The legume-specific transcription factor E1 controls leaf morphology in soybean

Background The leaf is a determinate organ essential for photosynthesis, whose size and shape determine plant architecture and strongly affect agronomic traits. In soybean, the molecular mechanism of leaf development is not well understood. The flowering repressor gene E1, which encodes a legume-specific B3-like protein, is known to be the gene with the largest influence on soybean flowering and maturity. However, knowledge of its potential other functions remains poor. Results Here, we identified a novel function of E1 protein in leaf development. Unifoliolate leaves of E1-overexpression (E1-OE) lines were smaller and curlier than those of wild type DongNong 50 (DN50) and Williams 82 (W82). Transverse histological sections showed disorganized cells and significantly elevated palisade tissue number, spongy tissue number, and bulliform cell number in E1-OE lines. Our results indicate that E1 binds to the promoters of the leaf- development-related CINCINNATA (CIN)-like TEOSINTE BRANCHED1/CYCLOIDEA/PROLIFERATING CELL FACTOR (TCP) transcription factor genes to negatively regulate their expression. Conclusions Our findings identify E1 as an important new factor in soybean leaf development. Supplementary Information The online version contains supplementary material available at 10.1186/s12870-021-03301-1.


Background
The shape of the leaf, the main photosynthetic organ in plants, varies based on species and developmental stage [1][2][3]. Leaf size and shape are important for leaf function and plant survival [4][5][6][7]. Genes that influence leaf formation have been described in rice, Cardamine, tomato, and Arabidopsis thaliana [1,3,[8][9][10][11]. For example, TCP3 gene regulates Arabidopsis leaf development via the jasmonate signaling pathway [12,13]. In rice, curled later1 (cur1) mutants have narrow leaves compared with wild type (WT) [3]. Little is known, however, about the molecular mechanisms of leaf development of in soybean (Glycine max), despite its status as an economically important plant oil and protein crop.
In 1927, a pair of genes controlling soybean maturity were detected and named E and e [14]. In 1971, Bernard confirmed that these genes are identical to E1 and e1, two alleles of a major locus that influencing maturity [15]. The E1 locus has a largest impact on flowering time and maturity in cultivated soybean [16][17][18]. Xia et al. mapped E1 through positional cloning, and discovered that it contained a putative nuclear localization signal and sequences related to the plant-specific B3 domain [19]. The E1 protein is assumed to be a transcription factor unique to legumes [18,19]. In cultivated soybean, it inhibits flowering: the leaky allele (e1-as) and the loss-offunction alleles (e1-fs and e1-nl) cause earlier flowering under long-day (LD) conditions [17,19,20]. The molecular mechanism of E1 regulation of soybean flowering has been well studied, with both genes acting upstream of E1 (such as E3, E4, J, LHYs, Tof11, and Tof12) [19][20][21][22] and genes acting downstream of E1 (such as FT2a and FT5a) reported [19,[23][24][25]. In soybean, some important flowering genes have different functions in other aspects of plant development [26,27]. For instance, FT5a have dual function in the regulation of post-flowering stem growth and flowering time [27]. Nonetheless, possible roles of E1 in other developmental processes are unknown.
As compared with DN50, E1-OE transgenic plants flowered significantly later under long-day conditions, and the plants were much shorter (Additional file 1: Fig.  S1a-b). Assessment of the unifoliolate leaves of E1-OE and DN50 plants at 7 DAE, revealed that E1-OE plants had smaller leaf areas and lower leaf weights than DN50 (Fig. 1c-e), they also curled downward (Fig. 1c). Higher E1 expression in the E1-OE lines was associated with increased curliness of the leaves (Fig. 1c).
We also observed the phenotypes of T 1 E1-OE transgenic lines in the Williams82 (W82) background. Consistent with our observations of E1-OE transgenics in DN50, leaf area and weight of E1-OE plants (E1-OE5-8) were smaller and lighter than those of W82 (Additional file 1: Fig. S2a-c). Thus, E1 may regulate leaf development in soybean. Since the genetic stability of T 7 E1-OE lines in DN50 was higher than that of T1 E1-OE lines in W82, the E1-OE transgenic lines in DN50 were selected for subsequent experiments.

