A N N A L S O F BIOMEDICAL E N G I N E E R I N G 1, 2 4 6 - 2 5 3
(1972)
Power Spectrum Analysis of EEG Synchronization Following Application of Serotonin to Area Postrema* JOSEPH D . BRONZINO, 1 J. BRUSSEAU 2 P. J. MORGANE AND W . C. STERN Worcester Foundation for Experimental Biology
Received March 27, 1972 The role of serotonin and of the area postrema in synchronization of the neocortical electroencephalogram (EEG) was investigated in the present studies by observing the effects of different drugs (serotonin, norepinephrine, acetylcholine, and Xylocaine) applied topically at the site of the area postrema in cats. Using Fast Fourier Spectral analysis and power density spectra techniques it was found that serotonin increased the low frequency components and decreased the high frequency components in the cortical EEG. Application of serotonin to the floor of the fourth ventrical 6 mm rostral to the area postrema never produced EEG synchronizing effects. Xylocaine applied directly to the area postrema, as well as lesions of this region, decreased the low frequency components of the EEG while norepinephrine and acetylcholine produced variable effects. These results indicate that a serotonergic-sensitive mechanism, which induces EEG synchronization, exists in the region of the area postrema. INTRODUCTION A diverse body of evidence has been accumulated which suggests that brain s e r o t o n i n is f u n c t i o n a l l y r e l a t e d to sleep. I t has b e e n d e m o n s t r a t e d t h a t w h e n s e r o t o n i n is a p p l i e d b y v a r i o u s r o u t e s [ i n t r a v e n o u s , B r a d l e y (1958), R o t h b a l l e r (1957); i n t r a - a r t e r i a l , R o t h et al. (1970), K o e l l a (1966); i n t r a v e n t r i c u l a r , B r a d l e y (1958), K o e l l a (1963, 1966)], e l e c t r o e n c e p h o g r a p h i c ( E E G ) o c u l a r a n d b e h a v ioral signs o f s l o w s l e e p a r e i n d u c e d . I n e x a m i n i n g t h e n e u r a l s y s t e m s t h r o u g h w h i c h s e r o t o n i n c o u l d e x e r t its i n f l u e n c e o n sleep, s o m e o f o u r e x p e r i m e n t a l r e s u l t s h a v e h i g h l i g h t e d t h e p o s s i b l e role o f t h e r e g i o n o f t h e n u c l e u s t r a c t u s solit a r i u s ( N T S ) , i n c l u d i n g t h e a r e a p o s t r e m a ( A P ) . M a g n e s et al. (1961) d e m o n s t r a t e d t h a t l o w f r e q u e n c y s t i m u l a t i o n in the r e g i o n o f t h e N T S r e s u l t e d in E E G s y n c h r o n i z a t i o n , w h i l e B o n v a l l e t et al. (1961, 1963) o b s e r v e d , a f t e r l e s i o n i n g t h e same region, that the arousal produced by reticular and nociceptive stimulation w a s m o r e i n t e n s e a n d p r o l o n g e d . T h e y p o s t u l a t e d t h a t a f e e d b a c k s y s t e m inhibiting e x c e s s i v e a r o u s a l e x i s t s b e t w e e n t h e r e g i o n o f t h e N T S a n d t h e m i d b r a i n reticular formation. Additional evidence supporting the existence of such a neural c i r c u i t b e t w e e n t h e s e s t r u c t u r e s h a s b e e n p r o v i d e d b y B r o n z i n o (1969, 1970) using evoked response techniques. * This research was supported by N.I.M.H. Grants 0221 and 10625. 1 Also of Department of Engineering, Trinity College, Hartford, Connecticut. Also of Western New England College, Springfield, Massachusetts. 246 Copyright 9 1972 by Academic Press, Inc. All rights of reproduction in any form reserved.
