Electroporation of CRISPR/Cas9 Targeting Neurogenin 3 ( NGN3 ) in Porcine Embryos and Its Effects on Mosaicism and Off-Target Effects by Next Generation Sequencing (NGS)

Abstract

Introduction:

CRISPR/Cas9-mediated editing of embryos through microinjection is limited in efficiency because of high rates of mosaicism and off-target mutations. Electroporation has been proposed as an easier and faster procedure compared to single embryo injection.

Methods:

Using a synthetic guide-RNA targeting porcine NGN3, we optimized the electroporation conditions to effectively introduce CRISPR/Cas9 in porcine zygotes and evaluate its effects on mosaicism and off-targets.

Results:

A numerical increase in the mutation rates was observed as embryos were electroporated with increasing concentrations of gRNA (5 ng/μl, 10 ng/μl, and 25 ng/μl) and Cas9 protein (10 ng/μl, 20 ng/μl, and 50 ng/μl) (1:2ratio) without compromising embryo development. Mosaicism results assessed by long-read targeted sequencing of embryos revealed that electroporation with the highest concentration of CRISPR/Cas9 resulted in 57.14% mosaic embryos, accompanied by a significant reduction in the average allele variants (2.43 alleles per embryo) compared to the lowest concentration (4.56 alleles per embryo). Off-target results suggested that no unintended indels were observed when electroporation with the highest CRISPR/Cas9 concentration (25 ng/μl:50 ng/μl) was used.

Conclusion:

Embryo electroporation with 25 ng/μl:50 ng/μl of CRISPR/Cas9 targeting porcine NGN3 resulted in a very high mutation efficiency, minimal mosaicism, and no off-target mutations.

Introduction

Pigs have been attractive candidates for studying human diseases due to their similarities to humans in organ size, physiology, genetics,1 –3 and for xenotransplantation.4 Several studies have successfully generated genetically engineered pig models to study human diseases, such as cancer,5 muscular dystrophy,6 hyperlipidemia,7,8 and immunodeficiency.9 These transgenic pig models were created by somatic cell nuclear transfer from fibroblasts genetically modified using standard gene targeting methods; however, this technique is difficult to perform.

Rapid advances of gene-editing technology using clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein 9 (Cas9) have enabled a relatively fast and easy generation of genetically engineered animals, such as mice10 and livestock,11 –16 by direct modification of 1-cell embryos. Most studies used cytoplasmic microinjection to deliver CRISPR/Cas9 into the embryos, which requires a time consuming and laborious process and limits the number of embryos that can be treated. Furthermore, several studies have shown that microinjection of CRISPR/Cas9 into the zygotes frequently resulted in genetic mosaicism in mice,17,18 cattle,19 and pigs.20 Mosaicism likely results from a delayed time of microinjection or a delayed activity of CRISPR/Cas9, resulting in modifications of the genome after the first DNA replication that in embryos happens in the first few hours post-fertilization.21 In line with this, microinjection of CRISPR/Cas9 into oocytes before insemination greatly reduced mosaicism rates.22 In addition, unintended off-target mutations can be caused by Cas9 activity leading to unwanted mutagenic insertions and deletions (indels) as well as chromosomal rearrangements.23,24

Alternatively, electroporation has been proposed as an efficient approach to introduce CRISPR/Cas9 components due to its relatively technically undemanding process in which the embryos can be treated as a group with a single electric pulse setting compared to the conventional microinjection procedure.25 –34 In addition, some studies have shown that the use of electroporation with a single gRNA along with Cas9 protein in zygotes may reduce mosaicism.28,35 However, the issues associated with mosaicism have not been resolved completely.36 –38 Since the creation of founder animals with invariable phenotypes is difficult because of mosaicism and potential off-target effects caused by the CRISPR/Cas9 system in embryos, deciphering and optimizing the system to make ideal homozygotic mutants is desirable.

Neurogenin 3 (NGN3) is responsible for generating pancreatic endocrine cells (islets of Langerhans) during embryonic development in mice and pigs. Knockout (KO) of mouse NGN3 resulted in deficiency of all four types of islets, thus leading to the development of diabetes and death 2–3 days after birth.39 A CRISPR/Cas9-induced NGN3 KO in pigs also resulted in loss of pancreatic endocrine hormones, glucagon, insulin, somatostatin, and pancreatic polypeptide, which are produced by alpha, beta, delta, and gamma cells, respectively. The KO newborn piglets lost significant weight during first 24–36 h after birth, possibly because of diabetes.39 Efficient generation of NGN3 KO pigs using the CRISPR/Cas9 system has great potential as a platform to study diabetes.

To generate consistent phenotypes, multiple experimental factors need to be adjusted to improve the efficiency of achieving nonmosaic, homozygotic gene-edited founder pigs and to further refine the use of CRISPR/Cas9 technology in zygotes. Here, we introduced different concentrations of porcine NGN3-targeted CRISPR/Cas9 (single gRNA along with Cas9 protein) by electroporation and determined the mutation efficiency, levels of mosaicism, and off-target effects using next generation sequencing in porcine embryos.

Materials and Methods

Oocyte collection

Oocytes were aspirated from antral follicles (3–6 diameters) from prepubertal gilt ovaries collected at a local slaughterhouse (Olson Meat Company, Orland, CA, USA). Cumulus-oocyte complexes (COCs) were washed in TCM199-HEPES-buffered medium (Sigma).

Invitro maturation (IVM)

For invitro fertilization (IVF), COCs with two to three layers of cumulus cells were cultured in maturation medium at 38.5°C in a humidified incubator containing 5% CO₂. The maturation medium was made by adding 1 mM dibutyryl cyclic adenosine monophosphate (AMP) (Sigma), 0.1% polyvinyl alcohol (PVA) (Sigma), 3.05 mM d-glucose (Sigma), 0.91 mM sodium pyruvate (Sigma), 0.5 μg/ml porcine follicle-stimulating hormone (pFSH), 0.5 μg/ml porcine luteinizing hormone (pLH), 10 ng/ml epidermal growth factor (EGF) (Invitrogen), 10 μg/ml gentamicin (Gibco), and 10% porcine follicle fluid (PFF). After 20–22 h of incubation, COCs were moved to maturation medium without cAMP, pFSH, and pLH for 20–22 h. For parthenogenetic activation (PA), maturation medium was made by adding 0.1% PVA, 3.05 mM d-glucose, 0.91 mM sodium pyruvate, 0.5 μg/ml pFSH, 0.5 μg/ml pLH, 10 ng/ml EGF, 10 μg/ml gentamicin (Gibco), and 10% PFF. COCs were cultured at 38.5°C in a humidified incubator containing 5% CO₂ for 40–44 h.

