Antidiuretic effects of purinoceptor agonists injected into the hypothalamic paraventricular nucleus of water-loaded, ethanol-anesthetized rats

Neuropharmacoiogy Vol.31,No. 6, pp. 585-592,1992 Great Britain. All rightsreserved

0028-3908/92 $5.00+ 0.00 Copyright0 1992PergamonPressLtd

Printedin

ANTIDIURETIC EFFECTS OF PURINOCEPTOR AGONISTS INJECTED INTO THE HYPOTHALAMIC PARAVENTRICULAR NUCLEUS OF WATER-LOADED, ETHANOL-ANESTHETIZED RATS M. MORI, H. TSUSHIMAand T. MATSUDA Department of Pharmacology, Nagoya City University Medical School, Kawasumi, Mizuho-ku, Nagoya 467, Japan (Accepted 27 November 1991)

Summary-The effects of injection of various purinoceptor agonists into the hypothalamic paraventricular nucleus in water-loaded and ethanol-anesthetized rats were investigated. Adenosine triphosphate (ATP), /?,y-methyleneadenosine 5’-triphosphate (AMP-PCP) and /I,y-imidoadenosine S&phosphate (AMP-PNP) potently decreased the outflow of urine in a time- and dose-dependent manner. The ED, values were approx 70 and 37 nmol for ATP and AMP-PCP, respectively. Adenosine diphosphate (ADP), AMP and adenosine reduced the outflow of urine much less than ATP. Adenosine triphosphate induced concomitant increases in the osmotic pressure of the urine and in the level of arginine-vasopressin (AVP) in plasma. The antidiuretic effect of ATP was blocked by prior injection of quinidine (a P,-purinoceptor antagonist) into the paraventricular nucleus, but not by the prior injection of theophylline (a P,-purinoceptor antagonist). The effect of ATP was also blocked by intravenous injection of an AVP(V,V,)-receptor antagonist, d(CH,),-D-Tyr(Et)VAVP. The results suggest that ATP injected into the paraventricular nucleus may stimulate a purinoceptor, releasing AVP and inducing the antidiuretic effect through renal AVP(V,) receptors. Key words-ATP,

antidiuresis, vasopressin, purinoceptor, hypothalamic paraventricular nucleus.

The hypothalamic paraventricular nucleus, as well as the supraoptic nucleus, contains cell bodies of vasopressinergic neurons which synthesize and release antidiuretic hormone, arginine-vasopressin (AVP), in most mammals including rats. Stimulation of vasopressinergic neurons by neurotransmitters and neuromodulators induces a release of AVP, eventually causing antidiuretic effects through the renal AVP(V& receptors (Reichlin, 1981; Schrier and Leaf, 1981; Zimmerman, 1983; Reid, 1983; Zimmerman, Hon-Yu, Nilaver and Silverman, 1984; Hays, 1985; Manning and Sawer, 1985). The innervation of the paraventricular nucleus by cholinergic and adrenergic neurons has been demonstrated histochemically (McNeil1 and Sladek, 1980; Kimura, MacGeer, Reng and MacGeer, 1981; Swanson and Sawchenko, 1983), which has suggested that these neuronal systems may play a role in regulation of the outflow of urine. Cholinergic (muscarinic) and adrenergic (a- and fi-subtypes) receptor mechanisms in the paraventricular nucleus have been reported by injecting directly various drugs into the nucleus and measuring the outflow of urine (Mori, Tsushima and Matsuda, 1984, 1989; Tsushima, Mori and Matsuda, 1985, 1986). On the other hand, ATP has been reported to be stored, together with acetylcholine (ACh) and noradrenaline (NA) in cholinergic and adrenergic nerve endings (Dowdall, Boyne and Whittaker, 1974; Lagercrantz and StjZirne, 1974). Adenosine triphos-

phate (ATP) acts at a specific receptor, modulating the release of neurotransmitters and causing pharmacological effects in a variety of tissues, including the peripheral and central nervous systems (Burnstock, 1978, 1981, 1986). Therefore, it was deemed interesting to test the effects of ATP and other purinoceptor agonists, injected into paraventricular vasopressinergic neurons and compare its effects with those of ACh and NA. In the present study, purinoceptor mechanisms in the paraventricular nucleus were investigated by injecting ATP and related compounds into the nucleus. The effects of ATP were studied on the outflow and the osmotic pressure of urine and on the concentration of AVP in the plasma. The influences of pretreatments with AVP receptor- and purinoceptorantagonists on the ATP-induced antidiuretic effects were also examined. METHODS Drugs

