• No results found

The rat hepatocyte plasma membrane organic anion binding protein is immunologically related to the mitochondrial F1 adenosine triphosphatase beta subunit

N/A
N/A
Protected

Academic year: 2020

Share "The rat hepatocyte plasma membrane organic anion binding protein is immunologically related to the mitochondrial F1 adenosine triphosphatase beta subunit"

Copied!
9
0
0

Loading.... (view fulltext now)

Full text

(1)

The rat hepatocyte plasma membrane organic

anion binding protein is immunologically

related to the mitochondrial F1 adenosine

triphosphatase beta-subunit.

T Goeser, … , R D Burk, A W Wolkoff

J Clin Invest.

1990;

86(1)

:220-227.

https://doi.org/10.1172/JCI114687

.

A 55-kD organic anion binding protein (OABP) was identified previously in liver cell plasma

membrane sinusoidal subfractions. Although this protein was localized to the surface of

hepatocytes by immunofluorescence, immunoblot analysis revealed reactivity toward both

plasma membrane and mitochondrial fractions. To clarify these findings, an immunoreactive

clone from a rat liver cDNA expression library was isolated, the 1,500-base pair cDNA insert

was sequenced, and the corresponding beta-galactosidase fusion protein was expressed

and purified. The resulting sequence corresponded to that of the rat mitochondrial

F1-adenosine triphosphatase (F1-ATPase) beta-subunit. This protein and OABP are of similar

size and are mutually immunologically cross-reactive. That the antigen was present on the

cell surface as well as in mitochondria was suggested from studies of immunoprecipitation

after cell-surface iodination, and light- and electron-microscopic immunocytochemistry.

Photoaffinity labeling of bovine F1-ATPase with high-specific-activity

[35S]sulfobromophthalein revealed binding only to the beta-subunit. Hepatocyte uptake of

bilirubin and sulfobromophthalein requires cellular ATP and mitochondria also transport

these organic anions, which at high doses inhibit respiration. The presence of an organic

anion binding site on the F1-ATPase beta-subunit suggests that it may play a role in these

processes.

Research Article

Find the latest version:

(2)

The Rat

Hepatocyte

Plasma Membrane Organic Anion Binding

Protein

Is Immunologically Related

to

the

Mitochondrial

F1

Adenosine Triphosphatase fl-Subunit

TobiasGoeser,* Ryo Nakata,* LawrenceF.Braly,*Alexander Sosiak,* Celeste G. Campbell,*RolfDermietzel,*

Phyllis M.

Novikoff,1

Richard J. Stockert,*II Robert D.

Burk,'***

and Allan W. Wolkoff*

Liver Research Center, Departments ofMedicine,*Pathology,§Biochemistry,11 Pediatrics,' Microbiologyand Immunology,**

andObstetrics and Gynecology,#Albert Einstein College ofMedicine,Bronx, NewYork10461;and

tAnatomical Institute, University ofRegensburg, Regensburg, Federal Republic ofGermanyD-8400

Abstract

A55-kD organic anion binding protein

(OABP)

was

identified

previously

in liver cell plasma membrane sinusoidal

subfrac-tions.

Although

this

protein

was localized to the surface

of

hepatocytes

by immunofluorescence, immunoblot analysis re-vealed

reactivity

toward both plasma membrane and

mito-chondrial fractions.

To

clarify

thesefindings, an

immunoreac-tive clone from

arat

liver

cDNA

expression library

was

iso-lated, the

1,500-base pair

cDNA insert was

sequenced,

and the

corresponding j-galactosidase

fusion

protein

was

expressed

and

purified.

The

resulting sequence corresponded

tothat

of

the rat

mitochondrial

Fl-adenosine

triphosphatase

(Fl-ATP-ase)

,-subunit.

This

protein

and OABP

areof similar size and are

mutually immunologically cross-reactive.

Thatthe

antigen

was

present

onthe

cell surface

as well as in

mitochondria

was

suggested

from

studies of immunoprecipitation

after

cell-sur-face iodination,

and

light-

and electron-microscopic

immuno-cytochemistry.

Photoaffinity

labeling of

bovine

FI-ATPase

with

high-specific-activity

V5Slsulfobromophthalein

revealed

binding only

tothe

#-subunit.

Hepatocyte

uptake

of

bilirubin

and

sulfobromophthalein requires

cellular

ATP and

mitochon-dria also transport these

organic

anions,

whichat

high

doses

inhibit respiration.

The

presence of

an

organic anion binding

site on the

F,-ATPase

#-subunit

suggests that

it

may

play

a

role in these

processes. (J. Clin.

Invest.

1990.

86:220-227.)

Key words: bilirubin *

liver

-

photolabeling

-

sulfobromophtha-lein

* transport

Introduction

The

organic

anions bilirubin

and

sulfobromophthalein

(BSP)'

circulate

tightly

boundtoalbumin

(1, 2).

The

avidity

of

albu-min

for these compounds is such that,

at

equimolar

concen-trations,

<0.1%

of

ligand

remains unboundatequilibrium

(3).

Despite

the

fact

that these

organic

anions

are

protein bound,

Address

reprint

requests to Dr. AllanWolkoff, Liver Research Center, Albert Einstein College of Medicine,1300MorrisParkAvenue,Bronx,

NY 10461.

Receivedfor publication20November 1989 andinrevisedform21

February1990.

1.Abbreviations usedinthis paper: BSP,sulfobromophthalein;OABP,

organicanionbindingprotein.

they are

rapidly extracted from their protein

carrier

by

the liver

(2). Previous studies revealed

that

hepatocyte

uptake

of

biliru-bin

and

BSP had

carrier-mediated

kinetics (1, 4).

In

particular,

recent studies performed

in

short-term cultured

rat hepato-cytes

revealed that this uptake

process was

saturable,

tempera-ture

sensitive, inhibited after

depletion

of

cellular

ATP,

and

dependent

upon

the

presence of C1-

in

medium (5, 6).

