Knockdown of GmD53a confers strigolactones mediated rhizobia interaction and promotes nodulation in soybean

Background Strigolactones (SLs) play a key role in modulating plant root growth, shoot branching, and plant-symbiont interaction. However, despite their significance, the components of SL biosynthesis and signaling in soybean and their role in soybean-rhizobia interaction is unknown. Methods In this study, we identified and functionally characterized the GmD53a from soybean. The GmD53a ORFs were amplified from root cDNA using primers for GmD53a RNA interference. To induce transgenic hairy roots of soybean, electric shock was used to transform pB7WG1WG2 vectors containing GmD53a knockdown and GUS into K599 strains of Agrobacterium rhizogenes. The hairy roots and nodules were collected and examined for root nodules ratio and RNA was extracted after 4 weeks of rhizobia inoculation. Results A tissue-specific expression assay showed that GmD53a was differentially expressed in plant parts, predominantly in the stem and nodule. Furthermore, its expression was significantly up-regulated during rhizobia infection and varied with nodule formation. The GmD53a-knockdown chimerical plants were produced to further check its role in soybean nodulation in comparison with control GUS. In knockdown lines, the GmD53a (suppressor of strigolactone MAX2) has a higher number of nodules compared to control lines. Furthermore, the expression levels of several nodulation genes essential for initiation and formation of nodules were altered in GmD53a-knockdown lines. Conclusion The results revealed that SL biosynthesis and signaling are not conserved but also have close interaction between SL and legume rhizobia.


INTRODUCTION
Strigolactone (SLs), a new class of terpenoid lactones is generated from carotenoids, and it was identified as a component of root secretion for parasitic witchweed germination (Xie, Yoneyama & Yoneyama, 2010). SLs are required for the formation of symbiotic arbuscularmycorrhizal fungi in plants which are effected by phosphorus deprivation (Akiyama, Matsuzaki & Hayashi, 2005;Bouwmeester et al., 2003;Waldie, McCulloch & Leyser, 2014), and has also been involved significantly in rhizobia-legume interaction (Foo et al., 2013;Liu et al., 2013;Liu et al., 2011). Rhizobia-legume interaction also activates defense responses and increases salt tolerance in soybean seedlings (Qu et al., 2016). In addition, under water deficit conditions, inoculation of rhizobia on soybean resulted in an increased number of nodules (Kibido et al., 2020).
The discovery of strigol supported the agricultural community in understanding and developing system to increase nitrogen fixation. Genetically, SLs are associated with numerous shoot mutants, such as Ramosus of pea (Pisum sativum) (Beveridge, Symons & Turnbull, 2000), More Axillary Growth (MAX ) of Arabidopsis thaliana (Booker et al., 2005), Decreased Apical Dominance (DAD) of petunia (Petunia hybrida) (Simons et al., 2007) and Dwarf or High-Tillering Dwarf(D/HTD) of rice (Oryza sativa) (Arite et al., 2007;Lin et al., 2009). The SLs are found to be synthesized from carotenoid pathways employing a carotenoid biosynthesis inhibitor and carotenoid metabolism mutants (Matusova et al., 2005). In addition, rice mutants d17 and d10 in CCD7 and CCD8 were found to have defective mechanisms, and ccd8/rms1 deficiency in SL synthesis was discovered in mutants of pea (Gomez-Roldan et al., 2008;Umehara et al., 2008). The roles of CCD7 and CCD8, MAX1, and P450 cytochrome in the production of SLs have also been discovered. However, strong evidence is required to prove that SLs are made up of carotenoids (Crawford et al., 2010). Lin et al. (2009) also found the Fe-containing protein DWARF27 (D27) in the biosynthesis of SLs. SLs biosynthesis genes D27, CCD7, and CCD8; the three primary biosynthetic enzymes that complete the sequential processes and produce CL as a product (Alder et al., 2012). D27 catalyzes the reversible isomerization of all-Trans-β-carotene at the C-9 position to create 9-cis-β -carotene. First CCD7 utilizes all-Trans-β-carotene as a substrate and generated all-trans-β-10-carotenol, and then the CCD7 cleaves the 9-cis-carotene and converts it to the 9-cis-apo-10-carotenol, which is consumed by CCD8 to make CL (Schwartz, Qin & Loewen, 2004). A P450 mono-oxygenase, encoded by Arabidopsis MAX1 and lotus LBO further converts carlactone into 5-deoxystrigol and other bioactive SLs (Abe et al., 2014;Breakspear et al., 2014). Furthermore, rice D14 or petunia DAD2, α/β-fold hydrolase that can hydrolyze SLs and act as an SL receptor. In the presence of SLs, MAX2/D3 group of F-box proteins binds with D14 for the production of D14/Skp1-Cullin-F-Box (SCF) E3 ubiquitin ligase complex D14-SCFD3/MAX2 (De Saint Germain et al., 2016;Hamiaux et al., 2012;Stirnberg, Furner & Ottoline Leyser, 2007;Yao et al., 2016). Notably, MAX2 and SCF (Skp1-Cullin-F-Box) play a critical role in SL-triggered protein degradation (Zhao et al., 2014). D53 proteins can create D53-D14-SCFD3 protein complexes by interacting with D14 and D3 proteins. These allow the proteasome system to preferentially degrade the ubiquitin D53 protein, and activate the expression of downstream target genes that lead to the regulation of tillering in cereal crops. This means that tillering ability can be influenced by modulating D53 expression, which can alter the SL signal transduction (Lv et al., 2019).
Although D53 protein belongs to the small group or family SMAX1-like (SMXL) and shares similarities with enzymes such as CIp ATPase enzymes, little is known about how D53 protein suppresses SLs signaling (Stanga et al., 2013). SLs are generated in roots and stems and transported upwards through the xylem to higher regions of the plant or extruded into the extracellular spaces (Kohlen et al., 2011). An ATP-binding cassette (ABC) transporter Pleiotropic Drug Resistance1 (PDR1) was identified as a SL exporter (Kretzschmar et al., 2012). Despite significant advances in understanding SL production and signaling, many SL-related events, such as complicated cross-talks or interactions between SLs and other hormones remain unknown.
Recently, it has been found that SL biosynthesis and signal transduction may play a significant role in soybean nodulation (Ahmad et al., 2020;Bennett & Leyser, 2014;Bennett et al., 2016;Rehman et al., 2018). Because the D53 gene is important for tillering regulation, we used RNAi techniques to knock down the homologous soybean gene of D53 (GmD53a), and examined the physicochemical properties and structure of GmD53 protein using bioinformatics tools. Furthermore, the role of GmD53a protein in soybean root nodulation was demonstrated in this study.

