A research team from the School of Biomedical Sciences, LKS Faculty of Medicine, The University of Hong Kong (HKUMed), has developed the first platform for assembling and barcoding protein-encoding sequences carrying multiple mutations en masse and has coupled it with next-generation sequencing to track all variants in an unprecedented throughput.
Background
The current challenge for engineering CRISPR-Cas enzymes and other proteins lies in identifying the optimal sets of amino acid changes without under- or over-engineering the protein.
Finding new ways to enhance the throughput in assembling and accessing the function of protein variants is important for understanding epistatic mutations and accelerating next-generation engineering of genome-editing enzymes with new or improved properties.
The HKUMed research team has developed the first platform for assembling and barcoding protein-encoding sequences carrying multiple mutations en masse and has coupled it with next-generation sequencing to track all variants in an unprecedented throughput.
Using its newly developed platform, named CombiSEAL, the HKUMed team successfully identified new high-fidelity variants of a genome-editing Cas9 protein that have enhanced gene-editing specificity without sacrificing potency and broad targeting range.
This high throughput-screening platform aims to lead discovery and optimisation of more clinically useful proteins. The low-cost, easy-to-implement and scalable strategy help researchers excel in the increasingly competitive environment of innovative bioproducts.
Study results
Streptococcus pyogenes Cas9 (SpCas9) nuclease is the most widely used CRISPR-based genome editing enzyme. Using CombiSEAL, a library of 948 SpCas9 variants, each of which contains one-to-eight mutations, was assembled and delivered into human cells, and their genome-editing efficiency and specificity were systematically quantified.
The total number of SpCas9 protein variants characterised in this study greatly exceeds the sum of all other major studies reported thus far.
The new HKUMed strategy will allow the team to further easily scale up the engineering process in future. The team also identified new high-fidelity SpCas9 variants, Opti-SpCas9 and OptiHF-SpCas9, which generate much fewer undesired off-target genome edits without sacrificing on-target editing efficiency and broad targeting range.
CombiSEAL greatly enhances the throughput for functional impact analysis of a massive number of mutation combinations by a simple one-pot reaction library assembly and a round of short-read sequencing of the variant-specific barcode combination.
Significance of the study
CombiSEAL, as developed by the HKUMed team, is the first-ever platform that breaks through limits to rapidly and simultaneously profile multiple mutations for protein engineering and map relationships between mutations. This platform can be easily implemented in many laboratories for the massively parallel engineering of proteins relevant to a multitude of biomedical, therapeutic and biotechnology applications.
The Lead researcher of the study stated that is a large-scale build-and-test platform for combinatorial optimisation of proteins. The one-pot assembly and barcoding strategy substantially reduce time and cost for the whole process starting from building to functional characterisation of the entire variant library.
The team identified new variants of SpCas9 with properties that are important for many of the applications enabled by genome editing.
It was noted that genome-editing tools need to be accurate to maximise safety for clinical use, while high editing efficiency and a broad target range are also important for applications like CRISPR screens.
Unlike other reported high-fidelity SpCas9 enzymes, Opti-SpCas9 is the only variant compatible with gRNAs containing an additional 5’ guanine for the transcription under the U6 promoter without comprising its on-target activity. This feature is important because U6 is the most widely used promoter in the many CRISPR-based genetic screens performed worldwide.
