| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
INSERM U-45 (F.A.E.F., P.N., M.D.-V.), Ecole Pratique des Hautes Etudes (D.P.), and Laboratoire d’Anatomie Pathologique (F.B.), Hôpital Edouard Herriot, 69437 Lyon, and Laboratoire d’Histologie-Embryologie, Faculté de Médecine Lyon-Sud (P.L.), 69921 Oullins, France
Address all correspondence and requests for reprints to: Dr. D. Pansu, Ecole Pratique des Hautes Etudes, INSERM U-45, Hôpital Edouard Herriot, 69437 Lyon Cedex 03, France.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Therefore, the goals of the present study were 3-fold. 1) As in peptide immunocytochemistry it is essential to stain sections with antibodies to several different regions of the molecule, antibodies to N-terminal, midportion, and C-terminal amino acid sequences of sorbin were evaluated by immunocytochemistry for the quality of the specific staining. In addition to the radioimmunological characterization of these antibodies, the specificity of the immunocytochemical labeling was ensured not only by the usual specificity tests, but also by extensive cross-reaction studies. 2) The distribution of sorbin was examined in detail in the porcine gastrointestinal tract and pancreas to obtain a greater understanding of its possible tissue-specific role. 3) Peptide expression was extensively studied by double staining cytochemistry to define which cells in the porcine digestive tract contribute to the production of sorbin.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Tissue processing was performed with previously validated procedures (6, 7). Briefly, the tissues were rinsed in ice-cold saline solution and immediately fixed by immersion in 4% paraformaldehyde plus 0.5% glutaraldehyde solution dissolved in 0.1 M Sörensen’s phosphate buffer, pH 7.4, for 2 h at 4 C. After postfixation in 4% paraformaldehyde in the same buffer for 2 days at 4 C, the tissues were washed in 0.1 M PBS, pH 7.6; dehydrated in a graded series of ethanol; and embedded in Paraplast. Sections of 4-µm thickness were mounted on glass slides precoated with 0.1% poly-L-lysine (Sigma Chemical Co., Saint Quentin Fallavier, France), deparaffinized in methylcyclohexane, and returned to water through graded ethanol solutions. The sections were pretreated for 5 min with 3% hydrogen peroxide solution before processing for immunocytochemistry.
Primary antisera
Several rabbit primary antisera were prepared using either the
whole molecule of sorbin or a smaller amino acid sequence. In
particular, primary antisera to N-terminal (synthetic fragment 8–18),
midportion (synthetic fragments 28–43 and 103–116), and C-terminal
(synthetic fragment 137–153) sequences of sorbin were tested; their
locations in the whole molecule are shown in Fig. 1
.
|
Other primary antisera were used for both the specificity controls and study of the interrelationship of cells immunoreactive to sorbin and other peptides to nonpeptide substances. They included rabbit cholecystokinin (CCK) antiserum (diluted 1:500; no. C8E), rabbit gastrin antiserum (diluted 1:200; no. 28E), rabbit gastric inhibitory polypeptide (GIP) antiserum (diluted 1:500; no. 59A), rabbit peptide YY (PYY) antiserum (diluted 1:100; no. A4D), rabbit somatostatin (SRIF) antiserum (diluted 1:500; no. 55A), and rabbit substance P antiserum (diluted 1:500; no. 80F and 81B). These primary antisera were developed and characterized in our laboratory (9, 10, 11, 12). The following primary antisera were also used: mouse chromogranin A antiserum (diluted 1:100; Dako, Glostrup, Denmark), rabbit gastrin antiserum (diluted 1:2000; Peninsula Laboratories Europe, Saint Helens, UK), rabbit Met-enkephalin antiserum (diluted 1:500; Peninsula Laboratories, Belmont, CA), mouse glucagon antiserum (diluted 1:2000; Sigma), guinea pig insulin antiserum (diluted 1:300; Peninsula), mouse serotonin (diluted 1:10; Dako), and rat serotonin (diluted 1:1000; Tebu, Le Perray-en-Yvelines, France) antisera; rat SRIF antiserum (diluted 1:1000, Tebu); and rabbit substance P antiserum (diluted 1:800; Peninsula).