Plant growth conditions
Seeds of soybean were surface sterilized with sodium hypochlorite (NaClo) and hydrochloric acid (HCL) and grown in three-gallon pots containing vermiculite soils at National Key Laboratories, Huazhong Agricultural University, Wuhan. The seedlings were grown at 26/20 • C (day/night) temperature, photoperiod of 14/10 h, 800 µmol m −2 s −1 light intensity and 60% humidity was maintained in growth chamber. Seeds, stems, roots, flowers, leaves, of and pods from soybean cultivar (Tianlong no. 1) were harvested at different growth phases. After treatments, all selected tissues were sensibly removed and instantly put in liquid nitrogen and then stored at −80 • C. Then, RNA was extracted and later on cDNA was synthesized according the protocol of supplier.

Construction of vector
The GmD53a (Glyma.11G230700.1) ORFs were amplified from root cDNA by using primers for GmD53a RNA interference (Table S1). Following that, RNA was isolated with TRIzole reagent (Invitrogen) or RNA kit (Biotech, Beijing) from soybean roots, and 10 g of RNA was taken in the cDNA synthesis utilizing the first-strand synthesis technique (Invitrogen) and cDNA was used as a template for the amplification of GmD53a as previously described (Ahmad et al., 2020). The amplified fragments were used for directionally cloning in pDONR221 vector via BP clonase and then recombined into the pB7GWIWG2 destination vector via LR clonase (Life Technologies, Rockville, MD, USA) after ORFs were cloned into T-easy vector (Promega, Madison, WI, USA) and sequenced. For comparison, the GUS gene was also recombined into both vectors. Then the constructs were transformed into A. rhizogenes K599 by electroporation and were utilized to induce hairy roots.

