Translational activation of oskar mRNA: reevaluation of the role and importance of a 5' regulatory element [corrected]

Abstract

Local translation of oskar (osk) mRNA at the posterior pole of the Drosophila oocyte is essential for axial patterning of the embryo, and is achieved by a program of translational repression, mRNA localization, and translational activation. Multiple forms of repression are used to prevent Oskar protein from accumulating at sites other than the oocyte posterior. Activation is mediated by several types of cis-acting elements, which presumably control different forms of activation. We characterize a 5' element, positioned in the coding region for the Long Osk isoform and in the extended 5' UTR for translation of the Short Osk isoform. This element was previously thought to be essential for osk mRNA translation, with a role in posterior-specific release from repression. From our work, which includes assays which separate the effects of mutations on RNA regulatory elements and protein coding capacity, we find that the element is not essential, and conclude that there is no evidence supporting a role for the element only at the posterior of the oocyte. The 5' element has a redundant role, and is only required when Long Osk is not translated from the same mRNA. Mutations in the element do disrupt the anchoring function of Long Osk protein through their effects on the amino acid sequence, a confounding influence on interpretation of previous experiments.

Fig 1. Mapping an RNA element required for osk activity.

A. Diagram of the 5' region of the osk mRNA bearing the M1R mutation, which eliminates translation of Long Osk. The extended 5' UTR is shown as a black line, and the Short Osk coding region as a rectangle. The oskM1RHA transgene has the M1R mutation and contains 3 copies of the HA epitope tag (shaded region), inserted after residue T140 (the Short Osk start codon is M139). Deletions are indicated. B. Patterning activity of osk transgenes, tested as single copies in the osk ^A87 /Df(3R)osk background (RNA null). The number of abdominal segments corresponds to the level of osk activity, with wild type embryos having eight. n values were all over 250, except for oskM1R INV121-182 (166), oskM1R ∆61–90 (203), and oskM1R ∆211–260 (177). No anterior patterning defects were observed for any of the transgenes, including oskM1R. C. Levels of osk mRNA produced from a single copy of the indicated transgenes. All values are normalized against the level of mRNA from a single copy of the oskHA transgene, which is identical to oskM1RHA except that it has the wild type M1 codon. Levels of rp49 were monitored to normalize for amount of RNA used in each assay. Error bars indicate standard error. An analysis of variance (ANOVA) demonstrates no significant difference between RNA levels for any of the transgenes tested (F = 1.066, df = 13, p = 0.417).

Fig 2. Sequence conservation in the 5' region of the osk gene.

The diagram at top shows the 5' region of osk, with the extended 5' UTR as a black line and the osk coding region as a rectangle. The AUG start codons for Long and Short Osk are shown, and the region containing the 5' activation element is shaded. The two analyses of conservation are shown below, with the phastCons output above and the consecCons output below, as indicated. For the latter, each vertical line indicates the presence of 2 consecutive positions that are perfectly conserved among the species analyzed (Methods and Materials). At bottom are segments of the osk sequence showing the short regions most highly conserved in the extended 5' UTR. Within the coding region, codons are indicated by spacing, and perfectly conserved positions are identified with asterisks. Endpoints of the indicated deletion mutations are marked.

Fig 3. The 5' element is required for translational activation.

A and B. Western blot analysis of transgenes expressed as single copies in the osk ^A87 /Df(3R)osk background. Tubulin is detected as a loading control. C. Diagram of the oskM1R 5' region, showing the positions of the two iORFs and how the ∆311–360 deletion fuses iORF2 to the Osk reading frame, and thus can produce the novel protein band detected in A. The partial sequence shown has the reading frame for Osk protein indicated by spaces, and the reading frame for iORF2 indicated by vertical hash marks. D-G. In situ hybridization to detect transgene mRNAs in the osk ^A87 /Df(3R)osk background (panels D'-G' are magnified views of the posterior region to better show the mRNA distributions). For the oskHA transgene, which makes both Long and Short Osk, the mRNA is tightly restricted to a posterior crescent (D,D'). The oskM1R transgene lacks Long Osk and its anchoring function, and the mRNA has a more punctate distribution (E,E'). Similarly, both of the mutants tested, one with normal osk activity (F,F'; the ∆61–90 deletion) and one largely lacking osk activity (G,G'; the ∆118–135 deletion), have the same punctate distribution of mRNA.

