0361-9230/91 $3.00 + II0
Brain Research Bulletin, Vol. 26, pp. 753-758. 0 Pergamon Press pk. 1991. Printed in the U.S.A.
Evidence Against Parenchymal Metabolites Directly Promoting Pial Arteriolar Dilation During Cortical Spreading Depression in Rabbits MASAAKI SHIBATA, CHARLES W. LEFFXER AND DAVID W. BUSIJA Department of Physiology and Biophysics, University of Tennessee, 894 Union Avenue, Memphis, TN 38163 Received 18 October 1990 SHIBATA, M., C. W. LEF’FLER AND D. W. BUSLJA. Evidence against parenchymal metabolites directly promoting pial arteriolar dilation during cortical spreading depression in rabbits. BRAIN RES BULL 26(5) 753-758, 1991. -The role of parenchyma1 metabolic factors in directly promoting pial arteriolar dilation during cortical spreading depression (CSD) in anesthetized rabbits was examined by direct measurement of periarachnoid cerebrospinal fluid (CSF) levels of a representative metabolite (i.e., K+) or superfusion of the cerebral cortical surface with artificial CSF. CSD was induced by KC1 microinjection or tissue puncture and its movement was monitored electrophysiologically. Pial arteriolar diameter was determined using a closed cranial window and intravital microscopy. CSD propagated across the cortex under the window with a velocity of 2.9k0.2 nunAn& and caused pial arteriolar diameter to increase from 8729 Pm to 133-t 11 pm (53%, n=23) for 1.620.1 mm. At the same time, Kf concentration increased from 3.0 kO.2 mM to a maximum of 4.6kO.3 mh4. Topical application of 6 mM Kf increased pial arteriolar diameter by only 8%. Continuous superfusion of the cortical surface with aCSF at a rate of 3.04.5 ml/mm (window volume = 0.5 ml) did not affect pial arteriolar dilation during CSD, but virtually abolished pial arteriolar dilation during inhalation of 10.2% CO,. These results suggest that pial arterioles dilate via a mechanism which does not involve diffusion of vasoactive metabolites released from the parenchyma during CSD. Pial arteriolar
dilation
Cortical spreading
depression
Potassium
ions
Cranial window
substances from reaching pial arterioles.
CORTICAL spreading depression (CSD) is a wave of neuronal and glial depolarization which propagates through cortical tissue at a velocity of 2-5 mm/min (1,lO) and leads to major changes in cerebral hemodynamics (2, 11, 18). Spatio-temporally related to the location of CSD is pial arteriolar dilation (57-71%) (16) and tissue hyperemia (256-487%, to be published). The mechanisms involved in cerebrovascular dilation are unknown, and it is also unclear whether pial and parenchymal arterioles are responding similarly to the same stimuli. Our recent observations (16) concerning the speed of onset and duration of pial arteriolar dilation during CSD have lead us to speculate that the response of these vessels is under the control of neurogenic rather than metabolic factors. In contrast, changes in tone of parenchymal arterioles during CSD may be dependent upon extracellular levels of several dilator stimuli, including K+, H+, and free arachidonic acid, which increase dramatically during CSD (9, 13, 14, 17). The possibility also exists that these substances could diffuse from the parenchyma into periarachnoid cerebrospinal fluid (CSF) where they would come into contact with and dilate pial arterioles. The aim of the present study was to evaluate the effect of metabolic stimuli from the parenchyma on pial arteriolar dilation during CSD in anesthetized rabbits. We used two different approaches. Fist, we measured levels of a representative vasodilator metabolite, namely K+ , in the vicinity of pial arterioles prior to and during CSD. Second, we superfused the cortical surface with artificial CSF to prevent parenchymal-derived vasoactive
METHOD Animal Preparations
