Effects of CRISPR/Cas9-mediated stearoyl-Coenzyme A desaturase 1 knockout on mouse embryo development and lipid synthesis

Animals and ethics statement

Ten-week-old male and female C57BL/6 and ICR mice were purchased from Air Force Medical University (Shaanxi, China) and housed in a specific-pathogen-free (SPF) animal facility with a 12-h light/dark cycle at a controlled temperature (22 ± 2 °C), humidity of 40%–60%, and ad libitum access to water and food. Same sex mice were housed together in individually ventilated cages with four mice per cage. All animal protocols were performed in accordance with the approval of the Animal Ethical and Welfare Committee of College of Animal Science and Technology, Northwest A&F University, Yang Ling, China (permit number: 15-516, date: 2019-8-1).

Isolation and cultivation of mouse zygotes

Female mice were selected randomly for zygote isolation. Superovulation was induced in female mice via intraperitoneal injection of 10 IU of pregnant mare serum gonadotropin (PMSG, Ningbo, China). After 46–48 h, 10 IU of human chorionic gonadotropin (hCG, Ningbo, China) was intraperitoneally injected in the mice. Mice were included in the study if they were in good health and in oestrus. Mice were excluded if they were ill or did not enter oestrus. There were 10 female mice treated for superovulation in the one experiment. The total number of animals used in the whole experiment was 200. Twelve hours later, female mice that had successfully mated with male mice were checked for the presence of a vaginal plug. If there was no vaginal plug, the mouse was not used in the experiment. Females were sacrificed by cervical dislocation; in detail, the mouse was removed from its cage and placed on a smooth, hard surface without releasing the tail. A metal rod was placed firmly behind the ears and across the neck, and the tail was pulled sharply to the rear while pressing down on the neck. Zygote-cumulus complexes were isolated from the most swollen oviduct in M2 medium (M7167, Sigma, St. Louis, MO, USA). The average number of zygotes isolated from each mouse was about 15, and the total number of zygotes was about 150 in one experiment. The zygotes were randomly allocated into different groups for microinjection, placed in hyaluronic acid in order to release granulosa cells, and cultured in M2 medium until microinjection. Injected zygotes were cultured in EmbryoMax^® KSOM Mouse Embryo Media (MR-121-D, Sigma) until the morula and blastocyst stages.

Specific single guide RNA design and vector construction

The sgRNAs that were targeted to exon 1 and exon 2 of the Mus musculus Scd1 gene (Gene ID: 20249) were designed using the online website CHOPCHOP (http://chopchop.cbu.uib.no/) and Guide Design Resources (https://zlab.bio/guide-design-resources). Three sgRNAs that were targeted to exon 1 and two sgRNAs targeted to exon 2 were selected according to their predicted scores. The sequences of sgRNAs and PAM motifs are shown in Fig. 1A. The sgRNAs were added to the Bsa I (R3733S, NEB, Ipswich, MA, USA) restriction enzyme site according to a previous study (Shen et al., 2013) and synthesized as single-strand DNA oligonucleotides by Sangon Biotech (Shanghai, China). The oligonucleotides were then annealed and inserted into a pUC57-sgRNA expression vector using T4 DNA ligase (2011A, Takara Bio Inc., Otsu, Japan) according to the manufacturer’s instructions. The negative control sgRNA (NC sgRNA) sequence (5′-ACCGGAAGAGCGACCTCTTCT-3′), which did not target to any sites in the mouse genome according to a previous report (Yang et al., 2017), was also synthesized and constructed into the vector, which was used as a control in the experiment.

