Mobility of the Sinorhizobium meliloti Group II Intron RmInt1 Occurs by Reverse Splicing into DNA, But Requires an Unknown Reverse Transcriptase Priming Mechanism

doi:10.1016/S0022-2836(03)00208-0

J. Mol. Biol. (2003) 327, 931–943

Mobility of the Sinorhizobium meliloti Group II Intron RmInt1 Occurs by Reverse Splicing into DNA, But Requires an Unknown Reverse Transcriptase Priming Mechanism Estefanı´a Mun˜oz-Adelantado1, Joseph San Filippo2 Francisco Martı´nez-Abarca1, Fernando M. Garcı´a-Rodrı´guez1 Alan M. Lambowitz2 and Nicola´s Toro1* 1

Grupo de Ecologı´a Gene´tica Estacio´n Experimental del Zaidı´n, Consejo Superior de Investigaciones Cientı´ficas Calle Profesor Albareda 1 18008 Granada, Spain 2

Institute for Cellular and Molecular Biology, Department of Chemistry and Biochemistry and Section of Molecular Genetics and Microbiology School of Biological Sciences University of Texas at Austin Austin, TX 78712, USA

The mobile group II introns characterized to date encode ribonucleoprotein complexes that promote mobility by a major retrohoming mechanism in which the intron RNA reverse splices directly into the sense strand of a double-stranded DNA target site, while the intronencoded reverse transcriptase/maturase cleaves the antisense strand and uses it as primer for reverse transcription of the inserted intron RNA. Here, we show that the Sinorhizobium meliloti group II intron RmInt1, which encodes a protein lacking a DNA endonuclease domain, similarly uses both the intron RNA and an intron-encoded protein with reverse transcriptase and maturase activities for mobility. However, while RmInt1 reverse splices into both single-stranded and double-stranded DNA target sites, it is unable to carry out site-specific antisense-strand cleavage due to the lack of a DNA endonuclease domain. Our results suggest that RmInt1 mobility involves reverse splicing into doublestranded or single-stranded DNA target sites, but due to the lack of DNA endonuclease function, it requires an alternate means of procuring a primer for target DNA-primed reverse transcription. q 2003 Elsevier Science Ltd. All rights reserved

*Corresponding author

Keywords: group II intron; maturase; reverse transcriptase; RNA; ribozyme

Introduction Group II introns are catalytic RNAs and mobile genetic elements found in bacteria and organelles. The mobility mechanism of these site-specific retroelements has been elucidated in studies of the yeast mtDNA aI1 and aI2 introns, and the Lactococcus lactis Ll.ltrB intron.1,2 These mobile introns encode proteins having reverse transcriptase (RT), RNA splicing (maturase), and DNA endonuclease activities. Mobility into an intronless allele (retrohoming) occurs by a target DNAPresent address: F. M. Garcı´a-Rodrı´guez, Institute of Molecular Plant Sciences, Clusius Laboratory, University of Leiden, Wassenaarseweg 64, 2333 AL Leiden, The Netherlands. Abbreviations used: IEP, intron-encoded protein; RNP, ribonucleoprotein; RT, reverse transcriptase; TPRT, target DNA-primed reverse transcription. E-mail address of the corresponding author: [email protected]

primed reverse transcription (TPRT) mechanism mediated by a ribonucleoprotein (RNP) complex containing the intron-encoded RT and the excised intron lariat RNA. The intron RNA reverse splices directly into the sense strand of a double-stranded DNA target site, while the intron-encoded protein (IEP) site-specifically cleaves the antisense strand and then uses the cleaved 30 end to prime reverse transcription of the inserted intron RNA. Group II introns also transpose, with low frequency, to ectopic sites that resemble the homing site, primarily by a DNA-target pathway similar to retrohoming.3 – 8 The proteins encoded by the yeast mtDNA and L. lactis Ll.ltrB introns consist of an N-terminal RT domain, which is followed by domain X, associated with maturase activity, and then a C-terminal DNA-binding region and conserved DNA endonuclease domain.3,9 – 11 The latter, sometimes referred to as the Zn domain, contains two pairs of conserved cysteine residues, similar to a

0022-2836/03/$ - see front matter q 2003 Elsevier Science Ltd. All rights reserved

932

zinc-finger motif, interspersed with amino acid sequences characteristic of the H –N – H family of endonucleases.12,13 Recent studies suggest that the first step in intron mobility is the recognition of a small number of specific nucleotide residues in the distal 50 -exon region of the DNA target site by the IEP.11 This recognition event involves major groove and phosphate-backbone interactions promoting local DNA unwinding, making it possible for the intron RNA to base-pair to the intron-binding sequence (IBS) and d0 sequences in the sensestrand of the DNA target site for reverse splicing. Antisense-strand cleavage occurs after a lag and requires additional interactions between the IEP and the 30 exon. The antisense strand is cleaved by the IEP in the 30 exon, 9 nt or 10 nt downstream from the intron-insertion site, depending on the endonuclease, and the 30 end of the cleaved antisense strand is then used as a primer for reverse transcription of the inserted intron RNA. The resulting cDNA copy of the intron is integrated into the host DNA either by repair in bacteria, or by repair and recombination mechanisms in yeast mitochondria.14 – 16 We have shown that the Sinorhizobium meliloti RmInt1 intron, which encodes an IEP with no recognizable DNA endonuclease domain, is an efficient mobile element, with a homing frequency approaching 100%, similar to that displayed by the fungal mtDNA introns and the lactococcal Ll.ltrB intron.17 – 19 The absence of a DNA endonuclease domain is relatively common among bacterial group II IEPs identified to date20 – 22 and is characteristic of a recently reported archaeal group II intron.23,24 It has been suggested that the DNA endonuclease domain may have been acquired later in evolution by a primordial bacterial group II intron lacking this domain.2,20,21 Furthermore, RmInt1 belongs to the IIB subclass of group II introns, whereas the lactococcal intron Ll.ltrB and the yeast mtDNA aI1 and aI2 introns belong to the IIA subclass.19 – 21,24 – 26 Thus, features of the retrohoming mechanism used by these three introns may differ for other group II introns, particularly those lacking a DNA endonuclease domain. The aim of this work was to gain insight into the RmInt1 mobility mechanism. Our results show that RmInt1 IEP has RT and maturase activity and that RmInt1 RNP particles catalyze the cleavage of both single-stranded and double-stranded DNA substrates by reverse splicing at the introninsertion site. However, the RmInt1 RNP particles are unable to carry out site-specific antisensestrand cleavage due to lack of the DNA endonuclease domain. Our results suggest that RmInt1 mobility is similar to that of the yeast mtDNA and L: lactis Ll.ltrB introns in involving reverse splicing into DNA target sites, but differs in requiring an alternate means of generating or accessing a primer for reverse transcription of the inserted intron RNA. This study constitutes the first biochemical analysis of the mobility reactions of a

S. meliloti Group II Intron Mobility

group II intron encoding a protein lacking a DNA endonuclease domain.

Results RmInt1 mobility depends on both the excised intron RNA and the IEP For the yeast aI1 and aI2 introns and the L. lactis Ll.ltrB intron, the C-terminal region of the IEP consists of an upstream variable region that contributes to DNA-binding, followed by a conserved DNA endonuclease domain (Figure 1).3,10,11 Sequence alignments show that the RmInt1 IEP lacks the DNA endonuclease domain20,21 but some alignments suggest that the RmInt1 IEP may have a C-terminal extension beyond domain X of 20 amino acid residues.27

Figure 1. RmInt1 intron and mutant derivatives. A scaled representation of the LtrA protein encoded by the L. lactis Ll.ltrB intron is shown. The open reading frame (ORF) is within intron domain IV. Protein domains are: reverse transcriptase, RT; maturase, X; variable DNA-binding region, D; and conserved DNA endonuclease domain, En. The putative C-terminal extension of 20 amino acid residues beyond domain X in the RmInt1 IEP is indicated with an asterisk (p ). Numbers in brackets indicate the nucleotide position of the stop codon from the intron 50 end.

