phosphorylation of tyrosine hydroxylase by cgmp-dependent protein kinase in intact bovine chromaffin...

Pbi

FB

R

ymtuapiipclTewRmbtetiltttendsatc

ddm

0CA

Archives of Biochemistry and BiophysicsVol. 366, No. 2, June 15, pp. 207–214, 1999Article ID abbi.1999.1199, available online at http://www.idealibrary.com on

hosphorylation of Tyrosine Hydroxylasey cGMP-Dependent Protein Kinase

n Intact Bovine Chromaffin Cells

ernando Rodrıguez-Pascual, Rut Ferrero, M. Teresa Miras-Portugal, and Magdalena Torres1

iochemistry Department, Veterinary Faculty, Complutense University of Madrid, Madrid, Spain

eceived October 6, 1998, and in revised form February 15, 1999

dpIdaitics©

dC

nlbtcdicppv

tnysta

The phosphorylation of the enzyme tyrosine hydrox-lase by the cGMP pathway was investigated in chro-affin cells from the bovine adrenal medulla. The ni-

ric oxide donor, sodium nitroprusside, and the natri-retic peptide, C-type natriuretic peptide, which areble to increase cGMP levels and cGMP-dependentrotein kinase activity, produced significant increases

n the phosphorylation level of tyrosine hydroxylasen a time- and concentration-dependent manner. Theretreatment of the cells with the soluble guanylylyclase inhibitor, 1H-[1,2,4]oxadiazolo[4,3-a]quinoxa-in-1-one blocked the effect of sodium nitroprusside.his result indicates that cGMP production by thisnzyme mediated this effect. Experiments performedith a cGMP-dependent protein kinase inhibitor, thep-isomer of 8-(4-chlorophenylthio)-cyclic guanosineonophosphorothioate, which blocked the effects of

oth sodium nitroprusside and C-type natriuretic pep-ide, demonstrated that the phosphorylation increasesvoked by both compounds were mediated by the ac-ivation of cGMP-dependent protein kinase. In cellsncubated with the adenylyl cyclase activator, forsko-in, an increase in the phosphorylation level of theyrosine hydroxylase was also found. When cells werereated simultaneously with forskolin and sodium ni-roprusside or C-type natriuretic peptide, an additiveffect on tyrosine hydroxylase phosphorylation wasot observed. This suggests that cAMP- and cGMP-ependent protein kinases may phosphorylate theame amino acid residues in the enzyme. Western blotnalysis of soluble extracts from chromaffin cells de-ected specific immunoreactivity for two differentommercial antibodies raised against cGMP-depen-

1 To whom correspondence should be addressed at Departamentoe Bioquımica, Facultad de Veterinaria, Universidad Complutense
e Madrid, E-28040 Madrid, Spain. Fax: 34-91-394 39 09. E-mail:[email protected].
De

003-9861/99 $30.00opyright © 1999 by Academic Pressll rights of reproduction in any form reserved.

ent protein kinase (both Ia and Ib isoforms). Electro-horetic mobility correlates with that of purified PKGa. Because the phosphorylation of the tyrosine hy-roxylase correlates with increases in its enzymaticctivity and thus with augmentation in the cell capac-ty to synthesize catecholamines, our results indicatehat a cGMP-based second messenger pathway partic-pates in catecholamine biosynthesis regulation inhromaffin cells, a mechanism which may be wide-pread in other catecholamine-synthesizing cells.1999 Academic Press

Key Words: chromaffin cells; cGMP; cGMP-depen-ent protein kinase; tyrosine hydroxylase; SNP,-type natriuretic peptide.

Signaling molecules, such as nitric oxide (NO),2 andatriuretic peptides play key roles in several physio-

ogical functions. These messengers exert their effecty elevation of intracellular cyclic GMP (cGMP) levelshrough activation of soluble and cell-surface guanylylyclases, respectively. Unlike cAMP, which acts pre-ominantly through PKA activation, cGMP has severalntracellular target proteins: PKG, cGMP-gated ionhannels, and cGMP-stimulated or cGMP-inhibitedhosphodiesterases (1, 2). cGMP pathway elementsarticipate in numerous physiological processes. Ele-ation of cGMP and activation of PKG are involved in

2 Abbreviations used: NO, nitric oxide; PKA, cAMP-dependent pro-ein kinase; PKG, cGMP-dependent protein kinase; SNP, sodiumitroprusside; CNP, C-type natriuretic peptide; TH, tyrosine hydrox-lase; L-DOPA, 3,4-dihydroxyphenylalanine; PMSF, phenylmethyl-ulfonyl fluoride; Rp-8pCPT-cGMPS, Rp-isomer of 8-(4-chlorophenyl-hio)-cyclic guanosine monophosphorothioate; ODQ, 1H-[1,2,4]ox-diazolo[4,3-a]quinoxalin-1-one; BSA, bovine serum albumin;
MEM, Dulbecco’s modified Eagle’s medium; PBS, phosphate-buff-red saline; Pi, inorganic phosphate; NOS, nitric oxide synthase.
207

lstCnpcpflctfwab

tc2btcifamasget

stahcAtdmcpamatyhdfatcSc

eahcviostpddyzyiMsi

tdzirhmt

E

wocCwUfhbwINf8[dlyUsftl2tC

208 RODRIGUEZ-PASCUAL ET AL.

owering cytosolic Ca21 concentration in neurones,mooth muscle cells, and platelets, this fact suggestinghat PKG is a major factor in regulating intracellulara21 levels in different cell types (3, 4). This mecha-ism has also been observed in catecholaminergic ex-erimental models such as PC12 and adrenomedullaryhromaffin cells, where the NO/cGMP pathway ap-ears to selectively inhibit voltage-dependent Ca21 in-ux (5–7). In addition, by activating soluble guanylylyclase, NO modulates the release of several neuro-ransmitters, such as noradrenaline and glutamate,rom different preparations (8, 9). The signaling path-ay involving cGMP may also play further roles, suchs the control of gene expression, as has been describedy several groups (10, 11).Despite the existence of a variety of receptor pro-

