| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Department of Medicine II, Technical University of Munich (S.M., N.N., W.S., M.C., C.P.), and the Department of Clinical Chemistry, Harlaching Municipal Hospital (H.J.K.), D-81675 Munich, Germany
Address all correspondence and requests for reprints to: Christian Prinz, M.D., Second Department of Medicine, Technical University of Munich, Ismaninger Strasse 22, D-81675 Munich, Germany. E-mail: christian.prinz{at}lrz.tum.de
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2 pM), transforming growth factor-
(TGF; 10–30 ng/ml), and basic FGF (bFGF; 1–10 ng/ml) increased the
total number of cultured ECL cells. bFGF augmented the gastrin (1
pM)-induced response. ß-Neuronal growth factor (10 ng/ml)
and bFGF (2 ng/ml) decreased the programed death of ECL cells.
Interleukin-1ß (100 pg/ml, 24 h) stimulated apoptosis 2- to
3-fold in ECL cells, and simultaneous incubation with TGF (20 ng/ml)
or bFGF (2 ng/ml) significantly inhibited this effect. ECL cells
express specific receptors for gastrin, epidermal growth factor,
neuronal growth factor, and FGF. bFGF prolonged ECL cell survival by
inhibiting spontaneous apoptosis. Our data further indicate that TGF
and bFGF increase ECL cell number by inhibiting cytokine-induced
programed cell death. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Moreover, ECL cells reveal prominent persistence during chronic atrophic gastritis. This type of gastritis is caused by autoimmune mechanisms (classical type A gastritis) or by long term Helicobacter pylori infection. Autoimmune gastritis is characterized by severe atrophy of parietal and peptic cells and is associated with the production of autoantibodies against parietal cells and intrinsic factor (10, 11, 12, 13, 14). Dysplasia, microclusters, and even carcinoid formations are observed during autoimmune gastritis (10, 11, 12, 13, 14, 15). Atrophy of gastric mucosal cells with persistence of endocrine cells is also observed during multifocal gastritis caused by certain subtypes of H. pylori-inducing atrophy and hypochlorhydria after years of infection (14, 16, 17, 18, 19, 20). The reason for the persistence of ECL cells is still unknown. Several investigators have proposed that ECL cell dysplasia and carcinoids evolve independently of the reversible ECL cell hyperplasia during antisecretory therapy and hypergastrinemia (13, 15).
In the chronically inflamed mucosa, there is increased production of
proinflammatory cytokines and growth factors paralleled by elevated
serum gastrin levels (16, 17). Therefore, the proliferative and
apoptotic responses of epithelial cells might be altered by
hypergastrinemia, growth factors, or cytokines released from
lymphocytes or epithelial cells. The antral hormone gastrin or mucosal
growth factors may act as direct stimulants of ECL cell growth by
enhancing cell mitosis and DNA synthesis. Active proliferation of rat
gastric ECL cells has been observed in response to gastrin under
in vivo and in vitro conditions and is mediated
by increased DNA synthesis (4, 6, 7, 8, 21, 22, 23, 24). Stimulation of
bromodeoxyuridine (BrdU) incorporation was also observed in isolated
ECL cells in the presence of transforming growth factor- TGF
(21). In the case of malignant transformation, ECL cells stimulate
their own growth in vivo by autocrine growth factor
production (25, 26), specifically reported for epidermal growth factor
(EGF)/TGF, insulin-like growth factor type I, and basic fibroblast
growth factor (bFGF). The receptor subtypes for these growth factors
and the corresponding ligands, however, have only been characterized in
malignant tissues (25, 26, 27, 28). Therefore, these growth factors and the
corresponding receptors might also play an important role in ECL cell
proliferation under normal conditions.
Alternatively, a relative increase in ECL cells may be due to enhanced cell survival after inhibition of programed cell death, i.e. apoptosis (29). Neuronal growth factor (NGF), for example, is an important inhibitor of programed cell death in chromaffin cells and plays a crucial role in the development of sympathetic and sensory neurons (30). Apoptosis is a process of active cellular self-destruction observed in all eukaryotic cells and reflects changes in the cytological and genetic structures (29). Induction of apoptosis is an active, genetically regulated process and requires the coordinated expression of specific genes. Necrosis, in contrast, is a passive process. Up until now, induction of apoptosis in ECL cells and its regulation by growth factors or proinflammatory cytokines have not been determined. As the ECL cell number in vivo or in vitro can be increased by active cellular proliferation or by enhanced cell survival, it is possible that growth factors also affect cell density by inhibition of programmed cell death.
