Temporal and Spatial Expression of RACK1
Our initial efforts were focused on defining the expression pattern of RACK1. To this end, we generated an anti-RACK1 polyclonal antibody that recognizes a polypeptide with apparent molecular mass of 31 kDa in Western immunoblots of wild-type fly lysates (Fig. 2). This size corresponds well with the molecular weight of RACK1 predicted from translation of the coding sequence.
Figure 2. Developmental Western blot of Receptor for Activated CKinase 1 (RACK1) expression. A: Samples of w1118 animals at the indicated developmental stages were collected and homogenized in 50 mM Tris pH 7.9, 150 mM NaCl, 0.1% Triton X-100. The protein content measured by Bradford assay (Bio-Rad) and equal protein (8 μg) was loaded in each lane. B: Whole males, females, females with their ovaries removed, and dissected ovaries were homogenized in Laemmli sample buffer and the equivalent of half of an animal or a single ovary was loaded in each lane.
Download figure to PowerPoint
A developmental time course shows that, in wild-type animals, RACK1 is expressed throughout embryogenesis and early larval stages with expression tapering off in third-instar larvae, pupae, and adults (Fig. 2a). RACK1 expression is much higher in adult females than males. To explore whether RACK1 expression in the ovary accounts for this sex-related difference, we determined the level of RACK1 in males, intact females, oophorectomized females, and isolated ovaries (Fig. 2b). These data indicate that RACK1 is highly enriched in the adult ovary. Indeed, our analysis reveals that the ovary represents the major site of RACK1 expression in adult females.
Immunofluorescence microscopy in wild-type embryos shows that RACK1 is ubiquitously expressed during embryogenesis (Fig. 3). The protein is cytosolic and is excluded from cell nuclei. It is present in all cells at the cellular blastoderm stage (Fig. 3A) and, likewise, is present throughout the embryo during the morphogenetic processes of germ band retraction (Fig. 3B) and dorsal closure (Fig. 3C). Presuming that Drosophila RACK1 functions in the same processes that have been described in mammalian cell culture—including cell migration, cell survival and apoptosis, and protein synthesis—widespread embryonic expression of RACK1 is not surprising.
Figure 3. Receptor for Activated CKinase 1 (RACK1) is ubiquitously expressed in the developing embryo. A–C: Confocal images of w1118 embryos at the cellular blastoderm stage (A), during germ band retraction (B), and during dorsal closure (C), using indirect immunofluoresence to visualize RACK1 expression pattern. D: Preimmune serum is shown as a control for specificity of the RACK1 antiserum. Boxed areas are shown in the enlargements, and more clearly demonstrate the nuclear exclusion of the RACK1 protein. Embryos are oriented anterior to the left and dorsal at the top. pc, pole cells; as, amnioserosa; gb, germ band; ep, epidermis; ps, posterior spiracles. Scale bar = 100 μm.
Download figure to PowerPoint
Considering the high levels of RACK1 expression in the ovary, we examined RACK1 localization in wild-type ovaries (Fig. 4). In the fly, each of the two ovaries consists of 15–17 ovarioles, tubular structures in which egg chambers develop (Fig. 4A). Within an ovariole, egg chambers show a continuum of growth. In the anterior germarium, egg chambers are initially formed from germ-line stem cells (GSCs) that reside there. A GSC daughter, the cystoblast, undergoes a series of four mitotic divisions (Fig. 4B) with incomplete cytokinesis to give rise to 16 large, germ-line–derived cells connected by cytoplasmic bridges. These connections between the germ cells are maintained by specialized structures called ring canals. One of the germ cells differentiates into the oocyte, and the remaining 15 cells become nurse cells (Fig. 4B). As it forms in the germaria, each egg chamber is enclosed in a somatically derived follicular epithelium. The egg chambers continue to develop as they travel posteriorly through the ovariole. In the late stages of oogenesis, the nurse cells transfer their cytoplasmic contents to the oocyte using the cytoplasmic bridges, and the oocyte increases in size to fill the entire egg chamber. Eventually, the mature egg passes into the oviduct to be fertilized and laid (Verheyen and Cooley,1994).
