Production of active oxygen species is not always inadvertent. For example, oxygen photoreduction (known as the Mehler peroxidase reaction) results from the transfer of electrons from photosystem I (PSI) to oxygen to form superoxides and eventually hydrogen peroxide (H₂O₂). The Mehler reaction thus serves to maintain electron flow through PSI and sustain its function. AOS is also produced in other cellular compartments such as the mitochondria, which have substantial electron transport activity, and in peroxisomes during the oxidation of glycollate.

To limit the deleterious effects of many active oxygen species resulting from physiological processes or from exposure to environmental pollutants, plants (like other organisms) use antioxidants. Ascorbic acid (vitamin C) is the most abundant antioxidant in plants, present in millimolar concentrations that range from 10 to 300 mM¹ and is used to detoxify H₂O₂, superoxide, hydroxyl radicals, and singlet oxygen. Glutathione, the other major soluble antioxidant, is typically present at only 10% of the concentration of ascorbate. The concentration of ascorbate often correlates with those environmental conditions that cause photo-oxidative stress. For instance, alpine plants that grow at high altitudes and thus are exposed to a combination of high light, cold temperatures, and elevated levels of UV-B contain high levels of ascorbate. The high concentration of ascorbate may be required to eliminate the active oxygen species that are produced in high amounts under these growth conditions.

Vitamin C is essential to animals to prevent diseases affecting the connective tissue such as scurvy. It also improves cardiovascular and immune cell function and is used to regenerate vitamin E. In contrast to most animals, humans cannot make vitamin C and therefore must obtain the vitamin regularly from dietary sources. Vitamin C is present at high levels in some fruits such as citrus and kiwi as well as in several green leafy vegetables but is present in low amounts, however, in those crops most important to humans such as grains.

Ascorbic acid (ASA) biosynthesis in plants differs from that in mammals and results from the oxidation of L-galactose to L-galactono-1,4-lactone, which in turn is oxidized to ascorbate by L-galactono-1,4-lactone dehydrogenase². Once used, ascorbate is oxidized to dehydroascorbate (DHA) and is rapidly regenerated by the enzyme dehydroascorbate reductase (DHAR), which requires glutathione as the reductant (Fig. 1).

If ascorbate is not salvaged by DHAR, it is quickly lost. Glutathione reductase uses NADPH produced principally from photosynthesis to regenerate glutathione from oxidized glutathione. Consequently, the detoxification of active oxygen species by this ascorbate-glutathione pathway involves the transfer of electrons from PSI to NADPH to glutathione to ascorbate to H₂O₂ in a series of reactions in which there is no net loss of ascorbate or glutathione (Fig. 1). The principal function of DHAR activity, therefore, is to maintain the existing pool of ascorbate in a reduced state and enable an organism to recycle ascorbate.

Because DHAR regenerates ascorbate from DHA and thus rescues it before DHA is further catabolized, an increase in DHAR activity might be expected to improve a plant's ability to recycle vitamin C and thus increase its cellular concentration of vitamin C. This would be true particularly if DHAR were limiting. In order to examine whether increasing the amount of DHAR in plants could increase their vitamin C content, we purified DHAR protein to near homogeneity from germinating wheat seedlings by assaying for the reduction of DHA to ascorbate in a glutathione-dependent assay. Antiserum was raised against the purified DHAR protein and was used to screen a wheat seedling cDNA expression library. We were successful in isolating a full-length DHAR cDNA encoding a protein that is 89.5% conserved with rice DHAR. We then introduced the wheat cDNA into corn where expression from the transgene increased the amount of total DHAR up to 100-fold³. This increase in DHAR elevated the amount of ascorbic acid in corn leaves and grain, showing that the vitamin C content of plants can be elevated by increasing expression of the enzyme responsible for recycling the vitamin. Similar results were achieved in tobacco leaves, which were used as a model for a leafy, non-grain crop, where leaves contained up to 4.6-fold more ascorbate relative to the control transgenic plants. Because ascorbate is the most abundant antioxidant in plants, it contributes significantly to the redox state of a cell. Consistent with its function, overexpression of DHAR resulted in a higher ratio of ascorbate (81%) relative to control plants (60%). Glutathione is used by DHAR to reduce DHA to ascorbate. Interestingly, the level of glutathione increased by up to 3.5-fold in leaves from DHAR-overexpressing plants, suggesting that increasing DHAR expression increases the level of ascorbate and glutathione and maintains a more reducing cellular environment.

