0013-7227/90/1273-1292$02.00/0 Endocrinology Copyright © 1990 by The Endocrine Society
Vol. 127, No. 3 Printed in U.S.A.
Preparation of Monoclonal Antibodies against Brain Natriuretic Peptide and Their Application to Radioimmunoassay and Passive Immunization* HIROSHI ITOH, KAZUWA NAKAO, MASASHI MUKOYAMA, YOSHIHIKO SAITO, TAKAYUKI YAMADA, GOTARO SHIRAKAMI, HIROSHI ARAI, KIMINORI HOSODA, SHIN-ICHI SUGA, IKUYO YOSHIDA, AND HIROO IMURA Second Division, Department of Medicine, Kyoto University School of Medicine, Kyoto 606, Japan
concentration in the medulla oblongata. These RIAs also detected BNP-LI in both the porcine and canine hearts and in the porcine plasma. The iv pretreatment of purified mAb[KY-BNPII] almost completely blocked the hypotensive action of iv administered BNP in rats, with the concomitant suppression of BNP-induced increase of the plasma cyclic GMP level. These results indicate that our mAbs against BNP will serve as a useful tool for the elucidation of the physiological and pathophysiological significance of BNP as a neuropeptide and as a hormone (Endocrinology 127: 1292-1300,1990)
ABSTRACT. Two monoclonal antibodies (mAbs) directed toward brain natriuretic peptide (BNP), KY-BNP-I and KYBNP-II, have been produced. Both mAbs against BNP possessed high affinities for BNP, with association constants (Ka) of 4.0 x 109 M"1 (KY-BNP-I) and 2.0 x 1010 M"1 (KY-BNP-II). With these mAbs, specific RIAs for BNP have been established. The least detectable quantities of BNP were 5 pg/tube (KY-BNP-I) and 1 pg/tube (KY-BNP-II). Cross-reactivities of a-human and rat atrial natriuretic polypeptides were less than 0.001%. These RIAs detected BNP-like immunoreactivity (BNP-LI) not only in the porcine brain but also in the canine brain, with the highest
M
ATSUO and his colleagues (1) have recently isolated a novel natriuretic peptide from the porcine brain, which has 26 amino acid residues with an intramolecular disulfide linkage and shows a remarkable sequence homology to atrial natriuretic polypeptide (ANP), although it is distinct from it. The authors designated the peptide brain natriuretic peptide (BNP). BNP also was reported to elicit natriuretic, diuretic, and hypotensive actions in anesthetized rats (1) and to inhibit aldosterone production in dispersed bovine adrenocortical cells (2), as potently as ANP. Natural occurrence of ANP and BNP in the mammalian brain opens up the possibility of the existence of a natriuretic peptide family. Using porcine renal tubular epithelial cells or brain slices, it has been suggested that BNP and ANP share the same Received May 2, 1990. Address all correspondence and requests for reprints to: Dr. Kazuwa Nakao, Second Division, Department of Medicine, Kyoto University School of Medicine, 54 Shogoin Kawahara-cho, Sakyo-ku, Kyoto 606, Japan. * This work was supported in part by research grants from the Japanese Ministry of Education, Science and Culture, the Japanese Ministry of Health and Welfare Disorders of Adrenal Hormone Research Committee, Japan, 1988, Life Science Research Project of Institute of Physical and Chemical Research (RIKEN), Japan Tobacco Inc., and Yamanouchi Foundation for Research on Metabolic Disorders, by research grants for cardiovascular diseases (62A-1 and 63C-2) from the Japan Ministry of Health and Welfare.
receptor(s) (3, 4). Using an RIA specific to ANP (5) and immunohistochemical methods, we demonstrated that ANP is a hormone not only secreted from the heart (6) but also a neuropeptide found in the brain (7, 8). It has been demonstrated that the intracerebroventricular (icv) administration of ANP suppresses water intake (9-11), salt appetite (12), pressor response (13), and secretion of vasopressin and ACTH (14, 15). We also reported the inhibitory effect of the icv injection of ANP on central angiotensin II (All)-induced ANP secretion from the heart (16, 17). These results suggest an antagonistic relationship between ANP and renin-angiotensin systems in the central nervous system, as well as in the periphery, affecting the modulation of body fluid and blood pressure homeostasis (18, 19). We have further demonstrated that the icv administration of BNP exhibited the equipotent inhibitory actions on central Allinduced water intake (20), pressor response (21), and vasopressin secretion (22), when compared to ANP. These findings raise the possibility that central actions of ANP are shared by BNP and that there exists a dual system, ANP and BNP, in the brain which controls blood pressure and fluid balance in a way antagonistic to AIL More recently, we have established an RIA specific to BNP using rabbit and mouse polyclonal antisera against
1292
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 18:54 For personal use only. No other uses without permission. . All rights reserved.
ANTI-BRAIN NATRIURETIC PEPTIDE MONOCLONAL ANTIBODY BNP. With the aid of these RIAs, we have demonstrated that the distribution of BNP is different from that of ANP in the brain. In the porcine brain, the highest concentration of BNP was observed in the caudate nucleus, whereas that of ANP was in the midbrain. We also showed that BNP with a small molecular weight (3 K) is detectable not only in the porcine brain but also in the canine brain, suggesting that BNP is a common neuropeptide among species (23). We also reported that BNP with a large molecular weight (12 K) is synthesized and stored in the porcine atrium and is secreted into the circulation, using isolated porcine perfused heart (24). These findings suggest that BNP is also a brain-heart peptide like ANP. Polyclonal antibodies, however, have several drawbacks in their use, such as the limited supply and the contamination with irrelevant antibodies not directed toward BNP. The monoclonal antibody technique has now come into widespread use for the study of various substances including biologically active peptides. We have produced two kinds of monoclonal antibodies (mAbs) against ANP with high affinity (25, 26). By neutralizing the increased plasma ANP in hypertensive animals with mAb against ANP, we reported the aggravation of hypertension, which suggests the pathophysiological significance of endogenous ANP to prevent the development of hypertension (27). In the present paper, we describe the preparation and characterization of mAbs directed toward BNP and examine their application to RIA and neutralization experiment in order to further elucidate the physiological and pathophysiological significance of BNP as a neuropeptide and as a hormone.
