Rapid desensitization of neonatal rat liver beta adrenergic receptors A role for beta adrenergic receptor kinase

Rapid desensitization of neonatal rat liver

beta-adrenergic receptors. A role for beta-beta-adrenergic

receptor kinase.

I García-Higuera, F Mayor Jr

J Clin Invest.

1994;

93(3)

:937-943.

https://doi.org/10.1172/JCI117099

.

Exposure of beta-adrenergic receptors (BAR) to agonists often leads to a rapid loss of

receptor responsiveness. The proposed mechanisms of such rapid receptor desensitization

include receptor phosphorylation by either cAMP-dependent protein kinase or the specific

beta-adrenergic receptor kinase (BARK), leading to functional uncoupling from adenylyl

cyclase and sequestration of the receptors away from the cell surface. To evaluate the

physiological role of such mechanisms, we have investigated whether rapid regulation of

BAR occurs in the neonatal rat liver immediately after birth, a physiological situation

characterized by a dramatic but transient increase in plasma catecholamines. We have

detected a rapid, transient uncoupling of liver plasma membrane BARs from adenylyl

cyclase (corresponding to a desensitization of approximately 45%) within the first minutes of

extrauterine life, followed by a transient sequestration of approximately 40% of the BARs

away from the plasma membrane. In agreement with such pattern of desensitization, we

have detected (by enzymatic and immunological assays) rapid changes in BARK specific

activity in different neonatal rat liver subcellular fractions that take place within the same

time frame of BAR uncoupling and sequestration. Our results provide new evidence on the

potential role of BAR desensitization mechanisms in vivo and suggest that they are involved

in modulating catecholamines actions at the moment of birth. Furthermore, our data indicate

that in addition to its suggested […]

Research Article

Find the latest version:

http://jci.me/117099/pdf

Rapid Desensitization of Neonatal Rat

Liver

,B-Adrenergic Receptors

A Role for fl-Adrenergic Receptor Kinase

Irene Garcia-Higuera and Federico Mayor,Jr.

Centro de Biologia Molecular "Severo Ochoa" (CSIC-UAM), UniversidadAut6noma,28049Madrid, Spain

Abstract

Exposure offl-adrenergic receptors (BAR) to agonists often leads to arapid loss of receptor responsiveness. The proposed mechanismsof such rapid receptor desensitization include re-ceptorphosphorylation by either cAMP-dependent protein ki-nase or the specific 8-adrenergic receptor kinase (BARK), leading to functional uncoupling from adenylyl cyclase and se-questration of the receptors away from the cell surface. To

eval-uatethephysiologicalrole of such mechanisms, we have

inves-tigatedwhetherrapid regulation of BAR occurs in the neonatal ratliver immediately after birth, a physiologicalsituation char-acterized by a dramatic but transient increase in plasma

cate-cholamines.Wehave detected a rapid, transient uncoupling of liver plasma membrane BARs from adenylyl cyclase

(corre-spondingto adesensitization of - 45%) within thefirst min-utesof extrauterinelife, followed by a transient sequestration of - _{40% ofthe BARs away fromthe plasma membrane. In} agreement with such pattern of desensitization, we have de-tected(by enzymatic and immunological assays) rapidchanges in BARKspecificactivity in different neonatal rat liver subcel-lular fractions that take place within the same time frame of BAR uncoupling and sequestration. Our results provide new

evidence onthepotential roleofBARdesensitization mecha-nisms in vivo and suggest that they are involved in modulating

catecholaminesactions at the momentofbirth. Furthermore,

ourdataindicatethat inadditiontoits suggested role as a rapid modulator of adrenergic receptor function at synapse, rapid BARK-mediated receptor regulation may have functional rele-vance inothertissues in response to high circulating or local levels ofagonists. (J. Clin. Invest. 1994. 93:937-943.) Key words: catecholamines *adenylyl cyclase * translocation * re-fractoriness * receptor regulation

Introduction

Itiswellknown that cellularresponsivenesstomessenger

stim-ulation isattenuatedintheface ofacute orsustained activa-tion. This generalbiological phenomenon isbasedonthe

regu-lationof signal transductionsystemsand has been termed

de-sensitization, tolerance, orrefractoriness. Aprototypicmodel

for the study of the molecular mechanisms of desensitization

Addresscorrespondence to Dr.FedericoMayor, Jr., Centro de Biologia

Molecular, UniversidadAut6nomadeMadrid, 28049 Madrid, Spain.

Receivedforpublication 30 March 1993 and in revisedform 4 Oc-tober 1993.

hasbeen the fl-adrenergic receptor(BAR)' adenylyl cyclase system, which mediates a variety of important physiological functions of the catecholamines adrenaline and noradrenaline indifferent tissues (1-3). Work from several laboratories has shown that the molecular mechanisms underlying rapid or short-term desensitization ofBAR involve both transient

se-questration ofreceptors awayfromthe plasma membrane and

functional uncouplingof the receptor from stimulatory G pro-tein. Afterlong-termexposure to fiagonists,othermechanisms

involvingchanges in receptorsynthesis and degradation also operate,usually leadingtoreceptordownregulation( 1-5).

Themechanisms involvedinrapidreceptorsequestration

arepoorly understood (6). On the other hand, therapid func-tionaluncouplingcanbetriggeredby BARphosphorylationby

cAMP-dependent proteinkinase(PKA)or

fl-adrenergic

recep-torkinase (BARK) (1,2, 7). BARKisacytoplasmicenzyme that specificallyphosphorylates the agonist-occupied form of

the BAR and other related Gprotein-coupledreceptors(2,7). BARK has been shown to rapidly translocate to the plasma membrane in response to the presence of, agonists in the

extracellularmedium(8-10);arolefor

fly

subunits of G

pro-teinsin BARKtargetingtothemembrane-boundreceptor has recently been reported ( I1). WhereasPKA-mediated regula-tiondirectly leadstoBAR uncoupling by phosphorylatingan

intracellular domain implicated in interaction with

stimula-tory Gprotein (3), phosphorylation byBARK promotes the

bindingtothe receptorofanothercytosolic protein,

fl-arrestin,

whichinhibits its coupling toa,and adenylyl cyclase (12, 13). Recent dataindicatethatadditional proteins,such as phosdu-cin, may alsoparticipateinthiscomplex regulatory network (14). Several studies (15-19), have suggested that

BARK-me-diated regulatory mechanismswould bespecially importantin the presenceof

high

levelsof

catecholamines,

suchas atneural synapses.Onthe contrary,PKA-mediated

modulation,

which

has been showntobeaslower process(18),would

predomi-nate atloweragonist

concentrations;

i.e., inresponsetonormal levelsof circulating catecholamines. This

interpretation

is

con-sistent with the fact that

expression

ofBARK- 1 mRNA is higherin thebrainandhighly innervated tissues (20),

although

thisenzyme has also been

reported

tobe

expressed

and

func-tionalin bloodcells suchasleukocytes(21).

