Use of the rhizobial type III effector gene nopP to improve Agrobacterium rhizogenes-mediated transformation of Lotus japonicus

Background Protocols for Agrobacterium rhizogenes-mediated hairy root transformation of the model legume Lotus japonicus have been established previously. However, little efforts were made in the past to quantify and improve the transformation efficiency. Here, we asked whether effectors (nodulation outer proteins) of the nodule bacterium Sinorhizobium sp. NGR234 can promote hairy root transformation of L. japonicus. The co-expressed red fluorescent protein DsRed1 was used for visualization of transformed roots and for estimation of the transformation efficiency. Results Strong induction of hairy root formation was observed when A. rhizogenes strain LBA9402 was used for L. japonicus transformation. Expression of the effector gene nopP in L. japonicus roots resulted in a significantly increased transformation efficiency while nopL, nopM, and nopT did not show such an effect. In nopP expressing plants, more than 65% of the formed hairy roots were transgenic as analyzed by red fluorescence emitted by co-transformed DsRed1. A nodulation experiment indicated that nopP expression did not obviously affect the symbiosis between L. japonicus and Mesorhizobium loti. Conclusion We have established a novel protocol for hairy root transformation of L. japonicus. The use of A. rhizogenes LBA9402 carrying a binary vector containing DsRed1 and nopP allowed efficient formation and identification of transgenic roots. Supplementary Information The online version contains supplementary material available at 10.1186/s13007-021-00764-z.


Background
Agrobacterium tumefaciens (Rhizobium radiobacter) causing formation of crown galls and Agrobacterium rhizogenes (Rhizobium rhizogenes) inducing hairy roots are powerful tools to express genes in plants. In these transformation systems, Agrobacterium harbors a plant expression vector (binary vector) with a given gene construct [1]. Like with A. tumefaciens, transfer DNA (T-DNA) flanked by defined border sequences is translocated by A. rhizogenes into host cells and then integrated into the plant genome [2]. A. rhizogenes has been widely used to obtain composite plants that express genes in formed hairy roots. These roots can either be transgenic (i.e., express genes derived from the binary vector) or non-transgenic. Thus, detection of transgenic hairy roots is facilitated by co-transformation of a marker gene that allows visualization of transgenic roots.
Various protocols for A. rhizogenes-mediated hairy root transformation of Lotus japonicus have been published previously [3][4][5][6][7][8][9][10][11][12]. However, relatively little efforts were made in the past to quantify and optimize hairy root transformation of this plant. L. japonicus is a model legume widely used to investigate plant development and interactions with nitrogen-fixing bacteria, mycorrhizal fungi and nematodes. The early-flowering ecotype MG20 is particularly suitable for indoor handling [13]. The genome of L. japonicus has been completely sequenced [14,15]. L. japonicus plants have been successfully used in various gene mapping studies. Most notably, genes of L. japonicus mutants were identified that play a crucial role in the nodule symbiosis between legumes and nitrogen-fixing rhizobia. Nod factor receptor genes, for example, were firstly identified and characterized in L. japonicus [16,17]. Rhizobial Nod factors are lipo-chitooligosaccharidic signals to trigger nodulation signaling in host legumes [18,19].
Pathogenic bacteria also often possess a T3SS to translocate T3 effectors into plant cells. AvrPto of Pseudomonas syringae, for example, is an effector that targets plant immune receptors, thereby suppressing plant defense responses [38,39]. A. tumefaciens-mediated transient transformation was increased in AvrPto expressing Arabidopsis thaliana plants, suggesting that suppression of plant defense responses stimulated the susceptibility to Agrobacterium [40]. We therefore wondered whether rhizobial T3 effectors can be used to promote hairy root transformation of L. japonicus. We examined four effectors of Sinorhizobium NGR234 and found that nopP expression stimulated formation of transgenic hairy roots. Finally, we show that nopP expressing roots can be well nodulated by Mesorhizobium loti.
For mobilization of binary vectors into A. rhizogenes, 0.5 μg vector DNA was added to 100 μL of pre-cooled competent cells in a sterile 0.2-cm electroporation cuvette. The electroporator (Ding Guo, Guangzhou, China) was set to 1500 V and 10 ms. After eletroporation, 1 mL of YMB medium (kept at room temperature) was added to the cuvette. The bacteria were transferred to 1.5-mL test tubes, manually mixed and then placed on a shaker (200 rpm, 27 °C). After incubation for 3 h, 100 μL of the bacterial suspensions were spread on YMB agar plates containing 100 mg L −1 kanamycin. The plates were incubated at 27 °C for two or three days. Finally, single colonies were selected and the presence of plasmids was confirmed by PCR tests using primers shown in Additional file 1: Table S2.

