Deep Learning Enables Discovery of a Short Nuclear Targeting Peptide for Efficient Delivery of Antisense Oligomers.

Therapeutic macromolecules such as proteins and oligonucleotides can be highly efficacious but are often limited to extracellular targets due to the cell's impermeable membrane. Cell-penetrating peptides (CPPs) are able to deliver such macromolecules into cells, but limited structure-activity relationships and inconsistent literature reports make it difficult to design effective CPPs for a given cargo. For example, polyarginine motifs are common in CPPs, promoting cell uptake at the expense of systemic toxicity. Machine learning may be able to address this challenge by bridging gaps between experimental data in order to discern sequence-activity relationships that evade our intuition. Our earlier data set and deep learning model led to the design of miniproteins (>40 amino acids) for antisense delivery. Here, we leveraged and expanded our model with data augmentation in the short CPP sequence space of the data set to extrapolate and discover short, low-arginine-content CPPs that would be easier to synthesize and amenable to rapid conjugation to desired cargo, and with minimal in vivo toxicity. The lead predicted peptide, termed P6, is as active as a polyarginine CPP for the delivery of an antisense oligomer, while having only one arginine side chain and 18 total residues. We determined the pentalysine motif and the C-terminal cysteine of P6 to be the main drivers of activity. The antisense conjugate was able to enhance corrective splicing in an animal model to produce functional eGFP in heart tissue in vivo while remaining nontoxic up to a dose of 60 mg/kg. In addition, P6 was able to deliver an enzyme to the cytosol of cells. Our findings suggest that, given a data set of long CPPs, we can discover by extrapolation short, active sequences that deliver antisense oligomers.

Parallel Automated Flow Synthesis of Covalent Protein Complexes That Can Inhibit MYC-Driven Transcription.

Dysregulation of the transcription factor MYC is involved in many human cancers. The dimeric transcription factor complexes of MYC/MAX and MAX/MAX activate or inhibit, respectively, gene transcription upon binding to the same enhancer box DNA. Targeting these complexes in cancer is a long-standing challenge. Inspired by the inhibitory activity of the MAX/MAX dimer, we engineered covalently linked, synthetic homo- and heterodimeric protein complexes to attenuate oncogenic MYC-driven transcription. We prepared the covalent protein complexes (∼20 kDa, 167-231 residues) in a single shot via parallel automated flow synthesis in hours. The stabilized covalent dimers display DNA binding activity, are intrinsically cell-penetrant, and inhibit cancer cell proliferation in different cell lines. RNA sequencing and gene set enrichment analysis in A549 cancer cells confirmed that the synthetic dimers interfere with MYC-driven transcription. Our results demonstrate the potential of automated flow technology to rapidly deliver engineered synthetic protein complex mimetics that can serve as a starting point in developing inhibitors of MYC-driven cancer cell growth.

Engineering Bioactive Dimeric Transcription Factor Analogs via Palladium Rebound Reagents.

Transcription factors (TF), such as Myc, are proteins implicated in disease pathogenesis, with dysregulation of Myc expression in 50% of all human cancers. Still, targeting Myc remains a challenge due to the lack of small molecule binding pockets in the tertiary structure. Here, we report synthetic covalently linked TF mimetics that inhibit oncogenic Myc-driven transcription by antagonistic binding of the target DNA-binding site. We combined automated flow peptide chemistry with palladium(II) oxidative addition complexes (OACs) to engineer covalent protein dimers derived from the DNA-binding domains of Myc, Max, and Omomyc TF analogs. Palladium-mediated cross-coupling of synthesized protein monomers resulted in milligram quantities of seven different covalent homo- and heterodimers. The covalent helical dimers were found to bind DNA and exhibited improved thermal stability. Cell-based studies revealed the Max-Max covalent dimer is cell-penetrating and interfered with Myc-dependent gene transcription resulting in reduced cancer cell proliferation (EC50 of 6 µM in HeLa). RNA sequencing and gene analysis of extracted RNA from treated cancer cells confirmed that the covalent Max-Max homodimer interferes with Myc-dependent transcription. Flow chemistry, combined with palladium(II) OACs, has enabled a practical strategy to generate new bioactive compounds to inhibit tumor cell proliferation.

Secondary Amino Alcohols: Traceless Cleavable Linkers for Use in Affinity Capture and Release.

Capture and release of peptides is often a critical operation in the pathway to discovering materials with novel functions. However, the best methods for efficient capture impede facile release. To overcome this challenge, we report linkers based on secondary amino alcohols for the release of peptides after capture. These amino alcohols are based on serine (seramox) or isoserine (isoseramox) and can be incorporated into peptides during solid-phase peptide synthesis through reductive amination. Both linkers are quantitatively cleaved within minutes under NaIO4 treatment. Cleavage of isoseramox produced a native peptide N-terminus. This linker also showed broad substrate compatibility; incorporation into a synthetic peptide library resulted in the identification of all sequences by nanoLC-MS/MS. The linkers are cell compatible; a cell-penetrating peptide that contained this linker was efficiently captured and identified after uptake into cells. These findings suggest that such secondary amino alcohol based linkers might be suitable tools for peptide-discovery platforms.