| Nucleic Acids Research | Pages |
Auxiliary downstream elements are required for efficient polyadenylation of mammalian pre-mRNAs
Introduction
Materials And Methods
Plasmids and transcripts
In vitro cleavage assays
UV cross-linking/label transfer analysis and immunoprecipitation
RNase T1 ladders
Expression and purification of recombinant GST-R17 coat protein fusion from Escherichia coli
Results
Mammalian polyadenylation signals contain AUX DSEs that affect the efficiency of 3[prime]-end processing
AUX DSEs may promote the efficiency of 3[prime]-end processing by maintaining the core elements in an unstructured domain
AUX DSEs may promote the efficiency of 3[prime]-end processing by forming a stable secondary structure that helps focus binding of CstF to the core downstream URE
Only selected proteins can stimulate 3[prime]-end processing by binding to auxiliary downstream sequences
Discussion
Acknowledgements
References
Auxiliary downstream elements are required for efficient polyadenylation of mammalian pre-mRNAs
ABSTRACT
INTRODUCTION
Maturation of the 3[prime]-end of most mammalian pre-mRNAs involves two coupled reactions that occur in the nucleus: a site-specific endonucleolytic cleavage event followed by addition of 150-200 adenylate residues to the newly formed 3[prime]-end in a template-independent manner (1,2). In addition to its general role in mRNA formation, 3[prime]-end processing is essential for transcription termination by RNA polymerase II (3,4). Moreover, polyadenylation signals can influence the efficiency of pre-mRNA splicing (5,6) and may affect gene expression through the inclusion or exclusion of a specific exon(s) in a tissue-specific or a development-specific manner (7-9). A thorough understanding of the process and regulation of 3[prime]-end formation, therefore, should give insights into the interplay of several areas of gene expression.
Two sequence elements, located ~10-25 bases upstream or downstream of the cleavage site, form the core elements of the polyadenylation signal. The core upstream AAUAAA element is highly conserved among mammalian polyadenylation signals and serves as the binding site for the 160 kDa protein of a general polyadenylation factor, cleavage polyadenylation specificity factor (CPSF) (10,11). The core downstream element, a four out of five base uridylate-rich tract in most mammalian polyadenylation signals (12,13), serves as the binding site for the 64 kDa protein of another general polyadenylation factor, cleavage stimulation factor (CstF) (14). Both core elements are required for efficient 3[prime]-end processing as well as to determine the site of cleavage (12,14). Moreover, it has been shown that the 160 kDa protein CPSF interacts with the 77 kDa subunit of CstF as well as poly(A) polymerase (11). Therefore, it appears that CPSF and CstF form a scaffold over the cleavage site which allows assembly of a core polyadenylation complex that facilitates efficient and accurate 3[prime]-end processing (14-17).
It is likely that sequences besides the core elements influence 3[prime]-end processing efficiency. In fact, we have noted that most regulated polyadenylation signals described to date have variant core upstream or downstream sequences that differ from the established consensus sequences. It is possible that additional sequence elements may assist in the assembly of the general processing factors on some polyadenylation signals. The identification of these putative auxiliary sequence elements, therefore, may be an important step in determining the underlying mechanisms of polyadenylation signal strength and site selection.
Numerous laboratories have previously demonstrated auxiliary elements located upstream of the core AAUAAA element that are equally active both in vivo and in unfractionated in vitro systems. Within the adenovirus major late transcription unit, auxiliary upstream sequences enhance processing of the L1 polyadenylation signal by promoting the binding of CPSF to this signal (18). Upstream sequences in the U3 region of most retroviral polyadenylation signals, including human immunodeficiency virus type 1 (HIV-1), are required for efficient 3[prime]-end processing (19,20). Moreover, the HIV-1 upstream element promotes binding of CPSF to the polyadenylation signal as well as minimizes secondary structure in the core element region to enhance processing efficiency (21-24). Finally, an upstream efficiency element has been identified in the SV40 late polyadenylation signal (25). The direct interaction of the upstream efficiency element with the U1 snRNP-A protein and the interaction between the U1 snRNP A protein and the 160 kDa protein CPSF may play a role in processing of this signal (26,27).
