OfinaMiRcfinonBbaCt

DuwatwHa

ct

FsKJ

P

PH

N

BASIC SCIENCE

9

MODULATION BY DESMOPRESSIN OF NEURONALACTIVITY IN BRAINSTEM MICTURITION CENTER

HIROSHI IWASAKI, YOSHIMASA KOYAMA, YOSHIYUKI TANAKA, AKIHIRO KAWAUCHI,EIICH JODO, YUKIHIKO KAYAMA, AND TSUNEHARU MIKI

ABSTRACTbjectives. To examine the effect of desmopressin (DDAVP) on bladder contraction and on the neurons thatre in relation to spontaneous bladder contraction (bladder-related neurons) in and around Barrington’sucleus, the micturition center. DDAVP is used for the treatment of nocturnal enuresis because of itsntidiuretic action, but the mechanism of this action has not been proved.ethods. Urethane-anesthetized Sprague-Dawley male rats (n � 20) were used. DDAVP was infused

ntravenously or as an intracerebroventricular infusion into the lateral ventricle.esults. We encountered three types of bladder-related neurons: those that fired before the start of theontraction (type E1), those that fired synchronous with the bladder contraction (type E2), and those thatred during bladder relaxation (type I). Intravenous infusion caused inhibition in three of five type E1eurons, excitation in two of five type E2 neurons, and excitation (one neuron) and inhibition (one neuron)f four type I neurons. With intracerebroventricular infusion into the lateral ventricle, two of four type E1eurons were inhibited, and one of seven type E2 neurons and three of four type I neurons were excited.ladder contraction was suppressed in 4 of 12 rats by intravenous infusion and in 2 of 8 rats by intracere-roventricular infusion into the lateral ventricle. In all cases, when the bladder contraction was suppressed,n electroencephalogram of larger amplitude and slower frequency appeared.onclusions. DDAVP seems to regulate bladder activity by affecting bladder-related neurons in the mic-urition center. UROLOGY 63: 994–998, 2004. © 2004 Elsevier Inc.

arcbteamu

mDiaaasmcp

esmopressin (DDAVP) a derivative of argi-nine vasopressin (AVP), has recently been

sed as a drug for nocturnal enuresis.1 Comparedith AVP, which has vasopressor and antidiuretic

ctions, DDAVP has a stronger antidiuretic ac-ion,2 rather than affecting the blood pressure,3hich is one of the reasons it prevents bed wetting.owever, the mechanisms of DDAVP on bladder

ctivity have not been elucidated.We recorded several types of neurons that dis-

harge in relation to spontaneous contraction ofhe urinary bladder (bladder-related neurons) in

rom the Department of Physiology, Fukushima Medical Univer-ity School of Medicine, Fukushima; and Department of Urology,yoto Prefectural University of Medicine, Kamigyo-ku, Kyoto,

apanH. Iwasaki is on leave from the Department of Urology, Kyoto

refectural University of Medicine.Reprint requests: Yoshimasa Koyama, M.D., Department of

hysiology, Fukushima Medical University School of Medicine, 1akari-ga-oka, Fukushima 960-1295, JapanSubmitted: September 2, 2003, accepted (with revisions):

tovember 25, 2003

© 2004 ELSEVIER INC.94 ALL RIGHTS RESERVED

nd around the brainstem micturition center, Bar-ington’s nucleus.4 Most neurons in this nucleusontain corticotropin-releasing hormone, which iselieved to be involved in the regulation of mic-urition.5 In the present study, we examined theffect of DDAVP on these bladder-related neuronsnd on bladder contraction and discuss the centralechanisms of DDAVP for the prevention of en-resis.

MATERIAL AND METHODS

Experiments were performed on 20 adult Sprague-Dawleyale rats weighing 250 to 500 g (obtained from Japan SLC).DAVP was infused intravenously (IV) in 12 rats or as an

ntracerebroventricular (ICV) infusion in 8 rats. They werenesthetized with urethane (1.0 g/kg, intraperitoneally), anddditional doses of the same anesthetic were given to maintainn appropriate level of anesthesia. The head was fixed in atereotaxic instrument, and stainless steel bolts (diameter 1.0m) were screwed into the skull over the frontal and parietal

ortices to record the electroencephalogram. The rectal tem-erature was kept at 37°C with a heating pad controlled

hrough a feedback circuit.

