Generation of Recombination Activating Gene-1-Deficient Neonatal Piglets: A Model of T and B Cell Deficient Severe Combined Immune Deficiency

Although severe combined immune deficiency (SCID) is a very important research model for mice and SCID mice are widely used, there are only few reports describing the SCID pig models. Therefore, additional research in this area is needed. In this study, we describe the generation of Recombination activating gene-1 (Rag-1)-deficient neonatal piglets in Duroc breed using somatic cell nuclear transfer (SCNT) with gene targeting and analysis using fluorescence-activated cell sorting (FACS) and histology. We constructed porcine Rag-1 gene targeting vectors for the Exon 2 region and obtained heterozygous/homozygous Rag-1 knockout cell colonies using SCNT. We generated two Rag-1-deficient neonatal piglets and compared them with wild-type neonatal piglets. FACS analysis showed that Rag-1 disruption causes a lack of Immunoglobulin M-positive B cells and CD3-positive T cells in peripheral blood mononuclear cells. Consistent with FACS analysis, histological analysis revealed structural defects and an absence of mature lymphocytes in the spleen, mesenteric lymph node (MLNs), and thymus in Rag-1-deficient piglets. These results confirm that Rag-1 is necessary for the generation of lymphocytes in pigs, and Rag-1-deficient piglets exhibit a T and B cell deficient SCID (T-B-SCID) phenotype similar to that of rodents and humans. The T-B-SCID pigs with Rag-1 deficiency generated in this study could be a suitably versatile model for laboratory, translational, and biomedical research, including the development of a humanized model and assessment of pluripotent stem cells.


Introduction
The V(D)J recombination generates antibody and T cell receptor diversity [1,2]. Recombination activating gene-1 and -2 (Rag-1, -2) were identified as activators of V(D)J recombination in NIH 3T3 fibroblasts grown on an artificial substrate [3]. Cotransfection of Rag-1 and Rag-2 into fibroblasts synergistically activates V(D)J recombination [4], and both Rag-1 and Rag-2 proteins are sufficient to perform V(D)J recombination, which requires DNA nicking (double-strand breaks) and hairpin formation at an early stage [5]. Moreover, it is established that Rag-1 and Rag-2 play a crucial role in lymphoid cell development. Both Rag-1-and Rag-2deficient mice lack mature B and T lymphocytes because of the blockage of lymphocyte differentiation early in development [6,7]. In humans, Rag mutations cause severe combined immunodeficiency (SCID) with a complete absence of both T and B cells (T-B-SCID) by a complete block of T and B cell differentiation and Omenn syndrome and granulomas by impaired V(D)J recombination [8][9][10][11][12].
Because of their anatomic, nutritional, physiologic, and genetic similarities to humans, pigs are essential for biomedical research [26,27]. The cloning of pigs by somatic cell nuclear transfer (SCNT) was established in 2000 [28][29][30]. Subsequently, SCNT combined with gene targeting has enabled the production of genetically modified pigs for use in xenotransplantation as human disease models and in regenerative medicine [31][32][33][34]. Some genetically modified pig models show greater similarities to human diseases than rodent models. For example, pig models of cystic fibrosis, retinitis pigmentosa, Type 2 diabetes, X-linked SCID, and familial adenomatous polyposis models have been established [35][36][37][38][39]. Regarding models of SCID, a T-B+NK-pig model of X-linked SCID was recently generated by targeted disruption of the interleukin 2 receptor gamma chain (Il2rg) gene [38,40]. Moreover, the T-B-NK+ SCID model pigs that had Rag-1/-2 inactivated using gene editing technology have been recently reported in 2014 [41,42]. Although studies about SCID model pigs are beginning to be reported, there are only four reports to date. Therefore, additional research is necessary to establish pig SCID models. Our objective was to generate Rag-1-deficient neonatal piglets of Duroc breed using SCNT with gene targeting, and analyze their peripheral blood mononuclear cells (PBMCs) and histology.

Ethics statement
All animal experiments were approved by the Animal Care Committee of the ZEN-NOH Central Research Institute for Feed and Livestock, Tsukuba, Japan (Permit Number: 20091224). All studies conducted in vivo were performed using sodium pentobarbital or midazolam/medetomidine with a combination of isoflurane and nitrous oxide anesthesia, and all efforts were made to minimize suffering.

