Germ cell formation in Drosophila relies on polar granules, which are large ribonucleoprotein complexes found at the posterior end of the embryo. The granules undergo characteristic changes in morphology during development, including the assembly of multiple spherical bodies from smaller precursors. Several polar granule components, both protein and RNA, have been identified. One of these, the protein Oskar, acts to initiate granule formation during oogenesis and to recruit other granule components. To investigate whether Oskar has a continuing role in organization of the granules and control of their morphology, we took advantage of species-specific differences in polar granule structure. The polar granules of D. immigrans fuse into a single large oblong aggregate, as opposed to the multiple, distinct, spherical granules of D. melanogasterembryos. D. immigrans oskar rescues the body patterning and pole cell defects of embryos from D. melanogaster oskar- mothers,and converts the morphology of the polar granules to that of D. immigrans. The nuclear bodies, which are structures that appear to be closely related to polar granules, are also converted to the D. immigrans type morphology. We conclude that oskar plays a persistent and central role in the polar granules, not only initiating their formation but also controlling their organization and morphology.

Pole plasm is a specialized cytoplasm found at the posterior pole of Drosophila oocytes. Polar granules form within this cytoplasm by stage 10 of oogenesis and are incorporated into the pole cells (germ cells) by stage 3 of embryogenesis. Classical experiments demonstrated the necessity of the pole plasm for forming germ cells(Geigy, 1931), and the ability of transplanted pole plasm to form ectopic pole cells(Illmensee and Mahowald,1974). Stronger evidence that the pole-cell-forming activity was contained within the polar granules, rather than posterior cytoplasm, was obtained once polar granule components were identified. The isolation of maternal effect mutants defective in embryonic body patterning revealed one group, the grandchildless-knirps genes, that have posterior patterning defects and fail to form pole cells (Boswell and Mahowald, 1985; Lehmann and Nüsslein-Volhard, 1986; Schüpbach and Wieschaus,1986). At least four members of this group, aubergine, oskar,tudor and vasa, encode proteins that are concentrated in polar granules (Bardsley et al.,1993; Hay et al.,1988; Lasko and Ashburner,1990; Breitwieser et al.,1996; Harris and Macdonald,2001). Notably, mislocalization of Osk protein to the anterior,using a transgene containing the osk coding region and bicoid anterior mRNA localization signal, induces the formation of anterior polar granules and pole cells. All of the known polar granule components are recruited to the site of ectopic polar granules, and the pole cells that form are functional germ cells(Ephrussi and Lehmann, 1992; Nakamura et al., 1996; Megosh et al., 2006). The behavior of the mislocalized Osk protein clearly demonstrates that Osk can nucleate polar granule formation. However, it remains uncertain if Osk serves only to recruit other factors, or if it has a more active or persistent role in the structure and organization of polar granules.

Polar granule organization has been revealed by ultrastructural studies(Mahowald, 1962; Mahowald, 1968; Mahowald et al., 1976). The electron-dense granules first appear at the posterior end of D. melanogaster oocytes as an abundance of small particles. These particles persist in the embryo, and as the pole cells mature they fuse to form larger,spherical structures with electron-lucid cores. By gastrulation, each pole cell contains a few dispersed polar granules, which remain until the pole cells migrate through the midgut primordium. In parallel, nuclear bodies, with the same spherical structure, form in nuclei(Mahowald et al., 1976; Harris and Macdonald, 2001). During pole cell migration, the polar granules begin to fragment and associate with the outer nuclear envelope, resembling nuage(Mahowald, 1971). The organization of polar granules differs among Drosophila species(Counce, 1963; Mahowald, 1968). Most notably,in D. immigrans, the polar granules fuse into one large aggregate(Mahowald, 1968; Mahowald et al., 1976), as opposed to the multiple, spherical, individual granules that are characteristic of D. melanogaster.

