Regulation of [3H]mazindol binding to subhypothalamic areas: Involvement in glucoprivic feeding

Regulation of [3H]mazindol binding to subhypothalamic areas: Involvement in glucoprivic feeding

Brain Research Bulletin, 0361-9230/86 $3.00 + .OO Vol. 17, pp. 873-877, 1986. 0 Ankho International Inc. Printed in the U.S.A. Regulation of 3H]Ma...

660KB Sizes 4 Downloads 53 Views

Brain Research

Bulletin,

0361-9230/86 $3.00 + .OO

Vol. 17, pp. 873-877, 1986. 0 Ankho International Inc. Printed in the U.S.A.

Regulation of 3H]Mazindol Binding to Subhypothalamic Areas: Involvement in Glucoprivic Feeding [

ITZCHAK

ANGEL, ALEXANDER KISS, JILL A. STIVERS, LANA SKIRBOLL, JACQUELINE N. CRAWLEY AND STEVEN M. PAUL’

Clinical Neuroscience

Branch, National Institute of Mental Health, Bethesda,

MD 20892

ANGEL, I., A. KISS, J. A. STIVERS, L. SKIRBOLL, J. N. CRAWLEY AND S. M. PAUL. Regularion of[%Yjmazindol hind& to subhypo~hulamic nrem: Invohwmenf in glucoprivic feeding. BRAIN RES BULL 17(6) 873-877, 1986.-The distribution of low-affinity sodium-sensitive binding sites of [3H]mazindol were studied in rat hypothalamic nuclei. Using microdissection methods, it was demonstrated that the highest level of [3H]mazindol binding is localized to the paraventricular nucleus (PVN) and the lowest binding is observed in the lateral hypothalamus. Following food deprivation, a significant decrease in [3H]mazindol binding in the PVN and ventromedial hypothalamus (VHM) were observed. Refeeding food-deprived rats resulted in restoration of the level of binding in the PVN, and this was correlated with changes in blood glucose levels. Thus, changes in the binding of [3H]mazindol in the PVN may reflect local changes in glucose levels. In related studies, the involvement of the PVN in the regulation of food deprivation or 2-deoxyglucose (2-DG)-induced food intake was studied. Application of amphetamine (20 pg) into the PVN had no effect on food deprivation induced feeding, but significantly inhibited 2-DG induced (glucoprivic) feeding. The PVN may play an important role in the glucostatic regulation of feeding and in mediating the anorectic action of amphetamine and related anorectic drugs on glucoprivic feeding. Hypothalamus Anoretic drugs

Glucoprivic feeding

Paraventricular

nucleus

WE have previously demonstrated the presence of low affinity, high capacity binding sties for the anorectic compound [3H]mazindol in synaptosomal membranes derived from rat hypothalamus [2,3]. Specific binding was rapidly reversible, temperature sensitive, labile to pretreatment with proteolytic enzymes and inhibited by physiological concentrations of sodium [3]. The binding of [3H]mazindol was also unevenly distributed in various brain regions, the hypothalamus and brainstem having the highest levels of binding, and the cerebellum the lowest. The potencies of a series of phenylethylamine derivatives in inhibiting specific binding of [3H]mazindol in hypothalamic membranes correlates well with their anorectic potencies but not with their motor stimulatory effect, suggesting that these sites may be involved in the appetite suppressant actions of anorectic compounds [2,3]. In comparing a number of biochemical and pharmacological properties, it appears that the binding site for [3H]mazindol is similar, if not identical to those previously reported for L3Hlamphetamine [9,181. In order to investigate the possible role of these binding sites in the regulation of food intake, we have studied the sub-hypothalamic distribution of these binding sites and the effects of fooddeprivation in discrete hypothalamic regions. In previous ‘Requests for reprints should be addressed 4N214, Bethesda, MD 20892.

