Mitochondrial small ribosomal RNA is present on polar granules in early cleavage embryos of Drosophila melanogaster

Authors


Abstract

In Drosophila, formation of the germline progenitors, the pole cells, is induced by polar plasm localized in the posterior pole region of early embryos. The polar plasm contains polar granules, which act as a repository for the factors required for pole cell formation. It has been postulated that the factors are stored as mRNA and are later translated on polysomes attached to the surface of polar granules. Here, the identification of mitochondrial small ribosomal RNA (mtsrRNA) as a new component of polar granules is described. The mtsrRNA was enriched in the polar plasm of the embryos immediately after oviposition and remained in the polar plasm throughout the cleavage stage until pole cell formation. In situ hybridization at an ultrastructural level revealed that mtsrRNA was enriched on the surface of polar granules in cleavage embryos. Furthermore, the localization of mtsrRNA in the polar plasm depended on the normal function of oskar, vasa and tudor genes, which are all required for pole cell formation. The temporal and spatial distribution of mtsrRNA is essentially identical to that of mitochondrial large ribosomal RNA (mtlrRNA), which has been shown to be required for pole cell formation. Taken together, it is speculated that mtsrRNA and mtlrRNA are part of the translation machinery localized to polar granules, which is essential for pole cell formation.

Introduction

In many animal groups, it has been proposed that factors for germline formation are localized in a histologically distinct region of the egg cytoplasm known as the germ plasm (Eddy 1975). In Drosophila, the germ plasm (also called polar plasm) is localized in the posterior pole region of early embryos and is partitioned into pole cells, the germline precursors. Several experiments have demonstrated that polar plasm contains factors sufficient for pole cell formation. Polar plasm is able to restore the ability to form pole cells in ultraviolet (UV)-irradiated embryos (Okada et al. 1974). Furthermore, the injection of polar plasm into the anterior pole region induces the formation of functional pole cells (Illmensee & Mahowald 1974).

Polar plasm is distinguished in the cytoplasm by the presence of electron-dense polar granules (Mahowald 1962). These granules first appear in the posterior pole of the late-stage oocytes and remain in the polar plasm of early cleavage embryos (Mahowald 1962, 1968, 1971a). At pole cell formation, polar granules are incorporated into pole cells. Because polar granules are present in germ line throughout most of the life cycle (Mahowald 1968, 1971a), they are regarded as the structures essential for germline development. This idea is further supported by the genetic analysis of oskar (osk), vasa (vas) and tudor (tud) genes (Boswell & Mahowald 1985; Lehmann & Nüsslein-Volhard 1986; Schüpbach & Wieschaus 1986). The mutation of any one of these genes affects polar granule assembly as well as pole cell formation (Boswell & Mahowald 1985; Lehmann & Nüsslein-Volhard 1986; Schüpbach & Wieschaus 1986). Conversely, mislocalization of osk mRNA causes embryos to form ectopic polar granules and pole cells (Ephrussi & Lehmann 1992; Bardsley et al. 1993). Furthermore, protein products from osk, vas and tud are all components of polar granules (Hay et al. 1988; Bardsley et al. 1993; Breitwieser et al. 1996).

It has been reported that polar granules in mature oocytes are composed of RNA(s) and protein(s), (Mahowald 1971b). The RNA component(s) remains detectable in polar granules after fertilization, but it disappears prior to pole cell formation (Mahowald 1971b). This data, and the finding that polar granules form well-developed polysomes on their surface (Mahowald 1968, 1971b) led to the idea that maternal mRNA(s) stored in polar granules is translated on the polysomes to produce protein(s) required for pole cell formation. Thus, polar granules are regarded as the site for translation. This idea is compatible with the observation that one of the two ribosomal RNA encoded by the mitochondrial genome, namely mitochondrial large ribosomal RNA (mtlrRNA) is localized on the surface of polar granules. The mtlrRNA is transported out of mitochondria to reach the granules at around oviposition and remains on the granules until pole cell formation (Kobayashi et al. 1993; Amikura et al. 1996). Its localization depends on the normal function of osk, vas and tud genes (Ding et al. 1994; Kobayashi et al. 1995). Furthermore, mtlrRNA is essential for pole cell formation (Kobayashi & Okada 1989; Iida & Kobayashi 1998).

