Current Neuropharmacology, 2012, 10, 139-158 139 Somatic Drugs for Psychiatric Diseases: Aspirin or Simvastatin for Depression? Juan Gibert Rahola* Department of Neurosciences, Faculty of Medicine, University of Cadiz, CIBER of Mental Health-CIBERSAM Abstract: The evolution in the understanding of the neurobiology of most prevalent mental disorders such as major depressive disorder (MDD), bipolar disorder or schizophrenia has not gone hand in hand with the synthesis and clinical use of new drugs that would represent a therapeutic revolution such as that brought about by selective serotonin reuptake inhibitors (SSRIs) or atypical antipsychotics. Although scientists are still a long way from understanding its true aetiology, the neurobiological concept of depression has evolved from receptor regulation disorder, to a neurodegenerative disorder with a hippocampal volume decrease with the controversial reduction in neurotrophins such as BDNF, to current hypotheses that consider depression to be an inflammatory and neuroprogressive process. As regards antidepressants, although researchers are still far from knowing their true mechanism of action, they have gone from monoaminergic hypotheses, in which serotonin was the main protagonist, to emphasising the anti-inflammatory action of some of these drugs, or the participation of p11 protein in their mechanism of action. In the same way, according to the inflammatory hypothesis of depression, it has been proposed that some NSAIDS such as aspirin or drugs like simvastatin that have an anti-inflammatory action could be useful in some depressive patients. Despite the fact that there may be some data to support their clinical use, common sense and the evidence advise us to use already tested protocols and wait for the future to undertake new therapeutic strategies. Keywords: Amygdala, anti-inflammatory drugs, antidepressants, aspirin, BDNF, cytokines, depression, hippocampus, inflammation, MDD, NDAIDS, simvastatin, statins. 1. INTRODUCTION 2. STATE OF THE ART What would you think if you were a patient suffering 2.1. The Neurobiology of Depression from a depressive disorder and your doctor prescribed you Although, today, it is accepted that MDD is a chronic, aspirin or simvastatin and, when you asked him why, your progressive and recurrent process related to structural and doctor told you that the cause of your symptoms was an functional changes in the brain [1-4], understanding of its inflammation of the brain? You would possibly think that it physiopathology, despite its prevalence, is infinitely below is difficult to understand current advances in psychiatry. But that of other chronic pathologies such as type 2 diabetes, this is neither a joke nor fiction as advances in the hypertension or atherosclerosis. This is due to the fact that understanding of the neurobiology of true depression or observation of pathological changes in brain is much more major depressive disorder (MDD) have led us to consider difficult than in the case other organs: techniques for that inflammation may play an important role in this disease. studying the brain are based on post-mortem (with their The evolution in the understanding of the neurobiology of many limitations) or neuroimaging studies that use indirect the most prevalent mental disorders such as MDD, bipolar markers of activation and to the fact that the experimental disorder or schizophrenia has not gone hand in hand with the models also give us information, the problem being synthesis and clinical use of new drugs which would the extrapolation of the results to human beings [5]. represent a therapeutic revolution such as that brought about Furthermore, the majority of depressions (depressive by selective serotonin reuptake inhibitors (SSRIs) or atypical disorders) is idiopathic and the limited knowledge of their antipsychotics. Researchers should consider the current aetiology is reflected in the risk factors: stressing events, hypotheses on the neurobiology of depression to reach the endocrine alterations, cancer and undesirable side effects of inflammatory hypothesis of depression and the possible drugs, among others, [6-8]. Genetic linkage studies are not usefulness of anti-inflammatory drugs such as aspirin and conclusive and no responsible genes, permitting us to other drugs like simvastatin. reproduce the disease in animals [8, 9, 10], that have been identified; genetic predisposition interacts with environmental risk factors and can trigger off depressive episodes in some *Address correspondence to this author at the Department of Neurosciences. patents [11]. Faculty of Medicina: University of Cádiz: CIBER of Mental Health- Diagnosis of MDD is subjective and is based on the CIBERSAM, Plaza Falla 9. 11011 CÁDIZ. Spain; Tel/Fax; +34956015225; E-mail: [email protected] evaluation of specific symptoms affecting functioning over a 1(cid:27)7(cid:24)-(cid:25)(cid:20)(cid:28)(cid:19)/12 $58.00+.00 ©2012 Bentham Science Publishers 140 Current Neuropharmacology, 2012, Vol. 10, No. 2 Juan Gibert Rahola certain period of time and these criteria overlap with other o Regulates autonomic and neuroendocrine responses. processes such as anxiety disorders [12, 13]. • The lateral orbital frontal cortex (OLPFC) [22, 23]: 2.2. Neural Circuits and Depression o Activity increases in depression, obsessive- Different regions and circuits of the brain control compulsive disorder (OCD), post-traumatic stress emotions, reward and executive function. Changes in the disorder (PTSD) and panic disorder. function of limbic regions that have many interconnections o Corrects and inhibits maladaptive, perseverative and [14], have been implicated in both depression and the action emotional responses. of antidepressants (ATDs). In post-mortem [7, 15] and neuroimaging [7, 16] studies decreases in grey matter and • The dorsolateral prefrontal cortex (DLPFC) [24]: glial density have been observed in prefrontal cortex (PFC) o Involved in cognitive control, the performance of and hippocampus, areas responsible for the cognitive aspects complex tasks and the manipulation