The founder animals were produced with a simplified DNA injection method of circular DNA plasmids into the cytoplasm of porcine zygotes [25,45]. We presume that the CPI techniques result in higher developmental rates of porcine zygotes, because the fractionation of cellular components by high speed centrifugation is omitted. To produce the transgenic founders, we combined the non-autonomous hyperactive Sleeping Beauty (SB) 100x-system [46], with CPI [45]. The components of a non-autonomous transposon system were separated on two different expression plasmids. The injected zygotes were surgically transferred to synchronized foster sows. The first litter was delivered at term (day 114) and had a size of 12 piglets, of which five were transgenic. All piglets had a normal birth weight (750–1,250 gram), were healthy and had normal behavior. A normal sex ratio was found, and no tendon problems or other malformations were detected [47,48]. Two Venus-transgenic boars (#503 and #505) were raised to sexual maturity. Figure 2 shows one of the F0 boars at the age of one month under specific Venus excitation. Southern blot analysis indicated that the transgenic founder boars carried three monomeric transposons [25]. Previously, we sequenced a total of 25 integration sites of the Venus transposon [25], however due to gaps and ambiguities in the current pig genome data, only five could be assigned to porcine chromosomes X, 3, 7, 8 and 13. Importantly, no random integrations of plasmid backbone sequences or of the SB transposase helper plasmid were found [25]. Here, we present a detailed report of reproductive parameters and germ line transmission of the founder boars, and cellular and molecular analyses of their offspring.

At the age of eight months, the founder boars (#503 and #505) were trained to use a dummy, both boars showed a good libido and ejaculated sperm could be recovered. Interestingly, all spermatozoa were uniformly Venus-positive. According to Mendelian rules, segregation of the three monomeric transposons should result in 12.5% triple-, 37.5% double-, 37.5% mono- and 12.5% non-transgenic sperm cells [49]. Apparently, the Venus protein was uniformly distributed between the developing sperm cells, which are connected via cytoplasmic bridges. This equal distribution of cytoplasmic components during spermatogenesis did not correlate with the genotype (Figure 3). A detailed analysis of ejaculated spermatozoa from the founder boar is given in reference [49]. Assessment of general parameters of sperm fertility showed that ejaculates from the transgenic boars fulfilled standard semen quality requirements, which included motility, velocity and morphological aspects [49].

To test the fertility of transgenic sperm, wild type sows were artificially inseminated with semen from the two Venus transgenic boars (#503 and #505), respectively. Six wild type sows were inseminated at natural heat and five became pregnant. Subsequently, three transgenic F1 sows were inseminated with founder semen; all three became pregnant and delivered healthy piglets at term. In total, 62 piglets were born, of which 81% were transgenic (Table 1). Molecular analysis confirmed segregation of the monomeric transposons in the offspring (see below). The piglets showed a normal birth weight, normal growth and social behaviour. The transgene expression did not affect health of the piglets. Previously, the most common clinical signs of disease associated with transgene expression in pigs included lethargy, lameness, uncoordinated gait, exophthalmos, and thickened skin [50]. Most likely, this phenotype was caused by ectopic expression of recombinant growth hormone. EGFP-transgenic piglets, produced with a lentiviral vector, were similar to wildtype animals with regard to reproductive parameters, behavior and general welfare criteria [37]. Here we extend these observations to transposon transgenic pigs, which showed an ubiquitous expression of the Venus fluorophore. Venus is an improved version of the yellow fluorescent protein (YFP) with stronger relative fluorescence, improved refolding after denaturation, and an increased pH resistance [51]. Compared to EGFP, the Venus protein carries several amino acids exchanges.

A direct correlation between the copy number of monomeric transposons and the fluorescence intensity in organs and cells was evident. Expression of the Venus reporter was observed in all organs (Figure 4) and almost all cell types of F1-animals. Only erythrocytes were found to be Venus-negative by fluorescence microscopy (Figure 4). To exclude the possibility that the specific fluorescence properties of Venus were quenched by haemoglobin or other red blood cell components, blood samples of F1-animals were fractionated in plasma, leucocytes and erythrocytes and used for immunoblotting. Western blot analyses revealed Venus protein in leucocytes, but not in erythrocytes. Thus the CAGGS-promoter did not seem to be active in porcine erythrocytes or their precursor cells (Figure 4). Albeit mammalian erythrocytes lack a nucleus, their precursor cells do have a nucleus and all proteins necessary during the lifespan of an erythrocyte have to be synthesized prior to degradation of the nucleus. Previously, we showed that functional Venus protein was also deposited in “extracellular” hair of these pigs, and that its fluorescent properties remained stable over periods of months [52].

A histological examination showed expression in all solid organs of transgenic F1 animals containing different transposon integrations, and all cells expressed the transgene (Figure 4, Figure 5). The level of expression varied between cell types, but specific cell types showed consistent expression levels. High expression levels were detected in kidney and pancreas, whereas heart, lung and liver showed moderate to low expression (Figure 5).

No case of variegated or silenced expression was found, whereas in previous studies of transgenic pig lines derived by random transgene integration, transgene silencing was a common finding [36,53].

Ubiquitous expression of Venus was confirmed by Northern blotting and Western blot detection of Venus protein [25,49]. Southern blot genotyping confirmed segregation of the monomeric transposons and supported the identification of F1 animals carrying identical integration events of the transposon (Figure 6). Thus the two founders (each with three monomeric transposons) can be used to derive six independent transgenic pig lines with a monomeric transposon. Germ line transmission of the SB transposons was not affected by the integrations site. In the absence of exogenously supplied Sleeping Beauty transposase, no changes in the integration sites (remobilisation) were found, consistent with the lack of endogenous SB-like transposons in the porcine genome.

