All animal experiments in this study were performed in strict accordance with the Guide for the Care and Use of Laboratory Animals of the Ministry of Science and Technology of China and were approved by the Institutional Animal Care and Use Committee of China Agricultural University (certificate of the Beijing Laboratory Animal employee, ID: 18086). All efforts were made to minimize animal suffering.

The Rosa26 locus, which produces three non-coding transcripts of unknown function (34, 35), was used as the target site for stable integration of the duplicate 4-1BB gene in pigs. The donor plasmid p4BOCNDR was constructed according to the following strategy. First, a 4-1BB expression cassette containing the coding sequence (CDS) of this gene (GenBank No. JQ663443) and its 5′ and 3′ regulatory elements were constructed. The FLAG-tag sequence was 24 bp in size and was added to the c-terminus of the 4-1BB CDS by PCR amplification. Second, the Oct-4 promoter, which is specifically activated during embryonic development (36), was cloned into a site upstream of the Cre recombinase gene. Third, two loxP sequences with the same orientation were ligated to the 5′ end of the Cre expression cassette and the 3′ end of the neo expression cassette, respectively. Fourth, 2- and 6-kb fragments of the first intron of Rosa26 were amplified and used as left- and right-homologous arms, respectively. Finally, the left-homologous arm, 4-1BB expression cassette, Cre expression cassette (with the 5′ loxP sequence), neo expression cassette (with the 3′ loxP sequence), right-homologous arm and negative selection marker diphtheria toxin A (DTA) expression cassettes were ligated together to produce donor plasmid p4BOCNDR (Figure 1A). p4BOCNDR was validated by restriction enzyme digestion and sequencing.

Peripheral blood mononuclear cell (PBMC) was isolated from a 2-month-old pig. Venous blood was collected in 5 mM anticoagulant EDTA (final concentration) and centrifuged on discontinuous Percoll gradients (Sigma, St. Louis, MO, USA) as previously described (37). PBMCs were >98% viable as confirmed by Trypan blue (Sigma) exclusion. PBMCs were seeded at a density of 1 × 10⁶ cells/mL of RPMI-1640 medium (Gibco, Carlsbad, CA, USA) supplemented with 10% heat-inactivated fetal bovine serum, 20 mM HEPES, 2 mM glutamine, and antibiotics (100 U/mL penicillin and 100 µg/mL streptomycin) at 37°C in a humidified atmosphere containing 5% CO₂.

Peripheral blood mononuclear cells were nucleofected with p4BOCNDR plasmid using Amaxa Nucleofector (Lonza Group AG Basel, Switzerland), following the instruction of the manufacturer. Transfected cells were stimulated using concanavalin A (ConA, 2 µg/mL) and incubated at 37°C with 5% CO₂. Cells were harvested at 24 h post-nucleofection. Indirect immunofluorescence assay (IFA) was performed following the standard protocol. Briefly, permeabilized and blocked cells were incubated with anti-FLAG monoclonal antibody (F1804, Sigma-Aldrich) and then with the goat anti-mouse IgG H&L (Cy3) secondary antibody (ab97035, Abcam, Cambridge, UK). Nuclei of cells were stained with 4′,6-diamidino-2-phenylindole (Sigma-Aldrich), and then slides were observed under a fluorescence microscope (Leica TCS SP5, Leica Microsystems, Germany).

sgRNAs targeting intron 1 of Rosa26 were designed using the CRISPRtool (http://crispr.mit.edu) to minimize off-target effects. The annealed oligonucleotides were inserted into the backbone of plasmid pX330 (CBh::SpCas9 + U6::chimeric guide RNA, 42230, Addgene) to construct pCas9-sgRNA as previously described (38). The cleavage efficiency of Cas9 at the targeting site was tested using a T7 endonuclease I (T7EI) assay. Briefly, porcine fetal fibroblasts (PFFs) were transfected with pCas9-sgRNA. Genomic DNA extracted 48 h post-transfection was subjected to PCR with primers P9 and P10. PCR products were purified and then digested with T7EI and electrophoresed on a native polyacrylamide gel. Gel images were analyzed using ImageJ 1.37 software as previously described (39). PCR products were sequenced to detect mutations within intron 1 of Rosa26.

