Vol.

188,

No.

Nevember

3,

AND BIOPHYSICAL

BIOCHEMICAL

1992

16, 1992

RESEARCH COMMUNICATIONS Pages 1205-l 213

SPECIFIC BLOCKADE OF BASIC FIBROBLAST GROWTH FACTOR GENE EXPRESSION IN ENDOTHELIAL CELLS BY ANTISENSE OLIGONUCLEOTIDE Hiroshi

Division Research

Masashi Mukoyama, Richard and Victor J. Dzau+

of Cardiovascular Center, Stanford Drive,Stanford,

*Second

Received

Itoh',

Division, School of

October

4,

E. Pratt

Medicine and Falk Cardiovascular University Medical Center,300 Pasteur CA 94305-5246, USA

Department of Medicine, Medicine Sakyo-ku, Kyoto,

Kyoto University 606, JAPAN

1992

SUMMARY: The migration and proliferation of endothelial cells play a pivotal role in various vascular diseases. To elucidate the role of endogenous basic fibroblast growth factor (bFGF) produced within endothelial cells on cell growth, we introduced the antisense oligonucleotide complementary to bFGF mRNA into cultured bovine aortic endothelial cells by cationic liposome to block the production of autocrine bFGF. The treatment of the endothelial cells with the specific antisense oligomer efficiently inhibited the synthesis of bFGF with the concomitant suppression of endothelial proliferation, indicating the significant role of bFGF as an endothelial growth promotor. The neutralizing antibody against bFGF had no inhibition on basal DNA synthesis of the endothelial cells, in contrast to marked suppressive action of bFGF antisense oligomer. The results provide the new analytic and therapeutic implications in the use of the antisense methodology for the study of vascular biology. c 1992 Academic press, Inc.

The migration important repair, factor

initial

and proliferation features

two processes (bFGF)

capillary

endothelial

formation

of

blood

to

involve

cells

basic

fibroblast

a potent

mitogen

cells

in vitro

and can stimulate

should

(angiogenesis)

are

and wound

bFGF is

capillaries

+To whom correspondence

endothelial

of neovascularization

thought

(1).

of

for

growth

vascular

in vivo

and the

(2,3).

We

be addressed.

1205

0006-291 X/92 $4.00 Copyright 0 1992 by Academic Press, Inc. All right.y of reproduction in any .fi,rm reserved

Vol.

188,

have

No.

3,

1992

recently

BIOCHEMICAL

demonstrated

AND

that

BIOPHYSICAL

a potent

II can act

as a growth

promotor

cells

(VSMC) and,

conversely,

a vasodilator,

(ANP)

endothelial and/or present

exerts

cell action

growth,

study

is

oligonucleotide

(7)

examine

role

al

growth.

via

cells

block

Materials

sites

antibody, of action

natriuretic

the

expression specific

muscle

on VSMC and

The purpose

by utilizing

and neutralizing

and potential

atria1

the

angio-

smooth

modulating

bFGF (4-6).

