Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
DEPARTMENT for ENVIRONMENT, FOOD and RURAL AFFAIRS Research and Development CSG 15 Final Project Report (Not to be used for LINK projects) Two hard copies of this form should be returned to: Research Policy and International Division, Final Reports Unit DEFRA, Area 301 Cromwell House, Dean Stanley Street, London, SW1P 3JH. An electronic version should be e-mailed to [email protected] Project title Non-target effects of transgenic crop plants resistant to virus diseases DEFRA project code CB02017 Contractor organisation and location Prof. Philip J. Dale and Dr. Nadia S. Al-Kaff John Innes Center, Colney Lane, Norwich NR4 7UH Total DEFRA project costs Project start date £ 209,100.00 16/07/00 Project end date 16/07/03 Executive summary (maximum 2 sides A4) Virus derived genes have been used successfully in genetically modified (GM) crops to generate resistance to virus diseases (reviewed by Lomonossoff, 1995; Baulcombe, 1996). Coat protein genes and virus genes associated with replication are the most common viral sequences used to date for this purpose. Nuclear inclusion of a protease (NIa) viral gene has also been employed successfully in developing protection against some plant potyviruses (Maiti et al., 1993; Vardi et al., 1993). Another example of viral sequences used worldwide in plant biotechnology, is the 35S promoter and terminator from the DNA plant pararetrovirus cauliflower mosaic virus (CaMV). The 35S promoter expresses abundantly and constitutively in different plant tissue types, and is therefore a very useful tool to regulate transgene expression in plants (reviewed by Schothof et al., 1996). Stability of transgene expression and evaluation of the potential ecological risks are aspects of important consideration in the assessment of GM crops containing viral sequences, especially those used commercially. In nature, CaMV is found mixed with other virus types in some plant species. A good example of a mixed infection is the occurrence of CaMV and the potyvirus turnip mosaic virus (TuMV) in populations of wild cabbage (B. oleracea) in Dorset in Southern England (Raybould et al., 1999). TuMV is a single-stranded plant RNA virus and has an extremely broad host range. The main aim of this project was to investigate unexpected changes in expression of transgenes conferring specific virus resistances resulting from infection by other viruses not targeted by the genetic modification. We investigated possible interactions between transgenic DNA sequences from either coding or non-coding regions of the virus. This included analysis of interactions following infection with a virus with no sequence homology to the viral sequences in the transgene construct. In addition, we simulated a field type virus infection by testing different virus variants and times of infection, and infected plants with more than one virus type. We used two transgenic crops, oilseed rape CSG 15 (Rev. 6/02) 1 Project title Non-target effects of transgenic crop plants resistant to virus diseases DEFRA project code CB02017 (B. napus) and pakchoi (B. rapa) containing a single copy of the NIa gene of TuMV flanked by the 35S promoter and terminator of CaMV and the bialophos herbicide tolerance bar gene flanked by the 35S promoter and the nopaline synthase (NOS) terminator. We also used transgenic oilseed rape plants containing the reporter gene gus flanked by the 35S promoter and its terminator in one plant line. In another line the NOS promoter was used instead of the 35S promoter. Both lines contained the antibiotic resistance gene nptll regulated by the 35S promoter. A series of CaMV variants was used, which caused a range of systemic symptoms in oilseed rape and pakchoi. The TuMV variant used, CDN1, is an oilseed rape isolate that causes symptoms in both Brassica species. To study the interaction between the non-coding regions of CaMV, 35S promoter and terminator, and the pathogenicity of a virus other than CaMV, transgenic oilseed rape lines containing transgenes regulated by the 35S promoter and/or terminator were infected with TuMV. It has been shown previously that transgenic and non-transformed oilseed rape plants respond to CaMV infection by initial development of systemic symptoms, followed by recovery from infection when the new asymptomatic leaves emerge. The recovery phenomenon is a plant antiviral response mechanism which cosuppresses viral genes and homologous transgenes at the post-transcriptional or transcriptional levels (Covey et al., 1997; Al-Kaff et al., 1998). However, infection with TuMV caused severe disease which led to a continued deterioration of the infected plants with no sign of recovery. Oilseed rape (B. napus) is formed from a natural hybrid between B. oleracea and B. rapa. Both hybrid and parental species are considered to be susceptible hosts to CaMV infection, but they differ in their response. B. oleracea (Covey et al., 1997) and B. napus (Al-Kaff et al., 1998) recover after showing systemic symptoms, whereas B. rapa does not recover from CaMV infection (Melcher, 1989; Al-Kaff and Covey, 1995). To test the ability of CaMV to influence transgene expression in the genetic background of B. rapa, we infected non-transformed and transgenic lines of pakchoi (B. rapa) containing the NIa gene of TuMV with the moderate variant of CaMV Cabb B-JI. Pakchoi was highly susceptible to CaMV and showed very severe distortion in leaves and flowers. Transgene expression patterns at different times post inoculation resulted in up-regulation rather than suppression. In a separate series of experiments, the pakchoi transgenic lines were infected with a range of virus variants, M11 (mild), C-M (moderate) Cabb B-JI (severe), and Aust (very severe). A clear relationship was found between CaMV virulence and an increase in transgenic expression. Thus, the effect induced by CaMV infection on transgene expression differed between the two Brassica species analysed, suggesting that the fate of CaMV-induced transgene instability may be controlled by both pathogen and host genetic factors. Different types of interaction have been observed between different virus types or variants of the same virus in several plant species ranging from no effect, cross protection or enhancement of pathogenicity of one or more of the infecting viruses (see review by Hammond et al., 1999). In order to assess plant responses to mixed viral infection, we studied susceptibility of transgenic and nontransformed plants to double viral infection (CaMV and TuMV) in oilseed rape and pakchoi. No differences were observed in pathogenicity between single or mixed infection of nontransformed and transgenic oilseed rape plants. Suppression of transgene expression was only detected in oilseed rape lines infected