We

identified

a 55-kD

protein in sinusoidal

surface-de-rived

subfractions

of

liver cell plasma

membrane

as a

putative

organic anion

transporter that

bound photoactivated

[35S]BSP

(7). Subsequently,

we

purified this organic anion binding

pro-tein

(OABP) from

deoxycholate-solubilized

rat liver cell

plasma membrane by using BSP-Sepharose affinity

chroma-tography (7). Through

the use

of

monospecific antibody,

OABP

was

localized

tothe

surface of hepatocytes

by

immuno-fluorescence (8).

However,

immunoblot analysis

revealed

re-activity towards both

plasma

membrane

and

mitochondrial

fractions (8).

To

clarify

these

findings,

we have

screened

a rat

liver

cDNA

expression library with this antibody.

DNA

inserts

of isolated clones

were

sequenced and expressed

as

f3-galacto-sidase fusion

proteins.

The cDNA sequence was

identical

to

the sequence of

the

3'

1,500

base

pairs of the ,8-subunit of

rat

mitochondrial F.-adenosine triphosphatase (F.-ATPase).

These proteins

are

of similar size

and are mutually

immuno-logically cross-reactive.

We present data

suggesting

that this

antigen resides

on

both the

plasma membrane

and the

inner

mitochondrial

membrane

of

the

hepatocyte.

Methods

Ratliver cDNAlibraryscreening. Aratliver

Xgt

1IcDNAexpression

library(Clontech Laboratories, Inc.,PaloAlto, CA)wasscreened for OABPexpressionby the method of Young and Davis(9).Inbrief,goat

anti-rat OABPantiserum, diluted 1:100in50mM Tris, pH7.8, 15 mM

NaCl

containing3%(wt/vol)bovineserumalbumin(BSA)was usedtoidentify OABP-expressingrecombinants. Nitrocellulose filters

werewashedandblocked with 3% (wt/vol) BSA and 0.3% (wt/vol) dry milk inTris-buffered salineat4VC.Positive plaqueswerevisualized using rabbitanti-goat IgG horseradishperoxidase conjugate (Sigma

ChemicalCo., St. Louis, MO) (1:400)followedby incubation with

4-chloro- I-naptholperoxidasesubstratesolution(Kirkegaard&Perry Laboratories, Inc., Gaithersburg, MD)accordingtothe manufacturer's directions. Positiveplaqueswererecoveredandpurified by rescreening

atlowerphage density.

Analysis ofclones. Positive plaques were liquid amplified and phageDNA waspurified. Insertsizewasdetermined afterEcoRI

di-gestionand agarose gelelectrophoresis. RecombinantDNA was di-gested with EcoRI, andthecDNA insertwas gel purified and

sub-cloned into the EcoRI site of pGEM2 plasmid (Promega Biotech, Madison, WI) for

restriction

analysesandsequencing.

Restriction endonuclease digestionswereperformed withBamHI,

Bgl II, EcoRI, EcoRV, Hind III, Kpm I, Pst

I,

Pvu

II,

and Sph I J.Clin.Invest.

©TheAmericanSocietyfor ClinicalInvestigation,Inc.

0021-9738/90/07/0220/08

$2.00

(3)

(Boehringer Mannheim Diagnostics,Inc.,Houston, TX) according to

themanufacturer's instructions. For sequencedetermination, restric-tion fragmentsof the OABP cDNA weresubclonedinto pGEM 3Z. The DNAsequence was determined from double-stranded DNA using

SP6 andT7primers by the dideoxy chain terminationprocedureusing [32P]dATP or [35S]dATP(Amersham Corp., Arlington Heights, IL) and a DNA sequencingkit(PromegaBiotec).

Preparation of recombinant fusion protein in lysogenic bacteria. Escherichia coli strain Y1089, infected with the recombinant encoding the proteinrecognized by goat anti-rat OABP antiserum, was grown

overnight in lysogenic bacteria media supplemented with 10 mM

CaCl2at320C. Aliquotsof theculture, sufficient to yield 1,000

colo-niesperdish,werestreakedon150-mmagarplates.Randomly selected colonieswereincubatedovernightat320Cor420C. Coloniesthat grew at 320Cbutnot at 420C werepresumedtobelysogenic.

Lysogenic colonieswere grown in Luria-Bertani media at 320C until OD260was0.6. Incubation wascontinued for20 min at420C.

Isopropyl

fl-D-thiogalactopyranoside

wasadded to a final

concentra-tion of 4.2MMandincubationwascontinuedfor 1 h at 37°C. Bacteria wereharvested bycentrifugationat 4,000 g for 5minin a rotor (model

GSA, DuPont-Sorvall,Newtown, CT). The pellet was resuspended in

20 mM sodium phosphate, pH 7.4, then frozen in liquid nitrogen,

thawed, sonicated (model W140 Sonifiercelldisrupter,Heat

Systems-Ultrasonics,Inc.,Plainview, NY), and refrozen untilused.

To purifythe recombinant protein, the thawed pellet was then washedwith2 mlof1%Triton X-100(Sigma Chemical Co.) in 0.02 M

sodium phosphate buffer (pH 7.4) followed by 2 mlof 1% sodium

deoxycholate (United States Biochemical Corp., Cleveland, OH) in0.2 Msodiumphosphate (pH7.4).The pellets were resuspended in 10%

glycerol, 5% mercaptoethanol, 2.3% SDS, 0.0625 M Tris (pH 6.8)

(sample buffer),andsubjectedto 7.5%SDS-PAGE.Therecombinant

fusionprotein bandat160kD wasexcised and electroeluted fromthe

gel (10).

Antibodystudies. A female New Zealand rabbit (Hazleton Re-search Products, Inc., Denver, PA)was injected subcutaneously, at

1-mo intervals, with 6-40

Mig

of electroeluted fusion protein(8). To

determine loss ofantibody activity after absorption with OABPor fusionprotein, 10mlof 1:1000 dilution ofrabbit anti-ratOABP serum oranti-fusionproteinserumin5% BSA, 50mMTris, 0. 15M NaCl,

pH7.6, wasabsorbedat40Covernight with - 5-10Migofpurified OABPorfusionprotein. Residual antibody activitywasassessedby immunoblotanalysis (8).