0090-4295/04/$30.00doi:10.1016/j.urology.2003.11.036

siteTsPbd

ei�Imf3tit

pcTaEt(sr

w3fwootocw

AoJL

ttAdcrwa

tro(cc

it

rsrJcsc(rttsoAsuc(ocmtHpOeOw

Ftcpsottt(

U

A polyethylene catheter (outer diameter 0.96 mm) was in-erted from the dome into the urinary bladder through a smallncision in the lower abdominal wall. The urethra was liga-ured to prevent voiding. The catheter had a bifurcation; onend was connected to a pressure transducer (Nihon-Koden,P400-T) and the other to a syringe filled with physiologicaline that was injected using an infusion pump (Harvard,ump 11) into the bladder at a rate of 0.2 mL/min, until reflexladder contractions under isovolumetric conditions were in-uced (range 0.4 to 1.0 mL).The IV infusion of DDAVP was performed through a poly-

thylene catheter (outer diameter 0.61 mm) that was insertednto the right inner jugular vein. DDAVP (10 �g/mL, 8.45M) was infused manually at a rate of about 0.05 mL/s. For the

CV infusion, a stainless guide cannula (outer diameter 0.7m) was directed to the lateral ventricle (1.0 mm posterior

rom the bregma, 1.6 mm lateral from the midline, and about.2 mm from the surface). DDAVP (84.5 �M) was infusedhrough an inner cannula (outer diameter 0.35 mm) insertednto the guide cannula by a syringe pump (Nihon-Koden) athe rate of 0.02 �g/min (1 �L/min).

Single neuronal activity was recorded through a glass pi-ette microelectrode filled with 0.5 M sodium-acetate solutionontaining 2% pontamine sky blue (impedance 10 to 25 M�).o avoid penetration of the venous sinus, the electrode wasngled posteriorly at 30° and inserted through the cerebellum.lectroencephalography, bladder pressure, and neuronal ac-

ivity data were fed into a computer using a CED interfaceCambridge Electronic Design, Cambridge, UK). Recordingites were marked by ejection of pontamine sky blue from theecording electrode.

After the recording, the animal was deeply anesthetizedith pentobarbital until dead and perfused transcardially with00 mL of physiologic saline followed by 300 mL of 4% para-ormaldehyde in 0.1 M phosphate buffer (pH 7.4). The brainas removed from the skull and postfixed in the same solutionvernight, immersed in 30% sucrose for several hours, and cutn a freezing microtome at 50 �m in the frontal plane. To stainhe cholinergic neurons, the sections were processed for nic-tinamide adenine dinucleotide phosphate-diaphorase, a spe-ific marker for brainstem cholinergic neurons.6 The sectionsere then counterstained by neutral red.The present study was carried out under the guidance of the

nimal Research Committee in accordance with the Guidelinen Animal Experiments in Fukushima Medical University andapanese Government Animal Protection and Managementaw.

RESULTS

A total of 168 neurons were encountered in theegmental area from the caudal mesencephalon tohe rostral pons, including Barrington’s nucleus.bout one third of them (n � 59, 35%) showedischarges modulated in relation to spontaneousontraction of the bladder (bladder-related neu-ons). Of the 59 bladder-related neurons, activityas maintained in 29 neurons long enough to ex-

mine the effect of DDAVP.According to the findings of our previous work,

hese neurons were divided into three types4: neu-ons that started firing before and finished beforer soon after the onset of bladder contractionstype E1), neurons that started firing before andontinued firing synchronous with the bladder

ontractions (type E2), and neurons that fired dur- n

ROLOGY 63 (5), 2004

ng bladder relaxation and showed a firing patternhe mirror image of the type E2 neurons (type I).About one half of the bladder-related neurons

esponded to IV infusion of DDAVP. Figure 1Ahows an example of a response of a type E1 neu-on to IV infusion of 1.5 �g (8.45 �M) DDAVP.udging from the time-expanded trace, this neuronould be classified as type E1 (Fig. 1B,C). After theecond contraction (Fig. 1A, arrow), which oc-urred about 80 seconds after the DDAVP infusionFig. 1A, triangle), the neuronal firing, which cor-elated well with the bladder contractions beforehe infusion, disappeared for about 3 minutes. Af-erward, the rhythmic firing seen before the infu-ion recovered. Similar inhibitory responses werebserved in two more of the five type E1 neurons.remarkable decrease in the firing rate was ob-

erved for 90 seconds in one neuron and for 5 min-tes in the other one. The neurons in Figure 2A,lassified as type E2 by the time-expanded scaleFig. 2B,C), increased its firing rate from 150 sec-nds after 1.5 �g of infusion (arrow). The in-reased firing continued for 11 minutes, with aean firing rate of 2.6 Hz, more than 50% greater

han that for 8 minutes before the infusion (1.7z), and then decreased to lower than during thereinfusion period (0.7 Hz) for about 4 minutes.f the five type E2 neurons, a second one was

xcited, and the remaining three were not affected.f four type I neurons, one was inhibited, and oneas excited (data not shown). In total, 7 of 14

