Brain Research 964 (2003) 128–135www.elsevier.com/ locate/brainres

Research report

D iabetes-induced neuroendocrine changes in rats: role of brainmonoamines, insulin and leptin

a a a bMatthew Barber , Badrinarayanan S. Kasturi , Maureen E. Austin , Kaushik P. Patel ,a a ,*Sheba M.J. MohanKumar , P.S. MohanKumar

aNeuroendocrine Research Laboratory, Department of Pathobiology and Diagnostic Investigation, College of Veterinary Medicine,Michigan State University, East Lansing, MI 48824,USA

bDepartment of Physiology and Biophysics, University of Nebraska Medical Center, Omaha, NE, USA

Accepted 20 November 2002

Abstract

Diabetes is characterized by hyperphagia, polydypsia and activation of the HPA axis. However, the mechanisms by which diabetesproduces these effects are not clear. This study was conducted to examine the effects of diabetes on the neuroendocrine system and to seeif treatment with insulin and/or leptin is capable of reversing these effects. Streptozotocin-induced diabetic adult male rats were subjectedto the following treatments: vehicle, insulin (2 U/day, s.c.), leptin (100mg/kg BW) or leptin1insulin every day for 2 weeks. Food intake,water intake, and body weight were monitored daily. We measured changes in monoamine concentrations in discrete nuclei of thehypothalamus at the end of treatment. Diabetes produced a marked increase in food intake and water intake and this effect was completelyreversed by insulin treatment and partially reversed by leptin treatment (P,0.05). Diabetes caused an increase in norepinephrine (NE)concentrations in the paraventricular nucleus with a concurrent increase in serum corticosterone. Treatment with insulin and leptincompletely reversed these effects. Induction of diabetes also increased the concentrations of NE, dopamine and serotonin in the arcuatenucleus and NE concentrations in the lateral hypothalamus, ventromedial hypothalamus (VMH) and suprachiasmatic nucleus (P,0.05).Although insulin treatment was capable of reversing all these changes, leptin treatment was unable to decrease diabetes-induced increasein NE concentrations in the VMH. These data provide evidence that hypothalamic monoamines could mediate the neuroendocrine effectsof diabetes and that insulin and leptin act as important signals in this process. 2002 Elsevier Science B.V. All rights reserved.

Theme: Endocrine and autonomic regulation

Topic: Neuroendocrine regulation: other

Keywords: Diabetes; Monoamine; Hypothalamus; Leptin; Insulin; HPLC–EC

1 . Introduction Changes in peripheral levels of glucose and insulin have adirect impact on these effects. Besides glucose and insulin,

Diabetes is a metabolic disorder that is known to hypothalamic neurotransmitters may also be involved inproduce various dysfunctions in the body including vascu- the regulation of these functions. Of the many neuro-lar disorders, retinopathy, cardiomyopathy, altered immune transmitters that are present in the hypothalamus, mono-functions, changes in intestinal functions, peripheral neuro- amines are likely candidates to be involved in this processpathy and dysfunctions of the central nervous system since they are known to modulate these functions. How-(CNS) [3,12,15]. Some of the diabetes-related CNS distur- ever, studies so far have not examined the role ofbances include hyperphagia, polydypsia and activation of monoamines in specific hypothalamic nuclei in precipi-the hypothalamo–pituitary–adrenal (HPA) axis [3,15]. tating diabetes-induced changes in central functions.

Besides neurotransmitters, a cascade of events involvinghormones, and other molecules probably mediate the*Corresponding author. Tel.:11-517-432-4680x34; fax:11-517-432-diabetes-induced dysfunctions in the body.7480.

E-mail address: mohankumar@dcpah.msu.edu(P.S. MohanKumar). The principal hormone that is affected in diabetes is

0006-8993/02/$ – see front matter 2002 Elsevier Science B.V. All rights reserved.doi:10.1016/S0006-8993(02)04091-X

