Depression is a serious mental disorder manifested by depressed mood, pessimistic thoughts, feelings of worthlessness, feelings of guilt, tearfulness, reduced or increased sleep, appetite loss or appetite disturbance, weight loss or weight gain, social restlessness, loss of interest, difficulty concentrating. Mania is characterized by abnormally elevated or irritable mood, arousal, and/or energy levels. Bipolar disorder features intermittent episodes of mania or hypomania and depressive episodes; rapid cycling, mixed states, and psychotic symptoms occuring in some cases. Depression and mania are thought to be heterogeneous illnesses that can result from dysfunction of several neurotransmitters or metabolic systems.
The predisposition to the disease is determined by genetic, psychosocial and biological factors; individual sensitivity to depressogenic effects during stressful life events is also a contributing factor. Pathophysiology of mood disorders is not sufficiently elucidated and about 1/3 of patients do not response to pharmacotherapy sufficiently. The exact molecular site and the primary cause of signal transduction disturbance associated with the symptoms of depression or mania are still unknown.
Recently, attention in the research of biological basis of mood disorders has been devoted to an overlapping set of molecular and cellular mechanisms of mood disorders, antidepressant response, neuroplasticity, and chronic stress , e.g. to changes in neuroprogression, inflammatory and cell-mediated immune response, antioxidant capacity, oxidative and nitrosative stress, and mitochondrial functions . Therefore, changes in the activities of compounds of these intracellular signalling pathways are studied with the aim of discovering new biological markers of mood disorders or predictors of response to antidepressant treatment [3-4]. Mitochondrial dysfunctions are assuming an increasingly important role in hypotheses of mood disorders, bipolar disorder mainly. Recently discussed biological hypotheses of mood disorders include the neurotrophic and neuroplasticity hypothesis of depression [1,5-8] and the mitochondrial hypothesis [9-11].
It is well-known that mitochondria strongly affect many intracellular processes coupled to signal transduction, neuron survival and plasticity. Impaired mitochondrial functions manifest themselves in various ways, they may be related to many psychiatric and neurodegenerative diseases, including bipolar disorder, major depressive disorder, schizophrenia, psychosis and anxiety [12-16]. Impaired functions of mitochondria can be assessed both in isolated mitochondria and in intact or permeabilized cells. Better insight into molecular mechanisms of cellular respiration, control of oxidative phosphorylation (OXPHOS) and effects of antidepressants and mood stabilizers on these processes is likely to lead to a better understanding of pathophysiology of neuropsychiatric disorders.
Mitochondria are small cellular structures consisting of an outer and inner membrane, an intermembrane space and an intracellular matrix. The outer membrane covers the organelle, the inner membrane folds and forms cristae. This settlement extends the surface and enables plenty of chemical reactions. In the mitochondrial matrix, the enzymes of the tricarboxylic acid cycle (TCA, also called citric acid cycle or Krebs cycle) are localized. It is the central pathway of metabolism; its main function is oxidation of acetyl-CoA derived from carbohydrates, amino acids and fatty acids (FAs). The TCA is organized into a supramolecular complex that enables interaction with mitochondrial membranes and the electron transport chain (ETC) in OXPHOS . Most of the TCA enzymes provide other additional “moonlighting” functions, e.g. they stabilize the mitochondrial DNA (mtDNA) or are associated with mitochondrial RNA (mtRNA) translation, oxidative stress, iron metabolism and tumour suppression .
In addition to their crucial role in generation of adenosine-5’- triphosphate (ATP), mitochondria are involved in other important processes, such as regulation of free radicals, neurotransmitters, calcium, and apoptosis. They are also involved in neuronal development - synaptogenesis, synaptic development and plasticity. Impaired function of mitochondria leads to impaired bioenergetics, decrease of ATP production, impaired calcium homeostasis, increased production of free radicals and oxidative stress [19-20]. Furthermore, monoamine oxidase (MAO), the enzyme responsible for the metabolism of monoamine neurotransmitters, is localized in the outer mitochondrial membrane.
Mitochondrial proteins are encoded by both nuclear and mitochondrial DNA. All 13 polypeptides encoded by mtDNA form subunits of respiratory chain complexes I, III, IV and V [21-22]. Furthermore, the mitochondrial genome encodes transfer RNA (tRNA) and ribosomal RNA (rRNA) used for RNA translation . Complex II is encoded only by nuclear DNA (nDNA). OXPHOS is under the control of the nuclear genome as well as the mitochondrial genome, which is only maternally inherited. Nevertheless, the dominant role in the regulation of mitochondrial activity has a nucleus; nuclear-encoded transcript factors control the activity of the mitochondrial genome and coordinate the expression of nuclear and mitochondrial genes to mitochondrial proteins [23-24].
Genetic defects or stress can cause mitochondrial dysfunctions, which leads to increased oxidative stress and/or altered calcium homeostasis . An excess of glutamate in the synapse  leads to an excess of cytosolic calcium, which produces overactivity of calcium-dependent enzymes and an overload of mitochondria by calcium; it leads to cytoskeletal degradation, protein malformation, decrease of ATP production, and increase of oxygen radical generation. These processes can lead to atrophy or death of neurons [27-28]. Different stimuli, such as hypoxia-ischemia, seizure and hypoglycemia, all activate this pathway. Thus, enhancing mitochondrial function may represent a critical component for the optimal treatment of stress-related diseases .
Eukaryotes synthetize ATP mainly by glycolysis in the cytosol and by OXPHOS in the mitochondria; i.e. the majority of cellular ATP is generated by glycolytic degradation of glucose to pyruvate in cytosol followed by aerobic cellular respiration. When pyruvate is converted to acetyl coenzyme A (acetyl-CoA), acetyl-CoA enters the TCA cycle and the result of this process is ATP production by OXPHOS in mitochondria . OXPHOS yields about 17 times more ATP than glycolysis. Therefore, it is considered as the main energy source and a key element of bioenergetics [30-31]. Integration of main metabolic pathways coupled to OXPHOS is illustrated in Figure 1.
