Twinkleトップページ - 東京女子医科大学学術リポジトリ

(1)

( g,T,3kYO, ,W, 2ii;l, ,Med,,8,o,il )

A Research Report Supported by Yayoi Yoshioka Award

ATRIAL NATRIURETIC PEPTIDE LEVELS AND ITS MOLECIULAR

FORM IN HUMAN CARDIAC TESSUE

Kiyoko NARUSE, Michiaki HIROEi), Mitsuhide NARUSE, Makoto NAGATA2),

Akimasa HASHIMOT03), Hitoshi KOYANAGI3), Morie SEKIGUCHI2),

Hiroshi DEMURA, Koushichiro HIROSAWA2)

and Kazuo SHIZUME

Department of Medicine, Institute of Clinical Endocrinology, i)Department of Radiology, 2)Department of Medicine, and 3)Department of Surgery, The Heart Institute of Japan,

Tokyo Women's Medical College

(Received March, 22th 1989)

Abstract

We investigated the levels and molecular form of atrial naturiuretic peptide (ANP) in human atrial tissue obtained frorn six autopsied patients without heart disease and 31 patients during

heart surgery. The levels of atrial ANP and its

three major variants -cr-hANP, B-hANP, and

7-hANP- were determined by a combination of

radioimmunoassay and high-performance gel

per-meation chromatography. The levels of right

atrial ANP were significantly higher in patients with mitral disease (12.2 ± 2.50 ptg/mg protein)

than that in the autopsy group or the patients with aortic disease, or atrial septal defect. These increased ANP levels were mainly accounted for by an increase in B-hANP. Both total ANP and

B-hANP levels in the right atrium correlated with the presence of atrial fibrillation and with right atrial, pulmonary arterial, and pulmonary capil-lary wedge pressures.

These results suggest that human atrium

responds to an imposed hemodynamic overload by

increasing ANP production and by a shift to the B-hANP form. This -dimeric peptide could be used

as a marker indicating adaptation to chronic pathologic conditions.

Introduction

The recent discovery of atrial natriuretic

poly-peptides (ANP) in human atrial tissue spurred

many studies, which have made it clear that the heart is a part of the endocrine system involved in the maintenance of body-fluid homeostasis (for review, see refereneesi}N4)>. ANP acts to correct

disturbances in body fluid volume and blood

pressure through potent natriuretic, diuretic, and vasodilator effects, and also by interacting with other regulatory system such as the renin-angiotensin-aldosterone system5)N8) and vasopres-sin9)lo).

Atrial ANP has been reported to be composed of

at least three major components: 7-human ANP

(7-hANP), a 126-amino-acid peptide; a-human ANP (a-hANP), a 28-amino-acid residue of the

carboxy-terminal of xhANP; and B-hANP, which is an

antiparallel dimer of cr-hANPii)i2). The

compo-nents 7-hANP and cr-hANP have been

charac-terized respectively as a prohormone and a major secretory form. The pathophysiologic significance

of B-hANP, which has a unique structure and a biologic activity different from or-hANP in its dura-tion and magnitudei2), is not known.

The present study was undertaken to

inves-tigate the biochemical changes in ANP production

in overloaded heart and the pathophysiologic significance of the different ANP molecular forms, especially B-hANP. We measured the levels of

total immunoreactive ANP and its moiecular

(2)

autopsy and open heart surgery.

Methods

Tissue collection

Human right and/or left atrial tissues were

obtained at autopsy from six patients with a mean

age of 60 ± 3 years, who died of noncardiac

disease: a 61-year-old man with thyroid and

esophageal cancer, 52-year-old and 70-year-old

men with esophageal cancer, a 53-year-old man

with liver cirrhosis, a 64-year-old man with lung cancer, and a 68-year-old man with a dissecting

aneurysm, respectively. None of them had left

ventricular hypertrophy as judged from electro-cardiograms and/or chest radiographs. All autop-tic specimens were obtained within 3 hours after death.

Surgical specimens were obtained at the time of open heart surgery from 31 patients. The mitral disease group consisted of nine patients: six with

predominant mitral stenosis and three with

predominant mitral regurgitation. These patients ranged in age from 46 to 63 years with a mean age

of 55 ± 2 years. The group with aortic valve disease consisted of eight patients, ranging in age from 34 to 68 years with a mean of 51 ± 5 years. Seven had aortic regurgitation and one had aortic

stenosis. The group with atrial septal defect

(ASD) consisted of seven patients, ranging in age from 19 to 61 years with a mean of 43 ± 6 years. Surgical specimens were also obtained from seven patients with miscellaneous cardiac diseases (left atrial aneurysm, tetralogy of Fallot with atrial

septal defect, congenital pulmaonary stenosis with ventricular septal defect, ventricular septal defect, endocardial cushion defect, tricuspid re-gurgitation, and left atrial myxoma) (Table 1). A small portion (approximately 5 mg) of the

right atrial appendage was incised during

can-nulation for cardiopulmonary bypass. A small amount of the left atrial appendage was also

trimed in some patients to provide access to the mitral valve according to routine operative

pro-cedure and to prevent postoperative thrombus

formation. The tissues obtained were immediately

frozen and stored at -800C until extracted.

