Identifying and Silencing Tomato Ripening Genes with Antisense Genes
Don Grierson
DOI: https://doi.org/10.1111/pbi.12463
IF: 13.8
2015-01-01
Plant Biotechnology Journal
Abstract:At Sutton Bonington, we demonstrated that ripening involves changes in gene expression. In partnership with Zeneca (now Syngenta), we cloned, sequenced and identified several important tomato ripening genes, including those encoding polygalacturonase (PG), pectinesterase, phytoene synthase and ACC oxidase, which catalyses the last step in the synthesis of the hormone ethylene. We were among the first to inhibit gene expression by silencing with sense and antisense constructs, characterized several gene promoters, made transgenic plants with altered traits, showed that the genetic modifications were stably inherited, and Zeneca developed the first genetically modified food approved for sale on both sides of the Atlantic. The majority of relevant papers are cited in Grierson et al., 1986a; Grierson and Schuch, 1993; and Grierson et al., 2012, which also clarifies some confusing nomenclature changes. I joined the Plant Physiology section at the University of Nottingham Sutton Bonington (SB) Campus, as an assistant lecturer, aged 25. Harry Smith had recently been appointed professor and gave me great encouragement. Later, he moved to Leicester University, but I remained at SB. The first 10 years were the hardest, but funding improved significantly after the first Research Assessment exercise in 1986 when our department was one of only three at Nottingham to be rated outstanding. In 1973, Nithiya Rattanapanone, at that time a PhD student in food sciences, asked whether I would help her study RNA metabolism in tomatoes. We worked together and showed that, after a quiescent period, there was a phase of active RNA synthesis at the onset of ripening. This was surprising to some, who thought ripening and senescence were essentially degradative processes. With the help of Jim Speirs, we used in vitro translation to show the presence of new mRNAs. These discoveries led to a few papers and enabled me to get a substantial grant from the Agricultural Research Council to understand the role of cell wall modifying enzymes produced during ripening, with the help of Neil Robertson and Greg Tucker. Later, appreciating the need to move into gene cloning and sequencing, I contacted a company called Imperial Chemical Industries (ICI), which metamorphosed over the years into Zeneca, then Syngenta. The ICI corporate laboratory in Runcorn had initiated a long-term strategy to collaborate with a few plant scientists (including the late Chris lamb and I) in order to explore the potential for combining genetic and chemical approaches in crops. I met with Wolfgang Schuch, Ian Bridges and later Simon Bright, to discuss working together. The partnership began when we gave ICI/Zeneca mRNA from different stages of fruit development and ripening, from normal fruits and mutants and Keith Edwards did the first cDNA cloning and John Ray and Colin Bird started the DNA sequencing. The key start date was late 1983, when the TOM series of cDNA clones was generated. Seven years of work by Adrian Slater, Andrew Hamilton, Chris Smith, Rupert Fray, Julie Gray, Kevin Davies, Martin Maunders, Steve Picton and others together with Zeneca staff, followed. Soon, we were exchanging expertise and personnel, getting collaborative research council grants, publishing joint papers and doing the conference circuits. I was electrified at a conference in Liege, Belgium, in 1974, when Jeff Schell described the molecular and genetic characterization of the Agrobacterium Ti plasmid. When Mike Bevan's Bin19 Ti plasmid system became available soon after the ICI collaboration began, we planned to overexpress our genes and test fruit-specific gene promoters. Later, my PhD student Rupert Fray sequenced BIN 19 and explained the origin of several unexpected bits (Fray et al., 1994). The main problem we faced, however, was in identifying the tomato DNA sequences that we had. Nowadays, of course, the sequencing is performed by machine in a flash and one just asks the computer what the genes are. In those days, sequencing was still a significant task and there was no database. The PG cDNA and gene were sequenced by Colin Bird; Greg Tucker had purified the enzyme and made an antibody, and we were able to get an N-terminal amino acid sequence done at Leeds. The other genes were harder to identify. We realized that we could not make sufficient mutants fast enough and this forced us to try something different, so we thought of antisense. All one needed to do to generate antisense RNA was to use a transgene with the coding sequence reversed downstream of the CaMV 35S promoter. Our first funding application for antisense experiments was turned down. I suspect that part of the argument against funding it was based on an incorrect assumption about the mechanism of action. Some people expected antisense RNA would