Due to a rising demand of porcine models with complex genetic modifications for biomedical research, the approaches for their generation need to be adapted. In this study we describe the direct introduction of a gene construct into the pronucleus (PN)-like structure of cloned embryos as a novel strategy for the generation of genetically modified pigs, termed “nuclear injection”. To evaluate the reliability of this new strategy, the developmental ability of embryos in vitro and in vivo as well as the integration and expression efficiency of a transgene carrying green fluorescence protein (GFP) were examined. Eighty percent of the cloned pig embryos (633/787) exhibited a PN-like structure, which met the prerequisite to technically perform the new method. GFP fluorescence was observed in about half of the total blastocysts (21/40, 52.5%), which was comparable to classical zygote PN injection (28/41, 68.3%). In total, 478 cloned embryos injected with the GFP construct were transferred into 4 recipients and from one recipient 4 fetuses (day 68) were collected. In one of the fetuses which showed normal development, the integration of the transgene was confirmed by PCR in different tissues and organs from all three primary germ layers and placenta. The integration pattern of the transgene was mosaic (48 out of 84 single-cell colonies established from a kidney were positive for GFP DNA by PCR). Direct GFP fluorescence was observed macro- and microscopically in the fetus. Our novel strategy could be useful particularly for the generation of pigs with complex genetic modifications.

Cancers are a leading cause of death worldwide and a major priority for biomedical research. Most animal models of solid cancers are in rodents, particularly genetically engineered mice. However, mice differ significantly from humans in size, lifespan, physiology, anatomy, and diet, limiting their usefulness for some studies. Pigs are increasingly recognised as a valuable adjunct to pre-clinical research. Our aim is to provide a series of genetically defined pigs that model serious and common human cancers. These will allow new diagnostic and therapeutic strategies to be investigated at human scale, and longitudinal studies under conditions that mimic the human patient. We are thus engaged in a program of gene targeting to replicate in pigs a series of genetic lesions known to underlie human cancers. Here, we describe results from two key tumour suppressor genes: adenomatous polyposis coli (APC) and p53 (TP53). Somatic mutations resulting in inactivation or altered p53 function are present in most human cancers, and germline TP53 mutations are responsible for Li-Fraumeni multiple cancer syndrome. TP53R175H is the most frequent missense mutation in many sporadic human cancers. We have created gene-targeted knockout pigs and pigs carrying a latent TP53R167H mutant allele orthologous to human mutant TP53R175H that can be activated by Cre recombination to model the occurrence of oncogenic mutant p53 in chosen tissues (Leuchs et al. 2012 PLoS One, in press). In vitro studies indicate that porcine TP53R167H resembles human TP53R175H in altered function, and homozygous knockout of porcine TP53 results in transformation of porcine MSCs. APC plays a vital initiating role in both sporadic colorectal cancer (CRC) and the inherited predisposition to colorectal cancer, familial adenomatous polyposis (FAP). We generated gene-targeted cloned pigs carrying two different nonsense mutations in APC (APC1061 and APC1311) at sites orthologous to human germline mutations responsible for FAP. At 1 year of age, the APC1311 mutation resulted in >100 lesions, including ~60 polyps, exclusively in the large intestine. Importantly, this accords with the location and onset of human FAP in early adulthood, and contrasts with equivalent mutations in mice where polyps develop in the small intestine. Histological and molecular analysis showed that the porcine model recapitulates all major features of early stage human FAP (Flisikowska et al. 2012 Gastroenterology, in press). Tumorigenesis involves multiple genetic alterations over time. It will now be possible to mimic this progression in pigs by combining these and other mutations. We are confident that pig models will make a significant contribution to human oncology.

Mutation of the tumor suppressor p53 plays a major role in human carcinogenesis. Here we describe gene-targeted porcine mesenchymal stem cells (MSCs) and live pigs carrying a latent TP53R167H mutant allele, orthologous to oncogenic human mutant TP53R175H and mouse Trp53R172H, that can be activated by Cre recombination. MSCs carrying the latent TP53R167H mutant allele were analyzed in vitro. Homozygous cells were p53 deficient, and on continued culture exhibited more rapid proliferation, anchorage independent growth, and resistance to the apoptosis-inducing chemotherapeutic drug doxorubicin, all characteristic of cellular transformation. Cre mediated recombination activated the latent TP53R167H allele as predicted, and in homozygous cells expressed mutant p53-R167H protein at a level ten-fold greater than wild-type MSCs, consistent with the elevated levels found in human cancer cells. Gene targeted MSCs were used for nuclear transfer and fifteen viable piglets were produced carrying the latent TP53R167H mutant allele in heterozygous form. These animals will allow study of p53 deficiency and expression of mutant p53-R167H to model human germline, or spontaneous somatic p53 mutation. This work represents the first inactivation and mutation of the gatekeeper tumor suppressor gene TP53 in a non-rodent mammal.

We have examined the use of RNA interference as a means of downregulating gene expression and provide the first comparison of shRNA and artificial miRNA constructs for transgenic livestock. Several in vitro assays were performed to identify the most effective RNAi constructs. shRNA and miRNA constructs achieved significant downregulation of two porcine target genes: the milk whey protein beta-lactoglobulin and the tumour suppressor p53. Results of different assays were, however, sometimes at variance, indicating that no one assay can be relied upon to predict the effectiveness of an RNAi construct. Our findings are that screening of RNAi constructs is most informative if carried out in primary cells that express the target gene and are competent for somatic cell nuclear transfer. Importantly, the use of miRNA constructs makes tissue specific gene knockdown in large animals a realistic possibility.