Plant material and seed germination
Seeds of Lotus japonicus (Regel) Larsen ecotype MG20 (Miyakojima MG20; [13]) were soaked in concentrated sulfuric acid for 10 min and washed with water for at least five times. The seeds were then incubated in 70% ethanol for 2 min and transferred to a tenfold diluted commercial bleach solution (~ 0.35% active chlorine; Langqi, Guangzhou, China). The seeds were vigorously shaken on a vortex mixer every 2 min for a period of 10 min. After five-times washing with distilled water, the surface sterilized seeds were evenly suspended on distilled water and incubated overnight at 4 °C. The seeds were then placed on 1.0% (w/v) water agar plates (~ 50 seeds per Petri dish; 15 cm in diameter), which were incubated at an angle of ~ 80° in the dark chamber of a temperature-controlled growth room (24 ± 2 °C). After 3 days, the plates with germinated seedlings were exposed to light/dark conditions (16-h photoperiod; 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 3:1). After incubation for 4 days, seedlings with unfolded cotyledons (green in color) were used for hairy root transformation (~ 80% of germinated seedlings; germination rate ≥ 50%).

Hairy root transformation
Transformation of L. japonicus ecotype MG20 was performed with a given A. rhizogenes strain carrying pISV-DsRed1 or pISV-DsRed1 containing an effector gene. All described steps were performed under sterile conditions. The bacteria were grown on 1.5% (w/v) YMB agar plates supplemented with kanamycin (100 mg L −1 ) and rifampin (25 mg L −1 ). After incubation for 60 h at 27 °C, the bacteria were used for inoculation of germinated seedlings with unfolded cotyledons. Roots of the seedlings were diagonally cut off by a scalpel. The wounded seedlings were then completely dipped into the A. rhizogenes colonies. After incubation at room temperature for 30 min in the dark, the inoculated seedlings were placed on agar plates prepared as follows: Round Petri dishes (15 cm in diameter) were filled with 0.9% (w/v) technical agar (HKM, Guangzhou, China) containing 1/2 strength Gamborg's B5 Salts and Vitamins medium (Sigma-Aldrich) supplemented with 200 μM acetosyringone (HKM, Guangzhou, China). A half of the solid agar was removed to provide more space for developing shoots. The agar plate with the seedlings was then covered with a filter paper of the same size (Whatman, Hangzhou, China) to stabilize the seedlings and reduce formation of condensate water. Each Petri dish contained 10 seedlings placed in a row (in the middle of the plate). The plates were sealed with parafilm with several incisions (using a sterile scalpel to allow air exchange) and incubated in the dark (23 ± 2 °C) at an angle of ~ 80°. All plates were partially covered with aluminum foil to protect the roots from light.
The following day, the plates with seedlings were transferred to the growth room and incubated at light/dark conditions as mentioned above (24 ± 2 °C; angle of ~ 80°). The plates were placed into dark 30 × 12 × 5 cm plastic boxes (without lid) to protect the roots from light effects. Every 7 days, seedlings were transferred to freshly-made agar plates as described above. Plants with formed hairy roots were analyzed 28 or 45 days post inoculation (dpi) with agrobacteria.

Effector gene expression analysis
Quantitative real time PCR (qRT-PCR) was performed to analyze expressed effector genes in formed hairy roots (28 dpi). Total mRNA (in triplicates) was extracted by an RNA extraction kit following the manufacturer's instructions (Takara, Tokyo, Japan) and treated with RNase-free DNase (Takara, Tokyo, Japan). The first-strand cDNA was synthesized using the HiScript II Q RT SuperMix for qPCR Kit (Vazyme, Nanjing, China). Reactions (in triplicate) were performed using the LightCycler ® 480 SYBR Green I Master Mix in a LightCycler 480 System (Roche Diagnostics, Mannheim, Germany). Primers are listed in Additional file 1: Table S2. Primers of the constitutively expressed Ubiquitin gene of L. japonicus (Gen-Bank: DQ249171) were used as a reference to normalize the transcript abundance values. Each PCR reaction consisted of 5 μL of cDNA template (500 ng), 10 μM of each primer and 2 μL of the SYBR Green I Master Mix in a final volume of 10 μL. Following thermocycling conditions were used: (i) denaturing: 95 °C for 2 min; (ii) 30 cycles: 95 °C for 30 s, 60 °C for 20 s, 72 °C for 20 s; (iii) melting curves: 95 °C for 30 s, 60 °C for 20 min; (iv) 72 °C for 5 min. Threshold cycles (C T values) were calculated with the Roche LightCycler 480 software. According to the manufacturer's suggestion, a threshold of 0.1 was defined as the C T detection limit value. Relative gene expression levels were calculated using the 2 −ΔΔCt method [49].