Figure 4. Receptor for Activated CKinase 1 (RACK1) is expressed in the ovary. A: Schematic diagram of a pair of Drosophila ovaries. A single ovariole, containing multiple egg chambers is shown in the enlargement. Egg chambers are formed in the germarium and develop as they move posteriorly through the ovariole. B: The 16 germ cells of each egg chamber are diagrammed. They arise from a series of four synchronous divisions (labeled 1–4) of the cystoblast (*). C–E: Confocal images of adult w1118 ovarioles using indirect immunofluoresence to visualize the RACK1 expression pattern. C: RACK1 is expressed throughout the ovariole in both germ-line and somatic tissue. There is a punctate distribution of RACK1 visible between the large nurse cells (arrows). The boxed area is shown in the enlargement. D: A higher magnification image of the germarium clearly shows the cytosolic localization and exclusion from the nuclei. E: Phosphotyrosine is a marker of cell–cell junctions and demonstrates that some RACK1 is colocalized with areas of tyrosine kinase signaling (arrows). g, germarium; fc, somatic follicle cells; gc, germ cells. Scale bars = 100 μm in C, 20 μm in D,E.
Download figure to PowerPoint
RACK1 is expressed in both the germ tissue and somatic tissue of each egg chamber (Fig. 4C–E). As observed in embryos, it is excluded from the nuclei and is found to have a diffuse cytoplasmic localization both in late stage egg chambers (Fig. 4C) as well as within the germarium (Fig. 4D). We also detected an enrichment of RACK1 in a punctate pattern at the cell membranes between nurse cells (inset in Fig. 4C). Colabeling phosphotyrosine (Fig. 4E) demonstrates that a pool of RACK1 is targeted to cell–cell contacts, where it colocalizes with enriched tyrosine kinase activity.
Lethal Phase of RACK1 Zygotic Mutants
All four RACK1 alleles are recessive, with no apparent heterozygous phenotypes. Homozygotes and transheterozygotes exhibit lethality at several developmental time points, with variability dependent upon the density of animals in the vials that are scored. RACK1EE, RACK1EY128, and Df(2)RACK1-mts homozygotes show complete larval lethality. Because RACK11.8 homozygotes and all transheterozygotes can survive past this time point, we speculate there may be additional mutations outside the RACK1 locus in the RACK1EE and RACK1EY128 alleles that have persisted despite repeated out-crossing. Among transheterozygotes and RACK11.8 homozygotes, as many as 75% of animals display larval lethality. Of those animals that do pupariate, up to 85% die before adult eclosion. For all transheterozygotes, as well as RACK11.8 homozygotes, a variable portion (3–35%) of the total progeny eclose as sterile adult escapers. Escapers display a range of phenotypes, including sluggish movement and difficulty exiting the pupal case, cuticle that does not harden, and wings that are blistered, abnormally small, or fail to unfold. RACK1EE/RACK1EY128 animals, which express mutant RACK1 protein at levels higher than the other transheterozygotes (Fig. 5B), produce the largest numbers of escapers (up to 35%), with the least severe phenotypes. These animals are sterile, but otherwise appear normal.
Lethal Phase of RACK1 Maternal Mutants
High RACK1 expression in ovaries, coupled with a published report of high levels of maternally inherited RACK1 transcript (Vani et al.,1997), suggest that RACK1 may have an essential role early in development, with maternal load allowing zygotic mutants to survive to later time points. Indeed, we have observed the persistence of maternally contributed RACK1 protein through embryogenesis. Western analysis shows that RACK1 mutant embryos, isolated at 18–22 hr by selecting against a balancer with an early embryonic twist:GFP reporter, have RACK1 levels that are not significantly reduced when compared with wild-type embryos (data not shown).
