By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Due to essential maintenance the subscribe/renew pages will be unavailable on Wednesday 26 October between 02:00- 08:00 BST/ 09:00 – 15:00 SGT/ 21:00- 03:00 EDT. Apologies for the inconvenience.
Trichosanthin (TCS) is the main bioactive component isolated from the root tubers of the Chinese medicinal herb Trichosanthes kirilowii Maximowicz, and has been used as an abortifacient for 1,500 years in China (Jin, 1985). TCS is a 27-kDa basic protein and possesses a single polypeptide chain with 247 amino acid residues. Amino acid sequence analysis showed that TCS cDNA encodes a polypeptide comprised of 289 amino acids, with a 23-residue signal peptide at the N-terminus and a 19-residue leader peptide at the C-terminus, which is cleaved from mature TCS (Collins et al., 1990). Based on the amino acid sequence, Pan et al., created a molecular structural model of TCS using 0.26-nm resolution X-ray diffraction (Pan et al., 1992). Functionally, TCS is a type I ribosome inactivating protein (RIP), with one subunit (A chain); type II RIPs consist of a toxic A chain and a lectin-like subunit (B chain). TCS has a similar mode of action as other RIPs; the N-glycosidase activity depurinates the adenine residue at position 4324 of 28S rRNA, resulting in the inhibition of cellular translation and protein synthesis.
For many years, it was believed that TCS-treated cells underwent cell death simply because TCS inhibits cellular protein synthesis and consequently induces necrosis. However, the current evidence suggests that TCS induces a typical apoptosis process in several cell lines (Zhang et al., 2001; Li et al., 2007a; Wang and Li, 2007). Apoptosis can also be induced by some type II RIPs such as ricin, shiga toxin 1 and mistletoe lectin. The role of the B chain in the apoptotic processes is debated (Hasegawa et al., 2000; Higuchi et al., 2004); of note, site-specific mutations in the A chain inhibit the onset of apoptosis (Langer et al., 1999; Smith et al., 2003). Collectively, these findings strongly suggest that the A chain is involved in the apoptotic effects of RIPs. Now, increasing attention is being paid to the role of the A chain of TCS on the induction of cellular apoptosis.
TCS has been used as a “wonder drug” for ectopic pregnancy, fetal death, hydatidiform mole and malignant hydatidiform mole because of its high toxicity on trophoblasts. Studies over the past 20 years have shown that TCS has a broad spectrum of biological and pharmaceutical activities including antitumor, immune regulatory and antiviral effects (Shaw et al., 2005). TCS was shown to exhibit selective cytotoxicity for specific tumor cell lines, and is more specific for tumor cell lines than normal cell lines, which has drawn researchers' attention to its antitumor activity. Interestingly, TCS was effective in treating chorionic epithelioma, and is also being considered for treatment of other tumors. Table 1 lists some of the studies showing these cytotoxic and selective antitumor effects. In vitro studies showed that TCS was effective in killing human gastric and colorectal cancer cells at different stages of differentiation in a time- and dose-dependent manner (Wang and Jin, 2000a). TCS was also cytotoxic on leukemia and lymphoma cells, however, it only had limited cytotoxic effects on peripheral blood lymphocytes, such as T cells and granulocytes (Kong et al., 1998). Furthermore, studies have shown that TCS can induce apoptosis of cervical adenocarcinoma HeLa cell (He and Li, 2006) and cervical squamous Caski cells (Zhang et al., 2008). In addition, TCS can inhibit the proliferation of breast adenocarcinoma cells in vitro and in vivo (Ding et al., 2008). Although numerous studies have focused on TCS-induced apoptosis of tumor cells, the mechanisms involved are poorly understood. The current review focused on the antitumor activities of TCS and, combined with our recent findings, we have tried to provide insights into the possible mechanisms by which TCS induces tumor cell apoptosis. The effects of TCS with those of other RIPs have also been compared in our review.