base pair with the target mRNA and prevent its translation, but PG mRNA was very abundant, so it might be difficult to make sufficient antisense RNA. Fortunately, a later application with ICI to the Science and Engineering Research Council was funded. Chris Smith was appointed as the postdoc to run the project and I still remember, after what seemed like a very long time, the morning in the autumn of 1987 when Chris showed me the fantastic result; a major reduction in fruit PG mRNA and protein in his transgenic tomatoes! (See Figure 1 and Smith et al., 1988.) We were able to show that antisense was remarkably effective and actually caused the mutual destruction of the antisense RNA and the target PG mRNA. Bill Hiatt and others at Calgene were also working on the PG problem, as were Allen Bennett and Dean DellaPenna. Both groups worked in Davis California, but did not seem to cooperate. Calgene's first tomato cloning paper cited, I think, eight of ours. Our group was the first to publish the PG cDNA gene sequence in 1986 (Grierson et al., 1986b), to characterize the PG gene promoter (Bird et al., 1988) and successfully knockout the PG mRNA by antisense (Smith et al., 1988); in this last paper, we wrote: ‘Our results indicate that PG antisense RNA generated in transgenic tomato plants causes a reduction in PG enzyme activity when the fruit ripen. …The precise mechanism of inhibition of enzyme synthesis is as yet unknown. The low level of PG mRNA … could be caused by the selective degradation of double-stranded RNA hybrids.. this approach is a practical general strategy for studies on development, gene expression and plant breeding’. We later proposed (Grierson et al., 1991v) that double-stranded RNA might be involved in both sense and antisense gene silencing and showed that inserting an inverted repeat into an antisense construct, which would generate double-stranded RNA, greatly enhanced the silencing effect (Hamilton et al., 1998). While we were doing this work, two other laboratories were doing similar experiments working on Petunia flower colour. We went on to identify and characterize several plant genes, in addition to PG and ACO, including pectinesterase (Julie Gray), phytoene synthase (Zeneca and Rupert Fray), and later lipoxygenases and alcohol acyl transferases. By knocking them out by gene silencing, we were able to study their roles, and showed that providing the selection and characterization was done carefully the silencing was stable and inherited. It is now clear that gene silencing can be achieved by antisense genes, sense genes and RNAi. All three methods depend on production of antisense RNA to generate RNA/RNA hybrids. These are then cut into small RNAs that amplify the antisense signal and target mRNAs for degradation in a sequence-specific manner. Sense and antisense methods were operating in plants 10 years before RNAi. Fire and Mello, who showed how to use it in worms without genetic transformation, were awarded a Nobel Prize in 2006. Polygalacturonase is involved in cell wall metabolism during ripening and was an obvious commercial target. Although the scientists at Calgene were generous in referencing many of our papers on fruit ripening in their publications, their managers were not so courteous, and referred to ‘those Bozos from Britain’ when our commercial partners visited the USA. There was a lot of activity regarding patents, regulations and agreements, largely taken care of by Zeneca people, but I did go to the US patent office on one occasion to give them a scientific presentation. Calgene had got to the US patent office several weeks before ICI, so they got the antisense patent and used it to develop the Flavr Savr tomato. The idea was that by reducing PG, tomatoes could be left longer before harvesting, to develop better flavour, while slowing the deterioration that accompanies ripening. Unfortunately, this was not a success, because although PG had seemed a promising candidate, it turned out that it was not the major determinant of fruit softening. Zeneca were granted a patent for silencing PG with sense genes so each side in the PG wars got a commercial opportunity. (Sense gene silencing, called co-suppression by others, achieved the same result as antisense, possibly because it can lead to antisense RNA production resulting from insertion events.) Zeneca showed tomato puree from PG-silenced tomatoes had improved viscosity (there was much less PG to cut the polygalacturonic acid) and reduced waste and could save energy and improve flavour because it was not necessary to heat-inactivate the PG at high temperatures during processing (Figure 2). This was the first GM food to be approved for sale on both sides of the Atlantic and at first was very successful. Around that time, NGOs began to criticize GM crops (I believe without any merit or real justification for their criticisms), with Monsanto as a major target. A host of critical stories appeared in the media, and some