Microcystins produced by cyanobacteria are potent and specific hepatotoxins; however, the mechanisms of microcystin-induced hepatotoxicity have not been fully elucidated. The induction of free radical formation and mitochondrial alterations are two major events found in microcystin-treated cultured rat hepatocytes. The mitochondrial alterations, i.e. loss of mitochondrial membrane potential and mitochondria permeability transition are now recognized as key steps in apoptosis. The activation of calpain and Ca2+/calmodulin-dependent protein kinase II is believed to be critical in the microcystin-induced apoptotic process.
Microcystins are a family of toxins produced by fresh water cyanobacteria (blue-green algae), primarily from Microcystis aeruginosa, but also from other Microcystis species and other genera, such as Anabaena, Oscillatoria and Nostoc. Francis (1878) was the first to report that toxic algal blooms contribute to the poisoning of cattle in Lake Alexandrina, Australia. Subsequently, numerous reports have demonstrated the adverse health effects of cyanobacterial toxins on wild and domestic animals, and also on humans, including a tragic incident that led to the death of 60 dialysis patients in Caruaru, Brazil in 1996 [2–4]. Microcystins are cyclic heptapeptides, composed of seven amino acids. So far, more than 50 types of microcystins have been identified, and among them microcystin-LR (MLR) is the most commonly encountered (Fig. 1). Using tritiated microcystin, it has been demonstrated that the liver is the prime target organ affected, characterized by fulminant intrahepatic hemorrhage and death of animals. Microcystin cannot penetrate the cell membrane through simple diffusion, but through the bile acid transport system. This mechanism of cell entry could thus explain the cell specificity and organotropism of microcystin. In addition to its potent hepatotoxicity, microcystin also appears to have tumor promotion activity. Epidemiologic studies in China have suggested that microcystins may play an important etiologic role in the higher incidences of primary human hepatocellular carcinoma (HCC).
So far, the exact mechanisms of microcystin-induced hepatotoxicity and tumor promotion activity have not been fully elucidated although changes in liver functions and programmed cell death have been considered as the main events. One of the most extensively studied mechanisms is that microcystins are potent inhibitors of protein phosphatases 1 (PP1) and 2A (PP2A), leading to increased protein phosphorylation [7,11].
Recently, others and our group have reported that microcystins are capable of initiating apoptosis in hepatocytes as evidenced by characteristic apoptotic morphological changes including membrane blebbing, cell shrinkage, externalization of membrane phosphatidylserine and chromatin condensation [12–14]. Several in vivo experiments have also demonstrated that apoptosis occurs in microcystin-induced liver injury. Although it has been implicated that protein phosphorylation and caspases may play a role in microcystin-induced apoptosis, the exact mechanism of microcystin-induced apoptosis remains unclear. Recent studies in our laboratory strongly suggest that oxidative stress and mitochondria play a critical role in microcystin-induced apoptosis. In this review we briefly summarize the recent evidence implicating that oxidative stress and mitochondria are involved in microcystin-induced apoptosis.
2Microcystin induces oxidative stress in hepatocytes
Oxidative stress can be defined as the exposure of molecule, cell or tissue to the excess level of oxidants, particularly to the free radicals, such as superoxide or hydroxyl radicals, which commonly refer to reactive oxygen species (ROS). Oxidative stress may occur either due to the overproduction of ROS or to the decrease of cellular antioxidant levels. Oxidative stress can lead to severe adverse effects on cells and tissues by causing lipid peroxidation and DNA damage. The membrane-active antioxidants vitamin E and silymarin, as well as other antioxidants including glutathione (GSH) and monoethyl ester glutathione have been shown to produce significant protection against animal death caused by microcystin [16,17]. In cultured hepatocytes, antioxidants such as silymarin, dithioerythritol, desferoxamine and N-acetylcysteine (NAC) significantly reduced the release of lactate dehydrogenase (LDH), cell detachment and cytoskeleton disruption in microcystin-treated cells [18–20]. Moreover, our previous studies demonstrated that microcystin increased the levels of fluorescence of dichlorofluorescein (DCF), a fluorochrome that indicates the excessive production of oxygen-derived free radicals and lipid peroxides. This increase occurs within 5 min after exposure of primary cultured hepatocytes to microcystin (Fig. 2). In addition, one of the lipid peroxidation products, malondialdehyde (MDA), was significantly increased in microcystin-treated hepatocytes. Notably, microcystin-induced ROS production apparently preceded the LDH leakage and MDA formation, suggesting a causative role of ROS production in initiation of lipid peroxidation and cytotoxicity. More recently, using electron spin resonance (ESR) spectroscopy and image-guided proton nuclear magnetic resonance (1H-NMR) spectroscopy, Towner et al. was able to show the lipid-derived radicals in rat livers after in vivo exposure to MLR.