Induction of transgenic hairy root and nodulation assay GmD53a
To induce transgenic hairy roots of soybean, electric shock was used to transform pB7WG1WG2 vectors containing GmD53a knockdown and GUS into K599 strains of Agrobacterium rhizogenes. A. rhizogenes strain K599 harboring pB7GWIWGII-Gm53a, cDNA fragments for knockdown constructs, or GUS were grown on LB-agar medium at 28 • C with spectinomycin and streptomycin as selection markers. The overnight Agrobacterium cultures were used for transformation of soybean cultivar. Seven days old seedlings of soybean were wounded at the hypocotyls region before being incubated in high humidity for 24 h carrying A. rhizogenes constructs (Ahmad et al., 2020). Hairy roots were emerged from the wounding sites after one week of infection, whereas non-transgenic roots were removed after one week. About 1 week after hairy root emergence when transgenic hairy roots were about to support the plants, the chimeric soybean plants were examined for transgene expression in hairy roots, and the main of non-transgenic roots are removed, before being inoculated with rhizobia strain USDA110 cultured in the YMA at 28 • C. About 25 mL rhizobia bacteria solution (OD 600 nm) was applied to each plant pot. The hairy roots and nodules were examined and collected for root nodules ratio, and RNA was extracted after 4 weeks of inoculation. Three independent experiments of transformation with 10 individual transgenic lines under equal treatments and growth conditions were carried out for each vector including GUS as a control. We determined the numbers of nodules from hairy roots per grams for normalization.

Rhizobia infection
For the rhizobia infection experiment, seedlings of about 14 days old were inoculated with ''USDA-110'' (O.D 600 = 0.08-0.1) strain of Bradyrhizobium japonicum in the YMA medium. Roots of infected plants were collected at 0, 6, 12, 24, 36, and 48 h after inoculation. These frozen tissues were used the extraction of RNA to synthesize cDNA.

QRT-PCR analysis of gene expression
Total RNA was isolated from several tissues i.e., seeds, leaf, stem, flower, root, and nodules), or transgenic hairy roots by using TRIzol (Invitrogen, Carlsbad, CA, USA) or RNA kit (Biotech, Beijing) according to instructions from the manufacturer. About 10 g of total RNA was digested with RNase-free DNaseI (Promega, Madison, WI, USA) to remove any genomic DNA contamination for each sample. A NanoDrop ND-2000 UV spectrophotometer (Thermo Scientific, Waltham, MA, USA) was used to measure the concentration of RNA. The first strand cDNA was prepared from 10 µg total RNA using the Superscript III first strand synthesis system (Invitrogen). All of cDNA samples were diluted with sterile water according to the requirements for qRT-PCR analysis where gene-specific primers were used (Table S1). The reaction mixture of 20 µL containing 2.5 µL SYBR Master Mix (Applied Biosystems), 1 µL primer mix (0.4 µL from each primer, 0.2 µL ddH2O), and 2 µL of 100 ng cDNA were put in 96-well plates (iQ5 Real Time PCR System; Bio-Rad) for all of selected tissues. The transcript levels were normalized by using GmACTIN1 as previously described (Ahmad et al., 2020).

Bioinformatics analysis
Protein sequences of SLs signaling genes reported from monocots and dicots were obtained from NCBI or phytozome database (phytozome.jgi.doe.gov/) for phylogenetic analysis and identification of SLs signaling genes. The phylogenetic tree was constructed by MEGA6, and Neighbor-joining was performed with 1,000 bootstraps (Tamura et al., 2013). Bioinformatics analyses on protein sequences of SL pathway genes were obtained from NCBI (https://www.ncbi.nlm.nih.gov/) and phytozome (phytozome.jgi.doe.gov/). The sequence similarities and identities were determined from Blast2 (http://pga2.mgh. harvard.edu:8080/rtpcr/blast/wblast2_cs.html). The RNA Seq-Atlas data from nine soybean tissues were extracted from public databases (http://soybase.org/soyseq/) to acquire the tissue-specific expression and stages of nodule development.

Statistical analysis
The data were recorded from three independent experiments for each trait, and Student's t -test was applied for analysis using Statistix 8.1 software. The confidence limit 95% represents the significant between two-tailed data (Rehman et al., 2018).

