell engagement data shift decision-making during hit validation and lead optimization?
Drug discovery is not about targeting proteins in isolation. Your drug will ultimately have to work inside a cell. The best thing we can do is try to get into a cellular environment as quickly as possible. The earlier you can get that data, the faster you de-risk your entire program.
It's very hard for me to rationalize why you'd be satisfied with biochemical data to tell you whether you have a lead drug molecule. I believe the vast majority of drug hunters would rather know they have a bioactive drug molecule that can bind to a target in a cell than simply bind to a fragment of a target biochemically. Getting past the cell hurdle and into a relevant environment as quickly as possible would be a big de-risker in your entire hit-to-lead campaign.
