Visualization of the Xenopus primordial germ cells using a green fluorescent protein controlled by cis elements of the 3′ untranslated region of the DEADSouth gene

Mechanisms of Development 123 (2006) 746–760 www.elsevier.com/locate/modo

Visualization of the Xenopus primordial germ cells using a green fluorescent protein controlled by cis elements of the 3 0 untranslated region of the DEADSouth gene Kensuke Kataoka, Takeshi Yamaguchi, Hidefumi Orii *, Akira Tazaki 1, Kenji Watanabe, Makoto Mochii Department of Life Science, University of Hyogo, 3-2-1 Koto, Kamigori, Akou-gun, Hyogo 678-1297, Japan Received 23 June 2006; received in revised form 14 July 2006; accepted 14 July 2006 Available online 21 July 2006

Abstract We succeeded in visualization of the primordial germ cells (PGCs) in a living Xenopus embryo. The mRNA of the reporter Venus protein, fused to the 3 0 untranslated region (UTR) of DEADSouth, which is a component of the germ plasm in Xenopus eggs, was microinjected into the vegetal pole of fertilized eggs and then the cells with Venus fluorescence were monitored during development. The behavior of the cells was identical to that previously described for PGCs. Almost all Venus-expressing cells were Xdazl-positive in the stage 48 tadpoles, indicating that they were PGCs. In addition, we found three sub-regions (A, B and C) in the 3 0 UTR, which were involved in the PGC-specific expression of the reporter protein. Sub-region A, which was identified previously as a localization signal for the germ plasm during oogenesis, participated in anchoring of the mRNA at the germ plasm and the degradation of the mRNA in the somatic cells. Sub-regions B and C were also involved in anchoring of the mRNA at the germ plasm. Sub-region B participated in the enhancement of translation.  2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Translation; mRNA stability; Imaging; Germ plasm; Vasa; DEAD-box

1. Introduction The choice of determination of cell lineage differentiation to either a germ cell or a somatic cell is one of the initial events occurring during early development of sexually reproductive animals. The primordial germ cells (PGCs) are distinguished from other somatic cells in early development, migrate to the developing gonad, and then differentiate to germ cells, sperm in a male or eggs in a female. There are two systems for generation of PGCs in early development. In mammals such as mouse, it is known that the PGCs are differentiated from embryonic cells depending *

Corresponding author. Tel./fax: +81 791 58 0187. E-mail address: [email protected] (H. Orii). 1 Present address: Max Plank Institute of Molecular Cell Biology and Genetics, 01307 Dresden, Germany.

on the surrounding conditions. In the other system, a distinctive cytoplasmic structure, called ‘germ plasm’, pre-exists in the oocytes, and the blastomeres carrying it differentiate into PGCs during development. The latter occurs in Xenopus, Drosophila and Caenorhabditis elegans. The germ plasm has been shown to be a ‘determinant’ for germ cell differentiation in Drosophila (Okada et al., 1974; Illmensee and Mahowald, 1974) and Xenopus (Ikenishi et al., 1986). However, these systems for germ cell determination are not evolutionarily conserved. Even among amphibians, there is germ plasm in anurans, but not in urodeles (reviewed in Wakahara, 1996). So far, many molecules localized in germ plasm have been identified in Xenopus and studied in order to understand the molecular mechanisms of germ cell development (reviewed in Zhou and King, 2004). The detailed localization of these molecules in the germ plasm has also been

0925-4773/$ - see front matter  2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.mod.2006.07.006

K. Kataoka et al. / Mechanisms of Development 123 (2006) 746–760

analyzed (Kloc et al., 2002). Xdazl which is one such germ plasm-specific gene, was shown to be required for migration of PGCs (Houston and King, 2000). Recently, Dead end was also shown to be required for proper migration and subsequent differentiation of PGCs (Horvay et al., 2006). However, generally, it is very difficult to determine the functions of germ plasm-specific genes in Xenopus, because a large amount of the transcript is stored in the unfertilized egg, so that it seems to be difficult to deplete the molecule in fertilized eggs (Machado et al., 2005). It is also not so easy to identify PGCs in the developing embryos. It is necessary to be able to label PGCs easily for functional analysis of germ plasm-specific genes. Recently, in Xenopus, Nishiumi et al. (2005) showed that the descendant cells, including PGCs, were labeled by microinjection of fluorescein-dextran lysine into a single vegetal blastomere of the 32-cell stage embryo and that PGCs migrated actively apart from other somatic descendant endodermal cells later than stage 24. In small fishes, Medaka killifish and zebrafish, there have been several reports showing the successful tracing of PGCs in a living organism using the green fluorescent protein (GFP) gene fused to the 3 0 untranslated region (UTR) of a germ cell-specific gene such as vasa or nanos (Tanaka et al., 2001; Ko¨prunner et al., 2001). Furthermore, the regions required for PGC-specific expression of the reporter protein have been identified by deletion analysis (Knaut et al., 2002). The PGC-specific expression of vasa seemed to be ensured by regulation at multiple steps such as degradation and translation of the mRNA (Wolke et al., 2002). The Xenopus DEADSouth gene was initially identified as Xcat-3, whose transcript is localized in the germ plasm of Xenopus eggs (Mosquera et al., 1993; MacArthur et al., 2000). It encodes a putative RNA helicase which belongs to the DEAD-box protein family. We focused on the 3 0 UTR of this gene as a regulator of PGC-specific expression based on analogy with fishes. In this study, we were successful in labeling the PGCs in a living Xenopus embryo by microinjection of venus mRNA fused to the 3 0 UTR of the DEADSouth gene. Furthermore, we identified three sub-regions in the 3 0 UTR participating in PGC-specific expression. Although one of the three has been already identified as a region participating in the localization of the transcript to the germ plasm during oogenesis, the others were newly identified. We also discuss the regulatory mechanisms of the transcript by the 3 0 UTR of the DEADSouth gene during development. 2. Results 2.1. Venus fused to 3 0 UTR of DEADSouth was useful for tracing the PGCs in a living embryo Although the 3 0 UTRs of various germ plasm-specific mRNAs were shown to be required for their localization to the germ plasm during oogenesis, it has remained

