Université Louis Pasteur Faculté des Sciences de la Vie

janvier 2008

Master Sciences du Végétal UE : Ingénierie métabolique des substances naturelles végétales Cours de M. François BERNIER Durée: 2 heures

Des plantes intra-géniques

Crop genetic engineering relies on the introduction of foreign DNA into plant genomes. Although genetically engineered traits provide valuable alternatives to those available through conventional breeding, there is public concern about the consumption of foods derived from transgenic plants. This concern raises the question of whether crops can be improved by inserting only native DNA into their genomes. Here, we discuss how rapid advances in molecular biology make it possible to use plants themselves as DNA sources. Native genes and regulatory elements can be reintroduced into plants without the need to use selectable markers. By also using transfer DNAs that are derived from within the targeted compatibility group, genetically engineered plants can now be produced that lack any foreign DNA. The transformation of wild species into domesticated crops, a process that started about 10 millennia ago, may be considered one of the greatest accomplishments of mankind. One of the methods employed by traditional plant breeding to enhance crop performance assesses numerous lines for hundreds of traits in multisite replicated plot field trials. A second more cumbersome and time-consuming approach captures or creates new traits and transfers them into existing varieties. This Mendelian aspect of plant breeding crosses varieties with wild relatives to produce F1 hybrids or, alternatively, self-fertilizes plants that were subjected to chemical mutagens to generate segregating M2 families. Individual plants that contain the new trait are extensively backcrossed to

remove unlinked wild or mutated DNA. Although initially considered as artificial, trait introgression and mutation breeding are currently perceived as acceptable methods in crop improvement. An additional method that dramatically affects the integrity of crop genomes but is accepted as one of the tools of traditional plant breeding is based on the fusion of somatic cells from related but sexually incompatible plant species. Such interspecies fusions result in the development of amphiploid hybrids, and considerable backcrossing and ploidy reductions are needed to develop varieties suitable for release. An important issue associated with traditional plant breeding arises from the fact that genetic variation, although randomly induced at the DNA level, is screened for phenotypically. Cultivars resulting from this practice will not only display most of the traits that the breeder selected for but also contain undesirable features and lack positive attributes that were not considered during the selection process. Eager to accelerate the process of crop improvement, plant biologists domesticated the crossspecies DNA transfer system of the plant pathogenic bacterium Agrobacterium tumefaciens and developed the first transgenic plants containing fungal and bacterial DNA. During the subsequent two decades, a multitude of genetically modified plants were generated that contain DNA that could not have been introgressed through any available breeding method. An initial lack of knowledge about the molecular biology of plants limited the ready exploitation of these sources for crop improvement. As a result, genes for agronomically important traits such as herbicide tolerance and insect resistance were mainly derived from other organisms, primarily bacteria and viruses, whose genomes can be more easily subjected to molecular analysis. In addition to the foreign genes of interest, most transgenic plants contain bacterial selectable marker genes that provide tolerance of antibiotics, herbicides or drugs. Although the stable integration of such genes makes it possible to identify the rare transformed cells and to regenerate plants from them, their lingering presence in crops complicates the regulation process and negatively affects public acceptance of the final products. Here, we will discuss these issues and indicate how some of them can be addressed by applying intragenic methods. Instead or relying on unpredictable genome modifications, these new methods specifically recombine native genetic elements in vitro and insert the linked DNA back into the plant using marker-free transformation. One of the elements, the plant-derived transfer (P-) DNA, replaces the Agrobacterium T-DNA by functioning as vehicle for DNA delivery into the plant cell. Thus, intragenic modification incorporates neither uncharacterized DNA (as is the case with traditional breeding) nor foreign DNA (as is typical for transgenic modification; see Figure 1) into a plant’s genome. We argue that intragenic applications produce GM crops that are inherently “low risk” and should be cleared through the regulatory process in a timely and cost-effective manner. Furthermore, we believe that intragenic crops may be considered more acceptable to consumers than transgenic plants.

