Lotus Accessions Possess Multiple Checkpoints Triggered by Different Type III Secretion System Effectors of the Wide-Host-Range Symbiont Bradyrhizobium elkanii USDA61

Abstract

Bradyrhizobium elkanii, a rhizobium with a relatively wide host range, possesses a functional type III secretion system (T3SS) that is involved in symbiotic incompatibility against Rj4-genotype soybean (Glycine max) and some accessions of mung bean (Vigna radiata). To expand our knowledge on the T3SS-mediated partner selection mechanism in the symbiotic legume-rhizobia association, we inoculated three Lotus experimental accessions with wild-type and T3SS-mutant strains of B. elkanii USDA61. Different responses were induced by T3SS in a host genotype-dependent manner. Lotus japonicus Gifu inhibited infection; L. burttii allowed infection, but inhibited nodule maturation at the post-infection stage; and L. burttii and L. japonicus MG-20 both displayed a nodule early senescence-like response. By conducting inoculation tests with mutants of previously reported and newly identified effector protein genes of B. elkanii USDA61, we identified NopF as the effector protein triggering the inhibition of infection, and NopM as the effector protein triggering the nodule early senescence–like response. Consistent with these results, the B. elkanii USDA61 gene for NopF introduced into the Lotus symbiont Mesorhizobium japonicum induced infection inhibition in L. japonicus Gifu, but did not induce any response in L. burttii or L. japonicus MG-20. These results suggest that Lotus accessions possess at least three checkpoints to eliminate unfavorable symbionts, including the post-infection stage, by recognizing different T3SS effector proteins at each checkpoint.

Legume plants and rhizobia establish symbiosis in a unique host plant organ, the root nodule, in which rhizobia convert atmospheric dinitrogen to ammonium. Rhizobia enter the plant root hairs and develop infection threads, and rhizobia are then released from the elongated infection threads into host cells in nodule primordia. In this process, host plant roots secrete flavonoids, which activate the rhizobial transcription factor NodD. This factor induces the expression of nodulation (nod) rhizobial genes that are needed to produce Nod factors (NFs) (Göttfert, 1993; Perret et al., 2000), which are lipochitooligosaccharides with various chemical modifications depending on the rhizobial species (Dénarié and Cullimore, 1993; Perret et al., 2000). The perception of NFs by host NF receptors induces signal transduction cascades that result in nodule formation (Limpens et al., 2003; Radutoiu et al., 2007). Thus, interactions through the recognition of flavonoids and NFs are important for mutual recognition. In addition, rhizobial surface polysaccharides, such as lipopolysaccharides and exopolysaccharides, and their receptors in the host plant play important roles in partner selection (Becker et al., 2005; Kawaharada et al., 2015; Kawaharada et al., 2017). Rhizobial proteins secreted by type III and IV secretion systems affect the efficiency of host plant infection (Saeki, 2011; Miwa and Okazaki, 2017).

In pathogenic bacteria, the type III secretion system (T3SS) directly injects T3SS effector proteins (T3SEs) into host cells to suppress host innate immune responses (Block et al., 2008). To counteract pathogenic T3SEs, host cells have resistance (R) genes, such as nucleotide-binding site-leucine-rich repeat (NBS-LRR)-type genes (Gassmann and Bhattacharjee, 2012). The encoded proteins recognize T3SEs directly or indirectly, and activate immune responses called effector-triggered immunity (Gassmann and Bhattacharjee, 2012). Genome sequencing revealed that T3SS-related genes are conserved in a wide range of rhizobia including Sinorhizobium fredii NGR234 (Freiberg et al., 1997), S. fredii USDA257 (Krishnan et al., 2003), S. fredii HH103 (de Lyra Mdo et al., 2006), Mesorhizobium japonicum MAFF303099 (reclassified from M. loti) (Kaneko et al., 2000), Bradyrhizobium diazoefficiens USDA110 (Kaneko et al., 2002), and B. elkanii USDA61 (Okazaki et al., 2009). The genes encoding the rhizobial T3SS machinery components are called rhc (rhizobium conserved), and T3SE genes are referred to as nop (nodulation outer protein). The rhc gene cluster and the majority of nop genes are generally located in a symbiotic island or symbiotic plasmid (Freiberg et al., 1997; Kaneko et al., 2000; 2002; 2011), and the expression of these genes is controlled by NodD through the induction of the transcriptional activator TtsI (Krause et al., 2002). TtsI activates the expression of rhc and nop genes through a unique cis element in their promoter regions called the tts box (Marie et al., 2004; Wassem et al., 2008). Rhizobial T3SEs have been reported to exert beneficial effects in the infection stage. For example, NopL of S. fredii NGR234 interferes with host mitogen-activated protein kinase signaling and suppresses defense reactions (Bartsev et al., 2004; Zhang et al., 2011). On the other hand, rhizobial T3SS is involved in incompatibility depending on the host plant genotype. In soybean (Glycine max), Rj alleles (Rj2/Rfg1, Rj3, and Rj4) restrict nodulation with specific rhizobial strains (Okazaki et al., 2009; Hayashi et al., 2012; Yasuda et al., 2016; Sugawara et al., 2018). The Rj2 allele makes soybean incompatible with B. diazoefficiens USDA122; this incompatibility is triggered by NopP (Sugawara et al., 2018). The Rj4 allele restricts soybean nodulation with B. elkanii USDA61 in a T3SS-dependent manner (Okazaki et al., 2009; Yasuda et al., 2016). The T3SS of B. elkanii USDA61 is also involved in incompatibility with mung bean (Vigna radiata) cultivar KPS1 (Okazaki et al., 2009). Despite accumulating evidence for partner selection depending on rhizobial T3SS, the underlying mechanisms remain unclear, such as the timing of effector recognition by host plants, whether T3SE conserved in several rhizobial strains affects partner selection in a single host plant, and if the same T3SE influences partner selection in different host plants.

B. elkanii is a microsymbiont with a relatively wide host range and induces nodules on soybean, V. radiata, Arachis hypogaea (peanut or groundnut), V. unguiculata (cowpea), and Macroptilium atropurpureum (siratro). B. elkanii USDA61 produces at least 10 types of NFs, including one with a similar structure to one of the NFs of the Lotus symbiont M. japonicum. (Carlson et al., 1993; Niwa et al., 2001). Although B. elkanii USDA61 cannot form nodules on L. japonicus Gifu accession, as reported previously (Kelly et al., 2018), we demonstrated that B. elkanii USDA61 induced nodules on L. japonicus MG-20 accession. To expand our knowledge on the T3SS-mediated partner selection mechanism, we herein focused on the model legume L. japonicus and a related species, L. burttii, and inoculated them with wild-type B. elkanii USDA61 and its T3SS-deficient mutant.

Materials and Methods

Bacterial strains

The bacterial strains and plasmids used in the present study are listed in Tables 1 and 2. B. elkanii strains and M. japonicum strains were cultured at 28°C in arabinose–gluconate (AG) medium (Sadowsky et al., 1987) or tryptone-yeast extract medium (Beringer, 1974) supplemented with appropriate antibiotics (Table 1). Escherichia coli strains were cultured at 37°C in Luria–Bertani medium (Sambrook and Russell, 2001) supplemented with appropriate antibiotics (Table 1).

Table 1. Bacterial strains used in the present study.

Bacterial strains	Characteristics or sequencea	Reference or source
Bradyrhizobium elkanii
USDA61	Wild-type strain, Polr	Keyserb
BErhcJ	USDA61 derivative harboring an insertion in the rhcJ region; Polr, Kmr, Tcr	Okazaki et al., 2009
BEnopL	USDA61 derivative harboring an insertion in the nopL region; Polr, Kmr, Tcr	This study
BEnopP1	USDA61 derivative harboring an insertion in the nopP1 region; Polr, Kmr, Tcr	This study
BEnopP2	USDA61 derivative harboring an insertion in the nopP2 region; Polr, Kmr, Tcr	This study
BEnopM	USDA61 derivative harboring an insertion in the nopM region; Polr, Kmr, Tcr	This study
BEbe61_78180	USDA61 derivative harboring an insertion in the BE61_78180 region; Polr, Kmr, Tcr	This study
BEnopF	USDA61 derivative harboring an insertion in the nopF region; Polr, Kmr, Tcr	This study
14k062	Field-isolated B. elkanii strain; Polr	This study
USDA61-DsRed	DsRed-labeled USDA61; Polr, Spr, Smr	Yasuda et al., 2016
BErhcJ-DsRed	DsRed-labeled BErhcJ; Polr, Kmr, Tcr, Spr, Smr	Yasuda et al., 2016
BEnopF-DsRed	DsRed-labeled BEnopF; Polr, Kmr, Tcr, Spr, Smr	This study
14k062-DsRed	DsRed-labeled 14k062; Polr, Spr, Smr	This study
BEnopF-C	BEnopF derivatives complemented with the plasmid pS18mob-nopF; Polr, Kmr, Spr, Smr,	This study
BEnopM-C_1 and BEnopM-C_2	BEnopM derivatives complemented with the plasmid Tn5::nopM; Polr, Kmr, Spr, Smr,	This study
Mesorhizobium japonicum
MAFF303099	Wild-type strain; Pmr	Saeki and Kouchi, 2000
M. japonicum-DsRed	DsRed-labeled MAFF303099; Pmr	Maekawa et al., 2009
M. japonicum-EV	MAFF303099 carrying pHC60; Pmr, Tcr	Kindly provided by Dr. Yoshikazu Shimoda, National Agriculture and Food Research Organization, Japan
M. japonicum-BenopF	MAFF303099 carrying BenopF-integrated pHC60; Pmr, Tcr	This study
M. japonicum-BenopM	MAFF303099 carrying BenopM-integrated pHC60; Pmr, Tcr	This study
DT3S	MAFF303099 derivative with genome deletions at positions 5,157,472 to 5,168,624 (mlr6342 to mlr8765), and inserted Kmr cassette; Pmr, Kmr	Okazaki et al., 2010
DT3S-BenopF	MAFF303099 DT3S carrying BenopF-integrated pHC60; Pmr, Kmr Tcr	This study
Escherichia coli
HB101	recA, hsdR, hsdM, pro; Smr	Invitrogen
DH5α	F-, Φ80dlacZΔM15, Δ(lacZYA-argF)U169, deoR, recA1, endA1, hsdR17(rK-, mK+), phoA, supE44, λ-, thi-1, gyrA96, relA1	Toyobo
S17-1	hsdR, pro, thi (RP4-2‍ ‍km::Tn7 tc::Mu, integrated into the chromosome); Smr, Spr	Simon et al., 1983

