optimization of a weak 39 splice site counteracts the...

JOURNAL OF VIROLOGY,0022-538X/00/$04.0010

July 2000, p. 5902–5910 Vol. 74, No. 13

Copyright © 2000, American Society for Microbiology. All Rights Reserved.

Optimization of a Weak 39 Splice Site Counteracts the Functionof a Bovine Papillomavirus Type 1 Exonic Splicing

Suppressor In Vitro and In VivoZHI-MING ZHENG,* JESSE QUINTERO, ERIC S. REID, CHRISTIAN GOCKE,

AND CARL C. BAKER

Basic Research Laboratory, Division of Basic Sciences, National Cancer Institute,National Institutes of Health, Bethesda, Maryland 20892

Received 28 January 2000/Accepted 7 April 2000

Alternative splicing is a critical component of the early to late switch in papillomavirus gene expression. Inbovine papillomavirus type 1 (BPV-1), a switch in 3* splice site utilization from an early 3* splice site atnucleotide (nt) 3225 to a late-specific 3* splice site at nt 3605 is essential for expression of the major capsid (L1)mRNA. Three viral splicing elements have recently been identified between the two alternative 3* splice sitesand have been shown to play an important role in this regulation. A bipartite element lies approximately 30 ntdownstream of the nt 3225 3* splice site and consists of an exonic splicing enhancer (ESE), SE1, followedimmediately by a pyrimidine-rich exonic splicing suppressor (ESS). A second ESE (SE2) is located approxi-mately 125 nt downstream of the ESS. We have previously demonstrated that the ESS inhibits use of thesuboptimal nt 3225 3* splice site in vitro through binding of cellular splicing factors. However, these in vitrostudies did not address the role of the ESS in the regulation of alternative splicing. In the present study, wehave analyzed the role of the ESS in the alternative splicing of a BPV-1 late pre-mRNA in vivo. Mutation ordeletion of just the ESS did not significantly change the normal splicing pattern where the nt 3225 3* splice siteis already used predominantly. However, a pre-mRNA containing mutations in SE2 is spliced predominantlyusing the nt 3605 3* splice site. In this context, mutation of the ESS restored preferential use of the nt 32253* splice site, indicating that the ESS also functions as a splicing suppressor in vivo. Moreover, optimizationof the suboptimal nt 3225 3* splice site counteracted the in vivo function of the ESS and led to preferentialselection of the nt 3225 3* splice site even in pre-mRNAs with SE2 mutations. In vitro splicing assays alsoshowed that the ESS is unable to suppress splicing of a pre-mRNA with an optimized nt 3225 3* splice site.These data confirm that the function of the ESS requires a suboptimal upstream 3* splice site. A surprisingfinding of our study is the observation that SE1 can stimulate both the first and the second steps of splicing.

Removal of introns from primary transcripts (pre-mRNAs)by RNA splicing is an essential step in maturation of mosteukaryotic mRNAs. The initial machinery of pre-mRNA splic-ing involves recognition of a 59 splice site by U1 snRNP and a39 splice site by U2 auxiliary factor (U2AF) and then U2snRNP. The conventional 39 splice site consists of three criticalelements: the branch point sequence (BPS), a polypyrimidinetract (PPT), and an AG dinucleotide at the 39 end of theintron. Although the yeast BPS (UACUAAC) is highly con-served and optimized for base pairing with a conserved region(GUAGUA) of U2 snRNA (26), the mammalian consensusBPS (YNYURAY) is fairly degenerate and yet does base pair-ing with U2 snRNA (3, 45, 54). However, many higher eukary-otic cellular and viral transcripts have a nonconsensus (subop-timal or weak) BPS. The PPT is a run of 15 to 40 pyrimidines(usually U’s) that lies between the BPS and the AG dinucle-otide (15, 30, 48). The PPT has binding sites for U2AF, aheterodimer of U2AF65 and U2AF35 (17, 27, 35, 46). Anypurines that are interspersed in the PPT will weaken the bind-ing of U2AF and make the PPT suboptimal (30, 31, 36, 47). Ithas been shown previously that a strong PPT can offset theeffect of a nonconsensus BPS or vice versa and make the

pre-mRNA be spliced more efficiently (3, 11, 22, 24, 34, 37, 38).Downstream of the PPT is an AG dinucleotide. The nature ofthe nucleotide preceding the AG has a profound effect onsplicing. It has been reported previously that GAG (but notCAG, UAG, or AAG) trinucleotides at the 39 splice junctioninhibit the second step of splicing (30, 36). Statistical analysisof the nucleotide preceding the AG dinucleotide has shownthat the yeast 39 splice site has solely CAG (21), whereasmammalian 39 splice sites have either CAG or UAG, but CAGis twice as common as UAG (48).

In general, suboptimal 39 splice sites are often subject toboth positive and negative regulation by a variety of cis splicingelements. Exonic splicing enhancers (ESEs) increase utiliza-tion of a suboptimal splice site by recruiting essential splicingfactors to that site (39, 55). Exonic splicing suppressors (ESS)or silencers are recently described cis elements that inhibitsplicing of pre-mRNAs. ESS elements share few sequencesimilarities. However, they are often found adjacent to an ESEand together form bipartite splicing elements. We have re-cently summarized most of the published ESS elements alongwith their functional core sequences (52). Studies of the mech-anisms by which an ESS functions in vitro and in vivo havedemonstrated that many cellular splicing factors are probablyinvolved in splicing suppression by an ESS. These include SRproteins (53) for the bovine papillomavirus type 1 (BPV-1)ESS, SC35 for the human immunodeficiency virus type 1(HIV-1) tat exon 3 ESS (20), hnRNP H for the rat b-tropo-myosin exon 7 ESS (6), and hnRNP A1 for the fibroblast

* Corresponding author. Present address: Bldg. 10/Rm. 13N240,National Cancer Institute, National Institutes of Health, HIV andAIDS Malignancy Branch, 9000 Rockville Pike-Bethesda, MD 20892.Phone: (301) 594-1382. Fax: (301) 480-8250. E-mail: [email protected].

5902

on May 22, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

http://jvi.asm.org/

growth factor receptor (FGFR) 2 K-SAM ESS and for theHIV-1 tat exon 2 ESS (5, 9). Moreover, the fibronectin EDAESS has been implicated in the maintenance of an RNA con-formation that facilitates display of the adjacent ESE SR pro-tein binding sequences (23).