Single antigen immunocytochemistry
The microwave irradiation consistently improved the
detectability of sorbin. We used a commercial microwave oven (Thompson)
with a maximum power of 650 watts. The slides were treated three times
at full power for 5 min each time in a beaker filled with 0.01
M citrate buffer (citric acid monohydrate, adjusted to pH
6.0 with NaOH). Then the slides were left in the beaker for 20 min at
room temperature and rinsed in 0.05 M Tris-HCl buffer
containing 0.15 M NaCl, pH 7.6 (TBS) (13).
The immunocytochemical procedure was previously described by our laboratory (14, 15). Briefly, the sections were processed at room temperature in a humid chamber by the indirect peroxidase-antiperoxidase technique (16). Microscope slides with fixed sections were treated for 20 min at room temperature with a nonimmune serum (goat; Sigma), diluted 1:40 in TBS, and then incubated overnight at 4 C with either of two primary region-specific antisera. The standard antibody dilutions used during immunocytochemistry were 1:2000 for C-terminal sorbin antiserum and 1:1000 for N-terminal sorbin antiserum. After washing, the sections were incubated with an unlabeled goat antirabbit IgG diluted 1:100 for 30 min at room temperature and then with the peroxidase-antiperoxidase complex (Jackson ImmunoResearch, Baltimore, MD) diluted 1:400 for 90 min at room temperature. The antibody dilutions were made with 1% BSA in TBS. The peroxidase was reacted with a mixture of 3,3'-diaminobenzidine tetrahydrochloride (0.025% in TBS; Sigma) and 0.015% H2O2. The slides were rinsed several times with deionized water, and then counterstained with hematoxylin, dehydrated through a series of alcohols and methylcyclohexanes, and coverslipped with Eukitt.
Double antigen immunocytochemistry
To examine the coexpression of two different peptides in the
same cell, various immunocytochemical approaches were applied. They
included the adjacent section method, in which consecutive sections
were incubated with different primary antisera. No cross-reaction
between antisera could occur, and consequently, there were no problems
of specificity due to interference between antibodies. However, a
disadvantage of this method was that the comparison of serial sections
of 4-µm thickness might result in equivocal pictures. Therefore, the
second approach was indirect double staining, using two primary
antisera raised in different species, mixed, and applied simultaneously
to the tissue preparation at their predetermined optimal dilutions.
Secondary antisera labeled with different chromogens (i.e.
green fluorescent fluorescein isothiocyanate and red fluorescent
tetramethyl rhodamine isothiocyanate or Texas Red), and directed
against IgG from the respective species were also mixed and applied
together (the details of each comparison are given in the legend to the
corresponding figure) (17).
Primary antiserum specificity controls
In the present study, the immunocytochemical localization of
sorbin was mainly investigated by use of the C-terminal sorbin
antiserum. Consequently, the region specificity of this antibody was
tested in both the radioimmunological and immunocytochemical systems.
On the other hand, the N-terminal sorbin antiserum was used in parallel
to confirm the sequence previously described (1), and the specificity
of this antiserum was examined by controls in immunocytochemistry.
Radioimmunological tests.
The C-terminal sorbin antiserum
was characterized using [125I-Tyr]C17-NH2 as
tracer (74 tetrabecquerels/mmol at the shipping date; Amersham, Les
Ulis, France) (9). The titration was performed using the tracer (2000
cpm) in the presence of antiserum dilutions varying from 1:50 to
1:100,000 in a buffer (pH 7.5) with 42 mmol/liter
Na2HPO4, 8 mmol/liter
NaH2PO4, 3.8 mmol/liter EDTA, 4.7 mmol/liter Na
azide, 2% equine serum, and 1.43% zymofren. The intraassay variation
was calculated with six duplicate determinations of 0.25 pmol
Y-C17-NH2 sorbin. The interassay variation was calculated
with the ID50 obtained from nine standard curves. The
regional specificity of the C-terminal sorbin antiserum was established
by monitoring the displacement obtained with the whole peptide and the
amidated C-terminal peptides of sorbin (i.e. C3, C5, C7,
C10, C17, and C20 fragments); in addition, the nonamidated C7 fragment
was used. The cross-reactivity was checked by incubation of
heterologous peptides (0.2–1 nmol/ml assay): angiotensin II, ACTH,
bombesin, calcitonin gene-related peptide, CCK-33, dynorphin A,
endothelin 1, 
-endorphin, galanin, GIP, gastrin 17-I,
Met-enkephalin, motilin, neurokinin-A (or substance K), neuromedin U,
neuropeptide Y (NPY), peptide histidine isoleucine, secretin,
serotonin-HCl, SRIF-14, substance P, and VIP. A concentration of 80
nmol/ml was used for serotonin (creatinine-sulfate complex).
