The regulatory mechanism of fruit ripening revealed by analyses of direct targets of the tomato MADS-box transcription factor RIPENING INHIBITOR

Abstract

The developmental process of ripening is unique to fleshy fruits and a key factor in fruit quality. The tomato (Solanum lycopersicum) MADS-box transcription factor RIPENING INHIBITOR (RIN), one of the earliest-acting ripening regulators, is required for broad aspects of ripening, including ethylene-dependent and -independent pathways. However, our knowledge of direct RIN target genes has been limited, considering the broad effects of RIN on ripening. In a recent work published in The Plant Cell, we identified 241 direct RIN target genes by chromatin immunoprecipitation coupled with DNA microarray (ChIP-chip) and transcriptome analysis. Functional classification of the targets revealed that RIN participates in the regulation of many biological processes including well-known ripening processes such as climacteric ethylene production and lycopene accumulation. In addition, we found that ethylene is required for the full expression of RIN and several RIN-targeting transcription factor genes at the ripening stage. Here, based on our recently published findings and additional data, we discuss the ripening processes regulated by RIN and the interplay between RIN and ethylene.

Keywords: :

• fruit ripening
• tomato
• RIPENING INHIBITOR
• MADS-box transcription factor
• chromatin immunoprecipitation
• ethylene

In a developmental process unique to plant species bearing fleshy fruits, ripening drastically changes fruit characteristics including color, texture, aroma and nutritional composition. Ripening is thus a key step to determine fruit quality, and the understanding of the genetic program regulating fruit ripening will provide useful information to facilitate breeding of high quality fruit crops. In tomato (Solanum lycopersicum), ripening is regulated mainly by ethylene and several transcription factors (TFs).^1,2 One of the earliest-acting TFs, the tomato MADS-box RIPENING INHIBITOR (RIN), is required for both ethylene-dependent and -independent ripening regulatory pathways.^2,3 The rin mutation causes a severe inhibition of a broad range of ripening processes including softening, carotenoid accumulation and ripening-associated ethylene production⁴ (Fig. 1). This suggests that RIN acts as a master regulator of ripening and recent studies revealed that RIN may directly regulate the expression of two dozen ripening-associated genes.^5-9 However, our knowledge of RIN target genes has been limited, considering the central role of RIN in regulation of ripening. In our recent work, published in The Plant Cell, we report a large-scale identification of direct RIN targets by chromatin immunoprecipitation coupled with DNA microarray (ChIP-chip) and transcriptome analysis.¹⁰ In the report, we analyzed RIN binding to DNA from ripening tomato fruits by ChIP-chip using anti-RIN antibodies and a microarray for tomato gene promoters. This ChIP-chip analysis detected 1,046 RIN binding sites and, in combination with transcriptome data for wild type and rin mutant tomato fruits,⁶ we identified 241 genes as direct RIN targets. These targets are positively or negatively regulated by RIN, generally contain a RIN binding site in their promoters, and include 11 known ripening-related genes.

Figure 1. The rin mutation inhibits tomato fruit ripening. Wild type tomato fruits turn red and soften at the ripening stage (left), whereas the rin mutant fruits do not exhibit these changes (right). The rin mutation also inhibits the ripening-associated ethylene production and rise in respiration rate.¹⁴ The induction of RIN at the onset of ripening is not influenced by exogenous ethylene.³ In addition, the rin fruits show induction of some ethylene-responsive genes but do not ripen in response to exposure to exogenous ethylene,¹⁴ suggesting that RIN is required for both ethylene-dependent and independent pathways.²

Functional classification of the direct RIN targets with gene ontology (GO)

Our results in the recent paper¹⁰ suggested that RIN participates in many biological pathways or processes during ripening, especially in ethylene production, lycopene accumulation and stress responses. For a more comprehensive interpretation of RIN involvement in ripening, we here provide an additional functional classification of the direct RIN targets based on similarity to Arabidopsis proteins with GO information.

