www.elsevier.com/locate/cardiores

Cardiovascular Research

Dow

nloa

Membrane association of nitric oxide-sensitive guanylyl

cyclase in cardiomyocytesB

Luis Agullo, David Garcia-Dorado*, Noelia Escalona, Marisol Ruiz-Meana,

Maribel Mirabet, Javier Inserte, Jordi Soler-Soler

Servicio de Cardiologıa, Hospital Universitari Vall d’Hebron, Pg Vall d’Hebron 129, 08035 Barcelona, Spain

Received 7 September 2004; received in revised form 11 April 2005; accepted 18 May 2005

Available online 13 June 2005

Time for primary review 19 days

by guest on June 1, 2013http://cardiovascres.oxfordjournals.org/

ded from

Abstract

Objective: Although the importance of the cyclic GMP (cGMP) signaling pathway in cardiac myocytes is well established, little is known

about its regulation. Ca2+-dependent translocation of nitric oxide (NO) sensitive guanylyl cyclase (GCNO) to the cell membrane has been

recently proposed to play a role. The aim of this study was to determine the possible functional relevance of GCNO bound to the cardiomyocyte

membrane.

Methods: Cytosolic and particulate fractions of adult rat cardiomyocytes were isolated and blotted, and their GCNO activity was assayed in

parallel experiments.

Results: In untreated cardiomyocytes, approximately 30% of h1-and a1-subunits of GCNO and a similar proportion of GCNO activity were

found in the particulate fraction. The dependence of GCNO activity on pH, Ca2+, GTP and NO donor concentrations was similar in

particulate and cytosolic fractions. Treatment of cardiomyocytes with the ionophore A23187 caused GCNO to translocate to the

sarcolemma, increased GCNO activity in this fraction, and potentiated NO-mediated cGMP synthesis. These effects appeared to be

mediated by Ca2+-dependent changes on the phosphorylation status of GCNO, since they were enhanced by the non-selective inhibitor

staurosporine and by the selective inhibitor of Ca2+/calmodulin-dependent protein kinase KN-93. The effect of drugs increasing

intracellular Ca2+ on cGMP synthesis was clearly correlated with their effects on membrane-associated GCNO activity but not with their

effects on cytosol-associated GCNO.

Conclusion: These results are the first evidence that 1) GCNO is associated with the cell membrane in cardiomyocytes, 2) the regulation of

membrane-associated GCNO differs from that of cytosolic GCNO, and 3) membrane association may have a crucial role in determining the

response of cells to NO.

D 2005 European Society of Cardiology. Published by Elsevier B.V. All rights reserved.

Keywords: Nitric oxide; Calcium (cellular); Signal transduction; Myocytes; Protein phosphorylation

1. Introduction

Cyclic GMP (cGMP) modulates important physiological

functions in the cardiovascular system as vasodilation, Ca2+

cycling, endothelium permeability, or myocardial contrac-

0008-6363/$ - see front matter D 2005 European Society of Cardiology. Publishe

doi:10.1016/j.cardiores.2005.05.021

i Ajay Shah (King’s College, London) served as Guest Editor for this

manuscript

* Corresponding author. Tel.: +34 93 489 40 38; fax: +34 93 489 40 32.

E-mail address: dgdorado@vhebron.net (D. Garcia-Dorado).

tility [1]. Abundant evidence indicates that cGMP can

modulate cell death during ischemia-reperfusion [2–7] and

cGMP has been described to mediate late preconditioning in

conscious rabbits [8].

cGMP can be synthesized by two different types of

guanylyl cyclases: a nitric oxide (NO)-sensitive guanylyl

cyclase (GCNO), generally known as cytosolic or soluble

guanylyl cyclase, and guanylyl cyclases that are integral

proteins in the plasmatic membrane of the cell and can be

stimulated by natriuretic peptides. GCNO is constituted of

68 (2005) 65 – 74

d by Elsevier B.V. All rights reserved.

L. Agullo et al. / Cardiovascular Research 68 (2005) 65–7466

by guest ohttp://cardiovascres.oxfordjournals.org/

Dow

nloaded from

two subunits, a and h, and two different isoforms of the a

subunit (a1 and a2) and of the h subunit (h1 and h2) have

been described. The a1h1 heterodimer is predominantly

found in the cardiovascular system, while a2h1 has been

mainly found in brain.

Little is known about how GCNO is modulated in vivo.

Rapid desensitisation of the signal [9] and phosphorylation

by different protein kinases, as protein kinase C (PKC) [11],

cyclic AMP-dependent protein kinase (PKA) [10,11] and

cGMP-dependent protein kinase (PKG) [12], have been

described.

Recent studies have challenged the classical concept of

an exclusive cytosolic location of GCNO. The isoform a2h1

has been found tightly associated to the neuron membrane

through a PSD-95 mediated interaction [13]. A histochem-

ical study suggested the presence of the more amply

distributed heterodimer, a1h1, in the sarcolemmal region

of skeletal muscle fibres [14]. This association to the

particulate fraction of a1 and h1 subunits has been recently

demonstrated in myocardial tissue, endothelial cells and

platelets [15], and in these later cells activation with ADP or

collagen has been correlated with enzyme translocation to

the membrane [15]. However, the mechanism of the

interaction of a1h1 with the membrane was unresolved. A

recent study has described a protein complex formed by

eNOS, Hsp90 and GCNO in endothelial cells, and that

bradykinin and vascular endothelial growth factor potentiate

the formation of this complex [16]. The contribution of

particulate GCNO to the cell response to NO remained

unknown.

In this study, we analyze how association of GCNO to

membrane affects its biochemical properties in cardiomyo-

cytes, the functional relevance of this association, and its

potential regulation by Ca2+.

n June 1, 2013

2. Materials and methods

The animal protocols conformed to the Guide for the

Care and Use of Laboratory Animals published by the US

National Institutes of Health (NIH publication No. 85-23,

revised 1996) and was approved by the Research Commis-

sion on Ethics of Hospital Vall d’Hebron.

2.1. Cardiomyocyte and platelet isolation

Cardiomyocytes were obtained from adult rat hearts as

previously described [17]. At the end of the procedure,

culture dishes contained >85% of quiescent rod-shaped

cells. Rat washed platelets were isolated from venous blood

collected with sodium citrate [18].

