I. Introduction

Top Previous Next References

Mitochondria play a central role in energy-generating processes within the cell. Apart from this important function, mitochondria are involved in such complex processes as apoptosis and cardioprotection. A rapidly expanding body of literature also suggests that mitochondrial dysfunctions play pivotal roles in neurodegenerative disorders ranging from Parkinson's to Huntington's to Alzheimer's diseases. Mitochondrial DNA mutations, whether inherited or acquired, cause impaired respiratory chain functioning. This, in turn, leads to decreased production of ATP, formation of free radicals, and alterations in cellular calcium handling. These events may initiate peroxidation of mitochondrial DNA, proteins, and lipids, and opening of the mitochondrial permeability transition pore, an event linked to apoptotic cell death. Mitochondria are also targets for drugs such as antidiabetic sulfonylureas, immunosupressants, some antilipidemic agents, etc. The aim of this review is to present interactions of various therapeutically applied substances with mitochondria as their primary or secondary targets. To make the mechanisms of these effects comprehensible, we will also briefly describe the metabolic routes in which the drugs in question interfere.

Medically applied substances that interact with mitochondria can be divided into two groups: 1) those that are specifically designed to affect mitochondrial functions, and 2) those for which primary targets are other cellular locations and their interactions with mitochondria are secondary. In any case, identification of mitochondria as primary or secondary targets of a drug may facilitate a better understanding of its mechanism of action and open new perspectives of its application. For example, recent studies on the cardioprotective action of potassium channel openers have revealed that cardiac mitochondria are more important as the primary targets of these drugs than is the plasma membrane. Recognition of drug interaction with mitochondria as secondary targets may help us to understand the mechanisms of side effects and to construct new drugs in which these side effects will be eliminated or minimized.

Mitochondria can also be affected by a number of toxins, especially those that interact with their respiratory and ATP-generating functions. This field is the subject of a recent review by Wallace and Starkov (2000) and will not be dealt with in the present article.

When dealing with particular classes of pharmaceuticals, we will often refer to original experimental work in which the drug is used in concentrations that may exceed those encountered under therapeutic conditions, especially for compounds interacting with mitochondria as their secondary targets. We consider such studies useful because they help us to understand the mechanisms of potential side effects, especially under conditions of chronic administration or overdose. In addition, some drugs may accumulate in particular tissues or organs to attain concentrations higher than those calculated for the whole body. Moreover, mitochondria are unique cellular organelles with alkaline and negatively charged interior, conditions that promote accumulation of some compounds.

	II. Mitochondria and the Cell

Top Previous Next References

The number of mitochondria per cell is roughly related to cell energy demands. Somatic tissues contain from a few dozen to several thousand mitochondria per cell. Human spermatozoa contain a fixed number of 16 mitochondria and oocytes up to 100,000. Organs that are very active metabolically, such as muscles, liver, brain, and cardiac and skeletal muscles, contain the largest number of mitochondria and are most susceptible to drugs acting on mitochondria and to mitochondrial pathologies.

In major mammalian tissues, 80 to 90% of ATP is generated by mitochondria in the process of oxidative phosphorylation. The mitochondrial respiratory chain, located in the inner mitochondrial membrane, is composed of enzymes and low molecular weight redox intermediates (coenzymes) that transport "reducing equivalents", in fact hydrogen atoms or just their electrons, from respiratory substrates to molecular oxygen, down the redox potential. This hydrogen/electron current forms three cascades in which the redox energy is high enough to be utilized to extrude protons from the mitochondrial inner compartment, the matrix, to the intermembrane space. The electrochemical proton gradient thus formed, also designated as the protonmotive force p, is the driving force for the back flow of protons through the ATP synthase complex (Fig. 1B). This gradient is composed of the electric component (²) and the proton concentration gradient. As a result, mitochondria are unique cellular organelles that can build up a transmembrane electric potential of up to 180 mV, negative inside, and whose internal milieu maintains a pH value of about 8 (Nicholls and Ferguson, 1992). As a consequence, they can not only accumulate membrane-permeable compounds of cationic character, but also trap weak acids in their anionic form. Both properties may be of importance in targeting specific drugs into mitochondria (see, for example, Section III.C.).

View larger version (71K): [in this window] [in a new window]   Fig. 1.   Schematic representation of the mitochondrion (A) and the respiratory chain (B). pyr, pyruvate; dicarb, dicarboxylate; UQ, ubiquinone (coenzyme Q); cyt. c, cytochrome c; Glib, glibenclamide; 5-HD, 5-hydroxydecanoate; RR, ruthenium red.

Because of a vital role of mitochondria in cell metabolism, a constant and intense flux of inorganic ions and metabolites occurs between the cytosol and mitochondria. Due to the presence of a pore protein, termed porin or voltage-dependent anion channel (VDAC) (Mannella et al., 1992), the outer mitochondrial membrane is permeable to polar molecules of up to 5 kDa. In contrast, the inner membrane is freely permeable to just a few compounds such as water, O₂, CO₂, and NH₃. Other hydrophilic metabolites and all inorganic ions of biological importance can cross the membrane due to the presence of specific channels and carrier proteins. Among the latter ones of special importance are carriers for phosphate (P_i), adenine nucleotides ADP and ATP, and the respiratory substrates mono-, di-, and tricarboxylates. In general, these transport proteins operate in the exchange mode, i.e., ADP is exchanged for ATP, P_i for OH, dicarboxylic anion for P_i anion, etc. (Palmieri et al., 1992; Kuan and Saier, 1993; Palmieri, 1994; Sluse, 1996).

Channels selective for major inorganic cations K⁺, Na⁺, Mg²⁺, and Ca²⁺ have been identified in mitochondria (Bernardi, 1999). Of particular importance are channels for K⁺ that will be dealt with later in this review (Section V.). Apart from these channels, which must be regulated in a subtle way to prevent collapse of the membrane potential, there are several cation exchangers, e.g., for K⁺/H⁺, Ca²⁺/H⁺, and Ca²⁺/Na⁺ exchange (Bernardi, 1999). An inner membrane anion channel has also been described (Beavis, 1992), but its molecular identity, characteristics, and control are less well recognized. Recently, a mitochondrial chloride channel has been cloned (Fernández-Salas et al., 1999).

Many lipophilic compounds penetrate the inner mitochondrial membrane freely. Among them, the most important are fatty acids. Undissociated molecules of long-chain fatty acids can easily penetrate the membrane in a "flip-flop" mode (McLaughlin and Dilger, 1980; Gutknecht, 1988). In contrast, a spontaneous crossing of the phospholipid bilayer by fatty acid anions is extremely slow (Kamp and Hamilton, 1992) because of the negatively charged polar carboxylic group. However, in the inner mitochondrial membrane, the transmembrane passage of fatty acid anions is facilitated, probably in an unspecific way, by several mitochondrial anion carrier proteins. This is a process that is essential for the protonophoric action of fatty acids in mitochondria (Wojtczak and Wickowski, 1999). Many lipophilic and amphiphilic drugs are good mitochondrial penetrants.

