key: cord-0966029-8oy18apx
authors: Chetverin, Alexander B; Chetverina, Helena V; Demidenko, Alexander A; Ugarov, Victor I
title: Nonhomologous RNA Recombination in a Cell-Free System: Evidence for a Transesterification Mechanism Guided by Secondary Structure
date: 1997-02-21
journal: Cell
DOI: 10.1016/s0092-8674(00)81890-5
sha: 9272fcf42ca7c921487b212793982f4d971df58b
doc_id: 966029
cord_uid: 8oy18apx

Abstract Extensive nonhomologous recombinations occur between the 5′ and 3′ fragments of a replicable RNA in a cell-free system composed of pure Qβ phage replicase and ribonucleoside triphosphates, providing direct evidence for the ability of RNAs to recombine without DNA intermediates and in the absence of host cell proteins. The recombination events are revealed by the molecular colony technique that allows single RNA molecules to be cloned in vitro. The observed nonhomologous recombinations are entirely dependent on the 3′ hydroxyl group of the 5′ fragment, and are due to a splicing-like reaction in which RNA secondary structure guides the attack of this 3′ hydroxyl on phosphoester bonds within the 3′ fragment.

Escherichia coli cells. There have also been indications that RNA can probably recombine in the in vitro Q␤ Intermolecular RNA recombination is a wide-spread replicase reaction (Biebricher and Luce, 1992) . phenomenon reported for a variety of animal, plant, and

The results show that RNA recombination did occur bacterial RNA-containing viruses (Lai, 1992 ; Chetverin in such a purified system, thereby providing direct eviand Spirin, 1995) . It has played an important role in the dence for the ability of RNA molecules to recombine evolution of RNA viruses (Lai, 1992) and might have without DNA intermediates and in the absence of host resulted in the exon-intron structure of modern genes cell components. Most of the recombinant RNAs were (Gilbert, 1986) . Viral RNA recombination is believed to be of a purely nonhomologous type, and the mechanism of quite different from the process of RNA trans-splicing, their generation was entirely different from copy choice. since it is not site-specific (Lai, 1992) and is apparently associated with the RNA synthesis by viral replicases (Kirkegaard and Baltimore, 1986) .

Results Since recombinations occur rarely and randomly, they are only observed if their products can be amplified, Strategy of RNA Recombination Experiments in the Cell-Free Q␤ System cloned, and distinguished against a large background of nonrecombinant RNAs. So far, RNA recombination In addition to the genomic RNA of Q␤ phage, Q␤ replicase can exponentially amplify in vitro a variety of small has been studied in vivo, almost exclusively on viruses or their defective interfering (DI) particles amplified in RQ RNAs that are natural phage satellites (Chetverin et al., 1991) , termed so for being Replicable by Q␤ repli-living cells under selective conditions. Thus, only those recombination events that render the particles infec-case. These subgenomic RNAs are the most suitable for in vitro studies on RNA recombination, since they tious have been detected. It is therefore not surprising that most reports on RNA recombination concern ho-are more efficient templates for Q␤ replicase than the genomic Q␤ RNA. Many RQ RNAs have themselves been mologous recombination, i.e., recombination between similar segments of related RNAs (Lai, 1992) . The homol-generated in vivo by intermolecular recombination, and their sequences can accommodate foreign inserts with-ogous recombination can most readily be accounted for by a copy choice (template switch) reaction mechanism out losing the ability to replicate (Chetverin and Spirin, 1995) . In contrast to viral genomes, however, RQ RNAs (Cooper et al., 1974) , which implies that recombination between RNA molecules occurs via copying of the 3Ј lack genetic markers facilitating the selection of recombinant molecules against a large excess of parent RNAs. portion of the first molecule by the replicase that eventually jumps, together with the nascent strand, to a homol-To overcome this problem, we devised the following experimental scheme ( Figure 1A ). ogous region of the second molecule where it resumes RNA synthesis. This hypothesis has become commonly Instead of the full-sized replicable molecules, the complementary strand does not appear unless recombination occurs. This eliminates the background hybridization problems even when relatively high concentrations of the nonreplicating RNA fragments are used in experiments. The recombinant molecules carrying foreign inserts are detected by hybridizing the colonies with labeled oligonucleotides complementary to the foreign sequences.

In this work, we used the 5Ј and 3Ј fragments of the (Ϫ)-strand of RQ135 Ϫ1 RNA, a 134 nt long subspecies of RQ135 RNA, which is one of the most potent Q␤ replicase templates (Munishkin et al., 1991) . Earlier, we found that this RNA can serve as a vector for amplification of short foreign inserts (V. I. U. and A. B. C., unpublished data). Since RNA recombination was expected to occur via the copy choice mechanism, the foreign extensions of the fragments were intentionally made almost entirely homologous to each other in order to facilitate the process ( Figure 1B ).

As expected, RNA colonies almost never grew when up these fragments cannot replicate. A few colonies that (B) Sequences of foreign extensions (white letters on a black backoccasionally appeared in these samples (and also in the ground) of the 5Ј and 3Ј fragments of the (Ϫ)-strand of RQ135 Ϫ1 samples where no RNA was added) did not hybridize RNA. A few base differences between the extensions (lowercase with an oligonucleotide targeted against the foreign exletters) serve as sequence markers. The fragment extensions consist of polylinker sequences providing for further manipulations of their tensions (cf. Figure 2C) , and seemed to result from the structure at the DNA level. airborne contamination by the wild-type RQ135 RNA (Chetverin et al., 1991; Chetverina and Chetverin, 1993) .

