All imaging experiments were performed on 6–16 weeks old C57Bl/6J wild type mice of both sexes (for EAE see below) (The Jackson Laboratory, Bar Harbor, ME, USA). Mice were held in the animal facilities of the Institute of Cell and Systems Biology (University of Hamburg) and the University Medical Center Hamburg-Eppendorf (UKE). Mice were kept in a 12/12 light/dark cycle and had access to food and water ad libitum. All experiments were performed in accordance with German law, the directive 2010/63/EU and were approved by the local ethics committee and the Behörde für Justiz und Verbraucherschutz, Lebensmittel und Veterinärwesen Hamburg.

To express the genetically encoded cAMP indicator Flamindo2 specifically in astrocytes, we used endotoxin-free recombinant adeno-associated viruses AAV^2/PhP.eBhGFAP-Flamindo2 and AAV^2/PhP.AX hGFAP-Flamindo2 that were prepared at the UKE vector facility. The plasmid pAAV-hGFAP-Flamindo2 was a gift provided by Hajime Hirase, pAAV-gfaABC1D-GRAB_ATP1.0 Addgene #167579 was a gift from Yulong Li, the capsid plasmid PhP-AX Addgene #195218 was kindly provided by Vivianna Gradinaru and Xinhong Chen and the capsid plasmid PhP.eB Addgene #103005 was a gift from V. Gradinaru (27–30). Virus suspensions were diluted with sterile saline and injected i.v. into the orbital sinus of isoflurane-anesthetized mice. Each animal received a total volume of 70 µl virus suspension containing 1 x 10¹¹ viral genomes (quantified by qPCR for WPRE sequence).

Experiments were performed in artificial cerebrospinal fluid (ACSF) that contained (in mM): NaCl, 120; NaHCO₃, 26; NaH₂PO₄, 1; KCl, 2.5; D-glucose, 2.8; CaCl₂, 2; MgCl₂, 1. For preparation of brain slices, Na⁺- and Ca²⁺-reduced ACSF was used (in mM): NaCl, 83; NaHCO₃, 26.2; NaH₂PO₄, 1; KCl, 2.5; CaCl₂, 0.5; MgSO₄, 2.5. In Na⁺- and Ca²⁺-reduced ACSF, D-glucose was increased to 20 mM and sucrose was added (70 mM) to compensate for reduced Na⁺ concentration and maintain osmolarity. The purinergic ligands ATP and adenosine were purchased from Merck/Sigma-Aldrich (Taufkirchen, Germany), N-Cyclopentyladenosine (N⁶-CPA), 2-Chloro-N-cyclopentyl-2’-methyladenosine (MeCCPA), 4-[2-[(6-Amino-9-b-D-ribofuranosyl-9H-purin-2-yl)thio]ethyl]benzenesulfonic acid ammonium salt (PSB 0777) from Bio-techne (Wiesbaden, Germany), forskolin and 4-(2-[7-Amino-2-(2-furyl) [1,2,4]triazolo [2,3-a][1,3,5]triazin-5-ylamino]ethyl) phenol (ZM241385) from Abcam (Cambridge, UK) and tetrodotoxin (TTX) from Biozol (Hamburg, Germany). Stock solutions were prepared according to the manufacturer’s instructions, dissolved in ACSF at the final concentrations given in the results section and applied with the perfusion system.

Mice were anesthetized using isoflurane (5% v/v in oxygen) and decapitated. Brains were quickly dissected and transferred into chilled Ca²⁺-reduced ACSF that was continuously gassed with carbogen (95% O₂/5% CO₂; buffered to pH 7.4 with CO₂/bicarbonate) during the entire preparation procedure. Two hundred micrometer thick slices of the bulbs were cut using a vibratome (VT1200S, Leica). Slices were stored in ACSF gassed with carbogen in darkness for 30 min at 30°C and then at room temperature until starting experiments.

