Distribution of ACBP family genes in plant kingdom
To identify the distribution of ACBP proteins in plant kingdom, the ACBPs sequences of Arabidopsis were used as probes to search the Phytozome database, and the numbers of genes encoding ACBP proteins in 22 representative species were obtained (Fig. 1). Although ACBP proteins are found in all plant kingdom with the evolution of species, the number and classification of ACBPs have likely undergone extensive changes (Fig. 1). In Chlorophyta before the evolution of land plants, Ostreococcus lucimarinus and Chlamydomonas reinhardtii have four and three ACBP genes, respectively (Fig. 1). There is four ACBP genes in S. moellendorffii, which belongs to the lycophyte and is the origin of seed plants. However, the number of ACBP genes are explosively increased to eight in the moss Physcomitrella patens, which evolved into the vascular plants (Fig. 1). Although monocots are higher in evolution than moss, the gene numbers of the ACBP sharply decrease to five in Brachypodium distachyon, six in rice, six in Sorghum bicolorFig. 1). However, there are nine ACBP genes in maize and 11 genes in Triticum aestivumFig. 1). Interestingly, it seems that there is no significant ploidy relationship between the number of ACBP and the size of the species genome, for example, the maize genome is eight-fold larger than that of B. distachyon whereas the number of ACBP genes only increases by four (Fig. 1). Compared with monocot plants, the numberof ACBP genes in eudicot plants varies greatly. For example, there are only three ACBP genes in the Medicago truncatula genome, however, as many as 11 ACBP genes exist in Glycine max (Fi. 1). Similarly, there are only six ACBP genes in Arabidopsis, but up to 12 in Brassica napus (Fig. 1).
Identification and characterization of ACBP genes in maize
A total of nine putative ZmACBP sequences were identified by searching against the Maize Genome Database (https://www.maizegdb.org),, and designated as ZmACBP1 to ZmACBP9, respectively. Table 1 lists the general characteristics of the ZmACBP gene family including ID, gene name, open reading frame size and exon number of each gene, and length, molecular weight and isoelectric point of each gene coding protein. The number of amino acids (aa) in ZmACBPs varies from 89 aa (ZmACBP1) to 783 aa (ZmACBP9). Similar to A. thaliana, the conserved ACB domain is located at the N-terminal segment in of most ZmACBP proteins, while ZmACBP5 and ZmACBP6 contain the C-terminal ACB domain (Additional file 1: Figure S1a). Nine ZmACBPs were divided into four classes based on the domain structures of the ACBPs (Additional file 1: Figure S1a). Class I containing only one ACB domain had two members, ZmACBP1 and ZmACBP2 (Additional file 1: Figure S1a). ZmACBP3 and ZmACBP4 possessed an ACB domain and ankyrin repeats, and belonged to Class II. ZmACBP5 and ZmACBP6 were in Class III, which is a large ACBP class (Additional file 1: Figure S1a). ZmACBP7, ZmACBP8, and ZmACBP9 possessed an ACB domain and a kelch motif, respectively, and belonged to class IV (Additional file 1: Figure S1a). Nine ZmACBP genes were unevenly distributed on five different chromosomes based on their data of the gene locus (Table 1, Additional file 1: Figure S1b). For example, three ZmACBP genes are localized on chromosome 1, two ZmACBP genes are distributed on chromosomes 5 or 10, whereas only one ZmACBP gene is located on chromosomes 2 and 9, respectively (Additional file 1: Figure S1b).
Phylogenetic relationship, gene structure and conserved motifs of Zm ACBP genes in maize
To analysis of the evolutionary relationships among different species, the sequences of nine ZmACBPs, six AtACBPs, six OsACBPs, 11 TaACBPs, five BdACBPs, and 11 GlACBPs were downloaded for alignment and used to construct an unrooted phylogenetic tree using the Neighbor-Joining method. A total of 48 ACBP protein sequences, including nine ZmACBPs, could be classified into four ACBP classes (Fig. 2). The result of phylogenetic tree clearly shown that every ZmACBP was clustered with those of other closely related species, such as rice，B. distachyon, and T. aestivum, but were further from those of A. thaliana, a distantly related species (Fig. 2).
Analysis of the exon/intron structures of the ZmACBP genes revealed that the number of exons differed among members of the ZmACBP genes family, mostly ranging from three to six, whereas ZmACBP7/–8/–9 had the greatest number of exons (up to 18 exons, Fig. 2). It is interesting to find a similar exon/intron pattern in each group, for instance, most of the ACBPs in Clade I containing four exons; the number of ACBPs with seven exons accounted the majority in Clade II or Clade III; and the vast majority of ACBPs in the Clade IV have 18 exons (Fig. 2). In addition, a total of 10 conserved motifs were detected from 48 ACBPs, and all of the ACBP proteins have two different conserved motifs, motif 1 and 2, which were ACB domain (Fig. 2). The ACBPs in class II have two conserved motifs (10 and 8) in C-terminal, that are ankyrin repeats, and the class IV ACBPs are multi-motifs proteins containing C-terminal kelch domain (Fig. 2). Interestingly, the ACBP genes clustered in the same clade have different gene structures and domain architectures, which providing some clues for the functional investigation of maize ZmACBP genes.
