Utilization of the bovine papillomavirus type 1 late-stage-specific nucleotide 3605 3' splice site is modulated by a novel exonic bipartite regulator but not by an intronic purine-rich element

Abstract

Bovine papillomavirus type 1 (BPV-1) late gene expression is regulated at both transcriptional and posttranscriptional levels. Maturation of the capsid protein (L1) pre-mRNA requires a switch in 3' splice site utilization. This switch involves activation of the nucleotide (nt) 3605 3' splice site, which is utilized only in fully differentiated keratinocytes during late stages of the virus life cycle. Our previous studies of the mechanisms that regulate BPV-1 alternative splicing identified three cis-acting elements between these two splice sites. Two purine-rich exonic splicing enhancers, SE1 and SE2, are essential for preferential utilization of the nt 3225 3' splice site at early stages of the virus life cycle. Another cis-acting element, exonic splicing suppressor 1 (ESS1), represses use of the nt 3225 3' splice site. In the present study, we investigated the late-stage-specific nt 3605 3' splice site and showed that it has suboptimal features characterized by a nonconsensus branch point sequence and a weak polypyrimidine track with interspersed purines. In vitro and in vivo experiments showed that utilization of the nt 3605 3' splice site was not affected by SE2, which is intronically located with respect to the nt 3605 3' splice site. The intronic location and sequence composition of SE2 are similar to those of the adenovirus IIIa repressor element, which has been shown to inhibit use of a downstream 3' splice site. Further studies demonstrated that the nt 3605 3' splice site is controlled by a novel exonic bipartite element consisting of an AC-rich exonic splicing enhancer (SE4) and an exonic splicing suppressor (ESS2) with a UGGU motif. Functionally, this newly identified bipartite element resembles the bipartite element composed of SE1 and ESS1. SE4 also functions on a heterologous 3' splice site. In contrast, ESS2 functions as an exonic splicing suppressor only in a 3'-splice-site-specific and enhancer-specific manner. Our data indicate that BPV-1 splicing regulation is very complex and is likely to be controlled by multiple splicing factors during keratinocyte differentiation.

FIG. 1

Mapping of the BP at the nt 3605 3′ splice site by Superscript II RT-PCR and sequencing. (A) Diagram of a lariat-containing splicing intermediate produced from the wt SE2 pre-mRNA 3 (Fig. 3A) and direction of RT by Superscript II. (B) Diagram of the pair of primers designed for PCR and sequencing. (C) Sequence of the PCR products obtained using oZMZ164 as the primer. The numbers are nucleotide positions in the BPV-1 genome. An arrowhead indicates the BP. Note that the BP A is converted to a T. (D) Sequence of the nt 3605 3′ splice site. The BPS is in a box, under which the mammalian BPS consensus sequence is displayed for comparison.

FIG. 2

BPV-1 SE2 functions in vitro as an intronic splicing repressor in an Ad late IIIa pre-mRNA. (A) The maps of the Ad IIIa pre-mRNAs with (pre-mRNAs 1 to 4) or without (pre-mRNAs 5 and 6) a U1 binding site (5′ splice site) (solid box) at the 3′ end. Exons (large boxes) and introns (lines) are indicated. The BP is shown as a vertical line. The 3RE or its substitutions are located within the intron immediately upstream of the BP and are individually labeled in the corresponding box. The numbers below the diagrams are the sizes in nucleotides of the exons and introns. Splicing efficiencies are shown on the right and were calculated as described previously (55) from the splicing gel in panel B. (B) Splicing gel, showing the identities of the corresponding splicing products on the right (from top to bottom: pre-mRNAs, splicing intermediates, fully spliced products, and 5′ exons). The products from each splicing reaction were analyzed by electrophoresis on an 8% polyacrylamide–8 M urea gel. The numbers at the top of the gel indicate the pre-mRNAs in panel A used for splicing.

