Abstract
The roles of EHDs in clathrin-mediated endocytosis (CME) in plants are poorly understood. Here, we isolated a maize mutant, designated as ehd1, which showed defects in kernel development and vegetative growth. Positional cloning and transgenic analysis revealed that ehd1 encodes an EHD protein. Internalization of the endocytic tracer FM4-64 was significantly reduced in ehd1 mutant and ZmEHD1 knock-out mutants. We further demonstrated that ZmEHD1 and ZmAP2 σ subunit physically interact in the plasma membranes. Cellular IAA levels were significantly lower in ehd1 mutant than in wild-type maize. Auxin distribution and ZmPIN1a-YFP localization were altered in ehd1 mutant. Exogenous application of 1-NAA but not GA3 rescued the seed germination and seedling emergency phenotypic defects of ehd1 mutants. Taken together, these results indicate that ZmEHD1 regulates auxin homeostasis by mediating CME through its interaction with the ZmAP2 σ subunit, which is crucial for kernel development and vegetative growth of maize.
Introduction
Endocytosis, which can be clathrin-dependent or clathrin-independent, is the internalization of plasma membrane (PM) proteins or uptake of extracellular materials for transport to the endosome (Murphy et al., 2005). Clathrin-mediated endocytosis (CME) is the major gateway into plant cells (Fan et al., 2015). Although detailed information is available on the roles of endocytosis in animals, researchers have only recently documented that CME is critical for various developmental processes in plants, including cell polarity determination (Van et al., 2011; Wang et al., 2013), male reproductive organ development (Blackbourn and Jackson, 1996; Kim et al., 2013), gametophyte development (Backues et al., 2010), and embryogenesis (Fan et al., 2013; Kitakura et al., 2011). CME is also important in plant responses to abiotic and biotic stresses by internalizing of PM-resident transporters or receptors, such as the boron transporter (BOR1) (Takano et al., 2010), the iron-regulated transporter (IRT1) (Barberon et al., 2011), the auxin efflux transporter PIN-FORMED (PINs) (Dhonukshe et al., 2007), the brassinosteroid receptor BRASSINOSTEROID INSENSITIVE 1 (BRI1) (Di et al., 2013), the bacterial flagellin receptor FLAGELIN SENSING 2 (FLS2) (Robatzek et al., 2006), and the ethylene-inducing xylanase receptor (LeEIX2) (Bar and Avni, 2009).
CME is a complex process that can be divided into at least four steps: the budding of a vesicle, the packaging of cargo into the vesicle, the release of the mature vesicle from the PM, and the fusing of the vesicle with the endosome (Sigismund et al., 2012). Because clathrin cannot directly bind to the PM or to cargoes, the formation of clathrin-coated endocytic vesicles initiates the association of adaptor protein complex 2 (AP2) with the PM, and thus AP2 plays a central role in the initiation of CME. The AP2 complex forms a heterotetrameric complex containing two large subunits (β1-5 and one each of γ/α/δ/ε/ζ), a medium subunit (μ1-5), and a small subunit (σ1-5) (McMahon and Boucrot, 2011). Our understanding of AP2 in plants has been enhanced by the following recent findings: knockdown of the two Arabidopsis AP2A genes or overexpression of a dominant-negative version of the medium AP2 subunit, AP2M, was shown to significantly impair BRI1 endocytosis and to enhance brassinosteroid signaling (Di et al., 2013); α and μ adaptins were demonstrated to be crucial for pollen production and viability, as well as elongation of staminal filaments and pollen tubes by modulating the amount and polarity of PINs (Kim et al., 2013); σ adaptin was found to play important roles in the assembly of a functional AP2 complex, and ap2 σ loss-of-function Arabidopsis exhibited defects in multiple aspects of plant growth and development (Fan et al., 2013).
In addition to the classical AP2 complex, other accessory proteins, including the C-terminal Eps15 homology domain (EHD) proteins, also connect the cargo and membrane lipid to form the CME vesicle. The EHD family contains one member in Drosophila and C. elegans and four orthologs in mouse and human. All members contain an EHD with two calcium-binding helix-loop-helix motifs (EF-hands), a P-loop with a predicted ATP/GTP binding site, a dynamin-N motif with a nucleotide-binding domain, and a coil-coil region (Bar et al., 2008). The involvement of EHD-containing proteins in CME was described in detail in mammalian cells, in which RME1 was demonstrated to be a key player controlling the recycling of internalized receptors from the endocytic recycling compartment to the PM (Lin et al., 2001). By homologous cloning, two EHD genes were isolated from Arabidopsis (Bar et al., 2008). The two AtEHD proteins regulated endocytosis in distinct patterns. Over-expression of AtEHD2 had an inhibitory effect on endocytosis, while down-regulation of AtEHD1 caused retardation of entry of endocytosed material into plant cells (Bar et al., 2008). However, the mechanisms by which EHD family members linked to CME in plants have not been characterized.
