Research Unit Leader
Plant Breeding and Cell Engineering Research Unit
RIKEN Advanced Science Institute
New varieties of fruit trees and ornamental plants created through painstaking work are being illegally cultivated and marketed at low prices—a serious problem that infringes on breeders’ rights. To enable the identification of different varieties and prevent illegal cultivation, a broad range of research and development at the DNA level has been conducted so far. In the case of fruit trees and ornamental plants propagated by grafting or cuttings, however, no useful method of identification has been established because new varieties and their parents often share the same DNA. Against this background, the DNA marking technique developed by Tomoki Matsuyama, Research Unit Leader in the Plant Breeding and Cell Engineering Research Unit at the RIKEN Advanced Science Institute, is drawing attention as a solution to the recently emerging issues concerning food safety, such as deceptive labeling of production centers.
Distinguishing among different varieties
Matsuyama says, “When I was a university student, I studied horticultural science. The theft of seedlings and shoots of excellent varieties of tangerine and other fruit trees has been problematic among their breeders and producers. Fruit trees permit propagation, a method of expanding plants easily and limitlessly by grafting. Even if great efforts are put into creating a new variety, in some cases the same variety is being produced and marketed by someone else.” And if you suspect a tree of being a progeny from your own, possibly stolen, shoots, no method has been available to prove it. “Wanting to solve this problem, I began developing a method of identifying different varieties.”
Although different varieties are distinguishable by examining the proteins produced by cells, this method does not ensure accurate identification because the proteins are altered by even slight changes in the environment. Matsuyama began by making trial-and-error attempts, recognizing the necessity for an index that is not influenced by the environment.
“In those days, DNA fingerprinting began to be used to identify people, often creating a media stir. I thought that this technique could be applied to the identification of fruit tree varieties.” The DNA fingerprinting method was proposed by Alec Jeffries of the University of Leicester, UK, in 1984. It is based on the fact that several to more than a dozen base sequences are repeated in the noncoding region of the DNA, and that the number of repeats varies from person to person, enabling personal identification by examination of the repeated units. Matsuyama promptly applied this method to the identification of closely related varieties of mandarins and apples, but was not successful.
To understand why, we can consider the hybridization of two conventional varieties of plants to produce two new varieties. Because the new varieties carry different genetic information, they are easily distinguishable by examining their DNA (Fig. 1). However, in the case of fruit trees and ornamental plants that permit propagation—by grafting and taking cuttings for example—any mutant with a slightly different but useful characteristic can be registered as a new variety. A mutation of a mandarin tree could, for example, produce a shoot that bears fruit one week earlier than the other shoots, and another shoot that bears sweeter fruit. If the two mutant shoots are grafted and cut, it is possible to produce two new varieties. “In this case, although they are handled as different varieties, they represent clones with almost the same DNA. They are indistinguishable. My investigation of variety identification at university ended at that point, and I shifted to chromosome research using repeated sequences.”
The virtual RLGS system for mutation analysis
Matsuyama has been engaged in developing the restriction landmark genomic scanning (RLGS) system at RIKEN since 1996. “This is the RLGS,” says Matsuyama, pointing to a film on the desk (Fig. 2). The RLGS system is a highly accurate method of DNA analysis that had been extensively investigated by Yoshihide Hayashizaki, Director of the RIKEN Omics Science Center, and his colleagues. The film indicated by Matsuyama shows black dots of different sizes and properties, representing several hundred DNA fragments, obtained by two-dimensional electrophoresis. If a mutation occurs, the number, size and properties of the DNA fragments change, resulting in an altered pattern of dot distribution. Comparing the analytical results between normal individuals and mutants, it is possible to easily locate the mutations. The RLGS system has so far been used mainly in mouse research.
Matsuyama then created a new version of the RLGS system specifically for plant research. In 2000 he developed his virtual RLGS system. “In those days, a project to decipher the whole genome of Arabidopsis, an experimental model plant, was in progress. I thought it must be possible to simulate the electrophoretic pattern if the entire base sequence that constitutes the genome were known. Comparing the virtual pattern and the real pattern, the site where the mutation has occurred can easily be identified, and the base sequence can be immediately determined.”