To test for an earlier requirement for RACK1, maternally contributed RACK1 was eliminated by the generation of RACK11.8 germ-line clones (Chou and Perrimon,1996). The germ tissue in these females is devoid of RACK1, and the surrounding somatic tissue is either heterozygous or mosaic. When mated to wild-type males, these females display a dramatically reduced rate of egg lay (data not shown) and the eggs that are produced are reduced in size by an average of 30% (Fig. 6A,B). The maternally deficient RACK11.8 eggs are competent to be fertilized (data not shown), but the fertilized embryos arrest very early in embryogenesis, with little discernable development upon gross examination. This finding demonstrates an early embryonic requirement for maternally contributed, germ-line RACK1, for which zygotic RACK1 cannot compensate. By comparing the RACK1 germ-line mutant to the RACK1 escaper females, we can also conclude that RACK1 has distinct functions in somatic cells. In the escapers, maternally contributed RACK1 protein is depleted throughout all the tissues of the animal, and as a result, escapers are incapable of laying eggs. The germ-line null mutant contains RACK1 in its somatic tissues, and we observe that somatic RACK1 allows oviposition to occur.
Figure 6. Embryos from RACK1 germ-line clone females are reduced in size. A: Comparison of the relative size of wild-type w1118 and RACK11.8 germ-line embryos (P = 0.004). B: Differential interference contrast microscopy images of representative embryos. RACK1, Receptor for Activated CKinase 1. Scale bar = 100 μm.
Download figure to PowerPoint
Ovary Phenotypes of RACK1 Mutants
The sterility of the RACK1 escaper females indicates a requirement for RACK1 in oogenesis and/or egg laying. The size of the ovary in the escaper and the germ-line clone females is reduced (Fig. 7), but overall body size is unaffected in the RACK1 mutants (data not shown). This finding suggests growth and metabolism in the somatic tissues are not impacted by these RACK1 mutations. Despite the small size of the RACK1 ovaries, all genotypes of RACK1 females examined contain properly specified reproductive tracts containing ovarioles and egg chambers. Furthermore, the muscle-containing sheaths surrounding the ovarioles contract rhythmically.
Figure 7. The ovaries of RACK1 females are reduced in size. Freshly dissected ovaries from 4-day-old adult females. A: w1118. B: RACK11.8/Df(2)RACK1-mts escaper. C: RACK1EE/RACK1EY128 escaper. D: RACK11.8 germ-line clone. RACK1, Receptor for Activated CKinase 1. Scale bar = 500 μm.
Download figure to PowerPoint
To explore the function of RACK1 in the ovary, we examined ovarioles from RACK11.8/Df(2)RACK1-mts escapers. Such escapers lack zygotic RACK1, and by eclosion, maternal RACK1 is depleted (Fig. 5B). These escapers are rare and feeble, and most do not survive the 3-day aging period necessary to allow facile dissection and isolation of ovaries. We tested for the presence of GSCs in the RACK11.8/Df(2)RACK1-mts mutants by immunolabeling the Sex-lethal protein (Fig. 8B). Within the germarium, Sex-lethal is present specifically in the cytoplasm of GSCs. It is also present in the GSC daughter cell, known as the cystoblast, which will differentiate into the egg. Cytoplasmic Sex-lethal expression persists through the first mitotic division of the cystoblast, but is absent in the cytoplasm at later developmental stages (Fig. 8A; Bopp et al.,1993). While some of the mutant germaria contain numbers of Sxl-positive cells within the normal range (three to eight cells), others are completely devoid of GSCs (Fig. 8B). Because we observe that RACK1 null ovarioles containing no GSCs do contain more mature egg chambers with germ-line cells (Fig. 8B), we conclude that the maintenance of GSCs is compromised in the RACK1 null ovaries.
Figure 8. Receptor for Activated CKinase 1 (RACK1) null ovarioles have severe germ-line defects. The 4′,6-diamidine-2-phenylidole-dihydrochloride (DAPI) staining (magenta) is included for orientation. A: Representative confocal image of Sex-lethal expression pattern (green) in a w1118 germarium. The germarium shown contains 6 Sxl positive cells (arrow). B: Example of a RACK11.8/Df(2)RACK1-mts ovariole showing a complete absence of germ-line stem cells (GSCs; arrow). C: Confocal image of hu li tai shao expression and DAPI staining in a RACK11.8/Df(2)RACK1-mts ovariole. In the ovariole shown, we observe seven ring canals and eight nuclei in one of the egg chambers (light blue arrow). Another egg chamber (yellow arrow), contains 16 nuclei, but only 13 ring canals. rc, ring canal; n, nurse cell nucleus; o, oocyte; fc, follicle cells. Scale bars = 20 μm in A,B, 50 μm in C.
