Introduction Compelling evidence obtained by different groups during the last years indicate that astrocytes play important roles in synaptic function [1]–[4]. In addition to their well-known passive homeostatic control of synaptic function, astrocytes sense synaptic activity responding with Ca2+ elevations to synaptically released neurotransmitters and, in turn, release gliotransmitters that regulate synaptic transmission and plasticity [5]–[14]. This evidence has led to the establishment of the Tripartite Synapse concept, in which astrocytes actively exchange information with the neuronal synaptic elements, suggesting that astrocytes may be considered as integral elements of the synapses being directly involved in synaptic physiology [1]–[4]. While this evidence has been largely obtained in brain slices, recent in vivo studies that used transgenic mice in which the gliotransmitter release of ATP was impaired have shown the participation of astrocytes in certain cortical network activity and in animal behaviour [2],[3],[15],[16]. However, the exact underlying cellular mechanisms are largely undefined. Furthermore, while the involvement of astrocytes in some forms of long-term potentiation (LTP) has been shown in hippocampal slices (e.g., [6],[11]), the active participation of astrocytes in specific forms of synaptic plasticity in vivo remains unknown. Cholinergic system is involved in many different processes of brain function [17]. In the hippocampus, cholinergic activity modulates neuronal excitability [18], network activity [19], as well as synaptic transmission and plasticity [20],[21]. In the CA1 region, acetylcholine (ACh) induces CA1 pyramidal neuron depolarization [18], theta rhythm generation [19], and LTP of glutamatergic CA3-CA1 synaptic transmission [20],[21], as well as astrocyte Ca2+ elevations [22],[23]. However, the physiological meaning of the cholinergic evoked astrocyte Ca2+ signal remains unknown. In the present work we have investigated two fundamental questions regarding the direct involvement of astrocytes in synaptic physiology, i.e, whether astrocytes actively participate in physiological processes underlying synaptic plasticity, and whether astrocyte synaptic modulation occurs in vivo. We have recently shown that the coincidence of astrocyte Ca2+ elevations evoked by Ca2+ uncaging and mild postsynaptic depolarization induces LTP in hippocampal synapses [11]. Therefore, we have investigated whether the astrocyte Ca2+ signal evoked by cholinergic activity [22] is involved in the generation of cholinergic-induced LTP of glutamatergic CA3-CA1 synapses. Using in vivo experimental approaches, we have found that cholinergic activity evoked by sensory stimulation or electrical stimulation of the septal nucleus, the main cholinergic input to the hippocampus, elevated Ca2+ in hippocampal astrocytes and induced LTP in CA3-CA1 synapses. Using hippocampal slices to investigate the underlying cellular mechanisms, we have found that stimulation of cholinergic axons evoked astrocyte Ca2+ elevations, depolarization of CA1 pyramidal neurons, and LTP in CA3-CA1 synapses. Like in vivo, astrocyte Ca2+ elevations and LTP required mAChR activation, and LTP also required mGluR activation. Cholinergic-induced astrocyte Ca2+ elevations and LTP were absent both in IP3R2 knock-out mice and in wildtype mice after loading astrocytes with BAPTA or GDPβS (which prevented astrocyte Ca2+ signalling). Notably, LTP was rescued by simultaneous astrocyte Ca2+ uncaging and postsynaptic depolarization. Taken together, these results indicate that astrocyte Ca2+ signal is necessary for cholinergic-induced hippocampal synaptic plasticity. In summary, present results show that cholinergic LTP requires the astrocyte Ca2+ signal, which stimulates the release