Oncogene
SEARCH     advanced search my account e-alerts subscribe register
Journal home
Advance online publication
Current issue
Archive
Press releases
For authors
For referees
Contact editorial office
About the journal
For librarians
Subscribe
Advertising
naturereprints
Contact NPG
Customer services
Site features
NPG Subject areas
Access material from all our publications in your subject area:
Biotechnology Biotechnology
Cancer Cancer
Chemistry Chemistry
Dentistry Dentistry
Development Development
Drug Discovery Drug Discovery
Earth Sciences Earth Sciences
Evolution & Ecology Evolution & Ecology
Genetics Genetics
Immunology Immunology
Materials Materials Science
Medical Research Medical Research
Microbiology Microbiology
Molecular Cell Biology Molecular Cell Biology
Neuroscience Neuroscience
Pharmacology Pharmacology
Physics Physics
Browse all publications
 
21 June 2001, Volume 20, Number 28, Pages 3683-3694
Table of contents    Previous  Article  Next   [PDF]
Original Paper
Modulation in vitro of H-ras oncogene expression by trans-splicing
Carles Codony1, Sònia Guil1, Concha Caudevilla2, Dolors Serra2, Guillermina Asins2, Adolf Graessmann3, Fausto G Hegardt2 and Montse Bach-Elias1

1IIBB-CSIC (Instituto de Investigaciones Biomédicas de Barcelona-Consejo Superior de Investigaciones Científicas), Dept. PMT, Unidad de Biología y Farmacología Molecular del Cáncer, c/Jorge Girona Salgado 18-26, 08034 Barcelona, Spain

2Department of Biochemistry and Molecular Biology, School of Pharmacy, University of Barcelona, Spain

3Institut für Molekularbiologie und Biochimie, Freie Universität Berlin, Arnimallee 22, 14195 Berlin, Germany

Correspondence to: M Bach-Elias, IIBB-CSIC, c/Jorge Girona Salgado 18-26, 08034 Barcelona, Spain

Abstract

In man, activated N-, K- and H-ras oncogenes have been found in around 30% of the solid tumours tested. An exon known as IDX, which has been described previously and is located between exon 3 and exon 4A of the c-H-ras pre-mRNA, allows an alternative splicing process that results in the synthesis of the mRNA of a putative protein named p19. It has been suggested that this alternative pathway is less tumorigenic than that which results in the activation of p21. We have used the mammalian trans-splicing mechanism as a tool with which to modulate this particular pre-mRNA processing to produce mRNA similar to that of mature p19 RNA. The E4A exon of the activated H-ras gene was found to be a good target for external trans-splicing. We reprogrammed the rat carnitine octanoyltransferase exon 2 to specifically invade the terminal region of H-ras. Assays performed with this reprogrammed trans-exon showed that the trans-splicing product was obtained in competition with cis-splicing of the D intron of the H-ras gene, and was associated with concomitant down-modulation of D intron cis-splicing. We also found that the exon 4A of the human c-H-ras gene underwent successive trans-splicing rounds with an external exon. Oncogene (2001) 20, 3683-3694.

Keywords

splicing; c-H-ras; trans-splicing

Introduction

The c-H-ras gene, and other members of the ras family (N- and K-ras) are proto-oncogenes, the activated forms of which are present in approximately 30% of all solid human tumours that have been examined (reviewed in Barbacid, 1987). The presence of H-ras activated genes varies from tumour to tumour, but it appears to be activated in a variety of induced cancers in different animal model systems, with an incidence of between 80 and 100%, depending upon the location of the tumour (mammary, skin or hepatocellular carcinomas). This suggests that H-ras and other ras-family oncogenes (N-ras and K-ras) are likely to participate in the initiation of tumour development (Barbacid, 1987). The p21 protein, which is the product of c-H-ras mRNA, belongs to the G-protein family and has an important function in the transduction of external mitogenic signals to the cell. p21 binds to the membrane through its C-terminal domain, in a manner that is dependent upon the Cys186 residue of c-H-ras exon (E) 4A (Barbacid, 1987; Lowy and Willumsen, 1993). As a result of the importance of the E4A exon in the activity of p21, any modification of the Cys186 residue will result in marked reduction in p21 activity (Barbacid, 1987; Lowy and Willumsen, 1993).

The regulatory pathway of the c-H-ras gene by alternative splicing was first described in 1989 (Cohen et al., 1989). The D intron of the c-H-ras gene contains an additional exon between (ras)E3 and (ras)E4A (see Figure 1), named IDX. Thanks to this exon, the pre-mRNA may be processed in at least two ways by an alternative splicing reaction that may yield two mature mRNAs: ras E0-E1-E2-E3-E4-E4B and ras E0-E1-E2-E3-IDX-E4A-E4B. The (ras)IDX (intron D exon, between exons E3 and E4A of the c-H-ras gene) includes a premature stop-codon that may yield a smaller truncated putative protein, p19, (see Figure 1). The presence of the (ras)IDX exon in mature RNA would result in the introduction of a termination codon upstream of two subsequent exons. Therefore p19 mRNA may be subjected to nonsense-mediated decay with consequent decreased baseline levels, yielding low levels of p19 protein (Cohen et al., 1989). Although the p19 protein has not yet been characterized, p19 cDNA has been detected in a retinoblastoma library, in human placenta and in rat brain, which confirms that p19 mRNA is present (Cohen et al., 1989). The results of in vivo transfection experiments suggest - as might be expected - that activated p19 does not exhibit transforming activity (Cohen et al., 1993), presumably because the protein does not contain (ras)E4A and is thus unable to anchor correctly to the inner membrane. Since the presence of the (ras)IDX inside the D intron of the c-H-ras gene is conserved in different species (human, rat, mouse and hamster), and two alternative mRNAs have been detected, it is possible that (ras)IDX participates in the fine tuning of the regulation of this gene (Cohen et al., 1989).

In a recent report it was demonstrated - by stable and transient transfection with expression vectors containing external cDNA (without introns) - that the p19 protein is expressed in human 293 cells (Huang and Cohen, 1997). These results suggest that the p19 protein may be a negative regulator of p21.

