Evidence for a Common Mechanism of Action for Fatty ...

0026-895X/94/061070-07$03.OO/O copyright © by The American Society for Pharmacology All rights of reproduction in any form reserved. MOLECULAR PHARMACOLOGY, 46:1070-1076, 1994.

and Experimental

Therapeutics

Evidence for a Common Mechanism of Action for Fatty Acids and Thiazolidinedione Antidiabetic Agents on Gene Expression Preadipose Cells AZEDDINE

IBRAHIMI,

MICHAEL

LYDIA

TEBOUL,

A. CAWTHORNE,

DANIELLE

and PAUL

GAILLARD,

state.

September

thiazolidinediones

of target

sensitivity diabetic

tissues

The

are able

of these

exerts,

49653,

drugs

preadipose

in

on phenotypic

cells, potent

expres-

have been compound,

effects

These effects of BRL 49653 in Ob 1771 preadipose of kinetics,

involved

on the

of genes

in terms

proteins

the

metabolism.

are similar,

in fatty

reversibility,

acid

specificity

of

genes affected, and requirement for protein synthesis, to those already described for natural or nonmetabolizable fatty acids.

Moreover,

when

used

at submaximally

Thiazolidinediones improve

insulin

dependent

Zucken

a new in

diabete

fatty

nediones

mellitus,

rats.

This

and

insulin

triglycerides, are

tulated

are sensitivity

most

that

action.

with

ciglitazone,

the

cose

metabolism

and

zone

administration transport

regulatable

or insulin-like work (UMR

and

was supported 134 CNRS).

the

play

KKA

by

INSERM

a central ob/ob

corrects

the

expression

factor-i, (Grant

gene

expression

L. T., 0G.,

effects

is lost when

act in an additive

in preadipose

cells,

one of the compounds

effective concentration. mechanisms of action

manner

but this

additivity

to of

is used at a maximally

These observations, for thiazolidinediones

suggesting similar and fatty acids, are

strongly supported by the demonstration that (i) both molecules activate, in a heterogolous trans-activation assay, the same nuclear receptor of the steroid/thyroid hormone nuclear receptor superfamily and (ii) transfection of 3T3-C2 fibroblasts with an expression vector for this nuclear receptor confers thiazolidinedione inducibility of adipocyte lipid-binding protein gene expression.

genic effects even in the absence of insulin (6). The mechanisms by which thiazolidinediones various effects are not well understood. It has strated that pioglitazone is able to regulate ALBP

been

pos-

role

in their

mice

treated

rates

of glu-

(2, 3). Pioglitathe

deficit

in

of insulin-

(4). In activators of cell differin combination with

4 in adipose

dramatically CRE920708

and 2-bromopalmitate

induce

and

mice

of thiazolidi-

mice

type

(Al.,

adipocyte differentiation the ciglitazone metabolite

in glucose,

it has

are increased

act as potent cells. Pioglitazone, growth

effects

from

up-regulating

transporter

db/db

insulin-stimulated

lipogenesis

2, France

that

non-insulin-

decreases

animals,

In adipocytes basal

by

glucose

thiazolidinediones entiation in 3T3-L1

This CNRS

must

to obese

vitro,

insulin

Because

of

and

with

in obese

tissue

pharmacological

glucose

(1).

agents

models

as ob/ob

such

Nice Cedex

concentrations,

of antidiabetic

rodent

is associated

marked

adipose

effective

class

Facult#{233} des Sciences, 06108 AL6 9AR, UK (P.Y., M.A.C.)

BRL49653

insulin

partially,

expression cells

encoding

Antipolis, Welwyn,

to improve

at least

sion in various tissues, including adipose tissue, reported. We report here that a new thiazolidinedione BRL

YOUNG,

26, 1994

and to reverse,

effects

PAUL

tissue

increases to P.A.G.)

and

by

(5), whereas ADD 4743

in STi3 exerts

pneadipose its potent

cells adipo-

exert their been demongene expres-

sion in 3T3-Li cells (7). We have reported that fatty acids are involved in the regulation of expression of adipose-related genes, such as ALBP and ACS, in Ob i771 preadipose cells (8, 9). Short term exposure of preadipose cells, i.e., committed, non-terminally differentiated cells that express early markers and in which genes such as ALBP or ACS are transcniptionally inactive, to long chain fatty acids led to induction of these genes and to an accumulation of the corresponding mRNAs. This effect of fatty acids was dose dependent and reversible upon removal of the inducer. A nonmetabolizable fatty acid, 2-bromopalmitate, was shown to be a better inducer than palmitate, demonstrating that fatty acid metabolism was not required for the induction of gene expression in preadipose cells (iO).

ABBREVIATIONS: ALBP, adipocyte lipid-binding protein (also called aP2, p422, and aFABP); ACS, acyl-CoA synthase; BRL 49653, 5-(4-[(N-methylN2-pyridylamino)ethoxy]benzyl)thiazolidine-2,4-dione; CAT, chloramphenicol acetyltransferase; FAAR, fatty acid-activated receptor; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GPDH, glycerophosphate dehydrogenase; hGR, human glucocorticoid receptor; PPAR, peroxisomal proliferator-activated receptor; SVF, stromal vascular fraction; MMTV, mouse mammary tumor virus; MALl , keratinocyte lipid-binding protein; ARE, adipocyte regulatory element. 1070

Downloaded from molpharm.aspetjournals.org at ASPET Journals on December 25, 2016

SUMMARY In diabetic rodents,

AILHAUD,

AMRI,

A. GRIMALDI

Centre de Biochimie, UMR 134 CNRS, Universit#{233}de Nice-Sophia E.-Z.A., GA., P.A.G.), and SmithKline Beecham Pharmaceuticals,

Received June 15, 1994; Accepted

GERARD

EZ-ZOUBIR

in

Mechanism It is now acids

well

are

the

established

mediated

steroid/thyroid

tons,

called

tissues

are

carboxylates

we have a new

hormone

receptor

have

been

activated

such

from

member

in various

and

mouse

of this

recep-

a cDNA

encoding

hormone

been termed be considered

pothesized

receptor

ing

in

mediating investigated

the

sion

(16),

in cells

mitate

increase

two

BRL

of the

it can

49653,

Ob

1771 BRL

FAAR, which tions strongly gene

expression

of gene

clonal

line. and

is able

in preadipose thiazolidinediones

cells.

in preadipose

cells

exthat

From

these

2-bromopal-

and

to activate

the

to activate

the

These observaand fatty acids

by similar

mecha-

Materials Cell culture.

