Variation in local cerebral blood flow response to high-dose pentobarbital sodium in the rat TADAHIRO OTSUKA, LING WEI, VIRGIL R. ACUFF, AKIRA SHIMIZU, KAREN D. PETTIGREW, CLIFFORD S. PATLAK, AND JOSEPH D. FENSTERMACHER Department of Neurological Surgery, Health Sciences Center, State University of New York, Stony Brook, New York 11794; and Division of Biometry and Applied Sciences, National Institute of Mental Health, Bethesda, Maryland 20892

OTSUKA, TADAHIRO, LING WEI, VIRGIL R. ACUFF, AKIRA SHIMIZU, KAREN D. PETTIGREW, CLIFFORD S. PATLAK, AND JOSEPH D. FENSTERMACHER. Variation in local cerebral blood

flow response to high-dose pentobarbital sodium in the rat. Am. J. Physiol. 261 (Heart Circ. Physiol. 30): HllO-H120, 1991.Microvascular bed structure and functions are known to vary throughout the brain. Microvascular responses to high doses of pentobarbital sodium might therefore differ among brain areas. This possibility was examined by measuring local cerebral blood in 52 brain areas at 5, flow (LCBF) with [‘*Cl* 10d oantipyrine 10, 25, and 60 min after intraperitoneal administration of pentobarbital (50 mg/kg). From 5 to 60 min, LCBF was significantly lowered in 17 of 25 forebrain gray matter areas but in only 1 of 18 hindbrain gray matter structures, the pontine nuclei. Smaller, shorter duration lowering of LCBF was also observed in ten other brain areas. In both control and treated rats, LCBF was found to vary within individual brain structures. The pattern of these LCBF variations was columnar in the cerebral cortex and the hippocampus but was patchy in the caudate-putamen, thalamus, and inferior colliculus. These results indicate that pentobarbital anesthesia more strongly alters LCBF in the forebrain than in the hindbrain and produces different patterns of changes in LCBF than in local cerebral glucose utilization, which was measured with 2-deoxyglucose in a companion study. blood flow regulation; organs

forebrain;

hindbrain;

circumventricular

SINCE THE DEVELOPMENT of the quantitative autoradiographic technique (QAR) by Sokoloff and co-workers (5, 15, Zl), measurements of local cerebral blood flow (LCBF) and of local cerebral glucose utilization (LCGU) have indicated that blood flow and glucose metabolism differ greatly among and within central nervous system (CNS) areas in normal animals (15, 23, 26). In certain experimental conditions, e.g., after ketamine or nicotine exposure or during auditory stimulation, LCGU and LCBF are known to change in some, but not all, brain areas (3,8,9, 16). Accordingly, the linkage among neural activity, metabolism, and blood flow must occur at the local level, and the coupling among these processes and their responses to drugs and other perturbations may vary throughout the brain areas. Barbiturates have been used as anesthetics and suppressors of neural activity for many years and are known to reduce cerebral blood flow; the latter reduction prob-

ably occurs mainly through a linkage to local metabolism (13, 17). Studies of the local response to barbiturates have, however, been limited (6, 15, 25). In an attempt to understand local coupling among blood flow, neural activity, and metabolism and the functional organization of the cerebrovascular system, we have postulated that the alterations in LCBF produced by pentobarbital sodium vary throughout the brain and that these changes may be highly localized, that is, may involve brain areas as small as hypothalamic nuclei and layers, columns, and zones within brain structures. In this study LCBF was measured in 52 brain areas of awake (control) and treated rats. The estimates of LCBF were made in the treated rats at 5, 10, 25, and 60 min after pentobarbital administration. Blood flow was determined by the [ 14C]iodoantipyrine (IAP) technique of Sakurada et al. (23). Tissue levels of 14C activity were assayed by QAR. These data were then used to 1) test the above hypothesis, 2) establish a cerebrovascular “steady-state” period for future studies with pentobarbital anesthetized rats, and 3) compare with LCGU findings from a companion investigation. METHODS Animal preparation. Male Sprague-Dawley rats weighing between 270 and 360 g were anesthetized with a mixture of halothane, nitrous oxide, and oxygen. The femoral artery and vein were catheterized on both sides. Catheters (PE-50 tubing) were also inserted into the peritoneal cavity through an abdominal incision in all animals. The wounds were infiltrated with lidocaine hydrochloride and closed with sutures. After surgery, the administration of anesthesia was stopped and the animal was rapidly wrapped in a loose-fitting plaster cast that extended from midthorax to lower abdomen. The rats were allowed 22 h to recover from surgery and anesthesia before beginning the measurement of blood flow. The physiological status of the animals was determined during the recovery period, immediately before beginning the experiment, and during the experiment. At these times arterial blood samples were obtained and analyzed for blood gases, hematocrit, plasma glucose concentration, and plasma osmolality. Rectal temperature was continuously measured and maintained between 36.5 and 38°C using a heat lamp. Arterial blood pressure

HllO

Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (137.154.019.149) on January 12, 2019.