The present study was designed to investigate the presence of various growth factor receptors and to point out the effects of the corresponding growth factors on ECL cell apoptosis. Isolated rat ECL cells were used as an in vitro model that enables us to study direct effects of added substances on a defined cell population. In this model, mast cells and other endocrine cells, such as enterochromaffin (EC-) cells, are absent and do not interfere (2). We investigated total ECL cell number in response to growth factors to determine their potential effects on ECL cell growth or enhanced cell survival. DNA fragmentation in ECL cells undergoing apoptosis was determined in single cells by TdT-mediated dUTP nick end labeling (TUNEL) reaction and by a semiquantitative enzyme-linked immunosorbent assay (ELISA). Our results suggest that growth factors act in two ways to increase ECL cell number: via direct effects on active cell proliferation or via indirect effects on cell survival. Growth factor effects become more evident in the presence of proinflammatory cytokines, which may parallel clinical findings during chronic atrophic gastritis in humans.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
PCR amplification using an ECL cell complementary DNA (cDNA)
library
RNA was extracted from enriched ECL cells after the
initial isolation step (2 x 107 cells, harvested from
20 preparations), and an ECL cell cDNA library was prepared as
previously described using the phage vector 
(3, 4). Amplification
of the gastrin, EGF, FGF, and NGF receptor messages was carried out
using specific primers for each of the known receptor sequences in
rats. Primers were as follows: gastrin receptor,
5'-GCAGCACCAGGGCCTGTCCACA-3' (sense) and 5'-TTCTCCAATCTCCCAACCCCTCAC-3'
(antisense); EGF receptor, 5'-GCTGGGGAGAGGAGAACT-3' (sense primer) and
5'-GCAGGGCCCGTCACATTTTT-3' (antisense primer); FGF receptor subtype I,
5'-T- TCTCCAATCTCCCAACCCC-3' (sense) and 5'-GCAGCACCAGGGCCTGTCCA-3'
(antisense); NGF receptor, 5'-AAGGGGCGGGGCATTGTGGTA-3' (sense) and
5'-CCGTGTTGGCTTCAGGCTTATGC-3' (antisense); and hepatocyte growth factor
receptor (HGF; c-met), 5'-GCCCGGCCCTCTTCAGTTTAG-3' (sense)
and 5'-TTGGAGGATGGGCCGTTGTGTT-3' (antisense). PCR was performed using
the Prime Zyme Kit (Biometra, Gottingen, Germany) and the following
temperature cycle profile: denaturation for 2 min at 95 C, followed by
30–35 cycles of annealing for 1 min at specific primer temperature,
extension for 1 min at 72 C, denaturation for 1 min at 95 C, and
termination by 7 min at 72 C. The exact cycle length and annealing
temperatures are specified in Results. Horizontal 1.9%
agarose gel electrophoresis (ultrapure agarose, Life Technologies,
Heidelberg, Germany) was performed. Primers were selected by software
analysis using the receptor sequences available on Gene Works
(IntelliGenetics, Campbell, CA). All amplification products were eluted
from the agarose gel using a QIAEX II Gel Extraction Kit (Qiagen,
Hilden, Germany) and were sequenced by a commercial institute
(MediGene, Martinsried, Germany). PCR products revealed homology to the
known receptor sequences. Negative controls were performed as follows.
In one of the controls performed during PCR analysis, no cDNA was added
because a cDNA library from isolated cells was used as a template
(commercially performed by Invitrogen, San Diego, CA). In this system,
amplification of DNA is excluded (32). In another negative control, no
cDNA but the remaining PCR constituents were added to exclude
contamination within the PCR reaction. RNA was isolated, followed by a
deoxyribonuclease digest, and PCR was performed without the addition of
reverse transcriptase. In the following step, RNA and the reaction
product obtained without reverse transcriptase addition were used as
substrate for PCR. None of these samples gave a base pair
product.
Immunocytochemistry
Isolated, purified ECL cells (4–5 x
104/slide) were incubated for 48 h on glass slides
coated with Cell-Tak (Becton Dickinson; 1:1 dilution with 0.5
M NaHCO3) in DMEM culture medium. Subsequently,
culture medium was removed, and staining was performed with fluorescein
isothiocyanate (FITC)-conjugated antibodies or with the
avidin-biotinylated enzyme complex method (Vector Laboratories,
Burlingame, CA). Staining for the EGF receptor was performed by fixing
cells in phosphate-buffered formaldehyde (3.7%) at room temperature
for 12 min and washing three times in PBS (0.15 M). Cells
were incubated with the polyclonal antihuman EGF receptor at a 1:50
dilution (Upstate Biotechnology, Lake Placid, NY) for 3 h at 37 C.
Slides were washed in PBS, and the secondary FITC-conjugated secondary
antibody (1:200) was added for 1 h at room temperature. PC12 cells
were used as a positive control for EGF receptor staining. Negative
controls were performed under identical conditions without adding the
primary antibody. Staining for the FGF receptor was performed in cells
fixed with 4% paraformaldehyde for 10 min, and incubation with
polyclonal antichicken FGF receptor antibody (1:25; Upstate
Biotechnology) was performed for 1 h at room temperature. Swiss
3T3 cells were used as a positive control. Specific antibody binding
was visualized by secondary staining with a FITC-conjugated antirabbit
antibody (Sigma). Immuocytochemistry for NGF receptor was performed by
fixing the cells in acetone at -20 C for 10 min and incubating with a
monoclonal antibody at a dilution of 1:10 (Becton Dickinson) at room
temperature for 1 h. Bound antibodies were localized using the
enzyme complex method. PC12 cells were used as a positive control.
