Gene activation in Caenorhabditis elegans using the Campylobacter jejuni CRISPR-Cas9 feeding system

Abstract Clustered regularly interspaced palindromic repeats-based activation system, a powerful genetic manipulation technology, can modulate endogenous gene transcription in various organisms through fusing nuclease-deficient Cas9 to transcriptional regulatory domains. At present, this clustered regularly interspaced palindromic repeats-based activation system has been applied to activate gene expression by microinjection manner in Caenorhabditis elegans. However, this complicated and time-consuming injection manner is not suitable for efficient and high-throughput gene regulation with clustered regularly interspaced palindromic repeats-Cas9 system. Here, we engineered a Campylobacter jejun clustered regularly interspaced palindromic repeats-Cas9-based gene activation system through bacteria feeding technique to delivering gene-specific sgRNA in C. elegans. It enables to activate various endogenous genes efficiently, as well as induce the corresponding phenotypes with a more efficient and labor-saving manner. Collectively, our results demonstrated that our novel dCjCas9-based activation feeding system holds great promise and potential in C. elegans.


Introduction
The clustered regularly interspaced palindromic repeats (CRISPR)-CRISPR-associated protein (CRISPR-Cas) system is most often based on the Streptococcus pyogenes Cas9 (SpCas9) nuclease, which is an RNA-guided genome editing tool for genetic manipulation in various organisms (Bevacqua et al. 2021;Farbiak et al. 2021;Yin et al. 2021). In addition to SpCas9, various Cas9 orthologs from different bacteria have been characterized and developed, such as Neisseria meningitidis Cas9 (NmCas9), Staphylococcus aureus Cas9 (SaCas9), Francisella novicida Cas9 (FnCas9), and Campylobacter jejuni Cas9 (CjCas9). To date, the CRISPR-Cas9 system has been widely used for introducing indels and generating knockouts in vitro and in vivo (Bevacqua et al. 2021;Farbiak et al. 2021;Yin et al. 2021). Moreover, when a missense substitution of the catalytic site was introduced to eliminate Cas9 nuclease activity that did not compromise its DNA binding activity, nuclease-deficient Cas9 (dCas9) was also able to upregulate RNAguided transcription by fusing with trans-activators (Mali et al. 2013). Among these various Cas9 orthologs, CjCas9, a newer CRISPR endonuclease, exhibits several unique features, including its smallest size (984 amino acid residues) for easier delivery, distinct target recognition of the 5 0 -NNNNACA-3 0 or 5 0 -NNNNRYAC-3 0 sequence, unique triple-helix tracrRNA structure, and potential for contact with the nucleotide sequences in both DNA strands of the target (Zhang et al. 2021).
In Caenorhabditis elegans, the CRISPR-Cas9 system, especially SpCas9, has been widely used for gene editing and transcriptional regulation, either alone or in combination with transcriptional regulatory domains (Long et al. 2015;Wei et al. 2019). Very recently, we also developed a robust CjCas9-based transcription activation system, miniCAFE, which combines a nuclease-deficient CjCas9 (dCjCas9) and the tripartite transcriptional activator VP64-p65-Rta (herpes simplex virus-derived VP64 activator, the human NF-KB p65 activator domain, and the Epstein-Barr-virus-derived R transactivator, VPR), and then microinjected it into C. elegans to activate the expression of various genes (Zhang et al. 2021). Microinjection in C. elegans is a reliable, versatile, and frequently used method for delivering genetic constructs. However, microinjection requires a specialized micromanipulator and a skilled microinjection operator (Berkowitz et al. 2008), so it is not suitable for efficient and highthroughput gene disruption or regulation in the CRISPR-Cas9 system. In C. elegans, feeding-based RNAi is the most convenient and powerful method for silencing gene function by feeding animals with bacteria expressing dsRNA (Timmons et al. 2001). This "feeding" method has also been used for delivering guide-RNA to achieve CRISPR-Cas9-based gene disruptions (Liu et al. 2014). However, the CRISPR-based feeding system for gene activation has not been reported. Therefore, we engineered a CRISPR-Cas9-based gene activation system using a bacteria feeding technique to deliver gene-specific sgRNA to determine whether genes could be activated using a bacteria feeding-dependent CRISPR-Cas9 system.
