Genome-wide analysis of DWD proteins in soybean (Glycine max): Significance of Gm08DWD and GmMYB176 interaction in isoflavonoid biosynthesis

A subset of WD40 proteins with DWD motif has been proposed to serve as substrate receptor of DDB-CUL4-ROC1 complex, thereby getting involved in protein degradation via ubiquitination pathway. Here, we identified a total of 161 potential DWD proteins in soybean (Glycine max) by searching DWD motif against the genome-wide WD40 repeats, and classified them into 20 groups on the basis of their functional domains and annotations. These putative DWD genes in soybean displayed tissue-specific expression patterns, and their genome localization and analysis of evolutionary relationship identified 48 duplicated gene pairs within 161 GmDWDs. Among the 161 soybean DWD proteins, Gm08DWD was previously found to interact with an isoflavonoid regulator, GmMYB176. Therefore, Gm08DWD and its homologue Gm05DWD were further investigated. Expression profile of both genes in different soybean tissues revealed that Gm08DWD was expressed higher in embryo, while Gm05DWD exhibited maximum transcript accumulation in leaf. Our protein-protein interaction studies demonstrated that Gm08DWD interacts with GmMYB176. Although Gm08DWD was localized both in nucleus and cytoplasm, the resulting complex of Gm08DWD and GmMYB176 was mainly observed in the nucleus. This finding is consistent with the functional localization of CUL4-E3 ligase complex. In conclusion, the survey on soybean potential DWD protein is useful reference for the further functional investigation of their DDB1-binding ability. Based on the functional investigation of Gm08DWD, we speculate that protein-protein interaction between Gm08DWD and GmMYB176 may lead to the degradation of GmMYB176 through CUL4-DDB1complex.


Introduction
The WD40 domain containing proteins are highly conserved and are found as tandem repeat units across a wide variety of eukaryotic organisms. They are also called as WD40-repeat PLOS ONE | https://doi.org/10.1371/journal.pone.0178947 June 6, 2017 1 / 20 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 biosynthesis in soybean [30]. GmMYB176 was found to interact with a DWD protein (Gm08DWD) in a yeast two hybrid (Y2H) screening where GmMYB176 was used as a bait protein and the soybean embryo proteins (50 and 60 days after pollination) as preys. In this study, we provide a genome-wide characterization of DWD proteins in soybean, and present a detailed characterization of Gm08DWD that interact with isoflavonoid regulator GmMYB176.

Plant materials
Soybean cultivar Harosoy63 seeds were planted at Agriculture and Agri-Food Canada experimental station in London following regular agronomic practices. Soybean tissues were collected from 5-10 random plants, frozen in liquid nitrogen and stored at -80˚C. Nicotiana benthamiana plants were grown in pots under 16 h light at 25˚C and 8 h dark at 20˚C with 70-80% relative humidity. Arabidopsis seeds were incubated at 4˚C in the dark for 3 days before moving them into the growth room. Plants were grown for 16 h in the light at 22˚C/8 h in the dark at 18˚C.

In silico analysis
The putative GmWD40 proteins were identified by the key word search 'WD40' against Phytozome 12 Glycine max Wm82.a2.v1 database (www.phytozome.net). This search used 201 ontologies from multiple databases including Pfam, Panther, KOG, EC, GO, KEGG Orthology and Cluster KEGG Orthology. Each candidate protein identified from Phytozome soybean database was confirmed for their WD40 domain annotation and architecture by using Simple Modular Architecture Research Tool (SMART) database available at http://smart.embl.de [31]. Following the methods used by Lee et al. [14] and Zhu et al. [15], we first confirmed for the presence of WD40 repeat in each WDR protein using SMART, and then DWD motif was manually identified by using the conserved amino acid sequence (

Multiple sequence alignment and phylogenetic analysis
The amino acid sequences of GmDWD proteins were aligned using ClustalW and the alignment was imported to MEGA7 for phylogenetic analysis. The neighbor-joining method was used to generate a phylogenetic tree based on the midpoint and the p-distance model [32]. To assess statistical significance of the phylogenetic trees, interior branch tests were conducted using 1000 replicates.