Blood samples were also obtained from a

peripheral vein in some of the patients before

cardiac surgery. Blood samples were collected in chilled tubes containing sodium

ethylenediamine-tetraacetate (Na2EDTA) (4 mM) and aprotinin

(500 kallikrein inactivator units/ml). The samples

were separated by centrifugation at 40C, and

plasma was stored at -700C until assayed.

Extraction of ANP

Atrial ANP was extracted in 1 M acetic acid containing 20 mM HCI after boiling for 10 min. Tissues were homogenized with a Polytron

hom-ogenizer (Brinkman Company, Westbury, NY),

centrifuged at 15,OOO g for 60 min at 40C, and the

clear supernatant was stored at -300C until

subjected to radioimmunoassay (RIA) or chro-matography.

Plasma ANP was extracted by the method

described previouslyi3}. In brief, 1 ml of each plasma was first acidified with acetic acid and then applied to a Sep-Pak Cis (Waters Associates, Milford, MA) cartridge. The adsorbed peptide was eluted with 2 ml of a mixture of acetonitrile and O.4% acetic acid (3:2). The eluates were evapo-rated, lyophilized, dissolved in the assay buffer,

and subjected to RIA or to the following

chro-matographic procedure.

High-performance gel permeation

chro-matography (HP-GPC)

Immunoreactive ANP components of a different molecular weight wefe separated and determined

by a combination of RIA and HP-GPC using a

TSK-GEL G2000 SW column (7.5 × 600 mm; Toyo Soda, Tokyo). The column was eluted with 10 mM trifluoroacetic acid containing O.2 M sodium chloride and 30% acetonitrile as a solvent at a flow rate of O.3 ml/min. Under the conditions

em-pioyed, the retention times of the standard

sub-stances used for molecular weight calibration

were as follows: 43.7 ± O.1 niin for bovine serum albumin, 48.2 ± O.1 min for carbonic anhydrase, 55.8 ± O.1 min for cytochrome C, 67.6 ± O.3 min

for synthetic B-hANP (Protein Research

(3)

Table 1 Clinical characteristics and hemodynamic variables

Case Sex Age Disease Rhythm

Mean (mmHg)pressure LAsize. (cm) CI (llminlm2) CTR(%)

NYHA

(yr)

RA

RA ==right atrium ; PCW=pulmonary capillary wedge ; PA=pulmonary artery ; AO=aorta ; LA= left atrium ; CI=cardiac index ; CTR-cardiothoracic ratio ; NYHA=New York Heart Association ; MSr =:mitral stenosis with mild mitral regur-getation ; TR== tricuspid regurregur-getation ; PH --pulmonary hypertension ; MS=mitral stenosis ; Tr=imild tricuspid

regurgeta-tion; MR=mitral regurgetaregurgeta-tion; HOCM=hypertrophic obstructive cardiomyopathy; AR=aortic regurgetaregurgeta-tion; AS=

aortic stenosis ; ASD=atrial septal defect ; TOF=tetralogy of Fallot ; PS=pulmonary stenosis, VSD=ventricular septal

defect; ECD==endocardial cushion defect; Af=:atrial fibrillation; SR=sinus rhythm; LVEDVI=left ventricular

end-diastolic volume index ; EF=ejection fraction ; QpfQs=ratio of pulmonary to systemic blood flow.

(Protein Research Foundation), and 82.4 ± O.1 min

for angiotensin I. The eluates was evaporated,

lyophilized, dissolved in the assay buffer, and subjected to RIA.

Radioimmunoassay of ANP

Immunoreactive ANP was determined by the

RIA method as described previouslyi3) except for

the anti-ANP antiserum. In brief, a-hANP was

radioiodinated by the chloramine-T method, and the iodinated tracer was purified by high-per-formance liquid chromatography on a ptBondapak Cis column (Waters Associates, Milford, MA). The antiserum used in this study was raised in New Zealand white rabbits using or-hANP as antigen'`).

It was used at a final dilution of 1:4 × 10`, yielding a maximum binding of approximately 35% of the

(4)

quantity of i25I-ANP. The crossreactivity with

a-rat ANP and B-hANP was 16% and 90% on a molar

basis, respectively. The crossreactivity with

an-giotensin II, arginine vasopressin, ACTH,

so-matostatin, vasoactive intestinal peptide, and

thyroglobulin was less than O.OOOI%.

The assay buffer in RIA was O.1 M Tris acetate,

pH 7.4, containing O.1% bovine serum albumin

(Amersham, Arlington Heights, IL) and 1 mM

Na2EDTA. The RIA incubation mixture consisted Table 2 Atrial and plasma ANP levels

AtrialANP(#g/mgprotein)

Patients Rightatrium Leftatrium

PlasmaANP

_(pg/ml)

