CRISPR/Cas9-Mediated Generation of Mutant Lines in Medicago truncatula Indicates a Symbiotic Role of MtLYK10 during Nodule Formation

Simple Summary

Clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein 9 (Cas9) technology has become an important research tool for targeted mutagenesis in plants. However, the construction of mutants in the model legume Medicago truncatula remains challenging as it requires the combination of an efficient CRISPR/Cas9-mediated gene editing system for explants and specific tissue culture techniques for regeneration to whole plants. Here, we describe the construction of M. truncatula mutants using the Agrobacterium tumefaciens-mediated transformation of leaf explants. Regeneration to whole plants was performed at high efficiency (80%), and a gene editing efficiency of up to 70% was reached in the obtained plants. CRISPR/Cas9-mediated gene editing was achieved for three M. truncatula genes. Three mutant lines with knockout mutations in the LysM receptor kinase gene MtLYK10 were further characterized. Inoculation experiments with nitrogen-fixing Sinorhizobium meliloti bacteria indicated that root nodule formation was impaired in the constructed MtLYK10 mutants. In conclusion, we show that the described CRISPR/Cas9 system is suitable for efficient gene editing in M. truncatula and that MtLYK10 plays a role in nodule symbiosis.

Abstract

CRISPR/Cas9 systems are commonly used for plant genome editing; however, the generation of homozygous mutant lines in Medicago truncatula remains challenging. Here, we present a CRISPR/Cas9-based protocol that allows the efficient generation of M. truncatula mutants. Gene editing was performed for the LysM receptor kinase gene MtLYK10 and two major facilitator superfamily transporter genes. The functionality of CRISPR/Cas9 vectors was tested in Nicotiana benthamiana leaves by editing a co-transformed GUSPlus gene. Transformed M. truncatula leaf explants were regenerated to whole plants at high efficiency (80%). An editing efficiency (frequency of mutations at a given target site) of up to 70% was reached in the regenerated plants. Plants with MtLYK10 knockout mutations were propagated, and three independent homozygous mutant lines were further characterized. No off-target mutations were identified in these lyk10 mutants. Finally, the lyk10 mutants and wild-type plants were compared with respect to the formation of root nodules induced by nitrogen-fixing Sinorhizobium meliloti bacteria. Nodule formation was considerably delayed in the three lyk10 mutant lines. Surprisingly, the size of the rare nodules in mutant plants was higher than in wild-type plants. In conclusion, the symbiotic characterization of lyk10 mutants generated with the developed CRISPR/Cas9 protocol indicated a role of MtLYK10 in nodule formation.

1. Introduction

Medicago truncatula (barrel medic) is a model legume and particularly suitable for research on plant–microbe interactions such as the nodule symbiosis with nitrogen-fixing bacteria (rhizobia) and root colonization by symbiotic arbuscular mycorrhizal (AM) fungi. In the nodule symbiosis, Sinorhizobium (Ensifer) bacteria enter roots via root hairs, and nitrogen-fixing bacteroids in formed root nodules supply M. truncatula with fixed nitrogen [1]. In the AM symbiosis, M. truncatula and many other plants benefit from enhanced access to mineral nutrients supplied by the AM fungi, especially phosphorus [2]. M. truncatula is an annual and self-pollinating legume related to the pasture legume M. sativa (alfalfa) [3,4]. The genome of M. truncatula is diploid and has been sequenced for ecotypes such as Jemalong A17 [5] and R108 [6,7,8,9]. Various research tools such as mutant collections containing Tnt1 retrotransposon insertions [10,11] and gene expression data for different tissues and mutants (e.g., MtExpress gene expression atlas [12]) are available. M. truncatula is amenable to genetic transformation using Agrobacterium tumefaciens carrying a given binary vector. The construction of stable transformants usually includes the regeneration of transformed leaf explants to whole plants [13,14,15,16], whereas floral dip transformation provides suboptimal results in M. truncatula [17]. Leaf explants of ecotype R108 exhibit high regeneration capacity and are thus most suitable for obtaining whole plants [13,14]. Furthermore, transgenic M. truncatula roots (hairy roots) can be obtained through transformation with the help of Agrobacterium rhizogenes strains [18,19].

Many bacteria and archaea possess an adaptive defense system that can act against viruses and other foreign DNA. Clustered regularly interspaced short palindromic repeats (CRISPR) derived from foreign DNA are incorporated into the prokaryotic genome and then transcribed. Then, the produced RNAs undergo processing steps resulting in short CRISPR RNAs (crRNAs) that guide CRISPR-associated (Cas) protein(s) to homologous nucleic acid target sequences. As a result, the prokaryotes can efficiently degrade undesired viral and other foreign DNA [20]. Taken advantage of these CRISPR-Cas systems, eukaryotic genomes can now be genetically modified, e.g., by the Cas9 nuclease gene of Streptococcus pyogenes in combination with a designed single guide RNA (sgRNA) construct. The sgRNA molecule consists of the target-specific crRNA fused to the trans-activating CRISPR RNA (tracrRNA). The Cas9-sgRNA complex is then guided to the desired target DNA region in the genome which contains a protospacer adjacent motif (PAM) required for cleavage by Cas9 [21]. The generated double-stranded DNA breaks are then repaired by cellular DNA repair mechanisms, often resulting in short deletions and/or insertions at the cleavage site [22]. CRISPR/Cas9-mediated gene editing has become a tool for targeted mutagenesis in plants [23], including M. truncatula transformed with the help of A. tumefaciens [24,25,26,27,28,29,30,31,32] or A. rhizogenes [33,34,35,36,37,38]. However, the generation of homozygous M. truncatula mutant lines remains challenging as it requires the combination of an efficient CRISPR/Cas9-mediated gene editing system for explants and specific tissue culture techniques for regeneration to whole plants.

Here, we describe the construction of M. truncatula mutants using the transformation of leaf explants with A. tumefaciens carrying constructed CRISPR/Cas9 binary vectors. Regeneration to whole plants was performed at high efficiency. Obtained M. truncatula lines with knockout mutations in the LysM receptor kinase gene MtLYK10 [39] were further characterized. Inoculation experiments with nitrogen-fixing S. meliloti bacteria indicated that the described CRISPR/Cas9 system is suitable for efficient gene editing in M. truncatula and that MtLYK10 plays a role in nodule symbiosis.

2. Materials and Methods

2.1. Bacterial Strains and Growth Media

Escherichia coli DH5α, Top10, and DH10B were grown in LB medium (5 g·L⁻¹ yeast extract, 10 g·L⁻¹ tryptone, 10 g·L⁻¹ NaCl) at 37 °C. Strain DH5α was used for the construction of plasmids pUC-tracrRNA-P_MtU6-1 and pUC-tracrRNA-P_MtU6-26. Strain Top10 allowed the maintenance of plasmids containing the ccdB gene (pBS-Cas9-ccdB, pISV-Cas9P_35S, and pISV-Cas9P_GmUbi). Constructed CRISPR/Cas9 binary vectors were transformed into strain DH10B suitable for harboring large plasmids. Agrobacterium tumefaciens EHA105 [40] was grown in LB medium at 27 °C. Sinorhizobium (Ensifer) meliloti strain 1021 was cultured in TY medium (5 g·L⁻¹ tryptone, 3 g·L⁻¹ yeast extract, 0.4 g·L⁻¹ CaCl₂) at 27 °C. Details on plasmids used in this study are shown in Supplementary Materials (Table S1). Bacterial culture media were supplemented with appropriate antibiotics (100 mg·L⁻¹ ampicillin, 50 mg·L⁻¹ kanamycin, 25 mg·L⁻¹ rifampicin, and 50 mg·L⁻¹ streptomycin).