Materials and Methods Peptides BNP was synthesized by the solid phase method (1). BNP32 (28) and BNP[17-26] were delivered by Dr. Ken Inouye and Dr. Yoshikazu Kambayashi (Shionogi Laboratories, Osaka, Japan). a-Human ANP (a-hANP) and a-rat ANP (a-rANP) were purchased from Protein Research Foundation (Minoh, Japan) and Peninsula Laboratories Inc. (Belmont, CA), respectively. The homogeneity of these peptides was confirmed by reverse phase HPLC and amino acid analysis. Preparation of mAbs Synthetic BNP (5 mg) was conjugated to bovine thyroglobulin (18.3 mg, Sigma Chemical, St. Louis, MO) using the carbodiimide (Nakarai Chemicals, Kyoto, Japan) coupling procedure (23). Adult female BALB/c mice (Shizuoka Animal Center, Shizuoka, Japan) were initially immunized by combined sc and ip injections with the conjugate containing 15 ^g BNP emulsified in complete Freund's adjuvant (Difco Laboratories, Detroit, MI). Two weeks later, they were boosted in
1293
the same manner, and thereafter, immunization was repeated twice at about 4-week intervals with the same amount of BNP in complete adjuvant sc. The mice were bled to test for the presence of antibody by RIA with [125I]BNP. Two mice giving the greatest response was further boosted iv with the conjugate containing 30 fig antigen 4 weeks after the last sc booster. Five days later, spleen cells were harvested from the mice for cell fusion. Fusion of spleen cells with nonproducing mouse myeloma cell line, X63-Ag8.653, screening, and cloning of hybridoma cells were performed according to the method previously described (25, 26). Positive clones were expanded in culture and injected ip into BALB/c mice pretreated with pristane (2,6,10,14-tetramethylpentadecane; Aldrich, Milwaukee, WI) to produce ascitic fluid with highly concentrated antibody. Isotyping of the monoclonal antibody was performed by the Ouchterlony technique (Mouse Monoclonal Typing Kit, Miles Laboratories, Elkhart, IN). RIA for BNP RIAs with mAbs against BNP were performed according to the method of the RIA with the polyclonal antiserum as previously described in detail (23). The RIA incubation mixture (500 ^I/tube) consisted of 100 ^1 standard BNP or sample, 100 /ul diluted antibody (antisera, culture media, or ascites), 100 >ul [125I]BNP (SA, 400-700 ixCi/^g; -10,000 cpm), and 200 n\ standard buffer [0.1 M phosphate buffer, pH 7.0, containing 0.5% gelatin (Merck, Darmstadt, West Germany), 1 mM Na2EDTA (Sigma Chemical Co., St. Louis, MO), 0.2 mM cystine, 0.1% Triton X-100, and 0.01% merthiolate]. The mixture was incubated for 24-48 h at 4 C. Bound and free ligands were separated by dextran-coated charcoal method (23). RIA for ANP ANP-like immunoreactivity (ANP-LI) was measured by RIA with a rabbit polyclonal antiserum that recognizes the Cterminal sequence of a-ANP, as previously described in detail (5-8). The cross-reactivity of BNP in the RIA was less than 0.002% (23). Tissues and extraction procedure Brains were obtained both from adult pigs (n = 4) and oneday-old pigs (n = 3) within 10 min after killing. Brains of dogs (beagles; n = 4), rats (n = 3), monkeys (n = 2), and humans (n = 2) were also studied. Brains were dissected on ice at 4 C into 11 regions referring to the rat brain dissection method of Glowinski and Iversen (7, 23). Hearts were excised from adult pigs (n = 4) and dogs (n = 3) immediately after killing. After obtaining right and left atria, right and left ventricular tissues also were dissected with special care to prevent atrial contamination (24). Brain and cardiac tissues were boiled for 5 min in 10 vol 0.1 M acetic acid and extracted as previously reported (23, 24). High performance-gel permeation chromatography (HP-GPC) HP-GPC was performed on a TSK-GEL G2000 SW (Toyo Soda, Tokyo, Japan) column (7.5 X 600 mm) and eluted with
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 18:54 For personal use only. No other uses without permission. . All rights reserved.