The highest number of mitochondria is present in organs demanding the most energy - brain, liver and muscles. Neurons usually utilize glucose as a source of energy. Since the brain stores only a very small amount of glycogen, it needs a steady supply of glucose. Neurons are known to have a lower glycolytic rate than astrocytes and when stressed they are unable to upregulate glycolysis. Following inhibition of mitochondrial respiration, neurons die rapidly, whereas astrocytes utilize glycolytically generated ATP. Glucose metabolism in neurons is directed mainly to the pentose phosphate pathway, leading to regeneration of reduced glutathione, which probably supports antioxidant controlled neuron survival . The regulative processes of OXPHOS are tightly related to reactive oxygen species (ROS) production, integrity of mitochondrial membranes, apoptosis, and intramitochondrial Ca2+ levels. Although this is known, the control mechanisms have not yet been sufficiently investigated.
2.1. Physiology of oxidative phosphorylation
The respiratory chain is localized in cristae, structures formed by the inner mitochondrial membrane and extending to the surface . ETC consists of complexes with supramolecular organization, where mitochondrial proton pumps (complexes I, III and IV) transport protons and generate a proton gradient [31,35]. Continuously, electrons are transported to complex III and finally complex IV enables the conversion of O2 to H2O. Most of the ATP synthesis comes from the electrochemical gradient across the inner membranes of mitochondria by ATP synthase (complex V). The CoQ cofactor is responsible for transferring electrons from complexes I and II to complex III; the second important cofactor is cytochrome
Energy saved in ATP is used in synaptic ion homeostasis and phosphorylation reactions. ATP is essential for the excitability and survival of neurons, OXPHOS is involved in synaptic signalling and is related to changes of neuronal structure and function. Therefore, mitochondria are included in neurotransmitter exocytosis, in recovery, and in ion homeostasis, and in presynaptic nerve terminals.
Oxidative phosphorylation enzymes and MAO are key mitochondrial enzymes studied in molecular psychiatry.
2.1.1. Oxidative phosphorylation enzymes
2.1.2. Monoamine oxidase
Monoamine metabolism by MAO involves oxidative deamination to corresponding aldehyde and free amine. Catalysis in MAO depends on the transfer of electrons to FAD, and mechanism-based inhibitors, such as the irreversible antidepressants, modify flavin . The aldehyde is rapidly metabolized by aldehyde dehydrogenase to acidic metabolites. Metabolism of monoamines by MAO is a major source of hydrogen peroxide (H2O2) in the brain. Normally the H2O2 is then inactivated by glutathione peroxidase but it can be converted, chemically, by Fe2+ ions (Fenton reaction) into the highly reactive hydroxyl radical. This radical has widespread deleterious effects which can cause neuronal damage and death and may account for associated health-related problems [51,54].
MAOs have important role in brain development and function, and MAO inhibitors (MAOIs) have a range of potential therapeutic uses . Generally, selective inhibitors of MAO-A and nonselective MAOIs seem to be effective in the treatment of patients with depression, panic disorder, and other anxiety disorders . It is supposed that MAO-B inhibition may slow the course of various neurodegenerative disorders; so, selective inhibitors of MAO-B may be efficacious in treating of Parkinson’s disease  and possibly Alzheimer's disease . MAO-B is the sole type in human platelets and the amino acid sequences of MAO-B in both platelets and brain are identical ; thus, platelet MAO can be adopted as a useful surrogate model for the study of aspects of central neuronal function related to monoaminergic neurotransmission .
2.2. Regulation of OXPHOS
There are five levels of OXPHOS regulation: 1. direct modulation of ETC kinetic parameters, 2. regulation of intrinsic efficiency of OXPHOS (by changes in proton conductance, in the P/O ratio or in the channelling of ECT intermediate substrates), 3. mitochondrial network dynamics (fusion, fission, motility, membrane lipid composition, swelling), 4. mitochondrial biogenesis and degradation, 5. cellular and mitochondrial microenvironment .
OXPHOS efficiency is dependent on delivery of reducing equivalents into ETC and on activities of participating enzymes or enzyme complexes. The optimal efficiency and flow ratios are determined by control of complex I (reflects integrated cellular pathway) and complex II (TCA cycle precedes) . Depletion of TCA cycle intermediates plays an important role in the OXPHOS flux control. In respirometry assays, supplies of complex I as well as complex II are required. Convergent electron input and reconstitution of the TCA cycle are needed to achieve maximal respiration . It is controlled also by the availability of adenosine 5´-diphosphate (ADP) for the adenine nucleotide transporter in the inner mitochondrial membrane .
Complex I is suggested to be responsible for adaptive changes and physiological set up of OXPHOS efficiency . The stoichiometric efficiency of OXPHOS is defined by the P/O ratio, or the amount of inorganic phosphate (Pi) incorporated into ATP per amount of consumed oxygen. P/O ratio was analysed in rat brain, liver and heart mitochondria. There were found tissue-specific differences and dependency of the P/O ratio on the respiratory rates with complex I, but not with complex II substrates . Metabolic control analysis, which compared ETC activities and oxygen consumption rates, determined the role of complex I in rat brain synaptosomes. Results of the study suggest complex I as rate-limiting for oxygen consumption and responsible for high level of control over mitochondrial bioenergetics .
