Brain Research, 101 (1976) 67-78 ,~} Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands
67
S I N G L E U N I T R E C O R D I N G IN H Y P O T H A L A M U S AND PREOPTIC AREA OF E S T R O G E N - T R E A T E D A N D U N T R E A T E D O V A R I E C T O M I Z E D F E M A L E RATS
JOSE BUENO AND DONALD W. P F A F F
Rockefeller University, New York, N.Y. 10021 (U.S.A.) (Accepted June 30th, 1975)
SUMMARY
Single unit activity was recorded with micropipettes in the medial hypothalamus and preoptic area of urethane-anesthetized ovariectomized female rats. Some females had received long-term estradiol treatment, while others had been left untreated. In the medial preoptic region and bed nucleus of the stria terminalis, estrogen-treated rats had fewer cells (compared to untreated rats) with recordable spontaneous activity, due primarily to a loss of cells with very slow firing rates. In the basomedial hypothalamus, estrogen-treated rats had more cells (than untreated rats) with recordable spontaneous activity, due primarily to an increase in the number of cells with slow firing rates. Responsiveness of neurons to somatosensory stimulation was generally low. If present it was depressed by estrogen treatment in medial preoptic area and bed nucleus of stria terminalis, while it tended to be elevated by estrogen treatment in medial anterior hypothalamus and basomedial hypothalamus. Differences in the effects of long-term systemic estrogen treatment on medial preoptic neurons compared to basomedial hypothalamus are paralleled by differences in the control of lordosis by these neurons in female rats.
INTRODUCTION
Microelectrode recording has allowed the demonstration of steroid hormone effects on hypothalamic electrophysiology. In some experiments local applications of steroids have been used to prove that the hormone effects are direct. Steiner et al. 31,34 showed effects of a synthetic glucocorticoid on hypotbalamic units, using iontophoretic application of the steroid. Testosterone can affect neurons in the preoptic area following either local or systemic application zs. Regarding ovarian hormones, several experiments have shown variations in
68 hypothalamic unit activity during the estrous cycle of the female rat3-~-:', ~l.~v~.~ These effects have been reviewed2~,mL However, during the estrous cycle blood levels of several hormones are changing simultaneously, making a complicated situation for the analysis of any individual hormone effect. The locations of estradiol-concentrating neurons in the hypothalamus, preoptic area and basal tbrebrain are known preciselye~L Short latency effects of intravenously injected estrogens have been recorded in antidromically identified preoptic neurons :~6. Since estrogens are never absent from the blood of the normally cycling female rat, one physiologically relevant comparison would seem to be between two groups of ovariectomized female rats: one group untreated and the other given daily injections of estradiol. Microelectrode recording for isolating single units is required, since it is known that hypothalamic neurons close to each other can have different physiological properties2'~, zS. For the present recording study, focussing on the mechanisms underlying the effects of estrogen on lordosis, progesterone injections were not used because ovariectomized female rats given longterm estrogen treatment will perform lordosis without progesteroner', ~. Special care was exerted in choosing peripheral stimuli for use during recording experiments. In order to focus on hypothalamic mechanisms of female rat reproductive behavior, we knew that somatosensory stimulation could be emphasized, because only somatosensory input is required for female rat lordosis ~4. Cutaneous stimuli on the female's flanks, rump and perineal regions were chosen to imitate stimulation provided by the male rat 27. The cutaneous stimuli used are known to be necessary 14 and sufficient6, '~9 for lordosis.
METHODS
Rats Recording was done on 42 female Sprague-Dawley rats, obtained from Hormone Assay Laboratories (Chicago, Ill.). All rats were ovariectomized at least 3 weeks prior to recording, and weighed between 250 and 350 g. Twenty females received daily injections o f 10/zg estradiol benzoate (subcutaneously in oil) per day, for at least 10 days prior to recording. This dosage was chosen to allow lordosis without the necessity of progesterone injections, and all 20 rats performed strong lordosis reflexes immediately before recording experiments. The other 22 rats did not receive estradiol, and none of these animals showed lordosis. Recording experiments were done without knowledge of whether or not a particular rat had been treated with estradiol.
Recording preparation Rats were anesthetized with urethane (1.1 g/kg, intraperitoneally), This low dose of urethane was chosen to help insure that hypothalamic units would be as active and responsive as possible. The head of the rat was stabilized with a stereotaxic instrument; To facilitate recording hypothalamic neurons close to the midline, where many estradiol-concentrating
69 cells can be found 26, the head was held at a left-right angle of 9 ° to 14° offthe horizontal plane. Body temperature was maintained by a heating pad throughout the experiment in the range of 36.5-37 °C. Rectal temperature was monitored only periodically, to avoid unnecessary somatosensory stimulation from the rear half of the body. After incision and retraction of the skin, the skull was thinned with a dental drill. The deepest layer of bone and the dura was then removed carefully with fine forceps. The cortical electroencephalogram (EEG) was recorded through two silver ball electrodes placed respectively in small holes through the frontal and parietal bones on the side opposite the opening for the micropipette. EEG potentials were amplified and displayed using a Grass Model 7 polygraph. During recording the trunk of the rat was suspended at the thorax and lower abdomen with two rubber bands to minimize inadvertent tactile stimulation. Single units were recorded with glass micropipettes filled with 2.5 M sodium chloride. Tip diameters were between 1 and 3 /~m, and DC resistances between 0.5 and 5 M ~ . The reference electrode was a silver ball electrode on the cortex. The micropipette was positioned using a micromanipulator (La Precision Cinematographique), and penetrations were made in the plane of the stereotaxic atlas of Kenig and Klippep 3. Input from the micropipette was led to a high input impedance preamplifier and then to a Tektronix 502 oscilloscope and an audio monitor. Frequencies below 400 and above 10,000 cycles/sec were filtered out. Each penetration was made with slow and systematic tracking. When the discharges of a single unit were isolated by small micropipette movements, they were quantitatively evaluated using a counting circuit equipped with a Schmitt trigger and a histogram generating circuit. At the end of each second this circuit puts out a voltage proportional to the number of spike discharges. The output of the counting circuit was displayed on-line as a histogram on one channel of the Grass Model 7 polygraph. This method provides a quantitative record for the evaluation of spontaneous activity, responses to stimulation, and relation to cortical EEG. Stimulation Somatosensory stimuli were chosen to test hypothalamic unit responses to input related to the lordosis reflex. The stimuli are designed to imitate those provided by the male rat during mating zv, and are known to be necessary 14 and sufficient6,29 for lordosis. Flank stimuli were composed of fast, repetitive bilateral stroking by the experimenter's thumb and index finger on the flanks, just anterior to the rear legs. Rump stimulation was gentle, repetitive palpation on the posterior rump and tailbase area. 'Fork' stimuli were brief applications of pressure on the perineal, tailbase and posterior rump areas by the experimenter's thumb, with the index and middle fingers 'forked' around the posterior end of the body on either side of the tailbase. This stimulus readily elicits the lordosis reflex in receptive female rats. Pain stimuli were intense pinches of the ear or paw. The precise time and duration of stimulus application was recorded with the signal marker on the polygraph, to be compared with possible alterations in unit activity or cortical EEG.
70
Data analysis For each cell recorded, average resting discharge rate was calculated from the histogram. The time for which average spontaneous discharge was calculated included the period long enough after the cell was isolated that stability of rate could be insured, and before any stimuli were applied. The fact that only a single unit was contributing to the record at any one time was insured during the experiment by monitoring the performance of the Schmitt trigger on the oscilloscope as well as by attending to the audio monitor. A response to stimulation was defined as a statistically significant change in firing rate, when the activity during stimulation was compared with 30 sec of immediately prior spontaneous discharge. Changes of firing rate of less than 30 o/~,were never counted as responses. Histograms of unit firing rates were compared to simultaneously recorded cortical EEG. Average firing rates during EEG synchrony were compared with those during EEG desynchrony, whenever both cortical E E G states were expressed during the time that unit was held. Also, possible changes in cortical EEG during stimulation were noted.
