Overlapping transcripts, double-stranded RNA and antisense regulation: a genomic perspective.

Bioinformatic analysis of the transcriptomes of diverse eukaryotes has demonstrated the ubiquity and structural diversity of complementary antisense RNAs. These include both trans-encoded microRNAs and a large population of cis-encoded antisense RNAs that encompasses both coding and non-coding RNAs. Antisense regulation has previously been characterized primarily as a post-transcriptional response affecting RNA stability, nuclear processing, export, or translation. However, the formation of double-stranded (ds) RNAs by base-pairing between complementary RNAs may elicit regulatory responses at the transcriptional level as well. Analysis of antisense transcription at several imprinted loci has suggested a number of other mechanisms that may not require formation of dsRNA. Understanding the integration of transcriptional and post-transcriptional regulatory mechanisms represents a major challenge for understanding antisense regulation in eukaryotes. Such insight is also essential for understanding general principles of genetic regulation within the complex genomes characteristic of mouse and man as well as those of other eukaryotes.

A purine-rich intronic element enhances alternative splicing of thyroid hormone receptor mRNA.

The mammalian thyroid hormone receptor gene c-erbAalpha gives rise to two mRNAs that code for distinct isoforms, TRalpha1 and TRalpha2, with antagonistic functions. Alternative processing of these mRNAs involves the mutually exclusive use of a TRalpha1-specific polyadenylation site or TRalpha2-specific 5' splice site. A previous investigation of TRalpha minigene expression defined a critical role for the TRalpha2 5' splice site in directing alternative processing. Mutational analysis reported here shows that purine residues within a highly conserved intronic element, SEa2, enhance splicing of TRalpha2 in vitro as well as in vivo. Although SEalpha2 is located within the intron of TRalpha2 mRNA, it activates splicing of a heterologous dsx pre-mRNA when located in the downstream exon. Competition with wild-type and mutant RNAs indicates that SEalpha2 functions by binding trans-acting factors in HeLa nuclear extract. Protein-RNA crosslinking identifies several proteins, including SF2/ASF and hnRNP H, that bind specifically to SEalpha2. SEalpha2 also includes an element resembling a 5' splice site consensus sequence that is critical for splicing enhancer activity. Mutations within this pseudo-5' splice site sequence have a dramatic effect on splicing and protein binding. Thus SEa2 and its associated factors are required for splicing of TRalpha2 pre-mRNA.

Selection of alternative 5' splice sites: role of U1 snRNP and models for the antagonistic effects of SF2/ASF and hnRNP A1.

The first component known to recognize and discriminate among potential 5' splice sites (5'SSs) in pre-mRNA is the U1 snRNP. However, the relative levels of U1 snRNP binding to alternative 5'SSs do not necessarily determine the splicing outcome. Strikingly, SF2/ASF, one of the essential SR protein-splicing factors, causes a dose-dependent shift in splicing to a downstream (intron-proximal) site, and yet it increases U1 snRNP binding at upstream and downstream sites simultaneously. We show here that hnRNP A1, which shifts splicing towards an upstream 5'SS, causes reduced U1 snRNP binding at both sites. Nonetheless, the importance of U1 snRNP binding is shown by proportionality between the level of U1 snRNP binding to the downstream site and its use in splicing. With purified components, hnRNP A1 reduces U1 snRNP binding to 5'SSs by binding cooperatively and indiscriminately to the pre-mRNA. Mutations in hnRNP A1 and SF2/ASF show that the opposite effects of the proteins on 5'SS choice are correlated with their effects on U1 snRNP binding. Cross-linking experiments show that SF2/ASF and hnRNP A1 compete to bind pre-mRNA, and we conclude that this competition is the basis of their functional antagonism; SF2/ASF enhances U1 snRNP binding at all 5'SSs, the rise in simultaneous occupancy causing a shift in splicing towards the downstream site, whereas hnRNP A1 interferes with U1 snRNP binding such that 5'SS occupancy is lower and the affinities of U1 snRNP for the individual sites determine the site of splicing.

Post-transcriptional regulation of thyroid hormone receptor expression by cis-acting sequences and a naturally occurring antisense RNA.

Thyroid hormone (T(3)) coordinates growth, differentiation, and metabolism by binding to nuclear thyroid hormone receptors (TRs). The TRalpha gene encodes T(3)-activated TRalpha1 (NR1A1a) as well as an antagonistic, non-T(3)-binding alternatively spliced product, TRalpha2 (NR1A1b). Thus, the TRalpha1/TRalpha2 ratio is a critical determinant of T(3) action. However, the mechanisms underlying this post-transcriptional regulation are unknown. We have identified a non-consensus, TRalpha2-specific 5' splice site and conserved intronic sequences as key determinants of TRalpha mRNA processing. In addition to these cis-acting elements, a novel regulatory feature is the orphan receptor RevErbAalpha (NR1D1) gene, which is transcribed from the opposite direction at the same locus and overlaps the TRalpha2 coding region. RevErbAalpha gene expression correlates with a high TRalpha1/TRalpha2 ratio in a number of tissues. Here we demonstrate that coexpression of RevErbAalpha and TRalpha regulates the TRalpha1/TRalpha2 ratio in intact cells. Thus, both cis- and trans-regulatory mechanisms contribute to cell-specific post-transcriptional regulation of TR gene expression and T(3) action.

