- Email: [email protected]
Locally applied progesterone metabolites alter neuronal responsiveness in the cerebellum
0361-9230/87 $3.00 + .OO
Brain Research Bulletin, Vol. 18, pp. 73%747. B Pergamon Journals Ltd., 1987. Printed in the U.S.A.
Locally Applied Progesterone Metabolites Alter Neuronal Responsiveness in the Cerebellum SHERYL
S. SMITH,’
Department
B. D. WATERHOUSE
AND D. J. WOODWARD
of Physiology and Biophysics, Hahnemann University Broad and Vine, Philadelphia, PA 19102-1192 Received
30 December
1986
SMITH, S. S., B. D. WATERHOUSE AND D. J. WOODWARD. LocoNy appliedprogesterone metabolites alter neuronnl responsiveness in the cerebellum. BRAIN RES BULL 18(6) 739-747, 1987.-Ongoing studies in this laboratory have
demonstrated that both systemically and locally administered sex steroids 17p estradiol (E,) and progesterone (P) alter cerebellar Purkinje cell responses to microiontophoretically applied amino acid neurotransmitters GABA and glutamate (GLUT) in the urethane-anesthetized, ovariectomized adult rat. In the present study, we have examined the effects of several locally pressure ejected P metabolites on Purkinje cell responsiveness to GABA and GLUT: .5a-pregnane-2tJ-one (5a DHP), 5o-pregnane-3a-ol-20-one (3a OH-DHP) and 5a-pregnane-3P-ol-20-one (3p OH-DHP). GABA-induced inhibition was markedly enhanced immediately after onset of local application of 3o OH-DHP or 5a DHP, unaccompanied by alterations in background discharge. Both metabolites also attenuated excitatory responses to GLUT by O-3 min after initiation of steroid application. In both cases, recovery to control levels of response was observed 6-9 min after termination of pressure application. These results are similar to those seen after local or systemic injection of P. In contrast, 3p OH-DHP did not produce any alteration in Purkinje cell responses to either amino acid. As 5a DHP and 3a OH-DHP can be localized in cerebellar tissue after P administration, the results presented here suggest that the neuronal effects of systemic P may be mediated by local membrane actions of P or its metabolites. Progesterone metabolites 3a OH-DHP Neuromodulation Anxiolytic
SLYOH-DHP
WE have previously shown [33,34] that both systemic and local administration of progesterone (P), at physiologic doses, results in marked augmentation of inhibitory responses of cerebellar Purkinje cells to microiontophoretitally applied gamma aminobutyric acid (GABA). In addition, by both routes of administration, P significantly attenuates excitatory neuronal responses to locally applied glutamate (GLUT). Both of these modulatory actions mimic actions of benzodiazepines on neuronal responsiveness, and, in a global sense, are consistent with the reported anxiolytic 17, 12, 15, 301, anticonvulsant [6,16] effects of the steroid. In neither case was the GABA enhancing or GLUT suppressing action of P immediate however, suggesting the possibility that conversion to other metabolites might precede the observed neuromodulatory actions of the steroid. In addition, a number of recent reports have emerged indicating that several P metabolites, including 5a pregnane-3cu hydroxy-20 one (3c~OH-DHP), can increase GABA binding 1261, as well as chloride conductance [14,26], both independently and synergistically with GABA, in cultured embryonic spinal cord and hippocampal neurons. Previous reports have
GABA
GLUT
Cerebellar Purkinje cell
shown that 3a OH-DHP, as well as 5a DHP, localize in cerebellar tissue to a greater extent than P itself following systemic injection of [H13-P [17]. Further, conversion of P to these metabolites can occur locally in the cerebellum [ 13,181. To test whether the observed effects of P on neurotransmitter responsiveness could be mediated by a metabolite, we have pressure ejected the progestins 3a OH-DHP and 5a DHP in the vicinity of individual cerebellar Purkinje cells and evaluated steroid-induced alterations in neuronal responses to iontophoretically applied GABA and GLUT. In order to test the stereospecificity of the observed metabolite effects on neuronal responsiveness, a third progestin, the 3p isomer of 3a OH-DHP, was also tested using an identical paradigm. METHOD
Animals Adult, ovariectomized (3-6 weeks) albino rats (SpragueDawley, 20&300 g) were housed in groups of 5-6 at a room temperature of 23-25°C. Animals were allowed free access to food and tap water and entrained to a 14: 10 light-dark cycle.
‘Requests for reprints should be addressed to Sheryl S. Smith. 2Department of Cell Biology and Anatomy, University of Texas Health Science Center, 5323 Harry Hines Blvd., Dallas, TX 75235.
739
SMITH, WATERHOUSE
740
AND WOODWARD
. 8
i
CONTROL
44%
/
M-OH-IJHP
O-3
MIN
3A-OH-OHP -
3-6 _
MIN _
RECOVERY
O-3
MIN
4-7
MIN
1----
10
20
30
RECOVERY
EJECTION PRESSURE ml)
B
ii” 6 0
3
6 mm
9
FIG. 1. Quantification of progesterone release by pressure ejection and extent of spread in cerebellar tissue. (A) Rate of sex steroid release: P (0.8 mM in 0.01% propylene glycol-saline) was pressure ejected into a buffer solution (0.01 M potassium phosphate buffer, pH 7.0) for a 3 minute period. Ejection pressures for individual samples varied from l-30 psi (pounds per square inch). Quantitication of the steroid was accomplished using Niswender’s antibody (GDN-337) and standard radioimmunoassay procedures [l]. Results were plotted as pg released in 3 minutes (ordinate) versus ejection pressure (abscissa). The steroid exhibited a linear release pattern up to 20 psi. At higher pressures, steroid release tended to level off. (B) Extent of spread: The radiolabeled sex steroid [1,2,6,7-WI-P (114 Wmmol. New England Nuclear) was pressure ejected at 4-5 psi for 3 minutes into lobule VI or VII (I-1.5 mm lateral to midline-see arrow) in the vicinity of a Purkinje cell, identified by its characteristic discharge pattern of simple and complex spikes. After the 3 minute ejection period, the rat was decapitated, the brain removed and frozen at -40°C in isopentane and cut into 20 CL sections using a cryostat. Sections were exposed to X-ray film for a 2 week period and the radius of spread determined to be approximately 100 p.
FIG. 2. 3u OH-DHP augments Purkinje cell responsiveness to GABA. Ratemetei- records (left) and drug-response histograms (right) indicate changes in Purkinje cell (P cell) responsiveness to GABA before, during and after local pressure injection of the P metabolite, 3q OH-DHP (1-2 psi for 6 mitt). Each histogram sums unit activity during 4-8 IO set GABA pulses (solid bar). occurring at 40 set intervals. GABA-induced inhibition is indicated as a percent reduction in firing rate relative to spontaneous discharge (numbers next to bars). Excerpts of control, drug effect and recovery data are presented. Marked increases in GABA responsiveness are seen immediately O-3 min after the onset of 3o OH-DHP application. These results are representative of 17 out of 20 cells. Local application of vehicle alone resulted in no alteration in GABA response.
Surgrr~ Urethane-anesthetized rats (1.2 g/kg, IP) were employed for all studies. Animals were placed in a stereotaxic apparatus and a craniotomy performed to expose parts of the anterior and posterior lobes of the cerebellum, as has been described [33]. Body temperature was maintained at 3637°C with a heating pad. Single Purkinje cells were recorded l-l.5 mm lateral to midline, in the paravermal area of the cerebellum. The saline-filled central barrel (3 M NaCl) of a 5 barrel micropipet (4-6 p tips) was used for recording extracellular discharge. Side barrels were filled by diffusion with either gamma aminobutyric acid (GABA, 1.0 M, pH 4.0, Sigma), glutamate (GLUT, 1.O M, pH 8.0, Sigma) or the steroid to be pressure ejected. GABA and GLUT were ejected as anions or cations, respectively, and retained by application of 15 nA current of opposite polarity [lo]. In addition, one barrel (saline filled) was used for current balancing. Solutions of progesterone (P, 0.8 mM) and various metabolites: 5a
BY PROGESTINS
NEUROMODULATION
741
a4
% ‘ONTRoL
GLUT
35nA
CONTROL -
-
GLUT
35nA -
3
wm;j?s;;,
,-
3A
OH
3A
OH
-1-
OHP
-
-
OHP
-
-
MIN -
3-6
MIN
-
RECOVERY
_l-
O-3
6-9
-
SA
OHP
O-2
MIN 1
33%
RECOVERY
-
2-4
-
MIN
26%
-
FIG. 4. 5u DHP attenuates GLUT excitation of Purkinje cell discharge. Control, drug effect and recovery data are depicted, illus-
MIN -
47x
. __ . . . FIG. 3. 3a OH-DHP suppresses GLUT excitation. Strip chart records (left) and peri-event histograms (right) illustrate the effect of local application of 3o OH-DHP (3 psi, 6 min) on P cell responses to iontophoretically applied GLUT (solid bar, 35 nA). The degree of GLUT excitation is indicated as a percent change from baseline (numbers next to bar). Unit discharge was monitored continuously before, during and after steroid application. A modest but significant decline in GLUT excitation is evident 3-6 min after initiation of steroid application. This result is similar to that seen in 14 of 15 cells tested. Local application of vehicle alone resulted in no alterations in GLUT response (n=12). Note: In this and subsequent figures, amino acid pulses consistently occur every 40 set; however, the number of pulses/histogram (and thus, ratemeter time base) may vary.