Here, we report the localization of mitochondrial small ribosomal RNA (mtsrRNA) on polar granules. The present results show that mtsrRNA is present on the surface of granules during the stages from oviposition to pole cell formation. Furthermore, the localization of mtsrRNA, like mtlrRNA, requires osk, vas and tud functions. Thus, both mitochondrial rRNA are localized on polar granules, suggesting the possibility that they form mitochondrial ribosomes on which polar granule-stored mRNA encoding the pole cell-forming factors are translated.

Materials and Methods

Embryos and oocytes

Wild-type embryos and mature oocytes were obtained from Oregon R Drosophila melanogaster. Unfertilized eggs were obtained from virgin females of a transformant line (YPhsSPg fly, g-10-3), which produces male sex peptide and ovulates eggs without mating (Chen et al. 1988; Aigaki et al. 1991). Mutant embryos were obtained from females homozygous for osk150, vasPD23, tudWC8 and nosBN. The full genotypes of these mutant stocks are st osk150 e/TM3 Sb, vasPD23cn bw/CyO, tudWC8 bw/CyO l(2)DTS and nosBN e/TM3 Sb Ser. Embryos in which osk mRNA is mislocalized at the anterior pole region were obtained from two independent transformant lines. These embryos have the osk-bcd 3′UTR fusion gene in the second and third chromosomes, respectively (Ephrussi & Lehmann 1992).

Probes and in situ hybridization with whole-mount embryos

A 0.8-kb DNA encompassing the entire length of mtsrRNA gene was amplified from D. melanogaster mitochondrial genomic DNA by polymerase chain reaction (PCR) using 5′-CGGAATTCATTCTAGATACACTTTCCAGTA-3′ and 5′-CCGGATCCTTAAAGTT TTATTTTGGCTTAAAAATT-3′ as a pair of primers. This fragment was cloned between EcoRI and BamHI sites in pBluescript. This insert was labeled by Digoxigenin (DIG) DNA labeling kit (Boehringer Mannheim, Mannheim, Germany).

In situ hybridization to whole-mount embryos was carried out according to the method by Amikura et al. (1993, 1996) and Kobayashi et al. (1993).

Visualization of RNA for electron microscopy

Visualization of hybridized signals for electron microscopy was carried out according to the method used by Amikura et al. (1993, 1996). We used 1 nm gold-conjugated anti-DIG antibody (BioCell, Cardiff, UK) for detection of the DIG-labeled probe. The antibody was pre-absorbed for 1.5 h with fixed embryos before use. The incubation was performed for 1 h at room temperature. To intensify the signal, the embryos were dipped in a silver-enhancing solution (BioCell) for less than 3 min. The embryos were postfixed in 1% OsO4 for 1 h and embedded in epoxy resin according to Spurr (1969). Thin sections were stained by uranyl acetate and lead citrate to be observed under electron microscope (JEM-2000EXII; Jeol Co. Ltd, Tokyo, Japan).

In situ hybridization of mature oocytes

In situ hybridization of sectioned oocytes was performed according to Amikura et al. (1996) except for using DNA probes instead of RNA probes. Ovaries were fixed in 4% paraformaldehyde in phosphate-buffered saline (PBS) for 2 h. Fixed materials were embedded in paraffin wax. Sections (7 mm) were cut and placed on silane-coated slides. Prehybridization was performed in hybridization solution (HS) (HS: 50% formamide, 5 × SSC (750 mM NaCl, 75 mM sodium citrate), 1 mg/mL tRNA, 100 mg/mL sonicated salmon sperm DNA, 1 × Denhardt’s, 5 mM ethylenediamine tetraacetic acid (EDTA), 0.1% 3-[(3-cholamidopropyl)- dimethylammonio]-l-propanesulfonate (CHAPS), 50 mg/ mL heparin, 0.1% Tween 20) for 1 h at 45°C. Samples were then incubated for hybridization in HS containing 2 mg/mL labeled DNA probe for 12 h at 45°C.

Oocytes cut into halves were also in situ hybridized. Ovaries were dissected in PBS and fixed in 4% paraformaldehyde in PBS for 20 min. During fixation, mature oocytes were cut into halves using a pair of tungsten needles. Half oocytes were then transferred to PBS and washed for 5 min. Then the oocytes were transferred to 70% ethanol and dechorionated using tungsten needles. Then, the oocytes were dehydrated in 80, 90% ethanol for 10 min each and 100% methanol twice for 5 min. The oocytes were stored in methanol for several days at – 20°C. Pretreatment and hybridization were performed according to Amikura et al. (1993, 1996) and Kobayashi et al. (1993).

