Mastering targeted genome engineering of GC-rich oleaginous yeast for tailored plant oil alternatives for the food and chemical sector

Abstract

Background: Sustainable production of triglycerides for various applications is a major focus of microbial factories. Oleaginous yeast species have been targeted for commercial production of microbial oils. Among all the oleaginous yeasts examined in a previous comparative study, Cutaneotrichosporon oleaginosus showed the highest lipid productivity. Moreover, a new lipid production process for C. oleaginosus with minimal waste generation and energy consumption resulted in the highest lipid productivity in the history of oleaginous yeasts. However, productivity and product diversity are restricted because of the genetic intractability of this yeast. To date, successful targeted genetic engineering of C. oleaginosus has not yet been reported.

Results: The targeted gene editing was successfully carried out in C. oleaginosus using CRISPR/Cas system. A tailored enzyme system isolated to degrade the C. oleaginosus cell wall enabled the isolation of viable spheroplasts that are amenable to in-cell delivery of nucleic acids and proteins. The employment of both Cas9 protein and Cas mRNA was effective in obtaining strains with URA5 knockout that did not exhibit growth in the absence of uracil. Subsequently, we successfully created several strains with enhanced lipid yield (54% increase compared to that in wild type) or modified fatty acid profiles comparable with those of cocoa butter or sunflower oil compositions.

Conclusion: This study establishes the first targeted engineering technique for C. oleaginosus using the CRISPR/Cas system. The current study creates the foundation for flexible and targeted strain optimizations towards building a robust platform for sustainable microbial lipid production. Moreover, the genetic transformation of eukaryotic microbial cells using Cas9 mRNA was successfully achieved.

Keywords: CRISPR/Cas; Cocoa butter; Cutaneotrichosporon oleaginosus; Fatty acid biosynthesis; Genome engineering; High-oleic sunflower oil; Oleaginous yeast; Tailored plant oil alternatives; Yeast oil.

Fig. 1

Targeted genetic engineering of Cutaneotrichosporon oleaginosus using CRISPR/Cas technology. a Schematic illustration of the spheroplasting procedure of C. oleaginosus and custom enzyme isolation. b Schematic illustration of the genetic engineering of C. oleaginosus using CRISPR/CAS system

Fig. 2

Overview of the URA5 knockout strategies. As a proof of principle for CRISPR/Cas-mediated genetic modifications, we attempted to knockout the orotate phosphoribosyltransferase gene (URA5) to counter select 5-fluoroorotic acid (5FOA). We selected three sgRNAs from the library, which showed no off-target activities within the C. oleaginosus genome through in-silico prediction, and selectively targeted URA5. We then followed three parallel strategies to implement the Cas9 platform in C. oleaginosus. Both spheroplast batches prepared by Glucanex and HEST were used to test all strategies. a Strategy one and two: genome editing by Cas nuclease delivered into spheroplasts by electroporation in two forms separately: protein (Cas:sgRNA ribonucleoprotein [RNP]) and mRNA. In both strategies one single guide RNA (sgRNA) was used to target URA5. A single stranded DNA (ssDNA) was simultaneously transferred to introduce the repair sequences, including base deletions and base substitutions. b Strategy three: genome editing using Cas nickase as an RNP. Here, two sgRNAs targeting the leading and lagging strands were delivered to create the double-strand break (DSB). The repair ssDNA included base insertions and substitutions. A non-cutting restriction site (HindIII) was also introduced in the URA5 loci of mutants. The protospacer adjacent motifs (PAMs) were mutated in all strategies to prevent further DNA cleavage after the repair. c–e Colonies on selection agar plates with URA5 knockout using Cas mRNA, nuclease protein, and nickase protein, respectively. f The agarose gel electrophoresis of digested URA5 gene from WT and Δura5 strains. The URA5 locus was PCR amplified from the genomic DNA of mutants and WT and subjected to fast digestion by HindIII restriction enzyme. The digestion resulted in appearance of two smaller bands in the gene isolated from the Δura5 strain, indicating the integration of repair DNAs by Cas nickase. The WT URA5 gene was not digested

Fig. 3

Metabolic engineering. a The URA5 gene, including its promoter, was deleted to generate the Δura5 strain (using single guide RNAs (sgRNAs 1 and 2). The 3’ end of the coding sequence (90 bp) and terminator were not deleted, as they contained the terminator elements of the downstream gene. The complete URA5 coding sequence, with its native promoter and terminator, was used as a selection marker. The D9FAD and D12FAD overexpression was accomplished by inserting a second copy fused to AKRp and AKRt, and the selection marker into the upstream region of URA5 locus in Δura5 strain (sgRNAs 3 and 4), thus generating the strains D9OE and D12OE, respectively. The D12FAD knockout was carried out by inserting the URA5 in the Δ12 desaturase locus (sgRNA9). The D9FAD promoter exchange was performed by separate insertion of AKRp or TEFp, and simultaneous deletion of the native promoter to modify their transcriptional regulation (sgRNAs 5 and 6), generating the AKRp-D9 and TEFp-D9 strains, respectively. The same strategy was used for D12FAD (sgRNAs 7 and 8), resulting in AKRp-D12 and TEFp-D12, respectively. b Fatty acid profile, c lipid contents and titres, and d growth obtained with the WT and engineered C. oleaginous strains in MNM + Glu in shake flasks after 96 h cultivation. All data and error bars represent average and standard deviation of biological triplicates. The WT yielded 9.2 ± 0.2 g/L biomass and 50 ± 1.5% [w_lipid/dw_biomass] lipids. The biomass and lipid accumulated by D9OE, D12OE, and TEFp-D9 are comparable to the WT (p > 0.05). In Contrast, the AKRp-D9 exhibited lower growth rate (DCW at 5.6 ± 0.3 g/L) but maintained the cellular lipid accumulation levels after 96 h (47 ± 3% [w_lipid/dw_biomass] (p > 0.05)). The D12FAD knockout and promoter exchange did not affect the ability of the strains to grow and accumulate lipid

Fig. 4

Fed-batch fermentation of C. oleaginosus strains. Left column: Fermentation using MNM + Glu. Right column: Fermentation using RM + AA + Glu. a, b Time course of fatty acid composition (FAC) in WT in MNM + Glu and RM + AA + Glu, respectively. c, d FAC in D9OE strain. e, f FAC in TEFp-D9. g, h FAC in AKRp-D9. i, j FAC in Δd12. k, l Lipid contents (yellow bars) and lipid titres (purple line), and m, n growth (dry cell weights) obtained with the WT and engineered C. oleaginous strains in MNM + Glu after 96 h of cultivation and in RM + AA + Glu after 72 h of cultivation, respectively. All data and error bars represent average ± standard deviation of biological triplicates. Statistically significant differences between the WT and each engineered C. oleaginosus strain were defined using the two-tailed Student’s t-test. * P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001