1294
ANTI-BRAIN NATRIURETIC PEPTIDE MONOCLONAL ANTIBODY
10 mM trifluoroacetic acid containing 0.3 M sodium chloride and 30% acetonitrile, as previously reported (5-8, 23, 24). Administration of mAbs to conscious unrestrained rats Conscious, unrestrained male Sic: Wistar rats (Shizuoka Animal Center, Shizuoka, Japan) weighing 250-350 g were used. A silastic cannula was implanted at the junction of the superior vena cava and the right atrium through the right external jugular vein under light ether anesthesia, as previously reported (29). mAb[KY-BNP-II] or mAb[KY-ANP-II] (26) in ascites was purified by Affi-Gel Protein A MAPS II kit (BioRad Laboratories, Richmond, CA) and dissolved in 500 ^1 physiological saline (500 fig immunoglobulin G). It was injected in 30 sec as a bolus via a venous silastic catheter, as previously reported (27). Blood pressure measurement A polyethylene catheter (id 0.50 mm, od 0.80 mm) was inserted into the left femoral artery under light ether anesthesia 24 h before the experiment, as previously reported (13). Blood pressure was monitored while the animals were conscious and freely moving in their home cage (17 X 24 X 12 cm) via an electronic manometer MP-24T (Nihon Kohden Co., Tokyo, Japan) and a multipurpose polygraph RM-150 (Nihon Kohden Co.). Blood sampling and determination of plasma BNP levels and plasma cyclic GMP levels Blood sampling was performed in adult pigs when they were killed. Blood (500 ^1) also was obtained through the venous catheter from conscious unrestrained rats and transferred into chilled polypropyrene tubes containing aprotinin (Ohkura Pharmaceutical Inc., Kyoto, Japan; 1,000 Kallikrein inactivator units/ml) and Na2 EDTA (10 mM), as previously reported (29). Porcine plasma BNP levels were determined by RIA with mAb[KY-BNP-II] with extraction, using a Sep-Pak C18 cartridge (Waters Associates, Inc., Milford, MA) as described previously in detail (6, 30). Rat plasma BNP levels after the administration of BNP were determined with 25 /A rat plasma without extraction by RIA with mAb[KY-BNP-II] as previously reported (29). Plasma cyclic GMP levels were determined with a commercially available RIA kit (YAMASA Cyclic GMP Assay Kit, Yamasa Shoyu Co., Chiba, Japan) according to the method of Steiner et al. (31). In vivo effectiveness of mAb against BNP Experiment 1. Rats in group 1-A (n = 5) were pretreated with the iv administration of purified mAb[KY-ANP-II] (500 ng) as a control mAb, which did not show any cross-reactivity with BNP (26), 15 min before the iv bolus administration of 10 ng synthetic BNP. Rats in group 1-B (n = 5) were pretreated with the same amount of purified mAb[KY-BNP-II], 15 min before the iv administration of 10 /ig BNP. Changes of mean arterial pressure (MAP) of groups 1-A and 1-B were compared. Experiment 2. Rats in groups 2-A (n = 7) were pretreated with
E n d o • 1990 Vol 127 • No 3
the iv administration of purified mAb[KY-ANP-II] (500 Mg) as a control 5 min before the iv administration of 10 ng BNP. Rats in group 2-B (n = 7) were pretreated with purified mAb[KY-BNP-II] (500 ng) 5 min before the administration of 10 ng BNP. Blood was collected before the mAb-administration, just before the injection of BNP and 2, 5, 10, and 20 min after the administration of BNP through the venous catheter. The plasma BNP level in group 2-A and cyclic GMP level in groups 2-A and 2-B at each sampling point were determined. Statistical analysis All values are expressed as mean ± SEM. Analysis of variance of repeated measures, with subsequent Duncan's test, was used to determine significant changes during time-dependent multiple observations (32). The statistical significance was defined as P < 0.05.
Results Preparation and characterization of mAbs against BNP After the fusion, hybridomas grew in nearly all of the wells (523 of 528). BNP antibody-producing cells were recognized in 9 out of 523 wells. After further culture and cloning, two clones were selected for expansion and characterization. The obtained two mAbs were named KY-BNP-I and KY-BNP-II. Both of the mAbs belonged to the immunoglobulin Gx subclass, as determined by the Ouchterlony technique. Analyses of a Scatchard plot revealed high affinity for BNP, with an association constant (Ka) of 4.0 x 109 M"1 (KY-BNP-I) and 2.0 x 1010 M"1 (KY-BNP-II), respectively (Fig. 1). Both mAbs against BNP were useful for RIA at final dilutions of ascitic fluid as much as 1:1.6 X 107 (KYBNP-I) and 1:1 X 108 (KY-BNP-II). Standard curves of BNP and cross-reactivities of its related peptides, ahANP and a-rANP in the RIAs with mAb[KY-BNP-I] and [KY-BNP-II] are shown in Fig. 1. The RIA with mAb[KY-BNP-I] showed the cross-reactivity of 55% with BNP-32, another form of endogenous BNP in the porcine brain (28), while the RIA with mAb[KY-BNPII] exhibited the cross-reactivity with BNP-32 of 33%, on a molar basis. The C-terminal fragment of BNP, BNP[17-26],,did not show any cross-reactivity in either RIA. Cross-reactivities of a-hANP and a-rANP were also less than 0.001% in both RIAs. RIA for BNP and detection of BNP-LI in porcine and canine tissues and porcine plasma In the RIA with mAb [KY-BNP-I], the minimal detectable quantity, in which 5% total tracer binding could be inhibited (IC5) was 5 pg/tube and the 50% binding intercept (IC50) was 80 pg/tube. In the RIA with mAb [KY-BNP-II], IC5 was as little as 1 pg/tube and IC50 was 15 pg/tube. The intraassay and interassay coeffi-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 18:54 For personal use only. No other uses without permission. . All rights reserved.
ANTI-BRAIN NATRIURETIC PEPTIDE MONOCLONAL ANTIBODY
1295
25 -
25 o 1 -// 10
102
10 J
Peptide ( p g / t u b e )
10 102 Peptide (pg/tube) FIG. 2. A standard curve of BNP (•) in the RIA with mAb[KY-BNPII] and dilution curves of the extract from porcine and canine tissues. A, brain (caudate nucleus), pig; B, right atrium, pig; C, right atrium, dog. °
10
10 2
10 3
"
1
10'
Peptide ( p g / t u b e )
FIG. 1. The standard curves of BNP (•) and dilution curves of BNPrelated peptides, a-hANP, and a-rANP in the RIAs with mAb[KYBNP-I] (top panel) and [KY-BNP-II] (bottom panel). [12BI]BNP and various concentrations of standard BNP or related peptides were incubated with mAb[KY-BNP-I] or [KY-BNP-II] for 48 h at 4 C. Insets, Scatchard plot analysis of the binding of BNP to mAb[KYBNP-I] or [KY-BNP-II]. Ascites containing mAb[KY-BNP-I] or [KYBNP-II] were incubated with increasing concentrations of [125I]BNP (1.0-104 pM, 500 /il/tube) for 48 h at 4 C, and the specific binding was detected after separation by the dextran-coated charcoal method. [B], Concentration of specifically bound BNP; B/F, bound/free ratio of the ligand. 30
cients of variation (for 10 pg/tube of BNP) were 9.7% (n = 6) and 13.2% (n = 6), respectively, in the RIA with mAb[KY-BNP-I], and 3.9% (n = 8) and 9.6% (n = 7), respectively, in the RIA with mAb[KY-BNP-II]. Serial dilution curves of brain and atrial tissues of the pig and dog were parallel to standard curves of BNP in the RIAs with mAb[KY-BNP-I] and [KY-BNP-II] (Fig. 2). The brain extracts of rats, monkeys, or humans, however, did not exhibit cross-reactivity in either RIA.