Results

Extra-mitochondrial mtsrRNA is localized in the polar plasm of early cleavage embryos

We have previously reported that mtsrRNA is not enriched in the polar plasm of cleavage embryos (Kobayashi et al. 1993). As the mitochondrial genomic fragment containing the mtsrRNA gene of Drosophila yakuba, which we used as a probe in the previous work, did not hybridize with the mtsrRNA of D. melanogaster, we cloned the mtsrRNA gene of D. melanogaster and used it as a probe for in situ hybridization in the present study. The temporal and spatial distribution of mtsrRNA in early embryos was determined by the whole-mount in situ hybridization method, which allows the detection of RNA only outside of the mitochondria (Amikura et al. 1993, 1996; Kobayashi et al. 1993). In early embryos immediately after oviposition, the mtsrRNA signal was concentrated in the polar plasm, but was very weak elsewhere in the egg cytoplasm (Fig. 1A). During the rest of the cleavage stage, the strong signal remained in the polar plasm (Fig. 1B). As the pole cell formation initiated, the weak signal became concentrated in the periplasm just beneath the pole buds (Fig. 1C). After the late syncytial blastoderm stage, the signal was no longer detectable (Fig. 1D). These results are consistent with the data reported by Ding et al. (1994). For a control, we used a probe for ND-1 mRNA, which remains within mitochondria during cleavage stages (Kobayashi & Okada 1989; Kobayashi et al. 1993), even though it is transcribed from the same strand of the mitochondrial genome as mtsrRNA (Clay & Wolstenholme 1985; Garesse 1988). No strong ND-1 signal was detected in the polar plasm of cleavage embryos (1Fig. 1E; Kobayashi et al. 1993; Ding et al. 1994). These results strongly suggest that mtsrRNA is transported outside of mitochondria in the polar plasm of cleavage embryos.

Figure 1.

. Distribution of mtsrRNA (A–D) and ND-1 mRNA (E) in early embryos. Whole-mount in situ hybridization of a mtsrRNA probe to embryos at (A) early cleavage stage (stage 1), (B) late cleavage stage (stage 2), (C) pole-bud stage (stage 3) and (D) syncytial blastodermal stage (stage 4) was carried out. ND-1 mRNA is not concentrated in the polar plasm of cleavage embryos (E). The stages of embryos are according to Campos-Ortega and Hartenstein (1985). Arrows indicate the mtsrRNA signal. Arrowheads show pole buds (C) and pole cells (D). In all panels, anterior is to the left. Bar (A), 100 μm.

Extra-mitochondrial mtsrRNA is localized on the surface of polar granules in early cleavage embryos

We next examined the subcellular distribution of mtsrRNA in polar plasm. We sectioned embryos for electron microscopy after whole-mount in situ hybridization. Table 1 shows that mtsrRNA signals were enriched in polar plasm. We found that 57% of the total signal in polar plasm was present on the surface of polar granules. The signals were always concentrated on one side of the granules (Fig. 2A,B). The rest of the signals in the polar plasm and the signal in the lateral region were evenly scattered throughout the cytoplasm (Fig. 2A–D). These results clearly show that enrichment of the mtsrRNA signal in the polar plasm was a result of the extra-mitochondrial signals in polar granules. The unlocalized mtsrRNA signal most likely represents background, as the number of unlocalized mtsrRNA signals was similar to that observed in control embryos. These embryos, which were not hybridized but treated with anti-DIG antibody (Table 1), displayed scattered signals throughout the cytoplasm, but no significant signal on the polar granules. Similar results were obtained when the embryos were hybridized with a ND-1 mRNA probe (1, 2Fig. 2E,F; Table 1).

Table 1.  . Distribution of mtsrRNA and ND-1 mRNA in polar plasm and lateral region of early cleavage embryos Thumbnail image of
Figure 2.

. Subcellular distribution of mtsrRNA and ND-1 in polar plasm (A,E) and the lateral region (C) of cleavage embryos hybridized with mtsrRNA (A,C) and ND-1 mRNA probe (E). (B,D,F) Micrographs at higher magnifications of the cytoplasmic region in A, C and E, respectively. Arrowheads show the signals on polar granules, arrows show the signals scattered in cytoplasm; pg, polar granules; mt, mitochondria. Bars, 1 μm (E); 200 nm (F).

