Granular Layer Neurons Control Cerebellar Neurovascular Coupling Through an NMDA Receptor/NO-Dependent System

Abstract

Neurovascular coupling (NVC) is the process whereby neuronal activity controls blood vessel diameter. In the cerebellum, the molecular layer is regarded as the main NVC determinant. However, the granular layer is a region with variable metabolic demand caused by large activity fluctuations that shows a prominent expression of NMDA receptors (NMDARs) and nitric oxide synthase (NOS) and is therefore much more suitable for effective NVC. Here, we show, in the granular layer of acute rat cerebellar slices, that capillary diameter changes rapidly after mossy fiber stimulation. Vasodilation required neuronal NMDARs and NOS stimulation and subsequent guanylyl cyclase activation that probably occurred in pericytes. Vasoconstriction required metabotropic glutamate receptors and CYP ω-hydroxylase, the enzyme regulating 20-hydroxyeicosatetraenoic acid production. Therefore, granular layer capillaries are controlled by the balance between vasodilating and vasoconstricting systems that could finely tune local blood flow depending on neuronal activity changes at the cerebellar input stage.

Significance statement: The neuronal circuitry and the biochemical pathways that control local blood flow supply in the cerebellum are unclear. This is surprising given the emerging role played by this brain structure, not only in motor behavior, but also in cognitive functions. Although previous studies focused on the molecular layer, here, we shift attention onto the mossy fiber granule cell (GrC) relay. We demonstrate that GrC activity causes a robust vasodilation in nearby capillaries via the NMDA receptors-neuronal nitric oxide synthase signaling pathway. At the same time, metabotropic glutamate receptors mediate 20-hydroxyeicosatetraenoic acid-dependent vasoconstriction. These results reveal a complex signaling network that hints for the first time at the granular layer as a major determinant of cerebellar blood-oxygen-level-dependent signals.

Keywords: NMDA receptor; capillaries; cerebellum; granule cells; neurovascular coupling; nitric oxide.

Figure 1.

Identification of capillary microvessels in the cerebellar granular layer. A, Examples of bright-field images of three capillaries in the cerebellar granular layer. Note that GrCs (*) are in close contact with nearby capillaries. Arrows indicate putative pericytes in close contact with the vessel wall. B, Confocal fluorescent images of cerebellar slices stained for IB4 (green), NG2 (red), and DAPI (blue) to reveal capillaries, pericytes, and cell nuclei, respectively. Ba, Top, Image showing a capillary surrounded by GrCs (*). Bottom, Same capillary (excluding DAPI) reveals the colocalization of IB4 and NG2 staining. The dashed line indicates the section analyzed for IB4-NG2 colocalization in Bb. Bb, Colocalization graph showing the overlap of the fluorescence signals originated by capillaries (green) and pericytes (red) staining. Bc, Granular layer capillary from another staining than in Ba, where the pericyte soma (arrow) and its bump-on-a-log morphology are visible in close contact with the capillary wall. C, Confocal acquisitions from a cerebellar slice stained for IB4 (green), NG2 (red), and DAPI (blue), as in Ba and Bc. Ca, Image showing the more regular microvessel organization in the molecular layer (ML) compared with the granular layer (GL) (20× objective). Cb, Image showing the granular layer (GL) and the MF bundle of another lamella at a higher magnification (40× objective).

Figure 2.

MF stimulation determines capillary vasodilation. A, U46619 perfusion restores the vascular tone in the granular layer capillaries. Aa, Average time course of the normalized capillary lumen diameter size in slices incubated for 1 h with U46619 (a thromboxane A2 agonist; 75 nm; n = 7). The gray line shows that capillary diameter did not change in the absence of U46619 perfusion. Ab, Image showing a capillary before (left) and after U46619 (right) perfusion. White bars indicate the location of lumen diameter measurements near the pericyte indicated by the arrow. B, MF stimulation causes vasodilation of preconstricted capillaries. Ba, Average time course of capillary diameter size (n = 13) before, during (gray filled rectangle) and after (black circles) MF stimulation (t = 0 s; 50 Hz); the gray trace shows the average time course of capillary diameter size in the absence of stimulation (n = 10) and the gray empty circles show the average time course of capillary diameter size in the presence of 4 μm TTX (n = 5) As a control, in five slices, a second stimulation (after 15 min) shows an average vasodilation not statistically different from the first one (inset). Bb, The same time course in Ba is shown on an expanded time scale to appreciate the initial dynamics of vessel dilation during MF stimulation. Note that vasodilation is evident as soon as 1 s after MF stimulation is started. Bc, Images showing capillary diameters before (left) and during (right) MF stimulation. White bars indicate the lumen diameter. In these and all subsequent experiments, capillaries were preconstricted using U46619.

Figure 3.

