• Nenhum resultado encontrado

Production of human recombinant proapolipoprotein A-I in Escherichia coli: purification and biochemical characterization

N/A
N/A
Protected

Academic year: 2021

Share "Production of human recombinant proapolipoprotein A-I in Escherichia coli: purification and biochemical characterization"

Copied!
8
0
0

Texto

(1)

MaryAnn Liebert, Inc., Publishers

Pp. 429-436

Production

of

Human

Recombinant

Proapolipoprotein

A-I

in Escherichia coli:

Purification

and Biochemical Characterization

NICOLE

MOGUILEVSKY,*

CORNELIS

ROOBOL.t

ROSETTE

LORIAU,*

JEAN-PAUL

GUILLAUME,*

PAUL

JACOBS,*

ALFREDO

CRAVADOR,*

ALBERT

HERZOG,*

LÉON

BROUWERS.t

ALAIN

SCARSO.t

PASCAL

GILLES.t

LEIF

HOLMQUIST.t

LARS A.

CARLSON.Î

and ALEX BOLLEN*

ABSTRACT

A human liver cDNA

library

was usedtoisolateaclone

coding

for

apolipoprotein

A-I

(Apo

A-I).

The clone

carries thesequence forthe

prepeptide (18

amino

acids),

the

propeptide

(6

amino

acids),

and themature pro-tein

(243

amino

acids).

A

coding

cassette forthe proapo A-I moleculewas reconstructed

by

fusing

synthetic

sequences,chosento

optimize expression

and

specifying

the amino-terminal methionine and aminoacids -6

to

+14,

to a

large fragment

of the cDNA

coding

for amino acids 15-243. The module was

expressed

in

pOTS-Nco,

anEscherichia coli

expression

vector

carrying

the

regulatable

X

P^

promoter,

leading

to the

pro-duction of

proapolipoprotein

A-I at up to 10% oftotal soluble

proteins.

The recombinant

polypeptide

was

purified

and characterizedintermsof

apparent

molecularmass, isoelectric

point,

and

by

bothchemical and

enzymatic

peptide

mapping.

In

addition,

it was

assayed

in vitro for the stimulation ofthe enzyme lecithin:

cholesterol

acyltransferase.

The data show

for

the first time thatproapo A-Icanbe

produced efficiently

in

E. colias a stable and

undegraded protein

having

physical

and functional

properties

indistinguishable

from

those ofthe natural

product.

INTRODUCTION is

responsible

for the formationofmost ofthe

cholesteryl

estersin

plasma

(Fielding

et

al,

1972).

Apo

A-I and HDL

Human

apolipoprotein A-I

(apo A-I),

the

major

pro- may

participate

in "reverse cholesterol

transport,"

from

pe-tein constituent of

high-density lipoproteins

(HDL),

ripheral

tissues to the liver for excretion from the

body

is

produced

in liver and intestine as a precursor

protein

(Glomset, 1968).

(preproapo A-I)

which

undergoes

intracellular

cleavage

Since accumulation of cholesterol in the arteries is the

andis released into the

plasma

and

lymph

asa

proprotein

hallmark and

presumably

an

important

cause for

athero-(Wu

and

Windmueller,

1979).

Proapo

A-Icarries six addi-

sclerosis,

stimulation of reverse cholesterol transport

by

tional amino acids

(Arg-His-Phe-Trp-Gln-Gln)

at the

supplementing

apo A-I

might delay

the atherosclerotic amino-terminalof the mature

protein.

It is cleaved inthe process.

Techniques

to

produce

this

protein

in

large

quan-vascular space

by

a

specific

protease to

yield

the mature tities have been

developed

using

blood

plasma

as

starting

molecule

(Gordon

et

al, 1983;

Zannis et

al,

1983).

Apo

material

(Ross

and

Carson,

1985; Breweret

al, 1986).

At-A-I is a

single

nonglycosylated polypeptide

ofknown se- tempts toexpressmatureapoA-I in recombinant bacteria

quence,

composed

of 243 amino acid residues

(Brewer

et

proved

difficult due to the

instability

of the

protein

ai, 1978).

