2.1. Ginsenoside Content in Studied Clones of Hairy Root Cultures of P. quinquefolium
Three clones of P. quinquefolium
hairy roots (A, B and G) were examined to determine their biological properties. Transformation was confirmed by a PCR analysis [16
]. This analysis confirmed that the rol B
and rol C
genes from the Ri plasmid of A. rhizogenes
became integrated with the genome of the P. quinquefolium
hairy roots and thus indicated the presence of integrated T-DNA in the hairy root cultures.
The studied clones differed in terms of morphology (Figure 1
) and content of active compounds–ginsenosides (Table 1
Line A demonstrated the morphology typical for hairy roots, with thin roots of a light-yellow colour. The roots from line B were also thin; however, their oldest part became brown. Additionally, they achieved a lower biomass production than those of clone A. The roots of clone G were partially thicker and had a callus-like appearance. The extracts in which the level of ginsenosides was examined were derived from the roots cultures that did not undergo an elicitation process (A, B and G), as well as those subjected to a MeJA elicitation (Ae
). The hairy roots that underwent elicitation contained more saponins than those untreated with MeJA (Table 1
The highest levels of total ginsenosides were determined in clone A and Ae (17.04 and 34.96 mg/g d.w., respectively). Both hairy root cultures were the richest in their Rb saponin content, expressed as the sum of Rb1, Rb2, Rb3, Rc and Rd; however, the protopanaxadiol derivatives content was 2.4-fold higher in Ae than clone A. In addition, the levels of the Rb group saponins also increased more than 4-fold in Be and 4.7-fold in Ge, i.e., after stimulation with MeJa, compared with the non-treated samples.
Slightly different findings were obtained for the Rg ginsenosides (expressed as sum of Rg1 and Re). Among the untreated cultures, clone A was found to express the greatest amount of protopanaxatriol derivatives. Among the treated cultures, clone Ge accumulated higher amounts of the Rg group saponins than Ae and Be. Additionally, clone Ae demonstrated lower Rg1 + Re than A.
An analysis of the individual saponins showed that the quantitatively dominant compounds were Rb1 and Rc (clone A), Rc and Rb1 (clone B) or Rb1 and Rg1 (clone G) in the cultures not subjected to elicitation. Further, metabolites Rb1 and Rd were found to predominate in all cultures (Ae
) after the MeJA treatment. Our results demonstrate that ginsenoside profiles varied significantly among the hairy root clones both with regard to the type of clone and elicitation status. The untreated clones demonstrated the following ginsenoside profiles: Rb1 > Rc > Re > Rd > Rb2/Rg1 > Rb3 for clone A, Rc > Rb1 > Rg1 > Re > Rd > Rb2 > Rb3 for clone B and Rb1 > Rg1 > Rc > Re > Rd > Rb2 > Rb3 for clone G. In contrast, the elicited cultures demonstrated quite different profiles: Rd > Rb1 > Rc > Rb2 > Rb3 > Re > Rg1 for Ae
, Rb1 > Rd > Rc > Rg1 > Re > Rb2 > Rb3 for Be
and Rb1 > Rd > Rg1 > Rc > Re > Rb2 > Rb3 for Ge
Elicitation, i.e., the treatment of a culture with an elicitor, is one of the most frequently applied methods used to increase the secondary metabolite production in in vitro cultures. It is based on the subjecting of the studied culture to the activity of the elicitor. An elicitor is a chemical compound that can enhance the synthesis of biologically active compounds in plants by causing defensive reactions. These compounds can be important ingredients from a commercial point of view [17
]. In this case, MeJA was used as the elicitor. Saponin production increased twofold in the Ae
line and threefold in the Be
lines of the P. quinquefolium
hairy roots compared with the non-elicited roots. This observation is not surprising considering previous studies [15
] indicating that MeJA boosted the expression of genes coding key enzymes involved in ginsenoside biosynthesis; more specifically, 250 µM MeJA was found to be the most optimal concentration for an effective ginsenoside accumulation [15
]. These observations are analogous to in vivo conditions where environmental factors very often strongly influence the production of secondary metabolites; hence, exposure to exogenous methyl jasmonate also influences the ginsenoside production.