It has been shown that a newly developed, natural splicing reaction (that is thought to involve unconventional splicing) might play a role in the expression of some mammalian genes, when performed in combination with the cis-splicing process (Akopian et al., 1999; Breen and Ashcroft, 1997; Caudevilla et al., 1998; Kawasaki et al., 1999; Li et al., 1999; Shimizu and Honjo, 1993; Sullivan et al., 1991; Vellard et al., 1992; Zaphiropoulos, 1999; Frantz et al., 1999). One example is rat carnitine octanoyltransferase (COT) pre-mRNA, which undergoes a natural trans-splicing (Caudevilla et al., 1998) process that yields three COT mRNAs: one with a duplication of the COT E2 exon (¼.E2-E2¼); a second with a duplication of the COT E2-E3 exons (¼E2-E3-E2-E3¼) and a third with no exon duplication. In mammals, trans-splicing can be used to reprogram or interfere with specific pre-mRNAs, thus modifying or completely changing the gene expressed (Puttaraju et al., 1999; Mansfield et al., 2000). In the reprogramming of the target pre-mRNA, it is first necessary to establish which exons of the target pre-mRNA are susceptible to involvement in a trans-splicing reaction. Many external trans-RNAs (i.e. a trans-donor or a trans-acceptor that will participate in a trans-splicing reaction) can be synthesized to interfere with other pre-mRNAs (Bruzik and Maniatis, 1995; Chiara and Reed, 1995; Eul et al., 1995). In the work described here, however, we have chosen a recently discovered natural COT trans-donor and trans-acceptor to study the effects of interference with the D intron of the H-ras gene. The COT trans-donors/acceptors were chosen because when tested against other trans-RNAs, they proved to be strong donors and acceptors (this paper and Codony et al., 2001, unpublished data). Thus, by using either (COT)E2 or (COT)E3 as a donor, and (COT)E2 as an acceptor (in a heterologous trans splicing-reaction), we were able to establish which segment of the H-ras terminal region can be invaded by trans-splicing. In these trans-splicing assays, we have used the E3-D1 intron-IDX-D2 intron-E4A sequence from the H-ras gene, since interruption of the D2 intron may yield a protein without (ras)E4A, and thus reduce the levels of transforming p21 protein. Thus, the invasion of pre-mRNAs by reprogrammed trans-RNAs (donors or acceptors) might allow the modulation of a harmful message. The primary aim of this study was to investigate whether natural or reprogrammed trans-exons can invade harmful mRNA specifically to modulate its abundance. The pre-mRNAs we chose to be deactivated were the terminal sequence of the human H-ras oncogene, D intron, and flanking sequences (see Figure 1). If more p19 mRNA is obtained (with the IDX therefore being included in the mature message by splicing of the D1 intron), activated mRNAs have less tumorigenic activity (Cohen et al., 1993). Therefore, if the degree of D splicing is reduced, the pathway to p19 mRNA activation should be favoured. The first step of this work was to establish whether some of the terminal exons of H-ras could participate in a heterologous trans-splicing reaction.

Results

An in vitro cis-splicing assay of the D intron of H-ras

In order to test how trans-RNAs (donors or acceptors, shown in Figure 2) invade the cis-splicing machinery, we first developed an in vitro cis-splicing assay with a mini-gene containing the E3-D intron-E4A sequence of the H-ras gene. Figure 3 shows, for the first time, in vitro cis-splicing of the D intron of the H-ras gene in HeLa nuclear extracts with the K pre-mRNA. We detected a greater abundance of the (ras)E3-(ras)E4A cis-product than of the (ras)E3-(ras)IDX-(ras)E4A cis-product (Figure 3, lane 2). This suggests that in these in vitro conditions in HeLa cell nuclear extracts, the excision of the D intron is more efficient than the excision of the D1+D2 introns (see Figure 1 and Cohen et al., 1993). All of the intermediate and product bands obtained from the first and second steps of the cis-splicing reaction (Figure 3, lane 2), including lariats, were mapped by RT-PCR and DNA sequencing.

The (ras) E4A acceptor trans-splices, performed in vitro with a heterologous trans-donor

Our first approach was to see whether the D intron (D1+D2) or the D2 intron of H-ras is invaded by trans-splicing with the natural (COT)E2 trans-donor, as determined by an in vitro trans-splicing assay. In the in vitro trans-splicing reactions shown in Figure 4a, the K and L pre-mRNAs were used as acceptors, and the A pre-mRNA was the donor. In this experiment we also assayed the (N4=G) mutant that is described elsewhere (Cohen et al., 1989). The (N4=G) mutant was isolated from a tumour expressing activated H-ras, and contains a mutation on the fourth base (N4=G) of the (ras)IDX 5'splice site (5'SS), where an adenosine residue was changed to guanosine. Activated (N4=G) mutant pre-mRNA was described as having a higher transformation capacity than the activated wild-type (wt) gene (N4=A). It has been suggested that the (N4=G) mutation affects p21 expression, by lowering D2 intron splicing (Cohen et al., 1989). We therefore wanted to study how this mutation affects the trans-splicing reaction. The mutation was present in both the K and L constructions. Thus we performed a trans-splicing attack of the natural donor (COT)E2 (A pre-mRNA) to the pre-mRNAs containing intron D (K pre-mRNAs) or to the pre-mRNAs containing intron D2 (L pre-mRNAs), which contained both the wt (N4=A) and (N4=G) mutant sequences. The A donor trans-spliced well with acceptor (ras)E4A (see Figure 4, lanes 3-17). This trans-splicing reaction had a higher efficiency with the L construction than with K pre-mRNA, probably because the 5'SS sequence of the (ras)IDX is weaker tha that of the (ras)E3 exon when it has to splice up the (ras)E4A exon [see also Figure 3, which shows that (ras)E3-(ras)IDX-(ras)E4A is less abundant than (ras)E3-(ras)E4A]. Therefore, the A donor competed better with D2 intron splicing than with D intron splicing. The product (COT)E2-(ras)IDX was not detected with any of the pre-mRNAs used, which suggested initially that the (ras)IDX acceptor did not splice well with this donor (this observation was studied further in the experiments described below). No significant differences were seen between the wt (N4=A) and mutant (N4=G) pre-mRNAs (Figure 4a: compare lanes 3-5 with lanes 6-9, and lanes 10-13 with lanes 14-17).

The (ras)E4A could be both an acceptor and a donor and is able to attack itself in successive rounds of in vitro trans-splicing

A second band was detected above the (COT)E2-(ras)E4A product band in this trans-splicing assay (Figure 4a, lanes 12, 13, 16 and 17). Sequencing of the upper band revealed that it corresponds to a second round of trans-splicing reaction in which the (COT)E2-(ras)E4A product re-attacked the L pre-mRNA, duplicating (ras)E4A and yielding the product (COT)E2-(ras)E4A-(ras)E4A (Figure 4a, lanes 12, 13, 16 and 17). This reaction is made possible by the fact that the L pre-mRNA contains a 5'SS sequence hanging from the (ras)E4A exon that is also present in (COT)E2-(ras)E4A. This band is hard to see in trans-splicing with K pre-mRNA, probably because this construct results in the production of a smaller amount of trans-splicing product (compare lanes 3-9 with lanes 10-17), and also because cis-splicing of the D intron, in vitro, is far more efficient than of the D2 intron (see Figure 3 above and Guil et al., 2001, unpublished results).

