(a) Slice and cell preparation

Mice (8–12 weeks) were anaesthetized by halothane and then decapitated, in accordance with UK legislation. Brains were removed rapidly after decapitation and placed into ice-cold physiological saline containing (mM): NaCl 130, KCl 3, CaCl2 0.5, MgCl2 2.5, NaH2PO4 1, NaHCO3 25, glucose 15, pH 7.4, gassed with 95% O2–5% CO2. Transverse slices (280 μm) were cut at 4°C and then placed in physiological saline containing (mM): NaCl 130, KCl 3, CaCl2 2.5, MgCl2 1, NaH2PO4 1, NaHCO3 22, glucose 15, pH 7.4, and kept for 1–4 h prior to cell isolation and recording.

Astrocytes and neurons were acutely isolated using the modified ‘vibrating ball’ technique [12,25,27,28]. The glass ball (200 µm diameter) was moved slowly some 10–50 µm above the slice surface, while vibrating at 100 Hz (lateral displacements 20–30 µm). This technique preserves the function of membrane proteins and therefore is devoid of many artefacts of enzymatic cell isolation and culturing procedures. The composition of the external solution for all isolated cell experiments was (mM): 135 NaCl; 2.7 KCl; 2.5 CaCl2; 1 MgCl2; 10 HEPES, 1 NaH2PO4, 15 glucose, pH adjusted with NaOH to 7.3. Astrocytes were identified by their morphology under differential interference contrast microscopy, EGFP fluorescence (astrocytes from dn-SNARE and GFAP-EGFP mice) or staining with sulforhodamine 101 (astrocytes from WT mice). After recordings, identification of astrocytes was confirmed by functional properties (high potassium conductance, low input resistance and strong activity of glutamate transporters) as described previously [25,27,28].

(b) Electrophysiological recordings

Whole-cell voltage-clamp recordings from neocortical neurons and astrocytes were made with patch pipettes (4–5 MΩ for neurons and 6–8 MΩ for astrocytes) filled with intracellular solution (in mM): 110 KCl, 10 NaCl, 10 HEPES, 5 MgATP, 10 EGTA, 1 CaCl2, pH 7.35. Currents were monitored using an AxoPatch200B patch-clamp amplifier (Axon Instruments, USA) filtered at 2 kHz and digitized at 4 kHz. Experiments were controlled by PCI-6229 data acquisition board (NI, USA) and winfluor software (Strathclyde University, UK); data were analysed by self-designed software. Liquid junction potentials were compensated with the patch-clamp amplifier. Series and input resistances were, respectively, 5–7 MΩ and 500–1100 MΩ in neurons and 8–12 MΩ and 50–150 MΩ in astrocytes; both series and input resistance varied by less than 20% in the cells accepted for analysis. For activation of synaptic inputs, axons originating from layer IV–VI neurons were stimulated with a bipolar coaxial electrode (WPI, USA) placed in the layer V close to the layer IV border, approximately opposite the site of recording; stimulus duration was 300 µs. The stimulus magnitude was set three to four times higher than minimal stimulus adjusted to activate the single-axon response in the layer II pyramidal neurons as previously described [27–29]. In order to trigger synaptically driven astroglial Ca2+-transients, the single episode of theta-burst stimulation (TBS; high-frequency stimulation, HFS) was delivered; an HFS episode consisted of five pulses of 100 Hz stimulation, repeated 10 times with 200 ms intervals (total 50 pulses per episode). For induction of long-term plasticity, two or five such HFS episodes were used.

(c) Multi-photon fluorescent Ca2+-imaging in astrocytes

To monitor the cytoplasmic free Ca2+ concentraton ([Ca2+]i) in situ, astrocytes of neocortical slices were loaded by 30 min incubation with 1 µM of Rhod-2AM or Oregon Green BAPTA-2 at 33°C. Two-photon imaging of neurons and astrocytes was performed using a Zeiss LSM-7MP multi-photon microscope coupled to a SpectraPhysics MaiTai pulsing laser; experiments were controlled by ZEN LSM software (Carl Zeiss, Germany). Images were further analysed off-line using ZEN LSM (Carl Zeiss) and ImageJ (NIH) software. The [Ca2+]i levels were expressed as ΔF/F ratio averaged over region of interest (ROI). For analysis of spontaneous Ca2+-transients in astrocytes, three ROIs located at branches and one ROI located at the soma were chosen. Overall Ca2+ response to agonists of eCB agonist or synaptic stimulation was quantified using ROIs covering the whole cell image.

