Cellular Target Engagement: Making Tricky Targets Druggable

Biochemical assays are a powerful tool for identifying compounds that bind to a purified protein in isolation, but that simplified environment can be misleading. Inside a living cell, a target protein may exist within a signaling complex with other proteins or be subject to post-translational modifications (PTMs) that impact its structure and function. This complexity is difficult to replicate with purified proteins.

Biochemical methods may also rely on truncated rather than full-length proteins, which can be challenging to express and purify. As a result, a compound that binds with high affinity to a recombinant protein in a microplate may not engage the native full-length protein inside a cell, where it may exist in complex with other proteins or carry PTMs. A further complication is that a compound may struggle to cross the cell membrane to reach its target at all.

Without a direct cellular readout of target engagement, researchers have no way to distinguish a genuinely promising compound from one that simply performs well in a cell-free system.

The result is wasted investment due to compounds that stall at the transition from biochemical to cellular validation.

The field has learned, sometimes at significant cost, that downstream functional or phenotypic readouts do not always confirm that a compound is working through its intended target. Demonstrating direct, on-target engagement in live cells early in the discovery process provides a much more reliable basis for compound prioritization during hit-to-lead phases. It also helps distinguish genuine target-mediated effects from off-target activity or non-specific cytotoxicity.

As drug discovery increasingly moves towards mechanistically complex and poorly characterized targets, this kind of direct cellular evidence is essential. Early target engagement data can reduce late-stage attrition of compounds by supporting more confident go/no-go decisions and helping R&D teams build a stronger case before progressing compounds into more costly, resource-intensive studies.

Estimates suggest that only around 20% of the human proteome (approximately 6,300 proteins) meets the criteria for conventional small molecule drug discovery based on structural and sequence similarity to known, validated targets. The remaining 80% includes uncharacterized proteins that lack well-defined binding pockets, have no established chemical probes, or belong to understudied families for which no reliable functional assays exist.

These so-called 'dark proteome' targets represent a substantial reservoir of potential therapeutic opportunity. Many are implicated in serious diseases but have remained out of reach simply because the tools to study them in a physiologically relevant context have not existed. Closing this gap, even partially, could open entirely new avenues for therapeutic intervention across disciplines like oncology, neurodegeneration, and rare diseases.

The TarSeer BRETSA Target Engagement System takes a fundamentally different approach to detecting compound–protein interactions. It measures ligand–protein interactions in intact cells using protein denaturation, via a bioluminescence resonance energy transfer (BRET)-based method. These cellular ligand–protein interactions are measured as a change in a target protein's thermal stability profile. This involves first expressing a target protein of interest fused to a luciferase tag in a cell. Next, a cell-permeable fluorescent probe is used to report on the protein's denaturation state when the cells are subjected to an increase in temperature. Importantly, this method differs from other cellular thermal shift assays that depend on protein aggregation. When a compound binds, it can alter protein unfolding, changing the BRET signal profile in a quantifiable way relative to the unliganded condition.

Depending on the compound and protein pair, binding can either destabilize or stabilize the protein. Using isothermal conditions, BRETSA excels at measuring compound potencies over a 5-log range, a sensitivity that makes it valuable for compound prioritization during hit-to-lead phases.

Crucially, this approach requires no prior knowledge of the target's binding pocket, target-specific tracer, or antibodies to detect the target protein—requirements that have historically made many proteins inaccessible. It complements the established NanoBRET® Target Engagement platform, which can quantify cellular affinity and binding mechanism for well-characterized targets, but depends on the availability of suitable fluorescent tracers.

BRETSA extends cellular target engagement capability to proteins where no such tracer can currently be made, effectively expanding the scope of accessible targets for early-stage drug discovery. The platform uses an addition-only method and is compatible with both 96- and 384-well formats, making it straightforward to incorporate into existing workflows.

A significant leap forward would be the ability to measure compound binding to dark proteome proteins—with no defined binding pocket, known ligands, structural data, or functional assays. BRETSA moves meaningfully in this direction, requiring only the sequence of the target proteins of interest and the ability to detect changes in the thermal stability of the protein upon compound binding. In principle, BRETSA could report on compound binding to many different types of target proteins that can be expressed with a luciferase tag in living cells. We continue to test proteins from different families to expand the range of target‒ligand interactions and potencies demonstrated with the system.

Expanding the range of targets accessible to cellular target engagement would represent a profound shift for drug discovery. Rather than being limited to well-characterized targets, R&D teams could advance novel therapeutic proteins into drug discovery projects with greater confidence. The ambition, then, is a cellular target engagement capability that is truly target-agnostic: one that works across a large breadth of the expressed proteome, regardless of whether a given protein has ever been studied in a drug discovery context before. Reaching that point would remove one of the fundamental bottlenecks between the genome and the clinic.

The introduction to this interview includes text that has been created with the assistance of generative AI and has undergone editorial review before publishing. Technology Networks' AI policy can be found here.