747

unknown whether they were also required for the specific expression in primordial germ cells (PGCs) in later development. We focused on such a gene, DEADSouth, whose 3 0 UTR contained the signal for localization of the mRNA to the germ plasm during oogenesis (Betley et al., 2002; Knaut et al., 2002). We amplified almost the full 3 0 UTR including the polyA tail (about 1.6 kbp) of the DEADSouth gene by PCR with EST cDNA clone (NIBB, XDB3, XL035o16) and inserted it just downstream of the reporter gene venus (a derivative of GFP). The mRNA of venusDEADSouth 3 0 UTR (termed DS3 0 UTR) was synthesized in vitro and microinjected at the vegetal pole of a fertilized egg. After several trials, it was determined that the appropriate amount of the injected mRNA to visualize PGCs was about 1.1 fmol/egg (Fig. 1A). When at least this amount of mRNA was injected, almost all PGCs could be visualized based on Venus fluorescence (see below). Almost all of the injected embryos developed normally, as did uninjected or water-injected embryos. The initial fluorescence of Venus protein was faintly detected in the vegetal blastomeres of early blastula embryos (about 128 cells at stage 7, Fig. 1B). The signal did not seem to be restricted to the primordial germ cells (PGCs). The distribution of the signal depended upon the embryo. At stage 10 or 11 (gastrula stage), faint but slightly increased fluorescence was observable in the vegetal region, including around the yolk plug (Fig. 1D). By comparison with the original expression pattern of the DEADSouth gene, the fluorescence of Venus was clearly observed in a region broader than the DEADSouth-expressing region at this stage (Fig. 1C and D). In some embryos, a few cells with strong fluorescence (probably PGCs) were visible externally among cells with faint fluorescence. We could detect several cells with strong fluorescence in the endodermal cell mass with faint fluorescence even in sections of the embryos in which cells with strong fluorescence were not observable externally (Fig. 2A). At stage 17, the external fluorescence became weak, which may have been due to continuous divisions and movement of cells. In the section of the embryo, we could find Venus-positive cells more easily in the endodermal cell mass, because of the weaker fluorescence in the surrounding cells (Fig. 2B). Little fluorescence was visible externally in the ventral region at stage 20 (neurula), because the PGCs were present in the depths of the opaque embryo. We found cells strongly positive for Venus, probably PGCs, when the endodermal cell mass of the embryo was dissociated (Fig. 2C). In the tailbud-stage embryo (stage 30), we could just detect externally the fluorescence in a few cells in the endodermal cell mass (Fig. 1F), because the embryo became relatively thin and transparent after elongation. The distribution pattern of these cells was very similar to that of the DEADSouth-expressing cells, indicating that they were PGCs (Fig. 1E and F). Then, Venus-expressing cells appeared to begin to aggregate in the endodermal cell mass at stage 34 (Fig. 1G) and migrated to the dorsal part of the body wall until stage 41 (Fig. 1H and I). This process of migration of

748

K. Kataoka et al. / Mechanisms of Development 123 (2006) 746–760

Fig. 1. Visualization of PGCs in a whole embryo/tadpole microinjected with venus mRNA fused to 1.6 kb of DEADSouth 3 0 UTR (DS3 0 UTR). (A) Schematic diagram of the microinjection of the mRNA into the vegetal pole of the fertilized egg (1 cell-stage). (B, D, and F–J) Expression of Venus protein (green) during development. (C and E) Whole-mount in situ hybridization for DEADSouth gene. (B–D) Vegetal view. (B and D) are the same embryo. Arrowheads in (C and D) indicate the pigmented dorsal lip. (E–I) Lateral view. (J) Dorso-lateral view. Each developmental stage is indicated (B–J). Arrows in (F–J) indicate the Venus-expressing cells. Scale bars are 250 lm in (B–D), 1 mm in (E–J).

Venus-expressing cells was also examined in detail in transverse sections of the embryos (Fig. 2E–G). At stage 34, these cells surrounded by cells with faint fluorescence were observable in the endodermal cell mass (Fig. 2E). They seemed to migrate laterally and dorsally in the endodermal cell mass, and then migrate to the most dorsal part of the mass (Fig. 2F). Finally, Venus-expressing cells accumulated on the dorsal body wall, which would form the dorsal mesentery later (Fig. 2G). At stage 43, they seemed to stay on the mesentery and settled in a region where the genital ridge formed (arrow in Fig. 1J). At this stage, we could find

the Venus-expressing cells in the living whole tadpole easily after removing the ventral part (Fig. 2D). At stage 48, paraffin-embedded transverse sections were prepared (inset of Fig. 3A) and stained immunohistochemically for Venus protein and Xdazl protein, which is a specific marker of PGCs (Mita and Yamashita, 2000; Houston and King, 2000). The Venus-positive cells were large in size and present in the genital ridges on both sides of the dorsal mesentery (Fig. 3A and C). They were also positive for antiXdazl antibody, indicating that they were obviously PGCs (Fig. 3B and D). To evaluate how many PGCs were

K. Kataoka et al. / Mechanisms of Development 123 (2006) 746–760

749

Fig. 2. Behavior of Venus-expressing cells inside the embryos. (A) Sagittal manual section of gastrula at stage 11. bl, blastocoel; dl, dorsal lip; vl, ventral lip. (B) Transverse manual section of the neurula at stage 17. ar, archenteron. (C) Endodermal cell mass of the embryo at stage 20 dissociated in 1.25-fold diluted PBS() (Kataoka et al., 2005). Arrowheads in (A–C) indicate the Venus-expressing cells. (D) High magnification images of the Venus-expressing cells at stage 43 (tadpole). (E–G) Manual transverse section of the tadpoles at stages 34, 38 and 41. Fluorescent images were superimposed on bright field images. Scale bars are 250 lm in (A and B), 100 lm in (C), 50 lm in (D) and 250 lm in (E–G).

Venus-positive, all serial transverse sections of the whole tapole at this stage were subjected to immunostaining with anti-GFP and anti-Xdazl. For 6 tadpoles at stage 48, we identified 37 to 59 PGCs per tadpole as Xdazl-positive cells. Among them, almost all cells were also Venus-positive (97.5 ± 1.6%, average ± standard error). Finally, PGCs settled in the gonad at stage 52 (Fig. 3E–G). Furthermore, we examined whether the Venus-fluorescence was maintained in the PGCs of later stage tadpoles. At stage 56, the gonads with mesonephros were taken out from the tadpole and immunostained by whole-mount using anti-Xdazl and anti-GFP antibodies (Fig. 4). Although we confirmed Venus fluorescence in the gonads in 5 out of 6 living tadpoles, we could find immunohistochemically GFP-positive gonads only in 2 of 5 tadpoles after fixation. In addition, at this stage, both PGCs with Venus (Xdazl-positive, GFP-positive; arrowheads in Fig. 4H) and without Venus (Xdazl-positive, GFP-negative) were also observed. These findings suggest that the Venus protein was degraded and/or diluted in the proliferating PGCs at around stage 56. Thus, the labeling system using venus-DEADSouth is very useful for tracing living PGCs until stage 56. 2.2. Identification of the region required for PGC-specific expression In order to identify the region required for the PGC-specific expression of the reporter Venus protein, we constructed the venus gene fused to the systematically deleted 3 0 UTR of DEADSouth (Fig. 5), synthesized these mRNAs in vitro and microinjected them at the vegetal pole of the fertilized eggs. To evaluate the ability of their mRNAs to direct PGC-specific expression, we observed the distribution pattern and intensity of Venus fluorescence in stage

40 or 41 tadpoles. More than 5 experiments were performed independently and 30 or more eggs were used for each experiment. The venus transcript fused to the first 1478 nucleotides (3 0 D1), 1298 nucleotides (3 0 D2) and 1135 nucleotides (3 0 D3) of 3 0 UTR of DEADSouth showed the same expression (in both distribution pattern and intensity) as the original transcript (DS3 0 UTR) (Figs. 1I and 6A and A 0 ). Venus fluorescence was observable in the PGCs in 54–73% of the injected tadpoles. These values were not significantly different from that for DS3 0 UTR (65%) by probability test. On the other hand, out of the tadpoles injected with the venus transcript fused to the first 1033 nucleotides (3 0 D4), 961 nucleotides (3 0 D5) and 806 nucleotides (3 0 D6) of 3 0 UTR of DEADSouth, 25%, 18% and 35% showed weak expression of Venus fluorescence in PGCs, respectively. These values were significantly lower than that of DS3 0 UTR (Fig. 5). In addition, little Venus fluorescence was observable in cells other than PGCs (Fig. 6B and B 0 ). Thus, it was suggested that at least the region downstream from nucleotide 1136 of the 3 0 UTR of DEADSouth was not required for the PGC-specific expression of the reporter Venus protein. Although the transcript of venus fused to 3 0 UTR of DEADSouth lacking the first 294 nucleotides (5 0 D1) showed the same expression as that of DS3 0 UTR (Fig. 6C and C 0 ), the transcripts lacking the first 551 nucleotides (5 0 D2) showed that PGC specificity was significantly reduced (28%) and weak fluorescence was observable not only in PGCs but also in the other region (Figs. 5 and 6D and D 0 ). Although mRNAs with further deletions, 5 0 D3, 5 0 D4 and 5 0 D5, also showed weak expression of Venus in the PGCs in addition to low specificity for expression in PGCs (11–30%), they showed strong fluorescence in regions other than PGCs (Figs. 5 and 6E and E 0 ). 5 0 D6, lacking nucleotides 1 to 953, showed no expression in

