Calcium signals in the nucleus accumbens: Activation of astrocytes by ATP and succinate
© Molnár et al; licensee BioMed Central Ltd. 2011
Received: 28 June 2011
Accepted: 3 October 2011
Published: 3 October 2011
Skip to main content
© Molnár et al; licensee BioMed Central Ltd. 2011
Received: 28 June 2011
Accepted: 3 October 2011
Published: 3 October 2011
Accumulating evidence suggests that glial signalling is activated by different brain functions. However, knowledge regarding molecular mechanisms of activation or their relation to neuronal activity is limited. The purpose of the present study is to identify the characteristics of ATP-evoked glial signalling in the brain reward area, the nucleus accumbens (NAc), and thereby to explore the action of citric acid cycle intermediate succinate (SUC).
We described the burst-like propagation of Ca2+ transients evoked by ATP in acute NAc slices from rat brain. Co-localization of the ATP-evoked Ca2+ signalling with immunoreactivities of the astroglia-specific gap junction forming channel protein connexin43 (Cx43) and the glial fibrillary acidic protein (GFAP) indicated that the responsive cells were a subpopulation of Cx43 and GFAP immunoreactive astrocytes. The ATP-evoked Ca2+ transients were present under the blockade of neuronal activity, but were inhibited by Ca2+ store depletion and antagonism of the G protein coupled purinergic P2Y1 receptor subtype-specific antagonist MRS2179. Similarly, Ca2+ transients evoked by the P2Y1 receptor subtype-specific agonist 2-(Methylthio)adenosine 5'-diphosphate were also blocked by MRS2179. These characteristics implied that intercellular Ca2+ signalling originated from the release of Ca2+ from internal stores, triggered by the activation of P2Y1 receptors. Inhibition by the gap junction blockers carbenoxolone and flufenamic acid and by an antibody raised against the gating-associated segment of Cx43 suggested that intercellular Ca2+ signalling proceeded through gap junctions. We demonstrated for the first time that extracellular SUC also evoked Ca2+ transients (EC50 = 50-60 μM) in about 15% of the ATP-responsive NAc astrocytes. By contrast to glial cells, electrophysiologically identified NAc neurons surrounded by ATP-responsive astrocytes were not activated simultaneously.
We concluded, therefore, that ATP- and SUC-sensitive Ca2+ transients appear to represent a signalling layer independent of NAc neurons. This previously unrecognised glial action of SUC, a major cellular energy metabolite, may play a role in linking metabolism to Ca2+ signalling in astrocytic networks under physiological and pathological conditions such as exercise and metabolic diseases.
In astrocytes of the brain reward area, the nucleus accumbens (NAc; ), γ-hydroxybutyric acid (GHB; ) evoked intracellular store-reliant Ca2+ transients, independently of neuronal activity . Previously, we also showed that binding sites for GHB are shared with citric acid cycle intermediate succinic acid (SUC) and the gap-junction blocker carbenoxolone hemisuccinate (CBX), as disclosed in NAc membrane homogenates isolated from rat and human brain tissues [4–6]. These findings raised the possibility that SUC, similarly to GHB may also evoke Ca2+ transients in NAc astrocytes. Further, it is conceivable that the rather specific sensitivity of the SUC/GHB target site to CBX might be a sign of its functional association with connexin channels. In order to study the effect and functional significance of SUC on the Ca2+ homeostasis of NAc astrocytes, we considered that the Ca2+ bursting activity was found ATP-responsive in vivo, i.e. in Bergmann glia networks activated by the motor behaviour of the awaken animal . Therefore, we sought to characterise first the ATP-responsive Ca2+ signalling amongst the astrocytes of the NAc.
ATP is known to evoke Ca2+ bursts by activation of purinergic G-protein-coupled receptors (GPCRs) in vitro [8–10] as well as in vivo [7, 11–13]. Different in vitro paradigms, including locally administered ATP stimuli (100 μM) were found effective to evoke Ca2+ transients [14–18]. In the present study, we investigated if locally ejected ATP (100 μM) could evoke Ca2+ bursting in NAc astrocytes. Measurements were performed by combined application of confocal Ca2+ imaging, immunohistochemistry and electrophysiology in acute NAc tissue slices prepared from the rat brain. Astrocytes were identified by co-localization of astrocyte-specific antibodies raised against the astroglial gap-junction protein connexin 43 (Cx43) and the glial fibrillary acidic protein (GFAP). Then, ATP-evoked Ca2+ bursts have been characterised by using of various drugs and agents, including gap-junction inhibitors (CBX, flufenamic acid: FFA), an antibody raised against the gating peptide segment of Cx43, purinergic P2 receptor agents such as the broad-spectrum P2X and P2Y receptor antagonist suramin (SUR), P2Y1 subtype-specific agonist 2-(Methylthio)adenosine 5'-diphosphate (2-Me-S-ADP) and antagonist MRS2179, the Na+ channel blocker tetrodotoxin (TTX) and the Ca2+ store depleting cyclopiazonic acid (CPA). Moreover, we also demonstrate for the first time the existence of SUC-responsive Ca2+ transients that overlay in a sub-population of NAc astrocytes.