The virtual RLGS system was completed with the aid of Chief Researcher Toshikazu Ebisuzaki in the Computational Astrophysics Laboratory at RIKEN. Currently, it is used not only for plant mutation analysis, but also for studying cancer cells and embryonic stem cells in mice.
Grievances from producers
Matsuyama took note of the plant breeding at RIKEN, which employs heavy ion beams as a means of inducing mutagenesis. Heavy ions (heavy atoms deprived of electrons) are accelerated to near the speed of light using a particle accelerator. When the shoots, seeds and other parts of plants are irradiated by the resulting ions, their DNA is damaged, resulting in a high probability of mutations. Using this approach, RIKEN researchers have created a number of new varieties, including Nishina Zao, a cherry tree with light-yellow flowers, and Olivia Pure White, a variety of dianthus characterized by snow-white blooms.
“When irradiated using a heavy ion beam, a wide variety of mutations occur, but it is not immediately possible to know where the mutations have occurred. Hence, I thought that the mutations might be located using RLGS. However, my attempt was unsuccessful. This is because DNA methylation—a type of chemical modification involved in gene expression regulation—occurs frequently in plants, which hampers the detection of mutations. Around that time I was asked by the Kagoshima Biotechnology Institute to provide consultation on difficulties they were having,” says Matsuyama in retrospect. That was in 2003.
In Kagoshima, a fall-blossoming chrysanthemum variety called Jinba is cultivated. Producing large pure-white flowers that last for a long time, this variety is in high demand for use at funerals. However, Jinba is not easy to cultivate. To prevent the formation of a large number of lateral branches, debudding is essential. This work cannot be mechanized and hence has to be performed manually. It is hot and humid in the greenhouses where Jinba are grown, and the market requires Jinba cut flowers to be about one meter in length. The work of the producers is therefore hard labor, involving repeatedly standing and crouching, hundreds of times a day.
Kagoshima Prefecture breeders then created a new variety, Arajin, characterized by the localized distribution of lateral branches only on the upper portion of the stem (Fig. 3), by irradiating the conventional Jinba variety with heavy ions using one of the Japan Atomic Energy Agency’s ion accelerators. By reducing the number of lateral branches, it freed producers from the need to repeatedly crouch and stand up and hence improved productivity. However, another problem arose.
Released onto the market after debudding, both Jinba and Arajin are free from lateral branches, so they are indistinguishable in appearance. Jinba may be cultivated freely because it is not registered as a new variety. Arajin, on the other hand, is registered as a new variety and must not be cultivated without permission because the rights of the breeder are protected by law. However, there are many cases where registered varieties are cultivated illegally and marketed, posing a major problem. If Arajin is illegally exported to foreign countries, grown on large scales at lower cost, and imported back to Japan at low prices, the breeder of the new variety will suffer considerable economic damage. The officials of Kagoshima Prefecture said repeatedly to Matsuyama that this is what they are most afraid of: “They said that they had tried many methods of distinguishing between Jinba and Arajin but always failed. They came to me with expectations for the potential of the RLGS system.”
As Matsuyama was listening to their worries, he was suddenly struck with an idea—a method that he had tried once but which had failed about 20 years earlier when he was a university student. “I told them about my unsuccessful attempt to locate mutations using RLGS, and suggested examining repeated sequences in the DNA, rather than mutated genes. This is how I discovered a mark that can be used for identification.”
Making DNA marks
Matsuyama focused on retrotransposon, a type of sequence that occurs repeatedly in the noncoding region of DNA. When a retrotransposon is damaged by a heavy ion beam, the number of repeats changes. Additionally, retrotransposon is present in the DNA of all organisms in the history of evolution, and the base sequences are remarkably similar. Although the base sequences of the chrysanthemum genome have not been deciphered at all, Matsuyama thought that closely related chrysanthemum varieties could be distinguished by means of the available genetic information on the known base sequences of corn and other plants. “We amplified the retrotransposon of both Jinba and Arajin, and examined their electrophoretic patterns. If a mutation has occurred, an electrophoretic band appears or disappears,” says Matsuyama (Fig. 3). “Here is a band in Jinba that it not present in Arajin. This can be used as a mark to distinguish between Jinba and Arajin.” Matsuyama named it a ‘DNA mark’.
