The murine IgM secretory poly(A) site contains dual upstream and downstream elements which affect polyadenylation
The murine IgM secretory poly(A) site contains dual upstream and downstream elements which affect polyadenylation
Cathy Phillips, Anders Virtanen*
Department of Medical Genetics, Uppsala University, Biomedical Centre, Box 589, SE-751 23 Uppsala, Sweden
Received March 14, 1997;Revised and accepted April 21, 1997
ABSTRACT
Regulation of polyadenylation efficiency at the secretory poly(A) site plays an essential role in gene expression at the immunoglobulin (IgM) locus. At this poly(A) site the consensus AAUAAA hexanucleotide sequence is embedded in an extended AU-rich region and there are two downstream GU-rich regions which are suboptimally placed. As these sequences are involved in formation of the polyadenylation pre-initiation complex, we examined their function in vivo and in vitro. We show that the upstream AU-rich region can function in the absence of the consensus hexanucleotide sequence both in vivo and in vitro and that both GU-rich regions are necessary for full polyadenylation activity in vivo and for formation of polyadenylation-specific complexes in vitro. Sequence comparisons reveal that: (i) the dual structure is distinct for the IgM secretory poly(A) site compared with other immunoglobulin isotype secretory poly(A) sites; (ii) the presence of an AU-rich region close to the consensus hexanucleotide is evolutionarily conserved for IgM secretory poly(A) sites. We propose that the dual structure of the IgM secretory poly(A) site provides a flexibility to accommodate changes in polyadenylation complex components during regulation of polyadenylation efficiency.
INTRODUCTION
The immunoglobulin (Ig) M secretory poly(A) site represents one of the most well-known examples of a regulated poly(A) site (Fig. 1A). Its usage is regulated during B cell differentiation (1-3), which appears to be controlled by modulation of polyadenylation activity at this site (4-19). It has been proposed that polyadenylation efficiency at this site may be regulated by modulating the abundance or binding activity of common components of the poly(A) complex (15,20,21). An inspection of the sequences of this poly(A) site revealed a duplicated structure in which upstream and downstream sequences normally involved in interactions with core components of the cleavage/poly(A) complex were repeated. We were therefore interested in examining the contribution of these repeated sequence elements to polyadenylation efficiency at the secretory poly(A) site, as this might provide insight into the regulation of polyadenylation activity at this site.
Figure 1 (A) The IgM heavy chain pre-mRNA can be alternatively processed into two forms of mRNA. (B) Schematic diagram showing the [mu] secretory poly(A) site and the mutant constructs used. The multiple hexanucleotide (light shaded boxes) and multiple GU-rich regions (dark shaded boxes) are as indicated. The cleavage site is indicated with a bold arrow. The 55 nt element is indicated between the arrows. Names of the mutant constructs indicating the mutations or deletions which they contain are indicated. Nucleotide positions are numbered according to the mouse IgM sequence accession no. V00818.
Recognition of a poly(A) site by the cleavage/poly(A) complex involves a network of multiple weak interactions between a number of sequence elements and different components of the 3[prime]-processing apparatus as well as those between the components of the complex themselves (22). Limiting the availability of any component of the polyadenylation complex in in vitro assays results in dependence upon that component for polyadenylation efficiency (see for example 23-26), thereby providing evidence for the possibility that regulation may simply occur by modulating the abundance of the basal polyadenylation factors. However, the structure of the cis-acting elements within a substrate will also determine the relative dependence for processing on those rate limiting factors (27). Thus, it is crucial to understand in detail the significance of the different participating trans-acting factors and cis-acting elements to be able to understand how polyadenylation can be regulated/modulated at a polyadenylation site.
Eukaryotic polyadenylation sites usually contain an AAUAAA hexanucleotide sequence element located [sim]20 nt upstream of the site of cleavage and poly(A) addition (reviewed in 28). In addition, polymorphic GU-rich regions are often present downstream of the cleavage site. The sequences of the GU-rich regions and their distance from the hexanucleotide element and the site of cleavage play a vital role in determining the strength of a poly(A) site (reviewed in 28-30). The multi-component processing complex which interacts with these sequences consists of at least five different factors: the cleavage polyadenylation specificity factor (CPSF), the cleavage stimulatory factor (CstF), two cleavage factors (CF I and CF II) and poly(A) polymerase (PAP) (31; reviewed in 32,33).