VARIATIONS

IN

LCBF

RESPONSES

and heart rate were monitored throughout the recovery and experimental periods. Treatment groups. Thirty-two rats were randomly divided into five groups of six to seven animals each. The control group received 0.5 ml of saline intraperitoneally before the measurement of LCBF and are referred to hereafter as controls. In contrast to the treated animals, the control rats were awake during the experimental period. For the other four rat groups, 0.5 ml of saline containing pentobarbital sodium (50 mg/kg) was infused through the intraperitoneal catheter. Blood flow measurements were made at 5 (n = 7), 10 (n = 6), 25 (n = 6), and 60 (n = 6) min after pentobarbital administration. Measurement of LCBF. LCBF was measured using the [ ‘“C]IAP technique of Sakurada et al. (23). Labeled IAP, in the form of 4[N-methyl-l*C]iodoantipyrine, was obtained from American Radiochemical (St. Louis, MO). Radiotracer purity was >98%. Before beginning the IAP infusion, an extracorporeal arteriovenous (a-v) shunt was formed by shortening the femoral artery and vein catheters on one side and connecting them with a 1.5-cm length of silicone rubber tubing (No. 2030-969; LKB Products, Sweden). At the start of the experimental period (t = 0), an intravenous infusion of -50 &i of [‘“c]IAP was begun. The rate of infusion was increased during the experiment according to a schedule that yielded a linearly rising concentration of IAP in the blood. Such a blood time course is best for accurate measurement of LCBF with this technique (19). Arterial blood samples were obtained every 5 s (starting at t = 0) by puncturing the silicone tube portion of the a-v loop with a 22-gauge needle on a plungerless syringe and collecting 60-80 ~1 of blood, which were driven into the syringe over the next 2-4 s by the “arterial” pressure in the loop. Radioactivity in the plasma obtained from each blood sample was measured by liquid scintillation counting. Thirty seconds after beginning the IAP infusion, the rat was decapitated. The brain was then rapidly removed, frozen in 2-methylbutane (Fisher Scientific, Pittsburgh, PA) cooled to -45”C, and stored in a sealed plastic bag at -8OOC. The time from decapitation to brain immersion in 2-methylbutane ranged from 30 to 40 s. The physiological and plasma radioactivity data were used to screen all experiments. If any of the monitored physiological variables was abnormal before beginning the experiment (i.e., was outside of the means t 2 SD of our historical controls) or changed appreciably during the infusion, then the experiment was discarded without further analysis. If the experiment was acceptable (virtually all were at this point), then the time course of IAP in the blood was examined. Because prior experience indicated that blood concentration rose linearly over time when the infusion of IAP and the sampling of arterial blood were properly done, the blood data were objectively tested for 1.inearity by the foll .owing procedure. A regression line through all the concentration-time points beyond 5 s was determined and the variability of the data points around the line (VL) was evaluated with the following equation VL = [(n - l)-lz

(y - Y)2]1’2/Y(25 s)

TO

PENTOBARBITAL

Hlll

where y is the measured plasma radioactivity at each time (t), Y is the “plasma” radioactivity on the regression line at the same time, n is the number of data points used, and Y(25 s) is the value of Y at 25 s. Based on data from 15 control rats, poor blood sampling or improper [‘*Cl IAP infusion yielded VL values >O. 1. An experiment was therefore dropped if VL was >O.l. The frozen rat brains from the acceptable experiments were serially cut into 20-pm-thick coronal sections at -17°C in a cryostat. The sections were picked up on cover slips, dried at 45-50°C on a hot plate, and placed into X-ray cassettes along with a set of [l*C]methylmethacrylate standards (American Radiochemical) and X-ray film (SB-5, Kodak, Rochester, NY). Tissue sections were taken at regular intervals, stained with cresyl violet, and prepared for histological examination. After 5-7 days of exposure, the X-ray films were developed and analyzed for radioactivity using an image analyzing system (MCID, Imaging Research, St. Catherines, Ontario, Canada). A curve relating optical density to radioactivity was constructed from the ‘*C standard data. The radioactivity of each structure was obtained by averaging 3-18 readings from a set of three adjacent autoradiographic images. LCBF was calculated from the average ‘*C radioactivity of each structure, the time course of plasma [‘“CIIAP, and the operational equation of Sakurada et al. (23). The tissue-blood partition coefficient was set at 0.8 for all brain areas in accordance with the measurements of Sakurada et al. (23). Densitometric readings and LCBF determinations were made for 52 brain areas. These brain areas were identified by using a rat brain atlas (20) and comparing the autoradiographic images with adjacent cresyl-violet-stained brain sections. Statistical analysis. The significance of the differences in the physiological measurements between the control and pentobarbital-treated groups was assessed with Student’s t test. The significance of the LCBF differences between the control and the experimental groups for each brain area was tested using the t test for pooled samples with a Bonferroni correction for multiple comparisons. Statistical significance was assumed when P c 0.05. Repeated measures analysis of variance was used to test for differences in the entire set of LCBF data among the various groups. All values are reported as the means t SE. RESULTS