The papillomaviruses are a family of small DNA tumorviruses which contain an 8-kb circular genome. Infection ofhumans and many animals causes epithelial and fibroepitheliallesions including benign warts or papillomas and invasive can-cers. It has been well documented that the virus life cycle of allfamily members requires a differentiating squamous epithe-lium. BPV-1 has served as the prototype for studies of papil-lomavirus molecular biology for more than 2 decades. Regu-lation of BPV-1 gene expression at the posttranscriptionallevel involves both alternative splicing and alternative polyad-enylation. The majority of the viral early transcripts are pro-cessed using a common 39 splice site at nucleotide (nt) 3225and the early poly(A) site at nt 4203 in the undifferentiatedkeratinocyte at early stages of virus infection. However, mat-uration of the viral late primary transcript to generate themajor capsid (L1) mRNA requires utilization of an alternative39 splice site at nt 3605 and the late specific poly(A) site at nt7175. Previous in situ hybridization studies in our laboratorydemonstrated that the nt 3605 39 splice site is utilized only inthe fully differentiated keratinocytes during late stages of thevirus life cycle (2). Further investigations into the molecularmechanisms that regulate BPV-1 alternative splicing lead us toidentify three cis-acting elements between the nt 3225 39 splicesite and the nt 3605 39 splice site. We demonstrated previouslythat two ESEs, SE1 and SE2, are essential for preferentialutilization of the nt 3225 39 splice site in vitro and in vivo andthat this function is mediated through interaction with cellularSR protein splicing factors (50, 51). Another cis-acting elementwas designated an ESS because it suppresses use of the nt 322539 splice site in vitro. The BPV-1 ESS is immediately down-stream of SE1 and 125 nt upstream of SE2 (see Fig. 1A). TheESS is 48 nt long and can be divided into three regions basedon sequence composition. Analysis of the proteins binding tothe ESS has shown that the U-rich 59 region of the ESS bindsU2AF65 and PTB, the C-rich central part binds 35- and 55-kDaSR proteins, and the AG-rich 39 end binds ASF/SF2 (53). Theactivity of the ESS maps to the central C-rich core (GGCUCCCCC), which, along with additional nonspecific downstreamnucleotides, is sufficient for partial suppression of splicing ofBPV-1 late pre-mRNAs at or before assembly of spliceosomecomplex A. Moreover, the inhibition of in vitro splicing by theESS can be partially relieved by excess purified HeLa SRproteins (53). We have also demonstrated using homologousand heterologous pre-mRNAs that the function of the ESS invitro requires a suboptimal upstream 39 splice site, but not anupstream ESE (52). Both the nt 3225 39 splice site and nt 360539 splice site have been shown to be functionally suboptimaland are characterized by nonconsensus BPSs and PPTs inter-spersed with purines. The weak nature of these sites allowsthem to be subject to regulation by ESEs and ESSs (50; Z.-M.Zheng et al., unpublished data).

Our previous studies demonstrated that the BPV-1 ESSfunctions in vitro in several different pre-mRNAs, all of whichcontain only one 39 splice site. These studies did not addressthe role of the ESS in the regulation of alternative splicing orits function in vivo, however. In this study, we have analyzedthe function of the ESS in vivo in a BPV-1 late pre-mRNA withtwo alternative splice sites. We have found that the splicingsuppressor activity of the ESS can be demonstrated in vivousing the late pre-mRNA in which SE2 has been mutated. Thissuggests that one role of SE2 is to counterbalance the effect of

the ESS. Furthermore, optimization of the suboptimal nt 322539 splice site counteracts the function of the ESS in vivo and invitro and leads to selection of the nt 3225 39 splice site even inthe absence of SE2. Finally, we have demonstrated usingBPV-1 late pre-mRNAs containing an optimized nt 3225 39splice site that a purine-rich ESE (SE1) stimulates not only thefirst step but also the second step of splicing.

MATERIALS AND METHODS

Plasmids and plasmid construction. Construction of BPV-1 late minigeneplasmids p3030 (pZMZ19-1), p3231 (pCCB458-2), and p3033 (SE2 mutationonly) has been described in our previous publications (50, 51). To replace theBPV-1 BPS at the nt 3225 39 splice site with the human b-globin BPS as well asto substitute pyrimidines for purines in the PPT, an overlap extension PCR (10,16) was performed on plasmid p3030. Basically, two pairs of primers were usedfor PCR. Primer Pr3169 (oZMZ162; 59-GTAAGAGATCAGGACAGAGTGTATGCTGGTCACTGACCCTCCTCTTCTTTTTTCAGAGATCGCCCAGACGGAGTCTGG-39) was complementary to Pr3246 (oZMZ191; 59-CCAGACTCCGTCTGGGCGATCTCTGAAAAAAGAAGAGGAGGGTCAGTGACCAGCATACACTCTGTCCTGATCTCTTAC-39). Pr3169 and Pr3246 contain theBPS and PPT mutations mentioned above in the middle of the oligonucleotidesequence and were combined, respectively, with primers Pr3715 (oZMZ161;59-TTTCAGCACCGTTGTCAGCAACTGTG-39) and Pr7276 (oFD127, a chi-meric T7/BPV-1 primer; 59-TAATACGACTCACTATAGGGA/GCGCCTGGCACCGAATCC-39) for separate PCRs (see diagram in Fig. 2A). The two PCRproducts were then gel purified and annealed together through their overlappingsequences. Finally, the annealed PCR mix was reamplified by PCR by using theprimer Pr7276 in combination with the primer Pr3715. The reamplified PCRproducts were digested with XhoI and Asp718. The XhoI and Asp718 fragmentwith a size of about 380 bp was cloned into the XhoI and Asp718 site of plasmidp3033 (51). The new plasmid, p3082 (pZMZ62-7), containing both human b-glo-bin BPS and mutant PPT, was verified by sequencing and used for in vitro andin vivo splicing analysis.

To create the ESS mutations or deletion in plasmid p3231, in vitro site-directed mutagenesis was carried out using a transformer site-directed mutagen-esis kit (Clontech Laboratories, Inc., Palo Alto, Calif.). Two mutagenic primersand one selection primer were used for the mutagenesis. The mutagenic primerESSm (59-GCCCAGCCTGTCTCTTCTTTGCTCATTGTGCAGTGCTTGCGCGTGCAAGAGAGCAGGCCTCGG-39) had extensive mutations in the cen-tral C-rich functional core sequence (53). The mutagenic primer ESSd (59-CCCAGCCTGTCTCTTCTTTGCCCTCGGTTGGGTACGGG-39) had an entiredeletion of both the central C-rich region and the AG-rich 39 end of the ESS.The selection primer (59-GGGACGCCAATTGGTAATCATGG-39) spansthe EcoRI site at the junction between BPV-1 and pUC18 and destroys thisrestriction endonuclease cleavage site. All mutations were verified by DNAsequencing. The new plasmid with the ESS mutations was named p3035 (pZMZTM5-3 or pESSm), and the new plasmid with the ESS deletion was named p3036(pZMZ TM6-1 or pESSd).

Four of our previously published plasmids, p3031, p3032, p3033, and p3034(51), were used to create a double mutation of both SE1 and SE2 in a BPV-1 latetranscript expression vector. Briefly, the XhoI and Asp718 fragments containingSE1 mutations in plasmid p3031 or an SE1 deletion in plasmid p3032 wereswapped, respectively, into the corresponding sites of plasmid p3033 (pSE2m)and p3034 (pSE2d). This strategy created a plasmid, p3078 (pZMZ58), with bothSE1 and SE2 mutations and a plasmid, p3079 (pZMZ59), with both SE1 and SE2deletions.