The influence of in vitro elicitation on the content of the secondary metabolites in the hairy root clones was also examined for Gentiana cruciata
or Psammosilene tunicoides
]. In the present study, the A, B and G root lines not only demonstrated differences in the ginseng saponin production but they were characterized by different morphologies. These disparities can be connected with the random integration of T-DNA into the Ri plasmid in plant tissue. Previous research indicated that even following a successful transformation, the length and copy number of T-DNA inserted into a plant cell varies, resulting in variation in the morphology, genetics, physiology and biochemistry of the resulting clone, i.e., with a different metabolic state and the capacity for the synthesis of secondary metabolites [21
]. Additionally, some reports on plants from the Araliaceae
(for which P. quinquefolium
indicate that the rol A
, rol B
and rol C
oncogenes, included in T-DNA, are capable to modulate plant growth, cell differentiation and be potential activators of secondary metabolism in transformed cells [21
2.2. Cytotoxic Activity of P. quinquefolium Extracts
The cytotoxic activity of the P. quinquefolium
extracts towards Caco-2 cells increased together with the rising extract concentration (Figure 2
). It could be observed that the highest concentrations of the tested extracts exerted the strongest metabolic inhibitory effect, while the lowest concentrations did not affect the cells significantly.
In the MTT assay, for extract A, the three highest concentrations of MeJa were associated with the greatest decrease in the metabolic activity (up to 98.7% ± 0.2%). The lowest values of cytotoxic effects were observed for concentrations lower than 0.136 mg/mL (Figure 2
a). Extract Ae
exerted a similar biological activity towards Caco-2 cells to extract A. In the case of extract B, a significant increase in cytotoxicity was observed between concentrations 0.137 mg/mL and 0.274 mg/mL (from 17.2% ± 2.6% to 93.6% ± 4.8%) (Figure 2
b), together with a significant decrease in cell viability. Extract Be
was found to be the greatest inhibitor of Caco-2 metabolic activity, demonstrating toxic effects from a minimum concentration of 0.035 mg/mL. The four highest concentrations demonstrated the greatest inhibition of cell viability (up to 98.9% ± 0.3%). Extracts G and Ge
demonstrated similar effects (Figure 2
c); however, at 0.51 mg/mL, Ge had a stronger effect than G. Extracts G and Ge
only exerted a strong cytotoxic activity (approximately 98%) when administered at the three highest concentrations, as well as the lowest influence on the metabolic activity.
In the PB assay, for extracts A and Ae
, a relevant increase in the cytotoxic activity (from 11.2% ± 3.9% to 86.9% ± 0.3%) was observed between concentrations 0.27 and 0.532 mg/mL. The strongest increase in cytotoxicity was observed between the concentrations of 0.27 and 0.54 mg/mL (Figure 2
d). Similar tendencies were observed for extract B and extract Be
: the strongest cytotoxicity was observed for the highest concentrations (up to approximately 86%). For extracts G and Ge
f), a rapid decrease in the metabolic activity (from 10.3% ± 4.0% to 85.8% ± 0.1% for extract G) was observed, starting from the concentration of 0.51 mg/mL. In the case of Ge
, a strong increase of cytotoxicity up to 75.1% ± 9.3% was observed, starting from the concentration 0.255 mg/mL. The highest concentrations of both extracts were the strongest inhibitors of the metabolic cellular activity (up to approximately 86%). Furthermore, extract G demonstrated the lowest cytotoxic activity among all the studied extracts. Generally, the P. quinquefolium
extracts derived from the plant cultures that underwent elicitation displayed a stronger influence on cellular viability than those that did not.
