Mechanisms of Development 118 (2002) 57–63 www.elsevier.com/locate/modo

Spatial and temporal ‘knock down’ of gene expression by electroporation of double-stranded RNA and morpholinos into early postimplantation mouse embryos Georg Mellitzer a,b, Marc Hallonet a, Lan Chen a, Siew-Lan Ang a,* a

Institut de Ge´ne´tique et de Biologie Mole´culaire et Cellulaire, CNRS/INSERM/ Universite´ Louis Pasteur, 67404 Illkirch cedex, CU de Strasbourg, France b INSERM U381, Ontogenesis and Pathology of the Digestive Tract. 3, Avenue Molie´re, 67200 Strasbourg, France Received 16 April 2002; received in revised form 28 May 2002; accepted 28 May 2002

Abstract Here we report the use of double-stranded RNA (dsRNA) and morpholino technologies to specifically ‘knock down’ gene expression in early postimplantation mouse embryos. Sequence specific interference mediated by either dsRNA or by morpholino has been a useful tool for studying gene function in several organisms. However, specifically for the dsRNA, doubts have been raised about whether it could successfully be applied on vertebrate embryos. We demonstrate that electroporation of dsRNA directed against Otx2 or Foxa2 into postimplantation mouse embryos results in specific knock down of the expression of the respective endogenous genes in a region- and germ-layer specific manner. We also show that electroporation of morpholino directed against Foxa2 into the node of mouse embryos leads to a specific down regulation of Foxa2 expression in the floor plate. Our results demonstrate for the first time that dsRNA and morpholino technologies can be successfully applied in early postimplantation mouse embryos to specifically knock down gene expression. q 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Otx2; Foxa2; Double stranded RNA; Morpholino; Electroporation; Mouse; Postimplantation; Node; Forebrain; Midbrain

1. Introduction Specific gene inactivation in the mouse by homologous recombination in ES cells has significantly enhanced our understanding of the genetic pathways that regulate embryogenesis in vertebrates (reviewed in, Ledermann, 2000). However, if a gene product functions widely in the embryo or is repeatedly used in space and time, this technique harbors several disadvantages. Moreover, if a gene has an essential early function its inactivation by knockout would not allow any further analysis of possible later functions during development. In theory, Cre-mediated gene inactivation circumvented some of these problems (reviewed in Nagy, 2000), however in practice this method is laborious and is limited by the availability of tissue-specific promoters. Some of these problems could potentially be overcome if one could apply the methods of double-stranded RNA (dsRNA) interference of gene expression or morpholino oligonucleotides suppressed translation of targeted genes, on postimplantation mouse embryos. Moreover, by electro* Corresponding author. E-mail address: [email protected] (S.-L. Ang).

porating dsRNA or morpholinos into specific regions of postimplantation mouse embryos one could generate region- and germ layer-specific ‘knock downs’ of genes expressed at a precise stage of development. Double-stranded RNA interference (RNAi) has been first developed in Caenorhabditis elegans and subsequently has been successfully applied in other eukaryotes, including Drosophila melanogaster, Trypanosoma brucei, planarians, plants and with limited success also in zebrafish embryos. The destruction of the targeted mRNA by dsRNA occurs prior to translation (reviewed in: Fire, 1999; Sharp, 1999; Bosher and Labouesse, 2000; Plasterk and Ketting, 2000; Sijen and Kooter, 2000; Zamore et al., 2000) and targets exon sequences. Genetic approaches in C. elegans have identified the molecular mechanism involved in the RNAi mediated pathway of mRNA degradation (Ketting et al., 1999; Tabara et al., 1999; Smardon et al., 2000). Morpholinos are chemically modified oligonucleotides (Summerton and Weller, 1997) which bind to and block translation of mRNA in vitro, in tissue culture cells (reviewed in Summerton, 1999; Summerton and Weller, 1997) and in whole embryos. The latter approach has already been proven to be a successful method to knock

0925-4773/02/$ - see front matter q 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0925-477 3(02)00191-0