Maize ranks first in total production among major staple cereals and is also an important raw material for biofuel and many other industrial products (Mclaren, 2005). As an important model system for basic biological research, maize has contributed significantly to our knowledge of plant development and evolution, and this understanding has been used to elucidate the developmental mechanisms of maize kernel morphogenesis. The earliest genetic studies performed in maize included analyses of defective kernel (dek) mutants. Although several hundred dek mutants have been isolated, relatively few corresponding genes have been molecularly cloned because embryo-lethal alleles are difficult to study (Scanlon and Takacs, 2009).
To better understand maize endosperm filling and maturation, we characterized a maize mutant ehd1 that is impaired in kernel development and vegetative growth. Positional cloning of ehd1 and transgenic analysis identified the causative locus as a gene that encodes an EHD-containing protein and that is herein designated as ZmEHDl. We show that ZmEHD1 is involved in CME through interaction with the ZmAP2 σ subunit. We further demonstrate that ZmEHDl affects the gene expression involved in auxin-related processes and that exogenous 1-NAA application rescues the fertility of the ehd1 mutant. Our results indicate that, by regulating auxin homeostasis, ZmEHD1-mediated CME is crucial for kernel development and vegetative growth of maize.
Results
The ehd1 mutant is impaired in kernel development and vegetative growth
The ehd1 mutant was originally isolated as a shrunken kernel mutant in the screening of natural mutagenesis defective in the filling of maize grains. At 14 days after pollination (DAP), the kernels of ehd1 homozygous mutant (ehdl/ehdl) and the wild-type (WT) (EHD1/EHD1 and EHD1/ehd1) resembled each other (Figure 1A). At 16 DAP, the WT kernels were canary-yellow while the ehd1 mutants were ivory-white (Figure 1A). At maturity, both the endosperm and embryo of ehd1 mutant were shrunken (Figure 1B), which led to a significant reduction in the 100-kernel weight (Figure 1C). The 100-kernel weight of the WT (26.9 g) was 3-times greater than that of the ehd1 mutant (Figure 1C).
To analyze the kernel development of the ehd1 mutant and the WT, we examined immature embryos at 20 DAP by light microscopy. Longitudinal sections indicated that embryo development was slower in the ehd1 mutant than in the WT (Figure 1D). Endosperm development and texture were also observed by light microscopy. Compared to WT endosperm cells, those of the ehd1 mutant had less dense cytoplasm and fewer starch granules (Figure 1E).
A paper-culture system was used to measure the germination rates of the WT and
the ehd1 mutant. The germination rate of the ehd1 mutant was only about 3%, which was far lower than that of the WT (Figure 2A). Relative to the WT seedlings, the germinated ehd1 seedlings had shorter and fewer primary roots (Figure 2B). Most of the ehd1 seedlings died before the first two leaves had completely expanded (Figure 2C). These results indicated that ZmEHD1 is essential for maize kernel development and vegetative growth.
The mutant was crossed to an inbred line Xun9058. The hybrid F1 displayed normal kernel as Xun9058. Due to the low germination rate of the ehd1 mutant, the normal kernels from F2 ears were planted. Among the 244 F2 individuals, 79 had the normal kernel phenotype (EHD1/EHD1), and 165 had the segregation phenotype (EHD1/ehd1). The segregation ratio followed 1:2 theoretical ratio predicted by a Chi-square test at a 0.05 level (Supplemental Table 1). For each F2 individual with kernel phenotype segregation, the pattern of shrunken kernel phenotype to normal kernel phenotype fit a 1:3 segregation ratio (Supplemental Table 2). A F3 population with 376 individuals derived from normal kernels of F2 selfed ears was also planted to validate the kernel phenotype segregation. Among them, 130 individuals had kernel phenotype (EHD1/EHD1), and 246 had the segregation phenotype (EHD1/ehd1) (Supplemental Table 1). These results indicate that this shrunken kernel mutant is controlled by a single recessive gene.
Positional cloning of ZmEHD1
Preliminary genetic mapping of ZmEHD1 was carried out by BSA using the F2 population, which contained 165 individuals that differed in seed size in the ears. The ZmEHD1 gene was first mapped to a 2,339-kb region between the simple sequence repeat (SSR) marker umc1650 (17 recombinants) and umc1716 (27 recombinants) on the long arm of chromosome 4 (Figure 3A). Because the germination rate of the ehd1 mutant was rather low, expanded F3 mapping population containing ~53,000 normal kernels (EHD1/EHD1 and EHD1/ehd1) was obtained to increase mapping resolution as described by Zuo et al. (2015). Thirteen new polymorphic SSR markers were developed (Supplemental Table 3), and ZmEHD1 was eventually mapped to a ~6-kb region between the SSR markers RM7 (56 recombinants) and RM13 (5 recombinants) (Figure 3A). This region contains two predicted open reading frames (ORFs) (Figure 3A). Sequence analysis revealed three deletions in the 5’ UTR and eight single-nucleotide polymorphisms (SNPs) in the ORF of GRMZM2G052740 (Figure 3B). Among the eight SNPs in the ORF of GRMZM2G052740, three led to amino acid replacement between the ehd1 mutant and the WT (Figure 3B). In addition to GRMZM2G052740, GRMZM2G052720 was also located in this region. Its candidacy as ZmEHD1 gene was excluded because there was no sequence difference between the WT and the ehd1 mutant. Thus, we inferred that GRMZM2G052740 was the candidate gene for the ZmEHD1 locus.