2.3. Estimation of Half Maximal Inhibitory Concentration (IC50)
is defined as the concentration of a compound which is required to reduce cell survival to 50% of the control values. The IC50
values of all the P. quinquefolium
extracts were determined on the basis of MTT and PB assays (Table 2
). The IC50
value for each extract, calculated based on the results obtained by the MTT and PB assays, were similar. The highest cytotoxic effect was documented for extract Be
(0.06 in MTT and 0.21 mg/mL for PB). The least cytotoxic appeared to be extract G, reaching an IC50
of 0.64 mg/mL (in MTT) and 0.77 mg/mL (in PB). According to the IC50
values, the cytotoxicity of the Panax
extracts ranked as follows: Be
and B > Ae
and A > Ge
and G. The extracts obtained by elicitation demonstrated lower IC50
values than those that were not, indicating that the elicited P. quinquefolium
plants demonstrate a higher cytotoxic activity. The Presto Blue and MTT assay results indicate comparable patterns of cytotoxicity. The higher sensitivity indicated by the MTT assay may result from the fact that it induces mitochondrial dysfunction, thus augmenting the effect of the extract [24
In general, our findings indicate that the P. quinquefolium
hairy root extracts derived from the cultures that underwent MeJA elicitation had stronger cytotoxic properties. The analysis of the IC50
values showed that these parameters are lower for the Ae
clones than for the A, B and G clones, respectively. This would suggest that higher saponin levels are associated with a stronger cytotoxic activity against Caco-2 cells. However, in contrast, extracts B and Be
were the most cytotoxic, even though they contained the lowest level of ginsenosides. This is strong evidence that the observed biological activity relies on the chemical composition rather than the total quantity of the compounds: the two extracts were the richest sources of Rc, Rb1 and Rg1 ginsenosides. In addition, these findings might be attributed to the presence of rare ginsenosides such as Rh2, Rh3, Rg2 or Rg5, which were not studied in the present study. The literature data indicated that these metabolites demonstrate cytotoxic, anti-cancer and anti-proliferative activities; however, they also appear in greater quantities after subjecting field-cultivated roots to high temperatures [25
A previous study [30
] examined the anti-proliferative activity of P. quinquefolium
extracts towards HCT-116 colorectal cancer cells by the modified trichrome stain (MTS) method. Higher concentrations of the extracts were found to be associated with lowered cell viability. At lower concentrations (0.1–0.25 mg/mL), the anti-proliferative activity was minimal, while a significantly higher (above 90%) activity was observed for the higher concentrations (0.5 mg/mL) [30
]. In our case, the pattern of results was similar.
A previous MTT-based study of the cytotoxicity of a P. quinquefolium
extract towards hepatocellular carcinoma cells (SMMC-7721) also found that the survival rate of cells decreased along with the increase in the extract concentration. The cells were incubated with different extract concentrations (0, 20, 40, 60 and 80 mg/mL) for 12 h. [31
]. The Rg3 ginsenoside level was also found to significantly decrease 375.S2 melanoma cell viability compared with controls (IC50
20 μM) [32
]. Li et al. [33
] observed that the total ginsenoside extract of Chinese ginseng containing a mixture of Rg1, Re, Rd and Rb1 induced stronger cytotoxicity against HT-29 human colon cancer cells than its individual ginsenoside components. After a 72-h treatment, the IC50
was equal to 0.105 mg/mL.