In contrast to the data for auxiliary upstream elements, only a single auxiliary downstream element (AUX DSE) has been previously characterized: a G-rich sequence (GRS) associated with the SV40 late (SVL) polyadenylation signal (28,29). The GRS element is located on the 3[prime]-side of the core downstream U-rich element (URE) and influences the efficiency of 3[prime]-end processing. The sequence and relative position of the GRS element affect its ability to function as a positive mediator of polyadenylation. Moreover, the GRS element serves as the binding site for a 50 kDa cellular protein called downstream element factor-1 (DSEF-1). We have suggested that the GRS element may promote processing efficiency by stabilizing assembly of the general polyadenylation factors on the core elements through DSEF-1 interactions (28). Whether other mammalian polyadenylation signals have AUX DSEs, however, has not been investigated, primarily due to the difficulty in identifying a true consensus for the core downstream element. Curiously, in vivo observations using polyadenylation signals containing deletions in their downstream regions have suggested that the downstream element region is large and rather diffuse (see for example 30,31). Since we have recently identified the core downstream element as an appropriately positioned four out of five base U-rich tract (12), this allows us to investigate the general significance of sequences 3[prime] of the core URE. Specifically, we addressed how AUX DSEs influence processing efficiency in mammalian polyadenylation signals using the well-established in vitro 3[prime]-end processing system derived from nuclear extracts. This in vitro system has been shown previously to faithfully reproduce in vivo observations on the influence of downstream sequences on 3[prime]-end processing efficiency (31).
In this study, we have identified novel AUX DSEs that promote the efficiency of 3[prime]-end processing in several mammalian and viral polyadenylation signals. In addition to serving as sites for specific AUX DSE binding proteins, we demonstrate that AUX DSEs may promote efficient 3[prime]-end formation by two other mechanisms. First, AUX DSEs may promote efficient processing by maintaining the core elements in a domain that contains little secondary structure, leaving the AAUAAA and URE non-base paired. This lack of structure may allow the core polyadenylation complex to assemble on the signal more efficiently. Second, AUX DSEs may promote efficient processing by forming a stable structure that helps focus binding of CstF to the core downstream element. CstF does not bind strongly to the core downstream element on its own. Perhaps adjacent structures may help stabilize the interaction of CstF with the URE by removing nearby single-strand regions that may compete with the core downstream element for CstF binding. These data suggest that AUX DSEs play a general role in efficient 3[prime]-end processing of mammalian polyadenylation signals.
MATERIALS AND METHODS
Plasmids and transcripts
All RNAs were transcribed in vitro using SP6 RNA polymerase and [32P]UTP. RNAs were purified from 5% acrylamide gels containing 7 M urea prior to use.
pAAV contains a 232 bp PstI-XbaI fragment containing the polyadenylation signal of adeno-associated virus inserted between the PstI-XbaI sites of pGEM 3 (Promega). Transcription of EcoRI-linearized template yielded a 286 base RNA (AAV-wt). Templates to generate AAV-mt, AAV-cAAUAAA, AAV-cCLEAV and AAV-cURE RNAs were prepared by PCR using pAAV, an SP6 promoter-specific primer (5[prime]-CATACGATTTAGGTGACACTATAG-3[prime]) and one of the following primers: AAV-mt, 5[prime]-CGACGATATCTTTAAACCCGGGCAGCTGAAGCTTGCATGCCTGCAAGAAAGAAATACGCAGAGA-3[prime]; AAV-cAAUAAA, 5[prime]-GTTAATCAATAAACCGTTTATCTACGTAGCCATGGAAACTAGAT-3[prime]; AAV-cCLEAV, 5[prime]-TTTCAGTTGAACTTTGGTCTTCTACGTAGCCATGGAAACTAGAT-3[prime]; AAV-cURE, 5[prime]-TGCGTATTTCTTTCTTATCTTCTACGTAGCCATGGAAACT-3[prime]. Amplification reactions were performed in a total volume of 100 µl using standard reaction mixtures for 35 cycles of 94 (1 min), 49 (1 min) and 72°C (1 min). PCR products were purified using centricon-100 filters (Amicon) prior to use. Transcription of these templates yielded 283 base RNAs.
pAdL3 contains a 252 bp KpnI-DraI fragment containing the L3 polyadenylation signal of adenovirus type 5 inserted between the KpnI and HincII sites of pGEM 4 (Promega) (32). Transcription of HindIII-cleaved templates yielded a 298 base RNA (AdL3-wt). DNA templates for all AdL3 RNA derivatives were generated by PCR using pAdL3, the SP6 promoter-specific primer and one of the primers listed below: AdL3-mt, 5[prime]-GGGTCGACGATATCTTTAAACCCGGGCAGCTGAAGCTTGCATGCCTGCAAATAATCACCC-3[prime]; AdL3-PSN, 5[prime]-TTTATCCCCCTTTTGAGTGGGGCTAGGGGACTGCCCCTTTACAAGAAATAATCACCCGAGAGTGTAC-3[prime]; AdL3-PSN mt, 5[prime]-TTTATGGGCCTTTTGAGTGGGGCTAGGGGACTGCCCCTTTACAAGAAATAATCACCCGAGAGTGTAC-3[prime]; AdL3-GRS, 5[prime]-TTTTGAGTGGCCTCCCACACCTCCCCCTTTACAAGAAATAATCACCCGAG-3[prime]; AdL3-R17, 5[prime]-TTTATGGGCCTTTTGAGTACATGGGTAATCCTCATGTTTTACAAGAAATAATCACCCGAG-3[prime]. Transcription of purified PCR products yielded 298 (AdL3-mt), 295 (AdL3-PSN, AdL3-PSN mt and AdL3-R17) and 285 base (AdL3-GRS) RNAs.
