Transgenic Research 9: 81–89, 2000. © 2000 Kluwer Academic Publishers. Printed in the Netherlands.
81
Effect of flanking matrix attachment regions on the expression of microinjected transgenes during preimplantation development of mouse embryos Alfonso Guti´errez-Ad´an∗ & Bel´en Pintado Departamento de Reproducci´on Animal y Conservaci´on de Recursos Zoogen´eticos, INIA, Ctra de la Coruña km 5,9 Madrid 28040, Spain Received 13 July 1999; revised 11 October 1999; accepted 5 November 1999
Key words: GFP, MAR, preimplantation embryo, selection, transgenic
Abstract The efficiency of transgenic animal production would increase if microinjected embryos with a successfully integrated transgene could be identified prior to transfer. It is possible to detect microinjected DNA in embryos. However, no reliable system is able to distinguish between transgenes merely present as extrachromosomal DNA and those that have been integrated into chromatin. The experiments reported here were designed to determine if the inclusion of matrix attachment regions (MARs) would enhance the efficiency of transgenic embryos identification using a selection scheme based on the expression of green fluorescent protein (GFP). Pronuclei of mouse embryos were microinjected with GFP reporter gene under the control of three different promoters and flanked or not by three different MAR elements. Transgene expression profiles were followed by direct visualization of GFP in cultured microinjected embryos. Embryos at different developmental stages were classified according to their GFP expression and groups with the same expression pattern were transferred into oviducts of pseudopregnant female mice. Fetuses were collected between days 12–15, and their genomic DNA was purified and analyzed to detect transgene integration. We did not find any statistically significant difference between the percentage of transgenic fetuses produced from GFP-positive or GFP-negative embryos transferred at 4-cell, morula, or blastocyst stage. However, when MAR elements were included in the construct, we found that GFP-positive embryos transferred at the 2-cell stage produced a significantly higher percentage of transgenic fetuses than GFP-negative embryos, but MAR sequences did not completely eliminate false positives.
Introduction Pronuclear microinjection of foreign DNA is the most commonly used technique to produce transgenic animals (Gordon et al., 1980). Unfortunately, with this procedure only about 10–20% of mice born from microinjected embryos are transgenic. The efficiency in farm animals is lower, varying from 0.1 to 4.45% (Pursel & Rexroad, 1993). Due to the high costs and the long gestation periods of large animals, attempts have been made to detect transgene integration prior ∗ Author for correspondence: Tel.: 34 91 3474026; Fax: 34 91
549 0956; E-mail:
[email protected]
to embryo transfer. One such approach was based on PCR analysis of embryo biopsies to detect the transgene, but this procedure yielded a high number of false positives presumably due to the coexistence of unintegrated transgenes in embryonic cells (Burdon & Wall, 1992; Horvat et al., 1993). Another approach used an antibiotic resistance selectable marker in the transgene to select transgenic embryos before transferring them to recipients, but this approach was also unsuccessful (Tada et al., 1995; Bondioli & Wall, 1996). It has also been suggested that the use of detection of luciferase activity by bioluminescence imaging could permit rapid detection of a transgene before embryo transfer (Menck et al., 1998; Nakamura et al.,
82
Figure 1. Constructs of microinjected transgenes. Plasmid IFN-CMV, APO-CMV, and HSP-CMV differed from plasmid CMV on the inclusion of MAR elements derived respectively from human elongation factor 1α, human apolipoprotein B, and Drosophila hsp-70 genes before the promoter. Plasmid IFN-EF and IFN-CMV-EF differed respectively from plasmid EF and CMV-EF on the inclusion of MAR elements derived from human elongation factor 1α before the promoter. Abbreviations used: GFP, green fluorescent protein; CMV, citomegalovirus; CMV-IE, enhancer element of CMV promoter; EF-1α, human elongation factor 1α; SV40, simian virus.