[3H]Mazindol binding

Amphetamine

studies, we have examined the regulation of the anorectic binding sites by glucose, both in vitro and in viva. We observed a stimulation of binding by parenteral administration of glucose, and a dose-dependent stimulation of binding by glucose in hypothalamic slices [Il. Furthermore, a good correlation between the levels of specific binding and blood glucose concentrations after feeding or food deprivation was also found [lo]. These data suggest that the anorectic binding sites as measured using either (+)[3H]amphetamine or [3H]mazindol may play an important role in the glucostatic regulation of food intake through specific hypothalamic sites. To study the direct CNS effects of anorectic drugs on feeding mediated by glucostatic signals, we employed 2-deoxyglucose-induced glucoprivation, which has been shown to selectively stimulate carbohydrate intake [ 121, and injected various anorectic drugs into discrete hypothalamic sites.

METHOD

Tissue Preparation

Adult

male

Sprague-Dawley

to Steven M. Paul, M.D., Chief, Clinical Neuroscience

873

rats (200-250

g) housed

Branch, NIMH, Building 10, Room

874

ANGEL ET AL.

under diurnal lighting conditions (12: 12) with free access to food and water were killed by decapitation and their brains rapidly removed. Hypothalamic nuclei were dissected by hand in ice-cold 0.32 M sucrose according to the method of Cue110 and Carson [7j. Rat brains were dissected on a linoleum plate, over cold 0.32 M sucrose solution to ensure iso-osmolarity and stability. Approximately 0.8-l mm thick slices were cut at A-P levels between - 1.8 and -3.5 mm from bregma. As shown in Fig. 1, the areas dissected contain in many cases more than one particular nucleus, and are therefore denoted as areas, rather than nuclei. The paraventricular area (PVN), contained both the magnocellular and parvocellular cells, and has residual fornix fibers. The anterior hypothalamus area (AH) contained also the periventricular nucleus. The lateral hypothalamic area (LH) was dissected in two different levels, and contained residual parts from the optic tract and supraoptic nucleus and fibers from the fornix and medial forebrain bundle. The dorsomedial hypothalamic area (DM or DMH) also contained residual posterior periventricular hypothalamic nucleus. The ventromedial hypothalamus (VMH) also contained the arcuate nucleus. Bilateral regions from 2-3 rats were pooled and homogenized in 4 ml of 0.32 M sucrose, using a glass-teflon homogenizer at moderate speed, centrifuged at 1OOOxg for 10 min and the supematant further centrifuged at 23,000 g for 20 min. The resulting crude synaptosomal pellet (P,) was gently resuspended in the same volume of sucrose and centrifuged at 23,OOOxg for an additional 20 min. The final pellet was resuspended in 40 vol of ice-cold 50 mM tris-HCl buffer (pH 7.4) using a Brinkman Polytron (set 7 for 5 set) and assayed immediately. Binding

Assay

The binding of [“Hlmazindol to crude synaptosomal membranes was assayed using a filtration assay as previously described [2,3]. Briefly, 8%200 pg of membrane protein (crude P, fraction), 50 ~1 of either unlabeled mazindo1 (600 PM) or buffer and 50 ~1 of [3H]mazindol (20-50 nM; specific activity 24.5 Ci/mmol, New England Nuclear, Boston, MA) were added in a total volume of 300 ~1. Following a 30 min incubation at 0-4°C the tubes were rapidly filtered and the radioactivity counted. Specific binding was defined as the difference between total binding and the binding in presence of unlabeled mazindol (100 PM). Protein was determined according to the method of Lowry et al. [ 161. Surgery

Rats were implanted with indwelling cannulae as previously described [5]. Coordinates aimed at the medial PVN were derived from Paxinos and Watson [19] and were: A-P -1.8 mm posterior and LAT +0.2 mm to bregma, VERT -7.3 mm from skull, incisor bar at -3.5 mm. Cannulae were implanted 1 mm dorsal to the intended site of injection. Examination of the cannulae tracing and placement were conducted histologically as previously described [5].