of information in of depression, although the results are not consistent. working memory. Neuroimaging studies carried out with fMRI (functional magnetic resonance imaging) or PET (positron emission o Its hypoactivity in depression is associated with tomography) show that the amygdala and the subgenual neuropsychological symptoms of depression. cingulate cortex (Cg25) are correlated with dysphoric • The amygdala: emotions [7, 13] while deep brain stimulation of white matter tracts around Cg25 and in nucleus accumbens is o Regulates cortical activation and neuroendocrine responses to surprising or ambiguous stimuli [25]. effective in resistant patients [17, 18]. These prosencephalic tracts, apart from controlling alertness and consciousness, o Role in emotional learning and memory. modulate the importance of emotional stimulus and, in o Its activation is related to the degree of depression turn, are modulated by monoaminergic projections from [7, 23]. mesencephalic and brainstem nuclei [6]. o Involved in the tendency to ruminate on negative The main areas of the brain involved in the regulation of memories. mood are (Fig. 1. Adapted from ref. no. [19]): • The hippocampus: • The ventromedial prefrontal cortex (VMPFC) [20, 21]: o Has a role in episodic and contextual learning and o Modulates pain and aggression as well as sexual and memory [26]. eating behaviors. • Involved in cognitive control, the pperformance of compplex tasks and DDLLPPFFCC the manipulation of information in workingmemory. • Its hypoactivity in depression is associated with neuropsychological ssyymmppttoommssooffddeepprreessssiioonn. Hippocampus VMPFC • Modulates pain and aggression as • Has a role in episodic and wweellll aass sseexxuuaall aanndd eeaattiinngg ccoonntteexxttuuaalllleeaarrnniinnggaannddmmeemmoorryy. behaviours. • Richincorticosteroidreceptors. • Regulates autonomic and • Has a role in the regulation by neuroendocrineresponses. feedback of the hypothalamic- pituitary-adrenalaxis. • IItsddysffunctiionmaybberesponsiibblle OLPFC for inappropriate emotional responses. • Modulatespainandaggressionaswell assexualandeatingbehaviours. • Regulates autonomic and neuroendocrineresponses. Amigdala • Regulates cortical activation and neuroendocrine responsestosurprisingorambiguousstimuli. •• RRoolleeiinneemmoottiioonnaalllleeaarrnniinnggaannddmmeemmoorryy. • Itsactivationisrelatedtothedegreeofdepression. • Involved in the tendency to ruminate on negative memories. Fig. (1). Brain areas involved in the regulation of mood. Somatic Drugs for Psychiatric Diseases Current Neuropharmacology, 2012, Vol. 10, No. 2 141 o Rich in corticosteroid receptors [27, 28]. triggers off depressive responses, which emphasizes the crucial specific regional actions of neurotransmitters and o Has a role in the regulation by feedback of the their postsynaptic effectors which are not contemplated in hypothalamic-pituitary-adrenal axis [25]. the simplistic models [38]. o Its dysfunction may be responsible for inappropriate Monoaminergic ATDs are still the basic treatment for emotional responses [29]. depression, but their long-latency effect [39] and low Having reviewed the main areas involved in the remission rates have encouraged the search for more neurobiology of depression we shall carry out a brief effective drugs [14, 40, 41]. analysis of the role of monoamines in this disorder. 2.4. Hippocampal Neuroplasticity: Neurotrophic Factors, 2.3. Monoamines and Depression Stress and Neurogenesis in the Adult Brain Several lines of evidence suggest an involvement of the Evidence suggests that depressive disorders are neurotransmitter serotonin (5-HT) in the pathophysiology or neurodegenerative disorders and there are studies providing pathogenesis of MDD [30] and that the noradrenergic or information on the decrease in the number of neurons and dopaminergic dysfunction would have a secondary role. glial cells in depressive patients. There is a decrease in the However, it is more plausible that the functional imbalance density of glial cells and the size of neurons in the DLPFC between serotonin and noradrenalin (5HT/NA) be responsible and caudal orbitofrontal cortex of depressed patients [42]. for the inadequate functioning of certain brain nuclei and Other quantitative studies demonstrated a decrease in the areas that would be reflected in certain symptomatic groups: density and size of neuronal glial cells in the anterior cingulate cortex and a low number of glial cells in the • Alterations of cortex regulation: concentration and amygdala in MDD [43-45]. memory alterations and an uncontrollable feeling of worry or guilt. Among the most important MDD structural alterations is decreased hippocampal volume, which would be directly • Hypothalamic dysfunction: changes in appetite, libido, related to the duration of untreated depression and the and autonomic symptoms. antidepressants would partially block or would reverse • Thalamus and brain stem: sleep and activation hippocampal volume decrease [46-49]. Furthermore, the alterations. hippocampus is a key structure in the regulation of the hypothalamic-pituitary-adrenal axis observed in MDD [50], • Alteration of the tracts connecting the cortex to the so that compromise of the hippocampus produces an hippocampus and the amygdala: chronic hypersensitivity alteration of neuroendocrine regulation which gives rise to to stress and fear that determines the characteristics of high cortisol levels which can act as toxins in the body and anxiety, anhedonia, aggression and lack of affective these high levels of cortisol can also affect neuronal control. plasticity and survival through the modulation of a neuro- However, it is possible that this monoamine deficiency trophic factor, brain-derived neurotrophic factor (BDNF) has only a marginal contribution to genetic vulnerability [31, [51] (Fig. 2. Adapted from ref. no. [19]). 