Identification of SL biosynthetic genes from soybean genome
The homologs of other organisms with 1 × 10 −3 permissive E cut-off values were discovered in tBLASTp searches to assess the evolutionary relationship of GmD53 with other plant species. A neighbor-joining phylogenetic tree was generated by using the D53 proteins from 12 dicotyledons and three monocotyledons by using MEGA6.0 software to compare the evolutionary relationships between GmD53a and other plant D53 proteins (Fig. 1A). The phylogenetic analysis revealed that all D53 proteins were divided into two main groups, i.e., dicotyledonous D53 proteins and monocotyledonous D53 proteins, indicating that dicotyledon D53 proteins were highly homologous. This group included all of dicotyledon D53 proteins which were separated into three sub-groups. One sub-group had D53 proteins from Orchidaceae, Palmae, and Musaceae, while other sub-groups were composed of D53 proteins from Gramineae. Thus, D53 proteins from Gramineous are closely related, and ScD53 protein was closely related to sorghum (Sorghum bicolor) D53 protein followed by maize (Zea mays) and millet (Panicum miliaceum). Furthermore, homology analysis by using Blastp program in NCBI showed that ScD53 had 93%, 82%, 80%, and 69% identity with D53 proteins of sorghum, maize, foxtail (Setaria italic), and rice (Oryza sativa), respectively. Besides, GmD53a was located on chromosome number 11, and exon/intron distribution analysis showed that GmD53a contains four exons and three introns with varying lengths (Fig. 1B). Further analysis of GmD53a protein sequence alignment revealed that all the 17 members were delimited by CLP-R and ATPase-AAA domain (Fig. 1C), suggesting that GmD53a is evolutionarily conserved in plant species.

Tissue-specific and altered expression in nodule stages of GmD53a
The expression patterns of the strigolactone repressor GmD53a gene were assessed in various tissues and organs. We performed a BLAST search of the rice D53 gene in Phytozome and obtained two close homologs GmD53a and GmD53b. We have compared gene expression patterns to public databases (http://soybase.org/soyseq/) GmD53a (Figs. 2: S1) to understand the role of these genes. Data on the expression levels of genes implicated in strigolactone signaling in nine different plant tissues (pod, root hair, root, nodule, stem, leaf, seed, shoot apical meristem (sam), stem, and flower) were studied. In soybean plants, we explored how the GmD53a gene was expressed in different tissues and organs. GmD53a was most abundant in the stem, pods, nodules, and leaves, while GmD53b was abundant in the stem, Shoot Apical Meristem (SAM), and leaves (Fig. S1). However, low expression was observed in the remaining tissues (Figs. 2, S1). GmD53a was more abundant in most of tissues and organs, particularly roots and nodules, than GmD53b (Fig. S1).
In addition, we also examined the expression of GmD53a gene, which is involved in SL signaling, throughout nodule development and senescence. We evaluated the RNA-Seq data from soybean nodules at various phases of development and validated it with qRT-PCR. The results revealed that GmD53a was differently expressed in nodules at various stages of development till nodule senescence (Fig. S2). The data were validated through qRT-PCR which was consistent with the RNA-Seq results. Interestingly, the transcript levels of GmD53a increased significantly during growth stages of nodule formation, and it was maximum at stage 2, and then declined afterward (Fig. 2).

Effect of GmD53a-knockdown on soybean nodulation
The knockdown chimerical soybean plants known as GmD53a-KD were developed to investigate the function of SLs Dwarf gene GmD53a in soybean nodulation, (Fig. 3A). The plants were infected with GmD53a containing RNAi vector and were collected after four weeks of rhizobium inoculation. The ratio of hairy roots to nodules was examined, and it was found that GmD53a had substantially more nodules per gram of fresh-weight hairy roots than GUS control (Fig. 3B). GmD53a-KD plants produced 60% more nodules than GUS control plants (P < 0.01). This change in number of nodules in GmD53a-KD plants was seen in three separate trials. However, the total size of mature root system did not differ among knockdown and GUS control plants, unlike the number of nodules on mature roots. The qRT-PCR was used to analyze the genetic background of GmD53a-KD, which revealed the considerable repression of GmD53a-KD plants (P < 0.05) (Fig. 3C). The number of nodules on GmD53a-KD plants differed significantly from the control (GUS) plants (Fig. 3). As a result, GmD53a in SL signaling pathway acted as an antagonist in soybean nodulation.