Chromosomal localization and gene duplication
The gene location and chromosome number for each of the putative GmDWD was obtained from Phytozome v11.0.7 database (www.phytozome.net). Based on their chromosomal locations GmDWDs were mapped manually on the chromosomes. As described in detail previously [33],GmDWDs in duplicated genomic regions and Ka/Ks values for each duplicated GmDWDs were obtained for syntenic mapping from batch download option of Plant Genome Duplication Database (http://chibba.agtec.uga.edu/). Segmental duplication was defined as the homologous genes located on duplicated chromosomal blocks, while two paralogs with less than 5 gene loci in-between were set as tandem duplication.

Gene expression analysis
The fragments per kilobase of transcript per million mapped reads (FPKM) values for each GmDWD were retrieved to assess the transcript level of GmDWDs by tracking nine tissue gene-level expression from Phytozome database (http://www.phytozome.net), and the data were normalized across tissues. The heatmap for GmDWD genes was produced using R'sheat-map2 function from the gplots CRAN library (http://CRAN.R-project.org/package=gplots).
For quantitative RT-PCR (qPCR) and RT-PCR analysis, total RNA was isolated according to Wang and Vodkin [34] from root, stem, leaf, flower buds, flower, embryo (30, 40, 50 days after pollination), pod wall and seed coat tissues collected from soybean plants. First strandcDNA was synthesized using a Quantitect Reverse Transcription kit (Qiagen Inc.). qPCR was performed using a CFX96 real-time PCR detection system (Bio-Rad Inc.) and QuantiTect SYBR Green PCR kit (Qiagen Inc.). The data were analyzed using a CFX manager, and Soybean ubiquitin-3 (SUBI3) was set as an internal reference for data normalization [35,36]. The experiment was conducted with three technical replicates for each independent biological replicate. The primer sequences are listed in S1 Table. Histochemical GUS assay Upstream regions from translation start site (~1.28 kb) for Gm05DWD and Gm08DWD were cloned into pMDC162 vector using the primers listed in S1 Table, and then transferred into Arabidopsis using Agrobacterium tumefaciens mediated floral dip method [37]. For GUS staining, seedlings and different tissues were immersed in GUS staining solution (0.5 mg/mL 5-bromo-4-chloro-3-indolyl-glucuronide, 20% methanol, 0.01 M Tris-HCl, pH 7.0) with vacuum for 15 min, and then incubated at 37˚C overnight [38]. The samples were cleared by sequential incubation in 75% and 95% ethanol.

Targeted yeast two hybrid assay
For the targeted Y2H assay, full-length cDNA of GmMYB176 and Gm08DWD were PCR amplified using gene-specific primers (S1 Table) and cloned into the Gateway entry vector pDONOR-Zeo by homologous recombination to obtain pDONZ-GmMYB176 and pDONZ-Gm08DWD and sequences were confirmed. The entry constructs were recombined into the Y2H destination vectors pGBKT7-DEST (bait) and pGADT7-DEST (prey) to obtain pGBKT7-GmMYB176, pGBKT7-Gm08DWD, pGADT7-GmMYB176 and pGADT7-Gm08DWD. As described in detail previously [33], the vectors in different combinations were co-transformed into yeast strain AH109 and selected on SD/-Leu/-Trp agar plates. Selected individual yeast transformants were grown in liquid medium and 5 μL AH109 culture with a series of 10-fold dilution was dropped onto SD/-Leu/-Trp and SD/-Ade/-His/-Leu/-Trp plates and grown for 8 days at 30˚C. Empty vectors were used as negative controls.