pregnane-20 one (5a DHP, 0.8 mM, Sigma), 5a: pregnane-3cr 01-20 one (3a OH-DHP, 0.8 mM, Sigma) or 5a pregnane-3/3 01-20 one (3/I OH-DHP, 0.8 mM, Sigma) were pressure ejected locally at l-5 psi. (Vehicle: 0.01% propylene glycolsaline, pH 7.4.) Purkinje neurons were identified by their unique electrophysiological signature [9] as previously described [33]. Unit discharge was monitored on an oscilloscope and converted to uniform voltage pulses by a window discriminator. Peri-drug histograms were then constructed using a microeclipse computer (Data General) and the Thoth program, written by Dr. John K. Chapin, Hahnemann Univ., Philadelphia. Average amino acid responses of a neuron were computed before, during (10-15 min) and after (l-2 hr) continuous sex steroid application. Steroid-alterations in background activity (spontaneous discharge), GABA inhibition/GLUT excitation (evoked discharge) and the evoked:spontaneous ratio (amino acid response relative to background) were calculated. Changes in the evoked:spontaneous ratio are suggestive of modulatory effects of a drug rather than simple inhibitory or excitatory effects which would be expected to alter
trating the suppressive effect of pressure ejected 5a DHP (2.5 psi, 2 min) on GLUT response. This effect is representative of 17 out of 20 cases. evoked and spontaneous discharge patterns to the same degree. The paired sample r-test or one-way analysis of variance and Student-Newman-Keuls tests were used to statistically assess significant differences between pre- vs. poststeroid or across steroid groups, respectively. Pressure
Ejection
This procedure has been described previously [23,31]. Briefly, polyethylene tubing (o.d. 2 mm, i.d. 1 mm) was glued into the steroid filled barrel of the recording micropipet. The appropriate steroid solution was ejected by compressed nitrogen gas, released with a pneumatic valve (Medical Systems Corp.). Quantification of the P ejected in this manner was accomplished using radioimmunoassay (RIA) procedures [l], kindly performed by Kathy Katz in the lab of Dr. Sergio R. Ojeda, Dallas. P was ejected at various pressures (l-30 psi) into test tubes containing 0.01 M potassium phosphate buffer (0.14 NaCl, 0.1% Knox gelatin and 0.01% merthiolate, pH 7.0: RIA assay buffer). The P (GDN 337) antiserum used was generously provided by Dr. Gordon D. Niswender (Colorado State Univ.). [1,2,6,7-H3]-P (114 Wmmol, New England Nuclear), standards, antiserum and samples for all assays were diluted in RIA assay buffer. Separations of antibody bound from free hormone were achieved with dextrancoated charcoal. Total incubation time was 90 min at room temperature, with the labeled hormone added after 30 min. After separation and centrifugation of the samples, the radioactivity in 0.5 ml of the supematant was assessed using a scintillation counter. Diffusion
of Pressure
Ejected
Steroid
Radiolabeled P [1,2,6,7-H3]-P (114 Wmmol, New England Nuclear) was pressure ejected at 4-5 psi for three minutes into lobule VI or VII of the paravermal cerebellum ( 1- 1.5 mm lateral to midline) in the immediate vicinity of a Purkinje neuron. After the 3 minute ejection period, the rat was de-
SMITH, WATERHOUSE $
5A
OHP
O-3
MIN
_
33%
40%
39
Cb
CPP
<-
5A
OHP
3-7
AND WOODWARD
2-i
I.’ ; :
-
-
MIN
RECOVERY
O-3
MIN
432
RECOVERY
3-7
MIN
35x
38
FIG. 5. 5a DHP potentiates Purkinje cell GABA responses. Using the same format as the previous figures, local 5a DHP effects on GABA-induced inhibition are demonstrated. Significant potentiation of the GABA response is seen by 3 min after the onset of continuous pressure ejection of this steroid (2.5 psi, 7 min), with recovery apparent by 7 min after termination of steroid administration. This effect was seen in 13 of 14 cases.
-I-
30
capitated, the brain removed, frozen at -40°C in isopentane and cut in 20 p sections using a cryostat. Sections were exposed to X-ray film for a 2 week period, and the radius of spread of the steroid determined. RESULTS
Quantification oj’ Sex Steroid Release by Pressure und Extent of Spread in Cerebellar Tissue
Ejection
In order to determine whether physiologic levels of P were being administered, release of the steroid was quantified with RIA procedures and extent of spread in cerebellar tissue determined autoradiographically. A range of ejection pressures (l-20 psi) released P in a linear fashion, when expressed as pg steroid ejected over a 3 min period (see Fig. 1A). Only at an ejection pressure of 20 psi did P release deviate from this linear relationship. However, for all experimental procedures, pressures of l-5 psi were used to apply the steroid. Because of the lipophilic nature of P, the extent of spread of the pressure ejected steroid was evaluated in order to estimate the neuronal components with which the steroid might come into contact. As depicted in Fig. lB,
1-
OH
OHP
-
-
OH OHP -
-
O-4
MIN
-
-
4-9 -
7
--
MIN -
-
41%
52% 1
-
FIG. 6. 3p OH-DHP does not alter Purkinje cell responses to either GABA or GLUT. This figure illustrates the inability of 3p OH-DHP to alter neuronal responses of P cells to iontophoretically applied GABA (A) or GLUT (B). In both cases, afier stable amino acid responses were obtained, continuous pressure application of 3p OH-DHP at 2 psi for 7-9 min failed to alter neuronal responses to either neurotransmitter. These results are representative of 13 of 15 cases (A) and 9 of 10 cases (B).
radiolabeled P exhibited a 100 p radius of spread, as determined by autoradiographic techniques. Thus, the pressure ejected steroid (45 psi, 3 min) would encounter no more than 20 Purkinje cells within one folium, with associated neuronal afferents and intemeurons. A rough estimate of the effective concentration of P would be l-3 x IO+ M, well within the physiologic range.
NEURGMODULA~ON
743
BY PROGESTINS
F GLUT RESPONSE
SPONTANEOUS , WGHARGE
”
FIG. 7. Steroid induced alterations in spontaneous discharge, amino acid response and the evoked: spontaneous ratio. Progestin effects on Purkinje cell responses to GABA (A) or GLUT 03) are presented. The percent change (from pre-steroid vafnes) in s~ntaneuus discharge, amino acid response and evoked response refative to spoutaa~ous firing was evafuated for all steroids tested, (*~
charge. No alterations ia background firing were noted after: local steroid administrat~~~~ Recovery was observed by 6-9 min after termination of steroid release. Local application of 3ar OH-DHP resulted in a marked and immediate potentiation of GABA-induced inhibition, which was increased by an average of 70% (n=20). The cell depicted in Fig. 2 illustrates this effect: An initial GABA i~~bition of 4% was increased by SS%, to a 7% inh~bi~on of b~~k~ound firing within seconds after the onset of local application of the steroid. Only a slight recovery from this GABA augmen~tion was noted. A variable recovery rate was noted for all cases tested, with onfy 6@% of the cells demanstrating significant recovery to control levels of GABA inhibition. Insignificant alterations in background discharge were noted during steroid application. 3a OH-DHP also produced a modest but significant suppression of GLUT-evoked excitation (n==121. As exemplified by the cell in Fig. 3, by 3-6 min after the onset of steroid application, a GLUT excitation of 3% was reduced by 3@% to a 25% GLUT-evoked increase above background dis-
The locally applied progestin, 5a DHP, decreased excitatory responses of Purkinje cells to GLUT by 45-W% (n=ZQ), as depicted by the cell in Fig. 4, In this example, 35 nA GLUT increased neuronal firing by 46% above b~ck~ound discharge. Cont~uons ejection of 5or DHP produced a W% suppression of the amino acid response, which was reduced to a 26% excitation, accompanied by only a minor decrease ia spon~~~us discharge, by O-2 min after initiation of steroid application. This locally applied metabolite also potentiated GABA inhibitory responses by an average of 50% (n=14). This effect is illustrated by the ceil presented in Fig. 5. fn this example, CABA inhibited background discharge by 33%.
SMITH, WATERHOUSE
744
Continuous application of 5cr DHP augmented this inhibitory response by 50%, an effect seen 3-7 min after the onset of pressure ejection of the steroid, with no concomitant alteration in the spontaneous tiring rate. After termination of pressure application, however, a small increase in background discharge was noted, which preceded by several min the return of GABA inhibitory responses of the Purkinje cell to pre-steroid levels. 3/3 OH-DHP Either GABA
Dovs Not Alter Parkinje or GLUT
Cell Responses
Steroid Induwd Alterations in Spontaneous Dischargr, Amino Acid Rrsponse and the Evoked:Spc,ntaneous Ratio
In order to determine to what extent observed alterations in amino acid responsiveness were due to changes in background discharge, the average maximal percent change in spontaneous discharge, amino acid response and the evoked: spontaneous ratio (i.e., alterations in amino acid responses relative to changes in background) discharge induced by the steroids from pre-steroid values were determined (Fig. 7). Alterations in the evoked:spontaneous ratio reflect differential actions of a drug on amino acid responses vs. background activity, rather than simple excitatory or inhibitory effects which would be expected to alter spontaneous and evoked discharge to an equivalent degree. A differential effect on evoked spontaneous discharge could be considered a modulatory action of a drug. i.e., indicating direct drug-induced changes in amino acid responses, rather than indirect alterations in this parameter secondary to changes in cell excitability. In Fig. 7, the progesterone meta~lite results are compared with progesterone-induced effects on these parameters. In Fig. 7A. progestin-GABA effects are summarized: In no case did P, ~CXDHP, 3a OH-DHP or 3/3 OH-DHP produce significant alterations in spontaneous discharge. However, only the 3/3 OH-DHP metabolite failed to change GABA responsiveness. 3a! OH-DHP, 5a DHP and P were all able to significantly augment GABA inhibition (by 52, 37 and 21%, respectively), as well as increase the evoked:spontaneous ratio (by 120, 76 and 92%, respectively). 3a OH-DHP was the most potent progestin in terms of alterations in both of these parameters. With regard to GLUT function (Fig. 7B), however, 5a DHP suppressed GLUT excitation to the greatest degree (2%) and increased the evoked:spontaneous ratio by 35%. P and 3crOH-DHP also exerted significant effects on the evoked: spontaneous ratio, with P the most potent in this capacity, yielding a 43% decrease in GLUT evoked:spontaneous changes in firing. Again, in no case did any steroid alter spontaneous firing rate significantly. in Neurotransmitter of P and Metabolites
A .-. p ..-..