In selecting the NAc region of interest, we first considered area-dependent distribution of Cx43 protein and its co-localization with GFAP. Next, we asked if Cx43-positive NAc astrocytes were responded to local administration of ATP by Ca2+ transients. Subsequently, the hypothesis that SUC may also activate Ca2+ transients playing part in the ATP-responsive Ca2+ signals was tested. Finally, responsiveness of NAc neurons to ATP was explored.
In addition to the number of ATP-responsive cells, we also analysed the magnitude of the response (Figure 4B and Methods section). Maximal dF/F0 values for individual cells in the absence and presence of the test compounds were determined. The calculated average change in (dF/F0)max values were compared to the Δ(dF/F0)max values obtained for two consecutive ATP applications both in the absence of test compounds (control), and were given as percentage of the first ATP application. This analysis concluded to results consistent with the cell number-based evaluation (Figure 4C). The obtained linear correlation between the cell number- and fluorescence intensity-based data (R = 0.94 ± 0.09, Figure 4C) indicates that the Ca2+ transients are produced by an on/off trigger, corroborating Ca2+ store-operated mechanisms.
We tested the hypothesis whether, in addition to ATP, SUC also activates astroglial Ca2+ signals. This was explored in measurements of the effect of SUC alone and in combination with the ATP puff-evoked Ca2+ transients characterized before. To this end, we applied the drug testing protocol described before (cf. previous and Methods sections) enabling the detection of the SUC-responsive Ca2+ transients before the second ATP application. This way, the observation of the SUC-responsive cells throughout the second ATP made possible comparison of the spatiotemporal characteristics of the SUC- and ATP puff-evoked Ca2+ transients.
The change in the number of cells showing SUC-evoked Ca2+ transients could be characterised by an EC50 value of 50-60 μM (Figure 5D, N = 35 from 16 rats). The same measurements on Ca2+ transients, however, provided less Δ(dF/F0)max data, the possible reason why the fluorescence intensity-based evaluation of the effective concentration for SUC was less conclusive (Figure 5D). These findings were at variance with the results of the inhibition of ATP puff-evoked Ca2+ transients (cf. Figure 4), suggesting that in addition to the oscillatory Ca2+ dynamics, SUC may also alter fluorescence intensity through some other mechanisms making these data more vulnerable to the threshold criterion applied (cf. Methods section).
Next, we asked if Ca2+ transients evoked by SUC application were occurred in astrocytes. It has already been demonstrated that the protocol used here for Fluo-4 AM loading preferentially labels astrocytes [3, 24]. For further identification of cells displaying SUC-responsive Ca2+ transients we applied double immunostaining for the glial marker proteins GFAP (green, Figure 5E left) and Cx43 (red, Figure 5E middle). Colocalization of Cx43 and GFAP proteins (yellow, Figure 5E right) identified the SUC-responsive cells as astrocytes.
Although the identified reporter neurons were flooded by the long ATP puff (cf. Figure 2B right) and surrounded by ATP-responsive NAc astrocytes (Figure 6B left images), astroglial burst-like Ca2+ signalling caused no significant changes in cytosolic Ca2+ (Figure 6A right images; N = 15, 11 rats) of these neurons. We measured neuronal Ca2+ (Figure 6A right images and green traces) and spontaneous post-synaptic currents (i.e., amplitude and frequency) under control, ATP and distinct washout periods (Figure 6C) and found no significant changes in any of these parameters (Figure 6C; N = 18 for 10 rats).