Table 1. Cytotoxicity and apoptosis induced by TCS in many tumors cell lines
Reactive oxygen species (ROS) appear to intersect with a multitude of intracellular pathways involved in cell proliferation, differentiation and apoptosis (Thannickal and Fanburg, 2000). ROS also play a regulatory role in drug-induced apoptosis of tumor cells (Nakazato et al., 2005; Nie et al., 2009), and participate in apoptosis induced by RIPs such as abrin and Korean mistletoe lectin (Shih et al., 2001; Kim et al., 2004). Zhang et al., reported TCS induced DNA condensation and fragmentation of human choriocarcinoma JAR cells by two-photon laser scanning microscopy and their appearance at sub-G0/G1 by flow cytometry analysis, suggesting TCS induced apoptosis in JAR cells. Accompanied with apoptosis, TCS increased ROS production in a time- and dose-dependent manner in JAR cells. Using ROS scavengers and the spin-trapping technique, it was confirmed that O·, OH· and H2O2, oxygen-centered species, were all involved in TCS-induced ROS formation in JAR cells (Zhang et al., 2001). Exposure to mutant TCS (Y55G and FYY140–142GSA) resulted in decreased ROS production and was less cytotoxic on JAR cells than wild-type TCS (Zhang et al., 2002), which suggested that ROS are involved in TCS-induced apoptosis in JAR cell. In nonphagocytic cells, ROS are produced in mitochondria, the plasma membrane and the endoplasmic reticulum, and production of ROS is regulated by the corresponding oxidative enzymes (Thannickal and Fanburg, 2000). After 10 min of TCS treatment, ROS were released rapidly and were distributed throughout the cytoplasm of JAR cells. Meanwhile, it has been demonstrated that TCS can interact with and enter JAR cells via low-density-lipoprotein receptor-related protein (Chan et al., 2000; Zhang et al., 2001). Therefore, it seems that the plasma membrane might be one site at which TCS induces ROS generation, but the target and mechanism remain to be elucidated. Studies carried out in our laboratory showed similar results that the production of ROS was involved in TCS-induced apoptosis in HeLa cells, which were observed by scanning electron microscope and transmission electron microscope, in a time-dependent manner. Compared with the rapid release of ROS in JAR cells, ROS fluorescence in HeLa cells increased significantly after 12 hours of TCS treatment (unpublished data).
It is worth mentioning that the production of ROS in both cell lines were associated with the intracellular calcium (Ca2+) concentration, including an influx of extracellular Ca2+ and a release of intracellular Ca2+. A Ca2+-chelating agent greatly decreased ROS production and inhibited apoptosis, indicating that Ca2+-dependent ROS generation is involved in TCS-induced apoptosis in HeLa and JAR cells. However, Ca2+-dependent ROS generation is not essential for TCS-induced apoptosis. In the TCS-treated K562 cells, the Ca2+ concentration increased instantly but ROS production increased slowly; a Ca2+-chelating agent reversed the increase of Ca2+ concentration, but failed to inhibit TCS-induced apoptosis, suggesting that Ca2+ signaling is not directly involved in TCS-induced apoptosis in K562 cells (Li et al., 2007b). Although there is close relationship between ROS production and the increase in intracellular Ca2+ concentration during apoptosis, ROS generation is independent of the Ca2+ concentration increase in some experimental models (Tan et al., 1998). For example, it was reported that pretreating K562 cells with ROS scavengers decreased ROS generation, but did not affect TCS-induced apoptosis, indicating that ROS production was not involved in TCS-induced apoptosis in K562 cells. Collectively, these findings suggest that there was no direct association between ROS generation and the Ca2+ concentration increase in K562 cells. Therefore, it appears that the role of Ca2+-dependent ROS generation in TCS-induced tumor cells apoptosis is a cell line-specific characteristic.