^a Pol^r, polymyxin resistant; Km^r, kanamycin resistant; Sm^r, streptomycin resistant; Sp^r, spectinomycin resistant; Pm^r, phosphomycin resistant; Tc^r, tetracycline resistant; Ap^r, ampicillin resistant.

^b United States Department of Agriculture, Beltsville, MD.

Plant materials, growth conditions, and inoculation

L. japonicus experimental accessions Gifu (B-129) and MG-20, L. burttii, and the nfr1 mutant (Gifu background) were used (Hashiguchi et al., 2018). Lotus seeds were scarified with sandpaper, sterilized with solution containing 2% (v/v) sodium hypochlorite and 0.02% (v/v) Tween-20 for 10‍ ‍min, rinsed five times with sterilized distilled water, and germinated on 0.8% (w/v) agar plates for 3‍ ‍d in the dark, followed by 1‍ ‍d under light with a 16-h day and 8-h night condition. Seedlings were transferred to autoclaved vermiculite in inoculation pots (7 to 16 plants pot^–1) with nitrogen-free B&D medium (Broughton and Dilworth, 1971).

Rhizobial cultures were incubated for 3 to 5‍ ‍d, centrifuged (8,000×g, room temperature, 2‍ ‍min), washed three times with sterilized distilled water, suspended in nitrogen-free B&D medium, and 20 mL of the inoculant (OD₆₀₀=0.1) was then added to each pot containing seedlings. Plants were grown in a growth chamber at 25°C with a 16-h day and 8-h night condition. Nodule numbers and nodule fresh weights were measured on day 30 post-inoculation.

Analysis of proteins secreted by T3SS

AG medium (Sadowsky et al., 1987) (120 mL) inoculated with a 1:100 dilution of the B. elkanii preculture was incubated at 28°C for 48 h in the presence of 10‍ ‍μM genistein, which activates the expression of the T3SS machinery and T3SE genes in Bradyrhizobium species (Hempel et al., 2009). Extracellular proteins were collected from the culture supernatant as follows. The bacterial culture was centrifuged twice at 4°C (4,000×g for 1 h; 8,000×g for 30‍ ‍min) to remove cells and exopolysaccharides, and 100 mL of the culture supernatant was collected. Aliquots of the supernatant (25 mL each) were dispensed into two 50-mL centrifuge tubes, and 7.5 mL of Tris-EDTA-saturated phenol and 1 mL of 1 M dithiothreitol were then added. The mixture was vortexed and centrifuged (10,000×g, room temperature, 30‍ ‍min). The water phase was removed, and the remaining culture supernatant (25 mL each) and 1 mL of 1 M dithiothreitol were added to the phenol phase. The mixture was vortexed and centrifuged again (10,000×g, room temperature, 30‍ ‍min). The phenol phase was collected and mixed with 20 mL methanol, 300‍ ‍μL 8 M ammonium acetate, and 400‍ ‍μL 1 M dithiothreitol. The mixture was incubated at –20°C overnight. Proteins were pelleted by centrifugation (10,000×g, 4°C, 1 h), washed with 70% ethanol, combined from both tubes into a 5-mL Eppendorf tube, and suspended in 50‍ ‍μL phosphate-buffered saline. Protein concentrations were measured using a Bradford-based method. Proteins (5‍ ‍μg) were then separated by SDS-PAGE in precast 5–20% gradient gels (HOG-0520; Oriental Instruments) and stained with Coomassie Brilliant Blue R250. T3SS-dependent bands were subjected to in-gel digestion with trypsin. A thin matrix layer was made with 1‍ ‍μL of α-cyano-4-hydroxy-cinnamic acid (CHCA) solution (1 mg mL^–1 CHCA in 50% acetonitrile containing 0.1% TFA and 25‍ ‍mM ammonium bicarbonate) on a sample plate (Sciex). Aliquots (1‍ ‍μL) of tryptic peptides were dropped onto the thin layer, air dried, and covered with 1‍ ‍μL of CHCA solution. Mass spectrometry was performed on a TOF/TOF 5800 mass spectrometer (Sciex). Database searches for protein identification were performed using MS-Fit (http://prospector.ucsf.edu) and the BE61 protein database (Kaneko et al., unpublished).

In a large-scale analysis of extracellular proteins with the iTRAQ system, B. elkanii USDA61 and the rhcJ gene disruptant (BErhcJ) were cultured in the presence of 10‍ ‍μM genistein. Extracellular proteins (20‍ ‍μg) from each culture supernatant were labeled using an iTRAQ Reagent Multi-Plex Kit (Sciex). Proteins were digested with trypsin and the labeled peptides were loaded onto a cation exchange spin column (Viva S Mini H; Sartorius) and eluted with 150‍ ‍mM or 1 M KCl in 10‍ ‍mM potassium phosphate and 20% (v/v) acetonitrile at pH 3.0. Acetonitrile was evaporated and aliquots were loaded onto a C18 tip column (Rappsilber et al., 2007) for desalting and stored on the column at –80°C until used. Peptides were separated on a C18W-3 column (DiNa nano-LC system; KYA Technologies). A mass spectrometric analysis was performed using TOF/TOF 5800. The database search and relative quantitation were performed using ProteinPilot (Sciex) and the BE61 protein database.

Construction of bacterial strains

DsRed-labeled rhizobial strains were constructed using a previously described method (Hayashi et al., 2014) and the DsRed transposon delivery vector pBjGroEL4::DsRed2, in which the DsRed-coding sequence was fused to the promoter region of the groEL4 gene from B. japonicum USDA110, and groEL4 promoter-driven DsRed was integrated into the mini transposon mini-Tn5.

B. elkanii nopL, nopP1, nopP2, and nopM mutants were constructed by single crossover recombination as described previously (Faruque et al., 2015) with the primers listed in Table 2. B. elkanii Be61_78180 and nopF mutants were constructed as follows. The internal regions of Be61_78180 and BenopF were amplified by a polymerase chain reaction (PCR) with the primer pairs BE61_78180int-NotI-L and BE61_78180int-SpeI-R, and nopFint-NotI-L and nopFint-SpeI-R, respectively (Table 2). PCR products were digested with the restriction enzymes NotI and SpeI and cloned into the NotI and SpeI sites of the pSUPSCAKm vector (Okazaki et al., 2009). The plasmids obtained (pSUPSCAKm-Be61_78180 and pSUPSCAKm-nopF, respectively) were introduced into E. coli DH5α (Toyobo). To transfer the two plasmids into B. elkanii USDA61, we used a bacterial conjugation system as follows. Each bacterial culture (1 mL) was centrifuged (8,000×g, room temperature, 2‍ ‍min), and the pellet was washed twice with sterilized distilled water and suspended in 1 ml of AG medium. One hundred microliters of the donor strain (E. coli harboring pSUPSCAKm-Be61_78180 or pSUPSCAKm-nopF), 100‍ ‍μL of the helper strain (E. coli harboring pRK2013), and 300‍ ‍μL of the recipient strain (B. elkanii USDA61) were mixed and centrifuged (8,000×g, room temperature, 2‍ ‍min). Each pellet was suspended in 60‍ ‍μL AG medium, dropped onto an AG plate, and incubated at 28°C for 2 d. Cells were collected and single-crossover mutants were selected on an AG plate containing 50‍ ‍μg mL^–1 polymyxin and 200‍ ‍μg mL^–1 kanamycin. The integration of pSUPSCAKm-BE61_78180 and pSUPSCAKm-nopF into the internal regions of BE61_78180 and BenopF, respectively, was confirmed by PCR.

Table 2. Plasmids and primers used in the present study.