750

K. Kataoka et al. / Mechanisms of Development 123 (2006) 746–760

Fig. 3. Location of Venus-expressing cells in the tadpoles at stages 48 (A–D) and 52 (E–G). Paraffin-embedded tadpoles injected with DS3 0 UTR were sectioned transversely at the indicated level (inset of A) and immunostained with anti-Xdazl (red) and anti-GFP (green). All images are merged with Nomarski images. (A) A low-magnification image merged with anti-Xdazl and anti-GFP. (B-D) High-magnification image of the genital ridges in the box in (A). (B and E) Anti-Xdazl staining (red). (C and F) Anti-GFP staining (green). (D and G) are merged images of (B and C), and (E and F), respectively. High-magnification image of the gonad at stage 52 is shown in the inset of (E–G), respectively. g, gonad; i, intestine; l, lung; m, muscle; nc, notochord; sc, spinal cord. Scale bars are 250 lm in (A), 20 lm in (B–D), 50 lm in (E–G) and 20 lm in insets of (E–G), respectively.

K. Kataoka et al. / Mechanisms of Development 123 (2006) 746–760

751

Fig. 4. Venus-expressing cells in the gonads at the tadpole stage (ventral view). The mesonephros of the tadpole at stage 56 was immunostained in wholemount simultaneously with anti-Xdazl (red in B and F) and anti-GFP (green in C and G). (A–D) Low-magnification images. (E–H) High-magnification images of the box in (A). (A and E) are bright-field images. (D) is a merged image of (B and C). (H) is a merged image of (E, F, and G). Auto-fluorescence was also observed. The arrowheads in (H) indicate the double positives for anti-Xdazl and anti-GFP (yellow). Note that several Xdazl-positive cells are Venus-negative. Scale bars, 500 lm in (A) and 200 lm in (E), respectively.

PGCs and only a little expression in the other region (Figs. 5 and 6F and F 0 ). These findings indicate that the first 294 nucleotides do not affect the expression pattern of DS3 0 UTR. Interestingly, Knaut et al. (2002) and Betley et al. (2002) reported independently that nucleotides 69 to 584 and nucleotides 300 to 631 of the 3 0 UTR of DEADSouth mRNA were required for its localization in the germ plasm during oogenesis, respectively. These results, taken together suggested that the localization signal for the germ plasm was located in nucleotides 300 to 584. This germ

plasm localization signal closely corresponds to our region defined by 5 0 D1 and 5 0 D2 (Fig. 5). This indicates that in addition to localization signal for the germ plasm, nucleotides 585 to 1135 are required for PGC-specific expression. Furthermore, we made constructs with various gaps inside the region identified above and examined them for PGCspecific expression (Fig. 5). Examination of D585–791, D585–856 and D585–953 showed that the fluorescence was drastically reduced in the PGCs and was not observable in other regions (Figs. 5 and 6G and G 0 ). Use

752

K. Kataoka et al. / Mechanisms of Development 123 (2006) 746–760

Fig. 5. Identification of sub-regions participating in PGC-specific expression in the 3 0 UTR of the DEADSouth gene. venus reporter gene fused to various regions of 3 0 UTR of DEADSouth gene (left) were microinjected into the vegetal pole of fertilized eggs. The embryos were developed until stage 40 or 41 and were examined for the intensity and location of the fluorescent signal of Venus protein. The numbers in the constructs indicate the nucleotide positions of the 3 0 UTR of the DEADSouth gene (DDBJ Accession No. AF190623, the first nucleotide of 3 0 UTR just after the stop codon of DEADSouth gene is position 1). pA indicates a transcribed poly(A) tail. n shows the number of independent experiments for each construct. At least 30 eggs were used for each experiment and more than 70% of them were normally developed. The intensity of fluorescent signal in PGCs of a typical example is graded by  (no signal), +, ++ and +++ (strong). The average ratio of tadpoles with fluorescent PGCs to total injected embryos is shown with the standard error. Probability was calculated vs. DS3 0 UTR by Student’s t-test. Asterisks indicate statistically significant differences (P < 0.05). Average ratio of tadpoles with fluorescence in regions other than PGCs to total injected tadpoles, also shown with the standard errors on the right. The intensity of fluorescent signal was graded in the same way. The regions involved in localization of DEADSouth mRNA during oogenesis were reported previously (Betley et al., 2002; Knaut et al., 2002) and aligned in the lowest panel with sub-regions identified in this study (see Fig. 7B).

of D585–1033 resulted in no expression (Figs. 5 and 6H and H 0 ). Based on these findings taken together, we tentatively divided nucleotides 295–1135 into three sub-regions: subregion A (295–662) closely corresponded to the localization signal for the germ plasm, as identified previously, sub-region B (663–953) was involved in strong expression in PGCs and other sites, and sub-region C (954–1135) was weakly involved in PGC-specific expression. 2.3. Characterization of sub-regions using venus-b-globin transcript We then examined if nucleotides 295–1135 (sub-region ABC) were sufficient for the PGC-specific expression of the Venus reporter protein. At first, we made a construct in which venus was fused to the 3 0 UTR of Xenopus b-globin including the poly(A) tail (bG), in the middle of which

individual or combined sub-regions were inserted (Fig. 7). The mRNA was synthesized in vitro using the construct linearized with XhoI as template (Fig. 7) and microinjected at the vegetal pole of the fertilized eggs as described above. As a control, the transcript of venus fused just to the 3 0 UTR of Xenopus b-globin (bG) was also examined. Venus protein was expressed throughout the tadpoles injected with bG, not in a PGC-specific manner (Figs. 7C and 8H and H 0 ). In contrast, Venus protein was expressed specifically in PGCs in the tadpoles injected with bG/ABC mRNA (Fig. 8A and A 0 ), which was similar to the finding for DS3 0 UTR (Fig. 1I). The specificity of bG/ABC for PGCs (62%) could not be distinguished significantly from that of DS3 0 UTR (65%, Fig. 5; P = 0.67 > 0.05), indicating that the sub-region ABC was sufficient for PGC-specific expression of the reporter Venus protein. Each sub-region, A, B, or C, gave significantly lower PGC-specificity than

