Stress, Microglial Activation, and Mental Disorders

Abstract

Microglia play a major role in immune response in the brain. Recent progress in studies for microglia suggests that stress causes morphological alterations in microglia and affects microglial humoral release and phagocytosis. In this review, we present a molecular mechanism by which stress impacts microglia. Then, we describe current findings for the involvement of microglia in stress-related mental disorders including posttraumatic stress disorder (PTSD), depression, and pain enhancement. We focus on preclinical and clinical studies. Preclinical PTSD studies using animal models with fear memory dysregulation show neuroinflammation by microglia and altered microglial phagocytosis, two imaging studies and a postmortem study assessing neuroinflammation in PTSD patients show contradictory results. Imaging studies suggest neuroinflammation in depressed patients, postmortem studies show no microglial inflammatory changes in non-suicidal depressed patients. Although it has been established that microglia in the spinal cord play a pivotal role in chronic neuropathic pain, several preclinical studies suggest microglia also participate in stress-induced pain. A clinical study with induced microglia-like (iMG) cells and an imaging study indicate neuroinflammation by microglia in fibromyalgia patients. We believe that progress in interactive research between humans and animals elucidates the role of microglia in the pathophysiology of stress-related mental disorders.

1. Introduction

Microglia is a glial cell that is widely distributed in the central nervous system occupying 10–15% of cells in the brain. Microglia are derived from the mesoderm as well as blood cells and peripheral immune cells. As other cells in the brain are derived from the ectoderm, microglia are similar in nature to peripheral immune cells and play an important role in immune response in the brain. In the steady-state, microglial protrusions are extended in a tree-shape to monitor the intracerebral environment, but when faced with infection, ischemia, exposure to harmful substances, trauma, etc., they are activated and changes morphologically. Activated microglia have enlarged cell bodies, thickened and shortened protrusions, and become amoebic. Microglia move to the target site and release humoral factors such as cytokines and neurotrophic factors [1]. In addition, microglia have a phagocytic ability similar to peripheral macrophages and have the function of digesting waste products in the brain.

Since 2006, rodent studies that report that stress causes microglial morphological changes in various areas of the brain have been accumulated [2, 3]. It has also been investigated how stress affects the release of cytokines and neurotrophic factors from microglia and the phagocytosis of synapses by microglia. Along with this, imaging studies focusing on microglia have been conducted in humans with stress-related mental disorders. Inflammatory changes in microglia in the human brain can be partially evaluated by positron emission tomography (PET) techniques targeting translocator proteins (TSPO). Depression and posttraumatic stress disorder (PTSD) are representative stress-related psychiatric disorders and are closely associated with suicide. Fibromyalgia is also closely related to stress in its onset and chronicity. Currently, the findings of PET studies targeting TSPO in patients with depression are accumulating, and PET studies targeting patients with PTSD and fibromyalgia have been reported since 2019. A few postmortem studies investigating the association between stress-related mental disorders and microglia have also been conducted.

In this chapter, we first describe the effects of stress on the release of cytokines and neurotrophic factors from microglia and the phagocytosis of synapses by microglia, and their molecular mechanisms. Second, we outline animal and human studies investigating the involvement of microglia in the pathologies of PTSD, depression, suicide, and stress-induced pain.

2. Microglia and stress

Some studies have reported that stress promoted the production and release of cytokines by microglia, while others have reported that it suppressed them. It seems to depend on the type and intensity of stress and the brain region where microglia are present. Water immersion restraint stress is a stress paradigm in which mice are confined to a conical tube and then immersed in water to the chest level. Ohgidani et al. reported that a single water immersion restraint stress for 2 h increased the production of tumor necrosis α (TNF-α), an inflammatory mediator, from microglia in the mouse hippocampus [4]. Chronic unpredictable stress (CUS) is a stress paradigm in which multiple stressors are applied daily, including cage rotation, radio noise, food or water deprivation, light on or off all day, single breeding, overcrowding, no bedding, and wet bedding. Wholeb et al. reported that 14-day chronic unpredictable stress reduced the production of TNF-α and interleukin-1β (IL-1β) in microglia in the mouse prefrontal cortex [5].

