Introduction
The production and application of genetically modified (transgenic) embryos and animals, despite its high cost, is still important. New methods are needed to increase efficiency of transgenesis (gene integration and expression) and decrease the final cost. One of the most promising approaches is to use GFP (green fluorescent protein) reporter genes for preimplantation-stage screening of embryos. Detection of GFP expression is possible using a fluorescence microscope (592 nm wavelength) with no deleterious effect on embryo vitality (Ikawa et al., Reference Ikawa, Kominami, Yoshimura, Tanaka, Nishimune and Okabe1995, Chrenek et al., Reference Chrenek, Vasicek, Makarevich, Jurcik, Suvegova, Bauer, Parkanyi, Rafay, Batorova and Paleyanda2005). On the other hand the controversial aspect of using GFP as a marker for the production of transgenic animal has been reported (Duszewska et al., Reference Duszewska, Lipinski, Piliszek, Slomski, Plawski, Wojdan, Gawrin, Juzwa, Zeyland, Wenta-Muchalska and Reklewski2004).
It was shown that the production of transgenic embryos by microinjection of foreign gene into pronuclei of fertilized eggs decreases survival of microinjected embryos (Makarevich et al., Reference Makarevich, Chrenek, Žilka, Pivko and Bulla2005). The other important factor limiting the efficiency of transgenic animal production is the quality of transgenic embryos resulting in poor development and low transgene integration and expression rates (as a result of random integration) or transgenic mosaicism. In transgenic mosaic embryos or animals not all cells carry the transgene (Wang et al., Reference Wang, Lin, Zhang and Chen2001; Duszewska et al., Reference Duszewska, Lipinski, Piliszek, Slomski, Plawski, Wojdan, Gawrin, Juzwa, Zeyland, Wenta-Muchalska and Reklewski2004).
Biological material developed in in vitro conditions differs in quality and vitality from that developed in vivo. Although the efficiency of the embryo production in vitro increased in the last decade, the quality of these embryos is still different from in vivo derived embryos, particularly in morphology and metabolism (Khurana & Niemann, Reference Khurana and Niemann2000), chromosomal abnormalities (Shi et al., Reference Shi, Dirim, Wolf, Zakhartchenko and Haaf2004) and ultrastructural changes (Popelkova et al., Reference Popelkova, Chrenek, Pivko, Makarevich, Kubovicova and Kacmarik2005). Cleavage-stage arrest of microinjected embryos related to apoptosis was also documented in rabbit embryos (Makarevich et al., Reference Makarevich, Chrenek, Žilka, Pivko and Bulla2005).
The aim of this study was to compare the quality of rabbit transgenic embryos obtained by microinjection of four GFP gene-bearing constructs. Developmental rate, total cell number in hatching blastocyst stage, number of apoptotic cells, diameter of embryos, transgene integration and transgenic mosaicism were examined in our study.
Materials and Methods
Gene constructs
The first GFP gene construct (CMVIE–EGFP, ClaI) consisted of full length (4.7 kb) plasmid pEGFP– N1 (Clontech, USA) linearized by ClaI and used for microinjection into fertilized rabbit eggs. The second gene construct (CMVIE–EGFP, DrdI) contained the same expression cassette CMVIE–EGFP–SV40 poly A on a short 1.99 kb DrdI fragment of pEGFP–N1. The expression cassette of this construct was flanked by short (ca 180 bp) vector backbone sequences on both sides. The third construct (PGK–EGFP) was a NotI linearized 6.3 kb plasmid PGKNEO PGK EGFP (INRA), and the fourth gene construct (CMVIE–hrGFP) containing humanized renilla GFP gene was ScaI linearized 5.0 kb pIRES–hrGFP–2a (Stratagene, USA). All DNA fragments were agarose gel-purified and diluted in endotoxin-free TE buffer for all microinjections at a final concentration of 4 μg/ml.
Embryo manipulation and microinjection
Three days before mating, New Zealand White rabbits were treated with PMSG (Werfaser) followed by hCG (Werfachor) 72 h later. At 19–20 h after mating, the pronuclear stage eggs were flushed with PBS from the oviducts of the animals. After the evaluation of flushed ova, the eggs with both pronuclei were subjected to microinjection in CIM medium + 10% fetal bovine serum (FBS, both from Gibco BRL) using an Olympus microscope equipped with micromanipulation units (Alcatel) and microinjector FemtoJet (Eppendorf). The eggs were fixed by suction with a holding pipette, and 4 μg/ml of the DNA (EGFP) in 1–2 pl was microinjected using air pressure (Pc – compensation and Pi – injection pressure, with injection time) into both pronuclei (Chrenek et al., Reference Chrenek, Vasicek, Makarevich, Jurcik, Suvegova, Bauer, Parkanyi, Rafay, Batorova and Paleyanda2005). Swelling of the pronuclei by 10% indicated successful microinjection. The eggs were cultured in k-DMEM medium supplemented with 10% FBS (Gibco BRL) at 5% CO2 and 39°C up to the morula stage for the analysis of transgene integration and transgenic mosaicism or up to the blastocyst stage for the diameter and number of cells.