Oh

cm2 and were mented with streptomycin, termed

cells

medium.

was

reached

(17)

were

plated

Media within

were

changed

5 days.

standard

This cells.

medium

to

obtain

medium Control

was prewarmed experiments

or

dimethylsulfoxide.

ethanol 2-bromopalmitate

were

of the cells with

the

of 2 x iO/

every

performed

Where

removed

from

concentration

as

and

to

then

exclude

indicated, the

day and

and 2-bro-

(18)

(2

x

culture

added any

BRL medium

to the

effects

of

49653 by

and

washing

standard

105/100-mm

plate)

maintained

in standard

medium

were

co-transfected with 1 tg of pMAMneo (Clontech) and 20 g of pSG5 without (control cells) or with (FAAR cells) the FAAR insert, using the DOTAP (N-[1-(2,3-dioleoyloxy)propyl]-N,N,N-triethylammonium methylsulfate) transfection method. Geneticin was added to the medium 48 hr later and geneticin-resistant clones were selected and tested for

responses

Wistar

to BRL

tissue

Adipose rats

were

49653.

culture. rinsed

Epidydimal

0.2

nM

fat

in phosphate-buffered

into approximately 150-mm3 pieces, ified Eagle’s medium supplemented tranferrin,

tniiodothyronin,

acids

acids

and

1-486)

137-440).

the

and

The

transfected

in which

into

pads

from

saline,

pH

and incubated with 5 g/ml 20 nM

sodium

the

chimeric

COS-1

expression

within

of

immobilized

as an

internal

standard.

The

hybrid

receptor

(15).

to the

D and the the

serum,

using

and

incubated

Cell

extracts

5 zg ofboth

plasmids

and

Mannheim,

Meylon,

France).

in the were

same

the

A/B

and

hGR/

C domains

domains

of

medium

MMTV-CAT, of the

glucocorti-

prepared

48

hr

reagent

8 hr, cells

presence

after

was perred-free fetal calf

transfection

After

in the

FAAR

by sequenc-

plasmid control

DOTAP

the

construction

was verified

with

is under

This

ElF

construct

cells

of CAT

from

coid-responsive MMTV promoter. Transient transfection formed in 60-mm dishes with cells maintained in phenol medium supplemented with 8% activated charcoal-treated were

washed

of various

transfection

and

ligands. assayed

for

CAT enzyme activity, using the CAT enzyme assay system (Promega, Charbonnieres, France). Control experiments were performed with cells transfected either with the MMTV-CAT plasmid alone or with the hGR expression plasmid and the reporter gene. Transfection experiments

were

performed

Materials. toise,

in triplicate.

Culture

France).

media

were

obtained

2-Bromopalmitate

was

from

from

the

MP were

from

random

priming

Amersham

kit,

(Les

Hybond

Ullis,

GIBCO

Aldrich

France). Bovine serum and other from Sigma Chimie (Saint-Quentin,

materials,

8-week-old 7.4,

and

100

of BRL 49653 1771 cells. The

confluent

Oh

M

sodium ascorbate, in the absence or presence of 0.8 M BRL 49653. Twenty hours later, adipose tissue fragments were digested with collagenase and cells ofthe SVF were isolated by centrifugation, as described previously (19). RNA analysis. RNA samples were prepared as described by Chomczynski and Sacchi (20). RNA was analyzed as described previously (21). Results were quantitated by densitometry using an LKB Ultroscan

1771

2-bromopalmitate genes

cells (iOO

were

(Cergy-PonChimie

(Saint-

chemical products were France). Radioactive membranes,

and

Hyperfilm

France).

to

either

by

were

because

on gene expression effects of exposure

were

level

in preadiof i-day post-

49653

(100

MM)

ofexpression

Northern

already

they

BRL

MM) on the

investigated

1, cells

blot

committed

analysis.

As

shown

to differentiate

expressing

the

early

or

of several at this

marker

p0b24/

A2COL6 (22). However, they were not yet terminally differentiated, as demonstrated by the absence of late markers such as ALBP, ACS, and GPDH mRNAs (Fig. i, lanes 1 ). Exposure to BRL 49653 for 24 hr led to emergence of ALBP and ACS mRNAs,

whereas

unchanged newly

the

described

tissues

level

1, lane

(Fig.

fatty

including This

of pOb24/A2COL6

3).