933

S. meliloti Group II Intron Mobility

We previously reported that the RmInt1 intron was correctly spliced in Escherichia coli and S. meliloti, and that this process requires the IEP.17,18 Here, we characterized the splicing of wild-type and mutant RmInt1 introns in vivo by primer extension analysis of whole-cell RNA using a primer complementary to a sequence 97 nt

from the 50 end of the intron (Figure 2(a)). With RNA isolated from S. meliloti strain RMO17 expressing the wild-type intron pKG2.5, the primer extension yields a cDNA product of 97 nt, corresponding to the excised intron RNA, indicative of splicing, along with larger bands presumably derived from unspliced precursor RNA (Figure 2(a)). The same 97 nt band with somewhat reduced intensity was observed with RNA from cells expressing pKG2.5-YAHH, which has a YADD to YAHH mutation in the RT domain, consistent with results for the yeast mtDNA introns aI1 and aI2 showing that RT activity is not required for splicing.14,28 However, the 97 nt extension product was not detected with RNA isolated from cells harboring pKG2.5X, which encodes an IEP truncated in the RT domain; pKG2.5-DC29, which has a Cterminal truncation of 29 amino acid residues; or pKG2.5D5-CGA, which has a mutation in the critical conserved pairing AGC-GUU of intron RNA domain V (GUU ! CGA). The same mutations affecting the RmInt1 IEP (pKG2.5X, pKG2.5YAHH, pKG2.5-DC29) or the intron RNA (pKG2.5D5-CGA) also block intron mobility (Figure 2(b)). Thus, we conclude that both the intron RNA and IEP are required for intron RNA splicing, which is necessary for mobility. The RmInt1 IEP has RT activity with poly(rA)/ oligo(dT)18

Figure 2. In vivo splicing and homing of wild-type and mutant RmInt1. (a) RNA splicing. Primer extension analysis was performed on total RNA extracted from S. meliloti RMO17 strain harboring pKG2.5D5-CGA (lane 1), pKG2.5-YAHH (lane 2), pKG2.5X (lane 3), pKG2.5 (lane 4) and pKG2.5-DC29 (lane 5). Total RNA was reverse transcribed with an intron-specific primer (P). The major cDNA product (97 nt) corresponds to the excised intron RNA. Larger products presumably derived from unspliced precursor RNAs were also detected (112, 126 and 327 nt). A representation of the primer extension assay is shown. Broken lines indicate major cDNA extension products. (b) Homing assays. Plasmid pools were analyzed by Southern hybridization with a DNA target probe. RMO17 containing pJB0.6þ recipient DNA was transformed with a compatible plasmid containing the wild-type RmInt1 (lane 1) and the derived mutant constructs indicated (lanes 2 – 5). Plasmid minipreps digested with Sal I show several DNA-hybridizing fragments: 7.6 kb for intron donor plasmid and 1.85 kb for recipient plasmid. Intron invasion of pJB0.6 þ results in a new 3.73 kb hybridizing fragment of DNA (homing product). Molecular mass markers (lambda phage DNA digested with HindIII) are indicated (M).

We characterized the RmInt1 RT activity by studying RNP particles isolated from S. meliloti strain RMO17 harboring the wild-type intron construct pKG2.5. When assayed with the exogenous substrate poly(rA)/oligo(dT)18, the RNP particles had high RT activity, which was dependent on the oligo(dT)18 primer, whereas no significant activity was detected with RNP particles from S. meliloti cells lacking the intron-containing plasmid (Table 1). In contrast to findings for the yeast aI2 intron,29 but as reported for the lactococcal Ll.ltrB intron,30 the RT activity of the RmInt1 RNP particle preparations was not increased in the presence of RNase A (data not shown), suggesting that the RmInt1 IEP can switch to an exogenous substrate. Table 1. RT assays with exogenous templates Strain RMO17 RMO17pKG2.5 RMO17pKG2.5X RMO17pKG2.5-YAHH RMO17pKG2.5D5-CGA RMO17pKG2.5-DC29 RMO17pKG-IEP

poly(rA)/oligo(dT)18

poly(rA)

2353 ^ 256 197,764 ^ 2164 1415 ^ 77 2315 ^ 361 11,167 ^ 479 2130 ^ 278 1649 ^ 356

2372 ^ 622 1463 ^ 52 1036 ^ 93 1492 ^ 96 1067 ^ 231 984 ^ 138 1232 ^ 135

RT assays were carried out by incubating RNP particles with the exogenous substrates poly(rA)/oligo(dT)18 or poly(rA) as described in Experimental Procedures. cDNA synthesis was quantified by measuring incorporation of [a-32P]dTTP into high molecular weight material retained on DEAE-paper. The values (cpm) are means ^ S.D.s for three determinations.

934

S. meliloti Group II Intron Mobility

As expected, the RT activity of RmInt1 was abolished by truncation of the IEP in the RT domain (pKG2.5X) or by mutation of the conserved YADD motif in the RT domain (pKG2.5-YAHH). However, other protein or intron RNA mutations that inhibited splicing led to strongly decreased RT activity in RNP particles, including the C-terminal truncation pKG2.5-DC29 and the mutation in the AGC-GUU pairing of intron domain V (pKG2.5D5-CGA). Furthermore, RNP particles from cells that produced the RmInt1 IEP in the absence of the intron RNA (pKG-IEP) lacked RT activity (Table 1). These findings demonstrate that the RmInt1 IEP has RT activity and suggest that either its recovery in RNP particle preparations is dependent on the functional binding of the IEP to the intron RNA or that spliced intron RNA is required to stabilize the RT activity.

Endogenous RT activity of RmInt1-derived RNP particles The RT activity of the yeast aI2 and lactococcal Ll.ltrB IEPs bound to endogenous RNA templates in RNP particles is stimulated by adding DNA primers, particularly 30 -exon primers that mimic the 30 end of the cleaved antisense strand normally used as the primer for TPRT.29 – 31 On the basis of these findings, it has been suggested that the RTs of aI2 and Ll.ltrB are bound to unspliced precursor RNA in a position to initiate cDNA synthesis in the 30 exon. The RmInt1 RNP particle preparations have low levels of RT activity with endogenous substrates (Figure 3). The addition of 30 -exon primers somewhat increased the RT activity with a maximum twofold increase for primers þ 31 and þ 52, much less than the factors of 30 and 10 observed for the yeast and lactococcal introns, with primers ending at positions þ 9 to þ 12, and þ 6 to þ 14, respectively.29,31 By contrast, a primer (Int2) complementary to a sequence in intron domain IV (intron position 1598 –1617) increased the level of incorporation by a factor of 8. Avian myeloblastosis virus (AMV) RT control reactions showed that the primers were accessible to their complementary sequences in the endogenous RNAs (Figure 3). The RmInt1 RNP particle preparations contain 2.5-fold more excised intron RNA than unspliced precursors as estimated by primer extension analyses (data not shown). Thus, the eightfold stimulatory effect of the Int2 primer on endogenous RmInt1 RT activity may be related to the position at which the IEP is bound to the excised intron lariat, while the smaller stimulatory effect with 30 -exon primers could reflect RT bound to unspliced precursor RNA. The finding that the stimulatory effect with 30 -exon primers is greatest for þ 52 rather than þ 9 to þ 12, as for the Ll.ltrB intron, could reflect that the RmInt1 RT is bound to the precursor RNA in a way that enables it to initiate at a wider range of distances farther downstream from the intron.

