Astrocyte inositol triphosphate receptor type 2 and cytosolic phospholipase A2 alpha regulate arteriole responses in mouse neocortical brain slices

Abstract

Functional hyperemia of the cerebral vascular system matches regional blood flow to the metabolic demands of the brain. One current model of neurovascular control holds that glutamate released by neurons activates group I metabotropic glutamate receptors (mGluRs) on astrocytes, resulting in the production of diffusible messengers that act to regulate smooth muscle cells surrounding cerebral arterioles. The acute mouse brain slice is an experimental system in which changes in arteriole diameter can precisely measured with light microscopy. Stimulation of the brain slice triggers specific cellular responses that can be correlated to changes in arteriole diameter. Here we used inositol trisphosphate receptor type 2 (IP(3)R2) and cytosolic phospholipase A(2) alpha (cPLA(2)α) deficient mice to determine if astrocyte mGluR activation coupled to IP(3)R2-mediated Ca(2+) release and subsequent cPLA(2)α activation is required for arteriole regulation. We measured changes in astrocyte cytosolic free Ca(2+) and arteriole diameters in response to mGluR agonist or electrical field stimulation in acute neocortical mouse brain slices maintained in 95% or 20% O(2). Astrocyte Ca(2+) and arteriole responses to mGluR activation were absent in IP(3)R2(-/-) slices. Astrocyte Ca(2+) responses to mGluR activation were unchanged by deletion of cPLA(2)α but arteriole responses to either mGluR agonist or electrical stimulation were ablated. The valence of changes in arteriole diameter (dilation/constriction) was dependent upon both stimulus and O(2) concentration. Neuron-derived NO and activation of the group I mGluRs are required for responses to electrical stimulation. These findings indicate that an mGluR/IP(3)R2/cPLA(2)α signaling cascade in astrocytes is required to transduce neuronal glutamate release into arteriole responses.

Figure 1. Astrocyte Ca²⁺ responses to mGluR agonist application are attenuated IP₃R2^−/− slices.

A. Neocortical brain slices from IP₃R2^+/+ (left panel) and IP₃R2^{− /−} (right panel) mice were loaded with the Ca²⁺-sensitive fluorophore Rhod-2/AM and astrocytes were identified by dye uptake, morphology and location. Ca²⁺ fluorescence was measured in the region of interest (green arrow) and is displayed at 3 time points in relation to 1S,3R-ACPD treatment: (a) before, (b) at peak response and (c) after. White outline indicates the region of magnification in C. Scale bar: 20 µm. B. Fluorescence intensity signals for the Ca²⁺ fluorescence measured in the soma of the indicated astrocytes. Signals were corrected for background that was measured in an identical area immediately adjacent to the region of interest. Representative single traces of the Ca²⁺ response in soma of IP₃R2^+/+ astrocytes (left trace) and IP₃R2^{− /−} (right trace) are shown and the duration of 1S, 3R-ACPD application is indicated below the traces. C. Z-stack of 12 images encompassing 12 mm of depth in IP₃R2^+/+ (left panel) and IP₃R2^{− /−} (right panel) brain slices. This demonstrates the ameboid shape of the astrocyte soma which extends a foot process near a neighboring arteriole. Scale Bar: 10 µm. D. Cumulative probability histograms of population responses are shown. Peak (left panel) and integrated (right panel) Ca²⁺ responses of IP₃R2^+/+ (open circles, 58 cells) and IP₃R2^{− /−} (filled circles, 63 cells) astrocytes with inset bar graphs indicating the mean ± S.E.M. Nine slices were prepared from four mice for each genotype.

Figure 2. Arteriole responses to mGluR agonist application are eliminated in IP₃R2^−/− neocortical slices.

A. Gradient contrast imaging was used to measure arteriole responses to treatment. Responses to 1S,3R-ACPD were quantified by defining 6 points (2 of the points are shown in this figure) at which to measure changes in arteriole diameter over time (described in Methods). The arteriole diameter is expressed as the average diameter of the points. Scale bar: 5 µm. B. Cortical slices from IP₃R2^+/+ (open circle, n = 18) and IP₃R2^{− /−} (filled circle, n = 24) were treated with 1S,3R-ACPD followed by PGE₂ and arteriole responses were measured. C. During continuous application of U-46619 the IP₃R2^+/+ and IP₃R2^{− /−} were treated with 1S,3R-ACPD while another group of IP₃R2^+/+ slices were treated with vehicle instead of 1S,3R-ACPD (red inverted triangle, n = 7). Inset shows the complete experiment from the time of application of U-46619. The dashed white box indicates the expanded graph. Treatment with U-46619 (100 nM) constricted arterioles of both IP₃R2^+/+ (n = 16) and IP₃R2^{− /−} slices (n = 17) to a similar extent. **, P<0.01 comparing IP₃R2^+/+ to IP₃R2^{− /−}10 min following 1S, 3R-ACPD application.

Figure 3. Neocortical arterioles of IP₃R2^{− /−} slices do not respond to electrical field stimulation.

