Agrobacterium rhizogenes

The human codon-optimized Cas9 gene [2 (link)] was obtained from Addgene (plasmid 41815). Two flanking primers with added NheI and SacII sites were used to amplify the coding sequence, including the SV40 nuclear localization signal, with the KAPA HiFi polymerase (KAPA BioSystems). The amplicon was digested with the two restriction enzymes and ligated to the vector, pM35S, between the double-enhancer 35S promoter and nopaline synthase (nos) terminator (Additional file 6). The entire cassette is flanked with I-SceI restriction sites, which were used to move the Cas9 cassette into p201N to create p201N:Cas9 (Addgene plasmid 59175). The p201N vector is a p201BK [37 (link)] vector modified to include an nptII selectable marker cassette and I-SceI and I-PpoI restriction sites (Additional file 6).
For biolistic transformation of soybean, a pSMART HC Kan (Lucigen Corporation, [GenBank: AF532107]) cloning vector was modified to contain a hygromycin phosphotransferase (hph) gene under the control of the Solanum tuberosum Ubi3 promoter and terminator [38 (link)] and the meganuclease I-PpoI site, and is referred to as pSPH2. The vector pSPH2 was digested with I-PpoI and DNA overhangs were removed with T4 DNA polymerase. To prepare the Cas9 insert, p201N:Cas9:gRNA-Glyma07g14530 was digested with SpeI and PmeI and DNA overhangs were removed with T4 DNA polymerase. The vector and insert were ligated to create the plasmid pSPH2:Cas9:gRNA-Glyma07g14530. The Glyma07g14530 gRNA was then replaced with the 01g + 11gDDM1 (Glyma01g38150 and Glyma11g07220) gRNA via I-PpoI to produce pSPH2:Cas9:gRNA-01g + 11gDDM1.
Additional binary Cas9 vectors were produced by replacing nptII from p201NCas9, with hph, bar (phosphinothricin resistance), or GFP. The hph cassette was moved from pSPH2 into the p201N Cas9 vector with the PacI and SpeI restriction sites to produce p201H:Cas9 (Addgene plasmid 59176). The bar and GFP cassettes (double-enhancer 35S promoter, nos terminator) were amplified with the SpeI 35SF and PacI nosR primers (Additional file 7), and moved into the p201N Cas9 vector with the PacI and SpeI restriction sites to produce p201B:Cas9 (Addgene plasmid 59177) and p201G:Cas9 (Addgene plasmid 59178).
The gRNA targets were designed as previously described [2 (link)], with the exception of the U6 promoter, which was replaced with the Medicago truncatula U6.6 polymerase III promoter [39 (link)] for efficient transcription in soybean. For the gRNA targets, 22-23-bp targets were chosen that had the GN_19-20GG motif as previously described [2 (link)]. The GFP 5′- and 3′-targets were chosen because they contain restriction sites that can be used for downstream analysis; however, given the high DNA-modification frequencies, such analyses were not performed. The GN_18-19 portion of the genomic target motif was incorporated into the gRNA target molecule. The GFP, Glyma07g14530, and DDM1 gRNA target sequences were synthesized by IDT using gBlocks. The gBlocks were amplified by PCR with flanking primers containing I-PpoI restriction sites. All primer sequences can be found in Additional file 7. The products were then digested with I-PpoI and inserted into the p201N vector. The MET1 (Glyma04g36150 and Glyma06g18790), miR1514, and miR1509 gRNA target sequences were produced with the pUC gRNA shuttle vector system described below. Plasmids were electroporated into Agrobacterium rhizogenes strain K599 and used for hairy-root transformation. Vectors containing both the Cas9 and gRNA target cassettes were combined by inserting the gRNA target cassette into the p201N Cas9 I-PpoI site.

To analyse the subcellular localization of SgPAP7, SgPAP10, and SgPAP26, the open reading frame (ORF) sequences of SgPAP7, SgPAP10, and SgPAP26 without stop codons were amplified from cDNA stock of TPRC2001-1 with specific primers (Supplementary Table S1). Subsequently, these three SgPAP sequences were separately cloned into the binary vector pLGFP and fused with GFP at the C-terminus. The SgPAP-GFP fusion vectors and GFP empty constructs were separately introduced into Agrobacterium tumefaciens strain GV3101. Transformed GV3101 cells were grown overnight at 28°C in liquid yeast extract peptone (YEP) medium, and suspended to an absorbance of 0.5 at 600nm in agroinfiltration buffer (pH 5.6), containing 10mM MES, 10mM MgCl₂, and 0.15mM acetosyringone. After incubating for 3h at 25°C in the dark, equal volumes of mixed suspensions were syringe-infiltrated into the abaxial side of near-fully expanded leaves of 5–6-week-old tobacco (Nicotiana benthamiana) plants. After 3 d, epidermal cells on the abaxial leaf side were imaged on a Zeiss LSM7 DUO confocal microscope (Zeiss, Germany). Co-localization experiments were performed using the mCherry-labelled plasma membrane marker AtPIP2A-mCherry.
For subcellular localization of SgPAPs in transgenic bean hairy roots, the SgPAP-GFP fusion and GFP empty constructs were separately transformed into Agrobacterium rhizogenes strain K599, which were then used to generate transgenic bean hairy roots with SgPAP-GFP or GFP overexpression. Transgenic bean hairy roots were generated as described by Liang et al. (2012) . Confocal images were taken on a Zeiss LSM7 DUO confocal microscope (Zeiss, Germany). GFP fluorescence was stimulated at 488nm and detected with filter sets at 500–530nm. For western blot analysis, soluble proteins and membrane proteins were separately extracted using the FOCUS™ Global Fractionation kit (G-Biosciences, USA) from transgenic hairy roots, which were generated as above. Western blotting analysis was carried out as described by Liang et al. (2010) (link). Briefly, extracted proteins were separated by 10% SDS-PAGE in a Mini-PROTEAN Tetra Cell (Bio-Rad, USA) and transferred onto polyvinylidene difluoride membranes (GE Healthcare Life Sciences, USA) using a Trans-Blot Cell (Bio-Rad, USA). Subsequently, the membranes were incubated overnight in 0.01M Tris buffer (pH 8.0) containing 5% (w/v) non-fat milk powder and 0.15M NaCl. Membranes were separately probed with a GFP antibody (HuaAn Biotechnology, China), a phosphoenolpyruvate carboxylase antibody (Agrisera, Sweden), and a plasma membrane proton ATPase antibody (Agrisera, Sweden). Antigenic polypeptides were visualized using alkaline-phosphatase-tagged secondary antibodies and BCIP/NBT substrates. For localization in Arabidopsis protoplasts, the SgPAP-GFP fusion or GFP empty constructs were transiently expressed in Arabidopsis mesophyll protoplasts following the methods of Wu et al. (2009) . The GFP fusion constructs were co-transfected with the plasma membrane marker AtPIP2A-mCherry. Confocal images were taken on a Zeiss LSM7 DUO confocal microscope (Zeiss, Germany). GFP fluorescence was stimulated at 488nm and detected with filter sets at 500–530nm. The mCherry fluorescence was excited at 568nm and emission captured at 580–630nm.

