Effects of the opiates on the paraventricular nucleus in genetically polydipsic mice

23

Brain Research, 598 (1992) 23-32 © 1992 Elsevier Science Publishers B.V. All rights reserved 0006-8993/92/$05.00

BRES 18317

Effects of the opiates on the paraventricular nucleus in genetically polydipsic mice Itsugi N a g a t o m o *, T o s h i h i k o K a t a f u c h i

**

and Kiyomi Koizumi

Department of Physiology, State University of New York, Health Science Center at Brooklyn, N Y 11203 (USA) (Accepted 7 July 1992)

Key words: Paraventricular nucleus; Polydipsia; Drinking; Morphine; Dynorphin; Naloxone; Nor-binaltorphimine; Brain slice

The inbred mice, STR/N, are known to possess extreme polydipsia with no known abnormality in vasopressin system and the kidney function. Our previous studies indicate that the opiate antagonists given intracerebroventricularly strongly attenuated spontaneous drinking. To determine the site(s) of action, the present study was undertaken. Microinjections of naltrexone methobromide and a selective r-receptor antagonist, nor-binaltorphimine (nor-BNI), into the paraventricular niacleus of the hypothalamus (PVN) greatly attenuated drinking of the S T R / N for 0.5 to 16 h after injections, while in the two control groups, non-polydipsic STR/1N and Swiss/Webster strains, drinking was not affected by the injections. Food intake was not much altered in all groups. Studies of PVN neurons in vitro (n = > 160 for each group) showed that basal firing rates and patterns were similar in the S T R / N and the control groups. Morphine added to the medium inhibited some but excited none in all strains tested. The threshold for the inhibitory action was higher in the polydipsic S T R / N mice (10 -8 M), compared to that in the control, S / W mice (10 -9 M). Further, a proportion of neurons inhibited by morphine in the PVN was significantly smaller ( P < 0.01) in the S T R / N (41.7%), compared to that in the control (64.9%). Dynorphin had very similar effect to that of morphine, but the proportion of cells inhibited was 25.4% in the STR/N, and 70.4% in the S/W. Prior applications of naloxone to the medium prevented the action of both morphine and dynorphin. Under the synaptic blockade (in a low Ca 2+ and high Mg 2+ medium) the inhibitory effect of the opiates persisted. We concluded that the PVN is at least one of the possible sites where the opiates are acting to cause the polydipsia in the S T R / N mice.

INTRODUCTION

In the late 1950's the extremely polydipsic inbred strain of mouse, STR/N, was discovered 6. The early studies by Silverstein and his associates 33'34 revealed that, though the polydipsia was always accompanied by polyuria with low electrolyte concentrations, under water restriction the mice could produce concentrated urine and responded well to exogenously applied pitressin. Moreover, plasma osmolarity and ionic concentrations of the polydipsic mice were normal; hypothalamic supraoptic and paraventricular nuclei contained abundant neurosecretory granules, and no gross abnormality was detected in several tissues including the salivary glands, the kidney and the adrenal. These authors concluded that the polydipsia was not neces-

sary for their survival but may be due to some brain abnormality causing the innate thirst. In the hope of elucidating brain mechanisms for the polydipsia, we began our studies on the S T R / N strain of m i c e 14'15'17-19'24. We have found that, though the inbred mice drink 5 to 8 times as much water per day as control mice, the diurnal pattern of drinking is preserved 17-19. Moreover, they respond normally to dehydration t8'19. Administration of angiotensin II inhibitors (captopril or saralasin) mildly reduced the polydipsia is, yet responses of neurons in the anteroventral region of the third ventricle (AV3V) and the subfornical organ to angiotensin II or other agents did not reveal a clear abnormality that could be attributed to factors that explain the excessive drinking in the S T R / N mice is. Recently we have found that the poly-

Correspondence: K. Koizumi, Department of Physiology, Box 31, SUNY Health Science Center at Brooklyn, 450 Clarkson Avenue, Brooklyn, NY 11203, USA. * Present address: Department of Neuropsychiatry, Kagoshima University, Faculty of Medicine, Kagoshima, Japan. ** Present address: Department of Physiology, Kyushu University, Faculty of Medicine, Fukuoka, Japan.

24 dipsia can be greatly attenuated by intracerebroventricular injections of the opiate antagonists, particularly the K-receptor antagonist, nor-binaltorphimine (norBNI), and its effect is very selective for suppressing excessive drinking in the S T R / N strain only without affecting normal drinking in non-polydipsic mice 17. In search for the site and the mechanism of action of the opiate antagonists on the polydipsia, we selected to study a possible role played by the paraventricular nucleus of the hypothalamus (PVN), because the nucleus is considered to be one of the important brain regions where the opiates act to alter feeding and drinking behavior. We approached this problem by studying (1) changes in drinking (and feeding) after injections of opiate antagonists locally into the PVN, and (2) effects of opiate agonists and antagonist on the PVN neuron activity in vitro.

10% formalin for a few days. Frozen sections of 4 0 / , m in thickness were prepared, Nissl stained and examined histologically.

Brain slice preparations for neuron studies The polydipsic mice, S T R / N , and their control, S / W mice, were stunned and decapitated. The brain was quickly removed, cooled in the perfusion medium at 4°C for 1 min, then placed on the cold plate. The meninges were removed under microscopic observation and the tissue was blocked with a razor blade. O n e or two coronal slices of 350 /zm in thickness containing the PVN were cut with a vibratome. Immediately after sectioning, slices were placed in the perfusion medium at least for 1 h, oxygenated with a mixture of 95% 0 2 and 5% CO 2 at room temperature. The perfusion medium was a modified Y a m a m o t o ' s solution 16 containing (in mM), NaCI 124, KCI 5, K H 2 P O 4 1.24, MgSO 4 1.3, CaCl 2 2.1, N a H C O 3 20, and glucose 10, with pH 7.3-7.5. In the perfusion medium Ca 2+ concentration was always lowered to 0.75 mM to increase spontaneous activity of PVN neurons 1<2s'26. When required, a low-Ca 2 + (0.5 mM) and high Mg 2~ (9 mM) solution was used to block synaptic transmission ~6. The pH of the perfusion medium was not affected by adding drugs at the concentrations used.

