Homology with Saccharomyces cerevisiae RNA14 suggests that phenotypic suppression in Drosophila melanogaster by suppressor of forked occurs at the level of RNA stability

Codon usage biases co-evolve with transcription termination machinery to suppress premature cleavage and polyadenylation

Codon usage biases are found in all genomes and influence protein expression levels. The codon usage effect on protein expression was thought to be mainly due to its impact on translation. Here, we show that transcription termination is an important driving force for codon usage bias in eukaryotes. Using Neurospora crassa as a model organism, we demonstrated that introduction of rare codons results in premature transcription termination (PTT) within open reading frames and abolishment of full-length mRNA. PTT is a wide-spread phenomenon in Neurospora, and there is a strong negative correlation between codon usage bias and PTT events. Rare codons lead to the formation of putative poly(A) signals and PTT. A similar role for codon usage bias was also observed in mouse cells. Together, these results suggest that codon usage biases co-evolve with the transcription termination machinery to suppress premature termination of transcription and thus allow for optimal gene expression.

The conserved intronic cleavage and polyadenylation site of CstF-77 gene imparts control of 3' end processing activity through feedback autoregulation and by U1 snRNP

The human gene encoding the cleavage/polyadenylation (C/P) factor CstF-77 contains 21 exons. However, intron 3 (In3) accounts for nearly half of the gene region, and contains a C/P site (pA) with medium strength, leading to short mRNA isoforms with no apparent protein products. This intron contains a weak 5′ splice site (5′SS), opposite to the general trend for large introns in the human genome. Importantly, the intron size and strengths of 5′SS and pA are all highly conserved across vertebrates, and perturbation of these parameters drastically alters intronic C/P. We found that the usage of In3 pA is responsive to the expression level of CstF-77 as well as several other C/P factors, indicating it attenuates the expression of CstF-77 via a negative feedback mechanism. Significantly, intronic C/P of CstF-77 pre-mRNA correlates with global 3′UTR length across cells and tissues. In addition, inhibition of U1 snRNP also leads to regulation of the usage of In3 pA, suggesting that the C/P activity in the cell can be cross-regulated by splicing, leading to coordination between these two processes. Importantly, perturbation of CstF-77 expression leads to widespread alternative cleavage and polyadenylation (APA) and disturbance of cell proliferation and differentiation. Thus, the conserved intronic pA of the CstF-77 gene may function as a sensor for cellular C/P and splicing activities, controlling the homeostasis of CstF-77 and C/P activity and impacting cell proliferation and differentiation.

Autoregulation is commonly used in biological systems to control the homeostasis of certain activity, and cross-regulation coordinates multiple processes. We show that vertebrate genes encoding the cleavage/polyadenylation (C/P) factor CstF-77 contain a conserved intronic C/P site (pA) which regulates CstF-77 expression through a negative feedback loop. Since the usage of this intronic pA is also responsive to the expression of other C/P factors, the pA can function as a sensor for the cellular C/P activity. Because the CstF-77 level is important for the usage of a large number of pAs in the genome and is particularly critical for expression of genes involved in cell cycle, this autoregulatory mechanism has far-reaching implications for cell proliferation and differentiation. The human intron harboring the pA is large and has a weak 5′ splice site, both of which are also highly conserved in other vertebrates. Inhibition of U1 snRNP, which recognizes the 5′ splice site of intron, leads to upregulation of the intronic pA isoform of CstF-77 gene, suggesting that the C/P activity in the cell can be cross-regulated by splicing, leading to coordination between these two processes.

Alternative mRNA polyadenylation in eukaryotes: an effective regulator of gene expression

Alternative RNA processing mechanisms, including alternative splicing and alternative polyadenylation, are increasingly recognized as important regulators of gene expression. This article will focus on what has recently been described about alternative polyadenylation in development, differentiation, and disease in higher eukaryotes. We will also describe how the evolving global methodologies for examining the cellular transcriptome, both experimental and bioinformatic, are revealing new details about the complex nature of alternative 3(') end formation as well as interactions with other RNA-mediated and RNA processing mechanisms.

A family of splice variants of CstF-64 expressed in vertebrate nervous systems

Background

Alternative splicing and polyadenylation are important mechanisms for creating the proteomic diversity necessary for the nervous system to fulfill its specialized functions. The contribution of alternative splicing to proteomic diversity in the nervous system has been well documented, whereas the role of alternative polyadenylation in this process is less well understood. Since the CstF-64 polyadenylation protein is known to be an important regulator of tissue-specific polyadenylation, we examined its expression in brain and other organs.

Results

We discovered several closely related splice variants of CstF-64 – collectively called βCstF-64 – that could potentially contribute to proteomic diversity in the nervous system. The βCstF-64 splice variants are found predominantly in the brains of several vertebrate species including mice and humans. The major βCstF-64 variant mRNA is generated by inclusion of two alternate exons (that we call exons 8.1 and 8.2) found between exons 8 and 9 of the CstF-64 gene, and contains an additional 147 nucleotides, encoding 49 additional amino acids. Some variants of βCstF-64 contain only the first alternate exon (exon 8.1) while other variants contain both alternate exons (8.1 and 8.2). In mice, the predominant form of βCstF-64 also contains a deletion of 78 nucleotides from exon 9, although that variant is not seen in any other species examined, including rats. Immunoblot and 2D-PAGE analyses of mouse nuclear extracts indicate that a protein corresponding to βCstF-64 is expressed in brain at approximately equal levels to CstF-64. Since βCstF-64 splice variant family members were found in the brains of all vertebrate species examined (including turtles and fish), this suggests that βCstF-64 has an evolutionarily conserved function in these animals. βCstF-64 was present in both pre- and post-natal mice and in different regions of the nervous system, suggesting an important role for βCstF-64 in neural gene expression throughout development. Finally, experiments in representative cell lines suggest that βCstF-64 is expressed in neurons but not glia.