Localization of mtsrRNA in polar plasm begins during a short period between the completion of oogenesis and oviposition

Distribution of extra-mitochondrial mtsrRNA in the mature oocytes (stage 14) was examined. To facilitate in situ hybridization, we cut the mature oocytes into halves and further removed the chorion from them. Subsequently, the oocytes were hybridized with the mtsrRNA probe. This method enabled us to detect osk mRNA localized in polar plasm of the mature oocytes (Fig. 3B; Ephrussi et al. 1991). However, no mtsrRNA signal in the polar plasm of the oocytes was detectable (Fig. 3A). We also performed in situ hybridization to the sectioned oocytes with the mtsrRNA and the osk mRNA probes. Even with this technique, no mtsrRNA signal was detectable in the polar plasm of mature oocytes (Fig. 3C), although the osk mRNA signal was tightly localized in the cytoplasm (Fig. 3D). These observations suggest that the transportation of mtsrRNA from mitochondria to polar granules began after the completion of oogenesis.

Figure 3.

.In situ hybridization of mtsrRNA (A,C) and osk mRNA (B,D) to the half-cut stage-14 oocytes (A,B) and the sectioned oocytes (C,D). The osk mRNA signal was localized in the posterior pole of stage-14 oocytes (arrows in B and D). (E) Subcellular distribution of mtsrRNA in the polar plasm of cleavage embryos. Note that mtsrRNA signals on the granules are enriched at the boundaries between polar granules (pg) and mitochondria (mt). (F) Whole-mount in situ hybridization of mtsrRNA to unfertilized eggs. The mtsrRNA signal was localized in the posterior pole (arrow). Bar (E), 200 nm.

In embryos immediately after oviposition, mtsrRNA signal was detected in the polar plasm using whole-mount in situ hybridization (Fig. 1A). In these embryos, the mtsrRNA signal was observed in polar granules by electron microscopy. Within 30 min after egg laying, polar granules and mitochondria are closely associated with each other (Mahowald 1968, 1971a). Almost all signals on the granules were enriched at the boundaries between these organelles (Fig. 3E). These observations suggest that transportation of mtsrRNA from mitochondria to polar granules initiates before oviposition. In Drosophila, fertilization occurs shortly before oviposition (Mahowald et al. 1983). We further examined whether fertilization was required for the localization of mtsrRNA in the polar plasm. We found that the mtsrRNA signal was enriched in the polar plasm of unfertilized eggs (Fig. 3F). The signal intensity in unfertilized eggs was almost identical to that observed in normal embryos. This observation indicates that fertilization is not a necessary step for the localization of mtsrRNA in the polar plasm.

Localization of mtsrRNA in polar plasm depends on the normal function of osk, vas and tud genes

We next examined whether the posterior localization of mtsrRNA was affected by mutations in osk, vas and tud genes, which disrupt polar granule assembly and pole cell formation. In cleavage embryos laid by mothers homozygous for any one of these mutations, the posterior localization of mtsrRNA was abolished (Fig. 4A–C). Thus, normal activities of these genes are all required for mtsrRNA localization. Among these genes, osk plays a key role in the pathway leading to polar granule assembly and pole cell formation (Ephrussi & Lehmann 1992). Mislocalization of osk mRNA to the anterior pole leads to the ectopic formation of polar granules and pole cells (Ephrussi & Lehmann 1992). In such embryos, the mtsrRNA signal was concentrated in the anterior as well as in the posterior (Fig. 4D). This clearly shows that the localization of mtsrRNA is under the regulation of osk.

Figure 4.

. Distribution of mtsrRNA in the cleavage embryos produced from females homozygous for maternal-effect mutations. The mtsrRNA signal was not enriched in polar plasm of the embryos from osk/osk (A), vas/vas (B) and tud/tud females (C). (D) In the embryos from females carrying the osk-bcd 3′UTR transgene, the mtsrRNA signal was concentrated at the anterior pole (arrowhead) as well as the posterior pole (arrow). The mtsrRNA signal was normally localized in the polar plasm of the embryo from nos/nos females (E).