2.2. Intracellular cGMP synthesis

After treatment, cells were stimulated for 1 minute with

SNAP 100 AM (unless otherwise indicated) in the presence

of 1 mM 3-isobutyl-1-methylxanthine (IBMX) as inhibitor

of cGMP degradation. cGMP was quantified by radio-

immunoassay using acetylated [3H]cGMP [17].

2.3. Membrane and cytosolic fractions

Following incubation with the different agents, cardio-

myocytes were homogenized with a straight wall grinder

(Radnoti Glass Technology) in ice-cold buffer A [in mM:

Tris.HCl 50 (pH 7.4), sucrose 250, EDTA 0.1, dithiotreitol

1, plus protease inhibitors [17], the protein kinase inhibitor

staurosporin 1�10�3, and the Ser/Thr protein phosphatase

inhibitors okadaic acid 1�10�3 and cypermethrin

5�10�4]. After clearing the homogenates by centrifugation

at 1000 �g for 15 min, the particulate and cytosolic

fractions were obtained by centrifugation (100,000 �g for 1

h). Membrane fractions were homogenized with buffer A

plus 10% glycerol. Triton X-100 solubility of particulate

GCNO was assessed as described [19]. Treated platelets were

centrifuged, resuspended in buffer A, frozen in liquid

nitrogen, thawed slowly on ice and then homogenized with

a straight wall grinder and ultracentrifuged as mentioned for

cardiomyocytes.

2.4. Guanylyl cyclase activity

Guanylyl cyclase activity was determined [17] by

measuring cGMP (radioimmunoassay) synthesized after

incubating the particulate and cytosolic extracts with no-

additions (basal) or 100 AM SNAP in assay buffer [final

concentrations (in mM): Tris.HCl 50 (pH 7.4), EGTA 1,

dithiotreitol 1, GTP 2.5, MgCl2 5, phosphocreatine 10,

IBMX 1, plus creatine kinase (50 U/ml)] at 37 -C for 30

min. GCNO activity, calculated by substracting basal

activity, was linear with respect to time in the assay period.

To analyze pH dependence of GCNO activity, Tris.HCl was

substituted by 30 mM PIPES (pH 6.0–6.8) or 30 mM

HEPES (pH 6.8–8.0). Free Ca2+ concentrations were

calculated by a modification of the RECIPC program (S.

Roberston, University of Cincinnati, 1981).

2.5. Western-blotting

Proteins were separated by electrophoresis on a 10% SDS

gel and transferred onto nitrocellulose membrane (Hybond-

ECL, Amersham). Membranes were incubated with rabbit

polyclonal antibodies to h1 (aminoacids 605–619; used at 1 /

2000 dilution; Sigma) and a1 (aminoacids 673–690; 1 /

20,000; Sigma) subunits of soluble guanylyl cyclase. A goat

anti-rabbit IgG horseradish peroxidase conjugated (1 /

50,000; Pierce) was used as secondary antibody. Specificity

of the immunostaining was assessed by displacing the

corresponding bands by incubating in the presence of their

respective immunization peptides (synthesized by Sigma

Genosys). Quantitative chemiluminiscence detection was

performed with SuperSignal West Dura Extended Substrate

L. Agullo et al. / Cardiovascular Research 68 (2005) 65–74 67

Dow

nloaded

(Pierce) and a 16-bit cooled CCD camara system (LAS-

3000, Fujifilm). Equal loading of the different samples was

confirmed by Ponceau Red staining.

2.6. Modulation of the membrane association of GCNO

The effect of increased cytosolic Ca2+ concentration on

the association of GCNO to the membrane fraction was

investigated by incubating cells 1 min with A23187 or 5

min with thapsigargin. Then, cells were: a) stimulated with

SNAP for determining NO-dependent cGMP synthesis, and

b) homogenized for determining content of h1 subunit and

GCNO activity in cytosolic and membrane fractions.

The potential role of changes in GCNO phosphorylation in

this effect was investigated by analyzing its modulation by

the protein kinase inhibitors staurosporin, Go-6976, H-89,

and KN-93. These drugs were added to the incubation media

4 min before stimulation with A23187, and maintained

during the time of incubation with the ionophore.

A

β1

Cyt. Memb. Cyt.

EXP 1 EX

β1

α1

Cytosol Membrane

IB: anti-β1

IB: anti-α1

3 µg

0.6

µg

15 µ

g

100µ

g

3 µg

0.6

µg

15 µ

g

Plq PlqCM C

Plq PlqCM C

B

Fig. 1. Presence of NO-sensitive guanylyl cyclase (GCNO) in the membrane fractio

activity. Immunoblot analysis (panel A) of the h1-subunit in the cytosol and memb

immunoblot analysis of the a1 and h1 subunits of GCNO in cytosolic and particula

platelets. An arrow points to the a1 band in cardiomyocytes to avoid confusion with

measurements (panel C) was confirmed by plotting GCNO activity in each fraction

cytosolic fraction, ˝ membrane fraction). Specific activity (panel C) was calcula

arbitrary units. *P <0.05 respect to the specific activity in the cytosolic fraction.

2.7. Detection of GCNO phosphorylation in intact cells

Cardiomyocytes or platelets were incubated for 3 h in

medium containing 0.1 mCi/ml [32P]ortophosphate (Amer-

sham), washed with Fcold_ medium, incubated for 5 min

with or without 1 AM staurosporin and stimulated for 1

min with 10 AM A23187 or with no drugs. At the end of

the stimulation period, cells were homogenized as

described, but including a phosphatase inhibitor cocktail

(Sigma) in the homogeneization medium. Homogenates

were fractionated by centrifugation at 100,000 �g, and

the h1 subunit of the cytosolic and particulate fractions

immunoprecipitated by incubation with Protein G-agarose

beads (Amersham Biosciences) previously bound to 15 Agof anti-h1 antibody. Phosphorylation of the h1 subunit in

the immunoprecipitates was assessed by Western-blotting

and phosphor screen (Fuji Photo Film Co.) autoradiog-

raphy with a red laser scanner (Typhoon 9400, Amersham

Biosciences). Analysis of a1 phosphorylation was not

Memb. Cyt. Memb.

