Vol.

175,

March

No.

29,

BIOCHEMICAL

3, 1991

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

Pages 1104-1111

1991

DIFFERENTIAL

INTERACTION

PURIFIED/RECONSTITUTED

OF FATTY ACIDS AND FATTY ACYL CoA ESTERS WITH THE

BROWN ADIPOSE TISSUE Sarvagya

Departments

February

19,

and Earl

of Medicine

University Received

S. Katiyar'

MITOCHONDRIAL Shrago'

and Nutritional

of Wisconsin,

Madison,

UNCOUPLING PROTEIN

Sciences,

Wisconsin

53706

1991

Proteoliposomes containing highly purified uncoupling protein generated by a modified purification/reconstitution procedure carried out active GDP dependent proton conductance. It was further established that long chain acyl CoA esters as well as fatty acids stimulated proton influx by the uncoupling protein, and, moreover, that the acyl CoA esters were partially effective in overcoming the inhibition by GDP. GDP binding to the purified uncoupling protein was inhibited by acyl CoA esters but not fatty acids. Phenylglyoxal which prevents GDP binding to the uncoupling protein eliminated the acyl CoA but not the fatty acid effect on proton conductance. These results substantiate the fact that nucleotides and acyl CoA esters act at the same regulatory site on the uncoupling protein, whereas, fatty acids act at a separate site. The properties of the purified/reconstituted uncoupling protein confirm they are identical to those inherent in brown adipose tissue mitochondria. 0 1991 Academic

Press,

Inc.

The thermogenic dependent

upon

uncoupling

function

a specialized

protein

CoA esters

have

membrane

been

they

by the maybe

tencies

shown

in the

retained scheme

observable

properties the its of

Copyright All rights

protein

the

could found

effects (3-6)

intrinsic

both

fatty

acid

(7,8).

and fatty

should

acyl

This

be addressed.

1104

report

CoA activation

of Chemistry,

$1.50

0 1991 by Academic Press, Inc. of reproduction in any form reserved.

We have

Indian

that

These if

a nearly

a modified

homogeneous the

inconsiswhich

developed is

conduc-

possibility

be resolved

the

in the

on proton

in proteoliposomes of a near

or

by measuring

to the

UCP which

first

documen-

of GDP-sensitive Institute

is

and acyl

However,

proteins.

of UCP might

be reconstituted

properties

acids

ligands

leading

(BATM) thermogenin

fatty

(1,2).

of these

in BATM.

called

UCP as determined

due to contaminating

in situ

mitochondria

BATM both

the

purification/reconstitution

whom correspondence

0006-291X/91

binding

characteristics

'Permanent Address: Department Kanpur - 208016, U.P., INDIA. *To

tissue

to mitochondria

inconsistent

and/or

preparation

for

retained tation

UCP has been

adipose

In isolated

to activate

system,

nonspecific

homogeneous

(1,Z).

and GDP binding

purified/reconstituted tance

nucleotide

(UCP)

potential

of brown

of Technology,

has

Vol.

175,

No.

proton shown

3, 1991

conductance that

EXPERIMENTAL

the

BIOCHEMICAL

in a highly

two ligands

act

AND

BIOPHYSICAL

purified/reconstituted at separate

sites

RESEARCH

UCP. on the

COMMUNICATIONS

Furthermore

it

is

protein.

PROCEDURES

Materials: Phosphatidylcholine (egg), phosphatidylethanolamine (egg) and cardiolipin (beef heart) were purchased from Avanti Polar Lipids, Inc. Octylglucoside, Triton X-100, Amberlite XAD-4, valinomycin, GDP, anthrone, phenylglyoxal, thiourea, palmitic acid, stearic acid and oleic acid were obtained from Sigma Chemical Company. Sephadex G-50, and long chain acyl CoA esters were purchased from Pharmacia Biotechnology Products. Carboxylcyanide p-triflouromethoxyphenlhydrazone was obtained from Aldrich Chemical Co. Bio-beads SM-2(20-50 mesh) and hydroxylapatite were from Bio-Rad. Radioisotopes were obtained from New England Nuclear. All other chemicals were highest purity commercially available.