E1 regulates cell number and size in the developing leaf
Proper balance of leaf tissue structure is critical for normal leaf development [34]. To further our understanding of the processes controlling leaf development, we analyzed transverse histological sections of E1-OE unifoliolate leaves. Compared with those in DN50 plants, cells were more disorganized in E1-OE plants (Fig. 2a). As an additional approach to examine the effects of E1-OE, we compared leaf functional traits such as leaf thickness, the cell tense ratio (CTR), spongy tissue ratio (SR), cell number and cell size ( Fig. 2b-i). Leaf thickness was similar in E1-OE1 and E1-OE2 lines, but increased in E1-OE3 and E1-OE4 lines (Fig. 2b). The CTR, SR, and spongy tissue size were similar in E1-OE and DN50 plants (Fig. 2g, h, i). In contrast, E1-OE plants had significantly higher palisade tissue number, spongy tissue number and bulliform cell number (Fig. 2c-e), and lower palisade tissue size (Fig. 2f ), confirming that the E1 could regulate leaf development by affecting the leaf tissue structure. We found that higher expression of E1 was associated with a more obvious cellular phenotype, confirming that E1 could influence the balance of different cells within the leaf tissue.

RNA-seq analysis of E1 overexpression
To identify the genes and signaling pathways related to E1-mediated leaf development, we performed RNA-seq analysis and analyzed the differentially expressed genes (DEGs) in the E1-OE transgenic and DN50 plants. The gene expression levels were similar between two biological replicates (Fig. 3a), but differed significantly between the E1-OE transgenic lines and DN50 plants. Genes involved in metabolic process, cellular process, singleorganism process, response to stimulus, and biological regulation were enriched in the DEGs (Fig. 3b). We compared the RNA-seq datasets and identified a total of 7407 DEGs (FDR P < 0.01) (Additional file 2: Data S1). Among these, 3966 genes were significantly upregulated and The soybean TUB (GmTubulin) gene was used as an internal control to normalize gene expression data. The experiment was performed using three biological replicates, each with three technical replicates. Differences between groups were statistically analyzed using Student's t-test (**P < 0.01). Bars indicate standard error of the mean. c. Top view of DN50 and E1-OE plants. Red boxes indicates that the close-up image of the leaf of DN50 and E1-OE plants. d. Quantification of leaf size in DN50 and E1-OE plants (n = 10). e. Quantification of leaf weight in DN50 and E1-OE plants (n = 10). All values are presented as mean ± standard error of the mean (s.e.m.) (n = 10 plants). Bars indicate the s.e.m. Significant differences were identified by Student's t-test (**P < 0.01) 3441 genes were significantly downregulated (Fig. 3c). The Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis demonstrated that some primary metabolic pathways that are essential for plant growth and development were significantly enriched; these included fatty acid metabolism, phenylpropanoid biosynthesis, cysteine and methionine metabolism, and starch and sucrose metabolism (Fig. 3d).
To examine the regulatory effect of E1 on its target genes, we performed transient expression assays, using TCP14 and TCP29 promoters fused to the LUC reporter (pTCP14:LUC and pTCP29:LUC; Fig. 5a). The effector construct harbored E1 under the control of the 35S promoter (p35S:E1; Fig. 5a). We transformed the reporter construct (pTCP14:LUC or pTCP29:LUC) and the effector construct (p35S:E1) into healthy N. benthamiana leaves and found that E1 significantly repressed TCP14 and TCP29 expression (Fig. 5b). Thus, E1 regulates leaf development by repressing CIN-type TCPs.
To determine whether E1 directly inhibits the expression of TCP genes, we performed a ChIP-qPCR assay to compare the relative enrichment of specific TCP14 and TCP29 sequences in E1-OE and DN50 plants using anti-Flag antibodies. We selected four sites in the 2027 bp and 2209 bp regions upstream of the TCP14 and TCP29 promoters, respectively (Fig. 5c). The E1 protein was highly enriched in the TCP14 promoter sites 1 and 4, and in the TCP29 promoter site 1, whereas it was present at  The transcription levels of TCP5, TCP7, TCP11, TCP13, TCP14, TCP15, TCP18, TCP19,  TCP29, TCP32, TCP47 and TCP49 in DN50 and E1-OE plants; data obtained by qRT-PCR. Significant differences were analyzed based on the results of three biological replicates, each with three technical replicates (Student's t-test: *P < 0.05, **P < 0.01). Bars indicate the standard error of the mean extremely low levels in the DN50 control (Fig. 5c). These results showed that E1 could directly bound the promoters of TCP genes.