eins, many of the physiological actions of intracellularGMP are mediated by the activation of PKG (EC.7.1.37). In mammalian tissues two types of PKG haveeen identified. Type I (a cytosolic form occurring aswo splicing variants of a single gene, Ia and Ib, withhain length of 670 and 684 amino acids, respectively)s widely expressed and significant concentrations areound in lung, cerebellum, platelets, smooth muscle,nd smooth muscle-like tissues. Type II (PKG II) is aembrane-bound form, which was originally described

s an intestine-specific form (12, 13). Molecular cloningtudies demonstrated that PKG II is indeed a distinctene product expressed predominantly in intestinalpithelial cells, but its mRNA has also been detected inhe kidney and brain (14).

Studies of the distribution of PKG in different tis-ues types by specific radioimmunoassay have reportedhe presence of PKG-I in various rat tissues includingdrenal tissue (15). Moreover, northern blot analysisas demonstrated a 7-kb mRNA for PKG Ib in raterebellum, cerebrum, lung, kidney, and adrenal (16).lthough no systematic immunological characteriza-

ion of PKG has been so far performed, there is evi-ence to suggest the presence of PKG in bovine chro-affin cells from adrenal medulla. Intracellular in-

reases in cGMP levels from bovine chromaffin cellsroduced an inhibition of both catecholamine releasend calcium entry elicited by secretagogues. Further-ore, activation of PKG seems to be involved in these

ctions because specific inhibitors of this enzyme effec-ively prevented them (6, 17). By measuring phosphor-lation of relatively specific substrates of PKG, such asistone F2b or the heptapeptide RKRSRAE, we haveetected kinase activity in soluble extracts obtainedrom bovine chromaffin cells activatable by cGMP an-logues and inhibitable by PKG inhibitors (6, 17). In-erestingly, pretreatment of chromaffin cells withGMP-increasing agents such as the nitric oxide donorNP, or the natriuretic peptide CNP, leads to an in-
rease in total catecholamine content (6). This latter
wi

ffect has also been described in PC12 cells where anctivation of tyrosine hydroxylase (TH, EC 1.14.16.2)as been observed in response to intracellular in-reases in cGMP (18). This enzyme catalyzes the con-ersion of tyrosine to L-DOPA, which is the rate-lim-ting step in the synthesis of catecholamines. A numberf different TH regulation mechanisms have been de-cribed in the literature, including inhibition feedback,ranscriptional regulation, and covalent modificationrocesses such as phosphorylation (19). Tyrosine hy-roxylase is phosphorylated in vitro by at least sevenifferent protein kinases, and many of these phosphor-lation events can be associated with increases in en-yme activity (19). Phosphorylation of tyrosine hydrox-lase by PKG has been demonstrated both in vitro andn vivo in response to natriuretic peptides (8, 20, 21).

oreover an activation of this enzyme occurs in re-ponse to intracellular increases in cGMP and cAMP inntact cells (18, 20).

In this experimental study we sought to characterizehe PKG present in chromaffin cells from adrenal me-ulla and to study whether the activation of this en-yme by increasing cGMP levels is accompanied by anncrease in phosphorylation protein. We selected ty-osine hydroxylase as a possible substrate becauseigh concentrations of this enzyme are present in chro-affin cells and its physiological importance is unques-

ionable.

XPERIMENTAL PROCEDURES

Materials. Culture medium and heat-inactivated fetal calf serumere obtained from GIBCO (Uxbridge, UK). Culture plates werebtained from Costar (Cambridge, MA). Collagen A was from Bio-hrom KG (Berlin, Germany). Collagenase A (EC 3.4.24.3) fromlostridium hystoliticum and phenylmethylsulfonyl fluoride (PMSF)ere purchased from Boehringer-Mannheim (Mannheim, Germany).rografin was from Schering Espana (Madrid, Spain). CNP was

rom Peninsula Laboratories (Belmont, CA). IP20-amide peptide in-ibitor of PKA, forskolin from Coleus forskohlii, PKG (isoform Ia,ovine recombinant), and anti-PKG Ia and Iß, (polyclonal, rabbit)ere purchased from Calbiochem (San Diego, CA). Anti-PKG Ia and

b (polyclonal, rabbit) was from Upstate biotechnology (Lake Placid,Y). Heptapeptide (RKRSRAE) substrate for PKG was obtained

rom Promega Corporation (Madison, WI). The cGMP analogue Rp--pCPT-cGMPS was from Biolog (Bremen, Germany). [g-32P]ATP,

32P]orthophosphate, a [3H]cGMP radioimmunoassay kit, autora-iography films (Hyperfilm-ßmax and Hyperfilm ECL), nitrocellu-ose membranes (Hybond ECL), and an ECL Western blotting anal-sis system were purchased from Amersham (Buckinghamshire,K). Protein molecular weight markers (MultiMark multicolored

tandard) were from Novex (San Diego, CA). ODQ was purchasedrom Tocris-Cookson (Langford Bristol, UK). SNP, leupeptin, pepsta-in, anti-tyrosine hydroxylase (monoclonal, mouse), anti-mouse IgGinked to agarose, nonimmune mouse IgG, Nonidet P-40, and Tween0 were from Sigma Chemical Co. (St. Louis, MO). Rabbit anti-yrosine hydroxylase polyclonal antibody was purchased fromhemicon International ICN (Temecula, CA).Isolation and culture of bovine chromaffin cells. Chromaffin cells
ere obtained after digestion of bovine adrenal glands by collagenase
n retrograde perfusion as previously described elsewhere (6).