Total 7-hANP S-hANP a-hANP Total 7-hANP fi-hANP

a-hANP

Autopsy 1.5e O.84 1,23 2.82 O.95 2,99 3.54 8.46 17,6 10,0 25.9 9.22 14,8 17.9 2.66 O.89 1.31 O.42 2,13 O.09 O.17 5.87 10.3 4,O 1.91 O.51 O.78 1.94 O.90 28.8 8.15 1.33 1.02 1.06 O.78 O,64 2.27 1,14 O.55 1.10 I,40 0,20 O.32 3,13 1,33 2.48 1.57 1.21 O.45 2.45 1.50 O,71 O.56 0.97 O.31 O.59 O.04 O.08 e.56 1,17 3.53 1,29 O.35 O.59 O.13 O.65 1.89 1.07 O.47 O.38 O.87 O.40 O.52 1.02 0,32 O.20 O.10 1.06 O.34 2.64 O,30 6,6I 14.6 7.48 22.7 8.18 11.5 13.7 1.62 0.25 O.14 O.07 1.32 O,02 O.05 4.63 8.87 O.16 e.4o O.15 O.10 1.64 O.12 23.4 4.54 O.80 0.42 O.19 O,11 O.05 O,68 O.02 O.03 O.Ol O.02 O.07 n,d, n.d. n.d. n.d. n.d. n.d. n,d. n,d. n.d. 0.05 O.02 O,04 O.Ol n.d. 0.01 e.o2 O.11 n,d. e.os O.15 0.01 O,04 O.03 O.04 O,52 e,49 n,d. 0.04 O,02 O.07 O.02 O,14 1.69 1.15 1.50 4.18 3.77 3.18 O.36 9.15 3.23 I,94 10,O 1.83 5.04 9,39 r---"-"--O.I2-2.13 O.72 O.43 O.94 1.81 2.24 1.21 O.28 1.08 O,21 0.29 O.29 O,20 O,35 1.68 O.95 O,43 0.19 1.95 1.23 1,25 o.e3 7,48 2.69 1.34 9.01 1.51 4.2I 6.60 n,d, O.14 O.19 O,13 O.13 O.17 O.03 n.d. n.d. n,d. n.d, n.d. n,d. 0.56 rm

(5)

of 100 pt1 of standard or-hANP, 100 pt1 of diluted antiserum (1:8,OOO), 100 pt1 of i25I-a-hANP (ap-proximately 10,OOO cpm) and 200 ptl of the assay buffer. The mixture was incubated for 24 h at 40C.

Antibody-bound and free tracer peptide were

separated by the addition of anti-rabbit

immu-noglobulin antiserum (1:20), normal rabbit serum

(1:200), and 15% polyethylene glycol in assay

buffer. The sensitivity of the RIA was 2 pg/tube, with 50% displacement at 30 pg/tube.

Protein concentration

The protein cencentration of the extract was measured by the Bradford dye-binding methodi5)

using crystalline bovine serum albumin as a

standard.

Cardiac catheterization

Right-sided and left-sided cardiac catheteriza-tion were performed with a No. 7F thermodilucatheteriza-tion

Swan-Ganz catheter (Edwards Laboratories,

Santa Ana, CA), which was inserted into the right internal jugular vein or the antecubital vein for right-sided catheterization, and the radial or fem-oral artery for left-sided catheterization. All

pressures -mean right atrial, mean pulmonary arterial, mean pulmonary capillary wedge, and mean aortic- were measured in each position.

Other hemodynamic parameters measured

in-cluded heart rate, cardiac index, left ventricular end-diastolic volume index, and ejection fraction (in patients with aortic disease), and left-to-right shunt and the ratio of pulmonary to systemic

blood flow (Qp/Qs ratio) (in patients with ASD).

Cardiac output was determined by thermodilu-tion. In addition, left atrial size was assessed by echocardiography using either the M-mode

echo-cardiographs or a two-dimentional estimate. Statistical analysis

Data are expressed as micrograms per

mil-ligram protein (mean ±SE), and were analyzed by

an unpaired and/or paired t-test. Regression analyses employed the linear and multivariate

least squares methods. Differences between

groups were considered to be significant at

P<O.05.

Results

The clinical characteristics and various hemo-dynamic data of each patient are shown in Table 1 and are in summarized in table 3. Seven of the nine patients with mitral disease also had atriai fibrillation. All except one (for whom data were not available) had left atrial enlargement accord-ing to echocardiography (in our laboratory, a value greater than 3.0 cm is considered enlarged). The right atrial pressure, pulmonary capillary wedge pressure, and/or mean pulmonary arterial pres-sure were elevated in most of the patients with mitral disease. The left ventricular end- diastolic volume index (204.6 ± 39.6 ml/m2 body surface area) was elevated and the ejection fraction (49.9 ± 3.8%) was abnormally low in patients with

aortic valve disease. The left-to-right shunt was 60 ± 6% and the Qp/Qs ratio (2.57 ± O.31) was

significantly elevated compared with normal

values (less than 1.5) in patients with ADS. As shown in Table 3, the pulmonary capillary

wedge pressure and pulmonary arterial pressure were significantly higher in patients with mitral disease than in patients with aortic disease or

ASD. In addition, although the cardiothoracic ratio (CTR) determined from chest radiographs

was abnormally high in all three cardiac disease groups, it was significantly higher in patients with mitral disease than in the other two groups. The atrial levels of total ANP and of its molec-ular variants are shown in Table 2. Right atrial

ANP levels were determined in all 31 patients

examined. As summarized in Table 3, the atrial total ANP level in mitral disease was 12.2 ± 2.50 ptg/mg protein, which was significantly higher

than in the autopsy, aortic disease, and ASD groups. Of the miscellaneous cardiac disease

group, the patient with left atrial aneurysm had the highes･t atrial ANP levels.

