M2 Macrophage Transplantation Ameliorates Cognitive Dysfunction in Amyloid-β-Treated Rats Through Regulation of Microglial Polarization

Abstract

Alzheimer’s disease (AD) is the most common neurodegenerative disorder in the elderly population. Neuroinflammation induced by amyloid-β (Aβ) aggregation is considered to be the critical factor underlying AD pathological mechanisms. Alternatively activated (M2) macrophages/microglia have been reported to have neuroprotective effects in neurodegenerative disease. In this study, we characterized the neuroprotective effects of M2 macrophage transplantation in AD model rats and investigated the underlying mechanisms. Intracerebroventricular injection of Aβ_1 - 42 to rats was used to model AD and resulted in cognitive impairment, neuronal damage, and inflammatory changes in the brain microenvironment. We observed an increased interferon regulatory factor (IRF) 5/IRF4 ratio, resulting in greater production of classically activated (M1) versus M2 microglia. M2 macrophage transplantation attenuated inflammation in the brain, reversed Aβ_1 - 42-induced changes in the IRF4-IRF5 ratio, drove endogenous microglial polarization toward the M2 phenotype, and ameliorated cognitive impairment. Nerve growth factor (NGF) treatment reduced the IRF5/IRF4 ratio and induced primary microglial polarization to the M2 phenotype in vitro; these effects were prevented by tyrosine Kinase Receptor A (TrkA) inhibition. M2 macrophage transplantation restored the balance of IRF4-IRF5 by affecting the expression of NGF and inflammatory cytokines in the brains of AD model rats. This drove microglial polarization to the M2 phenotype, promoted termination of neuroinflammation, and resulted in improved cognitive abilities.

Keywords

Alzheimer’s disease M2 macrophage microglial polarization nerve growth factor neuroinflammation

INTRODUCTION

Alzheimer’s disease (AD), which is one of the most devastating neurodegenerative disorders in the elderly population, is characterized by loss of cholinergic neurons and progressive memory impairments [1]. The incidence of AD increases sharply with aging, rising to 32% in people aged 85 or older [2]. AD places a heavy burden on society and the economy due to the long disease course, high incidence rate, and cognitive dysfunction in affected individuals, which seriously impacts their daily life [2, 3].

Previous studies have shown that the main pathologies of AD are deposition of amyloid-β (Aβ) plaques, accumulation of intracellular hyperphosphorylated tau protein, and development of neurofibrillary tangles [4, 5]. In addition, neuroinflammation induced by Aβ aggregation is considered to be a critical factor in AD etiopathology [6]. The main manifestations of neuroinflammation are the activation of microglia and astrocytes, and the increased release of proinflammatory mediators [7]. A large body of evidence has demonstrated that inflammatory cytokine levels are elevated in the brain of patients with AD and transgenic AD animal models, and the number of activated microglia surrounding amyloid plaques appears to be increased as well [8, 9]. Further suggesting that neuroinflammation is an important component of AD pathogenesis, an epidemiologic survey has indicated that the use of non-steroidal anti-inflammatory drugs (NSAIDs) can to a certain extent delay AD progression [10].

Microglia are the innate immune cells of the central nervous system (CNS) and act as macrophages, in keeping with their derivation from myeloid immune cells [11, 12]. When activated, macrophages/ microglia will polarize to either the classically activated (M1) phenotype or the alternatively activated (M2) phenotype [13, 14]. The main roles of M1 macro- phages/microglia are to secrete pro-inflammatory cytokines, resist pathogens, and promote tissue degradation, while cells of the M2 phenotype produce anti-inflammatory cytokines and promote remodeling of damaged tissue [15]. The opposing functions of M1 versus M2 macrophages/microglia create a unity of opposites, the balance of which determines the nature of the inflammatory processes at play in the microenvironment. The injection of M2 macrophages donated beneficial effects on learning and memory of immune deficient mice, but the administration of M1 macrophage show no such amelioration [16]. In aging, microglia are skewed toward the M1 phenotype, thereby inhibiting the termination of inflammation [17, 18], which might be important for the development and progression of AD. Therefore, we hypothesized that if we could bias activated microglia toward the M2 phenotype, microenvironmental homeostasis of the brain would be better maintained and as a result, inflammatory damage might be attenuated.

Moreover, it has been recently reported that interferon regulatory factors (IRFs) can regulate macrophage/microglia polarization. In mouse macrophages, IRF4 increased the expression of specific proteins, such as Arg1, Ym1, and Fizza1, and induced the M2 phenotype [19]. Contrastingly, IRF5 has been shown to activate genes encoding pro-inflammatory factors, increase the percentage of cell exhibiting the M1 phenotype, and reduce the ratio of M2 macrophages [20, 21]. Thus, the balance between IRF5 and IRF4 can affect the polarization of microglia, and this may be an effective therapeutic target for AD [15].

Recent studies have shown that when pro-inflammatory factors including targets of IRF4/5 induce CNS inflammation, peripheral blood mononuclear cells penetrate the blood-brain barrier, infiltrate the brain parenchyma, and are then recruited to lesion sites, where they inhibit the inflammatory response [22]. Therefore, we postulated that boosting the recruitment of monocytes and inducing them to polarize to the M2 phenotype might represent an effective AD therapy. In this study, we aimed to demonstrate that there is a neuroprotective effect of M2 macrophage transplantation in a rat model of AD and to determine the primary underlying mechanisms.

