Splicing Choices Research Articles

Intronic breakpoint definition and transcription analysis in DMD/BMD patients with deletion/duplication at the 5′ mutation hot spot of the dystrophin gene

Dystrophin mutations occurring at the 5′ end of the gene frequently behave as exceptions to the “frame rule,” their clinical severity being variable and often not related to the perturbation of the translation reading frame. The molecular mechanisms underlying the phenotypic variability of 5′ dystrophin mutations have not been fully clarified. We have characterized the genomic breakpoints within introns 2, 6 and 7 and identified the splicing profiles in a cohort of DMD/BMD patients with deletion of dystrophin exons 3–7, 3–6 and duplication of exons 2–4.Our findings indicate that the occurrence of intronic cryptic promoter as well as corrective splicing events are unlikely to play a role in exons 3–7 deleted patients phenotypic variability. Our data suggest that re-initiation of translation could represent a major mechanism responsible for the production of a residual dystrophin in some patients with exons 3–7 deletion.Furthermore, we observed that the out-of-frame exon 2a is almost constantly spliced into a proportion of the dystrophin transcripts in the analysed patients. In the exons 2–4 duplicated DMD patient, producing both in-frame and out-of-frame transcripts, this splicing behaviour might represent a critical factor contributing to the severe phenotype. In conclusion, we suggest that multiple mechanisms may have a role in modulating the outcome of 5′ dystrophin mutations, including recoding mechanisms and unusual splicing choices.

Sequence conservation, relative isoform frequencies, and nonsense-mediated decay in evolutionarily conserved alternative splicing

Studies of expressed sequence tag data sets have revealed large numbers of splicing variants for human genes, but it remains challenging to distinguish functionally important variants from aberrant splicing, clarify the nature of the alternative functions, and understand the signals that regulate splicing choices. To help address these issues, we have constructed and analyzed a large data set of 1,478 exon-skipping alternative splicing (AS) variants evolutionarily conserved in human and mouse. In about one-fifth of cases, one isoform appears subject to nonsense-mediated mRNA decay (NMD), supporting the idea that a major role of AS is to regulate gene expression; one-quarter of these NMD-inducing cases involve a conserved exon whose apparent sole purpose is to mediate destruction of the message when included. We explore sequence conservation likely related to splicing regulation, using in part a measure of the overall amount of conserved information in a sequence, and find that the increased conservation that has been observed within AS exons primarily affects synonymous sites, suggesting that regulatory signals significantly constrain synonymous substitution rates. We show that a lower frequency of the inclusion isoform relative to the exclusion isoform tends to be associated with weaker splice site signals, smaller exon size, and higher intronic sequence conservation, and provide evidence that all of these factors are under selection to control relative isoform frequencies. Some conserved instances of AS appear to represent aberrant splicing events that by chance have occurred in both species, and we develop a nonparametric likelihood approach to identify these.

Tau Exons 2 and 10, Which Are Misregulated in Neurodegenerative Diseases, Are Partly Regulated by Silencers Which Bind a SRp30c·SRp55 Complex That Either Recruits or Antagonizes htra2β1

Tau Exons 2 and 10, Which Are Misregulated in Neurodegenerative Diseases, Are Partly Regulated by Silencers Which Bind a SRp30c·SRp55 Complex That Either Recruits or Antagonizes htra2β1

Dscam-mediated Self- versus Non-Self-Recognition by Individual Neurons

Clearly, gene regulation (the turning on and off of the multitudinous genes in the genome) allows the specification of different parts of the nervous system as it allows specification of all parts of the developing animal. Even within a group of neurons, distinguishing similar neurons also can be accomplished by the specific turning on and off of various genes via differential activities of various transcription factors. Moreover, differences in levels of expression of different genes and both regulated and stochastic aspects of posttranscriptional and posttranslational events are all important. Alternative splicing represents another important mechanism that can render distinct cell populations different from one another. The mechanisms regulating the generation of alternative mRNA transcripts are known to have tissue specificity, and elucidation of this type of regulation is an important area of exploration. The generally accepted concept is that tissue-specific accessory splicing factors bind to cis-acting elements in the primary transcript and thus regulate the relative levels of different possible alternative splicing choices.

Targeting of a tropomyosin isoform to short microfilaments associated with the Golgi complex.