Inthepresenceof

synthetic liposomes,

it serves

(Lorenzetti

et

al,

1986;

Mallory

et

al., 1987);

moreover,

as cofactor for lecithin ¡cholesterol

acyltransferase,

which even as afusion

protein,

apoA-I wasstillverysensitiveto

•Service deGénétique Appliquée, Université Libre deBruxelles, 24rue del'Industrie, B-1400 Nivelles, Belgium.

tu.C.B.

and U.C.B. Bioproducts,chemin du Foriest, B-1420 Braine-l'AUeud,Belgium.

ÍKing

GustafVResearch Institute, KarolinskaInstitute, Box60004, S-10400Stockholm, Sweden.

(2)

degradation

and could not be cleaved

efficiently

from the

fusion

(Lorenzetti

et

al., 1986;

Monaco et

al,

1987).

To

circumvent these

problems,

we

expressed

apo A-I as the

proprotein.

We show here that proapo A-I can be

pro-duced

efficiently

in Escherichiacolias astable and

unde-graded molecule,

having physical

andfunctional

properties

identical to those of the natural

product.

MATERIALS AND METHODS

Plasmids, strains,

and media

Three

plasmids, pBR322 (Bolivar

et

al.,

1977),

pULB1221

(a

multi-cloning

site derivative of

pULB1219,

Cravador et

al., 1985),

pOTS-Nco

(Devare

et

al,

1984),

and three E. coli

strains,

MM294

(Lawn

et

al.,

1981),

AR58,

and AR120

(Mott

et

al,

1985),

were used. M33

minimal medium is described

by Fujimoto

et al.

(1988).

RNA

preparation; synthesis,

cloning,

screening,

and

sequencing of

cDNA

Procedurestoprepareandscreenthe humanlivercDNA

library

were as described before

(Jacobs

et

al,

1985).

Oligonucleotides

were

synthesized

by

the

phosphoramidite

method on an

Applied Biosystem

Synthesizer,

model380A

(Matteucci

and

Caruthers, 1981).

DNA sequence

analysis

was

performed by

the methods of Maxam and Gilbert

(1977)

and

Sanger

et al.

(1977).

Immunological

detection

of

proapoA-I in bacterialextracts

Transformed bacterial strains were grown in 20 ml of

rich

(LB)

or minimal medium

supplemented

with

ampicil-lin

(50

/ig/ml)

to an OD630 of 0.6. Induction of the X

PL

promoter was achieved either

by

shifting

the temperature from30°Cto

42°C,

or

by

adding

nalidixic acid

(60

fig/ml)

to cells grown at 37°C

(Mott

et

al., 1985).

Induction was

performed

forvarious

periods

of time as indicated in the

figure legends.

One-milliliter

aliquots

of cultures were collected and

centrifuged

at

10,000

x g for 5 min. Pellets were

resus-pended

in 50

¡A

ofNaDodS04

sample

buffer

(50

mM Tris

HC1

pH

6.8

containing

2%

NaDodS04,

6 Murea, 5%

2-mercaptoethanol,

and 10%

glycerol),

boiled for 3 minat

100°C,

and

centrifuged

for 10 minat 15,000 x g.

Samples

werethen fractionated

by electrophoresis

on 12.5or7.5%

NaDodS04-polyacrylamide gels

as described

by

Laemmli

(1970).

The

proteins

wereblotted ontonitrocellulose sheetsand tested for the presence of proapo A-I with antibodies raised

against plasma

apo A-I.The

procedure

follows Bol-len et al.

(1984),

except that the second

antibody,

goat

antimouse,

was labeled with alkaline

phosphatase

instead

of

peroxidase,

and that the

chromogenic

substrates were

BCIP

(5-bromo-4-chloro-3-indolyl-phosphate, Boehringer)

and NBT

(4-nitrotetrazolium-chloride blue, Sigma),

dis-solved in 100 mM Tris-HCl

pH 9.5,

100 mM

NaCl,

and 50 mM

MgCl2

at 66and 33

fig/ml, respectively.