K. Kataoka et al. / Mechanisms of Development 123 (2006) 746–760

753

Fig. 6. Typical expression pattern of venus fused to variously deleted 3 0 UTR of DEADSouth in stage 41 tadpoles. Whole-mount lateral views (left) are shown with high-magnification views of PGCs (right), respectively. Solid and dashed brackets indicate PGCs with and without fluorescence, respectively. The injected transcript is indicated on the upper right of each image. Refer to Fig. 5 for the transcripts. Scale bars, 1 cm and 250 lm in (H) and (H 0 ), respectively.

the complete sub-region ABC (Fig. 7C). The intensity of Venus fluorescence in these PGCs was also weaker than that of bG/ABC (Figs. 7C and 8B 0 , C 0 and D 0 ). Interestingly, in most tadpoles injected with bG/B or bG/C, Venus fluorescence was observable in regions other than the PGCs, but there was no expression there in the tadpoles injected with bG/A (Figs. 7C and 8B, C, and D). In regions other than PGCs, bG/B showed more Venus expression than bG/C or bG (Figs. 8C, D, and H). Next, we tested the transcript bG fused to any two subregions for their PGC-specific expression (Fig. 7). Although expression of the reporter protein was observable in the PGCs of tadpoles injected with bG/AB, the specificity seemed to be a little lower (not significantly) than that of bG/ABC (Figs. 7C and fig. 8E and E 0 ). The fluorescence in the PGCs also seemed to be weaker and it was observable in regions other than PGCs in 39% of the injected tadpoles (6% in bG/ABC). In the embryos injected with the transcripts of bG/BC and bG/AC, the PGC-specificity of the reporter protein expression was 29% and 25%, respectively and both were significantly lower than that of bG/ ABC. In addition, the Venus protein in regions other than PGCs was expressed much more strongly in bG/BC, in contrast with the lack of expression in bG/AC and weak

expression in bG/AB (Figs. 7C and 8E, F, and G). These results indicate that all of sub-regions A, B, and C are required and sufficient for PGC-specific expression of the reporter protein. 2.4. Stability of the injected transcripts As described above, each sub-region A, B or C participated in the PGC-specific expression of the reporter protein Venus, if there was any contribution. In particular, sub-region B seemed to participate in the high level of expression in general. In our conditions, expression of the reporter protein must be regulated at multiple steps, such as anchoring, stability and translation of the mRNA. In order to determine the stability of the injected mRNA, we performed Northern blot analysis using venus probe (Fig. 9). As a control, the transcript of bG seemed to be degraded gradually after stage 11 and was reduced to around one-fifth of the original level at stage 30. Interestingly, the vegetally injected mRNA of bG fused to sub-region ABC (bG/ABC) was drastically degraded after stage 11, and almost completely lost at stage 30. The mRNA for bG fused to sub-region A (bG/A) was also degraded and was undetectable at stage 30. In contrast, the

754

K. Kataoka et al. / Mechanisms of Development 123 (2006) 746–760

Fig. 7. Characterization of the requirement of sub-regions A, B and C for the PGC-specific expression. (A) The schematic diagram of the construct, bG/ ABC. The sub-regions ABC were inserted in the middle of b-globin 3 0 -UTR (bG) following the Venus coding region. (B) The cDNA sequence of 3 0 UTR of bG/ABC. The upper and lower lines are the nucleotide and amino acid (Venus) sequences, respectively. Thin and bold asterisks indicate stop codons from venus and b-globin genes, respectively. The small letters show the nucleotide sequence of the 3 0 UTR of the Xenopus b-globin gene (bold, DDBJ Accession No. V01433) and pCS2-venus vector (thin). The capital letters show the DEADSouth 3 0 UTR region (295–1135) required for the PGC-specific expression. The region was further divided into sub-regions A (pink), B (yellow) and C (blue) by the above analysis. The putative target sequences for miR-427 microRNA are shown in bold red. Vera protein binding sites (E2, E2-like and E4) are boxed and CAC-motif are wavy underlined. Some sites of restriction enzymes are also indicated. (C) The constructs of the transcript in combination with various sub-regions. pA indicates a transcribed poly(A) tail. Probability of each construct against bG/ABC was calculated by Student’s t-test. Probability of bG/ABC vs. DS3 0 UTR was 0.67. Probability of bG/ABC vs. bG was 0. Asterisks indicate statistically significant differences (P < 0.05). Refer to the legend of Fig. 5.

transcripts without sub-region A, bG/B and bG/C were not degraded drastically and one-fourth of the transcripts remained at stage 30. The drastic degradation seemed to occur in cells other than PGCs, because it also occurred in the embryos in which the mRNA was injected into the animal pole. In order to examine the distribution of

injected mRNA, we performed whole-mount in situ hybridization of stage 30-embryos. The transcript of bG remained in the endodermal region derived from the vegetal region of the fertilized egg where it was introduced. Few embryos (only 13%) showed signals in PGCs in addition to the endodermal region. In contrast, in most embryos injected

K. Kataoka et al. / Mechanisms of Development 123 (2006) 746–760

755

Fig. 8. Expression of Venus protein in the stage 41 tadpoles injected with the transcripts of venus-b-globin fused to various sub-regions. Whole-mount lateral views (left) are shown with high-magnification views of PGCs (right), respectively. Solid and dashed brackets indicate PGCs with and without fluorescence, respectively. The injected transcript is indicated on the upper right of each image. Refer to Fig. 7C for the transcripts. Scale bars, 1 cm and 250 lm in (H) and (H 0 ), respectively.

with the transcript of bG/A or bG/ABC, the RNA remained only in PGCs (brackets in whole-mount images of Fig. 9). Although transcripts of both bG/B and bG/C remained in PGCs (48% and 36%, respectively), they also were detected broadly in part of the endodermal region. The transcript of bG/B seemed to be present at higher levels and distributed more broadly in the endodermal region than that of bG/C. The transcript of bG/ABC was not detected anywhere in most embryos when it was injected into the animal pole of the fertilized eggs. This also indicated that the injected transcripts with sub-region A were degraded in the whole embryo. In addition, on Northern blots, we found that the injected mRNA became slightly longer at stage 11. The elongation of the injected mRNA seemed to be more drastic for bG/B in comparison with bG and bG/C (open and solid arrowheads in Fig. 9). This may be due to poly(A) tailing of the transcripts after injection. The poly(A) tailing did not seem to occur in the transcripts of bG/A and bG/ABC, because we did not see any change of these transcripts in size on the Northern blot. Taken together, these findings indicate that the Venus expression from the injected transcripts was regulated at least at the step of mRNA degradation.

