JOURNAL OF SURGICAL RESEARCH 78, 161-168
ARTICLE NO. JR975230
Hepatocyte
Growth
(1998)
@ ~/e/~-A/oMf.o
.~
Factor Stimulates Fetal Gastric
Gell Growth in Vitro 1
Epithelial
Wuyi Kong, Ph.D., Laurence F. Yee, M.D., and SeanJ. Mulvihill, M.D.
Gastrointestinal Research Laboratory, Department of Surgery, University of California at San Francisco,
533 Parnassus Avenue, San Francisco, California 94143-0788
Submitted for publication May 21, 1997
Background. The growth and development of the fetal gastrointestinal tract is likely mediated, in part, by
peptide growth factors. We compared the mitogenic
effects of graded doses of hepatocyte growth factor
(HGF) to epidermal growth factor (EGF), transforming
growth factor-a (TGF-a), and insulin-like growth factor-1 (IGF-1) on fetal rabbit gastric epithelial cells.
Materials and Methods. Fetal rabbit gastric epithelial cells were purified by mechanical dissociation and
selected culture and grown in short.term (24 h) and
long-term (12 days) culture. Stimulation offetal gastric
epithelial cell growth in response to individual peptide
growth factors was measured by [~Jthymidine incorporation and cell counting.
Results. In short-term culture, HGF stimulated [3HJ.
thymidine incorporation in a dose-dependent manner
from a threshold at 10 pM to a maximum at 100 pM.
For EGF and TGF-a, maximal stimulation occurred at
100 pM. For HGF, maximal [3HJthymidine incorporation was 3.6 :t 0.7 times basal. For EGF and TGF -a,
maximal [3HJthymidine incorporation was 4.3 :t 0.4,
and 3.6:t 0.4times basal, respectively. For IGF-1, maximal [3HJthymidine incorporation was only 70% of the
maximal effect observed for the other growth factors
tested. Rabbit amniotic fiuid increased [3HJthymidine
uptake in a dose-dependent manner. In long-term culture, purification to greater than 90% epithelial cells
was attained alter 12 days treatment. For HGF, EGF,
TGF-a, and 20% rabbit amniotic fluid, significant increases in cell number above control (P < 0.05) were
observed at 1 nM concentrations. None of these individual factors, however, increased cell growth as significantly as that of 10% fetal bovine serum.
Conclusions. Our results suggest that: (1)HGF stimulates [3HJthymidine uptake and cell proliferation in
fetal rabbit gastric epithelial cells in vitro, and (2)
HGF's mitogenic effect on fetal rabbit gastric epithelial cell growth is comparable to that observed for EGF
and TGF .a, but superior to the effect observed for IGF1. ~ 1998
Academic
Pre_Key Words: hepatocyte growth factor; epidermal
growth factor; transforming growth factor-a; insulin1This work was supported by National Institute of Health Grant
DK-24773.
like growth factor-!;
otic fluido
fetal gastric development;
amni-
INTRODUCTION
The growth and development ofthe fetal gastrointestinal tract is likely mediated, in part, by peptide growth
factors. Hepatocyte growth factor (HGF) was originally
identified as a serum factor mediating liver regeneration and appears to be the most potent identified mitogenic factor for hepatocytes [1, 2]. Recently, HGF has
been found to playa role in the regulation of complex
biological processes, including fetal development, angiogenesis, tissue regeneration, and malignant transformation [3]. HGF has been localized in many organs,
in~luding the gastrointestinal epithelium and salivary
glands (4). Furthermore, HGF stimulates a variety of
epithelial cell types other than hepatocytes, including
gastric epithelial cells from mature rabbit and rat [5,
6]. Recently, HGF has been found to accelerate wound
repair in rabbit gastric mucosal cells and rat gastric
mucosal cells [7,8]. Although the effects ofHGF on the
fetal gastrointestinal tract are unknown, it appears to
be essential for fetal development [9-11].
