BIOL PSYCHIATRY

269

Alterations of Local Cerebral Blood Flow and Glucose Uptake by Electroconvulsive Shock in Rats Amiram I. Barkai, lsak Prohovnik, William L. Young, Margaret Durkin, and Henry D. Ne.son

The effects of single and repeated electroconvulsive shock (ECS) treatments on regional cerebral blood flow ¢rCBF) and on rates of glucose flow from blood to local brain areas (rCGF), were investigated in pentobarbital-anesthetized rats, using q~ntitative autoradiographic techniques. Effects of single ECS on rCBF were assessed at two average time points of 15 and 55 sec after the application of the electric current, whereas the effects on rCGF were assessed at 70 and 110 sec. Effects of repeated ECS were assessed 24 hr a~er the last ECS in a series of eight daily treatments. Single ECS caused marked increases in rCGF in different brain structures, but no significant effects were observed after repeated ECS. Similarly. substantial increases in rCBF were seen during and immediately after the ECS-induced seizure, but not 24 hr after the last treatment of repeated ECS. Single ECS appeared to have differential effects on rCBF in hind-brain structures as compared to more anterior regions. A linear relationship between rCBF and rCGF values was established in control animals, indicating coupling of these two variables with a constant rCBF/rCGF ratio. ECS caused an apparent increase in the CBF/CGF ratio, which might be attributed to the different temporal resolution of the two methods used here to estimate rCGF and rCBF. Analysis of the increments of rCGF and rCBF extrapolated to the zame point in time after a single ECS (I0 sec), revealed that in many of the examined structures the CBF/CGF ratio was similar to that observed in control animals, indicating that the coupling of CBF and CGF is maintained during the seizure. But in some brain stem structures such as the dorsal raphe, inferior colHculi, superior olivary nucleus, and the vestibular nucleus there were large increases in CGF associated with a marked drop in the CBF/CGF ratio. This observation suggests that high metobolic demands can be met by increased local blood flow up to a given "'ceiling" keeping the glucose clearance from blood to brain tissue constant. However, when the metabolic demands exceed this upper limit, the additional demands could be met by. an increased clearance of glucose without a change in CBF.

From the New York State Psychia~c lnstitme ~AIB. IP. MID. HDN) and Departments of Psychiatry (A1B, IP). Neurology and Radiology (IP). and Anesthesiology (WLY). College of Physicians and S ~ e o n s , Columbia Umversity, New York, NY. Address reprint requests to Dr. Amham !. Barkai. Box 54, New York State Psychiatric Institute. 722 West !68th Street, New York. N.Y. 10032. R,-ceived December 13. 1990: revised March 30. 1991.

© 1991 Society of Biological Psychiatry

000~-3223/91/$03.50

270

A.I, Barkai et al

BIOL PSYCHIATRY • ,~ -~.~ 1991,o~.,~69~.~8.

Introduction Repeated electroconvulsive shock (ECS) is widely used in psychiatry for the treatment of depressive illness and schizophrenia (Medical Research Council 1965; Fink 1978; Malitz and Sackeim 1986). Until recently ~ therapeutic use of ECS has been based mainly on clinical experience with very little understanding of its mode of action. Animal studies have shown, however, that ECS alters the function of catecholaminergic and serotonergic synapses (reviewed in Barkai 1986; Sulser 1987), findings that are consistent with observations that implicate dysfunction of catecholamine and sem~onin systems in affective disorders or schizophrenia. A number of these animal studies produced data indicating that long-term ECS treatment was associated with changes in the responsiveness of certain monoamine receptors and that such ch~ges parallel fairly well the time course for observed clinical effects (Charney et al 1981; Sulser 1987). More recent studies that employed quantitative autoradiography demonstrated that the effects of repeated ECS on cerebral noradrenergic, serotonergic, or dopaminergic receptors are confined to discrete brain regions (Biegon and Israeli 1986, 1987; Barkai et ai 1990). One important goal in studying the efficacy of ECS is the search for brain structures that might be activated or supprressed during ictal and postictal states. It is well established that generalized seizures represent the most intense cellular work in the brain (Balazs 1969). The metabolic rate during the early stages of the seizure has been shown to increase by at least 2-3 fold (Meldrum and Niisson 1976). There is also an intense vasodilation accompanied by an increase in cerebral blood flow (CBF). The increase in CBF during the seizure is associated in some areas with a loss of the close coupling of cerebral glucose metabolism (CGM) and CBF (Horton et al 1980; Ingvar and Siesjo 1983). While global CBF and CGM were investigated quite extensively during the generalized seizure (usually chemically induced seizure), data on these variables during the postictai phase or following repeated ECS are scarce. Studies ,~n the postictal effect o¢ ECS showed acute reduction and apparent redistribution of regional CBF (rCBF) in humans (Prohovnik et al 1986). Sacks et al (1982) and Ackermann et al (1986) have demonstrated decreases of regional CGM (rCGM) following single or repeated ECS. Orzi et al (1987) also showed lower rates of rCGM 24 hr following a single ECS in rats, but reported a selectively increased CGM in the hippocampus following repeated ECS treatment. The purpose of the present study was to investigate the effects of single and reU

'~m~U

Vl

i

U V S i i

•

U i

~ I U

i ' ~ W I I ~ ,~ I

ill

I II'~,~l ~

g[

]llUa

L

L i i II II~llw I t 1 [~llf i '1~ i ~ J ~1

~ [ e l

ul~

u l ] ~ [

ul

ECS-induced seizures and at 24 hr following the last ECS application in a series of eight daily treatments. Quantitative autoradiography was employed and the experiments were directed to address the following questions: (1) What are the immediate effects of a single brief ECS on CGM and CBF in discrete brain regions? (2) What are the long-term effects of repeated ECS on rCBF and rCGM? and (3) What are the changes, if any, in the coupling of CGM and CBF following the onset of an ECSproduced generalized seizure?