Adenosine 5’-triphosphate sodium salt (ATP), /?,y-methyleneadenosine 5’-triphosphate tetralithium salt (AMP-PCP), /?,y-imidoadenosine 5’-triphosphate sodium salt (AMP-PNP), adenosine 5’-diphosphate sodium salt (ADP), adenosine 5’-monophosphate sodium salt (AMP), adenosine hemi-sulfate (Sigma Chemical Co., St Louis, Missouri, U.S.A.), quinidine

585

586

M. MORI et al.

sulfate (Tokyo Kasei, Tokyo, Japan) and theophylline (Wako Pure Chemical Industries, Ltd, Osaka, Japan) were purchased. For radioimmunoassay, [1251]-AVP (Amersham, Buckinghamshire, antiserum rabbit England), Arg8-vasopressin (Carbiochem, La Jolla, California, U.S.A.) and AVP (Grade VI, Sigma Chemical Co., St Louis, Missouri, U.S.A.) were used. The AVP(V,V,) antagonist, l-( pmercapto-/I,/?-cyclopentamethylene propionic acid), 2-(O-ethyl-o-tyrosine, 4-vahne) arginine vasopressin: d(CH,),-o-Tyr(Et)VAVP, was kindly provided by Professor K. G. Hofbauer (Department of Pharmacology, Heidelberg University, Fed. Rep. Germany and Cardiovascular Research Department, Pharmaceutical Division, Ciba-Geigy, Ltd, Basel, Switzerland). The other chemicals, such as drugs which were used for the artificial cerebrospinal fluid (NaCI, KCl, MgCl,, NaH*PO,, NaHCO,) and used for the radioimmunoassay of arginine-vasopressin (acetone, ether, Na2HP0,, K,HPO,) and ethanol were of the highest analytical grade available. Measurement of outflow and osmotic pressure of urine

The outflow of urine was measured by Dicker’s method, with some modifications (Dicker, 1953; Mori et al., 1984). Male Wistar rats (n = 59) weighing 280-320 g, were starved overnight for approx 17 hr but had free access to water. The animals were loaded orally through a catheter with a volume of water equivalent to 5% of the body weight. Then, they were anesthetized with oral administration of a volume of 12% ethanol, equivalent to 5% of the body weight. Cannulae were inserted into the trachea, bladder and external jugular vein. The rats were immobilized in a stereotoxic instrument for rats (Takahashi Co., Tokyo, Japan). Drops of urine, flowing from the urinary cannula, were counted using a photoelectric drop counter (DCT 102, Unique Medical Inc., Tokyo, Japan) and recorded as single pulses. Ethanol (3% in Locke solution) was infused at a constant rate of 0.10 ml/min through the cannula in the jugular vein, in order to maintain a constant level of anesthesia and a constant rate of urine outflow. The administration of ethanol used in the present study induced unconsciousness, analgesia, relaxation of the extremities and loss of the laryngeal and the peritoneal reflexes, showing some symptoms of the first phase to second phase of the stage of surgical anesthesia. Rats did not show nausea or vomiting during the anesthesia. Osmotic pressure of the urine was measured by the freezing point depression method, using the Fiske Osmometer (Model G-62, Fiske Associates, Inc., Uxbridge, Massachusetts, U.S.A.) (Towstoless, Congiu, Coghlan and Wintour, 1987). Urine, which was collected by a catheter in the bladder (dead space: 200 ~1) and was diluted to IO-fold by distilled water, was used (minimum volume:200 ~1) for measuring osmotic pressure. The initial control osmotic pressure was 237 + 43 mOsm.

Injection of drugs

A stainless steel cannula (o.d.: 200pm) was unilaterally inserted stereotaxically into the paraventricular nucleus, according to the atlas of Kiinig and Klippel (1963). Adenine nucleotides and adenosine were dissolved in saline, then the solutions were adjusted to approx pH 7 by NaOH. Injections of 1~1 of ATP, its analogs, ADP, AMP and adenosine were preformed when the outflow of urine reached a constant rate of approx 0.1 ml/min, which was usually within 1 hr after the animal was fixed in the stereotaxic instrument. Then, 2~1 of an artificial cerebrospinal fluid (C.S.F.: 128mM NaCl, 3.0mM KCI, 1.2 mM CaCI,, 0.8 mM MgCl*, 0.65 mM NaH,PO, and 4.8 mM NaHCO,, pH 7.4) was infused at a rate of approx 0.3 pI/min. There was a dead space of approx 1 ~1 between the tip of a cannula, inserted into the paraventricular nucleus and the tip of microsyringe, connected to the canula. An injection of 1 ~1 of a drug was followed by an infusion of 2 ~1 of C.S.F. in order to inject the solution of drug as completely as possible into the nucleus. Effects of drugs on the outflow of urine were measured at intervals of 10 min and expressed as a percentage of the initial control outflow. In the experiments to test the effect of pretreatment with a purinoceptor antagonist (quinidine sulfate or theophylhne), a first injection of ATP was followed by an injection of a purinoceptor antagonist and then a second injection of ATP was carried out. All the injections were performed at a time when the outflow of urine had recovered to the initial level, at 4070 min, at 30-40 min and at 20-30 min, after the injection of ATP, quinidine and theophylline, respectively, through a single cannula, inserted into the paraventricular nucleus. The controls for the multiple injection procedure: ATP (50 nmol) -+ quinidine (72 nmol) -+ ATP (50 nmol) were: ATP (50 nmol) - vehicle (saline) -+ ATP (50 nmol). The controls showed that pretreatment with vehicle alone had no significant effects on antidiuretic effects of the second injection of ATP. The inhibitory effects of the antagonists were estimated as changes in antidiuretic effects, caused by the injection of ATP, with and without the pretreatment. Pretreatment with an AVP antagonist

