A physical and functional link between splicing factors promotes pre-mRNA 3′ end processing

Polypyrimidine tract-binding protein (PTB) is a splicing regulator that also plays a positive role in pre-mRNA 3′ end processing when bound upstream of the polyadenylation signal (pA signal). Here, we address the mechanism of PTB stimulatory function in mRNA 3′ end formation. We identify PTB as the protein factor whose binding to the human β-globin (HBB) 3′ UTR is abrogated by a 3′ end processing-inactivating mutation. We show that PTB promotes both in vitro 3′ end cleavage and polyadenylation and recruits directly the splicing factor hnRNP H to G-rich sequences associated with several pA signals. Increased binding of hnRNP H results in stimulation of polyadenylation through a direct interaction with poly(A) polymerase. Therefore, our results provide evidence of a concerted regulation of pA signal recognition by splicing factors bound to auxiliary polyadenylation sequence elements.     

Downstream sequence elements with different affinities for the hnRNP H/H' protein influence the processing efficiency of mammalian polyadenylation signals

Auxiliary factors likely play an important role in determining the polyadenylation efficiency of mammalian pre-mRNAs. We previously identified an auxiliary factor, hnRNP H/H′, which stimulates 3′-end processing through an interaction with sequences downstream of the core elements of the SV40 late polyadenylation signal. Using in vitro reconstitution assays we have demonstrated that hnRNP H/H′ can stimulate processing of two additional model polyadenylation signals by binding at similar relative downstream locations but with significantly different affinities. A short tract of G residues was determined to be a common property of all three hnRNP H/H′ binding sites. A survey of mammalian polyadenylation signals identified potential G-rich hnRNP H/H′ binding sites at similar downstream locations in ~34% of these signals. All of the novel G-rich elements tested were found to bind hnRNP H/H′ protein and the processing of selected signals identified in the survey was stimulated by the protein both in vivo and in vitro. Downstream G-rich tracts, therefore, are a common auxiliary element in mammalian polyadenylation signals. Sequences capable of binding hnRNP H protein with varying affinities may play a role in determining the processing efficiency of a significant number of mammalian polyadenylation signals.     

Probabilistic prediction of Saccharomyces cerevisiae mRNA 3'-processing sites

We present a tool for the prediction of mRNA 3′-processing (cleavage and polyadenylation) sites in the yeast Saccharomyces cerevisiae, based on a discrete state-space model or hidden Markov model. Comparison of predicted sites with experimentally verified 3′-processing sites indicates good agreement. All predicted or known yeast genes were analyzed to find probable 3′-processing sites. Known alternative 3′-processing sites, both within the 3′-untranslated region and within the protein coding sequence were successfully identified, leading to the possibility of prediction of previously unknown alternative sites. The lack of an apparent 3′-processing site calls into question the validity of some predicted genes. This is specifically investigated for predicted genes with overlapping coding sequences.     

Statistical analysis of yeast genomic downstream sequences reveals putative polyadenylation signals

The study of a few genes has permitted the identification of three elements that constitute a yeast polyadenylation signal: the efficiency element (EE), the positioning element and the actual site for cleavage and polyadenylation. In this paper we perform an analysis of oligonucleotide composition on the sequences located downstream of the stop codon of all yeast genes. Several oligonucleotide families appear over-represented with a high significance (referred to herein as ‘words’). The family with the highest over-representation includes the oligonucleotides shown experimentally to play a role as EEs. The word with the highest score is TATATA, followed, among others, by a series of single-nucleotide variants (TATGTA, TACATA, TAAATA . . .) and one-letter shifts (ATATAT). A position analysis reveals that those words have a high preference to be in 3′ flanks of yeast genes and there they have a very uneven distribution, with a marked peak around 35 bp after the stop codon. Of the predicted ORFs, 85% show one or more of those sequences. Similar results were obtained using a data set of EST sequences. Other clusters of over-represented words are also detected, namely T- and A-rich signals. Using these results and previously known data we propose a general model for the 3′ trailers of yeast mRNAs.     

