Viola odorata, L., L.

Narayani, M., Varsha, M. K. N. Sai, Potunuru, Uma Rani, Beaula, W. Sofi, Rayala, Suresh Kumar, Dixit, Madhulika, Chadha, Anju & Srivastava, Smita, 2018, Production of bioactive cyclotides in somatic embryos of Viola odorata, Phytochemistry 156, pp. 135-141 : 136-140

2.1. Development of somatic embryos of V. odorata

Somatic embryos can differentiate either directly from the explant or indirectly from callus ( Williams and Maheswaran, 1986). Advantages of somatic embryos include the feasibility of cryopreservation ( Park et al., 1998), usage as synthetic seeds to help conserve rare and endangered species ( Kang et al., 2014; Krishna Kumar and Thomas, 2012) and easily amenable to genetic modification and bioprocess optimization for productivity enhancement of specialized products ( Mallón et al., 2014). Furthermore, large-scale production of somatic embryos is possible in bioreactors with controllable production rates ( Shohael et al., 2014). In this study, protocols for the induction of somatic embryos in Viola species via indirect somatic embryogenesis have been described for the first time.

For the induction of somatic embryos via indirect somatic embryogenesis, callus lines of V. odorata , VOL-2, VOP-4, VOL-44, VOP-11 and VOP-17 ( Narayani et al., 2017a), were cultivated on a solidified medium supplemented with different hormonal combinations ( Table 1 ). It was observed that VOP-4 ( Fig. 1a ), a petiole-derived callus, responded to somatic embryogenesis by revealing its embryogenic competence whereas no signs of somatic embryogenesis could be observed in other callus lines that were used in this study. Although in many studies, loss of embryogenic potential of callus is noted within a year of induction ( Cheema, 1989; Gaj et al., 2005), the callus line, VOP-4, continued to proliferate with repeated subcultures even after 2 years without any apparent loss in embryogenic competence.

The calli developed small green globular embryos as a lump on the surfaces of the callus after 45 days of the cultivation period ( Fig. 1b ). The medium, supplemented with 1 mg /l thidiazuron (TDZ), gave rise to the maximum frequency of induction and a maximum number of somatic embryos ( Table 1 ). Increasing the TDZ concentration from 0.5 to 1 mg /l in the growth medium led to an increase in the frequency of somatic embryogenesis from 46.7% to 86.7% as well as an increase in the number of embryos obtained (indicated as a mean number) from ∼1 to 3. This can be attributed to the fact that TDZ can modulate endogenous auxin metabolism, which is associated with the induction of somatic embryogenesis ( Guo et al., 2011), and therefore can substitute the auxin and cytokinin requirement of somatic embryogenesis ( Mithila et al., 2003). However, an increase in the TDZ concentration beyond 1 mg /l resulted in a decrease in the induction frequency of somatic embryos. This decrease at a higher concentration could be attributed to the inhibitory role of a combination of several endogenous factors ( Dewir et al., 2018; Umehara et al., 2007).

Induction of somatic embryogenesis is a response generally associated with auxins, notably 2,4-D, which is reported to play a major role in induction in many plant species ( Fehér et al., 2003; Karami et al., 2009). However, in this study, it was observed that there was no embryo formation when the callus lines were cultured in the medium supplemented with or without 2,4-D as mentioned in Table 1 . Similarly, the ratio of auxins to cytokinins is also known to induce somatic embryogenesis in some species ( Rocha et al., 2015; Gaj, 2004). However, in this study, it was observed that a combination of auxins and cytokinins, NAA + TDZ, resulted in a lower frequency of induction and lower mean number of embryos ( Table 1 ). Furthermore, they failed to regenerate into plants. Similarly, BAP + NAA developed a protocormlike body (PLB), which eventually regenerated into poorly developed plantlets (with chlorotic and distorted leaves).