Methods

Animals Male Sprague-Dawley rats (300-350g) from Taconic Breeding Farms (Germantown, New York) were used. Rats were kept under a controlled light--dark cycle. Food and water

ECS, Cerebral Metabolism, and Blood Flow

BiOL PSYCHIATRY

27 [

were ad iibitum, but food was removed 24 hr before the animal was anes~,etized for CGM or CBF measurements.

Administration of ECS Repeated ECS was appliea once daily for a period of eight days o The animal was placed in an isolated chamber and connected to saline-moistened ear-clip type electrodes. Elec~c current was delivered from a commercial ECT unit (Medcraft Mark II B-24L The stimulus consisted of a sinusoidal waveform (60 Hz; 1 ~ c duration). Voltage output was set at 150 volts to produce a current intensity between 20-30 mA ¢ d e t e ~ i ~ d by monitori~ voltage across a 1000 ohm resistor). The animal was returned to its home cage approximately 5 min after each ECS treatment. The control group for repeate8 ECS received similar treatment but electric current was not delivered, Single ECS was applied to anesthetized rats a short time prior to the ~ejection of the isotope for the measurements of rCGM or rCBF. The rat was maintained ender artificial respiration and connected to a polygraph for monitonng of blood press~e and electroencephalogram (EEG). Stimulus paramete~ were the s ~ e as for the repeated ECS,

Estimation of Cerebral Glucose Metabolism Rates of flow of glucose from blood to various cerebral structures (rCGF) were used to assess cerebral glucose metabolism, rCGF was estimated from the curve representing the decrease of plasma [ 14-C]-glucose specific activity with tree and from the concent:ation of [14-C] in the brain region (dpm/mg) me~ured 2 rain after the intravenous injection of a tracer amount of D-[ 1-14-C]-glucose. Values for [ ! 4-C] concentrations in brain tissue were obtained by quantitative autoradiography with the aid of an Amersham RAS-1000 Image Analyser, using appropriate [14-Cj standards. C~iculations of rCGF values were carried out in accord~ce with the semi-compartmental model of Baker and Heubotter (1972) as applied previously for i n c a s i n g the flow in vivo of glucose carbon to brain proteins in rats (Bahai et al 1974). Briefly, the rat was anesthetized with pentobarbital ( n e m b u ~ , 50 mg,qcg ip) and the left femoral vein and artery were cumulated with heparinized polyethylene tubing. The trachea was .i.,.~ .[ . #.l i., . l l . , I . Iat and ,.~.,,,u,,,,,.,~'~a for ,~,,~l,.~,u resph-avon ~,.,~,~,~t~'~,_,,,**,~,~~.,;.~Respirator, Harvard Apvaratus Co., South Natck, MA) set to maintain arterial pCO2 of 35--40 ton'. The surgical procedure was completed within 45 rain and the rat's head was placed under a guillotine. Exactly 60 min after the initiation of anesthesia the animal was injected IV with 25 ~Ci of 1-[ 14-C]-glucose (American Radiolabelled Chemicals, St. Louis, MO) in 0.2 ml saline. Arterial blood samples (8 samples; 50 ~tl each) were withdrawn at predetermined time intervaL, during a 2-min time period, to establish the plasma [14-C]-glucose specific activity time curve. At 2 rain after the isotope injection the rat was decapitated, the brain removed rapidly, frozen in cold isopentane and kept at -20°C. Coronal sections were cut at 20 ttm in a Hacker cryostat and representative slices were mounted on microscopic~ slides pretreated with gelatin. The slices were a/r-dried at room temperature and then apposed to Kodak X-omat AR5 films for one week to obtain autoradiographic images. Standards of known [14-C] radioa~.tivity were apposed to the same film to allow quantitative analysis. The blood samples were centrifuged (microcentrifuge) and plasma aliquots were taken •~ , . t ~ . ~ - ~ l

"

"

,.1 ~1 ,-7,z

BIOL PSYCHIATRY 1991 ;30: 269- 282

A.I. Barkai et al

to determine the glucose specific activity (SA). The plasma SA-time curve was analyzed by a nonlinear regression analysis to fit the equation: SA (t) = A e - " ' + B e - " '

Ill

Where SA(.t) is the specific activity of plasma glucose at time t, expressed as a fraction of the injected radioactivity, A and B are zero time intercepts, and m and n are the exponential slopes of the log SA versus time curve, The regression analysis was performed with the PCNONLIN computer program (Statistical Consultants, Lexington, KY) to obtain the values for A, B. m. and n. which were then u,~xl to calculate net flow of glucose from blood to brain. The size of the glucose miscible ~ 1 (Q) was calculated from the equation Q = I/(A + B) and was expressed in lxmols. The disposal rate of glucose from this pool during the whole experimental period (Rd) was obtained from the relationship Rd = Q/[A/(A + B ) . m + B/(A + B ) - n ] and was expressed in ixmol/min. The rate of flow of glucose from blood to brain (rCGF) was then calculated from the equation rCGF = Rd. q*b(t)/q*d(t)

121

Where q*b(t) is the radioactivity measured in 1 mg brain tissue at time t and expressed as a fraction of the injected radioactivity and q*d(t) is the total radioactivity that was transferred out of the glucose miscible pool during the experimental time period t, exo pressed also as a fraction of the injected radioactivity. Values of q*b(t) in various cerebral regions were measured experimentally at t = 2 min using quantitative autoradiography, whereas values for both Rd and q*d(t) were calculated from the plasma glucose SA-time curve according to equation l above (Barkai et al 1974).