MALi

tissue

mRNA

mRNA

mRNA,

acid-binding

adipose

of expression.

protein (23),

was

encodes

present

showed

already

remained

which a different

expressed

a

in various pattern

at a low,

but

level its expression

in 1-day postconfluent cells (Fig. 1, lane I), and was enhanced in cells exposed to BRL 49653 (Fig. 1, larw 3). In contrast to the aforementioned adiposerelated mRNAs, GPDH mRNA remained undetectable at this time. Additional experiments, not presented here, have shown that other adipose-related mRNAs, such as adipsin, hormonesignificant,

sensitive

chopped

in Dulhecco’s modinsulin, 10 zg/ml selenite,

Effects pose Ob

time,

indicated.

medium at 37’ (two washes of 15 mm each). Stable transfectants. An expression vector (pSG5/FAAR) was obtained by insertion of the coding sequence (nucleotides 1-1440) of FAAR into the BamHI site of the pSG5 vector (Stratagene). 3T3-C2 cells

(amino

in Fig.

of 50 mM in dimethylwere added immediately

at 37’ for 15 mm were

other

Thiazolidinediones

final

used

taken

a function

Results

at a density

mopalmitate were dissolved at a concentration sulfoxide and ethanol, respectively. Aliquots to

were

activation

and Methods

grown in Duihecco’s modified Eagle’s medium supple8% bovine serum, 200 units/ml penicillin, 50 sg/ml 33 iM biotin, and 17 zM pantothenate. This medium is

standard

confluence

1771

signals

as

1071

Expression

were

peaks

assay. was prepared by PCR mutagenesis a chimeric protein corresponding

Quentin, purchased

nisms.

measurements

integrated

mRNA

(Boehringer

expres-

set of genes,

manner

49653

to

hypothesis, we of the thiazolidi-

49653

identical

BRL

is expressed suggest that

this

of the same

in an

on gene

similar

To test compound

that

Furthermore,

effects

on the control

preadipose

act

their

a mechanism

the transcription

gene.

activate

by

be concluded

compounds

ALBP

exert

cells

fatty acid effects. the effects of a new

class

studies the

compounds

on Gene

type 24

lipase,

4 mRNA hr

with

apparent (Fig.

compounds set of genes, sion

of the

activation genes

BRL

when

49653

and

even

A similar

pattern

2-bromopalmitate

1, lanes exert

glucose

undetectable

49653.

effects ACS,

p0b24/A2COL6 be observed.

was lanes

2 versus

positive

i.e., ALBP, of GPDH

could

insulin-regulatable

remained

on the

and gene

on other

3),

MALl. was

terminal

not

in cells

transporter

treated

of expression

for was

used instead of BRL indicating that both expression

of the

Similarly,

the

modified,

whereas

differentiation-related

same

expresno


nedione

these

pneadipose

can be hy-

All

of the

receptor

hGR

(amino

super-

FAAR (i5). Because thiazolidinediones as amphipathic molecules, it could

that

pnession

GAPDH

of

family that is likely to be involved in the transcriptional effects of fatty acids in adipose tissue. Based upon these properties, it has also

RNA.

FAAR encoded

Recently,

of Thiazolidinediones

densitometer.

response

mammalian

(11-14).

preadipocytes

laser

linear

Hybrid

of amphipathic

fibrates

steroid/thyroid

of

Such

spectrum

KL

of fatty

receptors

superfamily.

identified

acids

effects

of nuclear

by a broad

as fatty

isolated

transcriptional

activation

PPARs,

and

that

through

of Action

Ibrahimi

1072

et aL

GAPDH 123123123

GAPDH

ACS

p0b24

p0b24

ALBP

12

12

12

GPDH Fig. 2. Northern blot analysis of SVF cell RNA from adipose tissue exposed to BRL 49653. Cell culture protocol, preparation of the SVF, and RNA analysis were described in Materials and Methods. Lanes 1, control medium; lanes 2, medium supplemented with 0.8 M BRL 49653. 40 _1

z

30

:

,

5-

z

20

-

0.

10

.-

Fig. 1. NOrthern blot analysis cells exposed to BRL 49653

of adipose-related

mRNAs

or to 2-bromopalmitate.

of Ob 1771

One-day

4

post-

were maintained for 24 hr in standard medium, in the absence 1) or in the presence of 1 00 M 2-bromopalmitate (lanes 2) or 1 00 tM BRL 49653 (lanes 3). RNA (20 zg/lane) was analyzed as described in Materials and Methods. The results are representative of

confluent

cells (lanes

five separate

rat

experiments.

Effects of BRL adipose tissue.

gene

was

rodent

49653 We

expressed

adipose

at very

tissue.

demonstrated

by

mRNA

(22).

ALBP

mRNA

on ALBP gene showed previously

To

low

This

the

levels

fraction

strong

investigate

contains

expression the

expression

Adipocytes

SVF

and

and

cells

centnifugation.

were

pieces

from

SVF

as

after

SVF

cells

tissue

defined (0.8 zM).

collagenase

observation

regulating rat adipose clonal lines

not

require Potency palmitate accumulation inducer

indicates

ALBP tissue, (Fig.

gene

that

as

it is in 1). Furthermore,

serum components. of BRL 49653, to induce ALBP of ALBP concentration

mRNA after

for

was

(pM)

for 24 hr to increasing

concentrations

of BRL 49653

(U),

pioglitazone (#{149}), or 2-bromopalmitate (0). RNA was analyzed as described in Materials and Methods. The ALBP mRNA signals were normalized to GAPDH mRNA signals and the results are expressed as the mean ± standard deviation of five separate experiments. confluent cells. The that of pioglitazone same series of cells. 49653

mRNA and

maximally Pioglitazone

potency of BRL 49653 was compared and 2-bnomopalmitate determined As shown in Fig. 3, the accumulation

was

significant

exposed to 1 M BRL BRL 49653. The halfconcentration was estimated to be 3 zM. a similar positive effect on ALBP mRNA

reached

a plateau

effective exerted

accumulation,

but

with in the of

this

in cells

at

occurred

30 tM

at

higher

concentrations

than

in

cells from preadipocyte the effect in SVF cells

did

differences in half-maximally effective concentrations, i.e., 10 M pioglitazone and 30 tM 2-bromopalmitate, compared with

was

effective

in

pioglitazone, and 2-bromomRNA accumulation. The a 24-hr

100

present

49653

in preadipocytes mouse

30

for BRL 49653. The effect of pioglitazone became significant at 3 zM and was maximal at 100 sM. In agreement with our previous data (10), 2-bnomopalmitate appeared to be effective at concentrations ranging from 10 MM to 100 sM. Despite