(d) Measurement of extracellular concentration of ATP and d-serine in the brain tissue

The concentration of ATP within cortical slices was measured using microelectrode biosensors obtained from Sarissa Biomedical Ltd (Coventry, UK). A detailed description of the properties of biosensors and the recording procedure has been published previously [30]. Briefly, biosensors consisted of ATP or d-serine metabolizing enzymes immobilized within a matrix on thin (25–50 µM) Pt/Ir wire. This allowed insertion of the sensors into the cortical slice and minimized the influence of a layer of dead surface tissue. Concentrations of ATP and d-serine were measured simultaneously. The concentration of ATP or d-serine has been calculated from difference in the signals of two sensors: a screened ATP- or d-serine sensor and screened null-sensor, possessing the matrix but no enzymes. This allowed compensation for release of any non-specific electro-active interferents. Biosensors show a linear response to increasing concentration of ATP and d-serine and have a rise time of less than 10 s [30]. Biosensors were calibrated with known concentrations (10 μM) of ATP and d-serine before the slice was present in the perfusion chamber and after the slice had been removed. This allowed compensation of any reduction in sensitivity during the experiment. Biosensor signals were acquired at 1 kHz with a 1400 CED interface and analysed using spike 6.1 software (Cambridge Electronics Design, Cambridge, UK).

(b) eCBs trigger vesicular release of ATP from acutely isolated astrocytes

To verify that CB1 receptors trigger a vesicular release of gliotransmitters, in the next line of experiments we used a ‘sniffer’-cell approach to detect a release of ATP from acutely isolated astrocytes. In our previous work, we demonstrated that neocortical neurons express functional P2X receptors [25,29] and these receptors can be activated by ATP released from astrocytes [29]. Thus, acutely dissociated cortical neurons could serve as a native sensor for ATP. We used a technique of non-enzymatic vibro-dissociation which allows retention of functional membrane receptors at the cell surface [25,28,29], including P2X receptors. Another benefit of this technique is the ability to retain a fraction of functional synapses as well [12], which can be verified by staining with FM1-43 and the presence of miniature spontaneous synaptic currents (figure 2a). Furthermore, adjustment of the parameters of vibro-dissociation allowed us to dissociate neurons with a few astrocytes attached (figure 2b–d), thereby retaining a certain proportion of intimate contact between thin astrocytic processes and neuronal membrane. Such a neuron–astrocyte ‘bundle’ could work as a good model of the glia–neuron interaction unit, enabling efficient activation of astrocytes and direct monitoring of neuronal response. In comparison to bath application of extracellular agents to the whole slice, the ‘neuron–astrocyte bundle’ has the advantage of better control of drug application and a lack of side-effects of massive activation of the astrocyte network and poly-synaptic connections.

Figure 2.

(Overleaf.) Detection of ATP release from astrocytes using modified ‘sniffer-cell’ method. Parameters of the technique of non-enzymatic cell dissociation [12,25,28,29] were optimized to yield either isolated neurons (a) or neurons attached …

We recorded whole-cell currents in the acutely dissociated neocortical pyramidal neurons at a membrane potential of −80 mV in the presence of CNQX (50 µM), D-APV (30 µM) and picrotoxin (100 µM). Similarly to our previous experiments [25,29], we observed residual non-glutamatergic miniature spontaneous synaptic currents (figure 2). These non-glutamatergic excitatory spontaneous currents (mEPSCs) were completely abolished by application of specific P2X receptors antagonists PPADS (10 µM) and 5-BDBD (5 µM) in all five neurons tested (data not shown). Based on these data as well as our previous work [25,29], the pulsatile inward currents observed in cortical neurons in the presence of glutamatergic and GABAergic antagonists can be confidently attributed to the ATP receptors.

In order to activate CB1 receptor-mediated Ca2+-signalling in astrocytes, we applied 500 nM AEA. For a positive control, we used 5 µM TFLLR, an agonists of the PAR-1 receptor, which activate Ca2+ signalling selectively in astrocytes [25]. As a negative control, we used fully isolated neurons (figure 2a) that were devoid of any astrocytes. These neurons exhibited purinergic mEPSCs with average amplitude of 9.8 ± 2.2 pA and average decay time constant of 8.9 ± 1.7 ms (n = 7). Neither TFLLR nor AEA produced marked changes in the amplitude or frequency of spontaneous purinergic mEPSCs in all of the seven fully isolated neurons tested.

