- Email: [email protected]
Distribution of bombesin binding sites in the rat gastrointestinal tract
Peptides, Vol. 9, pp. 643-649. ©Pergamon Press plc, 1988. Printed in the U.S.A.
0196-9781/88 $3.00 + .00
Distribution of Bombesin Binding Sites in the Rat Gastrointestinal Tract T I M O T H Y H . M O R A N , 1 T E R R Y W . M O O D Y , * A N N E M. H O S T E T L E R , P A U L H . R O B I N S O N , M I C H A E L G O L D R I C H A N D P A U L R. M c H U G H
Department of Psychiatry and Behavioral Sciences Johns Hopkins University School of Medicine, Baltimore, MD 21205 *Department of Biochemistry, George Washington University School of Medicine and Health Sciences, Washington, DC 20037 R e c e i v e d 14 S e p t e m b e r 1987 MORAN, T. H., T. W. MOODY, A. M. HOSTETLER, P. H. ROBINSON, M. GOLDRICH AND P. R. McHUGH.
Distribution of bombesin binding sites in the rat gastrointestinal tract. PEPTIDES 9(3) 643-649, 1988.--In an attempt to identify potential target sites for the satiety action of bombesin (BN), the distribution and pharmacological specificity of bombesin binding sites were examined in the rat gastrointestinal tract by in vitro autoradiography utilizing (tzsI-Tyr4) bombesin. Specific BN binding was localized to the circular muscle level of the gastric fundus and antrum, submucosal layer of the small intestine and longitudinal and circular muscle and submucosal layers of the colon. Pharmacological studies indicated that gastrin releasing peptide (GRP), Ac-GRP2°-27 and BN-like compounds, litorin and ranatensin, inhibited the binding of (125I-Tyr4)BN with high affinity while compounds which lacked COOH-terminal homology with BN demonstrated a low affinity for BN binding sites. The wide distribution of BN binding sites in the gastrointestinal tract provides a number of potential sites for the mediation of the satiety action of BN. Receptors
Peptides
Stomach
Intestine
were sacrificed and gastrointestinal segments from the gastric fundus, antrum, pyloric canal, pyloric sphincter and from the intestinal duodenum, jejunum, ileum and ascending, transverse and descending colon were rapidly removed and frozen in isopentane at -70°C for 15 sec. Frozen tissue segments were sectioned at a thickness of 25/zm at - 14°C, thaw mounted to cold gelatin-coated slides and dried in a desiccator under partial pressure. Slide mounted tissue sections were stored at -70°C until binding. Slides were incubated in 10 mM H E P E S , pH 7.4, 130 mM NaCI, 4.7 mM KC1, 5 mM MgCI2, 1 mM EGTA, 0.05% bovine serum albumin (BSA) 0.07 mM bacitracin and (125ITyr4)BN in the presence or absence of 1/zM unlabelled BN. Following incubation at 24°C, free radiolabelled peptide was removed by two consecutive 4-rain washes in 10 mM H E P E S (pH 7.4) containing 0.05% BSA at 4°C. Sections were either wiped from slides and assayed for radioactivity in a gamma counter or washed slides were dried with a stream of cold air and placed in a desiccator under partial vacuum pressure for 12 hr to insure total drying. Dried slides were placed in an x-ray cassette and opposed to sheets of L K B Ultrofilm. The binding conditions employed in these experiments were similar to those utilized by Wolfet al. [22] and Zarbin et al. [24] on brain sections. Binding under these conditions has been shown to be of high affinity, saturable, reversible on brain sections. Preliminary aspects of the present study
B O M B E S I N (BN), a tetradecapeptide originally isolated from amphibian skin, has been shown to be biologically active in mammalian gastrointestinal tract [1, 3, 9, 17] and brain [5,16]. Immunoreactivity to bombesin and structurally related compounds such as gastrin releasing peptide (GRP) has been demonstrated in neurons of the gastrointestinal tract [2,15], and BN receptors have been identified in brain tissue [11-13, 22-24]. BN has been shown to affect food intake [4,8] although the exact site of action or the mechanism for this effect of bombesin has not been identified. Neural disconnection of the gut via combined dorsal rhizotomy at T3-6, cord section at T6 and bilateral subdiaphragmatic vagotomy eliminated the satiety effect of bombesin suggesting a gastrointestinal site for this behavioral effect [ 19]. In order to identify potential sites for the satiety action of bombesin, we have mapped the distribution of BN binding sites in the rat gastrointestinal tract utilizing in vitro autoradiography with (125I-Tyr 4) bombesin and have pharmacologically characterized these binding sites. A preliminary report of these experiments has appeared [14]. METHOD BN was iodinated using the chloramine T procedure and purified using gel filtration techniques [11] to a specific activity of 1200 Ci/mmol. Male Sprague-Dawley rats 300-350 g
1Requests for reprints should be addressed to Timothy H. Moran, PhD., Department of Psychiatry and Behavioral Sciences, Meyer 4-119, Johns Hopkins University School of Medicine, 600 N. Wolfe St., Baltimore, MD 21205.
643
644
M O R A N E T AL.
1500-
6A Totol
Toto] o NonspeciFic • Specific
,, N o n s p e c i f i c
Q (~
i
"G
100O"
4-
v
Z-
500&
I
I
I
30
60
90
(TYR-4)BN (riM)
1
120
Time (min) FIG. 1. Association of (1z~I-TyrgBN to rat fundal sections. Total (filled triangles) and nonspecific (open triangles) binding were determined as a function of time after the addition of (~zSI-TyrgBN. 3
DISTRIBUTION OF BOMBESIN BINDING SITES 21.0"
•
ng 3
0.8'
1 (/) znl a
0.6 ¸
0
0.4
0 0
I 1
i
i
2
3
BN Bound (fmol/~ection)
0
FIG. 2. Top: Binding of (125I-TyrgBN as a function of radiolabelled peptide concentration. Total (open triangles) and nonspecific (open squares) binding were determined in response to increasing concentrations of (mI-TrygBN, Specific binding (filled circles) represents the difference between total and nonspecific. Bottom: Skatchard plot of the specific binding. The line represents the best fit assuming a single class of binding sites (Kd=l nM).
