SEARCH

SEARCH BY CITATION

Keywords:

  • Cre/loxP;
  • zp3;
  • maternal;
  • oocyte;
  • ovary;
  • GFP;
  • RFP;
  • germline

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. REFERENCES

In this communication, we report the generation of a cre transgenic zebrafish line under an oocyte-specific promoter, zp3. The transgenic line Tg(zp3:cre; krt8:rfp) also contains a co-integrated rfp transgene under the skin epithelial promoter krt8 to allow selection of cre transgenic fish based on RFP fluorescence in the skin. We demonstrated in this transgenic line that cre mRNA was specifically expressed in growing oocytes like endogenous zp3 mRNA. When Tg(zp3:cre; krt8:rfp) was crossed with a loxP transgenic line, the floxed DNA was specifically eliminated from female, but not male, germline. Tg(zp3:cre; krt8:rfp) fish also have maternal cre mRNA in early embryos to cause Cre-mediated recombination; this feature can be used to activate other loxP transgenic lines in early embryos. Furthermore, after crossing with another loxP transgenic line, Tg(EF:loxP-mCherry-loxP-egfp), we confirmed that our cre line was capable of activating a loxP-blocked EGFP reporter gene by both maternal and oocyte-expressed Cre. Developmental Dynamics 237:2955–2962, 2008. © 2008 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. REFERENCES

In genetic analyses of gene expression and function, it is important to develop conditional activation and inactivation systems as many genes may cause a pleiotropic or even lethal effect when they are constitutively expressed. So far several conditional transgenic systems have been popularly used, such as GAL4-UAS binary transgenic systems, Tet-on/Tet-off inducible system and Cre/loxP recombination system (Lewandoski,2001). The Cre/loxP system is currently the most powerful conditional system and it mediates site-specific DNA recombination between two loxP sites by a site-specific DNA recombinase, Cre, which recognizes the loxP sites and excises the floxed sequence between the two loxP sites with the same orientation (Sauer,1998). Now the Cre/loxP system has been widely used in genome manipulation in mouse genetics, such as conditional gene knock-out (Lobe and Nagy,1998), transgene excision from germline (Bunting et al.,1999; Lomeli et al.,2000; Lewandoski et al.,1997), tissue-specific activation or inactivation of targeted gene (Sauer,1998, Lasko et al., 1990; Kasim et al.,2004), and tracing cell lineage (Zinyk et al.,1998; Herrera,2000), etc. Recently the Cre/loxP system was also used to induce RNAi in a spatial, temporal manner for gene silencing, which may be potentially used for gene therapy (Kasim et al.,2004).

However, so far, little work has been carried out for applying the Cre/loxP system in other animal systems. Previously the Cre/loxP system has been tested and applied for tracing muscle cell lineage in tail regeneration in Xenopus (Ryffel et al.,2003; Gargioli and Slack,2004; Waldner et al.,2006). In zebrafish, we and others previously reported the excision of loxP-flanked transgenes by injection of cre mRNA into zebrafish embryos (Pan et al.,2005; Langenau et al.,2005). Later Cre-mediated site-specific recombination was also tested in a heat shock inducible Tg(HSP70:EGFP-cre) transgenic zebrafish line (Thummel et al.,2005; Le et al.,2007; Feng et al.,2007). Nevertheless, tissue-specific excision of loxP-flanked transgene has not been demonstrated in zebrafish because of the shortage of tissue-specifically expressed cre transgenic lines. In the present study, we attempted to test the possibility of germline excision of transgenes in zebrafish using the Cre/loxP system. cre transgenic zebrafish lines under the promoter of zebrafish zp3, an oocyte-specific gene, were generated. By in situ hybridization and reverse transcriptase-polymerase chain reaction (RT-PCR), we detected oocyte-specific expression of cre RNA in female transgenic fish. After crossing with a previously generated loxP transgenic line, Tg(mylz2:loxP-EGFP-loxP)gz3 (Pan et al.,2005), we observed the expected excision of the loxP-flanked egfp transgene through female, but not male, germline in F2 generation. Thus we demonstrated the feasibility of female germline excision of floxed genes in future in zebrafish.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. REFERENCES