Parthenogenetic activation

After IVM, matured oocytes were stripped of their cumulus cells by incubation in 1 mg/ml hyaluronidase (Sigma) and gentle pipetting. Denuded oocytes were washed with TCM199 containing 25 mM HEPES (Sigma) supplemented with 1% PFF and electrically activated with two pulses of 120 V/mmfor 40 μs, delivered by a BTX Electro Cell Manipulator 2001 (BTX, San Diego, CA, USA) in a 0.5 mmchamber containing 0.3 M mannitol, 0.05 mM CaCl₂, 0.1 mM MgSO₄, and 0.1% bovine serum albumin. After washing with porcine zygote medium (PZM)-5,40 the oocytes were incubated in the presence of 5 μg/ml cytochalasin B in PZM-5 for 3 h to prevent second polar body extrusion, thus generating diploid parthenogenetic embryos.

Invitro fertilization

Denuded oocytes were first placed in a 90 µl mTBM drop (20 oocytes/drop). Fresh semen for IVF was collected from the UC Davis swine facility. A total of 100 µl of fresh semen was added to a tube containing 1.9 ml of mTBM and centrifuged twice at 100 × G for 3 min. The spermatozoa pellet was resuspended in 1 ml of mTBM. Resuspended spermatozoa were counted and diluted to 2 × 10⁶ cells/ml. A volume of 10 µl of counted sperm was added to the denuded oocytes drops (20,000 cells/drop) and co-incubated for 5–6 h.

Single guide RNA (sgRNA) design

We used the online software (MIT CRISPR Design Tool: http://crispr.mit.edu) to design sgRNA targeting sequence (Exon 2) of the pig NGN3 gene. The target sequence (20 bp target including the underlined 3 bp protospacer adjacent motif (PAM) sequence) was 5′-CGTTGGATGCGCTGCGCGGG-3′.

Electroporation and invitro culture

Electroporation was performed using a NEPA21 Electroporator and CU500 Cuvette Adaptor (NEPAGENE, Ichikawa, Japan). For tetramethylrhodamine-labeled dextran (Sigma) experiments, parthenogenetically activated embryos and invitro fertilized embryos were washed three times with Opti-MEM (Gibco) and transferred to a drop containing 10 µl of Opti-MEM and 10 µl of tetramethylrhodamine-labeled dextran (10 mg/ml in phosphate buffered saline) immediately after the embryo production procedures. Then, the whole drop was gently mixed with pipetting and transferred to a 1 mmcuvette, which was then inserted in the CU500 cuvette adaptor. The electroporation condition was a transfer pulse of 30 V, 1 ms exposure, 100 ms pulse width (interval), 0% decay rate, and 10 pulses (bipolar). After electroporation, embryos were washed three times with PBS, and fluorescence was validated under the microscope (Nikon). For CRISPR/Cas9 experiments, putative porcine embryos were washed three times with Opti-MEM (Gibco). A drop with a total volume of 20 µl containing 10 µl of NGN3 sgRNA (Synthego) and Cas9 protein (PNA Bio) adjusted to a 1:2 ratio and 10 µl of Opti-MEM was prepared. Then, the same procedure with the same electroporation condition was conducted as described earlier. After electroporation, embryos were cultured in PZM-5 medium at 38.5°C in a humidified atmosphere of 5% CO₂, 5% O₂, and 90% N₂ for 7–8 days. The embryo cleavage rate and the blastocyst rate were observed on day 2 and on day 5–8 after IVF, respectively.

Preparation of DNA from a single blastocyst and genotyping

Genomic DNA derived from single blastocysts was prepared for use as a PCR template. Blastocysts were lysed in Epicenter DNA extraction buffer (Lucigen) by incubation at 65°C for 6 min and 98°C for 2 min. The resulting DNA solution was stored at −20°C until use. Lysed blastocyst DNA samples were analyzed by performing two rounds of PCR using GoTaq Hot Start Green Master Mix (Promega Biosciences LLC) with primers specific to NGN3 sequences in pig (accession #VGNC:90702): (F: 5′-CGCCCTAAAACGAGGTAACA-3′; R: 5′-GGAGGAACAAGTACGCCTGA-3′).

The PCR conditions were 95°C for 3 min, followed by 34 cycles at 95°C for 30 s, 58°C for 40 s, 72°C for 45 s, and a final step at 72°C for 5 min. PCR products were subjected to gel electrophoresis, and DNA bands were excised from the gel and purified using the QIAquick Gel extraction kit (QIAGEN). Amplicons were Sanger sequenced by Genewiz (https://www.genewiz.com).

Evaluation of mosaicism and off-target analysis

To analyze mosaicism rates, NGN3 PCR products from 30 IVF blastocysts and 27 parthenogenetic blastocysts (carrying two or more DNA peaks from the SnapGene program) analyzed for genotyping were used. A dual round PCR was performed to barcode each sample using GoTaq Hot Start Green Master Mix (Promega Biosciences LLC). Thesame primers for genotyping the NGN3 gene were used with the following 15-bp adapter sequence attached to the forward (AGATCTCTCGAGGTT) and reverse (GTAGTCGAATTCGTT) primers.

The second round of PCR was performed to add a barcode sequence to each sample to identify reads from pooled sequencing (Supplementary Table S9). PCR conditions were 95°C for 3 min, followed by 15 cycles at 95°C for 30 s, 60°C for 40 s, 72°C for 45 s, and a final step at 72°C for 5 min for the first PCR and 30 cycles at 95°C for 30 s, 51°C for 40 s, 72°C for 45 s, and a final step at 72°C for 5 min in the second round. DNA purification of PCR products was performed using NucleoSpin Gel and PCR clean-up kit (Macherey-Nagel).

