Efficient generation of B2m-null pigs via injection of zygote with TALENs

Donor major histocompatibility complex class I (MHC I) molecules are the main targets of the host immune response after organ allotransplantation. Whether and how MHC I-deficiency of pig donor tissues affects rejection after xenotransplantation has not been assessed. Beta2-microglobulin (B2M) is indispensable for the assembly of MHC I receptors and therefore provides an effective target to disrupt cell surface MHC I expression. Here, we report the one-step generation of mutant pigs with targeted disruptions in B2m by injection of porcine zygotes with B2m exon 2-specific TALENs. After germline transmission of mutant B2m alleles, we obtained F1 pigs with biallelic B2m frameshift mutations. F1 pigs lacked detectable B2M expression in tissues derived from the three germ layers, and their lymphocytes were devoid of MHC I surface receptors. Skin grafts from B2M deficient pigs exhibited remarkably prolonged survival on xenogeneic wounds compared to tissues of non-mutant littermates. Mutant founder pigs with bi-allelic disruption in B2m and B2M deficient F1 offspring did not display visible abnormalities, suggesting that pigs are tolerant to B2M deficiency. In summary, we show the efficient generation of pigs with germline mutations in B2m, and demonstrate a beneficial effect of donor MHC I-deficiency on xenotransplantation.

expression patterns, and their orthology or functional homology to human non-classical MHC I antigens, such as HLA-E, HLA-F and HLA-G, remains to be established 13 .
Human CD8+ positive cytolytic cell subpopulations as well as NK cell subpopulations have been shown to directly recognize SLA-1a molecules, leading to the lysis of target cells 14,15 . However, the relevance of porcine MHC I molecules in xenotransplantion is not fully understood and needs to be further explored by transplantation experiments with MHC I-deficient donor pig tissues. Porcine MHC I is a heterotrimeric complex consisting of a heavy α -chain, a light β -chain termed β 2-microglobulin (B2M), and short peptides. The α -chain is highly polymorphic and is encoded by several genes that exist inmultiple allelic variants. In contrast, the B2M molecule, which is indispensable to MHC I assembly on the cell surface, is non-polymorphic and encoded by a single gene. Therefore, B2m provides a simple and effective target to disrupt porcine MHC I expression on the cell surface.
Transcription activator-like effector nucleases (TALENs) are versatile genomic editing tools that have been successfully used in different species, including pigs [16][17][18][19][20][21][22][23] . More recently, clustered regularly interspersed short palindromic repeats (CRISPR)/CRISPR-associated 9 systems (CRISPR/Cas9) have been developed that can mediate effective genome editing in a range of species [24][25][26][27][28] . CRISPR/Cas9 systems can be readily modified and synthesized and allow for multiplex editing, whereas TALENs offer higher sequence specificity and are associated with a lower frequency of off-target effects 29 . Here, we used TALEN technology to obtain a targeted disruption of pig B2m. Pigs harboring frameshift mutations in B2m were efficiently generated via cytoplasmic injection of zygotes with TALEN mRNA. Following germline transmission of mutant alleles, we obtained B2m null-mutant pigs, which were devoid of B2M in tissues derived from all three germ layers and lacked MHC I assembly on the cell surface of lymphocytes. These MHC I-deficient pigs can be used as donors to investigate the biological effects on xenotransplantation, and also provide a non-rodent model to better explore how MHC I receptors regulate immunity in various species.

Results and Discussion
Design and validation of TALENs targeting B2m. To disrupt B2M expression in pig, we designed a pair of TALEN molecules that target exon2 of pig B2m (Fig. 1a). These B2m-TALENs did not have any predicted off-target sites (OTs) in the pig genome according to UCSC In-Silico PCR software (http://genome.ucsc.edu/ cgi-bin/hgPcr). To assess B2m-TALEN-affected early stage development in pig embryos, we injected mRNAs encoding the TALEN pair (10 ng/mL each) into the cytoplasm of parthenogenetically activated (PA) pig embryos at the one-cell stage, followed by culture of embryos in vitro to the blastocyst stage. Injected PA embryos exhibited similar rates of cleavage and development to the blastocyst stage as control (untreated) PA embryos ( Table 1). The slightly higher rate of blastocyst formation in microinjected embryos may be related to an increased influx of Ca 2+ during micromanipulation, and is consistent with our previous observations of preimplantation development in Cas9/sgRNA-injected pig PA embryos 28 .