Two cultivars of industrial chicory were assessed is this study. VL-70 seeds and sterile in vitro plants of K1793 were obtained from ILVO Flanders Institute for Agriculture, Fisheries and Food Research, Melle, Belgium. Seeds of VL-70 lines were surface sterilized with 70% ethanol (2 min), after which the seeds were transferred to 5% sodium hypochlorite solution (15 min). After this, the seeds were rinsed three times with UHP H₂O and then sown on solid Murashige and Skoog medium [20 (link)]. Hairy roots of C. intybus VL-70 line were initiated by Agrobacterium rhizogenes infection using altogether three bacterial strains: LBA9402, A4, and 15834. The agropine-type strains were A4 (kindly provided by Dr. David A. Tepfer, Versailles, France), LBA9402, and 15834 (kindly provided by Prof. Ulf Nyman, Copenhagen, Denmark). The strains LBA9402 and 15834 were cultivated in YMB (yeast mannitol broth) culture medium added with 100 ppm rifampicin. The strain A4 was cultivated in APM culture medium [21 (link)] added with 0.1 ppm biotin. Bacteria were cultivated on solid medium at +28 °C, 48 h prior to infection. Infections were performed by sterile syringe needle dipped in the bacterial mass and infecting a sterile explant.
For establishment of K1793 hairy roots, sterile in vitro plant leaves were infected with A. rhizogenes LBA9402 as described above. Explants were then placed on modified Gamborg B5-media [22 (link),23 (link)] and were incubated in the dark for a 48-h co-cultivation period.
After the co-cultivation, all the explants were transferred onto bacteriological media plates with cefotaxime sodium 500mg/L to kill the excess bacteria. After 10–14 days, hairy roots started to appear on the wound site, and they were excised from the explant and cultivated on solid modified B5 medium. The presence of rolB and the absence of virD genes were confirmed by PCR as earlier described [12 (link)]. For biomass determination, 100 mg FW of hairy roots from both cultivars were weighed in 100 mL Erlenmeyer flasks and 20 mL of modified B5 medium was added. Roots were cultivated in dark as described in [23 (link)]. Hairy roots were harvested by filtering after 7, 14, 21, or 28 days of cultivation and freeze-dried before analyses. Three in vitro plants of lines VL-70 and K1793 were transferred to soil and cultivated for a period of 3 months to form a taproot. Taproots were collected and freeze-dried before analysis.

Soybean hairy root transformation was done using Agrobacterium rhizogenes K599 carrying GmNINa-pMDC32 (GmNINa-OE), GmNINa-SRDX-pMDC32 (GmNINa-SRDX) or pMDC32 (Empty vector) plasmid according to methods described by Wang et al. [81 (link)]. The composite plants were inoculated with USDA110 (OD₆₀₀ = 0.08) at 10 days after transplantation (30 mL per plant). Three days after being inoculated with USDA110, the transgenic hairy roots were taken for analyzing gene expression. The primers of GmNINaCDS or GmNINaUTR were used for analyzing GmNINa expression in transgenic hairy roots having GmNINaOE or GmNINa-SRDX, respectively [82 (link)].

The plant materials were M. truncatula A17, soybean (G. max cv. Williams 82) and L. japonicus Gifu. Roots of these legumes were transformed using Agrobacterium rhizogenes strain ARqua-1 or K599 carrying specific vectors^{28 (link),29 (link)}. Nodules were induced on soybean by B. japonicum CB1809 when the NCR169 promoter activity was investigated and by B. japonicum USDA110 wild-type, its ΔbclA mutant, when the effects of NCR169 on bacteroids were tested. L. japonicus was inoculated with M. loti R7A and M. truncatula by S. medicae WSM419. Plants for nodulation were grown in vermiculite and fertilized once per week with 1 g l⁻¹ of Plant-Prod fertilizer (0–15–40, N–P–K; Brampton) in a growth chamber programmed for 16 h light and 8 h dark, at 28 °C/22 °C day/night for soybean and at 22 °C/20 °C day/night for M. truncatula and L. japonicus.