Immunocytochemistry for autocrine TGF or bFGF production was
performed by fixing cultured cells with acetone (50%)-methanol (50%)
for 2 min. Subsequently, cells were incubated with anti-TGF or
anti-bFGF antibody [Calbiochem (Cambridge, MA) or Upstate
Technologies, respectively] at a dilution of 1:10 for 2 h at room
temperature. Staining was visualized using a biotinylated antirabbit
IgG secondary antibody and subsequent peroxidase staining.
Measurement of ECL cell number
ECL cell number was measured using a commercially available
kit. This procedure determines mitochondrial formazan production in
living cells that are capable of reducing a yellow-colored tetrazolium
salt into a red-colored formazan derivative (EZ4Y kit, Biomedica,
Vienna, Austria). The EZ4Y kit determines the increase in cell quantity
after the addition of a growth factor of interest, but does not
differentiate between active proliferation or passive survival of
cells. The OD is linearly correlated with cell number; 0.05 U
corresponds to 5 x 103 cells, and 0.1 U corresponds
to 1 x 104 cells. These studies were performed with
ECL cells that were cultured in 96-well plates precoated with Matrigel
(dilution 1:5) at approximately 11,000 cells/well. After 24 h, the
initial growth medium was replaced by a serum-free DMEM medium
supplemented with 2% BSA, 1% N1 supplement, 1 pM gastrin,
10 mg/ml gentamicin, and 100 nM hydrocortisone. The
influence of gastrin, TGF, bFGF, or ßNGF was determined by
addition of the growth factor of interest to the culture medium.
Incubation was carried out over 72 h, and 25 µl of the dye
substrate were added subsequently to each well. OD was recorded at 450
nm using a microplate reader (Bio-Rad, Munich, Germany) after 4 h,
showing significant differences in metabolic capacity. Data are
expressed as the relative increase as a percentage of the basal
value.
Determination of ECL cell apoptosis
Two different techniques were used for the determination
of ECL cell apoptosis. First, the in situ cell death
detection kit, which is based on the TUNEL reaction, was used to
determine ECL cell death (POD, Boehringer Mannheim, Mannheim, Germany).
Isolated ECL cells were cultured over 24 h on Cell-Tak (dilution
1:1 with 0.5 M NaHCO3)-coated glass slides.
After 24 h, the initial culture medium was removed, and DMEM/F12
medium was added for another 24 h with vehicle, the growth factor
of interest, or, alternatively, with vehicle, interleukin-1ß
(IL-1ß; 100 pg/ml), or IL-1ß with the growth factor of interest.
Subsequently, cells were fixed in Bouin’s solution over 15 min. DNA
strand breaks, generated by apoptosis, were detected in ECL cells by
incubation with biotinylated nucleotides and end labeling with terminal
transferase reaction. Staining was visualized using peroxidase
reaction.
Second, a photometric immunoassay for in vitro determination of histone-associated DNA fragmentation was performed. This kit is based on a quantitative sandwich enzyme immunoassay using mouse monoclonal antibodies directed against histone complexes associated with DNA breaks, which can be detected during programmed cell death (cell death detection ELISA, Boehringer Mannheim, Ingelheim, Germany). Isolated ECL cells were cultured over 24 h on Matrigel-coated six-well plates, incubated with vehicle or stimulants over 24 h, trypsinated, washed with PBS, and lysed. Antibody binding with subsequent peroxidase reaction was measured in an ELISA reader at a wavelength of 405 nm.
Statistical analysis
Results are shown as the mean ± SEM. Data were
analyzed by Student’s t test. P < 0.05 was
considered significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2) PCR of ECL cell complementary DNA library (Fig. 1
)
The PCR reaction was performed using a ECL cell complementary DNA
library as template and specific primers for the rat gastrin, EGF, FGF
(subtypes I and IV), NGF, and HGF receptors. In lane 1 of Fig. 1, PCR
of the gastrin receptor is shown. We generated primers assuming the
presence of the long isoform receptor (33). The PCR reaction for the
rat sequence yielded a primer product of the expected size (563 bp)
and, therefore, determined the long isoform (452 amino acids) of the
rat gastrin receptor in the rat ECL cell cDNA library. This product
corresponded to the human sequence and to the previously published
sequence of the cholecystokinin-B receptor in rat brain (33, 34). Lane
2 shows the negative control for the rat gastrin receptor performed
without cDNA. Lanes 3, 5, and 7 show the PCR results with specific
primers for the rat EGF (lane 3), FGF (lane 5; subtype I), and NGF
(lane 7) receptors and using the identical cDNA library from ECL cells
as template (n = 3 independent experiments). Lanes 4, 6, and 8
present the corresponding negative controls. For the rat EGF receptor
(lane 3), the annealing temperature was 60 C, and 35 cycles were
performed, yielding a product of 404 bp. For the rat FGF receptor
subtype 1 (lane 5), the annealing temperature was 63 C, and
amplification was carried out over 30 cycles, revealing a product of
506 bp. Receptor subtype IV was not detected. For rat NGF receptor
subtype 1 (lane 7), the annealing temperature was 63 C at 30 cycles,
yielding a product of 481 bp (n = 3 for each PCR experiment). All
bands were eluted from the gels, sequenced, and found to be homologous
to the sequences known from other rat tissues. PCR was also performed
with primers specific for the rat HGF receptor (c-met
protooncogene), but no specific product was obtained (n = 3; not
shown).