The genomic DNA sequence of GRMZM2G052740 spanned ~4.4-kb and generated a transcript that included 16 exons (Figure 3B). The corresponding 2,209-bp cDNA sequence encoded a polypeptide of 547 amino acids with a molecular mass of ~61 kD (Supplemental Figure 1). BLASTP analysis in GenBank indicated that GRMZM2G052740 is closed related to the Arabidopsis EHD-containing proteins, AtEHD1 and AtEHD2 (Supplemental Figure 1). The ZmEHD1 protein shares the typical structure of the EHD family; it has an EH domain with two calcium-binding EF-hands (15-39 aa and 49-84 aa), a P-loop (GQYSTGKT), a dynamin-type guanine nucleotide-binding domain, and a coil-coil domain (Figure 3C).
According to the publicly available maize microarray database (www.qteller.com, Waters et al., 2011), ZmEHD1 is expressed in various tissues and its mRNA abundance is high in immature cobs (V18), embryos, and endosperms at 12 and 14 DAP. To verify the microarray data, we performed quantitative real-time RT-PCR using RNA isolated from various maize tissues, including roots, leaves, stems, endosperms, and embryos, and the results confirmed the public microarray data (Supplemental Figure 2). We also isolated RNA from the endosperms of the ehd1 mutant and the WT at 15, 20, 25, and 35 DAP to assess ZmEHD1 expression levels. The transcripts of ZmEHD1 were down-regulated during endosperm development in both the ehd1 mutant and the WT (Figure 3D). Unexpectedly, ZmEHD1 mRNA abundance at all sampling times in the ehd1 mutant was higher than in the WT (Figure 3D), indicating that the mutated ZmEHD1 might be only partly functional.
Validating that ZmEHD1 is the causative locus by ZmEHD1 loss-of-function mutants
To determine whether ZmEHD1 is responsible for the EHD1 locus, we used the CRISPR/Cas9 system to generate ZmEHD1 loss-of-function mutant (KO). The given gene sequencing in ZmEHD1 loss-of-function mutants revealed a deletion of 775-bp or 785-bp fragment at the coding sequence, which resulted in a frameshift (Supplemental Figure 3). ZmEHD1 loss-of-function also led to decreased germination rate as well as retarded vegetative development (Figure 4A and 4B).
An allelism test was performed by crossing KO-1 F1 (KO+/−) and ehd1 heterozygotes (ehd1+/−). The shrunken kernel phenotype (KO/ehd1) and the WT phenotype (EHD1/EHD1; EHD1/KO; EHD1/ehd1) in the hybrid F1 ears displayed a 1:3 segregation ratio (Figure 4C, 4D and Supplemental Table 4), indicating that KO cannot complement ehd1. Meanwhile, the germination rate of the KO/ehd1 mutant was much lower than that of the WT (Figure 4E). These results indicate that ZmEHD1 is the locus affected by ehd1.
The ehd1 and KO mutant are defective in endocytosis
EHD has usually been identified among proteins involved in endocytosis, vesicle transport, and signal transduction (Lin et al., 2001). To test whether the ZmEHD1 protein is involved in CME in maize, we examined the uptake of N-(3-triethylammonium-propyl)-4-(4-diethylaminophenylhexatrienyl) pyridinium dibromide (FM4-64), a commonly used endocytosis tracer, in the ehd1 mutant and the WT. Because maize seedlings are relatively large, we could not place the whole plant on the microscope platform as was previously done with 3-day-old Arabidopsis plants (Fan et al., 2013). Instead, we detached the similar parts of roots from the ehd1 mutant and the WT seedlings that had been treated with 5 μM FM4-64 for 10 min to monitor the FM4-64 endocytosis. At 30 minutes after labeling, FM4-64-labelled fluorescent puncta were detected in 28% of the WT root cells (21 of 75 cells), while no FM4-64-labelled fluorescent puncta were detected in ehd1 root cells (0 of 90 cells) (Figure 5A). Although FM4-64-labelled fluorescent puncta could be detected in root cells of both the WT and the ehd1 mutant at 60 minutes after labeling, FM4-64-labelled fluorescent puncta were detected in only 19% of the ehd1 root cells (22 of 117 cells) but in 45% of the WT root cells (45 of 101 cells) (Figure 5A). The effects of ZmEHD1 on endocytosis were further confirmed by the delayed internalization of FM4-64 in ZmEHD1 knock-out mutants (Figure 5B).
To further investigate whether ZmEHD1 was involved in the vesicle traffic pathway, we inhibited endocytic recycling of FM-64 using the fungal toxin brefeldin A (BFA). The accumulation of FM-64 in BFA bodies was clearly observed in the WT when treated with BFA (Figure 5C and 5D). In contrast, the aggregation of FM-64 in BFA bodies remarkably decreased in the ehd1 mutant and ZmEHD1 knock-out mutants (Figure 5C and 5D). Compared to the WT, the sizes of FM-64 labeled BFA bodies in the ehd1 mutant and ZmEHD1 knock-out mutants decreased by ~67% and ~76%, respectively (Fig. 5C and 5D). These results suggested that ZmEHD1 is important in endocytic transport.