of

METHODS hamsters, cold-adapted BATM were prepared from 5-6 week o Id male Syrian at 4" for 4-6 weeks (9). The mitochondrial pellet was washed and resuspended in 20 mM MOPS, 20 mM Na,SO,, 0.16 mM EDTA, pH 6.7. It was rapidly frozen in a dry ice/methanol bath and stored at -70°C. The UCP was initially extracted in Triton X-100 and purified by the method of Lin and Klingenberg (10). The extract was subjected to hydroxylapatite chromatography on a column (3.2x10 cm) previously equilibrated with 20 mM MOPS, 20 mM Na,SO,, 0.16 mM EDTA, pH 6.7 at 4°C. Protein containing fractions were pooled and concentrated to a final volume of about 3 ml by Phosphatidylcholine (5 mg/mg of protein) ultrafiltration on a PM-10 membrane. was added to the Triton X-100 concentrate containing l-l.5 mg UCP per ml. One ml of this solution was mixed with 1 ml of 80 mM octylglucoside, 20 mM MOPS, were 20 mM Na SO,, 0.16 mM EDTA, pH 6.7 and then 7 g of water moist Bio-beads added. fhe vial containing the slurry was shaken for 2 h at 0-4°C. The removal of Triton X-100 was determined from the absorbance measurement at 275 nm according to Holloway (11). Phenylglyoxal treatment of the purified UCP was carried out prior to reconstitution as previously described (8). The purified UCP solubilized in octylglucoside was reconstituted into liposomes for influx studies by a modification of the method of Strieleman et al (3). A mixture of phosphatidylcholine / phosphatidylethanolamine / cardiolipin in the ratio 49.5:49.5:1 was dried at room temperature under a slow stream of pure N for 4 h and then dispersed into the reconstitution buffer 50 mM MOPS, 106 mM KCl, pH 7.2 containing 40 mM octylglucoside. Purified UCP in 40 mM octylglucoside was added to the detergent/phospholipid suspension to bring the final volume to 4.0 ml. The final solution had the following composition: octylglucoside to phospholipid molar ratio lO:l, phospholipid concentration 3.13 mg/ml, and UCP concentration in the range 45-70 pgg/ml. Octylglucoside was removed by dialysis in Spectrapor 2 bags for 40-44 h at 4°C against 5 one liter buffer changes (8-10 h intervals each) of 50 mM MOPS, 100 mM KCl, pH 7.2. Proteoliposomes were also generated by fast removal of detergent octylglucoside by nonionic polymeric adsorbent Amberlite XAD-4, mesh-size 20-50. Four g of wet Amberlite XAD-4 were added to 4 ml of final solution containing octylglucoside, phospholipid and UCP and shaken for 2 h at 0-4°C. The proteoliposome preparation was separated by filtration on a sintered polypropyline filter. Proton influx was measured by a slight modification of the previously reported method (4). Medium external to proteoliposomes was exchanged for 0.5 mM TES, 20 mM Na SO,, pH 6.8 by gel permeation chromatography on a Sephadex G-50 column (1x36 cm) at 4°C. A 0.6-1.0 ml aliquot of the liposome/proteoliposomes was made up to a volume of 3.0 ml with 0.5 mM TES, 20 mM Na,SO,, pH

1105

Vol.

175,

No.

BIOCHEMICALAND

3, 1991

BIOPHYSICALRESEARCH

COMMUNICATIONS

The sample was placed in a water-jacketed vessel and stirred for 6.8 buffer. 5 min at 12°C under a gentle stream of water-saturated N . To measure H' influx valinomycin (final concentration 1.0 lg/ml) was a 3 ded to generate a K' diffusion potential across the liposome membrane and the resulting change in external pti was monitored. The binding of GDP to purified UCP and proteoliposomes was determined by the gel filtration chromatography procedure reported earlier (3). SDS gel electrophoresis was carried out on 15% acrylamide gels using the method of a Thomas and Kornberg (12). Protein was determined by the modified Lowry method with bovine serum albumin as a standard (13). RESULTS Triton the

purification

UCP (lo), tion

results

near

not

a number permit

X-100

UCP into

of mitochondrial

membrane

proteins

reconstitution,

whereas,

octylglucoside

presence

of phospholipids 1 appeared the

but

protein

extraction

followed

additional

protein

only for

a partially the

vesicles

X-100

with

step

as only

one band

purified by affinity bands

protein

modified must

on polyacrylamide method

chromatography

including

the extrac(3).

is then

gel

in

exchanged

procedure

is

be carried

out

Purified

using

the

electrophoresis, octylglucoside

on ATP-agarose

contained

(3).

kD kD

32 kD kkDD kD kD

A 1.

