Induction of Hairy Roots on Somatic Embryos of Rhizoclones from Typha domingensis Seedlings

A protocol for the induction of hairy roots on somatic embryos of rhizoclones from Typha domingensis seedlings grown in hydroponic rhizotron systems was established for the first time. Rhizogenesis was induced through the agrotransformation of somatic embryos in oblong and scutellar states of development using the K599, LBA9402, and A4 strains of Agrobacterium rhizogenes. The transfection to the embryos was performed by cocultivation of rhizoclones on a Murashige and Skoog mineral medium at 50% strength (MS0.5), in the dark, at 28 ± 2 °C for 72 h. In contrast to nontransformed embryos that did not exhibit any root tissue, transformed embryos presented hairy roots that varied in number, length, and density depending on the bacterial strain, and K599 was the most effective strain. After analysis via optical microscopy, the transformed embryos were collected and transferred to fresh culture media supplemented with 400 mg mL−1 cefotaxime and 10 mg L−1 ascorbic acid. The efficiency of transformation and survival of the oblong and scutellar embryos were similar among the three bacterial strains. The results show that agrotransformation of somatic embryos of rhizoclones from T. domingensis is a novel and viable strategy for the generation of genetic transformants of Typha that have potential applications in bioremediation technologies.


Introduction
The treatment of wastewater from water receptor ecosystems by using emergent aquatic macrophytes (EAM) plant species in phytoremediation-based processes is increasing as a result of the plasticity and physiological capacity of such plants to tolerate, accumulate, or remove both organic and inorganic environmental pollutants; moreover, this approach has led investigations of the genetic transformation of these species to improve the phytoremediation processes and the ecological recovery of the affected ecosystems [1][2][3][4][5][6] Typically, the plant species used for phytoremediation must be highly capable of adapting to the local environmental conditions of the polluted site, and as a consequence, stable genetic transformants from native species of the impacted sites are preferred [2,6].

Results
Hairy roots on the surface of SEs of T. domingensis were successfully induced via agrotransformation of in vitro rhizoclones grown in hydroponic rhizotron systems ( Figure 2). Based on the natural environmental and nutritional conditions of the sites where wild T. domingensis grows [34], three treatments for the selection of a culture medium that can to support the rhizotron system ( Figure 2) were implemented to establish adequate rhizoclone development via the in vitro germination of seeds. These treatments included growth on MS0.1 semisolid culture medium (Figure 2A), growth in liquid MS0.1 medium without growth regulators ( Figure 2B), or growth in purified sterile water alone ( Figure 2C); in addition, all the treatments included ascorbic acid. Both liquid systems produced friable tissue ( Figure 2B,C) with crumbly rhizoclones suitable for the induction of embryogenic callus with embryos in several developmental stages ( Figure 2D), as previously reported [33]. Moreover, further experiments revealed that the rhizoclones induced in rhizotron systems, incubated in the dark, yielded a large amount of unpigmented embryos in the scutellar and oblong morphological stages of development ( Figure 2D-F), which were suitable for studying the induction of hairy roots by cocultivation with A. rhizogenes ( Figure 2H).