Adult rabbits (2.2-2.8 kg, n=29) were initially anesthetized with thiopental sodium (25 mg/kg, IV, Abbott Lab.) followed by a bolus injection of urethane (1.6-1.8 g/kg, IP, Sigma Chem. Corp.). A femoral vein and artery were catheterized for, respectively, injection of fluids, or monitoring a blood pressure (BP), PO,, PCO, and pH. After tracheotomy, the animal’s head was mounted in a stereotaxic head holder, and the skin was cut open. An opening of the skull 20 mm in diameter was made over the parietooccipital cortex, and the dura mater was removed. A circular cranial window (20 mm in diameter and equipped with inlet and outlet ports) was placed over the opening and fixed to the skull with acrylic dental cement. Additionally, two small openings (1.0 mm in diameter) were made over the frontal and occipital cortices just outside the window (see Fig. l), and glass pipettes (2.0 mm in o.d., lo-15 mm in length with the tip diameter of ca. 150 pm) filled with 0.9% saline were lowered 1 .O mm into the cortex, and fixed to the skull with the cement. Two recording calomel electrodes were then connected via cotton wicks to open ends of the glass pipettes. A reference electrode of the same type was placed directly on the skull along the midline over the olfactory bulb. Slow potential
753
SHIBATA,
754
changes (SPC) accompanying CSD were then recorded between the frontal and occipital electrodes and the reference electrode. The distance between the two recording calomel electrodes varied from 20.0 to 22.0 mm. Another small opening of the skull was made 2 mm anterior to the frontal electrode, and CSD was elicited by microinjection of KC1 solution (5%, 4-7 pl) 1.0 mm below the cortical surface by an injection needle, or by repeated insertions of a needle into the cortex. Artificial Cerebrospinal
of Pial Arteriolar of CSD
Diameter
The cranial window, which contains ca. 500 pl aCSF, was filled with gassed and warmed aCSF and repeatedly flushed for a period of 30 min. Using a microscope and a video camera set above the window, images of the exposed cortical surface were projected onto a video monitor. Diameter of pial arteriole was measured by a dimensional analysis system (VPA 1000, FOR A Corp., Ltd.), and outputs were fed on-line into a pen-recorder. Warmed and gassed aCSF was infused under the window, and KC1 was microinjected into the frontal cortex. Successful induction of single CSD was confirmed by recording SPC from the nearby frontal electrode No. 1 which registered a large negative (upward deflection) followed by a small positive (downward deflection) DC shift (see trace O-l in Fig. 1). While the CSD propagated across the cortex under the window, SPC, changes in pial arteriolar diameter and BP were continuously recorded. Shortly after the arrival of the CSD at the occipital electrode No. 2 (10 min from the negative SPC peak recorded from the frontal electrode No. l), all recordings were terminated. Forty to 50 min elapsed before the 2nd CSD was elicited. Measurement of CSF Kt Levels During CSD and Determination of Pial Arteriolar Responses to Increased CSF K+ Levels
The window was filled with aCSF and pial arteriolar diameter was measured. After 10 min, a 100~pl sample of CSF from under the window was collected by gently flushing through one port while collecting from another. Another CSF sample was collected in a similar manner when the propagating CSD was in the vicinity of the pial arteriole of interest. To minimize dilution of parenchymally derived K+ by the bulk aCSF in the window, we also used a microsampling procedure. The open tip of a small tubing (PE 60) was placed on the cortical surface, and during CSD we collected a lOtI-pl sample when the adjoining arteriole dilated. CSF samples were frozen immediately at - 20°C for later assay of K+ levels by flame photometer (480 Flame Photometer, Coming Med. Sci.). Kf levels were determined twice from the same sample and the values averaged. Superfasion
of the Cortical Surface
We examined
conditions
SUPERFtJSION n,.