The pathway downstream of osk is also required to localize the abdominal determinant, nanos (nos) mRNA (Lehmann & Nüsslein-Volhard 1991; Wang & Lehmann 1991; Ephrussi & Lehmann 1992). Embryos from females homozygous for nos fail to form abdominal structures. However, both polar granule assembly and pole cell formation are intact in these embryos (Lehmann & Nüsslein-Volhard 1991). In such mutant embryos, mtsrRNA was normally localized in the polar plasm (Fig. 4E). Thus, mtsrRNA localization was independent of nos function.

Discussion

We have previously shown that mtlrRNA is essential for pole cell formation by the inhibition of this process via targeted ribozyme injection (Iida & Kobayashi 1998). The mtlrRNA has also been identified as a factor that restores the ability to form pole cells when injected into UV-irradiated embryos (Kobayashi & Okada 1989). However, mtlrRNA is unable to induce pole cells at the anterior pole unless it is co-injected with UV-irradiated polar plasm (Kobayashi & Okada 1989), suggesting a requirement for additional factor(s). Based on the following results, we propose that mtsrRNA is one of the candidates. First, mtsrRNA is transported from mitochondria to polar granules after the completion of oogenesis and remains on the granules during cleavage stages until pole cell formation. This temporal and spatial distribution pattern is indistinguishable from that of mtlrRNA (Kobayashi et al. 1993; Amikura et al. 1996). Second, a common mechanism directs the localization of mtsrRNA and mtlrRNA on polar granules. We show that the localization of mtsrRNA, like mtlrRNA, is under the regulation of osk, vas and tud, which are all required for pole cell formation.

The present results show that the two mitochondrial rRNA are concentrated on the surface of polar granules. This distribution pattern is completely different from that of the other RNA component of polar granules, namely Pgc RNA, which distributes all over the granules (Nakamura et al. 1996). Furthermore, the stage when the mitochondrial rRNA are present on the granules is limited to the cleavage stage until pole cell formation, while Pgc RNA remains detectable in the granules after pole cell formation. These differences may reflect the roles of these molecules in germline formation. Pgc RNA has an essential role in the differentiation of pole cells into functional germ cells, but reduction of this RNA has only a modest effect on pole cell formation (Nakamura et al. 1996). In contrast to Pgc RNA, mtlrRNA is essential for pole cell formation but not for pole cell differentiation (Kobayashi & Okada 1989). We have previously reported that reduction of mtlrRNA does not affect polar plasm assembly nor its maintenance (Iida & Kobayashi 1998). Thus, we speculate that both mtlrRNA and mtsrRNA on the surface of polar granules are required only for pole cell formation.

It is interesting to note that on the surface of polar granules there are well-developed helical polysomes (Mahowald 1968, 1971b). They are regarded as the structures that translate the polar granule-stored maternal mRNA (Mahowald 1968, 1971b). For the following reasons, we speculate that both mitochondrial rRNA participate in the formation of polysomes on the surface of the polar granules. First, the polysomes appear on polar granules after transportation of mitochondrial rRNA to the granules. Second, both the formation of the polysomes on polar granules and the transportation of mitochondrial rRNA initiate after oocyte activation, even in the absence of fertilization (Mahowald et al. 1983). Third, the polysomes are detectable on polar granules during the cleavage stages (Mahowald 1968, 1971a), when mitochondrial rRNA are stably associated with the granules (Kobayashi et al. 1993; Amikura et al. 1996; 2Fig. 2A). Finally, once pole cells are formed, both the polysomes and mitochondrial rRNA become undetectable on polar granules, while the granules themselves are partitioned into pole cells. Further analysis to test our idea that both mitochondrial rRNA form mitochondrial ribosomes on polar granules will provide a better understanding of the molecular mechanisms of pole cell formation.

Acknowledgements

We thank Drs A. Ephrussi, R. Lehmann, P. F. Lasko and T. Aigaki for providing osk-bcd 3′UTR lines, nosBN, vasPD23 and YPhsSPg fly stocks, respectively. This work was supported in part by a Grant-in-Aid from the Ministry of Education, Science and Culture, Japan, by the Tsukuba Advanced Research Alliance-Project, the Toray Science Foundation, the Sumitomo Foundation, and by a Research Project for Future Program from the Japan Society for the Promotion of Science. M.K. is a Research Fellow of the Japan Society for the Promotion of Science.

Footnotes

  1. Author to whom all correspondence should be addressed.

  2. E-mail: skob@sakura.cc.tsukuba.ac.jp

Ancillary