Layer-Specific Dilation of Penetrating Arteries Induced by Stimulation of the Nucleus Basalis of Meynert in the Mouse Frontal Cortex

Animal Preparation

Experiments were performed on seven adult male mice (C57BL/6NCr, 6 months old, 22 to 38 g). All animal experiments were conducted with the approval of and in accordance with the Guidelines for Animal Experimentation prepared by the Animal Care and Use Committee of Tokyo Metropolitan Institute of Gerontology and also National Institute of Radiological Sciences.

Long-term implantation of the stimulation electrode and constructing a cranial window were performed in animals anesthetized with pentobarbital (65 mg/kg), supplemented with 1% to 2% isoflurane in air if needed. The depth of anesthesia was assessed by testing the corneal reflex, and by monitoring breathing rates and body movements. A closed cranial window was made by removing a part of the skull (4 mm in diameter) over the left frontal cortex while leaving the dura intact, and the exposed cortex was sealed with a cover glass. ¹⁶ An additional hole, ~1 mm diameter, was drilled in the left parietal bone at 2.0 mm lateral to the midline and 3.2 mm posterior to the bregma to allow insertion of a coaxial metal stimulating electrode (outer diameter = 0.1 mm, USK-10, Unique Medical, Tokyo, Japan) at the angle of 45° from the vertical axis to the NBM. The stereotaxic coordinates of the tip of stimulation electrode were 0.9 mm posterior to the bregma, 2 mm lateral to the midline, and 3.5 mm vertical under the bregma height according to the atlas. ¹⁷ Implantation was performed at 1 to 12 days before the imaging experiment to avoid an influence of mechanical stimuli of cortical arteries during surgery and degradation of an implanted electrode for long-term implantation. Under these experimental conditions, no apparent influence of the time after surgery on the arterial responses to stimuli was observed.

Imaging of Cortical Vasculature by Two-Photon Microscopy

On the day of the imaging experiment, animals were anesthetized with urethane (0.8 to 1.1 g/kg, intraperitoneally or subcutaneously). Additional doses were administered as required to maintain the depth of anesthesia, which was assessed as described above. The trachea was cannulated and respiration was artificially maintained using a respirator (Minivent, Harvard, Holliston, MA, USA), set to 150 to 180 beats/min with a tidal volume of 150 to 180 μL. Rectal temperature was maintained at around 37.5°C using a heating pad system (TR-200, Fine Science Tools, Foster City, CA, USA). Mice were mounted on a stage for manipulation under the two-photon microscope, and the head was fixed on a stereotaxic instrument (SG-3N modified, Narishige, Tokyo, Japan).

To visualize the vasculature, rhodamine B isothiocyanate-dextran (MW 70 000, Sigma, St Louis, MO, USA) dissolved in saline (50 μL of 5% solution) was injected through the femoral vein just before the beginning of the imaging experiments. During imaging, sulforhodamine 101 (MP Biomedicals, Irvine, CA, USA) dissolved in saline (5 mM) was also injected intraperitoneally (4 μL/g body weight) as needed to intensify the signal, according to a previous study. ¹⁵ The imaging was conducted at National Institute of Radiological Sciences using a two-photon microscope (TCS-SP5MP; Leica Microsystems GmbH, Wetzlar, Germany) excited at 900 nm, as described previously.^15,18 An emission signal was detected through a band-pass filter (610/75 nm). We used a 20X water immersion objective (NA = 1.0, Leica Microsystems) to obtain a high-resolution image. A single image plane consisted of 1024-by-1024 pixels and the in-plane pixel size was 0.25 to 0.45 mm depending on an instrumental zoom factor.

Nucleus Basalis of Meynert Stimulation

Focal electrical stimulation of the NBM was performed by means of a stimulator (Master-8, A.M.P.I., Jerusalem, Israel) and stimulus isolation unit (ISO-Flex, A.M.P.I.). The NBM was electrically stimulated with repetitive rectangular current pulses of 30 to 50 μA intensity and 0.5 ms pulse duration. Direct current (50 μA) was applied for 30 to 60 seconds at the end of every experiment to localize the stimulated site. Mice were killed by an over dose of pentobarbital. The brains were removed, and histologic verification of the stimulating electrode was carried out using frozen transverse brain sections of 30 to 50 μm thickness.