I
Fluid (aCSF)
Artifical CSF for rabbits was prepared according to the following formula (in mM): KCl=2.9, MgC1,=0.6, CaCl,= 1.5, NaCl= 131.9, urea=6.7, dextrose=3.7 and NaHCO,=20.2. It was gassed with 6% 0,/6.3% CO,/87.7% N,, and exhibited PO,, PCO, and pH in the range of, respectively, 41-48 mmHg, 35-40 mmHg and 7.35-7.41. Artificial CSF containing higher KC1 concentration (6 mM and 12 mM) were prepared by balancing osmolarity of the medium by reducing NaCl. These were bubbled with the same gas mixture and showed similar ranges of PO,, PCO, and pH as those of standard aCSF. Determination and Induction
LEFFLER AND BUSIJA
With aCSF
pial arteriolar responses to CSD under normal and during superfusion with aCSF. For superfusion
0
2
4
6
810
FIG. 1. Diameter change of a pial arteriole induced
0
by
2
4
6
810 MIN
cortical spread-
ing depression (CSD) before (left) and during (right) superfusion. Experimental arrangements were schematically illustrated in the inset which represents dorsal view of the rabbit brain. A single CSD was elicited by KC1 microinjection in the left frontal cortex (KCl) and accompanying slow potential changes (SPC) were monitored between the two recording calomel electrodes (No. 1 and No. 2) and a reference calomel electrode (0). Successful induction of single CSD was proved by recording SPC at the nearby frontal electrode No. 1 (SPC trace, O-l) and subsequent propagation to the occipital electrode No. 2 (SPC trace, O-2). While CSD propagated the cortex toward the occipital boundaries across the cranial window (large open circle), diameter change of a pial arteriole (small filled circle) was recorded. Artificial cerebrospinal fluid was continuously superfused at a rate of 3.3 ml/mm in the retrograde direction with respect to CSD vector (from the occipital toward the frontal cortex). Time 0 min corresponds to the moment of the negative SPC peak recorded from the frontal electrode No. 1.
trials, the exposed cortex under the window was continuously flushed with warmed and gassed aCSF at a flow rate of 3.0-4.5 mUmin always in the retrograde direction with respect to CSD vector. Changes in arteriolar diameter, SPC and BP were continuously measured for 10 min. We also assessed effectiveness of superfusion in counteracting the dilator effect of CO, inhalation on pial arterioles. After several flushings of the window, changes in arteriolar diameter and BP were measured every 3 min for 30 min to establish the baseline level. For the control experiments, ventilating room air was switched to a gas mixture of 20.8% 0,/10.2% CO, in N, until pial arteriolar diameter attained a steady state, usually 12 min later. Changes in arteriolar diameter and BP were read every 3 min during 10.2% CO, inhalation. The gas mixture was then switched back to room air, and the reading continued for additional 18 min totalling 30 min. The window was flushed out and the return of arteriolar diarneter to the baseline level was confirmed within 60 min. Next, the cortex under the window was continuously superfused with gassed and warmed aCSF at a flow rate of 3.5-4.2 ml/min. Animals were similarly ventilated with scribed in the control experiments,
two different and changes
gases as dein arteriolar
diameter and BP were recorded every 3 min for 30 min during superfusion. We also tried to eliminate the possibility that K+ diffusing from the injection site was responsible for pial arteriolar dilation. For mechanical induction of CSD, a needle was repeatedly inserted into the frontal cortex until typical SPC accompanying CSD was recorded from the frontal electrode No. 1. To examine whether CSF Kf levels determined during CSD were sufficient to account for pial arteriolar dilation, we tested effects of 6 n&l and 12 mM K+ levels on arteriolar diameter. After repeated flushings of the window for 30 min, standard aCSF containing 2.9 n&l KC1 was infused under the window
PIAL ARTERIOLAR
A,
755
DILATION
tal and occipital electrodes,
and the frontal electrode and a tested pial arteriole was measured. Time for CSD propagation from the frontal electrode No. 1 to the occipital electrode No. 2 (between the two negative SPC peaks) and time from the negative SPC peak at the frontal electrode No. 1 to a moment of maximum arteriolar dilation (peak onset) was measured. These data were used to analyze spatio-temporal relationship between dilating pial arteriole and propagating CSD. Duration of arteriolar dilation was determined from the moment of its onset and that of a 80% return from a maximum size. Diameter changes of pial arterioles during 10.2% CO, inhalation before and during superfusion were expressed as percent change either above or below the baseline levels of corresponding recording at 3 min intervals. Arteriolar diameter change to aCSF containing differing Kf concentrations was determined when the diameter reached to a new steady state (usually between 3 min to 5 min after aCSF infusion), and expressed as percent change with respect to the preinfusion level. All data were expressed as mean k SEM and treated with two-tailed Student’s t-test. The 95% level of confidence @
Brain Research Bulletin, Vol. 26, pp. 753-758. 0 Pergamon Press pk. 1991. Printed in the U.S.A.
