Activation of the CB2 receptor system reverses amyloid-induced memory deficiency

Neurobiology of Aging 34 (2013) 791– 804 www.elsevier.com/locate/neuaging

Activation of the CB2 receptor system reverses amyloid-induced memory deficiency Jiang Wu1, Bihua Bie, Hui Yang, Jijun J. Xu, David L. Brown, Mohamed Naguib* Institute of Anesthesiology, Cleveland Clinic, Cleveland, OH, USA Received 17 December 2011; received in revised form 31 May 2012; accepted 15 June 2012

Abstract Cannabinoid type 2 (CB2) agonists are neuroprotective and appear to play modulatory roles in neurodegenerative processes in Alzheimer’s disease. We have studied the effect of 1-((3-benzyl-3-methyl-2,3-dihydro-1-benzofuran-6-yl) carbonyl) piperidine (MDA7)—a novel selective CB2 agonist that lacks psychoactivity— on ameliorating the neuroinflammatory process, synaptic dysfunction, and cognitive impairment induced by bilateral microinjection of amyloid-␤ (A␤)1– 40 fibrils into the hippocampal CA1 area of rats. In rats injected with A␤1– 40 fibrils, compared with the administration of intraperitoneal saline for 14 days, treatment with 15 mg/kg of intraperitoneal MDA7 daily for 14 days (1) ameliorated the expression of CD11b (microglia marker) and glial fibrillary acidic protein (astrocyte marker), (2) decreased the secretion of interleukin-1␤, (3) decreased the upsurge of CB2 receptors, (4) promoted A␤ clearance, and (5) restored synaptic plasticity, cognition, and memory. Our findings suggest that MDA7 is an innovative therapeutic approach for the treatment of Alzheimer’s disease. © 2013 Elsevier Inc. All rights reserved. Keywords: Amyloid; Cannabinoid receptor type 2 (CB2); CB2 agonist; MDA7; Microglia; Astrocyte

1. Introduction Alzheimer’s disease (AD) is an age-dependent neurodegenerative disorder characterized by progressive loss of memory and cognitive function. The brains of patients with AD are characterized by extensive deposits of extracellular aggregation of amyloid-␤ (A␤) peptides. These peptides form senile plaques and intracellular aggregation of hyperphosphorylated tau protein (Hardy and Selkoe, 2002). The abnormal accumulation of amyloid fibrils result in the progressive loss of neuronal circuitry, impairment of the synaptic plasticity in the brain, and eventual memory deficiency (Barry et al., 2011; Chacón et al., 2004; Klyubin et al., 2004; Li et al., 2011; Palop and Mucke, 2010; Ramírez et

1 Permanent address: College of Basic Medical Sciences, Dalian Medical University, Dalian, China. * Corresponding author at: Anesthesiology Institute, Cleveland Clinic, 9500 Euclid Avenue NE 6-306, Cleveland, OH 44195, USA. Tel.: ⫹1 216 444 6328; fax: ⫹1 216 636 2043. E-mail address: [email protected] (M. Naguib).

0197-4580/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.neurobiolaging.2012.06.011

al., 2005; Schmid et al., 2009; Srivareerat et al., 2011). In addition, amyloid fibrils activate the inflammatory pathway, characterized by the activated microglia and astrocytes seen in the brains of patients with AD (Akiyama et al., 2000; Eikelenboom et al., 2006; Wyss-Coray, 2006). It is possible that neuroinflammation could induce beneficial immune responses resulting in the phagocytosis of amyloid fibrils in an attempt to limit the development of the disease (WyssCoray, 2006). However, prevailing evidence suggests that neuroinflammation could be a driving force in accelerating AD through the production of the proinflammatory chemokines, cytokines, and neurotoxins by the activated microglia and astrocytes in the brain (Masliah, 2008; McGeer et al., 1988; Perry et al., 2010). The cannabinoid receptor (CB) family currently includes 2 cloned metabotropic receptors: CB1 (found predominantly in the brain) (Matsuda et al., 1990), and CB2 (found primarily in the peripheral immune system) (Munro et al., 1993) and to a lesser degree in the central nervous system (Van Sickle et al., 2005) and microglia (Núñez et al., 2004).

792

J. Wu et al. / Neurobiology of Aging 34 (2013) 791– 804

In vivo injection of A␤1– 42 into the hippocampus rat brain resulted in an increase in the endocannabinoid levels (van der Stelt et al., 2006), and anandamide has been shown to prevent A␤ toxicity in cell culture (Milton, 2002), suggesting that cannabinoids may play a neuroprotective role in AD. With the exception of a small population of neurons located in the brain stem and the cerebellum, healthy brain tissue does not express CB2 receptors (Van Sickle et al., 2005). Rather, CB2 receptors are upregulated in reactive microglial cells in AD, Huntington’s disease, simian immunodeficiency virus-induced encephalitis, HIV encephalitis, and multiple sclerosis (Benito et al., 2003, 2005, 2008; Ramírez et al., 2005). This finding suggests that the upregulation of CB2 receptors tends to attenuate the activation of early proinflammatory microglial signaling pathways associated with AD. In vitro studies demonstrated that CB2selective agonists blocked microglia-mediated neurotoxicity after A␤ is added to rat cortical cocultures (Ramírez et al., 2005). Furthermore, intracerebroventricular administration of nonselective cannabinoid receptor agonist WIN 55,212-2 to rats prevented A␤-induced microglial activation, cognitive impairment, and loss of neuronal markers (Ramírez et al., 2005). CB2 agonists are neuroprotective and they have the advantage of lacking the psychotropic adverse effects normally seen with CB1 agonists. In contrast, CB1 agonists appear to have no beneficial effects on AD neuropathology and behavioral deficits in a mouse model of AD (Chen et al., 2010). Taken together, studies show that in the settings of AD, microglia and astrocytes become fully reactive, initiating a proinflammatory cascade that results in the release of potentially neurotoxic substances, including cytokines, which lead to degenerative changes in neurons. CB2 is also upregulated during this process. We hypothesize that the CB2 receptor functions in a negative-feedback loop and that the administration of a CB2 agonist can (1) promote clearance of A␤, (2) ameliorate the neuroinflammatory response to A␤, and (3) prevent A␤-induced microglial and astrocyte activation, cytokine production, loss of synaptic plasticity, and cognitive impairment. Our previous work established 1-((3-benzyl-3-methyl2,3-dihydro-1-benzofuran-6-yl) carbonyl) piperidine (MDA7) as a novel, blood–brain barrier-permeant, and highly selective CB2 agonist (Fig. 1) (Diaz et al., 2009; Naguib et al., 2008, 2012). MDA7 lacks activity in CB2 knockout mice and its effects in rats and C57BL/6 wild type mice are antagonized by CB2 antagonists (Naguib et al., 2012). The neuroprotective effect of MDA7 was found to be mediated through prevention of glial activation in vivo and in in vitro models (Naguib et al., 2012). The potential effect of MDA7 in other glial-related pathophysiological conditions such as AD remains unexplored. In this study, we demonstrate that MDA7 exerts a neuroprotective effect by blunting neuroinflammatory processes that are pivotal for the development of A␤-induced neurotoxicity. We found that MDA7 can: (1) promote A␤

Fig. 1. Chemical structure of 1-[(3-benzyl-3-methyl-2,3-dihydro-1-benzofuran-6-yl) carbonyl] piperidine (MDA7), a selective cannabinoid type 2 (CB2) receptor agonist.

clearance, (2) prevent increases in levels of glial fibrillary acidic protein (GFAP) in astrocytes and of CD11b in microglia (indicative of neuroinflammatory process activation) that are typically seen in the hippocampus of A␤-injected rats; (3) decrease the production of interleukin (IL)-1␤; (4) ameliorate the A␤-mediated suppression of glutamatergic transmission in the hippocampus of A␤-injected rats; and (5) prevent cognitive impairment on spatial memory performance using the Morris water maze test in the A␤-injected rats. Because this CB2 agonist prevented A␤-induced neuroinflammation and its downstream consequence—synaptic plasticity and cognitive impairment—we hypothesize that the CB2 receptor functions in a negative-feedback loop and that its activation can induce A␤ clearance, and blunt neuroinflammatory responses and cognitive impairment induced by A␤ fibrils. 2. Methods 2.1. Animals and treatment protocol All animal procedures were approved by the Animal Care and Use Committee of Cleveland Clinic. Animals were housed the Institutional Biological Rodent Unit on a 12/12 hours light/dark cycle with water and food pellets available ad libitum. Adult male Sprague–Dawley (Charles River, Wilmington, MA, USA) rats weighing 200 –250 g were used, and all experiments were performed during the light cycle. Animals were divided into several groups (40 –50 rats per group). The A␤1– 40 group received bilateral intracerebral microinjection of A␤1– 40 fibrils once and saline intraperitoneally (i.p.) daily for 14 days. The A␤1– 40 ⫹ MDA7 group received bilateral intracerebral microinjection of A␤1– 40 fibrils once and 15 mg/kg MDA7 i.p. daily for 14 days. Other cohorts received smaller doses of MDA7 (5 or 1.5 mg/kg) i.p. daily for 14 days to determine the dose– response characteristics for MDA7. In other groups, AM630 (a CB2 antagonist) 5 mg/kg i.p. was administered daily for 14

