Improved antioxidant content for food applications

Miroarrays have been used to monitor changes in gene expression - During normal ripening of Microtom fruits and - In fruits of transgenic plants with altered phenolic profiles. The following experiments have been carried out: - Microtom fruit ripening. Microtom fruits, harvested at various stages of ripening were profiled for gene expression changes using microarrays. Based on multivariate analyses it was concluded that the green fruits were most deviating from the coloured fruit stages. Most dramatic gene expression changes occur in the transition from green to turning: groups of green genes are down-regulated, whereas many ripening-related genes are induced. In the over-ripe stage, a number of genes are down-regulated relative to ripe fruits. This experiment provided us with a good overview of the gene expression changes that occur during normal fruit ripening and the results confirmed the current insight in the molecular and biochemical changes which occur during tomato fruit development. - Gene expression changes in Ros and Ros/Del fruits (result 33971). Microarrays have been used to find target genes in microtom fruits over-expressing the Antirrhinum transcription factor genes Rosea (Ros) alone or both Ros and the Delila (Del) together. Metabolic analyses revealed that Ros fruits contained increased levels of phenylpropanoids and simple phenolic acids, whereas Ros/Del fruits contained increased levels of both phenylpropanoids, simple phenolic acids and strongly increased levels of delphinidin-type anthocyanins. Microarray analysis confirmed the metabolite analyses and revealed a strong up-regulation of phenylpropanoid and flavonoid genes in Ros and Ros/Del plants. Microarray results were confirmed and extended by real-time SYBR Green RT-PCR analyses. The gene expression results showed that Ros alone leads to up-regulation of phenylpropanoid genes and early flavonoid genes, but not the late flavonoid genes required for the production of anthocyanins. Ros and Del together are sufficient to up-regulate all the genes required for the production of delphinidin-type anthocyanins. The key gene required for the accumulation of anthocyanins in Ros/Del plants is F35H, which is required for the formation dihydromyricitin, the precursor of delphinidin-type anthocyanins. In addition to phenylpropanoid and flavonoid genes we surprisingly observed a strong upregulation of the gene encoding the antimicrobial protein gamma-thionin in Ros/Del tomatoes. Althoug we can not exclude the possibility that the upregulated gamma-thionin gene reveals a secundary effect of the transgenic plants, it is also possible that these trancription factors may activate other metabolic pathways involved in the production of defence-related metabolites. - Gene expression in XPT/FLS fruits. Comparison of gene expression profiles of XPT/FLS fruits with WT microtom fruits did not reveal any significant changes in gene expression in the transgenic plants analysed. - Gene expression in F3Hi and CHI fruits. F3Hi fruits are transgenic plants in which the F3H gene is downregulated by an RNAi approach. This resulted in plants whith strongly reduced levels of the flavonols quercetin and kaempferol (A. Bovy, personal communication). Microarray analyses of F3Hi fruit peel revealed that, except for the down-regulated F3H gene, no other significant effects on gene expression were observed in these plants. Similar results were obtained with CHI overexpressing tomato plants, which accumulate high levels of quercetin glucosides in the fruit peel. - Gene expression in CHSi fruits. In CHSi plants, the complete flavonoid pathway is blocked by RNAi-mediated suppression of both tomato CHS genes. This resulted in flavonoid-lacking tomato fruits with an altered fruit morphology. Microarray analysis revealed that blocking the flavonoid pathway leads to up-regulation of both upstream phenylpropanoid and downstream flavonoid genes, suggesting the existence of metabolic control mechanisms, which regulate the flux through the flavonoid pathway. In addition to phenylpropanoid and flavonoid genes, a number of NAM-family transcription factor genes were upregulated in CHSi plants as well. The microarray analyses of these transgenic tomato lines revealed a very good insight into the regulation of the flavonoid pathway and the possibilities and limitations of engineering metabolic pathways in general and the flavonoid pathway in particular. The results can be used to design novel strategies for engineering tomato plants with an optimal content of flavonoids, and with minimal pleiotrpoic side effects. In addition, identified key genes involved in fruit ripening and/or flavonoid production can be used as targets for marker development in breeding programs.