Increased production of damage-associated molecular patterns (DAMPs) in the brain such as high-mobility group box 1 (HMGB1), heat-shock protein 72 (HSP72), and ATP is a pivotal molecular mechanism by which stress promotes cytokine release from microglia. These DAMPs bind to toll-like receptors (TLRs) on microglial cell membranes to induce nuclear factor-κB (NF-κB) and increase the production of pro-IL-1β, IL-6, and TNF-α. In addition, DAMPs activate nucleotide-binding oligomerization domain, leucine-rich repeat, and pyrin domain protein 3 (NLRP3) inflammasomes in microglia that act on pro-IL-1β processing to increase IL-1β production [6]. Stress activates the hypothalamus-pituitary-adrenal (HPA) axis and sympathetic nerves, then glucocorticoid and noradrenaline increase in the brain. It is proposed that both glucocorticoid and noradrenaline regulate cytokine release from microglia, but their effects on microglia are complex. Glucocorticoid is considered to suppress cytokine release from microglia through suppression of NF-κB. On the other hand, a few studies have shown that administration of glucocorticoid to the hippocampus after inducing inflammation by kainic acid increased inflammatory cytokines and the number of microglia in the hippocampus [7, 8]. CXCR1 and CD200R are receptors that are expressed in microglia and act in a direction that suppresses microglial inflammatory changes. Glucocorticoid promotes inflammatory responses in microglia to future stress by reducing CXCR1 and CD200R expression and increasing HMGB1 release, which is referred microglial priming [6]. Noradrenaline acts on the β-receptor of microglia to promote their activation and stimulate cAMP/protein kinase leading to the release of IL-1β. On the other hand, when noradrenaline acts on the α-receptor of microglia, it works in the direction of suppressing their activation [9].

Brain-derived neurotrophic factor (BDNF), which is released from neurons and microglia, is involved in neurogenesis and neurite branching. It has been reported that various stress paradigms such as repeated restraint stress and CUS reduce BDNF expression in the hippocampus and prefrontal cortex of rodents [10, 11]. A recent study reported that when CUS was loaded into a rat model of stroke, CUS reduced BDNF release from microglia in the amygdala [12]. One of the possible mechanisms of this microglial-derived BDNF decrease is the decrease in binding of the cAMP response element-binding protein to the promoter region of the BDNF gene due to overactivation of NF-κB [13].

It has recently been suggested that the phagocytic capacity of microglia is also affected by stress. Several studies have reported that 14-day CUS increased the uptake of neuronal structures, including synapses, by microglia in the hippocampus and medial prefrontal cortex [5, 14, 15]. A recent study in which male mice were loaded with CUS extended to 28 days observed no increase of uptake of neuronal structures by microglia at 28-day loading despite an increase at 14-day loading [16]. This result indicates that repeated stress exposure changes microglial function dynamically. In addition to CUS, the effects of restraint stress exposure on microglial phagocytosis have been investigated. Piirainen et al. found that 10-day restraint stress enhanced the phagocytosis of pre-glutamatergic synapses by CD206-positive microglia in the hippocampus and reduced microglial-synaptic contact in the amygdala [17]. In another study with 7-day restraint, stress loading showed increased rat microglial process branching and contacts between microglial processes and synapses, using two-photon microscopy [18]. As a molecular mechanism behind stress-enhanced microglial synaptic phagocytosis, a pathway in which an increase in glucocorticoid enhances phagocytosis via a colony-stimulating factor 1 (CSF1) signal has been reported [15]. Given that CX3CR1 knockout mice were shown to inhibit stress-enhanced microglial phagocytosis [14], CX3CL1/CX3CR1 signaling may also be one of the candidates of the mechanism. We summarize the molecular mechanisms above mentioned in Figure 1.