2.2. Plant Material, Seed Germination, and Plant Growth Conditions

Medicago truncatula Gaertn. ecotype R108 (also known as R108-1 and spp. tricycla R108), Nicotiana benthamiana Domin, and Glycine max (L.) Merr. cv. Willams 82 (soybean) were used in this study. For germination, M. truncatula seed coats were scarified with a needle. The seeds were surface-sterilized through immersion in diluted sodium hypochlorite (commercial bleach; ~1.0% (w/v) active chlorine) on a shaker for 10 min. After incubation, the seeds were washed several times with sterilized H₂O. The seeds were then placed on 0.8% (w/v) water agar plates. The inverted plates were incubated at 4 °C in the dark for 48 h and then at 22 ± 1 °C for 18–20 h. The germinated seedlings were then planted into 2.5 L pots filled with soil containing vermiculite, expanded clay, and peat-based potting soil (Jiffy substrate TPS fine pH 5.8; Jiffy, Moerdijk, The Netherlands) at a ratio of 3:1:1 (v/v/v). The plants were grown in a growth room at 22 ± 1 °C under 16/8 h light/dark conditions (2000 lx light intensity; Philips Lifemax TL-D 36 W/54-765 and TL-D 36 W/29–530 daylight fluorescent tubes at a ratio of 3:1) [41]. N. benthamiana seeds were surface-sterilized with 70% (v/v) ethanol. The seedlings were planted into the Jiffy substrate and plants were kept in a temperature-controlled greenhouse at 25 ± 1 °C under 12/12 h light/dark conditions. G. max seeds were immersed in a 75% (v/v) ethanol solution for 2 min, washed several times with sterilized H₂O, and then incubated in a 20% (v/v) H₂O₂ solution for 5 min. After washing with sterilized H₂O, the seeds were incubated overnight at 4 °C and then placed onto 0.8% (w/v) water agar plates. The plates were incubated in the dark at 25 ± 1 °C for 3–4 days. The germinated seedlings were transferred into 8 L pots filled with vermiculite, expanded clay, and Jiffy substrate at a ratio of 3:1:2 (v/v/v). The plants were grown in a temperature-controlled greenhouse at 25 ± 1 °C under 16/8 h light/dark conditions.

2.3. Isolation of Genomic DNA from Plants

Plant DNA was extracted from leaves using the cetyltrimethylammonium bromide (CTAB) method [42]. Harvested leaf tissue (30–50 mg) was frozen with liquid nitrogen and ground in 1.5-mL microtubes using a plastic pestle. The powder was then mixed with 600 μL prewarmed (65 °C) CTAB extraction buffer (50 mM Tris-HCl pH 8.0, 0.6 M NaCl, 10 mM EDTA, 1% (w/v) CTAB, 0.1% (v/v) β-mercaptoethanol). The samples were then incubated at 65 °C for 30 min with gentle shaking. An equal volume of chloroform was then added and the microtubes were shaken well. Subsequently, a 2/3 volume of pre-chilled isopropanol was applied, and the samples were mixed gently. After incubation at −20 °C for 10 min, the microtubes were centrifuged (12,000 rpm; 10 min). The supernatant was discarded, and the precipitated DNA was washed twice with 70% (v/v) ethanol. Finally, the DNA was air-dried and then dissolved in H₂O.

2.4. Construction of CRISPR/Cas9 Vectors

The constructed plasmids are shown in Supplementary Materials (Table S1) and the primers used are listed in Supplementary Materials (Table S2). Firstly, a DNA fragment containing Cas9p (plant codon optimized Cas9 gene of S. pyogenes with an increased GC content) was PCR-amplified from pYLCRISPR/Cas9P_35S-B [43]. The PCR mixtures (50 μL) contained 50 ng template DNA, 0.25 μM of primers 1 and 2, 0.2 mM dNTP, 1.5 mM MgSO₄, 5 μL 10 × PCR Buffer, and 1 unit of KOD-Plus-Neo polymerase (Toyobo, Osaka, Japan). The PCR conditions were as follows: (i) 94 °C for 2 min; (ii) 35 cycles: 98 °C for 10 s, 68 °C for 2.5 min; and (iii) 68 °C for 10 min. The amplicon consisted of an enhancer sequencing, the coding region of Cas9p with N-terminal and C-terminal nuclear localization signal (NLS) sequences [44], and the cauliflower mosaic virus (CaMV) terminator. In parallel, primers 3 and 4 were used to amplify the ccdB expression cassette (with the lac promoter of E. coli) from pYLCRISPR/Cas9P_35S-B. The CcdB toxin targets DNA gyrase and thus can be used for the positive selection of recombinant DNA [45]. The amplified Cas9 and ccdB cassettes were then inserted into pBluescript II SK (+) (Stratagene, La Jolla, CA, USA), which was digested with SmaI and NotI using the One Step Cloning Kit from Novoprotein (Suzhou, China). The plasmid was named pBS-Cas9-ccdB and maintained in E. coli Top10.

To drive Cas9p expression under the control of a strong plant promoter, a soybean ubiquitin gene promoter sequence, known to be active in M. truncatula [24,27], was cloned from genomic G. max DNA (GenBank accession number EU310508.1; abbreviated as P_GmUbi in plasmids of this study). The PCR mixture (50 μL) contained 25 ng template DNA, 0.2 μM of primers 5 and 6, 200 mM dNTP, and 2.5 units of LA Taq polymerase (Takara Shuze, Osaka, Japan). The PCR conditions were as follows: (i) 95 °C for 5 min; (ii) 35 cycles: 95 °C for 30 s, 58 °C for 30 s, 72 °C for 1 min; and (iii) 72 °C for 10 min. The amplicon was inserted into the T-vector pMD19-T (Takara Shuze, Osaka, Japan) using the DNA ligase provided by the supplier. The resulting plasmid was named pMD19-P_GmUbi.

DsRed1 encoding a red fluorescent protein of the mushroom coral Discosoma sp. [46] was used to visualize the transformed M. truncatula plants. DNA was PCR-amplified from pRT104-DsRed1 [47] with primers 7 and 8, and the reaction conditions were identical to those for the construction of pMD19-P_GmUbi. The amplicon, consisting of a CaMV 35S promoter, DsRed1, and the polyA signal sequence of pRT104 [48], was cloned into pMD19-T. The BsaI restriction site in DsRed1 was then removed through site-directed mutagenesis [49] using KOD-Plus-Neo polymerase and the primers 9 and 10. DpnI was then used to eliminate the template DNA. The obtained plasmid was named pMD19-DsRed1(ΔB). Subsequently, the DsRed1(ΔB) expression cassette was ligated into the HindIII site of the binary vector pISV2678, constructed by Michael Schultze at the CNRS (Gif-sur-Yvette, France). This vector is a derivative of pGPTV-BAR [50] and contains a Bar (bilaphos resistance) gene under the control of a nopaline synthase promoter for glufosinate selection in transformed plants. The pISV2678 vector carrying the DsRed1(ΔB) expression cassette was named pISV-DsRed1(ΔB). Except for the missing BsaI site, pISV-DsRed1(ΔB) is identical to pISV-DsRed1 constructed previously [47] (GenBank accession number MW701373; Addgene ID 171024).