Histology The bottom of each micropipette track, at or near the bottom of the brain, was marked by a small electrolytic lesion. After each experiment the rat was killed by cardiac perfusion with 10 ~ formalin. Serial frozen sections, 100 # m thick, were cut in the plane of the atlas of Krnig and KlippeP 3, and then stained with cresyl violet. Using a microprojector the position of each electrode penetration was found and the location of each recorded neuron plotted on a drawing o f the appropriate section. Histology from preparations in which no single units were recorded was not considered, due to possible difficulties with electrodes, anesthesia, etc. The number of penetrations going through each anatomical structure was counted, and used for calculations in the results (i.e., some of the results were expressed on the normalized basis of 10 penetrations/brain structure). All anatomical terms and definitions of zones follow exactly the nomenclature of the KOnig and KlippeP 3 atlas o f the rat brain. We have combined unit samples from the arcuate nucleus, ventromedial hypothalamic nucleus and dorsomedial hypothalamic nucleus and the tissue between those nuclei, as defined by K r n i g and Klippel, to constitute a sample labeled basomedial hypothalamus (BM). A total of 328 single units were recorded in the anatomical'zones considered. For neurons in each anatomical zone, statistical comparisons between estrogentreated and untreated females were made. Differences in numbers of cells/penetration and in average resting discharge rates were evaluated parametrically using the t-test t° and non-parametrically using the Mann-Whitney U-tesOL Differences in proportions of cells responding to each stimulus were evaluated with a statistical test for differences between proportions 1°.
A.
,ol
71
B.
NST
E
~ ST
C. ( = A x B )
~-0
NST
(D 0 4-
+1 ~
0
~-
0
0
c0
C~
MPOA
POA
~) W
0
o3
c
2
~ -a o
-(~ (1)
E x o
0
o (D ~
-~
~
0
cn v
~
o
o
--
r"
09
0
50
0
__~ o
50
l/)
. . (D
.
.
o3
0
~.~ 0
c-
~,
50
×'~ 0
O3
E
"S_ c 0
x >
m w
~
-6 0
x o
~
I-=
50
x >
m w
Fig. 1. Comparison of single unit recordings from estradiol-treated (EB) and untreated (OVX) ovariectomized female rats. Numbers of neurons with recordable spontaneous activity (A) normalized according to number of electrode penetrations through each anatomical structure. Figure parts (B) and (C) are explained in text. Abbreviations: NST, bed nucleus of stria terminalis; MPOA, medial preoptic area, M A H A , medial anterior hypothalamus, BM, basomedial hypothalamus (combination of arcuate, ventromedial and dorsomedial nucleus recording sites). Definitions according to atlas of K6nig and Klippe113. Differences between estrogen-treated and untreated ovariectomized female rats: ** P < 0.01. RESULTS
Resting discharge rates Differences between unit recordings in estrogen-treated and untreated preparations were not the same in nucleus of the stria terminalis and medial preoptic area as in the basomedial hypothalamus. The numbers of cells/penetration with recordable spontaneous activity were greater for untreated than for estrogen-treated preparations in the nucleus of the stria terminalis and the medial preoptic area, while in the basomedial hypothalamus the opposite was the case (Fig. 1A). Among the cells recorded, when resting discharge rates were averaged for all ceils, without respect to the underlying firing rate distribution, there were no signifi-
72 cant differences between estrogen-treated and untreated preparations for any anatomical structure (Fig. I B). This is probably due to two factors. First, it was striking that even within the same preparation, units nearby each other (less than 200 Fm distance between them) could have very different resting discharge rates. Second, averaging spontaneous activity across units is a complicated calculation: such an average ca~t be increased either by a general elevation of firing rates or by the disappearance from the distribution of units with very low firing rates. To account for these two factors separately, distributions of unit liring rates were plotted (Fig. 2A). In agreement with previous results9,~6,1°, :~9the distribution shows that many cells in the hypothalamus have quite low firing rates (Fig. 2A). When distributions were plotted separately for estrogen-treated and untreated preparations in each anatomical structure, corrected for numbers of electrode penetrations through each structure, differences between
u~
A
40
so 2O
g
I0
o
ob o~ o~ o~ o~ c~ o
0
r~ u5
Resting discharge (spikes/sec.)
B.
Estradiol Irooted
8 Fq.
o--o Ovariex untreotecl
x
o L= ~
~ g
6 I
rate
~\
"E/%, 2/.:, , ~
NST
,
.,%
2
$ u~
9
MAHA
6V 18 J3 E Z
12 6
%, .
o
Q
BM ~,
o~ ~
o~ c~ c~ ~
Resting discharge (spikes/see.)
Q
rate
Fig. 2. Distributions of resting discharge rates. A: distribution for all hypothalamic and preoptic neurons recorded in all rats. B: distributions for each anatomical structure (abbreviations and definitions as in Fig. 1), plotted separately for estradiol-treated and untreated ovariectomized female rats. Numbers of neurons in each bin normalized according to the number of electrode penetrations through each anatomical structure. Differences between estradiol-treated and untreated female rat recording samples: * P - < 0.05; ** P - < 0,01.
73 hormone-treated and untreated groups appeared. In the nucleus of the stria terminalis and the medial preoptic area, there were significantly fewer neurons in the lowest firing rate category for estrogen-treated than for untreated preparations (Fig. 2B). Since this difference was not accompanied by the appearance of the same number of cells in higher resting discharge rate categories, estrogen must have suppressed the firing rate of these cells to rates so slow that they were not recorded with our sampling technique. The opposite occurred in the basomedial hypothalamus. Here there were significantly more cells in the lowest resting discharge rate categories for the estrogentreated preparations (Fig. 2B). Since this difference was not accompanied by a disappearance of an equal number of cells in higher resting discharge rate categories, it must have resulted from a facilitation by estrogen of very slow cells, bringing them into the recording distribution. To get an indication of the total neuronal output of each of the anatomical structures in hormone-treated and untreated preparations, the total number of cells recorded per penetration in each case was multiplied by the mean resting discharge rate. From this parameter, it would appear that the total number of spike discharges from the nucleus of the stria terminalis and the medial preoptic area would be less in estrogen-treated than untreated preparations, while in the medial anterior hypothalamus and basomedial hypothalamus there is a tendency in the opposite direction (Fig. IC). From the full distribution of firing rates (Fig. 2B) it appears that the most important effects of estrogen are on the slowest firing cells, and that the direction of the estrogen effect (after systemic injection) is different in the nucleus of the stria terminalis and medial preoptic area than in the basomedial hypothalamus. Responses to somatosensory stimuli The percentages of units responding to somatosensory stimuli were low (Fig. 3). There was a tendency for the responsiveness of neurons to be lower in more ventral and medial positions of the preoptic area and hypothalamus, but this trend was not statistically significant. Ninety-one per cent of the responses to somatosensory stimuli were excitations, and there were no significant differences between estrogentreated and untreated preparations or between anatomical structures in this regard. It was striking that even within individual preparations, neurons nearby each other would not be uniform in their tendency to respond: responsive cells were not all grouped together within an anatomical zone, and rather could be found very nearby (less than 200 #m distance) units which did not respond to any stimuli. Neuronal responsiveness to somatosensory stimuli related to the triggering of lordosis was different between estrogen-treated and untreated preparations, and the nature of the difference depended upon anatomical location. In the nucleus of the stria terminalis and the medial preoptic area there were significantly fewer units responding to somatosensory stimuli in estrogen-treated than in untreated preparations (Fig. 3). In the medial anterior hypothalamus and in the basomedial hypothalamus those differences in responsivity which were statistically significant were in the opposite direction: estrogen-treated preparations tended to have a greater number of responsive units (Fig. 3).