RT-PCR amplification of the RNA products obtained from in vitro cis-splicing reactions of L pre-mRNA (without adding the trans-donor A) yielded the (ras)IDX-(ras)E4 cis-product. However, little (ras)E4A-(ras)E4A trans-product was detected (data not shown). This indicates that the homologous (ras)E4Ax(ras)E4A trans-splicing of L pre-mRNA is very inefficient under these conditions. Therefore, duplication of the (ras)E4A exon is not favoured during the cis-splicing reaction, but the amount of duplicated (ras)E4A exon may be increased when a heterologous trans-donor starts a trans-splicing reaction with the (ras)E4A acceptor. This suggests that once an exon has started a trans-splicing event, the product of this reaction may undergo successive trans-splicing rounds if it still has a 5'SS sequence. In accordance with this hypothesis, we found a third band above the (COT)E2-(ras)E4A-(ras)E4A product (Figure 4a, arrowhead in lanes 16 and 17) which has a length that is compatible with a trans-spliced (COT)E2-(ras)E4A-(ras)E4A-(ras)E4A product, containing three (ras)E4A exons.

In order to better demonstrate that the trans-splicing (COT)E2-(ras)E4A product - which contains a 5'SS sequence hanging from the (ras)E4A exon (Figure 4a) - can be a donor in a second trans-splicing event, this band was cut from the gel and re-assayed as a trans-splicing donor. In addition, the (ras)E3-(ras)E4A cis-product (see Figure 3), which also contained a 5'SS sequence hanging from the (ras)E4A exon, was assayed in parallel as a trans-exon donor and compared with the (COT)E2-(ras)E4A trans-product (the product of the first trans-splicing round described in Figure 4a). In both cases the L pre-mRNA was used as the acceptor (see drawing in Figure 4b). Both donors performed a second trans-splicing round with the (ras)E4A 3'SS sequence of the L pre-mRNA acceptor (Figure 4b, lanes 3, 4, 7 and 8), yielding (COT)E2-(ras)E4A-(ras)E4A and (ras)E3-(ras)E4A-(ras)E4A. The trans-splicing products (COT)E2-(ras)E4A-(ras)E4A and (ras)E3-(ras)E4A-(ras)E4A yielded a double band which, after RT-PCR sequencing, proved to be the same product; the high mobility bands were the result of some nucleotide degradation at the 3' end. Again, some minor bands were observed above the trans-splicing products (Figure 4b, arrowheads in lanes 3, 4, 7 and 8) with molecular weights that were compatible with a third round of trans-splicing that may include a third (ras)E4A. No remarkable differences were observed between wt (N4=A) and mutant (N4=G); therefore, no more experiments were performed with the (N4=G) mutant.

To determine whether the capability of the (ras)E4A and its 5'SS sequence to be a trans-donor is due to its own sequences or to enhancer sequences present in (COT)E2 (in G pre-mRNA) or in (ras)E3 (in E pre-mRNA), both located upstream of (ras)E4A, trans-splicing reactions were performed with (ras)E4A alone. The F pre-mRNA, which contained neither the (COT)E2 nor (ras)E3 sequences, was reacted with the L pre-mRNA (see drawing in Figure 4c ) and the results were compared with those shown in Figure 4b. The (ras)E4A donor trans-spliced with the (ras)E4A acceptor of the L pre-mRNA (see lanes 3-6 in Figure 4c); thus the (ras)E4A and flanking regions are a trans-RNA that has the capacity to be either the donor or the acceptor in a heterologous trans-splicing reaction that may yield mRNAs with duplicated (ras)E4A. Again a minor band with a molecular weight that was compatible with a second trans-splicing round was observed above the first trans-splicing product.

We tested whether the (ras)E4A exon was duplicated in vivo in human mRNA. We performed PCR amplifications with cDNAs obtained from the HeLa S3 cell line ATCC CCL2 (Clontech). In all of the assays performed we found the p21 mRNA, but we detected no duplication of the (ras)E4A exon in the expressed cDNAs of this cell line (not shown). We also did not detect (ras)E4A duplications in genomic human DNA.

The (ras)IDX acceptor does not trans-splice with other heterologous donors

The observation that the (ras)IDX acceptor does not splice with the (COT)E2 donor was studied further in the following experiments. When we analysed the sequences of the (ras)IDX and the flanking regions, we observed that the (ras)IDX is 82 nucleotides (nt) long (Cohen et al., 1989). Work performed by Chiara and Reed (1995) with an adenovirus mini-gene has shown that an in vitro trans-splicing reaction may occur in human extracts, with an acceptor containing a 3'SS sequence and a 5'SS sequence separated by 50 nt. Although the distance between the 3'SS and 5'SS sequences in the (ras)IDX is greater than 50 nt, we first studied whether the (ras)IDX and flanking sequences could act as a 3'SS acceptor for several 5'SS donors. For this study, four trans-donors were chosen as partners of the (ras)IDX acceptor in trans-splicing reactions, for the following reasons: (1) previous studies have shown that (adenovirus) C pre-mRNA is an in vitro trans-donor (Chiara and Reed, 1995), (2) (COT)A and (COT)B pre-mRNAs are in vivo natural trans-donors that trans-splice in vitro (Caudevilla et al., 1998), and (3) (ras)E3 is the natural 5'SS partner of the (ras)IDX in a cis-splicing reaction.

The trans-splicing reactions between pre-mRNAs C and M, D and M, A and M, and B and M yielded no trans-splicing products (our unpublished results). This assay was also performed at several different donor concentrations, with the same result (data not shown). Thus, the (ras)IDX acceptor is a weak 3'SS partner for the four donors tested in an in vitro trans-splicing assay.

In order to determine whether the separation of 82 nt between the 3'SS and the 5'SS sequences in the (ras)IDX was the cause of its failure as a trans-acceptor, we created the (O) pre-mRNA, which contained a 5'SS that was 50 nt away from the 3'SS of the (ras)IDX. Figure 5a shows that radio-labelled O pre-mRNA did not function as a trans-acceptor of the adenovirus C pre-mRNA donor (see lanes 6-9). The expected product of a C´O trans-splicing would have been a 237-nt molecule, which was not detected in the present study. The C´O trans-splicing experiment was performed with other donors and under other conditions, but still no products were obtained (data not shown). These results indicate that the presence of a 5'SS near a 3'SS does not necessarily activate the in vitro trans-splicing of the (ras)IDX.

A (ras)IDX acceptor, modified with an enhancer sequence element, trans-splices in vitro

A second feature that is thought to be important for an in vitro trans-splicing reaction in mammalian cells with conventional pre-mRNAs is the presence of an enhancer sequence element (ESE), such as (GAR)n, near the 3'SS of the trans-acceptor (Bruzik and Maniatis, 1995; Chiara and Reed, 1995). The (ras)IDX contains no (GAR)n repeats in its sequence. Therefore, the ESE sequence of the ASLV (Tanaka et al., 1994) was added to the (ras)IDX, 50 nt away from its 3'SS, in order to determine whether this sequence could activate trans-splicing of the (ras)IDX. This molecule was named N pre-mRNA. It has been reported that the ESE sequence of ASLV is a trans-splicing activator (Bruzik and Maniatis, 1995; Chiara and Reed, 1995). The trans-splicing reaction of C´N pre-mRNAs (N pre-mRNA being radio-labelled) resulted in a radio-labelled band with a length of around 261 nt (Figure 5a, lanes 2-4), which corresponded to the (adenovirus)5'Ad-(ras)IDX-ESE trans-spliced product. This indicated that (ras)IDX 3'SS could be activated by proteins that bind to ESE boxes located in the modified (ras)IDX, which then becomes a trans-acceptor in an in vitro trans-splicing reaction.

