Excision of the Sinorhizobium meliloti Group II Intron RmInt1 as Circles in Vivo

Excision of group II introns as circles has been described only for a few eukaryotic introns and little is known about the mechanisms involved, the relevance or consequences of the process. We report that splicing of the bacterial group II intron RmInt1 in vivo leads to the formation of both intron lariat and intron RNA circles. We determined that besides being required for the intron splicing reaction, the maturase domain of the intron-encoded protein also controls the balance between lariat and RNA intron circle production. Furthermore, comparison with in vitro self-splicing products indicates that in vivo, the intron-encoded protein appears to promote the use of a correct EBS1/IBS1 intron-exon interaction as well as cleavage at, or next to, the expected 3′ splice site. These findings provide new insights on the mechanism of excision of group II introns as circles. Excision of group II introns as circles has been described only for a few eukaryotic introns and little is known about the mechanisms involved, the relevance or consequences of the process. We report that splicing of the bacterial group II intron RmInt1 in vivo leads to the formation of both intron lariat and intron RNA circles. We determined that besides being required for the intron splicing reaction, the maturase domain of the intron-encoded protein also controls the balance between lariat and RNA intron circle production. Furthermore, comparison with in vitro self-splicing products indicates that in vivo, the intron-encoded protein appears to promote the use of a correct EBS1/IBS1 intron-exon interaction as well as cleavage at, or next to, the expected 3′ splice site. These findings provide new insights on the mechanism of excision of group II introns as circles. Group II introns are large catalytic RNAs with a conserved secondary structure consisting of six domains, one of which (dIV) may incorporate the coding sequence of a reverse transcriptase (RT) 3The abbreviations used are: RT, reverse transcriptase; IEP, intron-encoded protein; RNP, RNA-protein; IBS, intron-binding sites; EBS, exon-binding sites; nt, nucleotide. 3The abbreviations used are: RT, reverse transcriptase; IEP, intron-encoded protein; RNP, RNA-protein; IBS, intron-binding sites; EBS, exon-binding sites; nt, nucleotide. (1Michel F. Ferat J.L. Annu. Rev. Biochem. 1995; 64: 435-461Crossref PubMed Scopus (488) Google Scholar). Although some group II introns self-splice in vitro, this reaction requires nonphysiological conditions, and in vivo, proteins are required to fold the intron RNA into a catalytically active structure. Group II intron-encoded proteins (IEPs) promote both splicing and mobility of the intron RNA through formation of a specific RNA-protein (RNP) complex (2Lambowitz A.M. Belfort M. Annu. Rev. Biochem. 1993; 62: 587-622Crossref PubMed Scopus (532) Google Scholar, 3Lambowitz A.M. Caprara M.G. Zimmerly S. Perlman P.S. Cech T.R. Atkins J.F. 2nd Ed. The RNA World. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY1999: 451-485Google Scholar, 4Lambowitz A.M. Zimmerly S. Annu. Rev. Genet. 2004; 38: 1-35Crossref PubMed Scopus (351) Google Scholar). The IEPs have two conserved domains, an N-terminal RT domain and domain X, a putative RNA-binding domain associated with maturase (RNA splicing) activity. In some cases the IEP also includes a C-terminal DNA-binding and a DNA-endonuclease domain (5San Filippo J. Lambowitz A.M. J. Mol. Biol. 2002; 324: 933-951Crossref PubMed Scopus (77) Google Scholar). Group II introns splice typically by the same two sequential transesterification reactions used in nuclear mRNA splicing (1Michel F. Ferat J.L. Annu. Rev. Biochem. 