Figure 1.

Schematic of the proposed molecular mechanism by which stress influences microglia.

3. Microglia and stress-related mental disorders

3.1 Microglia and PTSD

Fear memory of PTSD has several characteristics. First, fear extinction is impaired. Fear extinction refers to the diminished fear of a particular traumatic stimulus by learning that it is safe to be exposed to that stimulus. In case measuring the degree of fear extinction in animal experiments, after loading the foot electric shock stimulus in a chamber, researchers repeatedly expose the animal to the chamber or the sound that was heard when the stimulus was applied, without the stimulus, and observe whether the freezing behavior weakens or if it can be maintained once weakened. Lai et al. loaded single prolonged stress (restraint, forced swim, and ether anesthesia) to rats. Then, the authors gave a foot shock stimulus to rats in a specific chamber 7 days later to measure the degree of fear extinction to the chamber at a later date and examined changes in the inflammatory response and microglial cell number. SPS impaired fear extinction and increased the microglial cell number and the expression of HMGB1 and TLR4 in the amygdala. Intra-BLA administration of HMGB1 inhibitor or TLR4 antagonist normalized these behavioral and molecular changes [19]. A separate SPS study reported that the degree of fear extinction was inversely correlated with the number of IL-10 genes, which is anti-inflammatory, expressed by microglia in the prefrontal cortex [20]. Another recent study showed that microglial synaptic phagocytosis in the hippocampal dentate gyrus was enhanced by foot electric shock stimulation, which impaired fear extinction [21].

The second characteristic of PTSD fear memory is an excessive fear generalization. Fear generalization refers to showing fear response to stimuli that are similar to those that remind us of past traumatic experiences. Excessive fear generalization leads to perceiving an inherently neutral stimulus as dangerous and causes unnecessary anxiety and fear. In case measuring the degree of fear generalization in animal experiments, after loading the foot electric shock stimulus in a chamber, researchers put animals in a chamber that is different in color, shape, and odor from those used for electric shock stimulation, and measure the freezing behavior. Nguyen et al. focused on the interaction of neurons and microglia in the hippocampus via the IL-33 signal and investigated the effect of inhibition of that signal on fear generalization. The authors showed that when IL-33 released from neurons acted on the IL-33 receptor (IL1RL1) in microglia, it promoted extracellular matrix phagocytosis by microglia and increased synaptogenesis in the hippocampus, and that fear generalization was enhanced in IL-33 or IL1RL1 conditional knockout mice [22].

If the fear of a particular stimulus diminishes without going through the process of fear extinction, it is defined as forgetting of fear memory. Being not to be able to forget is the third feature of PTSD fear memory [23]. Memories are considered to be stored in engrams, a specific neuronal population. Wang et al. generated mice that express CD55 only in hippocampal engram cells.CD55 is supposed to suppress microglial phagocytosis by inhibiting complement pathways. In these mice, microglial phagocytosis for components of engram cells was reduced and fear forgetting was impaired [24].

The fourth feature of PTSD fear memory is that while the sensory information of the trauma is clearly preserved, the contextual information of when/where/why/how it happened is not well integrated. A protocol of animal experiments was proposed to investigate such characteristics. In the protocol, in an environment where the sound of a specific frequency is regularly generated, an electric foot shock stimulus is given to an animal in a chamber, and after a certain period of time, freezing time is measured for the chamber and the sound, respectively under a condition without a shock stimulus [25]. If the freezing response to sound is enhanced while the response to the chamber is not enhanced, it is considered to capture the fourth feature described above. Although the association between this feature and microglia has not been directly investigated, it was reported that glucocorticoid variability in the hippocampus and norepinephrine variability in the amygdala were associated with this behavioral change. Changes in microglial cytokine release by these stress hormones may be associated with the fourth feature of PTSD fear memory.