Other peptide growth factors, including epidermal
growth factor (EGF) and transforming growth factor-a
(TGF -a), are known to stimulate the proliferation of
gastric epithelial cells [12-15] and induce recovery
from gastric mucosal injury [16, 17]. Insulin-like
growth factor-1 (lGF-1) has been shown to have mitogenic effects on gastrointestinal tissue in rat [18] and
canine epithelial cells [13]. AlI these growth factors
have been found in amniotic Huid [19-22]. These
growth factors may have a role in the regulation of
gastric development, since the fetus swallows significant amounts of amniotic Huid during gestation. The
relative roles of peptide growth factors 6ft the development of the fetal gastrointestinal tract are unknown.
We hypothesized that HGF promotes growth offetal
gastric epithelial cells in culture. To study this, we compared HGF to other potential mitogens, -EGF, TGF-a,
and IGF-1, on fetal gastric epithelial cell growth in
primary culture.
161
0022-4804/98 $25.00
Copyright @ 1998 by Academic Press
Al! rights of reproduction in any form reserved.
162
JOURNAL OF SURGICAL RESEARCH:VOL. 78, NO. 2, AUGUST 1998
MATERIAlS AND METHODS
Reagents. Recombinant hurnan EGF, recombinant human TGFa, and recombinant human IGF-1 were purchased from Bachem Inc.
(Torrance, CA). Recombinant human HGF was a gift from Dr. R. H.
Schwall (Genentech Inc., South San Francisco, CA). Iscove's medium,
fetal bovine serum (FBS), bovine serurn albumin (BSA), penicillin,
streptomycin, and fungizone were purchased from Cell Culture Facility at the University of California San Francisco. Primary tissue
culture dishes and plates were purchased from Falcon Labware (Oxnard, CA). The monoclonal antibody against human cytokeratin
(AEl/AE3) was purchased from Boehringer Mannheim Biochemicals
(Indianapolis, IN). The second anti-mouse antibodyand avidin-biotin complex (ABC) kit were purchased from Vector Lab (Burlingame,
CA). [3H]Thymidine was purchased from Amersham (Arlington
Heights, IL). Time-mated New Zealand white rabbit does were obtained from a commercial breeder (Grimaud Rabbitry, Stockton, CA).
Rabbit amniotic fluid was collected from time-mated New Zealand
white rabbit does in our laboratory.
Purification ofgastric epithelial cells. Time-mated New Zealand
rabbits were anesthetized with sodium pentobarbital (50 mg/kg) via
an ear vein at day 25 of gestation. Fetuses were removed from the
uterus by cesarean section under sterile conditions. A transverse
laparotomy was made in the fetal abdomen with a scalpel. A gastrotomy was made in the fundus of stomach with fine scissors, and the
stomach was carefully everted through this opening. After washings
in Ca2+-and Mg2+-freeHank's buffer (HBSS), fue everted stomachs
were placed into fresh HBSS. The gastric mucosa was gently pipetted
for 10 min at room temperature. Detached cells were collected by
centrifugation at 1000 rpm for 5 min and resuspended in Iscove's
medium with 10% FBS. Cells were cultured on Falcon Primaria tissue culture plates in Iscove's mediurn with 10% FBS, penicillin (100
U/mI), streptomycin (100 j},g/mI), and fungizone (2.5 j},g/ml) at 37°C
and 5% CO2 in a humidified incubator. The plating density for fue
primary culture was < 1000 cells/cm2. Clear individual epithelial colonies were identifiable after approximately 5 days in culture. Each
epithelial colony was separated from fibroblasts by scraping the surrounding ares of the colony under a microscope. After washing off
the scraped cells, colonies were trypsinized and transferred onto
clean plates for the subculture. Using these methods, cultures of
epithelial cells with less than 2% mesenchymal cell contamination
were obtained for study.
Cell characterization. Immunohistochemical staining was performed to confirm the purity of the epithelial cells in culture. Cells
were incubated with monoclonal antibody AEl/AE3, as an epithelial
cell marker. Cells were then incubated with biotinylated goat antimouse antibody and ABC kit, stained with 3,3'-diaminobenzidine
(DAB) and counterstained with hematoxylin. The percentage ofpositively staining cells, denoting epithelial cells, was calculated by the
ratio of cell counts in an average of 500 cells per plate.