Measurements of rCBF The method described by Pulsinelli et al (1982) as applied by Young et al (1987) was used. Briefly, the rat was anesthetized with nembutal, the trachea was exposed, cannulated, and the animal was ventillation-controlled to maintain pCO2 between 35 and 40 ton'. Arterial blood gases were analyzed with a Coming Blood Gas Analyser. Temperature was maintained at 37°C by a heating lamp and monitored by a rectal thermistor. Polyethylene cannulae were placed in both femoral arteries and one femoral vein. Blood pressure and EEG were monitored on a Grass model 7 polygraph. Exactly 60 rain after the initial administration of nembutal, a bolus of 15 IxCi of [ 14-C]-iodoantipyrine (American Radiolabelled Chemicals, St. Louis, MO) was injected IV in 0.3 ml saline and a reference arterial blood sample was withdrawn simultaneously using a Harvard withdrawal pump at a rate of 0.786 ml/min. The experiment was terminated by decapitation lO sex: after the tracer injection; the arterial catheter was removed simultaneously from the femoral artery and its content added to the reference sample. The brain was rapidly removed and prepared for frozen coronal sections as described atmve. The distribution of [14-C] radioactivity was assessed by quantitative autoradiography. The reference blood sample was decolorized, treated with protosol and counted for radioactivity. Counting efficiency was determined with [l,~-C]-toluene internal standard. The rCBF values were obtained from the following equation: rCBF = Rwr. *Cb/*Qr

[3]

Where Rwr is the Reference withdrawal rate (ml/min). *Cb is the concentration of

ECS, Cerebral Metabolism, and Blood Flow

BI~OL~ Y C H r A ~ Y

radioactivity in brain (nCi/g) and *Qr is the quantity oi radioactivity in ~ sample (nCi).

273

reference

Statistical Analysis The values for rCGF and rCBF for brain regions examined after ECS were tested for significant differences from control values by analysis of vmance with subsequent testing by Dunnett's procedure for multiple comparisons ~Dunnett |955).

Results ECS-Produced Seizures Typical EEG recordings representing the response to each of the repeated ECS treatments were similar to those described previously [Perumal and B~kai 1982). Briefly, under the present conditions a primary after discharge (PAD~ activity of 12-16 sec duration was followed by a postictal depression of the EEG lasting about 40--60 sec. This postictal depression was frequently interrupted by a secondary after discharge (SAD) that lasted about 10 sec an~ consisted regularly of successive monophasic bursts of progressively increasing amplitudes until it ended abruptly. The behavioral concomitants consisted of a tonic convulsive phase during the PAD period, followed by clonic con~lsions of the hind legs and the forepaws for approximately one min. The SAD was frequently preceded by twitching jaw movements. The rat assumed alert behavior within 3--4 min after application of the electric current. Rats subjected to a single ECS during pentobarbital anesthesia responded with a PAD of 10-12 sec duration followed by a postictal depression of approximately 30 sec. In these animals the behavioral concomitants consisted of a tonic convulsive phase during the PAD period followed by clonic convulsions of the hind legs and twitching of the forepaws for a period of 20-30 sec.

Effects of ECS on rCGF The effects of single and repeated ECS treatments on rCGF are presented in Table 1. Because the control values of the single ECS group did not differ si~ificantly from the control values of the repeated ECS, group values of the two control groups were pooled. Effects of single ECS were assessed at two time intervals; one was between 10 and 130 sec (mean 70 sec) and the other between 50 and 170 sec (mean 1I0 sec)after the application of the elecmc current. The effec'~ of repeated ECS were assessed du~n~ a two-rain period starting 24 hr after the last in a series of eight daily ECS treatments. In control rats, the values for rCGF in various regions ranged from 0.78 ~mol/g/min in the corpus callosum to 2 p.mob'g/min in the vestibular nucleus. Application of a single ECS caused marked increases ]a ~ G F during and immediately after the seizure. When measured during the ame period of 10--!30 sec after the application of the electric current, the rCGF va~aes in different structures were elevated to 165-221% of the corresponding control values (mean 189 ± 13%), but when measured during the later time period of 50-170 sec ;ffter ECS, the rCGF values ranged between 116-148% of the corresponding control values (mean 132 ± 10%). There were no significant effects of repeated ECS on rCGF ve~lues measured 24 hr after the last treatment, although most s~cmres showed 10-20% reduction in their metabolic activity compared to controls (mean 85 _-. 8.3%).

274

A.I. B~rkai et al

BIOL PSYCHIATRY

T a b l e I. Effects o f E C S o n L o c a l Cerebi°at G l u c o s e H o w ( r C G F ) '~ rCGF (gmol/f,/min) Region

Control

70 "b

I !0 ~

Repeated ECS ~

Frontal cortex Motor cortex Sore, ,sen. cortex Caudate putamen Hippocampus Corpus callosum Thalamus Hypothalamus Substantia nigra

1,22 ± 0.31 1.37 ± 0.34 1,54 • 0.33

2.3 ± 0,83 (189%) 2.6 "" 0.84 (190%)

1.5 ± 0,22 (123%) 1.9 ± 0.33 (139%)

I.II ± 0.39 O1%) I.II ± 0.39 (81%)

2.9 ± 0.99 (188%)

2.1 ± 0.44 (136%)

1.33 ± 0.44 {86%)

!.22 ± 0.28

2.3 ± 0.78 (189%)

1.7 ± 0.50(139%)

!.05 ± 0.33

1.11 ± 0.33 0.78 ± 0.22

2.0 ± 0.67 (180%) !.5 ± 0.5 (192%)

1.5 ± 0,28 (135%) !.1 ± 0.17 (141%)

0.88 ± 0.22 (79%) 0,83 "*" 0,28 (!06%)

1.42 ± 0.33 1.22 ± 0.23 1.14 "- 0.28

2.8 ± 0.94 (197%) 2.3 ± 0.56 (18:8%) 2.2 ± 0.61 (193%)

2.1 ± 0.61 (148%) 1.8 ± 0.44 (148%) i.5 ± 0.33 (132%)