To analyze off-target effects, seven regions of the Sus Scrofa genome (Supplementary Table S10) were selected as possible off-target cut sites using Breaking-Cas software (BioinfoGP, CNB-CSIC, Spain).41 Fifteen IVF blastocysts electroporated with CRISPR/Cas9 ribonucleoprotein against NGN3 were used to evaluate off-target mutations. DNA extraction was performed as described earlier. Primers for NGN3 and off-target sequences were designed using Primer-Blast (NCBI, USA). To add a barcode in a final PCR, the same 15-bp adaptor sequences were attached to the 5′ end of these forward and reverse primers (Supplementary Table S10).

To sequence all the off-target regions from a single blastocyst, a three-step PCR was performed using GoTaq Hot Start Green Master Mix (Promega Biosciences LLC). For the first PCR, a multiplex PCR was performed using all primers. PCR conditions were 95°C for 3 min, followed by 30 cycles at 95°C for 30 s, 60°C for 40 s, 72°C for 45 s, and a final step at 72°C for 5 min.

For the second round, 1 µl of first PCR product and a single pair of primers were used per tube. PCR conditions were 95°C for 3 min, followed by 15 cycles at 95°C for 30 s, 60°C for 40 s, 72°C for 45 s, and a final step at 72°C for 5 min. Subsequently, a third PCR was performed to add the barcode sequence for each sample to identify reads from pooled sequencing. A total of 1 µl from the second round was used. PCR conditions were 95°C for 3 min, followed by 30 cycles at 95°C for 30 s, 51°C for 40 s, 72°C for 45 s, and a final step at 72°C for 5 min.

All the PCR samples targeting the gRNA cut sites were prepared according to the SMRTbell library preparation protocol (https://dnatech.genomecenter.ucdavis.edu) and sequenced on a PacBio Sequel II sequencer at the UC Davis Genome Center.42 Consensus sequences were called, sorted by barcode, and converted to individual a text-based sequencing data file format that stores both raw sequence data and quality scores (FASTQ) files using SMRT Link v8.0.0.80529 (https://www.pacb.com/support/software-downloads). Reads were aligned to each target site using BWA v0.7.16a.43 sequence alignment map (SAM) files were converted to compressed binary version of a SAM (BAM) files, sorted and indexed using SAMtools v1.9.44 CrispRvariants (GitHub-markrobinsonuzh/CrispRVariants) was used to analyze the sorted alignments. In the analysis, regions of approximately 40-bp in the predicted cut site were analyzed for the target gene (NGN3) and for off-target sites. To remove reads containing PCR/sequencing errors, we filtered the alleles using a frequency threshold to retain a maximum of two alleles within the 300 bp flanking regions upstream and downstream of the predicted cut site. The frequency threshold was, therefore, unique to each sample. All reads containing duplicate/inverted sequences of one or more genomic regions were combined for each sample. The codes used for the processing of the raw reads, the CrispRVariants analysis, and background removal can be found at GitHub-RaquelPinho/NGN3.

Statistical analysis

All the mean embryo development data from different time points and replicates were analyzed by one-way Analysis of Variance (ANOVA) using GraphPad Prism 8.0.1 (GraphPad). Values from at least three replicates of each experiment are presented as means ± standard error of the mean ormeans ± standard error of the proportion. Mutation efficiency from PacBio data and allele numbers for evaluating mosaicism were analyzed using t-test. Probability values less than 0.05 (P < 0.05) were considered statistically significant.

Results

Optimization of CRISPR/Cas9 introduction by electroporation

To test the feasibility of electroporation in pig zygotes, an electroporation condition was tested to assess membrane permeability (Fig.1A). Embryos were electroporated with tetramethylrhodamine-labeled dextran (3 kDa) immediately after PA or IVF and visualized under a fluorescence microscope. We observed that the electroporated embryos showed fluorescence localized in their cytoplasm, indicating the electroporation condition was able to sufficiently permeabilize zona pellucida and cytoplasm (Fig.1B).

FIG. 1.

Effective electroporation of porcine embryos. (A) Diagram representing the electroporation conditions of 30 V; 1-ms exposure; 100 ms interval (between pulses) with two sets of five pulses with inverted polarities. (B) Embryo uptake of tetramethylrhodamine-labeled dextran (3 kDa) after electroporation. PA: parthenogenetically activated embryos.

To further determine the efficiency of CRISPR/Cas9 delivery by electroporation and its impact on embryo development, a single guide RNA (CRISPR) targeting pig NGN3 was designed. Three different gRNA concentrations, 5 ng/µl, 10 ng/µl, and 25 ng/µl, retaining the CRISPR/Cas9 ratio at 1:2 were tested. Introduction of NGN3-targeted gRNA along with Cas9 protein in groups of about 35–50 embryos was performed by electroporation 9 h after activation. Cleavage rates of parthenotes for each of the CRISPR/Cas9 electroporated concentrations were similar to the untreated control group (Fig.2A). Furthermore, blastocyst rates were not affected by the electroporation procedure (Fig.2B). Mutation analysis of single blastocysts by Sanger sequencing showed that 5 ng/µl, 10 ng/µl, and 25 ng/µl CRISPR concentrations had indel rates of 14%, 44%, and 85% indels, respectively (Fig.2C and Supplementary Tables S1 and S3).

FIG. 2.

Electroporation of parthenogenetically activated embryos with increasing concentration of CRISPR/Cas9-targeting porcine NGN3. (A) Day 2 cleavage rates and (B) blastocyst rates of parthenotes electroporated with different concentrations of CRISPR/Cas9. (C) Gene editing rates of embryos electroporated with three different concentrations of CRISRP/Cas9. Numbers in parentheses represent mutated embryo numbers of total assessed.

Since the electroporation condition did not affect the development of parthenogenetically activated porcine embryos, we sought to determine if NGN3-targeted CRISPR/Cas9 introduction using electroporation would have similar results in invitro fertilized pig embryos. Electroporation of porcine zygotes was performed 9 h after insemination. In electroporated IVF-derived embryos, neither cleavage nor blastocyst rates were different from nontreated control embryos (Fig.3A and B). Mutation rates by Sanger sequencing were 78% for the 5 ng/µl group, 85% for the 10 ng/µl group, and 100% for the 25 ng/µl group (Fig.3C and Supplementary Tables S2 and S4).

FIG. 3.

Electroporation in invitro fertilized porcine embryos with increasing concentration of CRISPR/Cas9 targeting porcine NGN3. (A) Day 2 cleavage rates and (B) blastocyst rates of zygotes electroporated with different concentrations of CRISPR/Cas9. (C) Gene editing rates of embryos electroporated with different concentrations of CRISRP/Cas9. Numbers in parentheses represent mutated embryo numbers of total assessed.