to specifically

endothelial

of vascular

effects

possibly,

of autocrine

of bFGF in

cell

antiproliferative

COMMUNICATIONS

vasoconstrictor,

tensin

peptide

RESEARCH

production

of

the and action

antisense respectively

to

of bFGF on endotheli-

and Methods

Cell Culture: Bovine aortic endothelial cells (BAEC) were maintained in Dulbecco's modified Eagle's medium (DMEM) with 10% calf serum. In preparation for experiments, the cells at 80-9096 confluency were made quiescent by placing them for 3 days in Cell cultures from early pasDMEM with 0.5% calf serum (6). sages (three to eight) were used for the experiment. Determination of DNA Synthesis: The relative rate of DNA synthesis was assessed by determination of tritiated thymidine incorporation into trichloroacetic acid-precipitable material, as previously reported (4-6). Quiescent BAEC grown in 24-well Costar culture dishes were pulsed for 8 hours with tritiated thymidine (lO~Ci/ml) (20-28 hours after the transfection). Oligonucleotide Synthesis of oligomers and transfection: sequences utilized in this study and their relationshins to bFGF mRfiA are shown in Figure 1. Unmodified, 15-base deoxyribonucleotides were synthesized on an automated solid-phase synthesizer (Applied Biosystems Incorporated, Foster City, CA) using standard phosphoramide chemistry. Prior to use, the oligomers were purified by gel filtration, ethanol-precipitated, lyophilized to dryness and dissolved in the culture media. Antisense FGF oligonucleotide was complementary to human bFGF mRNA at the translation initiation region. Control oligonucleotides was the oligonucleotide with the same oligonucleotide sequence in a reversed 5'-3' orientation (reverse FGF). To introduce the oligonucleotides into BAEC, a cationic liposome-mediated transfection method (lipofection) was employed (8). Oligonucleotides dissolved in 50~1 media were mixed with LipofectinTM Reagent DOTMA (N[l-(2,3 dioleyloxy) propyll-N,N,N-trimethylammonium chloride) (BRL Life Technologies, Gaithersburg, MD) dissolved in the same volume of water in a ratio of 6/l (w/w) and incubated for 30 minutes at room temperature. The oligonucleotides/liposome complex (100,~l) was then added dropwise to each well. 1206

Vol.

188,

No.

3,

1992

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

human 5’-GCA

bFGF mRNA : GGG ACC AUG GCA GCC GGG AGC - 3’ Met Ala Ala Gly Ser antisense FGF : 3’-CCC TGG TAC CGT CGG-5’ control FGF : reverse FGF ; 5’-CCC TGG TAC CGT CGG-3’

Figure 1 . The sequences of the antisense oligonucleotide (antisense FGF) and the control oligonucleotide (reverse and their relationship to bFGF mRNA.

FGF)

BAEC were transfected with the oligonucleotide for 6 hours in serum-free culture media, on the basis that we have determined previously that the uptake of the oligomer by the cells reached plateau 8-12 hours after the administration in another series of experiments (uptake efficiency : l-2% of administered amount). The culture medium was then removed and replaced with fresh DMEM with 0.5% calf serum. Antibody characteristics: Anti-bFGF IgG (rabbit) was purchased from R & D Systems, Minneapolis, MN (Cat. No. AB-33-NA). This antibody was prepared against highly purified, natural bFGF isolated from bovine brain. Purified IgG was prepared by Protein A chromatography. The antibody does not neutralize the biological activity of acidic FGF or PDGF. The antibody (10 @g/ml) almost completely abolishes the biological activity of 1 rig/ml of bFGF. Bioassay for bFGF (Fig 2): Extraction of bFGF from BAEC and bioassay for bFGF activity using mouse 3T3 fibroblasts were performed as in a previous report (9). Confluent quiescent BAEC (1.3~10~) with or without previous treatment with antisense FGF-oligonucleotides were harvested from monolayer cultures by trypsinization, washed with phosphate-buffered saline, and resuspended in 2 ml of 1 M NaCl/O.Ol M Tris-HCl, pH 7.5, containing leupeptin (lfig/ml), pepstatin (4fiM) and phenylmethylsulfonyl fluoride (1 mM). After cells were disrupted by 3 cycles of freezing and thawing followed by sonication for 1 minute, the homogenate was centrifuged at 25,000xg for 30 minutes and the supernatant was dialyzed overnight against 0.1 M NaCl/O.Ol M Tris-HCl, pH 7.5. All procedures were performed at 4°C 1 and aliquots of cell extracts were stored at -80°C until use. For the measurement of bFGF activity, human bFGF standards (0.01-l rig/ml, Genzyme Corporation, Boston, MA) or samples preincubated (2 hours at 37°C) with either anti-bFGF IgG or nonimmune IgG at lOfigg/ml were incubated with quiescent Swiss 3T3 cells for 20 hours, after which the cells were pulse-labeled with 10nCi/m13 H-thymidine for 8 hours. Cellular bFGF content was estimated by antibody-suppressible mitogenic activity in the samples. Statistical Analysis: All -+SEM, n=6 to 12 in 3 separate ance with subsequent Duncan's icant differences in multiple ered significant.

values are expressed as mean experiments. Analysis of varitest was used to determine signifcomparisons. ~~0.05 was consid-

1207

Vol.