with CaMV alone or along with TuMV. Interestingly at a later stage of infection, an enhanced transgene expression was detected in plants which were suppressed at an early stage in the doubly infected plants. This indicated that the potyvirus TuMV suppressed transgene silencing induced by CaMV. Unlike oilseed rape, non-transformed and transgenic pakchoi plants infected with both viruses, CaMV and TuMV, showed more severe disease symptoms than when infected by either virus alone. In addition, transgenic pakchoi lines co-infected with both viruses had elevated levels of transgene expression. Transgene expression products accumulated at high levels in these plants compared with plants infected only with a single virus. Analysis of the virus titres indicated a higher accumulation of CaMV viral DNA and no change in TuMV in mixed infected plants, compared with those infected only by each virus singly. From these results, we concluded that CaMV and TuMV were interacting synergistically in pakchoi. This synergistic interaction was manifested by enhancement of pathogenicity and an increase of CaMV DNA accumulation. CSG 15 (Rev. 6/02) 2 Project title Non-target effects of transgenic crop plants resistant to virus diseases DEFRA project code CB02017 We also investigated the interaction between coding regions of TuMV and infection by CaMV. For this purpose, oilseed rape and pakchoi transgenic lines carrying the NIa gene from TuMV were infected with different isolates of CaMV ranging from mild to very severe isolates. In an attempt to simulate a field type mixed viral infection, the above transgenic lines, and crosses between them, were infected at different stages with a mixture consisting of both CaMV and TuMV. Generally, the results showed that there was a positive relationship between viral virulence and up-regulation of transgene expression. This was independent of virus type and time of infection. The enhancement of transgene expression as a result of stress conditions, such as virus infection, could be important not only in improving our understanding of pathogen and host interactions, but also in complementing current GM bio-safety assessment. The importance of this issue in bio-safety assessment will be influenced by the origin of the transgene and the nature of the crop plant carrying it. There is also a range of plant or environment factors which may influence host/pathogen interactions. It is therefore important to enhance our understanding of these factors, so that our biosafety assessment may continue to be well informed. CSG 15 (Rev. 6/02) 3 Project title Non-target effects of transgenic crop plants resistant to virus diseases DEFRA project code CB02017 Scientific report (maximum 20 sides A4) 1. Introduction Virus derived genes are used successfully in transgenic plants to produce resistance to virus diseases in plants (reviewed by Lomonossoff, 1995; Baulcombe, 1996). Coat protein and replication related viral genes are the most common sequences used to confer resistance against virus infection. Nuclear inclusion of protease (NIa) genes have also been employed successfully for developing protection against potyviruses (Maiti et al., 1993; Vardi et al., 1993). Another example of viral sequence used worldwide in plant biotechnology is the 35S promoter and terminator from CaMV. The 35S promoter expresses abundantly and constitutively in most types of plant tissues, therefore it is a very useful tool to regulate the expression of genes in plants (reviewed by Schothof et al., 1996). Understanding interaction between different plant viruses in a mixed infection can provide useful background information on the interaction between virus sequences in transgenic crops and infecting viruses. There are two main potential consequences of this interaction; introduction of a new viral disease and a change in susceptibility to an existing virus; this could be by recombination between transgenes and an infecting virus (Rubio, 1999 and Tepfer, 2002). Another is by affecting virus ecology and transmissibility (Hull, 1990; Tepfer, 1993) and a further possible interaction is synergy between viruses and viral sequences in transgenic crops. Plants are usually susceptible to more than one virus and naturally can be infected with more than one variant or distinct virus type (Falk and Bruening, 1994). In some cases of mixed infection, viruses can replicate and transmit independently of each other. However in other cases, interaction between viruses can result in a decrease in the pathogenicity of one of them (cross protection), or an enhancement in pathogenicity by increasing replication and movement (synergism) or transmission of one of the viruses in the presence of the others. Synergistic interaction between two virus types can lead to enhancement of replication and as a consequence increase in viral titre of one or both of these viruses. The potyvirus turnip mosaic virus (TuMV) has been recorded in naturally infected oilseed rape (Walsh and Tomlinson, 1989; Hardwick et al., 1994), collards (Khan and Dempski, 1982) and non-transformed brassicas (Raybould, et al., 1999) together with the pararetrovirus double standard DNA cauliflower mosaic virus (CaMV). Specific interaction between CaMV and the tobamovirus RNA virus turnip vein-clearing virus has also been recorded (Hii et al., 2002). Studying transgene stability in genetically modified crops for virus resistances or transgenes containing elements of virus sequence, e.g. 35S promoter under different stresses, is an important issue for GM assessment. We found that CaMV infection can suppress the expression of transgenes containing the 35S promoter or RNA homology with the virus in oilseed rape (Brassica napus) plants (Al-Kaff et al., 1998; Al-Kaff et al., 2000; see also the review by Covey and Al-Kaff, 2000). The main aim of this project was to investigate the effect of transgene conferring resistance to specific viruses on the response of the plant to infection by viruses not targeted by the genetic modification (i.e. non-target effects). We studied the interaction between non-coding regions of CaMV and the pathogenicity of TuMV in crops of both oilseed rape and pakchoi. In this scientific report, we present and discuss results obtained during the three-year period of the project. 