58

G. Mellitzer et al. / Mechanisms of Development 118 (2002) 57–63

down the function of specific genes in zebrafish, Xenopus and chick (Nasevicius and Ekker, 2000; Heasman et al., 2000; Kos et al., 2001). Both methods have also recently been successfully applied on preimplantation mouse embryos by injecting dsRNA or morpholinos into mouse oocytes These studies demonstrate that specific RNAi can phenocopy maternal cmos 2/2 and zygotic E-cadherin 2/2 phenotypes Wianny and Zernicka-Goetz, 2000; Svoboda et al., 2000). A morpholino directed against c-mos, when injected into mouse oocytes, also resulted in phenotypes similar to the mos mutation (Coonrod et al., 2001). These results prompted us to investigate whether these methods can also be used to down regulate gene expression in early postimplantation mouse embryos, by using in vivo electroporation (Akamatsu et al., 1999; Inoue et al., 2001; Osmui and Inoue, 2001; reviewed in Inoue and Krumlauf, 2001) to deliver either dsRNA or morpholino molecules into specific tissues of whole embryo, followed by whole-embryo culture (Osmui and Inoue, 2001). We show that electroporation of dsRNA or morpholinos into distinct regions of 7.5 d.p.c. mouse embryos successfully interferes with specific gene expression in a tissue- and germ-layer specific manner. dsRNA directed against Otx2 (Otx2 dsRNA) and Foxa2 (Foxa2 dsRNA) results in a specific down regulation of endogenous Otx2 and HNF3b expression, respectively. The phenotypes of dsRNA-treated embryos are consistent with known functions of these genes from the analysis of corresponding mouse mutants. In addition, we show that a morpholino directed against Foxa2 (Foxa2 morpholino) leads to down regulation of Foxa2 expression in the floor plate as Foxa2 dsRNA, albeit to a lesser extent.

2. Results and discussion 2.1. Electroporation of dsRNA specific for Foxa2 and Otx2 leads to specific down regulation of the corresponding endogenous gene To determine if dsRNA electroporation is effective to down regulate the expression of endogenous genes in postimplantation mouse embryos, we decided to knock down specifically Foxa2 and Otx2 expression. Specific antibodies to these genes are available (Filosa et al., 1997; Baas et al., 2000), and the phenotypes of mouse embryos carrying null mutations in Foxa2 and Otx2 are well characterized. Foxa2 is strongly expressed in the floor plate, notochord and definitive gut endoderm of 8.5 d.p.c. mouse embryos. Previous studies demonstrate that inactivation of Foxa2 activity leads to a loss of all midline notochord and floor plate cells (Ang and Rossant, 1994; Weinstein et al., 1994). Otx2 is a homeobox gene expressed in all three germ layers of anterior regions of the mouse embryo at 8.5 d.p.c. up to the level of the mid-hindbrain boundary (Ang et al., 1994). Inactivation of Otx2 by homologous recombination leads to

a truncation of the anterior brain rostral to rhombomere two (Acampora et al., 1995; Matsuo et al., 1995; Ang et al.,1996; Rhinn et al., 1998) in mouse embryos. In addition, further analysis showed that Otx2 is first required in the visceral endoderm for induction of the forebrain and midbrain. Subsequently, Otx2 is also required in embryonic tissues for the maintenance of the rostral brain, since these tissues although initially formed are missing by the 6–8 somite stage in chimeric embryos (Rhinn et al., 1998). To determine whether electroporation can be use to introduce DNA and RNA molecules specifically to the floor plate, we first electroporated a control LacZ expression construct into the dorsal layer of the node of 7.5 d.p.c. mouse embryos. The presence of double-labeled b-galactosidase and Foxa2 cells in the floor plate shows the specific electroporation of floor plate precursors. Moreover, the normal morphology of the cultured electroporated embryos indicates that the electroporation by itself is not harmful to early embryonic development (Fig. 1A, B). After having established the electroporation and whole-embryo culture conditions, an overall electroporation efficiency of 70– 80% was achieved. To investigate whether Foxa2 dsRNA is effective in down regulating endogenous Foxa2 expression, we co-electroporated embryos with a Foxa2 dsRNA and a LacZ expression construct. After 24–36 h culture, embryos were double-labeled for the expression of lacZ and Foxa2. Foxa2 expression was severely reduced or missing in the ventral part of the neural tube in electroporated embryos (Fig. 1C, D). In addition, Foxa2 expression in the notochord was frequently absent (data not shown), and in some cases the notochord was displaced ventrally away from its close proximity to the neural tube (Fig. 1D). Loss of the notochord is most likely a secondary consequence resulting from loss of the floor plate, since we have not observed LacZ-positive notochord cells in electroporated embryos. Different concentrations of dsRNA were tested, in the range of 10– 150 ng/ml (Table 1). The most efficient electroporation was obtained at a concentration of 100 ng/ml of dsRNA, where the majority of the embryos showed a strong down regulation of Foxa2 expression (Fig. 1C and Table 1). Higher concentrations of Foxa2 dsRNA did not enhance this phenotype. With concentrations lower than 100 ng/ml, about the same number of embryos showed a down regulation of Foxa2 expression, although the severity of the phenotype varied (data not shown). The interference with gene expression was specific, as shown by electroporation of Otx2 dsRNA, a gene that is not expressed in the node of wildtype embryos. Otx2dsRNA did not lead to a down regulation of Foxa2 expression or to the loss of the floor plate (Fig. 1E, F). To determine that the Otx2 dsRNA is effective in down regulating the endogenous expression of Otx2, we co-electroporated it together with a LacZ expression construct into one side of the presumptive forebrain and midbrain of mouse embryos at 7.5 d.p.c. In this way, the un-electropo-