Conclusion

This is the first report of a family of splice variants encoding a key polyadenylation protein that is expressed in a nervous system-specific manner. We propose that βCstF-64 contributes to proteomic diversity by regulating alternative polyadenylation of neural mRNAs.

Crystal structure of murine CstF-77: dimeric association and implications for polyadenylation of mRNA precursors

Cleavage stimulation factor (CstF) is a heterotrimeric protein complex essential for polyadenylation of mRNA precursors. The 77 kDa subunit, CstF-77, is known to mediate interactions with the other two subunits of CstF as well as with other components of the polyadenylation machinery. We report here the crystal structure of the HAT (half a TPR) domain of murine CstF-77, as well as its C-terminal subdomain. Structural and biochemical studies show that the HAT domain consists of two subdomains, HAT-N and HAT-C domains, with drastically different orientations of their helical motifs. The structures reveal a highly elongated dimer, spanning 165 A, with the dimerization mediated by the HAT-C domain. Light-scattering studies, yeast two-hybrid assays, and analytical ultracentrifugation measurements confirm this self-association. The mode of dimerization and the relative arrangement of the HAT-N and HAT-C domains are unique to CstF-77. Our data support a role for CstF dimerization in pre-mRNA 3' end processing.

The SU(VAR)3-9/HP1 complex differentially regulates the compaction state and degree of underreplication of X chromosome pericentric heterochromatin in Drosophila melanogaster

In polytene chromosomes of Drosophila melanogaster, regions of pericentric heterochromatin coalesce to form a compact chromocenter and are highly underreplicated. Focusing on study of X chromosome heterochromatin, we demonstrate that loss of either SU(VAR)3-9 histone methyltransferase activity or HP1 protein differentially affects the compaction of different pericentric regions. Using a set of inversions breaking X chromosome heterochromatin in the background of the Su(var)3-9 mutations, we show that distal heterochromatin (blocks h26-h29) is the only one within the chromocenter to form a big "puff"-like structure. The "puffed" heterochromatin has not only unique morphology but also very special protein composition as well: (i) it does not bind proteins specific for active chromatin and should therefore be referred to as a pseudopuff and (ii) it strongly associates with heterochromatin-specific proteins SU(VAR)3-7 and SUUR, despite the fact that HP1 and HP2 are depleted particularly from this polytene structure. The pseudopuff completes replication earlier than when it is compacted as heterochromatin, and underreplication of some DNA sequences within the pseudopuff is strongly suppressed. So, we show that pericentric heterochromatin is heterogeneous in its requirement for SU(VAR)3-9 with respect to the establishment of the condensed state, time of replication, and DNA polytenization.

An intronic polyadenylation site in human and mouse CstF-77 genes suggests an evolutionarily conserved regulatory mechanism

Human CstF-77 is one of the three subunits of cleavage stimulation factor (CstF) that is essential for mRNA polyadenylation. Its Drosophila homologue, suppressor of forked [su(f)], contains an intronic poly(A) site, which can lead to a short transcript without a stop codon. By both bioinformatic searches and validation with molecular biology experiments, we found that human and mouse CstF-77 genes also contain an intronic poly(A) site, which can be utilized to produce short CstF-77 transcripts lacking sequences encoding domains that are involved in many of the CstF-77 functions. The genomic sequence surrounding the poly(A) site is highly conserved among all vertebrates, but is not present in non-vertebrate species. Using public Serial Analysis of Gene Expression (SAGE) data, we found that the intronic poly(A) site is utilized in a wide range of tissues. This finding indicates that vertebrates may employ a similar alternative polyadenylation mechanism to modulate CstF-77, highlighting the importance of the regulation of CstF-77 in various species.

FY is an RNA 3' end-processing factor that interacts with FCA to control the Arabidopsis floral transition

The nuclear RNA binding protein, FCA, promotes Arabidopsis reproductive development. FCA contains a WW protein interaction domain that is essential for FCA function. We have identified FY as a protein partner for this domain. FY belongs to a highly conserved group of eukaryotic proteins represented in Saccharomyces cerevisiae by the RNA 3' end-processing factor, Pfs2p. FY regulates RNA 3' end processing in Arabidopsis as evidenced through its role in FCA regulation. FCA expression is autoregulated through the use of different polyadenylation sites within the FCA pre-mRNA, and the FCA/FY interaction is required for efficient selection of the promoter-proximal polyadenylation site. The FCA/FY interaction is also required for the downregulation of the floral repressor FLC. We propose that FCA controls 3' end formation of specific transcripts and that in higher eukaryotes, proteins homologous to FY may have evolved as sites of association for regulators of RNA 3' end processing.

