- Email: [email protected]
Impact of embryonic expression of enhanced green fluorescent protein on early mouse development
BBRC Biochemical and Biophysical Research Communications 313 (2004) 1030–1036 www.elsevier.com/locate/ybbrc
Impact of embryonic expression of enhanced green fluorescent protein on early mouse development Vikram Devgan,a,1 Manchanahalli R.S. Rao,b and Polani B. Seshagiria,* a
Department of Molecular Reproduction, Development and Genetics, Indian Institute of Science, Bangalore 560012, India b Department of Biochemistry, Indian Institute of Science, Bangalore 560012, India Received 20 November 2003
Abstract The impact of embryonic enhanced green fluorescent protein (EGFP)-expression on development is not clear. In this study, we comprehensively assessed EGFP-expression pattern and its effect on early mouse development, following pronuclear-microinjection of the EGFP-transgene, containing chicken-b-actin promoter and cytomegalovirus enhancer. Preimplantation embryos exhibited differential EGFP-expression patterns. While blastocyst development of non-expressing embryos was 77.3 1.8%, that of expressing embryos was only 43.9 1.6% (P < 0:0001). Developmental competence of embryos negatively correlated (r ¼ 0:99) with the levels of EGFP-expression. Faint-, moderate-, and intense-expressing embryos developed to 83.1 5.3%, 50 5%, and 9.5 3.9% blastocysts, respectively (P < 0:002). Interestingly, blastocysts expressing faint–moderate levels of EGFP were developmentally competent through the post-implantation period and delivered viable transgenic ‘green’ mice, following embryo transfer. These results indicate that hyper-expression of EGFP affects preimplantation development and faint–moderate level of its expression is compatible with normal embryogenesis in the mouse. Ó 2003 Elsevier Inc. All rights reserved. Keywords: Enhanced green fluorescent protein; Preimplantation embryo; Differential expression; Early development (mouse); Transgenesis
Mammalian preimplantation embryo development is a complex and critical phase during embryogenesis, which ensures a smooth maternal-to-zygotic transition and a timely initiation of differentiation. Molecular mechanisms that regulate gene expression during this phase of development are still not clearly understood. In this regard, transgenesis and the use of reporter genes such as b-galactosidase [1] and luciferase [2] are indispensable experimental approaches for evaluating the mechanism of early gene expression and regulation during development. Most reporter genes, in vogue, however, have a variety of limitations such as they require exogenous substrates or cofactors for detecting their expression and/or their assay methods are quite often not compatible with normal embryo development. *
Corresponding author. Fax: +91-80-360-0999. E-mail address: [email protected] (P.B. Seshagiri). 1 Present address: Cutaneous Biology Research Center, Massachusetts General hospital and Harvard Medical School, Charlestown, MA 02129, USA. 0006-291X/$ - see front matter Ó 2003 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2003.11.184
The enhanced green fluorescent protein (EGFP), a redshifted and humanized GFP variant [3], has become an extremely valuable and novel genetic reporter system in developmental studies and has been used in many different organisms, in view of its sensitivity, specificity, ease of detection, and non-invasive nature. There are a number of reports using EGFP as a reporter gene for studying embryogenesis [4,5] and for generating transgenic mice [6,7]. We and others have generated GFP-/EGFP-expressing transgenic mice by uterine transfer of selective transgene-expressing embryos to pseudopregnant recipients [8–11]. Although the GFP/EGFP is believed to be a biologically inert and non-toxic molecule [6,7], its expression in a few instances is shown to be cytotoxic in mammalian cell lines [12,13] and its overexpression was shown to cause dilated cardiomyopathy in transgenic mice [14] and toxicity in plants [15]. To the best of our knowledge, however, there has been no report precisely evaluating the developmental impact of EGFP-expression during early embryogenesis and neonatal development in any
V. Devgan et al. / Biochemical and Biophysical Research Communications 313 (2004) 1030–1036
mammalian species, including the mouse. There is a need to understand EGFP-expression pattern during preimplantation development in order to achieve maximum efficiency of transgenesis and to fully exploit EGFP reporter system for studying developmental regulation of gene expression. It was, therefore, of interest to comprehensively evaluate the expression pattern of the EGFP-transgene, under the control of a ubiquitous chicken-b-actin (CbA) promoter and cytomegalovirus enhancer (CMV-IE), during the entire period of preimplantation embryo development, and to assess its effect on embryogenesis in the mouse. We report, here, that the developmental competence of preimplantation mouse embryos negatively correlates with the level of EGFP-expression and that the EGFP-expression, at faint–moderate levels in developing embryos derived from transgenic ‘green’ mouse lines, is compatible with normal development.
Materials and methods Animals. Four- to six-week-old wild-type or EGFP-expressing transgenic [10] FVB/N and wild-type C57BL/6 mice, maintained on a 14 L:10 D lighting schedule (lights on at 06:00 h) and normal temperature (24–26 °C), from our transgenic mouse facility were used for the present study. Depending on experiments, embryos were recovered on 0.5–4 days post-coitus (dpc) as described earlier [8,9]. Procedures for handling and experimentation followed the Guidelines on the Use of Animals in Scientific Research (Indian National Science Academy, New Delhi, 1992). Transgene-microinjection and embryo culture. A pCAGGS-EGFPexpression vector was constructed as described earlier [10]. The transgene (CbA-EGFP; 3.2 kb), which included the CMV-IE enhancer, CbA promoter, b-actin intron, EGFP gene, and rabbit b-globin poly(A) signal, was excised with BamHI and SalI, electrophoresed, and gel purified using Qiagen gel elution kit (Qiagen, Chatsworth, CA, USA). Pronuclear-stage embryos were recovered from superovulated FVB/N females mated with proven stud males. The purified transgene (EGFP; 5 ng/ll) was microinjected (Transjector #5246, Eppendorf, Barkhausenweg, Hamburg, Germany) into the male pronuclei of embryos. Embryos, surviving microinjection, were cultured in 50 ll drops of the M16 (Sigma Chemical Company, St. Louis, MO, USA) medium overlaid with 2.5 ml of silicone oil in 35 mm plastic culture dishes (Greiner Labortechnik, Frickenhausen, Germany) in an incubator maintained at 37 °C with a humidified gas atmosphere of 5% CO2 in air [11]. Monitoring of EGFP fluorescence. The expression of EGFP in preand post-implantation embryos or in neonatal mice was examined under a fluorescence inverted microscope (IX-70, Olympus, Tokyo, Japan) and stereozoom microscope (SZX-RFL2, Olympus), respectively, with a GFP blue-filter set (470–490 nm). Fluorescent pictures were photographed using a 400 ASA Kodak film with the photomicrographic unit (PM-30, Olympus) attached to the microscopes. Embryonic EGFP fluorescence was monitored at every 12 h upto 120 h of culture. Fluorescence patterns were documented and EGFP-expressing ‘green’ embryos were separated from the non-expressing embryos and cultured individually along with ’9 non-injected similar stage embryos in separate culture drops. Categorization of EGFP-expressing embryos by visual and confocal microscopy methods. Cultured embryos were observed for green fluorescence and those exhibiting uniform green fluorescence were separated from the rest. They were visually classified into three categories
1031
based on their fluorescence intensity grades, i.e., faint, moderate, and intense and given subjective scores of 2, 4, and 6, respectively. Categorized embryos were cultured separately along with similar stage noninjected embryos. The total numbers of embryos were 10–15 in each 50 ll drop of M16 medium. We depended on assessing relative levels of EGFP in embryos only by fluorescence since it is a non-invasive method and the assessed embryos could be used for culture experiments, unlike immunoblot quantification, which would be a terminal experiment. Though the segregation of EGFP-expressing embryos could be performed by confocal microscopy, it is unsuitable for assessing blastocyst development during actively growing embryo cultures. Therefore, subjective method of fluorescence assessment was adopted as described above. However, to rule out the possibility of investigator’s bias in the assessment of fluorescence intensity, a double blind study was conducted on another set of embryos. For this, 2- and 4-cell embryos, developed from EGFP-injected pronuclear-stage-eggs, exhibiting uniform EGFP-expression, were transferred individually to 5 ll drops of M2 medium. Each embryo was assigned a particular identity number and classified by visual method as faint- (score: 2), moderate- (4), and intense- (6) fluorescence categories. These embryos were randomized by an independent investigator, given a different identity number and then subjected to confocal laser scanning microscopy (TCS-SP, Leica Microsystems, Heidelberger GmbH, Germany) with excitation at 488 nm and detection at 500–530 nm band pass filters. The fluorescence image of each embryo was recorded and its fluorescence intensity was quantified in arbitrary units (AU) using the Leica confocal software (LC Sv. 585). Based on the range of fluorescence intensities, embryos were categorized into three groups as low-, moderate-, and intense-fluorescing embryos and compared with those of the subjective method after decoding identity numbers. Statistical analysis. All experiments were performed for a minimum of three times. Results are expressed as means SEM. Percentage values were subjected to arcsin transformation before statistical analysis. Comparisons between groups were made by the unpaired Student’s t test (PRISM Graph Pad version 2; Graph Pad Software, San Diego, CA). To determine the correlation between the fluorescence intensities and developmental competence of EGFP-expressing embryos, a regression analysis was performed and correlation coefficient (r) was calculated; r value closer to )1 indicated a strong negative relationship.
