| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Department of Neurology and 2Laboratory of Neuro Imaging, David Geffen School of Medicine at University of California, Los Angeles, California
Submitted 9 January 2007; accepted in final form 21 February 2007
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
). These changes were confirmed by Van Harreveld and others, in rabbit, rat, and cat (van Harreveld and Ochs 1957; Wahl et al. 1987). Later studies using laser Doppler flowmetry focused on tissue perfusion rather than arterial diameter (Fabricius et al. 1995), but also came to the conclusion that CSD is associated with an initial increase in cortical parenchymal blood flow that was believed to correspond to arteriolar dilatation. Another long-standing observation is that CSD is followed by a sustained hypoperfusion (Lauritzen et al. 1982), concurrent with a reduction in reactivity to vasoactive stimuli. Arteriolar reactivity is reduced from 20 to 90 min to high and low pH, elevated potassium (Seitz et al. 2004), adenosine, bradykinin (Wahl et al. 1987), elevated CO2 (Lacombe et al. 1992), and papaverine (Florence et al. 1994). Multiple different mechanisms of CSD-mediated arteriolar changes have been proposed. Possible dilators include CGRP (calcitonin gene-related peptide), nitric oxide (NO; Colonna 1994), and parasympathetic innervation by acetylcholine (Reuter et al. 1998). Possible constrictors include prostanoids (Shibata et al. 1991) and potassium, although potassium also dilates cerebral vessels at lower concentrations (Horiuchi et al. 2002).
A common assumption is that the arteriolar changes accompany or follow the CSD wavefront (defined as the leading edge of the propagated EEG amplitude attenuation and DC shift) (Leão 1944; Wahl et al. 1987). However, the sequence of vascular changes relative to the physiological events in CSD has not been characterized in detail.
Specific conduction of vasodilatation distant to a localized stimulus was previously demonstrated in a variety of tissues in vivo (de Wit et al. 2006), including in brain, to synaptic activity (Iadecola et al. 1997). In vitro preparations of cerebral vessels show propagated dilatation in response to local application of a number of stimuli including KCl, acetylcholine, adenosine triphosphate (ATP), and adenosine (Horiuchi et al. 2002; Ngai et al. 2007). Mechanisms proposed for vascular conduction include electrotonic spread of hyperpolarization through gap junctions, activation of Na+/K+-ATPase, and activation of KCa and Kir channels (de Wit et al. 2006; Horiuchi et al. 2002). However, there has been comparatively little investigation of vascular propagation to physiological or pathological brain activation such as CSD in vivo.
Optical intrinsic signal imaging enables visualization of CSD at the cortical surface with high temporal and spatial resolution (Ba et al. 2002; Guiou et al. 2005; O'Farrell et al. 2000; Pouratian and Toga 2002). Previous work defined the OIS changes of CSD at multiple wavelengths and correlated these measures with laser Doppler and laser speckle flowmetry, intravascular dye imaging, and electrophysiology (Ayata et al. 2004; Ba et al. 2002; Guiou et al. 2005; O'Farrell et al. 2000; Tomita et al. 2002, 2005). The response differs depending on wavelength, but at visible wavelengths, where hemoglobin absorption is prominent, the dominant contribution to the OIS signal is thought to be changes in blood volume.
Here we have exploited the high spatial resolution and functional imaging capability of OIS to examine optical correlates of CSD in the superficial brain parenchyma while also tracking changes in the diameter of cortical surface vessels. We studied the phenomenon in both mice and rats because these animals can have differences in CSD characteristics (Ayata et al. 2004) and both are important model systems. We show that the surface vascular response to CSD propagates with temporal and spatial characteristics that are distinct from those of the underlying parenchyma, suggesting a specific mechanism for vascular conduction.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Experimental protocols were approved by the UCLA Chancellor's Animal Research Committee. Male C57Bl/6 mice (25–30 g) and Sprague–Dawley rats (300–350 g) were housed at 21°C on a 12-h light–dark cycle. Food and water were available without restriction. Anesthesia was achieved with isoflurane (induction 4%; maintenance 0.5–1.5% in a 70/30% nitrogen/oxygen mixture). The right femoral artery and tail vein were cannulated for physiologic monitoring and saline infusion, respectively. A silver/silver chloride wire was placed under the skin of the chest wall to monitor heart rate. An identical ground electrode was placed in the midline fascia of the neck. A pulse oximetry probe was placed on the left hindpaw (Nonin 8600V, Plymouth, MN). The animal was then placed in a stereotaxic frame (Kopf Instruments, Tujunga, CA) in preparation for imaging.