MATERIALS AND METHODS

Reagents

Recombinant rat macrophage colony stimulating factor (M-CSF) and recombinant rat interleukin-4 (IL-4) were purchased from PeproTech Corporation (New Jersey, USA). Aβ_1 - 42 peptides were obtained from Invitrogen Corporation (California, USA). Mouse anti-neuron-specific nuclear protein (NeuN) monoclonal, mouse anti-Iba1/AIF1 monoclonal, and rabbit anti-IRF4 polyclonal antibodies were purchased from Millipore (California, USA). Rabbit anti-CD206/FITC conjugated antibody was obtained from Beijing Biosynthesis Biotechnology Co. Ltd. (Beijing, China). Rabbit anti-iNOS, rabbit anti-CD206, and rabbit anti-IRF5 polyclonal antibodies were purchased from Abcam (Boston, Massachusetts, USA). The Bio-Plex Pro Rat cytokine 24-plex panel was produced by Bio-Rad (Hercules, California, USA). Rat β-nerve growth factor (NGF) DuoSet and Recombinant rat β-NGF were obtained from R&D (Minnesota, USA). GW441756 was purchased from Sigma-Aldrich (Missouri, USA).

Animal

Forty-five male F344 rats (8–9 weeks of age) weighing 210–220 g were purchased from Vital River Laboratory Animal Technology Co. Ltd. (Beijing, China). All rats were housed under standardized conditions (22±1°C, 60±10% humidity and 12 h light/dark cycle), and provided free access to food and water ad libitum. The rats were randomly divided into 3 groups of 15 rats each: Control group (i.c.v. PBS + i.v. PBS), Model group (i.c.v. Aβ_1 - 42 + i.v. PBS), and M2-transplantation group (i.c.v. Aβ_1 - 42 + i.v. M2 Macrophages). Animal husbandry, care, and all experiments were carried out in accordance with the guidelines established by the National Institutes of Health Guide for the Care and Use of Laboratory Animals (publication no. 85–23, revised 1996) and were approved by the Animal Care Committee of the Peking Union Medical College and Chinese Academy of Medical Science (No. ACUC-A02-2014-002).

Aβ_1 - 42 peptide was dissolved in sterilized bi-distilled water at a concentration of 6 mg/mL and diluted to 2 mg/mL with 0.01 M phosphate-buffered saline (PBS), then aliquoted, and stored at –80°C. The peptide was incubated at 37°C for 4 days to aggregate it before injection. Rats were anaesthetized before being placed on the stereotaxic apparatus (RWD Life Science Co., Ltd, Shenzhen, China). A hole was drilled into the skull at the following coordinates according to the stereotaxic atlas of Paxinos and Watson: 1 mm anteroposterior from bregma, 1.6 mm lateral from the midline, 4.0 mm deep from the dura [23]. Then, 5.0μL of sterile PBS (Sham group) or an aggregated Aβ_1 - 42 suspension (Model group and M2-transplantation group) was administered by intracerebroventricular injection.

Bone marrow cell isolation

Rats were anesthetized with an intraperitoneal injection of sodium pentobarbital (50 mg/kg) and bone marrow cells were collected from the tibial and femoral shafts of male F344 rats. Macrophages were isolated from the bone marrow cell suspensions, then cultured and differentiated in DMEM containing 10% heat-inactivated FBS and 10 ng/mL recombinant M-CSF for 6 days. On the seventh day, macrophages were skewed to the M2 phenotype by adding 15 ng/mL recombinant IL-4 for 48 h. On the ninth day, M2 macrophages were labeled with DiI [24, 25], and then detached with 0.25% trypsin for 10 min. Five days after surgery, 10⁶ M2 macrophages per rat were injected via the tail vein.

Morris water maze

The Morris water maze task was carried out from Day 12 to Day 14 after Aβ_1 - 42 injection, as previously described [26]. The water maze tank was a black circular pool, 180 cm in diameter and 50 cm deep with non-toxic black dyed water maintained at 20±1°C. The maze was divided into four equal imaginary quadrants, and the target platform (diameter, 10 cm) was placed 1 cm beneath the surface of the water and fixed in the middle of one quadrant. The maze was located in a room with several visual stimuli hanging on the wall to provide spatial cues. Rats were placed in the pool and allowed to swim freely to the platform. If the animals did not find the target within 60 s, they were guided to it and the time was recorded as 60 s. The animals were allowed to stay on the platform for 15 s after stopping the trial. This procedure was repeated two times from the other two quadrants (beside the platform quadrant). During each training session, the time required to find the hidden platform (latency) was recorded using a video camera-based Ethovision System (Noldus, Wageningen, the Netherlands). Then the test was performed 24 h after the training session and the whole experiment lasted four consecutive days [27].

Y maze

Spontaneous locomotor activity was tested by Y maze on Day 16, as described previously [28]. Each rat was placed in the center of the apparatus and allowed to move freely in the maze during an 8-min session. Animal movement between arms was monitored, and the sequence and total number of arms entered were recorded. One entry was considered to be finished when all four paws of the rats were placed in one arm. Percentage alternation was determined as the number of triads containing entries into all three arms divided by the maximum possible alternations (the total number of arms entered –2)×100 [29].