The pISV-DsRed1(ΔB) vector was further modified to construct a vector containing Cas9p and ccdB. DNA from pBS-Cas9-ccdB was amplified with primers 11 (with a PacI sequence) and 12, and the amplicon was cloned into the EcoRI site of pISV-DsRed1(ΔB). The resulting binary vector, named pISV-CRISPR/Cas9P_35S, contained the double CaMV promoter from pISV-DsRed1(ΔB), a translational enhancer sequence (5′ leader of tobacco etch virus), the Cas9p gene sequence with a CaMV 35S terminator, and the ccdB expression cassette flanked by BsaI restriction sites.

To obtain a similar vector with Cas9p driven by a ubiquitin gene promoter, the GmUbi promoter sequence amplified from pMD19-P_GmUbi (primers 5 and 6) was cloned into the generated PacI site of pISV-CRISPR/Cas9P_35S. The resulting plasmid was named pISV-CRISPR/Cas9P_GmUbi.

Next, pUC18 [51] derivatives containing tracrRNA and small nuclear RNA (snRNA) promoter sequences of M. truncatula (P_MtU6-1 and P_MtU6-26) were constructed. DNA containing the pUC18 plasmid backbone and the tracrRNA was PCR-amplified from pYLsgRNA-AtU6-1 [43] using primers 13 and 14. The P_MtU6-1 sequence was obtained from a previous publication [26]. The P_MtU6-26 sequence of M. truncatula R108 was the result of a BLAST search at the Medicago truncatula Genome Database [52,53] using the P_AtU6-26 sequence of A. thaliana [54] as a query sequence. The promoter sequences were amplified from the genomic DNA of M. truncatula R108 (primers 15 and 16 for P_MtU6-1; primers 17 and 18 for P_MtU6-26). The PCR mixtures (50 μL) contained 50 ng template DNA, 0.25 μM of each primer, 0.2 mM dNTP, 1.5 mM MgSO₄, 5 μL 10 × PCR Buffer, and 1 unit of KOD-Plus-Neo polymerase. The PCR conditions were as follows: (i) 94 °C for 2 min; (ii) 98 °C for 10 s, 68 °C for 1.5 min; 35 cycles; and (iii) 68 °C for 10 min. The amplicons were then fused using a Gibson Assembly Cloning Kit (Novoprotein, Suzhou, China). The resulting plasmids containing two BsaI restriction sites were named pUC-tracrRNA-P_MtU6-1 and pUC-tracrRNA-P_MtU6-26, respectively.

To design sgRNA sequences for a specific gene to be mutated, target sequences upstream of the PAM sequence (5′-NGG-3′) in a given gene sequence were identified using CRISPR-P 2.0 [55,56] and targetDesign [57,58]. In this study, target sequences were used for the β-glucuronidase (GUSPlus) gene of pCAMBIA1305 and three M. truncatula R108 genes [7]. In this article, these genes were named MtLYK10 (M. truncatula lysin-motif receptor-like kinase 10, Medtr5g033490), MtMFS1 (M. truncatula major facilitator superfamily transporter 1, Medtr3g093270), and MtMFS2 (M. truncatula major facilitator superfamily transporter 2, Medtr3g093290). The synthesized target-specific 20-bp oligonucleotides (crRNA), are listed in Supplementary Materials (Table S3). The oligonucleotide pairs (1 μM in 20 μL water) were then annealed to form a double strand with ATTG/CAAA overhangs (samples heated to 90 °C for 1 min and then cooled to room temperature).

The sgRNA expression cassettes were constructed as follows. (i) In a restriction-ligation reaction, the double-stranded oligonucleotide (crRNA) with the ATTG/CAAA overhangs was fused with tracrRNA-P_MtU6-1 or tracrRNA-P_MtU6-26 DNA digested with BsaI. The mixture (10 μL) contained 5 U BsaI-HFV2 (NEB, Ipswich, MA, USA), 1 μL 10 × BsaI reaction buffer, 20 U T4 ligase (Genestar, Beijing, China), 1 μL 10 × T4 ligation buffer, 20 ng plasmid DNA (pUC-tracrRNA-P_MtU6-1 or pUC-tracrRNA-P_MtU6-26), and 0.1 μM of a double-stranded oligonucleotide. The mixture was subject to 5 thermal cycles (37 °C for 5 min; 20 °C for 5 min). (ii) Then, the ligation product (3 μL) served as a template for a PCR to amplify a given sgRNA expression cassette containing the P_MtU6-1 or P_MtU6-26 promoter, the tracrRNA sequence, and crRNA sequence. The reaction mixture (25 μL) contained 0.2 μM of primers 19 and 20, 0.5 U of KOD-Plus-Neo polymerase, 2.5 μL 10 × PCR Buffer, 0.2 mM dNTPs, and 1.5 mM MgSO₄. The PCR was performed for 20 cycles (98 °C for 15 s, 58 °C for 15 s, 68 °C for 10 s). (iii) A total of 1 μL of the PCR product was then directly used as template for a second PCR (50 μL) to produce sgRNA expression cassettes flanked by BsaI restriction sites (primers 21 and 22 for sgRNA expression cassettes containing the P_MtU6-1 promoter; primers 23 and 24 for sgRNA expression cassettes containing the P_MtU6-26 promoter). The reaction mixture contained 0.2 μM of each primer, 1 U of KOD-Plus-Neo polymerase; 5 μL 10 × PCR Buffer, 0.2 mM dNTPs, and 1.5 mM MgSO₄ (98 °C for 15 s, 58 °C for 15 s, 68 °C for 10 s; 25 cycles). (iv) Using Golden Gate cloning [59], the constructed sgRNA expression cassettes with terminal BsaI restriction sites were assembled and inserted into a constructed binary vector containing a Cas9p expression cassette (pISV-CRISPR/Cas9P_35S and pISV-CRISPR/Cas9P_GmUbi). For comparison, pYLCRISPR/Cas9P_35S-B containing the pCAMBIA1300 backbone, Cas9p, under the control of a double CaMV promoter [43] was also used in initial studies. Each reaction mixture (10 μL), containing 60 ng DNA of the binary vector, 15 ng DNA of a given sgRNA expression cassette, 10 U BsaI-HFV2, and 1 μL 10 × BsaI reaction buffer, was incubated at 37 °C for 30 min. Subsequently, 35 U of T4 DNA ligase and 1 μL 10 × T4 ligation buffer were added and the samples were incubated at 37 °C for 5 min, 10 °C for 5 min and 20 °C for 5 min for a total of 15 cycles. Using the described procedure, two sgRNA expression cassettes, each with different promoters (P_MtU6-1 and P_MtU6-26, respectively), were assembled and ligated into the binary vector. Finally, the ligation products were transformed into competent E. coli DH10B cells which do not grow when transformed, with plasmids containing the ccdB expression cassette. The insertion of the sgRNA expression cassettes into the binary vectors was confirmed through sequencing (pYLCRISPR/Cas9P_35S-sgRNA(GUS), pISV-CRISPR/Cas9P_35S-sgRNA(GUS), pISV-CRISPR/Cas9P_GmUbi-sgRNA(GUS), pISV-CRISPR/Cas9P_GmUbi-sgRNA(LYK10), pISV-CRISPR/Cas9P_35S-sgRNA(MFS1/MFS2), and pISV-CRISPR/Cas9P_GmUbi-sgRNA(MFS1/MFS2)).