There is increasing interest in the ability of diets rich in flavonoids to modulate age-related diseases and promote healthy ageing. Thus, there is a growing interest in the production of flood plants with increasing amounts of flavonoids and consequently enhancement of flavonoid biosynthesis in chosen crops, such as tomatoes, is likely to increase human consumption of these compounds that have the potential to benefit human health. A large germplasm collection of tomato lines deposited in the germplasm collection at the Institute of Plant Genetics and Crop Pant Research, Gatersleben, has been screened for the profiles of flavonoids in the fruits. Tomato accessions have been grown in the field for initial screening, and a second round of flavonoid profiling was performed for selected lines grown under greenhouse conditions. Profiles of flavonoids and other soluble phenylpropanoids were obtained after HPLC separation of methanolic fruit extracts by diode array detection. HPLC analysis of tomato fruits revealed a large variability in the contents of flavonoids and other phenolic compounds. Accessions with overall low accumulation of phenolic compounds as well as with overall high contents of many compounds were observed. Assessment of the overall antioxidant capacity in vitro revealed a correlation with the accumulation of the flavonoid rutin, a major phenolic compound of tomato fruits. Therefore, with respect to the potential health benefits of a diet with high flavonoid intake, selected tomato accessions form the germplasm collections constitute a valuable resource for future breeding efforts to enhance the fruit flavonoid contents. Human studies have shown that increased intake of flavonoids lead to an increased accumulation of flavonoids and their metabolites in the circulation. However, this is unlikely to mean that the plasma antioxidant status is altered significantly even if greater amounts of rutin and its derivative metabolites are present in the circulation. Indeed, flavonoids are unlikely to express beneficial action in vivo through out-competing antioxidants, such as ascorbate, which are present at much higher concentrations. More likely is the possibility that flavonol metabolites derived from increased rutin intake may act favourably by exerting effects on specific signalling pathways through selective actions at different components of a number of protein kinase and lipid kinase sihgnalling cascades. These findings have been published (Free Radical Research September 2005).

During the funding period major objectives have been the creation and validation of genetic variability in flavonoid accumulation in tomato fruits, the biochemical and molecular characterisation of flavonoid-enriched tomatoes and feeding studies to validate possible health promoting effects of high-flavonoid tomatoes. To design high-flavonoid tomato fruits three general strategies, aimed at overcoming possible bottlenecks in biosynthetic capacity, were followed. These included (i) increased synthesis of precursors, (ii) inhibition of competing pathways and (iii) ectopic expression of regulatory proteins. To increase carbon flux towards flavonoid biosynthesis it was planned to increase the amount of phosphoenolpyruvat (PEP) and erythrose-4-phosphate in plastids of transgenic tomato plants. To increase PEP availability in plastids two novel PEP-synthesising pathways based on expression of PGM/enolase (partner 1) and PPDK (partner 3), respectively, and overexpression of the plastidic PEP-translocator (Partner 3) were achieved. PGM/enolase as well as PPDK expression lead to increased PEP levels in leaf plastids of transgenic plants. To increase plastidic erythrose-4-phosphat content in transgenic tomato plants, over-expression of transketolase (partner 7) and of XPT (partner 3) was attempted. To increase transketolase activity, respective coding regions were isolated from E. coli, spinach, P. patens and tested for functionality in E. coli, P patens and transgenic tomato plants. Activity could be demonstrated in E. coli, however, transgenic tomato plants did not show any increase in transketolase activity. XPT expression in transgenic tomato plants was verified by northernblotting. Besides attempts to increase PEP and E4P levels, enhancement of the first committed step in shikimate biosynthesis, the condensation of phosphoenoylpyruvate and erythrose-4-phosphate by the action of 3-deoxy-arabino-heptulosonate 7-phosphate synthase, was planned to be increased (partner 7). However, despite several attempts to increase activity of DAHPS in transgenic tomato plants, none of the regenerated plants revealed increased enzyme activities. Since plant and non-plant sources have been tested as gene donor in combination with various promoters, this result can only be explained by rigorous post-translational regulation of and DAHPS activity. In addition to the discussed approaches, transgenic tomato plants with ectopic expression of flavonol synthase, dehydroflavonol reductase (DFR), isoflavone synthase (IFS) and hydroxycinnamoyl CoA quinate transferase (HQT) have been developed. In conclusion, numerous transgenic lines altered in expression of individual pathway genes have been created and are available for further analysis.

