ETHYLENE-INSENSITIVE 3-LIKE 2 regulates β-carotene and ascorbic acid accumulation in tomatoes during ripening

Abstract

ETHYLENE-INSENSITIVE 3/ETHYLENE-INSENSITIVE 3-LIKEs (EIN3/EILs) are important ethylene response factors during fruit ripening. Here, we discovered that EIL2 controls carotenoid metabolism and ascorbic acid (AsA) biosynthesis in tomato (Solanum lycopersicum). In contrast to the red fruits presented in the wild type (WT) 45 d after pollination, the fruits of CRISPR/Cas9 eil2 mutants and SlEIL2 RNA interference lines (ERIs) showed yellow or orange fruits. Correlation analysis of transcriptome and metabolome data for the ERI and WT ripe fruits revealed that SlEIL2 is involved in β-carotene and AsA accumulation. ETHYLENE RESPONSE FACTORs (ERFs) are the typical components downstream of EIN3 in the ethylene response pathway. Through a comprehensive screening of ERF family members, we determined that SlEIL2 directly regulates the expression of 4 SlERFs. Two of these, SlERF.H30 and SlERF.G6, encode proteins that participate in the regulation of LYCOPENE-β-CYCLASE 2 (SlLCYB2), encoding an enzyme that mediates the conversion of lycopene to carotene in fruits. In addition, SlEIL2 transcriptionally repressed L-GALACTOSE 1-PHOSPHATE PHOSPHATASE 3 (SlGPP3) and MYO-INOSITOL OXYGENASE 1 (SlMIOX1) expression, which resulted in a 1.62-fold increase of AsA via both the L-galactose and myoinositol pathways. Overall, we demonstrated that SlEIL2 functions in controlling β-carotene and AsA levels, providing a potential strategy for genetic engineering to improve the nutritional value and quality of tomato fruit.

Figure 1.

Phenotypic observation of ERI1 and the eil2 mutants. A) The mRNA levels of SlEIL2 in 3 SlEIL2 RNAi lines (ERI1, ERI2, and ERI3) and the WT. Tomato ACTIN transcript levels were used for normalization. Data represent means ± Sd (n = 3). Statistical analysis was performed by Tukey-test. Columns with different letters are significantly different (P < 0.05). B) Immunoblot analysis of the expression of SlEIL2 in ERI lines. ACTIN was used as the loading control. C) Identification of the CRISPR/Cas9-induced mutations in the eil2 mutant (eil2-1 and eil2-2) by sequencing. In each case, an additional nucleotide was added at position 442 of the SlEIL2 coding sequence, creating a termination codon. D) Immunoblot analysis of the expression of SlEIL2 in 2 eil2 mutant lines. ACTIN was used as the loading control. E) The leaf epinasty phenotype of eil2-1 and eil2-2 plants. Scale bar = 5 cm. F) Ripening stages of fruits between 20 to 45 d after anthesis (DAA) in the WT, ERIs, and eil2-1 lines. ERIs and eil2-1 fruits displayed a significantly delayed onset of fruit ripening compared with the fruits of WT. Scale bar = 2 cm. G) The absolute values of β-carotene, phytoene, lycopene and β-cryptoxanthin in ERIs and WT fruits at 30DAA, 35DAA, 39DAA and 45DAA. Data represent means ± Sd (n = 3). Statistical analysis was performed by Tukey-test. Columns with different letters are significantly different (P < 0.05). H) Fruit firmness of WT and SlEIL2 transgenic fruits at different stages. Fifteen fruits were examined for each measurement; error bars represent ± Sd. P < 0.05 (Tukey-test).

Figure 2.

Analysis of carotenoid metabolism in the ERI lines and the WT. A) Gene Ontology (GO) term enrichment analysis of DEGs of ERIs and WT fruits at 45 DAA. B) Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis of DEGs of ERIs and WT fruits at 45DAA. C) Transcriptome and metabolome correlation analysis of ERIs and WT fruits at 45DAA. Model of the carotenoid metabolic pathway is shown. Blue heat map indicates the expression of selected carotenoid-associated DEGs from the ERI vs. WT comparison. Red type indicates the carotenoid-associated metabolite levels in ERI and WT fruits. Tomato gene ID and the used data are shown in Supplemental Table S2.

