Three sequences in the vicinity of poly (A) addition sites are conserved among vertebrate mRNAs. We analyze the effects of single base changes in each position of AAUAAA and in the nucleotide to which poly (A) is added on 3' end formation in vitro. All 18 possible single base changes of the AAUAAA sequence greatly reduce addition of poly (A) to RNAs that end at the poly (A) addition site, and prevent cleavage of RNAs that extend beyond. The magnitude of reduction varies greatly with the position changed and the base introduced. For any given mutation, cleavage and polyadenylation are reduced to similar extents, strongly suggesting that the same factor interacts with AAUAAA in both reactions. Mutations at and near the conserved adenosine to which poly (A) is added disturb the accuracy, but not the efficiency, of 3' end formation. For example, point mutations at the conserved adenosine shift the 3' end of the most abundant 5' half-molecule downstream by a single nucleotide. The mechanism by which these mutations might exert their effects on the precision of 3' end formation are discussed.

authors request that the following corrections be noted. It was accidentally stated that the studies by Kajita et al. (1) and Lee et al. (2) dealt with cinnamoyl-CoA reductase modified plants when in fact they concerned 4-coumarate:coenzyme A ligase (4CL) transgenic plants. Lignin concentration was reduced by down-regulation of 4CL activity in both studies (1, 2). In a subsequent article, Kajita et al. (3) reported a negligible decrease in lignin concentration and a decreased syringyl-toguaiacyl ratio for lignin composition of a sense-suppressed 4CL transgenic tobacco line. Kajita et al. (1) rather than Kajita et al. (3) was inadvertently cited when this later report was contrasted with the large decreases in lignin concentration and an increased syringyl-to-guaiacyl lignin ratio for anti-sense suppressed 4CL Arabidopsis transgenics (2). The authors apologize for the confusion these errors have created for readers of their Commentary and to the authors of the cited work for misrepresenting their research. November 10, 1998, of Proc. Natl. Acad. Sci. USA (95, 13612-13617), the authors request that the following correction be noted: In Fig. 2 appearing on page 13614, the genotype identification for testicular histology in panels C and D were shown reversed. The correct identification is Ϫ͞Ϫ for panel C and ϩ͞ϩ for panel D. The fifth sentence of the figure legend should read as follows: "Histological sections at lower (E) and higher (D) magnification of the seminiferous tubuli from a wild-type and mutant (F and C) mouse."Cell Biology. In the article "Efficient construction of a large nonimmune phage antibody library: The production of highaffinity human single-chain antibodies to protein antigens" by

Early in the development of many animals, before transcription begins, any change in the pattern of protein synthesis is attributable to a change in the translational activity or stability of an mRNA in the egg. As a result, translational control is critical for a variety of developmental decisions, including axis formation in Drosophila and sex determination in Caenorhabditis elegans. Previous work demonstrated that increases in poly(A) length can activate translation, whereas removal of poly(A) can prevent it. In this report we focus on the control of c-mos and cyclin A1, B1, and B2 mRNAs during meiotic maturation and after fertilization of frog eggs. We show that addition and removal of poly(A) from these mRNAs is extensively regulated: The time at which each mRNA receives or loses poly(A), as well as the number of adenosines it gains or loses, differ substantially. Signals in the 3'-untranslated region (UTR) of each mRNA are sufficient to reconstitute both the temporal and quantitative control of poly(A) addition: Chimeric mRNAs in which a luciferase-coding region is joined to the 3' UTRs of cyclin A1, cyclin B1, or c-mos mRNA, receive poly(A) of the same length and at the same time as do the endogenous mRNAs. Moreover, each 3' UTR also regulates translation of the chimeric mRNAs, determining when and how much translation of the luciferase reporter is stimulated during maturation. The magnitude of stimulation in luciferase activity varies from 5-to 100-fold, depending on the 3' UTR. Translational stimulation by each 3' UTR requires poly(A) lengthening, as it is prevented by mutations that prevent that process. These results suggest that the 3' UTRs of cyclin and c-mos mRNAs control not only whether or not an mRNA is turned on during maturation, but when that activation occurs and to what extent. Translational control of c-mos mRNA, which may be achieved through regulation of poly(A) length, may be critical in the activation of maturation, and in the onset of cleavage divisions. Our findings, as well as those of others, suggest that even quite complex patterns of translational activation in the early embryo can be attained through the differential control of a common mechanism.[Key Words: 3' UTR; poly(A); maternal mRNAs; translational control; cyclin; c-mos; oocyte maturation; development; Xenopus]Received December 28, 1993; revised version accepted March 2, 1994. In the earliest period of an animal's life, transcriptional control is irrelevant because the zygotic genome is inactive. Instead, any change in the pattern of protein synthesis during the earliest postfertilization period is attributable to regulation of mRNAs that were already present in the egg before fertilization--so-called maternal mRNAs. As a result, key decisions made during early development often rely on control of the translation, stability, or location of these mRNAs. For example, axis formation in Drosophila depends on proper localization This paper is dedicated to Howard M. Temin.