Purification of

recombinantproapoA-I

Packed bacterial cells

(usually

35

grams),

derived from 15-liter fermentations in M33

medium,

were

resuspended

intwovolumes ofice-cold extractionbuffer

(20

mM

Tris-HCl

pH

7.5, 100mM

NaCl,

1

mMEDTA,

1 mM

PMSF,

and 100

¡ig/ml merthiolate)

and

disrupted by

a

single

pas-sage

through

a French'spressure cell at 1,000 PSIG. The

extractwas

centrifuged

for 15 min at

4,000

x gto

sedi-ment membranes and aggregates then

applied

onto a

Phenyl Sepharose

column CL-4B

(bed volume,

25

ml;

sample,

100

ml).

The column wasthen

sequentially

eluted

with 250 mlof extraction

buffer,

250 mlof thesamebuffer

supplemented

with 60%

(vol/vol) propylene glycol,

and 250 ml of thesamebuffer without NaCl and

supplemented

with80%

(vol/vol) propylene glycol.

Fractions

containing

proapo A-Iwere

pooled

and then

applied

onto a PBE 74

chromatofocusing

column

(0.8

x 60

cm),

which was

elutedwithtwobed volumes ofabuffer

containing

25 mM

imidazole-HCl

pH

6.0 and 6 Murea, followed

by

10 bed

volumes of

polybuffer

74

(dilution

1/10) pH 4.0,

6 Murea. The columnwas

regenerated by washing

with 50 mM

Na-citrate

pH

4.0,

1 M

NaCl,

and 6Murea. Fractions were

assayed

for total

protein

content and apoA-I-related

im-munoreactivity,

and

analyzed by

isoelectric

focusing.

Chemical

peptide mapping

with BNPS-Skatol

Chemical

cleavage

with BNPS-Skatolwas

performed

ac-cording

to the

procedure

ofFontana

(1972). Briefly,

5-10 fig of

protein

were dissolved in 100

¡A

of a solution of

0.15%

(vol/vol) phenol

in50%

(vol/vol)

acetic acid.

Then,

50

fA

ofasolutionof 4.8mgBNPS-Skatolpermilliliter of

glacial

acetic acid were

added,

followed

by

an incubation

of 72 hr at25°C.

Subsequently,

50

¡A

of

2-mercaptoetha-nol was

added,

followed

by

a second incubationof 5 h at

37°C. The

samples

were

evaporated,

redissolved in 100

fA

of water, andextractedthree times with 200

^tl

of

ethyl

ace-tate.The

organic

phases

were

discarded,

whereas the aque-ous

phases

were

lyophilized

and

analyzed

by

acrylamide

gel electrophoresis.

Stimulation

of

lecithin:cholesterol

acyltransferase

activity by

recombinantproapo A-I

Human

plasma

apo A-I was

purified according

to

Holmquist

and Carlson

(1977)

and

Holmquist

and

Bro-strom

(1979).

Simple

bilayer

vesicles

containing

egg leci-thinand

7(/i)-3H-labeled

cholesterol

(molar

ratio

4:1)

were

(3)

purified

human

plasma

lecithin:cholesterol

acyltransferase

(35,000

units/mg)

was

prepared

according

to

Holmquist

(1987).

Incubations and enzyme

activity

determinations

werecarried outin 50 mM

phosphate

buffer

pH

7.4as

de-scribed

by

Piran and Morin

(1979).

Incubation mixtures

(250

fA)

contained perliter:240

/¿moles

lecithin,

60

/¿moles

cholesterol,

and 3 or 6

/¿moles apolipoprotein.

Miscellaneous

procedures

Protein concentrationsweredetermined

according

tothe

procedure

of Bradford

(1976),

using

ovalbumin asa

stan-dard. Partial

tryptic

hydrolysis

was

performed

according

tothe

procedure

ofAmonsetal.

(1983).

Human apo A-I-related

immunoreactivity

was

quantified

by

radial

immu-nodiffusion

(Cheung

and

Albers, 1977),

using

commercial

apo A-I

(Sigma)

as astandard. The relative

immunoreac-tivity profile

were determined

by

immunoturbidimetry

(Shapiro

et

al, 1980).

Isoelectric

focusing

was carriedout

according

to the

procedure

of

Catsimpoolas (1968)

in a

Bioradrod

electrophoresis

apparatus,

applying

a

self-gen-erating gradient

from

pH

4to

pH

6. Linear and nonlinear

regression analyses

were

performed

withtheStatView

pro-gram

(Brain

Power Inc.,

USA).