pGEM2µMPA contains a 199 bp RsaI fragment containing the µ heavy chain membrane-associated polyadenylation signal inserted into the SmaI site of pGEM 2 (Promega) (33). Templates for the wild-type and mutant immunoglobulin µM RNAs were generated by a PCR approach using pGEM2µMPA, the SP6 promoter-specific primer and either a wild-type (5[prime]-TATTCAGATAATCCTTGGGCTGTC-3[prime]) or a mutant (5[prime]-GGGTCGACGATATCTTTAAACCCGGGCAGCTGAAGCTTGCATGCAAAAACAAAACAAATTCC-3[prime]) second primer. Transcription of both µM templates yielded 184 base RNAs (µM-wt and µM-mt).
In vitro cleavage assays
Cleavage reactions were performed using equimolar amounts (15 fmol) of the indicated transcripts in the in vitro system of Moore and Sharp (34) as described previously (32). Nuclear extracts were prepared from HeLa spinner cells grown in 10% horse serum as described (36). A typical 10 µl reaction contained a final concentration of 4% polyvinyl alcohol, 1 mM [alpha],[beta]-methylene ATP, 1 mM EDTA, 12 mM HEPES (pH 7.9), 12% glycerol, 60 mM KCl, 0.3 mM DTT and 50 or 60% (v/v) HeLa nuclear extract. To test the effect of the binding site for the R17 coat protein on 3[prime]-end processing efficiency of the AdL3 polyadenylation signal, 1 µg GST protein or GST-R17 coat protein fusion was added to the standard in vitro cleavage reaction mixture. RNA products were analyzed on a 5% acrylamide gel containing 7 M urea.
UV cross-linking/label transfer analysis and immunoprecipitation
RNA-protein interactions were analyzed by UV cross-linking/label transfer analysis and immunoprecipitation as described previously (32). Briefly, RNAs were incubated in the in vitro cleavage system for 15 min, transferred into a 96-well culture dish and irradiated for 10 min at 4°C with a germical light (Sylvania G15T8) placed 4 cm from the sample. RNase A was added to a final concentration of 1 µg/µl and samples were incubated at 37°C for 15 min. Cross-linked CstF 64 kDa protein was isolated by immunoprecipitation using the 64 kDa specific monoclonal antibody 3A7 (36) and analyzed on a 10% SDS-polyacrylamide gel.
RNase T1 ladders
AAV-wt, AAV-cAAUAAA, AAV-cCLEAV and AAV-cURE RNAs were transcribed in vitro using SP6 RNA polymerase, labeled at their 3[prime]-ends with pCp and FPLCpure[trade] RNA ligase (Promega) (37) and purified from a 5% acrylamide gel containing 7 M urea prior to use. Equimolar amounts (1.5 fmol) of gel-purified RNAs were digested with 0.5 U RNase T1 at 30°C for 3 min under buffer conditions identical to those used in the in vitro cleavage reaction (without nuclear extract). RNA products were analyzed on a 10% acrylamide gel containing 7 M urea.
Expression and purification of recombinant GST-R17 coat protein fusion from Escherichia coli
A template to generate a GST-R17 coat protein fusion was constructed using a PCR approach. Amplification reactions were carried out in a total volume of 100 µl using standard reaction mixtures for 35 cycles of 94 (1 min), 60 (1 min) and 72°C (1 min). A 420 bp fragment was amplified from the plasmid pTCT (a kind gift from O.Uhlenbeck) (38) using a primer (5[prime]-CCCGGGGGATCCATGGCTTCTAACTTTACT-3[prime]) containing a BamHI site at its 5[prime]-end and a primer (5[prime]-CCGCGGCTCGAGCTATTAGTAGATGCCGGA-3[prime]) bearing an XhoI site at its 5[prime]-end. The amplified fragment was digested with BamHI and XhoI, purified from a 2% low melting point agarose gel and inserted into the BamHI and XhoI sites of pGEX2TZQ [a pGEX2T (Pharmacia) derivative containing the sequence (5[prime]-GATCCCTCGGGTCGACGGTACCTCGAG-3[prime]) inserted between the BamHI and EcoRI sites (39)]. GST-R17 coat protein fusion expression was induced in transformed HB101 cells by IPTG and fusion proteins were purified by passing the cell sonicate through a glutathione-Sepharose affinity column according to the manufacturer's recommendations (Pharmacia). The concentration of the GST-R17 coat protein fusion was measured using the Bio-Rad protein assay (Bio-Rad).