Two natural trans-donors (i.e. donors that perform natural trans-splicing in vivo) were also assayed with the N pre-mRNA containing the modified (ras)IDX-ESE acceptor: the SL RNA from Ascaris lumbricoides and COT donor A. A minor trans-splicing product was obtained when the SL RNA´N reaction was assayed (results not shown), but only a very small amount of trans-splicing product was detected in the trans-splicing of A´N (see Figure 5b, lanes 2-5, product indicated with an arrowhead on the right)) as compared with the natural trans-splicing reaction (COT)E2´(COT)E2, - the reaction of A´J pre-mRNAs, described in (Caudevilla et al., 1998) - which was also included as a control trans-splicing reaction in lanes 6-9 of Figure 5b. Again, double bands were attributable to nucleotide degradation at the product 3' end, as stated elsewhere (Caudevilla et al., 1998). These results indicate that, firstly, a 5'SS 50 nt away from the 3'SS (ras)IDX is not sufficient to activate it as an acceptor in a trans-splicing reaction. The activity of the 5'SS as an enhancer sequence in trans-splicing reactions is therefore dependent upon the acceptor sequences and the donor partner. Second, some combinations of partners are more efficient in yielding mammalian trans-splicing products (see Figure 5b, compare lanes 2-5 with lanes 6-9). Third, with some donors, the addition of an ESE enhancer box induces the (ras)IDX to become an acceptor in an in vitro trans-splicing reaction. From these results we concluded that with the donors tested, the (ras)IDX and its flanking sequences are not a good trans-splicing acceptor (at least in vitro), and that this exon is not the best choice for an external trans-RNA attack.

(ras)E3 acts as a trans-splicing donor in an in vitro trans-splicing reaction

The ability of (ras)E3 and its flanking 5'SS sequences to behave as a trans-donor in an in vitro trans-splicing reaction was also studied. As mentioned above, the wt (ras)IDX acceptor did not trans-splice with its natural cis-splicing 5' exon partner, (ras)E3 (unpublished results). Nevertheless, the artificial and modified (ras)IDX-ESE (N pre-mRNA) was a trans-acceptor for the (ras)E3 donor (D pre-mRNA). The D´N trans-reaction yielded the expected product (see Figure 5c, lanes 2-6), and the in vivo natural (COT)E2 also proved to be a trans-acceptor for the (ras)E3 donor (i.e. D´J trans-reaction, lanes 7-11). However, when both acceptors were used in competition in front of the (ras)E3 donor (by using increasing and equimolar amounts of N and J acceptors and 10.1 fmol of D donor; i.e. D´J´N), the D´N product remained almost undetected (Figure 5c, lanes 12-16), and the abundance of the D´J product was much greater (Figure 5c: compare lanes 1-11 with lanes 12-16). The inhibition effect seen by increasing the amounts of J acceptor (in lanes 9-11 and lanes 15 and 16) is a general behaviour of all of the different trans-splicing donor and acceptor pairs that we have tested so far. N acceptors also inhibit trans-splicing at higher concentrations than those shown in lanes 2-6 (our unpublished results), and this inhibition effect is dependent upon the sequence of each donor/acceptor pair. The differences in the abundance of product obtained indicate that J pre-mRNA is a stronger acceptor of the (ras)E3 donor (D pre-mRNA) than is N pre-mRNA. In an in vitro trans-splicing reaction, the (ras)E3 donor (D pre-mRNA) prefers to pair with the in vivo natural trans-acceptor (COT)E2 (i.e. with a heterologous exon) than with its proper 3'SS cis-exon (ras)IDX [in this case, the mutated (ras)IDX-ESE]. This result is in accordance with the results obtained with the (ras)IDX and flanking sequences, which proved to be weak, and remains quite weak even after the introduction of the enhancer sequences.

A reprogrammed (COT)E2 donor modulates the cis-splicing reaction of the H-ras D intron

As has been reported recently, trans-splicing may be a useful tool for gene reprogramming since it can modulate 'sick' pre-mRNAs via externally manipulated trans-RNAs (Puttaraju et al., 1999). Since the activated H-ras is a 'sick' message, we also investigated whether a trans-splicing reaction could modulate pre-mRNA cis-splicing of the H-ras D intron. To do this, the main point to establish is whether a targeted trans-donor can trans-splice to a H-ras D intron and, at the same time, decrease the abundance of (ras)E3-(ras)E4 product. With this aim, we reprogrammed the (COT)E2 so that it specifically targeted the H-ras D intron. This reprogrammed trans-RNA contained the natural (COT)E2 donor coupled to an additional sequence that recognized the (ras)D intron. This sequence was added to the 3' end of the A pre-mRNA, allowing a 70-nt base pairing with an internal region of the D intron (H pre-mRNA, see drawing in Figure 6). As a trans-donor control, a similar chimeric trans-donor was constructed, but with an inverted base-pairing region added, so that the trans-donor could perform trans-splicing but with no specificity for the D intron sequence. This latter construct did not contain the sequence that allowed 70-nt base pairing (I pre-mRNA; represented in Figure 6 by an inverted arrow). Both trans-exons were tested with the wt (N4=A) acceptor K pre-mRNA, which performed cis-splicing at the same time as the trans-splicing reaction (as indicated in Figure 3). Therefore, the trans-splicing reaction would compete with the cis-machinery. As shown in Figure 6, the (COT)E2 donor that allows base pairing with 70 nt of the D intron can compete with the cis-splicing reaction (ras)E3-(ras)E4A (compare lanes 2 and 4 in Figure 6), with a reduction in the abundance of the cis-product of around 50% (as quantified by densitometry). In three independent experiments, the (E2) donor that does not allow base pairing caused a reduction in the cis-splicing product of around 25%. This indicates that the (COT)E2 exon can exert some level of inhibition on cis-splicing, possibly by recruiting splicing factors from the available pool. This trans-donor also yielded the trans-splicing product (COT)E2-(ras)E4A (see lane 2 in Figure 6), while the trans-donor without the target sequence did not (compare lane 1 with lane 2 in Figure 6). The amount of radioactivity corresponding to the (COT)E2-(ras)E4A trans-product (Figure 6, lane 2) is not directly comparable with that of the (ras)E3-(ras)E4A cis-product (Figure 6, lane 2). The cis-product contains radio-labelled (ras)E3 and (ras)E4A exons, whereas the trans-product contains a non-radio-labelled (COT)E2 exon, which makes 50% of the non-radio-labelled trans-product. This result indicates that the base-pairing sequence chosen to target the D intron allows trans-splicing, and that the trans-exon down-modulates the cis-(ras)E3-(ras)E4A product. At the same time, as expected, the base-pairing target sequence increases the possibility that the trans-donor performs trans-splicing to (ras)E4A.

