The canonical cytoplasmic polyadenylation signals function in Drosophila. (A) Cytoplasmic polyadenylation in Drosophila embryo extracts. Nonadenylated full-length sop and bcd mRNAs, as well as the 3′ UTR of Toll, were incubated in 90-min embryo extracts, and the poly(A) tail was measured using the PAT assay. A schematic representation of the oligonucleotides used in this assay is shown in the top panel. For each transcript, a specific 5′ oligonucleotide was combined with either a specific 3′ primer to reveal the size of the nonadenylated RNA (3′ lanes), or with an oligo(dT) anchor to visualize the poly(A) tail (dT lanes). Molecular weight markers are also shown. (B) Cytoplasmic polyadenylation promotes translation of Toll mRNA. (Left panel) Polyadenylation time course of the Toll 3′ UTR, as measured by PAT assay. (Right panel) A firefly luciferase reporter containing the 3′ UTR of Toll (Luc-toll) was incubated in embryo extracts for different times, and the efficiency of translation was determined by measuring the luciferase activity. As controls, the translation efficiencies of nonadenylated Luciferase and cordycepin-treated Luc-toll mRNAs were determined. Curves represent the average of five independent experiments. (C) The CPE and the hexanucleotide are functional CPEs in Drosophila. (Left panel) Polyadenylation of radioactively labeled wild-type and hexanucleotide-mutated CycB1 3′ UTRs. The nature of the mutation is indicated in gray lowercase letters. RNAs were separated in a 6% denaturing acrylamide gel and visualized using the PhosphorImager. (Right panel) Polyadenylation of CAT reporter mRNAs containing either a wild-type CycB1 3′ UTR or a derivative with U-to-G transversions in all three CPEs. RNAs were visualized by Northern blot using a probe against the CAT ORF. A schematic representation of a Xenopus CycB1 3′ UTR is shown, with the three CPEs and the polyadenylation hexanucleotide (HN) depicted as gray and white boxes, respectively.

Polyadenylation of Toll mRNA is independent of the CPE and the hexanucleotide. (A, top panel) Schematic representation of a Toll 3′ UTR (1256 nt) and mutant derivatives. The location and sequence of the putative CPE and the hexanucleotide is detailed. (Bottom panel) Polyadenylation of these mRNAs was measured by Northern blot. (B) Polyadenylation competition assays. Polyadenylation of a ³²P-labeled Toll 3′ UTR or CAT-CycB1 after addition of excess Toll or CycB1 3′ UTRs. RNAs were separated in a 1% denaturing agarose gel. Input (i) RNA probes are also shown. The percentage of polyadenylated transcript with respect to total transcript within each lane is indicated (%pA). Lanes 16–21 in the bottom panel show samples of the same reactions as lanes 4–9 taken 15 min later.

Elements required for cytoplasmic polyadenylation of Toll mRNA. (A, left panel) Schematic representation of Toll 3′ UTR deletion derivatives. Nucleotide positions are shown according to the annotated Drosophila Toll sequence, taking as reference the first nucleotide of the 3′ UTR. The location of unique restriction sites and the PR are indicated. Restriction sites at both ends of Toll belong to the vector in which this sequence was cloned. Typical patterns of these RNAs are shown in the middle panel, before (−) or after (+) incubation in the Drosophila extracts. The sizes of the respective poly(A) tails are indicated. (Right panel) Quantification of the efficiency of polyadenylation, measured as the percentage of RNA that was polyadenylated versus the total signal (polyadenylated and nonadenylated) within the “+” lane. Values represent the average of at least three independent experiments. (B) The PR is essential for polyadenylation and translation of Toll. (Left panel) Polyadenylation time course of a wild-type Toll 3′ UTR or a mutant derivative lacking the PR. (Right panel) Translation efficiencies of reporter mRNAs containing either wild-type or ΔPR Toll 3′ UTRs. Renilla luciferase mRNA was cotranslated as an internal control. The firefly values were corrected for Renilla expression, and the data are represented relative to the translation of Luc-toll mRNA at the last time point. Curves represent the average of three independent experiments using a single batch of extract.

The PR is required for noncanonical cytoplasmic polyadenylation in vitro and in vivo. (A) Polyadenylation competition experiments using the RNAs schematically represented in the top panel as competitors. Experiments were performed as indicated in the legend for Figure 2B. (B) Excess PRs disrupt polyadenylation of endogenous Toll and reduce viability. (Left panel) Embryos (0–30 min) were microinjected with different amounts of the PR or an unrelated RNA of similar length (159 nt) at a concentration of 10 ng/μL. Samples were collected 1 h after injection, the RNA was extracted, and the poly(A) tail length was tested by PAT assay. Amplified products were visualized by Southern blot using a random-primed probe against the Toll 3′ UTR. The results for two independent injections are shown. (Right panel) Embryos were microinjected with RNA solutions at different concentrations or with water as control, and survival was scored as the percentage of hatched embryos as indicated in the Materials and Methods. The average of at least three independent injections of 100 embryos per injection is shown.