Evidence Against Parenchymal Metabolites Directly Promoting Pial Arteriolar Dilation During Cortical Spreading Depression in Rabbits MASAAKI SHIBATA, CHARLES W. LEFFXER AND DAVID W. BUSIJA Department of Physiology and Biophysics, University of Tennessee, 894 Union Avenue, Memphis, TN 38163 Received 18 October 1990 SHIBATA, M., C. W. LEF’FLER AND D. W. BUSLJA. Evidence against parenchymal metabolites directly promoting pial arteriolar dilation during cortical spreading depression in rabbits. BRAIN RES BULL 26(5) 753-758, 1991. -The role of parenchyma1 metabolic factors in directly promoting pial arteriolar dilation during cortical spreading depression (CSD) in anesthetized rabbits was examined by direct measurement of periarachnoid cerebrospinal fluid (CSF) levels of a representative metabolite (i.e., K+) or superfusion of the cerebral cortical surface with artificial CSF. CSD was induced by KC1 microinjection or tissue puncture and its movement was monitored electrophysiologically. Pial arteriolar diameter was determined using a closed cranial window and intravital microscopy. CSD propagated across the cortex under the window with a velocity of 2.9k0.2 nunAn& and caused pial arteriolar diameter to increase from 8729 Pm to 133-t 11 pm (53%, n=23) for 1.620.1 mm. At the same time, Kf concentration increased from 3.0 kO.2 mM to a maximum of 4.6kO.3 mh4. Topical application of 6 mM Kf increased pial arteriolar diameter by only 8%. Continuous superfusion of the cortical surface with aCSF at a rate of 3.04.5 ml/mm (window volume = 0.5 ml) did not affect pial arteriolar dilation during CSD, but virtually abolished pial arteriolar dilation during inhalation of 10.2% CO,. These results suggest that pial arterioles dilate via a mechanism which does not involve diffusion of vasoactive metabolites released from the parenchyma during CSD. Pial arteriolar
dilation
Cortical spreading
depression
Potassium
ions
Cranial window
substances from reaching pial arterioles.
CORTICAL spreading depression (CSD) is a wave of neuronal and glial depolarization which propagates through cortical tissue at a velocity of 2-5 mm/min (1,lO) and leads to major changes in cerebral hemodynamics (2, 11, 18). Spatio-temporally related to the location of CSD is pial arteriolar dilation (57-71%) (16) and tissue hyperemia (256-487%, to be published). The mechanisms involved in cerebrovascular dilation are unknown, and it is also unclear whether pial and parenchymal arterioles are responding similarly to the same stimuli. Our recent observations (16) concerning the speed of onset and duration of pial arteriolar dilation during CSD have lead us to speculate that the response of these vessels is under the control of neurogenic rather than metabolic factors. In contrast, changes in tone of parenchymal arterioles during CSD may be dependent upon extracellular levels of several dilator stimuli, including K+, H+, and free arachidonic acid, which increase dramatically during CSD (9, 13, 14, 17). The possibility also exists that these substances could diffuse from the parenchyma into periarachnoid cerebrospinal fluid (CSF) where they would come into contact with and dilate pial arterioles. The aim of the present study was to evaluate the effect of metabolic stimuli from the parenchyma on pial arteriolar dilation during CSD in anesthetized rabbits. We used two different approaches. Fist, we measured levels of a representative vasodilator metabolite, namely K+ , in the vicinity of pial arterioles prior to and during CSD. Second, we superfused the cortical surface with artificial CSF to prevent parenchymal-derived vasoactive
METHOD Animal Preparations