Evaluation of mutation, mosaicism rate, and off-target effects by next generation sequencing (NGS)

To further analyze the mosaicism rate following CRISPR/Cas9 zygote electroporation, 27 parthenogenetically activated blastocysts (one from the 5 ng/µl group; eight from the 10 ng/µl group; 18 from the 25 ng/µl group) and 30 IVF-derived blastocysts (nine from the 5 ng/µl group; 14 from the 10 ng/µl group; seven from the 25 ng/µl group) that had mutations and potential mosaicisms (multiple DNA peaks) at the target site were selected, barcoded by PCR amplification, and sequenced on a PacBio sequencer. Circular consensus sequencing (CCS) reads from PacBio data were sorted and analyzed by individual embryos.

Among the 27 parthenogenetically activated embryos, the single embryo from the 5 ng/µl group, four embryos from the 10 ng/µl, and 11 embryos from the 25 ng/µl groups were mosaic (Fig.4A and B). Mean mutation efficiencies were 38%, 52%, and 85% for the 5 ng/µl, 10 ng/µl, and 25 ng/µl groups, respectively. The 25 ng/µl CRISPR concentration group had a significantly higher mean mutation efficiency than that of the 10 ng/µl group (Fig.4C). Except for the lowest concentration group which had only one sample, mean numbers of alleles per embryo were 3.29 ± 0.68 for the 10 ng/µl group and 4.06 ± 0.95 for the 25 ng/µl group (Fig.4A and D). Calculation of embryos that carried a wildtype allele showed that 1 of 1, 6 of 7 (unedited embryo 2 excluded), and 9 of 18 embryos from the 5 ng/µl, 10 ng/µl, and 25 ng/µl electroporation groups, respectively, had wildtype variants (Fig.4A and E). Conversely, there was 0 of 1 embryo carrying only KO alleles from the 5 ng/µl electroporation group, and 1 of 7 and 9 of 18 embryos carried KO alleles without WT alleles from the 10 ng/µl and 25 ng/µl electroporation groups, respectively (Fig.4A and F and Supplementary Tables S5 and S6).

FIG. 4.

Mosaicism analysis of 27 electroporated parthenotes. (A) Bar graphs depicting the frequencies of alleles determined by PacBio sequencing in individual embryos electroporated with different concentrations of CRISPR/Cas9. Samples contained a mixture of wildtype allele (dark blue) and other indels (other colors). (B) Number of mosaic embryos (carrying more than two alleles). Embryo 2 (wildtype) from 10 ng/μl group was excluded. (C) Mutation efficiency determined with PacBio sequencing. (D) Mean number of alleles present in an embryo. (E) Number of embryos carrying wildtype allele. (F) Number of embryos carrying only KO alleles (**P < 0.01; two-tailed t-test). *Unedited embryo 2 in 10 ng/μl group was excluded for all the measurements except for mutation efficiency.

In IVF-derived embryos, all nine embryos from the 5 ng/µl group, 7 of 14 embryos from the 10 ng/µl group, and four of seven embryos from the 25 ng/µl group were mosaic (Fig.5A and B). Mean mutation efficiency rates by PacBio sequencing showed rates of 84.54%, 97.76%, and 95.99% for the 5 ng/µl, 10 ng/µl, and 25 ng/µl groups, respectively (Fig.5C). Mean allele numbers per group indicated that 4.56 ± 0.5 alleles, 3.64 ± 0.96 alleles, and 2.43 ± 0.43 alleles were detected for the 5 ng/µl, 10 ng/µl, and 25 ng/µl CRISPR groups, respectively, and the mean allele numbers for the 25 ng/µl CRISPR electroporated group were significantly lower than 5 ng/µl CRISPR group (Fig.5A and D). A total of 8 of 9 embryos, 6 of 14 embryos, and 1 of 7 embryos from the 5 ng/µl, 10 ng/µl, and 25 ng/µl CRISPR groups had wildtype alleles, respectively (Fig.5A and E). On the other hand, 1 of 9 embryos, 8 of 14 embryos, and 6 of 7embryos from the 5 ng/µl, 10 ng/µl, and 25 ng/µl CRISPR groups had only KO alleles, respectively (Fig.5A and F and Supplementary Tables S7 and S8).

FIG. 5.

Mosaicism analysis of 30 electroporated IVF zygotes. (A) Bar graphs depicting the frequencies of alleles determined by PacBio sequencing in individual embryos electroporated with different concentrations of CRISPR/Cas9. Samples contained a mixture of wildtype allele (dark blue) and other indels (other colors). (B) Number of mosaic embryos carrying more than two variants. (C) Mutation efficiency determined with PacBio sequencing. (D) Mean number of alleles present in an embryo. (E) Number of embryos carrying wildtype allele. (F) Number of embryos carrying only KO alleles (**P < 0.01; two-tailed t-test).

To determine if electroporation with 25 ng/µl porcine NGN3-targeted gRNA, the condition that appeared to be optimal for pig zygotes, resulted in off-target effects, seven potential off-target sites were predicted across six porcine chromosomes (Supplementary Table S10). The potential off-target sites were amplified, barcoded, and sequenced using PacBio SMRT technology for 15 IVF-derived electroporated blastocyst samples. SMRT-processed reads were aligned to seven possible off-target sites with a total of 847,037 reads. Among the 15 electroporated embryos, nine embryos showed no off-target effects at the predicted target sites, while genetic variations were found in six embryos with minimal indel frequencies ranging from 0.02% to 3% of total potential off-target reads per embryo (Fig.6 and Supplementary Tables S11 and S12). All the off-target indels were 1-bp insertions except for embryo 7 that carried an additional 1-bp deletion at off-target 7 site. The untreated control embryo had two off-target indels, both 1-bp insertions (Supplementary Table S12).

FIG. 6.

Evaluation of off-target effects following electroporation with 25 ng/µl (CRISPR): 50 ng/µl (Cas9 protein). Potential off-target regions (OFF1–7) were identified and analyzed with PacBio sequencing. (Top) PacBio read count of indels at off-target sites for individual embryos. (Bottom) Wildtype allele read count for potential off-target regions. Indel types at same target sites (same color bars) were classified. I = insertion; D = deletion; CTL = control blastocyst; BL = blastocyst.