Immunocytochemical tests.
The specificity of the
immunostaining was demonstrated not only by the usual specificity tests
(method controls), but also by extensive cross-reaction studies
(antibody controls), as recommended previously (18). Method controls to
validate the specificity of the binding of immunocytochemical reagents
with tissue included: 1) omission of the primary antiserum, 2)
replacement of the primary antiserum with nonimmune serum, 3) dilution
profile of the primary antiserum using doubling dilutions on serial
sections, 4) influence of the salt content (up to 0.5 M) of
buffer, and 5) complement-deprived antisera (19). In addition, controls
for the specificity of the double labeling were performed as follows:
1) using nonimmune serum as the first layer, and 2) using the labeled
antisera without the presence of one or both primary antisera (20).
Antibody controls to determine the specificity of interaction between
the primary antiserum and the tissue-bound antigen were as follows. The
primary C- or N-terminal sorbin antiserum was preincubated with 0.1–1
µmol of the corresponding antigen/ml undiluted antiserum. In
addition, the primary C-terminal sorbin antiserum was preincubated with
either 0.001 µmol of the pure sorbin extract or 0.1–1 µmol of the
amidated C-terminal peptides of sorbin (i.e. C3, C7, C10,
C15, and C20 fragments)/ml undiluted antiserum. Because both antibodies
revealed a positive reaction in the porcine gastrointestinal tract, the
primary C- or N-terminal sorbin antiserum was preincubated with
heterologous peptides known to be present in endocrine cells and
neurons of the gut, including angiotensin II; ACTH; bombesin;
calcitonin gene-related peptide; CCK-33; chromogranin A; Tyr-CRF;
dynorphin A; endothelin 1; -endorphin; galanin; GIP; gastrins 1–14,
14–17, 17-I, and 17-II; pentagastrin; Met-enkephalin; motilin;
neurokinin A; neurokinin B (or neuromedin K); neuromedin U;
neuropeptide K; NPY; peptide histidine isoleucine; secretin; serotonin
(creatinine-sulfate complex); serotonin-HCl; SRIF-28; substance P; TRH;
and VIP. Before staining, 1–10 µmol pure antigen were added per ml
undiluted antiserum. Because the primary C-terminal sorbin antiserum
(like the primary N-terminal sorbin antiserum) revealed a positive
reaction in the endocrine pancreas, this antibody was preincubated with
1–10 µmol heterologous peptides naturally occurring in the islets of
Langerhans per ml undiluted antiserum, including insulin, glucagon,
SRIF-14, pancreatic polypeptide (PP), serotonin (creatinine sulfate
complex), and TRH. The C-terminal sorbin antiserum was replaced by
other antibodies, including CCK, gastrin, GIP, Met-enkephalin,
glucagon, insulin, PYY, serotonin, SRIF, and substance P; in the case
of a dual label immunocytochemical localization, an excess amount of
Y-C17-NH2 sorbin (10 µmol/ml undiluted antiserum) was used in
preadsorption experiments. The substance P antisera were preincubated
with 10 µmol serotonin (creatinine-sulfate complex)/ml undiluted
antiserum because an unexpected cross-reactivity of substance P
antiserum with serotonin had been reported in gut endocrine cells in
the chick (21). These specificity tests were largely as previously
described (22).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). The ID50 was
approximately 2.5 pmol/ml for Y-C17-NH2 sorbin. The assay
sensitivity averaged 0.5 pmol/ml; intra- and interassay coefficients of
variation were 7% and 20%, respectively. The ID50 values
for the other amidated peptides were 2.7, 4.4, 41, and 590 pmol/ml for
the C-20, C-10, C-7, and C-5 forms, respectively. The ID50
was obtained with a pure sorbin concentration of approximately 220
pmol/ml. The primary antiserum recognized neither the C-3 fragment nor
the nonamidated C-7 fragment of sorbin. Finally, none of the
heterologous gastrointestinal peptides used here demonstrated
cross-reactivity with the C-terminal sorbin antiserum.