Here we show that Osk is persistently present in both polar granules and nuclear bodies from their formation to their fragmentation. Expression in D. melanogaster of the osk gene from D. immigranstransforms polar granules and nuclear bodies to the D. immigranstype. Our results show that Osk plays a continuing role in polar granule organization, and may be the primary determinant of the underlying arrangement of both polar granules and nuclear bodies.

Fly stocks

D. immigrans flies were obtained from the Tucson Drosophila Stock Center. P[uas-gfp-aub] and P[osk+] have been described previously(Harris and Macdonald, 2001; Kim-Ha et al., 1995).

Transgenes

D. immigrans osk (GenBank accession number: DQ823083) was isolated from a genomic library prepared from D. immigrans DNA in Lambda Fix II (Stratagene). P[oskimm] contains an 8.7 kb NotI-Asp718 genomic fragment including 5 kb of DNA upstream of the start codon, and 1 kb downstream of the polyadenylation site, all in a pCaSpeR vector (Pirotta, 1988)modified by insertion of NotI and Asp718 linkers at filled-in PstI and EcoRI sites, respectively. In P[oskimm3mel] the oskimm 3′UTR and 3′ flanking sequences are replaced with the equivalent region (a 1.6 kb HindIII-Asp718 fragment) from D. melanogaster osk. The P[M1L-oskimm3mel],P[M103L-oskimm3mel] and P[M103,106L-oskimm3mel]transgenes were all modified from P[oskimm3mel] using the QuikChange II XL Site-Directed Mutagenesis Kit (Stratagene).

Sequence analysis

Oskar protein sequences from different Drosophila species(http://flybase.net/blast/)were compared using ClustalW(http://www.ebi.ac.uk/clustalw/). To form the graph, each amino acid was given a score from 0-8 depending on how many sequences were identical. The average of every 5 amino acids was plotted in Excel (Microsoft).

Immunostaining and imaging

Embryos were collected and stained as described previously(Macdonald and Struhl, 1986; Macdonald et al., 1991) using secondary antibodies coupled to Cy5 (Jackson Immunoresearch Laboratories). Osk antibodies were used at 1:3000, with incubation at room temperature overnight. Stained embryos were mounted in Vectashield medium (Vector Labs) and imaged using a Leica TCS-SP confocal microscope.

To determine if Osk has an ongoing role in polar granule organization beyond the initial recruitment of other factors, we asked if replacing Osk with D. immigrans Osk (Oskimm) would convert the polar granules of D. melanogaster to the D. immigrans type. The D. immigrans osk transgene P[oskimm] was expressed in D. melanogaster flies lacking endogenous Osk protein[osk54/Df(3R)pXT103; referred to below as osk54/Df]. oskimm mRNA was localized to the posterior of the embryo (although not as tightly as osk mRNA;see Fig. S1 in the supplementary material), and rescued the body patterning and pole cell formation defects of osk mutants; fewer pole cells formed, but they were functional. A related transgene, P[oskimm3mel],bearing the 3′UTR and associated regulatory sequences of osk,behaved almost identically (see Fig. S2 in the supplementary material). These phenotypes, and those described below, are dependent only on the maternal genotype; for simplicity, the embryos are described by their mother's genotype.

In pre-cellular blastoderm D. melanogaster embryos, polar granule components, including Osk and the granule marker GFP-Aub, appear at the posterior in a crescent of small sand-like granules, approximately 0.2-0.3μm in diameter. At cellular blastoderm stage, the small granules have begun to coalesce to form spherical particles (0.7-1.4 μm in diameter). These spheres appear in confocal optical sections as donut shapes(Harris and Macdonald, 2001)(see Fig. S3 in the supplementary material).

