An Overview of the Heterogeneity of Major Depressive Disorder: Current Knowledge and Future Prospective.

Major depressive disorder (MDD) is estimated to impose maximum debilitating effects on the society by 2030, with its critical effects on health, functioning, quality of life and concomitant high levels of morbidity and mortality. Yet, the disease is inadequately understood, diagnosed and treated. Moreover, with the recent drastic rise in the pace of life, stress has materialized as one of the most potent environmental factors for depression. In this scenario, it is important to understand the modern pathogenetic hypotheses and mechanisms, and possibly try to shift from the traditional approaches in depression therapy. These include the elaboration of pathophysiological changes in heterogeneous systems such as genetic, epigenetic, serotonergic, noradrenergic, gammaaminobutyric acid, glutamatergic and endocannabinoid systems, neurotrophic factors, HPA axis, immune system as well as cellular stress mechanisms. These components interact with each other in a complex matrix and further elucidation of their mechanism and cascade pathways are needed. This might aid in the identification of MDD subtypes as well as the development of sophisticated biomarkers. Further, characterization might also aid in developing multitargeted therapies that hold much promise as compared to the conventional monoamine based treatment. New candidate pharmacons, refined psychotherapeutic modalities, advanced neuro-surgical and imaging techniques as well as the implementation of pharmacokinetic, pharmacogenetic prescribing guidelines constitute the emerging expanses of MDD treatment. Copyright© Bentham Science Publishers; For any queries, please email at epub@benthamscience.net.

INTRODUCTION

Depression is a disease that imposes a severe burden on the global population with a lifetime risk close to 15% [1]. Around 350 million people are estimated to be affected [2], with women twice diagnosed as compared to men [3, 4]. With its critical effects on the quality of day to day functioning [5], major depressive disorder (MDD) is set to create a greater societal and economic burden worldwide by 2030 [6]. Whereas the existing armamentarium of antidepressants contains numerous and widely used agents, the morbidity and mortality continue to increase, especially from suicide [7]. Even the effectiveness of serotonin and noradrenaline reuptake inhibitors (SSRI and SNRI, respectively) - the frontline class of antidepressant treatment, are delimited by the lower success rate; 30% of depressed patients are pharmacoresistant [8, 9]. Moreover, in depression, the path to successful treatment has to overcome hurdles like delayed onset of effectiveness and undesirable side effects of therapeutic agents (such as gastrointestinal symptoms, agitation, sleep and weight changes, and sexual dysfunction) [10, 11], along with the episodic course, with complete remission, and a recurrent course, with short periods of remission between episodes as well as progression to chronic form of depression that is resistant to treatment [6]. The risk of recurrence is partly associated with neurobiological vulnerabilities [12] and is higher with younger age and if there is a family history of depression [13]. Even untreated gestational depression, as well as depressive symptoms during pregnancy, can have detrimental effects on the developing fetus [14]. Thus, there is a need to develop sophisticated clinical markers along with better tolerated and effective pharmacotherapeutics for the timely diagnosis, prognosis and treatment of depression. This can be achieved only through a study of the mechanisms and pathogenesis of MDD.

THE PATHOGENESIS OF DEPRESSION

The term “depression” was first associated with the condition of low spirits by Samuel Johnson in the 1750s [15]. Through the years, numerous theories varying in scope were put forward and have eventually emerged with an environmental, social and evolutionary perspective according to which depression could be considered an evolutionary adaptation to deleterious social environment [16, 17]. Evidently, with the recent drastic rise in the pace of life, stress has become one of the most potent environmental factors for depression. In a healthy individual, a stress response is a homeostatic mechanism that aids optimum physiological and psychological functioning. When there are repeated stressor events, the individual gets affected through the unresolvable personal despair caused by the accumulation of psychological and physical demands, known as allostatic load [18]. The resulting induction of pathogenetic mechanisms is considered to disrupt the body’s internal homeostasis [6]. It is well understood that the neuropsychological aspect of allostasis is maintained by the mediators comprising of the autonomic nervous system, neuroendocrine system, immune system and the associated elements such as catecholamines, hormones and inflammatory mediators [6, 19]. Further, the diathesis-stress model proposed that a genetically vulnerable individual might be more inclined towards the development of depression on exposure to stressful events in the course of life [20]. Incidents affecting nervous system development such as prenatal, postnatal or peri- adolescent events might also play a role [21-23].

Recently, Gold proposed depression as a dysregulation of the adaptive response, an innate mechanism that tries to restore the homeostatic level altered by chronic stress [24]. Gold added that MDD could spread in a metastatic mode from the brain to peripheral areas via activation of a stress system cascade. Eventually, it could progress to multiple systemic pathologies such as inflammation, prothrombotic state and insulin resistance, resulting in the development of co-morbid illnesses like coronary artery disease, osteoporosis and diabetes. On this basis, Gold suggested the application of multiple interventions at various loci to effectively treat depression, especially when there is a failure of existing medications [24].

Animal models of depression based on various forms of stress elicitation such as chronic mild stress [25, 26], learned helplessness [27], repeated social defeat stress [28-30], separation from group housing [31], social dominance/subordination [32] and neonatal separation (early life stress model) [33, 34] corroborate with the above findings and are widely used for the evaluation of antidepressant candidates.

