Nucleic Acids Research

Plasmid constructs

PCR products from the secretory poly(A) site and surrounding sequences, containing mutations and a 5[prime] BglII site introduced as part of the synthetic primers, were cloned into pGem T vector. Inserts were subcloned into pPKLT55 (kindly provided by Walter Dietrich-Goetz) (24) between the BglII and NotI sites. The forward primer, 5[prime]-GACAGATCTGATCATGTCTGACACAGGCG-3[prime], was used in combination with either of the following reverse primers (nucleotide positions are numbered according to the mouse IgM sequence with accession no. V00818). For deletions (see also Fig. 1A): µspA1: 5[prime]-GCGTCTAGAGCAGGGATCAGCATTGTATA-3[prime]; µspA2: 5[prime]-GCGTCTAGATAGGGTGGAGGCAAGTATGC-3[prime]; Igsdel(2052): 5[prime]-GCAAGCAAAACACAACTGTCTCAGC-3[prime]; Igsdel(2058): 5[prime]-TGCAGAGCAAGCAAAACACAACTGT-3[prime]; Igsdel(2064): 5[prime]-GGTGTGTGCAGAGCAAGCAAAACAC-3[prime]; Igsdel(2071): 5[prime]-TATGCAGGGTGTGTGCAGAGCAAGC-3[prime]. For mutations in the stem (introduced point mutations are indicated with bold letters) (see also Fig. 2A): Igs(G2056U,C2061A), 5[prime]-TATGCAGGGTTTGTGAAGAGCAAGC-3[prime]; Igs(C2048A,G2042U), 5[prime]-TATGCAGGGTGTGTGCAGAGCAATCAAAAAACAACTGTCTCAGCAG-3[prime]; Igs-(G2056U,C2061A,C2048A,G2042U), 5[prime]-TATGCAGGGTTTGTGAAGAGCAATCAAAAAACAACTGTCTCAGCAG-3[prime]. For mutations of loops (see also Fig. 2A): Igs(mutloop): 5[prime]-TATGCAGGGTGTGTGCTTGCGCAAAACACAACTGT-3[prime]; Igs(C2059CA), 5[prime]-TATGCAGGGTGTTGTGCAGAGCAAGC-3[prime]; Igs(C2059AA), 5[prime]-TATGCAGGGTGTTTTGCAGAGCAAGC-3[prime]; Igs(C2064AA): 5[prime]-TATGCAGTTGTGTGTGCAGAGCAAGC-3[prime].

A

B

Figure 1. (A) Schematic diagram of the sequences surrounding the IgM heavy chain secretory poly(A) site and the deletion mutant constructs used. The multiple hexanucleotide and GU-rich regions have been indicated and are shown as follows: open box, upstream adjacent AU-rich region; black box, consensus hexanucleotide sequence; hatched boxes, proximal and distal GU-rich regions. The nucleotide sequences of the AU-rich region, the consensus hexanucleotide element and the GU-rich regions are shown. The location of the polyadenylation cleavage site is indicated with a vertical arrow. The 55 nt element is indicated with a horizontal arrow. Names of the mutant constructs and diagram indicating the position of their 3[prime] ends are depicted at the bottom. (B) 3[prime] deletion analysis of the 55 nt downstream element. Sequences were cloned into pPKLT55 and transfected into J558L cells, luciferase activity was adjusted for CAT activity and expressed as percentage of Lu µspA2 activity. Bars represent the means of luciferase activity of three transfections experiments ± standard error. Each triplicate experiment was performed three times with similar results. Samples are as indicated.

A

B

C

Figure 2. An intact stem and the two asymmetrically paired internal loops are necessary for the enhancement of polyadenylation activity in vivo. (A) Schematic diagram of the mutations in the stem-loop structure used for assessment of in vivo activity. (I) Those used for the compensatory mutations. (II) Those used for the mutations in the hairpin and asymmetric internal loops. (B) Compensatory mutations which destroy and reconstruct the stem-loop structure. Sequences were cloned into pPKLT55 and transfected into J558L cells and luciferase activity was adjusted for CAT activity and expressed as percentage of Lu µspA2 activity. Bars represent the means of luciferase activity of three transfection experiments ± standard error. Each triplicate experiment was performed three times with similar results. Samples are as indicated. (C) Hairpin and internal loops were mutated. Procedures were the same as for (B). Samples are as indicated.

To generate plasmids belonging to the pBsIgs category the forward primer 5[prime]-GACAGATCTGTTGCATTTTATAAAAATTAGAAATAAAAAAAAT-3[prime] was combined with the reverse primers used for Igs(G2056U,C2061A), Igs(C2048A,G2042U), Igs(G2056U,C2061A,C2048A,G2042U) and Igsdel(2071) respectively, during PCR amplication. The resulting PCR products were cloned into Bluescript II KS between the BamHI and SmaI sites. The plasmids obtained were named pBsIgs(G2056U,C2061A), pBsIgs(C2048A,G2042U), pBsIgs(G2056U,C2061A,C2048A,G2042U) and pBsIgswt respectively. An additional plasmid named pBsIgs(U1962,1968,1969,1975G) mutated in the hexanucleotide and the extended hexanucleotide sequence elements was generated similarly using primers 5[prime]-GACAGATCTGTTGCATTTTAGAAAAAGGAGAAAGAAAAAAAAT-3[prime] and the reverse primer Igsdel(2071) during PCR amplification.