Induction of Apoptosis by Cyclic Adenosine Monophosphate (cAMP) Signaling Pathways
Guanine nucleotide-binding protein (G protein) plays an important role in the transmission of extracellular signals into the cells and intracellular signaling transduction. However, little is known about the role of G proteins in RIP-induced cytotoxicity. A [35S]GTPγS binding assay revealed that TCS greatly activated G proteins on the cell membrane of JAR and K562 cells, but not in amniotic Wish cells (Wu et al., 1999), indicating that TCS or TCS segments could activate G proteins and regulate adenyl cyclase (AC) activities. AC is activated by stimulating G proteins (Gs) and inhibited by inhibiting G proteins (Gi); both of which could further increase or decrease the production of cyclic adenosine monophosphate (cAMP). Studies in our laboratory showed that both TCS and high concentrations of an AC inhibitor (SQ22563, 500 μM) could reduce the intracellular cAMP level in HeLa cells in a dose- and time-dependent manner (Wang et al., 2007). We assumed that Gi activation was involved in TCS-induced apoptosis. We also found that a sustained elevated intracellular concentration of Ca2+ was required for the inhibitory effect of TCS on cAMP production. Blocking the increase in the intracellular Ca2+ concentration significantly reversed the TCS-induced reduction in cAMP levels. Traditionally, cAMP mediates intracellular biological reactions in a protein kinase A (PKA)-dependent manner. However, recent studies have shown that cAMP can activate exocytosis, extracellular signal-regulated kinase (ERK) and H,K-ATPase in a PKA-independent manner (Hecquet et al., 2002; Seino and Shibasaki, 2005). It was reported that TCS inhibits protein kinase C (PKC) activity in HeLa and K562 cells, and the activation of PKC by PKC agonists inhibits TCS-induced apoptosis (Li et al., 2007b; Wang et al., 2009). Although activated PKA was detected, it was not involved in TCS-induced HeLa cell apoptosis (Wang et al., 2009). Studies of other RIPs showed that inactivation of PKA and PKC participated in mistletoe lectin II-induced apoptosis in HL60 cells, while activations of PKA and PKC attenuated apoptosis (Pae et al., 2000). Interestingly, exposure to a PKC activator not only weakened TCS-induced apoptosis but also reversed the TCS-induced reduction in cAMP levels; while a PKC inhibitor enhanced TCS-induced reduction of cAMP levels. In contrast, a PKA activator (FSK) and PKA inhibitor (H89) failed to affect TCS-induced reduction in cAMP (unpublished data). It was reported that the PKC signaling pathway is involved in regulating the levels of cAMP (Teitelbaum, 1993). However, further studies are needed to reveal the true mechanism involved in TCS-induced antitumor activities. At present, it can only be concluded that TCS elicited antitumor activities by inhibiting the cAMP/PKC pathway rather than the PKA pathway.
Apoptosis Induction Via the Caspase Family and the Mitochondrial Pathway
Caspases are a group of cysteine proteases critical for apoptosis in eukaryotic cells. Of these, caspase-3 has been identified as the key terminal executor. The activation of caspase-3 was detected in many TCS-treated tumor cell lines (Zhang et al., 2001; Li et al., 2007a; Li et al., 2007b; Wang and Li, 2007), while the other factors mentioned in this review, including PKC, ROS and Bcl-2, ultimately activate caspase-3 to promote cellular apoptosis (Zhang et al., 2001; Li et al., 2007b; Wang et al., 2007). A specific inhibitor of caspase-3 (Ac-DEVD-CHO) significantly inhibited TCS-induced apoptosis, indicating that TCS-induced antitumor activity was dependent on caspase-3. Later studies revealed that TCS activates caspase-8, caspase-9 and caspase-4 (He and Li, 2006; Li et al., 2007a), to cleave and activate the downstream caspases, resulting in a broader spectrum of cellular targets and cellular apoptosis. Li et al., (2007a) reported that when TCS induced the apoptosis of HL-60 cells, the mitochondrial membrane potential (MMP) was lost, which was followed by the release of cytochrome c and Smac from the mitochondria into the cytosol. The activation of caspase-9 activates caspase-3 and DNA fragmentation. Caspase-9 is the main executor of the mitochondrial apoptosis pathway, and inhibition of caspase-9 with z-LEHD-FMK blocked the cellular apoptosis processes, suggesting that the caspase-9-mediated mitochondrial pathway is involved in TCS-induced apoptosis (Li et al., 2007a). Caspase-8 is the initiator caspase upstream of the caspase signaling cascade and the initiator of the death receptor pathway. When the death receptor binds to its specific ligand in cells, various apoptotic proteins are recruited, resulting in caspase-8 activation and ultimately apoptosis. Studies have shown that TCS did not affect the level of Fas or Fas ligands; even the anti-Fas