Histochemical Journal 30, 425±434 (1998)

Production of monoclonal antibody against histamine and its application to immunohistochemical study in the stomach T A K A S H I D O I 1,2  , K A Z U H I R O T E R A I 1 , I K U O T O O Y A M A 1 , T O S H I I E S A K A T A 2 and HIROSHI KIMURA1 1

Division of Neuroanatomy, Institute of Molecular Neurobiology, Shiga University of Medical Science, Otsu 520-2192, Japan, and 2 Department of Internal Medicine I, School of Medicine, Oita Medical University, Oita 879-5593, Japan

Received 28 November 1997 and in revised form 3 March 1998

Summary A monoclonal antibody against histamine has been produced. A histamine±haemocyanin conjugate prepared using 1ethyl-3-(3-dimethylaminopropyl) carbodiimide as a coupling agent was used for immunizing mice. Immunized mice were sacri®ced to prepare monoclonal antibody using a hybridoma technique. On immunospot assay, the hybridoma culture supernatant containing a monoclonal antibody was capable of detecting 50 pmol of histamine. Using this antibody, we examined the cellular localization of histamine-like immunoreactivity in the stomach of normal or á¯uoromethylhistidine-treated rats and mice. Immunoreactive cells were abundant in the gastric mucosal layer. These positive cells were often located in the basal half of the fundic gland but were rare in the pyloric gland. The cells, small or medium in size, spindle or cone in shape, were intermingled with immunonegative epithelial cells. In the cytoplasm of the positive cells, granular reaction products were densely deposited. In addition, a few positive cells, identi®ed as mast cells by Toluidine Blue staining, were distributed mainly in the submucosal and muscular layer. The antibody preabsorbed with 10 mM histamine gave no positive immunostaining. For pharmacological study, some rats were injected six times with á-¯uoromethylhistidine every 8 h. In these rats, positive cells except mast cells were no longer detected. In conclusion, the monoclonal antibody produced appears to be highly speci®c for histamine. Its application in immunohistochemistry should provide a powerful tool for analysing the roles of histamine in enterochromaf®n-like or mast cells in the stomach. # 1998 Chapman & Hall

Introduction The cellular localization of histamine in the stomach and the brain has been demonstrated by immunohistochemical techniques using speci®c antibodies. Gastric histamine is stored predominantly in the enterochromaf®n-like (ECL) cells, which occur in the basal half of the fundic gland (Panula et al., 1985; HaÊkanson et al., 1986). Histaminergic neurons are localized exclusively in the tuberomammillary nucleus of the posterior hypothalamus (Panula et al., 1984). The antibodies used in previous studies are polyclonal antibodies raised from rabbit and have some problems with regard to tissue ®xation (Panula et al., 1988). Furthermore, these antibodies must be  Address for correspondence: Department of Pathology and Laboratory Medicine, University of Cincinnati Medical Center, 231 Bethesda Avenue, Cincinnati, OH 45267-0529, USA.

0018±2214 # 1998 Chapman & Hall

af®nity puri®ed before use for immunohistochemistry (Panula et al., 1984; Dohlsten et al., 1986). There are no reports of immunohistochemical study to localize histamine using monoclonal antibodies, and this is the goal of this study. Gastric histamine has been proposed as the common mediator of acid secretion (Code, 1965). The histamine secretion from the ECL cells is elevated by gastrin and vagus stimulation (Kahlson et al., 1973), stimulates the parietal cells through histamine H2 -receptor (Black et al., 1972) and, ®nally, activates the H ‡ ,K ‡ -ATPase (Forte & Lee, 1977; Fellenius et al., 1981). However, the physiological role of the gastric histamine and the ECL cells is still not fully understood. á-Fluoromethylhistidine (á-FMH) is a speci®c suicide inhibitor of the histamine-synthesizing enzyme, histidine decarbox-

426 ylase (Kollonitsch et al., 1978). á-FMH inhibits the decarboxylase more speci®cally and potently than any other inhibitor and has been shown to decrease the concentration of histamine in the brain and the stomach (Garbarg et al., 1980; Maeyama et al., 1982). Histamine depletion after á-FMH injection is useful for investigating the roles of histamine in the ECL cells (Andersson et al., 1992). However, the effect of á-FMH on the histamine stored in the ECL cells has not been well established. In the present study, we have produced polyclonal and monoclonal antibodies against histamine. To prove the speci®city of these antibodies, we used them in immunohistochemical studies in the stomach. Then, we applied them to á-FMH-treated rats and mice to investigate the changes in the histamine store in the ECL cells after á-FMH injection. Materials and methods Preparation of immunogen Keyhole limpet haemocyanin (KLH; Sigma, St Louis, MO, USA) was succinylated by adding 20 mg of succinic anhydride to 10 mg of protein dissolved in 2 ml of 0.1 M potassium phosphate buffer, pH 7.0, with constant stirring. The pH was kept between 6.7 and 7.3 for 1 h at room temperature. Then, the succinylated KLH (sKLH) was dialysed against distilled water at 48C overnight. The dialysed sKLH was then lyophilized. An immunogen was prepared by conjugating histamine to sKLH with carbodiimide. Ten milligrams of histamine2HCl (Nacalai, Tokyo, Japan) and 5 mg of sKLH were dissolved in 1.5 ml of distilled water, and 0.1 ml of 1ethyl-3-(3-dimethylaminopropyl) carbodiimide solution (100 mg mlÿ1 ; Nacalai) was added. The solution was kept at room temperature and pH 5.0±6.0 overnight. The conjugate was then dialysed and lyophilized. The method of conjugation is identical to those reported previously (Panula et al., 1984, 1988).