Immunohistochemistry

Five rats of each group were deeply anesthetized with pentobarbital (60 mg/kg, i.p.), and then perfused with 0.1 M PBS followed by 4% paraformaldehyde. The brains were removed and postfixed in 4% paraformaldehyde. Two weeks later, brains were immersed in 0.05 M PBS containing 15%, 20%, and 30% sucrose for two days at 4°C, successively. Then, the brains were coronally sectioned (35μm) and stored in storage solution at 4°C.

For immunohistochemistry, the sections were permeabilized with PBST (0.3% Triton X-100 in 0.01 M PBS) for 30 min and nonspecific binding was blocked with 10% bovine serum (Hyclone, Australia). Then the sections were incubated at 4°C overnight with anti-NeuN (1:200), anti-ChAT (1:300), anti-Iba1 (1:200), anti-CD206 (1:200), or anti-iNOS (1:50) antibodies, and incubated with the corresponding biotinylated or fluoresced secondary antibodies at room temperature (RT) or 37°C for 60 min, respectively. The sections incubated with biotinylated secondary antibodies were reacted with 0.02% 3, 3’-diaminobenzidine and 0.01% H₂O₂ for approximately 1 min. After each incubation step, sections were washed three times with PBS. Cell counting was conducted in cortex on 20x microscopic images using a BX-53 Olympus microscope. Positive cells were counted using the Image-Pro Plus software in three separate fields of each region in the slices.

Analysis of tissue homogenate cytokine levels

The remaining animals were decapitated and the cortex was isolated, homogenized, and diluted in RIPA lysis buffer containing protease inhibitors. These homogenates were analyzed using the Bio-Plex cytokine assay system (R&D, Hercules CA, USA) to quantify the concentrations of pro- and anti- inflammatory cytokine factors. β-NGF levels were evaluated with an ELISA kit (Rat β-NGF DuoSet, R&D, Hercules CA, USA).

Primary microglia culture and identification

Primary microglia were isolated from cortex of neonatal rats at postnatal day one and digested by trypsin, as previously described [30]. After digestion, centrifugation, and resuspension, cells were cultured in DMEM medium with 10% fetal bovine serum (FBS) in Poly-L-Lysine-coated culture flasks. The medium was changed every three or four days. Ten days later, microglia were separated from astrocytes by shaking (120 rpm) for 2 h at 33°C. Then the cells were identified by mouse anti-Iba1 antibody (1:100, Millipore, MABN92). Briefly, microglia were fixed in 4% paraformaldehyde for 20 min, blocked with 10% FBS in PBS for 30 min, incubated with the primary antibody over night at 4°C, and then incubated with corresponding FITC-conjugated secondary antibody for 2 h at 37°C. Nuclei were marked by Hoechst staining. The cells were used for the subsequent experiments only when the purity of primary microglia was higher than 95% [31].

Western blot analysis

After preparing tissue and cell lysates, the protein concentration was measured by Bradford protein assay, and the supernatants were mixed with loading buffer and boiled for 8 min. Then the protein was subjected to SDS-PAGE on 10% acrylamide gels and transferred onto polyvinylidene difluoride (PVDF) membranes. The blots were blocked with 5% FBS in 0.1% TBST (pH 7.6) at room temperature for 1 h and then incubated with primary rabbit anti-IRF5, anti-IRF4, anti-CD206, or anti-iNOS polyclonal antibodies and mouse anti-β-Actin monoclonal antibody at 4°C overnight. The membranes were subsequently incubated with the corresponding HRP-conjugated secondary antibodies at room temperature for 1.5 h. The blot was developed with the LAS400 mini chemiluminescence system (Fujifilm, Tokyo, Japan). The band intensities were quantified using ImageJ2x software and expressed as values relative to those of controls.

Statistical analysis

SPSS 18 (IBM Corp., New York, USA) was used for calculations and statistical evaluations. Data were analyzed using One-Way ANOVA with Tukey’s post-hoc test and presented as the mean±SEM. Values of p < 0.05 were considered to be statistically significant.

RESULTS

M2 macrophage transplantation improved learning and memory in AD model rats

We obtained M2 macrophages by stimulating the bone marrow-derived cells with M-CSF and IL-4, and then stained these cells with FITC-anti-CD206 antibody. We found that the percentage of CD206⁺ cells was higher than 95%. This met our purity threshold for transplantation and these cells were therefore injected into recipient AD model rats (M2-transplantation group) via the tail vein, after labeling them with DiI in order to distinguish transplanted cells from endogenous M2 macrophages/microglia.

The Morris water maze test was used to examine spatial and reference memory. Escape latency, which indicates spatial memory, is the key indicator used to measure task performance [32]. In contrast to spatial memory, working memory is based on the short-term memory, and it is also the basis and initial phase of higher cognitive functions [33]. Rodents have a tendency to explore new environments, and the Y maze was used in this study to test rodent working memory in this context by altering the presentation between the arms [29]. We found that with training spatial and working memory was gradually formed and strengthened in each group. In the Morris water maze, escape latency was prolonged in the Model group compared with the Sham group, indicating spatial memory deficits (p < 0.05, Fig. 1A, B). In the Y-maze experiment, the proportion of AD model rats that entered three consecutive different arms was significantly lower than that of the Sham controls. However, compared to the Model group rats, M2-transplantation rats displayed an increase in both tests (p < 0.05, Fig. 1C).