Changes in the intracellular cAMP concentration were visualized by the genetically encoded cAMP indicator Flamindo2, based on the circular permuted fluorescent protein citrine and the cAMP binding domain of mEPAC1 (27). Binding of cAMP to the cAMP binding site decreases the fluorescence intensity of citrine, hence an increase in cAMP concentration is reflected by a decrease in fluorescence intensity. Flamindo2 was excited at 488 nm and fluorescence was collected between 500 and 530 nm using a confocal microscope (eC1, Nikon). Time series of images were recorded at a frame rate of 1 frame every 3-5 seconds. To visualize increases in the extracellular ATP concentration, we expressed the EGFP-based fluorescent ATP sensor GRAB_ATP1.0 in astrocytes, which was excited at 488 nm and emitted light in the range of 500-530 nm (30). Increases in extracellular ATP were evoked by electrical stimulation of axons of olfactory receptor neurons that release ATP at their axon terminals. To stimulate these axons and trigger action potentials, a glass micropipette (GC150TF-10, Harvard Apparatus, USA) pulled with a patch-pipette puller (Flaming/Brown P97, Sutter Instruments, Novato, CA, USA) with electrical resistance of 1-3 MΩ was inserted into the olfactory nerve layer of a brain slice and trains of electrical pulses (0.1 ms pulse width, 2.3 mA amplitude, 20 Hz for 10 s) were applied (Digitimer DS3, Digitimer Ltd, Hertfordshire, England).

Successful infection and astrocyte-specific expression of Flamindo2 was verified by co-immunolabeling of Flamindo2 and astrocytic markers. Since Flamindo2 is based on citrine, a yellow-fluorescent variant of GFP, we used anti-GFP (chicken, #132 006, Synaptic Systems, Göttingen, Germany) to visualize Flamindo2 and a combination of anti-GFAP (rabbit, Z0334, Dako, Hamburg, Germany) and anti-S100 (rabbit, Z0311, Dako) to visualize astrocytes, all primary antibodies were used at a concentration of 1:500. Free floating slices (220 µm thick) were cut from brains fixed with 4% formalin using a vibratome (VT1000, Leica, Bensheim, Germany). Slices were rinsed with PBS (in mM: NaCl, 130; Na₂HPO₄, 7; NaH₂PO₄, 3) and blocked with 10% normal goat serum (NGS) and 0.5% Triton X-100 in PBS for one hour. Primary antibodies were incubated in 1.0% NGS and 0.05% Triton X-100 in PBS at 4°C for 36 hours. Secondary antibodies (Alexa Fluor 488 goat anti-chicken, #Ab150173, Abcam, Cambridge, UK; Alexa Fluor 555 goat anti-rabbit, #A21429, Life Technologies GmbH, Darmstadt, Germany) were incubated in PBS at 4°C overnight. After rinsing in PBS slices were mounted in Shandon Immu-Mount (Life Technologies GmbH) and recorded using a confocal microscope (eC1, Nikon, Düsseldorf, Germany). Since both Flamindo2 and Alexa Fluor 488 emission was recorded in the same range of wavelengths (500-530 nm), both signals add up to optimize the visualization of Flamindo2-expressing cells.

Mice were acclimatized two weeks before EAE induction in the experimental housing facility. EAE induction was performed as described before (31). Briefly, 13 to 14-week-old female C57Bl/6J mice were anesthetized with isoflurane (1% to 2% v/v in oxygen). Immunization was performed by a subcutaneous injection into the two lateral sides of the lower back area. Injections consisted of 200 µg myelin oligodendrocyte glycoprotein peptide (MOG35–55, peptides & elephants GmbH, Hennigsdorf, Germany) in saline mixed with complete Freund’s adjuvant (CFA, BD Difco) containing 4 mg/ml heat-killed Mycobacterium tuberculosis (BD Difco). On the day of immunization and two days later, 250 ng pertussis toxin (Merck Millipore) in 100 µl PBS was injected intraperitoneally. Control mice were not treated. Animals were weighed and scored daily for the presence of neurological signs from day 7 to the day of preparation. Mice were scored for clinical signs by the following system: 0: no clinical deficits; 1: tail weakness; 2: hind limb paresis; 3: partial hind limb paralysis; 3.5: full hind limb paralysis; 4: full hind and forelimb paralysis or plegia. Animals that reached a score of 4, or scored 3.5 for more than 7 days, or lost ≥ 25% of starting weight were euthanized according to the regulations of the local Animal Welfare Act. At the peak of the disease, the mean EAE score was 2.8 ± 1.16.