Acquisition and Analysis of the Time Course of Changes in Arterial Diameter After Nucleus Basalis of Meynert Stimulation

Of the three types of penetrating arteries reported, ¹⁹ we focused on relatively large penetrating arteries that travel to deep cortical layers, to measure precisely the diameter of the arteries, and to analyze the depth dependence of diameter changes. It was shown that small diameter arteries had a territory of only the upper layer cortex. ¹⁹ A total of six penetrating arteries from five animals were used for analysis of the time course of response to the NBM stimulation by cross-sectional imaging ²⁰ at each cortical depth, including the cortical surface, in a 50-μm step. A train of 500 electrical pulses (50 Hz for 10 seconds) were repeated with an interval of 1 minute, according to a previous result showing that increased blood flow after 10-second stimulation of the NBM returns to control level within 20 to 30 seconds after the end of stimulation. ⁴ A cross-sectional image was sampled at a rate of 2.2 frames per second, and acquired for a period of 30 seconds starting at 5 seconds before the onset of each train stimulation. During the interval between sequential acquisitions, we changed imaging depth up and down in a randomized order. Repeated measurements were performed 3 to 6 times per each depth, and were averaged. A time course of changes in the vessel diameter was quantified offline as described previously. ²⁰ To temporally smooth the data, values obtained at 2.2 Hz were either averaged over three time points or smoothed with a time constant of 1 second.

Acquisition and Reconstruction of Volume Image During Nucleus Basalis of Meynert Stimulation and During Hypercapnia

Volume images were acquired in six mice, four of which were also used for analysis of the time course of the response described above. A total of 91 to 265 images were acquired with a z-step size of 2.5 to 4.0 μm, depending on the visibility of the measurement location in each animal. Accordingly, depths of 360 to 800 μm below the cortical surface were covered in all six mice. Total acquisition time was 4 to 10 minutes, and the frequency of the acquisition was 2.5 seconds per frame. Three volume images, (1) at the resting control condition without stimulation, (2) during electrical stimulation of the NBM, and (3) during moderate hypercapnia, were acquired in a single location in each animal. The image acquisition was started at least 10 seconds after the start of either NBM stimulation or hypercapnia inhalation to ensure that the responses to stimulation reached a plateau, while the stimulation continued throughout the image acquisition for 5 to 15 minutes. Intermittent train pulse stimulation (1 second on/1 second off, 50 Hz) to the NBM was used because such stimulation to the NBM has been reported to sustain an increase in cortical blood flow for long-lasting stimulation.^12,21,22 Moderate hypercapnia was applied by introducing 3% CO₂ with air balanced gas in addition to the inspiratory room air. All volume images obtained were reconstructed to 3D volume images offline with LAS AF software (Leica Microsystems GmbH) or Imaris software (Bitplane, Zurich, Switzerland). For visual purposes, 2D maximum intensity projection image was obtained for the z-stack images. In each single slice image, a region of interest captured the target arteries, and the cross-section diameter for both penetrating and surface arteries was automatically determined, as described previously. ²⁰ Due to an uncertain circular edge of the target artery at the point of its branching in the image, the measurements of the vessel diameters around the branching locations (<60 μm length) were not included for further analysis. The values measured for diameters of the penetrating arteries were averaged at depths of every 100 μm, whereas those of the surface artery (a depth of 0 μm) were averaged across the site of the measured locations along the single artery. ²⁰

Statistical Analysis

For statistical analysis, Prism 6 software (GraphPad Software, La Jolla, CA, USA) was used. The changes in diameter during stimulation across different depths were analyzed by two-way analysis of variance (ANOVA) with repeated measures between diameter before (control) and during stimulation, followed by Fisher’s least significant difference test. The magnitude of the response by hypercapnia and that by NBM stimulation was compared by two-way ANOVA followed by least significant difference test. The time courses of changes in diameter induced by NBM stimulation were assessed separately for each depth by one-way repeated measures ANOVA followed by Dunnett’s multiple comparisons test. Two-way ANOVA with repeated measures at different time points was performed to compare the time course of the response between upper and lower layers. The statistical significance level was set at 5%. Data were expressed as mean ± standard error of the mean, unless otherwise stated.