Cell culture. Purified epithelial cells were plated at a density of
2 x lO. cells per well on 24-well plates in Iscove's medium with 10%
FBS. Cells were washed twice with serum-free medium (lscove's with
0.1% BSA, penicillin (100 U/mI), streptomycin (100 Jlg/ml), and fungizone (2.5 Jlg/ml) and incubated at 37°C for 12 h in serum-free medium to allow the dissociation of reversibly bound peptides.
[3H]Thymidine incorporation. Cells (2 x lO. cells/well) were incubated in serum-free medium overnight before replacing the medium
with fresh basal serurn-free medium with or without individual factors, which included HGF, EGF, TGF-a, IGF-1, rabbit amniotic fluid,
bovine amniotic fluid, and FBS. [3H]Thymidine (0.5 JlCi) was added
to each well after 24 h of treatmentwith growth factors. Incorporation was stopped after 24 h and cells were washed with PBS twice
before precipitation with 10% trichloracetic acid (TCA). After aspiration ofthe TCA, the precipitated material was solubilized with NaOH
and counted in a scintillation counter (Beckmann Instrurnents Inc.).
Cell growth study. Purified epithelial cells were plated on 24-well
plates (20,000 cells/well) in Iscove's medium with 10% FBS for 24 h.
Plates were washed twice with serurn-free medium and incubated
in serum-free mediurn overnight. The medium was then replaced by
fresh medium with or without individual factors. Cells from each
treatment were collected and counted after 12 days in culture.
Statistics. AlI data are presented as means j: SEM. For each
experiment, tested cells were plated in quadruplicate and these data
were averaged to obtain a single value. Experiments were repeated
with cells derived from separated litters to achieve a minimum of n
= 4. Statistical analysis was performed by ANOV A with P < 0.05
considered significant.
RESULTS
Epithelial GellPurification and ldentification
Fetal gastric epithelial cells were cultured in low
density for 5-7 days before epithelial cell-like colonies appeared. Each epithelial cell-like colony was
identified under microscopy and transferred to a new
dish. Epithelial cells formed monolayers 5-7 days
after the subculture. Figure lA shows that ayer 99%
of cells were stained with keratin antibody, denoting
an epithelial nature. In the long-term culture study,
purified epithelial cells were cultured in the medium
containing either 10% FBS or individual factors. Figure lB shows the anti-keratin antibody stained cells
after 12 days in medium containing 10% FBS. More
iban 80% of cells in this long-term culture were
stained positively for keratin.
Effects of Growth Factors on (3H]Thymidine
1ncorporation
HGF, EGF, TGF-a, and IGF-1 were added to serumfree medium in concentrations from 0.1 pM to 10 nM
and their effects on (3H]thymidine incorporation was
determined after 24 h in culture. HGF (Fig. 2A) increased fetal gastric epithelial cell (3H]thymidine uptake in a dose-dependent pattern. For HGF, a significant increase in (3H]thymidine uptake was first observed at concentrations of 10 pM, with maximal
(3H]thymidine uptake occurring at a concentration of
100 pM. For EGF, a significant increase in (3H]thymidine uptake was first observed at a concentration
ofO.1 pM, with maximal uptake occurring at a concentration of 100 pM (Fig. 2B). For TGF-a-treated cells, a
significant increase in (3H]thymidine uptake was first
observed at a concentration of 0.01 pM, with maximal
uptake occurring at a concentration of 100 pM (Fig.
2C). IGF-1 (Fig. 2D) was less potent compared to the
other growth factors. For IGF-1, a significant increase
in (3H]thymidine uptake was first observed at a concentration of 1 nM. At an IGF-1 concentration of 100 nM,
(3H]thymidine uptake was only approximately 70% of
the maxi!Jlal uptake observed for the other growth factors tested.
For rabbit amniotic fluid, a significant increase in
(3H]thymidine uptake was first observed at a concentration of 2.5% in serum-free medium (Fig. 3). Uptake
was increased over sixfold above basal at a concentration of 20%.