Physiological status. Before the intraperitoneal administration of pentobarbital or saline, the physiological data were similar for the control and experimental animals (Table 1). The physiological condition of the experimental animals was also determined after pentobarbital administration. Arterial blood pressure (means t SE in mmHg) in the anesthetized rats at the time of measuring LCBF was the following: 5 min, 91 t 3 (n = 7); 10 min, 92 t 3 (n = 6); 25 min, 104 t 4 (n = 6); 60 min, 107 t 8 (n = 6). Mean arterial blood pressure at 5, 10, and 25 min was significantly less than control (P c 0.01). Some of the pH, Pco~, and POT data from the experimental groups were misplaced after the initial review of

Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (137.154.019.149) on January 12, 2019.

H112

VARIATIONS

IN

LCBF

RESPONSES

Pentobarbital-Treated

BT, “C Hct, % Osm, mosmol/kg Glucose, mg/dl MABP, mmHg

Control, n=7

37.ltO.l 50&l 288t2 167k13 125t4

5, n=7

37.2t0.2 48&l 294t2 168klO 129t4

Values are means ~fr SE. BT, body Osm, osmolality; MABP, mean arterial

10, n=6

Group,

min

60,

25, n=6

n=6

36.7k0.2 50&l 291k2 170215 129s

36.8t0.2 47&l 289k2 165tll 120&3

temperature; Hct, blood pressure.

hematocrit;

36.9t0.2 51t1 29224 157t9 124t3

the physiological variables and cannot be reported. The “extant” data plus data taken from other experiments done at two of the same times (10 and 60 min) after pentobarbital administration give a fair indication of pH, Pco~, and PO, condition of these animals. The pH values (means t SE) are control, 7.48 t 0.01 (n = 7); 5 min, 7.46 t 0.01 (n = 3); 10 min (extant), 7.41 t 0.01 (n = 3); 10 min (other group), 7.4 t 0.01 (n = 9); 25 min, 7.41 t 0.01 (n = 4); 60 min (extant), 7.35 (n = 1); 60 min (other group), 7.4 t 0.01 (n = 10). The PCO~ values (mmHg) are control, 39.3 t 0.4 (n = 7); 5 min, 39.1 t 2.5 (n = 3); 10 min (extant), 45.2 t 1.2 (n = 5); 10 min (other group), 46.1 t 1.2 (n = 9); 25 min, 44.4 t 2.3 (n = 4); 60 min (extant), 51 (n = 1); 60 min (other group), 46.0 t 1.2 (n -- 10). The Po2 values (mmHg) are control, 88.6 t 2.1 (n = 7); 5 min, 77.1 t 0.7 (n = 3); 10 min (extant), 73.9 t 2.4 (n = 5); 10 min (other group), 76.8 t 2.7 (n = 9); 25 min, 77.7 t 3.9 (n = 4); 60 min (extant), 71 (n = 1); 60 min (other group), 76.0 t 3 (n = 10). Overall these data indicate that pH is -7.4 from 10 to 60 min after pentobarbital administration, PCO~ is high (44-46 mmHg) from 10 to 60 min, and PO:! is low (74-78 mmHg) from 5 to 60 min. Within 5 min of administering pentobarbital, and thereafter, the cornea1 reflex and the response to toe pinching were absent. These observations indicate that cardiovascular, respiratory, and nervous functions were depressed throughout the various experimental periods but that mean arterial blood pressure was still in the socalled autoregulatory range for the rat, namely, 80-160 mmHg (10). Local blood flow response. LCBF was measured in 52 brain areas. Within the hindbrain (brain stem plus cerebellum), LCBF was measured in 18 gray matter areas and one white matter area. These 19 hindbrain structures are perfused by the vertebrobasilar system (posterior cerebral circulation). Within the forebrain, LCBF was measured in 25 gray matter and 3 white matter structures. Most of these 28 forebrain areas are perfused by the carotid system (anterior cerebral circulation); however, some (e.g., the occipital cortex) are served by the posterior circulation. Blood flow was also estimated in five of the circumventricular organs. The rates of LCBF in the 52 brain areas were very similar between the control rats and a group of six normal, awake Sprague-Dawley rats that did not have intraperitoneal catheters and were studied during the same period for another proiect (Fig. 1). The placement