Adult rabbits (2.2-2.8 kg, n=29) were initially anesthetized with thiopental sodium (25 mg/kg, IV, Abbott Lab.) followed by a bolus injection of urethane (1.6-1.8 g/kg, IP, Sigma Chem. Corp.). A femoral vein and artery were catheterized for, respectively, injection of fluids, or monitoring a blood pressure (BP), PO,, PCO, and pH. After tracheotomy, the animal’s head was mounted in a stereotaxic head holder, and the skin was cut open. An opening of the skull 20 mm in diameter was made over the parietooccipital cortex, and the dura mater was removed. A circular cranial window (20 mm in diameter and equipped with inlet and outlet ports) was placed over the opening and fixed to the skull with acrylic dental cement. Additionally, two small openings (1.0 mm in diameter) were made over the frontal and occipital cortices just outside the window (see Fig. l), and glass pipettes (2.0 mm in o.d., lo-15 mm in length with the tip diameter of ca. 150 pm) filled with 0.9% saline were lowered 1 .O mm into the cortex, and fixed to the skull with the cement. Two recording calomel electrodes were then connected via cotton wicks to open ends of the glass pipettes. A reference electrode of the same type was placed directly on the skull along the midline over the olfactory bulb. Slow potential
753
SHIBATA,
754
changes (SPC) accompanying CSD were then recorded between the frontal and occipital electrodes and the reference electrode. The distance between the two recording calomel electrodes varied from 20.0 to 22.0 mm. Another small opening of the skull was made 2 mm anterior to the frontal electrode, and CSD was elicited by microinjection of KC1 solution (5%, 4-7 pl) 1.0 mm below the cortical surface by an injection needle, or by repeated insertions of a needle into the cortex. Artificial Cerebrospinal
of Pial Arteriolar of CSD
Diameter
The cranial window, which contains ca. 500 pl aCSF, was filled with gassed and warmed aCSF and repeatedly flushed for a period of 30 min. Using a microscope and a video camera set above the window, images of the exposed cortical surface were projected onto a video monitor. Diameter of pial arteriole was measured by a dimensional analysis system (VPA 1000, FOR A Corp., Ltd.), and outputs were fed on-line into a pen-recorder. Warmed and gassed aCSF was infused under the window, and KC1 was microinjected into the frontal cortex. Successful induction of single CSD was confirmed by recording SPC from the nearby frontal electrode No. 1 which registered a large negative (upward deflection) followed by a small positive (downward deflection) DC shift (see trace O-l in Fig. 1). While the CSD propagated across the cortex under the window, SPC, changes in pial arteriolar diameter and BP were continuously recorded. Shortly after the arrival of the CSD at the occipital electrode No. 2 (10 min from the negative SPC peak recorded from the frontal electrode No. l), all recordings were terminated. Forty to 50 min elapsed before the 2nd CSD was elicited. Measurement of CSF Kt Levels During CSD and Determination of Pial Arteriolar Responses to Increased CSF K+ Levels
The window was filled with aCSF and pial arteriolar diameter was measured. After 10 min, a 100~pl sample of CSF from under the window was collected by gently flushing through one port while collecting from another. Another CSF sample was collected in a similar manner when the propagating CSD was in the vicinity of the pial arteriole of interest. To minimize dilution of parenchymally derived K+ by the bulk aCSF in the window, we also used a microsampling procedure. The open tip of a small tubing (PE 60) was placed on the cortical surface, and during CSD we collected a lOtI-pl sample when the adjoining arteriole dilated. CSF samples were frozen immediately at - 20°C for later assay of K+ levels by flame photometer (480 Flame Photometer, Coming Med. Sci.). Kf levels were determined twice from the same sample and the values averaged. Superfasion
of the Cortical Surface
We examined
conditions
SUPERFtJSION n,.