RNA substrates

Templates for in vitro complex experiments were obtained by restriction of the `Bs' plasmid constructs with EcoRI. For the structural mapping, templates for in vitro transcription were obtained by PCR amplification of plasmid G1ms359 (25,26) using the oligonucleotide 5[prime]-GAAATTAATACGACTCACTATAGCCTGCTGAGACAGTTGT-3[prime] [containing the T7 RNA polymerase promoter at its 5[prime] end (27) and nucleotides 2025-2076 of the IgM secretory locus at its 3[prime] end] combined with either of the following reverse primers: wild-type stem-loop: 5[prime]-GGCAAGTATGCAGGGTG-3[prime]; open stem-loop: 5[prime]-GGCAAGTATGCAGGGTTTGTGAAGAGCAAGCAAAACA-3[prime]; closed stem-loop: 5[prime]-GGCAAGTATGCAGTTGTGTTTTGCAGAGCAAGCAAAACA-3[prime].

Capped uniformly labelled RNA substrates were synthesized by in vitro transcription using T7 RNA polymerase (Stratagene) according to the manufacturer's instructions in the presence of 50 µCi [[alpha]-³²P]UTP (40 Ci/mmol in the transcription mixture). Transcripts were purified by extraction after electrophoresis on 3.5% polyacrylamide gels containing 7 M urea. Capped and 3[prime] end labelled RNA substrates were synthesized by the same protocol omitting [[alpha]-³²P]UTP. Obtained RNA was purified by Sephadex G50 spun column chromatography, followed by phenol/chloroform extraction and ethanol precipitation. Subsequently the purified RNA was labelled by 5[prime]-³²pCp at the 3[prime] end using T4 RNA ligase. 3[prime] End labelled RNA was purified as above.

Cell culture and transfection

J558L (28) 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 (29) as previously described (24). Transfection efficiency (24) was measured by co-transfection of SVCATb (30).

Gel retardation assay

Reactions (10 µl) were assembled on ice having final concentrations: 25 mM Tris-HCl buffer system pH 8.3, 10% (v/v) glycerol, 1% (w/v) polyvinyl alcohol, 50 mM KCl, 2 mM MgCl₂, 0.1 mM EDTA, 0.2 mg/ml BSA, 0.01% NP-40, 2 mM DTT, 1 mM ATP and 5 U ribonuclease inhibitor. In each reaction ³²P-radiolabelled RNA substrate (400 000 c.p.m.) was added. Partially purified CPSF (31) was added to each reaction followed by HeLa cell nuclear extract as a source of CstF. Polyadenylation complexes were formed for 20 min at 30°C. Five microlitres of the reaction mixture were loaded on a native 3% polyacrylamide/agarose (0.5%) composite gel as described (7).

RNA secondary structure

All RNA substrates were first refolded by heating in reaction buffer at 70°C for 2 min followed by cooling to 37°C over a 30 min period.

For RNase T1 and lead (II) cleavage, 10 000 c.p.m. of 3[prime] end labelled RNA substrate was incubated at 37°C with either 0.001 U of RNase T1 in 20 mM Tris-acetate pH 7.5, 10 mM Ag(OAc)₂ and 100 mM NaCl for 5 min or 0.1-0.5 mM Pb(OAc)₂ in 50 mM Tris-HCl pH 7.5, 100 mM NH₄Cl and 10 mM MgCl₂ for 10 min. The reactions were stopped by the addition of 10 M urea, 25 mM EDTA. Samples were analyzed on 15% PAGE containing 7 M urea. A reference nucleotide ladder was made by boiling of RNA substrates with 50 mM NaOH, 10 mM EDTA for 10 s.

For RNase H probing, 100 000 c.p.m. of uniformly radiolabelled RNA substrate was refolded in 0.2 M NaCl, 10 mM MgCl₂, 25 mM Tris-HCl, pH 7.5, 1 mM DTT. 0-50 pM of oligo (5[prime]-GAGCAA-3[prime]) or 0-800 pM of oligo (5[prime]-AACACA-3[prime]) complementary to the RNA at positions 2049-2054 or 2039-2044, respectively, were added and allowed to hybridize for 10 min before the addition of 3 U of RNase H. Reaction mixtures were incubated for a further 30 min at 37°C and products were analyzed on 15% PAGE containing 7 M urea. For mobility on native gels, 100 000 c.p.m. of uniformly radiolabelled substrate were refolded in 30 mM Tris-HCl pH 7.8, 10 mM MgCl₂ and 270 mM KCl and immediately loaded onto 15% native polyacrylamide gels (80:1).

Sequence comparisons

Sequence searches were carried out by connection to the NCBI BLAST Email Server at `NCBI BLAST E-Mail Server' (blast@ncbi.nlm.nih.gov ) (32). Comparisons were carried out using the COMPARE and DOTPLOT programs of the `gcg' program (33).

