Withaferin A (WA) is a bioactive compound derived from Withania somnifera. The antitumor activity of WA has been well studied in human cancer models; however, its chemopreventive potential is unclear. In the present study, we used the skin epidermal JB6 P+ cells, a well-established model for tumor promotion, and demonstrated that WA suppressed the tumor promoter 12-O-tetradecanoylphorbol 13-acetate (TPA)-induced cell transformation and cell proliferation. Interestingly, TPA inactivated isocitrate dehydrogenase 1 (IDH1), which was reversed by WA. Similar results were also observed in mouse skin tissue. Therefore, we focused on metabolism as the potential mechanism of action. We found that mitochondrial functions were downregulated by TPA treatment, as indicated by reduced mitochondrial membrane potential, complex I activity and mitochondrial respiration. However, all of these downregulations were inhibited by WA. In addition, we examined the levels of α-ketoglutarate, a product of IDH1, and WA blocked its reduction upon TPA treatment. Finally, we detected the lactate level as a glycolysis marker, and WA suppressed its elevation caused by tumor promoter treatment. Altogether, these results suggest that WA might exert its chemopreventive activity via inhibiting not only oncogenic activation, but also IDH1 inactivation and mitochondrial dysfunction in early tumorigenesis.
Withaferin A (WA; Supporting Information Fig. S1) belongs to a group of biologically active constituents known as withanolides. The antitumor activity of WA has been reported, which shows that WA suppresses human breast cancer,[1, 2] prostate cancer, colon cancer, pancreatic cancer, glioma, renal cancer and leukemia. Moreover, a clinical trial of WA has been launched for treatment of metastatic melanoma. Not surprising, Withania somnifera is within the high-priority topics from the National Center for Complementary and Alternative Medicine. However, the chemopreventive potential of WA has not been well studied. Therefore, testing the mechanism of action of WA in chemoprevention using well-established skin cell transformation and skin carcinogenesis models is important.
Carcinogenesis is often associated with a metabolic shift. Cancer cells predominantly produce energy by a high rate of glycolysis followed by lactic acid fermentation even in the presence of oxygen, known as the “Warburg effect”. Although glycolysis is inefficient at producing ATP compared with oxidative phosphorylation (OXPHOS), the metabolic intermediates produced during glycolysis might provide a growth advantage for cancer cells. Associated with this metabolic switch is dysregulation of important metabolic enzymes. Isocitrate dehydrogenase (IDH), as a metabolic enzyme converting isocitrate to α-ketoglutarate, consists of three isoforms. Mutations in IDH1 and IDH2 have been found in glioma and leukemia, and IDH2 mutation is much less common than IDH1 mutation. Mutations in IDH1 impair its enzymatic activity, leading to decreased production of α-ketoglutarate; the latter serves as an inhibitor of hypoxia-inducing factor (HIF-1) via α-ketoglutarate-dependent dioxygenases. Recent studies also demonstrate a gain-of-function role of the IDH1 mutation. Instead of producing α-ketoglutarate, mutant IDH generates 2-hydroxyglutarate, one of the so-called “onco-metabolites”.
Although it is still controversial whether the shift from OXPHOS to aerobic glycolysis in tumor cells is because of malfunctions in mitochondria, we hypothesized that maintaining mitochondrial function and inhibiting aerobic glycolysis can be effective targets for chemoprevention.
Based on our previous study, IDH1, but not IDH2, was inactivated by tumor promoter TPA treatment in skin epidermal JB6 cells and skin tissues, and IDH1 inactivation promoted skin cell transformation. Whether IDH1 and mitochondrial function might be a potent target of WA, which leads to suppression of skin cell transformation, will be investigated using JB6 cells and skin epidermal tissues. These models are well established for screening cancer prevention agents and for mechanistic studies, and therefore suitable for our purpose.
Materials and Methods
Cell lines, reagents and treatment
Murine skin epidermal JB6 Cl-41 P+ cells (American Type Culture Collection, Manassas, VA, USA) were grown in Eagle's minimum essential medium (EMEM) containing 4% fetal bovine serum (FBS), 2 mM L-glutamine and 2.5 μg/mL penicillin and streptomycin in a 37°C incubator under 5% CO2.
Withaferin A, dissolved in dimethyl sulfoxide (DMSO; Sigma, St Louis, MO, USA), was purchased from ChromaDex (ASB-00023250; Irvine, CA, USA). Dimethylbenz[α]anthracene (DMBA; Sigma) and 12-O-tetradecanoylphorbol-13-acetate (TPA; Sigma) were dissolved in DMSO.
