Manipulation of phytoene levels in tomato fruit: effects on isoprenoids, plastids, and intermediary metabolism

Abstract

In tomato (Solanum lycopersicum), phytoene synthase-1 (PSY-1) is the key biosynthetic enzyme responsible for the synthesis of fruit carotenoids. To further our understanding of carotenoid formation in tomato fruit, we characterized the effect of constitutive expression of an additional tomato Psy-1 gene product. A quantitative data set defining levels of carotenoid/isoprenoid gene expression, enzyme activities, and metabolites was generated from fruit that showed the greatest perturbation in carotenoid content. Transcriptional upregulation, resulting in increased enzyme activities and metabolites, occurred only in the case of Psy-1, Psy-2, and lycopene cyclase B. For reactions involving 1-deoxy-d-xylulose5-phosphate synthase, geranylgeranyl diphosphate synthase, phytoene desaturase, zeta-carotene desaturase, carotene isomerase, and lycopene beta-cyclase, there were no correlations between gene expression, enzyme activities, and metabolites. Perturbations in carotenoid composition were associated with changes in plastid type and with chromoplast-like structures arising prematurely during fruit development. The levels of >120 known metabolites were determined. Comparison with the wild type illustrated that key metabolites (sucrose, glucose/fructose, and Glu) and sectors of intermediary metabolism (e.g., tricarboxylic [corrected] acid cycle intermediates and fatty acids) in the Psy-1 transgenic mature green fruit resembled changes in metabolism associated with fruit ripening. General fruit developmental and ripening properties, such as ethylene production and fruit firmness, were unaffected. Therefore, it appears that the changes to pigmentation, plastid type, and metabolism associated with Psy-1 overexpression are not connected with the ripening process.

Figure 1.

Profiles of Carotenoids and Other Isoprenoids Present in Both Psy-1 and Wild-Type Fruit at the Mature Green Stage. (A) and (B) Chromatograms for the wild type and Psy-1 variety [PS1-C26R2 (T₇)], respectively, recorded at 450 nm. (C) and (D) Wild type and Psy-1 chromatograms recorded at 400 nm, respectively. (E) and (F) Chromatograms recorded at 350 nm for the wild type and Psy-1 variety, respectively. (G) and (H) Chromatograms from the wild type and Psy-1 variety recorded at 286 nm. Chromatogram components are numbered, and those in bold represent components unique to Psy-1. For peak identity, UV/Vis spectral characteristics, and chromatographic properties, see Supplemental Table 1 online. The mature green stage used for this analysis represents 36 d after anthesis.

Figure 2.

Changes in Isoprenoid/Carotenoid Gene Expression Levels and Enzyme Activities Resulting from the Expression of an Additional Psy-1. The Psy-1 variety PS-1C26R2 (T₇) was used for the determination of gene expression and enzyme activity. One fruit (at the mature green stage, 36 to 37 d after anthesis) from three representative Psy-1 and wild-type plants was pooled and pulverized into a homogeneous powder as described in Methods. Total RNA was then extracted from an aliquot of this material. Quantitative real-time RT-PCR was performed with gene-specific primers for (1) Dxs, (3) Ggpps-1, (4) Ggpps-2, (5) Psy-1, (6) Psy-2, (7) Pds, (8) Zds, (9) CrtISO, (10) the Cyc-B gene, and (11) Lcy-B. Enzymes are numbered similarly, with the addition of (2) IPP isomerase. Expression data were normalized to the expression of actin. The data represent se ± (n = 3 to 8). Student's t tests illustrate statistical significance (* P < 0.05, ** P < 0.01, and *** P < 0.001). The black bars indicate wild-type determinations, and the checkered bars are for the Psy-1 transgenic variety. Enzyme activities were performed on extracts determined from the same tissue used for RNA analysis. Specific assays for DXS, GGPPS, PSY, PDS, and lycopene cyclase were performed as described in Methods. The protein levels used were as follows: 440 μg for the Psy-1 and wild-type extracts used in the assays of phytoene desaturation and lycopene cyclization; 200 and 100 μg for the wild type and Psy-1 extracts, respectively, used to assay DXS, IPI, GGPPS, and PSY. Typically, substrate levels were ∼50,000 dpm phytoene and 20,000 dpm lycopene formed from 1 μCi ¹⁴C-IPP. In all cases, activity was expressed as dpm incorporated/mg protein/h. Determinations were performed at least in triplicate. Data are shown as se (n = at least 3).