Vector construction
Targeting vector construction was performed as previously described [43]. The porcine Rag-1 gene targeting vector was designed using Rag-1 Exon 2 (Accession Number: AB091392, GenBank: AB091392.1) in the National Center for Biotechnology Information database. The 39 short homologous arm for the Rag-1 KO vector was generated from a 1.5 kilo base pair (kb) fragment by polymerase chain reaction (PCR) using the forward primer 59-GCGATGTGAAGTCAG-TGTGC-39 and the reverse primer 59-CCTCATATCTGTACTTGAACTTGG-39 and the pig genomic DNA of Duroc fetal fibroblasts. Likewise, the 59 long homologous arm was generated from a 6.5 kb Rag-1 gene fragment containing Exons 1 and 2 by PCR using the forward primer 59-GCAGATGCAACTC-CAATTCC-39 and the reverse primer 59-CTCAGACGGTGTTTCTGAGC-39. The Rag-1 heterozygous KO vector was constructed by the insertion of two homologous arms into the PGK-Neo/MC1-TK plasmid vector as previously described. The Rag-1 homozygous KO PGK-Neo vector was modified to contain the antibiotic resistance gene CAG-blasticidin resistance gene (bsr).

Preparation of Rag-1 KO cells
The preparation of Rag-1 KO cells was performed as previously described [43]. Original porcine fetal fibroblasts (PFF: T6-12) were isolated from a wild-type Duroc male fetus and cultured in minimum essential medium-a containing 10% fetal calf serum (10% FCS-MEMa; Invitrogen, Carlsbad, CA, USA) at 38.5˚C in 5% CO 2 . Transfection of the Rag-1 heterozygous KO vectors was performed by electroporation. PFFs (1610 7 ) were transfected with 5 mg of the Rag-1 heterozygous KO vector at 220 V and 950 mF using a Gene Pulser apparatus (Bio-Rad Laboratories, Hercules, CA, USA). Transfected cells were cultured in 10% FCS-MEMa in a 6-well plate for 48 h. After incubation, the transfected cells were suspended with 400 mg/ml Geneticin (Invitrogen, Carlsbad, CA, USA) and 20 mM gancyclovir (Nacalai Tesque, Inc., Kyoto, Japan) for positive-negative selection. For the preparation of the Rag-1 homozygous KO cells, Rag-1 heterozygous KO porcine fetal fibroblasts (Rag +/2 PFF) were isolated from a Rag-1 heterozygous Duroc male fetus produced by nuclear transfer. These cells were then transfected with the Rag-1 homozygous KO vector using the method described above. For the selection of homozygous KO cells, 10 mg/ml blasticidin (Invitrogen, Carlsbad, CA, USA) was added to the heterozygous KO selection method. After being cultured for 10 days, each cell colony was separated into two parts and then continuously cultured. After from 24 h to 48 h culture, one of the two-division cells was isolated and used for PCR analysis [43]. Positive KO cells were grown to confluence in a 75 cm 2 flask and cryopreserved until SCNT.

Oocyte collection and SCNT
All pigs of this study were raised in a specific pathogen free (SPF) environment. In vivo matured oocytes were collected from estrus-synchronized and superovulated gilts and sows treated with equine chorionic gonadotropin and human chorionic gonadotropin (Novartis Animal Health Inc., Tokyo, Japan) with/without prostaglandin F 2a (Asuka Pharmaceutical Co., Ltd., Tokyo, Japan) as described previously [44][45][46]. Oocytes were removed from cumulus cells in 1 mg/ml hyaluronidase in phosphate-buffered saline (PBS; Takara Bio Inc., Shiga, Japan) supplemented with 0.1% bovine serum albumin (Sigma-Aldrich, St.Louis, MO, USA). Enucleation and SCNT were performed using a procedure based on blind methods using a piezo-actuated system (PRIME TECH LTD., Ibaraki, Japan) previously described [28]. Oocytes injected with donor cells were electrically activated by a single direct current pulse at 1.5 kV/cm for 100 ms. Reconstructed embryos were transferred into porcine zygote medium-3 (PZM-3) supplemented with 5 mg/ml cytochalasin B (Sigma-Aldrich, St.Louis, MO, USA), and after a 2 h incubation at 38.5˚C in 5% CO 2 , 5% O 2 , and 90% N 2 , the embryos were cultured in PZM-3 until embryo transfer [47]. Reconstructed embryos with clear cytoplasts were selected and surgically transferred into the oviducts of estrus-synchronized recipient gilts. All the Rag-1-deficient neonatal piglets produced during this research were euthanized, along with age-matched two wild-type piglets, for analysis.