I.(}5 ÷ 0.44 (74%) 1.05 ± 0.50 (86%) ---

1.95 "- 0.56

4.3 ± 1.7 (221%)

2.5 ± 0.89 (128%)

1.55 "" 0.5{} (79~)

!.89 ± 0.39 !.72 ± 0.44

3,2 ± 0.94 (175%) 3.4 ± 1.6 (198%)

2.2 ± 0.44 (120%) 2.1 "*" 0.39 (123%)

--1.44 _ 0.38 (84%)

2.00 ± 0.61

3.3 ± 1.1 (165%)

2.3 -,'- 0.28 (116%)

!.66 -,- 0.56

Infcrior

colliculi Dorsal raphe Superior olivary nuc. Vestibular

(86%)

(83%)

nucleus

Walues represent mean and SD obtained from 6-1 ! rats. Number in parentheses ~ptesents percentage ofthe conesimn~ng control val~. Values at 70" and !10" were obtained during the first (IO-130 sec) and second (50-170 sec) time intervals, respectively, after a single ECS. See text Cot more detail. mall values in this column are signficantly (p < 0.01) higher than the corresponding controls. 'Values in this column are significantly (p < 0.05) h i g ~ than control except for the vestibular nucleus and the dorsal raphe where p values were t" 082 and 0.067 respectively. 'tAll values in this column are not significantly different from their con~sponding controls,

Effects of ECS on rCBF As with rCGF, the effects of single ECS on rCBF were assessed at two time intervals: The t;rst was between lO and 20 sec (mean 15 sec) and the second between 50 and 6 0 sec (mean 55 sec) after the application of the electric current. The effects of repeated ECS were assessed during a lO sec period, 24 hr after the last ECS. The results are presented in Table 2. Control values represent the single and repeated ECS control groups that did not differ significantly. Substantial increases in rCBF were observed during and immediately after the seizure. Values of rCBF in various structures were 277-761% (mean 540 _+ 168%) of the corresponding controls in the first time interval and 142383% (mean 245 _ 73%) of the controls during the second time i,,terval. There were no significant effects of repeated ECS on rCBF in any of the t~rain regions examined. Control rCBF values in brain-stem structures posterior to the level of the inferior colliculi were 2-3 fold higher than rCBF values in anterior cortical and subcortical structures. However, the relative increases in rCBF in those brain-stem structures were substantially smaller compared to increases observed in anterior regions after a single ECS. Thus, the mean percentage of increase of rCBF in posterior brain-stem structures (inferior colliculi, superior olivary nucleus; substantia nigra; vestibular nucleus; dorsal raphe) at the first time interval was 386 + 64% of the controls compared to a mean of 636 ± 136% of the controls in more anterior regions (Table 2).

ECS, Cerebral Metabolism, and Blood Flow

a~ot. ~svcmA~v tgOt :30: 2 ~ 282

275,

T a b l e 2. E f f e c t s o f E C S o n L o c a l C e r e b r a l B l o o d F l o w ( r C B F ~ '~ rCBF {wd~g/min ~ Region

Control

I 5 "~

Frontal cortex Motor cortex Soma~o sen. cortex Caudate putamen Hippocampu~ Corpus callosum Thalamus Hypothalamus Subs~antia nigra Inferior colliculi Dorsal raphe Superior olivary Vestibular nucleus

0.41 _~ 0.1 0 . 4 6 ~_ 0.1 0.58 ~ O~!

2,7 _~ 0~6 ~659q~1~ 3.5 _~ 0 8 ~ 7 6 1 ~ 3,8 -~ 0.5 { 6 5 5 ~

~.O3 ~ O.4 ~251~)~ !~55 _~ 0,6~337%}, ~.50 ~ 0 . 6 f259ck}

0 . ~ ± 0.! 0 . 3 9 ~ 0. l 0.50 ~ 0 . |

{9,8} {85} ~86~

0,49 _~ 0.1

3.2 ± 0 9 ~ 6 5 3 ' ~

| . 2 4 ~ 0.4 C253%}

0.42 ~" 0 . |

('~}

0.40-,- 0.1 O.3l = 0.1

2. l _-z 0.8 ~ 5 2 5 ~ l . | z ~3.4 ~355~;~

0.8t ~ 03~203%~ 0 . ~ "*- 0 . | (194%~

0 4 3 ~ 0.1 { | ~ 0.27 -'- 0 . | (87~

0.47 "" O. | Q33 ~ 0.| 0.95 ~ 0.3

3 A ___ 0 6 ~723rk~ 2.5 _-z 0 6 ~ 7 5 8 ~ 3.8 ~ 0 4 ~ 4 0 0 ~

1.80 ~_ 0.9 f383%) t . 2 0 ~ 0.4 ('364ck) t . 9 0 = 0.4 ~2~%~

0.41 ~_ O.l (87) 0.48 _-z 0,1 t145) ~

l . i O ± 0.4

4°7 ~ 0~5 ¢427%t

2.~

~ . | 0 ~ 3 (tOO%,~

! . | ! _ 0.3 i.48 -,- 0.3

4.3 ± 0,5 ~387%~ 4,1 -.- 0°7 ~277%)

2.20 ___ 0.7 ( | 9 8 % r 2.10 ~ 0.6 ~142%~

-1.66 _ 0.4 ( l l 2 j

1.04 ± 0.2

4.6 +_ 0.6 ~ 2 % ~

2.02 _ 0.6 (I94%)

1.02 -z 0.3

55 "~

Repea~ed ECS ~

_ 0.8 {2~8%r

(98)

~Values represent mean and SD obtained from 6- ] 1 rats. Number m parentheses represents the percentage of the corresponding control value. Values at |5" and 55" were o b ~ i ~ d ~ n g the fi~'st ( 10-20 sec) and second (50-60 sec) time intervals. respectively, after a single ECS. mAll values in this column are significantly ~p < 0.01 )htgher ~an the corresponding controls. 'All values m this column are not s~gmficantly different from then" correspo,ndmg controls.

