A stem-loop structure directs oskar mRNA to microtubule minus ends

Abstract

mRNA transport coupled with translational control underlies the intracellular localization of many proteins in eukaryotic cells. This is exemplified in Drosophila, where oskar mRNA transport and translation at the posterior pole of the oocyte direct posterior patterning of the embryo. oskar localization is a multistep process. Within the oocyte, a spliced oskar localization element (SOLE) targets oskar mRNA for plus end-directed transport by kinesin-1 to the posterior pole. However, the signals mediating the initial minus end-directed, dynein-dependent transport of the mRNA from nurse cells into the oocyte have remained unknown. Here, we show that a 67-nt stem-loop in the oskar 3' UTR promotes oskar mRNA delivery to the developing oocyte and that it shares functional features with the fs(1)K10 oocyte localization signal. Thus, two independent cis-acting signals, the oocyte entry signal (OES) and the SOLE, mediate sequential dynein- and kinesin-dependent phases of oskar mRNA transport during oogenesis. The OES also promotes apical localization of injected RNAs in blastoderm stage embryos, another dynein-mediated process. Similarly, when ectopically expressed in polarized cells of the follicular epithelium or salivary glands, reporter RNAs bearing the oskar OES are apically enriched, demonstrating that this element promotes mRNA localization independently of cell type. Our work sheds new light on how oskar mRNA is trafficked during oogenesis and the RNA features that mediate minus end-directed transport.

Keywords: Drosophila; mRNA localization; oocyte; oogenesis; osk; oskar.

FIGURE 1.

The central portion of oskar 3′ UTR is necessary and sufficient for oocyte transport. (A) Schematic representation of the reporter constructs analyzed in this panel and their ability to localize to the oocyte portion of stage 5–6 egg-chambers. Red rectangle: region 2b encompassing the oocyte entry signal (OES), gray rectangle: TLS of fs(1)K10. Numbers indicate nucleotide positions within the oskar 3′ UTR. (B–K) Stage 5–6 egg-chambers from oskar RNA null flies (osk^A87/Df(3R)p^XT103) expressing transgenic oskar 3′ UTR variants fused to the egfp open reading frame and under the control of a yeast UAS. Reporter mRNAs were expressed by pCog-Gal4 and nos-Gal4 drivers and detected by fluorescent in situ hybridization with an egfp-antisense probe. The asterisk indicates the position of the oocyte nucleus. Bar, 30 μm. The effects on localization shown were fully penetrant in all scored egg-chambers (n ≥ 20). Reporter mRNAs were fused to: the complete oskar 3′ UTR (B); region 1 + 2 (C); region 2 + 3 (D); region 1 (E); region 3 (F); region 2 (G); region 2b + 3 of the oskar 3′ UTR (H); the complete oskar 3′ UTR bearing a deletion of region 2b (I); region 2b of the oskar 3′ UTR (J); region 3 fused at its 5′ end to the TLS of fs(1)K10 mRNA (K). (L,M) Experimentally validated secondary structures of the 44-nt-long TLS of fs(1)K10 mRNA (L) (Bullock et al. 2010) and of the 67-nt-long OES of oskar mRNA (M) (Jambor et al. 2011).

FIGURE 2.

Region 2b forms a stem–loop important for oocyte entry of oskar mRNA. (A) Schematic of the secondary structure of the reporter mRNAs analyzed in this panel and their localization capacity during early oogenesis. Red indicates stretches with point mutations. (B) Alignment of the OES region in 12 Drosophila species with sequenced genomes. Highlighted in blue are nucleotides with >90% similarity. The experimentally validated secondary structure of the D. melanogaster OES is depicted below: A parenthesis indicates a base-paired nucleotide; an asterisk indicates a single-stranded nucleotide. Note that the terminal stem is almost entirely conserved among all species. (C) Primary sequence of the wild-type oocyte entry signal and the mutations analyzed in this panel. (D–K) Stage 5–6 egg-chambers from wild-type flies expressing transgenic oskar 3′ UTR-region 2 reporter mRNAs fused to the EGFP open reading frame under the control of mat-α4-tub-Gal4. Reporter RNAs are detected by fluorescent in situ hybridization using an egfp-antisense probe. A star indicates the position of the oocyte nucleus. Bar, 30 μm. Reporter mRNAs were fused to: region 2 of the oskar 3′ UTR (D); region 2 with deletion of proximal stem of region 2b: Δ2b-p (E); region 2 with a deleted distal stem of region 2b: Δ2b-d (F); region 2 with a deletion of the terminal loop (G); region 2 with a mutated 5′ stem (H); region 2 with a mutated 3′ stem (I); region 2 with mutations in both the 5′ and 3′ stem such that complementarity is restored (J); region 2 in which the quality of the terminal stem is changed from AU-rich to GC-rich (K).

FIGURE 3.

Region 2b directs apical localization of transcripts in blastoderm embryos. (A–H) RNA transcripts were injected into blastoderm embyos and fixed 8–10 min later (i.e., 8 min after injection of the last embryo). Images are oriented with apical to the top and basal to the bottom. Bar, 30 μm. The arrows in A and B indicate the apical cytoplasm of the blastoderm embryo. Schematic representations of the reporter RNAs are shown above the images. Thin black line: UTR RNAs, thick black line: ORF, dashed line: deletions, stem–loop: oocyte entry signal (OES) region, green rectangle: spliced oskar localization element (SOLE). (A′–H′) Categorization of RNA enrichment in the apical cytoplasm. Percentage of embryos showing different efficiencies of transport is shown: (++) strong, (+) weak, (+/−) very weak, (−) no enrichment. (n) Number of embryos scored for each RNA. The following reporter mRNAs were injected: fs(1)K10 RNA, containing the entire 3′ UTR and flanking 3′ sequences (A–A′); a mutated version of the fs(1)K10 RNA, containing a randomized TLS sequence (Bullock et al. 2010) (B–B′); full-length oskar mRNA (C–C′); region 2 + 3 of oskar 3′ UTR (D–D′); region 2 + 3 of oskar 3′ UTR bearing a deletion of the region 2b stem (E–E′); region 2 + 3 of oskar 3′ UTR with a deletion of the distal stem 2b (F–F′); region 2 + 3 of oskar 3′ UTR with a mutated 3′ portion of the terminal stem (G–G′); region 3 of oskar 3′ UTR (H–H′).

FIGURE 4.

The oocyte entry signal mediates anterior localization in stage 8 oocytes and apical localization in polarized epithelia. (A–O) Transgenic UAS-oskar 3′ UTR reporters fused to the egfp open reading frame were expressed under the control of the mat-α4-tub-Gal4 in the germline (A–F), GR1-Gal4 in follicular epithelial cells (G–L), or forkhead-Gal4 in salivary glands (M–O). Reporter RNAs were detected by fluorescent in situ hybridization using an egfp-antisense probe (RNA shown in green in M–O). For orientation, the salivary glands (M–O) were counter-stained with DAPI to reveal DNA (shown in magenta). A star indicates the position of the oocyte nucleus (A–F). Arrows indicate the apical membrane and arrowheads the basal membrane of salivary glands and follicle cells (G–O). Bar, 30 μm. Above each image is a schematic representation of the reporter RNA analyzed. Reporter mRNAs were fused to: the oskar 3′ UTR (A); region 1 + 2 (B); region 2 + 3 (C); wild-type region 2 (D,G,J,M); region 2 3′mut (E,H,K,N); region 2 5′3′mut (F,I,L,O).