Astrocyte β-Adrenergic Receptor Activity Regulates NMDA Receptor Signaling of Medial Prefrontal Cortex Pyramidal Neurons

Abstract

Glutamate spillover from the synapse is tightly regulated by astrocytes, limiting the activation of extrasynaptically located NMDA receptors (NMDAR). The processes of astrocytes are dynamic and can modulate synaptic physiology. Though norepinephrine (NE) and β-adrenergic receptor (β-AR) activity can modify astrocyte volume, this has yet to be confirmed outside of sensory cortical areas, nor has the effect of noradrenergic signaling on glutamate spillover and neuronal NMDAR activity been explored. We monitored changes to astrocyte process volume in response to noradrenergic agonists in the medial prefrontal cortex of male and female mice. Both NE and the β-AR agonist isoproterenol (ISO) increased process volume by ∼20%, significantly higher than changes seen when astrocytes had G-protein signaling blocked by GDPβS. We measured the effect of β-AR signaling on evoked NMDAR currents. While ISO did not affect single stimulus excitatory currents of Layer 5 pyramidal neurons, ISO reduced NMDAR currents evoked by 10 stimuli at 50 Hz, which elicits glutamate spillover, by 18%. After isolating extrasynaptic NMDARs by blocking synaptic NMDARs with the activity-dependent NMDAR blocker MK-801, ISO similarly reduced extrasynaptic NMDAR currents in response to 10 stimuli by 18%. Finally, blocking β-AR signaling in the astrocyte network by loading them with GDPβS reversed the ISO effect on 10 stimuli-evoked NMDAR currents. These results demonstrate that astrocyte β-AR activity reduces extrasynaptic NMDAR recruitment, suggesting that glutamate spillover is reduced.

Keywords: astrocyte; extrasynaptic NMDA receptors; glia modulation of synapses; norepinephrine; synaptic modulation; β-adrenergic receptors.

Figure 1.

Noradrenergic agonists increase astrocyte process volume in the mPFC. A, Confocal image of an astrocyte filled with Alexa Fluor 488 dextran (3,000 MW). B, Confocal image of multiple astrocytes labeled with SR 101 (magenta) and a single astrocyte filled with Alexa Fluor 488 dextran (cyan). The Alexa Fluor dye does not spread beyond the patched cell. C–F, Example fluorescence images of two dye-filled astrocytes. C, D, β-AR agonist ISO (5 μM) increases fluorescence intensity in the processes of an astrocyte. C, Last time point in aCSF. D, 40 min in ISO. E, F, Fluorescence intensity remains unchanged in an astrocyte loaded with GDPβS. E, Last time point in aCSF. F, 40 min in ISO. Pseudocolor scale indicates fluorescence intensity in arbitrary units. G, Astrocyte process brightness versus time from the experiment illustrated in C and D. H, Process brightness versus time for an astrocyte loaded with GDPβS, illustrated in E and F. In G and H, process brightness is measured in the ROIs indicated in C–F; the black bars indicate the time course of ISO incubation. I, Mean ± SEM astrocyte process fluorescence versus time for astrocytes initially in aCSF and exposed to experimental solutions for 40 min (black bar). Experimental solutions include ISO (purple), ISO with GDPβS-filled astrocyte (orange), aCSF control (blue), and aCSF control with GDPβS (green). J, Summary of astrocyte process fluorescence changes measured after 40 min incubation in noradrenergic agonists or control solution, with or without GDPβS loading of astrocytes. Mean ± SEM and data from individual mice are shown. NE (17.7 ± 2.4% increase; n = 7 mice), ISO (20.5 ± 4.6% increase; n = 5 mice), control aCSF (2.21 ± 1.86% increase; n = 8 mice), ISO-GDPβS (0.40 ± 1.67% increase; n = 6 mice), NE-GDPβS (4.88 ± 1.00% increase; n = 6 mice), aCSF GDPβS (3.71 ± 0.72% increase; n = 6 mice). Two-way ANOVA with Tukey–Kramer post hoc multiple comparisons, *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 2.

