¹Reprint requests: fax: (573)882-0588, email: schoelz@missouri.edu
²Present address: Department of Plant Pathology, 334 Plant Science Bldg. Cornell University, Ithaca NY 14853

SUMMARY

We have previously shown that cauliflower mosaic virus (CaMV) is able to recombine with Nicotiana bigelovii plants that contain a CaMV transgene under conditions of strong selection pressure. We constructed transgenic N. bigelovii that expressed gene VI of CaMV strain D4, a gene that determines systemic infection of solanaceous species including N. bigelovii. CaMV strain CM1841 could not systemically infect nontransformed N.bigelovii, but was able to recombine with the D4 gene VI coding sequence present in transgenic N. bigelovii plants and the recombinant CM1841 virus infected the transgenic N. bigelovii systemically. In the present study we have simulated conditions of moderate selection pressure by inoculating CaMV strain W260 to the transgenic N. bigelovii. Because W260 systemically infects nontransformed N. bigelovii, a recombinant virus formed between W260 and the transgene would have little selective advantage over the wild type W260 virus. W260 was inoculated to approximately 100 nontransformed and 100 transgenic plants and it systemically infected nearly all of the plants. An analysis of viruses recovered from 23 transgenic plants infected with W260 revealed that 20 infections resulted from the systemic movement of the wild type W260 virus, while a recombinant between W260 and the transgene was detected in three of the infections. Thus, recombinant viruses formed between a wild type virus and a transgene can be isolated from transgenic plants under conditions of moderate selection pressure.

Key words: Cauliflower mosaic virus, recombination, transgenes

INTRODUCTION

It has recently been demonstrated that plant viruses can acquire an assortment of viral genes from transgenic plants. Lommel and Xiong (1991) showed that a red clover necrotic mosaic virus (RCNMV) isolate defective in cell-to-cell movement recombined with a copy of its cell-to-cell movement gene present in transgenic Nicotiana benthamiana plants. Cauliflower mosaic virus was shown to acquire a copy of its gene VI, a gene identified as a translational transactivator, from either transgenic Brassica napus (Gal et al., 1992) or transgenic Nicotiana bigelovii (Schoelz and Wintermantel, 1993). Greene and Allison (1994) demonstrated that a cowpea chlorotic mottle virus (CCMV) isolate that contained a defective coat protein gene could acquire a functional copy through recombination with transgenic N. benthamiana plants that contained CCMV coat protein sequences.

One feature common to these studies is that strong selection pressure was used for the isolation of the recombinant from the transgenic plants. Strong selection pressure can be defined as the inability of a virus to infect a host systemically unless a recombination event occurs. In three of the studies, deletion mutants that were defective for systemic infection were inoculated to the transgenic plants (Lommel and Xiong, 1991; Gal et al., 1992; Greene and Allison, 1994). The only viruses capable of systemic infection were those formed by recombination between the mutant virus and the transgene. In one study, which documented recombination between CaMV strain CM1841 and transgenic N. bigelovii, the selective pressure was conferred by a difference in the host range determinants present in the viral inoculum and the host range determinants present within the viral transgene (Schoelz and Wintermantel, 1993). CM1841 causes systemic infections in a wide range of crucifers, but cannot infect nontransformed N. bigelovii plants systemically. This virus was inoculated to transgenic N. bigelovii that expressed the gene VI product of CaMV strain D4, a CaMV gene that determines systemic infection of solanaceous species including N. bigelovii (Schoelz et al., 1986). Because sequences within the CaMV transgene specified systemic infection of N. bigelovii, a recombinant that acquired the transgene also acquired the ability to systemically infect N. bigelovii, and therefore gained a selective advantage over CM1841.

Based upon these studies, it has been suggested that recombinant viruses may not become established in transgenic plants unless the transgene confers a significant selective advantage over the wild type virus (Schoelz and Wintermantel, 1993; Falk and Bruening, 1994). We now demonstrate, however, that recombinant viruses can be isolated from transgenic plants even when the initial virus can systemically infect the host, conditions that can be described as moderate to weak selection pressure.

