Molecular mechanisms underlying selective cytotoxic activity of ...

[Cancer Biology & Therapy 7:4, 577-586; April 2008]; ©2008 Landes Bioscience

Research Paper

Molecular mechanisms underlying selective cytotoxic activity of BZL101, an extract of Scutellaria barbata, towards breast cancer cells Sylvia Fong,1 Mark Shoemaker,1 Jaclyn Cadaoas,1 Alvin Lo,1 Wayne Liao,2 Mary Tagliaferri,1 Isaac Cohen1 and Emma Shtivelman1,* 1BioNovo

Inc.; Emeryville, California USA; 2Phalanx Biotech Group; Palo Alto, California USA

Key words: ROS, PARP, BZL101, glycolysis, scutellaria barbata

We studied the mechanism of the cytotoxic activity of BZL101, an aqueous extract from the herb Scutellaria barbata D. Don, which is currently in phase II clinical trial in patients with advanced breast cancer. The phase I trial showed favorable toxicity profile and promising efficacy. We report here that BZL101 induces cell death in breast cancer cells but not in non-transformed mammary epithelial cells. This selective cytotoxicity is based on strong induction by BZL101 of reactive oxygen species (ROS) in tumor cells. As a consequence, BZL101 treated cancer cells develop extensive oxidative DNA damage and succumb to necrotic death. Data from the expression profiling of cells treated with BZL101 are strongly supportive of a death pathway that involves oxidative stress, DNA damage and activation of death-promoting genes. In breast cancer cells oxidative damage induced by BZL101 leads to the hyperactivation of poly (ADP-ribose) polymerase (PARP), followed by a sustained decrease in levels of NAD and depletion of ATP, neither of which are observed in non-transformed cells. The hyperactivation of PARP is instrumental in the necrotic death program induced by BZL101, because inhibition of PARP results in suppression of necrosis and activation of the apoptotic death program. BZL101 treatment leads to the inhibition of glycolysis selectively in tumor cells, evident from the decrease in the enzymatic activities within the glycolytic pathway and the inhibition of lactate production. Because tumor cells frequently rely on glycolysis for energy production, the observed inhibition of glycolysis is likely a key factor in the energetic collapse and necrotic death that occurs selectively in breast cancer cells. The promising selectivity of BZL101 towards cancer cells is based on metabolic differences between highly glycolytic tumor cells and normal cells.

Introduction A growing number of women in the Western hemisphere, when diagnosed with breast cancer, seek alternative or complimentary treatments. Chinese herbal remedies are among the most popular *Correspondence to: Emma Shtivelman; BioNovo, Inc.; 5858 Horton Street; Ste. 375; Emeryville, California 94608 USA; Email: [email protected] Submitted: 11/26/07; Revised: 01/07/08; Accepted: 01/07/08 Previously published online as a Cancer Biology & Therapy E-publication: http://www.landesbioscience.com/journals/cbt/article/5535 www.landesbioscience.com

choices among the treatments available to cancer patients.1 However, there is a profound lack of knowledge about the efficacy and the active principals in the herbal extracts that have been prescribed in traditional Chinese medicine for centuries to treat cancer. Considering that well over a half of existing mainstream chemotherapeutic drugs were derived from botanical sources,2 there is a compelling reason to seek novel chemotherapeutics in the herbal extracts traditionally prescribed for cancer treatment. We have focused on the herb Scutellaria barbata (BZL101), that is widely used as an antitumor agent in China. We have recently concluded a phase I clinical trial, with promising results showing safety and efficacy in treating advanced breast cancer.3 At the same time, our efforts have been directed towards understanding the mechanism of the cytotoxic activity of BZL101. We have previously reported that treatment of breast cancer cells with BZL101 inhibits their growth and induces cell death.4,5 However, the mechanism of cell death induced by BZL101 in breast cancer cell lines was not understood. Only a relatively small fraction of cells died through apoptotic pathway, exhibiting binding of Annexin V, DNA fragmentation and caspase activation.3 The majority of cells dying after treatment with BZL101 did not exhibit any of these features. Moreover, we could not detect release of cytochrome c or other apoptogenic proteins from the mitochondria of treated cells. We have also described the lack of cytotoxicity of BZL101 towards normal human cells.3 These observations prompted us to consider a mechanism of BZL101-induced cell death that is different from the classical apoptotic process. It is well known that cancer cells are notoriously resistant to apoptotic death stimuli due to various deficiencies in the initiation, regulation and execution of apoptosis. Some of the fundamental differences between tumor and normal cells involve the different metabolic processes utilized for energy production. The so-called Warburg effect describes that the vast majority of tumors rely on glycolysis for their energy needs even under conditions of normal oxygen supply (reviewed in ref. 6). Most normal cells and tissues produce ATP through oxidative phosphorylation, with little or no need for glycolysis as a source of energy. Inhibition of glycolysis therefore seems to be a logical goal in developing drugs that will specifically target tumors because of their dependence on glycolysis, but will spare normal tissues. Relatively recently, approaches to cancer therapy indeed started to exploit the long-known metabolic

Cancer Biology & Therapy

577

Selective cytotoxicity of BZL101 towards breast cancer cells

differences between tumors and normal tissues with the goal of induction of tumorspecific necrosis. This report documents that the active compounds in BZL101 extract exert a selective cytotoxic activity towards breast cancer cell lines by inducing in them necrotic death, but fails to target normal breast cells. The specificity of BZL101 is apparently based on its ability to induce high levels of reactive oxygen species (ROS) in tumor cells, but only moderately so in normal cells. In tumor cells, high levels of endogenous ROS triggered by BZL101 lead to significant oxidative DNA damage and hyperactivation of poly(ADP-ribose) polymerase (PARP). This, in turn, results in depletion of NAD+/ NADH in tumor cells. These cells, dependent on aerobic glycolysis, respond to the PARP-mediated rapid decline in NAD+/ NADH levels by depletion of ATP stores, and undergo an energetic collapse followed by necrotic cell death. Our results demonstrate that cytotoxic activity of BZL101 is specific towards tumor cells because of the inherent differences in the metabolic preferences between tumor and normal cells.