Microscopic analysis
Roots transformed with pISV-DsRed1 and derivatives were microscopically analyzed for red fluorescence using a Zeiss fluorescence microscope ImagerZ1 (Carl Zeiss AG, Oberkochen, Germany) or a Lumar V12 fluorescence stereo-microscope (Carl Zeiss AG).
Red fluorescence conditions were used to distinguish between red fluorescent and non-fluorescent roots. All plant material was photographed under bright field and red fluorescence conditions as recommended by the user manuals of the microscopes. The computer software AxioVision Rel. 4.8 was used to record the pictures. Where indicated, the degree of root tissue showing red fluorescence was determined by the gridline intersection method originally established to quantify the degree of mycorrhizal root colonization [50]. Material from the whole hairy root system was randomly picked and dispersed in plates containing grid lines. Intersections between the grid lines and roots were designated as red fluorescent or non-fluorescent. Root samples from each plant were counted three times. The proportion (%) of red fluorescent tissue was calculated for each plant.

Nodulation test
Nodule formation was compared for hairy roots transformed with pISV-DsRed1 and pISV-DsRed1-

Hairy root formation induced by A. rhizogenes LBA9402 carrying pISV-DsRed1
We found in previous studies that pISV2678 is an effective binary vector for Agrobacterium transformation. The coding sequence of a given gene can be directly cloned into the multiple cloning site of this vector. To visualize transformed cells in hairy roots of L. japonicus, we cloned a DsRed1 expression cassette into the HindIII site of pISV2678. The resulting plasmid, named pISV-DsRed1, was then completely sequenced and submitted to the Genbank database (accession number MW701373). A schematic view of the T-DNA of pISV-DsRed1 is shown in Additional file 2: Figure S1. The binary vector was then mobilized into different A. rhizogenes strains (LBA1334, K599 and LBA9402) to test their capacity to transform L. japonicus roots. The protocol for hairy root transformation of L. japonicus described in this article is based on previously performed experiments [34,51,52] and L. japonicus transformation protocols from other laboratories [3][4][5][6][7][8][9][10][11][12]. Illustrative pictures of the different steps are shown in Fig. 1. Two to eighteen hairy roots reaching a length of 1-4 cm were formed on each plant at 20 dpi. The highest number of hairy roots per plant was observed when LBA9402 (carrying pISV-DsRed1) was used for inoculation (Fig. 2a).
Hairy roots expressing DsRed1 were then identified by fluorescence microscopy. Roots without DsRed1 expression showed negligible levels of background autofluorescence (Fig. 2b). Plants with at least one red fluorescent root were considered as transgenic. The percentage of obtained transgenic plants varied. When LBA9402 (carrying pISV-DsRed1) was used for transformation, 46 ± 4% of the inoculated plants were found to form transgenic hairy roots. A similar transformation frequency was obtained with K599 whereas a considerable reduced value was determined for LBA1334 (20 ± 2%). Significant differences were found for the transformation efficiency as determined by the percentage of transgenic (red fluorescent) roots per plant. Compared to the other strains, LBA9402 (carrying pISV-DsRed1) was more efficient in inducing red fluorescent roots (Fig. 2c). Based on these results, LBA9402 was further used for optimization of hairy root transformation.

