Major depressive disorder: hypothesis, mechanism, prevention and treatment

Abstract

Worldwide, the incidence of major depressive disorder (MDD) is increasing annually, resulting in greater economic and social burdens. Moreover, the pathological mechanisms of MDD and the mechanisms underlying the effects of pharmacological treatments for MDD are complex and unclear, and additional diagnostic and therapeutic strategies for MDD still are needed. The currently widely accepted theories of MDD pathogenesis include the neurotransmitter and receptor hypothesis, hypothalamic-pituitary-adrenal (HPA) axis hypothesis, cytokine hypothesis, neuroplasticity hypothesis and systemic influence hypothesis, but these hypothesis cannot completely explain the pathological mechanism of MDD. Even it is still hard to adopt only one hypothesis to completely reveal the pathogenesis of MDD, thus in recent years, great progress has been made in elucidating the roles of multiple organ interactions in the pathogenesis MDD and identifying novel therapeutic approaches and multitarget modulatory strategies, further revealing the disease features of MDD. Furthermore, some newly discovered potential pharmacological targets and newly studied antidepressants have attracted widespread attention, some reagents have even been approved for clinical treatment and some novel therapeutic methods such as phototherapy and acupuncture have been discovered to have effective improvement for the depressive symptoms. In this work, we comprehensively summarize the latest research on the pathogenesis and diagnosis of MDD, preventive approaches and therapeutic medicines, as well as the related clinical trials.

Fig. 1

An outline map of the hypotheses to explain MDD pathogenesis. (I) HPA axis dysfunction hypothesis: high levels of glucocorticoids (GCs) play a core role in the pathogenesis of MDD, and thyroid hormone (TH) and estrogen are also involved in functions of the HPA axis; (II) the monoamine hypothesis: the functional deficiency of serotonin (5-HT), dopamine (DA) and norepinephrine (NE) are the main pathogenesis of MDD; (III) the inflammatory hypothesis: the neuro-inflammation induced by reactive oxygen species (ROS), inflammatory cytokines and inflammasomes activation is suggested to promote the occurrence of MDD; (IV) the genetic and epigenetic anomaly hypothesis: some genes are susceptible in the patients with MDD, including presynaptic vesicle trafficking (PCLO), D2 subtype of the dopamine receptor (DRD2), glutamate ionotropic receptor kainate type subunit 5 (GRIK5), metabotropic glutamate receptor 5 (GRM5), calcium voltage-gated channel subunit alpha1 E (CACNA1E), calcium voltage-gated channel auxiliary subunit alpha2 delta1(CACNA2D1), DNA methyltransferases (DNMTs), transcription levels of somatostatin (SST), fatty acid desaturase (FADS); (V) the structural and functional brain remodeling hypothesis: the postmortem results of patients with MDD are mostly associated with the reduced densities of glial cells in the prefrontal cortex (PFC), hippocampus, and amygdala; (VI) the social psychological hypothesis: the traumatic or stressful life events are the high risks of the occurrence of MDD. Adobe Illustrator was used to generate this figure

Fig. 2

Schematic illustration of the pharmacological mechanism of fluoxetine in astrocytes. Acute treatment with fluoxetine at low concentrations (green arrows) stimulates Src, which phosphorylates EGF receptors by activating 5-HT_2B receptors (5-HT_2BR) and activates the PI3K/AKT signaling pathway. AKT phosphorylation induced by fluoxetine at low concentrations inhibits the expression of cFos and subsequently decreases the expression of caveolin-1 expression (chronic effects), which in turn decreases the membrane content of PTEN, induces phosphorylation and stimulation of PI3K and increases the phosphorylation of GSK3β, thus suppressing its activity. At higher concentrations, fluoxetine (red arrows) stimulates metalloproteinases (MMP) by activating 5-HT_2BR and induces the release of growth factors, which stimulates EGF receptors and activates the mitogen-activated protein kinases (MAPK)/ERK_1/2 signaling pathway. ERK_1/2 phosphorylation induced by fluoxetine at high concentrations stimulates the expression of cFos and subsequently increases the expression of caveolin-1 (chronic effects), which inhibits PTEN/PI3K/AKT/GSK3β, ultimately leading to MDD like behavior. At high concentration, fluoxetine can also stimulate the activation of cPLA_2a by the transactivation of EGFR/MAPK/ERK_1/2 pathway, and the activated ERK_1/2 can also increases the expression of cPLA_2a at chronic treatments. In addition, the increased expression of cFos induced by fluoxetine can further increases the RNA editing of GluK2 by increasing the expression of ADAR2 at the chronic treatments, the function of the edited GluK2 by fluoxetine is down-regulated, which causes the acute glutamated induced Ca²⁺-dependent ERK phosphorylation is suppressed. Adobe Illustrator was used to generate this figure