In vitro transcription of sgRNA, Cas9 mRNA, and analysis of sgRNA activity

pST1374-NLS-flag-linker-Cas9 vector (Addgene#44758) was linearized with Age I (R3552S, NEB) and used for Cas9 mRNA in vitro transcription using the T7 Ultra Kit (AM1345; Invitrogen, Waltham, MA, USA), and the Cas9 mRNA was purified using an RNeasy Mini Kit (74104; Qiagen, Hilden, Germany). For sgRNA in vitro transcription, the DNA templates were obtained using PCR amplification from pUC57-sgRNA expression vectors according to a previous study (Shen et al., 2014). In brief, the PCR was amplified using PrimeSTAR Max DNA Polymerase (R045A; Takara, Shiga, Japan) according to the manufacturer’s protocol and with the following primers: forward primer, 5′- TCTCGCGCGTTTCGGTGATGACGG-3′and reverse primer, 5′- AAAAAAAGCACCGACTCGGTGCCACTTTTTC-3′. The PCR conditions were as follows: 98 °C for 10 s, followed by 56 °C for 5 s, and 72 °C for 5 s, 33 cycles in total. The PCR product was run on 1.5% agarose gels and purified using an AxyPrep DNA Gel Extraction Kit (AP-GX-50, Axygen, NY, USA). sgRNA in vitro transcription was performed using a MEGAshortscript™ Kit (AM1354, Invitrogen, Waltham, MA, USA) and purified using a MEGAclear™ Kit (AM1908, Invitrogen, Waltham, MA, USA).

To detect the activity of sgRNAs obtained by in vitro transcription, Scd1 gene exon 1 and exon 2 fragments containing sgRNA target sites were obtained by PCR and used for sgRNA activity experiments. The primers used for exon 1 (E1) and exon 2 (E2) amplification were as follows: E1-F, 5′-AATACTGAACACGGTCATCCCA-3′; E1-R, 5′- ACTCATCTGCCCAAATTACAATC-3′; E2-F, 5′- ATCTGTTTTCCGATGGTCTT-3′; E2-R, 5′- ACAGGGACTCAGTATTCATGTTA-3′. PCR amplification was performed under the reaction conditions: 98 °C for 10 s, 55 °C for 5 s, and 72 °C for 10 s. Individual sgRNA at 100 ng and targeted DNA fragments at 300 ng were mixed with 0.3 µL Cas9 Nuclease, S. pyogenes (M0646M; NEB, Ipswich, MA, USA), and 10 × NEBuffer 3.1 at 20 µL volume in total. The mixture was then incubated at 37 °C for 30 min. The DNA fragments were purified using AxyPrep PCR Cleanup Kit (AP-PCR-50, Axygen) and agarose gel electrophoresis was performed to analyze whether the DNA fragments were cut off by sgRNA and Cas9 complex in order to examine sgRNA activity in vitro.

sgRNA/Cas9 cytoplasmic microinjection

In order to determine the optimal cell cycle phase for microinjection, RNase-free water was injected into the cytoplasm of mouse zygotes at the G1, S, and G2 phases using the FemtoJect system (Eppendorf, Hamburg, Germany) at 400 hPa injection pressure, 30 hPa compensatory pressure, and 0.3 s injection time. The G1, S, and G2 phases were determined to be 8 to 10 h, 14 to 16 h, and 18 to 20 h postfertilization, respectively, according to a previous study (Gu, Posfai & Rossant, 2018). The embryo survival rates were calculated immediately after microinjection. In order to select sgRNAs with higher gene-editing efficiency, each sgRNA (10 ng/µL) was separately mixed with Cas9 mRNA (20 ng/µL), the mixture was injected into the zygotes, and then the zygotes were cultured in KSOM at 37 °C with 5% CO₂ until morulae or blastocysts formed for genotyping. In order to explore the effects of Scd1 knockout on embryo development, zygotes were divided into three groups for microinjection at the optimal cell cycle phase: RNase-free water, NC sgRNA with Cas9 mRNA, and Scd1 sgRNA with Cas9 mRNA. The morulae and blastocysts were observed 3.5 days following fertilization and the rates of morula and blastocyst formation were calculated.

Mouse embryo transfer

The embryo transfer was conducted as previously described (Kolbe et al., 2012). The pseudo-pregnant foster mothers were prepared by mating ten-week-old ICR female mice with vasectomized male mice on the same day the zygote donor mice were mated. Successful mating was determined by the presence of a vaginal plug the following morning. Before embryo transfer, the surrogate mothers were checked again to determine whether ovulation had occurred. Females with a swollen ampulla (the sign of ovulation) were used as surrogate mothers. After zygotes were injected with Cas9 mRNA and sgRNA mixture for 24 h, the two-cell stage embryos were transferred into pseudo-pregnant foster mouse oviducts. Females were anesthetized via intraperitoneal injection and placed on a warm plate at 37 °C. The mouse’s back area near the ovary was shaved and disinfected with ethanol. An incision was made on the right side to pull out the ipsilateral ovary, oviduct, and uterine horn. The thin transparent connective tissue membrane was opened, and 16 two-cell embryos were transferred into the ampulla. The reproductive tract was then placed back inside and the incision was sutured. There were no adverse events during embryo transfer. The pups were delivered about 21 days after embryo transfer. The mice were weighed every week after birth. The surviving animals were kept alive at the conclusion of the experiment.