A history of poly A sequences: from formation to factors to function

Biological polyadenylation, first recognized as an enzymatic activity, remained an orphan enzyme until poly A sequences were found on the 3' ends of eukarvotic mRNAs. Their presence in bacteria viruses and later in archeae (ref. 338) established their universality. The lack of compelling evidence for a specific function limited attention to their cellular formation. Eventually the newer techniques of molecular biology and development of accurate nuclear processing extracts showed 3' end formation to be a two-step process. Pre-mRNA was first cleaved endonucleolytically at a specific site that was followed by sequential addition of AMPs from ATP to the 3' hydroxyl group at the end of mRNA. The site of cleavage was specified by a conserved hexanucleotide, AAUAAA, from 10 to 30 nt upstream of this 3' end. Extensive purification of these two activities showed that more than 10 polypeptides were needed for mRNA 3' end formation. Most of these were in complexes involved in the cleavage step. Two of the best characterized are CstF and CPSF, while two other remain partially purified but essential. Oddly, the specific proteins involved in phosphodiester bond hydrolysis have yet to be identified. The polyadenylation step occurs within the complex of poly A polymerase and poly A-binding protein, PABII, that controls poly A length. That the cleavage complex, CPSF, is also required for this step attests to a tight coupling of the two steps of 3' and formation. The reaction reconstituted from these RNA-free purified factors correctly processes pre-mRNAs. Meaningful analysis of the role of poly A in mRNA metabolism or function was possible once quantities of these proteins most often over-expressed from cDNA clones became available. The large number needed for two simple reactions of an endonuclease, a polymerase and a sequence recognition factor, pointed to 3' end formation as a regulated process. Polyadenylation itself had appeared to require regulation in cases where two poly A sites were alternatively processed to produce mRNA coding for two different proteins. The 64-KDa subunit of CstF is now known to be a regulator of poly A site choice between two sites in the immunoglobulin heavy chain of B cells. In resting cells the site used favors the mRNA for a membrane-bound protein. Upon differentiation to plasma cells, an upstream site is used the produce a secreted form of the heavy chain. Poly A site choice in the calcitonin pre-mRNA involves splicing factors at a pseudo splice site in an intron downstream of the active poly site that interacts with cleavage factors for most tissues. The molecular basis for choice of the alternate site in neuronal tissue is unknown. Proteins needed for mRNA 3' end formation also participate in other RNA-processing reactions: cleavage factors bind to the C-terminal domain of RNA polymerase during transcription; splicing of 3' terminal exons is stimulated port of by cleavage factors that bind to splicing factors at 3' splice sites. nuclear ex mRNAs is linked to cleavage factors and requires the poly A II-binding protein. Most striking is the long-sought evidence for a role for poly A in translation in yeast where it provides the surface on which the poly A-binding protein assembles the factors needed for the initiation of translation. This adaptability of eukaryotic cells to use a sequence of low information content extends to bacteria where poly A serves as a site for assembly of an mRNA degradation complex in E. coli. Vaccinia virus creates mRNA poly A tails by a streamlined mechanism independent of cleavage that requires only two proteins that recognize unique poly A signals. Thus, in spite of 40 years of study of poly A sequences, this growing multiplicity of uses and even mechanisms of formation seem destined to continue.

Chimeric human CstF-77/Drosophila Suppressor of forked proteins rescue suppressor of forked mutant lethality and mRNA 3' end processing in Drosophila

The Suppressor of forked [Su(f)] protein is the Drosophila homologue of CstF-77, a subunit of human cleavage stimulation factor (CstF) that is required for the first step of the mRNA 3' end processing reaction in vitro. We have addressed directly the role of su(f) in the mRNA 3' end processing reaction in vivo. We show that su(f) is required for the cleavage of pre-mRNA during mRNA 3' end formation. Analysis of the functional complementation between Su(f) and CstF-77 shows that most of the Drosophila protein (85%) can be exchanged for the human protein to produce chimeric CstF-77/Su(f) proteins that rescue lethality and cleavage defect during mRNA 3' end formation in su(f) mutants. Interestingly, we show that a domain in human CstF-77 is limiting for the rescue and that this domain is not able to reproduce protein interactions with the CstF subunits of Drosophila. We also show that chimeric CstF-77/Su(f) proteins that rescue lethality of su(f) mutants cannot restore utilization of a regulated poly(A) site in Drosophila. Taken together, these results demonstrate that CstF-77 and Su(f) have the same function in mRNA 3' end formation in vivo, but that these two proteins are not interchangeable for regulation of poly(A) site utilization.

Identification and characterization of a Drosophila homologue of ATBP

ATBP [(A+T)-stretch binding protein] activates the Sarcophaga defense protein genes in an (A+T)-stretch dependent manner as shown previously using a luciferase reporter assay [Aozasa et al. (2001) Eur J Biochem, 268, 2506-2511]. We identified the Drosophila homologue of ATBP. Drosophila ATBP has 388 amino acid residues and the amino acid sequence has high similarity with that of the Sarcophaga ATBP. In particular, residues 1-56 and 301-374 are highly conserved between Sarcophaga and Drosophila. Drosophila ATBP activated the Sarcophaga lectin gene promoter in SL-2 cells. The Drosophila ATBP gene is a single copy gene and is located in the 20E region of chromosome X.

Characterization of the flamenco region of the Drosophila melanogaster genome

The flamenco gene, located at 20A1-3 in the beta-heterochromatin of the Drosophila X chromosome, is a major regulator of the gypsy/mdg4 endogenous retrovirus. As a first step to characterize this gene, approximately 100 kb of genomic DNA flanking a P-element-induced mutation of flamenco was isolated. This DNA is located in a sequencing gap of the Celera Genomics project, i.e., one of those parts of the genome in which the "shotgun" sequence could not be assembled, probably because it contains long stretches of repetitive DNA, especially on the proximal side of the P insertion point. Deficiency mapping indicated that sequences required for the normal flamenco function are located >130 kb proximal to the insertion site. The distal part of the cloned DNA does, nevertheless, contain several unique sequences, including at least four different transcription units. Dip1, the closest one to the P-element insertion point, might be a good candidate for a gypsy regulator, since it putatively encodes a nuclear protein containing two double-stranded RNA-binding domains. However, transgenes containing dip1 genomic DNA were not able to rescue flamenco mutant flies. The possible nature of the missing flamenco sequences is discussed.

