SUMMARY

The genome of the bipartite red clover necrotic mosaic virus (RCNMV) consists of RNA-1 and RNA-2. RCNMV RNA-1 encodes proteins necessary for viral replication and encapsidation while RNA-2 encodes a single movement protein (MP) essential for the viral cell-to-cell spread. To study recombination between transgenically expressed viral RNA sequences and infecting viral RNA, transgenic Nicotiana benthamiana plants expressing a transcript of RCNMV RNA-2 lacking the most 5' terminal 26 nucleotides were generated. When the transgenic plants were inoculated with in vitro transcripts of RCNMV RNA-1, none of the inoculated plants developed local and systemic symptoms within one week. However, several inoculated transgenic plants became systemically infected four to six weeks later. Progeny viruses recovered from these symptomatic plants contained two species of RNA molecules (RNA-1 and RNA-2-like) and were infectious on non-transformed N. benthamiana plants. Run-off reverse transcription and direct RNA sequencing of the RNA-2-like molecules showed that the RNA-2-like molecules were recombinants between mRNA of the viral MP gene expressed in transgenic plants and the inoculated viral RNA-1. The 5' terminus of one recombinant RNA contains 10 nucleotides identical to the 5' terminus of the RNA-1. The 5' terminus of the second recombinant RNA contains 17 nucleotides that are complementary to an internal region of RNA-1. These data indicate that the MP transgene transcript recombined with both the plus strand and the minus strand of RCNMV RNA-1.

INTRODUCTION

Transgenic expression of viral genes or sequence fragments in plants has been demonstrated to confer resistance to infections by the same or related viruses (For reviews see Fitchen and Beachy, 1993; Grumet, 1990; Van der Elzen et al., 1989; Wilson, 1993). With few virus control strategies available, transgenic resistance is a promising tool in viral disease management. However, since transgenic plants are resistant only to specific viruses, they can be infected by other viruses. Consequently, transgene transcripts can become templates for RNA recombination with infecting viruses. Such recombination may generate viruses and strains with altered biological properties (De Zoeten, 1991).

RNA recombination among viruses occurs frequently. Evidence of RNA virus recombination has been compiled from laboratory studies as well as sequence analysis of naturally occurring viral isolates. Recombination was readily detected among the tripartite genome of brome mosaic virus (BMV) (Rao and Hall, 1990; 1993), between satellite RNA C and RNA D turnip crinkle virus (TCV) (Cascone et al., 1990; 1993), and within RNA-3 of cowpea chlorotic mottle virus (CCMV) (Allison et al., 1990). Evidence of recombination between RNA viruses and host RNA further suggests a potential for recombination between transgene transcripts and infecting RNA viruses. An insertion of a 28S ribosomal RNA sequence in the haemagglutinin gene was reported in influenza virus (Khatchikian et al., 1989). Subsequently, cellular sequences were identified in the RNA genomes of a bovine diarrhea virus isolate (Meyers et al., 1991) and a border disease virus isolate (Becher et al., 1996). Insertions of these cellular sequences changed non-cytopathogenic viral isolates into cytopathogenic ones. Recombination between RNA viruses and host RNA is not limited only to animal viruses. Sequence identical to an exon of tobacco chloroplast DNA ORF 196 was found at the 5' terminus of the RNA genome of a Scottish isolate of potato leafroll virus (Mayo and Jolly, 1991).