Results Embryonic EGFP-expression during preimplantation development, following microinjection The CbA-EGFP-transgene was microinjected into 453 pronucleate-stage-eggs. Of these, 372 (82%) eggs, surviving microinjection, were cultured and monitored for EGFP-expression. Green fluorescence was detected in 117 (32%) embryos, initially at 24–36 h post-injection, i.e., at the 2-cell stage; some non-starters also exhibited EGFP-expression (Fig. 1I). The proportion of EGFPexpressing embryos increased through development and was maximum (30%; 111/372) at 72 h post-injection, i.e., at the morula stage. Incidentally, 15% (18/117) of EGFP-expressing embryos showed only transient expression. Embryos exhibited variable EGFP-expression patterns. Out of 117 EGFP-expressing embryos, 75 (64%) showed uniform green fluorescence, while 26 (22%)
1032
V. Devgan et al. / Biochemical and Biophysical Research Communications 313 (2004) 1030–1036
Fig. 1. Embryonic EGFP-expression, following pronuclear-microinjection of the EGFP-transgene. Embryos exhibited (I) uniform, (II) mosaic, and (III) differential EGFP-expression patterns. Panel I shows photomicrographs of 1-cell (A,I), 2-cell (B,J), 4-cell (C,K), 8-cell (D,L), morula (E,M), and expanded (F,N), hatching (G,O), and hatched (H,P) blastocysts. Panels A–H and I–P correspond to differential interference contrast (DIC) and their corresponding fluorescence pictures, respectively. Panel II shows photomicrographs of 2-cell (A, J), 4-cell (B–D, K–M), 8-cell (E–F, N–O), morula (G,P), and expanded blastocyst (H–I, Q–R). Panels A–I and J–R correspond to DIC and their corresponding fluorescence pictures, respectively. Panel III shows photomicrographs of 2-cell (A–C, G–I, M–O, and S–U) and 4-cell (D–F, J–L, P–R, and V–X) embryos exhibiting faint (A,M and D,P) or moderate (B,N and E,Q) or high (C,O and F,R) green fluorescence intensity and differential EGFP-expression within sister blastomeres (G– L, S–X). Panels A–L and M–X correspond to bright field and their corresponding confocal fluorescence images, respectively. All the pictures are photographed with 20 objective. Magnification bars correspond to 50 lm and are shown in panels I—O (identical in panels A–G and I–O of I), I—P (identical in panel H of I), II—R (identical in all panels of II), and III—X (identical in all panels of III).
Fig. 3. A composite picture showing EGFP-expression during the entire period of EGFP-transgenic mouse development. EGFP-expression is shown in the oocyte (O) but the mature sperm (S) does not exhibit EGFP-expression. EGFP fluorescence is uniform in preimplantation embryos including 1-cell (0.5 dpc: days post-coitum), 2-cell (1.5 dpc), 4-cell (2 dpc), 8-cell (2.5 dpc), morula (3 dpc), and blastocyst (4 dpc) stages, post-implantation embryos including egg-cylinder- (6.5 dpc), late somite- (8.5 dpc), early limb bud- (9.5 dpc, 10.5 dpc), late limb bud- (11.5 dpc, 12.5 dpc), and late fetal(14.5 dpc, 16.5 dpc) stages and neonatal (D1: Day 1) and adult mice (1M: 1 month).
V. Devgan et al. / Biochemical and Biophysical Research Communications 313 (2004) 1030–1036
showed mosaicism, and 16 (14%) showed differential fluorescence within sister blastomeres. Photomicrographs of all stages of preimplantation embryos exhibiting uniform or mosaic EGFP-expression are shown in Figs. 1I and II, respectively. Interestingly, embryos, with uniform EGFP-expression, showed different levels of green fluorescence intensities. About 32% (24/75), 40% (30/75), and 28% (21/75) of these embryos showed faint, moderate, and intense levels of green fluorescence, respectively. Photomicrographs of 2-cell and 4-cell embryos exhibiting different levels of fluorescence intensities and differential fluorescence within sister blastomeres are shown in Fig. 1III. Effect of EGFP-expression on blastocyst development To evaluate the developmental potential and viability of EGFP-transgenic embryos, EGFP-transgeneinjected, T10 E0:1 buffer-injected, and non-injected pronuclear-stage-eggs were cultured and their blastocyst development was assessed after 120 h of culture (Table 1). Non-injected and buffer-injected embryos showed 91.3 2.3% and 81.3 2.3% blastocyst development, respectively. In contrast, only 69.4 1.9% EGFP-injected embryos developed to blastocysts. The percentage of blastocyst development of EGFP-injected embryos was significantly lesser than that of either buffer-injected or non-injected embryos (Table 1; P < 0:005). Moreover, when EGFP-injected embryos were exposed to blue light for 30 s at every 12 h of culture for observing fluorescence of EGFP-expression, 67.2 1.8% blastocyst development was achieved (Table 1). The percentage of blastocyst development was comparable in both categories of EGFP-injected embryos, i.e., without and with blue-light exposure during culture (Table 1; P > 0:05). Among EGFP-injected embryos, non-fluorescing embryos showed 77.3 1.8% blastocyst development, while EGFP-expressing, ‘green fluorescing’ embryos showed only 43.9 1.6% blastocyst development. The percentage of blastocyst development of EGFP-expressing embryos was significantly lesser (P < 0:0001) than that of non-expressing-EGFPinjected embryos (Table 1).
1033
Effect of differential level of EGFP-expression on blastocyst development To delineate relationship between the level of EGFPexpression and developmental competence of preimplantation embryos, uniform EGFP-expressing embryos were visually classified into faint (score 2), moderate (4), and intense (6) categories based on their fluorescence. Interestingly, a negative correlation (r ¼ 0:99) was observed between the percentages of blastocyst development and the levels of EGFP-expression. Embryos with faint fluorescence showed 83.1 5.3% (23/28) blastocyst development, while embryos with moderate and intense fluorescence showed 50 5% (21/42) and 9.5 3.9% (3/29) blastocyst development, respectively (Fig. 2A). The percentage of blastocyst development in the faint category was significantly higher (P < 0:003) than those in the moderate and intense categories. Data on the double blind study involving visual and confocal microscopy methods are shown in Fig. 2B. Confocal microscopy confirmed that embryos had varying extents of fluorescence intensities (1–225 AU). On the basis of quantified fluorescence intensity, embryos were divided into three groups, i.e., faint (1–75 AU), moderate (76–150 AU), and intense (151–225 AU) and their groups were compared with the visual method of categorization. Incidentally, 66 out of 75 (88%) were correctly assigned to the same categories by both methods. The corrected percentages of blastocyst development were calculated from the observed proportion of mismatch using inverting 3 3 matrix of coefficient and found to be negatively correlated with the levels of EGFP-expression (data not shown). Effect of EGFP-expression on post-implantation development To investigate the impact of embryonic EGFPexpression during post-implantation development, EGFP-injected embryos were cultured to blastocysts. EGFP-expressing ‘green’ or non-expressing ‘non-green’ blastocysts were segregated and transferred independently to uteri of day 3 pseudopregnant (C57BL/6)
Table 1 Development of blastocysts from EGFP-injected-pronuclear-stage mouse embryos Treatment
No. of embryos cultured
Blastocysts (%)
Non-injected T10 E0:1 buffer-injected EGFP-injected: Without observation at 480 nm With observation at 480 nm Non-expressing EGFP-expressing
180 163
164 (91.3 2.3a;b;c ) 131 (81.3 2.3a;d;e )
186 289 198 91
129 (69.4 1.9b;d ) 191 (67.2 1.8c;e ) 152 (77.3 1.8f ) 39 (43.9 1.6f )
Values in parentheses represent means SEM, from 10 replicate experiments. Values with identical superscripts differ significantly. a: P < 0:01; b,c,f: P < 0:0001; and d,e: P < 0:002.