Complex protein interactions within the human polyadenylation machinery identify a novel component

Polyadenylation of mRNA precursors is a two-step reaction requiring multiple protein factors. Cleavage stimulation factor (CstF) is a heterotrimer necessary for the first step, endonucleolytic cleavage, and it plays an important role in determining the efficiency of polyadenylation. Although a considerable amount is known about the RNA binding properties of CstF, the protein-protein interactions required for its assembly and function are poorly understood. We therefore first identified regions of the CstF subunits, CstF-77, CstF-64, and CstF-50, required for interaction with each other. Unexpectedly, small regions of two of the subunits participate in multiple interactions. In CstF-77, a proline-rich domain is necessary not only for binding both other subunits but also for self-association, an interaction consistent with genetic studies in Drosophila. In CstF-64, a small region, highly conserved in metazoa, is responsible for interactions with two proteins, CstF-77 and symplekin, a nuclear protein of previously unknown function. Intriguingly, symplekin has significant similarity to a yeast protein, PTA1, that is a component of the yeast polyadenylation machinery. We show that multiple factors, including CstF, cleavage-polyadenylation specificity factor, and symplekin, can be isolated from cells as part of a large complex. These and other data suggest that symplekin may function in assembly of the polyadenylation machinery.

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 poly-adenylation 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.

Role of the 3' untranslated region of baculovirus p10 mRNA in high-level expression of foreign genes

The p10 gene of Autographa californica nucleopolyhedrovirus has two putative AATAAA polyadenylation signals. The downstream signal is used predominantly, as was determined by analysing 3' cDNA ends. This downstream motif is followed by a GT-rich sequence, known to be important for efficient polyadenylation in mammalian systems. To analyse the importance of polyadenylation for p10 gene expression, recombinant viruses with altered 3' untranslated regions (UTRs) were tested using chloramphenicol acetyltransferase (CAT) as a reporter. Surprisingly, after inactivation of the downstream AATAAA motif, CAT expression remained at the same high level as observed with a wild-type 3' UTR. Polyadenylation occurred 24-28 nucleotides further downstream, probably due to an ATTAAA sequence motif. Replacing the p 10 3' UTR with the SV40 early terminator sequence as part of an hsp70-lacZ-SV40 gene cassette, which is commonly used in baculovirus expression vectors, resulted in a reduction in reporter gene expression. Polyadenylation occurred far more efficiently in the original p10 3' UTR than in the SV40 terminator sequence, as was shown by testing the SV40 terminator separately. These results indicate that in order to obtain high levels of foreign gene expression, vectors that provide a wild-type p10 3' UTR are to be preferred over those containing the hsp70-lacZ-SV40 gene cassette. Comparison of the p10 genes of various baculoviruses showed the presence of at least one AATAAA or ATTAAA motif in combination with a GT-rich sequence in the 3' UTR, suggesting an evolutionary conservation of these two elements, thereby maintaining the high level of p10 gene expression.

The suppressor of forked gene of Drosophila, which encodes a homologue of human CstF-77K involved in mRNA 3'-end processing, is required for progression through mitosis

The Suppressor of forked (Su(f)) protein of Drosophila melanogaster is a homologue of the 77K subunit of human cleavage stimulation factor required for cleavage of pre-mRNAs before addition of poly(A). We have previously shown that the Su(f) protein is not ubiquitously distributed: it accumulates in dividing cells at various stages of Drosophila development. In this paper, we show that phenotypes of su(f) temperature-sensitive mutants result from a defect in cell proliferation. Analysis of the mitotic phenotype of su(f) temperature-sensitive alleles in larval brain and in imaginal discs reveals an increase in the number of metaphases with overcondensed chromosomes and asymmetric or reduced mitotic spindles. In contrast, neural differentiation in eye imaginal discs of the same mutant flies does not appear to be affected. These results indicate that su(f) is required during cell division for progression through metaphase. Taken together, these data suggest that a decrease in su(f) activity preferentially affects 3'-end formation of particular mRNAs, some of which are involved in mitosis, and are in agreement with a role of su(f) in the regulation of poly(A) site utilization.

Down-regulation of RpS21, a putative translation initiation factor interacting with P40, produces viable minute imagos and larval lethality with overgrown hematopoietic organs and imaginal discs

Down-regulation of the Drosophila ribosomal protein S21 gene (rpS21) causes a dominant weak Minute phenotype and recessively produces massive hyperplasia of the hematopoietic organs and moderate overgrowth of the imaginal discs during larval development. Here, we show that the S21 protein (RpS21) is bound to native 40S ribosomal subunits in a salt-labile association and is absent from polysomes, indicating that it acts as a translation initiation factor rather than as a core ribosomal protein. RpS21 can interact strongly with P40, a ribosomal peripheral protein encoded by the stubarista (sta) gene. Genetic studies reveal that P40 underexpression drastically enhances imaginal disc overgrowth in rpS21-deficient larvae, whereas viable combinations between rpS21 and sta affect the morphology of bristles, antennae, and aristae. These data demonstrate a strong interaction between components of the translation machinery and showed that their underexpression impairs the control of cell proliferation in both hematopoietic organs and imaginal discs.