The dominant potato variety (in the United States) Russet Burbank was developed 131 years ago. Although suffering from multiple deficits, this potato’s excellent storage characteristics solidified its position as the preferred “white-fleshed” variety for French fry processing in the United States. One alternative variety that is considered to be the grower’s favorite but has not been able to challenge Russet Burbank is Ranger Russet. Market penetration of this 15-year-old variety has been limited to only about 20% of that of Russet Burbank. Although Ranger Russet combines superior yield with disease resistance, adaptability, tuber uniformity, and high levels of starch, it is particularly sensitive to tuber discolorations that are linked to impact-induced bruise. This phenomenon is caused by leakage of polyphenol oxidase (Ppo) from damaged plastids into the cytoplasm. The subsequent oxidation of polyphenols triggers a precipitation of black melanin that greatly affects tuber quality during prolonged storage. Furthermore, Ranger Russet accumulates high levels of glucose and fructose during cold storage. These reducing sugars react with free amino acids during high-temperature processing of the potato. Accumulation of the resulting Maillard reaction products (including the neurotoxin acrylamide) lowers consumer appeal by darkening French fries. Therefore, the quality of Ranger Russet is generally compromised if tubers are stored for longer than about 8 weeks. Both the storage and nutritional characteristics of potato can be improved by employing methods in genetic engineering. Silencing of the Ppo gene was shown to lower the extent of black spot bruise sensitivity, and the down-regulated expression of the starch-associated R1 or phosphorylase-L (PhL) genes lowered the accumulation of reducing sugars in coldstored potato tubers. Here, the simultaneous silencing of the Ppo, R1, and PhL genes is shown to provide black spot bruise tolerance and greatly reduced levels of cold-induced sweetening. Interestingly, French fries derived from the modified tubers displayed a strongly enhanced visual appearance and improved aroma while accumulating much lower levels of acrylamide. RESULTS AND DISCUSSION In an attempt to turn the storage deficits of Ranger Russet into strengths, we employed a multigene silencing approach. For this purpose, a DNA segment was produced that contains fragments of three “undesirable” genes. One of the fragments was derived from the 5’-untranslated trailer of the Ppo allele POT32 that is predominantly expressed in mature tubers. The other two fragments represent the 3’-untranslated leaders of the R1 and PhL genes, the down-regulated expression of which is known to slightly lower the extent of cold-induced sweetening. Two copies of the “triple-gene fragment” DNA segment were inserted as inverted repeat between the tuber-specific promoter of the granule-bound starch synthase (Gbss) gene and the terminator of the potato ubiquitin-3 (Ubi3) gene (Figure 2). Positioning of the resulting silencing construct between two tandemly repeated potato St01 elements that function as T-DNA border alternatives created an all-native potato transfer (P-) DNA. A plasmid carrying this P-DNA, designated pSIM371, was introduced into an Agrobacterium LBA4404 strain also harboring a conventional binary vector. This second “LifeSupport” vector, pSIM368, carried both the bacterial

neomycin phosphotransferase (nptII) positive selectable marker gene and the cytosine deaminase (codA) gene for negative selection inserted between conventional T-DNA borders. The resulting double-vector strain was used to infect 21900 Ranger Russet stem explants. Upon cotransfer of the P-DNA and T-DNA, the explants were subjected to kanamycin for 5 days to select for transient nptII gene expression. Proliferation of cells containing stably integrated nptII T-DNAs was subsequently prevented by transferring explants to media containing 5-fluorocytosine (5FC), a chemical that is converted into toxic 5-fluorouracil (5FU) by the codA gene product. A total of 3822 shoots that survived the double selection were genotyped by PCR for the presence of the desired PDNA and the absence of foreign DNA. This analysis found 85% of plants to contain plasmid backbone DNA. After elimination of both this large group of backbone-DNA-containing plants and an additional group that still carried the T-DNA, 256 marker-free and all native DNA (intragenic) plants were selected for propagation and planting in the greenhouse. Analyses demonstrated that 48 intragenic lines (19%) were effectively silenced for Ppo, often at levels below 15% of untransformed controls. Both the frequency and extent of silencing were comparable to those of plants that had been transformed with pSIM217, a construct designed to only down-regulate the expression of the POT32 Ppo gene. To confirm that reduced Ppo activity levels would provide black spot bruise tolerance, tubers were physically impacted and, after 2 weeks, used for processing. French fries obtained from both pSIM371 and pSIM217 tubers were found to be free of discolorations, including visible white spot impact damage. In contrast, control fries developed extensive black spot bruise symptoms that covered about half of the fry surface. Phenotypes of the 48 intragenic and bruise-tolerant lines were further analyzed by growing them in the field. Untransformed controls and three groups of transgenic “benchmark” lines were included in this trial. The first group of benchmarks consisted of five of the above-described Pposilenced pSIM217 lines. A second group of five lines was successfully silenced for the PhL gene through expression of the silencing construct of pSIM216, whereas the R1-targeted silencing construct of pSIM332 had been used to produce a third group of gene-suppressed lines (Figure 2). The extent of cold-induced sweetening and its impact on tuber quality was assessed by storing field-grown tubers for 3 months at 4 °C. Tuber analyses demonstrated that 43 of the bruise tolerant pSIM371 lines accumulated 30-60% of wild-type glucose levels. These levels were generally lower than those obtained with the benchmark lines that had been silenced for only one of the starchassociated genes (60-90%) (Figure 2). Thus, our results confirm that a single multigene silencing construct can be used effectively to down-regulate the expression of multiple genes. Furthermore, they indicate an either additive or synergistic effect of the combined R1/PhL gene-silencing approaches. By rating for color, color variation, and appearance defects, fries from tubers of five intragenic Ranger Russet lines were found to outperform those of both Ranger Russet and Russet Burbank. In contrast, the visual appearance of the benchmark lines was not significantly different from that of untransformed controls (data not shown). In addition to improved visual appearance and consistency, the intragenic fries from freshly harvested tubers also displayed a significantly enhanced overall aroma. Apart from aroma, the fries from intragenic lines were not different from control and benchmark fries in terms of texture, toughness, or mealiness. Collectively, our data indicate that the simultaneous suppression of the starch mobilization gene R1 or PhL enhances key quality and sensory attributes to deliver a more consistent and flavorful processed product. The low levels of reducing sugars in intragenic tubers prompted us to determine acrylamide levels in French fries obtained from these tubers. Field-grown tubers from a second-year trial, were cold stored, processed, and biochemically analyzed. French fries generated from intragenic tubers contained only about a third of the acrylamide that accumulates in control fries. Overall data from the two field trial seasons suggested that the incorporated traits had no negative effect on agronomic performance. Analyses of tubers from the second-year trial did not reveal any differences between untransformed and intragenic lines in terms of plant and tuber typeness.

