<scp>CRISPR</scp>/Cas9‐mediated generation of non‐motile mutants to improve the harvesting efficiency of mass‐cultivated <i>Euglena gracilis</i>

Abstract

Non-motile photosynthetic flagellates sediment well and thus can be harvested more easily under mass cultivation (Brennan and Owende, 2010). However, a precise genetic manipulation of microalgal motility remains challenging. The nutrient-rich microalga Euglena gracilis is widely used in the food, cosmetics and feed industries (Suzuki, 2017). This alga accumulates paramylon, a crystalline β-1,3-glucan with multiple industrial applications (Harada et al., 2020). Under hypoxia, E. gracilis degrades paramylon to generate energy and converts it into wax esters, primarily myristic acid (C14:0) and myristyl alcohol (C14:0) (Inui et al., 2017), which are a potential source of jet biofuel. To improve the harvesting efficiency of E. gracilis, a non-motile mutant ( M 3 − ZFeL ) was generated using Fe-ion beam irradiation. However, M 3 − ZFeL grew slower and produced less lipids than the wild type (WT), suggesting that the mutagenesis caused undesirable side mutations (Muramatsu et al., 2020). Although flagellar mutants with motility defects have been identified in model organisms such as Chlamydomonas reinhardtii, Trypanosoma brucei and Caenorhabditis elegans (Kamiya et al., 1991; Wingfield et al., 2018), a precise manipulation of motility in E. gracilis remains to be explored, despite its industrial usefulness. Here, using our Cas9 ribonucleoprotein (RNP)-based genome-editing technique (Nomura et al., 2019), we targeted the Bardet–Biedl syndrome (BBS) genes in E. gracilis to generate non-motile strains. We identified the genes encoding proteins homologous to Caenorhabditis elegans BBS-7 (NP_499585.1; tblastn E value 7e–72) or BBS-8 (NP_504711.2; tblastn E value 7e–123). These BBSome components are associated with the intraflagellar transport of particles and mediate the trafficking of cilium/flagellum membrane proteins in eukaryotic cells (Hammond et al., 2021; Nakayama and Katoh, 2018; Wingfield et al., 2018). We designed two pairs of single guide RNAs (sgRNAs) that target different regions of EgBBS7 and EgBBS8 and introduced Cas9–sgRNA RNPs into E. gracilis cells by electroporation (Method S1, Table S1). Assessing the motility of the electroporated E. gracilis cells, we identified four sets of target sequences and protospacer adjacent motifs (PAMs), which are useful for inducing the stable non-motility phenotype in E. gracilis (EgBBS7-A and EgBBS7-B; and EgBBS8-B and EgBBS8-D). To establish clonal bbs strains, we isolated two cell lines each for bbs7 and bbs8 (Method S1). PCR-based genotyping detected 500–1500 bp deletions within the EgBBS7 and EgBBS8 target regions in these strains (Method S2, Figure 1a,b), and Sanger sequencing verified that large genomic regions were deleted, including intronic regions (Figure 1c,d). When isolated single cells were cultured, all eight bbs mutant strains flocculated, whereas the WT spread (Figure 1e). We confirmed the non-flagellar phenotype of the mutant strains by scanning electron microscopy (Method S3, Figure 1f,g) and their non-motility by a trace momentum assay that quantifies swimming motion (Method S4, Figure 1h,i). The results indicate that EgBBS7 and EgBBS8 contribute to forming a full-length flagellum in E. gracilis and are thus required for motility. Next, we assessed gravitational sedimentation of the non-motile bbs mutants by time-lapse imaging in 1-min intervals (Figure 1j). We created 2D kymographs representing time-dependent changes in the transparent–sediment interface level in bbs and WT cells, revealing the rapid sedimentation of the bbs mutants (Method S5, Figure 1k). We examined time-series changes in the level of the transparent–sediment area in flasks segmented from binarized images. The WT cultures remained cloudy at 120 min after the flasks settled due to the presence of swimming cells, whereas bbs cells almost fully sedimented (Method S5, Figure 1l). Moreover, comparing the dry weights of bbs and WT sediments at 100 min after the flasks settled showed that the bbs mutants had 32–38% higher sedimentation rates than WT (Method S5, Figure 1m). We also examined the growth, biomass, paramylon content and lipid content of WT and the bbs mutants (Method S6). Using the cell density of cells cultured in KH medium as estimate for growth, the growth rates of all four bbs strains were not significantly different from WT until 4 days after culture initiation, but the cell concentration of the bbs8 mutants was slightly higher than that of WT (pbbs8-B #4 = 0.042, pbbs8-D #13 = 0.011, Dunnett's test) at 7 days after culture initiation (Figure 1n). We did not detect marked differences between the bbs mutants and WT in terms of biomass harvested at 7 days after culture initiation or for the lipid content of cells under aerobic or hypoxic conditions (Figure 1o,p). Interestingly, the paramylon content of cells under aerobic conditions was significantly higher in the bbs mutants than in the WT (Figure 1q), which might be related to cellular motility and paramylon biosynthesis and/or accumulation. These results suggest that mutations in the EgBBS genes that facilitate sedimentation do not negatively affect the production of biomass or high-value products. During the mass cultivation of microalgae, harvesting accounts for 20–30% of total production costs; therefore, improving harvesting procedures will enhance the economic viability of microalgal biomass production (Brennan and Owende, 2010). We estimated the harvesting efficiency of the bbs mutants to be 32–38% higher than that of the WT (Figure 1m), while all strains produced similar amounts of biomass (Figure 1o, aerobic). Therefore, our results demonstrate that knocking out EgBBS genes in E. gracilis could improve harvesting efficiency without negatively affecting productivity. This work was financially supported by the Japan Science and Technology Agency (JST)-OPERA Program to KM. The microchamber used for motion analysis was provided by Professor Simon Song (Hanyang University). We thank Aya Ide and Hiromi Ojima for their experimental support. This study was partially supported by a matching fund-based research programme between RIKEN and euglena Co., Ltd. MI, TN and ST contributed equally to this work. MI, TN, KY and ST conceived the study and designed the experiments. MI, TN, ST, KO, TS, TK and KH performed experiments and analysed the data. KM and KS supervised the project. KM and MI wrote the manuscript. All authors read and approved the final manuscript. LC644194, LC644196–LC644198, LC644202 (Method S1). Method S1 Euglena gracilis genome editing. Method S2 Genotyping of bbs mutants. Method S3 Microscopy observations. Method S4 Quantitative motion analysis. Method S5 Sedimentation analysis. Method S6 Growth, biomass, paramylon content, and lipid contents. Table S1 Oligos used in this study. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.