Although the outer and the inner mitochondrial membranes are well defined structures, each of them possessing different sets of enzymes and fulfilling different functions, intimate contacts between the two membranes have been identified on both morphological and functional grounds (Brdiczka, 1991).

Mitochondria contain circular DNA that encodes about two dozen polypeptide chains participating, as subunits, in mitochondrial respiratory chain complexes and other essential components of the energy-coupling machinery. However, the majority of mitochondrial proteins are encoded by nuclear DNA and synthesized outside the mitochondria. They are imported into these organelles by a complex multistep mechanism (Lill and Neupert, 1996; Schatz, 1996).

Although the end product of the respiratory chain is water that is generated in a four-electron reduction of molecular oxygen by cytochrome oxidase (complex IV), a minor proportion of O₂ can be involved in one-electron reduction processes, generating so-called "reactive oxygen species" (ROS), in particular, superoxide anion radical O₂, hydrogen peroxide H₂O₂, and the extremely reactive hydroxyl radical HO^·. Generation of ROS occurs mainly at complex III due to proton cycling between ubiquinone, cytochromes b and c₁, and iron-sulfur protein (Sugioka et al., 1988). Some contribution of complex I to this process has also been found. Reactive oxygen species are generally regarded as toxic metabolites and, as such, are decomposed by specialized enzymes: catalase, peroxidases, and superoxide dismutases. Nevertheless, a part of these reactive compounds that has sustained this catalytic removal may have a dramatic effect on mitochondria and the cell as a whole by eliciting a cascade of events leading to programmed cell death, so-called "apoptosis" (see Section III.).

This brief description of mitochondrial functions is aimed to point to multiple pathways interconnecting mitochondrial metabolic routes with those of the rest of the cell. Pharmacological agents, which affect either intramitochondrial metabolic processes or transport pathways connecting mitochondria with the cytosol, may therefore have a considerable influence on the total cell metabolism. Figure 1A presents the main mitochondrial processes and their locations within various mitochondrial compartments as well as routes of communication between them and the rest of the cell. It will also help to locate processes and points of pharmacological attack that will be dealt with in subsequent sections.

	III. Mitochondria in Chemotherapy-Induced Apoptosis

Top Previous Next References

Programmed cell death, or apoptosis, is a common process in multicellular organisms. It enables the elimination of single cells or their assemblies when their natural biological function has come to an end or when the cell has become damaged or mutated to such an extent that its further existence might be dangerous to the whole organism. In particular, apoptosis occurs in embryogenesis, in metamorphosis, and in the growth and maturation of individual organs. Being of such vital importance, apoptosis has attained, during evolution, a high degree of genetic and metabolic control. This broad subject is covered by numerous excellent reviews (e.g., Ellis et al., 1991; Kroemer et al., 1995; Vaux and Strasser, 1996; Hengartner, 2000). Therefore, the present considerations will be limited to the function that is played in apoptosis by mitochondria and to the pharmacological modulations of apoptosis in relation to cancer therapy (for a comprehensive review, see also Deigner and Kinscherf, 1999). It is generally believed that transformations that may lead to malignancy are rather common in cells of highly organized animals, including humans. However, such cells, as a rule, are efficiently eliminated by apoptosis put in operation by mechanisms deeply encoded in their genome. Only cells that have escaped those rescue systems give rise to malignant growth.

In general, two partly interdependent routes may lead to apoptosis (Hengartner, 2000; Kaufmann and Earnshaw, 2000). One of them is initiated by ligation of the so-called death receptors at the cell surface (Ashkenazi and Dixit, 1999), whereas the other one involves mitochondria (Petit et al., 1997; Green and Reed, 1998; Kroemer et al., 1998; Mignotte and Vayssiere, 1998; Susin et al., 1998; Kroemer, 1999; Desagher and Martinou, 2000; Halestrap et al., 2000). In the latter case, one of the early events leading to apoptosis is the release of cytochrome c from mitochondria (Liu et al., 1996; Kluck et al., 1997; Reed, 1997; Yang et al., 1997; Martinou et al., 2000). Along with another mitochondrial protein, the apoptosis-inducing factor (Daugas et al., 2000), it elicits in the cytosol a cascade of events leading to the activation of intracellular proteases of the caspase family (Earnshaw et al., 1999) and, eventually, to a partial self-digestion of the cell (Bossy-Wetzel and Green, 1999) (Fig. 2). The mechanism by which cytochrome c is liberated from mitochondria to the cytosol is debatable. According to some authors (e.g., Scarlett and Murphy, 1997; Vander Heiden et al., 1997; Petit et al., 1998), this is preceded by mitochondrial swelling that leads to disruption of the outer membrane.

View larger version (74K): [in this window] [in a new window]   Fig. 2.   Simplified scheme of the mitochondrial pathway of apoptosis. The pathway is triggered by various "death signals", such as ROS, DNA damage, etc., that promote binding of the proapoptotic protein Bax with the outer mitochondrial membrane, most likely at the contact sites between the two membranes, and its association with the PTP. This enables the release of cytochrome c () and the apoptosis-inducing factor (AIF; ) from the intermembrane compartment to the cytosol. An elevated intramitochondrial Ca2+ level and ROS production facilitate this process by promoting PTP opening. Once in the cytosol, cytochrome c and AIF, in cooperation with a cytosolic factor, Apaf-1 (not indicated), activate caspase-9 and subsequently other members of the caspase family, thus initiating self-digestion of the cell and nuclear DNA fragmentation, eventually leading to apoptotic cell death. Association of Bax with mitochondria is prevented by the antiapoptotic protein Bcl-2. ROS can be decomposed by Mn-containing (mitochondrial) and Cu, Zn-containing (cytosolic) superoxide dismutases (SOD), catalase, and glutathione peroxidase (GPx). Stimulation of ROS production is exemplified here by UV and ionizing radiation and by two anticancer drugs, Adriamycin and BMD188. Activation is indicated as  and inhibition as .