Similarly to ATP application (Figure 6C), postsynaptic currents of neurons within the astroglial Ca2+ signalling network did not exhibit significant changes in control frequency and amplitude (16.3 ± 1.5 pA; 3.0 ± 0.9 Hz) under 50 μM SUC (17.6 ± 2.9 pA, p > 0.05; 2.7 ± 1.0 Hz, p > 0.05), 50 μM SUC+ATP (18.1 ± 1.8 pA, p > 0.05; 2.6 ± 0.8 Hz, p > 0.05) and washout (16.4 ± 2.8 pA, p > 0.05; 3.0 ± 1.6 Hz, p > 0.05) periods (N = 3 from 3 rats, data not shown). These findings supported our conclusion that NAc neurons did not participate in astroglial Ca2+ signalling elicited by ATP and/or SUC.
We described that ATP evokes burst-like propagation of Ca2+ signals amongst astrocytes in the NAc. We disclosed for the first time that SUC does also induce Ca2+ transients. We will discuss the mechanisms of Ca2+ signalling amongst astrocytes in the NAc and their possible relation to neuronal activity. The potential importance of metabolites related to brain energy state regulating Ca2+ signals will also be speculated.
The ATP-evoked burst-like Ca2+ signal apparently propagates through astrocytes coupled by gap-junctions [26–28]. More extensive Ca2+ signalling in response to the localized ATP stimuli in the NAc as compared to nearby brain areas may possibly be due to the more abundant presence of Cx43 protein and gap-junctions thereupon. Our observations also suggest that Cx43 containing gap-junctions participate in burst-like Ca2+ signalling and were expressed only in astrocytes . Note, however, that co-localization of Cx43 and GFAP may also be associated with reactive astrocytes  or radial glia .
Astroglial burst-like Ca2+ signals are known to propagate via two mechanisms [26, 27]: transfer of cytosolic inositol (1,4,5)-trisphosphate (IP3) directly from cell to cell through gap junction channels and release of ATP onto extracellular purinergic receptors. The CBX-sensitive Ca2+ signal is coupled with the ATP-triggered enhancement of IP3 and its diffusion through gap-junction channels between neighbouring astrocytes . Another possible way of CBX- and FFA-sensitive propagation of the Ca2+ signal involves sequential activation of cation channels by ATP that is released through gap-junction hemichannels into the extracellular space [33, 34]. These CBX-sensitive Ca2+ waves could be blocked by SUR, a polysulfonated napthylurea used as a broad spectrum antagonist at purine-activated P2X channels and G-protein-coupled P2Y receptors [35–37]. In our experiments SUR (100 μM, 1 mM) did not block the ATP-activated Ca2+ bursts significantly, although it is known to antagonize P2X receptors in the 1-100 μM range . Nevertheless, our results corroborate previous data suggesting that SUR, the widely used antagonist of the effects of ATP at P2 purinoceptors, also induces calcium inducible calcium release (CICR) and increases open channel probability of ryanodine receptor channels [38–42]. These findings can explain the apparent ineffectiveness of SUR as a P2 antagonist by the opposing SUR-induced Ca2+ enhancements through the direct activation of intracellular Ca2+ stores.
In contrast, the blockade of both ATP- and the P2Y1-selective 2-Me-S-ADP-evoked Ca2+ transients by the P2Y1 selective antagonist MRS2179 indicated that the astroglial burst-like Ca2+ signal may possibly be evoked by activation of the P2Y1 receptor subtype in the NAc. The observed variability in MRS2179 effects may be related to developmental changes of P2Y1 receptors  resulting through receptor/cell heterogeneity within the shell/core regions of NAc slices prepared from 10-14 day old rats. In addition, endogenous SUC may have a co-stimulatory role in ATP-induced astroglial Ca2+ signalling and, by (partially) antagonizing the effects of the P2Y1 inhibitor MRS2179 as observed with human platelets , may contribute to the inter-slice/rat variability observed. It may be significant in this regard, that co-application of SUC and ATP apparently results in more durable Ca2+ transients, conjecturing coincidence detection.
As discussed before, Ca2+ store mobilization, underlying propagating astroglial burst-like Ca2+ transients can occur via P2Y1 receptor activation in the NAc. Both the inhibitor profile (blockade by CBX, FFA and Cx43 antibody) and dynamics of ATP-evoked Ca2+ waves suggest cell-to-cell signalling through the gap-junction-coupled astrocytes in the NAc. Low FFA efficacy may be due to the possible enhancement of intracellular Ca2+ by FFA acting at type 6 transient receptor potential canonical channel (TRPC6) [45, 46].