On the contrary, up to 150 RNA colonies could be seen when 2.7 ϫ 10 9 molecules (4.5 fmol) each of the scheme employs the mutually supplementing 5Ј and 3Ј 5Ј and 3Ј fragments were incubated together in a 10 l fragments of an RQ RNA that are prepared by transcripaliquot ( Figure 2A ), indicating that replicable RNAs are tion and that carry, at the truncated ends, artificial forformed by recombination between the fragments. Coloeign extensions whose sequence and secondary strucnies did not appear in control experiments where the ture may vary. Q␤ replicase can start at the 3Ј end of RNA fragments were replaced with the plasmids used the 3Ј fragment and copy it together with the foreign for their synthesis (data not shown), which excludes the extension, producing a complementary strand. The repossibility that recombination is due to traces of DNA action ceases after the first round, and RNA amplificain the fragment preparations. The number of colonies tion will not occur unless a complete replicable RNA is corresponded to an ‫01ف‬ Ϫ7 recombination frequency at formed by recombination between the fragments. If the the fragment concentration of 0.45 nM. The actual frerecombination occurs within the extensions, the product quency might have been higher, since only a fraction will comprise the original RQ RNA carrying the corre-‫)%01ف(‬ of RQ135 RNA molecules are capable of producsponding foreign insert.

ing colonies (Chetverina and Chetverin, 1993) and not The reaction products are then analyzed by the molecevery recombination product can be replicable. In more ular colony technique. According to this method, an RNA dilute samples, the yield of recombinant molecules desample is spread over a thin Q␤ replicase-containing creased roughly proportionally to the product of the agarose layer, which is then covered by a nylon memfragment concentrations, as expected for a bimolecular brane impregnated with all four ribonucleoside triphosreaction; that is to say, the recombination frequency phates. If there are replicable RNAs on the interface was proportional to the RNA concentration. At a higher between the agarose and the membrane, RNA colonies concentration of the fragments, the number of colonies appear within an hour, each colony containing up to 10 12 became too large to be counted. A large proportion of copies of a single progenitor template (Chetverina and the recombinant RNA colonies were hybridizable with Chetverin, 1993). While the replicable recombinant molan oligonucleotide targeted to the foreign extensions ecules will produce colonies, the nonreplicable frag-(cf. Figures 2B and 2C) , indicating that the extension ments of the original RQ RNA will not. This procedure sequences were incorporated into the recombination ensures positive selection of the recombination prodproducts. In additional experiments, we found that the ucts and allows the number of recombinant molecules complementary versions of the 5Ј and 3Ј fragments reto be directly assessed by simply counting the RNA combined at a similar frequency (data not shown), meancolonies hybridizable with the labeled 5Ј fragment. The ing that the ability to recombine does not depend on strand polarity. fragment is not copied by Q␤ replicase and thus its Shown in the order of appearance: the mixture incubated for 40 min at 55ЊC in the presence of 0.5 M KCl; the same sample brought into a 1 mM Na-EDTA solution by passing through a Sephadex G-25 microcolumn; the desalted sample after heating for 2 min at 96ЊC; the mixture incubated for 40 min at 37ЊC in the presence of 0.5 M KCl. The last two lanes contain separate fragments. (B) Colonies produced by the indicated number of molecules of each of the mixed 5Ј and 3Ј fragments that were annealed by heating to 96ЊC for 2 min in a Mg 2ϩ -free buffer containing 0.1 M KCl and then allowing the sample to cool down to 20ЊC during 1 hr (upper row), or preincubated in the same buffer at 20ЊC during 1 hr (lower row). Hybridization with the foreign extension-targeted oligonucleotide.

interaction between the fragments that precedes the creation of a recombinant molecule. Indeed, the 5Ј and 3Ј fragments were capable of forming a noncovalent associate (heteroduplex) stable enough to survive electrophoresis under nondenaturing conditions and gel- To obtain a deeper insight into the mechanism of RNA (B) Colonies produced by 10 10 molecules each of the mixed 5Ј and recombination, we sequenced a total of 17 recombinant 3Ј fragments, or the 3Ј fragment only (bottom row), preincubated in clones generated upon incubation of 10 10 each of the 5Ј the absence or in the presence of 10 mM MgCl2 during the specified time period. Hybridization with the labeled 5Ј fragment. and 3Ј fragments in the absence ( Figure 4A ) or in the (C) Same as (B), but hybridization with a labeled oligonucleotide presence ( Figure 4B ) of Mg 2ϩ . To this end, the total RNA targeted against the foreign extensions.

