Immunoreactive atrial natriuretic factor is present in both atria and ventricles

105

Molecular and Cellular Endocrinology, 52 (1987) 105-113 Elsevier Scientific Publishers Ireland, Ltd.

MCE 01684

Immunoreactive atria1 natriuretic factor is present in both atria and ventricles M. Cantin, J. Ding, G. Thibault, J. Gutkowska, L. Salmi, R. Garcia and J. Genest Laboratory of Pathobiology, Clinical Research Institute of Montreal and Depariment of Pathology, UniversitP de Mont&al, Montreal, Canada (Received 19 January 1987; accepted 16 March 1987)

Key words: Atria1 natriuretic factor; Atrial cardiocyte; Ventricular cardiocyte; (Culture)

A comparative study has been made of the amount and form of immunoreactive atrial natriuretic factor (ANF) present in atria versus ventricles in situ and in cultures of atrial and ventricular cardiocytes in rats of various ages: foetus (- 2 days of age), newborn (3 days of age), 10 days, 35 days and 70 days old. It was first established that ANF is present in the circulation at all ages investigated, the highest levels being found in the foetus. The atrial content of ANF increased gradually with time and was always in the microgram range. In ventricles, where it was in the nanogram range, total ANF increased with age. More ANF was found in the right than in the left atrium at all ages except in the foetus where the reverse was true. In ventricles, ANF was distributed equally between right ventricle, left ventricle and septum in the foetus, the newborn and in the adult. At 10 days and 35 days, ANF was unequally distributed: (right ventricle > septum > left ventricle). In cultured atria1 cardiocytes, ANF was more abundant than in cultured ventricular cardiocytes at all ages and the amount decreased with the age of the donor animals. Atria1 cardiocytes secreted more ANF than did ventricular cardiocytes at all ages investigated, the amount secreted generally decreasing with the age of the donor animals. The secretoty’activity of atrial cardiocytes per hour was also higher than that of their ventricular counterparts; over this time period, both types of cells secreted more ANF in the presence of serum. Purification of atrial and ventricular ANF by high pressure liquid chromatography revealed that ANF (as assessed by radioimmunoassay) is concentrated in an identical high molecular weight peak in both structures. These results suggest that immunoreactive ANF is present not only in the atria but also in the ventricles. The total content of ANF in the atria is higher by at least one order of magnitude than in ventricular tissue; the amount present in and secreted by ventricular cardiocytes of adult rats is very small.

Address for correspondence: Dr. Marc Cantin, Clinical Research Institute of Montreal, 110 Pine Ave West, Montreal, Quebec, Canada H2W lR7. Supported by a Group Grant from the Medical Research Council of Canada to the Multidisciplinary Research Group on Hypertension, by the National Research Council of Canada, 0303-7207/87/%03.50

The Ministere de la Science et de la Technologie du Quebec, BioMdga Inc., the Pfizer Company and the Canadian Heart Foundation. J. Ding is Visiting Scientist from Fuwai Hospital, Beijing, The People’s Republic of China.

0 1987 Elsevier Scientific Publishers Ireland, Ltd.

106

Introduction It is now well established that the cardiac atria are endocrine glands, secreting a peptide (atria1 natriuretic factor (ANF)) of 28 amino acids (Ser99-Tyr’26). This hormone produces a brief diuresis and natriuresis, vasodilatation, inhibition of aldosterone, cortisol and renin secretion and inhibition of vasopressin release (Atlas, 1985; Cantin and Genest, 1985; de Bold, 1985; Needleman and Greenwald, 1986). The atrial cells of newborn and adult rats in culture secrete large amounts of immunoreactive (IR-) ANF (Cantin et al., 1985). The aim of the present study was to investigate if the rat cardiac ventricular cells from animals of varying age also contain IR-ANF and if cultured ventricular cardiocytes, like atria1 cardiocytes, contain IR-ANF and secrete the immunoreactive peptide into the medium. Materials and methods Animals Throughout the course of the present study, Sprague-Dawley rats (Charles River, St-Constant, Quebec) were housed in cages at 22” C with 60% humidity and a 6 a.m.-6 p.m. light regimen. They were maintained on regular pellet chow (Purina, Richmond, IN) and tap water and, when necessary, anaesthetized with sodium pentobarbital. In experiments described below, rats of various ages were used: foetuses (19 days old) from timed pregnancies (Charles River, St-Constant, Quebec); newborn (3 days old, 10 g body weight), 10 days old (40 g), 35 days old (100 g) and 70 days old (200 g). Measurement of IR-ANF in plasma of control animals of various ages Foetuses were decapitated and trunk blood pooled from 20 animals for a single radioimmunoassay (RIA). Blood from all other animals was collected through the jugular vein: for newborns, the blood from six animals was pooled for a single RIA; for the lo-day-old rats, blood from two animals was pooled. IR-ANF levels in other rats were assessed in each animal separately. In all cases blood was collected in tubes cooled in crushed ice and containing EDTA lop5 M,