Environmental Effects on Genes and Epigenetic Interactions

While the heritability factor accounts for a 1/3rd of the risk for developing MDD, environmental influence is purported to the remaining 2/3rd [35, 36]. The gene-environment interactions could mediate catastrophic insults in the brain by altering structures including limbic and cortical brain areas, leading to functional abnormalities associated with alertness, arousal, behavior, thoughts, learning and memory. Epigenetic mechanisms, defined as chromatin changes capable of modifying gene expression without any modification in a DNA sequence, might be mediating these interactions. Recently, environmental events and behavioral experience have been reported to cause changes in various epigenetic markers, such as histone acetylation (via histone acetyltransferases and histone deacetylases (HDAC)) and methylation (via histone methyltransferases) as well as DNA methylation (via DNA methyltransferases) on specific gene loci controlling neuronal plasticity, thereby modulating the vulnerability or resilience to the development of MDD [36-39]. Exposure to various environmental factors can also modulate the expression and function of several miRNAs in the nervous system affecting adaptive mechanisms [40]. Epigenetic modifications in brain regions such as paraventricular nucleus (PVN) of the hypothalamus, prefrontal cortex, hippocampus, amygdala, nucleus accumbens, dorsal raphe, locus coeruleus and ventral tegmental area via experimental manipulations have been demonstrated to potently control stress responses, mood as well as antidepressant treatment response [37]. For instance, a significant increase in HDAC activity was reported in the nucleus accumbens of maternally deprived rats, which was found to be reversed by ketamine and imipramine treatment [41]. Moreover, epigenetic mechanisms might also be related to the individual differences in stress susceptibility, the intergenerational/transgenerational transmission of stress susceptibility, difference in psychiatric disease susceptibility between monozygotic twins as well as the difference in gender susceptibility to depression [42, 43], and the inconsistency in the genetic association studies of MDD [44]. Further, the genetic background can influence the consequence of chronic stress, which could be partially mediated by DNA sequence -directed alterations in epigenetic markers. In this regard, chronic mild stress was reported to induce depressive-like behavior in BALB/C mice, while C57BL/6 mice were not affected. The basis of such an observation was proposed to be the differences in methylation patterns at the glial cell-derived neurotrophic factor (GDNF) promoter [45]. Moreover, chronic administration of escitalopram in the Flinders Sensitive Line (FSL) genetic model of depression was reported to mediate antidepressant effects via epigenetic alterations in the P11 promoter region [46]. Another major revelation is the antidepressant potential of HDAC inhibitors [47], including sodium butyrate [48], trichostatin A [49], MS-275 [50] and vorinostat [51, 52]. Interestingly, the administration of HDAC inhibitors is reported to augment the effect of antidepressants, thus the combination is proposed as valuable for the treatment-resistant depression.

In the case of genome-wide association (GWA) studies, while, most of the studies were not successful in finding an association between susceptibility genes and MDD, a recent analysis has reported a strong association between MDD and 42 gene sets [53, 54]. These include proteins coding for known targets of antidepressant medications [55], clock gene to be one of them [56]. Moreover, genetic polymorphisms are reported in genes such as BMAL1 (ARNTL1), CLOCK, NPAS2, PER3, CRY1, RORA and TIMELESS [57-59], out of which only RORA has been supported in more powerful GWAS [60-62]. Chronic unpredictable stress-related changes in the suprachiasmatic nucleus might be influenced by clock-stress interactions as reported from a transient decrease in Per2 amplitude [63, 64]. Others include apolipoprotein E (APOE ԑ2 and APOE ԑ4), guanine nucleotide-binding protein (GND3), methylenetetrahydrofolate reductase (MTHFR 677T), dopamine transporter (SLC6A3), the serotonin transporter (SLC6A4) and the dopamine receptor gene (DRD4), have been reported to have significant relationships with MDD [65, 66]. These findings indicate that genetic, as well as epigenetic characterization, is imperative in the understanding of depression neurobiology.

Biochemical and Physiological Alterations

Ever since early observations, the influence of depression has focused predominantly on the biochemical and physiological alterations. These include changes in the levels and functions of neurotransmitters, neuromodulators, neurotrophic factors, HPA axis components and immune molecules.

Neurotransmitters/neuromodulators

Repeated exposure to stressful stimuli is reported to elicit depression through modulation of neural excitability and neurotransmitter systems. Accordingly, psychopharmacology of depression has conventionally focused on neurotransmitters including 5-hydroxytryptamine (5-HT), noradrenaline, dopamine and acetylcholine [67-69]; their synthesis, intracellular trafficking, degradation, re-uptake and effect on both presynaptic and postsynaptic receptors [70]. Originally, it was a serendipitous discovery that found increasing monoamine concentration in the synapses which could attenuate depressive symptoms, thereby linking the deficiency of monoamines to the depression pathogenesis. In this line, pharmacological agents were developed, and thus the first generation antidepressants, monoamine oxidase inhibitors (MAOI) and tricyclic antidepressants (TCA) came into use in the 1950s. Due to their limitations including low tolerance and toxicity, subsequently, SSRI, SNRI and atypical antidepressants were developed for better clinical application [71, 72].

Monoaminergic System

In fact, the monoaminergic hypothesis of depression is the most widely accepted one; predominantly, ‘5-HT hypothesis of depression’ according to which a deficiency of the synaptic 5-HT concentration could result from increased degradation by MAO, or an increase in its reuptake [73, 74]. Indeed, alterations in postsynaptic 5-HT signaling in corticolimbic structures might be strongly associated with the vegetative, emotional, endocrine and cognitive aspects of MDD [75, 76]. Recently, P11 was shown to mediate antidepressant effects through the amplification of serotonergic signaling via regulation of 5-HT receptor trafficking and regulation of gene transcription via p11/annexin A2/SMARCA3 [77]. However, 5-HT agonists do not work on all MDD patients [78]. Moreover, there is a delay in getting therapeutic benefits after initiation of treatment even though there is a rapid increase in extracellular 5-HT levels as well as the necessity of chronic treatment [79]. The deficiency of monoamines cannot be considered as the sole pathophysiology of MDD, as the same alone cannot precipitate a full MDD clinical phenotype, albeit the occurrence in cases with a previous history [18]. MDD may also be developed following a reduction in noradrenaline and dopamine production as well as increased cholinergic transduction [6]. Hence, it is imperative that studies should be explorative beyond the traditional hypotheses of monoamines in depression.