188, No. 3, 1992

BIOCHEMICAL

Mitogenic

Assay

AND BIOPHYSICAL

for bFGF

with

RESEARCH COMMUNICATIONS

Swiss

3T3 cells

Xlz!iXpQ5~5

(“*I

bFGF WmU

Ab: anti-bFGF NI : non-immune

x5 "+s Xl

Ab NI Ab NI WY (10) (10) WI IgG (#g/ml) IgG ( pglml)

":

“:

$l

Ab Ab NI (1) (10) (10)

($

(:i,

BAEC

extract

Figure 2 . Bioassay for bFGF. Confluent quiescent mouse fibroblast Swiss 3T3 cells incubated with DMEM and Ham's F12 medium containing insulin (5 x lO--'M), transferrin (5 fig/ml) and ascorbate (0.2 mM) were stimulated by synthetic bFGF (lOpg/ml-lng/ml) or BAEC cellular extract at the appropriate dilution (x125--1), which was preincubated with either anti at 37°C bFGF IgG (Ab) or control TgG from non-immune sera (NI) for 2 hours.

Results

In

order

endothelial say

for

the 2)

the

bFGF using of

the

dilution

standard the

examine

production

validity 1)

to

bFGF but concentration

mouse

curve curve

with not

of

bioassay

mitogenecity

preincubation

with of

the

effectiveness

bFGF,

of

we first

fibroblast

established

3T3 cells

was determined

synthetic

of

the

lOtig/ml

IgG from

cell of

the

(Fig

extract

was parallel

the

extract

was abolished

bFGF was estimated 1208

mitogenic

neutralizing

non-immune

bioasThe

2).

bFGF in

the

FGF on

as follows:

-of BAEC cellular

of

antisense

sera.

assay,

to be about

against

intracellular 7.5

and

by the

antibody The

to

ng/mg

pro-

Vol.

BIOCHEMICAL

188, No. 3, 1992

tein.

In

contrast,

we could

BAEC-conditioned In effects

ship

of

experiment,

antisense

intracellular

(reverse

FGF) had

in

synthesis the

we examined

Fig.

in

reduction

3, while

bFGF production

activity

the

these

cells

treatment

in

by the

the

study

and the

the

r

the

by the

paralleled o

control

BAEC. oligomcr

on bFGF production,

bFGF in

antisense

relationof

significantly

indicated

DNA synthesis

to

of

effect

autocrine is

parallel

incorporation

order

suppressed of

in

and DNA synthesis

no significant

The role

of

in

inhibition

FGF transfection

BAEC.

bFGF-like

FGF on =H-thymidine

bFGF production in

ty

detect

bFGF concentration

As shown

antisense

not

RESEARCH COMMUNKATIONS

media.

a separate of

AND BIOPHYSICAL

the

bFGF activicontrol

demonstration suppression

of

DNA that

of

gomer.

10 -

Lipofectin

+

oI(:.%! ?4rw

-

Figure 3 . The effect of bFGF antisense oligonucleotide production and DNA synthesis of BAEC. Quiescent BAEC T-150 flasks with DMEM containing 0.5% calf serum were 20 hours after the completion of the transfection for mination of cellular bFGF content or pulse-labeled for (20-28 hours after the transfection) to determine DNA rate in parallel. 1209

on bFGF grown in harvested the deter8 hours synthesis

Vol.

188, No. 3, 1992

Fig.4

depicts

tralizing sis

BIOCHEMICAL

the

antibody

of

BAEC.

comparison

and the

Although

the

administered

bFGF (10,

simultaneous

addition

the

had no effect

antibody

This

is

the

culture

effect

consistent

of

DNA synthesis

of

the

RESEARCH COMMUNICATIONS

effects

of bFGF neu-

antisense

FGF oligomer

mitogenic

action

the

our

of

neutralizing

FGF, which

exogenously abolished

antibody

basal

inability

the

on DNA synthe-

was completely

on the

using

of antisense

of

50 rig/ml)

with

medium

AND BIOPHYSICAL

bioassay,

and

caused

(50 fig/ml),

proliferation to detect is

by the

of BAEC. bFGF activity

in

contrast

a 60% reduction

in to the

of basal

BAEC.