RESULTS

We followed single penetrating arteries from an offshoot of a pial artery to a maximum cortical depth of 800 μm below the surface. The images in Figure 1A represent a typical 3D image of measured vasculature reconstructed as 2D horizontal projection images of every 100-μm thickness (i.e., maximum intensity projection for visual purposes) at a resting control condition. At the resting condition, the diameters of eight penetrating arteries measured in seven mice ranged from 9.6 and 27.9 μm over the cortical depths. They branched at depths of ~300 μm (range, 270 and 310 μm; see Figure 1A), 500 μm (range, 440 and 534 μm), and >600 μm (deeper cortex). These branching points were not included in the diameter measurements due to a blurred circular edge of the penetrating artery. Nevertheless, there was no clear correlation between the magnitude of dilation and branching location. For vessel comparisons among different imaging sessions, a depth location of the branching point was used as anatomical references to identify the same location of the vessel segment. The images in Figure 1B represent a longitudinal view of the same penetrating artery as shown in Figure 1A, compared across control, NBM stimulation, and hypercapnia conditions. Enlargement of the penetrating artery was induced consistently in all animals during both NBM stimulation and hypercapnia. However, the longitudinal length of the penetrating artery was unchanged across control, NBM stimulation, and hypercapnia conditions. In our experimental conditions, a defocus of the image due to a change of the blood volume in the parenchymal tissue was not recognizable across the three imaging conditions, compared with the depth resolution of the two-photon microscopy (< ~2 to 3 μm). This was seen consistently in our time-resolved image in which no detectable changes of the cross–section views of the capillaries distributed within the image field were observed between for resting and stimulation periods, and where only penetrating arteries showed a stimulation-induced increase in the cross-sectional area.

OPEN IN VIEWER

Figure 1

Typical 3D imaging of cortical microvasculature, shown as 2D reconstructed images. (A) Maximum intensity projection images of horizontal sections at the resting control condition reconstructed from 100-μm digital sections. Arrows indicate branching points of a penetrating artery. The first branch was seen at 285 μm and the second branch was at 534 μm in depth. (B) Longitudinally rendering images of a penetrating artery (100 μm thick) at the resting control condition, during stimulation of the nucleus basalis of Meynert (NBM) (0.5 ms, 30 μA, 50 Hz, 1-second on/1-second off), and during hypercapnia (induced by inhalation of 3% CO₂ in air).

Cross-sectional imaging of the penetrating artery, except for its branching point, allowed for the quantification of temporal dynamics of the enlarged diameter induced by 10-second NBM stimulation (Figure 2A). Figure 2B represents the diameter change of that penetrating artery at eight different depths including the surface location in response to NBM stimulation (10 seconds, 50 Hz, 30 to 50 μA). The artery showed only a marginal increase in diameter at the cortical surface after stimulation of the NBM in accordance with the previous report. ¹¹ In contrast, the artery at a depth of 100 μm from the surface showed an obvious increase in diameter during stimulation of the NBM. The increase in the diameter started within 1 second of the onset of NBM stimulation. Increases in the diameter were also observed at other cortical depths up to 700 μm from the cortical surface, except for a depth of 500 μm where the diameter scarcely increased during stimulation. The response was reproducible in successive trials in the same depth locations. Enlargement of penetrating arteries was observed selectively without apparent changes in other vascular structures, such as veins and capillaries, in the same in-plane image.

OPEN IN VIEWER

Figure 2

Representative dynamic changes in the diameter of a penetrating artery in response to nucleus basalis of Meynert (NBM) stimulation (0.5 ms, 30 to 50 μA, 50Hz, for 10 seconds). (A) Representative images of a penetrating artery at a depth of 250 μm before (left) and during (right) stimulation of the NBM. Signal intensity levels are shown as an 8-bit color scale (indicated right). (B) Temporal dynamics of diameter changes at various depths, including the surface for the single penetrating artery. Original values of the diameter (a scale indicated right) measured at 2.2 Hz (gray line) are temporally smoothed with a time constant of 1 second (black line).