The transcript levels of the CIN-type TCP genes in soybean tissues
To understand the functions of CIN-type TCP genes in soybean, we use an RNA-seq database and retrieved transcript levels of 10 of the TCP genes repressed by E1, in eight different tissues (flower, leaf, pod, shoot, nodule, cotyledon, seed and root; Machado et al., 2020). These genes presented similar expression profiles and were constitutively expressed to high levels in the leaf and cotyledon (Fig. 6a-k). In contrast, all CIN-type TCP genes displayed low transcript abundance in nodule and root, except TCP6 high expression level in seed and root (Fig. 6a-k). Moreover, TCP13, TCP47, and TCP49 presented similar expression profiles and were highly expressed in pod, flower and shoot (Fig. 6b, e, j, k). TCP7, TCP14, TCP19, TCP11, TCP18, and TCP29 were expressed in shoot and seed, seed, flower and shoot, shoot, pod and shoot, and flower, respectively (Fig. 6c, f,  h, d, g, i).
To determine the tissue-specific expression patterns of TCP genes, we assayed the transcript levels of 10 CIN-type TCP genes by qRT-PCR. The tissue-specific expression patterns in the qRT-PCR were similar to those in the RNA-seq data (Fig. 6l). Thus, the CIN-type TCP genes regulated by E1 play key roles in soybean leaf development.