To create a BPV-1 late pre-mRNA expression vector containing a doublemutation of both the ESS and SE2, the XhoI and Asp718 fragments containingESS mutations in plasmid p3035 or an ESS deletion in plasmid p3036 wereindividually swapped into the corresponding sites of plasmid p3033 (pSE2m).The new plasmids created are p3080 (pZMZ60 or pESSm1SE2m) and p3081(pZMZ61 or pESSd1SE2m).

DNA template preparation, in vitro splicing, and detection of splicing prod-ucts. Plasmid p3030, which transcribes a BPV-1 late pre-mRNA with a wild-type(wt) 39 splice site at nt 3225, and plasmid p3082, which transcribes a BPV-1 latepre-mRNA with a human b-globin BPS and optimal PPT at the nt 3225 39 splicesite, were used for preparation of DNA templates by PCR using a common 59sense primer, Pr7276 (oFD127, a chimeric T7/BPV-1 primer), in combinationwith one of the following 39 antisense primers: oZMZ127 (pre-mRNA 1 in Fig.3A), oZMZ84 (pre-mRNA 2 in Fig. 3A), or oZMZ102 (pre-mRNA 3 in Fig. 3A).The sequences of the three primers oZMZ127, oZMZ84, and oZMZ102 havebeen reported in our previous publication (52).

Pre-mRNAs were prepared using Promega’s riboprobe system (Promega,Madison, Wis.). In vitro runoff transcription was carried out with T7 RNApolymerase in the presence of the cap analog (m7GpppG) and [a-32P]rGTP asdescribed by the company protocol. HeLaSplice nuclear extracts used for in vitrosplicing assays were commercially obtained from Promega. The in vitro splicingassay and detection of splicing products have been described in our previousreport (49).

VOL. 74, 2000 EXONIC SPLICING SUPPRESSOR OF BPV-1 5903


http://jvi.asm.org/

Dow

nloaded from

http://jvi.asm.org/

Transfection of 293 cells and reverse transcription-PCR (RT-PCR) andNorthern blot analysis of spliced BPV-1 late mRNAs. Transfectam reagent(Promega) was used to transfect 2 mg of wt or mutant (mt) expression vectorDNA into human 293 cells plated in 60-mm-diameter dishes. Total cellular RNAwas prepared after 48 h using TRIzol (GIBCO BRL, Rockville, Md.) followingthe manufacturer’s instructions. Following DNase I treatment, 500 ng of totalcellular RNA was reverse transcribed at 42°C using random hexamers as primersand then amplified for 35 cycles using different pairs of primers as described inthe figure legends. An L1 cDNA and a mixture of L2-L and L2-S cDNAsgenerated from L2 mRNAs spliced using the nt 3225 39 splice site and nt 3605 39splice site, respectively, were used as PCR controls.

For Northern blot analysis, total cell RNA (40 mg) was poly(A) selected usingthe Promega PolyATtract mRNA Isolation System III. Equal amounts of thepoly(A)-selected RNAs were then denatured at 55°C for 15 min in formalde-hyde-formamide sample buffer (60% formamide, 2.2 M formaldehyde, 25 mg ofethidium bromide per ml, 13 MOPS [morpholinepropanesulfonic acid] buffer)and run on a 1% agarose-formaldehyde gel. After electrophoresis was com-pleted, the gel was washed for 5 min with water, partially hydrolyzed for 5 minin 50 mM NaOH–10 mM NaCl, neutralized for 5 min in 0.1 M Tris-HCl (pH 7.4)buffer, and transferred in 203 SSC (13 SSC is 0.15 M NaCl plus 0.015 M sodiumcitrate) for 1.5 h to a GeneScreen membrane using a Pharmacia vacuum blottingdevice. After UV cross-linking in a Stratalinker (Stratagene, La Jolla, Calif.) andphotography, the membrane was hybridized with 32P-labeled BPV-1 early probebased on Church’s protocol (8). The probe was prepared from a PCR-generatedBPV-1 DNA template (nt 3211 to nt 4230) and was labeled using a Prime-It IIrandom primer labeling kit (Stratagene) to a specific activity of ;4 3 109 cpm/mgand used at ;3 3 106 cpm/ml in the hybridization reaction. The 32P-labeledBPV-1 early probe should hybridize to all species of the alternatively splicedmRNAs.

RESULTS

The BPV-1 ESS inhibits selection of an upstream 3* splicesite in vivo. Our previous studies demonstrated that BPV-1late pre-mRNAs have three cis elements that reside betweentwo alternative 39 splice sites, the nt 3225 39 splice site and thent 3605 39 splice site (Fig. 1A). We have utilized a series ofBPV-1 late minigene expression vectors to analyze the functionof these cis elements in vivo. Since the BPV-1 late promoter isinactive in cell culture, the BPV-1 late transcription unit wascloned downstream of the cytomegalovirus IE1 promoter.Care was taken to ensure that the 59 ends of the BPV-1 latemRNAs expressed in transfection assays were similar to thoseexpressed from the late promoter in BPV-1-infected warts. Intransfection assays, the nt 3225 39 splice site is used predomi-nantly for splicing of the BPV-1 late pre-mRNA (51), (Fig. 1C,lane 6). Selection of this 39 splice site requires both SE1 andSE2 (51). Mutation of either one of these enhancers results insplicing of the pre-mRNA using the nt 3605 39 splice site (lanes2 to 5 in Fig. 1C). However, mutation or deletion of only theESS does not significantly affect BPV-1 late pre-mRNA splic-ing in 293 cells (Fig. 1C, lanes 10 and 11). These observationssuggest that two strong ESEs (SE1 and SE2) are required to

overcome suppression of the nt 3225 39 splice site by the ESS.If this assumption is correct, SE1 should be sufficient for se-lection of the nt 3225 39 splice site in the absence of the ESS.To address this fundamental question, we replaced or deletedthe most important functional region of the ESS (53) in thecontext of a BPV-1 late gene expression vector which alreadyhad SE2 mutations or deletions (51). These plasmids wereused to transfect 293 cells, and splicing of the BPV-1 late pre-mRNA was analyzed by RT-PCR and Northern blotting. Asexpected, wt BPV-1 late pre-mRNA is predominately splicedusing the nt 3225 39 splice site, and mutation or deletion ofboth SE1 and SE2 resulted in a switch to predominant utiliza-tion of the nt 3605 39 splice site (Fig. 1D and E). The effects ofdouble ESE mutations are similar to that seen for SE1 or SE2mutation alone (lanes 2 to 5 in Fig. 1C). However, point mu-tations in the ESS in the context of a pre-mRNA containingpoint mutations in SE2 (ESSm 1 SE2m) partially restoredutilization of the nt 3225 39 splice site (lane 8 in Fig. 1D; alsoFig. 1E). Even more dramatic was the effect of deletion of theESS (ESSd 1 SE2m) (lane 9 in Fig. 1D; also Fig. 1E). Two-thirds of this pre-mRNA was spliced at the nt 3225 39 splice site(Fig. 1E). This experiment suggests that SE1 enhances selec-tion of the nt 3225 39 splice site in the absence of the ESS. Weconclude from these in vivo experiments that the ESS sup-presses SE1-stimulated pre-mRNA splicing. In addition, SE2 isrequired to compensate for this suppression, even though SE2is positioned 125 nt downstream of the ESS and 252 nt down-stream of the nt 3225 39 splice site.

