A long prevailing dogma has been that plants only use inorganic nitrogen sources such as ammonium and nitrate for their nutrition. As early as the middle of the previous century, plant scientists showed, however, uptake of organic nitrogen sources by plants. Despite this, only in the last 10 years or so have organic nitrogen and preferably amino acids been recognized as important nitrogen forms for plant nutrition. Several studies have shown that amino acids are of importance for plant nutrition in natural habitats, especially in alpine, arctic, and boreal regions where mineralization of organic matter and production of inorganic N is slow. Other studies suggest the capacity for root absorption of different amino acids is ubiquitous to plants. Furthermore, studies of different amino acid transporters have indicated that several of these have a very broad substrate range, including also D-enantiomers of amino acids. We have discovered, while conducting research in the fields just described, important differences between optical isomers of amino acids. The nutritional value of individual amino acids was tested on sterile-grown Arabidopsis thaliana. In these tests we could show that several L-amino acids promoted growth while none of the D-enantiomers did so. We also found that some D-amino acids were very toxic to plants (e.g., D-alanine and D-serine), while others had slight negative effects, and still others had no effect at all (e.g., valine and isoleucine).

D-amino acids and their metabolism
All protein amino acids except glycine are chiral molecules and have at least two optical isomers. Optical isomers are nonsuperimposable mirror images having identical physical and chemical properties, such as melting and boiling points, and chemical reactivity against molecules that are not optical isomer themselves.

The metabolism of amino acids is described in detail for a large number of different organisms. Studies of amino acid metabolism are in many cases narrowed down to only L-amino acids, although many organisms metabolize the less abundant optical isomer, D-amino acids. One of the most studied metabolic routes involving D-amino acids is in bacteria, where the peptidoglycan layer of the bacterial cell wall contains amino acids of both optical isomers. Peptidoglycan consists of peptides with both L/D and D/D amino acids. Furthermore, D-amino acid metabolism is also described in many other organisms, such as humans, mammals, birds, insects, fish, algae, and yeast, but very little is known about D-amino acid metabolism in plants.

One of the best known enzymatic pathways for metabolism of D-amino acid is via oxidative deamination by D-amino acid oxidase (encoded by the DAO1 gene). D-amino acid oxidase is one of the most studied enzymes, discovered more than 60 years ago, and is a model flavoprotein for which the structure and catalytic mechanism have been described in great detail. Enzymes such as the D-amino acid oxidase, which is almost ubiquitous to organisms of many kingdoms, is missing in all plants investigated hitherto.

During the 1970s, a fair amount of research was carried out within the field of D-amino acid metabolism in plants. The majority of that research used pea seedlings or other leguminous plants as study objects. Radiotracer studies with D-amino acids showed that plant metabolism of these compounds preferably leads to the formation of N-malonyl and N-acetyl derivatives. Such conjugation is believed to be a way to inactivate potentially toxic compounds. The fact that plants cannot use the optical isomer of protein amino acids as nitrogen sources, and that they are conjugated into what is thought to be dead end products, suggests that plants treat D-amino acids as intrusive compounds. Thus, plants metabolize D-amino acids differently than other organisms.

D-amino acids in biotechnology
The potential for using D-amino acids and their metabolism in biotechnology is suggested from what is described (above). Because plants lack the metabolic pathway common to many other organisms, they are sensitive to several D-amino acids and cannot use such compounds as N sources. We hypothesized that this sensitivity could be alleviated through expression of a D-amino acid metabolizing enzyme in a plant. Furthermore, we hypothesized that introducing this metabolism into a plant would also enable it to use D-amino acids as N sources. A candidate for such an experiment was the dao1 gene. Hence, our first test was to see if the toxicity exerted by D-serine and D-alanine could be relaxed through expression of dao1 and if D-amino acids promoted growth of these plants. These experiments gave very promising results and suggested to us that dao1 could function as a selectable marker.

One of the unique features of the D-amino acid oxidase selectable marker system is, however, that it enables both positive and negative selection, simply by choosing different selective media. This feature of dao1 was discovered when we screened for effects of a range of D-amino acids on wild type and on dao1-expressing plants and found that some D-forms that were indifferent for wild type plants had a strong negative effect on the transgenic plants. Selecting for the transgenic plants/tissues/cells occurs through dao1 alleviation of the toxicity exerted by D-alanine and D-serine, for example. D-alanine is decomposed by D-amino acid oxidase into ammonium, hydrogen peroxide, and pyruvate while D-serine forms hydroxyl-pyruvate instead of pyruvate when metabolized by the enzyme. Selecting against dao1-carrying plants, tissues, or cells is possible through conversion of non-toxic D-amino acids (e.g., D-isoleucine, D-valine) into toxic products (3-methyl-2-oxo pentanoic acid and 3-methyl-2-oxo butanoic acid, respectively) by the action of D-amino acid oxidase. In Figure 1, the effect of D-alanine and D-isoleucine on dao1-expressing and wild type plants is shown.

Figure 1. Arabidopsis thaliana wild type plants (left half of the plate) and plants expressing dao1 (right half of the plate) sown on media containing D-isoleucine (upper half of the plate) and D-alanine (lower half of the plate). Wild type plants are killed by D-alanine but not by D-isoleucine. Transgenic plants are not harmed by D-alanine but suffer when exposed to D-isoleucine.

The use of selectable markers is currently debated. There are arguments raised against the use of antibiotic resistance markers due to the risk that it might compromise the therapeutic value of antibiotics. Whether this risk is real or not is still an open question; in any case, it is still a matter of public concern and it is likely that these markers will be phased out, at least in commercial crops, in the near future. Obviously, the best way to treat the problem of selectable markers is to delete it. One advantage with the D-amino acid oxidase marker is that it can be used for both positive and negative selection. Thus, in combination with existing techniques for marker removal, dao1 may find its primary function in the production of marker free transgenic plants. Preliminary studies on a range of crop plants show that all species tested react similarly as Arabidopsis upon exposure to D-amino acids. This suggests that dao1 may work as a selectable marker in commercially important crops as well as in the model plant, and future research will show if, and to what extent, dao1 may replace selectable markers used today.

Reference
Erikson O, Hertzberg M, & Näsholm T. (2004) A conditional marker gene allowing both positive and negative selection in plants. Nature Biotechnology, 22: 455-458.

Oskar Erikson & Torgny Näsholm
Umeå Plant Science Centre
Dept. of Forest Genetics and Plant Physiology
Umeå, Sweden
torgny.nasholm@genfys.slu.se