1034
V. Devgan et al. / Biochemical and Biophysical Research Communications 313 (2004) 1030–1036
EGFP-expression in pre- and post-implantation embryos of EGFP-transgenic mice Embryos were recovered, from wild-type females mated with hemizygous transgenic males, on 0.5. 1.5, 2, 2.5, 3, and 4 dpc and their EGFP-expression patterns were assessed. EGFP-expression was initially detected, in about half the embryos (54.2%; 26/48) at the compacting 8-cell stage (data not shown). By morula stage, EGFP-expression was clearly visible and thereafter, fluorescence was stable and uniform. On the other hand, all preimplantation stage embryos, recovered on 0.5, 1.5, 2, 2.5, 3, and 4 dpc of hemizygous transgenic females mated with wild-type males, exhibited EGFP-expression (Fig. 3). When freshly recovered day 1 embryos were cultured, about half the embryos (44.2%; 23/52) began to lose green fluorescence by 96 h of culture and they completely lost the fluorescence by 120 h (data not shown). Incidentally, in all fluorescing embryos, EGFPexpression was uniform and faint (Fig. 3). During post-implantation development, EGFP-expression was found to be ubiquitous in embryonic (Fig. 3) and extra-embryonic tissues (data not shown). Carcass analysis of fetus (14.5 dpc) and adult (1 month) transgenic mice showed that all organs derived from the three germ-cell lineages were ubiquitously green fluorescent (data not shown). Photomicrographs of all stages of mouse development exhibiting EGFP-expression are shown in Fig. 3.
Fig. 2. (A) Correlation between the fluorescence intensity and blastocyst development of uniformly EGFP-expressing embryos. Blastocyst development of embryos with faint (), moderate (P), and intense (j) green fluorescence intensity is shown. Results represent means SEM of five replicate experiments. Values with identical superscripts differ significantly. P values: a < 0:002; b; c < 0:0005. (B) Assessment of fluorescence intensity of uniformly EGFP-expressing embryos by confocal microscopy. Fluorescence intensities of green fluorescing embryos (n ¼ 77) were quantified in arbitrary units (AU) using Leica confocal software. These embryos were divided in three groups based on their quantified fluorescence intensity (FI), i.e., faint (FI ¼ 1– 75 AU), moderate (FI ¼ 76–150 AU), and intense (FI ¼ 151–225 AU).
recipients. In this regard, 84 EGFP-expressing and 208 non-expressing blastocysts were transferred to 8 and 10 pseudopregnant recipients, respectively. Of these, 3 (38%) and 5 (50%) females became pregnant and delivered at term 5 (15.4 4.8%; 5/32) and 23 (22.4 1.2%; 23/102) pups, respectively. The genotypic (PCR and genomic Southern) analysis confirmed that all pups born from green blastocysts–uterine-transfer were transgenic and none of the pups born from non-green blastocysts– uterine-transfer were transgenic [10]. The percentage of pups born per transferred embryos from EGFP-expressing blastocysts transfers was similar (P > 0:05) to that of non-expressing blastocyst transfers.
Discussion In the present study, we demonstrate, for the first time, that EGFP exhibits expression-level-dependent inhibition of preimplantation embryo development and hyper-expression reduces developmental competence and potential viability of mouse embryos. However, faint–moderate embryonic EGFP-expression is compatible with development of microinjected-embryos and of those derived from the EGFP-expressing transgenic “green” mice. Our data on the comprehensive analysis of embryonic expression patterns of the EGFP-transgene during mouse development are quite significant in terms of our understanding of its behavior during early development and its potential use as a reporter gene (for real-time imaging analysis) while studying regulation of gene expression during preimplantation embryo development. Developmental competence of preimplantation embryos correlates with the level of the transgene expression. Embryos with the lowest level of EGFP-expression show maximum blastocyst development and those expressing the highest level of EGFP show poorest blastocyst development. Though embryonic expression of EGFP has been studied earlier [8], the impact
V. Devgan et al. / Biochemical and Biophysical Research Communications 313 (2004) 1030–1036
(inhibition) on preimplantation embryo development has so far not been thoroughly investigated and the earlier work [8] has not reported such an effect in mammalian early embryos. However, our observation on the inhibition of development of preimplantation embryos is quite consistent with those of others who showed EGFP-cytotoxicity in a few mammalian cell lines [12,13] and cardiomyopathy in transgenic mice [14]. Incidentally, toxicity due to overexpression of GFP has been reported in transgenic plants [15]. It could be argued that sub-optimal culture conditions and/or microinjection-induced mechanical trauma [16] may result in the observed poor development of blastocysts, expressing high level of EGFP. We, however, discount this possibility since T10 E0:1 bufferinjected-eggs showed significantly higher blastocyst development when compared to that of EGFP-injected embryos (Table 1). Moreover, M16 medium, used in the study, supports comparable and maximal development of high quality blastocysts from EGFP-injected or noninjected pronuclear-stage mouse embryos [11]. Similarly, the possibility of detrimental effect of blue-light exposure is also ruled out, since the EGFP-injected embryos show similar percentages of blastocyst development without or with blue-light exposure (Table 1). Another possibility to be considered is that the observed inhibition of cleavage of embryonic cells by hyper-expression of EGFP (or any transgene) could be due to the change in the gene expression profile of cells [13]. But, how are these changes triggered by the transgene (EGFP) and what could be the consequences of these changes remain unclear and require further studies. To the best of our knowledge, there is no published report on the possible EGFP-induced metabolic perturbation or effect on organelle function in embryonic cells. The low blastocyst development may, in part, be attributable to either an insertional mutagenesis and/or repression of developmentally essential endogenous gene(s) by hyper-expression of the (EGFP) transgene, resulting in reduced developmental competence. Towards understanding these possibilities, it would be quite interesting to have a comparative study using Renilla reniformis GFP (hrGFP) vis- a-vis Aequoria victoria GFP, since the former is known to possess negligible or low cellular toxicity [17]. It is interesting that the developed blastocysts expressing EGFP did not affect post-implantation development as well as the reproductive outcome of transgenic animals. Following pronuclear-microinjection, the preimplantation phase of embryo development could be most vulnerable to possible embryotoxicity due to hyperexpression of EGFP. The preimplantation phase of embryo development could be acting as a selection strategy to allow only faint–moderate EGFP-expressing embryos to develop to blastocysts, which are capable of completing normal development. It may be necessary to screen/select
1035