POWER
SPECTRUM OF SEROTONIN
EFFECT
ON EEG
247
In addition, it has been noted by Koella et al. (1966) and our laboratory (Bronzino et al., 197 l) that topical application of serotonin to the area postrema in cats produced increased synchronization of the EEG. Our results were obtained using fast Fourier techniques to analyze the variation in the magnitude of the Fourier coefficients in six different frequency bands (0-1 Hz, 1.5-3.5 Hz, 4-7 Hz, 8-13 Hz, 14-18 Hz, and 20-30 Hz) as a function of time. Presented in this manner, it was consistently observed that a significant increase in the magnitudes of the Fourier coefficients occurred in the low frequency band (0-1 Hz) upon application of serotonin to the area postrema. Roth et al. (1970) showed that infusion of minute amounts of serotonin into the arterial circulation supplying the area postrema in rats also resulted promptly in low frequency shifts in the EEG. The above studies strongly suggest that one source of synchronization of the E E G originates in the ponto-medullary area, probably the area postrema, and possibly involves a serotonergic pathway. The present studies were carried out to further quantify these effects by using power density spectrum techniques. Most recent developments (Cooley and Tukey, 1965; Cooley et al., 1969; Theilheimer, 1969; Hord et al., 1965; Schallek et al., 1967) have been in the techniques used in performing the Fourier Transform to obtain the power density spectrum; however, the general framework and formulae follow closely the prescription given by Blackman and Tukey (1958).
EXPERIMENTAL PROCEDURES A series of twelve acute experiments were performed on adult cats of both sexes having body weights between two and four kilograms. The surgery was performed with the animals anesthetized with ether. Cannulae were inserted into the femoral vein and trachea following which the animal was mounted in a sterotaxic frame. After a posterior craniotomy the cerebellum was aspirated to expose the floor of the fourth ventricle which was kept moist with gauze pads saturated with physiological saline. In these cats, initially anesthetized with ether, the anesthesia was discontinued after completion of the surgical procedures. The animals were immobilized by Flaxedil (0.3 ml/kg/hr) and artificially respired with a Harvard respirator. Throughout the experiment the wound margins were infiltrated with Xylocaine. Body temperature was kept within normal range by use of a heating pad. Following exposure of the floor of the fourth ventricle, four small screw electrodes were installed on the skull over the right and left temporal and right and left frontal cortices while a fifth reference screw electrode was mounted in the frontal sinus. E E G signals were obtained from these electrodes and were simultaneously recorded on a Grass polygraph and on magnetic tape using a Hewlett-Packard 4channel magnetic tape recorder for later analysis. In each experiment, 30-45 rain of baseline E E G were recorded prior to the initial drug application. Drugs were applied by placing small cotton pellets saturated with the drug solution on the site of interest under direct observation. A 'drug period' of approximately 1-hr duration was then recorded. Between successive drug periods, the pellets were removed and the area washed with physiological saline solution and a saline-soaked pellet placed on the area postrema. Cortical
248
BRONZINO
ET AL.
activity was then recorded during a 30-min "saline wash-off" period prior to another drug application. In all, there were 22 drug periods which stimulated the area postrema, 12 using serotonin creatinine sulfate, 4 using norepinephrine bitartrate, 3 using acetylcholine chloride and 3 using Xylocaine (lidocaine hydrochloride 1%). In 4 additional drug periods, serotonin was applied approximately 6 mm rostral to the area postrema. Serotonin, norepinephrine and acetylcholine were dissolved in physiological saline in concentrations of 3/zg/ml as salt. Three different sets of experiments were performed during which these drugs were applied. In one group of experiments serotonin was the only drug administered, while in the second set of experiments acetylcholine, norepinephrine, Xylocaine, and serotonin were applied in a rotating sequence (i.e., one drug was initially used, followed by a saline wash-off period, followed by application of a second drug, etc.). T h e order in which the different drugs were administered was varied in separate experiments to minimize systematic bias in the results. In the third and final set of experiments, serotonin was applied rostral to the area postrema.
