Photolysis of diazo-2 stimulates ATP release and slowly propagating Ca²⁺ waves. (A) Time series of bioluminescence imaging of photolysis-evoked ATP release with (upper panel) or without (lower panel) diazo-2. Luciferinluciferase was added to the bath, and a highly light-sensitive liquid nitrogen–cooled camera detected photons released from hippocampal slices in response to UV photolysis of diazo-2 (eight pulses). White circles indicate the position of UV photolysis. Scale bars, 100 mm. (B) Bar graph comparing peak luminescence signaling in the presence or absence of diazo-2 or luciferase (n = 3 to 6 photolysis events, **P < 0.01 compared to presence of diazo-2 and luciferase). (C) Time series of astrocytic Ca²⁺ (rhod-2) and LFP signals in response to photolysis of diazo-2. Hatched circles define the 50- and 100-μm radius from the photolysis site. Traces below show Ca²⁺ changes (percent of fluorescence change from baseline, ΔF/F) in individual cells 50, 100, or 150 μm from the site of photolysis, as well as LFP recordings (50-μm distance) after diazo-2 photolysis from the same experiment. Lower traces show similar experiment in the absence of diazo-2. Small white arrowheads indicate astrocytes that exhibit increases in Ca²⁺. (D) Bar graph comparing astrocytic Ca²⁺ increases (upper) and LFP (lower) evoked by photolysis of diazo-2 in the presence of suramin (50 μM) or PPADS (50 μM) (n = 5 to 8 photolysis events, **P < 0.01). Scale bar, 50 μm. We could not perform bioluminescence imaging in the presence of P2 receptor antagonists because all agents tested interfered with the luciferinluciferase reaction.

Photolysis of MNI-glutamate evokes an initial fast Ca²⁺ increase followed by a slowly propagating Ca²⁺ wave. (A) Serial images of relative changes in Ca²⁺ (rhod-2) in response to photolysis of caged MNI-glutamate in a hippocampal slice. Hatched circles label the radius 50, 100, and 150 μm from the photolysis site. Two recording electrodes (E1, E2) were placed at different distances from the site of photolysis. Small white arrows indicate astrocytes. Scale bar, 50 μm. (B) Traces of Ca²⁺ signals in astrocytes located 50, 100, and 150 μm from the site of photolysis. As shown, MNI-glutamate photolysis induced an initial fast increase in Ca²⁺ that includes most cells in the field followed by a Ca²⁺ wave that slowly propagates from the site of photolysis. Red lightning bolt indicates the time of photolysis. LFP signals recorded at electrodes E1 and E2. Small black arrows indicate the passage of the second Ca²⁺ wave. Inset shows enlarged time scale of the depolarization (LFP deflection). Black arrows indicate the initiation of depolarization. (C) Bar graphs compared the amplitude of the increase in intracellular Ca²⁺ (left panel) and the LFP signal (right panel) in response to MNI-glutamate photolysis in controls and in slices exposed to TBOA (100 μM), APV (50 μM), CNQX (20 μM), APV + CNQX, MPEP (50 μM) + LY367385 (50 μM), or nifedipine (10 μM) (n = 11 to 40 photolysis events, *P < 0.05, **P < 0.01 compared to respective controls). (D) Ca²⁺ increases in astrocytes during the first and the second Ca²⁺ wave in control slices and slices exposed to suramin (100 μM), PPADS (30 μM), or Reactive Blue 2 (RB, 30 μM) (n = 16 photolysis events, **P < 0.01 compared to respective controls).