2.4. Basal Endogenous DNA Damage Induced by P. quinquefolium Extracts
The genotoxicity of the different concentrations of the P. quinquefolium
extracts was estimated by means of a comet assay. The mean percentage of DNA in the comet tail ± S.E.M. at the different concentrations of the extract is given in Table 3
. The choice of concentrations was based on the IC50
data analysis (close or lower than IC50
Negative control cells demonstrated 4.2% ± 0.3% DNA damage, while the positive controls demonstrated 40.6% ± 3.6%. The genotoxic activity of extracts was observed to be dose-dependent. The highest concentrations of the P. quinquefolium extracts were noticed to be the most genotoxic. The lowest extract concentrations displayed a slightly higher genotoxic activity than the medium ones. Extracts A and Ae at concentrations 0.017 and 0.068 mg/mL induced comparable results in DNA damage, i.e., up to 10.2% ± 0.6%. At a concentration of 0.27 mg/mL, extract A was found to be 2.5-times more genotoxic than Ae, resulting in 63.5% ± 1.9% DNA damage compared with 25.6% ± 2.3%. Extracts B and Be demonstrated similar genotoxicity at concentrations of 0.009 and 0.035 mg/mL. Extract Be seemed to display stronger genotoxic effects at a concentration of 0.137 mg/mL (40.9% ± 2.4%) than extract B (34.0% ± 3.3%). At a concentration of 0.51 mg/mL, extract G demonstrated 66.6% ± 1.8% genotoxicity, while at 0.255 mg/mL Ge exerted 41.6% ± 2.7% genotoxicity. Extract Ge was probably more genotoxic than G, indicated by the fact that half the concentration was needed to induce similar genotoxic effects to G. Those values cannot be exactly compared due to the different tested concentrations (chosen on the basis of the IC50 values).
No correlation was found between genotoxic effects and the source plant species. There is no data on the genotoxicity of P. quinquefolium
extracts on cell lines, but Zhang et al. [34
] found Rg3 ginsenoside to significantly increase DNA damage in a concentration-dependent manner in human osteosarcoma cells. Rg3 also induced double-strand breaks, which can lead to chromosome aberrations.
2.6. The Effect of P. quinquefolium Extracts on Intracellular ATP Level, Mitochondrial Membrane Potential, Intracellular Oxidative Stress and Apoptosis Induction
Further analyses at concentrations not exceeding the IC50
values were performed to determine the molecular mechanism of the B and Be
ginsenosides’ cytotoxicity against Caco-2 cells. It was found that both extracts influenced cellular ATP production (Figure 4
A). The ATP level in Caco-2 cells decreased by 20% following the treatment with 0.137 mg/mL extract of the plant following elicitation; this decreased to 50% at the higher concentration of 0.274 mg/mL. Ginseng B preparation reduced luminescence by 10–15% at all studied dosages.
Both extracts reduced the mitochondrial membrane potential in a concentration-dependent manner (Figure 4
B). While extract B diminished the potential by up to 15%, extract Be reduced the value by 20% to 60%. This observed decrease in the mitochondrial potential was accompanied by an intracellular increase in the ROS level for both B and Be
C); however, 0.137 mg/mL extract B was a stronger inducer of oxidative stress: it elevated fluorescence by nearly 20% compared to controls, whereas 0.274 mg/mL Be
extract increased the ROS level by nearly 45%. These quantitative results were confirmed by fluorescence microscopy observations: cells treated with extracts B and Be
demonstrated higher fluorescence than untreated cells due to the higher ROS concentration (Figure 5
). Extract Be demonstrated a stronger intracellular ROS accumulation.
The observed decrease in the mitochondrial potential and ATP level, as well as the intensive elevation of ROS, indicated that cellular death may be triggered, like apoptosis or necrosis. Therefore, the next part of the study investigated the impact of the ginseng extracts on apoptosis induction by the detection of externalized phosphatidylserine (PS) in the cell membrane using annexin-V-FITC/propidium iodide staining (Figure 6
A). Annexin V binds to externalised phosphatidylserine on the outer membrane leaflet of apoptotic cells, whilst the propidium iodide stains the nuclei of cells with perforated membranes. The highest number of apoptotic cells positive for annexin V staining was observed for 0.137 mg/mL of ginseng B (about 18%), whereas high levels of cells positive for both annexin V and propidium iodide were observed for the Be
extract at 0.137 mg/mL and the B extract at 0.274 mg/mL. Ginseng Be
at a 0.274 mg/mL concentration revealed a high red fluorescence quantity of cells with stained nuclei (about 58%) specific for necrosis or secondary necrosis of apoptotic bodies not engulfed by neighbouring cells. A subsequent DNA fragmentation analysis of the cytoplasmic mono- and oligonucleosomes revealed a significant increase in apoptosis induction by the B and Be
extracts at 0.137 mg/mL (Figure 6
B). Further investigation showed that cells treated with a 0.274 mg/mL concentration of both preparations showed predominantly necrotic death-type features due to the presence of cell-released nucleosomes in the culture medium.
