Inverse splicing of a group II intron ( ribozyme / drcular RNA / exon scrambling / exon shuflting ) KEVIN

I describe the self-splicing of an RNA that consists of exon sequences flanked by group II intron sequences. I find that this RNA undergoes accurate splicing in vitro, yielding an excised exon circle. This splicing reaction involves the joining of the 5' splice site at the end of an exon to the 3' splice site at the b nning of the same exon; thus, I term it inverse splicing. Inverse splicing provides a potential mechanism for exon scrambling, for exon deletion in alternative splicing pathways, and for exon shuffling in gene evolution. The removal of intron sequences from mRNA precursors by splicing is accomplished with high fidelity. However, exon scrambling, the joining of exons in an order different from that in the primary transcript, can occur at a low frequency (1, 2). Such scrambling could occur by an inverse splicing mechanism in which the 5' splice site of an exon is joined to the 3' splice site of an upstream exon. Inverse splicing could also delete specific exons from transcripts, providing a mechanism of alternative splicing. This reaction would involve joining the 5' splice site at the end of an exon to the 3' splice site at the beginning of the same exon, resulting in the excision of the exon as a circular RNA. Self-splicing group II introns, typically found in organellar RNA precursors, are spliced by a two-step mechanism analogous to that used in nuclearpre-mRNAs (3, 4). It is likely that the catalytic core of the group II intron is structurally similar to the core of the spliceosome (5-7). Group II introns are capable ofboth cis and trans splicing in vivo and in vitro (3, 4, 8-12), and distinct structural domains located at the 5' and 3' ends of the intron are required for splicing (13). In this report, I show that the splice sites flanking a single group II exon can be joined by splicing, to yield an exon circle. That efficient inverse splicing can occur on an appropriate substrate reveals that accurate splicing of normal RNA precursors involves mechanisms that prevent inverse splicing. In addition, the demonstration of inverse splicing has interesting implications for the exon-shuffling hypothesis of gene evolution. MATERIALS AND METHODS The RNA transcripts diagramed in the first three lines of Fig. 1 were transcribed from the plasmids pJD1, pJDI3'-673, and pJDI5'-75 (9), respectively. The (IVS 5,6)E3,E5(IVS 1-3) RNA, shown in the bottom line of Fig. 1, was synthesized from plasmid pINV1. To construct pINV1, the Sac I-HindIII fragnent of pJDI5'-75 was isolated and the HindIIl site was filled in with Klenow fragment. This DNA was ligated to pJDI3'-673 that had been cleaved with Sac I and Sma I. The RNA splicing substrates were made by in vitro transcription using T7 RNA polymerase (9). Transcription, RNA purification, and splicing reactions were as described (9). The E5-specific oligodeoxynucleotide (5'-GTAGGATTAGATGCAGATACTAGAGC-3') is identiThe publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. cal to 26 nucleotides (nt) of the E5 region of the (IVS 5,6)E3,E5(IVS 1-3) RNA. The E3-specific oligonucleotide (5'-GAGGACTTCAATAGTAGTATCCTGC-3') is homologous to 25 nt of the E3 region. To purify E3,E5(C) for the reverse transcription reaction, a standard 1004 transcription reaction was done (9), with pINV1 as a template. The (IVS 5,6)E3,E5(IVS 1-3) RNA was concentrated by ethanol precipitation and was then incubated under the (NH4)2SO4 splicing conditions for 1 hr. The E3,E5(C) RNA was gel purified and dissolved in 30 A4 of water. A 9-id annealing reaction mixture was incubated at 65°C for 3 min and then placed on ice. The annealing reaction mixture included 1 p4 ofthe E3,E5(C) RNA plus 100 ng ofthe E3-specific oligonucleotide. As a control, an identical annealing reaction was done, except E3,E5(C) was not added. A buffer (4 s.d) consisting of 0.25 M Tris HCl (pH 8.5), 0.25 M KCl, 0.05 M dithiothreitol, and 0.05 M MgCl2 was added to both annealing reaction mixtures. Deoxynucleoside triphosphates were each added to a fmal concentration of5 mM, followed by 40 units of RNasin (Promega) and 22 units of reverse transcriptase (Seikagaku America, Rockville, MD). The final volume was adjusted to 20 1d with water. The mixture was incubated at 42°C for 90 min. Two polymerase chain reaction (PCR) experiments were done using as templates either 1 p1 of the reverse transcription mixture that included E3,E5(C) or 1 p1 of the control reverse transcription mixture, which lacked E3,E5(C). The PCRs were performed as described (14) and were continued for 25 cycles. The E3and E5-specific oligonucleotides, 300 ng each, were used as PCR primers. DNA sequencing was done with Sequenase (United States Biochemical) according to the protocol provided by the manufacturer. RESULTS AND DISCUSSION The substrate used to test the possibility of inverse splicing is diagramed in Fig. 1 (bottom line). This RNA consists oftwo RNAs that were shown to trans splice (middle lines), joined in an inverse orientation to yield a circularly permuted version of the wild-type precursor. Group II intron excision can occur by transesterification (splicing) or by site-specific hydrolysis (cleavage). The former reaction is stimulated by (NH4)2SO4, the latter by KCl (15). Upon coincubation in the presence of (NH4)2SO4, the control RNAs [E5(IVS 1-3) plus (IVS 5,6)E3] trans spliced to yield spliced exons (E5-E3) and a Y-branched intron [IVS(Y)] (Fig. 2A, lane 2) as previously reported (9). Coincubation in the presence of KCI yielded free exons (E5 and E3) and a linear intron (IVS 1-3) as major products (lane 3). The products obtained in the presence of each salt have been extensively characterized (9, 15). The (IVS 5,6)E3,E5(IVS 1-3) precursor was also reactive. Most of the products could be identified based on their comigration with products of the control trans reaction. In the presence of (NH4)2SO4 (lanes 4-10), IVS(Y) and some linear intron were liberated; several novel products were also generated. Among these was an RNA (E3,E5) the expected