Auxin-activated MdARF5 induces the expression of ethylene biosynthetic genes to initiate apple fruit ripening

Abstract

The gaseous plant hormone ethylene induces the ripening of climacteric fruit, including apple (Malus domestica). Another phytohormone, auxin, is known to promote ethylene production in many horticultural crops, but the regulatory mechanism remains unclear. Here, we found that auxin application induces ethylene production in apple fruit before the stage of commercial harvest, when they are not otherwise capable of ripening naturally. The expression of MdARF5, a member of the auxin response factor transcription factor (TF) family involved in the auxin signaling pathway, was enhanced by treatment with the synthetic auxin naphthaleneacetic acid (NAA). Further studies revealed that MdARF5 binds to the promoter of MdERF2, encoding a TF in the ethylene signaling pathway, as well as the promoters of two 1-aminocyclopropane-1-carboxylic acid synthase (ACS) genes (MdACS3a and MdACS1) and an ACC oxidase (ACO) gene, MdACO1, all of which encode key steps in ethylene biosynthesis, thereby inducing their expression. We also observed that auxin-induced ethylene production was dependent on the methylation of the MdACS3a promoter. Our findings reveal that auxin induces ethylene biosynthesis in apple fruit through activation of MdARF5 expression.

Keywords: ARF5; DNA methylation; apple; auxin; ethylene; fruit ripening.

Fig. 1

Auxin‐induced ethylene production and expression of ethylene‐related genes in apple. (a–e) Apple fruit were harvested at 145 DAFB (days after full bloom) in 2018 and treated with naphthaleneacetic acid (NAA), 1‐methylcyclopropene (1‐MCP), or with 1‐MCP followed by NAA (1‐MCP + NAA) (a), stored at room temperature for 20 d and sampled every 5 d. Fruit not receiving any treatment were used as a control. Bars, 1 cm. Ethylene production was measured (b) and the expression levels of MdACS3a (c), MdACS1 (d) and MdACO1 (e) were detected by quantitative reverse transcription (qRT)‐PCR. DAH, days after harvest. (f–j) Apple fruit were harvested at 115 DAFB and treated with NAA (f), stored at room temperature for 20 d and sampled every 5 d. Bars, 1 cm. Ethylene production was measured (g) and the expression levels of MdACS3a (h), MdACS1 (i) and MdACO1 (j) were investigated by qRT‐PCR. Fruit not receiving any treatment were used as a control. NAA, fruit treated with NAA; 1‐MCP, fruit treated with 1‐MCP; 1‐MCP + NAA, fruit treated with 1‐MCP for 12 h followed by NAA treatment. Three biological experiments from independent RNA extractions for each group of fruit were analyzed. Values represent means ± SE. Asterisks indicate significant difference as determined by a Student’s t‐test (**, P < 0.01); ns, no significant difference.

Fig. 2

MdARF5 is required for naphthaleneacetic acid (NAA)‐induced ethylene biosynthesis in apple fruit. (a) MdARF5 expression detected by quantitative reverse transcription (qRT)‐PCR in apple fruit treated with or without NAA. Fruit tissues were the same as in Fig. 1(f). Control, untreated fruit; NAA, NAA‐treated fruit. Three biological replicates from independent RNA extractions for each group of fruit were analyzed. (b–h) Silencing of MdARF5 in apple fruit (MdARF5‐AN) at 110 DAFB (days after full bloom) using Agrobacterium‐mediated transient transformation. MdARF5‐AN fruit were harvested from the apple tree 5 d after infiltration and immediately treated with NAA (b), and stored at room temperature for 20 d. Fruit infiltrated with empty pRI101 vector were used as controls. Scale bars, 1 cm. Expression level of MdARF5 (c) was evaluated in MdARF5‐AN fruit by qRT‐PCR to confirm successful infiltration. Ethylene production was measured (d) and the expression of MdACS3a (e), MdACS1 (f), MdACO1 (g) and MdERF2 (h) were determined in MdARF5‐AN fruit by qRT‐PCR. Three biological experiments with independent RNA extractions were performed. Values represent means ± SE. Asterisks indicate significant differences as determined by a Student’s t‐test (**, P < 0.01); ns, no significant difference. DAH, days after harvest; DAI, days after infiltration