Identification and classification of GmDWD proteins in soybean
Since DWD motifs are found within WD40 repeats, we first retrieved all annotated WD40 proteins from Phytozome 12 Glycine max Wm82.a2.v1 using key word search. This search identified a total of 471 putative WD40 proteins in soybean genome. Subsequently, WD40 regions within each predicted WD40 protein were manually searched for the 16 conserved amino acid sequence for DWD motif: . A total of 161 putative DWD proteins were identified by this search in soybean (S2 and S3 Tables). Among them, 125 proteins contained single DWD domain, 34 contained 2 domains, and 2 proteins contained 3 DWD domains (Fig 1). A search for soybean DWD orthologs in Arabidopsis [14] found 130 DWD proteins, where their identity ranged from 43.6% to 89.1% (S2 Table). Among them, 62 GmDWDs corresponded to the DDB1 interactors in Arabidopsis that are experimentally validated [9,[16][17][18]. These results suggest that these DWD proteins are conserved in both legume and non-legume plants.
The candidate DWD proteins were further assessed for additional domains using SMART analysis. Based on the number of DWD motif and domain diversity, the 161 GmDWD proteins were classified into 20 types ranging from A to T (Fig 1). Majority of GmDWD proteins (132) contained either single or multiple DWD domain (s) whereas some GmDWDs (29) contained other known functional domains (Fig 1 and S2 Table). For example, Glyma.09G037300 and Glyma.15G142600 include BROMO domain that function in binding with acetyl-lysine during histone acetylation [42]. Similarly, 4 predicted DWD proteins contain a LisH motif that helps to regulate microtubule dynamics by assisting microtubule dimerization [43]. Some of other DWDs with additional domains are as following (Fig 1 and S2 Table): the TPR domain that facilitates protein-protein interaction was found in 2 DWD proteins. The E2-dependent ubiquitination-related domains were detected in 3 DWD proteins. Glyma.08G040500 contains the DENN domain implicated in the regulation of mitogen-activated protein kinase signaling pathways [44]. Glyma.20G131200 contains LRRcap facilitating the interaction with U2 snRNA. Five DWD proteins have the STYKc domain that possesses catalytic specificity for tyrosine kinase whereas two DWDs contain S_TKc domain that have serine/threonine protein kinase catalytic activity. The DWDs with BING4CT domain at the C-terminus end may function as nucleolar WDR proteins. Glyma.02G226300 and Glyma.14G193200 have ZnF_C2HC domain. These additional domains might be implicated in the recognition of a variety of substrates since the DWD motif binds DDB1 while other portions of the protein may bind substrates. On the basis of the functional annotation, we grouped the 161 GmDWD proteins into 8 categories (S1 Fig), where majority of DWDs are predicted to involve in RNA processing (22.4%) and signal transduction (23.0%).

Chromosomal localization and phylogenetic analysis of soybean DWD genes
To explore the evolutionary relationships of 161 predicted DWDs in soybean, a phylogenetic tree was generated. The results revealed that the putative DWD proteins could be categorized divided into 7 distinct groups (I to VII) (Fig 2). The groups VI and VII contained 51 and 57 candidate GmDWDs, respectively, while relatively fewer GmDWDs clustered in group I-V. It was observed that GmDWDs with the same or similar annotations were clustered together in the phylogenetic tree. For example, 8 CAF1/NURF55/MSI1 proteins clustered together with each other in the group VI including Glyma.05G131200, Glyma.08G085900, Glyma.09G063100, Glyma.11G091500, Glyma.12G033100, Glyma.13G350500, Glyma.15G024000 and Glyma.15G169800, while 4 mRNA export protein (Glyma.01G067900, Glyma.02G124400, Glyma.08G180700 and Glyma.15G051700) formed a clade in the group VII. Furthermore, 62 GmDWDs, whose orthologs in Arabidopsis showed interaction with  Table). It has been proposed that COP1 forms multiple complexes with SUPPRESSOR OF PHYA-105 (SPA1-4) family members to suppress photomorphogenic growth [19]. In the phylogenetic tree, the GmDWDs orthologs of SPAs, COP1, MSI and FY in Arabidopsis were clustered into a discrete clade (Fig 2), suggesting that they have close relationship during evolution.
The genomic distributions of all candidate GmDWD genes were determined by their location on soybean chromosome. The GmDWD genes were dispersed unevenly on all 20 chromosomes in soybean (Fig 3). Some chromosomes contain a relatively high number of GmDWD genes while others contain very few. For example, 19 DWD genes are present on chromosome 15, followed by 12 genes on chromosome 13 whereas only 2 GmDWD genes are present on chromosome 3 (Fig 3). Segmental duplication and tandem amplification of chromosomal regions are main contributors for gene extension during evolution [45]. Generally, tandem amplification is defined as two paralogs separated by less than five genes in the same chromosome [46]. Three pairs of GmDWD genes in the same chromosome are found close to each other such as Glyma.03G243800 and Glyma.03G244500; Glyma.19G241300 and Glyma.19G241800; Glyma.20G132800 and Glyma.20G133400 (Fig 3 and S2 Table). These genes share greater than 93.3% of sequence identity at both protein and nucleotide level, suggesting that they are likely derived from tandem amplification of chromosomal regions. Based on coordinates of GmDWD genes, we further investigated whether traceable genome duplications contributed to the expansion of GmDWD genes in soybean. As indicated in S4 Table, 45 sets of GmDWDs were mapped on 34 distinct duplicate blocks, and each set of GmDWDs were clustered into a discrete clade in phylogenetic tree with 75.5-99% sequence identity, suggesting that these pairs of GmDWDs on the same block are possibly derived from segmental duplication events during evolution. To investigate the selective evolutionary pressure on GmDWD gene divergence after duplication, the non-synonymous/synonymous substitution ratio (Ka/ Ks) was retrieved for the 46 duplicated pairs of GmDWD genes from Plant Genome Duplication Database. As shown in S4 Table, the Ka/Ks value of all the duplicated gene pairs ranged from 0.025 to 0.574. Since the Ka/Ks values were less than 1, these genes might have undergone a purifying selection with limited functional divergence after duplication.