Plasmids and primers	Characteristics or sequencea	Reference or source
Plasmids
pRK2013	ColE1 replicon carrying RK2 transfer genes; Kmr, tra	Figurski and Helinski, 1979
pSUPSCAKm	Derivative of pSUPPOL2SCA with a kanamycin-resistant gene in the DraI site, oriT of RP4; Kmr, Tcr	Okazaki et al., 2009
pS18mob	Derivative of pK18mob with a aadA in the position of kanamycin-resistant gene, oriT of RP4; Smr, Spr	Okazaki et al., unpublished
pS18mob-nopF	pS18mob carrying a 2.0-kb DNA fragment containing a non-coding region and nopF and its upstream tts box region; Spr, Smr	This study
pBjGroEL4::DsRed2	DsRed transposon delivery vector; Spr, Smr, Apr	Hayashi et al., 2014
Tn5::nopM	pBjGroEL4::DsRed2 carrying a 2.9-kb DNA fragment containing nopM and its upstream tts box region; Spr, Smr, Apr	This study
pHC60	GFP constitutively-expressing vector; Tcr	Cheng and Walker, 1998
pHC60-BenopM	pHC60 carrying BenopM and its upstream tts box promotor; Tcr	This study
pHC60-BenopF	pHC60 carrying BenopF and its upstream tts box promotor; Tcr	This study
Primers
nopL_F	5′-ACCGCGGTGGCGGCCAACTCAATCAGCCCAACG-3′	This study
nopL_R	5′-CGGGGGATCCACTAGTATGAAACGCTCGTCCTCGG-3′	This study
nopP1_F	5′-ACCGCGGTGGCGGCCTATTCCCTCGTGACCAAGCC-3′	This study
nopP1_R	5′-CGGGGGATCCACTAGCGCTATTCGTTGTCCATTTG-3′	This study
nopP2_F	5′-ACCGCGGTGGCGGCCATCGCTCTTCCTTCAATGAC-3′	This study
nopP2_R	5′-CGGGGGATCCACTAGTATCACCATCCCCTGCCTTG-3′	This study
nopM_F	5′-ACCGCGGTGGCGGCCGCACTCCTTCGGGAACTTC-3′	This study
nopM_R	5′-CGGGGGATCCACTAGAGGTCGGGCAGATTGGTC-3′	This study
BE61_78180int-NotI-L	5′-ACGAAGCGGCCGCGAGAGTTCCGCAAAGTCGAG-3′	This study
BE61_78180int-SpeI-R	5′-TATCTACTAGTCAATTGAGGGCCTATCGTTG-3′	This study
nopFint-NotI-L	5′-ACGAAGCGGCCGCAGGTGTGTCAGTCCGCCTAC-3′	This study
nopFint-SpeI-R	5′-TATCTACTAGTAAATGACAGTCCGCATTTCC-3′	This study
nopM-NotI-L	5′-ATTAAGCGGCCGCTCAGAATAGGTGGGGACTCG-3′	This study
nopM-NotI-R	5′-TATCTGCGGCCGCTTTCCTTCACCGGGTATCTG-3′	This study
nopM-SacI-L	5′-ACGTCGAGCTCTCAGAATAGGTGGGGACTCG-3′	This study
nopM-SacI-R	5′-ATTGCGAGCTCTTTCCTTCACCGGGTATCTG-3′	This study
nopF-NotI-L	5′-ATTAAGCGGCCGCGTAAAGGACCGGCTCATGC-3′	This study
nopF-NotI-R	5′-TATCTGCGGCCGCCCCTCAGGCGCACTCTTAC-3′	This study
pS18mob_EcoR1_inf	5′-CCATGATTACGAATTGATTTGGAATTGCGCTTGAT-3′	This study
nopF_inf_1	5′-GAGCCGGTCCTTTACTTGATGAGCCTGATGTGAG-3′	This study
nopF_inf_2	5′-GTAAAGGACCGGCTCATG-3′	This study
nopF_inf_3	5′-TACCGAGCTCGAATTCCCTCAGGCGCACTCTTA-3′	This study
nopF_out_F	5′-CAGATGGTGCTGCTTTTACG-3′	This study
nopF_out_R	5′-CTCCATCTCGCCCATAAGAA-3′	This study
nopM_out_F	5′-TCAGAATAGGTGGGGACTCG-3′	This study
nopM_out_R	5′-TTTCCTTCACCGGGTATCTG-3′	This study

^a Km^r, kanamycin resistant; Sm^r, streptomycin resistant; Sp^r, spectinomycin resistant; Pm^r, phosphomycin resistant; Tc^r, tetracycline resistant; Ap^r, ampicillin resistant.

The BenopM and BenopF genes were introduced separately into M. japonicum MAFF303099 or the T3SS-disrupted M. japonicum strain DT3S (Okazaki et al., 2010) as follows. The 2,939-bp fragment containing the coding region and tts box promoter region of BenopM and the 666-bp fragment containing those of BenopF were amplified by PCR using the primer pairs nopM-NotI-L and nopM-NotI-R, and nopF-NotI-L and nopF-NotI-R, respectively (Table 2). PCR products were digested with the restriction enzyme NotI and cloned into the NotI site of the GFP-expressing plasmid pHC60 (Cheng and Walker, 1998). The plasmids obtained (pHC60-BenopM and pHC60-BenopF, respectively) were introduced separately into E. coli DH5α and mobilized into M. japonicum MAFF303099 using the bacterial conjugation system described above. One-day post conjugation, transformants containing the BenopM or BenopF gene were selected on tryptone-yeast extract plates containing 100‍ ‍μg mL^–1 phosphomycin and 2.0‍ ‍μg mL^–1 tetracycline. The transfer of pHC60-BenopM or pHC60-BenopF was confirmed by PCR.

Regarding BenopM complementation, BenopM and its upstream tts box region were amplified by PCR with the primers nopM-SacI-R and nopM-SacI-L, and cloned into the mini-Tn5 region of the pBjGroEL4::DsRed2 plasmid. The resulting plasmid, Tn5::nopM, was mobilized into BEnopM using the bacterial conjugation system. The integration of mini-Tn5 containing BenopM into the USDA61 genome was confirmed by antibiotic resistance and PCR.

Regarding BenopF complementation, the BenopF gene, its promoter region, and a 1,281-bp non-coding region of the USDA61 genome were amplified by PCR using the primer sets pS18mob_EcoR1_inf, nopF_inf_1 and nopF_inf_2, and nopF_inf_3, and cloned into the EcoRI sites of the pS18mob plasmid (Okazaki et al., unpublished) using an In-fusion HD Cloning Kit (Takara Bio). The resulting plasmid, pS18mob-nopF, was mobilized into BEnopF using the bacterial conjugation system. The single-crossover recombination of pS18mob-nopF in the non-coding region was confirmed by antibiotic resistance and PCR.

Microscopy

Root nodules were observed under a stereomicroscope (SZ61; Olympus), and DsRed-fluorescent nodules, nodule sections, and infection threads under a fluorescence microscope (SMZ18; Nikon). Early infection events were observed under a confocal microscope (LSM800; Zeiss).

Data availability

Nucleotide sequences have been submitted to the DNA Data Bank of Japan (DDBJ) with the accession numbers LC471584 (Be61_78180), LC471585 (BenopM), and LC471586 (BenopF).

Results

T3SS of B. elkanii USDA61 induces three types of responses in Lotus accessions

To investigate the symbiotic potential of B. elkanii USDA61 for Lotus, we inoculated L. japonicus Gifu (B-129), L. japonicus MG-20 (Miyakojima), and L. burttii with wild-type USDA61. Only a few white nodules formed on Gifu, a few well-developed red nodules (mature nodules) formed on MG-20, and many small white nodules and few red nodules formed on L. burttii (Fig. 1A, B, C and G). Some of the developed and small nodules on L. japonicus MG-20 and L. burttii were brownish (Fig. 1H), resembling the phenotype of nodule early senescence (Yamaya-Ito et al., 2018). To test the effects of T3SS of B. elkanii USDA61, we inoculated Lotus accessions with BErhcJ, a strain carrying a mutation in the rhcJ gene encoding a T3SS machinery component (Okazaki et al., 2009). Mature nodules formed on all three Lotus accessions, indicating that phenotypic differences were caused by the T3SS of B. elkanii; however, nodule numbers and plant growth (fresh weights) were less than those induced by the inoculation with the Lotus symbiont M. japonicum MAFF303099 (Fig. 1D, E, F and S1). To investigate whether NFs are needed for nodulation by USDA61, we inoculated wild-type Gifu and the nod factor receptor 1 mutant (nfr1) with USDA61 or BErhcJ; the latter induced the formation of mature nodules in wild-type Gifu, but not in nfr1, indicating that nodulation by USDA61 depends on nod factor recognition (Fig. S2).

Fig. 1.

Phenotypes of Lotus accessions inoculated with B. elkanii USDA61, the T3SS-deficient BErhcJ mutant of B. elkanii, or M. japonicum MAFF303099. (A, B, and C) Nodule numbers and (D, E, and F) fresh weights of Lotus japonicus Gifu (A, D), L. burttii (B, E), and L. japonicus MG-20 (C, F) inoculated with the wild-type (USDA61) or T3SS-deficient mutant (BErhcJ) of B. elkanii USDA61, or Lotus symbiont M. japonicum MAFF303099 measured on day 30 post-inoculation. Measurements were performed three times with 6 to 16 plants each time. In panels B and C, brownish nodules were included in the count of white nodules. Error bars indicate standard deviations. The Student’s t-test was performed for fresh weight comparisons; ** P<0.01 vs. control (no inoculation). (G) Root nodules of the three Lotus accessions inoculated with the above bacteria. Scale bars=1 mm. (H) Brownish nodules of L. burttii and MG-20 inoculated with wild-type USDA61. Scale bars=1 mm.