Sequences 3[prime] of the distal GU-rich region are necessary for full polyadenylation activity in vivo

We have previously shown that a 55 nt element containing a distal GU-rich region enhances polyadenylation activity at the secretory poly(A) site (23,24) using transfection of luciferase reporter constructs. A similar reporter system has been previously used to probe the function of a `consensus' downstream sequence (34). In order to more precisely define the sequences necessary for the enhancing effect of the element spanning the distal GU-rich region of the IgM secretory poly(A) site, we generated a series of 3[prime] deletion mutants (Fig. 1). These constructs, as well as plasmids Lu pµspA1 and Lu pµspA2, were transfected into J558L cells and the luciferase activities were subsequently determined. Figure 1B shows that a sequence located 20 nt beyond the GU-rich sequence of this 55 nt element was necessary to retain the enhancing activity of the whole element, as deletions more proximal than position 2071 abolished the enhancing activity (Fig. 1B, compare samples 3 with 1 and 2).

An intact stem-loop structure is necessary for the enhancing activity

RNA folding predictions suggested that the distal GU-rich region could form a stem-loop structure with the nucleotides located downstream of it (Fig. 2A). We therefore examined whether the putative secondary structure might have a role in the biological function of this element since these nucleotides together with the GU-rich region were necessary for full luciferase activity. For this we generated mutants predicted to destroy and restore the stability of the stem-loop (Fig. 2). Mutations which destroyed the stem-loop structure on either side of the stem reduced the enhancing effect of the element to the level of Lu pµspA1 lacking the whole 55 nt element (Fig. 2B, compare samples 1 and 2 with samples 4 and 5). However, when these mutations were combined to reconstruct the predicted stem-loop structure the enhancing effect was restored (Fig. 2B, sample 3). Thus, we conclude that the presence of a stem-loop structure is important for the enhancing activity of the element. Furthermore, the compensatory mutational analysis suggested that nucleotides G2042 and C2048 interacted with nucleotides C2061 and G2056, respectively.

Two asymmetric internal loops contribute to in vivo polyadenylation efficiency

As hairpin and internal loops have been previously shown to be the main sites of protein binding (reviewed in 35-37), we examined how these structures might contribute to the activity of the putative stem-loop in vivo. We designed mutated substrates to either exchange the entire loop for the stable tetraloop sequence GCAA, or closed each internal loop in turn by pairing across the stem (Fig. 2A). The mutant constructs were transfected into J558L cells and luciferase activities were measured as before. Whereas changing the loop had little or no effect on the activity of the stem-loop structure (Fig. 2C, compare samples 1 and 2), partial or full closing of either internal loop reduce the enhancing activity of this element (Fig. 2C, samples 3, 4 and 5). These data suggest that the putative internal loops are important for full activity while the primary sequence of the hairpin loop is relatively less important.

The stem-loop structure contributes to polyadenylation complex formation in vitro

The importance of the stem-loop structure for CPSF/CstF polyadenylation complex formation was investigated by gel-retardation assays. In order to ensure that the formation of the CPSF/CstF complex was dependent on the downstream binding factor, CstF, and not the hexanucleotide interacting factor, CPSF, we provided CPSF in excess. To do this we first established the conditions in which a complex between the RNA substrate and CPSF could be formed, in the absence of CstF. HeLa cell nuclear extract was then added in increasing amounts to provide CstF necessary for stable complex formation. Under these conditions complex formation is dependent on the CstF factor binding to the downstream elements and differences among the mutated substrates could be detected. Figure 3 shows that complexes formed easily on the wild-type RNA substrate (BsIgswt), while the RNA substrate containing mutations in the descending arm of the stem-loop BsIgs(G2056U,C2061A) formed complexes poorly (Fig. 3, lanes 10 and 1-3 respectively). Furthermore, the effect of the two point mutations in BsIgs(G2056U,C2061A) could be suppressed by the compensatory mutations in RNA substrate BsIgs(G2056U,C2061A,C2048A,G2042U) restoring the stem-loop structure (Fig. 3, lanes 7-9). We conclude from this that the stem-loop structure of this element is important for complex formation. However, the RNA substrate BsIgs(C2048A,G2042U) in which the ascending arm of the stem-loop was mutated, formed complexes which were only slightly reduced in comparison with the wild-type substrate (Fig. 3, lanes 4-6 and 10). The formation of a complex on the hexanucleotide mutated RNA substrate, BsIgs(U1962,1968,1969,1975G) (Fig. 3, lane 11) was most likely due to the presence of excess CPSF in these reactions, since this complex was not formed unless CPSF was added in excess together with CstF (data not shown).

Figure 3. The stem-loop structure is important for CPSF/CstF complex formation in vitro. Sequences were cloned into bluescript KS and in vitro transcribed uniformly radiolabelled substrates were allowed to form complexes with excess semi-purified CPSF followed by decreasing amounts of HeLa cell nuclear extract as indicated. Complexes were visualized in gel-retardation assays. RNA substrates used are as follows: lanes 1-3, stem-loop mutation decending arm BsIgs(G2056U,C2061A); lanes 4-6, stem-loop mutation ascending arm BsIgs(C2048A,G2042U); lanes 7-9, stem-loop mutation compensatory BsIgs(G2056U,C2061A,C2048A,C2042U); lane 10, wild-type stem loop, wild-type hexanucleotide sequence BsIgswt; and lane 11, wild-type stem-loop, mutated hexanucleotide BsIgs(U1962,1968,1969,1975G).