In keeping with these results, immunohistochemical analysis of the density of total neurons (NeuN+ cells) in cortex and cholinergic neurons (ChAT+ cells) in nucleus basalis of Meynert (NBM) were significantly decreased in the Model group compared with rats in the Sham group (p < 0.01, Figs. 2 and 3). In the M2-transplantation group, these effects were reversed and the densities were restored to the level of the Sham group (p < 0.01). Taken together, these results demonstrated that M2 macrophage transplantation significantly improved learning and memory in AD model rats, and exerted neuroprotective effects.

M2 macrophage transplantation induced microglial polarization to M2 phenotype

Immunohistochemical analysis revealed that the densities of iNOS⁺ cells (M1 macrophage/microglia) and CD206⁺ cells in the cortex of rats in the Model group significantly increased and decreased, respectively (Fig. 4A–C). In the M2-transplantation group, these changes were dramatically reversed. Surprisingly, immunohistochemistry results showed no DiI⁺ cells in the brain (data not shown). Taken together, these results demonstrated that transplanted M2 macrophages were not trafficked to the brain, but were able to induce enhancement of the numbers of endogenous M2 microglia.

Transplanted M2 macrophages regulate the expression of IL-4, IL-5, and β-NGF in cortex of AD model rats

We tested the concentration of inflammatory cytokines and neurotrophic factors (NTFs) in the cortex of rats. By Bio-Plex and ELISA analysis, we found that compared with the Sham group, cortical IL-4, IL-5, and β-NGF levels of rats in the Model group decreased (p = 0.087, p < 0.05, p < 0.05, Fig. 5A–C). This indicates that administration of Aβ_1 - 42 led to the suppression of anti-inflammatory factors and triggered an inflammatory response and neuroinflammatory injury. All of these effects were reversed by M2 macrophage transplantation.

M2 macrophage transplantation influenced the expression of IRF5 and IRF4 in cortex of AD rats

In this study, western blot results showed that IRF5 levels in model rats were elevated dramatically (p < 0.05, versus sham group, Fig. 5D), and in the M2-transplantation group IRF5 levels decreased significantly. By contrast, cortical IRF4 expression significantly declined in AD Model rats (p < 0.01, versus Sham group, Fig. 5E), whereas M2 macrophage transplantation increased expression to the Sham level (p < 0.05).

Due to the opposing effects of IRF5 and IRF4 [19 –21], the ratio of IRF5/IRF4 is an important regulator of microglial polarization. We therefore compared the ratio of IRF5/IRF4 between groups, and found the IRF5/IRF4 ratio in cortex of rats in the model group was higher than that in the sham group (p < 0.05, Fig. 5F); this ratio decreased significantly in the M2-transplantation group (p < 0.05, Fig. 5F). All these observations indicate that transplantation promoted M2 phenotype polarization, confirming the above-mentioned results of immunohistochemical analysis.

NGF promoted primary microglial polarization to the M2 phenotype

These in vitro experiments demonstrated that compared with the control group, primary microglia iNOS and CD206 expression increased by administration of NGF or inhibition of its receptor, tyrosine Kinase Receptor A (TrkA). Microglia are very sensitive to environmental changes during activation and this influences polarization to different phenotypes, M1 or M2 [34]. We observed that the expressions of both iNOS and CD206 in each group increased. Several studies of immune cells outside of the nervous system have demonstrated that NGF is expressed by macrophages [35 –37], and in turn, at the same time, it affects the expression and function of macrophages [38]. The in vitro experiments in this study demonstrated that NGF induced changes in macrophage phenotype (both M1 and M2, Fig. 6A–C), as well as iNOS and CD206 expressions. Correspondingly, these phenomena were blocked by the TrkA inhibitor GW441756. We therefore speculate that NGF promotes microglial polarization to the M2 phenotype, resulting in a decreased M1/M2 ratio that promotes termination of the central inflammatory response. Furthermore, TrkA may mediate this process. Indeed, the variation of iNOS and CD206 levels further confirmed a role for NGF and TrkA in the regulation of microglial polarization, as we found that NGF and GW441756 treatment affected the expression of the proteins mentioned above. We speculate that this effect was related to the balance between the two microglial phenotypes. Together, these data demonstrate the importance of TrkA in NGF-mediated microglial polarization.

NGF may regulate the IRF5/IRF4 ratio

Based on the results of these in vivo and in vitro experiments, we assume that the IRF5-IRF4 balance plays a key role in the process of NGF-mediated microglial polarization. In further support of this conclusion, NGF significantly reversed the increase in IRF5 expression induced by Aβ_1 - 42 (p < 0.05, Fig. 6E, F). GW441756 inhibited this effect, indicating that TrkA mediates the effect of NGF on IRF5 expression. Aβ_1 - 42 did not have a significant impact on the level of IRF4 (Fig. 6E, G). Nevertheless, NGF had a tendency to increase IRF4 expression and this effect was blocked by GW441756. Moreover, NGF decreased the Aβ_1 - 42 induced increase of the IRF5/IRF4 ratio (Fig. 6H), and GW441756 inhibited this effect. Together, these results suggest that Aβ_1 - 42 induced a neuroinflammatory response that stimulated microglia to polarize into the M1 phenotype. In contrast, NGF, mediated by TrkA, induced polarization of microglia toward M2 and rebalanced the ratio of M1 and M2 phenotype cells.