Recordings from the PVN neurons in dtro MATERIALS

AND METHODS

Animals The polydipsic strain of mice, S T R / N , and two control strains, S T R / 1 N , and S w i s s / W e b s t e r ( S / W ) , were used. The non-polydipsic S T R / 1 N strain is genetically closely related to the S T R / N 34. T h e animals of both sexes and 6-12 m o n t h s of age were housed in individual cages in a room maintained at 22-24°C and 12/12 light/dark cycle (light on at 7.00 h). Tap water and food (laboratory chow) were given ad libitum. W a t e r and food intake and body weight were measured once (at 10.00 h) or twice (10.00 and 18.00 h) daily. Water intake was measured by weighing a small drinking tube fitted with a stainless steel spout. The body weight of animals ranged between 25 and 40 g and differences between the three strains of mice were not significant.

Central administration of opiate antagonists A stainless steel intracerebral guide cannula (25-gauge) was cut to an appropriate length and its tip was sharpened at 40 ° angle. Mice were anesthetized with 50 m g / k g pentobarbital sodium injected intraperitoneally. With the aid of a stereotaxic instrument, the guide cannula was inserted to a depth of 1.0 m m above the right PVN and secured in the position by fixing it to the skull with dental cement. A stainless steel wire was inserted to prevent any obstruction of the cannula and a protective polyethylene cap was placed on the top. The stereotaxic coordinates used were: A 3.0 (0.8 m m rostal to the bregma), L 0.25, H 1.3 (4.0 m m from the cortical surface), according to the atlas of M o n t e m u r o and Dukelow =. Mice were allowed to recover from the operation for 1 week during which period water and food intake was measured. The opiate antagonists were administered at the beginning of dark period, between 18.00 and 19.00 h. A n injector cannula (32gauge) attached to a microsyringe through polyethylene tubing (PE 10) was inserted to the PVN through a guide cannula. While an animal was gently restrained by the experimenter's hands, the amount of 0.1 p.l was injected within 2 s. The whole procedure took no more than 30 s. Animals were then returned to the home cage and drinking volume was measured at 0.5, 1.0, 1.5, 2, 3, 6 and 16 h after the injection. The amount of food intake after the injection until the next morning (16 h) was also measured. Some animals received two doses of the opiate antagonist with the interval at least of 1 week. Histological verification of the injection site was performed at the end of each experiment by injecting 0.1 /~1 Chicago Sky blue dye through the cannula. Mice were anesthetized with pentobarbital sodium and perfused intracardially with isotonic saline containing heparin followed by 10% formalin. Brains were excised and fixed in

Before recordings were done, each slice was carefully trimmed with a microsurgery knife to an area of less than 2 × 2 mm. The trimmed slice was placed on a Sylgard mat glued to the bottom of the recording chamber which had a volume of 0.8 ml and was held in place with a nylon net and platinum weights. The recording chamber was perfused with the medium oxygenated with 95% 0 2 and 5% CO 2 at a rate of 1 - 2 m l / m i n at 36_+_+0.5°C. The entire perfusate in the chamber could be completely exchanged within 1 min. The perfusate containing a drug at a desired concentration was given to the slice from separate bottles. Extracellular recordings were obtained from single neurons within the PVN using glass micropipettes filled with 0.5 M sodium acetate containing 2% Chicago Sky blue (Sigma). Their DC resistance was 20-35 MgL The PVN was readily distinguished from the surrounding structures under a microscope with a transmitted light. At the end of recordings, a current of 5 ~zA was passed through the electrode for 3 - 5 min (tip negative) to deposit blue dye. Tissues were fixed in 2% glutaraldehyde, frozen sectioned at 4 0 / z M and stained with Neutral red or Cresyl violet. The recorded sites were determined microscopically. Recorded action potentials were amplified and displayed on an oscilloscope and stored on magnetic tapes for further analyses. A window discriminator and an integrator were used for the continuous observation of the firing patterns of neurons which were displayed on a polygraph. After obtaining a stable recording from a single neuron for at least 10 min, a drug was added to the perfusion medium. Changes from the resting level (a m e a n firing rate during 5 min) to the peak or the trough (a mean firing rate during l-min period) following an application of a drug was calculated from the rate-meter records.

Drugs employed The opiate antagonists used for i.c.v. (intracerebroventricular) injections were dissolved in artificial cerebrospinal fluid (ACSF) whose composition was (in mM): Na ÷ 144, C I - 133, K ÷ 4, Ca 2÷ 2.1, Mg 2÷ 1.3, HCO~- 20, phosphate 1 and glucose 5. Naltrexone methobromide (kindly supplied by Boehringer and Ingelheim, Ridgefield, CT) at doses of 0.2, 0.5 and 1.0/zg in 0.1 /zl ACSF was used. This particular compound was chosen, rather than naltrexone hydrochloride, because this is a quaternary analog of naltrexone and is poorly lipophilic, thus it should remain at the site of injection for a longer period than other antagonists 3"43. Another drug used was the selective K-type receptor antagonist, nor-binaltorphimine (nor-BNI, Research Biochemicals, Nattick, MA) 28'29 at doses of 0.5, 1.0 and 2.5 /.~g dissolved in 0.1 tzl ACSF. Since the molecular weight of nor-BNI is twice as large as that of naltrexone, a double dose was used.

25

D

A

Opiate agonists and antagonist used for brain slice preparations were: morphine sulphate (Merck), dynorphin A (1-13) (Penninsula) and naloxone hydrochloride (Sigma).