Response to rhizobia infection of GmD53a-KD hairy roots
Apart from known activities in plant-mycorrhizal interactions, SLs have been reported to alter legume nodulation (Foo & Davies, 2011). The GmD53a was expressed after infection with B. japonicum (USDA110). To further explore how SL operates in soybean nodulation and expression of SL suppressor GmD53a in response to B. japonica infection, the GmD53a-KD and GUS control chimerical hairy roots were infected with rhizobia strain USDA110, and expression was assessed at various intervals after infection (Fig. 4). The results showed that GmD53a expression in hairy roots of GUS controls was reduced at 6 h and continued to decline significantly at 12 and 24 h (Fig. 4). Similarly, GmD53a expression was maximum at 36 h and it was declined at 48 h (Fig. 4). The expression level of GmD53a was increased considerably at every time point and peaked at 48 h as compared to GUS control (Fig. 4).

DISCUSSION
Soybeans are one of the important economic crops, and popularity is increasingly due to their role in nitrogen-fixing capabilities in animal feeds, the food sector, and sustainable agriculture. Understanding the role of SLs in soybeans is critical for the development of new accessions of soybean because the cited gene plays crucial roles in controlling the plant architecture, shoot branching, root growth, plant-mycorhization, and legume-rhizobium interactions. While the literature on studies on SLs in soybeans is scanty because these genes have not been investigated thoroughly by many researchers. Therefore, this research work was planned to investigate the role of SL signaling genes in the development, nodulation and architecture of soybean. Unlike diploid model legumes such as Medicago and Lotus, which have only one copy of homolog of the SL biosynthesis gene, while the tetraploid soybean genome has many copies of each MAX or Dwarf gene. Orthologs from pea, Arabidopsis, and rice are similar to each other. SLs are typically synthesized in roots and stems and transported upward to shoots A B Figure 5 GmD53a-KD altered expression levels of nodulation genes. Expression level of nodulation genes in GmD53a-KD and GUS transgenic hairy roots as control. Gene expression was determined by qRT-PCR with GmACTIN as an internal control. Data are expressed as means ± s.d. from at least 3 independent experiments with biological replicates. Differences were analyzed, *p < 0.05; **p < 0.01 in student's t -test. Full-size DOI: 10.7717/peerj.12815/fig-5 and leaves in other plants (Kohlen et al., 2011;Yoneyama et al., 2008). SLs biosynthesis genes have higher expression in roots and stems, but SL signaling genes can be expressed anywhere in plants (Kohlen et al., 2011;Yoneyama et al., 2008). Indeed, the expression of GmD53a in soybean confirms the conserved SL signaling components. In addition, rhizobia infections have significantly altered the expression of SL suppressor GmD53a. According to another study on pea, SLs play a key role in the development of infection threads after rhizobia infection by influencing early nodulation gene expression (McAdam et al., 2017). SLs are linked to legume nodulation, in addition to their physiological effects on root growth, shoot branching, and mycorrhizal branching (Foo, 2013;Foo & Davies, 2011;Foo et al., 2013). The pea SLs-deficient mutant rms1/CCD8 produces 40% fewer nodules than the wild type, although synthetic SLs analog GR24 partially reversed the phenotype (Foo & Davies, 2011). SLs were reduced by 80% in Lotus japonicus LjCCD7 -silenced plants, and nodules were reduced by 20% when compared to control plants (Liu et al., 2013;Liu et al., 2011). In GmD53a-KD soybean roots, the structural grounds for nodule initiation, development, and rhizobia interaction were significantly changed. As previously stated, root hair development and immediate responses of root hairs to rhizobial Nod factors are crucial for nodulation (Oldroyd, 2013). We also found that knocking down of GmD53a, a SL suppressor, boosted soybean nodulation. GmD53a-KD hairy roots had a 60% (P < 0.01) increase in nodule quantity compared to GUS, implying that GmD53a-KD influenced nodulation in chimerical transgenic hairy roots compared to control. The gene expression analysis revealed that GmD53a knockdown hairy roots also altered the expression of several key genes which are involved in nodulation. Similar to CCD7 and CCD8 in pea and Medicago (Liu et al., 2013;Liu et al., 2011;Van Zeijl et al., 2015), over-expression of GmMAX1a, 2a, 3a, and 4a were also involved in nodulation initiation and development as evidenced by genetic and molecular data (Ahmad et al., 2020);ur (Rehman et al., 2018). On the other hand, soybean GmD53a-KD had other opposite effects on nodulation in soybean. It was previously established that increase in the expression of D10 in d53 mutant explains a feedback control of the SL pathway (Zhou et al., 2013). The same researchers have also reported that that in d53 mutant, the expression of an axillary bud inhibitor FINE CULM 1 (FC1), which is an orthologue of TEOSINTE BRANCHED1 (TB1) and Arabidopsis BRANCHED 1 (BRC1) was down-regulated, implying that D53 is involved in SL biosynthesis or signaling (Zhou et al., 2013). Likewise, the accumulation of SLs in the roots of d53 mutant was significantly higher than compared to WT (Zhou et al., 2013). The increase in number of nodules in GmD53a-KD describes that SLs level might be higher in roots of GmD53a-KD compared to GUS control.
In particular, the d53 mutant plants had more number of tillers than the wild type (Zhou et al., 2013). The expression level in transgenics was linked with severity of tillering phenotype. In the same way, over-expression of D53 gene resulted in a slight increase in tillering compared to control plants (Zhou et al., 2013). These findings strongly suggest that D53 protein functions as a suppressor of the SL-mediated branching inhibition pathway, and the more tillering phenotype of d53 mutant was generated by a gain of function mutation in d53. RNA interference (RNAi) was used to develop D53 knockdown transgenic plants; as expected, the number of tillers was reduced (Zhou et al., 2013). These findings support the hypothesis that d53 mutation increases the D53 activity in repressing SL signaling (Zhou et al., 2013). Notably, a decreased expression of D53 significantly reduced the tiller number of d3 and d14 mutant lines, respectively, indicating that the reduction of D53 expression could partially rescue the signaling defects of d3 and d14 and the result is consistent with its negative role in SL signaling (Jiang et al., 2013). It also validated the increased nodule number in GmD53a-KD lines compared to GUS control. Downstream factors such as nodulin genes, and ENOD40 are connected to NFR1α genes which are involved in Nod factor perception and involved in Nod factor signal transductions, namely, DMI2α and DMI3β, NIN α, and NSP2β (Oldroyd, Engstrom & Long, 2001;Oldroyd, 2013). Our findings showed that GmD53a knockdown affects the nodulation signaling pathway.