ANALYSIS While the exact autocorrelation function for a random waveform is given by
c (7) = ~
1 fr/2
u J-~12 f ( t ) f ( t + r) dt
(1)
for all values of r, we are forced to limit ourselves to calculations involving relatively short signal samples. U n d e r these conditions we may calculate an apparent autocorrelation function given by
1 ((~,, -Irl)/2 Coo(7) -- T, - Irl a-<~,, - hi)/2f ( t -- r / 2 ) f ( t + 7/2) dt
(2)
for It] ~< T,, <~ T,, where T,, is the maximum lag time which we wish to use. This maximum lag time should usually be limited to five or ten per cent of the available sample length Tn (Blackman and T u k e y , 1958). The apparent autocorrelation function therefore is at best a reasonably good approximation to the true function o v e r the interval for which it is defined and is truncated at r = +-Tm. If we assume that the maximum lag time and the sample length are chosen judiciously, we may say that
Coo (r) = c ( r )
(--Tin ~< 7 <~ Tm).
(3)
I f now the apparent autocorrelation function is modified by multiplication by an appropriate window function D (r) which is zero for values of r outside the above interval and is symmetrical about 7 = 0, we will have D (7)Coo(7) ~ D (r) c ( 7 ) .
(4)
Fourier transformation of both sides of the above equation yields an apparent power density spectrum Pa ( f ) given by Pa(f) ~- QO c) * P ( f ) ,
(5)
POWER SPECTRUM OF SEROTONIN EFFECT ON EEG
249
where Q ( f ) is the transform of D (~-), P ( f ) is the true p o w e r density spectrum, and * denotes convolution. T h e window function chosen for this analysis is the Hanning window given by D (~') = g 1[ 1 -t- COS (Tr'r)/Tra], for 171 -<
Tin,
D(z) = 0 , for 171 > r ....
(6)
Q ( f ) for this function has side lobes whose heights fall off rapidly and are in general less than two percent of the height of the main lobe. T h e choice of a cosine window function of the form selected was simply to obtain an accurate estimate of the true p o w e r density. T h e analysis scheme used is shown in Fig. 1. A Saicor Correlation and Probability A n a l y z e r was e m p l o y e d to c o m p u t e the autocorrelation functions. This instrument accepts analog signals with a minimum rms value of 100 mV. T h e input signal is then quantized and digital circuitry is used in all of its remaining operation. T e n second samples (i.e., T , = 10 sec) of the E E G signals obtained from the frontal lobe were p r o c e s s e d using lag increments of 10 m s e c and a m a x i m u m lag time of 1 sec. A fast Fourier transform subroutine, utilizing the C o o l e y - T u k e y algorithm and a plotting subroutine were used with the I B M 1130 c o m p u t e r system to calculate the p o w e r density spectra and to plot these functions. All spectra were normalized to a total p o w e r of a nominal watt. D u e to limitations of the plotting routine of the
INPUT (EEG FROM POLYGRAPH)
]
TAPE RECORDER
FOURIERTRANSFORM AUTOCORRELATOR
I WINDOWFUNCTION
[
MODIFIER
PLOTTER
I OUTPUT
u COMPUTATIONAL SCHEME
FIG. 1. Data analysis scheme employing a Saicor Correlation and Probability Analyzer to compute the autocorrelation function and an IBM 1130 computer to calculate the power density spectra. The results were plotted out in the form presented in Figs. 2 and 3.
250
B R O N Z I N O ET AL.