Results

Figure 1. BZL101 induces DNA damage via generation of reactive oxygen species. (A) BZL101 induces DNA breaks that are repaired slower in breast cancer lines compared to normal cells. Cells were treated with BZL101 for 15 minutes, after which cells were harvested for comet assay immediately or allowed to recover in fresh medium without BZL101 for 1 and 3 hours. (B) Cells were treated as in (A), and images of cells subjected to comet assay and stained with SYBR green were taken. (C) BZL101 induces ROS as measured by conversion of CM-H2DCFDA to its fluorescent form. Shown are histograms of the FL1 fluorescence as determined by FACS analysis of live cells before and after 15 minute treatment (top), as well as BZL101 treated cells pre-incubated with 10 mM NAC (lower). The punctate line histograms show fluorescence levels in untreated cells, the solid lines show levels in cells treated with BZL101 (top) or with NAC and BZL101 (lower histogram). (D) BZL-induced DNA damage is alleviated in the presence of ROS scavengers. Cells were pretreated with 10 mM of pyruvate or NAC for 30 minutes prior to addition of BZL101 for an additional 15 minutes and processed for comet assay. All results are average of at least three experiments.

BZL101 induces DNA damage in breast cancer cells. We have reported previously that treatment of cells with BZL101 results in the accumulation of reactive oxygen species (ROS), in particular superoxide as detected by staining with hydroethidium.3 In addition, we have observed that in response to BZL101 treatment some of the breast cancer cell lines accumulate in the G2 phase of cell cycle,3 which is often a hallmark of DNA damage. It is well known that DNA damage frequently occurs as a consequence of oxidative stress. Together, these observations prompted us to examine if BZL101 induces DNA damage. To investigate this possibility we have employed comet assay that allows visual analysis of the DNA breaks in the individual cells after electrophoresis of cells suspended in agarose. Two breast cancer cell lines (SKBr3, highly sensitive to BZL101, and somewhat less sensitive to BZL101 line BT474) were analyzed for DNA damage along the immortalized non-transformed line of mammary epithelial cells MCF10A. Cells were treated with BZL101 for 15 minutes and analyzed for DNA damage immediately, or after recovery in fresh media without BZL101. Figure 1A shows that 80 to 100% of cells in cancer cell lines developed DNA breaks after a short 15 minute incubation with BZL101. Less than a half of the cells of the normal immortalized mammary epithelial cell line MCF10A accumulated DNA damage. In addition, as seen in micrographs in Figure 2B, the comet tails in SKBr3 cells treated with BZL101 were much larger that these in MCF10A. The quantitative analysis of 578

comet tail momentum (a reflection of the extent of DNA damage in each cell) showed very low levels of DNA damage in MCF10A cells, bordering on the sensitivity of the image analysis software used for the analyses (not shown). Similarly, comparison of tail momentum in SKBr3 cells and normal fibroblasts IMR90 treated with BZL101 showed a much lower level of DNA damage per cell in IMR90 cells (not shown). Significant differences have also emerged in the rate of repair of DNA breaks in different cells (Fig. 1A and B). In non-transformed cells most of the damage was repaired within the first hour after treatment, while in sensitive SKBr3 cells much of the damage remained unrepaired even three hours after the treatment. Inefficient repair of DNA damage in breast cancer cells could be a reflection of the more extensive damage induced in them by BZL101. DNA damage induced by BZL101 is caused by increased production of ROS. Next, we examined if differential induction of DNA damage by BZL101 in cancer versus non-transformed cells might be related to the different levels of oxidative stress generated


2008; Vol. 7 Issue 4


Table 1 Percentages of cell death caused by BZL101 correlates with the extent of oxidative stress and DNA damage

BT474

MCF10A

3.58

1.66

1.15

% cells with comets

97.80

80.30

58.07

% dead cells (24 hours treatment)

47.55

39.12

10.22

% dead cells (1 hour treatment)

39.45

16.02

7.03

Fold increase in ROS

SKBr3

Cells were treated with BZL101 for 15 minutes for quantification of increase in intracellular ROS and comet assays. For quantification of cell death cells were treated with BZL101 continuously for 1 day, or for 1 hour after which the BZL101 containing medium was removed and replaced by fresh medium for additional 23 hours. Percentages of dead cells represent percentage of cells that became permeable for propidium iodide, as quantified by FACS analysis of cultures harvested by trypsinization, and are average of two to three experiments.

Figure 2. BZL101 induces oxidative DNA damage. (A) Western blot analysis of Nrf2 expression in cells treated with BZL101 for different times. The lower band represent a non-specific protein band binding to the antibody. (B) SKBr3 cells were treated with BZL101 for 1 hour, fixed, permeabilized, stained for 8-oxoguanine with Avidin-FITC (FL1), and analyzed by FACS. (C) Determination of the number of apurinic/apyrimidinic sites in cellular DNA of BZL101 treated cells. The Y axis shows number of AP sites per 105 nucleotides of DNA (a.u., arbitrary units) in SKBr3 cells treated with either BZL101 or hydrogen peroxide at 25 μM for 1 hour.

by BZL101 treatment. We used the cell-permeable ROS-sensitive probe CM-H2DCFDA that detects both peroxide and superoxide after the latter is rapidly transmuted into peroxide within cells. As shown in Figure 1C, BZL101 induced a significant accumulation of ROS in SKBr3 cells. Incubation of cells with ROS scavenger www.landesbioscience.com

N-acetyl-cysteine (NAC) prior to addition of BZL101 prevented most of the increase in ROS generation, confirming that the conversion of non-fluorescent CM-H2DCFDA into fluorescent compound is indeed due to ROS. Table 1 shows the fold increase of ROS levels in cell lines treated with BZL101. It is of particular interest that the increase in ROS correlates well with the degree of DNA damage induced in these cells (Table 1). The lowest induction is seen in MCF10A cells which also have the lowest number of comets after treatment with BZL101, and the highest increase in ROS is observed in SKBr3 cells where the DNA damage is most extensive (Table 1). To further implicate ROS in the induction of DNA damage, we have analyzed comet formation in cells pretreated with the antioxidants NAC and pyruvate prior to the addition of BZL101. Both compounds significantly reduced the number of cells forming comets (Fig. 1D). DNA damage repair in cancer cells in the presence of NAC or pyruvate was also greatly accelerated (not shown). At the same time, pretreatment of cells with the nitric oxide scavenger PTIO had no effect on the numbers of cells with comets (not shown), indicating that most of DNA lesions induced by BZL101 are oxidative in nature. To confirm that BZL101 induces oxidative stress responses, we examined the levels of transcription factor Nrf2 in BZL101 treated cells. Nrf2 is a key regulator of the phase II detoxifying and antioxidant enzymes that are upregulated in response to oxidative stress.7 Western blot analysis showed a significant and sustained increase in Nrf2 levels in BZL101 treated BT474 cells (Fig. 2A) and SKBr3 cells (not shown), and a more transient increase in MCF10A cells. We have observed elevated levels of ROS in breast cancer cells at 6 to 18 hours after addition of BZL101 (not shown), perhaps relevant to the sustained activation of Nrf2. To verify that BZL101 leads to oxidative DNA damage, we have examined if the DNA of BZL101-treated cells contains 8-oxoguanine, the most ubiquitous marker of DNA oxidation. We have quantified formation of 8-oxoguanine through flow cytometric analysis of fixed permeabilzed cells incubated with avidin fluorescein, that was shown to bind relatively specifically to 8-oxoguanine.8 Figure 2B shows that there is a clear increase in binding of avidin to BZL101 treated SKBr3 cells versus untreated cells. This increase was abolished if cells were pretreated with NAC prior to addition of BZL101 (not shown), confirming the specificity of the observed staining. Similar results were obtained when a monoclonal antibody against 8-oxoguanine was used to stain BZL101 treated cells (not shown). We have also analyzed if BZL101 induced formation of apurinic/ apyrimidininc (AP) sites that result from the base excision and repair of oxidized, deaminated or alkylated bases. Figure 2C shows that treatment with BZL101 leads to the generation of AP sites in DNA, with a frequency comparable to that observed after treatment with hydrogen peroxide. These results strongly implicate induction of ROS as causative in DNA damage induced by BZL101. DNA damage induced by BZL101 treatment is responsible for its cytotoxic activity. We have examined the possible relationship between DNA damage induced by BZL101 and its cytotoxic activity. First, we analyzed if there is a correlation between the extent of DNA damage and induction of cell death after treatment with BZL101. As shown in Table 1, such a correlation is obvious in cell lines SKBr3, BT474 and MCF10A. Next, we have examined if BZL101 can induce cell death after a short treatment. The rationale for these