BfMNfAmwmwntpiFcamtftfv

ptCHT0t3wwcr

ispwctrrmßCU1cmb0iss3i2m

(1mss

aFnimn41mtabrUcutmPi

cafccCpct5mmsacc(cfi(pfibmgapd

Mye

bttfbisw1p

209TYROSINE HYDROXYLASE PHOSPHORYLATION BY cGMP-DEPENDENT PROTEIN KINASE

riefly, glands supplied by a local slaughterhouse were trimmed ofat, cannulated through the adrenal vein, and washed with Ca21/

g21-free saline buffer, containing (in mM) NaCl 154, KCl 5.6,aHCO3 3.6, glucose 5, and Hepes 5, pH 7.4. Digestion was per-

ormed with a 0.2% collagenase plus 0.5% BSA in the above medium.fter digestion, glands were halved, soft medulla were removed andinced, and dispersed cells were filtered through a nylon mesh. Cellsere purified through an Urografin density gradient according to theethod described by Wilson (22). Of the collected cells, .90–95%ere chromaffin cells, as they were massively and clearly stained byeutral red. Occasionally, the purity of the cultures was lower thanhis range; in this case, cells were then purified by differentiallating. Purified chromaffin cells were suspended in DMEM contain-ng 10% heat-inactivated fetal calf serum and standard antibiotics.or cyclic GMP measurements and phosphorylation experiments,ells were plated on collagen-treated 24-well Costar cluster dishes atdensity of 106 cells/well in culture medium supplemented with 10M cytosinearabinofuranoside and 10 mM fluorodeoxyuridine, main-ained at 37°C in 5% CO2/95% air. These cells were used 3–5 daysollowing cell isolation. For PKG determinations, cells were main-ained in suspension and kept at 4°C. These cells were used 2–3 daysollowing cell isolation, at this time the purity of the cultures wasery high.Intracellular cyclic GMP measurements. Cells were serum-de-

rived for 24 h before cGMP measurements, they were then washedwice with Locke’s solution (composition (in mM): NaCl 140, KCl 4.4,aCl2 2.5, MgSO4 1.2, KH2PO4 1.2, NaHCO3 4, glucose 5.6, andepes 10, pH 7.4) and preincubated for 30 min in the same medium.hey were then stimulated with the required agents (SNP or CNP) in.5 ml of the same solution for the indicated times at 37°C. Incuba-ions were terminated by aspiration of the medium and addition of00 ml of 6% trichloroacetic acid. Cells were then scraped out of theells and centrifuged. The supernatant fraction was neutralizedith 3 M KOH, and the cyclic GMP content was determined in the

rude extracts with the use of a commercial 3H-labeled cyclic GMPadioinmunoassay kit (Amersham) (23).cGMP-dependent protein kinase activity assays. In experiments

n which PKG was measured in vitro after specific treatments in situ,uspended chromaffin cells were washed twice with Locke’s solution,reincubated for 30 min in the same medium, and then stimulatedith 100 mM SNP or 100 nM CNP at 37°C. Aliquots (500 ml) of

hromaffin cells suspensions (about 5 3 106 cells) were taken out athe indicated times. The cells were harvested by centrifugation (3000pm, 1 min). The supernatants were aspirated and the cells wereapidly frozen by immersing the tubes in liquid N2. Then, 200 ml of 10M potassium phosphate, pH 6.8, containing 1 mM EDTA, 15 mM

-mercaptoethanol, 1 mM PMSF, and 10 mM leupeptin was added.ells were disrupted by sonication for 10 s with a Braun Micro-ltrasonic Cell Disrupter and the suspensions were centrifuged at3,000 rpm for 10 min at 4°C. Portions (30 ml) of the supernatantsontaining 60–70 mg of protein were taken for kinase activity deter-inations as previously described (6). Assays were carried out in a

uffer consisting of 40 mM Tris, pH 7.4, 20 mM magnesium acetate,.2 mM [g-32P]ATP (200–300 cpm/pmol), 50 nM IP20-amide peptidenhibitor of PKA, and 10 mg of heptapeptide (RKRSRAE) as sub-trate of PKG. The reaction was initiated by addition of 30 ml ofoluble extract (total volume 100 ml) and incubated for 10 min at0°C. The reaction was terminated by pipetting an aliquot of thencubation mixture onto filter paper squares (Whatman P81, 2.5 3.5 cm) and washing with 75 mM phosphoric acid (four times with 10l/paper). The filters were dried and the radioactivity was counted.Detection of PKG by immunoblotting. Cultured chromaffin cells