Figure 3 summarizes the arterial responses to stimulation of the NBM analyzed quantitatively in six penetrating arteries in five mice. Responses to NBM stimulation in these six arteries were similar to the results shown in Figure 2B, indicating a depth-dependent heterogeneity. The relationship between cortical depth at every 50 μm step and changes in artery diameter in response to NBM stimulation is presented in absolute values (Figure 3A) and percentages of prestimulus control values (Figure 3B). At a depth of 50 μm, the artery diameter of 15.6±1.6 μm before stimulation increased to 17.6±1.7 μm during NBM stimulation, representing an increase of 2.0 μm or 12.8% of the prestimulus control diameter. A significant increase in the diameter of penetrating arteries was observed at depths of 50 to 400 μm and 550 to 600 μm. Increases in diameter during NBM stimulation at these depths ranged between 9% and 13%. However, the increase in diameter during the stimulation was not statistically significant at depths of 450 and 500 μm. A depth-dependent reduction in the response magnitude was observed at depths between 350 and 500 μm. The relationship between cortical depth and changes in artery diameter in response to NBM stimulation is also presented in percentages of the maximum change in each artery (Figure 3C). The change in the diameter of penetrating arteries at depths of 50 to 350 mm and 550 to 600 mm was significantly larger than that at the surface. However, change in the diameter of penetrating arteries at depths of 400 to 500 mm was not significantly different from that at the surface. This result indicates that there was a similar trend of depth dependency in all arteries measured.

OPEN IN VIEWER

Figure 3

Relationship between cortical depth and changes in diameter of the artery during nucleus basalis of Meynert (NBM) stimulation (0.5 ms, 30 to 50 μA, 50 Hz, for 10 seconds). Changes in diameter during stimulation are expressed as an absolute value (A), as a percentage of prestimulus control diameter (B), and as percentage of the maximum change in each artery (C). The mean value for 5 seconds before the onset of stimulation was used as a prestimulus control diameter (gray box in A). The maximum value of temporally smoothed diameter during stimulation for 10 seconds was used as diameter during stimulation (white box in A). (A, B) Each column and vertical bar represents the mean ± s.e.m. The numbers in parenthesis indicate the number of arteries measured. ^*P<0.05, ^**P<0.01; significantly different from prestimulus control diameter. (C) Each point indicates individual data. Mean value at each depth was connected by a line. Vertical bar indicates s.d. Significant differences from the response at the surface are tested by one-way factorial analysis of variance (ANOVA) followed by least significant difference (LSD) test. _#P<0.05, ^#P<0.01.

The time course of diameter changes measured during NBM stimulation are summarized in Figure 4A and are also shown by all individual data in Figure 4B. In these graphs, the response of the penetrating arteries at each 100 μm depth are combined for simplification of the results. At a depth of 50 and 100 μm, the diameter was maximal at 5 seconds after the onset of stimulation, and reached 111% of the control diameter (Figure 4A). The increased diameter subsided gradually during the course of stimulation. After the cessation of stimulation, the diameter returned to prestimulus control levels, and then increased again slightly. The diameter remained slightly larger than baseline at 20 seconds after stimulus onset. However, we ensured that it had returned to prestimulus baseline at 30 seconds later (i.e., an onset of next image acquisition) before starting the next stimulus session. The time course of diameter changes after stimulation was similar from upper layers to a depth of 400 mm below the cortical surface, whereas the response at lower layers showed delayed responses. At a depth of 450 and 500 μm, the onset and peak latencies were delayed by > 1 second compared with that at a depth of 50 and 100 μm. In addition, the amplitude of the maximum response in the deeper layer was approximately half of that in the upper layer. At a depth of 550 and 600 μm, NBM stimulation induced a biphasic response, an initial slight decrease, and a subsequent large increase in the diameter during stimulation, with a peak time delayed by 3 seconds compared with that at a depth of 50 and 100 μm. The individual data showed consistently that there was a similar trend of depth dependency in all arteries measured (Figure 4B).