In the positively regulated subset of direct RIN targets, GO analysis showed a significantly higher frequency of genes classified into categories related to ethylene synthesis (GO:0008898 and GO:0009693) than the frequency in the whole genome (Table 1). A similar enrichment was found in categories related to responses to biotic or abiotic stresses (GO:0009816, GO:0051707, GO:0009611, GO:0016165 and GO:0040007) (Table 1), which may reflect increased susceptibility of ripening fruit to pathogens.¹ These results are consistent with those with the Arabidopsis protein annotation database provided by the Munich Information Center for Protein Sequences (MIPS) as reported.¹⁰ By contrast, it may seem strange that the GO analysis also found a significantly higher frequency in the category “Flower development” (GO:0009908) (Table 1). However, this reflects differences in function between tomato and Arabidopsis homologous genes, resulting in tomato genes involved in fruit ripening regulation, such as APETALA2a (AP2a), COLORLESS NON RIPENING (CNR) and RIN, being categorized in “Flower development” based on the annotation of their Arabidopsis homologs AP2, SQUAMOSA promoter binding protein and SEPALLATA genes. The implementation of functional gene annotation suitable for fleshy-fruit bearing plant species will enable more informative prediction of functional categories associated with RIN. This may help clarify the role of RIN-targeting genes or biological processes¹⁰ for which the association with ripening remains unclear.

Table 1. Overrepresented GO categories in the direct RIN target genes

GO ID	Category^1)	GO term	Positively-regulated (135 annotated genes)		Tomato genome (28,638 annotated genes)		P value	Examples of the classified direct RIN target genes
			Classified genes	ratio (%)	Classified genes	ratio (%)
GO:0009816	P	Defense response to bacterium, incompatible interaction	5	3.70	62	0.22	1.6E-05	GSH, LOXs, OSMs
GO:0051707	P	Response to other organism	5	3.70	97	0.34	1.2E-04	LOX, OSMs, UGTs
GO:0008898	F	Homocysteine S-methyltransferase activity	2	1.48	3	0.01	2.2E-04	HMTs
GO:0009611	P	Response to wounding	7	5.19	259	0.90	2.7E-04	C4H, 4CL, GH3, LOXs
GO:0009693	P	Ethylene biosynthetic process	4	2.96	29	0.10	1.7E-05	ACO6, ACS2, ACS4, E8
GO:0009908	P	Flower development ^2)	5	3.70	129	0.45	4.3E-04	AP2a, CNR, GSH, NAP, RIN
GO:0016165	F	Lipoxygenase activity	3	2.22	33	0.12	6.4E-04	LOXs
GO:0040007	P	Growth	3	2.22	35	0.12	7.6E-04	C4H, LOXs
GO ID	Category	GO term	Negatively-regulated (101 annotated genes)		Tomato genome (28,638 annotated genes)		P value	Examples of the classified direct RIN target genes
			Classified genes	Ratio (%)	Classified genes	Ratio (%)
GO:0003700	F	Sequence-specific DNA-binding transcription factor activity	17	16.83	1955	6.83	4.9.E-04	AP2s, HBs, IAAs, RAPs, WRKY
GO:0010093	P	Specification of floral organ identity	2	1.98	11	0.04	9.3.E-04	AP2s

