CRISPR/Cas9-mediated targeted mutagenesis of GmLCL genes alters plant height and internode length in soybean


 Background: Soybean ( Glycine max ) is an important economically crops for plant oil and protein in the world. The plant height as a key trait has significant effects on yield of soybean , however, the research on molecular mechanism for soybean plant height is still unclear. Recently, CRISPR (clustered regularly interspaced short palindromic repeat)/Cas9 (CRISPR-associated) system as a new technology for gene editing, has been rapidly utilized to edit the genomes of crop plants. Results: Here, we designed four gRNAs to mutate four LATE ELONGATED HYPOCOTYL ( LHY ) / CIRCADIAN CLOCK ASSOCIATED1 ( CCA1 )- LIKE ( LCL ) genes in soybean. In order to test whether the gRNAs could perform properly in transgenic soybean plants, we first tested the CRISPR construct in transgenic soybean hairy roots using Agrobacterium rhizogenesis strain K599. Once confirmed, we performed stable soybean transformation and obtained nineteen independent transgenic soybean plants. Subsequently, we obtained one T 1 transgene-free homozygous quadruple mutant of GmLCL by self-crossed. The phenotype of T 2 -generation transgene-free quadruple mutant plants were observed and the results showed that quadruple mutant of GmLCL displayed reduced plant height and shortened internodes. In addition, the relative expression levels of gibberellic acid (GA) metabolic pathway genes in the quadruple mutant of GmLCL were significantly decreased than wild type (WT). It suggests that GmLCLs encoding MYB transcription factor affect plant height through mediating the GA pathway in soybean. We also develop some genetic markers to identify mutant for assisting breeding studies. Conclusions: Our results indicate that CRISPR/Cas9-mediated targeted mutagenesis of four GmLCL genes reduce soybean plant height and shorten internodes. These findings suggest that manipulation of four GmLCL genes may improve yield by altered plant height and internode length in soybean.

4 findings suggest that manipulation of four GmLCL genes may improve yield by altered plant height and internode length in soybean.

Background
Soybean is one of the most important economically crops for plant oil and protein resources worldwide, and plant height, node number, internode length, branch number, seed size are important factors that affect the yield [1,2]. Plant height is a key trait of plant ideotypes, and a relatively shorter stem length contributes to obtaining higher yield in modern breeding programs [3,4,5]. Therefore, there are some plant height genes have been cloned by Map-based cloning in several plant species, such as maize [6,7,8], rice [9,10,11], tomato [12] and soybean [13,14].
For example, GA3 b-hydroxylase ( ZmGA3ox2) was cloned by candidate gene association mapping and genetic assay from the dwarf mutant d1-6016, and responded for the dwarf mutant in maize [7]. Xing et al have been cloned the Brachytic2 (Br2) gene from maize by mapping, which can significant effect plant height [8]. Recently research showed that GmDW1 (dwarf mutant) encodes an entkaurene synthase, and the mutant of GmDW1 displayed reduced plant height and shortened internodes in soybean [13]. In addition, a few transcription factors (TFs) families play important role in plant height. For instance, OsNAC2 is a NAC transcription factor, and constitutively expressed OsNAC2 had shorter internodes and shorter spikelet in rice [15].
Circadian clocks are endogenous 24 h oscillators that allow organisms to anticipate daily changing environmental conditions, and play critical roles in participating many biological processes and stress responses through the regulation of up to 80% of the transcriptome in plants [16,17,18]. LHY and CCA1 are key components of 5 the central oscillator, which can encode two morning-expressed MYB transcription factors in Arabidopsis [19,20]. AtLHY/CCA1 can bind to the evening element (EE; AAATATCT) of the promoter of TIMING OF CAB EXPRESSION 1 (TOC1) and act redundantly to repress the transcription of the AtTOC1 gene during the day [21].
AtTOC1 represses AtCCA1 and AtLHY from its induction at dusk until slightly before dawn [22]. Moreover, the other functions of LHY/CCA1 for flowering, stress response have been reported [23,24]. For example, silencing NaLHY abolished the vertical movement of flowers under continuous light conditions in Nicotiana [23]. Recently report showed that AtLHY can regulate the expression of ABA signalling components and downstream response genes to potentiate some ABA responses [24]. However, the potential functions of the LHY/CCA1 family members in soybean are still unclear.
In recent years, CRISPR/Cas9 systems have been engineered to create genetic manipulation in plants [25,26,27,28]. The use of CRISPR/Cas9 technology has attracted large attention and successfully applied in various crops to induce genome editing, such as wheat [29,30], maize [31,32], rice [33], barley [34], tomato [35,36] and soybean [37,38,39]. There are four LHY/CCA1-LIKE ( LCL) genes in soybean, named GmLCL1, GmLCL2, GmLCL3 and GmLCL4, but the functions of these genes are still unknow. Therefore, in the current study, the CRISPR/Cas9 system was used to target four GmLCL genes in soybean. We observed phenotype of T 2 -generation transgene-free quadruple mutant of GmLCL and found that the quadruple mutant plant height and internodes were significantly shorter than WT. Moreover, the relative expression levels of GA metabolic pathway genes in quadruple mutant of GmLCL were significantly lower than WT. These results suggested that GmLCLs were involved in the regulation of plant height directly or indirectly mediating key components of the GA pathway. We also developed some genetic markers to 6 identify mutant for assisting breeding studies. Our findings show that manipulation of these genes should facilitate improvements plant height and internodes in soybean.