Cis-regulatory elements in the promoters of ZmACBP genes
To further explore the function and the regulation patternsof ZmACBP genes, the cis-acting regulatory elements at the regions of 1,500 bp upstream from the initiation codons of nine ZmACBP genomic sequences were searched in the Plant CARE database. The results indicated that many conserved cis-regulatory elements resided within the ZmACBP promoters, and the number and distribution of the cis-regulatory elements were very different among different ZmACBPs (Fig. 3, Additional file 2: Table S1). The cis-acting regulatory elements could be divided into three important physiological processes: development-related, hormonal responses, and environment responses. The CAAT-box motif, involved in meristem expression element, was found in the promoters region of all nine ZmACBP, suggesting that ZmACBP genes play important roles in differentiation. What’s more, cis-elements involved in hormone responsiveness, including ABRE, AuxRR-core, CGTCA-motif, GARE motif, and TGA-element, resided within most ZmACBP gene promoters (Fig. 3), indicating that ZmACBP genes are involved in growth and development. In addition, some cis-elements involving in environment responses, such as G-Box, LTR, and TC-rich repeats, were only found in a few of ZmACBP gene promoters. For example, the LTR-motif, a cis-element involving in low-temperature responsiveness, could only be detected in ZmACBP4 and ZmACBP5 gene promoters (Fig. 3), while the TC-rich repeats involving defense and stress responsiveness were found in the promoters of ZmACBP1 and ZmACBP9 gene (Fig. 3).
Subcellular localization of ZmACBP proteins
To determine the location of ZmACBPs, eight ZmACBPs were selected for subcellular localization analysis by fusing to the N-terminus of GFP and transforming into N. benthamiana transiently. Notably, the fluorescence of the vector control was observed in the localization of nuclear and cytoplasmic (Fig. 4a). The GFP fluorescent signals of ZmACBP1::GFP and ZmACBP2::GFP were found in the cytoplasmic (Fig. 4b and c), indicating that they are located to cytosol. In N. benthamiana leaf epidermal cells, the GFP fluorescent signals of ZmACBP3::GFP, ZmACBP4::GFP, and ZmACBP5::GFP were distributed at the ER surrounding the nuclei (Fig. 4d, e, and f). To confirm above results, ZmACBP3::GFP, ZmACBP4::GFP, and ZmACBP5::GFP were co-transfected into N. benthamiana leaf epidermal cells with ER marker ZmBiP-RFP, and the signals of GFP (Fig. 4d, e, and f) and RFP (Fig. 4v1, w1, and x1) were overlapped (Fig. 4v, w, and x), indicating that ZmACBP3::GFP, ZmACBP4::GFP, and ZmACBP5::GFP were ER-associated proteins. The distribution of ZmACBP6::GFP, ZmACBP7::GFP, and ZmACBP8::GFP were similar to the GFP-vector control (Fig. 4g, h, and i), while disappearing in nuclear localization, indicating that they are targeted to the cytosol and plasma membrane.
Expression profiles of ZmACBP in different organs and developmental stages
To analyze the expression profiles of nine ZmACBP genes during maize development, their transcriptional patterns were detected using published microarray data (Maize eFP Browser, http://bar.utoronto.ca/efp_maize/cgi-bin/efpWeb.cgi) in 60 different tissues or developmental stages in maize. Based on the expression patterns of nine ZmACBPs in this database were represented as heatmaps (Fig. 5a). Some of these genes are preferentially expressed in particular organs. For example, four genes (ZmACBP2, 6, 5 and 8) show the highest expression levels in leaf compared with those in other tissues, three genes (ZmACBP3, 4 and 7) with the highest mRNA levels in endosperm, and ZmACBP1 and ZmACBP9 with the highest expression levels in embryo and husk, respectively (Fig. 5a).