Fig. 3

MdARF5 positively regulates ethylene biosynthetic genes through binding to their promoters. (a) ChIP (chromatin immunoprecipitation)‐PCR analysis showing MdARF5 binding to the MdACS3a promoter (1250 bp) in vivo. Cross‐linked chromatin samples were extracted from MdARF5‐FLAG‐overexpressing fruit calli treated with or without naphthaleneacetic acid (NAA) and precipitated with FLAG antibody. Eluted DNA was used to amplify sequences neighboring the AuxRE (auxin responsive element, ARF binding site) by quantitative PCR (qPCR). Five fragments (P1–P5) were analyzed. Fruit calli overexpressing the FLAG sequence alone were used as a negative control. (b) GUS (β‐glucosidase) activation assay showing that MdARF5 positively regulates the MdACS3a promoter. The MdARF5 effector vector and MdACS3a promoter reporter vector were co‐infiltrated into wild tobacco (Nicotiana benthamiana) leaves to analyze GUS activity. (c) ChIP‐PCR analysis showing MdARF5 binding to the MdACS1 promoter (1213 bp) in vivo. The ChIP assay was conducted as in Fig. 3(a). Three fragments (P1–P3) were analyzed. (d) GUS activation assay showing that MdARF5 positively regulates the MdACS1 promoter. The MdARF5 effector vector and the MdACS1 promoter reporter vector were co‐infiltrated into wild tobacco leaves to analyze GUS activity. (e) ChIP‐PCR analysis showing MdARF5 binding to the MdACO1 promoter (1000 bp) in vivo. The ChIP assay was conducted as in Fig. 3(a). Three fragments (P1–P3) were analyzed. (f) GUS activation assay showing that MdARF5 positively regulates the MdACO1 promoter. The MdARF5 effector vector and MdACO1 promoter reporter vector were co‐infiltrated into wild tobacco leaves to analyze GUS activity. For ChIP‐PCR, the ChIP assay was repeated three times and the enriched DNA fragments in each ChIP were used as one biological replicate for qPCR. For the GUS activation assay, three independent transfections were analyzed. Values represent means ± SE. Asterisks indicate significant differences as determined by a Student’s t‐test (**, P < 0.01); ns, no significant difference.

Fig. 4

MdARF5 together with MdERF2 regulate ethylene biosynthesis. (a) The expression level of MdERF2 was determined by quantitative reverse transcription (qRT)‐PCR in apples treated with or without naphthaleneacetic acid (NAA). Fruit tissues were the same as in Fig. 1(f). Control, untreated fruit; NAA, NAA‐treated fruit. Three biological experiments with independently extracted RNA for each group of fruit were analyzed. Values represent means ± SE. Asterisks indicate significant differences as determined by a Student’s t‐test (**, P < 0.01); ns, no significant difference. (b) ChIP (chromatin immunoprecipitation)‐PCR assay showing MdARF5 binding to the MdERF2 promoter (1380 bp) in vivo. ChIP‐PCR was conducted as in Fig. 3(a). Six fragments (P1–P6) were analyzed. (c) GUS (β‐glucosidase) activation assay showing that MdARF5 positively regulates the MdERF2 promoter. The MdARF5 effector vector and MdERF2 promoter reporter vector were co‐infiltrated into wild tobacco leaves to analyze GUS activity. (d) ChIP‐PCR assay showing MdERF2 binding to the MdACS3a promoter (1250 bp) in vivo. Cross‐linked chromatin samples were extracted from MdERF2‐GFP (green fluorescent protein)‐overexpressing fruit calli, which were treated with or without NAA and precipitated with GFP antibody. Eluted DNA was used to amplify sequence neighboring the DRE‐motif (dehydration responsive element, ERF binding site) by qPCR. Three fragments (P1–P3) were analyzed. Fruit calli overexpressing the GFP sequence were used as a negative control. (e) GUS activation assay showing that MdARF5 together with MdERF2 promotes MdACS3a promoter activity. The MdARF5 and MdERF2 effector vectors were co‐infiltrated into wild tobacco leaves with the MdACS3a promoter effector vector, before analysis of GUS activity. (f) GUS activation assay showing that MdERF2 reduces MdACS1 promoter activity promoted by MdARF5. The MdARF5 and MdERF2 effector vectors were co‐infiltrated into wild tobacco leaves with the MdACS1 promoter effector vector to analyze GUS activity. For ChIP‐PCR, the ChIP assay was repeated three times and the enriched DNA fragments in each ChIP were used as one biological replicate for qPCR. For the GUS activation assay, three independent transfections were analyzed. Values represent means ± SE. Asterisks indicate significant differences as determined by a Student’s t‐test (**, P < 0.01); ns, no significant difference.