RESULTS

Molecular

cloning of

Apo

A-IcDNA and

screening

of

the cDNA

library

Total

poly(A)*RNA

(25

fig)

fromhuman liverwas

tran-scribed into double-strandedcDNA

(ds

cDNA).

Molecules

resistant to S, nuclease were then fractionated on a

10-30% sucrose

gradient

to enrich for molecules

larger

than 0.8 kb. This ds cDNA fractionwasinserted into thePst I

siteof

pBR322 by

thedC:dG

tailing

procedure

andusedto

transform E. coli MM294 cells. The

library

was screened

by

the

colony hybridization

method

using

a 22-base

syn-thetic

oligonucleotide probe

derivedfromthesequence for apo A-I

(Cheung

and

Chan, 1983)

and

coding

for amino

acids -22 to -15 ofthe

signal

peptide.

Of the 30 clones

hybridizing

with the labeled

probe,

one,

pNIV1602,

was

chosen for further

analysis.

Characterization

of

theclone

pNIV1602

Clone

pNIV1602

carriedaninsertof 878

bp

whichcould

be isolated as a

single fragment by digesting

the

plasmid

with Pst I.The insert was

mapped

with variousrestriction

enzymesand

fully sequenced

oneach strand. Thedata

(not

shown)

showed that

pNIV1602

carries

(i)

a

19-bp

5'

non-coding

sequence,

(ii)

a

75-bp

sequence

corresponding

to

the known

signal

and

propeptide

region

of apo

A-I,

in-cluding

the ATGinitiation

codon,

(iii)

the

complete 729-bp

sequence

corresponding

tothemature

protein,

followed

by

aTGAstop

codon,

and

(iv)

a

55-bp

3'

noncoding

sequence

that did notcontain the

poly(A)

stretch. The sequence of thecDNAinsertwasidenticaltothesequence

published

by

Cheung

and Chan

(1983),

but differedfrom the sequence

of Seilhameretal.

(1984)

by

afew bases and

by

the

length

of the 5' and 3'

noncoding regions.

The deduced amino

acid sequence,

however,

wasconsistentwith the

published

data

(Cheung

and

Chan, 1983;

Seilhamer et

al.,

1984).

Construction

of

the recombinant

plasmid

pNIVl617

Theconstruction

proceeded

in twosteps. Thefirst used

the

multi-cloning

site

plasmid

pULB1221,

linearized with

Bal

I,

as anintermediatevehicletosubcloneablunt-ended

744-bp

Bal l-Pst I DNA

fragment

derivedfrom the clone

pNIV1602

and

coding

for amino acids 15-243 ofthe apo

A-I

protein.

The

resulting plasmid

was

digested

with Eco RI and Bal I and fused to

(i)

the

1,880-bp

Eco Rl-Nco I DNA

fragment

containing

the

regulatory regions

derived

fromthe

expression

vector

pOTS-Nco (Devare

et

al., 1984)

and

(ii)

a

65/61-bp

Nco l-Bal I

synthetic

DNA

adaptor

en-coding

theamino-terminal methionineand amino acids -6

to 14 of proapo A-I

(Fig.

1).

The

resulting

recombinant

plasmid,

pNIV1617,

thus carried the strong

regulatable

X

Pl

promoter upstreamfromthecassette

coding

for proapo A-I.

The

synthetic

adaptor

wasassembled fromfour

oligonu-cleotides

(Fig. 2)

whose base sequence was

designed by

computertominimize

secondary

structuresatthe5' endof

the

coding

sequence

(S.A.S.I.P.

software

package;

Clav-erie,

1984). Triplets

encoding

aminoacidresidues -6, -1, 1,

3, 4,

5, 6,

7, 10, 11,

and

14,

therefore differed fromthe natural ones

(Table 1),

although

the

resulting

amino acid

sequence was

preserved.

Plasmid

pNIV1617

wascharacterized

by

restriction

anal-ysis,

hybridization

with labeled

synthetic

DNA

fragments,

and

sequenced throughout

the

adaptor

region.

