Journal of Neuroscience Research 26:188-195 (1990)

Brain Regional Distribution of Physostigmine and Its Relation to Cerebral Blood Flow Following Intravenous Administration in Rats O.U. Scremin, A.M.E. Scremin, S.M. Somani, and E. Giacobini Veterans Administration Medical Center, Albuquerque, New Mexico (O.U.S., A.M.E.S.); Department of Pharmacology, Southern Illinois University School of Medicine, Springfield, Illinois (S.M.S., E.G.)

'H-labeled physostigmine (50 pg. kg- ') was administered intravenously to rats, and its concentration in brain tissue and spinal cord was assessed by quantitative autoradiography. Regional cerebral blood flow (rCBF) was measured with i~do-'~C-antipyrine autoradiography in control rats and in animals injected i.v. with a dose of physostigmine similar to that used for the distribution studies. Tissue concentration of 3H-physostigmine was correlated with rCBF for 37 brain regions. A high degree of correlation was found at 0.5 min after drug injection, r (correlation coefficient) = 0.87. This association decreased at later times (5 min r = 0.73, 12 min r = 0.24). Structures with high cholinesterase activity (caudate-putamen, amygdala, hippocampus) showed greater retention of physostigmine over time. The highest initial physostigmine concentrations were found in regions lacking a blood-brain barrier (pineal gland, median eminence, choroid plexus) (range = 10.4-23.8 nCi/mg) and the lowest in white matter (corpus callosum, internal capsule, hippocampus commisure, spinal cord dorsal column) (range = 1.2-2.6 nCi/mg). Initial concentrations of the drug in the areas in which physostigmine induced vasodilatation (motor, sensory, temporal and occipital cortex, claustrum, and superior collicullus) were not different from concentrations in areas of comparable basal rCBF in which no such effect was observed. Variations in drug access to brain regions, then, do not explain the topographical variations of the cerebrovascular action of physostigmine. Key words: cholinesterase inhibitors, drug distribution, cholinergic agents, autoradiography

Khalique, 1987) and dogs (Mattio et al., 1986). This drug is known to increase cerebral blood flow markedly, without a concomitant increase in metabolism (Scremin et al., 1973, 1982b). In rats, the physostigmine-induced vasodilatation is prominent in neocortex and associated areas (Scremin et al., 1987, 1988). It is expected that the distribution of physostigmine in the brain will be a function of local blood flow. There is experimental support for this concept (Somani and Khalique, 1987). The present experiments try to extend those studies by comparing the regional variations of local blood flow and physostigmine concentration with autoradiographic techniques that allow a high degree of spatial resolution. It was also hoped that this study could determine whether a differential permeability of the blood-brain barrier to physostigmine was responsible for the regional distribution of its cerebrovascular effect. This study was carried out in conscious rats in order to determine the concentration of physostigmine in different regions of the brain and its relationship to regional cerebral blood flow after i.v. administration of 3H-labeled drug.

MATERIALS AND METHODS Regional cerebral blood flow (rCBF) and regional concentration of 'H-physostigmine (rPC) were studied in male Sprague-Dawley adult rats (350-400 g body weight) with the autoradiographic iodo-'4C-antipyrine (IAP) technique (Sakurada et al., 1978) and quantitative dry autoradiography , respectively. Animals were implanted with arterial and venous catheters in the iliac vessels under halothane anesthesia (2.4% in air induction and 1.7% maintenance) and, after discontinuation of

INTRODUCTION Physostigmine is a cholinesterase inhibitor that induces central cholinergic effects. Its pharmacodynamics and pharmacokinetics following intravenous administration have been recently characterized in rats (Somani and 0 1990 Wiley-Liss, Inc.

Received October 4, 1989; revised November 17, 1989; accepted November 23, 1989. Address reprint requests to Oscar U. Scremin, MD, PhD, Research Service (151), Veterans Administration Medical Center, 2100 Ridgecrest Dr. S.E., Albuquerque, N M 87109.

