*4.1. Plant Material and Growth Condition*

In this article, tomatoes (*Solanum lycopersicum*, 'Ailsa Craig' AC++) from Laboratory of molecular biology of tomato, Bioengineering College, Chongqing University, Chongqing, China, were grown in controlled greenhouse conditions of a 16-h day (25 ◦C)/8-h night (18 ◦C) cycle, 80% humidity, a 250-µmol·m−<sup>2</sup> ·s −1 light intensity, and were managed routinely. The tomato flowers were tagged at anthesis and floral organs: sepals (Se), petals (Pe), stamens (St), and carpels (Ca) were collected from flowers at anthesis. The fruit color and days post-anthesis (DPA) were used to differentiate the ripening days of tomato fruits. We defined 20 DPA as the immature green (IMG), and 35 DPA as the mature green (MG), at which point the fruits are green and shiny and no obvious color change is observed. The 38-DPA tomato fruits with color change from green to yellow were characterized as breaker (B) fruits. Besides, the samples of B+4 (4 days after breaker) fruits and B+7 (7 days after breaker) fruits were also used in our study. All the samples that we used were frozen in liquid nitrogen immediately and stored at –80 ◦C.

## *4.2. Identification of MADS-Box Genes in Tomato*

The Sol Genomics Network (SGN, Available online: http://solgenomics.net/) and the National Center for Biotechnology Information (NCBI, Available online: https://www.ncbi.nlm.nih.gov/) database were used to comprehensively identify the whole MADS-box protein sequences of tomato. BLAST searches, using all the Arabidopsis and rice MADS-box protein sequences as queries, were performed to check the predicted tomato MADS-box protein sequences in the database. Subsequently, we further examined all the candidate protein sequences by the PROSITE (Available online: http://www.expasy.org/prosite/) and SMART (Available online: http://smart.embl-heidelberg.de/) databases for reliability. The tomato protein sequences, containing the typical conserved domain of the MADS-box protein family, were selected for amino acid sequence multiple alignment and phylogenetic tree analysis. Then, we obtained their DNA sequences according to their amino acid sequence from the SGN database. Additionally, the molecular weight and isoelectric points of tomato MADS-box proteins were detected by the ExPASy proteomics server.

## *4.3. Phylogenetic Analysis of Tomato MADS-Box Proteins*

Multiple sequence alignment for the two groups of all the 131 tomato MADS-box genes (Table 1) was generated using ClustalX 1.81. The alignment results were used to conduct a phylogenetic tree by the MEGA5.02 program, and the evolutionary history was inferred using the neighbor-joining method. [92].

## *4.4. The Analysis of Gene Structure and Conserved Motif*

The tomato MADS-box coding domain sequences (CDS) and corresponding genomic DNA sequences were collected from SGN and NCBI databases to predict gene structure. The online tool Gene Structure Display Server 2.0 (GSDS 2.0, Available online: http://gsds.cbi.pku.edu.cn/index.php), was used to construct an exon/intron map [93].

Conserved motifs of the tomato MADS-box protein sequences were identified by online software MEME Version 4.12.0 (Available online: http://meme-suite.org/tools/meme). It was performed with the following parameters: 10 different motifs, a motif width of 6–200 amino acids, and any number of repetitions. The SMART database was used to annotate the MEME motifs.