In mice, the parietal skull was thinned to expose an area 1 mm from bregma to 1 mm from lambda, and 1 mm from temporal ridge to 1 mm lateral to the sagittal suture. This enabled good imaging resolution without requiring a craniotomy. In rats, a craniotomy was performed and the dura was removed, exposing an area with the same boundaries as in mice. A well made with silicone gel was filled with artificial cerebrospinal fluid (ACSF, in mM: Na+ 151; Cl– 131; HCO3– 26; K+ 3.5; Ca2+ 2; Mg2+ 2; PO43– 1.25, glucose 10; pH 7.4). A field potential electrode and KCl ejection pipette were placed through two burr holes at opposite corners of the thinned area in mice (see Fig. 1A). In rats, these were placed at opposite corners of the cranial window. In both cases, the location of probes was kept consistent to allow reproducibility. Temperature was maintained at 37 ± 0.5°C, with a rectal temperature probe and homeothermic blanket. Adequate depth of anesthesia was assessed by lack of response to toe pinch, with preserved corneal reflex. Mean arterial pressure (MAP), heart rate (HR), respiratory rate (RR), and oxygen saturation (O2sat) were continuously monitored. Anesthetic levels were adjusted to maintain physiological variables within a constant range in mice and rats: RR 60–80 in mice, 50–60 in rats, MAP 90 ± 110 mmHg, HR 390–450 in mice, 280–300 in rats, pH 7.34 ± 0.01, PaO2 121 ± 7 mmHg; PaCO2 39 ± 1 mmHg, O2sat >95%. No significant difference in these parameters was noted during CSD compared with control conditions. Animals were killed at the end of the experiment with an overdose of anesthetic combined with nitrogen asphyxia.
|
CSD was induced by controlled release of 1 M KCl from a 34-gauge fused silica micropipette (MicroFil 34AWG, World Precision Instruments, Sarasota, FL) whose tip was positioned on the cortical surface, forming a seal. Controlled ejection of KCl was performed using a ramped series of pressure pulses from a pneumatic pico-pump (World Precision Instruments; 4–20 PSI, 100-ms pulse; volume released about 200–1,000 nL) until CSD threshold was achieved. This allowed a highly replicable induction of one CSD wave per application, while using a minimum of KCl to reach CSD threshold. To evaluate for the possibility of artifactual effects resulting from KCl ejection, five mice had CSD induced with DC cathodal stimulation (Rhodes MCE-100 concentric bipolar electrode, OD 150 µm, ID 25 µm, 100 µA x 1–10 s; Kopf Instruments). The stimulating electrode was placed just above the cortical surface in the same location used for KCl ejection. For continuous stimulation, five to eight KCl crystals were carefully placed on the cortical surface in rats, in the anteromedial corner of the cortical window. In separate experiments, we attempted induction of CSD using pulsed ejection of NaCl (0.9%) to control for fluid percussion effects of pulsed KCl application. Pulsed NaCl ejection did not induce CSD in the range of ejection parameters we used for these experiments. At much higher release pressures/volumes (50–100 PSI, corresponding to released volumes of 1.75–3.5 µL), we were able to induce CSD (with accompanying propagated vascular changes) but, unlike our lower pressure KCl ejections, these were accompanied by visible cortical trauma.
Imaging
The cortex was illuminated with broad-spectrum light (400–700 nm) by white light-emitting diodes (5500K, Lumiled Luxeon III, Philips Lumileds, San Jose, CA). Reflected light was collected with a custom objective consisting of two high-performance lenses (Cosmicar Pentax, f/0.95) joined front to front. Images were acquired at 2 Hz by a Watec 902K CCD camera (8-bit, sensitivity 0.00015 lux, spectral range, 320–1,100 nm; Watec, Tsuruoka, Japan), connected to a frame grabber (PCI-1409, National Instruments, Austin, TX). We did not filter the reflected light because in preliminary experiments (n = 3 mice) we demonstrated that there was no significant difference in the timing, amplitude, or duration of the OIS changes associated with CSD using broad-spectrum reflectance compared with filtering at 570 nm (full width at half-maximum 10 nm). Field of view was 3 x 2 mm, pixel size was 6.6 µm. The maximum depth in cortex at which 400- to 700-nm OIS changes were visualized was 100–200 µm for craniotomy preparation and 75–150 µm for thin skull preparation. Image acquisition and electrophysiology were coordinated by a workstation running a LabVIEW virtual instrument (National Instruments).
Treatment with dimethylsulfoxide (DMSO)
Five rats with open cranial preparation were treated with 2 and 10% (volume/volume) DMSO dissolved in ACSF. CSD was elicited at hourly intervals in either control solution (ACSF) or after 30 min of exposure to DMSO solution. Of the five animals tested, two were tested with 2 and 10% DMSO and three were tested with 10% only. The experimental protocol consisted of an initial control CSD induction, followed by one or two inductions in DMSO, and ending with a washout control CSD induction.
Electrophysiology
A capillary micropipette (resistance 0.5 M
) was advanced 500 µm under the cortical surface. The reference electrode was the silver–silver chloride ground wire placed in the midline neck fascia. Field potentials were acquired continuously, band-pass 0–1 kHz, amplified (A-M Systems 3000, Carlsborg, WA) and digitized (at 1 KHz) concurrently with optical imaging by a data acquisition board (PCI-6251, National Instruments).