P 2 EXP 3

100µ

g

O.D. of platelet β1 (a.u.)0 1 2

Plq

GC

NO a

ctiv

ity

(pm

ols

cGM

P/3

0 m

in)

0

2

4

6

Sp

ecif

ic a

ctiv

ity

(a.u

.)

1

2

*

CM Plq

CytosolMembrane

CytosolMembrane

M

M

C

n of cardiomyocytes (CM) and platelets (Plq), and estimation of its specific

rane fractions of 3 representative cardiomyocyte cultures. Panel B shows the

te fractions obtained from 100 Ag of cardiomyocytes or from 0.6–15 Ag of

the wider unspecific band located just above. Linearity of the densitometric

versus the optical density of the h1 band for the different Ag of sample (gted by dividing GCNO activity by the densitometry of h1 and expressed in

by guest on June 1, 2013http://cardiovascres.oxfordjournals.org/

from

L. Agullo et al. / Cardiovascular Research 68 (2005) 65–7468

Dow

nloa

possible since the antibody used against this subunit did

not significantly immunoprecipitate the protein.

2.8. Intracellular Ca2+

Changes induced by the Ca2+ ionophore A23187 and by

thapsigargin were monitored by ratio-fluorescence imaging

in cardiomyocytes loaded with fura 2-acetoxymethyl ester

(Molecular Probes) as previously described [20].

2.9. Data analysis and statistics

Differences between groups were evaluated by means of

paired Student’s t test when appropriated or one-way

analysis of the variance. Individual comparisons between

groups were performed using the Student–Newman Keuls

test. Values are expressed as meanTSEM. Nonlinear fitting

was performed using SigmaPlot (SPSS Inc.).

http://cardiovascres.oxfordjournalsded from

3. Results

3.1. Membrane association of GCNO in cardiomyocytes and

platelets

Cytosolic and particulate fractions (100,000 �g) of

adult rat cardiomyocytes were isolated and blotted.

Staining with an antibody against the h1 subunit of the

GCNO, demonstrated that 28T5% of h1 (65.1T0.2 kDa),

n =6, was in the membrane fraction (Fig. 1A). A similar

β1

Cyt. Memb. Cyt. Memb.

+ peptide β1

S

β1

A IB: anti-β1

α1+ peptide α1

B IB: anti-α1

C IB: anti-β1

Plq

Fig. 2. h1 (panel A) and a1 (panel B) immunoreactivity of 65.1 kDa (for the himmunization peptides in cytosolic and particulate fractions of platelets and cardio

kDa in the particulate fraction) were also displaced by peptides. Membrane associa

in a KCl containing buffer, but was completely disrupted by the presence of dete

proportion of the a1 subunit (75.7T2.9 kDa) seemed to be

associated to the particulate fraction, but the limited

sensitivity of the antibody against this subunit in cardio-

myocytes precluded precise analysis (Fig. 1B). Immuniza-

tion peptides displaced h1 and a1 bands both in platelets

and cardiomyocytes (Fig. 2A and B). However, as shown

in Fig. 2, an additional band in the cytosolic fraction (of

110–115 kDa) and in the membrane fraction (of 85–90

kDa) of cardiomyocytes were also displaced. Since the

identity of these other bands is unknown, they were not

used for the quantification of the h1 content. Throughout

the rest of the study only the antibody against the h1-

subunit was used for blot analysis.

Measurement of GCNO activity showed that 33T4% of

the activity was associated to the particulate fraction

(5.1T2.2 and 2.5T1.1 pmol/mg protein x min in the

cytosolic and membrane fractions, respectively, p =6). For

comparison with a cell model in which the association of

GCNO to the membrane had been previously described,

[15,21,22] cytosolic and membrane fractions were

obtained from rat platelets (Fig. 1B). According to

densitometric analysis of blots, platelets contained 100–

150 times more GCNO than cardiomyocytes. Specific

activity of membrane-associated GCNO, calculated by

dividing GCNO activity by the densitometry of h1 in this

fraction (in arbitrary units), was similar in platelets and

cardiomyocytes (Fig. 1C).

To rule-out an unspecific presence of GCNO in the

membrane fraction, cardiomyocyte homogenates were

extensively diluted (15 times) in a homogeneization

Cyt. Memb. Cyt. Memb.

+ peptide β1

Membrane

+KCltandard +TritonX-100

+ peptide α1

CM

1 band) and 75.7 kDa (for the a1 band) were blocked by their respective

myocytes. Other bands (110–115 kDa in the cytosolic fraction and 85–90

tion of GCNO in cardiomyocytes was not altered by diluting the homogenate

rgent (panel C).

by guest on June 1, 2013.org/

Log [GTP], mMC -4 -3

GC

NO a

ctiv

ity

(% G

TP

3 m

M)

0

50

100 A

CytosolMembrane

Log [SNAP], mMC -5 -4 -3

GC

NO a

ctiv

ity

(% S

NA

P 1

mM

)

0

50

100 B

P<0.01

P<0.05

Fig. 3. GCNO activity in the soluble (>) and membrane (?) fractions of

cardiomyocytes assayed at different concentrations of GTP (panel A) or

SNAP (panel B). P values refer to the differences in the EC50. C, GCNO

activity in the absence of GTP or SNAP for each fraction.

Log [free Ca2+], mMC -8 -7- 6 -5- 4- 3

GC

NO a

ctiv

ity

(% C

)0

50

100

B CytosolMembrane

pH of assay buffer6.0 6.5 7.0 7.5 8.0

GC

NO a

ctiv

ity

(% p

H 7

.4)

0

40

80

A

Fig. 4. Cytosolic (>) and membrane (?) associated GCNO activity assayed

at different pH values (panel A) or Ca2+ concentrations (panel B). C, GCNO

activity when no Ca2+ was added to the assay medium for each fraction.