RESULTS
Mammalian polyadenylation signals contain AUX DSEs that affect the efficiency of 3[prime]-end processing
We havepreviously identified a 14 base G-rich sequence as an AUX DSE in the SVL polyadenylation signal that affects the efficiency of 3[prime]-end processing (28,29). Whether the SVL polyadenylation signal is unique in its requirement for an AUX DSE for efficient processing or whether AUX DSEs are generally present in mammalian polyadenylation signals is unclear. Our recent description of the core downstream element as a four out of five base U-rich tract (12,13) allows us now to delineate regions where AUX DSEs may lie and examine their overall significance to the process of 3[prime]-end formation.
The cellular immunoglobulin µM polyadenylation signal and two viral polyadenylation signals [adenovirus type 5 major late transcription unit L3 (AdL3) and adeno-associated virus (AAV)] were selected to test whether we could identify AUX DSEs in other polyadenylation signals. For each of these signals, 30-40 base sequences 3[prime] of the core URE were substituted with several different polylinker sequences (Fig.
| Figure 1. Sequences in the auxiliary downstream regions of many mammalian polyadenylation signals influence the efficiency of 3[prime]-end processing. (A) Diagramatic representation of three wild-type and mutant RNAs. Sequences 3[prime] of the core downstream URE of the adenovirus L3 (AdL3), AAV and immunoglobulin µM polyadenylation signals were substituted with polylinker sequences. Numbers below the wild-type polyadenylation substrates refer to the positions of the URE and the 3[prime]-end of the transcript relative to the cleavage site. Stippled patterns downstream of the URE in mutant constructs refer to polylinker sequences. These sequences and other aspects of the constructs are described further in Materials and Methods. (B) Cleavage assays. RNAs were incubated in the in vitro cleavage system to assess the effect of substitutions in the auxiliary downstream region of polyadenylation signals on processing efficiency. RNA products were analyzed on a 5% polyacrylamide gel containing 7 M urea. The positions of uncleaved transcripts (input) and the 5[prime] cleavage product (arrow) are denoted to the left of the gel. |
|
We next addressed whether AUX DSEs were stimulating processing efficiency by influencing assembly of the general polyadenylation factors on the core AAUAAA and UREs. Since stable binding of CstF to the URE requires concomitant assembly of CPSF on the AAUAAA element (17,40), the association of CstF with the URE can be used as a measure of complex assembly. In addition, since substitution of the auxiliary downstream region did not alter the sequence of the core downstream URE, the level of UV cross-linking of the 64 kDa subunit of CstF (CstF-64) to variant polyadenylation signals should provide a direct reflection of its relative binding efficiency. Equimolar amounts of the wild-type and mutant polyadenylation signals described in Figure
Figure 2. The auxiliary downstream region of many polyadenylation signals functions to stabilize the interaction of CstF-64 with the RNA substrate. Equimolar amounts of wild-type and mutant RNAs were incubated in the in vitro cleavage system. Reaction mixtures were irradiated with UV light and treated with RNase A. Cross-linked CstF-64 was isolated by immunoprecipitation using a specific monoclonal antibody and analyzed on a 10% SDS-polyacrylamide gel. The position of the 64 kDa protein is indicated on the left.
AUX DSEs may promote the efficiency of 3[prime]-end processing by maintaining the core elements in an unstructured domain
We next attempted to precisely define the sequence and trans-acting factor requirements for the auxiliary downstream region of the L3 and AAV polyadenylation signals. Mutations in the auxiliary downstream region of these signals, however, had variable effects on processing efficiency (data not shown). This suggests that the functional element in the auxiliary downstream region of these signals was diffuse, reiterated or structural in nature and, therefore, difficult to identify. A comparison of the sequences present in the auxiliary downstream region of the four polyadenylation signals in which we have found AUX DSEs also failed to detect common sequence motifs or similarities (data not shown). Furthermore, while we have previously shown the AUX DSE of the SVL polyadenylation signal to increase processing efficiency through a titratable trans-acting factor (28), UV cross-linking studies failed to detect RNA-protein interactions with the AUX DSEs of the other polyadenylation signals tested (data not shown). These observations suggest that AUX DSEs may affect the efficiency of 3[prime]-end processing and assembly of the general polyadenylation factors by multiple mechanisms which can be dependent or independent of trans-acting factors. Since both the structural and functional consequences of mutations in the auxiliary downstream region of our polyadenylation substrate RNAs were difficult to predict, we attempted to identify mechanisms of AUX DSE action by destroying or restoring the ability of auxiliary downstream regions to function with insertions of defined sequences and structural elements. In this manner, the insertion of defined elements could be used to directly test a variety of models of AUX DSE action. Such an approach proved very useful in our previous efforts to define the core downstream URE (12,13).