Figure 3. Endogenous RT activity in RNP particle preparations from S. meliloti. Reverse transcription was carried out with RmInt1 RNP particles in the absence or in the presence of various primers complementary to the RmInt1 intron RNA or different positions in the 30 exon of unspliced precursor RNA. Filled bars indicate the amount of [a-32P]dCTP incorporated in 20 minutes. Controls (open bars) show incorporation with same primer-template combinations after addition of five units of AMV-RT. Data are means ^ standard deviations (indicated as error bars) for three determinations. The schematic at the bottom shows the positions of the different oligonucleotides.

Primer-mediated initiation of DNA synthesis by the RmInt1 RT The yeast aI1 and aI2 and the lactococcal RTs use the 30 end of the cleaved antisense strand as the primer for TPRT. We studied the initiation of DNA synthesis by the RmInt1 RT by carrying out DNA polymerase reactions with an exogenous 60mer DNA template containing the RmInt1 target site (position 2 30 to þ 30) annealed to a complementary 20 nt primer (þ 9; Figure 4(b)). Reactions were carried out in the presence or in the absence of dideoxynucleotides (ddNTPs). If the reaction mixture included only dNTPs, the RmInt1 RNP particles (pKG2.5) extended the þ 9 primer to the end of the 60-mer DNA template, giving a 58 nt product (Figure 4(a), lane 1). Other experiments showed that the size of the product varied as expected with the position of the primer, confirming that the 60-mer DNA was the template (not shown). If ddNTPs were included in the reaction mixture (Figure 4(a), lanes 2– 4), the þ 9 primer was extended to the site of incorporation of the corresponding ddNTP, resulting in DNA products of 22 nt for ddGTP, 23 nt for ddTTP and 26 nt for ddATP (Figure 4(b)). When RT mutant

S. meliloti Group II Intron Mobility

935

pKG2.5-YAHH RNP particles were used for the primer-mediated initiation of DNA synthesis, only small labeled products were observed, and these were not terminated at the expected position by ddNTPs, suggesting they result from other activities in the RNP preparations. As expected, in the absence of both the 60-mer DNA template and the þ 9 primer, no labeled products were observed with RNP particle preparations derived from cells containing pKG2.5 or the RT mutant pKG2.5YAHH (Figure 4(a), lanes 11 and 12). In the absence of the þ 9 DNA primer, RNP particles from cells containing pKG2.5, but not those from cells expressing the RT mutant, carried out residual DNA synthesis, which may reflect incorporation using endogenous DNA template or primers (Figure 4(a), lane 9). Together, these results provide further evidence that the RmInt1 IEP can switch from endogenous to exogenous templates and show that it can function as a DNA polymerase, as well as a reverse transcriptase.

Sense-strand cleavage by the RmInt1 intron on double-stranded DNA

Figure 4. DNA synthesis by the RmInt1 RT. (a) DNA polymerase assays. Reactions were carried out with RNP particles from S. meliloti RMO17 harboring pKG2.5 or the RT mutant pKG2.5-YAHH, in the presence or absence of a 60-mer DNA template containing the intron insertion site and a complementary 20-mer primer, whose 30 end is at position þ 9 of the target sequence. Labeled products were analyzed by electrophoresis in a denaturing, 8% (w/v) polyacrylamide gel. Molecular size markers are indicated (in nucleotides) at the left of the panel. An enlargement of the bottom of the gel corresponding to lanes 1 – 8 is shown. (b) A diagram of the 60-mer DNA template, þ9 primer and the synthesized products showing the expected products of DNA synthesis and positions corresponding to the addition of the ddNTPs. The arrow indicates the RmInt1 intron-insertion site. It should be noted that under these experimental conditions, and since the DNA target is unlabeled, reverse splicing cannot be detected.

To assess the DNA endonuclease activity of the RmInt1 intron, RNP preparations from S. meliloti RMO17, expressing the wild-type intron pKG2.5, were incubated with three labeled double-stranded DNA substrates of 100 bp, 148 bp and 196 bp containing the sequence between positions 2 50 to þ 50, 2 50 to þ 98 and 2 50 to þ 146 from the intron-insertion site (Figure 5(e)). The doublestranded DNA substrates were generated by PCR with 50 -end labeled sense-strand primers. After the incubation, the products were analyzed by electrophoresis in a denaturing, 6% (w/v) polyacrylamide gel to detect the 50 fragment of the sense strand. The RNPs from cells expressing pKG2.5 cleaved the sense-strand of the three double-stranded DNA substrates at the intron-insertion site, whereas no such cleavage was detected with RNP particles from cells expressing an RmInt1 intron with an IEP truncated in the RT domain (pKG2.5X) or in control lanes without RNPs (Figure 5(a)). The efficiency of cleavage appears lower with longer double-stranded DNA substrates, suggesting that the RmInt1 RNPs may have some difficulty carrying out DNA unwinding, and this difference was observed in three independent experiments (Figure 5(a) and data not shown). Double-stranded DNA substrates were also generated by PCR with 50 -end labeled primers to label separately each of the two strands. Even though the RNP preparations from S. meliloti RMO17 and GR4, and S. medicae RMS16, expressing pKG2.5 cleaved the sense strand of a 50 -end labeled double-stranded DNA substrate of 100 bp (positions 2 50 to þ 50 from the intron-insertion site), none of them detectably cleaved the antisense strand (Figure 5(b) and (c)).

936

S. meliloti Group II Intron Mobility

Figure 5. DNA reverse splicing and endonuclease assays with double-stranded DNA substrates. (a) Endonuclease assays with 50 -sense-strand labeled substrates of 100 bp (lanes 1 – 3), 148 bp (lanes 4 –6) and 196 bp (lane 7 – 9), position 2 50 to þ50, 2 50 to þ 98 and 2 50 to þ146, from the intron insertion site, respectively. RNP particles from S. meliloti RMO17 expressing wild-type intron pKG2.5, or the mutant pKG2.5X were incubated with the labeled DNA substrates, and the products were analyzed by electrophoresis in a denaturing, 6% (w/v) polyacrylamide gel. Control lanes without RNPs are shown in lanes 3, 6 and 9. (b) and (c) Endonuclease assays with 100 bp (position 250 to þ 50 from the intron insertion site) 50 -sense- and 50 -antisense-strand labeled substrates, respectively. RNP particles from the indicated strains were incubated with the labeled DNA substrates, and the products were analyzed by electrophoresis as indicated above, alongside sequencing ladders obtained from pGEM-t 0.6 with the same 50 -end labeled primer used to generate the substrate. Lanes 1, 2 and 4, RNP particles from S. medicae RMS16, S. meliloti RMO17 and S. meliloti GR4, expressing pKG2.5, respectively; lanes 3 and 5, RNP particles from intronless wild-type strain RMO17 and GR4, respectively; lane 6, boiled RNPs from RMS16 containing pGK2.5; lane 7, DNA substrate incubated in the absence of RNP particles. (d) Endonuclease assay with internally labeled 148 bp (position 250 to þ98 from the intron insertion site) DNA substrate. Lane 1, RNP particles from RMO17 cells expressing pKG2.5; lane 2, RNP particle from intronless wild-type strain RMO17; lanes 3 and 4, RNP particles from RMO17 cells expressing pKG2.5X and pKG2.5-YAHH, respectively; lane 5, boiled RNP particles from RMO17 cells containing pGK2.5. The reaction products shown in lane 1 were incubated with RNase A (lane 6). Bands below the major DNA substrate band reflect heterogeneity in the PCR products. An enlargement of the top of the gel corresponding to lanes 1 –6 is shown as well as a diagram of the product resulting from partial reverse splicing. (e) Sequence of the RmInt1 target site showing the location of the sense-strand cleavage (vertical arrow). The 50 exon (E1), 30 exon (E2), and the intron-binding sequences IBS1, IBS2 and IBS3 are indicated.