2.5. Electroporation of CRISPR/Cas9 Binary Vectors into A. tumefaciens

The constructed CRISPR/Cas9 binary vectors were transformed into competent A. tumefaciens EHA105 cells (Weidi, Shanghai, China) with the help of an electroporator (Ding Guo, Guangzhou, China). Pre-cooled cells (100 μL) were carefully supplemented with vector DNA (≈0.5 μg) in a sterilized 0.2 cm electroporation cuvette. After electroporation at 1500 V for 10 ms, the cells were manually mixed with 1 mL of LB medium (27 °C) and transferred to 1.5 mL plastic tubes. The bacteria were placed on a shaker (200 rpm, 27 °C) for 3 h, and 100 μL of each bacterial suspension was then spread on LB agar plates containing appropriate antibiotics (50 mg·L⁻¹ kanamycin and 25 mg·L⁻¹ rifampicin). After incubation (27 °C) for two days, single colonies were picked from the plates. The presence of the vectors in the bacteria was confirmed via PCR using primers listed in Supplementary Materials (Table S2).

2.6. Agrobacterium-Mediated Transformation of N. benthamiana Leaves

To test the functionality of the different CRISPR/Cas9 vectors binary vectors, they were examined for their capacity to induce mutations in the GUSPlus gene expressed in N. benthamiana leaf cells transformed with agrobacteria. The binary vector pCAMBIA1305 containing GUSPlus with a CaMV 35S promoter was used in these experiments. GUS activity in transformed tissue was visualized using histochemical staining with 5-bromo-4-chloro-3-indolyl β-D-glucuronide (X-Gluc) [60]. For Agrobacterium-mediated transformation, individual colonies of A. tumefaciens EHA105 containing pCAMBIA1305, pYLCRISPR/Cas9P_35S-sgRNA(GUS), pISV-CRISPR/Cas9P_35S-sgRNA(GUS), and pISV-CRISPR/Cas9P_GmUbi-sgRNA(GUS) were grown in 5 mL of LB medium (supplemented with 50 mg L⁻¹ kanamycin and 25 mg L⁻¹ rifampicin) overnight at 27 °C. The cultures were then harvested using centrifugation (6000 rpm for 5 min), washed with 10 mM MgCl₂, and then resuspended in the infiltration solution (10 mM MgCl₂ supplemented with 200 μM acetosyringone) to obtain an OD₆₀₀ ≈ 0.1. Next, the resuspended culture was shaken at 50 rpm (27 °C, in the dark) for 3 h. Each strain containing a CRISPR/Cas9 binary vector was then mixed with an equal volume of the strain carrying pCAMBIA1305. The bacterial suspensions were then infiltrated into the underside of 4–5-week-old N. benthamiana leaves using a 1-mL syringe (without needle). The plants were kept in a temperature-regulated greenhouse and leaves were harvested 7 days post infiltration [61].

2.7. Transformation of M. truncatula Leaf Explants and Regeneration to Whole Plants

M. truncatula R108 leaf explants were transformed with A. tumefaciens EHA105 carrying pISV-CRISPR/Cas9P_GmUbi-sgRNA(LYK10), pISV-CRISPR/Cas9P_35S-sgRNA(MFS1/MFS2), or pISV-CRISPR/Cas9P_GmUbi-sgRNA(MFS1/MFS2), according to previously published procedures with various modifications [13,14,15,16]. A single bacterial colony carrying a given CRISPR/Cas9 binary vector was inoculated into 4 mL of LB medium supplemented with 50 mg·L⁻¹ kanamycin and 25 mg·L⁻¹ rifampicin. The bacteria were cultured on a shaker (200 rpm, 27 °C) for 24 h, and 1 mL of the preculture was transferred into 30 mL LB medium containing the same antibiotics. The cells were then cultured on the same shaker for 12–16 h until the cells reached an OD₆₀₀ ≈ 1. The cells were centrifuged at 4000 rpm for 15 min and then resuspended (OD₆₀₀ = 0.3) in liquid SH-1 medium, which is defined as SH medium (0.185 g·L⁻¹ MgSO₄·7H₂O, 2.83 g·L⁻¹ KNO₃, 0.463 g·L⁻¹ (NH₄)₂SO₄, 0.4 g·L⁻¹ KH₂PO₄, 0.166 g·L⁻¹ CaCl₂·2H₂O, 1 mg·L⁻¹ MnSO₄·H₂O, 0.5 mg·L⁻¹ H₃BO₃, 0.1 mg·L⁻¹ ZnSO₄·7H₂O, 0.1 mg·L⁻¹ KI, 0.1 mg·L⁻¹ Na₂MO₄·2H₂O, 0.2 mg·L⁻¹ CuSO₄·5H₂O, 0.1 mg·L⁻¹ CoCl₂·6H₂O, 0.2 mg·L⁻¹ CuSO₄·5H₂O, 5 mg·L⁻¹ nicotinic acid, 5 mg·L⁻¹ thiamine HCl, 5 mg·L⁻¹ pyridoxine HCl, 140 mg·L⁻¹ FeNa-EDTA, and 100 mg·L⁻¹ myo-inositol; [62]) supplemented with 30 g·L⁻¹ sucrose, 4 mg·L⁻¹ 2,4-dichlorophenoxyacetic acid (2,4-D), and 0.5 mg·L⁻¹ 6-benzylaminopurine (BAP). The SH-1 medium was adjusted to pH 5.8 with KOH. Healthy leaves of 4–6-week-old M. truncatula R108 plants were surface-sterilized with diluted sodium hypochlorite solution (≈0.6% (w/v) active chlorine) for 15 min and then washed several times with sterilized H₂O. The leaves were cut into square pieces (≈1 cm²) with a sharp scalpel blade and immersed in the prepared A. tumefaciens suspension. For agroinfiltration, the bacterial suspension with the immersed leaf explants was exposed to a vacuum (≈−0.085 MPa) for 10 min. The explants were then gently shaken (60 rpm, room temperature) for 1 h, dried with autoclaved filter papers, and transferred to solid SH-1 medium plate containing 3 g·L⁻¹ phytagel. After 3 days of co-cultivation, the explants were washed in liquid SH-1 medium for 20 min, dried with a sterilized filter paper, and finally transferred to solid SH-2 medium, which is defined as SH-1 medium supplemented with 300 mg·L⁻¹ timetin (to suppress the growth of agrobacteria) and 3 mg·L⁻¹ of the herbicide glufosinate (phosphinothricin) ammonium salt. Glufosinate allows selection for transgenic M. truncatula cells expressing the Bar gene, which confers resistance to glufosinate. The plates were kept at 24 °C in the dark. The explants were transferred every two weeks to identical plates that were freshly prepared, and the formation of calli was observed at 6–7 weeks post agroinfiltration. About 4–5 weeks later, the calli were transferred to plates containing solid MSBK plates (pH 5.8), which is defined as 4.43 g·L⁻¹ Murashige and Skoog (MS) basal medium (Coolaber Beijing, China; [63]) supplemented with 30 g·L⁻¹ sucrose, 1 mg·L⁻¹ kinetin, 0.5 mg·L⁻¹ BAP, 2 mg·L⁻¹ glufosinate, 300 mg·L⁻¹ timetin, and 3 g·L⁻¹ phytagel [16]. The plates were incubated at 22 ± 1 °C under 16/8 h light/dark conditions (2000 lx light intensity; Philips Lifemax TL-D 36 W/54-765 and TL-D 36 W/29–530 daylight fluorescent tubes at a ratio of 3:1). About two weeks later, green somatic pro-embryos were observed and transferred to plates filled with phytohormone-free SH-3 medium (SH medium supplemented with 20 g·L⁻¹ sucrose, 150 mg·L⁻¹ timetin, 2 mg·L⁻¹ glufosinate, and 3 g·L⁻¹ phytagel). The plates were refreshed every two weeks. Plantlets with fully expanded trifoliate leaves were observed after 7–8 weeks growth on these plates. The plantlets were then transferred to 350-mL glass vessels which were partially filled with ≈150 mL SH-3 medium. To ensure gas exchange, the lids of the glass vessels had a hole of 0.79 cm², which was covered with filter paper. The plantlets were placed into fresh glass vessels every two weeks. Fully regenerated plants with well-developed roots were obtained after about 10 weeks growth in these vessels. Finally, the plants were transferred to sterilized 300-mL plastic jar units linked with a cotton wick (0.5 cm in diameter). The upper jar contained vermiculite and expanded clay (3:1 v/v), and the lower jar was filled with half-strength B&D [64] nutrient solution (125 μM CaCl₂·2H₂O, 62.5 μM KH₂PO₄, 1.25 μM Fe-citrate, 31.25 μM MgSO₄·7H₂O, 187.5 μM K₂SO₄, 0.125 μM MnSO₄·7H₂O, 0.25 μM H₃BO₃, 0.025 μM CuSO₄·5H₂O, 0.0125 μM CoSO₄·7H₂O, 0.0125 μM Na₂MoO₄·2H₂O, and 0.0625 μM ZnSO₄·H₂O) supplemented with 1 mM KNO₃. The upper jar was covered with a light transparent lid for several days to acclimatize the plants to the new environment. The jar units were kept in a temperature-controlled growth room as described above.

