Alzheimer’s disease (AD) is a progressive neurodegenerative disease and the leading cause of senile dementia in the United States. Accumulation of amyloid-β (Aβ) and the effects of this peptide on microglial cells contribute greatly to the etiology of AD. Experiments were carried out to determine whether the pan-selective σ-receptor agonist afobazole can modulate microglial response to the cytotoxic Aβ fragment, Aβ25–35. Treatment with afobazole decreased microglial activation in response to Aβ, as indicated by reduced membrane ruffling and cell migration. The effects of afobazole on Aβ25–35-evoked migration were concentration dependent and consistent with σ-receptor activation. When afobazole was coapplied with either BD-1047 [N-[2-(3,4-dichlorophenyl)ethyl]-N-methyl-2-(dimethylamino)ethylamine dihydrobromide] or rimcazole, which are σ-1- and σ-2-selective antagonists, respectively, the inhibition of Aβ25–35-induced migration by afobazole was reduced. Prolonged exposure of microglia to Aβ25–35 resulted in glial cell death that was associated with increased expression of the proapoptotic protein Bax and the death protease caspase-3. Coapplication of afobazole with Aβ25–35 decreased the number of cells expressing both Bax and caspase-3 and resulted in a concomitant enhancement in cell survival. Although afobazole inhibited activation of microglia cells by Aβ25–35, it preserved normal functional responses in these cells after exposure to the amyloid peptide. Intracellular calcium increases induced by ATP were depressed in microglia after 24-hour exposure to Aβ25–35. However, coincubation in afobazole returned these responses to near control levels. Therefore, stimulation of σ-1 and σ-2 receptors by afobazole prevents Aβ25–35 activation of microglia and inhibits Aβ25–35-associated cytotoxicity, suggesting that afobazole may be useful for AD therapeutics.
One of the pathologic hallmarks of Alzheimer’s disease (AD) is that the accumulation of amyloid-β (Aβ) comprised neuritic plaques (Glenner and Wong, 1984). The original observation of these plaques in AD brains and experiments in vitro with various forms of the peptide led to the amyloid cascade hypothesis of AD. This hypothesis suggests that Aβ directly injures brain neurons, resulting in AD pathology (Yankner, 1989, Hardy and Allsop, 1991, Hardy and Selkoe, 2002). However, in addition to these effects on neurons, the actions of Aβ on glial cells are a key component of the amyloid cascade hypothesis (Hardy and Selkoe, 2002). Aβ has been shown to initiate a sequence of events in microglia, including a proinflammatory cascade and enhanced hemichannel formation, which ultimately produces neuronal death (Orellana et al., 2011). In contrast, data also suggest that microglia may be beneficial in AD by removing Aβ plaques (Wilcock et al., 2004, Herber et al., 2007). It has been proposed that microglial senescence, not activation, is a contributing factor to late-onset Alzheimer’s disease (Streit et al., 2009). Given the importance of microglial cells in AD pathology, there is great interest in identifying molecular targets for modulating the responses of these cells to Aβ and for regulating their contribution to AD pathology in general.
One possible target that has been explored for therapeutics in other pathologic states of the central nervous system but is not well characterized in AD is σ receptors. σ Receptors are found in numerous cell types distributed throughout the mammalian body and modulate multiple signaling and regulatory pathways. Two subtypes of σ receptors have been pharmacologically identified, σ-1 and σ-2. To date, only the σ-1 receptor has been cloned (Walker et al., 1990, Quirion et al., 1992, Seth et al., 1998). σ-1 Receptors are intracellular chaperones that form a complex within the endoplasmic reticulum (ER) and the mitochondrion-associated ER membrane (Kourrich et al., 2012). Activation of σ-1 receptors promotes migration of these receptors away from the mitochondrion-associated ER membrane to the subplasmalemmal ER where the receptors interact with and regulate numerous membrane channels (Zhang and Cuevas, 2002, 2005, Herrera et al., 2008, Zhang et al., 2009, Su et al., 2010, Cuevas et al., 2011b). Less is known about σ-2 receptors, but these proteins seem to be involved in regulation of inflammation, cell survival, and calcium homeostasis (Vilner and Bowen, 2000, Crawford and Bowen, 2002, Zhang and Cuevas, 2002, Iniguez et al., 2013). Our laboratory has shown that activation of σ receptors in microglia decreases ATP-induced membrane ruffling and cell migration and lipopolysaccharide-induced cytokine production (Hall et al., 2009). Activation of σ receptors also protects microglia from ischemia-induced cell death (Cuevas et al., 2011b). These observations led our laboratory to propose that the neuroprotection provided by σ-receptor activation after stroke involves both a decrease in the inflammatory response mediated by microglia and enhanced survival of quiescent microglia (Cuevas et al., 2011b).
