High‐Efficiency Genome Editing of <i>Chlorella</i> sp. by <scp>CRISPR</scp> /Cas9

Chlorella spp. are among the most prominent microalgal species utilised commercially for human consumption. Since the 1950s, large-scale cultivation of Chlorella spp. has been implemented to address the growing demand for alternative protein sources, due to their rapid growth cycle, robust environmental tolerance and high protein content (Lu et al. 2021). These microalgae are used in various applications, including aquaculture, food substitutes and health supplements. Furthermore, their plant-like nature has made them an ideal model organism for investigating photosynthetic mechanisms and CO2 assimilation (Calvin-Benson cycle) since the 1910s (Nickelsen 2017). We previously isolated a saltwater Chlorella strain MEM25 (hereafter MEM25) with exceptional environmental adaption and high yields of proteins and polyunsaturated fatty acids (Lu et al. 2021). Additionally, a high-quality genome sequence of MEM25 reveals a compact genome (~50 Mb) and versatile metabolic pathways (Wang et al. 2025), reinforcing its promise as a model organism for both commercial applications and basic research, particularly in synthetic biology. Despite the significance, the development of genetic manipulation tools for Chlorella spp. remains limited, especially in comparison to other emerging model species (Ajjawi et al. 2017; Wang et al. 2016). Efforts have been made to establish gene overexpression (Run et al. 2016; Sharma et al. 2021) and CRISPR-mediated gene editing (Kim et al. 2021; Lin and Ng 2020) in Chlorella species; however, these studies have been exclusively conducted in freshwater strains. CRISPR/Cas9-mediated genome editing has been attempted using plant-derived vectors in two freshwater Chlorella strains, these approaches have been unstable, with gene knockout efficiencies too low to be used in practice (only two mutant colonies identified across three independent transformation attempts (Kim et al. 2021)). Moreover, the genetic modification protocols reported thus far have often been unstable and non-reproducible across different Chlorella strains, and transgenes are typically lost after a few generations. Therefore, it is essential to develop stable and high-efficiency CRISPR/Cas9-based genome-editing tools for Chlorella. We have established a robust transformation system for MEM25, characterised by shorter screening periods and greater genetic stability than those reported in other Chlorella spp. To enable efficient gene editing, we constructed a CRISPR/Cas9 gene knockout system using a custom-designed vector, pMEM-CP3 (Figure 1a). The Cas9/gRNA vector pMEM-CP3 harbours a full-length Cas9 gene originally from Streptococcus pyogenes, a sgRNA gene and a basta (i.e., glufosinate ammonium) resistance gene BarR, all driven by endogenous promoters and terminated by endogenous terminators (Figure 1a). Basta is a world's most commonly used knock-down herbicide with minimal environmental impact, and it is cost-effective and rapidly degraded. At concentrations as low as 800 μg mL−1 in solid F2 and 30 μg mL−1 in liquid medium significantly compromise the algal growth (Figure 1b). Therefore, BarR was used as selective marker which was driven by tRNA (guanine-N7-)-methyltransferase promoter (Pgmt) and terminated by the light-harvesting complex I chlorophyll a/b binding protein 4 termination (Tlhc4) region. To ensure optimal transcriptional regulation of both Cas9 and sgRNA, we characterised an endogenous bidirectional promoter Psece with a length of 477 bp (Supplementary methods). According to our transcriptome analysis, it drives the transcription of both upstream (pre translocase subunit SECE1) and the downstream (ATP-dependent RNA helicase DDX5) genes. In vivo experiments show that Psece drives robust expression of GFP upstream and mCherry downstream, with strong fluorescence signals observed in MEM25 (Figure 1c; Figure S1a). Moreover, to facilitate the nuclear localisation of Cas9, we added the SV40 nuclear localization signal to both the N- and C-termini of the Cas9 protein (Figure 1a). For efficient processing of the sgRNA, we incorporated self-cleaving ribozymes, fusing the gRNA scaffold with a hammerhead ribozyme (HH) at the 5′ end and the HDV ribozyme at the 3′ end (Figure 1a). Both Cas9 and sgRNA were driven by Psece and terminated by either the NF-X1-type zinc finger protein termination region (Tnfxl1) or cellulose synthase terminator (Tcs), respectively (Figure 1a). These constructs lack a sgRNA sequence for a target genomic locus. SapI was selected as restriction enzyme recognition sites for cloning sgRNAs. BarR and Cas9 genes were codon-optimised according to the codon frequency in MEM25. To evaluate the efficacy