Physostigmine Distribution in Brain

189

halothane, allowed to recover in a restraining device and cut in a cryostat in 20-pm-thick slices. These sections were then heat-dried on glass slides and exposed to (Bollman cage) for 2 hr. The rCBF technique was implemented by continu- Kodak AR-XOMAT film for I4C and LKB Ultrofilm for ous i.v. infusion of 0 . 6 ml saline containing 100 'H autoradiography , along with eight radioactive stanpCi*kg-' body weight of the tracer (i~do-'~C-methylan-dards. The resulting autoradiographic images of corona1 tipyrine, specific activity 58 mCi.mmol-'; Amersham brain sections contained enough anatomical information Corp.) over 30 sec. Timed blood samples were obtained to permit identification of regions by comparison to an every 2-3 sec from a free-flowing arterial catheter anatomical atlas of the rat brain (Paxinos and Watson, throughout the infusion time and processed for liquid 1982). Since there was no tissue shrinkage, accurate scintillation counting of radioactivity in a Packard Tri- measurements could be obtained in these specimens to Csrb liquid scintillation spectrometer. At the end of the define the designated regions. Details on regions' defi30 sec period, the heart was arrested with a bolus of nitions and anatomical coordinates are given in the legT-6 1 euthanasia solution-N-(2-[m-methoxy-phenyl] 1)- ends to Figures 2-4. Optical density of the films was gamma-hydroxybutiramide, 200 mg-ml- I ; 4,4'-methyl- determined with a Sargent Welch Densichron spot denene-bis(cyclohexy1-trimethyl-ammonium iodide), 50 sitometer, fitted with a sampling aperture of 0.2 mm in mg . m1-I; tetracaine hydrochloride, 5 mg ml-l (Amer- diameter. In some cases, the brain image was digitized ican Hoechst Co.)-infused through a second venous by a system consisting of a densitometer (Photovolt Dencatheter. This produced a precipitous fall of arterial sicord) positioned on the optical axis of a microscope. blood pressure within 3 sec. The exact timing of this This instrument was fitted with a motorized scanning event was determined from a continuous record of iliac stage driven by a Kaypro PC-I0 microcomputer. The artery blood pressure. rCBF was measured in four groups densitometer was interfaced to the computer with a of animals: controls (n = 5) = injected with saline; 0.5 DASH-8 Metrabyte analog-to-digital converter. The sysmin physostigmine (n = 3) = simultaneous i.v. infusion tem was programmed to generate linear scans or colorof IAP and physostigmine over 0.5 min; 5 min physo- coded maps of regional blood flow of physostigmine stigmine (n = 5) = i.v. infusion of physostigmine over concentration. The radioactivity level (essentially physostigmine 0.5 rnin followed 5 min later by the rCBF procedure; and 12 min physostigmine ( n = 5) = i.v. infusion of phy- in brain) measured from autoradiographs obtained at sostigmine over 0.5 min followed 12 min later by the timed intervals after 'H-physostigmine administration rCBF procedure. The dose of physostigmine was, in all was plotted against time in all regions of the brain. For those regions that showed declining radioactivity, the cases, 50 pg-kg-'. The determinations of rPC were performed by con- half-life (T1,2, min) and the rate of eliminiation (Khr, tinuous i . v. infusion of 'H-labeled physostigmine, spe- min- I) of physostigmine were determined by a leastcific activity 13 Ci-mmol-I, custom synthesized by Am- squares logistic regression analysis of log tissue radioacersham Corporation (Chicago, IL). Physostigmine is tivity on time. labeled with tritium on its aromatic ring, at both ortho positions to the carbamate side chain. The labeled drug was kept dissolved in ethanol at -80°C. Immediately RESULTS prior to use, the alcohol was evaporated under a stream Local variations in the concentration of 'H-physoof nitrogen passed through a molecular sieve, and the stigmine that translated into shades of gray in autoradiodrug was redissolved in saline. The equivalent of 50 grams allowed the imaging of the main anatomical repg.kg-' of physostigmine was infused i.v. over 0.5 gions in brain and spinal cord shown in Figure 1. min. The animals were sacrificed by i.v. pulse injection Animals were injected i.v. with 3H-labeled physostigof T-61 solution at 0.5 min (end of drug infusion), 5 min mine as described in Materials and Methods, the circuor 12 min after commencement of drug infusion. In the lation was stopped, and the brains were removed 0.5 last two groups there was then a period of 4.5 or 11.5 (left), 5 (middle), and 12 min later (right). The last row min between the end of physostigmine infusion and eu- corresponds to transverse sections of the cervical spinal thanasia. Samples of arterial blood were obtained from a cord. Quantitation of optical density of autoradiograms free-flowing catheter every 2 sec during infusion and and comparison with that induced by standards of known every minute afterwards until the experiment was con- radioactivity allowed determination of 'H activity in cluded. The exact time of circulatory arrest was obtained brain regions depicted in Figures 2-4. It is apparent that from a continuous record of arterial blood pressure. the relative optical densities of cerebral cortex and cauAt the end of the rCBF or rPC procedures, the date-putamen in autoradiograms change with time after brains were rapidly removed and frozen in methylbutane injection. At 0.5 min, 'H activity (higher density) preat -70°C. They were subsequently warmed to -20°C dominates in the cortex; at 12 min the opposite is true