After 80 days of incubation and growth on the induction medium, the somatic embryos were separated from the callus mass and allowed to mature independently on the same medium. The globular embryos matured into heart-shaped, torpedo and pre-cotyledonary and cotyledonary stages of development as shown in Fig. 1c–g . The cotyledonary stage embryo gave rise to a complete plantlet in the same medium after 10 days of cultivation period ( Fig. 1h ).

2.2. Qualitative analysis of cyclotides present in somatic embryos of V. odorata

One of the major limitations of cyclotide production via natural plants is that the cyclotide expression varies with geographical location, season and the type of plant tissue ( Narayani et al., 2017b; Trabi et al., 2004). Hence, to establish in vitro cultures of somatic embryos as an alternative to natural plant biomass for the production of bioactive cyclotides, identification and characterization of cyclotides were done using a Fourier transform mass spectrometer coupled with liquid chromatography (LC-FTMS).

Cyclotides range from 28 to 37 amino acids in size, which corresponds to approximately 2500–4500 Da ( Ireland et al., 2006). The crude extracts of somatic embryos had 298 cyclotide-like peptides in the expected mass range with long retention times, which is a characteristic feature of cyclotides due to the exposure of hydrophobic residues to the exterior ( Ireland et al., 2006; Craik et al., 1999). Among these, 41 masses were consistent with the known cyclotides available in the CyBase database ( Wang et al., 2008) and 26 masses were consistent with the cyclotides identified in the natural plant and in vitro cultures of V. odorata reported in our earlier work ( Narayani et al., 2017a, 2017b). Further, the presence of three disulfide bonds (yet another characteristic feature of cyclotides) in cyclotide-like peptides was determined by a stepwise treatment comprising of dithiothreitol (DTT) and iodoacetamide (IAA) and then observing the incurred mass shift. For each disulfide bond, carbamidomethylation (as a result of IAA treatment) yields an increase in the peptide mass by 116 Da. LC-FTMS analysis revealed a mass shift of +348 Da (116 × 3) in the putative cyclotides, indicating the presence of three disulfide bonds. Apart from the disulfide bonds, the cyclotides also possess a head-to-tail cyclized backbone and therefore the cyclic nature of these putative cyclotides was confirmed by linearizing them using endoproteinase Glu-C. This treatment resulted in a mass shift of +18 Da, indicating the cyclic structure of the putative cyclotides due to the hydrolysis of the peptide bond. As a result of cyclotide diagnostic mass shift analyses (reduction, alkylation and linearization of the cyclotide-like peptides), 55 putative cyclotides were identified, among which three were undescribed. As two or more cyclotides could have the same mass yet different amino acid sequences, identification of possible cyclotides was done by MS-based peptide sequencing. Thus, cyclotides were reduced, alkylated and digested (using trypsin/endo Glu-C/chymotrypsin) to make them susceptible to fragmentation in tandem FTMS. The sequences of the cyclotides were manually deciphered using the de novo sequencing approach. Further, the deciphered sequences were also validated by subjecting the tandem FTMS data to two different search engines, such as SEQUEST ( Table S1 ) and PEAKS DB ( Table S2 ). The combination of data from two search engines resulted in the identification of 31 cyclotides. Interestingly, vodo I98 reported in a callus line (VOP-4) but not detected in V. odorata natural plant extract ( Narayani et al., 2017a), was also found to be present in the somatic embryos of V. odorata . This further substantiates the importance of in vitro cultures for the production of unnatural cyclotides. It is evident (from Table S3) that the number of cyclotides identified in the somatic embryos (55) is more than that reported in the parent callus line (VOP-4) (25) from which they were induced ( Narayani et al., 2017a). The expression of a greater number of cyclotides seems to be associated with the physiological state of the cultures. This indicated that morphological differentiation and resulting biochemical maturation in an organized form of cultures could be linked with the production of cyclotides in plant cells. A close correlation between morphogenesis/organization of plant cells/tissues and cyclotide production has been reported in Oldenlandia affinis in vitro cultures ( Dörnenburg, 2008).