As mentioned above, mitochondria exhibit transmembrane potential across the inner membrane that is necessary for OXPHOS. Protons are transported outwardly and create proton motive force (Δ
ATP production is controlled by different mechanisms, depending on energy demands, thermogenesis, etc. . First mechanism of OXPHOS control has been called as “respiratory control”, and is based on feedback mechanisms controlling the rate of ATP synthesis, first of all by Δ
Recently, primary control has been implemented by secondary control mechanisms that are Δ
In the TCA cycle, glycerophosphate dehydrogenase, pyruvate dehydrogenase, isocitrate dehydrogenase, and α-ketoglutarate dehydrogenase are influenced by Ca2+ levels and their phosphorylation lead to increased ATP production, production of glycogen, and glucose oxidation . Reversible phosphorylation of pyruvate dehydrogenase complex mediated by calcium partly regulates the supply of reducing equivalents (NADH/NAD+ ratio). Activation of the TCA cycle enhances the NADH production that triggers movement of electrons down complexes I through to complex IV by initially donating of complex I .
|Low||Low-endogenous||Slow||Phosphate acceptor||Initial activity of the sample|
|High||Approaching zero||Slow||Substrate||1. Exhaustion of endogenous substrate utilized in OXPHOS of ADP|
2. Residual oxygen consumption (ROX)
|1. OXPHOS capacity at saturating ADP (State P)|
2. Electron transfer system capacity at optimum uncoupler concentration (State 3u)
|Low||High||Slow||Phosphate acceptor||1. Exhaustion of added ADP|
2. LEAK respiration (resting state when oxygen flux is maintained mainly to compensate for the proton leak after inhibition of ATP synthesis) (State 4o, L)
2. Antimycin A treatment
Regulation of complex I and COX subunits via specific protein kinases and protein phosphatases was observed. cAMP-dependent protein kinase catalyses phosphorylation of complex I subunit and stimulates ETC . At low Ca2+ levels, protein phosphatase dephosphorylates and inactivates complex I. It is presumed that COX is regulated by allosteric inhibition of ATP at high ATP/ADP ratios . Extramitochondrial ATP/ADP ratios regulate COX activity by binding to the cytosolic subunit of COX, whereas high mitochondrial ATP/ADP ratios cause exchange of ATP by ADP at COX and induce allosteric inhibition . Similarly, increased intracellular Ca2+ levels are suggested to activate mitochondrial phosphatase, which dephosphorylates COX and turns off the allosteric inhibition . This respiratory control by phosphorylated enzyme is assumed to keep the Δ
Thyroid hormone, mainly triiodothyronine (T3) and diiodothyronine (T2), has important effects on mitochondrial energetics and mitochondrial genome . Mechanism of allosteric inhibition of COX has been closely linked to regulation by thyroid hormones. 3,5-diiodothyronine (T2) mediates short term effects of thyroid hormones and increases immediately basal metabolic rate. T2 is formed by intracellular deiodination of T3 and binds to specific T2 binding sites, which were identified in the inner mitochondrial membrane . This binding to subunit Va of COX abolishes the allosteric inhibition of respiration by ATP  that could result in partial uncoupling of OXPHOS via increased Δ
2.3. Proton permeability of membranes
OXPHOS in cells is not fully efficient. Decrease of the proton gradient across the inner mitochondrial membrane by “proton leak” causes uncoupling of fuel oxidation from ATP generation, and some energy is lost as heat. The mechanism of the basal proton conductance of mitochondria (insensitive to known activators and inhibitors) is not understood. There is correlation between mitochondrial proton conductance and composition of inner membrane: phospholipid fatty acyl polyunsaturation correlates positively and monounsaturation correlates negatively with proton conductance .
Uncoupling proteins (UCPs) and adenine nucleotide translocator (ANT) are two types of mitochondrial carrier, which cause inhibitor-sensitive inducible proton conductance. UCPs themselves do not contribute to the basal proton conductance of mitochondria; however, they are important metabolic regulators in permitting fat oxidation and in attenuating free radical production . The amount of ANT present in the mitochondrial inner membrane strongly affects the basal proton conductance of the membrane and suggests that ANT is a major catalyst of the basal FA-independent proton leak in mitochondria .
2.3.1. Fatty acids
Long-chain fatty acids (FAs) are weak acids that can cross the membrane in both protonated and deprotonated forms. Effects of FAs are interrelated to 1. increase uncoupling, 2. increase ROS production, 3. opening mitochondrial permeability transition pores (MPTP) . Further, they can modulate effects of thyroid hormones as well as sex steroid hormones . FAs can act as like classic OXPHOS uncouplers with protonophoric action on the inner mitochondrial membrane and/or interactions of FAs with ADP carrier, COX and ATP synthase are presumed . Recent study suggests that FAs are not only inducers of uncoupling, but they also regulate this process. It supposes that transport of FA anions participates in both ADP/ATP antiport and aspartate/glutamate antiport, at the same time . On the other hand, studies using lipid membranes suppose that FAs are capable of spontaneous flip-flop . Since FAs move across the membrane spontaneously and rapidly, no protein transporters are necessary. Further, coupling/uncoupling effects depend on their concentrations pH gradient across the membranes [94-95].
2.3.2. Uncoupling proteins
Uncoupling diverts a significant proportion of energy to thermogenesis. UCPs are mitochondrial carriers catalysing a regulated proton leak across the inner membrane [96-97]. There are five types of UCP in mammals. UCP1 is presented exclusively in the inner mitochondrial membrane of brown adipose tissue, and its main function is to catalyse adaptive thermogenesis . It can be stimulated by FA and has synergic action of norepinephrine and thyroid hormones [49,99]. Concentrations of UCP2 and UCP3 in tissues are much lower than of UCP1, and their functions are not exactly known. They probably minimally contribute to basal metabolic rate, control of adaptive thermogenesis, preventive action against oxidative stress and ROS control, control of cellular energy balance, regulation of Ca2+ homeostasis, regulation of FA oxidation and ATP synthesis [100-103]. UCP2, UCP4 and UCP5 are present in the central nervous system (CNS); they have been suggested to have effects protecting neurons from the Ca2+ overload and/or oxidative stress [104-105].
