Regional Blood Flow in the Normal and Ischemic Brain Is Controlled by Arteriolar Smooth Muscle Cell Contractility and Not by Capillary Pericytes

Abstract

The precise regulation of cerebral blood flow is critical for normal brain function, and its disruption underlies many neuropathologies. The extent to which smooth muscle-covered arterioles or pericyte-covered capillaries control vasomotion during neurovascular coupling remains controversial. We found that capillary pericytes in mice and humans do not express smooth muscle actin and are morphologically and functionally distinct from adjacent precapillary smooth muscle cells (SMCs). Using optical imaging we investigated blood flow regulation at various sites on the vascular tree in living mice. Optogenetic, whisker stimulation, or cortical spreading depolarization caused microvascular diameter or flow changes in SMC but not pericyte-covered microvessels. During early stages of brain ischemia, transient SMC but not pericyte constrictions were a major cause of hypoperfusion leading to thrombosis and distal microvascular occlusions. Thus, capillary pericytes are not contractile, and regulation of cerebral blood flow in physiological and pathological conditions is mediated by arteriolar SMCs.

Figure 1. Distinct morphology of mural cells along the cortical vascular tree

(A) In vivo image captured from the cortex of an NG2cre:mT/mG transgenic mouse showing mGFP (green) expressing smooth muscle cells on arterioles and venules with mTomato (red) expression in all membranes. (B) In vivo images showing detailed morphology of mGFP labeled cells on pre- and post-capillary vessels (left), pre-capillary arterioles covered by single smooth muscle cells (middle, arrowheads), and capillaries covered by a single pericyte (right, arrow), numbers indicate vessel diameter. (C) In vivo two photon image showing mGFP expression on a penetrating arteriole and several branches (branch order indicated) demonstrating the transition from smooth muscle cell morphology (arrowheads) to capillary pericytes (arrows). Vessels are labeled via intravascular dye. (D–E) Images showing transitions from mGFP-labeled smooth muscle cell morphology (arrowheads) to capillary pericytes (arrows). (F) Image captured from the cortex of an NG2creER:mT/mG transgenic mouse showing isolated mGFP labeled cells after a single injection of tamoxifen. (G) Single cell circumferential morphology of smooth muscle cells on arterioles (top), pre-capillary arterioles (middle) and post-capillary venules (bottom), numbers indicate vessel diameter. (H) Single cell morphology of capillary pericytes extending longitudinally over hundreds of microns (arrows). See also Figure S1 and S2.

Figure 2. Awake imaging of spontaneous vasomotion along the cortical vascular tree

(A) In vivo images captured from the cortex of awake head fixed NG2cre:ZEG transgenic mice showing GFP (green) labeled smooth muscle (arrowheads), pericytes (white arrows), and NG2 cells (blue arrows) while vessels are labeled via intravascular dye (red). (B) Line-scan time series showing spontaneous diameter changes of the vessels shown in (A). (C) Diameter changes of four vessel segments with (magenta, green, orange) or without (blue) smooth muscle cells shown in (A–B). (D) Smooth muscle cells on a pre-capillary arteriole (arrowheads) and pericytes on capillaries (arrows) (E) Line-scan time series showing spontaneous diameter changes of the vessels shown in (D). (F) Diameter changes of the four vessel segments with (magenta and green) or without (orange and blue) smooth muscle cells shown in (D–E). (G) The relationship between vessel diameter and a vasomotion index defined as the area under the curve for percent spontaneous changes in vessel diameter with a ±5% cutoff threshold, n=47 vessel segments from 4 mice. See also Movie S1

Figure 3. Correlation between vasomotion and calcium transients in mural cells in vivo

(A) In vivo image from the cortex of an NG2cre:GCaMP3 transgenic mouse showing a penetrating arteriole and the transition from SMC covered (1 and 2) to pericyte covered (3) vessels. Time lapse sequences show the change in vessel diameter (dotted line) and corresponding changes in GCaMP3 fluorescence (arrowheads). (B) Traces of spontaneous changes in vessel diameter (red) overlayed with corresponding changes in GCaMP3 calcium fluorescence (green) in SMC covered arterioles and pericyte covered capillaries. (C) Time lapse sequence showing spontaneous GCaMP3 calcium changes in a single pericyte cell body (arrow) and processes. (D) Traces of changes in vessel diameter (red) overlayed with changes in pericyte GCaMP3 calcium signals (green). Arrows indicate changes in pericyte calcium with no changes in capillary diameter. (E) Example images and line-scan time series at each level of the vascular tree. Changes in vessel diameter occurred only on SMC covered vessels which corresponded to decreases in SMC GCaMP3 calcium signals (arrowheads). In contrast, peaks in pericyte calcium (arrow) correlated with no change in capillary diameter. (F) Quantification of the correlation between GCaMP3 calcium signals and changes in vessel diameter in SMCs (red) r = −0.49 ± 0.21 SD, and pericytes (blue) r = 0.02 ± 0.07 SD, n = 24 SMC covered vessels and 12 pericyte covered vessels from 3 mice, * indicates significant differences from 0 correlation determined by a 99% confidence interval. See also Movies S2 and S3