Northern blotanalysis.Rat livertotal,poly A(+),andpolyA(-)

mRNA wasprepared, electrophoresedon 1% agarose, andtransferred

toGene Screen(NewEngland Nuclear, Boston, MA) (1 1).Northern

blotanalysiswasperformedafter32p labelingof the 1,250-base pair

cDNA insert by primerextension using [32P]dCTP (Amersham Corp.)(11).

Cultureofisolatedrathepatocytes. Rathepatocyteswereisolated and culturedinmonolayeratdensities of 1.8X 106/60-mm dish in

modified Waymouth's 752/1 medium containing 5% fetal bovine

serum(6, 12). The mediumwaschanged2hafterplatingand cultures

wereused for studies 16hafterplating.

Surfaceiodinationofshort-term culturedrat

hepatocytes.

Cultured cellswerewashedtwice with ice-cold 20mMPBS, pH 7.4,and incu-bated in1 mlof PBScontaining25mMglucose, 100Mgof lactoperox-idase (8 U, Sigma Chemical Co.), 50 Mg of glucose oxidase (10 U,

Calbiochem-Behring Corp., San Diego, CA)and 500MCi of1251-Na

(13-17

mCi/Mg

I, Amersham Corp.) for45 minat4°C. Cellswerethen washed three timesin 2 ml of ice-cold PBS, anda finalwash was performed with2 mlof ice-coldPBScontainingKI(1 Mig/ml).

Immunoprecipitation. Surface-labeled cells were lysedin 1 mlof buffer containing0.1%ratserum albumin, 1%Nonidet P40, 1 mM

EGTA, 10 mM Tris, pH 7.6, 2 mM phenylmethylsulfonyl fluoride (PMSF), and 150mMNaCl. Celllysateswerecentrifugedfor2 min in an Eppendorf centrifuge (model 5414, BrinkmannInstrumentsCo.,

Westbury, NY) and the resultingsupernatantwasincubated for 30 minat4VCwithconstantmixingwith 50 Ml of 20% Protein

A-Sepha-rose(Sigma ChemicalCo.) suspended in PBS. Beads were removed by

centrifugation,the supernatant wasincubatedwith 10 Ml of nonim-mune rabbit serum for 30 min at 40C, and 50 Ml of20% Protein

A-Sepharosewasadded for30 min at40C. After centrifugationfor 2 min at40C, thesupernatant wasincubated with anti-OABPor

anti-fusionproteinserum(10Ml) for 3h at40C.Antigen-antibody com-plexes were recovered by adsorption to 50Mlof 20% Protein A-Sepha-rosefor30 min at40Cfollowed bycentrifugationfor 2 min. Protein

A-Sepharose beadswere thenwashedfive times with 1 ml of 10 mM

Tris,pH 7.2, 150mM NaCl, 1% Triton X-100, 1%sodium deoxycho-late, and0.1% SDS (13)and oncewith 0.05 MTris-saline,pH 7.4. The

antigen-antibodycomplexes were removedfromthe Protein A-Sepha-rose beads and denatured by boiling in SDS-PAGE sample buffer

containing5%fl-mercaptoethanolfor3 min. TheproteinA-Sepharose

beadswereremoved bycentrifugationandthesupernatant was

sub-jected to 10% SDS-PAGE. Fixed gels weredried and subjected to

radioautography (13).

Photoaffinity

labeling of

F,-ATPase

with

[35S]BSp.

[35S]BSP

at a specific activity of 3,400 mCi/mmolwasprepared by sulfonation of phenoltetrabromophthalein withH235SO4 (14). Purified bovine

Fl-ATPase was agift fromDr. HarveyPenefsky, SUNY Health Science

Center, Syracuse, NY). Approximately 23Mgofproteinwas photoaf-finity-labeledwith 22pmol of

[35S]BSP

in 100Mlof PBS, pH 7.4, as

previously described(7). Analiquot of this mixturewassubjected to

SDS-PAGEandfluorography.

Preparation ofliver tissuefor immunocytochemistry. Male

Sprague-Dawleyrats(250-300 g) (Marland Farms, Hewitt, NJ)were

anesthetized by ether, and the liverswerefixed either by perfusionor

immersion at 4°C with 4% paraformaldehyde in 0.1 M phosphate

buffer, pH 7.4,7.5% sucroseforatotalof4 h(15). Afteranovernight rinse in cold0.1 Mphosphate buffer,pH 7.4, 7.5%sucrose,nonfrozen sections- 30Mmthickwereprepared withanOxford Vibratome (Ted

Pella,Tustin, CA), usingaslowspeedandhigh amplitude setting,and placed in cold 7.5% sucrose. Sections wereexposed to rabbit anti-OABPserum at4°C for20-40 h, withintermittent,mildagitation. Single sections (-2mm2)wereplacedintissuecultureU-shaped

mul-tiwellplastictrays(Linbro,FlowLaboratories,Inc., McLean, VA)that

contained 0.2mlof antiserum diluted 1:100 with PBScontaining1%

BSA. After several rinsesin coldPBS,thesectionsweretreated with Protein A-horseradishperoxidase (0.1 ml/ml) (E.Y.Laboratories, San

Mateo, CA) for 1 h at room temperature in the dark. After several rinses in PBS and in 7.5% sucrose, the sections were incubated in

3,3'-diaminobenzidine tetrahydrochloride (SigmaChemicalCo.)atpH 7.6 (15)at roomtemperaturefor 10 mininthedarktoreveal peroxi-daseactivity,andexamined bylight microscopy.Somesectionswere processed for electron microscopyaspreviously described(15).

Immunogoldlocalization studies. Immunogold-labelingwas per-formedonLR-White (LondonResin,Ltd.,Hampshire, England)

ul-trathin sections. Fixation was done by vascular perfusion with 2%

paraformaldehydeplus 0.1% glutaraldehyde in PBS (pH 7.4) for5 min,

followed byimmersingdissectedcubesof liverin PBSat4°Cfor12h.

Thetissuewas thenrinsedinPBSat4°Covernight, dehydratedina graded series of ethanol concentrations,and immersed in 100% LR-White; thesamples were encapsulated in resinandpolymerized by irradiationusingablacklightsource(UVlight,282nm)at4°C.