Toprovideabetter

understanding

ofthe

physiological

role

ofreceptorregulatoryprocesses,ourlaboratory is interestedin

investigating

physiological

situationsinwhich such

desensiti-zation mechanisms

might

operate and could be detected. In

this regard,it isnoteworthythat the

perinatal period

is

charac-terized byadramaticsurge in

plasma

catecholamines

(22).

In

1. Abbreviations used in this paper: BAR,

fl-adrenergic

receptor; BARK,BARkinase;DHA,dihydroalprenolol; GppNHp, guanyl-5'-yl-imidodiphosphate;P3, microsomal internal membranes;PLSD, post

hocleastsignificant difference;PM, plasmamembrane;S3, cytosol.

J.Clin.Invest.

©TheAmericanSocietyforClinicalInvestigation,Inc.

0021-9738/94/03/937/07 $2.00

manyspecies, including therat, and to a somewhat lesser ex-tent,humans, sympatheticinnervation of autonomicorgansis absentornonfunctionalatbirth. Accordingly, inthefetus and neonate these messengers are released from the adrenal

me-dulla andparaganglia inresponse tothephysiological hypoxia associated with birth (23, 24), andpromotecriticalmetabolic,

cardiovascular, thermogenic, and respiratorychanges that are

necessarytosurvive the transitiontoextrauterine life (22, 25).

Such transient rise of catecholamines exceeds byanorderof magnitude the messenger levels detectedevenin themost ex-treme orpathological circumstances in the adult; in asphyx-iated neonates,evenhigher plasma catecholamine levelscanbe

attained(22, 23). An intriguing question is whether such im-portantchanges in hormone levels could triggerprocessesof adrenergicreceptordesensitization during the perinatal period. Thus, we hypothesized that rapid, catecholamine-induced changes in adenylyl cyclase coupling,

_{f3-adrenergic}

receptor

compartmentation, and

f3-adrenergic

receptorkinase

subcellu-lardistribution might beapparentinsuchphysiological

situa-tion. Inthis report, we show thatarapidprocessof

_{f3-adrenergic}

receptorregulation takes place in the neonatal rat liver immedi-ately afterbirth and explore its mechanisms.

Methods

Preparation of neonatal ratliversubcellularfractions. Term fetuses weredelivered with intact placentas byrapidhysterectomy ofcervically

dislocated mothers(timedpregnant Wistarrats)at22 d of gestation, exactlyasdescribed (24).Newbornswerequicklydetached after tying theumbilical cords and maintainedat 370C inahumidicrib witha

water-saturated atmosphere and withoutfeeding.Fetusesornewborn

rats werekilled bydecapitation immediatelyafter birthor<20 min after the induceddelivery;it is worthnotingthat the time 0 actually

correspondedto 45safter thebeginningof thehysterectomy

proce-dure.Using exactlythesameexperimental protocol, neonatalplasma catecholaminelevels of - 50nM(- 8ng/ml),have beenreportedat

themomentofdelivery,followedbyarapiddecreaseduringthe first 30 minoflife (to - 2ng/ml), reachinglevels of -1 ng/mlat1 h after

birth. Allsubsequentstepswereperformedat4VC.Livers(pooledfrom six fetuses from three different mothers) were extracted, weighed, diced,anddiluted in4volof 20 mMTris-HCl, pH 7.4,5mMEDTA,5 mMEGTA, 1 mMphenylmethylsulfonyl fluoride,20,ug/mlleupeptin, 20Mg/mlbenzamidine(buffer A) plus250mMsucrose, and

homoge-nizedwitheightstrokesofamotorized Teflonpestle.Subcellular frac-tionationwasperformed byestablishedprocedures (26, 27). Briefly,

the supernatant ofalowspeed centrifugation (250g, 4min)was cen-trifugedat3,000g for 10 mintoobtainaplasmamembranepellet.The

centrifugation of this supernatant (10,000g, 20 min) providedthe crude mitochondrialpellet.Finally,theinternal membranesor micro-somalfraction (P3) and thecytosolwere obtained aftera250,000g

centrifugation (60 min) (TL-100; Beckman Instruments, Fullerton,

CA). Subcellular fractionswerecharacterized by electronmicroscopy

asreported (reference 28, data not shown) and the distribution of marker enzymes for the plasma membrane (5' nucleotidase),

lyso-somes(,B-N-acetyl-glucoseaminidase) endoplasmicreticulum (glucose oxidase) and Golgi apparatus(a-mannosidaseII)wasmeasured. The plasma membrane fraction displayed a 6.3-fold increase in the specific activity of5'nucleotidasewith respecttothemicrosomalfraction(P3); P3 fractionswereenriched in endoplasmic reticulum, Golgi, and lyso-somal markers, with little contamination of plasma membrane. All

enzymatic markerswereassayed asdescribed (29). In other experi-ments,adifferent group of animalswasused. 1 d-oldrats were

sub-jectedto anoxia for 2-10 min exactlyaspreviouslydescribed (30). Afterallowing 30 s of recovery, the asphyxiated newborn rats andthe

appropriatecontrolswerekilledbydecapitationandliver subcellular fractions prepared exactlyasdescribed above.

Adenylyl cyclaseassay. Determination of cyclase activity was per-formedessentiallyasdescribed(31 ). Plasma membranes were washed three times in buffer A and the final pellet resuspended in 75 mM Tris-HCl, pH 7.5, 12.5 mM MgCl2, 1.5 mM EDTA, 2 mM DTT to a protein concentration of 2-3 mg/ml. Approximately 80 gof

mem-braneproteinwereaddedto anincubation mixturecontaining 30 mM Tris-HC1, pH 7.5, 5 mM MgCI2, 0.6 mM EDTA, 0.8 mM DTT, 10 mMcreatinine-phosphate, 0.5 mg/ml creatinine-kinase, 100

_ALM

ATP, 50 MMGTP,andvarious concentrations ofisoproterenol, 10 mM NaF

or 100 MM forskolin, in a final volume of 100 Ml. After 15 min of incubationat370C,thereaction was stopped and cAMP quantified by radioimmunoassay (Amersham Corp., Arlington Heights, IL) as re-ported (32). Experimental datawereanalyzed as previously described ( 17 ). Basal activitywasnotaffected by the different experimental con-ditions andwassubstractedfrom the values obtained in the presence of isoproterenoloractivators. The resulting isoproterenol-induced stimu-lationwasexpressedaspercent ofthe activity in the presence of 100

_AM

forskolintonormalize the data for effects thatoccur atthe levelof cyclase(i.e.,heterologousdesensitization).The extentof desensitiza-tionwascalculatedby measuring the loss of maximal stimulation by isoproterenolasdescribed ( 17).