Discussion

The present with

the

study

antisense

attenuated

the

=‘

4

endogenous

-

+

+

+

-

-

+

+

the

incubation

complementary

production

1 T

Lipofectin (2P9) Oligomer (4PW

that

oligonucleotide

Antisense oligonucleotide

f

E, ..O

demonstrated

of

of BAEC

to bFGF mRNA

bFGF with

the

concomi-

1 Antibody

Basal

bFGF (long/ml)

bFGF (50nglml) IgG : 50 @ml

of bFGF antisense oligonucleotide and Figure 4 . The comparison The transbFGF neutralizing antibody on DNA synthesis of BAEC. fection with the oligomers (4flM) combined with lipofectin (2Dg) or the administration of the anti bFGF IgG (50Dg/ml) or control IgG from non-immune sera with synthetic bFGF was performed on quiescent BAEC incubated with DMEM containing 0.5% calf serum.

1210

Vol.

188;

No.

tant the

3,

1992

BIOCHEMICAL

reduction control

of

of

oligonucleotide

cell

the

and restenosis role

in

endothelial

endothelial

cells

dependent

on the

the

balloon

erating (10).

presence

In addition,

the

endothelial

arteries

(11).

inhibits

endothelial

and/or

demonstrates bFGF in

normally

required

for

sial.

The cells

(14,15). tive

factor

bility

which is

the

nucleus

its

to be after

the

regen-

of

in

bFGF to

denuded

a potent

vasodilator

by suppressing

the

ANP action

The present property

of

study

of autocrine

the the

not

necessarily

the

neutralizing

(12)

the

which

is

site

of

The cellular have

been

controver-

bFGF appear

to

release

little

medium.

the

in

sequence

that

receptor

(13).

&text,

antibody are

of

synthesize

between

that

that

a signal

to be associated

In this

activities

artery, production

secretion.

role

into

the

shown

neutralizing ties

that

with

been

a key

that

of bFGF cDNA indicates

protein

of bFGF and the

also

to play

reported

carotid

proliferative

without

action

associated

rat

repair

appears

bFGF (6).

sequence

synthesized

or no growth

is

events

cells.

The nucleotide is

wound

and proliferation

reported

the

central

of bFGF was demonstrated

of endothelial

directly

the

repair

increased

infusion

significant

The ability

wound It

of

regrowth

endothelial

protein

for

proliferation

synthesis

one of

and proliferation.

showed

We also

the

bFGF appears

denuation cells

on bFGF concentration

of angiogenesis,

of bFGF.

endothelial

stimulate

is

to proliferate

with

proliferation.

proliferation

migration

COMMUNICATIONS

incubation

indicates

endothelial

angioplasty.

catheter

contrast,

herein

pathophysiology after

RESEARCH

had no effect

bFGF for

Endothelial in

In

The evidence

autocrine

involved

BIOPHYSICAL

DNA synthesis.

or DNA synthesis. role

AND

Intracellulary, On the

with

other

the

present

in

study

antibody

bFGF has

the

matrix

antiprolifera-

oligonucleotide

mutually

1211

hand,

extracellular

difference

antisense

bFGF is

and the

raises

two possibili-

exclusive.

One possi-

cannot

gain

full

access

Vol.

188, No. 3, 1992

to

the

AND BIOPHYSICAL

matrix-bound

or cell

extracellular

bFGF.