25 3 106) were resuspended in 500 ml of a buffer solution containing50 mM NaCl, 10 mM EDTA, 1 mM PMSF, 10 mg/ml leupeptin, 10g/ml pepstatin, and 20 mM Tris, pH 7.4, and homogenized by
onication. After centrifugation at 13,000 rpm for 10 min at 4°C, theupernatant was kept. Proteins were electrophoresed on 10% SDS–
dq

crylamide gels at 200 V for 45 min according to Laemmli (24).ollowing protein separation they were transferred to a Hybond ECLitrocellulose membrane at 300 mA for 2 h. The membrane was then

ncubated in PBS containing 0.1% Tween 20 and 5% nonfat driedilk for 1 h at room temperature on an orbital shaker to block

onspecific protein binding. Blots were then incubated overnight at°C with a commercial antibody against PKG I (Calbiochem) at a:300 dilution in PBS containing 0.1% Tween 20 and 5% nonfat driedilk. The antibody was raised against a synthetic peptide containing

he C-terminal region of cGMP-dependent protein kinase (types Iand Ib) (25). After several cycles of washing with PBS–Tween 20,lots were incubated with the secondary antibody linked to horse-adish peroxidase at a 1:1000 dilution. When the antibody frompstate biotechnology was used, the dilution was 1:1000 in PBS

ontaining 3% nonfat dried milk, and the secondary antibody wassed at a dilution of 1:1500. Then they were washed and the detec-ion was performed by using the enhanced cheminoluminiscenceethod as described by the manufacturer (Amersham Life Science).KG was quantitated by densitometric scanning (Molecular Dynam-

cs densitometer).Tyrosine hydroxylase phosphorylation experiments. Chromaffin

ells were washed once with 1 ml of phosphate-free Locke’s buffer,nd incubated with this medium for 1 h. They were then incubatedor an additional hour at 37°C with 200 ml of the same mediumontaining 0.1 mCi of [32P]orthophosphate. At this loading period,ells were stimulated with the required stimulatory agents (SNP,NP, forskolin, or basal) for the indicated times. The stimulationeriod was terminated by removing the medium and scraping theells from the plates into 300 ml of 20 mM Tris–HCl, pH 7.5, con-aining 1 mg/ml BSA, 1% Nonidet P-40, 4 mM EGTA, 10 mM EDTA,0 mM NaF, 1 mM Na3VO4, 1 mM PMSF, 10 mg/ml leupeptin, and 10g/ml pepstatin. This suspension was centrifuged at 10,000g for 15in at 4°C and the pellet discarded. The protein concentration of the

oluble material was determined using a Bradford assay (DC proteinssay, Bradford). These supernatants were subjected to immunopre-ipitation analysis as follows. Equal amounts of protein were pre-leared by incubation with goat anti-mouse IgG bound to agarose1:18) for 1 h at 4°C and then centrifuged at 200g for 1 min. Pre-leared samples (60–100 ml) were incubated overnight at 4°C in anal volume of 200 ml with either monoclonal anti-TH antibody

1:100) or nonimmune mouse IgG (1:100). Immune complexes wererecipitated using goat anti-mouse IgG-agarose (final dilution, 1:18)or 1 h at 4°C. The suspension was centrifuged at 200g for 1 min. Themmobilized immune complexes were washed five times with lysinguffer without BSA and inhibitors. The pellet was resuspended in 30l of electrophoresis loading buffer (62.5 mM Tris–HCl, pH 6.8, 10%lycerol, 5% ß-mercaptoethanol, 3% SDS, 0.005% bromophenol blue)nd heated at 95°C for 5 min. Finally, samples were centrifuged androteins were separated on 10% SDS–acrylamide gels. Gels wereried and exposed to hyperfilm-ßmax films for 4 to 6 days.Autoradiograms from one-dimensional gels were scanned using aolecular Dynamics densitometer. Densities of the tyrosine hydrox-

lase bands were determined using computer-assisted analysis andxpressed as a percentage of the basal samples run on the same gel.The quantity of TH loaded in each lane was verified by immuno-

lotting. In these cases nonlabeled samples were immunoprecipi-ated as described above. Following protein separation they wereransferred to a Hybond ECL nitrocellulose membrane, as describedor PKG detection. Immunodetection was performed by using a rab-it anti-tyrosine hydroxylase polyclonal antibody at a 1:2000 dilutionn PBS containing 0.1% Tween 20 and 5% nonfat dried milk. Aftereveral cycles of washing with PBS–Tween 20, blots were incubatedith the secondary antibody linked to horseradish peroxidase at a:10,000 dilution. Then they were washed and the detection waserformed by using the enhanced cheminoluminiscence method as
escribed by the manufacturer (Amersham Life Science). TH wasuantitated by densitometric scanning (Molecular Dynamics densi-

ta

R

cwScS(pkttsaiSTdcc

oecrraas(

d6sFi7trn

BSSSSSCCCCC

tkpd hatp t le

FcetIwallc


ometer). The differences in TH protein amount among samples werelways less than 5%.

ESULTS

cGMP-dependent protein kinase in bovine chromaffinells. Basal cGMP levels in bovine chromaffin cellsere 0.976 6 0.05 pmol/106 cells. Addition of 100 mMNP or 100 nM CNP produced time-dependent in-reases in cGMP levels (13 6 2-fold and 5 6 1-fold forNP and CNP, respectively, at 15 min of stimulation)

Table I). By employing a method based on the phos-horylation of an exogenous substrate of PKG, theinase activity was determined in soluble extractsaken from chromaffin cells incubated for differentime periods with SNP and CNP. Table I also showsignificant increases compared to basal values in PKGctivity when cells were challenged with these cGMP-ncreasing agents (1.9 6 0.2-fold and 2.0 6 0.2-fold forNP and CNP, respectively, at 15 min of stimulation).hese results, in agreement with previous studies, in-icate that chromaffin cells have the elements of theGMP system: soluble and membrane-bound guanylylyclase and a protein kinase with PKG properties.