32] since the depletion of amines has no effect on normal Decrease in the volume of the hippocampus and other subjects [33] and, in experimental studies with animals, prosencephalic structures in subgroups of depressed patients the dopamine (DA) and noradrenalin (NA) increase only supports the hypothesis of decrease of neurotrophic factors produces a maladaptive effect in paradigms related to stress, in depression, especially the BDNF, which is profusely as it strengthens the memory of adverse vital events [33, 34]. expressed in the adult limbic system. Studies with rodents Moreover, acute monoamine increases due to ATDs indicate that stress reduces hippocampus-mediated signals, induce medium-term secondary neuroplastic changes, which while chronic administration of ATDs increases them [6, 52]. Similar results have been found in the depressed involve transcriptional and translational changes, mediating hippocampus in human post-mortem studies [53] as well as molecular and cellular plasticity [6, 36]. For example, 5- an increase in plasma BDNF concentrations, whose origin is HT1B receptors interact with the p11 protein, a calcium binding protein, which increases in the cerebral cortex after controversial [52]. The antidepressant effect of direct chronic treatment with selective serotonin reuptake inhibitors infusion of BDNF into rodent hippocampus [54] and its (SSRIs) and decreases in the cingulate cortex of depressive blockade in knockout animals without the gene that encodes this factor in prosencephalic regions, provide more direct patients [37]. Moreover the specific transgenic over- proof of its causal participation in depression [55, 56]. expression of p11 in the brain produces an antidepressant phenotype, which means that p11 increase mediated by The discovery of a possible relationship between the SSRI is an important mechanism after the activation of the BDNF and MDD meant a great advance in the understanding receptor [36]. of the neurobiology of depression. BDNF is a neurotrophic Chronic administration of ATDs also up-regulates cAMP substance (it promotes the growth and survival of neurons) and there is considerable evidence that the levels of BDNF in response element-binding (CREB) transcription factor in the the brain decrease during depression and, in theory, this hippocampus (taking part in the signal cascade resulting could lead to atrophic brain changes, especially in the from the stimulation of different 5HT receptors and other receptors linked to G stimulating proteins) [35]. Furthermore, hippocampus [57]. the activation of CREB in nucleus accumbens due to stress 142 Current Neuropharmacology, 2012, Vol. 10, No. 2 Juan Gibert Rahola BDNF BDNF BDNF Stress Glucocorticoids Hippocampus Fig. (2). Stress, BDNF and apoptosis in hippocampus. The administration of ATDs normalizes the levels of Moreover, stress reduces hippocampal BDNF, which also BDNF, as has been observed both in studies with animals reduces neurogenesis [52, 62]. and in post-mortem studies on human brains of people A cellular effect of many (but not all) ATDs, including suffering from mood disorders, whether they were taking monoamine oxidase inhibitors (MAOIs), SSRIs, TCAs, medication at the time of death or not. Studies in humans SNRIs, noradrenergic and specific serotonergic ATDs, is the comparing levels of serum BDNF in depressed patients to induction of adult hippocampal neurogenesis, the process by non-depressed subjects reveal significantly lower BDNF means of which the neural progenitors of the hippocampal levels in depressed patients. It is most important to emphasize subgranular zone (SGZ) divide mitotically to form new that depressed patients treated with antidepressants have neurons, which differentiate and integrate into the dentate similar BDNF levels to those of non-depressed subjects gyrus [36, 63]. In rodent and primate hippocampus, the [58-60]. It has been demonstrated in animal models that adults generate neuronal cells, which arise from SGZ some antidepressants (i.e. SSRIs, serotonin-norepinephrine progenitor cells and migrate towards the granular cell layer reuptake inhibitors (SNRIs), tricyclic antidepressants where they differentiate into granular neurons [64]. (TCAs)), lithium and valproate increase BDNF levels in rat hippocampus. Typical antipsychotics can decrease BDNF Blockade of hippocampal neurogenesis slightly inhibits levels while atypical antipsychotics can increase its levels the "therapeutic" effect of the majority of antidepressant [52, 61]. treatments in rodents [64] but antidepressant treatments increase the concentrations of different hippocampal growth We could say that the hippocampus is the key to factors that influence neurogenesis, possibly through the depression as both 5-HT and NE influence the balance of actions of CREB or other transcriptional regulators [6, 36, activity between excitatory (glutamatergic) and inhibiting 64]. These include BDNF (which promotes neuronal (GABAergic) in PFC and limbic system. Excitatory neurons survival) as well as the vascular endothelial growth factor from the PFC have regulatory influence on the locus (VEGF) and the VGF nerve growth factor inducible (its coeruleus and the dorsal nuclei raphe [51]. A combination of name is non-acronymic), which result in antidepressant and excessive excitatory input from the VMPFC and an increase pro-neurogenic activities [65-67]. Activity dependent in glucocorticoid levels caused by stress can have a toxic increases in neurogenesis could increase activity propagation effect on the hippocampus. Modification of hippocampal in different hippocampal layers [68] and allow hippocampal function can contribute to a cognitive, emotional dysfunction networks to adapt and learn new experiences [69]. This and compromise neuroendocrine regulation in MDD [50]. would increase the possibility of the presence of intact Stress plays a very important part in the neurobiology of neurogenesis during stressful episodes producing most depressive patients, although depression without stress maladaptive learning and the appearance of depressive can exist. There are data indicating that stress can trigger or sequelae. While several types of stress reduce cell aggravate depression and that chronic stress leads to the proliferation in the