Recently,

action

of bFGF antibody

sponse

to vascular

were

BIOCHEMICAL

unable

Linder

and Reidy

injury

could

vascular

wall

with full

thickening.

antibody

injury

but

in

inhibitory

vivo

in

re-

However,

they

of DNA synthesis

invastigators free

the

catheter.

suppression

neutralize

associated

reported

balloon

These

only

surface-

on VSMC proliferation

to achieve

neointimal

(16)

RESEARCH COMMUNICATIONS

postulated

unbound

matrix-bound

or of that

bFGF released

the by

bFGF was probably

inacces-

sible. The other portion

of

regulator

ported

that

site

issue

mutant

of

of

biology.

excessive

the

vascular

angioplasty

study in

may have

translocation a nuclear

oligonucleotide circumventing

demonstrates the

proliferation,

study

further

treatment

of hypertension,

re-

the

action.

present

for

et al.

suggesting

bFGF thereby

methodology

strategies

as an "intra-

a nuclear

activity,

of

a significant

Imamura

The antisense

of

technology

pathophysiology after

FGF lacking

site(s)

antisense

This

gene-therapy

In deed,

synthesis

together,

the

growth.

of FGF (17).

cellular

that

intracellularly

of mitogenic

intracellular

is

bFGF acts

acidic

was devoid

Taken ness

of cell

of action

blocks

explanation

endothelial

crine"

sequence

attractive

of

which

the

of

vascular

application diseases is

in that

implicated

atherosclerosis

useful-

novel

involve in

the

or restenosis

(18). Acknowledgments

This work is supported by NIH grants HL35610, HL35252, HL 42663, and the University of California Tobacco Related Disease Program lRT215. Hiroshi Itoh is the recipient of Bristol Meyer-Squibb Japan-Stanford Fellowship Award. References 1. Schweigerer L, Neufeld G, Friedman and Gospodarowicz D (1987) Nature 1212

J, Abraham 325:257-259.

JA,

Fiddes

JC,

Vol.

188, No. 3, 1992

BIOCHEMICAL

AND BIOPHYSICAL

RESEARCH COMMUNICATIONS

2. Burgess WH, and Maciag T. (1989) Annu. Rev. Biochem. 58:575-606. 3. Sat0 Y, and Rifkin DB. (1988) J. Cell. Biol. 107:1199-1205. 4. Itoh H, Pratt R.E., and Dzau V.J. (1990) J. Clin. Invest. 86:1690-1697. 5. Itoh H, Pratt R.E, and Dzau V.J. (1991) Biochem. Biophys. Res. Commun. 176:1601-1609. 6. Itoh H, Pratt R.E., Ohno M, and Dzau V.J. (1992) Hypertension 19:758-761. 7. Marcus-Sekura CJ. (1988) Analytical Biochem. 172:289-295. 8. Behr JP, Demeneix B, Loeffler JP, Perez-Mutul J (1989) Proc. Natl. Acad. Sci USA 86:6982-6986. 9. Klagsbrun M, Sasse J, Sullivan R and Smith JA. (1986) Proc. Natl. Acad. Sci USA 83:2448-2452. 10. Lindner V., Reidy M.A., and Fingerle J. (1989) Lab. Invest. 61:556-563. 11. Lindner V., Majack R.A., and Reidy M.A. (1990) J. Clin. Invest. 85:2004-2008. 12. Abraham J.A., Whang JL, Tumolo A, Mergia A, Friedman J, Gospodarowicz D, Fiddes J.C. (1986) Science 233:545-548. 13. Speir E. Sasse J, Shrivastav S, and Casscells W. (1991) .J. Cell. Physiol. 1471362-373. 14. Vlodavsky I, Fridman R, Sullivan R, Sasse J, and Klagsbrun M. (1987) J. Cell. Physiol. 131:402-408. 15. Vlodavsky I, Fuks Z, Ishai-Michaeli R, Bashkin P, Levi E, Korner G, Bar-Shavic R, and Klagsbrun M. (1991) J. Cell. Biochem. 45:167-176. 16. Lindner V. and Reidy M.A. (1991) Proc. Natl. Acad. Sci USA. 88:3739-3743. 17. Imamura T, Engleka K, Zhan X, Tokita Y, Forough R, Roeder D, Jackson A, Maier JAM, Hla T, Maciag T. (1990) Science 249:1567-1570. 18. Dzau V.J., and Gibbons G.H. (1988) Am. J. Cardiol. G2:30G35G.

1213