Fruit-specific overexpression of lipoyl synthase increases both bound and unbound lipoic acid and alters the metabolome of tomato fruits

Abstract

Introduction: Lipoic acid (LA) is a key, yet overlooked player in primary metabolism, due to its role as a cofactor for various multi enzymatic complexes such as the E2 subunits of pyruvate dehydrogenase (PDH) and alpha-ketoglutarate dehydrogenase (kGDH). In recent years, this molecule has seen renewed interest given its strong antioxidant properties and its applications as a dietary supplement. The mechanisms behind the synthesis of LA in vivo have been elucidated, identifying lipoyl synthase (LIP1) as the key enzyme required for this process.

Methods: Therefore, in this work, we used the fruit-specific polygalacturonase (PG) promoter to guide Solanum lycopersicum (tomato) LIP1 (SlLIP1) overexpression in stably transformed tomatoes.

Results: The resulting plants presented higher transcript levels of SlLIP1 in a fruit-specific manner, accumulated more bound and unbound LA yet lacked major phenotypic defects at both the vegetative and reproductive growth stages. Furthermore, changes in the expression of genes related to LA synthesis were explored and a metabolomic study was carried out. Specific metabolite patterns were clearly distinguishable between untransformed and stably transformed lines. For instance, trehalose 6-phosphate, GABA and proline levels were generally higher, whilst glucose 6-phosphate and UDP-glucose levels were lower in fruits of the SlLIP1 transformants.

Discussion: In addition, as the overexpression of SlLIP1 results in lower transcript levels of E2 PDH and E2 kGDH, and enhanced amounts of LA-bound targets, we speculate that the proportion of unlipoylated E2 subunits of PDH and kGDH may have decreased. This work could assist in obtaining crops with a higher LA content and therefore improved health benefits.

Keywords: Solanum lycopersicum; TCA cycle; antioxidant; lipoylation; polygalacturonase promoter.

Figure 1

Generation of tomato plants overexpressing SlLIP1. (A) Diagram of the pCP vector used to transform tomato cv 'Micro-Tom'. (B) Three independent T0 lines (4-months old, L31, L4 and L3) and two in the T2 generation (3-months old, L31 and L4) exhibited normal development compared to WT. (C) Three representative fruits of PG-SlLIP1 lines at T0 and T2 generations.

Figure 2

SlLIP1 transcript levels and LA content in fruits of T0 and T2 PG-SlLIP1 transgenic tomato lines. (A) qRT-PCR analysis of SlLIP1 was performed on three different fruits, each with two technical replicates (n=6 ± SD), normalized against Actin7 and calibrated to WT levels. (B) Quantification of immunoblot signals of protein-bound LA using an anti-LA antibody, normalized to the signal of an anti-actin antibody in WT and PG-SlLIP1 transgenic fruits. All T0 and T2 samples were subjected to this analysis at least three times, whilst 5 replicates are considered for WT. Bars show means ± SD. (C) Detection of unbound LA in WT and T2 PG-SlLIP1 fruits (n=9 ± SD). p ≤ 0.05 (*), p ≤ 0.005 (**), and p ≤ 0.0005 (***) and not significant (n.s.).

Figure 3

Relative transcript levels of genes involved in LA metabolism in fruits of T0 and T2 PG-SlLIP1 transgenic tomato lines. The analyzed genes are: (A) Plastid lipoyl synthase (SlLIP1p): an isoform of lipoyl synthase that contains a signal peptide targeting to chloroplasts. (B) Octanoyl transferase (SlLIP2): participates in de novo LA biosynthesis. (C) Lipoate protein ligase (SlLplA): involved in the salvage pathway. (D) Mitochondrial E2 subunit of pyruvate dehydrogenase (E2 PDHm). (E) Plastidial E2 subunit of pyruvate dehydrogenase (E2 PDHc). (F) E2 subunit of alpha-keto glutarate dehydrogenase (E2 kGDH). (G) SAM synthetase (SAMS1): involved in SAM biosynthesis. Transcript levels were normalized against Actin7 and calibrated to WT levels. n=6 (T0 and T2) or 12 (WT) ± SD. p ≤ 0.05 (*), p ≤ 0.005 (**), p ≤ 0.0005 (***) and not significant (n.s.).

Figure 4

Targeted metabolomic analysis of fruits of T0 PG-SlLIP1 transgenic tomato lines. (A) Principal component analysis (PCA) was performed using detected metabolites as the predictor variable and genotype as the response variable. (B) Heatmap analysis of the top 50 significantly-altered metabolites identified by ANOVA. Columns represent the average of three biological replicates for each genotype (WT and PG-SlLIP1 lines L31, L4 and L3). The distance measure for clustering was Euclidean, and the clustering algorithm used was Ward.

Figure 5

Metabolic comparison of fruits of PG-SlLIP1 transgenic and WT tomato lines. The graphical representation illustrates the variance in the metabolic profile of each transgenic fruit (n=3 ± SD) compared to WT (n=3 ± SD). Green arrows beside metabolite names indicate an increase in accumulation in transgenic fruits compared to WT, while red arrows indicate a decrease, according to the t-test with significance set at p ≤ 0.05 (*), (**) and (***). Gray metabolites were not measured but are included in the scheme to facilitate understanding. G1P, Glucose 1-phosphate; UDPG, UDP-glucose; UDP-GlcA, UDP-glucuronic acid; UDP-Ara, UDP-arabinose; UDP-Xyl, UDP-xylose, UDP-Api, UDP-apiose; UDP-Gal, UDP-galactose; ADP-Glc, ADP-glucose; G6P, Glucose 6-phosphate; T6P, Trehalose 6-phosphate; F6P, Fructose 6-phosphate; M6P, Mannose 6-phosphate; R5P, Ribulose 5-phosphate; S7P, Sedoheptulose 7-phosphate; R1,5-bP, Ribulose 1,5-biphosphate; E4P, Erythrose 4-phosphate; Ery/Thre, Erythritol/Threitol; PGA, 2–3 phosphoglyceric acid; DHA-P, Dihydroxyacetone phosphate; GLY-P, Glycerol phosphate; PEP, Phosphoenolpyruvate; SAM, S-adenosyl methionine; GABA, gamma-aminobutyric acid; OH-proline, hydroxyproline; ArgSuc, Argininosuccinate.