BRL

expression

10

Fig. 3. Dose-response relationship for thiazolidinediones or 2-bromopalmitate effects on ALBP mRNA accumulation in preadipose Ob 1771 cells. One-day postconfluent cells maintained in standard medium were

ALBP

diges-

analyzed

the presence of ALBP mRNA, using GAPDH and p0b24/ A2COL6 mRNAs as internal controls. As shown in Fig. 2, incubation with BRL 49653 did not change the amount of GAPDH and p0b24/A2COL6 mRNAs, whereas the signal for ALBP mRNA was strongly increased in treated adipose tissue. This

3

on

of adipose

was

1

COMPOUND

exposed of

p0b24/A2COL6

20 hr in chemically of BRL 49653

isolated

RNA

the

of thiazolidinedione

cells,

were incubated for absence or presence

from

in ALBP

preadipocytes, of

effects

in SVF

from adult rats medium in the tion

in cells

expression that the

0

studied exposure

as

a function of

1-day

of post-

3 .sM BRL (70-80-fold) mally

Time 49653

effective

49653, in ALBP

all compounds mRNA

produced content

when

similar present

increases at

maxi-

concentrations.

course of induction and 2-bromopalmitate.

of

ALBP mRNA The time course

by BRL of ALBP


ALBP MALI I 23123123

Mechanism mRNA

accumulation

in cells

or 2-bromopalmitate

(100

to both compounds, early with time from thereafter

at

control culture

that

were

provided

medium.

after

a 24-hr

that

Removal

with a half-time of about turnover rate of this mRNA and

to addition mopalmitate whether,

hr

that to

were

(10),

the

1-day

49653

(30

(18

zM).

ALBP

in response

mRNA

cycloheximide,

for

mRNA

49653

the

of ALBP de novo

postconfluent

cells

tM), As

in

the

shown

hr

to

mRNA

were 4B,

49653

that

relationships

One-day

increasing

absence

between

postconfluent

or presence

by thia-

when

or

presence

of

induction

of

abolished was

compound was 2-bromopalmitate, in any reported

further

increase that

previously

a more A

than

additive

were

BRL

exposed

for

24

in

the

49653.

When used at 3 tM BRL 49653

i.e.,

were

active

expression.

concentration

In

of a single

30 tM BRL 49653 of the second inducer

of ALBP mRNA dexamethasone

manner

and

2-bromopalmitate,

effective

used, i.e., addition

re-

and dexTo inves-

thiazolidinediones

concentrations,

a maximally

for 6

the synthesis

cells of

of 3 or 30 tM

effective

or

synthesis.

was totally

protein

(9)

exposed

and 30 tM 2-bromopalmitate, the two compounds in an additive manner to induce ALBP mRNA contrast,

or did

100

tM result

not

content (Fig. 5A). We and fatty acids act in

to induce

ALBP

mRNA

expres-

B

!-

20

3Ogf%l

B, The same cells as in A were maintained

sion

in preadipose

BRL

49653

in ALBP zM) led

used

(9).

A similar

pattern

as a substitute

emerged

for fatty

(1 zM) for 24 hr resulted

mRNA level, whereas to a 75-fold induction.

acids.

in a 6-fold

exposure Treatment

when

Exposure induction

to BRL 49653 (30 of the cells with a

combination of the two inducers resulted in a 160-fold increase in the ALBP mRNA signal. This observation indicates that, even at maximally effective concentrations of BRL 49653, glucocorticoids

are

sion. BRL

49653

nuclear

receptor.

acids on mediated

able

to further

and

fatty

acid

We showed

of hormone

viously

for other

strated

that

binding domain transcriptional

of FAAR activation

assay,

cn

6

in

with

Fig. 6, exposure the MMTV-CAT

in

a dose-dependent

(15).

family and

gene. the

containing the

the

putative

ligand-

Using

the

possible

same

activation

by thiazolidinediones.

to BRL 49653 vector and manner,

pne-

we demon-

to exhibit fatty acid-dependent the glucocorticoid-responsive

CAT receptor

same

of fatty

As shown

(11-14),

(hGR/FAAR)

we investigated

chimenic

the

preadipocytes could be that is a member of the

hGR

is able of

expres-

the effects

receptors

the

promoter-driven

of

that

mouse FAAR,

of this

from

gene

activation

receptor

domain

ALBP

recently

nuclear

members

a hybrid

DNA-binding

hGR/FAAR

increase

gene expression in by a protein, named

superfamily

8

4

cells

was

to dexamethasone

L. zu, E

for 24 hr in control medium

or were exposed to 1 M dexamethasone (dexa.), to 30 M BRL 49653, or to a combination of both agents. In A and B, RNA was analyzed as in Fig. 2 and the results are the mean ± standard deviation of four separate experiments.

activation 30

1i!i

Fig. 5. Effects of BRL 49653 and other inducers of ALBP mRNA expression. A, One-day postconfluent cells were maintained for 24 hr in standard medium alone (0) or supplemented with 30 M () or 100 M (U) 2-bromopalmitate, with the indicated concentrations of BRL 49653.