Although the effects of σ-receptor activation on microglial responses to ischemia have been studied, there is little information on how σ receptors may affect microglial responses to Aβ. In the absence of glutamatergic transmission, activation of σ-1 receptors has been shown to decrease neuronal apoptosis caused by Aβ25–35 (Marrazzo et al., 2005), but it is unclear how σ-1-activation will affect survival of microglia exposed to Aβ. It has been reported that there is a reduction in the number of σ-ligand binding sites and σ-1 receptors in the brains of patients with AD (Jansen et al., 1993, Mishina et al., 2008). If these receptors prevent microglial activation and/or senescence, a loss of these receptors may contribute to AD progression. Thus, it is of significant interest to determine how σ-receptor activation influences microglial responses to Aβ.
Experiments were carried out to determine whether afobazole, an agonist at both σ-1 and σ-2 receptors (Cuevas et al., 2011b), can reduce activation and toxicity of microglia by Aβ25–35. Afobazole decreased membrane ruffling and migration evoked by Aβ25–35. The effects of afobazole were dependent on activation of both σ-1 and σ-2 receptors and were blocked by inhibition of these receptors with BD-1047 and rimcazole. Furthermore, afobazole inhibited increases in the levels of the proapoptotic protein Bax and the death protease caspase-3 induced by Aβ25–35, resulting in an increase in microglia survival. Finally, afobazole was shown to prevent disruption of ATP signaling in microglia incubated in Aβ25–35, indicating that afobazole preserves microglial function after Aβ exposure.
Materials and Methods
Primary Cultures of Microglia.
Primary cultures of microglia were prepared from Sprague-Dawley mixed sex rat pups (postnatal day 2–3) as previously described by our laboratory (Hall et al., 2009, Cuevas et al., 2011b). The mixed glial cultures were incubated for 7–12 days at 37oC before experiments were performed. Microglia were mechanically separated from the cultures by brief shaking. Isolated microglia were resuspended in DMEM PLUS containing the following: 500 ml Dulbecco’s modified Eagle's medium with 4.5 g/l glucose, l-glutamine, and sodium pyruvate (DMEM) (Corning Cellgro, Manassas, VA); 40 ml horse serum (Corning Cellgro); 12.5 ml heat-inactivated fetal bovine serum (Thermo Scientific Hyclone, Logan, UT); and 5 ml 10× antibiotic/antimycotic (Corning Cellgro). These cells were used immediately for migration and cytotoxicity assays or plated on glass coverslips for 1 day for imaging experiments.
Membrane Morphology and Migration Assay.
Morphologic changes to microglia induced by Aβ25–35 were assessed by plating the cells on poly-l-lysine-coated coverslips, serum starving the cells for 4 hours in DMEM, and exposing the cells to 25 μM Aβ25–35 for 10 minutes at 37°C. When afobazole or DTG were used, microglia were incubated in the σ ligands (in DMEM) for 10 minutes before Aβ25–35 exposure. Membrane ruffling was visualized by labeling the cells with the filamentous actin probe phalloidin (AlexaFluor 488 conjugated; Life Technologies, Grand Island, NY). Quantification of membrane ruffling was carried out as previously described (Hall et al., 2009). In brief, cell ruffling was scored as follows: “0,” no ruffling and multiple filopodia; “1,” ruffling and filopodia; and “2,” fully ruffled with no filopodia.
Migration assays were carried out using a 48-well chemotaxis chamber (Neuro Probe, Inc., Gaithersburg, MD) as previously described (Cuevas et al., 2011b). The bottom wells of the chamber were filled with 25 μM Aβ25–35 in DMEM, and the top wells contained 1 × 106 freshly isolated microglia. These chambers were separated by a fibronectin-coated polycarbonate membrane with 8 μm pores at a density of 1 × 103 pores/mm2. Each well had an exposed filter area of 8 mm2. Test compounds (i.e., σ ligands) were added to both the top and bottom wells at identical concentrations. The microglia were permitted to migrate for 2 hours while incubated at 37°C. The membranes containing microglia were processed as previously described and mounted on slides using Vectashield Hardset containing 4′-6-diamidino-2-phenylindole (DAPI) (Vector Laboratories, Burlingame, CA) (Cuevas et al., 2011b). DAPI-positive cells were counted in four random fields of view per well and averaged to determine migration for that well. For each experiment, a minimum of two to three wells were used, and at least three migration experiments were carried out for each condition. The n values given represent the total number of fields of view recorded.