In order to characterize as far as possible the isoformf PKG present in these cells, we employed two differ-nt commercial antibodies for PKG I which have re-ently become available. One of the antibodies wasaised against a synthetic peptide of the C-terminalegion of PKG type I and specifically identifies PKG Iand Ib from several species including cattle, humans,nd rats (25). The other one was raised against aynthetic peptide of the N-terminal cysteine added

TAB

Effect of Sodium Nitroprusside (SNP) and C-Type Natr

TreatmentcGMP levels

(pmol/106 cells) Percentage of ba

asal 0.976 6 0.050 100NP, 1 min 7.12 6 0.50 734NP, 2 min 8.12 6 0.84 837NP, 5 min 9.310 6 1.332 954NP, 15 min 12.911 6 2.299 1323NP, 30 min 11.277 6 0.507 1155NP, 1 min 3.58 6 0.12 369NP, 2 min 4.60 6 0.25 474NP, 5 min 4.898 6 0.514 502NP, 15 min 4.555 6 0.358 467NP, 30 min 4.774 6 0.360 489

Note. Cultured chromaffin cells were preincubated with Locke’s soimes with 100 mM SNP or 100 nM CNP in the same medium. After thinase activity were determined as described under Experimental Phorylation, respectively. cGMP levels are expressed as pmol/106 ceefined as the enzymatic activity able to incorporate 1 pmol phosperformed in triplicate and data are means 6 standard error from a

CDEPPPDDNSGWDIDF) corresponding to the resi-ef

ues 657–671 of the human PKG type Ia and residues72–682 of human PKG type Ib. This residue alsohows 15/15 homology to bovine PKG a and b isoforms.igure 1 shows immunoreactivity migrating at 74 kDa

n soluble extracts from chromaffin cells (lanes 6 and). Purified PKG Ia was used as a positive control inhis blot (lanes 1–5). Different amounts of purifiedecombinant PKG Ia (0.5–8 ng, lanes 1–5) were immu-odetected as a band that migrated at 74 kDa in each

I

etic Peptide (CNP) on cGMP Levels and PKG Activity

(%)PKG activity

(unit/mg protein) Percentage of basal (%)

441.2 6 89.4 100472.1 6 58.7 107539.2 6 79.5 122794.1 6 93.7 180837.4 6 112.3 190

1155.9 6 110.3 262525.5 6 48.7 118523.0 6 76.3 119750.6 6 63.0 170893.1 6 97.1 202816.2 6 129.3 185

ion at 37°C for 30 min. Then they were incubated for the indicatedincubation period the cGMP levels and the cGMP-dependent proteinedures through radioimmunoassay and exogenous substrate phos-and kinase activity as units per milligram of protein. One unit ise/min into the heptapeptide substrate at 30°C. Experiments wereast three different cellular preparations (n 5 6).

IG. 1. Immunodetection of PKG I in soluble extracts of bovinehromaffin cells. The presence of cGMP-dependent protein kinase inxtracts of chromaffin cells was studied using the immunoblottingechnique with a specific antibody (Calbiochem) against PKG Ia andb as described under Experimental Procedures. Purified PKG Iaas used as the positive control. Lanes 1–5 correspond to increasingmounts of purified recombinant PKG Ia (0.5, 1, 2, 4, and 8 ng). Inane 6, 1 ng of purified PKG plus 10 mg of cromaffin cell extracts wasoaded. Lane 7 corresponds to 10 mg of soluble extracts of chromaffinells (CC). The photograph shows the result of a representative

LE

iur

sal

lutisroclls

xperiment of four experiments perfomed with chromaffin extractsrom different cultures.

lefUote

comc2ytsz

[CTiacebpeiw(mssS1tbilt

pfaTtcspg(ptt

ipapne(

yes

Fhcaitbcstpiswir


ane. Figure 1 shows the results of a representativexperiment of four performed with chromaffin extractsrom different cultures which gave identical results.sing the peak area obtained for the different amounts

f purified PKG after scanning the film by densitome-ry, a PKG concentration of 52 6 4 ng/mg protein wasstimated in chromaffin cells extracts.Tyrosine hydroxylase phosphorylation in response to

GMP-increasing agents. Previous studies (includingurs) have shown that preincubation of bovine chro-affin cells with cGMP-increasing agents produces in-

reases in the cellular content of catecholamines (6,6). Because the basis of this effect may be a phosphor-lation by PKG of the enzyme tyrosine hydroxylase,he rate-limiting step in the catecholamine biosynthe-is, we studied the phosphorylation levels of this en-yme in cells preincubated with SNP or CNP.Bovine chromaffin cells were preincubated with

32P]orthophosphate and stimulated by either SNP orNP for varying lengths of time. 32Pi incorporated intoH was determined after immunoprecipitation by us-

ng a monoclonal anti-TH antibody. This techniquellowed accurate determination of radiophosphate in-orporation into TH because it made it possible tovaluate small changes in radioactivity on an uniqueand, avoiding many other proteins whose changes inhosphorylation might mask them. These experimentsstablished that some 32Pi incorporation into TH occursn basal chromaffin cells as shown in Fig. 2 (lane 2),hich was significantly enhanced in response to SNP

lanes 3–5) or CNP (lanes 6–8) (1.6 6 0.15-fold at 15in of stimulation) and in a time-dependent manner. A

ignificant increase in TH phosphorylation was ob-erved within 5 min of stimulation with either 100 mMNP or 100 nM CNP. Phosphorylation was maximal at5 min stimulation with SNP or CNP and decreasedhereafter, but remained high after 30 min relative toasal conditions. Figure 2B shows a Western blot formmunoprecipitated TH from chromaffin cells stimu-ated with 100 mM SNP or 100 nM CNP for differentimes.