was extracted from an agarose lawn containing more than 100 RNA colonies, subjected to reverse transcription and polymerase chain reaction (RT-PCR) amplifica-The yield of recombinant molecules increased when the recombining fragments were preincubated up to 2 tion, and the obtained DNA fragments were cloned in a plasmid vector. The plasmids purified from randomly hr prior to contacting them with Q␤ replicase (Figures 2B and 2C) , suggesting that the process of recombination picked clones were sequenced by the dideoxy chain termination method. Since each RNA colony contained includes a slow step occurring independently of the replicase action. Since this step does not depend on the at least 10 11 molecules (as estimated by normalizing the autoradiography spots to a known amount of RQ135 presence of Mg 2ϩ ( Figure 2B ), it could be a noncovalent RNA hybridized with the same probe), the extracted 5Ј and 3Ј fragments. This observation strengthens the above conclusion that these molecules were formed by material contained at least 10 13 recombinant RNA molecules, or 10 3 times as many as the number of the added genuine intermolecular recombination. (ii) In most cases, crossover sites lay within the foreign extensions and 5Ј and 3Ј fragments. Therefore, any subsequent recombination between the fragments during reverse tran-were precisely identified (recombinants A1-A5, B1, and B2). All these recombinants were of a purely nonhomolo-scription or PCR amplification could not appreciably contribute to the cloned sequences. In control experi-gous type. (iii) Surprisingly, there were no recombinants of a bona fide homologous type expected of copy ments, the fragment mixture was subjected to the same procedure, except the colony growth was omitted and choice: the nearly perfect homology between the foreign extensions was invariably ignored. Crossover sites were the fragments were mixed either before ( Figure 4C ) or after the reverse transcription step.

sometimes found within very short (1-3 nt) homologous stretches. Almost all of these crossovers involved the The following observations could be made from inspection of the Q␤ replicase-amplified sequences vector portion of either the 5Ј or the 3Ј fragment (recombinants B4-B6). Nevertheless, the recombinant mole-shown in Figures 4A and 4B . (i) None of the sequences coincided with the primary structure of wild-type cules retained the UCCCU/AGGGA helix of the vector RNA ( Figure 5 ), indicating that the relatively large propor-RQ135 Ϫ1 RNA (Munishkin et al., 1991) , and almost all included, at least partly, the foreign extensions of the tion of this type of recombinants might have resulted sites lay within the short stretches of homology between the fragments (see item [iii] ). Although the presence of Mg 2ϩ during the preincubation step did not noticeably influence the recombination frequency ( Figure 2B ), the latter observation suggests that it might have affected the recombination mechanism. (vi) Crossover sites were not evenly distributed along the 3Ј fragment sequence. As seen from Figure 5 , they tended to concentrate at or close to the 5Ј base of a helix of the putative intermolecular structure formed by the 3Ј and 5Ј fragments. (vii) Eight of the seventeen clones apparently were formed by adding the intact 3Ј terminus of the 5Ј fragment to the beginning of sequence UCUAGA in the 3Ј fragment (recombinants A1-A3, B1). This crossover site was located at the 5Ј base of the helix closest to the 3Ј terminus of the 5Ј fragment ( Figure 5 ). Although five of these clones were identical and had been isolated from one RNA pool (recombinant A1), the probability that all of them were descendants of a single original recombinant molecule was negligibly low, less than 10 Ϫ6 , as expected from a random distribution of the material contained in at least 100 RNA colonies among the 9 clones isolated from that pool. Moreover, a clone with the same nucleotide sequence had been independently isolated from another RNA pool (recombinant B1), and there were three more clones in which recombination had occurred in a close proximity to this site (recombinants A4, B2, B3).

To check the possibility that the relative abundance of these clones was due to selective amplification of ture also produced recombinant sequences. However, they migrated as a single band during gel electrophoresis, unlike the heterogeneous RT-PCR products synfrom their selective amplification at the expense of those thesized subsequent to the amplification of the remolecules in which this helix was distorted. (iv) Contribucombinant RNAs by Q␤ replicase (data not shown). tion of the 5Ј and 3Ј fragments to the recombinant se-Furthermore, their sequencing revealed a precisely hoquences was markedly asymmetric: 12 of the 17 clones mologous recombination expected for the copy choice included the intact 5Ј fragment (recombinants A1-A5, mechanism ( Figure 4C ). Since such sequences did ap-B1, B2), whereas the complete 3Ј fragment was prepear only if the 5Ј and 3Ј RNA fragments were mixed served in only one case (recombinant A5). The latter before the cDNA synthesis step, they seemed to result recombinant molecule was formed by end-to-end joinfrom a template switch occurring during reverse traning of the two fragments. Some of the recombinants scription (Negroni et al., 1995) . It follows that the RNA with the intact 5Ј fragment contained one to three extra fragments are prone to a copy choice reaction, provided nucleotides upstream from crossover sites (A2, A3, A5, that an appropriate enzyme is used, and that the fre-B2). These nucleotides were possibly added to the 3Ј quency of template switch by Q␤ replicase, if any, is terminus of the 5Ј fragment during its synthesis by T7 much lower than that by AMV reverse transcriptase used RNA polymerase (Milligan et al., 1987) or during the in these experiments. subsequent incubation with Q␤ replicase (Biebricher and Luce, 1992). (v) All recombinants in which the 5Ј fragment was abbreviated were generated in the experi-Role of the 3 Hydroxyl Groups ment in which the fragments were preincubated in the in RNA Recombination presence of Mg 2ϩ prior to contacting them with Q␤ repli-

The observed features of the recombinant sequences case (recombinants B4-B6 and, possibly, B3). Interestingly, these were the recombinants whose crossover indicate that the 3Ј terminus of the 5Ј fragment may the 5Ј and 3Ј fragments one of which was unmodified while the other stained with toluidine blue. Numbered letters above the gel refer to was either unmodified at the 3Ј terminus (... NPNOH) or modified by the type of recombinant sequence (Figure 4) , numbers at the bottom periodate oxidation and treatment with aniline (...NP) and subserefer to the clone number. quent dephosphorylation (...NOH). The mixtures were preincubated (A) Samples containing 1-2 g of transcripts synthesized by T7 RNA for 1 hr at 20ЊC in a Mg 2ϩ -free buffer. polymerase from the cloned recombinant plasmids digested at site (B) Colonies produced by a mixture of 6 ϫ 10 9 molecules each of SmaI.

