Phylogenetic and gene structural analysis

All the 122 MATE protein sequences in apple and Arabidopsis were used to perform phylogenetic and structural analysis. Multiple sequence alignments were conducted by ClustalW in MEGA X with default parameters and ProteinWeightMatrix = BLOSUM [77]. The alignment result was then used to construct a phylogenetic tree based on the neighbor-joining (NJ) method of MEGA X, with the following setups: Equal input model and partial deletion (60%). For the maximum likelihood (ML) tree, JTT + F + G was used as the best model for ML tree constructing which calculated by ProtTest 3.4.2 [78], the ML tree was constructed by MEGA X. Both NJ and ML trees were conducted 1000 bootstrap replications.

Gene structure analysis was performed using the Gene Structure Display Server (GSDS) [79] with default setting. Motifs in MATE proteins were identified using MEME Suite [25] (version 5.1.1, http://meme-suite.org/index.html) with default setting: site distribution is zero or one occurrence per sequence (zoops), motif width is 5 to 50 and the maximum number of motifs was set at 12. We extract the 1.5 kb upstream region of MdMATE genes and upload to PlantCARE [80] database to detect cis-regulatory elements in the promoter regions.