MMTV

10

40

-

3p%l IIRI. 49653

induction, the effects of treatment 49653 and 2-bromopalmitate were

concentrations

submaximally

1

I

0

determine acids

protein

absence

in Fig.

to BRL

acids in ALBP gene combinations of BRL

determined.

mRNA

or 2-bro-

to fatty

upon

indicating

further

fatty with

ii;J

the

in response

quired for this induction process. Effects of BRL 49653 on 2-bromopalmitateamethasone-induced ALBP mRNA expression. tigate

in

in

in ALBP

natural

induction

dependent

cycloheximide by

BRL

B ‘o

is consistent with the cells (24). The kinetics

performed

described

was

purpose,

A

1073

Expression

increase

of the

As shown

of COS-1 cells the hGR/FAAR

to an

trans-

tnansfected vector led,

in CAT

activity.

Q-

a

a .5

10

BRL

2 0

24

48 TIME

(hours)

72

BRL49653 Cycloheximide

than with

-

-‘ -

-

±

±

-

± ±

Fig. 4. Time course of induction of ALBP mRNA by BRL 49653 or 2bromopalmitate. A, One-day postconfluent cells were maintained in standard medium in the absence (D) or presence of 1 00 iM BRL 49653 (#{149}) or 2-bromopalmitate (A). Removal of BRL 49653 (0) or 2-bromopalmitate (1:1) was performed with two washes and incubation in standard medium. The results are presented as in Fig. 2, from five separate experiments. B, The same cells as in A were exposed or not to 30 M BRL 49653, in the absence or in the presence of 18 zM cycloheximide.

Six hours later, RNA was isolated and analyzed as in Fig. 2. The results are the mean ± standard deviation of four separate experiments.

49653

was

a more

was 2-bromopalmitate 10 tM BRL 49653,

10 M vatons

2-bromopalmitate). resulted in nearly

lation

of CAT

unclear

but

and

potent

activity. might

not

induce

CAT

vector

and

affect vector

CAT only.

CAT

vector

Higher complete The

with reported

normal

concentrations disappearance for this receptors

of both actiof the stimuobservation becoming

that

Control with

whereas

the BRL

activity in cells transfected with Moreover, in cells tnansfected

and

the unmodified

hGR,

exposure

the with

remain activated

dexamethasone

transfected vector,

hGR/FAAR

of CAT activity induction with

transcription.

showed

in cells

hGR/FAAR

of the

induction with 14-fold

reasons to other

here,

activity the

(26-fold compared

be due

interfering

ments,

activator

expenidid

not

MMTV-CAT 49653

did

MMTV-CAT the MMTV-

to dexametha-

not


For

those

present

of BRL

on Gene

to be similar.

as previously

zolidinediones

tM)

or 2-bromopalmitate

of ALBP

experiments

2-bromopalmitate

than

was

15 hn, which in cultured

respectively,

appeared

Additional

higher

49653

(100


In response

increased linwas maintained

in a decrease

disappearance

or removal,

4A.

mRNA and

inducer

resulted

49653

Fig.

in

80-fold the

of BRL

treatment

of appearance

to BRL

is shown

the level of ALBP 0 to 24 hr oftreatment

levels

cells,

exposed

tM)

of Action

1074

Ibrahimi

et a!.

30

control

transfected

cells

mRNA when maintained or exposed for 24 hr

C 0 25

expressed to

C.)

ALBP

mRNA

C

Similar clones

patterns were observed (data not shown). In

strong

signal

mRNA

(i.8

20

0

15

>

10

detectable exposure

signal.

0

1

3

10

1

30

Br-Palmitate

3

10

gene

are

in cells

a plateau

was

at 10

cate that, dinedione

form

cellular a very of FAAR

distinguished

from

(Fig. 7A, k.znes 3 and FAAR-transfected

in

were

obtained

cellular concentrations

clones

confluent

in Fig.

control

7B.

transfected

the

4). When cells did

BRL

at 1 pM

Taken

pM.

as previously BRL 49653

to this

49653;

the

together,

with

and

mRNA at

FAARremained

all

BRL

compound

49653

in FAAR-

BRL 49653 half-maximal these

three

(data not of BRL 49653

control ALBP cells

a response detectable

at 30 pM

observed

transfected

results

shown

whereas

transfected

ton

identical

expression

cells

concentrations,

Fig. 6. Activation of hGR/FAAR hybrid receptor by 2-bromopalmitate and BRL 49653. COS-1 cells were transfected with the MMTV-CAT vector and the hGR/FAAR expression vector. Transfected cells were maintained in phenol red-free medium containing 8% activated charcoaltreated fetal calf serum and were exposed for 40 hr to either 2-bromopalmitate or BRL 49653 at the indicated concentrations. CAT enzyme activity was determined as described in Materials and Methods. Results were calculated by taking as 1 the value obtained in cells maintained in control medium and are presented as the mean ± standard deviation of triplicate transfections.

the

FAAR-transfected effects of various

undetectable

BRL49653 (pM)

(pM)

ALBP

conditions.

amounts of ALBP mRNA (Fig. 7A, lane to 30 pM BRL 49653 led to the emergence

Nearly

transfected

30

both

and reached effect was

observations

mdi-

described for fatty acids, the thiazoliacts as an activator of the nuclear recep-

FAAR.

Discussion It is well established that the improvements in insulin produced by the thiazolidinedione pioglitazone are

sencon-

sitivity

sone

(0.1

pM)

for

40

hr

activity, whereas treatment any significant induction

Evidence

led

to

a 54-fold

with BRL of the enzymatic

induction 49653 did activity.

of not

CAT

lead

related

with

transcriptional

(5, 7). In the

to

and

49653

expression of FAAR in 3T3-C2 fibroblasts confers BRL 49653-responsive gene expression. We have recently shown that the constitutive expression of FAAR in 3T3-C2 fibroblasts is sufficient to confer to the

proteins

tnansfected

cell lines, because tissue maintained

acids

(15).

expression

that

BRL

present

cells

of FAAR

thiazolidinediones, 3T3-C2 fibnoblasts

expression sion

vector

A

the

To further

vector

inducibility

of the

investigate

the

and

the

regulation

gene

of gene

(control

transfected

34

cells).

cells) As

by

between

shown

in

directly

the

fatty

the

vation

is reminiscent

(8-iO,

25).

that

expres-

fatty

identical

7A,

cells

mechanisms.

