| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1,2
* Department of Biotechnology, Institute for Animal Breeding (FAL), Mariensee, Germany;
Klinik für Viszeral- und Transplantationschirurgie, Medizinische Hochschule Hannover, Hannover, Germany; and
Molecular Biotechnology, Gesellschaft für Biotechnologische Forschung mbH, Braunschweig, Germany
2Correspondence: K. W., Klinik f. Viszeral-und Transplantations Clinique, Medizinische Hochschule Hannover, Carl-Neuberg Str. 1, Hannover D-30623, Germany. E-mail: wonigeit.kurt{at}mh-hannover.de H. N., Institut for Animal Breeding (FAL) Department of Biotechnology Höltystr. 10, Neustadt D-31535, Germany. E-mail: niemann{at}tzv.fal.de
ABSTRACT
The applicability of tightly regulated transgenesis in domesticated animals is severely hampered by the present lack of knowledge of regulatory mechanisms and the long generation intervals. To capitalize on the tightly controlled expression of mammalian genes made possible by using prokaryotic control elements, we have used a single-step transduction to introduce an autoregulative tetracycline-responsive bicistronic expression cassette (NTA) into transgenic pigs. Transgenic pigs carrying one NTA cassette showed a mosaic transgene expression restricted to single muscle fibers. In contrast, crossbred animals carrying two NTA cassettes with different transgenes, revealed a broad tissue-independent and tightly regulated expression of one cassette, but not of the other one. The expression pattern correlated inversely with the methylation status of the NTA transcription start sites indicating epigenetic silencing of one NTA cassette. This first approach on tetracycline regulated transgene expression in farm animals will be valuable for developing precisely controlled expression systems for transgenes in large animals relevant for biomedical and agricultural biotechnology.—Kues, W.A., Schwinzer, R., Wirth, D., Verhoeyen, E., Lemme, E., Hermann, D., Barg-Kues, B., Hauser, H., Wonigeit, K., and Niemann, H. Epigenetic silencing and tissue independent expression of a novel tetracycline inducible system in double-transgenic pigs.
Key Words: autoregulative cassette • bicistronic • transgenic livestock • tissue independent expression • DNA methylation
TRANSGENIC MICE AND farm animals harboring the first generation of conditional promoter elements, which were responsive to heavy metals or steroid hormones, suffered from high basal expression levels and pleiotropic effects (1
2
3
4)
. In contrast, binary expression systems based on prokaryotic control elements, which are responsive to exogenous IPTG, RU-486, ecdysone, or tetracycline (tet) derivatives are compatible with a tightly controlled expression (5
, 6)
. Several transgenic mouse lines expressing modified tet-system components in various tissues have been generated and a rapid and reversible transgene regulation has been shown (7
8
9
10
11
12
13)
. Factors affecting the efficiency of tetracycline-dependent gene regulation include the strain of mice and the site of integration (14)
, as well as clearance of tetracycline derivatives (15)
. A high degree of variability in activator-dependent target gene expression is observed and high background expression makes it necessary to screen for appropriately responding lines (16)
.
The original tetracycline system requires two DNA constructs for expression of the transactivator and transactivator dependent expression of the target gene, respectively. These DNA-constructs are usually integrated in two different lines of transgenic mice. On breeding, offspring are obtained in which the target gene can be regulated by addition of doxycycline (7)
. The long-generation intervals make this approach unfeasible in other mammalian species, including livestock (17)
. One approach to overcome this limitation is to combine both transactivator and target genes, including individual promoters in a single construct. This approach was compatible with efficient control of gene expression in mammalian cells and mice (18
, 19)
. However, the near physical distance of two promoter sequences makes this system prone to elevated background expression. A tetracycline-controlled transcriptional silencer has been used to suppress background expression (9)
. Alternatively, autoregulatory expression systems have been established that overcome the problem of promoter interference. In these systems the transactivator is either expressed from a tet-responsive bicistronic expression cassette (20
21
22
23)
or from a tet-responsive bidirectional promoter (24
, 25)
. It has been shown that expression of biologically active proteins can be tightly controlled with these systems, albeit minimum basal expression levels of the transactivator are required (26)
. Another advantage of autoregulative cassettes is that transgene transcription is linked to ubiquitously present cellular cofactors, thereby circumventing cell-/tissue-specific expression as observed for classical promoters.