In the experiments to test the effects of pretreatment with the AVP antagonist, d(CH,),-D-Tyr(Et)VAVP, a first injection of ATP into the paraventricular nucleus was followed by an intravenous injection of the AVP antagonist (50 pg/kg, a dose which did not show any significant change in outflow of urine). Then a second injection of ATP was performed through the same cannula. The antagonist, dissolved in saline, was injected intravenously through a cannula in the external jugular vein, when the outflow of urine had recovered to the initial level, at approx 70 min after the first injection of ATP (100 nmol). A second injection of

Injections of ATP into the paraventricular nucleus ATP was carried out at 40min after the injection of the antagonist. The inhibitory effects of the antagonist were estimated as described above for the inhibitory effects of the purinoceptor antagonists.

Kohden Kogyo, Co.). Rectal temperature was moni:ored by a thermister probe (MGA 111-219, Nihon Kohden Kogyo, Co.), inserted into the rectum. ldentiJication of the sites of inserted can&a

Radioimmunoassay (RIA) of A VP in plasma

The position of the tip of the cannula, within the ?araventricular nucleus, was confirmed by the followng method: (1) functionally, by the appearance of an antidiuretic effect by the injection of depolarizing Jose of KC1 (400 nmol) through the cannula and 12) histochemically, by localization of the site of the tip of the cannula in a group of magnocellular cells In the paraventricular nucleus, positively stained by the method of Gomori (Bargman, 1949).

Water-loaded and ethanol-anesthetized rats were decapitated and trunk blood was collected into chilled and heparinized glassware. The plasma was separated immediately by centrifugation and the AVP in plasma was extracted by the acetone-ether method (Dunn, Brennan, Nelson and Robertson, 1973). The extract was dried with an evaporation head (EH-2, Taiyo Scientific Industrial Co., Ltd, Tokyo, Japan) by using a stream of nitrogen gas. The samples were stored at -20°C for no more than 1 month until measurement. The RIA of AVP was performed as previously described (Ishikawa, Saito and Yoshida, 1980). The minimum amount of AVP measurable by the method was approx 2pg/tube. The inter- and intra-assay coefficients of variation were 4.0 and 1.3%, respectively. The recovery rate of AVP (25 pg/ ml), extracted from plasma, was 69.7 + 2.1% (n = 8).

Statistical analysis

Student’s t-test was used for the statistical analysis of paired tests. One-way analysis of variance (ANOVA) was used for statistical analysis of multiple comparisons of data. Differences were considered significant at P < 0.05. The ED,, values and 95% confidence limit of the EDS, values, for ATP and AMP-PCP, were calculated from dose-effect curves (Fig. 2), drawn using the least squares method.

Measurement of blood pressure, heart rate, respiration rate and rectal temperature

Mean blood pressure and heart rate were measured through a cannula, inserted into the carotid artery, using, respectively, a pressure transducer (MPU-0.5 290-o-111, Nihon Kohden, Co., Tokyo, Japan) and an electrocardiograph (Fukuda, FD-14, Tokyo, Japan). Respiration rate was measured by a thermister probe (SR-115S, Nihon Kohden Kogyo, Co.), inserted into a tracheal catheter. These three indices were recorded simultaneously on a recticoder (RJG-3004-2, Nihon

0

20

40

RESULTS

Effects of injection of ATP and its analogs on the outflow of urine

Figure l(a) shows the effects of the injection of ATP (lo-100 nmol) into the paraventricular nucleus on the outflow of urine. Adenosine triphosphate decreased the outflow in a time- and dose-dependent manner. The outflow of urine usually decreased

60 Time

587

after

I 20

I 0 injection

(min

I 40

I 60

1

Fig. 1. Effects of various doses of ATP, AMP-PCP and AMP-PNP on the outflow of urine, as a function of time, after the injection into the paraventricular nucleus. In (a) n = control (n = 3); 0 = 10 (n = 3); A = 50 (n = 5); 0 = 100 (n = 8) nmol of ATP; in (b) l = control (n = 3); 0 = 10 (n = 5); A = 30 (n = 5); A = 50 (n = 7); q = 100 (n = 4) nmol of AMP-PCP and l = 100 (n = 4) nmol of AMP-PNP. Ordinate: rate of outflow of urine, during the preceding period of 10 min, expressed as a percentage of the rate of initial outflow of urine [(a): 0.097 + 0.012, (b): 0.080 _t 0.009 ml/min]. Abscissa: time in min after injection. Symbols are the mean f SEM of experiments (total number of rats: (a) n = 15; (b) n = 18).