STR/N

Data collection and analyses In experiments measuring water and food intake in response to injected opiate antagonists, all measured values were converted to ml and g consumed per 30 g body weight (BW). This unit was chosen rather than per kg body weight, because the weight of mice ranged between 25 and 40 g and values per 30 g were more closely associated with actual food or water consumed. Two and one-way ANOVA with repeated measures followed by Newman-Keuls' and Duncan's method were used for assessing statistically significant changes. In vitro studies of the PVN neurons, changes in firing rates (spikes/s) were compared before and after a drug administration, as described above. Neurons were regarded as having been excited or inhibited if an increase or a decrease of more than 20% in their firing rates followed a drug application.

Naltrex0ne ~

10

Effects of microinjections of naltrexone methobromide into the PI/N Injections of opiate antagonist, naltrexone methobromide, to the PVN attenuated the polydipsia. Cumulative water consumptions (n = 6 in each group) after each dose of the drug (0.2, 0.5 and 1.0/~g) are shown in Fig. 1A-C. A two-way ANOVA test indicated that microinjections of naltrexone significantly (F = 17.06, df 3/96, P < 0.001) reduced water intake of the polydipsic S T R / N mice (Fig. 1A). A multiple-range test further revealed that changes after naltrexone at all doses differed from that after ACSF for 0.5 through 6 h (0.2 and 0.5/zg, P < 0.01 at 0.5 and 6 h; P < 0.05 at 1 and 2 h; 1.0 Izg, P < 0.05 at 0.5, 1 and 2 h; P < 0.01 at 6 h). There was no significant difference among three doses of naltrexone for 0.5 through 6 h after injections. In the non-polydipsic STR/1N mice (Fig. 1B) the significant difference in drinking was indicated by an ANOVA test (F - 7.13, df 3/97, P < 0.001), and Newman-Keuls' multiple-range test showed a difference between the vehicle and 0.5 (P < 0.05) and 1.0 /zg (P < 0.01) of naltrexone at 6 h only. The S / W mice (Fig. 1C) showed no significant reduction in water intake after naltrexone. Fig. 1D,E,F also shows that in the S T R / N naltrexone at all doses significantly (F = 9.42; df 3/16; P < 0.001) decreased overnight (16 h) water intake (all doses, P < 0.01), while the drug had no significant effect in both the STR/1N and S/W. Overnight food intake was not affected by naltrexone methobromide at doses of 0.2 and 0.5 /zg in all three strains of mice (Fig. 3A). At the highest dose (1.0 ~g) of naltrexone food intake was reduced in the S T R / N (F = 7.82, df 3/16, P < 0.001) and STR/1N (F = 6.02, df 3/16, P < 0.001), compared to that after ACSF administration. The effect of 1.0 ~g differed also from that produced by 0.2 and 0.5/.~g.

ACSF

8

28

24 20

16

6

4 2 1

2

3

4

5

6

< ~. ~. o. O

B

O

~

E

¢D

RESULTS

STR/N

12.

/1N

STR/1N

El) 0 ("3

6~ 5-

ACSF .2 pg

E

5 ~g

0 /Jg

r-

1

2

3

4

5

4. 32~

1.

6

C

F

S/W

3

4ACSF

2 1

S/W

f

1

2

3

4

hours

5

0.2 ,ug 1.0 /Jg

32-

0.5 )Jg

1-

6 O

¢q

¢t)

O

¢; ,5 ,-: *p<0.05; **p<0.01

(ANOVA followed by Newman-Keuls test) Fig. 1. Effects of injection of naltrexone methobromide into the PVN at the beginning of dark period on water intake. A - C : cumulative water intake (mean + S.E.M.) for 6 h after injection at various doses in the polydipsic, S T R / N , and non-polydipsie controls, S T R / 1 N and

S / W (n = 6 for each group). A two- and one-way ANOVA followed by Newman Keuls' test. * P < 0.05; ** P < 0.01. D-F: overnight (16 h) water intake. Note differences in scales (ordinates of the top and two lower graphs) in this and the next figure. Significant reduction in water intake occurred after all doses of naltrexone only in S T R / N mice. ACSF, artificial cerebrospinal fluid.

Effects of microinjections of nor-BNI into the PI/N Fig. 2 shows cumulative water consumption after injections of nor-BNI to 6 each of the three groups of mice. Two-way ANOVA indicated that in the polydipsic S T R / N mice (Fig. 2A) microinjections of nor-BNI reduced water intake for 6 h following the injection (P = 17.50, df 3/96, P < 0.001). A multiple-range test showed that, though the effects of 0.5 and 1.0/zg did

26

D

A STR/N 10

/

nor-BNI

STR/N ACSF

22

8

14

6

10 6 2 1

2

3

4

5

6

~ m o m

E

B STR/1N 03 O co

4

"E 3 ~. t~

2

E

~

STR/1N

~ 1

2

ACSF 0 . 5 ug 1,0 ug 2 . 5 ug

4 3 2 1

3

4

5

6

c; -c ~,i

C

F S/W

ACSF j 0 . 5 ug ~ 1.0 ug

~ 1

S/W

2.5

2

3 4 hours

5

6

ug

3 2 1

In non-polydipsic S T R / 1 N and S / W strains, a multiple-range test showed no significant effect at all doses. A two-way A N O V A test showed the significant difference in 0 - 6 h water intake ( S T R / 1 N , F = 6.17, df 3/96, P < 0.001; S / W , F = 4.31, df 3/96, P < 0.01), but this is due to the fact that cumulative water intake increased with time regardless of the dosage of the drug injected. An overnight (16 h) water intake after nor-BNI administration to the control mice (Fig. 2E,F) was not altered at any dose tested. No significant effect was observed on overnight food intake after 0.5 and 2.5 /zg dose of nor-BNI, except a small reduction observed after 2.5 /zg only in the S T R / 1 N (Fig. 3B). To evaluate the strain and dose interaction, the 6 h cumulative water intake after naltrexone or nor-BNI was submitted to a 3 (strains)x 4 (doses) repeatedmeasures A N O V A (Fig. 4A,B). There was an overall difference between strains, with the S T R / N drinking more than the two control strains following injections of naltrexone ( F = 34.59, df 2/48, P < 0.001) or norBNI ( F = 47.59, df 2/48, P < 0.001). A multiple-range test indicated that water consumption of the S T R / N was higher than that of control groups at all doses except at 1.0/zg naltrexone and 2.5/zg nor-BNI. There was an overall dose effect (nalterexone, F = 10.33, df 3/48, P <0.001; nor-BNI, F = 8.90, df 3/48, P < 0.001). However, neither naltrexone nor nor-BNI administered to the control groups produced significant difference in water intake at all doses. There was also a strain × dose effect (naltrexone, F = 4.90, df 6/48, P < 0.005; nor-BN1, F = 4.47, df 6/48, P < 0.005).