Binding studies. Plasmamembranes(PM)or P3 obtained as de-scribed abovewerewashed threetimes andresuspendedinbufferAat a

proteinconcentration of 3 mg/ml. Total

_{fl-adrenergic}

receptornumber wasdetermined by incubating 300 Mgof membranes for 30 min at 30'C withasaturating concentration of 2 nM [3H]dihydroalprenolol ( [ 3H ] DHA) (Amersham Corp.), using 50MM (-)propanolol to define

nonspecificbinding; similar results were obtained when 100 MM (-)

isoproterenolwasusedinstead. For competition experiments, plasma membraneswereincubated in the presence of the radioligand and dif-ferent concentrations of ( -) isoproterenol in a medium containing 20 mMTris-HCl, pH 7.5, 10 mM MgCl2, 1 mM ascorbic acid, in the presenceorabsenceof 200MMguanyl-5'-yl-imidodiphosphate.

Com-petitioncurves werefittedtobothone-and two-site models using non-linear least squares regression analysis and the quality of the fits com-paredbyF test(33).

DeterminationofBARK. Since BARK has been shown to specifi-cally phosphorylate rhodopsin in an agonist-dependent fashion (34),

BARK activities in the different rat liver subcellular fractions were assessedbyusing purifiedurea-treated rod outer segments as substrate. This method has been previously used to determine BARK activity in cells and tissues and in BARK purification (9-11, 35). Rhodopsin kinase-free purified rod outer segments were prepared in the dark as reported (34), and aliquots containing 300-500 pmol of rhodopsin wereincubated for 20 min at 30°C in the presence or absence of light in

amediumcontaining25 mMTris-HCI, pH 7.5, 6 mM MgCI2, 5 mM NaF, 1 mM EDTA, 1 mM EGTA, 55MM [y-32P]ATP (2-3 cpm/

fmol)and extractsof rat liver subcellular fractions obtained from

neo-nates atdifferent times after delivery (final vol, 50Ml).Cytosol(S3) containingsolubleBARKactivity was assayed directly. However, for

determiningBARKactivity in membrane fractions, it first had to be extracted (10).PM orP3 were washedonce, resuspended inasmall volume ofbuffer A, and treated for 15 min at4°Cwith 200mM NaCl

inbufferAtoextractperipheral proteins. Membranes were then pel-leted and the supernatant containing the extracted proteins assayed for

BARKactivity asdescribed above. Phosphorylation reactions were

stopped bydiluting20-fold with ice-cold buffer A and centrifuging at 12,000 g for 15 min. The resultant pellets containing the phosphory-latedrhodopsinweresuspended in SDS-PAGE sample buffer and

elec-trophoresed as described (10, 34), followed by autoradiography.

BARKactivity was quantitated by measuring the radioactivity asso-ciated to the excised rhodopsin band in the dried gel by Cerenkov spectroscopy,andnormalized by the milligrams of protein present in the assay(9). In someexperiments, we investigated the presence of

BARKproteinin cytosolicormembrane-extracted fractions by

West-ernblotting. Thedifferent rat liver subcellular extracts (- 200Mgof

proteininSDS-PAGE sample buffer) were resolved by electrophoresis in 7.5% SDS-polyacrylamide gels and blotted to filters (Immobilon;

A

B Figure 1. Desensitization of

fl-adrenergic

receptors in the neonatal ratliver. (A)Adenylyl cyclase activities _ of

plasma

membranesobtained from the liver of

neo-< 60 0 300 Fsnatalratskilledimmediatelyafter birth(0min, o),

soL ;g;2* 260 5minafterbirth (x), or 20minafterdelivery(e)was

-50:: _measured~ _{in thepresenceof various concentrations}

R

.E

40

5

min

c

220l

of(-) isoproterenolandexpressedaspercent of the 30 mC"~~~~~~~' 8

d<E 30 r ,,9-9 0 Sin E 180 activityin the presence of 100;tM forskolin,as

de-v0 * _X

140

tailed in Methods. The values forforskolin

stimula-LJ 20 -14

-_j1

_{j>_}

_{10 I}

_-a|

_{E 100; M (B). Adenylyl cyclase}tion overbasalatthedifferent timesareshown in activities stimulated by 10mM

oLJ -9 -8 -7 -6 -5 -4 -3 0 520 052 NaF or100MgMforskolin in neonatal rat liver plasma

Log[ISOPROTERENOLJ Time (min) membranes obtained from animals that were killed

attheindicated times after

delivery.

InbothAandB,

resultsaremeans±SEM of threeindependent experiments performed in triplicate. There isasignificant variationof maximal adenylyl cyclase stimulation byisoproterenolwith time afterdelivery(one-wayANOVA,F(2,9)= 9.55, P<0.01).Significantdifferences(*P<0.05) for maximal stimulation comparedtothatobtainedat20 minafter deliveryareshown(Fisher's PLSD).

Millipore Corp., Bedford, MA). Filters were probed with AB-9 (1:1,000), a polyclonal antibody raised against the recombinant

BARK-l protein expressed in Sf-9 cells (kindly provided by Dr. Jeffrey L.Benovic,Jefferson Institute, Philadelphia, PA), or AB-792 (1:40),

anaffinity-purified polyclonal antipeptide antibody raised in our labo-ratoryagainst aminoacids 533-544 of the bovineBARK-l sequence (see reference36). Filterswerewashed and developedusing a chemilu-miniscent method (Luminol; Amersham Corp.).

Statisticalanalysis. Data are expressedasmean±SEM.Statistical evaluation wasperformed using individual Student's t test, when ap-propriate,or onewayANOVAwith Fisher's post hoc leastsignificant differencetest(PLSD), usingtheStatview program.

Results

Asafirststep toexamine theexistence ofapossible BAR de-sensitization process in theperinatal ratliver, weinvestigated the stimulation of adenylyl cyclase activity by different

concen-trations of the /3 agonistisoproterenol in plasma membrane

preparations obtained from animals killedat0, 5, or 20 min

afterdelivery. Fig.1 Ashows that themaximalresponse elicited

byisoproterenol(normalizedwith respect toforskolin-induced stimulation) was markedlydiminished at both the0- and 5-mintimepointswhencompared tomaximalstimulationat20 min afterdelivery, correspondingto adesensitization of 40-45%. These results indicate that uncoupling of BAR from adenylyl cyclase occurs

immediately

after birthand that a re-covery

(resensitization)

fromsuch

uncoupling

proceeds

there-after. Such

desensitization/recovery

wouldcorrespondto

/2-adrenergic

receptors,which is the

predominant

subtype in the

neonatal rat liver. Asshown in

Fig.