40
50
60
70
FRACTION NUMBER
FIG. 3. Gel filtration profile on a TSK-GEL G2000 SW column of BNP-LI determined by the RIA with mAb[KY-BNP-II] in extracts from the medulla oblongata in pigs. Arrows denote the elution positions of a series of myoglobins of a polypeptide mol wt calibration kit (Pharmacia Fine Chemicals, Uppsala, Sweden). Vo, void volume; Vt, total volume.
Figure 3 shows a gel filtration profile of BNP-LI in the porcine brain extract (medulla oblongata) determined by
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 18:54 For personal use only. No other uses without permission. . All rights reserved.
1296
ANTI-BRAIN NATRIURETIC PEPTIDE MONOCLONAL ANTIBODY
HP-GPC coupled with the RIA with mAb[KY-BNP-II]. BNP-LI was eluted as a single peak with an approximate mol wt of 3,000 at the elution position of synthetic BNP or BNP-32. The elution pattern of BNP-LI in the porcine brain is compatible with our previous report using the RIA for BNP with polyclonal antisera (23). The same elution pattern was observed also in the canine brain extract, as we previously demonstrated (23). In contrast, the major peak of BNP-LI in the porcine and canine atrial extract eluted at the position of the high molecular form with a mol wt of 12,000, which is also consistent with our previous finding (24). Table 1 gives regional distributions of ANP-LI and BNP-LI, determined by the RIA for ANP and the RIA for BNP with mAb[KY-BNP-II] in the brains of adult and neonatal pigs. In the adult pig, the rank order of BNP-LI concentrations were: caudate nucleus > medulla oblongata > lenticular nucleus > olfactory bulb > cerebral cortex > hypothalamus > midbrain > thalamus > pons. BNP-LI was not detected in the hippocampus, cerebellum, and pituitary gland. In the neonatal stage, the highest concentration of BNP-LI was observed in the medulla oblongata, followed by the lenticular nucleus and caudate nucleus. The concentration of ANP-LI was lower than that of BNP in the adult porcine brain. The distribution pattern of ANP-LI differs from that of BNP-LI. The highest concentration of ANP-LI was detected in the midbrain, followed by the olfactory bulb and hypothalamus. In the neonatal stage, however, the ANP-LI level was higher than the BNP-LI level, with the highest concentration in the olfactory bulb. In the canine brain, the highest concentration of BNPLI was detected in the medulla oblongata (0.30 ± 0.01 ng/g), followed by the pons (0.26 ± 0.05 ng/g), midbrain
Endo • 1990 Vol 127 • No 3
(0.23 ± 0.04 ng/g), and olfactory bulb (0.18 ± 0.07 ng/g). In contrast, the highest concentration of ANP-LI was detected in the olfactory bulb (0.30 ± 0.05 ng/g), followed by the caudate nucleus (0.24 ±0.11 ng/g) and hypothalamus (0.18 ± 0.01 ng/g). Table 2 shows the concentrations of BNP-LI and ANP-LI in the porcine heart. Like ANP, BNP levels were much higher in the atria than in the ventricles. The concentration of BNP-LI was 1-4% of that of ANP-LI both in the atria and the ventricles. Concentrations of BNP-LI in porcine and canine right atria determined by the RIAs with mAb[KY-BNP-I] and mAb[KY-BNP-II] were compared and shown in Table 3. In both porcine and canine atria, BNP-LI determined by the RIA with mAb[KY-BNP-I] was about 70% of that determined by the RIA with mAb[KY-BNP-II]. The serial dilution curve of the porcine plasma extract was parallel to the standard curve of BNP in the RIA with mAb[KY-BNP-II]. The porcine plasma level of BNP-LI was 4.2 ± 1.4 pg/ml (n = 4). The plasma level of ANP-LI determined simultaneously was 147.6 ± 56.4 pg/ml. The ratio of the plasma BNP-LI level to the plasma ANP-LI level was, therefore, 3.0 ± 0.8%. In HPGPC coupled with the RIA with mAb[KY-BNP-II], BNP-LI in the porcine plasma extract was eluted at the elution position of synthetic BNP or BNP-32. Neutralization of the effect of BNP by mAb[KY-BNP-II] Exp 1: effect on blood pressure. The administration of mAb[KY-ANP-II] or mAb[KY-BNP-II] caused no significant change of basal MAP (AMAP: - 2 ± 2 mmHg for mAb[KY-ANP-II] and +2 ± 2 mmHg for mAb[KYBNP-II]). The iv bolus administration of 10 ng BNP in rats of
TABLE 1. Regional distribution of BNP-LF and ANP-LI in the porcine brain Adult (n = 4)
Brain region Olfactory bulb Cerebral cortex Hippocampus Lenticular nucleus Caudate nucleus Thalamus Hypothalamus Midbrain Pons Medulla oblongata Cerebellum Cervical spinal cord Pituitary gland (pooled)
Neonate (n = 3)
BNP-LI
ANP-LI
BNP-LI
ANP-LI
1.12 ± 0.68 0.66 ± 0.34
Vol. 127, No. 3 Printed in U.S.A.