phenylmethylsulphonyl fluoride (PMSF) 10e5 M and pepstatin 5 PM. The RIA used for plasma IR-ANF has been described elsewhere (Horky et al., 1985; Gutkowska et al., 1986a, b). Briefly, RIA for plasma IR-ANF was performed with prior extraction on Vycor glass beads: 0.5 ml of 25 mg heat-activated Vycor glass beads (Coming Glassworks, NY) suspended in distilled water was added to 1 ml of plasma and rotated for 30 min at 4O C. After 5 min centrifugation at 3000 ‘pm, the supematant was aspirated by vacuum and the glass powder washed first with 1.5 ml distilled water and then with 1.0 ml of 1 N HCl. The adsorbed IR-ANF was eluted with 1.0 ml 60% acetone in water. Acetone was evaporated under a nitrogen stream and the aqueous solution lyophilized in a Speed-Vat. The lyophilized extracts were stored at - 70 o C until assayed. For determination of IR-ANF, the lyophilized extracts were dissolved in 500 ~10.1 M phosphate buffer pH 7.4 containing 0.1% Triton X-100 in 0.1% BSA, 0.01% NaN,, 0.05 M NaCl, and 50 ~1 and 100 ~1 aliquots in duplicate were used for RIA. In each assay, zero standard tubes (without unlabeled rat ANF (Arg10’-Tyr’26)), non-specific tubes (without unlabeled ANF and antibody, buffer blank) were processed; standard unlabeled ANF in RIA buffer (34.4 and 69 pg/ml), pooled rat plasma with and without standard ANF added (34.5 and 69 pg/ml) and a rat plasma pool (1 ml) ‘251-ANF (15000 cpm) were also excontaining tracted as controls for each series of unknown samples. Good linearity was observed between concentrations of ANF from 0.75 to 390 pg/tube. The detection limit was 0.75 pg/tube (minimum amount of ANF that could be statistically distinguished from zero at two standard deviations) or 7.5 pg/ml of plasma. Since plasma was concentrated two times during the extraction, as little as 4 pg/ml could be detected by the method. The standard curves were reproducible: analysis (mean f SE) of ten standard curves showed a slope -1.12 f 0.04, an ED,, of 13.8 + 3.4 pg/tube, a binding of ‘251-ANF to antibody of 36.9 + 6.5% and a non-specific binding of 1.9 k 0.5%. Good parallelism of the serial dilutions of rat plasma extracts with the standard curves was observed. The inter- and intra-assay coefficients of variation of the RIA were below 14%. The IR-ANF values