That the cleavage of the sense strand occurs by reverse splicing of the intron RNA is demonstrated by the experiment shown in Figure 5(d). In this experiment, precise sense-strand cleavage at the intron-insertion site was observed with an internally labeled double-stranded DNA substrate of 148 bp (positions 2 50 to þ 98), either with the wild-type pKG2.5 or RT mutant pKG2.5-YAHH RNPs (Figure 5(d)). Now, an additional labeled

band seen near the top of the gel corresponds to the product resulting from partial reverse splicing (attachment of intron lariat RNA to the labeled 30 exon). As expected, this product is sensitive to RNase A digestion, which degrades the attached intron lariat RNA, although the expected 98 nt product corresponding to the 30 exon DNA fragment could not be distinguished after RNase digestion as it is masked by a band of similar size in the

937

S. meliloti Group II Intron Mobility

DNA substrate preparation. With 50 -labeled DNA substrates, only a very light band corresponding to fully reverse splice product (linear intron RNA inserted between the two DNA exons) is seen at a similar position near the top of the gels (data not shown). Thus, as for the yeast mtDNA and L. lactis Ll.ltrB introns, the partial reverse spliced product predominates in in vitro reactions in the absence of dNTPs.30 Considered together, our results show that the RmInt1 RNP particles can recognize and promote reverse splicing into double-stranded DNA target sites, but lack site-specific antisensestrand cleavage activity, as expected from the absence of a C-terminal DNA endonuclease domain in the RmInt1 IEP. Single-stranded DNA cleavage by the RmInt1 intron We also assessed the activity of RmInt1 RNP particles with single-stranded DNA substrates by incubating RNP particle preparations from cells expressing pKG2.5 with a 50 -end labeled 60-mer extending from position 2 30 to þ 30 from the intron insertion site. With wild-type pKG2.5 RNP particles, the expected 30 nt product, corresponding to the 50 -exon fragment was detected in polyacrylamide gels, whereas the mutant pKG2.5X RNP particles did not detectably cleave the DNA substrate (Figure 6(a)). The pKG2.5 RmInt1 RNP particles also efficiently cleaved an internally labeled 196 nt single-stranded DNA substrate (Figure 6(a), lanes 3 and 4). Similar results were obtained with single-stranded DNA substrates of 100 bp and 148 bp (data not shown). We confirmed that the cleavage of singlestranded DNA occurred by reverse splicing. RmInt1 RNP particle preparations were incubated with the 60-mer DNA substrate labeled at the 50 or 30 end (61 nt), and the products were analyzed by polyacrylamide gel electrophoresis. The 50 -end labeled substrate gave the expected 30 nt cleavage product corresponding to the 50 exon (Figure 6(b), lane 1). However, the 31 nt 30 exon fragment expected for the 30 -end labeled substrate could be detected only after treatment with RNase A, indicating that the 30 exon is covalently linked to the intron RNA (Figure 6(b), lanes 2 and 3). The RNase treatment also results in a 39 nt band of unknown nature. These results suggest that cleavage of the single-stranded DNA substrates occurs predominantly by partial reverse splicing resulting in covalent attachment of intron lariat RNA to the 30 exon fragment (Figure 6(e)). We characterized the putative reverse-spliced products further by electrophoresis in an agarose gel. A product with a molecular size of 1.5 kb was detected only with the substrate labeled at the 30 end, which again suggests the attachment of the intron RNA to the 30 exon by partial reverse splicing (Figure 6(c), compare lanes 1 and 2). This product is shorter than the expected size of a fulllength intron RNA (1.9 kb), which may be

obscured by the background, likely reflecting intron RNA degradation, as found for the lactococcal intron.30 As expected, this product was sensitive to RNase treatment (lane 3). The reverse spliced product was observed with RNP particles from cells expressing a mutant RT (pKG2.5YAHH, lane 6), but not from cells producing a truncated IEP (pKG2.5X, lane 5). It was not detected with RNP particles pre-treated with phenol or with RNase A (lanes 8 and 9, respectively), or when the substrate was incubated in the absence of RNP particles (lane 7). Assays with the 196 nt, internally labeled, single-stranded DNA substrate showed that cleavage of this longer substrate occurred predominantly by partial reverse splicing (data not shown). Time-course experiments confirmed that single-stranded DNA cleavage and reverse splicing by RmInt1 RNP particles occurred within the first five minutes of incubation (Figure 6(d)). We conclude that the cleavage of single-stranded DNA by RmInt1 RNP particles, like cleavage of double-stranded DNA, occurs mainly by partial reverse splicing of the intron RNA and requires both the intron RNA and the IEP.

Discussion In this work, we characterized the biochemical activities of the S. meliloti group II intron RmInt1, which is efficiently mobile despite encoding a protein lacking a conserved DNA endonuclease domain. Our results show that the RmInt1 IEP has RT activity and that RNP particles containing the IEP and the excised intron RNA cleave singlestranded and double-stranded DNA at the introninsertion site by a reverse splicing reaction. Further, the RNP particles are unable to sitespecifically cleave the opposite strand of the DNA target site, reflecting the absence of the DNA endonuclease domain in the IEP. Our results suggest that RmInt1 mobility occurs by reverse splicing into DNA target sites, but with some alternative mechanism used to procure the primer for target DNA-primed reverse transcription of the inserted intron RNA. First, we showed that both the intron RNA and the IEP are required for RmInt1 splicing and mobility. As first found for the yeast aI2 intron, the mutation YAHH at the RT active site of the RmInt1 IEP completely abolishes RT activity, but the mutant IEP can still bind to the intron RNA to promote RNA splicing and forms RNP particles that can reverse splice into the sense-strand of the DNA target site.3,28,32 In yeast, these mutant RNP particles can promote , 40% residual mobility by an RT-independent double-strand break repair recombination pathway.14,16 By contrast, in L. lactis and S. meliloti, mutations in the RT domain abolish mobility completely, presumably reflecting the lack of host enzymatic machinery to promote mobility by double-strand break repair (this work).15

938

S. meliloti Group II Intron Mobility

Figure 6. Reverse splicing and DNA endonuclease assays with single-stranded DNA substrates. (a) Single-stranded DNA cleavage. RNP particles from RMO17 expressing pKG2.5 and pKG2.5X were incubated with single-stranded substrates of two different lengths: lanes 1 and 2, with a 60-mer oligonucleotide (2 30 to þ30 from the intron insertion site, 50 -end labeled), and lanes 3 and 4 with a 196 nt single-stranded DNA (2 50 to þ146, internally labeled), giving 50 exon cleavage products of different sizes. Products were analyzed by electrophoresis in a denaturing, 6% (w/v) polyacrylamide gel. The gel for the 196 nt substrate is shown in two parts for comparison with the shorter substrate. (b) Endonuclease assays. The 60-mer oligonucleotide (230 to þ 30 from the intron insertion site) labeled at the 50 (60 nt) or the 30 end (61 nt) as indicated was incubated with RNP particles from RMO17 cells containing pKG2.5. The products were analyzed by electrophoresis in a denaturing, 6% (w/v) polyacrylamide gel. Lane 3 shows the reaction product from lane 2 after treatment with RNase A. The positions of the DNA substrate and specific cleavage products are indicated. (c) Reverse splicing assays. The 60-mer oligonucleotide, used as a substrate, was labeled at the 50 (lane 1) or 30 end (lanes 2 –9). The labeled substrate was incubated with RNP particle preparations from RMO17 cells containing pKG2.5 (lanes 1 – 4), pKG2.5X (lane 5), pKG2.5-YAHH (lane 6) or without RNP particles (lane 7). The reaction products were untreated (lanes 1, 2, and 4 – 6) or digested with RNase A þ T1 (RNase, lane 3). Lanes 8 and 9, RNP particles treated before the reactions with phenol-CIA or RNase A, respectively. Products were denatured with glyoxal and analyzed by electrophoresis in a 1% (w/v) agarose gel. The numbers on the left indicate DNA size markers (in kb). Note that the background of unknown nature observed above the reverse spliced product is not dependent on reverse splicing. (d) Time-course of single-stranded cleavage and reverse splicing reactions. Reactions were carried out by incubating RNP particles from RMO17 cells containing pKG2.5 with 50 or 30 -end labeled 60-mer DNA substrate. The products were analyzed by electrophoresis in a denaturing 6% and 3.5% (w/v) polyacrylamide gels to monitor DNA cleavage and reverse splicing, respectively. Six independent reactions were performed in parallel and stopped at the indicated time-points. Open circles and filled circles show amounts of reverse splicing product and DNA cleavage product, respectively. (e) The diagrams show the single-stranded substrates used in these assays and the products resulting from partial reverse splicing and DNA cleavage with 50 and 30 -end labeled substrates.