Here, we describe tightly controlled transgene over-expression in the domestic pig by an autoregulative bicistronic tetracycline-responsive cassette (NTA). The NTA-cassette was designed to ensure high and ubiquitous expression levels of human regulators of complement activation (RCA) relevant for xenotransplantation (17)
, (i.e., decay accelerating factor (DAF)) and CD59. The rationale for this approach was twofold: (i) As proof of principle we wanted to demonstrate that single-construct cassettes can be successfully expressed in large animals; (ii) we intended to improve xenotransplantation by producing tet-off regulated human complement regulator transgenic animals. Unexpectedly, nine independent transgenic pig lines showed expression restricted to muscle fibers. However, in crossbred double-transgenic animals a broad tissue independent transgene expression of the CD59-NTA cassette, but not of the DAF-NTA cassette was discovered. This asymmetric expression pattern correlated with different methylation patterns of the NTA cassettes indicating that epigenetic mechanisms are critical for the function of the NTA system. Expression of human complement regulators and tet transactivator did not compromise animal health and fertility.
MATERIALS AND METHODS
Construction of the bicistronic NTA-cassettes
CD59 was amplified from pWTCD59 (27)
. DAF was amplified from pWTDAF (27)
. The transactivator tTA was amplified from pUHD15–1 (28)
, thereby flanking the tTA cDNA with NotI and PstI sites. The cDNAs were independently integrated into pTBC-1 (29)
and both plasmids were then fused via the unique BamHI and VspI sites. All polymerase chain reaction (PCR)-derived sequences were verified by DNA sequencing. Vector sequences are available on request.
Cell culture and transfection
NIH3T3 cells were cultivated in Dulbecco’s modified Eagle’s medium (DMEM). The RCA-NTA cassettes were cotransfected with a pAG60 by DNA-calcium phosphate coprecipitation (27)
. Cells were selected for resistance to G418 (1 mg/ml). Pools and cell clones were generated and cultivated in the presence and absence of doxycycline (2 µg/ml). Transfection of swine testis epitheloid (STE) cell line (kindly provided by A. Saalmüller, Wien, Austria) was carried out using essentially the same protocol as for NIH3T3 cells. Flow cytometry was performed as described earlier (27)
using specific monoclonal antibodies (PharMingen, Hamburg, Germany) to detect CD59 and DAF.
Determination of doxycyline plasma levels in pigs
A mouse-derived cell line NIH-TAluc with a stably integrated doxycycline-responsive luciferase cassette was used to determine doxycycline levels in plasma and fodder extracts.
Generation of transgenic pigs
Transgenic pigs were generated by pronuclear DNA microinjection as described (30)
. In brief, zygotes were collected from hormonally pretreated prepubertal German landrace donor gilts 15–18 h after the second mating. Animals were sacrificed, and zygotes were flushed from the excised oviducts. The NTA constructs were released from vector backbones by digesting the plasmids with BsrBI and Tth111I (New England Biolabs (NEB), Frankfurt, Germany) and isolated by gel electrophoresis. Approximately 10 pl of purified DNAs (4–10 ng/µl) were injected into one pronucleus of zygotes after centrifugation for 6 min at 14,500 g to visualize the pronuclei by concentrating the cytoplasmic lipids at one pole. A total of 653 microinjected zygotes were surgically transferred into the oviducts of 20 synchronized prepubertal gilts, resulting in 80 piglets from which 11 could be identified as transgenic by Southern blotting of the genomic DNA. For Southern blotting, the genomic DNA was digested with PstI (NEB) and blotted to Hybond membranes. The PstI digest released internal fragments from the NTA-cassettes, which were 2049 and 1658 bp for the CD59 and DAF-constructs, respectively. The complete constructs were isolated from vector backbones and used as probes after digoxigenin labeling. Hybridization products were visualized by chemiluminescence. Founder animals were bred with nontransgenic animals to test for germline transmission. Transgenic F1 and F2-animals were used for detailed analysis of the expression pattern and doxycycline responsiveness. Animals were treated according to institutional guidelines and experiments were approved by an external animal welfare committee.