2. Experimental procedures 2.1 Viruses and plants. The CaMV variants used in this study were M11 (mild), C-M (moderate), Cabb B-JI (severe) and Aust (very severe). For TuMV, the CDN1 variant was the most widely used isolate. Two transgenic lines of B. napus cv. Westar 10 (oilseed rape) were used. One of them (TL1, CSG 15 (Rev. 6/02) 4 Project title Non-target effects of transgenic crop plants resistant to virus diseases DEFRA project code CB02017 see Figure 2) carried a single copy of the reporter glucuronidase (gus) transgene regulated by the CaMV 35S promoter and its transcriptional terminator, and the other line (TL2, see Figure 2) contained the same reporter gus transgene flanked by the nopaline synthase promoter (NOSp) and the octopine synthase terminator (OCSt). Both lines carried the selectable marker gene neomycin phosphotransferase (NptII) transgene driven by the 35S promoter and terminated by the OCSt (Figure 2). The cross between these two B. napus transgenic lines was also studied. For B. rapa (pakchoi), we used two independent transgenic lines which carried a single copy of the nuclear inclusion; a (NIa) transgene from TuMV regulated by the 35S promoter and terminator, and the selective herbicide bialophos tolerance (bar) transgene flanked by the 35S promoter and the NOS terminator (Figure 1). We also used the reciprocal crosses between these two pakchoi transgenic lines in most of the experiments performed. 2.2 Plant inoculation. CaMV and TuMV were propagated in B. rapa plants (turnip, cv. Just Right). For inoculation with single viruses, inocula were prepared by grinding 1 g of CaMV or TuMV infected leaves in 2 ml (crude sap) 1x DNase buffer (100 mM Tris-HCl pH 7.4, 2.5 mM MgCl2) or 100 mM potassium-phosphate buffer (pH 7.5), respectively. Inocula consisting in both CaMV and TuMV were prepared by mixing equal aliquots of 1:2 dilutions of each crude sap. Plants were inoculated at the first true leaf stage by rubbing inocula onto celite-dusted leaves and subsequently rinsing the leaves with water. Plants were grown in a containment glasshouse supplemented with illumination for 16 h/day at 18-22ºC. Samples were taken weekly from 18 days post inoculation (d.p.i.) to at least 39 d.p.i. under Defra licences PHL 11D/3736 (7/2001), PHL 185/4141 (4/2002) & PHL 85A/4467 (3/2003) 2.3 Nuclei acid extraction. Leaf samples were collected at different d.p.i. and ground in liquid nitrogen. DNA extraction was carried out by the Kirby method (Hull and Covey, 1983), and total RNA by the RNeasy Plant Mini Kit from Qiagen (Valencia, CA). A total RNA concentration was measured using a spectrophotometer and run in an ethidium bromide agarose gel to estimate the quantity and the quality of the RNA using the ribosomal RNAs as a comparator. DNA and RNA were analysed by Southern and Northern hybridisation methods respectively, using the appropriate probes. 2.4 Virus accumulation measurement. For the estimation of CaMV accumulation levels in infected plants, leaves at different d.p.i. were ground in 10 mM Tris-HCl buffer, pH 7.5, containing 1 mM EDTA and 2% w/v triton, then treated with proteinase K (0.5 mg/ml) in 1% w/v SDS for 30 min at 37ºC. Total plant DNA was extracted twice by phenol/chloroform followed by ethanol precipitation. A hundred ng total DNA was blotted onto nylon membrane (Hybond-N+, Amersham) using Bio-Dot SF microfiltration apparatus (Bio-Rad), and hybridised to CaMV specific probe. Signal densities were compared with the signals given by known amounts of purified CaMV virion treated in the same way. TuMV accumulation in infected plants was estimated by hybridising, with TuMV specific probe, 1 μg Qiagen isolated total RNA blotted onto Hybond-N+ membrane using also Bio-Dot SF microfiltration apparatus. 3. Results: 3.1 Production of oilseed rape transgenic lines. 3.1.1 Production and characterisation of transgenic oilseed rape lines containing virus genes The transgene construct used (TL3) contained a nuclear inclusion protease (NIa) gene from turnip mosaic virus (TuMV), flanked by the cauliflower mosaic virus (CaMV) 35S promoter and terminator. It also contained the herbicide bialophos tolerance (bar) gene, driven by the 35S promoter and transcriptionally terminated by the nopaline synthase (NOS) terminator. The construct was incorporated into the T-DNA of an Agrobacterium transformation vector and used to transform oilseed rape (B. napus) plants (Figure 1A) using the vacuum infiltration method (Liu et al., 1998). One CSG 15 (1/00) 5 Project title Non-target effects of transgenic crop plants resistant to virus diseases DEFRA project code CB02017 transformant line containing the TL3 construct was produced. Southern blot analysis and segregation data revealed that this transgenic line contained a single copy of the transgene (Figure 1B). 3.1.2 Biological function of transgenes. The new oilseed rape transgenic line containing the NIa from TuMV and the bar genes (TL3 Figure 1A), as well as pakchoi transgenic lines already produced carrying the same transgene construct, were infected with the CDN-1 isolate of TuMV. The results showed that transgenic lines from both Brassica species were susceptible to TuMV. Chlorotic local lesions in the inoculated leaf (the first true leaf) were the first symptoms observed. The first fully systemically infected leaf was the third leaf, and symptoms were mainly chlorotic and dark green tissues. The severity of the disease in oilseed rape transgenic plants was comparable to those induced in non-transformed control plants. In the case of pakchoi, we observed milder symptoms in transgenic than in non-transformed plants. Expression of the bar transgenes was confirmed by Northern analysis (Figure 1C). The non transgenic plants which were used as a control in this experiment and all following experiments were the negative segregants for the transgene (i.e. those not containing the transgene) of both oilseed rape and pakchoi plant lines. 3.2. Interaction between non-coding regions of CaMV and the pathogenicity of other viruses We challenged previously characterised transgenic oilseed rape lines (TL1) carrying the reporter glucuronidase (gus) gene and the neomycin phosphotransferase (nptII) as a selectable marker gene (Figure 2). Both transgenes were driven by the CaMV 35S promoter, although the gus transgene was transcriptionally terminated by the 35S terminator and the nptII by the octopine synthase (OCS) terminator. As a control in these experiments, we used another transgenic line of oilseed rape carrying TL2 (Figure 2). TL2, the control plants also contained the gus reporter transgene, but regulated by the Agrobacterium nopaline synthase (NOS) promoter and the OCS terminator and the same nptII transgene. Five plants of each transgenic and non-transformed line were infected in the first true leaf, either with the Cabb B-JI variant of CaMV or with the CDN-1 variant of TuMV. 3.2.1. Disease development at different growth stages. At about two weeks after infection, plants showed systemic symptoms in the third or fourth leaf, characterised by vein clearing and then chlorotic vein banding with interveinal dark green islands, in the case of CaMV infection, or by a mosaic of green and chlorotic patches in TuMV- infected plants. Samples from those plants were collected at 18, 