DISCUSSION

Indeed, moderate inflammatory reactions can protect organisms from infection and injury, and the affected tissue will be restored to its original structure and functional state. But when inflammation is excessive and not terminated in a timely manner, it will lead to chronic disease. The main pathological mechanisms in these diseases have been considered to be secondary injury and abnormal conditions induced by inflammation [39]. Studies have shown that almost all neurodegenerative diseases, such as AD, Parkinson’s disease, and amyotrophic lateral sclerosis, are accompanied by neuroinflammation. Recently, researchers have increasingly realized that the termination of inflammation is an active process, determined by the milieu of inflammatory cells and cytokines secreted by them [40].

In the case of CNS inflammation, recent studies suggest that, monocytes will cross the blood-brain barrier and then differentiate into macrophages/microglia [22]. Depending on the internal environment, these cells will polarize to the M1 or M2 phenotype [13 , 42]. Our previous experiment demonstrated that M2 macrophage transplantation played no significant effect on the behavior and morphology of regular rats (data not shown), so the primary goal of this study was to determine if modulating the inflammatory state toward one in which an M2 state predominated would be beneficial in a rat AD model. In the present study, we demonstrated that M2 macrophage transplantation improved the cognitive abilities of AD model rats and this phenomenon was related to changes in the inflammatory environment of the brain.

Microglia are usually in a resting state in the healthy brain [43]. In the case of disease and injury, microglia are activated and polarize to either the M1 or M2 phenotypes. They then produce cytokines and chemokines that influence the cells in the surrounding microenvironment [44]. In this study, M2 macrophage transplantation elevated the percentage of M2 microglia in Aβ-treated rats. Many in vitro studies have shown that Aβ activates microglia and promotes the production of inflammatory cytokines, such as IL-1β, TNF-α, and reactive oxygen species; if not removed in a timely manner, these factors can cause neuroinflammatory injury [45 –47], just as we observed the decrease of neurons in cortex and cholinergic neurons in NBM in the model rats.

Moreover, when neuroinflammation develops, monocytes constantly repopulate the CNS; among these cells, T_H2 cells play a key role in neuroprotection and the termination of neuroinflammation [39]. During the process of microglial activation in the brains of patients with AD and AD model animals [48], T_H2 cytokines promote polarization to the M2 phenotype [49], in our study, this process was adjusted by M2 macrophage transplantation. In turn, M2 microglia are recruited to the inflammatory region to terminate inflammatory responses [50], and this maybe through the increased concentration of anti-inflammatory factors, such as IL-4 and IL-5 that we observed in our experiment.

In a previous study, we showed that there is crosstalk between inflammatory cytokines and NTFs in the brain of AD model rats, which was critically important for regulating homeostasis of the inflammatory environment [51]. By contrast, IL-4 and IL-5 are generated by T_H2 cells, and act as the anti-inflammatory factors [49], especially IL-4, which is an important mediator of T_H2 immune responses, and plays a key role in driving macrophage polarization to the M2 phenotype [26, 52]. After M2 macrophage transplantation, levels of these two factors were reversed. Meanwhile, the concentration of neurotrophic factor, nerve growth factor (NGF) increased, too. NGF and other NTFs were secreted by astrocytes, providing neurons with trophic and metabolic support [53] and protecting neurons from injury caused by pro-inflammatory cytokines [54].

Therefore, we speculate that Aβ_1 - 42 disturbed the balance between certain inflammatory factors and NTFs, and thereby greatly impaired microenvironmental homeostasis in the rat brain. As results microglial polarization was driven toward the M1 phenotype and accelerated neuroinflammation. However, M2 macrophage transplantation promoted homeostatic maintenance of the microenvironment in the CNS, and the polarization of microglia to the M2 phenotype resulting in the termination of the neuroinflammatory process. This likely explains the neuroprotective effect and improved learning and memory of AD rats that received M2 macrophage transplants.

We further evaluated how the inflammatory environment affected microglial polarization. The in vitro study showed that NGF affected the expression of IRF4 and IRF5, and these changes were consistent with CD206 and iNOS levels, which were biomarkers for M2 and M1 macrophage/microglia, respectively. It has been reported that, IRF5 is the main regulator of M1-polarization and is highly expressed in M1 macrophages [15 , 55]. IRF5 stimulates the release of pro-inflammatory cytokines, such as IL-12β and IL-23α, inhibits the expression of the anti-inflammatory cytokine IL-10 [56], and simultaneously promotes a decline in the proportion of M2 macrophages [15]. Studies have shown that silencing IRF5 in macrophages residing in injured tissue by nanoparticle-delivered siRNA can dramatically reduce expression of inflammatory M1 macrophage markers and promotes a resolution of inflammation [57]. Conversely, IRF4, which is mainly expressed in dendritic cells and macrophages [58], drives macrophage polarization to the M2 phenotype [19, 59]. Obviously, all these changes were confirmed in the in vivo experiment. Based on the above results, we speculated that M2 macrophage transplantation improved the CNS immune microenvironment and maintained the balance of IRF5 versus IRF4 levels, which ultimately led to improvements in learning and memory in AD model rats. And the results demonstrated that this process was mediated by neurotrophic factors, NGF.