SGZ, decreased neurogenesis does not atrophy or remodeling of CA3 hippocampal neurons. itself produce depression [69, 63]: Inhibition of hippocampal Somatic Drugs for Psychiatric Diseases Current Neuropharmacology, 2012, Vol. 10, No. 2 143 neurogenesis in rodents (through irradiation [70, 64] or truncated protein product of the Fosb gene) in the midbrain genetic techniques [71]) does not produce anxiety or periaqueductal grey nucleus was shown to promote a depressive behaviors. Taken together, these studies resilient phenotype. This effect was mediated through emphasize the weakness of a unified theory of depression: reducing expression of substance P, a neuropeptide released mechanisms which promote depressive symptoms in during stress [86]. response to stress differ markedly between different neuronal Mesolimbic DA-mediated signalling would participate in circuits and can also be distinct from changes that underlie emotional homeostatic mechanisms [37] so that vulnerability depression in the absence of external stress (endogenous to social avoidance and other negative consequences of depression). Moreover, neuroplastic processes that are stress would be mediated by the increased excitability of required for antidepressant efficacy do not need to reverse VTA DA neurons and their subsequent increased activity- the stress-induced alterations in plasticity and might function dependent release of BDNF onto nucleus accumbens through separate and parallel circuits [5]. neurons [86-88]. Resilient mice (with increased DeltaFOSB Neurogenesis is more prominent in sub-ventricular zones concentrations) [86] do not show this increase in VTA and in the hippocampal dentate gyrus SGZ. Early stressors neuronal excitability due to the upregulating of voltage-gated may induce abnormalities in the development of amygdala, potassium channels (molecular compensation to restore hippocampus, anterior cingulate cortex, corpus callosum and normal excitability and main maintain low levels of BDNF- other structures, which play a critical role in responses to mediated signalling in the nucleus accumbens). stress [72]. In adults, stress decreases neurogenesis in the Other possible mechanisms of resilience could be the hippocampal dentate gyrus subgranular zone [73] and release of neuropeptide Y from locus coeruleus nerve environmental enrichment [74-76] and ATDs [77] have a terminals onto the amygdala [85, 89]. There are studies positive effect on neurogenesis. The possible participation of which show stable individual differences in stress responses BDNF has already been mentioned, although other factors among genetically inbred mice, strongly implicating extra- may take part, since decrease has been observed in others genomic factors [36, 86, 90-92]. As these mice are kept in such as FGF (fibroblast growth factor) [78] or NCAM (neural identical environmental conditions, the findings imply the cell adhesion molecule), contributing to neurocognitive importance of epigenetic mechanisms during development. disorders in depressive processes related to stress [79]. The study of gene expression of stress-vulnerable and The BDNF hypothesis is, however, too simplistic and not stress-resilient mice revealed distinct transcriptional profiles exempt from criticism. Many preclinical studies have been in the VTA, nucleus accumbens [38] and hippocampus unable to demonstrate these changes produced by stress or suggesting that the resilient behavior represents a different, ATDs or have obtained opposing results [80, 81]: active neurobiological process (not simply the absence of conditional BDNF knockout show no depressive behavior vulnerability [36]). The understanding of such molecular [55], BDNF exerts a powerful pro-depressant action in the mechanisms of allostasis (efforts to maintain homeostasis) ventral tegmental area (VTA) and nucleus accumbens (its may be the way to the development of new drugs [93, 94]. expression in nucleus accumbens is increased by stress [82], direct infusion into these zones increases behaviors related to 2.6. Hippocampal-Neuroendocrine Interactions depression [35, 83] and a selective knockout without the In previous sections we have referred to the relationship gene encoding BDNF in this circuit has an antidepressant of stress with depression [1, 11, 37, 50, 52, 62, 82, 86, 88, effect) [82]. 89, 92-94], so that it is logical for alterations of stress axis These data suggest that BDNF and its TrkB receptor hormones such as increased serum glucocorticoid produce different effects on behavior depending on the area concentrations to be produced in depression [95, 96]. On the of the brain [52]. one hand, we know that physical or psychological stress increases serum glucocorticoid concentrations and that their 2.5. Cellular Resilience and Depression chronic administration can produce symptoms similar to Although it might seem that this aspect should have been depression [97] in rodents and, on the other, that excess included in the previous section, we would like to give it glucocorticoids can reduce cellular proliferation in the special attention. Human responses to stress and adversity hippocampal subgranular zone (SGZ) and produce atrophy are very heterogeneous which means that some types of in hippocampal sub-regions [93, 94], which may contribute depression could be due to stressing events, or that the to the reduction of its volume observed in depression. There events moderately increase the risk of depression [1, 11]. are various facts supporting this relationship, such as that However, reactive dysphoric states, as PTSD (post-traumatic patients with Cushing’s syndrome have depressive stress disorder), only appear in about 10-20% of patients symptoms and hippocampal atrophy [6, 93] and that some of exposed to trauma [84]. Although there is a lot of data the alterations of depression (insulin resistance and suggesting post-stress maladaptive neurobiological changes abdominal obesity) can be explained, at least in part, by the (decreased hippocampal neurogenesis and less BDNF) there increase in glucocorticoids [98]. is little information as to why most people adapt well (are Hypercortisolaemia in depression is manifested at several resilient) to adversity [85]. In studies carried out on animals, levels: analysing natural variations in the development of active escape in the learned-helplessness test, stress-induced • Alteration of glucocorticoid-receptor-mediated negative upregulation of the transcription factor ΔFOSB (a stable, feedback [98]. 