Suit-ableconsistencyof the blockswasobtained after 3-4 d of

polymeriza-tion.Ultrathin sectionswere cutbymeansofanUltratome

(Reichert

Scientific Instruments, Buffalo, NY),mountedon

Formavar/Carbon-coated Cugrids,andfurtherprocessedforimmunolabeling. Labeling

with theprimary antibodywasachieved byusinga 10-Mldroplet of polyclonal anti-OABP (1:50). Affinity-purified goat anti-rabbit

IgG

coupled tocolloidal gold(5-6 nm, Janssen Pharmaceutica, Beerse,

Belgium)wasusedasasecond marker.

Results

(4)

froma Xgtl 1 cDNAratliver library werescreenedwithgoat

anti-rat OABP antibody, and six positive plaqueswere identi-fied. These plaques were purified by three additional cycles of plating and screening with the same antibody. DNA was puri-fied from these phage, digested with EcoRI, and subjected to agarose gel electrophoresis. The DNA patterns from the six recombinant DNAs wereidentical, consisting of two bands of

' 1,250 and 300 base pairs, indicating the presence of an

internal

EcoRI site. The two EcoRI

insert

DNA fragments from one clone were subcloned into pGEM2.

Immunologic characterization of the recombinant protein

in E.

coli. Lysogenic

E.

coli

were

induced

with isopropyl

f3-D-thiogalactopyranoside to produce the

f3-galactosidase-OABP

fusionprotein. This recombinant protein was virtually insolu-ble in detergent at

4VC,

and waspurified by preparative 7.5% SDS-PAGE after extraction of thelysate sequentially with 1% Triton X-100 and 1% deoxycholate (Fig. 1, left). Immunoblot

analysis

of the E. coli

lysate

revealed

the

presence

of

an

- 160-kD bandreacting with both antibody to

fl-galactosidase

and

antibody

to

OABP

(Fig. 1, right). Nontransformed

E.

coli

lacked

this immunoreactivity (Fig.

1,

right).

Immunoblot analysis using antibody to the purified

recom-A B C

A

"MM,F

B

200

-95

-68

-43

-

200-97

-

68-

43-26

-

14-Figure 1.(Left) Expressionof recombinant

protein

inE.coli.E.coli strain Y1089wastransfected withXgtl1phage

coding

for immun-oactiveOABP,and

expression

of the

protein

as a

hybrid

with bacte-rial

fl-galactosidase

wasinducedasdescribed in Methods.This re-combinantproteinwasvirtuallyinsoluble indetergentat4VC,and

waspurifiedbypreparative7.5%SDS-PAGE afterextraction of the bacterial lysatesequentiallywith 1% TritonX-100and1%

deoxycho-late.LaneA,Coomassie-stainedSDS-PAGE of the bacteriallysate.

Lane B,Coomassie-stained SDS-PAGE of thepurifiedfusion pro-tein. Thispreparationwasusedtoimmunize rabbits.

(Right)

Immu-noblotanalysis of the recombinant protein.E.colilysatewas

sub-jectedto 10%

SDS-PAGE,

transferredtonitrocellulose,andprobed

witheitherantibodyto E.coli

fl-galactosidase

(lane A)oranti-OABP (laneC).Lane Brepresentsalysate from nontransformedE. coli

probed with anti-OABP. These results indicateasingleproteinwith

immunoreactivityto

fl-galactosidase

andOABP. The lower band in laneApresumably representssynthesisofanincomplete

fl-galactosi-daseprotein.Molecularmassmarkersatleft in bothpanelsarein ki-lodaltons.

A B C

-200

- 96

iMm

Figure 2.Immunologic identity of - 68 OABP and the recombinant protein.

Threeimmunoblots performed againstpurifiedOABPareshown. - 43 Lane A,reaction of anti-fusion

pro-teinantibodywith purifiedratOABP.

-36 Lanes B and C, antibody reactivity was markedly reducedafter preab-- 29 sorption with OABP (lane B) or

fu-sionprotein(lane C). Similar results wereobtained with the original

anti-- 18 OABPantibody. Molecular mass markersatright are in kilodaltons.

binant fusion protein revealed a similarly sized 55-kD band with purified OABP (Fig. 2) and rat liver homogenate (data not

shown). This immunoreactivity

was

markedly

reduced after preabsorption of the antibodywitheitherOABP orthe

fusion

protein (Fig. 2).

Similar results were obtained with the original anti-OABP antibody (datanotshown).

Nucleotide

sequence

and

deduced

amino

acid

sequence

of

OABP

cDNA. A

partial

restriction

map of the 1500-base

pair

Xgtl

1

cDNA

insert is shown in

Fig.

3. The five DNA

fragments

labeled

a-e were

subcloned

into either

pGEM-2 (b, e),

pGEM-3Z (c, d),

or

pGEM-3 (a).

Inaddition thetwoEcoRI

restriction fragments

were

subcloned

into

pGEM-2.

These

subclones

were

sequenced using

the strategy shown in

Fig.

3.

Computer analysis

of the DNA sequence revealed

identity

with bases 63-1558

of

the

published

sequence of the 13-subunit

of

rat

mitochondrial

FI-ATPase

(16).

Immunological identity

of

OABP

and

the

1-subunit

of

mi-tochondrial F,-ATPase. Rabbit

antibody

raised

against

the

13-subunit of yeast F1-ATPase

provided by

Dr.Michael

Douglas

(University

of

Texas, Dallas)

wasused for immunoblot

analy-(a)

l

l

(b)

(c)

(d)

(.)

P 11

1'

P1

T

T

(X34 -X

I-

(-H

(-HI

0.5 kb

Figure 3. Restriction map andsequencingstrategy of OABP cDNA.

Thefivefragmentslabeleda-e weresubcloned and inserted into

ei-therpGEM-2(b,e),pGEM-3Z (c,

d),

orpGEM-3 (a).Inaddition,

thetwoEcoRI restrictionfragmentsweresubcloned intopGEM-2.

(5)

sis of rat liver OABP, bovine heart F1-ATPase, and rat liver homogenate.A single immunoreactive protein bandwas de-tected (Fig. 4). Identical results were obtained when anti-rat liver OABP antibody was used (data not shown).

Northern blot analysis.

As seen in

Fig.

5, a

single

mRNA

of

- 1.7 kb hybridized with the 1,250-base pair [32P]cDNA.

