THE ANALYSIS OF INTER-SPIKE INTERVAL FLUCTUATION OF

The Japanese Journal of Physiology

19, pp.243-259, 1969

THE ANALYSIS OF INTER-SPIKE INTERVAL

FLUCTUATION OF HYPOTHALAMIC UNIT

ACTIVITY IN RESPONSE TO LUTEINIZING

HORMONE AND OXYTOCIN

Masazumi KAWAKAMI AND Hideo SAITO

2nd Department of Physiology, Yokohama City UniversitySchool of Medicine, Yokohama

The activity of the hypothalamic neurones is characterized by its suscepti-

bility to the influence of various neural and humoral factors. It is reported

that some neurones in the ventromedial or lateral hypothalamus show altera-

tion in the pattern of firing activity, paralleling the change in neocortical

activity such as appearance or disappearance of sleep spindle or desynchroni-

zation. On the other hand, neurones were found which increase or decreasedischarge activity, with the change in humoral factors, like luteinizing hormone

(LH), oxytocin or antidiuretic hormone (ADH)1,2,3,4,5). In fact there are not afew neurones that almost keep silent in normal conditions and only fire under

specific influence of humoral or neural factor. This is quite a contrast to mostof the thalamic or reticular neurones6,7). In order to understand this kind of

discharge activity, it seems necessary to introduce a concept of system con-

trolling this particular neurone. It may be possible to assume a change inthe way of regulation of these systems, perhaps hyperactivation of a certain

specific system among them, under a certain hormonal condition.

The traditional procedure of quantitative analysis of firing activity, such

as calculating discharge rate, seems unsatisfactory, then. MOUNTCASTLE8) has

pointed out the inadequacy of regarding the variability existing in what hadbeen thought to be steady states of neuronal activity as a statistical error and

proposed an idea to treat it as representing properties of the activity. Byanalyzing the thalamic activity with respect to the distribution of discharge

interval, he found that driven activity by sensory stimulation reveals an

entirely different mean-SD relation, i. e. a different distribution.The analyze the hypothalamic activity which seems to be "driven" by a

certain humoral or neural factor. Thus, statistical analysis was applied here

for the elucidation of action of LH or oxytocin on hypothalamic unit activity,

Received for publication July 1, 1968

川上正澄,斉藤英郎

243

244 M. KAWAKAMI AND H. SAITO

perhaps through activation of some unknown system.

MATERIALS AND METHODS

Experiments were made acutely on 85 mature female cats ranging from 2.0-3.5kgin body weight, which were overiectomized 3-4 weeks prior to the experiment . Inorder to produce the estrous state estradiol benzoate was administrated in two successivesubcutaneous injections of 0.08mg, 24 hours apart . The animals were anesthetized withether during surgical procedures and the cranial bone and duramate of both sides wereremoved to expose the brain surface and to allow insertion of recording microelectrodes .To prevent the movement of the cerebrum due to respiration the skull window wascovered with agar and an opening was made in the duramata of the occipital formento allow the outflow of cerebrospinal liquid . For intravenous injection a small poly-ethylene tube was inserted deeply into a femoral or brachial vein without stoppingcirculation. The animal's head was held firmly in place by a stereotaxic device andthe body rested on an electrical heating pad to maintain a rectal temperature of over37℃. Artificial respiration with a small stroke volume at a rate of 20 per minute was

used to minimize body movement and respiratory blood pressure changes throughoutthe experiment. After the operation , ether anesthesia was discontinued and the animalwas immobilized by a minimum dose of gallamine triethiodide . For cortical EEG re-cording silver bipolar electrodes were placed on the presigmoid gyrus . To record theunitary activity of a single neurone in the hypothalamus , stainless steel insect pin(size No. O-Japanese) electrodes were used, electropolished in a mixed solution ofpotassium chloride and hydrochloride acid and coated with insl-x except at the tipwhich was iess than 1μ in diameter. The microelectrode was connected to a cathode

follower preamplifier (Nihon kohden , type MZ-3B). The output was fed into a poly-graph (Nihon kohden, type RM-150) for the continuous registration of neural impulseswith presigmoidal and hippocampal EEG recording . The shape of unitary spikes wasmonitored simultaneously through a dual-beam oscilloscope (Nihon kohden type VC-6The discharge activity was recorded magnetically where required . The data wereanalyzed either through on-line computer (Nihon kohden , ATAC 501-20) or by hand.Electrodes were oriented according to the stereotaxic atlas of JASPER22) . Recordingwas started 6 hours after recovery from ether anesthesia . Isotonic physiological saline(0.3-0.5ml) was injected intravenously before and after oxytocin or LH administration ,and it was confirmed that no change occurred in any form of electrical activity due tosaline. A tapered glass rod with a rounded end was used for mechanical stimulationof vulva, vagina or cervix for one minute . The rod was lubricated with liquid paraffinbefore insertion into the reproductive tract. As a control for vaginal stimulation