Itwasthen

introduced into E. coliAR58

(for

thermal

induction)

and

AR120

(for

chemical

induction).

Expression

of

proapoA-I

Cells

carrying

either the control

plasmid,

pOTS-Nco,

or

the

pNIV1617 plasmid

weregrownat30°C inminimal

me-dium,

for thermal

induction,

or at 37°C in rich

medium,

forchemical induction.

Derepression

of the X

PL

promoter

wasachievedeither

by

shifting

thetemperatureto42°Cor

by

adding

nalidixic acid. Inductionwas

performed

for

var-ious

periods

of time as indicated in the

legend

to

Fig.

3.

Pelletedbacteriawere

lysed

andthe

resulting

extractswere

analyzed

by

NaDodS04-polyacrylamide gel

electrophore-sis,

Western

blotting,

andimmunodetection

using

antibod-ies raised

against plasma

apo A-I.

Figure

3 shows that a

single

product, specifically

recognized by

anti-apo

A-I

antibodies,

is

expressed

in the bacterial strains

carrying

plasmid

pNIV1617,

irrespective

ofthe modeofinduction.

(4)

apoA-l BalI "15 synthetic DNA ATG Ncol Met.aa_ T40NA ligase Antpr

Ball-EcoR Ifragment

13 07kbl

Amp

EcoRl-Nco Ifragment

(1 BBkb)

proapoA-l

"»-6 -aa243'

The

protein

hadamolecularmassofabout

28,000

daltons

which,

within

experimental

limits,

fittedthecalculated

mo-lecular

weight

ofproapo A-I.

Levels of proapo A-I

produced

in bacteria were

esti-mated

by

single

radial immunodiffusion. Under thermal

induction conditionsin M33 minimal

medium,

proapo A-I

was

expressed

at about 10% of total soluble

proteins

versus 3.5% inLB medium. Chemical induction

appeared

tobe less

efficient, yielding

expression

levels on theorder

of2% in LB mediumand no

expression

in M33 medium.

Purification of

recombinantproapo A-I

We used

hydrophobic

interaction

chromatography

to

purify

the recombinant proapo A-I. This

approach

al-lowed an excellent

separation

of recombinant molecules

from otherbacterial constituents

(Fig.

4,

peak

III).

Chro-matofocusing

on aPBE-74 columnwasusedto

purify

the

product

further. As shown in

Fig.

5, immunoreactive

ma-FIG. 1. Construction of

plasmid

pNIV1617.

A

744-bp

Bal l-PstI

fragment

derivedfromclone

pNIV1602,

coding

for amino acids 15-243 of apo

A-I,

was

ligated

into the

unique

Bal I site of

pULB1221.

From the

resulting

plas-mid,

the

large

Eco Rl-Bal I

fragment

was fused to the

1,880-bp

Eco Rl-Nco I

fragment

derived from

pOTS-Nco

andtothe

synthetic

adaptor

Nco l-Bal Itoregeneratea

se-quence

coding

for proapo A-I.

Hybrid

molecules were

usedto transformE. colicellsandtransformantswere

se-lected for

ampicillin

resistance. The nucleotide sequence of the

junctions

between

fragments

and ofthe full

synthetic

fragment

were determined.

Ncol site

5-- -35 mer

BalIsite

-3'5' -30

mer-CATG AGA CATTTCTGG CAG CAG GAC GAA CCT CCA C AA TCT CCTTGG GATAGA GTTAAGGAC TTG G TCT GTA AAGACC GTC GTC CTG CTTGGA GGT G TT AGAGGAACC CTATCT CAATTCCTGAAC C

3"-18 met--53-"- 43 mer-»5

Met Arg His Phe Trp Gin Gin

6 -3 1

NH-, terminal propeptide

methionine " 6 aa

1 '

Asp Glu Pro Pro Gin Ser Pro Trp Asp Arg Val Lys Asp Leu A(la)

»1 +3 +6 +9 +12 +14

mature apo A-l (aa.

-"14

cleavage site for apoA Ipropeptidase

FIG. 2.

Synthetic adaptor

usedtoreconstruct the proapo A-I cDNA. Four

oligomers

were

synthesized

andassembled

in vitrotogeneratea

65/61-bp

fragment

flanked

by

aNcoIandaBal Isitesonthe5'and3'

ends,

respectively.