monoclonal antibody could not inhibit TCS-induced apoptosis, indicating that Fas/FasL is not involved in TCS-induced apoptosis (Li et al., 2007a). That study also detected the activation of caspase-8, which was not only inhibited by its inhibitor z-IETD-FMK, but also blocked by inhibitors of caspase-9 and caspase-4. These data suggest that caspase-8 acts downstream of caspase-9 and caspase-4, rather than the death receptor pathway. However, in some cases, caspase-4 is required for endoplasmic reticulum stress-induced apoptosis (Momoi, 2004). Li et al., (2007a) also found that caspase-4 was activated and its inhibitor z-YVAD-FMK inhibited TCS-induced HL-60 cell apoptosis. Interestingly, caspase-3 activation and apoptosis induced by TCS were partially inhibited by the inhibition of either caspase-9 or caspase-4, and was almost completely inhibited by administration of both inhibitors, suggesting that the caspase-9-mediated mitochondrial pathway and the caspase-4-mediated endoplasmic reticulum pathway are both involved in TCS-induced HL-60 cell apoptosis (Li et al., 2007a). Endoplasmic reticulum stress-induced apoptosis was also investigated with other RIPs such as Shiga toxin I and saporin (Lee et al., 2008). However, because of differences in cell species used, the apoptotic signaling pathways of different tumor cells might differ. Nevertheless, the caspase family and mitochondrial pathways seem to play a primary role in TCS-induced antitumor apoptotic pathways. Clearly, the mechanisms involved in the endoplasmic reticulum-triggered apoptotic pathway need further investigation.
Apoptosis Induction by the Regulation of Apoptosis-Associated Genes
The induction of cell apoptosis is considered to be one of the most effective strategies involved in antitumor therapy. The activation and balance of apoptosis-associated genes determine the initiation of apoptosis and the fate of cells. Bcl-2, which is localized between the inner and outer mitochondrial membranes, is a key regulator of apoptotic processes. Bcl-2 is an antiapoptotic factor, and its reduction plays an important role in tumor cell apoptosis induced by ricin, arostin and other RIPs (Chiu et al., 2001). Recent studies showed that HSV-1 infection in Vero cells upregulated the expression of Bcl-2, while TCS effectively suppressed this elevation and induced more apoptotic cells in HSV-1 infected cells (Huang et al., 2006). Bax, another member of the Bcl-2 family, forms a homodimer to induce apoptosis, or forms a heterodimer with Bcl-2 to inhibit apoptosis (Diaz et al., 1997). Therefore, the ratio of Bcl-2/Bax determines cell fate during times of cellular stress. The higher the ratio, the fewer the apoptotic cells. Western blotting studies revealed that TCS decreased the Bcl-2 protein level and increased the Bax protein level, resulting in a decreased ratio of Bcl-2/Bax in HeLa cells and ultimately apoptosis (Huang et al., 2007). Studies carried out in our laboratory also revealed significant downregulation of Bcl-2 expression in TCS-induced HeLa cells, and the possible mechanism of Bcl-2 downregulation in TCS-induced apoptosis has been discussed (Wang et al., 2007). The Bcl-2 gene contains a cAMP-response element (CRE, TGACGTCA), which controls Bcl-2 levels via the activation of a transcription factor, CRE-binding protein (CREB) (Wilson et al., 1996). cAMP agonists such as CPT and FSK increased the phosphorylation of CREB, the expression of Bcl-2, and fully attenuated caspase-3 activation and apoptosis in TCS-treated HeLa cells. By contrast, using antisense oligonucleotide to knockdown the CREB gene expression and block the binding of CREB to the Bcl-2 CRE, CPT and TCS failed to increase Bcl-2 protein expression. These results suggest that TCS-suppressed CREB activation plays a crucial role in the down-regulation of Bcl-2 expression. It is worth noting that, by using a differential display RT-PCR technique, a novel apoptosis-related gene was found to be expressed at high levels in apoptotic leukemia U973 cells with typical DNA ladder exposed to TCS (Li et al., 2000). This gene was named Gene Related to TCS-induced Apoptosis (GRETA), and the role of GRETA in the antitumor effects of TCS need to be further investigated. Apoptosis is a complicated process involving altered expression of multiple genes. Studies have revealed that recombinant TCS-induced apoptosis in the human gastric adenocarcinoma cell line MCG803 was accompanied by up-regulation of p21 and down-regulation of p53 at the transcriptional and expression levels (Xu et al., 2009). It is known that the apoptotic and antiapoptotic genes for different cell lines are different; therefore, the mechanism of TCS-induced apoptosis might be identical in various cell lines. Therefore, further studies are needed to verify changes in gene expression induced by TCS in apoptosis.