Dopaminergic System

The dopaminergic hypothesis of depression relates the disorder to a deficiency of extracellular dopamine levels, especially a dysfunction of dopaminergic neurotransmission within the mesolimbic system contributing to symptoms such as anhedonia, psychomotor retardation and loss of motivation [80, 81]. Further, the expression and function of the D3 receptor, a dopamine receptor subtype located in pre-and postsynaptic membranes in brain areas regulating motivation and reward-related behavior, has also been reported to be downregulated in stress and depression [81, 82]. Moreover, the antidepressant effects of amineptine, bupropion and nomifensine were associated with selective inhibition of the dopaminergic uptake, suggesting the potential role of enhanced dopaminergic neurotransmission in the adaptive changes related to antidepressant activity [83-85]. Further, the antidepressant activity of aripiprazole and cariprazine might be mediated by their partial agonist activity at D3 receptors [86-88]. Furthermore, chronic antidepressant treatments have been reported to potentiate dopaminergic transmission in the mesolimbic system, characterized by an increased motor response to dopamine D2-like agonists [89]. Interestingly, dopaminergic neurotransmission is reported to be enhanced by other neural circuits [90]. For instance, the behavioral effects of antidepressant treatment could be mediated by serotonin-induced dopamine release as well as a brain-derived neurotrophic factor (BDNF) regulated expression of D3 receptors in brain areas [91, 92]. Thus, enhancing dopaminergic function, especially targeting D3 receptor might be a promising strategy to increase the therapeutic response in MDD [93].

Glutamate-GABA ergic System

Neurotransmission in the brain is regulated mostly by excitatory glutamate, then a relatively smaller inhibitory GABA component and a much smaller number of monoamine neurons [94]. As a corollary, both GABAergic deficit [95] and glutamate [94, 96] hypotheses were put forward to define the etiological sequel of MDD. GABAergic deficits in the brain contributing to MDD include the reduced concentration of GABA, altered function of GABAergic interneurons, changes in the levels and function of GABA receptors, along with alterations in changes in chloride homeostasis [97]. Whereas glutamate related changes involve malfunction in its release, clearance, metabolism, synaptic plasticity, postsynaptic receptor stimulation, intracellular signaling as well as associated changes in glial cell number and function [94]. Recent evidences suggest that monoamine-based antidepressant medications, as well as non-pharmacological interventions, might ultimately modify GABAergic and/or glutamatergic deficits also [24, 95]. While the conventional antidepressants take weeks or months, interestingly ketamine, a glutamate/N-methyl-D-aspartate (NMDA) receptor antagonist was shown to reverse symptoms of depression within 2 h of administration in treatment-refractory patients [98]. This might be the most promising development in the arena of the discovery of fast-acting antidepressants lately. The rapid effects of ketamine may be related to the mammalian target of rapamycin (mTOR) signaling and brain-derived neurotrophic factor (BDNF) [99, 100]. GLYX-13, a glycine site NMDA receptor partial agonist has also shown fast onset antidepressant action in clinical trials [101]. Moreover, a recent mouse study clearly indicated that at the metabolic level as well as at the molecular level, the pathways of glutamate and GABA, were affected in depression, even with a correlation between related gene expression [30]. These findings propose that GABA/glutamatergic system are key mediators of psychiatric pathology, hence should be given more focus in the treatment of depression.

Endocannabinoid (eCB) Signaling System

The endocannabinoid signaling (ECS) system mediates long and short term synaptic plasticity and is involved in reward, fear, memory extinction and stress [102]. As all of these fundamental processes are dysregulated in MDD, ECS deserves distinctive attention. A reduction in the ECS along with biochemical and genetic changes in ECS members (the cannabinoid receptors types 1 and 2 (CB1 and CB2), endocannabinoids anandamide (N- arachidonoyl-ethanolamine, AEA) and 2- arachidonoylglycerol (2-AG) and enzymes for their degradation, fatty acid amide hydrolase (FAAH) and monoacylglycerol lipase (MAGL) could underlie the depressive disorder. In particular, reduced CB1R-mediated ECS reduces drive toward reward, increases anxiety and decreases extinction of aversive memories; therefore pharmacological blockade of CB1R signaling is reported to predispose humans to depression, particularly those with previous depressive episodes [72]. The neuronal subpopulations also express the CB1 receptors, such as the dorsal raphe nucleus [103], the locus coeruleus [104], GABAergic interneurons (high CB1 receptor levels) and glutamatergic neurons (relatively low CB1 receptor levels) [105] and may be associated with the psychotropic and non- psychotropic effects of CB1 receptor activation. Moreover, increased ECS is hypothesized to mimic the effects of conventional antidepressants to enhance serotonergic and noradrenergic signaling, maintain hedonia during chronic stress, suppress HPA axis hyperactivity and increase hippocampal BDNF expression and neurogenesis. Further, acute or chronic treatment with pharmacological agents that modulate cannabinoid receptors, endogenous cannabinoids, synthetic nonspecific CB1/CB2 receptor agonists, selective CB1 receptor agonists and agents that inhibit endocannabinoids deactivation or cellular reuptake are reported to elicit antidepressant-like effects [72, 102]. Notably, CB1 receptor ligands may cause unwanted psychotropic effects while CB2 receptor activation may be safer, yet the evidences are controversial to commend application. FAAH inhibitors may have fewer adverse effects than direct CB1R agonists. Further, transient receptor potential vanilloid 1 (TRPV1) channel was proposed as an additional ECS “player” in mood regulation as they are found co-localized with CB1 receptors in multiple brain areas. Recently, N-Arachidonoylserotonin, a dual blocker of FAAH/TRPV1 was shown to reverse the stress-induced behavioral despair in rats through the normalization of the HPA axis dysregulation and in part with an elevation of anandamide levels [106]. Thus, studies have demonstrated that psychological and physiological stresses can induce changes in circulating endocannabinoids that are sensitive to the specifics of the stressor. The circulating endocannabinoids may be related to energy balance regulation, immune as well as endocrine responses to stress. For example, suppressed 2-AG could contribute to an increased inflammatory state seen in depression [102]. Further, endocannabinoids may be a mediator of the coincidence between cardiovascular morbidity and MDD [107]. Importantly, genetic polymorphisms in the ECS can confer resistance or vulnerability to depression. These data suggest that circulating endocannabinoids levels may be used in a personalized medicine approach to diagnose “low eCB” depression subtype that might be more amenable to ECS-mediated therapies [102].