As outlined above, the alternative ATP-driven way of signal propagation [26, 27, 32–34] may not be excluded purely on the basis of the apparent lack of SUR blockade of ATP-evoked Ca2+ signals. Binding of CBX to SUC receptor [4–6], however, could be accounted for some blockade of ATP-driven Ca2+ signal propagation as detailed below.
SUC binding sites have previously been disclosed in rat forebrain and human NAc membrane homogenates [4, 6], however, their functions have not been assigned yet. Identity of SUC-sensitive GHB  and GHB-sensitive SUC  binding sites raised the possibility to relate SUC and different GHB actions [47–50]. Importantly, the binding site of GHB/SUC interacts with CBX [4–6], a blocker of gap-junctions that are major players of astroglial Ca2+ signalling . We have also shown previously that GHB activates intracellular store-dependent Ca2+ transients in NAc astrocytes . Here we report, that SUC induced repetitive Ca2+ transients occur in a subpopulation of Cx43+ and GFAP+ cells responding to ATP with burst-like Ca2+ signals involving gap-junctions. SUC may bind to a purinergic G protein coupled receptor (GPCR) subtype or some other SUC-responsive membrane receptor such as SUCNR1 (GPCR91). Interestingly, the SUCNR1 gene is located on chromosome 3 as part of a cluster of seven GPCRs in close vicinity to the genes for P2Y1 . Also, direct effect of SUC on gap-junctions cannot be excluded. The SUCNR1 has recently been shown to regulate cellular functions implicated in renal blood pressure regulation [44, 53] and lipolysis of white adipose tissue . In addition, the SUCNR1 may also have a role for immunity, hyperglycemia, retinal neovascularization, ischemic liver injury and hematopoiesis as reviewed recently . The presence of SUCNR1 has been demonstrated in human platelets and megakaryocytes , in various cells of distal nephron  and in cardiomyocytes .
The concentration of SUC in plasma [57–59] increases with exercise , metabolic acidosis , hypertension and metabolic diseases  from 5 μM up to 125 μM. These data suggest that the tissue concentration of SUC can be high enough to induce astroglial Ca2+ transients characterised by the EC50 value for SUC-responsive cells within the range of 50-60 μM. Using arachidonic acid, ADP and SUC as platelet agonists, aggregation in response to SUC alone was highly variable with only 29% of donors showing a (mostly delayed) platelet response . In contrast, SUC reproducibly and concentration-dependently enhanced platelet aggregation in response to low concentrations of exogenous ADP . Assuming the presence of SUCNR1 and P2Y1 in astrocyte membrane, we conjecture co-stimulatory roles played by endogenous SUC and ATP within the brain. Such a coincidence detection performed by astrocytes could explain why astrocytic Ca2+ transients may be dramatically affected by pathological conditions associated with intense neuronal firing .
It has been reported, that mechanically evoked astrocytic Ca2+ waves mediated by the release of ATP and the activation of P2 receptors in hippocampal cultures down-regulate excitatory glutamatergic synaptic transmission in an ATP-dependent manner . Using astrocyte-specific inducible transgenic mice (dnSNARE mice), it has been evidenced that by the release of ATP, which accumulates as adenosine, astrocytes tonically suppress synaptic transmission in acutely isolated hippocampal slices . Causal linkage between reduced activation of adenosine receptors and surface expression of NMDA receptors has also been evidenced [ and references cited there]. We assumed therefore, that by using ATP puff, which may also reduce activity of NAc neurons, we could isolate 'autonomous' astrocytic Ca2+ signalling. Indeed, the loading pattern of the Ca2+ indicator Fluo-4 AM [3, 24] and the presence of Cx43 protein in cells responding to SUC/ATP by Ca2+ transients identified them as astrocytes.
Propagation of the Ca2+ signal in ellipsoidal and/or radial waveform at a speed of approximately 10 μm/s in NAc astrocytes is comparable to the time evolution of Bergmann glia Ca2+ transients, as previously recorded from fixed location in vitro [9, 10] and in vivo [7, 13]. The similarity suggests both phenomena involve ATP-triggered release of Ca2+ from intracellular stores, as found. The type of glial Ca2+ signalling classified as bursts persisted after application of TTX in the cerebellar Bergmann glia in vivo  and NAc astrocytes in vitro (this work). Unchanged postsynaptic currents and an invariable cytosolic Ca2+ level of neurons within the astrocyte network that exhibited concerted Ca2+ bursts support our conclusion that the ATP evokes astroglial burst-like Ca2+ signals independently of neuronal activity in the NAc. In this respect, ATP/SUC-evoked astrocytic Ca2+ transients differ from the ones dependent on the metabotropic Glu recptor subtype 5, leading to the activation of the NR2B subunit containing N-Me-D-Asp receptors (NMDARs) of medium spiny neurones in the NAc or the striatum . Our conclusion upholds previous findings by us  and others [, reviewed in ], suggesting the existence of "autonomous" activation of astroglial Ca2+ signals. Remarkably, glial Ca2+ bursting activity observed in the cerebellum of awake, behaving animal is also found to be independent of neuronal activity .