Immunoprecipitation with surface

iodinated

rat

hepato-cytes. The cell surface proteins of overnight cultures of rat hepatocytes were selectively radiolabeled with

1251

using

solu-ble lactoperoxidase/glucose oxidase at 40C for 45 min. A

major band

on SDS-PAGE migrating at 55 kD (Fig. 6) was

immunoprecipitated

with antibody against the recombinant protein. This band was not seen after immunoprecipitation with nonimmune serum or with antibody that had been

preabsorbed

with recombinant protein (Fig. 6).

Immunocytochemistry studies.

To

directly visualize

anti-genic

localization, sections of rat liver wereexamined using

immunocytochemistry.

Bound antibody was detected by im-munoperoxidase staining. As seen in Fig. 7, by light micros-copy, anti-OABP antibody reacted with the sinusoidal and

basolateral

cell surface but not with the canalicular membrane. Diffusegranular cytoplasmic staining, compatible with a mi-tochondrialsubcellular localization,wasalso seen.

Immunore-activity

was not seen with nonimmune serum or with

anti-OABP

that

had

been

preabsorbed

with the recombinant

pro-tein. By

electron

microscopy,

reaction

product

wasclearly seen on

sinusoidal and basolateral

butnot canalicular surfaces

of

hepatocytes (Fig. 8). Visualization

ofimmunoreactivityin

mi-tochondria

was

difficult. However, using

immunogold

meth-odology, antigen

was

clearly

detected within mitochondria

(Fig.

9).

Photoaffinity

labeling

of

F,-A

TPase

with

[35S]BSP.

Bovine

cardiac

F1-ATPase was

photoaffinity-labeled

with

high-spe-cific-activity

[35S]BSP. The labeled material

wasthen

subjected

to 15%

SDS-PAGE

and

fluorography.

These studies

revealed

that

[35S]BSP bound only

to the fl-subunit

(Fig. 10).

Discussion

In

previous studies,

we

found high-affinity binding of [35S]BSP

toa

sinusoidal surface-derived

ratliver cell

plasma

membrane

subfraction (7). Binding

was

eliminated by preincubation of

membrane

with

trypsin. Photoaffinity labeling of this

mem-brane

subfraction with [35S]BSP revealed

a

predominant

55-kD

protein,

and

this

was

purified

to

high

degree by

BSP-Sepharose affinity chromatography

(7).

Antibody raised

to

this

OABP

reacted

with the

surface of short-term cultured

rat

he-patocytes as

determined

by

immunofluorescence

(8).

A B C

-955 Figure 4.Immunologiccross-reactivity of OABP and the mitochondrialF,-ATPase fi-subunit.Antibodytothe#-subunitofyeast 43 F,-ATPasewasprovidedbyDr.Michael

-36

Douglas.

Itwasusedto

probe

by

immuno-blot OABP (lane A), bovine

F,-ATPase

(pro--29 vided by Dr. Harvey Panefsky, lane B), and ratliverhomogenate (lane C).Asingle band

-18.4 is seenin all threelanes, indicating

cross-re-activitybetween the

fl-subunit

of mitochon-drialF,-ATPaseandOABP.

A

B

28S-I

C Figure 5. Northern blot analysis of OABP cDNA.Ratlivertotal (lane

A), poly A(-) (lane B),andpoly

A(+) (lane C)mRNA was

electro-phoresedon 1%agarose and

trans-ferredtoGene Screen.Northern

blotanalysiswasperformedafter

32P labeling ofthe1,250-base pair

cDNA insert byprimer extension using

[32P]dCTP.

Hybridizationwas

overnight in 50% formamide, 5X SSC, and 2XDenhardt'ssolution. The blotwas washedtwiceat 65°C

with0.1%SDS in 0.1X SSC. A

sin-gle, - 1.7-kbpoly A(+) mRNA

hy-bridized with this probe.

Twoother

candidate organic anion

transport

proteins

have been

identified

(17, 18). Their relationship

to OABP is not

clear.

Antibody

to

the

protein

isolated by Stremmel

and

Berk

(17)

partially inhibited BSP uptake by isolated

hepatocytes. However,large amounts

of antibody

were necessary, and up-take was

determined only

at

high

concentrations of BSP in the

absence

of albumin.

It

is

questionable

whether

a

physiologic,

or a

higher

capacity,

lower

affinity

organic

anion

transporter was

studied

(5, 6). Sottocasa et al. (18)

reconstituted

BSP up-take

in

liposomes

with their candidate transporter.

The

rela-tionship

of this

protein

to

the

other

organic

anion

binding

proteins

and to

physiologic organic anion

transport

is

yet to be

established. Ultimate determination of function of

these

pro-teins

may

await transfection of transport-deficient

cells

with

specific full-length

cDNAs.

In thepresent study, the

antibody

to

purified

OABP was

used

to clone a cDNA

from

a rat

liver

expression library. This

cDNA coded

for

- 80%

of

theapparent

molecular weight of

mature

OABP

and was

expressed in

E.

coli

as a

fusion protein

with

fl-galactosidase.

Antibody

raised to the fusion protein or

antibody

to

native OABP

reacted

with

both

proteins

as

indi-cated

by loss of antibody activity after absorption against

ei-ther

protein. This immunologic cross-reactivity indicates

a

high degree of homology

between the

affinity

purified

OABP and the

fusion

protein encoded

by

the

cDNA.

Computer

anal-ysis of the

cDNA sequence

revealed

identity with

the

pub-lished

cDNA sequence

for the

fl-subunit

of

rat

mitochondrial

F,-ATPase (16). This

F,-ATPase

subunit

and

OABP

were

of

the

same

molecular

mass as

estimated by SDS-PAGE

and were

immunologically cross-reactive (Fig.

4).

A B C

JIM

k Figure 6.Immunoprecipitation of1251

surface-labeledrathepatocytes. Surfaceproteinsof

-97 overnightculturesofrathepatocyteswere ra-diolabeled with1251 usingsoluble

lactoperoxi--68 dase/glucoseoxidaseat4°Cfor45 min.The cell

lysate

wassubjectedto

immunoprecipita--43 tion,and SDSgel electrophoresiswas

per-formed.Immunoprecipitationwith anti-fu-sionproteinantibody revealedamajor55-kD band.Nobandwas seenafter

immunoprecipi--29 tation withantibody thatwaspreabsorbed

with fusionprotein (lane B)orwith

preim-muneserum(lane C). Molecularmass markersatrightarein kilodaltons.