In cooperation with BASF Plant Science (BPS) it was evaluated which animal system might be suitable for evaluating characterized tomato fruit material with elevated flavonoid content for disease preventing effects. Different systems were compared and the human CRP (C.reactive protein) transgenic mouse model was chosen. CRP appears to be a good risk marker (and risk factor) for cardiovascular diseases. Interleukin 1beta induces levels of CRP in human beings as well as in transgenic mice carrying the human CRP gene. Statins as well as fenofibrate decrease expression of transgenic CRP (positive control). A scientific plan was designed to elucidate whether flavonoid containing tomato material would reduce CRP expression as well. TNO Pharma (Leiden, The Netherlands) was identified as a partner running the mouse experiments. A contract for collaboration between partner 1 and TNO Pharma was signed. Two different extracts derived from transgenic tomato skin were compared with non-transgenic material and a fenofibrate diet. The tomato material was prepared by partner 5 and delivered to TNO pharma. As the feeding trials started only in November/December a neutral prolongation of the PROFOOD project was granted. Four groups of hCRPtg mice received chow diet either not supplemented (control) or supplemented with wildtype tomato peel (0.4% w/w, Tom-Con), or flavonoid-enriched tomato peel (0.4% w/w, Flavo), or fenofibrate (0.1% w/w, positive control) for a total of 7 weeks. After 6 weeks mice were intraperitoneally challenged with IL-1beta to raise hCRP expression. After recovery in week 7, mice were subjected to a 2 weeks-rebound period and fed a standard chow diet to test whether the tomato peel-induced effects on basal hCRP expression disappeared. RESULTS: Basal hCRP-levels rapidly decreased in both tomato peel-fed groups and remained significantly lower than starting values until the end of the intervention period (except for the Flavo group in week 4). Two animals of the Flavo group looked unhealthy from week 3 onwards and also displayed a high plasma fibrinogen level, which is an inflammation marker independent of hCRP. When these unhealthy animals were excluded, the maximal hCRP-lowering effect in the Flavo group was 55%(P<0.05; week 4) as compared to the chow-fed control group, whereas the hCRP-lowering effect in the Tom-Con group was less pronounced and only 44% (P<0.05; week 4). The additional effect of flavonoid-enriched tomato peel compared to wildtype tomato peel on basal hCRP levels became significant in week 6 and 7. Both types of tomato peel did not suppress an IL-1beta-induced expression of hCRP, whereas the positive control, fenofibrate, strongly suppressed this acute inflammatory effect of IL-1beta. After a 2-weeks rebound period in which all groups received standard chow diet, plasma hCRP levels recovered in all treatment groups and returned to the initial levels at t=0, except in the Tom-Con group, in which hCRP levels remained decreased. CONCLUSION: The relatively strong reduction of basal plasma hCRP levels achieved with tomato peel indicates that tomato peel-feeding strongly dampens low-grade chronic inflammation associated with cardiovascular disease. The anti-inflammatory potency of tomato peel is lower than that of fenofibrate and does not allow to suppress an acute, IL-1beta-based inflammatory stimulus. Compared to control tomato peel, flavonoid-enriched tomato peel is more advantageous because its inflammation-dampening effect is a) stronger and b) very specific (cf. reversibility of the hCRP-lowering effect). These beneficial properties of flavonoid-enriched tomato peel may have relevance for the prevention of atherosclerotic lesion development. A second study on possible health promoting effects of high-flavonoid toamtoes was carried out using the model system C. elegans. The tomato extracts displayed a remarkable wide spectrum of activities with some that depending on the flavonoid content and their particular composition reduce effectively ageing processes in the nematode C. elegans in vivo. It is highly suggestive that similar phenomena take place also in humans.