In early P[oskimm]/+; osk54/Df embryos, the progression of polar granule development was like that of D. melanogaster until the cellular blastoderm stage(Fig. 1I), when the smaller granules began to accumulate in one area of the cell, so that by gastrulation each pole cell contained a single large aggregate of granule material(Fig. 1J), as in D. immigrans embryos. This aggregate appeared in confocal sections as a series of loosely connected particles (Fig. 1L), whose linkage to one another was clearly revealed in a z series of sections (data not shown). This polar granule morphology persisted until the pole cells began to migrate. Expression of P[oskimm3mel] in osk54/Df embryos had essentially the same effect (data not shown). Thus, Oskimm confers the D. immigrans morphology on polar granules formed in a background where all components but Osk are of the D. melanogaster type. The control embryos, in which polar granule formation relies on a single copy of osk, had the typical appearance of D. melanogaster polar granules at all stages(Fig. 1A-E).

Osk appears not only in polar granules in the pole cell cytoplasm, but also in structures identified as nuclear bodies (on the basis of morphology and the presence of Vas), and persists in the nuclear bodies until their decay(Harris and Macdonald, 2001)(see Fig. S3 in the supplementary material). In D. melanogaster, the nuclear bodies first appear at syncitial blastoderm stage. They persist until the pole cells migrate to the gonad when they are no longer detectable. The nuclear bodies of D. immigrans are initially similar to those of D. melanogaster, displaying the characteristic donut appearance in cross-section. However, during the cellular blastoderm stage they develop discontinuities and distortions. Often, portions of multiple spheres aggregate(Mahowald et al., 1976).

To determine if Oskimm would alter D. melanogasternuclear body morphology, the transgenes expressing the protein were tested in a wild-type background (this is necessary because immunodetection of Osk is used to visualize the bodies; GFP-Aub is only cytoplasmic and we have no antibodies to detect Oskimm). In the presence of Oskimm,the nuclear bodies initially appeared with the regular spherical morphology shared by both species up to the cellular blastoderm stage(Fig. 2C). Subsequently, the nuclear bodies developed fissures, as seen in the donut cross-sections, and the spheres became irregular and flattened(Fig. 2F). Thus, the nuclear bodies had been transformed towards the D. immigrans morphology. It is not surprising that the transformation was not complete, as endogenous Osk was also present. We conclude that Oskimm is a determinant for the morphology of polar granules and nuclear bodies.

Osk protein is made in two forms, Long and Short, initiated from different methionine codons (Markussen et al.,1995; Vanzo and Ephrussi,2002). To determine which Oskimm isoform is responsible for polar granule morpology, three transgenes based on P[oskimm3mel], but with mutations to direct synthesis of the individual Oskimmisoforms, were tested in wild-type flies.

The P[M1L-oskimm3mel]transgene (with the first methionine codon mutated to leucine, and thus unable to make Long Oskimm) transformed granules to the D. immigrans type (Fig. 3B). Thus, the Short isoform of Oskimm is sufficient to specify polar granule morphology. The P[M103L-oskimm3mel]and P[M103,106L-oskimm3mel]transgenes differed in their activities. When only the first of the two methionines near the predicted start of Short Oskimm was mutated(M103L), the transgene could still transform the polar granules to the D. immigrans type (Fig. 3C). However, when both methionine codons that could initiate Short Oskimm were mutated (M103L and M106L), the ability to transform polar granule morphology was lost (Fig. 3D). These results strongly suggest that either methionine can initiate translation of Short Oskimm, and that the Long Oskimm isoform does not contribute to specification of polar granule morphology.

Alignment of Osk and Oskimm, as well as Osk from other species,revealed regions of higher and lower similarity (see Fig. S4 in the supplementary material). The variation between Osk and the predicted Oskimm protein sequences is similar to that seen in comparison of other species of similar evolutionary divergence. Thus, the feature responsible for the striking difference in the polar granules of D. melanogaster and D. immigrans cannot be readily predicted.