Discussion

The H-ras oncogene contains exons that could, potentially, be trans-spliced

The recent discovery that mammalian cells can perform trans-splicing naturally has incalculable consequences. It implies that the cell may be creating many RNA forms that have not yet been detected, and may be able to modify, create or modulate the expression or function of a protein. As described above, the H-ras oncogene is quite an interesting gene with regard to splicing. With the aim of invading the H-ras D intron, (ras)E3, (ras)IDX and (ras)E4A and flanking sequences were tested by in vitro trans-splicing with external trans-exons. As a 'fish-hook', we used the natural (COT)E2 exon to determine whether the (ras)E3, (ras)IDX and (ras)E4A exons and their flanking sequences could be trans-spliced with an external trans-RNA. As discussed above, the (ras)IDX and flanking sequences are not good trans-acceptors for all of the tested trans-donors. However, (ras)IDX is trans-spliced by adding an ESE purine-rich ehancer to this exon. Similar results have been obtained previously with other trans-RNAs (Bruzik and Maniatis, 1995; Chiara and Reed, 1995), which indicates that ESE purine-rich enhancers could activate trans-splicing events. The (ras)E3 and (ras)E4A exons and their flanking sequences proved to be good donors and acceptor/donors, respectively, for the COT trans-RNAs. The natural trans-spliced (COT)E2 contains three ESE sequences that follow the (GAR)n rule (Caudevilla et al., 1998), and more recent results demonstrate that at least one of these boxes could be essential for natural trans-splicing (Caudevilla et al., 2001, unpublished results). The (ras)E4A exon also contains purine-rich regions that might have a role as enhancer elements in cis- and trans-splicing, although this has not yet been studied. In conclusion, not all exons can be invaded by external trans-RNAs and, so far, the ESE boxes seem to play an important role in the activation of natural and externally designed trans-splicing events.

An interesting discovery was made with the trans-splicing of the (ras)E4A acceptor. After a first round of trans-splicing between the (COT)E2 donor and the (ras)E4A acceptor, a second round of trans-splicing was obtained between the product of the first trans-splicing reaction and the 3'SS of the (ras)E4A exon, a phenomenon we called successive trans-splicing rounds. Curiously, we also observed that it is even possible to obtain more trans-splicing rounds from the products of the previous trans-splicing reactions. Similar results have been found by one of us (A Graessmann) in vivo with the natural adenovirus trans-splicing reactions (data not shown), where successive trans-splicing rounds were also detected. In addition, Chiara and Reed (1995) also suggested that this phenomenon occurs in conventional adenovirus mini-genes. In explanation, since the nuclear machinery should splice out the introns, the 5'SS that hangs from the (COT)E2-(ras)E4A and (ras)E3-(ras)E4A products is committed to being spliced out. Therefore, as the available 3'SS exon may also contain a hanging 5'SS (in this work, the K and L pre-mRNAs), the trans-splicing machinery will perform successive trans-splicing rounds until the pre-mRNA concentration becomes limiting. Thus, with each successive round, the splicing machinery may decrease the amount of free 5'SS.

This raises the question as to which circumstances result in a 5'SS hanging from a pre-mRNA. This is not the normal situation, but the existence of natural trans-splicing implies that trans-donors and trans-acceptors should be obtained as products from an initial trans-splicing event between two cis pre-mRNAs. It should then be possible to obtain a trans-donor (with a hanging 5'SS) that can contribute to successive trans-splicing reactions. One event that may favour a trans-splicing reaction is the nuclear definition of terminal exons. If the first exon cannot be well defined, the second exon may become a trans-acceptor with a hanging 3'SS. The reverse may also occur; if the last exon is not well defined, the penultimate exon may become a trans-donor with a hanging 5'SS. A second event that may also create hanging 5'SS or 3'SS sequences would be co-transcriptional trans-splicing. In that case, two situations can be suggested: (a) a nearly nascent 5'SS may trans-splice with external 3'SS sequences before the next cis-3' exon is synthesized by RNA polymerase II, and (b), premature terminations of RNA polymerase II inside intron DNA sequences may yield molecules with a hanging 5'SS (Dye and Proudfoot, 1999).

Natural trans-exons may interfere with inconvenient pre-mRNAs

As expected, our results show that trans-splicing follows the cis-splicing rules; that is, some 5' donor and 3' acceptor pairs are more easily spliced together than others, and thus trans-splicing could join the most convenient trans-exons, resulting in a modulation of the expression of a message. However, there is a critical aspect that distinguishes trans- from cis-splicing (leaving alternative splicing aside): during cis-splicing it is clear that the 5'SS very often joins with the next downstream 3'SS, probably via the presence of an intron that links the 5' exon with the 3' exon. This situation is quite different in trans-splicing, as there is no intron that links the trans-donor with the trans-acceptor. Trans-molecules should meet, and thus the reaction is highly dependent upon the concentration of trans-molecules. Furthermore, trans-donors/acceptors should choose among several nearby sequences to splice to the strongest pair for this specific trans-donor/acceptor (see Figure 5c, lanes 12-16). This is equivalent to a lock-and-key system. Therefore, if we want to use a trans-RNA to interfere specifically with a target pre-mRNA, it will be necessary to add sequences that recognize specifically this pre-mRNA and allow the trans-splicing with the target exons. This will increase the concentration of trans-donors near the target pre-mRNA and will promote the formation of the selected donor/acceptor pair.

Reprogramming (COT)E2 at its 3' end in such a way that could allow base pairing with the (ras)D2 intron (H pre-mRNA) has demonstrated its ability to compete with the cis-splicing reaction to yield the (COT)E2-(ras)E4 trans-splicing product (Figure 6, lane 2). The (COT)E2 donor without the base-pairing stretch (I pre-mRNA), however, did not show any appreciable amounts of the (COT)E2-(ras)E4 trans-product (compare lane 1 with lane 2 in Figure 6). In conclusion, the discovery that (ras)E4A is capable of being a trans-donor and a trans-acceptor in vitro may allow the modulation of H-ras pre-mRNA cis-splicing by adding external trans-RNAs, such as the reprogrammed (COT)E2 donor, which is specifically designed to target the H-ras D intron. The second product of the trans-splicing reaction that is obtained in addition to (COT)E2-(ras)E4A is a pre-mRNA that contains (ras)E3-(ras)D1-(ras)IDX-(ras)D2 (the D2 intron should contain a branched 'Y' lariat on it, or it could be debranched). Although IDX behaves as a weak exon, this later pre-mRNA could be further cis-spliced to yield (ras)E3-(ras)IDX-(ras)D2 (with 736 nt; only barely detected in this gel; our unpublished results). These results indicate that it is possible (in vitro) to attribute this alternative pathway partially to mRNA that is similar to p19 mRNA (with a stop codon on IDX), via external donor attacks. This trans-splicing interference might yield the same consequences as the in vivo alternative splicing of the D intron - i.e. the activated H-ras gene will be less tumorigenic because mRNA similar to p19 mRNA is being synthesized (Cohen et al., 1989).