Strengthening the weak nt 3225 3* splice site counteracts thefunction of the ESS in vivo. Previously, we have shown that theBPV-1 ESS functions only in a suboptimal 39 splice site-con-taining pre-mRNA and that its negative effect on splicing isindependent of ESEs (52). We also demonstrated that theBPV-1 nt 3225 39 splice site contains a suboptimal BPS andPPT (50). This suggests that we may be able to neutralize thefunction of the ESS in a BPV-1 late pre-mRNA by strength-ening the weak nt 3225 39 splice site. To test this hypothesis, astrong BPS from the human b-globin pre-mRNA intron 1 wasintroduced to replace the suboptimal BPS at the nt 3225 39splice site of a BPV-1 late pre-mRNA expression vector con-taining SE2 point mutations (Fig. 2A). Human b-globin pre-mRNA intron 1 has a consensus BPS (CACUGAC) similar tothe mammalian consensus BPS (YNYURAC) (30). In addi-tion, three point mutations (R to U) in the PPT and a UAG-to-CAG mutation in the 39 splice junction (Fig. 2A) were alsomade in the nt 3225 39 splice site of the same plasmid (seeMaterials and Methods). These mutations in the nt 3225 39

FIG. 1. Double mutational analysis of both BPV-1 ESS and SE2 in vivo. (A) Diagrams of a BPV-1 late minigene expression vector (top) and the splicing pattern(bottom) of pre-mRNAs expressed from the vector. At the ends of the late minigene (open box) are the cytomegalovirus (CMV) IE1 promoter (solid box) and pUC18(grey box). Numbers below the lines are the nucleotide positions in the BPV-1 genome. Early and late poly(A) sites are indicated as AE and AL, respectively, abovethe lines. A large deletion in the intron region of the genome is shown by a heavy vertical line. Below the minigene diagram is the structure of the minigene transcript.The relative positions of SE1, ESS, and SE2 are shown as well as the nucleotide positions of the 59 splice site (59 ss) and 39 splice site (39 ss). Two pairs of primers usedfor RT-PCR in panels C and D are indicated by arrows under the transcript and named by the location of their 59 ends. (B) Nucleotide sequences of wt and mt SE1,ESS, and SE2 in the pre-mRNAs expressed from the following plasmids used for transfection of 293 cells: plasmids p3231 (wt), p3031 (SE1m), p3032 (SE1d), p3033(SE2m), p3034 (SE2d), p3035 (ESSm), p3036 (ESSd), p3078 (SE1m 1 SE2m), p3079 (SE1d 1 SE2d), p3080 (ESSm 1 SE2m), and p3081 (ESSd 1 SE2m) (seeMaterials and Methods for construction of the plasmids with single or double mutations). Unchanged nucleotides (dots) and deletions (horizontal lines) are indicated.Deletion endpoints are labeled above each sequence and correspond to positions in the BPV-1 genome. (C and D) RT-PCR analysis of BPV-1 late mRNAs transcribedand spliced in 293 cells from the expression vectors with single (C) or double (D) mutations as labeled on the top of each gel. Total cell RNA was extracted and digestedwith RNase-free DNase I before RT-PCR analysis. Several controls were included for the assays including p3231 (wt)-transfected 293 cell RNAs, untransfected 293cell RNAs, and water controls. Total cell RNAs extracted from p3231 (wt)-transfected 293 cells but not treated with RNase-free DNase I [WT (2RT)] and BPV-1 L1cDNA and a 10:1 mixture of L2-L and L2-S cDNAs were also used as templates for PCR amplification in panel D. Predicted sizes of the RT-PCR products differbetween panels due to the use of different sets of primers and are 658 bp (C) and 530 bp (D) for nt 3225 39 splice site usage and 278 bp (C) and 150 bp (D) for nt3605 39 splice site usage. Numbers at left of panel C and right of panel D are molecular sizes in base pairs. (E) Northern blot analysis of the BPV-1 late mRNAstranscribed and spliced in 293 cells from the expression vectors with double mutations. Poly(A)-selected total cell RNA extracted from 293 cells transfected with plasmidp3078 (SE1m 1 SE2m), p3080 (ESSm 1 SE2m), and p3081 (ESSd 1 SE2m) was run on a 1% agarose–formaldehyde gel, blotted onto a nitrocellulose membrane, andhybridized with 32P-labeled BPV-1 early probe. The ratio (percentage) of the nt 3225 39 splice site usage was calculated based on the formula: % 5 nt 3225 39 splicesite/(nt 3225 1 nt 3605 39 splice site). One representative experiment of three for each panel, C, D, and E, is shown.

5904 ZHENG ET AL. J. VIROL.


http://jvi.asm.org/

Dow

nloaded from

http://jvi.asm.org/



http://jvi.asm.org/

Dow

nloaded from

http://jvi.asm.org/

FIG. 2. Strengthening the nt 3225 39 splice site counteracts the function ofthe BPV-1 ESS in vivo. (A) Diagram of an overlap extension PCR strategy usedto introduce human b-globin BPS and PPT mutations at the nt 3225 39 splice sitein a BPV-1 late minigene. The schematic diagram, as in Fig. 1A, is of a BPV-1 lateminigene plasmid, p3030 (50). Promoters (black boxes) are T7 on the 59 end and SP6on the 39 end of the vector. Grey boxes at the 39 end are pUC18. Restriction sitesXhoI (X), Asp718 (A), and BamHI (B); 39 or 59 splice sites; and the early poly(A) site(AE) are indicated on the top of the vector. Two pairs of primers (Pr7276 and Pr3246as well as Pr3169 and Pr3715) were used for the overlap extension PCR (seeMaterials and Methods). The constructed plasmid p3082 has a pre-mRNA transcriptcontaining both human b-globin BPS and mt PPT at the nt 3225 39 splice site, whichare shown on the top of the diagram with branch points underlined and substitutionslowercased (mt). Mammalian BPS consensus YNYURAC and wt BPV-1 BPS andPT are shown for comparison. (B) A BPV-1 late minigene expression vector (top)and the splicing patterns of pre-mRNAs expressed from the vector (below) areschematically diagrammed as in Fig. 1A. Structures of pre-mRNAs 1, 2, and 3 andthe nucleotide sequences of SE1, ESS, and SE2 or SE2m are shown as in Fig. 1A andB. The wt BPV-1 BPS is shown in an open circle, and the human b-globin BPS isshown in a grey circle. The PPT region between BPS and the nt 3225 39 splice siteis diagrammed as a black line (wt) or a grey line (mt form). The nucleotidecomposition of the mt BPS and PPT in the pre-mRNA 3 is shown in panel A. Thepre-mRNAs 1, 2, and 3 were transcribed, respectively, from plasmids p3231, p3033(51), and p3082. Their splicing patterns were analyzed by RT-PCR using a 59sense primer, Pr7345 (oCCB48; 59-CAATGGGACGCGTGCAAAGC-39), com-bined with a 39 antisense primer, Pr3746 (oCCB57; 59-AAGGTGATCAGTATTTGTGC-39), as shown below the pre-mRNA 3. (C) A representative agarosegel of the RT-PCRs from three separate transfections. Total cell RNA was ex-tracted from 293 cells transfected by individual plasmids and digested with RNase-free DNase I before RT-PCR analysis. The pre-mRNAs 1, 2, and 3 (lanes 5 to 7,respectively) labeled above the agarose gel correspond to the pre-mRNAs 1, 2,and 3 in panel B. An untransfected 293 cell RNA control (lane 8) was also in-cluded. Mixtures of L2-S and L2-L cDNAs (spliced at nt 3605 and 3225, respec-tively) were prepared at the specified ratios and were used as PCR templates forquantitation and size controls. Predicted sizes of the RT-PCR products are 561bp for nt 3225 39 splice site usage and 181 bp for nt 3605 39 splice site usage.