|
) shows formalin-fixed cells incubated with a polyclonal
antihuman EGF receptor antibody (dilution 1:50). Binding was
visualized with an antisheep FITC-conjugated secondary antibody. A
positive control was performed with PC12 cells, stained by indirect
immunofluorescence (Fig. 2b). A negative control was performed by
omission of the first antibody (Fig. 2c). Figure 2d shows
paraformaldehyde-fixed cells stained by indirect immunofluorescence and
labeled with a polyclonal antichicken FGF receptor as primary antibody
(dilution 1:25). The positive control was performed in Swiss 3T3 cells
(Fig. 2e). Figure 2f shows the negative control for the FGF-R. Figure 2g presents acetone-fixed ECL cells stained with the antirat NGF
receptor monoclonal antibody (dilution 1:10). Staining was visualized
using a biotinylated antimouse antibody and peroxidase staining with
diaminobenzidine. The positive control was performed using PC12 cells
(Fig. 2h), and the negative control was performed without addition of
the first antibody (Fig. 2j).
|
and Table 1)). The metabolic capacity of cultured ECL cells was determined under
basal conditions or during gastrin or growth factor stimulation. The
increase in OD using the EZ4Y kit served as a parameter for the total
amount of living cells. After a total 96-h culture, OD ranged from
0.025–0.05 U, corresponding to 2.5–5 x 103 cells.
In Fig. 3, A–E, results are shown as the relative increase in the OD
above basal values. Gastrin (Fig. 3A) dose dependently stimulated ECL
cell number and increased the number of living cells 2- to 3-fold, as
indicated by increasing optical density at 450 nm. The maximal effect
was observed at 10-11 M gastrin, similar to
the results of previous experiments performed with FBS (4). The
EC50 was about 2 pM gastrin (n = 4
independent experiments).
|
|
also increased the number of ECL cells (tested at a
concentration of 10–30 ng/ml), with a maximal effect at 20 ng/ml (Fig. 3B
; n = 5) after 96 h of incubation. There was no additive
effect of TGF applied in combination with gastrin (data not shown;
n = 5). bFGF increased ECL cell number after 96 h of
incubation, and the maximal response was observed at a concentration of
2 ng/ml (P < 0.05 vs. basal; Fig. 3C;
n = 5 independent experiments). bFGF (5 ng/ml) produced additive
effects in combination with gastrin (10-12 M;
n = 5 independent experiments; Fig. 3D). No significant increase
in ECL cell quantity was observed after incubation with ßNGF (n
= 5; Fig. 3E; 1–100 ng/ml). Similarly, no significant increase was
observed after incubation with HGF (1–60 ng/ml; data not shown; total
of n = 6 independent experiments).
In a separate set of experiments (n = 5), ECL cells were isolated
and cultured on six-well plates. Cell numbers were determined after the
addition of vehicle, gastrin (10-11 M), bFGF
(2 ng/ml), and TGF (20 ng/ml) after 72 h of incubation.
Following cell removal from the six-well plates, the number of living
cells was counted after trypan blue exclusion. The total number of
trypan blue-negative cells was counted using a Neubauer chamber. All
three growth factors stimulated ECL cell number, corresponding to the
results using formazan incorporation (Table 1).
5) Control of spontaneous apoptosis by gastrin, TGF, ßNGF, or
bFGF (Fig. 4)
Programmed cell death of ECL cells was determined under
resting conditions and in the presence of growth factors using the
TUNEL reaction (Fig. 4A) or a DNA fragmentation ELISA (Fig. 4B). For
analysis of the TUNEL reaction, cells were counted in visual fields
with more than 100 cells and expressed as a percentage of cells with
positive staining. After 24 h of recovery, purified ECL cells were
incubated for another 24 h with vehicle, TGF (20 ng/ml),
gastrin (10 pM), ßNGF (10 ng/ml), or bFGF (2 ng/ml). As
shown in Fig. 4A, the spontaneous apoptotic rate ranged from 4–6% of
ECL cells in N1-supplemented medium. Gastrin and TGF had no
significant effect on ECL cell death. ßNGF significantly decreased
the basal rate (n = 6; P < 0.05). bFGF was the
most effective growth factor tested and reduced the apoptosis rate of
ECL cells to 3–4% (reduction of 33%; n = 6; P
< 0.01).