CPSF and CstF have been biochemically characterized and purified and both consist of multiple polypeptides (34,35). Furthermore, it has been shown that CPSF interacts with the hexanucleotide element while CstF interacts with the GU-rich downstream elements (reviewed in 30,31). Neither factor binds RNA avidly by itself, but interaction between the two and the RNA substrate results in formation of a stable complex known as the pre-initiation complex (24,36). It has been proposed that CstF is limiting for poly(A) site complex formation in B cells and regulates the usage of the secretory poly(A) site during B cell differentiation (21). Therefore, the structure of the GU-rich sequences which bind CstF plays an important role in this process. Sequences which bind CPSF, an essential component for stability of the pre-initiation complex involving CstF and RNA, would also presumably be important here.
In the IgM heavy chain secretory poly(A) site the consensus hexanucleotide sequence is embedded in an AU-rich sequence. Furthermore, two GU-rich regions, 8 and 11 nt in length, are located 21 and 58 nt respectively downstream of the hexanucleotide sequence (see Figs 1B and 3). By comparison with model substrates in which the distance between the cleavage site and the GU-rich regions were synthetically expanded or contracted (29), it was found that the location of both these GU-rich regions is suboptimal, one being closer and the other more distant from the hexanucleotide sequence than optimal. We have previously shown that the sequences spanning the distal GU-rich region enhance polyadenylation activity in vivo, using luciferase reporter constructs transfected into plasmacytomas (37). We have therefore examined the contribution of both the AAUAAA hexanucleotide sequence element and the upstream AU-rich regions as well as both GU-rich regions to polyadenylation activity at the [mu] secretory poly(A) site, both in vivo using transfection experiments and in vitro at the level of formation of the CPSF-CstF pre-initiation complex. We show that the multiple elements of the secretory poly(A) site all contribute to polyadenylation efficiency at this site.
MATERIALS AND METHODS
Plasmids constructs
Two parallel sets of plasmids were constructed using vectors pGem T (Promega) and pPKLT55, kindly provided by Walter Dietrich-Goetz (37). A series of synthetic oligonucleotides (for sequences see below) were synthesized. Recognition sites for restriction enzymes BglII or XbaI were included at the 5[prime]-ends of each oligonucleotide (a-e and x-z respectively). PCR products were generated using the following pairs of oligonucleotides, ax, bx, cx, dx, ex, ay, az, aq and px and plasmid G1[gamma]359 (38,39), containing the mouse IgM [mu] secretory poly(A) site, as the template. For introduction of mutations in the proximal GU-rich region, PCR products aq and px were gel purified, mixed and used as a template with primers a and x to produce a PCR product aqpx. For pictorial representation of mutants see Figure 1B. Nucleotide positions are numbered according to the mouse IgM sequence (DDBJ/EMBL/GenBank accession no. V00818). All PCR products except aq and px were cloned into the T site of pGem T. The pGem T inserts were isolated after restriction with BglII and XbaI and subcloned into pPKLT55 in place of the HSV-2 poly(A) site between the BglII and XbaI sites. Plasmid pLu p[mu]spA2 was similarly constructed and has been described previously and contains sequences from positions 1838-2085 (37). Plasmids pT3L3 and pT3L3G have been described previously (40). The following oligonucleotides were synthesized:
a, GACAGATCTGTTGCATTTATAAAAAATTAGAAATAAAAAAAAT; b, GACAGATCTGTTGCATTTATAAAAAATTAGAAAGAAAAAAAAT; c, GACAGATCTGTTGCATTTAGAAAAAAGGAGAAATAAAAAAAAT; d, GACAGATCTGTTGCATTTAGAAAAAAGGAGAAAGAAAAAAAAT; e, GACAGATCTGTTGCATTTATAAAAAATTAGAAAAAAT; f, GACAGATCTGTTGCATTTAGAAAAAAGGAGAAAAAAT; x, GCGTCTAGATAGGGTGGAGGCAAGTATGCAGGGTGTG; y, GCGTCTAGAGCAGGCATGAGCATTGTATAATCAAAACC; z, GCGTCTAGAAGTGACGTTTGAATGGATTT; p, CATTCAAACGTCACTAGATATAATTATACAATGCTCATG; q, CATGAGCATTGTATAATTATATCTAGTGACGTTTGAATG.