Preparation of Monoclonal Antibodies against Brain Natriuretic Peptide and Their Application to Radioimmunoassay and Passive Immunization* HIROSHI ITOH, KAZUWA NAKAO, MASASHI MUKOYAMA, YOSHIHIKO SAITO, TAKAYUKI YAMADA, GOTARO SHIRAKAMI, HIROSHI ARAI, KIMINORI HOSODA, SHIN-ICHI SUGA, IKUYO YOSHIDA, AND HIROO IMURA Second Division, Department of Medicine, Kyoto University School of Medicine, Kyoto 606, Japan
concentration in the medulla oblongata. These RIAs also detected BNP-LI in both the porcine and canine hearts and in the porcine plasma. The iv pretreatment of purified mAb[KY-BNPII] almost completely blocked the hypotensive action of iv administered BNP in rats, with the concomitant suppression of BNP-induced increase of the plasma cyclic GMP level. These results indicate that our mAbs against BNP will serve as a useful tool for the elucidation of the physiological and pathophysiological significance of BNP as a neuropeptide and as a hormone (Endocrinology 127: 1292-1300,1990)
ABSTRACT. Two monoclonal antibodies (mAbs) directed toward brain natriuretic peptide (BNP), KY-BNP-I and KYBNP-II, have been produced. Both mAbs against BNP possessed high affinities for BNP, with association constants (Ka) of 4.0 x 109 M"1 (KY-BNP-I) and 2.0 x 1010 M"1 (KY-BNP-II). With these mAbs, specific RIAs for BNP have been established. The least detectable quantities of BNP were 5 pg/tube (KY-BNP-I) and 1 pg/tube (KY-BNP-II). Cross-reactivities of a-human and rat atrial natriuretic polypeptides were less than 0.001%. These RIAs detected BNP-like immunoreactivity (BNP-LI) not only in the porcine brain but also in the canine brain, with the highest
M
ATSUO and his colleagues (1) have recently isolated a novel natriuretic peptide from the porcine brain, which has 26 amino acid residues with an intramolecular disulfide linkage and shows a remarkable sequence homology to atrial natriuretic polypeptide (ANP), although it is distinct from it. The authors designated the peptide brain natriuretic peptide (BNP). BNP also was reported to elicit natriuretic, diuretic, and hypotensive actions in anesthetized rats (1) and to inhibit aldosterone production in dispersed bovine adrenocortical cells (2), as potently as ANP. Natural occurrence of ANP and BNP in the mammalian brain opens up the possibility of the existence of a natriuretic peptide family. Using porcine renal tubular epithelial cells or brain slices, it has been suggested that BNP and ANP share the same Received May 2, 1990. Address all correspondence and requests for reprints to: Dr. Kazuwa Nakao, Second Division, Department of Medicine, Kyoto University School of Medicine, 54 Shogoin Kawahara-cho, Sakyo-ku, Kyoto 606, Japan. * This work was supported in part by research grants from the Japanese Ministry of Education, Science and Culture, the Japanese Ministry of Health and Welfare Disorders of Adrenal Hormone Research Committee, Japan, 1988, Life Science Research Project of Institute of Physical and Chemical Research (RIKEN), Japan Tobacco Inc., and Yamanouchi Foundation for Research on Metabolic Disorders, by research grants for cardiovascular diseases (62A-1 and 63C-2) from the Japan Ministry of Health and Welfare.
receptor(s) (3, 4). Using an RIA specific to ANP (5) and immunohistochemical methods, we demonstrated that ANP is a hormone not only secreted from the heart (6) but also a neuropeptide found in the brain (7, 8). It has been demonstrated that the intracerebroventricular (icv) administration of ANP suppresses water intake (9-11), salt appetite (12), pressor response (13), and secretion of vasopressin and ACTH (14, 15). We also reported the inhibitory effect of the icv injection of ANP on central angiotensin II (All)-induced ANP secretion from the heart (16, 17). These results suggest an antagonistic relationship between ANP and renin-angiotensin systems in the central nervous system, as well as in the periphery, affecting the modulation of body fluid and blood pressure homeostasis (18, 19). We have further demonstrated that the icv administration of BNP exhibited the equipotent inhibitory actions on central Allinduced water intake (20), pressor response (21), and vasopressin secretion (22), when compared to ANP. These findings raise the possibility that central actions of ANP are shared by BNP and that there exists a dual system, ANP and BNP, in the brain which controls blood pressure and fluid balance in a way antagonistic to AIL More recently, we have established an RIA specific to BNP using rabbit and mouse polyclonal antisera against
1292
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 18:54 For personal use only. No other uses without permission. . All rights reserved.
ANTI-BRAIN NATRIURETIC PEPTIDE MONOCLONAL ANTIBODY BNP. With the aid of these RIAs, we have demonstrated that the distribution of BNP is different from that of ANP in the brain. In the porcine brain, the highest concentration of BNP was observed in the caudate nucleus, whereas that of ANP was in the midbrain. We also showed that BNP with a small molecular weight (3 K) is detectable not only in the porcine brain but also in the canine brain, suggesting that BNP is a common neuropeptide among species (23). We also reported that BNP with a large molecular weight (12 K) is synthesized and stored in the porcine atrium and is secreted into the circulation, using isolated porcine perfused heart (24). These findings suggest that BNP is also a brain-heart peptide like ANP. Polyclonal antibodies, however, have several drawbacks in their use, such as the limited supply and the contamination with irrelevant antibodies not directed toward BNP. The monoclonal antibody technique has now come into widespread use for the study of various substances including biologically active peptides. We have produced two kinds of monoclonal antibodies (mAbs) against ANP with high affinity (25, 26). By neutralizing the increased plasma ANP in hypertensive animals with mAb against ANP, we reported the aggravation of hypertension, which suggests the pathophysiological significance of endogenous ANP to prevent the development of hypertension (27). In the present paper, we describe the preparation and characterization of mAbs directed toward BNP and examine their application to RIA and neutralization experiment in order to further elucidate the physiological and pathophysiological significance of BNP as a neuropeptide and as a hormone.