of this genome-editing system, we utilised the counter-selective marker adenine phosphoribosyltransferase gene (APT), which encodes a protein responsible for catalysing the conversion of adenine to adenylate. Disruption of the APT gene (APT-KO lines) confers resistance to growth inhibition in the presence of 2-fluoroadenine (2-FA). Two sgRNAs flanked by self-cleaving ribozymes (sgNR1 and sgNR2) were designed to independently target two distinct regions near the 5′ end of the APT gene (Figure 1d). We introduced the pMEM-CP3-sgNR1 (targeted region: TGGGCACTGCATCGATATTC) and pMEM-CP3-sgNR2 (targeted region: TTTGTACCTCTGAGAAAGCC) into MEM25 by electrotransformation (Figure 1d). Transformation with pMEM-CP3-sgNRs led to the generation of Basta-resistant lines, which were subsequently screened for APT knockout on F2 plates supplemented with both Basta and 2-FA (Figure S1b). Lines resistant to both Basta and 2-FA were confirmed for the presence of the Bar gene by PCR amplification and sequencing (Figure S1c). Mutations in the APT gene were then analysed by PCR amplification of the genomic locus, followed by sequencing (Figure 1e). Among the screened lines, 24% exhibited frame-shift mutations, 20% contained in-frame mutations, 16% showed no mutations and 40% produced low-quality sequencing results. The two sgRNAs demonstrated distinct efficiencies: sgNR1 (in APT-KO1 lines) exhibited high mutagenic efficiency, with 80% of colonies containing mutations and no wild-type sequences recovered. In contrast, sgNR2 (in APT-KO2 lines) had lower efficiency, with only 45% of colonies harbouring mutations and 20% showing no mutation (Table S1). Notably, low-quality sequencing reactions often produced readable chromatographs up to the target site, after which the sequence became unreadable. This suggests the presence of a heterogeneous mutant population, likely due to mutagenesis occurring after plating. Further purification of these populations by streaking to single colonies could yield pure genotypic lines. To assess the potential for off-target mutations, 100 randomly picked knockout colonies from APT-KO1 lines were selected for whole-genome sequencing (WGS). Cas-OFFinder software16 was used to identify potential off-target sites. With a threshold of ≤ 5 mismatches between the sgRNA and PAM sequences, 30 potential off-target sites (PAM: NGG) were identified (Figure 1f). Comprehensive sequence analysis, however, revealed no detectable off-target mutations in vivo, suggesting high specificity of the CRISPR/Cas9 system. In the knockout strains, 69.8% of single nucleotide polymorphisms (SNPs) and 84.2% of insertion/deletions (INDELs) are located within intronic regions (Figure S1d). This is widely believed to be a spontaneous genetic mutation that occurs during the cultivation process17. For SNP mutations, a slight preponderance of transversions over transitions was observed (p < 0.001), with the most common mutations being G to A (15.36%) and A to T (13.54%) (Figure S1e). INDELs were predominantly 3–5 base pairs in length (Figure S1f). In summary, a stable and high-efficiency CRISPR/Cas9-mediated genome-editing system has been successfully established for Chlorella sp., which could be applied to precisely edit virtually any gene or genomic region. Although this system has been applied to a single gene so far, it holds promise for advancing the use of Chlorella as a versatile chassis for synthetic biology and industrial applications. In contrast to previous studies (targeting the same gene), yielding only two distinct mutant colonies from three independent experiments (Kim et al. 2021)—our CRISPR/Cas9-based genome-editing system demonstrated significantly higher efficiency. Given the increasing demand for freshwater resources, the utilisation of marine Chlorella strains in industrial systems presents a promising alternative and strategically significant opportunity. Y.L. and Q.G. contributed to designing the experiments. X.C. and Y.X. contributed to performing the experiments related to transformation and off-target analysis. Y.L. wrote the manuscript. This work was supported in part by grants from the National Key R&D Program of China (2021YFA0909600 and 2021YFE0110100) and the National Natural Science Foundation of China (32560020, 32060061 and 32370380), the Key R&D Program of Hainan Province (ZDYF2024XDNY244 and ZDYF2022XDNY140), the Natural Science Foundation of Hainan Province (322QN250), and the Foreign Expert Foundation of Hainan Province (G20230607016E), and the Hainan Tropical Ocean University Joint Open Project for Aquatic South Breeding (2023SCNFKF04). The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions. Figure S1. Tables S1. Method S1. 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.