The plasmids were named in a binary fashion to denote which element was mutated as follows: ax, IgsH11G11; bx, IgsH10G11; cx, IgsH01G11; dx, IgsH00G11; ex, IgsH1delG11; fx, IgsH0delG11; ay, IgsH11G10, apqx, IgsH11G01, az, IgsH11G00 (see Fig. 1B). Those in vector pGem T were given the prefix `T' and those in pPKLT55 the prefix `Lu'.
RNA substrates
Templates for in vitro transcription of RNA substrates were prepared by restricting the T series of plasmids with XbaI, except those for RNA substrates IgsH11G10 and IgsH11G00, which were prepared by restricting the respective T plasmids with NsiI. Templates for L3(103) and L3G(103) were prepared by restricting plasmids pT3L3 and pT3L3G with DraI.
Capped RNA substrates were made by in vitro transcription following the protocol of the RNA polymerase (T3 or T7) supplier (Stratagene) as previously outlined (23).
Cell culture and transfection
J558L (41) was obtained from the European Collection of Animal Cell Cultures (Salisbury, UK). Plasmids were transfected into J558L cells in log phase using DEAE-dextran according to Grosschedl and Baltimore (42) as previously described (37). Transfection efficiency (see 37 for details) was measured by co-transfection of plasmid pSVCATb (43).
Preparation of nuclear extracts, CPSF and CstF
HeLa cell nuclear extracts were prepared according to Dignam et al. (44) with the modifications described by Moore and Sharp (45). The HeLa cell nuclear extract contained 20 mg/ml protein. CPSF and CstF were obtained as fractions IIIC and IA respectively according to the protocol of Åström et al. (23).
In vitro polyadenylation and complex formation
In vitro polyadenylation was performed according to Moore and Sharp (45). Complex formation reactions were performed according to Gilmartin and Nevins (46). CstF was depleted of the 64 kDa component using the monoclonal antibody 3A7 (47) and mock depleted using the monoclonal anti-poly(A) polymerase antibody 20:14 (48). Aliquots of 20 [mu]l of each antibody were added to separate 100 [mu]l aliquots of fraction 1A containing CstF and the supernatant rotated for 2 h in a cold room. To this was added 10 [mu]l gammabind (Pharmacia no 17-0885-01) and the mixture was rotated for a further 30 min. The beads were removed by microcentrifugation and the depleted fractions were recovered as the supernatants.
RESULTS
Multiple and conserved sequence elements at the [mu] isotype secretory poly(A) site
We compared the sequences of [mu] secretory poly(A) sites from a variety of higher eukaryotes to identify conserved cis-acting elements important for polyadenylation. The adjacent upstream AU-rich region was well conserved among mammals (Fig. 2). Furthermore, AU-rich sequences were found within the vicinity of the consensus hexanucleotide sequence in chicken, amphibians, fish and shark (see Fig. 2). Unfortunately, only cDNAs were available for many of these [mu] secretory sites, therefore, we were unable to examine the structure of the downstream regions. We next compared the secretory poly(A) sites of the murine immunoglobulin isotypes to look for conservation of the AU-rich region in the vicinity of the AAUAAA hexanucleotide element and the existence of dual GU-rich regions. No AU-rich region was present in the vicinity of the consensus hexanucleotide sequence of any of the other immunoglobulin isotypes, [gamma]3, [gamma]1, [gamma]2b, [gamma]2a, [epsis] or [alpha] (see Fig. 3). In addition, the GU-rich regions (marked by a box in Fig. 3) of [gamma]3, [gamma]1, [gamma]2b, [gamma]2a and [epsis], rather than having a repeated structure, consisted of one region optimally placed [sim]30 nt downstream of the hexanucleotide sequence. For [alpha], no obvious GU-rich region was near the hexanucleotide, although some G-rich regions are marked in Figure 3. Thus the repeated structure is a distinct feature of the [mu] secretory poly(A) site in mouse. The [mu] membrane poly(A) site shows a single AAUAAA hexanucleotide sequence element and a single optimally placed GU-rich region (data not shown). A similar pattern of cis-acting elements was found when human immunoglobulin isotypes were compared (data not shown).