OPEN IN VIEWER

Figure 4

Summarized time course of dynamic changes in diameter of a single penetrating artery response to the nucleus basalis of Meynert (NBM) stimulation at different depths of every 100 μm. Change in diameter is expressed as percentage of the prestimulus value (ordinate). The dashed lines and the heavy bar on the abscissa indicate the time during which the NBM was stimulated (0.5 ms, 30 to 50 μA, 50 Hz, for 10 seconds). The onset of electrical stimulation of the NBM is expressed as zero (abscissa). (A) Each point represents a mean ± s.e.m. Significant differences from prestimulus control diameter are indicated by a (P<0.05) and b (P<0.01). (B) Each line indicates individual data.

Figure 5 summarizes the diameter changes after NBM stimulation as an average of the responses across upper depths of 50 to 400 μm and lower depths of 450 to 600 μm, and compares the blood flow response induced by NBM stimulation, measured by laser speckle flowmetry (a reanalysis of previous data performed under identical experimental conditions and stimulation paradigms used in the present study ⁴ ). The time course of the diameter response of the upper and lower layer was significantly different (by two-way repeated ANOVA, Figure 5A). Further, a comparison of the diameter changes with flow measurements showed that a peak latency of enlargement of the penetrating artery was earlier than that of the blood flow response, at both upper and lower layers of the cortex. The increase in diameter at the upper layer precedes the blood flow response by > 1 second for onset latency and > 5 second for peak latency.

OPEN IN VIEWER

Figure 5

Changes in the diameter of single penetrating arteries (A) and accompanied changes of parenchymal blood flow (B) in the frontal cortex after nucleus basalis of Meynert (NBM) stimulation. Data are expressed as percentages of prestimulus control values. (A) Data from six arteries in five mice at upper layers (depths of 50 to 400 μm, n = 43) and lower layers (depths of 450 to 600 μm, n = 14) were summarized. Significant differences from prestimulus control diameter are indicated by a (P<0.05) and b (P<0.01). ^**P<0.01; significantly different interaction by two-way repeated analysis of variance (ANOVA). See Figure 4 for other details. (B) Time courses of the blood flow response measured by laser speckle flowmetry, acquired in a time constant of 1 second, of the previous study, ⁴ were resampled on the region of interest at frontal area every 1 second (mean ± s.e.m., n = 10 trials in 5 mice). Experimental conditions including anesthesia, respiration, and parameters of NBM stimulation were the same as those in the present study. Note that obvious delays of the response onset and sustained increases in parenchymal blood flow compared with the arterial responses shown in panel A.

The arterial response to NBM stimulation was compared with that to hypercapnia by quantitative analysis of volume images in six arteries obtained from six mice (Figure 6). Changes in diameter during NBM stimulation were similar to the result obtained by the dynamic imaging performed every 50 μm depth for time-course measurements shown in Figure 3, which consistently indicated a depth-dependent heterogeneity. In contrast, in the case of hypercapnia, obvious increases in the diameter of arteries were observed over all cortical layers (Figures 6A and 6B), including at the surface and a depth of 400 to 500 μm where NBM stimulation did not produce significant diameter changes (Figure 6C). The relationship between cortical depth at every 100-mm step and changes in artery diameter in response to hypercapnia is presented in absolute values (Figure 6A) and percentages of prestimulus control values (Figure 6B). A significant increase in diameter was observed at all depths measured up to 500 μm below the cortical surface. In these penetrating arteries measured at each cortical depth, increased diameter during hypercapnia reached around 20 μm irrespective of control diameter at rest. The magnitude of the increase in diameter during hypercapnia was similar from surface to a depth of 300 μm, such as 12.2 ± 5.4% at the surface and 10.5 ±2.4% at a depth of 100 to 200 μm. The increase tended to be larger in deeper layers, such as 24.8 ± 10.1% at a depth of 400 to 500 μm. The depth-dependent increase in the response magnitude was observed at depths between 300 and 500 μm, and the magnitude of the diameter response by hypercapnia was significantly larger than that by NBM stimulation at a depth of 400 to 500 mm.