1998; Thompson et al., 1995), but this approach relies on the use of sophisticated equipment, and till date, the technique has not substantially improved transgenic efficiency. Recently, monitoring green fluorescent protein (GFP) has been proposed (Takada et al., 1997) as a means for the selection of transgenic embryos. One advantage of GFP is that no substrate is required for detection, so it is a suitable marker for selection of those embryos expressing the microinjected transgene. However, it is important to point out that it is not possible to distinguish between expression from integrated or non-integrated transgenes with this procedure. Therefore, both false positives and false negatives are observed. DNA sequences containing scaffold attachment regions or matrix attachment regions (MARs) have been reported to improve expression of transgenes in eukaryotic cells, but only when those are integrated into the genome. These same sequences showed a neutral or even a negative effect on expression in transiently transfected cells (Bode & Maass, 1988; Klehr et al., 1991; Poljak et al., 1994; Kalos & Fournier, 1995;
Wang et al., 1996). In addition, increased expression has been observed in MAR flanked integrated transgenes, but not when the transgenes were transiently expressed (Thompson et al., 1994, 1995). Three MAR elements have been reported to exhibit these characteristics: an MAR element isolated from the human-interferon gene domain boundaries, hIFN-β (Bode & Maass, 1988; Klehr et al., 1991); an MAR isolated from human apolipoprotein B gene, hapoB (Kalos & Fournier, 1995); and an MAR from Drosophila heat shock protein-70 gene, dHSP-70 (Poljak et al., 1994). Our aim was to improve the selection strategy based on GFP expression by combining it with the reported enhancer effect of MAR sequences on stably integrated transgenes. We hypothesized that the number of false positives from expression of unintegrated transgenes would decrease when MAR sequences were included in the transgene construct. We have used three different promoters to drive the expression of GFP in early embryos. A viral human cytomegalovirus (CMV) immediate early promoter, which in-
83 duces high expression in specific cell types, including the ovary and testes (Baskar et al., 1996), a non-viral house keeping gene, 2.3 kb of the promoter of human elongation factor 1α (hEF-1α), and a combination of both promoters, 1 kb of hEF-1α promoter which has been used in combination with the enhancer element of CMV promoter to drive the expression of GFP in preimplantation mouse embryos (Takada et al., 1997). These three GFP expression vectors were combined with or without one of the three different MAR sequences mentioned above in order to test if the presence of MARs would improve the accuracy of transgenic embryo selection. Materials and methods Transgene construction The transgene constructs used in this study are summarized in Figure 1. The plasmid CMV containing the human CMV immediate early promoter and the enhanced GFP gene was obtained from Clontech (pEGFP-N1, Clontech Laboratories, Inc., Palo Alto, CA, USA). The polilinker of pEGFP-N1 was eliminated and a Bcl I-EcoR I adapter was inserted in the Afl III site. To make plasmids IFN-CMV, APO-CMV, and HSP-CMV one of three different MAR elements were subcloned before the CMV promoter. Construct IFN-CMV was obtained by subcloning a 3-kb Blg II MAR element isolated from the human β-interferon gene domain boundaries (hIFN-β) into the Bcl I adapter. To obtain APO-CMV construct, a 1-kb BamH IEcoRI MAR element isolated from the human apolipoprotein B gene (hapoB) was subcloned into the Bcl I-EcoRI adapter. Plasmid HSP-CMV was obtained by subcloning a 0.96-kb BamHI-EcoRI MAR element isolated from the Drosophila HSP-70 gene (dHSP70) into the Bcl I-EcoR I adapter. The IFN, APO and HSP MAR elements were kindly supplied by J. Bode (Gesellschaft für Biotechnologische Forschung, Braunschweig, Germany), M. Kalos (Fred Hutchinson Cancer Research Center, Washington, USA) and L. Poljak (Centre National de la Recherche Scientifique, Toulouse, France) respectively. Plasmid EF-GFP was kindly supplied by Kevin Wells (Gene Evaluation and Mapping Laboratory, USDA, Beltsville, USA). Plasmid IFN-EF contained a 2-kb Hind III MAR element isolated from the hIFN-β gene domain boundaries that was subcloned into the Hind III site that preceded the hEF-1α promoter. Plasmid CMV-EF was kindly supplied by T. Takada (Takada et al., 1997). To make the