, sti-mulations by other means such as stroking thigh or pinching tail , were performed. Todetermine the position of the electrode tip, a small amount of iron was deposited atthe recording point by passage of a 5μA current for one mimte at the con{nletion of

each experiment. The cat was then perfused by way of the carotid arteries withRinger's solution followed by a mixture of two solutions (one part straight formalinewith 10% trichloroacetic acid and 9 parts 3% potassium ferrocyanide with 3% potassiumferrocyanide with 3% potassium ferricyanide) following by Ringer's solution . WI-1cmhardened, frozen sections of about 50μ thickness were cut and counterstained with

carmine. The luteinizing hormone used was . Armour LH (Lot No. R 377279); oxytocinpreparations were Sandoz Syntocinon and Teikoku-zoki Atonin-S. The same amount ofAtonin-S solvent was injected intravenously as a control after the administration of thetest solution, and proved ineffective on electrical activity of the brain .

LH AND OXYTOCIN SENSITIVE NEURONE IN HYPOTHALAMUS 245

RESULTS

A. Firing frequency of hypothalamic neurones

Firing activity of the neurones were classified with respect to frequency

as 0-2, 2-4, 4-6, 6-8, 8-10 and over 10 pps. Histograms of neurone population

were made for each hypothalamic subdivisions, as in FIG.1. Total number of

hypothalamic neurones investigated was 354.

In general, the firing frequency of the hypothalamic neurones was rela-

tively low as compared with that of the cerebral cortex and thalamus8,12). The

microelectrode study on the midbrain reticular formation by these authors

(unpublished) revealed that the unit activity in this area also shows a dischargeof high frequency (10-50 pps). Majority of the hypothalamic neurones fire at

the frequency of 0.1-10.0 impulses per second and it is quite rare to fire oftener

than 20 pulses per second. Firing frequency in most of the neurones (35%)

was below 2 pps, though there was some regional difference. In some hypo-

thalamic portions 60% of the neurones examined fired less oftener than 2 pps

while in other areas only 30% of the neurones did so. Neurones firing at over30 pps was quite rare and only 3% even in the lateral hypothalamic (LHA)where neurones were found to fire at relatively high frequency.

As estrus in the anterior hypothalamic area (AHA) the number of neurones

firing at lower frequency below 2 pps somewhat decreased and there was anincrease in 4-6 pps firing neurones. Neurones of higher firing frequencies

above 10 pps did not increase or decrease. About 50% of the neurones in the

ventromedial nucleus (VMH) fired less oftener than 2 pps, and the neurones

firing at above 10 pps were very rare. As estrus, there was a tendency toward

slowing down of firing. This is quite interesting in view of the fact this

portion plays some important role in the process of estrus circle in reflectorilyovulating animal like cats. However, as will be described later, many of the

neurones in this area are profoundly influenced by general activation state,and therefore it was sometimes quite hard to isolate the change in firing due

to estrous cycle from that caused by general activation. In the LHA, more

than 60% of neurones observed fired at above 4 pps at estrus. In contrast to

the VMH neurones, they tended to fire at higher frequency than at anestrus

(FIG.1).Thus the discharge frequency of the hypothalamic neurones as well was

found to be, affected by the relatively gradual change in endocrine factor.On the other hand, there exist in the hypothalamus those neurones which

respond to phasic change in the general activation level. In 54 out of 204

neurones distributed in the AHA, LHA, VMH and dorsomedial nucleus of the

hypothalamus (DMH), a change in firing frequency was recognized which

corresponded to the neocortical (pre-sigmoid gyrus) EEG changes in sleep-

wakefulness.