G. Mellitzer et al. / Mechanisms of Development 118 (2002) 57–63

59

(Rhinn et al., 1998). To analyze for the loss of Otx2 expression at the cellular level, we co-electroporated Otx2 dsRNA together with a GFP expression construct. After 24–36 h culture, electroporated embryos were double-stained for Otx2 and GFP by immunohistochemistry on tissue sections. Otx2 dsRNA leads to a specific down regulation of endogenous Otx2 expression in electroporated cells (Fig. 2E) visualized by the absence of doublelabeled cells. Similar to the results obtained for Foxa2 dsRNA, Otx2 dsRNA showed a dose-dependant effect with optimum results obtained at a concentration of 150 ng/ml of dsRNA (Table 1). In contrast, electroporation of Foxa2 dsRNA into the same domain had no effect on the expression of Otx2 (Fig. 2H), or on the development of the forebrain and the midbrain (Fig. 2F, G). This result demonstrates that the loss of brain tissue and of Otx2 expression is not due to a non-specific effect of the electroporated dsRNA. Taken together, both dsRNA tested so far led to a very specific down regulation of the endogenous gene expression. The control electroporation, with dsRNA unrelated to the analyzed gene, further showed that dsRNA has no side effects and does not lead to a nonspecific down regulation of gene expression. 2.2. Electroporation of Foxa2 specific morpholino leads to down regulation of its endogenous expression Fig. 1. Electroporation of Foxa2-specific dsRNA into the node of mouse embryos results in loss of endogenous Foxa2 expression in the neural tube. Mouse embryos (7.5 d.p.c.) were electroporated either with a LacZ expression vector alone (A) or in combination with Foxa2 dsRNA (C) or together with Otx2 dsRNA (E). (A) A wild-type embryo showing co-expression of b-galactosidase (blue) and Foxa2 (brown) in the floor plate (arrowhead). (B) A transverse section of the embryo depicted in (A) showing similar expression of Foxa2 in the floor plate (arrowhead), as is observed in unelectroporated wild-type embryos (data not shown). (C) Embryos electroporated with Foxa2 dsRNA show a down regulation or loss of Foxa2 expression in the neural tube (arrowhead). (D) A transverse section of the embryo shown in (C) confirming the loss of Foxa2 expression in the neural tube (arrowhead). Electroporated embryos frequently show a ventral displacement of the notochord (D, small arrow). (E,F) In contrast, electroporation of Otx2 dsRNA had no effect on Foxa2 and b-galactosidase expression in the floor plate (arrowhead). Dotted lines in (A), (C), and (E) indicate the plane of transverse sections shown in (B), (D), and (F), respectively.

rated side of the embryo could serve as a control for the in vitro culture and electroporation conditions. After 24–36 h culture, electroporated embryos showed a severe reduction specifically of the forebrain and midbrain on the electroporated side (Fig. 2A, C), whereas the un-electroporated side developed normally (Fig. 2B). In addition, electroporated hindbrain region shows no apparent morphological abnormalities (Fig. 2A). These results indicate that the reduction of the forebrain and the midbrain is not due to the in vitro culture conditions or the electric pulse. The specific loss of the forebrain and midbrain is in accordance with results from chimeric studies of Otx2 function, since Otx2 is required in these tissues for their maintenance

To investigate if morpholino oligonucleotides are also effective in down regulating genes in early postimplantation mouse embryos, we electroporated a Foxa2 morpholino together with a LacZ expression construct into the node of 7.5 d.p.c. mouse embryos. After 24–36 h culture, electroTable 1 Frequencies of phenotypes of embryos electroporated with different concentrations of either Foxa2 dsRNA or Otx2 dsRNA a

a Loss of Foxa2 protein in midline tissues is the phenotype scored in Foxa2 dsRNA electroporated embryo. In the case of Otx2 dsRNA, the phenotype scored is a reduction of the forebrain and midbrain. Only embryos showing a moderate to strong expression of the co-electroporated LacZ were analyzed. n ¼ Total number of embryos analyzed.