The zona pellucida from rabbit transgenic embryo at morula stage was removed by treatment with 0.5% pronase (Sigma, USA). The rabbit zona-free embryos were incubated in PBS (Ca2+- and Mg2+-free medium) for 5 min at 38°C and separated into single blastomeres by gently pipetting through a fine glass pipette (20–30 μm inner diameter, Chrenek & Makarevich, Reference Chrenek and Makarevich2005).
Fluorescence analysis of EGFP expression
The EGFP expression in microinjected embryos at the morula or blastocyst stage was monitored using a Leica fluorescence microscope (excitation filter at 450–490 nm), as described by Chrenek et al. (Reference Chrenek, Vasicek, Makarevich, Jurcik, Suvegova, Bauer, Parkanyi, Rafay, Batorova and Paleyanda2005).
Cell number counting and differential staining
For the cell number determination, the blastocysts were stained for 20 min with 1 μg/ml of Hoechst 33342 (Sigma), mounted on a microslide in Vectashield (Vector Laboratories) and analysed under a Leica fluorescence microscope (excitation filter 340–380 nm) as was early reported by Chrenek et al. (Reference Chrenek, Vasicek, Makarevich, Jurcik, Suvegova, Bauer, Parkanyi, Rafay, Batorova and Paleyanda2005). For the ICM allocation, the blastocysts were differentially stained according to Fouladi-Nashta (Reference Fouladi-Nashta2005). Briefly, embryos were incubated in 0.2% Triton X-100 diluted in PBS with 0.2% BSA for 20 s. After twice washing in PBS solution with 0.2% BSA the embryos were stained in propidium iodide (30 μg/ml) diluted in PBS–BSA for 5 min. Following washing in PBS–BSA the embryos were fixed and stained in 4% paraformaldehyde containing bisbenzimide (Hoechst 33342, 10 μg/ml) for 30 min. After washing in PBS–BSA the embryos were incubated in a cooled solution of 0.1% Triton X-100 and 0.1% sodium citrate for 5 min, washed, covered with glycerol and mounted under coverslip. The embryos were examined under a Leica fluorescence microscope using excitation filters 340–380 nm (for Hoechst 33342) and 515–560 nm (for PI).
TUNEL assay of embryos
The embryos were washed three times for 5 min in washing solution PBS with 4 mg/ml polyvinylpyrrolidone (PBS–PVP, Sigma, USA). Then the embryos were fixed in 3.7% neutrally buffered formalin for 5 min and in 70% ethanol for 10 min. Permeabilization was carried out by 15 min incubation of embryos in 0.5% Triton X-100 in PBS. The embryos were processed for TUNEL using MEBSTAIN Direct Apoptosis Detection Kit (IM3171, Immunotech) according to the product's manual. Briefly, fixed and permeabilized embryos were incubated in TdT-labelling mixture (TdT buffer, FITC–dUTP and TdT) at 37 °C for 1 h. Following the incubation the TUNEL-reaction was stopped by three-times washing of embryos in PBS–PVP solution. Then the embryos were transferred onto a coverslip and covered with 5 μl of Vectashield anti-fade mounting medium, containing DAPI stain. The coverslip was fixed to a microslide using nail polish. The samples were stored at −20°C until fluorescence analysis (Makarevich et al., Reference Makarevich, Chrenek, Žilka, Pivko and Bulla2005).
Embryo diameter
Embryo diameters, excepting zona pellucida, were measured from the same images on the screen of the monitor using a scale bar micrometer (Leica, Germany), which was previously calibrated on a ×10 or ×20 objective and ×10 eyepiece. The diameter of the embryo without the zona pellucida was the mean of two measurements made perpendicularly to each other.
Statistics
Development of rabbit embryos up to blastocyst stage and the GFP expression were analysed using the chi-squared test. Cell numbers and TUNEL cells in embryos were calculated using analysis of variance (ANOVA).