1 B, only

slight

changes

(12-17%) inadenylyl cyclase activationbyNaF or forskolin (acting at the G protein andeffector levels,

respectively)

are

apparent in the same

experimental conditions,

and show a

somewhat

different,

slowertimecourse.For

instance,

the mini-mal stimulation by both activators isreached at 5 min after

delivery,

andforskolin-stimulated cyclase

activity

seems tobe not fully recovered at 20 min after birth, as opposed to the

isoproterenol-stimulatedactivity.Theseresultssuggestthatthe

rapid

regulation of this transduction system is

preferentially

taking placeupstreamoftheGprotein andthecyclase effector.

Uncoupling

atthereceptorlevelis further suggested

by

analysis

of

radioligand binding agonist competition

curves in the

ab-sence or in thepresenceof GppNHp,a

nonhydrolisable

ana-logue of GTP. Inplasma membranes obtained from animals immediately after birth (Fig.2A),the percentageofreceptors

showing high affinity agonist binding [IC50 of (-) isoproter-enol 1.3nM] in the absence of GppNHPwasonly

31+4%;

in thepresenceof GppNHp,aslight shifttotheright of the

com-petitioncurve occursbecause of the conversion ofmostofthe receptors tothe lowaffinity state [

IC50

of (-) isoproterenol 0.6

MM].

A

markedly

higher (about twofold) proportion

of

f3-adrenergic receptors in the high affinity state (59±5%)was de-tected inmembranes obtainedat20min after delivery (Fig. 2 B), in agreement with theincreased BAR-adenylylcyclase cou-pling detected at this timepoint compared to thesituationat birth (Fig. 2 A). The presence of GppNHpinthe experiments performed with the 20min plasma membranes also promoted the shift of receptors to a low affinity state

(IC50

of( -) isopro-terenol 0.8

/iM).

Furthermore, a very rapid process of BAR sequestration

canbe monitored inratliver during this early postnatal period. Asassessed by [3H

]

DHAbinding, about a 25% decreaseinthe number of BAR in the plasma membrane can be detected within 2 min of delivery (Fig.3, closed circles),and by5-10 min of life a significant decrease of- _{40% with respect to the}

Figure 2. Functional

100 A coupling off-adrenergic

80 receptors inratliver

'₆₀_ <\o

_plasma

membranes

de-E rived from animals

,x 40 killedatbirth(A)or20

E

20_ min after delivery (B).

//Iby, I

with,

S

QMembranes

were

incu-0 B batedwith increasing

hi100- * >>to B _{concentrations of the /3}

80

agonist(-)

isoproter-I 60

_60Am

enol to compete for the

bindingtothe BAR

I'

40 with the radiolabelled

20 0 antagonist [3H ]DHA

if 0 ^ (seeMethods) in the

-10 -9 -8 -7 -6 -5 -4 absence (.)orpresence

Log [ISOPROTERENOL] (o) of 200 gM GppNHp. The data shownarethemeansof threeindependentexperiments withtriplicate determinations. Datawerefitted and analyzed as described in Meth-ods.Isoproterenolcompetitioncurvesin the absenceof GppNHp were bestdescribed byatwo-site model at both time points (F test, P

- ₁₈₀

E ₁₆₀

,- 140 0

Z

120

L 100 ao

E 80 m

C

60

I 40 20

.i-\

\f*+

_N."

I I I

I.

0 5 10 15 20

Time

after

delivery

(min)

Figure 3. Timecourseofratliver,B-adrenergicreceptor redistribution

duringthefirst minutes afterdelivery.Plasma membrane(.)and internal membranes (o) fractionswereobtainedasdescribed in Methods from the liver of neonatalratssacrificedattheindicated times afterdelivery.Total,B-adrenergicreceptor number in each frac-tionwasestimatedbyradioligand bindingasdetailed in Methods. Thevalues obtainedimmediately afterbirth ( 18±2fmol/mg protein

for the internal membranes and 36±3fmol/mgproteinforplasma

membranes)weretakenas 100%. Dataaremeans±SEMof fourto

nineexperimentsassayed intriplicate.There isasignificantvariation of,B-adrenergicreceptor number in theplasmamembranewith time afterdelivery(one-wayANOVA,P(4,27)=5.74,P<0.01).

Signif-icantdifferences(*P<0.05) with respecttothe 0-mintimepoints

areshown(Fisher's PLSD).

levelsatbirthis attained. Conversely,anincreasein fl-adrener-gic receptors present in internal, microsomal membranes is

notedwithin the same time frame(Fig. 3, opencircles), thus

indicatingtheexistence ofasequestrationprocess. Such

redis-tribution of BAR istransient, since the internalized receptors

graduallyreturn totheplasma membrane,asindicatedbythe data obtained at 20 min afterdeliveryinbothplasmaand inter-nal membranes(Fig. 3).

The receptor uncouplingand sequestration datastrongly suggestedthat a very

rapid

processof

predominantly

homolo-gous

(i.e., agonist-specific)

BAR desensitization was taking placein the neonatal ratliverimmediately after birth. Homolo-gousdesensitization ofBARhas beenreportedtoinvolve

spe-A

S3 PM

%.

OPSIN

P3

,I

cific receptor phosphorylation by the BARK after

agonist-in-ducedkinase translocation from thecytoplasmtothe plasma membrane ( 1,2).Toinvestigateapossible role forBARK in this process of BAR desensitization detected in neonatal liver, we studied the presence of BARK inperinatal rat liver and the occurrence ofchanges in BARK subcellular distribution (i.e., BARK translocation).

Although BARK expression hasbeen mostlyascribed to

the brain and highly innervatedtissues(20)wehave found that BARK activity isclearlypresentindifferent subcellular frac-tions ofthe neonatalliver,asdemonstratedby the

light-depen-dent phosphorylation of rhodopsin (Fig. 4 A). Antibodies

raised against recombinant BARK-l specifically detected in

extracts from thedifferent subcellularfractionsaband of - 80

kD,whichcomigrates with recombinant bovineBARK- 1 ex-pressedinSf-9cells(Fig. 4B). Thesameresultsareobtained

withantipeptideantibodiesgeneratedagainstabovine BARK-1 sequence (data not shown, seealso reference 36).

Interest-ingly, importantchangesinBARKspecific activity in the dif-ferent rat liver subcellular fractions are detected within the

sametimeframe ofBARdesensitization.Asshownin

Fig.