(ESTRUS) (ANESTRUS)

FIG.1. Histogram showing resting firing frequencies of sample neuronesin various portions of the hypothalamus recorded from castrated female catsimmobilized with guallamine. The number in the parentheses indicates thesize of the sample. Note that, during estrus, the number of units with lowestdischarge frequency increased in the ventromedial hypothalamic nucleus (VMH),while in the lateral hypothalamic area (LHA) and anterior hypothalamic area

(AHA) the units of higher firing frequency increased.

TABLE 1 summarizes the distribution of neurones which are related toneocortical EEG activation level. In the AHA and the LHA, the increase in

discharge rate precedes the transition of neocortical EEG stages from spindle

or slow wave sleep to arousal. On the other hand a majority of the VMH

neurones tended to show increase in firing frequency during the episodes ofthe sleep spindle burst in EEG of the motor cortex. Increase or decrease in


TABLE 1.

spike discharges preceding the EEG change by 1-4sec could be observed in

these neurones when they are not too active.

B. Interval histogram of hypothalamic unit discharges

The discharges of neurone as a train of pulses can be analyzed into dis-charge interval histogram. As the sample population, the inter-spike intervals

of the unit activity recording of a definite length (1, 2 or 5 minutes) were

taken. The size of the sample was over 300 in any case. The interval class

was 1/5-1/10 of the mean, which had been found most suitable for classifying

the distribution, on the basis of preliminary analysis. Within the limitation

(1)

(2)

(3)

(4)

(5)

FIG.2. Five different types of firing interval histogram in hypothalamicneurones. 1) exponential distribution. 2) Gamma distribution . 3) Poissondistribution. 4) Gaussian distribution. 5) unclassified.


of analysis procedure as above, the distribution of hypothalamic unit activity

would be classified into five types.

1. exponential distribution

2. gamma distribution

3. Poisson distribution

4. normal distribution

5. unclassified

It is already reported that discharge interval histogram of the VMH and

LHA neuronal discharges show exponential or gamma distribution10) . In 90%of the hypothalamic neurones examined in the present experiment , the intervalhistogram was classified as either exponential or gamma types.

Some neurones showed normal, Poisson or unclassified distribution. It is

occasionally observed that interval histogram of identical neurone laterates

easily from exponential to gamma distribution v. v. But it was never observed

that the neurones of exponential or gamma distribution showed other types of

distribution.

It was already reported that in estrus and anestrus LH or oxytocin exerted

different actions on firing rate of the hypothalamic neurones3), here the pulse

train of discharge activity an response to LH or oxytocin was analyzed from

FIG.3. Plot of mean interval and standard deviation of samples (n:300-3000) of resting activity of hypothalamic neurones each sampling

period being 1, 2 or 5 minutes. Abbreviations: V: ventromedial hypo-thalamic nucleus, L: lateral hypothalamic area, A: anterior hypothalamicarea. In most of the neurones examined the estimate of regression coef-ficient was 0.9-1.4. In a few neurones as L-2-3 and V-8, however thecoefficient took a lower value.


statistical viewpoint. On administration of oxytocin the discharge pattern ofhypothalamic neurones shows change, if it ever does, starting at 20-80 seconds

after administration, continuing for 5-15 minutes.

On LH administration the change appears 10-30 minutes after injectionand continues for several hours. The time series of neuronal discharge, before

and after administration, were respectively divided into segments of equal time

length. The length of segment was 1 or 2 minutes for oxytocin administration

and 5 minutes for LH administration. Then mean and standard deviation (SD)

of discharge interval was calculated for each segment and then the regressionline of mean and SD of interval was obtained. FIG.3 represents the regression

line between mean and SD of discharge interval in spontaneous neuronal dis-

charge of different distribution.

It is obviously recognized that the neuronal discharge showing exponential

of gamma type interval histogram reveals a relatively abrupt regression line

having regression coefficient above 1.0. The regression line of the discharge

of Poisson and normal distribution showed a more gradual slope.

C. The change in discharge activity of hypothalamic neurone induced by LH

administration

In hypothalamic neurones the influence of LH on discharge activity appears

after 10-30 minutes'latency1,2,3,11,15). The discharge rate of AHA or perivent-

ricular arcuate neurones increases, while that of VMH and some LHA neurones

decreases by LH administration. There were some cases where the mean

discharge frequency did not seem to be influenced by LH but standard devi-

ation did show significant changes.

Here the effect of LH administration on hypothalamic discharge activity

was examined with respect to the change in the regression between SD and

mean of discharge interval.