The relative abundance of individual cyclotides was found to be varying as is evident from the LC chromatogram ( Fig. 2 ). In agreement with our previous report on cyclotide production in callus and cell suspension culture, the relative abundance of cyclotides in the in vitro culture of somatic embryos was also found to be higher than in the natural plants due to the presence of a lesser number of impurities (other compounds) ( Narayani et al., 2017a). The somatic embryos could produce cyclotides specific to both aerial and underground parts of the parent plant ( Narayani et al., 2017b). Therefore, there is no need to uproot the whole plant for the extraction of root and runner-specific cyclotides such as cycloviolacin O22, vodo I42 and hyfl I because the somatic embryos can serve as an alternative source for sustainable production of such cyclotides. Some of the cyclotides, such as circulin C, viba 2, and mram 5, that were not detected in the callus cultures or even in the parent plant were also found to be present in the somatic embryos. Furthermore, the somatic embryo exclusively produced undescribed cyclotides, vodos I102–104 ( Fig. S1 ) (which were detected based on the cyclotide diagnostic mass shifts alone and the sequences are not reported in this study). This indicates that in vitro cultures such as somatic embryos can also be used as model systems to study the cyclotide expression patterns and physiological effects in plants ( Slazak et al., 2015).

2.3. Bioactivity analysis of crude cyclotide extracts from natural plants of V. odorata and its somatic embryos

2.3.1. Cytotoxicity analysis

The cytotoxic activities in the crude cyclotides' extract of somatic embryos and the natural plant of V. odorata were tested against three different cancerous (HeLa, Caco-2 and UPCI: SCC131) and one non-cancerous (HEK293) human cell lines. It was found that the natural plant extract demonstrated cell proliferative effect at lower concentrations while higher concentrations exhibited a cytotoxic effect ( Fig. S2 ). However, the somatic embryo extract was found to be cytotoxic to all the cancerous cell lines tested except the Caco-2 ones, which demonstrated a proliferative effect at concentrations less than ∼400 μg/ml ( Fig. S2f ). Thus, half-maximal inhibitory concentration (IC 50) of the extracts against a cell line was calculated and the plots ( Fig. 3 ) obtained indicate that the cell viability (normalised) decreased in a concentration-dependent manner. The IC 50 values of the natural plant and somatic embryo extracts were comparable when they were tested against all cancer cell lines (nearly the same order of potent concentrations). The potencies of the somatic embryo extract against HeLa (IC 50 = 93.99 μg/ml) and UPCI: SCC131 (IC 50 = 93.60 μg/ml) cell lines were nearly the same. Interestingly, the somatic embryo extract was found to be less toxic (by ∼ fivefold) than the natural plant extract towards HEK293, indicating its specificity in targeting cancer cells vis-à-vis non-cancerous cells. This could be due to a more specific and targeted mode of action of cyclotides against cancer cells. The most abundant cyclotide in the extracts of somatic embryos was found to be cycloviolacin O2 ( Fig. 2 ). It is known to be a potent cytotoxic cyclotide whose activity is governed by the conserved Glu residue in its amino acid sequence ( Göransson et al., 2009). Cyclotides have a higher binding affinity towards phosphatidylethanolamine (PE) phospholipids and thus can preferentially target membranes of cancer cells over normal cells [as they have a higher proportion of PE exposed on the outer membrane] ( Wang et al., 2012).

2.3.2. Haemolytic activity analysis

As red blood cells (RBCs) are more susceptible to oxidative damage than other body tissues, and haemoglobin in RBCs acts as a strong catalyst which may initiate lipid peroxidation, RBCs can serve as a reliable model for testing any concentration-dependent drug toxicity ( Asgary et al., 2005). Thus, to establish the possible use of somatic embryos of V. odorata as an alternative to natural plant biomass for its ethnobotanical and therapeutic applications, the haemolytic activity of the crude cyclotide extracts obtained from somatic embryos and the whole plants were investigated and compared.