A similarity-based functional annotation of tomato predicted genes was conducted using the Arabidopsis protein database with gene ontology (GO) information provided by the Arabidopsis Information Resource (TAIR), as previously reported.⁶ Shown are the subcategories in which the ratio of classified genes was significantly (p < 0.001 by Fisher's exact test) higher in the sets of direct RIN target genes than that in the tomato genome. Gene symbols and products: ACO6, 1-aminocyclopropane-1-carboxylic acid (ACC) oxidase 6; ACS2, ACC synthase 2; ACS4, ACC synthase 4; AP2, APETALA2 homolog; AP2a, APETALA2a; C4H, cinnamate 4-hydroxylase; 4CL, 4-coumarate:coenzyme A ligase; CNR, COLORLESS NON RIPENING; E8, ripening-associated ACO homolog; GH3, auxin-responsive GH3 family protein; GSH, glutamate-cysteine ligase; HB, homoebox protein; HMT, homocysteine S-methyltransferase; IAA, auxin/indole-3-acetic acid (AUX/IAA) transcriptional regulator family protein; LOX, lipoxygenase; NAP, NAC family transcription factor protein; OSM, osmotin-like protein; RAP, related to AP2/ERF family protein; RIN, RIPENING INHIBITOR; UGT, uridine diphosphate (UDP)-glucosyl transferase; WRKY, WRKY family transcription factor protein. (1) F and P indicate molecular function and biological process, respectively. (2) Tomato genes involved in fruit ripening regulation (AP2a, CNR and RIN) were categorized in “Flower development” because their Arabidopsis homologs are included in this category.

In the negatively regulated subset of direct RIN targets, GO analysis showed a significantly higher frequency in categories associated with TFs (GO:0003700 and GO:0010093) than the frequency in the whole genome (Table 1). This implies that RIN plays a role in irreversibly promoting ripening via the negative regulation of TF genes.

The interplay between RIN and ethylene might play a key role in fruit ripening regulation

In our recent paper,¹⁰ we analyzed the ethylene responsiveness of TF genes directly and positively regulated by RIN, including NON RIPENING (NOR), CNR, TDR4, AP2a and RIN itself. The treatment with the ethylene signaling inhibitor 1-methylcyclopropene (1-MCP) substantially decreased the expression levels of RIN and eight TF genes, including NOR, TDR4 and AP2a, during ripening to 50% or less of the levels in the mock control. This indicated that the upregulation of RIN during ripening is affected by ethylene levels.

Based on our results, in the recent paper we proposed a model involving cooperative upregulation of the levels of RIN and ethylene during ripening.¹⁰ In this model: (1) at the onset of fruit ripening, RIN is initially induced by an unknown developmental factor; (2) RIN induces ethylene production by upregulating genes for ethylene synthetic enzymes such as 1-aminocyclopropane-1-carboxylic acid (ACC) synthase (ACS2 and ACS4) and ACC oxidase (E8 and ACO6 directly, and ACO1 indirectly);^5,8,11 (3) RIN-induced and subsequent autocatalytic ethylene stimulates ethylene signaling pathways; (4) these signaling pathways, in turn, enhance RIN expression in a positive feedback loop. The expression of RIN is also enhanced by its direct autoregulation.⁵ Thus, our model suggests that the full expression of RIN requires the ethylene-mediated feedback pathway in addition to the developmental controls. The increased RIN and ethylene might induce up- or downregulation of numerous ripening-related genes promptly at the early ripening stage, leading to synchronized initiation of ripening processes. Moreover, in our model, RIN modulates ethylene levels via the upregulation of AP2a, a negative regulator of ethylene production during ripening.^12,13 This RIN-AP2a pathway might contribute to maintaining the ripening stage by regulating ethylene levels. In RIN and other RIN-targeted TF genes such as NOR and TDR4, ethylene was required for full induction of expression; thus it is possible that unknown ethylene-inducible factors may be involved in the induction of these genes. Identification of these unknown ethylene-inducible factors will provide a clue as to how the ripening-specific upregulation of RIN and other TFs is controlled.

Abbreviations:
ACC	=	1-aminocyclopropane-1-carboxylic acid
CArG box	=	C-A/T-rich-G box
ChIP-chip	=	chromatin immunoprecipitation coupled with DNA microarray
GO	=	gene ontology
1-MCP	=	1-methylcyclopropene
MIPS	=	the Munich Information Center for Protein Sequences
TF	=	transcription factor
RIN	=	RIPENING INHIBITOR
ChIP	=	chromatin immunoprecipitation

Acknowledgment

This work was supported in part by the Program for Promotion of Basic and Applied Researches for Innovations in the Bio-oriented Technology Research Advancement Institution (BRAIN) of Japan to Y.I.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.