(6)

4V

&cA

~V~J

it<<

T_/+>

A!,

>ps w ?Vl...Jr

X,

;fr./

' W

'

C'.

Figure7.Hepatocytesurface localization of OABP afterabsorption

of

antibody

withrecombinant

protein. (a)

Asection of

aldehyde-fixedratliverwasexposedtoanti-OABP antiseradiluted

1:100,

treated withProtein

A-peroxidase,

andincubated in diaminobenzi-dineat

pH

7.6torevealsites of

peroxidase activity.

OABP

antigen

waslocalizedonlyinhepatocytesandtheir basolateralplasma

mem-branedomains. The

apical plasma

membranedomains

(bile

canalic-uli)

didnotdemonstrate the

antigen.

Punctateintracellular reaction

product,consistent with mitochondriallocalization,wasalsoseen. Arrowsindicate theportionof the

OABP-positive

membranefacing

thesinusoids. X400.

(b)

Sectionexposedtoanti-OABP antisera

ab-sorbed

against

fusion

protein

andprocessedasina.Essentiallylittle or noOABP

antigen

localization is evident inhepatocytebasolateral

plasmamembrane. X400.(c)Sectionexposedtononimmuneserum and

processed

asina.This control section showsnostainingatthe

hepatocyte

basolateralsurfaces. X400.

Although

the

presence

of OABP

in

sinusoidal

rat

liver

cell

subfractions could

be

due

to a

small

amount

of mitochondrial

contamination,

several

lines

of evidence

suggest

both

a

liver

cell

surface

and

inner mitochondrial

localization of

this

anti-gen. By ELISA, OABP was - 14-fold enriched insinusoidal

subfractions in which mitochondrial enzyme markers were

deenriched

(8).

Surface iodination of cultured

hepatocytes

at

4VC and

immunoprecipitation using

anti-OABP

(Fig. 6)

re-sulted

in

specific

recovery

of

a

55-kD

protein

(Fig. 6). These

results

suggesting hepatocyte surface localization of

OABP were confirmed by

immunocytochemistry

at both the

light-and

electron-microscopic

level. In rat

liver,

surface

localiza-tion of

OABP is seen

only

in

hepatocytes.

The

widespread

tissue homogenate distribution of the antigen reported

pre-viously (8), probably

derives from

its presence in mitochon-dria.

The

a-subunit

of F1-ATPase,

an

inner mitochondrial

membrane

protein,

has not been

previously described

on

the

plasma

membrane. The

63-subunit

precursor

is coded from

chromosomal DNA which is

thought

tobe translated on free

polysomes (19). The signal for mitochondrial localization of

this

protein

resides

at

the

NH2 terminus

(20). After

binding,

the

precursor

protein is translocated

into mitochondria

by

an

ATP-dependent

process

during which the

NH2-terminal

-

60

amino

acids

are

cleaved,

producing

the mature

form of the

protein (20-22).

Theseevents

differ from those characterizing

the usual

synthetic pathway

for

plasma

membrane

proteins,

involving

signal recognition protein-dependent synthesis

on

membrane-bound

polysomes (23-25). However, since

the

liver

cell

mitochondrial

membrane area

is

-

25

times that

of

the

plasma membrane

(26), it is

possible

that

a

small

yet

signif-icant

percent

of

mRNA

coding for the

#-subunit

of

F1-ATPase

is translated via the

signal recognition protein-dependent

translocation

mechanism. Alternatively,

there may be

hetero-geneity

in mRNA

coding

for the

signal

sequence

ofthis

protein

atthe 5' translated

region. Such heterogeneity

couldarise from a

mechanism

such as

tissue-specific alternate

splicing

of

RNA, or

alternative

start

sites.

These

possibilities

will

be

the

subject

of future studies. Although Northern blot

analysis

revealed a

single

band upon

hybridization of

rat

liver

mRNA

with

our

cDNA,

microheterogeneity

within this

band must be

consid-ered.

Both the cDNA

obtained by

usand

the

cDNA

obtained

by Garboczi

et

al.

(16) lack the NH2-terminal 200-250

nu-cleotides

as

judged

by

comparison

with the

sequence

of

the

human

F,-ATPase

#-subunit

(27).

The

presence

of identical

or

closely related

proteins

in

mi-tochondria and plasma membrane has

been

suggested in

other systems. Berk et al.

(28) reported

that a rat

hepatic plasma

membrane,

fatty

acid

binding protein and mitochondrial

glu-tamate-oxaloacetate

transaminase

located

in

the

mitochon-drial

matrix,

share a number

ofproperties. They found that the

sequence

of the

24

NH2-terminal

amino acids of

the

fatty

acid

binding protein

are

identical

to the

published

sequence

for

mitochondrial

glutamate-oxaloacetate transaminase.

Mono-specific

polyclonal antisera raised

against

each

of the

proteins

cross-reacted

with the other

protein.

Each

of the antisera

were

reported

to

exhibit

"cytoplasmic" (presumably mitochondrial)

and

plasma

membrane

immunofluorescence

staining.

Cell surface

localization of mitochondrial antigens

has also been

observed in

primary biliary

cirrhosis,

a

liver

disease of

unknown

etiology

characterized by slowly

progressive

destruc-tion of

biliary epithelium

and

signs of increasing cholestasis,

resulting

eventually

in

profound jaundice

and death.

Circulat-ing antibodies that react with mitochondria

(7)

Figure 8.Electron-microscopic immunocytochemical localizationofOABP in nonfrozen sections of aldehyde-fixed ratliver.(Top)Reaction

productispresent atboththesinusoidal(SP)and lateral (LP) hepatocyte plasmamembranebut isabsent from the bile canaliculus(BC). X44,000.(Bottom)Reactionproductis present only in the lateral plasma membrane(LP).No reaction product is detected in the bile

canalicu-lus(BC). X42,000.