|
). After 24 h of
recovery, ECL cells were treated with vehicle or the different growth
factors over 24 h and lysed. The lysates were transferred to
96-well plates in which the OD was measured at 405 nm (Fig. 4B; n
= 6 independent experiments). Similar results were obtained using the
DNA fragmentation ELISA compared with the data generated by the TUNEL
reaction. OD ranged from 0.1–0.2 relative units under basal conditions
and is shown as a percentage of the change vs. basal
density. Gastrin and TGF had no significant effect on ECL cell
death. ßNGF significantly decreased the basal rate (n = 6;
P < 0.01). Again, bFGF (2 ng/ml; P <
0.01) was the most effective factor tested, and this effect
corresponded to the inhibition of cell death measured by the TUNEL
reaction.
6) Apoptosis of ECL cells in the presence of proinflammatory
cytokines (Fig. 5)
Programmed ECL cell death was also investigated in the presence of
the proinflammatory cytokine IL-1ß and simultaneous addition of
gastrin or growth factors. Similar to the experiments performed in Fig. 4, ECL cell apoptosis was measured by the TUNEL reaction (Fig. 5A) or
by DNA fragmentation ELISA (Fig. 5B). After 24 h of recovery,
isolated and cultured ECL cells were treated over 24 h with
vehicle, IL-1ß (100 pg/ml), or IL-1ß (100 pg/ml) plus gastrin
(10-11 M), TGF (20 ng/ml), or bFGF (2
ng/ml). Staining for apoptotic cells was determined by the TUNEL
reaction (Fig. 5A) under basal conditions in 4–6% of ECL cells,
similar to the results shown in Fig. 4A. As shown in Fig. 5A, IL-1ß
(100 pg/ml) increased the number of apoptotic cells 3-fold to 12–16%
of the total cell number. Therefore, IL-1ß can be regarded as a
potent stimulant of ECL cell apoptosis. Incubation of ECL cells with
IL-1ß and gastrin (10 pM) decreased the stimulation of
apoptosis by approximately 40%, but this effect did not attain
statistical significance using Student’s t test. The
simultaneous presence of TGF (20 ng/ml; n = 6;
P < 0.01) in IL-1ß-treated samples significantly
inhibited the IL-1ß effect. Most prominently, bFGF (2 ng/ml) reduced
the amount of apoptotic ECL cells to 3–4%, corresponding to basal
values (n = 6; P < 0.01 vs. incubation
with IL-1ß alone).
|
). Identical results were achieved
using this technique. IL-1ß increased the amount of apoptotic cells
2- to 3-fold. Gastrin decreased this effect by approximately 30% (not
significant in six experiments). In contrast, TGF (20 ng/ml; n
= 6; P < 0.05) significantly inhibited the induction
of apoptosis, and bFGF (2 ng/ml) reduced the apoptotic rate below
basal (n = 6; P < 0.05).
7) Autocrine production of bFGF and TGF in ECL cells (Fig. 6)
For detection of autocrine growth factor production,
immunocytochemistry of 48-h isolated and cultured ECL cells was
performed. For determination of bFGF production, two different
antibodies and at least five different fixation methods were used. No
staining was detected in isolated ECL cells (total of eight
experiments). The Swiss 3T3 cell line served as a positive control. In
contrast, autocrine production of TGF was observed in 92% of the
counted cells, corresponding to the degree of enrichment. Figure 6a
shows isolated ECL cells incubated with an anti-TGF polyclonal
antibody and secondary antibody staining using a peroxidase reaction.
Figure 6b illustrates a positive control performed with a human
pancreatic cell line (LnCaP). Figure 6c demonstrates a negative control
performed without addition of the primary antibody (representative of
two additional experiments).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, have been detected in extracts of the
chronically inflamed gastric mucosa (25, 35, 36). Autocrine production
of bFGF and TGF has been detected in human ECL cell carcinoids
(25, 26, 27). These growth factors might affect ECL cell proliferation by
active induction of ECL cell growth or enhancement of cell survival
(25, 26). Finally, it may be possible that the interaction of gastrin
or growth factors with proinflammatory cytokines released in the
chronically inflamed mucosa (16) prevents ECL cell death during chronic
gastritis. In the present study, we used highly purified rat ECL cells to study the direct effects of agents and to minimize the influence of confounding factors. In rats, ECL cells are the most abundant and display a function similar to that in the human stomach. Moreover, mast cells are absent in rat gastric mucosa and do not interfere (2, 3). In this recently developed model of isolated ECL cells, it has been shown that gastrin directly stimulates ECL cell replication in vitro (3, 4). Previously used culture medium, however, contained FBS as an additive (3, 4). FBS has been shown to contain small amounts of gastrin and poorly standardized concentrations of growth factors, such as FGF and TGF, which complicates the interpretation of results obtained with FBS as a culture additive. Therefore, we modified our previous technique by using a new culture supplement free of FBS. We observed that progesterone, putresceine, and insulin were essential ingredients in the medium, similar to the conditions used for the culture of chromaffin cells (30).