Immunoreactive ANP in the atrial tissue

con-sisted of three major components. The first major peak was located at the elution position of cyto-chrome C with a molecular weight of 12.6 kilo-daltons, and was regarded as 7-hANP. The second

(6)

Table 3 Summary of the clinical, hemodynamic data and atrial ANP levels Autopsy Mitraldisease Aorticdisease

ASD

Age(yr) rneanpressureRA

(mmHg)PCW

PA

AO

CI(llmin/m2) CTR(O/.)

RATotalANP

v-hANP

P-hANP

a-hANP

LATotalANP

7-hANP

P-hANP

cr-hANP

n=6

60±3 n==9 55±2 9±2. 23±3t 40±5t 92±6 2.7±O.2 71±5. 12,2±2,50e 1,65±O,29g 9.63±2.31Q O.Ol±O.Ol n=8 5.12±1.37-. O.55±O.19** 4.11±1.15** O.30±O.27

n=8

51±5 4±1 12±2 17±1 89±5 3.4±O,4 59±1 2.65±1.28 O,54±O.14 1.92±1.14 O.02±O.Ol ---r

n=7

43±6 5±2 6±2 17±2 84±4 2.3±O.5 55±3 5,55±3,90 1.20±0,45 3,71±3.29 O.12±O,07

-m--Values are the means±SE. ANP levels are expressed as "gfmg protein, 'p<O.05 1 tp<O,O05 vs corresponding values in aortic disease and ASD. gp<O.05 i ep<O,Ol vs corresponding values in autopsy,

tp<O.05 i "p<O.Ol vs corresponding right atrial values. Abbreviations as in Table 1,

and third major peaks were located at the elution posi･tions of synthetic B-hANP and a-hANP, re-spectively. In the patients studied, atrial tissue always contained 7-hANP or B-hANP as the pre-dominant variant (Fig. IA, B).

The atrial 7-hANP level in the mitral disease group was 1.65 ± O.29 ptg/mg protein, which was significantly higher than in the autopsy and aortic disease groups (Table 3). The atrial B-hANP level in mitral disease was about 12 times higher than in the autopsy group. Atrial 6-hANP levels in the

aortic disease and ASD groups were not signi-ficantly different from those in the autopsy group.

Immunoreactive ANP corresponding to a-hANP

was not detected in most of the patients with mitral disease. In addition, there was no signifi-cant difference in atrial a-hANP levels among the different disease states. On the whole, total ANP levels in the right atrium showed a significant

correlation with B-hANP levles (r=O.992,

P<O.OOI).

ANP levels in the right atrium did not differ from those in left atrium in the autopsy group. In

ISO e n e loo 2g : 50 (A)' gk il

-a

zz

rt q

_{t ==t}

:･ " " 2･

i III

(B)

2:

i:- E. {･ {･S

II l Iii

tooo 7so

B

e

-a va 50o a

z

(

250 O 3o 4o so 6o 7o so go 3o 4o so 6o 7o eo go O Retention time{min> Retention tlrne(min)

Fig. 1 Representative HP-GPC gel filtration patterns of immunoreactive ANP from human atrial tissue (A) y-hANP-predominant type (patient 3 of the autopsy group). (B) B-hANP-predominant type (patient 4 of the

mitral disease group),

the mitral

atrium had

disease group, however, the

significantly higher levels of right

(7)

ANP, 7-hANP, and B-hANP than the left atrium. In addition, the plasma ANP level determined

from seven mitral disease patients was 131 ± 33.9

pg/ml, which is significantly higher than the

normal value for healthy subjects as found in our laboratory (22 ± 5 pg/ml).

Various clinical features were examined to

ascertain which had contributed to the increase of atrial total ANP and each ANP variant. When the subjects were divided into two groups on the basis

? 30 :. 9 4 a E -a 3 2o se y s S S 10 ･E-E RHYTHM o

-o

.

_oo

: . . ;

o

fotaL B-hANP lbtat B-hANP

hANP hANP

SR Af

(n=21) (n=10)

Fig. 2 RelationshipsbetweenrightatrialtotalhANPand

B-hANP levels and cardiac rhythm

Abbreviations as in Table 1. Mean is shown by bar, e : mitral disease, O: aortic disease, A: ASD, A: others.

tP<O.OOI vs corresponding values in patients with SR,

s

L* :

L*

_:

4:

-th

d ? 30 t'

s

J9 2o Sa : 10 t CTR g

i

: . . s .

-*

i o . N : : . -!-* . R

o

Totat B-hANP lbteL B-hANP

hANP hANP

E60ete 60eie<

(n=17) <n=14>

Fig. 3 RelationshipsbetweenrightatrialtotalhANPand B-hANP levels and CTR

Abbreviations as in Table 1. Mean is shown by bar. O: mitral disease, O: aortic disease, A: ASD, A: others, EP<O.Ol vs corresponding values in patients with CTR less than 60%,

of either age (more than 50 and less than 50 years old) or clinical severity according to the New York

Heart Association (NYHA) classification, no

statistically significant differences in atrial ANP levels were found. In contrast, the presence of atrial fibrillation (Fig. 2) and an increased CTR (Fig. 3) correlated respectiveiy with increased levels of total ANP and of B-hANP.