Relationship between rCGF and rCBF When control values of rCBF were plotted against corresponding rCGF values the relationship between these two variables agreed quite well with the linear function rCBF = 0.7 rCGF. This reiationship did not hold, however, after the application of a single ECS. At 15 and 55 sec after ECS the CBF/CGF ratio increased substantially due to a n n a r e , n t larger i n c r a m a n t c i n r C R I : : e . m n n m d t n rCGF ( F : i g u r ° i Because rCGF values were obtained during a 2-min period, they represent measurements at the average time of i rain after the IV injection of [14-C]-glucose. Values for rCBF, on the other hand, were obtained during a lO-sec period and therefore represent measurements at the average time of 5 sec after the IV injection of [ 14-Cl-iodoantipyrine. Thus, |.he different temporal resolution of the two methods could account for the higher increments in CBF compared to CGF as observed here. To examine whether or not the observed increase in the CBF/CGF ratio after ECS is merely a result of differences in the temporal resolution of the methods, it is necessary to obtain values for both of the variables at the same point in time after the application of ECS. Such values can be calculated from the present data if one assumes that the ECS-induced increment in CBF or CGF "decays" exponentially in accordar, ce with the following exponential function of time: I'lt

.

.

.

.

.

.

.

A(t) = A,o,e -~,

.

.

.

.

.

.

.

.

.

.

l"

-

•

""

" 7"

[41

where A(t) is the increment at time t and is expressed as the percentage of the control

276

A.I. Barkai et al

BIOL P S Y C H I A T R Y I t ~ l , ~a,n,~9 "~I~

rCBF (mllO/min) 0

O

0

x x

CI,,I~ O

T. . . . . . . . . . . . . . . . . . .

0

5

Control

o

E C S c o r r e c t e d 10"

- 1 - -

lo 15 rCGF (pmol/g/min) ~

-+--

-

i

20

2s

EC$ let lntervel

Regroee. line, c o n t .

Figure I. Relationship between rCBF and rCGF in pentobarbital-anesthetized rats under control conditions and shortly after a single ECS. Small squares represent control values. Xs represent values obtained during the first interval after ECS where rCBF was estimated during the period of 10-20 seconds and rCGF during the period of ! ~ ! 3 0 seconds after ECS. The open large squares represent values calculated according to equation 4 for t = i0" after ECS. The solid line represents the regression line obtained with the control values. See text and Table 3 for more detail.

value and k is the rate constant for the disappearance of this increment. If this assumption is correct, mean A(t) values at tl = 70 sec and t2 = ! !0 sec for CGF or tl = 15 sec and t2 = 55 sec for CBF can be substituted in equation 4 above and the constants A and k for CGF or CBF can be solved independently. Once these constants are found they can be used in equation 4 to caiculate CGF or CBF values for exactly the same time, t, after ECS and the "true" relationship between these variables can then be obtained. Equation 4 above was solved for A and k using the rCGF data of 70 and 110 seconds. Similarly mean rCBF data at t = 15 sec and t = 55 set: ,,,ere used to calculate A and k values for rCBF at various structures (Table 3). We have also calculated A and k values for the average increment percentage of both CGF and CBF in the "whole brain" and found that the function describing the mean change of the ECS-induced increment in "global" CGF appeared to be very similar to that obtained for CBF (Figure 2). The individual A and k values calculated for each brain region were then used to obtain values for rCGF or rCBF at t = !0 sec after the application of ECS (Table 3). When these values were plotted against each other, the relationship between the two variables agreed quite well with the CBF/CGF ratio obtair, ed for control animals, but this relationship held only for rCGF values lower than 7 I~mol/glmin (Figure 1). When rCGF values were higher, the CBF/CGF ratio dropped markedly as seen for the dorsal raphe, the inferior colliculi. ~he s u ~ , i o r olivary nucleus a.d the vestibular nucleus (Figure 1). Discussion The experimental approach taken in the present study to estimate cerebral glucose metabolism follows the automdiographic studies of Lu et al (1983). These authors based their calculations on previous studies (Hawkins et al 1974; Hawkins et al 1983) that were

ECS, Cerebral Metabolism. and Blood Flow

B ~ PSYCHIATRY |99| : ~ J : ~ 282

2~

Table 3. Calculated rCGF and tCBF Values at tO Seconds after ECS * rCGF

rCBF

lO~

lff,

Region

A

K

Percentage

wmoV gPmin

A

K

Frontal cortex Motor cortex Somato-sen. collex Caudate putamen Hippocampus Corpus callosum Thalamus Hypothalamus Substanfia nigra Inferior colliculi Dorsal raphe Superior olivary Vestibular nuclei

96| 39| 42 !

~,034 0,02| 0,022

684 316 336

9°56 5.69 6~74

~ 973 886

0,03[ 0,025 0,03,1

616 757 648

2.93 3.94 3.75

377

0.02 !

3~

4.9~

896

0.032

650

3.67

338 380

0.02 ! 0.020

274 310

4, | 5 3.19

720 369

O. 035 0.025

507 203

2.42 0.94

334 255 606

O.0l 8 0.0 i 5 0.027

280 2i8 463

5,30 3,87 6,4i

838 935 453

0.020 0.023 0.028

687 742 343

3.69 2.77 4.20

1590

0.037

! 100

23.40

488

0.027

374

5.21

765 ! 266

0.033 0,037

548 877

I i .87 16.80

430 303

0.027 0,036

328 211

4.75 4.60

765

0.035

537

! 2.74

557

0.032

402

5.22

P~cen~e

mV~ rain

~Values for A and K were obtained by substituting the percentage values from TaMes ! or 2 in equation 4, A and K were then used in equation 4 to obtain A(O at ~: = 10". which represents the ~perc:ntag~ W~rement at that time. A ~ rCGF or tCBF values at t = IO" w~erethen calculated. For exampte, in Frontal o~rtex, the rCGF ~,'~ue for A(t = 70") was 89% arm for A(t = | lff'~ it was 23~- (Table IL Substituting ~ese values it~ ~q~ation 4 a ~ solving for A and K resulted m A = 96I% and K = 0.034 per second. These v~ues were then u,ed to solve ~ a t i o n 4 for A(t = I0e) which resulted in an increment A = 684ck ever the control value of 1.22 ttmoFg/tmn. Thus. the value for rCGF at t = I0" was o ~ n e d as 9.56 V.mct'~min as shown here.