74 Stimuli" A. F l a n k s
B. R u m p
C. " F o r k "
35
=
35
o
E 0~
I MPOA
Oo
~
~
o
o
MPOA
of-] O
x..
D . A . or B. or C.
35
oil ol;
x~
o> w
NST
MPOA
0
.,O3
POA
o
O
O
MAHA
** MAHA
A
MAHA
"3 0
O ~
0
o BM
0
0 o
BM
0
~
o BM
0
8M
0
Fig. 3. Percentage of neurons responding to somatosensorystimuli. Plotted for each neuroanatomical structure (definitions and abbreviations as in Fig. 1). 'Fork' stimuli (C) were brief applications of pressure on the skin of the perineum, tailbase and posterior rump (see Methods). Part (D) shows percentages of neurons which would respond to at least one of the somatosensory stimuli, A, B, or C. Differences between responsiveness of neurons in estradiol-treated (EB) and untreated (OVX) female rat recording samples: * P < 0.05; ** P < 0.01.
Few units responsed to painful pinches of the ear or paw, less than 35 % in any anatomical zone of untreated or estrogen-treated preparations. All responses to painful stimuli were excitatory. It was of interest to determine the pattern of responses to pain and to the 'fork' somatosensory stimulus (on the rump, tailbase and perineum), among units tested with both types of stimuli. Most units (65 %) responded to neither, and 13 % responded only to pain. Very few units responded specifically to the 'fork' somatosensory stimulus, which induces lordosis. However, in the untreated preparations 3 units in the nucleus of the stria terminalis did so, and in estrogen-treated preparations 3 units in the medial anterior hypothalamus and 3 units in the basomedial hypothalamus responded only to this stimulus. Relation to cortical EEG Under these conditions of recording, with light urethane anesthesia, no cells showed statistically significant correlations between levels of resting discharge and the
75 state (synchrony vs. desynchrony) of the cortical EEG. Likewise, the EEG did not show rapid responses to somatosensory stimuli which were correlated with unit responses, with the exception of responses to pain. In 23 ~o of the cases in which a unit (without respect to location or hormonal state) responded to pain, the EEG changed from a synchronized to a desnychronized state. DISCUSSION
In comparing single unit recordings from ovariectomized female rats given either long-term estrogen treatment or not, differences between hormone-treated and untreated groups were not the same in all anatomical structures. In the medial preoptic area and nucleus of the stria terminalis estrogen treatment was followed by a decline in cells with recordable spontaneous activity, primarily because of a loss of cells with the slowest firing rates, and an overall decline in responsiveness to somatosensory stimuli. In the basomedial hypothalamus, and to some extent in the medial anterior hypothalamus, long-term estrogen treatment was followed by an increase in the number of cells recorded, primarily because of an increase in the number of cells with lowest resting discharge rates, and by some increase in responsiveness to somatosensory stimuli. A re-examination of the literature showed that comparisons of several previous studies also indicate that systemically injected estrogen is not necessarily followed by the same physiological effects in the preoptic area as it is in the basomedial hypothalamus and anterior hypothalamus. Kawakami e t al. 12 showed increases in firing rates of arcuate neurons following estrogen treatment, and Cross and Dyer 4 found increases in the medial anterior hypothalamus. In spayed female cats, responsiveness in the ventromedial nucleus and medial anterior hypothalamus was greater following estrogen treatment, and there was a shift in direction of response towards excitation 1. However, in the medial preoptic area Whitehead and Ruf 36 and Yagi 3v,3s found units which decreased their resting discharge rates to very low levels for a long time after estrogen administration. Lincoln TM also had shown that longer term estrogen treatment of ovariectomized rats was followed by lower spontaneous activity of preoptic units, of units at the very anterior border of the medial anterior hypothalamus, and in the lateral septum. Responses of preoptic neurons to probing of the vaginal cervix were shifted significantly in the direction of inhibition, following estrogen treatment iv. Radioactive estrogen is accumulated by neurons in all of the regions where recordings were made 26. It should be noted that neurochemical effects of estrogen are not the same in the preoptic area as they are in the basomedial hypothalamus (reviewed in ref. 23). The configuration of electrophysiological results in different preoptic and hypothalamic regions can be compared with effects of manipulation of the same tissue on the lordosis response to estrogen in ovariectomized female rats. Lesions restricted to the preoptic region enhance lordosis responding 15,3°, while lesions of the medial anterior hypothalamus or ventromedial region disrupt the lordosis response in estrogen-treated femalesV,lS, 33. In the preoptic region electrical stimulation has the opposite effect from lesionslS, 20. Thus, differences in the pattern of electrophysiological effects of estrogen between preoptic and hypothalamic structures
76 may be mirrored by differences in the physiological control of at least one estrogendependent function, lordosis in the female rat. A striking result was the heterogeneity of physiological properties of cells recorded nearby to each other. Two cells recorded within 100 or 200 #m of each other might have markedly different resting discharge rates, and one might respond to somatosensory stimulation while the other would not respond. In previous experiments with male rats ~'~' it was found that adjacent hypothalamic cells could have markedly different relationships between unit activity and the state of the cortical EEG. These findings underline the need for further physiological identification and characterization of hypothalamic units recorded. Much work by Cross and his collaborators has followed this strategy for studying neurons of the paraventricutar nucleus, and smaller numbers of studies have begun to elucidate properties of" preoptic and anterior hypothalamic neurons which can be antidromically stimulated through electrodes in the arcuate region TM. More neuroanatomical information on the projections of preoptic and anterior hypothalamic neurons would be of great help in furthering such physiological work, but stains for degenerating fibers have serious technical limitations in this regard. Projections of preoptic and anterior hypothalamic neurons as demonstrated by autoradiography following local injections of radioactive amino acids z will enlarge the list of places where electrodes for antidromic stimulation might be placed, to further characterize the cells in question. The fact that many units in this study had low rates of resting discharge confirms previousl2y published distributions of hypothalamic neuron resting discharge rates 9,~G. ~9,39. It was of interest that most preoptic and hypothalamic neurons did not respond to somatosensory stimuli which are involved in triggering lordosis in female rats. This might mean that these neurons do not participate in the control of lordosis by rapid and large responses to appropriate cutaneous stimuli. One concern, however', is that the effect of anesthesia might decrease responsiveness. There is evidence from previous experiments that urethane, as used here, does not have a massive depressing effect on hypothalamic cells. Recording in hypothalamic islands, Cross and Dyer 3 did not see large depressive effects of urethane on restirtg discharge rates. Regarding responsiveness of neurons to peripheral stimuli, Pfaff and Gregory ~4 recorded from rats anesthetized with 1.5 g/kg urethane (a higher dose than in the present study) and found responses to olfactory input in 50-90 o//,, of preoptic neurons, depending on the odor used. These findings indicate that low percentages of responsive neurons in the hypothalamus and preoptic area need not be ascribed solely to effects of urethane anesthesia. The means by which the net result of long-term estrogen treatment might be different when recording in the medial preoptic area compared to the basomedial hypothalamus are presently unknown. It should be emphasized, however, that since estradiol was administered systemically, recordings from preoptic and hypothalamic neurons might reflect a combination of direct and indirect estrogen effects. Thus, differences between preoptic area and basomedial hypothalamus need not imply differences in the mechanism of estradiol action on those very neurons, but might rather reflect differences in the impact of estrogen acting on neurons in other forebrain structures, reflected in afferent input to the cells recorded in these experiments.