Figure 2. (A) Diagram of constructs for multigene (pSIM371) and single-gene (pSIM216, 217 and 332) silencing. P>, promoter of the Gbss gene; PpoT, trailer of the Ppo gene; PhLL, leader of phosphorylase-L gene; R1L, leader of R1 gene; S, spacer from the ubiquitin-7 gene intron; T, terminator of the ubiquitin-3 gene.

L-Ascorbic acid (vitamin C) is synthesized from hexose sugars. It is an antioxidant and redox buffer, as well as an enzyme cofactor, so it has multiple roles in metabolism and in plant responses to abiotic stresses and pathogens. Plant-derived ascorbate also provides the major source of vitamin C in the human diet. A combination of radiolabelling, mutant analysis and transgenic manipulation provides evidence for multiple pathways of ascorbate biosynthesis (Figure 3). Since vitamin C was first isolated, there have been numerous reports on the physiological and metabolic processes in which it is involved. Ascorbic acid is crucial to the maintenance of a healthy immune system and is required for the synthesis of collagen, carnitine, and neurotransmitters. In general, it acts as an enzyme cofactor, free radical scavenger, and donor and acceptor in electron transfer reactions. As a consequence, its most vital role in the human body is as a water-soluble antioxidant. Vitamin C is the single most important specialty chemical manufactured in the world. Its industrial synthesis is a lengthy procedure involving microbial fermentation and a series of chemical steps. Cloning of a gene (AKR) encoding aldo-keto reductase from strawberry. The cultivated strawberry (Fragaria ananassa) is an important small fruit crop in temperate regions. Ripening of the fruit occurs over a short time period and is accompanied by a change in the expression of many genes. Ripe strawberry fruit are rich in ascorbic acid, containing an average of 60 mg per 100 g fresh weight, although this varies among cultivars. It has been reported that ascorbic acid content increases as fruit ripens. Early studies reported the conversion of Dgalacturonic acid to L-ascorbic acid in ripening strawberry fruit. A key enzymatic activity of this pathway was an NAD(P)H-dependent reductase that was present in the soluble fraction of some plant extracts.

Figure 3. The network of proposed biosynthetic pathways for ascorbate in plants. Enzymes: 1, GDP-D-Man pyrophosphorylase; 2, GDP-Man-30,50-epimerase; 3, GDP-L-Gal phosphorylase (GDP-L-Gal:orthophosphate guanylyltransferase); 4, L-Gal 1-phosphate phosphatase; 5, L-Gal dehydrogenase; L-GalL dehydrogenase; 7, GDPD- mannose-4,6-dehydratase; 8, GDP-4-keto-6deoxy-D-mannose 3,5-epimerase-4-reductase; 9, D-galacturonate reductase; 10, myo-inositol oxygenase; 11, D-glucuronate reductase; 12, aldonolactonase; 13, L-GulL oxidase or dehydrogenase. L-Fuc, L-fucose; L-Gal, L-galactose; L-GalL, L-galactono- 1,4-lactone; GDP, guanosine diphosphate; L-Gul, L-gulose; L-GulL, L-gulono-1,4-lactone; D-Man, D-mannose; UDP, uridine diphosphate.