UCP activities can be positively or negatively regulated by different factors. UCP are stimulated by FA and by ROS, generated by as a side reaction between CoQ and oxygen . UCP mediate the FA dependent proton influx that leads to uncoupled ATP synthesis and heat production . It is supposed that UCP and FA decrease Δ
2.4. Reactive oxygen species production
Reduction of O2 to water by aerobic respiration is accompanied by reactive intermediate formation. Generally, complex I and complex III are considered as the major O2•- sources .
Complex I releases O2•- to matrix, complex III can release O2•- to both sides of the inner mitochondrial membrane . Additionally, other ROS sources, e.g. MAO, present in the outer mitochondrial membrane, and α-ketoglutarate dehydrogenase (α-KGDH), the TCA cycle enzyme complex, are able to generate H2O2. MAO catalyses the oxidative deamination of biogenic and xenobiotic monoamines and increases the amount of ROS in mitochondria. H2O2 production by α-KGDH is dependent on NADH/NAD+ ratio. Higher NADH leads to higher H2O2 production, therefore, α-KGDH could significantly contribute to oxidative stress in mitochondria .
Physiologically generated H2O2 and O2•- from ETC are dependent on the magnitude of Δ
Complex I is considered to be the primary source of ROS in brain under physiological conditions, as well as in pathological processes (e.g. neurodegenerative disorders). ROS seem to be the key factors in brain aging processes and mitochondrial respiration with ROS production significantly contributes to functional changes in brain during aging. Study in isolated rat mitochondria found significantly increased H2O2 production and 30 % reduction of complex I activity in aged rats . Defective mitochondria release large amounts of ROS, similarly, decline of antioxidative enzyme activities (e.g. in elderly) enhances ROS production . Negative results of ROS can affect respiratory chain: complexes I, III and IV seem to be the most affected, whereas function of complex II appears to be unchanged .
Mitochondrial dysfunctions may accompany the clinical picture of neuropsychiatric disorders and contribute to neural apoptosis ; mitochondria play a pivotal role in intrinsic pathway of apoptosis . Several interrelated mitochondrial pathways facilitate cell death: mitochondrial permeability transition (MPT) and the release of apoptotic cell death promoting factors, cytochrome
2.6. Specific inhibitors of complexes of ETC
Rotenone is a specific complex I inhibitor, thenoyltrifluoroacetone (TTFA) specifically inhibits complex II. Both substances induce O2•- production that may result to major ROS production [45,123-124]. Pyrrolnitrin inhibits both complex I as well as complex II. It affects electron transport among NADH, CoQ and succinate, whereas COX remains unaffected .
Complex III inhibitors antimycin, myxothiazol and stigmatellin differ in their mechanism of action. Antimycin A inhibits the transfer of electrons from cytochrome
Complex IV inhibitors KCN and sodium azide decrease COX activity . Azide specifically blocks crossover between cytochrome
During the oxidation of complex I substrates (pyruvate, malate, glutamate), rotenone inhibition did not increase H2O2; contrary, oxidation of complex I and II substrates in the presence of antimycin A increased H2O2. Both myxothiazol and stigmatellin inhibited O2•- production and/or should inhibit the effect of antimycin [126,131]. The maximum of O2•- production has been observed in human skin fibroblasts with the prolonged treatment of rotenone, but not with antimycin A . Interestingly, rotenone prevented antimycin A to induce ROS production in complex I, but not in complex II . Qo side of complex III was found as the source of increased O2•- after transient exposure to hydrogen peroxide . KCN and sodium azide increase ROS formation . Oligomycin induces hyperpolarization of inner mitochondrial membrane and can increase O2•- levels .
2.7. Mitochondria and neuroplasticity
Mitochondrial distribution and activity are key factors in neuronal morphogenesis - synaptogenesis, developmental and synaptic plasticity and axogenesis. During the development, neuronal stem cells proliferate and differentiate into neurons; subsequently axons and dendrites form synapses [134-135]. The role of mitochondria in neuroplasticity is illustrated in Figure 3 . Due to ATP production and importance of mitochondria in synaptic ion homeostasis and phosphorylation reactions, mitochondria would be accumulated at sites where ATP consumption and Ca2+ concentration are higher. It was reported that mitochondria are more abundant in the regions of growing axons than in the non-growing axons. Mitochondrial net movement is anterograde in growing axons and is retrograde in non-growing axons. Shortly before axogenesis mitochondria congregate at the base of the neurite that is destined to become the axon. Nerve growth factor (NGF) was found as one of the signals inducing accumulation of mitochondria in the active growing cone . Interestingly, when the ATP production is impaired and cells provide alternative source of energy, axogenesis is abolished although growth of dendrites remains relatively unaffected .
There are changes in mitochondrial energy metabolism occurring in brain cells during CNS development. During embryonic and early postnatal development fats are primarily used, later on, glucose becomes as fuel. This fact supports the role of mitochondria in biochemical requirements of highly proliferative neuronal stem cells and post-mitotic neurons. During neuronal differentiation the number of mitochondria per cell increases, but the velocity at which individual mitochondria move decreases as neurite outgrowth slows and synaptogenesis occurs [20,137].
It was demonstrated that neuronal activity is influenced by the mitochondrial functions, defective trafficking and dysfunction of mitochondria from axon terminals is implicated in the pathogenesis of axonal degeneration [138-140]. In addition, dendritic mitochondria are essential in the morphogenesis and plasticity of spines and synapses . Recent findings suggest roles for mitochondria as mediators of at least some effects of glutamate and BDNF on synaptic plasticity . BDNF promotes synaptic plasticity, in part, by enhancing mitochondrial energy production. It increases glucose utilization and increases mitochondrial respiratory coupling at complex [62,142].