ALBP

in fatty

acid

proteins

ALBP

fatty

(i) to BRL

The 49653

pioglitazone

of genes

encoding

metabolism,

and

i.e.,

MALl

ACS

(Fig.

1).

is neither related to indirect nor confined to established

acid

(Fig.

effects

from rat adipose 2). This obser-

in preadipose

induction

process

is kinetically

cells

the hypothesis effects through in

Oh

identical

1771

to that

B2o Fig. 7. Effects of BRL 49653 on gene expression in FAAR-transfected 3T3-C2 fibroblasts. A, One-day postconfluent control transfected cells (lanes 1 and 2) or FAAR-transfected cells (lanes 3 and 4) were maintained for 24 hr in standard medium alone (lanes 1 and 3) or supplemented with 30 pM BRL 49653 (lanes 2 and 4). RNA (20 pg/lane) was analyzed as described in Materials

4 --i 5 zcn

N

activation

it occurs in preadipocytes in serum-free medium of

in adipocytes

that

Several lines of evidence support acids and BRL 49653 exert their

in response

-I

FAAR

acid-binding

by

Fig.

a striking

involved

studied in an empty

or a FAAR

elicit

of ALBP

we demonstrate

This effect of thiazolidinediones actions through serum components

fatty

expression

the response to BRL 49653 was stably transfected with either

(FAAR-transfected 12

ALBP

relationship

and

activation

report,

z 4

E

and Methods.

0

separate experiments. B, One-day postconfluent control transfected cells (0) or FAAR-transfected cells (U) maintamed in standard medium were exposed for 24 hr to increasing concentrations of BRL 49653. RNA was ana-

.1 4

GAPDH

30

BRL 49653

tpM I

100

The results

are representative

of three

lyzed as in Fig. 3 and the results are expressed

as the

mean ments.

expen-

±

standard

deviation

of three

separate


on

C)

of FAAR

(Fig. 7A, lane 1) (Fig. 7A, lane 2).

under

is easily

not

independent shown). The

5,

for which

form (3.1 kilobases) in standard medium,

of the

I.-

obtained

amounts

with five independent FAAR-transfected cells,

endogenous maintained express

small medium 49653

undetectable

kilobases),

3), whereas

C.)

remained

was

very

in standard 30 pM BRL

Mechanism in response

to fatty

a plateau acids,

after this

inducer

acids,

1 day.

occurring

phenomenon

from

the

within

(ii) As previously is

fully

medium.

hours

reaching

PPAR-y2

for fatty

moten,

called for

demonstrated

reversible

(iii)

and

removal

of

crucial

synthesis

is

recently

upon

Ongoing

protein

of Action


interacts

been

scniptional

used, cally

proliferaton-nesponsive

49653 (Fig.

and fatty acids act synergisti5 and Refs. 9 and 10).

We postulate that thiazolidinediones interact with a nuclear receptor that belongs to the steroid/thyroid hormone receptor superfamily. A subclass of these receptors, called PPARs, have been implicated as nuclear targets for fibrates and fatty acids (11-14), and we have recently described the identification in pneadipose

cells

of a member

of the

family

that

is activated

by

In

this

the the

paper,

we

hypothesis transcriptional

dinedione

BRL

FAAR Furthermore, gene

acts

a potent

tnansfection

of 3T3-C2

confers (Fig.

BRL

for

that

cells

(10),

potent BRL

molecules

insulin effects

compound

effective

concentrations

centnation

hGR/

(Fig.

with

6).

of

pioglitazone is also more

(10).

This is not because

Alternatively,

this

not

bound

to

MAL-1

the

diones

are

acids,

for this

are

the

not

molecules

chimenic

are

fully

for

to that

hormone

raises

the

that

26).

the

receptor level

described nuclear

(15).

cells

This

of other

issue

responsive

suggests

to fibrates

that nuclear

receptor

the

FAAR,

as of FAAR

that

FAAR

specificity

of the

PPAR

subfamily,

(11-14).

This

is also

the

including

determination of the

transcriptional nuclear recepon gene beneficial in rodent

mellitus.

A better

of thiazolidinedione of the physiological

motifs

in

action role of

mechanism

physiological

of acti-

ligand,

the

the

metabolism.

of thiazolidinediones

of the

data

and

promoter

iden-

regions

of

authors

are

MD),

grateful

P.

Krieg

(Tohoku

and

ACS for expert

to Drs.

M.

D. Lane

(Johns

(Krebsforschungzentrum,

University,

cDNAs,

Sendai,

respectively.

technical

and

Hopkins

Heidelberg,

Japan)

The

secretarial

for the

authors

thank

assistance,

University,

Germany), and kind gift of ALBP, L. Staccini and G.

respectively.

References 1. Chang, cemic

2.

A. V., B. M. Wyse, agent. II. Effect on

(1983). Ikeda,

H.,

tissue

and B. J. Gilchrist. Ciglitazone, a new hypoglyglucose and lipid metabolism and insulin binding of C57BL/6J-ob/ob and ±/? mice. Diabetes 32:839-845

S. Taketomi,

Y. Sugiyama,

T.

Sodha,

and

K.

Meguro.

Effects

of

pioglitazone on glucose and lipid metabolism in normal and insulin resistant animals. Arzneim. Forsch. 40:156-162 (1990). 3. Colca, J. R., and D. R. Morton. Antihyperglycaemic thiazolidinediones: ciglitazone and its analogues, in New Antidiobetic Drugs (C. J. Bailey and P. R. Flatt, eds.). Smith-Gordon, New York, 255-261 (1990). 4.

these 8.

of 9.