PENTOBARBITAL

of intraperitoneal catheters therefore seems to have no effect on cerebral blood flow. In the gray matter areas of the hindbrain of control rats, LCBF varied from 98 (cerebellar cortex) to -310 (superior olive and inferior colliculus) ml. 100 8-l. min-’ (Table 2). Among these brain areas, pentobarbital significantly lowered LCBF at all experimental times only in the pontine nuclei (Table 2). Blood flow was appreciably reduced relative to controls (P < 0.05) in the inferior colliculus at 5 and 10 min and in the inferior olive and the intermediate gray layer of the superior colliculus at 10 min. In brain areas such as the nucleus of the solitary tract and the nucleus of the facial nerve, LCBF was very similar in control and treated rats. In forebrain gray matter areas of control rats, LCBF ranged from 84 (periventricular nucleus) to -285 (dorsomedial and ventrolateral parts of anteroventral nucleus; lateral habenular nucleus; temporal cerebral cortex) ml 100 8-l. min-’ (Table 3). In contrast to the cerebellum and brain stem, LCBF was significantly lowered by pentobarbital in 18 of the 25 gray matter structures in all four experimental groups and in two additional areas in the lo-, 25-, and 60-min groups. Pentobarbital appeared to decrease LCBF in the arcuate and supraoptic nuclei of the hypothalamus and the lateral habenular nucleus, but these changes were not statistically significant. The greatest reductions in LCBF were observed throughout the thalamus (generally 30-40% of control, Tables 3 and 4) and the cerebral cortex (as low as 25% of control). The lowest rates of blood flow in both control and treated rats were found in white matter (Fig. 2). Small but significant decreases in LCBF were induced by pentobarbital in the pyramidal tract (hindbrain) and the internal capsule (forebrain) at 10 and 25 min and in the ventral hippocampal commissure (forebrain) at 10 min. Blood flow was slightly but not significantly lowered in any of the four white matter areas at either 5 or 60 min

1. Physiological variables in control and treated rats before intraperitoneal injection TABLE

Variable

TO

l

2 E 300 \

11

OY 0

100

UNOPERATED

I

I1

200

300

1

CONTROL

(ml/

lOOg/min)

1. Plot of local cerebral blood flow (LCBF) of control (awake rats that received intraperitoneal injection of saline 10 before flow measurement, n = 7) vs. LCBF of control rats (awake without intraperitoneal catheters, n = 6) for 47 noncircumventricular organ brain areas. Diagonal line has slope of 1 and intercept of 0 of identity). Differences in LCBF rates between 2 groups were significant for any of 47 areas. FIG.

4

rats min rats (line not

Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (137.154.019.149) on January 12, 2019.

VARIATIONS

TABLE

IN

LCBF

RESPONSES

TO

H113

PENTOBARBITAL

2. LCBF in 18 gray matter areas of hindbrain Brain Areas

Medulla SOL HPG IOV Pons DCN MVN FAC sov LLP PON Cerebellum CBC NIC Midbrain IC scs SC0 SC1 OCM SNR IPN

Pentobarbital-Anesthetized Control

Rats,

min

n 5

n

10

n

25

n

60

n

123klO 161k9 177t11

7 7 7

124klO 174&E 155213

7 7 7

117k12 147216 115ub

6 6 6

127t7 159t11 151k3

4 4 5

125k13 167k16 162k16

6 6 6

291t13 279t23 172-e6 313t20 264k27 163-elO

7 7 7 7 7 7

289t31 433k63 196215 434t118 221-el5 9Ok5”

7 7 7 7 7 7

202t24 268k36 183t14 326t40 191t21 7928”

6 6 6 5 6 5

334k36 504k82 184tlO 366-r-24 285t33 85t4”

5 5 5 5 4 4

363-+-53’ 407k55 185223 431k82 227k39 83klO”