In the current study, the total number of cultured ECL cells was measured by mitochondrial formazan formation using the EZ4Y kit. This technique determines the total number of living cells, which is a function of active proliferation and passive cell survival due to prevention of apoptosis. The EZ4Y kit was developed in analogy to standard thymidine or bromodeoxyuridine incorporation procedures. For gastrin, stimulation of formazan production was similar to previous results using BrdU incorporation in ECL cells, indicating that the kit determines an increment in cell number (4). In addition, the determination of cell number using the EZ4Y kit was consistent with the results of the actual cell counts. Therefore, this technique appears to reflect growth factor effects on cell number rather than only changes in mitochondrial function.
Using the EZ4Y kit, this technique revealed a 2- to 3-fold increase in ECL cell number in response to gastrin. The ED50 for this growth response to gastrin was 50-fold lower (i.e. 1 pM) than the gastrin concentration that elicited half-maximal stimulation of histamine release (3, 4). Gastrin did not significantly affect spontaneous or cytokine-induced cell death. Therefore, the increase in ECL cell number in response to gastrin appears to result from active induction of proliferation. Gastrin receptor distribution, receptor affinities, and gastrin effects on cell proliferation observed in the current study corresponded to the observations in vivo (5, 37). Gastrin stimulated cell proliferation 2- to 3-fold in vivo and in vitro within 2–7 days of incubation (4, 23). Maximally effective concentrations were observed at 50 pM in vivo (38), which is similar to the concentration used in the current protocol (10-11 M).
The focus of the present study was to identify EGF, FGF, and NGF
receptor subtypes on normal ECL cells and to investigate the effects of
the corresponding ligands, TGF, bFGF, and ßNGF. Antibody staining
of ECL cells revealed specific staining for the EGF receptor at the
plasma membrane. Staining was verified using PC12 cells as a positive
control. The EGF receptor is a 170-kDa transmembrane glycoprotein with
tyrosine kinase activity (39). Many gastric carcinoma tissues coexpress
EGF receptor and TGF genes at higher levels than normal gastric
mucosa (36, 40, 41). Overexpression of EGF receptors or mutation leads
to uncontrolled cell division, e.g. in breast tumors (36, 39, 40, 41). TGF specifically binds to the EGF receptor and plays an
important role as an autocrine growth factor in the progression of
gastric carcinomas (40), in naive ECL cells, and in ECL cell tumors
(21). In the current study, we observed the presence of the EGF
receptor and of TGF in isolated, untreated ECL cells, suggesting a
role for autocrine growth stimulation. After 96 h of incubation,
TGF increased the number of isolated ECL cells 2-fold, but failed to
augment the response to gastrin. The missing additive effect of TGF
in combination with gastrin suggests that these factors may share
similar intracellular transduction pathways and/or act in a similar way
to increase ECL cell number. TGF had no significant effect on the
spontaneous apoptosis of ECL cells, as demonstrated in Fig. 4.
Therefore, it may be speculated that in the presence of TGF, the
increase in absolute cell number results from induction of cell
proliferation, similar to gastrin and consistent with other reports
based on BrdU incorporation in ECL cells (21).
The distribution of the EGF receptor and the corresponding ligand and
the maximally effective TGF concentration found in isolated ECL
cells appear similar to those in vivo. TGF expression has
been detected in malignant and normal rat ECL cells in vivo
(37) and in other gastric mucosal or intestinal epithelial cells of
rats and guinea pigs (36, 42). TGF staining has also been observed
in the human gastric mucosa at the base of the gastric glands, which is
the presumable location of ECL and parietal cells, and has been
reported to be a potent mitogen for cells of the gastrointestinal tract
(42). Previous dose-response curves for growth stimulation of gastric
mucosal cells with TGF (ED50, 1–2 ng/ml) determined a
maximal effect between 1–100 ng/ml (43), which is comparable to our
present results. In addition, the presence of the EGF receptor has been
detected in the mammalian gastric mucosa (36) and in rat ECL cell
carcinoids (37). Therefore, TGF and the EGF receptor may be
important factors for ECL cell proliferation in vivo and
in vitro. TGF inhibited IL-1ß-induced cell death
(apoptosis), indicating that ECL cell apoptosis during conditions of
chronic gastric inflammation is also affected by this growth factor. As
we detected autocrine production of TGF in ECL cells, this factor
may be of crucial importance for the survival of ECL cells in the
presence of proinflammatory cytokines such as IL-1ß.