1995; 64: 435-461Crossref PubMed Scopus (488) Google Scholar). In a first step, the 2′-OH group of a branch point nucleotide residue, usually a bulged adenosine in dVI, attacks the 5′ splice junction resulting in cleavage of the 5′ exon and the formation of an intron-3′ exon-branched lariat intermediate. The released 5′ exon remains associated to the intron via base pairing of the intron-binding sites (IBS1 and IBS2) to the exon-binding sites (EBS1 and EBS2) located in domain dI. In the second step, the free 3′-OH of the 5′ exon attacks the 3′ splice junction leading to the release of the intron lariat and the ligation of the 5′ and 3′ exons. There also exists an alternate pathway, in which the first splicing step is initiated by a nucleophilic attack of water or an OH– ion, resulting in the formation of a linear intron-3′ exon intermediate that subsequently participates in a normal second step reaction. This hydrolytic pathway is observable in vitro and has been shown to be used in vivo by yeast mitochondrial introns carrying branch-site mutations (6Podar M. Chu V.T. Pyle A.M. Perlman P.S. Nature. 1998; 391: 915-918Crossref PubMed Scopus (86) Google Scholar). Moreover, even though most of the plant chloroplast group II introns follow the typical lariat-generating pathway, hydrolytic splicing has been reported for intron trnV, which lacks the conserved bulged A in dVI, whereas both the hydrolytic and branching pathways seem to coexist in the case of the barley trnK intron (7Vogel J. Börner T. EMBO J. 2002; 21: 3794-3803Crossref PubMed Scopus (72) Google Scholar). Circular DNA versions of group II introns are known to exist in Podospora anserina where they are somehow associated with cellular senescence (8Osiewacz H.D. Esser K. Curr. Genet. 1984; 8: 299-305Crossref PubMed Scopus (160) Google Scholar, 9Schmidt W.M. Schweyen R.J. Wolf K. Mueller M.W. J. Mol. Biol. 1994; 243: 157-166Crossref PubMed Scopus (30) Google Scholar, 10Begel O. Boulay J. Albert B. Dufour E. Sainsart-Chanet A. Mol. Cell. Biol. 1999; 19: 4093-4100Crossref PubMed Scopus (43) Google Scholar). In this particular case a model was proposed, by which the circular DNA molecules would be generated via transposition of the intron in front of itself, followed by excision of one of the tandem copies by homologous recombination (11Sellem C.H. Lecellier G. Belcour L. Nature. 1993; 366: 176-178Crossref PubMed Scopus (98) Google Scholar, 12Müeller M.W. Allmaier M. Eskes R. Schweyen R.J. Nature. 1993; 366: 174-176Crossref PubMed Scopus (85) Google Scholar, 13Sainsard-Chanet A. Begel O. Belcour L. J. Mol. Biol. 1994; 242: 630-643Crossref PubMed Scopus (16) Google Scholar). More recent data indicate that yeast intron aI5γ can be also excised as a true circular form in vitro (14Murray H.L. Mikheeva S. Coljee V.W. Turczyk B.M. Donahue W.F. Bar-Shalom A. Jarrell K.A. Mol. Cell. 2001; 8: 201-211Abstract Full Text Full Text PDF PubMed Scopus (44) Google Scholar). The circle seems to result from formation of a 2′-5′ phosphodiester bond between the last and first residues of the intron and it has been proposed that this reaction requires the 3′ exon to have been released first from precursor molecules by a trans-splicing reaction triggered by 5′ exon molecules previously generated by the so-called SER (spliced-exon reopening) reaction. For the yeast intron aI2, both RNA and DNA circles have been detected in vivo and it is thought that circular aI2 RNAs are copied into DNA, presumably by reverse transcription (14Murray H.L. Mikheeva S. Coljee V.W. Turczyk B.M. Donahue W.F. Bar-Shalom A. Jarrell K.A. Mol. Cell. 2001; 8: 201-211Abstract Full Text Full Text PDF PubMed Scopus (44) Google Scholar). In addition, the presence of circular forms of excised intron molecules in plant mitochondria has been recently reported (15Li-Pook-Than J. Bonen L. Nucleic Acids Res. 2006; 34: 2782-2790Crossref PubMed Scopus (58) Google Scholar). However, circular intron molecules have not been described so far for bacterial group II introns. RmInt1 is a bacterial group II intron identified in Sinorhizobium meliloti, the nitrogen-fixing symbiont of alfalfa (Medicago sativa). The RmInt1 IEP is required for intron splicing in vivo (16Muñoz-Adelantado E. San Filippo J. Martínez-Abarca F. García-Rodríguez F.M. Lambowitz A.M. Toro N. J. Mol. Biol. 2003; 327: 931-943Crossref PubMed Scopus (36) Google Scholar), but like those of many other bacterial group II introns, it lacks C-terminal DNA endonuclease and DNA-binding domains (5San Filippo J. Lambowitz A.M. J. Mol. Biol. 2002; 324: 933-951Crossref PubMed Scopus (77) Google Scholar, 17Martínez-Abarca F. García-Rodríguez F.M. Toro N. Mol. Microbiol. 