developmentally competent EGFP-expressing transgenic-embryos (blastocysts) prior to uterine-embryo-transfer, thereby improving the percentage of EGFP-transgenic offspring. Following this procedure, i.e., embryo-culturebased transgenesis (ECBT), we have successfully generated five EGFP-transgenic founder mice and observed that ubiquitously expressing EGFP-transgenic founders develop normally to fertile adults and produce transgenic lines [10,11]. The relatively lower rate of pregnancy with EGFP-expressing blastocysts when compared to that of the non-EGFP-expressing blastocysts (see results) could be due to a less number of embryos used for uterine transfer in the former case (10 vs. 20). However, the percentage of pups born per transferred embryos was similar in both categories. Despite using a universal and ubiquitous promoter and enhancer, we observed a number of interesting variations in the pattern of transgene expression in terms of the time of onset of gene expression and relative levels of EGFP-expression in various stages of preimplantation embryos (Fig. 2). The transient EGFPexpression may be due to the epichromosomal, non-integrated transgene, while the mosaic expression could be due to the integration of the transgene after the first round of DNA replication [18]. However, the observed difference in the levels of green fluorescence intensities is possibly due to different copy numbers of the transgene and/or to a chromosomal position effect [19]. Our earlier data show that the high copy number of EGFP in a single genomic locus may be responsible for repeat-induced gene silencing [10]. Our observation on the differential EGFP-expression within sister blastomeres is consistent with a recent report [20]. This pattern of expression could possibly be due to: (i) independent transgene integration events in individual blastomere during sequential cell division, (ii) EGFP-expression is developmentally regulated and each blastomere has different developmental fate, and (iii) non-integrated transgene distributed randomly during cell division. However, the second reason is unlikely, because, all blastomeres of mammalian embryos are totipotent upto the morula stage. We also observed that unlike spermatogenic cells, all oocytes (100%), derived from hemizygous EGFP-transgenic females, show EGFP-expression during the entire meiotic maturation process (data not shown). The green fluorescence in 100% embryos, recovered from hemizygous transgenic females mated with wild-type males, could be due to a residual, maternally expressed EGFP during oogenesis. But, about 50% of embryos lose green fluorescence during culture by 24 h (blastocyst; data not shown) or 120 h (1-cell), possibly due to degradation of maternally derived transcript and/or protein. Rest of the embryos retaining fluorescence through the ensuing developmental period is due to de novo embryonic expression of the inherited EGFP-transgene, possibly from
1036
V. Devgan et al. / Biochemical and Biophysical Research Communications 313 (2004) 1030–1036
the compacting 8-cell stage, similar to that observed with embryos having paternally derived EGFP gene. It is quite interesting to note that embryos derived from transgenic ‘green’ mouse lines show faint green fluorescence and the percentage of blastocyst development of embryos derived from EGFP-transgenic mice is quite similar to that of EGFP-injected embryos with faint green fluorescence (data not shown). Furthermore, during post-implantation development, all embryonic stages show ubiquitous EGFP-expression and throughout the neo-natal development. In conclusion, following microinjection, embryos exhibit variations in the pattern of the EGFP-expression illustrating the complex expression behavior of the transgene during early development. Hyper-expression of EGFP affects developmental ability and potential viability of preimplantation embryos. However, developed embryos, expressing EGFP at faint–moderate levels and/or those from EGFP-transgenic mice, are unaffected and are completely compatible with normal development. Our findings are quite valuable in providing new insights into using EGFP as an in vivo fluorescent marker for studying regulation of gene expression during mammalian preimplantation development. Besides, the EGFP-transgenic ‘green’ mouse is a rich source of fluorescent-marked (embryonic/somatic) stem cells for studying regulation of lineage-specific cell differentiation during mouse development.
Acknowledgments Financial support from the Department of Biotechnology, New Delhi, is gratefully acknowledged. The authors are thankful to Dr. S.M. Totey for advice and help; Ms. G.V. Sireesha, Ms. M. Sarkar, Mr. S. Nyati, and Dr. Uday kumar for their excellent technical support. Our thanks are also due to Prof. N. Joshi for advice on statistical evaluation of the data and to Ms. M.S. Padmavathi for help in the preparation of the manuscript.
References [1] S. Takeda, Y. Toyoda, Expression of SV40-lacZ gene in mouse preimplantation embryos after pronuclear microinjection, Mol. Reprod. Dev. 30 (1991) 90–94. [2] E.M. Thompson, P. Adenot, F.I. Tsuji, J.P. Renard, Real time imaging of transcriptional activity in live mouse preimplantation embryos using a secreted luciferase, Proc. Natl. Acad. Sci. USA 92 (1995) 1317–1321. [3] B.P. Cormack, R.H. Valdivia, S. Falkow, FACS-optimized mutants of the green fluorescent protein (GFP), Gene 173 (1996) 33–38.
[4] N. Kirchhof, J.W. Carnwath, E. Lemme, K. Anastassiadis, H. Scholer, H. Niemann, Expression pattern of Oct-4 in preimplantation embryos of different species, Biol. Reprod. 63 (2000) 1698– 1705. [5] K.A. Molyneaux, J. Stallock, K. Schaible, C. Wylie, Time-lapse analysis of living mouse germ cell migration, Dev. Biol. 240 (2001) 488–498. [6] M. Okabe, M. Ikawa, K. Kominami, T. Nakanishi, Y. Nishimune, ‘Green mice’ as a source of ubiquitous green cells, FEBS lett. 407 (1997) 313–319. [7] A.K. Hadjantonakis, M. Gertsenstein, M. Ikawa, M. Okabe, A. Nagy, Generating green fluorescent mice by germline transmission of green fluorescent ES cells, Mech. Dev. 76 (1998) 79–90. [8] T. Takada, K. Iida, T. Awaji, K. Itoh, R. Takahashi, A. Shibui, K. Yoshida, S. Sugano, G. Tsujimoto, Selective production of transgenic mice using green fluorescent protein as a marker, Nat. Biotechnol. 15 (1997) 458–461. [9] J.T. Keiser, P.M. Jobst, A.S. Garst, J.T. Boone, C.B. Geyer, C. Phelps, D.L. Ayares, R.L. Page, Preimplantation screening for transgenesis using an embryonic specific promoter and green fluorescent protein, Cloning 3 (2001) 23–30. [10] V. Devgan, M. Thomas, K.S. Ullas, M.R.S. Rao, P.B. Seshagiri, Embryo culture-based generation of enhanced green fluorescent protein-transgenic mice, Biochem. Biophys. Res. Commun. 303 (2003) 994–1001. [11] V. Devgan, P.B. Seshagiri, Successful development of viable blastocysts from enhanced green fluorescent protein transgenemicroinjected mouse embryos: comparison of culture media, Mol. Reprod. Dev. 65 (2003) 269–277. [12] H.S. Liu, M.S. Jan, C.K. Chou, P.H. Chen, N.J. Ke, Is green fluorescent protein toxic to living cells?, Biochem. Biophys. Res. Commun. 260 (1999) 712–717. [13] P. Vaillancourt, B. Rogers, Y. Wang, Do not let a toxic GFP to skew your results!, Strategies 14 (2001) 86–87. [14] W.Y. Huang, J. Aramburu, P.S. Douglas, S. Izumo, Transgenic expression of green fluorescence protein can cause dilated cardiomyopathy, Nat. Med. 6 (2000) 482–483. [15] J. Haselowff, B. Amos, GFP in plants, Trends Genet. 11 (1995) 328–329. [16] P. Chrenek, A. Makarevich, D. Vasicek, J. Laurincik, J. Bulla, T. Gajarska, J. Rafay, Effects of superovulation, culture and microinjection on development of rabbit embryos in vitro, Theriogenology 50 (1998) 659–666. [17] K. Felts, B. Rogers, K. Chen, H. Ji, J. Sorge, P. Vaillancourt, Recombinant Renilla reniformis GFP display low toxicity, Strategies 13 (2000) 85–87. [18] A.W.S. Chan, G. Kukolj, A.M. Skalka, R.D. Bremel, Timing of DNA integration, transgenic mosaicism, and pronuclear microinjection, Mol. Reprod. Dev. 52 (1999) 406–413. [19] T. Takada, K. Iida, K. Akasaka, H. Yasue, R. Torii, G. Tsujimoto, M. Taira, H. Kimura, Evaluation of heterologous insulator function with regard to chromosomal position effect in the mouse blastocyst and fetus, Mol. Reprod. Dev. 57 (2000) 232– 237. [20] S.Y. Medvedev, T. Tokunaga, R.M. Schultz, T. Furukawa, T. Nagai, M. Yamaguchi, M. Hosoe, A.F. Yakovlev, S. Takahashi, Y. Izaike, Quantitative analysis of gene expression in preimplantation mouse embryos using green fluorescent protein reporter, Biol. Reprod. 67 (2002) 282–286.