IBM 1130 the horizontal axis is labeled in terms of N, where frequency = ( N -- 1) 0.391 Hz. In order to check out the entire computational system, a square wave of frequency 4 H z and having a small dc offset was run through it. T h e resulting power spectrum consisted of scaled replicas of the Hanning spectral window at f = 4 Hz, f = 12 H z , f = 20 Hz, etc. (which are the harmonic frequencies of the square wave), plus one at f - 0 corresponding to the dc component. These scaled replicas of the window function centered about the harmonic frequencies 4, 12, 20 Hz, etc. had relative heights given by 1, 1/9, 1125, etc., respectively, which are correct
T. . . . .
1 0 0 1 ................
i 901 ................ i
B-
80 i......
I
MINUTE
AFTER
90 i .........
: .........
:
...........................
...............................................
7oi
60 ! ......... :. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
i. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
i E
I..u
: .........
J
......
~. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
60~ .............
: ......
A - CONTROL :. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
B- 10 M I N U T E S A F T E R S E R O T O N I N 80i ..................................................
:. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
:. . . . . . . . .
i
I
i
SEROTONIN
i
70~ . . . . . . . .
I O0 ;...............
! . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A - CONTROL i ....... . ....................................
......
S0~ . . . . . . . . . . . .
so~ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
~. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . p
-i-
............
40!
i. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
40 : - - ~ . - B
/.[
i i
i
C
i
i
;~
~---
T i
:,
..........................................
i
i',A~
zoi~
~__::
.
'i
i
i
o-. . . . ~ - - - - - ~ o
...... }~o----~o--
-~----~
. . . . . ~o--
0
N
UNITS
m
z
; }--
~oi ......... i ....... i ....... ~....... >-
i
..... i ~ - ~ - - - - } - - i
10
20
30
i--i 40
S0
i-60
70
uJ
s
I O0
........................
I00~
~. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A
CONTROL
....................
~ ........
T. . . . . .
- CONTROL : ....... ! ......
O
801~- ....
: .......
i .........
~. . . . . . .
:. . . . . . . . . . :. . . . . . . .
~. . . . . . . .
~....
80 i................
7o:i-~
i .........
! .........
i .......
i .........
i .........
i ....
70i ...........................
B
.....
:.......... :.
A 90 !........... : .......
u~
i
i.......
. . . . . . . . .
90~
. . . . .
: . . . . . . .
:
. . . .
AFTERS E R O T O N I N
20 M I N U T E S
i ........
B
s i.u
N
I ..J
60 i --#. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6 0 ~. . . . . . .
soi
<
.
O z
~o! ....... ~....... !-----i i ,, : .....
0 i ..... 0
..... i
!.....
i....
i--i--
:
i__!__i_
'
'
'
60
70
..... i..... i..... i.....
~ ......
I0
i ......
20
i .....
30
~ - - - - L - - ~
40
SO
.
.
~. . . .
30
T. . . . . . . . . . . . . . . . . . . . .
!. . . .
MINUTES
!.........
:----:
:----:
AFTER
. . . .
:--
SEROTONIN
..............
:
.
-'--
FREQUENCY
~ HZ
:
-:
2o
3o
4o
so
6o
7o
(N-I)(0.391)
FIG. 2. N o r m a l i z e d P o w e r D e n s i t y spectra analysis of the E E G w h e n serotonin (3 ~g/ml) was applied directly to the area p o s t r e m a of a cat; N repre s e nt s the nth term of the F a s t F o u r i e r Spectrum. The control graph (A) on each plot used for c o m p a r i s o n was calculated from the E E G record 1 rain before application of serotonin. It will be noted that there is a significant increase in the low frequency c o m p o n e n t s present in the E E G .
POWER
SPECTRUM
OF SEROTONIN
EFFECT
251
ON EEG
pulse heights for the power spectrum of a square wave. If we content ourselves with a frequency resolution of 1/TmHZ the system used is entirely satisfactory and corresponds closely with the resolution suggested by Blackman and Tukey
(1958). RESULTS
In all, more than 200 power density spectra were computed and plotted. Figures 2 and 3 show representative samples of the results. Following application of serotonin, the power density spectra always showed shifts toward the low 100
100:
, ,[
i A-'CONT.OL 90 ~ t Ill 80 --~i
~7o
!