Mitochondria are dynamic organelles; their function is modulated by fission, fusion and moving within the axons and dendrites . Their structure, functions and properties differ in axons and dendrites [141,143]. Transport and positioning of mitochondria are essential for neuronal homeostasis and the mitochondrial movement is a part of regulation by intracellular signals.
3. Advances in biological hypotheses of mood disorders
Findings about intracellular processes associated with mood disorders and long-term effects of antidepressants demonstrate an important role of signalling pathways primarily regulated by monoamine neurotransmitters; this was settled as the basis of many biochemical hypotheses [144-145]. While dysfunctions within monoaminergic neurotransmitter systems are likely to play an important role in pathophysiology of mood disorders, it probably represents the downstream effects of more primary abnormalities in signal transduction. Thus, new theories about the pathophysiology of depression and the action of antidepressant treatment proposes that mood disorders are caused by structural or functional changes in particular molecules and signalling pathways in the brain, and that antidepressants function by counteracting these molecular changes. It is supposed that structural and functional brain abnormalities in patients with depressive disorder may be associated with low levels of brain-derived neurotrophic factor (BDNF), abnormal function of hypothalamic-pituitary-adrenal (HPA) axis, glutamatergic toxicity, activation of inflammatory and cell-mediated immune response, decreased antioxidant capacity and increased oxidative and nitrosative stress, disturbed chronobiological rythms, and mitochondrial dysfunctions [2,146-148].
Research on the biological basis of mood disorders emphasises the changes of neural networks and synaptic plasticity. Evidence exists for impairment of neuroplasticity in major depression. Chronic stress is known to contribute both to development of major depression in vulnerable persons and to reduction of synaptic plasticity, induction of structural changes in dendrites, and impairment of neurogenesis . Mitochondria may be primary regulators of these processes, as they regulate not only neuronal survival and death, but also plasticity. There is mounting evidence for the role of mitochondrial dysfunction in the pathophysiology and treatment of bipolar disorder .
3.1. Monoamine hypothesis
Discovery of the first effective antidepressants, MAOIs and tricyclic antidepressants, implied hypothesis about significant role for the biogenic amine, particularly NE and 5-HT in the ethiopathogenesis of affective disorders. Classic monoamine hypothesis is an early milestone in the field of depression. It proposed that depression might be produced by a 5-HT or NE deficiency at functionally important receptor sites in the brain, i.e. that brain monoamine systems have a primary direct role in depression [149-150]. Soon it became evident that the monoamine hypothesis in its original form could not explain all of the effects of antidepressants [151-152]. In order to test this hypothesis, a series of studies was conducted to evaluate effects of monoamine depletion on depressive symptoms in depressed patients and in healthy controls. Relapse to 5-HT depletion or to catecholamine depletion was found to be specific to the type of antidepressant treatment and type of depletion. 5-HT or NE/dopamine depletion did not decrease mood in healthy controls and slightly lowered mood in healthy controls with a family history of major depressive disorder. In drug-free patients with major depressive disorder in remission, a moderate mood decrease was found for acute tryptophan depletion only. However, acute tryptophan depletion induced relapse in patients in remission who used serotonergic antidepressants . Depletion studies failed to demonstrate a causal relation between 5-HT and NE with depressive disorder [154-155]. The effects of acute tryptophan depletion on cognition in non-vulnerable participants are independent of mood changes . Even simultaneous disruption of 5-HT and catecholamine systems didn’t significantly alter mood in unmedicated depressed subjects . These findings forced a major revision of the classic monoamine hypothesis of depression. According to this revised monoamine theory of depression [148,157] monoamine systems are only modulating other brain neurobiological systems that have more primary role in depression.
3.2. Neurotrophic hypothesis
The neurotrophic hypothesis of depression [5-6,8] supposed that vulnerability to depression can arise as a result of neuronal damage, e.g. after chronic stress, long-term increased levels of glucocorticoids, hypoglycemia, ischemia, effects of neurotoxins or certain viral infections, etc. The therapeutic effects of antidepressants consist in the increased function of the noradrenergic or serotonergic system, leading to increased activity of transcription factor CREB (cAMP response element binding protein), higher expression of neurotrophin BDNF and its receptor trkB, and consequently to increased neuronal plasticity and resumption of cellular functions.
According to neurogenic hypothesis [158-159], depression may develop due to the decreased neurogenesis in hippocampus, and antidepressants takes effect through the stimulation of neurogenesis. Hypothesis of cellular plasticity  relate the neurotrophic and the neurogenic hypothesis to the statement that depression can be generally caused by damaged cellular plasticity leading to inadequate relations between structure and function. Molecular mechanisms leading to a disturbance of neuroplasticity are not known. The bioenergetic and neurochemical model of bipolar disorder attempts to identify these mechanisms and focuses attention on mitochondrial dysfunctions [9,161].
3.3. Inflammatory and neurodegenerative hypothesis
The central nervous system, endocrine and immune systems use neurotransmitters, cytokines and hormones to communicate among them . Now there is evidence that the activation of the immune system is associated with the symptoms of depression [163-164]. The inflammatory and neurodegenerative hypothesis of depression  supposes that depression is associated with both inflammatory processes, as well as with neurodegeneration and reduced neurogenesis. According to this hypothesis, enhanced neurodegeneration and impaired neurogenesis in depression are caused by inflammatory processes, related to the production of oxidative and nitrosative stress, tryptophan catabolites along the indoleamine-2,3-dioxygenase pathway, proinflammatory cytokines and lowered ω-3 polyunsaturated fatty acid status. Anti-inflammatory compounds should be able to counteract at least partly the enhanced neurodegeneration and decreased neurogenesis.