The de-duced aminoacidsequenceisindicated. The arrowshowsthe known

cleavage

site forapo A-I

propeptidase

(Gordon

et

(5)

Table 1. Relevant Changesin CodonsUsed toReconstruct

theProapo A-I cDNA Position in

the

protein

Aminoacid Naturalcodon

Modified

codon

-6 -1 1 2 3 and4 5 6 7 8 9 10 11 12 13 14

Arg

Gin

Asp

Glu Pro Gin Ser Pro

Trp

Asp

Arg

Val

Lys

Asp

Leu CGG CAÁ GAT GAA CCC CAG AGC CCC TGG GAT CGA GTG AAG GAC CTG AGA CAG GAC CCC and CCA CAA TCT CCT AGA GTT TTG

Modifications in the codon usageweredone in ordertominimizesecondary

structures in the 5' end of the proapo A-I cDNA. The software package

S.A.S.I.P.developed bythe Institut Pasteur(Paris)wasused(Claverie,1984).

11 10 987 65432 1 kDa

A

.43 13 12 11 10 9 8 7 6 5 4 3 2 1 kDa

B

_29 .18.4 29 .18.4

FIG. 3. Immunodetection of proapo A-I on Western blots. Protein extracts from induced and noninduced bacteria

carrying

plasmid pNIV1617

were

prepared

for

electrophoresis

on

NaDodS04-polyacrylamide gels.

Nitrocellulose blots wereincubated withmouseantiapoA-Iserum, followed

by

alkaline

phosphatase-labeled

goatanti-mouse and the

chro-mogenic

substrate. A. Timecourseof thermalinduction ofbacteria

carrying

pNIV1617

and growninM33minimal

me-dium. Lane

1,

Plasma derived apoA-I, 500 ng; lanes

2-6,

samples

takenat 1,

20, 40, 60,

and 120 minin noninduction conditions

(30°C);

lanes 7-11,

samples

taken at

1, 20,

40,

60,

and 120 minafter induction

(42°C).

B. Time courseof

chemicalinduction

(nalidixic acid)

ofbacteria

carrying

pNIV1617

andgrowninrichLBmedium.Lane 1,Plasma derived apo

A-I,

500 ng; lanes

2-7,

samples

taken at

1, 30, 60, 120, 180,

and240minin noninduction conditions

(37°C);

lanes

(6)

terial

(fractions 37-48)

was resolved from contaminants

andwasidentified

by

isoelectric

focusing

mainly

asproapo

A-I.

Moreover,

chromatofocusing

resolved most apo A-I

relatedisoforms.

Characterization

of

purified

recombinant

proapoA-I

Pure recombinant proapo A-I was

analyzed by

NaDodS04-polyacrylamide

gel

electrophoresis

and isoelec-tric

focusing.

Recombinant proapo A-Ibehavedasa

single

Material Eluted. Arbitrary Units

h.

I

Ccccccccctíí?

<ccccCrccCco°co_

20 30 FractionNumber -•rft>-40

FIG. 4.

Hydrophobie

interaction

chromatography.

A total of 100 ml ofclearedextractof bacterial cells derived

from 15-liter fermentation in M33 medium was

subjected

to

hydrophobic

interaction

chromatography

as described

inMaterials and Methods. The eluatewas

assayed

fortotal

protein (O)

and apo A-I-related

immunoreactivity

(•).

PeakIIIcontains the

majority

of therecombinant

protein.

polypeptide

chain with an apparent molecular mass of

(27.3 ±1.1)

kD.Matureapo A-Iwas

slightly

smaller,

hav-ing

an apparent molecular

weight

of

(26.8

±

1.1)

kD. In

isoelectric

focusing (data

not

shown),

the recombinant

protein

showedone

major

isoformandoneminor

isoform,

having

isoelectric

points

of 5.20 and

4.95,

respectively.

The

relative

migration

distances of these isoforms coincided

exactly

with those of

plasma

proapo A-I andapo

A-I,

re-spectively.