Materials and methods

Nuclear extracts and in vitro cis- and trans-splicing reactions

The active splicing extracts from HeLa cells were provided by 4C (Computer Cell Culture Center S.A., Mons, Belgium). The cis- and trans-splicing reactions have been described in detail elsewhere, (Caudevilla et al., 1998; Cicarelli et al., 1998), respectively. All of the trans-splicing reactions described here were carried out with one radio-labelled partner and increasing amounts of the other non-radio-labelled partner (in all Figures, an asterisk beside the pre-mRNA indicates which RNA substrate is radio-labelled).

Pre-mRNAs

All H-ras gene sequences used in these experiments were obtained from the human activated mutant 12 (Cohen et al., 1989), which contains a mutation in codon 12 of (ras)E1. This mutation does not affect the studies presented here, and is thus not localized in the mini-gene constructions described in Figure 2. All of the constructions used in this paper are summarized in Figure 2 as trans-donors, trans-acceptors and cis-splicing pre-RNAs. Since (ras)IDX is a real exon, the D intron was divided into two introns, named D1 and D2 (Cohen et al., 1993).

In this work, a 'donor' is a trans-RNA that contains an exon plus 5'SS sequences, and an 'acceptor' is an exon plus a branch point, a polypyrimidine track and 3'SS sequences (see Figure 2).

The constructs were prepared as follows: (A) was obtained from the rat COT gene, as described elsewhere (Caudevilla et al., 1998); (B) was obtained from the rat COT insert lambdagr5 (Caudevilla et al., 1998) using the Pfu polymerase and 11+12 primers (see below). A lambdagr5 fragment was obtained by polymerase chain reaction (PCR), and subcloned in the EcoRV restriction site of Bluescript SK(+) to obtain the pre-B construct. In order to reduce the polylinker region, the pre-B construct was first digested with SacI and EcoRI, and then incubated with T4 DNA polymerase and religated with T4 DNA ligase. The latter plasmid was linearized with HindIII and transcribed with T3 RNA polymerase. (C) was obtained from the Ad1 plasmid, which was obtained from Adenovirus (Konarska and Sharp, 1987); Ad1 was digested with HindIII, and the fragment was subcloned in HindIII cut Bluescript SK(-). This plasmid was linearized with KpnI and transcribed with T3 RNA polymerase. For (D), the K plasmid (described below) was linearized with NheI and transcribed with T7 RNA polymerase. (E) was obtained as a radio-labelled band (or splicing product) in a electrophoresis gel that was run following an in vitro cis-splicing reaction of pre-mRNA K in HeLa nuclear extract (see Figure 3). For (F), the fragment obtained by PCR from L plasmid (see below) with 7 (see below) and M13 (Stratagene) oligodeoxyribonucleotides and Pfu polymerase, was cut with BamHI and ligated to Bluescript (SK-), which had previously been cut with Ec/136II and BamHI. The plasmid was linearized with BamHI and transcribed with T3 RNA polymerase. (G) was obtained as a radio-labelled band or product of a electrophoresis gel that was run after an in vitro trans-splicing reaction between A´L pre-mRNAs (see Figure 4a). For (H) and (I), the L plasmid was first digested withPstI and BstEII and then incubated with T4 DNA polymerase. The 70-base-pair fragment was purified from the gel and ligated to the plasmid A, which had previously been digested with EcoRI and incubated with Klenow fragment. Both orientations of the 70-bp insert were obtained: one was complementary to the D2 intron (H pre-mRNA) and the other contained the 70-bp insert with the reverse sequence (I pre-mRNA). Both plasmids were linearized with XmaI and transcribed with T3 RNA polymerase. (J) was obtained from the rat COT gene, as described elsewhere (Caudevilla et al., 2001, unpublished results). (K) was obtained from the human H-ras gene as follows: -ile12N and ile12N(G) are the pre-mRNAs wt N4=A and mutant N4=G, respectively, as described in Results. The ile12N or ile12N(G) plasmids, which are described elsewhere (Cohen et al., 1989), were first digested with NcoI and NotI and then incubated with T4 DNA polymerase. This fragment was ligated to Bluescript SK (-) vector, which had previously been digested first with KpnI and SmaI and then incubated with T4 DNA polymerase. The plasmid was linearized with BamHI and transcribed with T7 RNA polymerase. For (l) plasmids ile12N or ile12N)G) were digested first with NheI and NotI and second with T4 DNA polymerase. The fragment was ligated to Bluescript SK (-) vector, which had already been digested with Acc651 and SmaI, and then with Klenow fragment. The plasmid was linearized with BamHI and transcribed with T7 RNA polymerase. For (M) plasmid L was linearized with PstI and transcribed with T7 RNA polymerase. For (N) the avian sarcoma-leukosis virus (ASLV) enhancer was obtained from pSP72-ASLV (env 3' exon) (Tanaka et al., 1994) that had been cut with XbaI and HindIII, and the fragment was subcloned in Bluescript SK (-) plasmid that had itself previously been cut with XbaI and HindIII. This plasmid, named pre-ASLV, was finally cut with XbaI. In order to obtain the fragment containing the (ras)IDX, the K plasmid was digested with NheI and XmaI, incubated with Klenow fragment, and the (ras)IDX containing the fragment was purified from a 1% agarose gel. This latter fragment was subcloned in the pre-ASLV plasmid digested with XbaI. The plasmid was cut with HindIII and transcribed with T3 RNA polymerase. For (O) 2 oligodeoxyribonucleotide primers (9+10, see below) corresponding to the mammalian 5'SS consensus sequence were subcloned in the Bluescript SK (-) plasmid that had been previously cut with NotI and BamHI. This vector was cut with BamHI and then treated with Klenow fragment to obtain pre-O. Plasmid ile12N was digested with NheI and XmaI, and then treated with T4 DNA polymerase, and this fragment was subcloned in pre-O plasmid. The plasmid O was linearized with NotI and transcribed with T7 RNA polymerase.