I......
‘
38
ot+m+p
3A OH-DHP
to
3p OH-DHP, a progesterone metaboiite with no anesthetic properties, failed to alter neuronal responsiveness to either amino acid neurotransmitter tested. As exemplified in Fig. 6, 7-9 min of continuous local application of this metabolite did not result in augmentation of GABA inhibition (Fig. 6A, n= 15), suppression of GLUT excitation (Fig. 6B, n= 10) or any changes in background discharge (Fig. 6A and B).
Tiww Course Q Alterations Responses: A Comparison
AND WOODWARD
Evoked
An evaluation of the temporal aspects of the neuromodulatory effects of the progestins, including progesterone, also
0
I
(MINI
O-3
3-6
6-9
O-3
9-12
\
/
3-6
12-15
<
STERDID EJECTKIN
/ POST-STEROID
B .-.
p
.--.
SA-DHP
o-3
3-6
..--*
6-9
3A OH-DHP
-‘.” 30 OH-DWP
O-3
9-12 /
\ STEROID EJECTIDN
3-6
12-15
.
/ POST-STEROID
FIG. 8. Time course of alterations in neurotransmitter evoked responses: A comparison of P and metabolites. Steroid-induced alterations (% change from pre-steroid values) in GABA-induced inhibition (A) or GLUT-evoked excitation (B) relative to background are presented for P and various metabolites, So DHP, 3a OH-DHP and 3p OH-DHP. Mean values are plotted for each of the groups at 3 min intervals during and after Iocai steroid application. Only 3/3 OH-DHP failed to significantly alter either parameter. Each point represents the mean of 15-25 cells. (Bars indicate SEM,) *p
revealed differences in terms of steroid effects on amino acid responses. 3o OH-DHP exerted the fastest potentiating effect on GABA inhibition (O-3 min, see Fig. 8A). P and 5a? DHP, however, exerted significant effects on this parameter with a longer latency to onset (3-6 min). Partial recovery of drug effect was observed for all three steroids, with Peffects displaying the most variable recovery rate. As in the previous figure, 3/3
NEUROMODULATION
BY PROGESTINS
OH-DHP had no effects on GABA responses at any point post-injection. An evaluation of steroid effects on GLUT responses demonstrated that, unlike steroid-induced GABA potentiation, attenuation of GLUT excitation was accomplished with the shortest latency by 5a DHP (O-3 min). P and 3a OH-DHP produced somewhat slower effects (3-6 min). Again, recovery from the neuro-modulatory actions of P were the most variable. Surprisingly, 3p OH-DHP produced a short-latency, transient increase in GLUT excitation. DISCUSSION
P-like effects on cerebellar Purkinje cell responses to amino acids, potentiation of GABA inhibition and suppression of GLUT excitation, were demonstrated in the present study for the P metabolites 3a OH-DHP and Sa! DHP. The observed steroid effects were also dependent upon the specific stereoconformation of the steroid, as a third metabolite, 3p OH-DHP (allopregnanolone), did not alter neuronal responsiveness to any degree. The results of this study raise the possibility that specific metabolites of P, and not the steroid itself, may mediate the observed neuromodulatory actions of P on Purkinie cell responses to GABA and/or glutamate (GLUT) 1331.This appears to be a viable possibility for several reasons. First, our previous reports have shown that both intravenous and local application of P result in neuromodulatory effects with a 5-10 minute latency [33]. However, 3o OH-DHP and 5a DHP resulted in immediate and robust augmentation of GABA responses and suppression of GLUT excitation, respectively. Both 3a OH-DHP and 5a DHP, collectively, have been localized to cerebellum in larger amounts than P alone by 10 minutes after IV injection of a physiologic dose of [H13-P, but are reduced to low levels by 30 minutes post-injection [ 171, a time course which closely parallels that observed for effects of systemic P on neurotransmitter responses [33]. Local conversion of P to 5a DHP could be accomplished within minutes, as the enzyme necessary for this process, 5a reductase, has been localized within cerebellar tissue [18]. In addition, infusion of [H13-P into cerebellar tissue in vitro results in accumulation of various P metabolites, including 3a OH-DHP and 5rwDHP [ 131, further suggesting that local conversion of P to these metabolites can be accomplished in this CNS site. Therefore, both IV and local administration of P would result in conversion to 3cr OH-DHP and 5a DHP and eventual localization in cerebellar tissue within the necessary time frame for the observed neuromodulatory actions of this steroid on GABA function. Additional evidence that 3a! OH-DHP may mediate the observed actions of P is provided by recent reports indicating that this metabolite can increase the membrane chloride ion current, either alone or synergistically with GABA, in cultured embryonic rat hippocampal and spinal cord neurons [14,26]. This was found to be an immediate, reversible effect, similar to that described in the present report. In addition, this metabolite was also shown to increase benzodiazepine/GABA binding [26], as has been shown previously for P [25], and stimulate chloride uptake into isolated brain vesicles [26]. Thus, several mechanisms exist by which 3a OH-DHP may mediate the GABA-enhancing effects of P. In contrast, 5a DHP in the present study did not potentiate GABA responses to as great a degree or as rapidly as did 3a OH-DHP. This finding is supported by Harrison et al. [14] who have specified structural re-
145
quirements of steroids able to increase chloride conductance using a voltage clamp paradigm; only 3a OH-DHP, but not 5a DHP, was potent in this regard. Because conversion of 5a DHP to 3a OH-DHP can also occur in the cerebellum, it may be that the observed GABA-potentiating effect of 5a! DHP is in fact due to the partial conversion of this compound to the more potent 3a OH-DHP. The longer latency to effect may in fact be a result of this type of conversion. However, 5~ DHP-induced suppression of GLUT excitation was immediate and prominent, unlike 3o OH-DHP effects on this parameter, which were less extensive and of longer latency. These data suggest that, in contrast to effects on GABA function, P-induced suppression of the GLUT response may be mediated by the 5a DHP metabolite. In addition, the fact that 3p OH-DHP failed to alter GABA responsiveness indicates the stereospecificity of this effect and is also consistent with the report of Harrison er al. [ 141which indicates that any hydroxy group present at C-3 of the steroid molecule must be in the (Yconfiguration in order to permit effects on chloride conductance. It is interesting, however, that this compound was also inactive in terms of GLUT modulation, which could not be predicted based on the differential effect of 3o OH-DHP and 5a DHP on these parameters. Mechanism
of Action
The steroid effects demonstrated in this and our previous studies have too rapid an onset to be genomic actions, and are most likely the result of interactions with membrane constituents. Membrane effects of P have been demonstrated using DPPC liposomes [4,35], erythrocyte membranes [19] and Xenopus faevis oocyte membranes [32]. In the latter, P has been shown specifically to release membrane phospholipid bound Ca++ prior to meiosis [32]. P has also been reported to act at the membrane level of the striatum to release dopamine [8] and at the hypothalamic membrane to release LHRH [20]. In the uterus, P exerts local effects on myometrial excitability [28]. It has been shown that all steroids are able to intercalate to some degree within membrane phospholipids [21,37], dependent to some extent on the presence of aliphatic side chains. However, a remarkable degree of stereospecificity is seen in terms of steroid H bond interaction [3] and anisotropic motion [16], such that an axial (i.e., 3a! OH-DHP) versus an equatorial (i.e., 3/3 OH-DHP) hydroxy group may be one limiting factor for such interactions. It has been speculated that interaction with the “hydrogen belt,” i.e., the carbonyl and hydroxyl groups of constituent membrane lipids, may in fact be the mechanism for anesthetic effects of steroids [3]. It may be that such interactions also contribute towards the neuromodulatory actions of a steroid, as a similar stereospecificity was noted in the present study in terms of steroid-induced alterations in both GABA and GLUT responsiveness. Although P itself has no OH group or aliphatic side chain, a recent study has determined that membranes containing 10% cholesterol are good solvents for P, and such a domain enables P to alter the phospholipid organization of the lipid bilayer [4]. Thus, synaptic membranes, which contain approximately 10% cholesterol [29], would be potential sites for this type of interaction. In addition, although the cerebellum contains only low levels of classic P receptors [24], membrane receptors for the steroid may exist [36], such as those
SMITH, WATERHOUSE
746
which have been reported for EZ 1361. Therefore. the present do not rule out the possibility that P itself may be able to exert direct effects on membrane parameters that lead to the observed neuromodulatory effects on GABA and GLUT funcdata
tion. Additional
mechanisms responsible for the observed neuromodulatory effects of P could involve indirect steroid actions on the cerebellar monoamine systems. Norepinephrine [27] and serotonin [22] have been shown to potentiate GABA responses and diminish GLUT responses of cerebellar Purkinje cells, respectively, in a manner similar to P. It is intriguing to consider the possibility that the various metabolites of P may subserve specific functions at different end organs. Only P exerts progestogenic effects on the uterus [l&38]. However, 5~yDHP, which exerted more potent effects on GLUT function relative to GABA in the present study, is also known to exert potent P-like effects on facilitation of gonadotropin release and lordosis at the level of the hypothalamo-pituitary-gonadal axis [ 18,381. On the other hand, 3a OH-DHP, a more potent GABA enhancing agent than P, as demonstrated in the present study, exerts only limited effects on gonadotropic regulation and reproductive behavior, and these are primarily of an inhibitory nature [38]. In addition, cyclic alterations in activity of the conversion enzymes, 5cu reductase and 3cu hydroxysteroid oxidoreductase, may determine to what extent the various metabolites are active. In general, CNS activity of these enzymes is highest on estrus and proestrus [l8]. Endogenous P levels are elevated on the afternoon of proestrus and again, to a lesser extent, on the evening of diestrus 1. Thus, differential neuromodulatory effects of P vs. the Sa-reduced metabolites may exist which are dependent both on substrate and
AND WOODWARD
enzyme levels at any given stage of the estrous cycle. It is also of interest that the effects of P. 3a OH-DHP and 5a DHP are similar to those reported for benzodiazepine action in terms of potentiation of GABA response [ 111, as well as attenuation of GLUT action [5], both of which have been demonstrated using the cerebellar Purkinje cell model in the intact rat. The behavioral consequences of P administration have also been shown to resemble those of the benzodiazepines in terms of their anxiolytic [7, 12, 15, 301 and anti-convulsant [6,16] properties. Although the GABAenhancing actions of both the progestins and the benzodiazepines is a sufficient mechanism to explain their anticonvulsant effects, potent GABA agonists do not necessarily exert anxiolytic effects [2]. Thus. suppression of excitatory synaptic events may also be a component of anxiolytic action. In summary, the present results suggest that two metabolites of P, 3~yOH-DHP and 5n DHP, may serve to mediate either some or all of the observed neuromodulatory actions of P. In addition, the two metabolites may subserve different functions, 3a OH-DHP exerting greater effects on GABA potentiution and 50 DHP. more pronounced suppression of GLUT excitation.