Discussion
The role of E1 in controlling soybean flowering time and maturity is well known; the molecular mechanisms have also been reported [19,21,22,35]. When E1 is knocked out in soybean, flowering is promoted by the derepression of two important FT genes (FT2a and FT5a) under long-day conditions [19,35]. To explore the functions of E1 in other development pathways in soybean, we created E1-OE transgenics and compared them with the DN50 wild type plants. E1-OE transgenic lines flowered significantly later under long-day conditions (Additional file 1: Fig. S1b). Consistent with previous reports [19,35], our findings demonstrate that E1 functions as a flowering repressor in soybean. We observed smaller, lighter, and curlier unifoliolate leaves in E1-OE transgenic plants (Fig. 1, Additional file 1: Fig. S1), suggesting that E1 might play an important role in leaf development.
The size and shape of a leaf are major traits that affect yield in soybean [36][37][38][39][40]. The narrow leaf trait is Luciferase activity under the control of TCP14 and TCP29 promoters, from three biological replicates. A Student's t-test was used to detect statistically significant differences. c. Location of the promoters of TCP14 and TCP29 and the amplicons targeted in ChIP-qPCR assay. d. Results of ChIP-qPCR on TCP amplicons in DN50 and E1-OE lines fused with Flag tags. A monoclonal Flag antibody was used for the ChIP assay associated with increased seed number than broad leaf trait in soybean [36]. The locus that controls leaf shape cosegregates with the locus that controls the seed number [40]. For example, Jeong et al. [40] recorded the numbers of 1-, 2-, 3-, and 4-seeded pods and leaf shape for each of the soybean plants, and found that narrow leaf linked with 4-seeded pods. Only one gene for leaf shape has been identified by map-based cloning in soybean -an allele of ln on chromosome 20, encoded by JAG-GED1 (Glyma20g25000.1), the ortholog of Arabidopsis JAGGED (JAG) [41]. JAG1 complements leaf shape and silique length in Arabidopsis mutants [41]. Although leaf traits are important for soybean yield, the molecular link between the two remains unknown. In this study, we show that overexpression of E1 in transgenic plants could affect unifoliolate leaf development and plant development (Fig. 1c-e, Additional file 1: Fig.S1a). Our data thus provide valuable information about the molecular basis of leaf development in soybean.
In plants, cell number and size affect leaf morphology [42][43][44][45]. In rice, the narrow leaf 7 (nal7) mutant has larger but fewer bulliform cells than the wild type  The expression  of TCP6, TCP7, TCP11, TCP13, TCP14, TCP18, TCP19, TCP29, TCP47, and TCP49 in different soybean tissues (flower, leaf, pod, shoot, nodule, cotyledon, seed and root) based on an RNA-seq database. The mean transcription values were visualized by TBtools; red represents high transcript level and green represents low transcript level. l. Expression levels of TCP6, TCP7, TCP11, TCP13, TCP14, TCP18, TCP19, TCP29, TCP47, and TCP49 in different soybean tissues (flower, leaf, pod, shoot, nodule, cotyledon, seed and root) by qRT-PCR. The mean transcription values were visualized by TBtools; green represents high transcript levels and blue represents low transcript levels. The mean expression value was calculated from three independent biological replicates (Haymasari), resulting in reduced leaf width [44]. The loss of function of ADL1 (Adaxialized leaf 1) increases the number of bulliform cells, which leads to a change in leaf shape in rice [42]. Our data showed that cells were more disorganized in E1-OE plants; the numbers of bulliform cells, palisade tissue cells, and spongy tissue cells were significantly higher (Fig. 2a, c-e). Thus, we propose that cell number affects the shape of the leaf. However, the molecular mechanisms underlying this effect require further study in soybean.
E1, a potential B3-like transcription factor, may directly target downstream genes involved in leaf development. We performed RNA-seq and found a significant enrichment in E1-OE transgenic soybean plants of metabolic pathways essential for plant growth and development (Fig. 3d). Therefore, E1 may regulate leaf development through metabolic signaling pathways in soybean. Furthermore, Furthermore, among the TF families previously identified by RNA-seq, such as the AP2/ERF-ERF, bHLH, MYB, WRKY, NAC, HB-HD-ZIP, C2H2, GRAS, bZIP, MYB-related, TCP and B3-ARF TFs (Fig. 4a). CINtype TCP genes in particular appear to regulate leaf development [9,12,13]. For example, CIN-like TCP proteins promote leaf differentiation by regulating the meristematic and auxin response genes in Arabidopsis [46]. In soybean, we found that most CIN-type TCP genes were down-regulated in E1-OE lines. E1 also repressed TCP14 and TCP29 expression in transient expression assay, and our ChIP-qPCR analysis demonstrated that E1 directly binds to the TCP14 and TCP29 promoter (Fig. 5d). Therefore, E1 may directly repress CIN-type TCP genes to regulate the leaf development in soybean.

Conclusions
Based on our data, we determined that the overexpression of E1 could affect leaf development in soybean by directly repressing a large number of leaf developmentrelated CIN-type TCP genes. E1, therefore, regulates leaf development and flowering time. Our findings provide important information into the molecular mechanism underlying leaf development in soybean.

Plasmid construction and soybean transformation
For overexpression and EI-Flag fusion constructs, we amplified the coding sequence (CDS) of E1 using the primer set E1Flag-F/E1Flag-R. The amplicon was inserted into PTF101-3Flag under the control of a CaMV35S promoter. The PCR conditions were as follows: 94 °C for 2 min followed by 30 cycles at 94 °C for30s, 55 °C for 30s, and 72 °C for 30s and a final extension at 72 °C for 10 min. Cotyledonary nodes from DN50 and W82 were used as explants for the Agrobacterium tumefaciens-mediated transformation method described by Paz et al. [47]. Transgenic soybean plants (T 1 ) were identified by PCR amplification and western blot hybridization, and then were advanced to T 7 for further analysis. All primers used for vector construction, PCR, and qRT-PCR assays for target genes are listed in Additional file 4: Table S1.