Since 2020, imaging studies and postmortem brain studies in PTSD patients have been reported. Bhatt et al. showed that expression of TSPO in the insula and ventromedial prefrontal cortex was reduced in patients with PTSD and suggested that less microglia that release neurotrophic factors might be behind it. In addition, they have genetically analyzed postmortem brain samples from female PTSD patients and found reduced expression of the microglial-related genes TNFRSF14 and TSPOAP1 in addition to TSPO in the prefrontal cortex [26]. Conversely, Deri et al. reported in a similar PET study that the severity of PTSD symptoms was positively correlated with TSPO expression in the hippocampus and prefrontal cortex [27]. We summarize animal, PET, and postmortem studies in Table 1.

Animal studies
Stress/manipulation	Behavioral change	Brain region	Cellular and molecular changes	References
SPS	Fear extinction ↓	AMY	Iba1 ↑, TLR 4 ↑, HMGB1 ↑
/HMGB1 inhibitor or TLR4 antagonist (intra BLA administration)	/Recovery		/Suppression	[19]
SPS	Fear extinction ↓	PFC	IL10 mRNA in microglia ↓	[20]
Foot shock	Fear extinction ↓	DG in HIP	Synaptic phagocytosis by microglia ↑
/CSF1R inhibitor	/Recovery		/Suppression	[21]
Foot shock /IL33 cKO IL1RL 1 cKO	/Fear generalization ↑	DG in HIP	/Phagocytosis of ECM by microglia ↓	[22]
Foot shock /CD55 expression in only engram cells	/Forgetting of fear ↓	HIP	/Phagocytosis by microglia for parts of engram cells	[24]
PET with TSPO
Subjects	Paper type	Brain region	TSPO availability	References
PTSD patients	Original article	Insula, VMPFC	Down	[26]
PTSD patients	Original article	HIP, PFC	Up	[27]
Postmortem study
Subjects	Brain region		Molecular changes	Reference
Female PTSD patients	PFC		TNFRSF14 ↓ TSPOAP1 ↓ TSPO ↓	[26]

Table 1.

List of papers of microglia and PTSD including animal studies, PET with TSPO, and a post mortem study.

AMY, amygdala; PFC, prefrontal cortex; DG, dentate gyrus; HIP, hippocampus; cKO, conditional knockout; ECM, extracellular matrix; VMPFC, ventromedial prefrontal cortex.

3.2 Microglia, depression, and suicide

CUS, social defeat stress, repeated restraint stress, and social isolation stress are known as stress-induced animal models of depression. In these models, immobility time during forced swimming or when hung upside down is evaluated as an index of depressive symptoms [28]. While many studies report that microglia cause neuroinflammation in these model rodents [29, 30, 31, 32], a few studies report reduced production of inflammatory mediators in microglia [5, 21]. However, depressive mood and suicidal ideation, which are important in the clinical setting of depression, are inherently subjective symptoms, and it is difficult to evaluate them from the behavior of model animals. In addition, it is not clear whether the phenomenon of suicide exists in animals other than humans, thus research on humans is indispensable for understanding the pathophysiology of depression and suicide.

Several PET studies using TSPO as a ligand have been conducted in depressed patients, and two systematic reviews have ever been reported. Gritti et al. examined nine original articles and reported that most studies suggested increased TSPO expression in the anterior cingulate gyrus, prefrontal cortex, hippocampal formation, and insula of depressed patients. In addition, the authors suggested treatment with antidepressants and cognitive-behavioral therapy might reduce TSPO expression [33]. Enache et al. performed a meta-analysis on six of the nine original articles above mentioned. The authors concluded that TSPO expression was increased in depressed patients in the anterior cingulate gyrus, prefrontal cortex, temporal lobe, insula, and hippocampus [34].

