Thrilled to share my freshly "baked" Baker lab paper. We took a bottom-up approach to designing transcriptional regulators. Instead of relying on natural activation domains, we designed compact proteins that recruit specific cofactors. Activators for now, repressors to come!
Excited to share that our MYC mutagenesis paper is now online at @CellCellPress! Free access link through Dec 24 authors.elsevier.com/a/1o316…. Special thanks to our handling editor Jude @JudeNicho and the reviewers for their support and feedback throughout the process!
Does the noncoding genome actually carry more genetic information than coding seqs? Motivated by this question we mutated every bp in the 10kb MYC locus. Results are even more exciting: Decoding the MYC locus reveals a druggable ultraconserved RNA element biorxiv.org/content/10.64898…
Our latest, sent on behalf of the outstanding but X-less Aidan Keith: OPTIMuS PRIME-Del uses paired prime editing to program targeted deletions at candidate enhancers, while concurrently replacing them w/ a molecular tag to verify & quantify deletions 1/7
biorxiv.org/content/10.64898…
Numerous biotech innovations started with adapting a protein to its non-native context, which often requires substantial experimental optimizations. Here’s a breakdown by our lead, @Xiaowei0402, on a versatile, efficient computational framework to facilitate such adaptations.
We deimmunized a clinical CD19 CAR binder with experimentally detected HLA-I epitope depletion while retaining tumor killing, and rescued a secreted protease to function in the cytosol.
Both done with one unified design framework: ProVADA.
How it works and what we tested 🧵 (1/11)
Avengers assemble: 1 year BakerLab + NVIDIA GenAIR = 🔥 rfd4-proteina.org
Presenting RFD4-Proteina, our latest protein design model, significantly better and faster than everything we compared against, both in silico and in the lab! 1/30
We hope this framework will open new opportunities to dissect and control both gene activation and silencing with precision, and, more ambitiously, program cell states and functions.
Our compact binders (~65–85 aa) also proved to be potent CRISPRa effectors, outperforming VPR at several endogenous loci. A combination of these binders further boosts transcription activity.
To test this experimentally, we designed p300 KIX binders with affinities ranging from nano- to micromolar and measured their transcriptional activity. We found a switch-like response around 500 nM! Stronger binding beyond this threshold provided little additional activation.
Kinetic simulations suggest that two TADs engaging distinct interfaces of the same cofactor or cofactor complex can cooperate to achieve efficient recruitment, even when individual interactions are weak.
Guided by these predicted interfaces, we used RFdiffusion to design proteins that specifically bind cofactors. Using a functional screen in cells, we identified more than 300 binders across 10 cofactors! These binders allow us to recruit cofactors individually or in combination.
We first used AF3 to systematically map interactions between TADs and transcriptional cofactors. Interestingly, the predictions revealed distinct cofactor-binding preferences. Some cofactors were highly selective, whereas others showed broader interaction profiles.
TFs recruit cofactors through transcriptional activation domains (TADs). These domains are often intrinsically disordered and engage cofactors through weak, promiscuous interactions. This makes it difficult to dissect and precisely control gene activation.
Every cell in our body comes from 1 fertilized egg — & now we can trace that dev. #HHMIInvestigators Jonathan Weissman & Jay Shendure & teams have created new tools that generate maps of the origin & outcome of vast numbers of cells in a developing mouse: bit.ly/4dVhnhe