190

Scremin et al. CINGULATE CORTEX MOTOR CORTEX CORPUS CALLOSUM SENSORY CORTEX CAUDATE-PUTAMEN SEPTUM

A

OLFACTORY TUBERCLE HIPP. COMMISSURE CAUDATE- PUTAMEN GLOBUS PALLIDUS

B

HIPFOCAMPUS THALAMUS

C

MAMILLARY BODIES

AREA 1 B A TEMPORAL CORTEX CENTRAL GREY HIPPOCAMPUS

D

PINEAL GLAND INFERl OR COLLl CU LLU S MEDIAL RAPHE

E

DORSAL COLUMN DORSAL HORN LATERAL COLUMN

F

VENTRAL HORN ANTERIOR COLUMN

0.5 min.

5 min.

12 min.

Fig. I . Autoradiographs obtained from rat brains and cervical spinal cords. The Animals were injected i.v. with 'H-labeled physostigmine as described in the text, and the circulation was stopped and the brains were removed 0.5 (left), 5 (middle), and 12 min (right) later.

(rows A and B). The same can be observed in rows C and D with the cerebral cortex, which appears darker than the hippocampus at 0.5 min and lighter at 12 min. The activity within the hippocampus appears uniform at 0 . 5 min, but at 5 and 12 min the inner and outer borders stand out with greater density. The inferior colliculus (row E) shows a remarkable change in activity from 0 . 5 min, when it is the region of highest density, to 12 min, when it ranks among the lowest. The pineal gland (not

shown for the 5 min specimen) stands out in row E as the region with highest activity in the entire brain. Activity in the cervical spinal cord is highest in gray matter and is just discernible from background in white matter. Immediately after completion of physostigmine infusion, rCBF did not show a significant change with regard to controls in any of the regions studied. For that reason, values of both groups were pooled together (Figs. 5-7). At 5 min after physostigmine, a large in-

Physostigmine Distribution in Brain Regional 3H-Physostigrnine

concentration

191

Regional 3H-Physostigmine concentration Variable times after i.v. injection

Variable t i m e s a f t e r i.v. i n j e c t i o n .

---

j

Y

0'01

'=0 3

0.G

-

,

.

-

00

Fig. 2. Radioactivity values obtained from densitometric measurements of autorddiograms. Values represent means of three determinations per animal in two animals per time period. Region name (graph abbreviation) and stereotaxic coordinates (mm) of the region’s center after the anatomical atlas of Paxinos and Watson ( 1982) (bregma coordinates: coronal plane, horizontal coordinate, distance from pial surface) are: frontoparietal cortex, motorarea (motor) -0.3,3,0.7 and - 1.3,3,0.7; frontoparietal cortex, somatosensory area (somatos) -0.3,5.5,0.7 and - 1.3, 5.5,0.7; temporal cortex, auditory area (temporal) -5.8,7,0.7; striate cortex, area I7(AREA 17) -5.8,4,0.7; striate cortex, area 18 (Area 18) -5.8,2,0.7; striate cortex, area 18a (AREA 18A) -5.8,6,0.7; primary olfactory cortex (olfactory) 1.3,5.5,0.7; anterior cingulate cortex (cingulate) 0.7,0.5,0.7; For the remaining two regions, coordinates listed are (plane, horizontal coordinate, vertical coordinate): field CAI of Ammon’s horn (hip.[CAl]) -5.3,3.5,3; field CA2 of Ammon’s horn (hip.[CA2]) -5.3,5.5,6. In motor and sensory cortex, an average was obtained from the two regions listed.