Anterior hypothalamus. Most neurones showed changes in discharge activity

paralleling the neocortical EEG activity. In anestrus spontaneous activity ofAHA neurones was classified into three types, that of exponential, gamma or

unclassified distribution. In all of them, the mean-SD regression revealed good

linearity with coefficient of regression of 1.0-1.4.

FIG.4 shows the examples of typical discharge interval histograms of two

neurones in this region. The mean discharge interval was 600-1000ms.

FIG.4, left, represents the influence of LH administration at anestrus upon

discharge activity of AHA neurones. LH administration induced shortage in

the mean interval and SD but little change was observed in the coefficient of

regression.

At estrus LH administration did not induce change in neurones which, at

control state, showed gamma or exponential distribution. However, as shown


ANESTRUS ESTRUS

CONTROL

FIG.4. The interval histogram and the plot of mean and SD in AHA neuronal

activity before and after LH administration at anestrus and estrus. The mean

interval declined by LH administration both at anestrus and estrus . However atanestrus the change occurred along one mean-SD regression line, while at estrus

LH administration resulted deviation downward from the control regression line .Note that the latter type of change was observed only in the neurones showing

unclassified distribution. The sample population was based on the discharge activity

of 5 minutes.


in FIG.4, right, in some neurones of uncrassified distribution, deviation from

the control regression line was observed.

Such neurones did not show corresponding change in discharge activity

with alteration in EEG of the pre-sigmoid gyrus.

Ventromedial hypothalamic nucleus. Interval histograms in most neurones were

gamma or exponential type, but a few neurones were found showing normal,

Poisson or unclassified distribution. Most neurones showed a correspondence

in the discharge activity with alterations in the neocortical EEG, but there

were some which did not show a distinct relation to the activation or sup-

pression of EEG activity. In spontaneous discharge activity, the coefficient

ANESTRUS ESTRUS

CONTROL

FIG.5. The interval histogram and plot of mean and SD in VMH neuronalactivity at anestrus (A) and estrus (B) before and after LH administration.Both of anestrus and estrus the mean interval increased after LH administration.Note that at anestrus the change occurred along the on regression line, while atestrus deviation from the control regression line resulted by LH.


of variation (C. V.) was constant for an individual neurone and the regressionshowed good linearity.

At anestrus in a majority of VMH neurones, LH administration inducedan increase in the mean interval and SD, but the coefficient of regression was

unchanged (FIG.5, left).

At estrus a few neurone were found which after LH administration showeda different SD-mean regression line. The mean interval extended but SD did

not follow it, hence the slope was remarkably decreased. Such neurones were

only found in those which revealed no correspondence with neocortical EEG

activity. It is to be noted that the interval histogram of these neurones were

of "unclassified" type (FIG.5, right).

In neurones of Poisson type or normal distribution, LH administration

proved ineffective.

Lateral hypothalamic area. Only one neurone out of 10 LHA neurones examinedhere showed a change in discharge activity by LH administration . It wasobserved in the case where the activity of two adjacent neurones was picked

up at a time from one electrode with discrete amplitude. One of the twoneuronal activities were of Poisson type distribution with a mean interval of

100-120ms and SD 50-60ms (CV: 0.5) and was not affected by LH administra-

tion. The other neuronal activity, with mean interval 200-250 ms and SD 170-270ms (CV: 0.7-1.2), showed an increase both in mean interval and SD by LH

administration. The coefficient of regression did not change, however .It is to be noted that these two neighboring neurones seemed to have

some interaction. As in FIG.6, cross-correlation analysis of the two neuronalactivities revealed a significant supression of the latter activity for 3-4ms

after discharge of the former neurone. On the other hand , discharge of thelatter neurone exerted no influence on the probability of firing of the former .

Thus the period of suppression of the latter neurone was found after the

former fired (FIG.6). Similar phenomenon was observed in some VMH neurone

(FIG.7).

D. The change in discharge activity of hypothalamic neurone induced by

oxytocin administration

Anterior hypothalamic area. As in FIG.8, left, at anestrus the neurone of

exponential distribution responded to oxytocin from 20 seconds to 6 minutes,

after administration, by a decrease in discharge internal, amounting to half

the control level, accompanied by a decrease in SD. The activity under influ-

ence of oxytocin occurred along the extension downward of the same regres-

sion line as the control.

Most of unclassified type neurones showed a decrease in the mean interval

and SD with no influence in coefficient of regression by oxytocin, as in the


(1)

(2)

FIG.6. The discharge interval histograms of two neighboring neuronesin LHA and the crosscorrelogram between them, under control condition.