http://jvi.asm.org/

Dow

nloaded from

http://jvi.asm.org/

splice site should convert this suboptimal splice site to anoptimal splice site which would allow us to test our assumption.The new construct was then used to transfect 293 cells, andtotal cell RNA was analyzed by RT-PCR for alterations insplice site usage.

As we showed in Fig. 1C, mutation of SE2 in the context ofa pre-mRNA containing a wt nt 3225 39 splice site (pre-mRNA2) switched splice site selection from nt 3225 to nt 3605 (lane6 in Fig. 2C). In contrast, mutation of SE2 had no significanteffect on nt 3225 39 splice site utilization in a pre-mRNAcontaining an optimal nt 3225 39 splice site (pre-mRNA 3)(lane 7 in Fig. 2C). These data suggest that splicing in vivo atthe nt 3225 39 splice site no longer requires SE2 because theESS cannot suppress an optimal 39 splice site. We concludethat a suboptimal nt 3225 39 splice site is a prerequisite for thefunction of the BPV-1 ESS in vivo.

In vitro splicing of the BPV-1 pre-mRNAs with an optimizednt 3225 3* splice site. To further confirm the observation fromour in vivo study described above, we used in vitro splicingassays to test if an optimized 39 splice site in a pre-mRNAcould counteract suppression of pre-mRNA splicing by theESS. These studies look at the absolute effect of the ESS on asingle 39 splice site rather than on alternative splicing. Two setsof pre-mRNAs with either a suboptimal (wt) or an optimized(mt) 39 splice site were prepared by in vitro transcription (see

Materials and Methods). Each set of pre-mRNAs consisted ofthree different transcripts in which exon 2 contained either amutant ESE, (Py3)2 (42, 50, 51); SE1; or SE1 plus the ESS(Fig. 3A, pre-mRNAs 1, 2, and 3, respectively). These pre-mRNAs were then spliced in vitro using HeLa nuclear extracts.The results are shown in Fig. 3B and summarized in Fig. 3A assplicing efficiency (percent spliced) for individual pre-mRNAspecies. As expected, the ESS suppressed splicing of a wt pre-mRNA with a threefold reduction of splicing efficiency (com-pare wt pre-mRNA 3 to wt pre-mRNA 2 in both Fig. 3A andB). However, when the pre-mRNAs had an optimized (mt) nt3225 39 splice site, both mt pre-mRNAs were spliced withsimilar efficiencies (compare mt pre-mRNA 3 to mt pre-mRNA 2 in both Fig. 3A and B). The data from these in vitroexperiments confirm that the ESS is not able to suppress anoptimal nt 3225 39 splice site.

The BPV-1 ESE SE1 stimulates the second step of splicingof a BPV-1 late pre-mRNA. We have reported previously thatwt BPV-1 late pre-mRNAs without SE1 are very inefficientlyspliced in vitro in comparison to wt pre-mRNAs containingSE1 (50, 52). In the experiment shown in lanes 4 and 5 in Fig.3B, pre-mRNA 1 containing a wt (suboptimal) nt 3225 39 splicesite and mutant ESE (Py3)2 (wt pre-mRNA 1 in Fig. 3A) wasspliced nine times less efficiently than the wt pre-mRNA con-taining SE1 (wt pre-mRNA 2 in Fig. 3A). In contrast, both pre-

FIG. 3. Conversion of the suboptimal nt 3225 39 splice site to an optimal 39splice site in BPV-1 late pre-mRNAs overcomes suppression of in vitro splicingby the ESS. (A) Structures of the pre-mRNAs used for in vitro splicing assay.Numbers above pre-mRNA exons (large boxes) and introns (lines) are thenucleotide positions in the BPV-1 genome. mt ESE (Py3)2 and BPV-1 SE1 andESS are shown as labeled boxes. Both wt or mt BPS and PPT at the 39 end ofeach intron are shown as a thick grey line, and their sequence compositions aredisplayed in Fig. 2A. Plasmid p3030, which transcribes a pre-mRNA with a wt 39splice site at nt 3225, and plasmid p3082, which transcribes a pre-mRNA with ahuman b-globin BPS and strong PPT at the nt 3225 39 splice site, were used forpreparation of DNA templates by PCR. Thus, the pre-mRNAs with or withoutSE1 and/or an ESS element transcribed from p3030 DNA templates have a wt nt3225 39 splice site, whereas the pre-mRNAs with the same structures transcribedfrom p3082 DNA templates have an mt nt 3225 39 splice site. Splicing efficiencyfor each pre-mRNA (% 5 spliced products/spliced products 1 unspliced RNA)was calculated from the splicing gel in panel B. (B) Electrophoresis of splicingproducts after in vitro splicing reactions carried out at 30°C for 2 h in thepresence of 40% HeLa nuclear extracts and 0.5 mM MgCl2. Electrophoresis wasperformed on an 8% polyacrylamide gel containing 8 M urea. The identities ofthe spliced products and splicing intermediates are shown on the right of the gel.One representative splicing gel of three is shown.



http://jvi.asm.org/

Dow

nloaded from

http://jvi.asm.org/

mRNAs, 1 and 2, were spliced efficiently when they containedan mt (optimal) nt 3225 39 splice site (compare mt pre-mRNAs1 and 2 in Fig. 3A). Surprisingly, a careful examination of thespliced products revealed that splicing of the mt pre-mRNA 1with (Py3)2 produced mostly splicing intermediates generatedfrom the first step of the splicing reaction (lane 1 in Fig. 3B).No fully spliced products resulting from the ligation of exon 1to exon 2 were detected. In contrast, mt pre-mRNA 2 with SE1produced not only splicing intermediates but also fully splicedproducts (lane 2 in Fig. 3B). The percentage of the startingpre-mRNA found in splicing intermediates was 56% for the mtpre-mRNA 1 with (Py3)2 and only 36% for the mt pre-mRNA2 with SE1. The smaller amount of splicing intermediates frommt pre-mRNA 2 with SE1 is most likely due to increasedefficiency of the second step of splicing for this pre-mRNA.These data suggest that BPV-1 SE1 can function not only bystimulating early steps in spliceosomal assembly but also byenhancing the second step of splicing. Thus, although a strong39 splice site is sufficient for the first step of splicing, in somecases a strong ESE may be required for the second step.