2000; 35: 1405-1412Crossref PubMed Scopus (58) Google Scholar, 18Zimmerly S. Hausner G. Wu X-C. Nucleic Acids Res. 2001; 29: 1238-1250Crossref PubMed Scopus (152) Google Scholar, 19Dai L. Zimmerly S. Nucleic Acids Res. 2002; 30: 1091-1102Crossref PubMed Scopus (144) Google Scholar, 20Toro N. Environ. Microbiol. 2003; 5: 143-151Crossref PubMed Scopus (73) Google Scholar). RmInt1 is nevertheless an efficient mobile element that has two retrohoming pathways for mobility, with predominant use of the nascent lagging strand at DNA replication forks for priming (21Martínez-Abarca F. Barrientos-Durán A. Fernández-López M. Toro N. Nucleic Acids Res. 2004; 32: 2880-2888Crossref PubMed Scopus (51) Google Scholar). Recently, we reported that RmInt1 self-splices in vitro in the absence of the IEP, but the in vitro activity of the intron is atypical in that the second step of splicing is unusually inefficient and several unconventional products are generated alongside the expected excised intron and ligated exons (22Costa M. Michel F. Molina-Sánchez M.D. Martinez-Abarca F. Toro N. Biochimie (Paris). 2006; 88: 711-717Crossref PubMed Scopus (17) Google Scholar). In this study, we have investigated the excision of the S. meliloti RmInt1 intron in vivo. Our data indicate that the RmInt1 group II intron is excised in vivo both as intron lariat and intron circles and that the IEP seems to determine the balance between two excision and used in this was S. meliloti This was at on for RNA and was used for the and of For the was at for and for E. RmInt1 and and have been previously reported (16Muñoz-Adelantado E. San Filippo J. Martínez-Abarca F. García-Rodríguez F.M. Lambowitz A.M. Toro N. J. Mol. Biol. 2003; 327: 931-943Crossref PubMed Scopus (36) Google Scholar, 17Martínez-Abarca F. García-Rodríguez F.M. Toro N. Mol. Microbiol. 2000; 35: 1405-1412Crossref PubMed Scopus (58) Google and are The RmInt1 IEP maturase generated by For the of the two of to the 5′ and 3′ of the IEP, A 5′ and a 3′ used to the a 5′ and a 3′ used to the The an of and promote the of two conserved at IEP by two through to in The was with and and used to the in which generated The maturase was by the II in vitro In which was from (16Muñoz-Adelantado E. San Filippo J. Martínez-Abarca F. García-Rodríguez F.M. Lambowitz A.M. Toro N. J. Mol. Biol. 2003; 327: 931-943Crossref PubMed Scopus (36) Google Scholar), RmInt1 located at to are by so that the conserved IEP is to The was generated by the resulting from of into F. García-Rodríguez F.M. Toro N. Mol. Microbiol. 2000; 35: 1405-1412Crossref PubMed Scopus (58) Google Scholar). on both RNA and reactions as described previously (16Muñoz-Adelantado E. San Filippo J. Martínez-Abarca F. García-Rodríguez F.M. Lambowitz A.M. Toro N. J. Mol. Biol. 2003; 327: 931-943Crossref PubMed Scopus (36) Google Scholar). on a RNA was from of S. meliloti or RmInt1 in of with they an of The and for at in of a by of at for the in the by of The was in of water and with of The RNA was with of a of and with of a of the RNA was with of and The RNA was with and in of For the shown in RNA with a few of a of was to the followed by the described or the was with and of for on and with of was as described previously (16Muñoz-Adelantado E. San Filippo J. Martínez-Abarca F. García-Rodríguez F.M. Lambowitz A.M. Toro N. J. Mol. Biol. 2003; 327: 931-943Crossref PubMed Scopus (36) Google Scholar). The was on a of and for in a at A with two to and and a the was For the shown in the two of the and The was with of and The resulting was in water and with of and The was with of a and The was on for and at for The was with followed by with of The resulting was in water and with of followed by and and strand was by of cellular RNA or an of and of to a sequence from the 5′ of RmInt1 in the presence of The was first at for for to being on for was triggered by of first strand reverse transcriptase of of and of II reverse transcriptase for of the for was for at of that reaction was used as a in