Of the various hemodynamic parameters, right atrial ANP levels showed a significant positive correlation with the right atrial pressure (r= O.654, p<O.OOI) (Fig. 4), pulmonary arterial pres-sure (r==O.468, P<O.Ol), and pulmonary capillary wedge pressure (r=O.447, P<O.O05). B-hANP levels also showed a similar positive correlation with these hemodynamic parameters (Fig. 5). In addi-tion, the presence of atrial fibrillaaddi-tion, elevated

rt 3o g 4 g -a-:. 20 :. E i :pt 10 E . s

I･

.

o

O 10 20 30

Righ± atriaL pressure (mmHg)

Fig. 4 Relationships between right atrial total hANP

levels and right atrial pressure

e: mitral disease, O: aortic disease, A: ASD, A:others.

Regression line calculated by least squares analysis.

O ..o: 4: rt 3o 6.de -,k i' 2o

E

l iO ,T' o . o : . i t': . .

e lo 2e 3o

Right etriat pressure CmmHg)

Fig. 5 Relationships between right atrial B-hANP levels

and right atrial pressure

e: mitral disease, O: aortic disease, A: ASD, A: others. Regression line calculated by least squares analysis.

(8)

pulmonary arterial pressure, and elevated CTR

correlated with the right atrial 7-hANP levels. Although right atrial ANP levels tended to cor-relate negatively with cardiac index, the relation was not statistically significant. There was no significant correlation between right atrial ANP levels and mean arterial pressure.

In contrast to the right atrial ANP levels, left atrial ANP levels did not correlate significantly

with any of the hemodynamic parameters inves-tigated. Although the left atrial ANP levels tended to show even a negative correlation with left atrial size, the relation was not statistically significant.

Discussion

The present study demonstrates that the atrial ANP levels are markedly elevated in patients with

mitral disease compared to patients with

non-cardiac disease, aortic disease, or ASD. Atrial pressure, which has been reported to be an im-portant factor regulating ANP secretioni6)Ni8), was significantly elevated in mitral disease but not in aortic disease or ASD. Therefore, we suggest that the increased atrial ANP levels could be the result of increased atrial pressure. The finding that

plasma ANP levels were also elevated suggests

that increased atrial pressure stimulates atrial production of ANP in mitral disease.

This conclusion is further supported by the

strong correlation observed between atrial ANP levels and right atrial pressure. We also found a less prominent but significant correlation of right

ANP levels with pulmonary arterial pressure

(r=O.468, P<O.Ol) and pulmonary capillary wedge

pressure (r=O.447, P<O.O05), suggesting that pressures in these areas could also stimulate

production of ANP in the right atrium. This

possibility is supported by findings of increased plasma ANP levels in hypoxia-induced pulmonary hypertension in ratsi9) and in patients with congen-ital pulmonary disease20}. In addition, we recently reported that plasma ANP levels correlate with

pulmonary arterial pressure in patients with

angina pectoris2i), which suggests that pulmonary arterial pressure may have some role in

modula-ting ANP secretion in certain pathologic states. The existence of vagal afferent fibers arising from

the nerve endings near the pulmonary

capil-laries22) indicates that central nervous system

mediated reflexes might be involved in the in-teraction between pulmonary arterial pressure

and ANP secretion. However, there is a significant correlation among right atrial, pulmonary

arte-rial, and pulmonary capillary wedge pressures, which can obviously be attributed to the

back-ward effect of the increased left atrial pressure. The correlation of right atrial ANP levels with

pulmonary arterial and pulmonary capillary

wedge pressures, therefore, may only reflect this pressure relation.

Another factor that correlates with increased right atrial ANP levels is the presence of atrial

fibrillation. Both fixed and paroxysmal atrial fibrillation is known to be associated with an increased plasma ANP levelsi3)23)24). It is possible that atrial fibrillation stimulates the production of ANP in the atrium as well as its secretion. Among the patients in this study, atrial fibrillation was mainly observed in patients with mitral disease, in whom atrial pressures are increased. In addi-tion, patient 7 with ASD and patient 1 with a left atrial aneurysm, who had the highest atrial ANP levels in their respective disease groups, exhibited both elevated atrial pressure and atrial fibrilla-tion. Therefore, it cannot be excluded that creased atrial ANP levels are related to the in-creased atrial pressure rather than to the presence of atrial fibrillation. Of the patients with mitral regurgitation, however, patients 7 and 8, who had atrial fibrillation, showed atrial ANP levels

sev-eral times higher than in patient 9, who had regular sinus rhythm, although atrial and

pul-monary arterial pressures were more elevated in patient 9 than in patients 7 and 8. This finding gives further support to the notion that atrial fibrillation is one of the important conditions

affecting ANP production. We cannot conclude

from the present study, however, whether atrial fibrillation itself can stimulate ANP production in the absence of chronically increased atrial

(9)

pres-sure, nor can we suggest the mechanism by which

this irregular movement of atrial wall might

affect ANP production.