Microcystin also enhances oxidative stress by altering antioxidant levels. It is known that GSH is the major intracellular antioxidant with multiple biological functions. GSH can participate as an antioxidant either indirectly or directly. In the first case, GSH serves as a substrate for GSH peroxidase to reduce hydrogen peroxide. In addition, GSH acts directly as a free radical scavenger to react with OH, HOCl, peroxynitrite, RO and carbon-centered radicals. Therefore, depletion of GSH often accompanies ROS generation. In the second case, GSH can conjugate with xenobiotics and plays an important role in the metabolic pathway leading to detoxification. Cellular GSH is also known to be important for the regulation of cytoskeletal organization. Perturbing of the cellular redox status by depleting intracellular GSH has been shown to disrupt the microfilament structures in human fibroblasts. Cellular GSH was depleted rapidly within 30 min after exposure to microcystin. Pretreatment of mice with GSH protected against microcystin lethality. It has been shown later that microcystin could conjugate with GSH and cysteine both in cell-free systems and under in vivo conditions through the Mdha moiety of microcystins. This conjugation reaction may occur under enzymatic activity by glutathione S-transferase (GST). In line with the above observation we also found that there is an earlier decrease of intracellular GSH levels after exposure to microcystin in cultured rat hepatocytes. Surprisingly, we further found that the GSH level increased dramatically at a later time point, due probably to the self-protective mechanisms of the cell. Pretreatment with NAC, a GSH precursor, significantly enhanced the intracellular GSH levels and decreased microcystin-induced cytotoxicity, as well as cytoskeleton changes. In contrast, buthionine-sulfoximine (BSO), a specific GSH synthesis inhibitor, increased the cell susceptibility to microcystin-induced cytotoxicity and cytoskeleton changes. Therefore, all the above evidence suggests that GSH plays a crucial role in the detoxification of cyanobacterial toxins.
3Mitochondrial membrane depolarization and permeability transition
Mitochondria are known to be vulnerable targets of various toxins because of their important role in maintaining cellular structures and functions. The functional alterations of mitochondria are usually manifested by the changes of mitochondrial membrane potential (MMP). The formation of MMP is due to the asymmetric distribution of protons and other ions on both sides of the inner mitochondrial membrane, giving rise to a chemical and electric gradient. MMP is essential for mitochondria function such as oxidative phosphorylation and ATP synthesis. The inner mitochondrial membrane is negatively charged. Thus MMP can be monitored using some cationic fluorescence probes such as rhodamine-123 (Rh123) or tetramethylrhodamine methyl ester (TMRM). We have examined the MMP alterations in rat hepatocytes exposed to microcystin using TMRM as an indicator of MMP. Microcystin has been shown to rapidly decrease the MMP in hepatocytes before the LDH leakage, suggesting that the membrane potential depolarization is an early event of microcystin-induced hepatotoxicity. In addition to the alterations of MMP, microcystin also induces the mitochondria permeability transition (MPT) changes. MPT, also called the mitochondrial megachannel, represents an abrupt increase of permeability of the mitochondrial inner membrane to solutes of molecular mass less than 1500 Da. Although the exact molecular composition of the pore complex remains elusive, it is thought to involve proteins from the outer membrane (voltage-dependent anion channel, VDAC), the inner membrane (the adenine nucleotide translocator, ANT) and the matrix (cyclophilin D), which interact to form the MPT pore complex. The MPT pore participates in the regulation of matrix Ca2+, pH, mitochondrial membrane potential (MMP) and volume. Opening of the MPT pore can provoke irreversible dissipation of MMP, and MPT inhibitors such as bongkrekic acid (a ligand of the ANT), cyclosporin A (a ligand of cyclophilin D) and Koenig's polyanion (a VDAC inhibitor) could prevent the loss of MMP. Using laser confocal microscopy we have demonstrated that the microcystin induced MPT in cultured hepatocytes (Fig. 3). Furthermore, this MPT can be inhibited by antioxidants and a mitochondrial Ca2+ uptake inhibitor. Evidently, the microcystin-induced oxidative stress is thought trigger of MPT and MMP changes in hepatocytes.