5, BARK-specific activity transiently increases in the plasma

membrane, reachinga peak at 5 min of life and decreasing thereafter, whereas it slightly decreases in the cytosol and signif-icantly increases in the internal membranesfraction during the first 20min afterdelivery.Itshould be noted that BARK-speci-fic activityinthelatterfraction is slightly higher than that in

plasma membrane and more than 10-fold higherthan thatof cytosol (see legendto Fig. 5). The subcellular distribution of

total BARKactivity isshown in Table I. Soluble BARK ac-counts for - 30% oftotalactivity atbirth, and tendsto

de-creasewithtime,whereas a

significant

increase isnotedin the kinase associated to microsomal membranes (from 50 to

65%);slight increasesarenoted in thesize ofthe BARKpool already associatedto plasmamembrane fractionsatthe mo-mentof birth. Itisinterestingto notethat theincrease in total

BARK activity noted in the plasma membrane and

micro-somalfractions ishigher thantheparallel decreaseinsoluble

BARK, thussuggestingtheexistence of additional factorsthat may modulate BARKactivity

(11).

To confirm that catecholamines could promote a very

rapidtranslocation ofBARKinratliver, weinvestigatedthe

effectof subjecting l-d-oldrats tobriefperiods ofanoxiaonthe subcellulardistribution ofthe enzyme. 1 d-oldrats arestill able

B

S3 PM P3 Rec

200

-97

-66- - -' - --BARK

45-Figure 4. Presence of,-adrenergic receptor kinase in the neonatalratliver. (A) S3, PM, or P3 were obtained from theliver of neonatal rats killed 10 min after de-livery, andBARKactivity was extracted from mem-branefractionsasdetailed in Methods. The different

extractswereincubated under phosphorylating con-ditions with urea-treatedpurifiedrodoutersegments either inthepresence(leftlanes under S3, PM, and P3) or absenceoflight (right lanes), followed by SDS-PAGEandautoradiography. The migration of

theopsinband(assessed by Coomassie blue staining) is markedwith anarrow.The autoradiogramis rep-resentative of three experiments. (B) S3, PM, or P3

extractsfrom neonatalratliverwerepreparedas de-scribed inAand Methods. These extracts and recombinantBARK-I expressed inSf-9 cells (Rec) wereresolved by 7.5%SDS-PAGE,blotted andprobed withapolyclonalantibody raised against recombinant bovine BARK-I(this figure) or with an affinity-purified anti-peptide antibody

raised againstabovine BARK-I sequence(data not shown). The antibodies recognize a band (marked with an arrow) with an apparent relative molecularmassof - 80 kD, similartothatreported for BARK-1 (20), which comigrates withrecombinant BARK (Rec).

Figure

5.

_Changes

in the

T'

.|>1-0

180 --- - r BARK activity

asso-- asso-- asso-- ciatedtodifferent liver

Z

160

--P3

subcellularfractions

140 ' duringthe first minutes

120_

l-?

2PM after

delivery.

Extracts

~120

100 were obtained as in Fig.

° \ 4from the liver ofrats

< 80 ~ t _{- -} s3 _{sacrificedatthe}

indi-m 60 catedtimesafter

deliv-0o 05 1010 1515 220 ery and BARK activity

~~measured

inanassay

Time after delivery Imin)

_mased

_in

_anhassay

based in light-depen-dentphosphorylationof rhodopsin as described (9, 10). BARK spe-cific activity was quantitated as detailed in Methods. The values ob-tained immediately after birth were taken as 100% (6±1, 59±8 and 64±6 pmol of phosphateincorporated into rhodopsin per milligram of protein for S3 [ *],PM [o], and P3 [x

_I

fractions, respectively). Results are means±SEMof four independentexperimentsperformed intriplicate. Thereis a significant variation of microsomal BARK activity with timeafter delivery (one-way ANOVA, F (4, 17)=2.98,

P <0.05). (a)significant difference (P < 0.05) compared to the 0-mintime point (Student's t test); *Significant difference compared

tothe0-min timepoint (Fisher's PLSD).

to release

huge

amounts of catecholamines from the adrenal medullainresponse to hypoxia (23). Preliminary results using thisexperimental system indicateamarked increase (- 2.5-fold) in liverplasmaandmicrosomal membrane BARK spe-cificactivity after2minofanoxia as compared to control ani-mals (data not shown). These results suggest that an increase in

circulating catecholaminesasa result ofeither induced delivery orexperimental hypoxiacanpromote very rapid changes in the

specific activity ofBARK indifferentsubcellularfractionsof the neonatal ratliver, the timecourseofsuchchanges being

consistent witharolefor BARK-mediated phosphorylation in theregulationof BARimmediately afterbirth.

Discussion

We haveinvestigatedwhether aphysiologicalprocessofrapid

f,-adrenergic

receptorregulationtakesplaceduringthe

perina-talperiod, which is characterized byadramatic, transient

in-creasein plasma catecholamines triggered by the hypoxia

asso-ciated with delivery (22, 25). Our results indicate thatarapid uncoupling of liver plasma membrane BAR from adenylyl

cy-claseoccursimmediately after birth, followed byarecoveryor

resensitization detectedat 20 min afterdelivery. The overall pattern and theestimated extent of desensitization (40-45%)

aresimilartothose described in cellular models of agonist-spe-cific BAR desensitization (2, 10, 17, 18, 37). Only slight changes in the adenylyl cyclase response to activatorsatthe levelof G proteinsortheeffectoritself are detected. This small heterologous componentis probably caused by PKA-mediated regulation of these proteins, as also suggested by the slower time course of such effects when compared to BAR uncoupling (18). In agreement with a process of predominantly homolo-gous BAR desensitization, binding studies indicated that plasma membrane adrenergic receptors were poorly coupledto

Gproteinsimmediatelyafterbirth, as has been described in the newbornrabbit liver(38) and in the lungs of adult rats infused withisoproterenol ( 31 ).

Furthermore, a very rapid, transient sequestration of plasma membrane receptors was noted in our experimental systemshortly after birth. Receptors seem to be translocated to internal membrane compartments, as has been reported with a similartime course in culturedcellsexposed to1 agonists(6, 10, 39-41 ) or in the lung of isoproterenol-treated adult rats (31 ). The exact intracellular fateofBARand the sequestra-tion-recycling mechanismsarepoorly understood, although

re-cent data suggest the implication of endosomes (41). It is worthnotingthat in our model, asinotherexperimental

situa-tions, theadenylyl cyclase response is clearly diminished before the maximal internalization is reached, thus reinforcing the

hypothesisthat BARuncouplingprecedessequestration (2, 3,

6, 39). Moreover, it is importanttostress the good temporal correlationbetween the recycling of internalized BAR back to the plasma membrane and the recovery of isoproterenol-sti-mulatedadenylylcyclaseactivity observed20 min after birth. It has beenrecently suggested that themainroleof sequestra-tion would be to allow thedephosphorylation andrecyclingof uncoupled,phosphorylated plasma membrane receptors (2, 6,

42).Our results are consistent with suchsuggestion,and

rein-TableI.Subcellular Distribution of TotalBARK Activity inthe Newborn Rat Liver