MATERIALS AND METHODS

Cauliflower mosaic virus strains, chimeric viruses, and transgenic plants. The cauliflower mosaic virus strains CM1841, W260, and D4 have been described previously. Each of these viruses has been cloned in infectious form into bacterial plasmid vectors (Wintermantel and Schoelz, 1996) and infections have been established in turnips from the cloned DNAs to eliminate natural variants in the populations. The wild-type viruses have been stored in lyophilized turnip leaf tissue at 4C for use as inoculum in experiments.

Transgenic N. bigelovii that express a chimeric gene VI product have been described previously (Schoelz et al., 1991). The gene VI coding region is derived primarily from CaMV strain D4, and contains the sequences of D4 required for systemic infection of solanaceous species, including N. bigelovii (Schoelz et al., 1986; Schoelz and Wintermantel, 1993). All transgenic N. bigelovii used in these studies were hemizygous for gene VI, and were obtained from a cross between homozygous transgenic plants and nontransformed N. bigelovii.

Virus inoculum consisted of purified virions (Hull et al., 1976) or a partially purified virion preparation prepared from infected turnips as described in Schoelz et al. (1986). The nontransformed Nicotiana bigelovii and transgenic N. bigelovii were inoculated approximately 5 weeks after seeds were sown. The inoculated plants were maintained for 42 to 56 days either in growth chambers (Percival Scientific Inc., Boone, IA; Conviron, Asheville, NC) with a 10 hr photoperiod at 18C at a light intensity between 200-350 E/m²/s¹, or in the greenhouse during the months of October through April.

Recovery of virion DNA from transgenic and nontransformed plants. Virion DNA was recovered from transgenic plants by 1) passage through turnips and subsequent purification from turnips or 2) directly from transgenic plants by use of PCR (Wintermantel and Schoelz, 1996). For passaging, turnips (Brassica rapa cv Just Right) were inoculated with extracts from systemically infected wild-type or transgenic N. bigelovii leaves. The infection was allowed to develop, and virion DNA was purified from the turnips approximately 5 weeks after inoculation by the procedure of Gardner and Shepherd (1980).

For competition studies between W260 and W260R, viral DNA was purified from 5-8 g of systemically infected nontransformed N. bigelovii leaves. The EcoRI-cleaved viral DNA was separated on a 1.5% agarose gel, transferred to nitrocellulose and probed with ³²P-labelled DNA as described in Maniatis et al. (1982). The probe consisted of equal proportions of the 459 bp EcoRI fragment of CM1841 and the 397 bp EcoRI fragment of D4.

Identification of recombination junctions. Recombination junctions were identified by sequencing of cloned viral DNA according to the dideoxy chain termination method of Sanger et al. (1977) using oligonucleotide primers complementary to viral sequences. Viral DNA was recovered from transgenic plants by PCR amplification as described in Wintermantel and Schoelz (1996). The 3172 bp viral DNA fragment that was amplified by PCR was cloned into the bacterial plasmid vector pGEM-7Zf (+) (Promega Corp., Madison, WI) at unique XbaI and ClaI restriction enzyme sites located in the viral DNA near the ends of the amplified fragment.

RESULTS

Recovery of recombinant viruses from transgenic plants under conditions of strong and moderate selection pressure. To determine whether recombinant viruses could be recovered from transgenic plants under conditions of moderate selection pressure, we examined the infections of transgenic N. bigelovii caused by CaMV strain W260, a virus that infects nontransformed N. bigelovii systemically (Schoelz and Shepherd, 1988). Hence, a recombinant virus formed between W260 and the transgene would presumably have little selective advantage over the initial viral inoculum. W260 was inoculated to approximately 100 nontransformed and 100 transgenic N. bigelovii plants and it infected almost all of the plants systemically.

To characterize the virus responsible for the infections, viral DNA was recovered from systemically infected leaves of 23 transgenic plants and cleaved with EcoRI. Restriction enzyme mapping with EcoRI revealed that viruses recovered from the transgenic plants could be divided into three groups: wild-type (the virus originally inoculated), recombinant, or a mixture of wild-type and recombinant. An analysis of the EcoRI sites of each type of virus recovered from the transgenic plants is illustrated in Figure 1. Wild type viruses could be distinguished from recombinants by the presence of an EcoRI site within gene VI at nucleotide position 6105 (Figure 1A). An EcoRI digest of a wild type viral DNA recovered from the transgenic plants by either PCR amplification or passage through turnips results in the appearance of a 459 bp DNA band (Figure 1B, lane 1), in addition to other bands. Viruses that arose through recombination with transgene sequences were distinguished by the presence of an EcoRI site within gene VI at nucleotide position 6043 (Figure 1A), a site present only in the transgene. Nucleotide sequencing of gene VI of recombinant viruses recovered from the transgenic plants has demonstrated that this EcoRI site is a reliable marker for recombination (Schoelz and Wintermantel, 1993). An EcoRI digest of a recombinant virus results in the appearance of a 397 bp DNA band (Figure 1B, lane 3). Mixtures of recombinant and wild type viruses recovered from a single plant could be identified through the presence of both the 459 bp and 397 bp DNA bands (Figure 1B, lane 2).