579


experiments was based on the observations described above that both increase in ROS and DNA damage occur very soon (within first 15 minutes) after addition of BZL101 (Fig. 1). If indeed ROS and DNA damage are critical for the cytotoxic activity of BZL101, then even a short treatment with BZL101 should lead to significant cell death. Measurements of cell death showed that a one hour treatment with BZL101 followed by a day-long incubation in fresh medium sufficed to induce death in a significant numbers of cancer cells, but not in MCF10A (Table 1). Figure 1D shows that ROS scavengers strongly reduce the extent of DNA damage induced by BZL101. We have next examined if NAC and pyruvate could also inhibit cell death induced by BZL101. Indeed, cell death was prevented by both NAC and pyruvate, once again confirming the pivotal role of ROS and ROS induced DNA damage in the cytotoxic activity of BZL101 (Fig. 3A). The preventive effect of antioxidants on BZL101 cytotoxic activity was observed in three additional breast cancer cell lines MCF7, MDA MB 361 and MDA MB 468. Treatment of these cells with BZL101 for one day resulted in 30% to 50% cell death, and death was entirely prevented by pre-treatment with either NAC or pyruvate (not shown). We have reported previously that DNA fragmentation, caspase activation and binding of Annexin V (all considered to be hallmarks of apoptosis) were observed only in a moderate percentage of breast cancer cells SKBr3 and BT474 treated with BZL101.3 Figure 3B shows that a relatively low numbers of cells treated with BZL101 enter apoptotic pathway, and this could be prevented by inclusion of NAC or pyruvate. We conclude that the increase in ROS induced by BZL101 and the resulting DNA damage are responsible for the tumor cell-selective cytotoxic activity of BZL101. PARP activation by BZL plays a prominent role in cell death induced by BZL. Our observations that ROS induced DNA damage is causative in the death of breast cancer cells induced by BZL101 prompted us to examine if hyperactivation of PARP is involved in the cellular responses triggered by BZL101. PARP is known to be activated by several types of genotoxic stress including oxidative DNA damage (reviewed in ref. 9), and the hyperactivation of PARP induces cell death classified as “programmed necrosis”.10 We have analyzed the levels of the PAR polymer synthesized by PARP in cells treated with BZL101 in the presence or absence of PARP inhibitors. A very strong transient activation of PARP was observed in SKBr3 cells within minutes of BZL101 treatment, but in MCF10A cells induction of PAR synthesis by BZL101 was insignificant (Fig. 4A). Addition of PARP inhibitor 3,4-dihydro-5[4-(1-piperidinyl)butoxyl]-1(2H)-isoquinolinone (DPQ) abolished accumulation of PAR on cellular proteins, confirming the role of PARP in the synthesis of PAR in BZL101 treated SKBr3 cells. A different PARP inhibitor, 3-amino-benzamide (3-AB), also had a strong inhibitory effect on PAR synthesis (not shown). Activation of PARP by BZL101 was additionally confirmed by immunofluorescence staining of SKBr3 cells for the accumulation of PAR polymer (Fig. 4B). A strong PAR immunoreactivity was observed in cells treated with BZL101 whereas it was not detectable in untreated cells (not shown) and in cells pre-treated with the PARP inhibitor 3-AB (or DPQ) prior to application of BZL101 (Fig. 4B). We then examined if PARP activation plays a role in cell killing by BZL101. Figure 4C shows that treatment with PARP inhibitor DPQ 580

Figure 3. Cell death induced by BZL is prevented by ROS scavengers. (A) Cells were treated with BZL101 for 24 hours in absence or presence of NAC or pyruvate, and analyzed by FACS after staining with Annexin V-Alexa 488 and propidium iodide. (A) Percentages of dead cells, quantified as all cells positive for PI uptake, irrespective of their AnnexinV-binding status (PI-positive cells could be either necrotic or late apoptotic cells). (B) Cells were treated as in (A), and early apoptotic cells were detected by Annexin V-Alexa 488 staining; only cells negative for PI staining were quantified because of the well known tendency of necrotic, PI positive cells, to bind Annexin V non-specifically.

strongly decreased numbers of dead (PI-positive) cells in p53-mutant cell lines SKBr3 and BT474 (Fig. 4D) as well as MDA MB468 (not shown). Unexpectedly, numbers of early apoptotic cells (Annexin V-single positive) were increased in SKBr3 and BT474 treated with BZL101 in presence of DPQ (Fig. 4D). Analysis of dead, PI-permeable cells revealed that in the presence of DPQ significantly more dead SKBr3 cells bind Annexin (Fig. 4E). Annexin V binding to dead cells is not always indicative of the apoptotic nature of death since necrotic cells tend bind Annexin V non-specifically. However, concomitant increase in DNA fragmentation in SKBr3 cells treated with BZL101 in presence of DPQ strongly suggest that inhibition of PARP in SKBr3 and BT474 cells (not shown) indeed augments apoptosis while decreasing necrotic death. Inhibition of PARP in p53 wild type lines MCF7 and MDA MB 361 treated with BZL101 did not result in an overall decrease in cell death (Fig. 4C). Similarly to p53 mutant cell lines, necrotic death was inhibited, as seen from the reduction in numbers of dead cells