The inoculation with DsRed-labeled USDA61 led to fluorescence of the entire nodules of L. japonicus MG-20 and a limited area of L. burttii nodules, but no clear fluorescence in L. japonicus Gifu (Fig. 2), as was reported previously for an inoculation of Rj4-genotype soybean in which infection inhibition was induced (Yasuda et al., 2016). The inoculation with DsRed-labeled BErhcJ led to fluorescence of the entire nodules in all three accessions (Fig. 2). By sectioning the nodules, it was confirmed that fluorescence observed in L. japonicus MG-20 and L. burttii inoculated with DsRed-labeled USDA61 as well as DsRed-labeled BErhcJ came from inside the nodules (Fig. S3). These results, together with those shown in Fig. 1, suggest that the T3SEs of USDA61 influence(s) responses at the post-infection stage, i.e., nodule maturation inhibition (rhizobia may infect, but nodules remain small and white even after day 30 post-inoculation) in L. burttii and a nodule early senescence–like response in L. burttii and L. japonicus MG-20.

Fig. 2.

Infection phenotypes of Lotus accessions inoculated with B. elkanii USDA61 or the T3SS-deficient mutant BErhcJ. L. japonicus Gifu, L. burttii, and L. japonicus MG-20 were inoculated with two DsRed-labeled rhizobial strains, and DsRed fluorescence in root nodules was observed under a fluorescence microscope on day 30 post-inoculation. Scale bars=1 mm.

NopM induces a nodule early senescence–like response

To identify the T3SEs involved in these checkpoint responses of the Lotus accessions, we investigated the interaction of the reported T3SEs of USDA61 with the Lotus accessions. Since NopP, NopL, and NopM were confirmed as T3SEs in USDA61 (Okazaki et al., 2009) at the time of the experiment, we constructed mutants of the two copies of BenopP, BEnopP1 and BEnopP2; a BenopL mutant, BEnopL; and a BenopM mutant, BEnopM. The inoculation with BEnopP1, BEnopP2, or BEnopL resulted in similar nodulation phenotypes of the three Lotus accessions to those induced by wild-type USDA61 (Fig. S4). However, L. burttii and MG-20 had fewer brownish nodules after the inoculation with BEnopM than with wild-type USDA61 (Fig. 3B, C, E, F, and G) or BEnopM complemented with the BenopM gene (Fig. S5). The inoculation with BEnopM slightly improved plant growth by L. burttii and L. japonicus MG-20 (Fig. S6). The inoculation with BEnopM did not change the nodulation phenotype of L. japonicus Gifu and did not alter the number of white nodules on L. burttii from those with the inoculation with wild-type USDA61 (Fig. 3A, D, E, and G). These results suggest that NopM induces a nodule early senescence-like response in L. burttii and L. japonicus MG-20, and that infection inhibition in L. japonicus Gifu and maturation inhibition in L. burttii are induced by T3SEs other than NopP, NopL, and NopM. The product of BenopM predicted from the genome sequence is composed of 610 amino acids. The domain organization and phylogenetic relationships of NopM are summarized in Fig. S7.

Fig. 3.

Characterization of the NopM protein of B. elkanii USDA61. Phenotypes of three Lotus accessions inoculated with wild-type B. elkanii USDA61, the BenopM mutant (BEnopM), or T3SS-deficient mutant (BErhcJ) were analyzed on day 30 post-inoculation. (A, B, and C) The number of brownish nodules and (D, E, and F) total number of mature and white nodules on the roots of (A, D) Gifu, (B, E) L. burttii, and (C, F) MG-20 are shown. ND means not detected. Inoculation tests were performed three times with 7 to 12 plants each time. In panels E and F, brownish nodules were included in the count of white nodules. Error bars indicate standard deviations. The Student’s t-test was performed for brownish nodule counts; ** P<0.01 vs. wild-type USDA61. (G) Root nodules of the three Lotus accessions inoculated with the above bacteria. Scale bars=1 mm.

Field-isolated B. elkanii strain lacking two effector proteins evades infection inhibition by L. japonicus Gifu

During the course of the large-scale field phenotyping of Lotus accessions (Shah et al., 2020), we isolated B. elkanii strains, confirmed by 16S rDNA sequences, from the nodules of L. japonicus accessions grown in the Kashimadai field (Osaki city, Miyagi, Japan) on which soybean had been cultivated over a three-year period. Strain 14k062 isolated from L. japonicus Gifu induced many white nodules on Gifu roots (Fig. 4A), resembling the phenotype of L. burttii inoculated with USDA61 (Fig. 1). The inoculation of Gifu with the DsRed-labeled 14k062 strain resulted in clear, but limited, areas of DsRed fluorescence within the nodules (Fig. 4B), similar to those in nodules on L. burttii inoculated with USDA61 (Fig. 2). On the other hand, the inoculation of L. burttii and L. japonicus MG-20 with 14k062 caused almost the same nodulation phenotypes as the inoculation with USDA61 (data not shown). These results suggest that 14k062 has the T3SS machinery, but lacks the T3SE(s) that trigger infection inhibition.

Fig. 4.

Symbiotic phenotype of B. elkanii 14k062. (A) Nodule number of L. japonicus Gifu inoculated with wild-type B. elkanii USDA61, 14k062 (a field-isolated strain of B. elkanii), or the T3SS-deficient mutant (BErhcJ) on day 30 post-inoculation. Nodulation tests were performed three times with 12 to 14 plants each time. Error bars indicate standard deviations. (B) Infection phenotypes of L. japonicus Gifu inoculated with DsRed-labeled USDA61, 14k062, or BErhcJ on day 30 post-inoculation. DsRed fluorescence was observed under a fluorescence microscope. Scale bars=1 mm.

To analyze T3SEs lacking in the 14k062 strain, we attempted to create a comprehensive list of T3SEs in USDA61 by comparing the extracellular proteins of wild-type USDA61 and BErhcJ using a MALDI-TOF-MS/MS analysis with the iTRAQ protein labeling system (Ross et al., 2004). We identified 9 candidate T3SEs based on their presence in wild-type USDA61 and low abundance or absence in BErhcJ (BErhcJ/wild-type USDA61 ratio <0.2) (Table 3 and matched peptides are shown in Table S1). These proteins included previously identified effector proteins (NopL, two isoforms of NopP, and NopM) and T3SS machinery components (NopA and NopX) (Okazaki et al., 2009). The new candidates were BE61_51850, BE61_76200, BE61_78180, BE61_78310, and BE61_91540 (Table 3). We then compared extracellular proteins between USDA61 and 14k062 by separating them electrophoretically, and confirmed the presence of most of the T3SE candidates in the 14k062 culture supernatant, indicating that 14k062 is not a mutant of a T3SS machinery component as expected (Fig. 5A). Two T3SE candidates were not detected in the 14k062 culture supernatant (Fig. 5A); using the MALDI-TOF-MS/MS analysis, we identified them as BE61_78180 and BE61_91540 (Table 3). The promoters of the corresponding two genes contained a typical tts box (Fig. 5B). A comparison of the amino acid sequence of BE61_91540 with the rhizobium genome data set in RhizoBase (http://genome.annotation.jp/rhizobase) revealed that BE61_91540 is conserved among B. diazoefficiens USDA110, B. diazoefficiens USDA122, and B. japonicum USDA6, and is annotated as T3SS-secreted protein NopF with no functional information (Hempel et al., 2009; Kimbrel et al., 2013; Tsukui et al., 2013). Based on these findings, we considered infection inhibition in L. japonicus Gifu to be triggered by BE61_78180 and/or NopF (BE61_91540).

Table 3. Extracellular protein analysis using the iTRAQ system.

Accession	Description	Molecular weight (kDa)	Total prot score	Coverage (%)ª	Peptidesb	Fold changec BErhcJ/WT
BE61_80730	Nodulation outer protein NopP	31.0	23.39	62.2	13	0.02
BE61_78180	Unknown protein	83.0	22.21	27.8	11	0.15
BE61_80150	Nodulation outer protein NopX	63.4	19.44	31.1	10	0.02
BE61_80320	Nodulation outer protein NopM	67.0	18.51	29.2	10	0.12
BE61_77110	Nodulation outer protein NopP	31.4	17.14	39.9	10	0.05
BE61_76200	Unknown protein	48.6	15.10	33.0	8	0.08
BE61_80180	Nodulation outer protein NopA	7.5	14.72	93.5	7	0.01
BE61_91540	Unknown protein (NopF)	19.2	8.01	45.3	4	0.02
BE61_78310	Unknown protein	19.9	8.00	25.0	3	0.04
BE61_51850	Unknown protein	33.1	7.72	30.8	3	0.11
BE61_80070	Nodulation outer protein NopL	24.6	6.16	29.7	3	0.01

^a Sequence coverage.

^b The total number of detected peptides (at the 95% confidence level) for each protein.

^c Fold changes in the T3SS mutant BErhcJ vs. wild-type USDA61.

Fig. 5.

Proteins secreted by B. elkanii strains. B. elkanii strains USDA61 and 14k062 and the T3SS-deficient mutant (BErhcJ) were grown in the presence of 10‍ ‍μM genistein. Supernatants containing extracellular proteins were collected, and proteins were separated by SDS-PAGE (5 to 20% gradient gel) and stained with Coomassie Brilliant Blue. Closed arrowheads indicate T3SS-dependent secreted proteins, and open arrowheads indicate proteins not detected in the 14k062 strain. (B) The tts box sequences of nopF and Be61_78180 of B. elkanii USDA61. In the consensus sequence, all invariant nucleotides are capitalized and lowercase letters are used for nucleotides conserved in at least 50% of the analyzed sequences (Krause et al., 2002). Nucleotides in common with the consensus sequence are shown in red.