Discussion

Efficient generation of nonmosaic animal embryos using CRISPR/Cas9 technology is important for not only the production of genome-edited founder animals but also for studying human diseases in swine models. Although the use of the CRISPR/Cas9 system has been described across many livestock species,45 most of the reports have used microinjection to deliver CRISPR/Cas9,14,33,46 –54 and a few studies have characterized its application in porcine embryos.22,55 –58 Electroporation of CRISPR/Cas9 in porcine embryos represents a less laborious and more efficient method than single embryo microinjection,25,55 but there has been no report scrutinizing mosaicism and off-target effects caused by CRISPR/Cas9 electroporation using next generation sequencing in pig embryos. Here, we identified a set of parameters that could efficiently deliver CRISPR/Cas9 targeting porcine NGN3 to one cell embryos without compromising further development to the blastocyst stage. Increasing amounts of gRNA and Cas9 protein significantly increased mutation rates in parthenotes and significantly decreased the number of variants in zygotes.

A potential problem when utilizing CRISPR/Cas9 editing in animal embryos is mosaicism, more than two alleles present in an individual; a high frequency of CRISPR-edited embryos has multiple variants. By Sanger sequencing analysis, we were able to identify mutation rates caused by CRISPR/Cas9 electroporation, and we observed many possible mosaic embryos when identifying mono- or bi-allelic mutations. Most studies characterized CRISPR/Cas9-mediated mosaicism by PCR amplicon sequencing of target regions and decomposing the chromatogram data results with the Tracking of Indels by Decomposition (TIDE) bioinformatics program.59 While this approach is rapid and cost-effective, adapting next generation sequencing of the PCR products offers a more precise characterization of the variants present in an embryo and their relative abundance.47 However, NGS approach raises some concerns with PacBio sequencing due to its susceptibility to errors such as indels and single-nucleotide polymorphisms (SNPs). To address these concerns, we adopted a strategy of generating CCS reads for the short sequences of the target amplicons to increase the confidence in the accuracy of the alleles.60 In the present study, we were able to precisely characterize CRISPR/Cas9-modified porcine blastocysts with this approach. Mosaicism data from CRISPR/Cas9-modified blastocysts indicated that approximately 50% of the analyzed embryos were mosaic at the 10 ng/µl and 25 ng/µl CRISPR concentrations following electroporation of either parthenotes or zygotes. One hundred percent of zygote-derived blastocysts were mosaic in the 5 ng/µl CRISPR group with only a single parthenote available for this concentration. Overall mutation efficiency was significantly higher with the highest CRISPR concentration tested in parthenotes, and numerically higher at the higher concentrations following electroporation of zygotes, suggesting that the higher concentration is needed to efficiently generate NGN3 KO embryos. In IVF-produced embryos, electroporation resulted in more than 78% mutation, even at the lowest CRISPR/Cas9 concentration, similar to the mutation rate in the 25 ng/µl:50 ng/µl parthenotes. Compared with CRISPR/Cas9 electroporated parthenotes, NGN3-targeted CRISPR/Cas9 electroporated IVF embryos had numerically higher mutation rates in the PacBio analyses. Parthenotes have higher homozygosity than IVF embryos. DNA repair might be different from naturally occurring DNA repair in zygotes.61 Furthermore, the decreased number of alleles present in each embryo and the decreased proportion of embryos carrying wildtype alleles in electroporated zygotes treated with the 25 ng/µl of porcine CRISPR suggested that a higher concentration of CRISPR/Cas9 during electroporation may reduce mosaicism with fewer alleles in developing porcine embryos. These results also suggest that there is a greater chance of obtaining animals without the wildtype allele of the target gene when using higher CRISPR-Cas9 concentration. Some embryo mutation types identified with Sanger sequencing results did not match NGS data due to different specificity and accuracy of the two analyses. Given that porcine parthenotes are considered an effective tool for confirming and improving the efficiency of CRISPR/Cas9 testing due to their mimicking capability of embryo development,52 it was crucial to assess whether the two different types of embryos, parthenotes, and IVF embryos exhibit similar effects when subjected to electroporation under identical conditions. The two types of embryos differed in mutation efficiencies, number of alleles, and the presence of wildtype alleles. This may be explained by parthenotes having a more resilient DNA repair system than zygotes. Frequent mosaicism in parthenotes was also found in a study with microinjection of CRISPR/Cas9 complex into porcine parthenotes.20 Different timing of DNA replication in the two different types of embryos could also be a factor because porcine parthenotes start their first DNA synthesis 5–6 h after activation,62 while IVF embryos begin DNA replication 12–14 h after insemination.63

Another concern related to CRISPR/Cas9 introduction into embryos is off-target effects. Our off-target data analyzed by PacBio for seven possible loci indicated that there was zero to minimal unintentional mutation frequencies accounted for in 15 porcine embryos electroporated with 25 ng/µl CRISPR and 50 ng/µl Cas9 protein, and all the mutated alleles observed in off-target regions were 1-bp insertions except for one 1-bp deletion found in embryo 7. Interestingly, we identified some 1-bp insertion reads in the control embryo as well. Given that a high level of natural insertion and deletion was identified throughout the whole pig genome in different breeds,64 all the indels detected in our off-target experiments were probably either a PCR error during sample preparation or a naturally occurring polymorphic variation. The small insertions and deletions observed in the embryos in the off-target study may simply be naturally occurring genomic variation since previous reports suggested that the off-target effects of CRISPR/Cas9-mediated mutagenesis were not different from the de novo mutations in rodents,65,66 rhesus monkeys,67 and humans.68,69

Conclusions

In summary, we demonstrated an effective set of electroporation parameters for delivering CRISPR/Cas9 targeting the pig NGN3 gene into zygotes. Zygote electroporation with 25 ng/μl CRISPR gRNA and 50 ng/μl Cas9 protein resulted in the highest mutation rate in blastocysts as determined by both Sanger sequencing and NGS. In addition, zygotes electroporated with 25 ng/μl CRISPR gRNA and 50 ng/μl Cas9 protein had significantly fewer modified alleles (less mosaicism) and fewer embryos carried wildtype alleles compared with the embryos electroporated with lower concentrations of CRISPR/Cas9. Our off-target analysis following targeting of porcine NGN3 gene indicated that there was little to no alteration at the potential sites. Although further studies are needed to optimize the system and to clarify the mosaicism and potential off-target effects caused by CRISPR/Cas9 technology, our results provide insight into understanding CRISPR/Cas9 electroporation in porcine embryos.