To assess for TALEN activity at the B2m target locus, we isolated genomic DNA from individual embryos and employed the T7 endonuclease I (T7EN1) cleavage assay that detects heteroduplex DNA in PCR products amplified across the target site. T7EN1 cleavage bands were identified in 5 of 6 PA embryos (Fig. 1b,c). Sequencing of PCR products from the 5 embryos with T7EN1 cleavage bands identified indels and overlapping peaks in sequencing chromatographs (Fig. 1d,e; Table 2), indicating that the B2m-TALENs cleaved the target site. These data demonstrate that TALENs targeting pig B2m had no adverse effects on pig embryo development in vitro, and exhibited target-specific nuclease activity.
One-step generation of B2m knockout pigs via injection of zygotes with TALEN mRNAs. We chose to pursue the B2m knockout pig model in the Chinese Bama minipig breed. The Bama minipig is widely used as a large animal model for biomedical research, has a relatively high degree of genetic stability, and is relatively higher inbred. A total of 118 fertilized 1-cell stage embryos were recovered from the oviducts of mated sows, followed by cytoplasmic microinjection of B2m TALEN mRNAs (10 ng/mL each) and transfer of injected zygotes into eight estrus-synchronized recipient sows. Three recipients became pregnant, and one aborted. The remaining 2 recipient sows delivered 7 full-term piglets (Fig. 2a, Table 3). We isolated genomic DNA from ear punch tissue from the 7 piglets and amplified the genomic region surrounding the B2m TALEN target site by PCR. In all samples, a single band was obtained (Fig. 2b), and T7EN1 cleavage bands were detected in PCR samples of 6 of the 7 animals (Fig. 2c). Sequencing of the PCR products revealed indel mutations in the individuals with T7EN1 cleavage bands as expected (Fig. 2d, Fig. S1). Of the 6 mutant founder pigs, bi-allelic frameshift mutations were detected in 1 founder (#T5), mono-allelic null mutations in 2 founders (#T4 and #T6), and mosaicism with at least 3 different genotypes in 3 founders (#T1, #T2 and #T3) (Fig. 2e, Table 4). Interestingly, no in-frame mutation was detected in any of the founders (Fig. 2e, Fig. S1 and Table 4), suggesting a potential bias towards double strand break (DSB)-induced mutations at this site.
Germline transmission of mutant B2m alleles. The two founder pigs (#T1 and #T5, Table 4), in which frameshift, but no other types of mutations were detected and affected both alleles, indicated by the absence of a wild-type allele, did not exhibit any obvious phenotypic abnormalities and grew normally into adults, suggesting that the lack of a B2M molecule had no apparent adverse effects on the development or health of these animals. However, gene-modified founder animals produced by the injection of customized engineered endonucleases into zygotes often exhibit mosaicism for the targeted mutations 20,23,25,28 . Therefore, genetically unmodified cells may be present in various tissues of the founders and functionally compensate for the B2m null mutant cells, masking potential phenotypic manifestations. To address whether pigs were truly tolerant of B2M-deficiency and whether the B2m mutations could be transmitted through the germline, we mated the female founder pig #T1, harboring three different frameshift mutations and no wild type target sequence according to ear tissue analysis, with the male founder #T6, which should produce offspring with bi-allelic B2m-null mutations. We obtained 4 piglets in the F1 generation (F1 piglets) and performed genetic analysis as before. No additional bands were detectable in PCR reactions amplifying across the B2mt arget area, and T7EN1 cleavage bands were detected  (Fig. 3a,b). Sequencing of the PCR products revealed that of the 4 F1 piglets, 1 (#P2) was bi-allelically mutant with two frameshift mutations, 2 (#P214 and #P216) were mono-allelically mutant with one frameshift mutation and the remaining piglet (#P215) did not carry a B2m mutation (Fig. 3c). The F1 piglet with bi-allelic frameshift B2m mutations (#P2) had a normal appearance and developed into a healthy adult. Of the 5 mutations that were detected in total between the 2 founder animals (#T1 and #T6), only 1 (− 1 bp (A)) was present in the F1 offspring. Lack of germline transmission of the other mutations may either be due to chimerism in the germline, or that the litter size of 4 piglets was not large enough for detection of those mutations to be transmitted through the germline. Furthermore, two novel B2m mutations (− 7 bp and −2 bp (CT)) were detected in the F1 piglet #P214, #P216 and #P2, whereas piglet #P215 lacked targeted B2m mutations. The presence of novel mutation(s) or absence of known mutation(s) of founders in individual offspringis were likely due to the high genetic mosaicism in the founder animals. As a result, different genotypes may exist in different tissues, and therefore a fraction of gametes in the founders may contain novel mutant allele(s) or the wild type allele, which were not detected in the ear tissues used for genetic screening of founders. Taken together, these data demonstrate germline transmission of the null mutations introduced into pig B2m via injection of zygotes with TALENs and suggest that pigs tolerate the lack of B2M.