Figure 4. Activation of PARP by BZL101 and its role in cell death. (A) MCF10A and SKBr3 cells were treated with BZL101 for the times indicated in presence or absence of PARP inhibitor DPQ (3,4-Dihydro-5-[4-(1-piperidinyl)butoxyl]-1(2H)-isoquinolinone at 20 μM. Cell lysates were electrophoresed and blotted with the anti-PAR antibody or with antibody to GAPDH for loading control. (B) SKbr3 cells were stained for presence of PAR polymer in the nuclei with an anti-PAR antibody 15 minutes after treatment with BZL101 with or without PARP inhibitor 3-AB. (C) Effect of PARP inhibition on cell death. The chart shows percentages of all PI-positive (dead) cells after 24 hours of treatment with BZL101 in absence or presence of 15 μM of DPQ. DPQ alone had no effect on cell viability (not shown). (D) Inhibition of PARP in breast cancer cells leads to initiation of apoptosis. The chart shows cells treated as in 4C, but that are still negative for PI uptake and already positive for Annexin V binding (early apopotic). (E) Inhibition of PARP augments apoptosis. Shown are the results of analysis of annexin binding in PI-positive (dead) cell populations shown in 4C, and of extent of DNA fragmentation. All results shown are average of three experiments, and the error bars represent SE.

that did not bind Annexin V in MDA MB 361 cells (Fig. 4E) and in MCF7 cells (not shown). However, unlike in p53 mutant cells lines, the increase in apoptotic cell death induced by addition of DPQ was much more pronounced as seen from a dramatic increase in Annexin positive cells and DNA fragmentation (Fig. 4E). In the non-transformed MCF10A cells there was little effect of PARP inhibition on the already insignificant levels of cell death induced by BZL101 (Fig. 4C and D). Given the negligible PARP activation in MCF10A cells (Fig. 4A), one would not expect the www.landesbioscience.com

PAR inhibitor to have an effect on the responses of these cells to BZL101. We conclude that inhibition of PARP has a strong inhibitory effect on necrotic death induction by BZL101 irrespective of p53 status. However, induction of apoptosis when PARP is inhibited in the absence of functional p53, is not efficient. Therefore, inhibition of PARP in p53 mutant cells treated with BZL101 has an overall positive effect on survival. In cells containing wild type p53 inhibition of PARP during the treatment with BZL101 results in a much


581


Figure 5. BZL101 causes a bioenergetic collapse and inhibition of glycolysis in SKBr3 but not MCF10A cells. (A) NAD+/NADH levels in cells treated with BZL101 Results are average of three experiments, and were normalized to the protein content of cell extracts used for analyses. (B) ATP levels in cells treated with BZL101. (C) Levels of secreted lactate in cells treated with BZL101. (D) Enzymatic activity of LDH and GAPDH after 6 hours treatment with BZL101. All results are average of at least three experiments performed in duplicates or triplicates. The Y-axis bars represent the standard error.

more robust activation of apoptotic death pathway. As a result, the overall numbers of cells entering a pathway to death in p53 functional lines are not decreased by PARP inhibitor. BZL101 inhibits glycolysis. Activation of PARP is known to result in the depletion of cellular stores of NAD+ as well as ATP, leading to the profound changes in cell redox status and depletion of energy stores.11,12 We have examined if treatment with BZL101 induces changes in levels of NAD+/NADH and ATP. Figure 5A shows that NAD+/NADH levels are reduced very quickly and significantly in SKBr3 cells, and that after the initial reduction by 60% within the first hour, the levels do not recover during the course of treatment. Addition of PARP inhibitor to the SKBr3 cells prior to BZL101 lead to a temporary delay in the reduction in NAD+/NADH levels. After 8 hours of incubation, NAD+/NADH levels were low in cells treated with BZL101 with or without PARP inhibitor (Fig. 5A). Remarkably, there was no significant or sustained change in levels of NAD+/NADH in MCF10A cells (Fig. 5A), consistent with the observed lack of PARP activation (Fig. 4A). Addition of PARP inhibitor had no effect on NAD levels in MCF10A cells (not shown). We conclude that activation of PARP by BZL101 induced DNA damage in breast cancer cells leads to a dramatic drop in the levels of nicotinamide dinucleotides, and this could have profound effects on the redox status of cells as well as their ability to sustain energy stores. Indeed, analysis of ATP levels showed a sharp depletion of ATP in SKBr3 cells already within the first four hours of BZL101 582

treatment (Fig. 5B). Addition of PARP inhibitor 3-AB delayed the energy collapse but did not prevent it. ATP levels are known to be rapidly depleted during necrotic death, but the decrease is much more gradual during apoptosis because the execution of apoptotic program requires energy.13,14 The effect of PARP inhibition on ATP levels in BZL101 treated cells is consistent with our observation that necrotic death is suppressed in SKBr3 cells in presence of DPQ (Fig. 4C) but apoptotic death is augmented (Fig. 4D). We have also examined the relative abundance of ATP and ADP in BZL101 treated SKBr3 cells, and have observed that the ratio was changed from approximately 10 fold excess of ATP in untreated cells to less than one fold in BZL101 treated cells after as little as four hours (not shown). The inversion of the ATP/ADP ratio confirms the mostly necrotic nature of cell death induced by BZL101.15 In contrast, ATP levels in MCF10A cells were not affected by treatment with BZL101, and even show a slight temporary but consistent increase in the presence of BZL101. Co-treatment of MCF10A cells with the PARP inhibitors has no significant effect on ATP levels (not shown). Published results made a casual connection between PARP super-activation, depletion of energy and NAD stores, and inhibition of glycolysis.10 The dramatic effect that BZL101 has on ATP levels in cancer but not normal cells indicates that inhibition of glycolysis in cancer cells could be causative in the observed drop in ATP levels, since cancer cells frequently rely on glycolysis for their energy needs.