Calcium Imaging Measurements.
Intracellular Ca2+ concentrations ([Ca2+]i) as a function of time were measured using fluorescent imaging techniques and fura-2 as previously reported (Hall et al., 2009, Cuevas et al., 2011b). Microglia plated on coverslips were incubated for 1 hour at room temperature in DMEM-PLUS containing 25 µM Aβ and 3 μM acetoxymethyl ester fura-2 and 0.3% dimethyl sulfoxide. Prior to the experiments, the coverslips were washed in Aβ and fura-2-free physiologic saline solution (PSS) containing the following (in mM): 140 NaCl, 5.4 KCl, 1.3 CaCl2, 1.0 MgCl2, 20 glucose, and 25 HEPES (pH to 7.4 with NaOH). For studies on ATP-evoked [Ca2+]i changes, ATP was applied in PSS using a rapid application system as described previously (Cuevas and Berg, 1998).
ROS and NO Imaging.
Intracellular ROS and NO concentrations were measured in isolated microglia using the fluorometric indicators di-hydroethidiumbromide (DHE) and diacetate (4-amino-5-methylamino-2′,7′-difluorofluorescein diacetate) (DAF-FM), respectively. Cells plated on poly-l-lysine-coated coverslips were incubated at 37°C in DMEM-PLUS containing 10 µM DHE with 0.25% dimethylsulfoxide for 30 minutes or 5 µM DAF-FM with 0.1% dimethylsulfoxide for 1 hour. Before beginning the experiments, the coverslips were rinsed with PSS. DHE- and DAF-FM-labeled microglia were measured by digital microscopy (CoolSnap HQ2; Photometrics, Surrey, BC, Canada) excited at 470 and 535 nm and emission captured at 525 and 610 nm, respectively. Excitation exposure times were autocalibrated to the control (absence of drugs or Aβ) using this function on the acquisition software (Nikon Elements; Nikon, Melville, NY) and used for all slides in that dish. Image analysis was performed using the Nikon Elements software. Images were acquired every second for 10 seconds, and the average of the acquisitions was used for the fluorescence intensities (arbitrary units). Statistical analysis was performed using SigmaPlot 11 (Systat Software, Inc., San Jose, CA).
Cell Death Assay.
Microglia plated on coverslips were incubated at 37°C for 72 hours in DMEM-PLUS with 25 µM Aβ25–35 in the absence (Control) and presence of 30 µM afobazole. The coverslips were rinsed with phosphate-buffered saline (PBS) and then incubated for 30 minutes in 4 µM EthD-1 solution at room temperature. The coverslips were washed with PBS and deionized water, air-dried, and mounted on a microscope slide with Vectashield Hardset mounting media (Vector Laboratories). EthD-1-positive cells were visualized at 400× using a Zeiss Axioskop 2 and counted using ImageJ software in four random fields per slide. For each experiment, a minimum of three slides were used, and the results of at least three experiments were averaged together. Coverslips were visualized in bright field to confirm homogeneous cell density.
Microglia plated on poly-l-lysine coverslips were incubated at 37°C for 24 hours in DMEM-PLUS with 25 µM Aβ25–35 in the absence (Control) and presence of 30 µM afobazole. The coverslips were rinsed three times, followed by a stepwise ethanol fixation, permeabilized with 0.1% Triton X, rehydrated, and blocked with 2% BSA. Primary antibodies were diluted in PBS with 0.5% BSA and incubated at 4°C for 24 hours. The following dilutions were used: anti-Bax 1:20, anti-activated caspase-3 1:25, and anti-Bcl-2 1:100. The secondary antibodies used were Alexa Fluor 488-conjugated anti-mouse or anti-rabbit, as appropriate, and diluted at a ratio of 1:300 in PBS with 0.5% BSA. Coverslips were mounted onto slides using VectaShield with DAPI. Excitation and emission wavelengths used were 359 and 461 nm (DAPI) and 485 and 530 nm (Alexa Fluor 488), respectively. Cells were imaged at 400× and counted in four random fields per slide using ImageJ software. The average for the fields was used as the value for each slide. For each experiment, a minimum of three slides was analyzed, and the results of at least three experiments were averaged together, with n values representing number of experiments.
Reagents and Antibodies.