Impact of embryonic expression of enhanced green fluorescent protein on early mouse development Vikram Devgan,a,1 Manchanahalli R.S. Rao,b and Polani B. Seshagiria,* a
Department of Molecular Reproduction, Development and Genetics, Indian Institute of Science, Bangalore 560012, India b Department of Biochemistry, Indian Institute of Science, Bangalore 560012, India Received 20 November 2003
Abstract The impact of embryonic enhanced green fluorescent protein (EGFP)-expression on development is not clear. In this study, we comprehensively assessed EGFP-expression pattern and its effect on early mouse development, following pronuclear-microinjection of the EGFP-transgene, containing chicken-b-actin promoter and cytomegalovirus enhancer. Preimplantation embryos exhibited differential EGFP-expression patterns. While blastocyst development of non-expressing embryos was 77.3 1.8%, that of expressing embryos was only 43.9 1.6% (P < 0:0001). Developmental competence of embryos negatively correlated (r ¼ 0:99) with the levels of EGFP-expression. Faint-, moderate-, and intense-expressing embryos developed to 83.1 5.3%, 50 5%, and 9.5 3.9% blastocysts, respectively (P < 0:002). Interestingly, blastocysts expressing faint–moderate levels of EGFP were developmentally competent through the post-implantation period and delivered viable transgenic ‘green’ mice, following embryo transfer. These results indicate that hyper-expression of EGFP affects preimplantation development and faint–moderate level of its expression is compatible with normal embryogenesis in the mouse. Ó 2003 Elsevier Inc. All rights reserved. Keywords: Enhanced green fluorescent protein; Preimplantation embryo; Differential expression; Early development (mouse); Transgenesis
Mammalian preimplantation embryo development is a complex and critical phase during embryogenesis, which ensures a smooth maternal-to-zygotic transition and a timely initiation of differentiation. Molecular mechanisms that regulate gene expression during this phase of development are still not clearly understood. In this regard, transgenesis and the use of reporter genes such as b-galactosidase [1] and luciferase [2] are indispensable experimental approaches for evaluating the mechanism of early gene expression and regulation during development. Most reporter genes, in vogue, however, have a variety of limitations such as they require exogenous substrates or cofactors for detecting their expression and/or their assay methods are quite often not compatible with normal embryo development. *
Corresponding author. Fax: +91-80-360-0999. E-mail address: [email protected] (P.B. Seshagiri). 1 Present address: Cutaneous Biology Research Center, Massachusetts General hospital and Harvard Medical School, Charlestown, MA 02129, USA. 0006-291X/$ - see front matter Ó 2003 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2003.11.184
The enhanced green fluorescent protein (EGFP), a redshifted and humanized GFP variant [3], has become an extremely valuable and novel genetic reporter system in developmental studies and has been used in many different organisms, in view of its sensitivity, specificity, ease of detection, and non-invasive nature. There are a number of reports using EGFP as a reporter gene for studying embryogenesis [4,5] and for generating transgenic mice [6,7]. We and others have generated GFP-/EGFP-expressing transgenic mice by uterine transfer of selective transgene-expressing embryos to pseudopregnant recipients [8–11]. Although the GFP/EGFP is believed to be a biologically inert and non-toxic molecule [6,7], its expression in a few instances is shown to be cytotoxic in mammalian cell lines [12,13] and its overexpression was shown to cause dilated cardiomyopathy in transgenic mice [14] and toxicity in plants [15]. To the best of our knowledge, however, there has been no report precisely evaluating the developmental impact of EGFP-expression during early embryogenesis and neonatal development in any
V. Devgan et al. / Biochemical and Biophysical Research Communications 313 (2004) 1030–1036
mammalian species, including the mouse. There is a need to understand EGFP-expression pattern during preimplantation development in order to achieve maximum efficiency of transgenesis and to fully exploit EGFP reporter system for studying developmental regulation of gene expression. It was, therefore, of interest to comprehensively evaluate the expression pattern of the EGFP-transgene, under the control of a ubiquitous chicken-b-actin (CbA) promoter and cytomegalovirus enhancer (CMV-IE), during the entire period of preimplantation embryo development, and to assess its effect on embryogenesis in the mouse. We report, here, that the developmental competence of preimplantation mouse embryos negatively correlates with the level of EGFP-expression and that the EGFP-expression, at faint–moderate levels in developing embryos derived from transgenic ‘green’ mouse lines, is compatible with normal development.
Materials and methods Animals. Four- to six-week-old wild-type or EGFP-expressing transgenic [10] FVB/N and wild-type C57BL/6 mice, maintained on a 14 L:10 D lighting schedule (lights on at 06:00 h) and normal temperature (24–26 °C), from our transgenic mouse facility were used for the present study. Depending on experiments, embryos were recovered on 0.5–4 days post-coitus (dpc) as described earlier [8,9]. Procedures for handling and experimentation followed the Guidelines on the Use of Animals in Scientific Research (Indian National Science Academy, New Delhi, 1992). Transgene-microinjection and embryo culture. A pCAGGS-EGFPexpression vector was constructed as described earlier [10]. The transgene (CbA-EGFP; 3.2 kb), which included the CMV-IE enhancer, CbA promoter, b-actin intron, EGFP gene, and rabbit b-globin poly(A) signal, was excised with BamHI and SalI, electrophoresed, and gel purified using Qiagen gel elution kit (Qiagen, Chatsworth, CA, USA). Pronuclear-stage embryos were recovered from superovulated FVB/N females mated with proven stud males. The purified transgene (EGFP; 5 ng/ll) was microinjected (Transjector #5246, Eppendorf, Barkhausenweg, Hamburg, Germany) into the male pronuclei of embryos. Embryos, surviving microinjection, were cultured in 50 ll drops of the M16 (Sigma Chemical Company, St. Louis, MO, USA) medium overlaid with 2.5 ml of silicone oil in 35 mm plastic culture dishes (Greiner Labortechnik, Frickenhausen, Germany) in an incubator maintained at 37 °C with a humidified gas atmosphere of 5% CO2 in air [11]. Monitoring of EGFP fluorescence. The expression of EGFP in preand post-implantation embryos or in neonatal mice was examined under a fluorescence inverted microscope (IX-70, Olympus, Tokyo, Japan) and stereozoom microscope (SZX-RFL2, Olympus), respectively, with a GFP blue-filter set (470–490 nm). Fluorescent pictures were photographed using a 400 ASA Kodak film with the photomicrographic unit (PM-30, Olympus) attached to the microscopes. Embryonic EGFP fluorescence was monitored at every 12 h upto 120 h of culture. Fluorescence patterns were documented and EGFP-expressing ‘green’ embryos were separated from the non-expressing embryos and cultured individually along with ’9 non-injected similar stage embryos in separate culture drops. Categorization of EGFP-expressing embryos by visual and confocal microscopy methods. Cultured embryos were observed for green fluorescence and those exhibiting uniform green fluorescence were separated from the rest. They were visually classified into three categories
1031
based on their fluorescence intensity grades, i.e., faint, moderate, and intense and given subjective scores of 2, 4, and 6, respectively. Categorized embryos were cultured separately along with similar stage noninjected embryos. The total numbers of embryos were 10–15 in each 50 ll drop of M16 medium. We depended on assessing relative levels of EGFP in embryos only by fluorescence since it is a non-invasive method and the assessed embryos could be used for culture experiments, unlike immunoblot quantification, which would be a terminal experiment. Though the segregation of EGFP-expressing embryos could be performed by confocal microscopy, it is unsuitable for assessing blastocyst development during actively growing embryo cultures. Therefore, subjective method of fluorescence assessment was adopted as described above. However, to rule out the possibility of investigator’s bias in the assessment of fluorescence intensity, a double blind study was conducted on another set of embryos. For this, 2- and 4-cell embryos, developed from EGFP-injected pronuclear-stage-eggs, exhibiting uniform EGFP-expression, were transferred individually to 5 ll drops of M2 medium. Each embryo was assigned a particular identity number and classified by visual method as faint- (score: 2), moderate- (4), and intense- (6) fluorescence categories. These embryos were randomized by an independent investigator, given a different identity number and then subjected to confocal laser scanning microscopy (TCS-SP, Leica Microsystems, Heidelberger GmbH, Germany) with excitation at 488 nm and detection at 500–530 nm band pass filters. The fluorescence image of each embryo was recorded and its fluorescence intensity was quantified in arbitrary units (AU) using the Leica confocal software (LC Sv. 585). Based on the range of fluorescence intensities, embryos were categorized into three groups as low-, moderate-, and intense-fluorescing embryos and compared with those of the subjective method after decoding identity numbers. Statistical analysis. All experiments were performed for a minimum of three times. Results are expressed as means SEM. Percentage values were subjected to arcsin transformation before statistical analysis. Comparisons between groups were made by the unpaired Student’s t test (PRISM Graph Pad version 2; Graph Pad Software, San Diego, CA). To determine the correlation between the fluorescence intensities and developmental competence of EGFP-expressing embryos, a regression analysis was performed and correlation coefficient (r) was calculated; r value closer to )1 indicated a strong negative relationship.