Neurotrophic Factors

Postmortem brains from depressed patients revealed volumetric decreases characterized by decreased dendritic branching and spine density in neurons of the hippocampus along with subgenual and dorsolateral prefrontal cortex [108]. Studies have linked these alterations to the decrement in neurotrophic factors that regulate plasticity within the adult brain [109-111]. Neuroplasticity is a dynamic mechanism of the nervous system comprising of structural and functional components that enables its adaptation to environmental challenges via neuronal remodeling, especially across synapses. The neurotrophic factors include nerve growth factor (NGF), BDNF, neurotrophin (NT- 3 and NT-4/5), GDNF, Insulin-like Growth Factor (IGF), Vascular Growth Factor (VGF), Fibroblast Growth Factor (FGF) and S100 calcium-binding protein B (S100B) [112-114]. On interaction with their specific receptors, they mediate activation of various signaling pathways including phospholipase Cγ (PLCγ), phosphoinositide 3 kinase (PI3K)/Akt, tyrosine kinase and extracellular signal-regulated kinase (ERK) pathways, thereby regulating cell viability and synaptic functioning in the peripheral and central nervous systems (CNS) [112, 115-117]. A deficiency in these adaptive mechanisms of neuroplasticity might result in increased susceptibility to environmental challenges and resultant psychopathological effects [118]. These anti-neuroplastic changes include a reduction in proliferation of neural stem cells, neurotrophin levels, the survival of neuroblasts and immature neurons, spine density as well as damage in neurocircuitry, especially cortical-striatal-limbic circuits. There is increasing evidence of neurotrophic factor associated alterations including single nucleotide polymorphisms (SNPs) in different tissue samples, such as post- mortem brain tissue, cerebrospinal fluid, plasma, serum and buccal samples from patients with MDD [113, 119-121].

BDNF

Among the entire neurotrophin family, BDNF has emerged as a key mediator (giving rise to BDNF hypothesis of depression), not only because of its significant role in brain plasticity, but also of its role in the regulation of expression and secretion [122], defense against cell death induction [123] and maintenance of neural circuits [124]. Due to the complexity in the genomic structure of BDNF, there exist distinct transcriptional, translational and post-translational regulatory mechanisms. Evidently, for a single BDNF protein, there are distinct mRNA transcripts [125]. These transcripts vary in their distribution as well as translation efficacy and are differentially regulated by stress and antidepressant treatments, resulting in contrasting functional effects [126]. BDNF promotes the survival of neurons in the CNS by binding to tyrosine kinase receptor B (TrkB) receptors [127]. Adult neurogenesis within the subgranular zone of the dentate gyrus of the hippocampus is instrumental in spatial learning [128], pattern discrimination [129], contextual memory and mood regulation [130] along with the recognition of stressors that are threatening [129]. While acute, mild, controllable stress is reported to induce neurogenesis [131]; prolonged uncontrollable stress causes depletion of BDNF, decreasing pattern recognition, thereby, compromising the capacity to deal with stressors and cognitive functioning.

Remarkably, antidepressant drug treatments, including serotonergic and non-serotonergic medications [132-134] as well as non-pharmacological interventions such as electroconvulsive shock [135], transcranial magnetic stimulation [136], and exercise [137, 138], have been reported to increase neurogenesis. Hence, it could be assumed that non-serotonergic pathways might also be associated with antidepressant-stimulated neurogenesis [18]. However, a recent 6-week placebo-controlled trial in MDD patients treated with transcranial direct current stimulation and sertraline (separately and combined), revealed an absence of significant change in plasma levels of NT-3, NT-4, NGF and GDNF even with significant improvement in depressive symptoms [139]. Moreover, blocking neurogenesis per se did not develop symptoms of depression [140]. Even hypotheses relating to depression vulnerability and an SNP in BDNF, causing an aminoacid substitution of valine (Val) to methionine (Met) at codon 66 (Val66Met) are met with contrasting observations [120, 126]. The discrepancies in the role of neurotrophic factors in MDD pathophysiology and antidepressant efficacy may be due to the methodological differences associated with different study models particularly due to varying sample size, disease subtypes, ethnicity, genetics, dosing schedules, scales of assessment, sampling time and sample source.

HPA Axis

Based on the conviction that early life trauma and repeated psychosocial stress are potent risk factors for depression; studies have identified that HPA axis dysregulation and concomitant rise in glucocorticoid (GC) and corticotropin-releasing hormone (CRH) levels along with the failure of dexamethasone suppression of GC and decrease in GC sensitivity, as the hallmark features of MDD [63]. Studies over the last 40 years have identified stress-responsive HPA axis dysregulation to be the most consistent biological mechanism, with the toxicity arising from excessive GC release [141, 142].