The EcoRI restriction enzyme patterns revealed that 20 plants developed infections resulting from the systemic movement of the wild type W260 virus, while a recombinant virus was detected in the three remaining plants (Table 1). Two of these three infections consisted of a mixed population of wild type W260 and a recombinant (Table 1; Figure 1B, lane 2). In one plant, only the recombinant virus was recovered from systemically infected leaves (Table 1; Figure 1B, lane 3), indicating that the recombination event had occurred early after the initial infection and that the wild type W260 had been excluded from the upper leaves by the recombinant.

To prove conclusively that W260 had recombined with the transgene, we characterized the recombination junctions of viral DNA recovered from the infection in which only the recombinant was detected (Table 1). The viral DNA of the W260 putative recombinant was cloned into the plasmid vector pUC18 and the recombination junctions identified by nucleotide sequencing of portions of gene VI. The recombination junctions of the W260 recombinant virus were localized to short stretches of homologous sequences that contained the transgene transcript initiation and termination sites (Figure 2), further evidence that the mechanism of recombination involves template switches between the 35S RNA and the transgene mRNA during reverse transcription (Gal et al., 1992; Schoelz and Wintermantel, 1993). The W260 recombinant virus arose from a template switch during reverse transcription from the 35S RNA to the transgene mRNA that occurred within 67 nucleotides of the 3' end of the transgene mRNA. A second crossover, from the transgene mRNA back to the 35S RNA, was identified within 73 nucleotides of the 5' end of the transgene mRNA. In essence, the entire gene VI coding sequence of W260 had been replaced with the corresponding sequence of the transgene.

We have previously demonstrated that CM1841 was able to recombine with CaMV sequences present in transgenic N. bigelovii, but at that time did not assess the percent recovery of recombinant viruses from transgenic plants. To determine the percent recovery of recombinants in the presence of strong selection pressure, CM1841 was inoculated to approximately 100 plants each of nontransformed and transgenic N. bigelovii. None of the nontransformed N. bigelovii became systemically infected with CM1841, which is in agreement with our previous findings (Schoelz and Shepherd, 1988; Schoelz et al., 1991). In contrast, 36% of the transgenic plants developed systemic symptoms (Table 1). To characterize the virus responsible for those infections, we isolated virus from systemically infected leaves of 24 transgenic plants and cleaved the viral DNA with EcoRI. The EcoRI digests revealed that only the recombinant virus was present in 22 of the infected plants, while a mixed population of wild type and recombinant viruses was isolated from the remaining two plants.

Comparison of the competitiveness of the wild type W260 virus versus the recombinant W260 virus. The observation that a recombinant virus could predominate over the wild type W260 virus might reflect differences in the aggressiveness of the recombinant relative to the initial inoculum. For example, the host specificity information present in the transgene, which is derived from CaMV strain D4, may confer a selective advantage to the recombinant virus over viruses in which gene VI is derived from W260. Although both D4 and W260 are capable of infecting N. bigelovii systemically, D4 systemic symptoms usually appear at 18-20 days postinoculation, 4-5 days before those of W260 (Schoelz and Shepherd, 1988).

To determine whether the W260 recombinant virus, designated W260R, was more competitive than the wild type W260, the two viruses were inoculated separately and in varying ratios to nontransformed N. bigelovii. To eliminate any potential carryover of wild type W260 virus in the W260R population, the complete W260R genome was first cloned at its unique SalI site into pUC18 and then the cloned W260R virus was inoculated to turnips. The W260 and W260R inoculum was subsequently prepared from infected turnip leaves and purified virions were inoculated to N. bigelovii. The competitiveness of the two viruses was examined by mixing W260 and W260R virions at ratios of 1:1, 5:1, and 1:5, respectively, at a total concentration of 75 g/ml.