Ten stimuli-evoked NMDAR EPSCs are modulated by astrocyte β-AR activity. A, Confocal image of experimental arrangement for EPSC recording. A patched Layer 5 pyramidal neuron is filled with Alexa Fluor 594 (red) while astrocytes are initially labeled with SR 101 (also red). In some experiments, an astrocyte was patched (asterisk) and filled with GDPβS and Alexa Fluor 488 hydrazide (green), which diffused widely through the astrocyte network. SR 101-labeled astrocytes that are filled with GDPβS and Alexa Fluor 488 are orange (arrowheads) and surround the neuron. A theta tube pipette in Layer 2/3 (white lines, labeled Stim) evokes EPSCs in the labeled neuron. B–E, NMDAR EPSCs recorded from neurons in control aCSF (blue), after 40 min in ISO (red), and after an additional 5 min in D-AP5 (green). B, EPSCs evoked by single stimuli. C, EPSCs evoked by 10 stimuli bursts. D, EPSCs evoked by single stimuli with the astrocyte network loaded with GDPβS. E, EPSCs evoked by 10 stimuli bursts with the astrocyte network loaded with GDPβS. F, Summary of NMDAR EPSC amplitudes (normalized to paired control aCSF EPSCs) elicited by single and 10 stimuli. Single stimulus groups include control, all conditions, ISO (0.91 ± 0.06; n = 6 mice), ISO with GDPβS (1.07 ± 0.06; n = 7 mice), and D-AP5 (0.08 ± 0.01; n = 5 mice). Ten stimuli groups include control, all conditions, ISO (0.82 ± 0.01; n = 6 mice), ISO with GDPβS (1.30 ± 0.10; n = 7 mice), and D-AP5 (0.06 ± 0.01; n = 5 mice). G, Summary of NMDAR EPSC charge transfer (normalized to paired control aCSF EPSCs) elicited by single and 10 stimuli. Single stimulus groups include control, all conditions, ISO (0.95 ± 0.13; n = 6 mice), ISO with GDPβS (1.17 ± 0.13; n = 7 mice), and D-AP5 (0.07 ± 0.01; n = 5 mice). Ten stimuli groups include control, all conditions, ISO (0.78 ± 0.02; n = 6 mice), ISO with GDPβS (1.16 ± 0.10; n = 7 mice), and D-AP5 (0.07 ± 0.02; n = 5 mice). In F and G, mean ± SEM and data from individual mice are shown; one-sample t tests compare normalized experimental groups with null control condition, and paired t tests compare D-AP5 with paired ISO trials. Comparisons between ISO with and without GDPβS use unpaired t tests. All t tests are Bonferroni corrected for multiple comparisons. *p < 0.05, **p < 0.01, ***p< 0.001. H, Summary of the fast decay time constant (τ_f) from 10 stimuli-evoked NMDAR EPSCs. EPSC decay was fit by the sum of two exponentials (R² = 0.9965 ± 0.0005; n = 26 traces from 13 mice). Groups include no GDPβS control (0.302 ± 0.035 s; n = 6 mice), no GDPβS ISO (0.328 ± 0.033 s; n = 6 mice), GDPβS control (0.314 s ± 0.043 s; n = 7 mice), GDPβS ISO (0.292 s ± 0.017 s; n = 7 mice). Mean ± SEM and data from individual mice are shown; two-way ANOVA with Tukey–Kramer post hoc multiple comparisons, n.s. between all groups p > 0.88.

Figure 3.

Single and 10 stimuli-evoked AMPAR EPSCs are not modulated by astrocyte β-AR activity. A–D, AMPAR EPSCs recorded from single neurons in control aCSF (blue), after 40 min in ISO (red), and after an additional 5 min in CNQX (green). A, EPSCs evoked by single stimuli. B, EPSCs evoked by single stimuli and the astrocyte network loaded with GDPβS. C, EPSCs evoked by 10 stimuli bursts. D, EPSCs evoked by 10 stimuli bursts and the astrocyte network loaded with GDPβS. E, Summary of single and 10 stimuli-evoked AMPAR EPSCs (normalized to paired controls). Mean ± SEM and data from individual mice are shown. Single stimulus groups include control, all conditions, ISO (1.25 ± 0.14; n = 5 mice), ISO with GDPβS (1.24 ± 0.16; n = 6 mice), and CNQX (0.14 ± 0.02; n = 11 mice). Ten stimuli-evoked Σ AMPAR EPSC groups include control, all conditions, ISO (1.28 ± 0.12; n = 5 mice), ISO with GDPβS (1.03 ± 0.05; n = 6 mice), and CNQX (0.15 ± 0.02; n = 11 mice). One-sample t tests compare normalized experimental groups with null control conditions, and paired t test compares ISO groups with paired CNQX trials. ISO with and without GDPβS use unpaired t tests. All t tests are Bonferroni corrected for multiple comparisons. ***p < 0.001.

Figure 4.

β-AR activity modulates extrasynaptic NMDAR EPSCs. A, Single stimulus-evoked NMDAR EPSCs tracking synaptic NMDAR blockade by MK-801, recorded from a single neuron (normalized to pre MK-801). Data points with error bars represent mean ± SEM. Data points during MK-801 treatment indicate individual EPSCs. Time course of 10 min MK-801 treatment is indicated by black bar. B, NMDAR EPSCs evoked by single stimuli, recorded from a single neuron. Pre MK-801 aCSF control (purple), post MK-801 aCSF (blue), post MK-801 after 40 min in ISO (red), and after an additional 5 min in D-AP5 (green). C, Summary of single stimulus NMDAR EPSCs (normalized to pre MK-801 control aCSF). Pre MK-801 control aCSF (purple), post MK-801 control aCSF (blue; 0.12 ± 0.01; n = 6 mice), ISO (red; 0.11 ± 0.01; n = 6 mice). One-sample t tests compare normalized post MK-801 groups with the pre MK-801 null control condition. Paired t test compares post MK-801 aCSF with post MK-801 ISO. All t tests are Bonferroni corrected for multiple comparisons. n.s., p > 0.05, ***p < 0.001. D, NMDAR EPSCs evoked by 10 stimuli, recorded from a single neuron. Post MK-801 aCSF (blue), post MK 801 after 40 min in ISO (red), and after an additional 5 min in D-AP5 (green). E, Summary of 10 stimuli-evoked NMDAR EPSCs and charge transfer (normalized to post MK-801 control aCSF, blue). Post MK-801 ISO NMDAR EPSC amplitude (red, 0.82 ± 0.02; n = 6 mice) and post MK-801 ISO NMDAR EPSC charge transfer (red, 0.90 ± 0.03; n = 6 mice). One-sample t tests compare normalized post MK-801 ISO with null post MK-801 control aCSF for NMDAR EPSC amplitude and charge transfer measures. All t tests are Bonferroni corrected for multiple comparisons. *p < 0.05, ***p < 0.001. Mean ± SEM and data from individual mice are shown in C and E.