4Role of mitochondria in microcystin-induced hepatocyte apoptosis
We have previously reported that microcystin was able to induce rapid apoptosis in primary cultured rat hepatocytes based on the following convincing evidence: (i) cell shrinkage, cell membrane blebbing and nuclear condensation in microcystin-treated cells; (ii) externalization of phosphatidylserine in microcystin-treated cells determined by annexin-V; and (iii) time-dependent increase of TUNEL positive cells. The typical apoptotic changes in microcystin-treated hepatocytes such as cell membrane blebbings, externalization of phosphatidylserine and positive TUNEL staining all occurred within 50 min. Another study demonstrated that microcystin could induce hepatocyte apoptosis within 2 min under a relatively higher concentration.
Among the apoptotic pathways, mitochondria have been recognized as the central executioner by the release of apoptotic factors such as cytochrome c, apoptosis inducing factor (AIF) and Smac/DIABLO. Once cytochrome c is released, it will activate caspase-9 and the subsequent caspase-3. The mitochondrial apoptotic pathway is also regulated by the Bcl-2 family proteins. These proteins consist of both anti-apoptosis (Bcl-2, Bcl-xL, Mcl-1, Bcl-w, A1, Boo/Diva) and pro-apoptosis members (Bax, Bak, Bid, Bad, Bik, Bim, Blk, Noxa, PUMA). The anti-apoptosis family proteins inhibit cytochrome c release and apoptosis in certain apoptotic models. On the other hand, the pro-apoptotic family proteins enhance the release of cytochrome c.
The critical role of mitochondria in microcystin-induced apoptosis is demonstrated by: (i) microcystin-induced onset of MPT and loss of MMP, (ii) microcystin-induced release of mitochondrial cytochrome c, and (iii) protective effects of cyclosporin A, a specific MPT inhibitor [14,32]. A time-course study clearly indicates that the ROS formation occurred before the onset of MPT. The antioxidants and the mitochondrial electron transport chain (ETC) inhibitors prevent the ROS formation, and also subsequently inhibited the onset of MPT and cell death. These findings suggest that mitochondria are a major source of ROS, and the disruption of ETC may serve as the initial trigger of MPT after the exposure of hepatocytes to microcystin. Recently it has been demonstrated that microcystin could bind to the ATP synthase, an important component in the mitochondria ETC. Whether this binding will cause the disruption of ETC and subsequent apoptosis needs to be further studied. On the other hand, in the stress-induced apoptotic pathway, activation of pro-apoptotic Bcl-2 family proteins such as Bax is known to play an important role. It will be very interesting to study whether microcystin will activate Bax.