ACKNOWLEDGEMENTS
Suppol-ted by NS25809 to S.S.S.: NS I8081 to B.D.W.: AA3901, DA02338 and a Biological Human& Grant to D.J.W. The authors would like to express their gratitude to Ms. Kathy Katz and Dr. Sergio R. Ojeda (Dallas) for their generous technical assistance in quantifying pressure ejected release of P using RIA procedures. The authors also thank Dr. Neil Harrison (NINCDS) for his helpful discussions during the course of the work.
REFERENCES
5.
6.
7.
8.
Andrews, A. A., J. P. Advis and S. R. Ojeda. The first proestrus in the female rat: Circulating steroid levels preceeding and accompanying the preovulatory LH surge. Proc, Sot, E.rp Biol Mccl 163: 305-310, 1980. Bartholini, G. Neuropsychopharmacological and therapeutic spectra of GABA receptor agonists. In: Gahn crnd Moc~cl I>i.\ordrrs. E.upcrimcntal cud Clinicul Rcscurch. edited by G. Bartholini, K. G. Lloyd and P. L. Morselli. New York: Raven Press, 1986, pp. l-7. Brockerhoff, H. Anesthetics may restructure the hydrogen belts of membranes. Lipidc 17: 100-1003, 1982. Carlson, J. C., M. Y. Gruber and J. E. Thompson. A study of the interaction between progesterone and membrane lipids. E/ldocrino/og.v 113: 19%194, 1983. Cheng, J. T. and B. D. Waterhouse. Interactions of flurazepam with glutamate evoked excitatory responses of cerebellar Purkinje neurons. Sot, Nrurosc? Ahsfr 9: 398, 1983. Craig, C. R. Anticonvulsant activity of steroids: separability of anticonvulsant from hormonal effects. J Phrrrmtud Exp Thor 153: 337-343, 1966. Dennerstein, L., C. Spencer-Gardner, G. Gotts, J. B. Brown, M. A. Smith and G. D. Burrows. Progesterone and the premenstrual syndrome: a double blind crossover trial. Br Mrcl J 290: 1617-1621, 1985. Dluzen, D. E. and V. D. Ramirez. Bimodal effect of progesterone on in vitro dopamine function of the rat corpus striatum. Neuroentloc.rirro/l~~~ 39: 149-155, 1984.
9. Eccles, J. C., M. Ito and J. Szentdgothai. 7’lrc (‘c,rc,hc//r//,r (I\ (I Ncuromrl Muclri~c,. New York: Springer-Verlag, 1967. IO. Geller. H. M. and D. J. Woodward. An improved constant current source for microiontophoretic drug application studies. Elec,rroc,nc.c~phohgr Clin Ncurophysiol 33: 43&432. 1972. II Geller, H. M., D. A. Taylor and B. J. Hoffer. Benzodiazepines and central inhibitory mechanisms. Norrrlyn Schmiet/rhrr,gt Arch Phrrrmtlud 304: 8 l-88. 1978. 12. Click, I. D. and S. Bennett. P\ychlatric effecta of progeztel-one and oral contraceptive\. In: P~ychitrrric Complictrtkms oj’Mrdictrl Dru,yc. edited by R. Shader. New York: Raven Press, 1972, pp. 295-332. I3 Hanukoglu, l., H. J. Karavolas and R. W. Goy. Progesterone metabolism in the pineal. brain stem, thalamus and corpus callosum of the female rat. BTN~‘/zRc.t 125: 313-324, 1977. 14 Harrison. N. I-., M. D. Majcw\ko. J. W. Harrington and J. 1.. Barkel-. Structure-activity relationships for steroid interaction with a gamma-aminobutyric acid, receptor complex. .I P/~tr,.nroc,r,/ ,5-\-p T/I<,/. 241: 346. 1987. 15. Herrmann, W. M. and R. C. Beach. Experimental and clinical data indicating the psychotropic properties of progestogens. Postgrd Md J 54: 82-87, 1978. 16. Holmes, G. L. and D. A. Weber. The effect of progesterone on kindling: A developmental study. &,\I Brcrin Res 16: 45-53. 1984.
NEUROMODULATION
BY PROGESTINS
17. Karavolas, H. J., D. R. Hodges, D. J. O’Brien and K. M. MacKenzie. In vivo uptake of [H3]-progesterone and [H3]5 alphadihydroprogesterone by rat brain and pituitary and effects of estradiol and time: Tissue concentration of progesterone itself or specific metabolites? Endocrino1og.v 104: 14181425, 1979. 18. Karavolas, H. J., P. J. Bet-tics, D. Hodges and N. Rudie. Progesterone processing by neuroendocrine structures. In: Metabolism of Hormonal Steroids in the Neurorndocrine Structures, edited by F. Celotti, F. Naftolin and L. Martini. New
York: Raven Press, 1984, pp. 149-170. 19. Kaya, H. and T. Saito. Effect of progesterone and its 17 alphahydroxyl derivative on human erythrocyte membrane. Jptz J Pharmacol
39: 299-306,
1985.
20. Ke, F. C. and V. D. Ramirez.
The plasma membrane of rat hypothalamic tissue is the primary site for progesterone stimulatory action of LHRH release. Sot Neurosci Ahstr 12:
604, 1986. 21. Lawrence, D. K. and E. W. Gill. Structurally
22.
23.
24.
25.
specific effects of some steroid anesthetics on spin-labeled liposomes. Mel Pharmacol 11: 280-286. 1975. Lee, M., J. C. Strahlendorfand H. K. Strahlendorf. Modulatory action of serotonin on glutamate-induced excitation of cerebellar Purkinje cells. Brain Res 361: 107-I 13, 1985. McCaman, R. E., D. G. McKenna and J. K. Ono. A pressure system for intracellular and extracellular ejections of picoliter volumes. Brain Rcs 136: 141-147, 1977. MacLuskey, N. J. and B. S. McEwen. Progestin receptors in rat brain: distribution and properties of cytoplasmic progestinbinding sites. Endoc~rinology 106: 192-202, 1980. Maggi, A. and J. Perez. Progesterone and estrogens in rat brain: Modulation of GABA receptor activity. Eur J Pharmacol 103: 165-168,
1984.
26. Majewska, M. D., N. L. Harrison, R. D. Schwartz, J. L. Barker and S. M. Paul. Steroid hormone metabolites are barbituratelike modulators of the GABA receptor. Science 232: 10041007, 1986. 27. Moises, H. C., D. J. Woodward, B. J. Hoffer and R. Freedman. Interactions of norepinephrine with Purkinje cell responses to putative amino acid neurotransmitters applied by microiontophoresis. Esp Nrurol 64: 493-515, 1979.
747 28. Osa, T. and T. Ogasawara. Effects of in vitro progesterone and estradiol-17/3 on the contractile and electrical responses in rat myometrium. Jpn J Physiol 34: 427-441, 1984. 29. Robertson, R. N. The Lively Membranes. New York: Cambridge University Press, 1980, pp. 152-153. 30. Rodriguez-Sierra, J. F., J. L. Howard, G. T. Pollard and S. E. Hendrickes. Effect of ovarian hormones on conflict behavior. Psyc.hopharmac.ol~~~y (Berlin) 9: 29>300.
, 1984. __-..
31. Palmer; M. R., S. M. Wuerthele and B. J. Hoffer. Physical and physiological characteristics of micropressure ejection of drugs from multibarreled pipettes. Neuropharmac,o/ogy 19: 931-938, 1980.
S., M. Schorderet and E. E. Baulieu. 32. Schorderet-Slatkine, Progesterone-induced meiotic reinitiation in vitro in Xenopus levis oocytes. D$ferentiation 9: 76, 1977. 33. Smith, S. S., B. D. Waterhouse, J. K. Chapin and D. J. Woodward. Progesterone alters GABA and glutamate responsiveness: A possible mechanism for its anxiolytic effect. Brain Rc.c 400: 353-359,
1987.
34. Smith, S. S.. B. D. Waterhouse and D. J. Woodward. Locally applied sex steroids alter neuronal responsiveness to GABA and glutamate in the cerebellum. Sot Neurosci Abstr 12: 990, 1986. 35. Sze, P. Y., E. C. Armstrong, G. G. Deanin and M. W. Gordon. Steroid hormones stabilize liposomal membrane from alteration by tubulin. Sot Nrurosci Abstr 11: 1135, 1985. 36. Towle, A. C. and P. Y. Sze. Steroid binding to synaptic plasma membrane: Differential binding of glucocorticoids and gonadal steroids. J Steroid Biochem 18: 135-143, 1983. 37. Trauble, H. and E. Sackmann. Studies of the crystalline liquid crystalline phase transmition lipid model membranes. III. Structure of a steroid-lecithin system below and above the lipid-phase ._ .-. transition. J Am Chum Sot 94: 4492-4498._, 1972. 38. Zanisi, M., E. Messi and L. Martini. Physiological role of the 5 a-reduced metabolites of progesterone. In: Metabolism of Hormonal Steroids in the Ncuroendocrinc Structures. edited by F. Celotti, F. Naftolin and L. Martini. New York: Raven Press, 1984, pp. 171-183.