144 Current Neuropharmacology, 2012, Vol. 10, No. 2 Juan Gibert Rahola • Adrenal hyper-responsiveness to ACTH [95]. been recommended and, especially, a recently published editorial by Maes et al. [110]. However, out of all the works • Hyper-secretion of Corticotropin-releasing hormone published there is one, Maes et al. [107], which establishes a (CRH), originally named corticotropin-releasing factor very logical sequence to explain the inflammatory hypothesis (CRF) [99, 6, 100]. of depression and which we have followed. However, hypercortisolaemia only occurs in very severe In previous sections we have tried to explain the data that depression [102] or accompanied by psychotic symptoms [6, strongly support the hypothesis that neurodegeneration 11] in which glucocorticoid antagonists show a certain would be added to a decrease in neurogenesis in MDD. amount of efficacy [102]. Volumetric changes have been observed in hippocampus, In atypical depression (hyperphagia and hypersomnia) amygdala, CPF, anterior cingulate and basal ganglia [111] as there would be hypocortisolaemia [101, 103, 104], which is well as hippocampal cellular changes and neuronal and glial also observed in fibromyalgia, chronic fatigue and PTSD cell modifications [112]. Selective loss of hippocampal [101]. High concentrations of glucocorticoids promote volume is due to both neuronal death and the decrease in the mobilization of energy during stress while low neurogenesis [113]. glucocorticoid concentrations help the immune system How could we relate decrease in neurogenesis and against infection or physical injury [105]. It is important to neurodegeneration to inflammation? There are various, point out that hippocampal dysfunction contributes to the stimulating hypotheses and there is more and more neuroendocrine alterations of depression [6] and that these information to support them [107-110]: have many systemic consequences [106]: the hypothalamus stimulates the pituitary gland and excess ACTH is produced, • Inflammation and neurodegeneration would play an constantly stimulating the suprarenal glands; the suprarenal essential role in depression. glands liberate an excessive amount of catecholamines and • The increase in neurodegenerative processes could be cortisol and the increase in catecholamines may cause due, at least in part, to inflammatory processes. myocardial ischemia, reduction in heart rate variability, contribute to ventricular arrhythmia and cause platelet • Multiple pro-inflammatory cytokines, lesions induced by activation; cytokine and interleukin increases can also oxygen radicals, tryptophan catabolites and neuro- contribute to atherosclerosis and possible hypertension; degenerative biomarkers are produced in depression. finally, the cortisol antagonizes insulin and contributes to • There are some vulnerability factors which predispose to dyslipidemia, type 2 diabetes and obesity; the increase in depression, increasing inflammatory reactions: cortisol also inhibits the immune system (Fig. 3). o Decrease in peptidase activity (dipeptidyl-peptidase 2.7. Inflammation, Neurogenesis and Neurodegeneration IV, DPP IV) Previous aspects had to be treated before we deal with o Decrease in omega-3 polyunsaturated fatty acid levels. this point. Several excellent reviews [107, 108, 109] have Depression Hippocampal disfunction Hypothalamus stimulates the pituitary gland and excess ACTH is produced, constantly stimulating the suprarenal glands The suprarenal glands liberate an excessive amount of catecholamines and cortisol The increase in Cytokine and Cortisol antagonizes The increase catecholamines may cause interleukin insulin and in cortisol myocardial ischemia, increases contribute contributes to inhibits the reduction in heart rate to atherosclerosis dyslipidemia, type 2 immune variability, contribute to and possible diabetes and obesity system ventricular arrhythmia and hypertension cause platelet activation Fig. (3). Hippocampal dysfunction and systemic consequences of the neuroendocrine alterations of depression. Somatic Drugs for Psychiatric Diseases Current Neuropharmacology, 2012, Vol. 10, No. 2 145 o Increased intestinal permeability 1. ω3 PUFAs reduce the production of pro-inflammatory cytokines such as TNFα (tumor necrosis factor-α) and IL- • The cytokine hypothesis considers that external 1β. (psychosocial) and internal (inflammations and postnatal period) could trigger off depression through inflammatory 2. ω3 PUFAs modulate membrane protein quaternary processes. structure as well as their fluidity, which determines serotonin fixation [138]. Serotonin, on its part, stimulates • Many ATDs have specific anti-inflammatory effects. neurogenesis [139]. • While restoration of neurogenesis, which could be 3. ω3 PUFAs influence the levels of neurotrophins such as induced by inflammatory processes, would be related to BDNF [141]. the therapeutic efficacy of the ATD treatment. Cytokines are important modulators of mood, their SNC Inflammation may produce oxygen radicals and this receptors being activated by both those produced centrally increase has also been observed in depression [115-117]. and peripherally [142]. Maes et al. [107] have recently Oxidative stress is an important factor in neurodegenerative proposed that the relationship between pro-inflammatory diseases [114]. It is responsible for programmed cell death, cytokines and depression would be based on 6 points: apoptosis, necrotic cell death and damage to DNA and membrane fatty acids which means that lipid signalling 1. In depression there is an increase in pro-inflammatory would be altered and lipid peroxidation increased. It cytokines (interleukine-1β (IL- 1β); IL-6, and IFNγ would affect gene expression and proteolysis contributing (interferon-gamma)) and the acute phase of depression is to neurodegeneration [114-118]. Furthermore, antioxidant due to high levels of pro-inflammatory cytokines such as enzymes (dismutase superoxide, catalase and glutathione IL-6 and IL-1β [144-146]. peroxidase) show therapeutic efficacy in neurodegenerative 2. They are able to cause depressive-like behaviors [143]. models [120, 121]. 