3.4. Mitochondrial hypothesis
Mitochondrial dysfunctions (leading to decreased ATP production, oxidative stress, and induction of apoptosis) occur in the early stages of different neurodegenerative diseases, associated often with mood disorders.
The role of mitochondrial dysfunction during bipolar disorder is supported both by observation of the changes of brain metabolism and by effects of mood stabilizers (lithium and valproate) on mitochondrial functions. Metabolic changes in brain were observed in bipolar disorder by magnetic resonance spectroscopy (MRS). It suggests the presumptions that mitochondrial dysfunctions include impaired OXPHOS, final shift to glycolytic production of energy, general decrease of energy (decreased ATP production), changed concentrations of phosphomonoesters and changed lipid metabolism .
mtDNA mutations in the brain, associations of mtDNA polymorphisms and bipolar disorder and changes in gene expression related to mitochondria in the brain were observed [10,166]. Mitochondrial dysfunction hypothesis of bipolar disorder is based on these observations. According to this hypothesis, mtDNA polymorphisms/mutations or mtRNA deletions caused by nuclear gene mutations can cause mitochondrial dysregulation of calcium leading to symptoms of bipolar disorder [10,161,167]. Mitochondrial hypothesis corresponds to, above mentioned, neurotrophic hypothesis because of an important role of calcium signalling pathway in synaptic plasticity regulation.
3.5. Biological markers of mood disorders
Biological markers are defined as a characteristic that is objectively measured and evaluated as an indicator of normal biologic processes, pathogenic processes, or pharmacologic responses to a therapeutic intervention. In medicine, a biomarker is an indicator of a particular disease state or a particular state of an organism.
Identification of biologic markers of mood disorders and factors capable of predicting the response to treatment with antidepressants has not been sufficiently successful [3,168-169]. In accordance to actual neurochemical hypotheses of mood disorders, biological markers have been primarily found at the level of neurotransmitter concentrations, their metabolites or precursors. Subsequently, attention was shifted to the receptor systems, and since the 1990´s, intracellular processes have become main interest. The chance to find sensitive and specific biological predictors of antidepressant treatment has been increased, because of introduction of new methods of molecular biology. These methods enable us better observation of cellular processes connected with the transduction of nervous signals in the brain. The choice of parameters, which should be studied as perspective biological markers of mood disorders, have been derived first of all from new findings of signalling pathways involved in neurotransmission and from above mentioned neurochemical hypotheses of mood disorders. From the view of intracellular processes, energetic metabolism, activities of PKC, CREB, BDNF, Bcl-2, glycogen synthase kinase-3, caspases or calcium could play a principal role in findings of biological markers of mood disorders. According to the complexity and connectivity of signalling pathways involved in etiopathogenesis of mood disorders, number of chosen parameters is not final.
4. Antidepressants, mood stabilizers and mitochondrial functions
Antidepressants are used mainly to alleviate mood disorders, such as major depression and dysthymia and anxiety disorders. Mood stabilizers are psychiatric medication used in treatment of mood disorders, which are characterized by intense and sustained mood shifts (e.g. bipolar disorder).
The antidepressant activity of the first generation of antidepressants, tricyclic antidepressants and MAOIs, was explained by their effects on availability of monoamine neurotransmitters. The next generations of antidepressants included selective serotonin reuptake inhibitors (SSRIs), norepinephrine reuptake inhibitors (NRIs), serotonin-norepinephrine reuptake inhibitors (SNRI), noradrenergic and specific serotonergic antidepressants (NaSSAs), norepinephrine-dopamine reuptake inhibitors (NDRIs), serotonin antagonist and reuptake inhibitors (SARIs), selective serotonin reuptake enhancer (SSRE), melatonergic agonists (MASSA), sigma receptor agonists etc. The therapeutic response to antidepressants occurs after long-term treatment; therefore, effects of antidepressants are linked to cellular adaptations including density and/or sensitivity of neurotransmitter receptors and transporters, regulation of signal transduction cascades, and changes in gene expression .
Most of mood stabilizers are anticonvulsants (valproate, carbamazepine, and lamotrigine), with an important exception of lithium, which is the oldest and the best known mood stabilizing drug. Some atypical antipsychotics (olanzapine, quetiapine, aripiprazole, risperidone, ziprasidone) have mood stabilizing effects, as well.
Although a wide range of pharmacologically different antidepressants and mood stabilizers is available, molecular mechanisms of their therapeutic effects haven’t yet been sufficiently clarified. Relatively little information is known about the association among therapeutic and/or adverse effects of drugs and mitochondrial enzyme activities. Incomplete data exist on the effect of pharmacologically selective antidepressants and mood stabilizers on MAO activity. Measurement of both mitochondrial respiration and membrane potential during action of appropriate endogenous and exogenous substances enables the identification of the primary sites of effectors and the distribution of control, allowing deeper quantitative analyses .
4.1. Inhibition of MAO
MAO inhibition is the best known direct action of some antidepressants on mitochondrial enzymes. The antidepressant effect of MAOIs has been established more than 50 years ago. Iproniazid became the first MAO inhibitor to be used successfully in the treatment of depression; it is an irreversible and nonselective MAO inhibitor . It is known to act as a pro-drug and can be converted into isopropyl hydrazine which binds covalently to MAO . Clorgyline is an irreversible inhibitor preferential for MAO-A, structurally related to pargyline (MAO-B inhibitor). It has antidepressant activity, and may potentially be useful in the treatment of Parkinson's disease. Selegiline (l-deprenyl) is an irreversible inhibitor preferential for MAO-B; it is used for the treatment of Parkinson's disease, depression and senile dementia. Inhibitors of MAO lose its selectivity at high doses. Moreover, there are feedbacks and interconnections of intracellular signalling pathways which lead to mutual interactions of monoaminergic and other systems . So, inhibiting of MAO-B should influence processes mediated primarily by substrates for MAO-A, and vice versa. The major disadvantage was the incidence of the cheese reaction with those early inhibitors .
