Rorippa indica Regeneration via Somatic Embryogenesis Involving Frog Egg-like Bodies Efficiently Induced by the Synergy of Salt and Drought Stresses

Frog egg-like bodies (FELBs), novel somatic embryogenesis (SE) structures first observed in Solanum nigrum, were induced in Rorippa indica. NaCl-mediated salt and mannitol-mimicked drought stresses induced FELBs in R. indica, which is very different from the induction by plant growth regulators (PGRs) under low light condition that was used in S. nigrum FELB induction. It demonstrated that NaCl or mannitol supplements alone could induce FELBs in R. indica, but with low induction rates, while the synergy of NaCl and mannitol significantly increased the FELB induction rates. For the combination of 5.0 g/L mannitol and 10.0 g/L NaCl the highest FELB induction rate (100%) was achieved. It suggests that the synergy of drought and salt stresses can replace PGRs to induce FELBs in R. indica. On medium supplemented with 1.0 mg/L gibberellic acid all the inoculated in vitro FELBs developed into multiple plantlets. Morphological and histological analyses confirmed the identity of FELBs induced in R. indica and revealed that FELBs originate from root cortex cells.

Rorippa indica is an annual cruciferous herb distributed worldwide. In China it is used as a kind of wild vegetable and traditional medicine in China 1 . R. indica is a rich resource of valuable medicinal components, among others of glucosinolates, flavonoids, roripamine and isothiocyanates [1][2][3] . Mustard aphid, Lipaphis erysimi, threatens the cultivation of the rapeseed-mustard (Brassica spp.) crop, and till present in cultivated Brassica germplasm no resistance was found. As a result, no resources are available for mustard aphid resistance breeding of Brassica crops. R. indica was found to be resistant to the aphid 4 , and therefore it could serve as a resistance resource.
R. indica, used as medicinal and vegetable in China, can be used as a model for research in cruciferous plants next to Arabidopsis thaliana, for its small adult stature, short life cycle, small genome size, prolific seed production and self-pollination 5,6 . In addition, the growth conditions for R. indica are less demanding than those for A. thaliana.
However, as R. indica has not much studied as a plant species, high-frequency regeneration and transformation systems in R. indica have not been established. Frog egg-like bodies (FELBs), which were first observed in Solanum nigrum 7 , are novel structures of somatic embryogenesis (SE). Regeneration via SE involving FELBs is highly efficient and labor saving, which makes it suitable for germplasm preservation and establishment of transformation systems 7 . Generally plant growth regulators (PGRs), like 2,4-dichlorophenoxyacetic acid (2,4-D) or 1-naphthaleneacetic acid (NAA), are necessary for the induction of SE. In the presence of PGRs, desiccation, drought, and salt stresses could promote regeneration efficiency [8][9][10][11] . However, the effects of these stresses on plant regeneration in the absence of PGRs have not been reported. In this study, an R. indica regeneration system via SE involving FELBs was first established via a new method. In this system the synergy of salt and drought stresses in the absence of PGRs resulted in a high FELB induction rate (100%) in R. indica, and further led to high-efficiency SE-mediated plant regeneration. To our knowledge it is the first report that high-efficiency plant regeneration achieved by the synergistic effects of salt and drought stresses in the absence of PGRs.

Results
Under low light condition PGRs (NAA and 2,4-D) induce FELBs in R. indica with low FELB frequency. In the dark, none of the different combinations of concentrations of NAA and 2,4-D induces FELBs in R. indica root explants (Table 1). Therefore, a low light condition was employed instead. Under low light condition less than 1% of the R. indica root explants on media with the tested concentrations of NAA formed FELBs, while 2,4-D concentrations of 0.5, 1.0, and 1.5 mg/L resulted in FELB induction rates less than 19% (Table 1). This indicated that in low light condition FELBs can be induced at low rates in R. indica root explants.

R. indica regeneration via SE involving FELBs on medium supplemented with mannitol and
NaCl. Two weeks after the induction of root explants (Fig. 1A), FELBs started to form under the epidermis of root explants, finally the epidermis burst by the development of FELBs (Fig. 1B-D). The induced FELBs in R.  The mean and standard error per treatment were calculated based on 300 explants from 30 petri dishes as 30 replicates (10 explants inoculated in each petri dish). Lowercase letters indicate highly significant differences at the 1% probability level. Highly significant differences were analyzed with one-way ANOVA by Tukey test using SPSS 16.0. indica root explants were surrounded with a kind of special translucent sticky callus, which resembles the typical FELBs in S. nigrum 7 . At late developmental stages, clusters of mature FELBs containing multiple individual somatic embryoids surrounded with translucent sticky callus were formed (Fig. 1E,F). In most cases, one FELB gave rise to 5-10 individual somatic embryoids. It showed that 100.00% of the in vitro FELBs induced into multiple plantlets on MS medium supplemented with 1.0 mg/L GA 3 within two weeks (Fig. 1G). The development of into multiple plantlets from R. indica root explants is similar in S. nigrum, but different from the regular patterns of traditional SE types, which one individual SE structure often develops into one plantlet 12,13 .