Apoptosis Induction by Regulation of the Cytoskeleton
The cytoskeleton is an important cellular component that maintains normal cell morphology and function. In apoptotic cells, the cytoskeleton undergoes breakdown and reformation, which results in changes in cell morphology and functional disabilities. Viscum album agglutinin-I (VAA-I), a type II RIP, has been used extensively in clinical fields as an anticancer adjuvant (Olsnes et al., 1982). It was reported that VAA-I could induce cytoskeleton breakdown (Lavastre et al., 2007). However, to date, very few studies have reported cytoskeleton changes in TCS-induced apoptosis. In our investigation, TCS was found to induce specific changes in cytoskeleton configuration in apoptotic HeLa cells. We also found depolymerized microfilaments (MF), which accumulated in the apoptotic body and coarsened cytoplasm; the microtubules (MT) were arranged in a ring beneath the plasma membrane, surrounding the fragmented nucleus and apoptotic bodies (Wang and Li, 2007). In physiological conditions, MF participates in the maintenance of plasma membrane integrity. During apoptosis, MF is depolymerized, and MT plays an important role in maintaining the membrane integrity. The ring-shaped MT structure induced by TCS was observed during the late stages of apoptosis, which might represent the shift from apoptosis to necrosis. Ca2+ is also involved in cytoskeleton arrangement. The presence of EDTA-AM disrupted the TCS-induced apoptotic ring-shaped MT structure, whereas no distinct effect was observed on the disorder of the MF network (unpublished data).The TCS-induced intracellular Ca2+ elevation is involved in the transduction of apoptosis in HeLa cells and also plays an important role in maintaining the integrity of the apoptotic cell membrane. Interestingly, TCS decreased the gene expression of cytoskeletal isoforms, but failed to affect cytoskeletal protein expression. The possible mechanism is that during TCS-induced apoptosis, MF rearrangement induced changes in gene expression, although the cytoskeleton was not affected at the level of translation and only resulted in the rearrangement of the cytoskeleton within the cells (Safavi-Abbasi et al., 2001).
Unknown Pathways Involved in TCS-Induced Apoptosis
TCS possesses N-glycosidase activity to inhibit protein synthesis, and also can induce cellular apoptosis. However, the possible correlation between these two events is not entirely clear (Fig. 1). Relevant studies of other RIPs showed that RIP-induced apoptosis is either dependent or independent of their N-glycosidase activity. Site-specific mutation of the A chain of type II RIPs, such as Shiga, toxin 1 and mistletoe lectin, significantly reduced the inhibition of protein synthesis and cell apoptosis, suggesting that N-glycosidase activity was involved in apoptosis induced by those RIPs (Langer et al., 1999; Smith et al., 2003). However, studies by Shih's group (2001) showed that abrin could directly interact with mitochondrial antioxidant protein-1 (APO-1) and decrease the antioxidant function of AOP-1, resulting in the production of ROS, the loss of MMP, the release of cytochrome c into the cytosol and finally the induction of apoptosis. They also revealed that the loss of MMP occurred before the inhibition of protein synthesis in abrin-induced apoptosis of Jurkat cells (Narayanan et al., 2004). These data suggest that the induction of apoptosis by abrin was independent of its N-glycosidase activity. In TCS-induced apoptosis of HL-60 cells, the loss of MMP was detected as early as 1.5 hours after treatment. However, further studies are needed to determine whether the TCS-induced loss of MMP occurs before the inhibition of the protein synthesis. In TCS-induced cellular apoptosis, it needs to be confirmed whether the apoptotic agents directly cause apoptosis, or requires N-glycosidase to initiate the apoptotic processes. Therefore, the relationship between TCS-induced apoptosis and its N-glycosidase is unclear.