The selective reversible MAO-A inhibitors such as moclobemide increase the content of 5-HT, NE and dopamine in the brain  but did not provoke the cheese reaction. Moclobemide has been extensively evaluated in the treatment of a wide spectrum of depressive disorders and social phobia. Overall, moclobemide appears to be safe and devoid of major side effects, although it is considered as a mild antidepressant, better tolerated by older patients [175-181]. Moclobemide undergoes extensive metabolism with less than 1 % of the dose being excreted unchanged. Metabolic pathways of moclobemide include mainly oxidative attack on the morpholine moiety . However, major metabolites in plasma were found to be less effective MAO-A inhibitors than moclobemide or pharmacologically inactive [183-184].
MAO inhibitors were developed as antidepressants but many drugs, including the oxazolidinone antibacterial agents, share similar molecular properties and have MAO inhibitory activity. These compounds were of interest as potential antidepressants because they could be selective inhibitors of either the A or B isoforms and were usually reversible .
Antidepressants which act primarily as 5-HT and/or NE reuptake inhibitors show inhibitory activity towards MAO also. It has been suggested that tricyclic antidepressants exert some of their therapeutic effect by inhibiting MAO . They are able to inhibit MAO-B both
4.2. Effects of antidepressants on mitochondrial functions
There is relatively little data about effects of antidepressants on mitochondrial functions as summarized in the Table 2.
Effects of antidepressants on apoptotic markers, e.g. cytochrome
4.3. Effects of mood stabilizers on mitochondrial functions
Mood stabilizers affect multiple sites in intracellular signalling pathways . Main targets of mood stabilizers are neurotrophin BDNF, ERK pathway, and pathways modulated by GSK-3 or Bcl-2 [8,226-227]. Molecular and cellular targets of mood stabilizers include enzymes inhibited by lithium (inositol monophosphatase, inositol polyphosphate 1-phosphatase, GSK-3, fructose 1,6-bisphosphatase, bisphosphate nucleotidase, phosphoglucomutase), enzymes inhibited by valproate (succinate semialdehyde dehydrogenase, succinate semialdehyde reductase, histone deacetylase), targets of carbamazepine (sodium channels, adenosine receptors, adenylate cyclase), and components of signalling pathways regulated by multiple drugs (PKC, cAMP, arachidonic acid) . Furthermore, lithium and valproate reduce transport of myo-inositol into the cells, which leads to reduced PKC activity. Lithium and valproate increase Bcl-2 concentrations  and inhibit GSK-3 activity (lithium directly, valproate indirectly). Valproate activates MAPK signalling pathway and regulates stress proteins of ER . Through the effects on Bcl-2 and p53 (proapoptotic protein), lithium affects mitochondria by stabilization of membrane integrity and prevention of MPTPs opening; i.e. by regulating the key process in cell death leading to at least temporary loss of Δ
Effects of mood stabilizers on monoaminergic activity have been studied; majority of data is about the effects of lithium. Lithium enhances the antidepressant effect both of MAOIs and inhibitors of the reuptake of 5-HT and/or NE [232-234]. The mode of action for the lithium augmentation of antidepressants is partly mediated by an increase of 5-HT neurotransmission [235-237]. However, lithium could not either inhibit MAO-A or MAO-B in the brain mitochondrial [195,201]. Unipolar and bipolar depressive patients showed significantly higher platelet MAO activity than controls, but there was no significant change in activity after the institution of lithium treatment .
|Imipramine||Isolated rat liver mitochondria|
Beef heart submitochondrial particles
|Uncoupling effects on OXPHOS (release of respiratory control, hindered ATP synthesis, enhanced ATP synthase activity)|
Inhibition NADH oxidation, inhibition of ATP synthase
|Imipramine||Rat brain mitochondria||Increased state 3 and state 4 respiratory rates|||
|Imipramine||Rat liver mitochondria||Increased state 3 and state 4 respiratory rates|||
|Imipramine, clomipramine, citalopram||Human peripheral lymphocytes|
|Dose-dependent induction of apoptosis||[214,215]|
|Imipramine, clomipramine, citalopram||Human acute myeloid leukaemia HL-60 cells||Loss in cell viability, increased ROS production, loss of Δ|||
|Clomipramine, desipramine, norfluoxetine,|
|Rat heart isolated mitochondria|
and CHO cells
|Reductions of Δ|
decrease in state 3 respiration, inhibition of activities of complexes I, II/III and IV
Insignificant change of Δ
decrease in state 3 respiration, inhibition of complex I activity
|Tianeptine||Rat liver mitochondria||Inhibited beta-oxidation and TCA cycle|||
|Fluoxetine||Rat liver mitochondria||Inhibition of state 3 respiration, stimulation of state 4 respiration, decrease of RCR and uncoupling effects on OXPHOS|||
|Fluoxetine||Rat brain mitochondria||Inhibition of OXPHOS, decreased activity of ATP synthase|||
|Amitriptyline, fluoxetine||Differentiated rat|
pheocytochroma PC12 cells
|Prevention of the loss of Δ|
of ROS induced by MPP+
|Amitriptyline, fluoxetine||Rat pheocytochroma cells||Attenuation of H2O2 neurotoxic effects, upregulation of superoxide dismutase|||
|Nortriptyline||*ALS mouse||Strong inhibitor of MPT|||
|Nortriptyline||Mouse model of ischemia||Inhibition of Δ|||
|Nortriptyline||Rat brain mitochondria||Inhibitor of MPT, inhibition of ETC, mild uncoupling|||
|Rat brain homogenates||Increased citrate synthase activity after acute, but not chronic treatment|||