L. Agullo et al. / Cardiovascular Research 68 (2005) 65–74 69

by guest on June 1, 2013http://cardiovascres.oxfordjournals.org/

Dow

nloaded from

medium containing 150 mM of KCl instead of sucrose, and

were incubated in agitation at 4 -C for 30 min before the

100,000 �g centrifugation. These manoeuvres had no effect

on the observed membrane association of the h1 subunit

(Fig. 2C). To examine the possible presence of GCNO in

glycolipid-rich domains, part of the homogenates was

incubated in a sucrose containing homogenization medium

(standard medium) with 1% Triton X-100 (30 min at 4 -C in

agitation). After this incubation, no significant immunos-

Table 1

Biochemical parameters of GCNO activity in soluble and particulate fractions

Cytosola

EC50 (IC50) AMb Hill coefficent

GTP 588.5T56.4 2.30T0.80

SNAP 160.2T11.5 1.28T0.42Ca2+(high affinity) 0.032T0.015 1.62T0.13

Ca2+(low affinity) 42.5T12.6 1.73T0.02a Nonlinear regression analyses were performed fitting each individual experime

or two-component model, for Ca2+).b Values are means of the effective concentrationsTSEM of the compound yieldi

experiments.c,d Significant differences ( P <0.05, c; P <0.01, d) between values in the particulat

test.

taining for the h1 antibody persisted in the particulate

fraction (Fig. 2C).

3.2. Biochemical differences between cytosolic and

particulate GCNO activity

Dependence of GCNO activity on GTP-, SNAP-, H+-and

Ca2+-concentrations was analyzed for GCNO associated to

the cytosolic and membrane fractions. Significant, although

Membrane n

EC50 (IC50) AM Hill coefficent

296.4T67.4c 1.90T0.50 6

55.2T9.4d 1.15T0.31 6

0.079T0.057 1.22T0.29 5

49.7T12.1 1.75T0.03 5

nt to the Hill Equation (one-component model, for GTP, and SNAP curves,

ng 50% of the response (stimulatory, EC50, or inhibitory, IC50) in n different

e and the cytosolic fractions as determined by a two tailed, paired Student’s t

Time (min)0 4

Fu

ra-2

rat

io (

340/

380)

1.0

1.2

1.4

1.6

Time (min)0 5 10 15

Fu

ra-2

rat

io (

340/

380)

1.0

1.2

1.4

1.6

A

B

2

Fig. 5. Changes in the intracellular Ca2+ concentration after the addition of

A23187 (10 AM, panel A) and thapsigargin (0.1 AM, panel B) to adult

cardiomyocytes in culture. Representative recordings from three individual

experiments (of a total n =5) are shown for both drugs. Maximal

fluorescence ratios reached after A23187 and thapsigargin treatments

correspond to intracellular Ca2+ concentration of 0.4–2.5 AM and 200–350

nM, respectively.

B

A

C

cGM

P (

% C

t)

50

100

150*

A23 (15 min)

A23 (5 min)

A23 (1 min)

Thap (5 min)

Ct

Cytosol Membrane

Ct A23 Ct A23

β1

Ct Thap Ct Thap

β1

GC

NO a

ctiv

ity

(% to

tal i

n C

t)

50

100

150

*

*

ThapCt A23Ct

Membrane

Cytosol

Fig. 6. Effect of increasing intracellular Ca2+ concentration on NO-

dependent cGMP synthesis (panel A), and on h1 immunoreactivity (panel

B) and GCNO activity (panel C) measured in cytosolic and membrane

fractions. After treating cardiomyocytes with no additions (Ct), thapsigargin

(0.1 AM for 5 min) or A23187 (10 AM for 1 min), cGMP synthesis was

measured after SNAP stimulation (panel A), or cultures were homogenized

and cytosolic and membrane fractions assayed for h1 immunoreactivity

(panel B) or GCNO activity (panel C). GCNO activity was expressed as

percentage of total activity in controls (cytosolic plus membrane-associated

GCNO). In panel B, a representative blot of n =5 is shown. The amount of

membrane-associated h1 in cardiomyocytes treated with A23187 increased

by 133% in this experiment (mean of 161T20%, respect to control, n =5;

P <0.05). *P <0.05 respect to control cardiomyocytes (for GCNO, respect to

the control value in the same cell fraction).

L. Agullo et al. / Cardiovascular Research 68 (2005) 65–7470

by guest on June 1, 2013http://cardiovascres.oxfordjournals.org/

Dow

nloaded from

small, differences were found for the EC50 values for GTP

(589 and 296 AM for the cytosolic and particulate activity,

respectively; P <0.05; Fig. 3A and Table 1) and for the NO

donor SNAP (160 and 55 AM, respectively; P <0.01; Fig.

3B and Table 1). In both cases, the concentration–response

curves were biphasic, and the last value was excluded in the

non-linear fitting. However, no differences were observed

regarding the dependence on pH (Fig. 4A) or Ca2+ (Fig.

4B). GCNO activity was maximal at pH 7.4 for the

particulate and cytosolic fractions (as previously described

for the cytosolic enzyme), and the decrease in activity

observed at basic or acidic pHs was identical for the two

fractions. Ca2+ exerted a profound inhibitory effect in both

cardiomyocyte fractions (no remanent GCNO activity was

observed at 1 mM Ca2+), that was best suited to a two-

component model (IC50 values calculated for the high and

low affinity effects are shown in Table 1).

3.3. Effect of increasing cytosolic Ca2+ concentration on

membrane-associated GCNO

Since Ca2+ has been proposed to be a regulating factor of

the GCNO association to membrane in platelets [15], the

effects of the ionophore A23187 (10 AM) and thapsigargin

(0.1 AM) on this association were analyzed. Both drugs

increased cytosolic Ca2+ concentration, although according

to clearly distinct patterns (Fig. 5). Addition of A23187

induced a rapid, marked and transient increase in the

intracellular Ca2+, while thapsigargin induced a small, but

sustained increase. The effects of incubating cardiomyo-

cytes with A23187 (for 1 min) or thapsigargin (for 5 min) on

NO-dependent cell synthesis of cGMP, distribution of the h1

subunit between the particulate and the cytosolic fraction,

and GCNO activity in both fractions, were analyzed.