Primary PFFs were isolated and cultured as described previously (40). Then, 1 × 10⁶ PFFs were co-transfected with 4 µg of linearized p4BOCNDR and 1 µg of pCas9-sgRNA and seeded onto 10 cm² dishes. After 10 days of positive selection with G418 (with the concentrations ranging from 500 to 1,000 µg/mL) and negative selection with DTA, cell clones visible by the naked eyes were isolated with cloning cylinders, trypsinized, and transferred to 24-well plates. Subcloned cells were cultured for another 48 h before transfer to a 6-well plate. One-fourth of the cells from each well of the 24-well plate were collected and used for PCR analysis using two primer pairs (P3/P4 and P5/P6). Cells from each correctly modified clone were cryopreserved in liquid nitrogen and used as donor cells for somatic cell nuclear transfer (SCNT). SCNT was conducted as described previously (41). Briefly, the nucleus of an unfertilized, haploid egg cell was extracted and replaced by a nucleus from a diploid donor cell. The reconstructed zygotes were treated in vitro with electrical fusion and chemical activation. Three hundred reconstructed embryos were transferred into both oviducts of each surrogate on day 1 of the onset of estrus. Pregnancy was determined by abdominal ultrasound examination 1 month after SCNT. Piglets were delivered by natural birth after approximately 114 days of gestation and housed in BSL-2 conditions at Taichang Agriculture and Animal Husbandry Co., Ltd. (Qinhuangdao, Hebei Province, China).

To detect the integration of 4-1BB into the Rosa26 locus, 10 µg of genomic DNA extracted from the tail samples of Tg or wild-type (WT) pigs were digested with NotI overnight and then resolved by 1% agarose gel electrophoresis. DNA samples were transferred from the agarose gel onto a nylon membrane and then hybridized with a digoxigenin (DIG)-labeled probe for amplification with primers P1 and P2 as indicated in Figure 1A. Then, the samples were incubated with anti-DIG AP (Roche), and the location of the probe was detected by chemiluminescent methods.

To detect integration of the 4-1BB expression cassette into the Rosa26 site, genomic DNA and four primer pairs (P1–P8, Table 1) were used to amplify the 4-1BB CDS, 5′ and 3′ homologous arms and neo gene by PCR.

To detect transcription of 4-1BB, the total RNA was extracted from PBMCs of Tg or WT pigs post-stimulation using TRIzol reagent according to manufacturer instructions (Invitrogen, Carlsbad, CA, USA). All RNA samples were treated with DNAse before cDNA synthesis. Total RNA was reverse-transcripted using an EasyScript First-Strand cDNA Synthesis SuperMix (Transgen, Beijing, China). Primers for real-time PCR are shown in Table 2. For PCR amplifications, 1 µL of cDNA was added to a mixture containing 10 µL of 2× SYBR Green Master Mix (Takara, Dalian, China), 0.4 µL of ROX reference dye, and 0.4 µL of each primer (50 pmol/μL). Three transgenic and five wild-type pigs were used to analyze the relative mRNA levels between the two groups with the 2^−ΔΔCT method (42). The mean ΔCT of five wild-type samples was used as the control (set to 1). The ΔΔCT was calculated by comparing the ΔCT of three transgenic pigs (n = 3) with controls. Transcripts of porcine GAPDH were used as an internal reference. All samples were tested in duplicate.