Fig. 3. Values obtained as in Figure 2 for a number of subcortical areas. Region name (graph abbreviation) and stereotaxic coordinates (mm) of the region’s center after the anatomical atlas of Paxinos and Watson (1982) (bregma coordinates: coronal plane, horizontal coordinate, vertical coordinate) are: claustruin (claustrum) 2.7,2,5; nucleus basalis (n basalis) 1.3,3,8; caudate-putamen (caudate-p) 0.7,2.5,6; anteromedial thalamic nucleus (ant thalam) - 1.3,0.5,6.2;posterior thalamic nuclear group (post thalam) -4.3,2,5.5; lateral preoptic area (preoptic a) -0.3,1.4,8.3; lateral hypothalamic area (lat hypoth) 1.3,2.2,8.8; dorsal hypothalamic area (dor hypoth) -2.3,0.5,8.5; dorsomedial hypothalamic nucleus (d m hypoth) -3.3,0.5,9; medial mammillary nucleus (m mamill n) -4.8,0,9; central amygdaloid nucleus (amygdala) - 1.8,3.8,8; interpeduncular nucleus (interped n) -5.8,0,9; superficial grey layer of the superior collicullus (s sup coll) -6.8,1,3; deep grey layer of the superior colliculus (d sup coll) -6.8,1,4.2; central nucleus of the inferior collicullus (infer coll) -8.8, I .5,4.4; pontine reticular nucleus, oral part (pont retic) -7.8, I .3,8.5; cerebellum (cerebellum) - 10.3,0,5.

crease in rCBF was observed in all areas of neocortex, claustrum, and superior colliculus. A small increase, marginally significant, was observed in olfactory cortex. rCBF in all these areas decreased with regard to 5 min, when measured at 12 min after physostigmine (Figs. 57). It was still higher than control in all regions but temporal cortex (Figs. 5 , 6). A significant, although small, increase in rCBF was observed in anterior thalamus and pontine reticular nuclei (Fig. 6) 12 min after drug infusion. At the end of 0.5 min of i.v. administration, the pattern of ‘H tissue radioactivity resembled closely that of blood flow distribution calculated from I4C-IAP autoradiograms. A significant correlation between these two variables was found (r = 0.87, P < 0.001) (Fig. 8) Paired rCBF and rPC data for all regions sampled (with

the exception of those lacking a blood-brain barrier) were fitted to a polynomial regression of the form

~

-

-

rPC

=

A

+ B . rCBF + C . (rCBF)2

Least-squares regression analysis yielded the following values and standard errors (S.E.) for the coefficients: A = 0.729, S.E. = 0.699;B = 4.97, S.E. = 0.927;and C = 0.693, S.E. = 0.268. Some structures showed a progressive decay in radioactivity over time (Figs. 2-4). The rate of elimination (Khr) and half-life (TIJ of physostigmine were calculated in these areas (Table I). Correlation coefficients for the decay slopes ranged from -0.95 to - 1 . Half-life of physostigmine in these regions ranged from 4.3 to 10 min. No correlation was observed between K,, and rCBF

Scremin et al.

192

Regional 3H-Physostigmine concentration Vorioble times after i.v. injection 0 0.5 min

6Y J m i n

I 12min

30.0

25.0

1A

*

7

I

.-C

6.0 t

E

*

7

I m

*

-

E

Fig. 4. Values obtained as in Figures 2 and 3. Region name (graph abbreviation) and stereotaxic coordinates (mm) of the regions center after the anatomical atlas of Paxinos and Watson ( 1982) (bregma coordinates: corona1 plane, horizontal coordinate, vertical coordinate) are: pineal gland (pineal g.), no coordinates; medial eminence (med. em.) -2.8,0,10; choroid plexus of the third ventricle (chor pl.) -2.3,0,4.5; periaqueductal grey (per gr.) - 7.3,0.5,5.5; medial geniculate body (m. genic.) -5.3,3.5,6; medial raphe nucleus (m. raphe) -7.3,0,8.5; pontine nuclei (pont. n.) -7.3,1,10; laminae I and 11 of the cervical spinal cord (lam. 1-11) no coordinates; ventral horn of cervical spinal cord (v. horn) no coordinates; corpus callosum (c. call.) 0.2,0,3.5; internal capsule (int. caps.) - I .3,2.5,4; ventral hippocampus comrnissure (hip. com.) - I .3,0,4; dorsal column of cervical spinal cord (dor. clm.) no coordinates.