107

plasma (Gutkowska et al., 1986a, b), with Sep-Pak cartridges. As can be seen in Table 1, this method gave values (obtained from the hearts of adult rats) for atrial IR-ANF which are not statistically different from those obtained by direct assay. In the ventricles, however, the total amount of IRANF measured after extraction was significantly lower than when the direct method was used. Because the total amount of protein measured after boiling is also smaller, the relative values obtained after extraction are not different from those obtained by the direct method. To eliminate all possibility of including atria1 tissue in the ventricular extracts the ventricles were cut 1 mm below the atrioventricular septum. The ventricles (with the interventricular septum) were minced in 1 N acetic acid containing 0.5% Nonidet P40 (10 ml/g tissue); EDTA 2.5 mM; pepstatin 0.1 mM; leupeptin 0.1 mM; PMSF 2.5 mM; Trasylol 100 U and captopril 0.01 mM. The tissues were boiled in the solution for 5 min then cooled in an ice bath. The tissues were homogenized and centrifuged at 8000 rpm for 10 min. A 10 ~1 aliquot of the supematant was then taken for protein determination (Lowry et al., 1951) and the material applied to a Sep-Pak cartridge (Waters Associate, Milford, MA) activated with 10 ml ethanol. The samples were washed with 10 ml H,O. The adsorbed ANF was eluted with 3 ml of 86% ethanol containing 4% acetic acid, and lyophilized in a Speed-Vat. The residue was taken in 1 ml of 1 N acetic acid and 5 ~1 and 20 ~1 aliquots taken for RIA as described for atria. The recovery for this method was 76 f 3.8% (mean ~fr SE). Results were not corrected for recovery.

were not corrected for recovery which was 83 + 3.7% (mean f SE). Measurement

of IR-ANF

in atria

Heart from foetuses, newborn rats, lo-day-old, 35day-old and 70-day-old were used. The right and left atria were dissected and used for measurement of IR-ANF as already described (Gutkowska et al., 1984). Briefly the tissues were homogenized in 1 ml of 0.1 N acetic acid with inhibitors as described for plasma for 30 s and centrifuged for 10 min at 16000 rpm. The pellet was discarded and the supernatant was collected and stored at - 70’ C until assayed. Statistical analysis of five randomly chosen standard curves showed a sensitivity as high as l-l.5 pg. In analytical recovery studies with two different quantities of synthetic ANF (Arg”’ -Tyr126) (5 and 20 pg) added to rat atria1 homogenates, recover ranged from 77 to 100% with a mean value (*SE) of 89.5 & 10%. The results were not corrected for recovery. The within-RIA coefficient of variance (CV) was 8.9% and the between-assay CV was 10.3%. Proteins in the extracts were measured by the method of Lowry et al. (1951). Measurement

of IR-ANF

in the ventricles

Since the amount of interfering substances in ventricles is potentially higher than in the atria and the amount of IR-ANF much smaller, a different technique for its measurement in ventricles was developed, based on the inhibition of a larger spectrum of enzymatic activity, on boiling to eliminate enzymatic activity altogether and on preliminary extraction of IR-ANF, as for human

TABLE

1

COMPARATIVE VENTRICULAR

STUDY OF RADIOIMMUNOASSAY IMMUNOREACTIVE (IR-) ANF

Technique

Atria Total protein (mg)

Direct method (n = 6) Extraction with Sep-Pak cartridges (n = 6) * P < 0.05.

TECHNIQUES

USED

FOR

MEASUREMENT

OF ATRIAL

AND

Ventricles IR-ANF

Total protein

IR-ANF

(mg)

Total (ng)

(ng/mg

10+1

76+1

176+18

2fO

851

34k2

Total (pg)

(ag/mg

2*0

2Ok3

2fO

19*1

protein)

*

1os*22

*

3fl

protein)