In frog oocytes, certain maternal mRNAs receive poly(A) in the cytoplasm during progesterone-induced maturation. To analyze this reaction and to compare it to poly(A) addition in the nucleus, we injected short, synthetic RNA substrates into Xenopus oocytes. These RNAs contain only portions of the 3'-untranslated regions of appropriate mRNAs and end at the natural poly(A) site. We demonstrate that the nuclear and maturation-specific polyadenylation activities are distinct in substrate specificity and subcellular location. The sequence AAUAAA, contained in virtually all pre-mRNAs, is necessary for both activities. A second sequence element, UUUUUAU, activates poly(A) addition during maturation. UUUUUAU and AAUAAA are both necessary and virtually sufficient for maturation-specific polyadenylation: Poly(A) tails of between 50 and 300 nucleotides are added during maturation to RNAs containing both sequences but not to RNAs that lack either sequence. Before maturation, RNAs that contain AAUAAA are extended by just 10 nucleotides, presumably adenosines. The maturation-specific activity first appears within 1 hr of the time the nucleus breaks down but apparently does not require a nuclear component, as it is unaffected by enucleation. These observations, combined with those of others, lead us to speculate that polyadenylation may be responsible for the translational activation of a family of mRNAs essential for maturation.

c-mos protein, encoded by a proto-oncogene, is essential for the meiotic maturation of frog oocytes. Polyadenylation of c-mos messenger RNA is shown here to be a pivotal regulatory step in meiotic maturation. Maturation is prevented by selective amputation of polyadenylation signals from c-mos mRNA. Injection of a prosthetic RNA, which restores c-mos polyadenylation signals by base pairing to the amputated mRNA, rescues maturation and can stimulate translation in trans. Prosthetic RNAs may provide a general strategy by which to alter patterns of mRNA expression in vivo.

We show that microRNA-427 (miR-427) mediates the rapid deadenylation of maternal mRNAs after the midblastula transition (MBT) of Xenopus laevis embryogenesis. By MBT, the stage when the embryonic cell cycle is remodeled and zygotic transcription of mRNAs is initiated, each embryo has accumulated ;10 9 molecules of miR-427 processed from multimeric primiR-427 transcripts synthesized after fertilization. We demonstrate that the maternal mRNAs for cyclins A1 and B2 each contain a single miR-427 target sequence, spanning less than 30 nucleotides, that is both necessary and sufficient for deadenylation, and that inactivation of miR-427 leads to stabilization of the mRNAs. Although this deadenylation normally takes place after MBT, exogenous miRNAs produced prematurely in vivo can promote deadenylation prior to MBT, indicating that turnover of the maternal mRNAs is limited by the amount of accumulated miR-427. Injected transcripts comprised solely of the cyclin mRNA 39 untranslated regions or bearing a 59 ApppG cap undergo deadenylation, showing that translation of the targeted RNA is not required. miR-427 is not unique in promoting deadenylation, as an unrelated miRNA, let-7, can substitute for miR-427 if the reporter RNA contains an appropriate let-7 target site. We propose that miR-427, like the orthologous miR-430 of zebrafish, functions to down-regulate expression of maternal mRNAs early in development.