0.2
0.0
,. =
m
'o
_o (3.
o
3,
~
=
<
~
F-
a
FIG. 3. Relative densities of (I~5I-TyrOBNbinding to tissue sections from various gastrointestinal levels demonstrating total and nonspecific binding. Values indicate the average density across the cellular layers within a segment expressed as optical density from computerized microdensitometry. Significant differences between total and nonspecific binding indicated by *p <0.05. evaluated the dynamics and equillibrium of bombesin binding to gastric fundal sections. Autoradiographic studies were conducted utilizing 1 nM (12~I-TyrgBN and incubating for 60 rain. Autoradiographic images from the developed films, demonstrating total and nonspecific binding, were directly compared with the corresponding tissue sections which were stained with toluidine blue. Superimposition of the stained slides with their corresponding autoradiographs allowed an accurate anatomical localization of the specific BN binding sites. Microden-
sitometry of autoradiographs was carried out on images utilizing an Amersham RAS1000 computerized microdensitometer. Sections demonstrating total binding and the adjacent sections in which binding had been inhibited with 1 /xM unlabelled BN were compared for levels of binding as indicated by optical density from the various tissue segments. Six samples of each segment were examined. In a second series of studies, tissue samples from the gastric fundus were sectioned at a thickness of 25 ~ m and thaw mounted to gelatin-coated slides. Four sections were mounted per slide. Slides were incubated as in the previous experiment in the presence or absence of various concentrations (10 -5 to 10-11 M) of BN analogs to determine the ability of these compounds to inhibit BN binding. These compounds included gastrin releasing peptide (GRP), acetylated GRP 2°-2r (Ac-GRp2°-zT), ranatensin, litorin, bombesin hydroxide (BN-OH), spantide and bombesin fragment 8--14 (BN 1-14). Following washes, sections were wiped from slides
G A S T R O I N T E S T I N A L B O M B E S I N B I N D I N G SITES
645
C
FIG. 4. Photomicrographs of toluidine blue stained gastric sections and corresponding autoradiographs showing total and nonspecific BN binding. (A) Cross section of gastric fundus. (B) Autoradiograph of section in (A) showing total ('2~I-Tyr4)BN binding. (C) Autoradiograph of section consecutive to that in A, showing nonspecific BN binding with 1/zM unlabelled BN added to the incubation buffer. (D) Cross section from gastric antrum• (E and F) Autoradiographs showing total and nonspecific BN binding, respectively corresponding to (D). Specific binding is localized to the circular muscle layer in both segments• CM, circular muscle; LM, longitudinal muscle; M, mucosa.
and assayed for radioactivity using a gamma counter. The ability of each concentration of analog to inhibit BN binding was examined in three separate experiments. RESULTS The kinetics of association of ('25I-Tyr4)BN to rat fundal slices is demonstrated in Fig. 1. Specific binding increased rapidly the first 60 min, slowly over the next 30 min and after 90 min, equilibrium was attained. In contrast, nonspecific binding in the presence of 1 /~M unlabelled BN increased only slowly through this period. The concentration dependence of ('25I-Tyr4)BN to rat fundal sections is demonstrated in Fig. 2A. Total binding and nonspecific binding increased with increasing concentrations of ('25I-Tyr4)BN, while specific binding was saturable. Scatchard analysis (Fig. 2B) indicated that BN binding was occurring to a single class of sites ( K d = l nM).
Distribution of Bombesin Binding Sites The relative density of (Tyr 4) BN binding across the various gastrointestinal segment is shown in Fig. 3. Overall, the
highest density of binding was evident in the gastric fundus. Significant levels of specific binding were also found in the antrum. Within the small intestine, significant binding was evident in the duodenum, jejunum, and ileum. All segments of the colon (ascending, transverse and descending) demonstrated specific binding. No specific binding was evident in the pyloric canal or pyloric sphincter. The specific localization of BN binding sites varied across the tissue segments in which they were found. In the fundus (Fig. 4A-C), high binding densities were localized to the circular muscle layer. No obvious increases in grain density were evident in the serosal, submucosal, or mucosal layers (Fig. 4B). Binding to the gastric antrum (Fig. 4D-F) was also localized primarily to the circular muscle. Nonspecific binding was evident in the submucosal and mucosal layers. In the duodenum (Fig. 5A-C), high levels of nonspecific binding were found in the mucosa (as in other small intestinal segments). Specific binding was localized to the circular muscle and the submucosal layer. Localization of BN binding sites to the submucosa was also evident in the jejunum (Fig. 5D-F) and ileum (Fig. 5G-I). Within the colon, similar distributions were found in the
M O R A N ET AL.
646
H
L
"
=Atr~.
FIG. 5. Photomicrographs of toluidine blue stained small intestinal sections and corresponding autoradiographs showing total and nonspecific BN binding. (A) Cross section from duodenum. (B and C) Autoradiographs showing total and nonspecific BN binding. Specific binding is localized to the submucosal and circular muscle layers. (D) Cross section of jejunum. (E and F) Autoradiographs showing total and nonspecific BN binding. Specific binding is localized to the submucosal layer. (G) Cross section of ileum. (H and I) Autoradiographs showing total and nonspecific BN binding. Specific binding is localized to the submucosal layer. CM, circular muscle; LM, longitudinal muscle; IG, intestinal glands; SM, submucosa. ascending (Fig. 6A-C) and transverse segments (Fig. 6D-F), with highest densities localized to the longitudinal muscle and the submucosal layer. Lower levels of specific binding were found in the circular muscle layer. In the descending colon (Fig. 6G-I), highest densities were evident in the circular muscle layer.
Pharmacological Specificity Figure 7 demonstrates that BN and structurally related peptides inhibited (125I-Tyr4)BN binding to the gastric fundus in a concentration dependent manner. The relative potencies of these compounds, as measured by the concentration required to inhibit 50% of the (n5I-Tyr4)BN binding (IC~0), are shown in
Table 1. Compounds such as GRP, Ac-GRP, litorin and ranatensin, with the C-terminal of bombesin essentially intact were equally or more potent than bombesin at the fundal sites• Compounds in which the C-terminal was altered such as BN 8-14, BN-OH and spantide, were not potent inhibitors of (125I-Tyr4) BN binding. DISCUSSION These results demonstrate a wide distribution of bombesin binding sites in the rat gastrointestinal tract. The particular cellular layer in which the binding sites are found depends upon the particular gastrointestinal segment• In the stomach, a gradient of receptors was evident from a high density in the
G A S T R O I N T E S T I N A L B O M B E S I N B I N D I N G SITES
647
°~ ' % e .