Soybean DWD genes display tissue-specific expression pattern
To investigate the expression patterns of candidate DWD genes, we mined the publicly available transcript profiling data of soybean tissues at the Phytozome database (http://www. phytozome.net). As shown in Fig 4, all DWD genes displayed tissue-specific expression patterns. None of the gene was expressed ubiquitously in all the tissues under study suggesting that these genes have unique role in the tissue that they are expressed in. Thirty five GmDWD genes showed higher transcript accumulation in seeds. Similarly, 12 GmDWD genes in pods, 30 genes in flower, 27 genes in leaf, 10 genes in stem, 17 genes in shoot apical meristem, 25 genes in root, and 5 in nodule accumulated highest level of transcripts (Fig 4 and S5 Table). The maximal fold change for each of DWD genes was calculated using the ratio of the maximal and minimal FPKM in different tissues. It was found that the maximal fold change among different tissues ranged from 1.34 to 119.62 (the expression of Glyma.15G138800 and Glyma.12G155700 were undetectable in some tissues), and 125 GmDWD genes showed more than 2.0 maximal fold changes (S5 Table). Noticeably, 10 GmDWD genes were highly expressed in one or few tissues including Glyma.06G046400, Glyma.06G185900, Glyma.06G242100, Glyma.07G058200, Glyma.12G155700, Glyma.13G184300, Glyma.13G329300, Glyma.15G138800, Glyma.16G027200, and Glyma.20G126400, suggesting a potential function in the specific tissue (s). For example, Glyma.15G138800 was mainly expressed in nodules, whereas its transcript accumulation was very low or undetectable in other tissues included in the study (Fig 4 and S5  Table), implying its potential function in nodule-related processes. Similarly, expression of Glyma.06G046400 was distinct in flower among other tissues, suggesting that this gene might play an important role in the regulation of flower-related processes. Further observation indicated that not all the closely-related DWDs in phylogenetic tree were clustered together showed similar expression pattern. For example, 4 GmDWD genes (Glyma.01G067900, Glyma.02G124400, Glyma.08G180700 and Glyma.15G051700) encoding putative mRNA export protein varied in their transcript accumulation profile (Fig 4).