NopF triggers infection inhibition in L. japonicus Gifu

To identify which of these two proteins triggers infection inhibition in L. japonicus Gifu, we constructed mutants of the Be61_78180 and BenopF genes, and used them to inoculate L. japonicus Gifu. The phenotype induced by the Be61_78180 mutant (BEbe61_78180) did not significantly differ from that induced by the USDA61 inoculation (Fig. 6A), whereas the BenopF mutant (BEnopF) induced many small white nodules (Fig. 6A), at a similar level to that induced by 14k062 (Fig. 4A). The inoculation of Gifu with DsRed-labeled BEnopF clearly showed the release of BEnopF into the nodules (Fig. 6B), whereas the nodules were rarely mature and most of them became brownish (Fig. 6C), as was the case in L. burttii inoculated with wild-type USDA61. These results suggest that L. japonicus Gifu inhibits nodule maturation and has an early senescence-like response. Consistent with this observation, the growth of L. japonicus Gifu inoculated with BEnopF was similar to that after the inoculation with wild-type USDA61 (Fig. S8). The inoculation of L. burttii with BEnopF, BEbe61_78180, or wild-type USDA61 resulted in a similar phenotype (Fig. S9), and L. burttii and MG-20 inoculated with BEnopF or BEbe61_78180 also showed a nodule early senescence-like response (Fig. S10). These results indicate that NopF, not BE61_78180 triggers infection inhibition in L. japonicus Gifu, and neither of these proteins triggers nodule maturation inhibition or an early senescence-like response. The complementation test on the BenopF gene confirmed that infection inhibition was induced by NopF (Fig. S11).

Fig. 6.

Characterization of NopF of B. elkanii USDA61. (A) Nodule numbers of L. japonicus Gifu inoculated with wild-type B. elkanii USDA61, Be61_78180 mutant (BEbe61_78180), BenopF mutant (BEnopF), or T3SS-deficient mutant (BErhcJ) on day 30 post-inoculation. Nodulation tests were performed at least twice with 9 to 12 plants each time. Error bars indicate standard deviations. (B) Infection phenotype of L. japonicus Gifu inoculated with DsRed-labeled BEnopF strain on day 30 post-inoculation. DsRed fluorescence was observed under a fluorescence microscope. Scale bars=1 mm. (C) Brownish nodules on L. japonicus Gifu inoculated with BEnopF. Scale bar=1 mm.

Nodulation restrictions against rhizobial T3SS are generally induced during infection thread formation (Yasuda et al., 2016). To investigate whether this is the case for infection inhibition induced by NopF, we inoculated Gifu with DsRed-labeled USDA61, BEnopF, BErhcJ, or M. japonicum MAFF303099, and counted infection threads on day 10 post-inoculation. Well-elongated infection threads were observed after the M. japonicum inoculation, whereas no infection threads were detected after the inoculation with B. elkanii strains, including BEnopF and BErhcJ (Fig. S12A, B, and C). Confocal observations of the nodule on day 14 post-inoculation with M. japonicum showed well-elongated infection threads on nodules, whereas the attachment of BEnopF and BErhcJ bacteria to the nodule surface and entry toward the nodule center were noted with no obvious infection threads (Fig. S12D). After the wild-type USDA61 inoculation, only the attachment of bacteria to the nodule surface was observed (Fig. S12D). This result suggests that B. elkanii infects Gifu by crack entry rather than through infection threads, and that NopF triggers infection inhibition in this process.

The predicted product of BenopF has 179 amino acids and belongs to the Acyl-CoA N-acyltransferase superfamily (InterPro ID; IPR016181). A BLASTP analysis showed that the amino acid sequence of NopF was identical to those of proteins encoded by the two gene copies in B. diazoefficiens USDA110 (Bll1862 and Bll8201) and USDA122 (BD122_09540 and BD122_41920), and by a single-copy gene in B. japonicum USDA6 (BJ6T_88790). A homolog of the nopF gene was not conserved in M. japonicum MAFF303099, similar to the nopM gene. NopF was 44% identical to the HopBG1 protein, a T3SE of the plant pathogen Pseudomonas syringae pv. maculicola ES4326 (Baltrus et al., 2011). The features of NopF are summarized in Fig. S13.

NopF secreted by M. japonicum MAFF303099 induces infection inhibition in L. japonicus Gifu

To elucidate the functions of NopM and NopF effector proteins, we introduced BenopM and BenopF cloned into the GFP constitutively-expressing plasmid pHC60 (Cheng and Walker, 1998) into M. japonicum MAFF303099 to generate M. japonicum-BenopM and M. japonicum-BenopF, respectively. We expected the two proteins to be functional because M. japonicum has the T3SS machinery (Kaneko et al., 2000; Okazaki et al., 2010), and B. elkanii USDA61 and M. japonicum MAFF303099 share a similar typical tts box promoter (Fig. S14). As a control, we used M. japonicum carrying empty pHC60 (M. japonicum-EV). On day 21 post-inoculation, the total nodule number and plant fresh weight of L. japonicus Gifu were markedly lower in plants inoculated with M. japonicum-BenopF than with M. japonicum-EV (Fig. 7A, B, and C). Microscopic observations revealed that the number of infection threads were lower with the M. japonicum-BenopF inoculation than with the M. japonicum-EV inoculation (Fig. S15). To confirm the T3SS dependence of NopF secretion by M. japonicum, we introduced the BenopF plasmid into the T3SS-disrupted M. japonicum strain, DT3S (Okazaki et al., 2010). Plants inoculated with DT3S-BenopF formed mature nodules and grew similarly to those inoculated with M. japonicum-EV (Fig. 7A, B, and C). These results indicate that introduced NopF was secreted into the host plant through M. japonicum T3SS, and its secretion triggered infection inhibition in L. japonicus Gifu. In L. burttii and L. japonicus MG-20, M. japonicum-BenopF induced mature nodules at the same level as M. japonicum-EV, and plant fresh weight did not significantly differ between the M. japonicum-BenopF and M. japonicum-EV inoculations, as expected from the phenotypes of wild-type USDA61 and the BEnopF inoculation (Fig. S16).

Fig. 7.

Symbiotic phenotypes of L. japonicus Gifu inoculated with M. japonicum MAFF303099 carrying NopF of B. elkanii USDA61. (A) Nodule numbers and (B) nodule fresh weights of plants inoculated with M. japonicum carrying the plasmid pHC60 (M. japonicum-EV), M. japonicum carrying pHC60-BenopF (M. japonicum-BenopF), or the M. japonicum T3SS-deficient mutant carrying pHC60-BenopF (DT3S-BenopF) on day 21 post-inoculation. All tests were performed three times with 11 to 12 plants each time. Error bars indicate standard deviations. ** P<0.01 vs. M. japonicum-EV in the Student’s t-test. (C) Plant growth phenotype of L. japonicus Gifu inoculated with the above bacteria. Scale bars=5 cm.

In contrast, mature nodules, but not brownish nodules, were induced by M. japonicum-BenopM in all three Lotus accessions (Fig. S17). This result suggests that the introduction of NopM did not occur in M. japonicum or that introduced NopM did not function in L. japonicus cells.

Discussion

In the present study, we performed inoculation tests on wild-type B. elkanii USDA61 and the T3SS machinery mutant BErhcJ against three Lotus accessions, and found accession-dependent responses triggered by T3SEs: infection inhibition in L. japonicus Gifu, nodule maturation inhibition in L. burttii, and a nodule early senescence-like response in L. burttii and L. japonicus MG-20. Although infection inhibition triggered by rhizobial T3SEs has been reported in soybean (Okazaki et al., 2009; Tsukui et al., 2013; Yasuda et al., 2016), we herein found nodulation restrictions triggered by rhizobial T3SEs at the post-infection stage. We identified NopF as a candidate trigger of infection inhibition and NopM as that of a nodule early senescence-like response. As indicated by plant phenotypes, B. elkanii and even its T3SS-deficient mutant BErhcJ exhibited a lower nitrogen fixation ability estimated by the plant growth phenotype than M. japonicum in combination with Lotus accessions (Fig. 1D, E, and F), suggesting that Lotus accessions use rhizobial T3SEs as markers of unfavorable rhizobial infection and have multiple checkpoints to eliminate rhizobia. B. elkanii strains, including 14k062, were isolated from the Lotus accessions grown in the field on which soybean had been cultivated over a three-year period. In the second year of the field experiment, all rhizobia strains isolated from the Lotus accessions grown in the same field became Mesorhizobium stains (unpublished data). This suggests that once the population of favorable rhizobia increased in the field, Lotus accessions may distinguish favorable and unfavorable rhizobium strains by recognizing T3SEs. A hypothetical model of the T3SS-mediated interaction between the Lotus accessions and B. elkanii USDA61 based on the results of the present study is shown in Fig. 8.

Fig. 8.

Model of the T3SS-mediated interaction between B. elkanii USDA61 and three Lotus accessions. Three types of Lotus responses—infection inhibition, nodule maturation inhibition, and a nodule early senescence-like response—may be caused by T3SS effectors of B. elkanii USDA61. L. japonicus Gifu has all three responses, L. burttii induces nodule maturation inhibition and nodule early senescence, and MG-20 induces nodule early senescence only. Infection inhibition is triggered by NopF; the nodule early senescence-like response is triggered by NopM of B. elkanii USDA61. Nodule maturation inhibition is triggered by other T3SS effector protein(s).