,t~ 7
'~2¢
2:
~ .
.
, ,e
-
v , ..
k ",.
v,
..~
. ,;-d 1."
~e,
I
I
f
•
~ f
4
te~
21 "~j
..
FIG. 6. Photomicrographs of toluidine blue stained large intestinal sections and corresponding autoradiographs showing total and nonspecific BN binding. (A) Cross section for ascending colon. (B and C) Autoradiographs showing total and nonspecific binding. Specific binding is localized to the circular and longitudinal muscle and the submucosal layers. (D) Cross section of transverse colon. (E and F) Autoradiographs showing total and nonspecific binding. Specific binding is localized to the longitudinal and circular muscle and submucosal layers. (G) Cross section of descending colon. (H and I) Autoradiographs of total and nonspecific binding. Specific binding is localized to the circular muscle layer.
circular muscle layer of the fundus, lower density in the antrum and a lack of discernible specific binding in the region of the pylorus. In the small intestine, receptors were predominantly localized to the submucosal layer, with some specific binding of the circular muscle layer in the duodenum and ileum. In the large intestine, high levels of specific binding were evident in the longitudinal muscle and submucosal layers with less dense binding in the circular muscle, although the descending colon had very dense binding in the circular muscle region. Bombesin receptors have previously been characterized in rat brain [11-13, 22-24], guinea pig pancreas [7] and in pituitary cell culture [21]. Although a similar af~nity for
bombesin is evident in these systems, differences are found in the relative affinity for BN-like peptides. In the present study, GRP, Ac-GRP 2°-27, litorin and ranatensin all were more potent than bombesin in inhibiting the binding of 1251T y # ) B N . In rat brain and guinea pig pancreas, litorin and ranatensin have been reported to have only 5-25% of the potency of B N [7,11], while in rat pituitary, these compounds were slightly more potent than BN [21]. Also, in the rat brain and pancreas, spantide is far more potent (IC50=1 /zM) than at fundal binding sites (IC50>10/~M). Thus, these fundal binding sites appear to be more similar to pituitary than brain or pancreatic B N receptors. However, fundal binding sites also demonstrate a greater affinity for GRP
648
MORAN
INHIBITIONOF 1251-TYR4 BN BINDING
TABLE 1 PHARMACOLOGY OF BOMBESIN AND RELATED PEPTIDES
100 i
80
-~-'-~ ~
.":' " L-~, _ ~'t'-.. "~ "~ ~ ~
ZXBN8-14 " BN-OH
20
" Utorln
O AcGRP • GRP
.c_
-11 -10
\ \ ~"~ \\ \
\
\
-9 -8 -7 -6 CONCENTRATION (log M)
Peptide Litorin Ac-GRP~°-~7 GRP Ranatensin Bombesin BN 8-14 BN-OH Spantide
60 40
ET AL.
IC5o (nM) 1.0 _+ 0.6 1.1 + 0.5 2.1 -+ 0.9 7.8 --+ 1.1 20.0 _+ 5.0 300.0 + 102 >10,000 > 10,000
-5
FIG. 7. Pharmacology of fundal (Tyr4)BN binding sites. The percentage of (lzSI-Tyr4)BN specifically bound is plotted as a function of unlabelled peptide concentration.
(10×) and Ac-GRP 2°-~7 (20x) than for BN, a result which was not evident for pituitary BN receptors where GRP was slightly less potent than BN [21]. This pattern of results suggests that these fundal BN binding sites may represent an additional subtype of BN receptor, different from those which have previously been characterized in brain, pancreas and pituitary, although the time course and Ko of (Tyr4)BN binding to fundal sections are similar to those which have been demonstrated for binding to brain sections [23,24]. Whether the BN binding sites in other gastrointestinal segments demonstrate similar affinities for these BN analogs has not been determined. The particular cellular element containing the BN binding sites cannot be discerned from this type of study. Binding sites may be directly on muscle fibers, or on neural elements presynaptic to these fibers. Furthermore, it may be that binding sites are on the muscle in some regions and on neural elements in others, such as in the submucosal layer. In vivo and in vitro studies have suggested all possibilities. A direct smooth muscle excitation of gastric circular muscle by BN was demonstrated by Mayer et al. [9] utilizing an in vitro preparation. In contrast, in vivo experiments have suggested that BN and GRP receptors are on nerves innervating the circular muscle of the stomach [3]. In the small intestine, receptors appear to be on inhibitory nerves, not on smooth muscle, as responses to BN and GRP are blocked by tetrodotoxin [3]. Bombesin immunoreactivity, localized to nerve fibers, has been demonstrated in various gastrointestinal segments in the rat and guinea pig [2,20]. In the rat stomach, these nerve fibers are found in high concentration in the gastric fundus and less so in the gastric antrum and appear to be localized to the mucosal layer. This pattern is also evident in levels of extractable BN [20]. The distribution of BN and BN binding sites, although similar in terms of relative density within the fundus and antrum, differ in that binding sites are localized to the muscle layer and BN itself is concentrated in the mucosal layer. In the small intestine, BN immunoreactivity is found in the myenteric plexus and within the muscle layers
Values indicate mean _+ SE from 3 determinations.