108

Measurement of IR-ANF in right and left ventricular walls and in the interventricular septum To determine the precise distribution of IRANF in the ventricles, the above experiment was repeated but, this time, the medial lateral part of the right ventricle, the medial portion of the interventricular septum and the medial anterolateral part of the left ventricle were dissected and treated as above. Culture of atria1 and ventricular cardiocytes Rats of the same age range as above were used. The technique of preparation of atrial and ventricular cell suspensions and their culture were as previously described (Cantin et al., 1980, 1981) with minor modifications (Moses and Claycomb, 1984; Cantin et al., 1985). Briefly, the right and left atria and the inferior two-thirds of both ventricles and interventricular septum were removed aseptically and immediately placed in modified Joklik fluid (Gibco Laboratories, NY). They were rinsed twice in the same fluid, minced and dissociated with collagenase (Worthington Type II (1 mg/ml)) for 20 min at 37” C and pipetted 40 times. Cold fetal calf serum (FCS) (20%) was then added to the suspension. After removal of the free cells and centrifugation, the remaining tissue was digested for a further period of 20 min in the same conditions, pipetted 20 times and 20% fetal calf serum added. The final cell suspension was then passed through a 100 pm nylon filter and centrifuged. Quasi-pure cultures of cardiocytes can be obtained by preplating cells on gelatinized flasks to permit more rapid, selective attachment of non-cardiocytes (Cantin et al., 1986). In the present experiment, this step was omitted because previous studies had shown that the amount of IR-ANF present in newborn and adult cardiocytes and their secretory activity can be increased several fold by culturing these cardiocytes in the presence of mesenchymal cells of cardiac origin (Cantin et al., 1986). The cell suspension (containing cardiocytes and a mixture of various mesenchymal cells) was then seeded at a density of lo6 in 35-mm diameter plastic wells (No. 3406 Mark II, Costar, MA) in 2 ml of Eagle’s minimal essential medium supplemented with 15% FCS, vitamins (1 ml/100 ml) (Flow Laboratories, Virginia), amino acids (2 ml/l00 ml) (Flow Laboratories) and 2 ml/100 ml

of antibiotics (penicillin, 10000 U/ml, fungizone 25 pg/ml, streptomycin 10000 pg/ml) (Gibco Laboratories, Ohio). The cells were maintained in an incubator at constant humidity and temperature (37°C) with 95% air and 5% CO,. Media were changed every other day. Radioimmunoassay of ANF in cells and culture media When the cells had reached confluence (between the 5th and 6th day of culture), media were collected and 24 h secretory activities (minus degradation) of either atria1 or ventricular cardiocytes were measured by RIA. Before harvesting the cells, fresh medium was added either with 15% FCS in half the wells or without FCS in the other half. The cells were then incubated for 1 h and media collected (1 h secretion). IR-ANF in culture media was measured exactly as for plasma as described above. In cells, IR-ANF was measured as for atria1 tissue by the direct method (i.e. without extraction). Proteins in culture media or cells were measured by the method of Lowry et al. (1951). Comparative HPLC pattern of atria1 and ventricular ANF in situ Five 70-day-old rats were used. The hearts were removed, rinsed in cold PBS and the atria and both ventricles with septum were dissected and treated as indicated above for measure of IR-ANF. The samples to be analyzed (either both atria or ventricles) were loaded on a C,, Vydac column (0.3 x 25 cm). The material was eluted at 1 ml/mm with a linear gradient (0.5%/min) of 15-55% acetonitrile in 0.1% trifluoroacetic acid. Two ml fractions were collected and an aliquot of each was used to measure ANF by RIA as described above for plasma. Statistical analysis of data One-way analysis of variance (ANOVA) and the Student-Newman-Keuls that were used throughout except for the comparison of data between the two radioimmunoassay techniques (Table 1) where the unpaired t-test was used and the comparison of data between the two atria at each time interval (Table 3) where the paired t-test was used. Results are expressed as mean k SE. A P > 0.05 was taken as not significant.

109

Results

100

Plasma IR-ANF

ATRIA

in control animals of various ages

YENTRICLES

Before starting the study on the possibility of secretion of IR-ANF by atria1 and ventricular cardiocytes, it was deemed necessary to find out if IR-ANF is present in the circulation at various stages in the rat life span. As can be seen in Table 2, IR-ANF was found to be present in plasma at all ages investigated, the highest values being present in foetuses. Total content of IR-ANF in atria and ventricles in situ in control animals of various ages

The total content of IR-ANF in atria increased regularly with time to reach a maximum at 70 days of age (Fig. 1). There was a marked difference in the total amount of IR-ANF in the ventricular walls in the foetus compared with other rats. The ratio of IR-ANF content between atria and ventricles is 20 : 1 in the foetus; 13 : 1 at 3 days of age; 60 : 1 at 10 days of age; 50 : 1 at 35 days and 152 : 1 at 70 days. Relative content of IR-ANF in right and left atria and ventricular walls in situ in control animals of various ages