Detection of CD59 and DAF expression
Transgenic offspring and wild-type control animals were sacrificed and tissue samples were obtained from the following organs: heart, lung, liver, kidney, spleen, thymus, skeletal muscle, pancreas, skin, brain, and aorta. Samples were used for primary cell cultures or snap-frozen in liquid nitrogen and used for RT–PCR, Northern blotting and immunohistological analysis. For RT–PCR, total RNA was isolated using the guanidinium thiocyanate-phenol-chloroform extraction. PCR detection and Northern blotting of CD59 and DAF transcripts were done as described recently (30)
. In brief, digoxigenin labeled 605 and 613 bp antisense cRNAs for CD59 and DAF were used for specific detection of the transgenic transcripts. The calculated sizes for the bicistronic mRNAs were 2.3 and 3.0 kb for the CD59 and DAF encoding transcripts, respectively. In addition to the 3.0 kb band a slightly faster migrating fragment was specifically detected in the DAF Northern. Apparently, the smaller fragment represented a degradation product of the 3.0 kb mRNA. In some experiments tissues collected from transgenic animals carrying a cytomegalovirus (CMV) promoter-CD59 construct were used as controls (30)
. A ß-actin (Actb) probe (31)
served as an internal control.
To detect CD59 and DAF, immunohistological staining was performed on cryostat sections using biotinylated anti-human mAB H19 (CD59) and mAB33572X (DAF/CD55), respectively (both antibodies were from BD PharMingen). Human heart tissue served as positive control. Specific antibody (Ab) binding was detected by incubation with streptavidine-peroxidase and 3-amino-9-ethylcarbazole (AEC) as substrate, cell nuclei were counterstained with hematoxylin. Tissue samples were evaluated at 100–200x magnifications.
Primary cell cultures of porcine fibroblasts were established from subdermal tissues as described (30
, 32)
and employed for expression analysis by Northern blotting or flow cytometry (FACS).
Genomic DNA was isolated from porcine tissues by proteinase K lysis. The following primers spanning the complete promoter regions of DAF-NTA and CD59-NTA constructs were used to amplify these regions by PCR: CMV-0, 5'-AAA ATA GGC GTA CAC GAG G; hDAF-1, 5'-ATC ACT GAG TCC TTC TCG CC and hCD59–1, 5'-CCC TCA AGC GGG TTG TGA CG. The PCR products were directly sequenced, employing the CMV-O primer.
Application of doxycycline
Doxycycline was fed in the form of tablets (doxy 200®, CT Arzneimittel GmbH, Berlin, Germany; one tablet contains 200 mg doxycycline). A daily dose of 3.3 mg/kg body weight was applied for periods from 6 to 45 d. In one feeding group, blood samples were taken during the course of the experiment to determine doxycycline levels in plasma. Muscle biopsies from the M. longissimus were obtained by a special biopsy device routinely used in meat quality testing program (33)
. Biopsies were taken from the longissimus muscle prior to doxycycline treatment, and at different time points thereafter and were analyzed by Northern blotting and immunohistology. In some experiments, subdermal tissue pieces from the biopsies were used to establish primary cell cultures.
Bioinformatic analysis of CpG islands
For CpG island prediction the MethPrimer program (http://www.urogene.org/methprimer) was employed. The NTA-cassette sequences were analyzed using the following criteria for CpG island prediction: i) a CG content greater than 60%, ii) a presence of CpG dinucleotides of greater than 0.6, and iii) a DNA region of greater than 300 bp.
Detection of methylated CpG dinucleotides by bisulfite sequencing
Bisulfite sequencing was done as described by Hajkova et al. (34)
. In brief, 20 ng genomic DNA was digested with EcoRI (NEB). Denatured DNA was embedded in 7 µl of 20 mg/ml low melting agarose (Biozym, Hess. Oldendorf, Germany), cooled to form agarose beads, and incubated in 2.5 M bisulfite-hydroquinone solution pH 5 (Roth, Hamburg, Germany) for 4 h at 50°C. PCR was carried out in a final vol of 100 µl with individual agarose beads containing embedded genomic DNA. PCR conditions were as follows: 1 x PCR reaction buffer (20 mM Tris-HCl pH 8.4, 50 mM KCl2), 1.5 mM MgCl2, 0.2 µM dNTPs (Amersham Biosciences Europe, Freiburg, Germany), and 0.6 µM of each primer for 40 cycles. The PCR fragments were separated by gel electrophoresis, isolated and directly sequenced. Only sequences with > 95% cytosine conversion of non-CpGs were analyzed. Methylation status of CpG dinucleotides was determined based on the ratio of cytosines to converted cytosines. The following primers were used to amplify the CMV minimal promoter of the CD59-NTA construct, CMV-2bi, 5'- GAA AGT TGA GTT CGG TAT TT, CD59–2bi, 5'- AAA AAT ATC CCA CCT TTT TC; for the DAF-NTA construct the CMV-2bi primer in combination with DAF-2bi, 5'- CCA TTA ACT ACC CTT AAA AC was used.