S. meliloti Group II Intron Mobility

939

Figure 7. Possible mobility mechanisms of RmInt1. The DNA target site may be double-stranded (right side) or transiently single-stranded due to DNA replication or transcription, or to a distorted DNA structure (left side). The doubled-stranded DNA target may be actively unwound directly by the RmInt1 RNP or with the help of a host enzyme (e.g. a DNA helicase)(right side). In either case, the IBS2, IBS1 and IBS3 elements are recognized by base-pairing with the intron RNA, with the IEP presumably contributing to DNA target site recognition by interacting with nucleotide positions 2 15 and þ 4.19Once the DNA target is single-stranded, the ribozyme cleaves at the introninsertion site by a reverse splicing reaction, and the IEP uses its RT activity to reverse transcribe the reverse spliced intron RNA. The primer for reverse transcription may be Okazaki fragments of the lagging strand or the leading strand during DNA replication, or random nicks generated in the antisense strand by host nucleases or non-specific cleavage by the IEP. Although RmInt1, like the Ll.ltrB intron, inserts mainly by partial reverse splicing in in vitro assays in the absence of dNTPs mobility in vivo likely occurs via complete reverse splicing followed by synthesis of a fulllength intron cDNA, which is integrated by a RecA-independent DNA repair mechanism without coconversion of flanking exon sequences.15,16

Next, we found that the RmInt1-derived RNP particles possess RT activity with the exogenous substrate poly(rA)/oligo(dT)18. Further, the RT could use a 60-mer DNA template/primer for DNA synthesis. These findings indicate that the IEP in RNP particles can switch to exogenous templates and has both RT and DNA polymerase activity. In addition, endogenous RT assays after the annealing of 20-mer DNA primers to different 30 -exon and intron positions suggest that some fraction of the RmInt1 RT may be bound to the precursor RNA in a way that enables it to initiate at a range of distances downstream from the intron in the 30 exon. This range appears to be wider than that for the yeast aI2 and lactococcal introns, which showed a maximum RT activity after annealing of primers ending at positions þ 9 to þ 12 and þ 6 to þ 14, respectively.29,31 DNA endonuclease assays show that the RmInt1 intron RNA in RNP particles can cleave doublestranded and single-stranded DNA target sites at the intron-insertion site primarily by a partial reverse splicing reaction in which intron lariat is covalently linked to the 30 exon. However, the

intron RNPs may have decreased efficiency with long double-stranded DNA substrates and are unable to site-specifically cleave the antisense strand as expected from the lack of DNA endonuclease domain. For the yeast mtDNA and lactococcal introns, the interaction of the IEP with the DNA target leads to local DNA unwinding, enabling the intron RNA to base-pair and reverse splice into the resulting single-stranded region.11 As RmInt1 RNP particles cleave the sense strand of double-stranded DNA substrates, they appear capable of recognizing the specific target site of the intron in double-stranded DNA. As in the case of the yeast and lactococcal introns, this may involve initial recognition of specific nucleotide residues by the IEP, followed by base-pairing of the intron RNA, which requires DNA unwinding. The S. meliloti RmInt1 IEP has only a short (20 amino acid residues) C-terminal extension after domain X, which is conserved among other phylogenetically related group II introns, but does not appear to be related to the DNA-binding region present in yeast mitochondria and lactococcal intron IEPs.27 This difference with

940

respect to the DNA-binding domain may account for the more limited DNA target site recognition by the RmInt1 IEP,19 and could potentially increase the difficulty of unwinding double-stranded DNA. In addition, the RmInt1 RNPs require an alternate means of generating the primer for reverse transcription. Several possible mobility mechanisms taking into account these considerations are illustrated by Figure 7. One possibility is that the initial target is in double-stranded DNA and that the RmInt1 RNP particle by itself or in conjunction with the intron RNA or the help of host factors, such as DNA helicases, promotes local DNA unwinding. A second possibility is that the RmInt1 intron reverse splices preferentially into a transiently singlestranded DNA target site at a replication fork or during transcription. Finally, the target site may have an unusual DNA structure with partially single-stranded character, either inherently or because of bound proteins. The streptococcal group II intron GBSi1 and other bacterial class C introns,33 which also lack the conserved DNA endonuclease domain, appear to insert downstream of putative transcription terminators, consisting of an inverted repeat followed by a poly(T) tract, which may have inherent single-stranded character. The RmInt1 target site has a run of four T residues at the beginning of IBS2, a region that may be critical to nucleate DNA unwinding, and the distal 50 -exon positions 2 19 to 2 14 potentially can base-pair with the IBS1 positions 2 2 to 2 7 to form a short inverted repeat. These features may facilitate local DNA unwinding by RmInt1 RNPs, compensating for more limited protein contacts with the DNA target site. After reverse splicing of the intron, cDNA synthesis may be initiated by using a primer generated by bottom-strand cleavage by an unknown host DNA endonuclease or by a non-specific cleavage activity of the IEP. Alternatively, priming could occur by using the 30 end of a nascent leading or lagging strand during DNA replication. In the case of the L. lactis Ll.ltrB intron, double-stranded DNA cleavage by the intron RNP leads to rapid degradation of target-containing plasmids by host nucleases.15 We have not observed such degradation in RmInt1 homing assays carried out either in S. meliloti or in heterologous hosts (F.M.-A. & N.T., unpublished results), a point favoring the second type of mechanism. Finally, we note that, like the L. lactis Ll.ltrB intron, the RmInt1 RNP particles carry out mainly partial reverse splicing in vitro in the absence of dNTPs, while the template for cDNA synthesis in vivo is likely to be the fully reverse spliced intron RNA.15,30In both cases, the use of fully reverse spliced intron RNA for synthesis of a full-length cDNA is consistent with the absence of co-conversion in the flanking exons and the RecA-independence of the homing process.5,15,18 The mobility mechanism proposed for RmInt1 resembles, in some respects, that for group II intron retrotransposition to ectopic sites. The latter

S. meliloti Group II Intron Mobility

appears to involve reverse splicing into doublestranded or transiently single-stranded DNA target sites that resemble the normal homing site, with an alternate means of generating the primer for reverse transcription, either non-specific nicking of the opposite strand or the use of nascent leading or lagging strand generated during DNA replication.4,6,8 However, ectopic transposition occurs at very low frequency, while RmInt1 homing occurs at high frequency (approaching 100%) and efficiency (20 – 45% of the recipient target).19 Thus, RmInt1 may have specific adaptations of this mechanism that make it more efficient in vivo. Overall, our results indicate that the homing mechanism of RmInt1 is similar to that of the yeast and lactococcal introns, in that the intron RNA reverse splices directly in a DNA target site and is then reverse transcribed by the IEP, but differs, in that RmInt1 must use an alternate mechanism to generate or access the primer for TPRT of the inserted intron RNA. The mobility mechanism used by RmInt1 may be used by other group II introns that lack a conserved DNA endonuclease domain.