B

SDS polyacrylamide gel electrophoresis of purified uncoupling The protein was extracted and purified in 5% Triton ~?~~'"P; hydroxylapatite chromatography and exchanged with 40 mM octylglucoside in the presence of phosphatidylcholine (5 mgjmg protein) as described under Experimental Procedures. Electrophoresis was carried out on 15% polyacrylamide gels by the method of Thomas and Kornberg (12). A: Uncoupling Protein (8 w). 6: Protein Standards.

1106

in the

UCP shown

kD

Fig.

An

of a

The UCP extracted

denaturation.

by an earlier

for

protein

chromatography which

successful

reconstitution

(7,8).

in this

octylglucoside

to prevent

purified

purification/

by hydroxylapatite The critical

of Triton

Fig.

very

phospholipid

and purified

exchange in

its

has proven

has been developed

octylglucoside.

whereas,

of BATM, which

in a reconstituted

method

homogeneous

Triton

extraction of

does

improved

with

X-100

five

Vol.

175,

No.

3, 1991

BlOCHEMlCALANDBlOPHYSlCALRESEARCHCOMMUNlCATlONS

TABLE

1

INFLUENCE OF FATTY ACIDS AND ACYL CoA ESTERS ON RATES OF PROTON CONDUCTANCE BY PURIFIED RECONSTITUTED UNCOUPLING PROTEIN Preparation

nmol

H+/min/mg

protein

110 f 40 f 110 f 250 _+ 300 f 210 t 200

Proteoliposomes t 100 pM GDP t 1 $4 Palmitic acid t 5 fl Palmitic acid t 10 fl Palmitic acid t 10 fl Stearic acid t 10 fl Oleic acid

1600

170 1850 2600 3280 2380 2330

10 $4 palmitoyl CoA 10 fl stearoyl CoA t 10 pM oleoyl CoA

2800 f 250 2650 f 250 2330 t 220

t t

100 fl

t t t

GDP t 5 @i palmitoyl 100 fl GDP t 10 fl palmitoyl 100 /rM GDP t 10 fl stearoyl 100 pM GOP t 10 pbl oleoyl

t

f

600 +

50 110 500 f 50 550 t 50

CoA CoA CoA CoA

910

+_

The proton influx measurements were done at 12°C as described under Experimental Procedures. UCP concentration in proteoliposomes was 40-62 Valinomycin was added after the addition of fatty acids or fatty acyl ~~g/ml. CoAs and the change was recorded. Data are expressed as mean f S.E. n=3.

The reconstituted effect

of either

On the

other

effects

fatty

hand

conductance

preparation

and or acyl

Strieleman

by fatty

of added

liposomes

acids

et al

acids

fatty

prepared

but

acids

by the

of fatty

acids

This

observation

(4),

which

fatty

acid

activation

protein. of a purified

al

control addition, inhibition inhibited

values

is

1 shows

mediated acid

liposomes

in proteo-

There

at

was no ionophore

results

of Strieleman

UCP, demonstrates

that

by a contaminating

on proton

has also

the

and oleate

a nonspecific

earlier pure

not

conductance.

stearate, out

the

of palmitic in

study,

at a 10 p acyl

by GDP.

palmitoyl,

the

were

The proton activated stearoyl,

H+ influx

concentration

CoA esters

by approximately

GDP, was variably As shown

effect

Table

by 30% to 100%. ruling

no

of proton

CoAs on H' influx

a partially

of H' conductance

proton

showed

transport

been

activity

recently

confirmed

(14). present

the

acyl

Palmitate,

confirming with

UCP reconstituted

et

In the

influx

besides

The stimulatory

by Jesek

chain

liposomes

obtained

on the

(5,6)

a stimulation

CoA esters.

H' influx

on control

were

CoA esters

method.

effect

and Winkler

reported

by acyl

and long

present

of 10 @l enhanced

et al

(4)

not

a concentration effect.

of Klingenberg

of UCP was stimulated of acyl

partially

conductance

90% of by acyl

its

original

and oleoyl

1107

(Table

in

value

in presence

CoA stimulated

the

In the

which

at a concentration

over

1).

overcoming

of proteoliposomes,

CoA esters

by 50%, 30%, and 35% respectively.