FACS
PBMCs were analyzed by FACS as previously described [48]. Peripheral whole blood was collected from each wild-type and Rag-1-deficient newborn piglet and immediately transferred to a collection tube supplemented with ethylenediaminetetraacetic acid to prevent coagulation. Gradient centrifugation was used to separate PBMCs. After PBMCs were washed with PBS, they were stained for 30 min at 4˚C using a phycoerythrin-conjugated mouse anti-porcine CD3e antibody (Beckman Coulter, Inc., Brea, CA, USA) and a fluorescein isothiocyanate-conjugated mouse anti-human CD19 antibody (Affimetrix, Inc., Santa Clara, CA, USA). Samples were analyzed on a flow cytometer (Becton, Dickinson and Company, Flanklin Lakes, NJ, USA).

Histological analysis
Spleen, MLNs, and thymus samples were collected from both wild-type and Rag-1-deficient neonatal piglets. These tissues were fixed in 4% paraformaldehyde (Wako Pure Chemical Industries, Ltd., Osaka, Japan) and embedded in paraffin blocks. Thereafter, tissue samples were cut at 4 mm and stained with hematoxylin and eosin (H&E). Histologic sections were evaluated under a microscope.

Preparation of Rag-1 KO fetus fibroblast cells
The Rag-1 targeting vectors constructed and the gene targeting scheme are shown in Figure 1. The vector construct was designed to produce a 27-base pair (bp) deletion in Exon 2 including open reading frame by homologous recombination as a result of the antibiotic resistance gene cassette. Therefore, translation is stopped approximately 1.3 kb downstream from the initiator codon, and truncated Rag-1 proteins are produced. Transfected cells possessing a DNA fragment from the inserted vector region of the antibiotic resistance gene to the 39 short homologous arm region were amplified by PCR, whereas nontransfected cells were not. Nine of the 384 cell colonies obtained were PCR-positive for Rag-1 The T-B-SCID Piglets with Rag-1 Deficiency heterozygous KO transfection. Among them, two colonies, #51-2 and #95-2, were successfully expanded and used for SCNT. Regarding Rag-1 homozygous KO transfection, of the 288 colonies, 12 PCR-positive cell colonies were obtained, six of which were used for SCNT.
Cloning and production of Rag-1 KO neonatal piglets Rag-1 heterozygous KO donor cells (#95-2) obtained from the original PFF (T6-12) by gene targeting were used for SCNT (Table 1: Step 1). The reconstructed embryos were cultured in PZM-3 medium for 2 days. Day 2 reconstructed embryos were transferred to one recipient gilt, and the pregnant gilt was used to obtain a Rag-1 heterozygous KO fetus for the generation of Rag-1 homozygous KO cells at a gestation of 33 days (Fig. 2: Rag +/2 PFF). Next Rag-1 homozygous KO donor cells (#3-55), derived from Rag-1 heterozygous KO PFF by homologous recombination, were used to obtain Rag-1 homozygous KO fetal fibroblast cells (Table 1: Step 2). The Rag-1 homozygous KO fetal fibroblast cells generated were used to produce Rag-1-deficient newborn piglets ( Fig. 2: Rag 2/2 PFF). A total of 526 reconstructed Day 1 and 2 embryos were mixed and transferred into four recipient gilts (Table 1: Step 3). Pregnancy was confirmed in three recipients using an ultrasound scanner at 26-28 days after embryo transfer. Eventually, two pregnant sows delivered two piglets by cesarean section.
Samples of genomic DNA were extracted from the original PFF (T6-12), heterozygous KO cells (#95-2, Rag +/2 PFF), and homozygous KO cells (#3-55, Rag 2/2 PFF). Moreover, the tail tissues were obtained from the Rag-1-deficient (Rag 2/2 No. 1 and No. 2) and newborn wild-type piglets (WT No. 1 and No. 2) for genomic DNA extraction. Genomic PCR was employed to confirm Rag-1deficiency and the presence of targeting vector transgenes ( Fig. 2A). The targeting vector transgene was detected in Rag-1-deficient newborn piglets, which was in agreement with Rag 2/2 PFF donor cells, but was not detected in wild-type piglets. Moreover, the Rag-1 Exon 2 region containing a 27 bp region deleted as a result of homologous recombination was present in wild-type piglets (Fig. 2B). On the other hand, Rag-1-deficient piglets did not possess the deleted 27 bp region of Rag-1 Exon 2. The genomic PCRs of Rag-1-deficient piglets perfectly matched that of donor cells (#3-55, Rag 2/2 PFF). These results imply that the piglets generated by SCNT were cloned from Rag-1-deficient donor cells and possessed an inactivated Rag-1 gene.