RESULTS
Evaluation of autoregulatory RCA–NTA expression cassettes in cell lines
We constructed an autoregulatory bicistronic tet-off based cassette that supports single step transduction (Fig. 1
A) and in which the mRNA is driven by the tet responsive promoter (PtTA). In two constructs carrying either the human CD59 or DAF cDNAs (designated CD59-NTA and DAF-NTA) the RCA reading frames are followed by an internal ribosomal entry site (IRES) that provides translation of the recombinant transactivator (tTA). Except for the human cDNAs, both cassettes were completely identical.
|
Mouse NIH 3T3 cells stably expressing the cassettes showed autoactivation and expression of the respective RCA (Fig. 1B
). Expression was efficiently suppressed by addition of doxycycline to the culture medium (Fig. 1B
). In addition, swine testis epitheloid (STE) cells were transfected with the RCA-NTA cassettes, and again autoactivation and conditional transgene regulation were found (not shown). Thus, autoregulatory bicistronic expression cassettes support tight regulation of human RCAs in xenogenic cells, suggesting that ubiquitous activation could be achieved in vivo.
Generation of RCA-NTA transgenic pigs
Five hCD59-NTA and six DAF-NTA transgenic founder animals were generated by pronuclear microinjection of the respective cassette into porcine zygotes, from which ten transgenic lines were established. Transgenic juvenile and adult F1 and F2-offspring were used for expression profiling in the absence of any tetracycline. Remarkably, an intensive immunostaining indicative for human RCA protein was found in single fibers of striated muscle in 9 of the 10 lines (Table 1
, Fig. 1C
; 2–3 animals investigated per line). Depending on the line, 5–30% of the muscle fibers stained positive. Expression in muscle was confirmed by Northern blotting with DAF and CD59 specific probes, respectively. Contrary to the tissue-independent design of the NTA cassette, Northern blotting, immunohistology, and FACS revealed that the animals did not express human RCAs in brain, heart, kidney, tongue, liver, skin, lymph nodes, blood and pancreas (summarized in Table 1
). Apparently, the (different) integration sites of the nine expressing lines had no or only a marginal effect on the muscle fiber-specific expression.
|
Integration of the entire bicistronic cassettes into the porcine genome was confirmed by Southern blotting of genomic DNA digested with restriction enzymes, cutting the NTA cassette internally near the ends (not shown). To rule out the possibility, that transgenic animals carried defective promoter sequences, the respective elements were amplified from porcine genomic DNA by PCR. Sequencing confirmed that heptamerized tet-operator and CMV basal promoter sequences were present in both, CD59-NTA and DAF-NTA cassettes. To further rule out putative contamination of the fodder with tetracycline and/or its derivatives, fodder and plasma samples of transgenic pigs were analyzed via a doxycycline sensitive reporter cell line. All samples proved to be negative for tetracycline activity (data not shown).
Enhanced tissue-independent expression of CD59 in crossbred animals carrying two NTA-RCA cassettes
To investigate whether the unexpected mosaic expression pattern in striated muscle was due to suboptimal transactivator expression levels, tTA was expressed in porcine primary fibroblasts from transgenic animals. Although cotransfected tTA-responsive reporter genes were activated, the chromosomal RCA-NTA constructs remained unresponsive (data not shown) indicating stable silencing of the chromosomal cassettes.