J. Wu et al. / Neurobiology of Aging 34 (2013) 791– 804

days either alone (A␤1– 40 ⫹ AM630 group) or 15 minutes before MDA7 (15 mg/kg i.p.) administration (A␤1– 40 ⫹ MDA7 ⫹ AM630 group) to rats injected with A␤1– 40 fibrils. MDA7 group and control group received bilateral intracerebral microinjection of artificial cerebrospinal fluid (130 mM NaCl, 2.6 mM KCl, 4.3 mM MgCl2 and 1.8 mM CaCl2) once and 15 mg/kg MDA7 or saline i.p. daily for 14 days, respectively. Intraperitoneal administration of AM630, MDA7 or saline started on the same day after intracerebral microinjection of A␤1– 40. 2.2. Microinjection of amyloid fibrils into the hippocampal CA1 area As described in a previous study, rats were anesthetized with sodium pentobarbital (45 mg/kg i.p.) and restrained in a stereotaxic apparatus (Bie et al., 2010). A␤1– 40 fibrils were formed as described previously (Chacón et al., 2004). A␤1– 40 fibrils (10 ␮g/3 ␮L) or 3 ␮L of artificial cerebrospinal fluid were injected stereotaxically and bilaterally into each hippocampus (coordinates; bregma ⫺3.5 mm anteroposterior, ⫾ 2.0 mm mediolateral, and ⫺3.0 mm dorsoventral) (Paxinos and Watson, 1998) using a 10 ␮L Hamilton syringe with a 27-gauge stainless steel needle at a rate of 0.5 ␮L/min. This experimental model has been used for studying AD (Ahmed et al., 2010; Chacón et al., 2004; Shin et al., 1997). 2.3. Morris water maze test A cohort of rats (n ⫽ 10 per group) was tested 15 days after surgery as described elsewhere (Chacón et al., 2004). Testing was conducted in all groups at the same time of the day (9:00 AM–12:00 noon). The experimental apparatus consisted of a circular pool (120 cm in diameter, 45 cm high). An invisible platform (15 cm in diameter, 35 cm high) was placed 1.5 cm below the surface of the water. The water temperature was kept at 24 °C–28 °C. The pool was located in a test room, and many clues external to the maze were visible from the pool, which could be used by the rats for spatial orientation. The position of the cues were kept constant throughout the task. Each rat underwent 4 trials per day for 5 days. During each trial (memory acquisition), the rats were placed randomly at 1 of 4 fixed starting points and were allowed to swim for 120 seconds or until they escaped from the water by reaching the platform. The platform was located in the same position throughout the test in the middle of 1 quadrant. The start location was moved to a different quadrant in each trial so that no single start location was used in consecutive trials. In each training session, the latency to escape to the hidden platform was recorded. After the rats reached the platform, they were allowed to rest there for 20 seconds, and they then were removed from the pool. An intertrial interval of at least 4 minutes was used to ensure that each rat’s performance was not impaired by fatigue. After completing 4 trials, the rats were removed from the pool, dried, and returned to their home cage. On

793

day 6, all rats were subjected to 1 probe trial (memory retrieval) in which the platform was removed, and each animal had 60 seconds to search the pool for the platform. All behavioral testing was performed by only 1 experimenter (Dr. Wu) who was blinded to the different treatment groups. After the behavioral test, the rats were sacrificed, and the hippocampus tissues were collected for further study. 2.4. Hippocampal slice preparation Brain slices containing hippocampus were prepared as previously described (Bie et al., 2009a, 2009b, 2010, 2011b). Brain slices used for electrophysiological recording were obtained from all rats (n ⫽ 10 –12 rats per group) that had been subjected to the Morris water maze test. All rats were anesthetized with pentobarbital and then euthanized by decapitation. The brain was quickly removed and cut on a vibratome (Leica VT1000S, Buffalo Grove, IL, USA) in cold (4 °C) physiological saline to obtain coronal slices (300 ␮m thick) containing the hippocampus. A single slice was submerged in a shallow recording chamber and perfused with warm (35 °C) physiological saline (126 mM NaCl, 2.5 mM KCl, 1.2 mM NaH2PO4, 1.2 mM MgCl2, 2.4 mM CaCl2, 11-mM glucose, and 25 mM NaHCO3, saturated with 95% O2 and 5% CO2, pH 7.3–7.4). Slices were maintained at approximately 35 °C throughout the recording experiment. 2.5. Whole-cell patch clamp recordings in hippocampal slices The hippocampal slices were allowed to recover for 1 hour after which they were visualized using an upright microscope with infrared illumination. Whole-cell voltageclamp recordings from the CA1 area were taken using an Axopatch 200B amplifier (Molecular Devices) with 2– 4-M⍀ glass electrodes containing the following internal solution (mM): K-gluconate or cesium methanesulfonate, 125; NaCl, 5; MgCl2 1; EGTA (ethylene glycol tetraacetic acid), 0.5; Mg–ATP (Magnesium adenosine triphosphate), 2; Na3GTP (Guanosine 5’-triphosphate trisodium), 0.1; HEPES (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid), 10; pH 7.3; 290 –300 mOsmol. A seal resistance of ⱖ 2 G⍀ and an access resistance of 15–20 M⍀ were considered acceptable. The series resistance was optimally compensated by ⱖ 70% and constantly monitored throughout the experiments. The membrane potential was held at ⫺70 mV, unless otherwise stated, throughout the experiment. Schaffer collateral commissural fibers were stimulated by ultra thin concentric bipolar electrodes (FHC, Inc., Bowdoinham, ME, USA), and the stimulus intensity was adjusted to evoke about 35% maximal stimulation unless specified. Excitatory postsynaptic currents (EPSCs) were recorded in the CA1 area in the presence of GABAA (gamma aminobutyric acid receptor type A) antagonist bicuculline (30 ␮M). The evoked EPSCs were filtered at 2 kHz, digitized at 10 kHz, and acquired and analyzed using Axograph X software

794

J. Wu et al. / Neurobiology of Aging 34 (2013) 791– 804

(Sydney, Australia). The amplitude of EPSCs was monitored for a baseline period of at least 15 minutes. If synaptic transmission was stable (⬍ 15% change in EPSC amplitude over 15 minutes), long-term potentiation (LTP) was induced by a single high-frequency electric stimuli train (100 Hz for 1 second) (Shipton et al., 2011). All electrophysiological experiments were performed at room temperature (23 ⫾ 2 °C). 2.6. Immunostaining Immunostaining of the hippocampal slices was performed as previously described (Bie et al., 2010). A cohort of rats (n ⫽ 10 –12 per group) were deeply anesthetized with 60 mg/kg sodium pentobarbital and perfused transcardially with 0.1 m phosphate-buffered saline (PBS) followed by 4% formalin. The brainstem was collected, postfixed in the same fixative for 4 hours, and then cryoprotected in 30% sucrose in PBS for 3 days. Serial sections (30 ␮m, 15–20 per rat) containing the hippocampal CA1 area were cut from the fixed brain. The sections were incubated for at least 1 hour in 0.01 M PBS with 0.3% Triton X-100 plus 5% normal donkey serum. Primary and secondary antibodies were diluted in 0.01 m PBS with 0.3% Triton X-100 plus 1% bovine serum. Sections were processed overnight at 4 °C for double-labeling immunofluorescence using rabbit antibodies directed against CB2 (1:500, Thermo Scientific, Rockford, IL, USA), mouse monoclonal antibody against the microglial marker CD11b (1:200, Abcam, Cambridge, MA, USA), and GFAP antibody (1:400; Abcam). Other sections were processed overnight at 4 °C for immunofluorescence using A␤1– 40 (11A50-B10) monoclonal antibody (1 ␮g/mL PBS, Covance, San Diego, CA, USA). Then, the sections were incubated with a mixture of fluorescein isothiocyanate (1:500, Jackson ImmunoResearch, West Grove, PA, USA) or Cy3-conjugated secondary antibodies (1:500, Jackson ImmunoResearch) for 1 hour. Omission of primary or secondary antibodies resulted in no immunostaining. Three to 6 sections from each rat were randomly selected and were analyzed using Leica DMLB fluorescence microscope (Leica Microsystems, Wetzlar, GmbH, Germany) by 1 investigator (Dr. Wu) who was blinded to the origin of tissue being examined. Images were quantified using MCID Core software Version 7.0 (InterFocus Imaging, Ltd., Cambridge, UK). 2.7. Protein extraction and immunoblotting The protocol for protein extraction and immunoblotting was generally based on previous reports (Bie et al., 2010, 2011a). The hippocampal CA1 tissues from the rats in all groups (n ⫽ 10 –12 rats per group) were gently homogenized in ice-cold lysis buffer containing 50 mM Tris-Cl, 150 mM NaCl, 0.02 mM NaN2, 100 ␮g/mL phenylmethyl sulfonyl fluoride, 1 ␮g/mL aprotinin, 1% Triton X-100, and proteinase inhibitor cocktail. The lysates were centrifuged at 14,000 rpm for 10 minutes at 4 °C, and the supernatant