More recent reports indicate, however, that cytochrome c is liberated from mitochondria by special mechanisms under conditions of preserved intactness of the outer membrane (Jürgensmeier et al., 1998; Doran and Halestrap, 2000). A decisive role in this process is played by the mitochondrial permeability transition pore (PTP) and the proapoptotic protein Bax (Marzo et al., 1998; Crompton, 1999). PTP is located in the contact sites between the outer and the inner mitochondrial membranes and, in its open state, enables a free passage of low molecular weight compounds, up to molecular weight 1500, between the mitochondrial inner compartment (matrix) and the cytosol. It constitutes a complex assembly of porin, adenine nucleotide translocase, and cyclophilin D (Fig. 3). The opening of PTP is favored by factors such as Ca²⁺ accumulation in mitochondria (Hunter et al., 1976), prooxidants (Byrne et al., 1999), and low mitochondrial transmembrane potential (Bernardi, 1992) (for more information see the reviews by Zoratti and Szabó, 1995; Bernardi, 1996; Bernardi et al., 1998; Fontaine and Bernardi, 1999; Halestrap, 1999).

View larger version (67K): [in this window] [in a new window]   Fig. 3.   Molecular composition of the mitochondrial permeability transition pore. The scheme on the right side represents the PTP as situated in the contact site, interconnecting the two mitochondrial membranes. The left-hand diagram shows the subunit composition of the contact site which also comprises the peripheral benzodiazepine receptor. HK, hexokinase; ANT, adenine nucleotide translocase; CK, creatine kinase (in muscle mitochondria); CpD, cyclophilin D. Binding of cyclophilin D by cyclosporin A and its removal from the membrane is indicated as .

PTP alone is too small to enable the release of cytochrome c (13 kDa). However, its association with Bax forms a channel specific for cytochrome c and apoptosis-inducing factor (Hirsch et al., 1997; Narita et al., 1998; Jacotot et al., 1999; Murphy et al., 2000). Association of Bax with PTP and its pore-forming activity are prevented by the antiapoptotic protein Bcl-2 (Kluck et al., 1997; Yang et al., 1997; Murphy et al., 2000). Thus, a subtle balance between these two (and other similar) proapoptotic and antiapoptotic proteins and their interaction with PTP are decisive for the survival or apoptotic death of the cell. This balance can be affected by a number of mitochondria-targeted drugs.

Because of the pivotal role that mitochondria play in initiating apoptotic cell decay, they are a vulnerable target for experimental and/or pharmacological intervention. As mentioned above, one of the main physiological factors that regulates the open probability of PTP is calcium ion. Therefore, the role of Ca²⁺ in the regulation of apoptosis has attained much attention, but the picture still remains obscure. Numerous observations have revealed an increased cytosolic Ca²⁺ concentration during apoptosis (McConkey and Orrenius, 1997). Manipulation of Ca²⁺ concentration within the intact cell can be performed, for example, by using thapsigargin, a potent inhibitor of endoplasmic reticulum Ca²⁺-ATPase (Thastrup et al., 1990), which prevents calcium accumulation in the endoplasmic reticulum. It has been observed (Zhu and Loh, 1995; Waring and Beaver, 1996; Bian et al., 1997) that thapsigargin induces apoptosis in various cell lines. Apoptosis induced in thymocytes by very low concentrations of thapsigargin can be prevented by cyclosporin A (Waring and Beaver, 1996), the well know inhibitor of mitochondrial PTP (Broekemeier et al., 1989), thus pointing to the involvement of mitochondria in this process. On the other hand, it has been observed that thapsigargin also initiates apoptosis after removal of extracellular Ca²⁺ (Bian et al., 1997), i.e., under conditions in which accumulation of calcium ions inside the mitochondria is prevented. Zhu et al. (2000) have recently come to the conclusion that mitochondrial Ca²⁺ depletion promotes apoptosis, whereas mitochondrial Ca²⁺ overload leads to necrosis of cardiac myocytes and neuroblastoma cells. Thus, the role of mitochondria as a target for calcium ions in the regulation of apoptosis still awaits its elucidation.

Another apoptosis-promoting factor is ROS (Buttke and Sandstrom, 1994; Slater et al., 1995; Rollet-Labelle et al., 1998; Jabs, 1999; Chandra et al., 2000; Matés and Sánchez-Jiménez, 2000). Excessive production of ROS in the cell can be induced by a number of xenobiotics, transition metal ions, and ultraviolet and ionizing radiations (Halliwell and Gutteridge, 1989, 1990). ROS action on mitochondria results in the opening of PTP (Kowaltowski et al., 1996; Vercesi et al., 1997) and thus triggers the mitochondria-related apoptotic pathway. It is also possible that peroxidative attack may directly damage the outer mitochondrial membrane, resulting in unspecific liberation of intermembrane proteins, including a fraction of cytochrome c. Apoptosis induced by ultraviolet radiation can be reduced or completely prevented by glutathione (Slyshenkov et al., 2001) that removes oxygen free radicals. This observation indicates that ultraviolet radiation initiates apoptosis by acting directly on mitochondria rather than on the genomic system. Ionizing radiation (X-ray and -radiation), often used in cancer treatment, also acts by inducing apoptosis. Being more energetic than ultraviolet radiation, it also affects DNA and thus puts into operation both the DNA- and the mitochondria-operated apoptosis pathways (Rupnow and Knox, 1999).

Because PTP opens upon collapse of , chemical and physical agents that discharge the mitochondrial membrane potential can be regarded as pro-apoptogenic. However, the picture is complicated by the fact that a complete deenergization of the cell leads to necrosis rather than to apoptosis (Leist and Nicotera, 1997). Therefore, most mitochondrial protonophoric uncouplers do not induce apoptosis. On the other hand, opening of PTP by other factors results in a decrease due to an increased proton influx through the pore. Thus, the PTP opening can be either the result or the causative agent of collapse. Having this in mind, one has to critically evaluate a long list of anticancer agents causing cell death and "disruption" of , as presented by Decaudin et al. (1998).

Recent studies have shown that a large number of anticancer drugs exert their therapeutic action by inducing the apoptosis of malignant cells (Debatin, 2000; Preston et al., 2001), mainly due to activation of the cytochrome c/caspase-9 pathway (Kaufmann and Earnshaw, 2000). Thus, etoposide, doxorubicin, 1--D-arabinofuranosylcytosine (Decaudin et al., 1997), lonidamine (Ravagnan et al., 1999), betulinic acid, arsenite, CD437, and several amphiphilic cationic -helical peptides (Decaudin et al., 1998; Costantini et al., 2000) induce apoptosis of malignant cells by activating PTP or otherwise affecting the mitochondrial membrane. The proapoptotic action of some of these compounds, e.g., that of lonidamine (Ravagnan et al., 1999), is counteracted by cyclosporin A (CsA) and prevented by overexpression of the antiapoptotic protein Bcl-2. This points to its mechanism of action by the opening of the PTP.