To measure expression levels, PBMCs, stimulated with ConA (2 µg/mL), were cultured for 24 h, lysed with RIPA buffer (10 mM Tris, pH 7.4; 150 mM NaCl; 0.2% Triton X-100; 2 mM EDTA; 1 mM PMSF; and 1× Protease Inhibitor Mixture), and then separated on 15% polyacrylamide gels. Separated proteins were transferred to nitrocellulose membranes (Amersham Pharmacia, UK) that were blocked with 5% skim milk in phosphate-buffered saline Tween 20 (PBST) at 25°C for 1 h, and then detected using anti-FLAG mouse monoclonal antibody (F1804, Sigma-Aldrich). The proteins were then incubated with goat anti-mouse IgG-HRP (ab97023, Abcam) for 1 h at 25°C, followed by washing three times with PBST. The membranes were subjected to luminol-based chemiluminescence detection using a commercial substrate (Millipore, USA) and Kodak film. IFA was also performed on PBMCs as described earlier.

For T cell proliferation analysis, PBMCs from Tg and WT pigs were labeled with carboxyfluorescein diacetate succinimidyl ester (CFSE) dye (BioLegend) as previously described (44). Briefly, cells were incubated with 5 µM CFSE in PBS containing 5% FBS at 37°C for 5 min and washed three times with ice-cold PBS-5% FBS. 2 × 10⁵ CFSE-labeled PBMCs/well of a 96-well plate were cultured in RPMI-1640 media supplemented with 10% heat-inactivated FBS, 1% penicillin/streptomycin in the presence of stimulation. PBMCs were harvested after a 3-day culture for subsequent analysis by the BD FACSVerse™ flow cytometer.

Peripheral blood mononuclear cells were cultured in 24-well round bottom plates (1 × 10⁶ cells/well) and stimulated for 0, 24, 48, and 72 h with ConA (2 µg/mL), recovered by centrifugation and re-suspended in 100 µL of PBS containing 10% porcine serum. After being washed twice with a cell-staining buffer (BD Biosciences), the cells were suspended in 50 µL of staining buffer and stained with mouse anti-pig 4-1BB monoclonal antibody at 4°C for 30 min. After two washes, they were stained with FITC anti-mouse IgG (BioLegend Cat. No. 406001). Then, the cells were washed two times, suspended in 200 µL of sterile PBS, and analyzed using the BD FACSVerse™ flow cytometer.

Both 4-1BB Tg and WT pigs were inoculated intramuscularly with a modified live PRRSV vaccine (JXA1-R, Qilu Animal Health Products Co., Ltd., Jinan, China) at 14 days of age. PBMCs, sampled 2 weeks postvaccination, were used to determine the expression levels of cytokines and a cytolytic mediator. The expression of Th1- (IL-2 and IFN-γ) and Th2-type (IL-4) cytokines were determined by real-time PCR. Furthermore, PBMCs were stimulated in vitro for 48 h with the PRRSV vaccine strain (2 × 10⁵ TCID₅₀, propagated by three rounds of plaque purification in MARC-145 cells) and then the production of Th1- (IL-2, IFN-γ, and TNF-α), Th2- (IL-4), Th17-type (IL-17) cytokines, transcription factor (Foxp3) of regulatory T cell and a cytolytic mediator (granzyme B) were measured.

The concentrations of IL-2, IFN-γ, and IL-4 in serum were measured using commercial enzyme-linked immunosorbent assay (ELISA) kits (BlueGene Biotech, Shanghai, China) at day 14 postvaccination. All ELISA procedures were performed as per the manufacturer’s instructions. Each sample was assayed in triplicate.

Interferon-γ ELISPOT was performed as previously described (45). Briefly, 96 wells of ELISPOT plates were coated with anti-porcine IFN-γ monoclonal antibody (5 µg/mL) (Mabtech, Mariemont, OH, USA). 5 × 10⁵ PBMCs from PRRSV-vaccinated three Tg and three WT pigs in 100 µL RPMI-1640 complete medium were added in each well. Cells were stimulated with PRRSV strain JXA1-R (2 × 10⁵ TCID₅₀) while the wells that received an equal amount of bovine serum albumin-coated beads or 2 µg/mL ConA were used as negative and positive controls, respectively. Cells were cultured for 36 h at 37°C in 5% CO₂. The plates were then washed five times with PBS (200 µL/well). Captured IFN-γ was detected with biotinylated anti-porcine IFN-γ antibody P2C11 and streptavidin–ALP. All procedures were conducted according to the instructions of the manufacturer (Mabtech, Mariemont, OH, USA).