Twenty DBA/2 female mice (6–8-weeks old), purchased from the Jackson Laboratory (Bar Harbor, ME, USA), were housed in the LSUHSC-S Animal Resource Facility under standard regulations. The LSUHSC-S Animal Facility is AAALAC approved and maintains a consultation team of two veterinarians.
Mice were separated into four groups: control (DMSO) group (five mice); WA group (five mice); DMBA + TPA group (five mice); WA + DMBA + TPA group (five mice). The hair on the back of the mice was shaved. Two days later, for the two DMBA/TPA groups, a single dose of 100 nmol DMBA was painted on the back of the mice. After 2 weeks, either 4 μg TPA or 20 μg WA (this dose of WA was selected from our preliminary study) plus 4 μg TPA was applied to the same area (WA was applied 30 min prior to TPA treatment). The control group received DMSO treatment and the WA group received 20 μg WA. Twenty-four hours later the mice were killed, skin tissues were removed and skin epidermal cells were collected as described previously.[17, 18]
Anchorage-independent growth assay in soft agar
A soft agar-based cell transformation assay was carried out in six-well plates. The bottom agar was made using 1.25% agar, 2 × EMEM medium, 10% FBS, PBS, glutamine and penicillin and was incubated in a 50°C water bath. The mixture (0.5% agar) was then divided and various treatment reagents were added. The top agar mix contained two fractions of the above 0.5% agar mixtures and one fraction of 1 × 105 JB6 cells. The agar was allowed to solidify and was incubated in a 37°C incubator under 5% CO2 for 7 days and stained with 0.25 mg/mL Neutral Red (Sigma, St Louis, MO, USA) for 24 h.
Preparation of whole cell lysate
Collected skin cells were suspended in 200 μL of PBS containing a proteinase inhibitor cocktail (Calbiochem, La Jolla, CA, USA). Cells were sonicated on ice for two strokes (10 s per stroke) using a Fisher Sonic Dismembrator (Scale 2; Pittsburgh, PA, USA). After incubating on ice for 30 min, the cell lyaste was centrifuged at 14 000g for 15 min, and the supernatant was collected and designated as whole cell lysate.
Western blot analysis
Whole cell lysate was used for the assay. Antibodies against IDH1 (sc-49996) and glyceraldehyde 3-phosphate dehydrogenase (GAPDH; sc-32233) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA).
NADP+-dependent IDH activity assay
NADP+-dependent IDH activity (including both IDH1 and IDH2) was measured in a 400-μL assay solution containing 33 mM of potassium phosphate buffer (pH 7.0), 0.167 mM NADP+, 3.3 mM Mg2+, 0.167 mM threo-DS-isocitrate, 10 μL (approximately 20 μg) of freshly isolated whole cell lysate or 10 μL of different concentrations of WA in DMSO and 5 μL (22.5 μg) IDH (10 units/mL) (Sigma). The activity of IDH was monitored through the reduction of NADP+ to NADPH, which was recorded using spectrophotometry at 340 nm for 5 min.
Measurement of oxygen consumption of JB6 cells
JB6 P+ cells (2 × 106/mL) in growth medium were suspended in a thermostated closed vessel at 37°C. Oxygen consumption was measured polarographically using a Clark–type O2 electrode (Yellow Spring Instruments, Yellow Springs, OH, USA). The rate of mitochondrial O2 consumption was determined as the antimycin-sensitive rate after addition of antimycin A to the final concentration of 1 μM from at least three experiments.
Detection of mitochondrial membrane potential
JB6 P+ cells (5 × 103) were seeded in 96-well plates with 150 μL growth medium. Twenty-four hours after plating, cells were treated as indicated in each experiment. After washing with PBS, cells were incubated in fresh medium containing 2 μg/mL of 5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraethylbenzimidazol-carbocyanine iodide (JC-1; Molecular Probes, Eugene, OR, USA) for 30 min. The dye was then removed and cells were washed with PBS. Fluorescence intensity was measured immediately using fluorescence spectrometry (Synergy HT, BioTek, Winooski, VT, USA). For JC-1 green, Ex = 485, Em = 528; for JC-1 red, Ex = 530, Em = 590. The fluorescence signals from the cells only (no JC-1 dye added) were used to subtract the sample values. Experiments were repeated three times and triplicate samples at least were included in each experiment.