Gell Growth Stlidy
Figure 4 shows the growth curve of fetal gastric epithelial cells after 12 days culture in 10% FBS. The cell
KONG, YEE, AND MULVIHILL: HEPATOCYTEGROWTH FACTOR
163
FIG. l. (A) Purified gastric epithelial cells were grown to confluence.Cells were labeled with anti-keratin antibody at 20.C for 1 h,
washed,processedusing a Vectorstain ABC kit, stained with DAB, and counterstainedwith hematoxylin. The cultures had low rates of
fibroblast contaminatiQn«2%). (B) Purified fetal gastric epithelial cells after 12daysof growth in mediumcontaining 10%FBS and stained
with anti-keratin antibody. Although Bornefibroblastsemerged,more than 80%ofthe cells were conflrmedto be epithelial.
.11
.11
164
JOURNAL OF SURGICAL RESEARCH:VOL. 78, NO. 2, AUGUST 1998
500 lA
~
a..
400
~
.,
300
500
Q
'"c
c
~
~
~
.,
~
~
c
~
~
c
...
~
200
'"
300
200
~
~
:1:
1;:1:
400
~
:r:
~
:r:
100
100
'"
o
o
Basa!
.12
-13
-10
-9
-8
Basal
-7
500
-
400
.Q
...,
~
200
~
-8
-7
..
~
200
]50
~
~
~
:I:
]00
1;-
1;-
~
-9
250
~
...
~
c
a
..10
~
300
..
-12
c
.,
~
!!
.13
300
~
c
..
c
.14
~
100
:I:
..,
M
Basal
.14
.13
12
-11
-10
-9
-8
50
Basal
-7
Concentration
-13
-12
-11
.10
-9
-8
(Iog M)
FIG. 2. (A) The effect of HGF on fetal gastric epithelial cell [3H]thymidine incorporation. Fetal gastric epithelial cells were incubated
in serum-free medium containing HGF at concentrations from 0.1 pM to 100 nM. After 24 h, [3H]thymidine was added to each well and
the cells were incubated at 37°C for an additional 24 h. The maximal [3H]thymidine incorporation was 3.6 times basa! at 100 pM of HGF.
Each data point represents the mean:!: SEM offour or greater separate experiments. *p < 0.05 compared to basa!. (B) The effect ofEGF
on fetal gastric epithelial cell [3H]thymidine incorporation. Maximal [3H]thymidine incorporation was 4.3 times basal at a concentration of
100 pM of EGF. Each data point represents the mean :!: SEM of four or greater separate experiments. *p < 0.05 compared to basa!. (C)
The effect ofTGF-a on fetal gastric epithelial cell [3H]thymidine incorporation. Maximal [3H]thymidine incorporation was 3.6 times basa!
at a concentration of 100 pM of TGF-a. Each data point represents the mean :!: SEM of four or greater separate experiments. *p < 0.05
compared to basal. (D) The effect ofIGF -1 on fetal gastric epithelial cell [3H]thymidine incorporation. Significant stimulation of[3H]thymidine
incorporation first occurred at a concentration of 1 nM. Each data point represents the mean :!: SEM offour or greater separate experiments.
*p < 0.05 compared to basal.
doubling time ranged from 40 to 48 h. Significant cell
growth in long-term culture was observed in response
to EGF and TGF -a, with doubling times approximating
96 h. After 6 days, the response of cells to these individual growth factors reached a plateau. Cells treated with
HGF or rabbit amniotic Huid significantly increased in
number, but did not double during the 12-day treatmento Conversely, no significant increase in cell growth
was observed in response to IGF-1 treatment.
Figure 5 shows cell growth after 10 days of culture
in serum-free medium with individual growth factors
at a concentration of 1 nM. After 10 days in medium
only (control), cell number was approximately 80% of
day O.Compared to control, treatment with HGF, EGF,
and TGF -a, resulted in a significant increase in cell
number. Conversely, treatment with IGF-l resulted in
no increase in cell number. Figure 6 shows cell growth
after 10 days of culture in 20% rabbit amniotic Huid
comparedto serum-free medium alone (control). The
20% rabbit amniotic fluid significantly increased cen
number comparedtoserum-free medium alone.