CONCLUSION
In this study, the soybean genome was exploited to find closely related homologs of SMXL6, 7, 8 from Arabidopsis, and D53 from rice to further figure out whether and how soybean SL signaling biosynthesis is involved in controlling the various growth and developmental phases in the soybean plant. The results showed that GmD53a is involved in the nodulation of soybean. This study not only demonstrated that SL signaling is conserved across the reported crop plants but also revealed that how GmD53a controls the developmental phases and legume-rhizobia interaction. However, further investigating the roles of these signal components may provide potential targets for directional improvement of crop traits, and further research that focuses on these aspects is highly desired.

ADDITIONAL INFORMATION AND DECLARATIONS Funding
This work was supported by Taif University Researchers Supporting Project number (TURSP-2020/123), Taif University, Taif, Saudi Arabia. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
• Ihteram Ullah conceived and designed the experiments, performed the experiments, analyzed the data, authored or reviewed drafts of the paper, and approved the final draft.
• Rana waseem Akhtar, Ahmad El Askary and Amany Salah Khalifa analyzed the data, authored or reviewed drafts of the paper, and approved the final draft.
• Muhammad Imran analyzed the data, prepared figures and/or tables, and approved the final draft.
• Arwa Abdulkreem AL-Huqail and Muhammad Tehseen Azhar analyzed the data, prepared figures and/or tables, authored or reviewed drafts of the paper, and approved the final draft.

Data Availability
The following information was supplied regarding data availability: The raw data is available in the Supplemental File.

Supplemental Information
Supplemental information for this article can be found online at http://dx.doi.org/10.7717/ peerj.12815#supplemental-information.