Data were evaluated with Nikon EZ-C1 Viewer (Nikon), processed using Excel (Microsoft, USA) and statistical tests were applied with Origin Pro 9.1 (OriginLab Corporation, Northampton, USA). The expression of Flamindo2 was highest in astrocytes in the glomerular layer. We therefore only analyzed astrocytes in the glomerular layer. Somata of Flamindo2-expressing astrocytes were marked by a region of interest (ROI) and the mean fluorescence intensity within the ROI was measured throughout the time series. In a subset of experiments, cell processes were analyzed in addition to somata. The mean fluorescence intensity (F) was normalized to the basal fluorescence intensity at rest which was set to 100%, hence changes in fluorescence intensity are stated as ΔF in %. Since an increase in the cAMP concentration in astrocytes is reported as a decrease in Flamindo2 fluorescence, we inverted the fluorescence traces in the figures to render perception of the results more intuitively.

Fluorescence changes of less than 5% ΔF were in the range of noise and were not counted as responses. All data values are given as mean ± standard error of the mean (SEM) with n representing the number of astrocytes analyzed. For each set of experiments, at least three animals were used. Means were tested for statistical differences using the Mann-Whitney-U test. For multiple comparisons the Kruskal-Wallis ANOVA followed by Dunn’s post hoc test was used. Statistical differences were indicated with *p < 0.05, **p < 0.01 and ***p < 0.001. Bar graphs and the sigmoidal fit of the dose-response curve were prepared using OriginPro 9.1. Bar graphs represent means + SEM. They were overlaid by the entire population of single data points that were normalized to the mean of the first application/stimulation.

The olfactory bulb is the most rostral part of the mouse brain and receives afferent input from olfactory receptor neurons (ORNs) located in the olfactory epithelium in the nasal cavity ( Figure 1A ) (32). Axons of ORNs terminate in the glomerular layer of the olfactory bulb where they release both ATP and glutamate as neurotransmitters (33) ( Figure 1B ). Astrocytes respond to glutamate and ATP as well as to adenosine derived from ATP with Ca²⁺ signaling, however, whether glutamate, ATP or its degradation product adenosine stimulates cAMP signaling in olfactory bulb astrocytes has not been shown before.

To investigate cAMP signaling and visualize cAMP dynamics in astrocytes of the olfactory bulb, we used the genetically encoded cAMP indicator Flamindo2. We expressed Flamindo2 in astrocytes by injection of AAV carrying the Flamindo2 gene ( Figure 1C ). Flamindo2 expression was under control of the gfaABC1D promoter to assure astrocyte-specific expression (34). We verified specific expression in astrocytes by co-labeling of Flamindo2 and astrocytes using anti-GFP and combined anti-GFAP/S100B immunostaining, respectively. Approximately 50% of astrocytes in the glomerular layer expressed Flamindo2 ( Figure 1D ).

Purinergic signaling is strongly involved in a plethora of neurological diseases (35) and we focused on the contribution of purinoceptors in cAMP signaling in olfactory bulb astrocytes. Adenosine activates both G_S-coupled receptors that stimulate adenylyl cyclases and hence increase the intracellular cAMP concentration and G_i-coupled receptors that inhibit adenylyl cyclases and thereby decrease the cAMP concentration (36). We applied adenosine to acute brain slices and detected a decrease in Flamindo2 fluorescence in glomerular astrocytes, indicative for an increase in cAMP concentration ( Figure 1E ). To better reflect the changes in cAMP concentration, we inverted the Y axis (-ΔF) ( Figure 1F ). We applied different concentrations of adenosine between 3 and 300 µM and plotted the results in a dose-response chart ( Figures 1G, H ). The results show an EC₅₀ value of 12.8 µM, as extrapolated from the sigmoidal fit, and that at 300 µM adenosine the response is close to saturation. For the following experiments, we used 30 µM adenosine, which evokes large non-saturating responses. Repetitive application of adenosine (10 min interval) resulted in cAMP transients with a small yet significant decrease in amplitude ( Figure 2A ). The first application of adenosine evoked a cAMP response with a mean amplitude of 42.5 ± 1.6% -ΔF (n = 238). The amplitude of the responses induced by the second and third applications decreased to 75.8% (p < 0.001) and 59.6% (p < 0.001) compared to the first application ( Figure 2B ). Inhibition of voltage-gated Na⁺ channels with tetrodotoxin (TTX) and thus suppression of neuronal activity had no effect on the adenosine-induced cAMP signals, indicating that the responses in astrocytes did not depend on neuronal neurotransmitter release but were directly evoked in astrocytes ( Figures 2C, D ).