Materials and Methods Bacterial strains and growth conditions S. meliloti (RMO17, GR4) and Sinorhizobium medicae (RMS16) strains were grown at 28 8C on TY medium.34 Antibiotics were added when required, at the following concentrations: kanamycin, 200 mg/ml; tetracycline, 10 mg/ml. Construction of RmInt1 mutant derivatives pKG2.5 and pKG2.5X have been described.18 Other mutants were constructed by site-directed mutagenesis, using the Altered sites II in vitro mutagenesis pAlter-1 system (Promega). To generate pAL2.5, the Sph I fragment containing RmInt1 from pGEMEX 2.518 was inserted into pAlter-1. pAL2.5-YAHH was derived from pAL2.5 by site-directed mutagenesis to replace the RmInt1 nucleotides GAC GAT by CAC CAT (positions 1234– 1239) and the corresponding amino acid residues D230 and D231 with H230 and H231. To generate pKG2.5-YAHH, pAL2.5-YAHH was digested with Bam HI and Spe I, and the RmInt1-containing fragment was inserted into pKG0 digested with Bam HI and Spe I.18 pAL-DC29 was derived from pAL2.5 by site-directed mutagenesis to insert the nucleotides TGAT at RmInt1 position 1716 (this converts codon R391 to a stop codon). To generate pKG2.5-DC29, pAL-DC29 was digested with Bam HI and Spe I, and the RmInt1-containing fragment was inserted into pKG0. pAL2.5D5-CGA was derived from pAL2.5 by site-directed mutagenesis, replacing the RmInt1 nucleotides GTT by CGA (positions 1840– 1842). To generate pKG2.5D5-CGA, pAL2.5D5-CGA was digested with Bam HI and Spe I, and the RmInt1-containing fragment was inserted into pKG0. pKG-IEP contains an 1281 bp insert generated by PCR using Pfu DNA Polymerase (Promega) with the complementary primers IEP-Spe I (50 -GGG GAC TAG TGG AAA CAG GAT GAC TTC GGA) and IEP-Sac I (50 -GGG

941

S. meliloti Group II Intron Mobility

GGA GCT CTC AGG TAA ACG TGT TCG TTCC). The amplicon was digested with Spe I and Sac I, and cloned between the Spe I and Sac I sites of pKG0 to generate pKG-IEP. Primer extension assay of intron RNA splicing in vivo Total cellular RNA was isolated from strain RMO17 containing plasmids expressing wild-type or mutant RmInt1. Cells were cultured in 10 ml of TY medium supplemented with kanamycin until they reached an A600 of 0.6. RNA was extracted as described.17 The primer annealing mixture contained 20 mg of total RNA in 10 mM Pipes (pH 7.5), 400 mM NaCl and 0.2 pmol (300,000 c.m) of (50 -32P)-labeled primer (50 -TGA AAG CCG ATC CCG GAG-30 ). The mixture (10 ml) was first heated at 85 8C for five minutes, then incubated at 60 8C for three minutes and cooled slowly to 44 8C. Extension reactions were started by adding 40 ml of 50 mM Tris – HCl (pH 8.0), 60 mM NaCl, 10 mM DTT, 6 mM magnesium acetate, 1 mM each of all four dNTPs, 60 mg/ml of actinomycin D (Sigma), two units of RNasin (Boehringer Mannheim) and seven units of AMV RT (Boehringer Mannheim). Reaction mixtures were incubated at 44 8C for 60 minutes. The reaction was stopped by adding 5 ml of 3 M sodium acetate (pH 5.2) and 150 ml of cold ethanol. Half of each sample was analyzed on a denaturing, 6% (w/v) polyacrylamide gel.

was resuspended in 20 ml of 0.1% (w/v) ice-cold Sarcosyl in TE (Tris – HCl 10 mM, EDTA 1 mM pH 8.0) and centrifuged in a Sorvall SS34 rotor (3000g for five minutes at 4 8C). The resulting pellet was washed with 20 ml of sterile water and then with 20 ml of ice-cold 150 mM NaCl. The cell pellet was resuspended in 4 ml of ice-cold buffer A (50 mM Tris – HCl (pH 7.5), 1 mM EDTA, 1 mM DTT, 10% (v/v (glycerol), with lysozyme (Boehringer Manheim) added to a final concentration of 2 mg/ml. The cells were then lysed by freezing (2 70 8C) and thawing (28 8C) three times. To this viscous solution, we added four volumes of HKCTD buffer (500 mM KCl, 50 mM CaCl2, 25 mM Tris – HCl (pH 7.5), 5 mM DTT) and sonicated at a power setting of 5 (Branson SONIFIER 250) for five bursts. The suspension was centrifuged in a Sorvall SS34 rotor (12,000g for 15 minutes at 4 8C) to pellet the cell debris, and the supernatant was layered on a 5 ml 1.85 M sucrose cushion/ HKCTD and centrifuged in a Beckman Ti50 rotor (50,000 rpm for 17 hours at 4 8C). The pellet containing the RNPs was washed three times with 1 ml of ice-cold distilled water and then dissolved in 100– 150 ml of icecold 10 mM Tris – HCl (pH 8.0), 1 mM DTT. The solution was transferred to a microfuge tube and spun at 20,000g for five minutes at 4 8C. The RNPs were stored frozen at 270 8C. The yield of RNP particles was 5 – 25 A260 units per 100 ml of culture. The RmInt1 IEP is not detected by SDS-PAGE and hence appears to be present in very small amounts in the RNP preparations, which contain mostly ribosomes.

Homing assays RmInt1 homing assays were performed in S. meliloti strain RMO17. We studied homing between compatible intron donor (pKG2.5 and mutant derivatives) and recipient (pJB0.6 þ ) plasmids containing a 640 bp fragment encompassing the intron-insertion site of ISRm2011-2.18 Plasmids isolated from transconjugants were digested with Sal I and analyzed for group II intron insertion by agarose gel electrophoresis and Southern hybridization. Hybridization was carried out using a DNA-target probe as described.18 Oligonucleotides The oligonucleotides used in this work were: 2 31, 50 GAAGCACTGCCCCGGAATTG; þ 1, 50 -CATGTTCGTC TTCGTCCAGG; þ9, 50 -AGCGGCGCCATGTTCGTCTT; þ31, 50 -GCCGCGAGGCGCCCAGCCCC; þ52, 50 -CGT AGCCCACCAGTCGTT; þ78, 50 -GGTCATGGTGTTCC AATGGCC; þ127, 50 -GAGGATAAAGGGAGCGCTGA; Int2, 50 -CCGTCCATAGTAGGCAATCC (complementary to positions 1617 to 1598 of the RmInt1 intron,17); 60mer, 50 -ATCCCGCCCGCCTCGTTTTCATCGATGAGAC CTGGACGAAGACGAACATGGCGCCGCTGC. Preparation of RNP particles RNP particles were isolated from bacterial strains essentially as described.30 A single colony from a plate of each bacterial strain was inoculated in 3 ml of TY medium and incubated in a shaker at 28 8C for two days until stationary phase was reached.Then 1 ml of the culture was added to 200 ml of TY medium in a 500 ml flask and incubated at 28 8C in a shaker until the culture reached exponential phase (A600 0.6– 0.9). Bacterial cells were harvested by centrifugation in a Sorvall GSA rotor (4000g for five minutes at 4 8C). The pellet