the with of to a sequence from the 3′ of of of and of in the step at for was at at at and a at for of was on a in and sequence of the products was by of the from The as the but previously ligated RNA to linear intron molecules that a 5′ of RNA or of from intron to a and of RNA in a of and the was at for RNA was by for and the reaction and followed by for at in of and used for reverse RT RT activity in was as previously described (16Muñoz-Adelantado E. San Filippo J. Martínez-Abarca F. García-Rodríguez F.M. Lambowitz A.M. Toro N. J. Mol. Biol. 2003; 327: 931-943Crossref PubMed Scopus (36) Google Scholar). RmInt1 in as by the RmInt1 has a bulged A located of the 3′ splice we expected the intron to splice as a lariat in vivo. we the RmInt1 splicing reaction by a to a sequence located from the 5′ of the intron (16Muñoz-Adelantado E. San Filippo J. Martínez-Abarca F. García-Rodríguez F.M. Lambowitz A.M. Toro N. J. Mol. Biol. 2003; 327: 931-943Crossref PubMed Scopus (36) Google Scholar). RNA and from S. meliloti intron two products with that by one nucleotide to the expected for the 5′ of the that the is a in both RNA and of two products detected with RNA or from the intron which has a in the conserved pairing of domain and we that the and products from excised intron RNA data not by it to the of the excised intron it is to that on a circular intron has been reported to result in two products that by the being the one (14Murray H.L. Mikheeva S. Coljee V.W. Turczyk B.M. Donahue W.F. Bar-Shalom A. Jarrell K.A. Mol. Cell. 2001; 8: 201-211Abstract Full Text Full Text PDF PubMed Scopus (44) Google Scholar). transcriptase is known to through the 2′-5′ in a RNA Nucleic Acids Res. 1995; PubMed Scopus Google Scholar), whereas with lariat the through the 2′-5′ (14Murray H.L. Mikheeva S. Coljee V.W. Turczyk B.M. Donahue W.F. Bar-Shalom A. Jarrell K.A. Mol. Cell. 2001; 8: 201-211Abstract Full Text Full Text PDF PubMed Scopus (44) Google Scholar). one of data that in to intron RmInt1 be excised in vivo as intron circles. of RmInt1 Excision in by and the of RmInt1 excision RNA and by reverse transcription with located of the 5′ splice followed by with and the located 5′ of the branch The products a 5′ and the products on a RNA used two products of and in the intron or the which has a to in the RT domain and linear introns The was also in the that it was not from the as shown this from the RNA the is that expected from a circular RNA in which the first of the intron is to the last residue, was RNA from the was used or reverse transcription was and These indicate that the of RNA is from an excised intron RNA The products also detected of from the intron used with the data reported the was not used This was in from the RT that has of the splicing activity in the absence of reverse transcription and the a of was detected in intron but was not in the splicing These to indicate the of DNA molecules generated from intron circles. In to the in the but not in the RNA we detected of nt, is that expected from a consisting of an intron lariat in which the branching point is the bulged A in This was in both intron and the RT but was in the or reverse transcription was and These that the of is generated from the of a second intron RNA excision RmInt1 Excision in and products from both RNA and and generated by the from intron and RT and RNA or of the 5′ exon to of the of the A result was by from a in the absence of reverse transcription from RT we that as this is generated from intron RNA precursor The products of generated from intron RNA also and of from the RT from intron RNA circles and the expected point of circular ligation between the first and last residues of the intron whereas the to a with an at the of the same was from intron of to the intron circle whereas the an findings are with the of intron RNA circles in vivo. The of in the absence of reverse transcription and intron used was also and six the same sequence as the of generated from the RNA We the molecules as that in to RmInt1 circular exist DNA circles. In the products of and and generated by of from intron or RT be not the products and of from