M. Mom et al.

588

(a) ‘T, 2

*

(b)

100

5 B $ 3 0 5

50

a zi .. ;

0

I

I

I

I

10

30

50

100

I 150

Dose f nmol) Fig, 2. Dose-effect curves of antidiuretic effects of ATP and AMP-PCP and the effect of a previous injection with quinidine on the effects of ATP, injected into the paraventricular nucleus. 0 and A: the effects of ATP without and with a previous injection of 72 nmol quinidine, respectively, l : the effects of AMP-PCP without the nrior injection. Since quinidine (72 nmol) alone induced a significant antidiuretic effect, ATP was injected at 3040 min after the previous injection of quinidine, when the outflow of urine returned approximately to the initial control level. Ordinate: minimal rate of outflow of urine during the preceding period of 10 min. at 20 and 30 mm after the injections of ATP and AMP-PCP, respectively, expressed as- a percentage of the initial rate of control outflow (0: 0.112 + 0.018. a: 0.090 5 0.044, 0: 0.079 F 0.010 ml/mm). Abs&sa: dose of ATP and AMP-PCP, nmol. Symbols are the mean k SEM with the number of experiments indicated in parentheses.

within 20 min after the injection of ATP (100 nmol),

with a minimum outflow at 20 min (n = 8, P < 0.05). The outflow returned to the initial control outflow at approx 1 hr (n = 8, P < 0.05, vs the values at 20 min). The difference between antidiuretic effects, induced by ATP at 50 nmol and those at 100 nmol, was not significant (n = 5 and 8, respectively; P > 0.1). The effects of AMP-PCP and AMP-PNP, injected into the paraventricular nucleus, are illustrated in Fig. l(b), demonstrating some stronger effects than those of ATP. Vehicle (saline) alone, when injected into the nucleus, did not change the outflow of urine (n = 3, Fig. 1). Figure 2 shows the dose-effect curves for ATP and AMP-PCP. The approximate median effective doses (ED, values), were estimated from the dose-effect curves (Fig. 2), to be 70 (values ranged from 44 to 111, n = 15) and 37 (values ranged from 30 to 46, n = 12) nmol for ATP and AMP-PCP, respectively. Eflects of inject~an of ATP on the osmotic pressure of urine Figure 3 shows the effects of ATP, injected into the paraventricular nucleus, on the osmotic pressure, compared with its effects on the outflow of urine. The osmotic pressure increased to approx 250% of the

c ATP

CSF 20

min

after

CSF 60

ATP mln

injection

Fig. 3. Effects of ATP, injected into the paraventricular nucleus on the outflow and osmotic pressure of urine. c]: Rate of the outflow of urine during the preceding period of 10 min, at 20 min (a) and 60 min (b) after the injection of C.S.F. or ATP (100 nmol), expressed as a percentage of the initial rate of control outflow (C.S.F.: 0.066 F 0.012, ATP: 0.126 + 0.024 mi/min. The rate of outflow was minimum at 20 min and recovered approximately to the initial control values at 60 min after the injection of ATP. tB: The osmotic pressure of urine at 20min (a) and 60min (b) after the injections, expressed as a percentage of the initial osmotic pressure (C.S.F.: 263 & 43, ATP: 237 + 43 mOsm). Abscissa: C.S.F. or ATP (100 nmol) injected into the paraventricular nucleus and time after the injections. Columns are the mean & SEM of 3 experiments. Significance compared with the value for the injection for C.S.F.: *P -z 0.05.

initial control value at 20min after the injection, when the outflow decreased to a minimum, approx 25% of the initial control value. In contrast, when the outflow recovered to the initial level, at 60 min after the injection, the osmotic pressure returned to approximately the initial control value. Vehicle (C.S.F.) alone, when injected into the nucleus, did not change the osmotic pressure, nor the outflow of urine (n = 3, Fig. 3). Egects

of injection of ATP

on ieuel of A VP in plasma

The level of AVP in plasma increased to approx 200% of the initial control level (control: 5.4 f 0.4, n = 7; injection of ATP: 10.7 + 1.3 pg/ml, n = 13, P < 0.05) at 20-30 min after the injection of ATP (100 nmol). Effects of pretreatment with purinoceptor antagonist on the antidiuretic e$ect of ATP Quinidine (72 nmol), a P,-purinoceptor antagonist when injected into the paraventricular nucleus,