Fig 4. The 5' element is required for the early phase of Osk protein accumulation.

A-B,A'-B'. Detection of transgenic OskHA protein expressed from single copies of the indicated transgenes in the osk ^A87 /Df(3R)osk background (RNA null). Panels A'-B' are magnified views of the posterior of the oocyte to better show the proteins. Green is OskHA and red is DNA detected with ToPro-3. C-J,C'-J'. Detection of transgenic Short OskHA protein expressed from single copies of the indicated transgenes in the presence of endogenous Long Osk for anchoring. Panels C'-J' are magnified views of the posterior of the oocyte to better show the proteins. I. Quantification of protein levels from the imaging experiments of C-J. OskHA signal intensities (Methods and Materials) are shown normalized to that from the oskHA transgene. The number of oocytes scored is indicated at the bottom of each bar. Error bars indicate standard error. J. Western blot analysis of transgenes expressed as single copies in the presence of a wild type copy of osk. Only the transgenic Osk protein is detected using anti-HA antibodies. Tubulin is detected as a loading control.

Fig 5. Effects of mutating the 5' element on translation and Long Osk function.

A. Diagram of the 5' region of the transgene mRNAs, using the conventions from Figs 1 and 2 but with the deletions marked by brackets and gaps. B. Western blot analysis of transgenes expressed as single copies in the osk ^A87 /Df(3R)osk background. C. Patterning activity of osk transgenes, tested as single copies in the osk ^A87 /Df(3R)osk background (RNA null). The number of abdominal segments corresponds to the level of osk activity, with wild type embryos having eight. n values were oskHA, 511; oskHA ∆121–150, 260; oskM1RHA ∆121–150, 324. D. Levels of osk mRNA produced from a single copy of the indicated transgenes. All values are normalized against the level of mRNA from a single copy of the oskHA transgene. Levels of rp49 were monitored to normalize for amount of RNA used in each assay. Error bars indicate standard error. E. Detection of transgenic OskHA expressed from single copies of the indicated transgenes. For the panels at left, the transgenes were tested in the osk ^A87 /Df(3R)osk background, revealing the anchoring defect of the OskHA∆121–150 mutant, which lacks aa 36–45. This defect is rescued when coexpressed with wild type Long Osk, as shown in the panels at right. F-G. The amino terminal domain of Osk confers anchoring on GFP. F shows the distribution of GFP, and G shows the distribution of the Osk::GFP fusion protein from transgene UAS-osk1-534::GFP, which includes the first 534 bp of the osk mRNA and encodes a fusion protein with the first 173 amino acids of Osk, including the entire Long Osk amino terminal domain. The fusion protein is highly enriched at the oocyte cortex and at nurse cell boundaries. White boxes outline the types of regions shown in panels H,J,L (solid lines) and I,K,M (dashed lines), although these are not the same egg chambers as in those panels. Green is GFP (or Osk::GFP), red is DNA (nuclei) stained with ToPro-3. H-M. Detection of Osk::GFP fusion proteins in stage 10 egg chambers. All panels are at higher magnification than in F and G to highlight anchoring at the oocyte cortex (H,J,L) or at nurse cell boundaries (I,K,M)(the scale bar is 5 μm). For panels I, K, M and M' the images show a portions of several nurse cells and the boundaries between them. Signal intensities can only be compared between panels J-M, which were imaged under identical conditions. Panels L' and M' are identical to L and M except that the green signal was enhanced to better show the absence of any anchoring. The level of protein from the UAS-GFP transgene is much higher than from the UAS-osk1-534::GFP transgenes, and lower intensity laser settings were used to obtain images in F, H and I with signal intensity comparable to G, J and K. Anchoring of the Osk::GFP protein is manifested in the enrichment at the cortex, along nurse cell boundaries, and the punctate appearance in the cytoplasm. Neither GFP alone nor the Osk INV121-182::GFP protein shows any similar anchoring. Green is GFP (or Osk::GFP), red is DNA stained with ToPro-3.