Adriamycin (doxorubicin) is a potent antitumor drug of the anthracycline antibiotic group the proapoptotic action of which is well documented (Müller et al., 1998; Kumar et al., 1999). Its application in chemotherapy is, however, strongly limited by its well known high cytotoxicity (cardio-, nephro-, and neurotoxicity) (Julka et al., 1993; Morgan et al., 1998; Singal et al., 2000). These toxic effects can be alleviated by antioxidants (Singal et al., 1997; DeAtley et al., 1999), thus confirming the notion (Olson et al., 1981; Sokolove, 1994) that Adriamycin and other quinoid anthracyclines are free radical generators (Fig. 2).

A novel anti-prostate cancer hydroxamic acid derivative, cis-1-hydroxy-4-(1-naphthyl)-6-octylpiperidine-2-one (designated as BMD188) (Tang et al., 1998; Li et al., 1999), is also an apoptosis-inducing agent. Although its mechanism of action is unclear and, at least in some cell lines, it does not involve cytochrome c liberation from mitochondria (Tang et al., 1998), more recent studies have provided evidence that its target is the mitochondrial respiratory chain and that it induces a rapid production of ROS (Joshi et al., 1999). An involvement of PTP seems, therefore, very likely (Fig. 2).

An ingenious way of destroying malignant tissues by synthetic doubly targeted peptides has been proposed (Arap et al., 1998; Ellerby et al., 1999). The authors designed helical amphipathic peptides containing mostly cationic amino acids that preferentially bound to the inner mitochondrial membrane (and prokaryotic cytoplasmic membranes) due to its high transmembrane potential and high content of anionic phospholipids. This binding resulted in the distortion of the lipid core of the membrane, manifested by mitochondrial swelling. These peptides manifested a much lower affinity toward the eukaryotic plasma membrane because of its lower membrane potential and content of mostly zwitterionic phospholipids. For that reason, they affected the integrity of the plasma membrane at concentrations hundreds of times higher than those needed to destroy bacterial or mitochondrial membranes. These peptides were then coupled to short cyclic peptides designed to be targeted toward angiogenic endothelial cells and some malignant cell lines (Pasqualini et al., 1997; Arap et al., 1998). Such hybrid peptides were internalized into the cytosol of angiogenic (but not angiostatic) endothelial cells and cancer cells, and their helical moieties associated with mitochondria, resulting in the swelling of these organelles. This led to classical symptoms of apoptosis. This treatment proved successful not only in killing cells in culture but also in hampering the growth of human breast cancer implanted into mice and in prolonging by several months the survival of such mice (Arap et al., 1998; Ellerby et al., 1999).

	IV. Mitochondria and Oxidative Stress, Aging, and Degenerative Diseases

Top Previous Next References

The biological importance of ROS has attracted an enormous interest during recent years due to their major role, both beneficial and noxious, in numerous vital processes. This subject is covered by thousands of original research papers and hundreds of review articles appearing each year. Only some of them can be referred to in this section. We will cite, in particular, the most recent review articles, as they will lead the reader to broader original literature and the most up-to-date information.

Mitochondria are the main source of the superoxide radical and other reactive oxygen species that may generate from them (Chance et al., 1979). The main mechanisms responsible for mitochondrial ROS production are the respiratory chain, in particular its complexes I and III (Beyer, 1992; Cadenas and Davies, 2000; Raha and Robinson, 2000), in the inner mitochondrial membrane, and monoamine oxidase in the outer membrane. Normally, ROS are decomposed or their peroxidation products are neutralized by natural defense systems mainly consisting of mitochondrial (manganese-containing) and cytosolic (containing Cu and Zn) superoxide dismutases (Mn- and CuZn-superooxide dismutase, respectively), glutathione peroxidase, and phospholipid hydroperoxide glutathione peroxidase (Neubert et al., 1962; Fridovich, 1974; Chance et al., 1979; Ursini et al., 1986; Augustin et al., 1997). However, under conditions of increased ROS generation, e.g., in ischemia-reperfusion, action of some xenobiotics, inflammation, aging, and ultraviolet or ionizing irradiation, or conditions of impaired antioxidant defense system, ROS may accumulate, exerting a potent damaging effect on the cell and the whole organism (Halliwell and Gutteridge, 1989, 1990; Mizuno et al., 1998; Schapira, 1999; Cadenas and Davies, 2000; Raha and Robinson, 2000). The noxious action of ROS mainly consists of the peroxidation of lipids, in particular phospholipids of biological membranes, and oxidative damage to proteins and DNA (Halliwell and Gutteridge, 1990; Lenaz et al., 1999; Cadenas and Davies, 2000). In particular, the aging of animals and humans is connected with increased mitochondrial production of ROS (Miquel et al., 1980; Lenaz, 1998; Cadenas and Davies, 2000; Lenaz et al., 2000; Sastre et al., 2000; Raha and Robinson, 2000). Mitochondria, being the main site of ROS generation in the cell, are also their primary target. This, in turn, results in damage to the mitochondrial respiratory chain and, as a consequence, a further increase in ROS generation. A vicious cycle is thus formed (Lenaz et al., 2000) that may be a causative agent of a number of age-associated dysfunctions of mitochondria and also one of the mechanisms inducing programmed cell death (see Section III.).

To protect mitochondria and the cell against such damaging effects, several measures can be applied. One of them is increasing the intracellular glutathione content. This can be done by supplying precursors for glutathione synthesis, e.g., N-acetylcysteine, which by itself has antioxidant properties (Benrahmoune et al., 2000). Increasing the cell content of CoA by supplying its precursor pantothenic acid (Slyshenkov et al., 1995) also increases the glutathione level (Slyshenkov et al., 1996), although the mechanism by which an increased content of CoA accelerates glutathione synthesis is not fully understood. Glutathione content and its reduction state can also be increased by incubating the cells with curcumin, the yellow pigment of the Indian spice curry (Jaruga et al., 1998), or with the analgesic drug flupirtine (Perovic et al., 1996); but the mechanisms of these processes are even more obscure. These ways of protecting mitochondria against damaging effects of ROS, although very effective, are not specifically mitochondria-directed because the glutathione content increases in the whole cell and not only in mitochondria.

Other compounds acting as general intracellular antioxidants are ascorbic acid (vitamin C), -tocopherol (vitamin E), -carotene (Sies et al., 1992), and -lipoic acid (Packer et al., 1995). All these compounds are naturally present in the cell, but their contents can be increased when they are additionally administered.

Ubiquinol, the reduced form of coenzyme Q, is a naturally occurring antioxidant which can be targeted more specifically to mitochondria and other cell membranes (Ernster, 1994; Forstmark-Andrée et al., 1995; Andrée et al., 1999; Shi et al., 1999). The redox couple of ubiquinone/ubiquinol plays a key role in the mitochondrial respiratory chain as a link between its complexes I, II, and III (Nicholls and Ferguson, 1992). Much attention has recently been paid to ubiquinone and ubiquinol as potential anti-aging remedies (Lenaz, 1998; Huertas et al., 1999; Lenaz et al., 2000). Some hope has also been expressed as to the beneficial effect of coenzyme Q in retarding and/or reducing symptoms of Alzheimer's, Huntington's and Parkinson's diseases (Beal, 1998; Beal et al., 1998; Cassarino and Bennett, 1999), considered to be caused by a damaging action of ROS in the central nervous system (Mizuno et al., 1998; Schapira, 1999).

