C-type natriuretic peptides stimulate chloride in the rectal gland of Squalus acanthias R. SOLOMON,

A. PROTTER,

G. MCENROE,

J. G. PORTER,

secretion

AND P. SILVA

New England Deaconess Hospital and Beth Israel Hospital, Boston, Massachusetts 02215; SCIOS, Mountain View, California 94043; and Mount Desert Island Biological Laboratory, Salsbury Cove, Maine 04672 Solomon, R., A. Protter, G.McEnroe, J. G. Porter, and P. Silva. C-type natriuretic peptides stimulate chloride secretion in the rectal gland of Squab acanthias. Am. J. Physiol. 262 (Regulatory Integrative Comp. Physiol. 31): R707R711, 1992.-Homologous shark C-type natriuretic peptide (sCNP) wasinfused asa bolusand asa constant infusion in the isolated perfused rectal gland of the same species,Squalus acanthias. sCNP was a potent stimulator of chloride secretion similar in its dose-responsecurve to vasoactive intestinal peptide. sCNP was equipotent with killifish CNP but more potent than human CNP (hCNP). Truncated and substituted, forms of hCNP were also capable of stimulation of chloride secretionin the order hCNP > hCNP (6-22) = [Glyg] hCNP > hCNP-( 7- 21). sCNP wasmorepotent than human atria1natriuretic peptide (hANP), which was more potent than porcine brain natriuretic peptide. hANP-(31-67) was without effect. These studiessuggestthat sCNP may be the physiological regulator of rectal gland function. The receptor in the rectal gland is unknown but basedon the order of potencies,position 4 in the NH,-terminal end and the ring itself are important for ligand effects. vasoactive intestinal peptide; atria1 natriuretic peptide; brain natriuretic peptide THERE HAS BEEN considerable

interest in the role of natriuretic peptides in fish osmoregulation (3) Although the majority of this work has focused upon the role of atria1 natriuretic peptides (ANPs), species-specific ANP has only recently been identified, and therefore most of the physiological studies have been conducted using heterologous (rat or human) peptides. In addition, it has recently been appreciated that other natriuretic peptides, e.g., brain natriuretic peptide (BNP) and C-type natriuretic peptides (CNPs), are also present in mammalian species and may contribute to the physiological function of this system. Both BNP and CNP have vasorelaxant and natriuretic properties similar to ANP (7,14,15). BNP is synthesized in both heart and brain, with higher tissue levels in the heart (17). CNP mRNA has only been found in the brain of mammalian and amphibian species to date (1, 7, 14, 19). The brain content of CNP is far greater than that of either ANP or BNP, and this suggests that CNP may be a neurohormone within the central nervous system and may be involved in the regulation of salt and water balance. Of particular interest, CNPs, which are the most highly conserved of the natriuretic peptides, have recently been identified in two species of dogfish, 0363-6119/92

canicula (16) and Squalus acanthias (10). In both cases, the peptide or its message was isolated from heart rather than brain, suggesting that the peptide may circulate as a natriuretic hormone. Using the homologous CNP for S. acanthias (sCNP), we performed studies in the isolated perfused rectal gland of this species to evaluate the physiological role of this peptide. The isolated perfused rectal gland is a model for active chloride transport (2) that can be stimulated to secrete chloride by exogenous (rat or human) ANP (12) or vasoactive intestinal peptide (13). In addition, we separately perfused the glands with human CNP (hCNP), killifish CNP (kCNP), porcine BNP (pBNP), human BNP (hBNP), human preproANP [hANP-(31-67)], and fragments and substitutions of the CNP peptide [ hCNP-(6-22), Glyg]hCNP-(7-21)) to evaluate the requirements for ligand-receptor interaction.