Identification and morphological analysis of FELBs induced in R. indica root explants. Microscopic
squash slides double stained with acetocarmine and Evans blue 7,14 were employed to distinguish embryogenic from non-embryogenic callus part of FELBs, as embryogenic cells will be stained in red, and callus cells in blue. The images showed that the induced FELBs in R. indica root explants were composed of ball-shaped embryoids stained in red surrounded with translucent sticky callus stained in dark blue ( Fig. 2A,B), similar to the FELBs induced in S. nigrum 7 . By using microscopic squash specimens together with borax-toluidine blue staining ( The stages, ranging from proembryo, globular, heart/torpedo-shaped embryos to cotyledon-shaped embryos were observed in FELBs at different developmental stages ( Fig. 3F-M). Vascular tissues developed in FELB embryoids at late developmental stage (generally from cotyledon-shaped embryos on), while no vascular bundles connecting FELBs with parental tissues were formed (Fig. 3N,O). This observation suggests that the FELB vascular tissues were separated from the parental vascular tissues, further confirming the nature of R. indica FELB embryoids.

Effects of mannitol-mimicked drought and NaCl-mediated salt stress on the induction of FELBs in R. indica.
In the absence of PGRs, different combinations of a concentration series of 0, 2.5, 5.0, and 7.5 g/L of mannitol, and a concentration series of 0, 5.0, 10.0, and 15.0 g/L of NaCl, were tested to evaluate their effects on the induction of FELBs. FELB induction rates were very low (< 1%) when only NaCl was supplemented to the media. FELB induction rates were lower than 19% if only mannitol was supplemented to the media (Table 2). However, the combination of mannitol and NaCl significantly increased the FELB induction rates ( Table 2). The combination of 5.0 g/L mannitol and 10.0 g/L NaCl resulted in FELB induction rates till 100% of the inoculated root explants, highly significantly higher than all other tested concentration combinations of mannitol and NaCl ( Table 2). The analysis of interaction effect between NaCl and mannitol indicated that there is interaction between NaCl and mannitol, and the combination of 5.0 g/L mannitol and 10.0 g/L NaCl also demonstrate highly significantly higher FELB induction rates than all other test concentration combinations (Table 3). These results suggest that mannitol and NaCl have a synergistic effect on FELB induction in R. indica root explants.

Discussion
Supplementation of the optimal combination of PGRs in culture media is crucial to successful SE induction and plant regeneration 15,16 . In our previous study, FELBs, which are novel SE structures, were first induced in S. nigrum. 2,4-D and culturing of the explants in the dark were important factors for FELB induction in S. nigrum, while NAA was not suitable for FELB induction in S. nigrum 7 . In this study FELBs were also successfully induced by 2,4-D in R. indica root explants, similar to S. nigrum, while NAA was also found not suitable for FELB induction. Therefore, it seems that the effect of 2,4-D on FELB induction in contrast to that of NAA is more general. However, with same 2,4-D concentration gave rise to a much lower FELB induction rate in R. indica root explants compared to that of S. nigrum, suggesting that optimal concentrations of 2,4-D for FELB induction differ between plant species. S. nigrum and R. indica belong to Solanaceae and Cruciferae, respectively, illustrating that FELBs can be induced in plants from different families, highlighting the potential of FELB induction in other plant species. In this study, we also found that a low light condition other than the dark condition used in S. nigrum FELB Since the tested concentrations of 2,4-D did not result in high FELB induction rates, a new method was employed. For the new method, in the absence of PGRs, mannitol-mimicked drought stress and NaCl-mediated salt stress were adopted and successfully induced FELBs in R. indica root explants, moreover the synergy of NaCl and mannitol resulted in highly significant higher FELB induction rates than either of them alone (P < 0.01). The synergistic effect of mannitol and NaCl on somatic embryogenesis induction in R. indica was also much better than that of mannitol alone in A. thaliana 17 , and that of NaCl alone in Vigna sinensis 18 . The FELB induction rate induced by the synergistic optimal concentration combination of mannitol and NaCl reached 100.00%. These results show that drought and salt stresses can induce FELBs in R. indica root explants instead of PGRs and have synergistic effect on R. indica FELB induction. The mechanism of this synergistic induction is not clear.
In addition, in media supplemented with mannitol and NaCl, less contamination by microbes was observed during FELB induction (not published). A possible explanation is that supplements of NaCl and mannitol decreased the growth of potential contaminating microbes. This advantage will benefit the large-scale induction of FELB using this method, especially for the large-scale suspension cultures of FELB cells as bioreactor 7 .
The unavailability of a suitable transformation system for R. indica hampers its application as a model plant. The establishment of FELB mediated regeneration system will benefit the establishment of a transformation system for R. indica, which will facilitate model-plant applications and molecular breeding of R. indica.  Table 2. Effects of mannitol and NaCl on FELB induction from root explants of R. indica. Note: FELB induction rate refers to the ratio of explants with induced FELBs to the total number of inoculated explants. The mean and standard error per treatment were calculated based on 300 explants from 30 petri dishes as 30 replicates (10 root explants inoculated in each petri dish). Lowercase letters indicate highly significant differences at the 1% probability level. Highly significant differences were analyzed with one-way ANOVA by Tukey test using SPSS 16.0.  Table 3. Interaction effects between NaCl and mannitol on FELB induction from root explants of R. indica. Note: FELB induction rate refers to the ratio of explants with induced FELBs to the total number of inoculated explants. The mean and standard error per concentration of NaCl or mannitol were calculated based on 4× 300 explants from 4× 30 petri dishes as 4× 30 replicates (10 root explants inoculated in each petri dish), which is the total number of explants treated by the referred concentration of one of the stress elements combined with four different concentrations of the other stress element (There are 300 explants per combination from 30 petri dishes). Lowercase letters indicate highly significant differences at the 1% probability level. Highly significant differences were analyzed with two-way ANOVA by Tukey test using SPSS 16.0.