Transgenic tomato lines expressing the regulatory proteins (transcription factors) Rosea1 and Delila (from Antirrhinum majus) specifically in ripening fruit. The initial lines produced were in MicroTom. The fruits express high levels of all the genes committed to anthocyanin biosynthesis (including transporters and genes encoding enzymes) as demonstrated by suppression subtraction hybridisation and analysis using the custom microarray developed by PRI Wageningen. They produce on average 1mg per gram fresh weight anthocyanins, which are principally delphinidin 5-glucoside (p-coumaroyl) rutinoside and petunidin 5-glucoside (p-coumaroyl) rutinoside, as confirmed by LC/MS. There arrre also increases in other phenolics(soluble) which are currently being identified. There is an overall increase in flavonoid content of at least 150-fold, and anthocyanins are produced throughout the flesh of the fruit as well as in the skin. There is a significant increase in anti-oxidant capacity as determined by the TEAC assay. This shows that the antioxidant capacity increases about 3-fold in the soluble (hydrophillic) fraction, whereas the antioxidant capacity of the hydrophobic fraction remains unaltered. The High Anthocyanin (purple) trait has been introduced into two commercial tomato backgrounds - Money Maker and Alisa Craig. The trait is fully heritable in a Mendelian fashion. These results have been described at 8 pleanary/keynote addresses at prestigous scientific meetings: Key Innovative Features: This is the first time that high levels of anthocyanins have been produced in tomato fruit. This trait is restricted to the fruit so that growth and yield are not impaired. These lines have significantly increased antioxidant capacities. The fruit also have significantly improved textural properties and shelf life. Current Status: We are introducing this trait into other tomato lines; specifically lines mutant in the enzyme DFR such that high levels of flavonols can be engineered in tomato fruit. We are investigating the basis for the improved textural properties and shelf life of the transgenic lines. Use of the result and its expected benefits: We hope that the lines we have produced will provide the material for breeding tomatoes with increased levels of health-promoting flavonoids. The feeding trials performed under PROFOOD with Caenorhabditis elegans showed no effects but are difficult to interpret with respect to health benefits in human. We have prepared large quantites of the tomatoes for mouse feeding studies to be performed on 3 models of disease (Cardio-vascular disease, Cancer and Age-related degeneration). Inclusion of control and purple tomatoes in the diet will allow tests on effects in disease progression, and cellular markers for disease. These analyses are being conducted in collaboration with clinicians and epidemiologists under the sponsorship of an EU-FP6 STReP project, FLORA.

Two promoters were analyzed for expression for expression in tomato fruits. The promoters of the soluble starch synthase 3 gene (SSS3) from potato and of the glucan phosphorylase gene (Pho1) from broad bean, respectively, were analysed during tomato fruit development. Fruit development was divided into 10 stages depending on size and colour of the fruits. First samples were taken one week after flowering (stage I). Stage VI corresponds to the breaker stage (green colour changes to yellow). Stage X represents a fully ripe fruit. From each stage slices were stained in X-Gluc and later compared with results from quantitative MUG analyses. For each time point 15 samples were collected from different fruits for a quantitative MUG assay. SSS3 and Pho1 promoters show a comparable expression pattern. During fruit development GUS activity increases and achieves highest values at stage VI (about 25 nmol MU/min mg protein). In stages VIIX GUS activity decrease to about 5 nmol MU/min mg protein. X-Gluc staining suggested that blue staining occurred throughout the whole fruit including the vasculature. These results reveal that both promoters can be used for metabolic engineering tomato fruits. The E8 promoter is frequently found in the literature and has been shown to be induced at the onset of ripening. This contrasts with SSS3 and Pho1 promoters as these drive highest GUS expression in green fruits which then decreases during ripening. Thus the promoters can be complementary used to the E8 promoter.

A dedicated tomato microarray was constructed containing 1000 oligonucleotides, each representing a gene putatively involved in regulatory and metabolic pathways important for fruit quality. These genes were selected from the existing public TIGR tomato database (LEGI) through a bioinformatics approach. In collaboration with all partners of the consortium a set of 608 unigenes were selected, representing metabolic and regulatory pathways which have relevance to aspects such as health and quality: Aromatic amino acids, sugar metabolism, flavonoid, stilbene and lignin, phenylpropanoid, ascorbate, folate and one carbon metabolism, sterol, vitamin E, vitamin K and carotenoids, terpenoid and MYB transcription factors. In addition, a set of 178 unigenes was selected that showed homology to members of the Arabidopsis transmembrane protein database "Agamemnon". The remaining 214 genes, required for a collection of 1000 genes, consisted of genes that were known to be induced or repressed during tomato fruit ripening, together with control genes required for background subtraction and normalisation. Unique 70-mer oligonucleotides have been designed for all of the 1000 selected genes. These oligonucleotides have been spotted onto 100 microarray slides. Conditions for spotting and hybridisation of these oligonucleotide arrays were optimised and hybridisation results could be confirmed using other detection methods such as realt-time RT-PCR, indicating that this microarray platform led to reliable results. Microarray slides have been used to monitor gene expression changes in tomato fruits, in order to obtain insight in (i) the process of tomato fruit ripening and (ii)the effect of transgene expression on quality-related metabolic pathways The results of these studies have been presented at several scientific conferences and have resulted in additional collaborations with partners in and outside PROFOOD. The oligonucleotides were synthesised in excess and may be used to design additional batches of microarrays which can be used in future collaborations for gene expression profiling studies of quality-related genes in tomato.