Gm08DWD interacts with GmMYB176
Previously we demonstrated that GmMYB176 regulates isoflavonoid biosynthesis by activating the GmCHS8 gene expression [30]. To identify the proteins that interact with GmMYB176, an Y2H screen was performed where GmMYB176 was used as the bait and protein from embryos  To validate the result from the Y2H screening, a targeted Y2H assay was performed where GmMYB176 was used as the bait and Gm08DWD as the prey protein. As shown in Fig 5A, yeast colony growth was observed on the selective medium lacking Leu/Trp/Ade/His when the above the heat map indicates gene expression levels, low transcript abundance indicated by green color and high transcript abundance indicated by red color. SAM, shoot apical meristem.
https://doi.org/10.1371/journal.pone.0178947.g004 Soybean DWD proteins and the interaction of Gm08DWD with GmMYB176 both the plasmids containing GmMYB176 as bait and Gm08DWD as prey were present, indicating the physical interaction between GmMYB176 and Gm08DWD proteins.
To confirm the interaction between GmMYB176 and Gm08DWD in planta, a BiFC analysis was carried out where split fluorescent protein segments were brought together to form a functional protein due to protein-protein interaction as described in detail previously [40]. Translational fusion of Gm08DWD or GmMYB176 was generated in the binary vector that contained either N-terminus half (YN) or C-terminus half (YN) of YFP. Tobacco leaves were co-infiltrated with A. tumefaciens containing pGm08DWD-YN and pGmMYB176-YC or pGmMYB176-YN and pGm08DWD-YC, and protein expression was monitored in leaf epidermal cells by confocal microscopy. The negative controls included the following combinations: (i) Gm08DWD-YN or -YC with the non-fusion half of YFP, (ii) GmMYB176-YN or -YC with the non-fusion half of YFP, and (iii) two non-fusion halves of YFP, YN and YC. As shown in Fig 5B, the interaction between Gm08DWD and GmMYB176 was confirmed in planta and the YFP signals were observed in the nucleus. The negative controls showed no signal (data not shown).

Characterization of Gm08DWD gene
The full-length Gm08DWD cDNA sequence (966 nucleotides) was predicted to encode a protein of 321 amino acid residues with a calculated molecular mass of 35.8 kDa and a pI of 5.2. It consists of 8 exons and 7 introns and is located on chromosome 8 in soybean (Fig 3). The SMART analysis revealed 7 WD40 domains in Gm08DWD, with a single DWD motif within its third WD40 domain (Fig 6, Type A in Fig 1).
The phylogenetic analysis grouped Gm08DWD together with Gm05DWD (Glyma.05G187400) that resides on the chromosome 5 (Figs 2 and 3). Both of them map on the duplication block 570 with 93.9% sequence identity at nucleotide level, suggesting they are duplicated gene pair. Both Gm05DWD and Gm08DWD were annotated as Mitogen-activated protein kinase organizer 1 (MORG1), and share 97.7% sequence identity at amino acid level. MORG1 has been reported to act as modular scaffold involved in various processes in human and animal [47,48]. To provide some functional clue of Gm08DWD and Gm05DWD, we obtained MORG1-like sequences from other species including human, and performed a multiple sequence alignment. The result indicated that they are highly conserved with 45.3-97.7% sequence identity (Fig 6).
To examine the subcellular localization of Gm08DWD, a translational fusion of Gm08DWD with YFP was created and transiently expressed in tobacco leaf epidermal cells. As shown in Fig  7, Gm08DWD was localized both in the nucleus and the cytoplasm. The intensities of YFP signals were similar in both the subcellular compartments.
Despite of their high sequence similarity, Gm05DWD and Gm08DWD were expressed differentially in soybean tissues as indicated by the qPCR analysis using gene-specific primers. As shown in Fig 8, Gm08DWD expression was detected in root, stem, leaf, flower buds, flower, pod walls and seed coat tissues, however, the transcripts accumulated to higher levels in the developing embryos. In contrary, Gm05DWD accumulation was highest in the leaf tissue followed by developing embryos (Fig 8). To determine the spatial expression pattern of Gm05DWD and Gm08DWD genes in detail, we cloned the promoter regions of each of the genes covering 1.28 kb upstream of translational start site to drive GUS reporter gene, and transformed into wild-type Arabidopsis Col-0. Transgenic lines were selected for Gm05DWDpro:GUS and Gm08DWDpro: GUS and analysis of GUS expression was conducted in T 2 generation by histochemical staining using multiple independent transgenic lines. The results indicated that strong GUS staining was detected in seedling, rosette leaf, stigma, filament, anther, flower stalk, pod wall for both transgenic plants (Fig 9), suggesting that Gm05DWD and Gm08DWD might perform similar  Soybean DWD proteins and the interaction of Gm08DWD with GmMYB176 functions in these tissues. Different GUS staining patterns were observed in seeds of two transgenic plants. As shown in Fig 9, GUS staining was clearly observed in young seeds of Gm05DWD transgenic plants, whereas relatively weak GUS staining was detected in mature seeds. In contrast, young seeds of Gm08DWD transgenic plants showed very weak GUS staining, while clear GUS staining was observed in mature seeds. These observations indicated that Gm05DWD and Gm08DWD might play different roles during seed development.