A. 1. Pamventricular nucleusIPVNI

, 2. AnteriorhypothabmusIAHI 13. Lateralhypothalamus(LHI

0. 4. DorsomedialhypothalamusIDMI 5. Ventromediilhypothalamus(VMHI i 6. LateralhypothalamusILH)

For the study of glucoprivic feeding, rats had free access to food and water prior to the experiment. All intracerebral injections were performed in awake rats as described, delivering 0.5 ~1 over a one minute period [5]. Rats were placed in single cages and 10 minutes after the intracerebral injection

I

,

; ,/’ 1 i i , 1

.1



FIG. 1. Diagramatic representation of the dissection of the different hypothalamic areas from the rat brain. Drawing taken from the Paxinos and Watson stereotaxic atlas [ 191. Areas were dissected from the fresh brain in the two levels demonstrated in A and B and were approximately I mm thick.

of either saline or amphetamine sulfate (20 pg) they were injected intraperitoneally with 300 mgikg of 2-deoxyglucose and returned to the cages. Thirty min after this injection Purina-chow pellets in a preweighed cup were introduced, and food intake was monitored every 30-60 min for up to 4 hr. For the study of intake after food deprivation, rats were food deprived for ls-24 hr (with free access to water) prior to the experiment. They then received intracerebral injections of drug or saline into the PVN, and were then placed in single cages. Food was present directly following injection and food intake was monitored every 30 min for 3-4 hr.

RESULTS In

Food Intuke

1

order

to

study

the

sub-hypothalamic

distribution

of

dissection method of Cuello and Carson [7]. Due to the low affinity and temperature sensitivity of this binding site 131, neither autoradiography nor binding in frozen tissue samples were successful (data not shown). The level of [3H]mazindol binding to crude synaptosomal membranes derived from the different brain regions is de[33H]mazindol

binding,

we

employed

the

fresh-tissue

[3H]MAZINDOL

BINDING

AND GLUCOPRIVIC

Ii VMH

875

FEEDING

ml

400

600

FIG. 2. Levels of specific [3H]mazindol binding in the different hypothalamic sites. Binding was processed as described in the Method section using 20 nM [3H]mazindol. Data represent mean+SEM quadruplicate determination of a representative experiment, repeated at least 4 times.

FIG. 3. Scatchard analysis of the effect of food deprivation (72 hr) on [3H]mazindol binding in the rat hypothalamus. Hypothalami from 4-6 rats were pooled together and assayed as described.

pitted in Fig. 2. Since there was no difference observed between the binding in the anterior and the posterior portion of the LH (data not shown), these were pooled and assayed together. As shown in Fig. 2, the lowest binding was obtained in the LH and the highest in the PVN. Intermediate levels of [3H]mazindol binding were observed in the other medial hypothalamic areas. Similar results were also obtained using (+)[3H]amphetamine binding (data not shown). To investigate the possible modulation of the [3H]mazindol binding site in the hypothalamus, rats were food deprived for different periods of time and the binding in both the whole hypothalamus and subhypothalamic regions was measured. As depicted in Fig. 3, food deprivation for 72 hr resulted in a significant reduction in [3H]mazindol binding in whole hypothalamus, e.g., the maximal number of binding sites (B,,,) was reduced from 851 pmol/mg protein to 487 pmol/mg protein, without a marked change in affinity (13.7 PM and 12.8 WM in fed control and food deprived, respectively). These results are in good agreement with the previously described changes in [3H](+)amphetamine binding following food deprivation [IO]. Analysis of the sub-hypothalamic sites (Table 1) revealed that the paraventricular nucleus had the largest change in [3H]mazindol binding (app. 30% reduction after 48 hr deprivation). A significant reduction was also observed in the ventromedial area. The dorsomedial area of the hypothalamus also appeared to have reduced binding, although this change was not statistically significant. Following further food-deprivation, up to 72 hr, some additional reduction in specific binding was observed in the PVN, but not in other areas (data not shown). When rats that were food deprived for 48 hr were allowed access to food for 3 hr, a reversal of the decreased binding in the PVN was observed. No significant changes were found in other hypothalamic areas (Table 1). In preliminary experiments (data not shown) we observed a good correlation between the level of binding in the PVN and blood glucose concentration. Blood glucose levels were reduced after food deprivation and elevated after refeeding. Similar results were previously obtained in whole hypothalamus of food deprived

