Gawky is a component of cytoplasmic mRNA processing bodies required for early Drosophila development

Abstract

In mammalian cells, the GW182 protein localizes to cytoplasmic bodies implicated in the regulation of messenger RNA (mRNA) stability, translation, and the RNA interference pathway. Many of these functions have also been assigned to analogous yeast cytoplasmic mRNA processing bodies. We have characterized the single Drosophila melanogaster homologue of the human GW182 protein family, which we have named Gawky (GW). Drosophila GW localizes to punctate, cytoplasmic foci in an RNA-dependent manner. Drosophila GW bodies (GWBs) appear to function analogously to human GWBs, as human GW182 colocalizes with GW when expressed in Drosophila cells. The RNA-induced silencing complex component Argonaute2 and orthologues of LSm4 and Xrn1 (Pacman) associated with 5'-3' mRNA degradation localize to some GWBs. Reducing GW activity by mutation or antibody injection during syncytial embryo development leads to abnormal nuclear divisions, demonstrating an early requirement for GWB-mediated cytoplasmic mRNA regulation. This suggests that gw represents a previously unknown member of a small group of genes that need to be expressed zygotically during early embryo development.

Figure 1.

A comparison of the GW protein family. (A) The product of CG31992, the Drosophila GW protein (GenBank/EMBL/DDBJ accession no. AE003843), contains three regions that are common to all human GW182-related proteins: an N-terminal GW-rich region, a C-terminal RRM domain, and a glutamine-rich region. It has a predicted ubiquitin-associated domain (UBA) that is also found in TNRC6C and a C-terminal serine-rich region that is not found in human GW proteins. Drosophila GW is 17.8–20% identical and 24–28.3% similar to the human GW protein family. It is most similar to TNRC6C. C. elegans AIN-1 is also suggested to be a member of the GW protein family (Ding et al., 2005) because it is GW rich and contains one region of significant (24%) amino acid similarity. (B) Predicted evolutionary relationships between GW proteins from vertebrates and invertebrates. Bar, 0.1 amino acid substitutions per site.

Figure 2.

Characterization of the gw mutation and localization of the GW protein. (A) gw ¹ is caused by a nonsense mutation of the tryptophan codon at position 967 to stop. (B) The gw ¹ mutation causes the loss of an NcoI restriction site and allowed rapid embryo genotyping by PCR. (C) Mutations were confirmed by DNA sequencing. (D) A polyclonal antibody raised against the N-terminal region of GW recognizes a 160-kD band representing the endogenous protein. (E) The anti-GW antibody also recognizes a 100-kD truncated form of GW in gw¹/gw¹ embryos that is not present in wild-type embryos.

Figure 3.

GW localization in normal Drosophila tissues and homozygous gw¹ mutant embryos. (A–C) Embryos were fixed 90–130 min AED. (A) In normal embryos undergoing cellularization (differential interference contrast [DIC]), GW (α-GW) localized to foci surrounding the cortical nuclei (DNA). The plasma membrane is visualized using antiphosphotyrosine (α-P-Tyr). (B) The boxed area in A is shown magnified. Note the presence of brightly staining GW foci in the cytoplasm surrounding the nuclei. (C) In homozygous gw ¹ mutant embryos, the DNA, anti-GW, and antiphosphotyrosine staining form disorganized aggregates. Bars, 100 μM.

Figure 4.

GW protein is expressed at varying levels during development. (A) Western blots showed high levels of GW protein during early embryonic development until ∼18 h and again during pupariation. (B) Relative GW protein levels are reduced at 60–70 min AED and subsequently increase at 70–80 min AED. Error bars represent the SD of the relative values obtained from three separate Western blots. (C) There is an increase in relative gw mRNA levels at ∼80–90 min AED compared with the mRNA encoding the RpL32 ribosomal protein. To confirm the accuracy of quantitation, the same sample was loaded at 0.5, 1.0, and 2.0× concentration.

Figure 5.

Colocalization of GW with markers associated with GWBs/PBs in Drosophila S2 cells. (A) A C-terminal fusion of RFP to PCM localized to discrete cytoplasmic foci. Several of these (arrows) colocalized with a GFP-GW fusion protein. (B) Another human GWB component, LSm4, localized to the nucleus (middle), but some signal was also detected in cytoplasmic foci (arrows). Some, but not all, Drosophila GWBs colocalized with the LSm4 foci. (C) AGO2, a RISC component, also colocalized with some cytoplasmic Drosophila GWBs (arrows). Notably, the cytoplasmic bodies containing GFP-GW and RFP-AGO2 were consistently larger than those containing only GFP-GW. (D–F) Protein fusions between RFP and the three major human GW182 family proteins transfected into Drosophila S2 cells were found in the same structures as Drosophila GW. The expression of human GW182 could not be detected without a coincident RNAi knockdown of endogenous GW. Bars, 5 μM.