Target site selection, construction and confirmation of target sites in soybean hairy roots
In order to identify the ortholog of AtLHY and AtCCA1 in soybean, we performed protein sequence alignment and identified four CCA1/LHY orthologs, GmLCL1 (Glyma.03G261800),, GmLCL2 (Glyma.19G260900),, GmLCL3 (Glyma.16G017400),, and GmLCL4 (Glyma.07G048500) in soybean. Phylogenetic analysis showed that GmLCL proteins closer to AtLHY than AtCCA1 ( Figure S1). To study their function of the four GmLCL genes in soybean, four target adaptors, target 1/2 were selected for targeting GmLCL1 and GmLCL2 genes, target 3/4 were selected for targeting GmLCL3 and GmLCL4 genes (Fig.1A). The target 1 in the second and third exon of GmLCL1 and GmLCL2 genes respectively, target 2 in the fifth and sixth exon of GmLCL1 and GmLCL2 genes respectively, and the target 3 in the first exon of GmLCL3 and GmLCL4, the target 4 in the fifth exon of GmLCL3 and GmLCL4 in soybean (Fig. 1A). The CRISPR vector used encodes Cas9 driven by the CaMV35S promoter and four gRNAs driven by the Arabidopsis U3b, U3d, U6-1 and U6-29 promoter, respectively ( Fig. 1B, C).
In order to test whether the CRISPR/Cas9 construct could edit properly these genes in transgenic soybean plants, we first tested the construct in transgenic soybean hairy roots usingA. rhizogenes K599 ( Figure S2A). The transgenic soybean hairy roots were generated by high-efficiency Agrobacterium rhizogenes-mediated transformation [40]. When the hairy roots generated at the infection site were approximately 2 cm long, the hairy roots were used for genotype detection. The genotype of transgenic hairy roots was detected by PCR using Cas9 gene-specific primers and GmLCL genes-specific primers. We detected mobility-shifted bands in six DNA bulked samples when specific primers of Cas9 gene-specific were utilized.
The result showed that five transgenic lines with the Cas9 gene product (Cas9 genepositive) ( Figure S2B). Sequencing analysis of GmLCL genes showed that the Cas9 gene-positive lines (R1-R5) produced superimposed peaks in target1/3 site, while the target 2/4 site no changed ( Figure S2C, Table S1). Together, these results indicated that the transgene-encoded Cas9 and gRNAs were able to efficiently induce double-strand breaks at target 1/3 sites in GmLCL genes.

The expression level of the GmLCLs in quadruple mutant and WT
LHY/CCA1 are key components of the circadian clock, which participate the temporal organization of biological activities and the regulation of gene expression [16,17,21]. Previous studies had shown the expression level of LHY/CCA1 were much higher in the morning than in the night [21]. However, the expression pattern of GmLCL

The quadruple mutant of GmLCL reduce soybean plant height and shorten internode
To examine the loss function of GmLCLs, the phenotype of T 2 -generation transgenefree quadruple mutant and WT plants were observed. We found that the plant height of quadruple mutant was significantly lower than WT under LD conditions for 20 days after emergence (DAE) (Fig. 4A, B). Subsequently, we examined that the node number andinternodal length, which could affect plant height [13,15]. As showed in
Compared with the WT plants these genes showed significantly decreased expression in quadruple mutant of GmLCL (Fig. 5A-F). These results suggested that GmLCLs might positively regulate the expression of these GA biosynthesis and GA responsive genes, thereby limiting soybean plant height.