FGF receptor subtype I was also detected in rat gastric ECL cells. At least four distinct receptor genes (FGF receptors 1–4) have been identified encoding four subtypes of FGF receptor proteins with many structural variations due to alternative splicing (44, 45). These receptor subtypes mediate different functions and are characterized by different affinities (44, 45). The high affinity receptors for FGF belong to a protein tyrosine kinase superfamily of proteins. FGF receptors 1 and 2 both bind acidic FGF and bFGF, an 18-kDa protein that has been shown to be a potent endothelial cell mitogen (46). The FGF receptor has not been described in human ECL cells to date. Gastrointestinal mucosal cells contain immunoreactivity for FGF receptor type 1 (47), but receptor distribution on mucosal cells has not been investigated in detail. For this point, our study reports a new finding of the presence of FGF receptor type 1 on isolated ECL cells in vitro.
Similar to the effects of TGF, bFGF might also play an important
role in the survival of ECL cells in the presence of proinflammatory
cytokines. bFGF production has been detected in hyperplastic human ECL
cells and has been implicated to play an important role in human ECL
cell growth in vivo (26). Due to the results demonstrated in
Fig. 3, bFGF increased the number of isolated ECL cells in culture (2-
to 4-fold) and showed an additive response in combination with gastrin.
The production of bFGF has been previously localized in human ECL cell
carcinoids and in human ECL cells in vivo, especially during
continuing hypergastrinemia (26). Dose-response curves for growth
stimulation of BALB or 3T3 cells with bFGF ranged between 1–50 ng/ml
(ED50, 0.1–0.25 ng/ml), with a maximal effect at 5–10
ng/ml, similar to the concentrations used in this study (48, 49).
Although the presence of bFGF shows a remarkable difference in isolated
ECL cells and in vivo, the FGF effects and dose-response
curves reported in the current study are consistent with previous
observations.
The dose-response curves for bFGF and TGF show an inverted U shape,
and similar results were observed in previous studies using ECL cells,
in which gastrin, TGF, or Pituitary Adenylate Cyclase Activating
Peptide stimulated ECL cell proliferation (4, 50, 51).
Furthermore, TGF or bFGF stimulated mitogenesis of gastric mucosal
cells (52) or Swiss 3T3 fibroblasts in a similar U-shaped concentration
curve (48, 49). Therefore, this effect may be a cell-specific finding
in response to the growth factors tested.
bFGF inhibited spontaneous and cytokine-induced apoptosis. This finding
is consistent with observations in chromaffin and ovarian cells, in
which FGF prevented programmed death and led to a prolonged survival
rate (53, 54). Therefore, bFGF may act in two ways to enhance ECL cell
numbers: 1) FGF may prolong ECL cell survival via inhibition of
programmed cell death; and/or 2) FGF may act as an active stimulant of
ECL cell growth. The prominent effects on cell apoptosis as well as the
additive effect of FGF in combination with gastrin might allow the
conclusion that the major effect of FGF on ECL cells (in contrast to
that of TGF or gastrin) is the inhibition of cell survival.
Autocrine production of bFGF was not detected in normal ECL cells. This finding is surprising because autocrine production of bFGF has been reported for ECL cell tumors and during gastrin-induced hyperplasia of ECL cells in humans (26, 45). One explanation for the missing detection in ECL cells may be that several isoforms of FGF processed by alternative splicing cannot be detected by the primary antibody, as reported previously (55). In addition, no rat-specific antibody was available. Finally, FGF may be localized only in the nucleus due to processing procedures and may not be accessible for the primary antibody, similar to observations obtained in neurons (55). Therefore, autocrine production of FGF cannot be excluded. Alternatively, bFGF may not be an important autocrine growth factor for normal ECL cells, and bFGF production may occur only in malignant tissues, as observed in human endocrine tumors (26). In this case, bFGF production may be an indicator for malignant transformation.
The NGF receptor in ECL cells was of subtype I. In other cell systems, two distinct NGF receptors have been identified (30). NGF receptor subtype I is a 140-kDa glycoprotein with high and low affinity status and is recognized as a TRK protooncogene, whereas subtype II is a 75-kDa low affinity receptor without full biological activity (30). In cultured ECL cells, we were able to detect the presence of the type I NGF receptor on ECL cells. ßNGF inhibited spontaneous ECL cell apoptosis significantly at a concentration of 10 mg/ml. In PC12 cells, NGF prevented programmed cell death and led to a prolonged survival rate at identical concentrations (54). ßNGF did not increase the number of ECL cells significantly and therefore appears to have no effect on cell proliferation, in contrast to chromaffin cells of the adrenal gland, in which proliferating effects of NGF were reported (30).
The HGF receptor was not detected in ECL cells by RT-PCR. The corresponding ligand of this receptor type is HGF. HGF has been implicated to play a role in the H. pylori-infected, chronically inflamed gastric mucosa (56). This factor, however, did not exert any effect on ECL cells. As the corresponding receptor was also missing in ECL cells, HGF does not seem to be of major importance for ECL cell proliferation.