In contrast to animals, land plants have adopted very different means by which they respond to environmental changes. This is partly a consequence of their more direct dependence on the physical environment for their survival, e.g., their light-driven mode of generating energy as well as their sessile nature. Because land plants cannot escape environmental stress, they must evolve means to minimize the damage resulting from exposure to a stress. Several abiotic stresses result in the generation of active oxygen species, including exposure to ozone, ultraviolet light, drought, salt, and cold. The ability to maintain ascorbate in a reduced state may likely be as important in mounting a rapid response to an acute demand on the antioxidant as is ascorbate synthesis. Because ascorbate is involved in reducing damage imposed by oxidative stress, increasing DHAR expression may well provide broad protection to several types of abiotic stresses, including drought and exposure to ozone. Determining to what extent DHAR expression and function is important in the response program to those stresses that generate active oxygen species will be critical to understanding the role that oxidative damage plays in these stresses and how DHAR is used to mount an effective response to limit stress-related damage. Engineering greater resistance to oxidative stresses or to drought would be highly advantageous to crop species important to U.S. agriculture, such as cereals, and may significantly improve plant performance in environments degraded by ozone, soils contaminated with salt, or where instances of increases in temperature or drought can occur. With the continued spread of development into agriculturally important regions and the resulting degradation of air quality as well as the continued salt contamination of farmland soils that results from irrigation practices, manipulating DHAR expression and thus the level and redox state of ascorbate may be an important means to improve crop performance under adverse growth conditions. Thus, increasing the vitamin C content in plant species will not only prove to be directly beneficial in improving the human diet but may provide an indirect effect by improving crop productivity in response to those environmental stresses that involve oxidative abiotic stress.

We have therefore developed technology that increases the amount of vitamin C in plants, including grains, by increasing the amount of the enzyme that is responsible for recycling vitamin C. The ability to increase the level of this important vitamin in plant foods will enhance their nutritive value by increasing the number of foods from which the vitamin can be obtained as well as increasing the level of the vitamin in those foods that are already good sources of vitamin C. The current recommended dietary allowance (RDA) of vitamin C for adults is 75 and 90 mg for adult females and males, respectively, which is sufficient to prevent diseases arising from severe vitamin C deficiency such as scurvy. Although the RDA can be obtained through a balanced diet, more than 30% of Americans fail to consume 60 mg a day. In addition, some studies have suggested that up to 200 mg may be necessary to ensure good cardiovascular health and immune cell function, increasing the likelihood that more people would fail to achieve the RDA with their present dietary habits. Our ability to increase the vitamin C content in plant foods such as green leafy crops as well as grains should make it easier for people to obtain enough of the vitamin for their optimal health.

References
1. Smirnoff, N. (2000) Ascorbate biosynthesis and function in photoprotection. Phil Trans Royal Soc London Series BBio Sci 355: 14551464.
2. Wheeler GL, Jones MA, and Smirnoff N. (1998) The biosynthetic pathway of vitamin C in higher plants. Nature 393: 365-369.
3. Chen Z, Young TE, Ling J, Chang SC, Gallie DR. (2003) Increasing vitamin C content of plants through enhanced ascorbate recycling. Proc Natl Acad Sci USA 100: 3525-3530.

CHLOROPLAST EVOLUTION, GENETIC MANIPULATION AND BIOSAFETY
Jeremy N. Timmis

The higher plant chloroplast (cp) has become an attractive arena for genetic manipulation because diverse proteins, or sets of proteins, may be produced by growing transplastomic plants on a field scale¹. Chloroplast genes are often highly expressed to produce large amounts of protein, and they are partially contained through stringent maternal inheritance that operates in many of the most important crop plants. However, our recent experiments² indicate that chloroplast genes occasionally transfer to the nucleus and therefore may escape in pollen. This transfer process has led to both popular and scientific discussion regarding the evolution of eukaryotic cells and the impact of the gene transfer process on the level of biological containment that has been claimed by proponents of chloroplast biotechnology³. Both sorts of discussion contain a mixture of fact and fiction.

The extant genomes of cytoplasmic organelles are far too small to encode the full spectra of different proteins that are required for the functions of chloroplasts and mitochondria. Since the advent of early eukaryotes, organellar genomes have evolved, throughout hundreds of thousands of years of endosymbiosis, from those of ancestral free-living prokaryotes. During this long period many ancestral organellar genes have been transferred to the nucleus and been deleted from mitochondrial and chloroplast DNA.