was used for SDS-polyacrylamide (sodium dodecyl sulfate polyacrylamide) gel electrophoresis. Protein concentrations were determined by using the Bio-Rad (Hercules, CA, USA) protein assay kit. The samples were treated with SDS sample buffer at 95 °C for 5 minutes, loaded on a 7.5% SDS-polyacrylamide gel, and blotted to a nitrocellulose membrane. The blots were incubated overnight at 4 °C with a rabbit polyclonal anti-CB2 primary antibody (1:100; Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA), monoclonal anti-CD11b antibody (1:1000; Abcam), monoclonal anti-GFAP antibody (1:1000; Abcam), goat polyclonal anti-IL-1␤ antibody (1:300; Abcam) or monoclonal anti-␤-actin antibody (1: 250; Santa Cruz Biotechnology, Inc.). The membranes were washed extensively with Tris-buffered saline and incubated with horseradish peroxidase-conjugated anti-mouse and antirabbit IgG or anti-goat IgG antibody (1:10,000; Jackson ImmunoResearch Laboratories, Inc.). The immunoreactivity was detected using enhanced chemiluminescence (ECL Advance Kit; Amersham Biosciences). The intensity of the bands was captured digitally and analyzed quantitatively with ImageJ software. The immunoreactivity of CB2, CD11b, and GFAP was normalized to that of ␤-actin. 2.8. Compounds A␤ peptide consisting of residues 1– 40 of the human wild type sequence (A␤1– 40) was purchased from Bachem (Torrance, CA, USA). D-2-amino-5-phosphonopentanoate, 6-cyano-2,3-dihydroxy-7-nitroquinoxaline, bicuculline, and other chemicals (NaCl, KCl, NaH2PO4, MgCl2, CaCl2, glucose, NaHCO3, EGTA, Mg–ATP, Na3GTP, HEPES, and Triton-X 100) were purchased from Sigma Aldrich (St. Louis, MO, USA) or Tocris (Ellisville, MO, USA). AM630 was purchased from Tocris. MDA7 (hCB1 Ki value ⬎ 10,1000 nM; hCB2 Ki value ⫽ 422 nM; rCB1 Ki value ⫽ 2565 nM; rCB2 Ki value ⫽ 238 nM; EC50 (half maximal effective concentration) at hCB1 ⫽ not active, EC50 at hCB2 ⫽ 128 nM; EC50 at rCB1 ⫽ not active; EC50 at rCB2 ⫽ 67 nM) was synthesized as previously described (Diaz et al., 2009; Naguib et al., 2008). 2.9. Data analysis and statistics The behavioral data were analyzed using repeated measures analysis of variance (ANOVA) followed by post hoc Student– Newman–Keuls multiple range test. Immunostaining and immunoblotting data were analyzed using 1-way ANOVA followed by post hoc Student–Newman–Keuls multiple range test for group means. The evoked EPSC and LTP data were analyzed using repeated measures ANOVA followed by post hoc Student–Newman–Keuls multiple range test. All analyses were performed using the BMDP statistical package (Statistical Solutions, Saugus, MA, USA). Results were expressed as mean and standard error of the mean, and they were considered significant when p ⬍ 0.05.

J. Wu et al. / Neurobiology of Aging 34 (2013) 791– 804

3. Results 3.1. Effect of activation of CB2 on the behavioral test First, we determined the effect of i.p. daily administration (for 14 days) of 3 different doses of MDA7 (1.5, 5, or 15 mg/kg) on cognitive impairment induced by A␤1– 40. Repeated measures ANOVA identified a significant dose-

795

dependent memory enhancing effect of MDA7 in the Morris water maze test (overall F(4,45) ⫽ 6.86, n ⫽ 50, p ⫽ 0.0002). Animals injected with A␤1– 40 fibrils and treated with 15 mg/kg MDA7 i.p. for 14 days had a significantly shorter escape latency in the Morris water maze test than the animals injected with A␤1– 40 fibrils and treated with i.p. saline for 14 days (Fig. 2A) at day 3 (p ⬍ 0.05), day 4 (p ⬍

Fig. 2. Administration of 1-[(3-benzyl-3-methyl-2,3-dihydro-1-benzofuran-6-yl) carbonyl] piperidine (MDA7) attenuated amyloid fibril-impaired performance in the Morris water maze test. (A) MDA7 treatment decreases escape latency scores in the Morris water maze test in a dose-dependent manner. Animals injected with amyloid-␤ (A␤)1– 40 fibrils and treated with MDA7 15 mg/kg intraperitoneally (i.p.) for 14 days had a significantly (p ⬍ 0.05) shorter escape latency in the Morris water maze test than the animals injected with A␤1– 40 fibrils and treated with saline i.p. for 14 days at days 3 and 5. In contrast, MDA7 1.5 mg/kg i.p. for 14 days did not result in any significant memory enhancement following A␤1– 40 fibrils administration, however, the effect of the 5 mg/kg MDA7 dose was limited only to days 3 and 5. (B) The probe trial was performed on day 6 to determine the time spent in the target quadrant (TQ or platform quadrant) compared with right quadrant (RQ), opposite quadrant (OQ), and left quadrant (LQ). Rats injected with A␤1– 40 fibrils and treated with MDA7 i.p. 15 mg/kg for 14 days spent the longest time in the TQ than animals injected with A␤1– 40 and treated with saline i.p. for 14 days (p ⬍ 0.05). (C) Rats injected with A␤1– 40 fibrils and treated with saline i.p. for 14 days had a significantly extended escape latency in the Morris water maze test compared with that of the rats that received bilateral intracerebral microinjection of artificial cerebrospinal fluid and treated with saline i.p. (controls) or animals injected with A␤1– 40 and treated with 15 mg/kg MDA7 i.p. for 14 days at days 3 to 5. Administration of AM630 prior to MDA7 treatment abrogated the observed effects of MDA7 indicating that the effects of MDA7 are mediated by stimulating cannabinoid (CB)2 receptors. Furthermore, rats injected with A␤1– 40 and received 5 mg/kg of AM630 i.p. for 14 days alone had cognitive impairment similar to those animals injected with A␤1– 40 and treated with either saline i.p. or 5 mg/kg AM630 plus 15 mg/kg MDA7 i.p. for 14 days. The rats that received bilateral intracerebral microinjection of artificial cerebrospinal fluid and treated with saline i.p. or 15 mg/kg MDA7 for 14 days showed no significant difference in their escape latency values. (D) During the probe trial, the rats injected with A␤1– 40 fibrils and treated with MDA7 15 mg/kg i.p. for 14 days spent the longest time in the TQ than animals injected with A␤1– 40 and treated for 14 days with either saline i.p., AM630 alone, or AM630 plus MDA7 (p ⬍ 0.01). Statistical significance was determined by repeated measures analysis of variance followed by Student–Newman–Keuls multiple range test. Each point represents the mean ⫾ standard error of the mean of each group (n ⫽ 10 per group). * p ⬍ 0.05, ** p ⬍ 0.01.

796

J. Wu et al. / Neurobiology of Aging 34 (2013) 791– 804

0.05), and day 5 (p ⬍ 0.05). Similar observation was noted during the probe trial, in that the rats injected with A␤1– 40 fibrils and treated with i.p. MDA7 15 mg/kg for 14 days spent the longest time in the target quadrant (TQ or platform quadrant) than animals injected with A␤1– 40 and treated with i.p. saline for 14 days (p ⬍ 0.05) (Fig. 2B). We noted that the effect of 1.5 mg/kg MDA7 i.p. for 14 days did not result in any significant memory enhancement following A␤1– 40 fibrils administration, however, the effect of the 5 mg/kg MDA7 dose was limited only to days 3 and 5 (Fig. 2A). Based on these data, the dose of 15 mg/kg of MDA7 was selected for further studies. Repeated measures ANOVA for the data reported in Fig. 2C identified significant differences among the 6 groups (overall F(5,54) ⫽ 8.10, n ⫽ 60, p ⬍ 0.0001). The rats in the A␤1– 40 group had a significantly extended escape latency in the Morris water maze test than those in A␤1– 40 ⫹ MDA7 group or the control group (Fig. 2C) at day 3 (p ⬍ 0.01), day 4 (p ⬍ 0.01), and day 5 (p ⬍ 0.01). During the probe trial (Fig. 2D), rats in the A␤1– 40 ⫹ MDA7 group spent significantly longer time in the TQ than those in the A␤1– 40 group (p ⬍ 0.01). This indicated that MDA7 treatment was able to prevent the cognitive impairment on spatial memory performance induced by A␤1– 40. To confirm the CB2-specific effect of MDA7, we administered 5 mg/kg of AM630, a CB2 antagonist, i.p. daily for 14 days either alone or 15 minutes prior to MDA7 (15

mg/kg i.p.) administration to groups of rats injected with A␤1– 40 fibrils. As shown in Fig. 2C, administration of AM630 prior to MDA7 treatment reversed the effect of MDA7 on the escape latency. Similar effect of AM630 was also noted in the probe test (Fig. 2D). Meanwhile, rats that received AM630 alone i.p. (in the A␤1– 40 ⫹ AM630 group) did not show any improvement in spatial memory or probe test compared with that in the A␤1– 40 group (Fig. 2C and D). The different behavioral performance noted in this study were not attributable to the presence of motor deficits because all groups of rats exhibited similar swimming speeds (data not shown). Our results indicated the effect of MDA7 on restoring cognition and memory was specifically mediated through the CB2 receptor system. 3.2. MDA7 effectively ameliorates A␤1– 40-induced glia activation in the hippocampal CA1 area The accuracy of the microinjection was histologically verified afterward (Bie et al., 2009a). Fig. 3A shows an India ink-marked microinjection site in the hippocampal CA1 area demonstrating the preciseness of the injection site. Because prior studies found an association between AD and glial cell activation (Akiyama et al., 2000; Masliah, 2008; McGeer et al., 1988; Perry et al., 2010; Wyss-Coray, 2006), we considered the possibility that MDA7’s preventive effect on A␤1– 40-induced microglial activation was