Results
Developmental rate and transgenic mosaicism
The developmental rate of rabbit embryos, microinjected with different GFP gene constructs, up to morula stage was significantly lower (p < 0.05) than in intact (non-microinjected) rabbit embryos (from 66–74% vs. 98%, Table 1). No significant differences in total cell number at morula stage among the groups were found.
Development rate and transgenic mosaicism of rabbit embryos
avsb significant differences at p < 0.05.
Based on fluorescence analysis, GFP gene expression was detected at the morula stage. The highest portion of GFP-positive embryos (15%) was found using the CMVIE–EGFP–N1 (DrdI) gene construct, the lowest one (about 4%) was exhibited using CMVIE–hrGFP (ScaI) gene construct. Significantly higher transgenic mosaicism (60%) was found in rabbit embryos microinjected with CMVIE–EGFP–N1 (DrdI) gene construct, as was determined by fluorescence analysis of individual dissociated blastomeres.
Apoptosis, embryo diameter and total cell number
No significant differences in total cell number and diameter of transgenic rabbit embryo with different GFP gene constructs at hatching blastocyst stage were found (Table 2). The lower cell number was recorded in rabbit transgenic embryos with CMVIE–hrGFP linearized by ScaI (115.0 ± 8.20), the higher cell number was detected in rabbit transgenic embryos at hatching blastocyst stage with PGK–EGFP linearized by NotI (134.0 ± 35.00). The diameter of rabbit transgenic embryos was in the range of 128.8–158.4 μm.
Apoptosis, embryo diameter and total cell number in rabbit embryos
Significant difference (p < 0.05) was observed in the number of apoptotic cells between transgenic rabbit embryos with different GFP gene constructs (Table 2). The higher number of apoptotic cells (2.6 ± 0.33) was detected in rabbit transgenic embryos at hatching blastocyst stage carrying the CMVIE–EGFP (ClaI) transgene.
Discussion
In our previous study (Chrenek et al., Reference Chrenek, Vasicek, Makarevich, Jurcik, Suvegova, Bauer, Parkanyi, Rafay, Batorova and Paleyanda2005), we showed that a double microinjection technique increased transgene integration rate in rabbit embryos with a significant difference in blastocyst survival rate between double and single microinjection. Our present results demonstrate a higher (80%) developmental rate of microinjected embryos up to the morula stage, but lower GFP gene expression (15%), when compared with our previous study (66% and 35%, resp.). Generally, GFP has become more popular as a living marker for positively transfected clones in many studies, but variation in the levels of GFP expression has been shown (Liu et al., Reference Liu, Trimarchi and Keefe1999). Variation of GFP expression in transgenic embryos was also reported. In about 20% of positive embryos mosaic green signal (some blastomeres were without signal) was found. Higher mosaicism (up to 50% of microinjected embryos) was reported, when the GFP gene was used with different promoters (Rosochacki et al., Reference Rosochacki, Kozikova, Korwin-Kossakowski, Matejczyk, Poloszynowicz and Duszewska2003). In our study, the CMVIE promoter (Clontech) was used for the expression of EGFP. The protein could be visualized in microinjected rabbit embryos after maternal/zygotic genome transition i.e., at the 8–16-cell stage. Transgenic mosaicism in our rabbit embryos may be explained by DNA integration related to cell cycle in embryos produced by the microinjection (Chan et al., Reference Chan, Kukolj, Skalka and Bremel1999), but also that the transgene is distributed randomly into every blastomere (Wang et al., Reference Wang, Lin, Zhang and Chen2001). The major factors influencing successful transgene integration and expression in all blastomeres of transgenic embryos are competency of DNA repair system, replication and transcriptional activity of target cells (Chan et al., Reference Chan, Kukolj, Skalka and Bremel1999), which may also explain transgenic mosaicism in our experiment. Even though we detected some embryos with mosaicism, the use of GFP gene enabled us to also select 100% positive rabbit embryos at early preimplantation (morula) stage, without any deleterious effect on their survival. Based on our results and some literature reports we may conclude that it is necessary to pre-select microinjected embryos with evident transgenic mosaicism. These embryos are not suitable for embryo transfer, because transgenic mosaic animals will be born. It will not be a problem if transgene integration of transgenic mosaic animal is integrated in germinal tissues, as this could provide transgene transmission on to the next generation. In our unpublished results we received nine transgenic mosaic rabbits exhibiting different levels of GFP mosaicism, but with no transgene transmission to a new generation.