The circular chloroplast genome of higher plants—the plastome—is approximately 150 kb in size and it encodes about 120 products. The remaining several thousand genes required to make functional chloroplasts are located in the nucleus. Although the plastome is small, each of several hundred chloroplasts within a mature leaf cell, in most species, contains tens of copies of an identical plastome. Therefore a chloroplast gene will usually be present thousands of times per cell with the result that chloroplast genes produce the most abundant proteins on the planet. The large number of chloroplast genomes per cell also means that cpDNA, despite its genetic simplicity, may contribute a very large proportion (up to 40%) of total cellular leaf DNA. A further interesting feature of organelle genes is that they are uniparentally inherited, with offspring usually inheriting only the maternal chloroplast genotype. This contrasts with Mendelian inheritance that characterises nuclear genes. The chloroplast genotype must therefore be considered relatively simple and homogeneous but present, in most species, in thousands of copies per cell.

A wide variety of studies has shown that most if not all of the nuclear genes that now encode chloroplast proteins were transferred from an ancestral prokaryote during endosymbiosis and then deleted from the plastome. This sounds like a long shot because a prokaryotic gene would not readily be expressed even if it were transferred to a nuclear environment. Neither would it be able to give rise to a product equipped to enter the organelle and usurp the function of, and eventually oust, its progenitor chloroplast gene. However unlikely this appears, we know that such events have happened repeatedly during endosymbiotic evolution⁴. The relocation of both chloroplast and mitochondrial genes can be tracked in multiple independent events in separate branches of phylogenetic trees. Some of the means by which such itinerant genes establish functionality in the nucleus have been characterized and the process occurs in a variety of different ways.

The ongoing nature of chloroplast-to-nucleus gene relocation is supported by the presence of tracts of present-day cpDNA in the nuclear genome of many plants^2,5. It has been concluded that the first step in the relocation to the nucleus of ancestral chloroplast genes involves the transfer of DNA, and that this process still continues even though the gene content of extant higher plant plastomes is relatively stable and invariable. It appears that most of the chloroplast genes that are able to transfer beneficially to the nucleus have already done so, leaving behind a minority that are apparently better placed in the chloroplast. Experimentally it is possible to move some of this latter group to the nucleus, suggesting that there are selective advantages that require the maintenance of separate genetic compartments. One obvious advantage of becoming a nuclear gene is the availability, not accessible in the chloroplast, of meiotic recombination and sexual reproduction that have been the most powerful drivers of evolution on earth.

We set out to measure the frequency with which cpDNA transfers to the nucleus, assuming that it would have to occur at an experimentally approachable rate to facilitate the evolutionary relocations described above². We designed a gene, neo, which confers resistance to kanamycin, so that it would be efficiently expressed only in the nucleus (Fig.1). Using the chloroplast transformation vector pPRV111A, we placed this gene into the two inverted repeats of the tobacco plastome. This vector contains the chloroplast-specific selectable marker gene aadA, which allows spectinomycin selection to identify cpDNA transformants. Prolonged spectinomycin selection pressure also leads to homoplasmy for the integration, meaning that regenerated plants become homogeneous for the additional cp transgenes. We aimed to render the neo gene inoperative in the chloroplast by adding an intron that we expected could only be spliced within the nucleus from a precursor mRNA transcript. Thus this gene was disabled in the chloroplast but poised to confer kanamycin resistance if it moved to the nucleus (Fig. 1).

Figure 1. Movement of DNA from the plastome to the nucleus. a) The tobacco plastome was transformed with two selectable marker genes, aadA and neo using a vector containing sequence similarity (regions C and D) to the plastome. b) The transgenes were incorporated into the plastome by biolistic transformation and homologous recombination within regions C and D of the vector and the corresponding regions of the native plastome. The aadA gene expressed () to confer spectinomycin resistance that allowed selection for homoplasmy of the transplastome. The neo gene was inefficiently expressed () in the transplastome. c) The neo gene was efficiently expressed () only after transposition to the nucleus in 1 in 16,000 pollen grains, at which time the co transposed aadA gene become inactivated ().

It became clear that even this blocked gene could be inefficiently expressed from its location in the chloroplast, so we backcrossed two independent homoplasmic plants to wild type female plants. We then scored 250,000 seedlings of this cross for DNA jumping events that had occurred at some time during the life cycle of the transplastomic male parents. To our surprise, 18 kanamycin resistant seedlings were obtained in this screen, suggesting that one in 16,000 pollen grains of the transplastomic parent contained a piece of DNA that had jumped into the nucleus.