The sequence AAUAAA is found near the polyadenylation site of eucaryotic mRNAs. This sequence is required for accurate and efficient cleavage and polyadenylation of pre-mRNAs in vivo. In this study we show that synthetic simian virus 40 late pre-mRNAs are cleaved and polyadenylated in vitro in a HeLa cell nuclear extract, and that cleavage in vitro is abolished by each of four different single-base changes in AAUAAA. In this same extract, precleaved RNAs (RNAs with 3' termini at the polyadenylation site) are efficiently polyadenylated. This in vitro polyadenylation reaction also requires the AAUAAA sequence.The 3' termini of eucaryotic mRNAs are formed by endonucleolytic cleavage of an mRNA precursor and the addition of approximately 250 adenylate residues [poly(A)] to the newly formed end (8,22,26,34). The structural features of the precursor that are required for cleavage and polyadenylation have not been completely defined, but are known to include two separable elements, the AAUAAA sequence (10,13,23,27,32) and the downstream element (or TG box) (7,12,19,20,28,29,33).The AAUAAA sequence, located 6 to 26 bases before the poly(A) addition site of mammalian mRNAs, is required for cleavage. Deletions (10) or single-base-pair changes (13,23,32) in AAUAAA prevent cleavage in vivo, but do not prevent polyadenylation; thus mRNA precursors containing mutant AAUAAA sequences are cleaved inefficiently, but all precursors that are cleaved also are polyadenylated (23,32). The downstream element is located 10 to 30 bases beyond the polyadenylation site, in the 3' flanking sequence. mRNA precursors from which this sequence has been deleted are cleaved inefficiently (7,12,19,20,29). The AAUAAA sequence is quantitatively more important than the downstream element. For example, in injected oocytes, cleavage is reduced more than 50-fold by point mutations in AAUAAA (32) but only 5-to 10-fold by deletion of the downstream element (7; D. Zarkower, unpublished data). Similarly, the AAUAAA sequence is more highly conserved than the downstream element (4, 20, 32).Recently Moore and Sharp (25) demonstrated cleavage and polyadenylation of a synthetic adenovirus pre-mRNA in an extract prepared from human cells (9, 25). To establish the physiological relevance of these cell-free reactions, they must be compared with the same reactions in vivo; the in vitro reactions must be accurate and efficient and must depend on the same sequences that are required in vivo. Hart et al. (12) cleavage, polyadenylation is dependent on the AAUAAA sequence.MATERIALS AND METHODS RNA substrates. In vitro transcription was performed as described previously (21), except that the GTP concentration was 300 ,uM and 1.2 mM GpppG (P-L Biochemicals) was included to produce capped transcripts (14). After the transcription reaction, DNA was removed with RNase-free DNase (P-L), and the transcripts were purified by repeated ethanol precipitation in the presence of 2 M ammonium

We show that, in Xenopus laevis oocytes and early embryos, double-stranded exogenous siRNAs cannot function as microRNA (miRNA) mimics in either deadenylation or guided mRNA cleavage (RNAi). Instead, siRNAs saturate and inactivate maternal Argonaute (Ago) proteins, which are present in low amounts but are needed for Dicer processing of pre-miRNAs at the midblastula transition (MBT). Consequently, siRNAs impair accumulation of newly made miRNAs, such as the abundant embryonic pre-miR-427, but inhibition dissipates upon synthesis of zygotic Ago proteins after MBT. These effects of siRNAs, which are independent of sequence, result in morphological defects at later stages of development. The expression of any of several exogenous human Ago proteins, including catalytically inactive Ago2 (Ago2mut), can overcome the siRNA-mediated inhibition of miR-427 biogenesis and function. However, expression of wild-type, catalytically active hAgo2 is required to elicit RNAi in both early embryos and oocytes using either siRNA or endogenous miRNAs as guides. The lack of endogenous Ago2 endonuclease activity explains why these cells normally are unable to support RNAi. Expression of catalytically active exogenous Ago2, which appears not to perturb normal Xenopus embryonic development, can now be exploited for RNAi in this vertebrate model organism.