Results
Hairy roots on the surface of SEs of T. domingensis were successfully induced via agrotransformation of in vitro rhizoclones grown in hydroponic rhizotron systems ( Figure 2). Based on the natural environmental and nutritional conditions of the sites where wild T. domingensis grows [34], three treatments for the selection of a culture medium that can to support the rhizotron system ( Figure 2) were implemented to establish adequate rhizoclone development via the in vitro germination of seeds. These treatments included growth on MS0.1 semisolid culture medium ( Figure  2A), growth in liquid MS0.1 medium without growth regulators ( Figure 2B), or growth in purified sterile water alone ( Figure 2C); in addition, all the treatments included ascorbic acid. Both liquid systems produced friable tissue ( Figure 2B,C) with crumbly rhizoclones suitable for the induction of embryogenic callus with embryos in several developmental stages ( Figure 2D), as previously reported [33]. Moreover, further experiments revealed that the rhizoclones induced in rhizotron systems, incubated in the dark, yielded a large amount of unpigmented embryos in the scutellar and oblong morphological stages of development ( Figure 2D-F), which were suitable for studying the induction of hairy roots by cocultivation with A. rhizogenes ( Figure 2H).  Three wild A. rhizogenes strains were tested for hairy roots induction on the SEs of rhizoclones from T. domingensis seedlings through genetic transformation of a SE line (YC1) in the oblong and scutellar states of development; these strains included the K599, LBA9402, and A4 strains ( Figure 3). The transformed embryos clearly showed the presence of hairy roots ( Figure 3A-E) emerging from single cells ( Figure 3B), whereas the nontransformed embryos did not show root tissue ( Figure 3F). Hence, transfection to the embryos by coculture in MS0.5 medium in the dark with any of the three bacteria was successful, although the number, density, and length of the hairy roots were significantly different between the bacterial strains ( Table 1). The transformed embryos showing the presence of hairy roots were collected and transferred to a fresh culture medium with 400 mg mL −1 cefotaxime and 10 mg L −1 ascorbic acid, a compound that is indispensable for the survival and health of embryos. Three wild A. rhizogenes strains were tested for hairy roots induction on the SEs of rhizoclones from T. domingensis seedlings through genetic transformation of a SE line (YC1) in the oblong and scutellar states of development; these strains included the K599, LBA9402, and A4 strains ( Figure 3). The transformed embryos clearly showed the presence of hairy roots ( Figure 3A-E) emerging from single cells ( Figure 3B), whereas the nontransformed embryos did not show root tissue ( Figure 3F). Hence, transfection to the embryos by coculture in MS0.5 medium in the dark with any of the three bacteria was successful, although the number, density, and length of the hairy roots were significantly different between the bacterial strains ( Table 1). The transformed embryos showing the presence of hairy roots were collected and transferred to a fresh culture medium with 400 mg mL −1 cefotaxime and 10 mg l -1 ascorbic acid, a compound that is indispensable for the survival and health of embryos. The infectivity of A. rhizogenes was evaluated by examining the typical appearance and morphology of the hairy roots that emerged on the treated embryos compared with the The infectivity of A. rhizogenes was evaluated by examining the typical appearance and morphology of the hairy roots that emerged on the treated embryos compared with the nontransformed tissue ( Figure 4) as suggested in numerous reports [2,4,5]. As observed, the formation of the hairy roots on the embryos was due to transfection by using any of the three Agrobacterium strains, but hairy roots were not visible on the noninfected embryos. Additionally, the structure and morphology of the transformed hairy roots were similar and independent of the bacterial strain used (Figures 3 and 4). In general, the hairy structures originated from single epidermal cells ( Figure 3B) that looked like tubular unicellular extensions in which the cells contained peripheral cytoplasm and a large central vacuole as reported for classical transformed hairy root tissue [5]. No similar root tissue was observed on the treatment with control embryos ( Figure 4A,B). It should be noted that greater abundance of Plants 2020, 9, 1679 5 of 12 hairy roots was observed in the apical and middle areas of the embryos ( Figure 4C,D), which are areas reported to have higher abundance of meristematic cells [4,5].
formation of the hairy roots on the embryos was due to transfection by using any of the three Agrobacterium strains, but hairy roots were not visible on the noninfected embryos. Additionally, the structure and morphology of the transformed hairy roots were similar and independent of the bacterial strain used (Figures 3 and 4). In general, the hairy structures originated from single epidermal cells ( Figure 3B) that looked like tubular unicellular extensions in which the cells contained peripheral cytoplasm and a large central vacuole as reported for classical transformed hairy root tissue [5]. No similar root tissue was observed on the treatment with control embryos ( Figure 4A,B). It should be noted that greater abundance of hairy roots was observed in the apical and middle areas of the embryos ( Figure 4C,D), which are areas reported to have higher abundance of meristematic cells [4,5].    The effects of the A. rhizogenes strain, and the morphological stage of the SE, on the agrotransformation process were evaluated and the results revealed that both factors affected the transformation frequency, survival, and hairy density (Table 1). Evidently, the efficiency of Plants 2020, 9, 1679 6 of 12 transformation for both the oblong and the scutellar embryos was similar for each of the three bacterial strains; however, survival was affected by the Agrobacterium strain but not the morphological stage. Interestingly, the SEs infected with the K599 strain produced a greater number, length, and density of hairy structures than those infected with the LBA9402 and A4 strains, and these results are consistent those discussed above for Figure 3C. However, in general, the Agrobacterium strains did not affect the effectiveness of the agrotransformation process of either the SEo or SEsc embryos (Kruskall-Wallis test; p > 0.05). It should be noted that the SEo stage showed lower frequencies of transformation than the SEsc stage, which affected the total number of SEs transformed (SEo plus SEsc). This finding is consistent with the observation that the efficiency of transformation was highly variable between the morphological stage of embryos (25-88%). Notably, frequent contamination was observed during cocultivation with the A4 and LB9402 strains ( Figure 2H, left flask), which may have resulted in the significantly different and lower level of survival of the SEo stage of each bacterial strain (92-54%) as confirmed by the Kruskall-Wallis test (p = 0.09), and the significantly different results for the SEsc stage (Kruskall-Wallis test; p = 0.003). Notably, the K599 strain stood out in all embryos regarding the frequency and survival of the transformed embryos, as well as the number of hairy roots (Table 1). Finally, the frequency of transformation in total SEs was substantially greater for the K599 strain. Moreover, the survival of the SEsc and the total SEs was significantly greater (Kruskall-Wallis test; p = 0.002) with the K599 strain.