intron the expected between the bulged A of and the first of the intron that they from a lariat whereas the two to intron circles. that in the products from lariat intron was of adenosine of as reverse transcriptase the nucleotide (7Vogel J. Börner T. EMBO J. 2002; 21: 3794-3803Crossref PubMed Scopus (72) Google Scholar, J. Börner T. Nucleic Acids Res. PubMed Scopus Google Scholar). with the of the in the RT the sequence expected from a lariat findings that excision of RmInt1 in vivo intron lariat as well as RNA and DNA circles. RmInt1 RNA by has been that intron circle formation by intron aI5γ result from the release of the 3′ exon by free 5′ exon molecules from the exon reaction reopening) (14Murray H.L. Mikheeva S. Coljee V.W. Turczyk B.M. Donahue W.F. Bar-Shalom A. Jarrell K.A. Mol. Cell. 2001; 8: 201-211Abstract Full Text Full Text PDF PubMed Scopus (44) Google Scholar). RmInt1 circles also be by ligation of the of a previously linear intron by the hydrolytic pathway in which the first splicing step in the formation of a linear intron-3′ exon intermediate. for the presence of in the RNA the RNA was RNA to reverse transcription with and with and (7Vogel J. Börner T. EMBO J. 2002; 21: 3794-3803Crossref PubMed Scopus (72) Google Scholar). the resulting products from both RNA and form a in they from the shown in sequence of from RNA and from a of molecules in which the 5′ exon was ligated to the 3′ of the intron or to one of the the 3′ splice of Moreover, the 5′ of the RNA precursor seems to be nucleotide other the intron by 5′ and 3′ exon presumably intron RNA precursor whereas the two to molecules at the 5′ or 3′ These are with the that the release of the 3′ exon be the first step for RmInt1 circle The for RmInt1 and group II the RT domain is followed by a domain X, the maturase (RNA splicing) activity. the of the maturase domain of RmInt1 in the splicing reaction in vivo and intron circle we the which has two residues of the conserved and and the in which the conserved is by an and both RNA and indicate that the splicing and Moreover, the from RT activity to that of the RT with the of to form active expected from mobility is in the not These findings indicate that as expected the RmInt1 maturase domain is required for the splicing reaction in vivo, and also for RNA intron circle shown in the point but not splicing the of have RT activity and the has mobility not from this an to intron circle formation the of products from the lariat and intron circles between the and the Moreover, the of the to intron circles appears to be one nucleotide that from the This was by the which an at the circle ligation The from lariat to intron circles only a of the for the lariat leading to a of lariat intron in the this the RNA from and the step also by In the case of the intron the from the lariat was the in both the and the of the whereas in the of the of products from lariat and intron circles was in the last DNA circles detected transcription and the in the the in the and in both and was from intron circles. These findings are with a from lariat to intron circle in the this was by the that in the RNA we only detected intron circles by we not the from the A is that the lariat RNA be or the in some the RNA this we also RNA that specific of or of shown in the use or or in the RNA to both lariat and circles in the RNA expected the from lariat molecules is the predominant in the intron but the from intron circles was the in the findings indicate that the maturase domain of the IEP controls the balance between lariat and RNA intron circle production. We that S. meliloti group II intron RmInt1 is excised in vivo both as intron lariat and intron circles and that the maturase domain of the IEP is not only required for intron RNA but a in the mechanism for intron RmInt1 as a but as in and we that excision of the RmInt1 intron from RNA precursor both lariat and RNA circular RNA intron circles are also by a in the RT active and by the maturase introns are not mobile and intron we that products in which the intron 5′ and 3′ are not generated by transcription of intron but the presence of RNA circles