Levels of polyadenylation factor CstF-64 control IgM heavy chain mRNA accumulation and other events associated with B cell differentiation

Cleavage stimulation factor (CstF) is one of the multiple factors required for mRNA polyadenylation. The concentration of one CstF subunit (CstF-64) increases during activation of B cells, and this is sufficient to switch IgM heavy chain mRNA expression from membrane-bound form to secreted form. To extend this observation, we disrupted the endogenous CstF-64 gene in the B cell line DT40 and replaced it with a regulatable transgene. Strikingly, a 10-fold decrease in CstF-64 concentration did not markedly affect cell growth but specifically and dramatically reduced accumulation of IgM heavy chain mRNA. Further reduction caused reversible cell cycle arrest in G0/G1 phase, while depletion resulted in apoptotic cell death. Our results indicate that CstF-64 plays unexpected roles in regulating gene expression and cell growth in B cells.

Autoregulation at the level of mRNA 3' end formation of the suppressor of forked gene of Drosophila melanogaster is conserved in Drosophila virilis

The Drosophila melanogaster Suppressor of forked [Su(f)] protein shares homology with the yeast RNA14 protein and the 77-kDa subunit of human cleavage stimulation factor, which are proteins involved in mRNA 3' end formation. This suggests a role for Su(f) in mRNA 3' end formation in Drosophila. The su(f) gene produces three transcripts; two of them are polyadenylated at the end of the transcription unit, and one is a truncated transcript, polyadenylated in intron 4. Using temperature-sensitive su(f) mutants, we show that accumulation of the truncated transcript requires wild-type Su(f) protein. This suggests that the Su(f) protein autoregulates negatively its accumulation by stimulating 3' end formation of the truncated su(f) RNA. Cloning of su(f) from Drosophila virilis and analysis of its RNA profile suggest that su(f) autoregulation is conserved in this species. Sequence comparison between su(f) from both species allows us to point out three conserved regions in intron 4 downstream of the truncated RNA poly(A) site. These conserved regions include the GU-rich downstream sequence involved in poly(A) site definition. Using transgenes truncated within intron 4, we show that sequence up to the conserved GU-rich domain is sufficient for production of the truncated RNA and for regulation of this production by su(f). Our results indicate a role of su(f) in the regulation of poly(A) site utilization and an important role of the GU-rich sequence for this regulation to occur.

Gene clusters and polycistronic transcription in eukaryotes

Sometimes genes are arranged nonrandomly on the chromosomes of eukaryotes. This review considers instances of gene clusters in which two genes or more are expressed from a single promoter. This includes cases in which a polycistronic pre-mRNA is processed to make monocistronic mRNAs in nematodes, as well as isolated examples of polycistronic mRNAs found in mammals, flies, and perhaps plants.

The suppressor of forked protein of Drosophila, a homologue of the human 77K protein required for mRNA 3'-end formation, accumulates in mitotically-active cells

The suppressor of forked (Su(f)) protein of Drosophila melanogaster is highly homologous to two proteins involved in mRNA 3'-end formation, the yeast RNA14 protein and the 77K subunit of human cleavage stimulation factor (CstF). This suggests a role for su(f) in mRNA 3'-end-processing, probably as part of Drosophila CstF. We have investigated the expression pattern of su(f) during Drosophila development and found that the su(f) gene product is not detected ubiquitously. The Su(f) protein accumulates in mitotically-active cells, but does not in non-dividing cells. This expression pattern corroborates earlier data suggesting that the phenotypes of su(f) mutants could result from a defect in cell proliferation. Our results suggest that, in Drosophila, Su(f) is involved in the regulatory function of CstF.

A 1.5 kb repeat sequence flanks the suppressor of forked gene at the euchromatin-heterochromatin boundary of the Drosophila melanogaster X chromosome

A 1.5 kilobasepair repeated DNA sequence is duplicated in direct orientation so as to flank the suppressor of forked gene in the euchromatin-heterochromatin transition region on the X chromosome of Drosophila melanogaster. These two copies are almost identical, but DNA blotting, analysis of cloned sequences and database searches show that elsewhere in the genome, homologous sequences are poorly conserved. They are often associated with other repeats, suggesting that they may belong to a scrambled and clustered middle repetitive DNA family. The sequences do not appear to be related to transposable elements and their location in different strains is conserved. In situ hybridization to metaphase chromosomes shows that homologous sequences are concentrated in the pericentric regions of the autosomes and the X chromosome. The sequences are not significantly under-represented in DNA from polytene tissue and must lie in the replicated regions of polytene chromosomes. The almost perfect conservation of the two repeats around suppressor of forked in D. melanogaster suggests they arose by duplication or gene conversion. Suppression of recombination in this chromosomal region presumably allows this unusual organization to be stably maintained. In the X-ray induced allele, suppressor of forked-L26, the sequence between the repeats, including the gene, and one copy of the repeat have been deleted.

The polyadenylation factor CstF-64 regulates alternative processing of IgM heavy chain pre-mRNA during B cell differentiation

The switch from membrane-bound to secreted-form IgM that occurs during differentiation of B lymphocytes has long been known to involve regulated processing of the heavy chain pre-mRNA. Here, we show that accumulation of one subunit of an essential polyadenylation factor (CstF-64) is specifically repressed in mouse primary B cells and that overexpression of CstF-64 is sufficient to switch heavy chain expression from membrane-bound (microm) to secreted form (micros). We further show that CstF-64 is limiting for formation of intact CstF, that CstF has a higher affinity for the microm poly(A) site than for the micros site, and that the microm site is stronger in a reconstituted in vitro processing reaction. Our results indicate that CstF-64 plays a key role in regulating IgM heavy chain expression during B cell differentiation.

