Generation of transgenic dogs that conditionally express green fluorescent protein



We report the creation of a transgenic dog that conditionally expresses eGFP (enhanced green fluorescent protein) under the regulation of doxycycline. Briefly, fetal fibroblasts infected with a Tet-on eGFP vector were used for somatic cell nuclear transfer. Subsequently reconstructed oocytes were transferred to recipients. Three clones having transgenes were born and one dog was alive. The dog showed all features of inducible expression of eGFP upon doxycycline administration, and successful breeding resulted in eGFP-positive puppies, confirming stable insertion of the transgene into the genome. This inducible dog model will be useful for a variety of medical research studies. genesis 49:472–478, 2011. © 2011 Wiley-Liss, Inc.


Dogs share a large number of disease types with humans (Mack,2005; Sutter and Ostrander,2004), and have been considered prime medical research models. However, lately dogs have lost some merit mainly due to the well-known difficulty in manipulating genetic information (Luvoni et al.,2006; Otoi et al.,2000). Transgenic animals are produced by two common methods of gene transfer: DNA insertion by pronucleus injection or homologous recombination in embryonic stem cells. So far, these methods are not available in dogs due to their unique reproductive physiology. It is particularly difficult to make matured canine oocytes and blastocysts in vitro, even though there have been a few reports (Luvoni et al.,2006; Otoi et al.,2004). This has limited manipulation of genetic information in this species by microinjection. Recently, dog ES cells have been reported (Hayes et al.,2008; Schneider et al.,2007), but germline competent dog ES cell lines have not yet been established and therefore traditional gene targeting methods using ES cells are not available in dogs.

To overcome this difficulty, we recently developed a novel transgenic procedure in dogs by applying a virus-driven gene insertion method to modulate genetic information in somatic cells, followed by substituting the whole genome with the targeted one using the technique of somatic cell nuclear transfer (SCNT). The transgenic dog line that expresses red fluorescent protein under a constitutively active promoter was generated by this procedure (Hong et al.,2009b).

In the current study, we aimed to generate transgenic dogs by employing an inducible gene expression approach. While a constitutively active promoter is effective to express a target gene, the uncontrollable expression often results in unwanted outcomes. To circumvent this problem, methods for inducible transgenic systems have been developed (Jaisser,2000), which include a tetracycline inducible vector system (Gossen and Bujard,1992). The system is composed of a tetracycline-dependent transactivator (tTA, also called Tet-off) driven by a specific promoter and a tetracycline-responsive element (TRE) consisting of a composite promoter that contains tet operator (TetO) sequences (Jaisser,2000). The resulting TetO promoter constitutively drives the expression of a gene of interest. However, the TetO promoter activity is shut off when treated with tetracycline or its analog, doxycycline (Gossen et al.,1995). An advanced system including a reverse tTA(rtTA or Tet-on) promoter has been developed that behaves in the opposite way (Roth et al.,2009). These systems however require the use of two transgenic animal lines that carry either transactivator or target genes, and the two lines need to be crossed to have both transgenes expressed in an animal. These requirements make it difficult to apply this system to large animals that take a long time to reach puberty and have prolonged gestation periods (Backman et al.,2004; Kues et al.,2006).

In this study, we used a universal doxycycline-inducible vector that contained both a driver and a target transgene (Kues et al.,2006), and an enhanced green fluorescent protein (eGFP) was used as the target gene to be expressed.


Tet-on eGFP cells containing an eGFP reporter gene and an rtTA2S-M2 transactivator sequence (Fig. 1a) were established. The induced transgene expression detected by epifluorescence microscopy (Fig. 1b) and by FACS analysis revealed a tight regulation of the transgene by doxycycline (Fig. 1c). The mean fluorescence intensity for Tet-on eGFP cells was increased approximately 42-fold over control during three days of incubation in doxycycline. Four days after ceasing the treatment, fluorescence intensity returned back to the pretreatment level (Fig. 1c).

Figure 1.

