Fibroblast growth factors modulate intestinal epithelial cell growth and migration

GASTROENTEROLOGY 1994;106:l2644262

Fibroblast Growth Factors Modulate Intestinal Growth and Migration AXEL U. DIGNASS,

SHOJI TSUNEKAWA,

and DANIEL

Epithelial Cell

K. PODOLSKY

Gastrointestinal Unit, Department of Medicine, and Center for the Study of Inflammatory Bowel Disease, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts

Background/Aims: Various peptide growth factors have been found to exert functional effects among epithelial cell populations. This study assessed the role of certain fibroblast growth factors (FGFs) (acidic FGF, basic FGF, and keratinocyte growth factor) in the regulation of intestinal epithelial cell proliferation and restitution. Methods: Recombinant growth factors were added to subconfluent cultures of IEC-6, Cace2, and HT-29 cell lines with subsequent assessment of [3H]thymidine incorporation. The effects on an in vitro model of restitution were assessed by quantitation of cells migrating into standard wounds established in confluent monolayers of IEC-6 cells. Transforming growth factor p (TGF-P) content of growth factortreated wounded monolayers was assessed by Notthern blot and bioassay. Results: Acidic FGF, basic FGF, and keratinocyte growth factor caused a modest increase in proliferation of IEC-6, Cace2, and HT-29 cell lines. Acidic FGF and basic FGF promoted intestinal epithelial cell restitution in vitro up to XI-fold, in conjunction with the enhanced expression of TGF-/3 messenger RNA and protein. Promotion of IEC-6 restitution by acidic and basic FGF could be blocked by addition of immunoneutralizing anti-TGF-P antisera. Conclusions: FGFs that exert effects on fibroblast cells also exert effects on intestinal epithelial cell populations and may help promote epithelial cell restitution, the initial step of intestinal wound healing through a TGF-P-dependent pathway.

epithelial

cell responses

and integrity Studies

over the past several

important

he epithelium of alimentary tract mucosa is composed of a highly dynamic cell population in which nearly complete turnover may occur every 24-96 hours in the small and large intestinal mucosa among different mammalian species.’ A number of observations suggest that a spectrum of structurally distinct regulatory peptides present within the mucosa of the gastrointestinal tract may collectively form a functionally integrated net-

work. The constituents of this network generally have multiple functional properties and show pleiotropism in their cellular sources and targets. Peptide growth factors may play an essential role in coordinating differential

normal

homeostasis

role of members

years have shown

of the transforming

the

growth

growth factor factor (TGF) a, TGF-P, and insulinlike families in the modulation of epithelial cell proliferation and differentiation.2-9 However, the effects of fibroblast growth factor (FGF) family peptides in the gastrointestinal tract have not been well characterized. peptides

FGF family

have been shown to act as mitogens

for mesoder-

mal-derived cell populations. In addition, they have neurotrophic and angiogenic properties and are among the most potent factors inducing new through both stimulation of endothelial tissue reorganization

including

lial cell migration, minogen

activator

production

basic

reports has been

and plas-

degrading

proteases),

matrix

constituents.‘”

examining

in gastrointestinal

FGF

of endothe-

of collagenase

of new extracellular

Although peptides

(two matrix

vessel formation proliferation and

stimulation

production

and

the role of FGF family

tract tissues remain

detected

in the colon

limited,

and

small

intestine as well as in various colon cancer cell lines.11-‘3 Basic FGF was found within the extracellular matrix in close apposition to epithelial cells as well as blood vessels. The distribution of acidic FGF was originally thought to be limited to neural tissue such as brain and retina, but recently acidic FGF has been found in other tissues including kidney, myocardium, and bone and in various cell types. l4 Keratinocyte

T

to preserve

of the mucosa.

growth

factor (KGF),

another

member of the FGF ligand family now recognized to encompass at least nine discrete peptides, is produced by stroma fibroblasts in many tissues. KGF is presumed to act largely or exclusively as a specific growth factor for epithelial cell populations.15 Little is known about the Abbreviations used in this paper: DMEM, Dulbecco’s modified Eagle medium; EGF, epidermal growth factor; FCS, fetal calf serum; FGF, flbroblast growth factor; GAPDH, glyceraldehydti-phosphate dehydrogenase; IFN?, interferon y; Il-lp, interleukin lp; KGF, keratinocyte growth factor: MvlLu, mink lung epithelial cells; SSC. standard saline citrate; TGF, transforming growth factor. 0 1994 by the American Gastroenterological Association 00165065/94/$3.00

MODULATION OF EPITHELIAL CELL FUNCTION BY FGF

May 1994

distribution

and biological

activities

recognized members of the FGF oncogene,

hst/Ks oncogene,

the gastrointestinal The potent proliferation

of more recently

peptide

FGF-5,

family

and FGF-6)

Effect of FGFs on Cell Proliferation

(int-2 within

tract.

Cells (IEC-6, HT-29,

as well as modulation

of fibroblast

and

function

Eagle

medium

approximately

washed

and then

three

times

after injury in general are also assumed to be important in repair after mucosal

supplemented

Recent

studies indicate

or surgical

that luminal

FGF may facilitate

repair of peptic ulcers in rodent models.” tic basis of this phenomenon fined.

Beneficial

on vascular present

elements

Paimela

of FGF propria

effects on epithelial et al. ” showed

be enhanced

may reflect

mesenchymal

or submucosa,

that epithelial

injury,

its effects cells

or unap-

cell populations.

by basic FGF

vitro model of gastric

The mechanis-

has not yet been well de-

or underlying

in the lamina

preciated could

effects

resection.