of glutamate from astrocytes that activates mGluRs on neurons. Then, cholinergic-induced hippocampal LTP results from the coincidence of astrocyte and postsynaptic activities simultaneously evoked by cholinergic signalling. Results Cholinergic Activity Evokes Astrocyte Ca2+ Elevations and LTP In Vivo We first assessed in vivo whether cholinergic activity regulates astrocyte Ca2+ signal and synaptic transmission (see Materials and Methods). In anesthetized rats, somatosensory stimulation by tail pinch, which stimulates cholinergic activity and hippocampal theta rhythm [24],[25], evoked Ca2+ elevations in hippocampal astrocytes (34 out of 66 astrocytes from n = 8 rats) that were abolished by the cholinergic muscarinic receptor (mAChRs) antagonist atropine (5 mg/kg) (n = 4 rats; Figure 1A–C). We analyzed hippocampal synaptic transmission in CA3-CA1 synapses, recording field EPSPs (fEPSPs) evoked by Schaffer collaterals (SC) stimulation in the CA1 pyramidal layer. Sensory stimulation also induced the LTP of fEPSPs (n = 7; Figure 1D and 1E). Similar LTP was also found after electrical stimulation of the medial septal nucleus (the main cholinergic input to the hippocampus) with a theta-like burst stimulation paradigm (TBS) (n = 9; see Materials and Methods; Figure 1F and 1G) [26],[27]. This LTP evoked by sensory or electrical stimulation was prevented in the presence of antagonists of either muscarinic receptors (mAChRs; 5 mg/Kg atropine) or metabotropic glutamate receptors (mGluRs; 1 mM MCPG) (n = 6 in each case) (Figure 1D–G), indicating that septohippocampal cholinergic activity induced the long-term potentiation (c-LTP) of CA3-CA1 synapses, which also required mGluR activation. Furthermore, because astrocyte responsiveness to sensory stimulation was similar in control and in the presence of MCPG (n = 3; Figure 1C), mGluR activation was downstream the astrocyte Ca2+ signal. 10.1371/journal.pbio.1001259.g001 Figure 1 Cholinergic activity induces astrocyte Ca2+ elevations and LTP in CA3-CA1 synapses in the hippocampus in vivo. (A) Schematic drawing of the experimental approach used to monitor Ca2+ levels in hippocampal astrocytes in vivo; representative images of astrocytes labeled with sulforhodamine 101 (SR101) and loaded with Fluo-4-AM; corresponding merge image; and image of Fluo-4-loaded astrocytes displaying regions of interests. Scale bar, 20 µm. (B) Fluorescence traces of Ca2+ levels in regions of interests in astrocytes showed in (A) evoked by tail pinch sensory stimulation (horizontal bars) in control and in the presence of atropine. (C) Proportion of astrocytes responding to sensory stimulation in control (66 astrocytes from n = 8 rats), atropine (32 astrocytes from n = 4 rats), and MCPG (15 astrocytes from n = 3 rats). (D) Schematic drawing of the in vivo experimental approach showing the stimulating electrode in the Schaffer collaterals (SC) and the extracellular recording electrode of fEPSPs placed in the hippocampal CA1 region, and a representative trace of a field potential showing hippocampal theta rhythm activity (bottom) during tail pinch sensory stimulation. Right, Relative fEPSP slope (from basal values) versus time. Zero time corresponds to the onset of stimulation (as in all other figures). Inset: mean fEPSPs before and 60 min after stimulation. (E) Average relative changes of fEPSP evoked 60 min after sensory stimulation in control (n = 7), atropine (n = 6), and MCPG (n = 6). (F) Schematic drawing showing the additional stimulating electrode in the medial septum nucleus. Right, Relative fEPSP slope (from basal values) versus time. Zero time corresponds to the onset of stimulation that lasted 90.7 s (horizontal bar). Inset: mean fEPSPs before and 60 min after septum stimulation. (G) Average relative changes of fEPSP evoked 