Phenotypic analysis of B2m-null pigs. Germline transmission of mutant B2m alleles provided an opportunity to verify absence of the B2M protein in various tissues and to evaluate the functional consequences of B2m null mutations on MHC I assembly and immune rejection after xenotransplantation. Using a monoclonal antibody specific for pig B2M protein, we performed immunohistochemical staining to investigate B2M expression in skin, liver, and intestine, which are derived from different germ layers. As shown in Fig. 4, B2M was not detectable in tissues from the F1 pig (#P2) with bi-allelic B2m mutations, but detected in wild-type control tissues, indicating that the frameshift B2m mutations effectively disrupted B2M protein expression. To further investigate whether the lack of B2M expression eliminated or reduced pig MHC I assembly on the cell surface, we collected peripheral lymphocytes from F1 animals and performed fluorescence-activated cell sorting (FACS) analysis of samples stained either using a FITC-conjugated mouse monoclonal antibody specific for pig MHC I (SLA-1) or, as a negative control, with a FITC-conjugated mouse IgG isotype antibody without known binding specificity. As shown in Fig. 5, the SLA-1 histogram of the sample from the F1 pig with bi-allelic null mutations of B2m completely overlapped with the isotype control peak, whereas distinct peaks were present in samples from a wild-type pig and a mutant pig with mono-allelic B2m deletions. These data indicate that lymphocytes from the mutant pig with biallelic B2m frameshift mutations were devoid of MHC I complexes on the cell surface. Interestingly, the SLA-1 peak of the sample from the pig with mono-allelic B2m deletions was shifted in intensity towards the isotype control peak, suggesting that haploid expression of B2M also reduced MHC I complex assembly. Thus, availability of B2M may affect pig MHC I assembly on the cell surface in a dose-dependent manner, similar to findings in B2m knockout mouse lymphocytes 30 . The above data demonstrated that the frameshift mutations caused by B2m-TALENs effectively disrupted B2M expression, and eliminated pig MHC I assembly on cell surface.
To assess the functional consequences of the absence of MHC I complex in pig tissues on xenotransplantation, we performed skin grafting from pigs to mice as described previously 31,32 . For improved graft survival, grafted skin pieces were covered with a dressing of several layers of gauze pieces and secured with a transparent film dressing (Fig. 6a). Skin grafts from the F1 bi-allelically mutant pig (#P2) exhibited remarkably prolonged survival compared to those from the littermate F1 pig lacking B2m mutation (#P214) (Fig. 6b,c and d). On day 6 post grafting, the majority of grafts without B2m mutation exhibited necrosis, and grafts were almost completely rejected on day 10 post grafting, demonstrating that acute immune rejection against grafts occurred (Fig. 6c). In contrast, the majority of skin grafts with bi-allelic B2m mutations exhibited no detectable necrosis on day 10 post grafting (Fig. 6b). These data suggest that donor MHC I complex deficiency may be beneficial for xenotransplantation.  However, compatibility may require the additional elimination of the genes encoding the alpha heavy chain of pig MHC I, i.e. pig classical MHC I molecules (SLA-1, 2, 3), to completely prevent intracellular MHC I expression.