Figure 3.

SlEIL2 regulates the expression of 4 ERF genes by binding to the EBS elements in their promoters. A) Diagram of the upstream regions of SlERF.J1, SlERF.G6, SlERF.H30, and SlERF.F12 and results of the yeast 1-hybrid assay. Yeast strain EGY48 was co-transformed with SlEIL2-GAD and SlERF.J1pro-EBS::LacZ, SlERF.G6pro-EBS::LacZ, SlERF.H30pro-EBS::LacZ, or SlERF.F12pro-EBS::LacZ; the corresponding EBS elements are shown in the upper diagram. Blue staining indicates binding. SlERFpro-EBS::LacZ, -GAD, -::LacZ, and SlEIL2-GAD were used as negative controls. B) EMSA. The respective EBS fragments of the SlERF genes were used as the probes (5′ biotin). The same probes without biotin were added as a competitive control (C-probes). Mutant probes, which were shown in the above diagram, were used as negative control. C) LUC reporter assay. Agrobacterium cells containing vectors expressing SlEIL2-FLAG or empty FLAG and Agrobacterium cells containing vectors expressing SlERFpro::LUC (the promoter region of SlERFs containing the SlEIL2 binding element) were co-injected into N. benthamiana leaves in the combinations shown at the bottom. After injection, the plants were cultured for 2 d prior to observation. The color bar indicates the intensity of the signal, with red as the maximum and purple as the minimum. Scale bar = 1 cm. D) pGreen0800 (a double LUC reporter system)-mediated LUC assay. The relative expression of SlERFpro::LUC was normalized to that of 35S::REN (internal control) (LUC/REN, mean ± Sd, n = 3). The combinations of Agrobacterium cells were as described above. **P < 0.01 (Student's t-test).

Figure 4.

SlERF.G6 and SlERF.H30 regulate SlLCYB2 expression. A) Above, diagram of the upstream regions of SlLCYB2, with the GCC element (G-BOX) marked in red; below, results of a yeast 1-hybrid assay to identify the binding of SlERF.G6 and SlERF.H30 to the G-BOX element in the promoter of SlLCYB2. SlERF.G6-GAD or SlERF.H30-GAD was co-transformed into yeast strain EGY48 with SlLCYB2pro-G-BOX::LacZ. Blue staining indicates binding. SlLCYB2pro-G-BOX::LacZ, -GAD, -::LacZ, and SlERF.G6/H30-GAD were used as the negative controls. B) EMSA. The GCC element described above was used as the probe (5′ biotin). The same probe without biotin was added as a competitive control (C-probes). Mutant probe, which were shown in the above diagram, was used as negative control. The left part of the gel shows the binding of SlERF.G6 to the GCC element, and the right part shows the binding of SlERF.H30 to the GCC element. C) LUC reporter assay. Agrobacterium cells containing vectors expressing SlERF.G6-FLAG, SlERF.H30-FLAG, or empty FLAG and Agrobacterium cells containing vectors expressing SlLCYB2pro::LUC (the promoter region of SlLCYB2 containing the GCC element) were co-injected into N. benthamiana leaves in the combinations shown at right. Scale bar = 1 cm. D) pGreen0800 (a double LUC report system)-mediated LUC assay. The relative expression of SlLCYB2pro::LUC was normalized to that of 35S::REN (internal control) (LUC/REN, mean ± Sd, n = 3). The combinations of Agrobacterium cells were as described above. **P < 0.01 (Student's t-test). E) An additional double LUC reporter assay was used to explore the role of SlEIL2 in SlERFs regulation of SlLCYB2. Agrobacterium cells containing vectors expressing SlEIL2-HA, SlERF.G6-FLAG, SlERF.H30-FLAG, or empty FLAG and Agrobacterium cells containing vectors expressing SlLCYB2spro::LUC were co-injected into N. benthamiana leaves. The combinations of Agrobacterium cells are shown at the bottom. The relative expression of SlLCYB2pro::LUC was normalized to that of 35S::REN (internal control) (LUC/REN, mean ± Sd, n = 3). The combinations of Agrobacterium cells were as described above. Columns with different letters are significantly different (P < 0.05, Tukey-test).