the 3Ј phosphoryl 5Ј fragment and the unmodified 3Ј fragment after (B) Products of 10 min Q␤ replicase reactions initiated with 1/400 preincubation for 1 hr at 37ЊC and indicated Mg 2ϩ concentration of the amount of RNA applied to gel A. The faster migrating band with or without 0.01 units of CIP. Prior to colony growth, all the is a double-stranded RNA, the slower migrating bands correspond samples received a sufficient amount of EDTA to chelate free Mg 2ϩ to single strands and partial duplexes. Lane marked "0" contains and were then heated for 2 min at 96ЊC to inactivate phosphatase. material from the reaction where no RNA was added.

RNA colonies were detected by hybridization with the foreign extension-targeted oligonucleotide.

(C) Silver-stained (Igloi, 1983) products of 10 min Q␤ replicase reacbe involved in the recombination process. Since the tions initiated with 2 ng of a recombinant RNA (type A5, clone 3, cf.

fragment contains the hydroxyl group at the terminal 3Ј the presence of the 3Ј hydroxyl is essential. To this end, the 3Ј terminal ribose was oxidized with periodate, and the terminal nucleoside dialdehyde was eliminated role in the recombination process. Similar results were by treatment with aniline. This resulted in a 1 nt shorter obtained for the complementary versions of the fragfragment that bears the phosphoryl group at the 3Ј end ments (data not shown), indicating that the mechanism (Steinschneider and Fraenkel-Conrat, 1966b) . Figure 7A of recombination is the same for the strands of either (upper row) shows that such a modification of the 5Ј polarity. fragment reduced the number of RNA colonies to zero, Figure 7B shows that recombination with the 3Ј phosand that the removal of the 3Ј terminal phosphate with phoryl 5Ј fragment could also be switched on by the calf intestine alkaline phosphatase (CIP) restored the Mg 2ϩ present in the preincubation mixture at a high ability of the fragment to recombine. Moreover, recombiconcentration, although to a lesser extent than by phosnation could be switched on by adding the phosphatase phatase. This effect can be explained by exposure of a directly into the sample containing a mixture of the 3Ј reactive hydroxyl group as a result of a Mg 2ϩ -catalyzed phosphoryl 5Ј fragment and the 3Ј fragment ( Figure 7B ). RNA cleavage, which gives a rationale for the observed In contradistinction, similar modification of the 3Ј frag-Mg 2ϩ -induced change in the sequence of some recombiment had no appreciable effect on the number of RNA nant molecules ( Figure 4B ). colonies and the number reduced, rather than increased, upon phosphatase treatment ( Figure 7A , lower row). A lesser number of colonies in this case correlates Discussion with a decreased capability of a 3Ј truncated recombinant RNA to be amplified by Q␤ replicase ( Figure 7C ).

In additional experiments, we found that other 3Ј terminal modifications eliminating the 3Ј hydroxyl group, Although intermolecular RNA recombination was first reported more than 30 years ago (Hirst, 1962; Ledinko, such as periodate oxidation and biotinylation, also annulled the recombination potential of the 5Ј fragment, 1963), its mechanism remains obscure due primarily to the absence of adequate in vitro models. Biebricher but not of the 3Ј fragment. These observations suggest that the 3Ј hydroxyl of the 5Ј fragment plays a crucial and Luce (1992) reported reproducible synthesis of a recombinant RQ RNA in the in vitro Q␤ replicase reac-heterotetramer, i.e., the translation factors EF-Tu and tions initiated with a different RQ RNA. However, since EF-Ts, and ribosomal protein S1 (Chetverin and Spirin, the synthesis was only observed under conditions selec-1995). Moreover, experiments in the cell-free system tive for amplification of the recombinant RNA, and since led us to an unexpected conclusion regarding the RNA the same RNA was studied in their laboratory earlier recombination mechanism that appeared to be entirely (Biebricher et al., 1982) , it is difficult to rule out the different from copy choice. possibility that their reactions were contaminated with According to the copy choice mechanism, a nascent the already existing recombinant RNA, given the ability strand synthesized by Q␤ replicase on the 3Ј fragment of Q␤ replicase to amplify even single airborne RQ RNA would serve as a primer for RNA synthesis on the 5Ј molecules (Chetverin et al., 1991; Chetverina and Chet- fragment. In order to facilitate replicase switching beverin, 1993).