The situation with ubiquinol as an antioxidant is, however, complicated by the fact that its operation in the respiratory chain may also be, under specific conditions, a source of free radicals (Beyer, 1992; Nohl et al., 1999; Cadenas and Davies, 2000; Raha et al., 2000). In this respect, ubiquinol resembles some other antioxidants; for example, ascorbic acid and -carotene, well known to both scavenge and generate free radicals.

An ingenious way of specifically introducing antioxidants to mitochondria within the intact cell has been recently proposed by Murphy and coworkers (Smith et al., 1999; Coulter et al., 2000; Murphy and Smith, 2000). These authors covalently coupled antioxidant moieties with the lipophilic triphenylphosphonium cation. Such compounds, being positively charged, are accumulated within mitochondria up to a thousand-fold, driven by the transmembrane electric potential of about 200 mV (negative inside) in fully energized mitochondria. Such accumulation can be further increased by the electric potential at the plasma membrane of 30 to 60 mV (also negative inside), so that the calculated inside to outside concentration ratio may amount up to 10⁴:1. Among the antioxidant moieties coupled to the lipophilic cation were natural products vitamin E and ubiquinol, fullerene derivatives, and a spin trap (Coulter et al., 2000). Tetraphenylphosphonium coupled to vitamin E appeared to exert a potent protective effect against the damage of isolated mitochondria by a hydroxyl radical-generating system (Smith et al., 1999).

Recently, it has been found (Schlüter et al., 2000) that the nonopioid analgesic drug, flupirtine (a fluor-containing triaminopyridine derivative), is an effective antioxidant in mitochondria. Being positively charged, flupirtine most likely accumulates in mitochondria where it efficiently scavenges free radicals. This action of flupirtine explains why this compound prevents apoptosis of cultured cells induced by oxidative stress (Lorenz et al., 1998) and protects rabbit retina against ischemic injury (Osborne et al., 1996).

Finally, it has to be mentioned that an effective way of diminishing mitochondrial production of ROS is a partial uncoupling, i.e., decreasing, of the mitochondrial transmembrane potential () and of the redox state of complex I of the respiratory chain (Skulachev, 1996). This can be obtained by increasing the permeability of the inner mitochondrial membrane to K⁺ or H⁺ by means of the opening of mitochondrial potassium channels (see Section V.) or by endogenous uncoupling proteins (see Section XII.).

	V. Interaction of Potassium Channel Openers with Mitochondria

Top Previous Next References

Potassium channel openers (KCOs) are agents, discovered in the early 1980s, that act by stimulating ion flux through K⁺ channels (Edwards and Weston, 1993; Challinor-Rogers and McPherson, 1994; Quast et al., 1994). Many drugs such as cromakalim, nicorandil, and diazoxide have been identified as KCOs (Duty and Weston, 1990; Edwards and Weston, 1995). KCOs act on two types of ion channels: ATP-regulated K⁺ channels (K_ATP channels) (Ashcroft, 2000) and Ca²⁺-activated K⁺ channels (BK channels) (Lawson, 2000). KCOs were first identified by their antianginal or antihypertensive mode of action (Edwards and Weston, 1993). Now, they are at various stages of development as antiasthmatic (Morley, 1994; Buchheit and Fozard, 1999; Prasad et al., 2000) and cardioprotective agents (Grover and Garlid, 2000). Preclinical and clinical evidence also supports the therapeutic role of KCOs in pulmonary (Wanstall and Jeffery, 1998) and vascular hypertension (Quast, 1992; Atwal, 1994; Quast et al., 1994; Schachter, 1995; Lawson, 1996), and the treatment of overactive bladder (Andersson, 2000).

Until recently, the effects of KCOs were believed to be attributable entirely to the modulation of K⁺ channels in cell surface membranes. It is now apparent, however, that new targets for KCOs exist in intracellular membranes including those of sarcoplasmic reticulum (Kourie, 1998), zymogen granules (Thevenod et al., 1992), and mitochondria (Garlid, 1994; 1996; Szewczyk et al., 1996a; Szewczyk, 1997). Mitochondria seem to be particularly important targets for KCOs because the interaction of KCOs with these organelles appears to mediate the cardioprotective action of these compounds (Szewczyk and Marban, 1999; Grover and Garlid, 2000). The protective role of mitochondrial ion channels was recently summarized in an excellent review article (O'Rourke, 2000). Mitochondrial targets for anti-ischemic drugs were recently described (Morin et al., 2001; Suleiman et al., 2001).

A small-conductance potassium channel, with properties similar to those of the K_ATP channel from the plasma membrane, was described in the inner membrane of rat liver and beef heart mitochondria and designated the mitochondrial ATP-regulated potassium channel (mitoK_ATP channel) (Inoue et al., 1991; Paucek et al., 1992). The mitoK_ATP channel was blocked not only by ATP, but also, similarly to the plasma membrane K_ATP channel, by antidiabetic sulfonylureas (Fig. 4) (Inoue et al., 1991; Paucek et al., 1992). These observations raised the question whether the mitoK_ATP channel could be activated by KCOs. In fact, an increased influx of K⁺ and depolarization of liver mitochondria in the presence of KCOs such as RP66471 was observed (Szewczyk et al., 1993, 1995). Also, other KCOs were shown to activate potassium ion transport into mitochondria (Belyaeva et al., 1993; Czy et al., 1995; Garlid et al., 1996b; Holmuhamedov et al., 1998). Moreover, ATP-inhibited K⁺ flux was restored by diazoxide (K_1/2 of 0.4 µM), cromakalim (K_1/2 of 1 µM), and two cromakalim analogs, EMD60480 and EMD57970 (K_1/2 of 6 nM) (Garlid et al., 1996b). KCOs such as pinacidil, cromakalim, and levcromakalim have been shown to depolarize cardiac mitochondria (Holmuhamedov et al., 1998). KCO-induced membrane depolarization was associated with an increase in the rate of mitochondrial respiration and decreased ATP synthesis (Holmuhamedov et al., 1998). Moreover, KCOs released calcium ions and cytochrome c from cardiac mitochondria (Holmuhamedov et al., 1998). Despite the effect on K⁺ transport, diazoxide also exhibits a direct effect on mitochondrial energy metabolism by inhibition of respiratory chain complex II in liver mitochondria (Grimmsmann and Rustenbeck, 1998). Recently, mitoK_ATP channel opener BMS-191095 with no peripheral vasodilator activity was described (Grover et al., 2001).