Authors’ Contributions: Experimental design (I.P., S.K.K., J.G., and P.J.R.). Embryo production and collection (I.P.). gRNA introduction and single embryo Sanger sequencing analysis (I.P.). Sample preparation for mosaicism and off-targets (S.N.S.). PacBio data analysis (R.M.P. and E.A.M.). Article writing (I.P., T.B., and P.J.R.).

Conflict of Interest: No competing financial interest exists.

Funding Information: This study was supported by the University of California, Davis Chancellor’s Fellowship (P.J.R.).

Supplementary Material

Supplementary Table S1

Supplementary Table S2

Supplementary Table S3

Supplementary Table S4

Supplementary Table S5

Supplementary Table S6

Supplementary Table S7

Supplementary Table S8

Supplementary Table S9

Supplementary Table S10

Supplementary Table S11

Supplementary Table S12

Footnotes

Acknowledgments

Assistance provided by UC Davis Swine facility manager Aaron Prinz with boar semen collection. Portions of this paper were previously published as a thesis by Insung Park. Park I. One-step generation of CRISPR/Cas9-mediated gene-edited porcine zygotes and fetuses. 2021. UC Davis. ProQuest ID: Park_ucdavis_0029D_21059. Merritt ID: Ark:/13030/m5gf7z5p. Available at .

References

Walters

, Prather

RS.

Advancing swine models for human health and diseases. Mo Med. 2013; 110(3): 212–215.

Whyte

, Prather

RS.

Genetic modifications of pigs for medicine and agriculture. Mol Reprod Dev. 2011; 78(10–11): 879–891. DOI: 10.1002/mrd.21333.

Kim

, Whitener

, Peiris

, et al. Molecular and genetic regulation of pig pancreatic islet cell development. Development. 2020; 147(6): DOI: 10.1242/dev.186213.

Phelps

, Koike

, Vaught

, et al. Production of alpha 1,3-galactosyltransferase-deficient pigs. Science. 2003; 299(5605): 411–414. DOI: 10.1126/science.1078942.

Sieren

, Meyerholz

, Wang

X-J

, et al. Development and translational imaging of a TP53 porcine tumorigenesis model. J Clin Invest. 2014; 124(9): 4052–4066. DOI: 10.1172/JCI75447.

Klymiuk

, Blutke

, Graf

, et al. Dystrophin-deficient pigs provide new insights into the hierarchy of physiological derangements of dystrophic muscle. Hum Mol Genet. 2013; 22(21): 4368–4382. DOI: 10.1093/hmg/ddt287.

, Fuchimoto

, Sudo

, et al. Development of human-like advanced coronary plaques in low-density lipoprotein receptor knockout pigs and justification for statin treatment before formation of atherosclerotic plaques. Am J Heart Assoc. 2016; 5(4): e002779. DOI: 10.1161/JAHA.115.002779.

Davis

, Wang

X-J

, Rohret

, et al. Targeted disruption of LDLR causes hypercholesterolemia and atherosclerosis in Yucatan miniature pigs. PLoS One. 2014; 9(4): e93457. DOI: 10.1371/journal.pone.0093457.

Ito

, Sendai

, Yamazaki

, et al. Generation of recombination activating gene-1-deficient neonatal piglets: A model of T and B cell deficient severe combined immune deficiency. PLoS One. 2014; 9(12): e113833. DOI: 10.1371/journal.pone.0113833.

10.

Yang

, Wang

, Shivalila

, et al. One-step generation of mice carrying reporter and conditional alleles by CRISPR/Cas-mediated genome engineering. Cell. 2013; 154(6): 1370–1379. DOI: 10.1016/j.cell.2013.08.022.

11.

Hai

, Teng

, Guo

, et al. One-step generation of knockout pigs by zygote injection of CRISPR/cas system. Cell Res. 2014; 24(3): 372–375. DOI: 10.1038/cr.2014.11.

12.

Whitworth

, Lee

, Benne

, et al. Use of the CRISPR/Cas9 system to produce genetically engineered pigs from in vitro-derived oocytes and embryos. Biol Reprod. 2014; 91(3): 78. DOI: 10.1095/biolreprod.114.121723.

13.

Park

K-E

, Powell

, Sandmaier

SES

, et al. Targeted gene knock-in by CRISPR/cas ribonucleoproteins in porcine zygotes. Sci Rep. 2017; 7: 42458. DOI: 10.1038/srep42458.

14.

Crispo

, Mulet

, Tesson

, et al. Efficient generation of myostatin knock-out sheep using CRISPR/Cas9 technology and microinjection into zygotes. PLoS One. 2015; 10(8): e0136690. DOI: 10.1371/journal.pone.0136690.

15.

Wang

, Yu

, Lei

, et al. Generation of gene-modified goats targeting MSTN and FGF5 via zygote injection of CRISPR/Cas9 system. Sci Rep. 2015; 5: 13878. DOI: 10.1038/srep13878.

16.

Zhou

, Wan

, Guo

, et al. Generation of beta-lactoglobulin knock-out goats using CRISPR/Cas9. PLoS One. 2017; 12(10): e0186056. DOI: 10.1371/journal.pone.0186056.

17.

Yen

S-T

, Zhang

, Deng

, et al. Somatic mosaicism and allele complexity induced by CRISPR/Cas9 RNA injections in mouse zygotes. Dev Biol. 2014; 393(1): 3–9. DOI: 10.1016/j.ydbio.2014.06.017.

18.

Horii

, Hatada

Genome editing of mouse by cytoplasmic injection. Methods Mol Biol. 2017; 1630: 55–66. DOI: 10.1007/978-1-4939-7128-2_5.

19.

Bevacqua

, Fernandez-Martín

, Savy

, et al. Efficient edition of the bovine PRNP prion gene in somatic cells and IVF embryos using the CRISPR/Cas9 system. Theriogenology. 2016; 86(8): 1886–1896.e1. DOI: 10.1016/j.theriogenology.2016.06.010.

20.

Sato

, Koriyama

, Watanabe

, et al. Direct injection of CRISPR/Cas9-related mRNA into cytoplasm of parthenogenetically activated porcine oocytes causes frequent mosaicism for indel mutations. Int J Mol Sci. 2015; 16(8): 17838–17856. DOI: 10.3390/ijms160817838.