A growing body of evidence suggests that the Golgi complex contains an actin-based filament system. We have previously reported that one or more isoforms from the tropomyosin gene Tm5NM (also known as gamma-Tm), but not from either the alpha- or beta-Tm genes, are associated with Golgi-derived vesicles (Heimann et al., (1999). J. Biol. Chem. 274, 10743-10750). We now show that Tm5NM-2 is sorted specifically to the Golgi complex, whereas Tm5NM-1, which differs by a single alternatively spliced internal exon, is incorporated into stress fibers. Tm5NM-2 is localized to the Golgi complex consistently throughout the G1 phase of the cell cycle and it associates with Golgi membranes in a brefeldin A-sensitive and cytochalasin D-resistant manner. An actin antibody, which preferentially reacts with the ends of microfilaments, newly reveals a population of short actin filaments associated with the Golgi complex and particularly with Golgi-derived vesicles. Tm5NM-2 is also found on these short microfilaments. We conclude that an alternative splice choice can restrict the sorting of a tropomyosin isoform to short actin filaments associated with Golgi-derived vesicles. Our evidence points to a role for these Golgi-associated microfilaments in vesicle budding at the level of the Golgi complex.

Voltage-gated Ca 2+ channel Ca V1.3 subunit expressed in the hair cell epithelium of the sacculus of the trout Oncorhynchus mykiss: cloning and comparison across vertebrate classes

Full-length sequence (>6.5 kb) has been determined for the Ca V1.3 pore-forming subunit of the voltage-gated Ca 2+ channel from the saccular hair cells of the rainbow trout ( Oncorhynchus mykiss). Primary structure was obtained from overlapping PCR and cloned fragments, amplified by primers based on teleost, avian, and mammalian sources. Trout saccular Ca V1.3 was localized to hair cells, as evidenced by its isolation from an epithelial layer in which the hair cell is the only intact cell type. The predicted amino acid sequence of the trout hair cell Ca V1.3 is approximately 70% identical to the sequences of avian and mammalian Ca V1.3 subunits and shows L-type characteristics. The trout hair cell Ca V1.3 expresses a 26-aa insert in the I–II cytoplasmic loop (exon 9a) and a 10-aa insert in the IVS2–IVS3 cytoplasmic loop (exon 30a), neither of which is appreciably represented in trout brain. The exon 9a insert also occurs in hair cell organs of chick and rat, and appears as an exon in human genomic Ca V1.3 sequence (but not in the Ca V1.3 coding sequence expressed in human brain or pancreas). The exon 30a insert, although expressed in hair cells of chick as well as trout, does not appear in comparable rat or human tissues. Further, the IIIS2 region shows a splice choice (exon 22a) that is associated with the hair cell organs of trout, chick, and rat, but is not found in human genomic sequence. The elucidation of the primary structure of the voltage-gated Ca 2+ channel Ca V1.3 subunit from hair cells of the teleost, representing the lowest of the vertebrate classes, suggests a generality of sensory mechanism for Ca V1.3 across hair cell systems. In particular, the exon 9a insert of this channel appears to be the molecular feature most consistently associated with hair cells from fish to mammal, consonant with the hypothesis that the latter region may be a signature for the hair cell.

Polypyrimidine Tract-binding Protein Represses Splicing of a Fibroblast Growth Factor Receptor-2 Gene Alternative Exon through Exon Sequences

The fibroblast growth factor receptor (FGFR)-2 gene contains two mutually exclusive exons, K-SAM and BEK. We made a cell line designed to become drug-resistant on repression of BEK exon splicing. One drug-resistant derivative of this line carried an insertion within the BEK exon of a sequence containing at least two independent splicing silencers. One silencer was a pyrimidine-rich sequence, which markedly increased binding of polypyrimidine tract-binding protein to the BEK exon. The BEK exon binds to polypyrimidine tract-binding protein even in the silencer's absence. Several exonic pyrimidine runs are required for this binding, and they are also required for overexpression of polypyrimidine tract-binding protein to repress BEK exon splicing. These results show that binding of polypyrimidine tract-binding protein to exon sequences can repress splicing. In epithelial cells, the K-SAM exon is spliced in preference to the BEK exon, whose splicing is repressed. Mutation of the BEK exon pyrimidine runs decreases this repression. If this mutation is combined with the deletion of a sequence in the intron upstream from the BEK exon, a complete switch from K-SAM to BEK exon splicing ensues. Binding of polypyrimidine tract binding protein to the BEK exon thus participates in the K-SAM/BEK alternative splicing choice.