This achievement was soon picked up by newspapers and other media. “I wanted to let the public know promptly and widely that even closely related varieties are distinguishable, since it would help prevent illegal cultivation of new varieties. I could not say that, however, because although there are people who are having problems with this issue, we could not disclose the fact until I had written a paper,” says Matsuyama. “I was happy that I was able to make use of my unsuccessful attempt of 20 years ago.”
In 2007 Matsuyama established his Plant Breeding and Cell Engineering Research Unit. Since then, many issues related to food safety have arisen, one after another, including the deceptive labeling of variety names or production centers. Hence, he advocated DNA marking for varieties of other plants as a theme for research at his unit, making use of his experience with Arajin. The Development of Technology for Identifying Varieties and Production Centers of Propagated Crops by DNA Marking project (2007–2009) proposed by the research unit was selected as part of the Project for Developing Practical Technologies to Promote New Policies in Agriculture, Forestry and Fisheries by Japan’s Ministry of Agriculture, Forestry and Fisheries. Currently, DNA marks are being developed for two chrysanthemum varieties bred by the Kagoshima Prefectural government, three cymbidium varieties from Mukoyama Orchids Co., Ltd, and the pear variety Akizuki from the National Institute of Fruit Tree Science.
Having formulated the Intellectual Property Strategic Program, the Japanese Government is implementing actions to enhance the protection of breeders of new varieties. “It is also necessary for the breeders to take their own protective measures,” Matsuyama points out. “I have proposed DNA marking to confer marks intentionally.” First, the variety of interest that is to be given DNA marks is irradiated with heavy ion beams or the like (Fig. 4). If the coding region of its DNA is damaged, changes in plant characteristics can occur. However, damage to the noncoding region is largely unrelated to character changes. Then, individuals with no character changes are selected. They have the same characteristics as the original, but their DNA is damaged at different sites. Using the same method as with Arajin, their characteristic DNA marks can be identified. “Although they are the same variety, they have different DNA marks. By distributing strains with different DNA marks corresponding to respective localities and producers, it will be possible to identify individual plants by production center and producer.”
He says that a knack for developing a DNA mark is required. “Now we use electrophoresis, but in the near future, it will possibly be easier, quicker and less expensive to read and compare target base sequences. In DNA marking, it is difficult to select individuals exhibiting no character changes after beam irradiation. Additionally, reading the base sequences generates enormous volumes of data. The DNA marking technique could not be brought into practical application without the aid of experts in field sciences and informatics. I hope that a research coalition system will be organized.”
To create new varieties
Matsuyama says quietly, “Something I have long wanted to do is to create new varieties.” His desire is reflected in his research unit’s name: Plant Breeding and Cell Engineering.
Matsuyama is targeting a specific gene. It is of a mutant of Arabidopsis discovered at the Plant Function Laboratory, to which he used to belong. Usually, plants cannot produce chlorophyll, and thus turn green, unless exposed to light. However, this mutant develops green cotyledons even when grown in the dark. Matsuyama says, “First, I will make an analysis at the gene level to determine why the mutant turns green without exposure to light. Then I will proceed to create new varieties using the gene.” We asked a question: “Will you create a vegetable that can grow in the dark?” He answered with a laugh, “I cannot give you an answer now; but the outcome will be evident in the future.” It is hoped that a new plant will be created at the Plant Breeding and Cell Engineering Research Unit.
About the researcher
Tomoki Matsuyama was born in Tokyo, Japan, in 1967. He graduated from the School of Agriculture, Meiji University, in 1991, and obtained his PhD in 1996 from the same university where he studied plant cytogenetics as a Research Fellow under the Japanese Junior Scientists scheme of the Japan Society for the Promotion of Science. After three years’ postdoctoral training as contract researcher and Special Postdoctoral Researcher at RIKEN, he became a research scientist in 1999. He worked on mutation research, including epigenetic alterations and genetic mutations in plants with physical mutagens and the development of a novel genome analyzing system—virtual restriction landmark genomic scanning (Vi-RLGS). At present, he leads the DNA Marking Project for the identification of closed cultivars promoted by the Japanese Ministry of Agriculture, Forestry and Fisheries. Having established the Plant Breeding and Cell Engineering Research Unit in the RIKEN ASI as Unit Leader, he continues to address cytogenetics for plant breeding, and is involved in analysis of greening mutants of Arabidopsis.