Lead (II) cleavage of the stem-loop structure

Lead (II) cleavage patterns have been used to monitor changes in RNA secondary structure induced by mutations (38-40) and as a technique to probe tertiary structures of RNA (41). Furthermore, lead (II) preferably cleaves at bulges of RNA structures independent of nucleotide sequence (42). Thus, lead cleavage patterns can provide more structural information than RNase T1, which cuts only after guanosine. To monitor the changes in folding brought about by the mutations designed to disrupt the putative secondary structure formed between the GU-rich region and the following 20 nt, we prepared three RNA constructs representing: (i) the wild-type stem-loop; (ii) an open stem-loop construct where the stem is predicted to be destroyed by two point mutations; and (iii) a closed stem-loop structure in which the two predicted internal loops are paired across the stem (Fig. 4C). The latter construct helps to locate the loop, as the loop is predicted to be the sole open structure in this structure. These RNA constructs were labelled at their 3[prime] ends and probed by lead (II) cleavage. The cleavage products were analyzed by PAGE and the obtained cleavage patterns of the three RNA constructs were compared. A summary of cleavage sites and putative secondary structures are pictorially represented in Figure 4C. Lead (II) can be seen to cleave at positions in the predicted hairpin and internal loops.

A

B

C

Figure 4. Lead (II) and RNase T1 cleavage of RNA constructs spanning the putative stem-loop structure. (A) 3[prime] end labelled wild-type, closed and open stem-loop RNA constructs, were incubated with increasing concentrations of lead (II) and run on 15% PAGE containing 7 M urea (lanes 1-12). A reference ladder was constructed by boiling of the wild-type RNA construct in 50 mM NaOH for 10 s (lane 13). The wild-type RNA construct was partially digested with 0.001 U of RNase T1 (sample 14). The positions of G residues have been indicated to the right. (B) 3[prime] end labelled RNA substrates were incubated with increasing concentration of lead (II) and run on 20% PAGE containing 7 M urea. The positions of three representative nucleotides have been indicated to the right. (C) A pictorial representation of RNase T1 (open arrow) and lead (II) cleavage (black arrows). Size of arrows represent strong and weak cuts. The putative secondary structures of the three indicated RNA constructs are shown.

Lead (II) was only able to cleave the closed stem-loop construct in the region of the hairpin loop (Fig. 4A, lanes 5-8; B, lanes 4-6; and C). This construct is 2 nt longer, resulting in a different mobility on PAGE. Furthermore, the cuts in the hairpin loop of the closed stem-loop construct are stronger on the 3[prime] side while the cuts in the hairpin loop of the wild-type stem-loop construct are stronger on both the 3[prime] and the 5[prime] sides of the loop and weaker in the center at UGC (Fig. 4A, lanes 1-4; B, lanes 7-9; and C]. The RNase T1 digest of the wild-type stem-loop shows a stronger cut at this G (Fig. 4A, lane 14). Thus, the RNase T1 cleavage pattern is consistent with the conclusion that this G is located within the loop of the wild type stem-loop structure.

Comparing the 5[prime] section of the wild-type stem-loop with that of the open stem-loop construct after 15% PAGE (Fig. 4A, compare lanes 1-4 with 9-12) shows very similar lead (II) cleavage patterns, although the cuts in the open stem-loop construct are stronger. However, comparing their 3[prime] sections after 20% PAGE (Fig. 4B, compare lanes 7-9 with lanes 1-3), the patterns differ. Nucleotides A2060 and C2061A of the open stem-loop are inaccessible to lead (II) even at 0.5 mM. Moreover, the 5[prime] adjacent GCAC sequence (position 2056-2059) is more accessible to lead (II) cleavage in the open stem-loop construct than in the wild-type stem-loop structure. Finally, the 5[prime] section of the wild-type stem-loop construct in general is more susceptible to lead (II) cleavage than the 3[prime] section, while this is not true for either of the open or closed stem-loop constructs.

RNase H probing of the stem-loop structure

In order to examine the relative openness of the wild-type stem-loop, the open stem-loop and the closed stem-loop RNA constructs, we performed RNase H cleavage using two hexamer DNA oligomers, complementary to nucleotides 2049-2054 of the hairpin loop, and nucleotides 2039-2044 of the stem between the two internal loops, as the probes (Fig. 4C).

Figure 5A shows that the accessibility of the oligonucleotide complementary to the hairpin loop sequence, as detected by RNase H cleavage, was ordered thus: the open stem-loop (lanes 9-12) being most accessible, the wild-type stem-loop (lanes 1-4) intermediary and the closed stem-loop (lanes 5-8) least accessible.

A

B

Figure 5. RNase H probing of the wild-type, closed and open stem-loop RNA constructs. Refolded RNA constructs were probed for susceptibility to oligonucleotide hybridization and subsequent RNase H cleavage using (A) oligonucleotide (5[prime]-GAGCAA-3[prime]) complementary to the RNA substrates at position 2049-2054 or (B) oligonucleotide (5[prime]-AACACA-3[prime]) complementary to position 2039-2044. Oligonucleotide concentrations are indicated above the lanes. Unreacted substrate and cleaved products are indicated with arrows.

Figure 5B shows that the oligonucleotide complementary to the paired section of the stem in the wild-type stem-loop construct could access the stem at high concentrations (Fig. 5B, lanes 1-4) while the stem of the open stem-loop was accessible already at lower concentrations of oligonucleotide (Fig. 5B, lanes 9-12). It appears that this oligonucleotide can bind to other GU-rich sequences of the open stem-loop structure, as a number of cleavage products appeared above and below the expected cleavage products (Fig. 5B, lanes 9-12). No prominent cleavage products were seen, even at high concentration of oligonucleotide, when the stem of the closed stem-loop structure was probed (Fig. 5B, lanes 5-7).