We have therefore examined the potential effects of BZL101 on several indicators of the glycolytic activity. First, we have measured levels of the lactate production in cells treated with BZL101. Figure 5C shows that levels of lactate secreted into growth media are significantly diminished in SKBr3 but not MCF10A cells treated with BZL101, indicating inhibition of glycolytic activity in SKBr3 cells. Next, we have measured the activity of two enzymes in the glycolytic pathway, glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and lactate dehydrogenase (LDH). Enzymatic activity of both was significantly reduced in SKBr3 but not MCF10A cells throughout the 24 hours course of treatment (only the 6 hour time point is shown in Fig. 5D). Combined, the results shown in Figure 5 support the hypothesis that treatment with BZL101 strongly inhibits glycolytic activity in cancer cells. The bioenergetic collapse that follows the hyperactivation of PARP could be further deepened by the inhibition of glycolysis. Neither the PARP activation nor the sustained inhibition of glycolysis are observed in non-transformed mammary epithelial cells MCF10A, providing an explanation for the lack of cytotoxicity of BZL101 in these cells. Treatment with BZL101 induces expression of genes involved in oxidative responses, DNA damage and cell death. We have performed expression array analysis of SKBr3 cells treated with BZL101 compared to untreated cells. A moderately high number of transcripts were found to change expression levels more than two-fold after 4 hours of BZL101 treatment (Suppl. Table 1). The most prominent groups according to function could be easily discerned among the upregulated transcripts: oxidative stress responses (GCLM, HMOX1, CBS, TR, ATF3, A20/TNFAIP3, OLK38); NFkB pathway (TNF, A20, SQSTM1, TRAF3, BIRC3, ICAM1, IL8, RELB, CLC, CCL2, CCL11, CXCL1, CXCL16 and more); DNA damage responses (TIPARP, ATF3, GADD45a, TOPOII); cell death (A20, PUMA, TNFRSF21, TNF); xenobiotic response (CYP1A1, CYP1B1, HSP70, CYP27B1). Among downregulated genes, most prominent group included transcripts related to cell cycle regulation, among them transcipition factors Id1, Id2 and Mad3, and a number of mitosis-related proteins (see Supplementary Table 1 for the full list of transcripts whose levels are affected by BZL101). Changes in expression levels of a number of these genes were verified by quantitative RT-PCR (Suppl. Data, Fig. 1). These analyses confirmed the veracity of the microarray results for all transcripts that were examined. Changes in transcript levels, either up or downregulated, were more pronounced in SKBr3 cells compared to MCF10A cells with the exception of ATF3 (Suppl. Fig. 1). For example, while DNA damage response genes GADD45A and CHOP were upregulated approximately 4 fold in SKBr3 cells, the increase in MCF10A was 2.4 fold for GADD45A and CHOP levels were not increased. Similarly, downregulation of Id1 and piroxiredoxin III was significant in SKBr3 but practically absent in MCF10A (Suppl. Fig. 1). Decrease in levels of PRXIII could contribute to cell death induced by oxidative DNA damage.16 Figure 6 shows the results of the Western blot analysis of the expression levels of some of the protein products for which corresponding RNA levels were found to be affected by exposure to BZL101. We found that though expression of transcription factor ATF3 is induced in both SKBr3 and MCF10A cells, the induction is transient in www.landesbioscience.com

MCF10A, while high levels of ATF3 are sustained in SKBr3 cells throughout the course of treatment. Oxidative stress response proteins cystathionine-beta-synthase (CBS), thioredoxin reductase (TR), glutamate cysteine ligase modifier subunit GCLM (all targets for transcriptional activation by Nrf2) are elevated in SKBr3 cells but not in MCF10A. These results document that BZL101 induces persistent oxidative stress responses in breast cancer cells. Western blot results confirmed that levels of A20 protein are increased during BZL101 treatment in agreement with the microarray and RT-PCR results. Interestingly, expression of A20 is upregulated in MCF10A cells only transiently, whereas higher levels of A20 are maintained in SKBr3 through the course of treatment. A20 is a de-ubiquitinase that terminates TNF-induced NFkB activation and inhibits apoptotic cell death (reviewed in ref. 7). However, A20 enhances necrotic cell death induced by oxidative stress.17 Sustained increase in expression of A20 in SKBr3 cells treated by BZL101 could therefore contribute to necrotic death. We have also examined expression of the transcription factor inhibitor of differentiation 1(Id1), focusing on it in particular because it has been shown to be a transcriptional target for downregulation by ATF3.18 In SKBr3 cells BZL101 induced significant suppression of Id1 (Fig. 6), while in untreated MCF10A cells the basal levels of Id1 were barely above limits of detection by Western blot, and made it difficult to discern any changes in expression levels (Fig. 6). Id family of helix-loop-helix proteins, and Id1 in particular, are known to promote cell survival, proliferation, invasion and angiogenesis.19 Inhibition of Id1 expression could therefore also contribute to cell death of SKBr3 cells.

Discussion The herbal extract BZL101 has completed phase I clinical trial with good indications of safety and efficacy.3 It is very important to understand the molecular basis of the BZL101 activity and the favorable toxicity profile observed in the clinical trial. This study was undertaken to address these questions. It is well documented that cancer cells in general produce more ROS than do normal cells.20 At least in part, this increase could be attributed to the mutations in the genes encoding components of the mitochondrial electron transport chain (reviewed in ref. 21). Inefficient electron transport leads to the capture of “loose” electrons in mitochondria by oxygen, leading to the formation of superoxide that is rapidly dismuted to yield hydrogen peroxide (H2O2). There could be presumably other genetic defects in transformed cells that will contribute to the intrinsically elevated levels of ROS. Increased levels of ROS are thought to be exploited by cancer cells to increase their proliferation rate (reviewed in ref. 22). However, the higher basal level of endogenous ROS could make tumor cells more susceptible to drugs that increase ROS levels further, to a lethal threshold (reviewed in ref. 23). Normal cells, in addition to lower ROS levels, could be better equipped to deal with oxidative stress because they better regulate expression of enzymes that detoxify oxidants. This has been shown to be a likely mechanism for selective cancer cell killing by β-phenylethyl isothiocyanate (PEITC), a constituent of cruciferous vegetables that has a chemopreventive and cancer cell selective cytotoxic activity.24 This report describes an activity derived from another botanical source, Scutellaria barbata, a Chinese medicinal herb used to treat