3. Discussion So far, there have been many reports on the behavior of PGCs during Xenopus development. They were almost all based on the identification of PGCs using their morphological features, such as being relatively large in size, round in shape, rich in yolk granule, and showing the presence of germ plasm. It is very laborious to identify dozens of the PGCs among a large number of embryonic cells. The number of PGCs is also highly dependent upon the embryo. Here, we were successful in establishing a method to identify and monitor the PGCs easily in a living embryo using the venus transcript fused to the 3 0 UTR of the DEADSouth gene. In general, it is well known that for many mRNAs, the localization and translation are controlled by the cis elements located in their 3 0 UTR. Although DEADSouth was identified as a gene encoding a putative DEAD-box RNA helicase, whose transcript was a component of the germ plasm like that of vasa in Drosophila, it is similar to MusDEAD5 rather than vasa, a member of the IF4A-like DEAD-box family (MacArthur et al., 2000). The transcript of DEADSouth was detected constantly as a single band of 3.1 kb during oogenesis, and was then decreased and lost at

756

K. Kataoka et al. / Mechanisms of Development 123 (2006) 746–760

Fig. 9. Stability of the injected mRNA of venus-b-globin sub-region A, B and C. (Left) Northern blot analysis with radioactive venus DNA probe for total RNA at stages 1 (just after microinjection), 11 (gastrula) and 30 (tailbud). 18S rRNA bands are shown as a loading control (lower band in each row, stained with methylene blue). Closed and open arrowheads indicate the band of injected transcripts and the upper shifted band, respectively. The ratio of band intensity at stage 30 to that at stage 11 is also shown. Each ratio was the mean of duplicate experiments. UD indicates undetectable level. (Right) Whole-mount in situ hybridization (WISH) with digoxigenin-labeled anti-sense venus RNA probe. The embryos injected with a given transcript were allowed to develop until stage 30. The hybridization signal was blue. Brackets indicate the venus-positive PGCs. The percentage of embryos with venuspositive PGCs shown on the right was calculated from a total of 3 independent injections. The lowest row shows the result from an experiment in which venus-b-globin transcript fused to sub-region ABC (bG/ABC) was injected at the animal pole of the fertilized eggs. Scale bar, 1 mm.

the end of gastrulation. The DEADSouth gene was expressed at a higher level in ovary than testis, in which it was expressed only in spermatogonia, spermatocytes and spermatids (MacArthur et al., 2000). We focused on the 3 0 UTR of this gene as a possible control element for PGC-specific expression. As expected, the PGC-specific expression of the reporter transcript could be detected initially around gastrulation and was observable externally around stage 30 using the venus transcript fused to the 3 0 UTR of the DEADSouth gene. We could easily distinguish the PGCs among a large number of the embryonic cells in the neurula (stage 20) if the embryo was dissociated (Fig. 2C). The PGCs were also

observable in the gonads, which formed on the mesonephros at stage 56 (Fig. 4). This indicated that it took several hours until translation of venus mRNA and maturation of the protein and that Venus protein was stable until stage 56 even if the venus mRNA was degraded. Thus, this system was useful for visualizing the PGCs during the developmental period. The PGCs were generated from the cells which incorporated germ plasm in the vegetal cortex of the fertilized egg. In later embryos, the number of PGCs with germ plasm was much less than would be expected if the germ plasm was inherited equally by two daughter cells of PGCs. It has been suggested that the germ plasm of PGCs is

K. Kataoka et al. / Mechanisms of Development 123 (2006) 746–760

inherited by one of the two daughter cells by unequal division (Whitington and Dixon, 1975). The microinjection of 3H thymidine into the embryos between stages 10 and 44 showed that relatively synchronous division of PGCs occurred three times, around stages 11, 23 and 38 (Dziadek and Dixon, 1977). We preliminarily found that the PGCs seemed to be slightly increased in number between stages 40 and 46, in contrast to being relatively constant in number between stages 30 and 40. Subsequently, the PGCs seemed to increase in number drastically between stages 48 and 56. It may be necessary to re-examine the previous experiments carefully using this PGC-visualizing technique. Based on electron microscopic observation, Kamimura et al. (1980) described that migrating PGCs (they called then ‘presumptive PGCs’) had morphological features corresponding to active migration, such as filopodia and loose contacts surrounding the cells. Recently, using a similar PGC-visualizing system in zebrafish, it was revealed that the migration of PGCs was guided by the chemokine SDF-1 through the receptor CXCR expressed in the PGCs (Doitsidou et al., 2002). Also in Xenopus, CXCR might be involved in the migration of PGCs (Nishiumi et al., 2005). Our visualization system should be very useful for analyzing the migration mechanism of the PGCs in Xenopus. To date, many RNAs have been identified as germ plasm components (Kloc et al., 2001). These RNAs were localized to the vegetal cortex of the oocyte via an ‘early (METRO)’ or ‘late’ pathway during oogenesis (reviewed in Kloc and Etkin, 2005). It is known that DEADSouth mRNA as well as Xcat2 mRNA, is localized to the germ plasm via the ‘early pathway’ (MacArthur et al., 1999, 2000). The cis element in the 3 0 UTR responsible for the localization was identified by a transport assay (Knaut et al., 2002; Betley et al., 2002). Knaut et al. (2002) found that in fish, a 180-nucleotide element was conserved among 3 0 UTRs of vasa genes of ostriophysans whose early embryos have germ plasm, but not conserved among those of euteleosts whose early embryos have no germ plasm. Surprisingly, a 180-nucleotide element in the sequence conferred localizing ability on the reporter transcript even in Xenopus oocytes. A sequence with low similarity to this element was also found in DEADSouth of Xenopus and functioned as a localization signal during Xenopus oogenesis (Knaut et al., 2002; nucleotides 65–584 of the 3 0 UTR of DEADSouth in this study, Fig. 5). Using a newly developed computer program, Betley et al. (2002) found that CACcontaining motifs were significantly present in the 3 0 UTRs of various transcripts localized to the vegetal cortex, including DEADSouth mRNA. Using the transport assay, they also showed that the 3 0 UTR containing the motif was important for the localization during oogenesis (nucleotides 300–631 of 3 0 UTR of DEADSouth in this study, Fig. 5). Thus, the element is sufficient for the localization of DEADsouth mRNA during Xenopus oogenesis. However, other elements must also be required for the PGC-specific expression in later development. Zhou and King

757

(2004) pointed out that RNA localization involved four basic steps: (1) selection of RNA for localization in germ plasm, (2) actual localization of the RNA via a motor-driven transport system, (3) immobilization of the RNA within the germ plasm, and (4) control of translation of the RNA. We do not know if a single cis element is required or if multiple elements are required for PGC-specific expression in later development. In this study, we defined a region of 3 0 UTR of DEADSouth gene (the nucleotides 295–1135), which was required for the PGC-specific expression. This region contained the common region between the localization elements which have been independently defined (Betley et al., 2002; Knaut et al., 2002), indicating that the localization element was insufficient for PGC-specific expression after fertilization. Based on deletion analysis, we defined tentatively two sub-regions (B and C) in addition to sub-region A as the common localization element, and examined if each sub-region conferred the PGC-specific expression on the reporter transcript of venus fused to the 3 0 UTR of b-globin gene (Figs. 7 and 8). Each of sub-regions A, B and C seemed to have ‘anchoring’ activity at the germ plasm rather than ‘localizing’ activity, because we examined the expression of the reporter transcript after microinjection into the vegetal pole of the fertilized egg, where germ plasm pre-existed. No PGC-specific expression was observed when the reporter transcript of venus fused to the 3 0 UTR of b-globin gene was injected as a control. This was also supported by the weak localization of the reporter transcript to PGCs when it was microinjected into the animal pole (Fig. 9). Sub-region A played a central role in anchoring of the reporter transcript in comparison with sub-regions B and C (Fig. 7C). Surprisingly, the mRNA with sub-region A was also degraded in regions other than PGCs (Fig. 9). This degradation in regions other than PGCs and the stability in PGCs must also occur for native DEADSouth mRNA, because it was not detected after gastrulation by RNase protection assays (MacArthur et al., 2000), but was detected only in the PGCs of tailbud (stage 30) by in situ hybridization (Fig. 1E). It was also reported that the 3 0 UTR of Vg1 mRNA, which is another mRNA localized to the vegetal cortex, had an element for translational repression, which was distinct from the localization element (Wilhelm et al., 2000; Otero et al., 2001). Translation of Vg1 mRNA was repressed in the cytoplasm of oocytes via the repression element, but occurred after localization to the vegetal cortex. The PGC-specific expression of DEADSouth mRNA was surely established at least by the degradation rather than the translational repression, in regions other than PGCs. Sub-region A seemed to be an anchor element and a degradation element as well as a localization element. Deshler et al. (1997) showed that Vera protein bound to the localization element of the 3 0 UTR of Vg1 mRNA, by which the mRNA was localized to the vegetal cortex during Xenopus oogenesis. There are five copies and one copy of the predicted binding sites of Vera protein in sub-regions