Subcellular localization of ZmEHD1 and ZmEHD1mut
To determine the subcellular localization of ZmEHD1 protein, yellow fluorescent protein (YFP) fused N-terminally to ZmEHD1 was transiently expressed in tobacco (N. benthamiana) leaves under the control of the CaMV 35S promoter. In tobacco epidermal cells expressing the YFP-ZmEHD1 fusion protein, the YFP signal was mainly detected in the PM (Supplemental Figure 4A). The localization was confirmed by staining lipid membranes with the red fluorescent probe, FM4-64 (Supplemental Figure 4A). We also cloned the ZmEHD1 CDS from the ehd1 mutant and examined the subcellular localization of ZmEHD1mut protein. YFP-ZmEHD1mut signals were aggregated and colocalized with FM4-64 signals in tobacco epidermal cells (Supplemental Figure 4B). This phenomenon was consistently observed in different biological replicates. Thus, we deduced that the mutation in ZmEHD1 affected the subcellular localization of ZmEHD1.
ZmEHD1 interacts with the ZmAP2 σ subunit in yeast and plant
In its drastically reduced fertility and endocytosis, the ehd1 mutant resembles Arabidopsis ap2 σ mutants, in which CME and multiple stages of plant development are impaired (Fan et al., 2013). We hypothesized that ZmEHD1 is involved in CME by virtue of its interaction with the ZmAP2 σ subunit. To test this hypothesis, we cloned the cDNA sequence of the maize AP2 σ subunit (GRMZM2G052713) and introduced it into Arabidopsis ap2 σ loss-of-function mutant (Fan et al., 2013). Arabidopsis ap2 σ homozygous for the T-DNA insertion was identified by PCR (Supplemental Figure 5A). RT-PCR showed that ZmAP2 σ subunit was expressed in the Arabidopsis ap2 σ mutant (Supplemental Figure 5B). The transformation of the Arabidopsisap2 σ mutant with ZmAP2 σ subunit rescued the developmental defects, such as reduced leaf size and fertility (Supplemental Figure 5C). These results suggested that ZmAP2 σ subunit has similar functions as AtAP2 σ subunit in CME.
The potential interaction between ZmEHD1 and the ZmAP2 σ subunit was first evaluated by a biomolecular fluorescence complementation (BiFC) assay in tobacco leaves and maize mesophyll protoplasts. The ZmEHD1, ZmEHD1mut, and ZmAP2 σ subunit constructs were introduced into N. benthamiana mesophyll cells or maize mesophyll protoplasts by A. tumefaciens-mediated infiltration. The vitality of the N. benthamiana mesophyll cells was determined by propidium iodide. Reconstituted YFP signals were observed in the PM when both nYFP-ZmAP2 σ subunit and cYFP-ZmEHD1 constructs were co-expressed (Figure 6A and Supplemental Figure 6). Reconstituted YFP signals were also observed when the nYFP-ZmAP2 σ subunit and cYFP-ZmEHD1mut constructs were co-expressed (Figure 6A and Supplemental Figure 6). Consistent with the subcellular localization of ZmEHD1mut, the reconstituted YFP signals were diffused throughout the cells when the nYFP-ZmAP2 σ subunit and cYFP-ZmEHD1mut constructs were co-expressed (Figure 6A and Supplemental Figure 6). This was further verified by the localization of ZmAP2σ subunit in WT maize and ehd1 mutant (Supplemental Figure 7).
We also used a split-ubiquitin membrane-based yeast two-hybrid system to verify the interaction between the ZmAP2 σ subunit and ZmEHD1 as well as ZmEHD1mut. Both ZmEDH1 and ZmEHD1mut interacted directly with ZmAP2 σ subunit in yeast. The interaction resulted in survival in a medium lacking His and β-galactosidase activity (Figure 6B).
Pull-down experiments using purified His-ZmEHD1/His-ZmEHD1mut to isolate GST-ZmAP2 σ subunit also confirmed the physical interaction between ZmEHD1/ZmEHD1mut and ZmAP2 σ subunit. His-ZmAP2 σ subunit was clearly detected in the pull-down fraction (Figure 6C). These results strongly suggested that ZmEHD1 directly interacts with ZmAP2 σ subunit at the plasma membranes in maize.
Transcriptome profiling of the ehd1 mutant
To gain insight into the molecular events involved in the ZmEHD1 -mediated signaling pathway, we compared the whole-transcriptome profiles of endosperms of the ehd1 mutant and the WT at 15 DAP using RNA-seq. Each sample was represented by two biological replicates, and the libraries were sequenced by Illumina high-throughput sequencing technology. These four RNA libraries yielded more than 0.5 billion raw reads, and approximately 97% of the raw reads remained after adaptor polluted reads, Ns reads, and low quantity reads were trimmed. Of the remaining reads, more than 0.12 billion could be perfectly mapped to maize B73 RefGen_V3.27 (ftp://ftp.ensemblgenomes.org/pub/plants/release-27/fasta/zea_mays). Sequences that could not be mapped to the maize genome were discarded, and only those perfectly mapped were analyzed further. The abundance of each gene was expressed by reads per kilo base million mapped reads (RPKM) (Wagner et al., 2012). We calculated the Pearson’s correlation coefficients (R value) of the two biological replicates for each treatment to investigate the variability between the replicates. The R value of both comparisons exceeded 0.99 (Supplemental Figure 8), indicating a high correlation between biological replicates.
