Divergent Roles of the Auxin Response Factors in Lemongrass (Cymbopogon flexuosus (Nees ex Steud.) W. Watson) during Plant Growth

Abstract

Auxin Response Factors (ARFs) make up a plant-specific transcription factor family that mainly couples perception of the phytohormone, auxin, and gene expression programs and plays an important and multi-faceted role during plant growth and development. Lemongrass (Cymbopogon flexuosus) is a representative Cymbopogon species widely used in gardening, beverages, fragrances, traditional medicine, and heavy metal phytoremediation. Biomass yield is an important trait for several agro-economic purposes of lemongrass, such as landscaping, essential oil production, and phytoremediation. Therefore, we performed gene mining of CfARFs and identified 26 and 27 CfARF-encoding genes in each of the haplotype genomes of lemongrass, respectively. Phylogenetic and domain architecture analyses showed that CfARFs can be divided into four groups, among which groups 1, 2, and 3 correspond to activator, repressor, and ETTN-like ARFs, respectively. To identify the CfARFs that may play major roles during the growth of lemongrass plants, RNA-seq was performed on three tissues (leaf, stem, and root) and four developmental stages (3-leaf, 4-leaf, 5-leaf. and mature stages). The expression profiling of CfARFs identified several highly expressed activator and repressor CfARFs and three CfARFs (CfARF3, 18, and 35) with gradually increased levels during leaf growth. Haplotype-resolved transcriptome analysis revealed that biallelic expression dominance is frequent among CfARFs and contributes to their gene expression patterns. In addition, co-expression network analysis identified the modules enriched with CfARFs. By establishing orthologous relationships among CfARFs, sorghum ARFs, and maize ARFs, we showed that CfARFs were mainly expanded by whole-genome duplications, and that the duplicated CfARFs might have been divergent due to differential expression and variations in domains and motifs. Our work provides a detailed catalog of CfARFs in lemongrass, representing a first step toward characterizing CfARF functions, and may be useful in molecular breeding to enhance lemongrass plant growth.

Keywords: ARF transcription factors; Cymbopogon; RNA-seq; expression analysis; gene duplication.

Figure 1

Phylogenetic analysis of the CfARF family. CfARFs, SbARFs, OsARFs, and AtARFs were used in the phylogenetic analysis, with ZmARFs orthologs annotated next to the corresponding SbARF geneIDs in the phylogenetic tree. CfARFs from the lemongrass hap1 and hap2 genomes are indicated with black and grey dots, respectively, on the tree branches. The four phylogenetic groups of CfARFs (i.e., groups 1, 2, 3, and 4) are indicated in the colored outermost circle in blue, orange, gray, and green, respectively. CfARF proteins specifically encoded by the lemongrass hap1 or hap2 genome are indicated with red arrowheads. OsARFs and AtARFs expanded due to ancient WGD events in monocots and dicots, respectively, are labeled with gray brackets. ARF genes expanded due to tandem duplication events (TD) are labeled with light green brackets.

Figure 2

Conservation of the CfARF DNA-binding domain. Amino acid alignment of the DNA-binding domain of CfARFs was performed with representative activator and repressor ARFs (MpARF1 and ZmARF35 as activators, AtARF1 and ZmARF13 as repressors). The phylogenetic clades for CfARFs are indicated in background colors (orange, light blue, gray, and green meaning groups 1, 2, 3, and 4, respectively). B3 DNA-binding domain (DBD) is indicated by the blue box, while flanking dimerization domains 1 and 2 are indicated by the red and gray boxes, respectively. The sequences of flanking dimerization domains 1 and 2 were adopted from those previously described [9,12,40]. Asterisks in the B3 DBD region indicate DNA-contacting residues described for AtARF1 [12]. Asterisks in the flanking dimerization domains indicate residues at the ARF dimer interface [14]. Amino acid residues conserved in the aligned ARFs for either group 1 or group 2 ARFs are highlighted with red background color.

Figure 3

Conservation of the CfARF C-terminal region. Amino acid alignment of the DNA-binding domain of CfARFs was performed with representative activator and repressor ARFs (MpARF1 and ZmARF35 as activators, AtARF1 and ZmARF13 as repressors). The phylogenetic clades for CfARFs are indicated in background colors, as described in Figure 2. The amino acid sequences similar to Aux/IAA motif III or IV are indicated in blue boxes.

Figure 4

Expression patterns of CfARFs in the leaf, stem, and root tissues during the four lemongrass growth stages. The expression patterns of CfARFs belonging to phylogenetic groups 1, 2, 3, and 4 are given for the four stages and three tissues (A, B, C, and D, respectively). To distinguish the CfARFs from groups 1, 2, 3, and 4, there figures are surrounded by orange, blue, grey and green broken lines, respectively. The leaf, stem, and root tissues are indicated in green, blue, and gray, respectively. The four stages are labeled as 1, 2, 3, and 4, and described in the Materials and Methods Section 3.5. “Transcriptomic of the lemongrass and coexpression network analysis”. (E) Quantitative RT-PCR validation of the expression of CfARF3 and CfARF10 at the T3 and T4 stages. qPCR was performed with three biological replicates, with the statistical difference determined by Student’s t-test (*, p < 0.05; **, p < 0.01; ***, p < 0.005). Within each gene and stage, statistical differences are indicated with black asterisks, while statistical differences in the expression levels from the same tissues between the T3 and T4 stages are shown with red asterisks, with “ns” indicating “not significant”. CfActin (geneID: hap1.evm.model.Chr01.4294) was used as the internal reference gene for qPCR with all of the qPCR primers provided in Table S2.

Figure 5

Allelic expression patterns of CfARFs. The allelic expression profiles of each pair of CfARF haplotypes are shown in the heat map, with the colors indicating the Z score of the expression level (in FPKM) and the number on each grid indicating the FPKM expression level. CfARF genes are sorted into phylogenetic groups.

Figure 6

Co-expression network analysis-identified modules with CfARFs and other genes related to auxin signaling and cell growth. (A) The heat map showing the 21 co-expression modules identified with WGCNA. The module names are labeled and indicated with corresponding colors. (B) Representative expression patterns (eigengenes) of the four typical CfARF-containing modules, with the CfARFs labeled for each module. In particular, the green and turquoise modules appear to be enriched with CfARFs, as determined by the hypergeometric test (p < 0.05), while P_{hypergeometric} for the turquoise module was close to the significant level (0.0894). (C) The enriched functional terms (including GO and KEGG) in the turquoise module, with the term no. and q values given. (D) In the brown, green, and turquoise modules, several genes homologous to the maize genes known to function in auxin biosynthesis, signaling, the cell cycle, and cell expansion were identified [44]. The heat map shows the relative expression of these genes in the leaf, stem, and root tissues of lemongrass, with their modules indicated with corresponding colors.