Partial chemical

cleavage

at

tryptophan

resi-dues,

was

performed

on recombinant proapo A-I

(major

isoform)

and on

plasma-derived

apo A-I.

Figure

6 shows

the

large peptides

of

expected

molecular

weight produced

in both cases

by cleavage

at

positions

8, 50, 72,

and 108 in

the

proteins.

In

addition,

digestions

of recombinant

proapo A-I and of

plasma-derived

apo A-I with

trypsin

in

verymild conditions

yielded

similarpatterns of

fragments

(data

not

shown).

Thus,

both types of

experiments

suggest that the recombinant molecule is

structurally

identical to

its natural counterpart.

Functional

analysis

of

recombinantproapoA-I

Recombinant proapo A-I

(pi 5.20)

and its minor iso-form

(pi 4.95)

were both ableto transform

synthetic

leci-thin-cholsterol

liposomes

into substrates for the enzyme

lecithin:cholesterol

acyltransferase.

AsseeninTable

2,

the

activation

capacity

of recombinant

apoproteins

wassimilar

totheoneobservedwithauthentic human

plasma

apo A-I.

kDa

J

Material Eluted, Arbitrary Units T-6 +5 30 40 50 Fraction Number

FIG. 5.

Phenyl-sepharose chromatography.

Peak III of the

phenyl sepharose

eluate was

subjected

to

chromatofo-cusing

as described in Materials and Methods. The eluate

was

assayed

for total

protein (O),

apoA-I-related

immu-noreactivity (•),

and

pH

(A).

Fractions 37-48 contain the

purified

recombinant

product.

43

30

20.1"

****'**

14-4-FIG. 6. Chemical and

enzymatic

characterizationof

pro-apo A-I. Purified recombinant proapo A-I and human

plasma

apoA-I were

subjected

to

partial

chemical

peptide

mapping

as described in Materials and Methods. Lane

1,

Molecularmass standards

(Pharmacia,

14,400-94,000

dal-tons)

indicated

by

arrows; lane

2,

human

plasma

apo

A-I,

untreated;

lane

3,

human

plasma

apo

A-I, cleaved;

lane4,

recombinant proapo

A-I, untreated;

lane

5,

recombinant

(7)

Table 2. Stimulation of Lecithin ¡Cholesterol Acyltransferase Activity by

Recombinant Proapo A-I Source

LCAT

Apolipoprotein

LCAT

activity

(nmoles/hr

x

liters)

Plasma apo A-I

Recombinant proapo A-I

Major

isoform

(pi 5.20)

Minor isoform

(pi 4.95)

25 25 100 100 100 100 100 100 0 3 0 3 6 0 3 6 2.5 250 100 850

1,600

100 700 700

Synthetic liposomes (lecithin/cholesterol, 4:1)wereincubatedat37°C for 2-4hr,inatotal vol-umeof 250(A,with LCAT(25units/ml)and variousamountsofapoA-I. The initialrateof

chole-sterylesterformation (LCAT activity)wasmeasuredaccordingtoPiran and Morin(1979).

DISCUSSION

Wehave cloned a cDNA

coding

for human preproapo

A-I;

its deduced amino acid sequence fits the

published

data

(Cheung

and

Chan, 1983;

Seilhameret

al.,

1984)

and

is almost identicaltothat

reported

by

Breweretal.

(1978).

Weconstructedamodule

coding

fortheproapoA-I mole-cule and

expressed

it in E. colt under the control of the

regulatable

X

PL

promoter.

Proapo

A-I,

produced

at

high

level,

was

stable,

largely undegraded,

and

indistinguishable

from the natural

product

in terms ofapparent molecular

mass,isoelectric

point,

and

partial

peptide

map. As

expected,

mostof the material

produced

in E. coliwasindie

propro-tein

form,

although

thepresenceofanamino-terminal

me-thioninecannot be excluded yet. The

purified

protein

ap

peared

to contain a minor isoform which coincided with

mature apo A-I

by

isoelectric

focusing. However,

it is

un-certain if this is duetodeamidationortoconversion ofthe recombinant

protein

to matureapo A-I. The

high

level of

expression

observed for proapo A-I inbacteriacontrasted

with the relative failure

(Lorenzetti

et

al, 1986;

our

data,

not

shown)

to achieve

significant expression

of a

stable,

undegraded

mature apo A-I

species

in the samehost. The

dramatically

increased

expression

associatedwith the

pres-enceofthe

propeptide

sequence

might

be

explained

in

sev-eral ways.