Figure 2 The extended hexanucleotide structure is conserved throughout evolution. A comparison of the sequences surrounding the hexanucleotide sequence of the [mu] secretory poly(A) site. The consensus hexanucleotide sequence is marked by shaded boxes and the adjacent AU-rich sequences are marked with hatched boxes. The [mu] poly(A) sequences were obtained from the following sources: shark (66); catfish (67); axolotl (68); toad (69); chicken (70); rabbit (71); mouse (2); hamster (72); sheep (73); pig (Bosch et al., unpublished results); human (74).
. Multiple hexanucleotides and GU-rich regions are a unique feature of the [mu] isotype. A comparison of poly(A) site structures between isotypes of the murine immunoglobulin genes. The consensus hexanucleotide sequence is marked by shaded boxes, the adjacent AU-rich region with ascending stripes (left of figure) and the downstream GU-rich regions are marked with boxes with descending stripes (right of figure). G-rich regions of the [alpha] sequences are underlined. The murine immunoglobulin isotype poly(A) sequences were obtained from the following sources: [mu] (2); [gamma]3 (75); [gamma]1 (76); [gamma]2b and [gamma]2a (77); [epsis] (78); [alpha] (79).
The AU-rich region, the AAUAAA hexanucleotide element and both GU-rich regions contribute to polyadenylation at the [mu] secretory poly(A) site in vivo
We examined the contribution of the AU-rich region as well as the AAUAAA element and the two downstream GU-rich regions to polyadenylation efficiency at the secretory poly(A) site in vivo.
We constructed a series of expression vectors containing the luciferase gene followed by the secretory poly(A) site modified by deletion or point mutation in each of the elements under investigation (see Fig. 1B and Materials and Methods for details). These constructs were transfected into J558L plasmacytomas and luciferase activity was subsequently determined.
Luciferase activity was reduced to 20% of that of the wild-type sequence when the central U was mutated to G in the consensus hexanucleotide sequence (Fig. 4, compare lanes 1 and 2). This luciferase activity was reduced significantly further by additionally mutating the three U residues to G in the AU-rich sequence upstream of the consensus hexanucleotide sequence element (Fig. 4, compare lanes 2 and 3). Similarly, when the whole consensus hexanucleotide sequence element was deleted, residual activity was retained (Fig. 4, lane 4), which again was reduced by additionally mutating the three U residues to G in the adjacent upstream AU-rich region (Fig. 4, compare lanes 4 and 5). These results suggest that the adjacent AU-rich sequence can maintain a low level of polyadenylation activity in the absence of a functional consensus hexanucleotide sequence. When the adjacent AU-rich sequence alone was mutated, leaving the consensus hexanucleotide intact, luciferase activity was not reduced (luciferase activity ± SE, 111.8 ± 15.2%).
Figure 3 Contribution of the hexanucleotide sequence and the proximal and distal GU-rich regions to polyadenylation efficiency in vivo. The respective sequences were inserted downstream of the luciferase gene at the poly(A) site of pPKLT55 and transfected into J558L plasmacytomas in triplicate. Transfection efficiency was standardized by co-transfection with SVCATb. Luciferase activity was measured for each construct and expressed as a percentage ± SE of Lu [mu]spA2 activity (lane 9). Bars represent per cent activity. Lanes are as indicated. Bars are shaded for easy reference: plain shaded boxes represent mutations in upstream regions whereas striped boxes represent mutations in downstream regions.