Brain Research Bulletin, Vol. 18, pp. 73%747. B Pergamon Journals Ltd., 1987. Printed in the U.S.A.
Locally Applied Progesterone Metabolites Alter Neuronal Responsiveness in the Cerebellum SHERYL
S. SMITH,’
Department
B. D. WATERHOUSE
AND D. J. WOODWARD
of Physiology and Biophysics, Hahnemann University Broad and Vine, Philadelphia, PA 19102-1192 Received
30 December
1986
SMITH, S. S., B. D. WATERHOUSE AND D. J. WOODWARD. LocoNy appliedprogesterone metabolites alter neuronnl responsiveness in the cerebellum. BRAIN RES BULL 18(6) 739-747, 1987.-Ongoing studies in this laboratory have
demonstrated that both systemically and locally administered sex steroids 17p estradiol (E,) and progesterone (P) alter cerebellar Purkinje cell responses to microiontophoretically applied amino acid neurotransmitters GABA and glutamate (GLUT) in the urethane-anesthetized, ovariectomized adult rat. In the present study, we have examined the effects of several locally pressure ejected P metabolites on Purkinje cell responsiveness to GABA and GLUT: .5a-pregnane-2tJ-one (5a DHP), 5o-pregnane-3a-ol-20-one (3a OH-DHP) and 5a-pregnane-3P-ol-20-one (3p OH-DHP). GABA-induced inhibition was markedly enhanced immediately after onset of local application of 3o OH-DHP or 5a DHP, unaccompanied by alterations in background discharge. Both metabolites also attenuated excitatory responses to GLUT by O-3 min after initiation of steroid application. In both cases, recovery to control levels of response was observed 6-9 min after termination of pressure application. These results are similar to those seen after local or systemic injection of P. In contrast, 3p OH-DHP did not produce any alteration in Purkinje cell responses to either amino acid. As 5a DHP and 3a OH-DHP can be localized in cerebellar tissue after P administration, the results presented here suggest that the neuronal effects of systemic P may be mediated by local membrane actions of P or its metabolites. Progesterone metabolites 3a OH-DHP Neuromodulation Anxiolytic
SLYOH-DHP
WE have previously shown [33,34] that both systemic and local administration of progesterone (P), at physiologic doses, results in marked augmentation of inhibitory responses of cerebellar Purkinje cells to microiontophoretitally applied gamma aminobutyric acid (GABA). In addition, by both routes of administration, P significantly attenuates excitatory neuronal responses to locally applied glutamate (GLUT). Both of these modulatory actions mimic actions of benzodiazepines on neuronal responsiveness, and, in a global sense, are consistent with the reported anxiolytic 17, 12, 15, 301, anticonvulsant [6,16] effects of the steroid. In neither case was the GABA enhancing or GLUT suppressing action of P immediate however, suggesting the possibility that conversion to other metabolites might precede the observed neuromodulatory actions of the steroid. In addition, a number of recent reports have emerged indicating that several P metabolites, including 5a pregnane-3cu hydroxy-20 one (3c~OH-DHP), can increase GABA binding 1261, as well as chloride conductance [14,26], both independently and synergistically with GABA, in cultured embryonic spinal cord and hippocampal neurons. Previous reports have
GABA
GLUT
Cerebellar Purkinje cell
shown that 3a OH-DHP, as well as 5a DHP, localize in cerebellar tissue to a greater extent than P itself following systemic injection of [H13-P [17]. Further, conversion of P to these metabolites can occur locally in the cerebellum [ 13,181. To test whether the observed effects of P on neurotransmitter responsiveness could be mediated by a metabolite, we have pressure ejected the progestins 3a OH-DHP and 5a DHP in the vicinity of individual cerebellar Purkinje cells and evaluated steroid-induced alterations in neuronal responses to iontophoretically applied GABA and GLUT. In order to test the stereospecificity of the observed metabolite effects on neuronal responsiveness, a third progestin, the 3p isomer of 3a OH-DHP, was also tested using an identical paradigm. METHOD
Animals Adult, ovariectomized (3-6 weeks) albino rats (SpragueDawley, 20&300 g) were housed in groups of 5-6 at a room temperature of 23-25°C. Animals were allowed free access to food and tap water and entrained to a 14: 10 light-dark cycle.
‘Requests for reprints should be addressed to Sheryl S. Smith. 2Department of Cell Biology and Anatomy, University of Texas Health Science Center, 5323 Harry Hines Blvd., Dallas, TX 75235.
739
SMITH, WATERHOUSE
740
AND WOODWARD
. 8
i
CONTROL
44%
/
M-OH-IJHP
O-3
MIN
3A-OH-OHP -
3-6 _
MIN _
RECOVERY
O-3
MIN
4-7
MIN
1----
10
20
30
RECOVERY
EJECTION PRESSURE ml)
B
ii” 6 0
3
6 mm
9
FIG. 1. Quantification of progesterone release by pressure ejection and extent of spread in cerebellar tissue. (A) Rate of sex steroid release: P (0.8 mM in 0.01% propylene glycol-saline) was pressure ejected into a buffer solution (0.01 M potassium phosphate buffer, pH 7.0) for a 3 minute period. Ejection pressures for individual samples varied from l-30 psi (pounds per square inch). Quantitication of the steroid was accomplished using Niswender’s antibody (GDN-337) and standard radioimmunoassay procedures [l]. Results were plotted as pg released in 3 minutes (ordinate) versus ejection pressure (abscissa). The steroid exhibited a linear release pattern up to 20 psi. At higher pressures, steroid release tended to level off. (B) Extent of spread: The radiolabeled sex steroid [1,2,6,7-WI-P (114 Wmmol. New England Nuclear) was pressure ejected at 4-5 psi for 3 minutes into lobule VI or VII (I-1.5 mm lateral to midline-see arrow) in the vicinity of a Purkinje cell, identified by its characteristic discharge pattern of simple and complex spikes. After the 3 minute ejection period, the rat was decapitated, the brain removed and frozen at -40°C in isopentane and cut into 20 CL sections using a cryostat. Sections were exposed to X-ray film for a 2 week period and the radius of spread determined to be approximately 100 p.
FIG. 2. 3u OH-DHP augments Purkinje cell responsiveness to GABA. Ratemetei- records (left) and drug-response histograms (right) indicate changes in Purkinje cell (P cell) responsiveness to GABA before, during and after local pressure injection of the P metabolite, 3q OH-DHP (1-2 psi for 6 mitt). Each histogram sums unit activity during 4-8 IO set GABA pulses (solid bar). occurring at 40 set intervals. GABA-induced inhibition is indicated as a percent reduction in firing rate relative to spontaneous discharge (numbers next to bars). Excerpts of control, drug effect and recovery data are presented. Marked increases in GABA responsiveness are seen immediately O-3 min after the onset of 3o OH-DHP application. These results are representative of 17 out of 20 cells. Local application of vehicle alone resulted in no alteration in GABA response.
Surgrr~ Urethane-anesthetized rats (1.2 g/kg, IP) were employed for all studies. Animals were placed in a stereotaxic apparatus and a craniotomy performed to expose parts of the anterior and posterior lobes of the cerebellum, as has been described [33]. Body temperature was maintained at 3637°C with a heating pad. Single Purkinje cells were recorded l-l.5 mm lateral to midline, in the paravermal area of the cerebellum. The saline-filled central barrel (3 M NaCl) of a 5 barrel micropipet (4-6 p tips) was used for recording extracellular discharge. Side barrels were filled by diffusion with either gamma aminobutyric acid (GABA, 1.0 M, pH 4.0, Sigma), glutamate (GLUT, 1.O M, pH 8.0, Sigma) or the steroid to be pressure ejected. GABA and GLUT were ejected as anions or cations, respectively, and retained by application of 15 nA current of opposite polarity [lo]. In addition, one barrel (saline filled) was used for current balancing. Solutions of progesterone (P, 0.8 mM) and various metabolites: 5a
BY PROGESTINS
NEUROMODULATION
741
a4
% ‘ONTRoL
GLUT
35nA
CONTROL -
-
GLUT
35nA -
3
wm;j?s;;,
,-
3A
OH
3A
OH
-1-
OHP
-
-
OHP
-
-
MIN -
3-6
MIN
-
RECOVERY
_l-
O-3
6-9
-
SA
OHP
O-2
MIN 1
33%
RECOVERY
-
2-4
-
MIN
26%
-
FIG. 4. 5u DHP attenuates GLUT excitation of Purkinje cell discharge. Control, drug effect and recovery data are depicted, illus-
MIN -
47x
. __ . . . FIG. 3. 3a OH-DHP suppresses GLUT excitation. Strip chart records (left) and peri-event histograms (right) illustrate the effect of local application of 3o OH-DHP (3 psi, 6 min) on P cell responses to iontophoretically applied GLUT (solid bar, 35 nA). The degree of GLUT excitation is indicated as a percent change from baseline (numbers next to bar). Unit discharge was monitored continuously before, during and after steroid application. A modest but significant decline in GLUT excitation is evident 3-6 min after initiation of steroid application. This result is similar to that seen in 14 of 15 cells tested. Local application of vehicle alone resulted in no alterations in GLUT response (n=12). Note: In this and subsequent figures, amino acid pulses consistently occur every 40 set; however, the number of pulses/histogram (and thus, ratemeter time base) may vary.
