Distribution and function of brain natriuretic peptide in the stomach and small intestine of the rat

Regulatory Peptides, 34 (1991) 61-70

61

© 1991 Elsevier Science Publishers B.V. 0167-0115/91/$03.50 REGPEP 01050

Distribution and function of brain natriuretic peptide in the stomach and small intestine of the rat Keith A. Sharkey, D. G r a n t Gall and Wallace K. M a c N a u g h t o n Neuroscience and Gastrointestinal Research Groups, University of Calgary, Calgary, Alberta (Canada)

(Received 4 December 1990; revised version received 5 March 1991; accepted 11 March 1991) Key words: Immunohistochemistry; Electrolyte transport; Chloride secretion;

Natriuretic peptide; Enteric nervous system

Summary The distribution and function of brain natriuretic peptide (BNP) was studied in the rat stomach and jejunum. B NP-like immunoreactive nerves were found in the myenteric plexus, circular muscle, submucosa and in the crypt region of the jejunum. In the stomach, BNP-like immunoreactivity was found in the myenteric plexus, circular muscle, submucosa and at the base of the gastric glands. In the submucosa, BNP-like immunoreactivity was often associated with blood vessels. In segments of rat jejunum mounted in Ussing chambers, serosal exposure to rat BNP caused a concentrationdependent increase in short circuit current. A maximal effect of 18 + 4 itA/cm 2 was observed with 1 # M BNP. The effect was quantitatively and qualitatively similar to that elicited by serosal exposure to equimolar atrial natriuretic peptide. The response to BNP was reduced by 88 ~o in chloride free Kreb's buffer, by 83 ~o in tissues pretreated with cinanserin, an antagonist of the 5-HT; subtype of the 5-hydroxytryptamine receptor, and by 96 To in tissues pretreated with tetrodotoxin, a blocker of axonal conduction. These results are consistent with a physiological role for BNP as a neuromodulator of gastrointestinal electrolyte transport.

Correspondence: K.A. Sharkey, Neuroscience and Gastrointestinal Research Groups, University of Calgary, 3330 Hospital Dr. N.W., Calgary, Alberta, Canada T2N 4NI.

62 Introduction

Brain natriuretic peptide (BNP) is a novel natriuretic peptide isolated first from porcine brain [ 1] and a homologous peptide has subsequently been isolated from rat brain and heart [2-4]. Cloning and sequence analysis ofcDNA encoding the rat, human and porcine forms of BNP reveal that it is a member of a well conserved family of peptides having a high degree of structural homology in a C-terminal 32 amino acid sequence that includes a ring structure formed by Cys-Cys disulphide bond [5,6]. BNP is also structurally homologous to atrial natriuretic peptide (ANP) and appears to share many pharmacological properties, suggesting that it acts through the same receptor [7-9]. Binding sites for ~25I-ANP have been demonstrated in the rat duodenum, jejunum and ileum [10,11] and immunoreactive material corresponding to ANP has been identified in guinea-pig and rat stomach, small intestine and colon [12,13 ]. Recently, Hardin et al. [ 14] demonstrated an effect of ANP on electrolyte transport in the rat small intestine. In those studies, ANP induced diuresis and natriuresis when injected intravenously and, when added to jejunum and ileum in Ussing chambers, increased short circuit current through a mechanism that resulted in an electrogenic chloride secretion [ 14]. However, the distribution and function of BNP has not been investigated in the rat gastrointestinal tract. In this study we have used immunohistochemistry to examine the distribution of BNP-like immunoreactivity in the stomach and jejunum and Ussing chambers to assess its effects on ion transport in the jejunum in vitro.

Materials and Methods

Animals Outbred female Hooded-Lister rats (150-200 g) were obtained from a breeding colony maintained at the University of Calgary. They were housed under constant photoperiod (14 h light, 10 h dark) and temperature (20 ° C), and were allowed food and water ad libitum. All procedures were carried out in accordance with the guidelines established by the Canadian Council on Animal Care.