The 50% haemolysis (HD 50) value was determined as 356.6 μg/ml and 756.7 μg/ml for the crude extracts of natural plant and the somatic embryos, respectively. The dose-dependent haemolytic activity is evident from the plot ( Fig. S3 ) and also from the echinocytosis (Fig. S4), which could be due to haemolytic compounds, including some of the cyclotides. However, the higher HD 50 value of somatic embryos (by ∼2.1 fold higher than the natural plant) extract demonstrated that it was less haemolytic than the natural plant extract. It should be noted that the most abundant compound in the extracts of the somatic embryo is cycloviolacin O2. Earlier studies report that cycloviolacin O2 has an HD 50 value of ∼113 μg/ml ( Ireland et al., 2006) which is ∼3.1 fold higher than the HD 50 value of the crude extract of somatic embryos. The crude extracts of the somatic embryos of V. odorata showed less than 10% haemolysis at a concentration range of 1–200 μg/ml, which is considered non-haemolytic ( Amin and Dannenfelser, 2006). Hence, this concentration range of the extract can be used favourably for in vivo testing for medicinal applications.

2.3.3. Antimicrobial activity analysis

The antimicrobial activities of the crude cyclotides' extract from somatic embryos and the natural plant of V. odorata were determined and are reported in Table 2 . The extract of somatic embryos was found to be most potent against Escherichia coli followed by Staphylococcus aureus . In contrast, the natural plant extracts were found to be better than the somatic embryo extracts against Pseudomonas aeruginosa ( Table 2 ). The most abundant cyclotide identified in the somatic embryos was cycloviolacin O2, which has been reported to be highly potent against E. coli (minimal inhibitory concentration (MIC) ∼ 0.007 mg /ml) ( Pränting et al., 2010). Moreover, a higher relative abundance of cycloviolacin O2 was observed in the somatic embryos in comparison to that in the natural plant extract, which could have led to its eightfold higher potency against E. coli ( Table 2 ). The MIC value obtained in this study for ampicillin (standard antibiotic) against E. coli was in accordance with the reported values [(MIC = 0.16 mg /ml] ( Kumar et al., 2008), supporting the reliability of the results obtained in this study.

3. Conclusion

An efficient protocol for the development of somatic embryos via indirect somatic embryogenesis was successfully established in the medicinal plant V. odorata . The methodology developed can be used for the germplasm conservation of the medicinal plant. The bioactive potential of a crude extract from the plant is generally associated with the synergy between its phytochemicals. However, in this study, despite the lack of small molecules, the somatic embryo extract demonstrated equivalent or better bioactive potential in most cases. This could be attributed to the higher relative abundance of cyclotides present in the somatic embryos thereby suggesting the significance of the bioactivity measurements of the in vitro culture extracts. Moreover, it is evident from the study that in vitro cultures of somatic embryos can be a source of undescribed cyclotides, which were not detected in the parent plant. These results indicate that somatic embryos can be used as an alternative to the natural plant biomass for uniform production of cyclotides for prospective commercial applications. However, further studies are needed to demonstrate conclusively a selective cyclotide production in the culture.

4. Experimental

4.1. Induction and maintenance of somatic embryos of Viola odorata L.

Petiole (VOP-4, VOP-11 and VOP-17) and leaf (VOL-2 and VOL-4) derived callus of Viola odorata L. ( Violaceae ) were used for the induction of somatic embryos ( Narayani et al., 2017a). Briefly, 0.5 g (FW) of callus was placed on a solidified medium of the following composition: Murashige and Skoog (MS) medium ( Murashige and Skoog, 1962) supplemented with different plant growth regulators [TDZ, NAA, 2,4-D, BAP] either alone or in combination ( Table 1 ), 3% (w/v) sucrose and 0.4% (w/v) CleriGar (HiMedia, Mumbai, India) at an initial pH 5.7. Further, the callus cultures were incubated at 23 ̊C with 70% relative humidity with a photoperiod of 16/8 h light/dark condition. The frequency of callus lumps forming somatic embryos and the number of embryos formed from each responding callus lump were counted and recorded. Each treatment with five replicates was performed thrice to check the statistical validity of the results obtained. Experimental values obtained from all treatments are presented as the mean ± SEM of three independent experiments. The data were analysed statistically by ANOVA and Duncan's multiple range test (at P <0.05 level of significance) using IBM SPSS 24.