PCR analysis of microinjected embryos at several developmental stages has repeatedly shown that DNA construct persists in most of the morula stage embryos (Krisher et al., Reference Krisher, Gibbons, Canseco, Johnson, Russell, Notter, Velander and Gwazdauskas1994). In morula and blastocyst stage embryos, the proportion of surviving embryos, in which detectable levels of microinjected DNA were maintained, dropped to 25% (O'Neill, Reference O'Neill and Houdebine1995). In our previous in vitro experiments we detected the hFVIII transgene in 38% of single microinjected embryos and in 43% of double microinjected embryos at blastocyst stage (Chrenek et al., Reference Chrenek, Vasicek, Makarevich, Jurcik, Suvegova, Bauer, Parkanyi, Rafay, Batorova and Paleyanda2005). Although Page et al. (Reference Page, Bulter, Subramanian, Gwazdauskas, Johnson and Velander1995) obtained a 13% transgene frequency in microinjected embryos using polylysine–DNA mixtures, so far no live transgenics have been reported by cytoplasmic microinjection of DNA alone.
The embryo diameter and cell number are non-invasive markers of embryo quality, as their determination does not require destructing the embryo, when vital dye staining is used (Makarevich et al., Reference Makarevich, Chrenek and Fl'ak2006). Although embryo diameter is assumed to be a potential marker for the viability testing of bovine expanded blastocyst (Mori et al., Reference Mori, Otoi and Suzuki2002), this statement has not been confirmed in rabbit transgenic embryos. In the study of Shu-Zhen Liu et al. (Reference Shu-Zhen, Jiang, Yan, Jiang, Ouyang, Sun and Chen2005) the total cell number in rabbit cloned embryos was significant lower than in in vivo derived rabbit embryos.
Apoptosis (programmed cell death) is an active physiological process and the result of this process is elimination of abundant, damaged or harmful cells. This process is genetically controlled (Schwarzman & Cidlowski, Reference Schwartzman and Cidlowski1993). The presence of various molecular components of the apoptotic cascade has been proved in mouse, human and bovine preimplantation embryos (Warner et al., Reference Warner, Exley, McElhinny and Tang1998; Gutierrez–Adan et al., Reference Gutierrez-Adan, Rizos, Fair, Moreira, Pintado, de la Fuente, Boland and Lonergan2004; Jurisicova & Acton, Reference Jurisicova and Acton2004). Liu et al. (Reference Liu, Trimarchi and Keefe1999) reported that the link exists between expression of GFP and induction of apoptosis. Our results confirmed the result of Makarevich et al. (Reference Makarevich, Chrenek, Žilka, Pivko and Bulla2005), that apoptosis is not always the primary cause of the decrease in embryo cell number. Apoptotic processes at earlier stages of preimplantation development showed obvious dissimilarities with apoptotic processes in the blastocyst. In this case, the occurrence of apoptosis was sporadic and its presence was noted only after reaching particular developmental stages. Percentage of apoptotic cells in mouse embryos was usually higher than in rabbit embryos (Fabian et al., Reference Fabian, Makarevich, Chrenek, Bukovská and Koppel2007). Occurence of apoptosis in vitro was indicative of suboptimal culture conditions or influence of experimental procedures (Schwarzman & Cidlowski, Reference Schwartzman and Cidlowski1993; Makarevich et al., Reference Makarevich, Chrenek, Žilka, Pivko and Bulla2005). Significant differences between both transgenic groups found in our study can be explained by the fact that transgenic GFP rabbit embryos were produced by microinjection into the pronucleus of eggs. Therefore the higher proportion of apoptotic cells in transgenic EGFP embryos can be caused by many factors associated with the microinjection itself, for example, mechanical damage by microinjection pipette, exposure of the zygote to a microscope light of a higher intensity, or their combination (Chrenek et al., Reference Chrenek, Vasicek, Makarevich, Jurcik, Suvegova, Bauer, Parkanyi, Rafay, Batorova and Paleyanda2005; Makarevich et al., Reference Makarevich, Chrenek, Žilka, Pivko and Bulla2005).
Conclusion
Basing on developmental rate, transgenic mosaicism, embryo diameter and number of cells, a more suitable gene marker for rabbit transgenic embryos production and selection seems to be the CMVIE–EGFP (ClaI) gene construct.
Moreover, before further use of microinjected embryos (for embryo transfer, vitrification or ESC isolation) it is necessary to pre-select microinjected embryos with evident transgenic mosaicism. These embryos are not suitable for embryo transfer, because transgenic mosaic animals will be born.
Acknowledgements
This work was supported by the Slovak Research and Development Agency under the contract No. LPP-0126-06 and LPP-0119-09.
