XlANG Guang-ming ,ZHANG Xiu-ling ,XU Chang-jiang ,FAN Zi-yao ,XU Kui ,WANG Nan ,WANG Yue,,CHE Jing-jing,XU Song-song,,MU Yu-lian,Ll Kui,,LlU Zhi-guo

1 State Key Laboratory of Animal Nutrition/Key Laboratory of Animal Genetics Breeding and Reproduction of Ministry of Agriculture and Rural Affairs of China,Institute of Animal Sciences,Chinese Academy of Agricultural Sciences,Beijing 100193,P.R.China

2 School of Life Science and Engineering,Foshan University,Foshan 528231,P.R.China

3 Genome Analysis Laboratory of the Ministry of Agriculture and Rural Affairs,Agricultural Genomics Institute at Shenzhen,Chinese Academy of Agricultural Sciences,Shenzhen 518120,P.R.China

Abstract Efficient and stable expression of foreign genes in cells and transgenic animals is important for gain-of-function studies and the establishment of bioreactors.Safe harbor loci in the animal genome enable consistent overexpression of foreign genes,without side effects.However,relatively few safe harbor loci are available in pigs,a fact which has impeded the development of multi-transgenic pig research.We report a strategy for efficient transgene knock-in in the endogenous collagen type I alpha 1 chain (COL1A1) gene using the clustered regularly interspaced short palindromic repeats/CRISPR-associated protein 9 (CRISPR/Cas9) system.After the knock-in of a 2A peptide-green fluorescence protein (2A-GFP) transgene in the last codon of COL1A1 in multiple porcine cells,including porcine kidney epithelial(PK15),porcine embryonic fibroblast (PEF) and porcine intestinal epithelial (IPI-2I) cells,quantitative PCR (qPCR),Western blotting,RNA-seq and CCK8 assay were performed to assess the safety of COL1A1 locus.The qPCR results showed that the GFP knock-in had no effect (P=0.29,P=0.66 and P=0.20 for PK15,PEF and IPI-2I cells,respectively)on the mRNA expression of COL1A1 gene.Similarly,no significant differences (P=0.64,P=0.48 and P=0.80 for PK15,PEF and IPI-2I cells,respectively) were found between the GFP knock-in and wild type cells by Western blotting.RNA-seq results revealed that the transcriptome of GFP knock-in PEF cells had a significant positive correlation(P＜2.2e–16) with that of the wild type cells,indicating that the GFP knock-in did not alter the global expression of endogenous genes.Furthermore,the CCK8 assay showed that the GFP knock-in events had no adverse effects(P24h=0.31,P48h=0.96,P72h=0.24,P96h=0.17,and P120h=0.38) on cell proliferation of PK15 cells.These results indicate that the COL1A1 locus can be used as a safe harbor for foreign genes knock-in into the pig genome and can be broadly applied to farm animal breeding and biomedical model establishment.

Keywords: COL1A1 gene,safe harbor,knock-in,CRISPR/Cas9,pig

1.lntroduction

Pigs are widely used in agricultural and medical research because of their important agricultural economic value and high similarity in terms of anatomy and physiology to humans.There are also fewer ethical issues in research using pigs than there are in human research(Perleberget al.2018;Yang and Wu 2018;Shinet al.2020).Traditionally,transgenic pigs have been produced by integrating a sequence or gene into the genome in a random manner,a process that often leads to unexpected genetic changes causing adverse effects,including gene silencing,gene disruption,and the generation of unstable phenotypes (Yanget al.2010;Ohtsuka 2014).Precise integration of exogenous genes into a specific safe harbor locus in the pig genome could solve these problems,advancing our understanding of the consequences of genetic modification (Maet al.2018;Zhouet al.2019).However,the paucity of safe and efficient integration sites and systems limits the application of precise integration in the pig genome.