A. Responses of arterioles in 95% O₂ following electrical field stimulation. Cortical slices from IP₃R2^+/+ (open circle) and IP₃R2^{− /−} mice (closed circle) were treated with electrical field stimulation (EFS) of 100 Hz trains of 200 ms at 0.2 Hz for 4 minutes as indicated by the dark bar. Arteriole diameter was measured every 5 min during the experiment. IP₃R2^+/+, n = 15; IP₃R2^{− /−} n = 15. **, P<0.01, IP₃R2^+/+ compared to IP₃R2^{− /−}. B. Pretreatment with U-46619 for 30 min was followed by electrical stimulation and arteriole diameters of IP₃R2^{− /−} were compared to IP₃R2^+/+. IP₃R2^+/+, n = 15; IP₃R2^{− /−} n = 14. **, P<0.01; ***, P<0.001. C. Blockade of Group I mGluR with MPEP and JNJ prevents arteriole responses to electrical stimulation in IP₃R2^+/+ slices (inverted triangles) while a 20% O2 environment does not alter responses of naïve IP₃R2^+/+ or IP₃R2^{− /−} arterioles. IP₃R2^+/+, n = 14; IP₃R2^{− /−} n = 12, IP₃R2^+/+ with MPEP/JNJ, n = 9. **, P<0.01; ***, P<0.001.

Figure 4. Arteriole responses to mGluR agonist application are eliminated in cPLA₂α^−/− neocortical slices.

A. Bath application of 1S,3R-ACPD 50 µM to slices equilibrated with 95% O₂ induced constriction of arterioles in cortical slices of cPLA₂α^+/+ (empty square, n = 29), but not cPLA₂α ^{− /−} mice (filled square, n = 15). Washout of 1S,3R-ACPD was followed by application of 10 µM PGE₂ which caused identical constriction of arterioles in both genotypes. B. In slices at equilibrium with 95% O₂, arterioles were preconstricted with 100 nM U-46619 supplemented ACSF. To compare acute cPLA₂α inhibition with gene deletion, cPLA₂α^+/+ slices were treated with 10 µM ATK (filled black triangle, n = 21) for the duration of the experiment. Inset shows the complete experiment from the time of application of U-46619. The dashed white box indicates the expanded graph. After 30 min equilibration in U-46619, 1S,3R-ACPD was added to the bath at a final concentration of 50 µM (time  = 0) and the responses of arterioles in cPLA₂α^+/+ (n = 21) and cPLA₂α^{− /−} (n = 25) cortical slices were compared to cPLA₂α^+/+ slices that were not treated with 1S,3R-ACPD (red empty triangle, n = 10). *, P<0.05; **, P<0.01; ***, P<0.001. C. When slices were equilibrated in 20% O₂, 1S,3R-ACPD treatment dilated cPLA₂α^+/+ (n = 18) but not cPLA₂α^{− /−} (n = 16) arterioles. Bath application of 10 µM PGE₂ caused dilation of both genotypes. ***, P<0.001.

Figure 5. Arterioles of cPLA₂α^−/− neocortical slices do not dilate in response to electrical stimulation.

A. Neocortical brain slices from cPLA₂α^+/+ (empty square, n = 23) and cPLA₂α^{− /−} mice (filled square, n = 23) were stimulated with 100 Hz trains of 200 ms at 0.2 Hz for 4 minutes as indicated by the dark bar (electrical field stimulation, EFS). Arteriole diameter was measured every 5 min during the experiment. *, P<0.05; **, P<0.01; *** P<0.001: cPLA₂α^+/+ compared to cPLA₂α^{− /−}. B. After treatment with 20 nM U-46619 for 30 min, electrical stimulation was applied to cortical slices from cPLA₂ ^+/+ and cPLA₂α^{− /−} mice and changes in arteriole diameter were compared to cPLA₂α^+/+ arterioles that did not have electrical stimulation (red triangle, n = 9). *, P<0.05; **, P<0.01; *** P<0.001: cPLA₂α^+/+ and cPLA₂α^{− /−} with electrical stimulation compared to cPLA₂α^+/+ without electrical stimulation at the same time points.

Figure 6. Inhibition of nNOS with L-NPA prevents dilation of arterioles following electrical stimulation.

cPLA₂α^+/+ slices were equilibrated in 20% O₂ and treated with ACSF (hollow square, n = 14; ) or ACSF with 10 µM L-NPA 60 min before electrical stimulation (filled red triangle, n = 9). cPLA₂α^{− /−} (filled square, n = 9) were treated with ACSF. Electrical stimulation was applied as indicated by the dark bar (electrical field stimulation, EFS). *, P<.05; **, P<.01.

Figure 7. Astrocyte Ca²⁺ responses in neocortical slices to mGluR agonist application are not altered by absence of cPLA₂α expression.

Cortical brain slices from S100β-EGFP/cPLA₂α^+/+ (upper panel) and S100β-EGFP/cPLA₂α^{− /−} (lower panel) mice were loaded with the Ca²⁺-sensitive fluorophore Rhod-2/AM. Regions of interest representing astrocyte soma (white circles) and vascular foot processes (white box) were identified by EGFP expression and histologic location. Regions of interest representing background fluorescence for soma and endfeet are depicted by blue outlines. Ca²⁺ fluorescence measured for the soma and endfeet are displayed at 3 times in relation to 1S,3R-ACPD treatment: (a) before, (b) at peak response and (c) after. Representative Ca²⁺ measurements for soma and endfeet for each genotype are depicted in the right panel. The time of the 1S, 3R ACPD application is indicated by the black bar. The Ca²⁺ responses of astrocyte populations are shown in B. Soma (+/+, n = 169; ^{− /−}, n = 166) and C. endfeet (+/+, n = 36; ^{− /−}, n = 33) and were measured as Ca²⁺ peak amplitude, area under curve, half width, rise time, decay time or decay tau (as defined in Methods). The graphs show cumulative probability histograms analysis of the astrocyte populations by parameters compared between cPLA₂α^+/+ (open circles and bars) and ^{− /−} (closed circles and bars) while inset bar graph shows the mean ± S.E.M. for each parameter. There were no significant differences between the genotypes.