Mitochondrial complex I activity assay
Complex I-specific activities were measured as described.[20, 21] Mitochondrial samples isolated from JB6 cells were subjected to three fast freeze–thaw cycles in hypotonic buffer. The protein concentration was measured and adjusted to 1.33 μg/μL before the assay. The assay mixtures, which contained 25 mM potassium phosphate buffer (pH 7.2), 5 mM MgCl2, 2 mM KCN, 2.5 mg/mL bovine serum albumin (fraction V), 0.13 mM NADH, 65 mM coenzyme Q1 and 2 mg/mL antimycin A, were incubated at 30°C for 1 min. Mitochondria were added to initiate the reaction, and the initial rate of NADH oxidation was monitored at 340 nm for 1 min (ΔA). The reaction was inhibited by 2 μL of 2 mg/mL rotenone and the rate of NADH oxidation was monitored for 1 min (ΔAr). The relative complex activity was calculated according to the following formula: ΔA-ΔAr.
Detection of mitochondrial reactive oxygen species (ROS) generation
The mitochondrial levels of ROS were assayed using mitoSOX Red dye (Grand Island, NY, USA). Briefly, 5000 JB6 P+ cells were seeded in a 96-well plate and incubated overnight. Twenty-four hours after plating, cells were treated as indicated in each experiment. After washing with warm PBS, cells were incubated with fresh medium containing 5 μM mitoSOX Red for 15 min at 37°C. The fluorescence intensity was measured at excitation/emission of 530/590 nm using fluorescence spectrometry. The cells-only sample was used as the background. Experiments were repeated three times and triplicate samples at least were included in each experiment.
Determination of α-ketoglutarate, ATP, lactic acid and pyruvate levels
The levels of α-ketoglutarate, ATP, lactate and pyruvate were determined using the Alpha-Ketoglutarate Assay Kit (BioVison; K677-100; Mountain View, CA, USA), ATP Assay Kit (BioVison; K354-100), Lactate Assay Kit (BioVison; K607-100) and Pyruvate Assay Kit (BioVison; K677-100), respectively, following the instructions provided by the manufacturer. Whole cell lysate from mouse skin tissues was diluted to 2 μg/μL in PBS and deproteinized by passing through a 10 kD cut-off membrane (VWR; 82031-348, Radnor, PA, USA) in α-ketoglutarate, lactate and pyruvate detection. For each sample, 50 μL of the whole cell lysate filtrate was used.
One-way anova followed by Newman–Keuls post-test was used for multi-group comparisons. All experiments were repeated at least three times. P < 0.05 was considered significant.
Withaferin A suppresses TPA-induced neoplastic transformation of JB6 P+ cells
Initially, to test whether WA suppresses tumor promotion, we performed a cell transformation assay using the well-characterized skin epidermal JB6 P+ cells. We treated JB6 P+ cells with WA at 0.625 μM; the concentration was selected as non-toxic to the cells based on MTT assays. As shown in Figure 1, WA significantly suppressed tumor promoter TPA (5 nM)-induced cell transformation. The WA caused a 90% decrease in the number of colonies compared with the TPA group. In addition, to intensify this data, we used another tumor promoter, ultraviolet C (UVC) (20 J/m2), and observed that WA made an 84% decrease in the number of colonies compared with the UVC group (Fig. S2; the experimental procedures are described in the Supporting Information Materials and Methods). These results indicate that WA inhibits skin cell neoplastic transformation.
Withaferin A inhibites TPA-induced decreases in IDH1 protein expression and enzymatic activity
As known, IDH1 converts isocitrate to α-ketoglutarate in the cytosol. Recent findings of IDH1 mutations in human glioblastoma and adult acute leukemias highlight the importance of metabolism in cancer. (,) However, the role of IDH1 in early cancer development is not completely understood. Our previous study revealed that decreased IDH1 expression enhances whereas increased IDH1 expression suppresses skin cell transformation, suggesting IDH1 activity is important for suppressing tumor promotion. To study whether WA can affect cellular metabolism and IDH1 activity in the early stage of carcinogenesis, we treated mouse skin with DMBA and TPA and isolated skin cells. We then detected the expression and activity levels of IDH1. As shown in Figure 2(a,b) both the expression and activity levels of IDH1 were decreased by tumor promoter TPA treatment, which were inhibited by TPA. Similar results were also observed in JB6 P+ cells treated with TPA (Fig. 2c,d). These results indicate that WA is able to suppress the decreases in IDH1 expression and activity during the early stage of skin carcinogenesis.
Next we determined whether WA enhances IDH1 enzymatic activity directly. Different concentrations of WA (0–10 μM) were loaded into the IDH1 activity assay solution directly. As shown in Figure 3, WA increased IDH1 activity at certain concentrations, suggesting that IDH1 might be a potential molecular target of WA for its chemoprevention-promoting activities.