Recently,

cell restitution

in an amphibian

but the mechanistic

trations

both

recent

studies,

basic FGF and the FGF receptor

several colon cancer-derived ulated cell proliferation autocrine

mechanism

in the regulation little

is known

transformed cancerous

in these cell lines, indicating

of epithelial

about

intestinal

in general.

epithelial family

In the present on intestinal

in an

may be involved

cell growth.

However, on non-

cells derived from non-

the effects of other ligands on intestinal

epithelial

study, we explored

of various FGF family peptides and KGF)

FGFs

the effects of basic FGF

tissues and about

of the FGF peptide

are expressed

cell lines.‘* Basic FGF stim-

by which

epithelial

cells

the effects

(acidic FGF,

basic FGF,

cell function.

In partic-

ular, the effects of acidic FGF and basic FGF on epithelial cell restitution, tegrity

a process that re-establishes

after superficial

mucosal in-

injury by rapid migration

of epi-

thelial cells across the wound margin, were assessed using a previously

described

from

0.1%

cells were

for an additional

FCS. Cultures

1 to 1000 pmol/L

of heparin (125 pg/mL).

24

were then

for acidic FGF and

After 20 hours at 37”, [‘H)thymidine

90 pCi/mL)

tion of radiolabeled

was added; after 4 hours, incorpora-

thymidine

Briefly,

previously.”

was determined

cells were washed

material

radioactivity

phosphate-buf-

acid (3:1, vol/vol).

was then lysed with

was counted

as described

with

fered saline and fixed in methanol-acetic Acid-insoluble

using a liquid

1N NaOH,

scintillation

and

counter.

Wounding Assays

action that

5% fetal calf

with acidic FGF, basic FGF, or KGF in concen-

ranging

(1.8 PCilwell;

in

observed

cultured

containing

Wounding

investigators

containing

50% confluent,

basic FGF and from 0.1 to 5 .O nmol/L for KGF in the presence

of basic FGF was not further defined. In other

in the presence of Dulbecco’s

(DMEM)

serum (FCS). When hours in DMEM

ulceration

and Caco-2) were seeded into 24-

well plates (1 - 5 X 1O* cells/well) modified

effects of FGFs on neovascularization

1255

viously

assays were essentially

described’8.21

using

a modified

Sato and Rifkin.‘” Confluent

monolayers

mm plastic dishes were wounded two wounds

approximately

by about

fresh serum-deprived

medium,

were further medium

cultured

FGF (1 - 1000 pmol/L).

by

of IEC-6 cells in 60-

mm across the dish were

1 cm. Cells were washed and the wounded

for 24 hours

in the presence

as pre-

described

with a razor blade; generally,

20-25

made and separated

performed

method

in fresh serum-deprived

or absence Migration

with

monolayers

of acidic FGF and basic

of IEC-6 cells was assessed

in a blinded fashion to avoid observer bias by determination of the number of IEC-6 cells observed across the wound border expressed

as the mean number

border in a standardized dardized

by taking

of cells found across the wound

wound area. Wound

photomicrographs

tion using an inverted

microscope

areas were stan-

at lOO-fold magnificaNikon

a Nikon N6006 camera. Experiments

Diaphor

TMS with

were performed

in tripli-

cates, and several wound areas per plate were used to quantitate migration.

Data presented

under results are expressed

value + SD of at least three independent

as mean

experiments.

in vitro model.‘83’9

Determination of TGF-Pl Messenger RNA

Materials and Methods

Total layers (intact

Materials

Messenger

Radiochemicals were obtained from New England Nu-

cellulose

cellular

RNA (mRNA) column

clear, Boston, MA. Human recombinant basic and acidic FGF

1 .O% formaldehyde

and porcine platelet TGF-P1

membranes

were obtained from R&D Sys-

tems, Minneapolis, MN; turkey anti-human TGF-P

was ob-

tained from Collaborative Research, Bedford, MA; and recombinant human KGF was kindly provided by Drs. S. Aaronson

was isolated by the method (4 pg/lane)

chromatography

and human

(GAPDH)

from IEC-6

mono-

of Chirgwin

et al.‘*

obtained

by oligo(dT)-

was electrophoresed

agarose gel and blotted

(MSI, Westboro,

man TGF-Pl drogenase

RNA

or wounded)

in a

onto nylon transfer

MA) by standard

methods.

Hu-

glyceraldehyde-3-phosphate-dehy-

probes were prepared

by random priming

and J. S. Rubin (Laboratory of Cellular and Molecular Biology,

of a 2.14-kilobase (kb) EcoRI insert of a human TGF-fll plementary DNA (cDNA) clone designated pHTGF$2

National Institutes of Health, Bethesda, MD). IEC-6 cells first

of a 0.78-kb

established in this laboratory by Dr. A. Quaroni were used at

designated

15-1&h obtained

Type Culture Collection. Hybridization was performed at 42°C in 50% formamide, 5 X standard saline citrate (SSC), 5 X Denhardt’s, and 10% dextran sulfate; blots were successively

MD.

passage.*’ MvlLu, Caco-2, and HT-29 from American Type Culture Collection,

cells were Rockville,

Pst I, Xba I insert of a human pHcGAP,

respectively,

obtained

GAPDH from

comand

plasmid American

1256

GASTROENTEROLOGY Vol. 106. No. 5

DIGNASS ET AL.

washed at 50°C in 2X SSC and 0.1% sodium three times. Blots were first hybridized and exposed for 16-48 then hybridized

as previously

TGF-Pl

transcript

probe

The same blots were

probe

to standardize

described.23*24 Relative

was assessed

ning, normalized

sulfate

hours at -80°C.

with GAPDH

loading

dodecyl

with the TGF-Pl

mRNA

abundance

by laser densitometric

to the density

of GAPDH

of

scan-

transcript.