Thapsigargin did not modify cGMP synthesis induced by

0.1 mM SNAP (Fig. 6A) nor the proportion of h1 associated

A

C

B

A23Ct A23+Stau

A23Ct A23+Stau

Cytosol Membrane

β1

cGM

P (

% C

t)

50

100

150

200*

*#

Stau + A23

A23 (1 min)

Stau (5 min)

Ct

GC

NO a

ctiv

ity

(% to

tal i

n C

t)

50

100

150

200

#

*

Membrane

Cytosol*

A23Ct A23+StauStau

Fig. 7. Effect of staurosporin on the effect of A23187 on NO-dependent

cGMP synthesis (panel A), and on h1 immunoreactivity (panel B) and

GCNO activity (panel C) measured in cytosolic and membrane fractions.

After treating cardiomyocytes without (Ct) or with staurosporin (Stau: 1 AMfor 5 min) and then incubated in the absence (Ct) or presence of A23187

(A23: 10 AM for 1 min), cGMP synthesis was measured after SNAP

stimulation (panel A), or cultures were homogenized and cytosolic and

membrane fractions assayed for h1 immunoreactivity (panel B) or GCNO

activity (panel C). GCNO activity was expressed as percentage of total

activity in controls (cytosolic plus membrane-associated GCNO). The

amount of h1 in the membrane fractions of cardiomyocytes treated with

staurosporin plus A23187 (panel B) increased by 132% in this experiment

respect to those cells receiving only A23187 (mean of 138T3%, n =3;

P <0.05). *P <0.05 respect to control cardiomyocytes (same cell fraction, if

applied). #P <0.05 respect to the effect of A23187.

Table 2

Effect of protein kinase inhibitors on A23187 potentiation of GCNO activity

in soluble and particulate fractions

Cytosola Membrane

Go-6975 1 AM (Ca2+-dependent PKC isozymes) 193T63b 88T26

H-89 10 AM (PKA and PKG) 23T2c 135T40KN-93 30 AM (CaMK) 75T18 266T50d

a Values are means of the effect of the different inhibitors usedTSEMexpressed as percentage of the effect of A23187 on SNAP-activated GCNO

in each individual experiment (n =3) and in the same fraction.b Inhibitors had no significant effect when added alone.

c,d Significantly different ( P <0.05) from the NO-dependent response in

the presence of A23187 alone in the same fraction as determined by

Student’s t test (two tailed).

β1

β1

A

B

Cyt. Memb. Cyt. Memb. Cyt. Memb.

No additions A23 + Stau A23 + Okad

Fig. 8. Activation with A23187 in the presence of either staurosporin or

okadaic acid leads to dephosphorylation of the h1 subunit in intact platelets.

Platelets were labelled with [32P]ortophosphate, stimulated with drugs,

homogenized, and the h1 subunit immunoprecipitated from the cytosolic

and membrane fractions obtained. 32P activity associated to the h1 subunit

(panel A) or h1 immunoreactivity (panel B) of the immunoprecipitated

platelet extracts are shown.

L. Agullo et al. / Cardiovascular Research 68 (2005) 65–74 71

by guest on June 1, 2013http://cardiovascres.oxfordjournals.org/

Dow

nloaded from

to membrane after homogeneization (Fig. 6B), but caused a

significant decrease of GCNO activity in the cytosolic

fraction (measured in the presence of Ca2+ chelating agents;

45T11% of the activity in non-treated-cells, n =6; P <0.05)

without apparent changes in the activity associated to the

particulate fraction (Fig. 6C).

On the other hand, in cardiomyocytes treated with

A23187, cGMP synthesis in response to SNAPwas enhanced

(152T17%, n=6; P <0.05; Fig. 6A), a slight increase in h1

associated to the membrane was found (161T20%, n =5;

P <0.05; Fig. 6B), and particulate GCNO activity was

increased more than fourfold (P <0.05) while cytosolic

GCNO activity did not increase significantly (Fig. 6C).

3.4. Role of changes in phosphorylation status as mediators

of the effects of increased cytosolic Ca2+ concentration

In cells treated with A23187, staurosporin at 1 AM (a

concentration that inhibits Ca2+/calmodulin-dependent

protein kinase or CaMK, myosin light chain kinase,

PKC, PKA and PKG), potentiated the cGMP response to

SNAP (165T21% of the ionophore effect, n =3; P <0.05;

Fig. 7A), and induced an increase in both h1 immunos-

taining (138T3%, n =3; P <0.05; Fig. 7B) and GCNO

activity (165T21%, n =3; P <0.05; Fig. 7C) in the

particulate fraction, without significant effects on cyto-

solic GCNO activity. In cells not treated with A23187,

staurosporin did not significantly increase h1 immunos-

taining (results not shown) or GCNO activity in the

membrane fraction (Fig. 7C). Go-6976, an inhibitor of

Ca2+-dependent PKC isozymes, at 1 AM had no

significant effect on GCNO activation by A23187 on both

cytosolic and membrane extracts (Table 2), and H-89, an

inhibitor of PKA and PKG, at 10 AM inhibited the effect

of A23187 on the cytosolic GCNO. Only KN-93, a

selective CaMK inhibitor, at 30 AM had a potentiating

L. Agullo et al. / Cardiovascular Research 68 (2005) 65–7472

http://cardiovD

ownloaded from

effect on particulate GCNO activity similar to that of

staurosporin (Table 2). On the other hand, the phospha-

tase inhibitors cypermethrin at 0.05 AM (selective

inhibitor of calcineurin) or okadaic acid at 1 AM(inhibitor of PP1 and PP2A) did not block the potentiat-

ing effect of A23187 on membrane GCNO activity; in

fact, okadaic acid enhanced it (290 T15% of the

ionophore effect, n =3; P <0.05).

We were not able to detect in vivo 32P-labelling of GCNO

in cardiomyocytes neither under control nor after stauro-

sporin plus A23187 treatment. In platelets, incubation with

the ionophore A23187 in the presence of staurosporin

decreased phosphorylation of the h1 subunit compared to

non-treated cells (to about 25% of the initial value; Fig. 8).

Stimulation with A23187 in platelets preincubated with

okadaic acid had a similar effect (Fig. 8).

3.5. cGMP synthesis versus GCNO activity in cells fractions

The effects on cGMP synthesis of the different treatments

assayed thorough the present manuscript were significantly

correlated with their effects on membrane-associated GCNO

activity, but not with the effects on GCNO activity in the

cytosolic fraction (P <0.001; Fig. 9).

