Microglia and the Aging Brain: Are Geriatric Microglia Linked to Poor Sleep Quality?

Poor sleep quality and disrupted circadian behavior are a normal part of aging and include excessive daytime sleepiness, increased sleep fragmentation, and decreased total sleep time and sleep quality. Although the neuronal decline underlying the cellular mechanism of poor sleep has been extensively investigated, brain function is not fully dependent on neurons. A recent antemortem autographic study and postmortem RNA sequencing and immunohistochemical studies on aged human brain have investigated the relationship between sleep fragmentation and activation of the innate immune cells of the brain, microglia. In the process of aging, there are marked reductions in the number of brain microglial cells, and the depletion of microglial cells disrupts circadian rhythmicity of brain tissue. We also showed, in a previous study, that pharmacological suppression of microglial function induced sleep abnormalities. However, the mechanism underlying the contribution of microglial cells to sleep homeostasis is only beginning to be understood. This review revisits the impact of aging on the microglial population and activation, as well as microglial contribution to sleep maintenance and response to sleep loss. Most importantly, this review will answer questions such as whether there is any link between senescent microglia and age-related poor quality sleep and how this exacerbates neurodegenerative disease.

Phagocytic elimination of synapses by microglia during sleep.

Synaptic strength reduces during sleep, but the underlying mechanisms of this process are unclear. This study showed reduction of synaptic proteins in rat prefrontal cortex (PFC) at AM7 or Zeitgeber Time (ZT0), when the light phase or sleeping period for rats started. At this time point, microglia were weakly activated, displaying larger and more granular somata with increased CD11b expression compared with those at ZT12, as revealed by flow cytometry. Expression of opsonins, such as complements or MFG-E8, matrix metalloproteinases, and microglial markers at ZT0 were increased compared with that at ZT12. Microglia at ZT0 phagocytosed synapses, as revealed by immunohistochemical staining. Immunoblotting detected more synapsin I in the isolated microglia at ZT0 than at ZT12. Complement C3- or MFG-E8-bound synapses were the most abundant at ZT0, some of which were phagocytosed by microglia. Systemic administration of synthetic glucocorticoid dexamethasone reduced microglial size, granularity and CD11b expression at ZT0, resembling microglia at ZT12, and increased synaptic proteins and decreased the sleeping period. Noradrenaline (NA) suppressed glutamate-induced phagocytosis in primary cultured microglia. Systemic administration of the brain monoamine-depleting agent reserpine decreased NA content and synapsin I expression in PFC, and increased expression of microglia markers, C3 and MFG-E8, while increasing the sleeping period. A NA precursor l-threo-dihydroxyphenylserine abolished the reserpine-induced changes. These results suggest that microglia may eliminate presumably weak synapses during every sleep phase. The circadian changes in concentrations of circulating glucocorticoids and brain NA might be correlated with the circadian changes of microglial phenotypes and synaptic strength.

Microglia may compensate for dopaminergic neuron loss in experimental Parkinsonism through selective elimination of glutamatergic synapses from the subthalamic nucleus.

Parkinson's disease (PD) symptoms do not become apparent until most dopaminergic neurons in the substantia nigra pars compacta (SNc) degenerate, suggesting that compensatory mechanisms play a role. Here, we investigated the compensatory involvement of activated microglia in the SN pars reticulata (SNr) and the globus pallidus (GP) in a 6-hydroxydopamine-induced rat hemiparkinsonism model. Activated microglia accumulated more markedly in the SNr than in the SNc in the model. The cells had enlarged somata and expressed phagocytic markers CD68 and NG2 proteoglycan in a limited region of the SNr, where synapsin I- and postsynaptic density 95-immunoreactivities were reduced. The activated microglia engulfed pre- and post-synaptic elements, including NMDA receptors into their phagosomes. Cells in the SNr and GP engulfed red fluorescent DiI that was injected into the subthalamic nucleus (STN) as an anterograde tracer. Rat primary microglia increased their phagocytic activities in response to glutamate, with increased expression of mRNA encoding phagocytosis-related factors. The synthetic glucocorticoid dexamethasone overcame the stimulating effect of glutamate. Subcutaneous single administration of dexamethasone to the PD model rats suppressed microglial activation in the SNr, resulting in aggravated motor dysfunctions, while expression of mRNA encoding glutamatergic, but not GABAergic, synaptic elements increased. These findings suggest that microglia in the SNr and GP become activated and selectively eliminate glutamatergic synapses from the STN in response to increased glutamatergic activity. Thus, microglia may be involved in a negative feedback loop in the indirect pathway of the basal ganglia to compensate for the loss of dopaminergic neurons in PD brains.