resulting from intron be that RmInt1 also RNA circles in a not which to the Our data also that not only RNA but DNA circular molecules are as well by the RmInt1 the molecules are generated through reverse transcription of the RNA circles by the IEP remains to be The circle in vitro by the yeast intron seems to result from formation a 2′-5′ bond and this may also be the case for RmInt1 RNA circles. Although the of intron circle lariat be from is to be by the of reverse transcriptase to through 2′-5′ on are of or (14Murray H.L. Mikheeva S. Coljee V.W. Turczyk B.M. Donahue W.F. Bar-Shalom A. Jarrell K.A. Mol. Cell. 2001; 8: 201-211Abstract Full Text Full Text PDF PubMed Scopus (44) Google Scholar, Nucleic Acids Res. 1995; PubMed Scopus Google Scholar), the that the lariat is the predominant excision of intron The of lariat intron molecules in the RNA but not the circular forms on the use of specific of or of proteins the RNA In addition, the of the and the the step that the lariat associated with the IEP and the circles have These may be by both the of of the excised intron molecules and a interaction of forms with the The of the RmInt1 IEP in and Excision as in in which the conserved residues of the RmInt1 IEP maturase domain by two residues formation of both RNA circles and intron that the maturase domain is required for the two excision mechanisms in vivo. The may RNA splicing by some of the of the IEP with the intron RNA or by the of the In the maturase splicing was to of the and was a of intron lariat This may also the IEP interaction with the intron which in may the intron excision Our that the maturase domain controls in some the balance of intron excision as lariat or intron circles. Moreover, we that some of the products presumably from RmInt1 RNA circles an at the circle ligation those products molecules from the maturase The of at the circle ligation point was for in circular aI5γ RNA (14Murray H.L. Mikheeva S. Coljee V.W. Turczyk B.M. Donahue W.F. Bar-Shalom A. Jarrell K.A. Mol. Cell. 2001; 8: 201-211Abstract Full Text Full Text PDF PubMed Scopus (44) Google and the residues to been by the reverse transcriptase used in the it the putative 2′-5′ phosphodiester However, a is also at of the RmInt1 3′ the in the RmInt1 RNA circular molecules from cleavage at in the 3′ being an of the reverse transcription reaction. This of that the maturase would to determine the of excision at the 3′ splice junction intron circle The of RmInt1 Excision as in RmInt1 is also to form RNA intron circles in the absence of protein by a sequence located to the correct 3′ splice in the 3′ exon (22Costa M. Michel F. Molina-Sánchez M.D. Martinez-Abarca F. Toro N. Biochimie (Paris). 2006; 88: 711-717Crossref PubMed Scopus (17) Google Scholar), the IEP seems to promote the correct interaction in vivo. of RNA RNA a of with a of 3′ exon cleavage located at or next to the 3′ splice site. in vivo the presence of the IEP seems to that cleavage at the correct 3′ splice or at of the 3′ exon intron RNA circle at or as in vitro (22Costa M. Michel F. Molina-Sánchez M.D. Martinez-Abarca F. Toro N. Biochimie (Paris). 2006; 88: 711-717Crossref PubMed Scopus (17) Google Scholar). the release of the 3′ which may result from a trans-splicing reaction as for intron aI5γ (14Murray H.L. Mikheeva S. Coljee V.W. Turczyk B.M. Donahue W.F. Bar-Shalom A. Jarrell K.A. Mol. Cell. 2001; 8: 201-211Abstract Full Text Full Text PDF PubMed Scopus (44) Google Scholar), the 2′-OH of the attack the 5′ splice the 5′ exon and an intron Excision of group II introns as circles been shown to in vivo for yeast introns like aI2 (14Murray H.L. Mikheeva S. Coljee V.W. Turczyk B.M. Donahue W.F. Bar-Shalom A. Jarrell K.A. Mol. Cell. 2001; 8: 201-211Abstract Full Text Full Text PDF PubMed Scopus (44) Google and for some plant mitochondria introns (15Li-Pook-Than J. Bonen L. Nucleic Acids Res. 2006; 34: 2782-2790Crossref PubMed Scopus (58) Google Scholar), but is reported for the first in this for a mobile bacterial group II is that this of excision is in it and we it as that it be to have a with to intron and We and Michel for and of the We are to for