HPA axis/CRH System

Hippocampus predominantly governs the release of CRH and arginine-vasopressin (AVP) from the PVN of the hypothalamus. Amygdala, bed nucleus of the stria terminalis (BNST) and other brain stem nuclei also provide input to the CRH neurons in the PVN [143]. Stress activates this neural circuit to release CRH into the hypothalamohypophyseal portal system, where it binds to CRH receptors on corticotrophs in the anterior pituitary, to further promote the synthesis of proopiomelanocortin (POMC) and release its post-translation products that include adrenocorticotropic hormone (ACTH) [144, 145]. ACTH is also released by the action of AVP on the vasopressin 1b (V1b) receptor and this pathway is predominant during chronic stress [146, 147]. Chronic stress also promotes CRH-mediated activation of locus ceruleus and sympathomedullary system to release norepinephrine and epinephrine, which in turn activates the CRH release [24]. CRH acts to coordinate the neuroendocrine, immune, autonomic, behavioral and cognitive responses of mammals to stress [145]. ACTH, in turn, stimulates the production of GCs (cortisol in humans and corticosterone in rodents) from the adrenal cortex into the bloodstream [148]. Once inside the cell, GCs interact with their receptors, the mineralocorticoid receptor (MR, type 1) and the glucocorticoid receptor (GR; type 2), to mediate numerous functions in almost every tissue of the body. It is interesting to note that around 5% of the genome are targets for GCs [148]. Moreover, GR is extensively expressed in neurons and glia. As GR possesses a low affinity towards GCs, under normal physiological concentrations, GR is barely occupied but becomes activated when there is rise in GC levels such as during diurnal maximum, ultradian pulses or a stress response [149-151]. The activated GR, in turn, stimulates an inhibitory feedback to HPA axis through the fornix, the bundle of axons linking hippocampus to hypothalamus [24, 152]. The feedback inhibition occurs on the level of CRF, AVP and ACTH, thereby leading to reduction of HPA axis activity [152, 153]. This regulatory feedback loop and resultant HPA axis function are found disrupted in ~50% of patients with affective disorders [154]. A decrease in hippocampal activity along with an increase in amygdala activity may cause the dysregulation of this negative feedback mechanism of GCs [6]. Further, BNST may be involved in the regulation of the HPA axis as its anterior section mediates activation causing CORT release while the posterior section mediates inhibition of HPA axis activation via GABAergic projections to the paraventricular nucleus. The associated changes in depressed patients include a rise in cortisol levels in plasma, saliva and urine as well as increased size and activity of adrenals and pituitary [155, 156]. There are reports that MDD patients with psychotic features, particularly, show elevated HPA axis activity as compared to nonpsychotic depressives or healthy controls [156]. The clinical manifestations, however, can be differential, and on this basis, MDD is categorized into two distinct subtypes, melancholic and atypical depression. 31-62% of MDD cases are reported to be atypical [157]. Melancholic and atypical subtypes are considered to be the clinical antithesis of each other [24].

Repeated administration of corticosterone (CORT) to rodents has shown to model chronic stress associated HPA axis dysregulation and concomitant depressive-like behavior; so is increasingly being utilized to evaluate potential antidepressant candidates [52, 158, 159]. These studies have revealed a significant association between hypercortisolemia and the reduction of BDNF levels in depression [160]. Excess GCs may be associated with hippocampal atrophy and volume reductions via GC receptor mediated decrease in subgranular zone proliferation [161]. Conversely, patients with Cushing’s syndrome exhibit a rise in circulating cortisol level as well as hippocampal atrophy and associated depressive symptoms [161]. Neurogenesis, in turn, can reduce the cortisol response to stress and its inhibition causes a delay in returning GC levels to its baseline [161]. In chronic uncontrollable stress, the reciprocal interactions of the HPA axis and neurogenesis can progress to a vicious cycle that can lead to depressive illness. Based on the light of the above evidences, the HPA axis is increasingly targeted for discovering new antidepressants such as antagonists of CRF1, GR and VIb receptors. However, successful clinical translation is yet to be achieved [153]. The major uncertainty is either to stimulate an impaired GR or to attenuate a hyperactive GR, to control MDD [162].

Immune-inducing Molecules

In the early 1990s, Maes et al. observed that there is an increase in proinflammatory cytokines in patients with MDD which correlated with the severity of illness and measures of HPA axis hyperactivity [163-166]. This led to the proposition of macrophage hypothesis of depression [167] (or cytokine hypothesis of depression [168]), according to which external and internal stressors prompt development of depressive behavior by immune hyperactivation characterized by the increase in levels of proinflammatory cytokines (interleukin-1 (IL-1), IL-6 and tumor necrosis factor-alpha (TNF-α)) and C-reactive protein (CRP). There can also be changes in function and number of natural killer (NK) cells, activation of cell-mediated immunity with the rise in CD4+/CD8+T cell ratio [169, 170], increase in number of activated T cells with markers CD2+ CD25+, CD3+ CD25+ and HLA-DR+ [171], rise in IL-2R level [172], increase in B cell subsets [173, 174] and changed Th1/Th2 cytokine and Treg balance [175, 176].

The immune molecules may be acting through multiple neural, cellular and humoral mechanisms in the brain to mediate either neuroprotective or neurodegenerative effects [177, 178]. Cytokines, which are humoral mediators of immunity, are produced primarily by microglia, but can also be produced by astrocytes and to some extent by neurons and oligodendrocytes, and can act as important modulators of mood [179]. External and internal stressors may also compromise the intestinal barrier leading to increased translocation of lipopolysaccharide (LPS) from gram-negative bacteria, which further mediates systemic as well as central inflammation, characterized by persistent activation of microglia and cytokines in the brain, resulting in sickness behavior (“leaky gut” hypothesis) [180]. Cytokine receptors within the CNS could be activated by peripherally as well as centrally synthesized cytokines [181]. Peripheral cytokines mediate the effects either by direct action on the brain after passive entry through the leaky regions in the blood-brain barrier (BBB) or active transport across BBB or binding to cell-surface proteins on endothelial cells, or via the transmission of cytokine signals through afferent vagus nerve fibers [182, 183]. Furthermore, interferon-α (IFN-α) injections were reported to cause depressive symptoms in 20–50% of cancer and hepatitis C patients [184], as well as a rise in the levels of sIL-2r, IL-6 and TNF-α in serum [185]. Accordingly, LPS, IL-1β, IFN-α or BCG injections in animals were used to develop inflammation-associated models of depression which further confirmed the relationship between inflammation, monoamine and neuroendocrine function [183]. In corroboration, pre-clinical and clinical studies revealed the effectiveness of conventional antidepressant treatments against cytokine-induced depression [182, 186].