77 REFERENCES
I ALCARAZ,M., GUZMAN-FLORES,C., SALAS,M., AND BEYER,C., Effect of estrogen on the responsivity of hypothalamic and mesencephalic neurons in the female cat, Brain Research, 15 (1969) 439 446. 2 CONRAD,L. A., AND PEAEF, D. W., Autoradiographic tracing of projections from preoptic area and anterior hypothalamus in the rat, Proe. Soc. Neurosei., (1974) 176, abstract 136. 3 CROSS, B. A., AND DYER, R. G., Cyclic changes in neurons of the anterior hypothalamus during the rat estrous cycle and the effect of anesthesia. In C. H. SAWYER AND R. A. GORSKI (Eds.), Steroid Hormones and Brain Function, UCLA Forum Or Medial Sciences, No. 15, Univ. of California Press, Los Angeles, Calif., 1971, pp. 95-102. 4 CROSS, B. A., AND DYER, R. G., Ovarian modulation of unit activity in the anterior hypothalamus of the cyclic rat, J. Physiol. (Lond.), 222 (1972) 25P. 5 DAVIDSON, J. M., RODGERS, C. H., SMITH, E. R., AND BLOCH, G. J., Stimulation of female sex behavior in adrenalectomized rats with estrogen alone, Endocrinology, 82 (1968) 193-195. 6 DIAKOW, C., PFAFF, D. W., AND KOMlSARUK, B., Sensory and hormonal interactions in eliciting lordosis, Fed. Proe., 32 (1973) 241 (abstract). 7 DORNER, G., DOCKE, F., AND HINZ, G., Homo- and hypersexuality in rats with hypothalamic lesions, Neuroendocrinology, 4 (1969) 2 ~ 2 4 . 8 DYER, R. G., An electrophysiological dissection of the hypothalamic regions which regulate the pre-ovulatory secretion of luteinizing hormone in the rat, J. Physiol. (Lond.), 234 (1973) 421-442. 9 DYER, R. G., PRITCHETT,C. J., AND CROSS, B, A., Unit activity in the diencephalon of female rats during the oestrous cycle, J. Endoer., 53 (1972) 151-160. l0 FREUND, J. E., Modern Elementary Statistics, Prentice-Hall, Inc., Englewood Cliffs, N.J., 1952. 11 KAWAKAMI,M., TERASAWA, E., AND IBUKI, T., Changes in multiple unit activity of the brain during the estrous cycle, Neuroendocrinology, 6 (1970) 30-48. 12 KAWAKAMI,M., TERASAWA,E., IBUKI, T., AND MANAKA, M., Effects of sex hormones and ovulation-b'ocking steroids and drugs on electrical activity of the rat brain. In C. H. SAWYERAND R. A. GORSKI (Eds.), Steroid Hormones and Brain Function, UCLA Forum in Medical Sciences, No. 15, Univ. of California Press, Los Angeles, Calif., 1971, pp. 79-93. 13 KONIG, J. F. R., AND KLIPPEL, R. A., The Rat Brain. A Stereotaxie Atlas o f the Forebrain and Lower Parts o f the Brain Stem, Williams and Wilkins, Baltimore, Md., 1963. 14 Kow, L.-M., AND PFAVV, D. W., Sensory requirements for the lordosis reflex in female rats, Brain Research, 101 (1976)47 66 15 LAW, T., AND MEAGHER, W., Hypothalamic lesions and sexual behavior in the female rat, Science, 128 (1958) 1626-1627. 16 LINCOLN, D. W., Unit activity in the hypothalamus, septum and preoptic area of the rat : characteristics of spontaneous activity and the effect of oestrogen, J. Endocr., 37 (1967) 177-189. 17 LINCOLN, D. W., AND CROSS, B. A., Effect of oestrogen on the responsiveness of neurones in the hypothalamus, septum and preoptic area of rats with light-induced persistent oestrus, J. Emtoer., 37 (1967) 191-203. 18 MALSBURY,C., AND PFA~F, D. W., Suppression of sexual receptivity in the hormone-primed female hamster by electrical stimulation of the medial preoptic area, Proe. Soe. Neurosei., (1973) 122 (abstract.) 19 Moss, R. L., AND LAW, O. T., The estrous cycle: its influence on single unit activity in the forebrain, Brain Research, 30 (1971) 435-438. 20 Moss, R. L., PALOUTZiAN, R. E., AND LAW, O. T., Electrical stimulation of forebrain structures and its effect on copulatory as well as stimulus-bound behavior in ovariectomized hormoneprimed rats, Physiol. Behav., 12 (1974) 997-1004. 21 PFAVF, D. W., Nature of sex hormone effects on rat sex behavior: specificity of effects and individual patterns of response, J. comp. physiol. Psychol., 73 (1970) 349 358. 22 PVAFF, D. W., Interactions of steroid sex hormones with brain tissue : studies of uptake and physiological effects. In S. SEGAL et al. (Eds.), The Regulation o f Mammalian Reproduction, Thomas, Springfield, |11., 1973, pp. 5-22. 23 PFAFF, D. W., Effects of steroid hormones on electrical activity of neurons. In N. ADLER (Ed.), Primer o[" Neuroendocrine Integration and Behavior, Plenum, New York, in press. 24 PFAEF, D. W., AND GREGORY, E., Olfactory coding in olfactory bulb and medial forebrain bundle of normal and castrated male rats, J. Neurophysiol., 34 (1971) 208-216. 25 PFAEV, D. W., AND GREGORY, E., Correlation between preoptic area unit activity and the cortical
78
26 27 28 29
30 31 32 33 34 35 36 37 38 39
EEG: difference between normal and castrated male rats, Eleetroenceph. elm. Ncurophv+i,L. ~ (19711 223-230. PFAPF, D. W., AND KEtNER, M., Atlas ofestradiol-concentrating cells in the central nervous s~,stem of the female rat, J. eomp. Neurol., 151 (1973) 121-158. PFAFV, D. W., AND I_+EWlS, C., Film analyses of lordosis in female rats, Horm. Behav., 5 ( t',v741 317-335. PFAFF, D. W., AND PFAFFMANN, C., Olfactory and hormonal influences on the basal forebrain of the male rat, Brain Research, 15 (19691 137-156. PVAPF, D. W., DIAKOW, C., Z~GMOND, R. E.+ AND KOW, L.-M., Neural and hormonal determinants of female mating behavior in rats. In F. O. SCHMtTT AND F. G. WORDEN (Eds+), The Neuroscienees, Vol. IlL M.[.T. Press, Cambridge, Mass., 1973, pp. 621-646. POWERS, B., AND VALENSTEIN, E. S., Sexual receptivity: facilitation by medial preoptic lesions in female rats, Science, 175 (1972) 1003-1005. RuF, K., AND S'rEINER, F. A., Steroid-sensitive single neurons in rat hypothalamus and midbrain: identification by microelectrophoresis, Science, 156 (1967) 667--669. SIEGEL, S., Nonparametric Statistiesfor the Behavioral Sciences, McGraw-Hill, New York, 1956. SINC;ER+J. J., Hypothalamic control of male and female sexual behavior in female rats, J. comp. physiol. Psyehol., 66 (1968) 738-742. STEJNER, F. A., RUF, K., AND AKERT, K., Steroid-sensitive neurones in rat brain: anatomical localization and responses to neurohumours and ACTH, Brain Research, 12 (1969) 74-85. TERASAWA, E., AND SAWYER, C. H., Changes in electrical activity in the rat hypothalamus related to electrochemical stimulation of adenohypophyseal function, Endocrinology, 85 (1969) 143+149. WmTEHEAD, S. A., AND RUV, K. B., Responses of antidromically identified preoptic neurons in the rat to neurotransmitters and to estrogen, Brain Research, 79 (1974) 185-198. YAGI, K., Effects of estrogen on the unit activity of the rat hypothalamus, J. Physiol. Soc. Japan, 32 (1970) 692-693. YAOI, K., Changes in firing rates of single preoptic and hypothalamic units following an intravenous administration of estrogen in the castrated female rat, Brain Research, 53 (1973) 343-352. YAGI, K., AND SAWAKI, Y., Feedback of estrogen in the hypothalamic control of gonadotrophin secretion. In K. YAGI AND S. YOSHIDA (Eds.), Neuroendocrine Control, Univ. of Tokyo Press, Tokyo, 1973, pp. 297-325.