Scyliorhinus

METHODS Rectal glandswere removedfrom live dogfishafter segmental transection of the spinal cord. The rectal gland artery, vein, and duct were immediately cannulatedwith PE-90 tubing, and perfusion was begun immediately with oxygenated (99% O,=l% C02) shark Ringer solution at 15°Cand a pressureof 40 mmHg. The shark Ringer solution contained (in mM) 290Na, 5 K, 290 Cl, 8 bicarbonate, 2.5 Ca, 1.2 Mg, 1 phosphate,0.5 sulfate, and 350 urea. The pH was 7.6, and glucose(5 mM) wasusedasthe soleexogenoussubstrate.The rectal gland wasplacedin a glass perfusion chamberkept at 15 &l”C by running seawater.Rectal gland secretion was collected at lo-min intervals in calibrated micropipettes or tared microcentrifuge tubes. Rectal gland venouseffluent was collected for similar timed periods. The peptides hANP, hANP-(31-67), and BNP were obtained from Peninsula Laboratories, kCNP was kindly supplied by Dr. D. Evans, and sCNP, hCNP, hCNP-(6-22), hCNP-(7-21), and [Glyg]hCNP were suppliedby SCIOS. Peptides were kept frozen until immediately before use,when they were diluted into 1 ml of shark Ringer solution. After three baselinecollection periods of 10 min each, the peptides were infused over 60 s into the perfusate line. Three additional lomin collectionswerethen performed. For eachdosein the doseresponsecurve, a minimum of 3 and as many as 14 separate perfusionswereperformed. In separateexperiments,a constant infusion of peptide wasadministeredfor 30 min. In theseexperiments, 0.1% albumin was addedto the perfusate to minimize lossof peptide to the glasswareand tubing.

$2.00 Copyright 0 1992 the American Physiological Society

R707

Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 18, 2019.

R708

CNP STIMULATES

THE RECTAL

Chloride concentration of the rectal gland secretionwasmeasured in a Buchler Cotlove chloridometer, and the chloride secretory rate was calculated as the product of the chloride concentration and the duct flow rate and expressed as microequivalentsper hour per gram wet weight. RESULTS

Effect

on chloride secretion. The

of sCNP-(1 -22)

effect of intra-arterial bolus injections of sCNP on chloride secretion by the shark rectal gland is shown in Fig. 1. A small stimulatory effect was seen at a concentration of 10S8 M. The maximal stimulation of chloride secretion observed with sCNP was seen at a concentration of 10S6 M. Because of limits in the amount of peptide available, concentrations >10V6 M were not tested. Within the limits of the concentrations of sCNP examined, half-maximal stimulation appears to be at 5 x loss M. Chloride secretion was similar at 10m7M vasoactive intestinal peptide and 10V6 M sCNP. Because of limited supplies of peptide, these doses may not represent maximal secretory rates. Figure 2 shows the effect of continuous infusions of sCNP on chloride secretion in the rectal gland. A large effect was observed at lo- lo M, followed by an even larger but not yet maximal effect at 10eg M. Again, limited amounts of peptide prevented a more extensive examination of the dose-response relation. Comparative effects of different CNP peptides. Figure 3 compares the effect of CNPs from different species on chloride secretion in the rectal gland. The effect of sCNP is similar to that of kCNP and is almost an order of magnitude greater than that of hCNP.

G 2 2

2200

r

2000

-.

1800 1600

I

I

I

I

I

I

0 VIP 0

sCNP

-

b i!i 4

-

600

-

400

-

Tii 1

800

g 2

700

ld d

600

g -PI 3

500

; -a tn a

400

‘E

200

;ti t-i

a-

8

t

,I

t 1

1I

300

1

T / I/

I

100 t

-

0’ lo-l3

lo-l2

10’”

lo-‘O

10-O

lo--*

sCNP [M] Fig. 2. Effect of a constant infusion of sCNP on chloride secretion by rectal gland. sCNP was given sequentially at increasing concentrations to 5 isolated perfused rectal glands. Perfusion at each concentration was maintained for 30 min. Significant stimulation of chloride secretion was observed at concentrations X0- lo M. Concentrations >lOBg M were not tested because of limits in the amount of peptide available. Each point represents mean t SE. 2000

-

1800

-

? d 1600 > z 1400 b ‘i 1200 W

0 v v

sCNP kCNP hCNP

-

1000

-

800

-

600

-

400

-

200

-

u

-

800

r

900

%i b : ; .d 23

-

OF THE SHARK

1000

lg

-

z0 1400 -c( 8 - 1200 d ; 1000

GLAND

I

01 1o-12

1o-”

A

1o-‘O

I

I

I

1o-9

1o-8

1o-7

I

lo-6

1

,o-5

l r(

zk u

200-

CNP t

,j”

,

,

1o-7

1o-6

i

0 ,o-12

1o-ll

1o-1O

1o-9

Peptide

1o-8

1o-5

[M]

Fig. 1. Effect of shark C-type natriuretic peptide (sCNP) and vasoactive intestinal peptide (VIP) on secretion of chloride by rectal gland. Both peptides were administered intra-arterially, as boluses, over a period of 1 min in amounts calculated to deliver the desired concentration to the gland. Stimulatory effect of sCNP was similar to that of VIP at concentrations w