Based on the criteria that the Log2 fold-change ratio was ≥ 1 and that the adjusted P value was ≤ 0.05, 4,760 genes, including ZmEHD1, were identified as differentially expressed genes (DEGs) by DEGseq software (Wang et al., 2010). Among the DEGs, 2,208 genes were up-regulated and 2,552 genes were down-regulated in the ehd1 mutant relative to the WT (Supplemental Table 5). The results of RNA-seq were confirmed by quantitative real-time RT-PCR. In agreement with our RNA-seq data, the expression levels of randomly selected GRMZM2G088273, GRMZM2G346897, GRMZM2G067929, and GRMZM2G418119 were lower in the ehd1 mutant than in the WT (Supplemental Figure 9A). As expected, GRMZM2G156877 and GRMZM2G420988 were expressed at higher levels in the ehd1 mutant than in WT (Supplemental Figure 9A), demonstrating the reliability of our RNA-seq data. We then performed Gene Ontology (GO; http://bioinfo.cau.edu.cn/agriGO/) analysis to determine the molecular events related to the DEGs. GO analysis indicated that the 4,760 DEGs were highly enriched for biological processed involved in response to stress (GO:0006950, P = 1.2e-12), integral to membrane (GO:0016021, P = 1e-5), intracellular membrane-bounded organelle (GO:0043231, P = 0.0002), vacuole (GO:0005773, P = 0.0007), starch metabolic processes (GO:0005982, P = 1.6e-5), and carbohydrate metabolic processes (GO:0005975, P = 4.16e-5).
Interestingly, some of the DEGs have known or presumed functions associated with auxin-mediated signaling pathway (GO:0009734, P = 1.2e-25) and response to auxin stimulus (GO:0009733, P = 2.1e-25), including AUXIN RESPONSE FACTOR (ARF), AUX/IAA transcription factor, small auxin up-regulated RNA (SAUR), indole-3-acetaldehyde oxidase, auxin transporters, and efflux carrier (Table 1). We randomly chose four of the DEGs listed in Table 1 (GRMZM2G082943, GRMZM2G365188, GRMZM5G809195, and GRMZM2G019799) and validated their difference in expression levels in the ehd1 mutant vs. the WT by quantitative real-time RT-PCR (Supplemental Figure 9B). These results indicated that ZmEHD1 might affect auxin homeostasis in maize.
Auxin distribution and ZmPIN1a-YFP localization in the ehd1 mutant
To test the hypothesis that ZmEHD1 might affect auxin homeostasis in maize, we first explored the gravitropic response by examining mesocotyl-coleoptiles. We placed the upright mesocotyl-coleoptiles in the horizontal direction. In contrast to less than 4 h for the recovered vertical growth of WT, it took about 5 h for the ehd1 mutant and ZmEHD1 knock-out mutants to recover vertical growth under the same condition (Figure 7A). The free IAA content in kernels of the ehd1 mutant and the WT at 15 DAP were determined. The free IAA content in the kernels of the ehd1 mutant was 2,558 pg mg-1 FW, which was ~30% lower than that in the WT (Figure 7B).
To further demonstrate that ZmEHD1 regulates auxin homeostasis in maize, we first crossed the auxin-responsive ZmDR5::RFP reporter maize with ehd1 mutant to examine the auxin distribution in the root tips of ehd1 mutant. The RFP signals were significantly reduced in the root caps in ehd1 mutant (Figure 7C). We also crossed the ZmPIN1a::YFP line with ehd1 mutant. The YFP signals in the roots of ehd1 mutant showed more diffuse localization than that in PIN1a::YFP EHD1 line, and could be detected in root cortex (Figure 7D). These results suggested that auxin homeostasis was altered in ehd1 mutant.
1-NAA application rescues the fertility of the ehd1 mutant
To further verify that the rather low fertility of the ehd1 mutant and ZmEHD1 knock-out mutant was caused by auxin homeostasis, we attempted to rescue the phenotypic defects of ehd1 mutant and ZmEHD1 knock-out mutants by exogenous application of the active auxin compound, 1-naphthaleneacetic acid (1-NAA). The seeds of the ehd1 mutant and ZmEHD1 knock-out mutants were rinsed in water or in four increasing concentrations of 1-NAA for 12 h, and then germinated in the paper-culture system. Little or no phenotypic rescue of the ehd1 mutant was observed when 30 mg L-1 1-NAA or water alone was applied (Figure 8A). However, the germination rate of ehd1 mutant and ZmEHD1 knock-out mutant was significantly increased when treated with low concentrations of 1-NAA (Figure 8A). Also, when treated with low concentrations of 1-NAA, the ehd1 mutant and ZmEHD1 knock-out mutant survived after its first two leaves had completely expanded (Figure 8B). In contrast to IAA, exogenous GA3 had no effect on the germination of the ehd1 mutant and ZmEHD1 knock-out mutant (Supplemental Figure 10). Overall, these results demonstrated that the phenotypic defects of the ehd1 mutant could be rescued by exogenous application of 1-NAA.