Complex alternative RNA processing generates an unexpected diversity of poly(A) polymerase isoforms

Multiple forms of poly(A) polymerase (PAPs I, II, and III) cDNA have previously been isolated from bovine, human, and/or frog cDNA libraries. PAPs I and II are long forms of the enzyme that contain four functional domains: an apparent ribonucleoprotein-type RNA-binding domain, a catalytic region that may be related to the polymerase module, two nuclear localization signals (NLSs I and 2), and a C-terminal Ser/Thr-rich region. PAP III would encode a truncated protein that lacks the NLSs and the S/T-rich region. To investigate further the structure and expression of these forms, we isolated the mouse PAP gene and an intronless pseudogene from a mouse liver genomic library. The structure of the gene indicates that different forms of PAP are produced by alternative splicing (PAPs I and II) or by competition between polyadenylation and splicing (PAP III). The pseudogene appears to reflect yet another form of long PAP, which we call PAP IV. Mouse PAP III and two additional truncated forms, PAPs V and VI, which would be produced by use of poly(A) sites in adjacent introns, were also isolated from a mouse brain cDNA library. RNase protection and reverse transcription-PCR analyses showed that PAP II, V, and VI are expressed in all tissues tested but that PAP I and/or IV and III are tissue specific. However, immunoblot analysis detected only the long forms, raising the possibility that the short-form RNAs are not translated. Purified recombinant baculovirus-expressed PAPs were tested in several in vitro assays, and the short forms were found to be inactive. We discuss the possible significance of this complex expression pattern.

Elimination of introns at the Drosophila suppressor-of-forked locus by P-element-mediated gene conversion shows that an RNA lacking a stop codon is dispensable

The suppressor of forked [su(f)] locus affects the phenotype of mutations caused by transposable element insertions at unlinked loci. It encodes a putative 84-kD protein with homology to two proteins involved in mRNA 3' end processing; the product of the yeast RNA14 gene and the 77-kD subunit of human cleavage stimulation factor. Three su(f) mRNAs are produced by alternative polyadenylation. The 2.6- and 2.9-kb mRNAs encode the same 84-kD protein while a 1.3-kb RNA, which terminates within the fourth intron, is unusual in having no stop codon. Using P-element-mediated gene replacement we have copied sequences from a transformation construct into the su(f) gene creating a su(f) allele at the normal genomic location that lacks the first five introns. This allele is viable and appears wild type for su(f) function, demonstrating that the 1.3-kb RNA and the sequences contained within the deleted introns are dispensable for su(f) function. Compared with studies on gene replacement at the white locus, chromosomal breaks at su(f) appear to be less efficiently repaired from ectopic sites, perhaps because of the location of su(f) at the euchromatin/heterochromatin boundary on the X chromosome.

Structure and expression of wild-type and suppressible alleles of the Drosophila purple gene

Viable mutant alleles of purple (pr), such as prbw, exhibit mutant eye colors. This reflects low 6-pyruvoyl tetrahydropterin (PTP) synthase activity required for pigment synthesis. PTP synthase is also required for synthesis of the enzyme cofactor biopterin; presumably this is why some pr alleles are lethal. The prbw eye color phenotype is suppressed by suppressor of sable [su(s)] mutations. The pr gene was cloned to explore the mechanism of this suppression. pr produces two PTP synthase mRNAs: one constitutively from a distal promoter and one in late pupae and young adult heads from a proximal promoter. The latter presumably supports eye pigment synthesis. The prbw allele has a 412 retrotransposon in an intron spliced from both mRNAs. However, the head-specific mRNA is reduced > 10-fold in prbw and is restored by a su(s) mutation, while the constitutive transcript is barely affected. The Su(s) protein probably alters processing of RNA containing 412. Because the intron containing 412 is the first in the head-specific mRNA and the second in the constitutive mRNA, binding of splicing machinery to nascent transcripts before the 412 insertion is transcribed may preclude the effects of Su(s) protein.

Effects of mutations in the Saccharomyces cerevisiae RNA14, RNA15, and PAP1 genes on polyadenylation in vivo

The RNA14 and RNA15 gene products have been implicated in a variety of cellular processes. Mutations in these genes lead to faster decay of some mRNAs and yield extracts that are deficient in cleavage and polyadenylation in vitro. These results suggest that the RNA14 and RNA15 gene products may be involved in both adenylation and deadenylation in vivo. To explore the roles of these gene products in vivo, we examined the site of adenylation and the rate of deadenylation for individual mRNAs in rna14 and rna15 mutant strains. We observed that the rates of deadenylation are not affected by lesions in either the RNA14 or the RNA15 gene. This result suggests that the proteins encoded by these genes are not involved in regulation of the deadenylation rate. In contrast, we observed that the site of adenylation for the ACT1 transcript can be altered in these mutants. Interestingly, we also observed that mutation of the poly(A) polymerase gene altered the site of ACT1 polyadenylation. These observations suggest that the RNA14, RNA15, and PAP1 proteins are involved in poly(A) site choice. This alteration in poly(A) site choice in the rna14 mutant can be corrected by the ssm4 suppressor, indicating that this suppression acts at the level of polyadenylation and not by slowing mRNA degradation.