Results Embryonic EGFP-expression during preimplantation development, following microinjection The CbA-EGFP-transgene was microinjected into 453 pronucleate-stage-eggs. Of these, 372 (82%) eggs, surviving microinjection, were cultured and monitored for EGFP-expression. Green fluorescence was detected in 117 (32%) embryos, initially at 24–36 h post-injection, i.e., at the 2-cell stage; some non-starters also exhibited EGFP-expression (Fig. 1I). The proportion of EGFPexpressing embryos increased through development and was maximum (30%; 111/372) at 72 h post-injection, i.e., at the morula stage. Incidentally, 15% (18/117) of EGFP-expressing embryos showed only transient expression. Embryos exhibited variable EGFP-expression patterns. Out of 117 EGFP-expressing embryos, 75 (64%) showed uniform green fluorescence, while 26 (22%)
1032
V. Devgan et al. / Biochemical and Biophysical Research Communications 313 (2004) 1030–1036
Fig. 1. Embryonic EGFP-expression, following pronuclear-microinjection of the EGFP-transgene. Embryos exhibited (I) uniform, (II) mosaic, and (III) differential EGFP-expression patterns. Panel I shows photomicrographs of 1-cell (A,I), 2-cell (B,J), 4-cell (C,K), 8-cell (D,L), morula (E,M), and expanded (F,N), hatching (G,O), and hatched (H,P) blastocysts. Panels A–H and I–P correspond to differential interference contrast (DIC) and their corresponding fluorescence pictures, respectively. Panel II shows photomicrographs of 2-cell (A, J), 4-cell (B–D, K–M), 8-cell (E–F, N–O), morula (G,P), and expanded blastocyst (H–I, Q–R). Panels A–I and J–R correspond to DIC and their corresponding fluorescence pictures, respectively. Panel III shows photomicrographs of 2-cell (A–C, G–I, M–O, and S–U) and 4-cell (D–F, J–L, P–R, and V–X) embryos exhibiting faint (A,M and D,P) or moderate (B,N and E,Q) or high (C,O and F,R) green fluorescence intensity and differential EGFP-expression within sister blastomeres (G– L, S–X). Panels A–L and M–X correspond to bright field and their corresponding confocal fluorescence images, respectively. All the pictures are photographed with 20 objective. Magnification bars correspond to 50 lm and are shown in panels I—O (identical in panels A–G and I–O of I), I—P (identical in panel H of I), II—R (identical in all panels of II), and III—X (identical in all panels of III).
Fig. 3. A composite picture showing EGFP-expression during the entire period of EGFP-transgenic mouse development. EGFP-expression is shown in the oocyte (O) but the mature sperm (S) does not exhibit EGFP-expression. EGFP fluorescence is uniform in preimplantation embryos including 1-cell (0.5 dpc: days post-coitum), 2-cell (1.5 dpc), 4-cell (2 dpc), 8-cell (2.5 dpc), morula (3 dpc), and blastocyst (4 dpc) stages, post-implantation embryos including egg-cylinder- (6.5 dpc), late somite- (8.5 dpc), early limb bud- (9.5 dpc, 10.5 dpc), late limb bud- (11.5 dpc, 12.5 dpc), and late fetal(14.5 dpc, 16.5 dpc) stages and neonatal (D1: Day 1) and adult mice (1M: 1 month).
V. Devgan et al. / Biochemical and Biophysical Research Communications 313 (2004) 1030–1036
showed mosaicism, and 16 (14%) showed differential fluorescence within sister blastomeres. Photomicrographs of all stages of preimplantation embryos exhibiting uniform or mosaic EGFP-expression are shown in Figs. 1I and II, respectively. Interestingly, embryos, with uniform EGFP-expression, showed different levels of green fluorescence intensities. About 32% (24/75), 40% (30/75), and 28% (21/75) of these embryos showed faint, moderate, and intense levels of green fluorescence, respectively. Photomicrographs of 2-cell and 4-cell embryos exhibiting different levels of fluorescence intensities and differential fluorescence within sister blastomeres are shown in Fig. 1III. Effect of EGFP-expression on blastocyst development To evaluate the developmental potential and viability of EGFP-transgenic embryos, EGFP-transgeneinjected, T10 E0:1 buffer-injected, and non-injected pronuclear-stage-eggs were cultured and their blastocyst development was assessed after 120 h of culture (Table 1). Non-injected and buffer-injected embryos showed 91.3 2.3% and 81.3 2.3% blastocyst development, respectively. In contrast, only 69.4 1.9% EGFP-injected embryos developed to blastocysts. The percentage of blastocyst development of EGFP-injected embryos was significantly lesser than that of either buffer-injected or non-injected embryos (Table 1; P < 0:005). Moreover, when EGFP-injected embryos were exposed to blue light for 30 s at every 12 h of culture for observing fluorescence of EGFP-expression, 67.2 1.8% blastocyst development was achieved (Table 1). The percentage of blastocyst development was comparable in both categories of EGFP-injected embryos, i.e., without and with blue-light exposure during culture (Table 1; P > 0:05). Among EGFP-injected embryos, non-fluorescing embryos showed 77.3 1.8% blastocyst development, while EGFP-expressing, ‘green fluorescing’ embryos showed only 43.9 1.6% blastocyst development. The percentage of blastocyst development of EGFP-expressing embryos was significantly lesser (P < 0:0001) than that of non-expressing-EGFPinjected embryos (Table 1).
1033
Effect of differential level of EGFP-expression on blastocyst development To delineate relationship between the level of EGFPexpression and developmental competence of preimplantation embryos, uniform EGFP-expressing embryos were visually classified into faint (score 2), moderate (4), and intense (6) categories based on their fluorescence. Interestingly, a negative correlation (r ¼ 0:99) was observed between the percentages of blastocyst development and the levels of EGFP-expression. Embryos with faint fluorescence showed 83.1 5.3% (23/28) blastocyst development, while embryos with moderate and intense fluorescence showed 50 5% (21/42) and 9.5 3.9% (3/29) blastocyst development, respectively (Fig. 2A). The percentage of blastocyst development in the faint category was significantly higher (P < 0:003) than those in the moderate and intense categories. Data on the double blind study involving visual and confocal microscopy methods are shown in Fig. 2B. Confocal microscopy confirmed that embryos had varying extents of fluorescence intensities (1–225 AU). On the basis of quantified fluorescence intensity, embryos were divided into three groups, i.e., faint (1–75 AU), moderate (76–150 AU), and intense (151–225 AU) and their groups were compared with the visual method of categorization. Incidentally, 66 out of 75 (88%) were correctly assigned to the same categories by both methods. The corrected percentages of blastocyst development were calculated from the observed proportion of mismatch using inverting 3 3 matrix of coefficient and found to be negatively correlated with the levels of EGFP-expression (data not shown). Effect of EGFP-expression on post-implantation development To investigate the impact of embryonic EGFPexpression during post-implantation development, EGFP-injected embryos were cultured to blastocysts. EGFP-expressing ‘green’ or non-expressing ‘non-green’ blastocysts were segregated and transferred independently to uteri of day 3 pseudopregnant (C57BL/6)
Table 1 Development of blastocysts from EGFP-injected-pronuclear-stage mouse embryos Treatment
No. of embryos cultured
Blastocysts (%)
Non-injected T10 E0:1 buffer-injected EGFP-injected: Without observation at 480 nm With observation at 480 nm Non-expressing EGFP-expressing
180 163
164 (91.3 2.3a;b;c ) 131 (81.3 2.3a;d;e )
186 289 198 91
129 (69.4 1.9b;d ) 191 (67.2 1.8c;e ) 152 (77.3 1.8f ) 39 (43.9 1.6f )
Values in parentheses represent means SEM, from 10 replicate experiments. Values with identical superscripts differ significantly. a: P < 0:01; b,c,f: P < 0:0001; and d,e: P < 0:002.