M2 macrophage transplantation decreased the IRF4-IRF5 ratio and attenuated inflammation in the CNS microenvironment. These effects were mediated through expression of NTFs and inflammatory factors. These molecules in turn regulated the balance of M1 versus M2 microglial phenotype polarization and promoted the termination of the neuroinflammatory response, resulting in an attenuation of AD-associated neuroinflammatory pathology and declines in cognitive abilities in AD model rats. Our findings provide a new idea for AD therapy that regulating the systemic immune may improve the inflammatory status in the brain and contribute to the improvement of pathology and cognitive ability of AD patients. However, the mechanism underlying IRF5-IRF4– mediated regulation of microglial polarization needs to be investigated further. An appropriate way, such as new drugs, to adjust the immune system of human needs to be developed as well, and actually it is our ongoing work in the future.

Footnotes

ACKNOWLEDGMENTS

This study was supported by National Natural Science Foundation of China (Grant NO. 81100801) and the Scientific Research Foundation of Graduate School of Peking Union Medical College.

Authors’ disclosures available online ().

References

, Liu

, Frost

, Lemere

(2010) Amyloid-beta immunotherapy for Alzheimer’s disease. CNS Neurol Disord Drug Targets 9, 197–206.

Thies

, Bleiler

(2013) 2013 Alzheimer’s disease facts and figures. Alzheimers Dement 9, 208–245.

Reitz

, Mayeux

(2014) Alzheimer disease: Epidemiology, diagnostic criteria, risk factors and biomarkers. Biochem Pharmacol 88, 640–651.

Neniskyte

, Neher

, Brown

(2011) Neuronal death induced by nanomolar amyloid beta is mediated by primary phagocytosis of neurons by microglia. J Biol Chem 286, 39904–39913.

Zara

, De Colli

, Rapino

, Pacella

, Nasuti

, Sozio

, Di Stefano

, Cataldi

(2013) Ibuprofen and lipoic acid conjugate neuroprotective activity is mediated by Ngb/Akt intracellular signaling pathway in Alzheimer’s disease rat model. Gerontology 59, 250–260.

Wyss-Coray

(2006) Inflammation in Alzheimer disease: Driving force, bystander or beneficial response?. Nat Med 12, 1005–1015.

Liu

, Chan

(2014) The role of inflammasome in Alzheimer’s disease. Ageing Res Rev 15, 6–15.

Bauer

, Strauss

, Schreiter-Gasser

, Ganter

, Schlegel

, Witt

, Yolk

, Berger

(1991) Interleukin-6 and alpha-2-macroglobulin indicate an acute-phase state in Alzheimer’s disease cortices. FEBS Lett 285, 111–114.

Koronyo-Hamaoui

, Ko

, Koronyo

, Azoulay

, Seksenyan

, Kunis

, Pham

, Bakhsheshian

, Rogeri

, Black

, Farkas

, Schwartz

(2009) Attenuation of AD-like neuropathology by harnessing peripheral immune cells: Local elevation of IL-10 and MMP-9. J Neurochem 111, 1409–1424.

10.

Cudaback

, Jorstad

, Yang

, Montine

, Keene

(2014) Therapeutic implications of the prostaglandin pathway in Alzheimer’s disease. Biochem Pharmacol 88, 565–572.

11.

Davoust

, Vuaillat

, Androdias

, Nataf

(2008) From bone marrow to microglia: Barriers and avenues. Trends Immunol 29, 227–234.

12.

Lucin

, Wyss-Coray

(2009) Immune activation in brain aging and neurodegeneration: Too much or too little?. Neuron 64, 110–122.

13.

Zhou

, Huang

, Lin

, Zhan

, Kong

, Fang

, Li

(2014) Macrophage polarization and function with emphasis on the evolving roles of coordinated regulation of cellular signaling pathways. Cell Signal 26, 192–197.

14.

, Wang

, He

, Jiang

, Huang

, Liao

, Li

, Saavedra

, Zhang

, Pang

(2015) Telmisartan prevention of LPS-induced microglia activation involves M2 microglia polarization via CaMKKbeta-dependent AMPK activation. Brain Behav Immun 50, 298–313.

15.

Krausgruber

, Blazek

, Smallie

, Alzabin

, Lockstone

, Sahgal

, Hussell

, Feldmann

, Udalova

(2011) IRF5 promotes inflammatory macrophage polarization and TH1-TH17 responses. Nat Immunol 12, 231–238.

16.

Derecki

, Quinnies

, Kipnis

(2011) Alternatively activated myeloid (M2) cells enhance cognitive function in immune compromised mice. Brain Behav Immun 25, 379–385.

17.

Varnum

, Ikezu

(2012) The classification of microglial activation phenotypes on neurodegeneration and regeneration in Alzheimer’s disease brain. Arch Immunol Ther Exp (Warsz) 60, 251–266.

18.

Maher

, Clarke

, Kelly

, Nally

, Lynch

(2006) Interaction between interferon gamma and insulin-like growth factor-1 in hippocampus impacts on the ability of rats to sustain long-term potentiation. J Neurochem 96, 1560–1571.

19.

Satoh

, Takeuchi

, Vandenbon

, Yasuda

, Tanaka

, Kumagai

, Miyake

, Matsushita

, Okazaki

, Saitoh

, Honma

, Matsuyama

, Yui

, Tsujimura

, Standley

, Nakanishi

, Nakai

, Akira

(2010) The Jmjd3-Irf4 axis regulates M2 macrophage polarization and host responses against helminth infection. Nat Immunol 11, 936–944.