Analysis
Arterioles and venules were identified by morphology and color. Arterioles had a smaller diameter, fewer branching points at less acute angles, and were a distinctly brighter red than venules. Arterioles also underwent marked constriction and dilatation during each CSD wave. They could thus be identified in an unbiased fashion by performing standard deviation (SD) analysis of the entire image stack (Z-project Tool, ImageJ, National Institutes of Health, Bethesda, MD). Using SD images, the total number of arteriolar segments in an image field was counted for each experiment. An arteriolar segment was defined as the region between two branch points or distal to a terminal branch point. Segments were counted separately even if they originated from the same vessel because responses between segments were often heterogeneous. Segments under 3 pixels (about 20 µm) in baseline diameter were not considered because below this diameter the pixel resolution of the camera did not allow precise measurement. The number of arteriolar segments responding before and after onset of CSD was recorded, as was the angle of incidence of the CSD wave at each arteriolar segment (Angle Tool, ImageJ). Arterioles oriented at 0° or 180° were oriented roughly parallel (formally, they were tangential) to the wavefront. Those at 0° were aligned with the proximal (wider) end to the left and the distal (narrower) end to the right. Those at 180° had the opposite orientation. Likewise, arterioles at 90° tapered toward the top of the image (distal up) and those at 270° tapered toward the bottom (distal down).
To sample vessels in an unbiased manner, we divided each imaging field into quadrants. The arteriolar segment closest to the center of each quadrant was chosen for analysis, yielding four arteriolar regions of interest (ROIs) per experiment. A 50-pixel (330-µm) line ROI was placed perpendicular to the long axis of the vessel. When plotted over the course of the experiment, this line ROI revealed the vessel's diameter, which could be computed by pixel count (Threshold and Measure Functions, ImageJ) (see Fig. 2). It also allowed determination of timing of dilatation relative to the CSD wavefront.
|
Velocity of the OIS CSD profile was measured by generating a linear ROI perpendicular to the advancing CSD wavefront. Changes in grayscale value that moved over time were revealed as angled lines in the resulting x–t plot or kymograph. Velocity of wave propagation was determined by dividing distance traveled in pixels over time. At least three different kymographs were generated and the velocity was computed as the average of the three measurements. Velocity of CSD propagation along arteriolar segments was measured by generating a ROI that followed the course of the vessel (see Fig. 3). A kymograph was generated as above and velocity computed. Because arterioles are at arbitrary and changing angles to the propagating CSD wavefront, the kymographs generated from vascular ROIs did not reveal the actual rate of propagation of CSD, generally underestimating it as a result of the longer path length of the kymograph through a vessel. Thus the velocity of propagation of arteriolar CSD changes was compared with parenchymal changes by displacing the original vascular ROI onto the parenchyma 10 pixels to the left or right of the vessel and generating a new kymograph. This allowed for comparison of relative velocity in arterioles and parenchyma using the same ROI.
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
CSD was associated with a highly consistent, multiphasic change in parenchymal reflectance that propagated concentrically from the point of stimulation. In mouse, there was a triphasic change (decrease, increase, decrease) in reflectance (Fig. 1, Supplementary Videos 1).1 In rats, a similar profile was noted, although the amplitude and duration of each phase was different (Supplementary Videos 2). Characteristics of the CSD OIS waveforms in mouse versus rat are summarized in Table 1. The negative DC shift began with the first OIS phase in both mice and rats and continued through the end of the second phase. It recovered during the third phase. In most experiments, CSD was induced with a KCl pulse, although in five mice where CSD was induced by cathodal stimulation, no significant difference in OIS waveform or rate of CSD propagation was noted (Table 1).
|
In all experiments, cortical surface arterioles showed changes in diameter associated with CSD, whereas the caliber of veins did not change. In both mice and rats, arterioles initially dilated in response to CSD. In mice, initial dilation was followed by a profound constriction before a larger dilatation and a return to a baseline more dilated than before CSD (Fig. 2). In rat, the initial dilation persisted throughout the first two phases of CSD and then was followed by sustained constriction relative to the baseline. In mouse, arterioles remained dilated relative to baseline diameter for 27 ± 5 min (mean ± SE; n = 5 animals) after passage of the CSD wave. In rat, arterioles were constricted relative to baseline diameter for 32 ± 7 min (mean ± SE; n = 4 animals) after CSD passage.
Before CSD induction, baseline vasomotion signal (V-signal) at about 0.1 Hz was noted in arterioles and surrounding parenchyma, which was attenuated during and after CSD, consistent with previously described spontaneous vasomotion changes associated with CSD (Guiou et al. 2005; Piper et al. 1991). Occasionally, there were also spontaneous segmental constrictions of small arterioles during the pre- and post-CSD baselines. However, propagated arteriolar changes like those described during CSD were never seen without CSD induction (n = 32 mice, n = 21 rats).