6 6 6 6 6 6

98k7 224&12

7 7

76k4 234t25

7 7

7Ok6 201t37

6 6

71k5 292t22

5 5

75t10 222k24

6 6

31Ok16 145t7 144k6 144t8 229t27 lllk5 262t19

7 7 7 7 7 7 7

185+5b 209k30 194t30 142k18 186t10 81k5 337k49

7 7 7 7 7 5 5

175t27” 103k16 95t16 79t8” 175k16 9OklO 284t22

6 6 6 6 6 6 5

383t90 194t38 182k37 125k16 243&49 81t7 358239

4 6 6 6 5 5 5

257&63 159k31 140t28 108t16 169t26 88t15 374k64

6 5 5 5 5 6 6

d

P values are Bonferroni corrected for multiple comparison t tests. LCBF, local cerebral Values are means k SE in ml. 100 g-’ 100 min. blood flow; SOL, solitary tract nucleus; HPG, hypoglossal nucleus; IOV, inferior olive; PON, pontine nuclei; DCN, dorsal cochlear nucleus; MVN, medial vestibular nucleus; FAC, facial nucleus; SOV, superior olive; LLP, lateral lemniscus; CBC, cerebellar cortex; NIC, nucleus interpositus; IC, inferior colliculus; SCS, superior colliculus, superficial gray; SCO, superior colliculus, optical layer; SCI, superior colliculus, intermediate gray; OCM, oculomotor nucleus; SNR, substantia nigra, reticular; IPN, interpeduncular nucleus. a P < 0.05, " P c 0.01, ’ P < 0.001, d P < 0.05, significantly different from 5-min group; e P c 0.05, significantly different from lo-min group. significantly different from control; l

after initiating pentobarbital anesthesia. Among the five circumventricular organs (CVOs), the control LCBF estimates ranged from 125 (subfornical organ) to 459 (neural lobe) ml 100 8-l. min-’ (Fig. 3). Pentobarbital reduced LCBF in the pineal gland after intraperitoneal administration, but this decrease was statistically significant only for the lo-min group. Of the 30 brain areas where appreciable flow changes were observed, LCBF was significantly lower in 18 of them at all four times of measurement (Table 4); this was seen for the pontine nuclei, the medial mammillary nucleus, all thalamic nuclei, the three parts of the caudate-putamen, the hippocampus, and the four regions of the cerebral cortex. Moreover, for 17 of these 18 brain areas, LCBF was fairly similar at 5, 10, 25, and 60 min after pentobarbital administration (Table 4)‘. Statistically, this is indicated by the lack of significant differences among the time groups for these 17 brain areas (Tables 2 and 3). For the sensorimotor cortex, however, LCBF was significantly greater at 5 than at 25 min (Table 3). By use of repeated measures analysis of variance, the “profile” of LCBF values for the 47 non-CVO brain areas was determined to be statistically similar for the four experimental groups; thus the effect of pentobarbital on LCBF was the “same” at 5, 10, 25, and 60 min after intraperitoneal injection. Overall, intraperitoneal pentobarbital lowers blood flow and maintains it fairly constant in these 18 brain areas from 5 to 60 min after administration and in the ventromedial hypothalamic nucleus and paraventricular nucleus, magnocellular, from 10 to 60 min. The experimental LCBF/control LCBF ratios varied widely among major brain divisions. The mean values of

these ratios for the hindbrain divisions were 0.94 (medullary gray matter), 1.03 (pontine gray matter including the pontine nuclei), 0.90 (cerebellum), and 0.96 (midbrain gray matter) and for the forebrain divisions were 0.64 (hypothalamic gray matter), 0.39 (thalamic gray matter), 0.53 (basal ganglia), 0.62 (hippocampus), and 0.35 (cerebral cortex). The lowest LCBF ratios were found in the high flow areas of the cerebral cortex [ratio, 0.22 for the temporal cerebral cortex at 10 min (Table 4)] and the thalamus [ratio, 0.28 for the anteroventral nucleus (dorsomedial and ventromedial) at 5 min (Table 4)]. The impact of pentobarbital on LCBF was thus much greater on the high flow areas of the forebrain than in the high flow areas of the brain stem and cerebellum. This difference in response between high flow areas of the forebrain and hindbrain can be seen readily from plots of treated vs. control LCBF data. For example (Fig. 4), the lo-min points for all forebrain areas with control flow rates between 200 and 300 ml. 100 g-l min-’ lie much farther below the line of identity than do the lomin points of the brain stem and cerebellar areas with control flow rates in this same range. The differences between forebrain and hindbrain are much less for areas with control LCBFs