(averaged between 0.5 and 12 rnin after administration) (r = 0.06). In contrast, occipital and olfactory cortex, hippocampus, nucleus basalis, caudate-putamen, thalamus, hypothalamus, and periaqueductal gray showed stable or increasing physostigmine concentration at 5 min after administration. The same behavior was observed in all white matter structures. Physostigmine concentration in hippocampus, caudate-putamen, and amygdala showed a minimal change between 0.5 and 12 min after drug administration. As a consequence of the different rate of disappearance of radioactivity among regions, the correlation of 'H radioactivity with blood flow observed initially decreased progressively at 5 min (r = 0.73, P < 0.001) and 12 min (r = 0.24, P n.s.) after i.v. administration (Fig. 8). Coefficients for the polynomial regressions, calculated as described above were for 5 min: A = 2.68, S.E. = 0.565; B = 1.87, S.E. = 0.509; C = -0.209,

Fig. 5 . Regional cerebral blood flow measured with the autoradiographic IAP technique in the same regions listed in Figure 2. Rats that did not receive physostigmine (0 min, n = 5 ) and those that had received the drug for 0.5 rnin before the completion of the rCBF procedure (n = 3) did not show differences in any of the regions listed in this or following figures. For that reason, values were pooled in a single group (010.5 min, n = 8). In the other two groups, animals received physostigmine either 5 min (n = 5 ) or 12 min (n = 5 ) before rCBF determination. Statistical significance of differences between means of the 010.5 min group and those of 5 and 12 rnin was tested by the Bonferroni procedure. * P < 0.05; **P < 0.025; ***P < 0.01. Regionol Cerebrol Blood Flow

Varioble times ofter Physostigrnine Inj. 0 0/0.5 min

E3 5 min

I 12 min

7

I

.-C E

-*

4.0

r _

Fig. 6. Regional cerebral blood flow measured in animal groups as described in Figure 5. Region coordinates are the same as in Figure 3.

S . E . = -0.395; and for 12 min: A = 1.52, S.E. = ,592; B = 1.03, S.E. = 0.694; C = 0.252, S.E. = 0.175.

193

Physostigmine Distribution in Brain Regional Cerebral Blood Flow Variable times after Physostigmine Inj. 00 / 0 . 5

4.0

r

rnin

[PJ 5 min

,--. c

15.0

I

I

Y = 0.74

0.5 rnin

T

I 12 min

+ 4.96 X

- 0.69 X 2

r = 0.87

4+

.U C

10.0

v

1

al

._C .-k

.-

L

E

-

c

m 0

*

2.0

I

*

-

E

r 2.

n I n

c7,

1.o

I

0.0

,--. -7-

on

0.0 1.0

2.0

F

Y = 2.68

4+

r = 0.73

C

5.0

6.0

15.0

I

.U

4.0

3.0

+ 1.87 X

- 0.21

X2

10.0

v 0)

.-C Fig. 7. Regional cerebral blood flow measured i n animal groups as described in Figure 5 . Region coordinates are the same as in Figure 4.

.-k Y

m

5.0

0

U)

J> z

a

The highest levels of radioactivity were found in structures lacking a blood-brain barrier (pineal gland, median eminence, and choroid plexus) (Figs. 1 , 4). The lowest levels were observed in white matter, which also showed the lowest levels of blood flow. The concentration of 'H-labeled physostigmine in spinal cord was considerably higher in gray matter than in the white matter columns (Fig. 1 , 4). Initial concentrations (0.5 min after administration) of physostigmine in the areas found to respond to this drug with vasodilatation (motor, sensory, temporal and occipital cortex, claustrum, and superior colliculus) were contrasted with those of areas of comparable basal rCBF that did not show such an effect (caudate-putamen, posterior thalamus, mammillary nucleus, inferior colliculus, cerebellum, and interpeduncular nucleus). Mean tissue radioactivity (SEM) in responding areas was 6.76(0.56) nCi/mg and in nonresponding areas 7.12(0.85) nCiimg. Thus the drug had reached comparable concentrations in both areas, and this disproves the hypothesis that variations in drug access underly the observed variations in the capacity of physostigmine to induce cerebral vasodilatation.