~.=., ~ -. _, ~ o ~ e,i *P" 0 05, * * p < 0 01

~ANOVA followed by Newman-Keuls test1

Fig. 2. Effects of nor-BNl (?,elective K-receptor antagonist) injected into the PVN at the beginning of dark period. A-C: cumulative water intake for 6 h; D-F: overnight (16 h) drinking in the polydipsic (STR/N) and control (STR/IN and S/W) mice (n=6 for each group). Mean + S.E.M. and two- and one-way ANOVA followed by Newman-Keuls' test. Significant reduction in water intake was observed at 2.5/zg for 2-6 h, and at all doses for 16 h drinking in the STR/N. No effect was seen in control groups. ACSF, artificial cerebrospinal fluid. not differ from the vehicle, a dose of 2.5 /xg significantly reduced water intake compared to that after ACSF for 2 through 6 h (2 h, P < 0.05; 3 and 6 h, P < 0.01). Further, the effect of 2.5 /zg differed from that of 0.5 /.tg in the cumulative 3 and 6 h drinking ( P < 0.05). For overnight water intake (16 h) in the S T R / N (Fig. 2D), significant attenuating effects ( F = 12.35, df 3/16, P < 0.001) were produced by all doses of nor-BNI (0.5 and 1.0/zg, P < 0.05; 2.5/xg; P < 0.01).

Effects of morphine on the PVN neurons One hundred and sixty-one spontaneously active PVN neurons of the S T R / N , and 165 of the S / W were recorded extracellularly. Though neurons of the S T R / 1 N were not included in control groups due to paucity of the particular strain of animals, our results on some S T R / 1 N neurons indicated that their responses were similar to those of the S / W . In the polydipsic strain ( S T R / N ) , 7 neurons (4.4%) showed a phasic patterns of firing, while 53 (32.9%) showed a pattern classified as continuous firing, and 101 (62.7%) an irregular firing patterns. The mean firing rates (mean + S.D.) for each group of neurons were: 3.2 +_ 0.7, 5.3 +_ 2.7 and 5.2 + 4.1 spikes/s, respectively. In the control S / W mice, 16 (9.7%) of the PVN neurons showed a phasic pattern, 60 (36.4%) continuous and 89 (53.9%) irregular pattern of firing. Mean firing rates were: 3.8 _+ 2.5, 6.1 _+ 3.0 and 4.9 + 2.7 spikes/s, respectively.

27 After making a stable recording from a single cell for at least 10 rain, morphine was applied to the slice at a concentration of 10 -6 M. As described previously, cells were classified as having been excited or inhibited if their firing rates after an application of morphine were changed by more than 20%. The dominant effect of morphine to PVN neurons was inhibitory in both polydipsic and control groups. In most cells the inhibitory response occurred within 3 min following the drug perfusion and reached its peak in 3 - 4 min. The magnitude of the effect was not dependent on the

A

B

11

STR/N 4

3 ¸

2

1 c~ c5

to c;

c) "~

d

Naltrexone

nor-BNI

--iil

4~

4-

rn O t'9

2-

STR/1N

Effects of dynorphin on the PVN neurons

3-

**

2-

1-

c
~

,,r,

d

{q

u~ o

m e,i

5-

4-

S/W

4-

2-

2-

l-

1-

L) <

¢q (5

~ (5

o "=

initial firing rate of a cell. After morphine perfusion was switched to normal perfusate, the firing rate returned to pre-treatment values in 5 - 2 0 rain (Fig. 5A,B). The inhibitory effect of morphine was reversible and tachyphylaxis was seldom observed with repeated applications of morphine, provided that a long interval (30 min) was used between the test. No neuron in either strain of mice was excited by morphine. The proportion of the inhibited neurons in the PVN differed between the two strains tested: in the S T R / N 43 out of 103 (41.7%) were inhibited, while in the S / W 72 out of 111 (64.9%) were inhibited. The difference was significant at P < 0.01 level. To investigate the d o s e - r e s p o n s e relationship, different concentrations of morphine, ranging from 10 -9 to 10 -5 M, were applied to the same neurons. D o s e response curves were constructed by plotting percentage changes in firing rate from the resting value against morphine concentrations. Fig. 5D shows averaged percentage decreases in firing rates taken from 2 to 20 neurons of the S T R / N and the S / W mice. Firing rates decreased with an increase in morphine concentrations in a dose-dependent manner. The threshold concentration of morphine was approximately 10 -8 M in the S T R / N and 10 -9 M in the S / W , indicating that the responsiveness to morphine was lower in the polydipsic mice than in the control.

<

*p<0.05, **p<0.01 (ANOVA followed by N e w m a n - K e u l ' s test)

Fig. 3. Effects of opiate antagonists, naltrexone methobromide (A) and nor-BNI (B) injections into the PVN at the beginning of dark period on overnight (16 h) food intake in the polydipsic, STR/N, (top) and non-polydipsic,STR/1N, (middle) and S/W (bottom) mice (n = 6 for each group). Only the highest dose had some effects on food intake. A one-way ANOVA followed by Newman-Keuls' multiple-range test.