The results of postmortem brain studies in depressed patients examining microglial changes are mixed. One study showed an increase in the number of Iba-1 positive amoeboid-like microglia in the ventrolateral prefrontal cortex of depressed patients [35], while another study showed that the number of HLA-positive microglia in the amygdala did not change [36]. In the tryptophan-serotonin alternative pathway, the tryptophan-kynurenine pathway, microglia synthesize neurotoxic quinolinic acid. It has ever been reported that quinolinic acid expression is reduced in the hippocampus and ventrolateral prefrontal cortex of depressed patients [35, 37]. Several studies observed microglial changes in the brains of suicide victims. Steiner et al. found increases in HLA-DR-positive microglia in the dorsolateral prefrontal cortex, anterior cingulate gyrus, and mediodorsal thalamus, of suicide victims [38]. On the other hand, Brisch et al. found a decrease in HLA-DR-positive microglia in the dorsal raphe nuclei of non-suicidal depressed patients [39]. Schneider et al. observed an increase in CD68 highly positive microglia in the ventral prefrontal white matter of suicide victims [40]. In another study, the number of IBA-1-positive microglia did not change in the dorsal anterior cingulate gyrus of depressed suicide victims, but microglia in suicide victims had wider cell bodies than control groups [41]. In a recent study, Snijders et al. isolated and extracted microglia from the medial frontal gyrus, superior temporal gyrus, thalamus, and subventricular zone in the postmortem brain of depressed patients, and investigated gene and protein expression changes extensively. No inflammatory changes in microglia were detected in these regions, the expression levels of CX3CR1 and TMEM119 increased, and the expression levels of CD14 and CD163 decreased [42]. The authors hypothesize that these results reflect changes in microglial homeostatic function other than inflammation in depressed patients.

We are conducting reverse translational research to elucidate the dynamics of microglia in depression at the molecular level using the peripheral blood of patients. We performed a blood metabolome/lipidome analysis in patients with first-time depressive episodes who are not receiving medication and found that multiple metabolites in the tryptophan-kynurenine pathway, which are closely associated with microglial activation, correlate with the severity of depressive symptoms and the intensity of suicidal ideation [43]. In a separate study of peripheral blood samples from depressed patients, we evaluated nerve-derived exosomes in blood by the sandwich ELISA (enzyme-linked immune sorbent assay) method and found that IL-34 was increased in the patient group and that synaptophysin and TNF-α correlated with the severity of depression [44]. IL-34 is a cytokine essential for maintaining the function of microglia. We envision a process in which activated microglia damage synapses and lead to the formation of depressive symptoms. Additionally, we are developing our own technology to generate induced microglia-like (iMG) cells from human peripheral blood monocytes and obtained a US patent in 2018. Human iMG cells can be produced in 2 weeks by separating monocytes from the collected human peripheral blood and adding two types of cytokines, granulocyte colony-stimulating factor and IL-34. We analyzed gene profiling patterns of iMG cells from three patients with rapid cycling bipolar disorder during both manic and depressive states, respectively. We revealed that CD206 gene expression was upregulated in the depressive state compared to the manic state among all three patients [45]. We summarize PET, postmortem, and iMG studies in Table 2.

PET with TSPO
Subject	Paper type	Brain region	TSPO availability	References
Depressed patients	Systematic review	ACC, PFC Hippocampal formation, insula	Up	[33]
Depressed patients	Systematic review meta-analysis	ACC, PFC, HIP temporal lobe, insula	Up	[34]
Postmortem study
Subject		Brain region	Molecular changes	References
Depressed patients		AMY	No change in the number of HLA positive microglia	[36]
Suicide victims		DLPFC, ACC, mediodorsal thalamus	HLA-DR positive microglia ↑	[38]
Depressed suicides		dACC white matter	Microglia having a wider cell body ↑	[41]
Suicide victims		Ventral prefrontal white matter	CD68 highly positive microglia ↑	[40]
Depressed patients		HIP	Quinolinic acid ↓	[37]
Non suicidal depressed patients		DRN	HLA-DR positive microglia ↓	[39]
Depressed patients		VLPFC	Iba-1 positive microglia ↑ quinolinic acid ↓	[35]
Subject		Brain region	Molecular changes	References
Depressed patients		Medial frontal gyrus, superior temporal gyrus, thalamus, subventricular zone	CX3CR1 ↑, TMEM119 ↑ CD163 ↓, CD14 ↓ no change in HLA-DRA, IL6, IL1β
			^*These are expressions in isolated microglia	[42]
iMG study
Subject			Molecular changes	Reference
Depressed state (bipolar patients)			CD206 ↑ (compared to manic state)	[45]

Table 2.