Development of genetic markers and inheritance of quadruple mutant alleles
Genetic markers are critical and effective method to identify mutant alleles for molecular-assisted studies, and the genetic markers could accelerate genotyping procedure in future generations [38]. Therefore, we developed three dCAPs (Derived Cleaved Amplified Polymorphic Sequences) markers to identify the GmLCLs mutant alleles (Fig. 6A). To identify the genotyping of GmLCL mutants, PCR amplifications were performed using GmLCLs-specific and dCAPs-specific primer pairs, respectively. The amplified products of GmLCL1, GmLCL2 and GmLCL4 on mutant genomic DNA templates could be cleaved by restriction endonuclease MspI, but not on WT genomic DNA templates (Fig. 6B). Meanwhile, the amplified products of GmLCL3 on mutant genomic DNA templates could be cleaved by restriction endonuclease RspRSII, but not on WT genomic DNA templates (Fig. 6B). These results showed that three dCAPs markers of GmLCLs could be used to identify the genotyping of GmLCL mutants and used for molecular breeding studies.

Discussion
Recently, CRISPR/Cas9 system has been rapidly and widely used to edit the genomes of various crops, such as soybean [37,38,39]. For example, Bao et al obtained GmSPL9 genes mutant through CRISPR/Cas9 system and stable soybean transformation, and they found that the mutant of GmSPL9sincreased node number on the main stem and branch number, consequently increased total node number per plants [38]. The CRISPR-edited soybean plants of both GmFAD2-1A and GmFAD2-1B genes showed dramatic increases in oleic acid content to over 80%, whereas linoleic acid decreased to 1.3-1.7% [39]. LHYand CCA1 are important circadian clock genes, which can encode two morning-expressed MYB transcription factors in Arabidopsis [19,20]. However, the functions of the LHY/CCA1 family members in soybean are still unknow. In this study, we designed four target adaptors (target 1, target 2, target 3 and target 4) to edit four GmLCL genes ( Fig.1A-C). In order to test whether the targets could perform properly in transgenic soybean plants, we firstly tested the CRISPR construct in transgenic soybean hairy roots using Agrobacterium rhizogenesis strain K599. We confirmed that target 1 and target 3 could perform, while target 2 and target 4 might not work properly ( Figure   S1). Then, we performed stable soybean transformation and obtained nineteen T 0 events. In previous CRISPR/Cas9 research, the production of chimeric mutations was a problem for reduced the heritable transmission of mutant alleles in soybean [43,44]. Therefore, in this study, we sought homozygous quadruple mutant of GmLCL lines without transgenes and screened T 1 plants derived from the T 0 transgenic lines. Luckily, we obtained one (T 1 -15) transgene-freehomozygous quadruple mutant of GmLCL (Fig. 2C-F; Table S2). Our findings showed that the CRISPR/Cas9 system offered great potential in soybean breeding.
The circadian clock plays a critical role in timing multiple biological processes and stress responses in some model crops [16,17,18]. As key components of the circadian clock, the LHY/CCA1 transcription factors have ability to initiate and set the phase of clock-controlled rhythms to produce some phenotype [16,23,24,45,46]. For example, the overexpression of NaLHY had elongated hypocotyls and flowered late compared with WT plants in N. attenuate [23]. The same phenotypes were found in Arabidopsis AtLHY-overexpressing lines [45,46]. Although the functions of LHY/CCA1 were shown to be involved in flowering and stress responses in model crops, little is known about the biological functions of LHY/CCA1 family members in soybean. To explore the molecular function of genes in soybean, we examined the phenotype of the loss function of GmLCLs in T 2 transgene-free mutant.
We found that the mutant shortens plant height in soybean at 20 to 35 DAE (Fig 4A-E). Our data demonstrated that the clock genes GmLCLs, as MYB transcription factors, function on regulating plant height in soybean.
It is generally known that plant height is a central yield trait for breeding in various crops [3,4,5]. GAs as a large group of tetracyclic diterpenoid plant hormones, regulate diverse biological processes in plant growth and development, such as the embryo, leaf primordia, flowering and plant height [47,48,49]. In recent years, a few GA metabolic pathway-related genes of plant height association have been reported in plants [13,14]. For example, SD1 encoding a gibberellin 20-oxidase gene (GA20oxs), and the sd1 mutantreduced endogenous GA levels and led to the short stature ofrice variety IR8 [49,50]. However, the research on molecular mechanism for transcription factors to regulate plant height in soybean is still unclear. In present study, we tested the expression levels of GA synthetic genes (GmDW1, GmGA1, GmGA2, GmCPS2) and GA response-related genes (GmGR2, GmGR8) in quadruple mutant of GmLCL and WT soybean plants (Fig. 5A-F), and determined that these genes substantially decreased expression in quadruple mutant of GmLCL. Overall, we speculat that GmLCLs might positively regulate the expression of these GA metabolic pathway-related genes to reduce soybean plant height.