Arteriolar vasodilatation propagates faster than parenchymal CSD
The velocity of propagated arteriolar dilatation was significantly greater than the corresponding parenchymal CSD wave velocity in both mice (4.3 ± 0.9 vs. 2.0 ± 0.3 mm/min; P < 0.001; n = 32 arteriolar segments, 11 experiments, mean ± SE) and rats (6.4 ± 0.6 vs. 2.9 ± 0.5 mm/min; t = 4.11; P < 0.001; n = 19 arteriolar segments, nine experiments) (Fig. 3). As a result, arteriolar dilatation significantly preceded the OIS and electrophysiological changes of CSD in the majority of mice (28/32) and rats (17/21) (Fig. 4). On average, vasodilatation preceded the parenchymal CSD wavefront by 507 ± 55 µm and 15 ± 2 s in mice (n = 52 arteriolar segments, 13 experiments) and 709 ± 106 µm and 16 ± 3 s in rats (n = 19 arteriolar segments, 9 experiments, mean ± SE). The separation between vascular and parenchymal responses was increasingly apparent with increasing distance from the stimulus. In mice, arterioles >1,500 µm from the stimulus dilated 20 ± 2 s ahead of the parenchymal CSD wavefront versus 10 ± 2 s (t = 3.38; P < 0.001) ahead of the wavefront for vessels <1,500 µm from the stimulus (n = 52 arteriolar segments, 13 experiments in mice). These differences were not dependent on the nature of the stimulus because they were seen with both KCl ejection and cathodal stimulation (n = 19 arteriolar segments, five experiments in mice).
|
2 = 4.23; P < 0.05). They also dilated earlier, an average of 588 ± 88 µm and 22 ± 3 s ahead of the wavefront, compared with 332 ± 85 µm (t = 1.94; P < 0.05) and 11 ± 2 s (t = 2.84; P < 0.01) ahead in vessels oriented closer to parallel. If the dilatation response arose simply from diffusion of an extracellular messenger resulting from CSD, arterioles parallel to the CSD wavefront would be expected to respond at the same time as those perpendicular to the wavefront—this was not the case. To investigate the possibility that arteriolar dilatation could be the result of a passive increase in blood volume caused by downstream or upstream vasoconstriction associated with CSD, we examined the dilatation response with respect to the orientation of blood flow relative to the CSD wave. The direction of blood flow in arterioles relative to CSD did not affect incidence, distance, or duration of dilatation ahead of the CSD wavefront, with no difference found between propagation anterograde or retrograde to blood flow. Finally, arteriolar size did not have an effect on incidence, distance, or duration ahead of the CSD wavefront (n = 52 arteriolar segments, 13 mice; n = 44 arteriolar segments, 11 rats).
Although in the majority of cases arteriolar dilatation rather than constriction preceded the CSD wave front, there were cases where constriction occurred. Constriction before OIS or electrophysiological CSD occurred in 6 of 32 mice and 4 of 19 rats. However, this change occurred exclusively in close proximity (within 500–1,000 µm) to the stimulus. We observed a similar spatially graded response to cathodal stimulation, indicating that the vascular constriction near the stimulus was not specific to KCl stimulation (n = 2/5 mice).
Arteriolar and parenchymal changes of CSD can be dissociated
Dissociation of vascular and parenchymal changes of CSD was observed under three conditions: 1) in areas distant to parenchymal CSD propagation; 2) during continuous stimulation that evoked frequent repetitive CSD; and 3) with pharmacological inhibition of the vascular response with DMSO.
In some experiments (six of 32 mice and three of 19 rats), the OIS and electrophysiological changes of CSD did not spread into the entire area of visualized cortex. Nevertheless, arteriolar dilatation was observed to extend into these regions where there was no evidence of parenchymal CSD (Fig. 5). Conversely, the vascular response could be abolished in the presence of an intact parenchymal response. Frequent repetitive CSD was induced by placing KCl crystals on the exposed cortex (n = 4 rats). This resulted in continuous waves of CSD [an average (±SE) of 17 ± 6 individual waves] that continued for 38 ± 7 min. The initial periodicity of CSD events was 101.5 ± 7 s over the first 10 min, after which periodicity increased to 153 ± 13 s until CSD episodes eventually ceased. Although rat arterioles reacted typically to the first episodes of CSD with dilatation followed by constriction, they eventually became persistently dilated with diminished reactivity. Vascular reactivity was absent by 22 ± 4 min of continuous stimulation despite intact OIS changes for another 16 min (n = 27 arteriolar branches, 27 parenchymal ROIs in four of four rats; nonsignificant change in parenchymal OIS amplitude: F = 0.58; P = 0.74, one-way ANOVA) (Fig. 6).
|
|
The arteriolar response could also be inhibited with DMSO, a solvent shown to affect vascular activity at high concentrations. We applied two different concentrations (2 and 10%) of DMSO dissolved in ACSF to the exposed cortex of rats (n = 5). We noted no difference in CSD-associated changes in arteriolar diameter or OIS signal during exposure to 2% DMSO as compared with controls (data not shown). At 10%, DMSO abolished the arteriolar dilatation response associated with CSD, whereas the parenchymal changes still occurred (although with amplitude decreased by roughly 40%; Fig. 7). Meanwhile, there was no significant difference in the amplitude or duration of DC shift. These data suggest that 10% DMSO disrupted vascular but not electrophysiological changes during CSD and disproportionately affected the arteriolar response compared with the parenchymal changes indicated by OIS.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
; van Harreveld and Ochs 1957; Wahl et al. 1987). Likewise, measures of cerebral blood flow and volume typically show a short-lasting increase, followed by a longer-lasting decrease, in blood flow or volume, during and after CSD. Several of these studies (Dreier et al. 1998; Fabricius et al. 1995; Osada et al. 2006; Sonn and Mayevsky 2000; Tomita et al. 2005; van Harreveld and Ochs 1957) also note a brief period of arterial constriction or decreased perfusion before the characteristic dilatation and perfusion increase. To our knowledge, however, no prior study has demonstrated vascular changes preceding the advancing CSD wavefront.
