Background
Asiasari radix (AR) has long been used in traditional Korean and Chinese medicine to treat cough, toothache, headache, neuralgia, gingivitis, asthma, and allergies due to its anti-bacterial and analgesic effects [
1]. Recently, the pharmacological roles of AR extracts (AREs) have been reported. The anti-allergic [
2,
3] and anti-inflammatory [
4,
5] activities of AREs have been confirmed in vitro and in vivo. AREs also exert an anti-carries activity by not only reducing acid production, but also by inhibiting the growth and adhesion of
Streptococcus mutans [
6]. In addition, AREs also possess skin renewal and hair follicle-generating activities [
7]. Furthermore, Jang et al. reported the possible skin-whitening role of ARE because it attenuates melanogenesis in rats [
8]. Due to its potent skin regeneration and hair loss-preventing activities, AREs have been widely used in many cosmetics. Nevertheless, the effects of ARE on various types of skin cancers were studied poorly.
Melanoma is a type of skin cancer that accounts for about 4% of all cancers; however, it is the most dangerous since it accounts for about 80% of skin cancer-related deaths [
9]. Although genetic risk factors contribute maximally to the development of melanoma, exposure to UV rays from the sun is directly or indirectly involved in the development of melanoma in 86% of the cases [
10]. Fortunately, overall survival rate for patients with melanoma has gradually improved over the last 35 years due to improvement in detection systems along with surgical strategies. However, due to the lack of active agents for the treatment of melanoma, prognosis in patients diagnosed with malignant melanoma (stage IV) has remained grave [
11]. One of the major goals of anti-cancer drug development is to selectively target cancer cells with high specificity [
12]. Although numerous anti-melanoma drugs have been identified, the need for cancer cell-selective drugs is increasing gradually.
In this study, G361 human melanoma cells were treated with an ethanolic ARE for testing its role on cell proliferation and death. Furthermore, to compare the effects of ARE on keratinocytes with those on melanoma cells, we used HaCaT human keratinocytes to test whether ARE induces selective toxicity on melanoma cells. Furthermore, ARE-mediated changes in cell signaling pathways related with cell cycle regulation and apoptosis were determined using western blot analysis. In addition, the effects of ARE on gene expression patterns in the two cell lines were analyzed using cDNA microarray and RT-PCR analyses. Taken together, the results of this study indicate that ARE selectively induces apoptosis in melanoma cells, and presents an attractive approach for melanoma treatment.
Discussion
Studies on the effects of AREs on the skin and hair have mainly focused on aesthetical benefits, including skin regeneration, skin whitening, and the prevention of hair loss. Therefore, AREs are commonly used in herbal skin-care products. However, their effects on various types of skin cancer cells have not been fully studied. Although they have been reported to exhibit anticancer effects on other types of cancers, including colon cancer [
16], lung cancer [
17,
18], and cervical cancer, specific mechanisms of action have not been uncovered.
Melanoma can be formed by malignant transformation of melanocytes, the cells in the skin responsible for pigment production [
19]. Although the early stage melanoma can be easily removed surgically, the late stage malignant melanoma cannot be treated easily because of their lack of responsiveness to currently available therapies [
20]. In this study, the possible anticancer effects of an ethanolic ARE was tested against malignant melanoma. For this, the most frequently used G361 human melanoma cell line, which is also capable of melanogenesis was used [
21]. To identify whether its actions were selective to cancer cells, its effects on HaCaT human keratinocytes were evaluated. Although HaCaT is an artificially constructed immortalized cell line, these cells can differentiate into epidermal cells of the skin [
22]. Although treatment with ARE for 24 h reduced the proliferation of both cells, its effects on G361 were more severe than those on HaCaT cells at ARE concentrations of 0.8 mg/ml and greater (Fig.
1). Although the effects of ARE at concentrations of 0.2–0.6 mg/ml on HaCaT and G361 cells were similar at 24 h after treatment, ARE preferentially downregulated the growth of G361 cells at 48 and 72 h after treatment (Fig.
2). However, ARE treatment at concentrations of 0.2 and 0.4 mg/ml for 72 h slightly enhanced the growth of HaCaT cells. This finding corroborates previous reports on hair growth-promoting effects of AREs based on the increased proliferation of HaCaT and dermal papilla cells after low-dose ARE treatment [
7]. These results indicate that the long-term exposure of the skin to AREs can selectively reduce the growth of melanoma cells. Although Lee et al. have reported that AREs preferentially inhibit the growth of A549 lung cancer cells [
17], the mechanism of its selective anticancer effect remains unexplored.
Therefore, the effects of the ethanolic ARE on the expression of apoptosis- and cell cycle-related proteins in G361 and HaCaT cells were determined. ARE effectively increased the cleavage of caspase-3 and PARP in G361 cells in a dose-dependent manner; however, no ARE-mediated change in these two proteins was detected in HaCaT cells (Fig.
3). Because the cleavage of these two proteins indicate apoptosis, it may be considered that the ARE-mediated growth inhibition of G361 cells is caused by apoptosis; however, ARE reduces the growth of HaCaT cells by other mechanisms. Although ARE treatment on the two cell lines did not modulate the expression of cell cycle-promoting proteins, the expression of p53 and p21, negative regulators of the cell cycle, were differently affected by ARE in G361 and HaCaT cells. In G361 cells, ARE treatment decreased the expression of p53 and p21 proteins, whereas in HaCaT cells, the expression of these two proteins was increased by ARE.