Both A₁ and A_2A receptors are the main adenosine receptors expressed in the olfactory bulb and modulating neuronal performance (31, 37, 38). Olfactory bulb astrocytes express A_2A receptors that mediate an increase in the cytosolic Ca²⁺ concentration (39). The canonical pathway downstream of A_2A receptor stimulation is activation of adenylyl cyclase resulting in an increase in the cAMP concentration. Therefore, we first tested the effect of the A_2A receptor antagonist ZM241385 on adenosine-mediated cAMP increases in olfactory bulb astrocytes in acute brain slices. ZM241385 entirely blocked adenosine-evoked cAMP transients in the majority of astrocytes, indicating that A_2A receptors are the main drivers of adenosine-induced cAMP signals ( Figures 3A, E ). This was confirmed by the agonistic effect of PSB0777, a highly specific A_2A receptor agonist (40) ( Figures 3B, E ). The PSB0777-evoked increase in cAMP was blocked by ZM241385. It is interesting that in some astrocytes the inhibition by ZM241385 was incomplete, which might be due to local concentration gradients and being outcompeted by the agonist, or might indicate that some astrocytes express an additional adenylyl cyclase-activating adenosine receptor such as A_2B ( Figure 3E ). Another widely expressed adenosine receptor found in astrocytes is the A₁ receptor, that is linked to G_i and hence triggers a decrease in cAMP concentration by inhibiting adenylyl cyclases (41). The A₁ receptor agonist N⁶-CPA, however, neither decreased the cAMP concentration in olfactory bulb astrocytes at resting conditions nor after increasing the cAMP concentration by the adenylyl cyclase activator forskolin ( Figures 3C, F ). Since N⁶-CPA possesses only weak adenosine receptor subtype selectivity, we additionally applied the highly selective and effective A₁ receptor agonist MeCCPA (42). MeCCPA also failed to induce a decrease in cAMP concentration ( Figure 3D ). Thus, in the majority of astrocytes adenosine elicits cAMP responses by stimulating A_2A receptors, while A₁ receptors are not involved.

The main source of extracellular adenosine in the CNS is synaptically released ATP that is degraded by ectonucleotidases, although alternative release mechanisms have also been proposed (43–45). The olfactory bulb receives input from axons of ORNs that are synaptically connected to mitral and tufted cells as well as to some interneurons (32). In the developing olfactory bulb, we have shown that release of ATP from ORN axons and subsequent degradation to adenosine leads to A_2A receptor activation and Ca²⁺ signaling in astrocytes of the glomerular layer (39). Therefore, we were interested in whether adenosine could be generated from ATP released from ORN axons and stimulate cAMP signaling in astrocytes in adult mice. We first tested the effect of bath-applied ATP on cAMP in astrocytes of acute brain slices of the olfactory bulb. Application of 30 µM ATP resulted in an increase in intracellular cAMP in astrocytes of the glomerular layer ( Figure 4A ). The mean amplitude of the ATP-evoked cAMP rise was 23.5 ± 1.2% -ΔF (n = 81), which decreased slightly upon repetitive ATP application. The second and third application of ATP resulted in cAMP signals with 79% and 73% (n = 57), respectively, when compared to the first application. The ATP-induced cAMP transients were entirely inhibited by ZM241385 (n = 23), indicating that ATP-evoked cAMP signals were mediated by adenosine acting on A_2A adenosine receptors ( Figure 4B ). We used the genetically encoded extracellular ATP sensor GRAB_ATP1.0, introduced into astrocytes by AAV injection, to visualize changes in extracellular ATP (30). We electrically stimulated ORN axons in the nerve layer of the olfactory bulb and could detect an increase in extracellular ATP in the glomerular layer ( Figure 4C ). We calibrated the GRAB_ATP1.0 signal by applying ATP at different concentrations via the perfusion system. When compared to the calibration values, the peak of ATP increases evoked by electrical stimulation of axons amounted to values between 3 and 10 µM ( Figure 4D ). On average, the extracellular ATP concentration reached 3.6 µM during axon stimulation (10 slices from 4 animals). We next assessed whether ATP released from ORN axons is able to increase cAMP in astrocytes of the glomerular layer. Electrical stimulation of ORN axons resulted in an increase in astrocytic cAMP of 25.2 ± 1.6% -ΔF (n = 76) ( Figure 4E ). The cAMP response evoked by a second stimulation was significantly smaller and amounted to 43.3 ± 1.8% compared to the first application (n = 76; p < 0.001). When the second stimulation was applied in the presence of ZM241385, no cAMP transients could be measured in the majority of astrocytes (82 out of 89 cells) ( Figure 4F ), while in 7 cells ZM241385 had no effect. In summary, our results show that adenosine produced from bath-applied and synaptically released ATP results in ZM241385-sensitive increases of intracellular cAMP in olfactory bulb astrocytes ( Figure 4G ).