Regulation of alternative pre-mRNA splicing during erythroid differentiation.

Although the mature enucleated erythrocyte is no longer active in nuclear processes such as pre-mRNA splicing, the function of many of its major structural proteins is dependent on alternative splicing choices made during the earlier stages of erythropoiesis. These splicing decisions fundamentally regulate many aspects of protein structure and function by governing the inclusion or exclusion of exons that encode protein interaction domains, regulatory signals, or translation initiation or termination sites. Alternative splicing events may be partially or entirely erythroid-specific, ie, distinct from the splicing patterns imposed on the same transcripts in nonerythroid cells. Moreover, differentiation stage-specific splicing "switches" may alter the structure and function of erythroid proteins in physiologically important ways as the cell is morphologically and functionally remodeled during normal differentiation. Derangements in the splicing of individual mutated pre-mRNAs can produce synthesis of truncated or unstable proteins that are responsible for numerous erythrocyte disorders. This review will summarize the salient features of regulated alternative splicing in general, review existing information concerning the widespread extent of alternative splicing among erythroid genes, and describe recent studies that are beginning to uncover the mechanisms that regulate an erythroid splicing switch in the protein 4.1R gene.

Splicing and the single cell.

The selection of alternative splice sites is an important component of cell-specific gene regulation in eukaryotic cells. Use of splice sites can be positively and negatively regulated, and often physiologically appropriate splice site choice is achieved by a balance of the two. RNA elements controlling splice site choice are found in both exons and introns, and these determine management by the cellular splicing machinery. However, the molecular basis of how the splicing machinery responds to these signals in different cells is somewhat of a paradox. Thus far the identified proteins which bind to tissue/cell-specific regulatory elements in mammals are expressed in many different tissues, and not just in the regulating tissue. Potential tissue-specific splicing regulators have been identified by non-biochemical means. However, alternative splicing choices are likely to be affected by subtle differences in the splicing machinery in different cells. In this review I suggest that one important factor is the ratio of proteins in different nuclear compartments, which might be established in a cell type specific fashion.

Concerted regulation and molecular evolution of the duplicated SNRPB'/B and SNRPN loci.

The human small nuclear ribonucleoprotein SNRPB ' /B gene is alternatively spliced to produce the SmB or SmB' spliceosomal core proteins. An ancestral duplication gave rise to the closely related SNRPN paralog whose protein product, SmN, replaces SmB'/B in brain. However, the precise evolutionary and functional relationship between these loci has not been clear. Genomic, cDNA and protein analyses presented here in chicken, two marsupials (South American opossum and tammar wallaby), and hedgehog, suggest that the vertebrate ancestral locus produced the SmB' isoform. Interestingly, three eutherians exhibit radically distinct splice choice expression profiles, producing either exclusively SmB in mouse, both SmB and SmB' in human, or exclusively SmB' in hedgehog. The human SNRPB ' /B locus is biallelically unmethylated, unlike the imprinted SNRPN locus which is unmethyl-ated only on the expressed paternal allele. Western analysis demonstrates that a compensatory feedback loop dramatically upregulates SmB'/B levels in response to the loss of SmN in Prader-Willi syndrome brain tissue, potentially reducing the phenotypic severity of this syndrome. These findings imply that these two genes encoding small nuclear ribonucleoprotein components are subject to dosage compensation. Therefore, a more global regulatory network may govern the maintenance of stoichiometric levels of spliceosomal components and may constrain their evolution.

Assembly of thick filaments and myofibrils occurs in the absence of the myosin head

We investigated the importance of the myosin head in thick filament formation and myofibrillogenesis by generating transgenic Drosophila lines expressing either an embryonic or an adult isoform of the myosin rod in their indirect flight muscles. The headless myosin molecules retain the regulatory light-chain binding site, the alpha-helical rod and the C-terminal tailpiece. Both isoforms of headless myosin co-assemble with endogenous full-length myosin in wild-type muscle cells. However, rod polypeptides interfere with muscle function and cause a flightless phenotype. Electron microscopy demonstrates that this results from an antimorphic effect upon myofibril assembly. Thick filaments assemble when the myosin rod is expressed in mutant indirect flight muscles where no full-length myosin heavy chain is produced. These filaments show the characteristic hollow cross-section observed in wild type. The headless thick filaments can assemble with thin filaments into hexagonally packed arrays resembling normal myofibrils. However, thick filament length as well as sarcomere length and myofibril shape are abnormal. Therefore, thick filament assembly and many aspects of myofibrillogenesis are independent of the myosin head and these processes are regulated by the myosin rod and tailpiece. However, interaction of the myosin head with other myofibrillar components is necessary for defining filament length and myofibril dimensions.