This differential effect of ARE on the cleavage of apoptotic proteins, and the expression of p21 and p53 in the two cell lines were confirmed in our time-dependent experiments (Fig.
4). In G361 cells, the cleavage of caspase-3 was induced at 2 h after ARE treatment, and PARP cleavage occurred after 4 h. However, ARE treatment on HaCaT cells failed to induce the cleavage of caspase-3 and PARP until 12 h after treatment. Caspase-3 activation occurs last in the apoptotic caspase activation cascade, and cleaved-caspase-3 actively promotes the cleavage of its substrate proteins, including PARP, leading to apoptosis [
23]. PARP is normally overexpressed when cell DNA is damaged, and actively participates in DNA repair process. However, when DNA damage is too severe, PARP undergoes caspase-3-mediated cleavage to form an apoptosis-promoting protein [
24]. Therefore, the effects of ARE on the cleavage of caspase-3 and PARP in G361 cells indicate the ARE-mediated activation of the apoptotic caspase cascade. Along with apoptotic protein regulation, the ethanolic ARE differently regulated the expression of p53 and p21 proteins in the two cell lines. In G361 cells, ARE treatment immediately reduced the expression of p53 and p21 protein; however, their expression was increased in HaCaT cells. p53 is a major tumor suppressor protein, which is reported to be mutated in several cancers including melanoma. As a multi-functional protein, p53 actively regulates the expression of DNA repair and apoptosis-promoting genes [
25].
p21 gene expression is positively regulated by p53. As a negative regulator of the cell cycle, p21 not only induces growth arrest, but also delays apoptosis under severe stress conditions [
26]. Therefore, as shown in Fig.
4, ARE differentially regulates the expression of p53 and p21, and causes apoptosis in G361 cells and growth arrest in HaCaT cells.
What kind of stress-mediated signal pathway is participating in ARE-mediated p53 and p21 decrease and apoptotic cell death in G361 cells? According to the results shown in Fig.
5a, JNK and p38 kinases, well-known stress-signal mediators, are not responsible for ARE-mediated cell death in G361 cells. However, treatment with p38 and JNK inhibitors enhanced the ARE-mediated cleavage of caspase-3 and PARP, and decreased the expression of p53 and p21 proteins. These effects might be because of the multi-functional properties of p38 and JNK. Although JNK and p38 proteins induce apoptosis under several stress conditions, they can be activated by growth factors to promote anti-apoptotic gene regulation [
27,
28]. Especially, the role of p38 in apoptosis regulation is dependent on the type of cells, and Yee et al. reported that it causes G1 phase cell cycle arrest by activating p21 [
29]. Accordingly, we observed that the inhibition of p38 further led to the ARE-mediated decrease in p21 activation, thereby causing enhanced ARE-mediated apoptosis in G361 cells.
In many kinds of cancer cells, excessive ROS can promote apoptotic cell death [
30], and several natural plant-derived compounds have been reported as potential anti-cancer drugs inducing ROS-dependent apoptosis [
31]. In this study, ARE-mediated apoptosis in G361 cells was ROS-dependent. Treatment with NAC, an ROS scavenger, significantly reduced ARE-mediated caspase-3 and PARP cleavage, and apoptosis, and decrease in p53 and p21 protein expression was prevented. These data indicate that an increase in cellular ROS decreases p53 and p21 protein expression to cause ARE-mediated apoptosis in G361 cells.
Finally, candidate genes responsible for the ARE-mediated differential regulation of apoptosis, and p53 and p21 expression in G361 and HaCaT cells were identified by performing a cDNA microarray analysis. Because ARE-mediated apoptosis occurred only in G361 cells, we measured changes in the expression of apoptosis-related genes (Fig.
6a). Because ARE induced apoptosis in G361 cells in an ROS-dependent manner, changes in the expression of ROS-related genes were also studied (Fig.
6b). Some genes showed similar expression patterns after ARE treatment in both the cell types; however, other genes were modulated differently by ARE. Despite being exposed to the same stress stimulus, ARE differentially regulated p53 and p21 expression, and apoptosis in G361 and HaCaT cells. As shown in Table
1, five genes were identified to be related to apoptosis and ROS that were differentially regulated by ARE in G361 and HaCaT cells. Among them, the expression of
MDM2 and
CFLAR genes is directly related with p53 protein level. As one of major negative regulators of p53, MDM2 protein promotes the ubiquitin-mediated degradation of p53 protein [
32,
33]. On the other hand,
CFLAR is one of the target genes of p53 protein, and it binds to the death receptor signaling complex and inhibits apoptosis by blocking the activation of caspases [
34,
35]. As shown in Fig.
6c and Table
1, ARE induces
MDM2 gene expression in G361 cells, whereas it decreases its expression in HaCaT cells. The expression pattern of
CFLAR gene after ARE treatment was similar to that of p53; its expression decreased in G361 cells, whereas it increased in HaCaT cells. Therefore, these results indicate that ARE-mediated increase in ROS induces the expression of
MDM2 in G361 cells, which can be directly linked to the decrease in p53 protein level and its target gene
CFLAR. Because CFLAR is a caspase inhibitor, caspase-mediated apoptosis in G361 cells upon ARE treatment is plausible. However, ARE reduced
MDM2 gene expression and stimulated p53 protein and
CFLAR gene expression in HaCaT cells, leading to the blockage of ARE-mediated caspase activation in these cells. In the future, we propose to elucidate the mechanism of the differential regulation of
MDM2 gene expression by ARE in the two types of cells.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.