Dose–response studies for SNP and CNP on THhosphorylation were performed on cells stimulatedor 15 min with cGMP-increasing compounds (Figs. 3And 3B, respectively). This figure shows that maximalH phosphorylation was achieved at SNP concentra-ions of 10–100 mM (lanes 4 and 5), but even at 1 mMoncentration an increase in 32P incorporation was ob-erved (lane 3) as shown in Fig. 3A. SNP-induced THhosphorylation was clearly prevented by the solubleuanylyl cyclase-selective inhibitor ODQ at 10 mMlane 6) and by the specific inhibitor for PKG (Rp-8-CPT-cGMPS) at 10 mM (lane 7). In the presence ofhese inhibitors, TH phosphorylation was even lower
han that observed in basal conditions (lane 2), thus s
ndicating that PKG may participate in basal TH phos-horylation. When CNP was employed as stimulatorygent, as shown in Fig. 3B, the maximum of TH phos-horylation was observed at concentrations of 10–100M (lanes 4 and 5). In this case, Rp-8-pCPT-cGMPSffectively abolished CNP-induced TH phosphorylationlane 7), whereas ODQ failed to prevent it (lane 6).

It has been reported that TH is effectively phosphor-lated by PKA in bovine chromaffin cells, and severalxperiments with purified TH from PC12 cells havehown that both PKA and PKG phosphorylated the

IG. 2. Time course of the effect of SNP and CNP on tyrosineydroxylase phosphorylation in bovine chromaffin cells. Chromaffinells preincubated for 1 h with [32P]orthophosphate were stimulatedt this loading period with 100 mM SNP or 100 nM CNP for thendicated times or basal (B). The phosphorylation level of the enzymeyrosine hydroxylase (molecular mass of '60 kDa) was determinedy immunoprecipitation with an anti-TH antibody and SDS–PAGEombined with autoradiography (A) and Western blot (B) as de-cribed under Experimental Procedures. The first lane correspondso a sample in which the anti-TH antibody was omitted (2). The THhosphorylation level and the potentiation reached by the cGMP-ncreasing agents were variable from culture to culture. The figurehows the mean 6 standard error of four experiments performedith different cultures. Autoradiogram show the result of one exper-

ment, but it is representative of the three performed with similaresults.

ame amino acid residues (18). As we have previously

ow3f4pcSrsau

D

ib2adpssssaat

FCCipw the results given as means 6 standard error of three experimentsp of a single representative experiment.

FlwomClwfioe


bserved that when chromaffin cells were stimulatedith forskolin maximal cAMP accumulation occurs at0 min, we selected this as the length of time to studyorskolin-induced phosphorylation. As is shown in Fig., 30 min stimulation with forskolin increased THhosphorylation up to 180% (lane 5). When cGMP andAMP were simultaneously increased by forskolin plusNP (lane 6) or forskolin plus CNP (lane 7) no furtheradiophosphate incorporation occurred over that ob-erved with forskolin alone. In these experiments SNPnd CNP were added during the latest 15 min of stim-lation with forskolin.

ISCUSSION

The existence of a cGMP-based intracellular signal-ng pathway in bovine chromaffin cells has alreadyeen reported in several other studies (6, 21, 23, 26,7). Some functional studies performed in these cellsnd other related cell types such as PC12 cells haveemonstrated the participation of a cGMP-dependentrotein kinase in important physiological processesuch as the modulation of catecholamine release or theynthesis of catecholamines (6, 20). Our study demon-trates that soluble extracts of bovine chromaffin cellshowed immunoreactivity for two different commercialntibodies which recognize both isoforms of PKG I (Iand Ib). One of these isoforms may be responsible for

IG. 3. Dose dependence of the effect of SNP and CNP on TH phromaffin cells preincubated for 1 h with [32P]orthophosphate werNP and TH phosphorylation levels determined as described unde

nhibitor, 10 mM ODQ, and the specific inhibitor of PKG, Rp-8phosphorylation evoked by 100 mM SNP and 100 nM CNP. The first las omitted (2). (B) Basal TH phosphorylation. The figure showserformed with different cultures. Autoradiograms show the result

hosphorylation: Effect of the inhibitors ODQ and Rp-8pCPT-cGMPS.e stimulated the last 15 min with increasing concentrations of SNP orr Experimental Procedures. The effect of the soluble guanylyl cyclaseCPT-cGMPS (abbreviated as Rp2; 10 mM), was studied on the THanes of each panel correspond to samples in which the anti-TH antibody

he observed effects of the cGMP-increasing agents onat

IG. 4. Effect of SNP or CNP and forskolin on the TH phosphory-ation levels in bovine chromaffin cells. Cultured chromaffin cellsere preincubated for 1 h with [32P]orthophosphate in the presence

f 100 mM SNP, 100 nM CNP (15 min), 10 mM forskolin (30 min), orixtures of these compounds. When pairs of drugs SNP/forskolin orNP/forskolin were assayed together, SNP or CNP was added the

ast 15 min of incubation with forskolin. TH phosphorylation levelsere determined as described under Experimental Procedures. Therst lane corresponds to a sample in which the anti-TH antibody wasmitted. The figure shows the results given as means 6 standardrror of three experiments performed with different cultures. The
utoradiogram correspond to a representative experiment of thehree performed with similar results.

kl

ravbfictsmfsir

trfptnrhToccf

rntoiTTcwccaT

tmtrpi(sttl

lmciUcT(rmbbuastphgphtmc

cptstbpssTytsptppwemtppac(

atS


inase activity measurements and TH phosphorylationevels.

Using the antibodies mentioned above and purifiedecombinant PKG Ia as a positive control, Western blotnalysis has shown specific immunoreactivity in bo-ine chromaffin cells extracts. As expected, this anti-ody stained a unique band in lanes loaded with puri-ed PKG Ia (apparent molecular mass, 74 KDa). Inhromaffin cells extracts one band was immunode-ected, and the molecular mass was equal to that ob-erved for the purified protein. Since the molecularass described for PKG Iß is higher than that reported

or PKG Ia (28), even though the antibody used in thistudy recognized both Ia and Iß isoforms, the bandmmunodetected in chromaffin cell extracts should cor-espond with PKG Ia (25).