Figure 9.Immunogold localization of OABP inratlivermitochondria.Ultramicrotome sections of LR-White ultrathin sections of2%

(8)

A B

-211

-107

-693

Figure

10.

[35S]BSP photoaffinity labeling

of bovine

F1-ATPase.

Approximately23

jtg

of -458 proteinwas

photoaffinity-labeled

with 22

pmol of

[35S]BSP

(3,400

mCi/mmol)

as de-scribed in Methods. SDS-PAGEon a 15%

polyacrylamidegelwasperformedand stained withCoomassie blue (lane A). Bands

corre-sponding

tothea-(upper),

,#-middle),

and y-subunits

(lower)

are seen.Fluorographyof this

gelwasthen

performed

and revealed localiza--18 2 tionof

radioactivity

onlyin the

,B-subunit

(lane B).

studies suggest that the

primary biliary

cirrhosis

specific

anti-gen(s)are inplasma membraneaswellasmitochondrial sub-cellular fractions(29). These studies also demonstrated by

im-munofluorescence

reactivity

of these sera with a

hepatoma

cell surface antigen(s); immunofluorescence was eliminated by preabsorption of antisera with bovine heart mitochondria (29). The functional

significance

of these

findings

is not as yet known. Previous studies that we performed in cultured rat hepatocytes revealed that

uptake

of bilirubin and BSP

required

cellularATP and the presence ofCl- (6). Other studies re-vealed that mitochondria transport

organic

anions such as BSP (30, 31), andin high doses, BSP and bilirubin inhibit mito-chondrial

respiration (32).

The

[35S]BSP photoaffinity-labeling

studies,

which were

performed

in an

approximately

threefold

molar excess of native bovine F1-ATPase

using

the molecular massof

371,135

daltonsasdetermined

by

Walker etal.

(33),

indicate

that

only

the (-subunit bound BSP

(Fig. 10).

The presence ofan

organic

anion

binding

site ontheF1-ATPase

d-subunit

suggeststhat itmay

play

arole in theseprocesses.

Acknowledgments

We acknowledge the assistance of Dr. Rui Liu in performingthe Northern blotanalysis.

Thisworkwassupported by National Institutes of Healthgrants

DK-23026 (Dr.Wolkoff),CA-06576 (Dr.Novikoff), DK-32972 (Dr.

Stockert), CA-45476 (Dr. Burk), and DK-41296 (LiverResearch

Center). Dr. Burk is also supported bya grant from the American

CancerSociety (JFRA-172) and the Sinsheimer Foundation.

Com-puter resourcesused tocarry out ourstudieswere providedby the BIONETNational ComputerResourceforMolecularBiology which is fundedbytheBiomedical ResearchTechnology Program,Division of

Research Resources, National Institutes of Health, grant

P41RRO1685.

References

1. Scharschmidt, B. F., J. G. Waggoner, and P. D. Berk. 1975.

Hepaticorganic anion uptake in therat.J.Clin. Invest. 56:1280-1292.

2.Stollman,Y.R.,U.Gartner,L.Theilmann,N.Ohmi, andA. W.

Wolkoff. 1983. Hepatic bilirubin uptake in theisolated perfused rat

liver isnotfacilitated by albuminbinding.J.Clin.Invest.72:718-723. 3.Wolkoff,A.W. 1987. The role ofanalbuminreceptorinhepatic

organicanion uptake: thecontroversycontinues.Hepatology. 7:777-779.

4. Goresky,C. A. 1964. Initial distributionandrateof uptake of

sulfobromophthaleinin theliver.Am. J.Physiol.207:13-26.

5. Lan, J. A., L. R. Chervu, K. L. Johansen, and A. W.Wolkoff. 1988. Uptake of technetium 99m hepatobiliary imagingagents by

culturedrathepatocytes.Gastroenterology. 95:1625-1631.

6.Wolkoff,A. W., A.C.Samuelson, K. L. Johansen, R. Nakata, D. M.Withers,and A.Sosiak. 1987.Influence ofCl- on organic anion transportinshort-term culturedrathepatocytes andisolated perfused

ratliver.J.Clin.Invest.79:1259-1268.

7.Wolkoff,A. W., andC. T. Chung. 1980. Identification,

purifica-tion, and partial characterization ofanorganic anion binding protein fromratlivercell plasma membrane. J.Clin.Invest. 65:1152-1161.

8.Wolkoff,A.W., A.Sosiak,H.C.Greenblatt,J. Van Renswoude,

andR.J.Stockert. 1985. Immunological studies ofanorganic

anion-binding protein isolated fromratlivercell plasma membrane. J.Clin.

Invest.76:454-459.

9.Young, R. A., and R. W. Davis. 1983. Yeast RNA polymerase II genes: isolation with antibody probe. Science (Wash. DC). 222:778-782.

10. Hunkapillar, M. W., E. Lujan, F. Ostrander, and L. E. Hood.

1983. Isolation of microgram quantities of proteins from

polyacryl-amide gels for amino acidsequenceanalysis. Methods Enzymol. 91:227-236.

11.Czaja,M. J., F. R.Weiner,M.Eghbali,M-A., Giambrone, M.

Eghbali, andM. A. Zern. 1987.Differential effects of y-interferonon

collagen andfibronectingeneexpression. J. Biol. Chem. 262:13348-13351.

12.Wolkoff, A.W., R. D.Klausner, G.Ashwell, and J. Harford. 1984. Intracellularsegregation of asialoglycoproteinsandtheir recep-tor: aprelysosomaleventsubsequenttodissociation of the

ligand-re-ceptorcomplex. J.Cell Biol. 98:375-381.

13.Paietta, E., R.J.Stockert,A.G. Morell, V.Diehl,and P. H.

Wiernik. 1986. Unique antigenofculturedHodgkin'scells. J. Clin.

Invest.78:349-354.

14.Kurisu, H., P. Nilprabbhasorn, andA.W.Wolkoff. 1989.

Prep-aration of35S-sulfobromophthaleinofhigh specific activity.Anal. Bio-chem. 179:72-74.