129M. Barber et al. / Brain Research 964 (2003) 128–135

insulin. Insulin levels decrease rapidly upon induction of received recombinant rat leptin (100mg/kg BW/day, s.c.)diabetes and this favors the accumulation of glucose in the and Group IV (n58) received a combination of insulin (2extracellular space [1]. Both these molecules can cross the U/day, s.c.) and leptin recombinant rat (R&D systems,blood brain barrier to affect specific brain areas [23]. The Minneapolis, MN, USA; 100mg/kg BW/day, s.c.). Therebrain in turn is capable of detecting the rise and fall in was also a group of non-diabetic animals that received theglucose or insulin levels [3] and therefore these two vehicle for STZ, which did not receive any treatmentmolecules can act as signals to the brain indicating the (control). Food intake, water intake and body weight of allmetabolic status of the animal. The hormone leptin, which these animals were measured daily between 8 and 10 a.m.is involved in metabolic signaling, could be another for 14 days.important mediator in diabetes. Since diabetes is markedby reductions in the levels of both insulin and leptin, this

2 .4. Brain microdissectionstudy was done to see if treatment with leptin, insulin or acombination of the two could reverse some of the central

At the end of treatment, the animals were sacrificed andand neuroendocrine changes caused by diabetes.

their brains were removed quickly and frozen on dry ice.For this purpose, diabetes was induced in adult male rats

Trunk blood was collected and serum was separated andwith streptozotocin and the resulting diabetic animals were

stored at220 8C until analyzed by radioimmunoassaytreated with leptin, insulin or a combination of both. The

(RIA). Serial sections (300mm thick) of the brain wereeffects of these treatments on water intake, food intake,

obtained using a cryostat maintained at210 8C. Theserum corticosterone, and neurotransmitters in specific

sections were transferred to cover slips, which were placedareas of the hypothalamus were measured.

on a cold stage set at210 8C. The paraventricular nucleus(PVN), arcuate nucleus (AN), suprachiasmatic nucleus(SCN), supraoptic nucleus (SON), lateral hypothalamus

2 . Methods(LH) and the ventromedial hypothalamus (VMH) werelocated with the help of a stereotaxic atlas [19] and

2 .1. Animalsmicrodissected using the Palkovits’ microdissection tech-nique [14,17]. Tissue samples were obtained bilaterally

Adult male Sprague–Dawley rats were obtained fromand included all subdivisions of the nuclei and stored at

Harlan Sprague–Dawley, Inc. (Indianapolis, IN, USA) and270 8C. They were analyzed for norepinephrine con-

were housed individually in temperature- (2362 8C) andcentrations by high-performance liquid chromatography

light-controlled (lights on from 0500 to 1900 h) rooms.with electrochemical detection (HPLC–EC).

They were given food and water ad libitum. The animalswere used in the experiment as described below following10 days after their arrival. All the protocols followed in 2 .5. HPLC–ECthis study were approved by the University Committee forAnimal Care and Use at Michigan State University. The HPLC–EC system has been described before [14].

It consisted of a LC-4C amperometric detector (Bioanalyti-2 .2. Induction of diabetes using streptozotocin (STZ) cal Systems, West Lafayette, IN, USA), a phase II, 5mm

ODS reverse phase, C-18 column (Phenomenex, Torrance,Diabetes was induced by administration of a 2% solu- CA, USA), a glassy carbon electrode, a CTO-10 AT/VP

tion of STZ (Sigma, St. Louis, MO, USA) in cold 0.1 M column oven, and an LC-10 AT VP pump (Shimadzu,citrate buffer, pH 4.5 at a dose of 65 mg/kg BW given Columbia, MD, USA). The composition of the mobileintraperitoneally as described before [18]. Control animals phase was as follows: monochloroacetic acid (14.14 g/ l),received the vehicle (0.1 M citrate buffer, i.p.). The next sodium hydroxide (4.675 g/ l), octanesulfonic acid di-day, all of the animals were anesthetized using halothane sodium salt (0.3 g/ l), ethylenediaminetetraacetic acidand blood from the tail vein was used to measure blood (0.25 g/ l), acetonitrile (3.5%), and tetrahydrofuran (1.4%).glucose levels using a glucometer (Accucheck, Boehringer- The mobile phase was made in pyrogen-free water andMannheim, Indianapolis, IN, USA). Animals with blood then filtered and degassed through a Milli-Q purificationglucose levels of 150 mg/dl or higher were considered system (Millipore, Bedford, MA, USA) and pumped at adiabetic. flow rate of 1.8 ml /min. The sensitivity of the detector was