IGURE 1. (A) Example of response of type E1 neurono IV infusion of 1.5 �g DDAVP and simultaneous re-ording of electroencephalogram (EEG) and bladderressure (BP). Rate � firing rate per second of spikes;pike � single neuronal spikes. Triangle indicates pointf infusion of DDAVP. Arrow indicates second contrac-ion after DDAVP infusion (see text). Dots indicate elec-rical artifacts when animal was touched. (B) Expandedime scale of activities during bar in leftmost part of (A).C) Averaged shape of action potential.

eurons responded to IV infusion of 1.5 to 7.5 �g

995

Ds

Itsf�aDnw3

rrtsam

bmdrtmga

fpemota

e(rSnAip

FtcpDcap

FIcpsaso

FDictepirtc

9

DAVP. In the remaining 7, no response was ob-erved after infusion of 1.8 to 14.5 �g.In each type of neuron, the response profile to

CV infusion was similar to that with IV infusion;wo of four type E1 neurons were inhibited, one ofeven type E2 neurons was excited, and three ofour type I neurons were excited after 0.04 to 0.12g of ICV infusion. Figure 3A shows an example ofresponse of a type I neuron to ICV infusion ofDAVP for 3 minutes (shown by broken bar). Thiseuron, showing a firing pattern that correlatedith bladder relaxation before infusion (Fig.

IGURE 2. (A) Example of response of type E2 neurono IV infusion of 1.5 �g DDAVP and simultaneous re-ording of electroencephalogram (EEG) and bladderressure (BP). Triangle indicates point of infusion ofDAVP. Dots indicate electrical artifacts. Arrow indi-ates start of excitation. (B) Expanded time scale ofctivities during bar in (A). (C) Averaged shape of actionotential.

IGURE 3. (A) Example of response of type I neuron toCV infusion of 0.06 �g DDAVP and simultaneous re-ording of electroencephalogram (EEG) and bladderressure (BP). Broken bar indicates period of ICV infu-ion of DDAVP. Arrow indicates third relaxation periodfter infusion of DDAVP (see text). (B) Expanded timecale of activities during bar in (A). (C) Averaged shapef action potential.

B,C), slightly increased its firing rate at the third c

96

elaxation period after the infusion (Fig. 3A, ar-ow) and continued to fire even during the con-raction period thereafter. The firing rate, mea-ured during the last 3 minutes (3.4 Hz), waslmost three times greater than that during the 3inutes before the infusion (1.1 Hz).IV infusion of DDAVP (1.5 to 6.0 �g) suppressed

ladder contractions in 4 of 12 rats for 5 to 13inutes. The suppression was accompanied with a

ecrease (not shown) or increase (Fig. 2) in neu-onal activity. After ICV infusion, the bladder con-raction was suppressed in 2 of 8 rats for 1 to 3inutes. On such occasions, an electroencephalo-

ram of a larger amplitude and slower frequencyppeared.The bladder-related neurons were distributed

rom the caudal mesencephalon to the rostralons, including Barrington’s nucleus, the lat-rodorsal tegmental nucleus, and the reticular for-ation ventral to the periaqueductal gray, which

verlapped with that shown previously.4 However,he localization of each type of neuron in specificreas was not recognized.We previously showed that the neurons that gen-

rated an action potential longer than 0.76 msbroad-spike neurons) were cholinergic, and theest (brief-spike neurons) were noncholinergic.7

imilarly, it has been reported that broad-spikeeurons in the locus coeruleus are noradrenergic.8

s shown in Figures 1C to 3C, all neurons shownn Figure 4 were brief-spike neurons. Thus, we canresume confidently that these neurons were non-

IGURE 4. Locations of neurons on which effect ofDAVP was examined. Of 29 neurons, 26 histologically

dentified neurons were plotted on diagrams of serialoronal sections around central gray of mesencephalono pons at intervals of 300 �m. Small dots representxact distribution of nicotinamide adenine dinucleotidehosphate-diaphorase-positive (ie, cholinergic) neurons

n one animal. Circles indicate type E1 neurons; quad-angles indicate type E2 neurons; upward triangles,ype I neurons. Bar � Barrington’s nucleus, LC � locusoeruleus; LDT � laterodorsal tegmental nucleus.

holinergic and non-noradrenergic.