Expression of nopP increases the transformation efficiency
The coding sequences of four T3 effector genes of Sinorhizobium sp. NGR234 were cloned into pISV-DsRed1 to examine effects of these genes on hairy root transformation of L. japonicus. The vectors containing nopL, nopM, nopP or nopT (pISV-DsRed1-nopL, pISV-DsRed1-nopM, pISV-DsRed1-nopP and pISV-DsRed1-nopT, respectively) were mobilized into LBA9402 and the  agrobacteria used for hairy root transformation. Microscopic analysis was performed at 28 dpi to detect red fluorescent roots. To confirm effector gene expression, total RNA was extracted from root material showing red fluorescence. Transcripts of nopL, nopM, nopP, and nopT were detected by qRT-PCR, indicating that the effector genes were co-expressed with DsRed1 (Fig. 3).
Further microscopic analyses revealed that the strength of red florescence in hairy roots varied. Some roots only showed partially red fluorescence, suggesting that not all root cells were transgenic. Such chimeric roots appeared to be less frequent when pISV-DsRed1-nopM or pISV-DsRed1-nopP were used for transformation. Furthermore, red fluorescence signals of roots transformed with pISV-DsRed1-nopM or pISV-DsRed1-nopP appeared sometimes stronger as compared to those obtained with the other vectors.
The transformation frequency of plants transformed with binary vectors containing effector genes was then compared to plants transformed with the pISV-DsRed1 control. In total, each binary vector was examined on 50 plants in 5 test runs. For pISV-DsRed1, 56 ± 4% of the inoculated plants were found to form transgenic (red fluorescent) hairy roots. Similar data were obtained for binary vectors containing nopL and nopP (50 ± 10% and 58 ± 5%, respectively) and lower values for the vectors containing nopM and nopT (38 ± 10% and 45 ± 6%, respectively). Remarkably, the use of pISV-DsRed1-nopP resulted in a significantly increased transformation efficiency as analyzed by the percentage of hairy roots showing red fluorescence. Nearly 70% of roots induced by pISV-DsRed1-nopP were fluorescent whereas values for pISV-DsRed1 and the other vectors were significantly lower (Fig. 4a). Hence, nopP expression in planta stimulated formation of transgenic hairy roots, while such an effect was not observed for nopL, nopM or nopT expression.
In an additional experiment with pISV-DsRed1 and pISV-DsRed1-nopP, the transformation protocol was slightly modified. After co-culture with A. rhizogenes, seedlings were directly placed into the agar plates to develop hairy roots (without use of a filter paper). In this experiment, plants were harvested at 45 dpi and the degree of root tissue showing red fluorescence was determined by the gridline intersection method [50]. Using this method, the ratio of red fluorescent to non-fluorescent tissue was determined for the whole hairy root system independently of the number or size of formed hairy roots. Compared to pISV-DsRed1, transformation with pISV-DsRed1-nopP resulted in a significant increase of red fluorescent tissue in this experiment (Fig. 4b).
Finally, we used the A. rhizogenes strains LBA1334 and K599 to examine the effect of nopP on hairy root transformation. Like with LBA9402, pISV-DsRed1-nopP was superior to pISV-DsRed1 when the percentage of hairy roots showing red fluorescence was determined (Additional file 2: Table S3).

Nodule formation on hairy roots is not affected by nopP expression
Expression of nopM in hairy roots of L. japonicus negatively affected nodule formation in a recent study [34]. We therefore wondered whether nopP expression also has an impact on nodule formation. LBA9402 carrying either pISV-DsRed1 or pISV-DsRed1-nopP were used for transformation and formed hairy roots were subsequently inoculated with M. loti MAFF303099. As in the previous experiments, transformation with pISV-DsRed1-nopP analyzed at 28 dpi resulted in a higher percentage of red fluorescent hairy roots as compared with pISV-DsRed1 (Fig. 5a). Selected transgenic plants (with at least one red fluorescent root) were transferred into plastic jars containing vermiculite and expanded clay. The plants were then inoculated with MAFF303099 and nodule formation was analyzed 28 days later. Most nodules showed red fluorescence (Fig. 5b). The total number of nodules per plant was similar for hairy roots induced by pISV-DsRed1 and pISV-DsRed1-nopP. Due to the higher transformation efficiency, more red fluorescent nodules were counted for plants transformed with pISV-DsRed1-nopP as compared to pISV-DsRed1. However, this difference was statistically not significant (Fig. 5c). Furthermore, the biomass of nodulated roots was similar for both binary vectors (pISV-DsRed1: 3.49 ± 0.32 mg DW per plant; pISV-DsRed1-nopP: 3.43 ± 0.23 mg DW per plant). These data indicate that nopP expression in host roots did not obviously affect the symbiosis between L. japonicus and MAFF303099 under the used experimental conditions.