Discussion
In the present research, we characterized the ehd1 mutant, which is impaired in kernel development and vegetative growth, and used positional cloning and transgenic validation to verify that the ehd1 gene encodes an EHD protein. The results of an FM4-64 uptake experiment, split-ubiquitin membrane-based yeast two-hybrid system, BiFC and pull-down assay indicated that ZmEHD1 is involved in CME through its interaction with the ZmAP2 σ subunit. Additionally, the transcriptome profiling of the ehd1 mutant, the auxin distribution and ZmPIN1a-YFP localization in the ehd1 mutant, and the rescue of the mutant phenotypes following the exogenous application of 1-NAA revealed that the ZmEHD1-mediated endocytosis mainly affects auxin homeostasis in maize.
In Arabidopsis, AP2 σ subunit is primarily recruited from the cytoplasm to the PM to initiate clathrin-coated endocytic vesicle formation (Fan et al., 2013; Fan et al., 2015). The developmental defects of ap2 σ loss-of-function Arabidopsis, such as reduced leaf size and fertility, could be rescued by ZmAP2 σ cDNA, suggesting that ZmAP2 σ subunit has similar functions as AtAP2 σ subunit in CME. Microarray data revealed that the AP2 σ subunit is ubiquitously expressed in various tissues throughout maize development (Sekhon et al., 2011), which is consistent with the expression pattern of ZmEHD1. Based on the direct interaction between the ZmAP2σ subunit and ZmEHD1 at the PM as demonstrated by BIFC, pull-down assay and split-ubiquitin membrane-based yeast two-hybrid system in the current study, we suspected that ZmEHD1 contributes to CME by interacting with the ZmAP2 σ subunit.
The naturally occurring ehd1 mutant had amino acid substitutions in three positions. Of these, the V/A substitution was in a linker region. The other two mutations were in the coil-coil domain of ZmEHD1. The coil-coil domain was responsible for protein-protein interactions (Bar et al., 2009; Sharma et al., 2008).
Here, we showed that the subcellular location of ZmEHD1mut was obviously different from that of ZmEHD1. Though ZmEHD1mut could directly interact with ZmAP2σ subunit, the PM was not the main location for interaction between ZmEHD1mut and the ZmAP2σ subunit. The impaired clathrin-coated endocytic vesicle formation and/or the recycling of the ZmAP2 σ subunit from the endosome to the PM should be the major contributor to phenotypes observed in the ehd1 mutant.
Arabidopsis ARF2 controls seed size by repressing cell proliferation (Schruff et al., 2006); in rice, activation of BIG GRAIN1 (BG1) significantly improves grain size by regulating auxin transport (Liu et al., 2015); TGW6, an IAA-glucose hydrolase, negatively regulates endosperm development of rice (Ishimaru et al., 2013). Analyses of a seed-specific viable maize mutant, defective endosperm 18 (de18), demonstrated that the ZmYUC1 gene (which is critical for IAA biosynthesis) is essential for normal endosperm development in maize (Bernardi et al., 2012). Thus, auxin biosynthesis, transport and signaling might coordinate to regulate vegetative and reproductive development in maize (Gallavotti et al., 2008; Li and Li, 2016). The current results with maize show that ARFs, indole-3-acetaldehyde oxidase, auxin transporters, and efflux carrier are differentially expressed in the ehd1 mutant vs. the WT; that the concentration of free IAA is lower in the ehd1 mutant than in the WT; that the auxin distribution and ZmPIN1a-YFP localization were altered in the ehd1 mutant; and that exogenous application of 1-NAA rescues the phenotypes of the ehd1 mutant. These results suggest that the growth defects of the ehd1 mutant result from the loss of auxin homeostasis.
Materials and methods
Plant materials
The ZmPIN1a::YFP and ZmDR5::mRFP lines were kindly provided by Lixing Yuan (China Agricultural University). The maize mutant ehd1 was isolated by screening for natural mutants defective in grain filling. To construct mapping population, the mutant was crossed with inbred line Xun9058 in the winter of 2011 in Sanya, Hainan Province. Xun9058 is the male parent of the elite hybrid Xundan20 and has been widely used in the breeding of hybrid maize in China. F2 individuals were obtained by selfing the F1 plants in the summer of 2012 in Zhengzhou, Henan Province. The site (113°42’E, 34°48’N) is located in central China and has an average annual temperature of 14.3 °C and an average annual rainfall of 640.9 mm. A F2 population with 165 individuals was used to map the preliminary location of ZmEHD1 gene. Normal kernels from F3 segregation ears were used for fine mapping of the candidate gene. The F3 population contains ~53,000 individuals. A small piece of each F3 kernel was chipped and genotyped before planting. At harvest stage, the ear phenotype was investigated to verify the kernel segregation of each individual. Two markers, bnlg589 and RM1, were closely linked to the locus to determine recombined individuals.