Figure 5.

SlEIL2 affects the ABA and ET contents in the tomato fruits. A, D) The ABA and ACC contents of WT and ERI fruits at 37DAA and 45DAA. The data are presented as means ± Sd of three biological replicates. P < 0.05 (Tukey-test). Columns with different letters are significantly different. B) Heat map comparisons of selected ABA biosynthesis-associated DEGs from the ERI vs. WT. Blue level indicates expression levels. The tomato gene ID is shown following each gene name. C) The ET contents of WT and ERI fruits at 30DAA, 33DAA, 37DAA and 45DAA. The data are presented as means ± Sd of three biological replicates. P < 0.05 (Tukey-test). Columns with different letters are significantly different. E) Heat map comparisons of selected ET biosynthesis-associated DEGs from the ERI vs. WT. The tomato gene ID is shown following each gene name.

Figure 6.

SlEIL2 is involved in the AsA biosynthesis in tomato fruits. Transcriptome and metabolome correlation analysis of ERIs and WT fruits at 45DAA. Model of the L-galactose and myo-inositol pathways of AsA biosynthesis were shown. Blue heat map comparisons indicate the expression of selected AsA biosynthesis-associated DEGs from the ERI vs. WT. Tomato gene ID and the used data are shown in Supplemental Table S2.

Figure 7.

SlEIL2 regulates the expression of SlGPP3/IMP3 and SlMIOX1. A) Above, diagram of the upstream regions of SlMIOX1 and SlGPP3/IMP3; below, results of a yeast one-hybrid assay to detect SlMIOX1 and SlGPP3/IMP3 regulated by SlEIL2. The EBS element is marked in red. SlEIL2-GAD was co-transformed into yeast strain EGY48 with SlMIOX1pro-EBS::LacZ and SlGPP3/IMP3pro-EBS::LacZ respectively. Blue staining indicates positive binding. SlMIOX1pro/ SlGPP3/IMP3pro -EBS::LacZ, -GAD, -::LacZ, and SlEIL2-GAD were used as the negative controls. B) Electrophoretic mobility shift assay (EMSA). The EBS elements described above were used as the probe (5′ biotin). The same probes without biotin were added as a competitive control (C-probes). Mutant probes, which were shown in the above diagram, was used as negative control. The left part of the gel shows the binding of SlEIL2 to the EBS element of SlMIOX1, and the right part shows the binding of SlEIL2 to the EBS element of SlGPP3/IMP3. C) Luciferase (LUC) reporter assay. Agrobacterium cells containing vectors expressing SlEIL2-FLAG or empty FLAG and Agrobacterium cells containing vectors expressing SlMIOX1pro/ SlGPP3/IMP3pro::LUC (the promoter region of SlMIOX1 or SlGPP3/IMP3 containing the EBS element) were co-injected into N. benthamiana leaves, respectively. After injection, the plants were cultured for 2 days prior to observation. Scale bar =1cm. D, E) mRNA levels of SlMIOX1 in WT and ERI fruits at 45DAA. Tomato ACTIN transcript levels were used for normalization. The data are presented as means ± Sd of three biological replicates. P < 0.05 (Tukey-test). Columns with different letters are significantly different.

Figure 8.

Model of the role of SlEIL2 in enhancing β-carotene and AsA accumulation. Model showing the mechanism underlying the role of SlEIL2 in regulating AsA and β-carotene accumulation in ripe tomato fruits. According to the model, SlEIL2 regulates the expression of ERF.H30 and ERF.G6 to influence SlLCYB2 transcript levels, thereby controlling the conversion of lycopene to β-carotene and then enhancing ABA accumulation. SlEIL2 also participates in the 2 AsA biosynthetic pathways by regulating SlMIOX1 and SlGPP3/IMP3 expression.