View larger version (79K): [in this window] [in a new window]   Fig. 4.   Interaction of potassium channel openers and inhibitors with mitochondria. Activation is indicated as  and inhibition as . PKC, protein kinase C-; mitoSUR, mitochondrial sulfonylurea receptor.

Using isolated mitochondria or proteoliposomes reconstituted with partly purified mitoK_ATP channel and measuring potassium flux, Garlid et al. (1996b) demonstrated that heart and liver mitoK_ATP channels share some pharmacological properties with the plasma membrane K_ATP channel, i.e., they are both activated by KCOs. The sensitivity of cardiac mitoK_ATP channels to diazoxide appeared to be 1000 times higher than that of plasma membrane K_ATP channels. This observation established that the interaction of KCOs with mitoK_ATP channels plays an important role in cardioprotection.

KCOs mimic cardiac ischemic preconditioning in the absence of ischemia, whereas K_ATP channel blockers, such as glibenclamide and 5-hydroxydecanoic acid, diminish the beneficial effects of short ischemic events on the cardiac tissue. The original hypothesis to explain these observations involved plasma membrane K_ATP channels. Recently, it has been proven that the action of KCOs such as diazoxide concerns, in fact, mitochondria and the mitoK_ATP channel (Garlid et al., 1997). In a complementary approach, it was shown that diazoxide induced oxidation of mitochondrial flavoproteins, due to the activation of mitoK_ATP channel, but did not activate plasma membrane K_ATP channels (Liu et al., 1998). The effects of diazoxide were completely and reversibly blocked by 5-hydroxydecanoic acid. Interestingly, exposure to phorbol-12-myristate-13-acetate potentiated and accelerated the effect of diazoxide (Sato et al., 1998). These studies established that the target for the protective effects of diazoxide in cardiac myocytes is the mitoK_ATP channel rather than the plasma membrane K_ATP channel (Fig. 5). Importantly, evidence for mitoK_ATP channels as effectors of myocardial preconditioning has also been demonstrated in human subjects (Ghosh et al., 2000).

View larger version (61K): [in this window] [in a new window]   Fig. 5.   Mitochondria and intracellular signaling cascade that leads to protection of cardiomyocytes against ischemic injury. The main target points of potassium channel openers are indicated with bold arrows. PKC, protein kinase C; MAPKs, mitogen-activated protein kinases; NF-B, transcriptional nuclear factor B; MnSOD, manganese-containing mitochondrial superoxide dismutase; DAG, diacylglycerol; PtdIns(4,5)P2, phosphatidylinositol-4,5-bisphosphate.

The initial observations on the cardioprotective action of KCOs on mitochondria were further confirmed and developed in a series of reports. It has been shown that other KCOs such as pinacidil, cromakalim, and nicorandil modulate mitochondrial membrane potential, respiration, ATP generation, and mitochondrial Ca²⁺ uptake (Holmuhamedov et al., 1998; Sato et al., 2000). Other data suggest that activation and diazoxide-induced translocation of the protein kinase C -isoform to mitochondria appears to be important for the protection mediated by the mitoK_ATP channel (Wang et al., 1999; Wang and Ashraf, 1999). Recently, it has been shown with the use of diazoxide that ischemic preconditioning depends on the interaction between actin cytoskeleton and mitochondria, and its protective action can be abolished by disruption of the cytoskeleton by cytochalasin D (Baines et al., 1999). Interestingly, diazoxide was also effective in improving the preservation of globally ischemic cold-stored hearts, as it occurs during cardiac transplantation (Kevelaitis et al., 1999, 2000; Ahmet et al., 2000).

The main question remains how the opening of the mitoK_ATP channel could protect cells against ischemic injury. First, opening of the mitoK_ATP channel followed by mitochondrial swelling could improve mitochondrial ATP production and/or handling (Garlid, 2000). In fact, diazoxide was found to preserve mitochondrial function in ischemic rat heart. It has been shown that hypoxia induces a decrease in the mitochondrial oxygen consumption rate to approximately 40% of the prehypoxic value, and treatment with diazoxide preserves the normal mitochondrial oxygen consumption rate during hypoxia (Iwai et al., 2000). Moreover, ATP concentration was significantly increased in diazoxide-treated hearts (Wang et al., 1999). Second, the protective effect of mitoK_ATP activation could be mediated by lowering Ca²⁺ overloading of mitochondria (Holmuhamedov et al., 1998; Crestanello et al., 2000). Third, it has been demonstrated that opening of the mitoK_ATP channel may increase ROS generation by mitochondria (Pain et al., 2000). This increase could lead to protein kinase C activation, which is known to be important during cardioprotection. Additionally, the mitoK_ATP channel seems to be involved in delayed preconditioning (Carroll and Yellon, 2000), probably due to an altered expression of "protective" proteins. It has been shown that pretreatment of hippocampal neurons with KCOs cromakalim and diazoxide increases the expression level of proteins involved in the control of apoptosis, such as Bcl-2 and Bcl-X_L (Jakob et al., 2000). Moreover, inhibition of apoptosis induced by oxidative stress in cardiac cells was observed (Akao et al., 2001). The presence of a mitochondrial target for diazoxide in hippocampal mitochondria recently has been observed (Dbska et al., 2001).

	VI. Sulfonylureas and Mitochondria

Top Previous Next References

It is well known that antidiabetic sulfonylureas such as glibenclamide (also known as glyburide) or glipizide bind to high affinity sulfonylurea receptors (SURs) in the plasma membrane of various cell types (Ashcroft and Ashcroft, 1992; Isomoto and Kurachi, 1997). In pancreatic -cells this causes a closure of the ATP-sensitive K⁺ (K_ATP) channel (Ashcroft and Ashcroft 1992; Lazdunski, 1994; Ashcroft, 2000) and initiates a chain of events that leads to the exocytotic release of insulin (Miki et al., 1999). The pancreatic -cell SUR was cloned (Aguilar-Bryan et al., 1995) and identified as an element composing, together with a K⁺ pore, the functional K_ATP channel (Inagaki et al., 1995). Similar channels are also present in the plasma membrane of smooth, skeletal, and cardiac muscle cells as well as in neurons (DeWeille, 1992). The channels are heterooctamers of four inwardly rectifying K⁺ channels (Kir) and four SURs. Two members of the Kir family, Kir6.1 and Kir6.2, appear capable of forming the pores of K_ATP channels (Aguilar-Bryan et al., 1998; Babenko et al., 1998; Seino, 1999). These Kir subunits coassemble with SURs encoded by either of two genes, SUR1 or SUR2, to form functional K_ATP channels. SURs belong to the superfamily of ATP-binding cassette proteins characterized by the presence of two nucleotide-binding folds within the molecule. Moreover, it is likely that various SUR subunits in combination with Kir6.x subunits contribute to the functional diversity of K_ATP channels and determine various pharmacological properties of these channels, including their regulation by KCOs.