The first alternative mechanism of action of AUX DSEs we tested was that they may contain sequences which have been selected for their ability to passively maintain the core polyadenylation elements in an unstructured domain. Recent observations by Gilmartin and colleagues suggest that unstructured RNA is a more efficient substrate for polyadenylation complex assembly (23). We inserted 20 base sequences into the auxiliary downstream region of the AAV poly(A) signal to create secondary structures with the regions encompassing the AAUAAA element, the cleavage site or the core downstream URE (Fig.
| Figure 3. Creation of secondary structures between the auxiliary downstream region of the AAV polyadenylation signal and various core elements. (A) Diagramatic representation of the AAV polyadenylation signal (AAV-wt) and its derivatives. Sequences of 20 bases (denoted 1[prime]-3[prime]) complementary to regions encompassing the AAUAAA element (denoted 1) (AAV-cAAUAAA), the cleavage site (denoted 2) (AAV-cCLEAV) or the core URE (denoted 3) (AAV-cURE) were inserted into the auxiliary downstream region of the AAV poly(A) signal. These sequences and other aspects of the constructs are described further in Materials and Methods. (B) RNase T1 ladders. RNAs were labeled at their 3[prime]-ends, incubated under the buffer conditions used for the in vitro cleavage assays and subjected to limited digestion with RNase T1 to generate RNA ladders. RNA products were analyzed on a 10% acrylamide gel containing 7 M urea. The regions protected by RNase T1 digestion due to the formation of secondary structures are bracketed. |
|
We next tested the effect of these secondary structures on the efficiency of 3[prime]-end processing. As shown in Figure
Figure 4. Secondary structures involving the auxiliary downstream region of the AAV polyadenylation signal and the regions encompassing the core elements or cleavage site dramatically decrease both 3[prime]-end processing efficiency and the level of assembly of CstF on the RNA substrate. (A) In vitro cleavage analysis. RNA products were analyzed on a 5% polyacrylamide gel containing 7 M urea. The positions of uncleaved transcripts (input) and the 5[prime] cleavage product (arrow) are denoted to the left of the gel. (B) UV cross-linking and immunoprecipitation analysis. Equimolar amounts of wild-type and variant RNAs were incubated in the in vitro cleavage system. Reaction mixtures were irradiated with UV light and treated with RNase A. Cross-linked CstF-64 was isolated by immunoprecipitation using a specific monoclonal antibody and analyzed on a 10% SDS-polyacrylamide gel. The position of the 64 kDa protein is indicated on the right.
AUX DSEs may promote the efficiency of 3[prime]-end processing by forming a stable secondary structure that helps focus binding of CstF to the core downstream URE
The second model for AUX DSE function we tested was whether some AUX DSEs may themselves form a stable secondary structure that does not involve the core polyadenylation elements. Such a structure may help focus binding of CstF to the core downstream URE. Such structures may prevent sliding of the relatively weak RNA binding subunit of CstF along the length of the RNA (14,40). Binding of the DSEF-1 protein, for example, may act in this manner by stabilizing the structure formed by base stacking interactions in the GRS element. An RNA pseudoknot from mouse mammary tumor virus was inserted into the auxiliary downstream region of the AdL3-mt polyadenylation signal (Fig.
| Figure 5. Insertion of an RNA pseudoknot in the auxiliary downstream region of the AdL3 polyadenylation signal promotes both efficient processing and assembly of CstF onto the RNA substrate. (A) Diagramatic representation of the AdL3 polyadenylation substrate (AdL3-wt) and its derivatives. An RNA pseudoknot was inserted into the AdL3-mt polyadenylation signal which lacks an AUX DSE. The AdL3-PSNmt RNA is a variant containing three base substitutions that specifically disrupt the pseudoknot structure. These sequences (represented by the stippled boxes) and other aspects of the constructs are described further in Materials and Methods. (B) In vitro cleavage analysis. RNA products were analyzed on a 5% polyacrylamide gel containing 7 M urea. The position of uncleaved transcripts (input) and the 5[prime] cleavage product (arrow) are denoted on the left of the gel. (C) UV cross-linking and immunoprecipitation analysis. Equimolar amounts of wild-type and variant RNAs were incubated in the in vitro cleavage system. Reaction mixtures were irradiated with UV light and treated with RNase A. Cross-linked CstF-64 was isolated by immunoprecipitation using a specific monoclonal antibody and analyzed on a 10% SDS-polyacrylamide gel. The position of the 64 kDa protein is indicated on the right. |
|
Only selected proteins can stimulate 3[prime]-end processing by binding to auxiliary downstream sequences
We have previously suggested a third model for AUX DSE-mediated efficient 3[prime]-end processing. The GRS element, the AUX DSE of the SVL polyadenylation signal, influences processing efficiency and assembly of the general polyadenylation factors through a titratable trans-acting factor, presumably DSEF-1 protein (28,29). In order to test whether the GRS element could substitute for the AUX DSE of the AdL3 polyadenylation signal, we inserted the GRS element into the auxiliary downstream region (Fig.