Fig. 3. Administration of 1-[(3-benzyl-3-methyl-2,3-dihydro-1-benzofuran-6-yl) carbonyl] piperidine (MDA7) attenuated amyloid fibrils-induced CD11b immunoreactivity in the hippocampal CA1 area. (A) Coronal rat brain section showing hippocampal ink injection: India ink was injected to confirm that the injection procedure was accurate and that the material was injected specifically into the hippocampal CA1 area of the rat. (B) Immunofluorescence images of CD11b immunoreactivity in various groups of rats in the hippocampal CA1 areas. (C) Analysis of CD11b immunofluorescence intensity (20 sections from 5 animals per group) showed that the amyloid-␤ (A␤)1– 40-injected rats receiving MDA7 treatment intraperitoneally (i.p.) for 14 days had significantly less (p ⬍ 0.01) immunofluorescence intensity than did the animals injected with A␤1– 40 and treated with saline i.p. for 14 days. (D) Immunoblots of CD11b reactivity in hippocampus CA1 areas. (E) Analysis of CD11b immunoreactivity revealed significant increases in CD11b (p ⬍ 0.01) intensities in animals injected with A␤1– 40 and treated with saline i.p. for 14 days compared with that in the control group (n ⫽ 5–7 per group). These changes were significantly (p ⬍ 0.01) attenuated in the A␤1– 40-injected rats receiving MDA7 treatment i.p. for 14 days. Note that treatment with MDA7 alone had no significant effect on the CD11B expression compared with the control rats. Statistical significance was determined by 1-way analysis of variane followed by Student– Newman–Keuls multiple range test. Data are shown as mean ⫾ standard error of the mean. Scale bar ⫽ 40 ␮m.

J. Wu et al. / Neurobiology of Aging 34 (2013) 791– 804

mediated through modulation of glial activation. Increased expression of specific markers such as CD11b and GFAP have been associated with activation of microglia (Eng 1985) and astrocytes (Garrison et al., 1991), respectively. We therefore examined the impact of 15 mg/kg of MDA7 i.p. daily for 14 days on A␤1– 40-induced microglial and astrocyte activation by comparing the levels of immunofluorescence staining and immunoblotting intensity for CD11b (Fig. 3) and GFAP (Fig. 4) in the hippocampal CA1 area of rats in the various treatment groups. Glial activation is characterized by an increase in the number and complexity of these cells (rounded cell bodies and thicker processes), resulting in an increase in both the number of labeled cells and the total integrated intensity of the labeled cells (Figs. 3B and 4A). When the hippocampal CA1 areas were examined in the rats after microinjection of A␤1– 40 and treatment with i.p. saline for 14 days, increased CD11b and GFAP immunoreactivity, enlarged cell mass, and increased cell complexity were noted in the microglia (Fig. 3B and C; F(3,76) ⫽ 243.2; n ⫽ 80 sections; 3–5 sections per animal from 5 animals per group; 1-way ANOVA followed by Student–Newman–Keuls multiple range test; p ⬍ 0.01) and astrocytes (Fig. 4A and B; F(3,76) ⫽ 225.3; n ⫽ 80 sections; 3–5 sections per animal from 5 animals per group; p ⬍ 0.01), respectively, compared with the control group. The protein expression of CD11b (Fig. 3D and E; n ⫽ 6 rats per group; F(1,10) ⫽ 19.9; 1-way ANOVA followed by

797

Student–Newman–Keuls multiple range test; p ⬍ 0.01) and GFAP (Fig. 4C and D; n ⫽ 5 rats per group; F(1,8) ⫽ 32.7; p ⬍ 0.01) was significantly higher in the rats injected with A␤1– 40 and treated with i.p. saline for 14 days than in the control group rats. These results confirmed that A␤1– 40induced brain inflammation is characterized by the activation of microglia and astrocytes in the hippocampal CA1 area. The changes in CD11b and GFAP expression were significantly attenuated in the A␤-injected rats that received 14 days of MDA7 i.p. treatment compared with the animals injected with A␤1– 40 and treated with i.p. saline for 14 days (Fig. 3D and E; n ⫽ 6 –7 rats per group; F(1,11) ⫽ 16.1; 1-way ANOVA followed by Student–Newman–Keuls multiple range test; p ⬍ 0.01) (Fig. 4C and D; n ⫽ 5 rats per group; F(1,8) ⫽ 46.2; p ⬍ 0.01). Treatment with MDA7 alone had no significant effect on the CD11B or GFAP expression compared with the control rats. 3.3. MDA7 modulated A␤1– 40-induced CB2 receptor upregulation in the hippocampal CA1 area Our previous studies established MDA7 as a potent and selective CB2 agonist (Naguib et al., 2008). Here we examined the effect of A␤1– 40 fibrils on the expression of CB2 receptor in the hippocampal CA1 area. We found that expression of CB2 receptors was enhanced with A␤1– 40 administration (Fig. 5A and B; F(3,76) ⫽ 48.45; n ⫽ 80

Fig. 4. Administration of 1-[(3-benzyl-3-methyl-2,3-dihydro-1-benzofuran-6-yl) carbonyl] piperidine (MDA7) attenuated amyloid fibrils-induced glial fibrillary acidic protein (GFAP) immunoreactivity in the hippocampal CA1 area. (A) Immunofluorescence images of GFAP immunoreactivity in various groups of rats in the hippocampal CA1 areas. (B) Analysis of GFAP immunofluorescence intensity (20 sections from 5 animals per group) showed that the amyloid-␤ (A␤)1– 40-injected rats receiving 15 mg/kg MDA7 treatment intraperitoneally (i.p.) for 14 days had significantly less (p ⬍ 0.01) immunofluorescence intensity than did the animals injected with A␤1– 40 and treated with saline i.p. for 14 days. (C) Immunoblots of GFAP reactivity in the hippocampus CA1 areas. (D) Analysis of GFAP immunoreactivity revealed significant increases in GFAP (p ⬍ 0.01) intensities in the rats injected with A␤1– 40 and treated with saline i.p. for 14 days compared with those in the control group (n ⫽ 5–7 per group). These changes were significantly (p ⬍ 0.01) attenuated in the A␤1– 40-injected rats receiving MDA7 treatment i.p. for 14 days. Note that treatment with 15 mg/kg MDA7 i.p. alone had no significant effect on the GFAP expression compared with control rats. Statistical significance was determined by 1-way analysis of variane followed by Student–Newman–Keuls multiple range test. Data are shown as mean ⫾ standard error of the mean. Scale bar ⫽ 40 ␮m.

798

J. Wu et al. / Neurobiology of Aging 34 (2013) 791– 804

Fig. 5. Administration of 1-[(3-benzyl-3-methyl-2,3-dihydro-1-benzofuran-6-yl) carbonyl] piperidine (MDA7) attenuated amyloid-␤ (A␤)1– 40 fibrilsupregulated cannabinoid (CB)2 receptor expression in the hippocampal CA1 area. (A) CB2 staining and localization in the in the hippocampal CA1 area. Immunofluorescence images of CB2 (red) and CD11b in microglia (green). The coregionalization of CB2 and microglia is shown in yellow. (B) No significant CB2 expression was observed in the control group, but marked CB2 expression was observed in the rats injected with A␤1– 40 and treated with saline intraperitoneally (i.p.) for 14 days (p ⬍ 0.01 vs. control) (n ⫽ 20 sections from 5 animals per group). CB2 expression was significantly decreased in the A␤1– 40-injected rats receiving 15 mg/kg MDA7 treatment i.p. for 14 days. (C) Representative immunoblotting bands to show the expression of CB2 receptor in hippocampal CA1 area in all groups; (D) plotted analysis of CB2 immunoreactivity in hippocampus CA1 tissues in all groups. Note that MDA7 treatment significantly reversed the upsurge of CB2 expression induced by A␤1– 40 (p ⬍ 0.01); n ⫽ 6 per group. Statistical significance was determined by 1-way analysis of variance followed by Student–Newman–Keuls multiple range test. Data are shown as mean ⫾ standard error of the mean. Scale bar ⫽ 40 ␮m.

sections; 3–5 sections per animal from 5 animals per group; 1-way ANOVA followed by Student–Newman–Keuls multiple range test; p ⬍ 0.01 vs. control group) and this expression is colocalized with the reactive microglia. Administration of 15 mg/kg MDA7 i.p. for 14 days to rats injected with A␤1– 40 resulted in reduced CB2 expression (p ⬍ 0.01 vs. rats injected with A␤1– 40 and treated with i.p. saline for 14 days). The immunoblotting studies revealed that CB2 receptor expression in the rats that received microinjection of A␤1– 40 and treatment with i.p. saline for 14 days (Fig. 5C and D; n ⫽ 5; F(1,8) ⫽ 10.3; 1-way ANOVA followed by Student–Newman–Keuls multiple range test; p ⬍ 0.05) was greater than that in the control rats. Furthermore, treatment with 15 mg/kg of MDA7 i.p. daily for 14 days significantly reversed this upsurge of CB2 receptor expression induced by A␤1– 40 fibrils (Fig. 5C and D; n ⫽ 5; F(1,8) ⫽ 19.1; 1-way ANOVA followed by Student–Newman–Keuls multiple range test; p ⬍ 0.01). MDA7 did have any effect on the CB2 receptor expression in the control rats. These results represent an adaptation of brain CB2 receptor induced by A␤1– 40 fibrils and its modulation by the CB2 agonist MDA7.