Taken together, results from Figure 5A and B are consistent with the prediction that the wild-type stem-loop construct contains a relatively accessible hairpin loop between positions 2049 and 2055 and a partially basepaired stem structure below this hairpin loop. This is based on the finding that the open stem-loop construct was more accessible to oligonucleotide binding both in the region of the hairpin loop and the stem compared to the wild-type stem-loop construct, and that the stem of the closed stem-loop construct was much less accessible to oligonucleotide hybridization compared with the wild-type stem-loop construct.

Mobility on native gels suggest that the stem-loop structure is bent

Asymmetric internal loops can often cause bending of the structure depending upon how the nucleotides within the loop are stacked (43). Different relative electrophoretic mobility of bent structures under denaturing or native conditions provide a simple assay for this possibility (44). We therefore compared the mobilities of the three RNA stem-loop constructs on denaturing and native PAGE. On denaturing PAGE the wild-type stem-loop construct ran with the same mobility as the open stem-loop construct and faster than the closed stem-loop structure which is 2 nt longer (Fig. 6A, compare lanes 1, 2 and 3). However, on native PAGE, the wild-type stem-loop construct ran with the slowest mobility (Fig. 6B, compare 1, 2 and 3). Remarkably, the wild-type stem-loop, in spite of being 2 nt shorter, ran with a slower mobility than the closed stem-loop structure. This suggests that the wild-type stem-loop structure is bent compared to the open or closed stem-loop constructs.

Figure 6. Mobility of wild-type, closed and open stem-loop RNA constructs on (A) denaturing and (B) native PAGE. Lanes are as indicated.

Structural probing of stem-loop structure within the context of the IgM secretory poly(A) site

We used partial RNase T1 digestion to examine the structure of the stem-loop within the context of the full IgM secretory poly(A) site from position 1838 to 2085 (Fig. 7). RNase T1 cleavage cannot provide the detailed information of secondary structure obtained by lead cleavage patterns of shorter structures. However, lead cleavage patterns becomes more difficult to interpret when longer substrates involving more than one structure are used. Partial RNase T1 digestion of the IgM secretory poly(A) site revealed that the hexanucleotide sequence, the cleavage site and the proximal GU-rich region were located in the most RNase T1 accessible region with the ascending GU-rich arm of the stem and the loop providing a further RNase T1 accessible region downstream of this (Fig. 7). Furthermore, the descending arm of the downstream stem-loop structure is relatively RNase T1 inaccessible compared with the ascending arm as was found to be the case for the isolated structure (compare Figs 4 and 7).

Figure 7. RNase T1 digestion of the IgM secretory poly(A) site from position 1838 to 2085. 3[prime] end labelled substrates were folded and subjected to partial RNase T1 digestion. Substrates in lanes 1-3 were subjected to OH-digestion for 10, 20 and 30 s respectively. Lane 4 is the undigested substrate. Substrates in lanes 5-10 were subjected to 0.4, 0.2, 0.1, 0.04, 0.02 and 0.01 U of RNase T1 for 5 min in a total volume of 5 µl. Stop buffer was added and samples run immediately on 15% PAGE.

The downstream sequence and the stem-loop structure are conserved

We were interested to compare sequences downstream of the hexanucleotide sequence of the µ-secretory poly(A) site with other mammalian species to look for conservation of particular sequence elements. Unfortunately, although cDNAs are available for a number of species, mammalian genomic sequences are only available for hamster and human in addition to mouse. A summary of these comparisons is presented in Figure 8. Overall the mouse and hamster sequences are 80% similar, while the mouse and human sequences are 63% similar. A highly conserved 14 nt long sequence element, TGCAC(G, A or C)CACCCTGC, was found in all three species immediately downstream of the distal GU-rich region. This 14mer element corresponds to the descending arm of the mouse stem-loop structure which basepairs with the preceding GU-rich region. The region preceding the highly conserved 14mer element, which includes the GU-rich region, is conserved between mouse and hamster. However, in human the corresponding sequence appears to have been deleted and the preceding sequence is highly mutated relative to the hamster and mouse (Fig. 8A).

A

B

Figure 8. Comparison of sequences downstream of the hexanucleotide sequence of the µ-secretory poly(A) site. (A) Alignment of the hamster, mouse and human primary nucleotide sequences downstream of the hexanucleotide element. Sequences which are conserved between species are connected by grey shading. The bold lines representing the sequences are broken in regions missing nucleotides compared with the hamster sequence. The black, diagonally hatched and vertically hatched boxes indicate the location of the hexanucleotide, GT-rich regions and the conserved downstream sequence element, respectively. The arrow indicates the location of the polyadenylation cleavage site. Below the overall comparision, the indicated region containing the downstream stem-loop is compared at the nucleotide level. The GT-rich regions are indicated by diagonally hatched boxes. The conserved downstream sequence is written in bold letters and marked by a vertically hatched box. (B) Predicted folding of the distal downstream element of the human, mouse and hamster µ-secretory poly(A) sites. The sequence which is highly conserved between mouse, hamster and human and is in the descending arm of the stem-loop is marked in bold. Nucleotides in the human and hamster sequence which differ from that of the mouse are marked with open letters.

