“Did you know that many well-known molecular glues are actually fragments?” 

Molecular glues have emerged as an exciting new class of agents in drug discovery, offering a novel mechanism to modulate protein function, both as protein degraders and for other therapeutic targets [1]. Unlike traditional drugs that inhibit or block protein activity, molecular glues enhance or induce new protein-protein interactions (PPIs), allowing them to activate biological pathways rather than suppressing them. For instance, the natural compound Rapamycin has been used as an immunosuppressant for decades. It binds at the interface of mTOR and FKBP12 and increases the interaction between these two proteins [2]. Similarly, Paclitaxel (Taxol®) acts as an allosteric molecular glue, enhancing the binding between alpha and beta tubulin, and is widely used in breast cancer treatment [3].  

Despite their potential, discovering new synthetic molecular glues has proven remarkably challenging. This is intriguing because some of the most impactful molecular glues are really small molecules. Thalidomide and Lenalidomide, which both target CRBN (cereblon) [4], are under 300 Da in molecular weight (Figure 1). Likewise, the plant hormone Auxin, a TIR1 glue degrader, is another example of a fragment-sized molecule with profound biological effects [5]. These examples underscore that molecular glues do not need to be large, complex natural products; they can be relatively simple, small molecules used in fragment-based drug design (FBDD) – provided we have the right methodologies tailored for finding them. 

Figure 1: Examples of well-known molecular glues. Natural products like Rapamycin and Paclitaxel are widely recognized as molecular glue. However, many small molecules less than 300 Da have profound biological effects, highlighting the diversity of molecular glues and the potential for small molecules as cooperative binders. (PDB: 1FAP[2], 1JFF[3], 5FQD[6], 2P1Q[5])

The main challenge in discovering new molecular glues is that the dynamics go beyond a simple 1:1 binding model. Unlike 1:1 binding events, molecular glues facilitate three-body interactions, where two proteins and the glue form a complex and all components can interact with each other. In this context cooperativity or enhancement of the binding affinity plays a crucial role [7]. The main idea in our approach is to prioritize cooperativity first in the search for new glues and then focus on raw binding affinity. Potent molecular glues exhibit high cooperativity, basically the amount of gluing potential, requiring some intrinsic affinity for at least one of the targets to observe this cooperativity [8]. This functions as a “double filter,” where hits need both high cooperativity and sufficient initial affinity. Promising glue-like molecules may be overlooked due to a lack of sensitivity. Since there are decades of experience in medicinal chemistry in optimizing binding affinity, it makes sense to screen for cooperativity and select for that first, and start with hits that have high cooperativity.  

Figure 2: Cascade of biophysical techniques optimized for molecular glue discovery. New Molecular glues can actually be found using the same biophysical methods as traditional hits. It only requires a slight adjustment in assay setup and data interpretation. By focusing on cooperativity first, a robust pipeline is formed combining SPR, NMR and X-ray crystallography. The combination of immobilized and in solution techniques, provide a robust workflow in profiling new molecular glues.  

At ZoBio, we employ a cascade of biophysical techniques (Figure 2), such as Surface Plasmon Resonance (SPR) and NMR, in a format we call TINS, to identify new molecular glues. These assays, that are widely used, are highly sensitive and have been optimized to specifically detect cooperative binding. Our process typically begins with a special SPR assay designed for screening cooperativity, followed by assessment of affinity and ligand-observed NMR for confirmation in solution. 2D-titrations (cooperativity assays) are performed for further, more detailed profiling and ultimately structural biology information is obtained via a combination of X-ray crystallography, NMR, or cryo-EM for binary and ternary structures leading then to structure-based drug design. With some small alterations, finding new molecular glues is not so different from finding traditional leads.