GCNO activity(% over basal)750

cGMP(% over basal)

100

A

B

750

100

GCNO activity(% over basal)

P=0.1223

P=0.0006

Fig. 9. The effects of treatments on cGMP synthesis in intact cells were not

significantly correlated with the effects on GCNO activity in the cytosolic

fraction (panel A), but were closely correlated with the effects on GCNO in

the membrane fraction (panel B). Data of cGMP synthesis shown in

previous figures (Figs. 6 and 7, A panels) were plotted versus GCNO

activity in particulate and cytosolic fractions of cardiomyocytes for each

treatment (Figs. 6 and 7, C panels). The different measurements were

expressed as percentage of their value in control cells.

by guest on June 1, 2013ascres.oxfordjournals.org/

4. Discussion

This study provides the first direct evidence supporting

the specific association of GCNO to the particulate fraction

of cardiomyocytes. Membrane-associated GCNO activity

showed similar concentration-dependence to GTP, NO

donors, Ca2+ and pH than GCNO in the cytosolic location.

As previously shown for the cytosolic GCNO, increasing

Ca2+ concentration in the assay medium inhibited partic-

ulate GCNO. However, treating cardiomyocytes with the

Ca2+ ionophore A23187 promoted translocation of GCNO to

the membrane fraction, increased cGMP synthesis in

response to stimulation of the cells with NO, and enhanced

GCNO activity in this fraction when assayed in vitro. These

effects were enhanced by staurosporin and the CaMK

inhibitor KN-93. These results suggest that cytosolic Ca2+

concentration regulates the intracellular distribution of

GCNO, and differentially regulates the activity associated

to the membrane and the cytosolic fraction, probably

through changes in its phosphorylation status. The obser-

vation that the effects of several treatments on NO-induced

cGMP synthesis in cardiomyocytes closely correlate with

their effects on GCNO activity in the particulate fraction, but

not with their effects on cytosolic GCNO, suggests that

membrane-associated GCNO largely determines NO-induced

cGMP synthesis.

Although different data indicate that changes in GCNO

activity greatly affect the function of the NO/cGMP path-

way in several tissues [17,23–30], little information is

available on the regulation of this enzyme. A previous study

[15], demonstrated the presence of a1 and h1 subunits in the

membrane fraction of rat myocardium, platelets and

endothelial cells. This study described enzyme translocation

to the membrane in activated platelets [15], and by means of

in vitro experiments of GCNO association/dissociation to

cell membranes suggested an important role of Ca2+.

However, although GCNO in the particulate fraction of rat

heart was responsive to NO, it was unclear whether GCNO

associated to cell membrane was important in the response

of the intact cell to NO.

We found that approximately 30% of the GCNO

immunostaining for a1 and h1 and a similar proportion of

GCNO activity were associated to the particulate fraction of

cardiomyocytes. h1 association to membranes resisted

extensive washing in a KCl buffer, that mimics the high

intracellular potassium concentration. However, no apparent

immunostaining persisted in the particulate fraction after

washing in a Triton X-100 medium. This is similar to what

was described by Zabel et al. [15] for the platelet particulate

GCNO. Given the scarce immunostaining found in cardio-

myocytes for GCNO subunits, we did not make any attempt

to further analyze their subcellular location.

Membrane-associated GCNO and cytosolic GCNO

showed similar biochemical characteristics. pH-dependence

of GCNO was identical for the two locations and similar to

that described previously [17,31]. Membrane GCNO had a

L. Agullo et al. / Cardiovascular Research 68 (2005) 65–74 73

by guest on June 1, 2013http://cardiovascres.oxfordjournals.org/

Dow

nloaded from

significantly lower EC50 value for GTP than the cytosolic,

but the difference was small and its physiological relevance

is doubtful. The difference in sensitivity to the NO donor

SNAP was more clear and similar to that described

previously in heart extracts [15]. However, a very recent

study [21] has suggested that contaminating myoglobin in

cytosolic extracts from heart tissue neutralizes a significant

part of the NO released by NO donors. According to this,

our results may underestimate GCNO activity in the cytosol.

Specific activity in cytosol could be thus higher than in the

particulate fraction, as observed in platelets.

As previously described [32,33], in the present study

Ca2+ inhibited cytosolic GCNO. We found that membrane-

bound GCNO was also inhibited by Ca2+ and that in the two

cell fractions GCNO showed a biphasic pattern very similar

to that recently described in GCNO purified from bovine

lung [34]. IC50 values calculated for the low affinity sites for

Ca2+ were the same for cytosolic and particulate GCNO and

comparable with those previously reported [34], and IC50

values for the high affinity sites of GCNO were also similar

in both cell fractions. But, besides this inhibitory effect of

Ca2+ when added to the assay medium (mediated by a direct

binding of Ca2+ to GCNO), we observed in this study effects

of increasing cytosolic Ca2+ in intact cells that had not been

previously described. These effects are persistent, and can

be detected after cell homogeneization in the presence of

Ca2+ quelating agents. Importantly, the different agents used

to increased cytosolic Ca2+ concentration have different

effects, suggesting distinct roles for different levels of

physiological concentrations of intracellular Ca2+ or for

different subcellular location of these increases. The

moderate and slow Ca2+ increase evoked by thapsigargin

did not alter GCNO activity associated to the particulate

fraction, and inhibited GCNO in the cytosolic fraction, while

the more marked Ca2+ increase elicited by the Ca2+

ionophore A23187 increased several times GCNO activity

in the particulate fraction without significant effects on

cytosolic GCNO. In parallel with the increase in activity in

the particulate fraction, A231287 increased the amount of

h1 subunit associated to this fraction. This is similar to the

translocation previously observed in activated platelets [15].

However, the change in quantity of h1 associated to the

particulate fraction was much smaller than the change

observed in GCNO activity. A critical point is that for the

different conditions assayed, cytosolic and membrane-

associated GCNO were found to respond differentially to

increased cytosolic Ca2+.

A potential explanation for the increase in specific activity

of membrane-associated GCNO induced byA23187 is that the

increase in cytosolic Ca2+ concentration induced by the drug

causes a modification in the phosphorylation status of the

enzyme. Few studies have analyzed GCNO regulation by

phosphorylation with conflicting results. Some studies

suggested that GCNO phosphorylation increases its activity.