DISCUSSION

The BPV-1 ESS was originally identified in an in vitro splic-ing assay using pre-mRNAs containing only the nt 3225 39splice site (50). In this context, the ESS blocks splicing en-hanced by the enhancer SE1. Here we have used these con-structs in transfection assays and demonstrated that the ESSalso acts as a splicing suppressor in vivo and plays an importantrole in the regulation of alternative splicing in a BPV-1 latepre-mRNA that contains both the nt 3225 39 splice site and thent 3605 39 splice site. Our data are consistent with the modelpresented in Fig. 4 where two strong ESEs (SE1 and SE2) arerequired to overcome the effect of the ESS and select the nt3225 39 splice site as the dominant splice site at early stages ofthe viral life cycle. Thus, when both SE1 and SE2 are present,mutation of the ESS has little effect on alternative splicing.Deletion of the ESS may give a slight increase in usage of thent 3225 39 splice site with a reciprocal decrease in utilization ofthe nt 3605 39 splice site, but this is difficult to accurately assessusing a competitive PCR assay (Fig. 1C, lanes 10 and 11). Incontrast, the splicing suppressor activity of the ESS can easilybe demonstrated when the balance between nt 3225 and nt

3605 39 splice site selection is altered by mutations in one ofthe splicing enhancers (Fig. 1D and E).

The data presented here also provide further evidence insupport of the notion that the function of the BPV-1 ESSrequires a suboptimal 39 splice site (52). Whereas two enhanc-ers are needed in vivo to overcome splicing suppression by theESS in the context of the wt suboptimal nt 3225 39 splice site,only one enhancer is required when the nt 3225 39 splice sitehas been optimized (Fig. 2). Optimization of the nt 3225 39splice site leads to a reciprocal increase in usage of the nt 322539 splice site and a decrease in usage of the nt 3605 39 splice sitedue to competition between the two sites. Likewise, in in vitroexperiments the ESS gave only a 20% reduction of splicing ofa pre-mRNA containing the optimized nt 3225 39 splice site(Fig. 3). Our strategy to optimize the nt 3225 39 splice site wasto strengthen both the BPS and the PPT. This was achieved byreplacing the suboptimal 39 splice site with a combination of ahuman b-globin BPS (CACUGAC) and a 16-nt pyrimidinestretch followed by a CAG trinucleotide (30). This 39 splice siteshould strongly bind both the U2 snRNP and U2AF65, even inthe absence of a splicing enhancer. These data, taken togetherwith our previous data showing that the function of the ESSdoes not require a splicing enhancer, suggest that the ESSblocks the binding of splicing factors to the suboptimal 39 splicesite. Several lines of evidence suggest that other ESS alsorequire a suboptimal 39 splice site for their function. In HIV-1tat ESS studies, mutation of the interspersed purines to pyri-midines in the PPT of the HIV-1 tat exon 2 39 splice site resultsin a significant increase in splicing efficiency of an ESS-con-taining pre-mRNA in vitro (34). Likewise, optimization of ei-ther the BPS or the PPT in the HIV-1 tat exon 3 39 splice sitealso eliminates most of the splicing suppression by the ESS3 invivo (38).

A surprising finding of our study is the observation that theBPV-1 ESE SE1 is capable of stimulating not only the first stepof splicing but also the second step of the two-step transesteri-fication reaction. While we were preparing the manuscript,Chew and coworkers (7) also demonstrated that an ASF/SF2-specific ESE promotes the second step in splicing. The firstcatalytic step involves cleavage at the 59 splice site, whichgenerates a free 59 exon and a lariat (intron)-containing 39exon. The second step is the cleavage at the 39 splice site with

FIG. 4. A proposed model for the function of the BPV-1 ESS in regulation of BPV-1 alternative splicing. Selection of the suboptimal nt 3225 39 splice site isregulated through three viral cis-acting elements, two ESEs (SE1 and SE2) and one ESS. The ESEs stimulate splicing at the nt 3225 39 splice site through recruitmentof splicing factors at early stages of spliceosome assembly (55). The ESS plays a negative role in the selection of the nt 3225 39 splice site through interaction withdifferent cellular splicing factors (53). In the normal context, the combination of two ESEs overcomes the negative effect of the ESS on the nt 3225 39 splice site anddrives splicing to use predominately the nt 3225 39 splice site. Cellular splicing factors such as SR proteins appear to play a key role in this regulation, and theirphysiological status might affect each cis element’s function and consequently alternative splicing of the BPV-1 late pre-mRNA. 39 ss, 39 splice site; 59 ss, 59 splice site;?, hypothetical factor(s). Circles, BPS.



http://jvi.asm.org/

Dow

nloaded from

http://jvi.asm.org/

simultaneous joining of the two exons and release of the intronas a lariat form (21, 25, 43). Splicing of mammalian pre-mRNAs, which frequently have suboptimal 39 splice sites, inmany cases depends upon ESEs in the 39 exon which bind SRproteins (29, 51). In currently accepted models for the functionof an ESE, the RS domain of ESE-bound SR proteins interactswith the RS domain of U2AF35, which then recruits U2AF65 tothe suboptimal PPT, promoting the binding of the U2 snRNPto the suboptimal BPS (3, 14, 55). However, it is unclear whatrole the ESE and SR proteins play during the second step ofsplicing. Our data suggest that a consensus BPS or optimalPPT or both promotes only the first step of splicing. Thisobservation indicates that the presence of an ESE in 39 exonsis essential for a complete splicing reaction and therefore sug-gests that ESEs should be universally present no matterwhether the 39 splice site is suboptimal or whether it is optimal.This ESE may not be limited to a purine-rich enhancer. Con-sistent with this hypothesis, a recent report indicates that splic-ing of the human b-globin intron 1, which contains a consensusBPS, also requires an ESE in the 39 exon (33). One possiblerole for SR proteins and ESEs in the second step of splicing isto directly promote bridging across the intron (40).