1 nA full scale, and the potential of the working electrode2 .3. Treatment was 0.65 V. The column oven maintained the temperature

of the column at 378C. At the time of HPLC analysis,Diabetic animals were divided into four groups and tissue samples were homogenized in 150ml of 0.1 M

treated as follows for 14 days. Group I (n58) served as the HClO using a micro-ultrasonic cell disruptor (Kontes,4

diabetic control and received the vehicle (s.c.). Group II Vineland, NJ, USA) and centrifuged at 10 0003g for 10(n58) received insulin (2 U/day, s.c.). Group III (n58) min. Fifty ml of the supernatant along with 25ml of the

M. Barber et al. / Brain Research 964 (2003) 128–135130

internal standard (0.05 M dihydroxybenzylamine) were increased their body weight from 296.764.3 on day 1 toinjected into the HPLC system. 343.463.2 on day 14 and was not different from the

control group. Treatment with both leptin and insulin also2 .6. Radioimmunoassay (RIA) increased the body weight of diabetic animals from

287.965.2 on day 1 to 336.966.6 on day 14 (P,0.05)Double antibody RIA was used to measure leptin and (Fig. 1). In contrast, treatment with leptin alone did not

corticosterone levels in the serum. RIA kits from Linco produce any significant changes in body weight comparedResearch (St. Charles, MO, USA) and the Coat-A-Count to the weight on day 1 (284.165.7).kit from Diagnostic Products Corp. (Los Angeles, CA,USA) were used to measure leptin and corticosterone 3 .2. Food intakelevels, respectively, as described previously [6].

Changes in food intake in the different groups over the2 .7. Statistical analysis entire observation period are shown in Fig. 2. The daily

food intake (mean6S.E.; g) of non-diabetic control ani-Changes in average daily food intake, body weight and mals on day 0 was 24.160.8 and remained at that level

water intake were analyzed by repeated measures ANOVA during the entire period of observation. In contrast, induc-followed by Fisher’s LSD test. Changes in neurotrans- tion of diabetes caused the food intake to increase frommitter concentrations, glucose, leptin and corticosterone 15.961 on day 1 to 44.861.8 on day 14. Food intake oflevels were analyzed by ANOVA followed by Fisher’s insulin-treated animals was similar to that seen in controlLSD test. animals. Treatment of diabetic animals with both leptin

and insulin also maintained the food intake around controllevels. However, leptin treatment alone attenuated dia-

3 . Results betes-induced hyperphagia, although not to control levels.After day 5, food intake in leptin-treated animals decreased

3 .1. Body weight significantly compared to diabetic animals until the end ofthe treatment period, although this was still higher than the

There were no significant differences in body weight control and insulin-treated animals (P,0.05).between the different groups at the beginning of treatment.Body weight (mean6S.E.; g) of non-diabetic control 3 .3. Water intakeanimals on day 0 was 312.665.5 and increased signifi-cantly to 352.168.1 on day 14. On the other hand, body Fig. 3 shows the changes in water intake in the differentweight of diabetic animals was 293.464.6 on day 1 and treatment groups during the entire period of observation. Inremained at that level during the rest of the period of non-diabetic control animals, water intake (mean6S.E.; oz)observation. Treatment of diabetic animals with insulin was 1.060.06 on day 1 and remained at that level during

Fig. 1. Changes in body weight after treatment of diabetic animals with the vehicle for leptin, or leptin (100mg/kg BW/day), or insulin (2 U/day) or acombination of leptin and insulin. Non-diabetic animals were used as controls. Body weight was measured daily after treatment of animals as describedabove.N 5 7–8 in each group; *P,0.05 as compared by repeated measures ANOVA.