200 PMthymidine uptake [% control]

100

Determination of Latent and Bioactive TGF-P Latent and bioactive TGF-P were determined assay using MvlLu

150

by a bio-

lung epithelial cells) as described

(mink

1

0

10

elsewhere.25 Briefly, samples were activated by acidification with 150 mmol/L HCI for 60 minutes tion before assay for TGF-P cells were used to initiate plates.

standardized

FCS-DMEM

quantities

Subconfluent

of porcine platelet (1.5 pCi/well)

was continued

methanol-acetic

for an additional

scintillation

and

(R&D Sys-

for 20 hours at

was added, and the incu-

acid (3:1, vol/vol).

using a liquid

test samples

TGF-Pl

4-hour

cells were washed with phosphate-buffered was then lysed with 1N NaOH,

MvlLu

assay. Then

per well were seeded

and cells were incubated

37°C. C3H]Thymidine

IPM]

by neutraliza-

inhibition

Five hours after seeding,

tems) were added bation

bioactivity. the growth

1 X lo5 cells in 0.2% on 24-well

followed

bFGF

period.

PHIthymidine uptake [% control]

The

200

150

100

saline and fixed in

Acid-insoluble

and radioactivity

material

was counted

0

counter.

0

10

1

100

1000

aFGF [PM]

Statistical Analysis Results

are expressed

nificance of differences the student’s significance

250

as mean

between

? SD. Statistical

I test for paired and unpaired levels represent

sig-

mean values was assessed with

two-tailed

data. All reported

P values.

thyzzne 150 p

Results Although tocrine

effects

uptake control]

100

previous studies identified potential of basic FGF

on human

colon

au-

cancer-

derived cell lines, the effects of basic FGF on nontransformed intestinal

epithelial

cells and the effects of other

ligands of the FGF peptide nant intestinal been defined. on intestinal

and/or colonic

epithelial

cells have not

To test the effect of FGF family peptides epithelial

cell proliferation,

three ligands of

the family, basic FGF, acidic FGF, and KGF, to subconfluent epithelial

cultures of nontransformed

cells (IEC-6)

cell lines (Caco-2

and HT-29).

cell proliferation

in the intestinal

IEC-6,

Caco-2,

can be stimulated

and HT-29

by basic FGF, acidic FGF, this laboratory thymidine

rat intestinal

As shown in Figure epithelial

cell

and KGF.

directly

cell lines used in this study.

HT-29. Intestinal epithelial cells were cultured in the presence of various concentrations of FGFs, and [3H]thymidine uptake was assessed as described in Materials and Methods. Data represent mean values ? SEM of at least three independent experiments.

1,

manner

Previous studies in

that the uptake of E3H]with proliferation

Figure 1. Effect of (A) basic FGF (bFGF), (13) acidic FGF (aFGF), and ine uptake in intestinal epithelial cell lines. ?? ,

lines

as assessed by E3H)thymidine in a dose-dependent

have confirmed

correlates

were added

and human colon cancer-derived

uptake

KGF [nM]

family on normal or malig-

in the

One of the hallmarks

that distinguishes

FGFs

from

many other growth factor families is their relatively high affinity for heparin. bioactivity mitogenic a lOO-fold,

of FGFs. activity

Heparin

is known

In certain

to modulate

fibroblast

of acidic FGF can be potentiated

thus rendering

it as potent

the

cell lines, the by

as basic FGF.26

May

1994

MODULATION OF EPITHELIAL CELL FUNCTION BY FGF

Two mechanisms

of FGF-heparin

interactions

are pre2.5

sumed to be important. First, FGFs may be bound onto the cell surface by interacting with cellular bound heparan

sulfate

and therefore

be locally

target cell. Second, heparin-FGF FGFs from the proteolytic tissue remodeling, thus allowing

inflammation,

that characterizes

and neovascularization,

them to function

in a metabolically

and basic FGF were evaluated

concentrations

from

of 12 5 PglmL

heparin

trations

degradation

for the

may protect

hostile

1 to 1000 pmol/L

1.5-

GAPDH mRNA

,_

dine could be stimulated pared with untreated and growth

over a range of in the presence

Incorporation

2 -5-fold

controls

(mean, 2.5-fold)

depending

that intestinal

epithelial

com-

on the cell line cell prolifera-

tion was enhanced

by FGFs, the effects of these peptides

on other epithelial

cell features were assessed. Studies

this laboratory

and elsewhere

p may be a powerful

growth

in

have suggested

that TGF-

factor regulating

intestinal

cell populations.2-4X8

to be expressed

by primary

strong inhibition

of proliferation

TGF-Pl enterocytes

has been found and

6

12 time [hours]

24

46

of thymi-

factor used.

After finding

0

and KGF over a range of concen-

from 0.1 to 5 .O nmol/L.

epithelial

relative

yyg..