Withaferin A reverses TPA-induced mitochondrial dysfunction
The cancer cell prefers to generate energy from glycolysis instead of oxidative phosphorylation and mitochondrial metabolism is often dampened. Therefore, maintaining mitochondrial metabolism might be a valid approach to suppressing tumorigenesis. To determine how mitochondrial metabolism is altered during tumor promotion, an electron respiration chain component, mitochondrial complex I activity was measured using the mitochondrial fractions isolated from the treated JB6 P+ cells. As shown in Figure 4(a), complex I activity was decreased upon tumor promoter TPA treatment, whereas WA reversed it. Other markers for mitochondrial functions were also detected after TPA treatment, which included decreased mitochondrial membrane potential (Fig. 4b), decreased oxygen consumption (Fig. 4c), and increased mitochondrial ROS generation (Fig. 4d). WA blocked all of these malfunctions. These results indicate that WA is effective for reversing tumor promoter TPA-induced mitochondrial dysfunction.
Withaferin A maintains the function of IDH1 in early skin tumorigenesis
As a cytosolic metabolic enzyme, the main function in metabolism of IDH1 is to convert isocitrate to α-ketoglutarate (α-KG). As we determined that WA inhibited the decreases of IDH1 expression and activity, to further study whether WA affects IDH1 functions in metabolism we detected the levels of α-ketoglutarate using mouse skin tissues. As shown in Figure 5(a), TPA treatment caused a decrease in α-ketoglutarate production and this decrease was clearly suppressed by WA treatment. In addition, we detected ATP levels in the same sample. As shown in Figure 5(b), TPA treatment reduced ATP levels, which were suppressed by WA. These results indicated that WA promoted the non-mutated IDH1 conversion of isocitrate to α-KG and maintained its normal metabolic enzyme activity.
Withaferin A suppressed TPA-induced lactic acid production in mouse skin tissue
Metabolism in cancer cells shifts from oxidative phosphorylation to aerobic glycolysis, which is accompanied by an increase in glucose consumption and lactate production. After examining the changes in mitochondrial functions and IDH1 expression and activity in TPA-induced early carcinogenesis models, we detected whether WA inhibited glycolysis in mouse skin. Figure 6(a) shows that in skin epidermal tissues, the levels of lactate were increased after treatment with DMBA/TPA compared with the control group and the increase was suppressed by WA treatment. Meanwhile, we detected levels of pyruvate, the end product of glycolysis. As shown in Figure 6(b), pyruvate almost remained at similar levels after all of these treatments, which is consistent with our previous study. These results indicate that WA suppressed lactate production during the early stage of carcinogenesis.
Withaferin A, a major withanolide found in the traditional Indian medicine Ashwagandha, has been widely studied for treating human cancers and currently is in clinical trial for melanoma therapy. In summary, WA suppresses tumor growth by inhibiting cell proliferation, the cell cycle, inflammation and angiogenesis, or by promoting apoptosis and oxidative stress. In particular, WA inhibits Notch-1 signaling and downregulates prosurvival pathways in colon cancer cell lines (HCT-116, SW-480 and SW-620). Withaferin A downregulates the expression of mammalian target of rapamycin signaling components, pS6K and p4EBP1, and activates JNK-mediated apoptosis in colon cancer cells. Withaferin A also causes G2 and M phase cell cycle arrest in human breast cancer cells, and induces apoptosis in human leukemia U937 cells and prostate cancer cells. The antiangiogenic effect of WA relies on targeting the intermediate filament protein vimentin. However, the chemopreventive activity of withaferin A is likely to be mediated by different mechanisms.
Neoplastic transformation of a normal cell requires many gene changes, mainly in two categories: activation of oncogenes and/or inactivation of tumor suppressor genes. Therefore, as neoplastic transformation occurs during the early stage of carcinogenesis, transformation inhibition could be an effective approach for chemoprevention. In our present study, we confirmed that WA significantly suppressed tumor promoters TPA and UVC-induced neoplastic transformation in JB6 P+ cells. In contrast, the balance between proliferation and cell death is pivotal for carcinogenesis, the uncontrolled and rapid proliferation of cells could be a characteristic of benign tumors. As a tumor promoter, TPA induces neoplastic transformation; in addition, accumulating evidence has shown that TPA could stimulate cell proliferation.[28, 29] We detected the expression levels of proliferating cell nuclear antigen, which is a widely used proliferation marker, and found that WA suppressed TPA-induced cell proliferation (Fig. S3a). Next we tested the possibility of WA inducing apoptosis as the mechanism action. As shown in Figure S3b,c, WA did not induce apoptosis in skin epidermal tissues. Based on the present study, the mechanism of the chemopreventive activity of WA might be mediated by inhibiting neoplastic transformation and cell proliferation but not by inducing apoptosis.