131M. Barber et al. / Brain Research 964 (2003) 128–135

Fig. 2. Changes in food intake after treatment of diabetic animals with the vehicle, or leptin (100mg/kg BW/day), or insulin (2 U/day) or a combinationof leptin and insulin. Non-diabetic animals were used as controls. Food intake was measured daily after treatment of animals as described above. *P,0.05as compared by repeated measures ANOVA.

the next 13 days. Diabetes caused the water intake to (3.660.4) until the end of the treatment period (5.260.3),increase significantly from 1.360.2 on day 1 to 7.560.4 when compared to diabetic rats (P,0.05).on day 14. Treatment with insulin and the combination ofinsulin and leptin reversed this effect and brought water 3 .4. Blood glucose levelsconsumption to control levels. Treatment with leptin alonereduced the water intake significantly in diabetic animals, Mean (6S.E.) blood glucose levels (mg/dl; Fig. 4) atalthough not to control levels. In these animals, water the time of sacrifice in non-diabetic control animals wereintake decreased significantly from day 4 onwards 85.2564. In contrast, diabetes produced a marked increase

Fig. 3. Changes in water intake after treatment of diabetic animals with the vehicle for leptin, or leptin (100mg/kg BW/day), or insulin (2 U/day) or acombination of leptin and insulin. Non-diabetic animals were used as controls. Water intake was measured daily after treatment of animals as describedbefore. *P,0.05 as compared by repeated measures ANOVA. Note that treatment with leptin alone decreases water intake significantly from untreateddiabetic animals.

M. Barber et al. / Brain Research 964 (2003) 128–135132

treatment with leptin alone produced a three-fold increasein leptin levels (2.660.4) compared to diabetic animals,this was still significantly lower than the control andinsulin-treated groups.

3 .6. Serum corticosterone

Corticosterone levels (mean6S.E.; ng/ml; Fig. 6) indiabetic animals at the time of sacrifice were significantlyelevated (228.4665.3) compared to non-diabetic controlanimals (91.52611.7; P,0.01). Treatment with insulin,

Fig. 4. Changes in blood glucose after treatment of diabetic animals with and the combination of insulin and leptin brought corticos-the vehicle for leptin (D), or leptin (100mg/kg BW/day; L), or insulin (2 terone levels to 69.33622.5 and 111.8632.1, respectively.U/day; I) or a combination of leptin and insulin (D1L1I). Non-diabetic Interestingly, treatment with leptin alone also decreasedanimals were used as controls. Blood glucose levels were measured at the

serum corticosterone (122.1618.3) to control levels.end of 2 weeks of treatment at the time of sacrifice. *P,0.05 whencompared to control, I and D1L1I. Note that treatment with leptin alonedoes not decrease blood glucose levels compared to untreated diabetic3 .7. Monoamines in the hypothalamusanimals.

Monoamine concentrations in various hypothalamicnuclei are shown in Table 1. NE levels (mean6S.E.;

in blood glucose (391.4615; P,0.01) which was reversed pg/mg protein) in the PVN of non-diabetic control animalsby treatment with insulin or the combination of leptin and were 24.866.2 and increased by about 200% upon induc-insulin. However, treatment of diabetic animals with leptin tion of diabetes. Treatment with insulin decreased it toalone had little effect on blood glucose levels (346.7610). 26.663.8 and the combination of leptin and insulin

brought it down further to 16.463.0. Interestingly, treat-3 .5. Serum leptin ment of diabetic animals with leptin alone reduced NE

levels to 34.264.6 when compared to diabetic animalsAt the end of treatment, the leptin level (mean6S.E.; (P,0.05). One- to three-fold increases in NE levels were

ng/ml; Fig. 5) in non-diabetic control animals was observed in the AN, VMH, LH and SCN of diabetic4.460.2. Diabetes caused a marked decrease in serum animals when compared to control animals. While treat-leptin levels (0.960.2; P,0.01) and treatment with insulin ment with insulin, leptin or both reversed these changes inbrought leptin levels back to the control range (4.860.9). most areas, treatment with leptin alone was not sufficientTreatment of diabetic animals with both leptin and insulin to reverse the increase in NE levels in the VMH. Besidesproduced a further increase in leptin levels (12.462.1) NE, diabetes also caused an increase in dopamine (DA)compared to the rest of the groups (P,0.01). Although

Fig. 5. Changes in serum leptin levels after treatment of diabetic animals Fig. 6. Changes in serum corticosterone levels after diabetic animalswith the vehicle (D), or leptin (100mg/kg BW/day; L), or insulin (2 were treated with the vehicle (D), or leptin (100mg/kg BW/day; L), orU/day; I) or a combination of leptin and insulin (D1L1I). Non-diabetic insulin (2 U/day; I) or a combination of leptin and insulin (D1L1I).animals were used as controls. Serum leptin levels were measured at the Non-diabetic animals were used as controls. Serum corticosterone levelsend of 2 weeks of treatment at the time of sacrifice. *P,0.05 when were measured at the end of 2 weeks of treatment at the time of sacrifice.compared to control and I. Note that treatment with leptin alone does not *P,0.05 when compared to the other groups. Note that treatment withincrease leptin levels to that seen in control animals or those treated with leptin alone decreases corticosterone levels compared to untreated dia-insulin. betic animals.