Induction of the N-methyl-D-aspartate receptor subunit 1 in the immortalized neuronal progenitor cell line HC2S2 during differentiation into neurons.

Conditionally immortalized neuronal progenitor cell line HC2S2 differentiates into mature neurons after suppression of the v-myc expression with tetracycline. Reverse transcription-polymerase chain reaction analyses were used to measure expression levels of N-metyl-D-aspartate receptor subunit 1 (NMDAR1) mRNAs encoding splice variants (NMDAR1a, -exon 5; NMDAR1b, +exon 5) in HC2S2 cells during the differentiation. Differential induction of NMDAR1a and NMDAR1b mRNAs was observed during the differentiation. Very low expression of NMDAR1 was observed in undifferentiated HC2S2 cells. NMDAR1a mRNA was induced coincidentally with the emergence of neurites, whereas NMDAR1b mRNA was induced at the time of network formation. Immunohistochemistry also demonstrated induction of NMDAR1 immunoreactivity in differentiated HC2S2 cells. In addition, expression of NMDAR2 mRNA and immunoreactivity was observed in undifferentiated and differentiated HC2S2 cells, suggesting that functional NMDA receptors are present in differentiated HC2S2 cells. While exposure to NMDA resulted in almost no cell death in undifferentiated HC2S2 cells, NMDA induced cell death in differentiated HC2S2 cells in a dose-dependent fashion. These findings suggest that the expression of NMDAR1 mRNA may be regulated by myc or its counterpart during neuronal terminal differentiation and that the splicing choice between NMDAR1a and NMDAR1b may vary according to the formation of neuronal network.

An intronic sequence element mediates both activation and repression of rat fibroblast growth factor receptor 2 pre-mRNA splicing.

Alternative splicing of fibroblast growth factor receptor 2 (FGF-R2) is an example of highly regulated alternative splicing in which exons IIIb and IIIc are utilized in a mutually exclusive manner in different cell types. The importance of this splicing choice is highlighted by studies which indicate that deregulation of the FGF-R2 splicing is associated with progression of prostate cancer. Loss of expression of a IIIb exon-containing isoform of FGF-R2 [FGF-R2 (IIIb)] accompanies the transition of a well-differentiated, androgen-dependent rat prostate cancer cell line, DT3, to the more aggressive, androgen-independent AT3 cell line. We have used transfection of rat FGF-R2 minigenes into DT3 and AT3 cancer cell lines to study the mechanisms that control alternative splicing of rat FGF-R2. Our results support a model in which an important cis-acting element located in the intron between these alternative exons mediates activation of splicing using the upstream IIIb exon and repression of the downstream IIIc exon in DT3 cells. This element consists of 57 nucleotides (nt) beginning 917 nt downstream of the IIIb exon. Analysis of mutants further demonstrates that an 18-nt "core sequence" within this element is most crucial for its function. Based on our observations, we have termed this sequence element ISAR (for intronic splicing activator and repressor), and we suggest that factors which bind this sequence are required for maintenance of expression of the FGF-R2 (IIIb) isoform.

Tissue‐specific alternative splicing of Shaker potassium channel transcripts results from distinct modes of regulating 3′ splice choice

Tissue‐specific alternative splicing of Shaker potassium channel transcripts results from distinct modes of regulating 3′ splice choice

Tissue-specific alternative splicing of Shaker potassium channel transcripts results from distinct modes of regulating 3? splice choice