Materials and Methods Peptides BNP was synthesized by the solid phase method (1). BNP32 (28) and BNP[17-26] were delivered by Dr. Ken Inouye and Dr. Yoshikazu Kambayashi (Shionogi Laboratories, Osaka, Japan). a-Human ANP (a-hANP) and a-rat ANP (a-rANP) were purchased from Protein Research Foundation (Minoh, Japan) and Peninsula Laboratories Inc. (Belmont, CA), respectively. The homogeneity of these peptides was confirmed by reverse phase HPLC and amino acid analysis. Preparation of mAbs Synthetic BNP (5 mg) was conjugated to bovine thyroglobulin (18.3 mg, Sigma Chemical, St. Louis, MO) using the carbodiimide (Nakarai Chemicals, Kyoto, Japan) coupling procedure (23). Adult female BALB/c mice (Shizuoka Animal Center, Shizuoka, Japan) were initially immunized by combined sc and ip injections with the conjugate containing 15 ^g BNP emulsified in complete Freund's adjuvant (Difco Laboratories, Detroit, MI). Two weeks later, they were boosted in
1293
the same manner, and thereafter, immunization was repeated twice at about 4-week intervals with the same amount of BNP in complete adjuvant sc. The mice were bled to test for the presence of antibody by RIA with [125I]BNP. Two mice giving the greatest response was further boosted iv with the conjugate containing 30 fig antigen 4 weeks after the last sc booster. Five days later, spleen cells were harvested from the mice for cell fusion. Fusion of spleen cells with nonproducing mouse myeloma cell line, X63-Ag8.653, screening, and cloning of hybridoma cells were performed according to the method previously described (25, 26). Positive clones were expanded in culture and injected ip into BALB/c mice pretreated with pristane (2,6,10,14-tetramethylpentadecane; Aldrich, Milwaukee, WI) to produce ascitic fluid with highly concentrated antibody. Isotyping of the monoclonal antibody was performed by the Ouchterlony technique (Mouse Monoclonal Typing Kit, Miles Laboratories, Elkhart, IN). RIA for BNP RIAs with mAbs against BNP were performed according to the method of the RIA with the polyclonal antiserum as previously described in detail (23). The RIA incubation mixture (500 ^I/tube) consisted of 100 ^1 standard BNP or sample, 100 /ul diluted antibody (antisera, culture media, or ascites), 100 >ul [125I]BNP (SA, 400-700 ixCi/^g; -10,000 cpm), and 200 n\ standard buffer [0.1 M phosphate buffer, pH 7.0, containing 0.5% gelatin (Merck, Darmstadt, West Germany), 1 mM Na2EDTA (Sigma Chemical Co., St. Louis, MO), 0.2 mM cystine, 0.1% Triton X-100, and 0.01% merthiolate]. The mixture was incubated for 24-48 h at 4 C. Bound and free ligands were separated by dextran-coated charcoal method (23). RIA for ANP ANP-like immunoreactivity (ANP-LI) was measured by RIA with a rabbit polyclonal antiserum that recognizes the Cterminal sequence of a-ANP, as previously described in detail (5-8). The cross-reactivity of BNP in the RIA was less than 0.002% (23). Tissues and extraction procedure Brains were obtained both from adult pigs (n = 4) and oneday-old pigs (n = 3) within 10 min after killing. Brains of dogs (beagles; n = 4), rats (n = 3), monkeys (n = 2), and humans (n = 2) were also studied. Brains were dissected on ice at 4 C into 11 regions referring to the rat brain dissection method of Glowinski and Iversen (7, 23). Hearts were excised from adult pigs (n = 4) and dogs (n = 3) immediately after killing. After obtaining right and left atria, right and left ventricular tissues also were dissected with special care to prevent atrial contamination (24). Brain and cardiac tissues were boiled for 5 min in 10 vol 0.1 M acetic acid and extracted as previously reported (23, 24). High performance-gel permeation chromatography (HP-GPC) HP-GPC was performed on a TSK-GEL G2000 SW (Toyo Soda, Tokyo, Japan) column (7.5 X 600 mm) and eluted with
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 18:54 For personal use only. No other uses without permission. . All rights reserved.