and refed rats, using [3H](+)amphetamine as a radioligand [lo]. These results indicate that the PVN may be a site where both glucose and anorectic drugs produce a signal that may be related to the glucostatic regulation of feeding. In order to test this hypothesis, rats were cannulated with indwelling cannulae and the direct effects of amphetamine on this site were studied. Two different feeding-paradigms were used: (1) glucoprivic feeding, elicited in sated rats with 2-deoxyglucose and (2) food deprivation-induced feeding. It has been previously demonstrated that 2-DG, which causes cerebral cellular glucopenia, induces feeding which is specific for carbohydrates [ 121. Food deprivation, on the other hand, stimulates overall food intake. As demonstrated in Fig. 3, rats that were injected intraperitonally with 2-DG had significantly higher food consumption than control rats. The rate of food intake was linear for the first l-2 hours and then leveled off. Injection of 20 Fg of amphetamine sulfate (approx. 100 nmol) into the PVN of satiated control rats had no significant effect on basal feeding (data not shown) but significantly inhibited the feeding elicited by 2-DG (Fig. 3A). These data suggest that local application of amphetamine in the PVN is capable of directly blocking the glucoprivic signal induced by 2-DG. Since Leibowitz et al. have previously shown that the same dose of amphetamine does not reverse food deprivation-induced feeding at this site [ 131, it was important to be able to compare the effects of amphetamine on a different, non-selective feeding paradigm. As seen in Fig. 4B, rats that were food deprived for 24 hr consumed large quantities of food in a short time; however, amphetamine sulfate (20 pg), applied directly into the PVN was unable to block this feeding. Doses as high as 50 pg amphetamine sulfate were similarly ineffective (data not shown). When these experiments were carried out as in the 2-DG-induced feeding paradigm, (i.e., PVN application of drug followed 10 min later by IP injection of saline and introduction of food 30 min later), no difference in the feeding or amphetamine responses were observed (data not shown). In some experiments a significant increase in food intake after amphetamine application into the PVN was observed, similar to previously published observations

Bound

(pmollmg

protelnl

876

ANGEL ET AL. TABLE

1

EFFECT OF FOOD DEPRIVATION AND REFEEDING ON 13HlMAZ1NDOL BINDING IN HYPOTHALAMIC NUCLEI

Area

Level (pmolimg protein)

LH PVN AH VHM DMH

4578 9986 8435 8757 8313

i t t t i

271 557 553

Food-Deprived (48 hr) % control I12 Ir 77 2 92+ 89 :+ 77t

677

404

12 2”: 10 1” 12

Refed (3 hr) % control 1092 7 89-+ It 96? 5 90+ 2 882 IO

Data represent mean ? sem of 2-l 1 experiments, each done in quadruplicate determinations. using 20 nM I”H]mazindol. Rats either had access to ad lib food. food-deprived for 48 hr or fooddeprived for 48 hr and allowed access to food for 3 hr prior to sacrifice. *p
[13]. In other experiments, using the anorectic drugs fen~uramine and mazindol, similar resuits were obtained. These drugs, when injected into the PVN (100 nmol) significantly blocked 2-DG induced feeding, but did not affect food-deprivation

induced

feeding

in this site (Angel et al., in

preparation).

DISCUSSION

have previously demonstrated the presence of specific, low affinity [3H]mazindol binding sites in rat brain, that appear to mediate the anorectic properties of various phenylethylamine derivatives [2,3]. These binding sites are enriched in synaptosomal membranes and are presumably post-synaptic [4J. Further analysis of the regional distribution of these binding sites revealed that the hypothalamus had the highest number of sites, followed by the brainstem and striatum. In peripheral tissues such as the liver or kidney, very low levels of binding were observed. However, in the adrenal gland, relatively high levels of specific [3H]mazindoi binding are present and the characteristics of these sites seem to be very similar, if not identical, to the CNS site [22]. The pharmacological significance of the recognition site was investigated by determining the affinities of a series of anorectic drugs for these sites. A good correlation was observed between the potencies of a series of phenylethylamine derivatives in inhibiting hypothalamic [3H]mazindol binding and their potencies as anorectic drugs, but not as motor stimulants [2,3]. These data are consistent with, and further support the findings for the previously described (+)[3H]amphetamine binding site, [9], suggesting that the (+)[3H]amphetamine and [3H]mazindol binding sites may mediate the anorectic action of those drugs and may play a role in the physiological regulation of food-intake. In order to further relate the changes in binding with food We

mle

,hl

FIG. 4. Effect of 2-deoxyglucose (A) or food deprivation for 24 hr 1B) on food intake after amphetamine (20 sg) application into the PVN. (A) Rats were injected with 0.5 ~1 of either saline (0) or 20 pg of amphetamine sulphate (0) into the PVN as described. Ten min later they were introduced with pre-weighed Purina chow pellets and the food intake was monitored. 03) After the rats were food deprived for 24 hr, they were injected with 0.5 ~1 of either saline or amphetamine sulfate (20 pg) into the PVN as described, and food intake was monitored immediately. Control rats (0) had ad lib access to food and were injected with saline similarly.