133M. Barber et al. / Brain Research 964 (2003) 128–135

Table 1Changes in norepinephrine (NE), dopamine (DA), and serotonin (5-HT) in the paraventricular nucleus (PVN), arcuate nucleus (AN), ventromedialhypothalamus (VMH), lateral hypothalamus (LH), supraoptic nucleus (SON) and suprachiasmatic nucleus (SCN) in control, diabetic and diabetic animalstreated with insulin (I), or leptin (L) or a combination of both insulin and leptin.

Neurotransmitters Treatment groupsin hypothalamus

Control Diabetes D1insulin D1leptin D1I1L(pg/mg protein)

(D) (I) (L)

PVNNE 24.866.2 48.264.4* 26.663.8 34.264.6 16.463.0DA 5.261.5 6.360.8 8.061.8 4.260.8 2.960.95-HT 6.061.2 9.160.9 7.061.3 5.961.1 5.160.7

ANNE 28.866.6 55.468.5* 15.561.9 20.863.8 30.9610.3DA 7.862.1 22.262.5* 9.465.7 8.062.9 7.462.95-HT 8.262.0 14.763.4* 4.161.1 4.962.0 1.760.4

VMHNE 13.063.2 30.567.1* 11.364.8 30.865.4* 11.964.0DA 1.260.4 2.860.8 1.260.5 4.761.4 2.061.15-HT 10.962.5 6.961.1 7.161.3 7.461.2 10.562.3

LHNE 20.963.5 31.264.8* 16.361.0 17.362.0 19.863.2DA 5.762.1 9.864.2 3.660.6 5.262.1 4.661.15-HT 4.960.8 8.061.2 6.161.1 7.761.6 6.060.8

SONNE 23.761.8 29.562.6 23.062.4 24.962.5 19.963.4DA 1.860.6 1.760.8 1.560.9 0.960.2 1.260.45-HT 4.360.8 4.760.6 3.060.6 4.160.2 3.260.6

SCNNE 41.269.6 65.3611.3* 30.965.9 34.566.2 25.469.8DA 7.464.5 5.962.1 2.060.5 3.661.8 3.961.55-HT 6.261.2 10.162.5 5.161.6 7.061.2 5.661.4

*P,0.05 compared to rest of the groups;n57–8 in each group.

and serotonin (5-HT) levels in the AN. Treatment with to control levels. It totally reversed the increase in corticos-insulin, leptin or a combination of the two reversed these terone levels along with concurrent changes in NE con-changes. There were no changes in monoamine concen- centrations in the PVN. These results indicate for the firsttrations in the SON with diabetes induction. time that leptin, besides insulin, can ameliorate diabetes-

induced increases in water intake, corticosterone secretionand neurotransmitters in the hypothalamus.

4 . Discussion Diabetes is a metabolic disorder that is known toproduce changes in various organs of the body including

Diabetes produces marked changes in the neuroendoc- several central nervous system (CNS) disturbances such asrine system besides affecting a number of body functions neurobehavioral changes, autonomic dysfunctions, alteredin a significant manner. The body weight of diabetic neuroendocrine functions and neurotransmitter alterationsanimals decreased significantly on the second day and [3,15]. Until recently, a reduction in insulin levels and theremained at that level for the rest of the observation period. associated increase in circulating glucose levels wereOn the contrary, food intake and water intake increased believed to be the prime peripheral signals linked to thissignificantly during the 2 weeks of observation. The disorder. With the discovery of leptin in 1994 [28], and thedecrease in body weight was followed by a significant observation that peripheral leptin levels also decrease inreduction in serum leptin levels by the end of 2 weeks. On diabetes [7], it has become clear that leptin could bethe other hand, there was a significant increase in corticos- another important mediator of diabetes-induced centralterone levels in these animals. Administration of insulin effects. In fact, leptin has been associated with a number ofcorrected all these changes associated with diabetes. The central functions including feeding. While increases ininteresting observation was that administration of recombi- circulating leptin levels are associated with reduced feed-nant rat leptin was capable of alleviating many of these ing [4,20], a decrease in leptin levels is expected tochanges. Leptin administration decreased both food intake increase food intake. This is probably the case in diabetes.and water intake significantly in diabetic rats, although not In fact, a recent study showed that replacing leptin in