1294
ANTI-BRAIN NATRIURETIC PEPTIDE MONOCLONAL ANTIBODY
10 mM trifluoroacetic acid containing 0.3 M sodium chloride and 30% acetonitrile, as previously reported (5-8, 23, 24). Administration of mAbs to conscious unrestrained rats Conscious, unrestrained male Sic: Wistar rats (Shizuoka Animal Center, Shizuoka, Japan) weighing 250-350 g were used. A silastic cannula was implanted at the junction of the superior vena cava and the right atrium through the right external jugular vein under light ether anesthesia, as previously reported (29). mAb[KY-BNP-II] or mAb[KY-ANP-II] (26) in ascites was purified by Affi-Gel Protein A MAPS II kit (BioRad Laboratories, Richmond, CA) and dissolved in 500 ^1 physiological saline (500 fig immunoglobulin G). It was injected in 30 sec as a bolus via a venous silastic catheter, as previously reported (27). Blood pressure measurement A polyethylene catheter (id 0.50 mm, od 0.80 mm) was inserted into the left femoral artery under light ether anesthesia 24 h before the experiment, as previously reported (13). Blood pressure was monitored while the animals were conscious and freely moving in their home cage (17 X 24 X 12 cm) via an electronic manometer MP-24T (Nihon Kohden Co., Tokyo, Japan) and a multipurpose polygraph RM-150 (Nihon Kohden Co.). Blood sampling and determination of plasma BNP levels and plasma cyclic GMP levels Blood sampling was performed in adult pigs when they were killed. Blood (500 ^1) also was obtained through the venous catheter from conscious unrestrained rats and transferred into chilled polypropyrene tubes containing aprotinin (Ohkura Pharmaceutical Inc., Kyoto, Japan; 1,000 Kallikrein inactivator units/ml) and Na2 EDTA (10 mM), as previously reported (29). Porcine plasma BNP levels were determined by RIA with mAb[KY-BNP-II] with extraction, using a Sep-Pak C18 cartridge (Waters Associates, Inc., Milford, MA) as described previously in detail (6, 30). Rat plasma BNP levels after the administration of BNP were determined with 25 /A rat plasma without extraction by RIA with mAb[KY-BNP-II] as previously reported (29). Plasma cyclic GMP levels were determined with a commercially available RIA kit (YAMASA Cyclic GMP Assay Kit, Yamasa Shoyu Co., Chiba, Japan) according to the method of Steiner et al. (31). In vivo effectiveness of mAb against BNP Experiment 1. Rats in group 1-A (n = 5) were pretreated with the iv administration of purified mAb[KY-ANP-II] (500 ng) as a control mAb, which did not show any cross-reactivity with BNP (26), 15 min before the iv bolus administration of 10 ng synthetic BNP. Rats in group 1-B (n = 5) were pretreated with the same amount of purified mAb[KY-BNP-II], 15 min before the iv administration of 10 /ig BNP. Changes of mean arterial pressure (MAP) of groups 1-A and 1-B were compared. Experiment 2. Rats in groups 2-A (n = 7) were pretreated with
E n d o • 1990 Vol 127 • No 3
the iv administration of purified mAb[KY-ANP-II] (500 Mg) as a control 5 min before the iv administration of 10 ng BNP. Rats in group 2-B (n = 7) were pretreated with purified mAb[KY-BNP-II] (500 ng) 5 min before the administration of 10 ng BNP. Blood was collected before the mAb-administration, just before the injection of BNP and 2, 5, 10, and 20 min after the administration of BNP through the venous catheter. The plasma BNP level in group 2-A and cyclic GMP level in groups 2-A and 2-B at each sampling point were determined. Statistical analysis All values are expressed as mean ± SEM. Analysis of variance of repeated measures, with subsequent Duncan's test, was used to determine significant changes during time-dependent multiple observations (32). The statistical significance was defined as P < 0.05.
Results Preparation and characterization of mAbs against BNP After the fusion, hybridomas grew in nearly all of the wells (523 of 528). BNP antibody-producing cells were recognized in 9 out of 523 wells. After further culture and cloning, two clones were selected for expansion and characterization. The obtained two mAbs were named KY-BNP-I and KY-BNP-II. Both of the mAbs belonged to the immunoglobulin Gx subclass, as determined by the Ouchterlony technique. Analyses of a Scatchard plot revealed high affinity for BNP, with an association constant (Ka) of 4.0 x 109 M"1 (KY-BNP-I) and 2.0 x 1010 M"1 (KY-BNP-II), respectively (Fig. 1). Both mAbs against BNP were useful for RIA at final dilutions of ascitic fluid as much as 1:1.6 X 107 (KYBNP-I) and 1:1 X 108 (KY-BNP-II). Standard curves of BNP and cross-reactivities of its related peptides, ahANP and a-rANP in the RIAs with mAb[KY-BNP-I] and [KY-BNP-II] are shown in Fig. 1. The RIA with mAb[KY-BNP-I] showed the cross-reactivity of 55% with BNP-32, another form of endogenous BNP in the porcine brain (28), while the RIA with mAb[KY-BNPII] exhibited the cross-reactivity with BNP-32 of 33%, on a molar basis. The C-terminal fragment of BNP, BNP[17-26],,did not show any cross-reactivity in either RIA. Cross-reactivities of a-hANP and a-rANP were also less than 0.001% in both RIAs. RIA for BNP and detection of BNP-LI in porcine and canine tissues and porcine plasma In the RIA with mAb [KY-BNP-I], the minimal detectable quantity, in which 5% total tracer binding could be inhibited (IC5) was 5 pg/tube and the 50% binding intercept (IC50) was 80 pg/tube. In the RIA with mAb [KY-BNP-II], IC5 was as little as 1 pg/tube and IC50 was 15 pg/tube. The intraassay and interassay coeffi-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 18:54 For personal use only. No other uses without permission. . All rights reserved.
ANTI-BRAIN NATRIURETIC PEPTIDE MONOCLONAL ANTIBODY
1295
25 -
25 o 1 -// 10
102
10 J
Peptide ( p g / t u b e )
10 102 Peptide (pg/tube) FIG. 2. A standard curve of BNP (•) in the RIA with mAb[KY-BNPII] and dilution curves of the extract from porcine and canine tissues. A, brain (caudate nucleus), pig; B, right atrium, pig; C, right atrium, dog. °
10
10 2
10 3
"
1
10'
Peptide ( p g / t u b e )
FIG. 1. The standard curves of BNP (•) and dilution curves of BNPrelated peptides, a-hANP, and a-rANP in the RIAs with mAb[KYBNP-I] (top panel) and [KY-BNP-II] (bottom panel). [12BI]BNP and various concentrations of standard BNP or related peptides were incubated with mAb[KY-BNP-I] or [KY-BNP-II] for 48 h at 4 C. Insets, Scatchard plot analysis of the binding of BNP to mAb[KYBNP-I] or [KY-BNP-II]. Ascites containing mAb[KY-BNP-I] or [KYBNP-II] were incubated with increasing concentrations of [125I]BNP (1.0-104 pM, 500 /il/tube) for 48 h at 4 C, and the specific binding was detected after separation by the dextran-coated charcoal method. [B], Concentration of specifically bound BNP; B/F, bound/free ratio of the ligand. 30
cients of variation (for 10 pg/tube of BNP) were 9.7% (n = 6) and 13.2% (n = 6), respectively, in the RIA with mAb[KY-BNP-I], and 3.9% (n = 8) and 9.6% (n = 7), respectively, in the RIA with mAb[KY-BNP-II]. Serial dilution curves of brain and atrial tissues of the pig and dog were parallel to standard curves of BNP in the RIAs with mAb[KY-BNP-I] and [KY-BNP-II] (Fig. 2). The brain extracts of rats, monkeys, or humans, however, did not exhibit cross-reactivity in either RIA.