There were many questions that needed to be answered at this stage of the experiments. Was this DNA incorporated into the nuclear genome? Was kanamycin resistance in the 18 plants inherited as would be expected if it were encoded by a nuclear gene? Did all the 18 plants result from a single rare incident or were they all independent events? It turned out that each event was indeed independent and that, in most cases, the relocated gene behaved exactly as would be expected of a diploid nuclear chromosomal gene. In some instances we were also able to confirm at the sequence level that the transposed DNA from the transplastome was linked to tobacco nuclear DNA. Most of the integrants contained the neo and aadA genes and extensive flanking native chloroplast genes that were not present in the construct (genes A, B, E and F in Fig. 1). We took pains to emphasize² that in no case did we observe the expression of the aadA gene after it was transposed to the nucleus. We concluded that, under our experimental conditions, there was watertight evidence that DNA transposes from the plastome to the nucleus at a frequency that was higher than most biologists, including ourselves, had thought possible. Of course there was absolutely no scientific basis for any such expectations!

More questions immediately arose from these results. Was this frequency we measured artificially high because of the techniques of the experiment and the material used? Was the plastome destabilised by the added transgenes? Is tobacco atypical because it is an allotetraploid? When and how does transposition occur? We cannot answer many of these questions at this time, but the results, in the context of our clearly defined and fully controlled experiments, are unequivocal.

What is the importance of these results for chloroplast biotechnology? Before addressing this question we need a little background to the chloroplast GM debate. We have known from strong evidence that has been reinforced time and again over the last 20 years that most plant nuclear genomes contain DNA that is clearly very recently derived from the plastome, necessarily implying that chloroplast transgenes are able to enter pollen nuclei and be disbursed. Plastid transgenes are not entirely contained for a second reason. Although the pollen of a transplastomic plant does not transmit the male plastome, weedy relatives (or non-transgenic crops) may act as male parents to produce hybrid seed. These hybrids would contain the trans-plastome and potentially could cause an environmental problem, though spread of the chloroplast transgenes would clearly be slower than their nuclear counterparts. Therefore there was never any doubt that chloroplast transgenes were not fully contained by uniparental inheritance, though they did enjoy a higher level of containment than nuclear genes. Our results demonstrate that, in tobacco, which is the major species targeted for chloroplast engineering, the level of potential escape is significant, particularly when large areas of cultivation are envisaged. It must be appreciated that we tracked the relocation of cpDNA by monitoring the expression of a tailor-made nuclear reporter gene. In none of the kanamycin-resistant plants screened using this selectable marker did we also observe spectinomycin resistance, despite the fact that aadA was concomitantly relocated in every instance. This is because the aadA gene is under the strict control of genetic elements that restrict expression to the prokaryote-like chloroplast compartment. It is possible that aadA could find a niche in nuclear DNA where it would be expressed, but we would expect this to be a very rare event, probably orders of magnitude less frequent than the transposition of DNA per se.

Therefore our results do not radically change what was previously known about chloroplast transgene containment. Genes designed for chloroplast expression are not likely to be functional when they move to the nucleus. This observation, together with the rarity of transfer to pollen and the phenomenon of uniparental inheritance, means that the level of biological containment offered by the chloroplast genetic compartment is high. However, scientists need continuously to evaluate the merits of transgenic technologies for crop improvement and biotechnology.

References
1. Maliga P. (2002) Engineering the plastid genome of higher plants. Curr Opin Plant Biol 5: 164-172.
2. Huang CY, Ayliffe MA, and Timmis JN. (2003) Direct measurement of the transfer rate of chloroplast DNA to the nucleus. Nature 422: 72-76.
3. Daniell H, Khan MS, and Allison L. (2002) Milestones in chloroplast genetic engineering: An environmentally friendly era in biotechnology. Trends Plant Sci 7: 84-91.
4. Millen RS et al. (2001) Many parallel losses of infA from Chloroplast DNA during angiosperm evolution with multiple independent transfers to the nucleus. Plant Cell 13: 645-658.
5. Yuan Q et al. (2002) Genome sequencing of 239-kb region of rice chromosome 10L reveals a high frequency of gene duplication and a large chloroplast DNA insertion. Mol Genet Genomics 267: 713-720.

JUMPING INTO THE NUCLEUS SILENCES PLASTID GENES
Anil Day

The potential benefits of biopharming are considerable. The use of transgenic plants to produce therapeutic proteins could provide a means to manufacture therapeutic compounds at lower cost and in greater amounts. On the other hand, many have been uneasy about the prospect of biopharm crops potentially mixing with conventional food crops. This apprehension was justified by the U.S. Department of Agriculture's November announcement that GM corn produced for ProdiGene, Inc. (College Station, TX) had been mixed with conventional soybeans. ProdiGene had reportedly engineered the corn to produce a protein used in a swine vaccine. The USDA slapped the company with a hefty fine, but some viewed this penalty as an insufficient deterrent to prevent future incidents.