M. Barber et al. / Brain Research 964 (2003) 128–135134

diabetic animals could indeed reverse diabetes-induced water intake down to non-diabetic levels. The mechanismhyperphagia [24]. In the present study, with the dose of by which diabetes causes polydipsia is unclear. Regulationleptin that we used (100mg/kg/day) we were able to of fluid balance in the body is a highly complicatedproduce a significant reduction in the hyperphagia com- process and all aspects of this phenomenon have not beenpared to diabetic animals, although it did not reach control studied in great detail. One of the possible neurochemicalslevels. This may be due to the fact that serum leptin levels that may have an important effect on this function iswere about 3- to 10-fold less in animals that were treated arginine vasopressin. The cell bodies that secrete thiswith leptin alone when compared to control animals and neurohormone are located in the PVN, the SON and thethose treated with insulin. This leads us to believe that the SCN and are in turn influenced by a variety of neuro-degree of hyperphagia in diabetes can be reduced substan- chemicals including NE, opiates, etc. [23]. Although NEtially by increasing peripheral leptin levels. Since the concentrations in the PVN and SCN increased withinsulin-treated animals had high levels of leptin, it raises diabetes, the effect on arginine vasopressin and otherthe possibility that the beneficial effects of insulin on neurochemicals needs further study.diabetes-induced hyperphagia may be mediated through It is well established that there is an increase in HPAleptin. activity in diabetes [12,15]. This has been marked by

The mechanism by which diabetes causes hyperphagia is increases in CRH mRNA in the PVN and increases inunclear. A number of neurochemicals and neuropeptides, serum ACTH and corticosterone levels [5,22]. Resultssuch as opiates, NPY, AGRP, CCK, CGRP etc., could be from our study are consistent with these reports and showinvolved in this phenomenon [11,16,25]. NE levels in the an increase in corticosterone levels with the induction ofPVN could be another important factor since it is known to diabetes. Treatment with insulin completely reversed thisregulate feeding behavior in animals [9]. Destruction of the effect. The interesting feature, however, was the ability ofPVN by electrolytic lesioning or injections of neurotoxins leptin to block the increase in corticosterone observed withcan abolish drug-induced stimulation of feeding [10,27]. diabetes. The mechanism by which diabetes increasesOn the contrary, NE injections into the PVN increases serum corticosterone could involve an increase in NEcirculating levels of vasopressin and glucose which are levels in the PVN. This is because the PVN of thebelieved to act synergistically with NE to potentiate hypothalamus has a large number of CRH cell bodies andfeeding [8]. Taken together, these studies clearly indicate receives rich noradrenergic innervation from the brain stemthat an increase in NE activity in the PVN stimulates through the ventral noradrenergic bundle [21]. Whilefeeding. The question now is whether diabetes affects NE neurotoxic blockade of the ventral noradrenergic bundlelevels in the PVN and if insulin and leptin can reverse causes reduction in CRH, direct administration of NE intothese changes. NE levels increased in the PVN of diabetic the PVN can stimulate CRH secretion [26]. Thus, it isanimals but decreased when these animals were treated clear that NE levels in the PVN play a stimulatory role inwith insulin or leptin alone or with the combination of CRH secretion and could therefore mediate the effects ofleptin and insulin. This could be a possible mechanism by diabetes. In the present study, the diabetes-induced in-which insulin and leptin decrease diabetes-induced hy- crease in serum corticosterone levels was accompanied byperphagia in these animals. an increase in NE concentrations in the PVN and a

