A. The Discovery of Trans -splicing

Trans-splicing, in which an identical short leader sequence, the spliced leader (SL), is spliced onto the 5′ends of multiple mRNAs (for reviews, see Agabian 1990; Donelson and Zeng 1990; Bonen 1993), was first discovered in the primitive eukaryotes, the trypanosomatids (Murphy et al. 1986; Sutton and Boothroyd 1986), and later shown to occur also in C. elegans and other nematodes (Krause and Hirsh 1987; for reviews, see Nilsen 1993; Davis 1996), in Euglena (Tessier et al. 1991), and in flatworms (Rajkovic et al. 1990; Davis et al. 1994). In trypanosomes, all splicing is trans-splicing; all mRNAs begin with the SL, and genes do not contain introns. Transcription is polycistronic, and trans-splicing is responsible for separating the long polycistronic transcripts into monocistronic units. In contrast, in nematodes, the genes do contain introns, and the pre-mRNA products of many genes are not subject to trans-splicing.

Trans-splicing in C. elegans was first found during molecular analysis of the actin genes (Krause and Hirsh 1987). It was discovered that mRNAs of three of the four actin genes begin with the identical 22-nucleotide sequence, a sequence that is not associated with the gene. Instead, the 22-nucleotide SL is donated by a 100-nucleotide small RNA, SL RNA, by a trans-splicing reaction. This trans-splicing process is closely related to cis-splicing (intron removal). A reasonable match to the 5′splice site consensus is present on the SL RNA, and a good match to the 3′splice site consensus is present at the site of SL addition (trans-splice site) on the pre-mRNA. Furthermore, the reaction proceeds by way of a branched intermediate similar to the lariat of cis-splicing (Bektesh and Hirsh 1988; Thomas et al. 1988; Hannon et al. 1990a). Trans-splicing also requires spliceosomal components including U2, U4, U5, and U6 snRNPs (Hannon et al. 1991; Maroney et al. 1996; see below).

B. The Spliced Leader snRNP

The donor in the trans-splicing reaction, the SL RNA, exists in the form of an snRNP (Bruzik et al. 1988; Thomas et al. 1988; Van Doren and Hirsh 1988). It is bound to the Sm proteins found associated with U1, U2, U4, and U5 RNAs, and it has the unusual modified cap structure, trimethylguanosine (TMG), found on these snRNAs. The secondary structure predicted for the SL RNA resembles that of other snRNAs. It has the 5′splice site base-paired to the upstream part of the SL sequence in a manner resembling the U1-5′splice site base pairing. It was hypothesized that this intramolecular interaction might replace in trans-splicing the interaction between U1 and the 5′splice site required for initiation of cis-splicing (Bruzik et al. 1988). Subsequent work has demonstrated that the U1 snRNP is indeed dispensable for nematode trans-splicing in vitro (see below).

C. Trans -splicing Signals

The trans-splice site consensus is the same as the intron 3′splice site consensus (Table 2), so it is not immediately obvious how the two reactions can be faithfully carried out. However, it is now clear that the signal for trans-splicing is simply the presence of an intron-like sequence at the 5′end of the mRNA with no functional 5′splice site upstream (Conrad et al. 1991, 1993b, 1995). The region of the pre-mRNA from the 5′end to the trans-splice site is called the outron (Conrad et al. 1991). Genes whose pre-mRNAs are subject to trans-splicing are distinguished from those that are not only by the presence of an outron. Considerable experimental evidence has been adduced to support this view. A conventional gene can be converted into a trans-spliced gene by placing at the gene's 5′end either an intron from another gene or an A + T-rich synthetic sequence followed by a canonical 3′splice site. Furthermore, a trans-spliced gene can be converted into a conventional gene by inserting a 5′splice site into its outron. These experiments demonstrate that the only difference between trans-spliced and conventional genes is the presence of an outron, and they show that no sequence-specific recognition is involved in the decision to trans-splice. They also show that the intron and outron 3′splice sites are the same; the choice between cis- and trans-splicing is based solely on the presence or absence of an upstream 5′splice site.

