A MADS-Box Gene Necessary for Fruit Ripening at the Tomato Ripening-Inhibitor (Rin) Locus

See allHide authors and affiliations

Science  12 Apr 2002:
Vol. 296, Issue 5566, pp. 343-346
DOI: 10.1126/science.1068181


Tomato plants harboring theripening-inhibitor (rin) mutation yield fruits that fail to ripen. Additionally, rin plants display enlarged sepals and loss of inflorescence determinacy. Positional cloning of the rin locus revealed two tandem MADS-box genes (LeMADS-RIN andLeMADS-MC), whose expression patterns suggested roles in fruit ripening and sepal development, respectively. Therin mutation alters expression of both genes. Gene repression and mutant complementation demonstrate thatLeMADS-RIN regulates ripening, whereasLeMADS-MC affects sepal development and inflorescence determinacy. LeMADS-RINdemonstrates an agriculturally important function of plant MADS-box genes and provides molecular insight into nonhormonal (developmental) regulation of ripening.

The maturation and ripening of fleshy fruits is a developmental process unique to plants and affects the quality and nutritional content of a significant portion of the human diet. Although specific fruit-ripening characteristics vary among species, ripening can be generally described as the coordinated manifestation of changes in color, texture, flavor, aroma, and nutritional characteristics that render fruit attractive to organisms receiving sustenance in exchange for assisting in seed dispersal (1, 2).

Fruit species are classically defined as one of two ripening types, climacteric and nonclimacteric, where the former display a burst in respiration at the onset of ripening, in contrast to the latter. Climacteric fruit typically increase biosynthesis of the gaseous hormone ethylene, which is required for the ripening of fruit such as tomatos, bananas, apples, pears, and most stone fruit. Nonclimacteric fruit, including strawberries, grapes, and citrus fruits, do not require climacteric respiration or increased ethylene for maturation. Molecular ripening research has focused primarily on ethylene, but little is known of control before ethylene induction, nor of common regulatory mechanisms shared by climacteric and nonclimacteric species (3).

The tomato is a model for analysis of ripening due originally to its significance as a food source and diverse germplasm, and more recently, the availability of molecular tools (4) and efficient transformation (5). A number of tomato-ripening mutants have been useful for research and breeding (3). Especially interesting is the recessiveripening-inhibitor (rin) mutation that inhibits all measured ripening phenomena, including the respiratory climacteric and associated ethylene evolution, pro-vitamin A carotenoid accumulation, softening, and production of flavor compounds (6). The rin mutant exhibits ethylene sensitivity, including the seedling triple response (7), floral abscission, and petal and leaf senescence. Nevertheless, rin fruit do not ripen in response to exogenous ethylene, yet they display induction of at least some ethylene-responsive genes, indicating retention of fruit ethylene sensitivity (8). We interpret these results to mean that theRIN gene encodes a genetic regulatory component necessary to trigger climacteric respiration and ripening-related ethylene biosynthesis in addition to requisite factors whose regulation is outside the sphere of ethylene influence. As such, RIN acts upstream of both ethylene and nonethylene-mediated ripening control. We previously reported mapping of the rin locus to tomato chromosome 5 (9). Here, we report cloning and characterization of two MADS-box genes at the rin locus. MADS-box genes encode transcription factors that, in plants, primarily regulate floral development (10, 11), yet have never before been demonstrated to regulate ripening.

The only reported mutation at the rin locus arose spontaneously in a breeding line developed by H. Munger (at Cornell University) (12). In addition to ripening inhibition, the mutant exhibits large sepals and a loss of inflorescence determinacy (Fig. 1, A to C). Genetic complementation of only the ripening phenotype of the original rin/rin mutant, with a recessive large sepal mutant (macrocalyx, mc), suggested that the lesion at rin affected two adjacent loci controlling ripening and sepal development, respectively (12).

Figure 1

Normal, mutant, and transgenic lines. (A) Developmental series of normal (cultivar Ailsa CraigRin/Rin, top) and mutant (nearly isogenic for rin/rin, bottom) fruit. The fruit shown on the right were mature green when exposed to 10 parts per million (ppm) ethylene for 12 hours and photographed 48 hours later. (B) Rin/Rindeterminate inflorescence. (C)rin/rin indeterminate inflorescence. (D) Fruit and sepals of T1 progeny from five independent transformants (rin/rin) with theLeMADS-RIN sense cDNA. The progeny shown at the top (complemented) contain the transgene, whereas those shown below have segregated it out. (E) Fruit and sepals of T1 progeny from two transformants of Rin/Rin with the LeMADS-RIN antisense cDNA. The fruit shown in the upper panel demonstrate ripening inhibition (+ transgene), whereas those on the bottom have segregated it out. (F) Normal inflorescence of a representative Rin/Rin line transformed with the LeMADS-MC antisense cDNA. A T1 that has segregated out the transgene is shown. (G) Indeterminate inflorescence and large sepals of a sibling of 1F harboring the transgene.