Conclusions
The CRISPR/Cas9 system can be used for multiplex gene editing to advance crop plant breeding. In present study, we used CRISPR/Cas9-based multiple genome editing, and successfully obtained quadruple mutant of GmLCL in soybean. Further, our results suggest that GmLCLs directly or indirectly improve the expression level of GA synthetic genes and GA response-related genes to regulate soybean plant height. Our findings offer efficiency by which gene editing was applied to generate non-transgenic soybean genotypes with an important yield trait and provide an insight into the mechanism underlying the plant height regulatory networks in crop plant species.
The pYLCRISPR/ Cas9P35S-B vector was gift from Ma et al [51]. The target sequences were subcloned into the different single guide RNA (sgRNA) expression cassettes and built into the pYLCRISPR/ Cas9P35S-B vector according to the protocol reported by Ma et al [51]. The positive plasmids were introduced into Agrobacterium tumefacien strain EHA101 for soybean stable transformation, and were introduced intoAgrobacterium rhizogenes strainK599 for soybean hairy roots transformation.

Stable soybean transformation
The transformation procedure was used according to the previous protocol [52,53].  (Table S3). RT-PCR amplifications were done once for each DNA sample.

Agrobacterium rhizogenes-mediated transformation of soybean hairy roots
Transgenic soybean hairy roots were generated by A. rhizogenes-mediated transformation as described by Kereszt et al. [40] and Cheng et al. [54] Table S3. qRT-PCR analysis Total RNA was isolated from WT and T 2 mutant soybean leaves using Trizol reagent (Invitrogen, Shanghai, China). cDNA synthesis was conducted using an M-MLV 15 reverse transcriptase kit (Takara, Dalian, China) according to the manufacturer's instructions. qRT-PCR analysis was performed to measure GmLCL genes, GmGA1, GmGA2, GmCPS2, GmGR2, GmGR8, GmDW1 transcript levels on a Roche LightCycler480 system (Roche, Germany) using a real-time PCR kit (Roche, Germany). The soybean housekeeping gene GmTubllin ( Glyma.05G157300) was used as an internal reference to normalize all data. The relative transcript level of the target gene was calculated using the 2 − ΔΔCT method. Three biological replications per line were performed in each test.

Molecular marker development
GmLCLs sequences of the Harosoy and mutant genome were obtained by sequencing. Primers were designed using Primer Premier 5.0, with a product size <200 bp. Three dCAPs marker was developed on the basis of variations in the target 1/3 site of the GmLCL genes. GmLCL1 and GmLCL2 shared a pair of makers, GmLCL3 and GmLCL4 used a pair of makers, respectively. Table S3 lists the dCAPs markers that were used for this study.
Additional files Figure S1. Phylogenetic tree of LHY and CCA1 from Arabidopsis and soybean.
Phylogenetic tree was inferred using the Neighbor-Joining method. The bootstrap consensus tree inferred from 1000 replicates is taken to represent the history of different LHY/CCA1 proteins analyzed.  Gel electrophoresis of PCR amplicons using specific primers for CRISPR/Cas9 vector.
B-E. The fragments containing the edited sites were amplified by PCR, and directly sequenced. The sequencing chromatograms with superimposed peaks derived from biallelic mutations of the targeted sites were decoded by DSD ecode program [51].
The red frames indicate the location of targets. Table S1. CRISPR/Cas9-meditated targeted mutagenesis of four GmLCL genes in transgenic soybean hairy roots.    The phenotype of the WT plants and T 2 homozygous quadruple mutant of GmLCL A The plan 30 Figure 5 The relative expression of GA metabolic pathway related genes in quadruple mutant of GmLC 31 Figure 6 Inheritance and segregation of GmLCL genes small deletions. A. An example of dCAPs specifi