Moreover, inflammatory molecules such as pro-inflammatory cytokines have the potential to be developed as a biomarker for MDD, especially for vascular depression where there is an increased risk for future cardiovascular diseases and depression [186]. For example, determination of the levels of CRP, IL-1β and TNF-α [186, 187] along with circulating leukocyte subsets [188] can serve as predictors of antidepressant treatment response in MDD. Thus, the association between inflammation and depression is possibly bidirectional, whose strength and direction may be determined by the specific cluster of symptoms, comorbidities, type of immune markers as well as population demographics. A divergent set of observation has also been reported which was based on immune suppression that includes reduction in levels and functions of NK cells and T cells [189]. Interestingly, in a recent study, clinical features of inflammation-associated depression were found to be age-related with young patients having a first depressive episode exhibiting a down-regulated monocyte gene expression profile; patients aged 28 years with a recurrent disease revealed upregulation [190]. Moreover, there are reports that immune suppression, as well as activation, can co-occur within the same MDD patient [191, 192].

Neurochemical, Neuroendocrine and Neuroimmune Interactions

The biochemical and physiological systems are not solitary mediators of MDD; rather interact with each other in a multidirectional network to mediate the progression of the disorder [193]. An elaboration of these mechanisms is needed to establish the linkup between different hypotheses of MDD and delineate the interactive matrix of pathophysiological components. The identification of the depressogenic role of indoleamine-pyrrole 2,3-dioxygenase (IDO) marked the fit of the cytokine and 5- HT hypothesis in MDD. Stress is reported to induce LPS, proinflammatory cytokines and oxidative stress, which in turn activates macrophages and other immunoregulatory cells to secrete IDO [186]. IDO catalyzes the rate-limiting step in the degradation of tryptophan to kynurenine, thus diverting tryptophan from the 5-HT synthetic route [183]. Kynurenine possess many immunomodulatory effects [194, 195] and is further metabolized into 3-hydroxykynurenine (generates free radicals causing oxidative stress), quinolinic acid (highly neurotoxic; directly damage neurons and function as a NMDA receptor agonist, provoke increase in oxidative and nitrosative stress, disrupt neuron glial communication and BBB integrity, inhibit ATP production by mitochondria), and kynurenic acid (neuroprotective; functions as antagonists of NMDA, α-amino-3-hydroxy-5-methyl-4- isoxazolepropionic acid and kainite receptors, also regulate glutamate and dopamine levels) [196]. These tryptophan catabolites (TRYCATs) are also formed following induction of tryptophan 2,3-dioxygenase (TDO) by GCs, which are found elevated in MDD. Thus the new ‘5- HT’ hypothesis’ of depression has been proposed according to which the activation of TRYCAT pathway was linked with the development of MDD through tryptophan depletion [197]. The reduction in plasma tryptophan level was also reported to be related to the decrease in serum albumin, to which it loosely binds. Kynurenine/tryptophan ratio and TRYCAT level were significantly associated with the affective symptoms observed in early puerperium. Further, in MDD, subchronic conditions with an activated counter anti-inflammatory response system (CARS) may exist, marked by a normal kynurenine/tryptophan ratio and reduced levels of plasma tryptophan [198-200].

Recent pathological evidences in MDD include activation of microglia and inflammasome mediated pathways [201, 202]. Microglia being the predominant immune cell in the brain, mediates immunosurveillance for pathogens, apoptotic cells and cellular debris [203]. The enduring presence of risk factors including oxidative stress, toxins, injury, vascular insults or genetic polymorphisms is reported to activate danger-associated molecular pattern (DAMP) detectors within the innate immune system, such as toll-like receptors (TLRs) and receptors for highly glycosylated end products (RAGEs), leading to chronic overproduction of proinflammatory cytokines, TNF- α, IL- 1β, and IL-6 [204]. Further, immune-mediated activation of pathogen-associated molecular pattern molecules (PAMPs) are also reported to contribute to the pathogenesis of MDD [183]. DAMPs signal through TLRs for the formation of the inflammasome, a multi-molecular signaling complex [205, 206]. Of particular relevance in MDD is the NLRP (Nucleotide-binding domain, Leucine-Rich Repeat, Pyrin domain-containing proteins)3 inflammasome, the only known inflammasome that requires a priming stimulus by GCs. This may serve as the link between stress, GC and neuroinflammation. NLRP3 inflammasome also requires a secondary stimulus by ATP [207]. NLRP3 acts by forming a scaffold with the adapter protein apoptosis-associated speck-like protein (ASC) which possesses a caspase recruitment domain, thereby mediating the formation of caspase-1, which in turn mediates formation and release of IL-1β [208]. IL-1β is an important mediator of sickness response and thereby coordinates peripheral and central host defense responses to danger [209]. Physiological alterations in ion fluxes, mitochondrial function, reactive oxygen species (ROS) levels and endosome integrity can activate NLRP inflammasomes [210]; suggestive of the involvement of these mechanisms in MDD [211]. Furthermore, DAMPs released from BBB endothelium and epithelial lining of the body may activate SNS as well as the HPA axis to mediate peripheral inflammatory response [212]. This pathway which is modulated by adrenergic and GC receptors on immune cells and associated inflammatory reflex is dysregulated in chronic stress due to altered GC receptor sensitivity [213-216]. It is important to note that, during a fight or flight response, an initial increase in GC level occurs which mediates anti-inflammatory response, followed by a persistent decrease of GC below the threshold level, promoting microglial sensitization, a mechanism to combat the subsequent immunologic threat. Further, psychological stress is reported to induce reactive granulopoiesis (increase in neutrophil number) and subsequent rise in the levels of circulating monocyte, especially, Ly6Chi cells. These monocytes are chemotactically attracted by the cytokines to distinct brain areas, activate the cytokine receptors and modulate synaptic plasticity [216, 217]. Based on these observations, it is highly feasible that the immune system and HPA axis function in an interactive matrix to integrate the physiological effects. Moreover, in MDD, inflammation may be linked to alterations in neuroplasticity via reduction of BDNF through direct action of cytokines [218-220] or cytokine-mediated effects on neurotransmitters such as glutamate [221] and GABA [222].