AKR2 encodes a D-galacturonic acid reductase. In the ascorbic acid biosynthetic pathways proposed in higher plants, three reactions can be identified as putatively catalyzed by an AKR enzyme: the oxidation of L-galactose using NAD+ as the cofactor and the reduction of either D-glucuronic or D-galacturonic acid, both using NADPH as the cofactor. We assayed the enzyme activity of a recombinant AKR2, obtained as a glutathione Stransferase (GST)-fused protein in Escherichia coli, with the three substrates but did not detect any activity. As an alternative strategy, we expressed AKR2 ectopically in A. thaliana plants. Western analysis of 17 independent transgenic lines revealed several lines with high levels of AKR2. We assayed the enzymatic activity in the crude extracts of three transgenic lines, using as controls both the wild type and a line transformed with the empty vector pSOV2. We found that D-galacturonic acid reductase activity was enhanced in the transgenic lines relative to the controls, whereas activities of D-glucuronic acid reductase or L-galactose oxidase activities were no different from those in the control plants. To demonstrate fully that AKR2 itself was responsible for this activity, we immunopurified the protein from line 17 to electrophoretic homogeneity. The purified enzyme showed very high and specific NADPH-dependent reduction of D-galacturonic acid, with a value 250-fold higher per milligram of protein than that of the crude extract. Hereafter we renamed AKR2 as GalUR. Vitamin C levels in strawberry fruit correlate with the expression of GalUR. GalUR expression was specific to the receptacle tissue of strawberry fruit, both at the mRNA and

protein levels, and its expression increased during ripening, being highest in fully mature red fruit. The amount of ascorbic acid in ripening strawberry fruit broadly correlates with the expression of GalUR. These results suggest that GalUR activity contributes to a substantial proportion of the ascorbic acid content of ripe strawberry fruit. Overexpression of GalUR increases vitamin C content in A. thaliana. We next set out to determine whether transgenic A. thaliana lines ectopically expressing GalUR showed increased levels of ascorbic acid. The ascorbic acid content (in young seedlings) of three lines was two to three times higher than that of A. thaliana control plants. In addition, feeding the A. thaliana plants with D-galacturonic acid resulted in increased ascorbic acid content of the transgenic lines, whereas the level in the control A. thaliana plants remained unchanged. EXPERIMENTAL PROTOCOL Preparation of transgenic plants. A 956 bp fragment containing the complete ORF of the GalUR gene was amplified by PCR using a high-fidelity polymerase enzyme and oligonucleotides containing PstI and HindIII sites. This product was subcloned in the pSOV2 binary vector, which contains the 35SCaMV promoter, after digestion with the PstI and HindIII restriction enzymes.

Questions: 1- Qu’est-ce qu’une plante intragénique? Expliquer brièvement le processus permettant d’obtenir une plante intragénique. 2- Quels sont les avantages et inconvénients des plantes intragéniques par rapport aux plantes transgéniques et aux plantes obtenues par selection traditionnelle? Croyez-vous que la réglementation des plantes transgéniques doivent s’appliquer aux plantes intragéniques ? 3- D’après vous, quel problème pourrait survenir chez une plante intragénique, lié à la nature même des gènes manipulés ? 4- Décrire le principe de sélection des lignées transformées permettant d’éviter la présence de gène de sélection. 5- Quels sont les avantages de la lignée de pommes de terre pSIM371 par rapport aux variétés Russet Burbank et Ranger Russet ? 6- Les lignées pSIM216, pSIM217 et pSIM332 sont-elles intragéniques ou transgéniques ? A quoi servent-elles ? 7- Pensez-vous que la voie de synthèse de la vitamine C à partir de D-galacturonate soit la voie principale dans toutes les parties de la plante ? Pourquoi ? 8- Comment les auteurs ont-ils démontré que le gène AKR2 codait l’enzyme GalUR ? Pourquoi la protéine purifiée à partir de E. coli ne présentait-elle aucune activité ? 9- La stratégie décrite pour augmenter la quantité de vitamine C chez A. thaliana est transgénique et non intragénique. Pourquoi ? Comment pourrait-on faire des plantes intragéniques contenant une quantité accrue de vitamine C ? Les approches seraient-elles les mêmes pour les différents organes de la plante ? 10- Décrire brièvement une autre application (de votre choix) des plantes intragéniques.