One key issue that remains is what splicing factors regulateBPV-1 alternative splicing during the viral life cycle. Theswitch from utilization of the nt 3225 39 splice site at earlytimes of the viral life cycle to use of the nt 3605 39 splice site atlate stages is an integral component of the early to late switchin viral gene expression and is intimately linked to keratinocytedifferentiation (2). Identification of the viral cis-acting splicingelements between the two alternative 39 splice sites and eluci-dation of their functions in vitro and in vivo have provided abasis for understanding the mechanism of splice site selection(Fig. 4). Our in vivo studies suggest that the early to late switchin splicing could result from either a decrease in enhanceractivity or an increase in activity of the ESS. As mentionedabove, the functions of SE1 and SE2 are mediated at least inpart by SR proteins (51). It is not yet known if the activity ofone or more SR proteins changes during keratinocyte differ-entiation. However, it has been shown that SR protein levelsand their differential expression as well as physiological statusare all important for regulation of alternative splicing (1, 4, 12,19, 28, 44). A recent study indicates that developmental regu-lation of SR protein activity in the nematode Ascaris lumbri-coides plays a role in activation of pre-mRNA splicing duringearly development (32). Other reports also suggest that alter-ations in the expression of a subset of SR proteins and alter-native splicing of a CD44 pre-mRNA accompany the transitionfrom normal cells to mouse mammary adenocarcinoma (41)and human colon adenocarcinomas (13). Viral proteins canalso regulate SR protein function. The adenovirus E4-ORF4protein is an early viral protein that activates protein phospha-tase 2A, leading to the dephosphorylation and inactivation ofSR proteins (18). An early to late shift in 39 splice site utiliza-tion results from decreased SR protein binding to an intronicsplicing suppressor. The function of the BPV-1 ESS is alsothought to be mediated through the binding of cellular splicingfactors, including SR proteins and perhaps other as-yet-uni-dentified proteins (53). Therefore, we predict that the differ-ential expression and/or modification of cellular splicing fac-tors during keratinocyte differentiation and virus infection willbe a key component of the regulation of BPV-1 gene expres-sion at the posttranscriptional level.

REFERENCES1. Bai, Y. D., D. Lee, T. D. Yu, and L. A. Chasin. 1999. Control of 39 splice site

choice in vivo by ASF/SF2 and hnRNP A1. Nucleic Acids Res. 27:1126–1134.2. Barksdale, S. K., and C. C. Baker. 1995. Differentiation-specific alternative

splicing of bovine papillomavirus late mRNAs. J. Virol. 69:6553–6556.3. Buvoli, M., S. A. Mayer, and J. G. Patton. 1997. Functional crosstalk between

exon enhancers, polypyrimidine tracts and branchpoint sequences. EMBO J.16:7174–7183.

4. Caceres, J. F., S. Stamm, D. M. Helfman, and A. R. Krainer. 1994. Regu-lation of alternative splicing in vivo by overexpression of antagonistic splicingfactors. Science 265:1706–1709.

5. Caputi, M., A. Mayeda, A. R. Krainer, and A. M. Zahler. 1999. hnRNP A/Bproteins are required for inhibition of HIV-1 pre-mRNA splicing. EMBO J.18:4060–4067.

6. Chen, C. D., R. Kobayashi, and D. M. Helfman. 1999. Binding of hnRNP Hto an exonic splicing silencer is involved in the regulation of alternativesplicing of the rat b-tropomyosin gene. Genes Dev. 13:593–606.

7. Chew, S. L., H. X. Liu, A. Mayeda, and A. R. Krainer. 1999. Evidence for thefunction of an exonic splicing enhancer after the first catalytic step of pre-mRNA splicing. Proc. Natl. Acad. Sci. USA 96:10655–10660.

8. Church, G. M., and W. Gilbert. 1984. Genomic sequencing. Proc. Natl. Acad.Sci. USA 81:1991–1995.

9. Del Gatto-Konczak, F., M. Olive, M. C. Gesnel, and R. Breathnach. 1999.hnRNP A1 recruited to an exon in vivo can function as an exon splicingsilencer. Mol. Cell. Biol. 19:251–260.

10. Fang, G., B. Weiser, A. Visosky, T. Moran, and H. Burger. 1999. PCR-mediated recombination: a general method applied to construct chimericinfectious molecular clones of plasma-derived HIV-1 RNA. Nat. Med.5:239–242.

11. Fu, X. Y., H. Ge, and J. L. Manley. 1988. The role of the polypyrimidinestretch at the SV40 early pre-mRNA 39 splice site in alternative splicing.EMBO J. 7:809–817.

12. Ge, H., and J. L. Manley. 1990. A protein factor, ASF, controls cell-specificalternative splicing of SV40 early pre-mRNA in vitro. Cell 62:25–34.

13. Ghigna, C., M. Moroni, C. Porta, S. Riva, and G. Biamonti. 1998. Alteredexpression of heterogenous nuclear ribonucleoproteins and SR factors inhuman colon adenocarcinomas. Cancer Res. 58:5818–5824.

14. Graveley, B. R., K. J. Hertel, and T. Maniatis. 1998. A systematic analysis ofthe factors that determine the strength of pre-mRNA splicing enhancers.EMBO J. 17:6747–6756.

15. Green, M. R. 1986. Pre-mRNA splicing. Annu. Rev. Genet. 20:671–708.16. Ho, S. N., H. D. Hunt, R. M. Horton, J. K. Pullen, and L. R. Pease. 1989.

Site-directed mutagenesis by overlap extension using the polymerase chainreaction. Gene 77:51–59.

17. Kanaar, R., S. E. Roche, E. L. Beall, M. R. Green, and D. C. Rio. 1993. Theconserved pre-mRNA splicing factor U2AF from Drosophila: requirementfor viability. Science 262:569–573.

18. Kanopka, A., O. Muhlemann, S. Petersen-Mahrt, C. Estmer, C. Ohrmalm,and G. Akusjarvi. 1998. Regulation of adenovirus alternative RNA splicingby dephosphorylation of SR proteins. Nature 393:185–187.

19. Krainer, A. R., G. C. Conway, and D. Kozak. 1990. The essential pre-mRNAsplicing factor SF2 influences 59 splice site selection by activating proximalsites. Cell 62:35–42.

20. Mayeda, A., G. R. Screaton, S. D. Chandler, X. D. Fu, and A. R. Krainer.1999. Substrate specificities of SR proteins in constitutive splicing are deter-mined by their RNA recognition motifs and composite pre-mRNA exonicelements. Mol. Cell. Biol. 19:1853–1863.

21. Moore, M. J., C. C. Query, and P. A. Sharp. 1993. Splicing of precursors tomessenger RNAs by the spliceosome, p. 303–357. In R. F. Gestland and J. F.Atkins (ed.), The RNA world. Cold Spring Harbor Laboratory Press, ColdSpring Harbor, N.Y.

22. Mullen, M. P., C. W. Smith, J. G. Patton, and B. Nadal-Ginard. 1991.Alpha-tropomyosin mutually exclusive exon selection: competition betweenbranchpoint/polypyrimidine tracts determines default exon choice. GenesDev. 5:642–655.

23. Muro, A. F., M. Caputi, R. Pariyarath, F. Pagani, E. Buratti, and F. E.Baralle. 1999. Regulation of fibronectin EDA exon alternative splicing: pos-sible role of RNA secondary structure for enhancer display. Mol. Cell. Biol.19:2657–2671.