intake we have investigated these changes in small areas of the hypothalamus. As demonstrated in Fig. 2, relatively high binding is observed in the medial hypothalamus and relative low levels in the lateral hypothalamus. The area with the highest binding was the PVN, which in all experiments had significantly higher specific [3H]mazindol binding followmedial hypothalamic areas. Furthermore, the PVN showed the most robust changes in [3H]mazindol binding following food deprivation and refeeding (Table 1). In related experiments (Angel ef trl., in preparation) we have found significant changes in [3H]mazindol binding in the PVN following 2-deoxyglucose administration as well as a good correlation between binding and blood glucose concentration. Previous studies have shown that the PVN is an important site of food-intake regulation [S]. Lesions of the PVN results in obesity, similar to the VMH overeating syndrome 1141. Lesions or knife cuts around the PVN disrupt the satiety action of intraperitoneally adminstered cholecystokinin [6]. Injection of norepinephrine [ 15,17), opioids [ 17,221 neuropepride Y [20] and other substances directly into the PVN stimulate feeding. The high level of [3HJmazindol binding in this site and its sensitivity to manipulations in feeding, may indicate the presence of specific receptors in this area that mediate the inhibition of food intake. Previous studies have indicated that the hy~thalamic binding site may be directly regulated by blood glucose levels and/or glucoprivic stimuli [ l,lO]. Consequently we have postulated that this recognition site may play an important role in the glucostatic regulation of feeding. In the present study, we have applied amphetamine directly into the PVN, and investigated its effect on glucoprivic feeding, elicited by 2-DC. Since it was previously demonstrated (131 that amphetamine failed to block fooddeprivation induced feeding when injected into the PVN, we compared the same dose of amphetamine on food deprivation-induced feeding as well. As demonstrated (Fig. 4A), a

[‘HJMAZINDOL

BINDING

AND GLUCOPRIVIC

FEEDING

robust and significant blockade of 2-DG induced feeding was observed by application of 20 Fg of amphetamine into the PVN. However, amphetamine injected into the PVN had no effect on food-deprivation induced feeding (Fig. 4B). These data support the hypothesis that the PVN could respond differently to different food-intake stimuli. It was previously demonstrated that feeding induced in the PVN by norepinephrine is specific for carbohydrate, and may be mediated through alpha-2 adrenergic receptors [ lS]. The present data

may suggest the presence of an anorectic recognition site in the PVN, which is specifically involved in the regulation of carbohydrate intake, an action which is antagonistic to the one mediated by alpha-2 receptors. Finally, the ability of fenfluramine and mazindol to decrease 2-DG induced feeding following local application into the PVN, further suggests that the inhibition of carbohydrate intake by anorectic drugs may involve a common mechanism.

REFERENCES 1. Angel, I., R. L. Hauger, M. D. Luu, B, Giblin, P. Skolnick and S. M. Paul. Glucostatic regulation of (+)[8H]amphetamine binding in the hypothalamus: correlation with Na+K+ATPase ac-