All test compounds were applied in DMEM-PLUS for 10 minutes prior to Aβ25–35 exposure and throughout the Aβ25–35 incubation. The following drugs were used in this investigation: DTG and BD-1047 (Tocris Biosciences, Ellisville, MO); rimcazole (Sigma-Aldrich, St. Louis, MO); fura-2 acetoxymethyl ester, DHE, DAF-FM, Live/Dead Viability/Cytotoxicity Kit, Alexa Fluor 488 Phalloidin (A-12379), Alexa Fluor 488 anti-mouse (A-11001), and anti-rabbit (A-11008) (Invitrogen); and Bax (ab5714), activated caspase-3 (ab32351) and Bcl-2 (ab32370) (Abcam, Cambridge, MA). Afobazole was generously provided by IBC Generium (Moscow, Russian Federation). Aβ25–35 was obtained from (American Peptides, Sunnyvale, CA).
Fura-2 fluorescence intensities as a function of time were recorded from microglia in the same manner as previously described (Katnik et al., 2006, Cuevas et al., 2011a, 2011b) and converted to [Ca2+]i using an in situ calibration kit (Invitrogen). Data were analyzed using SigmaPlot 11 (Systat Software, Inc., San Jose, CA). Data points represent peak means ± S.E.M. One- or two-way analysis of variance was used for determining significant differences for multiple group comparisons, as appropriate, followed by post hoc analysis using the Holm-Sidak test. Results were considered statistically significant if P < 0.05.
σ Receptors have been shown to reduce activation of microglia in response to various stimuli, including lipopolysaccharide, ATP, and monocyte chemoattractant protein-1 (Hall et al., 2009, Cuevas et al., 2011b). Experiments were carried out to see if the pan-selective σ-receptor agonist afobazole can also block microglial activation caused by application of amyloid β25–35 (Aβ25–35). One of the events associated with microglial activation that has been shown to be regulated by σ receptors is membrane ruffling (Hall et al., 2009). Thus, we examined how afobazole affects morphologic changes induced by the amyloid peptide in these cells. In cultured, quiescent microglia, filopodia are readily observed, and these membrane processes are not directly affected by incubation of the cells in 30 μM afobazole or a second pan-selective σ agonist, DTG (30 μM) (Fig. 1A). Incubation of the microglia in 25 μM Aβ25–35 for 10 minutes (37°C) resulted in retraction of filopodia and increased membrane ruffling (Fig. 1Aii). When afobazole (30 μM) was applied along with the Aβ fragment, there was a decrease in membrane ruffling compared with Aβ25–35 alone (Fig. 1Aiv). Pronounced membrane ruffling was still observed when DTG was added with Aβ25–35 (Fig. 1Avi). Quantification of the degree of membrane ruffling in microglia in experiments identical to those of Fig. 5A shows that Aβ25–35 increases membrane ruffling by more than fivefold (Fig. 1B). Incubation of the cells in either afobazole or DTG significantly reduced the degree of membrane ruffling caused by Aβ25–35 (Fig. 1B). However, the effect of afobazole on Aβ25–35-induced membrane ruffling was significantly greater than that produced by DTG, with the membrane ruffling being decreased by 70 ± 7% and 21 ± 9% by these drugs, respectively (Fig. 1B).
In addition to membrane ruffling, Aβ25–35 induces chemotaxis of microglia, and σ-receptor activation has been shown to disrupt migration of microglia in response to other stimuli (Hall et al., 2009, McLarnon, 2012). Experiments were therefore carried out to determine if application of afobazole can inhibit microglial migration caused by Aβ25–35. A concentration-response relationship for afobazole reduction in microglial migration when Aβ25–35 is used as the chemoattractant is shown in Fig. 2A. Afobazole, at concentrations as low as 10 μM, reduced the number of microglia migrating in response to Aβ25–35. Half-maximal inhibition of Aβ25–35-induced migration was observed at 122 μM afobazole, and maximal recorded block was achieved at 1 mM afobazole (Fig. 2A).
To verify that σ receptors are involved in afobazole suppression of microglial activation by Aβ25–35, the σ-receptor subtype-specific antagonists BD-1047 and rimcazole were used to block the effects of afobazole on migration. Afobazole was applied at 30 μM in the absence and presence of the selective σ-1- and σ-2-receptor antagonist BD-1047 (10 μM) (Matsumoto et al., 1995) and rimcazole (300 nM) (Gilmore et al., 2004), and migration in response to Aβ25–35 was assessed. Afobazole alone decreased microglial migration caused by Aβ25–35 by 29 ± 1%. This inhibition was reduced to 10 ± 1 and 8 ±1% when afobazole was combined with BD-1047 and rimcazole, respectively (Fig. 2B). Therefore, inhibition of either σ-1 or σ-2 significantly reduced afobazole-mediated suppression of microglial activation by Aβ25–35.