Transient Expression of pCMV-CRE and pKRT8-LoxP-EGFP-LoxP-RFP

To develop Cre/loxP transgenic zebrafish, two plasmid DNA constructs were made and analyzed by transient transgenic expression assay in zebrafish embryos. The first one was pKRT8-loxP-EGFP-loxP-RFP (pKRT8-LGLR for short), in which a 2.2-kb keratin8 (krt8) promoter (Ju et al.,1999; Gong et al.,2002) was placed in front of the floxed EGFP reporter gene that was followed by the RFP reporter gene (Fig. 1A). The second construct, pCMV-CRE, was made by combination of the ubiquitous CMV promoter and cre cDNA (Fig. 1A). When pKRT8-LGLR was injected into zebrafish embryos at 1–2 cell stage, expression of GFP, but not RFP, was observed mostly in the skin epithelial cells (Fig. 1B), similar to our earlier observation by direct injection of pKRT8-GFP (Ju et al.,1999; Gong et al.,2002). When it was co-injected with pCMV-CRE, skin-specific expression of both GFP and RFP was observed in >83% of the co-injected embryos (Fig. 1C,D); thus, the floxed egfp gene has been successfully excised in the injected construct by the transiently expressed Cre recombinase under the CMV promoter and the downstream RFP gene was activated. This experiment demonstrated the functionality of Cre-mediated recombination of pKRT8-LGLR.

thumbnail image

Figure 1. Transient expression of pKRT8-LGLR in zebrafish embryos by co-injection of pCMV-CRE or pZP3-CRE. A: Schematic representation of chimeric DNA constructs, pKRT8-LGLR, pCMV-CRE, and pZP3-CRE. B: A 48 hours postfertilization (hpf) embryo injected with pKRT8-LGLR only, showing only GFP expression. C,D: A same 48 hpf embryo co-injection of pKRT8-LGLR and pCMV-CRE, showing both green fluorescent protein (GFP) expression under a blue light (C) and red fluorescent protein (RFP) expression under a yellow light (D). E,F: A same 48 hpf embryo co-injection of pKRT8-LGLR and pZP3-CRE, showing both GFP (E) and RFP (F) expression. Scale bars = 500 μm.

Download figure to PowerPoint

Generation of cre Transgenic Zebrafish Under the Oocyte-Specific zp3 Promoter

To investigate the feasibility of germline excision of transgene in zebrafish, we attempted to generate a cre transgenic zebrafish line under the oocyte-specific zp3 promoter. Thus, a chimeric plasmid pZP3-CRE, in which the zebrafish zp3 promoter (Liu et al.,2006) was used to drive the cre gene, was made. To develop cre transgenic lines, pKRT8-LGLR was co-injected with pZP3-CRE. Based on our and other's experience, when two DNA constructs are co-injected, they usually form heterogeneous concatemers and co-integrate into host chromosomes (e.g., Muller et al.,1997, 1999; Wan et al.,2002). Thus, the use of pKRT8-LGLR was both to facilitate screening of transgenic offspring by observation of living color reporter gene expression and to test the Cre/loxP system in cre transgenic lines directly. If the two DNA constructs are transmitted to the F1 generation through a female founder, the floxed egfp will be excised by the oocyte-specific expression of Cre and RFP expression in the skin epithelia is expected. If they are inherited to the F1 generation through a male founder, the floxed egfp will not be excised as there is no Cre expression in the male germ cells, and thus the GFP expression in the skin epithelia will be expected.

When pZP3-CRE and pKRT8-LGLR were co-injected into zebrafish embryos, co-expression of both GFP and RFP were observed in approximately 85% of the injected embryos after 24 hours postfertilization (hpf; Fig. 1E,F), similar to embryos co-injected with pKRT8-LGLR and pCMV-CRE. As no RFP was observed in embryos injected with pKRT8-LGLR alone (Fig. 1B), the RFP expression in the co-injected embryos was likely due to the excision of floxed egfp by leaked expression of Cre from the injected pZP3-CRE construct.

To develop stable cre transgenic lines, the injected embryos with GFP/RFP expression were raised to adulthood and crossed with each other or with wild-type fish to screen for F1 offspring by both observation of GFP/RFP expression and PCR detection of cre DNA. Finally, two female fish from over 160 founders were identified to be capable of producing cre-positive F1 offspring and thus two stable cre transgenic lines (T22 and T24) were obtained. However, no GFP expression was observed in the two transgenic lines. RFP expression was observed in line T24 but not in line T22. Thus, only line T24 was focused in the present study.