1400

0 b --

1200

d

!! Q 4 .Pl g

ANP

0 pBNP + hBNP H pre-pro

t

!J g .r( 3

0

t

ANP

I 1000 t

800

-

600

-

400

-

200

-

8

0 ,o-12

,o-1l

,o-10

,o-9

,o-8

,o-7

,o-6

1o--‘2 1o-l1

,o-5

1o-lO

1o-Q

Io-8

1o-7

Io-8

Io-S

NP W

hCNP [M] Fig. 4. Effect of modifications of hCNP structure on chloride secretion. Peptides were given as described in Fig. 1 legend. Removal of NH2terminal end [hCNP-(6-22)] shifts dose-response curve to the right, as does substitution of leucine in position 9 by glycine ( [Glyg] hCNP). Opening the ring [ hCNP- (7 - 2 l)] substantially reduces stimulatory effect but does not abolish it. Total number of observations for hCNP(6-22), [Glyg]hCNP, and hCNP-(7-21) was 38, 40, and 36, respectively. Each point represents mean of:SE.

Fig. 5. Comparison of effect of sCNP with other natriuretic peptides (NPs). Peptides were given as described in Fig. 1 legend. Effect of sCNP was at least 2 orders of magnitude greater than that of human atria1 natriuretic peptide-( l-28) (hANP) or porcine brain natriuretic peptide BNP (pBNP). Human BNP (hBNP) and preproANP-(31-67) had no effect. Total number of observations for hANP, pBNP, hBNP, and preproANP-(31-67) was 38, 18, 26, and 19, respectively. Each point represents mean t SE.

to the right, although the maximal stimulatory effect is similar to that of hCNP. Substitution of the leucine in position 9 by a glycine also modestly reduced the functional effect of the molecule ( [Glyg] hCNP vs. hCNP). The ringless structure [hCNP-(7-21)] made up of amino acids 7 to 21 had a minimal stimulatory effect. Functional effects of different natriuretic peptides. Figure 5 compares the effect of sCNP with that of hANP, pBNP, hBNP, and hANP-(31-67). As can be seen, the effect of sANP is considerably greater than that of hCNP and pBNP. hBNP and hANP-(31-67) had virtually no effect.

less effective than ANP in inducing diuresis when injected into the femoral vein of anesthetized rats (17). The effect of the l-22 peptide has not been examined. The results reported above represent the first demonstration that CNP-( 1- 22) stimulates electrolyte transport in a secretory epithelium. These results support a physiological effect of sCNP to provoke salt secretion by the shark rectal gland. sCNP is a 22-amino acid peptide containing a W-member ring on the COOH-terminal end. Although sCNP has beenisolated from the shark heart, evidence for its presence in the circulation is lacking. Until such evidence is obtained, a definitive role in the- control of rectal gland function remains speculative. However, the steep dose-response curve with stimulation occurri ng at lo- 10 M level during continuous perfusion strongly suggests a physiologically important ligand. The continuous infusion protocol is most analogous to in vitro models involving incubation of tissue with ligand. In such in vitro studies, concentrations

DISCUSSION

To date, the natriuretic properties of only hCNP(l-53) and pCNP-( l-53) have been examined [these peptides differ in two amino acids, whereas CNP-( 1 - 22) peptide is identical in human, pig, and rat]. Both peptides were found to be equipotent and two orders of magnitude C-1

.

. nat.rwebc

.

-tIdeI

.

S-L-R-R-S-S-C-F-G-G-R-M-D-R-I-G-A-Q-S-G-L-G-C-N-S-F-R-Y-COOH

. . . (brw . nat.ruet+c, neDt+ge) S-P-K-T-M-R-D-S-G-C-F-G-R-R-L-D-R-I-G-S-L-S-G-L-G-C-N-V-L-R-R-Y-COOH PORCINE S-P-K-M-V-Q-G-S-G-C-F-C-R-K-M-D-R-I-S-S-S-S-G-L-G-C-K-V-L-R-R-H-COOH 12

34

5

6

7

8

9

10

11 12

13 14

15

16 17

18 19 20

21 22

G-L-S-K-G-C-F-G-L-K-L-D-R-I-G-S-M-S-G-L-G-C-COOH

KILLIFISH DOGFISH

G-W-N-R-G-C-F-G-L-K-L-D-R-I-G-S-M-S-G-L-G-C-COOH G-P-S-R-S-C-F-G-L-K-L-D-R-I-G-A-M-S-G-L-G-C-COOH

Fig. 6. Amino acid sequence of shark, killifish, and human CNP, human ANP, and human and porcine BNP.

Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 18, 2019.

R710

CNP STIMULATES

THE RECTAL

>10S8 M were necessary before an effect of CNP was observed (4,5,8). The finding of a physiological effect at lo-lo M of CNP is therefore particularly compelling. The mechanism by which sCNP stimulates chloride secretion is presently unknown. Evidence for both direct and indirect effects on the secretory epithelium has been presented by us and other investigators (6, 11). It is known, however, that the hCNP interacts preferentially with the human ANPR-B receptor, whereas ANP interacts preferentially with the human ANPRrA receptor. BNP is capable of interacting with both the receptor subtypes. Both receptors are coupled to guanylate cyclase and result in the accumulation of guanosine 3’,5’-cyclic monophosphate within target cells (8). Whether similar ligand-receptor specificities exist in other species is unknown. kCNP and sCNP differ in three amino acids (Fig. 6) in the NH&erminal end, positions 2, 3, and 5, and in position 16 in the ring structure. Both peptides had similar dose-response curves, suggesting that these amino acid differences were not important. On the other hand, hCNP differs from sCNP at positions 2,4,5, and 16 and from kCNP at positions 2-4 and is one order of magnitude less potent than sCNP or kCNP. This suggests that position 4 in the NHa-terminal end of the molecule is of functional significance. Further support for the importance of the NH2-terminal end of the molecule is suggested by the marked reduction in response to hCNP(6-22) containing the ring structure alone compared with intact hCNP. In addition to the importance of the NHz-terminal end of the molecule, the ring may also be of some functional significance. The linear fragment, hCNP- (7 - 2 1), and the intact ring, hCNP-(6-22), were still capable of stimulating chloride secretion albeit at markedly blunted maximal levels and only at very high concentrations (~5 x 10e7 and > 10V7 M, respectively) of peptides. In bovine aorta smooth muscle cells that contain the ANPR-B receptor subtype, hCNP and hCNP-(6-22) have similar dose-response curves for the generation of cGMP. The linear fragment, hCNP-(7-21), on the other hand, is inactive (A. Protter, personal communication). The substitution of glycine for leucine at position 9 in the ring also reduced the stimulatory effect ( [Glyg]hCNP vs. hCNP), suggesting a functional role for this site as well. In the bovine aorta smooth muscle preparation, [Glyg] hCNP is also inactive (A. Protter, personal communication). Both hANP-(31-67) and pBNP are known to circulate in mammals and therefore may represent additional natriuretic hormones (9, 18). Both peptides appear to have vasorelaxant and natriuretic effects in mammals (7, 14, 15). BNP interacts primarily with the ANPR-A receptor. In the isolated perfused rectal gland it has a dose-response curve shifted well to the right of that of hANP. It is not known which receptor is stimulated by hANP-(31-67) in mammals. Little effect was seen with hANP-(31-67) in the rectal gland. Taken together with the results obtained with the various CNPs. we sug-

GLAND

OF THE SHARK

gest that the biologically active receptor for sCNP in shark rectal gland resembles the ANPR-B receptor. The authors express their appreciation to Heather Brignull, Sonja Hornung, Judd Landsberg, Marcela Silva, Hadley Solomon, Melissa Taylor, and Doug Wolf, who performed the majority of the rectal gland perfusions. The work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grant DK-18078. Address for reprint requests: R. Solomon, 110 Francis St., Suite 9D, Boston, MA 02215. Received 21 October 1991; accepted in final form 23 January 1992. REFERENCES 1. Arimura, J. J., N. Minamino, K. Kangawa, and H. Matsuo. Isolation and identification of C-type natriuretic peptide in chicken brain. Biochem. Biophys. Res. Commun. 174: 142-148, 1991.

2. Epstein, F. H., J. S. Stoff, and P. Silva. Mechanism and control of hyperosmotic NaCl-rich secretion by the rectal gland of Squalus acanthias. J. Exp. Biol. 106: 25-41, 1983. 3. Evans, D. H. An emerging role for a cardiac peptide hormone in fish osmoregulation. Annu. Rev. Physiol. 52: 43-60, 1990. 4. Furuya, M., M. Takehisa, Y. Minamitake, Y. Kitajima, Y. Hayashi, N. Ohnuma, T. Ishihara, N. Minamino, K. Kangawa, and H. Matsuo. Novel natriuretic peptide, CNP, potently stimulates cyclic GMP production in rat cultured vascular smooth muscle cells. B&hem. Biophys. Res. Commun. 170: 201-208, 1990.