Previous mapping indicated that rin resides on a tomato 365-kb yeast artificial chromosome (YAC) clone, [Yrin9 (9)], and the analysis summarized in Fig. 2 confirmed this hypothesis (13). Yrin9 was used as a hybridization probe to identify putative rin cDNAs (13). Gene-expression analysis revealed two Yrin9-derived cDNA clones with altered transcripts in mutant fruit (Fig. 3A). C34 showed constitutive expression during maturation of normal fruits, whereas C43 was induced coincident with ripening. Both sequences hybridized to mRNA that was identically expressed in maturing rin fruit. This result, in combination with increased transcript size for both probes in rin, suggested that the rin lesion resulted in the fusion of gene sequences represented by C34 and C43. Reverse transcription–polymerase chain reaction (RT-PCR) ofRin/Rin and rin/rin fruit RNA (14) confirmed this hypothesis (Fig. 3B). Sequencing of both cDNAs, in addition to the rin RT-PCR product (15), verified that a deletion occurred in the mutant, yielding a chimeric RNA (Fig. 3C). PCR ofRin/Rin genomic DNA (16) indicates that 2.6 kb separates the C43 and C34 coding regions (Fig. 3C).

Figure 2

Genetic map of the rin region of chromosome 5. Mapping of the YAC end (Yrin9L) and a subclone of Yrin9 (RS173) provided genetic evidence that Yrin9 containsrin.

Figure 3

Characterization ofLeMADS-RIN and LeMADS-MC. (A) RNA gel-blot hybridizations with 3′–untranslated region probes of LeMADS-RIN (C43) andLeMADS-MC (C34). Total RNA was isolated from normal (Rin/Rin) and mutant (rin/rin, identical age post-pollination as normal) fruit pericarp tissue. Mature green (M), M after 12 hours of exposure to 10 ppm ethylene (E), “breaker” showing initial external signs of lycopene (red pigment) accumulation (B), and ripe [B + 7 days (R)]. LeMADS-MC was additionally probed to RNA from normal sepals (S), petals (Pe), stamen (St), styles/stigma (SS), and carpels (C) of anthesis flowers. rS denotesrin/rin sepals. A tomato ribosomal (18S) probe confirmed equal RNA loading. (B) Chimeric MADS-box transcript in the rin mutant. RT-PCR of total RNA from R-stage Rin/Rin andrin/rin fruit pericarp using primers toLeMADS-RIN (RIN), LeMADS-MC(MC), or both (RIN/MC) (14). (C) Schematic of MADS-box genes at the rin locus in addition to RNAs defined in (B). The arrows above the diagram indicate primers used in (B) (14), whereas those below it were used to estimate intergenic distance (16). C43F and C34R were used to amplifyrin/rin RNA. The gray boxes represent nontranscribed regions. The black and crosshatched boxes correspond to the LeRIN-MADS andLeMADS-MC transcribed regions, respectively. The white boxes indicate MADS domains. The right-angle arrows denote transcription start sites.

C34 and C43 sequencing revealed that both are members of the MADS-box family of transcription factors (15). MADS-box genes encode proteins characterized by the conserved MADS-box DNA binding domain and have been isolated from numerous eukaryotic organisms (17). Plant MADS-box genes are primarily associated with the regulation of floral development (10, 11, 18). The best described plant MADS-box family is from Arabidopsis, with at least 47 genes identified, although many remain to be defined functionally (19).

We named the genes corresponding to C43 and C34,LeMADS-RIN and LeMADS-MC, respectively, based on analyses described herein. To define the activities of each gene, sense and antisense constructs of full-length cDNAs under the direction of the CaMV35 promoter were created (13). Transgene integration into theRin/Rin and rin/rin genomes was performed by Agrobacterium-mediated transferred DNA (T-DNA) transformation (5), and 10 to 50 independent transformants were recovered for each construct.

The expression of LeMADS-RIN in therin/rin genotype resulted in complementation of the ripening deficiency but did not affect the sepal or indeterminacy phenotypes (Fig. 1D). This result demonstrates thatLeMADS-RIN encodes theRIPENING-INHIBITOR gene and, furthermore, the function of this gene is specific to ripening control. Although homeotic conversion of tomato sepals to carpeloid structures by ectopic expression of the tomatoAGAMOUS1 gene (TAG1) demonstrated the obvious link between tomato carpel development and eventual ripening (20), elucidation of the rin mutant establishes that a MADS-box gene directly regulates ripening. Consistent with the complementation results, antisense repression ofLeMADS-RIN in normal (Rin/Rin) tomatoes resulted in a phenotypic mimic of the ripening aspect of rin with no effect on sepals or determinacy (Fig. 1E).

Antisense of LeMADS-MC resulted in indeterminate inflorescences with large sepals and normally ripening fruit (Fig. 1, F and G). Together, these results confirm thatLeMADS-MC affects tomato inflorescence determinacy and sepal development, whereasLeMADS-RIN is a previously undescribed regulator of ripening.