In line T24, strong maternal RFP was observed in blastomere of early embryos. RFP expression in skin epithelial cells was observed in 1–3 days postfertilization fish embryos, similar to our previously reported transgenic lines Tg(krt8:egfp) (Gong et al.,2002). Cre DNA was detected in embryos with RFP expression but not in embryos without RFP expression (Fig. 2B). It is interesting to note that no GFP expression was observed in these F1 embryos, indicating that Cre-mediated recombination has already occurred. This could occur either during oogenesis in the female founder or even before the transgene integration caused by leaked expression of Cre in injected embryos (Fig. 1F).

thumbnail image

Figure 2. Co-existence of red fluorescent protein (RFP) expression and cre DNA in Tg(zp3:cre, krt8:rfp) F1 embryos. A: Skin-specific RFP expression in Tg(zp3:cre, krt8:rfp) embryos: 24 hours postfertilization (hpf; upper) and 72 hpf (lower). B: Presence of cre DNA in RFP-positive embryos. Genomic DNAs were prepared from 5 days postfertilization embryos individually and cre DNA was detected by polymerase chain reaction (PCR). The picture shows nine RFP-positive embryos and one RNA-negative embryo. P, positive PCR control with pZP3-CRE template; N, negative PCR control without template; M, 100-bp DNA marker (Promega). C: Demonstration of the Cre-mediated excision of floxed egfp in Tg(zp3:cre, krt8:rfp) transgenic fish. PCR was performed for both RFP-positive and -negative embryos using primers P1 and R1 (upper) as indicated in the diagram with expected sizes of PCR products before and after Cre-mediated excision. Presence of cre DNA was also amplified by PCR using a pair of cre-specific primers (lower). The 850-bp recombined fragment and cre DNA are co-existed in skin RFP expressing embryos (RFP+) but not in the non-RNA expressing embryos (RFP−). Primers P1+G1 amplified 750-bp unexcised gfp fragment. P, positive PCR control with templates pKRT8-LGLR (upper) or pZP3-CRE (lower); N, negative PCR control without template; M, 1-kb DNA marker. In the diagram, LoxP, EGFP, and RFP DNAs are indicated by blue triangles, green and red boxes, respectively. Scale bars = 500 μm.

Download figure to PowerPoint

To determine Cre-mediated recombination in F1 fish, fin DNA was prepared from F1 adult fish and PCR was carried out to examine egfp excision with forward primer (P1) from the keratin8 promoter region and reverse primer (R1) from the end of rfp coding sequence (Fig. 2C). The expected bands were 1.6 kb and 850 bp, respectively, before and after Cre-mediated recombination (Fig. 2C). In all F1 fish with RFP expression, both the 850-bp rfp fragment and cre DNA were detected, confirming the Cre-mediated recombination as well as the co-presence of the two transgenic constructs (Fig. 2C). In contrast, no amplified DNA fragment was observed in non-RFP expressing fish because they did not contain either of the two transgenes. Furthermore, when primers (P1 and G1) targeted for amplification of un-excised GFP DNA was used, no fragment was amplified from both RFP-expressing and non-RFP expressing fish (Fig. 2C). Thus, the transgenic line presented here is named Tg(zp3:cre; krt8:rfp).

Oocyte-specific Expression of cre mRNA in Tg(zp3:cre, krt8:rfp) Transgenic Fish

To determine cre expression in the Tg(zp3:cre, krt8:rfp) fish, RT-PCR was used to examine embryonic expression of cre mRNA and it was found that cre mRNA was detectable from 2 hpf to 48 hpf but it disappeared by 72 hpf in embryos derived from female Tg(zp3:cre, krt8:rfp) fish (Fig. 3A). This is likely due to maternal persistence of cre mRNAs from oocytes as there was no maternal cre mRNA detected in early embryos collected from wild-type female crossed with transgenic males (data not shown). Consistent with this, we also detected maternal zp3 mRNA up to 24 hpf (data not shown). In adult transgenic fish, cre mRNA was detected only in female but not in male (Fig. 3B). In female transgenic fish, the ovary was the only organ to express cre mRNA and there was no detectable expression in nonovary tissues examined, such as the skin, muscle, gut, and liver (Fig. 3B). Whole-mount in situ hybridization of ovary from female adult transgenic fish indicated that cre expression was detected predominantly in small, developing oocytes but not in large, mature oocytes (Fig. 3C), consistent with the expression pattern of endogenous zp3 mRNA (Fig. 3D; Liu et al.,2006). Thus, the oocyte-specific expression of cre in Tg(zp3:cre, krt8:rfp) fish faithfully recapitulated the endogenous zebrafish zp3 expression.