60

G. Mellitzer et al. / Mechanisms of Development 118 (2002) 57–63

Fig. 2. Electroporation of Otx2 dsRNA leads to a down regulation of Otx2 expression and a reduction of the forebrain and midbrain (A–E). Mouse embryos (7.5 d.p.c.) were co-electroporated either with Otx2 dsRNA together with a LacZ expression vector (A,B) or in combination with a GFP expression vector (C–E), respectively. (A,B) Whole mount in situ hybridization showing a severe reduction of forebrain and midbrain tissue expressing Otx2 on the electroporated side of the embryo (A), compared to the domain of Otx2 expressing cells in the brain of the un-electroporated contralateral side of the same embryo (B). Pictures in (A) and (B) were taken at the same magnification, the brackets indicate the forebrain and midbrain region. (C) Bright field image of the embryo shown in (D), showing a reduction of the forebrain (dotted line) on the electroporated side of the embryo, whereas the un-electroporated side developed normally (arrowhead). (D) Expression of GFP in an Otx2 dsRNA/CMV-GFP electroporated embryo is apparent under UV light. (E) A transverse section of the embryo shown in (C) and (D) double stained with an Otx2 specific antibody (red) and an anti-GFP antibody (green). The absence of cells double-labeled for both Otx2 and GFP (yellow) indicates the successful down regulation of endogenous Otx2 expression by Otx2 dsRNA. (F–H) UV/bright field (F) and UV (G) images showing normal development of the forebrain and midbrain in embryos electroporated with Foxa2 dsRNA. (H) A transverse section of the embryo shown in (F) and (G) double-labeled with the Otx2 specific antibody (red) and the anti GFP antibody (green). Co-expression of Otx2 and GFP (yellow cells) demonstrates that the Foxa2 dsRNA has no effect on Otx2 expression. Dotted lines in (D) and (G) indicate the plane of the transverse sections shown in (E) and (H), respectively.

porated embryos were double-labeled for b-gal and Foxa2 expression. The discontinuous expression of Foxa2 protein in the axial midline of electroporated embryos indicates that the Foxa2 morpholino leads to a down regulation of endo-

genous Foxa2 expression in the electroporated tissue (Fig. 3A). Analysis of sections of these embryos confirmed on a cellular level, that in areas where the floor plate has been positively electroporated, as shown by the LacZ-positive cells, a down regulation of Foxa2 protein could be observed (Fig. 3B). Different concentrations were tested ranging from 10 to 200 mM. The best results where obtained with 150 mM, increasing this concentration did not lead to a stronger down regulation of Foxa2 expression (Table 2). In contrast to the Foxa2 dsRNA, where complete loss Foxa2 expression was observed in the neural tube of electroporated embryos, only in a few embryos electroporated with Foxa2 morpholino showed partial loss of Foxa2 protein in the neural tube. We also never observed a down regulation of Foxa2 protein in the notochord or a loss of notochord cells. Electroporation of control morpholino had no effect on the expression of Foxa2 protein (Fig. 3C, D). These results indicate that the Foxa2-specific morpholino leads to specific down regulation of endogenous Foxa2 expression in early postimplantation mouse embryos, albeit less efficiently than Foxa2 RNAi. There could be several reasons for the different efficiency of Foxa2 RNAi and Foxa2 morpholino on the down regulation of Foxa2 expression. An explanation for the lower efficiency could be that the sequence, which we have chosen for the Foxa2 morpholino, is not the most effective one, therefore it might only lead to partial down regulation of Foxa2 expression. Alternatively, another reason for the higher efficiency of dsRNA compared to morpholinos may be explained by their different mechanisms of down regulation of gene expression. Recent biochemical and genetic evidences from Drosophila and C. elegans demonstrate that a RNA-directed RNA polymerase chain reaction, primed by single stranded RNA, amplifies the interference caused by a small amount of dsRNA (Lipardi et al., 2001; Sijen et al., 2001). A similar mechanism might exist in vertebrates, which could lead to an amplification of RNAi by small amounts of electroporated dsRNA, thereby resulting in a strong phenotype. Since morpholinos are chemically modified oligonucleotides (Summerton and Weller, 1997), it is highly unlikely that a similar amplification step is involved in the down regulation of gene expression. These different mechanisms may also account for the fact that a rapid increase in resulting phenotypes was obtained with increasing concentrations of Foxa2 dsRNA, whereas only a gradual amplification of observed phenotypes was found with increasing concentrations of Foxa2 morpholino (Tables 1 and 2). Another form of dsRNA, small interfering RNAs (siRNAs) of 21–25 nucleotides, have also recently been successfully used to down regulate the expression of specific genes in different mammalian cells (Elbashir et al., 2001; Brummelkamp et al., 2002). It will be interesting in future experiments to compare the efficiency of siRNAs versus long dsRNA in knocking down gene expression in early postimplantation mouse embryos.