Production of polyclonal antibody against histamine Two female Japanese white rabbits (Kiwa Laboratories, Wakayama, Japan), weighing 2.5±3.0 kg, were used for immunizations. Before immunization, preimmune serum was collected. The animals were injected intradermally with 2 ml of an emulsion containing 250 ìg of the immunogen in 0.15 M saline and 1 ml of Freund's complete adjuvant (Difco, Detroit, MI, USA). Subsequent injections were carried out at 2-week intervals with 2 ml of an emulsion containing 250 ìg of the immunogen and 1 ml of Freund's incomplete adjuvant (Difco). Antisera were collected 1 week after the third booster injection. Af®nity column chromatography was used to obtain puri®ed antibody. Five millilitres of 0.2 M histamine-2HCl dissolved in 0.1 M 3-[N-morpholino]propane-sulphonic acid (pH 7.5) and a 5 ml suspension of Af®-Gel 10 (BioRad, Hercules, CA, USA) were mixed and gently incubated with shaking for 4 h at 48C. Then, 0.5 ml of 1 M ethanolamine-HCl (pH 8.0) was added in order to block

DOI et al. remaining active binding sites of the gel. The gel was packed in a column (6 3 100 mm) and washed with distilled water, 4 M MgCl2 and 0.1 M phosphate-buffered 0.15 M saline (PBS; pH 7.4). After the gel was taken out of the column, 5 ml of the antiserum was added and incubated overnight at 48C. Then, the gel was repacked in the column and washed with 0.1 M PBS. The bound antibody on the column was eluted with 4 M MgCl2 . The elution pro®le was monitored at 280 nm, and a fraction eluted with a void volume contained speci®c polyclonal antibody to histamine. The speci®city of the puri®ed antibody was examined by immunospot assay and immunoabsorption test.

Production of monoclonal antibody against histamine Six female BALB=c mice (Clea Japan, Tokyo, Japan), 5 weeks old, were immunized. The animals were injected intraperitoneally (i.p.) with 0.2 ml of the emulsion containing 100 ìg of the immunogen and 0.1 ml of Freund's complete adjuvant. Subsequent injections at 2-week intervals with 0.2 ml of the emulsion containing 100 ìg of the immunogen and 0.1 ml of Freund's incomplete adjuvant were performed. Blood samples were collected 1 week after each booster injection. The immunoreactivity of the antisera was examined by immunohistochemical study. Three days after the sixth booster with 100 ìg of the immunogen in 0.2 ml of 10 mM PBS, the spleen cells were isolated. Two 3 108 spleen cells were fused by polyethylene glycol with 2 3 107 P3=X63-Ag8.U1 (P3U1) myeloma cells (Yelton et al., 1978). A monoclonal antibody (HAKTD) was produced using a hybridoma technique (Schreier et al., 1980). Hybridoma cells were screened by immunohistochemistry and immunospot assay as described below. The speci®city of the culture supernatant was examined by immunospot assay and immunoabsorption test.

Immunospot assay The immunospot assay was performed according to the method of Larsson (1981). Brie¯y, a 1-ìl solution of a test compound, in an amount ranging from 39 pmol to 10 nmol, was spotted on a gelatin-coated nitrocellulose membrane. The substances tested were histamine-2HCl, Lhistidine, 3-methylhistamine-2HCl, imidazole-4-acetateHCl, serotonin (5HT)-creatinin sulphate, ã-amino-n-butyric acid (GABA), DL-noradrenaline-HCl, dopamine-HCl, Ladrenaline bitartrate, â-alanyl-L-histidine, thyrotropin-releasing hormone, L-methionine, â-alanine, L-asparagine monohydrate, L-glutamine, L-ornithine-HCl, L-isoleucine, L-glutamate monosodium salt, glycine and L-lysine-HCl. The membrane was then ®xed for 1 h with paraformaldehyde vapour at 808C and washed with 0.1 M PBS containing 0.3% Triton X-100 (PBST; pH 7.4). The membrane was incubated overnight at room temperature with (1) the puri®ed rabbit polyclonal antibody against histamine (diluted 1:1000) or (2) the hybridoma culture supernatant containing monoclonal antibody against histamine (diluted 1:10). The membrane was then stained by the same procedure as the immunohistochemistry described below. The exact amount of histamine-2HCl adhered to the membrane was con®rmed by the use of a radioisotope

Histamine monoclonal antibody and immunohistochemistry in stomach ([2,5-3 H]histamine-2HCl; UK).