Oligodeoxyribonucleotides used as primers forreverse-transcriptase-PCR mapping of splicing bands (described in the Figure legends) or plasmid constructions (described above in Materials and methods)

(1) (COT)E2, sense, described elsewhere (Caudevilla et al., 1998); (2) (COT)E2, antisense, described elsewhere (Caudevilla et al., 1998); (3) (Adenovirus)Ad1, sense: TTGCATGCCTGCAGGTCGAG; (4) (ras)IDX, sense: GGCAGCCGCTCTGGCTCTAGCTCC; (5) (ras)E3, sense: ACGCACTGTGGAATCTCGGC; (6) (ras)E4, antisense: AGTCCCCCTCACCTGCGTCA; (7) (ras)E4 sense: GGAGTGGAGGATGCCTTCT; (8) ASLV enhancer, antisense: CCTTCTTGCTTGTTGCTGGCG; (9) Consensus 5' splice site (5'SS), sense: GATCCGGGGTGAGTGGC; (10) Consensus 5'SS, antisense: GGCCGCCACTCACCCCG; (11) (COT)E3, sense: GGCGTACACATCGATTGAAGCCATTTGCAAATGAAGATTTT; (12) (COT)INTRON 3, antisense: AAGTATACAAAATAAATACCCAG.

RNA gels and mapping of RNA bands

RNA denaturing gels were run in an electrophoresis system as described elsewhere (Cicarelli et al., 1998). Gels of between 5 and 8% acrylamide were used - the specific percentages of acrylamide used are indicated in each Figure legend. The RNA bands from the radio-labelled products of the cis- and trans-splicing reactions were characterized by reverse transcriptase (RT)-PCR and DNA sequencing, as described elsewhere (Caudevilla et al., 1998). The molecular weight markers used were pBR322 digested with MspI

Acknowledgements

We thank all members of the group headed by M Bach-Elias for their comments on the manuscript. This work was supported by the Asociación Española contra el Cáncer, La Marató de TV3 and Fundación Ramón Areces. S Guil was a recipient of a BEFI fellowship. We also thank Martí Cullell for revising this manuscript. We thank Dr AD Levinson, Dr Y Shimura and Dr I Mattaj for donating the c-H-ras genes, the ASLV ESE sequence and the adenovirus sequence, respectively.

References

Akopian AN, Okuse K, Souslova V, England S, Ogata N, Wood JN. (1999). FEBS Lett. 445, 177-182. MEDLINE

Barbacid M. (1987). Ann. Rev. Biochem. 56, 779-827. MEDLINE

Breen MA, Ashcroft SJ. (1997). FEBS Lett. 409, 375-379. MEDLINE

Bruzik JP, Maniatis T. (1995). Proc. Natl. Acad. Sci. USA 92, 7056-7059. MEDLINE

Caudevilla C, Serra D, Miliar A, Codony C, Asins G, Bach M, Hegardt FG. (1998). Proc. Natl. Acad. Sci. USA 95, 12185-12190. Article MEDLINE

Chiara MD, Reed R. (1995). Nature 375, 510-513. MEDLINE

Cicarelli RMB, Khaouja A, Codony C, Espuny R, Eritja R, Bach-Elias M. (1998). Biochem. Biophys. Res. Commun. 247, 204-206. MEDLINE

Cohen JB, Broz SD, Levinson AD. (1989). Cell 58, 461-472. MEDLINE

Cohen JB, Broz SD, Levinson AD. (1993). Mol. Cell. Biol. 13, 2666-2676. MEDLINE

Dye MJ, Proudfoot NJ. (1999). Mol. Cell. 3, 371-378. MEDLINE

Eul J, Graessmann M, Graessmann A. (1995). EMBO J. 14, 3226-3225. MEDLINE

Frantz SA, Thiara AS, Lodwick D, Ng LL, Eperon IC, Samani NJ. (1999). Proc. Natl. Acad. Sci. USA 96, 5400-5405. MEDLINE

Huang MY, Cohen JB. (1997). Oncol. Res. 9, 611-621. MEDLINE

Kawasaki T, Okumura S, Kishimoto N, Shimada H, Higo K, Ichikawa N. (1999). Plant J. 18, 625-632. MEDLINE

Konarska MM, Sharp PA. (1987). Cell 49, 763-774. MEDLINE

Li BL, Li XL, Duan ZJ, Lee O, Lin S, Ma ZM, Chang CC, Yang XY, Park JP, Mohandas TK, Noll W, Chan L, Chang TY. (1999). J. Biol. Chem. 274, 11060-11071. Article MEDLINE

Lowy DR, Willumsen BM. (1993). Annu. Rev. Biochem. 62, 851-891. MEDLINE

Mansfield SG, Kole J, Puttaraju M, Yang CC, Garcia-Blanco MA, Cohn JA, Mitchell LG. (2000). Gene Therapy 7, 1885-1895. MEDLINE

Puttaraju M, Jamison SF, Mansfield SG, GarciaBlanco MA, Mitchell LG. (1999). Nat. Biotechnol. 17, 246-252. Article MEDLINE

Shimizu A, Honjo T. (1993). FASEB J. 7, 149-154. MEDLINE

Sullivan PM, Petrusz P, Szpirer C, Joseph DR. (1991). J. Biol. Chem. 266, 143-154. MEDLINE

Tanaka K, Watakabe A, Shimura Y. (1994). Mol. Cell. Biol. 14, 1347-1354. MEDLINE

Vellard M, Sureau A, Soret J, Martinerie C, Perbal B. (1992). Proc. Natl. Acad. Sci. USA 89, 2511-2515. MEDLINE

Zaphiropoulos PG. (1999). Nucleic Acids Res. 27, 2585-2590. MEDLINE

Figures

Figure 1 Alternative splicing of the D intron in c-H-ras pre-mRNA

Figure 2 Pre-mRNAs tested in this work. K and L pre-mRNAs contained identical 5' splice sites (5'SS). The (ras)E4a exon has been abbreviated in all the figures, to rasE4. Constructs K and L are basically equivalent in sequence, except for the addition of (ras)E3 and complete D1 intron in K

Figure 3 In vitro cis-splicing assay of the D intron of H-ras. Lane 1, control without ATP; lane 2, with ATP [K pre-mRNA, wt (5 fmol)]. The two bands of 285 and 390 nucleotides (nt) correspond to the (ras)E3-(ras)E4A and (ras)E3-(ras)IDX-(ras)E4A products, respectively. The lariat product was mapped by reverse transcriptase-polymerase chain reaction (RT-PCR), and it can be observed on the upper part of the gel. The molecular weight markers are shown to the left of the figure. All bands were excised from the gel, eluted and then mapped by RT-PCR (with 4+6 or 5+6 primers) and DNA sequencing. Intron D products, either branched or debranched, were sequenced with intron D specific primers (our unpublished results). A 5% acrylamide gel was used here