Genotyping of mouse embryos and tails

The single embryos that developed to morulas or blastocysts were collected for genomic DNA extraction using a REPLI-g Single Cell Kit (150343, Qiagen, Hilden, Germany). PCR was performed to amplify the exon 1 and exon 2 of the Scd1 gene using PrimeSTAR Max DNA Polymerase (R045A, Takara, Shiga, Japan) with the reaction conditions as follows: 98 °C for 10 s, 55 °C for 5 s, and 72 °C for 5 s. The primers were the same as those used for the sgRNA in vitro activity analysis experiment. The PCR product was used for Sanger sequencing and inserted into the pMD19-T vector to verify the sequence mutation in the genome.

Mice tails were isolated by biopsy three weeks after birth. Mouse tail genomic DNA was extracted using a Universal Genomic DNA Kit (CW2298S, CW Biotech, Beijing, China). The PCR products of exon 1 and exon 2 were obtained using the same method as that of the embryos and purified using a PCR Clean-Up Kit (AP-PCR-50, Axygen, NY, USA). Purified DNA was annealed for T7EN1 cleavage assay (M0302L, NEB, Ipswich, MA, USA) and analyzed using agarose gel electrophoresis according to a previous study (Vouillot, Thelie & Pollet, 2015). The DNA was then inserted into a pMD19-T vector. A single bacterial colony was picked up for sequencing by Sangon Biotech (Shanghai, China) to evaluate the nucleotide modifications.

Protein isolation and Western blot

The mice born after embryo transfer that were identified as wild type were used as the control group to be compared with Scd1 gene modified mice in the following experiments. For each animal, three different investigators were involved in the experiment as follows: the first investigator collected the samples and was the only person aware of the group allocation. A second investigator was responsible for the sample analysis. Finally, a third investigator (also unaware of the mouse’s group) assessed the data analysis.

Mouse tail tissue was placed in liquid nitrogen to be ground and then lysed in ice-cold RIPA buffer (R0010; Solarbio, Beijing, China) with protease inhibitor (04693132001; Roche Diagnostics Ltd., Mannheim, Germany). A BCA protein assay kit (23227; Thermo Fisher Scientific, Rockford, IL, USA) was used for protein concentration measurement. Protein was separated using 10% SDS-PAGE and transferred onto a nitrocellulose membrane (HATF00010; Millipore, Burlington, MA, USA). After being blocked for 1.5 h using 5% skim milk, the membranes were probed with primary antibodies, monoclonal mouse anti-SCD1 (ab19862, 1:1000; Abcam, Cambridge, MA, USA), and monoclonal mouse anti- β-tubulin (CW0098, 1:1000; CW Biotech, Beijing, China). The membranes were washed three times with TBST, and HRP-conjugated goat anti-mouse-IgG (CW0102, 1:2000; CW Biotech) was used as a secondary antibody. Signals were detected using an ECL Western blot system (1705061, Bio-Rad, Hercules, CA, USA). ImageJ software was used to analyze the densities of the bands and SCD1 relative expression was normalized to β-tubulin.

Measurement of total serum triacylglycerol (TAG) and cholesterol

Total serum TAG and cholesterol were detected using a GPO-Trinder TAG assay kit (E1003; Applypen Technologies Inc., Beijing, China) and a cholesterol assay kit (E1005; Applypen Technologies Inc.). The tip of the mouse tail was clipped, and blood was collected from the tail using a hand-held pipette when the mice reached four and eight weeks old. Serum was collected from clotted blood and used for TAG and cholesterol assays at 550 nm on a Biotek microplate reader (Winooski, VT, USA).