|
shows
that the labeling was observed in endocrine-type cells and neurons in
the porcine gastrointestinal tract and pancreas. As a first step in the
distribution analyses of the sorbin, we verified that the labeling
obtained was specific in the digestive tract; this was exemplified with
the use of gastric, duodenal, and pancreatic sections. The specificity
of the binding mechanisms involved at the cellular level was assessed
by the method controls. No staining was observed after omission of the
C-terminal sorbin antiserum or its substitution by nonimmune serum.
Sorbin immunoreactivity was independent of the dilution of the primary
antiserum within a wide range (up to 1:2000); in addition, staining was
not modified by increasing the salt content of buffer up to 0.5
M or by heating the C-terminal sorbin antiserum at 56 C for
30 min just before use. The specificity of the double labeling was
corroborated by appropriate controls. Thus, we used a nonimmune serum
as a first layer, which resulted in the absence of any staining; on the
other hand, omission of one or both primary antisera resulted in a lack
of specific staining, thus indicating that the detecting antibodies
were free of species cross-reactivity. The antibody controls carried
out established the occurrence of an immunoreactant(s) to sorbin in
both endocrine-type cells and the enteric nervous system. Indeed,
preliminary data indicated that a significant reduction in staining was
observed with the use of 1 nmol pure sorbin/ml undiluted antibody. Not
surprisingly, all staining was abolished after adsorption of the
C-terminal sorbin antiserum with the synthetic homologous peptide,
Y-C17-NH2 sorbin (1 µmol/ml undiluted antibody or more;
Fig. 3) and with other amidated fragments, specifically C20, C15, C10,
and C7 forms (1 µmol/ml undiluted antibody or more). In contrast, the
synthetic C3-NH2 fragment of sorbin did not affect the
signal. On the other hand, addition of heterologous antigens to
C-terminal sorbin antiserum had no effect on the immunostaining in
either endocrine-type cells or the enteric nervous system.
|
Test tissues were stained with an immune serum toward an antigen different from that investigated. Serial sections alternately labeled with sorbin and serotonin, Met-enkephalin, substance P, or insulin antiserum demonstrated that sorbin was present in multiple endocrine-type cells. However, the staining obtained with sorbin antibodies was unaffected by adsorption with serotonin, Met-enkephalin, substance P, and insulin, and the competition with sorbin neither suppressed nor diminished the immunoreactivity of serotonin-, Met-enkephalin-, substance P-, and insulin-containing cells. It should be noted that serotonin failed to inactivate staining obtained with substance P antisera, in contrast to a previous report (21). Not surprisingly, nerves immunoreactive to CCK, Met-enkephalin, serotonin, SRIF, and substance P were not observed with our fixation method. Therefore, the topographic distribution of these peptide-containing neuronal systems could not be used for the specificity controls.
Distribution of cells immunoreactive to sorbin antiserum
Localization of sorbin to the stomach.
By incubating tissues
with the C-terminal sorbin antiserum, positive cells were restricted to
the gastric mucosa. In the fundic mucosa, a sparse population of
immunoreactive cells was identified. They were scattered in glands,
especially in the midportion, and appeared to lack luminal contact
(Fig. 4A). Sorbin-immunoreactive cells
were moderate in number in the pyloric antrum (Fig. 3A
), but were
numerous in the pyloric canal on both sides of the pyloric sphincter
(Fig. 4B). They predominated in the middle part of the glands; in
contrast, the deepest part of the glands and their upper part or the
gastric pits near the luminal surface were immunonegative for sorbin.