An InterPro scan analysis identified leucine-rich repeats (InterPro ID; IPR001611) and a novel E3 ubiquitin ligase domain (InterPro ID; IPR029487) in B. elkanii NopM and a BLASTP analysis showed that its homologs were conserved in Bradyrhizobium strains and Sinorhizobium strains, but not in Mesorhizobium strains (Fig. S7). The E3 ubiquitin ligase activity of NopM of S. fredii NGR234 was previously shown to reduce the flg22-triggered accumulation of reactive oxygen species (ROS) in Nicotiana benthamiana leaves, and the same NopM increased nodule numbers in Lablab purpureus (Xin et al., 2012). These findings imply that the E3 ubiquitin ligase activity of NopM of S. fredii NGR234 promotes symbiosis by reducing harmful ROS generation in host roots during nodule maturation or senescence. Since ROS accumulate in senescent nodules (Alesandrini et al., 2003; Cam et al., 2012), USDA61 may use NopM as a positive effector to counteract ROS accumulation during nodule senescence. On the other hand, Lotus accessions may detect NopM as a post-infection marker of unfavorable rhizobial infection and induce a nodule early senescence-like response. However, the function of NopM in nodule development has not yet been elucidated in detail.

The NopF protein is identical in different Bradyrhizobium species. Two copies of nopF genes were identified in the B. diazoefficiens USDA110 and USDA122 genomes. One copy (Bll1862 and BD122_09540, respectively) is located on symbiosis island A and the other (Bll8201 and BD122_41920) is located in a genome region highly conserved between B. diazoefficiens and B. japonicum USDA6 and reported as locus C (Kaneko et al., 2011), in which a single-copy nopF gene (BJ6T_88750) is located. BenopF is also located in the genome region corresponding to locus C. Conjugal transfer protein genes and the replication protein A gene have been identified in this locus (Kaneko et al., 2011); this implies that locus C is transferred between Bradyrhizobium species, similar to the symbiotic island. NopF conservation in Bradyrhizobium species, including those with two copies in the genome, indicates a strong selection pressure on this T3SE with possible significance in the life cycle of Bradyrhizobium. This conservation feature may also be advantageous for the use of NopF by host plants as a signal molecule for infection inhibition.

In soybean with the Rj4 allele, B. elkanii USDA61 was eliminated in a T3SS-dependent manner, and BEL2-5 was identified as a candidate T3SE triggering infection inhibition (Okazaki et al., 2009; Faruque et al., 2015). BEL2-5, encoded by Be61_51970, was identified in our MALDI-TOF-MS/MS analysis of extracellular proteins; however, we did not select it as a candidate T3SE because of the high BErhcJ/wild-type ratio (0.64). The InnB protein was recently identified as the T3SE triggering infection inhibition in mung bean (V. radiata cv. KPS1) (Nguyen et al., 2018). The InnB protein is encoded by Be61_78180, which we confirmed did not induce infection inhibition in L. japonicus Gifu. These results suggest that different T3SEs of B. elkanii USDA61—BEL2-5 (BE61_51970) in soybean, InnB (BE61_78180) in mung bean, and NopF (BE61_91540) in L. japonicus—are recognized by host plants and induce infection inhibition.

In the present study, we demonstrated not only infection inhibition, but also novel responses triggered by T3SS at the post-infection stage, i.e., nodule maturation inhibition and a nodule early senescence-like response, in Lotus accessions. Although these checkpoints have not been reported as T3SS-triggered reactions, nodule maturation inhibition is a typical phenotype observed in cases of nitrogen fixation deficiency caused by mutations in host plants or rhizobia (Krusell et al., 2005; Kumagai et al., 2007; Daubech et al., 2017). A previous study showed that the nifA and nifH mutants of S. meliloti died prematurely after bacteroid elongation in the host plant (Berrabah et al., 2015), suggesting that the host plant monitors the nitrogen fixation ability of symbionts and punishes those with nitrogen fixation deficiencies. A nodule early senescence-like response has been reported in L. japonicus MG-20 inoculated with Rhizobium etli CE3, which has a lower nitrogen-fixing ability than M. japonicum (Banba et al., 2001). Therefore, these post-infection checkpoints may be conserved for monitoring the nitrogen fixation level of symbionts and T3SE recognition.

The T3SEs involved in nodule maturation inhibition in the Lotus accessions remain to be identified; however, some T3SEs that inhibit nodule development have been reported. NopT of S. fredii NGR234, a C58 cysteine protease with amino acid sequence similarity to AvrPphB of P. syringae pv. phaseolicola, reduced its nodule number and nodule dry weight following an inoculation with Crotalaria juncea (Dai et al., 2008). NopE of B. diazoefficiens USDA110, a T3SE containing two EF-hand–like calcium-binding motifs, reduced nodulation efficiency in V. radiata KPS2 (Hempel et al., 2009; Wenzel et al., 2010). Among the 9 candidate T3SEs of B. elkanii USDA61 identified in the present study (Table 3), none of the single gene disruptants tested (BenopF, BenopL, BenopM, BenopP1, BenopP2, or Be61_78180) affected the nodule maturation inhibition phenotype in L. burttii and MG-20. Thus, a future disruption analysis of the remaining three candidate genes and/or multiple gene disruption may contribute to identifying the T3SE(s) triggering nodule maturation inhibition.

The inoculation of L. japonicus Gifu with M. japonicum carrying BenopF, but not with T3SS-disrupted M. japonicum carrying BenopF markedly reduced nodule numbers and nodule fresh weights (Fig. 7), suggesting that NopF is produced and secreted by the T3SS of M. japonicum. Infection inhibition caused by NopF suppressed the stable symbiont, M. japonicum, although one or two mature nodules in each plant were occasionally observed. The presence of NopF alone is sufficient to trigger infection inhibition, implying that this T3SE functions in different rhizobial strains. Although the present results may reflect the functional expression of a T3SE in a different rhizobial genus, the secretion of B. japonicum Bll8244 has been reported in S. fredii HH103 (Yang et al., 2010). As demonstrated by the introduction of BenopF into M. japonicum, it may be possible to exchange T3SEs between different rhizobial species. For example, it may be feasible to increase the symbiotic potentials of target strains by introducing NopL of S. fredii NGR234, which interferes with host mitogen-activated protein kinase signaling and suppresses defense reactions (Bartsev et al., 2004; Zhang et al., 2011).

The inoculation of M. japonicum carrying BenopM did not induce early senescence in the Lotus accessions tested. Based this result, we propose two hypotheses. The first hypothesis is that the expression timing of tts box-regulated genes may differ between M. japonicum and B. elkanii after infection. Okazaki et al. showed that B. elkanii USDA61 secreted T3SEs without the addition of genistein (Okazaki et al., 2009), suggesting that the flavonoid signal is not essential for T3SS activation in this strain, and the continuous activity of T3SS in B. elkanii USDA61 may be expected at the post-infection stage. While B. elkanii constitutively secretes T3SEs, M. japonicum did not secrete T3SEs at the post-infection stage due to the strict regulation of tts box-regulated genes. Another hypothesis is that NopM may function together with additional B. elkanii T3SE(s). In a previous study, T3SS of S. fredii NGR234 positively affected symbiosis with T. vogelii, and the nopL and nopP double mutant reduced nodule numbers more than a nopP single effector mutant (Skorpil et al., 2005), suggesting that the positive effects of this interaction were induced by at least two T3SEs. The absence of the effects of NopM on the early senescence-like response in M. japonicum indicates the requirement for additional B. elkanii T3SE(s).

In the present study, we demonstrated that Lotus accessions have at least three checkpoints to eliminate B. elkanii USDA61, and they are regulated by different T3SEs. In addition to infection inhibition, we revealed that nodule maturation inhibition and a nodule early senescence-like response were triggered by T3SEs at the post-infection stage. The present results indicate that leguminous plants continue to recognize rhizobial T3SEs after intracellular infection and attempt to eliminate unfavorable rhizobial strains. In nature, there are risks to host plants associated with infection by inefficient rhizobia that have low or no nitrogen-fixing ability, but produce host-compatible NFs. The present results suggest that host plants use rhizobial T3SEs to monitor unfavorable rhizobia throughout nodulation.

Acknowledgements

Accessions of L. japonicus were provided by the National BioResource Project ‘Lotus/Glycine’. We thank Dr. Masayoshi Kawaguchi (National Institute for Basic Biology, Japan) for L. japonicus nfr1 mutant seeds, Dr. Hisayuki Mitsui (Tohoku University, Japan) for the GFP-labeling vector pHC60, and Dr. Yoshikazu Shimoda (National Agriculture and Food Research Organization, Japan) for GFP-labeled M. japonicum strains. We thank Dr. Tomomi Nakagawa (National Institute for Basic Biology, Japan) for valuable discussions and comments on the study. We thank Ms. Chikako Mitsuoka and Ms. Sakuya Nakamura (Tohoku University, Japan) for their technical assistance. This work was supported by JSPS KAKENHI Grants JP 616J020580 (to SK) and JP 26650089 (to SS).