Genetic enhancement of RNA-processing defects by a dominant mutation in B52, the Drosophila gene for an SR protein splicing factor

SR proteins are essential for pre-mRNA splicing in vitro, act early in the splicing pathway, and can influence alternative splice site choice. Here we describe the isolation of both dominant and loss-of-function alleles of B52, the gene for a Drosophila SR protein. The allele B52ED was identified as a dominant second-site enhancer of white-apricot (wa), a retrotransposon insertion in the second intron of the eye pigmentation gene white with a complex RNA-processing defect. B52ED also exaggerates the mutant phenotype of a distinct white allele carrying a 5' splice site mutation (wDR18), and alters the pattern of sex-specific splicing at doublesex under sensitized conditions, so that the male-specific splice is favored. In addition to being a dominant enhancer of these RNA-processing defects, B52ED is a recessive lethal allele that fails to complement other lethal alleles of B52. Comparison of B52ED with the B52+ allele from which it was derived revealed a single change in a conserved amino acid in the beta 4 strand of the first RNA-binding domain of B52, which suggests that altered RNA binding is responsible for the dominant phenotype. Reversion of the B52ED dominant allele with X rays led to the isolation of a B52 null allele. Together, these results indicate a critical role for the SR protein B52 in pre-mRNA splicing in vivo.

The 160-kD subunit of human cleavage-polyadenylation specificity factor coordinates pre-mRNA 3'-end formation

Cleavage-polyadenylation specificity factor (CPSF) is a multisubunit protein that plays a central role in 3' processing of mammalian pre-mRNAs. CPSF recognizes the AAUAAA signal in the pre-mRNA and interacts with other proteins to facilitate both RNA cleavage and poly(A) synthesis. Here we describe the isolation of cDNAs encoding the largest subunit of CPSF (160K) as well as characterization of the protein product. Antibodies raised against the recombinant protein inhibit polyadenylation in vitro, which can be restored by purified CPSF. Extending previous studies, which suggested that 160K contacts the pre-mRNA, we show that purified recombinant 160K can, by itself, bind preferentially to AAUAAA-containing RNAs. While the sequence of 160K reveals similarities to the RNP1 and RNP2 motifs found in many RNA-binding proteins, no clear match to a known RNA-binding domain was found, and RNA recognition is therefore likely mediated by a highly diverged or novel structure. We also show that 160K binds specifically to both the 77K (suppressor of forked) subunit of the cleavage factor CstF and to poly(A) polymerase (PAP). These results provide explanations for previously observed cooperative interactions between CPSF and CstF, which are responsible for poly(A) site specification, and between CPSF and PAP, which are necessary for synthesis of the poly(A) tail. Also supporting a direct role for 160K in these interactions is the fact that 160K by itself retains partial ability to cooperate with CstF in binding pre-mRNA and, unexpectedly, inhibits PAP activity in in vitro assays. We discuss the significance of these multiple functions and also a possible evolutionary link between yeast and mammalian polyadenylation suggested by the properties and sequence of 160K.

mRNA stability in mammalian cells

This review concerns how cytoplasmic mRNA half-lives are regulated and how mRNA decay rates influence gene expression. mRNA stability influences gene expression in virtually all organisms, from bacteria to mammals, and the abundance of a particular mRNA can fluctuate manyfold following a change in the mRNA half-life, without any change in transcription. The processes that regulate mRNA half-lives can, in turn, affect how cells grow, differentiate, and respond to their environment. Three major questions are addressed. Which sequences in mRNAs determine their half-lives? Which enzymes degrade mRNAs? Which (trans-acting) factors regulate mRNA stability, and how do they function? The following specific topics are discussed: techniques for measuring eukaryotic mRNA stability and for calculating decay constants, mRNA decay pathways, mRNases, proteins that bind to sequences shared among many mRNAs [like poly(A)- and AU-rich-binding proteins] and proteins that bind to specific mRNAs (like the c-myc coding-region determinant-binding protein), how environmental factors like hormones and growth factors affect mRNA stability, and how translation and mRNA stability are linked. Some perspectives and predictions for future research directions are summarized at the end.

Effects of a transposable element insertion on alcohol dehydrogenase expression in Drosophila melanogaster

Variation in the DNA sequence and level of alcohol dehydrogenase (Adh) gene expression in Drosophila melanogaster have been studied to determine what types of DNA polymorphisms contribute to phenotypic variation in natural populations. The Adh gene, like many others, shows a high level of variability in both DNA sequence and quantitative level of expression. A number of transposable element insertions occur in the Adh region and one of these, a copia insertion in the 5' flanking region, is associated with unusually low Adh expression. To determine whether this insertion (called R142) causes the low expression level, the insertion was excised from the cloned R142 Adh gene and the effect was assessed by P-element transformation. Removal of this insertion causes a threefold increase in the level of ADH, clearly showing that it contributes to the naturally occurring variation in expression at this locus. Removal of all but one LTR also causes a threefold increase, indicating that the mechanism is not a simple sequence disruption. Furthermore, this copia insertion, which is located between the two Adh promoters and their upstream enhancer sequences, has differential effects on the levels of proximal and distal transcripts. Finally, a test for the possible modifying effects of two suppressor loci, su(wa) and su(f), on this insertional mutation was negative, in contrast to a previous report in the literature.