Plants of the T₀ generation were examined for the emission of red fluorescent signals. For seed production, the transgenic plants were transferred to 2.5 L pots filled with vermiculite and expanded clay (3:1 v/v). The plants were watered with half-strength B&D nutrient solution containing 0.5 mM KNO₃ (≈300 mL per pot every 3 days).

2.8. Detection of Red Fluorescent Signals in Transformed M. truncatula Plants

A Zeiss ImagerZ1 fluoresce microscope (Carl Zeiss, Oberkochen, Germany) was used to detect red fluorescent signals in transgenic M. truncatula plants expressing DsRed1 [46]. The excitation wavelengths were 540–552 nm and the emission window 575–640 nm. Typically, small root pieces were harvested from T₀ generation plants before they were transferred from jar units to pots.

2.9. Identification of Mutations in Transformed Plants

Genomic DNA was extracted from the infiltration zone of N. benthamina leaves (5 leaves from 5 different plants per A. tumefaciens strain) to identify sgRNA-mediated mutations in GUSPlus sequences. Likewise, leaves of 10 transformed and regenerated M. truncatula plants were used for the extraction of genomic DNA to detect mutations in the MtLYK10, MtMFS1, and MtMFS2 genes. The DNA was then used as a template for a PCR (primers 31 and 32 for GUSPlus; primers 33 and 34 for MtMFS1; primers 35 and 36 for MtMFS2; and primers 37 and 38 for MtLYK10). The reaction mixture (50 μL) contained 200 ng template DNA, 0.25 μM of gene-specific primers, 0.8 mM dNTP, 5 mM MgSO₄, 5 μL 10 × PCR Buffer, and 2.5 U of LA Taq (2 × LA taq mix; Takara, Osaka, Japan). The annealing temperature was 60 °C and the extension time was 1 min. PCR products amplified from M. truncatula DNA were then directly subjected to Sanger sequencing. When the sequencing results showed double peak signals, the corresponding amplicon was inserted into pMD19-T and sequencing data with single peak signals were obtained from transformed E. coli colonies. PCR products amplified from N. benthamiana leaves were directly cloned into pMD19-T. For each transformed leaf, three randomly sampled E. coli colonies were sequenced to estimate the mutation frequency at the GUSPlus target sites.

2.10. Analysis of Generated lyk10 Mutant Plants

T₀ generation plants with identified mutations in MtLYK10 were propagated by selfing. The resulting T₁ generation plants were then subjected to PCRs (primers 37 and 38; Supplementary Materials, Table S2) and sequencing to identify homozygous (biallelic) mutant plants in three independent lyk10 mutant lines (lyk10-1, lyk10-1, and lyk10-1). The prediction of possible off-target mutations in the lyk10 mutants was performed with offTarget software [57,65] using the corresponding crRNA-PAM sequences as query sequences. The nucleotide sequences of possible off-target regions in various genes (MTR_2g063570, MTR_7g029350, MTR_4g127480, MTR_2g069950, MTR_5g024910, MTR_3g112020, MTR_1g054795, MTR_4g116370, and MTR_7g056073) were analyzed by sequencing the amplicons of PCRs using primers 41–58.

Furthermore, the nodule formation of homozygous lyk10 mutant plants was analyzed. T₁ generation plants (25–30 per identified mutation) and wild-type plants were grown in sterilized 300 mL plastic jar units (two jars were connected with a cotton wick). The upper jar was filled with vermiculite and expanded clay at a ratio of 3:1 (v/v). The lower jar contained B&D nutrient solution [64] (250 μM CaCl₂·2H₂O, 125 μM KH₂PO₄, 2.5 μM Fe-citrate, 62.5 μM MgSO₄·7H₂O, 375 μM K₂SO₄, 0.25 μM MnSO₄·7H₂O, 0.5 μM H₃BO₃, 0.05 μM CuSO₄·5H₂O, 0.025 μM CoSO₄·7H₂O, 0.025 μM Na₂MoO₄·2H₂O, and 0.125 μM ZnSO₄·H₂O) supplemented with 1 mM KNO₃. The growth conditions were the same as described above. After one week, plants (1 plant per jar unit) were inoculated with S. meliloti 1021. The bacteria were centrifuged and resuspended in 10 mM MgSO₄ (OD₆₀₀ adjusted to 0.2). Each jar system was inoculated with 2 mL bacteria. Genomic DNA was isolated from a harvested leaf at 7 days post inoculation (dpi) to identify homozygous lyk10 mutant plants using sequencing, as described above. The roots of three homozygous mutant lines were harvested at two different time points (10 and 21 dpi) to determine the number of formed nodules for each plant. Longitudinal sections of nodules (harvested at 21 dpi) were analyzed using a Zeiss Lumar.V12 microscope (Carl Zeiss, Oberkochen, Germany). Image J2 software [66] was used to calculate the area of the sectioned nodules. Statistical analysis (Student’s t-test) was performed with GraphPad Prism (version 8.0.2 (263); Prism software, Vestal, NY, USA).

