HKUMed researchers discover more efficient CRISPR-Cas9 variants for gene therapy applications

A analysis workforce from the LKS Faculty of Medicine, The University of Hong Kong (HKUMed) found more efficient CRISPR-Cas9 variants that may very well be helpful for gene therapy applications. By establishing a brand new pipeline methodology that implements machine studying on high-throughput screening to precisely predict the exercise of protein variants, the workforce expands the capability to investigate as much as 20 instances more variants without delay with out the necessity for buying extra experimental information, which vastly accelerates the pace in protein engineering. The analysis workforce has efficiently utilized the pipeline in a number of Cas9 optimizations and engineered newStaphylococcus aureusCas9 (SaCas9) variants with enhanced gene modifying effectivity. The findings at the moment are printed in Nature Communications and a patent software has been filed based mostly on this work.


Staphylococcus aureusCas9 (SaCas9) is a good candidate for in vivo gene therapy resulting from its small dimension permitting packaging into adeno-associated viral vectors to be delivered into human cells for therapeutic applications. However, its gene-editing exercise may very well be inadequate for some particular illness loci. Further optimizations of SaCas9 are essential in precision medication earlier than it may be used as a dependable device to deal with human illnesses. Such optimizations include boosting its effectivity and precision by altering the Cas9 protein. Standard protocol for modifying the protein entails saturation mutagenesis, the place the variety of potential modifications that may very well be launched to the protein far exceeds the experimental screening capability of even the state-of-art high-throughput platforms by orders of magnitudes.

In this work, the analysis workforce explored if combining machine studying with structure-guided mutagenesis library screening might allow the digital screening of many more modifications to precisely determine the uncommon and higher performing variants for additional in-depth validations.

Research findings

The analysis workforce examined the machine studying framework on a number of beforehand printed mutagenesis screens on Cas9 variants and illustrated that machine studying might robustly determine one of the best performing variants by utilizing merely 5-20% of the experimentally decided information.

The Cas9 protein incorporates a number of components, together with protospacer adjoining motif (PAM)-interacting (PI) and Wedge (WED) domains to facilitate its interplay with the goal DNA duplex. The analysis workforce coupled the machine studying and high-throughput screening platforms to design activity-enhanced SaCas9 protein by combining mutations in its PI and WED domains surrounding the DNA duplex bearing a (PAM). PAM is important for Cas9 to edit the goal DNA and the concept was to scale back the PAM constraint for wider genome concentrating on while securing the protein construction by reinforcing the interplay with the PAM-containing DNA duplex by way of the WED area.

In the display screen and subsequent validations, the researchers recognized new variants, together with one named KKH-SaCas9-plus, with enhanced exercise by as much as 33% at particular genomic loci. The subsequent protein modeling evaluation revealed the brand new interactions created between the WED and PI domains at a number of places throughout the PAM-containing DNA duplex, attributing to KKH-SaCas9-plus’s enhanced effectivity.

Research significance

Structure-guided design has been dominating the sector of Cas9 engineering; nevertheless, it solely explores a small variety of websites, amino-acid residues, and mixtures. In this research, the analysis workforce confirmed that screening with bigger scale and fewer experimental efforts, time and price may be performed utilizing the machine learning-coupled multi-domain combinatorial mutagenesis screening method, which led them to determine a brand new high-efficiency variant KKH-SaCas9-plus.

‘This method will enormously speed up the optimization of Cas9 proteins, which might enable genome modifying to be utilized in treating genetic illnesses in a more efficient method,’ mentioned Dr Alan Wong Siu-lun, Assistant Professor of the School of Biomedical Sciences, HKUMed.
Source:The University of Hong KongJournal reference:Thean, D.G.L., et al. (2022) Machine learning-coupled combinatorial mutagenesis permits resource-efficient engineering of CRISPR-Cas9 genome editor actions. Nature Communications.

Recommended For You