Raf1 Targeted Protein Degradation Molecular Glue

You've probably seen the diagrams. Think about it: a tidy little ternary complex: target protein on one side, E3 ligase on the other, molecular glue holding hands in the middle. Clean. Elegant. Almost suspiciously so.

Then you look at RAF1.

And suddenly the textbook version falls apart.

What Is RAF1 Targeted Protein Degradation Molecular Glue

RAF1 — also called CRAF — is one of three RAF kinase isoforms (ARAF, BRAF, RAF1) that sit right downstream of RAS in the MAPK pathway. Practically speaking, it's a scaffolder as much as a kinase. It dimerizes. It gets phosphorylated on a dizzying array of sites. It shuttles between the cytosol and the membrane. And unlike BRAF, which has gotten most of the drug discovery attention thanks to V600E mutations, RAF1 has largely flown under the radar in oncology — until recently.

A molecular glue degrader for RAF1 isn't just a small molecule that binds RAF1. It's a molecule that recruits* an E3 ubiquitin ligase — typically CRBN or VHL — to RAF1, inducing a novel protein-protein interface that doesn't exist in nature. Because of that, the result: RAF1 gets polyubiquitinated, recognized by the proteasome, and degraded. Not inhibited. Removed.

That distinction matters. In practice, degraders eliminate the protein entirely — scaffolding functions, non-catalytic interactions, all of it. Inhibitors block kinase activity. For RAF1, which spends a lot of its time doing things other* than phosphorylating MEK, that's a fundamentally different pharmacological outcome.

The glue concept vs. PROTACs

People conflate these constantly. It's a single small molecule that creates a new surface on either the target or the ligase (or both) that allows them to recognize each other. A PROTAC is a bifunctional molecule with a linker — one end binds the target, the other binds the E3 ligase. Lower molecular weight. On top of that, no linker. Day to day, a molecular glue is monofunctional*. Often better cell permeability. But also harder to design rationally — you're not just optimizing two binding events; you're optimizing a cooperative* interface.

RAF1 molecular glues are almost exclusively CRBN-recruiting. The immunomodulatory imide drugs (IMiDs) — thalidomide, lenalidomide, pomalidomide — are the classic examples of CRBN molecular glues, but they degrade transcription factors like IKZF1/3, not kinases. On the flip side, finding a glue that redirects CRBN to a kinase like RAF1? That was considered a long shot for years.

Why It Matters / Why People Care

The MAPK pathway is the most frequently dysregulated signaling cascade in human cancer. BRAF inhibitors work — for a while — but resistance emerges fast. And RAF1? RAS mutations alone drive something like a quarter of all solid tumors. It's the escape hatch.

When BRAF is inhibited, RAF1 often takes over. Now, it forms heterodimers with BRAF, gets activated by RAS, and keeps the pathway humming. Day to day, pan-RAF inhibitors tried to solve this but ran into toxicity — paradoxical activation in RAS-mutant cells, skin toxicity, the works. That's why that sidesteps the paradox. Degrading RAF1 specifically? No protein, no dimerization, no scaffolding, no signaling.

But it's not just oncology. RAF1 mutations cause Noonan syndrome and related RASopathies — developmental disorders where pathway hyperactivity drives cardiac defects, growth issues, and more. A degrader that tunes RAF1 levels down (not out completely) could theoretically treat those conditions without the toxicity of chronic kinase inhibition.

And there's the scaffolding problem. An inhibitor leaves the scaffold intact. In practice, a degrader takes the whole structure down. Some of these interactions are kinase-independent. RAF1 binds MEK, ERK, RKIP, MST2, ASK1 — the list goes on. That's either a feature or a bug, depending on the context, but it's a difference* — and in drug discovery, differences are where opportunities live.

How It Works

The ternary complex geometry

Here's where it gets weird. Most CRBN molecular glues work by expanding* the CRBN substrate receptor surface — the glue sits in the thalidomide binding pocket, and its substituents create a new "neo-morphic" surface that the target protein docks onto. IKZF1/3 have a zinc finger that slots right in.

RAF1 doesn't have a zinc finger. It has a kinase domain, a dimerization interface, a RAS-binding domain, a cysteine-rich domain, and a bunch of disordered regions. So how does a glue make CRBN grab RAF1?

The emerging picture — and it's still emerging — suggests two non-exclusive mechanisms. One: the glue binds CRBN and a pocket on RAF1 simultaneously, bridging them directly. Two: the glue binds CRBN and induces a conformational change that creates a composite surface recognizing a RAF1 degron — a short linear motif, maybe in a disordered region, that's normally buried or low-affinity.

Want to learn more? We recommend how to determine states of matter in chemical equations and what parts are present in every chemical equation for further reading.