Recently Green and Allison (1994) reported transgenic recombination in cowpea chlorotic mottle virus (CCMV). When transgenic Nicotiana benthamiana plants expressing a truncated CCMV capsid protein (CP) gene were inoculated with a CCMV CP deletion mutant defective in systemic movement, a virus capable of systemic infection was recovered. It was shown that a complete CCMV CP gene was restored by a recombination event between the viral RNA and the transgene transcript. Recombination between a transgene and its cognate virus has also been documented in a plant pararetrovirus, cauliflower mosaic virus (CaMV). Transgenically expressed CaMV gene VI, a gene that determines systemic infection of CaMV in solanaceous hosts, recombined with infecting CaMV under strong (Schoelz and Wintermantel, 1993) or moderate (Wintermantel and Schoelz, 1996) selection pressures. Although CaMV is a DNA virus, the transgenic recombination in CaMV was reported to involve the transgene transcript and the RNA replication intermediate of the pararetrovirus. It is important to note that these recombination events occurred under the selection pressure to recover systemic infection function. It is unclear whether a low probability recombination event would be detected or selected without a selection pressure. Regardless, these studies have demonstrated that recombination between a transgenically expressed viral gene and an infecting virus is possible.

In this report, we describe a recombination event between the cognate viral movement protein (MP) gene expressed in transgenic plants and the infecting viral RNA-1 of red clover necrotic mosaic virus (RCNMV). RCNMV is a small bipartite RNA virus (Murphy et al., 1995; Xiong and Lommel, 1989). Its genome consists of two single-stranded, positive-sense RNAs. RCNMV RNA-1 contains the CP gene and genes essential for virus replication (Xiong and Lommel, 1989; Xiong et al., 1993b). RNA-2 encodes a single 35 kd MP required for RCNMV cell-to-cell movement (Lommel et al., 1988; Xiong et al., 1993a). In protoplasts, RNA-1 alone is capable of replication (Paje-Manalo and Lommel, 1989). However, both RNA-1 and RNA-2 are required for systemic infection in planta (Gould et al., 1981; Okuno et al., 1983).

We generated transgenic plant lines expressing a RCNMV RNA-2 transcript lacking the 5' terminal 26 nucleotides. The transgenic lines were neither resistant to wild-type RCNMV infection nor capable of supporting local and systemic infection by RNA-1. However, systemic infection by RNA-1 was observed after a long delay in several of the transgenic plants. Analysis of the progeny virus recovered from the symptomatic transgenic plants indicated that systemic infection was the result of recombination between the infecting RNA-1 and the MP transgene mRNA. Sequence analysis of the recombinant RNA further indicated that the MP transgene transcript recombined with both the plus strand and the minus strand of RCNMV RNA-1. A template-switching event during viral RNA replication is proposed to explain the observed transgenic recombination.

MATERIALS AND METHODS

Plant transformation. The 5' terminal 26 nucleotides of RCNMV RNA-2 were deleted from a cDNA clone, pRC2 (Xiong and Lommel, 1991) by digestion with exonuclease III (Sambrook et al., 1989). The resulting viral cDNA fragment was inserted into a binary transformation vector, pMON410, between the CaMV 35S promoter and the nos polyadenylation signal. Transformation of Nicotiana benthamiana with the engineered vector, pMRC2D26, was performed as described (Horsch et al., 1985).

Transgenic plants were initially identified by polymerase chain reaction (PCR) with oligonucleotide primers corresponding to nucleotides 161-178 and complementary to nucleotides 643-659 of the RNA-2. Plants harboring RNA-2 sequences, as determined by PCR, were further analyzed by Southern blot hybridization (Sambrook et al., 1989). Total DNA from N. benthamiana was digested with either BamHI or EcoRI, fractionated by agarose gel electrophoresis, and blotted to nylon membranes. ³²P-labeled RCNMV RNA-2 probes were prepared with PCR and used for hybridization. To identify the transgene transcripts, total RNA was extracted from transgenic and non-transformed N. benthamiana plants using the TRIzol nucleic acid extraction kit (GIBCO-BRL, Gaithersburg, Maryland). Polyadenylated mRNA was purified with the Poly(A) Tract system as instructed by the manufacturer (Promega, Madison, WI). Both the total RNA and the polyadenylated RNAs were fractionated by agarose gel electrophoresis and blotted onto nylon membranes. ³²P-labeled riboprobes complementary to the genomic strand of RNA-2 were used to detect RCNMV RNA-2 specific RNA in Northern hybridization as described previously (Xiong et al., 1993a). The riboprobes were prepared in a 20 ml in vitro transcription reaction containing 40 mM Tris-HCl, pH 8.0, 8 mM MgCl2, 2 mM spermidine-(HCl)₃, 25 mM NaCl, 10 mM DTT, 0.5 mM each of ATP, CTP, and GTP, 15 µM UTP, 50 mCi -³²P-UTP (800 Ci/mmol), 25 U T3 RNA polymerase, and 0.5 mg linearized pRC2 plasmid DNA (Xiong and Lommel, 1991). After incubation at 37C for 1.5 hr, DNA templates were removed by digestion with 3 U RNase-free DNase at 37C for an additional 15 min.

