Biology subjects

Silverstein, R. A.

Publications and source records attributed to Silverstein, R. A..

5 recordsLinked to original sources

bioRxiv2026

Mechanistic machine learning enables interpretable and generalizable prediction of prime editing outcomes

Although prime editing (PE) can effect virtually any specified local change to genomic DNA in living systems, its efficient application currently requires extensive optimization of prime editing guide RNA (pegRNA) sequences. We present OptiPrime, a machine learning model of PE efficiency based on our current understanding of the mechanism of prime editing. OptiPrime achieves state-of-the-art accuracy on PE efficiency prediction and also enables prediction of nicking guide RNA (PE3) and dual pegRNA (twinPE) outcomes. We validated that OptiPrime has learned the determinants of mammalian mismatch repair (MMR), and is therefore well suited for nominating MMR-evasive silent edits that improve PE efficiency. We demonstrate the utility of OptiPrime in a variety of prospective therapeutic contexts, including in primary human and mouse cells. Finally, we show how OptiPrime can be used to achieve highly streamlined and efficient in vivo correction of a pathogenic mutation in the brain of a mouse model of KIF1A-associated neurological disorder.

biochemistry↗

bioRxiv2025

Engineering of CRISPR-Cas PAM recognition using deep learning of vast evolutionary data

CRISPR-Cas enzymes must recognize a protospacer-adjacent motif (PAM) to edit a genomic site, significantly limiting the range of targetable sequences in a genome. Machine learning-based protein engineering provides a powerful solution to efficiently generate Cas protein variants tailored to recognize specific PAMs. Here, we present Protein2PAM, an evolution-informed deep learning model trained on a dataset of over 45,000 CRISPR-Cas PAMs. Protein2PAM rapidly and accurately predicts PAM specificity directly from Cas proteins across Type I, II, and V CRISPR-Cas systems. Using in silico deep mutational scanning, we demonstrate that the model can identify residues critical for PAM recognition in Cas9 without utilizing structural information. As a proof of concept for protein engineering, we employ Protein2PAM to computationally evolve Nme1Cas9, generating variants with broadened PAM recognition and up to a 50-fold increase in PAM cleavage rates compared to the wild-type under in vitro conditions. This work represents the first successful application of machine learning to achieve customization of Cas enzymes for alternate PAM recognition, paving the way for personalized genome editing.

synthetic biology↗

bioRxiv2024

In vivo Treatment of a Severe Vascular Disease via a Bespoke CRISPR-Cas9 Base Editor

Genetic vascular disorders are prevalent diseases that have diverse etiologies and few treatment options. Pathogenic missense mutations in the alpha actin isotype 2 gene (ACTA2) primarily affect smooth muscle cell (SMC) function and cause multisystemic smooth muscle dysfunction syndrome (MSMDS), a genetic vasculopathy that is associated with stroke, aortic dissection, and death in childhood. Here, we explored genome editing to correct the most common MSMDS-causative mutation ACTA2 R179H. In a first-in-kind approach, we performed mutation-specific protein engineering to develop a bespoke CRISPR-Cas9 enzyme with enhanced on-target activity against the R179H sequence. To directly correct the R179H mutation, we screened dozens of configurations of base editors (comprised of Cas9 enzymes, deaminases, and gRNAs) to develop a highly precise corrective A-to-G edit with minimal deleterious bystander editing that is otherwise prevalent when using wild-type SpCas9 base editors. We then created a murine model of MSMDS that exhibits phenotypes consistent with human patients, including vasculopathy and premature death, to explore the in vivo therapeutic potential of this base editing strategy. Delivery of the customized base editor via an engineered SMC-tropic adeno-associated virus (AAV-PR) vector substantially prolonged survival and rescued systemic phenotypes across the lifespan of MSMDS mice, including in the vasculature, aorta, and brain. Together, our optimization of a customized base editor highlights how bespoke CRISPR-Cas enzymes can enhance on-target correction while minimizing bystander edits, culminating in a precise editing approach that may enable a long-lasting treatment for patients with MSMDS.

bioengineering↗

bioRxiv2022

Precise DNA Cloning via PAMless CRISPR-SpRYgests

While restriction enzymes (REs) remain the gold-standard for manipulating DNA in vitro, they have notable drawbacks including a dependence on short binding motifs that constrain their ability to cleave DNA substrates. Here we overcome limitations of REs by developing an optimized molecular workflow that leverages the PAMless nature of a CRISPR-Cas enzyme named SpRY to cleave DNA at practically any sequence. Using SpRY for DNA digests (SpRYgests), we establish a method that permits the efficient cleavage of DNA substrates at any base pair. We demonstrate the effectiveness of SpRYgests using more than 130 gRNAs, illustrating the versatility of this approach to improve the precision of and simplify several cloning workflows, including those not possible with REs. We also optimize a rapid and simple one-pot gRNA synthesis protocol, which reduces cost and makes the overall SpRYgest workflow comparable to that of RE digests. Together, SpRYgests are straightforward to implement and can be utilized to improve a variety of DNA engineering applications.

biochemistry↗

bioRxiv2021

A systematic genotype-phenotype map for missense variants in the human intellectual disability-associated gene GDI1

Next generation sequencing has become a common tool in the diagnosis of genetic diseases. However, for the vast majority of genetic variants that are discovered, a clinical interpretation is not available. Variant effect mapping allows the functional effects of many single amino acid variants to be characterized in parallel. Here, we combine multiplexed functional assays with machine learning to assess the effects of amino acid substitutions in the human intellectual disability-associated gene, GDI1. We show that the resulting variant effect map can be used to discriminate pathogenic from benign variants. Our variant effect map recovers known biochemical and structural features of GDI1 and reveals additional aspects of GDI1 function. We explore how our functional assays can aid in the interpretation of novel GDI1 variants as they are discovered, and to re-classify previously observed variants of unknown significance.

genetics↗