The immunoreactive cells did not reach the lumen, and they could be
considered as belonging to the closed type; however, a few cells
appeared to reach the lumen.
|
). Organized in small clusters,
positive cells become gradually fewer distally in the jejunum (Fig. 4D), they were rarely present in the ileum and were absent in the large
intestine. Immunoreactive cells were mainly located in the crypts of
Lieberkühn, with only rare cells occurring at the base of the
villi. As in the stomach, most immunoreactive cells were of the closed
type. A small number of positive cells were also found in Brunner’s
glands (Fig. 4C).
Localization of sorbin to the pancreas.
Three apparently
different cell types were stained within the pancreas: cells within the
islets of Langerhans, cells scattered throughout the exocrine pancreas,
and cells associated with the neuroendocrine complex. The first
category comprised sorbin-containing cells associated predominantly
with a subpopulation of the cells in the endocrine pancreas. More
precisely, these immunoreactive cells tended to accumulate in the
central portion of the islet and were surrounded by a rim of unstained
cells (Figs. 3G and 4E). Like the insulin-containing cells, such cells
were observed throughout the pancreatic gland, and their distribution
was homogeneous in the lower posterior part of the head, the body, and
the tail. The second category comprised sorbin-containing cells widely
distributed in the pancreatic parenchyma. Grouped or single positive
cells could be seen intercalated between acinar cells, as well as in
the simple columnar epithelium of the intralobular and interlobular
pancreatic ducts. At the junction with the ampulla of Vater,
specifically, positive cells were visible in the lining epithelium of
the duct of Wirsung, intermingled with ordinary columnar cells but also
with numerous goblet cells (Fig. 4F). It should be noted that small
mucous glands associated with the duct of Wirsung also contained
occasional positive cells. In all cases, cell apexes appeared either to
reach the lumen or to lack luminal contact; these cells were of the
open and closed types. The third category comprised sorbin-containing
cells present in the neuroendocrine (or neuroinsular) complex. This
structure corresponds to an intimate association between islet cells
and nerve cells and/or nerve fibers (for the occurrence and nature of
such associations, see Refs. 23–26). In the present study, occasional
positive cells were seen in close juxtaposition with large
sorbin-negative ganglion cells (not shown), suggesting that certain
endocrine cells in these particular structures contain sorbin.
Distribution of enteric nerve fibers immunoreactive to C-terminal
sorbin antiserum
The occurrence of sorbin-containing nerve structures was
unexpected and observed only with the C-terminal sorbin antiserum in
the intestine. Sorbin-immunoreactive nerve fibers predominated in the
myenteric (Figs. 3E and 5C) and submucous (Fig. 5A) plexus, but were rare in the circular
smooth muscle layer (Fig. 5B). This distribution pattern was similar in
the small and the large intestine. No staining was found in the stomach
or pancreas, particularly within and around pancreatic islets.
|
Sorbin in the porcine gastrointestinal tract.
Numerous
enterochromaffin (EC) cells stained with serotonin antiserum were
observed along the whole gastrointestinal tract. The majority of EC
cells were negative for sorbin (Fig. 6, C
and D), with only a relatively small fraction reacting with antisorbin
serum. In the stomach, however, all of the sorbin-containing cells
proved to contain serotonin, indicating that sorbin is present in a
subpopulation of EC cells in pyloric glands (not shown). Interestingly,
Met-enkephalin-containing cells, which constitute a subpopulation of
the EC cells, contained immunoreactive sorbin (Fig. 7, A and B); however, numerous cells
reacted only with the Met-enkephalin antiserum and not with sorbin
antiserum. In the proximal small intestine, a similar distribution of
sorbin in occasional EC cells was revealed (Fig. 6, A and B). As in the
stomach, a subpopulation of Met-enkephalin-containing cells was found
to store sorbin in the duodenum; it was obvious that the cells reactive
to Met-enkephalin antiserum largely outnumbered those labeled by the
sorbin antiserum, as Met-enkephalin-containing cells were numerous in
the distal small intestine where sorbin-containing cells were rare.