Timeafter delivery

0 min 5min 20 min

Totalfraction Percentof total Totalfraction Percent of total Totalfraction Percent of total

activity tissueactivity activity tissueactivity activity tissue activity

pmol/g oftissue pmol/g oftissue pmol/g oftissue

Cytosol(S3) 114±21 31±4 100±4 23±3 91±1.6 20±4

Plasma membranefractions 42±2 14±2 63.5±2 17±1 52±12 62±2

Microsomalfraction (P3) 161±9 50±7 241±13* 64±5 210±37 59±5

TotalBARKactivity in each fractionandtimepointwascalculated in eachindependentexperimentfrom thespecific

activity

values

(expressed

inpicomoles of phosphateincorporatedintorhodopsinpermilligramofprotein), takingintoaccountthevolume of each fraction obtainedper

gramof tissue (4mlof cytosol, 1.5 mlof resuspended plasma membranefractions,and 1mlofresuspendedmicrosomalfraction)and theprotein

concentration of thesamplesused in therhodopsinphosphorylationassay(5-6 mg/mlcytosol, 0.6-0.8mg/mlextracted

plasma

membranes,

and2-2.5mg/mlextracted microsomalmembranes).Resultsaremeans±SEM ofthreetofour

independent experiments performed

in_triplicate.

Thepercent of total tissueactivityatdifferent times afterdeliverywasdetermined in eachexperiment

taking

thesumof total BARK activities

as 100%. The resultsaremeans±SEM of threetofourindependentexperiments. Thereisa

significant

variationof microsomal BARK

activity

force the fact thatwe areobserving a desensitization /

resensiti-zationprocessvs apossiblepostnatal supersensitization of

3-adrenergicreceptors.

BARK-mediated receptor phosphorylation has been

sug-gestedtoplayakey roleinthemechanisms of agonist-specific,

homologous BARdesensitization.Thus, we next sought to

es-tablishwhetherthis kinasewaspresent in the neonatalratliver

and if BARK translocation and activation was taking place

within the sametime frame ofthe observed process ofBAR

desensitization. ThepresenceofBARKactivityintheneonatal ratliverisdemonstrated by thefactthat extractsfromdifferent subcellularfractionsareabletophosphorylate rhodopsinina

light-dependentway. Suchactivity is inhibited bylow

concen-trations of Zn 2+orheparin (datanotshown),whichareknown

negative modulators ofBARK(18, 35). Furthermore,two dif-ferentantibodies generated againstbovine BARK-1 (20, 36)

recognized a band of80 kDin neonatal liver extracts, thus

indicatingthatthis kinase isoform or aclosely related,

cross-reacting isoenzyme is expressed in this tissue.

With regard to BARKtranslocation, our results indicate

theoccurenceofsignificant changesin BARKspecific activity

in plasma membrane and otherliver subcellular fractions with

aclosetemporalrelationshiptotheprocess of BAR

uncoupling

andsequestration.Theextentof such changesin

specific

activ-ity are similar to those reported for BARK translocation in response to fiagonists in C6glioma cells ( 10)orhuman

periph-eral blood leukocytes (21 ). Since

0-adrenergic

receptors are

already uncoupled from adenylyl cyclase atthe moment of birth (time point 0) and BARK translocationis

required

for agonist-inducedBARphosphorylationandhomologous

recep-tor

desensitization,

itcould beargued that

significant

BARK

translocation (i.e., an increase in plasma membrane BARK

specific activity) should alreadyhadoccurredatthis

point.

The

question is difficult to address in our in vivo

experimental

model. First, no control animals(i.e., animals unaffected by

the surgeof plasma catecholaminesatdelivery)areavailableto

establishthebasal valuesofBARK-specific

activity

in the dif-ferent cellular fractions. Second, sinceas

explained

in Meth-ods, time

point

0correspondsto45safter

initiating

the

surgical

delivery

ofthe

fetuses,

and BARKtranslocationand receptor

phosphorylation inresponsetof3agonistsoccursveryrapidly inA43 1 cells ( 18 ) and in C6gliomacells (10), it is likelythat 45safterdelivery,BARKtranslocationhas

already

taken

place

to a

significant

extent, and thatwedetect

only

afractionofthe

BARK "peak." The rapidityand relatively higher extent of

BARK translocation with respect to adequate controls ob-served in 1 -d-old ratssubjected toanoxiaare alsoconsistent with this interpretation.Infact,Table Iindicatesthat a

signifi-cant amount of total BARK activity is already present in plasma membranefractions immediately after birth. It still

re-mainstobeestablishedwhether such changes in BARK subcel-lular localization are caused by actual protein translocation,

kinaseactivation, or a combination of both mechanisms.

Inconclusion, our results are consistent with rapid BARK

translocation preceding BAR uncoupling and sequestration, thusstrongly suggestingakeyrolefor the kinase in this

physio-logical process ofBAR regulation triggered by the surge in plasmacatecholaminesassociated with birth. However, given thefactthat BARK can regulate other G protein-coupled re-ceptors(2,7, 9, 21 ), thepossibilitythat part of BARK

translo-cation

is caused by theconcurrentactivation ofother receptors

different fromBAR cannotbecompletelyruled out.

PreviousinvivostudieshavedetectedBARdesensitization

and downregulation in different tissues afterlong-term

infu-sion withfi

agonists

orchronic

pathological

increasesin

circu-lating catecholamines,such as inheart

failure, hypertension,

or

pheochromocytoma (43-45). On thecontrary,our

experimen-tal model is based intransient, physiological changesin

cate-cholamine concentrations.The

rapid

timecourseof desensiti-zationthatweobserve correlates wellwithclinical observations in other systems

showing

reduced

sensitivity

to

,3 agonists

within minutes ofparenteral administration of antiasthmatic drugs

(reference

30 and references

therein).

Our results are

alsoconsistent withtherefractorinessto

j3-adrenergic

stimula-tionof

glycogenolysis

in theneonatal rabbit andratliverin the

firsthoursof

life, although

additional factorsmay beinvolved

inorderto

explain

suchlong-term

unresponsiveness

(24, 37,

46,

47).

Furtherstudies will be needed in ordertoascertain the

physiological

role ofsuch very

rapid

desensitization ofBAR

andto

identify possible

functional correlates

(gene expression,

induction of mitochondrial

maturation)

for such transient liver BARactivationat

birth,

in thecontextof thecatecholamines actions

during

the

perinatal period.

A different

sensitivity

to

desensitizationwould allow the tissues and cellsto

respond

toa different extent or with variable rates to the same surge in

plasma catecholamines.

Thus,

it is

tempting

to _suggest that thesemechanisms ofreceptor

regulation play

a

_key

role in

limit-ingand

modulating

the extent,kindand time frame

ofcatechol-amine actions

during

the

perinatal period.