DISCUSSION A blood flow-related distribution of radioactivity was observed in brain within 0.5 min of' i.v. infusion of 'H-labeled physostigmine. This is to be expected since physostigmine, a lipid-soluble tertiary amine, crosses the blood-brain barrier easily (Somani and Khalique, 1987).

I

n

I

7

0.0 J0.0

I

I

F i

.U C

1.0

3.0

4.0

Y = 1.52

+ 1.04 X

2.0

5.0

6.0

15.0 I

10.0

.

5.0

1

12 min

- 0.25 X2

r = 0.24

v

01

.-C .-

s

c

m 0

L x

a I n

I

0.0 1

0.0

1.0

2.0 rCBF (ml

3.0 g-1

4.0

*

5.0

E 0

min-1)

Fig. 8 . Correlation between regional cerebral blood flow and regional concentration of tritium-labeled physostigmine. Blood flow was measured with the autoradiographic I4C-IAP technique and tritiuni-labeled physostigmine concentration with quantitative autoradiography . The regions sampled are those listed in Figures 2-7, with the exception of those lacking a blood-brain barrier (pineal gland, choroid plexus, and median eminence) that were not included in the correlation analysis. Standard errors of the coefficients of the polynomial regression fitted to the data are given in the text.

brain tissue seems to be imposed, however, by the blood-brain barrier, since pineal gland, median eminence, and choroid plexus, regions lacking this property (Gross et al., 1987), showed higher levels of radioactiv-

The blood-tissue exchange of a highly permeable mole-

ity than any other region.

cule like this one becomes blood flow dependent (Kety, 1951). Some restriction to exchange between blood and

Because physostigmine is slowly metabolized in brain, at short intervals, most of the radioactivity can be

194

Scremin et al.

TABLE I. Physostigmine Pharmacokinetics in Different Brain Reeions* Regions Cingulate M. genic. D. sup. coll. Motor s. sup. coll. Pont. n. Interped. Somato. M. niamill. Med. em. Lam 1-11 V . horn Cerebellum Pont. ret. M . raphe Temporal Infer. col. Ch. plexus

TIiZ(min)

K,, (min-’)

10.03 7.53 7.53 7.53 7.33 6.02 6.02 6.02 6.02 6.02 6.02 6.02 5.02 5.02 5.02 5.02 4.30 4.30

0.069 0.092 0.092 0.092 0.094 0.115 0.1 15 0.115 0.115 0.115 0. I15 0.115 0.138 0.138 0.138 0. I38 0.161 0.161

*Half-life (T,,>)and rate of elimination (Kbr) from brain regions were calculated from least-squares analysis of the regression of log tissue radioactivity on time after drug administration.

assumed to be still in the original molecule. Somani and Khalique (1987) have found that, 2 min after i.v. injection, 93% of the activity in brain was still associated with physostigmine. A rapid decline in physostigmine concentration with time was observed in a number of areas. The most remarkable example of this phenomenon was in the inferior collicullus, a region of high blood flow (Sakurada et al., 1978; Scremin et al., 1988) and low cholinesterase activity (Yamamura et al., 1974) (Figs. 1, 3). A high rate of clearance by the circulation and few cholinesterase molecules available for reaction with the drug can explain the rapid disappearance of physostigmine in these regions. Five minutes after i.v. administration, some brain areas showed a higher concentration than at 0.5 min. Several mechanisms could explain this phenomenon. First, most such areas are known to have a high level of cholinesterase activity (Hoover et al., 1978). The drug reacts with cholinesterase by interaction between the C atom of its carbamyl group and the serine hydroxyl group in the esteratic site of the enzyme. The rest of the physostigmine molecule is then split off, leaving the carbamylated enzyme that reacts with water to regenerate the active enzyme and produce methylcarbamic acid (Koelle, 1975). Although the label resides in the aromatic ring of physostigmine, a region that does not combine with the enzyme, it is possible that the rate of dissociation of this portion of the physostigmine molecule from the carbamic group could be slow enough to permit a “sequestration” of the radioactivity by this associa-