Fig. 3

The molecular signaling schematic of cytokine hypothesis in the pathogenesis of MDD. The rodent performed the depressive like behaviors are impaired by some widely accepted risk factors, such as long-term sleep deprivation (SD), oxidative stress, lipopolysaccharide (LPS), ischemic damage and so on. Long-term SD can increase the extracellular ATP level, the latter inhibits the activation of AKT and the followed phosphorylation of FoxO3a by stimulating P2X7 receptors (P2X7R), the dephosphorylated FoxO3a translocates into the astrocytic nucleus, then the increased FoxO3a decreases the expression of 5-HT_2BR expression, which results the reduced phosphorylation of STAT3 which increases the activation of cPLA2 and the followed release of arachidonic acid (AA) and prostaglandin E2 (PGE2), finally causing the depressive-like behaviors. Thus, antidepressant fluoxetine activates ERK_1/2/cFos pathway by stimulating 5-HT_2BR and AC/cAMP/PKA pathway by activating GPCRs in order to increase the activation of CREB and the level of BDNF and TrkB, which can alleviate the depressive like behaviors induced by long-term SD.^, As well as, imipramine, other SSIRs, and TCAs can also play antidepressive roles by increasing BDNF mRNA expression in astrocytes. Ischemic stroke can trigger the increase of reactive oxygen species (ROS) which can induce the activation of NLRP3 inflammasome and the release of IL-1β/18, resulting in the neuroinflammation, however, Li⁺ salt inhibits the activation of GSK3β and increases the phosphorylation of FoxO3a by activating AKT, which promotes the more FoxO3a transportation from nucleus into cytoplasm, and the reduced FoxO3a in nucleus lacks the competition with TCF4, the increased complex level of β-catenin and TCF4 further stimulates the expression and the phosphorylation of STAT3, which further induce the mRNA and protein expression of UCP2, then in mitochondrion, the increased UCP2 suppresses the production of ROS and results in the deactivation of NLRP3 inflammasomeincreases. Superoxidation of Fe²⁺ stimulates an increase in ROS, resulting in the production of inflammatory cytokines (including IFN-γ, TNF-α, IL-1β, IL-6) and inducible nitric oxide synthase (iNOS). While, the treatments of oxidative stress (OS) can produce a large number of ROS, such as OH• and H₂O₂, resulting in neuronal impairments, while astrocytes can play their neuroprotective role by antioxidation. Additionally, LPS can also increase TNF-α, IL-1β, and IL-6 by TLR-4/NFkB/AP-1 pathway and cause depressive-like behavior. Adobe Illustrator was used to generate this figure

Fig. 4

The pathogenesis of MDD is closely related to synapses, astrocytes, microglia, and their interactions as well as interactions among organ. Genetic factors, stress and comorbidities are considered the most common pathogenic factors of MDD³. The traditional monoamine theory contends that MDD may cause by the deficits in monoamine neurotransmitters. Moreover, the other abnormal increase of neurotransmitters in the synaptic cleft, such as glutamate, GABA and ATP, has the high relationship with the pathogenesis of MDD.^, The interaction between neurons and glial cells can induce the oxidative stress, pro-inflammatory cytokines released, the reduction of neurotrophic factors. The microbiota-gut-brain axis is clearly disrupted in MDD.^, When liver dysfunction occurs and causes OS and neuroinflammation in the brain, which also contribute to the pathophysiology of MDD. Adobe Illustrator was used to generate this figure