Discussion
DWD proteins are encoded by a large multigene family in plants. In Arabidopsis, 85 DWD proteins were identified based on their conserved 16-amino acid residues [14]. Among them, 27 DWD proteins have been experimentally verified for their function as substrate receptors of CUL4-DDB1 [9,[16][17][18]. In this study, we identified 161 putative DWD protein encoding genes in soybean genome. It has been proposed that the DWD box within the WDR proteins binds to DDB1 proteins [4][5][6]. Although we did not experimentally define their interactions with DDB1, 62 GmDWDs were found to be orthologous to the DDB1 interactors in Arabidopsis such as COP1, SPAs, MSIs, DWAs (Fig 3, S1 and S2 Tables). Besides, 130 putative GmDWD proteins have their orthologs in Arabidopsis (S1 Table), which have identified as DWD protein by Lee et al [14]. These findings suggest that the DWD proteins identified in this study may have potential to bind to CUL4-DDB1 complex. However, additional protein lacking DWD motif can also bind to the CUL4-DDB1 complex [4,10,49], implying the underestimation of the number of soybean DDB1-binding proteins.
It has been proposed that DWD proteins can interact with DDB1 and serve as the substrate-recognition subunits of the CUL4-DDB1 ubiquitin E3 complex, and DWD motif is required for efficient DDB binding. Majority of putative DWD proteins in soybean (77.6%) contain single DWD motif (Fig 1), suggesting that one copy is sufficient to bind DDB1. However, 34 putative DWD proteins consisted of 2 DWD motifs, and 2 putative DWD proteins consisted of 3 DWD motifs within the sequence. This finding is consistent with the previous report that DWD proteins usually possess 1 and sometimes 2 but rarely 3 DWD motifs [5]. It has been suggested that the additional DWD motifs may enhance DDB1 binding or may interact with other proteins [14]. Besides WD40 domain, 29 DWD proteins contain several other known functional domains (Fig 1), which might be involved in interaction with substrate or other components. The presence of additional domains in the 29 soybean DWD proteins supports the hypothesis that the DWD motif binds DDB1 while other portions of the protein may bind substrates [14]. The remaining 132 putative GmDWDs do not possess any additional domains. It is possible that these DWDs bind their substrates with some unidentified motif. The features of DWD proteins such as the large number, diverse domains and multiple DWD motifs suggest functional diversity in some aspects, and is consistent with the proposal that CUL4-mediated ubiquitin E3 complex regulates diverse processes such as RNA processing, protein assembly and degradation, signal transduction, epigenetic regulation, cell cycle progression, cytoskeletal dynamics (S1 Fig). Although 26.1% GmDWDs have not been annotated or characterized yet, their Arabidopsis orthologs identified in this study provides clues for their possible functions. Additionally, DWD genes displayed tissue-specific expression patterns. Ten putative GmDWD genes were highly expressed in one or few tissues only while others were expressed in multiple tissues (Fig 4). It has been proposed that gene expression provides functional specificity in certain tissues [50]. Diverse expression patterns in soybean tissues indicated functional diversification among GmDWD genes.
Soybean DWD genes were unevenly distributed on the 20 chromosome. Evidently, some chromosomes exhibited dense distribution of DWD genes, whereas others contained sparse distribution of them (Fig 3). Similar DWD gene distribution on chromosomes was reported in foxtail millet [12] and tomato [15]. The uneven distribution of the DWD genes suggested diverse contributions of soybean chromosomes to the formation and expansion of DWD gene family. It has been estimated that soybean genome has undergone two whole genome duplication events approximately 56.5 and 19.2 million years ago [51]. This has resulted into duplication of at least 75% of gene in soybean genome [52]. Here, we identified 48 