The amount of IR-ANF in atria, when expressed per mg of protein, was in the microgram range at all time intervals (Table 3). It was higher in the right than in the left atrium except in the foetus where the reverse was true. It increased regularly with time to reach the highest levels in adult animals. The amount found in the ventricular walls was in the nanogram range at all time intervals. The amount found in ventricular walls decreased regularly with age to reach the lowest values in adult animals. At age 10 and 35 days, IR-ANF was much more abundant in the right ventricle than in the left or in the septum. There

TABLE

AGE

Fig.

(DAYS)

1. Values of immunoreactive ANF in rat atria ventricles in situ (n = 10). Bars represent SE.

was very little difference between the amount found in the various ventricular walls at other times. IR-ANF in cultured cardiocytes and in medium

As visualized by phase-contrast microscopy, the cells reached confluence between the 5th and 6th day of culture. Cardiocytes formed rhythmically beating islands amongst non-beating (mesenchymal) cells. Electron microscopy (data not shown) revealed relatively mature cardiocytes with a well-formed contractile apparatus. In all cases, cardiocytes formed between 40 and 60% of the total cell population as visualized by phase-contrast microscopy and as verified by electron microscopy. The cells were always harvested at confluence. As can be seen in Table 4, the amount of IR-ANF found in cultured cardiocytes from

2

IMMUNOREACTIVE

ANF

IN THE PLASMA

Age (days)

-2

IR-ANF

300*43 (n=S)

@g/ml)

* P i 0.05 as compared

*

with other values.

and

OF RATS OF VARIOUS

AGES

3

10

35

70

15Ok23 (n =lO)

157*:6-l (n = 10)

ES+17 (n =lO)

13k29 (n =lO)

110 TABLE

3

LEVELS

OF IMMUNOREACTIVE

Age (days)

ANF

IN VARIOUS

-2

IR-ANF (pg/mgprotein) Right atrium (n = 10) Left atrium (n = 10)

3

4& 4+

IR-ANF (ng/mg protein) Right ventricle (n = 10) Septum (n = 10) Left ventricle (n = 10)

CARDIAC

1 1

10

5+ 5+

120 f 28 180+22 171+ 29

STRUCTURES

0 0

35

6+ 5*

54+14 93*19 64+13

70

7+1

0 1

lOf1 7+1*

7*0

59+11 9+ 2* 4& o*

35*4 10*2 1*0

1*0 1+0 1*0

* *

* P < 0.05.

TABLE

4

IMMUNOREACTIVE

Age (days)

ANF CONTENT

OF CULTURED

-2

IR-ANF (ng/mgprotein) a 1831+604 Atrium Ventricle 93f 3

CARDIOCYTES

3

10

35

705 + 50 87* 5

431 f 16 5+ 0 **

57+2 2*0

70

* **

102+4 1*0

* **

a The results are from at least ten wells in each of three replicate experiments. * P i 0.05 as compared with atrial values obtained from cultured cardiocytes of younger animals. * * P c 0.05 as compared with ventricular values obtained from cultured cardiocytes of younger animals.

TABLE IR-ANF

5 IN CULTURE

Age (days) IR-ANF Atrium

MEDIUM

-2

(pg/mgprotein) a 8960+174

(24 h SUPERNATANT) 3

*** 17 ***

OF ATRIAL

AND VENTRICULAR

10

24680+11240 64105

** 70 ****

35

1729*70

**

Ventricle

1460+

a The results * P < 0.05 as * * P < 0.05 as * * *P -c 0.05 as * * * *P < 0.05 as

are from at least ten wells in each of three replicate experiments. compared to atrial values of lo-, 35- and 70-day-old rats. compared to ventricular values. compared to ventricular values of lo-, 35- and 70-day-old rats. compared to all other ventricular values.