3. They may explain the multicausal aetiology of depression Another important factor is nitrosative stress. One of the by which psychosocial stressors and internal stressors modifications caused by nitric oxide metabolism imbalance (medical diseases) may trigger depression [146]. is S-nitrosylation of protein cysteine residues [122]. Overproduction of NO may compromise neural energy and 4. They may explain the serotonergic alterations of lead to neurodegeneration [122]. It inhibits cellular depression [148, 145]. respiration and superoxide anions would be released, 5. The may explain alterations of the hypothalamic- interacting with NO superoxide anions produced by the pituitary-adrenal axis in depression [149] mitochondria producing peroxynitrite which is a powerful oxidant causing neurotoxicity [123]. 6. Antidepressants block pro-inflammatory cytokines [150, 151] Another common characteristic shared by depression and inflammation is IDO (indoleamine 2,3-dioxygenase) However, studies directed towards relating depression to activation, an enzyme that induces the catabolism of serum cytokine increases are inconsistent [152] and this tryptophan into TRYCATs (tryptophan catabolites along the might suggest that immunological activation would only be IDO pathway), such as kynurenine with the consequent produced in depressive subgroups, especially when there is increase in TRYCATs [124-126]. Some TRYCATs, such as comorbidity with autoimmune process (rheumatoid arthritis) quinolinic acid and kynurenine, are extremely neurotoxic in which generalized inflammation may increase the risk of [127, 128], as quinolinic acid causes acute tumefaction and cardiac arrest as well as producing depressive symptoms the destruction of post-synaptic elements, induces nerve cell [153]. degeneration, including hippocampal cell death and selective necrosis of granular cells, among others. Quinolinic acid Could we, in a quite concise manner, relate external (or causes a dose-dependent decrease in cholinergic circuits internal) stressors to inflammatory processes and cell and may empty dopamine, choline, GABA and encephalin damage, which occur in depression? External stressing deposits [129-133]. The areas most affected by the factors in animals, such as mild chronic stress and learned- neurotoxic effects of quinolinic acid would be the striatum, helplessness, are accompanied by behavior that we interpret the pallidal formation and the hippocampus [134]. The as being depressive. These behaviors are accompanied by neurotoxic effects of quinolinic acid may be due to: peripheral and central inflammation, with increased levels of pro-inflammatory cytokines, such as IL-1β, TNFα) and IL-6. 1) Agonism at NMDA receptor level [135, 136] External trigger factors and cytokines may induce NFkB 2) Pro-oxidant effect through the formation of ferrous (nuclear factor kB), which in turn induces the expression of quinolinate chelates inducing lipid peroxidation pro-inflammatory cytokines; O &NS (oxidative, and nitrosative) stress pathways and COX-2 (cyclooxygenase) 3) Exacerbation of the neurotoxic effects of corticosterone pathways. Through these pathways stress may cause a higher and IL-1 (interleukine-1) [135, 137]. number of ROS and RNS (reactive oxygen and nitrogen According to Maes et al. [107], omega-3 polyunsaturated species), including O2 and NO, which results in peroxynitrite fatty acids (ω3 PUFAs) influence neurogenesis [138] generation. COX-2 may generate PG (prostaglandins), such through their anti-inflammatory and serotonergic effects and as PGE2 and PGJ2. External stressing factors also increase those they have on neurotrophins [139] in the following way: the expression of TLR4 (Toll-like receptors), raising the 146 Current Neuropharmacology, 2012, Vol. 10, No. 2 Juan Gibert Rahola sensitivity of internal stressing factors, including PAMP neuroprogressive", since the term "neuroprogression" is (pathogen-associated molecular patterns), such as LPS more appropriate to describe the participation of apoptosis and DAMPs (lipopolysaccharide, and damage-associated and anti-neurogenic and neurodegenerative progresses than molecular patterns). External stressing factors increase the term "neurodegeneration" [155]. glucocorticoids and the liberation of glutamate and, The effect of antidepressants, antagonizing the effect or consequently, provoke NMDA (N-methyl-D-aspartate) the liberation of pro-inflammatory cytokines has been known neuronal receptor activation. External stressing factors cause for some time [150, 151] and would justify their anti- antineurogenic effects by down-regulating neurotrophic inflammatory effect. There seems to be no doubt that factors such as BDNF and VGF. Finally, external stressing the administration of antidepressants produces anti- factors cause apoptosis, with low levels of Bcl-2 (B-cell inflammatory effects (see the reviews of Maes et al. [156] lymphoma 2) and BAG1 (Bcl-2 associated athanogene 1), and Kubera et al. [154]) as it reduces the production of and an increase in caspase-3 levels. Cytokines, as well as O TNFα, IL-12 and IFNγ and increases that of one of the most and NS stress, NFkB, COX-2, PGs (PGE2), excitotoxic important anti-inflammatory cytokines, IL-10. The majority glutamatergic effects, apoptosis pathways and the decrease of antidepressants, including ISRS, tricyclic, IMAO, lithium in neurotrophic substances contribute to neurodegenerative and even atypical antidepressants, such as tianeptine, have process and the decrease in neurogenesis observed in anti-inflammatory effects [157]. For a more detailed study of depressive behavior [154] (Fig. 4). inflammatory process in depression see the reviews by Gardner et al. [158], Maes et al. [159], Maes [160], Maes 2.8. Anti-inflammatory Effects of Antidepressants et al. [161], Son et al., [162], Szewczyk et al. [163] and The evolution of the understanding of the neurobiology Zunszain et al. [164]. of depression means that the concepts and names of There are studies relating the effect of antidepressants to hypotheses are constantly changing. Until a short time ago their action on cytokines. We shall only mention one of these scientists were talking about the "inflammatory and studies [165] during which 100 patients (36 men and 65 neurodegenerative hypothesis of depression" [98] and they women) were treated with escitalopram 10-20 mg / day for included a group of processes among which were the 12 weeks. Responders and non-responders were identified oxidative and nitrosative stress pathways, pro-inflammatory according the Montgmery-Asberg scale (MADRS). Cytokine cytokines, decrease in antioxidants, including zinc, levels were measured at the commencement and during coenzyme Q and glutathione, the formation of TRYCATs, weeks 4 and 12 of treatment and compared to cytokine levels decrease in ω3 PUFAs and the increase in glucocorticoid in healthy volunteers ((n = 45, 19 men and 26 women). The levels. This hypothesis, now known as "inflammatory and Peripheral and pro-inflammatory central cytokines (IL-1β, inflammation TNFα, IL-6 O &NS stress NFkB pathways ROS & COX-2 PG (PGE2 & RNS pathways PGJ2) External TLR4 sensitivity of internal stressing factors stressing factors (PAMP: LPS & DAMPs) glucocorticoids and the liberation of glutamate NMDA neuronal receptor activation antineurogenic effects by down-regulating neurotrophic BDNF & VGF apoptosis, Bcl-2, BAG1 + caspase-3 levels Contribute to neurodegenerative process and the decrease in neurogenesis observed in depressive behaviour Fig. (4). External and internal stress, inflammation and cell damage. Somatic Drugs for Psychiatric Diseases Current Neuropharmacology, 2012, Vol. 10, No. 2 147 results obtained indicate that a higher level of TNF-α could information about possible interactions of antidepressants, predict the lack of response to treatment with escitalopram cytokines, p11 and non-steroidal anti-inflammatories and that changes in sIL-2R concentration during the (NSAIDS), the researchers carried out experiments on mice treatment were different in responders and non-responders. and reanalysed the STAR*D data [204]. Selective serotonin, citalopram and fluoxetine reuptake inhibitors increased p11 Table 1 shows the main studies on the anti-inflammatory levels in rat frontal cortex, but co-administration of effect of ATDs. ibuprofen (IBU) or acetylsalicylic acid (ASA) blocks this 3. ASPIRIN OR OTHER ANTI-INFLAMMATORIES increase. IBU decreases plasma citalopram levels. The FOR DEPRESSION? administration of desipramine produced a small p11 increase, which was not affected by IBU or ASA. Anti- This is the key point of our review and an aspect that depressants related to increased p11 depend on the signalling may surprise the clinical psychiatrist. Should we administer of two cytokines (IFNγ and TMF-α). In a rat depression NSAIDS like aspirin for depression? How and when, in model IBU, ASA and acetaminophen impeded behavioral monotherapy, associated with antidepressants? Theoretically, response to SSRI, but not to other types of antidepressants. if depression is an inflammatory disease, the administration During the STAR * D study, in the patients who took of anti-inflammatories should be beneficial. citalopram for 12 weeks, the remission was significantly 3.1. Aspirin lower if they were taking NSAIDS than if they were not (45% vs. 55%). The findings were similar when other Observation of the effects of aspirin on the mood is not analgesics were used (37% vs. 54%). There is a difference something new and it was proposed before the inflammatory between serotonergic and noradrenergic antidepressants hypothesis of depression as, in 1996 [200], it was observed, because SSRI increase cytokines, which increase p11, for the first time, that the administration of low doses of resulting in an antidepressant response. NSAIDS (and aspirin to patients about to undergo coronary angiography acetaminophen) inhibit passage to cytokine activation. was associated with less anxiety, depression or guilt. STAR*D has shown that the worst antidepressant response is associated with the use of analgesics. 3.1.1. Preclinical Trials 3.1.2. Clinical Trials We shall only mention two studies with animals during which aspirin has been used to increase the effect of There is only one study that confirms the clinical efficacy fluoxetine. The first one [201] evaluated the effect of the co- of aspirin in depression and it is an open study. Mendlewicz administration of acetylsalicylic acid (ASA, 45 mg / kg or et al. [205], based on their group's preclinical trials, [200], 22.5 mg / kg) and fluoxetine (FLX, 5 mg /kg) in the chronic published a study that demonstrated that the administration escape deficit model of depression. In this model, FLX of ASA, 160 mg / day shortened the onset of the action of needed 3 weeks to reverse behavioral changes. In this study fluoxetine. Another study [206] included seventy-seven it was observed that combined treatment of fluoxetine and patients with MDD, divided into two groups. The first group, ASA reverted the condition of escape deficit in 7 days, a consisting of 52 patients, received fluoxetine 20 mg, and the partial effect being observed after 4 days, and was second one, in addition to fluoxetine 20 mg, received 150 mg maintained after 14 and 21 days of treatment. ASA alone of ASA. The activity of antioxidative enzymes, copper-zinc was ineffective and the effect of fluoxetine was significant superoxide dismutase (CuZnSOD, SOD1), catalase (CAT), only at 21 days. The aim of the second study [202] was to glutathione peroxidase (GPSH-x) and the concentration of investigate whether aspirin can be used as an augmentation malonyldialdehyde (MDA) was determined in erythrocytes, agent in fluoxetine treatment resistant depressive rats and the total antioxidant status (TAS) was determined in induced by chronic unpredictable mild stress (CUMS). In the plasma. All parameters were measured before and after this study, the effects of CUMS regimen and antidepressant three months of treatment. The results obtained indicate treatment were assessed by behavioral testing, hippocampal a significant decrease in the activity of SOD1, CAT and expression of COX-2 and PGE2. 