Empirical optimization of dCas12a-based CRISPRi/a for pooled screening in Yarrowia lipolytica

Abstract

Yarrowia lipolytica is an important host for biotechnology and metabolic engineering, valued for its ability to accumulate lipids and produce a wide range of industrially relevant compounds. This creates a growing need for genome-scale tools that can accelerate the identification of genes underlying complex phenotypes. Although Cas9-based CRISPR interference (CRISPRi) and activation (CRISPRa) tools are well established in Y. lipolytica, Cas12a offers a complementary platform with a T-rich PAM that can expand access to promoter regions and facilitate multiplexed regulation. Despite these advantages, Cas12a-based CRISPRi/a remains comparatively underdeveloped in Y. lipolytica, particularly for large-scale functional screening. Here, we establish and systematically benchmark dCas12a-based CRISPRi/a platforms for pooled screening in Y. lipolytica. We first compared two widely used effectors, dLbCas12a and denAsCas12a, across single-gene assays and a pooled CRISPRi library targeting nine essential or conditionally essential genes with 752 crRNAs. Both orthologs supported robust transcriptional repression, but their relative performance was crRNA- and condition-dependent, with dLbCas12a showing more consistent and stronger late-stage depletion across targets. dLbCas12a also enabled enrichment-based CRISPRi and CRISPRa screens, recovering expected drug-resistance and lipid-associated phenotypes. The broad design of the pooled library further enabled identification of crRNA features associated with CRISPRi efficacy. Highly active crRNAs were enriched near transcription start sites, with an effective targeting interval extending from approximately -260 to +50 bp. Intermediate GC content was associated with higher efficacy, while PAM identity showed a weaker but consistent effect, with TTTA and TTTT PAMs linked to stronger depletion than TTTC and TTTG PAMs. Spacer length showed the strongest association with CRISPRi efficacy, with 18-nt crRNAs consistently outperforming matched longer 20-, 23-, and 25-nt variants. Finally, we tested dHyperLbCas12a, an engineered LbCas12a variant with enhanced activity, and found that it produced substantially stronger CRISPRi and CRISPRa phenotypes than dLbCas12a. Together, these results establish Cas12a as a versatile platform for pooled CRISPRi/a screens in Y. lipolytica, identify practical crRNA prioritization rules for improved library performance, and highlight dHyperLbCas12a as a promising effector for future large-scale gain- and loss-of-function screens.