Hyperphosphatasemia in leiomyosarcoma of the uterus: Two case reports and a literature review.

Some tumors are known to produce alkaline phosphatase (ALP). Seven cases of uterine leiomyosarcoma were identified from the clinical records of Sakai City Medical Center from January 2006 to December 2014. Patients' ages ranged from 47 to 75 years (median: 58). Clinical stages were IB, IIB, IIIA, and IVB in four, one, one, and one cases, respectively. Of these, two were found to have hyperphosphatasemia before surgery, and the elevated ALP level decreased after surgery, and increased with disease recurrence. In the cases of hyperphosphatasemia, leiomyosarcoma cells showed positive staining for ALP. There was no correlation between serum ALP and lactate dehydrogenase, or ALP and cancer antigen 125. The combination of ALP, lactate dehydrogenase, and cancer antigen 125 may be more useful for diagnosis and follow-up of leiomyosarcoma.

Microglia modulate sleep/wakefulness under baseline conditions and under acute social defeat stress in adult mice.

Although sleep is tightly regulated by multiple neuronal circuits in the brain, nonneuronal cells such as glial cells have been increasingly recognized as crucial sleep regulators. Recent studies have shown that microglia may act to maintain wakefulness. Here, we investigated the possible involvement of microglia in the regulation of sleep quantity and quality under baseline and stress conditions through electroencephalography (EEG)/electromyography (EMG) recordings, and by employing pharmacological methods to eliminate microglial cells in the adult mouse brain. We found that severe microglial depletion induced by the colony-stimulating factor 1 receptor (CSF1R) antagonist PLX5622 (PLX) reversibly decreased the total wake time and the wake episode duration and increased the EEG slow-wave power during wakefulness under baseline conditions. To examine the role of microglia in sleep/wake regulation under mental stress, we used the acute social defeat stress (ASDS) paradigm, an ethological model for psychosocial stress. Sleep analysis under ASDS revealed that microglial depletion exacerbated the stress-induced decrease in the total wake time and increase in anxiety-like behaviors in the open field test. These results demonstrate that microglia actively modulate sleep quantity and architecture under both baseline and stress conditions. Our findings suggest that microglia may potentially provide resilience against acute psychosocial stress by regulating restorative sleep.

Metabolomic and pharmacologic analyses of brain substances associated with sleep pressure in mice.

Sleep pressure, the driving force of the homeostatic sleep regulation, is accumulated during wakefulness and dissipated during sleep. Sleep deprivation (SD) has been used as a method to acutely increase animal's sleep pressure for investigating the molecular changes under high sleep pressure. However, SD induces changes not only reflecting increased sleep pressure but also inevitable stresses and prolonged wake state itself. The Sik3Sleepy mutant mice (Sleepy) exhibit constitutively high sleep pressure despite sleeping longer, and have been useful as a model of increased sleep pressure. Here we conducted a cross-comparison of brain metabolomic profiles between SD versus ad lib slept mice, as well as Sleepy mutant versus littermate wild-type mice. Targeted metabolome analyses of whole brains quantified 203 metabolites in total, of which 43 metabolites showed significant changes in SD, whereas three did in Sleepy mutant mice. The large difference in the number of differential metabolites highlighted limitations of SD as methodology. The cross-comparison revealed that a decrease in betaine and an increase in imidazole dipeptides are associated with high sleep pressure in both models. These metabolites may be novel markers of sleep pressure at the whole-brain level. Furthermore, we found that intracerebroventricular injection of imidazole dipeptides increased subsequent NREM sleep time, suggesting the possibility that imidazole dipeptides may participate in the regulation of sleep in mice.