animals of different ages was always higher in atria1 cells but was also present in the nanogram range at all time intervals in ventricular cardiocytes. In both atria1 and ventricular cells, the amount of IR-ANF decreased with the age of the donor animals. The ratio between atria1 and ventricular IR-ANF varied from 20 : 1 in the foetus to 8: 1 at 3 days of age, to 90: 1 at 10 days, to 35 : 1 at 35 days to 146 : 1 at 70 days. The amount of IR-ANF secreted made a quantum jump when

135+

CARDIOCYTES

6

70

482*24 8&

** 1

1534+18 9*

** 0

atria1 and ventricular cardiocytes were cultured from 3-day-old rats as compared to values obtained from foetuses (Table 5). From then on, it decreased with age, more from ventricular than from atria1 cardiocytes to reach low levels (9 pg/mg protein) in ventricular cardiocytes of adult animals. The ratio between IR-ANF secreted by atria1 and ventricular cardiocytes varied from 6 : 1 in the foetus to 4: 1 at day 3 to 13 : 1 at day 10 to 61: 1 at day 35 and to 170 : 1 at day 70. The

111

TABLE

6

IMMUNOREACTIVITY CARDIOCYTES

ANF

MEDIUM

(1 h SUPERNATANT)

OF

ATRIAL

AND

VENTRICULAR

3

10

35

IO

3860+ 67 4548 f 105

4096f 80 10 392 + 276

1002k41 1317+67

38Ok18 428 f 15

1106+30 1588+25

a

Atria Without serum With serum Ventricle Without serum With serum a The results exposed or cardiocytes ventricular

CULTURE

-2

Age (days)

ANF (m/mU

IN

676* 868*

12 16

567k 157&

are from at least ten wells in not to serum from animals exposed or not to serum from cardiocytes exposed or not to

21 21

101+ 6 115 f 10

55* 49+

4 6

11+ 21+

7 7

each of three replicate experiments. The differences observed in ventricular cardiocytes of 10, 35 and 70 days of age are not significant. The differences observed in atria1 animals of 35 days of age are not significant. All the other differences between atria1 or serum are significant

amount of IR-ANF secreted by cardiocytes over 1 h (Table 6) was always higher in serum-containing medium but was present in measurable amounts at all ages studied.

15

60

1

Comparative HPLC pattern lar ANF in situ

of atria1 and ventricu-

As can be seen in Fig. 2, the majority of IR-ANF in both atria and ventricles eluted at a similar position following immunological identification of HPLC purified material: ANF in both atria and ventricles represent quasi-identical high molecular weight material; much smaller peaks of low molecular weight material were also present in both atria and ventricles. Discussion

0.012-

T‘

,g

0.010

-

0.008

-

-60

VENTRICLE /* ,.=

z 9 s 2

0.006

-

0.004-

/ _

0002

.*

.’

.*

IX

.’

,A

,’

/

,A

,/’

-40

;‘ ;

c’ 3 :, -20

c’

;

--\ 20

40 ELUTION

TIME

60

80

(mm)

Fig. 2. HPLC pattern of immunoreactive ANF in atrium and ventricle of adult (‘lo-day-old) Sprague-Dawley rats. The low Mr form of ANF elutes at - 30-35% of acetonitrile while the high M, form elutes at 40-458 (Thibault et al., 1986).

The present results indicate that foetal plasma levels of IR-ANF are much higher than those found in newborn or adult animals. Whether the high levels are due to increased production by the foetal heart, transplacental passage of the peptide or decreased degradation remains to be determined. The fact that the placenta has been found to harbour ANF receptors (Sen, 1986) suggests a possible role for the peptide in the regulation of placental blood flow or fluid exchange. The present results also indicate both atria1 and ventricular cardiocytes contain IR-ANF of similar high molecular weight, and that cultured myocytes secrete the immunoreactive peptide in the medium at all ages. As already shown (Thibault et al., 1987) the major high M, form of ANF (13.6 kDa; ANF 1-126) from rat atria elutes at - 42% acetonitrile, while much lesser amounts elute at