(above) and after LH administration (below). One showed exponential dis-tribution (1) while another showed Poisson distribution (2).

exponential type. However two cases were found in which a change of entirely

different pattern occurred. During the control state they showed distribution

showing two peaks with a mean interval 30-40ms and coefficient of regressionabout 1.4. From 2 to 6 minutes after oxytocin injection, the mean intervals

significantly extended but SD did not increase in proportion to the increase in

mean. Therefore the slope greatly reduced (FIG.8, right).

In estrus neither type of neurone was influenced by oxytocin.


(1)

(2)

FIG.7. The crosscorrelogram between the discharge activity of twoVMH units (1) (2) recorded simultaneously from one electrode. Theinterval histogram of the unit (1) was exponential type while that ofthe unit (2) was normal. Note the decline in the probability of firingby the unit (1) for 10 ms after the unit (2) fired.

Ventromedial hypothalamic nucleus. In anestrus by oxytocin administrationneurones of gamma type distribution showed an increase in mean discharge

interval and SD from 20-30 seconds through 15minutes after injection. Therewere a few neurones among them which entirely ceased to discharge at about

2minutes after administration. After several minutes, then, the neuronerecovered its control state of activity. These changes were observed in neu-

rones connected to neocortical EEG activity. In a few exceptional cases, there

occurred a decrease in the mean interval and SD. However it was found that

these changes occur along the extention of the same regression line as the

control. The neurones of normal and Poisson distribution were not influenced

by oxyto cin injection.


MEAN INTERVAL

FIG.8. Interval histogram of successive 2 minutes' samples and the plot of itsmean interval and standard deviation in two AHA neurones before and after oxytocinadministration at anestrus. In the neurone shown in the left of the figure, the meanfiring interval and standard deviation decreased after oxytocin administration keepinga linear relationship. In the neurone, however, as exemplified in the right half, thelinearity of the mean-SD regression could not be observed after oxytocin injection.The interval histogram before oxytocin injection consisted of two peaks. After oxy-tocin administration, the peak of the shorter interval became less remarkable whilethe peak of longer interval became more dominant. The character "C" indicatessampling before administration.


In most cases oxytocin administration at estrus induced no remarkablechange in neuronal activity. However there were a few neurones which showed

a transient increase in discharge rate for 10 seconds, directly after oxytocin

injection. Only one out of 10 neurones showed an increase in the discharge

rate, keeping the same mean-SD relationship, as the control state. The neuronal

activity was of gamma type distribution and did not show a corresponding

change with pre-sigmoidal EEG pattern.

Lateral hypothalamic nucleus. In the dorsolateral neurones, most of which

showed no correspondence with neocortical EEG activity, by oxytocin admin-istration in anestrus, the mean discharge interval and SD decreased along the

extension of the control regression line. The changes were less durable than

those observed in the VMH neurones. On the other hand at estrus there weresome neurones in this area which showed an increase in mean and SD from

1 to 8 minutes after oxytocin administration. This change was also on the

same regression line as the control.

On the other hand, in the VMH neurones, which was found to have cor-

respondence with neocortical EEG activity, there were a few which responded

to oxytocin by a decrease in mean interval and SD in anestrus and an increase

in estrus.None of the neurones with normal or Poisson distribution responded to

oxytocin.

DISCUSSION

The attempt of regarding the neuronal activity of the diencephalon as a

statistical process and defining it, by the shape of distribution, is quite new

and data has not been satisfactorily accumulated. Among them POGGIO and

VIERNSTEIN12) reported that some of the inter-spike interval histograms ofthalamic neurones are of Poisson type. RODIECK9) noted in the cochlear nucleus

neurones showing Poisson, quasi-Gaussian and bimodal distribution. There

were also several neurones of unclassified distribution. With respect to hypo-thalamic unit activity, OMURA13) reported that the interval histogram could beclassified as either gamma or exponential distribution.

The present experiment also showed that majority of hypothalamic unit

discharges were of gamma or exponential type, but at the same time revealed

the existence of not a few neurones of Poisson, normal and unclassified dis-

tribution. The latter types of neurones are few in number and they seem to

belong to an entirely different neural system. For instance, while neurones

of gamma or exponential distribution reveal an alteration in firing activity in

correspondence with the change in neocortical EEG activity, those of Poisson,

exponential and unclassified distribution revealed no such correspondence.