plasmid IFN-CMV-EF a 3-kb Blg II MAR element isolated from the hIFN-β gene domain boundaries was subcloned into the BamH I site. All constructs contained the SV40 polyadenylation signal sequences. Microinjection and embryo culture All gene constructs linearized by a unique restriction endonuclease digestion were purified using NaCl gradient (Fink, 1991). The gradients were fractionated and those fractions containing the construct were desalted using NAP-25 columns (Pharmacia Biotech, Barcelona, Spain). Each linearized transgene fragment was diluted in 10 mM Tris–HCl (pH 7.5) and 0.1 mM EDTA to a concentration of approximately 5 ng/µl. Transgenes were microinjected into pronuclearstage mouse embryos obtained from superovulated (C57BL6×CBA) F1 mice 19 h after the administration of hCG. Microinjected embryos were cultured in vitro in KSOM medium (Erbach et al., 1994) in a humidified atmosphere of 5% CO2 in air, until they were analyzed for GFP expression and transferred into pseudopregnant recipient mice. Fluorescent analysis of preimplantation embryos and embryo transfer Embryos were placed individually in 2–3 µl microdrops of PBS medium under mineral oil (Sigma). GFP expression in each embryo was examined under a fluorescent microscope (Nikon, Optiphot-2), using a standard FITC filter set (BP470-490 and BA515) at 100x magnification. Embryos were considered as positive when a clear and distinct green colour was visible in at least 50% of the blastomeres regardless of its intensity. Embryos expressing GFP (GFP-positive) were separated from GFP-negative embryos. GFP-positive and GFP-negative embryos were transferred separately into the oviducts of pseudopregnant female mice for further development. Twelve days after blastocyst transfers, or 15 days after 2-cell embryo transfers, recipients were killed, and both fetuses and reabsortions were recovered. Genomic DNA was purified and analyzed for detection of transgene integration. Slot blot analysis Transgenic animals were identified by slot blot analysis of the DNA extracted from implanted and reabsorbing fetuses. Blots were probed with a digoxigenindUTP (DIG/GeniusTM System, Boehringer Mannheim) 340 bp PCR fragment (described below) of the
84 GFP DNA. Primers used in the PCR analysis were GFP1 (50 -TGA ACC GCA TCG AGC TGA AGG G30 ) and GFP2 (50 -TCC AGC AGG ACC ATG TGA TCG C-30 ), which specifically amplified a 340 bp GFP DNA. Amplification was carried out in a total volume of 25 µl (1 of PCR mix containing 1 U Taq polymerase (Promega), 2.5 µl 10x buffer (Promega), 100 µM each dNTP, 0.1 µM each primer and 2.5 mM MgCl2 . The PCR reaction was performed for 30 cycles at 92◦ C (35 s), 60◦ C (45 s) and 72◦ C (35 s). In the first PCR cycle denaturation was at 92◦ C for 2 min. The CMV-EF construct contained the GFP (S65T), which exhibits less intense expression (Clontech Laboratories) and has a different sequence to the other GFP present in the above construct. In this case the system of identification was similar except for the primers used were GFPS65T-1 (50 -GAG TGC CAT GCC CGA AGG TTA TG-30 ) AND GFPS65T-2 (50 -GGC AGA TTG TGT GGA CAG GTA ATG G-30 ). Experiment 1 In order to evaluate the effects of the presence of MAR sequences on expression of GFP induced by CMV, hEF-1α, or the combination of both promoters during preimplantation mouse embryo stage, constructs CMV, IFN-CMV, EF, IFN-EF, CMV-EF, and IFN-CMV-EF (Figure 1) were microinjected into pronuclear stage mouse embryos that were then cultured to the blastocyst stage. GFP expression was monitored at 2-cell, 4-cell, morula, and blastocyst stages of development. Data was analyzed by Chi-square test.
purified and analyzed to detect transgene integration. Data was analyzed by Chi-square test.