Immunohistochemistry Rats were killed and the stomach and jejunum were removed. Tissues were washed in phosphate-buffered saline (PBS), fixed by immersion in Zamboni's fixative [15], washed in dimethylsulpboxide (3 × 10 rain) and PBS (3 x 10 min), and finally in PBS containing 20~o sucrose. The tissues were then sectioned (15 #m) in a cryostat and then processed for indirect immunofluorescence. Sections were washed in PBS containing 0.1Fo Triton X-100 for 30 min at room temperature, incubated in the primary anti BNP-antibodies (1 : 500, Ref. 16) or the same antibodies pre-absorbed with rat BNP (1 or 10 nmol/ml, Bachem) or rat ANP (10 nmol/ml, Sigma) for 24-48 h at 4 °C in a moist chamber, and were washed (3 × 10 min) in PBS and incubated with secondary antibody (swine anti-rabbit IgG conjugated to FITC, Dako) for a further 1 h at room temperature. Finally, they were washed in PBS containing 0.1~o Triton X-100

63 (3 x 10 min) and mounted in bicarbonate-buffered glycerol (pH 8.6). Sections were examined using a Leitz Orthoplan fluorescence microscope fitted with filter system 12. Photographs were taken with Kodak TMax 400 ASA film.

Effects on jejunal electrolyte transport in vitro Experiments were conducted to determine the effects of BNP on jejunal electrolyte transport in vitro. Rats were killed and a 10 cm segment of jejunum was removed beginning at the ligament of Treitz and gently flushed with cold (4 ° C) Kreb's buffer. The Kreb's buffer was composed of (mM): NaC1 (115.0), KH2PO 4 (2.0), MgCI z (2.4), CaC12 (1.3), N a H C O 3 (25.0) and KC1 (8.0). The external muscle layers, including the myenteric plexus, were removed by blunt dissection and 1.5 to 2.0 cm segments, devoid of Peyer's patches, were mounted between two halves of a standard Ussing flux chamber. The serosal side of the tissue was bathed with Kreb's buffer containing 10 mM glucose, while the mucosal side was bathed with Kreb's buffer containing 10 mM mannitol. Solutions were maintained at 37 °C and were aerated and circulated with a gas lift system (95 ~ 02, 5 ~o CO2). The electrical potential difference (PD) across the tissue was maintained at 0 V using a voltage clamp apparatus (World Precision Instruments). The short circuit current (Isc) required to maintain zero PD was recorded as the indicator of active electrolyte transport. After mounting, tissues were allowed to equilibrate until a stable baseline Isc was recorded. The serosal side of the tissue was then exposed to rat BNP (10-7, 5 . 1 0 - 7 or 10- 6 M ; B achem), ANP (10- 6 M; Sigma) or vehicle (0.0025 ~o (v/v) acetic acid). All concentrations are the final concentration of the peptide or the vehicle in the bathing solution. The Isc change which occurred in response to the peptide was recorded. It has been previously shown that the jejunal Isc response to rat ANP is chloride dependent and can be inhibited by the 5-HT 2 receptor antagonist, cinanserin [ 17]. To determine if the mechanism of BNP-induced electrolyte transport was similar to that of ANP, the following experiments were conducted. In one series of experiments, tissue was mounted in Kreb's buffer in which chloride ion was replaced with equimolar isethionate (sodium salt) and gluconate (magnesium, calcium and potassium salts). After a 15 min equilibration period, the serosal side of the tissue was exposed to 5 . 1 0 - 7 M BNP and the Isc response recorded. In another series of experiments,jejunal segments were mounted in normal Kreb's buffer as described above and allowed to equilibrate for 10 min. Pairs of tissues were matched on the basis of basal conductance (conductances within 20~o), with one member of the pair exposed serosally to either 10 - 5 M cinanserin (Squibb) or 10 - 6 M tetrodotoxin (Sigma). The other member of the pair was exposed to the vehicle (100 #1 distilled H 2 0 ). After a 15 min incubation, the serosal side of each tissue was exposed to 5 . 1 0 - 7 M BNP and the Isc response recorded. Statistics Dose-response data were compared by one-way analysis of variance followed by the Newman-Keuls test. Comparison of two groups was made by Student's t-test for paired or unpaired data where appropriate. An associated probability (P) value of less than 0.05 was considered to be significant.