Our present in vitro data support the idea that ECL cell quantity can be affected by stimulation or inhibition of programmed cell death in the presence of cytokines and growth factors. We have previously reported that short term exposure (60 min) of ECL cells to IL-1ß results in sustained functional impairment and inhibits histamine synthesis (9). Due to the current results, IL-1ß induced DNA fragmentation in ECL cells after 24 h of incubation. DNA fragmentation is a characteristic finding occurring during programmed cell death (29, 57). DNA fragmentation in ECL cells was determined by the TUNEL reaction and histone-associated ELISA. These experimental procedures are independent techniques of measurement of apoptosis, but gave similar results. These results are in accordance with previously published results in islet cells of the pancreas (29). In this cell type, apoptosis can be induced by the addition of IL-1ß after 24 h of incubation at concentrations of 10–100 pg/ml, similar to the conditions applied in the present protocols (9, 29). Determination of apoptosis by the TUNEL reaction indicated that ECL cells undergo programmed cell death at a low rate (4–6%) under spontaneous conditions and more prominently in the presence of the proinflammatory cytokine IL-1ß. ECL cell apoptosis and ECL cell DNA strand breaks were detected by specific antibody labeling in 12–16% of ECL cells after incubation with IL-1ß. Therefore, IL-1ß can be regarded as an active inducer of apoptosis in ECL cells.
We speculate that this interaction detected in isolated ECL cells
in vitro may parallel clinical observations in the human
gastric mucosa in vivo. During chronic atrophic gastritis,
interaction of growth factors and cytokines may play an important role
in the survival rate of ECL cells. TGF and bFGF inhibited cell
survival more prominently in the presence of IL-1ß. Therefore, it may
be speculated that in addition to gastrin, bFGF and TGF should be
regarded as important growth factors for ECL cell proliferation and
survival and might play an important role in prolonged ECL cell
survival during chronic atrophic gastritis.
|
Footnotes |
|---|
Received January 29, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. J Surg Res 60:293–302[CrossRef][Medline]
and its receptor in human neuroendocrine cell
line. Regul Pept 40:000 (Abstract 217)
: expression, regulation, and biological
activities. Pharmacol Rev 47:51–85[Medline]
is a potent mitogen for primary cultures of guinea pig gastric
mucous epithelial cells. Am J Physiol 265:G361–G369
in the
enterochromaffin-like cell tumor autonomy in an African rodent
mastomys. Gastroenterology 111:1212–1223[CrossRef][Medline]
This article has been cited by other articles:
 |
|
 |
|
  D. M. Pritchard, D. Berry, S. M. C. Przemeck, F. Campbell, S. W. Edwards, and A. Varro Gastrin increases mcl-1 expression in type I gastric carcinoid tumors and a gastric epithelial cell line that expresses the CCK-2 receptor Am J Physiol Gastrointest Liver Physiol, October 1, 2008; 295(4): G798 - G805. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. N. Jain and L. C. Samuelson Differentiation of the Gastric Mucosa II. Role of gastrin in gastric epithelial cell proliferation and maturation. Am J Physiol Gastrointest Liver Physiol, November 1, 2006; 291(5): G762 - G765. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Zanner, M. Gratzl, and C. Prinz Expression of the endocytic proteins dynamin and amphiphysin in rat gastric enterochromaffin-like cells J. Cell Sci., May 1, 2004; 117(11): 2369 - 2376. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Kazumori, S. Ishihara, M. A. K. Rumi, C. F. Ortega-Cava, Y. Kadowaki, and Y. Kinoshita Transforming growth factor-{alpha} directly augments histidine decarboxylase and vesicular monoamine transporter 2 production in rat enterochromaffin-like cells Am J Physiol Gastrointest Liver Physiol, March 1, 2004; 286(3): G508 - G514. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Hui, A. Nourparvar, X. Zhao, and R. Perfetti Glucagon-Like Peptide-1 Inhibits Apoptosis of Insulin-Secreting Cells via a Cyclic 5'-Adenosine Monophosphate-Dependent Protein Kinase A- and a Phosphatidylinositol 3-Kinase-Dependent Pathway Endocrinology, April 1, 2003; 144(4): 1444 - 1455. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A Varro, P J Noble, L E Wroblewski, L Bishop, and G J Dockray Gastrin-cholecystokininB receptor expression in AGS cells is associated with direct inhibition and indirect stimulation of cell proliferation via paracrine activation of the epidermal growth factor receptor Gut, June 1, 2002; 50(6): 827 - 833. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Zanner, G. Hapfelmeier, M. Gratzl, and C. Prinz Intracellular signal transduction during gastrin-induced histamine secretion in rat gastric ECL cells Am J Physiol Cell Physiol, February 1, 2002; 282(2): C374 - C382. [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 |