67
S I N G L E U N I T R E C O R D I N G IN H Y P O T H A L A M U S AND PREOPTIC AREA OF E S T R O G E N - T R E A T E D A N D U N T R E A T E D O V A R I E C T O M I Z E D F E M A L E RATS
JOSE BUENO AND DONALD W. P F A F F
Rockefeller University, New York, N.Y. 10021 (U.S.A.) (Accepted June 30th, 1975)
SUMMARY
Single unit activity was recorded with micropipettes in the medial hypothalamus and preoptic area of urethane-anesthetized ovariectomized female rats. Some females had received long-term estradiol treatment, while others had been left untreated. In the medial preoptic region and bed nucleus of the stria terminalis, estrogen-treated rats had fewer cells (compared to untreated rats) with recordable spontaneous activity, due primarily to a loss of cells with very slow firing rates. In the basomedial hypothalamus, estrogen-treated rats had more cells (than untreated rats) with recordable spontaneous activity, due primarily to an increase in the number of cells with slow firing rates. Responsiveness of neurons to somatosensory stimulation was generally low. If present it was depressed by estrogen treatment in medial preoptic area and bed nucleus of stria terminalis, while it tended to be elevated by estrogen treatment in medial anterior hypothalamus and basomedial hypothalamus. Differences in the effects of long-term systemic estrogen treatment on medial preoptic neurons compared to basomedial hypothalamus are paralleled by differences in the control of lordosis by these neurons in female rats.
INTRODUCTION
Microelectrode recording has allowed the demonstration of steroid hormone effects on hypothalamic electrophysiology. In some experiments local applications of steroids have been used to prove that the hormone effects are direct. Steiner et al. 31,34 showed effects of a synthetic glucocorticoid on hypotbalamic units, using iontophoretic application of the steroid. Testosterone can affect neurons in the preoptic area following either local or systemic application zs. Regarding ovarian hormones, several experiments have shown variations in
68 hypothalamic unit activity during the estrous cycle of the female rat3-~-:', ~l.~v~.~ These effects have been reviewed2~,mL However, during the estrous cycle blood levels of several hormones are changing simultaneously, making a complicated situation for the analysis of any individual hormone effect. The locations of estradiol-concentrating neurons in the hypothalamus, preoptic area and basal tbrebrain are known preciselye~L Short latency effects of intravenously injected estrogens have been recorded in antidromically identified preoptic neurons :~6. Since estrogens are never absent from the blood of the normally cycling female rat, one physiologically relevant comparison would seem to be between two groups of ovariectomized female rats: one group untreated and the other given daily injections of estradiol. Microelectrode recording for isolating single units is required, since it is known that hypothalamic neurons close to each other can have different physiological properties2'~, zS. For the present recording study, focussing on the mechanisms underlying the effects of estrogen on lordosis, progesterone injections were not used because ovariectomized female rats given longterm estrogen treatment will perform lordosis without progesteroner', ~. Special care was exerted in choosing peripheral stimuli for use during recording experiments. In order to focus on hypothalamic mechanisms of female rat reproductive behavior, we knew that somatosensory stimulation could be emphasized, because only somatosensory input is required for female rat lordosis ~4. Cutaneous stimuli on the female's flanks, rump and perineal regions were chosen to imitate stimulation provided by the male rat 27. The cutaneous stimuli used are known to be necessary 14 and sufficient6, '~9 for lordosis.
METHODS
Rats Recording was done on 42 female Sprague-Dawley rats, obtained from Hormone Assay Laboratories (Chicago, Ill.). All rats were ovariectomized at least 3 weeks prior to recording, and weighed between 250 and 350 g. Twenty females received daily injections o f 10/zg estradiol benzoate (subcutaneously in oil) per day, for at least 10 days prior to recording. This dosage was chosen to allow lordosis without the necessity of progesterone injections, and all 20 rats performed strong lordosis reflexes immediately before recording experiments. The other 22 rats did not receive estradiol, and none of these animals showed lordosis. Recording experiments were done without knowledge of whether or not a particular rat had been treated with estradiol.
Recording preparation Rats were anesthetized with urethane (1.1 g/kg, intraperitoneally), This low dose of urethane was chosen to help insure that hypothalamic units would be as active and responsive as possible. The head of the rat was stabilized with a stereotaxic instrument; To facilitate recording hypothalamic neurons close to the midline, where many estradiol-concentrating
69 cells can be found 26, the head was held at a left-right angle of 9 ° to 14° offthe horizontal plane. Body temperature was maintained by a heating pad throughout the experiment in the range of 36.5-37 °C. Rectal temperature was monitored only periodically, to avoid unnecessary somatosensory stimulation from the rear half of the body. After incision and retraction of the skin, the skull was thinned with a dental drill. The deepest layer of bone and the dura was then removed carefully with fine forceps. The cortical electroencephalogram (EEG) was recorded through two silver ball electrodes placed respectively in small holes through the frontal and parietal bones on the side opposite the opening for the micropipette. EEG potentials were amplified and displayed using a Grass Model 7 polygraph. During recording the trunk of the rat was suspended at the thorax and lower abdomen with two rubber bands to minimize inadvertent tactile stimulation. Single units were recorded with glass micropipettes filled with 2.5 M sodium chloride. Tip diameters were between 1 and 3 /~m, and DC resistances between 0.5 and 5 M ~ . The reference electrode was a silver ball electrode on the cortex. The micropipette was positioned using a micromanipulator (La Precision Cinematographique), and penetrations were made in the plane of the stereotaxic atlas of Kenig and Klippep 3. Input from the micropipette was led to a high input impedance preamplifier and then to a Tektronix 502 oscilloscope and an audio monitor. Frequencies below 400 and above 10,000 cycles/sec were filtered out. Each penetration was made with slow and systematic tracking. When the discharges of a single unit were isolated by small micropipette movements, they were quantitatively evaluated using a counting circuit equipped with a Schmitt trigger and a histogram generating circuit. At the end of each second this circuit puts out a voltage proportional to the number of spike discharges. The output of the counting circuit was displayed on-line as a histogram on one channel of the Grass Model 7 polygraph. This method provides a quantitative record for the evaluation of spontaneous activity, responses to stimulation, and relation to cortical EEG. Stimulation Somatosensory stimuli were chosen to test hypothalamic unit responses to input related to the lordosis reflex. The stimuli are designed to imitate those provided by the male rat during mating zv, and are known to be necessary 14 and sufficient6,29 for lordosis. Flank stimuli were composed of fast, repetitive bilateral stroking by the experimenter's thumb and index finger on the flanks, just anterior to the rear legs. Rump stimulation was gentle, repetitive palpation on the posterior rump and tailbase area. 'Fork' stimuli were brief applications of pressure on the perineal, tailbase and posterior rump areas by the experimenter's thumb, with the index and middle fingers 'forked' around the posterior end of the body on either side of the tailbase. This stimulus readily elicits the lordosis reflex in receptive female rats. Pain stimuli were intense pinches of the ear or paw. The precise time and duration of stimulus application was recorded with the signal marker on the polygraph, to be compared with possible alterations in unit activity or cortical EEG.
70
Data analysis For each cell recorded, average resting discharge rate was calculated from the histogram. The time for which average spontaneous discharge was calculated included the period long enough after the cell was isolated that stability of rate could be insured, and before any stimuli were applied. The fact that only a single unit was contributing to the record at any one time was insured during the experiment by monitoring the performance of the Schmitt trigger on the oscilloscope as well as by attending to the audio monitor. A response to stimulation was defined as a statistically significant change in firing rate, when the activity during stimulation was compared with 30 sec of immediately prior spontaneous discharge. Changes of firing rate of less than 30 o/~,were never counted as responses. Histograms of unit firing rates were compared to simultaneously recorded cortical EEG. Average firing rates during EEG synchrony were compared with those during EEG desynchrony, whenever both cortical E E G states were expressed during the time that unit was held. Also, possible changes in cortical EEG during stimulation were noted.