64

Fig. 1. Fluorescence micrographs of BNP-like immunoreactivity (BNP-LI) in the rat jejunum ( A - C ) and stomach (D). In the jejunum, BNP-LI was largely confined to the myenteric plexus, circular muscle, submucosa (associated with blood vessels), crypt region and the base of the villi (A-C). In the stomach, relatively sparse BNP-LI was found in the myenteric plexus, circular muscle, submucosa (associated with blood vessels) and at the base of the gastric glands (D). lm, longitudinal muscle; cm, circular muscle; sin, submucosa; m, mucosa (crypt region); v, villi. Scale bar: 50 pm.

65

Fig. 2. Fluorescence micrographs of BNP-LI in the rat stomach. (A) Staining of BNP-immunoreactive nerves was completely abolished by pre-absorption of the antibodies with BNP (1 nmol/ml), however, immunoreactivity was unaffected by pre-absorption with rat ANP (10 nmol/ml). (B)In the stomach, BNP-LI was mostly found associated with blood vessels in the submucosa, cm, circular muscle; sm, submucosa; m, mucosa. Scale bar: 50/~m.

66 Results

Immunohistochemistry Brain natriuretic peptide (BNP)-like immunoreactivity was found in nerve fibres throughout the wall of the stomach and jejunum (Figs. 1 and 2). Higher densities of immunoreactive nerves were found in the jejunum compared to the stomach (Fig. 1). BNP-like immunoreactivity was abolished in sections stained using antisera preabsorbed with BNP (1 or 10 nmol/ml) but staining was unaffected by pre-absorption with rat ANP (10 nmol/ml) (Fig. 2A). In the jejunum, BNP-like immunoreactivity was found in the myenteric, circular muscle, submucosa, crypt region and base of the villi (Fig. 1). In the stomach, occasional fibres in the longitudinal muscle showed BNP-like immunoreactivity with denser staining in the myenteric plexus, circular muscle, submucosa and base of gastric glands. In both tissues BNP-like immunoreactivity in the submucosa was largely associated with blood vessels (Fig. 2B). Immunoreactive material was not distributed in other cells in the stomach or jejunum of the rat. However, BNP-like immunoreactivity was found only in entero-endocrine-like cells in human duodenal villi (taken by endoscopic biopsy, data not shown). Effects on jejunal electrolyte transport in vitro Serosal exposure of the rat jejunum to the rat sequence of BNP resulted in an increase in Isc within 1 min (Fig. 3). The Isc response peaked within 2 min and, in most cases, remained elevated above baseline for several minutes. The vehicle (0.0025 ~o acetic acid) had no effect on baseline Isc. The Isc response to BNP was concentration-dependent (Fig. 4), reaching a maximum at 5 . 1 0 - 7 M (16 + 4 #A/cm2). The peak responses to 10 - 6 M BNP and 10 - 6 M ANP were similar (18 + 4 and 17 + 9 #A/cm 2, respectively; Fig. 4). The time to reach the peak Isc response after the addition of peptide was also similar for BNP (1.8 + 0.1 min) and ANP (1.7 + 0.1 min). The Isc response to 5 . 1 0 - 7 M BNP was significantly reduced by 88~o in chloridefree Kreb's buffer ( P < 0.05; Fig. 5A). Pretreatment of the tissue with the 5-HT 2 receptor antagonist, cinanserin (10 - 5 M), reduced the Isc response to 5 . 1 0 - 7 M BNP by 83 + 17% (P < 0.05; Fig. 5B). This concentration of cinaserin was shown to reduce the Isc increase caused by serosal application of 10 -5 M 5-HT by 78~o (data not shown). Pretreatment with the neural blocker, TTX (10 -6 M), almost abolished the responses both to 5 . 1 0 - 7 M BNP (96 + 4~o inhibition, P < 0.05; Fig. 5C) and

,oo] ~'Icm2 63

A BNP

2 min

Fig. 3. Representativetrace of the short circuit current (Isc) response to the serosal exposure of the rat jejunum to 5.10- 7 M rat BNP (arrowhead). BNP was added followinga 15 min equilibrationperiod (not shown).