thumbnail image

Figure 3. Cre RNA expression in Tg(zp3:cre, krt8:rfp) transgenic zebrafish. A,B: cre RNA expression in 2–72 hours postfertilization (hpf) embryos (A) and in adult transgenic fish (B) as detected by reverse transcriptase-polymerase chain reaction (RT-PCR). Embryonic stages in hpf (A) and tissue sources (B) are indicated at the top of each lane. Whole-male transgenic fish (WF♂) and selected tissues from female transgenic fish were used for RNA preparation. P, positive control with pZP3-CRE plasmid; N, negative control without template. β-actin probes were used for RT-PCR control. C,D: Ovary expression of cre mRNA in Tg(zp3:cre, krt8:rfp) transgenic female (C) and zp3 mRNA wild-type female (D). Whole-mount in situ hybridization detection was performed with isolated ovary tissues and both cre and zp3 mRNAs were detected in developing oocytes (arrows) but not in the mature oocyte.

Download figure to PowerPoint

Female Germ Line Excision of loxP-Flanked Transgene

To test the capability of female germ line excision of floxed transgene, Tg(zp3:cre, krt8:rfp) zebrafish were crossed with Tg(mylz2:loxP-EGFP-loxP)gz3 zebrafish that possesses a floxed egfp reporter gene under the muscle-specific mylz2 promoter (Pan et al.,2005). When a female hemizygous Tg(zp3:cre, krt8:rfp) fish was crossed with a male hemizygous Tg(mylz2:loxP-EGFP-loxP)gz3 fish, we expected 50% of GFP-expressing embryos if there was no maternally expressed Cre. However, we actually observed approximately 19% of embryos showing muscle-specific GFP expression, which was below the expected 50%, suggesting partial excision of floxed egfp transgene by maternally expressed Cre in early embryos, as confirmed by PCR assay (data not shown). Consistent with this, we observed both strong (∼20%) and weak (∼80%) GFP expression in these GFP-positive embryos (Fig. 4A). However, when a male hemizygous Tg(zp3:cre, krt8:rfp) transgenic fish was crossed with a female hemizygous Tg(mylz2:loxP-EGFP-loxP)gz3 transgenic fish, anticipated ∼50% offspring showed strong GFP expression in the muscle. As confirmed by PCR assay, there was no Cre-mediated excision detected (data not shown).

thumbnail image

Figure 4. Test of Cre-mediated recombination at the chromosome level by breeding of Tg(zp3:cre, krt8:rfp) and Tg(mylz2:loxP-EGFP-loxP)gz3. A: The 48 hpf embryos with strong, weak, and nil green fluorescent protein (GFP) expression from a cross between female Tg(zp3:cre, krt8:rfp) and male Tg(mylz2:loxP-EGFP-loxP)gz3. Percentages of each type of embryos based on GFP expression are indicated. B: Polymerase chain reaction (PCR) detection of Cre-mediated recombination in offspring of female double transgenic fish Tg(zp3:cre, krt8:rfp)/Tg(mylz2:loxP-EGFP-loxP)gz3 and male wild-type fish. In both skin RFP-expressing (cre transgenic) and non–RFP-expressing (cre non-transgenic) embryos (4 days postfertilization), approximately half of them showed Cre-mediated recombination as anticipated from the female germline excision while the other half were loxP non-transgenic individuals. The expected PCR products are 2.1 kb and 1.2 kb, respectively, before and after the Cre-mediated recombination. P, positive PCR control with pMYLZ2-loxP-EGFP-loxP plasmid that is used to generate Tg(mylz2:loxP-EGFP-loxP)gz3; N, negative PCR control without a template DNA; M, 1-kb DNA ladder. Scale bar = 500 μm.