We next investigated whether a stem-loop structure similar to that in mouse could be found downstream of the hamster and human secretory poly(A) sites. As can be seen in Figure 8B, both the hamster and human sequences could form similar stem-loop structures to that of the mouse. In all cases the GU-rich regions are basepaired with the conserved 14mer element. The hamster stem-loop is almost identical to the mouse counterpart, while the human stem-loop structure differs. In the human structure the ascending arm is now formed by the mutated G-rich sequence. Thus, stem-loop structures can be folded downstream of the mouse, hamster and human IgM secretory poly(A) sites.

DISCUSSION

We have shown here that an intact stem-loop structure is necessary for the biological activity of the distal downstream element of the murine IgM secretory poly(A) site, which include a GU-rich element and a conserved element as ascending and descending arms of the stem-loop respectively. The stem-loop structure contains two asymmetric internal loops both of which contribute to activity. Furthermore, CPSF/CstF complex formation in vitro is disrupted by mutations in the descending arm of the stem-loop which destroy the structure and restored by compensatory mutations which reform the structure. Structural probing, using lead (II) cleavage, RNase H probing or RNase T1 digestion of the stem-loop structure, are consistent with the predicted folding. Native gel mobility analysis suggested that the stem-loop structure may bend, possibly due to the stacking of the nucleotides within the two asymmetric internal loops.

There are numerous examples where hairpin loops, internal loops and bulges represent targets for site-specific recognition of RNA by proteins (45,46). Such structures appear to function to open the major groove and expose unpaired nucleotides for recognition by proteins (36,47). The unpaired nucleotides in the internal asymmetric loops may function to increase the accessibility of these sequences to interact with components of the polyadenyl-ation machinery. Furthermore, the distortability introduced by the bend in the stem-loop structure may make the protein-RNA interaction thermodynamically more favorable (36,37,48). In addition, the strong downstream stem-loop structure may help to maintain the area surrounding the hexanucleotide sequence in a non-basepaired state, increasing its accessibility for CPSF binding (49).

The GU-rich sequence is surprisingly underconserved between mouse and human. A clue to this lack of conservation may be its sequence similarity to a part of the U2-IR2 domain of Epstein-Barr virus (EBV), a tumourigenic virus which invades human B cells. The sequence TTGTGTTTTGCTT in the EBV genome, which is identical to the distal GU-rich region of the murine µ-secretory poly(A) site, is located downstream of potential 3[prime] splice sites and immediately precedes the AUG codon in the open reading frame of the U2 protein which has been implicated in tumourigenesis (50). There is evidence that selective pressure by EBV has induced mutations in other human genes (reviewed in 51). Thus it is possible that it was advantageous to modify the distal GU-rich region of the human µ-secretory poly(A) site once EBV became endemic in humans.

The most conserved sequence between mouse, hamster and human within the 55 nt enhancing sequence is the sequence TGCAC(G, A or C)CACCCTGC immediately following the GU-rich sequence in mouse and hamster and predicted to form the descending arm of the stem-loop structure. The conservation of the TGCAC(G, A or C)CACCCTGC sequence between species including humans, in which the ascending arm of the stem-loop is not conserved, raises the possibility that the formation of a secondary structure may not be the only function of this conserved sequence. The two point mutations on the 3[prime] arm of the stem which disrupt the conserved sequence reduce polyadenylation activity in vivo as well as reduce complex formation in vitro, whereas the two point mutations on the 5[prime] side of the stem, while reducing in vivo activity, have only a minor affect on complex formation. This raises the possibility that the conserved sequence may play an extra role in complex formation. Alternatively, the stem-loop structure may form differently in vivo and in vitro under the conditions used here. In this case, it is possible that mutations on the 5[prime] arm may disrupt this structure less severely in vitro than in vivo than those in the 3[prime] arm. Furthermore, activity is restored both in vivo and in vitro when these mutations are combined even though this construct still contains mutations in the conserved sequence. This argues that it is the secondary structure which is important here. Furthermore, if the GU-rich region has been deleted in the human under selective pressure, it is relevant that the sequence preceding the GU-rich sequence deletion is heavily mutated relative to the surrounding sequences. These heavily mutated sequences can form a stem-loop structure with the conserved sequence resulting in a structure which bears a great deal of resemblance to the murine structure.

The IgM secretory poly(A) site contains dual upstream and downstream sequence elements, the latter consisting of two GU-rich elements which are both suboptimally placed and both required for the formation of the pre-initiation complex in vitro (23). That the distal GU-rich element is presented as the ascending arm of a bent stem-loop structure may be an important feature making the region more efficiently recognized by the polyadenylation machinery, thus compensating for its suboptimal location. We have proposed that the dual structure may provide a flexibility to accommodate regulatory events at this particular poly(A) site (23). The stem-loop structure may also provide novel possibilities for regulatory events as the target for structure modifying activities such as RNA helicases (52), single-strand RNA binding proteins, such as HnRNP proteins (53) or strand-annealing activities (54). It will be an important task for future work to identify the protein or protein factors that specifically interact with the stem-loop structure. Obvious candidates are the basal polyadenylation factor CstF, as the 55 nt element spanning this stem-loop structure is necessary for CPSF/CstF complex formation in vitro (23), and the inducible 30 kDa polypeptide that we previously have identified (24).