pregnane-20 one (5a DHP, 0.8 mM, Sigma), 5a: pregnane-3cr 01-20 one (3a OH-DHP, 0.8 mM, Sigma) or 5a pregnane-3/3 01-20 one (3/I OH-DHP, 0.8 mM, Sigma) were pressure ejected locally at l-5 psi. (Vehicle: 0.01% propylene glycolsaline, pH 7.4.) Purkinje neurons were identified by their unique electrophysiological signature [9] as previously described [33]. Unit discharge was monitored on an oscilloscope and converted to uniform voltage pulses by a window discriminator. Peri-drug histograms were then constructed using a microeclipse computer (Data General) and the Thoth program, written by Dr. John K. Chapin, Hahnemann Univ., Philadelphia. Average amino acid responses of a neuron were computed before, during (10-15 min) and after (l-2 hr) continuous sex steroid application. Steroid-alterations in background activity (spontaneous discharge), GABA inhibition/GLUT excitation (evoked discharge) and the evoked:spontaneous ratio (amino acid response relative to background) were calculated. Changes in the evoked:spontaneous ratio are suggestive of modulatory effects of a drug rather than simple inhibitory or excitatory effects which would be expected to alter
trating the suppressive effect of pressure ejected 5a DHP (2.5 psi, 2 min) on GLUT response. This effect is representative of 17 out of 20 cases. evoked and spontaneous discharge patterns to the same degree. The paired sample r-test or one-way analysis of variance and Student-Newman-Keuls tests were used to statistically assess significant differences between pre- vs. poststeroid or across steroid groups, respectively. Pressure
Ejection
This procedure has been described previously [23,31]. Briefly, polyethylene tubing (o.d. 2 mm, i.d. 1 mm) was glued into the steroid filled barrel of the recording micropipet. The appropriate steroid solution was ejected by compressed nitrogen gas, released with a pneumatic valve (Medical Systems Corp.). Quantification of the P ejected in this manner was accomplished using radioimmunoassay (RIA) procedures [l], kindly performed by Kathy Katz in the lab of Dr. Sergio R. Ojeda, Dallas. P was ejected at various pressures (l-30 psi) into test tubes containing 0.01 M potassium phosphate buffer (0.14 NaCl, 0.1% Knox gelatin and 0.01% merthiolate, pH 7.0: RIA assay buffer). The P (GDN 337) antiserum used was generously provided by Dr. Gordon D. Niswender (Colorado State Univ.). [1,2,6,7-H3]-P (114 Wmmol, New England Nuclear), standards, antiserum and samples for all assays were diluted in RIA assay buffer. Separations of antibody bound from free hormone were achieved with dextrancoated charcoal. Total incubation time was 90 min at room temperature, with the labeled hormone added after 30 min. After separation and centrifugation of the samples, the radioactivity in 0.5 ml of the supematant was assessed using a scintillation counter. Diffusion
of Pressure
Ejected
Steroid
Radiolabeled P [1,2,6,7-H3]-P (114 Wmmol, New England Nuclear) was pressure ejected at 4-5 psi for three minutes into lobule VI or VII of the paravermal cerebellum ( 1- 1.5 mm lateral to midline) in the immediate vicinity of a Purkinje neuron. After the 3 minute ejection period, the rat was de-
SMITH, WATERHOUSE $
5A
OHP
O-3
MIN
_
33%
40%
39
Cb
CPP
<-
5A
OHP
3-7
AND WOODWARD
2-i
I.’ ; :
-
-
MIN
RECOVERY
O-3
MIN
432
RECOVERY
3-7
MIN
35x
38
FIG. 5. 5a DHP potentiates Purkinje cell GABA responses. Using the same format as the previous figures, local 5a DHP effects on GABA-induced inhibition are demonstrated. Significant potentiation of the GABA response is seen by 3 min after the onset of continuous pressure ejection of this steroid (2.5 psi, 7 min), with recovery apparent by 7 min after termination of steroid administration. This effect was seen in 13 of 14 cases.
-I-
30
capitated, the brain removed, frozen at -40°C in isopentane and cut in 20 p sections using a cryostat. Sections were exposed to X-ray film for a 2 week period, and the radius of spread of the steroid determined. RESULTS
Quantification oj’ Sex Steroid Release by Pressure und Extent of Spread in Cerebellar Tissue
Ejection
In order to determine whether physiologic levels of P were being administered, release of the steroid was quantified with RIA procedures and extent of spread in cerebellar tissue determined autoradiographically. A range of ejection pressures (l-20 psi) released P in a linear fashion, when expressed as pg steroid ejected over a 3 min period (see Fig. 1A). Only at an ejection pressure of 20 psi did P release deviate from this linear relationship. However, for all experimental procedures, pressures of l-5 psi were used to apply the steroid. Because of the lipophilic nature of P, the extent of spread of the pressure ejected steroid was evaluated in order to estimate the neuronal components with which the steroid might come into contact. As depicted in Fig. lB,
1-
OH
OHP
-
-
OH OHP -
-
O-4
MIN
-
-
4-9 -
7
--
MIN -
-
41%
52% 1
-
FIG. 6. 3p OH-DHP does not alter Purkinje cell responses to either GABA or GLUT. This figure illustrates the inability of 3p OH-DHP to alter neuronal responses of P cells to iontophoretically applied GABA (A) or GLUT (B). In both cases, afier stable amino acid responses were obtained, continuous pressure application of 3p OH-DHP at 2 psi for 7-9 min failed to alter neuronal responses to either neurotransmitter. These results are representative of 13 of 15 cases (A) and 9 of 10 cases (B).
radiolabeled P exhibited a 100 p radius of spread, as determined by autoradiographic techniques. Thus, the pressure ejected steroid (45 psi, 3 min) would encounter no more than 20 Purkinje cells within one folium, with associated neuronal afferents and intemeurons. A rough estimate of the effective concentration of P would be l-3 x IO+ M, well within the physiologic range.
NEURGMODULA~ON
743
BY PROGESTINS
F GLUT RESPONSE
SPONTANEOUS , WGHARGE
”
FIG. 7. Steroid induced alterations in spontaneous discharge, amino acid response and the evoked: spontaneous ratio. Progestin effects on Purkinje cell responses to GABA (A) or GLUT 03) are presented. The percent change (from pre-steroid vafnes) in s~ntaneuus discharge, amino acid response and evoked response refative to spoutaa~ous firing was evafuated for all steroids tested, (*~
charge. No alterations ia background firing were noted after: local steroid administrat~~~~ Recovery was observed by 6-9 min after termination of steroid release. Local application of 3ar OH-DHP resulted in a marked and immediate potentiation of GABA-induced inhibition, which was increased by an average of 70% (n=20). The cell depicted in Fig. 2 illustrates this effect: An initial GABA i~~bition of 4% was increased by SS%, to a 7% inh~bi~on of b~~k~ound firing within seconds after the onset of local application of the steroid. Only a slight recovery from this GABA augmen~tion was noted. A variable recovery rate was noted for all cases tested, with onfy 6@% of the cells demanstrating significant recovery to control levels of GABA inhibition. Insignificant alterations in background discharge were noted during steroid application. 3a OH-DHP also produced a modest but significant suppression of GLUT-evoked excitation (n==121. As exemplified by the cell in Fig. 3, by 3-6 min after the onset of steroid application, a GLUT excitation of 3% was reduced by 3@% to a 25% GLUT-evoked increase above background dis-
The locally applied progestin, 5a DHP, decreased excitatory responses of Purkinje cells to GLUT by 45-W% (n=ZQ), as depicted by the cell in Fig. 4, In this example, 35 nA GLUT increased neuronal firing by 46% above b~ck~ound discharge. Cont~uons ejection of 5or DHP produced a W% suppression of the amino acid response, which was reduced to a 26% excitation, accompanied by only a minor decrease ia spon~~~us discharge, by O-2 min after initiation of steroid application. This locally applied metabolite also potentiated GABA inhibitory responses by an average of 50% (n=14). This effect is illustrated by the ceil presented in Fig. 5. fn this example, CABA inhibited background discharge by 33%.
SMITH, WATERHOUSE
744
Continuous application of 5cr DHP augmented this inhibitory response by 50%, an effect seen 3-7 min after the onset of pressure ejection of the steroid, with no concomitant alteration in the spontaneous tiring rate. After termination of pressure application, however, a small increase in background discharge was noted, which preceded by several min the return of GABA inhibitory responses of the Purkinje cell to pre-steroid levels. 3/3 OH-DHP Either GABA
Dovs Not Alter Parkinje or GLUT
Cell Responses
Steroid Induwd Alterations in Spontaneous Dischargr, Amino Acid Rrsponse and the Evoked:Spc,ntaneous Ratio
In order to determine to what extent observed alterations in amino acid responsiveness were due to changes in background discharge, the average maximal percent change in spontaneous discharge, amino acid response and the evoked: spontaneous ratio (i.e., alterations in amino acid responses relative to changes in background) discharge induced by the steroids from pre-steroid values were determined (Fig. 7). Alterations in the evoked:spontaneous ratio reflect differential actions of a drug on amino acid responses vs. background activity, rather than simple excitatory or inhibitory effects which would be expected to alter spontaneous and evoked discharge to an equivalent degree. A differential effect on evoked spontaneous discharge could be considered a modulatory action of a drug. i.e., indicating direct drug-induced changes in amino acid responses, rather than indirect alterations in this parameter secondary to changes in cell excitability. In Fig. 7, the progesterone meta~lite results are compared with progesterone-induced effects on these parameters. In Fig. 7A. progestin-GABA effects are summarized: In no case did P, ~CXDHP, 3a OH-DHP or 3/3 OH-DHP produce significant alterations in spontaneous discharge. However, only the 3/3 OH-DHP metabolite failed to change GABA responsiveness. 3a! OH-DHP, 5a DHP and P were all able to significantly augment GABA inhibition (by 52, 37 and 21%, respectively), as well as increase the evoked:spontaneous ratio (by 120, 76 and 92%, respectively). 3a OH-DHP was the most potent progestin in terms of alterations in both of these parameters. With regard to GLUT function (Fig. 7B), however, 5a DHP suppressed GLUT excitation to the greatest degree (2%) and increased the evoked:spontaneous ratio by 35%. P and 3crOH-DHP also exerted significant effects on the evoked: spontaneous ratio, with P the most potent in this capacity, yielding a 43% decrease in GLUT evoked:spontaneous changes in firing. Again, in no case did any steroid alter spontaneous firing rate significantly. in Neurotransmitter of P and Metabolites
A .-. p ..-..
I......