Figure 4. Smooth muscle actin is expressed by arteriole but not capillary mural cells in mouse and human neocortex

(A) In vivo image captured from the cortex of a smooth muscle actin (SMA)-mCherry:Tie2-GFP transgenic mouse showing GFP (green) labeling in all vessels and bright mCherry (red) expression in arterioles and pre-capillary vessels. (B) A single GFP labeled arteriole enveloped by a layer of SMA-labeled mCherry. (C) Branching of a pre-capillary vessel showing the termination of SMA-mCherry expressing cells (arrowheads) with vessel diameter indicated adjacent to single vessels. (D) SMA-mCherry expression in surface and penetrating arterioles with vessels labeled via intravascular dye (cyan). Vessel branch orders indicated in yellow. (E) High magnification image of the vessel outlined in (D) showing the termination of the last SMA-mCherry expressing cell (arrowheads) on a 4^th branch order vessel, numbers in lower panel indicate vessel diameter. (F–G) SMA-mCherry expression on pre-capillary vessels at different vascular tree levels. The last SMA-mCherry expressing cells (panels on the right indicate zoomed regions of terminal SMA+ cells shown on the left) are indicated with cell bodies (arrows) and terminal processes (arrowheads), numbers indicate vessel diameter. (H) Correlation of vessel branch order with that of the distribution of vessel diameters where the last SMA-mCherry expressing smooth muscle cell is located. (I) The percentage breakdown for the branch order on which terminal SMA+ cells are located. (J) Correlation of vessel diameter with the distance from a vessel branch point where the last SMA-mCherry expressing smooth muscle cell is located. (K) Immnohistochemical staining of NG2cre:ZEG mouse necortex with antibodies for SMA (red) and PDGFRβ (white) showing that SMA locates only at SMCs on arterioles and pre-capillary arterioles (yellow arrowheads) while PDGFRβ is expressed by both arteriolar SMCs and capillary pericytes (yellow arrows). GFP labeled cells in the parenchyma (blue arrows) are oligodendrocyte lineage cells. (L) Immunohistochemical staining of human neocortex with antibodies for collagen IV (white) to label blood vessels and SMA (red) showing the transition from SMA-covered pre-capillary arterioles to SMA-lacking capillaries (arrowheads), numbers indicate vessel diameter. See also Figure S3.

Figure 5. Single-cell optogenetic activation causes vessel constriction and decreases blood flow in smooth muscle but not pericyte-covered microvessels

(A) In vivo images captured from the cortex of an NG2cre:ChR2-YFP:SMA-mCherry mouse showing SMA (red) expression on arterioles and pre-capillary arterioles and Chr2-YFP expression (green) on all vessels and a subpopulation of astrocytes. (B) Line scan time series images showing changes in vessel diameter after line scan activation of ChR2 at various levels of the vascular tree with and without SMA (indicated by numbers with location identified with white lines (A)). (C) Paired examples showing locations of ChR2 activation (white lines) on vessels with and without SMA with corresponding line-scan time series images demonstrating that only SMA covered vessels show vessel constriction. (D) The average vessel diameter change in response to ChR2 activation (at time 0) on vessels with smooth muscle (arterioles – red line and pre-capillary arterioles – orange line) and without smooth muscle (capillaries – blue line). Traces show mean values ± SEM, n = 58 arterioles, 18 pre-capillary arterioles, and 39 capillaries from 8 mice. * indicates significance determined by a 99% confidence interval. (E) The distribution of vessels that showed contractility in response to ChR2 activation in relation to vessel diameter. (F) Method for determination of changes in blood flow velocity using the distortion of the image of a red blood cell as it moves (dotted line) during laser scanning (relative velocity field scanning). (G) A decrease in vessel diameter and blood flow on a SMC covered vessel after optogenetic activation. (H) No change in blood flow velocity after optogenetic activation in a pericyte covered vessel. (I) Quantification of the average change in blood flow velocity during optogenetic activation in SMC+ pre-capillary arterioles (red line) and SMC− capillaries (blue line). Traces show mean values ± SEM, n = 10 pre-capillary arterioles and 30 capillaries from 3 mice.* indicates significance determined by a 99% confidence interval.