ACKNOWLEDGEMENTS

We would like to thank Walter Dietrich Goetz for the donation of the pPKLT55 luciferase construct and Clinton MacDonald for the 3A7 antibodies. We would also like to thank Leif Kirsebom, Gerhart Wagner and Devapriya Choudhury for helpful discussions and suggestions throughout the project, and Hans Johansson for helpful comments on the manuscript. C.P. was supported by a Wellcome Travelling Fellowship and as a gästforskare by the Swedish Medical Research Council. C.B.K. was supported by an EU-TMR fellowship. The work was supported by the Wellcome Trust, the Swedish Cancer Society and the Swedish Natural Science Research Council.

REFERENCES

1. Manley,J.L. (1988) Biochim. Biophys. Acta, 950, 1-12. MEDLINE Abstract

2. McDevitt,M.A., Hart,R.P., Wong,W.W. and Nevins,J.R. (1986) EMBO J., 5, 2907-2913. MEDLINE Abstract

3. Proudfoot,N. (1991) Cell, 64, 671-674. MEDLINE Abstract

4. Wahle,E. and Keller,W. (1996) Trends Biochem. Sci., 21, 247-250. MEDLINE Abstract

5. Gilmartin,G.M. and Nevins,J.R. (1991) Mol. Cell. Biol., 11, 2432-2438. MEDLINE Abstract

6. Murthy,K.G.K. and Manley,J.L. (1992) J. Biol. Chem., 267, 14804-14811. MEDLINE Abstract

7. Rüegsegger,U., Beyer,K. and Keller,W. (1996) J. Biol. Chem., 271, 6107-6113. MEDLINE Abstract

8. Keller,W. (1995) Cell, 81, 829-832. MEDLINE Abstract

9. Manley,J.L. (1995) Curr. Opin. Gen. Dev., 5, 222-228.

10. Alt,F.W., Bothwell,A.L.M., Knapp,M., Siden,E., Mather,E., Koshland,M. and Baltimore,D. (1980) Cell, 20, 293-301. MEDLINE Abstract

11. Early,P., Rogers,J., Davis,M., Calame,K., Bond,M., Wall,R. and Hood,L. (1980) Cell, 20, 313-319. MEDLINE Abstract

12. Rogers,J., Early,P., Carter,C., Calame,K., Bond,M., Hood,L. and Wall,R. (1980) Cell, 20, 303-312. MEDLINE Abstract

13. Perry,R.P., Atchison,M.L., Kelley,D.E. and Peterson,M.L. (1988) Ann. N.Y. Acad. Sci., 546, 25-33. MEDLINE Abstract

14. Peterson,M.L. and Perry,R.P. (1986) Proc. Natl Acad. Sci. USA, 83, 8883-8887. MEDLINE Abstract

15. Peterson,M.L. and Perry,R.P. (1989) Mol. Cell. Biol., 9, 726-738. MEDLINE Abstract

16. Peterson,M.L., Gimmi,E.R. and Perry,R.P. (1991) Mol. Cell. Biol., 11, 2324-2327. MEDLINE Abstract

17. Peterson,M.L. (1992) Gene Expr., 2, 319-327. MEDLINE Abstract

18. Peterson,M.L., Bryman,M.B., Peiter,M. and Cowan,C. (1994) Mol. Cell. Biol., 14, 77-86. MEDLINE Abstract

19. Peterson,M.L. (1994) Mol. Cell. Biol., 14, 7891-7898. MEDLINE Abstract

20. Galli,G., Guise,J.W., McDevitt,M.A., Tucker,P.W. and Nevins,J.R. (1987) Genes Dev., 1, 471-481. MEDLINE Abstract

21. Galli,G., Guise,J., Tucker,P.W. and Nevins,J.R. (1988) Proc. Natl Acad. Sci. USA, 85, 2439-2443. MEDLINE Abstract

22. Takagaki,Y., Seipelt,R.L., Peterson,M.L. and Manley,J.L. (1996) Cell, 87, 941-952. MEDLINE Abstract

23. Phillips,C. and Virtanen,A. (1997) Nucleic Acids Res., 25, 2344-2351. MEDLINE Abstract

24. Phillips,C., Schimpl,A., Dietrich-Goetz,W., Clements,J.B. and Virtanen,A. (1996) Eur. J. Immunol., 26, 3144-3152. MEDLINE Abstract

25. Berberich,I. and Schimpl,A. (1990) Eur. J. Immunol., 20, 445-448. MEDLINE Abstract

26. Berberich,I. and Schimpl,A. (1992) Eur. J. Immunol., 22, 525-529. MEDLINE Abstract

27. Uhlenbeck,O.C. (1987) Nature, 328, 596-600. MEDLINE Abstract

28. Oi,V.T., Morrison,S.L., Herzenberg,L.A. and Berg,P. (1983) Proc. Natl Acad. Sci. USA, 80, 825-829. MEDLINE Abstract

29. Grosschedl,R. and Baltimore,D. (1985) Cell, 41, 885-897. MEDLINE Abstract

30. Larsson,S., Svensson,C. and Akusjärvi,G. (1992) J. Mol. Biol., 225, 287-298. MEDLINE Abstract

31. Åström,A., Åström,J. and Virtanen,A. (1991) Eur. J. Biochem., 202, 765-773. MEDLINE Abstract

32. Altschul,S.F., Gish,W., Miller,W., Myers,E.W. and Lipman,D.J. (1990) J. Mol. Biol., 215, 403-410. MEDLINE Abstract

33. Genetics Group (1994) Program Manual for the Winconsin Package, v.8.

34. McLauchlan,J., Gaffney,D., Whitton,J.L. and Clements,J.B. (1985) Nucleic Acids Res., 13, 1347-1368. MEDLINE Abstract

35. Wyatt,J.R. and Tinoco,I. (1993) In Gesteland,R.F. and Atkins,J.F. (eds), RNA World. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 465-496.