3. Results

3.1. Construction of CRISPR/Cas9 Binary Vectors Suitable for M. truncatula Transformation

In previous studies, pISV2678, a binary vector with a Bar gene, was found to be very effective for the A. tumefaciens-mediated transformation of M. truncatula leaf explants [41,67,68,69]. We therefore aimed to construct a CRISPR/Cas9 vector with a pISV2678 backbone and a DsRed1 expression cassette for the visual confirmation of transformants. The Cas9p gene with N-terminal and C-terminal NLS sequences was PCR-amplified from pYLCRISPR/Cas9P_35S-B [43] along with the transcriptional enhancer 5’UTR of Tobacco Etch Virus and the CaMV terminator. Furthermore, a ccdB expression cassette was amplified from pYLCRISPR/Cas9P_35S-B. Both amplicons were then cloned into pBluescript II SK (+), resulting in pBS-Cas9-ccdB (Figure 1a). Next, the binary vector pISV2678 was modified by inserting a DsRed1 expression cassette (lacking the undesired BsaI restriction site) into the HindIII site of pISV2678, resulting in pISV-DsRed1(ΔB) (Figure 1b). The insert of pBS-Cas9-ccdB was then cloned into the EcoRI site of pISV-DsRed1(ΔB), yielding pISV-CRISPR/Cas9P_35S (Figure 1c). In this vector, the Cas9p gene expression is under the control of a double CaMV promoter, while the ccdB expression cassette is flanked by unique BsaI restriction sites. In addition, a similar binary vector, pISV-CRISPR/Cas9P_GmUbi, was constructed in which Cas9p gene expression is driven by a strong ubiquitin gene promoter of soybean (Figure 1d). Furthermore, plasmids containing tracrRNA and snRNA promoter sequences of M. truncatula (P_MtU6-1 or P_MtU6-26) were constructed. The pUC18 plasmid backbone and the tracrRNA sequence were PCR-amplified from pYLsgRNA-AtU6-1 [43] and the P_MtU6-1 or P_MtU6-26 promoter sequences from genomic DNA of M. truncatula. The plasmids containing the amplified DNA fragments were named pUC-tracrRNA-P_MtU6-1 and pUC-tracrRNA-P_MtU6-26, respectively (Figure 1e).

Using CRISPR-P 2.0 and targetDesign software, target sequences upstream of the PAM sequence (5′-NGG-3′) were identified for the GUSPlus gene in the binary vector pCAMBIA1305 and the M. truncatula genes MtLYK10, MtMFS1, and MtMFS2. Synthesized oligonucleotides (listed in Supplementary Material, Table S3) were then annealed to form double strands with ATTG/CAAA overhangs (crRNA), which were then cloned into pUC-tracrRNA-P_MtU6-1 or pUC-tracrRNA-P_MtU6-26 digested with BsaI. The obtained DNA was used as a template for a PCR to amplify a given sgRNA expression cassette consisting of the P_MtU6-1 or P_MtU6-26 promoter, the tracrRNA, and a synthesized crRNA. In a second PCR, terminal sequences with a BsaI restriction site were introduced into the sgRNA expression cassette. Finally, using a restriction-ligation reaction with BsaI, the sgRNA expression cassettes were inserted into pISV-CRISPR/Cas9P_35S (sgRNA expression cassettes to edit GUSPlus, MtMFS1, and MtMFS2) and pISV-Cas9P_GmUbi (sgRNA cassettes to edit GUSPlus, MtLYK10, MtMFS1, and MtMFS2), respectively. For comparison, pYLCRISPR/Cas9P_35S-B [43], a pCAMBIA1300 derivative containing Cas9p under the control of a double CaMV 35S promoter, was also used in initial GUSPlus experiments. The procedure for cloning sgRNA expression cassettes and inserting them into the binary vectors is illustrated in Supplementary Materials (Figure S1). A schematic view of the obtained CRISPR/Cas9 binary vectors is shown in Figure 1f. All constructed binary vectors contained two sgRNA expression cassettes with different promoters (P_MtU6-1 and P_MtU6-26, respectively).

3.2. Testing for the Functionality of Constructed CRISPR/Cas9 Binary Vectors in N. benthamiana

Transient gene expression experiments with N. benthamiana plants were performed to examine the functionality of the constructed CRISPR/Cas9 binary vectors. The vectors (pISV-CRISPR/Cas9P_35S-sgRNA(GUS) and pISV-CRISPR/Cas9P_GmUbi-sgRNA(GUS)) contained two sgRNA expression cassettes to induce mutations at two different target sites of a co-transformed GUSPlus gene. For comparison, the two sgRNA expression cassettes targeting GUSPlus were also introduced into pYLCRISPR/Cas9P_35S-B [43] (vector pYLCRISPR/Cas9P_35S-sgRNA(GUS)). Prior to agroinfiltration, suspensions of A. tumefaciens EHA105 harboring a given CRISPR/Cas9 binary vector were mixed with an equal volume of EHA105 carrying pCAMBIA1305, which contains a GUSPlus expression cassette. Each bacterial suspension was injected into five leaves of five different N. benthamiana plants and leaves were harvested 7 days post infiltration. GUS activity tests with X-Gluc showed blue coloration in the infiltration zones, indicating the successful transformation of the GUSPlus gene. Subsequently, DNA was isolated from the infiltration zones and used as a template to amplify a GUSPlus gene fragment containing the two target sites. As the sequencing of amplicons resulted in double peaks, the amplicons were cloned into pMD19-T, which was then transformed into E. coli cells to obtain single peaks from isolated plasmid DNA. Sequence comparisons showed that all the CRISPR/Cas9 binary vectors tested can cause mutations at the GUSPlus target sites. In a number of sequences, gene knockout was observed due to a frameshift mutation at a single target site. Mutations at two target sites led to DNA deletion between the target sites. Examples of obtained GUSPlus sequence variants are shown in Figure 2a and the corresponding Sanger sequencing chromatograms in Figure 2b. The editing efficiency was then quantified for each binary vector using data from a total of 194 sequences. The use of the pISV-CRISPR/Cas9P_GmUbi-sgRNA(GUS) vector resulted in the highest frequency of simultaneous mutations at both target sites (Figure 2c). These data indicated that the constructed CRISPR/Cas9 binary vectors were functional and thus could be used for experiments to edit specific genes in the M. truncatula genome.