To generate a plasmid expressing gene-specific sgRNA driven by the T7 promoter, a T7 promoter-lac operator-target sequence-sgRNA (FþE) scaffold-T7 terminator cassette was constructed and cloned into the L4440 vector to replace the convergently oriented T7 promoter fragment. The modified sgRNA scaffold [sgRNA (FþE) ] was amplified from the PU6::unc-119 sgRNA (FþE) plasmid (Zhao et al. 2016), and the T7 promoter-lac operator and the T7 terminator sequences were directly added to the forward and reverse primers, respectively. The sequences of gene-specific sgRNAs of interest were inserted into the sgRNA plasmid by nested overlap PCR, and all sgRNAs were designed using http://crispor.tefor.net/. The guide RNA sequences and their primers are listed in Table 1. The full sequence of the T7 promoter-lac operator-target sequence-sgRNA (FþE) scaffold-T7 terminator cassette is shown in Table 2.

The sgRNA bacteria feeding assay
The empty vector or gene-specific sgRNA plasmids were transformed into HT115(DE3) chemically competent cells using CaCl 2 transformation protocols. The gene-specific sgRNA sequences were verified by Sanger sequencing. All sgRNA bacteria feeding experiments were performed at 20 C using an RNAi-like feeding protocol . Briefly, in IPTG-supplemented plates, the VPR-dCjCas9 or VP64-dCjCas9 transgenic worms were fed HT115 bacteria carrying empty vector or gene-specific sgRNA plasmids starting at the L1 larval stage until the young adult stage. The sgRNA-fed worms were used for subsequent experiments. Table 1. The sgRNAs and primers in this study.

RNA extraction and quantitative RT-PCR
Worms were fed HT115 bacteria carrying empty vector or genespecific sgRNA plasmids until age-synchronized young adults and were then collected with M9 buffer and washed several times. Worm pellets were resuspended using AG RNA ex PRO reagent (Accurate Biology, Changsha, China). Total RNA was isolated by chloroform extraction and isopropanol precipitation. Afterward, 500 ng of total RNA was used for reverse transcription with a high-capacity cDNA transcription kit (RK20400, ABclonal, Wuhan, China). Quantitative real-time PCR was performed using SYBR Green Select Master Mix (RK21203, ABclonal, Wuhan, China) on a CFX96 real-time system (Bio Rad, CA, USA), and each experiment was repeated at least 3 times. Quantification of transcripts was normalized to the cdc-42 gene, and results were computed using the 2 -DDCt method. P-values were calculated using the 2-tailed Student's t-test. The primers used in this study are shown in Table 3.

Fluorescence microscopy and image analyses
To assess the fluorescence intensity of myo-2::GFP, we fed VPR-dCjCas9 and VP64-dCjCas9 transgenic worms HT115 bacteria carrying an empty vector or sgRNA targeting the myo-2 promoter. We mounted the worms on 2% agar pads after they were anesthetized using M9 buffer containing NaN 3 (50 mM) and then observed fluorescence using a Nikon Ti2-U (Zhang et al. 2021), The GFP fluorescence intensity of each worm was analyzed using ImageJ as previously described. Each experiment used at least 30 animals. P values were calculated using the 2-tailed Student's t-test.

Lifespan assay
All lifespan experiments were performed using standard protocols, as previously described (Zhang et al. 2021). Briefly, synchronized L1 animals were fed HT115 bacteria carrying empty vector or gene-specific sgRNA plasmids until the L4 larval stage. Then, approximately 100-120 young adult worms were transferred to IPTG-supplemented plates to conduct survival analyses. The corresponding HT115 bacteria were seeded before use. To prevent progeny production, 10 lM 5-fluoro-2 0 -deoxyuridine (FUDR, Sigma) was also added. Death events were scored daily, and experiments were repeated at least twice. The SPSS package was used for statistical analysis, and the log-rank (Mantel-Cox) method was used to determine the significant difference. A P-value < 0.05 was considered statistically significant.