CoA esters

effective

20-30%

of

was 100 /.&i

of 10 $4.

GDP inhibited

H'

vol.

175, No. 3, 1991

BIOCHEMICALAND

BIOPHYSICALRESEARCH

COMMUNICATIONS

TABLE 2 INFLUENCE OF FATTY ACIDS AND FATTY ACYL CoA ESTERS ON RATES OF PROTON CONDUCTANCEBY PHENYLGLYOXAL MODIFIED PURIFIED RECONSTITUTED UNCOUPLING PROTEIN Preparation

nmol H'/min/mg

Proteoliposomes t 100 pM GDP t 10 fl Palmitic t 10 fl Stearic 5 fl 10 ti

1

1550 1440 2170 1940

acid acid

Palmitoyl-CoA Palmitoyl-CoA

protein

f f f f

110 110 200 200

1580 t 110 1600 + 110

UCP was modified with phenylglyoxal prior to incorporation into liposomes. UCP solubilized in 40 mM octylglucoside (0.41 mg protein/ml) was reacted with 15 mM phenylglyoxal for 15 min at 25°C in 20 mM MOPS-NaOH/20 mM Na$OJ 0.16 mM EDTA, pH 7.5. It was then incorporated into liposomes as described under Experimental Procedures. The proton influx experiments were done at 12°C. UCP concentration in proteoliposomes was 40-62 &ml. Valinomycin was added after the addition of fatty acids and acyl CoA, and proton influx was measured as change in external pH. Data are expressed as mean f S.E. n=3. When phenylgloxal, the

purified

UCP, it

the

present

experiments

prior

previously separation

present

study

conductance

binding

and stearic the

acid

activation

(Table

1).

retained

was not

of the

acyl

reagent

conductance their

quite

The stimulatory

and

on H' affirmed

sites.

stimulation

as great

effect

of GDP, was completely

similarity

of the

here

of H' conductance

and proton

In the

of H'

as with

the

of palmitoyl

obliterated

In

phenylglyoxal

As shown

effect

with

(8).

CoA,

like

by phenylglyoxal

CoA and GDP sites

of interaction

with

UCP.

nucleotide BATM.

Whereas,

In the

which

to the the

appeared

present

GDP for

of fatty

binding

CoA esters

site

show the

UCP have of action

to compete

purified

whereas,

palmitoyl

been

reported

fatty

acids

with

{Table

interaction

to the

(1,15) and their

acids

experiments

binding

has no effect, the

was no direct

was incubated GDP binding

with

liposome.

nucleotide

The effects

time

in the

of the

effect the

protein

there

groups, block

UCP was modified

of GDP inhibition

palmitic

inhibitory

2) the

arginine completely

prevention

protein

indicating the

(8),

with

the

although

unmodified the

of the

reported However,

reacts

to almost

(Table

to incorporation

influx. the

which was found

CoA thioesters only

of fatty UCP.

the

was somewhat

GDP and ATP for 3) results

with

have

acids

binding been

CoA significantly

use of

the

UCP (17).

for

that

inhibits

isolated

to the

CoA esters

evident

on

unclear,

obtained

and acyl

It is clearly

(16,17)

the with

acyl first [3H]

palmitic

GDP binding

acid to

protein.

DISCUSSION These tion

experiments

of a purified/reconstituted

provide

the

first

unambiguous

UCP by both 1108

fatty

results acids

on the

and acyl

regula-

CoA esters.

Vol.

175,

No.