' ' ' I MINUTE AFTER
~
. . . . . . . . . .
60
i---ii i i. . . . . i
. . . . . . . . . . . . ~ i i i
A
-"CONTROL'
B
10 M I N U T E S
t
! .
~o
-i ....
i---i---i
....
40i. . . . .
~__
2~
.... i--
"N
i
0"0
I0
20
30
40
50
60
. SEROT(~NIN ...h
I
30i. . . . .
~Z
.
i
~---i---
g ~o ~ - i - ~ i - - ~ < - i - - i
.
60i---
i
50j-----
.
AFTER
'o i 7oi~
!----
'- ..... i......... i i i
""
,oi
SEROTONIN
\l
i
i
I
I
i
!
40
50
60
1
0L 0
70
L __[ . . . .
10
20
30
70
1.1.1
UNITS I O0 . . . .
:".....
i
O
a,.
u.i
N m
,,-I
<
90i"
i
1
i
; ......
i .....
B-20
'~701. ........i--[ i i
J
~......
i----~
A - CONTROL :. . . . AFTER
.....
MINUTES
901
:----~'--
B
70
i.
J i
._L_i~_i ' dL_.!~..L ....
,,,
!I _ L ~.... L_
30MINUTES
Ai -
l.g~
, i___Lr_\l__[__!
'-I
i
%(, \
,'I,, . . _ _ ~ A
-0
10
20
30
40
i
i
I i
60
70
........
50
FREQUENCY
i
i
j
iI
I
I
.,1,
!
q
= (N-I)
,
t
1--
_-1--1--
l_
!XlA'
0 HZ
--~ SEROTONIN
I
,0 d~l o!
I AFTER
"
50
i---
r--
i........ -: , :
6~ ' '
i
i
SEROTONIN
'~ i
SO L'-/---|k'----"
O z
IO0~
: ......................
A - CONTROL
1 !__]
1
I0
1"
20
30
40
I t__t--L 50
60
70
(0.391)
FIG. 3. N o r m a l i z e d P o w e r D e n s i t y s p e c t r a a n a l y s i s of the E E G w h e n s e r o t o n i n (3 jxg/ml) was topically applied to the area p o s t r e m a . It will be noted that the s e r o t o n i n effect was o v e r after approximately 40 rain. e v e n t h o u g h the cotton pellet containing s e r o t o n i n w a s still applied to the area postrema.
252
BRONZINO ET AL.
frequencies. The control curves (A) in the figures used for comparison were calculated from the E E G record 1 min before each separate drug run. These figures illustrate the increased low frequency components present in the E E G during the first 30 min after application of serotonin. It will be noted in these figures that the serotonin effect was over after approximately 40 min even though the cotton pellet containing serotonin caused a marked increase in relative power at the low frequencies as well as a very noticeable increase in the periodic component of the signal. It was noted that in cases where the E E G was already somewhat synchronized before drug application, application of serotonin invariably led to increased synchronization as defined by Roth e t al. (1970). Application of Xylocaine produced an increase in relative power densities at higher frequencies. Norepinephrine and acetylcholine showed mixed results, sometimes producing an increase in relative power densities at lower frequencies, and sometimes at higher frequencies. None of these showed the consistent results obtained with serotonin. In addition, when serotonin was applied rostral to the AP, synchronization did not occur. DISCUSSION These experimental results indicate that the topical application of serotonin to the area postrema produces greater synchronization of the EEG. It is important to note that this was apparent even in those animals in which the control E E G was already somewhat synchronized. In this case, our analysis clearly demonstrated a shift from a synchronized to a m o r e synchronized state. Only serotonin consistently produced shifting of the power density towards low frequencies, an effect not apparent in visual examination of the E E G record. This signal analysis technique provides a sensitive measurement of frequency changes in the E E G that may be brought about by chemical stimulation. As noted, serotonin appears to be an important neuro-chemical involved in the processes