Perhaps because vasomotor changes have always been identified during or after CSD, they have been widely considered to be a passive response to sequential increased and decreased metabolic activity (Leão 1944). If this were the case, vascular changes would be expected to lag the CSD wavefront and travel with the same or slower velocity. They would also be expected to change in the same direction as what would be expected for neurovascular coupling—that is, increased blood volume in response to increased metabolism and decreased volume in response to decreased activity. In contrast, our results suggest that there is a dissociation of metabolic demand and vascular response and that the cortical surface arteriolar changes associated with CSD appear to have independent, and possibly active, mechanisms of propagation.
Independent conduction of vasodilatation in cortical surface arterioles
Dilatation of surface arterioles propagated along individual vessels with intrinsic velocities that were significantly greater than the velocity of the parenchymal CSD wave measured by OIS and electrophysiology. This resulted in vasodilatation that increasingly preceded the CSD wavefront (by up to hundreds of microns and tens of seconds) with increasing distance from the stimulus.
The pattern of propagation of the arteriolar dilatation in response to CSD is most consistent with specific vascular conduction. In addition to having a distinct intrinsic velocity, vasodilatation was always continuous along individual vessels and never "skipped" across an area of cortex without an intervening vessel. This observation was confirmed quantitatively, in that arterioles oriented perpendicular to the CSD wave consistently showed dilatation earlier in relation to arrival of the parenchymal CSD wave than those oriented at acute angles to the parenchymal wave. The propagation of arteriolar dilatation into regions that were not reached by optical or electrophysiological changes of CSD also argues for an independent propagation of CSD-related changes along arterioles.
Experimental dissociation of arteriolar and cortical activity
Consistent with the hypothesis that vascular conduction during CSD was independent of underlying cortical activity, we were able to dissociate the vascular response from the responses of the underlying parenchyma. Frequent repetitive CSD evoked by continuous stimulation with KCl crystals caused the response of surface arterioles to become refractory despite ongoing parenchymal CSD events. In addition, in the process of investigating the effects of hydrophobic compounds on CSD using DMSO as a vehicle, we found that DMSO itself inhibited vascular reactivity associated with CSD. Although DMSO did attenuate the OIS signal change of CSD, the DC shift was preserved, suggesting that DMSO was acting primarily on vascular as opposed to neuronal or glial signaling. The mechanism of DMSO's action on arterioles is not known with certainty; it has multiple actions including antioxidant effects, inhibition of platelet aggregation, local anesthesia, protection from ischemia and glutamate-mediated cell death, and vasodilatation (Santos et al. 2003). A microapplication study of DMSO on pial arterioles (Pitts et al. 1986) showed that DMSO at concentrations comparable to ours significantly increased arteriolar diameter. These concentrations resulted in increased osmolarity in vitro and similar dilatation was observed when osmolarity was increased in vivo by other methods (Pitts et al. 1986; Wahl et al. 1973). Regardless of its mechanism, DMSO was able to reversibly abolish the vascular response to CSD in spite of preserved parenchymal changes. This observation provides further evidence of a distinct pathway of vascular activation in CSD.
Dissociation of vasomotor activity from metabolic demand in CSD
The traditional view of the vascular responses to CSD is that they represent passive responses to metabolic demand. Our results are not consistent with this hypothesis. First, the existence of parenchymal hypoperfusion during the period of maximum metabolic demand (the negative peak of the DC shift) in our preparation and that of others (Ayata et al. 2004; Dreier et al. 1998; Fabricius et al. 1995; Osada et al. 2006; Sonn and Mayevsky 2000; Tomita et al. 2005; van Harreveld and Ochs 1957) is inconsistent with a rapid increase in perfusion to meet metabolic demand. Although it is possible that the perfusion response to neuronal and glial activity is simply delayed, this delay (tens to hundreds of seconds) would be orders of magnitude greater than the delay in perfusion that occurs in response to physiological neuronal activation (Narayan et al. 1994). At the parenchymal level, it is more likely that neurovascular coupling is disrupted and profound neuro-glial depolarization causes swelling that constricts the capillary bed (Tomita et al. 2003).
The response of cortical surface arterioles to CSD is also inconsistent with a direct response to metabolic demand. In both mouse and rat, arteriolar dilatation precedes not only the OIS wavefront but also the prominent increase in parenchymal perfusion during the third OIS phase of CSD. Furthermore, in rat, surface arteriolar dilatation persists into the second OIS phase, which was shown to be associated with hypoperfusion in similar preparations (Tomita et al. 2005).