Both in vitro phosphorylation by PKC and PKA [11] and in

vivo phosphorylation by PKA [35] have been described to

increase GCNO activity, while dephosphorylation of the h1

subunit has been associated to a decrease in GCNO activity

[23]. Avery recent study has described, in contrast, a decrease

in GCNO activity associated to an increase in GCNO

phosphorylation in response to PKG activation [12]. In the

present study, the protein kinase inhibitors staurosporin and

KN-93 activated membrane-associated GCNO sinergically

with A23187 suggesting that in cardiomyocytes intracellular

Ca2+ potentiates GCNO activity probably by promoting GCNO

dephosphorylation. Although, direct evidence of GCNO

dephosphorylation in response to A23187 could not be

obtained, the evidence obtained in platelets supports this

hypothesis. However, the fact that the protein phosphatase

inhibitor okadaic acid also enhanced the response to A23187

suggest that regulation of GCNO activity by phosphorylation

may be complex. A protein phosphatase activated by

phosphorylation, as found in chromaffin cells [23], or an

additional regulatory site (in the a1 subunit or in some of

regulatory proteins recently described: as Hsp90 [16], Hsp70

[36], or CCTD [37]) that would increase GCNO activity after

phosphorylation could explain the results. To sum up, our

observations suggest that the increase in cytosolic Ca2+

induced by A23187 has two opposite effects on NO-mediated

cGMP synthesis: a direct inhibitory effect and an indirect

stimulatory effect mediated by h1 dephosphorylation result-

ing in membrane-associated GCNO activation.

The present study provides information that strongly

suggests an important functional role of GCNO localized in

the membrane fraction of cardiomyocytes. Our results show

that a profound inhibition of cytosolic GCNO do not

significantly affect the cell response to SNAP, while

activation of the particulate fraction markedly increases it.

As shown in Fig. 9, cGMP synthesis in the entire cell

correlates well with changes in membrane GCNO activity,

but not with changes in cytosolic GCNO activity. This is in

agreement with recent results in bovine aortic endothelial

cells, indicating a decrease cell-response to NO stimulation

when the formation of a membrane-associated protein

complex between eNOS, HSP90 and GCNO is inhibited

[16].

The association of a fraction of GCNO to cell membrane

in cardiomyocytes, the important role of membrane-asso-

ciated GCNO on the cell response to NO, and the fact that

the regulation of its activity differs from that of the cytosolic

enzyme, may be of great relevance for the better under-

standing of pathophysiological conditions in which the NO/

cGMP-pathway is compromised, and in the design of new

therapies for these conditions.

Acknowledgments

Supported by Fondo Investigaciones Sanitarias (Grant

01/3008 and RECAVA), and Comision Interministerial de

Ciencia y Tecnologıa (Grant SAF2002/0759). We thank

Angeles Rojas for her excellent technical work.

L. Agullo et al. / Cardiovascular Research 68 (2005) 65–7474

by guest on June 1, 2013http://cardiovascres.oxfordjournals.org/

Dow

nloaded from

References

[1] Casadei B, Sears CE. Nitric-oxide-mediated regulation of cardiac

contractility and stretch responses. Prog Biophys Mol Biol

2003;82:67–80.

[2] Padilla F, Garcia-Dorado D, Agullo L, Barrabes JA, Inserte J,

Escalona N, et al. Intravenous administration of the natriuretic peptide

urodilatin at low doses during coronary reperfusion limits infarct size

in anesthetized pigs. Cardiovasc Res 2001;51:592–600.

[3] D’Souza SP, Yellon DM, Martin C, Schulz R, Heusch G, Onody A,

et al. B-type natriuretic peptide limits infarct size in rat isolated hearts

via KATP channel opening. Am J Physiol Heart Circ Physiol

2003;284:1592–600.

[4] Ockaili R, Salloum F, Hawkins J, Kukreja RC. Sildenafil (Viagra)

induces powerful cardioprotective effect via opening of mitochondrial

KATP channels in rabbits. Am J Physiol Heart Circ Physiol

2002;283:1263–9.

[5] Ahluwalia A, Foster P, Scotland RS, McLean PG, Mathur A, Perretti

M, et al. Antiinflammatory activity of soluble guanylate cyclase:

cGMP-dependent down-regulation of P-selectin expression and

leukocyte recruitment. Proc Natl Acad Sci U S A 2004;101:1386–91.

[6] Izumi T, Saito Y, Kishimoto I, Harada M, Kuwahara K, Hamanaka I,

et al. Blockade of the natriuretic peptide receptor guanylyl cyclase-A

inhibits NF-nB activation and alleviates myocardial ischemia/reperfu-

sion injury. J Clin Invest 2001;108:203–13.

[7] Hempel AM, Friedrich M, Schluter KD, Forssmann WG, Kuhn M,

Piper HM. ANP protects against reoxygenation-induced hypercon-

tracture in adult cardiomyocytes. Am J Physiol 1997;273:244–9.

[8] Kodani E, Xuan YT, Takano H, Shinmura K, Tang XL, Bolli R. Role

of cyclic guanosine monophosphate in late preconditioning in

conscious rabbits. Circulation 2002;105:3046–52.

[9] Bellamy TC, Garthwaite J. Sub-second kinetics of the nitric oxide

receptor, soluble guanylyl cyclase, in intact cerebellar cells. J Biol

Chem 2001;276:4287–92.

[10] Kostic TS, Tomic M, Andric SA, Stojilkovic SS. Calcium-independent

and cAMP-dependent modulation of soluble guanylyl cyclase activity

by G protein-coupled receptors in pituitary cells. J Biol Chem

2002;277:16412–8.

[11] Zwiller J, Revel MO, Malviya AN. Protein kinase C catalyzes

phosphorylation of guanylate cyclase in vitro. J Biol Chem

1985;260:1350–3.

[12] Murthy KS. Modulation of soluble guanylate cyclase activity by

phosphorylation. Neurochem Int 2004;45:845–51.

[13] Russwurm M, Wittau N, Koesling D. Guanylyl cyclase/PSD-95

interaction. Targeting of the nitric oxide-sensitive a2h1 guanylyl

cyclase to synaptic membranes. J Biol Chem 2001;276:44647–52.