Results MARs effect on expression of GFP induced by CMV, hEF-1α, and a combination of both promoters during preimplantation mouse embryo development As the GFP protein is relatively stable (Bokman & Ward, 1981) it persists in the embryo for a relatively long time. This attribute makes it more difficult to discern transient from integrated expression in microinjected embryos. Thus, in order to determine the onset of GFP expression in positive blastocysts, embryos were checked through out development. We evaluated GFP expression at the 2-cell, 4-cell, morula and blastocyst stages. When the CMV promoter containing constructs without a MAR (CMV, Figure 1) was microinjected, GFP was detectable in 20% of 2-cell embryos. GFP-positive embryos increased to 49% of morulae and 78% of blastocysts (Figure 2A).
Experiment 2 To assess MAR effects on selection of preimplantation transgenic embryos expressing GFP induced by CMV, hEF-1α, or combination of both promoters, construct CMV, IFN-CMV, APO-CMV, HSP-CMV, EF, IFN-EF, CMV-EF, and IFN-CMV-EF (Figure 1) were microinjected into pronuclear stage embryos. After culture to the 2-cell stage GFP-positive and GFP-negative embryos were transferred into recipient females. Other injected embryos were further cultured to 4-cells, morula, and blastocyst stages, segregated based on GFP selection and then transferred into pseudopregnant recipient mice. In order to evaluate the transgenic production rate when no selection was performed, a control group of embryos injected with the IFN-CMV construct were also transferred at the 2-cell stage without selection. Fetuses were collected between days 12–15, and their genomic DNA was
Figure 2. Effect of MAR elements on the pattern of preimplantation transgene expression (A, B, C and D construct are described in Figure 1). After microinjection of GFP transgene with (B and D) or without (A and C) a MAR element, the expression of GFP in different developmental stages of preimplantation embryos were identified using a fluorescent microscopy.
85 Table 1. Efficiency of transgenic mice production by the MAR-GFP selection method with CMV promoter Construct
Stage of transfer
GFP expression
Number transferred
Number tested
CMV
2-cells
Positive Negative Positive Negative Positive Negative
37 20 20 11 40 28
12 12 7 5 22 23
1 (8)b 3 (25)a,b 1 (14)a,b 1 (20)a,b 3 (14)a,b 5 (22)a,b
Positive Negative Positive Negative Positive Negative Positive Negative
42 69 20 40 34 14 46 85
19 25 9 37 32 6 27 35
8 (43)a 5 (20)a,b 3 (33)a,b 6 (22)a,b 7 (22)a,b 0 (0)a,b 4 (15)b 4 (11)b
Positive Negative Positive Negative
18 30 27 50
14 23 17 32
3 (21)a,b 2 (9)b 2 (12)b 3 (10)b
Positive Negative Positive Negative
30 35 34 34
12 8 15 20
3 (25)a,b 1 (13)a,b 1 (7)b 3 (15)a,b
Positive Negative Positive Negative
90 124 108 169
45 56 59 87
Not analyzed
114
39 fetuses
4-cells Blastocyst
IFN-CMV
2-cell 4-cells Morulae Blastocyst
APO-CMV
2-cells Blastocyst
HSP-CMV
2-cells Blastocyst
Cumulative MAR-CMV
2-cell Blastocyst
IFN-CMV
2-cell
Number transgenic (%)
14 (31)a 8 (11)b 7 (12)b 8 (11)b 8 (21)a,b
Different superscripts within the number of transgenics indicate statistically significant differences (a6 =b, p 6 0.05, chi-square test).