Deletion of a 55 nt element containing the distal GU-rich region reduced luciferase activity to 30% (Fig. 4, compare lanes 1 and 6). Similarly, mutation of the proximal GU-rich element while retaining the distal GU-rich region also reduced luciferase activity, although to a lesser extent (Fig. 4, compare lanes 1 and 7). However, deletion of both GU-rich regions abolished activity completely (Fig. 4, lane 9). These results suggest that both GU-rich regions contribute to polyadenylation and both are necessary for full activity.
The consensus hexanucleotide element and both GU-rich regions are necessary for pre-initiation complex formation in vitro
The hexanucleotide sequence element and the downstream GU-rich regions together with CPSF and CstF participate in formation of the pre-initiation complex (36). We therefore examined the effect of mutation in these sequences on complex formation in vitro. For this we used fractions containing CPSF and CstF (23). The 103 nt RNA substrates L3(103) and L3G(103), containing the wild-type or a U[rarr]G mutated hexanucleotide sequence element of the human adenovirus type 2 L3 poly(A) site respectively, were used to identify hexanucleotide-specific complex formation (40). A hexanucleotide-dependent complex was identified (Fig. 5A, compare lanes 1 and 2). A number of non-specific complexes of higher mobility were also visible, but these did not differ between the mutant and the wild-type L3 substrate.
Figure 4 The contribution of poly(A) sequence elements to pre-initiation complex binding. Fractions containing the indicated components of the cleavage/poly(A) complex were incubated with wild-type or mutant substrates of the [mu] secretory poly(A) site or the adenovirus late 3 poly(A) site [L3 and L3(G)], run on 4% native PAGE as indicated in Materials and Methods and analysed by phosphorimaging. (A) The [mu] secretory poly(A) site forms complexes with CPSF- and CstF-containing fractions, dependent on the presence of both fractions and the 64 kDa component of CstF. Lanes are as indicated. (B) Contribution of the multiple elements. Lanes are as indicated.
A complex of the same mobility as the L3-specific hexanucleotide-dependent complex was formed on the 133 nt RNA substrate spanning the wild-type [mu] secretory poly(A) site (Fig. 5A, lane 3). The extra 30 nt of the IgM substrate did not affect the mobility of the specific complex; small variations in RNA substrate length make a minor contribution to the molecular weight of the RNA-CPSF-CstF complex. This complex was only formed when both CPSF and CstF fractions were present (Fig. 5A, compare lane 3 with 4 and 5). The amount of CPSF used here was not sufficient to form the less stable hexanucleotide-dependent CPSF-RNA complex (46; Fig. 5A, lane 4).
Depletion of the CstF fraction of the 64 kDa component of CstF with antibody 3A7 (47) by immunoprecipitation abolished the ability of this fraction to participate in complex formation (Fig. 5A, lanes 6 and 7). Mock depletion with monoclonal antibody 20:14 (48) did not abolish complex formation (Fig. 5A, lanes 8 and 9). Taken together these experiments identify this complex as the specific hexanucleotide-dependent CPSF-CstF complex, similar to that formed on the L3 wild-type substrate. Once again, a number of complexes of faster mobility could be seen. However, none of these correlated with the specific complex formed on the L3 RNA substrate.
We next examined the effects of mutations in the extended hexanucleotide sequence element upon specific complex formation. Whereas mutation of the three U residues to G of the adjacent AU-rich region did not abolish formation of this specific complex (Fig. 5B, lane 5), a U[rarr]G mutation in the consensus hexanucleotide sequence element abolished the ability to form a complex of this mobility (Fig. 5B, lane 6). Furthermore, substrates in which all four U residues were mutated to G residues were not able to form specific complexes (Fig. 5B, lane 7). Finally, we investigated the effect of each GU-rich region on specific complex formation. The specific complex was not formed when either of the distal or proximal GU-rich regions was deleted and replaced by a section of polylinker (Fig. 5B, lane 2) or mutated (Fig. 5B, lane 3). Furthermore, when both GU-rich regions were deleted and replaced by polylinker, specific complex formation was abolished (Fig. 5B, lane 4), although the higher mobility non-specific complexes were still present. Thus, we conclude that both GU-rich regions and the consensus hexanucleotide element are necessary for formation of the specific hexanucleotide-dependent CPSF-CstF complex.