To date,only a few safe harbor loci have been identified in pigs (Ruanet al.2015;Xieet al.2017).In 2014,Liet al.(2014) found that reverse orientation splice acceptorβ-geo 26 (ROSA26) is widely expressed in embryonic and adult pig tissues,and confirmed that exogenous genes inserted at this site could be expressed ubiquitously.Subsequently,Ruanet al.(2015) reported another safe harbor locus,pig Hipp11 (pH11),which allows an exogenous gene to be inserted into the locus with stable and robust expression in cells,embryos,and adult animals.Maet al.(2018)distinguished thePIFS501site,located in actively transcribed intergenic regions,as a new safe harbor locus.Recently,Liet al.(2020) found that the centrosomal protein 112 (CEP112) locus supported exogenous gene expression driven by a tissue-specific promoter,with a targeting efficiency higher than that of theROSA26locus in porcine embryonic fibroblasts (PEFs).Hanet al.(2019)and Xionget al.(2020) found that the housekeeping genes glyceraldehyde-3-phosphate dehydrogenase (GAPDH)and beta-actin (ACTB) could be used as alternative safe harbors.Although these loci have been discovered and verified in pigs,they are still not enough for the need of multiple transgenes site-specific integration in more complex models.Therefore,identifying more safe harbor loci in the pig genome is important in the development of transgenic pig models.Gene knock-in mediated by homologous recombination (HR) allows the precise insertion of foreign fragments,but its efficiency was quite low (Laiet al.2002;Capecchi 2005).The clustered regularly interspaced short palindromic repeats/CRISPR-associated protein 9(CRISPR/Cas9) technology has become the most popular tool for genetic modification because of its high efficiency and low cost.It has been applied to genetic improvements in pigs,such as increasing meat quality and disease resistance (Wanget al.2015,2017;Whitworthet al.2016;Xuet al.2020).It has also been used to precisely integrate foreign genes into animal genomes with high efficiency(Yaoet al.2017;Tranet al.2019;Huanget al.2020;Ranawakageet al.2021),indicating that the combination of the CRISPR/Cas9 system and HR may be a good strategy for site-specific integration.

Genes expressed in most or all tissues are good candidates for universal genome safe harbor loci (Sadelainet al.2012;Maet al.2018).The collagen type I alpha 1 chain (COL1A1) gene,located on porcine chromosome 12.This is a ubiquitous extracellular matrix protein that is widely expressed in many tissues,including the stomach,omentum,skeletal muscle,bone,cornea,dermis,tendon,and cartilage (Karsenty and Park 1995).Previous reports demonstrated that widely expressed housekeeping geneGAPDHandACTBcould be used as alternative safe harbors by CRISPR/Cas9-mediated homologous recombination and 2A peptide (Hanet al.2019;Xionget al.2020).Therefore,we hypothesized that by using 2A peptide and homologous recombination strategy,theCOL1A1locus of pig could also be used as a safe harbor.Here,a single guide RNA (sgRNA) expression vector and a donor vector containing the 2A peptide-green fluorescent protein (2A-GFP) fragment without a promoter,targeting the stop codon of theCOL1A1gene,were designed for the CRISPR/Cas9 system.Then expression levels ofCOL1A1andGFP,the total transcriptome and the proliferation of2A-GFPsite-specific knock-in cell lines were assessed.

2.Materials and methods

2.1.Vector construction

A sgRNA targeting theCOL1A1locus in this experiment was designed using the web-based software CRISPOR(http://crispor.tefor.net/) (Concordet and Haeussler 2018).Two oligonucleotides coding for the sgRNA were synthesized and denatured using incubation at 98°C for 10 min,followed by natural cooling to 20°C.Then introduced into aBbsI-digested pX330 vector (#42230,Addgene,Watertown,MA,USA) and named pX330-COL1A1-sgRNA.All primer pairs and oligonucleotides are listed in Appendix A.

The donor vector was constructed using a pUC57-Kan backbone.The2A-GFPsequences which were mammalian codon-optimized,and their flanking homology arms were synthesized (Genewiz,Suzhou,China) and inserted into the pUC57-Kan vector byEcoRV enzyme ligation,and named pUC57-COL1A1-GFP-KI-donor(Appendix B),and its detail sequences are listed in the Appendix C.The left and right homology arms for HR events at theCOL1A1locus were 900 bp in size.All plasmids were extracted in large quantities using EndoFree Plasmid Midi Kits (Cwbiotech,Beijing,China).