In addition to morphine, effects of the endogenous K-receptor agonist in the brain, dynorphin A(1-13), on the PVN neurons were examined, since the effect of dynorphin on the AV3V neurons in the S T R / N was found to be quite different from that in the controls 14. Dynorphin at 10 - 6 M in the bathing medium inhibited 15 out of 58 neurons (25.4%) in the S T R / N strain, while in the control, S / W mice, 38 out of 54 neurons (70.4%) were inhibited. No PVN neuron in either strain of mice was excited by dynorphin. Thus, the proportion of inhibited neurons in the S / W was significantly higher than that of the S T R / N ( P < 0.01). The latency and the time course of dynorphin's effects were similar to those of morphine. Although d o s e - r e s p o n s e relationship was obtained successfuly only in 4 neurons for each strain, the curves indicated that the threshold for the inhibitory action was higher (10 -8 M) in the S T R / N , compared to that in the S / W , (10 -9 M). Again, the inhibitory effect of dynorphin was reversible and tachyphylaxis was not observed.

Effects of naloxone on morphine and dynorphin action Naloxone is known to be a specific opioid antagonist. Its action on morphine- or dynorphin-induced

28

"STR/IN • STR/N OS/W

~ 1 ~ C~ O (,~

10

JD

10

O) O

E

E

4) *~

.~

E ~

P<0.01

4==k

o. 2 o. 5 1. 0 (~¢) Naltrexone concentration (ANOVA followed by Duncan's test)

0

0

*P<0.05, **P<0.01

P<0.01

4==1=

0. 5

1. 5

2. 5 (t~o)

nor-BNI concentration (ANOVA followed by Duncan's test)

**P
Fig. 4. Effects of different dosage of naltrexone (A) and nor-BNl (B) on the 6-h cumulative water intake of the three strains of mice. 3 (strains) × 4 (doses) repeated-measures ANOVA followed by multiple-range test. e, STR/N; , , STR/1N; II, S/W.

inhibition of the PVN neuron activity was examined. Examples taken from the S T R / N and S / W mice (Fig. 6A,B) show that naloxone (at 10 -6 M) reversiby blocked dynorphin-induced (at 10 -6 M) inhibition of PVN neurons. Similar blocking effect of naloxone on morphineinduced inhibition was also observed in both strains of mice. We did not find any difference in strength of blocking action of naloxone between the two strains of mice. An application of naloxone alone had no effect on the activity of PVN neurons.

Effects of morphine and dynorphin after synaptic blockade of PVN neurons In order to find whether or not morphine or dynorphin is acting directly on PVN neurons, we used a technique to produce synaptic blockade commonly employed 16'25"26. Under a low Ca 2+ (0.5 mM) and high Mg 2+ (9 mM) medium which was shown to block synaptic drive almost completely ~6, the inhibitory action of either morphine or dynorphin could still be demonstrated, even though the spontaneous activity of PVN neurons of either strain of mice had been strongly reduced in the medium. No difference between the S T R / N and S / W mice was detected in this case (Fig. 6C,D). These results suggest that the inhibitory action of the opiates is directly on these neurons, rather than through interneurons situated in the neighboring area. DISCUSSION In search of the mechanisms of the polydipsia of inbred S T R / N mice, we have found that excessive drinking can be attenuated mildly by angiotensin II inhibitors (captoprii and saralasin) and strongly by the opiate antagonists (naltrexone and nor-BNI) adminis-

tered subcutaneously or intracerebroventricularly (i.c.v.) 17-19. Their effect on drinking was observed only in the polydipsic mice, without affecting control mice at dosage used in our studies. Our aim here was to find a possible site or sites in the brain where the opiate antagonists exert the attenuating action on the polydipsia. Certain areas in the hypothalamus having a high density of opioid peptides and receptors have been shown to play a role in food and water intake 7"9'35. As shown in our present results, naltrexone and nor-BNI applied to the PVN both attenuated drinking in the polydipsic mice, S T R / N , in a manner very similar to what we observed after the i.c.v, injections of the same drugs, except that a smaller dose was sufficent in this case. The effective dose of naltrexone to reduce spontaneous drinking of a satiated mouse was 2.65 nM (1.0 gg) when given i.c.vJ 7, while 0.46 nM (0.2 /xg) when injected into the PVN, nearly one fifth of that needed by i.c.v, injection. Similarly, the selective K-receptor type antagonist, nor-BNI, needed to attenuate drinking was 1.24 nM (1.0 /xg) for i.c.v, injection JT, but it was 0.62 nM (0.5 /xg) when injected into the PVN. In this case the ratio was one half, probably because nor-BNI was thought to diffuse more slowly to the site of the opioid receptors 39 than other antagonists, like naloxone. These comparisons are interesting, yet they do not necessarily indicate that the opiate antagonists' action at the PVN is more effective than at other sites where the i.c.v, administered drugs reach, because the chemicals injected into the lateral ventricle spread to a much wider area compared to that injected locally into the brain tissue. At this stage we wish only to suggest that the PVN is at least one of the sites where the opiate antagonists have an effect on the polydipsia.

29 A

STR/N

Morphine 10 - 5 M

~

OL

10 min

~

Morphine 1 0 - 6 M

1Or ~ L a - 111v ....... ..... " I p - ,• pl¶,ir,T

oL

10 min

L ""

""'

, ' , ~ ' ~ ' . ~

5 ~,n

Z ............

Morphine 10-7 M

B

S/W a

b

15 rnln

Morphine 10 - ° M m -~¢ 5

.

c

• |..~,

o~

Morphine 1 0 - 7 M

5 rain

Morphine 1 0 - 8 M

5 rain 5 rain

C

a

b

c

IJlJllllll IIIIl 1 J 0.5 mV 0.5 s rr (4

t-

0

~

5)

I.L. r-

50

(16)I

m c

~100

I-_~±

10~

I

1041 10-7 lJ0-6 10-5 Morphine Concentration (MI Fig. 5. A,B: responses to morphine of PVN neurons taken from the S T R / N (A) and S / W (B) mice studied in vitro. Morphine hydrochloride was added to the medium (indicated by open bars) at varied concentrations. Rate meter records. C: oscilloscope tracings of neuron discharges taken at the time marked in B, top tracing. D: dose-response curves of the S T R / N ( = ) and the S / W (e) to morphine. Values are means S.E.M. Numbers in the parenthesis indicate number of neurons tested. n