67 3O 25 20

a.

_L

II

Ol

0.1

0.5

[BNP]

1.0

1.0

[ANP] OzM)

Odd)

Fig. 4. Concentration-response relationship o f B N P (n = 4 rats; open bars) and A N P (n = 4; hatched bar). The peak change in short circuit current (Isc) which occurred following exposure to BNP was maximal at 5- 10 - 7 M. The response to BNP was quantitatively similar to that elicited by ANP. *P < 0.05 compared to other groups. All groups were significantly different from vehicle control, which produced no change in Isc. Z5

20 15

i 10-

~en

5-

0 25-

I 0 ~

Normal Krebs

e l - Free Krebs

20

~

E

~

N

m ?~ < w n

i

'~t

B

0

25-

Vehicle

Cinanserin

20. 15,

10. 5.

i

0 Vehicle

TTX

Fig. 5. (A) Effect of chloride replacement on the short circuit current (Isc) change caused by exposure to 5 • 10 - 7 M rat BNP. n = 4; *P < 0.05 by t-test for unpaired data. (B) Effect of pretreatment of tissue with the 5-HT2 receptor antagonist, cinanserin ( 1 0 - 5 M), on the Isc response induced by 5' 10-7 M rat BNP. n = 4; *P < 0.05 by t-test for paired data. (C)Effect of pretreatment of tissue with T T X ( 1 0 - 6 M) on the Isc response to 5- 10 - 7 M rat BNP. n = 3; *P < 0.05 by t-test for paired data. The change in Isc is expressed as the difference from basal Isc in the presence of the antagonist immediately prior to exposure to BNP and the peak response to BNP.

68 5 • 10 7 M ANP (92 + 6~o inhibition, P < 0.05). Neural blockade by this concentration of TTX was complete, as shown by the absence of a response to BNP in tissues pretreated with TTX and subsequently subjected to electrical field stimulation (30 V at a pulse duration of 500 #s and frequency of 10 Hz for a total duration of 1 s, data not shown). Both cinanserin and TTX caused a slight decrease in basal Isc when added to the serosal bathing solution.

Discussion In the present study we have demonstrated the presence of BNP-like immunoreactive nerves in the stomach and jejunum of the rat. In an in vitro bioassay we showed that BNP stimulated a chloride-dependent electrogenic ion transport that was apparently mediated through activation of a receptor that caused 5-HT release and subsequent activation of a 5-HT 2 receptor. The intestine plays an important role in salt and water homeostasis and would, therefore, be an important site of action of natriuretic peptides. Indeed, binding sites for ANP have been identified in the jejunum, particularly in fibroblast-like cells in the lamina propria and at the bases of mature enterocytes [10]. This configuration is compatible with an electrolyte transport function for ANP. Since it has been suggested that ANP and BNP react with equal potencies at ANP binding sites [7-9], it is possible that BNP localized in the gut may also play a role in intestinal function, particularly electrolyte transport. The recent demonstration of BNP binding sites in the gut [ 18] further supports this hypothesis. The nature of the material cross-reacting with the BNP antibodies used in this study is not yet clear. In our liquid-phase preabsorption controls it was apparent that BNP, but not rat ANP, could completely abolish staining of immunoreactive nerves. That the cross-reactivity is not ANP is further supported by studies in which the localization of ANP has been examined directly [ 12]. These studies show that ANP-like immunoreactivity is localized in a population of cells with the morphological characteristics of plasma cells and not to any neural tissues. However, using radioimmunoassay (RIA) two groups have been unable to detect BNP-like immunoreactivity in the rat gastrointestinal tract [2,3]. Although these results conflict with our own it was apparent that the RIAs could not detect BNP in many regions of the rat brain where it has been localized by immunohistochemistry [ 16]. Hence, it is possible that two (or more) forms of BNP-like material are present in the rat, an atrial form and a neural form. This is supported by recent evidence for two C-type natriuretic peptides found in porcine brain [19,20]. These peptides are homologous with ANP and BNP and may represent the material identified by immunohistochemistry in this study. We have previously shown that ANP can stimulate electrogenic chloride secretion in the rat jejunum in vitro [14,17]. This effect has been shown to be dependent upon stimulation of 5-HT2 receptors, since it could be blocked by the 5-HT 2 receptor antagonist, cinanserin. Evidence exists for 5-HT 2 binding sites in the rodent small intestine [21]. In the present study, rat BNP elicited a transport response in the rat jejunum in vitro. Several lines of evidence suggest that this effect is mediated through