Isocitrate dehydrogenases convert isocitrate to α-KG to generate NADPH, which supplies reducing energy in enzymatic reactions as a cofactor. They consist of three isoforms, IDH1, IDH2 and IDH3; the former two act in the cytoplasm and mitochondria, respectively, and the latter acts in the TCA cycle. Our previous study showed that tumor promoter decreased the expression and activity levels of IDH1, but not IDH2, in the tumor promotable JB6 P+ cell model. In the present study, our data show that WA suppresses the decreases in IDH1 protein expression and activity during the early stage of skin carcinogenesis. Importantly, we found that WA directly increased IDH1 activity, suggesting IDH1 might be a potential target of WA. However, how WA and IDH1 might interact is not known, which will be examined in our future studies.
Can WA affect IDH1 at the transcription level? Next we treated JB6 P+ cells with WA and/or TPA and found that the mRNA levels of IDH1 remained the same (Fig. S4), and similar results were found for IDH2. These data suggest that WA may act on IDH1 at the post-translational level instead of the transcriptional level, or WA might directly interact with IDH1 and increase its activity (as suggested by Fig. 3). We will work on this interesting hypothesis in our future studies.
In the JB6 cell model, TPA treatment has been shown to activate the ERK pathway, which contributes to neoplastic transformation. In addition, the activated AKT signaling pathway by TPA also contributes to skin tumor promotion. We detected the phosphorylation status of ERK, JNK (both are the members of the MAPK family) and AKT. As shown in Figure S5, WA did not affect TPA-induced phosphorylation of ERK, JNK or AKT. These data suggest that the MAPK and AKT signaling pathways might not contribute to the mechanism of action of WA during the early stage of skin carcinogenesis. As a pivotal sensor of energy status, AMP-activated protein kinase (AMPK) plays an important role in cancer cell metabolism. In our future studies, the link between IDH1 and the AMPK pathway will be addressed.
Cell growth requires energy. How to obtain the energy makes a difference between cancer cells and normal cells. Even with ample oxygen, cancer cells prefer glycolysis over oxidative phosphorylation to produce ATP. On one hand, accumulating evidence has shown that there is mitochondrial malfunction in cancer cells, so do mutations in mtDNA and metabolic enzymes. In particular, Warburg first hypothesized that there is mitochondrial dysfunction in cancer cell development. mtDNA mutations have been found in various kinds of cancer, such as breast and colorectal cancer. Mutations of metabolic enzymes have also been identified, including succinate dehydrogenase, fumarate hydratase and IDH.[11, 37] On the other hand, some studies have shown that there is natural mitochondrial function in several cancer cells.[38, 39] Although the role of mitochondrial metabolism in cancer development remains a disputed question, our in vitro studies show that tumor promoters induced mitochondrial dysfunction and WA could reverse the effects in the early stage of skin carcinogenesis.
Wild-type IDH1 converts isocitrate to α-KG, which activates dioxygenases. Mutant IDH1 gains a new function, which produces 2-hydroxyglutatarate. Based on our previous studies, both UV irradiation (in vitro) and tumor promoter TPA (in vitro and in vivo) cause downregulation of IDH1 expression and activity in skin epidermal cells. In addition, knockdown of IDH1 enhances whereas overexpression of IDH1 suppresses skin cell transformation, suggesting a tumor suppressive role for IDH1 in the early stage of skin carcinogenesis. In addition, the expression and activity levels of IDH2 are not altered by tumor promoter treatment. In the present study, we found that WA promotes the conversion of isocitrate to α-KG. However, as an important metabolic enzyme, glutamate dehydrogenase (GDH) converts glutamate to α-KG as well. We detected levels of GDH activity using JB6 P+ cells. As shown in Figure S6, GDH activity was not significantly altered by either the TPA or WA treatment. Furthermore, our result (Fig. S7) demonstrated that WA did not alter the protein expression of IDH2. These results suggest that WA exerts its chemopreventive activity by maintaining IDH1 activation during early tumorigenesis.
In summary, our studies indicate that IDH1 and mitochondrial dysfunction will be induced during early skin tumorigenesis. Withaferin A can preserve IDH1 activity and mitochondrial function, which might provide a novel mechanism for WA as a chemopreventive agent.
The authors thank Drs Tammy Dugas, Tak Yee Aw and Magadas Citrus and student workers Alexandreia Wilson and Alexis Ellis at LSUHSC-Shreveport for technical support.