The expression levels of the nine ZmACBP mRNAs were detected in different tissues by using qPCR, including root, stem, leaf, silk, immature cob, anther, kernel (10 DAP) (Fig. 5b). The expressions of nine ZmACBP transcripts were all detected in all different tissues, but the patterns were dramatically different. ZmACBP1, ZmACBP2, and ZmACBP5 were more highly expressed in the leaf than in other organs or tissues (Fig. 5b). ZmACBP3 was more highly expressed in kernel, and had almost the same expression level in others organs (Fig. 5b). ZmACBP4 was relatively highly expressed in root, moderately expressed in 10 DAP of kernel and showed low expression in leaf (Fig. 5b). ZmACBP6 had the highest expression level in cob (Fig. 5b). ZmACBP7 and ZmACBP8 had similarly expression patterns with relatively high expressions in kernel (Fig. 5b). ZmACBP9 was relatively highly expressed in silk and 10 DAP of kernel, followed by stem, and showed low expression in leaf (Fig. 5b). The qRT-PCR results were consistent with the microarray data, suggesting that different ZmACBP might play different roles in the development of different organs.
Rigional association mapping for ZmACBP genes
To explore the potential functions of the ZmACBP genes for agronomic traits in maize, an association mapping panel from different maize inbred lines was applied to explore correlation between ZmACBP genes and maize agronomic traits. A total of eight ZmACBP genes were found to be significantly correlated with more than one agronomic traits at the P ≤ 0.05 level, such as plant height (PH), ear height (EH), ear diameter (ED), 100 grain weight (HGW), silking time (ST), heading date (HD), kernel width (KW), and so on (Additional file 3: Table S2). What’s more, some ZmACBP genes were found to associate with some important agronomic traits at the P ≤ 0.01 level, such as grain yield, flowering time, and kernel-related traits. For example, ZmACBP3 was significantly correlated with EH (Fig. 6a), ZmACBP4 significantly affected on KW (Fig. 6b), and ZmACBP7 significantly effected on HD (Fig. 6c). These results suggested that these ZmACBP genes in different maize inbred lines could be regarded as important candidate genes related to maize development and kernel-related traits for subsequent functional validation in maize.
Expression profiling of ZmACBP genes in response to various abiotic and biotic stress
To gain further insight into the role of ZmACBPs in response to abiotic or biotic stress, the expression patterns of ZmACBPs were analyzed using qRT-PCR under different treatments, including 200 mM NaCl, 20% PEG6000, wounding, 4°C，20 μmol/L Cu2+ and fungus infection (Fig. 7). The results shown that ZmACBPs genes were induced by high salinity or osmotic treatments, peaking at 6 or 12 h after two treatments and then remained at relatively high levels (Fig. 7a and b). ZmACBP1, ZmACBP2, ZmACBP5 and ZmACBP6 were rapidly induced and peaked at 0.5 h after wounding treatment, following by a decrease in expression to levels lower than untreated, however, the expression of ZmACBP3, ZmACBP4, ZmACBP7, ZmACBP8, and ZmACBP9 were suppressed (Fig. 7c). ZmACBP1, ZmACBP4, ZmACBP7, ZmACBP8 and ZmACBP9 mRNAs were reduced after cold treatment and then recovered to minimum level in 12 or 24 h (Fig. 7d). While, ZmACBP2, ZmACBP3, ZmACBP5 and ZmACBP6 mRNAs were induced with cold treatment, and peaked within 12 h, then remained at relatively high levels at 24 h (Fig. 7d). In copper ion treatments, ZmACBP1 and ZmACBP8 mRNAs were suppressed within 48 h, and ZmACBP3 mRNAs was almost unaffected (Fig. 7e). ZmACBP4, ZmACBP5, ZmACBP6, ZmACBP7 and ZmACBP9 were rapidly induced by copper ion treatments, and peaked at 3 or 6 h after salt treatment, then remained at relatively high levels at 48 h (Fig. 7e). Fungus infection (Ustilago maydis) induced the expression of ZmACBP1, ZmACBP2, ZmACBP3, ZmACBP5, ZmACBP6 and ZmACBP8, while the expression of ZmACBP4, ZmACBP7 and ZmACBP8 were suppressed (Fig. 7f). These results suggest ZmACBPs play important roles in stress defense responses.
The constitutive expression in A. thaliana of two ZmACBPs had a positive effect on stress tolerance
To confirm ZmACBPs response to stress, ZmACBP1 and ZmACBP3 were selected for further overexpression (OE) in A. thaliana. Given the inducibility of ZmACBP1 and ZmACBP3 by salinity and osmotic (Fig. 7a and b), the OEs were exposed to 100 mM NaCl or 200 mM mannitol. There was no observable phenotypic effect between OE and VC plants with respect to either leaf or root growth on a medium containing no abiotic stress agent (Fig. 8a and e), however, in the presence of 100 mM NaCl, the OE lines heterogeneously expressing ZmACBP1 or ZmACBP3 grew more vigorously than VC (Fig. 8b and f), forming larger leaves and longer roots (Fig. 8d and h). In addition, the OE and VC lines performed similarly when challenged with 200 mM mannitol stress (Fig. 8c and d).