Oil Red O staining and quantification
Synchronized worms were collected in M9 buffer. Then, Oil Red O (ORO) staining was conducted according to standard protocols (Zhang et al. 2021). ORO-stained animals were mounted onto 2% agar pads and imaged using a Nikon Ti2-U fluorescence microscope at 20Â magnification. The mean intensity values in arbitrary units (a.u.) of per worm were graphed by using Image-Pro-Plus processing software and the lipid levels were determined using GraphPad Prism using a 2-tailed Student's t-test. At least 30 animals were used in each experiment. The experiment was repeated at least twice.
To ensure efficient transcription of sgRNA in HT115 bacteria, a T7 promoter was used to drive an sgRNA cassette, including a target-specific sequence and a structurally modified sgRNA [sgRNA (FþE) ; Zhao et al. 2016] scaffold sequence. A lac operator was added between the T7 promoter and the target sequence to ensure sgRNA expression induced by isopropyl-beta-D-galactoside (IPTG). Furthermore, a T7 RNA polymerase terminator was added to the 3 0 end of the scaffold to ensure the correct termination of sgRNA transcription (Fig. 1a). Afterward, the sgRNA cassette was cloned into the L4440 vector to replace the convergently oriented T7 promoter fragment and then transformed into HT115 bacteria (Fig. 1a).
We fed VPR-dCjCas9 transgenic worms grown on plates containing IPTG using HT115 bacteria carrying gene-specific sgRNA for the promoter or empty vector L4440 (without the sgRNA cassette as a control) from L1 larvae to the young adult stage. The worms fed sgRNA were then collected to determine transcriptional levels and corresponding phenotype analyses to assess the efficacy of our VPR-dCjCas9 activation feeding system. Quantification of transcripts was normalized using the housekeeping gene cdc-42 (Hoogewijs et al. 2008), whose mRNA levels The T7 promoter-lac operator-target sequence-sgRNA (F1E) scaffold-T7 terminator cassette sequence (5 0 -3 0 ) taatacgactcactataggggaattgtgagcggataacaattccGAACCCGTTGCCGA-ATACACGTTTAAGAGCTATGCTGGAAACAGCATAGCAAGTTTAA-ATAAGGCTAGTCCGTTATCAACTTGAAAAAGTGGCACCGAGTCG-GTGCTTTTTTTctagcataaccccttggggcctctaaacgggtcttgaggggttttttg  Fig. 1. Transcriptional activation of multiple targeted genes and corresponding phenotype induction in C. elegans using the C. jejuni CRISPR-Cas9 activation feeding system. a) Schematic of the VRP-dCjCas9 activation feeding system. VPR-dCjCas9 transgenic worms were generated by injection of the Pdpy-30::VPR-dCjCas9 plasmid. By delivering the engineered sgRNA plasmid containing the gene-specific gRNA sequence through bacterial feeding, the VRP-dCjCas9 activator can activate gene expression in C. elegans. b) Pmyo-2::GFP fluorescence in the pharyngeal region of the VPR-dCjCas9 and 4 | G3, 2022, Vol. 12, No. 6 (continued) were not affected by sgRNA-bacteria, which was verified using other internal reference genes (acl-1 and csq-1; Hoogewijs et al. 2008; data not shown). In addition, to quantify transcripts, we used both empty vector L4440 and an sgRNA targeting an intergenic sequence as the negative control (Fig. 1f). During injection of plasmid Pdpy-30::VPR-dCjCas9, Pmyo-2::GFP::H2B was used as a coinjection maker to ensure successful microinjection. To test the capacity of the VPR-dCjCas9 feeding system, we first examined whether our VPR-dCjCas9 activator could up-regulate the pharyngeal GFP fluorescence expression by sgRNA targeting the myo-2 promoter. We fed the VPR-dCjCas9 transgenic worms HT115 bacteria carried with myo-2-specific sgRNA for promoter. We observed a significant increase in the fluorescence intensity of pharyngeal-GFP (Fig. 1b) and mRNA expression level when compared with the VPR-dCjCas9-only control worms (Fig. 1f), suggesting that the VPR-dCjCas9 feeding system could be used to activate the target gene in C. elegans.