A doxycycline-controllable eGFP vector system and doxycycline responsiveness of Tet-on eGFP cells. (a) The retrovirus vector construct used for the generation of Tet-on eGFP dogs. LTR, long terminal repeat; HygR, hygromycin B-resistance gene; TRE-tight, modified version of the tetracycline-response element (TRE); mCMV, minimal cytomegalovirus promoter; eGFP, enhanced green fluorescent protein; PGK, phosphoglycerate kinase promoter; rtTA2s-M2, reverse tetracycline transactivator; WPRE, woodchuck hepatitis virus post-transcriptional regulatory element sequence. (b) Tet-on eGFP cells were cultured in the absence or presence of doxycycline (Dox, 1 μg/ml) for 24 h. (c) Cells were cultured in the presence of doxycycline (1 μg/ml) for five days and then in its absence for seven days. Mean fluorescence intensity for the infected cells was assessed by FACS.

A total of 203 in vivo oocytes were recovered and 182 oocytes were subjected to enucleation for somatic cell nuclear transfer. From these, 139 fused embryos were produced (fusion rate = 76.4%) and 135 embryos were transferred to nine surrogates. On Day 22 after embryo transfer, three females were diagnosed as pregnant (pregnancy rate = 33.3%) and three cloned beagles (Tet-on eGFP 1, Tet-on eGFP 2, and Tet-on eGFP 3) were delivered on Day 60 (rate of development to term = 2.2%, Table 1, Fig. 2a). We confirmed by microsatellite and mitochondrial DNA analyses that all 3 are cloned beagles originating from the Tet-eGFP cells (data not shown). The transgene insertion was detected from the skin DNA of the 3 Tet-on eGFP dogs and from the DNA samples that were isolated from a number of organs of Tet-on eGFP 2 (Fig. 2b and c).

Figure 2.

Detection of eGFP transgene in the genomic DNA of Tet-on eGFP dogs. (a) Photograph of Tet-on eGFP 2 and 3 on Day 3 after birth. (b) Insertion of the eGFP gene in the Tet-on eGFP dogs was confirmed by Southern blotting. Genomic DNA samples extracted from three Tet-on eGFP dogs were used. (c) Integration of the eGFP gene was confirmed in several internal organs of Tet-on eGFP 2. As the controls, genomic DNA extracted from Tet-on eGFP cells (P) and nontransgenic beagle (N) were included in the Southern assay.

Table 1. Characteristics of Tet-on eGFP Dogs Produced by SCNT
Pregnancy numberID of cloned dogsGestation (days)Birth weight (g)Delivery statuseGFP expression at birth
  • a

    Tet-on eGFP 1 was born with malformation.

  • b

    Tet-on eGFP 2 was died at day 10 after the birth due to interstitial pneumonia.

  • c

    Tet-on eGFP 3 alive without any sign of health problems (>11 month-year-old).

1Tet-on eGFP 160DeadaNegative
2Tet-on eGFP 260240AlivebNegative
3Tet-on eGFP 360290AlivecNegative

Emission of green fluorescence was not detectable in any of the transgenic beagles at birth (Table 1) or before doxycycline treatment (Fig. 3a), suggesting that leaky expression did not occur. Obvious fluorescence was observed in the whole body of Tet-on eGFP 3 (Fig. 3a) with no adverse clinical signs during the two weeks of doxycycline administration. The results of Western blot analysis also showed a robust eGFP band signal which appeared after doxycycline administration (Fig. 3b). After ceasing treatment, fluorescence intensity decreased back to the pretreatment level (Fig. 3a). Western blotting also showed a reduction of eGFP production during the first three weeks after ceasing doxycycline administration and it gradually decreased to pretreatment level by nine weeks after ceasing treatment (Fig. 3b).

Figure 3.