Immunological detection and concentration estima-ions of PKG have been previously reported for variousat and bovine tissues (15) and the amount describedor rat adrenal gland is 0.14 pmol PKG per milligram ofrotein. Assuming a molecular weight of 150,000 forhe holoenzyme (28), this value would correspond to 21g PKG per milligram of protein. Thus the value weeported here (52 6 4 ng/mg protein) is significantlyigher than that obtained in the rat adrenal gland.his difference might indicate a higher concentrationf PKG in bovine chromaffin cells in relation to otherellular components of the adrenal gland, or simplyould be due to differences between species or the dif-erent techniques employed in each study.

In this study we show that radiophosphate incorpo-ation into TH was augmented by the nitric oxide do-or and the natriuretic peptide which activated PKGhrough increases in cGMP levels. It is worth pointingut that although SNP was more potent than CNP inncreasing cGMP levels, its effects on PKG activity andH phosphorylation were very similar to those of CNP.his observation could be explained by the fact that theGMP concentration reached after 15 min stimulationith CNP, which can be calculated considering a intra-

ellular volume of water of 1.38 ml per million bovinehromaffin cells (29), is high enough to produce a fullctivation of PKG. Time courses for SNP and CNP onH phosphorylation and PKG activation showed that

32Pi incorporation was maximum at 15 min, droppinghereafter, whereas both cGMP and PKG activity re-ained heightened. The presence of protein phospha-

ases may explain these differences because it has beeneported that okadaic acid, a well-known inhibitor ofrotein phosphatases, enhances phosphorylation of THn several serine residues in PC12 and chromaffin cells30, 31). On the other hand, it has been described thatome biological actions of PKG are mediated by activa-ion of protein phosphatase 2A (32, 33). Although TH ishe substrate for various protein kinases, different
ines of evidence demonstrate that PKG phosphory- n
ates TH in response to cGMP increases in bovine chro-affin cells: (i) Stimulation with SNP or CNP was

arried out in the absence of any phosphodiesterasenhibitor in order to avoid any effect on cAMP levels.nder these experimental conditions no changes in

AMP levels were observed as described before (6). (ii)he specific inhibitor for PKG, Rp-8-pCPT-cGMPS

12), effectively prevented 32Pi incorporation into TH inesponse to both CNP and SNP. Pretreatment of chro-affin cells with this compound even diminished the

asal radiophosphate incorporation, indicating thatasal PKG activity participates in TH phosphorylationnder rest conditions. In a previous study basal NOSnd soluble guanylyl cyclase activities were demon-trated in bovine chromaffin cells, and they were foundo contribute to maintaining basal levels of cGMP andutatively tonic PKG activity (34). (iii) On the otherand, as expected, the specific inhibitor for the solubleuanylyl cyclase isoform, ODQ (35), only abolishedhosphorylation induced by SNP whereas it did notave any effect on CNP stimulation, indicating thathe effect of the nitric oxide donor was specificallyediated by cGMP production by soluble guanylyl cy-

lase.Tyrosine hydroxylase catalyzes the rate-limiting and

ommitted step of the catecholamine biosynthesisathway, and its activity is regulated by phosphoryla-ion by different protein kinases (19). Roskosky et al.howed that cGMP and cAMP second messenger sys-ems enhanced TH activity in both PC12 cells andovine chromaffin cells. These authors proposed thathosphorylation of TH by PKA or PKG might be re-ponsible for the observed effect. The same authorsucceed in finding in vitro phosphorylation of purifiedH from PC12 cells by purified PKG, but the phosphor-lation of TH by the endogenous kinase was not inves-igated either in chromaffin or PC12 cells (20). In thistudy, we have described a cGMP- and PKG-mediatedhosphorylation of TH in intact chromaffin cells, andhese results correlate well with those previously re-orted (21). However, Haycock (30) did not find THhosphorylation after stimulating chromaffin cellsith nitroprusside 100 mM for 20 min. This lack of

ffect might be explained because we observed a maxi-un at 15 min and a fall in 32P incorporation thereaf-

er, and it is possible that at 20 min the increase inhosphorylation would be very low. The results of theresent study may explain both the increase of THctivity and catecholamine content in response toGMP-increasing agents reported in different studies6, 26).

Phosphorylation sites on TH have been identified,nd all of the modified serine residues were found inhe amino-terminal regulatory domain (Ser8, Ser19, ander40) (30). There is significant heterogeneity in the
ature of the kinases that modify a given residue and

mpswyPb

PcPfseattcs

A

tgCfD

R

1

1

1

11

11

1

1

12

2

22

22

2

2

2

233

3

3

3


ost of the traditional second messenger systemshosphorylate Ser40 (19). In this study we show thatimultaneous stimulation of cAMP and cGMP path-ays does not have an additive effect on TH phosphor-lation. These results might indicate that PKA andKG phosphorylate the same residues of TH, as haseen shown by Roskoski et al. (20).In summary, our results demonstrate a cGMP- and

KG-mediated phosphorylation of TH in chromaffinells. Immunological experiments show the presence ofKG I immunoreactivity in soluble extracts obtained

rom these cells. This protein kinase might be respon-ible for the observed phosphorylation of TH and alsoxplain the augmentation of catecholamine contentnd activation of TH activity described in the litera-ure. These results suggest that PKG plays an impor-ant role in regulating catecholamine synthesis inhromaffin cells, a mechanism which may be wide-pread in other catecholamine-synthesizing cells.