40
50
60
70
FRACTION NUMBER
FIG. 3. Gel filtration profile on a TSK-GEL G2000 SW column of BNP-LI determined by the RIA with mAb[KY-BNP-II] in extracts from the medulla oblongata in pigs. Arrows denote the elution positions of a series of myoglobins of a polypeptide mol wt calibration kit (Pharmacia Fine Chemicals, Uppsala, Sweden). Vo, void volume; Vt, total volume.
Figure 3 shows a gel filtration profile of BNP-LI in the porcine brain extract (medulla oblongata) determined by
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 18:54 For personal use only. No other uses without permission. . All rights reserved.
1296
ANTI-BRAIN NATRIURETIC PEPTIDE MONOCLONAL ANTIBODY
HP-GPC coupled with the RIA with mAb[KY-BNP-II]. BNP-LI was eluted as a single peak with an approximate mol wt of 3,000 at the elution position of synthetic BNP or BNP-32. The elution pattern of BNP-LI in the porcine brain is compatible with our previous report using the RIA for BNP with polyclonal antisera (23). The same elution pattern was observed also in the canine brain extract, as we previously demonstrated (23). In contrast, the major peak of BNP-LI in the porcine and canine atrial extract eluted at the position of the high molecular form with a mol wt of 12,000, which is also consistent with our previous finding (24). Table 1 gives regional distributions of ANP-LI and BNP-LI, determined by the RIA for ANP and the RIA for BNP with mAb[KY-BNP-II] in the brains of adult and neonatal pigs. In the adult pig, the rank order of BNP-LI concentrations were: caudate nucleus > medulla oblongata > lenticular nucleus > olfactory bulb > cerebral cortex > hypothalamus > midbrain > thalamus > pons. BNP-LI was not detected in the hippocampus, cerebellum, and pituitary gland. In the neonatal stage, the highest concentration of BNP-LI was observed in the medulla oblongata, followed by the lenticular nucleus and caudate nucleus. The concentration of ANP-LI was lower than that of BNP in the adult porcine brain. The distribution pattern of ANP-LI differs from that of BNP-LI. The highest concentration of ANP-LI was detected in the midbrain, followed by the olfactory bulb and hypothalamus. In the neonatal stage, however, the ANP-LI level was higher than the BNP-LI level, with the highest concentration in the olfactory bulb. In the canine brain, the highest concentration of BNPLI was detected in the medulla oblongata (0.30 ± 0.01 ng/g), followed by the pons (0.26 ± 0.05 ng/g), midbrain
Endo • 1990 Vol 127 • No 3
(0.23 ± 0.04 ng/g), and olfactory bulb (0.18 ± 0.07 ng/g). In contrast, the highest concentration of ANP-LI was detected in the olfactory bulb (0.30 ± 0.05 ng/g), followed by the caudate nucleus (0.24 ±0.11 ng/g) and hypothalamus (0.18 ± 0.01 ng/g). Table 2 shows the concentrations of BNP-LI and ANP-LI in the porcine heart. Like ANP, BNP levels were much higher in the atria than in the ventricles. The concentration of BNP-LI was 1-4% of that of ANP-LI both in the atria and the ventricles. Concentrations of BNP-LI in porcine and canine right atria determined by the RIAs with mAb[KY-BNP-I] and mAb[KY-BNP-II] were compared and shown in Table 3. In both porcine and canine atria, BNP-LI determined by the RIA with mAb[KY-BNP-I] was about 70% of that determined by the RIA with mAb[KY-BNP-II]. The serial dilution curve of the porcine plasma extract was parallel to the standard curve of BNP in the RIA with mAb[KY-BNP-II]. The porcine plasma level of BNP-LI was 4.2 ± 1.4 pg/ml (n = 4). The plasma level of ANP-LI determined simultaneously was 147.6 ± 56.4 pg/ml. The ratio of the plasma BNP-LI level to the plasma ANP-LI level was, therefore, 3.0 ± 0.8%. In HPGPC coupled with the RIA with mAb[KY-BNP-II], BNP-LI in the porcine plasma extract was eluted at the elution position of synthetic BNP or BNP-32. Neutralization of the effect of BNP by mAb[KY-BNP-II] Exp 1: effect on blood pressure. The administration of mAb[KY-ANP-II] or mAb[KY-BNP-II] caused no significant change of basal MAP (AMAP: - 2 ± 2 mmHg for mAb[KY-ANP-II] and +2 ± 2 mmHg for mAb[KYBNP-II]). The iv bolus administration of 10 ng BNP in rats of
TABLE 1. Regional distribution of BNP-LF and ANP-LI in the porcine brain Adult (n = 4)
Brain region Olfactory bulb Cerebral cortex Hippocampus Lenticular nucleus Caudate nucleus Thalamus Hypothalamus Midbrain Pons Medulla oblongata Cerebellum Cervical spinal cord Pituitary gland (pooled)
Neonate (n = 3)
BNP-LI
ANP-LI
BNP-LI
ANP-LI
1.12 ± 0.68 0.66 ± 0.34