OPEN IN VIEWER

Figure 6

Relationship between cortical depth and changes in diameter of the artery during hypercapnia (A, B) and nucleus basalis of Meynert (NBM) stimulation (C). Changes in diameter during stimulation are expressed as an absolute value (A) and as a percentage (B, C) of corresponding control diameter at rest (without any stimuli), over every 100-mm depth. Each column represents the mean±s.e.m. The numbers in parenthesis indicate the number of arteries measured. ^*P<0.05, ^**P<0.01; significantly different from control diameter. Significant differences between diameter response to hypercapnia and to NBM stimulation are indicated by a (P<0.05).

DISCUSSION

With high-resolution two-photon imaging of single penetrating arteries in the cortical parenchyma (Figure 1), we report that electrical stimulation of the NBM produces dilation in the cortical penetrating arteries. The temporal dynamics of the arterial diameter response was quantified, which showed a different time course and magnitude of the dilation in a depth-dependent manner (Figures 2 to 4), and also differed from the previously known time course of flow responses (Figure 5). These layer-specific dilations of the penetrating arteries to the NBM stimulation were verified further by comparing the response with hypercapnia inhalation (Figure 6).

The magnitude of the changes in penetrating arteries, ~ 11% of the basal diameter, was well in accord with the magnitude of the increase in diameter of parenchymal microvessels at 60 μm below the cortical surface measured histologically using electron microscopy (mean inner diameter was 4.9 μm in nonstimulated control rats and 5.5 μm in NBM-stimulated rats, approximately a 12% increase ¹² ). Dilation of the penetrating artery would result in a similar extent of enlargement in microvessels connected to the artery. Although the present result does not exclude the possibility of active dilation of branched smaller arteries and capillaries after NBM stimulation, the present study clearly shows that NBM stimulation, which affects the surface arteries to a lesser degree, induces active dilation of the penetrating artery that drives an increase in cortical blood flow. Furthermore, the present result of a more rapid response of the penetrating arteries in the upper layers than in the lower layers during NBM stimulation supports a regulation of blood flow initiated in penetrating arteries irrigating a larger volume of the tissue, rather than a possible regulation exerted at the level of the capillaries propagated upstream to microarteries. ²³

The fronto-parietal cortical parenchymal blood vessels, including capillaries and arteries, are in close contact with cholinergic nerve terminals originating in the basal forebrain.^1,2 Individual cholinergic neurons have axonal arbors that ramify through nearly the full thickness of the cortex extending across >2 mm² of the mouse cortical surface. ²⁴ However, layer heterogeneity of the density of nerve terminals from the NBM has been shown.^1,13 In the frontal cortex, major projections with a relatively high density of terminal boutons are observed in layers I, II, and V or VI in rats^1,13 and layers I to III and VI in mice (Figure 6E of Rotolo et al. ²⁴ ). Our results indicate depth-related differences in arterial response to the NBM stimulation. In the mouse frontal cortex, the thickness of layers I, II/III, and V to VI corresponds to the approximate cortical depths of 50, 400, and 700 μm, respectively, according to our postmortem histologic analysis (not shown) and a previous report. ²⁵ Therefore, our present observation of parenchymal vasculature was superficial to layer V. Depths of 50 to 400 μm, where responses of large amplitude with the earliest onset were observed, correspond to layers I to III. Depths of 550 to 600 μm, where a similar amplitude, but delayed response was observed, correspond to the middle of layer V. Depths of 450 to 500 μm, where the amplitude of response to NBM stimulation, but not to hypercapnia was much smaller than that at other areas, correspond to the upper part of layer V. Our results appear to relate to the layer heterogeneity of density of nerve terminals from the NBM. Future studies should colocalize the effector site of the vasodilation and terminals of cholinergic fibers originating from NBM using transgenic mice in which target neurons are labeled specifically with genetically expressed fluorescent proteins.