758

K. Kataoka et al. / Mechanisms of Development 123 (2006) 746–760

A and B, respectively (Fig. 7B). Although it is not known whether Vera protein is involved in the transport of DEADSouth mRNA, it is possible that Vera protein is involved not only in transport, but also in anchoring of the mRNA. This might be supported by the fact that throughout sub-region ABC, there are many CAC-motifs related to the Vera binding sites which are found in 3 0 UTRs of different transcripts localized to the vegetal cortex of the egg (Fig. 7B). Recently, in zebrafish, it has been revealed that the maternally supplied mRNAs were degraded via the micro-RNA miR-430, which began to be expressed at the onset of zygotic transcription (2.5 h after fertilization), accumulated during the maternal-to-zygotic transition and decreased later. miR-430 is the most abundant miRNA family and is ubiquitously expressed (Chen et al., 2005; Giraldez et al., 2006). In Xenopus, miR-427 seems to correspond to miR-430 in zebrafish, judging by the sequence similarity, which is one of the most abundant microRNA families, and shows the approximate expression timing (Watanabe et al., 2005; Chen et al., 2005). We found three copies of the possible target sequence for miR-427 only in sub-region A (Fig. 7B). It is an attractive hypothesis that miR-427 is involved in the degradation of DEADSouth mRNA in cells other than PGCs to ensure the PGC-specific translation. However, it will be necessary to examine if miR-427 is involved directly in degradation of the reporter mRNA with sub-region A. It also remains necessary to determine the detailed spatial and temporal expression pattern of miR-427 with special reference to the PGCs in Xenopus embryos. It will be necessary to perform deletion and base substitution analyses of sub-region A in order to identify cis elements for anchoring and degradation functions. Sub-region B obviously had enhancing activity for translation in addition to anchoring activity (Figs. 7–9), because the fluorescence of bG/B was stronger than that of bG and the stability of the injected bG/B transcript did not seem to be different from that of bG. This activity was observable not only in the PGCs, but also in the other cells. Interestingly, the injected transcript with sub-region B was obviously elongated until stage 11, probably due to poly(A) tailing. Thus, although the translational activity of sub-region B might be enhanced via poly(A) tailing, the elongation did not seem to be related to the stability of the mRNA (Fig. 9). It is interesting that elongation did not seem to occur in the bG/A or bG/ABC transcript, which also contained sub-region B. The elongation of the transcript might be suppressed by sub-region A. Unfortunately, it was not clear if the elongation of the transcript occurred in the PGCs, because we could not detect the injected transcript distributed in the PGCs on Northern blots. Sub-region C had also anchoring activity as well as subregion B, but not enhancing activity for translation (Figs. 7C and 8). The elongation of the transcript with sub-region C seemed to be similar to that of control transcript (bG),

and shorter than that of the transcript containing sub-region B (Fig. 9). Although the anchoring activity of sub-region C was not so strong, it was required for complete PGC-specific expression. Xenopus eggs are useful tools for analyzing early development, including PGC development, because they can be easily microinjected with various molecules. In this study, the application of a reporter gene fused to control elements of the PGC-specific gene enabled us to label the PGCs in development from the gastrula stage (stage 10) to the tadpole stage (stage 56). This visualization technique will enable the preparation of Xenopus PGCs on a large scale in combination with fluorescence-activated cell sorter technology. Actually, in mouse and Drosophila, such research has already resulted in successful profiling of genes expressed in the PGCs (for example, Abe et al., 1996; Shigenobu et al., 2006). The PGC visualization system is also very useful for analyzing the behavior of PGCs when function of a PGC-specific gene is deleted. Although it might not be so easy to observe the PGCs in a living, thick and pigmented early embryo, it may be possible to observe them continuously for a short period in the sliced embryo. 4. Experimental procedures 4.1. Eggs Sexually mature male and female wild-type Xenopus laevis were purchased commercially and maintained at 22 C in our laboratory. The females were injected with 50 U of pregnant mare serum gonadotropin (Teikoku Zouki) 2 or 4 days before use. Eggs were manually squeezed from the adult females into a dry dish a half day after induction of spawning by injection of 800 U of human chorionic gonadotropin (Teikoku Zouki) into the dorsal lymph sac. A testis was dissected from a male and suspended in a microcentrifuge tube containing 0.5 · Marc’s Modified Ringer solution (MMR; 100 mM NaCl/2 mM KCl/2 mM CaCl2/1 mM MgCl2/5 mM Hepes buffer (pH7.5). A few drops of the testis suspension were placed on the squeezed eggs, mixed and allowed to settle in 0.1 · MMR at 14 C for about 20 min. Then, the in vitro fertilized eggs were de-jellied with gentle agitation in 3% cysteine hydrochloride (pH 7.8) and then washed three times in autoclaved tap water. The fertilized eggs were kept in 0.1 · MMR at 14 C until microinjection.

4.2. Preparation of constructs The 3 0 UTR of the DEADSouth gene was amplified by polymerase chain reaction (PCR) with EST cDNA clone (XL035o16; Xenopus Data Base XDB3 in National Institute for Basic Biology, http://xenopus. nibb.ac.jp/) as a template and the following set of primers: forward primer with EcoRI site: 5 0 -GCCGAATTCAGTGATGAGTCGGCATGATCA3 0 , reverse primer with T7 primer according to the manufacturer’s protocol (TaKaRa ExTaq polymerase). The PCR product was cloned into the pCS2-venus vector (Nagai et al., 2002) using the EcoRI and XhoI sites and the construct was named DS3 0 UTR. The DS3 0 UTR construct was linearized by cutting with XhoI and used as a template for in vitro synthesis of the venus RNA with almost all of the 3 0 UTR of the DEADSouth gene. The deletion mutants from the 3 0 end of the 3 0 UTR were constructed by PCR using SP6 primer (5 0 -CGGAGCAAGCTTGA TTTAGGTGACAC TATA-3 0 ) and each reverse primer (3 0 D1–D6). The PCR product was purified with a QIAquick PCR kit (Qiagen) and used as a template for mRNA synthesis. The deletion mutants from the 5 0 end of 3 0 UTR were