RT assays For exogenous RT activity, we added RNP particles (0.040 A260 unit) to 10 ml of reaction medium containing 10 mM KCl, 25 mM MgCl2, 50 mM Tris – HCl (pH 7.5), 5 mM DTT, 1 mg of poly(rA)/oligo(dT)18 or poly(rA) and 10 mCi of [a-32P]dTTP (3000 Ci/mmol; Amersham) and incubated for 20 minutes at 37 8C. We then spotted 8 ml of reaction products onto DEAE paper and counted Cerenkov radioactivity.28 Endogenous reverse transcription reactions were performed in 10 ml of reaction medium containing 10 mM KCl, 25 mM MgCl2, 50 mM Tris – HCl (pH 7.5), 5 mM DTT, 200 mM each of dATP, dTTP and dGTP and 10 mCi of [a-32P]dCTP (3000 Ci/mmol; Amersham), 20 pmol of DNA oligonucleotide primer complementary to the sense endogenous RNA and 0.040 A260 unit of RNPs for 20 minutes at 37 8C.30 Primer-mediated initiation of DNA synthesis reactions Primer-mediated initiation of DNA synthesis was done by incubating 0.080 A260 unit of RNPs in 10 ml of a reaction medium containing 10 mM KCl, 25 mM MgCl2, 50 mM Tris –HCl (pH 7.5), 5 mM DTT, 40 pmol of substrate (60-mer oligonucleotide annealed or not to a complementary 20-mer primer), 10 mCi of [a-32P]dCTP, 0.2 mM one ddNTP and 0.2 mM the other two dNTPs (if no ddNTP was used, the mixture contained 0.2 mM each of the three dNTPs). After ten minutes at 37 8C, the reactions were chased with 0.2 mM unlabeled dCTP for another ten minutes. Products were extracted and precipitated in ethanol as described above, and analyzed by electrophoresis in a denaturing, 8% (w/v) polyacrylamide gels, alongside molecular markers or sequencing ladders generated from pGEM-t 0.6.

942

Preparation of DNA substrates for DNA endonuclease and reverse splicing reactions Double-stranded DNA substrates for DNA endonuclease and reverse splicing assays were prepared by PCR, using 50 -end-labeled primers to label separately the sense and antisense strands. We labeled 40 pmol of oligonucleotide with phage T4 polynucleotide kinase (New England Biolabs) and 200 mCi of [g-32P] ATP (3000 Ci/mmol; Amersham) in a reaction volume of 50 ml. For most experiments, primers were labeled with 100 mCi of [g-32P]ATP (3000 Ci/mmol; Amersham) in a total volume of 40 ml. Oligonucleotides were purified by extraction with phenol-CIA (phenol/chloroform/isoamyl alcohol, 25:24:1, by vol.) and chromatography on G-25 Sephadex columns and were precipitated in ethanol or used directly. For PCR, we added 10 ng of template containing 640 bp of target site for RmInt1 (pGEM-t 0.6), 40 pmol of unlabeled primer, 0.8 ml of 25 mM dNTPs and 0.2 unit of Vent polymerase (New England Biolabs) to a final volume of 50 ml. The amplification conditions were as follows: initial denaturation at 94 8C for two minutes; 25 cycles at 94 8C for 30 seconds, 60 8C (67 8C for primers 231/ þ 31) for 30 seconds and 72 8C for 30 seconds, followed by final extension at 72 8C for ten minutes. Double-stranded DNA substrates for reverse splicing assays were labeled internally with 50 mCi of [a-32P]dTTP or [a-32P]dCTP (3000 Ci/mmol; Amersham) in 50 ml PCR reactions containing 10 ng of template, 50 pmol of each primer, 200 mM each of dATP and dGTP, with 200 mM dCTP and 30 mM dTTP (if [a-32P]dTTP was used) or 200 mM dTTP and 30 mM dCTP (if [a-32P]dCTP was used) and 0.2 unit of Vent polymerase (New England Biolabs). The 60-mer DNA oligonucleotide for endonuclease assays was 50 -end labeled as indicated above. The substrate was 30 -end labeled with terminal deoxynucleotidyl transferase (Roche) and 50 mCi of [a-32P]ddATP (3000 Ci/mmol; Amersham), according to the manufacturer’s instructions. The single-stranded DNA of 196 nt was generated by asymmetric PCR with internal labeling, using as a template a product from a standard PCR purified by electrophoresis in a 2% (w/v) agarose gel. PCR was carried out in a final volume of 50 ml, containing 200 mM each dATP, dGTP, dTTP and 30 mM dCTP, 50 mCi of [a-32P]dCTP, 100 pmol of sense primer, and 0.2 unit of Vent polymerase. The annealing temperature was 57 8C. All the above substrates were further purified by electrophoresis in a non-denaturing, 6% (w/v) polyacrylamide gel. DNA endonuclease and reverse splicing assays DNA endonuclease activity was assayed by incubating RNP particles (0.080– 0.200 A260 unit) in 10 ml of reaction medium containing 10 mM KCl, 25 mM MgCl2, 50 mM Tris – HCl (pH 7.5), 5 mM DTT and 150,000 – 370,000 cpm of the DNA substrate. Reactions were carried out for 20 minutes at 37 8C. Products were extracted with phenol-CIA in the presence of 0.3 M sodium acetate (pH 5.2), 8 mM EDTA and linear acrylamide carrier (BIO RAD) at a final concentration of 0.2% (w/v). They were precipitated in ethanol and analyzed by electrophoresis in a denaturing 6% polyacrylamide gel.30 Molecular size markers or DNA sequence ladders generated from pGEM-t 0.6 (containing the 640 bp fragment encompassing the intron-insertion site of

S. meliloti Group II Intron Mobility

ISRm2011-2) with the corresponding 50 -end labeled primer were used. The gels were dried and autoradiographed with a Molecular Dynamics Imager. In Figure 5(d) (lane 6) and Figure 6(b) (lane 3) reaction products were treated with 1 mg of RNase A for 30 minutes at 37 8C. For analysis of products in agarose gels, reverse splicing reactions were performed as described above.30 In Figure 6(c) (lane 3), reaction products were digested with 0.4 mg of RNase A and 0.5 unit of RNase T1 for 15 minutes at 37 8C. RNPs were extracted with phenol-CIA or incubated before the reaction with 1 mg of RNase A for 15 minutes at 37 8C (lanes 8 and 9, respectively). Reaction products were extracted, precipitated as described above, denatured with glyoxal and analyzed by electrophoresis in a 1% (w/v) agarose gel.

Acknowledgements We thank Drs Marlene Belfort and Manabu Matsuura for critical reading of the manuscript, Lcda Dolores Molina for preparing RNP particles and carrying out RT assays, and Dr Jose´ Ignacio Jime´nez-Zurdo for critical reading of the manuscript and the design of Figure 7. E. Mun˜oz-Adelantado is supported by an MEC grant. This work was funded by the research project BIO99-0905 and BIO2002-02579 from the Ministerio de Ciencia y Tecnologı´a to N.T., and NIH grant GM37949 to A.M.L.

References 1. Lambowitz, A. M., Caprara, M. G., Zimmerly, S. & Perlman, P. S. (1999). Group I and group II ribozymes as RNPs: clues to the past and guides to the future. In The RNA World (Gesteland, R. F., Cech, T. R. & Atkins, J. F., eds), 2nd edit., pp. 451– 485, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 2. Belfort, M., Derbyshire, V., Parker, M. M., Cousineau, B. & Lambowitz, A. M. (2002). Mobile introns: pathways and proteins. In Mobile DNA (Craig, N. L., Craigie, R., Gellert, M. & Lambowitz, A. M., eds), 2nd edit., pp. 761– 783, ASM Press, Washington, DC. 3. Zimmerly, S., Guo, H., Eskes, R., Yang, J., Perlman, P. S. & Lambowitz, A. M. (1995). A group II intron RNA is a catalytic component of a DNA endonuclease involved in intron mobility. Cell, 83, 529– 538. 4. Yang, J., Mohr, G., Perlman, P. S. & Lambowitz, A. M. (1998). Group II intron mobility in yeast mitochondria: target DNA primed reverse transcription activity of aI1 and reverse splicing into DNA transposition sites in vitro. J. Mol. Biol. 282, 505– 523. 5. Martı´nez-Abarca, F. & Toro, N. (2000). RecA-independent ectopic transposition in vivo of a bacterial group II intron. Nucl. Acids Res. 28, 4397– 4402. 6. Dickson, L., Huang, H.-R., Liu, L., Matsuura, M., Lambowitz, A. M. & Perlman, P. S. (2001). Retrotransposition of a yeast group II intron occurs by reverse splicing directly into ectopic DNA sites. Proc. Natl Acad. Sci. USA, 98, 13207 –13212.