Hofman,

C.,

K.

Lorentz,

and

J.

R.

Colca.

Glucose

transport

deficiency

in

is corrected by treatment with the oral antihyperglycaemic agent pioglitazone. Endocrinology 1 29: 1915-1925 (1991). Kletzien, R. F., S. D. Clarke, and R. G. Ulrich. Enhancement of adipocyte differentiation by an insulin-sensitizing agent. MoL PharmacoL 41:393-398 (1992). Hiragun, A., M. Sato, and H. Mitsui. Preadipocyte differentiation in vitro: identification of a highly active adipogenic agent. J. Cell. PhysioL 134:124130 (1988). Kletzien, R. F., L. A. Foelimi, P. K. W. Harris, B. M. Wise, and S. D. Clarke. Adipocyte fatty acid-binding protein: regulation of gene expression in vivo and in vitro by an insulin-sensitizing agent. MoL PhorrnacoL 42:558-562 (1992). Amri, E., B. Bertrand, G. Ailhaud, and P. Grimaldi. Regulation of adipose cell differentiation. I. Fatty acids are inducers of aP2 gene expression. J. Lipid Res. 32:1449-1456 (1991). Amri, E., G. Ailhaud, and P. Grimaldi. Regulation of adipose cell differentiation.

in 3T3-

acid

diabetes

characterization

it is ARE

genes.

The

7.

(27-30).

action,

may contribute to the on insulin sensitivity

mechanism knowledge

of DNA-binding

(15). with

thiazolidinediones controlling

in fatty

non-insulin-dependent of the complete

peroxisomal also

to activate of the same

effects

is

Acknowledgments

6.

inducens, compared of the hGR/FAAR

such

of

understanding awaits more

5.

recep-

members

family

that

models

molecules

and/or of the

cells, activity

in preadipose cells of this class of drugs

diabetic

expla-

these

for some

as well

members

Possible

be that

clearly act as weaken thiazolidinediones,

in COS-i

cells from

this

(11-14, could

phosphonylation

similar

receptor

is different

phtalates

ligands the

fibrates acids and

C2-tnansfected

molecules,

of specificity

steroid/thyroid

Interestingly, with fatty

which

and

lack

actual affect

ton in a process the

diverse

of

of ligand-FAAR interactions, but a similar has already been observed for other PPARs, by a broad spectrum of molecules including

fibrates,

nations

It is possible

ARE

gene,

interact

of FAAR

involved

the

ACS

mechanisms

directly

more tnan-

elements

that, in preadipose fatty acid-like

expression actions

Oillaux

of

in

FAAR

fatty

ton.

and

Because

to

the

could

and BRL 49653 appear of these genes by means

MALl,

con-

exact

(33).

binds

from

Fatty acids expression

in the adipose

shift

of a higher

FAAR

element

the

T. Yamamoto

of

gene

of genes

of the

concentrations

that

demonstrate a powerful

(32)

pnoliferator-responsive

expression

by preadipose

be a reflection

of thiazolidinediones

structurally

Whatever

gene

proas being

pioglitazone-induced

ALBP

to speculate that FAAR ALBP gene expression.

Baltimore,

a better

affinities

various

medium.

Thiathan

difference between due to changes in we have demon-

metabolized

to

could

is itself

of different

leading

in inducing potent than

in vivo (16). concentrations

which

is not

in the

of the specificity lack of specificity which are activated

tempting to regulate herein activate

of the

ALBP

identified

of the

in the

peroxisomal

demonstrated

of the was

expression

implicated

to

tification

ALBP

and, in turn in the nucleus, and/or a better affinity for BRL 49653. Because fatty acids and thiazolidine-

cytosol

we

vation,

a FAAR

inducibility

sensitivity at lower

be a reflection

albumin,

unbound

cells

fibroblasts

2-bnomopalmitate,

could

for

of the

COS-1

than 49653

2-bromopalmitate but

activator

49653

inducer than natural fatty acids BRL 49653 and 2-bromopalmitate medium fatty acid concentration, strated

supporting mediator of The thiazoli-

and

target

in improving exert their

required

data

7).

49653 is more gene expression.

those

as

in transfected

expression

pioglitazone zolidinediones

experimental

is involved as a nuclear of thiazolidinediones.

receptor

vector

BRL ALBP

some

FAAR effects

49653

hybrid

expression

present

that

similar

element

previously

1075

Expression

animals

II. Kinetics

of induction

of the

aP2

gene

by fatty (1991).

by dexamethasone. J. Lipid Res. 32:1457-1463 10. Grimaldi, P. A., S. M. Knobel, R. Whitesell, the aP2 gene by nonmetabolized USA 89:10930-10934 (1992).

long

chain

and fatty

acids

and

modulation

N. Abumrad. acids.

Proc.

Induction NatL

Aced.

of Sci.

11.

the

case for the adipose-specific form of PPAR-y2 recently described as being involved as a transcriptional mediator of fatty acid and fibrate activation of ALBP gene expression (31).

Issemann, I., and S. Green. Activation of a member of the steroid hormone receptor superfamily by peroxisome proliferators. Nature (Lond.) 347:645650 (1990). 12. Gottlicher, M., E. Widmark, Q. Li, and J. A. Gustafsson. Fatty acids activate a chimera of the clofibric acid-activated receptor and the glucocorticoid 13.

receptor.

Proc.

Schmidt,

A., N. Endo,

NatL

Aced.

Sci.

USA

S. J. Rutledge,

89:4653-4657 R. Vogel,

(1992).