‘
38
ot+m+p
3A OH-DHP
to
3p OH-DHP, a progesterone metaboiite with no anesthetic properties, failed to alter neuronal responsiveness to either amino acid neurotransmitter tested. As exemplified in Fig. 6, 7-9 min of continuous local application of this metabolite did not result in augmentation of GABA inhibition (Fig. 6A, n= 15), suppression of GLUT excitation (Fig. 6B, n= 10) or any changes in background discharge (Fig. 6A and B).
Tiww Course Q Alterations Responses: A Comparison
AND WOODWARD
Evoked
An evaluation of the temporal aspects of the neuromodulatory effects of the progestins, including progesterone, also
0
I
(MINI
O-3
3-6
6-9
O-3
9-12
\
/
3-6
12-15
<
STERDID EJECTKIN
/ POST-STEROID
B .-.
p
.--.
SA-DHP
o-3
3-6
..--*
6-9
3A OH-DHP
-‘.” 30 OH-DWP
O-3
9-12 /
\ STEROID EJECTIDN
3-6
12-15
.
/ POST-STEROID
FIG. 8. Time course of alterations in neurotransmitter evoked responses: A comparison of P and metabolites. Steroid-induced alterations (% change from pre-steroid values) in GABA-induced inhibition (A) or GLUT-evoked excitation (B) relative to background are presented for P and various metabolites, So DHP, 3a OH-DHP and 3p OH-DHP. Mean values are plotted for each of the groups at 3 min intervals during and after Iocai steroid application. Only 3/3 OH-DHP failed to significantly alter either parameter. Each point represents the mean of 15-25 cells. (Bars indicate SEM,) *p
revealed differences in terms of steroid effects on amino acid responses. 3o OH-DHP exerted the fastest potentiating effect on GABA inhibition (O-3 min, see Fig. 8A). P and 5a? DHP, however, exerted significant effects on this parameter with a longer latency to onset (3-6 min). Partial recovery of drug effect was observed for all three steroids, with Peffects displaying the most variable recovery rate. As in the previous figure, 3/3
NEUROMODULATION
BY PROGESTINS
OH-DHP had no effects on GABA responses at any point post-injection. An evaluation of steroid effects on GLUT responses demonstrated that, unlike steroid-induced GABA potentiation, attenuation of GLUT excitation was accomplished with the shortest latency by 5a DHP (O-3 min). P and 3a OH-DHP produced somewhat slower effects (3-6 min). Again, recovery from the neuro-modulatory actions of P were the most variable. Surprisingly, 3p OH-DHP produced a short-latency, transient increase in GLUT excitation. DISCUSSION
P-like effects on cerebellar Purkinje cell responses to amino acids, potentiation of GABA inhibition and suppression of GLUT excitation, were demonstrated in the present study for the P metabolites 3a OH-DHP and Sa! DHP. The observed steroid effects were also dependent upon the specific stereoconformation of the steroid, as a third metabolite, 3p OH-DHP (allopregnanolone), did not alter neuronal responsiveness to any degree. The results of this study raise the possibility that specific metabolites of P, and not the steroid itself, may mediate the observed neuromodulatory actions of P on Purkinie cell responses to GABA and/or glutamate (GLUT) 1331.This appears to be a viable possibility for several reasons. First, our previous reports have shown that both intravenous and local application of P result in neuromodulatory effects with a 5-10 minute latency [33]. However, 3o OH-DHP and 5a DHP resulted in immediate and robust augmentation of GABA responses and suppression of GLUT excitation, respectively. Both 3a OH-DHP and 5a DHP, collectively, have been localized to cerebellum in larger amounts than P alone by 10 minutes after IV injection of a physiologic dose of [H13-P, but are reduced to low levels by 30 minutes post-injection [ 171, a time course which closely parallels that observed for effects of systemic P on neurotransmitter responses [33]. Local conversion of P to 5a DHP could be accomplished within minutes, as the enzyme necessary for this process, 5a reductase, has been localized within cerebellar tissue [18]. In addition, infusion of [H13-P into cerebellar tissue in vitro results in accumulation of various P metabolites, including 3a OH-DHP and 5rwDHP [ 131, further suggesting that local conversion of P to these metabolites can be accomplished in this CNS site. Therefore, both IV and local administration of P would result in conversion to 3cr OH-DHP and 5a DHP and eventual localization in cerebellar tissue within the necessary time frame for the observed neuromodulatory actions of this steroid on GABA function. Additional evidence that 3a! OH-DHP may mediate the observed actions of P is provided by recent reports indicating that this metabolite can increase the membrane chloride ion current, either alone or synergistically with GABA, in cultured embryonic rat hippocampal and spinal cord neurons [14,26]. This was found to be an immediate, reversible effect, similar to that described in the present report. In addition, this metabolite was also shown to increase benzodiazepine/GABA binding [26], as has been shown previously for P [25], and stimulate chloride uptake into isolated brain vesicles [26]. Thus, several mechanisms exist by which 3a OH-DHP may mediate the GABA-enhancing effects of P. In contrast, 5a DHP in the present study did not potentiate GABA responses to as great a degree or as rapidly as did 3a OH-DHP. This finding is supported by Harrison et al. [14] who have specified structural re-
145
quirements of steroids able to increase chloride conductance using a voltage clamp paradigm; only 3a OH-DHP, but not 5a DHP, was potent in this regard. Because conversion of 5a DHP to 3a OH-DHP can also occur in the cerebellum, it may be that the observed GABA-potentiating effect of 5a! DHP is in fact due to the partial conversion of this compound to the more potent 3a OH-DHP. The longer latency to effect may in fact be a result of this type of conversion. However, 5~ DHP-induced suppression of GLUT excitation was immediate and prominent, unlike 3o OH-DHP effects on this parameter, which were less extensive and of longer latency. These data suggest that, in contrast to effects on GABA function, P-induced suppression of the GLUT response may be mediated by the 5a DHP metabolite. In addition, the fact that 3p OH-DHP failed to alter GABA responsiveness indicates the stereospecificity of this effect and is also consistent with the report of Harrison er al. [ 141which indicates that any hydroxy group present at C-3 of the steroid molecule must be in the (Yconfiguration in order to permit effects on chloride conductance. It is interesting, however, that this compound was also inactive in terms of GLUT modulation, which could not be predicted based on the differential effect of 3o OH-DHP and 5a DHP on these parameters. Mechanism
of Action
The steroid effects demonstrated in this and our previous studies have too rapid an onset to be genomic actions, and are most likely the result of interactions with membrane constituents. Membrane effects of P have been demonstrated using DPPC liposomes [4,35], erythrocyte membranes [19] and Xenopus faevis oocyte membranes [32]. In the latter, P has been shown specifically to release membrane phospholipid bound Ca++ prior to meiosis [32]. P has also been reported to act at the membrane level of the striatum to release dopamine [8] and at the hypothalamic membrane to release LHRH [20]. In the uterus, P exerts local effects on myometrial excitability [28]. It has been shown that all steroids are able to intercalate to some degree within membrane phospholipids [21,37], dependent to some extent on the presence of aliphatic side chains. However, a remarkable degree of stereospecificity is seen in terms of steroid H bond interaction [3] and anisotropic motion [16], such that an axial (i.e., 3a! OH-DHP) versus an equatorial (i.e., 3/3 OH-DHP) hydroxy group may be one limiting factor for such interactions. It has been speculated that interaction with the “hydrogen belt,” i.e., the carbonyl and hydroxyl groups of constituent membrane lipids, may in fact be the mechanism for anesthetic effects of steroids [3]. It may be that such interactions also contribute towards the neuromodulatory actions of a steroid, as a similar stereospecificity was noted in the present study in terms of steroid-induced alterations in both GABA and GLUT responsiveness. Although P itself has no OH group or aliphatic side chain, a recent study has determined that membranes containing 10% cholesterol are good solvents for P, and such a domain enables P to alter the phospholipid organization of the lipid bilayer [4]. Thus, synaptic membranes, which contain approximately 10% cholesterol [29], would be potential sites for this type of interaction. In addition, although the cerebellum contains only low levels of classic P receptors [24], membrane receptors for the steroid may exist [36], such as those
SMITH, WATERHOUSE
746
which have been reported for EZ 1361. Therefore. the present do not rule out the possibility that P itself may be able to exert direct effects on membrane parameters that lead to the observed neuromodulatory effects on GABA and GLUT funcdata
tion. Additional
mechanisms responsible for the observed neuromodulatory effects of P could involve indirect steroid actions on the cerebellar monoamine systems. Norepinephrine [27] and serotonin [22] have been shown to potentiate GABA responses and diminish GLUT responses of cerebellar Purkinje cells, respectively, in a manner similar to P. It is intriguing to consider the possibility that the various metabolites of P may subserve specific functions at different end organs. Only P exerts progestogenic effects on the uterus [l&38]. However, 5~yDHP, which exerted more potent effects on GLUT function relative to GABA in the present study, is also known to exert potent P-like effects on facilitation of gonadotropin release and lordosis at the level of the hypothalamo-pituitary-gonadal axis [ 18,381. On the other hand, 3a OH-DHP, a more potent GABA enhancing agent than P, as demonstrated in the present study, exerts only limited effects on gonadotropic regulation and reproductive behavior, and these are primarily of an inhibitory nature [38]. In addition, cyclic alterations in activity of the conversion enzymes, 5cu reductase and 3cu hydroxysteroid oxidoreductase, may determine to what extent the various metabolites are active. In general, CNS activity of these enzymes is highest on estrus and proestrus [l8]. Endogenous P levels are elevated on the afternoon of proestrus and again, to a lesser extent, on the evening of diestrus 1. Thus, differential neuromodulatory effects of P vs. the Sa-reduced metabolites may exist which are dependent both on substrate and
AND WOODWARD
enzyme levels at any given stage of the estrous cycle. It is also of interest that the effects of P. 3a OH-DHP and 5a DHP are similar to those reported for benzodiazepine action in terms of potentiation of GABA response [ 111, as well as attenuation of GLUT action [5], both of which have been demonstrated using the cerebellar Purkinje cell model in the intact rat. The behavioral consequences of P administration have also been shown to resemble those of the benzodiazepines in terms of their anxiolytic [7, 12, 15, 301 and anti-convulsant [6,16] properties. Although the GABAenhancing actions of both the progestins and the benzodiazepines is a sufficient mechanism to explain their anticonvulsant effects, potent GABA agonists do not necessarily exert anxiolytic effects [2]. Thus. suppression of excitatory synaptic events may also be a component of anxiolytic action. In summary, the present results suggest that two metabolites of P, 3~yOH-DHP and 5n DHP, may serve to mediate either some or all of the observed neuromodulatory actions of P. In addition, the two metabolites may subserve different functions, 3a OH-DHP exerting greater effects on GABA potentiution and 50 DHP. more pronounced suppression of GLUT excitation.