Induction of eGFP expression in a Tet-on eGFP dog. (a) Expression of the eGFP gene in the pads of an age-matched control dog and before and after doxycycline treatment in Tet-on eGFP 3. Apparent eGFP fluorescence was visible after administering doxycycline (Switch on, two weeks), and the fluorescent intensity was reduced to the pretreatment level by nine weeks after ceasing doxycycline administration (Switch off, 9 weeks). (b) Skin samples were collected before (Pretreatment) and after (Switch on) the doxycycline treatment. Skin tissues were also collected three, six, and nine weeks after switch off, and Western blot was performed Recombinant eGFP (P) and skin tissues from a non-transgenic beagle (N) were used for the positive and negative control, respectively.

When Tet-on eGFP 3 reached puberty, artificial insemination was performed with fresh semen from a wild type male beagle. On Day 60 after embryo transfer, Tet-on eGFP 3 delivered naturally four pups that weighed 140 (pup1, male), 180 (pup2, female), 180 (pup3, female), and 230 (pup4, female) g, respectively (Fig. 4a). The presence of eGFP transgene in three (pup1 ∼ 3) out of four pups was confirmed by PCR (Fig. 4b) and Southern blotting (supporting data). Green fluorescence was induced in the cultured PBMC of three pups after adding doxycycline to the culture media (supporting data).

Figure 4.

Confirmation of germ line transmission of the eGFP gene by PCR. (a) Four puppies were delivered from Tet-on eGFP 3. (b) The transgene eGFP and hygromycin B-resistance gene (HygR) were detected by PCR. For the controls, genomic DNA was also extracted from Tet-on eGFP 3 (P) and a nontransgenic beagle (N).


The objective of this study was to establish a doxycycline-inducible dog model for the temporal expression of a transgene. Unlike early versions of the Tet-on system that requires the use of two lines of transgenic animals that have either the regulator (operon) or target gene (Backman et al.,2004), the newly designed vector has both of them, rtTA2s-M2 and eGFP, on a single vector (Fig. 1a). This vector was successfully used for generating inducible eGFP dogs, indicating that the vector can be used for generating other transgenic dogs that conditionally produce proteins of interest by replacing the eGFP sequence with the genes that encode those proteins.

The kinetics of eGFP expression in the Tet-on system show a very tight regulation dependent on doxycycline as the expression of the transgene was progressively increased and maintained at a high level and decreased by doxycycline treatment (Szulc et al.,2006). Unlike the reported high level of transgene expression without induction in other Tet-on systems (Chtarto et al.,2003; Roth et al.,2009), our Tet-on eGFP dog did not show any sign of ‘leaky’ expression. Furthermore, eGFP gene expression in this study showed a pattern of rapid reduction to the pretreatment level within 5 days after doxycycline removal as demonstrated by rapid reduction of fluorescence intensity in skin (Fig. 3a) as well asin peripheral mononuclear cells (data not shown)and the eGFP band signal in Western blotting (Fig. 3b).

Unexpectedly and interestingly, we found that only 1/10 (100 mg/kg/2 days) of the doxycycline dose (560 mg/kg/day) currently used in mouse Tet-on systems (Urlinger et al.,2000) was effective in inducing eGFP in our dogs. We initially intended to determine a dose that would be high enough to induce eGFP expression but not adversely affect the dogs. The therapeutic dose of doxycycline in dogs is 10 mg/kg/day (Harrus et al.,1998; Sainz et al.,2000) and the LD50 dose is reportedly higher than 500 mg/kg/day (, We first administered doxycycline to age-matched control beagles with increasing doses beginning from the therapeutic dose until they showed clinical signs such as nausea and vomiting, the most commonly reported adverse effects of oral therapy of doxycycline in dogs (Plumb,2002). This led to our finding that administration of 100 mg/kg every other day was sufficient. The commonly used doxycycline dose in rodents (560 mg/kg/day) to induce transgene expression (Urlinger et al.,2000) is 56 times higher than the therapeutic dose (10 mg/kg/day) (Plumb) and one third of the LD50. Considering the potential toxicity that may be caused by this high dose of doxycycline in rodents (Holst and Phemister,1971), it is of interest to determine if lower doses than the currently used ones or the dose that we administered in this study would be sufficient to induce genes of interest in other species.