| Figure 6. Protein interaction with the GRS element, but not with the bacteriophage R17 coat protein binding site, functionally substitutes for the AUX DSE of the AdL3 polyadenylation signal. (A) Diagramatic representation of the AdL3 polyadenylation signal and its variants. The GRS element and the binding site for the bacteriophage R17 coat protein were inserted into the downstream region of the AdL3-mt polyadenylation signal which lacks an AUX DSE (AdL3-GRS and AdL3-R17 respectively). These sequences (represented by the stippled boxes) and other aspects of the constructs are described further in Materials and Methods. (B) In vitro cleavage analysis. RNA products were analyzed on a 5% polyacrylamide gel containing 7 M urea. The positions of uncleaved transcripts (input) and the 5[prime] cleavage product (arrow) are denoted to the left of the gel. (C) UV cross-linking and immunoprecipitation analysis. Equimolar amounts of wild-type and variant RNAs were incubated in the in vitro cleavage system. Reaction mixtures were irradiated with UV light and treated with RNase A. Cross-linked CstF-64 was isolated by immunoprecipitation using a specific monoclonal antibody and analyzed on a 10% SDS-polyacrylamide gel. The position of the 64 kDa protein is indicated on the right. (D) The effect of R17 coat protein on in vitro cleavage efficiency. RNAs were incubated in the in vitro cleavage system and products were analyzed on a 5% polyacrylamide gel containing 7 M urea. The last three lanes show the results of incubating the AdL3-R17 variant polyadenylation signal alone in the in vitro cleavage system (RNA only) or in the presence of a GST-R17 coat protein fusion (GST-R17CP) or GST protein (GST). The positions of uncleaved transcripts (input) and the 5[prime] cleavage product (arrow) are denoted to the left of the gel. |
|
We next tested whether other stable RNA-protein interactions in the auxiliary downstream region of a polyadenylation signal could stimulate 3[prime]-end processing. The binding site for the bacteriophage R17 coat protein (R17CP) was inserted into the auxiliary downstream region of the AdL3 polyadenylation signal at a similar location to the GRS element discussed above (Fig.
DISCUSSION
We have demonstrated that the region downstream of the core URE significantly affects the efficiency of 3[prime]-end processing of several mammalian and viral polyadenylation signals. All AUX DSEs we have tested influence processing efficiency by affecting the assembly of CstF, and perhaps all the general polyadenylation factors, on the RNA substrate. Although determination of the precise identity of AUX DSEs by mutational analyses has been difficult, evidence for three possible mechanisms by which AUX DSEs stimulate 3[prime]-end processing has been obtained. First, some AUX DSEs may promote efficient 3[prime]-end processing by passively maintaining the core elements of the polyadenylation signal in an unstructured domain to allow more efficient assembly of the general polyadenylation factors. Second, AUX DSEs may promote efficient processing by forming a stable structure that helps the weak RNA binding component of CstF focus on the core downstream URE of the polyadenylation signal and prevent its sliding along the RNA substrate. Finally, AUX DSEs may stimulate 3[prime]-end processing in an active fashion through interactions with specific trans-acting factors.
The data presented in this study, in conjunction with recent work on upstream auxiliary elements, outline several important principles which govern the organization of 3[prime]-end processing signals. Appropriately spaced AAUAAA and U-rich core elements provide the assembly sites for the general polyadenylation factors. These sequence elements alone, however, do not fully determine the strength of a 3[prime]-end processing signal. The core elements must be presented to the general polyadenylation factors in an appropriate structural context for efficient assembly. The structure of the core elements of the polyadenylation signal is influenced by its local sequence context on both the 5[prime]- and 3[prime]-sides of the region. Furthermore, functionally significant protein binding sites or structural elements may also be present in these surrounding sequences. Since the RNA binding components of the general polyadenylation factors interact with the core elements rather weakly on their own (14,40), a network of cooperative protein-protein interactions is required to stabilize the assembly of these factors onto the RNA substrate. It is reasonable to assume that the assembly of this multifactorial complex will not only be affected by the structural context of the core elements of the polyadenylation signal, but also by direct or indirect influences of RNA-protein interactions or structural elements present within the sequences surrounding the core elements. The sequences, structure and protein binding sites within these auxiliary regions, therefore, need to be seriously considered when evaluating a processing signal for mechanisms of efficiency, strength and regulation.
The core elements of a polyadenylation signal often function efficiently in vitro when the auxiliary downstream region is deleted from the RNA substrate (12,13). Presentation of the core elements to the general polyadenylation factors in this fashion is clearly artificial, however, as RNA polymerase II continues well beyond the cleavage/polyadenylation site of the nascent transcript before transcription termination occurs. Auxiliary downstream elements, therefore, appear to be an integral part of viral and mammalian polyadenylation signals. Just as auxiliary upstream sequences have been shown to affect CPSF-AAUAAA element interactions (18,21,24,27), auxiliary downstream sequences are likely to influence the association of CstF with the URE. For most signals, these auxiliary elements probably function by providing a proper structural context for the core AAUAAA and UREs. It, would be difficult to identify precise AUX DSE sequence requirements, therefore, because the sequence of the auxiliary element would depend on the sequence context of the adjacent core elements to prohibit the formation of short range or long range inhibitory structures in the pre-mRNA. In addition, it would also be unlikely to uncover a consensus sequence for AUX DSEs. In these cases, AUX DSEs are an innate part of polyadenylation signals and are probably not subject to many regulatory influences.