Injections of ATP into the paraventricular nucleus

(al

150

589

Tt

(bl

t

,*T

Time

after

injection

(min)

Fig. 4. Effects of a previous injection of quinidine (a) or theophylline (b) on the antidiuretic effect of ATP injected into the paraventricular nucleus. In (a) ordinate: rate of outflow of urine during the preceding period of 10 min expressed as a percentage of the initial rate of control outflow (0: 0.089 f 0.007 (n = 4), 0: 0.092 + 0.006 (n = 6) ml/min). Abscissa: time after the first (0) and second (0) injection of ATP (SOmnol). Seventy-two nmol quinidine was injected into the same nucleus, at 3@40 min before the second injection of ATP. In (b), ordinate: the same as in (a), except the initial rate of control outflow (0: 0.118 f 0.030 (n = S), 0: 0.127 + 0.037 (n = 5) ml/min). Abscissa: time after the first (0) and second (0) injection of ATP (100 nmol). Fifty nmol theophylline was injected into the same nucleus, at 20 min before the second injection of ATP. Symbols are the mean & SEM of experiments. Significance compared with the effect of the first injection of ATP. at the same time: *P c 0.05.

almost completely blocked the antidiuretic effect of ATP (50nmol), as illustrated in Fig. 4(a). The dose-effect curve for the antidiuretic effect of ATP

was shifted towards the right by the pretreatment with the purinoceptor antagonist (Fig. 2). Quinidine (72 nmol) alone induced a significant decrease in the outflow of urine. In contrast, a pretreatment with the injection of theophylline (50 nmol), a P,-purinoceptor antagonist, did not inhibit the antidiuretic effect of ATP (50 nmol, Fig. 4b). Theophylline (SOnmol) alone did not change the outflow of urine. The outflow of urine, expressed as a percentage of control, was 103 f 7, 90 f 17, 94 + 13 and 98 + 10 at 10, 20, 30 and 40 min after injection of theophylline into the paraventricular nucleus, respectively (n = 5). Inhibition of ATP-induced antidiuretic effect intravenous injection of an AVP antagonist

20

0 Time

after

40 injection

60 of

ATP

(min)

Fig. 5. Effect of intravenous injection of an AVP(V,V,) receptor antagonist, d(CH,),-o-Tyr(Et)VAVP, on the antidiuretic effect of ATP injected into the paraventricular nucleus. Ordinate and abscissa are as in Fig. 4. 0: First injection, 0: second injection, of ATP (100 nmol). The initial rate of control outflow were 0: 0.083 +O.Oll, 0: 0.101 f 0.028 mI/min. The AVP receptor antagonist (SOpg/kg), was injected intravenously at 40min before the second injection of ATP (0). Symbols are the mean f SEM of 4 experiments. Significance compared with the effects of the first injection of ATP, at the same time: *P < 0.05.

by

As shown in Fig. 5, the AVP (V,V,) receptor antagonist, d(CH,),-o-Tyr(Et)VAVP (50 pg/kg), injected intravenously, completely blocked the antidiuretic effect of ATP (100 nmol), injected into the paraventricular nucleus. The AVP antagonist alone, when injected intravenously, did not significantly change the outflow of urine: the outflow of urine expressed as a percentage of initial control values, was 129f5, 137k11, 108f3, 103+4, lOOf and 104 + 2 at 10, 20, 30, 40, 50 and 60 min after intravenous injections of the AVP antagonists, respectively (n = 3). Efects of injection of various adenine nucleotides and adenosine Figure 6 compares the effects of ADP, AMP and adenosine with the effects of ATP and its analogs

M.

MORI et

=

L?

‘;

100

00 ‘i;

E

4

$

50

a e ._ ; n CSF

n

7

AMP

AMP

-PCP

-PNP

4

4

ATP

ADP

8

7

AMP

5

ADO

4

Fig. 6. Effects of injection of adenine nucleotides and adenosine on the outflow of urine, compared with the effects of ATP and its analogs. Ordinate: minimal rate of urine outflow during the preceding 10 min period, expressed as a percentage of the initial rate of control outflow (C.S.F.: 0.066 + 0.012, AMP-PCP: 0.065 f 0.005, AMP-PNP: 0.086 + 0.006, ATP: 0.085 f 0.021, ADP: 0.116 + 0.017, AMP: 0.101 & 0.018, ADO (adenosine): 0.129 + 0.014 ml/min). Abscissa: C.S.F. and drugs (100 nmol) injected into the paraventricular nucleus and n: numbers of experiments. The minimum rates of outflow of urine were at 20 min after the injections of adenine nucleotides and adenosine and at 30min after the injections of AMP-PCP and AMP-PNP, respectively. Symbols are the mean + SEM. Significance compared with the values for injection of C.S.F.: *P < 0.05.

on the outflow of urine when

they were injected into the paraventricular nucleus, demonstrating the maximum effects at 20-30 min after the injections.