2.2.Cell culture and transfection

Porcine kidney epithelial cell line (PK15) was purchased from ATCC (CCL-33) and maintained in Dulbecco’s modified Eagle’s medium (DMEM) (Gibco,GrandIsland,NY,USA) supplemented with 10% fetal bovine serum(FBS) (Gibco),1% penicillin/streptomycin (Gibco),1% Glutamax (Gibco),1% non-essential amino acids(Gibco),and 1% sodium pyruvate (Gibco).PEF cells were isolated from a single 35-day-old Large White pig fetus by our lab and maintained in DMEM supplemented with 15% fetal bovine serum,1% penicillin/streptomycin,1% non-essential amino acids,1% Glutamax,and 1%sodium pyruvate,and passaged four generations prior to transfection.The porcine intestinal epithelial (IPI-2I)cells,a cell line established from porcine ileum,was a kind gift from Prof.Xiao Shaobo (Huazhong Agricultural University,China) and were cultured in DMEM supplemented with 10% FBS,1% penicillin/streptomycin,and 1% Glutamax.All cell lines were maintained at 37°C and 5% CO2in a humidified incubator (Forma Series 3111,Thermo Fisher Scientific,Waltham,USA).PK15,PEF and IPI-2I cells are mycoplasma-free.

Cells were transferred to a 10-cm dish one day before transfection.When the cells were 70–80%confluent,5 μg of pX330-COL1A1-sgRNA vector and 5 μg of pUC57-COL1A1-GFP-KI-donor vector were cotransfected per 106cells using Basic Primary Fibroblasts Nucleofector Kits (Lonza,Basel,Switzerland) following the manufacturer’s guidelines.The optimal transfection procedures for PK15,PEF,and IPI-2I were T-020,U-023,and S-005,respectively.Subsequently,cells were transferred to six-well plates and cultured at 37°C in an incubator with 5% CO2.The medium was replaced 5–6 h after transfection.

2.3.The T7 endonuclease 1 (T7EN1) detection assay and sequencing

T7EN1 (New England BioLabs,Beverly,MA,USA) assay was performed to detect the editing efficiency ofCOL1A1-sgRNA.At 48 h post-transfection,the genomic DNA of cells was extracted using a TIANapm Genomic DNA Kit(Tiangen,Beijing,China).Then PCR was performed to amplify the targeted region from wild type and sgRNA transfected cells.PCR products were purified using the TIANquick Midi Purification Kit (Tiangen) following the manufacturer’s protocol.After purification,PCR products were melted and reannealed using the temperature program of 95°C for 10 min,95 to 85°C at–1°C s–1,85 to 25°C at–0.25°C s–1,and 20°C for 2 min (Hanet al.2019).Then the products were digested with 0.5 μL of T7EN1 at 37°C for 30 min and analyzed on a 2% agarose gel.ImageJ Software (NIH,Bethesda,MD,USA) was used to quantify the densitometry of digested products.The editing efficiency of sgRNA was calculated using the formula: Indel (%)=100×[1–(1–Fraction cleaved)1/2].

2.4.Fluorescence-activated cell sorting (FACS)

GFP-positive cells were sorted using a BD FACS AriaTM(BD Biosciences,FranklinLakes,NJ,USA) three days post-transfection.Knock-in efficiencies were analyzed using FlowJo Software (TreeStar,Woodburn,OR,USA).Single GFP-positive PK15 cells were sorted into 96-well plates and then cultured in an incubator with 5% CO2at 37°C for 7–9 days.

2.5.Genomic PCR detection

When the clones of PK15 cells were transferred to 48-well plates,some of the cells were used to extract DNA for the identification of knock-in events.The junction PCR programs were as follows: 94°C for 2 min;36 cycles of 98°C for 10 s,60°C for 30 s,68°C for 1 min 30 s,and one cycle of 72°C for 2 min.The junction PCR-positive clones were collected and expanded for subsequent experiments.