4.2. Qualitative analysis of cyclotides present in somatic embryos of V. odorata

V. odorata plants [Voucher number: 122002, Foundation for Revitalisation of Local Health Traditions (FRLH) herbarium, Bengaluru, India] were maintained in the horticulture unit at the Indian Institute of Technology Madras (12̊ 59′ 27″ N, 80̊ 14′ 2″ E), Chennai (Tamilnadu, India) and the aerial parts (leaves and petiole) were harvested during April, 2016 (dry season) for the study. The collected plant parts were rinsed with water, blotted and lyophilized. Similarly, the somatic embryos were also harvested and lyophilized to dryness for further use. The protocols for the extraction, identification and characterization of cyclotides were adopted from Narayani et al. (2017b). Briefly, the lyophilized material was homogenized and macerated in 60% (v/v) ethanol. The extracts were centrifuged and the supernatant was lyophilized until dryness. For identification of cyclotides, the crude extracts were reduced and alkylated with 100 mM DTT and 200 mM IAA, respectively. This was followed by digestion using either endo Glu-C, TPCK-treated trypsin or chymotrypsin. The samples were analysed on the EASY-nLC™ liquid chromatography and coupled online to an Orbitrap Elite Mass Spectrometer via a Nano-Electrospray ion source. The monoisotopic masses of cyclotides were determined both manually and also using Protein Deconvolution Software (version 1.0; Thermo Fisher Scientific, San Jose, CA, USA). For cyclotide sequence match, the SEQUEST algorithm available in Proteome Discoverer software (version 1.4.0.228; Thermo Fisher Scientific, San Jose, CA, USA) and PEAKS DB in Peaks (version 7; Bioinformatics Solutions Inc., Waterloo, Canada) were used. The sequences of known cyclotides from the CyBase database ( Wang et al., 2008) (http://www.cybase.org.au) and the manually deciphered sequence of a cyclotide, vodo I98, was used for database searching. In both the search engines, precursor tolerance of 15 ppm and fragment mass tolerance of 0.5 Da were used. Cysteine carbamidomethylation and methionine oxidation were used as static and dynamic modifications, respectively. Data were filtered using 1% protein and peptide false discovery rate that requires at least one unique peptide per protein. The sequences of cyclotides were also verified manually by a de novo sequencing approach.

4.3. Bioactivity analysis of crude cyclotide extracts from the natural plant of V. odorata and its somatic embryos

4.3.1. Assay for cytotoxicity analysis

The cytotoxic effect of the crude extracts (1–100 μg/ml) was evaluated using an Alamar blue assay as per the protocol reported by Rani et al. (2013) and Venugopalan et al. (2016). Alamar blue assay is simple, fast and sensitive, and the dye used is non-toxic to the cells ( O'Brien et al., 2000), and hence this method was employed for indirect measurement of the cell viability. Cell lines used in this study were HEK293 [embryonic kidney cell line], HeLa [cervical cancer cell line], Caco-2 [colon cancer cell line] and UPCI: SCC131 ( White et al., 2007) [oral cancer cell line]. These cell lines were cultured in Dulbecco's Modified Eagle's Medium with 10% foetal bovine serum (5 ml medium for 25 cm 3 culture flask) for routine maintenance. The cells were seeded in a 48-well cell culture plate at a density of 2–2.5 × 10 4 cells per well, containing 0.2–0.25 ml of growth medium for the experiment. The normalised cell viability was determined by measuring the optical density of the test (a culture medium containing cells and crude extract), blank (culture medium) and vehicle control in a multimode microplate reader at dual wavelengths of 595 nm (oxidized state) and 570 nm (reduced state). Each experiment was repeated thrice in duplicate wells and in every repetition a fresh batch of cells was used. Statistical analysis was performed using one-way ANOVA, followed by Dunnett's multiple comparisons test. For each cell line, a non-linear regression analysis was done, using GraphPad Prism software to obtain IC 50 values.