of the duodenum, jejunum, ileum and colon [20], a similar distribution to that of BN binding sites. The distribution of BN binding sites in the rat gastrointestinal tract differs from that found for cholecystokinin (CCK), another peptide which can potently inhibit food intake. In the rat, gastrointestinal C C K receptors are limited to the circular muscle layer of the pyloric sphincter, and a mechanism for the satiety action of CCK through the interaction of CCK with this receptor population has been suggested [18]. No such specific localization or potential function is evident for BN. The distribution of BN binding sites presents a wide range of possibilities of sites and mechanisms for the mediation of the satiety action of bombesin. Preliminary data has suggested a variety of sites and potential mechanisms for BN satiety. Complete gastrointestinal deafferentation involving bilateral subdiaphragmatic vagotomy, rhizotomy and cord section at the level of T6 eliminates the satiety action of BN [19]. Although this finding supports a peripheral site of action for BN, it does not identify a specific pathway or mechanism of action. Furthermore, partial gastrointestinal deafferentation by either extensive rhizotomy (T3-T10) alone or rhizotomy and cord transection without vagotomy result in a partial loss of BN satiety [10] suggesting the possibility for multiple sites of action or additive effects. The satiety action of BN may be direct or may be secondary to some BN mediated gastrointestinal action. BN has been reported to inhibit gastric emptying in the rat following both central [5,16] and peripheral administration [1,17], although the inhibition following peripheral administration is not a universal finding and may be strain specific [6,16]. Some component of the satiety action of bombesin may be secondary to this gastric inhibitory effect of BN, but, as yet, no mechanism for the satiety action of BN has been demonstrated. Thus, any or all the gastrointestinal binding sites demonstrated in the present work may be among the candidate target sites for the mediation of BN's satiety action. In summary, the present results identify a wide distribution of bombesin binding sites in the rat gastrointestinal tract, through different cellular layers in different tissues. This pattern of results suggests a general role for BN in coordinating gastrointestinal function. In contrast to the gastrointestinal localization of CCK receptors, the extensive distribution of BN binding sites does not clearly identify a locus for the peripheral mediation of the satiety action of BN.
GASTROINTESTINAL
BOMBESIN
BINDING SITES
649
REFERENCES 1. Bertaccini, G., M. Impicciatore. Action of bombesin on the 14. Moran, T. H., T. W. Moody, M. S. Goldrich, P. H. Robinson and P. R. McHugh. Autoradiographic localization of bombesin motility of the stomach. Naunyn Schmiedebergs Arch Pharbinding sites within the gastrointestinal tract. Soc Neurosci macol 289: 149-156, 1975. 2. Costa, M., J. B. Furness, N. Yanalhara, C. Yanaihara and T. Abstr 11: 903, 1985. 15. Polak, J. M., S. R. Bloom, S. Hobbs, E. Solcia and A. G. E. W. Moody. Distribution of projections of neurons with immunoPearse. Distribution of a bombesin-like peptide in human gasreactivity for both gastrin-releasing peptide and bombesin in the guinea-pig small intestine. Cell Tissue Res 235: 285-293, 1984. trointestinal tract. Lancet 1:1109-1110, 1976. 16. Porreca, F. and T. F. Burks. Centrally administered bombesin 3. Fox, J. E. T. and T. J. McDonald. Motor effects of gastrin releasing peptide (GRP) and bombesin in the canine stomach affects gastric emptying and small and large bowel transit in the rat. Gastroenterology 85: 313-317, 1983. and small intestine. Life Sci 35: 1667-1673, 1984. 17. Scarpignato, C. and G. Bertacinni. Bombesin delays gastric 4. Gibbs, J., D. J. Fauser, E. A. Rowe, B. J. Rolls, E. T. Rolls and S. P. Maddison. Bombesin suppresses feeding in rats. Nature emptying in the rat. Digestion 21: 104-106, 1981. 282: 208-210, 1979. 18. Smith, G. T., T. H. Moran, J. T. Coyle, M. J. Kuhar, T. L. 5. Gmerek, D. E. and A. Cowan. Pituitary-adrenal mediation of O'Donohue and P. R. McHugh. Anatomical localization of bombesin-induced inhibition of gastrointestinal transit in rats. cholecystokinin receptors in the pyloric sphincter. Am J Physiol Regul Pept 9: 299-304, 1984. 246: R127-R130, 1984. 19. Stuckey, J., J. Gibbs and G. P. Smith. Neural disconnection of 6. Hostetler, A. M., T. H. Moran and P. R. McHugh. Bombesin: Gastric emptying and feeding. Proc East Psychol Assoc 57: 57, gut from brain blocks bombesin-induced satiety. Peptides 6: 1986 (abstract). 1249-1252, 1985. 20. Walsh, J. H., J. R. Reeve and S. R. Vigna. Distribution and 7. Jensen, R. T., T. W. Moody, C. Pert, J. E. Rivier and J. D. molecular forms of mammalian bombesin. In: Gut Hormones, Gardner. Interaction of bombesin and litorin with specific memedited by S. R. Bloom and J. M. Polak. Edinburgh: Churchill brane receptors on pancreatic acinar cells. Proc Natl Acad Sci Livingstone, 1981, pp. 414-418. USA 75: 6139-6143, 1978. 21. Westendorf, J. M. and A. Schonbrunn. Characterization of 8. Kulkosky, P. J., J. Gibbs and G. P. Smith. Behavioral effects of bombesin receptors in a rat pituitary cell line. J Biol Chem 258: bombesin administration in rats. Physiol Behav 28: 505-512, 1982. 7527-7535, 1983. 22. Wolf, S. S., T. W. Moody, T. L. O'Donohue, M. A. Zarbin and 9. Mayer, E. A., J. Elashoff and J. H. Walsh. Characterization of M. J. Kuhar. Autoradiographic visualization of rat brain sites bombesin effects on canine gastric muscle. A m J Physiol 243: for bombesin-like peptides. Eur J Pharmacol 87: 163-164, 1983. G141-G147, 1982. 23. Wolf, S. S. and T. W. Moody. Receptors for GRP/Bombesin10. Mindell, S., J. DiPoala, S. Wiener, J. Gibbs and G. P. Smith. like peptides in the rat forebrain. Peptides 6: 111-114, 1985. Satiety effect on bombesin is attenuated by dorsal rhizotomy 24. Zarbin, M. A., M. J. Kuhar, T. L. O'Donohue, S. S. Wolf and and spinal cord transection. Proc East Psychol Assoc 57: ll, R. W. Moody. Autoradiographic localization of (~25I-Tyr4) 1986. bombesin binding sites in the rat brain. J Neurosci 5: 429-437, 11. Moody, T. W., C. B. Pert, J. Rivier and M. R. Brown. Bombe1985. sin: Specific binding to rat brain membranes. Proc Natl Acad Sci USA 75: 5372-5376, 1978. 12. Moody, T. W., T. L. O'Donohue and D. M. Jacobowitz. Biochemical localization and characterization of bombesin-like peptides in discrete regions of rat brain. Peptides 2: 75-79, 1981. 13. Moody, T. W., J. N. Crawley and R. T. Jensen. Pharmacology and neurochemistry of bombesin-like peptides. Peptides 3: 559-563, 1982.