15.Novikoff,P. M., N. F., LaRusso, A. B.Novikoff,R. J.Stockert,

A.Yam, andG.D.LeSage. 1983.Immunocytochemical localization of lysosomal

fl-galactosidase

inratliver.J. Cell Biol. 97:1559-1565.

16.Garboczi,D.N., A. H. Fox,S.L.Gerring, andP. L. Pedersen.

1988.

#3-subunit

ofratlivermitochondrialATPsynthase:cDNA clon-ing, amino acidsequence, expression in Escherichiacoli, and

struc-turalrelationshiptoadenylate kinase.Biochemistry.27:553-560.

17.Stremmel,W.,and P. D. Berk.1986. Hepatocellular uptakeof

sulfobromophthaleinandbilirubin isselectivelyinhibited byan anti-body to the liverplasma membrane sulfobromophthalein/bilirubin binding protein.J. Clin.Invest.78:822-826.

18. Sottocasa, G. L., G. Baldini, G. Sandri, G. Lunazzi,andC. Tiribelli. 1982. Reconstitution in vitro ofsulfobromophthalein trans-portby bilitranslocase. Biochim.Biophys.Acta.685:123-128.

19.Mihara, K., T. Omura,T.Harano, S. Brenner, S. Fleischer,

K.V.Rajagopalan,andG. Blobel. 1982. Rat liverL-glutamate dehy-drogenase, malate dehydrogenase,

D-f-hydroxybutyrate

dehydroge-nase,andsulfite oxidaseareeachsynthesizedaslargerprecursorsby cytoplasmic free polysomes.J.Biol. Chem. 257:3355-3358.

20.Douglas, M.G.,M.T.McCammon, andA.Vassarotti. 1986.

Targeting proteinsintomitochondria. Microbiol.Rev.50:166-178. 21. Hay, R.,P. Bohni, and S.Gasser. 1984. How mitochondria importproteins. Biochim.Biophys.Acta.779:65-87.

22.Schatz,G.,R. A.and Butow.1983.How areproteins imported into mitochondria? Cell. 32:316-318.

23. Walter, P., and G. Blobel. 1981. Translocation ofproteins

acrosstheendoplasmic reticulum. III. Signal

recognition protein

(SRP) causessignal sequence-dependent and

site-specific

arrest of chainelongationthatis releasedbymicrosomal membranes.J. Cell Biol. 91:557-561.

24. Walter, P., and G. Blobel. 1981. Translocation ofproteins

acrosstheendoplasmicreticulum.II.Signal recognitionprotein (SRP)

(9)

vitro-assembled polysomessynthesizing secretory proteins. J. CellBiol.

91:551-556.

25. Walter, P.,I.Ibrahimi,andG.Blobel. 1981. Translocation of

proteinsacrosstheendoplasmic reticulum. I. Signal recognition

pro-tein (SRP) bindstoin vitro-assembledpolysomessynthesizing secre-toryproteins.J.Cell Bio. 91:545-550.

26.Weibel, E.R., W.Staubli,H. R.Gnagi,and F. A. Hess. 1969.

Correlatedmorphometric and biochemical studiesonthelivercell.I. Morphometricmodel,stereologicmethods, and normal morphomet-ric dataforratliver.J.Cell Bio. 42:68-91.

27.Ohta, S.,andY.Kagawa. 1986.HumanF1-ATPase:molecular

cloning ofcDNAfor the beta subunit.J.Biochem.99:135-141. 28. Berk, P. D., B. J. Potter, D. Sorrentino, W. Stremmel, D.

Stump, C.-L.Kiang, and S.-L.Zhou.1989. Characteristicsof organic

anionbinding proteins fromratliver sinusoidalplasma membranes. In

Hepatic Transport ofOrganic Substances. E. Petzinger, R. K. H.

Kinne,and H.Sies, editors. Springer-Verlag, Berlin. 195-2 10.

29. Ghadiminejad, I., andH. Baum. 1987. Evidence for the

cell-surface localization ofantigenscrossreactingwiththe"mitochondrial antibodies" ofprimary biliary cirrhosis.Hepatology.7:743-749.

30. Burr, R.,M. Schwenk, andE.Pfaff. 1977. Interaction of

bro-mosulfophthalein with mitochondrial membranes: uptake of

bromo-sulfophthaleinandeffectonANS-fluoresence. Biochem. Pharmacol. 26:457-460.

31. Laperche, Y., and P. Oudea. 1976. Inhibition by sulfobro-mophthalein of mitochondrial translocation of anions and adenine nucleotides: effects upon liver adenosine triphosphate and possible correlation withinhibition of bile flowintherat.J.Pharmacol. Exp.

Ther. 197:235-244.

32. Burr,R., M. Schwenk, and E. Pfaff. 1977.Interaction of

bro-mosulfophthalein with mitochondrial membranes: inhibition of respi-ration.Biochem. Pharmacol. 26:461-466.

33.Walker, J. E., I. M. Fearnley, N. J.Gay,B.W.Gibson,F. D.

Northrop, S. J. Powell, M. J. Runswick, M. Saraste, and V. L. J. Tybulewicz. 1985. Primary structure and subunit stoichiometry of

References

Related documents

Proceedings of the Workshop on Sentiment and Subjectivity in Text Proceedings of the Workshop on Sentiment and Subjectivity in Text, pages 47?54, Sydney, July 2006 c?2006 Association

This study aimed for the first time the identification of the antioxidant and the antimicrobial activities of the total tannins extracted from aerial part of Sedum pubescens

But if we could get a way of converting the unstructured text documents into a structured form, we could then simply apply the traditional data mining techniques

In this paper, we have both generalized and extended the study carried out by [1] on the estimation of the popu- lation ratio ( R ) of the population means of two variables ( y and

According to Rag- bir Kaur (2010), individual differences involve differences in religion and culture (spiritual), physical (physical), the wisdom (intellectual), emotional and

In this proposed modification of LEACH’s cluster-head selection algorithm is to reduce energy consumption where an optimal cluster head selection algorithm prolongs the

Numerical simulation result shows that ventilation system of case 2 is safer than case 1 in terms of contaminated particles strikes on the critical areas of Operating

reveals that k-means clustering has a high Davies–Bouldin index and low Dunn index, which is not suited to assess the quality of an image based on internal