The microglia-mediated inflammatory response caused by Aβ is in part mediated via upregulation in the production of ROS and NO (Weldon et al., 1998, Bianca et al., 1999, Combs et al., 2001). Therefore, experiments were carried out to determine if afobazole affects ROS or NO levels in microglia after incubation in Aβ25–35. Incubation in afobazole alone produced a decrease in basal levels of ROS in isolated microglia (Fig. 3A). As predicted, incubation of microglia for 24 hours in Aβ25–35 increased ROS in the cells. However, afobazole failed to block this increase in ROS. Unlike ROS, basal NO levels in microglia were not significantly affected by afobazole (Fig. 3B). Aβ25–35 did produce a small but statistically significant increase in NO, and as with ROS, afobazole failed to block this effect of the amyloid peptide (Fig. 3B).
Given that microglial cell senescence has been suggested to contribute to AD pathology and that Aβ has been shown to produce microglial apoptosis (Korotzer et al., 1993, Streit et al., 2009), experiments were carried out to determine how activation of σ receptors with afobazole affects microglial survival after application of Aβ25–35. Figure 4A shows representative photomicrographs of microglia labeled with EthD-1 after 72-hour (37°C) incubation in media alone or media containing Aβ25–35 in the absence and presence of 30 μM afobazole. Few apoptotic microglia were detected when the cells were incubated in media alone (Fig. 4Ai), but there was a significant increase in EthD-1-positive cells after Aβ25–35 application (Fig. 4Aii). Afobazole alone did not visibly alter the number of EthD-1-positive cells (Fig. 4Aiii), but there was a decrease in the number of apoptotic cells in the afobazole- Aβ25–35 group (Fig. 4Aiv) relative to Aβ25–35 alone (Fig. 4Aii). In identical experiments, Aβ25–35 was found to produce a nearly threefold increase in the number of apoptotic cells relative to control (no Aβ25–35) (Fig. 4B). However, afobazole application significantly counteracted the proapoptotic effects of Aβ25–35 such that no significant increase in EthD-1 labeling was observed in the Aβ25–35 + afobazole group (Fig. 4B).
Microglial cell death induced by Aβ25–35 has been linked to upregulation of the proapoptotic gene Bax (Jang et al., 2005). Immunocytochemical experiments were conducted to determine if levels of Bax expression increase in our rat microglia model and if afobazole can affect expression of this protein. Representative photomicrographs of microglia double labeled with DAPI and anti-Bax antibody are shown in Figs. 5A. Whereas Bax labeling was rarely observed in control (untreated) microglia, Bax immunolabeling was prominent in cultures treated for 24 hours (37°C) with Aβ25–35 (Figs. 5A, i and ii). Incubation of microglia in afobazole alone had no effect on Bax expression levels, but application of the σ agonist inhibited Aβ25–35 -induced elevated Bax expression (Fig. 5A, iii and iv). Compilation of data from several identical experiments indicates that approximately 20% of control microglia expressed Bax but that this number increases to over 80% after Aβ25–35 incubation (Fig. 5B). Although an increase in Bax expression was noted when microglia were incubated in a combination of afobazole and Aβ25–35, there was a 54 ± 4% reduction in the number of cells showing Bax reactivity when compared with microglia incubated in Aβ25–35 alone (Fig. 5B).
Caspase-3 has also been implicated in Aβ25–35-induced microglial apoptosis, and σ-1-receptor activation has been shown to induce caspase-3 downregulation in these cells (Jang et al., 2005, Tchedre and Yorio, 2008). Thus, we examined the effects of afobazole on the levels of the death protease, caspase-3, after microglial cell were exposed to Aβ25–35 for 24 hours (37°C). Incubation of microglia in Aβ25–35 increased the number of cells expressing the active form of caspase-3 relative to control (Fig. 6A, i and ii). Although afobazole alone had no significant effect on activated caspase-3 expression, afobazole inhibited the increases in active caspase-3 evoked by Aβ25–35 (Fig. 6A, iii and iv). In identical experiments, comparing the number of cells expressing activated caspase-3 in cultures treated with Aβ25–35 alone to those treated with Aβ25–35 + afobazole indicates that afobazole reduces caspase-3 expression by 47 ± 5% (Fig. 6B).