UROLOGY 63 (5), 2004

wiirD1e(bitflwEirtTmib

srsarepDbnroottstsBtssrsd

sVrobTh

DabcbricbDntop

wrspa0ttofhsrcgt

htaTotMDtt

itrspantvFvcs

U

COMMENT

The effect of DDAVP on bladder-related neuronsas observed in 7 (50%) of 14 neurons after IV

nfusion and in 6 (40%) of 15 neurons after ICVnfusion. No increase in blood pressure has beeneported with IV infusion of 0.5 to 8.0 �g/kg ofDAVP in rats.9 In the preliminary experiment,.5 �g of DDAVP, which was used in the presentxperiment as a single IV infusion, induced a smallless than 10 mm Hg), but prompt, increase inlood pressure. However, the time course of thencrease was so fast and quite different from that ofhe changes in neuronal firing. Therefore, the in-uence of blood pressure on the recorded neuronsould be small, if any. Each type of neuron (E1,2, and I) showed a similar response to both IV

nfusion and ICV infusion, that is, the type E1 neu-on response was inhibitory, and the type E2 andype I neuron responses were mainly excitatory.hese facts suggest that DDAVP, which was ad-inistered through different routes, acts in a sim-

lar way on the bladder-related neurons in therainstem.On the basis of the findings of our previous

tudy, we believe that of the three types of bladder-elated neuron activity recorded in the presenttudy, type E2 neurons may directly regulate thectivity of the bladder, and type E1 and type I neu-ons modulate the activity of type E2 neurons byxcitatory and inhibitory effects, respectively.4 Theresent results, taken together, suggest thatDAVP has a tendency to suppress bladder activityy suppressing type E1 neurons or activating type Ieurons. The unexpected response of type E2 neu-ons (excitation to DDAVP) suggests that factorsther than DDAVP are involved in the suppressionf bladder activity. It is known that the neurons inhe micturition center (Barrington’s nucleus) con-ain corticotropin-releasing hormone and send de-cending fibers to the sacral segment to innervatehe bladder.5 Because most of the neurons re-ponding to DDAVP are distributed widely outsidearrington’s nucleus, they are likely to project tohe micturition center neurons rather than de-cending directly to the sacral segment. Additionaltudies are needed to determine whether the neu-ons recorded from Barrington’s nucleus send de-cending projecting, antidromic stimulation to theescending fibers.DDAVP has an affinity for the V2 receptor10 and

hows antidiuretic action acting selectively on the2 receptor in the kidney.11 In contrast to the V1

eceptor, which is localized in the smooth musclef the urinary bladder,12 the V2 receptor has nevereen confirmed to exist in bladder smooth muscle.he present results, which show that about one

alf of bladder-related neurons responded to w

ROLOGY 63 (5), 2004

DAVP, raise the possibility that the V2 receptorlso exists in the central nervous system, and, thaty affecting some bladder-related neurons, DDAVPonsequently modifies the activity of the urinaryladder. Because DDAVP was administered pe-ipherally or ICV, and the brainstem areas includ-ng the bladder-related neurons or Barrington’s nu-leus receive widespread afferents from variousrain regions,5 the possibility remains that theDAVP effect is not directly to the bladder-relatedeurons but is mediated through other neural sys-ems. Antagonists for V1 and V2 receptors and forther transmitters should be used to clarify thisoint.Suppression of bladder contraction by DDAVPas observed in a limited number of cases (4 of 12

ats after IV infusion and 2 of 8 rats after ICV infu-ion). In the rat whose bladder was never sup-ressed (nonresponsive rats), the DDAVP dosespplied were up to 14.5 �g in the IV infusion and.22 �g in the ICV infusion, 5 to 10 times greaterhan those applied to the responsive rats. Althoughhe cause of this difference is unknown, the effectf DDAVP might largely depend on individual dif-erences among the rats as has been reported inumans.13 In most cases, when the bladder wasuppressed (3 of 4 rats with IV infusion and 2 of 2ats with ICV infusion), the suppression was ac-ompanied by changes in neuronal activity, sug-esting that the DDAVP effect was mediatedhrough the bladder-related neurons.It has been reported that some enuretic children

ave low nightly serum levels of AVP,14,15 and thathe enuretic children have more difficulties withrousal from sleep than nonenuretic children.16

herefore, it is possible that enuresis is due notnly to high urine volumes but also to the inabilityo wake up because of the low levels of AVP. Di-

ichele et al.17 reported that the ICV injection ofDAVP increased locomotor activity, suggesting

hat DDAVP, infused into the central nervous sys-em, may have an arousal effect.Arousal is regulated by the cholinergic neurons

n the laterodorsal tegmental and pedunculopon-ine tegmental nuclei and the noradrenergic neu-ons in the locus coeruleus.18 Judging from thehort positive spikes, all neurons examined in theresent study were noncholinergic and non-nor-drenergic. Recently, we found that the cholinergiceurons in the laterodorsal tegmental increasedheir activity after direct injection of DDAVP in theicinity of the neurons (unpublished observation).rom these results, it is probable that DDAVP pre-ents enuresis by inducing arousal through theholinergic neurons. To confirm this hypothesis, atudy using nonanesthetized, free-moving animals

ill be necessary in the future.