Histology The bottom of each micropipette track, at or near the bottom of the brain, was marked by a small electrolytic lesion. After each experiment the rat was killed by cardiac perfusion with 10 ~ formalin. Serial frozen sections, 100 # m thick, were cut in the plane of the atlas of Krnig and KlippeP 3, and then stained with cresyl violet. Using a microprojector the position of each electrode penetration was found and the location of each recorded neuron plotted on a drawing o f the appropriate section. Histology from preparations in which no single units were recorded was not considered, due to possible difficulties with electrodes, anesthesia, etc. The number of penetrations going through each anatomical structure was counted, and used for calculations in the results (i.e., some of the results were expressed on the normalized basis of 10 penetrations/brain structure). All anatomical terms and definitions of zones follow exactly the nomenclature of the KOnig and KlippeP 3 atlas o f the rat brain. We have combined unit samples from the arcuate nucleus, ventromedial hypothalamic nucleus and dorsomedial hypothalamic nucleus and the tissue between those nuclei, as defined by K r n i g and Klippel, to constitute a sample labeled basomedial hypothalamus (BM). A total of 328 single units were recorded in the anatomical'zones considered. For neurons in each anatomical zone, statistical comparisons between estrogentreated and untreated females were made. Differences in numbers of cells/penetration and in average resting discharge rates were evaluated parametrically using the t-test t° and non-parametrically using the Mann-Whitney U-tesOL Differences in proportions of cells responding to each stimulus were evaluated with a statistical test for differences between proportions 1°.
A.
,ol
71
B.
NST
E
~ ST
C. ( = A x B )
~-0
NST
(D 0 4-
+1 ~
0
~-
0
0
c0
C~
MPOA
POA
~) W
0
o3
c
2
~ -a o
-(~ (1)
E x o
0
o (D ~
-~
~
0
cn v
~
o
o
--
r"
09
0
50
0
__~ o
50
l/)
. . (D
.
.
o3
0
~.~ 0
c-
~,
50
×'~ 0
O3
E
"S_ c 0
x >
m w
~
-6 0
x o
~
I-=
50
x >
m w
Fig. 1. Comparison of single unit recordings from estradiol-treated (EB) and untreated (OVX) ovariectomized female rats. Numbers of neurons with recordable spontaneous activity (A) normalized according to number of electrode penetrations through each anatomical structure. Figure parts (B) and (C) are explained in text. Abbreviations: NST, bed nucleus of stria terminalis; MPOA, medial preoptic area, M A H A , medial anterior hypothalamus, BM, basomedial hypothalamus (combination of arcuate, ventromedial and dorsomedial nucleus recording sites). Definitions according to atlas of K6nig and Klippe113. Differences between estrogen-treated and untreated ovariectomized female rats: ** P < 0.01. RESULTS
Resting discharge rates Differences between unit recordings in estrogen-treated and untreated preparations were not the same in nucleus of the stria terminalis and medial preoptic area as in the basomedial hypothalamus. The numbers of cells/penetration with recordable spontaneous activity were greater for untreated than for estrogen-treated preparations in the nucleus of the stria terminalis and the medial preoptic area, while in the basomedial hypothalamus the opposite was the case (Fig. 1A). Among the cells recorded, when resting discharge rates were averaged for all ceils, without respect to the underlying firing rate distribution, there were no signifi-
72 cant differences between estrogen-treated and untreated preparations for any anatomical structure (Fig. I B). This is probably due to two factors. First, it was striking that even within the same preparation, units nearby each other (less than 200 Fm distance between them) could have very different resting discharge rates. Second, averaging spontaneous activity across units is a complicated calculation: such an average ca~t be increased either by a general elevation of firing rates or by the disappearance from the distribution of units with very low firing rates. To account for these two factors separately, distributions of unit liring rates were plotted (Fig. 2A). In agreement with previous results9,~6,1°, :~9the distribution shows that many cells in the hypothalamus have quite low firing rates (Fig. 2A). When distributions were plotted separately for estrogen-treated and untreated preparations in each anatomical structure, corrected for numbers of electrode penetrations through each structure, differences between
u~
A
40
so 2O
g
I0
o
ob o~ o~ o~ o~ c~ o
0
r~ u5
Resting discharge (spikes/sec.)
B.
Estradiol Irooted
8 Fq.
o--o Ovariex untreotecl
x
o L= ~
~ g
6 I
rate
~\
"E/%, 2/.:, , ~
NST
,
.,%
2
$ u~
9
MAHA
6V 18 J3 E Z
12 6
%, .
o
Q
BM ~,
o~ ~
o~ c~ c~ ~
Resting discharge (spikes/see.)
Q
rate
Fig. 2. Distributions of resting discharge rates. A: distribution for all hypothalamic and preoptic neurons recorded in all rats. B: distributions for each anatomical structure (abbreviations and definitions as in Fig. 1), plotted separately for estradiol-treated and untreated ovariectomized female rats. Numbers of neurons in each bin normalized according to the number of electrode penetrations through each anatomical structure. Differences between estradiol-treated and untreated female rat recording samples: * P - < 0.05; ** P - < 0,01.
73 hormone-treated and untreated groups appeared. In the nucleus of the stria terminalis and the medial preoptic area, there were significantly fewer neurons in the lowest firing rate category for estrogen-treated than for untreated preparations (Fig. 2B). Since this difference was not accompanied by the appearance of the same number of cells in higher resting discharge rate categories, estrogen must have suppressed the firing rate of these cells to rates so slow that they were not recorded with our sampling technique. The opposite occurred in the basomedial hypothalamus. Here there were significantly more cells in the lowest resting discharge rate categories for the estrogentreated preparations (Fig. 2B). Since this difference was not accompanied by a disappearance of an equal number of cells in higher resting discharge rate categories, it must have resulted from a facilitation by estrogen of very slow cells, bringing them into the recording distribution. To get an indication of the total neuronal output of each of the anatomical structures in hormone-treated and untreated preparations, the total number of cells recorded per penetration in each case was multiplied by the mean resting discharge rate. From this parameter, it would appear that the total number of spike discharges from the nucleus of the stria terminalis and the medial preoptic area would be less in estrogen-treated than untreated preparations, while in the medial anterior hypothalamus and basomedial hypothalamus there is a tendency in the opposite direction (Fig. IC). From the full distribution of firing rates (Fig. 2B) it appears that the most important effects of estrogen are on the slowest firing cells, and that the direction of the estrogen effect (after systemic injection) is different in the nucleus of the stria terminalis and medial preoptic area than in the basomedial hypothalamus. Responses to somatosensory stimuli The percentages of units responding to somatosensory stimuli were low (Fig. 3). There was a tendency for the responsiveness of neurons to be lower in more ventral and medial positions of the preoptic area and hypothalamus, but this trend was not statistically significant. Ninety-one per cent of the responses to somatosensory stimuli were excitations, and there were no significant differences between estrogentreated and untreated preparations or between anatomical structures in this regard. It was striking that even within individual preparations, neurons nearby each other would not be uniform in their tendency to respond: responsive cells were not all grouped together within an anatomical zone, and rather could be found very nearby (less than 200 #m distance) units which did not respond to any stimuli. Neuronal responsiveness to somatosensory stimuli related to the triggering of lordosis was different between estrogen-treated and untreated preparations, and the nature of the difference depended upon anatomical location. In the nucleus of the stria terminalis and the medial preoptic area there were significantly fewer units responding to somatosensory stimuli in estrogen-treated than in untreated preparations (Fig. 3). In the medial anterior hypothalamus and in the basomedial hypothalamus those differences in responsivity which were statistically significant were in the opposite direction: estrogen-treated preparations tended to have a greater number of responsive units (Fig. 3).