The infections induced by W260 and W260R were indistinguishable until systemic symptoms developed. There was no difference in the temporal appearance of primary lesions induced by W260 and W260R. Both viruses induced chlorotic primary lesions in N. bigelovii leaves at 7-8 days postinoculation. There was also no difference in the specific infectivity of the two virion preparations. W260 induced an average of 82.9 ± 33.7 primary chlorotic lesions per leaf (n = 10), while W260R induced 89.0 ± 33.2 primary lesions per leaf (n = 10). Although primary lesions of W260 and W260R were initiated at the same time, there was a striking difference in the appearance of systemic symptoms between W260 and W260R. The systemic symptoms of W260R began 4 days prior to those of W260 (Figure 3), indicating that W260R was more aggressive than W260. Furthermore, the N. bigelovii plants inoculated with mixtures of W260 and W260R developed systemic symptoms at the same rate as those inoculated with W260R alone, suggesting that W260R could predominate in mixed infections even when a minor component of the initial virus population.

To identify the virus responsible for the systemic infection in plants inoculated with mixtures of the W260 and W260R, viral DNA was purified from each of the N. bigelovii plants, cleaved with EcoRI, and the restriction enzyme patterns revealed after gel electrophoresis and southern blotting. The restriction enzyme digests indicated that W260R was more competitive than the wild type W260. Only W260R was recovered from each of the 10 plants inoculated at ratios of 5:1 or 1:1 (W260R vs. W260). Of the plants inoculated at a ratio of 1:5 (W260R vs. W260), W260R alone was recovered from 9 plants, while W260 alone was recovered from one plant (data not shown). A second experiment in which viruses were inoculated at approximately 40 g/ml yielded the same results.

DISCUSSION

We have demonstrated that virulent recombinant viruses formed between a virus and a viral transgene can be isolated from transgenic plants under conditions of moderate selection pressure. A recombinant which consisted of the W260 strain of CaMV and a transgene derived from gene VI of the D4 strain of CaMV was detected in three of 23 transgenic N. bigelovii plants exhibiting symptoms. In one plant the recombinant was the only virus detected in systemically infected leaves.

Because recombination between CaMV and transgenic plants has been documented only recently, very little is known about competition between wild type and recombinant viruses in infected transgenic plants. In contrast, previous studies have shown that two CaMV isolates inoculated to the same plant are in competition for the plant's resources, and that there can be clear winners and losers. For example, the Cabbage S isolate has a competitive advantage over other isolates such as W, D/H and UM130 (Melcher et al., 1986; Zhang and Melcher, 1989). If turnip plants are inoculated with equal amounts of any of these viruses and Cabbage S, only Cabbage S is recovered from the infected plant. An equal mixture of Cabbage S and UM130 was recovered from infected plants only when the ratio of Cabbage S virions to UM130 virions in the inoculum was 2:23. Although mixtures of two isolates can be isolated from infected plants, it is generally believed that CaMV isolates recovered from infected plants in nature consist of one predominant isolate and several populations of minor variants (Vaden and Melcher, 1990; Riederer et al., 1992; Al-Kaff and Covey, 1994).

The observation that a recombinant virus could predominate over the wild type W260 virus was unexpected, but it can be explained by an increase in aggressiveness conferred by the transgene. An experiment involving direct competition between W260 and the recombinant W260R clearly showed that W260R was more aggressive in N. bigelovii than the wild type W260. Even when five times as much W260 was present in the inoculum as W260R, W260 was recovered from systemically infected leaves of only one of 10 N. bigelovii plants. This study demonstrates that recombination with a transgene can actually enhance the competitiveness of a virus if the transgene confers even a slight selective advantage and recombination occurs between closely related strains. It is important to note, however, that a selective advantage in one host may not mean that the virus has a selective advantage in all hosts. Significantly, mixtures of wild type and recombinant detected in the transgenic plants remained mixtures even after passage through turnips, an indication that the recombinant may not have a selective advantage over the wild type in that host.