Given the ability of the VPR-dCjCas9 system to upregulate reporter gene expression, we speculated that it might also induce endogenous gene transcription in C. elegans. Therefore, we used the VPR-dCjCas9 feeding system to activate a series of endogenous genes and conducted phenotype analyses.
First, we chose aak-2 and pha-4 to examine the ability of the VPR-dCjCas9 system to induce endogenous gene transcription. Both the AMPK a-catalytic subunit AAK-2 and the human FoxA transcription factor ortholog PHA-4 play important roles in lifespan regulation (Greer et al. 2007;Panowski et al. 2007). By delivering specific sgRNA targeting the aak-2 or pha-4 promoter using bacterial feeding, we observed a significant extension in mean lifespan (Fig. 1c) and greatly increased mRNA levels of the target (Fig. 1f), compared with control. Furthermore, we used lipid metabolism-regulating genes (lipl-4 and lipl-5) to assess our VPR-dCjCas9 system. The lysosomal acid lipases LIPL-4 and LIPL-5 have been reported to regulate lipid storage and longevity in C. elegans (Folick et al. 2015;Buis et al. 2019). Consistent with previous findings, using ORO staining, we found that the fat storage levels of C. elegans were significantly decreased when the VPR-dCjCas9 feeding system was used to target the lipl-4 or lipl-5 promoter, respectively (Fig. 1d). Moreover, up-regulation of the mRNA expression levels of lipl-4 or lipl-5 was also detected (Fig. 1f). Similar to aak-2 and pha-4, we also observed a significant lifespan extension when targeting lipl-4 promoter (Fig. 1c). Recently, researchers have used engineered dCas9::VP46 by microinjection manner, to successfully induce the mammalian REST orthologue spr-4 transcriptional activation and lifespan extension in worms (Zullo et al. 2019). Consistently, using our VPR-dCjCas9 feeding system with spr-4-specific-sgRNA, we also observed upregulation of the mRNA level of spr-4 with corresponding lifespan extension (Fig. 1, e and f). Altogether, these results suggest that the feeding-based VPR-dCjCas9 activation system enables efficient activation of endogenous genes in C. elegans.
Compared with the traditional microinjection activation system, our VPR-dCjCas9 feeding system, although not as efficient as microinjection systems, is more efficient and less laborintensive than injection systems (Zhang et al. 2021). Another gene activation system using herpes simplex virus-derived VP64 as an activator has been reported in C. elegans (Long et al. 2015;Zullo et al. 2019). To compare the activation efficiency of VP64 and the VPR activator, we generated VP64-dCjCas9 transgenic worms and then assessed the ability of VP64 and the VPR activator to induce gene expression using the same sgRNA targeting the myo-2 promoter by feeding manner. Compared with the control worms, we observed a significant increase in the fluorescence intensity of pharyngeal-GFP both in VPR-dCjCas9 and VP64-dCjCas9 worms, but those of VP64-dCjCas9 worms were weaker, suggesting that the activation efficiency of the VPR activator was obviously stronger than that of VP64 (Fig. 1b).
Furthermore, to determine the activation efficiency of our VPR-dCjCas9 feeding system in different tissues, we attempted to activate tissue-specific genes including pie-1 (germline), ges-1 (intestine), and rab-3 (neuron), by delivering specific-sgRNAs of targets through bacteria feeding. We found significant up-regulation of the mRNA levels of ges-1 and rab-3, but not pie-1, by feeding the gene-specific sgRNA bacteria (Fig. 1h). These results suggest that our feeding-based VPR-dCjCas9 system exhibits different efficiencies in different tissues, with a high activation efficiency in the intestine, pharynx (confirmed by myo-2) and neuron, but poor efficiency (even nonfunctional) in the germline.