tion. It is also possible that, after dissociation from the enzyme, the fraction of the drug that contains the label may enter intracellular compartments and not be available for back diffusion into the blood. By this mechanism, regions rich in cholinesterase activity will be expected to accumulate the label and to show higher levels of activity at later times after drug administration. Intracellular accumulation of ’H-physostigmine labeled similarly to that used in the present experiments has been previously demonstrated (King and Somani, 1987). A second possible mechanism relates to the metabolism of the free drug in the organism. Twelve minutes after infusion, we expect a significant portion of the brain radioactivity to be in the form of physostigmine metabolites (Somani and Khalique, 1987). It is possible that some of these compounds may differ from the parent drug in their ability to cross the blood-brain barrier and be trapped or slowed down in their exit from brain tissue to blood. Finally, it is possible that some of the circulating metabolites, which are known, from previous work (Somani and Khalique, 1987), to represent more than half of the activity in plasma at 12 min after i.v. administration, could permeate the blood-brain barrier at a slower rate than physostigmine. This could give rise to a diffusionlimited, rather than blood flow-limited, distribution that may help explain the lack of correlation between tissue radioactivity and rCBF observed at 12 min after injection. The three regions that showed the lowest rates of decrease in radioactivity over time (hippocampus, caudate-putamen, and amygdala) have been reported to contain the highest levels of cholinesterase activity (Hoover et al., 1978). The spinal cord showed a marked difference in physostigmine-related radioactivity between gray and white matter at all times after drug administration. Two factors may contribute to this phenomenon. Blood flow level of spinal gray matter is approximately five times higher than that of white matter. The ratio of blood flow between gray matter and white matter is similar to that of ‘H radioactivity. On the other hand, gray matter is also richer in cholinesterase activity (Paxinos and Watson, 1982), and retention of the radioactive label by the mechanisms outlined above may explain why a higher level of radioactivity in gray matter was also observed long after drug administration. The rapid increase in brain tissue concentration and the rapid elimination of ’H-labeled physostigmine observed in most brain regions upon i.v. dosing are in agreement with the dynamic effect of this drug observed in vivo. One clinical use of physostigmine is the treatment of coma induced by drugs with central anticholinergic activity (Duvoisin and Katz, 1968; Rosenberg, 1974; Nattel et al., 1979; Daunderer, 1980; Newton,

Physostigmine Distribution in Brain

1975) or in the reversal of anesthesia (Rupreht, 1987). An almost immediate reversal of symptoms is reported following an i.v. bolus of physostigmine. The patients recover consciousness and appear normal but lapse into coma again within 30 min (Rosenberg, 1974). The cerebral blood flow effects of physostigmine observed in the present experiments are, in general, similar to those reported previously in rats (Scremin et al., 1988). In the present series, however, new information has been obtained regarding the latency and time course of physostigmine cerebrovascular actions. The lack of effect of this drug on rCBF immediately after i . v . administration is in agreement with previous findings in rabbits. In this species, i.v. pulse injection is attended, after 2-3 min, by an increase in internal carotid artery blood flow, a vessel that supplies almost exclusively the brain in this species (Scremin et al., 1982a). The effect peaks in 6 min (Scremin et al., 1983). Half-life of disappearance of physostigmine vasodilatation in the parietal cerebral cortex of the rat, measured locally by hydrogen gas clearance, is approximately 15 min (Scremin and Scremin, 1986). This is only slightly lower than the average half-life of drug elimination in motor and sensory (parietal lobe) regions observed in the present experiments. The predominance of the cerebral dilator effect of physostigmine in neocortex does not seem to be explained by its regional cerebral distribution. At the time of the peak effect, about the same levels of physostigmine-related radioactivity can be observed in regions that show a prominent vasodilatation as well as in those that lack this effect entirely. Moreover, regions lacking a blood-brain barrier and in which physostigmine concentration was maximal did not show any change in rCBF. In summary, the autoradiographic quantitative evaluation of physostigmine distribution in brain permits an analysis of drug distribution with a high degree of spatial resolution. Combination of this approach with autoradiographic measurement of blood flow has permitted a fine analysis of the role of this variable in physostigmine pharmacokinetics.

ACKNOWLEDGMENTS The authors thank Margaret O’Neal for skillful technical assistance and Deborah Heuser for help in manuscript preparation and illustrations. This work was supported by a Merit Review Grant from the Veterans Administration (O.U.S., A.M.E.S.) and U . S . Army Contract DAMD 17-83-C-3195 (S.M.S., E.G.).