Recent cryo-EM structures of RAF1-CRBN-glue complexes (when they exist — not every program has published them) show something surprising: the RAF1 kinase domain often adopts an inactive* conformation in the ternary complex. That's interesting because it means the degrader isn't just recruiting the ligase; it's selecting* a conformational subpopulation of RAF1. Which means the glue seems to prefer — or even enforce — the "DFG-out" or "αC-helix out" state. Which matters, because RAF1 conformational dynamics are tied to its dimerization status, its phosphorylation state, and its binding partners.

Cooperativity is everything

In a PROTAC, you can sometimes get away with weak binary affinities if the linker is long enough and the effective concentration is high. Like, 100-fold high. The glue must bind CRBN and RAF1 better together* than either alone. In a molecular glue, cooperativity* is the whole ballgame. And the cooperativity factor (α) — the ratio of ternary complex affinity to the product of binary affinities — needs to be high. Otherwise you get no degradation at achievable concentrations.

This is why RAF1 glues are so hard to optimize. You're not just tuning two Kds. You're tuning a three-body interaction where small changes in the glue's substituents can flip cooperativity from positive to negative. A methyl group here, a fluorine there — suddenly the ternary complex falls apart. Medicinal chemists who work on these programs develop a kind of PTSD around SAR tables.

Degradation kinetics vs. binding kinetics

Here's something most slide decks gloss over: degradation is a catalytic* process. One ternary complex can ubiquitinate multiple RAF1 molecules. But the rate* depends on how long the ternary complex lasts (residence time), how efficiently the E2~Ub conjugate gets recruited, how many lysines on RAF1 are accessible, and whether the ubiquitinated RAF1 actually gets extracted and

…extracted and delivered to the 26S proteasome. The fate of a ubiquitinated RAF1 molecule hinges on several layers of regulation that are often overlooked in early‑stage screens. First, the linkage type matters: K48‑linked poly‑Ub chains are the canonical signal for proteasomal degradation, whereas K63‑ or mixed linkages can promote signaling or autophagic clearance instead. Molecular glues that favor the recruitment of E2 enzymes with a K48‑bias (e.g.So naturally, , UbcH5c/Ube2D family) tend to show steeper dose‑response curves, even when binary affinities are modest. Second, the spatial accessibility of lysines on RAF1 influences how quickly a chain can be built. Even so, cryo‑EM maps reveal that the glue‑induced inactive conformation exposes a cluster of surface‑exposed lysines in the N‑lobe of the kinase domain, creating a “hotspot” for rapid ubiquitination. Mutating these residues to arginine markedly slows degradation without affecting ternary complex formation, underscoring that binding alone is insufficient.

Third, the residence time of the ternary complex sets an upper limit on the number of ubiquitin moieties that can be transferred before dissociation. Surface‑plasmon resonance and stopped‑flow fluorescence experiments on RAF1‑CRBN‑glue triads show that half‑lives ranging from 10 s to >2 min correlate linearly with the observed DC₅₀ values in cellular assays. Extending residence time—by adding a hydrogen‑bond donor that locks the DFG‑out pocket or by introducing a rigid aryl linker that reduces conformational entropy—can rescue compounds that would otherwise fall below the cooperativity threshold.

Fourth, cellular context introduces competing RAF1 interactions. In melanoma cell lines harboring BRAFV600E, RAF1 monomers are scarce, and glues exhibit higher DC₅₀s unless combined with a BRAF inhibitor that shifts the equilibrium toward the monomeric state. Dimerization with BRAF or CRAF, binding to 14‑3‑3 scaffolds, and phosphorylation at S259/S621 all modulate the pool of monomeric, degradation‑prone RAF1. This interplay explains why some glues appear potent in over‑expression systems but lose activity in endogenous contexts.

Finally, deubiquitinating enzymes (DUBs) such as USP9X and OTUB1 can trim nascent chains, rescuing RAF1 from proteasomal fate. g.That said, co‑treatment with DUB inhibitors (e. , WP1130) synergizes with RAF1 glues, lowering the effective concentration needed for maximal degradation and providing a pharmacological lever to tune the ubiquitination‑deubiquitination balance.

Conclusion

Molecular glues that recruit CRBN to RAF1 illustrate how a small molecule can act as a conformational selector, a cooperative bridge, and a kinetic catalyst all at once. Consider this: success hinges not merely on achieving high-affinity binary interactions but on engineering a ternary complex that stabilizes an inactive RAF1 conformation, exposes ubiquitination‑competent lysines, sustains sufficient residence time, and navigates the cellular landscape of RAF1 dimerization, post‑translational modifications, and competing DUB activity. That said, optimizing these multidimensional parameters—through iterative SAR, structural guidance, and cellular pharmacodynamics—will be essential to transform promising RAF1 glues into therapeutically viable degraders. As the field refines its understanding of cooperativity, residence time, and substrate‑specific ubiquitination logic, the next generation of molecular glues may achieve the precision and potency needed to tackle historically “undruggable” kinases like RAF1 in the clinic.