Guide RNA Design for Genome-Wide CRISPR Screens in Yarrowia lipolytica

Advances in CRISPR-enabled genome-wide screens in yeast

Clustered regularly interspaced short palindromic repeats (CRISPR)-Cas genome-wide screens are powerful tools for unraveling genotype-phenotype relationships, enabling precise manipulation of genes to study and engineer industrially useful traits. Traditional genetic methods, such as random mutagenesis or RNA interference, often lack the specificity and scalability required for large-scale functional genomic screens. CRISPR systems overcome these limitations by offering precision gene targeting and manipulation, allowing for high-throughput investigations into gene function and interactions. Recent work has shown that CRISPR genome editing is widely adaptable to several yeast species, many of which have natural traits suited for industrial biotechnology. In this review, we discuss recent advances in yeast functional genomics, emphasizing advancements made with CRISPR tools. We discuss how the development and optimization of CRISPR genome-wide screens have enabled a host-first approach to metabolic engineering, which takes advantage of the natural traits of nonconventional yeast-fast growth rates, high stress tolerance, and novel metabolism-to create new production hosts. Lastly, we discuss future directions, including automation and biosensor-driven screens, to enhance high-throughput CRISPR-enabled yeast engineering.

Balanced Training Sets Improve Deep Learning-Based Prediction of CRISPR sgRNA Activity

CRISPR-Cas systems have transformed the field of synthetic biology by providing a versatile method for genome editing. The efficiency of CRISPR systems is largely dependent on the sequence of the constituent sgRNA, necessitating the development of computational methods for designing active sgRNAs. While deep learning-based models have shown promise in predicting sgRNA activity, the accuracy of prediction is primarily governed by the data set used in model training. Here, we trained a convolutional neural network (CNN) model and a large language model (LLM) on balanced and imbalanced data sets generated from CRISPR-Cas12a screening data for the yeast Yarrowia lipolytica and evaluated their ability to predict high- and low-activity sgRNAs. We further tested whether prediction performance can be improved by training on imbalanced data sets augmented with synthetic sgRNAs. Lastly, we demonstrated that adding synthetic sgRNAs to inherently imbalanced CRISPR-Cas9 data sets from Y. lipolytica and Komagataella phaffii leads to improved performance in predicting sgRNA activity, thus underscoring the importance of employing balanced training sets for accurate sgRNA activity prediction.

Optimized genome-wide CRISPR screening enables rapid engineering of growth-based phenotypes in Yarrowia lipolytica

CRISPR-Cas9 functional genomic screens uncover gene targets linked to various phenotypes for metabolic engineering with remarkable efficiency. However, these genome-wide screens face a number of design challenges, including variable guide RNA activity, ensuring sufficient genome coverage, and maintaining high transformation efficiencies to ensure full library representation. These challenges are prevalent in non-conventional yeast, many of which exhibit traits that are well suited to metabolic engineering and bioprocessing. To address these hurdles in the oleaginous yeast Yarrowia lipolytica, we designed a compact, high-activity genome-wide sgRNA library. The library was designed using DeepGuide, an sgRNA activity prediction algorithm and a large dataset of ∼50,000 sgRNAs with known activity. Three guides per gene enables redundant targeting of 98.8% of genes in the genome in a library of 23,900 sgRNAs. We deployed the optimized library to uncover genes essential to the tolerance of acetate, a promising alternative carbon source, and various hydrocarbons present in many waste streams. Our screens yielded several gene knockouts that improve acetate tolerance on their own and as double knockouts in media containing acetate as the sole carbon source. Analysis of the hydrocarbon screens revealed genes related to fatty acid and alkane metabolism in Y. lipolytica. The optimized CRISPR gRNA library and its successful use in Y. lipolytica led to the discovery of alternative carbon source-related genes and provides a workflow for creating high-activity, compact genome-wide libraries for strain engineering.

Genome Editing, Transcriptional Regulation, and Forward Genetic Screening Using CRISPR-Cas12a Systems in Yarrowia lipolytica

Class II Type V endonucleases have increasingly been adapted to develop sophisticated and easily accessible synthetic biology tools for genome editing, transcriptional regulation, and functional genomic screening in a wide range of organisms. One such endonuclease, Cas12a, presents itself as an attractive alternative to Cas9-based systems. The ability to mature its own guide RNAs (gRNAs) from a single transcript has been leveraged for easy multiplexing, and its lack of requirement of a tracrRNA element, also allows for short gRNA expression cassettes. To extend these functionalities into the industrially relevant oleaginous yeast Yarrowia lipolytica, we developed a set of CRISPR-Cas12a vectors for easy multiplexed gene knockout, repression, and activation. We further extended the utility of this CRISPR-Cas12a system to functional genomic screening by constructing a genome-wide guide library targeting every gene with an eightfold coverage. Pooled CRISPR screens conducted with this library were used to profile Cas12a guide activities and develop a machine learning algorithm that could accurately predict highly efficient Cas12a gRNA. In this protocols chapter, we first present a method by which protein coding genes may be functionally disrupted via indel formation with CRISPR-Cas12a systems. Further, we describe how Cas12a fused to a transcriptional regulator can be used in conjunction with shortened gRNA to achieve transcriptional repression or activation. Finally, we describe the design, cloning, and validation of a genome-wide library as well as a protocol for the execution of a pooled CRISPR screen, to determine guide activity profiles in a genome-wide context in Y. lipolytica. The tools and strategies discussed here expand the list of available synthetic biology tools for facile genome engineering in this industrially important host.

Genome-wide functional screens enable the prediction of high activity CRISPR-Cas9 and -Cas12a guides in Yarrowia lipolytica

Genome-wide functional genetic screens have been successful in discovering genotype-phenotype relationships and in engineering new phenotypes. While broadly applied in mammalian cell lines and in E. coli, use in non-conventional microorganisms has been limited, in part, due to the inability to accurately design high activity CRISPR guides in such species. Here, we develop an experimental-computational approach to sgRNA design that is specific to an organism of choice, in this case the oleaginous yeast Yarrowia lipolytica. A negative selection screen in the absence of non-homologous end-joining, the dominant DNA repair mechanism, was used to generate single guide RNA (sgRNA) activity profiles for both SpCas9 and LbCas12a. This genome-wide data served as input to a deep learning algorithm, DeepGuide, that is able to accurately predict guide activity. DeepGuide uses unsupervised learning to obtain a compressed representation of the genome, followed by supervised learning to map sgRNA sequence, genomic context, and epigenetic features with guide activity. Experimental validation, both genome-wide and with a subset of selected genes, confirms DeepGuide's ability to accurately predict high activity sgRNAs. DeepGuide provides an organism specific predictor of CRISPR guide activity that with retraining could be applied to other fungal species, prokaryotes, and other non-conventional organisms.