20.

Ouyang

, Negishi

, Takeda

, Fujita

, Taniguchi

, Honda

(2007) Cooperation between MyD88 and TRIF pathways in TLR synergy via IRF5 activation. Biochem Biophys Res Commun 354, 1045–1051.

21.

Koltsova

, Hedrick

, Ley

(2013) Myeloid cells in atherosclerosis: A delicate balance of anti-inflammatory and proinflammatory mechanisms. Curr Opin Lipidol 24, 371–380.

22.

Schwartz

, Shechter

(2010) Systemic inflammatory cells fight off neurodegenerative disease. Nat Rev Neurol 6, 405–410.

23.

Paxinos

, Watson

(2007) The rat brain in stereotaxic coordinates (6th edition), Elsevier Inc, Amsterdam.

24.

Cho

, Jung

, Lim

, Kwon

, Her

, Park

(2012) Bone marrow-derived, alternatively activated macrophages enhance solid tumor growth and lung metastasis of mammary carcinoma cells in a Balb/C mouse orthotopic model. Breast Cancer Res 14, R81.

25.

Gibbings

, Elkins

, Fitzgerald

, Tiao

, Weyman

, Shibao

, Fini

, Wright

(2011) Xanthine oxidoreductase promotes the inflammatory state of mononuclear phagocytes through effects on chemokine expression, peroxisome proliferator-activated receptor-{gamma} sumoylation, and HIF-1{alpha}. J Biol Chem 286, 961–975.

26.

Kawahara

, Suenobu

, Yoshida

, Koga

, Hyodo

, Ohtsuka

, Kuniyasu

, Tamamaki

, Sugimoto

, Nakayama

(2012) Intracerebral microinjection of interleukin-4/interleukin-13 reduces beta-amyloid accumulation in the ipsilateral side and improves cognitive deficits in young amyloid precursor protein 23 mice.. Neuroscience 207, 243–260.

27.

, Zhao

, Xu

, Wang

, Chen

, Wang

(2012) Isoflurane exposure during mid-adulthood attenuates age-related spatial memory impairment in APP/PS1 transgenic mice. PLoS One 7, e50172.

28.

Yabuki

, Shioda

, Maeda

, Hiraide

, Togashi

, Fukunaga

(2014) Aberrant CaMKII activity in the medial prefrontal cortex is associated with cognitive dysfunction in ADHD model rats. Brain Res 1557, 90–100.

29.

Devi

, Ohno

(2012) 7,8-dihydroxyflavone, a small-molecule TrkB agonist, reverses memory deficits and BACE1 elevation in a mouse model of Alzheimer’s disease. Neuropsychopharmacology 37, 434–444.

30.

Chhor

, Le Charpentier

, Lebon

, Ore

, Celador

, Josserand

, Degos

, Jacotot

, Hagberg

, Savman

, Mallard

, Gressens

, Fleiss

(2013) Characterization of phenotype markers and neuronotoxic potential of polarised primary microglia in vitro. Brain Behav Immun 32, 70–85.

31.

Guo

, Shi

, Tang

, Liang

, Xu

, Wang

, Qiao

(2014) The suppressive effects of gx-50 on Abeta-induced chemotactic migration of microglia. Int Immunopharmacol 19, 283–289.

32.

Vorhees

, Williams

(2006) Morris water maze: Procedures for assessing spatial and related forms of learning and memory. Nat Protoc 1, 848–858.

33.

, Ouyang

, Liu

, Bai

, Yi

, Tian

(2014) Increase of spike-LFP coordination in rat prefrontal cortex during working memory. Behav Brain Res 261, 297–304.

34.

Trial

, Cieslik

, Haudek

, Duerrschmid

, Entman

(2013) Th1/M1 conversion to th2/m2 responses in models of inflammation lacking cell death stimulates maturation of monocyte precursors to fibroblasts. Front Immuno 4, 287.

35.

Chiba

, Amagai

, Tanaka

, Katakura

, Matsuda

(2014) Nerve growth factor promotes killing of Leishmania donovani by macrophages through the induction of hydrogen peroxide. Microbes Infect 16, 702–706.

36.

Stoppiello

, Mapp

, Wilson

, Hill

, Scammell

, Walsh

(2014) Structural associations of symptomatic knee osteoarthritis. Arthritis Rheumatol 66, 3018–3027.

37.

Hepburn

, Prajsnar

, Klapholz

, Moreno

, Loynes

, Ogryzko

, Brown

, Schiebler

, Hegyi

, Antrobus

, Hammond

, Connolly

, Ochoa

, Bryant

, Otto

, Surewaard

, Seneviratne

, Grogono

, Cachat

, Ny

, Kaser

, Torok

, Peacock

, Holden

, Blundell

, Wang

, Ligoxygakis

, Minichiello

, Woods

, Foster

, Renshaw

, Floto

(2014) Innate immunity. A Spaetzle-like role for nerve growth factor beta in vertebrate immunity to Staphylococcus aureus. Science 346, 641–646.

38.

Fields

, Dumaop

, Langford

, Rockenstein

, Masliah

(2014) Role of neurotrophic factor alterations in the neurodegenerative process in HIV associated neurocognitive disorders. J Neuroimmune Pharmacol 9, 102–116.