14,26

environment. Acidic

available

interactions

1257

to effect

in the IEC-6 cell line.394

flgure 2. Effect of acidic FGF (m) and basic FGF (A) on the expression of TGF-Pl mRNA in IEC-6 cells. Subconfluent layers of IEC-6 cells maintained for 24 hours in DMEM containing 0.1% FCS were cultured for various periods in DMEM containing 0.1% FCS and 100 pmol/L acidic or basic FGF. Northern blots were prepared as detailed in Materials and Methods. The relative expression of TGF-Pl standardized to GAPDH was quantitated by laser densitometry. Data represent mean values 2 SEM of four independent experiments.

this model, IEC-6

a contact-inhibited

cells is wounded

razor blade,

and migration

reconstitute

the continuity

confluent

in a standard into

monolayer fashion

the wound

of

with

a

defect

to

Recent studies in this laboratory have provided evidence of interrelated autocrine and paracrine expression of TGF-Pl and TGF-CY in these cells.8 We speculated that

mined. FGFs are known in general to play an important role in facilitating repair after injury, particularly after

a similar

mucosal

interrelationship

exists between

FGFs and TGF-

ulceration

of the monolayer

or surgical

resection

is deter-

of intestinal

tis-

p. The IEC-6 cell line was chosen because extensive studies have shown the similarity between this untransformed

sues. However,

small intestinal crypt intestinal

FGFs that have been shown to be present at the basolatera1 surface of the intestinal epithelium in vivo promote intestinal wound healing through enhanced intestinal

differences

between

counterpart capacity

rat epithelial cell line and the normal rat epithelial ce11.2,4,8,20Although important IEC-6

and

have to be assumed

of terminal

the primary

epithelial

in view of the limited

differentiation

into phenotypically

mature villus enterocytes, vides a model intestinal

this cell line nonetheless proepithelial cell line that offers

the opportunity

epithelial

the mechanistic

has not yet been

epithelial

basis of this phenomenon

well defined.

cell restitution,

acidic

To determine

whether

and basic FGFs

were

added to wounded IEC-6 cell monolayers. As shown in Figures 3 and 4, addition of acidic and basic FGF to

As shown in Figure 2, culture of subconfluent monolayers of IEC-6 cells in the presence of acidic and basic

serum-deprived medium enhanced the migration of IEC6 cells into the denuded area of a model wound in a dosedependent manner compared with migration of IEC-6 cells cultured in serum-deprived medium alone. Migra-

FGF led to a moderate but steady-state level of TGF-Pl sion of TGF-Pl was apparent

significant increase in the mRNA. Increased expreswithin 6 hours after addi-

tion of cells into the wound area could be observed as early as 5 hours after producing the wound in the presence of FGFs. To obtain reliable numbers of cells across

tion of FGFs and persisted as long as 24 hours. Enhancement of TGF-Pl expression ranged from 2-fold to 4-fold and averaged 2.5-fold in four separate experiments when mRNA levels were standardized relative to the constitu-

the wound edge, standardized incubation times of 24 hours were used for all subsequent experiments. All additional studies were performed with concentrations of acidic and basic FGF that caused maximal stimulation of epithelial cell migration in dose-response studies. As shown in Figure 4, significnnt enhancement of epithelial restitution was observed at concentrations of 10 pmol/L; also, migration of IEC-6 cells was significantly (P < 0.001) enhanced for acidic FGF at a concentration of

tive marker

to study

transcript

cell responses.

for GAPDH.

Previous studies in this laboratory have shown that several cytokines and conventional growth factors present at the basolateral surface of the epithelium in vivo ptomote epithelial cell restitution in an in vitro model.*l In

1258

GASTROENTEROLOGY Vol. 106, No. 5

DIGNASS ET AL.

Figure 3. Cell migration in an in vitro model of epithelial restitution. Standard wounds were made with a razor blade in confluent monolayers of IEC-6 cells as previously described.l’ After washing with fresh medium, wounded monolayers were cultured for 24 hours in (A) control media or media supplemented with (13)100 pmol/L acidic FGF or (C) 100 pmol/L basic FGF. Cells were fixed in 100% methanol, and photomicrographs were taken (original magnification x100). The scratch indicates the original margin of the wound.

100 pmol/L concentration stimulation concentration

(average, sixfold) and for basic FGF at a of 100 pmol/L (average, sixfold). Further of epithelial

restitution

was observed

at a

of 1000 pmol/L.

through

a TGF-P-dependent

addition

found to stimulate

intestinal

To assess whether lial restitution

Previous studies” had shown that promotion of intestinal epithelial cell restitution in vitro by various cytokines was mediated

way. Furthermore,

path-

dent

pathway,

FGF-induced

is also mediated the expression

of TGF-P epithelial

itself

enhancement through

has been

cell migration.‘* of epithe-

a TGF-P-depen-

of TGF-Pl

mRNA

was

evaluated in wounded and unwounded monolayers. As shown in Figure 5, wounding of confluent monolayers

B

1

Figure 4. Effects of (A) basic and (13)acidic FGF on epithelial restitution in wounded lEC-6 monolayers. Wounds were established in confluent monolayers of IEC-6 cells as detailed in Materials and Methods, and wounded monolayers were cultured for 24 hours after addition of control media or media containing acidic or basic FGF (1,10,100, or 1000 pmol/L). Cells migrating across the wound margin (Figure 2) were quantitated by blinded determination as detailed in Materials and Methods. Results represent mean value ? SD of at least three separate experiments.