K. Kataoka et al. / Mechanisms of Development 123 (2006) 746–760 also constructed by PCR (Vallejo et al., 1995). The Venus region was amplified with SP6 primer and Venus 3 0 end reverse primer (5 0 AAGAATTCTTACTTGTACAGCTCGTCCATG-3 0 ). The 5 0 -deleted 3 0 UTR was amplified by PCR using the forward primer tagged with a part of the 3 0 end of the Venus coding region (5 0 D1–D6) and 3 0 D1 primer. The PCR fragments of Venus and 5 0 -deleted 3 0 UTR were mixed and used as a template for PCR using SP6 primer and 3 0 D1 primer. The combined fragment of Venus and 5 0 -deleted 3 0 UTR was checked on an agarose gel and subjected to mRNA synthesis. The mutants with a gap were also constructed in the same way using the reverse primer corresponding to nucleotides 570–584 tagged with a part of each forward primer such as 5 0 D4 (for example, D585–791 construct), instead of Venus 3 0 end reverse primer. As described above, b-globin 3 0 UTR (DDBJ Accession No. V01433) was also amplified using pXbG-ev1 (Preston et al., 1992) as a template and a set of primers (forward 5 0 -GGCAGATCTGAATTCCACTAAA CCAGCCTC-3 0 : reverse 5 0 -GGGGGGCTCGAGTTTTTTTTTTTTT TTTTT-3 0 ). The fragment was inserted into pCS2-venus using the EcoRI and XhoI sites (named bG). The regions A, B, C, AB, BC, AC and ABC were amplified by PCR and inserted directly into the middle site of b-globin 3 0 UTR of bG using PCR technology (Vallejo et al., 1995) (Fig. 7). These constructs were confirmed by sequencing, linearized by cutting with XhoI and used as a template for RNA synthesis. Although the sequence of our region ABC cloned by PCR from the EST clone (NIBB, XDB3, XL035o16: Kitayama et al. unpublished) was different at several nucleotides from that of the original sequence (GenBank Accession No. AF190623), our sequence was functional for PGC-specific expression (Fig. 7B).

759

solutions (Ito et al., 2001). For whole-mount specimens, the gonads were picked up with mesonephroses from stage 56 tadpoles. After elimination of fat bodies, they were fixed in the same way. The sections or whole-mount specimens were washed in PBS(), blocked in PBS() containing 0.1% Triton X100 (TPBS()) supplemented with 10% goat serum and reacted with mouse anti-Xdazl (Mita and Yamashita, 2000) and rabbit anti-GFP (A. v. peptide antibody, Clontech) antibodies. Then they were washed in TPBS() and reacted with Alexa488 conjugated anti-rabbit IgG (Molecular Probes) and Cy3-conjugated antimouse IgG (Jackson Immuno Research Laboratories). After washing in TPBS(), they were mounted in 50% glycerol and observed under a fluorescent microscope BX60 (Olympus). The whole-mount specimens were observed under a binocular SZX12 (Olympus) without mounting.

4.7. Northern blot analysis Total RNA was purified from the embryos at a given stage as described previously (Orii et al., 1999). The RNA (10 or 15 lg per lane) was glyoxylated, separated on a 1.5% agarose gel, and transferred to a nylon membrane (Hybond N+, Amersham). The membrane was stained with 0.04% methylene blue/0.5 M sodium acetate (pH 5.2) to check the quality and quantity of the separated RNA by examining the ribosomal RNA bands. Then, it was subjected to hybridization as described previously (Kataoka et al., 2005). The relative intensity of the radioactive bands was quantified using an image analyzer FLA3000 with Image Gauge software (Fuji Film).

Acknowledgements 4.3. Preparation and microinjection of mRNA Venus mRNA with various 3 0 UTRs was synthesized using mMESSAGE mMACHINE SP6 Kit (Ambion) with a template construct as described above. After purification by extraction with phenol, precipitation with isopropylalcohol and rinsing with 70% ethanol, the RNA was dissolved in water at a concentration of about 0.12 pmol/lL. The RNA (4.6 or 9.2 nL) was injected into the cortical region at the vegetal pole of a fertilized egg (1-cell stage embryo) in 0.5 · MMR containing 3% Ficoll (Amersham Biosciences) (Fig. 1A). The injected embryos were kept in the same solution at 18 C. At the blastula stage, they were transferred into 0.1 · MMR containing 3% Ficoll. The embryos were observed under a fluorescent binocular SZX12 (Olympus). The embryos were staged according to Nieuwkoop and Faber (1994).

4.4. Whole-mount in situ hybridization The embryos were fixed in 10% methanol/20% formalin/phosphate buffered saline without Ca2+ and Mg2+ (PBS(-)) at 4 C overnight. Whole-mount in situ hybridization was performed using digoxigeninlabeled antisense RNA probe as described previously (Kataoka et al., 2005). The hybridization signal was detected using Anti-Digoxigenin-AP (Roche) and BM purple AP substrate (Roche). The specimens were bleached with 10% H2O2/PBS() after signal detection.

4.5. Manual sectioning of fixed embryos The embryos were fixed with 2% paraformaldehyde/0.1 M MOPS (pH 7.5)/0.5 M NaCl at 4 C in the dark overnight and sectioned manually with a razor blade.

4.6. Immunohistochemistry For sections, the embryos were fixed in 4% paraformaldehyde/PBS() at 4 C for 1 day, dehydrated through a series of ethanol-butanol solutions, embedded in paraffin, sliced at 6-lm thickness with a microtome and mounted on slide glasses. The specimens were de-paraffinized through a series of xylene, and hydrated through a series of ethanol–water

We are grateful to Atsushi Miyawaki (RIKEN, BSI) for a generous gift of pCS2-Venus and to Koichi Mita and Masakane Yamashita (Graduate School of Science, Hokkaido University) for a generous gift of anti-Xdazl monoclonal antibody. We also thank Takuji Sugiura, Yuka Taniguchi and other members of our lab for their support. References Abe, K., Hashiyama, M., Macgregor, G., Yamamura, K., 1996. Purification of primordial germ cells from TNAPbeta-geo mouse embryos using FACS-gal. Dev. Biol. 180, 468–472. Betley, J.N., Frith, M.C., Graber, J.H., Choo, S., Deshler, J.D., 2002. A Ubiquitous and conserved signal for RNA localization in chordates. Curr. Biol. 12, 1756–1761. Chen, P.Y., Manninga, H., Slanchev, K., Chien, M., Russo, J.J., Ju, J., Sheridan, R., John, B., Marks, D.S., Gaidatzis, D., Sander, C., Zavolan, M., Tuschl, T., 2005. The developmental miRNA profiles of zebrafish as determined by small RNA cloning. Genes Dev. 19, 1288– 1293. Deshler, J.D., Highett, M.I., Schnapp, B.J., 1997. Localization of Xenopus Vg1 mRNA by Vera protein and endoplasmic reticulum. Science 276, 1128–1131. Doitsidou, M., Reichman-Fried, M., Stebler, J., Ko¨prunner, M., Do¨rries, J., Meyer, D., Esguerra, C.V., Leung, T., Raz, E., 2002. Guidance of primordial germ cell migration by the chemokine SDF-1. Cell 111, 647–659. Dziadek, M., Dixon, K.E., 1977. An autoradiographic analysis of nucleic acid synthesis in the presumptive primordial germ cells of Xenopus laevis. J. Embryol. Exp. Morphol. 37, 13–31. Giraldez, A.J., Mishima, Y., Rihel, J., Grocock, R.J., Van Dongen, S., Inoue, K., Enright, A.J., Schier, A.F., 2006. Zebrafish miR-430 promotes deadenylation and clearance of maternal mRNAs. Science 312, 75–79. Horvay, K., Claussen, M., Katzer, M., Landgrebe, J., Pieler, T., 2006. Xenopus Dead end mRNA is a localized maternal determinant that