DISCUSSION
In the present study, HGF was foundto have significant mitogenic effects on fetal gastric epithelial cells
in both short- and long-term culture. In short-term culture, HGF treatment resulted in significant stimulation of [3H]thymidine uptake and its maximal effect
was similar to that of EGF and TGF-a. In short-term
culture, EGF and TGF -a treatment first demonstrated
significant growth at concentrations 100- to 1000-fold
lower than that observed for HGF. In long-term culture, HGF, EGF, and TGF -a treatment resulted in significant stimulation of cell growth. Although IGF -1 significantly stimulated [3H]thymidine incorporation in
KONG, YEE, AND MULVIHILL: HEPATOCYTE GROWTH FACTOR
165
800
la
(/)
la
.c
600
-~
G>
~
la
a.
~
400
200
o
o
1.25
2.5
5
10
20
Rabbit A.F. (%)
FIG. 3. Fetal gastric epithelial cells were incubated in serum-freemedium containing rabbit amniotic fluid at concentrationsranging
from 1.25to 20%. In short-term culture, rabbit amniotic fluid resulted in dose-dependent
increasesin (3H]thymidineuptake. Significantly
increasedincorporationwas presentat concentrationsof 2.5%and greater. *p < 0.05 comparedto 0%amniotic fluido
short-term culture of fetal gastric epithelial cells, its
potency was far less than that observed for HGF, EGF
or TGF -Q. In long-term culture, no significant cell
growth was observed with IGF-l as a single agent.
In our fetal gastric epithelial cell model, we found
t}:lat HGF, EGF, and TGF -a stimulated cell growth in
both short- and long-term culture. The stimulatory effects of HGF on the fetal gastric epithelial cell culture
O'
>la
"tJ
~
--"al
..o
E
~
C
al
U
Days
FIG. 4. Effect of 10%fetal bovine serum on growth of fetal gastric epithelial cells in long-termculture. Medium was changedevery 2
days.The doubling time of the cells was40-48 h. Eachdata point representsthe mean :t SEM of nine separateexperiments.
166
JOURNAL OF SURGICAL RESEARCH:VOL. 78, NO. 2, AUGUST 1998
300
o
>-
la
'C
200
~
~
(/)
c
ou
100
al
U
o
Control
HGF
EGF
TGFa
IGF-1
FIG. 5. The effect of mitogenic peptide factors on growth of fetal gastric epithelial cells in long-term culture. The cell counts after 12
days of culture are shown. Medium was changed every 2 days. Cell counts with medium only (control group) slightly decreased after 10
days in culture. HGF (1 nM). EGF (1 nM), and TGF-a (1 nM) treatment led to significantly increased cell counts compared with control,
whereas fue IGF-1 (1 nM)-treated group showed no significant growth. Each data point represents the mean :!: SEM of four or greater
separate experiments. *p < 0.05 compared with control.
may not be limited to gastric cellular growth. HGF may
be essential for many other aspectsof fetal gastrointestinal developmentand differentiation [9-11] since it
has been localized in many organs, including the gastrointestinal epithelium [4, 11], and has been found to
stimulate a variety of epithelial cell types, including
gastric epithelial cells from adult rabbits and rats [5,
6]0 We algo found that EGF and TGF-a had similar
stimulatory effects on fetal gastric epithelial cellso Although this finding is similar to one on canine gastric
epithelial cells [13], EGF has been found to be more
potent than TGF -a in rabbit gastric smooth muscle cell
culture [14], while TGF-a had been found to be a more
potent mitogen than EGF in pig gastric epithelial cells
250
O'
>la
200
~
150
"ti
In
C
j
o
()
100
Q)
()
50
o
Control
20% R-AF
FIG. 6. Effect of rabbit amnioticfluid on growth of fetal gastric epithelial cells in long-termculture. Cell countsafter 10days of culture
in 20% rabbit amniotic fluid or serum-freemedium (control)are shown. Medium was changedevery2 days. Each data point represents
the mean :t SEM offour or greaterseparateexperiments.*P < 0.05comparedwith control.