60 min after stimulation in control (n = 9), atropine (n = 6), and MCPG (n = 6). ***p 1 h at room temperature (21–24°C) in ACSF that contained (in mM): NaCl 124, KCl 2.69, KH2PO4 1.25, MgSO4 2, NaHCO3 26, CaCl2 2, and glucose 10, and was gassed with 95% O2 / 5% CO2 (pH = 7.3). Slices were then transferred to an immersion recording chamber and superfused with gassed ACSF including 0.05 mM Picrotoxin and 5 µM CGP 55845 to block GABA receptors. To prevent possible NMDAR-mediated plasticity, experiments were performed in the presence of 50 µM AP5. Cells were visualized under an Olympus BX50WI microscope (Olympus Optical, Tokyo, Japan). Electrophysiology In Vivo For rats, electrodes were placed stereotaxically according to [42]. Field potentials were recorded through tungsten macroelectrodes (1 MΩ) placed in the CA1 layer (A, −3.8; L, 1; V, 2.5 mm from Bregma). For mice, electrodes were placed stereotaxically according to [43]. Recording electrodes were placed at the CA1 area (1.2 mm lateral and 2.2 mm posterior to Bregma; depth from brain surface, 1.0–1.5 mm) and bipolar stainless steel stimulating electrodes aimed at the right Schaffer collateral–commissural pathway of the dorsal hippocampus (2 mm lateral and 1.5 mm posterior to Bregma; depth from brain surface, 1.0–1.5 mm). Extracellular excitatory postsynaptic field potentials (fEPSPs) were amplified (DAM80; World Precision Instruments, Sarasota, FL), bandpass filtered between 0.1 Hz and 1.0 kHz, and digitized at 3.0 kHz (CED 1401 with Spike 2 software; Cambridge Electronic Design, Cambridge, UK). SC fibers continuously stimulated with single pulses (100 µA, 0.3 ms, 0.5 Hz) using a bipolar stainless steel stimulating electrode (0.1 mm diameter) placed in the stratum radiatum (A, −3.8; L, 4; V, 4 mm from Bregma). The medial septum (A, −0.2; L, 0; V, 7 mm from Bregma) was stimulated using a similar electrode. The initial phase of the fEPSP was used to quantify SC synaptic transmission. To mimic theta activity (theta burst stimulation, TBS), the medial septum was stimulated with four trains at 5 Hz of 5 stimuli (at 40 Hz) delivered 10 times at 0.1 Hz. Electrophysiology in Slices Electrophysiological recordings from CA1 pyramidal neurons and astrocytes located in the stratum radiatum were made using the whole-cell patch-clamp technique. Patch electrodes had resistances of 3–10 MΩ when filled with the internal solution that contained (in mM) for pyramidal neurons: KGluconate 135, KCl 10, HEPES 10, MgCl2 1, ATP-Na2 2 (pH = 7.3); and astrocytes were patched with 4–9 MΩ electrodes filled with an intracellular solution containing (in mM): MgCl2 1, NaCl 8, ATP-Na2 2, GTP 0.4 , HEPES 10, and either 40 mM BAPTA or 20 mM GDPβS, titrated with KOH to pH 7.2–7.3 and adjusted to 275–285 mOsm. Recordings were obtained with PC-ONE amplifiers (Dagan Corporation, Minneapolis, MN). Fast and slow whole-cell capacitances were neutralized and series resistance was compensated (≈70%). Recordings were rejected when the access resistance increased >20% during the experiment. Recordings from pyramidal neurons were performed in voltage-clamp conditions and the membrane potential was held at −70 mV to record SC-evoked EPSCs. During alveus TBS, recordings were performed in current-clamp conditions, unless stated otherwise (e.g., Figure 5C and 5D). Signals were fed to a Pentium-based PC through a DigiData 1440 interface board (Axon Instruments). The pCLAMP 10 software (Axon Instruments) was used for stimulus generation, data display, acquisition, and storage. Experiments were performed at room temperature (21–24°C). To stimulate cholinergic axons, theta capillaries (10–30 µm tip; WPI, Sarasota, FL) filled with ACSF were used for bipolar stimulation. The electrodes were connected to a stimulator S-900 through an isolation unit (S-910, Dagan Corporation) and placed in the stratum oriens/alveus near the subiculum area (for simplicity herein termed alveus), which contains cholinergic axons from the diagonal band of Broca and septum [22],[44]. For TBS, four trains at 5 Hz of 5 stimuli (at 40 Hz) were delivered 10 times at 0.1 Hz. To stimulate SC fibers, electrodes were placed in the stratum radiatum of the CA1 region. Single pulses (250 µs duration) or paired pulses (50 ms interval) were delivered at 0.33 Hz. Basal EPSC values were recorded 10 min before the stimulus, and the relative mean amplitudes of 10 consecutive EPSCs from basal values were plotted over time (e.g., Figure 2F). Long-term changes of synaptic transmission were assessed from the relative amplitude of 30 consecutive EPSCs recorded 54–60 min after the stimulus (e.g., Figure 2G). Paired-pulse facilitation was quantified as PPF = [(2nd EPSC−1st EPSC)/1st EPSC]. For minimal stimulation of SC, the stimulus intensity (10–50 mA) was adjusted to meet the conditions that putatively stimulate a single, or very few, synapses (cf. [7],[11],[45],[46]) and was unchanged during the experiment. The recordings that did not meet these criteria [7],[11],[45],[46] and synapses that did not show amplitude stability of EPSCs were rejected. The synaptic current parameters analyzed were: synaptic efficacy (mean peak amplitude of all responses including failures), synaptic potency (mean peak amplitude of the successes), probability of release (Pr, ratio between number of successes versus total number of stimuli), and paired-pulse facilitation. The responses and failures were identified by visual inspection. Calcium Imaging In Vivo Adult animals were craniotomized and the cortical tissue above the hippocampus was removed by aspiration to expose the dorsal hippocampus (see [47],[48]), which was bathed with 4 µl of Fluo-4 AM (2 mM) and sulforhodamine 101 (SR101, 125 µM), for 30–60 min and covered with 2% agar and a glass coverslip. Most of the Fluo-4-loaded cells were astrocytes as indicated by their SR101 staining (Figure 1A) (cf., [49],[50]). Cells were imaged with an Olympus FV300 laser-scanning confocal microscope. Ca2+ variations recorded at the soma of 5 to 11 astrocytes in the field of view were estimated as changes of the fluorescence signal over the baseline (ΔF/F0). Astrocytes were considered to respond to the stimulation when ΔF/F0 increased two times the standard deviation of the baseline during the stimulus or with a delay ≤15 s after the end of the stimulus, and the proportion of responding astrocytes in different conditions was compared. Calcium Imaging in Slices Ca2+ levels in astrocytes located in the stratum radiatum of the CA1 region of the hippocampus were monitored by fluorescence microscopy using the Ca2+ indicator fluo-4 (Molecular Probes, Eugene, OR). Slices were incubated with fluo-4-AM (2–5 µl of 2 mM dye were dropped over the hippocampus, attaining a final concentration of 2–10 µM and 0.01% of pluronic) for 20–30 min at room temperature. In these conditions, most of the cells loaded were astrocytes [23], as confirmed in some cases by their electrophysiological properties [22],[51]–[53]. Astrocytes were imaged using a CCD camera (ORCA-235, Hamamatsu, Japan) attached to the microscope. Cells were illuminated during 100–500 ms with a xenon lamp at 490 nm using a monochromator Polychrome V (TILL Photonics, Gräfelfing, Germany), and images were acquired every 0.5–1 s. The monochromator and the camera were controlled and synchronized by the IP Lab software (BD Biosciences, MD) that was also used for quantitative epifluorescence measurements. Astrocyte Ca2+ levels were recorded from the astrocyte cell body and Ca2+ variations were estimated as changes in the fluorescence signal over the baseline. Astrocytes