Introduction
ost people in the world depend on rice or wheat as their staple food. However, rice and wheat alone cannot sustain health. There are more than one hundred different kinds of vegetables in the world, and they have become a necessary part of our daily diet.
There are two main reasons for this. Firstly, they supply us with many kinds of nutrients needed for our health, e.g., vitamins, trace elements, amino acids, fiber and oil. Secondly, they appeal to the five senses. The taste of vegetables may be sharp, hot, pungent, bitter, sweet or sour. Pickled vegetables, processed with the assistance of microbes, have a different taste than the natural one. Moreover, the flavor, color and crispness of vegetables is enjoyed by our nose, eyes, ears, teeth and tongue. The numerous food cultures of the world use the various characteristics of vegetables in a wide variety of ways.
In vegetable breeding, there are various breeding objectives for consumers, such as eating quality, color and nutrient content. For producers, there are breeding objectives such as high yields, early maturity and disease resistance. However, there is one more objective in vegetable breeding to consider.
Even if a variety has a good flavor and high yields and is disease resistant, unless the seeds are stable and can be relied on to produce the same characteristics every year, it cannot become a good variety. Seed is also an agricultural product, and a good variety must have the characteristic of good seed production.
In recent years, most of the vegetables cultivated over a large part of the world have been F1 hybrid varieties, that realize high yields and early maturity through heterosis. Heterosis breeding is only possible if there is an efficient method of producing F1 seed on a large scale. For this, some control of reproduction or pollination is needed.
In controlling pollination, self-incom-patibility (SI) has been used for the Cruciferae. These include many important kinds of vegetables, such as cabbage, radish, Chinese cabbage, turnip and broccoli. The study of SI in crucifer crops began in Japan, where it still continues. In 1949, a Chinese cabbage F1 hybrid variety, "Nagaoka Kohai I Go", was produced by Shojiro Ito, and in 1961 a radish F1 hybrid variety, "Harumaki Minowase", was produced by a commercial seed company. Most cruciferous vegetables grown in Japan and in other countries are now F1 hybrid varieties, whose seeds are produced using SI.
However, SI is not always stable. It can be easily overcome under various external and physiological conditions. Therefore, stable F1 seed production has been the subject of study for many years.
SI is governed by a series of multiple alleles (hereafter referred to as the S-gene) (Bateman 1955). However, there are genetic variations in the level of SI (hereafter referred to as LSI). Therefore, it can be assumed that SI is also regulated by genes other than the S-gene.
There are two major seed production methods using the SI system for crucifer crops. One method is a single cross, in which SI is overcome in parental seeds by treating them with CO2 gas. The other method is a double cross, in which parental seeds are produced using a pair of isogenic lines for the S-gene.
There are genetic variations in the reaction level of self-incompatibility (RLSI) to a 4% CO2 gas treatment. Thus, the parental lines in the parental seed production in a single cross have to show a marked reaction to CO2. In F1 seed production on both a single and a double cross, the parental lines have to show a high LSI. It is therefore important to know the genetic relationship between these characteristics.
The aim of our study was to divide SI into these three characteristics, perform a genetic analysis for selection of the reproductive traits, and apply the DNA marker assisted selection for the S-gene to radishes.
Genetic Diversity of the S-Gene
S-Gene Cloning and Sequencing
PCR amplifications were performed using a primer set derived from Brassica oleracea SLG genes, and the total DNA extracted from the radishes as a template. Some of the radish SLG genes were cloned. S201 shared an 88% homology at the DNA level and an 80% homology at the amino acid level with SLG6 of B. oleracea (Niikura and Matsuura 1997). It also has twelve cysteine residues and the conserved and variable domains common in the SLG genes. Furthermore, northern blot analysis using S201 as a probe showed that the major transcript was detected in the leaf and anther in the various developmental stages. The transcript in the stigma appeared one day before anthesis, at the same time as the onset of the SI response. A maximum level of expression was observed at the day of flower opening (Niikura and Matsuura 1997).
PCR-RFLP
Identification of the S-alleles by PCR-RFLP was performed, using the primers which originated in the radish SLG genes. The accordance between the results of the PCR-RFLP and test crossing was confirmed, using a pair of isogenic lines for the S-gene and an F2 population which was segregated for the S-alleles. Thus, the PCR amplifications were performed in 37 S-alleles (S201 to S237) and a single fragment with the expected size of around 900bp was amplified in 29 S-alleles. No PCR products were found in the other 8 S-alleles. When the PCR products from the former 29 S-alleles were digested with MspI, RFLP patterns unique to 21 of these S-alleles were observed (Niikura and Matsuura 1998, Fig. 1; Niikura and Matsuura 2001).
Genetic Variation
Genetic variation in the reaction level of self-incompatibility to a 4% CO2 gas and the level of self-incompatibility (LSI) was studied in the landraces.