Varied results have been reported as to the central sites and mode of actions of opioid agonists and antagonists on drinking in non-polydipsic rats. Most agree that the antagonists generally suppress drinking as well as feeding. Centrally administered naloxone or naltrexone reduced both drinking and feeding in deprived and non-deprived animals 4,5,35,43. On the other hand, conflicting findings have been reported regarding effects of opiate agonists on drinking. Some found that the

agonists decreased drinking 1°'3°'36, while others found they increased drinking 3°'3L42. It has been speculated by some that the effect of an opiate antagonist on drinking was mediated by its antagonizing action at the K-receptor 2°, though fl-endorphin, the endogenous /~and tS-ligands, was also suggested as more effective stimulator of water intake than the K-ligand, dynorphin =2. Injections of the opiates or its antagonists into the PVN were also carried out in non-polydipsic animals in some studies. Ukai and Holtzman 43 injected naltrexone methobromide into various regions in the rat brain and found that post-deprivational water intake was suppressed when the least amount of the drug was injected to the PVN and a slightly larger dose for the SON. In other studies using again rats, Gosnell et al. ~2 found that an injection of naloxone into the PVN and VMH (the ventromedial nucleus of the hypothalamus) reduced feeding, while dynorphin into the PVN and VMH caused a large increase in feeding but not drinking. They also found that following /3-endorphin injection to the PVN drinking increased significantly for 4 h in rats. Results are complicated because fooddeprived but not water-deprived animals were used in the study. As these studies suggest, the opiate's action on drinking or feeding behavior may be more complicated than expected. It is possible to conclude, however, that the PVN is one of a few important sites containing opiate receptors that participate in ingestive behavior, though some other brain regions, such as the preoptic, the lateral hypothalamic areas and zona incerta were also suggested as possible sites 13'32'43. This brings up a question of the opiate's action on PVN neurons. In brain slice preparations of normal rats morphine, met-enkephalin analogues and fl-endorphin all produced reduction in their firing frequencies of PVN neurons 23"e7. In guinea pigs a /x-type agonist only was found to be effective, while ~- and K-agonists did not show the inhibitory action on PVN neurons 45. It was suggested that this inhibition was due to direct action of the opioid peptide on/~-receptors to hyperpolarize the neurons through an increased K ÷ conductance8a L45. Such finding, however, does not give us an answer as to why these PVN neurons of the polydipsic mice are less sensitive to the opiate agonists compared to controls. We found that the threshold concentration of the neurons of the polydipsic mice to morphine and dynorphin was higher (10 -8 vs. 10 -9 M), and moreover, the proportion of neurons responded to the opiate was less in the polydipsic than in controls (41.7% vs. 64.9% for morphine; 25.4% vs. 70.4% for dynorphin). These findings are very comparable to what we found in neurons in the AV3V =4. The

30 results can be interpreted by assuming that in the polydipsic mice presence of more opiate, particularly the K-receptor type opiate, dynorphin, is abundant and this must create desensitization of neurons to the opiate. Since the occurrence of the dynorphin-induced inhibition of neurons (25.4%) is significantly lower (X2=4.06) than that of the morphine-induced inhibiton (41.7%), it is likely that neurons are desensitized more to dynorphin than to morphine in the S T R / N mice. There is much evidence to show that the antagonistic effect of naloxone is enhanced when the animal is pretreated with morphine 2'38'4°. Moreover, the increased efficacy of naloxone is at a maximum in animals that are highly tolerant to chronic exposure to morphine. The affinity constant of the receptors for naloxone is found to increase 8 times in the tolerant state, which may be due to a qualitative rather than a quantitative change in receptors4k Therefore, in our studies it is reasonable to find that the S T R / N mice whose brain probably contains more opioid are more sensitive to the opioid antagonists compared to the controls, while their neurons are less sensitive to the opioid agonists due to desensitization. Such assumption is supported by our recent findings that vaso-

A

pressinergic neurons in the hypothalamus of the S T R / N mice are more numerous (Koizumi, K., Yamashita, H., and Kawata, M. et al., unpublished observation). Dynorphin is known to co-exist with vasopressin in the hypothalamic neurons. Watson et al. 44 have reported that the two peptides appear generally, but not always, in the same cell. Moreover, an augmented secretion of vasopressin is also accompanied by an increase in dynorphin accumulation 2~. It is thus possible that more dynorphin is present in the brain of the polydipsic than normal mice. Such studies are now being pursued. The above hypothesis assumes that presence of abundant dynorphin in the hypothalamus is a cause or, at least, contributes to the polydipsia. Others found that dynorphin A injected into the preoptic area in rats increased water intake, but not feeding, for 2-4 h after injection and the action was prevented by pretreatment of naloxone 32. The mechanisms of dynorphin's action on drinking may be more than one. Leander and Hyness 2° have explained that dynorphin's action on drinking is due to its inhibitory action on release of vasopressin (and oxytocin) from the pituitary ~, thus causing polyuria and concomitant polydipsia. Although such hypothesis may

STR/N

C STR/N 10

Before

5 rain

Oynorphin 10-6M

Oynorphin 10-6M

5 min

v~ 5 F LOW Ca+¢'- High Mg**

105E 'v'~,.~ ~

o

Naloxone 10-6 M

====

Dynorphin 10-6 M

Oynofphin 10-6 M

5 min

After 10

5 Oynorphin 10-6 M

~ C)ynorphin 10-6M

5 min

5 rain 5 min

5 rain

B

D

S/W

SlW Before

fO

Dynorphin 10-6 M

10rain

Dynorphin 10-6M

w

5 rain

Low Ca'* - High Mg*÷ ~

OL

f

~ o

,

i NaJoxone 10-6 M Dynorphin 10-6 M

Oynorph~n 10

M

~=~ Dynorphin 10-6 M

lO"-'~in

Dynorphin 10-6M

5 rnin

After

10 min 5 rain

5 rnin

Fig. 6. A,B: rate meter records of PVN neurons showing the effect of naloxone on the dynorphin-induced inhibitory action. Single neurons taken from the S T R / N (A) and the S / W (B) mice in vitro. A prior (5 min) addition of naloxone to the medium blocked the inhibitory effect of dynorphin (middle traces), and naloxone's effect was reversible (bottom traces). C,D: persistence of the inhibitory action of dynorphin on PVN neurons under synaptic blockade (low Ca 2+ and high Mg 2÷ ), observed in both S T R / N (C) and S / W (D) mice. Open bars indicate the time of applications of dynorphin or naloxone.