Cellular Stress Mechanisms

Oxidative and Nitrosative Stress

Oxidative and nitrosative stress (O&NS) is a result of the imbalance between the levels of ROS/ reactive nitrogen species (RNS) and antioxidants, leading to the damage of cellular constituents. According to past studies, it is assumed that metabolic stress together with O&NS, poses a greater risk of morbidity and mortality in MDD patients [223, 224]. In MDD, the levels of endogenous antioxidants such as reduced glutathione (GSH), catalase, superoxide dismutase (SOD), glutathione reductase (GR), glutathione peroxidase (GPx), coenzyme Q10, zinc and vitamin E are reported to be decreased in plasma and brain. Interestingly, MDD patients in a (sub)acute phase of illness showed a rise in peroxide levels, while the levels tend to become normal in chronic cases [225]. Moreover, susceptibility to oxidative stress in MDD may be affected by gender as higher plasma peroxide levels were more found in depressed males as compared to females [226]. Furthermore, O&NS may be causing (auto)immune responses by converting inactive autoepitopes to immunogenic neoantigens. Evidently, the rise in levels of plasma IgG antibodies against oxidized low-density lipoprotein as well as IgM-mediated responses against membrane fatty acids and NO modified amino-acids, are observed in MDD. The role of O&NS in MDD pathogenesis is further strengthened by its association with polymorphisms in O&NS genes, like catalase, SOD, cyclooxygenase (COX)-2 and myeloperoxidase (MPO). In corroboration with these findings, antidepressants have demonstrated the ability to improve antioxidant levels and reverse O&NS induced damage whereas numerous antioxidants have exhibited antidepressive effects [223]. It is noteworthy that ROS also has the role of secondary messenger and is fundamental for neurobiological processes such as cell growth, proliferation, differentiation, migration and adhesion, regulation of gene expression, immune responses, control of the intracellular redox-related signaling pathways and cell death [227]. Increased oxidative stress can result in mitochondrial dysfunction and MDD has been associated with dysfunctional mitochondria [228] and damaged mitochondrial ultrastructure [229]. As ROS is a by-product of monoamine oxidase activity, it has got an important role in the inactivation of monoaminergic neurotransmitters; hence might be significant in the pathogenesis of MDD [230].

While O&NS leads to the loss of adaptive potential of cells to the alterations in redox homeostasis, causing eventual cell death, inflammatory damage is reported to cause additive effects in the neuroprogression of MDD [227]. The two transcription factors, nuclear factor (erythroid-derived 2)-like 2 (Nrf2) and nuclear factor-κB (NF-κB) are key components of the O&NS and inflammation cascade; both are implicated in MDD [227, 231, 232]. Various signaling pathways, including mitogen-activated protein kinases (MAPK) [233] and PI3K/ AKT/GSK3/mTOR pathways [234] are also important in the neuroprogression of MDD. Further, in MDD, ROS generated by IFN-γ from macrophages is reported to decrease the levels of 5,6,7,8-tetrahydrobiopterin, a cofactor involved in the biosynthesis of monoamine neurotransmitters [233].

Endoplasmic Reticulum Stress

Psychological stress associated HPA axis activation, rise in cytokine level, dopamine auto-oxidation and glutamatergic receptor activation are reported to elicit cellular stress response [235]. A highly intense stressful situation may enhance glutamate neurotransmission that can result in neuronal hyperexcitability which in turn demands extra protein synthesis. This can result in a critical situation, called endoplasmic reticulum (ER) stress, if the ER fails to effectively fold normal proteins as per the increased demand in function [236, 237]. To overcome ER stress and to promote cell survival, ER normally possesses a sophisticated stress response mechanism, called the unfolded protein response (UPR) [238]. Over time, misfolded proteins are removed to the cytoplasm from ER for ER-associated degradation (ERAD). Thus, the ERAD and its molecular chaperones promote cellular survival under ER stress by preventing the accumulation of toxic protein aggregates [239]. However, an elevated systemic and brain ER stress response is reported to underlie depression in humans and animals [240, 241]. Moreover, ER stress-associated ailments such as cardiovascular disease have shown a higher incidence of depression [240], which is further elaborated by Gold et al., linking ER stress and para-inflammation in the development of affective illness through the dysregulation of a stress response. Therapeutic strategies such as valproate and lithium are reported to modulate ER stress-related targets in favour of cell rescue [237]. Novel targets for MDD therapy may include central insulin, klotho and peroxisome proliferator-activated receptor-γ (PPAR-γ) systems, as they are reported to contain ER stress [237]. Further, agonists of the sigma-1 receptor, a novel ligand-operated ER chaperone that regulates bioenergetics, free radical production, UPR and cytokine signaling, have been reported to improve depressive symptoms [235]. These findings imply that targeting ER stress associated components may provide novel targets for MDD drug development.

Mitochondrial Alterations

Mitochondrial dysfunction is also important in the pathogenesis of MDD and forms the basis of the “mitochondrial psychiatry model of depression”. The significant correlation came from the observations of psychiatric symptoms in mitochondrial illness and mitochondrial alterations in psychiatric diseases [241]. Modifications in translational products linked to the function of mitochondria reduced gene expression of mitochondrial DNA (mtDNA)-encoded transcripts and changes in components of electron transport chain might be the important links between mitochondrial dysfunction and MDD [242-244]. Notably, susceptibility to develop MDD can be maternally inherited through mtDNA as mtDNA sequence variants are capable of inducing mitochondrial dysfunction, which in turn predispose individual towards MDD development [245]. A mitochondrial function could be pharmacologically modified, as shown by the ability of imipramine and harmine to modulate creatine kinase and mitochondrial respiratory chain activities in rat brain areas in a dose and treatment-related manner [246].