G. Mellitzer et al. / Mechanisms of Development 118 (2002) 57–63

61

electroporation of dsRNAs or morpholinos into postimplantation mouse embryos combined with whole-embryo culture should be a useful approach for studying gene functions in vivo. The recent success in introducing DNA into later stage mouse embryos via in utero electroporation will further extend the applicability of RNAi and morpholino technologies for analyzing gene function to later stages of mouse embryonic development (Fukuchi-Shimogori and Grove, 2001; Saito and Nakatsuji, 2001; Tabata and Nakajima, 2001). 4. Materials and methods 4.1. Generation of dsRNA and morpholino

Fig. 3. Electroporation of Foxa2 morpholino leads to the down regulation of Foxa2 expression. Mouse embryos (7.5 d.p.c.) were electroporated with a LacZ expression vector together with either a Foxa2 or a control morpholino. (A) An electroporated embryo, stained for b-galactosidase (blue) and Foxa2 (brown), indicates that Foxa2 morpholino leads to a down regulation of Foxa2 expression in the floor plate (arrowhead). (B) Cross section of the embryo shown in (A) confirming the down regulation of Foxa2 in the floor plate (arrowhead). (C,D) In contrast, the electroporation of a control morpholino has no effect on the expression of Foxa2 in the floor plate (arrowhead). The difference in the number of Foxa2 expressing cells may be due to the slightly different stages of the embryos. Dotted lines in (A) and (C) indicate the plane of section shown in (B) and (D), respectively.

3. Conclusions In summary, our results indicate that dsRNA and morpholino can be successfully used to knock down gene expression in postimplantation mouse embryos. It has been suggested that the use of RNAi in vertebrates might lead to a general blockage of protein synthesis. Low concentrations of dsRNA in viral infections trigger an interferon response (Marcus, 1993) resulting in a general suppression of translation, which then induces specific elimination of the infected cells by apoptosis. However, we demonstrate here that the electroporation of dsRNA into postimplantation mouse embryos leads to a specific down regulation of the expression of the targeted gene. In addition, the use of electroporation to target RNAi or morpholinos into a specific region of a germ layer at a particular stage resulted in precise spatial knock down of gene activity. This approach should also in theory be feasible to inactivate two members of a gene family or of different families simultaneously by electroporating a combination of different gene specific dsRNA or morpholinos. This would give the possibility to address the issue of redundant functions and/or genetic interactions of genes in a tissue and stage specific manner. The generation of double mutant mice is laborious and sometimes not feasible because of heterozygous lethality of the single mutants. Given these different applications,

Chimeric primers containing T7 promoter-specific sequence and Otx2 or Foxa2 sequence were used to amplify an Otx2 (nt 380–879) or a Foxa2 (nt 3–549) fragment, respectively, using PCR. The Ambion T7 mMessage Kit was used to simultaneously generate sense and antisense strands using the PCR-generated fragments as templates. The DNAse treatment and subsequent purification of the dsRNA was done as described in the Manual. The morpholino oligonucleotides were purchased from Gene-Tools, LLC (Corvallis, OR, USA). The Foxa2 sequence was as follows; 5 0 -GCTAAGGTTGTTTTTAGGCGCAAAG-3 0 . The standard control oligo from Gene-Tools, LLC has been used as a control morpholino. 4.2. Microinjection and electroporation of embryos Embryos were obtained from CD1 mice at 7.5 d.p.c. and placed in Tyrode’s solution. A DNA solution containing 0.8 mg/ml of pCCALL-lacZ (Niwa et al., 1991) or 1 mg/ml Table 2 Frequencies of embryos showing reduction of Foxa2 protein in the midline after electroporation of Foxa2 morpholino a

a

Loss of Foxa2 protein in midline tissues is the phenotype scored in Foxa2 morpholino electroporated embryo. Only embryos showing a moderate to strong expression of the co-electroporated LacZ were analyzed. n ¼ Total number of embryos analyzed.