Alternatively, auxiliary elements could provide important regulatory influences on the efficiency of polyadenylation in some cases. Regulation of 3[prime]-end processing efficiency solely through the core elements would appear to be difficult to accomplish, since these elements are so well conserved. We have noted that most known polyadenylation signals which are subject to regulation contain non-consensus sequences in one or both core elements. In these signals, the influence of auxiliary sequences probably plays an important role in stabilizing the interaction of the general polyadenylation factors with these variant core elements. While dramatic changes in the concentration of one component of CstF have recently been shown to affect poly(A) signal usage (42), AUX DSEs could provide a subtle, signal-specific way of altering or fine tuning the effective concentration of CstF available to a pre-mRNA.
The observations described in Figure
DSEF-1 protein, which specifically interacts with the GRS element located in the auxiliary downstream region of the SV40 late polyadenylation signal, is the only candidate AUX DSE binding factor identified to date (28). The inability of gel shift and UV cross-linking analyses to identify proteins binding to other AUX DSEs (data not shown) may be a reflection of the stringency of these individual assays and cannot be taken as definitive data that other AUX DSE binding factors do not exist. The lack of AUX DSE consensus elements, however, suggests that specific AUX DSE binding proteins may be rather diverse, possess minimal sequence requirements for RNA binding or are relatively rare. Studies underway to address the requirements and mechanism of stimulation of 3[prime]-end processing by DSEF-1 should provide insight into the role of AUX DSE binding proteins.
Strong structural elements, such as an RNA pseudoknot placed downstream of the URE, stimulate 3[prime]-end processing efficiency. Such elements have been previously shown to alter another post-transcriptional process, namely ribosomal frameshifting (41). The observation that subtle mutations which disrupt the pseudoknot structure but have only minor impact upon the overall sequence in the region dramatically decrease processing efficiency argues strongly that the structure itself may be playing a role. Stable RNA structures may cause CstF to pause and prevent it from migrating too far from the URE prior to stabilization of assembly by cooperative interactions with CPSF. Structural elements have beeen noted in the upstream auxiliary regions of the HIV and HTLV-1 polyadenylation signals (21). Whether structural elements are a common phenomenon in auxiliary downstream regions remains to be elucidated.
In conclusion, we have demonstrated that auxiliary downstream sequences play an important and general role in determining the efficiency of 3[prime]-end processing. Understanding the mechanism of this stimulation of processing efficiency should provide insight not only into determinants of poly(A) signal strength, but also into possible roles of these elements in regulation.
ACKNOWLEDGEMENTS
We wish to thank Martha Peterson and Olke Uhlenbeck for plasmid constructs and C.Schaefer, L.Ford and C.Williams for helpful suggestions on the manuscript. This work was supported by grants from the Pew Charitable Trusts, the Foundation of UMDNJ and the National Institutes of Health (GM 56434) to J.W. F.C. was supported in part by training grant CA09663.
REFERENCES
This article has been cited by other articles:
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 4 Jun 1998
Copyright©Oxford University Press, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
S. Millevoi, A. Decorsiere, C. Loulergue, J. Iacovoni, S. Bernat, M. Antoniou, and S. Vagner
A physical and functional link between splicing factors promotes pre-mRNA 3' end processing
Nucleic Acids Res.,
August 1, 2009;
37(14):
4672 - 4683.
[Abstract]
[Full Text]
[PDF]
M. Dalziel, N. M. Nunes, and A. Furger
Two G-Rich Regulatory Elements Located Adjacent to and 440 Nucleotides Downstream of the Core Poly(A) Site of the Intronless Melanocortin Receptor 1 Gene Are Critical for Efficient 3' End Processing
Mol. Cell. Biol.,
March 1, 2007;
27(5):
1568 - 1580.
[Abstract]
[Full Text]
[PDF]
D. Liu, J. M. Brockman, B. Dass, L. N. Hutchins, P. Singh, J. R. McCarrey, C. C. MacDonald, and J. H. Graber
Systematic variation in mRNA 3'-processing signals during mouse spermatogenesis
Nucleic Acids Res.,
January 12, 2007;
35(1):
234 - 246.
[Abstract]
[Full Text]
[PDF]
J. E. Wilusz and K. L. Beemon
The Negative Regulator of Splicing Element of Rous Sarcoma Virus Promotes Polyadenylation
J. Virol.,
October 1, 2006;
80(19):
9634 - 9640.