To test if the typical mosaic expression pattern in transgenic pigs could be altered by increased threshold levels of tTA in vivo, heterozygous transgenic animals carrying either CD59-NTA or DAF-NTA were crossbred to obtain double-transgenic animals. From three litters with a total of 24 piglets the expected mendelian distribution of transgene combinations was obtained; i.e., 5 piglets (21%) were double-transgenic (CD59-NTA + DAF-NTA), 12 (50%) carried either the CD59 or DAF cassette and 7 (29%) were nontransgenic. In addition, a female and male double-transgenic sibling (CD59-NTA + DAF-NTA) were mated and produced nine piglets, i.e., four piglets carrying both cassettes, four piglets had one cassette and one nontransgenic offspring was obtained. From the nine double-transgenic animals, three were sacrificed at the age of 6 wk and 6 mo, respectively, and organs were analyzed for expression of the transgenes. Muscle biopsies and blood samples were analyzed from the other double-transgenic animals. Interestingly, a selective up-regulation of CD59 expression was found in double-transgenic animals derived from crossbreeding L13 and L20 (Table 1
). The CD59 mRNA levels were significantly (50–100 fold) elevated in three different muscles, i.e., Musculus (M.) longissimus, front and hind leg muscles, in double-transgenic animals over transgenic siblings carrying only one CD59-NTA cassette (Fig. 2
A). Immunohistological staining revealed that virtually all fibers in striated muscle stained positive for CD59 (Fig. 3
A). In addition, other organs, including liver, pancreas, kidney and lung expressed the transgene at the mRNA (Fig. 2B
) and protein levels (Fig. 3A
). RNA and protein expression pattern of CD59 correlated well, e.g., muscle and liver, which showed the highest mRNA levels (Fig. 2A
) also displayed the highest proportion of anti-CD59 Ab stained cells (Fig. 3A
). Tissues with lower mRNAs levels, such as lung, showed only few positive cells in the immunohistology. Importantly, in all examined tissues the positive cells were strongly stained for CD59, suggesting that the autoregulative up-regulation worked well in this population of cells. However, DAF mRNA and protein levels were not increased in the same double-transgenic animals and immunoreactivity remained confined to single muscle fibers (Fig. 3A
). Seven out of nine double-transgenic pigs, derived from crossing transgenic lines L13 with L20 showed this asymmetric over-expression of CD59.
|
|
The remaining two double-transgenic pigs, derived from crossing L15 with L20, mimicked expression patterns of their parental lines showing mosaic expression in single muscle fibers. Individual muscle fibers coexpressed both CD59 and DAF (Fig. 3B
), while some fibers expressed only CD59 (Fig. 3B
, arrows), but not DAF (Fig. 3B
, circles), highlighting that in these fibers high transactivator levels alone were not sufficient to activate the DAF-NTA gene. Fibers with exclusive expression of DAF were never detected.
Since none of the single-transgenic animals showed this enhanced transgene expression, we argue that the presence of two cassettes is necessary, but not always sufficient for tissue independent expression. This finding was confirmed by the widespread expression of CD59 in a CD59-NTA homozygous animal (CD59-NTA +CD59-NTA) (data not shown).
Health status, development, and fertility of animals bearing either one or two NTA-cassettes were not compromised. All animals were not distinguishable from their nontransgenic counterparts.
Conditional RCA-NTA regulation in single- and double-transgenic animals
To evaluate the exogenous control of the autoregulated tet-off cassettes, 12 single-transgenic and four double-transgenic animals were fed with doxycycline and muscle biopsies were taken before, during and after antibiotic treatment. A daily dose of 3.3 mg doxycycline/kg body weight for 6 d was found to be effective to shut down transgene expression (Fig. 4
). This dose is well below the recommended effective dose of 12 mg doxycycline/kg/day for antibiotic treatment of pigs. RCA-NTA transcripts were virtually eliminated by doxycycline feeding within 2–6 d. Thus, the expression cassettes located at different chromosomal sites could be readily switched off. Re-expression of RCA-NTAs took unexpectedly long, and did not resume until eight weeks after termination of the doxycycline application (Fig. 4)
.
|
Tightly controlled expression of human RCA in white blood cells
To analyze the regulatory capacity of the NTA cassettes, we studied the expression patterns of DAF and CD59 by flow cytometry on white blood cells collected from seven double-transgenic animals (L13xL20), and 2 single-transgenic (1 DAF, 1 CD59) pigs. No significant expression of DAF or CD59 was found in lymphocytes (resting and concanavalin A (Con A)-activated cells) from the two single-transgenic pigs. (Fig. 5
A). Absence of DAF and CD59 was also typical for resting cells from double-transgenic animals (Fig. 5B
), although in one animal a small number of CD59 expressing cells were detected (data not shown). However, expression of DAF and CD59 could be induced in cells from all double-transgenic animals by activation with ConA. The intensity of ConA-mediated up-regulation of the two molecules was asymmetric showing strong expression of CD59 and weak but significant up-regulation of DAF. Expression of both molecules could be terminated by culturing the cells with tetracycline for 48 h (Fig. 5B
), again providing evidence for a tight control of gene expression from these cassettes.
