N.B. This page is under construction (awaiting images).
Pia Malnoë, Gŕbor Jakab, Eric Droz and Fabian Vaistij Station Fédérale de Recherches en Production Végétale de Changins, CH 1260 Nyon et l'Université de Genčve
Artwork: Joanna Painter for Nestle Alimentarium, Vevey
I would like to start by giving you an introduction on the life cycle of Potato Virus Y (PVY), then move on to the resistance mechanism in the transgenic virus resistant plants and finally cover RNA recombination between two different strains of PVY able to infect transgenic plants.
Under the electron microscope, PVY appears as long filaments containing a nucleic acid (RNA) molecule protected by about 2,000 coat protein molecules. When the virus enters a plant cell, having been transmitted by an aphid or by mechanical inoculation, its coat is removed. The virus immediately initiates the synthesis of several proteins necessary for its survival in the plant. One of the most important ones is the replicase, which allows the virus to copy itself. When the viral genome has been copied to a sufficient number, some copies are coated again and some remain as naked RNA molecules. The latter can move into an adjacent cell and start the whole process over again. Coated virus particles can travel up in the plant to infect its upper regions. This is called a primary infection and does not usually produce any serious symptoms. In August, when the plant has almost finished its vegetative cycle, the virus migrates down to the tubers and hibernates.
In spring, the tubers germinate and the virus wakes up. It is now able to take over the whole plant and rapidly give rise to a secondary infection where the symptoms are far more serious. Aphids will transfer the virus produced in the secondarily infected plants to healthy plants where the whole cycle is repeated.
The best way to control a viral infection in the field is to use plants that are resistant to the virus. Such plants can be obtained by introducing natural resistance genes by classical breeding methods. However, these genes are often linked to undesirable character traits. Some ten years ago, it was suggested that genetic engineering could be used to increase virus resistance in plants. The idea was based on the observation that inoculation with a non-virulent virus strain conferred protection from infection by a virulent strain. This phenomenon is called cross protection and has been used to protect crops in the field. Instead of inoculating the whole virus into the plant, the insertion of only part of the viral genome into the plant chromosomes was expected to provide a protection against the virus. This was confirmed and transgenic virus resistant plants expressing a viral sequence are now starting to appear on the market.
The PVY genome consists of an RNA molecule of 10,000 bases. In the plant cell it acts as messenger RNA that directs the synthesis of a polyprotein which is then cleaved by proteases into 9 different polypeptides. Each of these polypeptides plays a specific role in virus life cycle. Among these are the capsid or coat protein, the replicase, several proteases and so on. We cloned the cDNA sequence corresponding to the coat protein gene and introduced this sequence into the potato chromosomes using Agrobacterium tumefaciens as a vector. After having regenerated transformed plants, we monitored their resistance in the greenhouse. Some transgenic lines showed very good resistance to PVY. One of these, called Bt6, was tested in the field in 1991/1992. Its appearance was normal and we could not observe any differences at the level of the plant itself. However in the beginning, we could see a slight difference in the tuber's shape, but that tended to disappear later on. We measured the amount of virus in the transgenic and control Bintje potatoes. In the untransformed plants there was a high accumulation of virus in almost 100 % of the plants compared to 22% of the transgenic plants. The tubers obtained from these plants were germinated and tested for the virus. 100% of the untransformed tubers were infected but no virus could be detected in the transgenic tubers. The conclusion of this work is that the presence of the PVY coat protein gene in the potato genome protects the plant from both primary and secondary PVY infection.
We wanted to know why the transgenic plants were protected. To study the resistance mechanism, we have been using tobacco instead of potato because it is easier to work with this plant and it also belongs to the Solanaceae family. We obtained a transgenic tobacco line (4B5) which is resistant to all PVY strains tested so far. The 4B5 line was self-pollinated and we analysed 50 transgenic R1 plants for PVY resistance. The ELISA test, which detects the virus, showed that five of the R1 plants had lost the resistance although they still contained the transgene. These 5 plants were studied in more detail by Southern and Northern hybridisations using the PVY coat protein gene as a probe. We found that in these plants the transgene was heavily expressed but in plants that still showed resistance there was no PVY mRNA detectable. This was an unexpected result but it is important to realise that in resistant plants there is very little transgenic viral mRNA present. The full significance of this will emerge later.
A similar phenomenon referred to as gene silencing has been described in the literature. Several years ago, when a group of scientists were introducing supplementary copies of the chalcone synthase gene into petunia in order to increase the violet colour of the flowers, in some cases they obtained flowers with a more intense violet colour but also transgenic lines with white flowers. In these plants, both the endogenous and the transgenic chalcone synthase genes were turned off. This is an example of gene silencing which has since been observed for other genes in many other transgenic plants. Silencing, or inactivation of homologous genes or transgenes, is obtained through a transcriptional regulation. In some cases it has been shown that methylation of the transgene hinders its transcription. In other cases, there is a post transcriptional regulation involving a specific degradation of the transgenic messenger RNA in the cytoplasm.
The resistance to PVY seems to be regulated at a post transcriptional level because the coat protein gene is heavily transcribed in the nucleus but no coat protein mRNA can be detected in the cytoplasm by Northern hybridisations in the resistant plants. However, in the plants that had lost resistance, a large amount of coat protein mRNA could be observed in the cytoplasm. Several scientists, like David Baulcombe at the Sainsbury laboratory in Norwich have proposed a cellular compound, such as an RNAse, which would destroy the coat protein mRNA. A specific sequence at the 3' end of the mRNA seems to be recognised by this " RNAse". When the virus enters a resistant cell the RNAse - (whose existence still has to be demonstrated ) - which normally destroys the transgenic mRNA also attacks the viral RNA infecting the cell and inactivates it. This is our working hypothesis for the time being.