39.

Schwartz

, Baruch

(2014) The resolution of neuroinflammation in neurodegeneration: Leukocyte recruitment via the choroid plexus. EMBO J 33, 7–22.

40.

Buckley

, Gilroy

, Serhan

, Stockinger

, Tak

(2013) The resolution of inflammation. Nat Rev Immunol 13, 59–66.

41.

Zhang

, Lund

, Mia

, Parsa

, Harris

(2014) Adoptive transfer of cytokine-induced immunomodulatory adult microglia attenuates experimental autoimmune encephalomyelitis in DBA/1 mice. Glia 62, 804–817.

42.

Sudduth

, Schmitt

, Nelson

, Wilcock

(2013) Neuroinflammatory phenotype in early Alzheimer’s disease. Neurobiol Aging 34, 1051–1059.

43.

Doens

, Fernandez

(2014) Microglia receptors and their implications in the response to amyloid beta for Alzheimer’s disease pathogenesis. J Neuroinflammation 11, 48.

44.

Kettenmann

, Hanisch

, Noda

, Verkhratsky

(2011) Physiology of microglia. Physiol Rev 91, 461–553.

45.

Fernandez

, Britton

, Rao

(2013) Potential immunotargets for Alzheimer’s disease treatment strategies. J Alzheimers Dis 33, 297–312.

46.

Zaheer

, Zaheer

, Thangavel

, Wu

, Sahu

, Yang

(2008) Glia maturation factor modulates beta-amyloid-induced glial activation, inflammatory cytokine/chemokine production and neuronal damage. Brain Res 1208, 192–203.

47.

Van Eldik

, Thompson

, Ralay Ranaivo

, Behanna

, Martin Watterson

(2007) Glia proinflammatory cytokine upregulation as a therapeutic target for neurodegenerative diseases: Function-based and target-based discovery approaches. Int Rev Neurobiol 82, 277–296.

48.

Zhang

, Li

, Yu

, Guo

, Meng

, Liu

, Xie

, Feng

, Xiao

, Ma

(2013) Rho kinase inhibitor fasudil regulates microglia polarization and function. Neuroimmunomodulation 20, 313–322.

49.

Lopez-Bravo

, Minguito de la Escalera

, Dominguez

, Gonzalez-Cintado

, del Fresno

, Martin

, Martinez del Hoyo

, Ardavin

(2013) IL-4 blocks TH1-polarizing/inflammatory cytokine gene expression during monocyte-derived dendritic cell differentiation through histone hypoacetylation. J Allergy Clin Immunol 132, 1409–1419.

50.

Shechter

, London

, Varol

, Raposo

, Cusimano

, Yovel

, Rolls

, Mack

, Pluchino

, Martino

, Jung

, Schwartz

(2009) Infiltrating blood-derived macrophages are vital cells playing an anti-inflammatory role in recovery from spinal cord injury in mice. PLoS Med 6, e1000113.

51.

, Song

, Zuo

(2011) The mechanism of memory impairment induced by Abeta chronic administration involves imbalance between cytokines and neurotrophins in the rat hippocampus. Curr Alzheimer Res 8, 410–420.

52.

, Xu

, Jin

, Yang

, Wang

, Yu

, Cao

(2012) The E3 ubiquitin ligase neuregulin receptor degradation protein 1 (Nrdp1) promotes M2 macrophage polarization by ubiquitinating and activating transcription factor CCAAT/enhancer-binding Protein beta (C/EBPbeta). J Biol Chem 287, 26740–26748.

53.

Medeiros

, LaFerla

(2013) Astrocytes: Conductors of the Alzheimer disease neuroinflammatory symphony. Exp Neurol 239, 133–138.

54.

, Zhao

, Gao

, Wei

, Yin

, Lau

, Chui

, Hoi Yu

(2011) Astrocytes: Implications for neuroinflammatory pathogenesis of Alzheimer’s disease. Curr Alzheimer Res 8, 67–80.

55.

Seki

, Brenner

(2008) Toll-like receptors and adaptor molecules in liver disease: Update. Hepatology 48, 322–335.

56.

Weiss

, Blazek

, Byrne

, Perocheau

, Udalova

(2013) IRF5 is a specific marker of inflammatory macrophages in vivo. Mediators Inflam 2013, 245804.

57.

Courties

, Heidt

, Sebas

, Iwamoto

, Jeon

, Truelove

, Tricot

, Wojtkiewicz

, Dutta

, Sager

, Borodovsky

, Novobrantseva

, Klebanov

, Fitzgerald

, Anderson

, Libby

, Swirski

, Weissleder

, Nahrendorf

(2014) In vivo silencing of the transcription factor IRF5 reprograms the macrophage phenotype and improves infarct healing. J Am Coll Cardiol 63, 1556–1566.

58.

Akbari

, Honma

, Kimura

, Miyakoda

, Kimura

, Matsuyama

, Yui

(2014) IRF4 in dendritic cells inhibits IL-12 production and controls Th1 immune responses against leishmania major. J Immunol 192, 2271–2279.

59.

Lee

, French

, Morgan

, French

(2014) The liver is populated by a broad spectrum of markers for macrophages. In alcoholic hepatitis the macrophages are M1 and M2. Exp Mol Pathol 96, 118–125.