Alternative splicing of precursor RNA enables a single gene to encode multiple protein isoforms with different functional characteristics and tissue distributions. Differential splicing of Drosophila Shaker (Sh) gene transcripts regulates the tissue-specific expression of kinetically distinct potassium ion channels throughout development. Regulation of Sh alternative splicing is being examined in germline transformants using lacZ as a reporter gene. P-element constructs were generated in which one or both of the two mutually exclusive Sh 3' acceptor sites were positioned in the same translational reading frame as the lacZ coding sequences. The constructs were introduced into the germline and the transgenic animals examined for tissue-specific beta-galactosidase expression patterns. Some tissues exhibit "promiscuous" splicing; these tissues are competent to splice to either 3' acceptor even when both are present on the same pre-mRNA. In other tissues splice choice results from competition between the two 3' sites; these tissues can splice to either site when it is the only available 3' acceptor, but when given a choice will splice to only one of the two 3' acceptors. In some tissues, splicing occurs exclusively at only one of the 3' acceptor sites; these tissues are not competent to splice to one of the sites even if it is the only 3' acceptor present on the pre-mRNA. These results suggests that multiple, distinct regulatory modes are operating to control tissue-specific alternative splicing of Sh 3' domains and are discussed in terms of potential underlying mechanisms for regulating the tissue-specific expression of alternatively spliced genes.

Reflections on STAT3, STAT5, and STAT6 as fat STATs.

The current issue of the Proceedings contains an article entitled Defective STAT signaling by the leptin receptor in diabetic by Ghilardi et al. (1). The reported results suggest how leptin, the recently discovered weight control hormone (2), may signal cells through its cognate receptor by activation of STATs, proteins that serve the dual function of signal transducers and activators of transcription in cells exposed to signaling polypeptides (3, 4). Mice that produce no leptin (obese or ob mutants) weigh up to 60 g instead of the usual 15-20 g for a normal mouse. The human protein is virtually identical to mouse leptin, suggesting that the control of body weight in humans may also be regulated by this hormone (2). The effect of leptin on ob mice is to control food intake, so that weight loss ensues (5-7). In addition, the mice exhibit increased mouse-like exploratory activity. Thus, the description of the first molecule in the weight control pathway opens up the chance to explore in molecular detail the control of a complex behavior. The Ghilardi et al. paper (1) reports confirmatory results showing the presence in cells of widely scattered tissues, including the hypothalamus, the putative control center for feeding behavior, of a and leptin receptor (8-10). The leptin receptor has considerable sequence similarity to the gp130 transmembrane receptor chain that pairs as the signaling molecule with a number of other transmembrane proteins to constitute the receptor for many ligands including interleukin (IL)-6, ciliary neurotrophic factor, leukemiainhibitory factor (11, 12). The leptin receptor appears not to function normally in the mouse mutant termed diabetes (db) because of a base change in an intron that leads to a frequent aberrant splice choice; the resulting mRNA retains a translation stop codon producing a truncated protein lacking approximately 270 amino acids of the cytoplasmic domain of the transmembrane receptor (9). The omission of these amino acids was hypothesized to prevent intracellular signaling occasioned by leptin binding to its cell surface receptor. Based on the homology between the leptin receptor and the gp130 transmembrane protein (8-12), the pathway through which the leptin receptor seemed likely to signal is the recently recognized JAK/STAT pathway (3, 4). All of the known receptors that contain gp130 have JAK kinases (tyrosine kinases) bound to their intracellular tails (11). After ligand-mediated receptor assembly, the JAKs become phosphorylated on tyrosine and thereby activated as tyrosine kinases. The intracellular tail of one or more receptor chains is then phosphorylated on one or more tyrosine residues (13-15), offering binding sites to the Src homology 2 groups of latent cytoplasmic proteins called STATs (15). The attached STATs become phosphorylated on tyrosine by the activated Jak kinases. The STATs then dimerize, translocate to the nucleus, and participate in transcriptional regulation by binding to specific DNA sites. In the mutant db receptor both the putative STAT-binding sites and the JAK-binding sites are missing. There are six mouse and human STATs known at present (seven if the duplicated STAT5A and STAT5B genes are considered as two) and at least STATI, STAT3, and STAT5 exhibit differentially spliced forms. Over 30 different polypeptides have been recorded that cause STAT activation in various mammalian cells (3, 4, 11). The most potent activation of STATs through the gp130-containing receptors is of STAT3 (15-18) and by sequence comparison the wild-type leptin receptor has potential docking sites for STAT3 molecules. Ghilardi et al. (1) now report success in showing STAT activation dependent on the long form of the leptin receptor. In a comprehensive set of experiments they cotransfected COS cells with either the long or short versions of the leptin receptor together with, individually, each of six mouse or human STAT proteins. Leptin treatment of the cells transfected with the long but not the short receptor resulted in activation of DNAbinding complexes containing, individually, STAT3, STAT5, or STAT6; STATI, STAT2, and STAT4 were not detectably activated. It will now be crucial to show whether the same set of STATs is activated by leptin in the hypothalamus, which is hypothesized to be the center for weight control in the Coleman model (19). In this model the ob gene product, a circulating hormone, would operate by binding to a hypothalamic receptor, the db gene product, to regulate feeding.