25, 32, 39 and 46 d.p.i. Transgenic and nontransformed oilseed rape plants were found to be susceptible to both CaMV and TuMV, although the plant responses to each of these two viruses varied substantially. TuMV caused a severe disease which led to a continuous deterioration of the infected plants (Figure 3A). In contrast, CaMV infection of oilseed rape plants developed initial systemic symptoms followed by recovery. The new leaves of the recovered CaMV infected plants were asymptomatic (Figure 3B). To rule out the possibility that transgenes might interact with the replication of the infecting virus, we analysed the CaMV DNA and the TuMV RNA accumulations in transgenic and non-transformed plants. As shown in Figure 4, there was no clear difference in virus accumulation of either virus between transgenic and non-transformed oilseed rape plants. We also studied CaMV replication in those plants and found that they exhibited a similar pattern, with the presence of intermediate replicative forms (R) of the virus at early stages of infection (18-25 d.p.i.). The main DNA form accumulated at later stages of infection 39-46 d.p.i. was supercoiled viral DNA (Sc) (Figure 5). 3.2.2. Effect of infection on transgene expression CSG 15 (1/00) 6 Project title Non-target effects of transgenic crop plants resistant to virus diseases DEFRA project code CB02017 A major aim of our research was to study transgene expression in the oilseed rape transgenic lines following virus inoculation at different days post-inoculation (18, 25, 39, 46 d.p.i.). We used Northern blot analysis to study expression of both nptII and gus transgenes (Figure 6), and histo-chemical staining to detect the activity of the gus reporter gene (Figure 7). Transgenic plants were infected with both TuMV and CaMV. The results showed that only CaMV infection suppressed transgenes in oilseed rape with promoter and terminator homology to the virus. However, transgene expression was unaffected in plants infected with TuMV (Figures 6 and 7). The selectable marker nptII transgene expression was partially suppressed after CaMV infection even at a later stage of infection (46 d.p.i. – see Figure 6). In this and previous work, transgenes with promoter homology to CaMV sequences showed down-regulation of the transgene. The suppression of expression in this case was found not to be complete and we always detected traces of transgene transcripts (Al-Kaff et al, 1998). In most experiments with transgenic plants infected with CaMV and containing both promoter and terminator homology to the virus, transgene transcription was found to be silenced at the post transcriptional level (expressed at a lower level), but in other experiments for the same transgene, expression, was downregulated, as in the plants containing only the 35S promoter. Even though in the control line there was no homology between the gus transgene, regulated by the NOS promoter and CaMV, the GUS mRNA levels increased in those plants over time after CaMV infection (Figures 6 and 7). These results have already been observed and published (Al-Kaff et al., 1998). Our previous findings (Al-Kaff et al., 1998, 2000) were consistent with the silencing induced by CaMV on transgenes driven by the 35S promoter in oilseed rape. However we found that expression instability also occurred when there was no homology between the transgene and the infecting virus in the case of TL2, where the gus reporter gene was driven by the NOS promoter. 3.3 Interaction between virus coding regions of TuMV and CaMV infection In this study we used the two independent pakchoi transgenic lines carrying the NIa gene of TuMV (see TL3 in Figure 1A) regulated by the CaMV 35S promoter and terminator, and the selectable bar gene (herbicide tolerant gene), driven by the 35S promoter and transcriptionally terminated by the NOS terminator. Both lines were homozygous and had a single copy of the transgenes construct. Non-transformed pakchoi was included in these experiments as a control plant. Plants from each transgenic and non-transformed line were infected with the Cabb B-JI variant of CaMV. As a virus control, we used the CDN-1 variant of TuMV. This experiment was repeated twice. In order to establish a comparison with oilseed rape, we also analysed transgenic oilseed rape plants containing the same transgene construct (Figure 1A) after infecting them by Cabb B-JI or CDN-1, variants of CaMV and TuMV respectively. This was an attempt to study CaMV induced instability of transgene expression in a host that is considered to be very susceptible to CaMV infection and does not recover. 3.3.1 Symptoms induced by CaMV on infected transgenic lines of pakchoi and oilseed rape Systemic symptoms induced by CaMV infection were first observed in pakchoi plants after 7-10 d.p.i. and at about 10-12 d.p.i. in oilseed rape plants. By contrast with oilseed rape, pakchoi plants did not recover from CaMV infection. Rather, these plants developed very severe leaf chlorosis and in general they were stunted and distorted, including the flower heads (Figures 8 and 13A). It was also observed that transgenic pakchoi plants were more susceptible to CaMV in comparison with the nontransformed plants (Figure 13A). However, there were no differences observed in symptoms induced by CaMV in oilseed rape flowers between transgenic and non-transformed plants. We therefore studied the replication and accumulation in relation to symptoms in leaves and flowers of both oilseed rape and pakchoi to detect if there were changes at the molecular level (see 3.3.4). 3.3.2 Virus accumulation in transgenic lines of oilseed rape and pakchoi We used both Southern blot and slot blot analysis to study the replication and accumulation of the virus, respectively. Leaf samples were collected at different d.p.i., ground and then divided into CSG 15 (1/00) 7 Project title Non-target effects of transgenic crop plants resistant to virus diseases DEFRA project code CB02017 three samples, to study 1) replication, 2) virus accumulation and 3) transgene expression. To study CaMV replication 10 µg of total DNA was loaded into 1% agarose gel and then blotted and probed with the CaMV specific probe. Analysis of CaMV replication in the non-transformed and transgenic pakchoi infected plants (Figure 9A) by Southern blotting showed that there was no high accumulation of the supercoiled DNA (Sc) form, as had been observed in infected oilseed rape plants (Figure 9B). The total amount of viral DNA accumulated in transgenic and non-transformed plants of both oilseed rape and pakchoi was studied by slot blot and measured by comparing it with CaMV standards. DNA viral standards were extracted from virion viral preparations and measured by a spectrophotometer and slotted in replicate at different amounts from 0.10 ng