MYC, BPTF, HDAC6: The 'Undruggable' List Keeps Shrinking

MYC finally has binders

In a bioRxiv preprint posted October 16, McNamara-Bordewick and colleagues — with David Baker among the authors — report de novo designed proteins that bind the MYC oncoprotein with nanomolar affinity. The designed binders disrupt MYC-MAX interactions and remodel MYC-associated protein complexes. Binder expression decreased cell proliferation across diverse MYC-dependent tumor models — hematopoietic, liver, and pancreatic — and in orthotopic pancreatic cancer models in mice, binder-expressing tumor cells were selectively depleted. m:…6673

BPTF gets a bivalent

A second preprint attacks an epigenetic reader protein. The BPTF PHD-bromodomain cassette specifically recognizes the H3K4me3K14ac histone modification, and ligand binding induces a conformational change between the PHD and bromodomain modules. The authors designed MJ-1, a bivalent peptide–small molecule conjugate that inhibits both reader domains by exploiting that ligand-induced rearrangement. The new part of this story, offered as reasoning: instead of inhibiting one reader domain, MJ-1 engages both at once by exploiting the rearrangement. m:…4323

HDAC6 gets a picture

The third preprint supplies atomic detail: a 1.9 Å crystal structure of the HDAC6 CD2 domain bound to macrocyclic nonapeptide 1, which inhibits HDAC6 with nanomolar potency. The structure shows zinc coordination by a noncanonical thiolate-containing amino acid and a network of direct and water-mediated hydrogen bonds that matches previously reported tetrapeptide and octapeptide macrocycles — a shared structural scaffold, now with a high-resolution picture to design against. m:…6485

What thins is the pitch, not the work

Step back and the week looks coordinated even though it is not. Three different hard-target classes — the MYC transcription factor, a dynamic tandem epigenetic reader, a histone deacetylase — each fell to a different modality: a de novo designed protein, a bivalent peptide–small molecule conjugate, and a macrocyclic peptide solved at 1.9 Å. The common thread is that none of the three is a conventional small molecule doing the conventional thing. MYC was bound by something built for it from scratch; BPTF was gripped at two domains at once; HDAC6 was inhibited by a macrocycle whose zinc-coordination chemistry a medicinal chemist would not have drawn on a whiteboard. The undruggable label, in each case, described the limits of one modality, not of the target. m:…6673 m:…4323 m:…6485

The reasoning a VC should take from the trio: undruggable is increasingly a statement about modality imagination rather than chemistry. De novo binders, bivalent conjugates, and macrocycles each cracked something a conventional small molecule could not — but none of this is a drug yet. All three are preprints with no clinical validation: the MYC work is in mouse and cell models, the BPTF paper reports no in vivo efficacy data, and the HDAC6 paper is a structure. The diligence question shifts from whether anyone can bind the target to whether the modality can reach it, selectively, in a patient — a harder and more expensive question, and exactly the one worth underwriting. m:…6673 m:…4323 m:…6485

For diligence, the practical screen this suggests: when a pitch deck says undruggable, ask which modalities have actually been tried. If the answer is small molecules, the target may be undrugged rather than undruggable — these three preprints are the exhibit. Then ask the harder questions these papers leave open: delivery, selectivity across protein families, and safety in animals. The MYC binders were expressed in the tumor cells rather than delivered as drugs; the BPTF conjugate has no animal data; the HDAC6 structure is a picture, not a pharmacokinetic profile. The alpha is in the modality; the risk is in everything after binding. m:…6673 m:…4323 m:…6485