Mice Lacking Cerebellar Cortex and Related Structures Show a Decrease in Slow-Wave Activity With Normal Non-REM Sleep Amount and Sleep Homeostasis.

In addition to the well-known motor control, the cerebellum has recently been implicated in memory, cognition, addiction, and social behavior. Given that the cerebellum contains more neurons than the cerebral cortex and has tight connections to the thalamus and brainstem nuclei, it is possible that the cerebellum also regulates sleep/wakefulness. However, the role of the cerebellum in sleep was unclear, since cerebellar lesion studies inevitably involved massive inflammation in the adjacent brainstem, and sleep changes in lesion studies were not consistent with each other. Here, we examine the role of the cerebellum in sleep and wakefulness using mesencephalon- and rhombomere 1-specific Ptf1a conditional knockout (Ptf1a cKO) mice, which lack the cerebellar cortex and its related structures, and exhibit ataxic gait. Ptf1a cKO mice had similar wake and non-rapid eye movement sleep (NREMS) time as control mice and showed reduced slow wave activity during wakefulness, NREMS and REMS. Ptf1a cKO mice showed a decrease in REMS time during the light phase and had increased NREMS delta power in response to 6 h of sleep deprivation, as did control mice. Ptf1a cKO mice also had similar numbers of sleep spindles and fear memories as control mice. Thus, the cerebellum does not appear to play a major role in sleep-wake control, but may be involved in the generation of slow waves.

Synaptic elimination by microglia and disturbed higher brain functions.

Microglial cells in normal mature brains have long been considered to be cells that are resting until pathological events take place, activating the microglial cells. However, it is currently well known that the microglia that have resting ramified morphology in normal mature brains move actively in the brain parenchyma and phagocytose synapses, thus forming and maintaining neural circuits. This review summarizes recent findings on the roles of microglia in mature brains, with special reference to phagocytosis of synapses and higher brain functions. Phagocytic elimination of synapses by microglia may affect the balance between excitatory and inhibitory synaptic transmission, termed the E/I balance. When impaired synaptic elimination by microglia leads to disturbed E/I balance, various problems may follow in brain functions: in memory and cognitive functions, sleep, movement, social behaviors, and thinking. In addition to the roles of microglia in normal developing and mature brains, impaired microglial phagocytosis functions also correlate with disturbances to these higher brain functions that are caused by neurological, mental, and developmental disorders; Parkinson's and Alzheimer's diseases, autism spectrum disorder, attention deficit/hyperactivity disorder, and schizophrenia.

[A case of gastric cancer with bilateral ovarian metastasis after gastrectomy performed ovarian resection].

The patient was a 51-year-old female, who underwent radical surgery for cancer of remnant stomach in May 2006 (f-T4N0M0P0H0CY0, por 2, Stage IIIA, Cur B). Bilateral ovarian resection was performed in March 2009 for bilateral ovarian metastasis, so called "Krukenberg tumor" with peritoneal dissemination detected with CT scan after one-year adjuvant chemotherapy with S-1 (80 mg/m2, 4 weeks on and 2 weeks off). As of June 2010, she is alive and maintain her status quo after 6 courses of S-1 plus CDDP combination therapy (S-1 80 mg/m2, 3 weeks on, CDDP 60 mg/m2, started at day 8, ended 35 days later) followed by S-1 for residual peritoneal dissemination detected at operation. In ovary metastasis of gastric cancer, even if accompanied by peritoneum metastasis, ovarian resection as a reduction surgery followed by chemotherapy may improve survival.

Lister hooded rats as a novel animal model of attention-deficit/hyperactivity disorder.