Often transgenic animals are infertile, mostly due to the insertion of the transgene into a gene that is associated with fertility regulation (Meng et al.,2002). We therefore performed a fertility test of the Tet-on eGFP dog by artificially inseminating with fresh semen of a wild type beagle. The female gave birth to four puppies of which three were eGFP positive, proving its fertility as well as stable insertion of the transgene into the genome.

In conclusion, this study is the first report of the successful generation of a transgenic dog line that conditionally expresses a target gene. The transgenic vector used in this study as well as the procedures and dosage that were established will open a new avenue for applying transgenesis in canines and therefore for developing novel canine animal models for the study of human diseases.


Animal Use

Mixed-breed female dogs weighing 20–35 kg were used as oocyte donors and recipients. All dogs were cared for in separate indoor facilities and managed following a standard procedure established by the Committee for Accreditation of Laboratory Animal Care (approval number: SNU-090508-5) and according to the Guideline for the Care and Use of Laboratory Animals of Seoul National University.

Establishment of a Fibroblast Cell Line Containing an eGFP Expression Construct

Dog fibroblasts infected with with an eGFP expression vector was performed as previously reported (Hong et al.,2009b). Briefly, the plasmid used in this study (Fig. 1a) was generated by ligating the eGFP reporter gene, PGK promoter and rtTA2S-M2 transactivator sequence into pTRE-Tight plasmid (Clontech, Mountain View, CA). After PT67 cells (Clontech) were transiently infected with pTet2-GPTW, the resultant viruses were added to GP2-293 cells (Clontech). The pTet2-GPTW-infected GP2-293 cells were selected with hygromycin B for 2 weeks and the resulting cells were transfected with pVSV-G (Clontech). Forty-eight hours post transfection, the virus-producing cells were cultured in DMEM (Dulbecco's Modified Eagle Medium, Invitrogen, Carlsbad, CA) with glucose (Invitrogen), FBS (fetal bovine serum, Invitrogen), penicillin, and streptomycin. The virus-containing medium harvested from the cell culture was filtered and used to infect female beagle fetal fibroblasts. The infected cells (Tet-on eGFP cells) were frozen and maintained at −150°C until use for fluorescence-activated cell sorter (FACS) analysis and SCNT. To analyze that the Tet-on vector system worked in the fibroblast cells, mean fluorescence intensity was evaluated by FACS. The Tet-on eGFP cells were cultured with DMEM supplemented with 10% FBS and 1% nonessential amino acid (NEAA, Sigma-Aldrich Corp., St. Louis, MO) in the presence of 1 μg/ml of doxycycline for five days and then cultured without the doxycycline for seven days. For FACS preparation, cells were retrieved by trypsinization, neutralized, and suspended in phosphate-buffered saline (Invitrogen). Analyses were performed with FACSCalibur (Becton-Dickinson, NY, USA).

Generation of Doxycycline Inducible Transgenic Cloned Dogs

To collect in vivo matured oocytes, oviducts of oocyte donors were flushed with Hepes-buffered TCM-199 (Invitrogen) (Lee et al.,2005). Cumulus cells were removed and the oocytes were enucleated by micromanipulation (Jang et al.,2007). For the preparation of donor cells, 1 μg/ml of doxycycline was added to Tet-on eGFP cells the day before SCNT. Tet-on eGFP cells expressing the eGFP gene were transferred into the enucleated oocytes. The resulting couplets were electrically fused and chemically activated (Hong et al.,2009b), then transferred into the oviducts of surrogates which were naturally synchronized (Kim et al.,2010). Pregnancy diagnosis was performed by ultrasonography at least 26 days after embryo transfer. Parentage analysis was performed using microsatellite analysis to confirm the genetic identity between the nuclear donors and cloned dogs (Hong et al.,2009a). Canine mitochondrial DNA (GenBank accession no. U96639) was used to identify that the mitochondrial DNA of the cloned dogs was inherited from the oocyte donors (Jang et al.,2008).

