Recombinant DNA Molecule Research Potential

For Agricultural Crop Plants

David N. Duvick presentation at National Research Council 11apr77

Recombinant DNA molecular research probably will have little direct impact on the development of useful new crop varieties. Complex, delicately balanced interactions among many genes determine the phenotypes of successful crop varieties. Changes of only one or a few genes have had important useful effects in relatively few instances; the total benefit of single gene changes in crop varieties is negligible compared to the importance of complexly inherited traits. But the in vitro recombinant DNA molecular techniques deal basically with one or at most a few genes. They are concerned with isolating and moving individual short segments of DNA; they are not suited to assorting and recombining very large numbers of genes into optimum genomic combinations. (This latter process has been accomplished within species and varieties in evolutionary time; modern plant breeding techniques are concerned largely with developing optimum recombinations of internally balanced chromosome segments - large blocks of genes. In general, breeding processes which reshuffle the genes result in useless genomes as far as productive crop varieties are concerned.) Thus, it is not likely that recombinant DNA molecular techniques can be used to rebuild genomes in wholesale fashion.

An important accomplishment of recombinant DNA molecular research is movement of operational DNA fragments from one organism into another entirely alien organism. Whether or not this can be done reliably in higher plants awaits demonstration, but it may be possible. A hypothetical example would be to transfer the gene(s) allowing for fixation of atmospheric nitrogen from bacteria to wheat. Another example might be transfer from sorghum to corn, of a gene(s) for immunity to corn root worm attack.

But even if these gene transfers could be made, and stable inheritance of the transferred genes could be expected, such hypothetical transfers would have very low likelihood of being useful to plant breeders. In most cases so-called "single-gene" effects actually depend on a very complex background of supporting genes to produce their phenotype, even though the null form of the gene may cause complete disappearance of the phenotype. Moving given genes from variety to variety often changes the intensity or even the nature of their effect; sometimes the gene produces unexpected deleterious effects. When genes are moved from one species or genus to another these unexpected and often deleterious effects can be even greater. The only genes that seem to move easily with minimum unexpected effects are those acting on terminal reactions such as pigment formation, or deposition of other chemical end-products, and even they are affected by background genotype. Therefore, using recombinant DNA molecular techniques to move genes across wide taxonomic distances probably will give few direct practical benefits to plant breeding.

To list the reasons why in vitro DNA recombinant research may not help produce new varieties does not, however, mean that it should not be tried. Movement of nitrogen fixation genes from bacteria to wheat might give us valuable new insights into the nature of the fixation process, which might then be used to improve nitrogen-fixing bacteria or their symbiosis with higher plants, or other completely unexpected pieces of useful knowledge about crop plants and their symbiosis might emerge from the work. Movement of genes producing products toxic to insect pests occasionally might be successful, because such products often are the ends of biochemical pathways and might not upset the functioning of adapted, productive varieties.

However, it is likely that the real use of in vitro recombinant DNA research in plant breeding will be to allow determination of nucleotide sequences and thus the gene products of key genes. It also should allow for more complete gene mapping (especially in organelles) and eventually it also might give new clues to gene action in development. Thus for plant breeders the technique is most useful as a tool to study genes, rather than as a tool to create new organisms.

Of course improved understanding of individual genes probably will suggest to plant breeders ways in which they might improve those genes' action, structure or numbers, in specific instances. Modification, substitution or multiplication of genes for high amylose production in corn for, boll weevil resistance in cotton or for rust resistance in wheat could give highly useful and valuable improvements. But as noted earlier, it is likely that the total effect of all such changes would be minor in comparison to changes in total productivity that can be achieved by selecting in vivo genetic recombinations of delicately balanced sets of genomes or chromosome segments. Meiosis and the very large scale on which its products can be selected, is still the most powerful tool available to plant breeders. It may be no accident that it also was selected as the chief tool of natural evolution in higher plants.

Nevertheless, business firms engaged in the plant genetic supply industry will be interested in developing techniques and products with molecular DNA research, even if not, in general, for immediate use (for the reasons noted above). Because our survival depends on maintaining physical possession or legal protection of the products we develop, we will need to have either the right to hold these products within our organization or the ability to protect them by means of something like variety protection laws. Likewise, if we develop novel techniques we need to be able to either hold them within our organization or protect them with patents. In both cases, if we cannot use one of the two alternatives, there likely will be little financial incentive for doing the work.

The suggested guidelines for recombinant DNA research in connection with plants seem to be reasonable, but it is important that they remain open to modification, as new knowledge is gained.

I would guess that pathogens require a delicate balance of many interacting genes for their survival, just like higher plants, and that the danger of creating super-pathogens by means of long-distance gene transfer is not at all great, in fact it probably is impossible. At any rate, pathogens of all sorts do very well now, in creating novel and highly efficient races, whenever a new evolutionary niche appears. It may be that we are presumptuous to suppose we can move a few genes at random into strange genetic backgrounds and surpass pathogen's present capabilities.

Dr. Don Duvick, director of plant breeding research in Pioneer Hi-Bred International, Inc.