3.4. MDA7 attenuates IL-1␤ protein expression induced by A␤1– 40 in the hippocampal CA1 area To better understand the basis for MDA7’s interference with A␤-induced neuroinflammation, we examined the impact of MDA7 treatment (15 mg/kg i.p. daily for 14 days) on proinflammatory cytokine IL-1␤ production in the hippocampal CA1 area. In the brain, IL-1␤ is mostly synthesized and secreted from the activated microglia and astrocytes (Kettenmann et al., 2011), and it is implicated in the development of brain inflammation (Gabay et al., 2010) and amyloid-induced memory deficiency (Schmid et al., 2009). In the present study, animals injected with A␤1– 40 and treated with i.p. saline for 14 days showed significantly higher expression of the IL-1␤ protein in the hippocampal CA1 tissue (Fig. 6; n ⫽ 5; F(1,8) ⫽ 63.3; 1-way ANOVA followed by Student–Newman–Keuls multiple range test; p ⬍ 0.01) than the control rats (n ⫽ 5). Meanwhile, MDA7 treatment for 14 days in the rats injected with A␤1– 40 significantly attenuated the upsurge of IL-1␤ induced by microinjection of amyloid fibrils (Fig. 6; n ⫽ 5; F(1,8) ⫽ 34.6; p ⬍ 0.01). Our results indicate that the levels of IL-1␤ from activated glial cells were significantly increased in the hippocampal CA1 area after microinjection of A␤1– 40, and

J. Wu et al. / Neurobiology of Aging 34 (2013) 791– 804

799

Fig. 6. Administration of 1-[(3-benzyl-3-methyl-2,3-dihydro-1-benzofuran-6-yl) carbonyl] piperidine (MDA7) attenuated amyloid-␤ (A␤)1– 40 fibrilsupregulated interleukin (IL)-1␤ level in the hippocampal CA1 area. (A) Immunoblots of IL-1␤ reactivity in hippocampus CA1 tissues in all groups; (B) plotted analysis of IL-1␤ immunoreactivity in hippocampus CA1 tissues in all groups. Note that 15 mg/kg MDA7 treatment intraperitoneally (i.p.) for 14 days significantly reversed the increase of IL-1␤ protein expression induced by A␤1– 40 (p ⬍ 0.01; n ⫽ 5 per group; 1-way analysis of variance followed by Student–Newman–Keuls multiple range test). Data are shown as mean ⫾ standard error of the mean.

these increases were blunted by MDA7 treatment for 14 days. 3.5. MDA7 effectively promotes the clearance of A␤1– 40 injected in the hippocampal CA1 area In vitro studies (Tolón et al., 2009) have shown that the CB2 activation induces removal of A␤ and this effect was blocked by CB2 antagonists. Therefore, we reasoned that MDA7 would enhance clearance mechanisms of the injected A␤1– 40 in rats. MDA7 treatment for 14 days (15 mg/kg per day) in the rats injected with A␤1– 40 induced removal of A␤1– 40 (Fig. 7; F(3,76) ⫽ 707.07; n ⫽ 80 sections; 3–5 sections per animal from 5 animals per group; 1-way ANOVA followed by Student–Newman–Keuls multiple range test; p ⬍ 0.01) and this effect was abrogated by prior administration of 5 mg/kg AM630 i.p. for 14 days. We also observed that administration of 5 mg/kg AM630 i.p.

alone has no effect on the clearance of the injected A␤1– 40. Our results indicate that MDA7-mediated cognitive improvements are correlated with the enhanced clearance of A␤ peptide. 3.6. MDA7 effectively ameliorates A␤1– 40-impaired glutamatergic transmission in the hippocampal CA1 area Then, we compared the input (stimulation intensity)output (current amplitude) responses of the evoked EPSCs in the CA1 neurons among all groups. As shown in Fig. 8A and B, the amplitudes of evoked EPSCs were significantly less in the A␤1– 40 group than that in the control group at all 3 stimulus intensities, indicating attenuated basal glutamatergic strength in the CA1 neurons. However, the amplitude of evoked EPSCs in the CA1 neurons from the rats injected with A␤1– 40 in rats was restored by 15 mg/kg MDA7 i.p. treatment for 14 days in the A␤1– 40 ⫹ MDA7 group at all

Fig. 7. The cannabinoid (CB)2 agonist 1-[(3-benzyl-3-methyl-2,3-dihydro-1-benzofuran-6-yl) carbonyl] piperidine (MDA7) induces amyloid-␤ (A␤)1– 40 clearance. (A) Immunofluorescence images of A␤1– 40 deposits in the hippocampal CA1 areas in different groups. (B) Rats injected with A␤1– 40 and treated with 15 mg/kg MDA7 intraperitoneally (i.p.) for 14 days exhibited a reduction of the deposited A␤ peptide (p ⬍ 0.01 vs. the rats injected with A␤1– 40 and treated with saline i.p. for 14 days; n ⫽ 20 sections from 5 animals per group) and this effect was abrogated by prior administration of 5 mg/kg AM630 i.p. for 14 days. Administration of 5 mg/kg AM630 i.p. alone has no effect on the clearance of the injected A␤1– 40. Statistical significance was determined by 1-way analysis of variance followed by Student–Newman–Keuls multiple range test. Data are shown as mean ⫾ standard error of the mean. Scale bar ⫽ 80 ␮m.

800

J. Wu et al. / Neurobiology of Aging 34 (2013) 791– 804

Fig. 8. Systemic administration of 15 mg/kg 1-[(3-benzyl-3-methyl-2,3-dihydro-1-benzofuran-6-yl) carbonyl] piperidine (MDA7) intraperitoneally (i.p.) for 14 days ameliorated amyloid-␤ (A␤)1– 40-impaired basal glutamatergic strength and long-term potentiation (LTP) in the hippocampal CA1 slices. LTP was induced by the electric stimuli on the Schaffer collateral– commissural fibers at 100 Hz for 1 second. (A) Representative traces of evoked excitatory postsynaptic currents (EPSCs) at 3 graded stimulus intensities in all groups. (B) Input (stimulus intensity)-output (EPSC current) curve of evoked EPSCs in the CA1 neurons in all 4 groups (n ⫽ 9 –10 neurons per group). (C) Representative traces to show the evoked EPSCs at baseline, 30, and 60 minutes after electric induction in all 4 groups. (D) Time course of the LTP induction in the hippocampal CA1 neurons in all 4 groups (n ⫽ 9 –12 neurons from 4 to 5 rats per group). Statistical significance was determined by repeated measures analysis of variance followed by Student–Newman–Keuls multiple range test. Data are shown as mean ⫾ standard error of the mean. * p ⬍ 0.05; ** p ⬍ 0.01 versus A␤1– 40 and A␤1– 40 ⫹ MDA7 groups.

3 stimulus intensities (Fig. 8A and B). The basal glutamatergic strength in the hippocampal slices was not different between the control and MDA7 groups. These results indicated that administration of MDA7 significantly ameliorated the impaired basal glutamatergic strength induced by microinjection of A␤1– 40 fibrils into the hippocampus. In the hippocampus slice of the saline-injected rats, high frequency electric stimuli on the Schaffer collateral– commissural fibers induced significant synaptic potentiation in CA1 neurons that lasted more than 1 hour. The average LTPs were 210.5 ⫾ 18.6% and 224.7 ⫾ 2.2% at 30 and 60 minutes after induction, respectively (n ⫽ 12; repeated measures ANOVA followed by Student–Newman–Keuls multiple range test; p ⬍ 0.01 when compared with baseline; Fig. 8C and D). Meanwhile, in the hippocampus slice of the A␤1– 40-injected rats, the intensity of LTP induced by high frequency electric stimuli was significantly attenuated. The average LTPs were 107.1 ⫾ 14.0% and 110.4 ⫾ 15.6% at 30 and 60 minutes after induction, respectively (n ⫽ 10; repeated measures ANOVA followed by Student–Newman–Keuls multiple range test; p ⬍ 0.01 when compared

with that in the saline-injected rats; Fig. 8C and D). However, in the hippocampus slices from the A␤ ⫹ MDA7 group, the average LTPs were 214.5 ⫾ 23.6% and 195.7 ⫾ 21.8% at 30 and 60 minutes after induction, respectively (n ⫽ 9; repeated measures ANOVA followed by Student– Newman–Keuls multiple range test; p ⬍ 0.01 when compared with that in the A␤1– 40-injected rats; Fig. 8C and D). Note that the administration of MDA7 alone failed to significantly affect the induction of LTP in the hippocampal slices of the control rats. These results indicted that administration of MDA7 significantly ameliorated the impaired hippocampal synaptic plasticity induced by microinjection of A␤1– 40 fibrils. 4. Discussion Our results reveal a promising potential role for the CB2 agonist MDA7 in (1) promoting A␤ clearance; (2) ameliorating A␤-induced glial activation and cytokine production; and (3) restoring of synaptic plasticity, cognition, and memory. The effects of MDA7 were abrogated by prior administration of a CB2 antagonist AM630. The administration of