In contrast to fully isolated neurons, purinergic mEPSCs recorded in dissociated neurons that had WT astrocytes attached (figure 2b) exhibited bimodal amplitude distributions with peaks at 5.9 ± 1.4 pA and 9.9 ± 2.6 pA (n = 6). The distributions of mEPSC decay time in these neurons had peaks at 8.8 ± 1.1 ms and 15.1 ± 2.2 ms. Application of TFLLR caused a significant increase in the frequency of purinergic mEPSCs and decrease in their average amplitude (figure 2b,d) in all of six experiments. This effect was associated with selective increase in the number of events with smaller amplitude and slower decay kinetics (figure 2b, right column). Application of TFLLR did not cause marked increase in the number of purinergic currents in neurons attached to the dn-SNARE-expressing astrocytes (figure 2c,d). This observation closely agrees with our previous experiments carried out in the neocortical slices where we demonstrated that TFLLR-induced purinergic mEPSCs of smaller amplitude and slower kinetics originated from the vesicular release of ATP from astrocytes [25].

Similarly to TFLLR, application of AEA significantly increased the frequency of purinergic currents of slower kinetic and smaller amplitude (figure 2b,d) in the neurons connected to the WT but not dn-SNARE astrocytes. It has to be noted that neurons were perfused with extracellular solution containing 10 mM EGTA which should effectively inhibit putative retrograde release of neurotransmitters, as shown in Duguid et al. [12]. Thus, our data provide strong evidence that glial CB1 receptors are able to trigger vesicular release of gliotransmitters.

(c) CB1 receptors contribute to control of astroglial exocytosis in situ

The above data were obtained upon activation of astrocyte CB1 receptors by exogenous agonist, and it would be natural to ask whether endogenous eCBs could activate exocytosis from cortical astrocytes. We demonstrated previously that a short burst of HFS of intracortical afferents was able to elevate cytosolic Ca2+ level in the cortical astrocytes and trigger vesicular release of ATP. To explore a plausible involvement of CB1 receptors in this effect, we monitored purinergic spontaneous currents in the pyramidal neurons of neocortical slices at membrane potential of −80 mV in the presence of CNQX and picrotoxin (figure 3). Currents recorded in the neocortical neurons under these conditions were shown to be mediated solely by P2X receptors [25,29].

Figure 3.

(Overleaf.) Attenuation of Ca2+-signalling in cortical astrocytes inhibits the release of ATP. Ca2+-signalling was monitored in cortical astrocytes loaded with Rhod-2AM using multi-photon fluorescent microscopy simultaneously with whole-cell recording …

The short HFS train caused significant elevation of the frequency of purinergic mEPSCs in the pyramidal neurons of WT mice, this elevation being associated with increased number of mEPSCs of smaller amplitude and slower kinetics. We have shown previously that such mEPSCs originate from dn-SNARE-dependent release of ATP from astrocytes [25]. Thus, similarly to our present data obtained in the dissociated neurons (figure 2), appearance of slower purinergic mEPSCs can be used as a good read-out for glial exocytosis. We found that inhibition of CB1 receptors with AM251 considerably reduced the Ca2+ elevation triggered in the astrocytes by HFS train (figure 3a) and decreased the following burst of slow purinergic mEPSCs (figure 3b–d). This result indicates that eCB-mediated signals from neurons can activate exocytosis from astrocytes in physiological conditions. To estimate the putative impact of eCBs on glial exocytosis, we compared the modulatory effect produced by AM251 to the effect produced by antagonists of mGluR5 and NMDA receptors, reported to contribute to astrocytic Ca2+ signalling [27,28,31]. The effect of eCB antagonist was comparable to that of MPEP, an antagonist of mGluR5 [31], and UBP141, a selective antagonist of glial NMDA receptors [27,28]. This result suggests that eCB receptors can bring notable contribution to the Ca2+-dependent modulation of exocytosis (at least, of ATP) in the neocortical astrocytes (figure 3e).

The authors thank V. Parpura, J. Bains and P. Haydon for early discussion and helpful comments on the manuscript. S.R-N., O.P. and U.L. carried out the experiments and data analysis, S.R-N. and U.L. participated in the design of the study and helped to draft the manuscript, Y.P. conceived and coordinated the study, participated in the data analysis and drafted the manuscript. All authors gave final approval for publication.