FIG. 3

In vitro analysis of the role of BPV-1 SE2 in the regulation of the nt 3605 3′ splice site in BPV-1 late pre-mRNAs. (A) BPV-1 SE2 does not function as an intronic splicing repressor in BPV-1 late pre-mRNAs containing no nt 3225 3′ splice site. Structures of the pre-mRNAs with or without a downstream exon 2 5′ splice site (5′ ss) used for in vitro splicing are shown at the top of the panel. The nt 3764 5′ splice site is indicated by an arrowhead, with its splice junction sequence, AG|GC, AG|GU, or AG|GG, shown. Each pre-mRNA has three versions containing a wt (from p3072) or mt (from p3073) SE2 or with SE2 deleted (from p3074). The numbers above the pre-mRNA structures are nucleotide positions in the BPV-1 genome. The numbers below the structures indicate the sizes in nucleotides of exons (open boxes) and introns (lines). The sequences of SE2 and its substitutions, including the Ad 3RE, are shown in the middle. At the bottom of the panel are two splicing gels. The corresponding splicing products are displayed on the right. The products from each splicing reaction were analyzed by electrophoresis on an 8% polyacrylamide–8 M urea gel. The numbers at the top of the gel correspond to each pre-mRNA used for splicing. Splicing efficiencies are shown at the bottom of the gel and were calculated as described previously (55). (B) BPV-1 SE2 and Ad 3RE stimulate cryptic splicing in BPV-1 late pre-mRNAs. A consensus 5′ splice site is located at nt 3646 in these pre-mRNAs. The pre-mRNAs were transcribed from p3072(wt SE2), p3073(mt SE2), and p3074(Ad 3RE) DNA templates, respectively. Other descriptions are the same as for panel A.

FIG. 4

In vivo functional analysis of the BPV-1 SE2 and Ad 3RE in BPV-1 late pre-mRNAs. (A) Structure of the BPV-1 late minigene transcription unit expression vectors and the splicing patterns of the pre-mRNAs expressed from these vectors. At the ends of the late minigene (open box) are the cytomegalovirus (CMV) IE1 promoter (solid box) and pUC18 (shaded box). The numbers below the lines are the nucleotide positions in the BPV-1 genome. Early and late poly(A) sites are indicated as A_E and A_L above the lines. A large deletion in the intron region of the genome is shown by a heavy vertical line. Below the minigene diagram are the structures of the minigene transcripts. The relative positions of SE1, ESS, and SE2 are shown, as well as the nucleotide positions of the 5′ splice site (5′ss) and 3′ splice site (3′ss). Two pairs of primers used for the RT-PCR assays shown in panels B and C are indicated by arrows under the transcripts and named by the locations of their 5′ ends. The following plasmids were used for transfection of 293 cells: p3231(wt 3225 3′ ss + wt SE2), p3033(wt 3225 + SE2m), p3084(wt 3225 3′ ss + 3RE), p3072(Δ3225 3′ss + wt SE2), p3085(Δ3225 3′ss + 3RE), p3073(Δ3225 3′ss + SE2m), and p3074(Δ3225 3′ss + SE2d) (see Materials and Methods for construction of the plasmids). The nucleotide sequences of wt and mt SE2 and Ad 3RE in the pre-mRNAs expressed from the plasmids described above are shown in Fig. 3A. (B and C) RT-PCR analysis of BPV-1 late mRNAs transcribed and spliced in 293 cells from the expression vectors containing wt or mt SE2 or 3RE with or without the nt 3225 3′ splice site as indicated above each gel. Total cell RNA was extracted and digested with RNase-free DNase I before RT-PCR analysis. Several controls were included for the assays, including p3231(wt)-transfected 293 cell RNAs, untransfected 293 cell RNA, and water controls. Total cell RNA extracted from p3231(wt)-transfected 293 cells but not treated with RNase-free DNase I [WT (−RT)] and BPV-1 L1 cDNA and a 10:1 mixture of L2-L and L2-S cDNA were also used as templates for PCR amplification in panel B. The predicted sizes of the RT-PCR products differ between the panels due to the use of different sets of primers: Pr7345 and Pr3715 for panel B and Pr7345 and Pr3746 for panel C.