In

conclusion,

ourresultssuggestafunctionalrelevancefor BARK-mediated

phosphorylation

and

uncoupling

ofBARin vivo.Inagreementwithrecent results(21 ),ourdataindicatea

physiological

role for this kinase in

peripheral

tissues in

re-sponsetoplasma

catecholamines,

inadditiontoits

suggested

roleas

rapid

regulator of

adrenergic

receptorfunction in the brainand other

highly

innvervated

tissues,

where cellsare

ex-posed

to

high

and

rapidly

changing

concentrations of catechol-aminesreleasedatthe synapse

(2, 18).

Acknowledgments

The authors thankDr.R. J.Lefkowitz andDr. I.Sandoval for critical

readingofthemanuscript,Dr.J.L.Benovicforprovidingrecombinant BARK-1, theAb-9 polyclonal antibodyandhelpfulsuggestions,and

Dr.J.M.Cuezvafor hiscomments.C.MurgaandP.Penelaare

grate-fully acknowledgedforhelpinraisingandcharacterizingaBARK-I

antipeptide antibody,Prof.F.Mayorfor continous encouragement, C. San Martin fortheelectronmicroscopy studies,Dr.

Vatzquez

forhelp

with the statisticalanalysis, and Mrs.M. Sanzfor skillful secretarial assistance.

This workwassupported byComisionInterministerialCienciay

TecnologiagrantsPM890060andPB92-0135, CAMC105/91,

Boehr-inger Ingelheim,andFundacionRamonAreces. I.Garcia-Higuerawas recipientofpredoctoralfellowshipsfrom Andromaco and the Basque Government.

References

1. Benovic,J. L.,M. Bouvier, M. G.Caron, and R. J. Lefkowitz. 1988. Regulationof adenylyl cyclase-coupled ,B-adrenergicreceptors. Annu. Rev.Cell Biol. 4:405-428.

2.Hausdorff,W.P.,M. G.Caron,and R. J.Lefkowitz. 1990.Turning offthe

signal:desensitization of/3-adrenergicreceptorfunction. FASEB(Fed.Am.Soc. Exp. Biol.)J. 4:2881-2889.

3.Dohlman,H.G.,J.Thorner,M.G.Caron, and R.J.Lefkowitz. 1991. Modelsystemsfor the study of seven-transmembrane-segmentreceptors. Annu. Rev.Biochem.60:653-688.

4.Hadcock, J. R., and C. C. Malbon. 1991. Regulation of receptorexpression byagonists: transcriptional and post-transcriptional controls. Trends Neurosci. 14:242-247.

5.Collins, S.,M.G.Caron, and R. J. Lefkowitz. 1992. From ligand bindingto

gene expression:newinsights into the regulation of G-protein-coupled receptors. Trends Biochem. Sci. 17:37-39.

6.Kobilka,B. 1992.Adrenergic receptorsasmodelsfor Gprotein-coupled receptors. Annu. Rev. Neurosci. 15:87-114.

7.Palczewski, K., and J.L.Benovic. 1991.G-protein-coupled receptor ki-nases. Trends Biochem. Sci. 16:387-39 1.

8. Strasser,R.H., J.L.Benovic,M.G. Caron, andR.J.Lefkowitz. 1986. /3-Agonists-andprostaglandinEl-induced translocationofthefl-adrenergic re-ceptorkinase: evidence thatthekinase mayact onmultiple adenylate cyclase-coupled receptors. Proc. Nat!. Acad. Sci. USA. 83:6362-6366.

9. Mayor, F., Jr., J. L. Benovic, M. G. Caron, and R. J. Lefkowitz. 1987. Somatostatin induces translocation ofthefl-adrenergicreceptor kinase and de-sensitizes somatostatin receptors in S49 lymphoma cells. J. Biol. Chem. 262:6468-6471.

10.Garcia-Higuera, I., and F. Mayor, Jr. 1992. Rapid agonist-induced beta-adrenergic receptor kinase translocation in C6 glioma cells. FEBS (Fed. Eur. Biochem.Soc.) Lett. 302:61-64.

11.Pitcher, J. A., J. Inglese, J. B. Higgings, J.L.Arriza, P. J. Casey, C. Kim, J. L.Benovic, M. M. Kwatra, M. G. Caron, and R. J. Lefkowitz. 1992. Role offy subunits ofGproteins intargetingthe,3-adrenergicreceptor kinase to membrane-boundreceptors. Science(Wash. DC). 257:1264-1267.

12. Lohse, M. J., J. L. Benovic, J. Codina, M. G. Caron, and R. J. Lefkowitz. 1990.#l-Arrestin:aprotein that regulatesf,-adrenergicreceptorfunction. Science (Wash.DC). 248:1547-1550.

13. Lohse, M. J., S. Andexinger, J. Pitcher, S. Trukawinski, J. Codina, J. P. Faure, M. G. Caron, and R. J. Lefkowitz. 1992. Receptor-specific desensitization withpurified proteins.J.Biol. Chem. 267:8558-8564.

14.Bauer,P.H., S. Muller,M.Puzicha, S. Pippig, B. Obermaier, E. J. M. Helmreich, and M. J. Lohse. 1992. Phosducin is a protein kinase A-regulated G-proteinregulator. Nature(Lond.).358:73-76.

15.Hausdorff, W. P., M. Bouvier, B. F. O'Dowd, G. P. Irons, M. G. Caron, andR.J. Lefkowitz. 1989.Phosphorylationsites ontwodomains ofthe

#2-adren-ergicreceptorareinvolved indistinct pathways of receptor desensitization. J. Biol. Chem. 264:12657-12665.

16. Lohse, M. J., R.J.Lefkowitz, M.G. Caron, andJ. L.Benovic. 1989. Inhibition of ,B-adrenergic receptor kinase prevents rapid homologous desensiti-zation ofj2-adrenergicreceptors. Proc. Natl.Acad.Sci. USA. 86:3011-3015.

17. Lohse,M.J., J. L. Benovic,M.G. Caron, andR.J.Lefkowitz. 1990. Multiple pathways of rapidj02-adrenergicreceptor desensitization. J. Biol. Chem. 265:3202-3209.

18. Roth,N.S., P.T.Campbell,M.G.Caron, andR. J.Lefkowitz. 1991. Comparativeratesofdesensitizationof ,-adrenergicreceptorsby the ,B-adrener-gic receptor kinase and thecyclicAMP-dependentproteinkinase. Proc.Natl. Acad.Sci. USA.88:6201-6204.

19.Clark,R.B.,M.W.Kunkel, J. Friedman, T. J. Goka, and J.A.Johnson. 1988. Activation ofcAMP-dependentproteinkinase isrequiredfor heterologous desensitization of adenylyl cyclase in S49wild-type lymphomacells. Proc.Natl Acad.Sci. USA. 85:1442-1446.