FACS analysis of PBMC
We used FACS to analyze differentiated lymphocyte populations among PBMCs using an anti-porcine CD3 antibody for mature T cells and an anti-human CD19 antibody for Immunoglobulin M (IgM)-positive B cells. The anti-human CD19 antibody used in FACS analysis reacted with porcine B cells. Wild-type (Rag-1 +/+ ) piglets possessed more than 30% CD3-positive T cells and 45% IgM-positive B cells among their PBMCs (Fig. 4). In contrast, CD3-positive T cells and IgMpositive B cells were not present (,1%) in Rag-1 2/2 piglets. Therefore, Rag-1 deficiency prevents the differentiation of lymphocytes and causes a lack of IgMpositive B and mature T lymphocytes in pigs.

Histological analysis of Rag 2/2 neonatal piglets
Spleens, MLNs, and thymuses were histologically investigated by H&E staining (Fig. 5). In wild-type neonatal piglets, spleens clearly possessed white pulp, clusters of lymphocytes had formed around the central artery (Fig. 5A), and numerous lymphoid follicles were evident in MLNs (Fig. 5C). In contrast, the white pulp of the spleen and lymphoid follicles of MLNs were unclear and virtually devoid of small lymphocytes in Rag-1 2/2 neonatal piglets (Fig. 5B, D). Likewise, in wild-type piglets, the lobules of the thymus consisted of cortex and medulla and many lymphocytes were extant in the cortex region (Fig. 5E, G). On the other hand, there was no division between the cortex and medulla in the lobules of the thymus in Rag-1 2/2 piglets, and polygonal cells with irregular nuclei and large cytoplasm existed in the gaps between interdigitating reticulum cells (Fig. 5F, H). These polygonal cells were similar to cells observed in the spleens and MLNs of Rag-1-deficient pigs and exhibited morphologic characteristics reminiscent of immature lymphocytes.
Immunohistochemical analysis of Rag 2/2 neonatal piglets Furthermore, we confirmed lymphocyte development in the thymus, spleen, and MLNs by immunohistochemistry using CD3 as a lymphocyte marker (Fig. S1). There were numerous CD3 + T cells in the medulla of the thymus (Fig. S1A), the periarterial lymphatic sheaths of the spleen (Fig S1C), and the surrounding lymphoid follicles of MLNs (Fig. S1E) in wild-type piglets. In contrast to wild- Table 1. Production of Recombination activating gene-1-knockout pig fetuses or piglets by somatic cell nuclear transfer.
Step Objective type piglets, the thymus, spleen, and MLNs of the Rag-1 KO piglets showed loss of CD3 + T cells (Fig. S1B, D and F).