The mechanism of recombination between CaMV and transgenic plants is similar to the recombination mechanism between RNA viruses and transgenic plants. We have shown in this study that the mechanism of recombination between CaMV and transgenic plants involves two template switches during reverse transcription of the CaMV 35S RNA to circular, dsDNA, which is in agreement with previous studies (Gal et al., 1992, Schoelz and Wintermantel, 1993). The first template switch occurs from the 5' end of the viral RNA to the 3' end of the transgene mRNA produced by the transgenic plants. A second switch occurs at the 5' end of the transgene mRNA back to the viral 35S RNA. In essence, the gene VI transgene completely replaces the gene VI coding region present on the viral inoculum. Thus, CaMV recombination can occur between two RNA molecules and is mediated by the viral reverse transcriptase. This recombination mechanism is similar to that of the RNA viruses, in which the RNA polymerase switches from one template to another location on the same template, or to a completely different template (Simon and Bujarski, 1994).

Although similar mechanisms are utilized by RNA viruses and CaMV for recombination with viral genes in transgenic plants, the isolation frequency of recombinant viruses is much higher with CaMV. For example, a recombinant form of CCMV was recovered from 3% of transgenic N. benthamiana plants containing CCMV sequences under conditions of high selection pressure (Greene and Allison, 1994). In contrast, we found that CM1841 infected 36% of the transgenic N. bigelovii inoculated, and all of the infected plants that we examined contained a recombinant virus. The high recovery rate of CaMV recombinants may be a consequence of the replication strategy of CaMV and the high degree of homology at the 3' ends of the transgene mRNA and the 35S RNA. During first strand DNA synthesis, the viral reverse transcriptase must switch templates from the 5' end of the 35S RNA to the 3' end for replication to continue. There is a 180 nucleotide terminal redundancy in the 35S RNA, and the template switch during viral replication occurs within this 180 nucleotide stretch (Mason et al., 1987). The gene VI mRNA produced in transgenic N. bigelovii has the same 3' end as the 35S RNA produced by the viral inoculum, and this may facilitate the template switch from the 35S RNA to the transgene mRNA. In fact, it has been suggested that a template switch between the 35S RNA and the 19S RNA (the viral gene VI mRNA) might be a common occurrence in mixed infections of CaMV (Dixon et al., 1986). Further research will be directed toward determining whether CaMV is more likely to recombine with transgenic plants that express the gene VI mRNA than with transgenic plants that express other CaMV genes.

ACKNOWLEDGMENTS

We thank David Pinkerton for the preparation of figures. This research was supported by a grant from the Food for the 21st Century program at the University of Missouri and by U.S. Department of Agriculture/National Research Initiative Competitive Grants No. 92-37303-7862 and 95-33120-1854. All figures and tables were used with permission from Academic Press.

REFERENCES

Al-Kaff, N., and Covey, S. N. (1994). Variation in biological properties of cauliflower mosaic virus clones. J. Gen. Virol. 75, 3137-3145.

Dixon, L., Nyffenegger, T., Delley, G., Martinez-Izquierdo, J., and Hohn, T. (1986) Evidence for replicative recombination in cauliflower mosaic virus. Virology 150, 463-468.

Falk, B. W., and Bruening, G. (1994) Will transgenic crops generate new viruses and new diseases? Science, 263, 1395-1396.

Gal, S., Pisan, B., Hohn, T., Grimsley, N., and Hohn, B. (1992) Agroinfection of transgenic plants leads to viable cauliflower mosaic virus by intermolecular recombination. Virology, 187, 525-533.

Gardner R. C., and Shepherd, R. J. (1980) A procedure for rapid isolation and analysis of cauliflower mosaic virus DNA. Virology, 106, 159-161.

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

Hull, R., Shepherd, R. J., and Harvey, J. D. (1976). Cauliflower mosaic virus: an improved purification procedure and some properties of the virus particles. J. Gen. Virol. 31, 93-100.

Lommel, S. A., and Xiong, Z. (1991) Reconstitution of a functional red clover necrotic mosaic virus by recombinational rescue of the cell-to-cell movement gene expressed in a transgenic plant. J. Cellular Biochemistry, 15A, 151.

Maniatis, T., Fritsch, E. F., and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Mason, W. S., Taylor, J. M., and Hull, R. (1987) Retroid virus genome replication. Adv. Virus Res. 32, 35-96.

Melcher, U., Choe, I. S., Lebeurier, G. Richards, K. Essenberg, R. C. (1986) Selective allele loss and interference between cauliflower mosaic virus DNAs. Mol. Gen. Genet. 203, 230-236.