POTENTIAL THERAPEUTIC CONSIDERATIONS

The search for novel therapeutic strategies is a “hot topic” in neuroscience and the first step forward is to validate and integrate the emerging pathogenetic mechanisms and newer targets into the current therapeutic understanding of MDD. Brain neuropeptides including CRF, urocortin, AVP, substance P, neuropeptide Y, neuropeptide S, galanin, oxytocin and pituitary adenylate-cyclase activating polypeptide; not only participate in stress physiology, but also have immense clinical relevance as novel targets in stress-induced MDD [247]. Further, the circadian rhythm is reported to be bidirectionally linked with depression. In particular, the development of symptoms or recurrence of seasonal depression is associated with changes in circadian rhythms [248]. Melatonin or bright light is reported to modulate circadian rhythm and can be used as an adjunct to classical antidepressants. Another novel approach is the development of antidepressants with intrinsic chronobiotic properties. One such promising candidate is agomelatine, which possesses joint MT1 and MT2 antagonism and 5- HT2C receptor inhibition [249]. An additional remarkable finding is the common transcriptional responses implicating distinct elements of the circadian clock by two rapid-acting antidepressant strategies, low-dose ketamine and sleep deprivation therapies; thereby projecting new targets for rapid and improved treatment based on chronopharmacological strategies [250]. Another possible approach that needs further scientific elaboration and validation is the use of omega-3 polyunsaturated fatty acids (PUFA) for both the prevention and treatment of depression [251]. The beneficial role of PUFA includes maintaining the brain structures and preserving their function by interacting with phospholipid metabolism and, hence, the modulation of signal transduction; as well as preventing or decreasing the inflammatory status occurring during depression [252]. In addition, these biological systems may be targeted to develop reliable biomarkers which in conjunction with new in vivo imaging techniques can be used to diagnose depression subtypes. More importantly, preclinical studies need to shed their exclusive focus on male subjects, incorporate females and examine sex as a variable to develop an effective treatment for all, since the higher prevalence of depression and comorbidities are found in women than men [43]. Stress-regulated transcriptional changes in the hypothalamus [26] and other related gene expressions in the prefrontal cortex [253] in the mouse are already indicative of the fact that these changes are sex-specific and ovarian hormone-dependent, warrants the sex variable in MDD therapy.

Novel therapeutics for treatment-resistant MDD are emerging such as modulators on glutamatergic, cholinergic, monoaminergic and opioid systems as well as nutraceuticals such as L-methylfolate and S-adenosyl-methionine [254]. Apart from previously mentioned ketamine and GLYX-13, pharmacological and non-pharmacological treatment methods such as electroconvulsive therapy, sleep deprivation, scopolamine and pindolol have been pursued a fast-onset antidepressant action as well as to be used in conjunction with classical antidepressants to decrease the delay in antidepressant onset [255]. Psychotherapeutic modalities such as cognitive-behavioral therapy (CBT) [2], cognitive behavioral analysis system of psychotherapy [256] and short term psychodynamic psychotherapy [256] are also being evolved into promising applications including treatment-resistant depression. Advances in neurosurgical techniques such as deep brain stimulation to the nucleus accumbens are also emerging to treat patients suffering from treatment-resistant depression [257]. Further, the development of treatment-specific biomarkers can guide antidepressant treatment selection via its potential to predict an individual’s improvement to a specific treatment and non-responsiveness to an alternative treatment [258]. Furthermore, considering the association of genetic variation with the differential risk to benefit ratio of antidepressants, implementation of pharmacokinetic pharmacogenetic guidelines for prescribing antidepressants can bring precision medicine to psychiatry [259].

CONCLUSION WITH FUTURE PROSPECTIVE

The overall discussion reveals that MDD, a multifactorial disorder, is strongly associated with stress-related pathophysiological components (Fig. ). Each component has its own role in MDD phenotypes (Fig. ). The heterogeneity in molecular adaptations, signaling pathways, cells, brain areas as well as the neurocircuitry implicated, is integral in the pathophysiological changes associated with stress and MDD. In light of this, it could be stated that the damaging effects of stress on biology and behavior might be caused cumulatively by molecular, structural and functional alterations. Therefore, it is imperative that the diagnosis as well as treatment of MDD, be characterized based on the multitude of systems involved viz. serotonergic, noradrenergic, gamma-aminobutyric acid, glutamatergic and endocannabinoid systems, neurotrophic factors, HPA axis, immune system as well as cellular stress mechanisms including sex difference. These can aid in developing multitargeted therapies with more promise as compared to the standard treatment. In addition, brain stimulation therapies like deep brain stimulation and vagal nerve stimulation along with new classes of antidepressant-like vortioxetine, a multimodal antidepressant that acts as 5-HT transporter inhibitor, a 5-HT3, 5-HT1D and 5-HT7 receptor antagonist, 5-HT1B receptor partial agonist and 5-HT1A receptor agonist, may hold better assurance than the conventional antidepressants [260-262]. As a breakthrough in decades, esketamine, the S-enantiomer of racemic ketamine was recently approved by the FDA to use in conjunction with an oral antidepressant, for the treatment of treatment-resistant depression in adults [263]. Further, the paradigm shift from standard ‘one-size-fits-all’ treatment plan to individualized and personalized treatment may become the state of the art direction considering the assessment of individual biological properties and treatment response of the patients to MDD. Additionally, these characterizations may aid in the identification of MDD subtypes and the development of sophisticated clinical, endocrine, immune, inflammatory, imaging and genetic biomarkers. This might enable diagnostic and therapeutic modalities beyond the currently utilized symptom-based approach. Over the next few years, significant advances are expected to occur in the understanding and treatment of MDD, which hold much promise as compared to conventional monoamine based treatment.