Iordanov et al., (Iordanov et al., 1997) found that RIPs such as ricin and anisomycin could activate the SAPK/JNK1 signaling pathway and evoke the ribotoxic stress response. The activation of SAPK/JNK1 by RIPs was not only associated with the inhibition of protein synthesis, but was also affected by 28S rRNA signaling. It has been demonstrated that the ribotoxic stress response was mediated by N-glycosidase activity, resulting in the activation of SAPK/JKN and p38 MAPK, finally leading to apoptosis (Narayanan et al., 2005). However, not all of the RIPs with N-glycosidase activity elicit the ribotoxic stress response (Narayanan et al., 2005). It is unclear whether ribotoxic stress is elicited by TCS. It was reported that TCS inhibited PKC/ERK1/2 and activated p38 mitogen-activated protein kinase (MARK), resulting in the proliferative inhibition of HeLa cells (Wang et al., 2009). This seems to indicate that TCS is not associated with the ribotoxic stress response. Obviously, further studies are needed to investigate the effects and mechanisms of SAPK/JNK in TCS-induced apoptosis.
As shown in Table 1, TCS is cytotoxic and can induce apoptosis in several tumor cell lines. However, the characteristics of TCS itself limit its activity in some respects. Since TCS is a type I RIP lacking a B chain, which assists the transport of type II RIPs into cells, the IC50 value of TCS is much higher than that of type II RIPs (Narayanan et al., 2005). Researchers have tried to investigate the mechanism of TCS entry into cells and discover the specific receptor involved in this process (Chan et al., 2000; Chan et al., 2002), which revealed the need to lower the TCS dose in treatment. Nevertheless, animal experiments showed that the IC50 value of TCS (0.5 mg/kg) in human gastric cancer SGC-7901 cell in nude mice was much lower than that of the LD50 value (13.4 mg/kg) (Wang and Jin, 2000b), suggesting that TCS is a potential antitumor agent. It is worth noting that TCS can be conjugated to a monoclonal antibody, which enhances the targeting therapy of antitumor drugs and decreases side effects. Thus, it was considered as a potential biological warfare agent (Wang et al., 1991). Compared with the low-molecular-weight RIPs of 8–14 kDa, the molecular weight of TCS is 27 kDa is larger and is associated with immunogenicity, which limits the efficacy of TCS in clinical applications. Researchers have tried to identify the epitopes of TCS-induced antigenicity, and modify the structure in the hope of reducing immunogenicity without lowering its biological activity (Shaw et al., 2005).
Accumulated evidence indicates that oxidative stress, cAMP signaling pathway, mitochondrial and endoplasmic reticulum stress signaling pathways, altered expression of apoptosis-related genes and regulation of cytoskeleton structure are all involved in TCS-induced apoptosis of tumor cells. As shown in Fig. 1, mitochondria, endoplasmic reticulum, ribosome, microfilaments and microtubules play respective and cooperative roles in TCS-induced apoptosis. The caspase family and mitochondrial pathways are believed to play the primary roles in the TCS-induced apoptotic pathways in tumor cells, while Ca2+ is involved in various pathways in TCS-induced apoptosis. Although several mechanisms involved in TCS-induced apoptosis have been hypothesized, but more studies are needed to reveal the true mechanism. It is still unknown whether clathrin is involved in receptor-mediated TCS-induced endocytosis. In TCS-induced tumor cell apoptosis, the site of ROS production remains to be confirmed, and may include the plasma membrane, mitochondria or endoplasmic reticulum (Fig. 1). The possible correlation between the N-glycosidase activity and TCS-induced apoptosis is also worthy of further exploration. However, because of differences between cell lines, tumor cells may possess differential sensitivity to TCS, whose underlying mechanisms vary accordingly. The elucidation of the mechanism and the molecular basis of TCS-induced apoptosis will provide a theoretical basis for the clinical administration of TCS for antitumor therapy.
The authors thank Dr. Yun Xe, Dr. Tian Zhang, and Ms. Dongmei Li for their valuable review and revision of the manuscript.