62

G. Mellitzer et al. / Mechanisms of Development 118 (2002) 57–63

of the pMES-GFP (Swartz et al., 2001) expression plasmid and either Foxa2 dsRNA, Otx2 dsRNA, Foxa2 morpholino, or control morpholino (for the exact concentration see Tables 1 and 2) was prepared in 2 £ PBS and mixed 1:1 with 0.01% fast green dye (Sigma) and microinjected into the amniotic cavity of a dissected embryo. Immediately after the injection of the DNA solution, the embryo was placed into the electroporation-chamber (distance between the electrodes ¼ 5 mm, TR Tech Company) in Tyrode’s solution and electroporated by using a BTX Electro Square Porator ECM 830 (BTX, San Diego, CA, USA). For the electroporation of the floor plate, embryos were positioned in the electroporation chamber with their node (distal end of the embryo) pointing to the positive pole, and then electroporated with three pulses of 30 ms with 26 V. To electroporate the anterior end of the embryos the presumptive forebrain and midbrain area was positioned parallel to the positive pole of the electroporation chamber. For unilateral electroporation, embryos were tilted slightly to one side. Embryos were then electroporated with three pulses of 30 ms with 29 V. After electroporation, embryos were cultivated for 24–36 h in a roller culture system in Tyrode’s solution supplemented with 50% rat serum and equilibrated with 5% CO2, 5% O2 (described in detail in Osumi-Yamashita et al., 1997) and subsequently fixed in 4% paraformaldehyde for 45 min on ice.

4.3. LacZ staining, in situ hybridization, whole-mount antibody staining and histology After fixation, embryos were washed three times in PBT (PBS 1 0.1% Tween 20), and stained for b-galactosidase as described in Tajbakhsh and Houzelstein (1995). The whole mount in situ hybridization was performed as described previously (Conlon and Herrmann, 1993), using an Otx2 probe (Ang et al., 1994). Whole-mount immunocytochemistry was performed according to published procedure (Davis et al., 1991) using an antiFoxa2 antibody (Filosa et al., 1997) at a dilution of 1:500. After immunocytochemistry embryos were postfixed overnight in 2.5% glutaralaldehyde in PBS, rinsed in PBS, embedded using the JB-4 embedding Kit (Polysciences, Inc) and 5 mm sections taken. For the Otx2 and GFP detection, immunocytochemistry was performed on 10 mm Cryostat section, as described in Muhr et al. (1997). A rabbit polyclonal antibody for Otx2 (Baas et al., 2000) and a polyclonal chicken antibody for GFP (Chemicon, Inc, AB16901) were used at 1:1000 and 1:250 dilutions, respectively. For the detection of these specific antibodies, different fluorescent conjugated secondary antibodies against rabbit IgG (Cy3-conjugated AffiniPure Donkey Anti-Rabbit IgG, Jackson ImmunoReserach, Inc.) and chicken IgG (Alexa Fluor 488 goat antichicken IgG, Molecular Probes) were used at a 1:500 dilution.

Acknowledgements We thank members of the laboratory for critical reading of the manuscript. We are grateful to Michael Labouesse for technical advice on the generation of dsRNA, and Giles Duval for preparing the rabbit serum. The pMES-GFP expression plasmid was a kind gift from Cathy Krull. The Otx2 and Foxa2 antibodies were generously provided by Drs Flora Vaccario and Hiroshi Sasaki, respectively. G.M. was supported by a postdoctoral fellowship from INSERM and from HFSP. This work was supported by grants from the European Community Biotech Programme and the Association pour la Recherche sur le Cancer, to S.L.A., and by funds from the Institute National de la Sante´ et de la Recherche Me´ dicale, the Centre National de la Recherche Scientifique, and the Hoˆ pital Universitaire de Strasbourg.