Download figure to PowerPoint

The F1 double transgenic larvae Tg(zp3:cre, krt8:rfp)/Tg(mylz2:loxP-EGFP-loxP)gz3 as indicated by both skin-RFP and muscle-GFP expression were obtained from a cross between male Tg(zp3:cre, krt8:rfp) and female Tg(mylz2:loxP-EGFP-loxP)gz3 and grown up to breed with wild-type fish to test germline excision. When F1 male double transgenic fish were crossed with wild-type female, approximately 50% F2 offspring showed GFP fluorescence in the muscle as expected from their genotype without Cre-mediated excision (Table 1). When F1 female double transgenic fish were crossed with wild-type male, in eight pairs of the cross, all of their offspring had no GFP expression in muscle (Table 1). To detect Cre-mediated excision, PCR was performed for both cre-positive and -negative fry as indicated by presence and absence of visible skin-RFP expression respectively (also confirmed by PCR analysis, see Fig. 2B). As shown in Figure 4B, in both groups, approximately 50% of fry had the mylz2-gfp transgene and all of them showed the recombined form (1.2 kb) as compared to the nonrecombined form (2.1 kb). Thus, these observations indicated the successful Cre-mediated excision in female germline but not in male germline. However, we also found incompleted female germ line excision in two of 10 F1 double transgenic females as there were approximately 10% of their offspring observed with GFP expression in the muscle (Table 1).

Table 1. Offspring of F1 Double Transgenic Fish [Tg(zp3:cre, krt8:rfp)/Tg(mylz2:loxP-EGFP-loxP)gz3] Crossing With Wild-Type Fish
Fish pairGFP+GFP−% GFP+
  1. GFP, green fluorescent protein.

♂ F1 transgenic fish x ♀ wild-type fish
112412649.6
227427050.4
318619149.3
426025650.4
528228449.8
♀ F1 transgenic fish x ♂ wild-type fish
103680
204250
304620
403580
504300
604780
705130
804320
93834110.0
10262459.6

To further confirm the maternal and oocyte-specific excision, the Tg(zp3:cre, krt8:rfp) was also crossed with a newly generated loxP line, Tg(EF:loxP-mCherry-loxP-egfp), in which dual fluorescent protein reporter genes (mCherry and egfp) was under the control of Xenopus elongation factor 1 alpha promoter (EF; Emelyanov et al., unpublished observations). The cherry gene was flanked by two loxP elements and egfp would be activated when the mCherry fragment was excised. When a female Tg(zp3:cre, krt8:rfp) was crossed with a male Tg(EF:loxP-cherry-loxP-egfp), both mCherry and EGFP were detected (Fig. 5A,B). In Tg(zp3:cre, krt8:rfp)/Tg(EF:loxP-mCherry-loxP-egfp) double transgenic embryos, ∼24% of them show GFP expression at variable levels, thus confirming the activation of egfp gene by maternally expressed Cre. In comparison, no GFP expression was detected in offspring when a wild-type female or a Tg(zp3:cre, krt8:rfp) female was crossed with a male Tg(EF:loxP-mCherry-loxP-egfp) (Fig. 5C,D).

thumbnail image

Figure 5. Activation of loxP-blocked reporter gene expression. A,B: A 48 hours postfertilization (hpf) embryo from a cross between a female Tg(zp3:cre, krt8:rfp) and male Tg(EF:loxP-mCherry-loxP-egfp) to show mCherry (A) and enhanced green fluorescent protein (EGFP; B) expression. C,D: A 48-hpf embryo from a cross between a female wild-type fish and male Tg(EF:loxP-mCherry-loxP-egfp) to show mCherry (C) and GFP (D) expression. E,F: View of mCherry (E) and GFP (F) expression in the ovary from a 1.5-month-old female double transgenic fish Tg(Zp3:Cre, Krt8:RFP)/ Tg(EF:loxp-mCherry-loxp-egfp. G,H: View of mCherry (G) and GFP (H) expression in the ovary from a 1.5-month-old female Tg(EF:loxp-mCherry-loxp-egfp fish. Scale bars = 100 μm.