Interestingly, others and our results demonstrate that microcystin can induce nearly all the apoptotic changes in hepatocytes in less than 1 h. Such a rapid process is different from other traditional apoptotic inducers such as TNF-α or TGF-β1, which usually stimulate apoptosis after more than 6 h of treatment. Thus the cyanobacterial toxins, such as microcystin, may have evolved to acquire certain abilities to short-cut the apoptotic process, most likely by triggering the mitochondrial changes as the apoptotic initiator. On the other hand, alteration of protein phosphorylation has also been suggested to be an important modulator of apoptosis. Nodularin, another cyanobacterial toxin with potent inhibition of protein phosphatases 1 and 2A, has been demonstrated to cause protein phosphorylations which preceded the cellular apoptotic events. Thus, it is possible that protein phosphorylation may be another important factor in microcystin-induced apoptosis. Moreover, since microcystin induced marked ROS formation in rat hepatocytes, it will be interesting if the phosphoprotein(s) could be identified that are related to the redox status of hepatocytes. However, there is no such phosphoprotein(s) that has been found to be related to the regulation of cellular redox status. On the other hand, it is also not known whether microcystin could phosphorylate Bcl-2 family proteins that regulate the mitochondrial changes. Therefore, the link between protein phosphorylation and microcystin-induced apoptosis needs to be explored further.
While mitochondria have been considered as the central executioner of apoptosis, activation of caspases has been shown to be another important factor in apoptosis. Nevertheless, there is controversy on whether activation of caspases is involved in microcystin-induced apoptosis. Other and our previous reports have demonstrated that ZVAD-fmk, a general caspase inhibitor, can inhibit microcystin-induced apoptosis [13,32], suggesting that caspase may be involved. However, this interpretation should be taken cautiously when using a caspase inhibitor due to the specificity of the inhibitor. ZVAD-fmk has been demonstrated to be able to inhibit other proteases such as calpain, a calcium-dependent cysteine proteinase, which also has been shown to be involved in apoptosis. In line with this, there is no report to directly measure the caspase activity after exposure to microcystin. In our previous study, we also failed to detect any increased activity of caspase-9 and caspase-3 by both caspase substrate cleavage assay and Western blot analysis of PARP cleavage, a hallmark of caspase-3 activation. However, the participation of other caspases such as caspase-2 and caspase-6 should not be excluded in microcystin-induced apoptosis. Notable, nodularin, another cyanobacterial toxin, has been shown to be able to cleave and activate caspase-3. Thus, in order to further elucidate the role of caspase in microcystin-induced apoptosis, besides the caspase inhibitor and cleaved substrate assay, a more specific technique such as Western blot using specific antibody to recognize cleaved caspase bands should be considered in future studies.
Calpain, a calcium-dependent protease, has recently been shown to be involved in apoptosis. Indeed, we demonstrated that calpain was activated in microcystin-treated hepatocytes by the following two pieces of evidence: (i) the time-dependent activation of calpain activity, and (ii) the inhibition of microcystin-induced calpain activation and cell death by the calpain inhibitors, ALLN or ALLM. Based on the data we have obtained, it seems that calpain activation, rather than caspase activation plays an important role in MLR-induced apoptosis. At present, how microcystin activates calpain is not clear. However, the onset of MPT could lead to release of mitochondrial Ca2+, and Ca2+ could further activate calpain. Indeed, the specific MPT inhibitor, cyclosporin A, is able to inhibit MLR-induced calpain activation. To further support the role of intracellular Ca2+, it is recently demonstrated that Ca2+/calmodulin-dependent protein kinase II is required for microcystin-induced apoptosis in hepatocytes. Although how microcystin activates the Ca2+/calmodulin-dependent protein kinase II is not completely clear in that study, it is very likely that the release of mitochondrial Ca2+ may play an important role. Thus, all the above evidence supports the notion that mitochondria play a key role in microcystin-induced apoptosis. The possible cellular events after the microcystin exposure are summarized in Fig. 4.
Recent studies have provided evidence that microcystin caused oxidative stress in hepatocytes. Mitochondria changes, including loss of MMP and onset of MPT are two critical events that have been observed after exposure of hepatocytes to microcystins. This leads to the activation of calpain and Ca2+/calmodulin-dependent protein kinase II and resulted in apoptosis. These findings suggest that oxidative stress and mitochondria changes have a pivotal role in microcystin-induced apoptosis.