ACKNOWLEDGEMENTS
Suppol-ted by NS25809 to S.S.S.: NS I8081 to B.D.W.: AA3901, DA02338 and a Biological Human& Grant to D.J.W. The authors would like to express their gratitude to Ms. Kathy Katz and Dr. Sergio R. Ojeda (Dallas) for their generous technical assistance in quantifying pressure ejected release of P using RIA procedures. The authors also thank Dr. Neil Harrison (NINCDS) for his helpful discussions during the course of the work.
REFERENCES
5.
6.
7.
8.
Andrews, A. A., J. P. Advis and S. R. Ojeda. The first proestrus in the female rat: Circulating steroid levels preceeding and accompanying the preovulatory LH surge. Proc, Sot, E.rp Biol Mccl 163: 305-310, 1980. Bartholini, G. Neuropsychopharmacological and therapeutic spectra of GABA receptor agonists. In: Gahn crnd Moc~cl I>i.\ordrrs. E.upcrimcntal cud Clinicul Rcscurch. edited by G. Bartholini, K. G. Lloyd and P. L. Morselli. New York: Raven Press, 1986, pp. l-7. Brockerhoff, H. Anesthetics may restructure the hydrogen belts of membranes. Lipidc 17: 100-1003, 1982. Carlson, J. C., M. Y. Gruber and J. E. Thompson. A study of the interaction between progesterone and membrane lipids. E/ldocrino/og.v 113: 19%194, 1983. Cheng, J. T. and B. D. Waterhouse. Interactions of flurazepam with glutamate evoked excitatory responses of cerebellar Purkinje neurons. Sot, Nrurosc? Ahsfr 9: 398, 1983. Craig, C. R. Anticonvulsant activity of steroids: separability of anticonvulsant from hormonal effects. J Phrrrmtud Exp Thor 153: 337-343, 1966. Dennerstein, L., C. Spencer-Gardner, G. Gotts, J. B. Brown, M. A. Smith and G. D. Burrows. Progesterone and the premenstrual syndrome: a double blind crossover trial. Br Mrcl J 290: 1617-1621, 1985. Dluzen, D. E. and V. D. Ramirez. Bimodal effect of progesterone on in vitro dopamine function of the rat corpus striatum. Neuroentloc.rirro/l~~~ 39: 149-155, 1984.
9. Eccles, J. C., M. Ito and J. Szentdgothai. 7’lrc (‘c,rc,hc//r//,r (I\ (I Ncuromrl Muclri~c,. New York: Springer-Verlag, 1967. IO. Geller. H. M. and D. J. Woodward. An improved constant current source for microiontophoretic drug application studies. Elec,rroc,nc.c~phohgr Clin Ncurophysiol 33: 43&432. 1972. II Geller, H. M., D. A. Taylor and B. J. Hoffer. Benzodiazepines and central inhibitory mechanisms. Norrrlyn Schmiet/rhrr,gt Arch Phrrrmtlud 304: 8 l-88. 1978. 12. Click, I. D. and S. Bennett. P\ychlatric effecta of progeztel-one and oral contraceptive\. In: P~ychitrrric Complictrtkms oj’Mrdictrl Dru,yc. edited by R. Shader. New York: Raven Press, 1972, pp. 295-332. I3 Hanukoglu, l., H. J. Karavolas and R. W. Goy. Progesterone metabolism in the pineal. brain stem, thalamus and corpus callosum of the female rat. BTN~‘/zRc.t 125: 313-324, 1977. 14 Harrison. N. I-., M. D. Majcw\ko. J. W. Harrington and J. 1.. Barkel-. Structure-activity relationships for steroid interaction with a gamma-aminobutyric acid, receptor complex. .I P/~tr,.nroc,r,/ ,5-\-p T/I<,/. 241: 346. 1987. 15. Herrmann, W. M. and R. C. Beach. Experimental and clinical data indicating the psychotropic properties of progestogens. Postgrd Md J 54: 82-87, 1978. 16. Holmes, G. L. and D. A. Weber. The effect of progesterone on kindling: A developmental study. &,\I Brcrin Res 16: 45-53. 1984.
NEUROMODULATION
BY PROGESTINS
17. Karavolas, H. J., D. R. Hodges, D. J. O’Brien and K. M. MacKenzie. In vivo uptake of [H3]-progesterone and [H3]5 alphadihydroprogesterone by rat brain and pituitary and effects of estradiol and time: Tissue concentration of progesterone itself or specific metabolites? Endocrino1og.v 104: 14181425, 1979. 18. Karavolas, H. J., P. J. Bet-tics, D. Hodges and N. Rudie. Progesterone processing by neuroendocrine structures. In: Metabolism of Hormonal Steroids in the Neurorndocrine Structures, edited by F. Celotti, F. Naftolin and L. Martini. New
York: Raven Press, 1984, pp. 149-170. 19. Kaya, H. and T. Saito. Effect of progesterone and its 17 alphahydroxyl derivative on human erythrocyte membrane. Jptz J Pharmacol
39: 299-306,
1985.
20. Ke, F. C. and V. D. Ramirez.
The plasma membrane of rat hypothalamic tissue is the primary site for progesterone stimulatory action of LHRH release. Sot Neurosci Ahstr 12:
604, 1986. 21. Lawrence, D. K. and E. W. Gill. Structurally
22.
23.
24.
25.
specific effects of some steroid anesthetics on spin-labeled liposomes. Mel Pharmacol 11: 280-286. 1975. Lee, M., J. C. Strahlendorfand H. K. Strahlendorf. Modulatory action of serotonin on glutamate-induced excitation of cerebellar Purkinje cells. Brain Res 361: 107-I 13, 1985. McCaman, R. E., D. G. McKenna and J. K. Ono. A pressure system for intracellular and extracellular ejections of picoliter volumes. Brain Rcs 136: 141-147, 1977. MacLuskey, N. J. and B. S. McEwen. Progestin receptors in rat brain: distribution and properties of cytoplasmic progestinbinding sites. Endoc~rinology 106: 192-202, 1980. Maggi, A. and J. Perez. Progesterone and estrogens in rat brain: Modulation of GABA receptor activity. Eur J Pharmacol 103: 165-168,
1984.
26. Majewska, M. D., N. L. Harrison, R. D. Schwartz, J. L. Barker and S. M. Paul. Steroid hormone metabolites are barbituratelike modulators of the GABA receptor. Science 232: 10041007, 1986. 27. Moises, H. C., D. J. Woodward, B. J. Hoffer and R. Freedman. Interactions of norepinephrine with Purkinje cell responses to putative amino acid neurotransmitters applied by microiontophoresis. Esp Nrurol 64: 493-515, 1979.
747 28. Osa, T. and T. Ogasawara. Effects of in vitro progesterone and estradiol-17/3 on the contractile and electrical responses in rat myometrium. Jpn J Physiol 34: 427-441, 1984. 29. Robertson, R. N. The Lively Membranes. New York: Cambridge University Press, 1980, pp. 152-153. 30. Rodriguez-Sierra, J. F., J. L. Howard, G. T. Pollard and S. E. Hendrickes. Effect of ovarian hormones on conflict behavior. Psyc.hopharmac.ol~~~y (Berlin) 9: 29>300.
, 1984. __-..
31. Palmer; M. R., S. M. Wuerthele and B. J. Hoffer. Physical and physiological characteristics of micropressure ejection of drugs from multibarreled pipettes. Neuropharmac,o/ogy 19: 931-938, 1980.
S., M. Schorderet and E. E. Baulieu. 32. Schorderet-Slatkine, Progesterone-induced meiotic reinitiation in vitro in Xenopus levis oocytes. D$ferentiation 9: 76, 1977. 33. Smith, S. S., B. D. Waterhouse, J. K. Chapin and D. J. Woodward. Progesterone alters GABA and glutamate responsiveness: A possible mechanism for its anxiolytic effect. Brain Rc.c 400: 353-359,
1987.
34. Smith, S. S.. B. D. Waterhouse and D. J. Woodward. Locally applied sex steroids alter neuronal responsiveness to GABA and glutamate in the cerebellum. Sot Neurosci Abstr 12: 990, 1986. 35. Sze, P. Y., E. C. Armstrong, G. G. Deanin and M. W. Gordon. Steroid hormones stabilize liposomal membrane from alteration by tubulin. Sot Nrurosci Abstr 11: 1135, 1985. 36. Towle, A. C. and P. Y. Sze. Steroid binding to synaptic plasma membrane: Differential binding of glucocorticoids and gonadal steroids. J Steroid Biochem 18: 135-143, 1983. 37. Trauble, H. and E. Sackmann. Studies of the crystalline liquid crystalline phase transmition lipid model membranes. III. Structure of a steroid-lecithin system below and above the lipid-phase ._ .-. transition. J Am Chum Sot 94: 4492-4498._, 1972. 38. Zanisi, M., E. Messi and L. Martini. Physiological role of the 5 a-reduced metabolites of progesterone. In: Metabolism of Hormonal Steroids in the Ncuroendocrinc Structures. edited by F. Celotti, F. Naftolin and L. Martini. New York: Raven Press, 1984, pp. 171-183.