References

•  Alesandrini,  F.,  Mathis,  R.,  Van de Sype,  G.,  Hérouart,  D., and  Puppo,  A. (2003) Possible roles for a cysteine protease and hydrogen peroxide in soybean nodule development and senescence. New Phytol 158: 131–138.
•  Baltrus,  D.A.,  Nishimura,  M.T.,  Romanchuk,  A.,  Chang,  J.H.,  Mukhtar,  M.S.,  Cherkis,  K., et al. (2011) Dynamic evolution of pathogenicity revealed by sequencing and comparative genomics of 19 Pseudomonas syringae isolates. PLoS Pathog 7: e1002132.
•  Banba,  M.,  Siddique,  A.B.,  Kouchi,  H.,  Izui,  K., and  Hata,  S. (2001) Lotus japonicus forms early senescent root nodules with Rhizobium etli. Mol Plant-Microbe Interact 14: 173–180.
•  Bartsev,  A.V.,  Deakin,  W.J.,  Boukli,  N.M.,  McAlvin,  C.B.,  Stacey,  G.,  Malnoë,  P., et al. (2004) NopL, an effector protein of Rhizobium sp. NGR234, thwarts activation of plant defense reactions. Plant Physiol 134: 871–879
•  Becker,  A.,  Fraysse,  N., and  Sharypova,  L. (2005) Recent advances in studies on structure and symbiosis-related function of rhizobial K-antigens and lipopolysaccharides. Mol Plant-Microbe Interact 18: 899–905.
•  Beringer,  J.E. (1974) R factor transfer in Rhizobium leguminosarum. J Gen Microbiol 84: 188–198.
•  Berrabah,  F.,  Ratet,  P., and  Gourion,  B. (2015) Multiple steps control immunity during the intracellular accommodation of rhizobia. J Exp Bot 66: 1977–1985.
•  Block,  A.,  Li,  G.,  Fu,  Z.Q., and  Alfano,  J.R. (2008) Phytopathogen type III effector weaponry and their plant targets. Curr Opin Plant Biol 11: 396–403.
•  Broughton,  W.J., and  Dilworth,  M.J. (1971) Control of leghaemoglobin synthesis in snake beans. Biochem J 125: 1075–1080.
•  Cam,  Y.,  Pierre,  O.,  Boncompagni,  E.,  Hérouart,  D.,  Meilhoc,  E., and  Bruand,  C. (2012) Nitric oxide (NO): a key player in the senescence of Medicago truncatula root nodules. New Phytol 196: 548–560.
•  Carlson,  R.W.,  Sanjuan,  J.,  Bhat,  U.R.,  Glushka,  J.,  Spaink,  H.P.,  Wijfjes,  A.H., et al. (1993) The structures and biological activities of the lipo-oligosaccharide nodulation signals produced by type I and II strains of Bradyrhizobium japonicum. J Biol Chem 25: 18372–18381.
•  Cheng,  H.P., and  Walker,  G.C. (1998) Succinoglycan is required for initiation and elongation of infection threads during nodulation of alfalfa by Rhizobium meliloti. J Bacteriol 180: 5183–5191.
•  Dai,  W.J.,  Zeng,  Y.,  Xie,  Z.P., and  Staehelin,  C. (2008) Symbiosis-promoting and deleterious effects of NopT, a novel type 3 effector of Rhizobium sp. strain NGR234. J Bacteriol 190: 5101–5110.
•  Daubech,  B.,  Remigi,  P.,  Doin de Moura,  G.,  Marchetti,  M.,  Pouzet,  C.,  Auriac,  M.C., et al. (2017) Spatio-temporal control of mutualism in legumes helps spread symbiotic nitrogen fixation. eLife 6: e28683.
•  de Lyra Mdo,  C.,  López-Baena,  F.J.,  Madinabeitia,  N.,  Vinardell,  J.M.,  Espuny,  M.R.,  Cubo,  M.T., et al. (2006) Inactivation of the Sinorhizobium fredii HH103 rhcJ gene abolishes nodulation outer proteins (Nops) secretion and decreases the symbiotic capacity with soybean. Int Microbiol 9: 125–133.
•  Dénarié,  J., and  Cullimore,  J. (1993) Lipo-oligosaccharide nodulation factors: a minireview new class of signaling molecules mediating recognition and morphogenesis. Cell 74: 951–954.
•  Faruque,  O.M.,  Miwa,  H.,  Yasuda,  M.,  Fujii,  Y.,  Kaneko,  T.,  Sato,  S., and  Okazaki,  S. (2015) Identification of Bradyrhizobium elkanii genes involved in incompatibility with soybean plants carrying the Rj4 Allele. Appl Environ Microbiol 81: 6710–6717.
•  Figurski,  D.H., and  Helinski,  D.R. (1979) Replication of an origin-containing derivative of plasmid RK2 dependent on a plasmid function provided in trans. Proc Natl Acad Sci U S A 76: 1648–1652.
•  Freiberg,  C.,  Fellay,  R.,  Bairoch,  A.,  Broughton,  W.J.,  Rosenthal,  A., and  Perret,  X. (1997) Molecular basis of symbiosis between Rhizobium and legumes. Nature 387: 394–401.
•  Gassmann,  W., and  Bhattacharjee,  S. (2012) Effector-triggered immunity signaling: from gene-for-gene pathways to protein-protein interaction networks. Mol Plant-Microbe Interact 25: 862–868.
•  Göttfert,  M. (1993) Regulation and function of rhizobial nodulation gene. FEMS Microbiol Rev 10: 39–63.
•  Hashiguchi,  M.,  Tanaka,  H.,  Muguerza,  M.,  Akashi,  R.,  Sandal,  N.N.,  Andersen,  S.U., and  Sato,  S. (2018) Lotus japonicus genetic, mutant, and germplasm resources. Curr Protoc Plant Biol 3: e20070.
•  Hayashi,  M.,  Saeki,  Y.,  Haga,  M.,  Harada,  K.,  Kouchi,  H., and  Umehara,  Y. (2012) Rj (rj) genes involved in nitrogen-fixing root nodule formation in soybean. Breed Sci 61: 544–553.
•  Hayashi,  M.,  Shiro,  S.,  Kanamori,  H.,  Mori-Hosokawa,  S.,  Sasaki-Yamagata,  H.,  Sayama,  T., et al. (2014) A thaumatin-like protein, Rj4, controls nodule symbiotic specificity in soybean. Plant Cell Physiol 55: 1679–1689.
•  Hempel,  J.,  Zehner,  S.,  Göttfert,  M., and  Patschkowski,  T. (2009) Analysis of the secretome of the soybean symbiont Bradyrhizobium japonicum. J Biotechnol 140: 51–58.
•  Kaneko,  T.,  Nakamura,  Y.,  Sato,  S.,  Asamizu,  E.,  Kato,  T.,  Sasamoto,  S., et al. (2000) Complete genome structure of the nitrogen-fixing symbiotic bacterium Mesorhizobium loti. DNA Res 7: 331–338.
•  Kaneko,  T.,  Nakamura,  Y.,  Sato,  S.,  Minamisawa,  K.,  Uchiumi,  T.,  Sasamoto,  S., et al. (2002) Complete genomic sequence of nitrogen-fixing symbiotic bacterium Bradyrhizobium japonicum USDA110. DNA Res 9: 189–197.
•  Kaneko,  T.,  Maita,  H.,  Hirakawa,  H.,  Uchiike,  N.,  Minamisawa,  K.,  Watanabe,  A., and  Sato,  S. (2011) Complete Genome Sequence of the Soybean Symbiont Bradyrhizobium japonicum Strain USDA6T. Genes 2: 763–787.
•  Kawaharada,  Y.,  Kelly,  S.,  Nielsen,  M.W.,  Hjuler,  C.T.,  Gysel,  K.,  Muszyński,  A., et al. (2015) Receptor-mediated exopolysaccharide perception controls bacterial infection. Nature 523: 308–312.
•  Kawaharada,  Y.,  Nielsen,  M.W.,  Kelly,  S.,  James,  E.K.,  Andersen,  K.R.,  Rasmussen,  S.R., et al. (2017) Differential regulation of the Epr3 receptor coordinates membrane-restricted rhizobial colonization of root nodule primordia. Nat Commun 8: 14534.
•  Kelly,  S.,  Mun,  T.,  Stougaard,  J.,  Ben,  C., and  Andersen,  S.U. (2018) Distinct Lotus japonicus transcriptomic responses to a spectrum of bacteria ranging from symbiotic to pathogenic. Front Plant Sci 9: 1218.
•  Kimbrel,  J.A.,  Thomas,  W.J.,  Jiang,  Y.,  Creason,  A.L.,  Thireault,  C.A.,  Sachs,  J.L., and  Chang,  J.H. (2013) Mutualistic co-evolution of type III effector genes in Sinorhizobium fredii and Bradyrhizobium japonicum. PLoS Pathog 9: e1003204.
•  Krause,  A.,  Doerfel,  A., and  Göttfert,  M. (2002) Mutational and transcriptional analysis of the typeIII secretion system of Bradyrhizobium japonicum. Mol Plant-Microbe Interact 15: 1228–1235.
•  Krishnan,  H.B.,  Lorio,  J.,  Kim,  W.S.,  Jiang,  G.,  Kim,  K.Y.,  DeBoer,  M., and  Pueppke,  S.G. (2003) Extracellular proteins involved in soybean cultivar-specific nodulation are associated with pilus-like surface appendages and exported by a type III protein secretion system in Sinorhizobium fredii USDA257. Mol Plant-Microbe Interact 16: 617–625.