NMDARs, NO, and cGMP mediate vasodilation. A, Capillary vasodilation depends on NMDAR activation. Aa, Average time course of capillary diameter changes during the same MF stimulation protocol shown in Figure 2B in control (open circles) and during (filled circles) APV + 7ClKyn perfusion (NMDAR inhibitors; 100 and 50 μm, respectively; n = 7; the vessels diameter at the end of the control stimulation is renormalized to show the relative size changes after the stimulation in APV + 7ClKyn). NMDARs blockage turns the vasodilation into a vasoconstriction. Ab, Images showing capillary diameters before (left), during MF stimulation in control (center), and during MF stimulation in the presence of APV + 7ClKyn (right). White bars marks lumen diameter. B, NO production is necessary for vasodilation. Ba, Average time course of the capillary diameter changes during the MF stimulation protocol before (black empty circles) and after (filled circles) l-NAME perfusion (a NOS inhibitor; 200 μm, n = 9) and after l-NAME and 10 μm ODQ perfusion (gray empty circles). The stimulus-evoked vasodilation is turned into vasoconstriction in both cases. Note that, during the hour of l-NAME perfusion, the vessel diameter decreases (black circles, n = 8; compared with controls without l-NAME perfusion, gray line, n = 5) as shown in the inset. The vessel diameter relative size reported in the graph is therefore renormalized to the prestimulus condition in l-NAME. Bb, Images showing capillary diameter before (left) and after (right) MF stimulation in slices treated with l-NAME (1 h). White bars indicate the lumen diameter. C, sGC activation by NO is involved in vasodilation. Ca, Average time course of capillaries diameter size during the stimulation protocol before (empty circles) and after (filled circles) ODQ perfusion (an sGC inhibitor; 10 μm; n = 10). The blockage of sGC activity abolishes vasodilation that is partially reversed toward vasoconstriction. Cb, Images showing the same capillary before (left) and during MF stimulation in control (center) and in the presence of ODQ (right). White bars indicate the lumen diameter.

Figure 4.

DAF-FM fluorescence induced by synaptic activity in the granular layer. A, Average trace showing the peak in DAF-FM fluorescence in response to 4 s MF stimulation (n = 20 slices). B, Average traces of DAF-FM fluorescence signals evoked as in A in cerebellar granular layers before (black) and after 1 h of l-NAME perfusion (green; 200 μm; n = 7) and after APV + 7ClKyn perfusion (orange; 100 and 50 μm, respectively; n = 6). The superimposed traces are on the same amplitude scale as in A, whereas the time scale is dilated to appreciate the fluorescence peaks. C, Scatter plot showing the distances from the stimulating electrode of each vessel and DAF-FM fluorescence signal analyzed (filled circles). The empty circles show the average values of the single points.

Figure 5.

20-HETE and mGluRs mediate vasoconstriction. A, 20-HETE does not influence vasodilation. Aa, Average time course of capillary diameter size in cerebellar slices during the same MF stimulation protocol as in Figures 2B and 3 before (empty circles) and after 1 h of HET-0016 perfusion (a CYP ω-hydroxylase inhibitor; 1 μm; n = 4). The presence of HET-0016 does not change vasodilation significantly. Ab, Images showing the diameter variation of a capillary before (left) and during MF stimulation in the absence (center) or presence (right) of HET-0016. White bars indicate the lumen diameter. B, C, 20-HETE is the main vasoconstriction agent. Ba, Average time course of capillary diameter size in slices exposed to 1 h l-NAME during the same MF stimulation protocol as in A before (empty circles) and during (filled circles) HET-0016 perfusion (a CYP ω-hydroxylase inhibitor; 1 μm; n = 9). Note that the blockage of 20-HETE synthesis reduced vasoconstriction. Bb, Images showing the diameter variations of the same capillary after 1 h of l-NAME perfusion before (left) and during MF stimulation in the absence of (center) and in the presence of (right) HET-0016. White bars indicate the lumen diameter. Ca, Average time course of capillary diameter size in slices perfused with APV + 7ClKyn (as in Fig. 3A) during MF stimulation before (empty circles) and during (filled circles) HET-0016 perfusion. Cb, Images showing the changes in capillary diameter after perfusion with APV + 7ClKyn before (left) and during MF stimulation in the absence (center) or presence (right) of HET-0016. White bars indicate the lumen diameter. D, Capillary vasoconstriction depends on mGluR activation. Da, Average time course of capillary responses in cerebellar slices exposed to 1 h of l-NAME after the MF stimulation protocol in the absence (empty circles) or presence (filled circles) of MCPG + CPPG (mGluR blockers; 500 and 300 μm, respectively; n = 15). The perfusion of mGluR antagonists abolishes stimulus-evoked vasoconstriction. Db, Images showing the changes in capillary diameter before (left) and during MF stimulation in the absence (center) or presence (right) of MCPG + CPPG. White bars indicate the lumen diameter.