While this manuscript was in preparation, pigs with disruption of the classical MHC I molecules (SLA-1, 2 and 3) were reported 33 . However, the presence of a porcine B2m gene in these pigs would hamper any attempted humanization of MHC I, because endogenous porcine B2M could combine with human MHC I heavy alpha chains in hybrid complexes as observed in rodent models 34 . Furthermore, the presence of pig non-classical MHC I molecules would be immunogenic after xenotransplantation. Therefore, the combined disruption of B2M and MHC I alpha heavy chains in pigs will be required to investigate the functional consequences of pig MHC I deficiency on xenotransplantation. Such MHC I-null pigs could be used for MHC I humanization with transgene encoding the human counterparts, especially non-classical MHC I molecules such as HLA-G, HLA-E and HLA-F, which are of limited immunogenicity for the human immune system due to their limited polymorphism and inhibitory function on human NK cell activation. In summary, we demonstrate the efficient generation of B2m-null pigs after germline transmission of mutant alleles that were generated by the injection of zygotes with TALENs targeting exon2 of B2m. We find that the absence of B2M protein prohibits the assembly of MHC I on the cell surface but has no detectable adverse effects on pig development and health. Furthermore, our study also showed a possible beneficial effect of donor MHC I complex deficiency on xenotransplantation success.

Materials and Methods
Animals. The  Vector constructs and in vitro transcription. B2m-TALENs were designed using TALEN-NT software and constructed by Golden Gate methods as previously described 35 . pCS2-PEAS and pCS2-PERR were utilized as upstream and downstream TALEN-assembling backbones, respectively, as described 36 . The CDS region of the TALEN vectors were cut out using SpeI and NheI, and subcloned into pcDNA3.1 vector to be under the drive of T7 promoter. The TALEN-expressing plasmid was linearized using PmeI, of which the cutting site was located at the 3′ -end of the FokI domain. B2m-TALEN mRNA was prepared via in vitro transcription using the linearized plasmid as the templates with T7 U1tra Kit (Ambion, Austin, TX), and further purified by RNeasy Mini Kit (Qiagen).

TALEN efficacy test via pig parthenogenetic embryo injection. The efficacy of B2m
TALENs was tested in pig parthenogenetically activated (PA) embryos. To prepare pig PA embryos, the cumulus-oocyte complexes (COCs) were collected from slaughter house and cultured for in vitro maturation as described 10    mRNAs of the designed TALEN pair were mixed and diluted to be the final concentration of 10 ng/uL each using RNase-free deionized water. The pig oocytes were freed of cumulus, and the matured oocytes with extruded polar body were selected out and subjected to cytoplasmic microinjection with diluted TALEN mRNAs as described 28,37 . The injected oocytes were activated by direct current electrical pulses (1.2 KV/cm, 30 μ s, two times, 1 sec interval) and the activated oocytes (PA embryos) were cultured in PZM-3 media as described by Wang et al. 37 . The cleavage rate of PA embryos was counted at 48 h post activation and blastocystes were harvested at 144 h post activation. Pig genomic DNA was extracted from individualPA blastocysts by incubating individual embryos in lysis buffer as described 28,37 . Using the genomic DNAs as templates, a primer pair set (B2m-TAL-F: 5-CGGTGAAATCCTCTGGCG-3; B2m-TAL-R2: 5-GCCTGTGCTTCCCTGAGACT-3; product size: 658bp) were used to amplify modified B2m alleles in injected embryos by PCR, and the amplification products were subjected to T7 endonuclease 1 (T7EN1, NEB) cleavage assay or Sanger sequencing after purification using gel extraction kit (Qiagen).
Production of B2m knock-out pigs via zygote injection with TALENs. The TALEN mRNAs were mixed and diluted as described above. Pig zygotes were surgically collected from the oviducts of mated sows. The collected zygotes were subjected to cytoplasmic microinjection with the diluted TALEN mRNA mixture in the same way as that for parthenogenetic embryos as described above. Shortly after injection, the injected zygotes were transferred into estrus-synchronized foster mother sows as described 28,37 . Pregnancy was investigated by observing the oestrus behaviors of recipient sows at every ovation circle.