May 1994

MODULATION OF EPITHELIAL CELL NNCTION

led to significantly TGF-Pl

mRNA

observed

increased compared

previously.

steady-state

expression

with unwounded

lEf2’ Furthermore,

of as

controls,

the absolute

in-

mL) added

by enhanced

slightly,

TGF-Pl

significantly,

enhanced

compared

monolayers cultured in serum-deprived To assess further the role of TGF-P FGF-enhanced

epithelial

monolayers, culture

the concentration

media

of IEC-6

6, the concentration increased pmol/L)

6 cells cultured

peptide

obtained

was

from wounded

IEC-6

cells

in serum-deprived

and basic FGF (100

media of wounded medium

as previously

The intestinal

TGF-P

with control

duced TGF-P

IEC-

alone (P <

0.01). The functional importance of increased levels of bioactive TGF-P peptide in wounded IEC-6 monolayers after

dynamic

epithelial

ferentiation maintain

of the extracellular make

TGF-bl. Previous

anti-

though

many

homeostasis.

studies

important

in which proliferating from a gradient

of in-

mechanisms

exist to bal-

with both commitment

elements regulating intestinal liferation, and differentiation

monolayers

of immunoneutralizing

complex

and loss of mature mucosal

of a highly

cells present along the longituor colonic mucosal surface. It is

activity

treatment with acidic and basic FGF was evaluated by determination of the migration of IEC-6 cells in wounded in the presence

is composed

cell population

that highly

ance proliferative

pro-

observed.”

cells in the crypt are segregated

assumed

of anti-

of the endogenously

mucosa

creasingly differentiated dinal axis of the villus

The reduction

in the presence

reflects neutralization

in the

incubation

of establishing

TGF-fil.

rate of migration

Discussion

and bioactive

of acidic

levels of bioactive

in the baseline

IEC-6

was assessed. As shown in Figure

in the presence compared

in wounded

of TGF-P

of latent

in media

medium alone. as a mediator of

cells after a 24-hour

period after wounding

cultured

restitution

with

at the time

1259

the wound. This observation supports the conclusion that the observed increase in IEC-6 cell migration was caused

crease of TGF-Pl mRNA expression in wounded monolayers in the presence of acidic and basic FGF was but

to the medium

BY FGF

to dif-

cells from the villus The understanding

epithelial is still

have suggested

of the

cell growth, incomplete. that

to

proAl-

constituents

matrix

and cell to cell interactions

contributions

to these processes,

it is

to the

clear that many peptide growth factors are also essential. Over the past several years, studies of both primary intestinal epithelial cells and established cell lines derived

culture medium, as assessed by 13Hlthymidine uptake assays and wounding assays. As shown in Figure 7, the

from the intestinal epithelium and human colon carcinomas have suggested that soluble factors may be essential

stimulation

in regulating

studies had shown the ability

to immunoneutralize

exogenous

of IEC-6 restitution

could be blocked

completely

of this antibody

TGF-Pl

added

by acidic and basic FGF

by anti-TGF-pl

(27.5 pg/

nal epithelial

proliferation

and differentiation in intesticells.3*4*8S27-i1These regulatory peptides en-

compass structurally as peptide

growth

colony-stimulating

+

??

Although

diverse peptides

identified

factors, interleukins,

variously

interferons,

and

factors.

a number

of reports

have shown

an im-

1.5 relative expression of TGFPl GAPDH mRNA

,

’

DMEM

bFGF

aFGF +bFGF

Rgure 5. Effect of acidic and basic FGF on the expression of TGF$l mRNA in unwounded (0) and wounded (m) IEC-6 monolayers. Total RNA was prepared from confluent monolayers either intact or 24 hours after wounding. Monolayers were cultured in the presence or absence of acidic and basic FGF (100 pmol/L). The relative expression of TGFpl standardized to GAPDH was quantitated by laser densitometry. Columns represent mean values 2 SEM of four Northern blots obtained in independent experiments and used for densitometric analysis. *P < 0.03 for wounded vs. unwounded; ‘P < 0.03 vs. wounded control (DMEM).

+aFGF

flgure 6. Effect of acidic and basic FGF and wounding on production of latent and bioactive TGF-P by IEC-6 monolayers. Confluent mono layers of IEC-6 cells maintained for 24 hours in DMEM containing 0.1% FCS were wounded and then cultured for 24 hours after replace ment of medium with fresh DMEM containing 0.1% FCS supplemented with acidic or basic FGF (100 pmol/L). Media were collected after 24 hours, and latent ( ) and bioactive (W) TGF-6 was determined as outlined in Materials and Methods. Results are presented as mean 2 SD. *P < 0.01 vs. respective control (DMEM).

1260

GASTROENTEROLOGY Vol. 106, No. 5

DIGNASS ET AL.

effects of FGFs on intestinal in the nontransformed

epithelial

IEC-6

colon cancer-derived

cell lines HT-29

gest the presence of functional in these cell lines, New and Yeoman

60

receptor

40 20 bFQF

DMEM

aFGF

bFGF *anti TGFP

rFGF +mti TGFp

lating acidic

expression

role of members

insulinlike

growth

of the TGF-a,

only limited

cell function,2-’

TGF-P,

factor family in modulating knowledge

and

epithelial

is available

on

the role of FGF family peptides in the gastrointestinal tract. FGFs have been shown to exert a variety of pleiotropit activities derm

and

in a number the

of cells derived from the meso-

neuroectoderm.1*,26

Whereas

basic

and

acidic FGFs were originally discovered on the basis of their mitogenic activity on cultured fibroblasts, it is now evident that these factors act as mitogens on a variety of mesenchymal or neuroectodermal cells. FGFs also modify some nonmitogenic

cell functions

in vitro including

che-

by and

colon cancer-

epithelia

to the effect on epithelial

and basic FGF growth

induced

factor,

in vivo.

cell proliferation,

the expression

of a key

in the IEC-6

cell line.