20.Benovic,J.L., A. DeBlasi,W.C.Stone,M.G.Caron, andR. J.Lefkowitz. 1989. f,-Adrenergicreceptor kinase:primarystructuredelineatesamultigene family.Science(Wash. DC).246:235-240.

21.Chuang,T.T.,M.Sallese, G.Ambrosini,G.Parruti, andA.DeBlasi. 1992.Highexpressionoffl-adrenergicreceptor kinase in humanperipheralblood leukocytes.J.Biol. Chem. 267:6886-6892.

22.Lagercrantz,H., andT. A.Slotkin. 1986. The "stress" ofbeingborn.Sci.

Am.254:100-107.

23.Seidler, F. J., andT. A.Slotkin. 1985.Adrenomedullary function in the neonatalrat:responses to acutehypoxia.J.Physiol.358:1-16.

24.Cuezva,J.M.,E.S.Burkett,D.S.Kerr,H.M.Rodman,andM.S. Patel. 1982. The newborn of diabeticrat.I.Hormonal andmetabolic changes inthe

postnatalperiod.Pediatr. Res. 16:632-637.

25.Slotkin,T. A., andF.J. Seidler. 1988.Specializedresponses of the

sym-pathoadrenal andcardiac-sympatheticaxes in the newborn. InProgress in Cate-cholamineResearch.T. A.Slotkin and F. J.Seidler,editors.Alan R.Liss, Inc., New York.pp.441-444.

26.Traub, L. M., and R.Sagi-Eisenberg.1991. Purification of plOO,aprotein antigenicallyrelatedtothesignal transducingGproteinsGt and Gi.J. Biol. Chem.266:24642-24649.

27.Graham, J. 1984. Isolationofsubcellularorganellesandmembranes.In Centrifugation.APractical Approach.D.Rickwood,editor.IRLPressLtd., Ox-ford. pp. 161-182.

28. Lasa,I.,J. R.Cast6n,L. A.Fernindez-Herrero,M.A.DePedro,and J. Berenguer. 1992. Insertionalmutagenesisintheextremethermophilic eubac-teria.ThermusthermophilusHB.Mol. Microbiol. 6:1555-1564.

29. Bonay,P., and R. C. Hughes. 1991. Subcellular distribution inratliverof

anovelbroad-specificity(a1-2,a1-3 anda1-6)mannosidase activeon

oligoman-noseglycans. Eur. J. Biochem. 197:229-238.

30. Mayor,F., Jr.,M.Pages, J.Diez-Guerra,F.Valdivieso, andF.Mayor. 1985. Effect ofpostnatal anoxia on bilirubin levels inratbrain.Pediatr. Res. 19:231-236.

31.Strasser, R. H., G. L. Stiles, and R. J. Lefkowitz. 1984. Translocation and uncouplingof thef3-adrenergicreceptor in ratlung after catecholamine-pro-moteddesensitizationinvivo.Endocrinology. 115:1392-1400.

32.Zaremba, T. G., and P. H. Fishman. 1984. Desensitization of catechol-amine-stimulated adeylate cyclase and down-regulation of beta-adrenergic

recep-torsinratglioma C6 cells. Mol.Pharmacol.26:206-213.

33. Motulsky,H.J., and L.A. Ransnas. 1987. Fitting curves to data using nonlinearregression: a practical nonmathematicalreview. FASEB (Fed. Am. Soc. Exp. Biol.)J. 1:365-374.

34. Benovic, J. L., F,Mayor, Jr., R. L. Somers, M. G. Caron, and R. J. Lefkowitz. 1986.Light-dependent phosphorylation of rhodopsin byfl-adrenergic receptor kinase. Nature (Lond.). 321:869-872.

35.Benovic, J. L., F, Mayor, Jr., C. S. Staniszewski, M. G. Caron, and R. J. Lefkowitz. 1987. Purification and characterizationofbeta-adrenergic receptor kinase.J.Biol.Chem.262:9026-9032.

36.Garcia-Higuera,I.,P.Penela, C. Murga, G. Egea, P. Bonay, J. L.Benovic, andF.Mayor,Jr. 1994.Association of the regulatory ,B-adrenergic receptor ki-nase with rat liver microsomal membranes.J.Biol. Chem. 269:1348-1355.

37.Su,Y-F., T.K.Harden, and J. P. Perkins. 1980.Catecholamine-specific desensitization of adenylate cyclase. J. Biol.Chem.255:7410-7419.

38.Kawai,Y., S.M.Graham,C.Whitsel,andI.J.Arinze. 1985.Hepatic adenylatecyclase. J.Biol.Chem.260:10826-10832.

39.Waldo, G. L., J. K. Northup, J. P. Perkins, and T. K. Harden. 1983. Characterization ofanalteredmembraneform of the beta-adrenergic receptor produced during agonist-induced desensitization. J. Biol. Chem. 258:13900-13908.

40.Staehlin,M., andP.Simons.1982.Rapidandreversibledisappearanceof

fl-adrenergiccell surfacereceptors. EMBO (Eur. Mol. Biol. Organ.) J. 1:187-190.

41.Zastrow, M.,vonandB.Kobilka. 1992.Ligand-regulated internalization

andrecyclingof humanf32-adrenergicreceptors between the plasma membrane and endosomescontainingtransferrin receptors. J Biol. Chem. 267:3530-3538. 42. Yu, S. S., R. J. Lefkowitz,andW. P.Hausdorff. 1993.

f3-Adrenergic

receptorsequestration:apotential mechanism of receptor resensitization.J.Biol. Chem. 268:337-341.

43.Bristow,M.R., R. Gingsburg, W.Minobe,R.S.Cubicciotti,W.S. Sage-man,K. Lurie,M. E. Billingham, D.C. Harrinson,andE. B.Stinson. 1982. Decreased catecholaminesensitivityandfl-adrenergicreceptordensityinfailing

human hearts.N.Engl.J. Med.307:205-211.

44.Snavely,M.D.,L.C.Mahan,D. T.O'Connor,andP. A.Insel. 1983. Selectivedown-regulationofadrenergicreceptorsubtypesintissues fromrats

withpheochromocytoma. Endocrinology. 113:354-361.

45. Vatner,D.E., S.F.Vatner,J.Nejima,N.Uemura, E.E.Susanni,T.H.

Hintze, and C.J.Homcy. 1989. Chronicnorepinephrineelicitsdesensitization by uncouplingthe,B-receptor. J. Clin.Invest.84:1741-1748.

46.Garcia-Sainz,J.A., andB.Michel. 1987.Homologousfl-adrenergic de-sensitizationin isolatedrathepatocytes. Biochem.J.246:331-336.