The Osk, Vas, Aub and Tud proteins are present in polar granules and are required for their formation. Only for Tud has a specific function in polar granule assembly been suggested. Several mutants in Tud result in smaller and fewer granules, raising the possibility of a role in the fusion of granule material (Boswell and Mahowald,1985; Thomson and Lasko,2004). Recently, a detailed dissection of Tud revealed a crucial role for Tud domain 1 (tudA36 mutant) in germ cell formation and polar granule formation, and led to the proposal that the Tud domains serve as docking platforms for polar granule assembly. Notably,mutation of a single Tud domain altered polar granule morphology from the small, sand-like spherical granules seen in wild-type embryos at egg-lay, to string-like granules (Arkov et al.,2006). Thus, Osk, with its central role in specifying granule morphology, may perform this role via interactions with Tud. We compared, by yeast two-hybrid assays, the interactions of the short isoforms of Osk and Oskimm proteins with Tud and with the other known D. melanogaster polar granule components(Breitwieser et al., 1996), but found no substantial differences in the level of interactions displayed by the Osk and Oskimm proteins (data not shown).

Evidence of a central role for Osk in organizing polar granules and nuclear bodies focuses attention on the question of how it performs these functions,and on what features of Osk determine which type of granule will form. The answer to the first question must involve the interactions of Osk with other granule components, but how molecular interactions of proteins whose volume is measured in Angstroms can direct formation of seemingly perfect spheres that are hundreds of nanometers in diameter is uncertain. The extreme regularity of the structures invites comparisons to other highly regular spherical structures, such as viral capsids. The principles underlying the assembly of icosahedral virions could also apply to spherical polar granules, with an important difference: virions are assembled from a well-defined number of subunits, usually organized as a series of hexamers and pentamers(Morgan, 2003). A similar assembly on the scale of polar granules would require subunits several orders of magnitude larger than the viral capsid proteins. Thus the subunits would have to consist of a very large number of individual polypeptides; these subunits could possibly correspond to the large number of small, sand-like polar granules seen very early in embryogenesis. Extending the analogy to viral capsid assembly to address the different morphologies of polar granules,the spherical polar granules could be comparable to wild-type viral capsids,with the D. immigrans-type granules comparable to polymorphic aggregates of the coat proteins. Whereas only the spherical assemblies could function to encase a viral genome, either type could contain integral germ cell determinants. Our results show that either type of polar granule can mediate germ cell formation in D. melanogaster.

We thank Anne Ephrussi for providing the Osk antibody; the Tucson Drosophila Stock Center for providing Drosophila immigrans flies; Nan Yan for injecting P[osk+]; and Mark Snee,Cuiyun Geng and Daniel Jones for comments on the manuscript. This work was supported by a grant (GM54409) from the National Institutes of Health.