36. Steitz,T.A. (1993) In Gesteland,R.F. and Atkins,J.F. (eds), RNA World. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 219-237.

37. Varani,G. and Pardi,A. (1994) In Nagai,K. and Mattaj,I. (eds), RNA-Protein Interactions. IRL Press, Oxford, pp. 1-24.

38. Behlen,L.S., Sampson,J.R., DiRenzo,A.B. and Uhlenbeck,O.C. (1990) Biochemistry, 29, 2515-2523. MEDLINE Abstract

39. Sampson,J.R., Sullivan,F.X., Behlen,L.S., DiRenzo,A.B. and Uhlenbeck,O.C. (1987) Cold Spring Harbor Symp. Quant. Biol., 52, 267-275. MEDLINE Abstract

40. Peterson,E.T. and Uhlenbeck,O.C. (1992) Biochemistry, 31, 10380-10389. MEDLINE Abstract

41. Hüttenhofer,A., Westhof,E. and Böck,A. (1996) RNA, 2, 354-366. MEDLINE Abstract

42. Pan,T. and Uhlenbeck,O.C. (1992) Nature, 358, 560-564.

43. Aboul-Ela,F., Murchie,A.I.H., Homans,S.W. and Lilley,D.M.J. (1993) J. Mol. Biol., 229, 173-181. MEDLINE Abstract

44. Bhattacharya,A., Murchie,A.I.H. and Lilley,D.M.J. (1990) Nature, 343, 484-486.

45. Rould,M.A., Perona,J.J. and Steitz,T.A. (1989) Science, 246, 1135-1142. MEDLINE Abstract

46. Ruff,M., Krishnaswamy,S., Boeglin,M., Poterszman,A., Mitschler,A., Podjarny,A., Rees,B., Thierry,J.C. and Moras,D. (1991) Science, 252, 1682-1689. MEDLINE Abstract

47. Turner,D.H. (1992) Curr. Opin. Struct. Biol., 2, 334-347.

48. Riordan,F.A., Bhattacharya,A., McAteer,S. and Lilley,D.M.J. (1992) J. Mol. Biol., 226, 305-313. MEDLINE Abstract

49. Chen,F. and Wilusz,J. (1998) Nucleic Acids Res., 26, 2891-2898. MEDLINE Abstract

50. Dambaugh,T., Hennessy,K., Chamnankit,L. and Kieff,E. (1984) Proc. Natl Acad. Sci. USA, 81, 7632-7636. MEDLINE Abstract

51. Masucci,M.G. and Ernberg,I. (1994) Trends Microbiol., 2, 125-130. MEDLINE Abstract

52. Bass,B.L. (1993) Semin. Dev. Biol., 3, 425-433.

53. Dreyfuss,G., Matunis,M.J., Pinol-Roma,S. and Burd,C.G. (1993) Annu. Rev. Biochem., 62, 289-321. MEDLINE Abstract

54. Portman,D.S. and Dreyfuss,G. (1994) EMBO J., 13, 213-221. MEDLINE Abstract

---

*To whom correspondence should be addressed. Tel: +46 18 4714908; Fax: +46 18 526849; Email: anders.virtanen@medgen.uu.se
⁺Present address: Department of Molecular Biology and Biochemistry, Rutgers University, Nelson Laboratories, Room A322, Piscataway, NJ 08855, USA

---

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: 23 Dec 1998
Copyright©Oxford University Press, 1998.
CiteULike    Connotea    Del.icio.us    What's this?

This article has been cited by other articles:

				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]

				D. Oberg, J. Fay, H. Lambkin, and S. Schwartz A Downstream Polyadenylation Element in Human Papillomavirus Type 16 L2 Encodes Multiple GGG Motifs and Interacts with hnRNP H J. Virol., July 15, 2005; 79(14): 9254 - 9269. [Abstract] [Full Text] [PDF]

				J. C. Loke, E. A. Stahlberg, D. G. Strenski, B. J. Haas, P. C. Wood, and Q. Q. Li Compilation of mRNA Polyadenylation Signals in Arabidopsis Revealed a New Signal Element and Potential Secondary Structures Plant Physiology, July 1, 2005; 138(3): 1457 - 1468. [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]

				M. Paulus, M. Haslbeck, and M. Watzele RNA stem-loop enhanced expression of previously non-expressible genes Nucleic Acids Res., May 26, 2004; 32(9): e78 - e78. [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]

				H. Hans and J. C. Alwine Functionally Significant Secondary Structure of the Simian Virus 40 Late Polyadenylation Signal Mol. Cell. Biol., April 15, 2000; 20(8): 2926 - 2932. [Abstract] [Full Text]

				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 (511K) 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 (17) Commercial Re-use Guidelines for Open Access NAR Content Google Scholar Articles by Phillips, C. Articles by Virtanen, A. Search for Related Content PubMed PubMed Citation Articles by Phillips, C. Articles by Virtanen, A. Social Bookmarking          What's this?