J. Wu et al. / Neurobiology of Aging 34 (2013) 791– 804

AM630 alone did not result in any beneficial effect on A␤-related pathology. The present in vivo study confirmed that microinjection of A␤1– 40 fibrils significantly (p ⬍ 0.01) induced the activation of astrocytes and microglia and upregulation of CB2 receptors in the hippocampal CA1 area and that MDA7 treatment reversed this process. The finding that microinjection of A␤1– 40 in rats treated with i.p. saline was associated with increased CB2 expression in the hippocampal CA1 area (Fig. 5) whereas treatment with the CB2 agonist MDA7 ameliorated the effects of amyloid fibrils raises the possibility that the CB2 receptor acts as a negative feedback regulator and its activation by MDA7 can serve to limit the extent of the neuroinflammatory response and the subsequent development of A␤-induced neurotoxicity. In the present study, microglial (CD11b) and astrocyte (GFAP) activation markers (measured on day 15) were significantly increased (p ⬍ 0.01) in the hippocampal CA1 area after microinjection of A␤1– 40 in rats treated with i.p. saline compared with controls (Figs. 3 and 4). A␤ aggregates are potent neurotoxins and microglial activators (Cameron and Landreth, 2010). Our results are consistent with previously published reports in which intracerebral microinjection of A␤1– 42 in the hippocampus (Chacón et al., 2004), frontal cortex (Esposito et al., 2007), or intracerebroventricular administration of A␤25–35 (Ramírez et al., 2005) in that the central administration of A␤ is associated with microglia (Ramírez et al., 2005) and astrocyte activation (Chacón et al., 2004; Esposito et al., 2007; Ramírez et al., 2005). This neuroinflammatory process appears to contribute to and/or reflect neuronal dysfunction in the brain and is found to be inversely correlated with the cognitive function in AD (Edison et al., 2008; Perry et al., 2010). Furthermore, in brain tissues from patients with AD, microglia and astrocyte activation, evidenced by cellular hypertrophy, increases in the expression of GFAP, astroglial S100B, and CD11b proteins, are routinely observed (Griffin et al., 1989; Meda et al., 2001). Although micromolar concentrations of A␤ can induce direct toxicity to neurons, concentrations at nanomolar levels can induce neuronal loss through a microglia-mediated mechanism (Neniskyte et al., 2011). It seems that the deposition of A␤ represents important trigger factors in glial activation (possibly via interaction with microglial surface receptors; El Khoury et al., 1996; Yan et al., 1996) leading to an inflammatory reaction in the brain (Meda et al., 2001). Although it has been reported that the administration of SR144525, a CB2 antagonist, could have a role in decreasing A␤1– 40-induced astrocyte activation (Esposito et al., 2007), our data indicate that administration of 5 mg/kg AM630 (a CB2 antagonist) i.p. for 14 days to rats injected with A␤1– 40 failed to improve cognitive functions or induce A␤ clearance. Our results showed increased levels of IL-1␤ after bilateral microinjection of A␤1– 40 in the hippocampus, and this increase was prevented by MDA7 treatment. CB2 receptor activation has been shown to decrease the production of

801

proinflammatory molecules in vitro in rat microglial cells (Facchinetti et al., 2003; Naguib et al., 2012), human microglial cells (Stella, 2004), and human astrocytes (Sheng et al., 2005), and in animal models of perinatal hypoxia– ischemia (Fernández-López et al., 2006), Huntington’s disease (Fernández-Ruiz et al., 2007), and paclitaxel-induced neuroinflammation (Naguib et al., 2012). IL-1␤ is synthesized and released from the activated microglia and astrocytes in the brain (Gabay et al., 2010; Sims et al., 1988) and is actively involved in the development of amyloid-induced brain inflammation, impaired glutamatergic transmission and memory deficiency (Griffin et al., 1989). Intracerebroventricular administration of IL-1␤ significantly induced memory deficiency evidenced by impaired performance in several memory tasks (Moore et al., 2009; Song et al., 2004). In the primary culture of hippocampal neurons, IL-1␤ (but not IL-10 or tumor necrosis factor-␣) significantly downregulated the surface expression and Ser831 phosphorylation of the AMPA (␣-amino-3-hydroxy-5methyl-4-isoxazolepropionic acid) receptor subunit GluR1 (Hsieh et al., 2006; Lai et al., 2006), which plays an important role in synaptic plasticity. Antagonism of IL-1 receptor in the brain alleviated the impaired synaptic plasticity induced by A␤ (Schmid et al., 2009). These previous reports established IL-1␤, released from the activated microglia and astrocytes, as an important factor for the pathogenesis and development of Alzheimer’s disease. In the present study, we also noted an increased expression of CB2 receptors in the hippocampal CA1 tissue in the rats following microinjection of A␤1– 40 fibrils. MDA7 appears to be effective in suppressing microglial and astrocyte activities and IL-1␤ production through activation of the CB2 receptor system. Previous studies have demonstrated that CB2 receptors expression in glia-associated plaques was increased in the postmortem AD brains (Núñez et al., 2004; Ramírez et al., 2005). Intracerebroventricular administration of a nonselective cannabinoid receptor agonist WIN 55,212-2 was effective in preventing A␤25–35-induced microglial activity (Ramírez et al., 2005). Different cannabinoids (WIN 55,212-2, HU-210, and JWH-113) blocked A␤1– 40-induced activation of cultured microglial cells (Ramírez et al., 2005). In primary microglial culture, A␤1– 42-induced activation of CB2 receptors and administration of JWH-015 (a CB2 agonist) markedly suppressed microglial cytokines and nitric oxide production and attenuated CD40mediated inhibition of microglial phagocytosis of A␤1– 42 (Ehrhart et al., 2005). In vitro activation of CB2 receptor facilitated the removal of native A␤ from human frozen tissue sections as well as removal of synthetic pathogenic peptide by a human macrophage cell line (Tolón et al., 2009). Our data showed that activating CB2 receptor system by MDA7 promoted A␤1– 40 clearance (Fig. 7), possibly by restoring microglial phagocytic function. Microglia play an important role in promoting the clearance and phagocytosis of A␤ (Rogers et al.,

802

J. Wu et al. / Neurobiology of Aging 34 (2013) 791– 804

2002; Streit, 2004; Yan et al., 1996, 2006) and there is an inverse relationship between cytokine production and A␤ clearance (Hickman et al., 2008). However, as the AD progresses, microglia continue to produce proinflammatory cytokines, but lose their A␤-clearing capabilities (Hickman et al., 2008). Hence, the upregulation of microglial CB2 receptor in the Alzheimer brain potentially positioned the CB2 receptor as an endogenous protective mechanism to limit the neuroinflammation and pathologic development in this disease. The beneficial effects of CB2 agonist may not only rely on their ability to block A␤-induced microglial activation, but also to restore microglial abilities to remove A␤ deposits. The behavioral deficits we observed following hippocampal microinjection of A␤1– 40 fibrils (Fig. 2B) correlated well with alteration in the hippocampal glutamatergic transmission (Fig. 8). The present study, for the first time, demonstrated that systemic administration of CB2 agonist significantly ameliorated the impaired basal glutamatergic strength and electric stimuli-induced synaptic plasticity in the hippocampal CA1 area and the memory deficiency induced by the local microinjection of A␤1– 40 fibrils. It was previously reported that microinjection of amyloid fragments significantly attenuated basal glutamatergic strength in the hippocampus of rodents (Stéphan et al., 2001). Intracerebroventricular injection of A␤-containing aqueous extracts of Alzheimer’s disease brain significantly inhibited high-frequency electric stimuli-induced LTP—a form of synaptic plasticity—in the rat hippocampus (Barry et al., 2011; Schmid et al., 2009). Hippocampal LTP is used as a correlate for learning and memory (Bliss and Collingridge, 1993; Lynch, 2004) and has emerged as a valuable model for studying mechanisms involved in cognitive deficits related to AD (Barry et al., 2011; Klyubin et al., 2004; Schmid et al., 2009). Administration of exogenous A␤ into the hippocampus significantly lowered the performance in the memory tasks, such as the water maze, in rats (Chacón et al., 2004; Li et al., 2011; Ramírez et al., 2005; Srivareerat et al., 2011). Consistently, in the present study, an impaired electric stimuli-induced LTP and basal glutamatergic strength and extended escape time in the Morris water maze test were observed in the rats 15 days after the microinjection of A␤1– 40 fibrils in the rats treated with i.p. saline for 14 days. Interestingly, the impaired synaptic plasticity and memory deficiency were normalized by the systemic administration of 15 mg/kg MDA7 for 14 days. In conclusion, activation of central microglial CB2 receptors by MDA7 (1) promoted A␤ clearance, (2) ameliorated A␤-induced glia activation and production of IL-1␤, and (3) restored synaptic plasticity, cognition, and memory. The presence of CB2 receptors in microglia in the human AD brain suggests that MDA7 may represent a novel therapeutic target in the treatment of AD.