3.3. Efficient CRISPR/Cas9-Mediated Generation of Mutant Lines in M. truncatula

To determine whether specific M. truncatula genes can be mutated using the established CRISPR/Cas9 system, we constructed binary vectors to edit three M. truncatula genes, namely, the LysM-type receptor kinase gene MtLYK10 [39] and two MFS transporter genes related to N-acetylglucosamine transporters in rice [70]. The exon/intron structure and selected target sites in these three M. truncatula genes are illustrated in Figure 3a. The constructed CRISPR/Cas9 binary vectors (pISV-CRISPR/Cas9P_35S-sgRNA(MFS1/MFS2), pISV-CRISPR/Cas9P_GmUbi-sgRNA(MFS1/MFS2), and pISV-CRISPR/Cas9P_GmUbi-sgRNA(LYK10)) were then used for the A. tumefaciens-mediated transformation of leaf explants. Pictures illustrating the different cultivation steps from leaf disks to whole transgenic plants are shown in Figure 3b–g. The whole procedure lasted 6–9 months. The expression of transformed DsRed1 in regenerated glufosinate-resistant plants (T₀ generation) was confirmed through the fluorescence microscopy analysis of the roots (Figure 3h). A PCR analysis confirmed that the transformed plants contained the Bar gene (Figure 3i). A quantitative analysis indicated that most leaf disks could form a callus (callus formation efficiency of about 95%). The regeneration efficiency from calli to whole plants was also high. On average, transgenic plants could be obtained from about 85% of the calli (Figure 3j). The sequencing results of PCR-amplified DNA from the leaves of T₀ generation plants indicated double peaks in the target regions, indicating that one allele of the target gene was successfully edited (Supplementary Materials, Figure S2). For confirmation, the amplicons were cloned into pMD19-T and sequenced. Only single peaks were observed in chromatograms obtained from the plasmid DNA of individual E. coli colonies.

The variations in MtLYK10 sequences (single peaks obtained from the plasmid DNA of a single E. coli colony) are shown in Figure 4. These sequences were obtained from 10 regenerated plants derived from the transformation of 10 different leaf explants with pISV-CRISPR/Cas9P_GmUbi-sgRNA(LYK10). This vector contains sgRNA expression cassettes to edit MtLYK10 at two different target sites. Mutations were observed at target site 1 for two plants and at target site 2 for four plants. In three plants (P1, P8, and P9), the open reading frame of MtLYK10 (exon 1) was disrupted. In one plant (P3), mutations at both target sites caused a deletion of 654 bp. Accordingly, the analysis of amplified MtLYK10 DNA from P3 by agarose gel electrophoresis revealed two bands (Supplementary Materials, Figure S3). Overall, six mutations in the MtLYK10 gene were obtained at the two target sites, resulting in an editing efficiency of 2/10 (=20%) for target site 1 and 4/10 (=40%) for target site 2.

The variations in single-peak sequences of the two examined MFS transporter genes in 10 regenerated plants transformed with pISV-CRISPR/Cas9P_35S-sgRNA(MFS1/MFS2) and pISV-CRISPR/Cas9P_GmUbi-sgRNA(MFS1/MFS2) are shown in Figure 5. The two vectors contain identical sgRNA sequences (one target site in the MtMFS1 gene and another target site in the MtMFS2 gene), but differ with respect to the promoter controlling Cas9p expression (CaMV 35S promoter and GmUbi promoter, respectively). For the plants transformed with pISV-CRISPR/Cas9P_35S-sgRNA(MFS1/MFS2), no mutation was observed at the target site in MtMFS1 while mutations at the target site of MtMFS2 were identified in five plants (50% editing efficiency). Four of these mutations (plants P3, P4, P6, and P8) resulted in a disrupted frameshift of MtMFS2 (exon 1) (Figure 5a,b). For the plants transformed with pISV-CRISPR/Cas9P_GmUbi-sgRNA(MFS1/MFS2), two plants showed mutations at the target site of MtMFS1 (20% editing efficiency) and seven plants at the target site of MtMFS2 (70% editing efficiency). In one plant (P1), the open reading frame of MtMFS1 (exon 1) was altered. Four plants (P4, P5, P8, and P9) exhibited sequences with a disrupted open reading frame of MtMFS2 (exon 1). Furthermore, one plant (P2) showed simultaneous mutations in MtMFS1 and MtMFS2. In this plant, the insertion of a nucleotide disrupted the open reading frame of MtMFS1 while the mutated MtMFS2 sequence exhibited a deletion of 18 bp that caused the elimination of six amino acids while the reading frame remained unaffected (Figure 5c,d).

3.4. Analysis of Homozygous lyk10 Mutants

For T₀ generation plants, the sequencing analysis of MtLYK10 indicated the presence of frameshift mutations in three independent plants (P1, P3, and P8 in Figure 4). After selfing of these plants, seeds were collected and used for further analysis of the T₁ generation. MtLYK10 sequencing data were obtained from leaf DNA of 2-week-old plants. In about one quarter of T₁ generation plants, no wild-type allele sequences could be observed, indicating the generation of homozygous MtLYK10 mutant plants (Figure 6a). The mutations in these plants were identical to those of the T₀ generation plants P1, P3, and P8 (Figure 4). The mutant lines were named lyk10-1 (derived from P1), lyk10-2 (derived from P3), and lyk10-3 (derived from P8). No off-target mutations were identified in these three lyk10 mutant lines when the DNA of nine genes with predicted off-target sites was sequenced (Supplementary Materials, Table S4). Finally, the identified homozygous lyk10 mutant and wild-type plants were compared with respect to the formation of root nodules induced by S. meliloti. In contrast to wild-type plants, no nodules were observed on the roots of the three mutant lines at 10 dpi (Figure 6b). At 21 dpi, nodules on the roots of mutant plants were observed but their numbers were statistically lower than in wild-type plants (Figure 6c). Surprisingly, the size of the rare nodules in mutant plants was higher than in wild-type plants (Figure 6d). In conclusion, nodule formation was considerably delayed in the three constructed lyk10 mutant lines.

4. Discussion

In this study, we used CRISPR/Cas9 technology to optimize gene editing in the model legume M. truncatula. The procedure is based on the transformation of leaf explants with the help of A. tumefaciens EHA105 carrying a constructed derivative of pISV-CRISPR/Cas9P_GmUbi. The binary vectors contained a plant codon-optimized Cas9 gene (Cas9p), Bar, DsRed1, and sgRNA expression cassettes with snRNA promoter sequences of M. truncatula (P_MtU6-1 and P_MtU6-26). We found that the high transformation and regeneration efficiency of leaf explants under glufosinate selection is probably critical for the construction of mutant plants. Small technical details such as the preparation of surface-sterilized leaf explants (≈1 cm²) from healthy leaves and the use of timetin to suppress the growth of agrobacteria were found to be crucial for the regeneration to whole plants. Furthermore, we observed that kinetin (at a concentration of 1 mg·L⁻¹ [16]) and the used light conditions [41] are important factors for efficient somatic pro-embryogenesis.

The efficiency of CRISPR/Cas9 systems also depends on strong promoters. Ubiquitin promoters are constitutive promoters that are frequently active in a number of different plants. For example, a ubiquitin promoter from Lotus japonicus was routinely used to overexpress genes in M. truncatula in previous studies [71,72,73]. When comparing different binary vectors in transformation experiments with N. benthamiana, we found the highest mutation frequency at target sites of the co-transformed GUSPlus gene for pISV-CRISPR/Cas9P_GmUbi-sgRNA(GUS), which contains Cas9p under the control of a soybean ubiquitin gene promoter [24,27]. Likewise, the editing efficiency in M. truncatula was considerably higher for this vector as compared to the same vector with a CaMV 35S promoter controlling Cas9p expression (Figure 5). We therefore recommend using pISV-CRISPR/Cas9P_GmUbi for future gene editing studies with M. truncatula.