References Acampora, D., Mazan, S., Lallemand, Y., Avantaggiato, V., Maury, M., Simeone, A., Brulet, P., 1995. Forebrain and midbrain regions are deleted in Otx2-/- mutants due to a defective anterior neuroectoderm specification during gastrulation. Development 121 (10), 3279–3290. Akamatsu, W., Okano, H.J., Osumi, N., Inoue, T., Nakamura, S., Sakakibara, S., Miura, M., Matsuo, N., Darnell, R.B., Okano, H., 1999. Mammalian ELAV-like neuronal RNA-binding proteins HuB and HuC promote neuronal development in both the central and the peripheral nervous systems. Proc. Natl Acad. Sci. USA 96 (17), 9885–9890. Ang, S.L., Rossant, J., 1994. HNF-3 beta is essential for node and notochord formation in mouse development. Cell 78 (4), 561–574. Ang, S.L., Conlon, R.A., Jin, O., Rossant, J., 1994. Positive and negative signals from mesoderm regulates the expression of mouse Otx2 in ectoderm explants. Development 120, 2979–2989. Ang, S.L., Jin, O., Rhinn, M., Daigle, N., Stevenson, J., Rossant, J., 1996. A targeted mouse Otx2 mutation leads to severe defects in gastrulation and formation of axial mesoderm and to deletion of rostral brain. Development 122 (1), 243–252. Baas, D., Bumsted, K.M., Martinez, J.A., Vaccarino, F.M., Wikler, K.C., Barnstable, C.J., 2000. The subcellular localization of Otx2 is cell-type specific and developmentally regulated in the mouse retina. Brain Res. Mol. Brain Res. 78 (1-2), 26–37. Bosher, J.M., Labouesse, M., 2000. RNA interference: genetic wand and genetic watchdog. Nat. Cell Biol. 2 (2), E31–E36. Brummelkamp, T.R., Bernards, R., Agami, R., 2002. A system for stable expression of short interfering RNAs in mammalian cells. Science 296, 550–553. Coonrod, S.A., Bolling, L.C., Wright, P.W., Visconti, P.E., Herr, J.C., 2001. A morpholino phenocopy of the mouse mos mutation. Genesis 30 (3), 198–200. Conlon, R.A., Herrmann, B.G., 1993. Detection of messenger RNA by in situ hybridization to postimplantation embryo whole mounts. Methods Enzymol. 225, 373–383. Davis, C.A., Holmyard, D.P., Millen, K.J., Joyner, A.L., 1991. Examining pattern formation in mouse, chicken and frog embryos with an Enspecific antiserum. Development 111, 287–298. Elbashir, S.M., Harborth, J., Lendeckel, W., Yalcin, A., Weber, A., Tuschl, T., 2001. Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 411, 494–498. Filosa, S., Rivera-Perez, J.A., Gomez, A.P., Gansmuller, A., Sasaki, H., Behringer, R.R., Ang, S.L., 1997. Goosecoid and HNF-3beta genetically interact to regulate neural tube patterning during mouse embryogenesis. Development. 124 (14), 2843–2854.

G. Mellitzer et al. / Mechanisms of Development 118 (2002) 57–63 Fire, A., 1999. RNA-triggered gene silencing. Trends Genet. 15, 358–363. Fukuchi-Shimogori, T., Grove, E.A., 2001. Neocortex patterning by the secreted signaling molecule FGF8. Science 294 (5544), 1071–1074. Heasman, J., Kofron, M., Wylie, C., 2000. Beta-catenin signaling activity dissected in the early Xenopus embryo: a novel antisense approach. Dev. Biol. 222 (1), 124–134. Inoue, T., Krumlauf, R., 2001. An impulse to the brain using in vivo electroporation. Nat. Neurosci. 4 (Suppl), 1156–1158. Inoue, T., Tanaka, T., Takeichi, M., Chiska, O., Nakamura, S., Osumi, N., 2001. Role of cadherins in maintaining the compartment boundary between the cortex and striatum during development. Development 128 (4), 561–569. Ketting, R.F., Haverkamp, T.H., van Luenen, H.G., Plasterk, R.H., 1999. Mut-7 of C. elegans, required for transposon silencing and RNA interference, is a homolog of Werner syndrome helicase and RnaseD. Cell 99, 133–141. Kos, R., Reedy, M.V., Johnson, R.L., Erickson, C.A., 2001. The wingedhelix transcription factor FoxD3 is important for establishing the neural crest lineage and repressing melanogenesis in avian embryos. Development 128, 1467–1479. Ledermann, B., 2000. Embryonic stem cells and gene targeting. Exp. Physiol. 85 (6), 603–613. Lipardi, C., Wei, O., Paterson, B.M., 2001. RNAi as random degradative PCR: siRNA primers convert mRNA into dsRNAs that are degraded to generate new siRNAs. Cell 107 (3), 297–307. Marcus, P.I., 1993. Interferon induction by viruses: one molecule of dsRNA as the threshold for interferon induction. Interferon 5, 115–180. Matsuo, I., Kuratani, S., Kimura, C., Takeda, N., Aizawa, S., 1995. Mouse Otx2 functions in the formation and patterning of rostral head. Genes Dev. 9 (21), 2646–2658. Muhr, J., Jessell, T.M., Edlund, T., 1997. Assignment of early caudal identity to neural plate cells by a signal from caudal paraxial mesoderm. Neuron 19, 487–502. Nagy, A., 2000. Cre recombinase: the universal reagent for genome tailoring. Genesis 26, 99–109. Nasevicius, A., Ekker, S.C., 2000. Effective targeted gene ‘knockdown’ in zebrafish. Nat. Genet. 26 (2), 216–220. Niwa, H., Yamamura, K., Miyazaki, J., 1991. Efficient selection for highexpression transfectants with a novel eukaryotic vector. Gene 108 (2), 193–199. Osmui, N., Inoue, T., 2001. Gene transfer into cultured mammalian embryos by electroporation. Methods 24 (1), 35–42. Osumi-Yamashita, O., Ninopiya, Y., Eto, A., 1997. Mammalian craniofacial embryology in vitro. Int. J. Dev. Biol. 41, 187–194. Plasterk, R.H., Ketting, R.F., 2000. The silence of the genes. Curr. Opin. Genet. Dev. 10 (5), 562–567. Rhinn, M., Dierich, A., Shawlot, W., Behringer, R.R., Le Meur, M., Ang,