The role of A_2A receptors in astrocytes in neurodegenerative diseases such as Alzheimer’s and Parkinson’s disease is well described (46–48). Less is known about astrocytic A_2A receptors in neuroinflammation and we used EAE mice as a model to study purinergic signaling in neuroinflammation. Purinergic signaling in astrocytes is able to mediate both pro- and anti-inflammatory responses and we were interested in whether purinergic cAMP signaling in astrocytes is altered in olfactory bulb astrocytes in the acute phase of EAE. We applied adenosine, the A_2A-specific agonist PSB0777 and ATP and compared cAMP responses in astrocytes in acute slices of healthy control mice with those of mice suffering from EAE at the peak of the disease, i.e. day 14 to 16 post immunization. To gain additional information of the cAMP responses, we analyzed cAMP signals in somata and cell processes of olfactory bulb astrocytes separately ( Figure 5A ). The amplitude of cAMP rises in somata was significantly larger compared to the cell processes using either adenosine or ATP ( Supplementary Table 1 ; Figures 5B, C ). Similar results were found in brain slices from animals with EAE ( Figures 5D, E ; Supplementary Table 1 ). We compared the amplitudes of the cAMP responses between healthy mice and mice with EAE but did not detect any significant differences, regardless whether we tested the different agonists or whether either values from somata or cell processes were compared ( Figure 5F ). Hence, at least in olfactory bulb astrocytes, EAE appears not to have a major impact on purinergic cAMP signaling.

Bath application of adenosine reliably evoked cAMP transients in olfactory astrocytes that were not affected when neuronal activity was blocked using TTX, indicating that adenosine directly stimulated astrocytes. Adenosine-induced cAMP increases were entirely blocked by ZM241385, an A_2A-preferring adenosine receptor antagonist (49), in the majority of astrocytes. However, at the concentration used (1 µM), ZM241385 could also inhibit other subtypes of adenosine receptors. Therefore, we applied the selective A_2A agonist PSB0777 to verify the presence of A_2A receptors in olfactory bulb astrocytes (40). PSB0777 also evoked increases in cAMP, indicating the presence of A_2A in olfactory bulb astrocytes. In addition, in some astrocytes, ZM241385 reduced adenosine-evoked cAMP responses only partially, which could suggest the contribution of another adenosine receptor such as the A_2B receptor. However, this is unlikely as cAMP responses evoked by the A_2A-selective antagonist PSB0777 were also only partially reduced by ZM241385 in some cells favoring an incomplete inhibition of the A_2A receptor. A₁ receptors are linked to inhibition of adenylyl cyclase and hence lead to a reduction of the cAMP concentration, thereby counteracting A_2A receptors. In our study, we applied two A₁ receptor agonists, N⁶-CPA and MeCCPA, that did not decrease the cAMP concentration. Even when the resting cAMP concentration was artificially increased by the adenylyl cyclase activator forskolin, stimulation of A₁ receptors had no effect on the cAMP concentration, indicating that glomerular astrocytes do not express A₁ receptors linked to adenylyl cyclase inhibition. We cannot exclude the contribution of A₃ receptors to adenosinergic cAMP dynamics in the astrocytes (50, 51), but the strong effect of ZM241385 on adenosine-evoked cAMP signals suggests that A_2A receptors are the main players in purinergic cAMP signaling in these cells. We found an EC₅₀ value of 12.8 µM of the adenosinergic cAMP response, which is about twentyfold higher than published EC₅₀ values of A_2A receptor activation. However, adenosine is rapidly degraded to inosine by extracellular adenosine deaminase or taken up by adenosine transporters (52), hence the adenosine concentration reaching astrocytic A_2A receptors may be much lower than the adenosine concentration in the perfusion saline. In summary, our results indicate that glomerular astrocytes express A_2A receptors that stimulate adenylyl cyclase and increase the cAMP concentration. Astrocytes in the olfactory bulb have been shown to affect neuronal performance and mediate neurovascular coupling in the olfactory bulb, mostly triggered by Ca²⁺ signaling (53–58). Which role astrocytic cAMP signaling plays in the olfactory bulb neuronal network still needs to be shown.