High-frequency stimulation triggers a decrease in [Ca²⁺]_e and stimulates the release of ATP and a slowly propagating Ca²⁺ wave. (A) Representative traces of [Ca²⁺]_e and LFP evokedby an increasing number of stimulating pulses (10 to 200) in hippocampal slices. Lower trace shows expanded time scale for the stimulation. Rapid downward and upward peaks (indicated by arrows)before the slower downward evoked EPSPs represent stimulus artifacts. (B) Representative traces of [Ca²⁺]_e and LFP evoked by increasing number of stimulating pulses (10 to 200) in the presence of 50 μM APV and 20 μM CNQX. (C) Bar graph comparing changes in [Ca²⁺]_e in response to HFS in the absence or presence of APV and CNQX (n = 5 to 6 HFS events; *P < 0.05, **P < 0.01 compared to 0 pulse; ^#P < 0.05, ^## P < 0.01 between vehicle and APV/CNQX). (D) Bioluminescence imaging of ATP release in response toHFS (200 pulses). White dotted lines indicate the position of stimulation electrode. Scale bar, 50 μm. (E) Time series of astrocytic Ca²⁺ (rhod-2) in response to HFS. Hatched circles indicate the radius 50 and 100 μm from the photolysis site. Traces on the right display Ca²⁺ changes in individual cells 50, 100, and 150 μm from the site of diazo-2 photolysis in the same experiment. Small white arrowheads indicate astrocytes. Scale bar, 50 μm. (F) Time series of astrocytic Ca²⁺ in response to HFS in the presence of APV and CNQX. (G) Comparison of Ca²⁺ responses and LFP changes in the presence of vehicle (control) or APV and CNQX (n = 16 to 20 for Ca²⁺ responses, n = 5 to 6 for LFP HFS events; **P < 0.001 between vehicle and APV/CNQX).

Photolysis of diazo-2 triggers a dose-dependent decrease in [Ca²⁺]_e. (A) Experimental setup used for photolysis of diazo-2 in hippocampal slices. The Ca²⁺-sensitive microelectrode was placed at a distance of 75 μm from the site of photolysis. (B) Traces of changes in [Ca²⁺]_e evoked by photolysis of diazo-2 in response to 1, 2, 4, 6, 8, or 10 UV pulses. In the absence of diazo-2, 10 UV pulses did not elicit a change in [Ca²⁺]_e. (C) Bar graph displaying the changes in [Ca²⁺]_e evoked by photolysis of diazo-2 plotted as a function of increasing number of UV photolysis. Omission of diazo-2 eliminated photolysis-induced decreases in [Ca²⁺]_e (n = 3 to 4 photolysis events, *P < 0.05 compared to 0 pulse).

Photolysis of MNI-glutamate triggers a decrease in [Ca²⁺]_e and release of ATP. (A) Experimental setup used for measuring bioluminescence and [Ca²⁺]_e after photolysis of MNI-glutamate in hippocampal slices. (B) Representative traces of [Ca²⁺]_e and LFP in response to increasing numbers of UV pulses (1 to 10) in cortical slices. (C) Bar graph comparing changes in [Ca²⁺]_e and LFP (black bars) in response to MNI-glutamate photolysis and increasing the number of 1 to 10 UV pulses. Omission of MNI-glutamate eliminated changes in [Ca²⁺]_e and LFP (n = 5 to 6 photolysis events, P > 0.05). (D) Bioluminescence imaging of ATP release in response to photolysis of MNI-glutamate. Lower panel shows photon emission in response to photolysis in the absence of MNI-glutamate. Scale bars, 100 μm. (E) Bar graph comparing the effect of omitting MNI-glutamate or luciferase on the peak bioluminescence signal evoked by photolysis (n = 12 photolysis events, **P < 0.01).