0196-9781/88 $3.00 + .00
Distribution of Bombesin Binding Sites in the Rat Gastrointestinal Tract T I M O T H Y H . M O R A N , 1 T E R R Y W . M O O D Y , * A N N E M. H O S T E T L E R , P A U L H . R O B I N S O N , M I C H A E L G O L D R I C H A N D P A U L R. M c H U G H
Department of Psychiatry and Behavioral Sciences Johns Hopkins University School of Medicine, Baltimore, MD 21205 *Department of Biochemistry, George Washington University School of Medicine and Health Sciences, Washington, DC 20037 R e c e i v e d 14 S e p t e m b e r 1987 MORAN, T. H., T. W. MOODY, A. M. HOSTETLER, P. H. ROBINSON, M. GOLDRICH AND P. R. McHUGH.
Distribution of bombesin binding sites in the rat gastrointestinal tract. PEPTIDES 9(3) 643-649, 1988.--In an attempt to identify potential target sites for the satiety action of bombesin (BN), the distribution and pharmacological specificity of bombesin binding sites were examined in the rat gastrointestinal tract by in vitro autoradiography utilizing (tzsI-Tyr4) bombesin. Specific BN binding was localized to the circular muscle level of the gastric fundus and antrum, submucosal layer of the small intestine and longitudinal and circular muscle and submucosal layers of the colon. Pharmacological studies indicated that gastrin releasing peptide (GRP), Ac-GRP2°-27 and BN-like compounds, litorin and ranatensin, inhibited the binding of (125I-Tyr4)BN with high affinity while compounds which lacked COOH-terminal homology with BN demonstrated a low affinity for BN binding sites. The wide distribution of BN binding sites in the gastrointestinal tract provides a number of potential sites for the mediation of the satiety action of BN. Receptors
Peptides
Stomach
Intestine
were sacrificed and gastrointestinal segments from the gastric fundus, antrum, pyloric canal, pyloric sphincter and from the intestinal duodenum, jejunum, ileum and ascending, transverse and descending colon were rapidly removed and frozen in isopentane at -70°C for 15 sec. Frozen tissue segments were sectioned at a thickness of 25/zm at - 14°C, thaw mounted to cold gelatin-coated slides and dried in a desiccator under partial pressure. Slide mounted tissue sections were stored at -70°C until binding. Slides were incubated in 10 mM H E P E S , pH 7.4, 130 mM NaCI, 4.7 mM KC1, 5 mM MgCI2, 1 mM EGTA, 0.05% bovine serum albumin (BSA) 0.07 mM bacitracin and (125ITyr4)BN in the presence or absence of 1/zM unlabelled BN. Following incubation at 24°C, free radiolabelled peptide was removed by two consecutive 4-rain washes in 10 mM H E P E S (pH 7.4) containing 0.05% BSA at 4°C. Sections were either wiped from slides and assayed for radioactivity in a gamma counter or washed slides were dried with a stream of cold air and placed in a desiccator under partial vacuum pressure for 12 hr to insure total drying. Dried slides were placed in an x-ray cassette and opposed to sheets of L K B Ultrofilm. The binding conditions employed in these experiments were similar to those utilized by Wolfet al. [22] and Zarbin et al. [24] on brain sections. Binding under these conditions has been shown to be of high affinity, saturable, reversible on brain sections. Preliminary aspects of the present study
B O M B E S I N (BN), a tetradecapeptide originally isolated from amphibian skin, has been shown to be biologically active in mammalian gastrointestinal tract [1, 3, 9, 17] and brain [5,16]. Immunoreactivity to bombesin and structurally related compounds such as gastrin releasing peptide (GRP) has been demonstrated in neurons of the gastrointestinal tract [2,15], and BN receptors have been identified in brain tissue [11-13, 22-24]. BN has been shown to affect food intake [4,8] although the exact site of action or the mechanism for this effect of bombesin has not been identified. Neural disconnection of the gut via combined dorsal rhizotomy at T3-6, cord section at T6 and bilateral subdiaphragmatic vagotomy eliminated the satiety effect of bombesin suggesting a gastrointestinal site for this behavioral effect [ 19]. In order to identify potential sites for the satiety action of bombesin, we have mapped the distribution of BN binding sites in the rat gastrointestinal tract utilizing in vitro autoradiography with (125I-Tyr 4) bombesin and have pharmacologically characterized these binding sites. A preliminary report of these experiments has appeared [14]. METHOD BN was iodinated using the chloramine T procedure and purified using gel filtration techniques [11] to a specific activity of 1200 Ci/mmol. Male Sprague-Dawley rats 300-350 g
1Requests for reprints should be addressed to Timothy H. Moran, PhD., Department of Psychiatry and Behavioral Sciences, Meyer 4-119, Johns Hopkins University School of Medicine, 600 N. Wolfe St., Baltimore, MD 21205.
643
644
M O R A N E T AL.
1500-
6A Totol
Toto] o NonspeciFic • Specific
,, N o n s p e c i f i c
Q (~
i
"G
100O"
4-
v
Z-
500&
I
I
I
30
60
90
(TYR-4)BN (riM)
1
120
Time (min) FIG. 1. Association of (1z~I-TyrgBN to rat fundal sections. Total (filled triangles) and nonspecific (open triangles) binding were determined as a function of time after the addition of (~zSI-TyrgBN. 3
DISTRIBUTION OF BOMBESIN BINDING SITES 21.0"
•
ng 3
0.8'
1 (/) znl a
0.6 ¸
0
0.4
0 0
I 1
i
i
2
3
BN Bound (fmol/~ection)
0
FIG. 2. Top: Binding of (125I-TyrgBN as a function of radiolabelled peptide concentration. Total (open triangles) and nonspecific (open squares) binding were determined in response to increasing concentrations of (mI-TrygBN, Specific binding (filled circles) represents the difference between total and nonspecific. Bottom: Skatchard plot of the specific binding. The line represents the best fit assuming a single class of binding sites (Kd=l nM).