The anti-apoptotic gene product, Bcl-2, has been shown to be upregulated by interventions that decrease microglial cell death caused by Aβ25–35 (Jang et al., 2005). In neurons, the expression levels of this protein increase after 24-hour exposure to Aβ and then decreases after 48 hours (Kim et al., 1998). Thus, to explore how afobazole protects microglia from Aβ25–35-induced apoptosis, we used the 48-hour time point to assess how afobazole affects Bcl-2 expression. Bcl-2 was readily detected at 48 hours in microglia under all conditions tested (Fig. 7A). Quantification of the data shows that afobazole application produced a small, but significant, increase of 14 ± 1% in the number of microglia expressing Bcl-2 (Fig. 7B). Similarly, Aβ25–35 alone increased the number of cells expressing Bcl-2 by 11 ± 1%. Although afobazole coapplication did not increase Bcl-2 expression significantly above afobazole alone, the combination of afobazole and Aβ25–35 yielded a small but significant increase (4 ± 1%) in the number of microglia expressing Bcl-2 after 48 hours compared with Aβ alone (Fig. 7B).
Although our observations indicate that activation of σ receptors with afobazole decreases microglial cell death, it is unclear if the surviving cells have been compromised such that they fail to exhibit responses to normal stimuli. Thus, functional studies were carried out to evaluate microglia using ATP to stimulate changes in [Ca2+]i after Aβ25–35 application in the absence and presence of afobazole. Figure 8A shows a series of traces of [Ca2+]i as a function of time recorded from four microglia incubated in Aβ25–35 for 24 hours and exposed to ATP in the absence and presence of afobazole. Focal application of 100 μM ATP produced transient elevations in [Ca2+]i in microglia under control conditions and when the cells were incubated in 25 μM Aβ25–35 (Fig. 8A, top traces). However, the ATP-induced elevations in [Ca2+]i were greatly reduced in the microglia incubated in Aβ25–35. In contrast, microglia incubated in 30 μM afobazole displayed similar ATP-induced increases in [Ca2+]i in absence and presence of the Aβ fragment (Fig. 8A, bottom traces). In identical experiments, ATP evoked comparable increases in [Ca2+]i in microglia incubated in media alone (control) and 30 μM afobazole (Fig. 8B). However, microglia incubated in Aβ25–35 showed a statistically significant 34 ± 4% reduction in ATP-evoked [Ca2+]i responses (Fig. 8B). Although afobazole preincubation did not have a long-term effect on ATP-evoked responses, coapplication of 30 μM afobazole along with the 25 μM Aβ25–35 rescued the ATP-evoked responses such that the increases in [Ca2+]i observed in this group were statistically similar to both the control group and the afobazole alone group. The [Ca2+]i responses in the afobazole + Aβ25–35 group were 44 ± 8% greater than the Aβ25–35 only group (Fig. 8B).
The most salient finding reported here is that activation of both σ-1 and σ-2 receptors by afobazole rescues microglia from the deleterious consequences of Aβ25–35 exposure. Afobazole application reduced the activation of microglia caused by incubation of the cells in Aβ25–35 as evident by a decrease in membrane ruffling and cell migration induced by the Aβ fragment. The apoptosis of microglia induced by Aβ25–35 was also reduced by afobazole. This decrease in cell death is likely attributed to a reduction in the expression of the proapoptotic gene Bax and the death protease caspase-3. Afobazole also preserved functional responses of the microglia to ATP after Aβ25–35 incubation, which has significant implications for Alzheimer’s disease progression.
Previous studies have shown that microglia rapidly respond to Aβ by withdrawing their filopodia and assuming an amoeboid, ruffled morphology (Araujo and Cotman, 1992). This early event is indicative of microglial activation and precedes the inflammatory response that includes release of neurotoxic substances from the microglia (Combs et al., 2001, Garcao et al., 2006, Orellana et al., 2011). Afobazole (30 μM) effectively blocked membrane ruffling in response to Aβ25–35. Although a second pan-selective σ-receptor agonist, DTG (30 μM), also decreased membrane ruffling caused by Aβ25–35, this drug was significantly less potent than afobazole. It has been shown that higher concentrations of DTG (100 μM) are also required to block ATP-induced microglial membrane ruffling (Hall et al., 2009).