directed to measure CGM in experiments o f brief duration. The use of [14-C]-glucose

as the labeling precursor was preferred over the commonly used [!4-C]-deoxyglucose method (Sokoloff et al 1977) mainly because we intended to measure rCGM d u n g short time periods to allow temporal evaluation of ECS effects during and following the ECSinduced seizure. While a basic assumption in both methods is that rCGM is constant during the entire experimental period, this assumption is not ~ways valid, especially during states of perturbations such as seizures. The results obtained with each of the two methods represent the sum of all activities occurring during the experiment. A change in rCGM over a time period of l rain, for example, is expected to be "masked" to a larger degree during an experimental period of 45 rain, such as requ~ed for the [14-C]deoxyglucose method, than during an experimental period lasting only 2 nrfin as conducted here with the [14-C]-glucose technique. It should be mentioned, however, that using the autoradiographic approach with [14C]-glacose as the labeling precursor, one actually measures the rate of glucose flow from blood to various cerebral regions (tCGF). In experiments of brief duration (1-5 rain), as performed here, the measured values of rCGF closely represent the rate of glucose influx (Cremer et a! 198!; Hawkins et a! 1983) or K ! .Cp ip the compartmental m~del ~resented by Lu et al (1983). Hawkins et al (1983) and Lu et al (1983) found that the rate constant

278

A.I. Barkai et al

BIOL PSYCIIIATRY 1 9 9 1 .~.,.69~,8,. '"~ ~'~

Inorement

1000

100

--....

,

I0

_..

0

t

20

.

,~

40

L

60

I

ao

_..__.,t

i00

[

120

140

TIME (Seconds) --"-- r C G F

--+-- r C B F

Figure 2. Changes with time of the average percentage increment of CGF and CBF after a single ECS. Mean rCBF values were obtained from Table 2 for the average time points of 15" and 55" after ECS, whereas mean rCGF values were obtained from Table ! for the average time points of 70" and ! 10" after ECS. The exponential functions for the decay of CGF and CBF increments are very. similar: The equation for CGF is 534 x exp (-0.0256t~ and for CBF it is 668 × exp (-0.0278t). See text for further detail.

for the unidirectional glucose influx (K! in their model) was related to the cerebral metabolic rate (CMR) by the nonlinear relationship KI = 13.87 (rCMR)°3911(Cp+8) where Cp is the concentration of plasma glucose (mM). This relationship was established only for normal physiological conditions and was used in the autoradiographic study of Lu et al (1983) to obtain CMR values from measured KI values in conscious normal rats. It is unlikely, however, that this relationship remains valid under abnormal conditions, such as after ECS. For example, if rCMR has a value of 5 Ixmol/glmin, the calculation of K i .Cp (at Cp = 7 raM) produces a value of 3. I ~mol/g/min that is paradoxical, because the influx is smaller than the utilization rate. We have therefore chosen to present our results in terms of the cerebral glucose flow (rCGF) without the: conversion to rCMR as done by Lu et al (1983). The use of rCGF values as measures of glucose utilization is supported by the high correlation that was observed between the rate of total glucose influx and the rate of glucose phosphorylation for various brain regions of rats in different physiological states (Cremer et al 1981). The experiments described here were conducted under nembutal anesthesia, which has been shown to decrease rCGM by about 50%-70% compared to the awake state (Sokoloff et al 1977). The study of ECS effects under anesthesia was deemed important because in the clinical situation ECS is applied under barbiturate anesthesia. Possible variations in rCGM in anesthetized animals may be related not or, ty to the level of the anesthetic agent in the brain but also to changes in CBF resulting from alterations in blood pCO2 during various stages of anesthesia, compared to awake animals. For comparisons between experimental and control groups it was important to reduce such variations to a minimum by controlling the level of anesthesia and maintaining a relatively narrow range of pCO2. Therefore, the I-[ 14-C]-glucose injection was given at a fixed time (60 min) after the