To assess, whether the Venus protein behaves “neutral” in transgenic pigs, or whether (detrimental) effects become obvious above a critical expression level, three F1 sows, carrying two transposon copies, were inseminated with sperm from founders #503 or #505 (Table 1). The sows delivered 18 piglets, which carried up to five monomeric transposons. Again, a direct correlation between transposon copy number of the carrier, and the specific fluorescence intensity was apparent (Figure 7). Detrimental effects of the Venus reporter with respect to vital parameters, such as birth weight, litter size and animal health were not found. Piglets, which carried five different Venus transposons at permissive loci, exhibited the highest fluorescence, and no critical expression level could be determined. Apparently breeding to F2 generation neither affected expression, nor animal health.

The mean littersize of 7.8 piglets for the germ line transmission experiments (Table 1) is smaller than the mean littersize of wild type pigs in the Institutes pig facility. However, this may be due to different selection criteria for these two populations of pigs. Whereas wild type animals are strongly selected for high fecundity, and animals which give rise to small litter sizes are slaughtered, the transposon founders were solely selected on the presence of the transgene. Also a higher ratio of male (62%) than female offspring (38%) was recorded (Table 1). This was found both in the transgenic and in the non-transgenic offspring of transposon transgenic parental animals, and did not correlate with the transgenic status. The smaller litter size and the altered sex ratio warrant further studies into the role of transposon transgenesis on porcine fertility.

Importantly, the active transposition in porcine zygotes did not require any antibiotic selection cassettes, which are un-wanted DNA elements according to guidelines provided by regulatory authorities like the European Medicines Agency (EMA) and the Food and Drug Administration (FDA) in the USA. Recent experimental data in transgenic mice and cattle support the notion that the presence of an antibiotic selection cassette may result in variegated expression of the gene of interest [28,54]. Thus gene transfer strategies, which omit or allow removal of unwanted “transgenic” DNA sequences, are necessary to fully exploit transgenic animal models.

Methodological improvements in gene transfer have led to a rapidly increasing list of biomedical livestock models during recent years [55,56,57,58,59,60,61,62]. Transgenic pigs have been shown to mimic genetic and metabolic human diseases [1,5,55,56,57,58,60,61,62,63]. An important example is the porcine cystic fibrosis model, which shows a similar disease phenotype as for human patients [64], whereas transgenic mouse models failed to exhibit lung, pancreatic and intestinal obstructions. Huntington´s disease is a neurodegenerative disorder characterized by expression of a mutated huntingtin. A transgenic pig model expressing the N-terminal huntingtin with a polyglutamine tract seems to accumulate misfolded proteins in neurons, and initiated apoptosis in the affected neurons [56]. Photoreceptor topography in the pig retina is highly similar to that in humans, as it includes a cone rich, macula-like area centralis. Transgenic pigs expressing macular dystrophy-causing mutation in the ELOVL4 (elongation of very long chain fatty acids-4) gene showed photoreceptor loss, disorganised inner and outer segments, and diminished electro-retinography responses, suggesting that the transgenic pigs mirror macular degeneration and provide a unique model for therapeutic interventions [65]. Recently, the first immune deficient pigs were cloned [66,67,68], which is promising for large animal models for cell transplantation experiments. Our results suggest that Sleeping Beauty transposon-catalyzed stable transgene delivery is a useful addition in the toolkit for improved genetic engineering of the porcine genome for the generation of novel disease models [69].

Animals were maintained and handled according to German laws according animal welfare and genetically modified organisms (GMO). All animal experiments were approved by an independent ethics committee. Sperm rich fractions were collected using a dummy and by the gloved hand technique. Sperm samples were extended with Androhep (1:1). For artificial insemination, semen was diluted to a final concentration of 10⁸ sperm cells/mL. For genotyping, blood and ear biopsies were collected, and DNA was isolated with the proteinase K method.

Southern blot was done according to standard protocols. The genomic DNA was digested with NcoI. Hybridization with a CAGGS-Venus probe resulted in a constant internal fragment of around 1.4 kb, and a variable external fragment of >1.4 kb per integration [25]. Southern blotting was done according to standard protocols.

Flow cytometry analysis of spermatozoa was performed using a FACScan (BD Bioscience) equipped with an argon laser (488 nm, 15 mW). Samples were diluted to 0.5 × 10⁶ cells/mL and measured in duplicates acquiring 10,000 cells per sample.

Cells were extracted in RIPA buffer and 10 microgram of protein per slot was separated on a SDS-PAGE gel, blotted to a PVDF membrane, blocked and probed with a rabbit polyclonal anti-EGFP antibody (Santa Cruz) in 1:1000 dilution, followed by a secondary anti-rabbit antibody in 1:10 000 dilution (Sigma-Aldrich). For detection an ECL+ kit (GE Healthcare) and an image acquisition system (Vilber Lourmat, Fusion SL 3500) were used. For complete Western blotting protocol see references [49,52].

Venus transgenic and wildtype animals were excited with a blue LED and images were recorded with a digital camera and an emission filter (Lee Filter). Isolated organs were recorded in an identical manner. Organ samples were fixed in neutralized 4% formaldehyde overnight, then transferred to a 20% sucrose solution, and frozen in embedding medium (Micom Laborgeräte, Walldorf, Germany) on dry ice. This treatment excellently preserved the Venus fluorescence. In a cryotome, 10 μm sections were cut, mounted in Vectashield and viewed an an Olympus BX60 equipped with epifluorescence (excitation: 450–490 nm; emission: 515–550 nm; dichroic mirror: 505 nm). Normalized images were recorded with a scientific digital camera (Olympus DP71).