Amersham,

Buckinghamshire,

Tissue preparations for immunohistochemistry Male ICR strain mice, weighing 35±40 g, and male immature and mature Wistar rats, weighing 90±100 g and 300±400 g respectively (Clea Japan), were used. The animals were housed on a 12 h light±dark cycle (08.00± 20.00 h) and given free access to food and water. Under deep anaesthesia with sodium pentobarbital (50 mg kgÿ1 i.p.; Abbott, North Chicago, IL, USA), the animals were perfused via the ascending aorta with 10 mM PBS followed by a cold ®xative containing 2% paraformaldehyde, 1% glutaraldehyde in 0.1 M phosphate buffer (PB, pH 7.4). Food was withheld from the animals for 24 h before the perfusion. Thirty minutes after the perfusion, the stomach was removed, cut open along the major curvature and stretched out on a rubber plate. After washing for 2 days at 48C with several changes of PB containing 15% sucrose, the specimens were dipped in PB containing 10% gelatin for 6 h at 378C. Gelatin-embedded specimens were immersed at 48C for 2 h in PB containing 4% paraformaldehyde and then washed for 2 days at 48C with several changes of PB containing 15% sucrose. Each stomach segment was frozen on dry ice and cut into 12-ìm-thick sections. The sections were collected in PBST.

Immunohistochemical procedures The sections were treated in a free-¯oating state with 3% H2 O2 in PBST for 20 min to inactivate endogenous peroxidase. After washing, the sections were incubated for 2 days at 48C with (1) the puri®ed rabbit polyclonal antibody against histamine (diluted 1:1000); or (2) the hybridoma culture supernatant containing monoclonal antibody against histamine (diluted 1:10). Sections were washed and incubated for 2 h at room temperature with (1) biotinylated anti-rabbit IgG (diluted 1:1000; Vector, Burlingame, CA, USA); (2) biotinylated anti-mouse IgG, rat adsorbed (diluted 1:500; Vector) or biotinylated anti-mouse IgM (diluted 1:1000; Vector), washed and then incubated for 1 h at room temperature with the avidin±biotin± peroxidase complex (diluted 1:4000; Vector). PBST was used for all washes and for dilution of the antibodies. Colour was developed by reacting the sections for 5 min with a mixture containing 0.02% 3,39±diaminobenzidine (DAB), 0.0045% H2 O2 and 0.3% nickel ammonium sulphate in 50 mM Tris-HCl buffer (pH 7.6). The stained sections were mounted on gelatin-coated glass slides, air dried, counterstained with Kernechtrot, dehydrated and coverslipped.

Double immunostaining Some sections were double immunostained using the monoclonal antibody against histamine and rabbit antiserum against 5HT. Brie¯y, after completion of the ®rst DAB± nickel reaction using the monoclonal antibody against histamine yielding a blue±purple reaction product, sections were treated with 3% H2 O2 in PBST for 20 min to destroy any residual peroxidase activity. Sections were then incubated for the second cycle with rabbit antiserum against 5HT (diluted 1:20 000; Fujimiya et al., 1986).

427

Biotinylated anti-rabbit IgG (diluted 1:1000) was used as the secondary antibody. The 5HT antibody binding was detected similarly to the ®rst cycle, except the nickel ammonium sulphate was eliminated from the DAB solution, yielding a yellow±brown reaction product.

Toluidine Blue staining Mast cells were identi®ed by virtue of their metachromasia when stained with Toluidine Blue. After photography of the immunoreaction, the coverslips were removed and the sections were incubated for 1 h in 1% Toluidine Blue (in 60% ethanol solution), dehydrated and coverslipped again.

Immunoabsorption test Histamine was diluted in PBST at several concentrations ranging from 0.1 mM to 50 mM. Solutions containing 5HT at concentrations of 10 mM were used as controls. The solutions were added to the puri®ed antibody or the culture supernatant at working dilution. The mixture was incubated overnight at 48C and applied for immunohistochemistry.

Distribution of histamine-like immunoreactive (HAIR) cells in the rat gastric mucosa and quantitative morphometric analysis The stomachs from three mature rats were equally divided into nine portions each, along the major curvature. HAIR cells in gastric mucosa were stained by immunohistochemistry using the monoclonal antibody and morphometrically quantitated using ®ve sections from each portion of the stomach. Quanti®cation was carried out with respect to the numerical density (NV ) of HAIR cells in the gastric mucosa on the basis of the mathematical equation given by Floderus (Weibel, 1979).