The hexanucleotide-adjacent AU-rich region functions in in vitro polyadenylation assays
We tested the ability of the hexanucleotide-mutated substrates to function in in vitro polyadenylation assays using HeLa cell nuclear extracts. The results are presented in Figure 6. Polyadenylated products were clearly formed when the wild-type substrate or the substrate containing the three U[rarr]G mutations in the upstream hexanucleotide-adjacent AU-rich region were used (Fig. 6, lanes 1 and 3). There were no polyadenylated products visible when all four U residues were mutated to G (Fig. 6, lane 4). However, when substrates containing only the U[rarr]G mutation in the consensus hexanucleotide sequence were used, polyadenylated products were still discernible (Fig. 6, lane 2). Taken together, this shows that the upstream hexanucleotide-adjacent AU-rich region can function as a hexanucleotide sequence, albeit at a lower level than the consensus hexanucleotide sequence.
Figure 5 Contribution of the consensus hexanucleotide and the upstream adjacent AU-rich region to cleavage/polyadenylation in vitro. Polyadenylation assays were performed in vitro using 2 [mu]l HeLa cell extracts containing 40 [mu]g protein. Processed RNA was recovered by proteinase K digestion followed by extraction in phenol/chloroform and precipitation with ethanol. Samples were run in 10% denaturing PAGE and analysed by phosphorimaging. Lanes are as indicated. Positions of the unreacted substrates and poly(A) tails are indicated
DISCUSSION
We have examined the contribution of sequence elements which interact with components of the polyadenylation complex at the [mu] secretory poly(A) site, focusing our studies on the hexanucleotide element and the downstream GU-rich regions. We found that the murine IgM poly(A) site consists of repeated cis-acting elements, i.e. two downstream GU-rich regions (Figs 1 and 3) and an extended hexanucleotide element consisting of the AAUAAA hexanucleotide element and an adjacent AU-rich region (Figs 1 and 2). While the hexanucleotide-adjacent AU-rich region can function in in vitro and in vivo polyadenylation assays, both GU-rich regions and the consensus hexanucleotide sequence contribute to the stability of the pre-initiation complex. This multi-element structure may confer flexibility, enabling the secretory poly(A) site to accommodate changes which may occur during the regulation of polyadenylation efficiency at this site.
In the [mu] secretory poly(A) site, neither GU-rich region is optimally placed (29). We have shown that both GU-rich regions are necessary in the [mu] secretory poly(A) site to restore wild-type luciferase activity in vivo (Fig. 4). However, each GU-rich region can confer a residual reduced activity when present alone and it is not until both are removed that activity is abolished in vivo (Fig. 4). Indeed, it has been shown that substrates including only the proximal GU-rich region are sufficient for cleavage and polyadenylation in vitro using HeLa cell nuclear extracts (27). However, both GU-rich regions appear to be necessary for pre-initiation complex formation involving only CstF and CPSF in vitro. No specific complex was formed on substrates containing either GU-rich region alone (Fig. 5B). It is therefore possible that further components of the cleavage/polyadenylation complex [e.g. the cleavage factors or poly(A) polymerase] are necessary to stabilize the complex on the sequences containing only one suboptimally placed GU-rich region.