ECS, Cerebral Metabolism, and Blood Flow

B~

PSYCH|ATRY

279

administration of nembutal and the animals were maintained under controlled ar.dficial respira).ion. At that time the level of anesthesia was relatively light as judged by ~ return of some nociceptive reflexes and of the presence of EEG PAD following the application of ECS. It should be mentioned that the level of anesthesia is very important for the intensity and perhaps for the distribution of seizure activity ~ is kept light during clinical application of ECS. The rCGF values obtained under these conditions in control aniw~s ~ g e d from 0.78 p.mol/g/min in the corpus callosum to 2.0 I.gmol/~rmin in the vestibul~ nucMus. As already discussed, these rCGF values overestimate the CMR val~s ~mined by either the l l4-Cl-glucose or the [14-C]-deoxyglucose techniques, because ~ y me essentiMly equivalent to the total influx of glucose from blood to b ~ n . As expe~ed, the values obtained here under control conditions in the anesthetized rat are lower than the values for glucose influx obtained by Hawkins et al (1983) in conscious rats ( r a g e bbetween 1.65 p.moVgdmin in the corpus callosum and 3.2 pmoVg/min in the inferior colliculi). Cremer et al (1981 ) reported glucose influx values that ranged from 1.21 tzmoF~min in the hippocampus to 2.13 p,moV~min in the inferior colliculi in control rats. The measurements of rCGF and tCBF at two time points after a single ECS allowed estimations of changes in these variables at 10 sec after the application of t ~ elec~c current, during the perieM of the EEG seizure. These estimations revealed that increments in rCGF ranged from 218% over the control value in the h ~ M m n u s to I I ~ % in inferior colliculi, whereas rCBF increments ranged from 203% in ~ corpus callosum to 757% in the motor cortex (Table 3). It is interesting to note that s~ctures locat~ posterior to the subs:antia nigra responded to ECS with rCGF increments ~ g i n g from 463%-1100% (mean 703 _+ 122%) whereas the rCGF increments in more anterior regions Hanged from 218% to 684% (mean 340 _ 50%); the difference of the means between the posterior and anterior structures was significant (p < 0.01). On the other h ~ , the same posterior structures showed a significantly lower incremem percentage in rCBF compared to the anterior structures (331 +_ 32% versus 601 + 62%, Table 3). It is therefore unlikely that such differences between anterior and posterior regions were simply related to the site of electrical stimulation and the anatomical s~ctures; factors that play an important role in determining the electric current pathway. These differences between brain stem structures and anterior structures, taken together with the observed drop in CBF/CGF ratio in such structures as the dorsal raphe, the inferior coHiculi, the superior h i ; .y~l,,,, . . . . g&lnlnIJt tl h ~ e .Vlh,,"~wiRql,~lP~,,liln~ . . . gh..l,~.. Hu~.l~.u~ IK';m.t-~ !g ] X ;t,~ttit-at~ t~ gh~ ia. t tI gng g t~ ah* ~. t r o t ma~ da, l t c ~g g ~ ~gig~g~Cd£L~

Ulna

ht-.aln o f tmh. i a~~ ~ c t h, ~ tiTed L,ulgU.lgEa ~u,a~vJz, z..

rat. high metabolic demands can be met by increased local blood flow up to a "ce~ng" of approximately 5 ml/g/min while the glucose clearance from plasma to b ~ rem~,fins constant. However, when the metabolic demands exceed 7 I~mol/g/wfm and the CBF response has already reached its apparent upper limit (approximately 5 ml/g/min), the additional metabolic demands will be met by an increased glucose clearance from plasma to brain tissue. The relationship between cerebral energy metabolism and blood flow has been investigated under various conditions in numerous studies (reviewed in Sokoloff 1981; Lou et al 1987). While the results of many of these studies generally confirm the hypothesis of Roy and Sherrington (1890) that the cerebral vascular supply can be varied | o c ~ y in correspondence with local variations of functional activity, it has become clear that there are exan~ples of disproportionate changes in CBF compared to CMR during states of increased brain activity (Lou et al 1987). For example, Nilsson et al (1978) have shown that during bicucuiline-induced seizures CBF increases more than CMR, suggesting that

2:8,0

BIOL ~YCHIA'I'RY I ~1 ;30~,~,.8,.

A.I, Barkai et al

under extreme hypermetabolic conditions the relationship between CBF and metabolic activity may become nonlinear. Fox and Reichle (1986), working with humans, found discrete focal increases in rCBF and rCMR after vibratory stimulation of neuronal activity, but rCBF increased significantly more than rCMR. In the present study, the observed increments in rCBF compared to rCGF were also disproportionate during or immediately after the ECS-induced seizure (Figure 1), but the higher CBF/CGF ratio was probably erroneous and reflected differences in the temporal ~ l u t i o n of the methods used to determine rCGF and rCBF. Sacks et al (1982), using 114-C]-glucose and autoradiography, showed that a single ECS increased the [ 14-C] content of the brain within 1 rain after the application of the electric current, but when measured at time periods of IO-30 rain after ECS the | 14-C] content was reduced dramatically compared to c o n ~ l s . ~ d u c t ion in the incorporation of ! 14-C] to various brain regions was also apparent after a series of 9 ECS treatments. Orzi et al (1987) have used the [14-C]-deoxyglucose technique ~ observed a diffuse metabolic depression one day after a single ECS whereas repeated ECS had no significant effects on CMR in most brain structures with the exception of a 31% increase in the hippocampus and a 25% decrease in the olfactory cortex. Inasmuch as we have no data on glucose metabolism at times later than 2 min after a single ECS, our results appear to be consistent with the observations of Sacks et al (1982~ showing an early increase in glucose metabolism after a single ECS and an a p ~ n t decrease after repeated ECS treatment. in summar3,, this study has demonstrated that the immediate effects of a single, brief, ECS on both rCGF and rCBF, consist of inhomogeneous increases followed by a rapid decline towards the corresponding control value. During the seizure, the percentage of increase in rCGF appears to be higher in brain-stem structures compared to mesencephalic and telencephalic structures whereas the percentage of increase in rCBF in these brainstem structures appears to be lower in comparison with the more anterior regions. However, there were no significant effects of repeated ECS on either rCGF or rCBF.

References Ackermann RF, Engel J, Baxter L (1986): Positron emission tomography and autoradiographic studies of glucose utilization following electroconvulsive seizures in humans and rats. Ann NY Acad Sci 462:263-269. Baker N, Ituebotter RJ (1972): Compartmental and semicompartmental approaches for measuring glucose carbon flux to fatty acids and other products in vivo. J Lipid Res 13:716-724. Balazs R ( 1969): Carbohydrate metabolism, in Lajtha A (ed), Handbook ofNeurochemistry, Voi 3. New York: Plenum Press, pp i-36, Barkai A! (1986): Interaction of drugs and electroshock treatment on cerebral monoaminergic systems. Ann NY Acad Sci 462:147-162. Barkai AI, Durkin M, Nelson HD (1990): Localized alterations of dopamine receptor binding in rat brain by repeated electroconvulsive shock: An autoradiographic study. Brain Res 529:208213. Barkai Al, Mahadik S, Rapport MM (1974): Flow in vivo of glucose carbon to brain proteins in rats: Effect of starvation. J Neurochem 22:511-516. Biegon A, ~sraeli M (1986): Localization of the effects of ECS on beta receptor binding sites in the ra', brain. Eur J Pharmacol 123:329-334.