31 explain dynorphin's facilitatory action on drinking in normal rats in certain experimental conditions, our study showed that basal level of plasma vasopressin in the polydipsic mice was similar to that of normal mice (Yamada, Yamashita and Koizumi, unpublished observation). Thus, it is unlikely that the polydipsia in these mice is induced by the polyuria. Another attractive idea is to assume the inhibitory system in control of ingestive behavior 37, i.e., once drinking starts it generates stimulus which activates the 'central inhibitory system' that suppressess further drinking. In the polydipsic mice the opiate may be inhibiting this central inhibitory system, and therefore each drinking, once it begins, lasts much longer, resulting in the polydipsia. We indeed found that in nocturnal drinking the polydipsic mouse not only drinks more frequently but also drinks a far larger amount of water at each sip t9. We do not know, however, where the central inhibitory system is located, and how it works. The present working hypothesis then is that in the polydipsic mice, STR/N, there is an overaction of vasopressinergic neurons that also produce dynorphin in the brain. The particular opiate acts to increase spontaneous drinking, possibly through inhibiting the 'inhibitory system' on water intake, though its real mechanisms are not clear at present. An increase in dynorphin in certain regions in the brain results in desensitization of neurons in the AV3V as well as in the PVN. Though some studies suggest that dynorphin in the hypothalamus and in the pituitary inhibits vasopressin release t, in the S T R / N this apparently does not occur. In addition, the possibility still exists that brain angiotensin II may play some role in the polydipsia directly, or through interaction between that and the opiates 18. Acknowledgements. The authors express their appreciation to National Institute of Health, USPHS for Grant support (NS 00847) for this study, to the Japanese Government for Investigator Fellowship (to I.N.), to Dr. Emanuel Silverstein for the original supply of the special strains of mice, and to Boehringer and Ingelheim Pharmaceuticals, Inc., Ridgefield, CT for generous supply of naltrexone methobromide.

REFERENCES 1 Bicknell, R.J., Optimizing release from peptide hormone secretory nerve terminals, J. Exp. Biol., 139 (1988) 51-65. 2 Brase, D.A., Iwamoto, E.T., Loh, H.H. and Way, E.L., Reinitiation of sensitivity to naloxone by a single narcotic injection in postaddicted mice, J. Pharmacol. Exp. Ther., 197 (1976) 317-325. 3 Brown, D.R. and Goldberg, L.I., The use of quaternary narcotic antagonists in opiate research, Neuropharmacology, 24 (1985) 181-191. 4 Brown, D.R. and Holtzman, S.G., Suppression of deprivation-induced food and water intake in rats and mice by naloxone, Pharmacol. Biochem, Behac., 11 (1979)567-573. 5 Calcagnetti, D.J., Helmstetter, F.J and Fanselow, M.S., Central

and peripheral injection of quaternary antagonist, SR58002C, reduces drinking, Physiol. Behac., 40 (1987) 573-575. 6 Carter, T.C., Dunn, L.C., Falconer, D.S., Gruneberg, H, Heston, W.E and Sell, G.D., Standardized nomenclature for inbred strains of mice, Cancer Res., 12 (1952) 602-613. 7 Chang, K-J., Cooper, B.R., Hazum, E and Cuatrecasas, P., Multiple opiate receptors: different regional distribution in the brain and differential bindings of opiate and opioid peptides, Mol. Pharm., 16 (1979) 91-104. 8 Charpak, S., Dubois-Dauphin, M., Raggenbass, M and Dreifuss, J.J., Direct inhibition by opioid peptides of neurones located in the ventromedial nucleus of the guinea pig hypothalamus, Brain Res., 450 (1988) 124-130. 9 Czech, D.A., Stein, E.A. and Blake, M.J., Naloxone-induced hypodipsia: a CNS mapping study, Life Sci., 33 (1983) 797-803. 10 DeCaro, G., Micossi, L.G. and Venturi, F., Drinking behaviors reduced by intracerebroventricular administration of enkephaline to rats, Nature, 277 (1979) 51-55. 11 Dudek, R.D., Tasker, J.G. and Wuarin, J.-P., Intrinsic and synaptic mechanisms of hypothalamic neurons studied with slice and explant preparations, J. Neurosci. Methods, 28 (1989) 59-69. 12 Gosnell, B.A., Morley, J.E. and Levine, A.S., Opioid-induced feeding: localization of sensitive brain sites, Brain Res., 369 (1986) 177-184. 13 Grossman, S.P., Thirst and Sodium Appetite: Physiological Basis, Academic Press, New York, 1990. 14 Hattori, Y., Katafuchi, T. and Koizumi, K., Characterization of opioid-sensitive neurons in the anteroventral third ventricle region of polydipsic inbred mice in vitro, Brain Res., 538 (1991) 283-288. 15 Hattori, Y. and Koizumi, K., Sensitivity to angiotensin II of neurons in the circumventricular organs of polydipsic inbred mice, Brain Res., 524 (1990) 181-186. 16 lnenaga, K. and Yamashita, H., Excitation of neurones in the rat paraventricular nucleus in vitro by vasopressin and oxytocin, J. Physiol., 370 (1986) 165-180. 17 Katafuchi, T., Hattori, Y., Nagatomo, I. and Koizumi, K., KOpioid antagonist strongly attenuates drinking of genetically polydipsic mice, Brain Res., 546 (1991) 1-7. 18 Katafuchi, T., Hattori, Y., Nagatomo, I., Koizumi, K and Silverstein, E., Involvement of angiotensin II in water intake of genetically polydipsic mice, Am. J. Physiol., 260 (1991) Rl152-Rl158. 19 Koizumi, K., Hattori, Y., Katafuchi, T and Silverstein, E., Studies of spontaneously polydipsic inbred mice. In S. Yoshida and L. Share (Eds.), Recent Progress in Posterior Pituitary Hormones, Elsevier, Amsterdam, 1988, 403-410. 20 Leander, J.D. and Hynes III, M.D., Opioid antagonists and drinking: evidence of K-receptor involvement, Ear J. Pharmacol., 87 (1983) 481-484. 21 Lightman, S.L., The neuroendocrine paraventricular hypothalamus: receptors signal transduction, mRNA and neurosecretion, J. Exp. Biol., 139 (1988) 31-49. 22 Montemuro, D.G. and Dukelow, R.H., A Stereotaxic Atlas of the Diencephalon and Related Structures of the Mouse, Futura, Mount Kisco, N.Y., 1972. 23 Muehlenthaler, M., Gaehwiler, B.H. and Dreifuss, J.J., Enkephalin-induced inhibition of hypothalamic paraventricular neurons, Brain Res., 197 (1980) 264-268. 24 Nagatomo, I., Katafuchi, T. and Koizumi, K., Role of paraventricular nucleus mediating opiates' action on polydipsia, Abstract, Soc. Neurosci., 15 (1989) 897. 25 Okuya, S., Inenaga, K., Kaneko, T and Yamashita, H., Angiotensin II sensitive neurons in the supraopitc nucleus, subfornical organ and anteroventral third ventricle of rats in vitro, Brain Res., 402 (1987) 58-67. 26 Okuya, S. and Yamashita, H., Effects of atrial natriuretic polypeptide on rat hypothalamic neurones in vitro, J. Physiol., 389 (1987) 717-728. 27 Pittman, Q.J., Hatton, J.D. and Bloom, F.E., Morphine and opioid peptides reduce paraventricular neuronal activity: studies in the rat hypothalamic slice preparation, Proc. Natl. Acad. Sci. USA, 77 (1980) 5527-5531.