More interestingly, substance P-containing cells, which constitute
another subpopulation of EC cells in the duodenum, contained
immunoreactive sorbin (Fig. 7, C and D). However, a relatively small
fraction of substance P-containing cells reacted with the sorbin
antiserum. CCK, gastrin, or GIP immunoreactivity was never found
concomitantly within the cells immunoreactive to sorbin antiserum.
|
|
, E and F). Glucagon, PYY, or SRIF
immunoreactivity was never associated with the sorbin immunoreactivity.
In pancreatic ducts, colocalization of sorbin and serotonin
immunoreactivities in the same duct cells was also found (not shown).
In the neuroendocrine complex, unfortunately, only very rarely were
cells labeled by sorbin antiserum; therefore, the possible
colocalization of sorbin and serotonin could not be examined.
C- and N-terminal sorbin sequences.
C- and N-terminal
sorbin-immunoreactive cell profiles belonged to the same cells with the
same topographical distribution, in particular in the antral glands,
crypts of Lieberkühn (Fig. 7, E and F), and pancreatic
islets.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Distribution of sorbin
Using these antisorbin antibodies, our aim was to examine in more
detail the distribution pattern of sorbin as a first step in the
identification of novel functions for this peptide. This approach
provides considerably complementary information obtained from ligand
binding studies (5) in which the contribution of different cell types
cannot be evaluated. We demonstrate that 1) sorbin predominates in
endocrine-type cells preferentially localized in the porcine stomach,
the upper small intestine, and the pancreas; and 2) sorbin occurs also
in nerves of the intestine tract.
The diversified distribution of sorbin in endocrine-type cells suggests that this peptide may participate in regulating specific functions. Originally isolated from porcine duodenum (1), the immunocytochemical localization of sorbin in the porcine intestine was thus expected and recently confirmed in the human and rat (31). This suggests that sorbin may have a role during the first phases of digestion and absorption, a physiological function agreeing with both the accumulation sites and the regulation of ion transport by sorbin previously reported in the rat; sorbin stimulates intestinal absorption or inhibits secretion and acts without vasomotility or motility, in contrast with other proabsorptive and antisecretory agents, such as angiotensin II, NPY, SRIF, and opioid peptides (2, 3, 4). The expression of sorbin in the endocrine pancreas is also consistent with the recent discovery of sorbin in human endocrine pancreatic tumors (32). This location points to a possible involvement in the physiology of the endocrine pancreas. On the other hand, it is interesting to note that sorbin is found in sites that provide the digestive juices and the mucus necessary for the function of the small intestine, i.e. the pancreas; islet cells, but also in pancreatic ducts; the glands of Brunner (mucus-secreting); distributed throughout the acini, as reported for gastrin, SRIF, and PYY (12, 33), but also in another mucous glands associated with the duct of Wirsung in the ampulla of Vater; and the crypts of Lieberkühn. This supports the involvement of sorbin in controlling distinctive aspects of the digestive process.
Although the immunocytochemical investigations of the sorbin-endocrine system may be satisfactorily interpreted, further studies, in contrast, are required to clarify the possible presence of the histochemically distinguished neuronal cell types containing sorbin. Undeniably, sorbin is present in neuronal structures, suggesting that this peptide may be a potential transmitter. However, these findings are only preliminary, mainly because the fixation and section preparation used here are not the best tools for the light microscope localization of neuropeptides. Localization of the C-terminal sorbin fragment only in enteric neurons is subject to more than one interpretation; either the technical procedures used here are not adequate for the detection of sorbin-immunoreactive material present in the neurons and fibers, or this sorbin-like material may constitute a novel neuropeptide occurring in a distinct form, in contrast to the sorbin in endocrine cells containing the complete sequence previously described (1).
Coexistence of sorbin and other nonpeptide and peptide
substances
This study was carried out to define which cells in the porcine
digestive tract contribute to the production of sorbin. We show that 1)
in the gastrointestinal tract, sorbin is present in EC cells reacting
with antiserotonin serum and, more precisely, with the gastric-type EC
cells exhibiting Met-enkephalin and with the duodenal type EC cells
exhibiting substance P; and 2) in the islets of Langerhans, sorbin is
present in insulin-containing ß-cells.