A potassium channel, with properties similar to those of the K_ATP channel of the plasma membrane, was described in the inner mitochondrial membrane and designated as the mitoK_ATP channel (see Section V.). This channel is also blocked by antidiabetic sulfonylureas (Inoue et al., 1991; Paucek et al., 1992). The interaction of antitumor sulfonylureas with mitochondria has also been described (Fig. 6) (Howbert et al., 1990).

View larger version (49K): [in this window] [in a new window]   Fig. 6.   Chemical structure of antidiabetic and antitumor sulfonylureas.

The effects of antidiabetic sulfonylureas on mitochondrial K⁺ transport have been observed both in intact mitochondria (Szewczyk et al., 1997) and in proteoliposomes reconstituted with partly purified mitoK_ATP channel (Paucek et al., 1992). The mitoK_ATP channel became sensitive to glibenclamide only when opened by Mg²⁺, ATP, and physiological activators such as GTP, or by the KCO diazoxide. In such an induced open state of the mitoK_ATP channel, glibenclamide inhibited the channel activity with a K_1/2 of 1 to 6 µM (Jabrek et al., 1998).

Equilibrium binding studies performed with [³H]glibenclamide reveal a single class of low-affinity binding sites in intact rat liver mitochondria, with a K_d of 4 µM (Szewczyk et al., 1996b). In beef heart mitochondria the K_d for glibenclamide binding is much lower, 300 nM (Szewczyk et al., 1997). Glibenclamide binding to mitochondria is modulated by SH reagents such as N-ethylmaleimide and mersalyl (Szewczyk et al., 1999).

It is important to mention that, due to the hydrophobicity of its protonated form, glibenclamide is able to increase the proton conductance of the mitochondrial membrane (Szewczyk et al., 1997). Additionally, antidiabetic sulfonylureas such as glibenclamide and tolbutamide affect fatty acid oxidation due to the inhibition of carnitine palmitoyltransferases (Patel, 1986; Cook, 1987) and also block pyruvate carboxylase activity (White et al., 1988).

Diarylsulfonylureas are antitumor agents shown to have therapeutic activity against both rodent solid tumors and xenografts of human tumors in mice (Howbert et al., 1990; Mohamadi et al., 1992; Houghton and Houghton, 1996). Their mechanism of action is unknown but does not appear to be the result of nonselective destruction of actively dividing cell populations. In isolated liver mitochondria, both N-(5-indanylsulfonyl)-N'-(4-chlorophenyl)urea and its N-4-methyl analog uncouple oxidative phosphorylation (Thakar et al., 1991). At concentrations below 50 µM, both compounds exhibited a deleterious effect, causing damage to mitochondrial functions. These data confirm that diarylsulfonylureas may lower cellular ATP by uncoupling mitochondrial oxidative phosphorylation (Thakar et al., 1991).

Diarylsulfonylureas, such as N-(4-chlorophenyl)aminocarbonyl-2,3-dihydro-1-indene-5-sulfonamide (sulofenur) and N-(4-chlorophenyl)aminocarbonyl-4-methylbenzene sulfonamide (LY181984), have also been shown to be effective antitumor agents (Fig. 6) (Stagg and Diasio, 1990). Mitochondria have been shown to accumulate sulofenur and therefore may be targets of drug action. Many of the diarylsulfonylureas that were effective antitumor agents in animal models were also uncouplers of mitochondrial oxidative phosphorylation (Rush et al., 1992). The mechanism of uncoupling appeared to be related to a dissociable hydrogen ion, inasmuch as these molecules had pK_a values that ranged from 6.0 to 6.2 and were hydrophobic. However, the mechanism of antitumor activity does not appear to be the result of uncoupling, because no correlation was evident between the inhibition of cell growth and the uncoupling action of a variety of active and inactive antitumor diarylsulfonylureas.

	VII. The Mitochondrial Benzodiazepine Receptor

Top Previous Next References

Benzodiazepines are among the most widely prescribed drugs due to their pharmacological actions in relieving anxiety, and as anticonvulsants, muscle relaxants, or sedative hypnotics. These effects are mediated in the central nervous system through postsynaptic plasma membrane GABA_A receptors that are -aminobutyric acid-gated chloride channels. In addition to these central-type benzodiazepine receptors, binding sites were also identified in peripheral tissues, and this second class of sites was termed the peripheral benzodiazepine receptor (PBR). The PBR first characterized by Braestrup and coworkers (Braestrup and Squires, 1977; Braestrup et al., 1977) is present in peripheral tissues such as adrenal glands, kidney, and heart, as well as in the brain. The density of the PBR is the highest in endocrine tissues such as adrenal gland, testis, ovary, uterus, and placenta. PBR is also abundant in kidney, heart, and platelets, but densities in these tissues are approximately five times lower than that in adrenal gland. The PBR has been localized in mitochondrial membranes, and nonmitochondrial localizations have been observed in heart, liver, and testis. Additionally, the PBR was found in mature erythrocytes, which lack mitochondria. The properties and role of PBR were described in several review papers (McEnery, 1992; Parola et al., 1993; Gavish et al., 1999).

There is some terminological inconsistency concerning the mitochondrial PBR. Some authors (e.g., Tatton and Olanow, 1999) use this term to describe an 18-kDa polypeptide with high affinity toward benzodiazepine derivatives and isoquinoline carboxamides. This peptide has been isolated from rat adrenal gland and characterized by Antkiewicz-Michaluk et al., (1988a,b). However, other authors (e.g., McEnery et al., 1992; Gavish et al., 1999) consider the isoquinoline-binding protein (IBP) as one of at least three subunits of the mitochondrial PBR. The other two components are the mitochondrial pore protein, porin, also known as voltage-dependent anion channel (VDAC), with a molecular mass of 32,000; and the adenine nucleotide translocase, with a molecular mass of 30,000. The reactivity of the 18-kDa polypetide toward benzodiazepines usually requires the interaction of all three subunits.

IBP is a protein with five transmembrane domains usually associated with the mitochondrial outer membrane. The cDNA for the 850-nucleotide IBP mRNA has been cloned from a number of species. A detailed description of the IBP gene was recently reviewed (Gavish et al., 1999).