Appropriate animal models are necessary to determine the molecular and cellular mechanisms underlying attention-deficit/hyperactivity disorder (ADHD). This study used a battery of behavioral tests to compare Lister hooded rats (LHRs), an old outbred strain frequently used for autistic epilepsy research, with Wistar rats and spontaneously hypertensive rats (SHRs), a commonly used ADHD model. The open field, elevated plus maze, light/dark box, and drop tests demonstrated that LHRs were the most hyperactive animals and displayed the most inattentive- and impulsive-like behaviors, which are characteristics of ADHD. The radial arm maze, social interaction, and Morris water maze tests showed that LHRs did not display deficits characteristic of autism or intellectual disability. Although LHRs did not show different monoamine contents, the mRNA expression levels of various genes linked to ADHD (Cdh13, Drd5, Foxp2, Maoa, Sema6d, Slc9a9, and St3gal3) and tyrosine hydroxylase protein expression levels were lower in the prefrontal cortex of LHRs compared with that of Wistar rats or SHRs. c-Fos, synapsin I, and tau protein expression levels in the prelimbic region of the medial prefrontal cortex were also increased in LHRs compared with Wistar rats. Atomoxetine and guanfacine, commonly used non-stimulant treatments for ADHD, ameliorated ADHD-like behaviors in LHRs. These results suggest that LHRs can serve as a better ADHD model to develop novel pharmacological interventions.

Aggravating effects of treadmill exercises during the early-onset period in a rat traumatic brain injury model: When should rehabilitation exercises be initiated?

Physical exercise is one of the best interventions for improving traumatic brain injury (TBI) outcomes. However, an argument has been raised regarding the timing at which physical exercise should be initiated. In this study, male Wistar rats were subjected to stab wounding of the right hemisphere to develop a TBI model and were forced to walk once on a treadmill at a 5-m/min pace at 24 h or 48 h after TBI for 10 min. Injured brain tissue was dissected after TBI to evaluate the effects of exercise. Behavioral abnormalities and motor impairment were assessed by various behavioral tests between 2 and 3 weeks after TBI. Exercise did not affect the circulating corticosterone levels and the weight of the adrenal glands. Exercise particularly that at 24 h, worsened the motor impairment of the left forelimbs. Quantitative reverse-transcription polymerase chain reaction showed that exercise at 24 h increased proinflammatory cytokines and chemokines on the third day while suppressing the proinflammatory reactions on the fourth day. Exercise at both time points decreased expression of transforming growth factor (TGF) ß1 and its receptor TGFßR1. Exercise at 24 h increased phosphorylation of IκB kinase on the fourth day, which may be correlated with the decreased effects of TGFß1. Even a low-intensity exercise activity could cause deleterious effects when it is initiated within 48 h after the onset of severe TBI, probably because of the resulting proinflammatory effects. Therefore, rehabilitation exercise programs should be initiated after 48 h of TBI onset.

Behavioral tests predicting striatal dopamine level in a rat hemi-Parkinson's disease model.

Parkinson's disease (PD) is a frequent neurodegenerative disease causing bradykinesia, tremor, muscle rigidity and postural instability. Although its main pathology is progressive dopaminergic (DArgic) neuron loss in the substantia nigra, motor deficits are thought not to become apparent until most DArgic neurons are lost, probably due to compensatory mechanisms that overcome the decline of DA level in the striatum. Even in animal PD models, it is difficult to detect motor deficits when most DArgic neurons are functional. In this study, we performed various behavioral tests (apomorphine-induced rotation, cylinder, forepaw adjustment steps (FAS), beam walking, rota-rod, and open-field), using 6-hydroxydopamine (OHDA) and lipopolysaccharide (LPS)-induced hemi-PD model rats with various striatal DA levels, to find the best way to predict the DA level from earlier disease stages. Different from the 6-OHDA-induced model, reduction in the striatal DA levels in the LPS-model was less significant. Among the behavioral tests, data from cylinder and FAS tests, which evaluate forelimb movements, best correlated with decline of the DA level. They also correlated well with decreased body weight gain. The beam and apomorphine tests showed less significant correlation than the cylinder and FAS tests. Open-field and rota-rod tests were not useful. Expressional levels of mRNA encoding tyrosine hydroxylase (TH), a marker of DArgic neurons, correlated well with the DA level. Metabotropic glutamate receptor 4 mRNA expression correlated with the striatal DA level and may be related to compensatory mechanisms. These results suggest that motor impairments of PD should be evaluated by forelimb movements, or hands and forearms in clinical settings, rather than movement of the body or large joints. The combination of cylinder and FAS tests may be the best to evaluate the rat PD models, in which many DArgic neurons survive.