KONG, YEE, AND MULVIHILL: HEPATOCYTEGROWTH FACTOR
167
[15]. An important difference in OUTmodel, possibly epithelial cell growth because it is known that the fetus
explaining this discrepancy, is that we used fetal, not swallows significant amounts of amniotic fluid, particuadult, gastric epithelial cells.
larly in the third trimester of gestation.
Comparedto HGF, EGF, andTGF-a, treatmentwith
The present study, using highly purified epithelial
IGF-1 resulted in only modest stimulation offetal gas- cells, demonstrates that HGF, EGF, and TGF -o:as inditric epithelial cell growth. It is known that IGF-1 recep- vidual factors stimulate both DNA synthesis and fetal
tors are present in adult rabbit gastrointestinal tissue gastric epithelial cell growth in both short- and long[23]. However, the concentration ofIGF-1 required for term cultures. These three peptide growth factor are
significant growth has been reported to be 100 times potential mediators ofthe growth promoting properties
higher than that for EGF [13]. An extremely high con- of amniotic fluid on fetal gastric epithelial cells. OUT
centration ofIGF-1 was required to increase growth of results suggest that: (1) HGF stimulates [3H]thymidine
gastrointestinal tissue in young rats [18]. The insensi- uptake and cell proliferation in fetal rabbit gastric epitivity of cells to IGF -1 in OUTstudy may be explained thelial cells in vitro, and (2) HGF's mitogenic effect on
by (1) down-regulation of IGF -1 receptors by the pres- fetal rabbit gastric epithelial cell growth is comparable
ence of autocrine IGF-1, (2) reduced expression ofIGFto that observed for EGF and TGF-o:, but superior to
1 receptors in the fetal gastric epithelial cells, or (3) the effect observed for IGF-1.
indirect stimulation of growth by IGF-1 through secondary mediators. This is a present focus of study in
A CKN O WLED GMENTS
OUTlaboratory.
In contrast to most studies using only short-term
We thank Dr. Ralph Schwall of Genentech,Inc. for bis kind gift
cultures, we tested the effects of potential mitogens in of recombinanthuman hepatocyte growth factor (rhHGF, Lot No.
18181-01),Edna Q. Calaustro for expert technical assistance,and
both short- and long-term culture. Although different
Leveau for preparation of the manuscript. The vertebrate
methods of isolating gastric epithelial cells have been Eleanor
animal studies in this study have beenapprovedby the University
reported for the adult rabbit [24], adult human [25], ofCalifornia, SanFranciscoInstitutional Animal Careand UseComand fetal rabbit [26], most investigators have used mittee. This work was supported by National Institute of Health
Grant DK-24773.
methods of scraping cells or mincing tissues to initiate
the isolation of gastric cells. This has the potential
problem ofincluding significant numbers offibroblasts
REFEREN CFS
in the culture, especially problematic in long-term cull. Michalopoulos,G., Houck, K. A., Dolan, M. L., and Luetteke,
tures. To minimize this problem, we modified preN. C. Control of hepatocytereplication by two serum factors.
viously described methods for the isolation offetal rabCancerRes.44: 4414-4419, 1984.
bit gastric epithelial cells [12]. OUTcells were derived
2.
Nakamura,
T., Nawa, K., and Ichihara, A. Partial purification
from third-trimester fetal rabbit stomachs and mainand characterizationofhepatocytegrowth factor from serum of
tained in both short-term (24 h) and long-term (12
hepatectomizedrats. Biochem. Biophys. Res. Commun. 122:
days) culture. The methods we utilized included use of
183-192, 1984.
the everted fetal stomach and washing offcells in Ca2+ 3. Matsumoto,K., and Nakamura, T. Emerging multipotent aspects of hepatocytegrowth factor. J. Biochem.119: 591-600,
and M~+-free Hank's buffer with gentle pipetting.