1034
V. Devgan et al. / Biochemical and Biophysical Research Communications 313 (2004) 1030–1036
EGFP-expression in pre- and post-implantation embryos of EGFP-transgenic mice Embryos were recovered, from wild-type females mated with hemizygous transgenic males, on 0.5. 1.5, 2, 2.5, 3, and 4 dpc and their EGFP-expression patterns were assessed. EGFP-expression was initially detected, in about half the embryos (54.2%; 26/48) at the compacting 8-cell stage (data not shown). By morula stage, EGFP-expression was clearly visible and thereafter, fluorescence was stable and uniform. On the other hand, all preimplantation stage embryos, recovered on 0.5, 1.5, 2, 2.5, 3, and 4 dpc of hemizygous transgenic females mated with wild-type males, exhibited EGFP-expression (Fig. 3). When freshly recovered day 1 embryos were cultured, about half the embryos (44.2%; 23/52) began to lose green fluorescence by 96 h of culture and they completely lost the fluorescence by 120 h (data not shown). Incidentally, in all fluorescing embryos, EGFPexpression was uniform and faint (Fig. 3). During post-implantation development, EGFP-expression was found to be ubiquitous in embryonic (Fig. 3) and extra-embryonic tissues (data not shown). Carcass analysis of fetus (14.5 dpc) and adult (1 month) transgenic mice showed that all organs derived from the three germ-cell lineages were ubiquitously green fluorescent (data not shown). Photomicrographs of all stages of mouse development exhibiting EGFP-expression are shown in Fig. 3.
Fig. 2. (A) Correlation between the fluorescence intensity and blastocyst development of uniformly EGFP-expressing embryos. Blastocyst development of embryos with faint (), moderate (P), and intense (j) green fluorescence intensity is shown. Results represent means SEM of five replicate experiments. Values with identical superscripts differ significantly. P values: a < 0:002; b; c < 0:0005. (B) Assessment of fluorescence intensity of uniformly EGFP-expressing embryos by confocal microscopy. Fluorescence intensities of green fluorescing embryos (n ¼ 77) were quantified in arbitrary units (AU) using Leica confocal software. These embryos were divided in three groups based on their quantified fluorescence intensity (FI), i.e., faint (FI ¼ 1– 75 AU), moderate (FI ¼ 76–150 AU), and intense (FI ¼ 151–225 AU).
recipients. In this regard, 84 EGFP-expressing and 208 non-expressing blastocysts were transferred to 8 and 10 pseudopregnant recipients, respectively. Of these, 3 (38%) and 5 (50%) females became pregnant and delivered at term 5 (15.4 4.8%; 5/32) and 23 (22.4 1.2%; 23/102) pups, respectively. The genotypic (PCR and genomic Southern) analysis confirmed that all pups born from green blastocysts–uterine-transfer were transgenic and none of the pups born from non-green blastocysts– uterine-transfer were transgenic [10]. The percentage of pups born per transferred embryos from EGFP-expressing blastocysts transfers was similar (P > 0:05) to that of non-expressing blastocyst transfers.
Discussion In the present study, we demonstrate, for the first time, that EGFP exhibits expression-level-dependent inhibition of preimplantation embryo development and hyper-expression reduces developmental competence and potential viability of mouse embryos. However, faint–moderate embryonic EGFP-expression is compatible with development of microinjected-embryos and of those derived from the EGFP-expressing transgenic “green” mice. Our data on the comprehensive analysis of embryonic expression patterns of the EGFP-transgene during mouse development are quite significant in terms of our understanding of its behavior during early development and its potential use as a reporter gene (for real-time imaging analysis) while studying regulation of gene expression during preimplantation embryo development. Developmental competence of preimplantation embryos correlates with the level of the transgene expression. Embryos with the lowest level of EGFP-expression show maximum blastocyst development and those expressing the highest level of EGFP show poorest blastocyst development. Though embryonic expression of EGFP has been studied earlier [8], the impact
V. Devgan et al. / Biochemical and Biophysical Research Communications 313 (2004) 1030–1036
(inhibition) on preimplantation embryo development has so far not been thoroughly investigated and the earlier work [8] has not reported such an effect in mammalian early embryos. However, our observation on the inhibition of development of preimplantation embryos is quite consistent with those of others who showed EGFP-cytotoxicity in a few mammalian cell lines [12,13] and cardiomyopathy in transgenic mice [14]. Incidentally, toxicity due to overexpression of GFP has been reported in transgenic plants [15]. It could be argued that sub-optimal culture conditions and/or microinjection-induced mechanical trauma [16] may result in the observed poor development of blastocysts, expressing high level of EGFP. We, however, discount this possibility since T10 E0:1 bufferinjected-eggs showed significantly higher blastocyst development when compared to that of EGFP-injected embryos (Table 1). Moreover, M16 medium, used in the study, supports comparable and maximal development of high quality blastocysts from EGFP-injected or noninjected pronuclear-stage mouse embryos [11]. Similarly, the possibility of detrimental effect of blue-light exposure is also ruled out, since the EGFP-injected embryos show similar percentages of blastocyst development without or with blue-light exposure (Table 1). Another possibility to be considered is that the observed inhibition of cleavage of embryonic cells by hyper-expression of EGFP (or any transgene) could be due to the change in the gene expression profile of cells [13]. But, how are these changes triggered by the transgene (EGFP) and what could be the consequences of these changes remain unclear and require further studies. To the best of our knowledge, there is no published report on the possible EGFP-induced metabolic perturbation or effect on organelle function in embryonic cells. The low blastocyst development may, in part, be attributable to either an insertional mutagenesis and/or repression of developmentally essential endogenous gene(s) by hyper-expression of the (EGFP) transgene, resulting in reduced developmental competence. Towards understanding these possibilities, it would be quite interesting to have a comparative study using Renilla reniformis GFP (hrGFP) vis- a-vis Aequoria victoria GFP, since the former is known to possess negligible or low cellular toxicity [17]. It is interesting that the developed blastocysts expressing EGFP did not affect post-implantation development as well as the reproductive outcome of transgenic animals. Following pronuclear-microinjection, the preimplantation phase of embryo development could be most vulnerable to possible embryotoxicity due to hyperexpression of EGFP. The preimplantation phase of embryo development could be acting as a selection strategy to allow only faint–moderate EGFP-expressing embryos to develop to blastocysts, which are capable of completing normal development. It may be necessary to screen/select
1035