2.6.Absolute quantitative real-time PCR

To detect the copy number of integratedGFPin junction PCR-positive clones,an absolute quantitative real-time PCR assay was performed.One of the GFP-positive,junction PCR-positive,andGFPhomozygous PK15 cell clones (PK15-GFP) was collected and expanded for subsequent experiments.

2.7.Off-target assay

Potential off-target sites (OTSs) were predicted using the web-based CRISPOR Software (Concordet and Haeussler 2018).Nine potential OTSs were selected by their cutting frequency determination (CFD) offtarget score.Pairs of primers (Appendix A) were designed to amplify potential OTSs from the PK15-GFP cells.Then the PCR products were sequenced to determine whether off-target mutations had occurred.

2.8.Quantitative PCR (qPCR)

To detect whether theGFPknock-in event at theCOL1A1locus had an effect on the expression of theCOL1A1gene at the transcription level,we conducted qPCR to detect the expression of theCOL1A1gene in the PK15-GFP cells and wild type PK15 cells.qPCR was performed using TB Green?Premix Ex Taq II (TaKaRa,Tokyo,Japan) on an Applied Biosystems Real-Time PCR Detector (QuanStudio 3,Thermo Fisher Scientific)following the manufacturer’s instructions.And the qPCR primers are listed in Appendix A.qPCR was performed as follows: 95°C for 2 min;40 cycles at 95°C for 5 s,60°C for 34 s;95°C for 15 s,60°C for 1 min,and 95°C for 1 s.ACTBwas used as a reference gene.

2.9.Western blotting

To verify the protein expression of GFP and COL1A1,the PK15-GFP cells and wild type PK15 cells were collected and lysed in 10 mL of M-PERTMmammalian protein extraction reagent (Thermo Fisher Scientific)with one PierceTMprotease and phosphatase inhibitor mini-tablet (Thermo Fisher Scientific).Total protein concentrations were quantified using BCA kits (Beyotime,Shanghai,China).Equal weight of the total protein was separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (EpiZyme,Shanghai,China) and transferred to a polyvinylidene difluoride membrane(Millipore,Massachusetts,MA,USA).Membranes were then blocked with 5% nonfat milk (BD,FranklinLakes,NJ,USA) for 2 h and incubated with COL1A1 (Absin,Shanghai,China) and GFP (Proteintech,Rosemont,IL,USA) primary antibodies at dilutions of 1:1 000 and 1:2 000,respectively,overnight at 4°C.After washing with TBST (EpiZyme,Shanghai,China),anti-rabbit IgG,HRP-linked antibody (Cell Signaling Technology,Boston,MA,USA),and anti-mouse IgG,HRP-linked antibody (Cell Signaling Technology,Boston,MA,USA) were used as the secondary antibody.Chemiluminescence imaging was performed using a Tanon Instrument (YPHBio,Beijing,China).

2.10.CCK8 assay of cell proliferation

To investigate the effect ofGFPknock-in events at theCOL1A1locus,the PK15-GFP cells and wild type PK15 cells were used to evaluate cell proliferation.PK15 cells were seeded at 4×103cells/well in 96-well plates.After incubation for 24,48,72,96,or 120 h,the number of cells was determined using a CCK8 Proliferation Assay Kit (Dojindo,Kumamoto,Japan).For each well,PK15 cells were mixed with 10 μL of CCK-8 reagent and incubated for 2 h at 37°C.Then the optical densities of each well were determined at 450 nm using a microplate reader and were used to calculate cell proliferation rates.

2.11.RNA sequencing

Sequencing library construction and RNA-seq were performed at annoroad gene technology.Paired-end(PE) libraries for sequencing were prepared according to the Illumina PE library preparation protocol (Illumina,San Diego,CA,USA).The qualified libraries were sequenced on an Illumina NovaSeq 6000 Sequencing Platform(Illumina) to generate 150 PE reads.Clean data were aligned to theSus scrofareference genome (v.11.1) using HISAT2 v2.1.0.Expression levels of mRNAs in each sample were quantified as FPKM (Fragments Per Kilobase Millon Mapped Reads) using the HTSeq v0.6.0 Package.