|
Identification of CpG islands and determination of the methylation status
Epigenetic mechanisms were hypothesized to be causatively involved in the asymmetric expression of the two RCA-NTA cassettes in double-transgenic animals and the muscle fiber-specific expression in single-transgenic pigs. Bioinformatic analysis of the DAF-NTA cassette revealed that the minimal CMV promoter within PtTA and the 5'coding region of the DAF cDNA together contained a putative CpG island (Fig. 6
A). In contrast, no CpG island was predicted for the PtTA and the 5'coding region of the CD59 gene (Fig. 6A
).
|
Genomic DNA from muscle biopsies of single and double-transgenic pigs was isolated, and analyzed by bisulfite sequencing to assess the methylation status of the NTA-promoter regions in the pig genome.
The promoter region of the CD59-NTA cassette was completely methylated in single-transgenic animals, while in double-transgenic animals only 
50% of the CpG dinucleotides were fully methylated (Fig. 6B
). In contrast, this region was essentially nonmethylated in CD59-NTA cassettes from stably transfected STE cells.
An almost complete methylation of the CpG dinucleotides around the transcription initiation site of the DAF-NTA cassette was discovered either in single or double-transgenic animals (Fig. 6C
). Primary fibroblasts isolated from biopsies also maintained this methylation pattern (Fig. 6C
). For comparison, the methylation status of the DAF-NTA cassette was determined in a pool of stably transfected NIH 3T3 cells and in two cell clones, which expressed high levels of the DAF-NTA construct. These cells possessed a relative low concentration of methylation (Fig. 6C
). Similarly, stably transfected porcine STE cells showed weak methylation of the promoter regions (Fig. 6C
). Thus an inverse correlation between the observed gene expression and the methylation status of these cassettes is obvious. These data indicate that the PtTA is prone to methylation in transgenic pigs, in particular in combination with the 5'coding region of the human DAF gene.
DISCUSSION
Here, we report the first conditional and broad transgene expression in a domestic animal species using an autoregulative tet-off system. The methylation of transgene promoters correlated well with expression levels, suggesting de novo methylation as a crucial mechanism for expression of RCA-NTA constructs in these pigs. The bicistronic, autoregulative cassettes were designed to be compatible with stable and widespread expression and carried either CD59 or DAF, followed by an IRES and the tTA coding sequence. The target genes were chosen, since human CD59 and DAF are critically involved in the suppression of the hyperacute rejection of xenotransplanted porcine organs (35)
. For autoactivation, a minimum basal expression of the minimal promoter is required. The original axiome of the tet system predicts that in the absence of tetracycline, tTA binds to the tet operator sequences and activates transcription (5
, 6
, 36)
. The presumed autoregulated function of the NTA cassettes was found to be fulfilled in NIH3T3 and porcine STE cells.
Unexpectedly, transgenic pigs carrying one NTA cassette displayed confined expression in single striated muscle fibers, irrespective of the transgene or the integration site. Mosaic expression patterns have also been described in tet-transgenic mice and, in some cases, the highest expression levels were found in skeletal muscle (7
, 16
, 18)
. In the present study, confinement to the muscle compartment could be overcome by crossbreeding selected double-transgenic pigs carrying both the DAF-NTA and CD59-NTA cassettes. This resulted in a widespread expression of CD59 in muscle, liver, pancreas, lung and other tissues, whereas DAF expression remained restricted to individual muscle fibers.
The lack of a widespread expression of DAF/NTA correlated with the methylation of critical parts of the tTA dependent promoter and neighboring sequences and is thus epigenetic in nature. Methylation is usually associated with additional changes in histone conformation and chromatin structure, rendering the DNA region into a compressed state and leading to gene silencing (37
38
39)
.