The adjacent AU-rich region was able to support residual luciferase activity in vivo. This residual activity did not appear to be at the level of pre-initiation complex formation, as we were unable to detect any CstF-CPSF pre-initiation complex formation on these sequences in the absence of the consensus hexanucleotide sequence in vitro (Fig. 5B). Nevertheless, the adjacent AU-rich sequence could support residual polyadenylation activity in vitro in unfractionated HeLa cell nuclear extracts. Caution should be employed in interpreting in vivo polyadenylation assays involving mutations in sequences upstream of the cleavage site, as these may affect reactions subsequent to polyadenylation. Most notably, AU-rich sequences have been shown to be involved in RNA stability (reviewed in 49). However, detection of residual polyadenylation activity in vitro in HeLa cell nuclear extracts, which is dependent on the presence of the adjacent AU-rich region (Fig. 6, compare lanes 2 and 4), shows that the adjacent AU-rich region can promote polyadenylation, in agreement with the residual activity in vivo. Thus it is possible that other components of the complex are necessary for function of this sequence. These differences between pre-initiation complex formation in vitro, in vitro polyadenylation and in vivo activity highlight the potential for regulation of polyadenylation activity at multiple levels and the possibility that poly(A) site structure can determine which factor is limiting in each case.
Taken together, our results suggest that two GU-rich regions and an extended hexanucleotide sequence are involved in polyadenylation at the secretory poly(A) site. The simplest interpretation of our polyadenylation-specific complex formation experiments (Fig. 5) is that only one CstF interacts with one CPSF factor and both GU-rich sequences simultaneously, each contributing to stabilize the interaction. In support of this suggestion, the AAUAAA-dependent specific complex formed on the L3 substrate co-migrated with the complex formed on the IgM substrate including both GU-rich regions, thus providing no evidence for an extra CstF in the IgM-specific complex (Fig. 5A, lanes 1 and 3). We propose that the bipartite nature of the CstF binding site in the secretory poly(A) site may be a mechanism to modulate the affinity of binding of CstF at different stages of cell differentiation.
In this respect, it has recently been reported that the abundance of the 64 kDa component of CstF may be regulated in primary cells during B cell growth (21). As an essential component of CstF it thus regulates the abundance of functional CstF. It can be shown that titration of CstF in an in vitro purified reconstituted system is sufficient to regulate polyadenylation efficiency at the [mu] secretory poly(A) site (21). These experiments highlight the importance of the nature of the interaction of CstF with the RNA substrate in polyadenylation efficiency. It can be envisioned that transacting factors which function to increase or decrease the binding of CstF to the RNA substrate would have the same effect. A number of studies have detected changes in the polyadenylation complex which correlate with differing levels of usage of the secretory poly(A) site. These include an inhibitory factor which specifically targets the [mu] secretory poly(A) site and a factor which enhances polyadenylation activity at a number of poly(A) sites (15,20). We have previously found a 28-32 kDa polypeptide whose induction correlates with increased usage of the secretory poly(A) site and whose binding is dependent upon the 55 nt element spanning the distal GU-rich element (37).
It has previously been found that CPSF may also contact sequences outside the consensus hexanucleotide sequence, as exemplified by the HIV-1 poly(A) site (50-52), the SV40 poly(A) site (53) and the adenovirus major late poly(A) site (54; reviewed in 55). In this site, polymorphic upstream sequences can compensate for a suboptimal hexanucleotide sequence. Thus the HIV-1 case demonstrates that sequences other than the AAUAAA element can be important for interaction between CPSF and its RNA substrate. In addition, sequences both upstream and downstream of the consensus sequence affect polyadenylation in plants (56-59). However, in the cases mentioned above the upstream elements do not resemble the consensus hexanucleotide sequence, suggesting that they may bind different polypeptide sequences. In contrast, the IgM secretory poly(A) upstream regions do resemble the hexanucleotide sequence, which suggests that they may bind the same polypeptide sequences. Our in vivo analysis shows that the AU-rich region has a residual function in the absence of the consensus hexanucleotide sequence in in vivo polyadenylation assays. Tsurushita et al. (13) have previously found that a U[rarr]G point mutation of the consensus AAUAAA was not sufficient to block processing at the IgM secretory poly(A) site in plasmacytomas, but that complete blockage of polyadenylation at this site required simultaneous removal of the AAUAAA and the adjacent AU-rich region. The residual activity of the adjacent AU-rich regions when the hexanucleotide sequence is artificially removed presupposes that CPSF can interact with the adjacent AU-rich region to some extent, raising the possibility that CPSF may simultaneously interact with it and the AAUAAA sequence when both are present. However, the exact effect that the upstream AU-rich region has on binding of CPSF to the AAUAAA sequence would most likely depend upon constraints imposed by which other factors are present at the time.