63

S.L., 1998. Sequential roles for Otx2 in visceral endoderm and neuroectoderm for forebrain and midbrain induction and specification. Development 125 (5), 845–856. Saito, T., Nakatsuji, N., 2001. Efficient gene transfer into the embryonic mouse brain using in vivo electroporation. Dev. Biol. 240 (1), 237–246. Sharp, P.A., 1999. RNAi and double-strand RNA. Genes Dev. 13, 139–141. Sijen, T., Kooter, J.M., 2000. Post-transcriptional gene-silencing: RNAs on the attack or on the defense? Bioessays 22 (6), 520–531. Sijen, T., Fleenor, J., Simmer, F., Thijssen, K.L., Parrish, S., Timmons, L., Plasterk, R.H., Fire, A., 2001. On the role of RNA amplification in dsRNA-triggered gene silencing. Cell 107 (4), 465–476. Smardon, A., Spoerke, J., Stacey, S., Klein, M., Mackin, N., Maine, E., 2000. EGO-1 is related to RNA-directed RNA polymerase and functions in germ-line development and RNA interference in C. elegans. Curr. Biol. 10, 169–178. Summerton, J., Weller, D., 1997. Morpholino antisense oligomers: design, preparation, and properties. Antisense Nucleic Acid Drug Dev. 7 (3), 187–195. Summerton, J., 1999. Morpholino antisense oligomers: the case for an RNase H-independent structural type. Biochim. Biophys. Acta 1489 (1), 141–158. Swartz, M., Eberhart, J., Mastick, G.S., Krull, C.E., 2001. Sparking new frontiers: using in vivo electroporation for genetic manipulations. Dev. Biol. 233 (1), 13–21. Svoboda, P., Stein, P., Hayashi, H., Schultz, R.M., 2000. Selective reduction of dormant maternal mRNAs in mouse oocytes by RNA interference. Development 127 (19), 4147–4156. Tabara, H., Sarkissian, M., Kelly, W.G., Fleenor, J., Grishok, A., Timmons, L., Fire, A., Mello, C.C., 1999. The rde-1 gene, RNA interference, and transposon silencing in C. elegans. Cell 99, 123–132. Tabata, H., Nakajima, K., 2001. Efficient in utero gene transfer system to the developing mouse brain using electroporation: visualization of neuronal migration in the developing cortex. Neuroscience 103, 865– 872. Tajbakhsh, S., Houzelstein, D., 1995. In situ hybridization and beta-galactosidase: a powerful combination for analysing transgenic mice. Trends Genet. 11, 42. Weinstein, D.C., Ruiz, I., Altaba, A., Chen, W.S., Hoodless, P., Prezioso, V.R., Jessell, T.M., Darnell Jr, J.E., 1994. The winged-helix transcription factor HNF-3 beta is required for notochord development in the mouse embryo. Cell 78 (4), 575–588. Wianny, F., Zernicka-Goetz, M., 2000. Specific interference with gene function by double-stranded RNA in early mouse development. Nat. Cell Biol. 2 (2), 70–75. Zamore, P.D., Tuschl, T., Sharp, P.A., Bartel, D.P., 2000. RNAi: doublestranded RNA directs the ATP-dependent cleavage of mRNA at 21 to 23 nucleotide intervals. Cell 101, 25–33.