Download figure to PowerPoint
Wild-type egg chambers invariantly contain 16 germ-line cells—15 nurse cells plus 1 oocyte, derived from the four mitotic divisions of the cystoblast (see Fig. 4B). These germ cells can be readily counted after 4′,6-diamidine-2-phenylidole-dihydrochloride (DAPI) labeling of the nuclei. Additionally, labeling with an antibody to the hu li tai shao (hts) protein specifically labels the 15 ring canals that rim the cytoplasmic bridges between the 16 germ cells (Robinson et al.,1994). RACK11.8/Df(2)RACK1-mts egg chambers sometimes contain fewer germ cells per egg chamber (Fig. 8C). A large portion of the aberrant egg chambers contains either four (not shown) or eight (Fig. 8C, light blue arrow) nuclei. One explanation for this observation is that only two or three of the cystoblast's four mitotic divisions were successfully completed. We also observed that the number of ring canals sometimes does not correspond to the number of nuclei present. For instance, Figure 8C (yellow arrow) shows an egg chamber with 16 nuclei, but only 13 ring canals. In the RACK11.8/Df(2)RACK1-mts ovarioles that we examined, many egg chambers contained wild-type numbers of germ cells, but approximately one egg chamber per ovariole was found to contain either aberrantly low numbers of germ cells or ring canals. These data suggest that RACK1 may function in cell division. This suggestion is consistent with data from a large-scale RNAi screen in Caenorhabditis elegans, which identified RACK1 among the components of the cytokinesis machinery (Skop et al.,2004).
Because of the rarity of RACK1 null escapers, a hypomorphic allele was used for examination of larger numbers of animals (Fig. 9). RACK1EE/RACK1EY128 escapers are frequent and have milder phenotypes than other allelic combinations. Therefore, large numbers of hypomorphs can easily be collected, and these animals survive the 3- to 7-day aging period necessary for facile ovary dissection. At 3 days posteclosion, ovaries from RACK1EE/RACK1EY128 females are reduced in size compared with wild-type (Fig. 7A vs. C), but these ovarioles have no gross morphological defects. At 7 days posteclosion, the mutant ovarioles are crowded with small but mature, stage 14 eggs (Fig. 9B). This finding is in contrast to wild-type ovaries of the same age, which contain a normal continuum of eggs at all 14 developmental stages (Fig. 9A). These data suggest that RACK1 may belong to a class of female-sterile mutants that retain apparently normal stage 14 egg chambers within their ovaries (Schupbach and Wieschaus,1991). Very little is known about mutations of this type, but it is speculated that the genes involved may control egg maturation that takes place in the oviducts and/or uterus. Alternatively, defects in oviposition could lead to the accumulation of mature eggs observed in hypomorphic RACK1 ovaries. As mentioned earlier, the muscular sheaths surrounding the mutant ovarioles do contract, but subtle defects in muscle function or innervation can impact oviposition (Middleton et al.,2006). The sluggish movement of RACK1 escaper adults may be consistent with muscle defects that compromise egg-laying ability. Furthermore, as discussed earlier, ability of the RACK1 germ-line clone females, in which the soma is not mutated, to lay eggs suggests that it is somatically derived RACK1 that functions in oviposition.
Figure 9. RACK1EE/RACK1EY128 hypomorphic escapers have milder ovary phenotypes. A: Seven-day-old wild-type ovarioles contain eggs chambers distributed normally across all developmental stages. B: The RACK1 hypomorph shows an accumulation of mature, stage 14 eggs (*), with crowding of the more anterior, less mature egg chambers. C: Histogram showing the distribution in the number of Sxl-positive cells in the RACK1EE/RACK1EY128 mutant versus wild-type w1118 germaria. Unpaired Student's t test indicates P < 0.0001. g, germarium; nc, nurse cells; o, oocyte; fc, follicle cells. Scale bar = 200 μm in B.