Download figure to PowerPoint

To demonstrate the activation of egfp gene by oocyte-specific Cre expression, offspring of male Tg(zp3:cre, krt8:rfp) and female Tg(EF:loxP-mCherry-loxP-egfp) were grown to approximately 1.5 months old. In the double transgenic female [Tg(zp3:cre, krt8:rfp)/ Tg(EF:loxP-mCherry-loxP-egfp)], GFP expressing oocytes were observed while there was no mCherry expression in these oocytes (Fig. 5E,F). In comparison, in Tg(EF:loxP-mCherry-loxP-egfp) female, no GFP expression was observed in the ovary or in any other tissue and only mCherry expression was observed (Fig. 5G,H). The mCherry expression is stronger in the ovary than in other tissues in Tg(EF:loxP-mCherry-loxP-egfp) (Fig. 5G). All of these observations from the crosses of Tg(zp3:cre, krt8:rfp) and the dual reporter loxP line further confirmed the possibility of activation of loxP-blocked transgene by both maternal Cre and oocyte-specific Cre.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. REFERENCES

Previously several studies demonstrated the success of Cre-mediated transgene excision in zebrafish (Pan et al.,2005, Thummel et al.,2005, Langenau et al.,2005; Le et al.,2007). In many of these studies, cre RNA was injected into fertilized eggs to achieve Cre-mediated excision of loxP-flanked transgene in early embryos. For example, Pan et al. (2005) detected the Cre-mediated excision of a floxed egfp transgene in the Tg(mylz2:loxP-EGFP-loxP)gz3 transgenic line by injection of cre mRNA. Again by injection of cre mRNA into eggs of rag2-loxP-dsRED2-loxP-EGFP-mMyc transgenic zebrafish lines, Langenau et al. (2005) demonstrated the conditional activation of the c-myc oncogene after Cre-mediated excision of floxed dsRed and thus induced T cell lymphoblastic leukemia. So far only one cre stable transgenic zebrafish line has been reported for conditional activation of transgene in zebrafish and this transgenic line was constructed using the heat-shock inducible hsp70 promoter (Thummel et al.,2005). Recently a similar Tg(hsp70:cre) transgenic line has been used successfully to activate a ras oncogene by crossing with Tg(B-actin-LoxP-EGFP-LoxP-kRASG12D) fish and observed multiple tissue tumors (Le et al.,2007). Similarly, the same (hsp70:cre) transgenic line was demonstrated to activate a c-myc oncogene and caused T-cell lymphoma/leukemia (Feng et al.,2007). In the present work, we added another functional cre transgenic zebrafish line under the female germline specific promoter from a zp3 gene. Our present transgenic line should be valuable for conditional expression of transgenes in oocytes and this may be particularly useful in analysis of maternal effect genes.

In the newly generated Tg(zp3:cre, krt8:rfp) transgenic line, the cre gene was driven by our previously characterized oocyte-specific zp3 promoter (Liu et al.,2006) and an rfp reporter gene was under a skin-specific promoter krt8 (Gong et al.,2002). The two gene constructs were co-injected and apparently they were co-integrated as we always observed their co-segregation in all transgenic offspring for multiple generations. The presence of krt8-rfp transgene is useful in identification of cre transgenic fish by observation of RFP expression conveniently in the skin.

Previously we demonstrated that the zp3 promoter is oocyte-specific and activated predominantly in early stages of oocytes (Liu et al.,2006). In the present work, we observed the same expression pattern, further confirming the faithful function of the zp3 promoter in transgenic zebrafish. Interestingly, we observed the presence of maternal cre mRNA in early embryos of our transgenic line. The presence of maternal GFP fluorescence was also observed in our previously reported gfp transgenic line under the same zp3 promoter (Liu et al.,2006) as well as in a similar gfp transgenic line under another zebrafish zona pellucid protein gene promoter, zpc (Onichtchouk et al.,2003). Because of the presence of maternal cre RNA and likely Cre recombinase in early embryos, this apparently caused Cre-mediated recombination in early embryos when a female Tg(zp3:cre, krt8:rfp) was used to cross with a male loxP fish (Fig. 4A). Previously loxP-blocked c-myc oncogene has been activated by injection of cre mRNA to induce lymphoblastic leukemia (Laugenau et al.,2005). By using the feature of maternally expressed Cre in early embryos, our Tg(zp3:cre, krt8:rfp) transgenic line may provide a convenient tool to activate loxP-blocked genes to allow analyses of function of these genes. However, for female germline recombination experiments, it is important to start with a male Tg(zp3:cre, krt8:rfp) crossing with a female loxP fish to avoid early excision of floxed DNA in embryonic stages.