Phylogenetic analysis of LeMADS-MC revealed substantial similarity to the Arabidopsis APETALA1 (AP1) protein in addition to putative AP1 orthologs from additional species (Fig. 4). AP1 is a class A MADS-box gene whose expression is restricted to the outer whorls (sepals and petals) of flowers, and gene knockouts of AP1 lead to the homeotic conversion of sepals to leaf or bractlike organs, loss of petals in most flowers, and inflorescence indeterminacy (21).LeMADS-MC is expressed in sepals, petals, and carpels (Fig. 3A) more similar to the SQUAMOSA(SQUA) AP1 ortholog from Antirrhinum(22). LeMADS-MC expression is absent in rin mutant sepals, presumably because the chimeric gene is directed by the LeMADS-RIN promoter. Our strongest repression of LeMADS-MC causes homeotic conversion from sepals to leaflike structures (Fig. 1G), although no effect on petals was observed. This phenotype contrasts the loss of petals in ap1, yet is similar to the phenotype of thesquamosa mutant of Antirrhinum. In summary, phylogeny, gene expression, and both the rin mutant and antisense of LeMADS-MC suggest thatLeMADS-MC, SQUA, and AP1are orthologs.

Figure 4

Relation of RIN and MC to other plant MADS-box proteins. Maximum parsimony map of LeMADS-RIN, LeMADS-MC,Arabidopsis (AP1, SEP1, AGL3, AGL8, PI, AGL29),Antirrhinum (SQUA), petunia (FBP4), pepper (MADS1), tobacco (NAP1-2), and strawberry (FvMADS-9) MADS-box sequences (13).  

LeMADS-RIN phylogenetic analysis revealed the greatest similarity to FPB4 of petunia and a pathogen-infected pepper fruit EST, MADS1 (Fig. 4). The most similarArabidopsis genes were SEP1 and AGL3. SEP1 is required for manifestation of B and C class organ-identity functions (petals, stamen, and carpel development) and distinct fromLeMADS-RIN, where only the ripening aspect of carpel development is affected (23). The exact function of AGL3 remains uncertain, although expression has been observed in all aerial tissues (24). Functional characterization of AGL3 and continued isolation and analysis of tomato MADS-box genes should elucidate these relations. Gene expression analysis indicates that LeMADS-RIN is expressed primarily in fruits, and its induction coincides with ripening. Unlike many previously characterized tomato fruit-ripening genes,LeMADS-RIN expression is not significantly influenced by ethylene (Fig. 3A).

The rin mutation reveals a function for plant MADS-box genes as regulators of fruit ripening. LeMADS-RIN is required to initiate climacteric respiration and associated ethylene biosynthesis in addition to ripening factors that cannot be complemented by supplemental ethylene. Consequently, LeMADS-RIN is upstream of ethylene in the regulatory cascade and may represent a global developmental regulator of ripening potentially shared among climacteric and nonclimacteric species. In support of this hypothesis, we have isolated a cDNA (FvMADS-9) from strawberries (a nonclimacteric fruit), usingLeMADS-RIN as a probe (25). FvMADS-9 displays fruit-specific expression (25) and clusters close to LeMADS-RIN in phylogenetic analysis (Fig. 4).

LeMADS-RIN is closely related to sequences derived from petunias and peppers (Fig. 4) but for which no functional characterization has been reported. The near identity of these sequences from additional members of the Solanaceae suggests potential involvement in ripening. Interestingly, unlike tomatos and peppers, petunia fruit are not fleshy and undergo a maturation process ending in senescence, dehydration, and dehiscence similar to that of Arabidopsis siliques. Functional analysis of FBP4 could lead to additional insights regarding maturation of different fruit types.

Ripe fruits serve as a significant portion of the human diet, directly affecting human nutrition and health.LeMADS-RIN represents a molecular bridge between the extensively studied phenomena of floral development and fruit ripening/ethylene response with regard to the cascade of ethylene-regulated events associated with climacteric ripening being dependent on a member of the floral development–associated MADS-box family. This discovery opens a new research frontier in fruit ripening. For example, because MADS-box genes are known to act as multimers (26), one could logically predict that additional MADS-box genes might affect ripening.

From a practical perspective, the rin mutation is widely used in tomato hybrid cultivars to yield fruit with a long shelf life and acceptable quality. Tomatoes heterozygous for therin allele remain firm and ripen over a protracted period (presumably due to reduced levels of functional RIN protein) permitting industrial-scale handling and expanded delivery and storage opportunities. LeMADS-RIN is a rare example of a gene whose effects are documented a priori, suggesting excellent potential for practical genetic modification of fruit ripening and quality characteristics.

  • * To whom correspondence should be addressed. E-mail: jjg33{at}


View Abstract

Navigate This Article