The current results are consistent with the microscopic observations of the cellular morphology changes occurring after cellular death induction. DAPI staining allows morphological changes in cell nuclei to be assessed. The nuclear morphology of Caco-2 cells was evaluated after 48 h of exposure to 0.137 mg/mL of the B and Be
extracts. Numerous apoptotic bodies, chromatin condensation and nuclear fragmentation could be observed (Figure 7
). AO/PI staining analyses were also conducted according to the criteria given by Baskić et al., 2006 [35
] and Salim et al., 2013 [36
]. The control cells (viable) exhibited a green fluorescence with a light-green nucleus with an intact structure of the chromatin (Figure 8
). An orange colour, chromatin fragmentation, cell shrinkage and cell membrane blebbing were symptoms of late apoptosis, while bright-red nuclei with condensed chromatin indicated direct necrosis. Kim et al., 2019 [32
] demonstrated that the Rg3 ginsenoside induced apoptosis in A375.S2 melanoma cells related to the mitogen-activated protein kinase signalling pathway. The authors also observed morphological changes in cells such as membrane blebbing. Li et al., 2018 [33
], in DAPI staining, observed nuclear changes in the colon cancer cell HT-29 typical for apoptosis, such as karyopyknosis, chromatin condensation and nuclear fragmentation. These observations were made for the total ginsenosides of Chinese ginseng containing Rg1, Re, Rd and Rb1.
There are many reports indicating that Rg3, Rh2, Rg5, Rk1 and Rh4 ginsenosides act as apoptosis inducers in different types of cell lines [37
]. It is supposed that the most potent apoptosis activators among saponins are those with less polar chemical structures [40
]. Recently, it was demonstrated that not only ginsenosides but also their metabolites secreted by intestinal bacteria, like compound K, are able to activate apoptosis via the induction of intracellular reactive oxygen species and mitochondria membrane potential loss [41
]. Remarkably, apoptosis as a programmed cell death (implicated in the removal of defective or unwanted cells without inflammation induction) is one of the tools used in cancer prevention. Studies performed on a BALB/c nude mouse model of human breast cancer demonstrated that Rg5 activates caspase-dependent apoptosis via the activation of the extrinsic death receptor and intrinsic mitochondrial signalling pathways [42
]. We examined the ginsenoside involvement in Caco-2 cell death induction via oxidative stress generation; however, a more detailed evaluation of specific markers connected with cellular death, such as caspases−3/−9 activation or the appearance of specific proteins, i.e., t-Bid, cytochrome c, Bax and Bak, is required [40
]. Such a molecular identification is very important because in human colorectal cancer HCT116 cells, there have been demonstrated studies identifying ginsenosides Rh2 and Rg3 as not only inducers of apoptotic-type cellular death but also as activators of paraptosis [43
]. That type of cell death is independent of caspase activation and is characterized by cytoplasmic vacuole formation, mitochondrial swelling and clumping.
The present study also examines the ability of the extracts to induce necrosis. However, it is important to mention that some proteins involved in the intrinsically regulated type of cell death, which shares features of apoptosis and necrosis, are responsible for the induction of the cellular death type known as necroptosis [44
]. On the other hand, the treatment of H9c2 cardiomyocytes with the deglycosylated ginsenoside compound Mc1 significantly increased the levels of catalase and superoxide dismutase and reduced the elevation of the proapoptotic Bax/Bcl2 ratio and caspase-3 activity [45
]. Due to these facts, the identification of the detailed mechanism of the biological activity of P. quinquefolium
ginsenosides requires further investigation.