Download figure to PowerPoint
In the hypomorphic RACK1 escapers, we also analyzed the number of GSCs within the germaria. By counting the number of cells with positive, cytoplasmic Sex-lethal staining in large numbers of individual germaria (Fig. 9C), we show that RACK1EE/RACK1EY128 mutants have significantly reduced, but not absent, GSC populations (mean = 5.5 in wild-type vs. 3.9 in the mutant, P < 0.0001). We also examined the number of germ-line nuclei in each egg chamber. However, in contrast to the null mutant, there is no evidence for aberrant numbers of germ-line derived cells in hypomorphic egg chambers (data not shown), suggesting that residual RACK1 function in the hypomorph is sufficient for completion of the four mitotic divisions of the cystoblast.
In summary, RACK1 is essential at multiple, specific points in Drosophila development. Analysis of germ-line clones reveals that maternally contributed (germ-line) RACK1 is required for early embryonic development, and zygotic RACK1 cannot compensate for its absence. Maternally loaded RACK1 enables zygotic RACK1 mutants to survive to later stages, presumably until RACK1 protein is depleted and a RACK1-dependent larval or pupal event fails. Partial lethality at multiple developmental stages may arise because of variability in the rate at which maternal RACK1 is turned over, with enough RACK1 persisting to allow for pupariation, metamorphosis, and subsequent adult eclosion in only a subset of the mutant animals. The most frequent escapers contain the RACK1EE allele, a hypomorphic point mutant that allows for a low level of mutant RACK1 expression. This finding is in contrast to RACK11.8, RACK1EY128, and Df(2)RACK1-mts, which produce no detectible RACK1 protein and, hence, lower numbers of adult escapers.
Through a series of experiments in which we analyzed sterile female escapers containing either the null or hypomorphic RACK1 alleles, we have uncovered essential roles for RACK1 at several points in oogenesis. RACK11.8/Df(2)RACK1-mts null escaper females have severe ovary phenotypes, including depletion of germ-line stem cell populations, implicating RACK1 as a participant in GSC maintenance. Moreover, there are abnormal numbers of germ-line–derived cells and/or ring canals in a subset of the RACK1 null egg chambers. This finding suggests that RACK1 may play a role in the production, partitioning, or maintenance of the 16 germ cells in each egg chamber. This finding is consistent with data from a large RNAi screen in C. elegans that included RACK1 and demonstrated a requirement for RACK1 in cytokinesis and fertility (Skop et al.,2004).
Intriguingly, despite the importance of RACK1 for the germ-line, the RACK11.8 germ-line clones are competent to produce mature eggs that can be fertilized and subsequently arrest early in embryogenesis. Analysis of germ-line clone ovarioles did not reveal deficits in germ-line stem cell numbers or in the number of germ cells within each egg chamber (data not shown). This finding demonstrates that germ-line derived RACK1 is not required for oogenesis, but rather, somatic RACK1 is required for proper germ-line function in a non–cell-autonomous manner.
Because the hypomorphic escaper has milder phenotypes, we are able to analyze ovaries at later time points. In the hypomorph, we observe that oogenesis can proceed relatively normally to stage 14, but these eggs are not laid and accumulate in the ovariole. This finding demonstrates an essential role for RACK1 in egg maturation or oviposition. Again, the germ-line clone data confirm a requirement for somatically derived RACK1 in this process.
We have shown that RACK1 mutants have female germ-line defects. In mammalian systems, a physical interaction between protein kinase C and RACK1 has been well characterized (McCahill et al.,2002). Interestingly, Drosophila aPKC (DaPKC) mutations in the female germ-line lead to defects in germ cell anterior–posterior polarity and maintenance of oocyte cell fate (Cox et al.,2001). If DaPKC physically interacts with RACK1 in the fly, the RACK1 mutants we have characterized here can be valuable tools for genetically determining whether DaPKC and RACK1 function together in oogenesis. Likewise, using the Drosophila counterparts for a host of reported RACK1 binding partners, further genetic interactions with RACK1 can be uncovered. In this way, a more complete picture of the in vivo function of RACK1 will emerge.