Virus inoculation and purification. Transgenic N. benthamiana expressing the truncated RCNMV RNA-2 transcript and non-transformed plants were inoculated with RCNMV in vitro transcripts. The RNA-1 and RNA-2 transcripts were synthesized from infectious cDNA clones pRC1IG and pRC2IG, respectively, by T7 RNA polymerase as previously described (Xiong and Lommel, 1991). RNA-1 transcript alone, RNA-2 transcript alone, or RNA-1 and RNA-2 transcripts together were inoculated to transgenic and non-transformed N. benthamiana plants. Inoculated plants were kept in a temperature-controlled greenhouse at a temperature between 20-25C. Purification of RCNMV from infected plant tissues and subsequent extraction of virion RNA were carried out as described previously (Xiong and Lommel, 1989).

Sequence analysis. Wild type RCNMV RNA and viral RNA recovered from RNA-1 inoculated, MP transgenic plants were subjected to runoff transcriptional mapping and direct RNA sequencing at the 5' termini. Both reactions were carried out with Superscript II reverse transcriptase (GIBCO-BRL) as described by Xiong and Lommel (1991). A 17 base oligonucleotide complementary to RCNMV RNA-2 nucleotides 45-61 was used to prime the RNA sequencing as well as the runoff transcription reactions. Unique sequences obtained at the 5' termini of the recombinant RNA-2 molecules were compared with RCNMV RNA-1 and RNA-2 sequences using the FASTA program in the University of Wisconsin Genetic Computer Group software. Sequence complementarity at the junction of recombination was analyzed with the same program.

RESULTS AND DISCUSSION

Characterization of transgenic plants. The transformation vector, pMRC2D26, contained nearly the complete RCNMV RNA-2 sequence except for the 5' terminal 26 nucleotides. The movement gene in the transgenic plants thus contains the entire ORF, the intact 3' sequence, and a truncated 5' non-coding region of RCNMV RNA-2. The predicted mRNA transcript derived from this construct presumably can not be replicated by RCNMV replicase because it lacks a highly conserved 5' terminal ACAAAC sequence that is thought to be required for replication.

Transgenic plants were identified by PCR with a pair of primers that delineate a region between nucleotide 161 and nucleotide 659 within the coding region of RCNMV RNA-2. Most of the R2 transgenic plants produced a single DNA fragment of approximately 498 bp in length (data not shown). In order to obtain homozygous transgenic lines expressing the truncated RCNMV RNA-2, R2 plants that were PCR-positive were selfed. Their progeny was further screened by PCR. R2 transgenic plants that produced all PCR-positive progenies were considered as homozygous and were used in subsequent experiments.

Integration of the truncated RCNMV RNA-2 sequence into N. benthamiana plants was confirmed by Southern hybridization. Total DNA from transgenic lines was digested with either BamHI or EcoRI and subsequently hybridized to a PCR generated RCNMV RNA-2 probes. As predicted, a 1.4 kb DNA fragment from BamHI-digested DNA and a 12 kb DNA fragment from EcoRI-digested DNA hybridized with the RNA-2 specific probe. No DNA fragment from the total DNA preparation of non-transformed plants hybridized with the same probe (data not shown). Transcriptional expression of the movement gene in the transgenic plants was assayed by Northern hybridization (Sambrook et al., 1989). Total and poly(A) plus RNAs were extracted from non-transformed and transgenic N. benthamiana plants. A single species of RNA molecule of approximately 1.4 kb was readily detected in both the total RNA and the poly(A) RNA fraction.