Afobazole also effectively blocked microglial migration induced by Aβ25–35. It was shown previously that various Aβ fragments, including Aβ25–35, are microglial chemotactic agents (Davis et al., 1992). The Aβ-induced chemotaxis is in part due to release of ATP from microglia after Aβ25–35 application and subsequent activation of P2Y2 receptors on these cells (Kim et al., 2012). We previously showed that afobazole blocks microglial migration induced by activation P2Y receptors (Cuevas et al., 2011b). Afobazole appears to inhibit microglial migration caused by UTP via σ-receptor-mediated disruption in P2Y signaling, because afobazole treatment also inhibited P2Y-induced elevation in [Ca2+]i in microglia (Cuevas et al., 2011b). The IC50 for afobazole inhibition of microglial migration caused by Aβ25–35 is 120 μM, which is similar to the low-affinity site reported for afobazole inhibition of UTP-induced microglial migration (102 μM) (Cuevas et al., 2011b). Therefore, afobazole may in part prevent microglial chemotaxis caused by Aβ25–35 via σ-receptor-mediated inhibition of P2Y receptors in these cells.
Data shown here indicate that both σ-receptor subtypes are involved in afobazole inhibition of Aβ25–35-evoked microglial activation. Afobazole-mediated inhibition of microglial migration was reduced by ∼50% by 10 μM BD-1047, which is indicative of a σ-1-receptor-mediated effect (Matsumoto et al., 1995). An identical concentration of BD-1047 had a similar effect on afobazole inhibition of ATP-induced microglial cell migration (Cuevas et al., 2011b). Likewise, involvement of σ-2 receptors is confirmed by results obtained using the σ-receptor antagonist rimcazole, which binds to σ-2 receptors with nanomolar affinity (Ferris et al., 1986, Rybczynska et al., 2008). Nanomolar concentrations of rimcazole also blocked afobazole inhibition of microglial migration in response to ATP (Cuevas et al., 2011b).
Given that one of the consequences of microglial activation is increased oxidative and nitrosative stress, which leads to Aβ25–35 toxicity (Weldon et al., 1998, Bianca et al., 1999, Combs et al., 2001), we examined how afobazole affects Aβ25–35-induced increases in cellular ROS and NO levels. Consistent with previous reports, both ROS and NO levels were increased significantly after Aβ25–35 application. However, afobazole failed to attenuate the increases in ROS and NO evoked by the Aβ fragment. Activation of σ receptors has been shown to decrease oxidative injury and nitrosative stress in various models. It was recently shown, using σ-receptor knockout mice, that σ-1 has a dual function of both reducing oxidative stress and activating antioxidant response elements (Pal et al., 2012). In microglia, stimulation of σ receptors by DTG reduces NO production in response to lipopolysaccharide (Hall et al., 2009). Taken together, however, our data suggest that the reduction in Aβ-evoked microglial cell toxicity provided by afobazole application is not the result of σ-receptor-mediated decrease in oxidative or nitrosative stress.
Although decreased activation of microglia will be beneficial in AD by reducing the neuroinflammatory response, recent data suggest that enhanced microglial survival is also important for improved outcomes given the ability of these cells to phagocytose Aβ plaques (Streit et al., 2009, Solito and Sastre, 2012). Application of 30 μM afobazole significantly reduced the number of microglia undergoing apoptosis after 72 hours of incubation in Aβ25–35. The concentrations of afobazole shown to be effective here are identical to those previously shown to decrease microglial cell death caused by ischemia in a σ-receptor-dependent manner and consistent with the binding affinity of afobazole for σ-1 receptors (Seredenin et al., 2009, Cuevas et al., 2011a,b). This observation suggests that afobazole is similarly acting through σ-receptor activation to prevent microglial cell death. Both melatonin and N-acetylcysteine have been shown to decrease Aβ25–35-evoked apoptosis in microglia via a reduction in ROS (Jang et al., 2005). Given that afobazole does not appear to affect Aβ25–35-elicited increases in ROS, it is unlikely that afobazole provides cytoprotection via a similar mechanism to these compounds.