NV ˆ NA =(d ‡ t ÿ 2h), 2h ˆ d ÿ (d2 ÿ d2 )1=2 where NV is the number of cells per unit volume (numerical density), NA is the number of cells per unit area, d is the mean diameter of the cells, d is the smallest diameter of the cells, h is the cap height and t is the thickness of the section.

In practice, we used 12-ìm-thick cryostat sections, so that t ˆ 12 ìm. Therefore, the values we had to measure were NA , d and d. These parameters were measured using a computer-assisted image analyser (Qube 6000, Nexus, Tokyo, Japan) attached to a light microscope and a highresolution colour video camera.

Bolus injection of á-FMH to mice á-FMH was kindly donated by Dr J. Kollonitsch, Merck Sharp & Dohme Research Laboratories, USA. Matched on the basis of body weight, 13 mice were divided into four groups. The mice were injected i.p. with 0.2 ml of á-FMH solution at doses of 12.5, 25 and 50 mg kgÿ1 . The test solutions were freshly dissolved in 0.15 M saline and adjusted to pH 7.0±7.5 with 1 N sodium hydroxide. For control testing, 0.2 ml of saline was injected. Four hours after the injection, the mice were perfused with the ®xative. The stomach was cut into longitudinal sections

428

DOI et al.

composed of forestomach±corpus±antrum±pylorus. HAIR cells in gastric mucosa were stained by immunohistochemistry using the monoclonal antibody. HAIR cells in the fundic gland area were analysed by quantitative morphometry in ®ve sections from each mouse. Statistical analysis was carried out by one-way analysis of variance. For multiple comparisons, Fisher's PLSD test was used.

Sustained injection of á-FMH to rats Two immature rats were injected i.p. with 0.5 ml of á-FMH solution at a dose of 50 mg kgÿ1 , and the injection was repeated six times at 8-h intervals. Identical injections were carried out using saline instead of á-FMH as procedural controls. Four hours after the last injection, the rats were perfused with the ®xative, and HAIR cells in the stomach were stained by immunohistochemistry using the monoclonal antibody.

Results Speci®city of polyclonal antibody against histamine The highest titre puri®ed polyclonal antibody was capable of detecting was 6.25 nmol of histamine. This antibody cross-reacted with histidine, 3-methylhistamine, 5HT, GABA, noradrenaline, dopamine, alanyl-histidine, methionine, alanine, ornithine, glutamate and lysine, but no cross-reactivity was observed with imidazole acetate or other substances (data not shown). Speci®city of monoclonal antibody against histamine The results of immunospot assay using the monoclonal antibody are shown in Fig. 1. The hybridoma culture supernatant was capable of detecting 156 pmol of histamine when biotinylated anti-mouse IgG was used as the secondary antibody. When antimouse IgM was used instead of anti-mouse IgG, histamine was not detected by the supernatant. The cross-reactivity with 3-methylhistamine or 5HT was 6.25%, and there was no cross-reactivity with other substances. The exact amount of histamine bound to the membrane was con®rmed by the use of a radioisotope. The amount was 1.03 nmol, 0.51 nmol, 169 pmol, 50 pmol or 18 pmol when the spotted amount was 10 nmol, 2.5 nmol, 625 pmol, 156 pmol or 39 pmol respectively. Therefore, the sensitivity of this assay was 50 pmol. Immunostaining by polyclonal antibody The puri®ed polyclonal antibody detected many cells and structures. The immunopositive cells were often located in the basal half of the fundic gland. A few positive cells, identi®ed as mast cells by Toluidine Blue staining, were distributed in the submucosal and muscular layer. Immunopositive structures in the gland lumen were also observed, possibly caused

Fig. 1. Immunospot assay probed with monoclonal antibody (HA-KTD). HA, histamine; mHA, 3-methylhistamine; His, histidine; 5HT, serotonin; IA, imidazole acetate. Test compound solution (1 ìl) in an amount ranging between 39 pmol and 10 nmol was spotted on a gelatincoated nitrocellulose membrane. The hybridoma culture supernatant (diluted 1:10) recognized up to 156 pmol of histamine and up to 2.5 nmol of mHA and 5HT, whereas it fails to recognize as much as 10 nmol of His, IA and other substances.

by cross-reactivity. In the immunoabsorption test, when the antibody was preabsorbed with 50 mM histamine, the immunostaining was weak, but similar immunostaining features were observed. Immunostaining by monoclonal antibody Figs 2 and 3 show the immunohistochemical localization of histamine in the stomach of rats using the monoclonal antibody. HAIR cells were abundant and often located in the basal half of the fundic gland, with very few HAIR cells seen in the pyloric gland (Fig. 2A and C). The cells were small or medium in size, spindle or cone in shape, intermingled with immunonegative epithelial cells and lay directly on the basal lamina of the glands (Fig. 3A and E). In the cytoplasm of the positive cells, granular reaction products were densely deposited (Fig. 3B±D and F). A few immunopositive cells in the submucosal and muscular layer (Fig. 3G) were identi®ed as mast cells by Toluidine Blue staining (Fig. 3H). A few weakly immunopositive mast cells were also observed (Fig. 3G). In the immunoabsorption test, the monoclonal antibody preabsorbed with 10 mM histamine gave no positive immunostaining (Fig. 2B and D). 5HT (10 mM) had no effect on the intensity of the immunohistochemical reaction.