Differences between mouse and rat
We noted several differences in CSD between mouse and rat. CSD propagation was slower and duration of OIS and field potential changes was longer, in mouse than in rat. The relative amplitude of the second OIS phase was also significantly larger and that of the first and third phases smaller. In contrast to rat, mouse arterioles showed marked constriction after the CSD wavefront. The larger second phase and arteriolar constriction are consistent with the work of Ayata et al. (2004), who noted a profound hypoperfusion associated with CSD in mouse, attributed to a greater sensitivity of mouse arterioles to elevated potassium. The slower propagation and longer duration of the CSD event was an unexpected finding. It is possible that the slower propagation of mouse CSD might contribute to its longer duration, perhaps by more extreme elevations of excitatory mediators such as potassium, which in turn could account for the distinct vascular profile.
Possible mechanisms for conducted vasodilatation ahead of CSD
The initial dilatation of surface arterioles to CSD might be a purely mechanical process caused by profound microvascular constriction at the OIS wavefront. This mechanism would be in accord with recent work (Tomita et al. 2005) that showed capillary flow stop at the CSD wavefront in rats. Constriction near the CSD wavefront could lead to upstream or downstream passive dilatation of the involved arteriole, depending on the orientation of the arteriole to CSD. Our data are not consistent with this mechanism. The orientation of arterioles relative to direction of blood flow did not affect either incidence, duration, or amount of dilatation ahead of the CSD wavefront. If passive mechanical changes were involved, one would expect to see a difference in these parameters in CSD propagating with or against the flow of blood from the heart. However, these observations do not rule out active changes arising from stretch- or shear-induced signaling in the vessel.
Because we sampled electrophysiologically from only one location, it is possible that undetected electrophysiological changes of CSD preceded the OIS wavefront to trigger the surface vascular changes. However, DC potential shift consistently occurred simultaneously with the first phase of the OIS change (n > 100 experiments; data not shown). This consistency between optical and electrophysiological changes was previously shown in other preparations, both in vitro and in vivo (Basarsky et al. 1998; Guiou et al. 2005; Kunkler and Kraig 1998; Peters et al. 2003). Thus the first phase of the OIS wave was a highly reliable indicator of the onset of the electrophysiological changes of CSD.
Although it is known that CSD is associated with the release of multiple diffusible vasoactive substances (Gorji 2001; Somjen 2001), the diffusion of a vasodilatory substance is not alone adequate to explain the pattern of propagation we observed. If vasodilatation was attributable solely to the local presence of a diffusible substance, it should propagate circumferentially and independently of vascular orientation. The same argument applies to parenchymal astrocyte calcium waves, which were shown to propagate ahead of electrophysiological correlates of CSD in vitro (Basarsky et al. 1998; Kunkler and Kraig 1998; Peters et al. 2003). However, this does not exclude the possibility of either of these processes triggering an independently conducted vascular process.
Specific vascular conduction was previously observed in a variety of tissues including brain (de Wit et al. 2006; Iadecola et al. 1997) and isolated middle cerebral artery branches similar to the vessels we sampled demonstrated propagated changes in diameter in response to local stimuli, indicating that they have the independent capacity to conduct vasomotor activity over distances (Horiuchi et al. 2002; Ngai et al. 2007). The vasoactive compounds shown to induce conducted vasodilatation (acetylcholine, K+, ATP, NO, and adenosine, among others) are all known to be released during CSD (Gorji 2001; Somjen 2001). It is possible that one or more of these agents are responsible for triggering the conducted vasodilatations we observed.
One potential mechanism for this vascular conduction is electrotonic spread of changes in membrane potential through gap junctions between vascular cells (de Wit et al. 2006). In addition, endothelial cells, smooth muscle cells, and astrocytes (whose end feet ensheath cerebral vessels) each show intercellular waves of increases in intracellular calcium that could coordinate a propagated response to a localized stimulus (Charles et al. 1991; Domenighetti et al. 1998; Yashiro and Duling 2000).
Others have noted conducted vascular changes associated with CSD (Osada et al. 2006). However, these changes—consisting of spindle-shaped constrictions followed by conducted dilatation—were observed only after passage of the CSD wavefront. The authors did not observe dilatation ahead of the CSD wavefront. This difference may be explained by the experimental preparation (craniotomized cats with an optical fiber inserted into the brain parenchyma). It may also be explained by the local concentration of K+. Although the authors do not report the amount of K+ released to elicit CSD, previous studies by the same group used manual injection of comparatively large volumes of KCl to elicit CSD [<10 µL of 300–500 µM KCl, compared with 200–1,000 nL of 1 M KCl in our studies (Tomita et al. 2002, 2005)]. We showed that conducted dilatation ahead of the CSD wavefront was more likely to occur and of longer distance and duration, further away from the site of KCl ejection, suggesting it is less favored at high KCl concentrations. Furthermore, in some animals, within 500–1,000 µm from the stimulus, propagated constriction was noted ahead of the CSD wavefront. This biphasic conducted response is consistent with what was observed earlier in isolated cerebral vessels (Horiuchi et al. 2002) and further underlines the dissociation of arteriolar events from the underlying CSD event.