Recombinational capture of the movement gene mRNA from transgenic plants. When the transgenic N. benthamiana plants were inoculated with RCNMV RNA-1 in vitro transcripts, none of the plants developed any local or systemic symptoms in seven days. In comparison, both transgenic and non-transformed N. benthamiana showed pronounced systemic symptoms within one week after being inoculated with RNA-1 and RNA-2 in vitro transcripts. These results indicated that the MP-transgenic plants were not resistant to infection by RCNMV and were unable to complement infection by RCNMV RNA-1. It is not clear why the MP-transgenic plants failed to complement the systemic infection by RCNMV RNA-1. The failure might be explained by an insufficient level of the MP expression or MP expression at inappropriate times. Western blot using antibodies made against a fragment of MP protein expressed in E. coli did not detect any MP protein in the transgenic plants.

However, four out of 20 transgenic plants inoculated with RCNMV RNA-1 transcripts developed systemic symptoms after six to eight weeks. The negative controls, non-transformed N. benthamiana plants inoculated with RNA-1 only, did not show any symptoms. The symptoms on the RNA-1 inoculated transgenic plants were wild type in appearance. The virus recovered from the RNA-1 inoculated, transgenic plants was infectious when inoculated to healthy, non-transformed plants. Normal virions were observed in RNA-1 inoculated, symptomatic transgenic plants.

RNA extracted from these virions contained a RNA similar in electrophoretic mobility to wild type RNA-2 in addition to RNA-1. The infectious nature of the RNA-2-like RNA suggests that it is not an encapsidated movement gene mRNA. For confirmation, we analyzed this RNA for the presence of poly(A) tail by oligo(dT) cellulose column chromatography. The RNA-2 recovered from the transgenic plants were present only in the flow-through fraction along with the RNA-1, which does not have a poly(A) tail. As a positive control, a 2.3 kb mRNA supplied with a cDNA synthesis kit (GIBCO BRL) was present in both the flow-through and the bound fractions. The result clearly indicated that the recovered RNA-2 was not bound to the oligo(dT) cellulose and thus did not have a poly(A) tail.

As predicted from the transformation construct, the MP mRNA expressed in the transgenic plants should not have the 26 nucleotides at the 5' terminus of the wild type RNA-2. To determine the nature of the recovered RNA-2-like molecules, the 5' termini of these RNA molecules were mapped by run-off reverse transcription. The run-off transcriptional mapping revealed the presence of two RNA species in the RNA-2-like RNA. The smaller RNA (S-RNA) is 16 nucleotides shorter and the larger RNA (L-RNA) is 17 nucleotides longer than the wild type RNA-2. The exact sequences at the termini of the two recombinant RNAs were subsequently determined by direct RNA sequencing. Analysis of the sequences showed that the 5' terminal 10 nucleotides of the S-RNA were identical to 10 nucleotides at the 5' terminus of RCNMV RNA-1 and that the 5' terminal 17 nucleotides of the L-RNA were completely complementary to nucleotides 1995-2111 of the RNA-1 within the polymerase coding region (Figure 1). These results suggest that these RNAs are products of heterologous, intermolecular recombination. Both the plus-strand and the minus strand of RCNMV RNA-1 were involved in the transgenic recombination.