To ensure high gene editing efficiency in M. truncatula, our sgRNA expression cassettes contained P_MtU6-1 and P_MtU6-26, which are snRNA promoters of M. truncatula. The P_MtU6-1 promoter has been used in previous works [26,30,31], while P_MtU6-26 was found to be as active as P_MtU6-1 in the current study. The combination of these promoters allowed the simultaneous editing of two target sites in M. truncatula. With these promoters, an editing efficiency of up to 70% (per target site) was achieved (Figure 5), which is comparable to or higher than the mutation frequencies reported in previous publications on CRISPR/Cas9-mediated gene editing in M. truncatula (e.g., references [24,26,29,31]). Future gene editing in M. truncatula could be performed with pISV-CRISPR/Cas9P_GmUbi containing more than two sgRNA expression cassettes (e.g., by assembling them with the help of Esp3I in addition to BsaI). For such constructs, additional M. truncatula promoters such as P_MtU6-6 [31,38] and P_MtU6-8 [36] could be used. Alternatively, multiple sgRNAs, each separated by RNA cleavage sequences, could be expressed from a single promoter and then processed to sgRNAs, e.g., with the endoribonuclease Csy4 [25].

In general, the probability to obtain a gene knockout mutant increases when the binary vector contains multiple sgRNA sequences for several targets in the same gene. This, however, may also increase the risk of mutations in other genes. In our study, off-target mutations were not identified in the obtained lyk10 mutant plants. To minimize the risk of off-target mutations, we recommend mutating a given M. truncatula gene using a maximum of two sgRNA sequences. Off-target effects could be also reduced by other means, including the use of CRISPR/Cas9 systems with a nuclease-deactivated variant of Cas9 (dCas9) [74].

In our transient transformation experiments with N. benthamiana, co-transformation with the two A. tumefaciens strains carrying pISV-CRISPR/Cas9P_GmUbi-sgRNA(GUS) and pCAMBIA1305 (containing GUSPlus) resulted in a surprisingly high number of mutations in the GUSPlus gene. It is therefore possible to use N. benthamiana leaves as a test system to rapidly verify the functionality and gene editing activity of pISV-CRISPR/Cas9P_GmUbi derivatives containing sgRNA expression cassettes. M. truncatula DNA fragments with target sites could be cloned into pCAMBIA1305 and co-transformation experiments could be performed as described for GUSPlus in our study. Results from such a pre-test could provide information on the editing activity of a given sgRNA within a few days (post infiltration with agrobacteria), while the regeneration of leaf explants to whole M. truncatula plants takes several months. Indeed, our study shows that the sgRNA sequence chosen for editing MtMFS1 was suboptimal for unknown reasons and that preliminary transformation experiments with N. benthamiana plants are particularly recommended when the mutation of multiple genes is envisaged. The pre-testing of binary vectors with sgRNA expression cassettes could also be performed through the CRISPR/Cas9-mediated reconstitution of a fluorescence protein in N. benthamiana [75].

The constructed CRISPR/Cas9 binary vectors contained the fluorescence marker gene DsRed1. Accordingly, all regenerated plants were found to be fluorescent when analyzed under red fluorescence conditions. To simplify the removal of the Cas9p gene in regenerated M. truncatula plants, the DsRed1 marker will be helpful as non-fluorescent plants can be selected conveniently to obtain transgene-free (T-DNA-free) mutant plants through outcrossing. Furthermore, the DsRed1 marker allows the convenient identification of transgenic hairy roots obtained through A. rhizogenes-mediated transformation. However, the efficiency of CRISPR/Cas9-mediated gene editing in hairy roots and their regeneration to whole M. truncatula plants appears to be relatively low in various studies published so far [33,35,36,37,76]. According to our experience, the large-scale regeneration of hairy roots to whole M. truncatula plants under glufosinate selection conditions is challenging and less efficient than the regeneration procedure with A. tumefaciens-transformed leaf explants described in this study (unpublished observations).

The MtMFS1 and MtMFS2 genes of M. truncatula mutated in this study encode major facilitator superfamily transporters [77], which are predicted to import or secrete N-acetylglucosamine. This sugar has been hypothesized to act as a symbiotic signal in the association between AM fungi and monocots (maize and rice), in which N-acetylglucosamine transporters (homologous to MtMFS1 and MtMFS2) have been identified [70]. It is worth mentioning in this context that M. truncatula mutants carrying a Tnt1 retrotransposon in a β-N-acetylhexosaminidase gene (MtHEXO2) exhibit reduced colonization by AM fungi. MtHEXO2 is an extracellular enzyme that releases N-acetylglucosamine from fungal chitooligosaccharide signals [69]. Future research will reveal whether MtMFS1 and MtMFS2 are N-acetylglucosamine transporters and whether they are required for the establishment of AM symbiosis.

The MtLYK10 gene encodes a LysM-type receptor kinase showing homology to exopolysaccharide/glycan receptors of L. japonicus involved in nodule formation (LjEPR3 receptor; [78,79,80]) and in symbiosis with AM fungi (LjEPR3a receptor [81]). Our results that homozygous lyk10 mutants exhibit delayed nodule formation in M. truncatula are consistent with findings on a MtLYK10 mutant reported previously [39]. However, since only a single Tnt1 retrotransposon insertion mutant was analyzed in that earlier study, the conclusions regarding the symbiotic role of MtLYK10 had to be interpreted with caution. In our study, three independent lyk10 mutants inoculated with S. meliloti showed a similar symbiotic phenotype. Thus, our results clearly show that MtLYK10 is a symbiotic gene that positively affects nodule development during early symbiotic stages. Interestingly, when compared to wild-type plants, the size of the rare nodules formed by the three lyk10 mutant plants was increased, suggesting that MtLYK10 negatively affects nodule size in mature nodules. Alternatively, the increased nodule size was the result of a compensation reaction, as the nodules of poorly nodulated root systems may tend to have a larger volume and more infected cells containing nitrogen-fixing bacteroids. Future experiments are required to investigate whether MtLYK10 can recognize rhizobial glycans such as exopolysaccharide (succinoglycan) [82], cyclic β-glucan [83], and mixed-linkage β-glucan [84].

5. Conclusions

This article shows that MtLYK10 plays a role in nodule symbiosis and that the used CRISPR/Cas9 technology is suitable for efficient gene editing in the model legume M. truncatula. The CRISPR/Cas9-mediated generation of mutant lines makes it possible to characterize, with relatively little effort, the function of M. truncatula genes for which no or only one gene knockout mutant was obtained in earlier mutagenesis programs [85]. The constructed pISV-CRISPR/Cas9P_GmUbi, pUC-tracrRNA-P_MtU6-1, and pUC-tracrRNA-P_MtU6-26 vectors were deposited in the Addgene repository [86] and are available to the scientific community. Our data show that the leaf transformation, glufosinate selection, and regeneration procedures from leaf explants to whole M. truncatula plants are highly efficient but require several months to obtain seeds. Future optimization steps should focus on technical parameters that could further standardize and accelerate these processes. Finally, the pre-testing of binary vectors with different sgRNA sequences in N. benthamiana should be considered in future experiments, especially if the simultaneous editing of multiple genes is planned.