To evaluate the ability of upregulated 4-1BB signaling to enhance the antiviral cellular immune responses, a CTL assay was performed with porcine alveolar macrophages as target cells that were infected with PRRSV (JXA-1R strain, multiplicity of infection = 10) and PBMCs from PRRSV-vaccinated pigs restimulated in vitro with PRRSV for 7 days as a source of PRRSV-specific CTLs. The viability of target cells following treatment was assessed using an MTT Cell Proliferation and Cytotoxicity Assay Kit following the manufacturer’s instruction (Beyotime Biotechnology, China).

All experimental data were analyzed using SPSS software (V. 19.0; SPSS Inc., Chicago, IL, USA). Data were expressed as mean ± SEM. Statistical significance was determined using a two-tailed Student’s unpaired t-test or one-way ANOVA, and P values less than 0.05 were considered significantly different.

The size of vector p4BOCNDR is 20.3 kb (Figure 1A). As expected, digestion with PacI, NotI, and XhoI fragmented the vector into three bands of 12.6, 6.1, and 1.6 kb, respectively, whereas digestion with AgeI produced three fragments of 9.0, 6.6, and 4.7 kb (Figure 1B). Expression of the cDNA copy of 4-1BB was confirmed by staining with monoclonal antibody against FLAG tag, which was fused in-frame to 4-1BB on the surface of transfected PBMCs (Figure 1C).

The sgRNA (5′-ATCTTGACTACCACTGCGAT-3′) targeting the Rosa26 locus with the best rank in CRISPRtool was tested for cutting efficiency. Two cleaved bands were generated after the T7EI assay, indicating that our pCas9-gRNA was efficient in generating mutations at the Rosa26 locus. Out of 20 sequenced clones, 4 had deletion or insertion of random nucleotide(s), confirming efficient cleavage at this site. Interestingly, a nucleotide deletion of up to 31 bp occurred downstream of the target site (data not shown).

As a consequence of G418 selection, an obvious reduction in non-transfected cells and the visible formation of living cell foci were observed (Figure 2A). Cell clones with integration and expression of the 4-1BB gene yielded PCR products of 2.5 kb for the 5′ fragment and 8 kb for the 3′ fragment as expected (Figures 2B,C). PCR products were then sequenced to confirm targeting. In total, 10 out of 48 drug-resistant clones showed the expected recombination post co-transfection with p4BOCNDR and Cas9/gRNA, with the transgene inserted specifically at the Rosa26 locus, whereas 2 out of 19 clones with the single plasmid p4BOCNDR showed the correct recombination. The efficiency of homologous recombination increased strikingly with CRISPR/Cas9-mediated editing.

A correctly targeted PFF cell line obtained by selection was used as a donor cell line for SCNT. One out of five surrogate mothers became pregnant, and three cloned piglets were born 114 days later. PCR results showed that the sizes of amplicons were 0.5 kb with primers P1 and P2, 2.5 kb with primers P3 and P4, 8 kb with primers P5 and P6, and 0.7 kb with primers P7 and P8 (Figures 3A–D). As expected, CRISPR/Cas9-mediated site-specific homology recombination occurred at the Rosa26 locus. DTA, as the counter-selection marker, further improved the odds of successful recombination. The neomycin gene was removed in vivo following the expression of Cre, regulated by the Oct4 promoter, during early embryogenesis. Site-specific insertion was also confirmed by Southern blot analysis, with a single 6-kb DNA band indicating correct targeting (Figure 3G). To verify efficient expression of 4-1BB at the Rosa26 locus in the cloned piglets, IFA and western blot were also performed as described earlier. Expression of 4-1BB-FLAG was detected in PBMCs from all three piglets (Figures 3E,F). The results of real-time PCR showed that the levels of 4-1BB mRNA in Tg pigs were twofold higher than those in WT pigs (P = 0.033, Figure 3H). It should be noted that all Tg piglets developed normally, and they showed no significant histological abnormalities (Figure 3I).