112

lower concentrations. Analogous elution patterns of intracellular ANF have been found in cultured atria1 cardiocytes (Bloch et al., 1985; Glembotski and Gibson, 1985; Vuolteenaho et al., 1985). These results are thus in agreement with the presence of ANF mRNA not only in atria1 but also in ventricular cells of the rat (Nemer et al., 1986) and, from a phylogenetic viewpoint, with the presence of secretory granules (Cantin and Genest, 1985), ANF immunoreactivity (Chapeau et al., 1985) and bioactive ANF (de Bold and Salerno, 1983) in both atria1 and ventricular cardiocytes of nonmammalian vertebrates. There is a paucity of data on the developmental ultrastructure of the secretory apparatus of the rat atria1 and ventricular cardiocytes. Secretory granules are absent in rat heart embryo atria1 cardiocytes before the 10 mm stage (Jamieson and Palade, 1964; Bencosme and Berger, 1971) but are found at the 15 mm stage (Bencosme and Berger, 1971) and in late foetal and neonatal rats (Jamieson and Palade, 1964) where they persist for life. Secretory granules are said to appear in ventricular cells of rats and mouse early embryo and to disappear in late embryonic life (Nanot and Le Douarin, 1970). The presence of secretory granules in a few cardiocytes of adult rat ventricular cells has been described but, without cytochemical and immunocytochemical data, the images are impossible to distinguish with certainty from lysosomes (Bencosme and Berger, 1971). The Golgi complex of ventricular cardiocytes is also much smaller and less elaborate than that of their atria1 counterparts (Fawcett and McNutt, 1969). These morphological observations suggest that the secretion of ANF by ventricular cardiocytes may be of the constitutive type, in contrast with that of atria1 cardiocytes which would appear regulated (Tartakoff et al., 1978; Gumbiner and Kelly, 1982; Willingham and Pastan, 1984). The levels of IR-ANF are always higher by at least one order of magnitude in the atria than in the ventricles. While the relative atria1 levels increase with age, the ventricular levels decrease so that the ratio between the two increases with time. This is probably why ventricular extracts of adult rats do not induce diuresis and natriuresis in assay animals (de Bold et al., 1981). These results are confirmed by culture. Atria1 cardiocytes in culture

contain and secrete much more IR-ANF than ventricular ones at all time points investigated. The amounts contained and secreted by both atria1 and ventricular cells generally decrease with time. This is true whether the cell secretory activity is monitored for long or short intervals. In the latter case, the presence of serum stimulated the secretion of IR-ANF, particularly from atrial cells and, to the greatest degree, in atria1 cardiocytes of 3-day-old rats. Here again, the difference in reactivity between atria1 and ventricular cells may be due to the regulated character of the former. Why atrial cardiocytes of 3-day-old rats are more responsive to serum remains to be determined. Since ventricular cardiocytes secrete ANF, it is possible that, in vivo, they participate in the maintenance of ANF plasma levels. While this is plausible in the foetus and the neonate, it is unlikely to occur to any significant degree in normal adult rats given the low levels present in ventricles. Ventricular cardiocytes of adults may, however, participate in the maintenance of high plasma levels of IRANF in certain pathologic situations. In cardiomyopathic hamsters with heart failure, for instance, where the plasma levels of IR-ANF are extremely elevated, the levels of IR-ANF in the atria are significantly lowered while those of the ventricles are increased several fold, so that the ratio of IR-ANF between atria and ventricles, which is extremely high in control animals, becomes very small in animals with severe congestive heart failure (Ding et al., in press 1987). References Atlas, S. (1985) Recent Prog. Horm. Rex 42, 207-249. Bencosme, S.A. and Berger, J.M. (1971) In: Functional Morphology of the Heart. Methods Achiev. Exp. Pathol. Vol. 5, Eds.: E. Bajusz and G. Jasmin (S. Karger, Basel) pp. 173-213. Bloch, K.D., Scott, J.A., Zisfein, J.B., Fallon, J.T., Margolies, M.N., Seidman, C.E., Matsueda, G.R., Homey, C.J., Graham, R.M. and Seidman, J.G. (1985) Science 230, 1168-1171. Cantin, M. and Genest, J. (1985) Endocr. Rev. 6, 107-127. Cantin, M., Tautu, C., Ballak, M., Yunge, L., Benchimol, S. and Beuzeron, J. (1980) J. Mol. Cell. Cardiol. 12,1033-1051. Cantin, M., Ballak, M., Beuzeron-Mangina, J., Tautu, C. and Anand-Srivastava, M.B. (1981) Science 214, 569-570. Cantin, M., Dagenais, N., Salmi, L., Gutkowska, J., Ballak, M., Thibault, G., Garcia, R. and Genest, J. (1985) Clin. Exp. Hypertens. A7, 685-705.