[14] Feussner M, Richter H, Baum O, Gossrau R. Association of soluble

guanylate cyclase with the sarcolemma of mammalian skeletal muscle

fibers. Acta Histochem 2001;103:265–77.

[15] Zabel U, Kleinschnitz C, Oh P, Nedvetsky P, Smolenski A, Muller

H, et al. Calcium-dependent membrane association sensitizes

soluble guanylyl cyclase to nitric oxide. Nat Cell Biol 2002;4:

307–11.

[16] Venema RC, Venema VJ, Ju H, Harris MB, Snead C, Jilling T, et al.

Novel complexes of guanylate cyclase with heat shock protein 90 and

nitric oxide synthase. Am J Physiol Heart Circ Physiol

2003;285:669–78.

[17] Agullo L, Garcia-Dorado D, Escalona N, Ruiz-Meana M, Inserte J,

Soler-Soler J. Effect of ischemia on soluble and particulate guanylyl

cyclase-mediated cGMP synthesis in cardiomyocytes. Am J Physiol

Heart Circ Physiol 2003;284:2170–6.

[18] Massaguer A, Engel P, Perez-del-Pulgar S, Bosch J, Pizcueta P.

Production and characterization of monoclonal antibodies against

conserved epitopes of P-selectin. Tissue Antigens 2000;56:117–28.

[19] Garcıa-Cardena G, Oh P, Liu J, Schnitzer JE, Sessa WC. Targeting of

nitric oxide synthase to endothelial cell caveolae via palmitoylation:

implications for nitric oxide signaling. Proc Natl Acad Sci U S A

1996;93:6448–53.

[20] Ruiz-Meana M, Garcia-Dorado D, Lane S, Pina P, Inserte J, Mirabet

M, et al. Persistence of gap junction communication during myocar-

dial ischemia. Am J Physiol Heart Circ Physiol 2001;280:2563–71.

[21] Wykes V, Garthwaite J. Membrane-association and the sensitivity of

guanylyl cyclase-coupled receptors to nitric oxide. Br J Pharmacol

2004;141:1087–90.

[22] Kempfert J, Behrends. Analysis of nitric oxide-sensitive guanylyl

cyclase in human platelets before and after aggregation. Platelets

2003;14:429–35.

[23] Ferrero R, Rodrıguez-Pascual F, Miras-Portugal MT, Torres M. Nitric

oxide-sensitive guanylyl cyclase activity inhibition through cyclic

GMP-dependent dephosphorylation. J Neurochem 2000;75:2029–39.

[24] Baltrons MA, Saadoun S, Agullo L, Garcıa A. Regulation by calcium

of the nitric oxide/cyclic GMP system in cerebellar granule cells and

astroglia in culture. J Neurosci Res 1997;49:333–41.

[25] Agullo L, Garcia-Dorado D, Escalona N, Inserte J, Ruiz-Meana M,

Barrabes JA, et al. Hypoxia and acidosis impair cGMP synthesis in

microvascular coronary endothelial cells. Am J Physiol Heart Circ

Physiol 2002;283:917–25.

[26] Papapetropoulos A, Abou-Mohamed G, Marczin N, Murad F,

Caldwell RW, Catravas JD. Downregulation of nitrovasodilator-

induced cyclic GMP accumulation in cells exposed to endotoxin or

interleukin-1 beta. Br J Pharmacol 1996;118:1359–66.

[27] Kloss S, Bouloumie A, Mulsch A. Aging and chronic hypertension

decrease expression of rat aortic soluble guanylyl cyclase. Hyper-

tension 2000;35:43–7.

[28] Laber U, Kober T, Schmitz V, Schrammel A, Meyer W, Mayer B, et al.

Effect of hypercholesterolemia on expression and function of vascular

soluble guanylyl cyclase. Circulation 2002;105:855–60.

[29] Mulsch A, Oelze M, Kloss S, Mollnau H, Topfer A, Smolenski A,

et al. Effects of in vivo nitroglycerin treatment on activity and

expression of the guanylyl cyclase and cGMP-dependent protein

kinase and their downstream target vasodilator-stimulated phospho-

protein in aorta. Circulation 2001;103:2188–94.

[30] Sinnaeve P, Chiche JD, Nong Z, Varenne O, Van Pelt N, Gillijns H,

et al. Soluble guanylate cyclase a1 and h1 gene transfer increases NO

responsiveness and reduces neointima formation after balloon injury

in rats via antiproliferative and antimigratory effects. Circ Res

2001;88:103–9.

[31] Grabow M, Chakraborty G, Ledeen RW. Characterization of guanylyl

cyclase in purified myelin. Neurochem Res 1996;21:457–62.

[32] Parkinson SJ, Jovanovic A, Jovanovic S, Wagner F, Terzic A,

Waldman SA. Regulation of nitric oxide-responsive recombinant

soluble guanylyl cyclase by calcium. Biochemistry 1999;38:6441–8.

[33] Serfass L, Carr HS, Aschenbrenner LM, Burstyn JN. Calcium ion

downregulates soluble guanylyl cyclase activity: evidence for a two-

metal ion catalytic mechanism. Arch Biochem Biophys 2001;387:

47–56.

[34] Kazerounian S, Pitari GM, Ruiz-Stewart I, Schulz S, Waldman SA.

Nitric oxide activation of soluble guanylyl cyclase reveals high and

low affinity sites that mediate allosteric inhibition by calcium.

Biochemistry 2002;41:3396–404.

[35] Kostic TS, Andric SA, Stojilkovic SS. Receptor-controlled phosphor-

ylation of a1 soluble guanylyl cyclase enhances nitric oxide-dependent

cGMP production in pituitary cells. Mol Endocrinol 2004;18:458–70.

[36] Balashova N, Chang FJ, Lamothe M, Sun Q, Beuve A. Character-

ization of a novel type of endogenous activator of soluble guanylyl

cyclase. J Biol Chem 2005;280:2186–96.

[37] Hanafy KA, Martin E, Murad F. CCTeta, a novel soluble guanylyl

cyclase-interacting protein. J Biol Chem 2004;279:46946–53.