69 the same mechanism as that of A N P . First, the responses to 1 # M B N P and 1 # M A N P were quantitatively and qualitatively similar with respect to short circuit current. Second, the Isc response to B N P was chloride dependent, since it was reduced in tissue bathed with chloride-free Kreb's buffer. Third, the Isc response to B N P was significantly inhibited by pretreatment o f the tissue with cinanserin at a concentration which significantly inhibited the Isc response to serosal application of 10- 5 M 5-HT. Finally, the Isc responses to both B N P and A N P were significantly reduced by pretreatment with TTX, indicating a neural mechanism of action in stimulating electrolyte transport. In conclusion, this study provides the first link between the localization o f a natriuretic peptide in gastrointestinal nerves and its role in intestinal physiology. This work provides the impetus for further characterization of the role o f B N P in gastrointestinal physiology, and its potential contribution to transport anomalies in intestinal disease.

Acknowledgements This work was supported by grants from the Alberta Heritage Foundation for Medical Research ( K A S ) and the Medical Research Council of C a n a d a ( D G G ) . W . K . M . is the recipient o f postdoctoral fellowships from the Medical Research Council o f C a n a d a and the Alberta Heritage Foundation for Medical Research. The authors are very grateful to Drs. C.B. Saper and P. Needleman who provided the B N P antibody used in this study, to Dr. C.B. Saper for his helpful comments on the manuscript, and to Lorraine Oland for technical assistance.

References 1 Sudoh, T., Kangawa, K., Minamino N. and Matsuo, H., A new natriuretic peptide in porcine brain, Nature, 332 (1988) 78-81. 2 Aburaya, M., Minamino, N., Hino, J., Kangawa K and Matsuo, H., Distribution and molecular forms of brain natriuretic peptide in the central nervous system, heart and peripheral tissue of rat, Biochem. Biophys. Res. Commun., 165 (1989) 880-887. 3 Ogawa, Y., Nakao, K., Mukoyama, M., Shirakami, G., Itoh, H., Hosoda, K., Saito, Y., Arai, H., Suga, S., Jougasaki, M., Yamada, T., Kambayashi, Y., Inouye, K. and Imura, H., Rat brain natriuretic peptide. Tissue distribution and molecular form, Endocrinology, 126 (1990) 2225-2227. 4 Nakao, K., Itoh, H., Kambayashi, Y., Hosoda, K., Saito, Y., Yamada, T., Mukoyama, M., Arai, H., Shirakami, G., Suga, S., Jougasaki, M., Ogawa, Y., Inouye, K. and Imura, H., Rat brain natriuretic peptide. Isolation from rat heart and tissue distribution, Hypertension, 15 (1990) 774-778. 5 Kojima, M., Minamino, N., Kangawa, K. and Matsuo, H., Cloning and sequence analysis of eDNA encoding a precursor rat brain natriuretic peptide, Biochem. Biophys. Res. Commun., 159 (1989) 1420-1426. 6 Sudoh, T., Maekawa, K., Kojima, M., Minamino, N., Kangawa, K. and Matsuo, H., Cloning and sequence analysis of cDNA encoding a precursor for human brain natriuretic peptide, Bioehem. Biophys. Res. Commun., 159 (1989) 1427-1434. 7 Zhou, H.-L. and Fiscus, R. R., Brain natriuretic peptide (BNP) causes endothelium-independent relaxation and elevation of cyclic GMP in rat thoracic aorta, Neuropeptides , 14 (1989) 161-169. 8 Torda, T., Nazarali, A.J. and Saavedra, J.M., Brain natriuretic peptide receptors in the rat peripheral sympathetic ganglia, Biochem. Biophys. Res. Commun., 159 (1989) 1032-1038.