113

Cantin, M., Thibault, G., Ding, J., Gutkowska, J., Salmi, L. and Genest, J. (1986) 11th Scientific Meeting Int. Society of Hypertension, Heidelberg, Aug. 31-Sept. 7 (Abstract). Chapeau, C., Gutkowska, J., Schiller, P.W., Milne, R.W., Thibault, G., Garcia, R., Genest, J. and Cantin, M. (1985) J. Histochem. Cytochem. 33, 541-550. de Bold A.J. (1985) Science 230, 767-770. de Bold, A.J. and Salerno, T.A. (1983) Can. J. Physiol. Pharmacol. 61, 127-130. de Bold, A.J., Borenstein, H.B., Veress, A.T. and SoMenberg, H. (1981) Life Sci. 28, 89-94. Ding, J., Thibault, G., Gutkowska, J., Garcia, R., Karabatsos, T., Jasmin, G., Genest, J. and Cantin, M. (1987) Endocrinology (in press). Fawcett, D.A. and McNutt, N.W. (1969) J. Cell Biol. 42, l-45. Glembotski, C.C. and Gibson, T.R. (1985) B&hem. Biophys. Res. Commun. 132,1008-1017. Gumbiner, B. and Kelly, R.B. (1982) Cell 28, 51-59. Gutkowska, J., Thibault, G., Januszewicz, P., Cantin, M. and Genest, J. (1984) Biochem. Biophys. Res. Commun. 122, 593-601. Gutkowska, J., Bonan, R., Roy, D., Bourassa, M., Garcia, R., Thibault, G., Genest, J. and Cantin, M. (1986a) B&hem. Biophys. Res. Commun. 139, 287-295. Gutkowska, J., Horky, K., Schiffrin, E.L., Thibault, G., Garcia, R., De Lean, A., Hamet, P., Tremblay, J., Anand-Srivastava,

M.B., Januszewicz, P., Genest, J. and Cantin, M. (1986b) Fed. Proc. 45,2101-2105. Horky, K., Gutkowska, J., Garcia, R., Thibault, G., Genest, J. and Cantin, M. (1985) B&hem. Biophys. Res. Commun. 129, 651-657. Jamieson, J.D. and Palade, G.E. (1964) J. Cell Biol. 23,151-172. Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 183, 265-275. Moses, R.L. and Claycomb, W.C. (1984) Am. J. Anat. 171, 191-206. Nanot, J. and Le Douarin, G. (1970) C.R. Sot. Biol. (Paris) 164, 890-893. Needleman, P. and Greenwald, J.E. (1986) N. Engl. J. Med. 314, 828-834. Nemer, M., Lavigne, J.P., Drouin, J., Thibault, G., Gannon, M. and Antakly, T. (1986) Peptides 7, 1147-1152. Sen, I. (1986) B&hem. Biophys. Res. Commun. 135, 480-486. Tarn&off, A., Vassali, P. and Detraz, M. (1978) J. Cell Biol. 79, 694-707. Tbibault, G., Garcia, R., Gutkowska, J., Bilodeau, J., Lazure, C., Seidah, N.G., Chretien, M., Genest, J. and Cantin, M. (1987) B&hem. J. 241, 265-272. Vuolteenaho, O., Ajamaa, 0. and Ling, N. (1985) Biochem. Biophys. Res. Commun. 129, 82-88. Willingham, M.C. and Pastan, I. (1984) Int. Rev. Cytol. 92, 51-92.