Acknowledgements This study was supported by funds from the Institute of Anesthesiology, Cleveland Clinic, Cleveland, OH, USA.

Disclosure statement Mohamed Naguib is named on 2 patent applications on CB2 modulators submitted by the University of Texas MD Anderson Cancer Center, Houston, TX, USA. All other authors have nothing to disclose. All animal procedures were approved by the Animal Care and Use Committee of Cleveland Clinic.

References Ahmed, T., Enam, S.A., Gilani, A.H., 2010. Curcuminoids enhance memory in an amyloid-infused rat model of Alzheimer’s disease. Neuroscience 169, 1296 –1306. Akiyama, H., Barger, S., Barnum, S., Bradt, B., Bauer, J., Cole, G.M., Cooper, N.R., Eikelenboom, P., Emmerling, M., Fiebich, B.L., Finch, C.E., Frautschy, S., Griffin, W.S., Hampel, H., Hull, M., Landreth, G., Lue, L., Mrak, R., Mackenzie, I.R., McGeer, P.L., O’Banion, M.K., Pachter, J., Pasinetti, G., Plata-Salaman, C., Rogers, J., Rydel, R., Shen, Y., Streit, W., Strohmeyer, R., Tooyoma, I., Van Muiswinkel, F.L., Veerhuis, R., Walker, D., Webster, S., Wegrzyniak, B., Wenk, G., Wyss-Coray, T., 2000. Inflammation and Alzheimer’s disease. Neurobiol. Aging 21, 383– 421. Barry, A.E., Klyubin, I., Mc Donald, J.M., Mably, A.J., Farrell, M.A., Scott, M., Walsh, D.M., Rowan, M.J., 2011. Alzheimer’s disease brainderived amyloid-␤-mediated inhibition of LTP in vivo is prevented by immunotargeting cellular prion protein. J. Neurosci. 31, 7259 –7263. Benito, C., Kim, W.K., Chavarría, I., Hillard, C.J., Mackie, K., Tolón, R.M., Williams, K., Romero, J., 2005. A glial endogenous cannabinoid system is upregulated in the brains of macaques with simian immunodeficiency virus-induced encephalitis. J. Neurosci. 25, 2530 –2536. Benito, C., Núñez, E., Tolón, R.M., Carrier, E.J., Rábano, A., Hillard, C.J., Romero, J., 2003. Cannabinoid CB2 receptors and fatty acid amide hydrolase are selectively overexpressed in neuritic plaque-associated glia in Alzheimer’s disease brains. J. Neurosci. 23, 11136 –11141. Benito, C., Tolón, R.M., Pazos, M.R., Núñez, E., Castillo, A.I., Romero, J., 2008. Cannabinoid CB2 receptors in human brain inflammation. Br. J. Pharmacol. 153, 277–285. Bie, B., Brown, D.L., Naguib, M., 2011a. Increased synaptic GluR1 subunits in the anterior cingulate cortex of rats with peripheral inflammation. Eur. J. Pharmacol. 653, 26 –31. Bie, B., Brown, D.L., Naguib, M., 2011b. Synaptic plasticity and pain aversion. Eur. J. Pharmacol. 667, 26 –31. Bie, B., Zhang, Z., Cai, Y.Q., Zhu, W., Zhang, Y., Dai, J., Lowenstein, C.J., Weinman, E.J., Pan, Z.Z., 2010. Nerve growth factor-regulated emergence of functional delta-opioid receptors. J. Neurosci. 30, 5617– 5628. Bie, B., Zhu, W., Pan, Z.Z., 2009a. Ethanol-induced delta-opioid receptor modulation of glutamate synaptic transmission and conditioned place preference in central amygdala. Neuroscience 160, 348 –358. Bie, B., Zhu, W., Pan, Z.Z., 2009b. Rewarding morphine-induced synaptic function of delta-opioid receptors on central glutamate synapses. J. Pharmacol. Exp. Ther. 329, 290 –296. Bliss, T.V., Collingridge, G.L., 1993. A synaptic model of memory: longterm potentiation in the hippocampus. Nature 361, 31–39. Cameron, B., Landreth, G.E., 2010. Inflammation, microglia, and Alzheimer’s disease. Neurobiol. Dis. 37, 503–509.

J. Wu et al. / Neurobiology of Aging 34 (2013) 791– 804 Chacón, M.A., Barría, M.I., Soto, C., Inestrosa, N.C., 2004. Beta-sheet breaker peptide prevents Abeta-induced spatial memory impairments with partial reduction of amyloid deposits. Mol. Psychiatry 9, 953–961. Chen, B., Bromley-Brits, K., He, G., Cai, F., Zhang, X., Song, W., 2010. Effect of synthetic cannabinoid HU210 on memory deficits and neuropathology in Alzheimer’s disease mouse model. Curr. Alzheimer Res. 7, 255–261. Diaz, P., Phatak, S.S., Xu, J., Fronczek, F.R., Astruc-Diaz, F., Thompson, C.M., Cavasotto, C.N., Naguib, M., 2009. 2,3-Dihydro-1-benzofuran derivatives as a series of potent selective cannabinoid receptor 2 agonists: design, synthesis, and binding mode prediction through ligandsteered modeling. Chemmedchem 4, 1615–1629. Edison, P., Archer, H.A., Gerhard, A., Hinz, R., Pavese, N., Turkheimer, F.E., Hammers, A., Tai, Y.F., Fox, N., Kennedy, A., Rossor, M., Brooks, D.J., 2008. Microglia, amyloid, and cognition in Alzheimer’s disease: An [11C](R)PK11195-PET and [11C]PIB-PET study. Neurobiol. Dis. 32, 412– 419. Ehrhart, J., Obregon, D., Mori, T., Hou, H., Sun, N., Bai, Y., Klein, T., Fernandez, F., Tan, J., Shytle, R.D., 2005. Stimulation of cannabinoid receptor 2 (CB2) suppresses microglial activation. J. Neuroinflammation 2, 29. Eikelenboom, P., Veerhuis, R., Scheper, W., Rozemuller, A.J., van Gool, W.A., Hoozemans, J.J., 2006. The significance of neuroinflammation in understanding Alzheimer’s disease. J. Neural Transm. 113, 1685–1695. El Khoury, J., Hickman, S.E., Thomas, C.A., Cao, L., Silverstein, S.C., Loike, J.D., 1996. Scavenger receptor-mediated adhesion of microglia to beta-amyloid fibrils. Nature 382, 716 –719. Eng, L.F., 1985. Glial fibrillary acidic protein (GFAP): the major protein of glial intermediate filaments in differentiated astrocytes. J. Neuroimmunol. 8, 203–214. Esposito, G., Iuvone, T., Savani, C., Scuderi, C., De Filippis, D., Papa, M., Di Marzo, V., Steardo, L., 2007. Opposing control of cannabinoid receptor stimulation on amyloid-beta-induced reactive gliosis: in vitro and in vivo evidence. J. Pharmacol. Exp. Ther. 322, 1144 –1152. Facchinetti, F., Del Giudice, E., Furegato, S., Passarotto, M., Leon, A., 2003. Cannabinoids ablate release of TNFalpha in rat microglial cells stimulated with lypopolysaccharide. Glia 41, 161–168. Fernández-López, D., Martínez-Orgado, J., Nuñez, E., Romero, J., Lorenzo, P., Moro, M.A., Lizasoain, I., 2006. Characterization of the neuroprotective effect of the cannabinoid agonist WIN-55212 in an in vitro model of hypoxic-ischemic brain damage in newborn rats. Pediatr. Res. 60, 169 –173. Fernández-Ruiz, J., Romero, J., Velasco, G., Tolón, R.M., Ramos, J.A., Guzmán, M., 2007. Cannabinoid CB2 receptor: a new target for controlling neural cell survival? Trends Pharmacol. Sci. 28, 39 – 45. Gabay, C., Lamacchia, C., Palmer, G., 2010. IL-1 pathways in inflammation and human diseases. Nat. Rev. Rheumatol. 6, 232–241. Garrison, C.J., Dougherty, P.M., Kajander, K.C., Carlton, S.M., 1991. Staining of glial fibrillary acidic protein (GFAP) in lumbar spinal cord increases following a sciatic nerve constriction injury. Brain Res. 565, 1–7. Griffin, W.S., Stanley, L.C., Ling, C., White, L., MacLeod, V., Perrot, L.J., White, C.L., 3rd, Araoz, C., 1989. Brain interleukin 1 and S-100 immunoreactivity are elevated in Down syndrome and Alzheimer disease. Proc. Natl. Acad. Sci. U. S. A. 86, 7611–7615. Hardy, J., Selkoe, D.J., 2002. The amyloid hypothesis of Alzheimer’s disease: progress and problems on the road to therapeutics. Science 297, 353–356. Hickman, S.E., Allison, E.K., El Khoury, J., 2008. Microglial dysfunction and defective beta-amyloid clearance pathways in aging Alzheimer’s disease mice. J. Neurosci. 28, 8354 – 8360. Hsieh, H., Boehm, J., Sato, C., Iwatsubo, T., Tomita, T., Sisodia, S., Malinow, R., 2006. AMPAR removal underlies Abeta-induced synaptic depression and dendritic spine loss. Neuron 52, 831– 843. Kettenmann, H., Hanisch, U.K., Noda, M., Verkhratsky, A., 2011. Physiology of microglia. Physiol. Rev. 91, 461–553.