CKNOWLEDGMENTS

This research was supported by a grant from the Spanish Minis-erio de Educacion y Cultura, DGES (PM96-0053). Fernando Rodri-uez-Pascual was supported by a fellowship from the Universidadomplutense de Madrid. Rut Ferrero was supported by a fellowship

rom the Spanish Ministerio de Educacion y Cultura. We thankuncan Gilson for help in preparation of the manuscript.

EFERENCES

1. Schmidt, H. H. W., Lohmann, S. M., and Walter, U. (1993) Circ.Res. 77, 841–848.

2. Francis, S. H., and Corbin, J. D. (1994) Annu. Rev. Physiol. 56,237–272.

3. Cornwell, T. L., and Lincoln, T. M. (1989) J. Biol. Chem. 264,1146–1155.

4. Wang, X., and Robinson, P. J. (1997) J. Neurochem. 68, 443–456.5. Rodrıguez-Pascual, F., Miras-Portugal, M. T., and Torres, M.

(1994) Neurosci. Lett. 180, 269–272.6. Rodrıguez-Pascual, F., Miras-Portugal, M. T., and Torres, M.

(1996) Mol. Pharmacol. 49, 1058–1070.7. Simard, J. M., Tewari, K., Kaul, A., Nowicki, B., Chin, L. S.,

Singh, S. K., and Perez-Polo, J. R. (1996) Neurosci. Res. 45,216–225.

8. Schwarz, P., Diem, R., Dun, N. J., and Forstermann, U. (1995)Circ. Res. 77, 841–848.

9. Sistiaga, A., Miras-Portugal, M. T., and Sanchez-Prieto, J. (1997)Eur. J. Pharmacol. 321, 247–257.

3

0. Haby, C., Lisovoski, F., Aunis, D., and Zwiller, J. (1994) J. Neu-rochem. 62, 496–501.

1. Gudi, T., Huvar, I., Meinecke, M., Lohmann, S. M., Boss, G. R.,and Pilz, R. B. (1996) J. Biol. Chem. 271, 4597–4600.

2. Butt, E., Eigenthaler, M., and Genisser, H. G. (1994) Eur.J. Pharmacol. 269, 265–268.

3. DeJonge, H. R. (1981) Adv. Cyclic Nucleotide Res. 14, 315–333.4. Jarchau, T., Hausler, C., Markert, T., Pohler, D., Vandekerck-

hove, J., DeJonge, H. R., Lohmann, S. M., and Walter, U. (1994)Proc. Natl. Acad. Sci. USA 91, 9426–9430.

5. Walter, U. (1981) Eur. J. Biochem. 118, 339–346.6. Sandberg, M., Natarajan, V., Ronander, I., Kalderon, D., Walter,

U., Lohmann, S. M., and Jhansen, T. (1989) FEBS Lett. 255,321–329.

7. Rodrıguez-Pascual, F., Miras-Portugal, M. T., and Torres, M.(1995) Neuroscience 67, 149–157.

8. Roskoski, R., Jr., and Roskoski, L. M. (1987) J. Neurochem. 48,236–242.

9. Kumer, S. C., and Vana, K. E. (1996) J. Neurochem. 67, 443–462.0. Roskoski, R., Jr., Vulliet, P. R., and Glass, D. B. (1987) J. Neu-

rochem. 48, 840–845.1. Yanagihara, N., Okazaki, M., Terao, T., Uezono, Y., Wada, A.,

and Izumi, F. (1991) Naunyn-Schiedeberg’s Arch. Pharmacol.343, 289–295.

2. Wilson, S. P. (1987) J. Neurosci. Methods 19, 163–171.3. Rodrıguez-Pascual, F., Miras-Portugal, M. T., and Torres, M.

(1995) Biochem. Pharmacol. 50, 763–769.4. Laemmli, U. K. (1970) Nature 227, 680–685.5. Keilbach, A., Ruth, P., and Hofmann, F. (1992) Eur. J. Biochem.

208, 467–473.6. Tsutsui, M., Yanagihara, N., Minami, K., Kobayashi, H., Na-

kashima, Y., Kuroiwa, A., and Izumi, F. (1994) J. Pharmacol.Exp. Ther. 268, 584–589.

7. Babinski, K., Haddad, P., Vallerand, D., McNicoll, N., De Lean,A., and Ong, H. (1995) FEBS Lett. 313, 300–302.

8. Butt, E., Geiger, T., Jarchau, T., Lohmann, S. M., and Walter, U.(1993) Neurochem. Res. 18, 27–42.

9. Haycock, J. W. (1990) J. Biol. Chem. 265, 11682–11691.0. Haycock, J. W. (1993) Neurochem. Res. 18, 15–26.1. Miras-Portugal, M. T., Torres, M., Rotllan, P., and Aunis, D.

(1986) J. Biol. Chem. 261, 1712–1719.2. Zhou, X-B., Ruth, P., Schossmann, J., Hofmann, F., and Korth,

M. (1996) J. Biol. Chem. 271, 19760–19767.3. White, R. E., Lee, A. B., Shcherbatko, A. D., Lincoln, T. M., Schon-

brunn, A., and Armstrong, D. L. (1993) Nature, 361, 263–266.4. Schwarz, P. M., Rodrıguez-Pascual, F., Koesling, D., Torres, M.,

and Forstermann (1998) Neuroscience 82, 255–265.
5. Garthwaite, J., Southam, E., Boulton, C. L., Nielsen, E. B.,
Schmidt, K., and Mayer, B. (1995) Mol. Pharmacol. 48, 184–188.

phosphorylation of tyrosine hydroxylase by cgmp-dependent protein kinase in intact bovine chromaffin...

Documents