The dual hexanucleotide structure is reminiscent of the yeast polyadenylation signal, which, although not as defined as the mammalian poly(A) site, usually consists of an A-rich positioning element and an upstream AU-rich efficiency element (reviewed in 60). The yeast polyadenylation complex shows a surprising homology to the mammalian factors, despite the different sequence recognition (61-63). IgM represents the most ancient of isotypes from which the others have evolved (reviewed in 64,65). The fact that the multiple structure has been retained suggests that it has an important function for this isotype. This also raises the intriguing possibility that the upstream AU-rich regions of the IgM secretory poly(A) site are a remnant from polyadenylation signals in the earliest eukaryotes.
The [mu] secretory poly(A) site appears to be unique in having multiple elements, as the [gamma], [epsis] and [alpha] loci contain a single hexanucleotide sequence and a single optimally placed downstream GU-rich element, if it is present. These different structures raise the possibility that the alternative processing of IgM heavy chain pre-mRNA is regulated in a different way to that of isotypes in which the production of high affinity secreted antibody is the major consideration. IgM represents the first line of defence in the immune response in which both membrane-bound and secreted immunoglobulin play crucial roles. It is conceivable that B cells during a primary immune response have a broader function (e.g. as antigen-presenting cells) as a first line of defence before isotype switching has occurred.
ACKNOWLEDGEMENTS
We would like to thank Walter Dietrich Goetz for donation of the pPKLT55 luciferase construct and Clinton MacDonald for the 3A7 antibodies. Cathy Phillips was supported by a Wellcome Travelling Fellowship and as a Visiting Scientist by the Swedish Medical Research Council. The work was supported by the Wellcome Trust and the Swedish Natural Science Research Council.
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]
J. MA, S. I. GUNDERSON, and C. PHILLIPS Non-snRNP U1A levels decrease during mammalian B-cell differentiation and release the IgM secretory poly(A) site from repression
RNA,
January 1, 2006;
12(1):
122 - 132.
[Abstract][Full Text][PDF]
C. Phillips, N. Pachikara, and S. I. Gunderson U1A Inhibits Cleavage at the Immunoglobulin M Heavy-Chain Secretory Poly(A) Site by Binding between the Two Downstream GU-Rich Regions
Mol. Cell. Biol.,
July 15, 2004;
24(14):
6162 - 6171.
[Abstract][Full Text][PDF]
C. Phillips and S. Gunderson Sequences Adjacent to the 5' Splice Site Control U1A Binding Upstream of the IgM Heavy Chain Secretory Poly(A) Site
J. Biol. Chem.,
June 6, 2003;
278(24):
22102 - 22111.
[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]
Y. Aissouni, C. Perez, B. Calmels, and P. D. Benech The Cleavage/Polyadenylation Activity Triggered by a U-rich Motif Sequence Is Differently Required Depending on the Poly(A) Site Location at Either the First or Last 3'-Terminal Exon of the 2'-5' Oligo(A) Synthetase Gene
J. Biol. Chem.,
September 20, 2002;
277(39):
35808 - 35814.
[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]
K. L. Veraldi, G. K. Arhin, K. Martincic, L.-H. Chung-Ganster, J. Wilusz, and C. Milcarek hnRNP F Influences Binding of a 64-Kilodalton Subunit of Cleavage Stimulation Factor to mRNA Precursors in Mouse B Cells
Mol. Cell. Biol.,
February 15, 2001;
21(4):
1228 - 1238.
[Abstract][Full Text]
J. H. Coyle and D. A. Lebman Correct Immunoglobulin {alpha} mRNA Processing Depends on Specific Sequence in the C{alpha}3-{alpha}M Intron
J. Immunol.,
April 1, 2000;
164(7):
3659 - 3665.
[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]
CiteULike 
Connotea 
Del.icio.us What's this?