In addition to the direct innervation of parenchymal vasculature, fibers from the NBM terminate on cortical neurons and astrocytes. Partial involvement of cortical pyramidal neurons, interneurons, and astrocytes to the blood flow response during basal forebrain stimulation has been suggested.^26,27 Their studies, using cFos protein as a marker for increased neuronal activity, showed that both excitatory pyramidal neurons and inhibitory interneurons are recruited by basal forebrain stimulation in rats. Activated neurons are distributed throughout the different layers of the ipsilateral cortex, except for pyramidal neurons in layer V.^26,27 Our result does not exclude the possibility of a partial contribution of such neuronal activity or astrocytes to the dilation of penetrating arteries. Neuronal response to acetylcholine has also been shown to have layer heterogeneity. The delay of the vasodilative response in layer V may be related to the acetylcholine-induced hyperpolarization of pyramidal neurons preferentially observed in layer V. ²⁸

In addition to the NBM stimulation, we also examined the effect of moderate hypercapnia (induced by inhalation of 3% CO₂), which is well known to increase cortical blood flow. The gentle increase in PaCO₂ (PaCO₂ = 45 ± 7 mm Hg) has been shown to be sufficient to produce robust increases in arterial diameter of the parenchyma measured up to a 200-mm depth of rat’s cortex. ²⁹ We confirmed an obvious increase in diameter of penetrating arteries during hypercapnia, over all layers examined (Figure 6). Thus, the animal’s physiologic condition in the present experiment seems appropriate to detect vasodilation in the cortical parenchyma. Moreover, it is noteworthy that the layer dependence of the arterial responses by NBM stimulation was different from that by hypercapnia. This suggests that the regional difference in reactivity to NBM stimulation is specific to this system, and not related to differences in the cortical vasoreactivity itself. This result further supports our implication that layer dependency of the arterial response to NBM stimulation appears to be due to layer heterogeneity of the density of cholinergic nerve terminals from the NBM.

Layer heterogeneity of the arterial response induced by NBM stimulation might be important for the survival and function of cortical neurons. Dilation of specific penetrating arteries, known as a ‘bottle neck’ of cortical blood flow regulation, ³⁰ may produce effective increases in regional blood flow to a specific neural column of the neocortex. Enlargement of the penetrating artery at upper layers of the cortex, where both dilation onset and peak latency was earlier than that of blood flow response reported previously (Figure 5), may contribute primarily to the increase in cortical parenchymal blood flow due to NBM stimulation. The difference of time course between responses to NBM stimulation of penetrating arteries and of parenchymal blood flow appears to be explained by considering the capacity of the downstream microvascular bed. Furthermore, the present finding of predominant vasodilation in layers I to III would produce spatially distinct specific increases in blood flow to the arteries at around 300 μm depth.

Alzheimer’s disease, known to be quite striking degeneration of cells in the basal forebrain, also shows layer-specific cortical pathologic changes. In the neocortex, mainly neurons in layers III and V are affected.^31,32 It is unknown whether the layer heterogeneity of neuronal vulnerability is due to intrinsic properties of cortical neurons or due to environmental factors. Layer heterogeneity of the arterial response to NBM stimulation found in the present study may be an environmental factor. Pathologically, the microstrokes observed in human patients are often centered around penetrating arteries,^33,34 in which cerebral ischemia was most severe when a clot was located among three different levels in rats (surface artery, penetrating artery, and deep microvessels). ³⁰ In our previous study, we found that NBM stimulation protects ischemia-induced delayed neuronal death after transient occlusion of a large artery. ²² Additionally, we predict from the present finding that activation of NBM may protect against ischemia due to occlusion of the penetrating artery, such as by micro thrombus.

CONCLUSION

Our results show in mice that stimulation of the NBM produces dilation of penetrating arteries of the cerebral cortex in a layer-specific manner. In contrast, obvious vasodilation comparable to those measured over the cortical layers, from surface of the cortex to a depth of 500 μm, was induced by hypercapnia. Together, these findings indicate that the layer-specific dilation of penetrating arteries represent a layer specificity of cholinergic nerve terminals from the NBM.