ECS, Cerebral Metabolism, and Blood Flow

BIOl. I~¥CttLATAY |99| ~30:2f~282

28 i

Biegon A, Israeli M (1987): Quantitative autoradiographic analysis of the effects of ECS on 5 ~ 2 receptors in male and female rats. J Neurochem 48:1386-1391. Charney DS, Menkes DB, Phil M, Heninger GR (198 !): Receptor sensitivity and the mechanism of action of antidepressant treatment. Implication for the ethiology ~ therapy of depress~. Arch Gen Psychiat~ 38:! 160-1180. Cremer JE, Ray DE, Soma GS, Cunningham VJ (! 98 i): A study of the kinetic behavior of glucose based on simultaneous estimations of the influx and phos~rylatioa in brain regions of rats in different physiological states. Brain Res 221:331-342. Dunnett CW (1955): A multiple comparison procedure for comparing s e v e ~ treatments with a control. J Am Slat Assoc 50:!096-1121~ Fink M (1978): Efficacy and safety of induced seizures tEST) in man. Comp Ps:rchiat¢¢ |9:t-18. Fox PT, Reichle ME (I 986): Focal physiological uncoupling of cerebral b l ~ flow and oxidative metabolism during somatosensory stimulation in human sub)eels. Proc Nail Acad Sci (USA~ 83:! 140--I 144. Hawkins RA, Miller AL, Cremer JE, Veech RL (1974): Measuremen~ of the rate of glucose utilization by rat brain in vivo. J Neurochem 23:917-~23. Hawkins RA, Mans AM, Davis DW, Hibbard LS, Lu DM (1983): Glucose availability to individual cerebral structures is correlated to glucose metabolism. J Neurochem 40:1013-10l 8. Horton RW, Meldrum BS, Padley TA, McWilliam JR (1980): Regional cereb~ blood flow in the rat during prolonged seizure activity. Brain Res 192:399-412. lngvar M, Siesjo BK (1983): Local blood flow and glucose metabolism in the rat b ~ n during sustained bicuculline-induced seizures. Acta Neurol Scand 68:129-144. Lou HC, Edvinsson L, MacKenzie ET ( 1987): The concept of coupling blood flow to b ~ n hncfion: Recision required? Ann Neurol 22:289-297. Lu DM, Davis DW, Mans AM, Hawkins RA ( 1983): Regional cerebral glucose utilization measures with [ 14C]-glucose in brief experiments. Am J P~'siol 245:C428-C438. Malitz S, Sackeim HA (eds) (1996): Electroconvu!sive therapy: Clinical and basic research issues. Ann NY Acad Sci 462:424 Medical Research Council ( 1965): Clinical mat of the treatment of depressive illness. Br Med J !:~86-888. Meldrum BS, Nilsson B ( 1976): Cerebral blood flow and metabolic rate early and late in prolonged epileptic seizures induced in rats by bicuculline. Brain 99:523-542. •

.an

at.B.

~

twa,~

,v

a.,w.a'l,

= ,i

r,t.i

.

'~,,..~..FUti,~l~ltltS ~

v~

~..~,,a~,'t.~lg¢,.U

llga,a..t.u,t.,ra~l;ig,.vl.glt

¢autt, t,~

&t,t.~v,T

epileptic seizures, hypoxia and hyperg|ycemia. In Elliott K, O'Connor M (eds), Cerebral Vascular Smooth Muscle andhs Control. CIBA Foundation Syrup, New York: Elsevier, 56:199214. Orzi F, Passarelli F, Diana G, Fieschi C (1987): Effects of single and repeated ECS on local cerebral glucose utilization in the conscious rat. Brain Res 423:144-148. Perumal AS, Barkai A! (1982): Beta-adrenergic receptor binding in different regions of rat brain after various intensities of electroshock: Relationship to postictal EEG. J Neurosci Res 7:289296. Prohovnik 1, Sackeim HA, Decina P, Malitz S ( 1986): Acute reductions of regional cerebral blood flow following ECT. Ann N¥ Acad Sci 462:249-262. Pulsinelli WA, Levy DE, Duffy TE (1982): Regional cerebral blood flow and glucose metabolism following transient forebrain ischemia. Ann Neurology 11:499-509. Roy CW, Sherrington CS (! 890): On the regulation of the blood supply of the brain. J Physiol (Lond) ! !:85-108.

BlOlo PSYCHIATRY I~1.30:269-282

A,I. Barkai et al

Sacks W, Sacks S, Badalamenti A, Fleischer A (1982): A proposed m e ~ for the determination of cerebral regional intermediary glucose metabolism in humans in vivo using specifically labelled I I IC]-glucose and PETT. I. An animal model with [14C]-glucose and rat brain autoradiography. J Neurosci Res 7:57-69. Sokoloff L (1981): Relationships among local functional activity, energy metabolism and blood flew in the central nervous system. Fed Proc 40:2311-2316. Sokoloff L, Reivich M, Kennedy C, et al (1977): The |14C]-deoxyglucose method for the measurement of local cerebral glucose utilization: theory procedure and normal values in the conscious and anesthetized albino rat. J Neurochem 28:897-916. Sulser F (1987): Serotonin-norepinephrine receptor interactions in the brain: Implications for the pharmacology and pathophysiology of affective d i ~ r s . J Clin Psychiatry 48:12-18. Young WL, Josovitz K, Morales O, Chien S (1987): The effect of N i ~ i p i n e on postischemic cerebral glucose utilization and blood flow in the rat. Anesthesiology 57:54-59.