The finding in patient 6 with tricuspid regur-gitation is interesting. Although this patient had increased atrial pressure and atrial fibrillation,

atrial ANP levels were no higher than in the

autopsy group. At least two possible explanations for this finding must be considered. First,

monary capillary wedge pressure and/or

pul-monary arterial pressure (which were in the

normal range in this patient) may be more critical than right atrial pressure in stimulating right

atrial ANP production. However, in our series ANP ieveis correlated most strongly with right

atrial pressure. An alternative explanation

in-volves the duration of increased right atrial pressure. In our mitral disease group, the duration of the disease estimated from the history ranged from 8 months to 36 years, with an average of 17 years. Although there was no direct correlation between the duration and the ANP levels, patient

9 who had the shortest duration also had the

lowest atrial ANP levels, and patient 5 who had the longest duration had the highest atrial ANP ieveis. The duration of disease in the patient with tricuspid regurgitation -estimated at 3 years from the history- might be too short to induce increased atrial ANP levels.

The clinical severity of congestive heart failure was determined according to the NYHA Funcional

Classification, which has been reported to cor-relate with plasma ANP levels25}"'27). Eight patients

fell in class I, 18 in class II, 4 in class III, and 1 in

class IV. Although the atrial ANP level in the class III and IV patients (9.02 ± 3.60 ptg/mg protein) was higher than in the class Iand II patients (5.09 ± 1.43 ug/mg protein), the difference was not sta-tisticaily significant. The reason for the

discre-pancy with the previous studies25)'J27) is not

known. However, in the present study,

classifi-cantion according to NYHA was determined

within 1 week before the heart surgery, and

therefore could have been influenced by

treat-ments including medication, and restriction of

diet and excercise since admission to the hospital.

The immunoreactive ANP in our specimens

comprised three major components corresponding

to 7-hANP, B-hANP, and or-hANP, in agreement with previous findingsii)i2). In the autopsy group, who died of noncardiac disease, atrial 7-hANP and B-hANP levels were almost identical, and the a-hANP level was much lower. In contrast, in eight of the nine patients with mitral disease, atrial

ANP was composed mainly of B-hANP. B-hANP

was also the predominant molecular form in the atrium of the patients with other cardiac diseases in whom atrial ANP levels were high: patients 4, 7, and 8 with aortic disease, patient 7 with ASD, and the patient with a left atrial aneurysm. There was a significant correlation between total ANP

and B-ANP levels. Furthermore, 6-hANP levels correlated with the same clinical feature and hemodynamie variables as the total ANP levels

-the presence of atrial fibrillaiton, and right atrial, pulmonary arterial, and pulmonary capil-lary wedge pressures. These results indicated that B-hANP becomes the main molecular form of ANP in chronically overloaded hearts.

6-hANP, an antiparallel dimer of a-hANP, was

reported to have a biological activity different from a-hANP. Its natriuretic activity is weaker but longer-lasting than that of a-hANP, whereas its hypotensive effect is comparabie to that of ev-hANPi2}. Such a dimeric peptide with bioactivity has never been found in the tissue of any other animal species. How and why B-hANP is increased

in the chronically overloaded atrium is not

known. However, it seems improbable that

B-hANP is produced from two a-B-hANP molecules,

since the disulfide linkage between the two cys-teine residues at positions 7 and 23 is chemically stable. Therefore, we suggest that the dimeric form may be generated before the disulfide linkage is formed in the or-hANP molecule- i.e., early in ANP processing. If so, we predict the presence of

an immunoreactive ANP with higher molecular

weight than y-hANP. Consistent with this hy-pothesis, we have sometimes found on HP-GPC a very small peak of immunoreactive ANP with an

(10)

apparent molecular weight of 25,OOO daltons, although further studies are required to elucidate the nature of this high-moleuclar-weight ANP.

It has been reported the B-hANP is rapidly converted to a-hANP in the plasma28)29). It is not known, however, whether B-hANP itself is also

secreted into the circulation. We have compared

the molecular form of atrial ANP with that of

plasma in seven patients with mitral disease in

whom a B-hANP was the predominant molecular

form in the atrium. In six of the seven patients,

B-hANP was present as a minor component in

addition to a-hANP in the plasma. Thus, all the

patients with B-hANP in their plasma also had B-hANP in their atrial tissue29).

We recently reported the molecular form of

plasma ANP before and after atrial pacing at 150 beats/min for 30 min in two patients with sick

sinus syndrome29). Plasma ANP was increased

significantly by atrial pacing, and the increase involved B-hANP as well as cr-hANP. In addition, B-hANP has been found in the plasma of patients with renal and heart disease30). Taken together, these results suggest that B-hANP is secreted into the circulation under certain conditions. 6-hANP

could be a precursor substance that quickly

supplies two molecules of a-hANP, or it might exhibit unique biological activities without being

converted to or-hANP. Whichever its mode of

action, an increase in B-hANP may mark

adap-tation to a chronic hemodynamic overload.

ANP is also produced in the left atrium.

Although both right and left atrial pressures were elevated above normal to a'similar extent in the mitral disease group, right atrial ANP levels were two times higher than in the left atrium. Fur-thermore, left atrial ANP levels did not correlate with any of the factors that correlated with right atrial ANP levels. It is noteworthy that not even pulmonary capillary wedge pressure or left atrial size correlated with left atrial ANP level. In fact, there was an insignificant negative correlation between left atrial size and left atrial ANP level (r= -O.469). The exact explanation for this finding is not known. However, it can be anticipated that

histologic deterioration and fibrosis of the tissue may be more prominent in the left atrium than in the right atrium in mitral disease in which left atrial pressure is primarily increased. Extensive tissue destruction may reduce the number of the atriocytes that produce ANP.