Concerning the risks associated with the use of viral resistant transgenic plants, two different types of risks have been identified. One is heterologous encapsidation and the other is RNA recombination. I would like to focus on RNA recombination which is the only one giving rise to a permanent change in the viral genome.
When a virus infects a plant that synthesises a transgenic mRNA, the viral replication process can also be initiated on the transgenic mRNA if this molecule contains a polymerase binding site. At a certain moment, the replicase might switch template to the viral genomic RNA, and this way produce a recombined virus. However, if the replicase binding site is missing from the transgenic mRNA, the replication will start on the viral genomic RNA, then switch to the transgene and finally return to the viral RNA template. Again a recombined virus results. This kind of template switching also happens when two closely related viruses infect the same plant. This is a normal process in nature and a way for the virus to get additional genetic information to ensure its evolution. The environment will determine which combination is the most fitted.
We have been studying viral recombination between a transgenic mRNA and a genomic viral RNA. In our model system, we need a double recombination event in order to obtain an infectious recombinant virus. We used a transgenic plant expressing the coat protein gene (hence sensitive) and a virus with a deletion in the same gene. This modified virus cannot migrate upwards in the plant because it cannot be encapsidated. Therefore, to infect the plant systemically, a recombination event is needed between the transgenic mRNA and the viral RNA. In order to have enough mRNA available for the recombination event, we cannot use a resistant plant as a host because the viral RNA is degraded in such a plant. Instead we have to use a plant that has lost its resistance but still contains the coat protein gene. Such a plant will synthesise a large amount of coat protein mRNA which will be translated into the coat protein and encapsidate the mutated virus. This means that a systemic infection can develop in the transgenic plant. To be able to demonstrate that the virus with the gene deletion has replaced the missing part of its coat protein gene by recombination, we then need to transfer this infection to a normal untransformed plant which is not able to complement the dysfunctional coat protein. This experiment has now been going on for over six months in the greenhouse and we have not yet been able to detect such a recombination event. However the experiment indicates that a transgenic plant which synthesises a functional coat protein creates a new ecological environment. Put more clearly, this means that a mutated virus will be able to survive in a transgenic plant if the transgene can complement the viral mutation. However, as soon as it is transferred to a normal plant it cannot survive. This means that growing transgenic plants on a very large scale modifies the ecological environment.
In order to avoid this phenomenon, we only introduced parts of the coat protein gene into the genome of a potato variety called Matilda. The part which includes the 3' end of the RNA polymerase gene and the first bases of the coat protein gene produces a good resistance against the N strain of PVY. This means that one can indeed introduce only a small parts of the coat protein gene into the plant genome and still get a good protection. In these plants no coat protein synthesis is possible and hence there is no risk of heterologous encapsidations.
Both untransformed and transformed plants can be infected by two different strains of PVY. Although these two strains might show several differences in the amino acid sequence of the coat protein, they are closely related. In this experiment we have been using a transgenic tobacco (R1-28), which synthesises very little transgenic mRNA and is resistant to the PVY-N but not to the PVY-O strain. A co-infection was carried out with the two strains. Unexpectedly, 4 out of 13 infected plants developed symptoms of a PVY N infection. These results could be explained if a recombination had happened between the two PVY strains. The sequence in the PVY-N strain which is recognised by the RNAse (described above) could have been exchanged for the corresponding PVY-O sequence which is not recognised by that RNAse since PVY-O is able to infect the R1-28 plant. This was confirmed by a straight forward polymerase chain reaction (PCR). The a and b primers can detect the N strain, the c and d primers the O strain and the combination of the primers a and d a recombined virus. As expected, all the plants were infected with PVY O. Only one plant, which had lost the resistance, was infected with PVY-N (this plant was then left aside). With the primers a and d, four plants gave a positive result and among these, three showed PVY-N symptoms. To determine whether a second recombination event had taken place, we tested for the presence of an additional part of the N genome by using primers e and f. This region could not be detected in one of the four recombinants, namely the one that did not show any N symptoms, this means that a double recombination event had taken place in this plant
Among the four different recombination events, two were very close to the 5' end of the coat protein gene and the other two were further downstream. We have looked for recombination between PVY-N and other PVY-O strains, and found the same high frequency of recombination but not at the same sites.
It is important to realise that, in nature, recombination between related viral strains occurs all the time. Such recombination events increase the sequence variability and allow the virus to adapt to a new situation. However, under identical external conditions the viral sequence is remarkably stable. The observation of recombined virus particles in four out of thirteen double infected transgenic plants indicates a high frequency of recombination. In theory, we do not expect the recombination frequency to be different in an untransformed and a transformed plant. However the external selection pressure is different in the two types of plants. The recombined PVY-N has the advantage of having eliminated the sequence which is recognised by the "specific RNAse" and hence is not degraded in the transgenic plant.
In conclusion: transgenic virus resistant plants create a new ecological environment, and new viruses will appear since there is a strong selection for recombinants. This is most probably also true for "natural" virus resistant plants. However, it has not been studied as carefully as for transgenic plants.
This raises the following questions: how could this knowledge be integrated into a responsible risk assessment programme and which kind of cultural strategies would be appropriate in order to avoid the establishment of new viruses in our ecosystem?