The PBR complex is located in the contact sites between the outer and the inner mitochondrial membranes. Its subunit composition roughly coincides with that of the mitochondrial permeability transition pore (Brustovetsky and Klingenberg, 1996; Beutner et al., 1996; Halestrap et al., 1997b) that opens under specific conditions and enables unselective passage of molecules of up to 1.5 kDa between the mitochondrial matrix and the cytoplasm (Bernardi et al., 1994; Zoratti and Szabó, 1995). Recently, much attention has been directed toward this pore, also called the "mitochondrial megachannel", because of its postulated role in events leading to programmed cell death (apoptosis) in multicellular organisms (Kroemer et al., 1997, 1998; Tatton and Olanow, 1999). The role of the 18-kDa IBP in the opening/closing transition of the pore and in its function in eliciting apoptosis is not clear. On the other hand, other components, such as cyclophillin D, have been described as obligatory functional elements of the permeability transition pore (Halestrap et al., 1997a) (for more information see Section III. and Fig. 3).

Recently, using a cytoplasmic domain of IBP as a bait in the yeast two-hybrid system, a new protein that specifically interacts with IBP has been cloned (Galigue et al., 1999). This protein, named PRAX-1 (peripheral benzodiazepine receptor-associated protein 1), exhibits several domains involved in protein-protein interactions such as three proline-rich domains and three leucine-zipper motifs (Galigue et al., 1999).

In contrast to the central benzodiazepine receptor, PBR exhibits nanomolar affinity to benzodiazepine Ro5-4864 (4'-chlorodiazepam) and the isoquinoline carboxamide derivative PK11195, and low affinity to benzodiazepine clonazepam. Isoquinoline carboxamide derivatives, such as PK11195, bind specifically to the 18-kDa IBP subunit, whereas PBR-specific benzodiazepine ligands, such as Ro5-4864, bind to a site consisting of porin, as well as adenine nucleotide translocase and the 18-kDa IBP subunit (Garnier et al., 1994). Other ligands for PBR, such as 2-aryl-3-indoleacetamide (Romeo et al., 1992) and N-(2,5-dimethoxybenzyl)-N-(5-fluoro-2-phenoxyphenyl)acetamide (Chaki et al., 1999), were also described.

Benzodiazepine Ro5-4864 and isoquinoline carboxamide PK11195 are the two most widely used PBR ligands. Based on the entropy-driven and enthalpy-driven nature of ligand-receptor interactions, PK11195 has been classified as an antagonist and Ro5-4864 as an agonist (Le Fur et al., 1983). The functional significance of such classification was recently confirmed in studies on the antiapoptotic activities of these PBR ligands (Bono et al., 1999).

Protoporphyrin IX binds to PBR with nanomolar affinity and has been suggested to be an endogenous ligand for PBR (Snyder et al., 1987). During heme biosynthesis, cytosolic coproporphyrinogen III traverses the mitochondrial outer membrane and is converted via protoporphyrin IX to heme, which is subsequently exported from mitochondria. This suggests an involvement of PBR in mitochondrial heme synthesis (Verma et al., 1987; Woods and Williams, 1996). Another candidate for endogenous ligand for PBR is a 104-amino acid neuropeptide known as diazapam binding inhibitor (Corda et al., 1984). Also, a 16-kDa protein called anthralin inhibits the specific binding of Ro5-4864 to PBR (Mantione et al., 1988).

PBR has been implicated in several mitochondrial functions, but its exact physiological role is still unclear. In vitro studies using isolated cells, mitochondria, and submitochondrial fractions demonstrated that PBR is present in steroid-synthesizing cells and is involved in this kind of cell in the regulation of cholesterol transport from the outer to the inner mitochondrial membrane, known to be the rate-determining step in steroid biosynthesis (Culty et al., 1999). The postulated role of PBR was recently reviewed in detail (Gavish et al., 1999). PBR agonist Ro5-4864 was found to strongly protect against apoptosis induced by tumor necrosis factor- in human lymphoblastoid cell line U937 (Bono et al., 1999). The potent antiapoptotic effect might represent a major function for this receptor as demonstrated by the lack of antiapoptotic activity of Ro5-4864 in wild-type Jurkat cells (lacking the PBR receptor) and the reappearance of this effect in PBR-transfected cells (Bono et al., 1999). Additionally, the blockade of the antiapoptotic effect of PBR agonist by selective PBR antagonist PK11195 was observed (Bono et al., 1999).

	VIII. Immunosuppressant Drugs and Mitochondria

Top Previous Next References

Cyclosporin A has potent immunosuppressive properties due to its ability to block the transcription of cytokine genes in activated T cells (Rovira et al., 2000). It is well established that CsA forms a complex with cyclophylin D, inhibiting the peptidyl-prolyl cis-trans isomerase activity of this protein. Additionally, the CsA-cyclophilin complex inhibits the activity of calcineurin (Rusnak and Mertz, 2000). Calcineurin is a Ca²⁺- and calmodulin-dependent serine/threonine phosphatase (protein phosphatase 2B) that regulates nuclear translocation and subsequent activation of the nuclear factor of activated T cells known as NFAT transcription factor. Prevention of NFAT dephosphorylation, which is an important step for its translocation to the nucleus, blocks cytokine production. In addition to the calcineurin/NFAT pathway, recent studies indicate that CsA also blocks the activation of stress-activated protein kinase JNK (c-jun NH₂-terminal kinase) and the mitogen-activated protein kinase p38 signaling pathway triggered by antigen recognition (for review see Matsuda and Koyasu, 2000). Despite this beneficial role of CsA in organ transplantation, CsA has significant side effects such as hypertension, and renal and muscle toxicity. Probably, these effects of CsA are related to ROS generation (Buetler et al., 2000).

CsA blocks the opening of the PTP (see Section III.A.) because of its high affinity to cyclophilin D. This effect is independent of the inhibition of calcineurin and the immunosuppressive action of CsA, because the CsA analog N-methyl-Val-4-cyclosporin lacks immunosuppressive properties and is still able to block PTP (Zamzami et al., 1996; Hortelano et al., 1997; Matsumoto et al., 1999; Vergun et al., 1999). The blocking of PTP opening is most likely the mechanism underlying the protective action of CsA against ischemic and ischemic/reperfusion injuries. Such beneficial effects have recently been described in ischemic and traumatic brain injury (TBI) in experimental animals (Scheff and Sullivan, 1999; Sullivan et al., 1999, 2000; Albensi et al., 2000; Li et al., 2000). For example, animals subjected to forebrain ischemia for 30 min exhibited extensive neuronal necrosis and failed to survive, whereas injection of CsA (in combination with an intracerebral lesion to open the blood-brain barrier) prolonged their survival time, ameliorating brain damage and preventing secondary mitochondrial dysfunction (Li et al., 2000).

Although TBI often results in impaired learning and memory functions, the underlying mechanisms are unknown, and there are currently no treatments that can preserve such functions. Recently, the plasticity at CA3-CA1 synapses in hippocampal slices from rats subjected to controlled cortical impact TBI was studied (Albensi et al., 2000). Long-term potentia