21.

Navarro-Serna

, Piñeiro-Silva

, Luongo

, et al. Effect of aphidicolin, a reversible inhibitor of eukaryotic nuclear DNA replication, on the production of genetically modified porcine embryos by CRISPR/Cas9. Int J Mol Sci. 2022; 23(4): 2135. DOI: 10.3390/ijms23042135.

22.

Navarro-Serna

, Hachem

, Canha-Gouveia

, et al. Generation of nonmosaic, two-pore channel 2 biallelic knockout pigs in one generation by CRISPR-Cas9 microinjection before oocyte insemination. CRISPR J. 2021; 4: 132–146. DOI: 10.1089/crispr.2020.0078.

23.

Kosicki

, Tomberg

, Bradley

Repair of double-strand breaks induced by CRISPR-Cas9 leads to large deletions and complex rearrangements. Nat Biotechnol. 2018; 36(8): 765–771. DOI: 10.1038/nbt.4192.

24.

Shin

, Wang

, Lee

, et al. CRISPR/Cas9 targeting events cause complex deletions and insertions at 17 sites in the mouse genome. Nat Commun. 2017; 8: 15464. DOI: 10.1038/ncomms15464.

25.

Tanihara

, Takemoto

, Kitagawa

, et al. Somatic cell reprogramming-free generation of genetically modified pigs. Sci Adv. 2016; 2(9): e1600803. DOI: 10.1126/sciadv.1600803.

26.

Kaneko

, Sakuma

, Yamamoto

, et al. Simple knockout by electroporation of engineered endonucleases into intact rat embryos. Sci Rep. 2014; 4: 6382. DOI: 10.1038/srep06382.

27.

Qin

, Dion

, Kutny

, et al. Efficient CRISPR/Cas9-mediated genome editing in mice by zygote electroporation of nuclease. Genetics. 2015; 200(2): 423–430. DOI: 10.1534/genetics.115.176594.

28.

Chen

, Lee

AY-F

, et al. Highly efficient mouse genome editing by CRISPR ribonucleoprotein electroporation of zygotes. J Biol Chem. 2016; 291(28): 14457–14467. DOI: 10.1074/jbc.M116.733154.

29.

Camargo

LSA

, Owen

, Van Eenennaam

, et al. Efficient one-step knockout by electroporation of ribonucleoproteins into zona-intact bovine embryos. Front Genet. 2020; 11: 1047. DOI: 10.3389/fgene.2020.570069.

30.

Hashimoto

, Takemoto

: Electroporation enables the efficient mRNA delivery into the mouse zygotes and facilitates CRISPR/Cas9-based genome editing. Sci Rep. 2015; 5: 11315. DOI: 10.1038/srep11315.

31.

Tröder

, Ebert

, Butt

, et al. An optimized electroporation approach for efficient CRISPR/Cas9 genome editing in murine zygotes. PLoS One. 2018; 13(5): e0196891. DOI: 10.1371/journal.pone.0196891.

32.

Teixeira

, Py

, Bosc

, et al. Electroporation of mice zygotes with dual guide RNA/Cas9 complexes for simple and efficient cloning-free genome editing. Sci Rep. 2018; 8(1): 474. DOI: 10.1038/s41598-017-18826-5.

33.

Tanihara

, Hirata

, Nguyen

, et al. Generation of a TP53-modified porcine cancer model by CRISPR/Cas9-mediated gene modification in porcine zygotes via electroporation. PLoS One. 2018; 13(10): e0206360. DOI: 10.1371/journal.pone.0206360.

34.

Lin

, Van Eenennaam

AL.

Electroporation-mediated genome editing of livestock zygotes. Front Genet. 2021; 12: 648482.

35.

Hashimoto

, Yamashita

, Takemoto

Electroporation of Cas9 protein/sgRNA into early pronuclear zygotes generates non-mosaic mutants in the mouse. Dev Biol. 2016; 418(1): 1–9. DOI: 10.1016/j.ydbio.2016.07.017.

36.

, Yang

, Yan

, et al. Promoting Cas9 degradation reduces mosaic mutations in non-human primate embryos. Sci Rep. 2017; 7: 42081. DOI: 10.1038/srep42081.

37.

Sato

, Kosuke

, Koriyama

, et al. Timing of CRISPR/Cas9-related mRNA microinjection after activation as an important factor affecting genome editing efficiency in porcine oocytes. Theriogenology. 2018; 108: 29–38. DOI: 10.1016/j.theriogenology.2017.11.030.

38.

Zuo

, Cai

Y-J

, Li

, et al. One-step generation of complete gene knockout mice and monkeys by CRISPR/Cas9-mediated gene editing with multiple sgRNAs. Cell Res. 2017; 27(7): 933–945. DOI: 10.1038/cr.2017.81.

39.

Gradwohl

, Dierich

, LeMeur

, et al. Neurogenin3 is required for the development of the four endocrine cell lineages of the pancreas. Proc Natl Acad Sci USA. 2000; 97(4): 1607–1611. DOI: 10.1073/pnas.97.4.1607.

40.

Yoshioka

, Suzuki

, Onishi

Defined system for invitro production of porcine embryos using a single basic medium. J Reprod Dev. 2008; 54(3): 208–213. DOI: 10.1262/jrd.20001.

41.

Oliveros

, Franch

, Tabas-Madrid

, et al. Breaking-Cas-interactive design of guide RNAs for CRISPR-Cas experiments for ENSEMBL genomes. Nucleic Acids Res. 2016; 44(W1): W267–W271. DOI: 10.1093/nar/gkw407.

42.

Kanwar

, Blanco

, Chen

, et al. PacBio sequencing output increased through uniform and directional fivefold concatenation. Sci Rep. 2021; 11(1): 18065. DOI: 10.1038/s41598-021-96829-z.

43.

Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM. arXiv:13033997 [q-bio]. 2013. DOI: 10.48550/arXiv.1303.3997.

44.

, Handsaker

, Wysoker

, et al. The sequence alignment/map format and SAMtools. Bioinformatics. 2009; 25(16): 2078–2079. DOI: 10.1093/bioinformatics/btp352.

45.

Mclean

, Oback

, Laible

Embryo-mediated genome editing for accelerated genetic improvement of livestock. Front Agr Sci Eng. 2020; 7(2): 148–160. DOI: 10.15302/J-FASE-2019305.