to 25.00 ng (Figures 10A & 10C). The results showed that more CaMV DNA was detected in transgenic pakchoi plants infected only with CaMV compared with the non-transformed (Figure 10A), whereas in oilseed rape no significant differences were observed between non-transformed and transgenic plants (Figure 10C). However, comparison between nontransformed and transgenic pakchoi plants infected with TuMV also showed no difference in the accumulation of the viral TuMV RNA (Figure 10B). In general, mix infected pakchoi plants accumulated higher amounts of CaMV viral DNA in both non-transformed and transgenic plants. Transgenic and non-transformed pakchoi plants responded to the mix infection of both CaMV and TuMV by showing severe symptoms and high accumulation of CaMV DNA (Figures 8 & 10A). Increased susceptibility in transgenic pakchoi plants infected with CaMV could only result from an interaction between the NIa gene of TuMV and CaMV. This was also true when we infected nontransformed pakchoi plants with TuMV and CaMV (Figure 10A). 3.3.3 Instability of transgene expression after virus infection Northern blots and slot blots were performed on total RNA extracted from pakchoi and oilseed rape leaves at 18, 25 and 39 d.p.i. to assess the effect of virus infection on transgene expression (Figures 11 and 12). In all experiments total RNA was extracted using the Qiagen RNeasy Plant Mini Kit. Total RNA concentrations was then measured using a Spectrophotometer and also run in an ethidium bromide agarose gel to estimate the amount to be loaded and also to check the quality of the RNA using the ribosomal RNAs as an indication. In our previous work to study the effect of CaMV infection on transgene expression on oilseed rape plants, comparison of the RNA amount was estimated by the radioactive signal produced of the endogenous gene, tubulin as well as ribosomal RNAs amount. However, when we started comparing infected and non-infected transgenic pakchoi with oilseed rape after infection with TuMV, CaMV or both, we found that tubulin was affected by viral infection in the susceptible host pakchoi infected with both viruses. We also observed a similar effect in oilseed rape infected with severe CaMV isolates or TuMV. We therefore began to use actin, but also found similar results after viral infection. The increase of the expression in both tubulin and actin endogenous genes could be due to the severe distortion induced in pakchoi or oilseed rape by virulent variants of both viruses. As a result, for all expression studies in this project we estimated the loaded amount of total RNA using the Spectrophotometer readings as well as the intensity of the ribosomal RNAs in ethidium bromide agarose gels. The general observation for both bar and NIa transgenes was down-regulation in oilseed rape (Figures 12A and 12C) and up-regulation in pakchoi plants (Figures 11A and 11B) infected with CaMV, in comparison with the non-infected plants. The up-regulation of transgene expression detected in pakchoi plants was independent of virus type. Both CaMV and TuMV were able to increase the rate of transgene expression with time. Moreover, we found that in plants infected with both viruses, the transgene RNAs were accumulated at a higher level than in those infected only with one of the viruses at a time (Figures 11 A and 11B). We also found a relationship between the increase of transgene expression rate and time post-infection. At 39 d.p.i. for example, more transgene RNA was accumulated in comparison with earlier times post inoculation. As before, pakchoi plants infected with both viruses showed severe symptoms compared to those infected with only one of the viruses (Figure CSG 15 (1/00) 8 Project title Non-target effects of transgenic crop plants resistant to virus diseases DEFRA project code CB02017 8). We also detected more CaMV RNA in mix infected plants compared with plants infected only with CaMV at 39 d.p.i. (Figure 11A). For the NIa transgene it was not possible to determine transgene expression level, especially in plants infected with TuMV or both TuMV and CaMV, therefore we performed a Northern blot containing only non-infected plants and plants infected only with CaMV and then the blot was probed with a NIa specific probe. However, the up-regulation of NIa expression was not as significant as we found for the bar gene (Figure 11B). When the same experiment was performed on oilseed rape plants by Northern blot analysis (Figure 12C) we detected a clear reduction in transcription of the NIa gene. We also studied the expression of viral RNA for both CaMV and TuMV in pakchoi and oilseed rape. In both plant species, TuMV stimulated CaMV replication and accumulation at a late stage of infection (Figures 11A, 12A) However no clear stimulation of TuMV by CaMV infection was found (Figure 12B and 12 D) in oilseed rape. 3.3.4 Viral accumulation and transgene expression in pakchoi flowers The most severe symptoms observed in pakchoi plants infected with CaMV were located in the flower heads (Figure 13A). Pakchoi plants infected with TuMV also showed severe symptoms but with no severe distortion in the flower heads. Therefore, we analysed virus replication and transgene expression in pakchoi flowers. It was interesting to detect higher accumulation of inactive supercoiled (Sc) DNA forms and no replicative (R) forms in flowers in comparison with leaves of the same plant (Figure 13B). The level of the inactive supercoiled DNA accumulation in those flowers was similar to that found in leaves of oilseed rape infected with CaMV at the recovery stage. In oilseed rape, at the recovery stage, accumulation of supercoiled DNA and PTGS of transgenes induced by CaMV were found to be related (Al-Kaff et al., 1998). NIa transgene in TL3 plants had RNA homology to CaMV, therefore we thought that flowers of pakchoi could display the same phenomenon as found in leaves of oilseed rape, even though leaves of pakchoi showed transgene up-regulation. The results of the Northern blot analysis (Figure 13C) of transgenes showed high accumulation of bar but a reduction of the NIa gene. 3.3.5 Up-regulation of some endogenous genes in plants infected with CaMV and TuMV. In our work published over the past few years on oilseed rape we used actin and tubulin as internal positive controls for the expression in Northern blots, run-on assay and RT-PCR. From our experience, those genes do not change during CaMV infection in oilseed rape. However, we found changes on the expression of these genes when we started working with pakchoi. Actin and tubulin were both up-regulated in pakchoi plants infected with CaMV or TuMV (Figure 11A and 14). As we mentioned before, pakchoi is a very susceptible host to CaMV and plants do not recover from infection, whereas oilseed rape responds initially by showing systemic symptoms and then recovering from infection. CaMV induces a severe distortion in pakchoi but not in oilseed rape. Therefore, there will be more physiological and morphological changes in pakchoi than in oilseed rape that could affect the expression of endogenous genes. RNA samples were first measured by a spectrophotometer, then run in a normal agarose gel to estimate amount and quality of the RNA. The Northern blots were probed with both tubulin and actin, as endogenous controls. 