5. Furuya, M., M. Yoshida, Y. Hayashi, N. Ohnuma, N. Minamino, K. Kangawa, and H. Matsuo. C-type natriuretic peptide is a growth inhibitor of rat vascular smooth muscle cells. Biochem. Biophys. Commun. 177: 927-931, 1991. 6. Karnaky, K. J., Jr., J. D. Valentich, M. Currie, and W. Oehlenschlager. Atriopeptin stimulates chloride secretion by cultured shark (Squab acanthias) rectal gland epithelium. Bull. Mt. Desert Isl. Biol. Lab. 29: 86-87, 1990. 7. Kojima, M., N. Minamino, K. Kangawa, and H. Matsuo. Cloning and sequence analysis of a cDNA encoding a precursor for rat C-type natriuretic peptide (CNP). FEBS Lett. 276: 209-213, 1990.

8. Koller, K. J., D. G. Lowe, G. L. Bennett, N. Minamino, K. Kangawa, H. Matsuo, and D. V. Goeddel. Selective activation of the B natriuretic peptide receptor by C-type natriuretic peptide (CNP). Science Wash. DC 252: 120-123, 1991. 9. Mukoyama, M., K. Nakao, Y. Saito, Y. Ogawa, K. Hosoda, S-I. Suga, G. Shirakami, M. Jougasaki, and H. Imura. Increased human brain natriuretic peptide in congestive heart failure (Abstract). N. Engl. J. Med. 323: 757, 1990. 10. Schofield, J. P., D. S. C. Jones, and J. N. Forrest, Jr. Identification of C-type natriuretic peptide in heart of spiny dogfish shark (Squab acanthias). Am. J. Physiol. 261 (Renal Fluid Electrolyte Physiol. 30): F734-F739, 1991. 11. Silva, P., J. S. Stoff, R. J. Solomon, S. Lear, D. Kniaz, R. Greger, and F. H. Epstein. Atria1 natriuretic peptide stimulates salt secretion by shark rectal gland by releasing VIP. Am. J. Physiol. 252: (Renal Fluid Electrolyte Physiol 21): F99F103,

1987.

Solomon, R., M. Taylor, D. Dorsey, P. Silva, and F. H. Epstein. Atriopeptin stimulation of rectal gland in SquaZus acanthias. Am. J. Physiol. 249 (Regulatory Integrative Comp. Physiol. 18): R348-R354, 1985. 13. Stoff, J. S., R. Rosa, R. Hallac, P. Silva, and F. H. Epstein. Hormonal regulation of active chloride transport in the dogfish rectal gland. Am. J. Physiol. 237 (Renal Fluid Electrolyte Physiol. 12.

6): Fl38-F144,

1979.

14.

Sudoh, T., K. Kangawa, N. Minamino, and H. Matsuo. A new natriuretic peptide in porcine brain. Nature Lond. 332: 78-81,

15.

Sudoh, T., N. Minamino, K. Kangawa, and H. Matsuo. C-type natriuretic peptide (CNP): a new member of natriuretic peptide family identified in porcine brain. Biochem. Biophys. Res. Commun. 168: 863-870, 1990.

1988.

Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 18, 2019.

CNP STIMULATES

THE RECTAL

R., R. Takahashi, N. Hazon, and Y. Takei. Isolation of high-molecular-weight C-type natriuretic peptide from the heart of a cartilaginous fish (European dogfish, Scyliorhinus canicula). FEBS Lett. 282: 321-325, 1991.

16. Suzuki,

17. Tawaragi, Kangawa,

Y., K. Fuchimura, and H. Matsuo.

human C-type natriuretic Commun. 175: 645-651, 1991.

S. Tanaka,

N. Minamino,

K.

Gene and precursor structures of peptide. Biochem. Biophys. Res.

GLAND

R711

OF THE SHARK

18. Winters, C. J., A. L. Sallmann, J. Meadows, D. M. Rico, and D. L. Vesely. Two new hormones: prohormone atria1 natri-

uretic peptides l-30 and 31-67 circulate in man. Biochem. Biophys. Res. Corm-nun. 150: 231-236, 1988. 19. Yoshihara, A., and H. Matsuo.

C-type natriuretic 591-598,

H.

Kozawa,

N.

Minamino,

K.

Kangawa,

Isolation and sequence determination of frog peptide. Biochem. Biophys. Res. Commun. 173:

1990.

Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 18, 2019.