List of papers of microglia in depression and suicide including PET with TSPO, postmortem studies.

ACC, anterior cingulate cortex; dACC, dorsal anterior cingulate cortex; PFC, prefrontal cortex; HIP, hippocampus; AMY, amygdala; DLPFC, dorsolateral prefrontal cortex; DRN, dorsal raphe nuclei, VLPFC, ventrolateral prefrontal cortex.

3.3 Microglia and stress-induced pain

It has been established that microglia in the dorsal horn of the spinal cord play major roles in the mechanism of chronic neuropathic pain. Activated microglia highly express P₂X₄ and P₂X₇ receptors, which enhance ATP/P₂ receptors signaling and increase IL-1β, TNFα, BDNF release, leading to increased glutamatergic receptor function and decreased GABA receptor function in dorsal horn neurons of the spinal cord. This is the mechanism of chronic neuropathic pain by microglial activation [46].

As a mechanism by which pain is enhanced by stress, changes at the respective levels of the spinal cord and the central nervous system can be considered. Several animal studies have investigated the relationship between stress-induced pain and changes in microglia. Sawacki loaded social defeat stress for 6 days to mice and evaluated pain behavior, gene expressions of inflammatory mediators in the spinal cord. Social defeat stress enhanced mechanical allodynia and increased the number of microglia and expressions of IL-1, TNF, TLR4, CC chemokine ligand2. Selective removal of microglia by CSF-1 inhibitor attenuated these changes [47]. Another series of studies using SPS reported that SPS also enhances mechanical allodynia, increases the number of microglia and expressions of inflammatory mediators in the spinal cord, and induces microglial priming there. Administration of respectively, angiotensin II type 1 receptor antagonist, alpha-7 nicotinic acetylcholine receptor agonist, and glucocorticoid receptor antagonist were shown to attenuate rat pain sensitivity and inflammatory changes in the spinal cord [48, 49, 50]. Thus, these receptors may be involved in the mechanism of stress-induced pain.

Activation of microglia in the central nervous system as well as in the spinal cord may also be involved in stress-induced pain. Intrahippocampal injection of minocycline was shown to normalize SPS-induced hippocampal microglial inflammatory changes and mechanical allodynia [51]. The rostral ventromedial medulla (RVM) is considered to be one of the brain regions that directly project to the dorsal horn of the spinal cord to regulate pain. In pathological pain, serotonin neurons in the RVM are excited to increase serotonin release in the spinal cord. Wei et al. induced postoperative chronic pain in rats by skin/muscle incision and retraction. It caused inflammatory changes in microglia and elevated serotonin levels in the RVM. Inhibiting microglial inflammatory changes in the RVM reduced rat pain sensitivity and serotonin levels in the spinal cord [52]. Given that it has been reported that chronic restraint stress increases serotonin levels in the RVM and increases pain [53], inflammatory changes in microglia in the RVM may be relevant to the stress-induced pain.

Fibromyalgia is characterized by a wide range of pain for a long period of 3 months or more, strong stiffness, and various symptoms such as severe fatigue, insomnia, headache, and depressed mood. Psychosocial stress is related to its onset and chronicity. A PET study targeting TSPO in patients with fibromyalgia was reported in 2019. The expression level of TSPO in the anterior and posterior middle cingulate cortex increased in the patient group and was also correlated with clinical symptoms [54].