11. Hulian-Giblin, B., R. L. Hauger, A. Janowsky and S. M. Paul. Dopaminergic denervation increases [3H](+)-amphetamine binding in the rat corpus striatum. Submitted to Eur J Phar-

tivity. Proc Nat1 Acud Sci USA 82: 6320-6324, 1985. 2. Angel, 1. and S. M. Paul. Hypothalamic [3H]mazindol binding

mucol, 1986. 12. Kanarek, R. B., R. Marks-Kaufman, R. Ruthazer and L. Gualtieri. Increased carbohydrate consumption by rats as a function of 2-deoxy-D-glucose administration. Pharmacol Biochem Behav 18: 47-50, 1983. 13. Leibowitz, S. F. Amphetamine: possible site and mode of action for producing anorexia in the rat. Brain Res 84~ 160-167, 1975. 14. Leibowitz, S. F., N. J. Hammer and K. Chang. Hypothalamic paraventricular nucleus lesions produce overeating and obesity in the rat. Physiol Behav 27: 1031-1040, 1981. 15. Leibowitz, S. F., 0. Brown, J. R. Tretter and A. Kirschgessner. Norepinephrine, clonidine and tricyclic antidepressants selectively stimulate carbohydrate ingestion through noradrenergic system in the paraventricular nucleus. Pharmacol Biochem

correlates with the anorectic properties of phenylethylamines. 113: 133-134, 1985. 3. Angel, I., M. D. Luu and S. M. Paul. Characterization of f3H]mazindol binding in the rat brain: sodium sensitive binding correlates with the anorectic potencies of phenylethylamines. Submitted to J ~~~u~~r~e~. 4. Angel, I., A. Janowsky and S. M. Paul. The effects of serotonergic and dopaminergic lesions and sodium ions on [3H]mazindol binding in rat hypothalamus and corpus striatum. Submitted to Brain Ras, 1986. .5 . Crawley, J. N., J. A. Olschowa, D. I. Diz and D. M. Jacobowitz. Behavioral investigation of the coexistance of substance P, corticotropin releasing factor and acetylcholinesterase in lateral dorsal tegmental neurons projecting to the medial frontal cortex of the rat. Peptides 6: 891-901, 1985. 6. Crawley, J. N. and J. 2. Kiss. Paraventricular nucleus lesions abolish the inhibition of feeding induced by systemic cholecystokinin. Pcptides 6: 927-935, 1985. 7. Cuello, A. C. and S. Carson. Microdissection of fresh rat brain tissue slices. in: Bruin ~i~rodis.~e~tj~)n Techniques, edited by A. C. Cuello. Chichester: John Wiley and Sons, 1983, pp. 37125. 8. Gold, R. M., A. P. Jones and P. E. Sawchenko. Paraventricular area: Critical focus of a longitudal neurocircuity mediating food intake. Physiol Behav 18: 1111-I119, 1977. 9. Hauger, R. L., B. Hulihan-Giblin, P. Skolnick and S. M. Paul. Characterization of [3H](+)amphetamine binding sites in the rat central nervous system. Life Sci 34: 771-782, 1984. IO. Hauger, R. L., B. Hulihan-Giblin, P. Skolnick and S. M. Paul. Glucostatic regulation of hypothalamic and brainstem [3H](+)amphetamine binding during food deprivation and refeeding. Ear J Pllctrmocnl, 1986, in press. Eur J Phurmacol

Behav 23: 541-550, 1985. 16. Lowry, 0. H., N. J. Rosenbrough,

A. L. Farr and R. J. Randall. Protein measurements with the Folin phenol reagent. J Biol Chem 193: 165-175, 1951. 17. McLean, S. and B. G. Hoebel. Opiate and norepinephrineinduced feeding from the paraventricular nucleus of the hypothalamus are d&sociable. Life Sci 32: 2379-2382, 1982. _ 18. Paul. S. M.. B. Hulihan-Giblin and P. Skolnick. (+tAmphetamine.binding to rat hy~th~amus: relation to anorexic potency of phenylethylamines. Science 218: 487-490, 1982. 19. Paxinos, G. and C. Watson. The Rat Brain in Stereotaxic Coordinates. New York: Academic Press, 1982. 20. Stanley, B. G. and S. F. Leibowitz. Neuropeptide Y injected in the paraventricular hypothalamus: a powerful stimulant of feeding behavior. Proc Nat1 Acud Sci USA 82: 3940-3943, 1985. 21. Vocci, F., S. M. Paul and I. Angel. Characterization of 13H]mazindol binding in the rat adrenal gland. Submitted to Life Sci, 1986. 22. Woods, J. S. and S. F. Leibowitz.

Hypothalamic site sensitive to morphine and naloxone: effects on feeding behavior. Phar-

macol Biochem Behav 23: 431-438,

1985.