Another factor that should be taken into

ac-count is the difference of the conditions under which the left and right atrial tissue specimens

were taken. The right atrial appendage was

trimmed during cannulation for cardiopulmonary bypass, whereas the left atrial appendage was trimmed after the completion of the

cardiopul-monary bypass. Accordingly, the hemodynamic

conditions could be quite different when the left atrial tissue was obtained, and the value of

cor-relating left atrial ANP levels with the

hemo-dynamic variables determined before surgery may be limited.

It has been reported that normal rat ventricle can adapt to an imposed systoilc load by

decreing the proportion of Vi isomyosin and the as-sociated ATPase activity, resulting in a lower

velocity of muscle shortening and probably a more

energy-efficient contraction3i). More recent

studies32)33) have found that human atrial tissue, which is biochemically similar to rat ventricle in isomyosin distribution, has a similar capacity to alter the ratio of Vi to V3 isomyosins in response to hemodynamic loads. The increase in ANP and its

molecular shift to B-hANP could be another

adaptation performed by the atrial tissue in the pressure/volume overloaded heart. In contrast to the myosin isozymes34), changes in ANP molecular form may reflect hemodynamic overload of longer duration (although this will not necessarily hold in advanced stages of cardiac disease, as seen in the left atrium in our mitral disease group). Taken together, these findings lead to the intriguing hypothesis that human atrial tissue adapts to overload conditions through two dif-ferent mechanisms-an increase in ANP

produc-tion accompanied by a shift to the B-hANP form,

and changes in isomyosin distribution. The

(11)

the peripheral tissues, whereas the isomyosin

ratio affects the heart directly. In addition, our findings give further support to the clinical im-portance of performing a biochemical analysis of the atrial tissue specimen to evaluate the bio-chemical adaptation of the failing heart, although more comparative studies of biopsied atrial tissue are needed.

Acknowledgement

The authors thank Dr. M. Takeishi and the staff of

the Department of Pathology I and Dr. A. Kajita and the

staff of the Department of Pathology II, Tokyo Women's

Medical College for their assistance in obtaining autoptic

tissue specimens and Ms. Fumiko Numata and Ms. Yuko

Kato for their able technical assistance.

This work was supported in part by a Grant-in-Aid for Scientific Research of the Yayoi Yoshioka Award, a

Grant-in Aid for Scientific Research from the Ministry of

Education, Science, and Culture, Japan, and by a research grant from the Foundation for Growth Science

in Japan.

References

1) Cantin M, Genest J: The heart and the atrial uretic factor. Endocrine Rev 6: 107-127, 1985

2) Cole BR, Needleman P: Atriopeptins; velume regulatory hormones, Clin Res 33: 389-394, 1985

3) DeBold AJ: Atrial natriuretic factor: a hormone

produced by the heart. Science 230: 767--770, 1985

4) Laragh JH: Atrial natriuretic hormone, the

aldosterone axis, and blood pressure-electrolyte eostasis. N Engl J Med 313: 1330-1340, 1985

5) Burnett JC Jr, GrangerJP, Opgenorth TJ: Effects

of synthetic atrial natriuretic factor on renal function and renin release. Am J Physiol 247: F863-F866, 1984

6) Maack T, Marion DN, Camargo MJF et al: Effects

of auriculin (atrial natriuretic factor) on blood pressure,

renal function, and the renin-aldosterone system in dogs. Am J Med 77: 1069-1075, 1984

7) Atarashi K, Mulrow PJ, Franco-Saenz R et al: Inhibition of aldosterone production by an atrial

extract. Science 24: 992-994, 1984

8) Kudo T, Baird A: Inhibition of aldosterone production in the adrenal glomerulosa by atrial natriuretic factor.

Nature 312: 756-757, 1984

9) Samson WK: Atrial natriuretic factor inhibits

dration and hemorrhage-induced vasopressin release, Neuroendocrinology 40: 277-279, 1985

10) Obana K, Naruse, M, Inagami T et al: Atrial

natriuretic factor inhibits vasopressin secretion from

rat posterior pituitary, Biochem Biophys Res Commun 132: 1088-1094, 1985

11) Kangawa K, Matsuo H: Purification and complete amino acid sequence of a-human atrial' natriuretic polypeptide (a-hANP). Biochem Biophys Res Commun 118: 131-139, 1984

12) Kangawa K, Fukuda A, Matsuo H: Structural identification of beta- and gamma-human atrial

uretic polypeptides (B- and 7-hANP). Nature 313: 397-400, 1985

13) Naruse M, Naruse K, Obana K et al:

tive a-human atrial natriuretic polypeptide in human plasma. Peptides 7; 141-145, 1986

14) Yoshinaga K, Yamaguchi K, Abe K et al: mination of atrial natriuretic polypeptide (ANP) in human plasma. Biomed Res 7: 173-179, 1986

15) Bradford MM: A rapid and sensitive method for the

quantitation of microgram quantities of protein ing the principle of protein-dye binding. Ann Biochem