I
Fluid (aCSF)
Artifical CSF for rabbits was prepared according to the following formula (in mM): KCl=2.9, MgC1,=0.6, CaCl,= 1.5, NaCl= 131.9, urea=6.7, dextrose=3.7 and NaHCO,=20.2. It was gassed with 6% 0,/6.3% CO,/87.7% N,, and exhibited PO,, PCO, and pH in the range of, respectively, 41-48 mmHg, 35-40 mmHg and 7.35-7.41. Artificial CSF containing higher KC1 concentration (6 mM and 12 mM) were prepared by balancing osmolarity of the medium by reducing NaCl. These were bubbled with the same gas mixture and showed similar ranges of PO,, PCO, and pH as those of standard aCSF. Determination and Induction
LEFFLER AND BUSIJA
With aCSF
pial arteriolar responses to CSD under normal and during superfusion with aCSF. For superfusion
0
2
4
6
810
FIG. 1. Diameter change of a pial arteriole induced
0
by
2
4
6
810 MIN
cortical spread-
ing depression (CSD) before (left) and during (right) superfusion. Experimental arrangements were schematically illustrated in the inset which represents dorsal view of the rabbit brain. A single CSD was elicited by KC1 microinjection in the left frontal cortex (KCl) and accompanying slow potential changes (SPC) were monitored between the two recording calomel electrodes (No. 1 and No. 2) and a reference calomel electrode (0). Successful induction of single CSD was proved by recording SPC at the nearby frontal electrode No. 1 (SPC trace, O-l) and subsequent propagation to the occipital electrode No. 2 (SPC trace, O-2). While CSD propagated the cortex toward the occipital boundaries across the cranial window (large open circle), diameter change of a pial arteriole (small filled circle) was recorded. Artificial cerebrospinal fluid was continuously superfused at a rate of 3.3 ml/mm in the retrograde direction with respect to CSD vector (from the occipital toward the frontal cortex). Time 0 min corresponds to the moment of the negative SPC peak recorded from the frontal electrode No. 1.
trials, the exposed cortex under the window was continuously flushed with warmed and gassed aCSF at a flow rate of 3.0-4.5 mUmin always in the retrograde direction with respect to CSD vector. Changes in arteriolar diameter, SPC and BP were continuously measured for 10 min. We also assessed effectiveness of superfusion in counteracting the dilator effect of CO, inhalation on pial arterioles. After several flushings of the window, changes in arteriolar diameter and BP were measured every 3 min for 30 min to establish the baseline level. For the control experiments, ventilating room air was switched to a gas mixture of 20.8% 0,/10.2% CO, in N, until pial arteriolar diameter attained a steady state, usually 12 min later. Changes in arteriolar diameter and BP were read every 3 min during 10.2% CO, inhalation. The gas mixture was then switched back to room air, and the reading continued for additional 18 min totalling 30 min. The window was flushed out and the return of arteriolar diarneter to the baseline level was confirmed within 60 min. Next, the cortex under the window was continuously superfused with gassed and warmed aCSF at a flow rate of 3.5-4.2 ml/min. Animals were similarly ventilated with scribed in the control experiments,
two different and changes
gases as dein arteriolar
diameter and BP were recorded every 3 min for 30 min during superfusion. We also tried to eliminate the possibility that K+ diffusing from the injection site was responsible for pial arteriolar dilation. For mechanical induction of CSD, a needle was repeatedly inserted into the frontal cortex until typical SPC accompanying CSD was recorded from the frontal electrode No. 1. To examine whether CSF Kf levels determined during CSD were sufficient to account for pial arteriolar dilation, we tested effects of 6 n&l and 12 mM K+ levels on arteriolar diameter. After repeated flushings of the window for 30 min, standard aCSF containing 2.9 n&l KC1 was infused under the window
PIAL ARTERIOLAR
A,
755
DILATION
tal and occipital electrodes,
and the frontal electrode and a tested pial arteriole was measured. Time for CSD propagation from the frontal electrode No. 1 to the occipital electrode No. 2 (between the two negative SPC peaks) and time from the negative SPC peak at the frontal electrode No. 1 to a moment of maximum arteriolar dilation (peak onset) was measured. These data were used to analyze spatio-temporal relationship between dilating pial arteriole and propagating CSD. Duration of arteriolar dilation was determined from the moment of its onset and that of a 80% return from a maximum size. Diameter changes of pial arterioles during 10.2% CO, inhalation before and during superfusion were expressed as percent change either above or below the baseline levels of corresponding recording at 3 min intervals. Arteriolar diameter change to aCSF containing differing Kf concentrations was determined when the diameter reached to a new steady state (usually between 3 min to 5 min after aCSF infusion), and expressed as percent change with respect to the preinfusion level. All data were expressed as mean k SEM and treated with two-tailed Student’s t-test. The 95% level of confidence @