803

Klyubin, I., Walsh, D.M., Cullen, W.K., Fadeeva, J.V., Anwyl, R., Selkoe, D.J., Rowan, M.J., 2004. Soluble Arctic amyloid beta protein inhibits hippocampal long-term potentiation in vivo. Eur. J. Neurosci. 19, 2839 –2846. Lai, A.Y., Swayze, R.D., El-Husseini, A., Song, C., 2006. Interleukin-1 beta modulates AMPA receptor expression and phosphorylation in hippocampal neurons. J. Neuroimmunol. 175, 97–106. Li, J., Wang, Y.J., Zhang, M., Fang, C.Q., Zhou, H.D., 2011. Cerebral ischemia aggravates cognitive impairment in a rat model of Alzheimer’s disease. Life Sci. 89, 86 –92. Lynch, M.A., 2004. Long-term potentiation and memory. Physiol. Rev. 84, 87–136. Masliah, E., 2008. Neuropathology: Alzheimer’s in real time. Nature 451, 638 – 639. Matsuda, L.A., Lolait, S.J., Brownstein, M.J., Young, A.C., Bonner, T.I., 1990. Structure of a cannabinoid receptor and functional expression of the cloned cDNA. Nature 346, 561–564. McGeer, P.L., Itagaki, S., Boyes, B.E., McGeer, E.G., 1988. Reactive microglia are positive for HLA-DR in the substantia nigra of Parkinson’s and Alzheimer’s disease brains. Neurology 38, 1285–1291. Meda, L., Baron, P., Scarlato, G., 2001. Glial activation in Alzheimer’s disease: the role of Abeta and its associated proteins. Neurobiol. Aging 22, 885– 893. Milton, N.G., 2002. Anandamide and noladin ether prevent neurotoxicity of the human amyloid-beta peptide. Neurosci. Lett. 332, 127–130. Moore, A.H., Wu, M., Shaftel, S.S., Graham, K.A., O’Banion, M.K., 2009. Sustained expression of interleukin-1beta in mouse hippocampus impairs spatial memory. Neuroscience 164, 1484 –1495. Munro, S., Thomas, K.L., Abu-Shaar, M., 1993. Molecular characterization of a peripheral receptor for cannabinoids. Nature 365, 61– 65. Naguib, M., Diaz, P., Xu, J.J., Astruc-Diaz, F., Craig, S., Vivas-Mejia, P., Brown, D.L., 2008. MDA7: A novel selective agonist for CB2 receptors that prevents allodynia in rat neuropathic pain models. Br. J. Pharmacol. 155, 1104 –1116. Naguib, M., Xu, J.J., Diaz, P., Brown, D.L., Cogdell, D., Bie, B., Hu, J., Craig, S., Hittelman, W.N., 2012. Prevention of paclitaxel-induced neuropathy through activation of the central cannabinoid type 2 receptor system. Anesth. Analg. 114, 1104 –1120. Neniskyte, U., Neher, J.J., Brown, G.C., 2011. Neuronal death induced by nanomolar amyloid ␤ is mediated by primary phagocytosis of neurons by microglia. J. Biol. Chem. 286, 39904 –39913. Núñez, E., Benito, C., Pazos, M.R., Barbachano, A., Fajardo, O., González, S., Tolón, R.M., Romero, J., 2004. Cannabinoid CB2 receptors are expressed by perivascular microglial cells in the human brain: an immunohistochemical study. Synapse 53, 208 –213. Palop, J.J., Mucke, L., 2010. Amyloid-beta-induced neuronal dysfunction in Alzheimer’s disease: from synapses toward neural networks. Nat. Neurosci. 13, 812– 818. Paxinos, G., Watson, C., 1998. The Rat Brain in Stereotaxic Coordinates, fourth ed. Academic Press, New York. Perry, V.H., Nicoll, J.A.R., Holmes, C., 2010. Microglia in neurodegenerative disease. Nat. Rev. Neurol. 6, 193–201. Ramírez, B.G., Blázquez, C., Gómez del Pulgar, T., Guzmán, M., de Ceballos, M.L., 2005. Prevention of Alzheimer’s disease pathology by cannabinoids: neuroprotection mediated by blockade of microglial activation. J. Neurosci. 25, 1904 –1913. Rogers, J., Strohmeyer, R., Kovelowski, C.J., Li, R., 2002. Microglia and inflammatory mechanisms in the clearance of amyloid beta peptide. Glia 40, 260 –269. Schmid, A.W., Lynch, M.A., Herron, C.E., 2009. The effects of IL-1 receptor antagonist on beta amyloid mediated depression of LTP in the rat CA1 in vivo. Hippocampus 19, 670 – 676. Sheng, W.S., Hu, S., Min, X., Cabral, G.A., Lokensgard, J.R., Peterson, P.K., 2005. Synthetic cannabinoid WIN55,212–2 inhibits generation of inflammatory mediators by IL-1beta-stimulated human astrocytes. Glia 49, 211–219.

804

J. Wu et al. / Neurobiology of Aging 34 (2013) 791– 804

Shin, R.W., Ogino, K., Kondo, A., Saido, T.C., Trojanowski, J.Q., Kitamoto, T., Tateishi, J., 1997. Amyloid beta-protein (Abeta) 1– 40 but not Abeta1-42 contributes to the experimental formation of Alzheimer disease amyloid fibrils in rat brain. J. Neurosci. 17, 8187– 8193. Shipton, O.A., Leitz, J.R., Dworzak, J., Acton, C.E., Tunbridge, E.M., Denk, F., Dawson, H.N., Vitek, M.P., Wade-Martins, R., Paulsen, O., Vargas-Caballero, M., 2011. Tau protein is required for amyloid {beta}-induced impairment of hippocampal long-term potentiation. J. Neurosci. 31, 1688 –1692. Sims, J.E., March, C.J., Cosman, D., Widmer, M.B., MacDonald, H.R., McMahan, C.J., Grubin, C.E., Wignall, J.M., Jackson, J.L., Call, S.M., et al., 1988. cDNA expression cloning of the IL-1 receptor, a member of the immunoglobulin superfamily. Science 241, 585–589. Song, C., Phillips, A.G., Leonard, B.E., Horrobin, D.F., 2004. Ethyleicosapentaenoic acid ingestion prevents corticosterone-mediated memory impairment induced by central administration of interleukin1beta in rats. Mol. Psychiatry 9, 630 – 638. Srivareerat, M., Tran, T.T., Salim, S., Aleisa, A.M., Alkadhi, K.A., 2011. Chronic nicotine restores normal A␤ levels and prevents short-term memory and E-LTP impairment in A␤ rat model of Alzheimer’s disease. Neurobiol. Aging 32, 834 – 844. Stella, N., 2004. Cannabinoid signaling in glial cells. Glia 48, 267–277. Stéphan, A., Laroche, S., Davis, S., 2001. Generation of aggregated betaamyloid in the rat hippocampus impairs synaptic transmission and plasticity and causes memory deficits. J. Neurosci. 21, 5703–5714.

Streit, W.J., 2004. Microglia and Alzheimer’s disease pathogenesis. J. Neurosci. Res. 77, 1– 8. Tolón, R.M., Núñez, E., Pazos, M.R., Benito, C., Castillo, A.I., MartínezOrgado, J.A., Romero, J., 2009. The activation of cannabinoid CB2 receptors stimulates in situ and in vitro beta-amyloid removal by human macrophages. Brain Res. 1283, 148 –154. van der Stelt, M., Mazzola, C., Esposito, G., Matias, I., Petrosino, S., De Filippis, D., Micale, V., Steardo, L., Drago, F., Iuvone, T., Di Marzo, V., 2006. Endocannabinoids and beta-amyloid-induced neurotoxicity in vivo: effect of pharmacological elevation of endocannabinoid levels. Cell. Mol. Life Sci. 63, 1410 –1424. Van Sickle, M.D., Duncan, M., Kingsley, P.J., Mouihate, A., Urbani, P., Mackie, K., Stella, N., Makriyannis, A., Piomelli, D., Davison, J.S., Marnett, L.J., Di Marzo, V., Pittman, Q.J., Patel, K.D., Sharkey, K.A., 2005. Identification and functional characterization of brainstem cannabinoid CB2 receptors. Science 310, 329 –332. Wyss-Coray, T., 2006. Inflammation in Alzheimer disease: driving force, bystander or beneficial response? Nat. Med. 12, 1005–1015. Yan, P., Hu, X., Song, H., Yin, K., Bateman, R.J., Cirrito, J.R., Xiao, Q., Hsu, F.F., Turk, J.W., Xu, J., Hsu, C.Y., Holtzman, D.M., Lee, J.M., 2006. Matrix metalloproteinase-9 degrades amyloid-beta fibrils in vitro and compact plaques in situ. J. Biol. Chem. 281, 24566 –24574. Yan, S.D., Chen, X., Fu, J., Chen, M., Zhu, H., Roher, A., Slattery, T., Zhao, L., Nagashima, M., Morser, J., Migheli, A., Nawroth, P., Stern, D., Schmidt, A.M., 1996. RAGE and amyloid-beta peptide neurotoxicity in Alzheimer’s disease. Nature 382, 685– 691.