In conclusion, our results indicate that vascular propagation of CSD should be considered as a distinct component of this complex CNS event. CSD is believed to occur in the setting of migraine, stroke, and traumatic brain injury. Changes in vascular caliber may be of primary importance in the pathogenesis of each of these disorders. Distinct mechanisms of activation and propagation of the vascular response could also represent specific therapeutic targets in the clinical settings where CSD occurs.
|
GRANTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
1 The online version of this article contains supplemental data. 
Address for reprint requests and other correspondence: A. C. Charles, David Geffen School of Medicine at UCLA, Neurocience Research Building, Room 575, 635 Charles E. Young Drive South, Los Angeles, CA 90095 (E-mail: acharles{at}ucla.edu)
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Ba AM, Guiou M, Pouratian N, Muthialu A, Rex DE, Cannestra AF, Chen JW, Toga AW. Multiwavelength optical intrinsic signal imaging of cortical spreading depression. J Neurophysiol 88: 2726–2735, 2002.
Basarsky TA, Duffy SN, Andrew RD, MacVicar BA. Imaging spreading depression and associated intracellular calcium waves in brain slices. J Neurosci 18: 7189–7199, 1998.
Charles AC, Merrill JE, Dirksen ER, Sanderson MJ. Intercellular signaling in glial cells: calcium waves and oscillations in response to mechanical stimulation and glutamate. Neuron 6: 983–992, 1991.[CrossRef][ISI][Medline]
Colonna DM. Nitric oxide promotes arteriolar dilation during cortical spreading depression in rabbits. Stroke 25: 2463–2470, 1994.[Abstract]
de Wit C, Hoepfl B, Wölfle S. Endothelial mediators and communication through vascular gap junctions. Biol Chem 387: 3–9, 2006.[CrossRef][ISI][Medline]
Domenighetti AA, Beny J-L, Chabaud F, Frieden M. An intercellular regenerative calcium wave in porcine coronary artery endothelial cells in primary culture. J Physiol 513: 103–116, 1998.
Dreier JP, Korner K, Ebert N, Gorner A, Rubin I, Back T, Lindauer U, Wolf T, Villringer A, Einhaupl KM, Lauritzen M, Dirnagl U. Nitric oxide scavenging by hemoglobin or nitric oxide synthase inhibition by N-nitro-L-arginine induces cortical spreading ischemia when K+ is increased in the subarachnoid space. J Cereb Blood Flow Metab 18: 978–990, 1998.[CrossRef][ISI][Medline]
Fabricius M, Akgoren N, Lauritzen M. Arginine-nitric oxide pathway and cerebrovascular regulation in cortical spreading depression. Am J Physiol Heart Circ Physiol 269: H23–H29, 1995.
Florence G, Bonvento G, Charbonne R, Seylaz J. Spreading depression reversibly impairs autoregulation of cortical blood flow. Am J Physiol Regul Integr Comp Physiol 266: R1136–R1140, 1994.
Gorji A. Spreading depression: a review of the clinical relevance. Brain Res Rev 38: 33–60, 2001.[CrossRef][Medline]
Guiou M, Sheth S, Nemoto M, Walker M, Pouratian N, Ba A, Toga AW. Cortical spreading depression produces long-term disruption of activity-related changes in cerebral blood volume and neurovascular coupling. J Biomed Opt 10: 11004–11011, 2005.[CrossRef][Medline]
Horiuchi T, Dietrich HH, Hongo K, Dacey RG Jr. Mechanism of extracellular K+-induced local and conducted responses in cerebral penetrating arterioles. Stroke 33: 2692–2699, 2002.
Iadecola C, Yang G, Ebner TJ, Chen G. Local and propagated vascular responses evoked by focal synaptic activity in cerebellar cortex. J Neurophysiol 78: 651–659, 1997.
Kunkler PE, Kraig RP. Calcium waves precede electrophysiological changes of spreading depression in hippocampal organ cultures. J Neurosci 18: 3416–3425, 1998.
Lacombe P, Sercombe R, Correze JL, Springhetti V, Seylaz J. Spreading depression induces prolonged reduction of cortical blood flow reactivity in the rat. Exp Neurol 117: 278–286, 1992.[CrossRef][ISI][Medline]
Lauritzen M, Balslev Jøsrgensen M, Diemer NH, Gjedde A, Hansen AJ. Persistent oligemia of rat cerebral cortex in the wake of spreading depression. Ann Neurol 12: 469–474, 1982.[CrossRef][ISI][Medline]
Leão AAP. Pial circulation and spreading depression of activity in the cerebral cortex. J Neurophysiol 7: 391–396, 1944.
Narayan S, Santori E, Toga A. Mapping functional activity in rodent cortex using optical intrinsic signals. Cereb Cortex 4: 195–204, 1994.