Figure 6. Smooth muscle but not pericyte-covered microvessels undergo active sensory-evoked vasodilatation in awake mice

(A–B) In vivo images captured from the somatosensory barrel cortex of awake head fixed NG2cre:GCaMP3 transgenic mice showing GCaMP3 (green) baseline fluorescence labeling of smooth muscle (arrowheads) and neurons (blue arrows) but very little in capillary pericytes (white arrows). Vessels are labeled via intravascular dye (red). (C) A single neuronal calcium response (box) to a whisker puff detected by a fluorescence intensity increase of the GCaMP3 signal. (D) Trace showing the percent change in neuronal GCaMP3 fluorescence to five repeated whisker puffs. (E) Image showing examples of GCaMP3 positive and negative vessels analyzed from time lapse images during thirty seconds of whisker stimulation. (F) Line-scan time series images showing the GCaMP3 neuronal response and whisker evoked dilation (arrows) of the vessels shown in (E). (G) Quantification of the neuronal and vascular responses to whisker stimulation (black bar) with the average neuronal GCaMP3 fluorescence signal (green), the average change in diameter of pre-capillary vessels with a smooth muscle layer (SMC+, red), capillaries lacking a smooth muscle layer (SMC−, blue), and GCaMP3 fluorescence signal in pericytes (magenta) and SMCs (orange). Traces show mean values ± SEM, n = 57 SMC+ vessel segments, 83 SMC− vessel segments from 4 mice for neuronal response and vessel diameter and n = 24 SMCs and 12 pericytes from 3 mice for GCaMP3 traces. * indicates significance compared to baseline determined by a 99% confidence interval coded by color for corresponding cell type (SMC or pericyte) and signal (vessel diameter or GCaMP3).

Figure 7. Cortical spreading depolarization induces vasomotion of smooth muscle but not pericyte-covered microvessels

(A) In vivo image captured from the cortex of an SMA-mCherry (red) transgenic mouse with neurons and astrocytes labeled with the Oregon Green Bapta-1AM (OGB-1AM) calcium indicator (green) and vessels labeled via intravascular dye (blue). (B) Region depicted in (A) of the transition from pre-capillary SMA covered vessel (arrows) to capillary pericyte covered vessel (arrowheads), numbers indicate baseline vessel diameters. (C) Time sequence showing the OGB-1AM fluorescence intensity and vessels at baseline, during KCl induced spreading depolarization, and during the recovery period; numbers indicate vessel diameters. (D) Line-scan time series images showing KCl induced change in vessel diameter in an SMA+ vessel but not in the downstream SMA− vessel. (E) Quantification of the neuronal and vascular responses to spreading depolarization (gold bar) with the average neural OGB-1AM fluorescence signal (green), the average change in diameter of pre-capillary vessels with a smooth muscle layer (SMA+, red), and capillaries lacking a smooth muscle layer (SMA−, blue). Traces show mean values ± SEM, n = 10 SMA+ vessels and 10 SMA− vessels from 3 mice. * indicates significance compared to baseline determined by a 99% confidence interval.

Figure 8. Transient MCA occlusion induces focal SMC constriction and reperfusion deficits

(A) Timeline for experimental transient (90 min) middle cerebral artery occlusion (tMCAO) with in vivo imaging. In vivo image from an SMA-mCherry transgenic mouse (red) with intravascular dye (green) showing an example of a transition from SMC to pericyte covered vessels depicted in B. (B–C) Repeated in vivo images captured before, during, after and 4 hours after tMCAO showing prolonged (white arrowheads) or transient (blue arrowheads) SMC focal constrictions with maintained subsequent vessel blockage in downstream capillaries. Right side panels in B are zoomed from boxed region. (D–E) Example vessel perfusion outcomes (blocked, plasma perfused, or RBC perfused) after SMC focal constrictions (arrowheads) due to tMCAO (F) Diagram showing the measurement locations on the vascular tree at the transition points between SMCs and pericytes for quantification in G–I. (G) Vessel perfusion outcomes 4 hours after tMCAO for all SMC to pericyte transitions, n = 59 transitions from 5 mice. (H) Average vessel diameter measurements ± standard deviation at each location along the vascular tree before, during, after and 4 hours after tMCAO specifically in blocked, non-collapsed vessels, n is indicated in table from 5 mice, p values obtained from unpaired t-test. (I) Analysis of vessel perfusion during, after and 4 hours after tMCAO at distinct locations along the vascular tree, n is indicated from 5 mice. See also Figures S6–S8, S10 and Movies S4–6