TGF-P,

has been found to be expressed by primary inhibitor

epithelial

of intestinal

cell proliferation

in the IEC-6 cell line.‘.4S8 These observations

in a manner

similar

ship between

TGF-a

entero-

cell lines and acts as a

epithelial

the effects of FGFs on intestinal tion may be modulated

portant

studies

basic FGF response in human

of gastrointestinal

cytes and in intestinal strong

previous

cell lines.‘* Thus, FGFs may play a role in regu-

peptide TGF-fl

who showed

proliferation

In addition Figure 7. Effect of immunoneutralizing anti-TGF-6 on acidic and basic FGF stimulation of IEC-6 cell migration. Confluent monolayers of IEC6 ceils were wounded as described in Materials and Methods. Medium was replaced by DMEM containing 0.1% FCS supplemented with acidic and basic FGF (100 pmol/L), and monolayers were incubated for a 24-hour period in the presence or absence of anti-TGF-P. Results are presented as mean 2 SD obtained from at least three independent experiments. *P < 0.001 vs. control; **P < 0.001 vs. corresponding FGF in the absence of TGF-6.

and Caco-2 sug-

receptors for these ligands

thus extending

and ligand

derived

0

cell proliferation

cell line and the human

epithelial

by induction

to that observed

suggest

that

cell prolifera-

of TGF-P

synthesis

in the interrelation-

and TGF-P.

Accumulating evidence suggests a role for FGF family peptides in the repair of epithelial injuries. Migration of both endothelial and epithelial cells is significantly enhanced by FGF.32X33 Basic FGF has been shown to be present in the mucosa of the mammalian gastrointestinal tract both by biochemical and histochemical studies. 11*13z16 The expression of other members of the FGF peptide family has not yet been extensively evaluated, but it is likely that other FGF family peptides are expressed within the mammalian gastrointestinal tract. Recent

work in our laboratory

had shown that several

motactic activity, induction or suppression of cell-specific protein synthesis or secretion, and regulation of cellular differentiation. ‘*s They also play an important role as

peptide growth factors and classical cytokines enhance intestinal epithelial restitution in an in vitro wounding

angiogenic

model through

factors in wound

healing,

tissue repair,

and

tissue regeneration. In addition, some FGFs have neuronatrophic properties and may play a role in embryonic development and differentiation, modulation of endocrine function, tion.‘4,26

and diseases with enhanced

cell prolifera-

The studies in this report show that three members of the FGF peptide family (acidic FGF, basic FGF, and KGF) cause a moderate stimulation of intestinal epithelial cell proliferation in vitro. It is also possible that other members of the FGF family exert effects on gastrointestinal epithelium. Stimulation of intestinal epithelial cell proliferation in vitro was observed in the presence of concentrations of FGFs that are assumed to be present in proximity to intestinal epithelial cells in vivo. The effects on intestinal epithelial cell proliferation caused by FGFs are less marked than those induced by TGF-a and epidermal growth factor (EGF), which are potent stimulators of intestinal epithelial cell proliferation.4 The

a TGF-P-dependent

pathway.

restitution is a process by which epithelial rapidly re-established after various forms

Epithelial

continuity of injury

is by

migration of viable epithelial cells from areas adjacent to or just beneath the injured surface to cover the denuded area. It occurs both in vivo and in vitro within minutes to hours and does not require cell proliferation. It is assumed to be the initial process of epithelial wound healing followed by a more delayed wound healing that requires cell proliferation and cell differentiation.34-‘” Recent work by Folkman et al. suggested that the healing of cysteamine-induced chronic duodenal ulcers in rats was retarded by inactivation of basic FGF and enhanced by administration of exogenous basic FGF.” These findings were attributed to the angiogenic properties of basic FGF. However, the relative contribution of these factors on wound contraction or epithelial restitution apart from angiogenesis was not evaluated. More recently, Paimela et al. reported that basic FGF enhances the rapid re-

May 1994

MODULATION OF EPITHELIAL CELL FUNCTION BY FGF

epithelialization

after superficial

gastric

in an in vitro model of gastric injury, action of basic FGF was not further The studies FGF promote

in this report intestinal

in vitro wounding FGF on epithelial potent

show that acidic and basic

Maximum

observed

after addition

(IFN-y),

and interleu-

gamma

m t h e same in vitro wounding

stimulation

of epithelial

IFN-y,

greater

model.*l

cell restitution

acidic and basic FGF was on mean 4.2-, 8.5-fold

in an

than that by TGF-a,

by

2.3-, 2.7-, and

EGF, IL-lb,

and

respectively.