•  Krusell,  L.,  Krause,  K.,  Ott,  T.,  Desbrosses,  G.,  Krämer,  U.,  Sato,  S., et al. (2005) The sulfate transporter SST1 is crucial for symbiotic nitrogen fixation in Lotus japonicus root nodules. Plant Cell 17: 1625–1636.
•  Kumagai,  H.,  Hakoyama,  T.,  Umehara,  Y.,  Sato,  S.,  Kaneko,  T.,  Tabata,  S., and  Kouchi,  H. (2007) A novel ankyrin-repeat membrane protein, IGN1, is required for persistence of nitrogen-fixing symbiosis in root nodules of Lotus japonicus. Plant Physiol 143: 1293–1305.
•  Kumar,  S.,  Stecher,  G., and  Tamura,  K. (2016) MEGA7: Molecular evolutionary genetics analysis version 7.0 for bigger datasets. Mol Biol Evol 33: 1870–1874.
•  Limpens,  E.,  Franken,  C.,  Smit,  P.,  Willemse,  J.,  Bisseling,  T., and  Geurts,  R. (2003) LysM domain receptor kinases regulating rhizobial nod factor–induced infection. Science 302: 630–633.
•  Maekawa,  T.,  Maekawa-Yoshikawa,  M.,  Takeda,  N.,  Imaizumi-Anraku,  H.,  Murooka,  Y., Hayashi, M. (2009) Gibberellin controls the nodulation signaling pathway in Lotus japonicus. Plant J 58: 183–194.
•  Marie,  C.,  Deakin,  W.J.,  Ojanen-Reuhs,  T.,  Diallo,  E.,  Reuhs,  B.,  Broughton,  W.J., and  Perret,  X. (2004) TtsI, a key regulator of Rhizobium species NGR234 is required for type III-dependent protein secretion and synthesis of rhamnose-rich polysaccharides. Mol Plant-Microbe Interact 17: 958–966.
•  Miwa,  H., and  Okazaki,  S. (2017) How effectors promote beneficial interactions. Curr Opin Plant Biol 38: 148–154.
•  Nguyen,  H.P.,  Ratu,  S.T.N.,  Yasuda,  M.,  Göttfert,  M., and  Okazaki,  S. (2018) InnB, a novel type III effector of Bradyrhizobium elkanii USDA61, controls symbiosis with Vigna species. Front Microbiol 9: 3155.
•  Niwa,  S.,  Kawaguchi,  M.,  Imazumi-Anraku,  H.,  Chechetka,  S.A.,  Ishizaka,  M.,  Ikuta,  A., and  Kouchi,  H. (2001) Responses of a model legume Lotus japonicus to lipochitin oligosaccharide nodulation factors purified from Mesorhizobium loti JRL501. Mol Plant-Microbe Interact 14: 848–856.
•  Okazaki,  S.,  Zehner,  S.,  Hempel,  J.,  Lang,  K., and  Göttfert,  M. (2009) Genetic organization and functional analysis of the type III secretion system of Bradyrhizobium elkanii. FEMS Microbiol Lett 295: 88–95.
•  Okazaki,  S.,  Okabe,  S.,  Higashi,  M.,  Shimoda,  Y.,  Sato,  S.,  Tabata,  S., et al. (2010) Identification and functional analysis of type III effector proteins in Mesorhizobium loti. Mol Plant-Microbe Interact 23: 223–234.
•  Perret,  X.,  Staehelin,  C., and  Broughton,  W.J. (2000) Molecular basis of symbiotic promiscuity. Microbiol. Microbiol Mol Biol Rev 64: 180–201.
•  Radutoiu,  S.,  Madsen,  L.H.,  Madsen,  E.B.,  Jurkiewicz,  A.,  Fukai,  E.,  Quistgaard,  E.M., et al. (2007) LysM domains mediate lipochitin–oligosaccharide recognition and Nfr genes extend the symbiotic host range. EMBO J 26: 3923–3935.
•  Rappsilber,  J.,  Mann,  M., and  Ishihama,  Y. (2007) Protocol for micro-purification, enrichment, pre-fractionation and storage of peptides for proteomics using StageTips. Nat Protoc 2: 1896–1906.
•  Ross,  P.L.,  Huang,  Y.N.,  Marchese,  J.N.,  Williamson,  B.,  Parker,  K.,  Hattan,  S., et al. (2004) Multiplexed protein quantitation in Saccharomyces cerevisiae using amine-reactive isobaric tagging reagents. Mol Cell Proteomics 3: 1154–1169.
•  Sadowsky,  M.J.,  Tully,  R.E.,  Cregan,  P.B., and  Keyser,  H.H. (1987) Genetic diversity in Bradyrhizobium japonicum serogroup 123 and its relation to genotype specific nodulation of soybean. Appl Environ Microbiol 53: 2624–2630.
•  Saeki,  K., and  Kouchi,  H. (2000) The Lotus symbiont, Mesorhizobium loti: molecular genetic techniques and application. J Plant Res 113: 457465.
•  Saeki,  K. (2011) Rhizobial measures to evade host defense strategies and endogenous threats to persistent symbiotic nitrogen fixation: a focus on two legume-rhizobium model systems. Cell Mol Life Sci 68: 1327–1339.
• Sambrook, J., and Russell, D.W. (2001) Molecular Cloning: A Laboratory Manual, 3rd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press.
•  Shah,  N.,  Wakabayashi,  T.,  Kawamura,  Y.,  Skovbjerg,  C.K.,  Wang,  H.,  Mustamin,  Y. et al. (2020) Extreme genetic signatures of local adaptation during Lotus japonicus colonization of Japan. Nat Commun 11: 253.
•  Simon,  R.,  Priefer,  U., and  Pühler,  A. (1983) A broad host range mobilization system for in vivo genetic engineering: transposon mutagenesis in gram negative bacteria. Nat Biotechnol 1: 784–791.
•  Skorpil,  P.,  Saad,  M.M.,  Boukli,  N.M.,  Kobayashi,  H.,  Ares-Orpel,  F.,  Broughton,  W.J., and  Deakin,  W.J. (2005) NopP, a phosphorylated effector of Rhizobium sp. strain NGR234, is a major determinant of nodulation of the tropical legumes Flemingia congesta and Tephrosia vogelii. Mol Microbiol 57: 1304–1317.
•  Sugawara,  M.,  Takahashi,  S.,  Umehara,  Y.,  Iwano,  H.,  Tsurumaru,  H.,  Odake,  H., et al. (2018) Variation in bradyrhizobial NopP effector determines symbiotic incompatibility with Rj2-soybeans via effector-triggered immunity. Nat Commun 9: 3139.
•  Tsukui,  T.,  Eda,  S.,  Kaneko,  T.,  Sato,  S.,  Okazaki,  S.,  Kakizaki-Chiba,  K., et al. (2013) The Type III secretion system of Bradyrhizobium japonicum USDA122 mediates symbiotic incompatibility with Rj2 soybean plants. Appl Environ Microbiol 79: 1048–1051.
•  Wassem,  R.,  Kobayashi,  H.,  Kambara,  K.,  Le Quéré,  A.,  Walker,  G.C.,  Broughton,  W.J., and  Deakin,  W.J. (2008) TtsI regulates symbiotic genes in Rhizobium species NGR234 by binding to tts boxes. Mol Microbiol 68: 736–748.
•  Wenzel,  M.,  Friedrich,  L.,  Göttfert,  M., and  Zehner,  S. (2010) The type III-secreted protein NopE1 affects symbiosis and exhibits a calcium-dependent autocleavage activity. Mol Plant-Microbe Interact 23: 124–129.
•  Xin,  D.W.,  Liao,  S.,  Xie,  Z.P.,  Hann,  D.R.,  Steinle,  L.,  Boller,  T., and  Staehelin,  C. (2012) Functional analysis of NopM, a novel E3 ubiquitin ligase (NEL) domain effector of Rhizobium sp. strain NGR234. PLoS Pathog 8: e1002707.
•  Yamaya-Ito,  H.,  Shimoda,  Y.,  Hakoyama,  T.,  Sato,  S.,  Kaneko,  T.,  Hossain,  M.S., et al. (2018) Loss-of-function of ASPARTIC PEPTIDASE NODULE-INDUCED 1 (APN1) in Lotus japonicus restricts efficient nitrogen-fixing symbiosis with specific Mesorhizobium loti strains. Plant J 93: 5–16.
•  Yang,  Y.,  Zhao,  J.,  Morgan,  R.L.,  Ma,  W., and  Jiang,  T. (2010) Computational prediction of type III secreted proteins from gram-negative bacteria. BMC Bioinformatics 11: S47.
•  Yasuda,  M.,  Miwa,  H.,  Masuda,  S.,  Takebayashi,  Y.,  Sakakibara,  H., and  Okazaki,  S. (2016) Effector-triggered immunity determines host genotype-specific incompatibility in Legume-Rhizobium symbiosis. Plant Cell Physiol 57: 1791–1800.
•  Zhang,  L.,  Chen,  X.J.,  Lu,  H.B.,  Xie,  Z.P., and  Staehelin,  C. (2011) Functional analysis of the type 3 effector nodulation outer protein L (NopL) from Rhizobium sp. NGR234: symbiotic effects, phosphorylation, and interference with mitogen-activated protein kinase signaling. J Biol Chem 286: 32178–32187.