The second Ca²⁺ wave depends on astrocytic hemichannels and is potentiated by a point mutation that increases the number of open Cx43 hemichannels in astrocytes. (A) Time series of Ca²⁺ signaling evoked by photolysis of MNI-glutamate in a hippocampal slice prepared from a Cx43/30KO mouse. Hatched circles define the 50-, 100-, and 150-μm radius from the photolysis site. Scale bar, 75 μm. (B) Traces of Ca²⁺ increases and LFP at 50, 100, and 150 μm in experiment shown in (A). Inset shows enlarged time scale for the depolarization. Black arrows indicate the initiation of depolarization. (C) Comparison of the maximal amplitudes of LFP depolarization (left) and intracellular Ca²⁺ increases (right) evoked by photolysis of MNI-glutamate in hippocampal slices prepared from Cx30KO or Cx43/30KO mice, or Cx30KO slices treated with CBX (50 μM) (n = 14 to 21 photolysis events, **P < 0.01). (D) Ca²⁺ signaling evoked by MNI-glutamate photolysis in a hippocampal slice prepared from a Cx43G138R mouse. Small white arrows point to astrocytes. Scale bar, 75 μm. (E) Traces of Ca²⁺ signals and LFP in experiment shown in (D). (F) Histogram comparing the maximal amplitudes of LFP depolarization (left) and increased intracellular Ca²⁺ (right) in hippocampal slices prepared from Cx43WT (wild type) or Cx43G138R mice in response to photolysis of MNI-glutamate (n = 14 to 21 photolysis events, *P < 0.05, **P < 0.01). (G) Histograms comparing the maximal radius (n = 28 photolysis events; **P < 0.001 compared to Cx43WT) and durations (n = 5 to 9 photolysis events; **P = 0.004 compared to Cx43WT) of the Ca²⁺ wave evoked by photolysis of MNI-glutamate.

Diazo-2 photolysis evokes interneuronal depolarization and bursting. (A) Experimental setup used for photolysis of diazo-2 combined with whole-cell recordings of interneurons located ~40 to 80 μm from the site targeted by the UV beam in hippocampal slices. Time series images of the slow Ca²⁺ wave evoked by photolysis of diazo-2. The pipette contained Alexa Fluor 488 (green) to visualize the impaled interneuron. Small white arrows point to astrocytes. Hatched circles define the radius 50, 100, and 150 μm from the photolysis site. Scale bar, 50 μm. (B) Representative recordings of interneurons (current clamp) in response to diazo-2 photolysis without (left panel) or with TTX (1 μM, right panel). Of a total of 25 recordings, 17 interneurons exhibited a transient depolarization (membrane potential shift >2 mV), and 8 of these cells exhibited spiking activity. In the presence of TTX, 12 of a total of 18 interneurons exhibited a transient depolarization (>2 mV), whereas the remaining 6 failed to respond to diazo-2 photolysis. (C) Comparison of the amplitude, delay, and duration of diazo-2–induced depolarization in the presence and absence of TTX (n = 6 to 8 photolysis events, P > 0.05). (D) Representative traces of recordings in a CA1 pyramidal neuron (current clamp) located ~40 to 80 μm from diazo-2 photolysis site. Enlarged time scales for the indicated periods are shown below. (E) Changes in the frequency of IPSCs induced by diazo-2 photolysis plotted as a function of time before or after diazo-2 photolysis (n = 5 photolysis events, **P < 0.01). Inset: Bicuculline reduced the frequency of IPSCs during baseline conditions (20 mM n = 9 slices, **P < 0.01). (F) Comparison of the effect of photolysis of diazo-2 on IPSC frequency during control conditions (10 to 50 s), after addition of the P2Y1 receptor antagonist MRS2179 (50 μM), or in slices from mice lacking Cx43 and Cx30 (Cx43/30KO) (n = 9 slices, **P < 0.01). (G) Proposed model for [Ca²⁺]_e as a mediator of neuron-glia signaling. Glutamatergic signaling (or MNI photolysis) triggers Ca²⁺ influx into neurons through AMPA- or NMDA-type glutamate, leading to a localized decrease in [Ca²⁺]_e. Connexin hemichannels in astrocytes open in response to low [Ca²⁺]_e, enabling the efflux of ATP. ATP triggers two events: (i) ATP activates slowly propagating Ca²⁺ waves in astrocytes by binding to astrocytic P2YR1, P2YR2, and P2YR4; (ii) ATP is degraded to ADP, and in turn, ADP activates interneuronal P2Y1 receptors, stimulating the depolarization and increased firing of a subpopulation of interneurons. This mode of neuron-glia signaling may act as a negative feedback mechanism to increase inhibitory transmission during excessive glutamatergic activity.