Figure 4 (a) The (COT)E2 donor trans-splices in vitro with the H-ras D intron. Radio-labelled A (11.5 fmol) donor was incubated with increasing amount of the wt (N4=A) or mutant (N4=G) forms of the cis-splicing K and L pre-mRNAs. The mutant (N4=G) form has an adenosine-to-guanosine mutation in position 4 of the 5'SS sequence of the D2 intron. Lane 1 is the negative control (without ATP). Lane 2 is a non-related RNA (J in Figure 2, but containing an additional stretch of the polylinker region of Bluescript SK-) that was used as a length marker of 352 nt. Lanes 3-5 are increasing amounts of K (wt, N4=A) pre-mRNA (0.16, 0.31 and 0.62 pmol, respectively); lanes 6-9 are increasing amounts of K mutant (N4=G) pre-mRNA (0.08, 0.16, 0.31 and 0.62 pmol, respectively); lanes 10-13 are increasing amounts of L (wt, N4=A), and lanes 14-17 are increasing amounts of L mutant (N4=G) pre-mRNAS (0.1, 0.2, 0.4 and 0.8 pmol, respectively). The figure shows the bands corresponding to the (COT)E2-(ras)E4A and (COT)E2-(ras)E4A-(ras)E4A products. Products were mapped by RT-PCR (with 1+6 primers) and DNA sequencing. An 8% acrylamide gel was used here. (b) Second trans-splicing round of the (ras)E4A exon. The radio-labelled bands corresponding to the products, and (ras)E3-(ras)E4A, E pre-mRNA, (COT)E2-(ras)E4A and G pre-mRNA (from Figures 3 and 4a, respectively) were cut from the gels, eluted and purified. The purified bands were incubated with non-radio-labelled wt (N4=A) L pre-mRNA or mutant (N4=G) L pre-mRNA. Lanes 1-4 correspond to (COT)E2-(ras)E4A donor, and lanes 5-8 to (ras)E3-(ras)E4A donor. Lanes 1 and 5 are negative controls without ATP. Lanes 2 and 6 are negative controls without acceptor. Lanes 3 and 7 were incubated with wt (N4=A) L pre-mRNA (1.4 pmol), and lanes 4 and 8 with mutant (N4=G) L pre-mRNA (1.4 pmol). The figure shows trans-splicing products with both donors (E and G pre-mRNAS and acceptors - wt (N4=A) and mutant (N4=G). All bands were excised from the gel, eluted and then mapped using RT-PCR (with the same primers as described in the legends of Figures 3 and 4a) and DNA sequencing. Doublet bands, as shown elsewhere (Caudevilla et al., 1998), were mapped by RT-PCR and contained the same sequence, with the exception of some nucleotide degradation at the 3' end. An 8% acrylamide gel was used here. (c) The (ras)E4A donor trans-splices in vitro with the (ras)E4A acceptor. Radio-labelled F donor (11.5 fmol) was incubated with increasing amounts of the wt L pre-mRNA, lanes 3-6 (0.18, 0.36, 0.73 and 1.46 pmol, respectively). Lane 1 is the negative control without ATP, and lane 2 is the negative control without acceptor. All bands were excised from the gel, eluted and then mapped by RT-PCR (with 7+6 primers) and DNA sequencing. A 6% acrylamide gel was used here

Figure 5 (a) An enhancer sequence element (ESE) - but not a 5'SS enhancer - activated the (ras)IDX and flanking sequences to participate as an acceptor in a trans-splicing reaction. Increasing amounts of radio-labelled N (205 nt) and O (210 nt) acceptors were tested against non-radio-labelled C donor. Lane 1 was the negative control without ATP, with 440 fmol of the N acceptor. Lanes 2-5 were increasing amounts of radio-labelled N acceptor (55, 110, 220 and 440 fmol, respectively) with C donor (1.75 pmol). The trans-splicing product had 261 nt, as confirmed by RT-PCR (with 3+8 primers) and DNA sequencing. Lanes 6-9 were a similar experiment with radio-labelled O trans-acceptor (the expected product of the C´O reaction should have 237 nt). An 8% acrylamide gel was used here. (b) The (ras)IDX modified with an ESE sequence is a poor acceptor compared with the natural (COT)E2 acceptor. Radio-labelled A donor (14.4 fmol) was incubated with increasing amounts of non-radio-labelled N acceptor (75, 150, 300 and 600 fmol in lanes 2-5, respectively) or J acceptor (75, 150, 300 and 600 fmol in lanes 6-9, respectively). Lane 1 is the control (without ATP) of the A´N reaction (600 fmol of N). The products were confirmed by RT-PCR (with 1+8 primers in the A´N reaction, and with 1+2 in the A´J reaction) and DNA sequencing. The double bands are due to some nucleotide degradation at the product 3' end (Caudevilla et al., 1998). An 8% acrylamide gel was used here. (c) Competition of the (ras)IDX-ESE and (COT)E2 acceptors with the (ras)E3 donor. (ras)E3 trans-spliced better with the (COT)E2 acceptor than with the (ras)IDX-ESE acceptor. Radio-labelled D donor (10.1 fmol in all lanes) was incubated with increasing amounts of non-radio-labelled N or J acceptors. Lane 1 was the negative control without ATP. Lanes 2-6 were increasing amounts of the N acceptor (0.75, 1.5, 3, 6 and 12 pmol, respectively). Lanes 7-11 were increasing amounts of the J acceptor (0.5, 1.1, 2.2, 4.4 and 8.9 pmol, respectively). Lanes 12-16 were increasing, but always equimolar amounts of both N and J acceptors (each 0.5, 1.1, 2.2, 4.4 and 8.9 pmol, respectively). The bands were confirmed by RT-PCR (with 5+8 primers in the D´N reaction, and with 5+2 primers in the D´J reaction) and DNA sequencing. The double bands are due to some nucleotide degradation at the product 3' end (Caudevilla et al., 1998). An 8% acrylamide gel was used here

Figure 6 A reprogrammed (COT)E2 donor inhibits the D intron cis-splicing of H-ras. The radio-labelled K pre-mRNA (wt, 2.12 fmol) was incubated with 0.55 pmol of non-radio-labelled I and H donors (lanes 1 and 2, respectively). Lane 4 was the positive control with no added trans-donor. The Figure shows the (ras)E3-(ras)E4A cis-splicing product in lanes 1, 3 and 4, (ras)E3-(ras)IDX-(ras)E4A product in lane 4, and the (COT)E2-(ras)E4A trans-splicing product in lane 2. Lane 3 was the negative control without ATP and added donors. The bands were cut from the gel and mapped by RT-PCR, with 5+6 primers for the (ras)E3-(ras)IDX-(ras)E4A cis-splicing products and with 1+6 primers for the (COT)E2-(ras)E4A trans-spliced product, and DNA sequencing. A 5% acrylamide gel was used here. The autoradiography was scanned with a Molecular Dynamics laser densitometer using the integration function in the ImageQuant programme to determine peak areas. The integrated areas of the bands corresponding to the products were used to calculate the percentage of inhibition (described in Results) compared to the product obtained in lane 4. Since the amount of the pre-mRNA input may fluctuate, the amount of pre-mRNA K added was taken into account when the percentage inhibition was calculated

Received 4 January 2001; revised 27 February 2001; accepted 15 March 2001
21 June 2001, Volume 20, Number 28, Pages 3683-3694
Table of contents    Previous  Article  Next    [PDF]