Several thousand individual T-DNA activation-tagged lines from Arabidopsis containing an immobilised element (4x35S Enhancer, Basta Resistance, Inverted repeats) have been screened in co-operation with Partner 1 for high phenylpropanoid content by HPLC-Analysis using fluorescence and UV techniques. In the first round of screening, about ten mutant lines, which show modified HPLC profiles under greenhouse conditions have been defined as putants and selected for further studies. Three out of ten lines showing a high phenolics phenotype were studied in detail in a new set of recapitulation lines. Thus, accumulation of new chemoprotective antioxidants in mutant lines were determined by overexpression of genes with unknown functions - At1g18570, At2g20095 and At3g10820. Constructs for the expression of those genes in tomato plants have been created and sent to Partner 1. Results have been published in A. Schneider, T. Kirch, T. Gigolashvili, H-P. Mock, U. Sonnewald, R. Simon, U-I. Flügge, W. Werr (2005) A Transposon-Based Activation-Tagging Population in Arabidopsis thaliana (TAMARA) and its Application in the Identification of Dominant Developmental and Metabolic Mutations. FEBS Lett. 29;579(21):4622-8.

We aimed at reducing lignin biosynthesis to enhance availability of these precursors and thereby stimulate the production of soluble, potentially health-promoting, phenolic compounds in tomato (Lycopersicon esculentum Mill.). We first identified and characterized two tomato genes encoding cinnamoyl-CoA reductase (CCR), a key enzyme in the formation of lignin monomers. Transgenic plants exhibiting reduced lignin content were subsequently obtained through an RNAi strategy targeting one of these genes. As anticipated, the total level of soluble phenolics was higher in stems and leaves of the transformants as compared to control plants. This was correlated with an increased antioxidant capacity of the corresponding plant extracts. Analysis of the soluble phenolic fraction by HPLC-MS revealed that vegetative organs of CCR down-regulated plants contained higher amounts of chlorogenic acid and rutin, and accumulated new metabolites undetectable in the wild type, such as N-caffeoyl putrescine and kaempferol rutinoside. In fruits, CCR down-regulation triggered the moderate accumulation of two new compounds in the flesh, but the total phenolic content was not affected. Although the prospects of exploiting such a strategy for crop improvement are limited, our results provide further insight into the control of the phenylpropanoid pathway in Solanaceae. This result is to be published in a scientific journal (submitted 21st September 2005).

Changes of transcripts and metabolites have been measured in transgenic lines altered in flavonoid metabolism by different partners of the PROFOOD consortium. All data have been combined and stored in a database located at the MPI-Golm (Partner 7. Genes used encode proteins invovled in the transport of precursors across the plastid envelope membrane, and transcription factors involved in the regulation of the expression of genes involved in flavanoid synthesis. In addition, a data visualisation software tool has been developed, which allows tomato expression data to be visualised in a user-driven manner on diagrams or maps. The MapMan visualization software was developed previously for Arabidopsis thaliana transcript and metabolite profiling. It offers the possibility to paint out microarray or metabolite profiling experiments onto diagrams of pathways or processes, and visualize the responses of gene expression in a biological context at either a gene for gene level or at different hierarchical levels of the functional categorisation system. This plant specific ontology is a key element of MapMan. The principle of the MapMan ontology is a hierarchical BIN-based structure. Each BIN comprises items of similar biological function e.g. glycolysis, and can be further split into subBINs corresponding to submodes of the biological function. It presently contains over 700 hierarchical categories. In contrast to the GO ontology (The Gene Ontology Consortium, 2000), the MapMan BIN structure minimizes the redundancy between BINs. The hierarchical structure is reflected in its BINCode by separating dots similar to EC numbers. Thus BIN 1 (photosynthesis) can, for example, be subdivided to 1.1 (Photosynthesis.light reaction) and 1.2 (Photosynthesis.Calvin cycle). This allows the relationship between subBINs and BINS to be deduced immediately from the name of the BIN as well as from its numerical BINCode. A very important goal was to extend this software platform to allow it to be applied to crop plants, which will allow knowledge generated by analysis of the model system Arabidopsis to be used to support crop genomic research. We have developed a classification of the spots represented on the tomato TOM1 array according to the MapMan ontology. This allows the software systems developed for Arabidopsis to be used for the visualisation system of tomato and other Solanaceous transcript profiling data. A similar approach has been taking to display metabolite data on diagrams of metabolic pathways. The results have been accepted for publication.

Deliverables

Share this page

Download