Histamine monoclonal antibody and immunohistochemistry in stomach

429

Fig. 2. Histamine-like immunoreactive (HAIR) cells in horizontal (A and B) and longitudinal (C and D) sections of rat stomach detected by monoclonal antibody against histamine (A and C) and the same antibody preabsorbed with 10 mM histamine (B and D). Many HAIR cells were seen in the basal half of the fundic gland. The HAIR cells were completely abolished after histamine preabsorption. Bar ˆ 200 ìm.

430

DOI et al.

Fig. 3. HAIR cells in rat stomach detected by monoclonal antibody against histamine. (A±D) Horizontal sections of fundic mucosa. (E and F) Longitudinal sections of fundic mucosa. The cells were small or medium in size, spindle (arrows) or cone (arrowheads) in shape, appear singly and lie directly on the basal lamina of the glands. In the cytoplasm, granular reaction products were densely deposited. The cross-reactivity in the gland lumen, seen previously with our polyclonal antibody, was not observed. (G and H) Mast cells in submucosa detected by immunostaining (G) and Toluidine Blue staining (H). Immunopositive (arrows) and immunoweak (arrowheads) mast cells were observed. Bars: A, E, G and H ˆ 50 ìm; B, C, D and F ˆ 25 ìm.

Histamine monoclonal antibody and immunohistochemistry in stomach

Changes in histamine-like immunostaining in the stomach by bolus injection of á-FMH to mice The number of HAIR cells in mice gastric mucosa 4 h after bolus á-FMH i.p. injection is shown in Fig. 5. The number of HAIR cells was signi®cantly decreased to about 45% by a 12.5 or 25 mg kgÿ1 áFMH injection ( p , 0:01 for each) compared with the control group, but did not decrease further with a 50 mg kgÿ1 á-FMH injection. The immunostaining intensity of á-FMH stomachs was weak, but the distribution was not changed by the administration of á-FMH. Effect of sustained injection of á-FMH to rats Fig. 6 shows the changes in HAIR staining by sustained injection of á-FMH. Because the rats used in this experiment were immature, fewer HAIR cells were observed than in the mature rats (Fig. 6A). In sustained á-FMH-treated rats, the HAIR cells in the gastric mucosa were completely abolished (Fig. 6B). Mast cells in the submucosal and muscular layer

1

NA (/mm2)

200

2 3 4

5 6 7 8

9

10

(n 5 3)

150 100

5

NV (3 103/mm3)

15

250

50 0

1

2

3

4

5

6

7

8

9

0

area

Fig. 4. Regional distribution of HAIR cells in rat gastric mucosa. The stomach was equally divided into nine portions along the major curvature. NV , number of cells per unit volume (numerical density). NA , number of cells per unit area. Each column and vertical bar shows the mean  SD. 50 40 NA (/mm2)

Fig. 4 shows the regional distribution of HAIR cells in rats gastric mucosa. HAIR cells were mainly distributed in the corpus region, but sparsely located in the pyloric region. In the corpus region, the density of the cells gradually increased according to the oral side, but the density decreased remarkably near the forestomach. Some sections were double immunostained using monoclonal antibody against histamine and rabbit antiserum against 5HT. The 5HT-like immunoreactive cells were distributed mainly in the intermediate region of the fundic gland and pyloric gland. Their distribution was not similar to that of HAIR cells.

431

30

n54

20

*

10

n53

n53

12.5

25

0

0

* n53

50

α-FMH dose (mg/kg)

Fig. 5. Number of HAIR cells in mice gastric mucosa 4 h after bolus á-¯uoromethylhistidine (á-FMH) injection. NA , number of cells per unit area. Each value represents the mean  SD.  p , 0:01 compared with the saline injection group.

Fig. 6. HAIR cells in the stomach of sustained á-FMH-treated rats. (A) Saline-treated rats as controls. (B±D) á-FMHtreated rats. (D) Stained by Toluidine Blue. HAIR cells in the fundic glands were completely abolished after á-FMH treatment. Mast cells in the submucosa (arrowheads) were detected after á-FMH treatment. Bars: A and B ˆ 200 ìm; C and D ˆ 50 ìm.