Discussion
The SEs of T. domingensis developed typical hairy roots through the transfection of rhizoclones with the A. rhizogenes K599, LBA9402, and A4 strains. The phenotype of these hairy roots was similar to the phenotype of hairy roots reported in the transgenic embryos of Bouteloua gracilis [35]. The virulence was different for each of the three A. rhizogenes strains; however, the K599 strain stood out in terms of its hypervirulence, and the A4 strain stood out in terms of its virulence. Hypervirulent strains of Agrobacterium spp., have been proven effective in other species, such as Cyclamen persicum, in which embryogenic callus cocultivated with the A. tumefaciens EHA105 hypervirulent strain resulted in a transformation efficiency twice as high as that achieved with the A. tumefaciens LBA4404 virulent strain [36]; moreover, similar results were achieved with the A. rhizogenes K599 and LBA9402 strains in this study. The K599 strain, which is equivalent to the NCPPB2659 cucumopine biovar, harbors the pRi2659 plasmid (ADN-T) that transports the aux1 and aux2 genes for auxin autotrophic cell division and biosynthesis and the onset of hairy roots, including root loci oncogenic genes (mainly rol A, B, and C genes), and other genes with unidentified function, such as open reading frames [24,37,38]. The genes cluster aux and rol A, B, and C play a part in the infection and induction of hairy roots [39].
The root hairs in the SEs of T. domingensis in both the oblong and scutellar embryogenic states developed on the epidermis towards the apex and middle of the cotyledon area. Similarly, the root hairs on the root of A. thaliana emerge from epidermal cells and form short hairs on the basal area oriented towards the root apex [40], thus showing acropetal development of root hairs, which is characteristic of axonomorphic roots [41].
The critical requirements to confirm the genetic agrotransformation of a plant tissue include the identification of transformed cells or target tissue, evidence of the presence and/or expression of the transferred T-DNA in the nuclear genome, and the regeneration of transgenic plants [42,43]. In this work, the first requirement was confirmed by the selection of embryos showing hairy roots with the typical morphological pattern of such tissue, such as branching and growth on the medium without plant hormones. The biological effect and confirmation for the genetic transformation by the presence of the transferred DNA (T-DNA) of A. rhizogenes to the target plant tissue is still under study, as there are certain limitations in the case of recalcitrant species [39,44]. The infectivity of the strains has been confirmed via morphologic markers of the hairy phenotype in the Agrobacterium-treated SEs [35] as well as by the coculture technique performed in this study. Similar to the study that induced transgenic embryos of Bouteloua gracilis, polymerase chain reaction (PCR), Southern blot, and Northern blot analyses are commonly used to confirm the presence of the rol A genes in hairy SEs; however, these molecular techniques are not infallible for confirming genetic transformation. For example, the secondary embryos of Coffea canephora transfected with the A4 rhizogenic wild strain molecularly revealed a stable transformation, but without the hairy phenotype [45]. In contrast, the presence of the hairy root morphological phenotype in agrotransformed tissue with A. rhizogenes wild strains provides proof for the success of the genetic agrotransformation [20,24,35,37,[44][45][46].
Knowledge of the interaction between Agrobacterium and the structure and composition of the outer cell wall of the protodermic cells that cover SEs is limited [47]. The transfective susceptibility of A. tumefaciens in root hairs may be explained by the biochemical changes that take place on the cell surface [48]. The biochemical mechanism was revealed by the presence of polar binding polysaccharides and an unknown receptor molecule based on the greater number of adhered bacteria. As previously reported [48], the adherence of Agrobacterium to the surface of embryos might have been improved by the acid pH (5.5) of the culture medium used in the coculture, the absence of accessory structures such as cutin in this type of tissue [47], or the addition of ascorbic acid to the culture system [49]. In addition, Agrobacterium has been reported to show enhanced virulence in the presence of phenolic compounds [50], such as those produced by T. domingensis embryos [33]. Thus, consistent with these results, the acid pH, addition of ascorbic acid, and liquid state of the coculture medium all favored biochemical adherence of the bacteria to the embryo surface, thereby enhancing the agrotransformation process.
The high variability in the frequency of transformation and survival of the transformed tissue observed in this work agree with studies of transformations mediated by Agrobacterium of the monocotyledon species, such as Triticum aestivum (wheat), Oryza sativa (rice), and Zea mays (maize), for which high variable frequencies of transformation have been reported [51,52]. Additionally, a general variability for the efficiency of the agrotransformation of immature (14.9%) or mature (9.8%) embryos of wheat, and low efficiencies of transformation of zygotic embryos of maize (1.21-10.96%) with the A. rhizogenes K599 strain, with or without extra copies of virulence genes, have been reported [25]. Furthermore, high variability in the transformation efficiency has also been reported by using the electroporation technique for some rice species, and it produced more than 80% efficiency for some varieties but less than 2% for others [53].
Overall, since hairy roots can emerge only from single transformed cells expressing of transferred T-DNA [42,43,54], the present results suggest that the hairy root specific phenotype may become a practical morphologic feature for the early identification of an agrotransformation event in embryos because it might not require additional molecular or biochemical techniques. In addition, this approach provides a simple and rapid method for the genetic transformation of Typha species through the coculture of embryos with Agrobacterium, although molecular techniques should be used to validate the relationship between the hairy phenotype and the genetic transformation success. Such strategy can be combined with the current machine learning algorithms technology, which have emerged as promising tools for prediction of hairy root cultures, among other in vitro cultures types [55,56], along with the optimization of environmental conditions to achieve maximum productivity and efficiency of hairy root cultures of Typha domingensis as a valuable plant for biotechnological applications. On the other hand, it is still uncertain why the presence of ascorbic acid enhanced the survivance and health of embryos as previously reported [36]; however, this compound has been used in genetic transformation events of several plant species due to its antioxidant properties [57,58] Additionally, the frequency of infection with the A. rhizogenes strains used in this work should be studied for the transformation of other dicotyledonous species. Moreover, although the regeneration of transformed plants is still a pending issue in this work, whether the agrotransformation of other plant species by using the present approach results was successful, this could be a good advance in the area of plant biotechnology.
In conclusion, infection of T. domingensis somatic embryos, in the oblong and scutellar stage of development, with the three A. rhizogenes strains was an efficient strategy for inducing hairy roots, and the K599 strain was the most effective. The results provide new insights in the area of plant