0.2
0.0
,. =
m
'o
_o (3.
o
3,
~
=
<
~
F-
a
FIG. 3. Relative densities of (I~5I-TyrOBNbinding to tissue sections from various gastrointestinal levels demonstrating total and nonspecific binding. Values indicate the average density across the cellular layers within a segment expressed as optical density from computerized microdensitometry. Significant differences between total and nonspecific binding indicated by *p <0.05. evaluated the dynamics and equillibrium of bombesin binding to gastric fundal sections. Autoradiographic studies were conducted utilizing 1 nM (12~I-TyrgBN and incubating for 60 rain. Autoradiographic images from the developed films, demonstrating total and nonspecific binding, were directly compared with the corresponding tissue sections which were stained with toluidine blue. Superimposition of the stained slides with their corresponding autoradiographs allowed an accurate anatomical localization of the specific BN binding sites. Microden-
sitometry of autoradiographs was carried out on images utilizing an Amersham RAS1000 computerized microdensitometer. Sections demonstrating total binding and the adjacent sections in which binding had been inhibited with 1 /xM unlabelled BN were compared for levels of binding as indicated by optical density from the various tissue segments. Six samples of each segment were examined. In a second series of studies, tissue samples from the gastric fundus were sectioned at a thickness of 25 ~ m and thaw mounted to gelatin-coated slides. Four sections were mounted per slide. Slides were incubated as in the previous experiment in the presence or absence of various concentrations (10 -5 to 10-11 M) of BN analogs to determine the ability of these compounds to inhibit BN binding. These compounds included gastrin releasing peptide (GRP), acetylated GRP 2°-2r (Ac-GRp2°-zT), ranatensin, litorin, bombesin hydroxide (BN-OH), spantide and bombesin fragment 8--14 (BN 1-14). Following washes, sections were wiped from slides
G A S T R O I N T E S T I N A L B O M B E S I N B I N D I N G SITES
645
C
FIG. 4. Photomicrographs of toluidine blue stained gastric sections and corresponding autoradiographs showing total and nonspecific BN binding. (A) Cross section of gastric fundus. (B) Autoradiograph of section in (A) showing total ('2~I-Tyr4)BN binding. (C) Autoradiograph of section consecutive to that in A, showing nonspecific BN binding with 1/zM unlabelled BN added to the incubation buffer. (D) Cross section from gastric antrum• (E and F) Autoradiographs showing total and nonspecific BN binding, respectively corresponding to (D). Specific binding is localized to the circular muscle layer in both segments• CM, circular muscle; LM, longitudinal muscle; M, mucosa.
and assayed for radioactivity using a gamma counter. The ability of each concentration of analog to inhibit BN binding was examined in three separate experiments. RESULTS The kinetics of association of ('25I-Tyr4)BN to rat fundal slices is demonstrated in Fig. 1. Specific binding increased rapidly the first 60 min, slowly over the next 30 min and after 90 min, equilibrium was attained. In contrast, nonspecific binding in the presence of 1 /~M unlabelled BN increased only slowly through this period. The concentration dependence of ('25I-Tyr4)BN to rat fundal sections is demonstrated in Fig. 2A. Total binding and nonspecific binding increased with increasing concentrations of ('25I-Tyr4)BN, while specific binding was saturable. Scatchard analysis (Fig. 2B) indicated that BN binding was occurring to a single class of sites ( K d = l nM).
Distribution of Bombesin Binding Sites The relative density of (Tyr 4) BN binding across the various gastrointestinal segment is shown in Fig. 3. Overall, the
highest density of binding was evident in the gastric fundus. Significant levels of specific binding were also found in the antrum. Within the small intestine, significant binding was evident in the duodenum, jejunum, and ileum. All segments of the colon (ascending, transverse and descending) demonstrated specific binding. No specific binding was evident in the pyloric canal or pyloric sphincter. The specific localization of BN binding sites varied across the tissue segments in which they were found. In the fundus (Fig. 4A-C), high binding densities were localized to the circular muscle layer. No obvious increases in grain density were evident in the serosal, submucosal, or mucosal layers (Fig. 4B). Binding to the gastric antrum (Fig. 4D-F) was also localized primarily to the circular muscle. Nonspecific binding was evident in the submucosal and mucosal layers. In the duodenum (Fig. 5A-C), high levels of nonspecific binding were found in the mucosa (as in other small intestinal segments). Specific binding was localized to the circular muscle and the submucosal layer. Localization of BN binding sites to the submucosa was also evident in the jejunum (Fig. 5D-F) and ileum (Fig. 5G-I). Within the colon, similar distributions were found in the
M O R A N ET AL.
646
H
L
"
=Atr~.
FIG. 5. Photomicrographs of toluidine blue stained small intestinal sections and corresponding autoradiographs showing total and nonspecific BN binding. (A) Cross section from duodenum. (B and C) Autoradiographs showing total and nonspecific BN binding. Specific binding is localized to the submucosal and circular muscle layers. (D) Cross section of jejunum. (E and F) Autoradiographs showing total and nonspecific BN binding. Specific binding is localized to the submucosal layer. (G) Cross section of ileum. (H and I) Autoradiographs showing total and nonspecific BN binding. Specific binding is localized to the submucosal layer. CM, circular muscle; LM, longitudinal muscle; IG, intestinal glands; SM, submucosa. ascending (Fig. 6A-C) and transverse segments (Fig. 6D-F), with highest densities localized to the longitudinal muscle and the submucosal layer. Lower levels of specific binding were found in the circular muscle layer. In the descending colon (Fig. 6G-I), highest densities were evident in the circular muscle layer.
Pharmacological Specificity Figure 7 demonstrates that BN and structurally related peptides inhibited (125I-Tyr4)BN binding to the gastric fundus in a concentration dependent manner. The relative potencies of these compounds, as measured by the concentration required to inhibit 50% of the (n5I-Tyr4)BN binding (IC~0), are shown in
Table 1. Compounds such as GRP, Ac-GRP, litorin and ranatensin, with the C-terminal of bombesin essentially intact were equally or more potent than bombesin at the fundal sites• Compounds in which the C-terminal was altered such as BN 8-14, BN-OH and spantide, were not potent inhibitors of (125I-Tyr4) BN binding. DISCUSSION These results demonstrate a wide distribution of bombesin binding sites in the rat gastrointestinal tract. The particular cellular layer in which the binding sites are found depends upon the particular gastrointestinal segment• In the stomach, a gradient of receptors was evident from a high density in the
G A S T R O I N T E S T I N A L B O M B E S I N B I N D I N G SITES
647
°~ ' % e .