Riederer, M. A., Grimsley, N. H., Hohn, B., and Jiricny, J. (1992). The mode of cauliflower mosaic virus propagation in the plant allows rapid amplification of viable mutant strains. J. Gen. Virol. 73, 1449-1456.

Sanger, F., Nicklen, S., and Coulson, A. R. (1977) DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA, 74, 5463-5467.

Schoelz, J. E., and Shepherd, R. J. (1988) Host range control of cauliflower mosaic virus. Virology, 162, 30-37.

Schoelz, J., Shepherd, R. J., and Daubert, S. (1986) Region VI of cauliflower mosaic virus encodes a host range determinant. Mol. Cell Biol. 6, 2632-2637.

Schoelz, J. E., Goldberg, K.-B., and Kiernan, J. (1991) Expression of cauliflower mosaic virus (CaMV) gene VI in transgenic Nicotiana bigelovii complements a strain of CaMV defective in long distance movement in nontransformed N. bigelovii. Mol. Plant-Microbe Interact. 4, 350-355.

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

Simon, A. E., and Bujarski, J. J. (1994) RNA-RNA recombination and evolution in virus-infected plants. Annu. Rev. Phytopathol. 32, 337-362.

Vaden, V. R., and Melcher, U. (1990) Recombination sites in cauliflower mosaic virus DNAs: Implications for mechanisms of recombination. Virology 177, 717-726.

Wintermantel, W. M., and Schoelz, J. E. (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.

Zhang, X. S., and Melcher, U. (1989) Competition between isolates and variants of cauliflower mosaic virus in infected turnip plants. J. Gen. Virol. 70, 3427-3437.

Table 1. Isolation of recombinant viruses from transgenic N. bigelovii plants.

# Plants w/systemic symptomsa

# Infections analyzedb

Virus	No. of tests	Wild-type N. bigelovii	Transgenic N. bigelovii	W	R	M	Total
W260	12	101/102	104/106	20	1	2	23
CM1841	13	0/102	38/106	0	22	2	24

Figure 1. EcoRI restriction enzyme digest of W260 viral DNAs recovered from transgenic N. bigelovii. (A) Partial EcoRI maps of W260, D4, and CaMV sequences present in transgenic N. bigelovii. Numbers within parentheses indicate nucleotide positions. (B) EcoRI restriction enzyme analysis of the W260 viral DNAs recovered from transgenic N. bigelovii. Lane 1, W260 wild type DNA. Lanes 2 and 3, W260 viral DNAs after passage through transgenic N. bigelovii and subsequent passage through turnip. Lane 4, D4 plasmid DNA. The DNA fragments were separated on a 1.5% (w/v) agarose gel and visualized by staining with ethidium bromide. The arrows indicate the EcoRI fragment polymorphism between D4 and W260 sequences. The 459-bp DNA fragment is indicative of W260 sequences within gene VI while the 397-bp DNA segment is indicative of D4 sequences within gene VI.

Figure 2. Location of recombination junctions between CaMV and transgene sequences. Transgenic plants contained a XbaI - ClaI DNA segment of CaMV and produced a mRNA from the CaMV 19S promoter and polyadenylation signals. A SacI - HgiAI DNA segment of the transgene was derived from CaMV strain D4. This segment contained the essential D4 sequences necessary for systemic infection of solanaceous species. Flanking regions of the transgene were derived from strain CM1841. Virion DNAs were recovered from transgenic plants by PCR using primers that flanked the transgene sequences such that only virion DNA, and not chromosomal DNA, would be amplified (Wintermantel and Schoelz, 1996). A XbaI - ClaI DNA segment was subsequently cloned into pGEM-7Zf (+). The double arrows indicate the regions that were sequenced to identify the recombination junctions. The unassigned sequences cannot be identified as being derived from the transgene or the CaMV viral inoculum because of a lack of sequence polymorphisms.

Figure 3. Appearance of CaMV systemic symptoms in N. bigelovii after inoculation with W260, W260R, or varying mixtures of the two viruses. Ten N. bigelovii plants were inoculated with W260R (q), W260 (n), a 5:1 mixture of W260 vs. W260R (l), a 1:1 mixture of W260 vs. W260R (s), or a 1:5 mixture of W260 vs. W260R (m).