Fig. 5

Schematic of prevention strategies for the occurrence and reoccurrence of MDD. An outline of various prevention strategies for MDD includes sleep improvement, dietary adjustment, exercise, and social intervention. Sleep disturbances have the high relationship with the occurrence of MDD, the anhedonia, anxiety and insomnia are the main symptoms of patients with MDD. The behavioral and educational strategies, cognitive reconstructing therapy and circadian rhythm support can be applied to improve sleep quality.^, Dietary adjustments are also suggested to have the potential effects to prevent the occurrence or re-occurrence of MDD, the improvement mechanism of diet may involve in the regulated immune-inflammatory responses, the improved gut-microbe-brain axis and synaptic plasticity.^,, In addition, xxercise is an effective way to improve neuroplasticity, to maintain neuroendocrine homeostasis, and to regulate neuroinflammation, in order to effectively prevent the occurrence or re-occurrence of MDD.^, Importantly, getting social support from family members, friends, coworkers and community members can be helpful for the MDD patients’ recovery, these social interventions can let patients get emotional support and improve their self-awareness.^, Adobe Illustrator was used to generate this figure

Fig. 6

The molecular mechanisms of tricyclic antidepressants (TCAs), selective serotonin reuptake inhibitors (SSRIs), serotonin/norepinephrine reuptake inhibitors (SNRIs) and ketamine. TCAs can inhibit the protein kinase C (PKC) pathway by blocking the H1 receptors (H1Rs),^, TCAs decreases the reuptake of dopamine (DA) by inhibiting dopamine transporters (DATs) in the presynaptic membrane, and increases the DA concentration in the synaptic gap, increase the effect of DA on dopamine receptors (DARs) of postsynaptic membrane. The activated DARs increase Ca²⁺ dependent CaMKII and CaMK4, as well as, the secretion of CREB.^, In another way, the stimulated DARs by DA can also activate the cAMP-PKA pathway, which in turn activates the levels of CREB and BDNF by stimulating MAPK/ERK_1/2 pathway. TCAs, SSRIs, and SNRIs can all inhibit the reuptake of 5-HT by SERTs, specially SSRIs have the selective inhibition on SERTs, which increase the concentration of 5-HT in the synaptic gap and play antidepressive roles by effecting on 5-HTRs in postsynaptic membrane,^, which also activate the cAMP-PKA pathway.^, Moreover, TCAs and SNRIs can also inhibit the reuptake of NE by NETs, which also increase the concentration of NE in the synaptic gap, and in turn activate the effect of NE on adrenoceptors (ADRs) and activate the cAMP-PKA pathway in postsynaptic membrane. Besides of the AC/cAMP/PKA pathway, the effect of NE on ADRs can also activate protein kinase B (Akt) phosphorylation and mammalian target of rapamycin complex 1 (mTORC1) by stimulating TrkB, in order to promote the secretion of postsynaptic density 95 (PSD95) and glutamate receptor 1 (GluR1). Ketamine works as the antagonist of NMDAR on GABAergic interneurons, it suppresses the excitation of subsets of GABAergic interneurons, which reduces the gamma aminobutyric acid (GABA) effects on gamma aminobutyric acid type B receptor (GABABR), and relieves the inhibition of GABAergic interneurons on the release of glutamate, the latter further stimulates AMPAR on postsynaptic membrane and increases the level of BDNF, even the release of BDNF stimulates the above TrkB/AKT/mTORC1 pathway.^, Adobe Illustrator was used to generate this figure

Fig. 7

The protocol and stressors used for CURD and CUMR. In order to establish the CUMR model, a combination of various stressors includes interference of constraint (a), damp bedding (b), cage shaking (c), tail suspension (d), forced swimming (e), and cage tilting (f). Among these six stressors, two were randomly selected and administered daily for a duration of 3 weeks. On the other hand, to establish the CUMR model, a set of behavioral constraints includes circadian rhythm (g), sleep deprivation (h), interference of cone light (i), interference of followed spotlight (j), high temperature stress (k), stroboscopic illumination (l), noise disturbance (m), and foot shock (n). Similarly, two out of these eight constraints were randomly chosen and applied daily for a period of 3 weeks