The antidiuretic effects of the adenine nucleotides and adenosine were much less than the effects of ATP and its analogs. Cardiovascular and respiratory effects of injection of ATP

Some visceral functions, which might be expected to change by the injection of ATP into the paraventricular nucleus and which might influence the outflow of urine, were monitored during the experiments. However, no significant changes in blood pressure, heart rate, respiration rate and rectal temperature were observed by the injection of ATP (100 nmol) into the paraventricular nucleus (data not shown). DISCUSSION

Adenosine triphosphate, when injected into the hypothalamic paraventricular nucleus in waterloaded and ethanol-anesthetized rats, induced a potent antidiuretic effect: a decrease in the outflow, with a concomitant increase in the osmotic pressure of urine, in a time- and dose-dependent manner (Figs la and 3). The time-course of the ATP-induced antidiuretic effect, which appeared maximally at approx 20 min and disappeared at SO-60min, after the injection

al.

(Fig. la), was similar to that of the antidiuretic effects of intravenously injected AVP (10 ng) and the effects of ACh (Mori et al., 1984) and NA (Tsushima et al., 1986), injected into the paraventricular nucleus. The approximate ED, value for ATP was 70 nmol (Fig. 2), whereas the values for ACh and NA were 20 and 4.4 nmol, respectively (Mori et al., 1984; Tsushima et al., 1986). Therefore, the affinity of ATP in the nucleus was approximately a quarter and one-twentieth of the affinities of ACh and NA, respectively. The slowly degradable analogs of ATP, AMP-PCP and AMP-PNP, induced a slightly more potent antidiuretic effect (Fig. lb than the effect of ATP. However, ADP, AMP and adenosine, induced effects several times less potent than those of ATP. The presence of the phosphate at the y-position in the molecule of ATP but not phosphorylation reactions of the y-phosphate nor metabolites of the injected ATP, is likely to play important roles in the effect of ATP. No significant changes could be found in the cardiovascular and respiratory effects by the injection of ATP. The level of AVP in plasma increased approximately 2-fold of the initial control levels during the maximum antidiuretic effect of ATP. In addition, the osmotic pressure of the urine increased by the injection of ATP, as well as an intravenous injection of AVP (3 ng/kg, Tsushima et al., 1986). The antidiuretic effect of ATP was completely blocked by an intravenous injection of the AVP (V,V,) receptor antagonist (50 pg/kg), which is more selective for V2receptors among the AVP antagonists. Intravenous injections of 8 and 100 pgg/kg of the AVP antagonist were sufficient to inhibit the antidiuretic effects of 4 and 25 ng/kg AVP (i.v.) for 3 hr, respectively (Sawer, Pang, Seto, McEnroe, Lammek and Manning, 1981; Ishikawa, Kim and Schrier, 1983; Hofbauer, Mah and Opperman, 1986). Therefore, the present findings suggest that ATP stimulated the vasopressinergic neurons in the paraventricular nucleus, thereby releasing AVP from the neurohypophysis into the circulation, which induced an antidiuretic effect mediated through the renal AVP(V,) receptors. The antidiuretic effect of ATP (50 nmol) was antagonized by injection of a little larger dose of quinidine (72 nmol), a P,-purinoceptor antagonist (Burnstock, 1981) (Figs 2 and 4), than that of ATP but not by the injection of theophylline, a P,-purinoceptor antagonist, at an equimolar dose of ATP (50nmol) (Burnstock, 1978, 1981). The potency of the effect of ATP was much larger than that of adenosine (Fig. 6). The data suggest that the effect of ATP may be mediated through a purinoceptor of the P,-type (Jahr and Jessell, 1983; Salt and Hill, 1983; Salter and Henry, 1985). However, further studies on the type of purinoceptor remain to be investigated. Quinidine, at the dose used in the present study (72 nmol), did not block the antidiuretic effects of NA (40 nmol), injected into the paraventricular nucleus,