583


cancers. The S. barbata extract, or BZL101, selectively kills cancer cells but is not cytotoxic towards normal cells.3 We show that the selectivity of BZL101 is based on its ability to induce oxidative stress and extensive DNA damage in cancer cells but not in normal cells. The cytotoxicity of BZL101 could be entirely prevented by preloading cells with ROS scavengers, demonstrating that killing indeed depends on the increased ROS production in cancer cells treated with BZL101. The source of ROS induced by BZL101 remains to be determined. However, we have found that inhibition of one of the major cellular sources of ROS, flavoenzyme NADPH oxidase,25 by diphenylene iodonium did not rescue cells from the BZL101 induced cell death, increasing the likelihood of the mitochondrial source for the BZL101 induced ROS. The mode of killing by BZL101 does not conform to a classical apoptotic model, as it does not involve release of apoptogenic factors from mitochondria, shows only a limited activation of caspases and limited chromosomal DNA fragmentation,3 that does not reflect the actual extent of cell death taking place. We described here that BZL101 induces severe DNA damage in cancer cells that is repaired inefficiently, as opposed to non-transformed cells where BZL101 induces a mild oxidative stress and much less extensive DNA damage that is repaired quickly. The effects that BZL101 has on breast cancer cells are reminiscent of the cellular response to treatment with a high dose of oxidative reagents such as hydrogen peroxide: ever a brief exposure to high concentrations of H2O2 induces necrotic cell death without caspase activation or DNA fragmentation. This mode of death involves activation of PARP, is referred to as “regulated necrosis” and is observed in response to severe oxidative, nitorosative or alkylating DNA damage. This paradigm appears to be applicable to our observations of the differential effects that BZL101 exerts on normal versus cancer cells. Immortalized mammary epithelium cells respond to BZL101 by a low level induction of ROS, limited and efficiently repaired DNA damage, and continuing proliferation. Breast cancer cells respond to the same dose of BZL101 by inducing much higher levels of ROS, extensive DNA damage, PARP activation and necrosis. We presented evidence that the hyperactivation of PARP in breast cancer cells treated with BZL101 is largely responsible for the necrotic cell death induced by BZL101. PARP-1 acts as a DNA damage sensor and is activated by DNA breaks to cleave NAD+ into nicotinamide and ADP-ribose to synthesize long branching poly(ADP-ribose) polymers (PAR) covalently attached to nuclear acceptor proteins (reviewed in ref. 11). Activation of PARP-1 by mild DNA damage facilitates DNA repair and cell survival. However, severe DNA damage leads to hyperactivation of PARP, whereas instead of activating DNA repair PARP triggers the necrotic death pathway.10,26,27 Our results show persistent DNA damage in SKBr3 cells even hours after exposure to BZL101 (Fig. 1). Normally oxidative DNA lesions are removed very quickly. Apparently, hyperactivation of PARP by BZL101 in breast cancer cells has an inhibitory effect on DNA damage repair. Moreover, it leads to a cascade of events starting with depletion of NAD, followed by inhibition of glycolysis, loss of ATP and finally mostly necrotic death. None of this was observed in non-transformed cells. Normal cells rely on oxidative phosphorylation for their energy needs and can maintain their ATP production even when cytosolic NAD is depleted by PARP during genotoxic stress. Oxidative 584

Figure 6. BZL101 modulates expression of genes responsive to oxidative damage. Western blot analysis of the indicated proteins in SKBr3 and MCF10A cells treated with BZL101 for different times.

hosphorylation utilizes mitochondrial NAD, which is not depleted p by PARP activation.12,28 As described earlier,10 normal cells might not activate necrotic death program even when faced with fairly extensive DNA damage. In cancer cells, however, depletion of cytosolic NAD by hyperactivated PARP triggers the execution of necrotic death program, due to inhibition of glycolysis and following energetic collapse. All these results strongly indicate that selective death of cancer cells after treatment with BZL101 could be attributed not only to the selective induction of ROS, but also to the different metabolic consequences of the attempt to repair DNA damage. Out results show that inhibition of PARP in BZL101 treated cells delays depletion of ATP (Fig. 5B). We suggest that this delay is relevant to the observed inhibition of necrotic death that usually occurs as a result of the rapid loss of ATP.15 Presumably, higher levels of ATP when PARP is inhibited allow the observed activation of the apoptotic cell death (Fig. 4D). We suggest that PARP is instrumental in necrotic death induced by BZL101, but other factors might determine the choice between survival and death once PARP is inhibited. Tumor suppressor p53 appears to be just such a factor, determining the fate of cancer cells treated with BZL101 while PARP is inhibited. We have demonstrated that both p53 wild type and mutant cell lines succumb to death after treatment with BZL101. Therefore, the cytotoxicity of BZL101 does not depend on the p53 status. However, p53 influences the outcome when activation




of PARP (causative in necrotic death) is inhibited. Overall, the net result of PARP inhibition during BZL101 treatment in p53 mutant cells is a significant reduction in cell death (Fig. 4C). This increase in survival results from the inhibition of necrotic cell death, while apoptotic death is increased by PARP inhibition, but not sufficient to “compensate” for the pervention of necrosis. On the contrary, overall death of p53 wild type breast cancer cells was not diminished by inhibiting PARP, because the concomitant increase in the number of cells entering apoptotic pathway was very significant. One possible explanation for this is that activated PARP was shown to ADP-ribosylate p53 leading to the inactivation of transcriptional activity.29,30 Therefore, hyperactivated PARP could inhibit the ability of p53 to mediate DNA damage induced apoptosis. Indeed, apoptosis accounts for only a minor fraction of death observed in MDA MB 361 and MCF7 cells treated with BZL101 (Fig. 4C and D). However, when PARP is inhibited, unmodified p53 protein could potentially execute its transcriptional function as a mediator of apoptosis. There is a possible additional (though not contradictory to the one above) explanation for the observed differences in the responses of cells with different p53 status to the inhibition of PARP during treatment with BZL101. Inhibition of PARP obviously prevents necrotic death, and because in p53 deficient cells the apoptotic program is severely affected, the net outcome of PARP inhibition is protection from death. In cells expressing wild type p53, the p53 protein “steps in” when PARP is inhibited. Because inhibition of PARP delays the acute loss of ATP, p53 can efficiently initiate apoptotic death program that depends on availability of ATP. It remains to be explored what elements of p53 governed apoptotic signal transduction are activated during the course of BZL101 treatment when PARP-induced necrosis in inhibited. In any case our results strongly suggest that inhibition of PARP could have different outcomes in tumor cells treated with DNA damaging drugs depending on their p53 status. Results of expression profiling of BZL101 treated cells revealed transcriptional changes that are entirely consistent with our observations of the cellular events triggered by BZL101, and ultimately leading to cell death. In general, the observed differences in the transcriptional responses between cell lines reflect a strong oxidative DNA damage and induction of death-promoting genes in SKBr3 cells but not in MCF10A treated with BZL101. BZL101 induces changes in expression of transcripts involved in oxidative stress responses, DNA damage responses, cell death pathways, and cell cycle, in agreement with our experimental observations. A significant number of transcripts from these functional groups could be regulated by the NFkB pathway. In this context, it is perhaps relevant that activation of PARP has been linked to the transcriptional potential of NFkB.31 The roles of NFkB in regulating cell death are complex, but in the case of BZL101 treated breast cancer cells, it appears that the balance of the consequences of NFkB activation is tilted heavily towards promoting necrosis. In the recent years, a number of compounds were isolated from S. barbata, some of them with interesting biological activities. Among them are flavonone compounds scutellarin, scutellarein, carthamidine, isocarthamidin and wogonin as well as clerodane diterpenoids, and sterol glucosides.32-37 Antiproliferative activity of Scutellaria barabat extracts in cancer cell lines other than breast cancer was also reported.32,34,38,39 Our efforts are now aimed at the purification www.landesbioscience.com

and chemical identification of the activity within BZL101 extract that is responsible for the regulated necrosis induced by BZL101. We have tested the anti-cancer activity of some of the S. barbata flavanones purified in our laboratory, but none have shown promising cytotoxic characteristics at relevant dozes. We believe that the active compound(s), once identified, would represent a novel class of chemotherapeutic drugs with a high selectivity towards cancer cells that is based on inherently different metabolic profiles of cancer versus normal cells.