Ngai AC, Nguyen T-S, Meno JR, Britz GW. Postischemic augmentation of conducted dilation in cerebral arterioles. Stroke 38: 124–130, 2007.
O'Farrell AM, Rex DE, Muthialu A, Pouratian N, Wong GK, Cannestra AF, Chen JW, Toga AW. Characterization of optical intrinsic signals and blood volume during cortical spreading depression. Neuroreport 11: 2121–2125, 2000.[ISI][Medline]
Osada T, Tomita M, Suzuki N. Spindle-shaped constriction and propagated dilation of arterioles during cortical spreading depression. Neuroreport 17: 1365–1368, 2006.[CrossRef][ISI][Medline]
Peters O, Schipke CG, Hashimoto Y, Kettenmann H. Different mechanisms promote astrocyte Ca2+ waves and spreading depression in the mouse neocortex. J Neurosci 23: 9888–9896, 2003.
Piper RD, Lambert GA, Duckworth JW. Cortical blood flow changes during spreading depression in cats. Am J Physiol Heart Circ Physiol 261: H96–H102, 1991.
Pitts L, Young A, McCulloch J, MacKenzie E. Vasomotor effects of dimethyl sulfoxide on cat cerebral arteries in vitro and in vivo. Stroke 17: 483–487, 1986.
Pouratian N, Toga AW. Optical imaging based on intrinsic signals. In: Brain Mapping: The Methods, edited by Toga AW. San Diego, CA: Academic Press, 2002, p. 97–140.
Reuter U, Weber JR, Gold L, Arnold G, Wolf T, Dreier J, Lindauer U, Dirnagl U. Perivascular nerves contribute to cortical spreading depression-associated hyperemia in rats. Am J Physiol Heart Circ Physiol 274: H1979–H1987, 1998.
Santos NC, Figueira-Coelho J, Martins-Silva J, Saldanha C. Multidisciplinary utilization of dimethyl sulfoxide: pharmacological, cellular, and molecular aspects. Biochem Pharmacol 65: 1035–1041, 2003.[CrossRef][ISI][Medline]
Seitz I, Dirnagl U, Lindauer U. Impaired vascular reactivity of isolated rat middle cerebral artery after cortical spreading depression in vivo. J Cereb Blood Flow Metab 24: 526–530, 2004.[ISI][Medline]
Shibata M, Leffler CW, Busija DW. Prostanoids attenuate pial arteriolar dilation induced by cortical spreading depression in rabbits. Am J Physiol Regul Integr Comp Physiol 261: R828–R834, 1991.
Somjen GG. Mechanisms of spreading depression and hypoxic spreading depression-like depolarization. Physiol Rev 81: 1065–1096, 2001.
Sonn J, Mayevsky A. Effects of brain oxygenation on metabolic, hemodynamic, ionic and electrical responses to spreading depression in the rat. Brain Res 882: 212–216, 2000.[CrossRef][ISI][Medline]
Tomita M, Schiszler I, Tomita Y, Tanahashi N, Takeda H, Osada T, Suzuki N. Initial oligemia with capillary flow stop followed by hyperemia during K+-induced cortical spreading depression in rats. J Cereb Blood Flow Metab 25: 742–747, 2005.[CrossRef][ISI][Medline]
Tomita M, Tanahashi N, Takeda H, Takao M, Tomita Y, Amano T, Fukuuchi Y. Astroglial swelling in the neuronal depolarization ensemble. Acta Neurochir Suppl 86: 219–222, 2003.[Medline]
Tomita Y, Tomita M, Schiszler I, Amano T, Tanahashi N, Kobari M, Takeda H, Ohtomo M, Fukuuchi Y. Repetitive concentric wave-ring spread of oligemia/hyperemia in the sensorimotor cortex accompanying K+-induced spreading depression in rats and cats. Neurosci Lett 322: 157–160, 2002.[CrossRef][ISI][Medline]
van Harreveld A, Ochs S. Electrical and vascular concomitants of spreading depression. Am J Physiol 189: 159–166, 1957.
Wahl M, Kuschinsky W, Bosse O, Thurau K. Dependency of pial arterial and arteriolar diameter on perivascular osmolarity in the cat: a microapplication study. Circ Res 32: 162–169, 1973.
Wahl M, Lauritzen M, Schilling L. Change of cerebrovascular reactivity after cortical spreading depression in cats and rats. Brain Res 411: 72–80, 1987.[CrossRef][ISI][Medline]
Yashiro Y, Duling BR. Integrated Ca2+ signaling between smooth muscle and endothelium of resistance vessels. Circ Res 87: 1048–1054, 2000.
This article has been cited by other articles:
|
|
 |
|
  T. H. Murphy, P. Li, K. Betts, and R. Liu Two-Photon Imaging of Stroke Onset In Vivo Reveals That NMDA-Receptor Independent Ischemic Depolarization Is the Major Cause of Rapid Reversible Damage to Dendrites and Spines J. Neurosci., February 13, 2008; 28(7): 1756 - 1772. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||