Injections of ATP into the paraventricular nucleus which are mediated through a-adrenoceptors (Tsushima et al., 1986). Therefore, the blocking effect of quinidine on the effect of ATP is likely to be a specific effect on purinoceptor. However,. the antidiuretic effect of quinidine alone is difficult to interpret at the present time. In a previous experiment (unpublished) in the supraoptic nucleus, which also contains vasopressinergic neuronal cell bodies, an injection of 50 nmol of theophylline blocked the effect of ATP (100 nmol). However, as shown in Fig. 4(b), the pretreatment with theophylline (50nmol) did not block the effect of ATP (100 nmol) in the present study. The effect of ATP appeared to be mediated through P,-type of purinoceptors in the supraoptic nucleus, but not in the paraventricular nucleus. By the control injection of 3 ~1 of C.S.F. (n = 10) into the paraventricular nucleus, no significant changes in outflow of urine were observed. It seems that damage to the paraventricular nucleus by infusing a volume of 3 ~1 of injectate itself, if present, may not be as great as the changes in outflow of urine. Injection of 1 ~1 volume of effective doses of antidiuretic agonists, such as oxotremorine (0.2 nmol) or NA (80 nmol) at a distance of 1 mm outside (superior or anterior to) the paraventricular nucleus, did not induce any significant effects on the outflow of urine. Therefore, spread of drug to sites in the brain outside the paraventricular nucleus by injection of 1 ~1 volume of drug, seemed to be less than approx 1 mm. Repeated injection of ATP into the paraventricular nucleus induced reproducible antidiuretic effects. The lack of antidiuretic effects of ATP, after pretreatment with quinidine or an AVP receptor antagonist given intravenously, therefore, may not be due to a tolerance developing to repeated administration of ATP. The physiological significance for the antidiuretic effect of ATP, injected into the paraventricular nucleus, is unknown at the present time. However, ATP is known to coexist with ACh (Dowdall et al., 1974) and NA (Lagercrantz and StjBrne, 1974). If ATP was released together with neurotransmitters (StjBrne and Astrand, 1985; Burnstock, 1986), from cholinergic and adrenergic nerve endings in the paraventricular nucleus (McNeil1 and Sladek, 1980; Kimura et al., 1981; Swanson and Sawchenko, 1983), it could modulate the antidiuretic effect of ACh or NA (Mori et al., 1984; Tsushima et al., 1986). In summary, ATP induced a potent antidiuretic effect, when injected into the paraventricular nucleus. In the nucleus there may be purinoceptors, probably of P,-type, which stimulate the release of AVP into the circulation. Acknowledgements-The study was supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture of Japan and from the Research Foundation for Oriental Medicine, Nagoya, Japan. We would like to thank Professor K. G. Hofbauer, Department of Pharmacology, Heidelberg University, Fed.

591

Rep. Germany and Cardiovascular Research Department Pharmaceutical Division, Ciba-Geigy Ltd, Base], Switzerland, for providing the AVP antagonist. REFERENCES

Bargman W. (1949) tiber die neurosekretorische Verkniipfung von Hypothalamus und Neurohypophyse. Z. Zellforsch. 34: 610-634. Burnstock G. (1978) A basis for distinguishing two types of purinergic receptor. In: Cell Membrane Receptors for Drugs

and Hormones:

A

Multidisciplinary

Approach

(Straub R. W. and Bolis L., Eds), pp. 107-118. Raven Press, New York. Burnstock G. (1981) Neurotransmitters and trophic factors in the autonomic nervous system. J. Physiol. 313: l-35. Burnstock G. (1986) Purines as contransmitters in adrenergic and cholinergic neurones. Prog. Brain Res. 68: 193-203.

Dicker S. E. (1953) A method for the assay of very small amounts of antidiuretic activity with a note on the antidiuretic titre of rat’s blood. J. Physiol., Lond. 122: 149-157.

Dowdall M. J., Boyne A. F. and Whittaker V. P. (1974) Adenosine triphosphate: A constituent of cholinergic synaptic vesicles. Biochem. J. 140: l-12. Dunn F. L., Brennan T. J., Nelson A. E. and Robertson G. L. (1973) The role of blood osomolality and volume in regulating vasopressin secretion in the rat. J. clin. Invest. 52: 3212-3219.

Hays R. R. (1985) Agents affecting the renal conservation of water. In: The Pharmacological Basis of Therapeutics (Goodman A. G., Goodman L. S., Rall T. W. and Murad F., Eds), 7th edn, pp. 908-919. MacMillan, New York. Hofbauer K. G., Mah S. C. and Opperman J. R. (1986) Chronic blockade of vasopressin receptors in rats. J. Cardiovasc. Pharmac. 8: Suppl. 7, S56-60. Ishikawa S.-E., Kim J. K. and Schrier R. W. (1983) Further in vivo evidence for antagonist-to-antidiuretic action of arginine vasopressin. Am. J. Physiol. 245: R713-719. Ishikawa S.-E., Satio T. and Yoshida S. (1980) The effect of osmotic pressure and angiotensin II on arginine vasopressin release from guinea-pig hypothalamo-neurohypophyseal complex in organ culture. Endocrinology 106: 1571--1578. Jahr C. E. and Jesse11 T. M. (1983) ATP excites a subpopulation of rat dorsal horn neurones. Nature 304: 730-733.