[Abstract]
[Full Text]
[PDF]
M. L. Peterson, G. L. Bingham, and C. Cowan
Multiple features contribute to the use of the immunoglobulin m secretion-specific poly(a) signal but are not required for developmental regulation.
Mol. Cell. Biol.,
September 1, 2006;
26(18):
6762 - 6771.
[Abstract]
[Full Text]
[PDF]
T. Hall-Pogar, H. Zhang, B. Tian, and C. S. Lutz
Alternative polyadenylation of cyclooxygenase-2
Nucleic Acids Res.,
May 4, 2005;
33(8):
2565 - 2579.
[Abstract]
[Full Text]
[PDF]
B. Tian, J. Hu, H. Zhang, and C. S. Lutz
A large-scale analysis of mRNA polyadenylation of human and mouse genes
Nucleic Acids Res.,
January 12, 2005;
33(1):
201 - 212.
[Abstract]
[Full Text]
[PDF]
D. T. Fritz, D. Liu, J. Xu, S. Jiang, and M. B. Rogers
Conservation of Bmp2 Post-transcriptional Regulatory Mechanisms
J. Biol. Chem.,
November 19, 2004;
279(47):
48950 - 48958.
[Abstract]
[Full Text]
[PDF]
A. Hajarnavis, I. Korf, and R. Durbin
A probabilistic model of 3' end formation in Caenorhabditis elegans
Nucleic Acids Res.,
June 24, 2004;
32(11):
3392 - 3399.
[Abstract]
[Full Text]
[PDF]
J. Qiu, R. Nayak, and D. J. Pintel
Alternative Polyadenylation of Adeno-Associated Virus Type 5 RNA within an Internal Intron Is Governed by both a Downstream Element within the Intron 3' Splice Acceptor and an Element Upstream of the P41 Initiation Site
J. Virol.,
January 1, 2004;
78(1):
83 - 93.
[Abstract]
[Full Text]
[PDF]
M. I. Zarudnaya, I. M. Kolomiets, A. L. Potyahaylo, and D. M. Hovorun
Downstream elements of mammalian pre-mRNA polyadenylation signals: primary, secondary and higher-order structures
Nucleic Acids Res.,
March 1, 2003;
31(5):
1375 - 1386.
[Abstract]
[Full Text]
[PDF]
B. J. Natalizio, L. C. Muniz, G. K. Arhin, J. Wilusz, and C. S. Lutz
Upstream Elements Present in the 3'-Untranslated Region of Collagen Genes Influence the Processing Efficiency of Overlapping Polyadenylation Signals
J. Biol. Chem.,
November 1, 2002;
277(45):
42733 - 42740.
[Abstract]
[Full Text]
[PDF]
M. L. Peterson, S. Bertolino, and F. Davis
An RNA Polymerase Pause Site Is Associated with the Immunoglobulin {micro}s Poly(A) Site
Mol. Cell. Biol.,
August 1, 2002;
22(15):
5606 - 5615.
[Abstract]
[Full Text]
[PDF]
G. K. Arhin, M. Boots, P. S. Bagga, C. Milcarek, and J. Wilusz
Downstream sequence elements with different affinities for the hnRNP H/H' protein influence the processing efficiency of mammalian polyadenylation signals
Nucleic Acids Res.,
April 15, 2002;
30(8):
1842 - 1850.
[Abstract]
[Full Text]
[PDF]
D. P. Tran, S. J. Kim, N. J. Park, T. M. Jew, and H. G. Martinson
Mechanism of Poly(A) Signal Transduction to RNA Polymerase II In Vitro
Mol. Cell. Biol.,
November 1, 2001;
21(21):
7495 - 7508.
[Abstract]
[Full Text]
[PDF]
L. C. Chao, A. Jamil, S. J. Kim, L. Huang, and H. G. Martinson
Assembly of the Cleavage and Polyadenylation Apparatus Requires About 10 Seconds In Vivo and Is Faster for Strong than for Weak Poly(A) Sites
Mol. Cell. Biol.,
August 1, 1999;
19(8):
5588 - 5600.
[Abstract]
[Full Text]
[PDF]
J. Zhao, L. Hyman, and C. Moore
Formation of mRNA 3' Ends in Eukaryotes: Mechanism, Regulation, and Interrelationships with Other Steps in mRNA Synthesis
Microbiol. Mol. Biol. Rev.,
June 1, 1999;
63(2):
405 - 445.
[Abstract]
[Full Text]
[PDF]
This Article 
Abstract
Print PDF (229K)
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (26)
Request Permissions
Commercial Re-use Guidelines
for Open Access NAR Content
Google Scholar
Articles by Chen, F.
Articles by Wilusz, J.
Search for Related Content
PubMed
PubMed Citation
Articles by Chen, F.
Articles by Wilusz, J.
Social Bookmarking
What's this?