Distinct functions of the closely related tandem RNA-recognition motifs of hnRNP A1.

hnRNP A1 regulates alternative splicing by antagonizing SR proteins. It consists of two closely related, tandem RNA-recognition motifs (RRMs), followed by a glycine-rich domain. Analysis of variant proteins with duplications, deletions, or swaps of the RRMs showed that although both RRMs are required for alternative splicing function, each RRM plays distinct roles, and their relative position is important. Surprisingly, RRM2 but not RRM1 could support this function when duplicated, despite their very similar structure. Specific RNA binding and annealing are not sufficient for hnRNP A1 alternative splicing function. These observations, together with phylogenetic and structural data, suggest that the two RRMs are quasi-symmetric but functionally nonequivalent modules that evolved as components of a single bipartite domain.

Expression of the thyroid hormone receptor gene, erbAalpha, in B lymphocytes: alternative mRNA processing is independent of differentiation but correlates with antisense RNA levels.

The erbAalpha gene encodes two alpha-thyroid hormone receptor isoforms, TRalpha1 and TRalpha2, which arise from alternatively processed mRNAs, erbAalpha1 (alpha1) and erb alpha2 (alpha2). The splicing and alternative polyadenylation patterns of these mRNAs resemble that of mRNAs encoding different forms of immunoglobulin heavy chains, which are regulated at the level of alternative processing during B cell differentiation. This study examines the levels of erbAalpha mRNA in eight B cell lines representing four stages of differentiation in order to determine whether regulation of the alternatively processed alpha1 and alpha2 mRNAs parallels the processing of immunoglobulin heavy chain mRNAs. Results show that the pattern of alpha1 and alpha2 mRNA expression is clearly different from that observed for immunoglobulin heavy chain mRNAs. B cell lines display characteristic ratios of alpha1/alpha2 mRNA at distinct stages of differentiation. Furthermore, expression of an overlapping gene, Rev-ErbAalpha (RevErb), was found to correlate strongly with an increase in the ratio of alpha1/alpha2 mRNA. These results suggest that alternative processing of erbAalpha mRNAs is regulated by a mechanism which is distinct from that regulating immunoglobulin mRNA. The correlation between RevErb and erbAalpha mRNA is consistent with negative regulation of alpha2 via antisense interactions with the complementary RevErb mRNA.

Function of conserved domains of hnRNP A1 and other hnRNP A/B proteins.

hnRNP A1 is a pre-mRNA binding protein that antagonizes the alternative splicing activity of splicing factors SF2/ASF or SC35, causing activation of distal 5' splice sites. The structural requirements for hnRNP A1 function were determined by mutagenesis of recombinant human hnRNP A1. Two conserved Phe residues in the RNP-1 submotif of each of two RNA recognition motifs appear to be involved in specific RNA-protein interactions and are essential for modulating alternative splicing. These residues are not required for general pre-mRNA binding or RNA annealing activity. The C-terminal Gly-rich domain is necessary for alternative splicing activity, for stable RNA binding and for optimal RNA annealing activity. hnRNP A1B, which is an alternatively spliced isoform of hnRNP A1 with a longer Gly-rich domain, binds more strongly to pre-mRNA but has only limited alternative splicing activity. In contrast, hnRNP A2 and B1, which have 68% amino acid identity with hnRNP A1, bind more weakly to pre-mRNA and have stronger splice site switching activities than hnRNP A1. We propose that specific combinations of antagonistic hnRNP A/B and SR proteins are involved in regulating alternative splicing of distinct subsets of cellular premRNAs.

Heterogeneous nuclear ribonucleoprotein A1 catalyzes RNA.RNA annealing.

Within the nucleus, pre-mRNA molecules are complexed with a set of proteins to form heterogeneous nuclear ribonucleoprotein complexes. A1, an abundant RNA binding protein present in these complexes, has been shown to bind selectively to single-stranded RNAs and destabilize base-pairing interactions. In this study.A1 is shown to promote the rate of annealing of complementary RNA strands greater than 300-fold under a wide range of salt concentration and temperature. Maximal annealing is observed under saturating or near saturating concentrations of protein, but annealing decreases sharply at both higher and lower concentrations of A1. Kinetic analysis shows that the rate of annealing is not strictly first or second order with respect to RNA at a ratio of protein/RNA that gives optimal rates of annealing. This result suggests that A1 protein may affect more than one step in the annealing reaction. Two polypeptides representing different domains of A1 were also examined for annealing activity. UP1, a proteolytic fragment that represents the N-terminal two-thirds of A1, displays very limited annealing activity. In contrast, a peptide consisting of 48 amino acid residues from the glycine-rich C-terminal region promotes annealing at a rate almost one-quarter that observed with intact A1. The RNA.RNA annealing activity of A1 may play a role in pre-mRNA splicing and other aspects of nuclear mRNA metabolism.