Kimura H., MacGeer P. L., Reng J. H. and MacGeer E. G. (198 1) The central cholinergic system studied by choline acetyl-transferase immunohistochemistry in the cat. J. camp. Neural. 200: 151-201. KBnig J. F. R. and Klippel R. A. (1963) A stereotaxic atlas of the forebrain and lower parts of the brain stem, In: The Rat Brain. Williams and Wilkins, Baltimore. Lagercrantz H. and Stjlrne L. (1974) Evidence that most noradrenaline is stored without ATP in sympathetic large dense core nerve vesicles. Nafure 249: 843-845. Manning M. and Sawer W. H. (1985) Development of selective agonists and antagonists of vasopressin and oxytocin. In: Vasopressin (Schrier R. W., Ed.), pp. 131-144. Raven, New York. McNeil1 T. H. and Sladek J. R. Jr (1980) Simultaneous monoamine histofluorescence and neuropeptide immunocytochemistry. II Correlative distribution of catecholamine varicosities and magnocellular neurosecretory neurons in rat supraoptic aid paraventricular nuclei: J. camp. Neural. 193: 1023-1033. Mori M., Tsushima H. and Matsuda T. (1984) Antidiuretic effects of oxotremorine microinjected into the hypothalamic supraoptic and paraventricular nuclei in a waterloaded and ethanol-anesthetized rat. Jap. J. Pharmac. 35: 27-36.

592

M. Mom er al.

Mori M., Tsushima H. and Matsuda T. (1989) Effect of vasopressin antagonist on antidiuresis by oxotremorine microinjected into the hypothalamic supraoptic and paraventricular nuclei in a water-loaded and ethanolanesthetized rat. Jap. J. Pharmac. 49: 357-364. Reichlin S. (1981) Neuroendocrinology. In: Textbook of Endocrinology (Williams R. H., Ed.), pp. 589-645. Saunders, Philadelphia. Reid I. A. (1983) Salt and water regulation. In: Brain Peptides (Krieger D. T., Brownstein M. J. and Martin J. B., Eds), pp. 333-347. Wiley, New York. Salt T. E. and Hill R. G. (1983) Excitation of single sensory neurones in the rat caudal trigeminal nucleus by iontophoretically applied ATP. Neuroici. Left. 35: 53-58. Salter M. W. and Henrv J. L. (1985) Effects of cvclic AMP and ATP on functionally identified units in the cat spinal dorsal horn. Evidence for a differential effect of ATP on nociceptive versus non-nociceptive units. Neuroscience 15: 815-825.

Sawer W. H., Pang P. K. T., Seto J., McEnroe M., Lammek B. and Manning M. (1981) Vasopressin analogs that antagonize antidiuretic response by rats to the antidiuretic hormone. Science 212: 49-51. Schrier R. W. and Leaf A. (1981) Effects of hormones on water, sodium, chloride and potassium metabolism. In: Textbook of Endocrinology (Williams R. H., Ed.), pp. 1032-1046. Saunders, Philadelphia. Stjiirne L. and Astrand P. (1985) Relative pre- and postjunctional roles of noradrenaline and adenosine 5’-triphos-

phate as neurotransmitters of the sympathetic nerves of guinea-pig and mouse vas deferens. Neuroscience 14: 929-946. Swanson L. W. and Sawchenko P. E. (1983) Hypothalamic integration: organization of the paraventricular and supraoptic nuclei. A. Rev. Neurosci. 6: 269-324. Towstoless M. K., Congiu M., Coghlan J. P. and Wintour E. M. (1987) Placental and renal control of plasma osmolality in chronically cannulated ovine fetus. Am. J. Physiol. 253: R389-395.

Tsushima H., Mori M. and Matsuda T. (1985) Antidiuretic effects of alpha- and beta-adrenoceptor agonists microinjected into the hypothalamic supraoptic nucleus in a -water-loaded and -ethanol-anesthetized rat. Jap. J. Pharmac. 39: 365-374.

Tsushima H., Mori M. and Matsuda T. (1986) Antidiuretic effects of alpha- and beta-adrenoceptor agonists microinjected into the hypothalamic paraventricular nucleus in a water-loaded and ethanol-anesthetized rat. Jap. J. Pharmac. 40: 319-328.

Zimmerman E. A. (1983) Oxytocin, vasopressin and neurophysins. In: Brain Peptides (Krieger D. T., Brownstein M. J. and Martin J. B., MS), pp. 597-611. Wiley, New York. Zimmerman E. A., Hon-Yu A., Nilaver G. and Silverman A.-J. (1984) Anatomy of pituitary and extrapituitary vasopressin secretory systems. In: The Neurohypophysis, Physiological and Clinical Aspects (Reichlin S., Ed.), pp. 5-33. Plenum Medical Book, New York.