It was evaluated on the basis of the rate of self-seed setting and the degree of the pollen-tube penetrations treated by a CO2 gas incubator [grade 1 = low to grade 5 = high]. The LSI in these lines ranged from 0 to 100%. The self-incompatibility to CO2 gas ranged from grade 1 (11 lines; 32% of all the tested lines) to grade 5 (4 lines; 12% of all the tested lines). No correlation between the S-allele, LSI and self-incompatibility to CO2 gas was detected (Table 2 and Table 3).
Genetic Analysis
Genetic analysis of the level of self-incompatibility (LSI) and the reaction level of self-incompatibility to a 4% CO2 gas was carried. For the latter, an F2 population and F3 lines were analyzed in which the S-alleles and self-incompatibility to CO2 were expected to segregate. The results showed that a high reaction level of self-incompatibility to CO2 is controlled by a recessive gene (Fig. 2, Niikura and Matsuura 2000).
To examine which organs respond to CO2 gas, the same S-homozygotes with different levels of self-incompatibility to CO2 were used from the above F2 population. Reciprocal crosses were performed among these plants. Only when using female plants which showed a high reaction level of self-incompatibility to CO2 was this trait inherited by the cross combinations (Niikura and Matsuura 2000).
A simple differential display method (Yoshida et al. 1994) was performed to isolate the genes related to LSI. When an arbitrary primer was used, a specific PCR product (PA-1) common to lines with a low LSI was amplified (Fig. 3). The sequence of PCR productions had a high homology with S-adenosyl methionine synthase. Furthermore, northern blot analysis using PA-1 as a probe showed that the faint transcript was also detected in the high LSI lines. This result shows that a regulatory gene of PA-1 transcript is a trigger of the LSI. It is necessary to isolate this type of gene in future.
Conclusions
The biotechnology used in this study is based on the use of the PCR-RFLP method. It is very useful, because it allows breeders to know the S-genotype before harvest. In the past, this could not be known until pollination tests were performed at the reproductive stage. The use of this kind of marker assisted selection (MAS) will greatly increase selection efficiency for reproductive characteristics which appear only after bolting and flowering.
Another important result from this study is to divide the SI into the three characteristics, the "S-gene", "Reaction level of self-incompatibility to CO2", and "LSI", and to carry out a genetic evaluation of these characteristics independently. The success or failure of MAS depends on whether an exact evaluation of the target characteristics, and genetic analysis based on that, is possible. It will be necessary in future to perform a genetic analysis of the LSI, and to isolate the gene governing the reaction level of self-incompatibility to CO2 and LSI, or screen a linkage DNA marker.
As far as the practical implications of SI are concerned, there are other areas of SI that have not been elucidated. The process of seed production also includes such steps as bolting, flowering, pollination, fertilization and seed setting. Fertilization is the most important step, because this step connects directly with the F1 purity of the seeds produced. Our study focuses on this step in order to realize high-quality seed production.
Concerning the S-gene, a number of genetic and molecular biological studies have been carried out (Nettancourt 2001). The success of the PCR-RFLP method depends largely on these studies.
However, in order to realize stable seed production on a large scale, there is one more topic we need to study. We must examine the unknown parts of SI by the reduction method, as we have done in this study. We also need to study SI in its ecological context, and its biological role in pollination.
This is because in F1 seed production, the exchange of pollen between parental lines has to be done by pollinating insects.
Firstly, it is important to synchronize the time of flowering of parental lines, by changing the sowing time and transplanting time, and trimming off the flowers which open first. This is included in the "bolting and flowering" step. Secondly, it is necessary to clarify how to prevent insects from preferential pollination of one or other of the parental lines.
This has to be done by following an ecological approach to the "pollination" step. That is, the optimal density of the parental lines should be decided, taking into account the length of their flowering time, their plant shape, the quantity of their flowers and their population structure.
Thirdly, the form of floral organs that are easily pollinated should be considered. This is also included in the "pollination" step. Namai has published many reports on "pollination biology" (e.g. Namai and Ohsawa 1988). By applying these studies to F1 seed production, it will be clear whether there are congenial floral organ forms in the parental lines of crops as diverse as primrose and buckwheat, or even whether there is an ideal floral organ form irrespective of which parental line is the partner. Besides these studies, further research into "seed setting", and into interactions between these different steps, must be carried out.
A few decades ago there were classes on seed production in most Japanese agricultural universities. These contributed a great deal to improved technology for seed production. Unfortunately, such classes have now disappeared because students prefer biotechnology, with its smart "high-tech" image. Even if they wanted to attend traditional seed production classes, there are no longer any teachers to teach them.
However, since cruciferous vegetables continue to be important in the world's food supply, not only should good varieties be raised, but also seed production techniques should be handed down. Biotechnology in agricultural research has no meaning until it is used to offer or improved genetic material to farmers. Scientists working on the breeding of cruciferous vegetables must not forget this.
Satoshi Niikura
agnet.org
9.17.2008
Self-Incompatibility in Vegetable Seed Breeding