Statistical analysis

All data were analyzed using SPSS 19.0 (SPSS, Inc., Chicago, IL, USA) and the results were presented as means ±SEM with a 95% confidence interval. Significance across the three groups was analyzed using one-way ANOVA and a general linear model procedure, and lowercases (a to c) denoted significant differences (P < 0.05). The statistical analysis of the other parameters between the two groups was performed using the Student’s t-test and statistical significance was considered at *P < 0.05 and **P < 0.01.

Zygote microinjection during the S phase showed the highest embryonic survival rate

The mouse zygotes that were injected with RNase-free water at different cell cycle phases and microinjected during the S phase had the highest embryonic survival rate (P < 0.05) compared with those injected during the G1 and G2 phases (Table 1). This result indicated that the S phase was the optimal time for microinjection with less damage to zygotes.

Four sgRNAs targeted to Scd1 revealed in vitro activity

Of the five sgRNAs (Fig. 1A), sgRNA2 and sgRNA3 targeted to exon 1 of the Scd1 gene, and sgRNA6 and sgRNA7 targeted to exon 2, showed in vitro activity (Fig. 1B). Based on these results, sgRNA2, 3, 6, and 7 were used for Scd1 knockout in mouse zygotes.

Gene editing efficiency of different sgRNAs and Scd1 gene sequence modification in mouse embryos

Individual sgRNA was mixed with Cas9 mRNA, and four groups of RNA mixture were used for cytoplasmic microinjection in the S phase. The gene editing efficiencies of sgRNA2, 3, 6, and 7 were 9.1%, 36.4%, 45.6%, and 20.0%, respectively (Table 2). DNA sequences at the sgRNA targeted sites were also analyzed. One embryo had two genotypes in the sgRNA3 group, and the sgRNA6 group had one embryo with two genotypes and one embryo with three genotypes, indicating that these embryos were mosaicisms (Fig. 2).

Knockout of Scd1 prevented blastocyst formation

Zygotes injected with NC sgRNA and Cas9 mRNA had a lower blastocyst formation rate (P < 0.05) compared with the RNase-free water group. The embryo images of the three groups’ morulae and blastocysts are shown in Fig. 3. Compared with the two control groups, the blastocyst formation rate decreased (P < 0.05) after injection with Scd1 sgRNA and Cas9 mRNA (Table 3). This indicated that knockout of Scd1 prevented zygote development into blastocyst, and microinjection of sgRNA and Cas9 mRNA had a negative effect on embryo development.

Sequence modifications at Scd1 gene locus in mice

We used sgRNA3 targeted to exon 1 and sgRNA6 targeted to exon 2, which had relatively higher gene editing efficiency, for Scd1 knockout mice generation. Six mice were born after embryo transfer, and PCR was performed to obtain the fragments of exon 1 and exon 2, as shown in Fig. 4A. Of these mice, three were identified with Scd1 sequence modifications at sgRNA target sites by T7EN1 assay and sequencing (Figs. 4B and 4D). The photograph of the knockout mice is shown in Fig. 4C. Two mice had nucleotide mutations only in sgRNA3 target sites, and another mouse genome had sequence modifications at both sgRNA3 and sgRNA6 target sites (Fig. 4D). The wild type genotype could be detected in three mice with mutations, suggesting that they were Scd1 monoallelic knockout mice. Three wild type mice that were born after embryo transfer were used as the control group in the following experiment.

Scd1 monoallelic knockout mice showed decreased body weights, serum TAG, and cholesterol

The Scd1 knockout mice were inactive compared with the wild type mice. However, it took a longer time for the knockout mice to be mated. Scd1 knockout (n = 3) and wild type mice (n = 3) were analyzed for protein, serum TAG, and cholesterol detection. The SCD1 protein level decreased (P <  0.05) by 40% in Scd1 sequence modified mice (Figs. 5A and 5B), which also indicated monoallelic knockout of Scd1 in mice. The body weights of Scd1 knockout mice showed a continuous decrease (P < 0.05) from four weeks compared with the wild type mice (Fig. 5C). Additionally, the serum TAG and cholesterol reduced in Scd1 knockout mice (P < 0.01) (Figs. 5D and 5E).