As so far, a modular and flexible platform for gene activation in vivo has been built using the dCas9 protein combined with trans-activator domains (Bö hm et al. 2020;Chiarella et al. 2020). In this study, we developed a CRISPR-CjCas9-based system to efficiently activate transcription through bacteria feeding to deliver gene-specific sgRNA in C. elegans, which is cost-effective and efficient. However, it is worth noting that targeting either the different genes or the same gene in different sites using our VPR-dCjCas9 feeding system exhibited markedly different activation efficiencies (Table 4). Therefore, a preliminary screening to obtain suitable and efficient sgRNAs is required before corresponding studies using the VPR-dCjCas9 feeding system. To better select the functional sgRNAs, we analyzed the distribution of all designed sgRNAs in the promoter region of the different target genes and found that the functional sgRNAs were primarily distributed between 400 and 700 bp upstream of the 5 0 UTR of target genes (Fig. 1g). Moreover, a previous study indicated that overexpression SID-1 and SID-2 transgenic worms could be selected as a powerful genetic background to increase gRNA uptake (Liu et al. 2014). In further study, we will attempt to use the SID-1 Fig. 1. Continued VP64-dCjCas9 strains bearing the Pmyo-2::GFP::H2B transgene in the presence or absence of myo-2-specific sgRNA targeting the promoter region. c) Survival analyses of the VPR-dCjCas9 worms in the presence or absence of aak-2-, pha-4-, or lipl-4-specific sgRNA targeting the promoter region. d) Fat levels in the intestinal region of the VPR-dCjCas9 strains in the presence or absence of lipl-4-or lipl-5-specific sgRNA targeting the promoter region. e) Survival analyses of the VPR-dCjCas9 worms in the presence or absence of the spr-4-specific sgRNA targeting the promoter region. f) Relative mRNA expression of target genes in the VPR-dCjCas9 worms in the presence or absence of the specific sgRNA targeting the target gene promoter region. sgRNA (À), empty vector L4440; sgRNA (þ), gene-specific sgRNA plasmid; NC sgRNA, negative control sgRNA plasmid. g) Schematic of the distribution of all designed sgRNAs in the promoter regions of the target genes. h) Relative mRNA expression of tissue-specific genes, including pie-1, ges-1, and rab-3, in the VPR-dCjCas9 worms in the presence or absence of the specific sgRNA targeting the target gene promoter region. In (b) and (d), data are displayed as the mean 6 SD, n ! 30. In (c) and (e), lifespan analyses were performed using the Kaplan-Meier plotter, and the p value was determined by the log-rank test. The detailed lifespan values are listed in Table 5. In (f) and (h), the data are displayed as the mean 6 SD of 3 independent experiments. In (b), (d), (f), and (h), ***P < 0.001, *P < 0.05, ns, not significant (Student's t-test). and SID-2 over-expression transgenic background to optimize our VPR-dCjCas9 feeding system. Furthermore, the expression of most genes could be regulated by transcription factors or epigenetic modifiers. Therefore, combining dCjCas9 with other functional domains, such as an epigenetic modifier, would greatly expand the applicability of our CRISPR-Cas9 feeding system in C. elegans. Overall, this novel dCjCas9-based feeding system may hold great promise for genome editing, transcriptome modulation, and other applications in C. elegans.

Data availability
The C. elegans strains and plasmids are available upon request. Table 1 contains the sgRNAs and primers sequence.  contains the full sequence of feeding sgRNA cassette. Table 3 contains the primers used in qRT-PCR. Table 4 contains the sgRNA screening data of all candidate genes. Table 5 contains the statistics data of all lifespan assays. The authors affirmed that all data necessary for confirming the conclusions of the article are present within the article. QZ, Q-LW, and ZL designed the study, ZL, WD, CW, and QY conducted the experiments. Q-LW, ZL, and CW analyzed the data. ZL and WD wrote the manuscript. QZ and Q-LW reviewed and edited the manuscript.