REFERENCES Daunderer M (1980): Physostigmine salicylate as an antidote. Int J Clin Pharmacol Ther Toxicol 18523-535, Duvoisin RC, Katz R (1968): Reversal of central anticholinergic syn-

195

drome in man by physostigmine. J Am Med Assoc 206: 19631965. Gross PM, Blasberg RG, Fenstermacher JD, Patlak CS (1987): The microcirculation of rat circumventricular organs and pituitary gland. Brain Res Bull 18:73-85. Hoover DB, Muth EA, Jacobowitz DM (1978): A mapping of the distribution of acetylcholine, choline acetyltransferase and acetylcholinesterase in discrete areas of the rat brain. Brain Res 1S3:295-306. Kety SS (1951): The theory and applications of the exchange of inert gas at the lungs and tissues. Pharmacol Rev 3:l-41. King BF, Somani SM (1987): Distribution of physostigmine and metabolites in brain subcellular fractions of the rat. Life Sci 41: 2007-2015. Koelle GB (1975): Anticholinesterase agents. In Goodman LS, Gilman A (eds): “The Pharmacological Basis of Therapeutics.” New York: Macmillan Publishing Co., Inc, pp. 445-466. Mattio T, Mcllhany M, Giacobini E, Hallak M (1986): The effects of physostigmine on acetylcholinesterase activity of CSF, plasma and brain. A comparison of intravenous and intraventricular administration in beagle dogs. Neuropharmacology 25: 1 1671177. Nattel S , Bayne L, Ruedy J (1979): Physostigmine in coma due to drug overdose. Clin Pharmacol Ther 25:96-102. Newton RW (1975): Physostigmine salicylate in the treatment of tricyclic antidepressant overdosage. J Am Med Assoc, 23 l:94l943. Paxinos G, Watson C (1982): “The Rat Brain in Stereotaxic Coordinates.” Sidney: Academic Press. Rosenberg H (1974): Physostigmine reversal of sedative drugs. J Am Med Assoc 229:1lh8. Rupreht J (1987): Physostigmine reversal of general anesthesi ciated EEG changes. Anesth Analg 66: 1050. Sakurada 0, Kennedy C, Jehle J , Brown JD, Carbin GL, Sokoloff L (1978): Measurement of local cerebral blood flow with i~do[’~Clantipyrine. Am J Physiol 234:HSY -H66. Scremin OU, Allen K, Torres CD, Scremin AME (1988): Physostigmine enhances blood flow metabolism ratio in neocortex. Neuropsychopharmacology 1 :2Y7-303. Scremin OU, Rovere AA, Raynald AC, Giardini A (1973): Cholinergic control of blood flow in the cerebral cortex of the rat. Stroke 4232-239. Scremin OU, Scremin AME (1986): Physostigmine induced reversal of ischemia following middle cerebral artery occlusion in the rat. Stroke 17:1004-1009. Scremin OU, Sonnenschein RR, Rubinstein EH (1Y82a): Cerebrovascular anatomy and blood flow measurements in the rabbit. J Cereb Blood Flow Metab 255-56. Scremin OU, Sonnenschein RR, Rubinstein EH (1982b): Cholinergic cerebral vasodilatation in the rabbit: Absence of concomitant metabolic activation. J Cereb Blood Flow Metab 2:24 1-247. Scremin OU, Sonnenschein RR, Rubinstein EH (1983): Cholinergic cerebral vasodilatation: Lack of involvement of cranial parasympathetic nerves. J Cereb Blood Flow Metab 3:362-368. Scremin OU, Torres CD, Scremin AME (1987): Differential effects of physostigmine on blood flow-metabolism coupling in cerebral cortex and superior colliculus. J Cereb Blood Flow Metab 7: S264. Somani SM, Khalique A (1987): Phamiacokinetics and pharmacodynamics of physostigmine in the rat after intravenous administration. Drug Metab Dispos 15:627-633. Yamamura HI, Kuhar MJ, Greenberg D, Snyder SH (1974): Muacarinic cholinergic receptor binding: Regional distribution in monkey brain. Brain Res 66541-546.

Brain regional distribution of physostigmine and its relation to cerebral blood flow following intravenous administration in rats.

3H-labeled physostigmine (50 micrograms.kg-1) was administered intravenously to rats, and its concentration in brain tissue and spinal cord was assess...
813KB Sizes 0 Downloads 0 Views