3, 1991

BIOCHEMICALAND

BIOPHYSICALRESEARCH

TABLE EFFECT OF PALMITIC ACID CAPACITY OF PURIFIED

3

AND PALMITOYL RECONSTITUTED

Addition

COMMUNICATIONS

CoA ON THE GDP BINDING UNCOUPLING PROTEIN

GDP binding nmol/mg protein

Palmitic

acid

20

ti FM

6.8 7.1 6.8 6.9

1:

5

K

I5

Palmitoyl-CoA

20

PM 30 Ml

% of

control

reports

were

of GDP inhibitable ing

effects

retained

the

of

fatty moreover,

For

a number

effective

mitochondria liposome

system

carrier

(21),

detected labeling

the

Because

was not

reagent,

effect

of

different triton

than

was with

interact

with

ADP/ATP

GDP binding

transferase The present

site

on the

from

the

an acyl

to covalently

not

distinct

vesicles

provide

present

and continuing the

evidence

as was shown

for

further

binding

nucleotides

conductance

and fatty

evidence and acyl

acid 1109

binding

was

ADP/ATP the

by the

the Fiol

properties

with

purification

CoA effect that

two

detergent

beginning the

acyl

in the

of the

BATM (17).

in the triton

ADP/ATP

of BATM with

scheme,

is

preserved esters and Bieber

of carnitine

or octylglucoside. for

a common regulatory

CoA esters site

a

photo-

to the

generated

is

in either

in

However,

nature

only

differences

purified

were

isolated

CoA like bind

there

but

in

The difference

upon the

on

UCP to the

(22).

CoA (4).

sites

on GDP binding

by extraction

extraction

carrier, sites

results

UCP for

proton

recently,

In the

chromatography, the

of the

conditions

are

reconstituted

More

to be dependent

octylglucoside

have demonstrated

palmitoyl

homology

to palmitoyl

BATM

CoA esters

carrier

that

modulat-

isolated

acyl

CoA effect

specific

the

transport

an acyl

and purification.

hydroxylapatite

as it

close

UCP except

at separate

that

UCP in reconstituted

appears

extraction rather

ADP/ATP

CoA according 10 fl.

that in

that

purified

UCP prepared

respond

CoA on the

procedures

for

known nucleotide

ACT CoA, was demonstrated

not

observed

to interact

has been

of the

of the

now apparent,

of adenine

BATM (16,17).

did acyl

is

appear

surprising

purified/reconstituted

octylglucoside

(23)

and with

and or UCP in BATM under

partially

with

it

activities

CoA esters

ligands

of years

in isolated

carrier

used

the

(20). it

It

and acyl

as inhibitors

(18,19)

on all

conductance.

acids

and, UCP.

equally

in disagreement

proton

96 86 58 26

4:o 1.8

Purified UCP (5-25 w protein) in octylglucoside containing phosphatidyl choline (5 mg/mg protein) was preincubated with palmitic acid/palmitoyl for 25 min at 25°C. r3H]-GDP binding capacity of UCP was determined to method described under Experimental Procedures. GDP concentration

Previous

99 102 99 100

(s).

which

is

separate

Phenylglyoxal

Vol.

175,

which

No.

completely

arginine

inhibits

group

ductance

(8),

by acyl

was postulated compete long It

chain

only

acyl

the

with of the

acyl

nucleotide

which by the

acyl

CoA.

the

acyl

CoA can alter

present

open

the

this

question.

to the

directly

Based

(25),

binding acids

mediated

the

porposed

proposed

and liver ADP/ATP

protein

that

this

mitochondria carrier

and the

will

the

(26,27),

and that protein

alone

stimulate and is with

which

then

to answer

of the fatty

binding

by the

UCP

and GDP

that

nucleotide

acids site.

UCP, Garlid

may represent of an uncoupling was postulated

indicating

H'

UCP to further

activators

channel

they

no nucleo-

modification

a report

if

may be some endogenous

contention

at the

a

(24).

conditions,

be required

transport is

site

of the

phenylglyoxal

interest

to

that

to be the major

of chloride

was sufficient

purification

conformation

and not

it

to the

be considered studies

with

conductance

Of related the

UCP during also

carrier,

present there

that

of H' con-

ADP/ATP

ligand

the

is

UCP seem to support

recently

in muscle this

might

examination

site.

through between

to the

experiments

the

acting

Additional

been

purified

on a thorough

fatty ities

have

on proton

co-workers acid

acids

It

port.

The present

(1,151. binding

bound

H' conductance

Fatty

act

remains

an essential

binding

CoA esters

One possibility

displaced tide

atractylate

why under

COMMUNICATIONS

activation the

CoA molecule

to anchor

to explain

UCP.

the with

on the

at the

GDP binding,

RESEARCH

by blocking

to prevent

on studies

group

was required

difficult

interfere

shown

adenine

BIOPHYSICAL

presumably

Based

nucleotides

group

however,

conductance

now also

CoA esters.

adenine

AND

GDP binding,

is

that

with

is,

BIOCHEMICAL

3, 1991

the

and fatty

effect

by

to be

additional

similar-

UCP.