70 9 Ermirio, R., Avanzino, G.-L., Ruggeri, P., Bergaglio, M. and Calaresu, F.R., Cardiovascular effects of microinjection of ANF and brain natriuretic peptide into ventrolateral medulla, Am. J. Physiol., 259 (1990) R32-R37. 10 Bianchi, C., Thibault, G., DeLean, A., Genest, J. and Cantin, M., Atrial natriuretic factor binding sites in the jejunum, Am. J. Physiol., 256 (1989) G436-G441. 11 Bianchi, C., Gutkowski, J., Thibault, G., Garcia, R., Genest, J. and Cantin, M., Radioautographic localization of ~25I-atrial natriuretic factor (ANF) in rat tissues, Histochemistry, 82 (1985) 441-452. 12 Vollmar, A.M., Friedrich, A., Sinowatz, F. and Schulz, R., Presence of atrial natriuretic peptide-like material in guinea-pig intestine, Peptides, 9 (1988) 965-971. 13 Vuolteenaho, O., Arjamaa, O., Vakkuri, O., Maksniemi, T., Nikkila, L., Kangas, J., Puurunen, J., Ruskoaho, H. and Leppaluoto, J., Atrial natriuretic peptide (ANP) in rat gastrointestinal tract, FEBS Lett., 233 (1988) 79-82. 14 Hardin, J.A., Catto-Smith, A. G., Patrick, M.K. and Gall, D.G., Characterization of the regulation of small intestinal electrolyte transport by atrial natriuretic factor, Clin. Invest. Med., 12 (1989) B34 (abstract). 15 Stefanini, M., De Martino, C. and Zamboni, L., Fixation of ejaculated spermatozoa for electron microscopy, Nature, 216 (1967) 173-174. 16 Saper, C.B., Hurley, K.M., Moga, M.M., Holmes, H.R., Adams, S.A., Leahy, K.M. and Needleman, P., Brain natriuretic peptides: differential localization of a new family of neuropeptides, Neurosci. Lett., 96 (1989) 29-34. 17 Catto-Smith, A. G., Hardin, J. A. and Gall, D. G., Regulation of small intestinal electrolyte transport by atrial natriuretic peptide (ANP), Gastroenterology, 98 (1990) A538 (abstract). 18 Konrad, E.M., Thibault, G., Pelletier, S., Genest, J. and Cantin, M., Brain natriuretic binding sites in rats: in vitro autoradiographic study, Am. J. Physiol., 259 (1990) E246-E255. 19 Sudoh, T., Minamino, N., Kangawa, K. and Matsuo, H., C-type natriuretic peptide (CNP): a new member of natriuretic peptide family identified in porcine brain, Biochem. Biophys. Res. Commun., 168 (1990) 863-870. 20 Minamino, N., Kangawa, K. and Matsuo, H., N-terminally extended form of C-type natriuretic peptide (CNP-53) identified in porcine brain, Biochem. Biophys. Res. Commun., 170 (1990) 973-979. 21 Siriwardena, A. K. and Kellum, J. M., Characterization of an enteric mucosal binding site for 5-hydroxytryptamine, Gastroenterology, 98 (1990) A524 (abstract).