46.

Wei

, Wagner

, Lu

, et al. Efficient introgression of allelic variants by embryo-mediated editing of the bovine genome. Sci Rep. 2015; 5: 11735. DOI: 10.1038/srep11735.

47.

Vilarino

, Rashid

, Suchy

, et al. CRISPR/Cas9 microinjection in oocytes disables pancreas development in sheep. Sci Rep. 2017; 7(1): 17472. DOI: 10.1038/s41598-017-17805-0.

48.

Zhang

, Li

, Wu

, et al. Disruption of the sheep BMPR-IB gene by CRISPR/Cas9 in in vitro-produced embryos. Theriogenology. 2017; 91: 163–172.e2. DOI: 10.1016/j.theriogenology.2016.10.025.

49.

Wang

, Du

, Shen

, et al. Efficient generation of gene-modified pigs via injection of zygote with Cas9/sgRNA. Sci Rep. 2015; 5: 8256. DOI: 10.1038/srep08256.

50.

Lamas-Toranzo

, Galiano-Cogolludo

, Cornudella-Ardiaca

, et al. Strategies to reduce genetic mosaicism following CRISPR-mediated genome edition in bovine embryos. Sci Rep. 2019; 9(1): 14900. DOI: 10.1038/s41598-019-51366-8.

51.

Petersen

, Frenzel

, Lucas-Hahn

, et al. Efficient production of biallelic GGTA1 knockout pigs by cytoplasmic microinjection of CRISPR/Cas9 into zygotes. Xenotransplantation. 2016; 23(5): 338–346. DOI: 10.1111/xen.12258.

52.

Tao

, Yang

, Wang

, et al. Efficient biallelic mutation in porcine parthenotes using a CRISPR-Cas9 system. Biochem Biophys Res Commun. 2016; 476(4): 225–229. DOI: 10.1016/j.bbrc.2016.05.100.

53.

Tanihara

, Hirata

, Nguyen

, et al. Effects of concentration of CRISPR/Cas9 components on genetic mosaicism in cytoplasmic microinjected porcine embryos. J Reprod Dev. 2019; 65(3): 209–214. DOI: 10.1262/jrd.2018-116.

54.

Proudfoot

, Carlson

, Huddart

, et al. Genome edited sheep and cattle. Transgenic Res. 2015; 24(1): 147–153. DOI: 10.1007/s11248-014-9832-x.

55.

, Tanihara

, Wittayarat

, et al. Comparison of the effects of introducing the CRISPR/Cas9 system by microinjection and electroporation into porcine embryos at different stages. BMC Res Notes. 2021; 14(1): 7. DOI: 10.1186/s13104-020-05412-8.

56.

Tanihara

, Hirata

, Thi Nguyen

, et al. Generation of viable PDX1 gene-edited founder pigs as providers of nonmosaics. Mol Reprod Dev. 2020; 87(4): 471–481. DOI: 10.1002/mrd.23335.

57.

Yamashita

, Kogasaka

, Hiradate

, et al. Suppression of mosaic mutation by co-delivery of CRISPR associated protein 9 and three-prime repair exonuclease 2 into porcine zygotes via electroporation. J Reprod Dev. 2020; 66(1): 41–48. DOI: 10.1262/jrd.2019-088.

58.

, Chen

, Cai

, et al. Production of non-mosaic genome edited porcine embryos by injection of CRISPR/Cas9 into germinal vesicle oocytes. J Genet Genomics. 2019; 46(7): 335–342. DOI: 10.1016/j.jgg.2019.07.002.

59.

Brinkman

, Chen

, Amendola

, et al. Easy quantitative assessment of genome editing by sequence trace decomposition. Nucleic Acids Res. 2014; 42(22): e168–e168. DOI: 10.1093/nar/gku936.

60.

Wenger

, Peluso

, Rowell

, et al. Accurate circular consensus long-read sequencing improves variant detection and assembly of a human genome. Nat Biotechnol. 2019; 37(10): 1155–1162. DOI: 10.1038/s41587-019-0217-9.

61.

Somfai

, Ozawa

, Noguchi

, et al. Diploid porcine parthenotes produced by inhibition of first polar body extrusion during invitro maturation of follicular oocytes. Reproduction. 2006; 132(4): 559–570. DOI: 10.1530/rep.1.01216.

62.

Liu

, Lee

, Moor

RM.

DNA synthesis, microtubule and nuclear dynamics in porcine parthenotes. Zygote. 1996; 4(2): 139–144. DOI: 10.1017/s0967199400003014.

63.

Gioia

, Barboni

, Turriani

, et al. The capability of reprogramming the male chromatin after fertilization is dependent on the quality of oocyte maturation. Reproduction. 2005; 130(1): 29–39. DOI: 10.1530/rep.1.00550.

64.

Chen

, D’Alessandro

, Murani

, et al. SINE jumping contributes to large-scale polymorphisms in the pig genomes. Mobile DNA. 2021; 12(1): 17. DOI: 10.1186/s13100-021-00246-y.

65.

Iyer

, Shen

, Zhang

, et al. Off-target mutations are rare in Cas9-modified mice. Nat Methods. 2015; 12(6): 479–479. DOI: 10.1038/nmeth.3408.

66.

Willi

, Smith

, Wang

, et al. Mutation frequency is not increased in CRISPR-Cas9-edited mice. Nat Methods. 2018; 15(10): 756–758. DOI: 10.1038/s41592-018-0148-2.

67.

Luo

, He

, Zhang

, et al. Trio deep-sequencing does not reveal unexpected off-target and on-target mutations in Cas9-edited rhesus monkeys. Nat Commun. 2019; 10(1): 5525. DOI: 10.1038/s41467-019-13481-y.

68.

Veres

, Gosis

, Ding

, et al. Low incidence of off-target mutations in individual CRISPR-Cas9 and TALEN targeted human stem cell clones detected by whole-genome sequencing. Cell Stem Cell. 2014; 15(1): 27–30. DOI: 10.1016/j.stem.2014.04.020.

69.

Yang

, Grishin

, Wang

, et al. Targeted and genome-wide sequencing reveal single nucleotide variations impacting specificity of Cas9 in human stem cells. Nat Commun. 2014; 5(1): 5507. DOI: 10.1038/ncomms6507.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.05 MB