72: 248-254, 1976

16) Dietz JR: Release of natriuretic factor frorn rat

lung preparation by atrial distension. Am J Physiol 247: RI093--RI096, 1984

17) Lange RE, Th61ken, H, Ganten D et al: Atrial

natriuretic factor-a circulating hormone stimulated by volume loading. Nature'314: 264-266, 1985

18) Sato F, Kamoi K, Wakiya Y et al: Relationship

between plasma atrial natriuretic peptide levels and

atrial pressure in man. J CIin Endocrinol Metab 63: 823-827, 1986

19) Mckenzie JC, Tanaka T, Inagami T et al:

tions in atrial and plasma atrial natriuretic factor (ANF)

content during development of hypoxia-induced

monary hypertension in the rat. Proc Sco Exp Biol Med

181: 459-463, 1986

20) Lang RE, Unger T, Ganten D et.al: a-atrial

natriuretic peptide concentrations in plamsa of children with congenital heart and pulmonary diseases, Br MedJ

291: 1241, 1985

21) Naruse K, Naruse M, Honda T et al: Atrial

triuretic peptide correlates with pulmonary arterial pressure and cardiac output. Peptides 8: 285-290, 1987 22) Paintal AS: Mechanism of stimulation of type J monary receptors. J Physiol 203: 511-532, 1969

23) Yamaji T, Ishibashi M, Nakaoka H et al: Possible

role for atrial natriuretic peptide in polyuria associated

with paroxysmal atrial arrhythmia. Lancet i: 1211, 1985 24) Roy D, Paillard F, Cassidy D et al: Atrial natriuretic factor during atrial fibrillation and supraventricular tachycardia. J Am Coll Cardial 9: 509-514, 1987

25) Nakaoka H, Imataka K, Amano M et al: Plasma

levels of atrial natriuretic factor in patients with

congestive heartfailure. N Engl J Med 313: 892-983, 1985

26) Shenker Y, Sider RS, Ostafin EA et al: Plamsa

levels of immunoreactive atrial natriuretic factor in

healthy subjects and in patients with edema. J CIin Invest 76: 1684-1687, 1985

(12)

Plasma atrial natriuretic peptide in cardiac disease and during infusion in healthy volunteers． Lancet ii：66−69，

1985

28）Itoh H， Nakao K， Shiono S et a1：Conversion ofβ一

human atrial natriuretic polypeptide intoα一human

atrlal natriuretic polypeptide in human plasma in vitro．

Biochem Biophys Res Commun 143：560−569，1987 29）Naruse M， Naruse K， Hifumi S et al：Molecular

heterogeneity of immunoreactive ANP in human

plasma． Biologically active atrial peptides． ASH sym−

posium series 2：183−186，1988

30）Miyata A， Toshimori T， Hashiguchi T et al：

Molecular forms of atrial natriuretic polypeptides cir−

culating in human plasma． Biochem Biophys Res Commun l42：461−467，1987

31）Mercadier JJ， Lompre AM， Wisnewsky C et ah

Myosin isoenzymatic changes in several models of rat

cardiac hypertrophy． C重rc Res 49：525−532，1981

32）Gorza L， Mercadier JJ， Schwartz K et al：Myosin types in the human heart． An immunofluorescence

study of normal and hypertrophied atrial and ventri−

cular myocardium． Circ Res 54：694−702，1984 33）Yazaki Y， Tsuchimochi H， Kuroo M et al：Dis・ tribution of myosin isozymes in human atrial and

ventricular myocardium：comparison in normal and

overloaded heart． Eur Heart∫5（suppL F）：103−110，1984

34）Buttr韮ck PM， Malhotra A， Brodman R et al： Myosin isoenzyme distribution in overloaded human

atrial tissue． Circualtion 74：477−483，1986

ヒト心房組織中の心房性Na利尿ペプチド濃度とその存在様式について

東京女子医科大学第2内科，探勝線科，2循環器内科，3循環器外科

成瀬清子・広江道昭1

橋本明政3・小柳仁3

広沢弘七郎2・鎮目和夫

・成瀬光栄・永田まこと2

・関口守衛2・出村博

ヒト心房組織中心房性Na利尿ペプチド（ANP）の組織中含量とその存在様式を非心疾患（n＝6）

の剖検および各種心疾患例の開心手術（n＝31）に際して得られた心房組織において検討した．ANP

の組織中含量および3種の存在様式すなわちα一hANP，β一hANP，γ一hANPは，ラジオイムノアッセ

イと高速ゲル濾過の組み合わせにより測定した．僧帽弁疾患における心房組織中ANP含量は，

12．2±2．5μg／mg proteinで，剖検例や大動脈弁疾患，心房中隔欠損でのそれに比し，有意に高値を示

し，その増加は主にβ一hANPのそれによることが明らかになった．心房組織中総ANP含量および

β一hANP含量は共に，心房細動の存在と関連すると共に，右心房圧，肺動脈圧，肺動脈懊入圧などと有意な正相関を示した．以上のことから，ヒト心房組織は，慢性的な負荷状態において，ANP産生の

増加とそのプロセッシングのβ一hANPへの偏位にて対応することが示竣された．このα一hANPの二

量体であるβ一hANPは，心臓の慢性的な病的状態における適応を示す良い指標になると考えられた．