In our laboratory, we generated iMG cells from patients with fibromyalgia and investigated the details of their activation at the cellular level. We found that patient-derived iMG cells have an increased ability to release TNF-α and it correlates with the degree of pain [55]. We summarize animal, PET, and iMG studies indicating microglial involvement in stress-induced pain and fibromyalgia in Table 3.

Animal studies
Stress/manipulation	Behavioral change	Region	Cellular and molecular change	References
Social defeat stress	Allodynia ↑	Spinal cord	IL-1β ↑, TNF-α ↑, CCL2 ↑ TLR4 ↑
/CSF1 inhibitor	/Suppression		/Suppression	[47]
SPS	Allodynia ↑	Spinal cord	Iba1 positive cells ↑ TNF-α ↑, IL-1β ↑, NFκB in microglia ↑, microglial priming ↑
/α7 nAChR agonist (intrathecal injection)	/Suppression		/Suppression	[49]
/AT1R antagonist (intrathecal injection)	/Suppression		/Suppression	[48]
/GR antagonist	/Suppression		/Suppression	[50]
SPS	Allodynia ↑	HIP	Iba1 positive cells ↑ TNF-α ↑, IL-1β ↑
/minocycline (intrahippocampal injection)	/Suppression		/Suppression	[51]
PET with TSPO
Subjects	Brain region		TSPO availability	Reference
Fibromyalgia	Anterior and posterior middle cingulate cortices		Up	[54]
iMG study
Subject			Molecular change	Reference
Fibromyalgia patients			TNFα release ↑	[55]

Table 3.

List of papers regarding microglia in stress induced pain, which include animal studies, PET with TSPO, and an study with iMG cells.

HIP, hippocampus.

4. Conclusion

Given that microglia play major roles in brain inflammation and express many receptors for neurotransmitters in addition to stress hormones, we speculate that microglia are deeply involved in the pathophysiology of stress-related psychiatric disorders. In this chapter, we described findings regarding microglia ever obtained from animal and human studies of stress-related psychiatric disorders. The animal study is an indispensable research method for stress-related psychiatric disorders because it can control the type and intensity of stress to be applied and the molecular mechanism related to the pathological condition and therapeutic mechanism can be investigated in detail. There is a limit to extrapolating human higher brain dysfunction from animal behaviors. On the other hand, the greatest advantage of a human-samples study is the ability to capture detailed psychopathological symptoms. There are also drawbacks in human imaging studies, postmortem brain studies, and studies using peripheral blood. Current PET imaging with TSPO ligands can measure only a small portion of microglial changes and has an inadequate resolution. Postmortem brain study is not easy to carry out due to various technical and ethical constraints. It is not clear how far we can infer events in the central nervous system from the results of studies using peripheral blood. It has been shown that human iMG cells made from monocytes in our laboratory share many features with microglia [56, 57]. We expect that iMG cells play a role as a biomarker for psychiatric disorders, and have the potential to reproduce in vitro the dynamics of microglia actually occurring in the brain. We are planning to generate iMG cells in patients with stress-related psychiatric disorders and perform morphological and molecular analysis, and further analyze the correlation with clinical findings such as diagnosis, various test scores, severity. We hope that progress in bidirectional research between animals and humans by making the best use of the strengths of each research method and improving the weaknesses elucidates the role of microglia in stress-related psychiatric disorders and develops treatment targeting them.

Acknowledgments

Our studies shown in this chapter were partially supported by Grant-in-Aid for Scientific Research on (1) Innovative Areas “Will-Dynamics” of The Ministry of Education, Culture, Sports, Science, and Technology, Japan (JP16H06403 to T.A.K.), (2) The Japan Agency for Medical Research and Development (AMED) (JP19dk0307047 & JP19dk0307075, JP19dm0107095, and JP21wm0425010 to T.A.K.), (3) KAKENHI - the Japan Society for the Promotion of Science (“Wakate A” JP26713039 and “Kiban A” JP18H04042 to T.A.K.) and (4) SENSHIN Medical Research Foundation (to T.A.K.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Conflict of interest

The authors declare no conflict of interest.