duplicated gene pairs within the 161 GmDWDs. Among them, 3 pairs of GmDWD genes were likely derived from tandem amplification and 45 pairs of GmDWD genes from segmental duplication (S4 Table). Each pair of duplicated genes formed a discrete clade in the phylogenetic tree with 75.5-99% sequence identity (Fig 2), indicating a closely-related evolutionary relationship. Evidently, the origins of these putative GmDWD genes are in agreement with soybean evolutionary history. In addition, their low Ka/Ks values implied that these duplicated genes might have undergone a purifying selection with limited functional divergence after duplication. Taken Arabidopsis and selected T2 transgenic plants were used for analysis in vegetative and reproductive tissues during various stages of development. https://doi.org/10.1371/journal.pone.0178947.g009 Soybean DWD proteins and the interaction of Gm08DWD with GmMYB176 together, the gene duplication and sequence identity together with low Ka/Ks values suggest that each pair of GmDWD duplications or amplifications possibly share similar function with each other. Intriguingly, 14 pairs of duplicated genes shared similar expression pattern in different tissues (Fig 4 and S4 Table). Since functional role of a gene can be reflected by their temporal and spatial expression [50], we speculate that the 14 pairs of duplicated genes may display functional redundancy in soybean.
It has been demonstrated that plant DWD proteins play important roles in diverse processes, including the regulation of photomorphogenesis and flowering time [19,24], signal transduction [22,23], chromatin modification [20], stress response [9,53], as well as gametophyte [21], embryo and endosperm development [17]. In this study, we investigated the interaction of Gm08DWD with an isoflavonoid regulator GmMYB176 and speculated the possible consequence in isoflavonoid biosynthesis. Previously we demonstrated that GmMYB176 regulates CHS8 gene expression and affects isoflavonoid biosynthesis in soybean seeds [30]. We also revealed that 14-3-3 proteins regulate the intracellular localization of GmMYB176 thereby affecting isoflavonoid biosynthesis in soybean [40]. To further dissect the mechanisms of GmMYB176-mediated regulation of isoflavonoid biosynthesis of soybean seeds, an Y2H assay was performed using GmMYB176 as the bait and proteins from developing embryo as the prey. Here we demonstrate the interaction of GmMYB176 with Gm08DWD ( Fig 5). Furthermore, Gm08DWD showed a high expression level in soybean embryo and Arabidopsis mature seeds (Figs 8 and 9), which is consistent with the previous report that the expression of GmMYB176 is relatively high in soybean embryo. The similar expression patterns may imply an association between these two genes during seed development and maturation. Based on these evidences, we speculated that Gm08DWD might be involved in the regulation of isoflavonoid biosynthesis in soybean seeds through its interaction with GmMYB176. Since Gm08DWD is putatively a DWD protein and has 81.2% identity with its Arabidopsis homologue which is predicted as a component of CUL4-DDB1complex and G protein complex in Arabidopsis, we speculate its similar function in soybean. In Mammalian cells, MORG1 was found to interact with PHD3, thereby regulating the protein levels of HIF-1α via proteasome degradation pathway [47]. MORG1 has not been functionally characterized in plants, however, sequences similar to MORG1 has been identified in many plant species including Gm08DWD in soybean (Fig 6). The fact that GmMYB176 and Gm08DWD interact with each other and the site of interaction is nucleus (Fig 5), is consistent with the functional localization of CUL4-E3 ligase complex [53]. Thus, we speculate that CUL4-DDB1 complex or other protein complex might specifically recognize the GmMYB176 via its specific substrate, Gm08DWD, thereby regulating the degradation of GmMYB176 through the ubiquitination pathway. Further experimental evidences are required to support this speculation.