Arkov, A. L., Wang, J. Y., Ramos, A. and Lehmann, R.(
2006
). The role of Tudor domains in germline development and polar granule architecture.
Development
133
,
4053
-4062.
Bardsley, A., McDonald, K. and Boswell, R. E.(
1993
). Distribution of tudor protein in the Drosophilaembryo suggests separation of functions based on site of localization.
Development
119
,
207
-219.
Boswell, R. E. and Mahowald, A. P. (
1985
). tudor, a gene required for assembly of the germ plasm in Drosophila melanogaster.
Cell
43
,
97
-104.
Breitwieser, W., Markussen, F.-H., Horstmann, H. and Ephrussi,A. (
1996
). Oskar protein interaction with vasa represents an essential step in polar granule assembly.
Genes Dev
.
10
,
2179
-2188.
Counce, S. J. (
1963
). Developmental morphology of polar granules in Drosophila including observations on pole cell behavior and distribution during embryogenesis.
J. Morphol.
112
,
129
-145.
Ephrussi, A. and Lehmann, R. (
1992
). Induction of germ cell formation by oskar.
Nature
358
,
387
-392.
Geigy, R. (
1931
). Action de l'ultra-violet sur le pole germinale dans l'oeuf de Drosophila melanogaster.
Rev. Suisse Zool.
38
,
187
-288.
Harris, A. N. and Macdonald, P. M. (
2001
). Aubergine encodes a Drosophila polar granule component required for pole cell formation and related to eIF2C.
Development
128
,
2823
-2832.
Hay, B., Jan, L. Y. and Jan, Y. N. (
1988
). A protein component of Drosophila polar granules is encoded by vasa and has extensive sequence similarity to ATP-dependent helicases.
Cell
55
,
577
-587.
Illmensee, K. and 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.
Kim-Ha, J., Smith, J. L. and Macdonald, P. M.(
1991
). oskar mRNA is localized to the posterior pole of the Drosophila ooctye.
Cell
66
,
23
-35.
Kim-Ha, J., Kerr, K. and Macdonald, P. M.(
1995
). Translational regulation of oskar mRNA by bruno,an ovarian RNA-binding protein, is essential.
Cell
81
,
403
-412.
Lasko, P. F. and Ashburner, M. (
1990
). Posterior localization of vasa protein correlates with, but is not sufficient for, pole cell development.
Genes Dev
.
4
,
905
-921.
Lehmann, R. and Nüsslein-Volhard, C.(
1986
). Abdominal segmentation, pole cell formation, and embryonic polarity require the localized activity of oskar, a maternal gene in Drosophila.
Cell
47
,
141
-152.
Macdonald, P. M. and Struhl, G. (
1986
). A molecular gradient in early Drosophila embryos and its role in specifying the body pattern.
Nature
324
,
537
-545.
Macdonald, P. M., Luk, S. K.-S. and Kilpatrick, M.(
1991
). Protein encoded by the exuperantia gene is concentrated at sites of bicoid mRNA accumulation in Drosophila nurse cells but not in oocytes or embryos.
Genes Dev
.
5
,
2455
-2466.
Mahowald, A. P. (
1962
). Fine structure of pole cells and polar granules in Drosophila melanogaster.
J. Exp. Zool.
151
,
201
-216.
Mahowald, A. P. (
1968
). Polar granules of Drosophila. II. Ultrastructural changes during early embryogenesis.
J. Exp. Zool.
167
,
237
-262.
Mahowald, A. P. (
1971
). Polar granules of Drosophila. III. The continuity of polar granules during the life cycle of Drosophila.
J. Exp. Zool.
176
,
329
-344.
Mahowald, A. P., Illmensee, K. and Turner, F. R.(
1976
). Interspecific transplantation of polar plasm between Drosophila embryos.
J. Cell Biol.
70
,
358
-373.
Markussen, F.-H., Michon, A.-M., Breitwieser, W. and Ephrussi,A. (
1995
). Translational control of oskar generates Short OSK, the isoform that induces pole plasm assembly.
Development
121
,
3723
-3732.
Megosh, H. B., Cox, D. N., Campbell, C. and Lin, H.(
2006
). The role of PIWI and the miRNA machinery in Drosophila germline determination.
Curr. Biol
.
16
,
1884
-1894.
Morgan, G. J. (
2003
). Historical review:viruses, crystals and geodesic domes.
Trends Biochem. Sci.
28
,
86
-90.
Nakamura, A., Amikura, R., Mukai, M., Kobayashi, S. and Lasko,P. F. (
1996
). Requirement for a noncoding RNA in Drosophila polar granules for germ cell establishment.
Science
274
,
2075
-2079.
Pirotta, V. (
1988
). Vectors for P-mediated transformation in Drosophila. In
Vectors: A Survey of Molecular Cloning Vectors and their Uses
(ed. R. L. Rodriguez and D. T. Denhardt), pp.
437
-456. Boston:Butterworths.
Schüpbach, T. and Wieschaus, E. (
1986
). Maternal-effect mutations altering the anterior-posterior pattern of the Drosophila embryo.
Rouxs Arch. Dev. Biol
.
195
,
302
-317.
Thomson, T. and Lasko, P. (
2004
). Drosophila tudor is essential for polar granule assembly and pole cell specification, but not for posterior patterning.
Genesis
40
,
164
-170.
Vanzo, N. F. and Ephrussi, A. (
2002
). Oskar anchoring restricts pole plasm formation to the posterior of the Drosophila oocyte.
Development
129
,
3705
-3714.

Supplementary information