Electrical stimulation of the midbrain reticular formation, by high frequency,

induced facilitation of discharge only in the former type of neurones. Thus

the neurones of gamma and exponential distribution seems to be at least

related to the reticular activating system, while other types are not.

The responses to LH or oxytocin also varied among neurones of different

distribution. The neurones of Poisson or normal distribution did not show any

kind of change in firing activity either by LH or oxytocin. The neurones of

gamma and exponential distribution as well as some of unclassified distribution,

in most cases, show an increase or decrease in mean discharge interval and

standard deviation. However the change in mean interval and SD was such

as could be plotted on an extension of one regression line obtained before

administration. That the mean and SD increased, but along one regression

line, means that the statistical properties of discharge activity was essentially

the same, was not affected by LH or oxytocin. In other words, a change may

have occurred in the threshold or excitability of the neurone itself, but the

systems controlling the neurone are the same. The previous findings that LH

implantation into the basal hypothalamus induced appreciable reduction in

pituitary LH content and plasma LH13,14) may support the assumption of the

existence of hypothalamic neurones directly affected by LH.

Among the neurones of unclassified distribution, there were a few which

showed a change by LH or oxytocin not only in mean interval and SD but

also in the relation between mean interval and SD, i. e. a definite deviation

from the control regression line. The sort of change can be only explained

by a hyperactivation of a few specific systems controlling that neurone, which

had not been active before administration. That this sort of change was found

only in the neurones of unclassified interval histogram may be understood in

terms of less stability in the regulating systems connected to it.

By the previous experiments, employing evoked potential technique and

EEG activation technique, it was found that at estrus midbrain reticular acti-

vating system, thalamo-cortical activating system and hypothalamo-pituitary

system are in functionally different state from at anestrus and show different

responses to LH and oxytocin11,15,16,17). In the present experiment existence of

hypothalamic neurones was confirmed which undergoes change in the statistical

characteristics of discharge activity by LH administration only when the

animal was in estrus state. The mechanism in which these neurones are affected

by LH in estrous state is unknown, but it is suggested that a new system,

discrete from the midbrain reticular, thalamo-cortical or hypothalamo limbic

system, may play a role here, since electrical stimulation of the midbrain

reticular formation does not affect the activity of these neurones according to

our supplementary experiment. It seems important that again activity of these

types of neurones are without exception categorized as unclassified. Elucida-

tion of the physiological and morphological basis of this kind of neurone,


belongs to the future investigation, as well as, its fiber connections and therelation to LH releasing factor.

SUMMARY

Non-anesthetized ovariectomized cats were employed for recording hypo-thalamic unit activity. The discharge pattern of the neurones which responded

to blood LH or oxytocin were analyzed from statistical view-point.

1. The pulse train of neuronal activity was analyzed into discharge interval

histogram. In addition the statistical dependency between mean discharge

interval and standard deviation was calculated. In three fourths of the hypo-

thalamic neurones the discharge activity showed exponential or gamma type

distribution of discharge interval and the rest showed either normal, Poissonor unclassified distribution. In the control condition the coefficient of regres-

sion was between 0.9 and 1.4 and the linearity was good. Although no essential

difference existed in the distribution of inter-spike interval between estrus and

anestrus, an increase in the discharge rate was observed in anterior hypo-thalamic neurones and a decrease in LH neurones.

2. Both at estrus and anestrus LH administration caused mean discharge

interval of the anterior hypothalamic neurones to decrease and of the VMH

neurones to increase. The change in discharge interval by LH at anestrus

was accompanied by an increase in standard deviation and the same mean-SD

relationship was kept as before administration.

On the other hand LH administration at estrus resulted in deviation ofmean-SD relation from the regression line of the control condition. The latter

type of change could be observed in a few neurones of unclassified distribution

in the VMH and AHA.

3. By oxytocin administration at anestrus the discharge rate of AHA and

LHA neurones increased and of VMH neurones decreased. On the contraryat estrus oxytocin injection caused discharge rate of lateral hypothalamic

neurones to decrease and of ventromedial hypothalamic neurones to increase.

In the majority of these cases, the change could be plotted on an extension

of mean-SD regression line obtained before administration. But there were a

few neurones which showed a deviation downward from the control regression

line. This type of change was observed only in neurones of unclassifieddistribution.

