(563b) Application and Validation of a Genome-Wide CRISPR-Cas9 Library for the Oleaginous Yeast Yarrowia Lipolytica

Schwartz, C., UC Riverside
Wheeldon, I., UC Riverside
Engineering of the nonconventional oleaginous yeast Yarrowia lipolytica is limited by an incomplete set of synthetic biology tools for genome editing and strain construction. Our lab has made progress towards overcoming these limitations by developing a number of CRISPR-Cas9-based tools to enable (i) gene disruption via indel formation, (ii) integration of expression cassettes into predefined genomic loci without the need for integration of a selectable marker, (iii) repression of native gene expression with CRISPR interference (CRISPRi), (iv) activation of native gene expression with CRISPR activation, and (v) genome-wide knockout screening. Our lab found that straightforward adaptation of CRISPR-Cas9 technologies from other eukaryotes did not work well in Y. lipolytica, resulting in low gene disruption rates. To improve CRISPR-Cas9 function, we constructed novel hybrid RNA polymerase III promoters and applied them for sgRNA expression. This allowed higher than 90% gene disruption rates and gene integration rates exceeding 50%. The CRISPR-Cas9 system was also adapted for repressing gene expression by rendering Cas9 catalytically inactive and fusing it to the Mxi1 domain. Nonhomologous end-joining was repressed using this CRISPRi system, and a significant increase in homologous recombination was achieved. By fusing catalytically inactive Cas9 to an activation domain (VPR) instead of a repressor domain, we were able to activate expression of natively silent β-glucosidase genes and enable growth of Y. lipolytica on cellobiose.

To achieve genome-wide knockout screening, we designed approximately 48,000 sgRNAs that targeted each gene with 6-fold coverage. The sgRNAs were designed to (i) maximize on-target cutting efficiency, (ii) target in the first 20% of each gene’s coding sequence, (iii) avoid targeting intronic sequences, and (iv) avoid targeting more than once within the genome. With the help of DOE’s Joint Genome Institute, we compared four different cloning strategies and constructed a pooled library that contained over 99% of the designed sgRNAs. In order to determine the efficacy of our library design and quantify the effectiveness of the sgRNAs, we transformed the pooled library into a strain of Y. lipolytica that expressed Cas9 and was deficient in DNA repair. In this strain, a double-stranded break in genomic DNA leads to cell death, meaning efficient sgRNAs are depleted and poor cutting sgRNAs are enriched with extended culture time. Isolation and sequencing of the plasmid library after several days of outgrowth enabled us to identify and rank the cutting efficiency of each designed sgRNA. We found that approximately 63% of the designed sgRNAs were highly efficient cutters, and that over 97% of the genes in the Y. lipolytica genome were targeted by at least one highly efficient sgRNA. The library was then transformed into a strain of Y. lipolytica expressing Cas9 and competent for DNA repair, so that a double-strand break in genomic DNA lead to an indel mutation to inactivate the targeted gene. By quantifying only the most efficient sgRNA targeting each gene, we identified approximately 1,500 genes that had at least a 10-fold reduction in sgRNA abundance. This work is one of the first examples of genome-wide engineering of a nonconventional yeast and provides a blueprint for how genome-wide CRISPR screens can be effectively implemented in other organisms.