2.12.Statistical analysis

Two groups were set up in this study,including wild type and GFP-positive cells.Differences between the two groups (each group contains three biological replicates,each replicate served as an experimental unit) of samples were assessed using unpairedt-tests,assuming unequal variances;P-values less than 0.05 were considered significant.The data were analyzed with GraphPad Prism6.0.0(La Jolla,CA,USA) for Windows.

3.Results

3.1.Generation and characterization of the reporter system

To determine whether the endogenousCOL1A1gene can be used as a safe harbor locus,a site-specific integration system was constructed.One sgRNA targeting the stop codon of theCOL1A1gene was designed and introduced into the CRISPR/Cas9 expression vector pX330 to generate the pX330-COL1A1-sgRNA vector (Fig.1-A and B),and confirmed by Sanger sequencing (Shanghai).To evaluate the editing efficiency ofCOL1A1-sgRNA,the pX330-COL1A1-sgRNA vector was transfected into PEF cells,and then T7EN1 assay was performed.Compared with the wild type cells,theCOL1A1-sgRNA-transfected group showed apparent editing activity (Fig.1-C),with a mutation percentage of 22%.The PCR products were also sequenced,and the sequencing curves showed that multiple peaks were presented after the target site(Fig.1-D),which were consistent with T7EN1 results.

An HR donor vector,pUC57-COL1A1-GFP-KI-donor,carrying a promoter-less2A-GFPsequence flanked by two homology arms,was obtained by inserting the synthesized homology arms and2A-GFPfragment into the pUC57-Kan vector (Fig.1-A and B;Appendix B).When CRISPR/Cas9-mediated site-specific double stands break events occurred,the2A-GFPfragment was inserted in frame with theCOL1A1coding sequenceviahomologous recombination DNA repair pathway.AndCOL1A1andGFPcan be expressed separately,due to the presence of self-cleaving 2A peptides (Fig.1-B).

Fig.1 Schematic overview depicting the targeting strategy for the COL1A1 locus and single guide RNA (sgRNA) activity measurement.A,schematic overview used in this study.Cas9/sgRNA expression vector and donor vector co-transfected into PEF,PK15 and IPI-2I cells,then GFP-positive cells were sorted by FACS for the subsequent experiments.One of the GFPpositive,junction PCR-positive,and GFP homozygous PK15 cell clones (PK15-GFP) was selected for gene expression and cell proliferation analysis.B,targeting strategy.Exons of COL1A1 are shown as gray boxes (E),the black triangle in Exon 51 (E51)represents the sgRNA targeting site.The yellow arrow box represents the COL1A1 promoter.A targeting vector was created corresponding to the cleavage site of Cas9 and carried 900 bp regions of homology to the COL1A1 sequence astride the cleavage site.C,measurement of COL1A1-sgRNA activity by T7EN1 cleavage assay.NC,negative control;M,DL100 marker.D,Sanger sequencing showed the presence of multiple peaks after the targeted site in the sequencing curves that distinguish non-targeted cells (I) from mutants (II).The sgRNA sequence has a shaded background and the PAM sequence is highlighted by a red box.

3.2.ldentification of the HR-mediated GFP knock-in in multiple porcine cell lines

The pX330-COL1A1-sgRNA vector and pUC57-COL1A1-GFP-KI donor vector were co-transfected into PK15,PEF,and IPI-2I,respectively.Green fluorescence was observed in partial cells 48 h after transfection (Fig.2-A).Due to the GFP expression,we could directly assess the knock-in efficiencies using FACS analysis.In the different porcine cell lines,the average knock-in efficiencies for PK15,PEF,and IPI-2I cells were 2.32,4.77,and 1.38%,respectively (Fig.2-B).