tween the recombining fragments, the fragments were Here, we describe an efficient cell-free assay system provided with homologous foreign extensions (Figure for RNA recombination; the sensitive molecular colony 1B). Thus, sequence crossovers were expected to occur technique is used to observe individual recombination between the extensions resulting in molecules with chievents in a reaction containing only homogeneous Q␤ meric foreign inserts in which the extension sequence replicase, pure ribonucleoside triphosphates, and two of the 5Ј fragment was partially replaced from the 3Ј types of purified RNA molecules. This is a cell-free verside by the homologous segment of the 3Ј fragment, as sion of the Q␤ phage system whose ability to perform it was observed for the reverse transcriptase-generated RNA recombinations in vivo was demonstrated earlier recombinants ( Figure 4C) . Surprisingly, none of the Q␤ (Munishkin et al., 1988; 1991; Palasingam and Shaklee, replicase-amplified recombinants displayed the ex-1992). The recombination substrates used in these expected structure ( Figures 4A and 4B ). The homology periments were fragments of a Q␤ phage satellite RNA between the extension sequences was invariably ig- (Chetverin et al., 1991) , which itself is a product of an nored. Moreover, in most cases, a part of the 3Ј fragment intermolecular recombination that had occurred in vivo extension was attached to, rather than did replace, the (Munishkin et al., 1991) . The fragments were derived extension of the 5Ј fragment, which was preserved intact from the RNA sequence by breaking it at the junction in most of the recombinant molecules. In the extreme site between the in vivo duplicated segments (Chetverin case, the recombinant molecule was generated by the et al., 1992), i.e., these RNA molecules may resemble end-to-end joining of the intact 5Ј and 3Ј fragments the natural recombination substrates. The experimental (recombinant A5). conditions (temperature, pH value, concentrations of Figure 7 demonstrates that recombination between mono-and divalent cations) were close to the normal the RNA fragments was entirely dependent on the availphysiological conditions; this suggests that the results ability of the 3Ј hydroxyl group at the 5Ј fragment termimay be physiologically relevant.

nus. This result not only rules out the possibility that Of course, the reported system has certain limitations.

the classical copy choice mechanism contributes to the Thus, only those recombinants could be detected and observed recombination events, but also rejects any studied that were amplified by Q␤ replicase. The results variation of this mechanism in which an RNA fragment show that in the isolated recombination products the serves as a primer for its own extension on another crossover sites were exclusively located within or close template. For example, one might argue that elimination to the introduced foreign sequences, and that all the of the terminal hydroxyl simply prevents extension of the recombinants retained the secondary structure ele-5Ј fragment on the 3Ј fragment complementary strand, ments of the original RQ RNA vector, suggesting that which can be synthesized by Q␤ replicase. This could selection pressure rejects all molecules in which the not prevent, however, a reciprocal extension of the structure of the vector is altered.

newly synthesized strand on the 5Ј fragment, since its At the same time, a foreign insert as long as 52 nt 3Ј terminus was not modified, and only a slight (some (Figure 4 , recombinant A5) did not eliminate the ability 2-fold) reduction in the recombination frequency could of the RQ135 RNA vector to replicate (Figure 6) , implying be expected. The absence of recombination in this case that the system allows the recombination between for-(as well as in the reciprocal experiment with the compleeign sequences of this size to be studied. In any case, mentary versions of the fragments) means that the newly the selection pressure in the in vitro system is much synthesized strand lacked its recombination partner, less pronounced than in any in vivo system, since there which must be the missing complementary copy of the is no need for the ability of the recombinant RNAs to 5Ј fragment. encode proteins, to be encapsidated, or to infect the It follows that the recombination occurs through a cell. In addition, the in vitro system allows the recombireaction between the fragments of the same polarity. nation mechanism to be explored by changing the reac-This is further supported by the observation that prelimition conditions, varying composition of the reaction menary annealing of the 5Ј and 3Ј fragments increased the dium, utilizing modified RNAs or nucleotides, and by recombination frequency by orders of magnitude (Figure step-by-step monitoring of the recombination process.

3). Together with the distribution of crossover sites In fact, the system presents a remarkable example of within the putative secondary structure of the annealed quantitative biochemical assay for reactions occurring fragments ( Figure 5) , the latter observation also indibetween single molecules.

cates that RNA recombination is a secondary-structure This system allowed us to demonstrate directly that guided process. RNAs can recombine without the formation of DNA inter-

The most likely mechanism for RNA recombination, mediates and without the assistance of host cell proteins, except possibly those included in the Q␤ replicase totally consistent with the above observations, can be pictured as follows. First, the recombining molecules mentioned previously, but at that time no experimental form a noncovalent complex by means of intermolecular proof in support of this mechanism existed (Lai, 1992) . hydrogen bonding. Then, the 3Ј terminal hydroxyl of Such a proof is provided by the results reported here. the 5Ј partner attacks a phosphate group in the sugar-At the same time, only nonhomologous recombinations phosphate backbone of the 3Ј partner, resulting in the were observed in this cell-free system, contrasting the ligation of one part of the target RNA with a concomitant in vivo observations that predominantly revealed homolrelease of the other. Of course, the 3Ј terminal hydroxyl ogous recombination (Lai, 1992) . This discrepancy may of the 3Ј fragment could also attack the 5Ј fragment, be partially due to selection pressure that rejects in but this would result in a nonreplicable molecule where vivo most nonhomologous recombinants. However, a the fragments are in a wrong order. Obviously, this majority of the poliovirus recombinants appear to be mechanism is very similar to that operating in RNA splichomologous even in the absence of selection (Jarvis ing (Cech and Bass, 1986) . We would like to stress that and Kirkegaard, 1992) , and are presumably generated our results do not rule out in principle other possible via the copy choice mechanism (Kirkegaard and Baltimechanisms, including copy choice. Rather, they show more, 1986). that those mechanisms do not appreciably contribute A clue to the controversy seems to lie in the extremely to the RNA recombination in our cell-free system. different frequencies of homologous recombination in This paper does not address the question of whether poliovirus (6 ϫ 10 Ϫ3 per 1900 nt; Jarvis and Kirkegaard, the observed RNA recombination required any Q␤ repli-1992) and in Q␤ phage from which the cell-free system case activity. At present, it seems unlikely, given the was derived (only 10 Ϫ8 per a similar genome length; known ability of RNA to self-splice and the absence of Palasingam and Shaklee, 1992) . A backward extrapolaindications that Q␤ replicase can ligate RNA fragments.