Germination of Typha Seeds and Establishment of Rhizoclones on Somatic Embryos (SEs)
Seeds were isolated from the achenes of T. domingensis and germinated in rhizotron systems to establish rhizoclones. The achenes were collected from the tropical wetlands of Villahermosa, Tabasco (17 • 59 9.91" N and 17 • 57 48.98" E), Mexico, and deposited in the dry plant collection of the Herbarium in the Universidad Juárez Autónoma de Tabasco (UJAT-Herbarium 35227). The seeds used for this work were stored in a desiccator at room temperature, and 5-15 seeds were sowed in each rhizotron unit (sRU) established in test tubes with 20 mL of Murashige and Skoog basal medium [59] at one-tenth ionic strength (MS0.1), in either semisolid (agar 4% w/v) or liquid format with pH 6.5 and formulated with MS vitamins, without growth regulators in an autotrophic regime. Additionally, a sRU with deionized sterile water was prepared as a control. The rhizotron systems in the semisolid medium were covered with a water film of 1 cm. After 15 days germination, the root system of seedlings with 1 cm length and with similar development (pair of leaves, radicle with adventitious root) were separated from the caulinar base and transferred into a new sRU for the establishment of rhizoclones (Figure 2). Individual roots were positioned vertically on the surface of the semisolid culture medium, avoiding injuries, and were incubated under photoperiod (16/8 h light/darkness) by using cold white light (Philips, USA), at a photon flux density of 20 µmol photons m −2 s −1 (Quantum Light Meter; Spectrum Technologies, Aurora, IL, USA), and a temperature of 28 ± 2 • C. The rhizoclones were used for the induction of SEs by incubation in the dark in MS culture medium at half ionic strength (MS0.5), and supplemented with vitamins, sucrose (3% w/v), ascorbic acid (10 mg L −1 ), and 0.5 mg L −1 2,4-D, as described previously [33]. The SEs in the oblong (SEo) and scutellar (SEsc) stages of development were isolated from the embryogenic line YC1 of T. domingensis that was established and maintained as described above.