Interestingly, cell restitution pathway

acidic and basic FGFs enhance through

the

shown for TGF-a,

same

epithelial

TGF-P-dependent

EGF, IFN-7,

tial studies showed the ability

and IL-lfi.

of TGF-P

of restitution

Ini-

itself to enhance

restitution when added to wounded monolayers.” deed, FGFs did not effect any marginal additional hancement

when added in conjunction

Inenwith

exogenous TGF-P. The functional importance of this TGF-P-mediated modulation of intestinal epithelial restitution TGF-P

6. Park JH, Vanderhoof JA, Blackwood D. Macdonald RG. Characterization of type I and type II insulin-like growth factor receptors in an intestinal epithelial cell line. Endocrinology 1990; 126:29983005. 7. Park JHY, McCusker RH, Vanderhoff JA, Mohammadpour H, Harty RF, MacDonald RG. Secretion of insulin-like growth factor II (IGFII) and IGF-binding protein-2 by intestinal epithelial (IEC)-6 cells: implications for autocrine growth regulation. Endocrinology 1992;131:1359-1368.

in vitro were even more

than the effects previously

of TGF-CC, EGF, interferon kin lp (IL-lp)

cell restitution

The effects of acidic and basic

cell restitution

like growth factor-i and insulin-like growth factor-11 mRNAs in rat fetal and adult tissue. J Biol Chem 1986;261:14539-14544.

injury

but the mechanistic explored.”

epithelial

model.

mucosal

8. Suemori S, Ciacci C, Podolsky DK. Regulation of transforming growth factor expression in rat intestinal epithelial cell lines. J Clin Invest 1991;87:2216-2221. 9. Young GP, Taranto TM, Jonas HA, Cox AJ, Hogg A, Werther GA. Insulin-like growth factors in the developing and mature rat small intestine: receptors and biological actions. Digestion 1990; 46:240-252. 10. Flaumenhaft R, Abe M, Mignatti P, Rifkin DB. Basic fibroblast growth factor-induced activation of latent transforming growth factor beta in endothelial cells: regulation of plasminogen activator activity. J Cell Biol 1992; 118:901-909. 11. Cordon-Cardo C, Vlodavsky I, Haimovitz-Friedman A, Hicklin D, Fuks Z. Expression of basic fibroblast growth factor in normal human tissues. Lab Invest 1990;63:832-840. 12. New BA, Yeoman LC. Identification of basic fibroblast growth factor sensitivity and receptor and ligand expression in human colon tumor cell lines. J Cell Physiol 1992;150:320-326. 13.

was shown by the ability of immunoneutralizing antibody to block the FGF-enhanced epithelial

restitution.

In contrast

EGF, IFN-Y, and ILl-0, tion of TGF-Pl,

to our previous

work on TGF-CX,

which enhanced

the bioactiva-

acidic and basic FGF enhanced

both the

that these actions are mediated in part through modulation of the extracellular matrix. Recent studies have shown alteration of intestinal epithelial cell migration in vitro by the underlying matrix.42,4’ Further studies will be needed to define the integrated soluble lating

growth growth

coordination

factors and extracellular

matrix

between in regu-

and restitution.

References 1. Potten CS, Nellet M, Roberts SA, Revi DA, Wilson GD. Measure-

2.

3.

4.

5.

ment of in vivo proliferation in human colorectal mucosa using bromodeoxyuridine. Gut 1992;33:71-78. Barnard JA, Beauchamp RD, Coffey RJ, Moses HL. Regulation of intestinal epithelial cell growth by transforming growth factor type 13.Proc Nat1 Acad Sci USA 1989;86:1578-1582. Koyama S, Podolsky DK. Differential expression of transforming growth factors a and p in rat intestinal epithelial cells. J Clin Invest 1989;83:1768-1773. Kurokawa M, Lynch K, Podolsky DK. Effects of growth factors on an intestinal epithelial cell line: transforming growth factor p inhibits proliferation and stimulates differentiation. Biochem Biophys Res Commun 1987; 142:775-782. Lund PK, Moats-Staats BM, Hynes MA. Somatomedin-C/insulin-

Nice EC, Fabri L, Whitehead RH, James R, Simpson RJ, Burgess AW. The major colonic cell mitogen extractable from colonic mucosa is an N-terminally extended form of basic fibroblast growth factor. J Biol Chem 1991;22:14425-14430.

14. Baird A, Biihlen P. Fibroblast growth factors. In: Sporn MB, Roberts AB, eds. Peptide growth factors and their receptors. I. New York: Springer-Verlag. 1991:369-418. 15.

bioactivation of TGF-P and also the expression of TGFp mRNA and production of latent TGF-P peptide. This might reflect a different mechanism by which acidic and basic FGF modulate the synthesis of TGF-P. It is possible

1261

Rubin JS, Osada H, Finch PW, Taylor WG, Rudikoff S, Aaronson SA. Purification and characterization of a newly identified growth factor specific for epithelial cells. Proc Natl Acad Sci USA 1989;86:802-806.

16. Folkman J, Szabo S, Stovroff M, McNeil P, Li M, Shing Y. Duodenal ulcer: discovery of a new mechanism and development of angiogenic therapy that accelerates healing. Ann Surg 1991; 214:414-427. 17. Paimela H, Goddard PJ, Carter K, Khakee R, McNeil PL, Ito S, Silen W. Restitution of frog gastric mucosa in vitro: effect of basic fibroblast growth factor. Gastroenterology 1993;104: 1337-1345. 18. Ciacci C, Lind SE, Podolsky DK. Transforming growth factor p regulation of migration in wounded rat intestinal epithelial monolayers. Gastroenterology 1993; 105:93-101. 19. Sato Y, Rifkin DP. Inhibition of endothelial cell movement by pericytes and smooth muscle cells: activation of latent transforming growth factor pl like molecules by plasmin during coculture. J Cell Biol 1989; 109:309-315. 20. Quaroni A, Wands J, Trelstad TL, lsselbacher KJ. Epithelial cell cultures from rat small intestine. J Cell Biol 1979;80:245-265. Dignass AU, Podolsky DK. Cytokine modulation of intestinal epithelial cell restitution: central role of transforming growth factor p. Gastroenterology 1993; 105:1323-1332. 22. Chirgwin JM, Przybyla AF, Rutters WJ. Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 1979: 18:5295-5299. 21.