tion of the frequency of homologous recombination in However, we cannot exclude the possibility that Q␤ rep-Q␤ phage to the RNA concentrations used in our in vitro licase acts as a protein cofactor that binds the RNA experiments would give the value of 10 Ϫ13 per 30 nt. Of fragments and provides favorable conditions for a selfcourse, this is a very rough estimate, but it helps to catalyzed splicing reaction.

demonstrate that even if the copy choice mechanism operated in the in vitro system as efficiently as in Q␤ phage, it would not, except fortuitously, show up on Implications for RNA Recombination In Vivo the million-fold higher background of nonhomologous Although the frequency of nonhomologous recombinarecombinations. tion observed in our in vitro assay system is exceedingly

The following concept helps to reconcile the in vivo low, it may nonetheless be comparable to that of RNA and in vitro data. Most homologous recombinations ocrecombination occurring in vivo during bacteriophage cur by a copy choice mechanism, and their frequency and viral infection. The frequency is 10 Ϫ7 per a 30 nt is a virus-specific parameter, inasmuch as it is deter-RNA segment at the 10 Ϫ9 M fragment concentration and mined by the properties of viral replicases, such as the is roughly proportional to the RNA concentration. A enzyme processivity. At the same time, nonhomologous straightforward extrapolation to the concentration of Q␤ RNA recombinations mainly occur via splicing-like reac-RNA in infected E. coli cells (up to 10 Ϫ5 M; Weissmann, tions, probably nonenzymatic ones, and their frequency 1974) would give the value of 10 Ϫ3 , i.e., similar to the is relatively virus-nonspecific, since it is determined by frequency of the most efficient homologous recombinathe general properties of RNA, such as its secondary tions in vivo (Lai, 1992) . Thus, the efficiency of the nonhoand/or tertiary structure. Given the fact that viruses with mologous recombination reaction does not rule out, and a high overall recombination potential (such as picornamay well argue in favor of, the physiological relevance and coronaviruses) predominantly display homologous of the in vitro reaction.

recombination, one could predict that the relative contri-General properties of the in vitro recombination bution of nonhomologous recombinations to the totality closely resemble those of the viral RNA recombination of recombination events in different viruses would inin vivo. Indeed, RNA recombination in vivo (i) can occur crease as the overall recombination rate decreases, and when one (Kirkegaard and Baltimore, 1986; this largely conforms to the available in vivo data (Lai, et al., 1988) or even both of the recombining molecules 1992). For example, most of the nongenomic (satellite) (Munishkin et al., 1991; Moody et al., 1994; Hajjou et al., RNAs of Q␤ phage, which are less subject to selection 1996) do not replicate; (ii) seems to be guided by the than the genomic RNA, consist of a mosaic of fragments secondary structures formed between RNA molecules of both viral and cellular origin, i.e., they have formed via (Kuge et al., 1986; Romanova et al., 1986; Nagy and nonhomologous recombination (Chetverin and Spirin, Bujarski, 1993; Carpenter et al., 1995); (iii) predominantly 1995) . This hypothesis is further supported by the obseroccurs within unpaired regions of RNA (Romanova et vation that the RNA fragments used in this work proal., 1986; Tolskaya et al., 1987; Carpenter et al., 1995) ; duced only nonhomologous recombinants when ampli-(iv) is promoted by preannealing of the recombination fied by Q␤ replicase, whereas the same fragments partners (Dzianott et al., 1995); and (v) has an RNA conproduced only homologous recombinants when they centration dependence similar to that of the in vitro were copied by reverse transcriptase whose template recombination (Jarvis and Kirkegaard, 1992) . However, switching ability seems to be even higher than that of these properties do not distinguish between the copy poliovirus replicase (Negroni et al., 1995) . choice and splicing-like mechanisms.

Thus, homologous and nonhomologous recombina-The possibility of a breaking and joining mechanism of a trans-splicing type in viral RNA recombination was tions may be brought about by different mechanisms.

Also, they seem to play different roles in biological sys-genomes include the 3Ј terminus of rat tRNA Asp attached to the beginning of the viral genome sequence (Monroe tems. As has been pointed out (Lai 1992) , the homologous recombination may be a mechanism to eliminate and Schlesinger, 1983) . Also, one half of the experimental recombinants between the genome of turnip crinkle replication errors, and this could be particularly advantageous for viruses with large nonsegmented genomes virus and its satellite (sat-D) RNA have been generated by the addition of the 3Ј terminus of sat-D to the 5Ј and for those encoding large polyproteins. Therefore, the elevated capability of template switch may have base of a hairpin in the genomic 3Ј-terminal untranslated region (Carpenter et al., 1995) , which is reminiscent of been acquired by the replicases of picorna-and coronaviruses as a result of evolution. At the same time, the secondary structure requirements for the in vitro recombination. nonhomologous recombinations provide for big and instant evolutionary jumps by enabling a virus to create Thus, the splicing-like mechanism does not seem to be incompatible with the existing in vivo data, and it new genes or to borrow them from other viruses or from the host cell, and they are also responsible for was observed in vitro under conditions not too different from those existing in the cell. We believe that such a the generation of DI genomes that strongly affect viral infection. The low capability of Q␤ replicase to switch mechanism does operate in vivo, bringing about nonhomologous recombination between RNA molecules, both between templates does not allow homologous recombination to be observed in our cell-free system. At the viral and cellular. Although recombination between cellular RNAs has not yet been experimentally demon-same time, it makes the system perfectly suited to studying nonhomologous recombinations.