When an MAR element from the hIFN-β gene was included in the construct (IFN-CMV, Figure 1), there was a decrease in the percentage of GFP-positive embryos (35% and 54% for morula and blastocysts, respectively; Figure 2B). With the construct containing the hEF promoter without MAR (EF, Figure 1), GFP was detected in 6% of 2-cell embryos, 40% of 4-cell embryos, 63% of morulae and 80% of blastocysts (Figure 2C). The pattern of expression differed from that of the CMV promoter. The use of the EF construct resulted in fewer GFP-positive embryos at the 2-cell stage, and in a higher number of GFP-positive embryos at the 4-
cell stage. Unlike constructs with the CMV promoter, no significant decrease in the percentage of GFP expression was observed when an MAR element from hIFN-β gene was included in the construct (IFN-EF, Figure 2D). The combination of both hEF and CMV promoters (CMV-EF, Figure 1), showed a GFP expression pattern similar to the pattern seen with EF constructs. Less than 8% of embryos were GFP-positive at the 2-cell stage, 32% of GFP-positive embryos were detected at the 4-cell stage, 54% of morulae, and 65% of blastocyst (Figure 2E). As for the EF promoter, the inclusion of an MAR element form hIFN-β gene (IFN-
86
Figure 3. Effect of MAR elements on selection of preimplantation embryos expressing GFP microinjected with CMV vs MAR-CMV constructs (IFN-CMV, APO-CMV, HSP-CMV). The figure shows the cumulative percentage (±SD) of transgenic fetuses obtained after transferring embryos at 2-cells, 4-cells, blastocyst and combined 4-cells+ morulae+ blastocyst stages (Table 1). P < 0.05 (Chi-square test).
CMV-EF, Figure 1), did not decrease significantly the percentage of GFP-positive embryos (45% and 54% for morula and blastocysts, respectively; Figure 2F). MAR effect on selection of preimplantation transgenic embryos expressing GFP induced by CMV promoter Embryos microinjected with the CMV, APO-CMV, IFN-CMV, and HSP-CMV transgenes (Figure 1) were cultured for 24, 48, 72, and 96 h. At those times, embryos were separated into GFP-positive and GFPnegative groups and all the GFP-positive and a control group of GFP-negative embryos were transferred separately into oviducts of pseudopregnant female mice. The remaining GFP-negative embryos were cultured for another 24 h and then those embryos which started to express GFP at the next developmental stage (4-cell, morula or blastocyst stage) were also transferred and healthy and reabsorbing fetuses were collected as previously described. PCR and slot blot analysis of fetuses did not reveal a significant difference between the percentage of transgenic animals from transferred GFP-positive or GFP-negative embryos (Table 1). Furthermore, no statistical differences were found between the proportion of the transgenic animals obtained from reabsorbing fetuses (56%) or from healthy ones (44%) (Chi-square analysis). Overall, MAR sequences did not improve the proportion of transgenic mice obtained from GFP-positive
blastocysts. However, when the MAR element from the hIFN-β gene was included in the construct, a significantly higher proportion of transgenic fetuses was obtained from 2-cell GFP-positive embryos (43% vs 8%; p < 0.05). This percentage of transgenics obtained decreased when GFP-positive 4-cell, morula or blastocyst embryos were transferred (43%> 33%> 22%> 15%, respectively, Table 1). Also, when the MAR elements from hapoB gene or dHSP70 gene were included, a higher proportion of transgenic fetuses was obtained after transferring the 2-cell GFP-positive embryos. Using the additive property of chi-square we grouped the results from transfers of 2-cell GFP-positive embryos microinjected with APO-CMV, IFN-CMV and HSP-CMV. This analysis indicated a significant increase in the number of transgenic animals after transferring 2-cell GFP-positive embryos when the transgene was combined with MAR (P < 0.05, Figure 3). In order to compare the real efficiency of selection on transgenic production, a control group consisting of microinjected embryos with the IFN-CMV construct was transferred at the 2-cell stage without selection. Eight of the 39 fetuses recovered were transgenic (21%). This efficiency did not differ statistically from the 43% transgenic production obtained after GFP selection (p = 0.08; Table 1). MAR effect on selection of preimplantation transgenic embryos expressing GFP induced by EF-1α promoter The results of selection based on expression of EF and IFN-EF (Figure 1) are summarized in Table 2. After analysis of transgene integration in fetuses, we found that selection for GFP-positive embryos did not increase significantly the proportion of transgenic mice. Even though a lower number of transgenic animals were achieved after transfer of GFP-negative embryos, differences were not significant. Similar results were obtained with the selection based on expression of CMV-EF and IFN-CMV-EF (Figure 1, Table 3).