FIG. 5

Identification of a novel bipartite splicing regulatory element downstream of the nt 3605 3′ splice site. (A) Maps of the pre-mRNAs with various lengths of exon 2. The numbers above the pre-mRNAs are nucleotide positions in the BPV-1 genome. The numbers below the pre-mRNAs indicate the sizes in nucleotides of exons (open boxes) and introns (lines). Each pre-mRNA has either a wt SE2 (from p3072) or an mt SE2 (from p3073). The splicing efficiency of each pre-mRNA was calculated as described previously (53) from the splicing gel in panel B. (B) Splicing gel, showing the corresponding splicing products on the right. The products from each splicing reaction were analyzed as for Fig. 2B. The numbers at the top of the gel correspond to each pre-mRNA with wt or mt SE2 in panel A used for splicing.

FIG. 6

Functional analysis of SE4 and its downstream sequence ESS2 for selection of a heterologous nt 3225 3′ splice site for in vitro splicing. (A) Structures of the pre-mRNAs transcribed from plasmid p3030-derived DNA templates. The following pre-mRNA-specific 3′ primers were used: oZMZ84 (pre-mRNA 1), oZMZ102 (pre-mRNA 2), oZMZ168 (pre-mRNA 3), oZMZ169 (pre-mRNA 4), oZMZ170 (pre-mRNA 5), oZMZ171 (pre-mRNA 6), oZMZ172 (pre-mRNA 7), and oZMZ173 (pre-mRNA 8). Pre-mRNA 1 differs from pre-mRNAs 3 to 7 only in its SE1 region. The latter five pre-mRNAs had SE1 replaced by individual sequences as shown below the map. The ACE motif is underlined. The numbers above the sequences indicate nucleotide positions in the BPV-1 genome. The dots indicate unchanged nucleotides. Pre-mRNA 2 differs from pre-mRNA 8 by two different fragments downstream of SE1, boxed in the map. Other descriptions are the same as for Fig. 5A. (B) Splicing gel, showing the corresponding splicing products on the right. The products from each splicing reaction were analyzed as in Fig. 2B. The numbers at the top of the gel correspond to each pre-mRNA in panel A used for splicing. The splicing efficiency (% spliced) at the bottom of the gel was calculated as described previously (53).

FIG. 7

Identification of a UGGU splicing suppressor motif. (A) Structures of two different sets of BPV-1 late pre-mRNAs used for splicing. Two sets of the pre-mRNAs transcribed, respectively, from PCR templates generated from plasmid p3072 (top; the pre-mRNAs with a 3605 3′ splice site) and p3030 (bottom; the pre-mRNAs with a 3225 3′ splice site) share the same exon 1 but differ in their introns and exon 2. The following pre-mRNA-specific 3′ primers were used: oZMZ160 (pre-mRNA 1), oZMZ161 (pre-mRNA 2), oZMZ185 (pre-mRNAs 3 and 9), oZMZ186 (pre-mRNAs 4 and 10), oZMZ187 (pre-mRNAs 5 and 11), oZMZ184 (pre-mRNAs 6 and 12), oZMZ168 (pre-mRNA 7), and oZMZ183 (pre-mRNA 8). The descriptions of the diagrams are the same as for Fig. 6B. The SE4 in exon 2 is boxed in each diagram. The nucleotide sequence of the SE4, with an italic ACE motif, is shown along with that of the ESS2 or its derivatives in which the UGGU motifs are in boldface and underlined. The splicing efficiency (% spliced) was calculated from the gel in panel B as described previously (53). (B) Splicing gel, showing the corresponding splicing products on the right. The products from each splicing reaction were analyzed as for Fig. 2B. The numbers at the top of the gel correspond to each pre-mRNA in panel A used for splicing.

FIG. 8

Proposed model for the function of the BPV-1 ESEs and ESSs in regulation of BPV-1 alternative splicing. 3′ ss, 3′ splice site; 5′ ss, 5′ splice site; ?, hypothetical factor(s); solid circles, BPS; +, enhances splicing; −, suppresses splicing. Arrows indicate sites at which elements function.