1996.
After 5-7 days, each epithelial cell-like colony was
4. Wolf,H. K., Zarnegar,R.,and Michalopoulos,G. K. Localization
identified under microscopy and transferred to a new
ofhepatocytegrowth factor in humanand rat tissues:an immudish. Epithelial cells formed monolayers approximately
nohistochemicalstudy. Hepatology14: 488-494, 1991.
5- 7 days after the subculture. Using these methods,
5. Takahashi, M., Ota, S., Terano, A., et al. Hepatocyte growth
nearly pure epithelial cells were isolated. In OUTnovel
factorinducesmitogenicreactionto the rabbit gastric epithelial
culture system, epithelial cells grown in 10% FBS for
cells in primary culture. Biochem.Biophys.Res.Commun.191:
528-534, 1993.
12 days were contaminated with less than 20% fibro-
blasts.
Amniotic fluid has mitogenic effects on fetal rabbit
gastric epithelial cells [27], and these effects may be
due to the presence of growth factors in the amniotic
fluid [28]. Amniotic fluid is known to contain mitogenic
peptide growth factors, including EGF, TGF-a, IGF-1,
and HGF [19-22]. Some inhibitory factors have algo
been identified in amniotic fluid, including a factor
which has a {3-chain is homologous to that ofTGF-{3 in
bovine amniotic fluid [29]. IGF -binding protein-1, a cell
growth inhibitor [30], has been identified in human
and avine amniotic fluid [31, 32]. In the present study,
rabbit amniotic fluid had significant mitogenic effects,
including stimulation of [3H]thymidine uptake and cell
growth. These results suggest that growth factors present in amniotic fluid may act to regulate fetal gastric
6. Fukamachi, H., Ichinose, M., Tsukada, S., et al. Hepatocyte
growth factor specificallystimulatesgastro-intestinalepithelial
growth in primary culture. Biochem.Biophys.Res. Commun.
205: 1445-1451,1994.
7. Watanabe,S., Hirose, M., Wang, X., et al. Hepatocytegrowth
factor acceleratesthe wound repair of cultured gastric mucosal
cells. Biochem.Biophys. Res.Commun.19~,,1453-1460,1994.
8. Tsuji, S., Kawano, S., Tsujii, M., Fusamoto,H., and Kamada,
T. Rolesof hepatocytegrowth factor and its receptorin gastric
mucosa.Dig. Dis. Sci. 40: 1132-1139, 1995.
9. Warn, R. The scatteredjigsaw. Nature 376: 723,1995.
10. Uehara, Y., Minowa, O., Mori, C., et al. Placental defect and
embryonic lethality in mice lacking hepatocytegrowth factor/
scatterfactor. Nature 373: 702-705, 1995.
11. Kermorgant, S., Walker, F., Hormi, K., Dessirier, Y., Lewin,
M. J., and Lehy,T. Developmentalexpressionand functionality
ofhepatocytegrowth factor and c-Met in human fetal digestive
tissues. Gastroenterology
112: 1635-1647, 1997.
168
JOURNAL OF SURGICAL RESEARCH:VOL. 78, NO. 2, AUGUST 1998
12. Nakajima, N., and Kuwayama, H. Effects of transforming
growth factor a and fJ on rabbit gastric epithelial cell proliferation: a preliminary reportoJ. Clin. Gastroenterol.
17: 8121-124,
1993.
13. Chen,M. C., Lee, A. T., and 8011,A. H. Mitogenic responseof
canine fundic epithelial cells in short-term culture to transforming growth factor alpha and insulinlike growth factor l. J.
Clin. Invest. 87: 1716-1723, 1991.
14. Yuan, Q. X., McRoberts,J. A., Lakshmanan,J., Yagi, H., and
Hyman, P. E. Newborn rabbit gastric smoothmuscle cell culture: EGF and TGF-a are potent mitogens.J. Pediatr. Gastroenterol.Nutr. 17: 153-160, 1993.