Southern Blot Analysis

Samples for Southern blot analysis were collected from skin tissues of cloned puppies and internal organs of a puppy. Southern blotting was performed as previously described (Hong et al.,2009b). Genomic DNA (20 μg) was digested with BamHI and then separated on a 1% agarose gel. A DNA fragment of 713 bp spanning the 3,201–3,913 region of pTet2-EGFP-PTW vector was amplified using the following primer sets: 5′- AGCAAG GGCGAGGAGCTGTT-3′ and 5′-GACTTGTACAGCTCGT CCATG-3′. This fragment was used to synthesize a Southern probe using the PCR DIG Probe Synthesis kit (Roche, Mannheim, Germany). The resulting probe was labeled with digoxin alkaline phosphatase and purified by agarose gel electrophoresis before hybridization. A positive Southern signal was detected using a DIG luminescent detection kit (Roche).

Doxycycline Administration

A dose of 100 mg/kg of doxycycline was administered to one cloned dog for inducing eGFP expression. The doxycycline was dissolved in 5% sucrose to mask the bitter taste and orally administered every other day for two weeks.

eGFP Expression in Cloned Dog

Blood samples (5.5 ml) were collected from the jugular vein of the cloned dog every other day and used for serum enzyme assay and peripheral blood mononuclear cell (PBMC) isolation. During the doxycycline treatment, eGFP expression of cloned dog was observed in the PBMC via epifluorescence microscopy. Emission of green fluorescence was also monitored from the whole body of the cloned dog using a GFsP-5 head light source (Biochemical Laboratory Services, Budapest, Hungary) and an emission filter with a maximal transmittance wavelength of 500–650 nm.

Western Blot Analysis

Skin tissue samples were collected before and after the doxycycline treatment and three, six, and nine weeks after the last doxycycline treatment. The skin tissues were stored at −80°C until used. Western blotting was performed as described previously (Choi et al.,2006). In short, after lysis of the specimens, supernatants were collected and loaded onto a 12% SDS-polyacrylamide gel. The proteins were electrophoresed and transferred to a polyvinylidene fluoride membrane. The membrane was blocked with 5% skim milk in TBS with 0.1% Tween-20 (MTBST) and incubated for 16 h at 4°C in MTBST containing anti-EGFP antibody (Clontech), and alpha Tubulin (Abcam, Cambridge, MA) antibody. The membrane was washed three times and incubated for 1 h in MTBST containing HRP-conjugated goat anti-mouse IgG (Pierce, Rockford, IL). Then, SuperSignal West Pico Substrate (Pierce) was added and developed after exposing the membrane to X-ray film.

Stable Transmission of the eGFP Gene to the Next Generation

When the female cloned dog reached puberty, serum progesterone concentration was monitored for detecting timing of ovulation (Hong et al.,2009b). Approximately 72–79 hr after ovulation, the dog was surgically inseminated into both uterine horns as previously described (Park et al.,2009). On Day 30 after insemination, pregnancy was confirmed by ultrasonography. To prove integration of the eGFP gene into the offspring, skin samples were collected from the pups on Day 4 after birth. Polymerase chain reaction (PCR) was performed for genotyping. For PCR analysis, a primer pair for the eGFP gene, upstream (5′-CAGTGCTTCAGCCGC TACCC-3′) and downstream (5′- AGTTCACCTTGATGC CGTTCTT-3′) primers, were used to amplify a 289 bp DNA fragment. To detect the HygR gene in the retrovirus vector, upstream (5′-GCTCTCGA TGAGCTGATGC TTTG-3) and downstream (5′-TCTGCTGCTCCATACA AGCCAAC-3′) primers were used to amplify a 208 bp DNA fragment. The initial denaturation was performed at 95°C for 10 min, followed by 35 cycles at 95°C for 30s (denaturation), 54°C (eGFP) or 57°C (HygR) for 30 s (annealing), and 72°C for 30 s (extension), and a final incubation at 72°C for 10 min to ensure complete strand extension.