74 Stimuli" A. F l a n k s
B. R u m p
C. " F o r k "
35
=
35
o
E 0~
I MPOA
Oo
~
~
o
o
MPOA
of-] O
x..
D . A . or B. or C.
35
oil ol;
x~
o> w
NST
MPOA
0
.,O3
POA
o
O
O
MAHA
** MAHA
A
MAHA
"3 0
O ~
0
o BM
0
0 o
BM
0
~
o BM
0
8M
0
Fig. 3. Percentage of neurons responding to somatosensorystimuli. Plotted for each neuroanatomical structure (definitions and abbreviations as in Fig. 1). 'Fork' stimuli (C) were brief applications of pressure on the skin of the perineum, tailbase and posterior rump (see Methods). Part (D) shows percentages of neurons which would respond to at least one of the somatosensory stimuli, A, B, or C. Differences between responsiveness of neurons in estradiol-treated (EB) and untreated (OVX) female rat recording samples: * P < 0.05; ** P < 0.01.
Few units responsed to painful pinches of the ear or paw, less than 35 % in any anatomical zone of untreated or estrogen-treated preparations. All responses to painful stimuli were excitatory. It was of interest to determine the pattern of responses to pain and to the 'fork' somatosensory stimulus (on the rump, tailbase and perineum), among units tested with both types of stimuli. Most units (65 %) responded to neither, and 13 % responded only to pain. Very few units responded specifically to the 'fork' somatosensory stimulus, which induces lordosis. However, in the untreated preparations 3 units in the nucleus of the stria terminalis did so, and in estrogen-treated preparations 3 units in the medial anterior hypothalamus and 3 units in the basomedial hypothalamus responded only to this stimulus. Relation to cortical EEG Under these conditions of recording, with light urethane anesthesia, no cells showed statistically significant correlations between levels of resting discharge and the
75 state (synchrony vs. desynchrony) of the cortical EEG. Likewise, the EEG did not show rapid responses to somatosensory stimuli which were correlated with unit responses, with the exception of responses to pain. In 23 ~o of the cases in which a unit (without respect to location or hormonal state) responded to pain, the EEG changed from a synchronized to a desnychronized state. DISCUSSION
In comparing single unit recordings from ovariectomized female rats given either long-term estrogen treatment or not, differences between hormone-treated and untreated groups were not the same in all anatomical structures. In the medial preoptic area and nucleus of the stria terminalis estrogen treatment was followed by a decline in cells with recordable spontaneous activity, primarily because of a loss of cells with the slowest firing rates, and an overall decline in responsiveness to somatosensory stimuli. In the basomedial hypothalamus, and to some extent in the medial anterior hypothalamus, long-term estrogen treatment was followed by an increase in the number of cells recorded, primarily because of an increase in the number of cells with lowest resting discharge rates, and by some increase in responsiveness to somatosensory stimuli. A re-examination of the literature showed that comparisons of several previous studies also indicate that systemically injected estrogen is not necessarily followed by the same physiological effects in the preoptic area as it is in the basomedial hypothalamus and anterior hypothalamus. Kawakami e t al. 12 showed increases in firing rates of arcuate neurons following estrogen treatment, and Cross and Dyer 4 found increases in the medial anterior hypothalamus. In spayed female cats, responsiveness in the ventromedial nucleus and medial anterior hypothalamus was greater following estrogen treatment, and there was a shift in direction of response towards excitation 1. However, in the medial preoptic area Whitehead and Ruf 36 and Yagi 3v,3s found units which decreased their resting discharge rates to very low levels for a long time after estrogen administration. Lincoln TM also had shown that longer term estrogen treatment of ovariectomized rats was followed by lower spontaneous activity of preoptic units, of units at the very anterior border of the medial anterior hypothalamus, and in the lateral septum. Responses of preoptic neurons to probing of the vaginal cervix were shifted significantly in the direction of inhibition, following estrogen treatment iv. Radioactive estrogen is accumulated by neurons in all of the regions where recordings were made 26. It should be noted that neurochemical effects of estrogen are not the same in the preoptic area as they are in the basomedial hypothalamus (reviewed in ref. 23). The configuration of electrophysiological results in different preoptic and hypothalamic regions can be compared with effects of manipulation of the same tissue on the lordosis response to estrogen in ovariectomized female rats. Lesions restricted to the preoptic region enhance lordosis responding 15,3°, while lesions of the medial anterior hypothalamus or ventromedial region disrupt the lordosis response in estrogen-treated femalesV,lS, 33. In the preoptic region electrical stimulation has the opposite effect from lesionslS, 20. Thus, differences in the pattern of electrophysiological effects of estrogen between preoptic and hypothalamic structures
76 may be mirrored by differences in the physiological control of at least one estrogendependent function, lordosis in the female rat. A striking result was the heterogeneity of physiological properties of cells recorded nearby to each other. Two cells recorded within 100 or 200 #m of each other might have markedly different resting discharge rates, and one might respond to somatosensory stimulation while the other would not respond. In previous experiments with male rats ~'~' it was found that adjacent hypothalamic cells could have markedly different relationships between unit activity and the state of the cortical EEG. These findings underline the need for further physiological identification and characterization of hypothalamic units recorded. Much work by Cross and his collaborators has followed this strategy for studying neurons of the paraventricutar nucleus, and smaller numbers of studies have begun to elucidate properties of" preoptic and anterior hypothalamic neurons which can be antidromically stimulated through electrodes in the arcuate region TM. More neuroanatomical information on the projections of preoptic and anterior hypothalamic neurons would be of great help in furthering such physiological work, but stains for degenerating fibers have serious technical limitations in this regard. Projections of preoptic and anterior hypothalamic neurons as demonstrated by autoradiography following local injections of radioactive amino acids z will enlarge the list of places where electrodes for antidromic stimulation might be placed, to further characterize the cells in question. The fact that many units in this study had low rates of resting discharge confirms previousl2y published distributions of hypothalamic neuron resting discharge rates 9,~G. ~9,39. It was of interest that most preoptic and hypothalamic neurons did not respond to somatosensory stimuli which are involved in triggering lordosis in female rats. This might mean that these neurons do not participate in the control of lordosis by rapid and large responses to appropriate cutaneous stimuli. One concern, however', is that the effect of anesthesia might decrease responsiveness. There is evidence from previous experiments that urethane, as used here, does not have a massive depressing effect on hypothalamic cells. Recording in hypothalamic islands, Cross and Dyer 3 did not see large depressive effects of urethane on restirtg discharge rates. Regarding responsiveness of neurons to peripheral stimuli, Pfaff and Gregory ~4 recorded from rats anesthetized with 1.5 g/kg urethane (a higher dose than in the present study) and found responses to olfactory input in 50-90 o//,, of preoptic neurons, depending on the odor used. These findings indicate that low percentages of responsive neurons in the hypothalamus and preoptic area need not be ascribed solely to effects of urethane anesthesia. The means by which the net result of long-term estrogen treatment might be different when recording in the medial preoptic area compared to the basomedial hypothalamus are presently unknown. It should be emphasized, however, that since estradiol was administered systemically, recordings from preoptic and hypothalamic neurons might reflect a combination of direct and indirect estrogen effects. Thus, differences between preoptic area and basomedial hypothalamus need not imply differences in the mechanism of estradiol action on those very neurons, but might rather reflect differences in the impact of estrogen acting on neurons in other forebrain structures, reflected in afferent input to the cells recorded in these experiments.