432

DOI et al.

were detected as before (Fig. 6C) and identi®ed by Toluidine Blue staining (Fig. 6D). Discussion More than 10 antibodies against histamine have been reported (Table 1). The antibodies that were produced for the measurement of histamine (Peyret, 1986; Guesdon et al., 1986; Morel & Delaage, 1988) cannot be used for immunohistochemistry, as they reacted with histamine derivatives but not with histamine itself. The antibodies for immunohistochemistry (Wilcox & Seybold, 1982; Steinbusch & Mulder, 1984) have cross-reactivity with presumably histidine-containing peptides, such as luteinizing hormone-releasing hormone (Berkenbosch & Steinbusch, 1987). The antibodies produced by Panula et al. (1984) and HaÊkanson et al. (HaÊkanson et al., 1986; Dohlsten et al., 1986) have been used as reliable histochemical tools. Both antibodies need special ®xation to enhance detection sensitivity, that is carbodiimide ®xation for the former (Panula et al., 1988) and diethylpyrocarbonate vapour ®xation for the latter (HaÊkanson et al., 1986). Moreover, as both antibodies are polyclonal antibodies raised from rabbit antisera, they need af®nity puri®cation before immunohistochemical use (Panula et al., 1984; Dohlsten et al., 1986). Therefore, our goal was to produce monoclonal antibodies that were highly sensitive for detecting histamine for immunohistochemical study. In advance of the production of monoclonal antibody, we produced polyclonal antibody against histamine. The polyclonal antibody was con®rmed as detecting histamine according to the results of an immunospot assay. However, the antibody crossreacted with many substances. In immunohistochemistry, the antibody cross-reacted presumably

with mucous in the gastric gland lumen, and the histamine preabsorption could not abolish the staining. To avoid the problem of cross-reactivity, we produced a monoclonal antibody against histamine. The staining by the monoclonal antibody was better de®ned, the glandular luminal staining disappeared and histamine preabsorption abolished the staining completely. According to the results of the immunospot assay, the monoclonal antibody cross-reacted with 5HT. However, by immunohistochemistry, the antibody could not detect the 5HT-containing endocrine cells in the gastric mucosa. With the immunoabsorption test, the monoclonal antibody was absorbed completely by histamine, whereas even very high concentrations of 5HT failed to absorb the antibody. Double immunostaining showed that the histamine-positive cells, except mast cells, were a different population of cells from the 5HT-immunopositive cells. In sustained á-FMH-treated rats, the monoclonal antibody could not detect any positive cells in the gastric mucosa. These ®ndings indicate that the monoclonal antibody is sensitive and selective to histamine and will serve as a powerful tool in immunohistochemical staining for histamine. Gastric histamine in rats is stored mainly in the gastric endocrine cells, so-called ECL cells, in the mucosa and partly in mast cells in the mucosa, submucosa and muscular layer (HaÊkanson et al., 1970, 1986). The gastric endocrine cells appear as single, small to medium-sized cells and lie directly on the basal lamina of the glands. The ECL cells are scattered in the fundic glands, especially in their deep and intermediate regions; very few ECL cells have been found in the neck of the glands. Like other endocrine cells of the fundic mucosa, ECL cells lack any contact with the lumen of the glands,

Table 1. Antibodies against histamine Reference

Year

p=m

Antigen=carrier protein=coupling agents

Animal

Purpose

KovaÂcs & Csaba Wilcox & Seybold Haydik Steinbusch & Mulder Osborne & Patel Panula et al. Plunkett et al. HaÊkanson et al. Peyret

1980 1982 1983 1984 1984 1984 1985 1986 1986

poly poly poly poly poly poly mono poly poly

Rabbit Guinea pig Rabbit Rabbit Rabbit Rabbit Mouse Rabbit Rabbit

Histochemistry Histochemistry RIA Histochemistry Histochemistry Histochemistry RIA Histochemistry RIA

Guesdon et al. Morel & Delaage

1986 1988

mono mono

Aminobenzoyl-azoHA=HSA=diazotization HA=methylated BSA=none Not shown HA=BSA=FA, GA, CDI HA=methylated BSA=none HA=succinylated KLH=EDCDI Not shown HA=HSA=EDCDI HA=BSA or HSA or Hb= hexamethylene diisocyanate, dimethylformamide HA=BSA=1,4-benzoquinone Succinylated HA=BSA=glycyl residues

Mouse Mouse

RIA, ELISA RIA

p=m, polyclonal or monoclonal antibody; poly, polyclonal antibody; mono, monoclonal antibody; HA, histamine; HSA, human serum albumin; BSA, bovine serum albumin; FA, formaldehyde; GA, glutaraldehyde; CDI, carbodiimide; KLH, keyhole limpet haemocyanin; EDCDI, 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide; Hb, haemoglobin; RIA, radioimmunoassay; ELISA, enzyme-linked immunosorbent assay.