opponents attempted to destroy the market for GM crops, reasoning that if they stirred up enough controversy supermarkets might become worried about overall loss of sales. The GM puree sold particularly well in university towns, however, and Simon Bright says ‘The GM puree actually continued to sell in the midst of the media firestorm, but the product lapsed after failure to get clearance for growing the tomatoes in the EU’. Our GM research taught us a very great deal about plant physiology, plant biochemistry and genetics, and during this period, I appeared on a dozen or so TV shows and did around 40 radio interviews, explaining the science behind our GM work. I shall write elsewhere about the anti-GM campaigners, some of whom said, incorrectly, that the puree was not labelled (Figure 3). Even before the PG result, a PhD student, Andrew Hamilton, started to use antisense to try and identify a gene involved in the synthesis of the hormone ethylene. This work was carried out concurrently with the PG project and led to the identification of the ACC oxidase (ACO) genes, encoding the enzyme that catalyses the final step in the synthesis of the hormone ethylene. At that time, ACO was called the ethylene-forming enzyme. Its identification by antisense was important because the enzyme had never been purified, there was no amino acid sequence and no antibody. It was the first time that any gene had been identified by an antisense knockout. The work began with a systematic search by Chris Smith, Adrian Slater and Kevin Davies, looking for mRNAs expressed with the right kinetics and location to be a possible candidate sequence associated with ethylene synthesis. We only found one, called TOM13. Andrew Hamilton then knocked it out in transgenic tomatoes and showed ethylene production was inhibited by around 95%. It took Andrew 2 years before he got a result and he published in Nature (Hamilton et al., 1990) and the following year, with Mondher Bouzayen, he expressed the enzyme in yeast and confirmed the catalytic activity and stereospecificity (Hamilton et al., 1991). The ACC oxidase gene (ACO) was a major discovery as ethylene is important throughout plant development, stress responses, ripening and leaf and flower senescence and spoilage. During question time after speaking at a conference in Europe about Andrew's result, Michael Bevan asked me to comment on the fact that ethylene researchers believed ACO was associated with membranes but our result predicted a soluble protein. I took a deep breath and replied: ‘Well, Mike, either we have made a mistake, or everyone else is wrong!’ Michael Holdsworth showed there were several different genes for ACO (TOM13), and later, we identified a transcription factor that controls the expression of the ACC oxidase gene 1 in ripening fruit. We also showed that by reducing ethylene synthesis, we could reduce the rate of deterioration of fruit, and leaves, prolong storage life and reduce waste. But, preventing ethylene completely has an adverse effect on flavour, so it is important not to reduce it too much. I was lucky enough to have over 100 scientists, more than half of them PhD students, working with me at SB, supported by around 50 competitive grants, which meant that I had sufficient funds to pay salaries, buy equipment, and pay the cost of expensive consumables that we needed for experiments. I personally gave over 200 invited talks to Universities, Institutes, and national and international conferences. With Zeneca personnel, I was named as an inventor on over a dozen plant biotechnology patents. Not everyone cited our papers, however. An extreme example occurred in 2012, when over 20 years after, we showed we had identified the genes for ACC oxidase, knocked them out, and inhibited ethylene synthesis in transgenic plants, a paper was published (see Grierson et al., 2012) reporting the cDNA sequence (again!) and predicting how it might be possible to use it to inhibit ethylene production and achieve what we had already done! The SB Campus is now four to five times the size it was when I joined. The University built us a new Plant Sciences building, opened in 2002 by the then UK Minister for Science, Lord David Sainsbury. I became Pro-Vice-Chancellor for research and retired from full-time work in 2010, as an emeritus professor. Liz and I now spend about 25% of our time in China, where I work part-timve at Zhejiang, China Agricultural, and Shanxi Agricultural universities. I believe our collaboration with industry was mutually beneficial to academia and industry, involving excellent science, with an eye to applications. This was a hallmark of the SB set-up and instilled in some an interest in a career in industry. Above all, it was a lot of fun! I am greatly indebted to the many colleagues who worked with me and contributed to our success during 40 years at SB. Thanks are also due to Claire Grierson and Simon Bright for helpful comments and suggestions on the manuscript.