3.4 Effect of virus type, virus variants and time of infection of CaMV on transgene expression 3.4.1. Effect of more than one virus type on pathogenicity and transgene expression. The occurrence of more than one infecting virus is common in plants and therefore the disease phenotypes displayed are normally caused by the interaction of more than one of the co-infecting viruses. Some of these interactions may have a synergistic effect, in which one of the viruses involved increases the replication and/or the transmission of others (Palukaitis and Kaplan, 1997), or cross protects against the infection by other viruses. To determine what sort of interaction occurs between CaMV and TuMV in our plant systems, we infected non-transformed and transgenic pakchoi and oilseed rape plants with mixed inocula of both viruses. We compared mix infected plants with the two CSG 15 (1/00) 9 Project title Non-target effects of transgenic crop plants resistant to virus diseases DEFRA project code CB02017 virus types to those infected with either virus alone. Results from these experiments are shown in Figures 6, 7, 8, 9, 10, 11A and 12. Plants were infected on the first true leaf with both viruses, and symptoms, virus replication and accumulation and transgene expression were studied. Transgenic plants of both plant species contained the NIa of TuMV and the herbicide tolerance gene bar (TL3 in Figure 1A). We also used oilseed rape transgenic plants containing gus and nptII genes (TL1 and TL2 in Figure 2). Pakchoi plants infected with CaMV and TuMV showed more severe symptoms than those infected with only one of the viruses (Figure 8). However, in oilseed rape infected with mix inocula we observed similar symptoms to those in plants infected only by TuMV (Figure 3A). These plants did not recover from TuMV infection, unlike the plants infected only by CaMV, which usually recovered after 21 d.p.i. It was also clear that CaMV titre and replication in most of the pakchoi mix infected plants tested was higher than those found in plants infected only with CaMV at late stage of infection (Figures 10A, 11A). In the case of oilseed rape, no significant difference in virus accumulation was detected between plants infected with both CaMV and TuMV, and plants infected only with CaMV (Figure 10C). In general, virus titre enhancement for CaMV and no accumulation changes for TuMV were detected in pakchoi plants infected with both viruses. We also studied NIa and bar gene expression in pakchoi and oilseed rape infected with CaMV and TuMV (see Figures 6, 11A, 12A and 12C). The results showed that in pakchoi plants infected with both viruses, expression of transgenes was up-regulated at a higher level than in plants infected with CaMV or TuMV. However, in oilseed rape we found that TuMV suppressed CaMV-induced transgene silencing. Therefore, we detected RNA of transgenes in plants infected with both viruses (Figures 6 and 12A). This could be due to the HCpro protein encoded by TuMV, which is a gene silencing suppressor. 3.4.2 Effect of different CaMV variants on transgene expression. We have previously shown that in oilseed rape CaMV-mediated transgene suppression depends on the viral pathogenic variant and time of infection (Al-Kaff et al, 2000). Surprisingly, transgenes with CaMV sequence homology in plants infected with a very severe CaMV isolate were less affected by the CaMV-induced transgene silencing. To investigate further the effect of virus genotype on transgene instability, we infected pakchoi transgenic plants with two CaMV isolates, Cabb B-JI and Aust, which caused moderate to severe and severe to very severe symptom responses, respectively (AlKaff and Covey, 1995). We also used another CaMV deletion mutant, M11, which had been tested and defined as a mild variant of CaMV. Systemic symptoms induced in all infected plants with CaMV variant, M11 (Mild), C-M (moderate), Cabb- BJI (severe) and Aust (very severe) developed as initial vein clearing and later as chlorotic vein banding with interspersed green tissues. However, virus-induced stunting and distortion were observed, especially in pakchoi infected with all CaMV variants and in oilseed rape infected only with Aust. In general, infected plants were more stunted than non-infected healthy plants and the most severe disease was caused by Aust CaMV variant in both pakchoi (Figures 15A and 15C) and oilseed rape (Figures 15B and 15C). Interestingly, the variant M11, although it was considered a mild variant, caused more severe symptoms in our plants than Cabb B-JI. One possible explanation for this observation is that the original CaMV mutant M11 reverted spontaneously to the wild type, since spontaneous mutagenic processes appear to be frequent in viral genomes (Steinhauer and Holland, 1987; Pennington and Melcher, 1993). We also found that transgenic plants of pakchoi were more susceptible to the CaMV variants than non-transformed plants. This was especially clear from observing the distortion induced in the flower heads of the infected transgenic pakchoi plants (Figure 16A). We observed similar kinds of distortion in oilseed rape, but only in plants infected with Aust variant of CaMV (Figure 16B). The effect induced by the CaMV variants on transgene expressions in both pakchoi and oilseed rape is shown in Figures 17A, 17B and 17C. The Northern blot analysis of total RNA from transgenic pakchoi CSG 15 (1/00) 10 Project title Non-target effects of transgenic crop plants resistant to virus diseases DEFRA project code CB02017 and oilseed rape plants at 18, 25, and 39 d.p.i. showed a clear relationship between CaMV virulence and the transgene up-regulation induced (Figures 17A and 17B). Plants infected with the very severe variant Aust presented the highest expression of the bar transgene. In oilseed rape we also found reversion of transgene silencing, as previously observed (Al-Kaff et al., 2000) with the severe variant. When we compared transgene expression with CaMV accumulation in plants where the expression was up-regulated, the virus also accumulated at higher level (Figures 18A and 18B). 