Fig.2 Analysis of knock-in efficiencies in different cell lines transfected with donor and sgRNA expression vectors.A,the knock-in efficiencies were visualized by fluorescence microscopy.B,the knock-in efficiencies were measured by FACS.NC,non-transfected cells;REP,replicate.

To characterize the site-specific insertions ofGFPat theCOL1A1locus,we sorted,cultured,and identified single clones of GFP-positive PK15 cells.Green fluorescence could be observed in almost all single clones of GFP-positive PK15 cells (Fig.3-A).Two pairs of primers spanning the homology arms were designed to amplify the junctions at the two ends between the transgene fragment and genomic DNA (Fig.3-B and C).Genomic PCR and sequencing of single clones showed that the2A-GFPfragment was precisely integrated before the stop codon of theCOL1A1gene as the designed HR donor vector (Fig.3-C).To avoid that the2A-GFPfragment was inserted also into other places of the genome,the copy number of2A-GFPin the single clones was quantified by absolute quantitative real-time PCR assay.We found that a homozygous2A-GFPinserted PK15 clone (Appendix D) had the same copy number of2A-GFPas that of theGAPDHgene (Table 1),which indicates that the2A-GFPfragment was specifically inserted into both alleles of theCOL1A1gene and no random integration of2A-GFPfragments occurred in this clone.Therefore,this clone was named PK15-GFP and used for the subsequent experiments.

Fig.3 Validation of knock-in events in single clones of GFP-positive PK15 cells.A,single clones of GFP-positive PK15 cells visualized by fluorescence microscopy.B,junction PCR confirming GFP knock-in of PK15-GFP cells.M,DL2000 marker;S,PK15-GFP cells;N,wild type PK15 cells.C,sequence analysis of junctions between endogenous and exogenous DNA corresponding to homologous recombination (HR) events.P1/P2 and P3/P4 primers that are used to amplify specific regions for the left-and right-homology arm junctions are indicated by black arrows.Primary nucleotide sequence data corresponding to the junctions between the homology arms and exogenous 2A-GFP sequences,and between the homology arms and endogenous COL1A1 sequences are indicated by red lines.

Then,off-target analysis was performed on the PK15-GFP cells.Potential OTSs were predicted using the online website tool CRISPOR (Concordet and Haeussler 2018).Nine predicted OTSs with fewer than four mismatches with theCOL1A1-sgRNA were selected for further analysis.Genomic PCR and sanger sequencing showed no mutations in these potential OTSs (Table 2).

Table 1 The detection of GFP copies in PK15-GFP cells

Table 2 Analysis of potential off-target sites (OTS)1)

3.3.ldentification of the expression of the COL1A1 gene

The strategy of this study was to integrate the2A-GFPfragment precisely before the stop codon of theCOL1A1gene,so thatGFPcould be expressed in frame with theCOL1A1CDS under the control of theCOL1A1promoter without affecting or disrupting the expression of theCOL1A1gene.Therefore,we measured the expression of theCOL1A1gene at the transcriptional and translation level in the PK15-GFP cells,which is aGFPhomozygous PK15 cell clone,and wild type PK15 cells (PK15-NC)were used as the control group.The results of qPCR and Western blotting showed that no significant differences(P=0.29 for mRNA level,andP=0.64 for protein level)were found between the PK15-GFP and PK15-NC groups(Fig.4;Appendix E).In addition,we also detected the expression ofCOL1A1gene in PEF and IPI-2I cells.Data from qPCR presented no significant differences (P=0.66 andP=0.20 for PEF and IPI-2I cells,respectively) for the mRNA expression ofCOL1A1gene (Appendix F).Similarly,no significant differences (P=0.48 andP=0.80 for PEF and IPI-2I cells,respectively) were found between theGFPknock-in and wild type cells by Western blotting(Appendix F).