Over the last decade, the number of candidate substances in the digestive tract has increased exponentially (reviewed in Refs. 34 and 35). A careful identification of regulatory peptides is, therefore, required as a basis for classification of the endocrine-type cells. In the stomach, the pyloric glands contain several endocrine-type cells, which to some extent have been distinguished from each other by various criteria. Notwithstanding species differences, the cell types and their constituent distinguishing content include G (gastrin), D (SRIF), and EC2 (serotonin/Met-enkephalin) cells; the remaining cells are F-like (PP) and X (unknown) cells. In the oxyntic glands, EC-like (histamine and gastrocalcin) cells represents a major fraction of the endocrine cells; D and EC2 cells are also represented (reviewed in Refs. 36 and 37). In the present study, the distribution of sorbin-containing cells in the fundic and pyloric mucosa corresponds by several criteria to that of EC2 cells.
In the intestine, endocrine cells display three different patterns of distribution (reviewed in Ref.36). The cells that contain secretin, motilin, CCK, and GIP are numerous in the duodenum; in certain species (e.g. the pig), EC2 (serotonin/Met-enkephalin) cells also display this regional distribution. The second category comprises SRIF cells and EC (intestinal-type) cells, which occur throughout the small and large intestines, and a relatively small fraction of intestinal EC cells (duodenal EC1 cell type) exhibits substance P. The third category comprises neurotensin and glicentin/PYY cells, which are numerous in the distal small intestine; the glicentin/PYY cells (and EC cells) account for most of the endocrine cells in the large intestine. In the present study, a superimposable cell profile is noted between sorbin and EC2 cells and between sorbin and EC1 cells. This latter finding is in agreement with previous observations demonstrating the presence of EC cells storing substance P in the porcine duodenum (38, 39), although this peptide predominates in the midgut.
In the pancreas, the islet of Langerhans consists of at least four
hormone-producing cell types: , ß, 
, and F cells, which
synthesize glucagon, insulin, SRIF, and PP, respectively (reviewed in
Ref.34). In the present study, insulin cells have been found to store
sorbin. In addition, the majority of the sorbin cells are positive for
serotonin in the endocrine and exocrine pancreas. It should be noted
that guinea pig ß-cells are also considered a serotonin-storing cell
type (40), and our data are in agreement with earlier observations of
EC cells scattered among endocrine and exocrine pancreatic tissues
(41).
An intriguing observation is the coincident presence of sorbin with an endogenous content of serotonin in gastric, intestinal, and pancreatic EC cells and in pancreatic ß-cells. This observation can be interpreted in two ways. The first interpretation is that sorbin plays a putative role in the release of serotonin, as documented for opioid peptides (42). However, sorbin has been found in pancreatic ß-cells immunonegative for serotonin in the human (31), and this result seems to refute this hypothesis. The second is that this coexistence does not represent a physiological property, but is related to various unifying concepts considered for the integration of gastro-entero-pancreatic endocrine cells into a common system (reviewed in Ref.43). Recent studies using transgenic animals overexpressing PYY have challenged hypotheses about the development of endocrine cells in the digestive tract. One model is that colonic endocrine cells and islets originate from a multipotent stem cell expressing PYY; interestingly, PYY expression continues for most of the enteroendocrine cells, with the exception of serotonin-producing cells (44, 45). The second step of differentiation might be determined by the position and migration of the endocrine cells (46, 47); it seems plausible that sorbin-containing cells migrate downward following the substance P/serotonin pathway.
In conclusion, the localization of sorbin in serotonin-containing cells confirms the special place of this peptide in the neuroendocrine digestive system. Further studies are now required to confirm the presence of a well defined sorbin neuronal system and to elucidate the role(s) of sorbin as a regulatory peptide in the digestive tract.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received April 8, 1997.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  F A. El Fadil-Nicol, F Berger, M Descroix-Vagne, and D Pansu Presence of sorbin in human digestive tract and endocrine digestive tumours Gut, February 1, 2000; 46(2): 182 - 190. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
), C7 (
), C10 (
), and C20 (
) and Y-C17-NH2 (
). Antiserum
dilution, 1:50,000.