From F2 generation, the normal kernels from segregation ears were analyzed by closely linked marker with ZmEHD1 gene and the recombined individuals were selfed. This process was continued for five generations to construct nearly isogenic lines (NILs). The NILs were used as WT.
Molecular markers
Bulked segregant analysis (BSA) was used to detect the genetic linkage of the ZmEHD1 gene (Michelmore et al., 1991) and 1082 SSR markers from the maize genome database (www.maizegdb.org) were tested for polymorphism in the two parents and the two bulks. To develop new markers for fine mapping, the sequence of the B73 reference genome between markers bnlg589 and umc2289 on chromosome 4 was downloaded from MaizeGDB (http://www.maizegdb.org/). SSR-Hunter software was used find simple sequence repeats (SSRs). After BLAST was performed (http://blast.ncbi.nlm.nih.gov/Blast.cgi), SSRs with single copy were developed to new markers with PRIMER 5.0. The newly designed SSRs were tested by polyacrylamide gel electrophoresis (PAGE) for polymorphism in the ehd1 mutant, Xun9058, pooled samples of normal kernels (EHD1/EHD1 and EHD1/ehdl) or shrunken kernels (ehdl/ehdl). SSRs with polymorphism were used for subsequent fine mapping. The sequences of the primers used for mapping are listed in Supplemental Table 3.
Transgenic functional validation
The CRISPR/Cas9-mediated ZmEHD1 editing was performed as described by Xing et al. (2014) with some modifications. In brief, two gRNAs that direct target sequences located at nucleotides 16 and 791 of ZmEHD1 were produced using primers listed in Supplemental Table 3. Thereafter, the fragments were cloned into the pBUE411 vector using the BsaI restriction site by T4 Ligase reaction. The plasmid contained the Streptomyces hygroscopicus phosphinothricin acetyltransferase gene (bar) under the control of a CaMV 35S promoter and was electroporated into A. tumefaciens EHA105. Immature embryos of maize inbred line ZZC01 were transformed by co-cultivation with EHA105 at Life Science and Technology Center of China National Seed Group Co., LTD. Transformants were selected with gradually increasing concentrations of Bialaphos. T2 homozygous lines were sequenced to ensure that ZmEHD1 was knocked out.
For Arabidopsis transformation, the cDNAs of ZmAP2 σ subunit without 3’ UTR were amplified by PCR. The corresponding products were introduced into the pENTRTM/D-TOPO vector (Invitrogen) and cloned into pMDC32 by LR reactions (Invitrogen). The plasmid was electroporated into A. tumefaciens GV3101 and was transformed into Arabidopsis by the floral dip method (Clough and Bent, 1998). Transgenic plants were selected with the use of 35 μg ml-1 hygromycin. The sequences of the primer pairs used in the experiments are listed in Supplemental Table 3.
Expression analysis of ZmEHD1
Total RNA was extracted from the WT and the ehd1 mutant with the RNeasy plant mini kit. The RNA was used for the synthesis of first-strand cDNA with SuperScript III first-strand synthesis supermix (Invitrogen). Quantitative real-time RT-PCR was carried out in an ABI 7500 system using the SYBR Premix Ex TaqTM (perfect real time) kit (TaKaRaBiomedicals), with 18S rRNA and Actin as controls. Primers designed to detect the transcription level of ZmEHD1 were listed in Supplemental Table 3. The relative expression level was calculated using the comparative Ct method. Each experiment was replicated at least three times.
Subcellular localization of ZmEHD1
The full-length ZmEHD1 and ZmEHD1mut coding region was amplified from the WT and the ehd1 mutant. To generate the ZmEHD1 expression plasmids, the 1644-bp fragments were cloned into the pCPB vector, which was fused to an YFP protein in the N terminus using the BamHI restriction site by In-Fusion reaction. The plasmid was electroporated into A. tumefaciens GV3101 and was transiently expressed in tobacco epidermal cells as described previously (Li et al., 2008). To stain the PM of epidermal cells, 5 μM FM4-64 was infiltrated into tobacco leaves for 3 h before observation. YFP/FM-64 images were collected with a Zeiss LSM700 confocal microscope.
Split-ubiquitin yeast two-hybrid assay
The split-ubiquitin two-hybrid system was used to detect the interactions between membrane proteins. The assay was carried out according to the instructions provided with the DUALmembrane Kit (Dualsystems Biotech). The full-length coding regions of the ZmAP2 σ subunit, ZmEHD1 and ZmEHD1mut were amplified and cloned into the vectors pST3-STE and pPR3-N NubG using the SfiI restriction site by In-Fusion reaction. Vectors were co-transformed into yeast strain NMY51. The interactions were assessed by the growth of yeast colonies on synthetic minimal medium containing 7.5 mM 3-aminnotriazole without Leu, Trp, His and Ade and also by chloroform overlay β-galactosidase plate assay (Duttweiler, 1996).