S. meliloti Group II Intron Mobility

7. Mun˜oz, E., Villadas, P. J. & Toro, N. (2001). Ectopic transposition of a group II intron in natural bacterial populations. Mol. Microbiol. 41, 645– 652. 8. Ichiyanagi, K., Beauregard, A., Lawrence, S., Smith, D., Cousineau, B. & Belfort, M. (2002). Retrotransposition of the Ll.ltrB group II intron proceeds predominantly via reverse splicing into DNA targets. Mol. Microbiol. 46, 1259– 1272. 9. Mohr, G., Perlman, P. S. & Lambowitz, A. M. (1993). Evolutionary relationships among group II intronencoded proteins and identification of a conserved domain that may be related to maturase function. Nucl. Acids Res. 22, 4991– 4997. 10. Guo, H., Zimmerly, S., Perlman, P. S. & Lambowitz, A. M. (1997). Group II intron endonucleases use both RNA and protein subunits for recognition of specific sequences in double-stranded DNA. EMBO J. 16, 6835– 6848. 11. Singh, N. N. & Lambowitz, A. M. (2001). Interaction of a group II intron ribonucleoprotein endonuclease with its DNA target site investigated by DNA footprinting and modification interference. J. Mol. Biol. 309, 361– 386. 12. Gorbalenya, A. E. (1994). Self-splicing group I and group II introns encode homologous (putative) DNA endonucleases of a new family. Protein Sci. 3, 1117 – 1120. 13. Shub, D. A., Goodrich-Blair, H. & Eddy, S. R. (1994). Amino acid sequence motif of group I endonucleases is conserved in open reading frames of group II introns. Trends Biochem. Sci. 19, 402– 404. 14. Eskes, R., Yang, J., Lambowitz, A. M. & Perlman, P. S. (1997). Mobility of yeast mitochondrial group II introns: engineering a new site specificity and retrohoming via full reverse splicing. Cell, 88, 865– 874. 15. Cousineau, B., Smith, D., Lawrence-Cavanagh, S., Mueller, J. E., Yang, J., Mills, D. et al. (1998). Retrohoming of a bacterial group II intron: mobility via complete reverse splicing, independent of homologous recombination. Cell, 94, 451– 462. 16. Eskes, R., Liu, L., Ma, H., Chao, M. Y., Dickson, L., Lambowitz, A. M. & Perlman, P. S. (2000). Multiple homing pathways used by yeast mitochondrial group II introns. Mol. Cell. Biol. 20, 8432– 8446. 17. Martı´nez-Abarca, F., Zekri, S. & Toro, N. (1998). Characterization and splicing in vivo of a Sinorhizobium meliloti group II intron associated with particular insertion sequences of the IS630-Tc1/IS3 retroposon superfamily. Mol. Microbiol. 28, 1295–1306. 18. Martı´nez-Abarca, F., Garcı´a-Rodrı´guez, F. M. & Toro, N. (2000). Homing of a bacterial group II intron with an intron-encoded protein lacking a recognizable endonuclease domain. Mol. Microbiol. 35, 1405– 1412. 19. Jime´nez-Zurdo, J. I., Garcı´a-Rodrı´guez, F. M., Barrientos-Dura´n, A. & Toro, N. (2003). DNA target requirement for homing in vivo of a bacterial group II intron encoding a protein lacking the DNA endonuclease domain. J. Mol. Biol. 326, 413– 423. 20. Martı´nez-Abarca, F. & Toro, N. (2000). Group II introns in the bacterial world. Mol. Microbiol. 38, 917–926.

943

21. Zimmerly, S., Hausner, G. & Wu, X.-C. (2001). Phylogenetic relationships among group II intron ORFs. Nucl. Acids Res. 29, 1238– 1250. 22. Dai, L. & Zimmerly, S. (2002). Compilation and analysis of group II intron insertions in bacterial genomes: evidence for retroelement behaviour. Nucl. Acids Res. 30, 1091– 1102. 23. Galagan, J. E., Nusbaum, C., Roy, A., Endrizzi, M. G., Macdonald, P., FitzHugh, W. et al. (2002). The genome of Methanosarcina acetivorans reveals extensive metabolic and physiological diversity. Genome Res. 12, 532– 542. 24. Toro, N. (2003). Bacteria and Archaea group II introns; additional mobile genetic elements in the environment. Environ. Microbiol. 5, 143– 151. 25. Mills, D. A., Mckay, L. L. & Dunny, G. M. (1996). Splicing of a group II intron involved in the conjugative transfer of pRS01 in lactococci. J. Bacteriol. 178, 3531 –3538. 26. Michel, F., Umesomo, K. & Ozeki, H. (1989). Comparative and functional anatomy of group II catalytic introns—a review. Gene, 82, 5 – 30. 27. San Filippo, J. & Lambowitz, A. M. (2002). Characterization of the C-terminal DNA-binding/DNA endonuclease region of a group II intron-encoded protein. J. Mol. Biol. 324, 933– 951. 28. Moran, J. V., Zimmerly, S., Eskes, R., Kennell, J. C., Lambowitz, A. M., Butow, R. A. & Perlman, P. S. (1995). Mobile group II introns of yeast mitochondrial DNA are novel site-specific retroelements. Mol. Cell. Biol. 15, 2828– 2838. 29. Zimmerly, S., Moran, J. V., Perlman, P. S. & Lambowitz, A. M. (1999). Group II intron reverse transcriptase in yeast mitochondria. Stabilization and regulation of reverse transcriptase activity by the intron RNA. J. Mol. Biol. 289, 473– 490. 30. Matsuura, M., Saldanha, R., Ma, H., Wank, H., Yang, J., Mohr, G. et al. (1997). A bacterial group II intron encoding reverse transcriptase, maturase, and DNA endonuclease activities: biochemical demonstration of maturase activity and insertion of new genetic information within the intron. Genes Dev. 11, 2910 –2924. 31. Wank, H., San Filippo, J., Singh, R. N., Matsuura, M. & Lambowitz, A. M. (1999). A reverse transcriptase/maturase promotes splicing by binding at its own coding segment in a group II intron RNA. Mol. Cell, 4, 239– 250. 32. Zimmerly, S., Guo, H., Perlman, P. S. & Lambowitz, A. M. (1995). Group II intron mobility occurs by target DNA-primed reverse transcription. Cell, 82, 545 –554. 33. Granlund, M., Michel, F. & Norgren, M. (2001). Mutually exclusive distribution of IS1548 and GBSi1, an active group II intron identified in human isolates of group B streptococci. J. Bacteriol. 183, 2560 –2569. 34. Robertsen, B. K., Aiman, P., Darvill, A. G., McNeil, M. & Alberstein, P. (1981). The structure of acidic extracellular polysaccharides secreted by Rhizobium leguminosarum and Rhizobium trifolii. Plant Physiol. 67, 389–400.

Edited by M. Belfort (Received 18 June 2002; received in revised form 23 January 2003; accepted 27 January 2003)