controlling sleep. The question arises, however, as to the manner in which this substance affects synchronization of the EEG. It is possible that increases in serotonin levels in blood or in cerebrospinal fluid may result in E E G synchronization by acting on serotonin-receptor sites located in the area postrema. This area is well suited for producing these effects since the permeability of the blood-brain barrier is considerably increased at this site. The present experiments thereby suggest that the area postrema may be the site of serotonin receptors. REFERENCES BLACKMAN, R. B., AND T1JKEY, J. W. The measurement o f power spectra. Dover, New York. BONVALLET, M., AND BLOCH, V. Bulbar control of cortical arousal. Science 1961, 133, 1133-1134. BONVALLET, M., AND ALLEN, JR., M. B. Prolonged spontaneous and evoked reticular activation following discrete bulbar lesions. Electroencephalography and Clinical Neurophysiology 1963, 15, 969-988. BRADLEY, P. B, The effects of 5-hydroxytryptamine on the electrical activity of the brain and on behavior in the conscious cat. In G. P. Lewis (Ed.), 5-Hydroxytryptamine. London: Pergamon, 1958. Pp. 214-220.
POWER SPECTRUM OF SEROTONIN EFFECT ON EEG
253
BRONZINO, J. D. Verification of neural pathway between the reticular formation and the nucleus tractus solitarius. Proceedings o f 8th International Conference on Medical and Biological Engineering. 1969, 11, 4.3. BRONZINO, J. D. Effect of serotonin and Xylocaine upon evoked responses established in neural feedback circuits associated with sleep-waking process. Biological Psychiatry 1970, 3, 217-226. BRONZINO, J. D., MORGANE, P. J., AND STERN, W. EEG Synchronization following application of serotonin to area postrema. American Journal o f Physiology, 1972, 223, 376-383. COOLEY, J. W., AND TUgEY, J. W. An algorithm for the machine calculation of complex Fourier series. Mathematics o f Computation 1965, 19, 297-301. COOLEY, J. W., LEWIS, P. A. W., AND WELCH, P. D. The Finite Fourier Transform. IEEE Transactions on Audio and Electroacoustics 1969, AU-17, 77-85. HORD, D. J., JOHNSON, L. C., LUBIN, A., AND AUSTIN, M. T. Resolution and stability in the autospectra of EEG. Electroencephalography and Clinical Neurophysiology 1965, 19, 305-308. KOELLA, W. P., AND CZICMAN, J. S. Influence of serotonin upon optic evoked potentials, EEG, and blood pressure of cat. American Journal o f Physiology 1963, 204, 873-878. KOELLA, W. P., AND CZICMAN,J. S. Mechanism of the EEG synchronizing action of serotonin. American Journal o f Physiology 1966, 211, 926-934. MAGNES, J., MORUZZI, G., AND POMPEIANO, O. Synchronization of the EEG produced by lowfrequency electrical stimulation of the region of the solitary tract. Archives ltaliennes de Biologie 1961, 99, 33-67. ROTH, G. I., WALTON, P. L., AND YAMAMOTO, W. S. Area postrema: Abrupt EEG synchronization following close intraarterial peffusion with serotonin. Bruin Research 1970, 23, 223-233. ROTHBALLER, A. B. The effect of phenylephrine, methamphetamine, cocaine, and serotonin upon the adrenaline sensitive component of the reticular activating system. Electroencephalography and Clinical Neurophysiology 1957, 9, 409-417. SCHALLEK,W., LEWINSON, T., AND THOMAS,J. Power Spectrum analysis of drug effects on electroencephalogram of cat. International Journal o f Neuropharmacology 1967, 6, 253-264. THEIEHEIMER, F. A matrix version of the Fast Fourier transform. IEEE Transactions on Audio and Electroacoustics 1969, AU-17, 158-161.