CstF-64 and 3′-UTR cis-element determine Star-PAP specificity for target mRNA selection by excluding PAPα

Almost all eukaryotic mRNAs have a poly (A) tail at the 3′-end. Canonical PAPs (PAPα/γ) polyadenylate nuclear pre-mRNAs. The recent identification of the non-canonical Star-PAP revealed specificity of nuclear PAPs for pre-mRNAs, yet the mechanism how Star-PAP selects mRNA targets is still elusive. Moreover, how Star-PAP target mRNAs having canonical AAUAAA signal are not regulated by PAPα is unclear. We investigate specificity mechanisms of Star-PAP that selects pre-mRNA targets for polyadenylation. Star-PAP assembles distinct 3′-end processing complex and controls pre-mRNAs independent of PAPα. We identified a Star-PAP recognition nucleotide motif and showed that suboptimal DSE on Star-PAP target pre-mRNA 3′-UTRs inhibit CstF-64 binding, thus preventing PAPα recruitment onto it. Altering 3′-UTR cis-elements on a Star-PAP target pre-mRNA can switch the regulatory PAP from Star-PAP to PAPα. Our results suggest a mechanism of poly (A) site selection that has potential implication on the regulation of alternative polyadenylation.     

ElrA binding to the 3'UTR of cyclin E1 mRNA requires polyadenylation elements

The early cell divisions of Xenopus laevis and other metazoan embryos occur in the presence of constitutively high levels of the cell cycle regulator cyclin E1. Upon completion of the 12th cell division, a time at which many maternal proteins are downregulated by deadenylation and destabilization of their encoding mRNAs, maternal cyclin E1 protein is downregulated while its mRNA is polyadenylated and stable. We report here that stable polyadenylation of cyclin E1 mRNA requires three cis-acting elements in the 3′ untranslated region; the nuclear polyadenylation sequence, a contiguous cytoplasmic polyadenylation element and an upstream AU-rich element. ElrA, the Xenopus homolog of HuR and a member of the ELAV gene family binds the cyclin E1 3′UTR with high affinity. Deletion of these elements dramatically reduces the affinity of ElrA for the cyclin E1 3′UTR, abolishes polyadenylation and destabilizes the mRNA. Together, these findings provide compelling evidence that ElrA functions in polyadenylation and stabilization of cyclin E1 mRNA via binding these elements.     

The Drosophila homologue of the 64 kDa subunit of cleavage stimulation factor interacts with the 77 kDa subunit encoded by the suppressor of forked gene

During mRNA 3′ end formation, cleavage stimulation factor (CstF) binds to a GU-rich sequence downstream from the polyadenylation site and helps to stabilise the binding of cleavage-polyadenylation specificity factor (CPSF) to the upstream polyadenylation sequence (AAUAAA). The 64 kDa subunit of CstF (CstF-64) contains an RNA binding domain and is responsible for the RNA binding activity of CstF. It interacts with CstF-77, which in turn interacts with CPSF. The Drosophila suppressor of forked gene encodes a homologue of CstF-77, and mutations in it affect mRNA 3′ end formation in vivo. A Drosophila homologue for CstF-64 has now been isolated, both through homology with the human protein and through protein–protein interaction in yeast with the suppressor of forked gene product. Alignment of CstF-64 homologues shows that the proteins have a conserved N-terminal 200 amino acids, the first half of which is the RNA binding domain with the second half likely to contain the CstF-77 interaction domain; a central region variable in length and rich in glycine, proline and glutamine residues and containing an unusual degenerate repeat motif; and then a conserved C-terminal 50 amino acids. In Drosophila, the CstF-64 gene has a single 63 bp intron, is transcribed throughout development and probably corresponds to l(3)91Cd.     

Knock-down of 25 kDa subunit of cleavage factor Im in Hela cells alters alternative polyadenylation within 3'-UTRs

Alternative polyadenylation leads to mRNAs with variable 3′ ends. Since a 3′-untranslated region (3′-UTR) often contains cis elements that impact stability or localization of mRNA or translation, selection of poly(A) sites in a 3′-UTR is regulated in mammalian cells. However, the molecular basis for alternative poly(A) site selection within a 3′-UTR has been unclear. Here we show involvement of cleavage factor Im (CFIm) in poly(A) site selection within a 3′-UTR. CFIm is a heterodimeric 3′ end-processing complex, which functions to assemble other processing factors on pre-mRNA in vitro. We knocked down 25 kDa subunit of CFIm (CFIm25) in HeLa cells and analyzed alternative poly(A) site selection of TIMP-2, syndecan2, ERCC6 and DHFR genes by northern blotting. We observed changes in the distribution of mRNAs in CFIm25 depleted cells, suggesting a role for CFIm in alternative poly(A) site selection. Furthermore, tissue specific analysis demonstrated that the CFIm25 gene gave rise to 1.1, 2.0 and 4.6 kb mRNAs. The 4.6 kb mRNA was ubiquitously expressed, while the 1.1 and 2.0 kb mRNAs were expressed in a tissue specific manner. We found three likely poly(A) sites in the CFIm25 3′-UTR, suggesting alternative polyadenylation. Our results indicate that alternative poly(A) site selection is a well-regulated process in vivo.     