strated, apparently because of the absence of a suitable experimental system, the sequencing of some RQ RNA molecules has shown that they consist of fragments of RNA Secondary Structure Requirements cellular RNAs only (Munishkin et al., 1991; Moody et al., for Nonhomologous Recombination 1994) . Our data support the hypothesis that the modern In agreement with the proposed splicing-like mechasplicing systems, together with the exon-intron strucnism, most of the crossover sites were located within ture of genes, have evolved from the ability of cellular the putative intermolecular secondary structure close RNA to recombine (Gilbert, 1986) . The described in vitro to the attacking 3Ј terminus of the 5Ј fragment, the attack system could provide for direct evolutionary experioccurring most frequently at the 5Ј base of the nearest ments on this matter. helix ( Figure 5 ). In contrast to the classical splicing reaction, the sites of attack were scattered along the target Experimental Procedures sequence and the reaction was orders of magnitude less efficient. This was not unexpected, since the recom-

The 5Ј and 3Ј fragments were constructed at the DNA level utilizing bining sequences were not selected for the highest perthe selective PCR amplification of the corresponding regions of formance.

plasmid pT7RQ135 Ϫ1 (Ϫ) comprising a T7 promoter/RQ135 Ϫ1 (Ϫ) One interpretation of our data is that nearly any RNA cDNA cartridge inserted into plasmid pUC18 (Morozov et al., 1993) . that positions its 3Ј hydroxyl in the vicinity of a stem/loop The 5Ј fragment was amplified with primer 1 (5Ј-CTGCAGGCATGCA structure will lead to some level of transesterification.

AGCTTAATACGACT-3Ј, partially overlapping the sequence of the Indeed, this interpretation would explain why the recom-T7 promoter and containing site HindIII) and primer 2 (5Ј-AGACTCGA GCTGCAGAAGGGACGCACG-3Ј, complementary to positions 41-52 bination mechanism totally ignores homology between of the RQ135 Ϫ1 (Ϫ) sequence and containing site PstI). The 3Ј fragthe foreign extensions, why a 3Ј hydroxyl can add to a ment was amplified with primer 3 (5Ј-CGCTGCAGCTCGAGTCTAGA 5Ј-terminal triphosphate (recombinant A5), and why the GGATCCTACGAGGGATTTGA-3Ј, matching positions 53-66 of the splicing-like reaction even tolerates addition of CCC to RQ135 Ϫ1 (Ϫ) sequence and introducing the foreign extension) and the 3Ј end of the attacking molecule (recombinant A3). primer 4 (5Ј-CCGCGGATATCGATCCCGGGCTAACAGTG-3Ј, com-Thus, RNAs may be intrinsically recombinogenic under plementary to the 3Ј terminus of the RQ135 Ϫ1 (Ϫ) sequence and containing site SmaI), and then extended at the 5Ј end in two addi-physiological conditions, but the resulting low level of tional PCRs by employing, consecutively, primer 5 (5Ј-AAGCTTAAT nonhomologous recombinants can only be appreciated ACGACTCACTATAGGCGCTGCAGCTCGAGT-3Ј, introducing the T7 when a strong selective advantage allows the recombipromoter and restriction site HindIII) and primer 6 (5Ј-CTGCAGGCAT nant molecules to outreplicate the nonrecombinants.

GCAAGCTTAATACGACT-3Ј, extending the sequence upstream By using the strategy worked out in the in vitro experifrom HindIII), each in combination with primer 4. The amplified fragments, it is now possible to perform direct experiments ments were digested by restriction endonucleases HindIII and PstI (5Ј fragment) or HindIII and SmaI (3Ј fragment), and cloned within on the role of splicing-like reactions in nonhomologous the pUC18 vector between the corresponding sites. The primary virus recombination. At the same time, a body of indirect structures of the resulting constructs were checked by sequencing. evidence for the existence of such a mechanism in vivo RNA fragments used for recombination experiments were synthecan be found in the literature. A hallmark of the splicing sized with T7 RNA polymerase (Noren et al., 1990) after digesting the mechanism is that the 3Ј terminus of an attacking subplasmids at site BamHI located in the pUC18 polylinker downstream strate is preserved within the recombinant sequence.