In the developing mouse olfactory bulb, it has been shown that axons of ORNs release ATP that triggers Ca²⁺ signaling in two types of glial cells, namely astrocytes and olfactory ensheathing cells, by activation of P2Y₁ receptors (39, 59, 60). In addition, in the glomerular layer ATP is degraded to adenosine that stimulates Ca²⁺ signals in developing astrocytes via A_2A receptors (39). We now studied the effect of ATP release on astrocytes in olfactory bulbs of adult mice. Using the fluorescent ATP sensor GRAB_ATP we were able to visualize extracellular ATP and to prove that also in adult mice, ORNs release ATP at their terminals in the glomerular layer. The peak of the extracellular ATP increase was approximately 10 µM which was in the range of the concentrations of ATP and adenosine used in the present study for bath application. Responses in cAMP evoked by both bath-applied ATP and synaptically released ATP were entirely blocked by ZM241385 in the majority of astrocytes, indicating that synaptically released ATP is degraded to adenosine that stimulates A_2A receptors, while other adenosine receptors or P2Y receptors play only a minor role in purinergic cAMP signaling in olfactory bulb astrocytes.

In neuroinflammation, astrocytes undergo a severe morphological transformation, hallmarked by proliferation, growth of cell processes and up-regulation of the expression of intermediate filaments such as glial fibrillary acidic protein (GFAP) and vimentin (6, 61–63). In addition, in a neuroinflammatory and neurodegenerative environment, purinergic Ca²⁺ signaling can be enhanced (15, 64, 65). We checked whether also purinergic cAMP signaling might be affected by neuroinflammation in the olfactory bulb. We studied neuroinflammation in EAE, a model for MS, since it has been shown that astrocytes are involved in pathogenesis and progression of the disease in the spinal cord and brain of both MS and EAE (8, 10). For example, inhibition of A_2A receptors in EAE ameliorated neurological and behavioral deficits in diseased mice (66, 67). We did not find any differences between A_2A-mediated cAMP signals in olfactory bulb astrocytes of healthy mice compared to mice suffering from EAE, indicating that adenosinergic cAMP signaling in astrocytes is not up- or down-regulated in EAE in this brain region. A_2A receptors on other cell types or pathways downstream of cAMP in olfactory bulb astrocytes are therefore more likely contributing to disease severity in EAE. We can also not exclude an involvement of astrocytic A_2A-dependent cAMP signaling in other brain regions or in the spinal cord in EAE. Furthermore, other components of the purinergic transmitter system could be affected by EAE and differences in the amount of extracellular adenosine and in the expression levels of ATP-degrading enzymes could also account for an impact on purinergic signaling in EAE but have not been addressed in the present study (68). Several studies have shown an important role of A_2A receptors in EAE. Of note, A_2A receptors can have both beneficial and exacerbating effects on the severity of EAE. A_2A stimulation inhibits lymphocyte proliferation and release of proinflammatory cytokines that promote MS progression, while inhibition of A_2A in EAE mice attenuated the progression of EAE and A_2A-deficient mice developed a more severe acute phase of EAE (66, 67, 69, 70). In other neurological diseases and disease models, an up-regulation of A_2A in astrocytes and contribution of A_2A to the disease progression has been demonstrated (46–48). In a mouse model of Alzheimer’s disease (AD), up-regulation of A_2A receptors in hippocampal astrocytes was shown which can lead to dysregulation of gene expression and connexin43-dependent gliotransmitter release in astrocytes (71–73). In addition, in an animal model of chronic cerebral hypoperfusion inducing neuroinflammation, activation of astrocytic A_2A receptors reduced the pro-inflammatory signal transducer and activator of transcription 3 (STAT3)/Chitinase 3-like-1 (YKL-40) pathway, resulting in decreased release of pro-inflammatory cytokines (74). These examples imply that A_2A receptors in astrocytes play an important role in neurodegeneration and neuroinflammation, however, the role of astrocytic A_2A receptors in MS and EAE needs still to be elucidated. Our results indicate that A_2A receptors in olfactory bulb astrocytes significantly contribute to cAMP dynamics but do not provide evidence for dysregulation of A_2A receptor expression and function in EAE.