Findings obtained with the NAc model of glia activation described in this work do not exclude, however, that Ca2+ signalling involving ATP/SUC sensing cells might be a unique feature of the NAc. We may also assume such a signalling route in other brain areas exhibiting extensive Cx43 expression. The way of Ca2+ signalling is probably characteristic of a given brain region . Ca2+ bursts have been shown in cultured astrocytes , white matter tract , cerebellum  and cortical brain slices [28, 68, 69]. Ca2+ bursts in the neocortex depend on gap-junction coupling of astrocytes and are not influenced by neuronal activity while in the corpus callosum Ca2+ signalling requires ATP-release mechanism but not gap-junction expression . These pathways are not independent from each other and most probably linked to the cell types participated in Ca2+ signalling . In the retina, astrocytic Ca2+ wave is dependent on gap-junctions, while the wave propagation between astrocytes and Müller cells requires ATP-release mechanism . The expression of the major gap-junction protein of astrocytes Cx43 influences purinergic receptor expression in the spinal cord astrocytes . These examples rather support the view that glial Ca2+ bursts may represent a signalling layer independent of neurons.
The issue of independence as opposed to interdependence of glial and neuronal activities is related to the question of astrocytic release of signalling molecules that may also modulate synaptic transmission, a process called gliotransmission. As outlined before, there are several indications, that astrocytes do affect synaptic activity through the release of ATP, Glu, D-Ser, tumour necrosis factor alpha [[63–66, 72] and references cited] or uptake of Glu [73, 74] and GABA . It is to note, that the physiological relevance of gliotransmission in long-term potentiation (LTP) was called into question [[75, 76], see also  for additional references]. In contrast, Ca2+-dependent release of D-Ser from an astrocyte affecting NMDAR-dependent plasticity was demonstrated by clamping internal Ca2+ in individual astrocytes of the CA1 area of the hippocampus .
Diverse cellular pathways of astroglial Ca2+ signals are differently pronounced in various brain regions and they can also co-occur. Besides, the initial stimuli are likely a determinant of the relevant physiological function being involved.
The findings of the present study describe for the first time the Ca2+ signals evoked in astrocytes of the brain reward area, the nucleus accumbens, by the citric acid cycle intermediate SUC. We also identify these SUC-responsive Ca2+ signals in ATP-evoked burst-like intercellular Ca2+ signalling. We provide evidence that ATP- and SUC-responsive glial Ca2+ signals are independent on neuronal activity therefore apparently represent a signalling layer independent of NAc neurons.
ATP-induced calcium responses have previously been shown in cortical and hippocampal astrocytes whereas succinate-induced calcium signals had not been described in astrocytes. This previously unrecognized glial action of the major cellular energy metabolite SUC may represent a link between brain energy states and Ca2+ signalling in astrocytic networks.
Animal care and preparations were in accordance with the Helsinki declaration, the European Council Directive of 24 November 1986 (86/609/EEC), the Hungarian Animal Act 1998, and were associated local guidelines, as approved by the Institutional Animal Care and Use Committee (approval ID: MÁB 1.51.4.).
Slice preparing buffer contained in mM: 250 sucrose, 2 KCl, 1.25 KH2PO4, 10 MgSO4, 2 CaCl2, 16 NaH2CO3 and 10 glucose. Artificial cerebrospinal fluid (ACSF) contained in mM: 129 NaCl, 2 KCl, 1.25 KH2PO4, 1 MgSO4, 2 CaCl2, 16 NaHCO3 and 10 glucose.
The following drugs were applied via the ACSF perfusion: SUC, GHB, CBX, another gap-junction blocker FFA (purchased from Sigma-Aldrich, Budapest, Hungary); P2X/P2Y receptor antagonist SUR, the P2Y1 receptor antagonist MRS2179, the endoplasmic Ca2+-ATPase inhibitor CPA, and TTX (from Tocris, Bristol, UK). Agonists ATP and 2-Me-S-ADP were locally applied through a glass micropipette (Sigma-Aldrich, Budapest, Hungary).