Transgenic recombination model. Based on the early work on RNA virus recombination (Cascone et al., 1990; 1993; Jarvis and Kirkegaard, 1993; Kirkegaard and Baltimore, 1986; Zhang et al., 1991), we propose a model to explain the generation of the S-RNA and L-RNA (Figure 2). When the transgenic plant is inoculated, RCNMV RNA-1 could replicate in the initially infected cells, but is unable to move to adjacent cells due to the lack of complementation by MP expressed in the transgenic plants. Nevertheless, MP transgene mRNA is transcribed in the nuclei of the transgenic plants and exported to cytoplasm, where the mRNA is recognized as a template for RNA replication by RCNMV replicase. The complete RNA-2 3' sequence present in the mRNA presumably contains a viral RNA replication promoter that can be recognized by the replicase. The replicase then initiates synthesis of a (-) strand RNA from the 3' end of the mRNA. During normal viral RNA replication, the viral replicase would recognize a (+) strand RNA synthesis promoter located at the 5' terminal of the viral RNA and start the synthesis of (+) strand viral RNA using the new (-) strand RNA as a template. However, the 5' promoter signal is not present in the MP mRNA as the first 26 nucleotides of RCNMV RNA have been removed. Instead of initiating synthesis of (+) strand RNA, the replicase along with the attached nascent (-) strand RNA dissociates from the original template as the replication fork approaches the 5' terminus of the MP mRNA. The replicase-nascent (-) RNA complex may subsequently encounter RCNMV RNA-1 and switches the template to either the 5' end of a positive strand or to an internal region of a negative strand of RNA-1. The template-switch event may be assisted by partial local hybridization between the nascent (-) strand RNA and the RNA-1 templates in this heterologous recombination (Figure 2). Tight association of RCNMV replicase with (-) strand RNA templates may have also contributed to the transgenic recombination event. Purified RCNMV replicase is tightly associated with endogenous RNA templates. Analysis of the endogenous templates showed that more than 95% of them are (-) strand viral RNA (Z. Xiong, unpublished data).

After the template-switch, RCNMV replicase continues the (-) strand RNA synthesis using the new templates, resulting in the capture of RCNMV RNA-1 sequence to the 5' end of the MP mRNA. Two scenarios that generate the recombinant S-RNA and L-RNA are illustrated in Figure 2. The replicase-nascent (-) strand RNA complex may land near the 5' terminus of a (+) strand RNA-1 and continues the synthesis of (-) RNA, leading to the capture of the 5' terminal 10 nucleotide of RNA-1 to the 5' end of the recombinant S-RNA (Figure 2D). Alternatively, the replicase-nascent RNA complex may jump to an internal region of a (-) strand RNA-1, leading to the capture of 17 nucleotides complementary to RCNMV RNA-1 to the 5' end of the recombinant L-RNA (Figure 2E). Synthesis of (-) strand RNA in this case may continue beyond the boxed sequence. The L-RNA is likely to be generated by the initiation of (+) RNA synthesis at the boxed sequence ACAAAC, which presumably contains the (+) strand synthesis signal. The recombinant RNA with a reconstructed 5' promoter sequence is then quickly amplified by RCNMV replication. Expression of MP by the recombinant RNA-2 allows RCNMV cell-to-cell movement and systemic infection which further amplifies the recombinant RNA-2. The delay of the systemic infection observed in RNA-1 inoculated, transgenic plants probably reflects the time required for the proposed recombination.

Two important factors may have contributed to the observed transgenic recombination. One is an extremely high selection pressure for the recombinant virus. RCNMV RNA-1 encodes genes necessary for viral replication and encapsidation, but cannot establish a successful infection in plants because it lacks a cell-to-cell movement function encoded by RCNMV RNA-2. Since the transgene in N. benthamiana was unable to complement infection by RCNMV RNA-1, only recombinants that have captured the MP gene were able to establish a successful infection. The second factor is the presence of the complete RCNMV 3' terminal sequence in the transgene. According to the proposed model, this sequence plays a pivotal role in the transgenic recombination. The terminal sequence allows the transgene mRNA to be recognized and used as a template by RCNMV replicase. Subsequent template-switching by the replicase resulted in the transgenic recombination. Additional experiments are being conducted to determine transgenic recombination frequency under little or no selection pressures and transgenic recombination frequency with transgenes without the 3' terminal viral sequence.