Materials and Methods
Plant materials and explant preparation. R. indica seeds were treated with 75% (v/v) ethanol and 2.5% (v/v) sodium hypochlorite for sterilization, according to our published protocol 7 . For germination, the sterilized seeds were sown on 1/2 Murashige and Skoog (MS) medium 19 supplemented with 1.0 mg/L gibberellic acid (GA 3 ), 30 g/L sucrose, and 7.8 g/L agar (pH 5.8) 7 , then incubated in a germination chamber (25 °C in the dark) until the seeds were fully germinated. Seedlings were transplanted onto MS medium for the preparation of explants and cultured at 25 °C with a photoperiod of 16 h light (180 μ mol·m −2 s −1 ) and 8 h dark.

Induction of FELBs.
Following the protocol of FELB induction in S. nigrum 7 NAA and 2,4-D with the concentration series 0, 0.5, 1.0, and 1.5 mg/L were employed to induce FELB in R. indica. The effect of mannitol-mimicked drought and NaCl mediated salt stress on R. indica root explant regeneration via FELBs mediated SE was tested. Root explants were placed on MS media with 30 g/L sucrose and 3.6 g/L gellan gum, pH 5.8, supplemented with different concentration combinations of mannitol (0, 2.5, 5.0, and 7.5 g/L) and NaCl (0, 5.0, 10.0, and 15.0 g/L). The explants were incubated at 25 °C and under a low light condition (36 μ mol·m −2 s −1 ) to induce FELBs. Different developmental stages of SE were recorded by using a digital camera (EOS 600D, Canon Inc., Japan) and a stereomicroscope (SMZ800, Nikon Corporation, Japan). To evaluate FELB induction rates, three hundred of root explants inoculated in thirty petri dishes (10 root explants inoculated in each petri dish) were calculated for each treatment.
Histological and histochemical analyses of FELBs. Double staining with acetocarmine and Evans blue 20 was employed to distinguish embryogenic cells from non-embryogenic ones. Embryogenic cells of FELBs were stained bright red, and non-embryogenic cells of callus and explants were stained dark blue 7 . Different SE stages were recorded by using a digital camera (EOS 600D, Canon Inc., Japan), a digital fluorescence microscope (BX 61, Olympus Corporation, Japan), and a digital optical microscope (BX 41, Olympus Corporation, Japan), respectively. Staining with 4' ,6-diamidino-2-phenylindole (DAPI) was used to detect the nuclei of embryonic and callus cells, following a previously published method 21 . Cell outlines were detected by using borax-toluidine blue staining 21 . Microscopic images were recorded by using a digital optical microscope (BX 41, Olympus Corporation, Japan). The frozen sections (thickness 8 μ m) of FELBs at different developmental stages were made with a cryostat microtome (CM1850, Leica Microsystems, Germany) following a previously published method 5 . The sections were observed and recorded by using a digital optical microscope (BX 41, Olympus Corporation, Japan).