Fluorescence indicators Fluo-4 AM, Fluo-4 tetrapotassium salt, OGB-1 hexapotassium salt, propidium iodide (PI), SR101 and pluronic acid were purchased from Molecular Probes (Eugene, USA). Stock solutions of ester fluorescence indicators prepared in DMSO were diluted to 0.1% DMSO in the staining solution.
Coronal slices from the forebrain through the NAc and the caudate putamen (CP) were prepared for the imaging experiments. Young, 10-14 day-old male Wistar rats were decapitated and the brains were quickly removed. The forebrains were serially cut into 300 μm thick coronal sections (Vibratome, Technical Products International Inc., St. Louis, MO, USA). The slices were collected in ice-cold preparation buffer and incubated for one hour under humidified gas-mixture carbogen (5% CO2 + 95% O2) atmosphere in an interface-type holding chamber containing warmed (35°C) ACSF. After preincubation in 2% pluronic acid containing ACSF for 2 minutes, slices were further incubated with 5 μM Fluo-4 AM in ACSF at 35°C in the dark under humidified carbogen (5% CO2 + 95% O2) atmosphere, for one hour . In order to monitor cell death, several slices were exposed to double dye-loading protocol, performed by adding 7.5 μM PI (excitation: 534 nm, emission: 570-600 nm) to the Fluo-4 AM containing ACSF. In order to allow for cleavage of the AM ester group of Fluo-4, slices were transferred to dye-free ACSF at least 30 minutes before the start of the experiment .
We used Fluo-4 AM for quantifying astroglial Ca2+ concentrations in the 100 nM to 1 microM range . Fluo-4 offers high fluorescence emission due to its greater absorption near 488 nm, and a large dynamic range for reporting [Ca2+] around a Kd(Ca2+) of 345 nM . Fluorescence recordings of changes in the intracellular Ca2+ ion level in cells loaded with Fluo-4 AM were performed, with an upright epifluorescent microscope (Olympus BX61WI, Olympus, Budapest, Hungary) equipped with the FluoView300 confocal laser-scanning system (Olympus, Budapest, Hungary) using 20× (0.5 numerical aperture) or 60 (0.9 numerical aperture) water immersion objectives (for details, see ). Image acquisition rate was controlled by a computer running Tiempo software for FluoView300 (Olympus, Budapest, Hungary).
Freshly isolated slices were transferred to the submerge-type recording chamber mounted on the stage of the microscope and were superfused with carbogenated (5% CO2 + 95% O2) ACSF (3 ml/min, room temperature). Serial scanning of slices were made at 488 nm excitation wavelengths and emitted green fluorescence was collected through a 510-530 nm bandpass filter. Ca2+ transients were initiated by application of SUC in the perfusion or by pressure-ejection of 100 μM ATP in ACSF through a glass micropipette (5-10 μm diameter) on the slice surface. Fluorescence intensity changes within a 355 × 355 μm field containing approximately 100 Fluo-4 AM loaded cells around the area of the ATP puff, were followed over a 10-minute interval (2 s/image).
Images taken during the ATP application were summed and the average of images taken during the control period was subtracted as a background. The Fluo-4 loaded cells were identified by using custom written algorithms in Matlab under strict visual control. The intensity changes in individual cells were calculated as dF/F0. To identify ATP responsive cells, the standard deviances of the control periods of the dF/F0 traces (SDcontrol) were calculated and the traces with (dF/F0)max,ATP > 5*SDcontrol were selected. The number of cells with fluorescence intensity changes above the threshold was counted to determine the number of responsive cells (N) and the average of their maximal dF/F0 values ((dF/F0)max) during the ATP application period was used to assess the magnitude of the response. The changes in N and (dF/F0)max values (ΔN and Δ(dF/F0)max, respectively) obtained in the presence of test compounds were normalized to the corresponding N and (dF/F0)max values obtained in the absence of the test compounds (Figure 7C). We observed that the number of ATP responsive cells and their maximal response were reduced in the second ATP application even if the test compounds were not present. Therefore to take this reduction into account, the calculated ΔN and Δ(dF/F0)max values for each condition were compared (ANOVA) to ΔN and Δ(dF/F0)max values obtained for two consecutive ATP applications both in the absence of test compounds (control in Figure 4).