,t~ 7
'~2¢
2:
~ .
.
, ,e
-
v , ..
k ",.
v,
..~
. ,;-d 1."
~e,
I
I
f
•
~ f
4
te~
21 "~j
..
FIG. 6. Photomicrographs of toluidine blue stained large intestinal sections and corresponding autoradiographs showing total and nonspecific BN binding. (A) Cross section for ascending colon. (B and C) Autoradiographs showing total and nonspecific binding. Specific binding is localized to the circular and longitudinal muscle and the submucosal layers. (D) Cross section of transverse colon. (E and F) Autoradiographs showing total and nonspecific binding. Specific binding is localized to the longitudinal and circular muscle and submucosal layers. (G) Cross section of descending colon. (H and I) Autoradiographs of total and nonspecific binding. Specific binding is localized to the circular muscle layer.
circular muscle layer of the fundus, lower density in the antrum and a lack of discernible specific binding in the region of the pylorus. In the small intestine, receptors were predominantly localized to the submucosal layer, with some specific binding of the circular muscle layer in the duodenum and ileum. In the large intestine, high levels of specific binding were evident in the longitudinal muscle and submucosal layers with less dense binding in the circular muscle, although the descending colon had very dense binding in the circular muscle region. Bombesin receptors have previously been characterized in rat brain [11-13, 22-24], guinea pig pancreas [7] and in pituitary cell culture [21]. Although a similar af~nity for
bombesin is evident in these systems, differences are found in the relative affinity for BN-like peptides. In the present study, GRP, Ac-GRP 2°-27, litorin and ranatensin all were more potent than bombesin in inhibiting the binding of 1251T y # ) B N . In rat brain and guinea pig pancreas, litorin and ranatensin have been reported to have only 5-25% of the potency of B N [7,11], while in rat pituitary, these compounds were slightly more potent than BN [21]. Also, in the rat brain and pancreas, spantide is far more potent (IC50=1 /zM) than at fundal binding sites (IC50>10/~M). Thus, these fundal binding sites appear to be more similar to pituitary than brain or pancreatic B N receptors. However, fundal binding sites also demonstrate a greater affinity for GRP
648
MORAN
INHIBITIONOF 1251-TYR4 BN BINDING
TABLE 1 PHARMACOLOGY OF BOMBESIN AND RELATED PEPTIDES
100 i
80
-~-'-~ ~
.":' " L-~, _ ~'t'-.. "~ "~ ~ ~
ZXBN8-14 " BN-OH
20
" Utorln
O AcGRP • GRP
.c_
-11 -10
\ \ ~"~ \\ \
\
\
-9 -8 -7 -6 CONCENTRATION (log M)
Peptide Litorin Ac-GRP~°-~7 GRP Ranatensin Bombesin BN 8-14 BN-OH Spantide
60 40
ET AL.
IC5o (nM) 1.0 _+ 0.6 1.1 + 0.5 2.1 -+ 0.9 7.8 --+ 1.1 20.0 _+ 5.0 300.0 + 102 >10,000 > 10,000
-5
FIG. 7. Pharmacology of fundal (Tyr4)BN binding sites. The percentage of (lzSI-Tyr4)BN specifically bound is plotted as a function of unlabelled peptide concentration.
(10×) and Ac-GRP 2°-~7 (20x) than for BN, a result which was not evident for pituitary BN receptors where GRP was slightly less potent than BN [21]. This pattern of results suggests that these fundal BN binding sites may represent an additional subtype of BN receptor, different from those which have previously been characterized in brain, pancreas and pituitary, although the time course and Ko of (Tyr4)BN binding to fundal sections are similar to those which have been demonstrated for binding to brain sections [23,24]. Whether the BN binding sites in other gastrointestinal segments demonstrate similar affinities for these BN analogs has not been determined. The particular cellular element containing the BN binding sites cannot be discerned from this type of study. Binding sites may be directly on muscle fibers, or on neural elements presynaptic to these fibers. Furthermore, it may be that binding sites are on the muscle in some regions and on neural elements in others, such as in the submucosal layer. In vivo and in vitro studies have suggested all possibilities. A direct smooth muscle excitation of gastric circular muscle by BN was demonstrated by Mayer et al. [9] utilizing an in vitro preparation. In contrast, in vivo experiments have suggested that BN and GRP receptors are on nerves innervating the circular muscle of the stomach [3]. In the small intestine, receptors appear to be on inhibitory nerves, not on smooth muscle, as responses to BN and GRP are blocked by tetrodotoxin [3]. Bombesin immunoreactivity, localized to nerve fibers, has been demonstrated in various gastrointestinal segments in the rat and guinea pig [2,20]. In the rat stomach, these nerve fibers are found in high concentration in the gastric fundus and less so in the gastric antrum and appear to be localized to the mucosal layer. This pattern is also evident in levels of extractable BN [20]. The distribution of BN and BN binding sites, although similar in terms of relative density within the fundus and antrum, differ in that binding sites are localized to the muscle layer and BN itself is concentrated in the mucosal layer. In the small intestine, BN immunoreactivity is found in the myenteric plexus and within the muscle layers
Values indicate mean _+ SE from 3 determinations.