Inhibition of c-erbA mRNA splicing by a naturally occurring antisense RNA.

The rat erbA alpha locus encodes two overlapping mRNAs, alpha 1 and alpha 2, which are identical except for their most 3' exons. alpha 1 mRNA encodes a thyroid hormone receptor, while alpha 2 encodes an altered ligand binding domain of unknown function. Previous studies have shown that the ratio of alpha 1 to alpha 2 is highest in cells expressing a high level of a third RNA, Rev-ErbA alpha mRNA, which is transcribed in the opposite direction and is complementary to alpha 2 but not alpha 1 mRNA. It was hypothesized that base pairing with Rev-ErbA alpha blocks splicing of alpha 2 mRNA, thereby favoring formation of the non-overlapping alpha 1. To test this model, a system was developed in which alpha 2 pre-mRNAs were accurately spliced in vitro. Splicing was inhibited by the addition of a 5-fold excess of antisense RNAs containing the 3' end of Rev-ErbA alpha mRNA. Both an antisense RNA extending across the 3' splice site and a shorter RNA complementary only to exon sequences efficiently blocked splicing. However, splicing was only inhibited by complementary RNAs. These observations are consistent with a mechanism in which base pairing with a complementary RNA regulates alternative processing of alpha 1 and alpha 2 mRNAs.

Antisense RNA inhibits splicing of pre-mRNA in vitro.

Antisense RNAs complementary to human beta-globin pre-mRNA or to a chimeric globin/adenovirus E2a pre-mRNA specifically and efficiently inhibit pre-mRNA splicing in vitro. The level of inhibition depends on the length, position and concentration of the antisense RNA relative to the pre-mRNA substrate. Antisense RNAs complementary to sequences greater than 80 nucleotides downstream of the globin 3' splice site inhibit at least as efficiently as those extending across the splice sites. Thus splicing is sensitive to perturbations involving exon sequences some distance from the splice sites. Inhibition is mediated by factors which affect the annealing of antisense and substrate RNAs. Direct analysis of RNA duplex formation demonstrates the presence of an activity in HeLa cell nuclear extract which promotes the rapid annealing of complementary RNAs in an ATP-independent manner. Both annealing and inhibition are greatly reduced when antisense RNA is added to the splicing reaction greater than or equal to 5 min after substrate. This result may reflect a transition between an open structure, in which annealing of antisense RNA with pre-mRNA is facilitated, and a closed complex in which pre-mRNA is sequestered at an early stage of spliceosome assembly.

Splice site consensus sequences are preferentially accessible to nucleases in isolated adenovirus RNA.

The conformation of RNA sequences spanning five 3' splice sites and two 5' splice sites in adenovirus mRNA was probed by partial digestion with single-strand specific nucleases. Although cleavage of nucleotides near both 3' and 5' splice sites was observed, most striking was the preferential digestion of sequences near the 3' splice site. At each 3' splice site a region of very strong cleavage is observed at low concentrations of enzyme near the splice site consensus sequence or the upstream branch point consensus sequence. Additional sites of moderately strong cutting near the branch point consensus sequence were observed in those sequences where the splice site was the preferred target. Since recognition of the 3' splice site and branch site appear to be early events in mRNA splicing these observations may indicate that the local conformation of the splice site sequences may play a direct or indirect role in enhancing the accessibility of sequences important for splicing.

Secondary structure of splice sites in adenovirus mRNA precursors.

In order to investigate the possible role of RNA secondary structure in determining the efficiency and specificity of mRNA splicing, the structures of sequences at three acceptor splice sites in adenovirus were studied. Transcripts spanning intron-exon junctions were synthesized using SP6 RNA polymerase and analyzed using single and double-strand specific nucleases. Distinctive patterns of nuclease cleavage were observed for each of the 3 sites examined. At both sites in the E2a region sequences adjacent to the splice sites were particularly susceptible to digestion with T1 and S1 nucleases. In contrast, a splice site for hexon mRNA was largely resistant to these nucleases. The results obtained suggest that the conformation of the RNA at some, but not all, acceptor sites may enhance the accessibility of these sites to factors involved in splicing nuclear RNA and confirm the presence of a large, previously predicted hairpin structure centered on the acceptor site at 67 map units.

A large inverted repeat sequence overlaps two acceptor splice sites in adenovirus.

The distribution of nucleotide sequences resembling functional sites for mRNA splicing was examined by computer-directed searches in order to determine what factors may influence splice site selection in nuclear precursors. In particular, the distribution of large potentially stable hairpin structures or regions of extensive dyad symmetry was studied in adenovirus sequences. One region, spanning 106 nucleotides, was found at 66.4 map units, overlapping back-to-back acceptor sites for two mRNA molecules, those coding for the 100K protein and the 72K DNA binding protein, which are transcribed from opposite strands. This region displays exceptional dyad symmetry and is potentially capable of forming a single, highly stable hairpin when transcribed. It seems likely that the secondary structure as well as the primary structure of RNA plays a role in determining the correct splicing of these mRNA molecules.