Fig.4 Expression of COL1A1 gene in PK15-NC and PK15-GFP.A,transcription level of COL1A1 gene in PK15-NC and PK15-GFP.Different passages of PK15-GFP cells were used for total RNA extraction and cDNA synthesis.ACTB was used as the reference gene for qPCR assay.PK15-NC,wild type PK15 cells;PK15-GFP,a GFP homozygous PK15 cell clone.B,Western blotting for GFP and COL1A1 protein expressions in PK15-NC and PK15-GFP.C,histogram showing the quantitative calculation of protein expression levels from Western blotting images with ImageJ Software (n=3).All data are presented as the mean±standard error(n=3).P-values less than 0.05 were considered significant.

To further explore whether the insertion of2A-GFPfragment alters the global expression of endogenous genes,the mRNA expression profile of wild type PEF cells (PEF-NC) and GFP-positive PEF cells (PEF-GFP)were analyzed by mRNA sequencing.After strict quality control,more than 20.1 and 19.5 G clean bases were retained from PEF-NC and PEF-GFP cells respectively.Each biological replicate sample retained 44.12 M clean read on average,with a read length of 150 bp.The clean read data had an average Q30 of 92.55% (Appendix G).Between 95.80 and 96.42% of the clean reads could be mapped onto theSus scrofareference genome (v.11.1).The Pearson correlation coefficient revealed that the mRNA expression profile of all the six PEF samples were very similar,as that the Pearson correlation coefficient of any two samples was equal to or higher than 0.98(Fig.5-A).Moreover,the significance test demonstrated that PEF-NC and PEF-GFP groups were significantly and positively correlated (R=0.99,P＜2.2e–16;Fig.5-B).The Human Genome Project recommends that under ideal sampling and test conditions,the Pearson correlation coefficient between samples within the same group should not be less than 0.95 (ENCODE Project Consortium 2012).According to this recommendation,the PEF-NC and PEF-GFP samples can be recognized as biological replicates for each other.Similarly,the FPKM distribution data also demonstrated that the mRNA expression variation was reasonably low (Fig.5-C) and these data were consistent with the qPCR results.

Fig.5 Global mRNA expression profile between PEF-GFP and PEF-NC cells.A,Pearson correlation coefficient among PEF-GFP and PEF-NC samples.B,correlation of PEF-GFP and PEF-NC groups (each group contains three biological replicates).C,FPKM distribution among PEF-GFP and PEF-NC cells.PEF-NC,wild type PEF cells;PEF-GFP,GFP-positive PEF cells;R,Pearson correlation coefficient.P-values less than 0.05 were considered significant.

Finally,we measured the proliferation of the PK15-GFP and wild type PK15 cells using CCK8 assay.The results showed that the growth rate of PK15-GFP group is similar (P24h=0.31,P48h=0.96,P72h=0.24,P96h=0.17,andP120h=0.38) to that of wild type PK15 cells (Fig.6).Taken together,these results demonstrated thatCOL1A1could be used as an alternative safe harbor locus for integrating exogenous genes into the pig genome.

Fig.6 Growth curves of PK15-NC and PK15-GFP by CCK8 assays.PK15-NC,wild type PK15 cells;PK15-GFP,a GFP homozygous PK15 cell clone.All data are presented as the mean±standard error (n=3).

4.Discussion

Efficient,stable transgene expression in cells is important for the construction of animal models and the application of transgenic products (Wakasaet al.2009;Maet al.2018).However,only a few safe harbor loci have been found in the pig genome,and some identified safe harbors,likepH11,PIFS501,andCEP112,were located in intergenic regions or intron regions.At these loci,exogenous promoters and regulatory elements were needed (Liet al.2020) to achieve high-level expression.This is laborious and may introduce uncertainty (Liet al.2014;Ruanet al.2015).Therefore,exploring and identifying more safe harbor loci is important for the development of site-specific integration technology and the application of transgenic pigs.Type I collagen,encoded by theCOL1A1gene,strengthens and supports many tissues in the body.It is widely expressed in different tissues and is closely related to cell growth and differentiation,hyperplasia,the repair of tissue damage,inflammatory reactions,sclerosis,and fibrosis(Bou-Gharioset al.2004;Shintaniet al.2006;Jia and Wang 2020).COL1A1locus is well studied and highly conserved in several species and supports a high-level expression of exogenous transgenes,even in cells that do not normally express theCOL1A1gene (Beardet al.2006).Previous studies have shown that theCOL1A1locus is safe in mouse and sheep (Mccreathet al.2000;Beardet al.2006),however,whether theCOL1A1locus could act as a safe harbor in the pig genome still needs to be studied.