Methods and Materials Reagents and antibodies. 3,4-dihydro-5-[4-(1-piperidinyl) butoxyl]-1(2H)-isoquinolinone (DPQ), 3-amino-benzamide (3AB), N-acetyl-cystein and diphenylene iodonium were purchased from Sigma. CM-H2DCFDA was purchased from Molecular Probes. Antibodies to ATF3, Nrf2 were from Santa Cruz Biotechnology, to CBS and GCLM from Abnova, to Id1 from Biocheck, to TR from Imgenex, to PAR for western blotting form Becton Dickinson, to PAR for immunofluoresecence from Trevigen, to A20 from EMD Biosciences. Cell culture and treatments. All cell lines were obtained from the ATCC, and propagated according to the instructions provided. Cells were treated with BZL101 at a concentration of 0.5 mg/ml (dry weight per volume) or with water (labeled as “untreated” throughout the paper). The aqueous extract of BZL101 was prepared as described.3 Comet assay. Comet assays were performed using the Comet assay kit from Trevigen according to the manufacturer’s instructions. Briefly, cells were harvested, washed and resuspended with PBS. The cells were combined with molten, low melting point agarose at 37°C and pipetted unto Comet slides. The agarose was allowed to solidify at 4°C for 30–40 min and immersed in cold lysis solution (Trevigen, Inc.) for 30 min at 4°C. The slides were immersed into freshly prepared alkali solution (300 mM NaCl and 1 mM EDTA) for 20 min and subjected to electrophoresis in the same alkaline buffer at 300 mA for 30–40 min. Slides were rinsed in water and then fixed in 70% ethanol for 5 min. After air-drying, the nuclei were stained with Sybr green (Trevigen, Inc.) and viewed under a fluorescence microscope. Percentages of cells with comets were quantified by an observer blinded to the identity of the slides. Analysis of the comet tail momentum was performed using the CometScore program. Analysis of the 8-oxoguanidine and apurininc/apyrimidinic (AP) B/bases in DNA. The 8-oxo kit from EMD Chemicals, Inc., or anti-8-oxoguanine monoclonal antibody (QED Biosciences) were used for staining of the fixed and permeabilized cells according to the manufacturer’s protocols. Quantification of apurinic/apyrimidininc nucleotides in genomic DNA was performed using the DNA damage quantification kit from BioVision Research Products. Flow cytometry. Cells were collected by trypsinization into their culture media, washed with PBS and stained with Annexin V- Alexa Fluor 488 (InVitrogen) and propidium iodide (PI), and analyzed immediately after 15 minute incubation using the CellQuest software on the FACScan (Becton Dickinson). All PI-positive cells were considered to be dead; cells single positive for Annexin V binding were categorized as early apoptotic. Immunofluorescence for PAR. Cells on chamber slides were washed with PBS and fixed with a 1:1 mixture of methanol and


585


acetone on ice for 10 minutes. After a 10 minute permeabilization with PBS containing 0.1% Triton X-100, cells were incubated in PBS with 2% BSA for 1 hour, and then with 0.25 μg/ml of anti-PAR antibody (Trevigen) overnight at 4°C. After three washes in PBS, cells were incubated with the anti-mouse Alexa568 conjugated antibody for 1 hour, washed again, and counterstained with 0.2 μg/ml of DAPI. Images were recorded on a Nikon Eclipse 80i. Metabolic and enzymatic analyses. Combined NAD+ and NADH content, ATP content and the concentration of secreted lactate were quantified using the appropriate assay kits from the Biovision Research Products. Enzymatic activity of LDH and GAPDH was quantified using assays kits from AnaSpec and Ambion respectively. Western blot bnalysis. Whole cell lysates were electrophoresed on SDS-PAGE and transferred to nitrocellulose membranes. Membranes were blotted with antibodies at recommended concentrations overnight at 4°C and the bound primary antibodies were detected using peroxidase-conjugated secondary antibodies. Blots were developed using SuperSignal enhanced chemiluminescence kit (Pierce) and imaged on Kodak Imager ISR2000. Note

Supplementary materials can be found at: www.landesbioscience.com/supplement/FongCBT7-4-Sup.pdf References 1. Boon H, Wong J. Botanical medicine and cancer: a review of the safety and efficacy. Expert opinion on pharmacotherapy 2004; 5:2485-501. 2. Cragg GM, Newman DJ. Plants as a source of anti-cancer agents. Journal of ethnopharmacology 2005; 100:72-9. 3. Rugo H, Shtivelman E, Perez A, Vogel C, Franco S, Tan Chiu E, Melisko M, Tagliaferri M, Cohen I, Shoemaker M, Tran Z, Tripathy D. Phase I trial and antitumor effects of BZL101 for patients with advanced breast cancer. Breast Cancer Res Treat 2007; 105:17-28. 4. Campbell MJ, Hamilton B, Shoemaker M, Tagliaferri M, Cohen I, Tripathy D. Antiproliferative activity of Chinese medicinal herbs on breast cancer cells in vitro. Anticancer Res 2002; 22:3843-52. 5. Shoemaker M, Hamilton B, Dairkee SH, Cohen I, Campbell MJ. In vitro anticancer activity of twelve Chinese medicinal herbs. Phytother Res 2005; 19:649-51. 6. Gatenby RA, Gillies RJ. Why do cancers have high aerobic glycolysis? Nature reviews 2004; 4:891-9. 7. Hayes JD, McMahon M. Molecular basis for the contribution of the antioxidant responsive element to cancer chemoprevention. Cancer letters 2001; 174:103-13. 8. Struthers L, Patel R, Clark J, Thomas S. Direct detection of 8-oxodeoxyguanosine and 8oxoguanine by avidin and its analogues. Analytical biochemistry 1998; 255:20-31. 9. Kim MY, Zhang T, Kraus WL. Poly(ADP-ribosyl)ation by PARP-1: ‘PAR-laying’ NAD+ into a nuclear signal. Genes & development 2005; 19:1951-67. 10. Zong WX, Ditsworth D, Bauer DE, Wang ZQ, Thompson CB. Alkylating DNA damage stimulates a regulated form of necrotic cell death. Genes & development 2004; 18:1272-82. 11. Burkle A. Poly(ADP-ribose). The most elaborate metabolite of NAD+. Febs J 2005; 272:4576-89. 12. Ying W, Alano CC, Garnier P, Swanson RA. NAD+ as a metabolic link between DNA damage and cell death. J Neurosci Res 2005; 79:216-23. 13. Eguchi Y, Shimizu S, Tsujimoto Y. Intracellular ATP levels determine cell death fate by apoptosis or necrosis. Cancer research 1997; 57:1835-40. 14. Leist M, Single B, Castoldi AF, Kuhnle S, Nicotera P. Intracellular adenosine triphosphate (ATP) concentration: a switch in the decision between apoptosis and necrosis. The Journal of experimental medicine 1997; 185:1481-6. 15. Bradbury DA, Simmons TD, Slater KJ, Crouch SP. Measurement of the ADP:ATP ratio in human leukaemic cell lines can be used as an indicator of cell viability, necrosis and apoptosis. Journal of immunological methods 2000; 240:79-92. 16. Mukhopadhyay SS, Leung KS, Hicks MJ, Hastings PJ, Youssoufian H, Plon SE. Defective mitochondrial peroxiredoxin-3 results in sensitivity to oxidative stress in Fanconi anemia. J Cell Biol 2006; 175:225-35. 17. Storz P, Doppler H, Ferran C, Grey ST, Toker A. Functional dichotomy of A20 in apoptotic and necrotic cell death. The Biochemical Journal 2005; 387:47-55. 18. Kang Y, Chen CR, Massague J. A self-enabling TGFbeta response coupled to stress signaling: Smad engages stress response factor ATF3 for Id1 repression in epithelial cells. Molecular Cell 2003; 11:915-26. 19. Perk J, Iavarone A, Benezra R. Id family of helix-loop-helix proteins in cancer. Nature Reviews 2005; 5:603-14.