Spatial constraints on polyadenylation signal function

Efficient cleavage and polyadenylation of eukaryotic messenger RNAs require at least two signal elements: an AAUAAA or closely related sequence located 7-30 base pairs (bp) upstream of the site of processing, and a G/U- or U-rich sequence located 3' to the cleavage site. The herpes simplex virus type 1 thymidine kinase (tk) gene contains two copies of the AATAAA hexanucleotide and a GT-rich region. We have shown that the first AATAAA and the GT-rich region are essential for efficient processing, both in vivo and in vitro, whereas the second AATAAA does not appear to play any role in the formation of tk mRNA 3' ends. The failure of a signal containing only the second AATAAA and the GT-rich element to signal cleavage and polyadenylation suggested that these two elements might be too close together to constitute a functional polyadenylation signal. The experiments described in this report were directed at determining the effects on mRNA 3' end formation of alterations in spacing between signal elements. Wild-type tk contains 19 bp between these two elements. Constructs were made in which an AATAAA and the GT-rich region were separated by various distances ranging from 7 to 43 bp. The quantity and location of 3' ends of the tk mRNA produced by these constructs in Cos-1 cells were measured by S1 nuclease protection analysis. Signal efficiency was gradually reduced as the separation between the two signal elements was increased; with a separation of 43 bp, the signal functioned at approximately one-eighth the efficiency of the parental construction. Bringing the two signals closer together resulted in decreased signal efficiency; with a separation of 7 or 9 bp, no tk mRNA polyadenylated within the normal region was produced. Altering the sequences between these two elements without changing the distance had small effects on processing efficiency.     

Selection of sequence elements that substitute for the standard AATAAA motif which signals 3'' processing and polyadenylation of late simian virus 40 mRNAs.

A method is described which allows selection of sequences which can substitute for the normal AATAAA hexanucleotide involved in polyadenylation of SV40 late mRNAs. Plaques were generated from viral DNA lacking the motif, forcing acquisition of substitute sequences. Four variants were characterized. All displayed wild-type growth kinetics and produced normal levels of late mRNAs and proteins. Two variants had reacquired AATAAA elements and one acquired an ATTAAA sequence. The last variant carried an ATTTTTTAAA segment, suggesting this novel sequence, or some portion of it, can also signal poly A addition.     

The gene structure and expression of human ABHD1: overlapping polyadenylation signal sequence with Sec12

Background

Overlapping sense/antisense genes orientated in a tail-to-tail manner, often involving only the 3'UTRs, form the majority of gene pairs in mammalian genomes and can lead to the formation of double-stranded RNA that triggers the destruction of homologous mRNAs. Overlapping polyadenylation signal sequences have not been described previously.

Results

An instance of gene overlap has been found involving a shared single functional polyadenylation site. The genes involved are the human alpha/beta hydrolase domain containing gene 1 (ABHD1) and Sec12 genes. The nine exon human ABHD1 gene is located on chromosome 2p23.3 and encodes a 405-residue protein containing a catalytic triad analogous to that present in serine proteases. The Sec12 protein promotes efficient guanine nucleotide exchange on the Sar1 GTPase in the ER. Their sequences overlap for 42 bp in the 3'UTR in an antisense manner. Analysis by 3' RACE identified a single functional polyadenylation site, ATTAAA, within the 3'UTR of ABHD1 and a single polyadenylation signal, AATAAA, within the 3'UTR of Sec12. These polyadenylation signals overlap, sharing three bp. They are also conserved in mouse and rat. ABHD1 was expressed in all tissues and cells examined, but levels of ABHD1 varied greatly, being high in skeletal muscle and testis and low in spleen and fibroblasts.

Conclusions

Mammalian ABHD1 and Sec12 genes contain a conserved 42 bp overlap in their 3'UTR, and share a conserved TTTATTAAA/TTTAATAAA sequence that serves as a polyadenylation signal for both genes. No inverse correlation between the respective levels of ABHD1 and Sec12 RNA was found to indicate that any RNA interference occurred.