77 REFERENCES
I ALCARAZ,M., GUZMAN-FLORES,C., SALAS,M., AND BEYER,C., Effect of estrogen on the responsivity of hypothalamic and mesencephalic neurons in the female cat, Brain Research, 15 (1969) 439 446. 2 CONRAD,L. A., AND PEAEF, D. W., Autoradiographic tracing of projections from preoptic area and anterior hypothalamus in the rat, Proe. Soc. Neurosei., (1974) 176, abstract 136. 3 CROSS, B. A., AND DYER, R. G., Cyclic changes in neurons of the anterior hypothalamus during the rat estrous cycle and the effect of anesthesia. In C. H. SAWYER AND R. A. GORSKI (Eds.), Steroid Hormones and Brain Function, UCLA Forum Or Medial Sciences, No. 15, Univ. of California Press, Los Angeles, Calif., 1971, pp. 95-102. 4 CROSS, B. A., AND DYER, R. G., Ovarian modulation of unit activity in the anterior hypothalamus of the cyclic rat, J. Physiol. (Lond.), 222 (1972) 25P. 5 DAVIDSON, J. M., RODGERS, C. H., SMITH, E. R., AND BLOCH, G. J., Stimulation of female sex behavior in adrenalectomized rats with estrogen alone, Endocrinology, 82 (1968) 193-195. 6 DIAKOW, C., PFAFF, D. W., AND KOMlSARUK, B., Sensory and hormonal interactions in eliciting lordosis, Fed. Proe., 32 (1973) 241 (abstract). 7 DORNER, G., DOCKE, F., AND HINZ, G., Homo- and hypersexuality in rats with hypothalamic lesions, Neuroendocrinology, 4 (1969) 2 ~ 2 4 . 8 DYER, R. G., An electrophysiological dissection of the hypothalamic regions which regulate the pre-ovulatory secretion of luteinizing hormone in the rat, J. Physiol. (Lond.), 234 (1973) 421-442. 9 DYER, R. G., PRITCHETT,C. J., AND CROSS, B, A., Unit activity in the diencephalon of female rats during the oestrous cycle, J. Endoer., 53 (1972) 151-160. l0 FREUND, J. E., Modern Elementary Statistics, Prentice-Hall, Inc., Englewood Cliffs, N.J., 1952. 11 KAWAKAMI,M., TERASAWA, E., AND IBUKI, T., Changes in multiple unit activity of the brain during the estrous cycle, Neuroendocrinology, 6 (1970) 30-48. 12 KAWAKAMI,M., TERASAWA,E., IBUKI, T., AND MANAKA, M., Effects of sex hormones and ovulation-b'ocking steroids and drugs on electrical activity of the rat brain. In C. H. SAWYERAND R. A. GORSKI (Eds.), Steroid Hormones and Brain Function, UCLA Forum in Medical Sciences, No. 15, Univ. of California Press, Los Angeles, Calif., 1971, pp. 79-93. 13 KONIG, J. F. R., AND KLIPPEL, R. A., The Rat Brain. A Stereotaxie Atlas o f the Forebrain and Lower Parts o f the Brain Stem, Williams and Wilkins, Baltimore, Md., 1963. 14 Kow, L.-M., AND PFAVV, D. W., Sensory requirements for the lordosis reflex in female rats, Brain Research, 101 (1976)47 66 15 LAW, T., AND MEAGHER, W., Hypothalamic lesions and sexual behavior in the female rat, Science, 128 (1958) 1626-1627. 16 LINCOLN, D. W., Unit activity in the hypothalamus, septum and preoptic area of the rat : characteristics of spontaneous activity and the effect of oestrogen, J. Endocr., 37 (1967) 177-189. 17 LINCOLN, D. W., AND CROSS, B. A., Effect of oestrogen on the responsiveness of neurones in the hypothalamus, septum and preoptic area of rats with light-induced persistent oestrus, J. Emtoer., 37 (1967) 191-203. 18 MALSBURY,C., AND PFA~F, D. W., Suppression of sexual receptivity in the hormone-primed female hamster by electrical stimulation of the medial preoptic area, Proe. Soe. Neurosei., (1973) 122 (abstract.) 19 Moss, R. L., AND LAW, O. T., The estrous cycle: its influence on single unit activity in the forebrain, Brain Research, 30 (1971) 435-438. 20 Moss, R. L., PALOUTZiAN, R. E., AND LAW, O. T., Electrical stimulation of forebrain structures and its effect on copulatory as well as stimulus-bound behavior in ovariectomized hormoneprimed rats, Physiol. Behav., 12 (1974) 997-1004. 21 PFAVF, D. W., Nature of sex hormone effects on rat sex behavior: specificity of effects and individual patterns of response, J. comp. physiol. Psychol., 73 (1970) 349 358. 22 PVAFF, D. W., Interactions of steroid sex hormones with brain tissue : studies of uptake and physiological effects. In S. SEGAL et al. (Eds.), The Regulation o f Mammalian Reproduction, Thomas, Springfield, |11., 1973, pp. 5-22. 23 PFAFF, D. W., Effects of steroid hormones on electrical activity of neurons. In N. ADLER (Ed.), Primer o[" Neuroendocrine Integration and Behavior, Plenum, New York, in press. 24 PFAEF, D. W., AND GREGORY, E., Olfactory coding in olfactory bulb and medial forebrain bundle of normal and castrated male rats, J. Neurophysiol., 34 (1971) 208-216. 25 PFAEV, D. W., AND GREGORY, E., Correlation between preoptic area unit activity and the cortical
78
26 27 28 29
30 31 32 33 34 35 36 37 38 39
EEG: difference between normal and castrated male rats, Eleetroenceph. elm. Ncurophv+i,L. ~ (19711 223-230. PFAPF, D. W., AND KEtNER, M., Atlas ofestradiol-concentrating cells in the central nervous s~,stem of the female rat, J. eomp. Neurol., 151 (1973) 121-158. PFAFV, D. W., AND I_+EWlS, C., Film analyses of lordosis in female rats, Horm. Behav., 5 ( t',v741 317-335. PFAFF, D. W., AND PFAFFMANN, C., Olfactory and hormonal influences on the basal forebrain of the male rat, Brain Research, 15 (19691 137-156. PVAPF, D. W., DIAKOW, C., Z~GMOND, R. E.+ AND KOW, L.-M., Neural and hormonal determinants of female mating behavior in rats. In F. O. SCHMtTT AND F. G. WORDEN (Eds+), The Neuroscienees, Vol. IlL M.[.T. Press, Cambridge, Mass., 1973, pp. 621-646. POWERS, B., AND VALENSTEIN, E. S., Sexual receptivity: facilitation by medial preoptic lesions in female rats, Science, 175 (1972) 1003-1005. RuF, K., AND S'rEINER, F. A., Steroid-sensitive single neurons in rat hypothalamus and midbrain: identification by microelectrophoresis, Science, 156 (1967) 667--669. SIEGEL, S., Nonparametric Statistiesfor the Behavioral Sciences, McGraw-Hill, New York, 1956. SINC;ER+J. J., Hypothalamic control of male and female sexual behavior in female rats, J. comp. physiol. Psyehol., 66 (1968) 738-742. STEJNER, F. A., RUF, K., AND AKERT, K., Steroid-sensitive neurones in rat brain: anatomical localization and responses to neurohumours and ACTH, Brain Research, 12 (1969) 74-85. TERASAWA, E., AND SAWYER, C. H., Changes in electrical activity in the rat hypothalamus related to electrochemical stimulation of adenohypophyseal function, Endocrinology, 85 (1969) 143+149. WmTEHEAD, S. A., AND RUV, K. B., Responses of antidromically identified preoptic neurons in the rat to neurotransmitters and to estrogen, Brain Research, 79 (1974) 185-198. YAGI, K., Effects of estrogen on the unit activity of the rat hypothalamus, J. Physiol. Soc. Japan, 32 (1970) 692-693. YAOI, K., Changes in firing rates of single preoptic and hypothalamic units following an intravenous administration of estrogen in the castrated female rat, Brain Research, 53 (1973) 343-352. YAGI, K., AND SAWAKI, Y., Feedback of estrogen in the hypothalamic control of gonadotrophin secretion. In K. YAGI AND S. YOSHIDA (Eds.), Neuroendocrine Control, Univ. of Tokyo Press, Tokyo, 1973, pp. 297-325.