Data presented are mean ± S.E.M., with N denoting the number of slices in a given experimental condition. Statistical analysis was performed using the non-parametric Mann-Whitney test with Bonferroni post hoc test (OriginLab Co., Northampton, UK) and p < 0.05 was considered statistically significant. If otherwise stated, the effects of different treatments were compared to the control.
Whole cell patch clamp recordings were performed both in the shell and the core areas of the NAc by using a MultiClamp 700A amplifier (Axon CNS, Molecular Devices, Sunnyvale, California, USA). Signals were low-pass filtered at 2 kHz and digitized at 10 kHz (Digidata1320A, Axon Instruments). Cells selected by their visual appearance were identified as neurons by the presence of voltage activated fast Na+ currents during application of a voltage ramp protocol -40 to +50 mV (pClamp8, Axon Instruments). 6-8 Mc pipettes pulled from borosilicate glass capillaries were filled with an intracellular solution (containing in mM: 135 K-Gluconate, 10 NaCl, 0.05 CaCl2, 2 adenosine-triphosphate Mg2+ salt, 10 4-(2-hydroxyethyl)piperazine-1-ethanesulfonic acid (HEPES); pH 7.3 (set with 1N KOH). In order to visualize and monitor cytosolic Ca2+ of patched neurons, 200 μM Fluo-4 tetrapotassium salt or 50 μM OGB-1 hexapotassium salt (excitation: 488 nm, emission: 510-530 nm for both dyes) was added to the intracellular solution. Cells (input resistance: 166 ± 13 MΩ) were clamped to -70 mV without corrections for liquid junction potential (-15 mV). At this potential inward spontaneous postsynaptic currents were recorded. If signs of seal deterioration or cell closure occurred, the recordings were discarded. The first 10 min of each recording were used for monitoring stabilization of the baseline. Then, a 1-minute duration control period, followed by ATP application (1 min), and a washout period (5-10 min) were recorded.
Spontaneous postsynaptic currents (PSCs) were analyzed using Mini Analysis software (Synaptosoft, Decatur, USA; http://www.synaptosoft.com) in 4 data segments (1 min, each) recorded prior to, at the ATP application, washout after ATP application and late-washout periods. The automatically detected events (threshold: 10 pA) were verified by visual inspection. Multiple comparisons were done with the non-parametric Mann-Whitney test (OriginLab Co., Northampton, UK). Statistical significance was defined as p < 0.05. The data are presented as mean ± S.E.M.
Adult, male Wistar rats (n = 3) (200-250 g body weight; Charles Rivers Laboratories, Hungary) were deeply anesthetized with a mixture containing 0.2 ml/300 g b.w. ketamine (100 mg/ml) and 0.2 ml/300 g b.w. xylazine (20 mg/ml), and perfused transcardially with 150 ml saline followed by 300 ml of ice-cold 4% paraformaldehyde in phosphate buffer, pH 7.0 (PB). Brains were removed and postfixed in the same fixative solution for 24 h, and transferred to PB containing 20% sucrose for 2 days. Serial coronal sections were cut at 50 μm on a sliding microtome (SM 2000R, Leica Microsystems, Nussloch, Germany) between +4.0 mm to 0 mm from the level of the bregma. Sections were collected in PB containing 0.1% sodium azide and stored at 4°C until further processing.
Free-floating sections were immunolabeled for Cx43 using an affinity-purified rabbit polyclonal antiserum raised against the carboxy terminal 362-382 peptide segment KPSSRASSRASSRPRPDDLEI of Cx43 (Abcam, Cambridge, UK, catalogue number: ab11370). Brain sections were pretreated in PB containing 0.5% Triton X-100 and 3% bovine serum albumin for 1 h. Then, they were incubated with a primary antibody against Cx43 (1:1250) in PB containing 0.5% Triton X-100 and 3% bovine serum albumin and 0.1% sodium azide for 48 h at room temperature. Sections were then incubated in biotin-conjugated donkey anti-rabbit secondary antibody at 1:1000 (Jackson Immunoresearch, West Grove, PA) for 2 h, followed by incubation in avidin-biotin-horseradish peroxidase complex (ABC) at 1:500 (Vectastain ABC Elite kit, Vector Laboratories, Burlingame, CA) for 2 h. Then, sections were treated with either 0.02% diaminobenzidine (Sigma) or fluorescein isothiocyanate (FITC)-tyramide (1:8000) with 0.003% H2O2 in Tris-HCl buffer (0.05 M, pH 8.2) for 6 min. After washing, the sections were mounted on positively charged slides (Superfrost Plus, Fisher Scientific) and coverslipped with either Cytoseal 60 (Stephens Scientific, Riverdale, NJ, USA) or in antifade medium (Prolong Antifade Kit, Molecular Probes).