Agrobacterium Rhizogenes Strains
The A. rhizogenes K599 (cucumopine-type), LBA9402 (agropine-type), and A4 (agropine-type) strains were provided by the Metabolic Engineering group at CINVESTAV-IPN. The bacterial strains were kept at 4 • C in Luria-Bertoni semisolid culture medium (LB) at pH 7. The bacteria were first activated by transferal to the LB liquid culture medium using commercial yeast extract and incubation in an orbital shaker at 30 • C and 130 rpm for 48 h. The concentration of bacteria for infection of the embryogenic tissue was adjusted at DO 600 0.8-1 to proceed to inoculation as described by Jimenez-Antaño et al. [60].

Induction of Transformed Hairy Roots on SE of Typha Domingensis
The inoculation of the SE of T. domingensis with the A. rhizogenes strains was performed by coculture. A bacterial suspension with 48 h of growth was mixed at a ratio of 1:5 (final volume 30 mL) with the T. domingensis embryo MS0.5 medium culture as described for embryogenic induction, although the medium was not supplemented with phytoregulators. The mixture was incubated in the dark, at 28 ± 2 • C, for 72 h, and it was stirred constantly (130 rpm, 60 min) (n = 4). The embryogenic callus and the SEo and SEsc in the embryogenic mixture of each culture unit were filtered through a sterile gauze, transferred to 30 mL of MS0.5 medium with 400 mg/L cefotaxime (n = 4) to eliminate the residual A. rhizogenes, and kept under the previous environmental conditions for 28 days.

Experimental Model
The effect of the A. rhizogenes K599, LBA9402, and A4 strains on the efficiency of transformation of T. domingensis EL-YC1 was studied. The observational unit was the viable SEs (n = 40 per treatment), which were randomly sampled from the MS medium with cefotaxime. The transformed embryos were separated following a previously reported protocol [33]. The survival, hairy morphology, and transformation frequency of all viable SEs, SEos, and SEscs were analyzed at the end of the experiment.

Agrotransformation Quantification
Survival was estimated considering the excessive growth of bacteria [51] since the number of viable embryos per treatment decreased with the three strains. The visual indicators used to calculate the frequency of agrotransformation were the hairy SEo and SEsc phenotypes observed under a stereomicroscope (Carl Zeizz ® , Göttingen, Germany). The magnitude of hairiness in the SEo and SEsc stages was quantified through an image analysis (n = 3) of the number (No.), length (µm), and density (No. µm) of the hairy units [60]. The SE images were digitized using an Axio Scope A1 optical microscope (Carl Zeizz ® ) equipped with an AxioCam ERc5s camera (Carl Zeizz ® ). The fields of view of interest were selected using the 10×, 40×, and 100× + 1.25× objective lens in the optovar (Carl Zeizz ® ) and the SEN/2011 program (Carl Zeizz ® Microscopy GmbH, 2011).

Statistical Analyses
Kolmogorov-Smirnov and Cochran C tests were used to verify the assumptions of normality and homoscedasticity of the data. The Fisher's least significant difference test (LSD) or the Kruskal-Wallis H test were used to estimate the statistical difference between treatments by using a significance level of 0.05. Statistica V8 (Stat Soft Inc., Tulsa, OK, USA) software was used for statistical analyses.