Besides the PVN, a few other hypothalamic areas such decrease in serum leptin levels. When exogenous insulin oras AN, VMH, LH and SCN have also been implicated in leptin was administered to diabetic animals it caused afeeding [13]. All these areas also showed marked increases stark reduction in NE concentrations in the PVN and thisin NE levels with diabetes. In the VMH, however, treat- was accompanied by a reduction in corticosterone levels.ment with leptin alone was not sufficient to bring NE to This indicates that NE levels in the PVN could play ancontrol levels. This agrees well with our results on food important role in mediating the subduing effects of insulinintake, since leptin treatment did not totally reverse the and leptin on corticosterone levels in diabetic rats. Besides,diabetes-induced increase in food intake. In the AN, NE and other neurochemicals may also be involved in thisdiabetes also produced an increase in DA and 5-HT. This phenomenon. In contrast to the effect observed with foodeffect was totally reversed by treatment with insulin, leptin intake, water intake and serum corticosterone, treatment ofor a combination of both. The roles of these neuro- diabetic rats with leptin failed to decrease blood glucose.transmitters and the function of these nuclei in diabetes There are several lines of evidence to indicate the contraryneed to be studied further. [2]. The failure of leptin treatment to decrease blood

Following the effects on food intake, diabetes also glucose levels in the present study may be attributed to thecaused an increase in water intake starting from day 2. As dose of leptin and the mode of administration that wasexpected, treatment with insulin completely reversed water used. This is supported by our observations with food andintake. Once again, the water intake in animals treated with water intake in this study.leptin alone did not drop down to the level seen in control In summary, diabetes not only produced changes inor insulin-treated animals. This is another indicator that the body weight, food intake, and water intake but also hasdose of leptin that we used was not sufficient to bring the marked effects on the neuroendocrine system. These

135M. Barber et al. / Brain Research 964 (2003) 128–135

[11] C.G. MacIntosh, J.E. Morley, J. Wishart, H. Morris, J.B. Jansen, M.effects were completely reversed by insulin treatment,Horowitz, I.M. Chapman, Effect of exogenous cholecystokininwhile treatment with leptin produced partial to complete(CCK)-8 on food intake and plasma CCK, leptin, and insulin

reversal. It is unclear at the present time if leptin treatment concentrations in older and young adults: evidence for increasedcould provide long-term benefits. This will be the focus of CCK activity as a cause of the anorexia of aging, J. Clin.future studies. Endocrinol. Metab. 86 (2001) 5830–5837.

[12] A.L. McCall, The impact of diabetes on the CNS, Diabetes 41(1991) 557–570.

[13] M.L. Moal, P. Mormede, L. Stinus, The behavioral neuroendocrinol-A cknowledgements ogy of arginine vasopressin, adrenocorticotropic hormone and

opioids, in: C.B. Nemeroff (Ed.), Neuroendocrinology, CRC Press,This study was partially supported by NIH AG05980 to Boca Raton, FL, 1992, pp. 365–396.

[14] S.M.J. MohanKumar, P.S. MohanKumar, S.K. Quadri, Specificity ofP.S.M. Matthew Barber and Maureen Austin were partiallyinterleukin-1b-induced changes in monoamine concentrations insupported by a Merck Merial Summer Research Fellow-hypothalamic nuclei: blockade by interleukin-1 receptor antagonist,

ship for Veterinary Students at Michigan State University. Brain Res. Bull. 47 (1998) 29–34.[15] A.D. Mooradian, Diabetic complications of the central nervous

system, Endocr. Rev. 9 (1988) 346–356.[16] G.J. Morton, M.W. Schwartz, The NPY/AgRP neuron and energyR eferences

homeostasis, Int. J. Obes. Relat. Metab. Disord. Suppl. 5 (2001)S56–S62.

[1] J.D. Bagadade, E.L. Bierman, D. Porte, The significance of basal [17] M. Palkovits M, Isolated removal of hypothalamic or other braininsulin levels in the evaluation of the insulin response to glucose in nuclei of the rat, Brain Res. 59 (1973) 449–455.diabetic and nondiabetic subjects, J. Clin. Invest. 46 (1967) 1549– [18] K.P. Patel, P.L. Zhang, Reduced renal sympathoinhibition in re-1557. sponse to acute volume expansion in diabetic rats, Am. J. Physiol.

[2] N. Barzilai, J. Wang, D. Massilon, P. Vuguin, M. Hawkins, L. 267 (1994) R372–R379.Rossetti, Leptin selectively decreases visceral adiposity and en- [19] G. Paxinos, C. Watson, The Rat Brain in Stereotaxic Co-ordinates,hances insulin action, J. Clin. Invest. 100 (1997) 3105–3110. 2nd Edition, Academic Press, New York, 1987.