Figure 6.

Ultrastructural analysis of Drosophila GWBs and the effect of GW loss on embryos. (A) Thin sections of embryos do not show appreciable immunogold localization when preimmune serum is used. (B and C) Sections stained with α-GW antibodies show appreciable immunogold signal in irregular, electron-dense structures. These are not membrane bound or associated with any other known cytoplasmic structure. Boxed area in B represents a single structure; a representative example is shown at higher magnification in C. (D) Thin sections of wild-type 3-h embryos show characteristic structures (including nuclei) surrounded by a distinct bilayer membrane, which is continuous with the rough endoplasmic reticulum, as well as mitochondria. (E) Homozygous gw ¹ 3-h mutant embryos have few recognizable nuclei and darkly staining membrane-bound vesicles, presumably corresponding to yolk particles in the embryo cortex, from which they are usually excluded at this later stage of development. Large multivesicular bodies (closed arrowhead and large box) are seen and are shown in higher magnification in F. (G) A higher magnification of the aggregates of filamentous structures indicated by the open arrowhead and small box in E. Bars, 0.2 μm.

Figure 7.

Cytoplasmic GWBs require the presence of intact RNA. (A) Drosophila GWBs were detected in S2 cells expressing a GFP-GW protein fusion. (B) 5 min after RNase treatment, punctate GWBs were no longer present, and the GFP-GW signal became diffuse throughout the cytoplasm. (C) In 10% of RNase-treated and 4% of untreated cells, an alternate perinuclear pattern of GFP-GW was seen. Each image represents a maximum projection of a three-dimensional stack of confocal images encompassing the entire cell. (D–F) Endogenous GW (α-GW) and mitochondria (Mitotracker) was also observed to ensure that the RNase treatment did not cause general organelle breakdown. (G) Quantification of the number of cells displaying each of the patterns of GFP-GW (A–C) with or without RNase treatment (n = 320). Error bars represent the SD from three separate experiments. Bars (A–C), 10 μm; (D–F) 20 μM.

Figure 8.

Lack of functional GW protein leads to nuclear breakdown caused by defects in mitosis. (A) In wild-type embryos of the same age, a regular array of nuclei, each with a pair of centrosomes, can been seen immediately below the embryo cortex. (B) In homozygous gw ¹ embryos 90 min AED stained with anticentrosomin (Cnn, red) antibody and antitubulin (blue), severe defects are observed after NC10. Fewer nuclei (PicoGreen) are seen, and the majority of these have improperly localized centrosomes. Large, brightly staining DNA aggregates are also seen (arrowheads). Maximum projection of 125 slices that are 10 μm deep. Bars, 5 μM.

Figure 9.

Loss of GW causes breakdown of the cortical cytoskeleton and nuclear expansion. (A) An embryo at NC10 immediately after injection with anti-GW antibody. Actin surrounds each dividing nucleus, and no obvious differences in this pattern are observed at the site of injection (arrowheads in all panels). (B) At 6 min after injection, the majority of the actin has formed apical “caps” over the interphase nuclei. However, at the anti-GW injection site, actin remains in the honeycomb pattern indicative of mitotic nuclei. (C) This stabilization of actin into the pseudocleavage furrows spreads from the site of injection. (D and E) 10–15 min after injection, the stabilized actin network elongates, and the region of stabilized actin enlarges with time. Areas more distant from the injection site (arrowheads) still form interphase caps for one more NC. (F) By 30 min after injection, the majority of the actin cytoskeleton is in a stabilized pattern, and structures nearest the injection site are beginning to break down. (G) An embryo after NC14 (2 h 10 min AED) expressing an NLS-GFP fusion that localizes to the nuclei injected with guinea pig preimmune serum at 1 h AED. No significant alterations in the morphology or spacing of the nuclei are seen, and posterior pole cells are observed. (H) Injection of anti-GW antibody at 1 h AED produces embryos with ∼200 enlarged nuclei (approximately four times normal size) at 2 h 10 min AED. These nuclei migrate to the cortex but are not anchored there and subsequently move freely within the embryo cytoplasm. (I) Anti-GW injection at 1 h 40 min AED shows a stepwise nuclear enlargement phenotype that is greatest proximal to the injection site at 2 h 10 min AED. Measurements of the nuclear diameter reveal that they are on average two (anterior), four, and eight times larger than those found in wild-type stage 14 embryos. (J) Injection at 1:50 produces a region of nuclei twice the normal size proximal to the injection site at 2 h 10 min AED. (K) Injection of normal rabbit serum into embryos 1 h AED has no effect on nuclear-GFP localization at 2 h 10 min AED. (L) Injection of polyclonal anti-AGO2 antibody into 1-h AED embryos produces a similar phenotype to the injection of anti-GW shown in H.