By the end of the year, The Genetically Engineered Food Alert (Washington, D.C.), a coalition of seven organizations, filed a petition requesting that the USDA initiate an immediate moratorium on the planting of food crops engineered to produce pharmaceuticals or industrial chemicals. The coalition also requested the USDA to prepare environmental impact statements for all biopharm crops.

Another coalition requested a similar ban in early 2003. This time, the group represented food manufacturers and restaurants. The Grocery Manufacturers of America, Inc. (Washington, D.C.) took the position that the USDA should create a presumption against the use of food or feed crops for pharmaceutical production unless the company developing the drug clearly demonstrates that it is not feasible to use nonfood crops.

So far, the USDA has not initiated a ban on biopharming. But on March 6, the agency's Animal and Plant Health Inspection Service (APHIS) announced proposed rules for the regulation of field-testing of plants designed to produce pharmaceutical and industrial compounds. Under the new rules, the USDA will increase the frequency of GM crop site inspections, and the agency will prohibit farmers from growing conventional corn within one mile of a field test site for the duration of any field test that involves open-pollinated GM corn. APHIS will also require farmers to operate planters and harvesters dedicated to use in the permitted test sites for the duration of the tests and dedicated facilities for the storage of equipment and regulated articles.

The proposed regulations include a prohibition against the planting of a food crop the year after the land was used for biopharming. This ban might have been inspired by the ProdiGene incident. Following a standard crop rotation practice, farmers had planted conventional soybeans on land previously used to grow ProdiGene's GM corn. The result: corn seed left from the transgenic crop grew into the soybean fields.

During a pre-briefing on the proposed rules, Bobby Acord, the administrator for APHIS, said that the proposed changes are designed "to make absolutely certain that there are no ProdiGenes in the future." Acord assured reporters that APHIS has the resources to meet the challenge of increased inspections, even though APHIS transferred over 2600 of its agriculture quarantine inspectors to the Department of Homeland Security.

For a short time, the Biotechnology Industry Organization had proposed a ban on GM corn biopharming in the Corn Belt. To a certain extent, this geographic isolation approach is built into the proposed rules. Cindy Smith, APHIS's deputy administrator for Biotechnology Regulatory Services, told reporters that the agency anticipates fewer corn field trials in the Corn Belt, because it may be difficult to secure land with the required one-mile isolation distance in an area where land is best suited for corn. Furthermore, the proposed rules would take the most productive land out of use by prohibiting farmers from cultivating conventional food or feed crops on ground used for a GM crop test during the following growing season.

Food industry representatives and biotechnology opponents labeled the proposals inadequate. The rule that the food industry and certain environmental groups wanted was a provision that only nonfood crops could be used to produce pharmaceuticals. The USDA's position is that such a rule would not prevent contamination of nonfood GM plants with conventional crops. Rather, the agency chose to focus on methods intended to prevent any intermingling of GM plants and conventional plants.

Genetically Engineered Forest and Fruit Trees: Public Meeting
July 8 - 9, 2003

The Animal and Plant Health Inspection Service, USDA has released a notification for parties involved in those fields associated with the environmental release of genetically engineered trees, as well as other interested persons, that a public meeting will be held to provide a forum for discussion on the environmental safety, potential benefits, and risks of genetically engineered trees relative to traditional varieties.

The meeting will be held on Tuesday, July 8, 2003, from 8 a.m. to 4 p.m., and Wednesday, July 9, 2003, from 8:30 a.m. to 4 p.m. at the USDA Center at Riverside, 4700 River Road, Riverdale, MD.

This is an international meeting that will bring together representatives from the leading laboratories in transgenic animal research. This meeting is a follow up to the highly successful meetings held at Granlibakken in 1997, 1999 and 2001.

The conference will again focus on state-of-the-art science in the field of transgenic research. Presentations will address cutting-edge methodology, technical improvements, and current progress towards producing transgenic animals for biomedical and agricultural applications. The intent of these meetings is to bring together scientists to discuss progress, problems, and potential application of transgenic technology for animal applications.

The goal of this conference is to provide technical and content updates for high school and college biology educators. New applications in medicine, agriculture, and the environment continue to be discovered and discussed. Biotechnology 2003 will feature: pre-conference biotechnology boot camp, roundtable discussions, workshops, keynote speakers and much more. The Conference is organized by Fralin Biotechnology Center, Virginia Polytechnic Institute and State University.