Discussion
In neonatal pigs, B cell development occurs earlier than T cell development and begins in the yolk sac [49,50]. T cells are first detected in the thymus at a gestation of 40 days [49]. Peripheral blood IgM-positive B cells begin to appear at 40 days of gestation [49]. On the other hand, peripheral T cells are generated at 45 days of  gestation [49]. In association with B lymphocyte development, the Rag-1 gene is strongly expressed and V(D)J rearrangement starts at 20 days of gestation [50], and the rapid elevations of mRNA levels of Rag-1, -2, and CD3e in the thymus increase from 40 to 65 days of gestation [51]. In this study, we generated Rag-1deficient model pigs using gene targeting and SCNT and confirmed the contribution of the Rag-1 gene to lymphogenesis and the phenotype of Rag-1deficient pigs. In mice, previous experiments have revealed that disruption of Rag-1 or Rag-2 blocks the differentiation of lymphocytes and causes T-B-SCID [6,7]. Moreover, Rag-deficient blastocyst complementation allowed the generation of somatic chimeras possessing mature B and T cells, all of which were derived from injected ES cells [52]. In addition, studies on Rag activity in humans have revealed that Rag gene mutations cause T-B-SCID or Omenn syndrome [8][9][10][11]. Rag-1 and Rag-2 interact and the Rag complex functions as a sequence-and structure-specific nuclease during V(D)J recombination [53]. These reports suggest that Rag-1 and  The T-B-SCID Piglets with Rag-1 Deficiency suggests that the inactivation of Rag-1 blocks the differentiation and proliferation of T and B cells and Rag genes are important for the differentiation and proliferation of T and B lymphocytes in pigs, which is similar in mice and humans. These results confirm that Rag-1-deficient pigs exhibit the T-B-SCID phenotype. The T-B-SCID Piglets with Rag-1 Deficiency In a recent study, Rag-1-deficient rats were reported by two groups [54,55]. Although their Rag-1 mutant rats completely lacked Rag-1 protein and exhibited a hypoplastic thymus, they retained a small quantity of CD3-positive T cells or IgMpositive B cells [54,55]. These differences in lymphocytes differentiation were ascribable to translation from an internal methionine without an N-terminal BIAA in the case of frame shift-mutated Rag-1 protein truncated in front of the nonamer binding region [56]. Our gene targeting scheme disrupted the coding sequence in Exon 2, 1.3 kb downstream from the initiator codon, leading to the inactivation of the porcine Rag-1 gene through antibiotic resistance genes similarly to Rag-1 KO mice [6]. Our results show that the function of Rag-1 gene was completely blocked on Rag-1 KO piglets.
Consistent with FACS analysis, histologic and immunohistochemistry analyses also showed a lack (abnormal development) of mature lymphocytes and structural defects in the spleen, MLNs, and thymus of the Rag-1-deficient piglets. Immunoglobulin joining region gene (JH) KO piglets devoid of B cells show a complete lack of follicular structure and germinal center organization in MLNs [57]. Moreover, Il2rg-disrupted pigs exhibit greatly reduced T cell development and hypoplastic white pulp in the spleen [38]. Furthermore, Rag-1/-2 knockout pigs were recently reported to show a hypoplastic periarterial lymphatic sheath and loss of white pulp in the spleen as well as hypoplastic corpuscles in the thymus [41,42]. Consistent with these lymphoid-defective pigs and previous reports on rodents, our Rag-1 2/2 neonatal piglets lacking T and B cells demonstrated hypoplasia of the white pulp and lymphoid follicles. Regarding the thymus, the architecture of the medulla and cortex and the lymphocytes were completely lost in Rag-1 2/2 piglets similar to the recent report [42]. Moreover, our Rag-1 2/2 neonatal piglets showed no CD3-positive cells in their thymus, spleen, or MLNs. These histologic findings also clearly prove that the characteristics of our Rag-1 2/2 neonatal piglets did not result from a leaky phenotype.
Unlike in rodents and Rag-1 and -2-biallelic mutant pigs, coat color changed in our Rag-1 KO piglets. We used donor cells derived from Duroc fetuses. As shown in the wild-type figure ( Figure 3A), Duroc piglets have a dark brown coat. On the other hand, the coat of Rag-1-deficient piglets was discolored. It is known that DNA methylation and epigenetic changes can be caused by SCNT [58][59][60][61][62], and phenotypic and epigenetic variation resulting in depigmentation has been reported in cloned pigs [63,64]. In a study of piglets cloned using Duroc donor cells [64], one of nine cloned piglets exhibited depigmentation of the skin and hair. It is possible that the downregulation of certain genes induced this depigmentation, although the changes occurred during puberty. However, both our Rag-1-deficient piglets generated showed the same coat color accompanying the T-B-SCID phenotype. Therefore, the discoloration in Rag-1-deficient piglets may result from the Rag-1 gene KO rather than an effect of cloning in colored pig breeds. The cause of this coat discoloration phenomenon is unknown, and further study is necessary to elucidate the relationship between coat coloration and Rag-1 disruption in colored pigs.
Animals with SCID are a versatile model for laboratory, translational, and biomedical research. Furthermore, they may represent a platform for the development of humanized animals and the assessment of pluripotent stem cells. Although rodent models are widely used, human-scale models and long-term testing are still required to bridge the gap between animal models and humans. In this study, our findings provide additional evidence that the Rag-1 gene is essential for lymphocyte development in pigs, which is similar in rodents and humans, and we confirmed that Rag-1-deficient pigs display a T-B-SCID phenotype with no leaks. Therefore, Rag-1 KO pigs represent a promising model and provide important research resources that are similar to those described in recent reports. Future studies are required to confirm their utility as well as the fundamental characteristics of the phenotypic, immunologic, and morphologic changes in other breed and growth process.