T7EN1 cleavage assay and sequencing. Pig tissue samples were digested in lysis buffer (0.4 M NaCl, 2 mM EDTA, 1% SDS, 10 mMTris-HCl, and100 mg/ml Proteinase K) overnight. Genomic DNAs were extracted from the lystes after treatment with phenol-chloroform and recovered via alcohol precipitation. For performing T7EN1 cleavage assay, the DNA fragment covering the target site was amplified from thegenomic DNAs using PrimerSTAR HS DNA polymerase (TaKaRa, DR010A)with another primer pair set ssB2m-TAL-F/R, of which the sequences were: 5-GGAAGCTCATTTGGCCTGAAGGG-3 (forward) and 5-CTCTCAGAAGGTGCTACTAGACG-3 (reverse), respectively, and the product size was 565bp. After purification with a PCR cleanup kit (Axygen, AP-PCR-50), the purified PCR product was denatured and re-annealed in NEBuffer 2 (NEB)using a thermocycler. Thereafter, the re-annealed PCR products were digested with T7EN1 (NEB, M0302L) for 30 min at 37 °C and separated on a 2.5% agarose gel. The PCR products exhibiting additional band(s) after T7EN1 cleavage assay were sub-cloned into T vector (Takara, D103A). For each sample, the no less than 16 colonies were randomly picked up and sequenced using M13F(− 47) primer.
Immunohistochemical assay. The freshly collected tissues were fixed in 4% paraformaldehyde solution overnight, washed with PBS for 3~4 times and stored in 70% alcohol at 4 °C before use. The fixed tissues were embodied in paraffin and 5-μ m sections were conventionally prepared. The sections were subjected to immunohistochemical staining using the monoclonal antibody of murine origin specific for porcine B2M protein (LifeSpan, Cat.: LS-1858) and HRP-conjugated rabbit anti-mouse IgG Fc antibody (PIERCE, USA) as the primary and secondary antibody, respectively. After stained with DAB kit (Zhongshan Biotech, China), the sections were examined and photographed under microscope (Olympus, Japan).
Fluorescence-activated cell sorting (FACS) assay. Pig whole blood sample (~50 uL) was placed into Eppendorf tube containing 1 mL of PBS with 2 mM EDTA, and then subjected to centrifuge at 1200 rpm, room temperature for 5 min. After centrifuge, the blood cells were re-suspended with 5 mL of PBS. 1 mL of ACK solution was added into the sample and mixed thoroughly to lyse the red blood cells. The lysis process of red blood cells was repeated twice, and then the remaining blood cells were stained with 50 uL of FACS buffer (   Fresh peripheral lymphocytes were collected from the piglets with bi-allelic mutation, mono-allelic mutation and no mutation (wild type pig), and stained with FITC-conjugated SLA-1 antibody and FITCconjugated IgG isotype control antibody in parallel. The red histogram is for lymphocytes stained with SLA-1 antibody, and the blue one for IgG isotype control antibody. To wash the stained cells and eliminate the unbound antibodies, the blood cells were re-suspended with 1 mL of FACS buffer and spin at 1400 rpm for 5 min. This washing process was repeated twice, and then the stained blood cells were subjected to flow cytometry analysis.
Skin grafting. Skin grafting was performed as previously described 31,32 . Briefly, Split-thickness skin grafts were made from the full thickness skin pieces surgically collected from pigs after anesthesia. The hair on the back of anesthetized recipient mice of FVBN strain was shaved off, and a 1 × 1 cm 2 wound was made on the back by shaving off the dorsal skin carefully while leaving the subcutaneous vasculature intact. Porcine skin graft with the size of 1 × 1cm 2 was planted into the wound and sutured with the host skin. To fasten the touch of skin graft with the wound, the porcine skin grafts were covered with a dressing of several layers of gauze pieces and secured with a transparent medical film dressing to provide a sustainable pressure. The survival of skin grafts was observed every other day, and the survival time was defined as the first day on which more than 90% area of skin grafts was necrotic as described by Wang et al. 32 .