24. Norton, P. A. 1994. Polypyrimidine tract sequences direct selection of alter-native branch sites and influence protein binding. Nucleic Acids Res. 22:3854–3860.

25. Padgett, R. A., P. J. Grabowski, M. M. Konarska, S. Seiler, and P. A. Sharp.1986. Splicing of messenger RNA precursors. Annu. Rev. Biochem. 55:1119–1150.

26. Parker, R., P. G. Siliciano, and C. Guthrie. 1987. Recognition of the TAC-TAAC box during mRNA splicing in yeast involves base pairing to theU2-like snRNA. Cell 49:229–239.

27. Potashkin, J., K. Naik, and K. Wentz-Hunter. 1993. U2AF homolog re-quired for splicing in vivo. Science 262:573–575.

28. Prasad, J., K. Colwill, T. Pawson, and J. L. Manley. 1999. The protein kinaseClk/Sty directly modulates SR protein activity: both hyper- and hypophos-phorylation inhibit splicing. Mol. Cell. Biol. 19:6991–7000.

29. Ramchatesingh, J., A. M. Zahler, K. M. Neugebauer, M. B. Roth, and T. A.Cooper. 1995. A subset of SR proteins activates splicing of the cardiac



http://jvi.asm.org/

Dow

nloaded from

http://jvi.asm.org/

troponin T alternative exon by direct interactions with an exonic enhancer.Mol. Cell. Biol. 15:4898–4907.

30. Reed, R. 1989. The organization of 39 splice-site sequences in mammalianintrons. Genes Dev. 3:2113–2123.

31. Roscigno, R. F., M. Weiner, and M. A. Garcia-Blanco. 1993. A mutationalanalysis of the polypyrimidine tract of introns. Effects of sequence differ-ences in pyrimidine tracts on splicing. J. Biol. Chem. 268:11222–11229.

32. Sanford, J. R., and J. P. Bruzik. 1999. Developmental regulation of SRprotein phosphorylation and activity. Genes Dev. 13:1513–1518.

33. Schaal, T. D., and T. Maniatis. 1999. Multiple distinct splicing enhancers inthe protein-coding sequences of a constitutively spliced pre-mRNA. Mol.Cell. Biol. 19:261–273.

34. Si, Z. H., B. A. Amendt, and C. M. Stoltzfus. 1997. Splicing efficiency ofhuman immunodeficiency virus type 1 tat RNA is determined by both asuboptimal 39 splice site and a 10 nucleotide exon splicing silencer elementlocated within tat exon 2. Nucleic Acids Res. 25:861–867.

35. Singh, R., J. Valcarcel, and M. R. Green. 1995. Distinct binding specificitiesand functions of higher eukaryotic polypyrimidine tract-binding proteins.Science 268:1173–1176.

36. Smith, C. W., E. B. Porro, J. G. Patton, and B. Nadal-Ginard. 1989. Scanningfrom an independently specified branch point defines the 39 splice site ofmammalian introns. Nature 342:243–247.

37. Staffa, A., and A. Cochrane. 1994. The tat/rev intron of human immunode-ficiency virus type 1 is inefficiently spliced because of suboptimal signals inthe 39 splice site. J. Virol. 68:3071–3079.

38. Staffa, A., and A. Cochrane. 1995. Identification of positive and negativesplicing regulatory elements within the terminal tat-rev exon of human im-munodeficiency virus type 1. Mol. Cell. Biol. 15:4597–4605.

39. Staknis, D., and R. Reed. 1994. SR proteins promote the first specific rec-ognition of pre-mRNA and are present together with the U1 small nuclearribonucleoprotein particle in a general splicing enhancer complex. Mol. Cell.Biol. 14:7670–7682.

40. Stark, J. M., D. P. Bazett-Joness, M. Herfort, and M. B. Roth. 1998. SRproteins are sufficient for exon bridging across an intron. Proc. Natl. Acad.Sci. USA 95:2163–2168.

41. Stickeler, E., F. Kittrell, D. Medina, and S. M. Berget. 1999. Stage-specificchanges in SR splicing factors and alternative splicing in mammary tumor-igenesis. Oncogene 18:3574–3582.

42. Tanaka, K., A. Watakabe, and Y. Shimura. 1994. Polypurine sequenceswithin a downstream exon function as a splicing enhancer. Mol. Cell. Biol.14:1347–1354.

43. Umen, J. G., and C. Guthrie. 1995. The second catalytic step of pre-mRNAsplicing. RNA 1:869–885.

44. Wang, J., and J. L. Manley. 1995. Overexpression of the SR proteins ASF/SF2 and SC35 influences alternative splicing in vivo in diverse ways. RNA1:335–346.

45. Wu, J., and J. L. Manley. 1989. Mammalian pre-mRNA branch site selectionby U2 snRNP involves base pairing. Genes Dev. 3:1553–1561.

46. Zamore, P. D., J. G. Patton, and M. R. Green. 1992. Cloning and domainstructure of the mammalian splicing factor U2AF. Nature 355:609–614.

47. Zhang, L., and C. M. Stoltzfus. 1995. A suboptimal src 39 splice site isnecessary for efficient replication of Rous sarcoma virus. Virology 206:1099–1107.

48. Zhang, M. Q. 1998. Statistical features of human exons and their flankingregions. Hum. Mol. Genet. 7:919–932.

49. Zheng, Z. M., and C. C. Baker. 2000. Parameters that affect in vitro splicingof bovine papillomavirus type 1 late pre-mRNAs. J. Virol. Methods 85:203–214.

50. Zheng, Z. M., P. J. He, and C. C. Baker. 1996. Selection of the bovinepapillomavirus type 1 nucleotide 3225 39 splice site is regulated through anexonic splicing enhancer and its juxtaposed exonic splicing suppressor. J. Vi-rol. 70:4691–4699.

51. Zheng, Z. M., P. J. He, and C. C. Baker. 1997. Structural, functional, andprotein binding analyses of bovine papillomavirus type 1 exonic splicingenhancers. J. Virol. 71:9096–9107.

52. Zheng, Z. M., P. J. He, and C. C. Baker. 1999. Function of a bovine papil-lomavirus type 1 exonic splicing suppressor requires a suboptimal upstream39 splice site. J. Virol. 73:29–36.

53. Zheng, Z. M., M. Huynen, and C. C. Baker. 1998. A pyrimidine-rich exonicsplicing suppressor binds multiple RNA splicing factors and inhibits spliceo-some assembly. Proc. Natl. Acad. Sci. USA 95:14088–14093.

54. Zhuang, Y., and A. M. Weiner. 1989. A compensatory base change in humanU2 snRNA can suppress a branch site mutation. Genes Dev. 3:1545–1552.

55. Zuo, P., and T. Maniatis. 1996. The splicing factor U2AF35 mediates criticalprotein-protein interactions in constitutive and enhancer-dependent splic-ing. Genes Dev. 10:1356–1368.



http://jvi.asm.org/

Dow

nloaded from

http://jvi.asm.org/

optimization of a weak 39 splice site counteracts the...

Documents