from PstI (5Ј fragment) or SmaI (3Ј fragment), and purified by electrophoresis through denaturing polyacrylamide gel (Milligan and Uhlen- This is especially easy to discern if the sequence in- beck, 1989 ). cludes the 3Ј terminus of an intact RNA, and such recom-Where indicated, the fragments were modified at the 3Ј end as binants have been reported for viruses of all types of follows. For oxidation (Steinschneider and Fraenkel-Conrat, 1966a), organisms. Q␤ phage satellite RQ120 RNA has been 4 g of RNA were incubated in the dark for 60 min at 25ЊC in 40 l generated by the addition of the 3Ј terminus of E. coli of a buffer (100 mM Na-acetate [pH 5.3], 10 mM EDTA) containing tRNA Asp 1 to the last quarter of the Q␤ coat protein cistron 35 mM NaIO4. The extent of modification was nearly 100%, judging by a streptavidin-induced gel-shift of the RNA biotinylated at the (Munishkin et al., 1988) . A number of Sindbis virus DI oxidized end. The oxidized RNA was used for ␤-elimination, essen-Prisyazhnoy for help and advice in the experiments on terminal RNA modifications, and N. I. Androsova and I. I. Naumova for technical tially according to Steinschneider and Fraenkel-Conrat (1966b) : 2 g RNA were incubated for 3 hr at 25ЊC in 60 l of a buffer containing assistance. This work was supported by grants MTO 000 and MTO 300 from the International Science Foundation and the Russian Gov-10 mM Na-acetate and 0.25 M aniline, and adjusted to pH 5.0 with acetic acid. The 3Ј phosphate was removed by incubation of 1 g ernment, grants 93-04-06550 and 96-04-48331 from the Russian Foundation for Basic Science, grant INTAS-RFBR 95-1365, and by of the aniline-treated RNA for 1 hr at 50ЊC in 20 l of a buffer (25 mM Tris-HCl [pH 8.0], 0.1 mM EDTA) containing 2.5 units of calf an International Research Scholar's award from the Howard Hughes Medical Institute to A. B. C. intestine alkaline phosphatase (molecular biology grade, Boehringer Mannheim) with a subsequent phenol extraction and ethanol precipitation.

Received March 29, 1996; revised January 13, 1997.

RNA recombination in animal and plant viruses

Genetic recombination with poliovirus type 1: studies of crosses between a normal horse serum-resistant mutant and several guanidine-resistant mutants of the same strain

Oligonucleotide synthesis using T7 RNA polymerase and synthetic DNA templates

Determination of RNAprotein contacts using thiophosphate substitutions

RNAs from two independently isolated defective interfering particles of Sindbis virus contain a cellular tRNA sequence at their 5Ј-ends

Evolution of host cell RNA into efficient template RNA by Q␤ replicase: the origin of RNA in untemplated reactions

Synergism in replication and translation of messenger RNA in a cellfree system

An in vivo recombinant RNA capable of autocatalytic synthesis by Q␤ replicase

Efficient templates for Q␤ replicase are formed by recombination from heterologous sequences

Targeting the site of RNA-RNA recombination in brome mosaic virus with antisense sequences

Homologous recombination by reverse transcriptase during copying of two distinct RNA templates

In vitro suppression of an amber mutation by a chemically aminoacylated transfer RNA prepared by runoff transcription

Reversion of Q␤ phage mutants by homologous RNA recombination

The primary structure of crossover regions of intertypic poliovirus recombinants: a model of recombination between RNA genomes

Studies of nucleotide sequences in tobacco mosaic virus ribonucleic acid. III. Periodate oxidation and semicarbazone formation

Studies of nucleotide sequences in tobacco mosaic virus ribonucleic acid. IV. Use of aniline in stepwise degradation

Studies on the recombination between RNA genomes of poliovirus: the primary structure and nonrandom distribution of crossover regions in the genomes of intertypic poliovirus recombinants

The making of a phage

Optimal computer folding of large RNA sequences using thermodynamics and auxiliary information

References RNA fragments were preincubated in a buffer (80 mM Tris-HCl [pH 7.8], 1 mM EDTA, 10% glycerol, 100 mM KCl, Ϯ 10 mM MgCl2) at Berestowskaya, N.H., Vasiliev, V.D., Volkov, A.A., and Chetverin, 20ЊC during the indicated time period. A 10 l aliquot containing A.B. (1988) . Electron microscopy study of Q␤ replicase. FEBS Lett. the specified number of RNA molecules was spread over a 0.37 228, 263-267. mm-thick square (18 ϫ 18 mm) of a 2% Type IX agarose (Sigma) Biebricher, C.K., Diekmann, S., and Luce, R. (1982) . Structural analyslab prepared in a buffer (80 mM Tris-HCl [pH 7.8], 2 mM MgCl2, 1 sis of self-replicating RNA synthesized by Q␤ replicase. J. Mol. Biol. mM EDTA, 20% glycerol) containing, in a volume of 120 l, 4.2 g 154, 629-648. of a homogeneous preparation of Q␤ replicase isolated as described (Blumenthal, 1979; Berestowskaya et al., 1988 (1995) . Foreign complerandomly picked clones were subjected to alkaline denaturation mentary sequences facilitate genetic RNA recombination in brome and sequenced by the dideoxy chain termination method using the mosaic virus. Virology 208, 370-375. Sequenase Version 2.0 (USB) DNA sequencing kit and protocol. Gilbert, W. (1986) . The RNA world. Nature 319, 618.

Hajjou, M., Hill, K., Subramaniam, S.V., Hu, J.Y., and Raju, R. (1996) . Plasmids carrying recombinant sequences were digested by restric-Nonhomologous RNA-RNA recombination events at the 3Ј nontranstion endonuclease SmaI and used as templates for RNA synthesis lated region of the Sindbis virus genome: hot spots and utilization by T7 RNA polymerase (Noren et al., 1990