23. Melton DA, Krieg DA, Rebagliati MR, Maniatis T, Zinn H, Green MR. Efficient in vitro synthesis of biologically active RNA and RNA hybridization probes from plasmids containing a bacteriophage sp, promoter. Nucleic Acids Res 1984;12:7055-7056. 24. Scheving LA, Shurba RA, Nguyen TD, Gray GM. Epidermal growth

1262

factor receptor of the intestinal enterocyte. Localization to latercbasal but not brush border membrane. J Biol Chem 1989;264:1735-1741. 25.

GASTROENTEROLOGY Vol. 106, No. 5

DIGNASS ET AL.

Danielpour D, Dart LL, Flanders KC, Roberts AB, Sporn MB. Immunodetection and quantitation of the two forms of transforming growth factor 6 (TGF81 and TGFj32) secreted by cells in culture. J Cell Physiol 1989; 138:79-86.

R. Repair of rabbit duodenal mucosa after acid injury in vivo and in vitro. Gastroenterology 1987;92:1973-1986. 35. Lacy ER. Epithelial restitution in the gastrointestinal tract. J Clin Gastroenterol 1988;1O(Suppl):72-77. 36. Moore R, Carlson S, Madara JL. Rapid barrier restitution in an in vitro model of intestinal epithelial injury. Lab Invest

26. Gospodarowicz D, Neufeld G, Schweigerer L. Fibroblast growth factor: structural and biological properties. J Cell Physiol Suppl 1987;5:15-26.

37.

27. Carroll KM, Wong lT, Drabik DL, Chang EB. Differentiation of rat small intestinal epithelial cells by extracellular matrix. Am J Physiol 1988;254:G355-G360.

38.

28. Coffey RJ, Shipley GD, Moses HL. Production of transforming growth factors by human colon cancer cells. Cancer Res 1986;46:1164-1169.

39.

40.

29. Coffey RJ, Goustin AJ, Soderquist AM, Shipley GD, Wolfshohl J, Carpenter G, Moses HL. Transforming growth factor CLand 8 expression in human colon cancer lines: implications for an autocrine model. Cancer Res 1987;47:4590-4594.

41.

30.

Hahn G, Stallmach A, Hahn EG, Riecken OE. Basement membrane components are potent promoters of rat intestinal epithelial cell differentiation in vitro. Gastroenterology 1990;98:322335.

42.

31.

Hoosein NM, Brattain DF, McKnight MK, Levine AE, Brattain MG. Characterization of the inhibitory effects of transforming growth factor j3 on a human colon carcinoma cell line. Cancer Res 1987;47:2950-2954.

43.

32.

Mustoe TA, Pierce GF, Morishima C, Duvel TV. Growth factor induced acceleration of tissue repair through direct and inductive activities in a rabbit dermal ulcer model. J Clin Invest 1991;87:694-703.

33. Sato Y, Rifkin DB. Autocrine activities of basic fibroblast growth factors: regulation of endothelial cell movement, plasminogen activator synthesis and DNA synthesis. J Cell Biol 1988; 107:1199-1205. 34.

Feil W, Wentzl E, Vattay P, Starlinger M, Sogukoglu R, Schiessel

1989;60:237-244. Morris GP, Wallace JL. The roles of ethanol and of acid in the production of gastric mucosal erosions in rats. Virchows Arch B 1981; 38:23-38. Nusrat A, Delp C, Madara J. Intestinal epithelial restitution. J Clin Invest 1992; 89:1501- 1511. Rutten MJ, Ito S. Morphology and electrophysiology of guinea pig gastric mucosal repair in vitro. Am J Physiol 1983;244:G171G182. Silen W. Gastric mucosal defense and repair. In: Johnson LR, ed. Physiology of the gastrointestinal tract. Volume 2. 2nd ed. New York: Raven, 1987:1044-1069. Waller DA, Thomas NW, Self TJ. Epithelial restitution in the large intestine of the rat following insult with bile salts. Virchows Arch A 1988;414:77-81. Basson MD, Modlin JM, Flynn SD, Jena BP, Madri JA. Independent modulation of enterocyte migration and proliferation by growth factors, matrix proteins, and pharmacologic agents in an in vitro model of mucosal healing. Surgery 1992; 112:299-308. McCormack SA, Viar MJ, Johnson LR. Migration of IEC-6 cells: a model for mucosal healing. Am J Physiol 1992;263(3:426-435.

Received October 4, 1993. Accepted December 21, 1993. Address requests for reprints to: Daniel K. Podolsky, M.D., Gastro intestinal Unit, Jackson 7, Massachusetts General Hospital, 32 Fruit Street, Boston, Massachusetts 02114. Fax: (617) 7263673. Supported by grants from the National Institutes of Health (DK 41557, DK 43351) and by the Deutsche Forschungsgemeinschaft (Di 477/l-l). The authors thank D. Franchimont for technical assistance and Drs. S. Aaronson and J. S. Rubin for generously providing recombinant keratinocyte growth factor.