were considered to respond to the stimulation when ΔF/F0 increased two times the standard deviation of the baseline. In some cases, Ca2+ levels in single neurons or astrocytes were monitored by including 50 µM fluo-4 in the corresponding internal solution and recording pipette. The astrocyte Ca2+ signal was quantified from the probability of occurrence of a Ca2+ spike, which was calculated from the number of Ca2+ elevations grouped in 5-s bins recorded from 5 to 20 astrocytes in the field of view [7], and mean values were obtained by averaging different experiments. To test the effects of alveus stimulation on Ca2+ spike probability under different conditions, the respective mean basal (15 s before the stimulus start) and maximum Ca2+ spike probability (i.e., 5–10 s after) from different slices were averaged and compared. Local application of ACh (1 mM) was delivered by 30-s duration pressure pulses through a micropipette. Calcium Uncaging by UV-Flash Photolysis In photo-stimulation experiments, single astrocytes were electrophysiologically recorded with patch pipettes filled with the internal solution containing (in mM): MgCl2 1, NaCl 8, ATP-Na2 2, GTP 0.4, HEPES 10, GDPβS 20, NP-EGTA 5, and 50 µM fluo-4 (to monitor Ca2+ levels). Ca2+ uncaging was achieved by delivering 10 trains at 0.1 Hz of 5 pulses (1-ms duration, 6–15 mW) at 5 Hz of UV light (340–380 nm) to the soma and processes of the recorded astrocyte (optical window of 15–25 µm diameter) using a flash photolysis system (Rapp OptoElectronic, Hamburg, Germany). Drugs and Chemicals D-(-)-2-Amino-5-phosphonopentanoic acid (D-AP5), (S)-α-Methyl-4-carboxyphenylglycine (MCPG), (2S)-3-[[(1S)-1-(3,4-Dichlorophenyl)ethyl] amino-2-hydroxypropyl](phenylmethyl)phosphinic acid hydrochloride (CGP 55845), and 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetate (BAPTA) were purchased from Tocris Cookson (Bristol, UK). Fluo-4-AM, o-nitrophenyl EGTA, tetrapotassium salt (NP-EGTA), and sulforhodamine B were from Molecular Probes, Eugene, Oregon. All other drugs were from Sigma. For in vivo experiments, atropine sulfate (5 mg/kg) was intraperitoneally injected and its effects tested 10–15 min after the injection. For in vivo electrophysiological experiments, MCPG (100 nl, 1 mM) was injected into the hippocampus with a Hamilton microliter syringe. For in vivo Ca2+ imaging, MCPG (0.8 mM) was included in the solution bathing the dorsal hippocampus. Data are expressed as mean ± s.e.m. Results were compared using a two-tailed Student's t test (α = 0.05). Statistical differences were established with p<0.05 (*), p<0.01 (**), and p<0.001(***). Supporting Information Figure S1 Astrocyte-mediated c-LTP was associated with changes in paired-pulse facilitation index. (A) Representative mean EPSCs (10 consecutive traces) evoked by paired-pulse stimulation of SC before (basal) and 60 min after alveus TBS stimulation, and scaled trace of the basal EPSCs. (B) Summary of PPF index before (basal) and 60 min after alveus TBS (n = 10). *p<0.05. Data are presented as means ± s.e.m. (TIF) Click here for additional data file. Figure S2 Astrocyte Ca2+ elevations induce LTP of transmitter release at single hippocampal synapses. (A–D) Relative changes in synaptic efficacy (i.e., mean amplitude of responses including successes and failures of neurotransmission), probability of neurotransmitter release (Pr), and synaptic potency (i.e., mean EPSC amplitude excluding failures) (bin width, 2 min) over time in basal non-stimulated slices (n = 5), UV-flash astrocyte stimulation (n = 6), alveus TBS (n = 5), and pairing both stimuli (n = 4). Zero time corresponds to the onset of the stimulation (UV Ca2+ uncaging and alveus TBS are depicted by arrows and horizontal bars, respectively). (TIF) Click here for additional data file.