4-weeks of fluoxetine GSHP-x, as well as in MDA concentration after the treatment antagonized the behavioral changes in combined therapy. A significant TAS increase was also approximately 70-80% of depressive rats. This means that, observed. The study demonstrated that combined therapy 20-30% of depressive rats were resistant to fluoxetine. In the with fluoxetine and ASA is characterized by the same hippocampus of fluoxetine treatment resistant depressive efficacy and clinical safety as fluoxetine monotherapy, rats, a significant upregulation of COX-2 level and PGE 2 but with an additional improvement of oxidative stress concentration was produced. However, in these rats, parameters in patients treated for depression. adjunctive aspirin treatment significantly improved the depressive behaviors and downregulated the COX-2 level 3.2. Cyclooxygenase-2 Inhibitors and PGE2 concentration in the hippocampus. COX-2 is expressed and has important functions in the There is a very important study that provides data leading central nervous system. It interacts with neurotransmitters in the opposite direction [203]. Scientists know that cytokines such as acetylcholine, serotonin and glutamate as well as produced by glial cells regulate the serotonergic and participating in the regulation of immune mechanisms in the noradrenergic brain systems and activate the hypothalamic- central nervous system and of the inflammatory response by pituitary-adrenal axis. Antidepressants increase the levels of prostaglandins, especially PGE2. COX-2 inhibitors have p11, a specific protein that regulates experimental rat models been used as coadjuvants in the treatment of schizophrenia and interacts with serotonin receptors. In order to obtain [207, 208] and it is logical that, since they are the most 148 Current Neuropharmacology, 2012, Vol. 10, No. 2 Juan Gibert Rahola Table 1. Anti-inflammatory Effects of Antidepressants Level References Preclinical In ex vivo cultures, TCAs highly significantly blocked the production of IL-1β, TNFα and IL-6 in human purified [150] study monocytes, the production of IFNγ and IL-2 in human lymphocytes Preclinical Imipramine: CMS-exposed Wistar rats: ↓Con A-stimulated IL-1, IL-2 production; CMS-unexposed Wistar rats: ↔ [166] study Con A-stimulated IL-1, IL-2 production (splenocytes) Preclinical TCAs, like clomipramine and imipramine, reduce the number of Th1 cells secreting IFNγ in rats with experimental [167] study autoimmune neuritis. Preclinical Desipramine ↓ LPS-induced TNF-α levels ↑ LPS-induced IL-10 levels (plasma) [168] study Paroxetine, venlafaxine ↔ LPS-induced TNF-α, IL-10 levels (plasma) Preclinical Desipramine: in bulbectomized Rats ↓LPS-induced TNF-α, IL-1β levels; Sham operated ↓ LPS-induced IL-1β levels; [169] study ↔ LPS-induced TNF-α (plasma) Preclinical In animal models, antidepressants decrease inflammation-induced brain cytokine production and actions as well as [170] study depressive-like symptoms. Imipramine and fluoxetine: ↔ LPS-induced TNF-α and IL-β mRNA; (spleen) Preclinical Tianeptine ↓ LPS-induced TNF-α levels ↔ LPS-induced IL-1β, IL-10 levels (plasma), ↓LPS-induced TNF-α, IL-β [171] study RNAm (spleen) . Tianeptine ↓ LPS-stimulated IL-1β, TNF-α, IL-6 mRNAs and a tendency for ↑ IL-10 mRNA in the hypothalamus without changes in the hippocampus and pituitary in rats Preclinical Desipramine: CMS-exposed: ↑ Con A-induced IL-10 production; CMS-unexposed C57BL/6 mice: ↔ Con A-induced [172] study IL-2, IL-4, IFN-γ production (splenocytes) Preclinical The effects of antidepressants on the production of IL-6 and TNFα are partly modulated by serotonergic mechanisms. [173] study Preclinical Bupropion lowers production of tumor necrosis factor-alpha and interferon-gamma in mice [174] study Preclinical Mirtazapine Single injection in NET-KO mice: Production stimulated by Con A (splenocytes): ↓ ↓ IL-6, ↓ IFN-γ, ↑↑ [175, 176] study IL-4; in C57BL/6J mice: 7 days, ↓ ↓ IL-6, ↓ IFN-γ, ↑ IL-4, Single injection: ↓ IL-6, ↔ IFN-γ, ↑ IL-4 Preclinical Amitriptyline ↓ LPS-stimulated IL-1β, TNF-α release in primary mixed glial cells and nortryptyline in microglia cells [177] study Preclinical Imipramine, fluoxetine and reboxetine: ↓ IFN-γ-stimulated IL-6 release in microglia 6-3 cells [178] study Preclinical Paroxetine: ↓ IFN-α-induced IL-β, TNF-α and ↑ IL-10 production stimulated by Con A+LPS (whole blood culture) [179] study Paroxetine Normalized IFN-α-stimulated IL-1β, IL-10 levels in the rat hypothalamus Preclinical Desipramine: ↓ LPS-stimulated IL-1β, IL-6, TNF-α release in hippocampus-derived adult neuronal stem cells [180] study Preclinical Clomipramine, imipramine: ↓ LPS-stimulated TNF-α levels, ↓ LPS-stimulated IL-1β, TNF-α mRNAs without any [181] study changes in the unstimulated gene expression in BV-2 microglia cells Preclinical Desipramine: ↓ LPS-stimulated IL-1β, TNF-α mRNAs in the rat cortex [182] study Desipramine: ↔ LPS-stimulated IL-1β, TNF-α mRNAs in cortical mixed glia cells Preclinical In brain cell cultures, TCAs and SSRIs significantly suppressIL-1β, IL-6 and TNF-α production [183] studies Clinical study In some but certainly not all studies, antidepressant treatments were shown to normalize the initially increased IL-6 [184-186] plasma levels in depressed patients Clinical study A 6-week treatment with antidepressants significantly decreased the initial increases in the stimulated production of [187] IFNγ, significantly decreased the initially elevated plasma concentrations of a number of APPs (acute phase proteins), such as C-reactive protein, haptoglobin and alpha 2-macroglobulin.