3.4.3 Effect of CaMV on transgene expression when infection was carried out at different stages of plant development. In these experiments, we infected the 4th, 5th and 7th leaf of pakchoi transgenic and nontransformed plants with CaMV. Samples were collected at two time points;18 and 39 d.p.i (Figures 19A and 19B). Systemic symptom expression was compared in plants infected in the 4th, 5th and 7th leaf with plants infected in the 1st leaf. The results showed that there is a relationship between time of infection and symptom severity. The earlier we infected plants with CaMV, the more severe were the symptoms observed. Transgene expression was found to be up-regulated in plants infected on the 4th and 5th leaf in comparison to those infected on the 7th. This was also reflected in the amount of viral RNAs, 35S and 19S, accumulated at both 18 and 39 d.p.i. (Figures 19A and 19B). Plants infected in leaf 7 did not show up-regulation of the transgene. This could be due to the undetectable activity of the virus in those plants at both 18 and 39 d.p.i. (Figure 19B). 3.5 Determining cross-protection and the effect of virus type on transgene instability We infected plants in the first true leaf with one of the viruses, either CaMV or TuMV. Then after 10 days, when the plants showed symptoms of the corresponding virus, the same plants were inoculated with the respective second virus on the third or the fourth leaf. In three independent experiments involving a total of 30 non-transformed and transgenic pakchoi plants, a very low proportion of plants infected with the first virus supported infection by the second virus. This was true whether the infected virus was CaMV and then by TuMV or visa versa. Fairly similar results were observed with oilseed rape plants. It is therefore unlikely that CaMV and TuMV were protecting the plants from each other in either of the Brassica species studied. 3.6 Effect of virus infection on stacked transgenes We crossed two independent transgenic lines of pakchoi containing the TL3 construct (see Figure 1A). This cross contained two copies of each of the bar herbicide tolerance gene and the NIa of TuMV. All transgenes in this cross contained sequence homology with CaMV. We also used another cross between the oilseed rape containing constructs TL1 and TL2 (see Figure 2). This cross contained two copies of the reporter gene gus but only one of them had homology with CaMV. It also contained two copies of the selectable marker nptII with promoter homology to CaMV. Plants were infected in the first true leaf with one of the viruses, either CaMV or TuMV, and also with both viruses. The results are shown in Figures 6, 11 A and 11B, 14 and 21. In pakchoi plants infected with CaMV, TuMV or both viruses we detected a high accumulation of transgene RNAs in all plants tested (Figures 11A and 11B). However, in oilseed rape, CaMV induced transgene down-regulation affected only the transgene (gus) with promoter or terminator homology to CaMV. Interestingly we did not detect silencing on the second gus copy, which is driven by the NOS promoter. The expression of the gus transgene with the NOS promoter was higher in the infected than in the non-infected plants (Figure 6). This was interesting, since there was no homology between CaMV and the NOS promoter. Discussion CSG 15 (1/00) 11 Project title Non-target effects of transgenic crop plants resistant to virus diseases DEFRA project code CB02017 One of the main challenges in risk assessment is how we determine non-targeted or unexpected changes in the new crops or their products. The main aim of this project was to study non-target transgene expression instability in plants containing viral genetic transgenic material after virus infection. Plants in general are hosts for more than one virus and can be infected with more than one virus during their growing season. Pathogens, including viruses, induce a range of physiological and morphological changes when invading a plant and as a result gene expression patterns can also be affected. CaMV is a double stranded DNA virus and mainly infects members of the cruciferous plant family. It is important in plant biotechnology mainly because of the use of its 35S promoter. The 35S promoter has been used extensively to express genes in a range of crop plants some of which are specific hosts of CaMV. When CaMV infects transgenic oilseed rape (B. napus) plants containing transgenes driven by the 35S promoter, transgene expression is suppressed (Al-Kaff, et al., 1998 and Covey and Al-Kaff, 2000). Cruciferous species differed in their susceptibility to CaMV infection (Saunders et al., 1990 and Al-Kaff and Covey, 1995). In particular, oilseed rape is considered susceptible to CaMV at an early stage of infection but it recovers at a late stage (Al-Kaff et al., 1998). The recovery from CaMV infection was first discovered in B. oleracea which is one of the parents of oilseed rape (Covey et al., 1997). In this project we studied another Brassica crop plant, pakchoi (B. rapa). This species is the other parent of oilseed rape. It is considered as a susceptible host to CaMV and plants do not recover from infection. When a transgene with the 35S promoter in the pakchoi genetic background interacted with CaMV, we detected up-regulation of transgene expression. CaMV infection can induce transgene instability in at least two ways, by suppressing or by upregulating transgene expression, depending upon the host species. These kinds of interactions could be the result of sequence homology between the virus and the transgene. In this case, the infected host recognises the virus and transgene in a similar way. Therefore, when e.g. oilseed rape recovers from CaMV infection, the virus is suppressed at the PTGS level (Covey et al., 1997). This was also true for transgenes with RNA homology to CaMV (Al-Kaff et al., 1998). However, we found that the upregulation induced in pakchoi plants was not specific to CaMV infection because TuMV was also able to induce it. Therefore, the viral inducible transgene up-regulation observed was not due to the sequence homology between the transgene and the infecting virus. Such interactions can be referred to as non-target or unexpected effects on transgenes. In infected plants, as we presented in the results section, there were severe morphological changes in infected pakchoi and oilseed rape, although pakchoi was more severely affected by either CaMV or TuMV than oilseed rape. Moreover, virus virulence was even more pronounced when both viruses coinfected pakchoi plants, and as a result we detected in these plants increased transgene expression in comparison with plants infected with only one of the two viruses. It was also interesting to observe that there was a relationship between virus virulence and transgene up-regulation. Plants with very severe systemic symptoms had higher a transgene expression. This was clear when we infected plants with virus variants causing a different range