Histamine monoclonal antibody and immunohistochemistry in stomach usually being covered by oxyntic or peptic cells; hence, they are called `closed' type (Solcia et al., 1975). In the present study, the HAIR cells in the gastric glands were similar to the ECL cells with regard to the mucosal distribution and the cell morphology. Immunostaining by chromogranin A is necessary to identify the endocrine cells (HaÊkanson et al., 1986; Wilson & Lloyd, 1984), although the staining was not performed in the present study. The mast cells were identi®ed by Toluidine Blue staining. To investigate the changes in histamine stores of ECL cells manipulated by á-FMH, we injected áFMH into the animals and counted the HAIR cells in the gastric mucosa. The histidine decarboxylase activity and histamine concentrations in mice stomach after bolus á-FMH injection reach their lowest level between 1 and 4 h and 4 and 12 h after the infusion respectively (Maeyama et al., 1982). Accordingly, we immunostained the stomach sections that were ®xed 4 h after bolus á-FMH injection. The HAIR cell numbers decreased to the lowest level, about 45% of control group, after 12.5 or 25 mg kgÿ1 á-FMH injections. This result is similar to the previous biochemical study (Maeyama et al., 1982). A bolus á-FMH injection could not decrease the number of HAIR cells dose dependently, but sustained á-FMH injection completely abolished the HAIR cells in the gastric mucosa. These ®ndings indicate that sustained but not bolus á-FMH injection is necessary to abolish the histamine in ECL cells. The reason why the histamine in mast cells was detected even after sustained á-FMH injection seems to be that there is a difference in the histamine turnover between ECL cells and mast cells (Bouclier et al., 1983; Andersson et al., 1992). The present study demonstrated that the monoclonal antibody we have produced is highly speci®c and sensitive for the detection of histamine. The immunohistochemical application of this antibody was useful for analysing the morphological localization of histamine in the stomach of normal or áFMH-treated rats and mice. This antibody should be a powerful tool for investigating the roles of histamine in ECL cells or mast cells.

433

BLACK, J.W., DUNCAN, W.A.M., DURANT, C.J., GANELLIN,

(1972) De®nition and antagonism of histamine H2 -receptors. Nature 236, 385±90. B O U C L I E R , M . , J U N G , M . J . & G E R H A R T , F . (1983) Effect of prolonged inhibition of histidine decarboxylase on tissue histamine concentrations. Experientia 39, 1303±5. C O D E , C . F . (1965) Histamine and gastric secretion: a later look, 1955±1965. Fed. Proc. 24, 1311±21. Ê K A N S O N , R . (1986) DOHLSTEN, M., HALLBERG, T. & HA Production and characterization of rabbit antibodies against histamine. Mol. Immunol. 23, 799±806. C.R. & PARSONS, E.M.

FELLENIUS,

E.,

ELANDER,

BERGLINDH, T., SACHS, G., OLBE, L., È STRAND, S.E. & WALLMARK, B. SJO

B.,

(1981) Substituted benzimidazoles inhibit gastric acid secretion by blocking (H ‡ ‡K ‡ )ATPase. Nature 290, 159±61. F O R T E , J . G . & L E E , H . C . (1977) Gastric adenosine triphosphatases: a review of their possible role in HCl secretion. Gastroenterology 73, 921±6. F U J I M I Y A , M . , K I M U R A , H . & M A E D A , T . (1986) Postnatal development of serotonin nerve ®bers in the somatosensory cortex of mice studied by immunohistochemistry. J. Comp. Neurol. 246, 191±201. GARBARG, M., BARBIN, G., RODERGAS, E. & SCHWARTZ, J . C . (1980) Inhibition of histamine synthesis in brain by á-¯uoromethylhistidine, a new irreversible inhibitor: in vitro and in vivo studies. J. Neurochem. 35, 1045±52.

G U E S D O N , J . L . , C H E V R I E R , D . , M A Z I EÂ , J . C . , D A V I D , B . &

(1986) Monoclonal anti-histamine antibody: preparation, characterization and application to enzyme immunoassay of histamine. J. Immunol. Methods 87, 69±78.

AVRAMEAS, S.

Ê KANSON, R., OWMAN, C., SJO È BERG, N.O. & SPORRONG, HA

(1970) Amine mechanisms in enterochromaf®n and enterochromaf®n-like cells of gastric mucosa in various mammals. Histochemie 21, 189±220.

B.

Ê KANSON, R., BO È TTCHER, G., EKBLAD, E., PANULA, P., HA SIMONSSON, M., DOHLSTEN, M., HALLBERG, T. &

(1986) Histamine in endocrine cells in the stomach. Histochemistry 86, 5±17. H A Y D I K , I . B . (1983) Histamine determination by radioimmunoassay (RIA) [abstract]. J. Allergy Clin. Immunol. 71, 152. K A H L S O N , G . , R O S E N G R E N , E . & S V E N S S O N , S . E . (1973) Histamine and gastric secretion with special reference to the rat. In: Pharmacology of Gastrointestinal Motility and Secretion, Vol. 1 (edited by H O L T O N , P . ) pp. 41± 102. Oxford: Pergamon Press. SUNDLER, F.