REFERENCES

Allison, R. F., C. Thompson, and P. Ahlquist. 1990. Regeneration of a functional RNA virus genome by recombination between deletion mutants and requirements for cowpea chlorotic mottle virus 3a and coat genes for systemic infection. Proc. Natl. Acad. Sci. U S A 87:1820-1824

Becher, P., G. Meyers, A. D. Shannon, and H. J. Thiel. 1996. Cytopathogenicity of border disease virus is correlated with integration of cellular sequences into the viral genome. J. Virol. 70:2992-2998.

Cascone, P. J., Carpenter, C. D., Li, X. H., and A. E. Simon. 1990. Recombination between satellite RNA of turnip crinkle virus. EMBO 9: 1709-1716.

Cascone, P. J., T. F. Haydar, and A. E. Simon. 1993. Sequences and structures required for recombination between virus-associated RNAs. Science 260:801-805.

De Zoeten, G. A. 1991. Risk assessment: Do we let history repeat itself? Phytopathology 81:585-586.

Fitchen, J. H., and R. N. Beachy. 1993. Genetically engineered protection against viruses in transgenic plants. Ann. Rev. Microb. 47:739-763.

Gould, A. R., R. I. B. Francki, T. Hatta, and M. Hollings. 1981. The bipartite genome of red clover necrotic mosaic virus. Virology 108:499-506.

Green, A. E., and R. F. Allison. 1994. Recombination between viral RNA and transgenic plant transcripts. Science 263:1423-1425.

Grumet, R. 1990. Genetically engineered plant virus resistance. HortScience 25:508-513.

Horsch, R. B., J. E. Fry, N. L. Hoffman, S. G. Rogers, and R. T. Fraley. 1985. A simple and general method for transferring genes into plants. Science 277:1229-1231.

Jarvis, T. C., and K. Kirkegaard. 1992. Poliovirus RNA recombination - Mechanistic studies in the absence of selection. EMBO J. 8:3135-3145.

Khatchikian, D., M. Orlich, and R. Rott. 1989. Increased viral pathogenicity after insertion of a 28S ribosomal RNA sequence into the haemagglutinin gene of an influenza virus. Nature 340:156-157.

Kirkegaard, K., and D. Baltimore. 1986. The mechanism of RNA recombination in Poliovirus. Cell 47:433-443.

Lommel, S. A., M. Weston-Fina, Z. Xiong, and G. P. Lomonossoff. 1988. The nucleotide sequence and gene organization of red clover necrotic mosaic virus RNA-2. Nucl. Acids Res. 16:8587- 8602.

Mayo, M. A., and C. A. Jolly. 1991. The 5'-terminal sequence of potato leafroll virus RNA: Evidence of recombination between virus and host RNA. J. Gen. Virol. 72:2591-2595.

Meyers, G., N. Tautz, E. J. Dubovi, and H. J. Thiel. 1991. Viral cytopathogenicity correlated with integration of ubiquitin coding sequences. Virology 180:602-616.

Murray, M. G., and W. F. Thompson. 1980. Rapid isolation of high molecular weight plant DNA. Nucleic Acids Res. 8:4321-4325.

Okuno, T., C. Hiruki, D. V. Rao, and G. C. Figueiredo. 1983. Genetic determinants distributed in two genomic RNAs of sweet clover necrotic mosaic, red clover necrotic mosaic and clover primary leaf necrosis viruses. J. Gen. Virol. 64:1907-1914.

Paje-Manalo, L., and S. A. Lommel. 1989 . Independent replication of red clover necrotic mosaic virus RNA-1 in protoplasts. Phytopathology 79:457-461.

Rao, A. L. N., and T. C. Hall. 1990. Requirement for a viral trans-acting factor encoded by brome mosaic virus RNA-2 provides strong selection in-vivo for functional recombinants. J. Virol. 64:2437-2441

Rao, A. L. N., and T. C. Hall. 1993. Recombination and polymerase error facilitate restoration of infectivity in brome mosaic virus. J. Virol. 67:969-979.