As the zp3 promoter is active in early developing oocytes (Stage I–III) before the first cleavage of meiosis (Stage IV) (Liu et al.,2006), the Cre-mediated recombination should occur in all of the oocytes even if the transgenic female are hemizygous. Thus, the excision of floxed egfp was expected in all offspring from Tg(zp3:cre, krt8:rfp) female. This was demonstrated by crossing Tg(zp3:cre, krt8:rfp) and a loxP transgenic line Tg(mylz2:loxP-EGFP-loxP)gz3 (Table 1). The efficiency of female germline excision was quite high as 8 out of 10 crossing pairs [female Tg(zp3:cre, krt8:rfp)/Tg(mylz2:loxP-EGFP-loxP)gz3 X male wild-type)] achieved 100% excision and only two of the 10 crossing pairs showed incomplete excision as 10% of their offspring retained the floxed egfp transgene. In contrast, the previously used Tg(hsp70:cre) displayed Cre-mediated recombination in only approximately 30% of genomes after heat shock induction, as assayed by quantitative real-time PCR (Le et al.,2007). However, a direct comparison of the efficiency of the two cre transgenic lines by the same approach is not possible. In our cre transgenic line, we targeted only a specific cell lineage (oocytes) for Cre-mediated recombination and direct quantification of recombined genomes in the ovary tissue is not appropriate as the ovary also contain overwhelmingly somatic cells.

Finally, by crossing the Tg Tg(zp3:cre, krt8:rfp) with a loxP line with dual reporter genes, Tg(EF:loxP-cherry-loxP-egfp), we also demonstrated that the newly made cre transgenic line is capable of activating loxP-blocked genes by both maternal Cre and oocyte-specifically expressed Cre. Thus, our present Tg(zp3:cre, krt8:rfp) has several obvious applications. First, it is useful to develop conditional gene activation and inactivation systems in oocytes or female germline for investigation of gene function and cell lineage. Second, The Tg(zp3:cre, krt8:rfp) line will help to activate some conditional transgenic lines that display early mortality or defects because of constitutive expression of certain functional genes. Unlike other tissue-specific cre transgenic lines that execute recombination in somatic tissues and thus activate conditional transgenes only in certain tissues, the Tg(zp3:cre, krt8:rfp) line should be useful to activate loxP-blocked genes in all tissues in the next generation and tissue-specificity of the targeted genes can be achieved by the promoters used. This characteristic is particularly useful to analyze function of transgene in early development in some transgenic lines that are difficult to maintain because of mortality and defects after the transgene is activated. Third, it is useful to study the function of maternal effect genes as described by De Vries et al (2000). Like all oviparous species, zebrafish largely rely on maternal genes in early development and it should be an important topic to investigate the function of maternal effect genes in zebrafish. Finally, the Tg(zp3:cre, krt8:rfp) may serve as a transgenic fish model to investigate the feasibility of self-excision of trangenes from transgenic fish to address the proprietary issue as well as ecological concerns on transgene contamination. As recently demonstrated in transgenic plant (Verweire et al.,2007), it is possible to engineer transgenic fish to retain a beneficial phenotype from the transgene while the transgene is programmed to be excised from the germline by the germline-specific Cre/loxP system; thus the released transgenic fish will not be able to produce transgenic offspring even they are accidently bred with wild fish.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. REFERENCES

Zebrafish and Maintenance

Spawning adult zebrafish were originally purchased from a local aquarium farm and maintained in our departmental zebrafish aquarium. Breeding and rearing of zebrafish were performed according to standard methods (Westerfield,2000). The loxP-GFP reporter transgenic line Tg(mylz2:loxP-EGFP-loxP)gz3 was previously generated in our lab (Pan et al.,2005).