KOLLONITSCH, J., PERKINS, L.M., PATCHETT, A.A., DOLDOURAS, G.A., MARBURG, S., DUGGAN, D.E., MAY-

References Ê KANSON, R., MATTSSON, H. ANDERSSON, K., CHEN, D., HA

(1992) Enterochromaf®n-like cells in the rat stomach: effect of á-¯uoromethylhistidineevoked histamine depletion. Cell Tissue Res. 270, 7±13. B E R K E N B O S C H , F . & S T E I N B U S C H , H . W . M . (1987) Histamine-immunostaining in the rat median eminence: an unexpected form of cross-reactivity with LH-RH. Brain Res. 405, 353±7. & SUNDLER, F.

C O C K , A . L . & A S T E R , S . D . (1978) Selective inhibitors of biosynthesis of aminergic neurotransmitters. Nature 274, 906±8. Â C S , P . & C S A B A , G . (1980) An immunocytochemical KOVA method for histamine detection. Acta Histochem. 66, 50±8. L A R S S O N , L . I . (1981) A novel immunocytochemical model system for speci®city and sensitivity screening of antisera against multiple antigens. J. Histochem. Cytochem. 29, 408±10.

434

DOI et al.

MAEYAMA, K., WATANABE, T., TAGUCHI, Y., YAMATODA-

(1982) Effect of á-¯uoromethylhistidine, a suicide inhibitor of histidine decarboxylase, on histamine levels in mouse tissues. Biochem. Pharmacol. 31, 2367±70. M O R E L , A . M . & D E L A A G E , M . A . (1988) Immunoanalysis of histamine through a novel chemical derivatization. J. Allergy Clin. Immunol. 82, 646±54. O S B O R N E , N . N . & P A T E L , S . (1984) Localisation of histamine-immunoreactivity in speci®c neurones of the gastropod CNS including a cell which has been identi®ed as histaminergic. Histochemistry 81, 207±8. P A N U L A , P . , Y A N G , H . Y . T . & C O S T A , E . (1984) Histaminecontaining neurons in the rat hypothalamus. Proc. Natl Acad. Sci. USA 81, 2572±6. NI, A. & WADA, H.

È CKLIN, M. & COSTA, E. PANULA, P., KAARTINEN, M., MA

(1985) Histamine-containing peripheral neuronal and endocrine systems. J. Histochem. Cytochem. 33, 933±41.

È LA È PPO È , O., AIRAKSINEN, M.S., AUVINEN, PANULA, P., HA È K I , A . (1988) Carbodiimide as a tissue S. & VIRKAMA

®xative in histamine immunohistochemistry and its application in developmental neurobiology. J. Histochem. Cytochem. 36, 259±69.

PEYRET, L.M., MOREAU, P., DULLUC, J. & GEFFARD, M.

(1986) Antibodies to histamine: speci®city studies and radioimmunological assay. J. Immunol. Methods 90, 39±45.

PLUNKETT,

G.,

DAILEY,

F.,

FICALORA,

J.,

LALLY,

P.,

M U E L L E R , B . , N A U G L E R , D . & B U C K L E R , R . (1985) An immunoassay for measuring histamine (abstract). Ann. Allergy 55, 230. SCHREIER, M., KONLER, G., HENGARTNER, H., BEREK, C., TRUCCO, M., FORNI, L., STAEHELIN, T., STOCKER, J. &

B . (1980) Hybridoma Techniques. EMBO, SKMB Course, Basle. S O L C I A , E . , C A P E L L A , C . , V A S S A L L O , G . & B U F F A , R . (1975) Endocrine cells of the gastric mucosa. Int. Rev. Cytol. 42, 223±86. S T E I N B U S C H , H . W . M . & M U L D E R , A . H . (1984) Immunohistochemical localization of histamine in neurons and mast cells in the rat brain. In: Handbook of Chemical Neuroanatomy. Vol. 3. Classical Transmitters and Transmitter Receptors in the CNS, part II (edited by B J OÈ R K È K F E L T , T . & K U H A R , M . J . ), pp. 126±140. LUND, A., HO Amsterdam: Elsevier. W E I B E L , E . R . (1979) Elementary introduction to stereological principles. In: Stereological Methods, Vol. 1, pp. 9±62. London: Academic Press. W I L C O X , B . J . & S E Y B O L D , V . S . (1982) Localization of neuronal histamine in rats brain. Neurosci. Lett. 29, 105±10. W I L S O N , B . S . & L L O Y D , R . V . (1984) Detection of chromogranin in neuroendocrine cells with a monoclonal antibody. Am. J. Pathol. 115, 458±68. TAKACS,

YELTON, D.E., DIAMOND, B.A., KWAN, S.P. & SCHARFF,

(1978) Fusion of mouse myeloma and spleen cells. Curr. Topics Microbiol. Immunol. 81, 1±7.

M.D.