Discussion Until recently (Schnieke et al., 1997), transfer of foreign genes into farm animals was performed by microinjection of DNA into the pronuclei of fertilized eggs. Pronuclear microinjection in farm animals is labor intensive, costly and inefficient. Compared
87 Table 2. Efficiency of transgenic mice production by the MAR-GFP selection method with the EF promoter. Microinjected embryos were cultured to 2-cell or blastocyst stage and were transferred according to expression or non expression of GFP Construct
Stage of transfer
GFP expression
Number transferred
Number tested
Number transgenic (%)
EF
Blastocyst
Positive M+B Positive B Negative
56 22 29
28 7 14
5 (18) 1 (14) 1 (7)
IFN-EF
2-cell
Positive Negative Positive M+B Positive B Negative
12 8 90 27 51
8 6 42 15 31
1 (13) 1 (17) 9 (21) 3 (20) 3 (10)
Blastocyst
Positive M+B: Expressing GFP at morula and blastocyst stages. Positive B: Expressing GFP only at blastocyst stage. No significant differences were found between percentages of transgenic progeny (p < 0.05; chisquare analysis). Table 3. Efficiency of transgenic mice production by the MAR-GFP selection method with CMV-EF promoter. Microinjected embryos were cultured to blastocyst stage and were transferred according to expression or non expression of GFP Construct
Stage of transfer
GFP expression
Number transferred
Number tested
Number transgenic (%)
CMV-EF
Blastocyst
Positive M+B Positive B Negative
37 35 42
18 17 37
4 (22) 5 (29) 4 (11)
IFN-CMV-EF
Blastocyst
Positive M+B Positive B Negative
45 39 50
21 18 40
5 (24) 6 (33) 4 (10)
Positive M+B: Expressing GFP at morula and blastocyst stages. Positive B: Expressing GFP only at blastocyst stage. No significant differences were found between percentages of transgenic progeny (p < 0.05; chi-square analysis).
to the success rate of transgenic mice, the success rate in transgenic farm animals is low, probably due to the poor viability of embryos after DNA microinjection and the low integration rate of the transgene (Wall, 1996). The identification of preimplantation embryos that have successfully integrated the microinjected transgene into their genome would substantially reduce the costs and increase the efficiency of transgenic farm animal production. With this purpose in mind, we tested the described role of MAR elements to stimulate expression of a heterologous construct in stably transformed cells, applying it to the identification of integrated transgenes after microinjection in preimplantation embryos. In the present study, we have used two different promoters and a combination of both in order to drive
the expression of GFP in early embryos. GFP expression was detected in preimplantation mouse embryos using the three constructs, with slightly different expression patterns. With the hEF promoter, GFP is expressed in a very small percentage (6%) of 2-cell embryos. Probably this is because this promoter needs the transcriptional activation of embryos that takes place at the 2-cell stage in mouse. After this stage, the hEF promoter produces a higher number (40% 4cell embryos) of GFP-positive embryos, (Figure 2). The third construct, consisting in a short hEF promoter and an enhancer element of the CMV promoter, drove the expression of GFP in a similar pattern to the hEF construct but with a lower level of expression. This difference may have been a consequence of the kind of GFP used in the third construct. We
88 used the exact plasmid which had provided a positive selection scheme in preimplantation embryos (Takada et al., 1997) but GFP-S65T has lower level of expression when compared to the enhanced GFP used in the two other constructs. In contrast to those reports in which selection was successful at the blastocyst stage (Takada et al., 1997; Thompson et al., 1995), we did not find a correlation between GFP expression and transgenic fetuses. We were unable to differentiate between transient expression and expression of integrated transgenes at this developmental stage, obtaining similar percentages of transgenic animals after transferring GFP-positive and GFP-negative