Table 2

Comparison of cis- and trans-3′splice sites.

Because trans-splicing is a relatively efficient reaction (like cis-splicing), it is generally impossible to isolate outron-containing pre-mRNAs, and so very few natural outrons have been defined. Nevertheless, in a few cases, the promoters of trans-spliced genes or start sites of outrons have been identified (e.g., col-13 has a 64-bp outron [Park and Kramer 1990]). It might be possible to determine trans-spliced gene start sites by deleting the trans-splice site and analyzing the RNA product from a transgenic strain carrying the mutated gene. However, in the few cases in which this technique has been attempted, it has been unsuccessful because trans-splicing occurred at an alternative site. One proven successful technique is to introduce a 5′splice site consensus sequence into the outron. In this case, the introduced splice site splices to the trans-splice site, and the outron length can be calculated from the length of the 5′-untranslated region (5′UTR). This procedure was used to determine outron length (173 bp) for rol-6 (Conrad et al. 1993). Although few natural outrons have been characterized, synthetic sequences have been introduced into the 5′UTR of a gene that is not normally trans-spliced in order to determine whether they can function as outrons. A + U-rich sequences of 51 nucleotides or longer resulted in efficient trans-splicing, whereas sequences that were 41 nucleotides or shorter (or not A + U-rich) were much less effective (Conrad et al. 1995). Thus, the same constraints that set the lower size limit on introns may be acting on outrons.

D. Function of Trans -splicing

Trans-splicing occurs throughout the nematode phylum, and there is a remarkable degree of conservation of the SL sequence (although the portions of the SL RNAs downstream from the splice site have diverged widely) (Bektesh and Hirsh 1988; Tackacks et al. 1988; Nilsen et al. 1989; Zeng et al. 1990). In the many free-living species, as well as animal and plant parasitic nematodes, that have been examined, only one single-base change has been found in the SL sequence (Ray et al. 1994). It is not known what selection pressure has kept this sequence so stable. In fact, it is not yet known precisely what function the SL has in the cell. In C. elegans, SL tends to be spliced very close to the initiating methionine codon (often immediately adjacent) (Fig. 5), so it seems likely to play a part in translation initiation. The unusual cap structure, trimethylguanosine (TMG), present at the 5′end of the SL becomes the 5′end of trans-spliced mRNAs, and this cap remains on the mRNA during translation (Liou and Blumenthal 1990; Van Doren and Hirsh 1990). A TMG cap is known to inhibit translation in mammalian extracts (Darzynkiewicz et al. 1988), but it may actually stimulate translation in a nematode extract, at least when it is present at the 5′end of the SL sequence (Maroney et al. 1995).

Figure 5

Distance from the trans-splice site to the translation initiation site. The survey includes 83 trans-spliced genes, both SL1- and SL2-acceptors. Each bar represents the number of (more...)

In Ascaris lumbricoides, an animal parasite, the SL sequence in the DNA is needed for transcription of the SL RNA gene, which may be one reason why it has been so highly conserved (Hannon et al. 1990b). Although it is not known precisely what roles the SL sequence itself may perform, trans-splicing is in fact required for viability (Ferguson et al. 1996). An embryonic lethal mutation in the rrs-1 gene is a deletion of all 100 tandem copies of the 1-kb sequence that encodes both 5S ribosomal RNA and SL RNA (see below). Remarkably, the embryonic lethality is rescued by a tandem array carrying the SL RNA gene alone (presumably the maternal supply of 5S RNA can carry the homozygous mutants through embryogenesis). Mutations in the SL RNA gene that eliminate the Sm-binding site prevent rescue, so it is fair to conclude that the SL snRNP is required for embryogenesis. Its required role could be a positive effect such as providing a sequence needed for translation initiation, mRNA stability, or localization, or it could be required for suppression of a negative effect such as inhibition of translation initiation by AUG codons in the outron. At least some of the mRNAs that normally are trans-spliced to SL1 have been found to be trans-spliced to the alternative spliced leader, SL2 (see below), in the rrs-1 mutant strain (Ferguson et al. 1996).