997

tl

ca1

b

tu

wgs

deL

pel1

ir1

r

rc

oV

rchm

tA

d

a1

dp

cN

vm1

n1

9

REFERENCES1. Knudsen UB, Rittig S, Norgaard JP, et al: Long-term

reatment of nocturnal enuresis with desmopressin: a fol-ow-up study. Urol Res 19: 237–240, 1991.

2. Sawyer WH, Acosta M, Balaspiri L, et al: Structuralhanges in the arginine vasopressin molecule that enhancentidiuretic activity and specificity. Endocrinology 94: 1106–115, 1974.

3. Robinson AG: DDAVP in the treatment of central dia-etes insipidus. N Engl J Med 294: 507–511, 1976.

4. Tanaka Y, Koyama Y, Kayama Y, et al: Firing of micturi-ion center neurons in the rat mesopontine tegmentum duringrinary bladder contraction. Brain Res 965: 146–154, 2003.

5. Valentino RJ, Page ME, Luppi PH, et al: Evidence foridespread afferents to Barrington’s nucleus, a brainstem re-ion rich in corticotropin-releasing hormone neurons. Neuro-cience 62: 125–143, 1994.

6. Vincent SR, Satoh K, Armstrong DM, et al: NADPH-iaphorase: a selective histochemical marker for the cholin-rgic neurons of the pontine reticular formation. Neurosciett 43: 31–36, 1983.

7. Koyama Y, Honda T, Kusakabe M, et al: In vivo electro-hysiological distinction of histochemically-identified cholin-rgic neurons using extracellular recording and labelling in rataterodorsal tegmental nucleus. Neuroscience 83: 1105–1112,998.

8. Aghajanian GK, and Vandermaelen CP: Intracellulardentification of central noradrenergic and serotonergic neu-ons by a new double labeling procedure. J Neurosci 2: 1786–792, 1982.

9. Tyagi MG, and Thomas M: Enhanced cardiovasculareactivity to desmopressin in water-restricted rats: facilitatory

98

ole of immunosuppression. Methods Find Exp Clin Pharma-ol 21: 619–624, 1999.

10. Phillips PA, Abrahams JM, Kelly JM, et al: Localizationf vasopressin binding sites in rat tissues using specific V1 and2 selective ligands. Endocrinology 126: 1478–1484, 1990.11. Butlen D, Guillon G, Rajerison RM, et al: Structural

equirements for activation of vasopressin-sensitive adenylateyclase, hormone binding, and antidiuretic actions: effects ofighly potent analogues and competitive inhibitors. Mol Phar-acol 14: 1006–1017, 1978.12. Thibonnier M, Snajdar RM, and Rapp JP: Characteriza-

ion of vasopressin receptors of rat urinary bladder and spleen.m J Physiol 251: H115–H120, 1986.13. Neveus T, Bader G, and Sillen U: Enuresis, sleep and

esmopressin treatment. Acta Paediatr 91: 1121–1125, 2002.14. Norgaard JP, Pedersen EB, and Djurhuus JC: Diurnal

nti-diuretic-hormone levels in enuretics. J Urol 134: 1029–031, 1985.15. Rittig S, Knudsen UB, Norgaard JP, et al: Abnormal

iurnal rhythm of plasma vasopressin and urinary output inatients with enuresis. Am J Physiol 256: F664–F671, 1989.16. Watanabe H, and Kawauchi A: Nocturnal enuresis: so-

ial aspects and treatment perspectives in Japan. Scand J Urolephrol Suppl 163: 29–38, 1994.17. DiMichele S, Sillen U, Engel JA, et al: Desmopressin and

asopressin increase locomotor activity in the rat via a centralechanism: implications for nocturnal enuresis. J Urol 156:

164–1168, 1996.18. Kayama Y, and Koyama Y: Brainstem neural mecha-

isms of sleep and wakefulness. Eur Urol 33(suppl 3): 12–15,998.

UROLOGY 63 (5), 2004