embryos. However, when 2-cell GFPpositive embryos carrying an MAR element and the CMV promoter were transferred, we obtained a higher proportion of transgenic fetuses. Our disagreement with previous reports (Takada et al., 1997) may be due to the selection system followed. We selected embryos as GFP+ when at least 50% of cells showed distinct fluorescence at 100x regardless intensity. We chose to include embryos not expressing GFP in all their blastomeres because it has been demonstrated that most transgenic founders are mosaic (Whitelaw et al., 1993), showing that the majority of DNA injected integrates after the first round of chromosomal DNA replication. This might lead to an overestimation of positive embryos, but it would avoid false negatives; but these were still present in our results, showing that a significant percentage of transgenic animals failed to express GFP at preimplantation stages. We did not select embryos based on intensity of GFP expression because our aim was not directed to evaluate expression level but integration percentage. Our results show that episomal expression of the construct is always present interfering with the reliability of the technique, and only at the 2-cell stage, were the MARs sequences able to provide a significant increase in the proportion of transgenics obtained. When transgenes are microinjected into one of the pronuclei, a small number (range between 200–500) of identical DNA molecules are introduced. It has been proposed that this DNA forms extrachromosomal concatemers (arrays), by rounds of homologous recombination, and the concatemer molecules integrate into the chromosomes, more or less at random, by illegitimate recombination. (Bishop, 1996). In cell culture transfections, a burst of transcription is usually observed before integration (transient expression). This has been ascribed to the ability of DNA to recruit transcription factors before its final assembly into
nucleosomes. At this stage MAR sequences generally have adverse effects, probably because the attachment of a soluble template to the scaffold is disadvantageous (Bode & Maass, 1988; Stief et al., 1989; Klehr et al., 1991; Kalos & Fournier, 1995). We have found that this effect is transient and/or saturable in earlymicroinjected embryos, because after the 4-cell stage, no positive selection was achieved by transferring GFP-positive embryos. Our results show that it is impossible to determine if GFP expression comes from an integrated or non-integrated transgenes. More studies are necessary to identify DNA elements that may allow a selective expression of microinjected DNA after stable integration into the genome, without being confused with transient expression.
Acknowledgements We thank Kevin Wells (Gene Evaluation and Mapping Laboratory, USDA, Belsville, USA) for providing the EF-GFP construct; J.Bode (Gesellschaft für Biotechnologische Forschung, Braunschweig, Germany) for providing us with the MAR elements isolated from the human β-interferon gene domain boundaries; M. Kalos (Fred Hutchinson Cancer Research Center, Washington, USA) for providing us with the MAR elements from human apolipoprotein B gene; L. Poljak (Centre National de la Researche Scientifique, Toulouse, France) for providing us with the MAR elements isolated from Drosophila HSP-70 gene and T. Takada for providing us with the plasmid CMV-EFGFP. This work has been supported by the Comunidad Autónoma de Madrid.
References Baskar JF, Smith PP, Nilaver G, Jupp RA, Hoffmann S, Peffer NJ, Tenney DJ, Colberg-Poley AM, Ghazal P and Nelson JA (1996) The enhancer domain of the human cytomegalovirus major immediate-early promoter determines cell type-specific expression in transgenic mice. J Virol 70: 3207–3214. Bishop JO (1996) Chromosomal insertion of foreign DNA. Reprod Nutr Dev 36: 607–618. Bode J and Maass K (1988) Chromatin domain surrounding the human interferon-beta gene as defined by scaffold-attached regions. Biochemistry 27: 4706–4711. Bokman SH and Ward WW (1981) Renaturation of Aequorea green fluorescent protein. Biochem Biophys Res Commun 101: 1372– 1380.