DNA Constructs

The pKRT8-LGLR (pKRT8-loxP-EGFP-loxP-RFP) DNA construct was made by placing two loxP sites flanking the egfp sequence under the skin-specific krt8 promoter (Gong et al.,2002). The loxP-flanked egfp DNA fragment was obtained by PCR amplification using the following two long primers with synthetic loxP sequences: Loxp5 (forward), 5′-GCGGATCCATAACTTCGTATAGCATACATTATACGAAGTTATGCCACCATGGTGAGCAAG; Loxp3 (reverse), 5′-GCGGATCCATAACTTCGTATAATGTATGCTATACGAAGTTATACCACAACTAGAAGCAGTG. In both primers, a BamHI site (GGATCC, underlined) was introduced for facilitating subsequent ligation and the 34-bp LoxP sequences are in italic. The amplified sequence was placed between the 2.2-kb krt8 promoter and rfp gene (DsRed-express-1) in the pDsRed-Express-1 plasmid backbone (Clontech) (Fig. 1). To construct pCMV-CRE and pZP3-CRE, the cre DNA (∼1.2 kb) was amplified from the Cre plasmid pACN (Bunting et al.,1999) by cre-specific primers: cre-5 (forward), 5′-ATGCGGATCCATGCCCAAGAAGAAGAGGAA; cre-3 (reverse), 5′-AAGCGGCCGCCTAATCGCCAT. A BamH I (GGATCC) and a Not I (GCGGCCGC) restriction enzyme cut sites (underlined) were introduced respectively in the 5′ end of the two primers to facilitate cloning. The PCR product was cut with NotI and BamHI, and inserted into the two plasmids with the CMV or zp3 promoters.

Microinjection and Development of Transgenic Fish

To test the reporter construct, pKRT8-LGLR and pCMV-CRE were mixed with 0.25% phenol red (dissolved in 0.1 M Tris-HCL, PH 7.6) and co-injected into zebrafish embryos of 1–2 cell stage. The injected embryos were observed at 24 hpf and 48 hpf under a Zeiss Axiovert 25 fluorescence microscopy. GFP was observed under a blue filter (450–490 nm), and RFP under a yellow filter (546 nm). To develop stable cre transgenic lines, both pKRT8-LGLR and pZP3-CRE DNA constructs were linearized by EcoRI and injected into zebrafish embryos at the one- to two-cell stage. Expression of GFP and RFP was observed under a fluorescence microscope. The injected embryos with GFP/RFP expression were reared to spawning stage (founder F0). To screen transgenic fish, F1 embryos were collected from pairs of founder fishes (F0), and genomic DNA was prepared from at least 100 embryos and examined by PCR with cre-specific primers, cre-5 and cre-3. F1 embryos from positive founders were raised to adulthood. Fin-clip DNAs were prepared from F1 fish at 3 months old and screened by PCR for presence of cre DNA.

RT-PCR and Whole-Mount In Situ Hybridization

RNAs from 2–72 hpf embryos and various adult tissues were prepared with Trizol (Invitrogen). Cre mRNA was examined by RT-PCR with one-step RT-PCR Kit (QIAGEN, USA) in 20-μl volume according to manufacturer's protocol. For whole-mount in situ hybridization analysis, the ovary of Tg(zp3:cre, krt8:rfp) transgenic fish (F1) was isolated and fixed in 4% PFA (paraformaldehyde) at 4°C overnight. Whole-mount in situ hybridization with a DIG-labeled cre antisense RNA probe was carried out as described previously (Gong et al.,2002). The ovary of wild-type female with a zp3 probe was used as a control for in situ hybridization.

Detection of Cre/loxP Excision by PCR

To detect excision of floxed egfp in Tg(zp3:cre, krt8:rfp) transgenic fish, Three primers were designed and used for PCR: Primer P1 (forward), 5′-GCTCCACCCTCTCAAGAATG, against zebrafish krt8 promoter; Primer G1 (reverse) 5′-CTGCTGGTAGTGGTCGGC, designed from the gfp coding region; and Primer R1 (reverse), 5′-CTGCTCCACGATGGTGTAGT, designed from 3′ end of rfp gene (Fig. 3A). To detect Cre-mediated recombination in Tg(mylz2:loxP-EGFP-loxP)gz3 after crossing with Tg(zp3:cre, krt8:rfp), PCR was carried out with a forward primer from mylz2 promoter region and a reverse primer from rfp region as described by Pan et al. (2005).

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. REFERENCES