760

K. Kataoka et al. / Mechanisms of Development 123 (2006) 746–760

serves a conserved function in germ cell development. Dev. Biol. 291, 1–11. Houston, D.W., King, M.L., 2000. A critical role for Xdazl, a germ plasmlocalized RNA, in the differentiation of primordial germ cells in Xenopus. Development 127, 447–456. Ikenishi, K., Nakazato, S., Okuda, T., 1986. Direct evidence for the presence of germ cell determinant in vegetal pole cytoplasm of Xenopus laevis and in a subcellular fraction of it. Dev. Growth Differ. 28, 563– 568. Illmensee, K., Mahowald, A.P., 1974. Transplantation of posterior polar plasm in Drosophila. Induction of germ cells at the anterior pole of the egg. Proc. Natl. Acad. Sci. USA 71, 1016–1020. Ito, H., Saito, Y., Watanabe, K., Orii, H., 2001. Epimorphic regeneration of the distal part of the planarian pharynx. Dev. Genes Evol. 211, 2–9. Kamimura, M., Kotani, M., Yamagata, K., 1980. The migration of presumptive primordial germ cells through the endodermal cell mass in Xenopus laevis: a light and electron microscopic study. J. Embryol. Exp. Morphol. 19, 1–17. Kataoka, K., Tazaki, A., Kitayama, A., Ueno, N., Watanabe, K., Mochii, M., 2005. Identification of asymmetrically localized transcripts along the animal-vegetal axis of the Xenopus egg. Dev. Growth Differ. 47, 511–521. Kloc, M., Bilinski, S., Chan, A.P., Allen, L.H., Zearfoss, N.R., Etkin, L.D., 2001. RNA localization and germ cell determination in Xenopus. Int. Rev. Cytol. 203, 63–91. Kloc, M., Dougherty, M.T., Bilinski, S., Chan, A.P., Brey, E., King, M.L., Patrick Jr., C.W., Etkin, L.D., 2002. Three-dimensional ultrastructural analysis of RNA distribution within germinal granules of Xenopus. Dev. Biol. 241, 79–93. Kloc, M., Etkin, L.D., 2005. RNA localization mechanisms in oocytes. J. Cell Sci. 118, 269–282. Knaut, H., Steinbeisser, H., Schwarz, H., Nusslein-Volhard, C., 2002. An evolutionary conserved region in the vasa 3 0 UTR targets RNA translation to the germ cells in the zebrafish. Curr. Biol. 12, 454–466. Ko¨prunner, M., Thisse, C., Thisse, B., Raz, E., 2001. A zebrafish nanosrelated gene is essential for the development of primordial germ cells. Genes Dev. 15, 2877–2885. MacArthur, H., Bubunenko, M., Houston, W.D., King, M.L., 1999. Xcat2 RNA is a translationally sequestered germ plasm component in Xenopus. Mech. Dev. 84, 75–88. MacArthur, H., Houston, W.D., Bubunenko, M., Mosquera, L., King, M.L., 2000. DEADSouth is a germ plasm specific DEAD-box RNA helicase in Xenopus related to eIF4A. Mech. Dev. 95, 291–295. Machado, R.J., Moore, W., Hames, R., Houliston, E., Chang, P., King, M.L., Woodland, H.R., 2005. Xenopus Xpat protein is a major component of germ plasm and may function in its organization and positioning. Dev. Biol. 287, 289–300. Mita, K., Yamashita, M., 2000. Expression of Xenopus Daz-like protein during gametogenesis and embryogenesis. Mech. Dev. 94, 251–255. Mosquera, L., Forristall, C., Zhou, Y., King, M.L., 1993. A mRNA localized to the vegetal cortex of Xenopus oocytes encodes a protein with a nanos-like zinc finger domain. Development 117, 377–386.

Nagai, T., Ibata, K., Park, E.S., Kubota, M., Mikoshiba, K., Miyawaki, A., 2002. A variant of yellow fluorescent protein with fast and efficient maturation for cell-biological applications. Nat. Biotechnol. 20, 87–90. Nieuwkoop, P.D., Faber, J., 1994. Normal Table of Xenopus laevis (Daudin): A Systematical and Chronological Survey of the Development from the Fertilized Egg Till the End of Metamorphosis. Garland Publications. Nishiumi, F., Komiya, T., Ikenishi, K., 2005. The mode and molecular mechanisms of the migration of presumptive PGC in the endoderm cell mass of Xenopus embryos. Dev. Growth Differ. 47, 37–48. Okada, M., Kleinman, I.A., Schneiderman, H.A., 1974. Restoration of fertility in sterilized Drosophila eggs by transplantation of polar cytoplasm. Dev. Biol. 37, 43–54. Orii, H., Kato, K., Umesono, K., Sakurai, T., Agata, K., Watanabe, K., 1999. The planarian HOM/HOX homeobox genes (Plox) expressed along the anteroposterior axis. Dev. Biol. 210, 456–468. Otero, L.J., Devaux, A., Standart, N., 2001. A 250-nucleotide UArich element in the 3 0 untranslated region of Xenopus laevis Vg1 mRNA represses translation both in vivo and in vitro. RNA 7, 1753–1767. Preston, G.M., Carroll, T.P., Guggino, W.B., Agre, P., 1992. Appearance of water channels in Xenopus oocytes expressing red cell CHIP28 protein. Science 256, 385–387. Shigenobu, S., Arita, K., Kitadate, Y., Noda, C., Kobayashi, S., 2006. Isolation of germline cells from Drosophila embryos by flow cytometry. Dev. Growth Differ. 48, 49–57. Tanaka, M., Kinoshita, M., Kobayashi, D., Nagahama, Y., 2001. Establishment of medaka (Olyzias latipes) transgenic lines with the expression of green fluorescent protein exclusively in germ cells: a useful model to monitor germ cells in a live vertebrate. Proc. Natl. Acad. Sci. USA 98, 2544–2549. Vallejo, A.N., Pogulis, R.J., Pease, L.R., 1995. Mutagenesis and synthesis of novel recombinant genes using PCR. In: Dieffenbach, C.W., Dveksler, G.S. (Eds.), PCR Primer: A Laboratory Manual. Cold Spring Harbor Laboratory Press, pp. 603–612. Wakahara, M., 1996. Primordial germ cell development: Is the urodele pattern closer to mammals than to anurans? Int. J. Dev. Biol. 40, 653– 659. Watanabe, T., Takeda, A., Mise, K., Okuno, T., Suzuki, T., Minami, N., Imai, H., 2005. Stage-specific expression of microRNAs during Xenopus development. FEBS Lett. 579, 318–324. Whitington, P.McD., Dixon, K.E., 1975. Quantitative studies of germ plasm and germ cells during early embryogenesis of Xenopus laevis. J. Embryol. Exp. Morphol. 33, 57–74. Wilhelm, J.E., Vale, R.D., Hegde, R.S., 2000. Coordinate control of translation and localization of Vg1 mRNA in Xenopus oocytes. Proc. Natl. Acad. Sci. USA 97, 13132–13137. Wolke, U., Weidlinger, G., Ko¨prunner, M., Raz, E., 2002. Multiple levels of posttranscriptional control lead to germ line-specific gene expression in the zebrafish. Curr. Biol. 12, 289–294. Zhou, Y., King, M.L., 2004. Sending RNAs into the future: RNA localization and germ cell fate. IUBMB Life 56, 19–27.