developmentally competent EGFP-expressing transgenic-embryos (blastocysts) prior to uterine-embryo-transfer, thereby improving the percentage of EGFP-transgenic offspring. Following this procedure, i.e., embryo-culturebased transgenesis (ECBT), we have successfully generated five EGFP-transgenic founder mice and observed that ubiquitously expressing EGFP-transgenic founders develop normally to fertile adults and produce transgenic lines [10,11]. The relatively lower rate of pregnancy with EGFP-expressing blastocysts when compared to that of the non-EGFP-expressing blastocysts (see results) could be due to a less number of embryos used for uterine transfer in the former case (10 vs. 20). However, the percentage of pups born per transferred embryos was similar in both categories. Despite using a universal and ubiquitous promoter and enhancer, we observed a number of interesting variations in the pattern of transgene expression in terms of the time of onset of gene expression and relative levels of EGFP-expression in various stages of preimplantation embryos (Fig. 2). The transient EGFPexpression may be due to the epichromosomal, non-integrated transgene, while the mosaic expression could be due to the integration of the transgene after the first round of DNA replication [18]. However, the observed difference in the levels of green fluorescence intensities is possibly due to different copy numbers of the transgene and/or to a chromosomal position effect [19]. Our earlier data show that the high copy number of EGFP in a single genomic locus may be responsible for repeat-induced gene silencing [10]. Our observation on the differential EGFP-expression within sister blastomeres is consistent with a recent report [20]. This pattern of expression could possibly be due to: (i) independent transgene integration events in individual blastomere during sequential cell division, (ii) EGFP-expression is developmentally regulated and each blastomere has different developmental fate, and (iii) non-integrated transgene distributed randomly during cell division. However, the second reason is unlikely, because, all blastomeres of mammalian embryos are totipotent upto the morula stage. We also observed that unlike spermatogenic cells, all oocytes (100%), derived from hemizygous EGFP-transgenic females, show EGFP-expression during the entire meiotic maturation process (data not shown). The green fluorescence in 100% embryos, recovered from hemizygous transgenic females mated with wild-type males, could be due to a residual, maternally expressed EGFP during oogenesis. But, about 50% of embryos lose green fluorescence during culture by 24 h (blastocyst; data not shown) or 120 h (1-cell), possibly due to degradation of maternally derived transcript and/or protein. Rest of the embryos retaining fluorescence through the ensuing developmental period is due to de novo embryonic expression of the inherited EGFP-transgene, possibly from
1036
V. Devgan et al. / Biochemical and Biophysical Research Communications 313 (2004) 1030–1036
the compacting 8-cell stage, similar to that observed with embryos having paternally derived EGFP gene. It is quite interesting to note that embryos derived from transgenic ‘green’ mouse lines show faint green fluorescence and the percentage of blastocyst development of embryos derived from EGFP-transgenic mice is quite similar to that of EGFP-injected embryos with faint green fluorescence (data not shown). Furthermore, during post-implantation development, all embryonic stages show ubiquitous EGFP-expression and throughout the neo-natal development. In conclusion, following microinjection, embryos exhibit variations in the pattern of the EGFP-expression illustrating the complex expression behavior of the transgene during early development. Hyper-expression of EGFP affects developmental ability and potential viability of preimplantation embryos. However, developed embryos, expressing EGFP at faint–moderate levels and/or those from EGFP-transgenic mice, are unaffected and are completely compatible with normal development. Our findings are quite valuable in providing new insights into using EGFP as an in vivo fluorescent marker for studying regulation of gene expression during mammalian preimplantation development. Besides, the EGFP-transgenic ‘green’ mouse is a rich source of fluorescent-marked (embryonic/somatic) stem cells for studying regulation of lineage-specific cell differentiation during mouse development.
Acknowledgments Financial support from the Department of Biotechnology, New Delhi, is gratefully acknowledged. The authors are thankful to Dr. S.M. Totey for advice and help; Ms. G.V. Sireesha, Ms. M. Sarkar, Mr. S. Nyati, and Dr. Uday kumar for their excellent technical support. Our thanks are also due to Prof. N. Joshi for advice on statistical evaluation of the data and to Ms. M.S. Padmavathi for help in the preparation of the manuscript.
References [1] S. Takeda, Y. Toyoda, Expression of SV40-lacZ gene in mouse preimplantation embryos after pronuclear microinjection, Mol. Reprod. Dev. 30 (1991) 90–94. [2] E.M. Thompson, P. Adenot, F.I. Tsuji, J.P. Renard, Real time imaging of transcriptional activity in live mouse preimplantation embryos using a secreted luciferase, Proc. Natl. Acad. Sci. USA 92 (1995) 1317–1321. [3] B.P. Cormack, R.H. Valdivia, S. Falkow, FACS-optimized mutants of the green fluorescent protein (GFP), Gene 173 (1996) 33–38.
[4] N. Kirchhof, J.W. Carnwath, E. Lemme, K. Anastassiadis, H. Scholer, H. Niemann, Expression pattern of Oct-4 in preimplantation embryos of different species, Biol. Reprod. 63 (2000) 1698– 1705. [5] K.A. Molyneaux, J. Stallock, K. Schaible, C. Wylie, Time-lapse analysis of living mouse germ cell migration, Dev. Biol. 240 (2001) 488–498. [6] M. Okabe, M. Ikawa, K. Kominami, T. Nakanishi, Y. Nishimune, ‘Green mice’ as a source of ubiquitous green cells, FEBS lett. 407 (1997) 313–319. [7] A.K. Hadjantonakis, M. Gertsenstein, M. Ikawa, M. Okabe, A. Nagy, Generating green fluorescent mice by germline transmission of green fluorescent ES cells, Mech. Dev. 76 (1998) 79–90. [8] T. Takada, K. Iida, T. Awaji, K. Itoh, R. Takahashi, A. Shibui, K. Yoshida, S. Sugano, G. Tsujimoto, Selective production of transgenic mice using green fluorescent protein as a marker, Nat. Biotechnol. 15 (1997) 458–461. [9] J.T. Keiser, P.M. Jobst, A.S. Garst, J.T. Boone, C.B. Geyer, C. Phelps, D.L. Ayares, R.L. Page, Preimplantation screening for transgenesis using an embryonic specific promoter and green fluorescent protein, Cloning 3 (2001) 23–30. [10] V. Devgan, M. Thomas, K.S. Ullas, M.R.S. Rao, P.B. Seshagiri, Embryo culture-based generation of enhanced green fluorescent protein-transgenic mice, Biochem. Biophys. Res. Commun. 303 (2003) 994–1001. [11] V. Devgan, P.B. Seshagiri, Successful development of viable blastocysts from enhanced green fluorescent protein transgenemicroinjected mouse embryos: comparison of culture media, Mol. Reprod. Dev. 65 (2003) 269–277. [12] H.S. Liu, M.S. Jan, C.K. Chou, P.H. Chen, N.J. Ke, Is green fluorescent protein toxic to living cells?, Biochem. Biophys. Res. Commun. 260 (1999) 712–717. [13] P. Vaillancourt, B. Rogers, Y. Wang, Do not let a toxic GFP to skew your results!, Strategies 14 (2001) 86–87. [14] W.Y. Huang, J. Aramburu, P.S. Douglas, S. Izumo, Transgenic expression of green fluorescence protein can cause dilated cardiomyopathy, Nat. Med. 6 (2000) 482–483. [15] J. Haselowff, B. Amos, GFP in plants, Trends Genet. 11 (1995) 328–329. [16] P. Chrenek, A. Makarevich, D. Vasicek, J. Laurincik, J. Bulla, T. Gajarska, J. Rafay, Effects of superovulation, culture and microinjection on development of rabbit embryos in vitro, Theriogenology 50 (1998) 659–666. [17] K. Felts, B. Rogers, K. Chen, H. Ji, J. Sorge, P. Vaillancourt, Recombinant Renilla reniformis GFP display low toxicity, Strategies 13 (2000) 85–87. [18] A.W.S. Chan, G. Kukolj, A.M. Skalka, R.D. Bremel, Timing of DNA integration, transgenic mosaicism, and pronuclear microinjection, Mol. Reprod. Dev. 52 (1999) 406–413. [19] T. Takada, K. Iida, K. Akasaka, H. Yasue, R. Torii, G. Tsujimoto, M. Taira, H. Kimura, Evaluation of heterologous insulator function with regard to chromosomal position effect in the mouse blastocyst and fetus, Mol. Reprod. Dev. 57 (2000) 232– 237. [20] S.Y. Medvedev, T. Tokunaga, R.M. Schultz, T. Furukawa, T. Nagai, M. Yamaguchi, M. Hosoe, A.F. Yakovlev, S. Takahashi, Y. Izaike, Quantitative analysis of gene expression in preimplantation mouse embryos using green fluorescent protein reporter, Biol. Reprod. 67 (2002) 282–286.