The choice between alternative IIIb and IIIc exons of the FGFR-3 gene is not strictly tissue-specific

An essential feature of fibroblast growth factor receptors (FGFR) is the existence of multiple possibilities for alternative splicing. One of these concerns sequences of the mRNA coding for the C-terminal half of Ig domain 3 which correpponds to a part of the ligand-binding site. Two alternative exons, IIIb and IIIc, encode the C-terminal half of Ig domain 3. We show here that the alternative splicing choice between IIIb and IIIc exons of the FGFR-3 is not strictly tissue-specific: epithelial cells show exclusively IIIb transcripts while fibroblastic cells show a mixture of IIIb and IIIc transcripts. This is in contrast with the strictly exclusive alternative choice between IIIb or IIIc exons of the FBFR-2 gene: epithelial cells make only the IIIb choice while fibroblastic cells make only the IIIc choice.

Tissue-specific alternative splicing of hybrid Shaker/lacZ genes correlates with kinetic differences in Shaker K + currents in vivo

Alternative splicing of the Shaker (Sh) locus of Drosophila generates several transcripts with divergent 5′ and 3′ domains that produce kinetically distinct K + currents in Xenopus oocytes. Although suggestive that alternative splicing may be involved in generating K + channel diversity, clear tissue-specific differences in the distribution of particular Sh gene products have not been demonstrated. Using IacZ as a reporter gene for accurate splicing of variable Sh 3′ domains, we observe differences in β-galactosidase expression patterns in transgenic animals that indicate both temporal and spatial regulation of 3′ splice choice. The differences in 3′ splice choice can account for variation in recovery kinetics of Sh-encoded K + currents recorded in adult flies. The results indicate that tissue-specific expression of functionally distinct Sh K + channels is regulated, in part, at the level of pre-mRNA splicing and implicate sequences in or around the 3′ splice sites in regulating the choice of 3′ domain.

ONCOPROTEIN FOS ACTIVATION IN EPITHELIAL-CELLS INDUCES AN EPITHELIOMESENCHYMAL CONVERSION AND CHANGES THE RECEPTOR ENCODED BY THE FGFR-2 MESSENGER-RNA FROM K-SAM TO BEK

Activation of c-Fos, by using an inducible c-Fos estrogen receptor fusion protein, triggers the epitheliofibroblastoid cell conversion of mouse mammary epithelial cells. We show that this change in phenotype is accompanied by a definitive switch of the fibroblast growth factor receptor 2 from K-SAM to BEK. This splicing switch occurs a few hours after estrogen stimulation. Our data suggest that Fos proteins could be important in modulating the FGFR-2 splicing choice. Moreover, these observations reinforce previous evidence that the BEK/K-SAM choice is strictly tissue-specific: the K-SAM exon is expressed exclusively in epithelial cells, the BEK exon in cells of the fibroblastic type.

Splice site selection in polyomavirus late pre-mRNA processing

Polyomavirus late pre-mRNAs contain one 5' splice site and two message body 3' splice sites, which are not used at equal frequencies. As a result of alternative splicing, the total late mRNA population consists of about 5% mVP2 (no message body splice chosen), about 15% mVP3 (promoter-proximal 3' splice site chosen), and about 80% mVP1 (promoter-distal 3' splice site chosen). To determine whether it is splice site strength that determines the ratio of spliced products, constructs containing duplicated or rearranged 3' splice sites were created. In construct VP1,1, 160 bp surrounding the VP3 3' splice site was substituted with the corresponding region of the VP1 3' splice site. This construct resulted in the duplication of the VP1 3' splicing signal. VP3,3 (two identical VP3 3' splice sites) and VP1,3 (VP1 and VP3 3' splice sites reversed) were similarly created. Each construct maintained wild-type spacing between the 3' splice sites. Analysis of RNAs from transfections showed that in each construct, the 3' splice closest to the polyadenylation site was used preferentially. Analysis of a number of additional constructs indicated that there are no strong cis-acting positive or negative regulators of polyomavirus late splicing; rather, splicing choices appear to be determined largely by relative position of splice sites.