89 Bondioli KR and Wall RJ (1996) Positive selection of bovine embryos in culture. Theriogenolgy 45: 345. Burdon TG and Wall RJ (1992) Fate of microinjected genes in preimplantation mouse embryos. Mol Reprod Dev 33: 436–442. Erbach GT, Lawitts JA, Papaioannous VE and Biggers JD (1994) Differential growth of the mouse preimplantation embryo in chemically defined media. Biol Reprod 50: 1027–1033. Fink PS (1991) Using sodium chloride step gradients to fractionate DNA fragments. BioTechniques 10: 437–440. Gordon J, Scangos G, Plotkin D, Barbosa J and Ruddle F (1980) Genetic transformation of mouse embryos by microinjection of purified DNA. Proc Natl Acad Sci USA 77: 7380–7384. Horvat S, Medrano JF, Behboodi E, Anderson GB and Murray JD (1993) Sexing and detection of gene construct in microinjected bovine blastocysts using the polymerase chain reaction. Transgenic Res 2: 134–140. Kalos M and Fournier REK (1995) Position-independent transgene expression mediated by boundary elements from the apolipoprotein B chromatin domain. Mol Cellular Biol 15: 189–207. Kim DW, Uetsuki T, Kaziro Y, Yamaguchi N and Sugano S (1990) Use of the human elongation factor 1α promoter as a versatile and efficient expression system. Gene 91: 217–223. Klehr D, Maass K and Bode J (1991) Scaffold-attached regions from the human interferon ( domain can be used to enhance the stable expression of genes under the control of various promoters. Biochemistry 30: 1264–1270. Menck MC, Mercier Y, Campion E, Lobo RB, Heyman Y, Renard JP and Thompson EM (1998) Prediction of transgene integration by noninvasive bioluminiscent screening of microinjected bovine embryos. Transgenic Res 7: 331–341. Nakamura A, Okumura J-I and Muramatsu T (1998) Quantitative analysis of luciferase activity of viral and hybrid promoters in bovine preimplantation embryos. Mol Reprod Dev 49: 368–373. Poljak L, Seum C, Mattioni T and Laemmli U (1994) SARs stimulate but do not confer position independent gene expression. Nucleic Acids Res 22: 4386–4394.
Powell AM, Bondioli KR and Rexroad CE (1996) The sheep uterus as host for IVM-IVF bovine embryos from day 7– day 15. Theriogenology 45: 217. Pursel VG and Rexroad CE (1993) Status of research with transgenic farm animals. J Anim Sci 71(3): 10–15. Schnieke AE, Kind AJ, Ritchie KM, Scott AR, Ritchie M, Wilmut I, Colman A and Campbell HS (1997) Human factor IX transgenic sheep produced by transfer of nuclei from transfected fetal fibroblasts. Science 278: 2130–2133. Stief A, Winter DM, Stratling WH and Sippel AE (1989) A nuclear DNA attachment element mediates elevated and positionindependent gene activity. Nature 341: 343–345. Tada N, Sato M, Hayashi K, Kasai K and Ogawa S (1995) In vitro selections of transgenic mouse embryos in the presence of G418. Transgenics 1: 535–540. Takada T, Lida K, Awaji T, Itoh K, Takahashi R, Shibui A, Yoshida K, Sugano S, and Tsujimoto G (1997). Selective production of transgenic mice using green fluorescent protein as a marker. Nature Biotech 15: 458–461. Thompson EM, Adenot P, Tsuji FI and Renard J-P (1995) Real time imaging of transcriptional activity in live mouse preimplantation embryos using a secreted luciferase. Proc Natl Acad Sci USA 92: 1317–1321. Thompson EM, Christians E, Stinnakre M-G and Renard J-P (1994) Scaffold attachment regions stimulate HSP70.1 expression in mouse preimplantation embryos but not in differentiated tissues. Mol Cellular Biol 14: 4694–4703. Wall RJ (1996) Transgenic livestock: progress and prospects for the future. Theriogenology 46: 57–68. Wang D-M, Taylor S and Levy-Wilson B (1996) Evaluation of the function of the human apolipoprotein B gene nuclear matrix association regions in transgenic mice. J Lipid Res 37: 2117–2124. Whitelaw CBA, Springbett AJ, Webster J, Clark AJ (1993) The majority of G0 transgenic mice are derived from mosaic embryos. Transgenic Res 2: 29–32.