Acknowledamentr The expert technical assistance appreciated. This work was supported

of Mr. Neal S. Rhutasel by NIH Grant DK 32686.

is

gratefully

REFERENCES

1. 2. 3. 4.

Nicholls, D.G. and Locke, R.M. (1984) Physiol. Rev. 64, l-64. Shrago, E. and Strieleman, P.J. (1987) in World Review of Nutrition Dietetics (8ourne, G.H., ed.) vol. 53, pp. 171-217, Karger Medical co., Switzerland. Strieleman, P.J., Schalinske, K.L. and Shrago, E. (1985) Biochem. Biophys. Res. Commun. 127. 509-516. Strieleman, P.J., Schalinske, K.L. and Shrago, E. (1985) J. Biol.

and Publ.

Chem.

260, 13402-13405.

i: 7. 8. 9.

Klingenberg, M. and Winkler, E. (1985) EMBO. J. 4, 3087-3092. Klingenberg, M. and Winkler, E. (1986) Meth. Enzymol. 127, 772-779. Shrago, E., McTigue, J., Katiyar, S. and Woldegiorgis, G. (1989) In Hormones; Thermogenesis and Obesity. H. Lardy and F. Stratman Eds. Elsevier Sci. Pub. Co. Inc., pp. 129-136. Katiyar, S. and Shrago, E. (1989) Proc. Natl. Acad. Sci. USA 86, 2559-2562. Cannon, B. and Lindberg, 0. (1979) Meth. Enzymol. 55, 65-78.

1110

Vol.

10. 11. 2 14. 15.

:;: 18. 19. 20. 2: 23. 24.

27.

175,

No.

3, 1991

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

Lin, C.S. and Klingenberg, M. (1982) Biochemistry 21, 2950-2956. Holloway, P.W. (1973) Anal. Biochem. 53, 304-308. Thomas, J.O. and Kornberg, R.D. (1978) Meth. in Cell Biol. 28, 429-440. Peterson, G.L. (1977) Anal. Biochem. 83, 346-356. Jesek, P., Drahota, Z., and Ring, K. (1990) J. Lipid Med. 2, 85-94. Rial, E. and Nicholls, D.G. (1989) in Anion Carriers of Mitochondrial Membranes (Azzi, A. et al Eds.) pp. 261-268, Springer-Verlag, Berlin Heidelberg. Cannon, B., Sundin, U. and Romert, L. (1977) FEBS Lett. 74, 43-46. Strieleman, P.J. and Shrago, E. (1985) Am. J. Physiol. 248, E699-E706. Shug, A., Lerner, E., Elson, C. and Shrago, E. (1971) Biochem. Biophys. Res. Commun. 43, 557-563. Pande, S.V. and Blanchaer, M.C. (1971) J. Biol. Chem. 246, 402-411. Woldegiorgis, G., Shrago, E., Gipp, J. and Yatvin, M. (1981) J. Biol. Chem. 256, 12297-12300. Aquila, H., Link, T.A. and Klingenberg, M. (1987) FEBS Lett. 212, l-9. Woldegiorgis, G., Duff, T., Contreras, L., Shrago, E. and Ruoho, A.E. (1989) Biochem. Biophys. Commun. 161, 502-507. Fiol, C.J. and Bieber, L.L. (1988) Lipids 23, 120-125. Shrago, E., Shug, A., Elson, C. and Lerner, E. (1972) in the Role of Membranes in Metabolic Regulation. M. Mehlman and R. Hanson, Eds Academic Press, N.Y., pp. 165-182. Jezek, P., and Garlid, K.D. (1990) J. Biol. Chem. 265, 19303-19311. Andreyev, A.V., Bondareva, T.O., Dedukhova, V.I., Mokhova, E.N., Skulachev, V.P., Tsofina, L.M., Volkov, N.I. and Vygodina, T.V. (1989) Eur. J. Biochem. 182, 585-592. Schonfeld, P. (1990) FEBS Lett. 2, 246-248.

1111