586

20. Szatrowski TP, Nathan CF. Production of large amounts of hydrogen peroxide by human tumor cells. Cancer Research 1991; 51:794-8. 21. Brandon M, Baldi P, Wallace DC. Mitochondrial mutations in cancer. Oncogene 2006; 25:4647-62. 22. Behrend L, Henderson G, Zwacka RM. Reactive oxygen species in oncogenic transformation. Biochemical Society Transactions 2003; 31:1441-4. 23. Martindale JL, Holbrook NJ. Cellular response to oxidative stress: signaling for suicide and survival. Journal of Cellular Physiology 2002; 192:1-15. 24. Trachootham D, Zhou Y, Zhang H, Demizu Y, Chen Z, Pelicano H, Chiao PJ, Achanta G, Arlinghaus RB, Liu J, Huang P. Selective killing of oncogenically transformed cells through a ROS-mediated mechanism by beta-phenylethyl isothiocyanate. Cancer Cell 2006; 10:241-52. 25. Lambeth JD. NOX enzymes and the biology of reactive oxygen. Nat Rev Immunol 2004; 4:181-9. 26. Szabo C, Dawson VL. Role of poly(ADP-ribose) synthetase in inflammation and ischaemiareperfusion. Trends in Pharmacological Sciences 1998; 19:287-98. 27. Yu SW, Wang H, Poitras MF, Coombs C, Bowers WJ, Federoff HJ, Poirier GG, Dawson TM, Dawson VL. Mediation of poly(ADP-ribose) polymerase-1-dependent cell death by apoptosis-inducing factor. Science (New York, NY) 2002; 297:259-63. 28. Kobayashi K, Neely JR. Control of maximum rates of glycolysis in rat cardiac muscle. Circulation Research 1979; 44:166-75. 29. Mendoza-Alvarez H, Alvarez-Gonzalez R. Regulation of p53 sequence-specific DNAbinding by covalent poly(ADP-ribosyl)ation. The Journal of Biological Chemistry 2001; 276:36425-30. 30. Valenzuela MT, Guerrero R, Nunez MI, Ruiz De Almodovar JM, Sarker M, de Murcia G, Oliver FJ. PARP-1 modifies the effectiveness of p53-mediated DNA damage response. Oncogene 2002; 21:1108-16. 31. Oliver FJ, Menissier-de Murcia J, Nacci C, Decker P, Andriantsitohaina R, Muller S, de la Rubia G, Stoclet JC, de Murcia G. Resistance to endotoxic shock as a consequence of defective NFkappaB activation in poly (ADP-ribose) polymerase-1 deficient mice. The EMBO Journal 1999; 18:4446-54. 32. Chan JY, Tang PM, Hon PM, Au SW, Tsui SK, Waye MM, Kong SK, Mak TC, Fung KP. Pheophorbide a, a major antitumor component purified from Scutellaria barbata, induces apoptosis in human hepatocellular carcinoma cells. Planta Medica 2006; 72:28-33. 33. Dai SJ, Chen M, Liu K, Jiang YT, Shen L. Four New neo-Clerodane Diterpenoid Alkaloids from Scutellaria barbata with Cytotoxic Activities. Chemical & Pharmaceutical Bulletin 2006; 54:869-72. 34. Dai SJ, Wang GF, Chen M, Liu K, Shen L. Five new neo-clerodane diterpenoid alkaloids from Scutellaria barbata with cytotoxic activities. Chemical & Pharmaceutical Bulletin 2007; 55:1218-21. 35. Ducki S, Hadfield JA, Lawrence NJ, Liu CY, McGown AT, Zhang X. Isolation of E-1-(4'Hydroxyphenyl)-but-1-en-3-one from Scutellaria barbata. Planta Medica 1996; 62:185-6. 36. Sato Y, Suzaki S, Nishikawa T, Kihara M, Shibata H, Higuti T. Phytochemical flavones isolated from Scutellaria barbata and antibacterial activity against methicillin-resistant Staphylococcus aureus. Journal of Ethnopharmacology 2000; 72:483-8. 37. Wang WS, Zhou YW, Ye YH, Du N. [Studies on the flavonoids in herb from Scutellaria barbata]. China Journal of Chinese Materia Medica 2004; 29:957-9. 38. Chui CH, Lau FY, Tang JC, Kan KL, Cheng GY, Wong RS, Kok SH, Lai PB, Ho R, Gambari R, Chan AS. Activities of fresh juice of Scutellaria barbata and warmed water extract of Radix Sophorae Tonkinensis on anti-proliferation and apoptosis of human cancer cell lines. International Journal of Molecular Medicine 2005; 16:337-41. 39. Goh D, Lee YH, Ong ES. Inhibitory effects of a chemically standardized extract from Scutellaria barbata in human colon cancer cell lines, LoVo. Journal of Agricultural and Food Chemistry 2005; 53:8197-204.