Sambrook, J., E. F. Fritisch, and T. Maniatis. 1989. Molecular cloning, A laboratory manual. 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York.

Schoelz, J. E., and W. M. Wintermantel. 1993. Expansion of viral host range through complementation and recombination in transgenic plants. Plant Cell 5:1669-1679.

Van der Elzen, P. J. M., Huisman, M. J., Willink, D. P., Jongedijik, E., Hoekma, A., and B. J. Cornelissen. C. 1989. Engineered virus resistance in agricultural crops. Plant Mol. Biol. 13: 337-346

Wilson, T. M. A. 1993. Strategies to protect crop plants against viruses - pathogen derived resistance blossoms. Proc. Nat. Acad. Sci.U S A 90:3134-3141.

Wintermantel, W. M., and J. E. Schoelz. 1996. Isolation of recombinant viruses between cauliflower mosaic virus and a viral gene in transgenic plants under conditions of moderate selection pressure. Virology 223:156-164.

Xiong, Z., and S. A. Lommel. 1989. The complete nucleotide sequence and genome organization of red clover necrotic mosaic virus RNA-1. Virology 171:543-554.

Xiong, Z., and S. A. Lommel. 1991. Red clover necrotic mosaic virus infectious transcripts synthesized in vitro. Virology 182:388-392.

Xiong, Z., D. Giesman-Cookmeyer, K. H. Kim, and S. A. Lommel. 1993a. The roles of the red clover necrotic mosaic virus capsid and cell-to-cell movement proteins in systemic infection. Virology 192:27-32.

Xiong, Z., K. H. Kim, T. L. Kendall, and S. A. Lommel. 1993b. Synthesis of the putative red clover necrotic mosaic virus RNA polymerase by ribosomal frameshifting. Virology 193:213-221.

Zhang, C., P. J. Cascone, and A. E. Simon. 1991. Recombination between satellite and genomic RNAs of turnip crinkle virus. Virology 184:791-794.

Figure 1. Alignment of the 5' terminal sequences of the recombinant S-RNA and L-RNA with the respective homologous sequences in RCNMV RNA-1 and with the 5' terminal sequence of the mRNA of the movement protein gene (MP) expressed in the transgenic plants. Sequences at the 5' termini of the L- and S-RNA were obtained by direct sequencing of the purified RNAs. Sequence at the 5' termini of the MP mRNA is predicted from the transformation construct. Box A highlights the MP 5' sequence which is also present in RNA-2, S-RNA and L-RNA (not included). Box B shows 10 identical nucleotides at the 5' termini of RNA-1 and S-RNA. Box C shows 17 nucleotides at the 5' terminus of the L-RNA that are complementary to an internal RNA-1 sequence within the putative polymerase coding region (solid box in RNA genome representation).

Figure 2. Proposed RCNMV transgenic recombination model. A. RCNMV replicase initiates (-) strand synthesis at the 3' end of the movement protein (MP) mRNA. B. When the replicase reaches at the 5' terminus of the MP mRNA , it dissociates from the MP mRNA template. C. The polymerase switches to either the 5' terminus of the (+) strand RNA-1 or to the internal region of the (-) strand RNA-1. The template switching is presumably assisted by partial local hybridization between the nascent (-) RNA and the new RNA-1 template. D and E. Putative recombination cross over junctions of the recombinant S-RNA and L-RNA are enlarged to illustrated the partial local hybridization between the recombining RNA molecules and the proximity of the polymerase to the sequences acquired by S-RNA and L-RNA. Sequences in (+) and (-) RNA-1 templates identical to the 5' termini of the S-RNA and L-RNA are boxed. In D, the (-) MP-RNA is partially hybridized with the (+) RNA-1 to generate the S-RNA. In E, the (-) MP-RNA is partially hybridized with the (-) RNA-1 to generate the L-RNA.