Materials and growth conditions
DN50 and W82, were obtained from the Innovative Center of Molecular Genetics and Evolution, School of Life Sciences, Guangzhou University, Guangzhou, and subsequently used for transformation and experiments. The E1-OE transgenics and wild types (W82, DN50) were grown in a chamber maintained at 25 °C and 70% relative humidity with a 16 h light/ 8 h dark cycle. Plants were phenotyped 7 days after emergence (DAE).
Nicotiana benthamiana was obtained from the Innovative Center of Molecular Genetics and Evolution, School of Life Sciences, Guangzhou University, Guangzhou, and used for transactivation assays. The N.benthamiana seeds were grown in a chamber maintained at 22 °C and 70% relative humidity with a 12 h light/ 12 h dark cycle. Twenty days after planting, the leaves were used for transient transformation.

Quantitative reverse transcription-PCR (qRT-PCR) analysis
qRT-PCR analysis was performed to determine the transcript abundance of E1. Total RNA was isolated from DN50 soybean leaf, cotyledon, pod, flower, shoot, seed, nodule and root using TRIzol (Invitrogen, Shanghai, China) according to the manufacturer's protocol. cDNA was synthesized using Oligo (dT) 18 primer and the First cDNA transcriptase kit (Takara, Dalian, China). qRT-PCR was performed using a LightCycler 480 SYBR Green I Master (Roche, Mannheim, Germany) in Roche LightCycler480 system (Roche, Mannheim, Germany). The soybean housekeeping gene Tubulin was used as the internal control. The relative transcript level of the target gene was calculated using the 2 −ΔΔCt method. Three biological replicates with three technical replicates each were performed.

RNA-seq analysis
Two independent EI-OE transgenic lines and two DN50 plants grown for 20 days in the greenhouse were used for RNA-sequencing (RNA-seq) analysis. Total RNA was extracted from leaves using the Spectrum Plant Total RNA Kit (Sigma-Aldrich). The RNA was sequenced on an Illumina HiSeq 2500 platform to generate paired-end reads. DEGs between samples were defined by DESeq using two separate models [48], based on fold change > 2 and false discovery rate-adjusted P value < 0.05. DEGs were then subjected to enrichment analysis of Gene Ontology (GO) functions and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways.

Transient expression assay
A transient dual-luciferase assay was performed as previously described [49]. Briefly, p35S:E1-Flag was used in effector constructs and the 2027 bp and 2209 bp promoter sequences of TCP14 and TCP29 were cloned using genespecific primers pTCP14/29luc-F/R and inserted into the pGreen-0800-LUC vector. The reporter construct pTCPs:LUC and the effector constructs p35S:E1-Flag were transformed into A. tumefaciens strain GV3101 and transfected into healthy leaves of 21-d-old N. benthamiana tobacco leaves by agroinfiltration as described previously [26,50]. The plants were placed under continuous white light for 3 d after infiltration, leaf samples were collected for the Dual-Luciferase Reporter Assay System kit (Promega, USA). Relative LUC activity was normalized against the renilla luciferase (REN) activity, and the data presented are the averages of at least three independent replicates.

Protein extraction and immunoblotting
To analyze protein expression in transgenics, total proteins were extracted according to the protocol of Cheng et al. [26]. Total proteins were transferred to polyvinylidene difluoride membranes (Millipore, Germany) and probed using anti-Flag antibodies (Sigma).

Chromatin immunoprecipitation-qPCR assays
For chromatin immunoprecipitation (ChIP)-qPCR assays, DN50 and E1-OE transgenic lines were subjected to chromatin extraction and immunoprecipitation as described by Saleh et al. [51]. The precipitated DNA was recovered and analyzed by qRT-PCR with LightCycler 480 SYBR Green I Master (Roche, Mannheim, Germany). The precipitated input DNA samples were analyzed by qPCR using gene-specific primers. The data were normalized to input transcript levels and the means represent three biological replicates.

Tissues expression profile analysis
The expression data for TCP genes in different tissues, including leaf, shoot, root, flower, seed, pod, cotyledon, and nodule, were available in the RNA-seq database [52]. TBtools [53] was used to display the expression profile of TCP genes in the heatmap.