A polyadenylation factor subunit is the human homologue of the Drosophila suppressor of forked protein

Polyadenylation of messenger RNA precursors is a complex process that requires multiple protein factors (for reviews, see refs 1, 2). Cleavage stimulation factor (CstF) is one of these, functioning together with cleavage-polyadenylation specificity factor, two cleavage factors, and poly(A)+ polymerase. CstF is composed of three subunits of M(r) 77, 64 and 50K. The 64K and 50K subunits contain, respectively, an RNP-type RNA-binding domain that contacts the pre-mRNA and transducin repeats characteristic of G-protein beta-subunits. Here we report the cloning and characterization of the 77K subunit of human CstF (referred to as 77K). We show that the 77K subunit is required for formation of active CstF and bridges the 64K and 50K subunits. Sequence analyses indicate that the 77K subunit is the homologue of the protein encoded by the Drosophila melanogaster suppressor of forked (su(f)) gene. Mutations in su(f) can enhance or suppress the effects of transposable element insertions, and our data indicate that this is due to changes in polyadenylation. Both the 77K subunit and the su(f) protein share homology with Saccharomyces cerevisiae RNA14, previously shown to be involved in mRNA metabolism. Our results thus also indicate that components of the complex polyadenylation machinery are conserved from yeast to man.

A rosy future for heterochromatin

The demonstration by Zhang and Spradling (1) of efficient P element transposition into heterochromatic regions will aid ongoing studies of heterochromatin structure and function. P element insertions will provide entry points for further molecular analysis of heterochromatin and will allow the isolation of small and large heterochromatic deficiencies. The generation of heterochromatic P insertions also will aid the study of heterochromatic genes. Of the heterochromatic insertions isolated by Zhang and Spradling, five were homozygous lethal, and one of these defined a lethal locus not previously uncovered by heterochromatic deficiencies. P elements have previously been used to mutagenize and clone specific heterochromatic genes (14, 19, 26). New methods, like those described here (1, 32), should allow the efficient identification and molecular isolation of other single-copy heterochromatic genes. Furthermore, since position-effect suppression allowed the recovery of heterochromatic P insertions, it may also allow the recovery of insertions in euchromatic regions previously refractory to P mutagenesis. Studies of position-effect variegation show that genes normally found in heterochromatin require a heterochromatic context for normal expression and that heterochromatin is inhibitory to euchromatic gene expression (16). The physical basis of these related phenomena--chromatin assembly, nuclear positioning, and/or heterochromatin elimination--can be resolved only with a more thorough understanding of heterochromatin structure and functions. Analyzing heterochromatin also will help define the chromosomal components responsible for inheritance processes such as chromosome pairing, sister chromatid adhesion, and centromere function. These efforts will be facilitated by the effective use of P elements combined with other current molecular-genetic approaches.

Insertional mutagenesis of Drosophila heterochromatin with single P elements

Insertional mutagenesis with transposable P elements has greatly facilitated the identification and analysis of genes located throughout the 70% of the Drosophila melanogaster genome classified as euchromatin. In contrast, genetically marked P elements have only rarely been shown to transpose into heterochromatin. By carrying out single P element insertional mutagenesis under conditions where position-effect variegation was suppressed, we efficiently generated strains containing insertions at diverse sites within centromeric and Y-chromosome heterochromatin. The tendency of P elements to transpose locally was shown to operate within heterochromatin, and it further enhanced the recovery of heterochromatic insertions. Three of the insertions disrupted vital genes known to be present at low density in heterochromatin. Strains containing single P element insertions will greatly facilitate the structural and functional analysis of this poorly understood genomic component.

Cellular localization of RNA14p and RNA15p, two yeast proteins involved in mRNA stability

RNA14 and RNA15 were originally identified by temperature-sensitive mutations that cause a rapid decrease in poly(A)-tail length and overall mRNA levels at the restrictive temperature. We have raised antibodies to the RNA14 and RNA15 proteins, and used subcellular fractionation and immunofluorescence to localize these proteins within the yeast cell. RNA14p is a 73 kDa protein found in both the nucleus and the cytoplasm, whilst RNA15p is a 42 kDa protein detected only in the nucleus. The observation that both proteins are found in the nucleus is in agreement with previous genetic data which suggest an interaction between RNA14p and RNA15p. Also the joint nuclear localization is consistent with the biochemical data suggesting a role in polyadenylation. The detection of significant amounts of RNA14p in the cytoplasm opens the possibility of a second function for this protein, either in cytoplasmic regulation of mRNA deadenylation or, more interestingly, in mRNA stability.

The Drosophila forked gene encodes two major RNAs, which, in gypsy or springer insertion mutants, are partially or completely truncated within the 5'-LTR of the inserted retrotransposon

Mutations in the forked (f) gene of Drosophila cause deformation of bristles and hairs. Our molecular analysis showed the f gene to span more than 30 kb, and to encode two major RNAs, 6.0 and 2.5 kb long, both of which are prematurely terminated in gypsy and springer insertion mutants. These truncated RNAs were polyadenylated using putative polyadenylation signals within the 5'-LTR of the inserted retrotransposon. No evidence was found for effects of the retrotransposon insertions on the promoters for transcription of the 6.0 and 2.5 kb RNAs. In f1 and fx, a single gypsy element was found to be inserted at identical sites in the second intron of region encoding the 2.5 kb f RNA and both truncated and wild-type sized RNAs were detected. Recessive mutations at suppressor of forked (su(f)) increased the fraction of wild-type sized RNAs considerably, suggesting that the wild-type su(f) product either stimulates premature termination at the gypsy LTR or inhibits normal splicing. In f36a, a springer element inserted in the third exon of the region encoding the 2.5 kb f RNA completely suppressed the formation of apparently wild-type transcripts.