4.3.2. Assay for haemolytic analysis

The haemolytic assay was performed as per the protocol reported by Poth et al. (2011). Human blood from healthy volunteers was collected and centrifuged at 905 g for 10 min. The suspension of RBCs [0.5% (v/ v)] was prepared in phosphate buffered saline (PBS). The stock solution (200 mg /ml) of the extracts (natural plants and somatic embryos) was serially diluted to obtain a concentration range of 2–2000 μg/ml in PBS. Further, 60 μl of the serially diluted extract sample and 60 μl of RBCs suspension was added in 200 μl microcentrifuge tube. The tube was first incubated at 37 ̊C for 1 h and then was centrifuged. The absorbance of the supernatant was determined at 415 nm (λ max). The level of haemolysis was calculated as the percentage of maximum lysis (1% (v/v) Triton X-100 control) after adjusting for minimum lysis (PBS with 0.3% (v/v) ethanol control) ( Ireland et al., 2006). PBS served as a negative control. The experiments were conducted in triplicate and the concentration of the crude extract causing 50% haemolysis (HD 50) was calculated using GraphPad Prism software.

4.3.3. Assay for antimicrobial analysis

The antimicrobial activity of the crude extracts was determined by resazurin microtiter assay ( Sarker et al., 2007). Bacterial strains ( Staphylococcus aureus MTCC 737, Pseudomonas aeruginosa MTCC 2297 and Escherichia coli MTCC 443) used in this study were procured from the Microbial Type Culture Collection (MTCC) IMTECH, Chandigarh, India. An inoculum, equal to a 0.5 McFarland turbidity standard, was prepared from each bacterial isolate in Mueller–Hinton (MH) broth. Sterile PBS (50 μl) was dispensed in each well of a sterile 96-well flat bottom plate. The test material (50 μl of 400 mg /ml crude extract) was added to the first well, and twofold serial dilutions were performed. Similarly, twofold serial dilution was performed using 50 μl of the vehicle control [40% (v/v) ethanol] and 50 μl of the positive controls [ampicillin (5 mg /ml) and cefotaxime (6.25 mg /ml)] in separate wells. Separate wells with negative control (culture with no plant extract) and sterility control were also included. To each well, 10 μl of resazurin indicator solution was added, followed by the addition of 30 μl of triple strength MH broth. Further, 10 μl of the bacterial suspension was added to all the wells except the sterility control well. The plates were incubated at 37 ̊C in a BOD incubator for 12–18 h for further growth if any was observed. Please note, change in colour of the dye, added to the wells from blue (oxidized state) to pink (reduced) indicates growth of bacteria. The MIC was defined as the lowest concentration of each test compound or drug that prevented any colour change (observed visually), i.e. which completely inhibited microbial growth.

Figures

Fig. 1. Stages of somatic embryogenesis in Viola odorata. (a) Embryogenic-competent callus (b) induction of somatic embryos via indirect somatic embryogenesis (after 45 days). Stages of somatic embryos maturation: (c) globular (after 55 days), (d) heart-shaped (after 65 days); (e) torpedo (after 70 days); (f) pre-cotyledonary (80 days) (g) cotyledonary (after 85 days); (h) V. odorata plantlet regenerated from somatic embryos (after 90 days).

Fig. 2. Total ion chromatogram depicting the relative abundances of cyclotides detected in somatic embryos of Viola odorata. VI-Vodo I, K-Kalata, Cy- Cycloviolacin.

Fig. 3. Dose-response plot generated using GraphPad Prism software for calculating IC50 value of the crude extracts of plant (––●––) and somatic embryos (––▼––) against the cell lines, a,b) HEK293; c,d) HeLa; e,f) Caco-2; g,h) UPCI: SCC131. The error bars represent the mean ± SEM for each concentration value.