Discussion was made on the mechanisms underlying the changes in hypo-

thalamic neuronal activity induced by administration of LH and oxytocin.

These studies were supported by a grant from the Ministry of Education, Japan,and a grant from the National Institutes of Health, U. S. Public Health Service (NB-03860-5).


RESERENCES

1) BARRACLOUGH, C. A. AND CROSS, B. A. (1963.) Unit activity in the hypothalamusof the cyclic female rat: effect of genital stimuli and progesterone. J. Endocrinol.,26: 339-359.

2) RAMIREZ, V. D., KOISARUK, B. R., WHITMOYER, D. L. AND SAWYER, C. H. (1967)Effects of hormones and vaginal stimulation on the EEG and hypothalamic unitin rats. Amer. J. Physiol., 212: 1376-1384.

3) KAWAKAMI, M. AND SAITO, H. (1967) Unit activity in the hypothalamus of thecat: effect of genital stimuli, luteinizing hormone and oxytocin. Jap. J. Physiol.,17: 466-486.

4) BROOKS, C. McC., ISHIKAWA, T., KOIZUMI, K. AND Lu, H-H. (1966) Activity ofneurones in the paraventricular nucleus of the hypothalamus and its control. J.Physiol., 182: 217-231.

5) CROSS, B. A. AND GREEN, J. D. (1959) Activity of single neurones in the hypo-thalamus: effect of osmotic and other stimuli. J. Neurophysiol., 27: 878-892.

6) EVARTS, E. V. (1961) Effects of sleep and waking on activity of single units inthe unrestrained cat. In: The Nature of Sleep. pp.171-182. ed. by G. E. W. Wol-stenholme and M. O'Connor. J. & A. Churchill Ltd. London.

7) VERZEANO, M. AND NEGISHI, K. (1961) Neuronal activity in wakefulness and insleep. In: The Nature of Sleep. pp.108-126. ed. by G. E. W. Wolstenholme andM. O'Conner. J. & A. Churchill Ltd., London.

8) WERNER, G. AND MOUNTCASTLE, V. B. (1963) The variability of central neuralactivity in a sensory system and its implications for the central reflection of sen-sory events. J. Neurophysiol., 26: 958-977.

9) RODIECK, R. W., KIANG, N. T. S. AND GERSTEIN, G. L. (1962) Some quantitativemethods for the study of spontaneous activity of single neurones. Biophysic. J.,2: 351-368.

10) OMURA, Y., OYAMA, H., YAMAMOTO, T., NAKA, F., KOBAYASHI, N. AND ONO, T.

(1967) Neuronal mechanism of feeding. Progr. in Brain Res., 27: 1-33.11) KAWAKAMI, M. AND TERASAWA, E. (1967) Differential control of sex hormone

and oxytocin upon evoked potentials in the hypothalamus and midbrain reticularformation. Jap. J. Physiol., 17: 65-93.

12) PIGGO, G. F. AND VIERNSTEIN, L. J. (1964) Time series analysis of impulse sequ-ences of thalamic somatic sensory neurones. J. Neurophysiol., 27: 517-545.

13) CORBIN, A. AND COHEN, A. I. (1966) Effect of median eminence implants of LHon pituitary LH of female rats. Endocrinology, 78: 41-46.

14) DAVID, M. A., FRASCHINI, F. AND MARTINI, L. (1966) Control of LH secretion:role of a "short" feedback mechanism. Endocrinology, 78: 55-60.

15) KAWAKAMI, M., TERASAWA, E., TSUCHIHASHI, S. AND YAMANAKA, K. (1966)Differential control by sex hormones of brain activity in the rabbit and its physi-ological significance. In: Steroid Dynamics. pp.237-302. ed. by K. Nakao, G.Pincus and J. F. Tait. Academic Press, New York.

16) KAWAKAMI, M. AND SAWYER, C. H. (1959) Neuroendocrine correlates of changesin brain activity thresholds by sex steroids and pituitary hormones. Endocrino-logy, 65: 652-668.

17) KAWAKAMI, M., SETO, K., TERASAWA, E. AND YOSHIDA, K. (1967) Mechanismsin the limbic system controlling reproductive functions of the ovary with specialreference to the positive feedback of progestin to the hippocampus. Progr. inBrain Res., 27: 69-102.

Documents

THE ANALYSIS OF INTER-SPIKE INTERVAL FLUCTUATION OF