Our findings demonstrate that afobazole mitigates Aβ25–35-induced apoptosis of microglia by decreasing expression of the proapoptotic gene Bax and the death protease caspase-3 while preserving the levels of the antiapoptotic gene Bcl-2. An increase in Bax levels has been observed in both neurons and microglia of AD brains (Su et al., 1997). Moreover, interventions that decrease Aβ25–35-induced microglial apoptosis have been shown to lower Bax expression (Jang et al., 2005, Shang et al., 2012). Likewise, caspase activity has been detected in plaque-associated neurons and microglia in AD brains (Yang et al., 1998). Caspase-3 levels have been shown to be upregulated in microglia after Aβ administration, whereas reduced levels of this death protease have been associated with inhibition of Aβ toxicity (Shang et al., 2012). Conversely, expression of the antiapoptotic gene Bcl-2 is inversely correlated with disease severity in AD. Increases in Bcl-2 levels promote microglial survival after exposure to Aβ25–35 (Satou et al., 1995, Jang et al., 2005). Therefore, afobazole likely enhances microglial survival and resistance to Aβ25–35 toxicity via a complex response involving both antiapoptotic and proapoptotic genes. This multifaceted effect is consistent with afobazole acting as a σ agonist, because σ-receptor activation has been shown to modulate Bax, caspase-3, and Bcl-2 levels (Tchedre and Yorio, 2008, Meunier and Hayashi, 2010).
Importantly, not only did afobazole protect microglia from Aβ25–35-evoked apoptosis, but application of the σ agonist preserved the functional capacity of these cells, as evident from the continued [Ca2+]i responses to ATP. Ultimately, one of the theoretical benefits to increased microglial survival would be continued clearance of Aβ, because this peptide has been shown to be phagocytosed by microglia (Wilcock et al., 2004, Herber et al., 2007). This hypothesis is further supported by the recent observation that microglial clearance of Aβ directly correlates with expression of the gene CD33, such that a CD33 deficiency reduces brain levels of both insoluble Aβ1–42 and amyloid plaque burden in a mouse AD model (Griciuc et al., 2013). Moreover, there is a decrease in both CD33 expression and insoluble Aβ1-42 in the brains of individuals expressing the minor allele of the CD33 (SNP rs3865444) gene, which was identified previously as imparting protection against AD (Naj et al., 2011, Griciuc et al., 2013). The fact that responses to ATP itself are preserved is significant, because the clearance of Aβ has been shown to be dependent of the activation of P2Y2 receptors. Microglia respond to Aβ by releasing ATP, and this nucleotide functions in a paracrine manner, acting on P2Y2 receptors expressed on the microglia to enhance Aβ uptake by these cells (Kim et al., 2012). We previously showed that in our model these increases in [Ca2+]i are due to activation of P2Y receptors, because they can be elicited by UTP (Cuevas et al., 2011b). Therefore, by preserving ATP responses in microglia, afobazole will permit these cells to respond to Aβ1-42 in vivo and potentially decrease amyloid burden.
In conclusion, our study demonstrates that afobazole can mitigate activation of microglial cells in response to Aβ25–35. Afobazole also reduced microglial toxicity caused by Aβ25–35, but this cytoprotection does not involve regulation of either ROS or NO production caused by the amyloid peptide. The observed decrease in apoptosis is due to afobazole-mediated regulation of multiple genes, with a noted decrease in Bax and caspase-3 and a concomitant increase in Bcl-2 expression. In addition to the decrease in cell death, afobazole preserved functional activity of the cells as evident by the conserved ATP-induced increases in [Ca2+]i after prolonged incubation in Aβ25–35. Taken together our data suggest that afobazole may prevent microglia from entering the activated state associated with neurotoxicity after exposure to Aβ and increasing their survival, which will facilitate clearance of the amyloid peptide. These properties make afobazole an attractive drug for potential Alzheimer’s therapeutics.
The authors thank Christopher Katnik, Ph.D., and Nivia Cuevas, R.Ph., for comments on a draft of this manuscript.
Participated in research design: Behensky, Yasny, Shuster, Seredenin, Petrov, Cuevas.
Conducted experiments: Behensky.
Performed data analysis: Behensky.
Wrote or contributed to the writing of the manuscript: Behensky, Yasny, Cuevas.
- Received July 25, 2013.
- Accepted August 20, 2013.
This study was supported by a grant from IBC Generium (to J.C.).
- amyloid-beta peptide fragment 25–35
- Alzheimer’s disease
- N-[2-(3,4-dichlorophenyl)ethyl]-N-methyl-2-(dimethylamino)ethylamine dihydrobromide
- bovine serum albumin
- intracellular calcium concentration
- diacetate (4-amino-5-methylamino-2',7'-difluorofluorescein diacetate)
- Dulbecco’s modified Eagle’s medium
- endoplasmic reticulum
- ethidium homodimer-1
- nitric oxide
- phosphate-buffered solution
- physiological saline solution
- reactive oxygen species
- Copyright © 2013 by The American Society for Pharmacology and Experimental Therapeutics