To evaluate the proliferation capacity of 4-1BB transgenic T cells, we performed an in vitro proliferation assay. A CFSE division profile of 4-1BB transgenic T cells during 3 days showed more rapid cell division than WT control, and the percentage of cells in each division cycle compared with non-stimulated control was shown in Figure 4A. This revealed that T cells from 4-1BB transgenic pigs had greater proliferation capacity than those from WT control, consistent with the 4-1BB-deficient mouse phenotype showing impaired T cell proliferation (27). Flow cytometry analysis of 4-1BB expression determined that naïve T cells did not express detectable levels of 4-1BB in both groups. There was a profound increase in 4-1BB expression until 48 h, and by 72 h levels of 4-1BB decreased (Figure 4B).

To assess the role of the 4-1BB co-stimulatory pathway on the immune responses to PRRSV, Th1- (IL-2 and IFN-γ) and Th2-type (IL-4) cytokines were measured. Higher expression levels of IL-2 (P = 0.013) were observed in 4-1BB Tg pigs, relative to those in WT controls (Figure 4C). However, the remarkable increase in IFN-γ production (P = 0.008) and the slight but no significant change in IL-4 production (P = 0.692) in Tg pigs (Figures 4D,E) suggested that the 4-1BB signal enhances the induction of immunity by promoting a Th1-dominant response. There was no significant difference in the titer of anti-PRRSV antibodies between Tg and WT pigs (data not shown).

We next determined whether co-stimulatory signaling through 4-1BB could augment specific cellular immune response to PRRSV. We found that the IL-2 mRNA expression levels in 4-1BB Tg pigs were nearly threefold higher than those in WT pigs (P = 0.016), which indicated that greater T cell expansion occurred in 4-1BB Tg pigs than in WT pigs post in vitro stimulation with PRRSV vaccine (Figure 5A). A significant increase in the expression of IFN-γ (P = 0.009) and TNF-α (P = 0.027), along with a slight decrease in the production of IL-4 (P = 0.652), was observed in 4-1BB Tg pigs (Figures 5B–D). This cytokine secretion profile suggested that 4-1BB co-stimulation enhanced the induction of cellular immune responses to PRRSV. We also found that the cytolytic mediator, granzyme B dramatically upregulated in 4-1BB Tg pigs (P = 0.008, Figure 5E). No significant differences in the levels of Foxp3 (P = 0.082, Figure 5F) and IL-17 (P = 0.623, Figure 5G) between Tg and WT pigs. Compared with the WT control, 4-1BB transgenic PBMCs showed a significant increase in lysis of porcine alveolar macrophages at the highest E:T cell ratio (100:1, P = 0.0025, Figure 5H), with high levels of lysis observed at an E:T cell ratio of 50:1 (P = 0.0046, Figure 5H). Moreover, the production of IFN-γ by PBMCs stimulated with PRRSV was also examined using a standard ELISPOT assay. In accordance with qPCR results, the level of PRRSV-specific IFN-γ secreting cells derived from Tg pig was significantly increased (P = 0.0038, Figures 6A,B). The serum concentrations of Th1 cytokines IL-2 (P = 0.026) and IFN-γ (P = 0.016) in Tg pigs were significantly higher than those of WT control, whereas there is no difference in the level of IL-4, the Th2 cytokine, between both groups (Figure 6C). These data indicate that 4-1BB plays a critical role in CTL-mediated specific anti-PRRSV immune responses.