Bimolecular fluorescence complementation (BiFC) assays
The coding sequences of the ZmAP2 σ subunit, ZmEHD1 and ZmEHD1mut were amplified using specific primers listed in Table S1, and were cloned into the pSPYNE-35S or pSPYCE-35S binary vectors using the BamHI restriction site. The various combinations of ZmAP2 σ subunit, ZmEHD1, and ZmEHD1mut expression vectors were transiently expressed in 3-week-old N. benthamiana leaves by Agrobacterium-mediated infiltration (strain EHA105). YFP images were obtained 3 days after infiltration with a Zeiss LSM700 confocal microscope.
Pull-down analysis
The full-length coding sequences of ZmEHD1, ZmEHD1mut and ZmAP2 σ subunit were individually subcloned into the pET-28a(+) and pGEX-4T-1 vectors using the BamHI restriction site by In-Fusion reaction. The resulting constructs were verified by sequencing. His-ZmEHD1, His-ZmEHD1mut, GST and GST-ZmAP2 σ subunit were expressed in Escherichia coli BL21.
For the in vivo pull-down analysis, 10 μg of purified His-ZmEHD1, His-ZmEHD1mut were incubated with GST-ZmAP2 σ subunit or GST bound to glutathione-Sepharose™ 4B (GE Healthcare) for 4 h at 4 °C on a rotary shaker. Precipitated beads were washed six times with washing buffer. Washed beads were boiled with 50 μL of 1 × SDS sample buffer for 5 min and subjected to SDS-PAGE and immunoblot analysis.
FM4-64 internalization assay
FM4-64 internalization assays were carried out to evaluate the rate of endocytosis as described by Fan et al. (2013) with some modifications. WT and ehd1 seedlings were incubated in half-strength Hoagland’s nutrient solution containing 5 μM FM4-64 for 10 min at room temperature. The roots were then cut and transferred to glass slides. The FM4-64 internalization was monitored at indicated durations at room temperature with a Zeiss LSM700 confocal microscope.
RNA-seq analysis
At 15 DAP, total RNA was extracted from the endosperms of the ehd1 mutant and the WT with TRIZOL reagent (Invitrogen), and 3 μg of total RNA was used as input material for construction of the RNA libraries. The RNA-sequencing libraries were constructed with NEBNext® Ultra™ RNA Library Prep Kit for Illumina® (NEB, USA). In brief, mRNA was purified from total RNA using poly-T oligo-attached magnetic beads. The enriched mRNA was then fragmented. The random hexamers were used for first-strand cDNA synthesis. After second-strand cDNA synthesis, terminal repair, and poly(A) tail and sequencing oligonucleotide adaptors ligation, the fragments were purified and subsequently amplified by PCR. The insert size was assessed with the Agilent Bioanalyzer 2100 system (Agilent Technologies, CA). Finally, the libraries of inserting cDNAs with 200-bp in size were generated and sequenced with the IlluminaHiseq 2500 platform (ANROAD, Beijing, China).
The raw reads were produced after exclusion of low quality reads and 5’ and 3’ adaptor contaminants. The unique RNAs were aligned to the maize RefGen_V3.27 (ftp://ftp.ensemblgenomes.org/pub/plants/release-27/fasta/zea_mays). Only perfectly matching sequences were considered for further analysis. The count information was used to determine normalized gene expression levels as RPKM (Wagner et al., 2012). Multiple testing with the Benjamini-Hochberg procedure for false discovery rate (FDR) was taken into account by using an adjusted P-value. Genes were considered to be differentially expressed if the Log2 fold-change ratio was ≥ 1 and if the adjusted P value was < 0.05 according to the DEGseq method (Wang et al., 2010).
Free IAA analysis
At 15 DAP, kernels of the WT and the ehd1 mutant were collected and frozen in liquid nitrogen. A 200-mg (fresh weight) sample of WT and ehd1 kernels was finely ground in liquid nitrogen and then extracted with 1 ml of methanol containing antioxidant and 2H2-IAA (internal standard, CDN Isotopes) at 4°C for 24 h. After centrifugation, the extract was purified with an Oasis Max solid phase extract cartridge (150 mg/6 cc; Waters). IAA was quantified using UPLC-MS/MS consisting of a UPLC system (ACQUITY UPLC, Waters) and a triple quadruple tandem mass spectrometer (Quattro Premier XE, Waters) as described by Wang et al. (Wang et al., 2015). Four independent biological replicates and two technical repeats were performed for the WT and the ehd1 mutant.
Phytohormone treatments
For phytohormone treatments, WT and ehd1 mutant seeds were immersed in water and the indicated concentrations of 1-NAA for 12 h or GA3 for 24 h. The seeds were then kept at 22 °C in the dark for 120 h for germination. A seed was scored as germinated if its radicle protruded through the seed coat.
Additional information
Funding
Author contributions
J.T. and W.X.L. designed the research; Y.W., W.L., H.W., Q.D. and Z.F. performed the research; W.X.L. and J.T. analyzed the data and wrote the article.