[3] G.J. Biessels, A.C. Kappelle, B. Bravenboer, D.W. Erkelens, W.H. [20] M.A. Pelleymounter, M.J. Cullen, M.B. Baker, R. Hecht, D.Gispen, Cerebral function in diabetes mellitus, Diabetologia 37 Winters, T. Boone, F. Collins, Effects of the obese gene product on(1994) 643–650. body weight regulation in ob/ob mice, Science 269 (1995) 540–

[4] L.A. Campfield, F.J. Smith, Y. Guisez, R. Devos, P. Burn, Recombi- 543.nant mouse OB protein: evidence for a peripheral signal linking [21] P. Petrusz, I. Merchenthaler, The corticotropin-releasing factoradiposity and central neural networks, Science 269 (1995) 546–549. system, in: C.B. Nemeroff (Ed.), Neuroendocrinology, CRC Press,

[5] O. Chan, S. Chan, K. Inouye, M. Vranic, S.G. Matthews, Molecular Boca Raton, FL, 1992, pp. 129–184.regulation of the hypothalamo–pituitary–adrenal axis in strep- [22] F.P. Pralong, R.C. Gaillard, Neuroendocrine effects of leptin,tozotocin-induced diabetes: effects of insulin treatment, Endocrinol- Pituitary 4 (2001) 25–32.ogy 142 (2001) 4872–4879. [23] M.W. Schwartz, D.P. Figlewicz, D.G. Baskin, S.C. Woods, D.D.

[6] J. Francis, S.M.J. MohanKumar, P.S. MohanKumar, Correlations of Porte Jr., Insulin and the central regulation of energy balance:norepinephrine release in the paraventricular nucleus with plasma update, Endocr. Rev. Monogr. 2 (1994) 109–113.corticosterone and leptin after systemic lipopolysaccharide: blockade [24] D.K. Sindelar, P.J. Havel, R.J. Seeley, C.W. Wilkinson, S.C. Woods,by soluble IL-1 receptor, Brain Res. 867 (2000) 180–187. M.W. Schwartz, Low plasma leptin levels contribute to diabetic

[7] P.J. Havel, J.Y. Uriu-Hare, T. Liu, K.L. Stanhope, J.S. Stern, C.L. hyperphagia in rats, Diabetes 48 (1999) 1275–1280.Keen, B. Ahren, Marked and rapid loss of circulating leptin in [25] T.W. Stephens, Fat regulation. Life without neuropeptide Y, Naturestreptozotocin diabetic rats: reversal by insulin, Am. J. Physiol. 274 381 (1996) 377–378.(1998) R1482–R1491. [26] A. Szafarczyk, V. Guillaume, B. Conte-Devolx, G. Alonso, P.

[8] S.F. Leibowitz, Hypothalamic paraventricular nucleus: interaction Malaval, N. Pares-Herbute, C. Oliver, I. Assenmacher, Centralbetween alpha 2-noradrenergic system and circulating hormones and catecholaminergic system stimulates secretion of CRH at differentnutrients in relation to energy balance, Neurosci. Biobehav. Rev. 12 sites, Am. J. Physiol. 255 (1988) E463–E468.(1988) 101–109. [27] G.F. Weiss, S.F. Leibowitz, Efferent projections from the paraven-

[9] S.F. Leibowitz, L.L. Brown, Histochemical and pharmacological tricular nucleus mediating alpha 2-noradrenergic fooding, Brain Res.analysis of noradrenergic projections to the paraventricular hypo- 347 (1985) 225–238.thalamus in relation to fooding stimulation, Brain Res. 201 (1980) [28] Y. Zhang, R. Proenca, M. Maffei, M. Barone, L. Leopold, J.M.289–314. Friedman, Positional cloning of the mouse obese gene and its human

[10] S.F. Leibowitz, N.J. Hammer, K. Chang, Fooding behavior induced homologue, Nature 372 (1994) 425–432.by central norepinephrine injection is attenuated by discrete lesionsin the hypothalamic paraventricular nucleus, Pharmacol. Biochem.Behav. 19 (1983) 945–950.