of symptom severity. There were also other unexpected changes in transgene expression where sequence homology was not involved. Transgenes in oilseed rape driven by the NOS promoter were also up-regulated in plants infected with CaMV. Interestingly, in oilseed rape plants containing two copies of the gus gene, one driven by the 35S promoter and other by the NOS promoter, only the transgene with RNA homology to 35S promoter CaMV was silenced. The similarity between the two cases of transgene up-regulation presented above is that the 35S and the NOS promoters are both pathogenic promoters and they could be recognized by the host defence machinery after pathogen invasion, in this case, CaMV and TuMV. If this is true, it would be interesting to study the effect of other pathogens, e.g. fungus and bacteria, on the activity of these promoters or other pathogen-related promoters controlling transgene expression. CSG 15 (1/00) 12 Project title Non-target effects of transgenic crop plants resistant to virus diseases DEFRA project code CB02017 Another unexpected result found in this research was that CaMV-mediated PTGS did not silence homologous transgenes. We had an oilseed transgenic line containing two copies of the gus reporter gene, the first flanked by the 35S and the terminator of CaMV and the second by the NOS promoter and the OCS terminator. After CaMV infection we still detected the NOS:gus but not the 35S:gus. It is known that PTGS mechanisms in other systems recognise homologous sequences and as a result they are silenced. This work showed the importance of studying stress conditions, including pathogen infections, on transgene expressions in crops improved by genetic manipulation. Pathogen elements or genes are a powerful tool which can be employed to improve crop characters, but their use should be based on sound scientific knowledge, to inform their scientific assessment. References Al-Kaff, N.S. and Covey, S.N. (1995) Biological diversity of cauliflower mosaic virus isolates expressed in two Brassica species. Plant Pathol. 44, 516-526 . Al-Kaff, N. S., Covey, S. N., Kreike, M. M., Page, A. M., Pinder, R. and Dale, P. J. (1998). Transcriptional and posttranscriptional plant gene silencing in response to a pathogen. Science 279, 2113-2115. Al-Kaff, N. S., Kreike, M. M., Covey, S. N., Pitcher, R., Page, A. M. and Dale, P. J. (2000). Plants rendered herbicide-susceptible by cauliflower mosaic virus-elucited suppression of a 35S promoterregulated transgene. Nature Biotechnol. 18, 995-999. Baulcombe, D. C. (1996). Mechanisms of pathogen-derived resistance to viruses in transgenic plants. Plant Cell 8, 1833-1844. Covey, S. N. and Al-Kaff, N.S. (2000). Plant DNA viruses and gene silencing. Plant Molecular Biology.43: 307-322. Covey, S. N., Al-Kaff, N. S., Langala, A. and Turner, D. S. (1997). Plant combat infection by gene silencing. Nature 385, 781-782. Falk, B.W. and Bruening, G. (1994). Will transgenic crops generate new viruses and new diseases? Science 263:1395-1396 Hammond, J., Lecoq, H. and Raccah, B., 1999. Epidemiological risks from mixed virus infections and transgenic plants ex-pressing viral transgenes. Advances in Virus Research 54:189-314. Hardwick, N.V., Davies, J.M.L. and Wright, D.M. (1994) The incidence of three virus diseases of winter oilseed rape in England and Wales in the 1991/92 and 1992/93 growing seasons. Plant Pathol. 43, 1045-1049. Hii, G, Pennington, R., Hartson, S., Taylor, C.D., Lartey, R., Williams, A., Lewis, D. and Melcher, U. (2002) Isolate-specific synergy in disease symptoms between cauliflower mosaic and turnip veinclearing viruses. ArchVirol.147(7):1371-84. Hull, R. (1990) Virus resistant plants: potential and risks. Chem. Ind. 17, 543-546. Hull, R. and Covey, S. N. (1983). Characterization of cauliflower mosaic virus DNA forms isolated from infected turnip leaves. Nucleic Acids Res. 11, 1881-1895. CSG 15 (1/00) 13 Project title Non-target effects of transgenic crop plants resistant to virus diseases DEFRA project code CB02017 Khan, M.A. and Demski, J.W. (1982). Identification of turnip mosaic and cauliflower mosaic viruses naturally infecting collards. Plant Disease 66, 253-256. Lomonosoff, G. P. (1995). Pathogen-derived resistance to plant viruses. Annu. Rev. Phytopathol. 33, 323-343. Liu, F., Cao, M. Q., Yao L., Li, Y., Robaglia, C. and Tourneur, C. In planta transformation of Pakchoi (Brassica campestris L. ssp Chinesis) by infiltration of adult plants with Agrobacterium. Acta Hortic. 467, 187–193 (1998). Maiti, I. B., Murphy, J. F., Shaw, J. G. and Hunt, A. G. (1993). Plants that express a potyvirus proteinase gene are resistant to virus infection. Proc. Nat. Acad. Sci. USA 90, 6110-6114. Melcher, U. (1989) Symptoms of cauliflower mosaic virus infection in Arabidopsis thaliana and turnip. Botanic Gazette 150, 139-147. Palukaitis, P. and Kaplan, I. B. (1997). Synergy of virus accumulation and pathology in transgenic plants expressing viral sequences. In Virus-resistant transgenic plants: Potential ecological impact. M. Tepfer and E. Balazs (Eds.), 77-84. Pennington, R. E. and Melcher, U. (1993). In planta delection of DNA inserts from the large intergenic region of cauliflower mosaic virus DNA. Virology 192, 188-196., Raybould, A. F., Maskell, L. C., Edwards, M. L., Cooper, J. I. and Gray, A. J. (1999). The prevalence and spatial distribution of viruses in natural populations of Brassica oleracea. New Phitol. 141, 265275. Rubio, T., Borja, M., Scholthof, H.B., Feldstein, P.A., Morris, T.J. and Jackson, A.O. (1999) Broad-spectrum protection against tombusviruses elicited by defective interfering RNAs in transgenic plants. J. Virol. 73, 5070-5078. Saunders, K., Lucy, A.P. and Covey, S.N. (1990) Susceptibility of Brassica species to cauliflower mosaic virus infection is related to a specific stage in the virus multiplication cycle. J. Gen. Virol. 71, 1641-1647. Scholthof, H. B., Scholthof, K. B. G. and Jackson, A. O. (1996). Plant virus gene vector for transient expression of foreign proteins in plants. Annu. Rev. Phytopathol. 34, 299-323. Steinhauer, D. A. and Holland, J. J. (1987). Rapid evolution of RNA viruses. Ann. Rev. Microb., 41, 409-433. Tepfer, M. (2002) Risk assessment of virus-resistant transgenic plants. Annu. Rev. Phytopathol. 40, 467-491. Vardi, E., Sela, I., Edelbaum, O., Liben, O., Kuznetsova, L. and Stram, Y. (1993). Plants transformed with a cistron of a potato virus Y protease (NIa) are resistant to virus infection. Proc. Nat. Acad. Sci. USA 90, 7513-7517. Walsh, J.A. and Tomlinson, J.A. (1989).Viruses infecting winter oilseed rape (Brassica napus ssp. oleifera). Annals of Applied Biology 110, 661-681 CSG 15 (1/00) 14 Project title Non-target effects of transgenic crop plants resistant to virus diseases CSG 15 (1/00) 15 DEFRA project code CB02017