15. Rutten, M. J., Dempsey,P. J., 8010mon,
T. E., and Coffey,R. J.,
Jr. TGF-a is a potent mitogen for primary cultures of guinea
pig gastric mucousepithelial cells.Am. J. Physiol. 265: G361369,1993.
16. Polk Jr., W. H., Dempsey,P. J., Russell,W. E., et al. lncreased
productionof transforming growth factora following acute gastric injury. Gastroenterology
102: 1467-1474,1992.
17. Dembinski,A. B., and Johnson,L. R. Effectofepidermalgrowth
factor on the developmentof rat gastric mucosa.Endocrinology
116: 90-94, 1985.
18. 8teeb,C. B., Trahair, J. F., Tomas,F. M., and Read,L. C. Prolonged administration of IGF peptidesenhancesgrowth of gastrointestinal tissues in normal rats. Am. J. Physiol. 266:
G1090-G1098,1994.
19. D'80uza,8. W., Haigh, R., Micklewright, L., Donnai, P., and
Keys,A. Amniotic fluid epidermalgrowth factor and placental
weight at term [letter]. Lancet2: 272-273, 1985.
20. Goodyer,P. R., Mulligan, L., and Goodyer,C. G. Expressionof
growthrelated genesin human fetal kidney.Am. J. KidneyDis.
17: 608-610, 1991.
21. Merimee,T. J., Grant, M., and Tyson,J. E. Insulin-like growth
factors in amniotic fluido J. Clin. Endocrinol. Metab. 59: 752755, 1984.
22. Rosen,E. M., Meromsky, L., Romero,R., Setter, E., and Goldberg, l. Human placentacontainsan epithelial scatter protein.
Biochem.Biophys. Res.Commun.168: 1082-1088, 1990.
23. Termanini, B., Nardi, R. V., Finan,T. M., Parikh, l., and Korman, L. Y. Insulin-like growth factor 1 receptorsin the rabbit
gastrointestinal tract. Gastroenterology
99: 51-60, 1990.
24. Ota, S., Terano, A., Hiraishi, H., et al. A monolayerculture of
gastric mucouscells from adult rabbits. Gastroenterol.Jpn. 25:
1-7,1990.
25. Terano, A., Mach, T., Stachura,J., Sekhon,S., Tarnawski, A.,
and Ivey, K. J. A monolayerculture ofhuman gastric epithelial
cells. Dig. Dis. Sci. 28: 595-603, 1983.
26. Logsdon,C. D., Bisbee,C. A., Rutten,M. J., and Machen,T. E.
Fetal rabbit gastric epithelial cellscultured onfloating collagen
gels. In Vitro 18: 233-242, 1982.
27. Mulvihill, S. J., Stone,M. M., Fonkalsrud,E. W., and Debas,
H. T. Trophic effect of amniotic fluid on fetal gastrointestinal
development.J. BurgoRes.40: 291-296,1986.
28. Mulvihill, S. J., Stone,M. M., Debas,H. T., and Fonkalsrud,
E. W. The role of amniotic fluid in fetal nutrition. J. Pediatr.
Burgo20: 668-672, 1985.
29. Chertov,O. Y., Krasnosel'skii,A. L., Bogdanov,M. E., and Hoperskaya,O. A. Mesoderm-inducingfactor from bovineamniotic
fluid: Purification and N-terminalaminoacidsequencedetermination. Biomed. Sci. 1: 499-506, 1990.
30. Liu, L., Brinkrnan, A., Blat, C., and Harel, L. IGFBP-l, an
insulin like growth factorbinding protein,is a cell growth inhibitor. Biochem.Biophys.Res.Commun.174: 673-679, 1991.
31. Drop,S. L., Kortleve,D. J., and Guyda,H. J. Isolationora somatomedin-binding protein from pretermamniotic fluidoDevelopment of a radioimmunoassay.J. Clin. Endocrinol. Metab. 59:
899-907, 1984.
32. Wang,J. F., Fraher, L. J., and Hill, D. J. Characterization of
insulin-like growth factor-binding protein in avine amniotic
fluido J. Endocrinol. 127: 325-333, 1990.