D. Shinar,

and

G. A. Rodan.


fatty acids. This FAAR protein is expressed in various tissues, i.e., adipose tissue, intestine, muscle, heart, brain, and lung, and is induced during the process of adipose cell differentiation.

that

activation

very

a distal

ARE,

tissue-specific

has

required for the effects of both compounds (Fig. 4 and Refs. 9 and 10). (iv) The additive effects of BRL 49653 and fatty acids are lost when a maximally effective concentration of each is whereas both BRL with dexamethasone

with

the

on Gene

1076

Ibrahimi

et aL

Identification of a new member that is activated by a peroxisome 6:1634-1641

of the steroid hormone receptor superfamily proliferator and fatty acids. Mol EndocrinoL

(1992).

25.

14. Keller, H., C. Dreyer, J. Medin, A. Mahfoudi, K. Ozato, and W. Wahli. Fatty acids and retinoids control lipid metabolism through activation of peroxisome proliferator-activated receptor-retinoid X receptor heterodimers. Proc. NatL Aced.

Sci. USA

90:2160-2164

(1993).

15. Amri, E. Z., F. Bonino, G. Ailhaud, N. A. Abumrad, and P. A. Grimaldi. Cloning of a protein that mediates transcriptional effects of fatty acids in preadipocytes. J. BiOL Chem., in press. 16. Cawthorne, M. A., C. A. Lister, J. C. Holder, D. M. Kirkham, P. Young, B. C. C. Cantello, R. M. Hindley, and S. A. Smith. Anti-hyperglycaemic efficacy of a highly potent thiazolidinedione in animal models of non-insulin dependent diabetes. Diabetes 42 (Suppl. 1):204A (1993). 17. Doglio, A., C. Dani, P. Grimaldi, and G. Ailhaud. Growth hormone regulation of the expression of differentiation-dependent genes in preadipose Ob 1771 cells. Biochem. J. 238:123-129 (1986). 18. Green, H., and 0. Kehinde. Sublines of mouse 3T3 cells that accumulate lipid. Cell 1:113-116 (1974). 19.

Deslex, S., serum-free

R. Negrel,

mRNAs in 3T3 adipocytes: 100:514-520 (1985).

and

G. Ailhaud.

Development

of a chemically-defined, precursor cells. Exp.

24.

Cook,

(1993).

K. S., C. R. Hung,

B. M. Spiegelman.

Developmentally

regulated

B. M. Spiegelman.

Fatty

acid

regulation

of

expression: transcriptional and post-transcriptional mechanisms. J. BiOL Chem. 267:5937-5941 (1992). 26. Green, S. Receptor-mediated mechanisms of peroxisome proliferators. Biochem. PharmocoL 43:393-401 (1992). 27. Denner, L. A., N. L. Weigel, B. L. Maxwell, W. T. Schrader, and B. W. O’Malley. Regulation of progesterone receptor-mediated transcription by phosphorylation. Science (Washington D. C.) 250:1740-1743 (1990). 28. Power, R. F., J. P. Lydon, 0. M. Conneely, and B. W. O’Malley. Dopamine activation of an orphan of the steroid receptor superfamily. Science (Washington D. C.) 252:1546-1548 (1991). 29. Hirata, Y., K. Kiuchi, H. C. Chen, J. Milbrandt, and G. Guroff. The phosphorylation and DNA binding of the DNA-binding domain of the orphan receptor

30.

NGFI-B.

J. BiOL Chem.

268:24808-24812

(1993).

Hsieh, J. C., P. W. Jurutka, S. Nakajima, M. A. Galligan, Shimizu, G. K. Whitfield, and M. R. Haussler. Phosphorylation vitamin

D

receptor

by

protein

kinase

C.

J. BiOL Chem.

C. A. Haussler,

Y.

ofthe human 268:15118-15126

(1993).

31.

Tontonoz,

P., E. H. Hu, R. A. Graves,

mPPAR--y2: 32.

33.

tissue-specific

8:1224-1234 (1994). Ross, S. R., R. A. Graves, and B. M. Spiegelman. of gene expression for 87:9590-9594 (1990). Harris, P. K., and R. F.

element 445

Send CNRS), 06108

and

and

J. Cell BiOI.

control.

gene

Res. 168:15-30

268:17362-17369

R. J., G. S. Robinson,

of transcriptional

in the adipocyte

A. I. Budavari, of an adipocyte

regulator

A. Greenstein, A fat-specific adipocyte Kletzien.

fatty

and B. M. Spiegelman. enhancer. Genes Dev.

K. A. Platt, H. L. Shyu, is the primary

B. Mellovitz,

enhancer P2

in

vivo.

Proc.

Localization

acid-binding

of

protein.

NatL

Aced.

determinant Sci. USA

a pioglitazone PharmacoL

response

Mo!

45:439-

(1994).

reprint

requests to: Paul Universit#{233} de Nice-Sophia Nice c#{233}dex 2, France.

A. Grimaldi, Antipolis,

Centre

de Biochimie

Facult#{233}des Sciences,

(UMR Parc

134 Valrose,


medium for differentiation of rat adipose Cell (1987). 20. Chomczynski, P., and N. Sacchi. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. AnaL Biochem. 162:156-159 (1987). 21. Amri, E. Z., C. Dani, A. Doglio, J. Etienne, P. Grimaldi, and G. Ailhaud. Adipose cell differentiation: evidence for a two-step process in the polyaminedependent 0b1754 clonal line. Biochem. J. 238:115-122 (1986). 22. Ailhaud, G., P. Grimaldi, and R. Negrel. Cellular and molecular aspects of adipose tissue development. Annu. Rev. Nutr. 12:207-233 (1992). 23. Krieg, P., 5. Feil, G. Furstenberg, and G. T. Bowben. Tumor-specific overexpression of a novel keratinocyte lipid-binding protein. J. BiOL Chem.

Distal,

analysis