Discussion
Hairy root transformation is a powerful technique to rapidly express genes in roots of legumes. In this study, we present a simple and convenient method for A. rhizogenes-mediated transformation of the model legume L. japonicus. We used DsRed1 as a marker to ascertain the transgenic nature of obtained hairy roots. Remarkably, we found that nopP effector gene expression in L. japonicus stimulates formation of transgenic root tissue. Compared to the strains LBA1334 and K599, A. rhizogenes LBA9402 was most suitable for hairy root transformation in accordance with published L. japonicus transformation protocols [3,11] and previous work of our laboratory [34,51,52]. Strain LBA1334 has been frequently used for L. japonicus transformation [5,8,10]. Under our experimental conditions, however, LBA1334 carrying pISV-DsRed1 showed a suboptimal transformation efficiency. K599 was included into our analysis because it has been used for transformation of Lotus corniculatus [53]. In our study with L. japonicus, induction of transgenic hairy roots by K599 carrying pISV-DsRed1 was not different from that of LBA1334 and considerably higher than reported previously for K599 containing another binary vector [3].
NopP of Sinorhizobium sp. NGR234, an effector originally identified by a phage display approach [54], was found to promote hairy root transformation of L. japonicus. In plants transformed with LBA9402 carrying pISV-DsRed1-nopP, significantly more red fluorescent roots were obtained as compared to pISV-DsRed1. In contrast, nopL, nopM, and nopT did not show such stimulatory effect. The molecular basis for this phenomenon remains to be explored. We suggest that expression of nopP in L. japonicus suppresses stress-related plant reactions that likely occur in response to A. rhizogenes inoculation as reported previously [8]. NopP may possess the capacity to suppress defense reactions in L. japonicus plants. In line with this hypothesis, a pathogenesis-related protein and a mitogen-activated protein kinase of soybean were identified as putative targets for NopP of Sinorhizobium fredii HH103 [55]. Furthermore, TRAPPC13 (trafficking protein particle complex subunit 13-like protein) of Robinia pseudoacacia was recently found to interact with NopP of Mesorhizobium amorphae CCNWGS0123. However, NopP in this legume rather seems to be associated with induction of plant defense responses at an early symbiotic stage [56]. Likewise, expression analysis of pathogenesis-related genes in certain soybeans indicated that strain specific NopP variants stimulate defense gene expression [57,58]. Remarkably, specific NopP protein variants trigger nodulation blockage depending on the soybean genotype [59,60]. Two cultivar-specific soybean resistance proteins, GmNNL1 and Rj2 (Rfg1), have been identified to be crucial in this process [58,61]. GmNNL1 was found to interact directly with NopP of Bradyrhizobium diazoefficiens USDA110 [58]. The NopP protein of strain NGR234 used in this study is closely related to NopP proteins produced by S. fredii strains forming an incompatible interaction with Rfg1 soybeans. Based on these findings, we expect that pISV-DsRed1-nopP will not be applicable for transformation of all soybean genotypes. Future work will be required to identify effectors that improve hairy root transformation of soybeans and other legumes such as the model legume Medicago truncatula.
Our study shows that nopP expression in L. japonicus does not obviously influence nodule formation. M. loti MAFF303099 was used for the nodulation experiment as this strain efficiently nodulates L. japonicus and lacks a nopP gene in its genome [62]. Prior rhizobial inoculation, transgenic L. japonicus seedlings (showing red fluorescence in at least one root) were selected. Considering the fact that L. japonicus roots possess strong autofluorescence under green fluorescence conditions, DsRed1 can be considered as a powerful tool to detect transformed roots. Red fluorescent proteins were also found to be good selection markers for hairy root transformation of other legumes such as soybean [63] and M. truncatula [64]. For our nodulation test, hairy roots lacking fluorescence were not removed. Nodule formation on red fluorescent and non-fluorescent roots therefore allowed a direct comparison of transgenic and non-transformed root tissue of a single plant. Using this non-destructive approach, we aimed to keep plant stress, known to negatively affect nodulation, on a minimal level. We also recommend transferring the seedlings to test jars as nodulation on agar plates, although possible, is often suboptimal, perhaps due to ethylene production and other stress factors [8]. Furthermore, to avoid any plant stress, we did not apply antibiotics to eliminate A. rhizogenes bacteria.
The genome editing technique CRISPR/Cas9 has been successfully used to knockout specific genes in the L. japonicus genome [65]. However, the A. tumefaciens-mediated transformation procedure to obtain whole transgenic plants is relatively time-consuming and could be accelerated by the use of A. rhizogenes in future. In fact, hairy roots of L. japonicus can be regenerated to whole transgenic plants [3]. This opens the possibility to generate mutations in hairy roots using a CRISPR/Cas9 construct. Single red fluorescent hairy roots could be further analyzed for mutations and the bar gene expression cassette in the T-DNA region of pISV-DsRed1 opens the possibility to select for transgenic plants resistant to the herbicide Basta during the shoot induction procedure.

Conclusion
This article shows that effector genes can be screened for their capacity to improve A. rhizogenes-mediated transformation of a given plant species. Expression of the rhizobial effector gene nopP in L. japonicus roots resulted in a significantly increased transformation efficiency while nopL, nopM, and nopT did not show such an effect. The hairy root transformation protocol for L. japonicus described in this article recommends the use of A. rhizogenes LBA9402 carrying pISV-DsRed1-nopP. The set of constructed pISV-DsRed1 derivatives containing different effector genes opens the possibility to test effector activities on plant species recalcitrant to A. rhizogenes transformation.