The hypnotic bromovalerylurea ameliorates 6-hydroxydopamine-induced dopaminergic neuron loss while suppressing expression of interferon regulatory factors by microglia.

The low molecular weight organic compound bromovalerylurea (BU) has long been used as a hypnotic/sedative. In the present study, we found that BU suppressed mRNA expression of proinflammatory factors and nitric oxide release in lipopolysaccharide (LPS)-treated rat primary microglial cell cultures. BU prevented neuronal degeneration in LPS-treated neuron-microglia cocultures. The anti-inflammatory effects of BU were as strong as those of a synthetic glucocorticoid, dexamethasone. A rat hemi-Parkinsonian model was prepared by injecting 6-hydroxydopamine into the right striatum. BU was orally administered to these rats for 7 days, which ameliorated the degeneration of dopaminergic neurons in the substantia nigra pars compacta (SNpc) and alleviated motor deficits. BU suppressed the expression of mRNAs for interferon regulatory factors (IRFs) 1, 7 and 8 in the right (lesioned) ventral midbrain as well as those for proinflammatory mediators. BU increased mRNA expression of various neuroprotective factors, including platelet-derived growth factor and hepatocyte growth factor, but it did not increase expression of alternative activation (M2) markers. In microglial culture, BU suppressed the LPS-induced increase in expression of IRFs 1 and 8, and it reduced LPS-induced phosphorylation of JAK1 and STATs 1 and 3. Knockdown of IRFs 1 and 8 suppressed LPS-induced NO release by microglial cells. These results suggest that suppression of microglial IRF expression by BU prevents neuronal cell death in the injured brain region, where microglial activation occurs. Because many Parkinsonian patients suffer from sleep disorders, BU administration before sleep may effectively ameliorate neurological symptoms and alleviate sleep dysfunction.

Treadmill exercise ameliorates ischemia-induced brain edema while suppressing Na⁺/H⁺ exchanger 1 expression.

Exercise may be one of the most effective and sound therapies for stroke; however, the mechanisms underlying the curative effects remain unclear. In this study, the effects of forced treadmill exercise with electric shock on ischemic brain edema were investigated. Wistar rats were subjected to transient (90 min) middle cerebral artery occlusion (tMCAO). Eighty nine rats with substantially large ischemic lesions were evaluated using magnetic resonance imaging (MRI) and were randomly assigned to exercise and non-exercise groups. The rats were forced to run at 4-6m/s for 10 min/day on days 2, 3 and 4. Brain edema was measured on day 5 by MRI, histochemical staining of brain sections and tissue water content determination (n=7, each experiment). Motor function in some rats was examined on day 30 (n=6). Exercise reduced brain edema (P<0.05-0.001, varied by the methods) and ameliorated motor function (P<0.05). The anti-glucocorticoid mifepristone or the anti-mineralocorticoid spironolactone abolished these effects, but orally administered corticosterone mimicked the ameliorating effects of exercise. Exercise prevented the ischemia-induced expression of mRNA encoding aquaporin 4 (AQP4) and Na(+)/H(+) exchangers (NHEs) (n=5 or 7, P<0.01). Microglia and NG2 glia expressed NHE1 in the peri-ischemic region of rat brains and also in mixed glial cultures. Corticosterone at ~10nM reduced NHE1 and AQP4 expression in mixed glial and pure microglial cultures. Dexamethasone and aldosterone at 10nM did not significantly alter NHE1 and AQP4 expression. Exposure to a NHE inhibitor caused shrinkage of microglial cells. These results suggest that the stressful short-period and slow-paced treadmill exercise suppressed NHE1 and AQP4 expression resulting in the amelioration of brain edema at least partly via the moderate increase in plasma corticosterone levels.