Figure 3.

Visualization of Plastids Present in Wild Type and Psy-1 Variety. The Psy-1 variety PS-1C26R2 (T₈) was used for this analysis. pl, plastoglobules; t, thylakoid grana; Cl, chloroplasts; Cr, chromoplasts. (A) and (B) A typical wild-type cell from mature green fruit (36 d after anthesis) visualized under bright-field (A) and fluorescence emission (B) magnification ×10. (C) and (D) A typical cell from Psy-1 mature green fruit (bottom section of fruit) visualized under bright-field and fluorescence emission, respectively, ×10. (E) and (F) The higher magnification (×40) of the Psy-1–derived cell from the top section of the mature green fruit under bright-field and fluorescence emission, respectively. Green arrow shows plastids autofluorescing due to the presence of chlorophyll, and red arrow indicates collection of plastids in which chlorophyll is absent. (G) and (H) Transmission electron micrographs of the wild type and Psy-1 variety, respectively.

Figure 4.

Metabolites Detected by Metabolomic Analysis and Displayed onto Schematic Representations of the Biochemical Pathways. Changes arising from wild-type mature green fruit compared with wild-type ripe fruit. Data correspond to those displayed in Table 2. Green indicates an increased level of metabolite, with a significant to threefold increase in pale green, a threefold to eightfold increase in green, and more than eightfold is dark green. Gray indicates no significant change, while blue indicates that the metabolite was not detected in the samples. White indicates that the compound cannot be detected using the analytical parameters. Red coloration has been used to represent decreased metabolite levels; dark red is below eightfold, red is below twofold to fivefold, and pale red is below twofold. Aco, aconitic acid; l-Asc, ascorbic acid; citramal, citramalic acid; Cit, citric acid; dehydroasc, dehydroascorbic acid; Fum, fumaric acid; Mal, malic acid; 2-oxoglut, 2-oxoglutaric acid; Succ, succinic acid; Thre, threonic acid; 5HT, 5-hydroytryptamine; 5-OxoPRO, 5-oxo-proline; Arab, arabinose; DXP, deoxyxylulose-5-phosphate; F6P, fructose-6-phosphate; G6P, glucose-6-phosphate; 3-CaQuinic, 3-caffeoylquinic acid; CGA, chlorogenic acid; FPP, farnesyl diphosphate; GPP, geranyl diphosphate.

Figure 5.

Metabolites Detected by Metabolomic Analysis and Displayed onto Schematic Representations of the Biochemical Pathways. Changes arising from wild-type mature green fruit compared with Psy-1 mature green fruit. Abbreviations are the same as in Figure 4.

Figure 6.

PCA of Metabolites Present in Wild-Type and Psy-1 Mature Green and Ripe Fruit. Data analysis of the variables was performed as described in Methods. Components 1 and 2 were responsible for 40 and 16% of the total variance, respectively.

Figure 7.

Comparison of General Ripening-Related Parameters Found in the Wild Type and Psy-1 Variety. The Psy-1 variety PS-1C26R2 (T₁₀) was used for this analysis. Data sets have been represented as se and statistical differences determined using Student's t test. IM, immature fruit; MG, mature green; BR, breaker; RR red ripe. (A) Ethylene production of detached fruits. Three fruit at each stage were picked randomly from six plants, and three to four technical replications made. (B) Determination of fruit firmness over fruit development and ripening. Four fruit were picked randomly from six plants and two to five technical replicates taken at the equator of each fruit. (C) The pH of each fruit homogenate; three fruit were harvested randomly from six plants. Each fruit was quartered and a determination performed on four random quarters. (D) The days after anthesis (DAP) required to reach observationally defined developmental and ripening stages. Typically, four determinations were made from flowers on the first three trusses. (E) and (F) Fruit weight and diameter on four fruit from the same truss over development and ripening, for the immature, mature, and breaker stages. In the case of the ripe fruit, 22 wild-type fruit were used to determine the average fruit weight and 21 Psy-1 fruit used to determine the average fruit weight. Ripe fruit diameters were measured on four fruit.