Comparative Analysis of the MADS-Box Genes Revealed Their Potential Functions for Flower and Fruit Development in Longan (Dimocarpus longan)

Abstract

Longan (Dimocarpus longan Lour.) is an important economic crop widely planted in tropical and subtropical regions, and flower and fruit development play decisive effects on the longan yield and fruit quality formation. MCM1, AGAMOUS, DEFICIENS, Serum Response Factor (MADS)-box transcription factor family plays important roles for the flowering time, floral organ identity, and fruit development in plants. However, there is no systematic information of MADS-box family in longan. In this study, 114 MADS-box genes were identified from the longan genome, phylogenetic analysis divided them into type I (Mα, Mβ, Mγ) and type II (MIKC*, MIKC ^C ) groups, and MIKC ^C genes were further clustered into 12 subfamilies. Comparative genomic analysis of 12 representative plant species revealed the conservation of type II in Sapindaceae and analysis of cis-elements revealed that Dof transcription factors might directly regulate the MIKC ^C genes. An ABCDE model was proposed for longan based on the phylogenetic analysis and expression patterns of MADS-box genes. Transcriptome analysis revealed that MIKC ^C genes showed wide expression spectrums, particularly in reproductive organs. From 35 days after KClO₃ treatment, 11 MIKC genes were up-regulated, suggesting a crucial role in off-season flower induction, while DlFLC, DlSOC1, DlSVP, and DlSVP-LIKE may act as the inhibitors. The gene expression patterns of longan fruit development indicated that DlSTK, DlSEP1/2, and DlMADS53 could be involved in fruit growth and ripening. This paper carried out the whole genome identification and analysis of the longan MADS-box family for the first time, which provides new insights for further understanding its function in flowers and fruit.

Keywords: ABCDE model; KClO3; MADS-box; flower; fruit development; longan.

FIGURE 1

The MADS-box gene family in 15 species. The evolutionary relationship of 15 species on the left of the figure was obtained from NCBI (https://www.ncbi.nlm.nih.gov/Taxonomy/CommonTree/wwwcmt.cgi). The right of the figure shows the number detail of the MADS-box family in each species. Type I MADS-box genes contain three groups: Mα, Mβ, Mγ, and type II genes are subdivided into MIKC^C and MIKC* groups.

FIGURE 2

Phylogenetic analysis of type II MADS-box transcription factors. The maximum-likelihood (ML) tree was constructed with MIKC^C proteins sequences from Arabidopsis (At), longan (Dl), litchi (LITCHI), and yellowhorn (EVM).

FIGURE 3

Chromosomal location and synteny analysis of the MADS-box genes. (A) A total of 114 MADS-box genes are located in 15 chromosomes with different colors. Gene pairs of WGD or segmental duplication are linked using blue lines. Tandem duplication genes are marked by red stars. Gene labels of five types: Mα, Mβ, Mγ, MIKC*, and MIKC^C are denoted by blue, green, yellow, red, and black, respectively. (B) Synteny relationship of type I MADS-box genes from longan, litchi, and yellowhorn genomes. Gray lines in the background indicate the collinear blocks within two genomes. Red lines highlight the syntenic type I MADS-box gene pairs.

FIGURE 4

Conserved motif compositions of longan type II MADS-box proteins. The neighbor-joining tree was constructed with the aligned protein sequences of longan type II MADS-box genes. 10 motifs were identified and displayed in different colors.

FIGURE 5

Prediction of cis-elements and transcription factors of DlMADS. (A) The number of cis-elements related to light-responsive, defense and stress-responsive, circadian, hormone-responses and meristem. The upstream 1.5 kb sequences of all longan MADS-box genes were analyzed through PlantCARE. (B) The cis-motifs and transcription factors potentially target 36 MIKC^C genes. Red indicates the transcription factor binding sites were found in promoter motifs of MIKC^C genes.

FIGURE 6

The deduced ABCDE model in longan. (A) The heat map shows the number of ABCDE classes genes in Amborella, Arabidopsis, rice, and longan. (B) Mature male (left) and female (right) flowers of longan. Five types of floral organs including sepal (S), petal (P), stamen (St), carpel (C), ovary (O) were collected for gene expression analysis. (C) Expression analysis of ABCDE classes genes in floral organs in (B) by using qRT-PCR. (D) The ABCDE model in longan is proposed based on the gene expression levels (bar heights) from (C).

FIGURE 7

The expressional pattern of MIKC^C genes in longan. (A) Expression heat maps of MIKC^C genes in nine tissues (root, stem, leaf, dormant bud, flower bud, flower, fruit, pericarp, and seed) of “SJ” longan Expression values were normalized by log₂(FPKM+1). (B) Detection of MADS-box genes in longan based on qRT-PCR. Six MADS-box genes were selected to detect the expression level in leaf (L), dormant bud (DB), floral bud (FB), flower (F), and pulp (F) of “XC” longan. Data were normalized to Actin1 and Actin2 genes, and the vertical bars indicate standard deviation.

FIGURE 8

Transcript analysis of longan MADS-box genes in off-season flower induction. (A) Expression profile of differentially expressed MADS-box genes during 10 stages (day 0∼54) under CK and KClO₃ treatment. (B) Co-expression networks of differential expression MADS-box genes during off-season flower induction. Each node represents a gene and the lines between nodes represent co-expression correlations.

FIGURE 9

The expression pattern of differentially expressed DlMADS during the fruit development stage. (A) Heat maps based on FPKM for differentially expressed DlMADS at three fruit developmental stages (80∼140 days after flowering). (B) Detection of five differentially expressed DlMADS at three selected stages by using qRT-PCR. The bar and line graphs are derived from RNA-seq and qRT-PCR data, respectively.