of the duodenum, jejunum, ileum and colon [20], a similar distribution to that of BN binding sites. The distribution of BN binding sites in the rat gastrointestinal tract differs from that found for cholecystokinin (CCK), another peptide which can potently inhibit food intake. In the rat, gastrointestinal C C K receptors are limited to the circular muscle layer of the pyloric sphincter, and a mechanism for the satiety action of CCK through the interaction of CCK with this receptor population has been suggested [18]. No such specific localization or potential function is evident for BN. The distribution of BN binding sites presents a wide range of possibilities of sites and mechanisms for the mediation of the satiety action of bombesin. Preliminary data has suggested a variety of sites and potential mechanisms for BN satiety. Complete gastrointestinal deafferentation involving bilateral subdiaphragmatic vagotomy, rhizotomy and cord section at the level of T6 eliminates the satiety action of BN [19]. Although this finding supports a peripheral site of action for BN, it does not identify a specific pathway or mechanism of action. Furthermore, partial gastrointestinal deafferentation by either extensive rhizotomy (T3-T10) alone or rhizotomy and cord transection without vagotomy result in a partial loss of BN satiety [10] suggesting the possibility for multiple sites of action or additive effects. The satiety action of BN may be direct or may be secondary to some BN mediated gastrointestinal action. BN has been reported to inhibit gastric emptying in the rat following both central [5,16] and peripheral administration [1,17], although the inhibition following peripheral administration is not a universal finding and may be strain specific [6,16]. Some component of the satiety action of bombesin may be secondary to this gastric inhibitory effect of BN, but, as yet, no mechanism for the satiety action of BN has been demonstrated. Thus, any or all the gastrointestinal binding sites demonstrated in the present work may be among the candidate target sites for the mediation of BN's satiety action. In summary, the present results identify a wide distribution of bombesin binding sites in the rat gastrointestinal tract, through different cellular layers in different tissues. This pattern of results suggests a general role for BN in coordinating gastrointestinal function. In contrast to the gastrointestinal localization of CCK receptors, the extensive distribution of BN binding sites does not clearly identify a locus for the peripheral mediation of the satiety action of BN.
GASTROINTESTINAL
BOMBESIN
BINDING SITES
649
REFERENCES 1. Bertaccini, G., M. Impicciatore. Action of bombesin on the 14. Moran, T. H., T. W. Moody, M. S. Goldrich, P. H. Robinson and P. R. McHugh. Autoradiographic localization of bombesin motility of the stomach. Naunyn Schmiedebergs Arch Pharbinding sites within the gastrointestinal tract. Soc Neurosci macol 289: 149-156, 1975. 2. Costa, M., J. B. Furness, N. Yanalhara, C. Yanaihara and T. Abstr 11: 903, 1985. 15. Polak, J. M., S. R. Bloom, S. Hobbs, E. Solcia and A. G. E. W. Moody. Distribution of projections of neurons with immunoPearse. Distribution of a bombesin-like peptide in human gasreactivity for both gastrin-releasing peptide and bombesin in the guinea-pig small intestine. Cell Tissue Res 235: 285-293, 1984. trointestinal tract. Lancet 1:1109-1110, 1976. 16. Porreca, F. and T. F. Burks. Centrally administered bombesin 3. Fox, J. E. T. and T. J. McDonald. Motor effects of gastrin releasing peptide (GRP) and bombesin in the canine stomach affects gastric emptying and small and large bowel transit in the rat. Gastroenterology 85: 313-317, 1983. and small intestine. Life Sci 35: 1667-1673, 1984. 17. Scarpignato, C. and G. Bertacinni. Bombesin delays gastric 4. Gibbs, J., D. J. Fauser, E. A. Rowe, B. J. Rolls, E. T. Rolls and S. P. Maddison. Bombesin suppresses feeding in rats. Nature emptying in the rat. Digestion 21: 104-106, 1981. 282: 208-210, 1979. 18. Smith, G. T., T. H. Moran, J. T. Coyle, M. J. Kuhar, T. L. 5. Gmerek, D. E. and A. Cowan. Pituitary-adrenal mediation of O'Donohue and P. R. McHugh. Anatomical localization of bombesin-induced inhibition of gastrointestinal transit in rats. cholecystokinin receptors in the pyloric sphincter. Am J Physiol Regul Pept 9: 299-304, 1984. 246: R127-R130, 1984. 19. Stuckey, J., J. Gibbs and G. P. Smith. Neural disconnection of 6. Hostetler, A. M., T. H. Moran and P. R. McHugh. Bombesin: Gastric emptying and feeding. Proc East Psychol Assoc 57: 57, gut from brain blocks bombesin-induced satiety. Peptides 6: 1986 (abstract). 1249-1252, 1985. 20. Walsh, J. H., J. R. Reeve and S. R. Vigna. Distribution and 7. Jensen, R. T., T. W. Moody, C. Pert, J. E. Rivier and J. D. molecular forms of mammalian bombesin. In: Gut Hormones, Gardner. Interaction of bombesin and litorin with specific memedited by S. R. Bloom and J. M. Polak. Edinburgh: Churchill brane receptors on pancreatic acinar cells. Proc Natl Acad Sci Livingstone, 1981, pp. 414-418. USA 75: 6139-6143, 1978. 21. Westendorf, J. M. and A. Schonbrunn. Characterization of 8. Kulkosky, P. J., J. Gibbs and G. P. Smith. Behavioral effects of bombesin receptors in a rat pituitary cell line. J Biol Chem 258: bombesin administration in rats. Physiol Behav 28: 505-512, 1982. 7527-7535, 1983. 22. Wolf, S. S., T. W. Moody, T. L. O'Donohue, M. A. Zarbin and 9. Mayer, E. A., J. Elashoff and J. H. Walsh. Characterization of M. J. Kuhar. Autoradiographic visualization of rat brain sites bombesin effects on canine gastric muscle. A m J Physiol 243: for bombesin-like peptides. Eur J Pharmacol 87: 163-164, 1983. G141-G147, 1982. 23. Wolf, S. S. and T. W. Moody. Receptors for GRP/Bombesin10. Mindell, S., J. DiPoala, S. Wiener, J. Gibbs and G. P. Smith. like peptides in the rat forebrain. Peptides 6: 111-114, 1985. Satiety effect on bombesin is attenuated by dorsal rhizotomy 24. Zarbin, M. A., M. J. Kuhar, T. L. O'Donohue, S. S. Wolf and and spinal cord transection. Proc East Psychol Assoc 57: ll, R. W. Moody. Autoradiographic localization of (~25I-Tyr4) 1986. bombesin binding sites in the rat brain. J Neurosci 5: 429-437, 11. Moody, T. W., C. B. Pert, J. Rivier and M. R. Brown. Bombe1985. sin: Specific binding to rat brain membranes. Proc Natl Acad Sci USA 75: 5372-5376, 1978. 12. Moody, T. W., T. L. O'Donohue and D. M. Jacobowitz. Biochemical localization and characterization of bombesin-like peptides in discrete regions of rat brain. Peptides 2: 75-79, 1981. 13. Moody, T. W., J. N. Crawley and R. T. Jensen. Pharmacology and neurochemistry of bombesin-like peptides. Peptides 3: 559-563, 1982.