Our hypothesis that by using 2A peptide and homologous recombination strategy,theCOL1A1locus of pig could also be used as a safe harbor has been supported by the results of the present study.In this study,the HR-mediated integration system worked well in the pig genome,which was illustrated by green fluorescence detected in PK15,PEF,and IPI-2I cells after sgRNA and donor vector co-transfection.Results of genomic PCR and absolute quantitative real-time PCR assay indicated that homozygous knock-in clones without random integration could be generated by our method.Most importantly,assays of qPCR,Western blotting and RNA-seq showed that theGFPknock-in at theCOL1A1locus of pig neither affect the expression ofCOL1A1itself nor alter the global expression of other endogenous genes.Furthermore,the CCK8 assay showed that theGFPknock-in events had no adverse effects on cell proliferation.Taken together,these results suggested that theCOL1A1locus of pig could be used as a valuable safe harbor in which to precisely integrate exogenous genes into the pig genome.This finding broadens the use ofCOL1A1locus and offers a new safe harbor for multitransgenic pig research.

Moreover,we found that the knock-in efficiencies of PK15,PEF,and IPI-2I cells were 2.32,4.77,and 1.38%,respectively,which is within the normal scope of HDR efficiency (0–15%) mediated by CRISPR/Cas9 system without drug selection (Liet al.2014,2020;Ruanet al.2015;Maet al.2018,2020).We also noticed that although the same sgRNA/Cas9 expressing vector,donor vector,and transfection method were used in all three cell lines,the knock-in efficiency differed from cell lines.We suppose that the differences in knock-in efficiency were mainly caused by the types of cells.Cell types can affect the transfection efficiency,sgRNA targeting efficiency and HR mechanism,and all these factors are responsible for the difference in knock-in efficiency (Maet al.2018;Liet al.2020).The different knock-in efficiencies among PEF,PK15 and IPI-2I cell lines also suggested that the HR-mediated integration system used in our study worked well in the pig genome.

According to the gene expression data of Expression Atlas Database (https://www.ebi.ac.uk/gxa/home),COL1A1is widely expressed in different tissues of pigs,and is highly expressed in the corpus callosum,stomach,epididymis,omentum,mesenteric lymph node,and skeletal muscle.COL1A1is also highly expressed in PEF cells,and PEF cells were widely used in transgenic and gene-editing research of pigs and generating geneediting pigs by somatic cell nuclear transfer.Therefore,exogenous genes integrated at theCOL1A1locus may achieve high-level expression and facilitate gene function research and exogenous protein expression in gene-editing PEF cells or pigs.Because of their high similarity in terms of anatomy and physiology to humans,pig models are often favored over rodent models.Sitespecific knock-in at theCOL1A1locus could be valuable for generating pigs models of human disease and shed light on clinical research.In a word,COL1A1locus has great potential for generating gene-editing pigs in the application of agriculture and medicine.

5.Conclusion

We constructed a knock-in system that permits the normal expression of exogenousGFPand the endogenousCOL1A1gene using the CRISPR/Cas9.TheCOL1A1locus appears to be valuable as a safe harbor to precisely integrate exogenous genes into the pig genome.These results are promising for agricultural breeding and biomedical model applications.

Acknowledgements

This work was financially supported by the Major Scientific Research Tasks for Scientific and Technological Innovation Projects of the Chinese Academy of Agricultural Sciences(CAAS-ZDRW202006),the National Transgenic Breeding Project (2018ZX08010-10B) and the Agricultural Science and Technology Innovation Program of Chinese Academy of Agricultural Sciences (ASTIP-IAS05).

Declaration of competing interest

The authors declare that they have no conflict of interest.

Appendicesassociated with this paper are available on http://www.ChinaAgriSci.com/V2/En/appendix.htm