Fig. 5

MdACS3a expression is crucial for auxin‐induced ethylene biosynthesis in apple fruit. (a–g) Apple fruit harvested at 95 DAFB (days after full bloom) were treated with naphthaleneacetic acid (NAA), stored at room temperature for 20 d, and sampled every 5 d (a, fruit phenotypes). Fruit not receiving any treatment were used as control. Bars, 1 cm. Ethylene production was measured (b) and the expression levels of MdACS3a (c), MdARF5 (d), MdERF2 (e), MdACS1 (f) and MdACO1 (g) were determined by quantitative reverse transcription (qRT)‐PCR. DAH, days after harvest. (h–l) MdACS3a silencing in on‐tree apple fruit (MdACS3a‐AN) at 110 DAFB performed using Agrobacterium‐mediated transient transformation. Fruit were harvested 5 d after infiltration and immediately treated with NAA, stored at room temperature for 20 d and sampled. Fruit infiltrated with empty pRI101 vector were used as a control (h, fruit phenotypes). Bars, 1 cm. MdACS3a expression (i) was determined to confirm successful infiltration as determined by qRT‐PCR. Ethylene production was measured (j) and expression levels of MdACS1 (k) and MdACO1 (l) were determined by qRT‐PCR. For qRT‐PCR, three biological experiments from independent RNA extractions were performed. Values represent means ± SE. Asterisks indicate significant differences as determined by a Student’s t‐test (**, P < 0.01); ns, no significant difference.

Fig. 6

DNA methylation analysis of the MdACS3a promoter in apple fruit. (a) Detection of MdACS3a promoter methylation. Genomic DNA was extracted from the 0 and 10 DAH samples in Fig. 5(a), digested with McrBC and used as template for PCR‐based methylation detection. The MdACS3a promoter was divided into five fragments (S1–S5), and its methylation level was investigated by standard PCR. DNA incubated without McrBC was used as a control. Un, DNA incubated without McrBC; M, DNA incubated with McrBC; DAH, days after harvest; NAA10, fruit treated with NAA and sampled at 10 DAH. Numbers below the PCR bands indicate the shift in intensity. (b–f) MdACS3a promoter methylation level detected using BSP (bisulfite sequencing PCR). The same genomic DNA as in (a) was used for bisulfite modification, and the product was used as a template to amplify the five MdACS3a promoter regions: S1 (b), S2 (c), S3 (d), S4 (e) and S5 (f). Control, nontreated DNA; NAA, NAA‐treated DNA. DAH, days after harvest. Three biological experiments from independent DNA extractions were performed. Values represent means ± SE. Asterisks indicate significant differences as determined by a Student’s t‐test (**, P < 0.01); ns, no significant difference.

Fig. 7

Model of auxin‐activated MdARF5 inducing ethylene biosynthesis in apple fruit. After demethylation of the MdACS3a promoter, its expression is initiated. Auxin‐activated MdARF5 binds to the MdACS3a and MdERF2 promoters to enhance their expression, and MdERF2 then upregulates MdACS3a expression. Moreover, auxin‐activated MdARF5 binds to the MdACS1 and MdACO1 promoters and induces their expression. MdERF2 suppresses MdACS1 expression through promoter binding in order to buffer the auxin‐activated ethylene biosynthesis. Symbols/abbreviations: ‘+’, promotion; ‘−’, suppression; solid arrow, direct regulation; dotted arrow, unclear regulation mechanism; AuxRE, auxin responsive element (ARF binding site); DRE, dehydration responsive element (ERF binding site); SAM, S‐adenosyl methionine; ACC, 1‐aminocyclopropane‐1‐carboxylic acid.