Although the CD59-NTA transgene did not possess a predicted CpG island, bisulfite sequencing revealed that the CpGs around the transcription start site were fully methylated in single-transgenic animals and that approximately half of the CpG dinucleotides were methylated in double-transgenic animals. Studies with Epstein-Barr virus have shown that even methylation of few sites is sufficient to suppress transcription of a promoter, and experiments with episomes have demonstrated that methylation of as few as 7% of the CpG sites can induce quiescence of a gene locus (40
, 41)
. Why individual skeletal muscle fibers showed a relaxed transgene expression control of the NTA cassettes in single-transgenic animals remains to be investigated. We cannot rule out that simple gene dose effects led to the widespread expression in double-transgenic animals. However, we assume that unknown molecular mechanisms during critical time windows have resulted in the observed widespread expression in double-transgenic pigs. It is well known that genomic methylation and histone modifications undergo major changes during development (42)
.
Our data indicate that de novo methylation may interfere with proper expression of tetracycline sensitive promoters with significant implications for the generation of tet-transgenic animals as most of the employed promoter sequences are based on the CMV minimal promoter. Investigations of DMNT1 over-expressing cell lines favor the concept that the sequence context might play a crucial role in the methylation sensitivity of CpG islands (43)
. This might explain the high degree of variability in activator-dependent target gene expression requiring screening for appropriately responding mouse lines, as well as the here observed asymmetric expression in pigs. Along this line it is interesting that CpG depleted CMV promoter constructs gave rise to long term expression in vivo (44)
.
Effective conditional transgene expression could be demonstrated in transgenic pigs and in white blood cells. Feeding of transgenic pigs with a daily dose of 3.3 mg doxycycline per kg body weight terminated transgene transcription in vivo. Re-expression of human RCAs after termination of doxycycline application was resumed by autoactivation after 8 wk. Whether this delayed re-expression is due to a sluggish clearance of doxycycline (15)
or the architecture of the autoregulative expression cassette remains to be determined. At present we do not have evidence for specific metabolic features or a tetracycline reservoir in the pig that could explain the long interval from termination to re-expression of the cassette. Measurements of doxycyline levels in porcine plasma indicated a rapid decline after the application. Analysis of the fodder for doxycycline contaminations did not reveal any doxycycline residues. Overall, these data show that the regulation of the RCA-NTA cassettes is tight, once the expression is turned off by doxycycline. Probably, time consuming stochastic events re-express RCA-NTA cassettes and induce autoactivation. However, in cell culture RCAs were re-expressed within 6–10 d after withdrawal of doxycycline.
In conclusion, we have generated transgenic pigs bearing a bicistronic tetracycline-responsive cassette and demonstrated tightly controlled expression in a large animal model for the first time. The usage of the autoregulative bicistronic cassette supports efficient single step transduction, which is of utmost importance in view of the long generation intervals in large domestic animals. Expression of RCAs and transactivator did not compromise health status and fertility of transgenic pigs. Design of the next generation of expression cassettes will take into account potential methylation prone sequences and should thus be compatible with true ubiquitous transgene expression already in the F0-generation. The use of optimized autoregulative constructs in combination with improved gene transfer techniques, such as somatic nuclear transfer or lentiviral vectors will pave the way toward precisely controlled transgene expression in farm animals for biomedical and agricultural application (45)
.
ACKNOWLEDGMENTS
The critical discussions and comments by J.W. Carnwath are gratefully acknowledged. The authors thank K-G. Hadeler, H.H. Döpke (Mariensee) and L. Schöberlein (Leipzig) for expert help with the muscle biopsy device and A. Saalmüller (Wien) for gift of the STE cell line. The excellent technical assistance of A. Kanwischer (Hannover) and W. Mysegades, the introduction into bisulfite sequencing by C. Gebert, as well as the expert animal caretaking by H. Hornbostel and T. Peker are kindly acknowledged. This work was supported by grants of the DFG (SFB 265, Forschergruppe 535) and BMBF to H.N. Support of the Kuratorium für Dialyze und Nierentransplantation to K.W. is acknowledged.
Competing financial interests: The authors declare that they have no competing financial interests in relation to this paper.
FOOTNOTES
1 Present address: Envelope retrovirale et ingeniérerie retrovirale, INSERM 758, ENS de Lyon, France. 
Received for publication November 15, 2005. Accepted for publication February 14, 2006.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  F. Park Lentiviral vectors: are they the future of animal transgenesis? Physiol Genomics, October 19, 2007; 31(2): 159 - 173. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