32 28 Portoghese, P.S., Lipkowski, A.W and Takemori, A.E., Bimorphimines as highly selective, potent K-opioid receptor antagonist, J. Med. Chem., 30 (1987) 238-239. 29 Portoghese, P.S., Lipkowski, A.W and Takemori, A.E., Binaltorphimine and nor-binaltorphimine, potent and selective K-opioid receptor antagonists, Life Sci., 40 (1987) 1287-1292. 30 Sanger, D.J. and McCarthy, P.S., Differential effects of morphine on food and water intake in food deprived and freely-feeding rats, Psychopharmacology, 72 (1980) 103-106. 31 Sanger, D.J. and McCarthy, P.S., Increased food and water intake produced in rats by opiate receptor agonists, Psychopharmacology, 74 (1981) 217-220. 32 Shimizu, H., Shimomura, Y., Negishi, M., Kobayashi, I. and Kobayashi, S., Dynorphin A (1-13), microinjected into the preoptic area, stimulates water intake in rats, Life Sci., 45 (1989) 25-30. 33 Silverstein, E., Effect of hybridization on the primary polydipsic trait of an inbred strain of mice, Nature, 191 (1961) 523. 34 Silverstein, E., Sokoloff, L., Mickelsen, O. and Jay Jr., G.E., Primary polydipsia and hydronephrosis in inbred strain of mice, Am. J. Pathol., 38 (1961) 143-159. 35 Sivily, S.M., Bermudez-Rattoni, F., Rockwood, G.A., Dargie, C.M. and Reid, L.D., Intracerebral administration of naloxone and drinking in water deprived rats, Pharm. Biochem. Behat,., 15 (1981) 257-262. 36 Spencer, R.L., Deupree, D., Hsiao, S., Mosberg, H.I, Hruby, V., Burks, T.F and Porreca, F., Centrally-administered opioid selective agonists inhibit drinking in the rat, Pharm. Biochem. BehaL'., 25 (1986) 77-82.

37 Stricker, E.M., Homeostatic origins of ingestive behavior. In E.M. Stricker (Ed.), Neurobiology of Food and Fluid Intake, Plenum, New York, 1990, 45-60. 38 Takemori, A.E., Oka, T. and Nishiyama, N., Alteration of analgesic receptor-antagonist interation induced by morphine, J. PharmacoL Exp. Then, 186 (1976) 261-265. 39 Takemori, A.E., Schwartz, M.M. and Portoghese, P.S., Suppression by nor-binaltorphimine of K-opioid-modified diuresis in rats, J. PharmacoL Exp. Then, 247 (1988) 971-974. 40 Tulunay, F.C. and Takemori, A.E., The increased efficacy of narcotic antagonists induced by narcotic analgesics, Z Pharmacol. Exp. Then, 190 (1974) 395-400. 41 Tulunay, F.C. and Takemori, A.E., Further studies on the alteration of analgesic receptor-antagonist interaction induced by morphine, J. Pharmacol. Exp. Ther., 190 (1974) 401-407. 42 Turkish, S. and Cooper, S.J., Effects of a kappa receptor agonist, ethylketocyclazocine, on water consumption in water-deprived and nondeprived rats in diurnal and nocturnal tests, Pharm. Biochern. Behat,., 21 (1984) 47-51. 43 Ukai, M. and Holtzman, S.G., Suppression of deprivation-induced water intake in the rat by opioid antagonists: central sites of action, Psychopharmacology, 91 (1987) 279-284. 44 Watson, S.J. Akil, H., Fischli, W,, Goldstein, A., Simmerman, E. and Nilaver, G., Dynorphin and vasopressin: common localization in magnocellular neurons, Science, 216 (1982) 85-87. 45 Wuarin, J.-P, Dubois-Dauphin, M., Raggenbass, M and Dreifuss, J.J., Effect of opioid peptides on the paraventricular nucleus of the guinea pig hypothalamus is mediated by /x-type receptors. Brain Res., 445 (1988) 289-296.