Free-floating sections were first immunolabeled for Cx43 using FITC-tyramide amplification immunohistochemistry, as described above. Then, sections were incubated overnight in a mouse monoclonal anti-GFAP, a marker of astrocytes (1:250; catalogue number: sc-33673, Santa Cruz Biotechnology, Santa Cruz, CA). Subsequently, sections were incubated in Alexa 594 donkey anti-mouse secondary antibody (1:500; Molecular Probes) for 2 h, then mounted and coverslipped, as described above. Sections were examined using an Olympus BX60 light microscope also equipped with fluorescent epi-illumination. Images were captured at 2048 × 2048 pixel resolution with a SPOT Xplorer digital CCD camera (Diagnostic Instruments, Sterling Heights, MI) using 4-20× objectives. Confocal images were acquired with a Nikon Eclipse E800 confocal microscope equipped with a BioRad Radiance 2100 Laser Scanning System using 20-60× objectives (at an optical thickness of 1-3 μm). Contrast and sharpness of the images were adjusted using the "levels" and "sharpness" commands in Adobe Photoshop CS 8.0. Colours were adjusted so that Cx43 appeared red and GFAP was green. Full resolution was maintained until the photomicrographs were cropped and assembled for printing, at which point images were adjusted to a resolution of 300 dpi.
In order to identify the cell types involved in Ca2+ bursts, the brain slices were immunostained with antibodies for astrocyte marker proteins (Cx43 and GFAP). However, Fluo-4 signal could not be preserved through fixation with either 0.4% paraformaldehyde or 40 mg/ml EDAC (1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide, Sigma-Aldrich; Dawling and Deitmer, 2002). Therefore, co-localization between SUC/ATP-responsive cells and cell type markers could not be resolved. Hence we opted to perform the in situ immunostaining in non-fixed slices directly after the Ca2+ imaging protocol. Slices used previously to measure Ca2+ changes in response to SUC/ATP application were treated as follows: upon completion of the calcium-imaging experiments, each slice was kept in its original position in the recording chamber (using a ballast) and incubated with Cx43 (1:300) and GFAP (1:200) primary antibodies for 30 min, at room temperature. After 3 × 10 minute washing in ACSF, the slice was incubated with Chromeo 546 goat anti-rabbit (1:100 Abcam, Cambridge, UK, catalogue number: ab60317) secondary antibody and Alexa 488 donkey anti-mouse (1:100, Molecular Probes) secondary antibody in ACSF for 30 min, at room temperature. This was followed by 3 × 10 minute washing in ACSF. Serial Z-scans of Cx43- (excitation: 543 nm, emission: 570-600 nm) and GFAP- (excitation: 488 nm, emission: 510-530 nm) labelled slices were acquired between the slice surface and the maximal penetration depth of the antibodies (approximately 60-70 μm from the surface) through a 20× objective (1 μm/step). Since the fluorescence emission of both Fluo-4 and Alexa-488 dyes are collected in the 510-530 nm range, GFAP-specific staining was obtained by subtracting the Fluo-4 fluorescence from the GFAP immunolabelling signal, at each Z depth. Optical sections from identical depths of Fluo-4 and Cx43 images were merged along the Z axis. Single cells and glia filaments that showed double immunolabelling were recorded through a 60× objective and Z-scans were performed by alternating the excitation wavelengths between 543 nm (Cx43) and 488 m (GFAP) using Tiempo software for FluoView300, at each depth (0.1 μm/step). Images were processed using ImageJ 1.44  and Adobe Photoshop CS 8.0 image analysis software.
The authors thank Erzsébet Kútiné Fekete and Ildikó Pál (Department of Neurochemistry, Institute of Biomolecular Chemistry, Chemical Research Center, Hungarian Academy of Sciences) for excellent technical assistance and interpretation of data, respectively. Authors TM, GNY, PB, LH, ZSE and JK were supported by grants 1/A/005/2004 MediChem2, GVOP-3.2.1.-2004-04-0210/3.0 Transporter Explorer AKF-050068, TECH-09-AI-2009-0117 NanoSen9 and CRC-HAS-2009-Nanotransport. Authors AD and MP were supported by the grant TECH-09-AI-2009-0117 NanoSen9.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.