|Isolated rat liver mitochondria||Severe inhibition of oxygen consumption, inhibition complexes I and IV|
Modest inhibition of oxygen consumption, inhibition of complex I
|Isolated rat liver mitochondria||Complex I and complex IV inhibitor|
|Fluoxetine||Hippocampal synaptic plasma membranes||Increased ATP synthase activity|||
|Sertraline||Isolated rat liver mitochondria||Uncoupling effects on OXPHOS, inhibition of complex I and complex V activities, induction|
of Ca2+ mediated MPT
|Venlafaxine, paroxetine, nortriptiline||Rat brain homogenates (after 15 days of drug administration)||Differences in brain areas: increased or unchanged citrate synthase and SDH activities|||
|Paroxetine, fluoxetine, klomipramine||Rat glioma and human neuroblastoma cell lines||Increased cyt |||
|Desipramine||Rat glioma cells||Activation of caspases 3 and 9, no changes of Δ|||
Studies have shown effects of mood stabilizing drugs on mitochondria. In isolated brain mitochondria lithium caused desensitisation to calcium, antagonized permeability transition, and diminished cytochrome
Chronic treatment with lithium, valproate and carbamazepine protects against NMDA-mediated toxicity . Interestingly, recent study performed with epileptic children examined the influence of carbamazepine and lamotrigine on mitochondrial functions - both drugs influenced respiratory chain complexes and significantly affected ATP production, carbamazepine decreased the production, oppositely to stimulatory effect of lamotrigine . Carbamazepine interferes in adenylate cyclase pathway: inhibits adenylate cyclase and the synthesis of cAMP . Lamotrigine prevented the toxicity caused by rotenone and MPP+ in rat PC12 cells by suppressing the MPT formation, which leads to cytochrome
Effects of mood stabilizers on mitochondrial functions are summarized in the Table 3.
|Valproate||Rat liver mitochondrial|
|Inhibition of oxygen consumption rate, sequestration of intramitochondrial CoA|||
|Valproate||Isolated rat liver mitochondria||State 3 rates of oxygen consumption inhibited|||
|Valproate||Isolated beef brain α-KGDH||Inactivation of α-KGDH complex|||
|Valproic acid||Isolated rat hepatocytes||CoA, acetyl-CoA and long chain acyl-CoA fractions decreased (accumulation of valproyl-CoA; without any evidence of this metabolite in brain tissue)|||
|Valproate and its metabolites||Submitochondrial particles from rat liver||Inhibition of pyruvate uptake|||
|Valproate||Rat liver mitochondria. Digitonin permeabilized rat hepatocytes||Inhibition of pyruvate-driven OXPHOS. Inhibition of the rate of ATP synthesis (pyruvate as substrate used, no inhibitory effects caused by succinate and glutamate as substrates)|||
|Valproate and lithium||Rat brain tissue obtained|
from animals pretreated by d-amphetamine
|No modification of complex I, II, III and IV activities after the treatment with valproate and lithium in controls|||
|Valproate and lithium||Rat brain tissue obtained|
from animals pretreated by d-amphetamine
|Treated animals with lithium and valproate prevented inhibition caused by d-amphetamine|||
|Valproate and lithium||Rat brain tissue obtained|
from animals pretreated by d-amphetamine
|Reversed ATP synthase activity (increased after d-amphetamine) after lithium and valproate treatment|||
|Valproate and lithium||Human neuroblastoma and glioma cells||Protective effects against H2O2 or rotenone induced cytotoxicity in neuroblastoma cells|||
|Valproate and lithium||Human neuroblastoma cells||Reduction of methamphetamine-induced reduction of cyt |||
|Lithium||Plasma synaptic membrane from rat brain||Impaired function of ATP synthase was modulated (reversed by lithium, and prevented by lithium pretreatment)|||
|Lithium||Isolated brain mitochondria||Desensitisation to calcium, antagonized MPT, diminished cytochrome |||
|Lithium||Postmortem human brain cortex||Dose-depedent increased activities of complexes I+III, II+III and succinate dehydrogenase|||
|Lithium||Human neuroblastoma SH-SY5Y cells||Attenuation of rotenone-induced caspase-3 activation|||
|Carbamazepine||Rat liver mitochondria||Decreased state 3 respiration, RCR, ATP synthesis, Δ|||
|Carbamazepine||Rat brain mitochondria||Protection against rotenone induced complex I inhibition|||
|Human white blood cells||Carbamazepine decreased ATP production, stimulatory effect on production by lamotrigine|||
|Suppression of MPT formation, attenuation of rotenone-toxicity, inhibition of ROS production|||
Biological markers of depression, predictors of the response to the drug administration and molecular targets of new antidepressants are searched on the basis of recently known hypotheses of affective disorders. We come out mostly from stimuli of neurotrophic hypothesis and mitochondrial hypothesis. According to these hypotheses, the leading role in the pathophysiology of mood disorders and therapeutic effects of antidepressants has mitochondria, which are destined for changes in energetic metabolism of cells. Mitochondrial dysfunctions and thereby impaired neuronal metabolism can lead to disturbances in neuronal function, plasticity and brain circuitry. Impaired functions of mitochondria contribute to a wide range of diseases; the role of mitochondria in the pathophysiology of schizophrenia, bipolar disorder, and major depressive disorder is supported by studies investigating genomic differences, changes of energy metabolism and mitochondrial changes included.
This. research was supported by grant no. 41310 given by Grant Agency of Charles University in Prague, by project MSM0021620849 given by Ministry of Education, Youth and Sports and by project PRVOUK-P26/LF1/4 given by Charles University in Prague, Czech Republic.