Transcriptional

or translational efficiencies

might

be increased due to the extra sequence.

Alterna-tively,

the

proprotein species

might

fold more

stably

in

bacteria thanthe mature

protein

andhencebemore

resis-tant to

degradation.

This last

hypothesis

seems more

probable

since

pulse-chase

experiments

have shown that thehalf-life ofmature apo A-I inbacteria is less than 10

min

(Mallory

et

al, 1987).

Our

data,

in

addition,

offer

di-rect evidence for the first time that recombinant proapo

A-I is able to stimulate in vitro the

enzymatic

activity

of lecithin:cholesterol

acyltransferase,

as

efficiently

as

plasma-derived

apoA-I. In this respect,proapo A-I

might

constitute an

appropriate

substituteto matureapo A-I for

replacement

therapy, provided

that

pharmacological

stud-ies in animal models confirm its functional

efficiency.

ACKNOWLEDGMENTS

This work was

performed

under a contract between U.C.B. and the

University

ofBrussels and

supported by

a

grant from the Swedish Medical Research Council

(19X-04)

The authorsthank Drs. A. Shatzmanand M. Rosen-L i

¡Smith

Klineand

French,

Philadelphia,

PA)

for

pro-viding

the

pOTS-Nco

vector and the E. coli AR58 and AR120strains.

REFERENCES

AMONS, R., PLUYMS, W., ROOBOL, C,andMOLLER, W.

(1983). Sequence homology between EF-la, the a-chain of

elongationfactor1 from Artemia salina andelongation factor EF-Tu from Escherichia coli. FEBS Lett. 153,37-42.

BATZRI, S.,andKORN,E.(1973).Single bilayer liposomes

pre-pared without sonication. Biochim. Biophys. Acta298, 1015-1019.

BOLIVAR, F., RODRIGUEZ, R.L., GREENE, P.J.,

BET-LACH, M.C., HEYNECKER, H.L., BOYER, H.W., CROSA, J.H., and FALKOW, S. (1977). Construction and

(8)

characterization of new cloning vehicles. IL A multipurpose cloning system. Gene2,95-113.

BOLLEN, A., LORIAU, R., HERZOG, A., and

HÉRION,

P.

(1984). Expressionof humana,-antitrypsininEscherichiacoli. FEBS Lett. 166,67-70.

BRADFORD,M.M. (1976).Arapidand sensitivemethodfor the

quantitation of microgram quantities of protein utilizing the

principleofprotein-dyebinding.Ann. Biochem. 72, 248-254.

BREWER, H.B., FAIRWELL, T., LARUE, A., ROÑAN, R., HOUSER,A.,andBRONZERT,T.(1978).The amino acid

se-quence of human apo A-I, an apolipoprotein isolated from

high density lipoproteins. Biochem. Biophys. Res. Commun.

80, 623-630.

BREWER, H.B., RONAN, R., MENG, M., and BISHOP, C.

(1986). Isolation and characterization ofapolipoproteinsA-I,

A-II and A-IV. MethodsEnzymol. 128,223-246.

CATSIMPOOLAS, N. (1968). Microisoelectrofocusing in

poly-acrylamidegels columns. Ann. Biochem.26,480-482.

CHEUNG, M.C., and ALBERS, J.J. (1977). The measurement

ofapolipoproteinA-IandA-IIlevels inmenandwomenby im-munoassay. J. Clin. Invest. 60,43-50.

CHEUNG, P., and CHAN, L. (1983). Nucleotide sequence of

clonedcDNA of humanapolipoproteinA-I.NucleicAcids Res.

11, 3703-3715.

CLAVERIE, J.-M. (1984). A common philosophy and

FOR-TRAN77softwarepackageforimplementingandsearching

se-quence databases. Nucleic Acids Res. 12, 397-407.

CRAVADOR, A., JACOBS, P., VANELSEN, A., LACROIX,

C, COLAU, B., VAN ALPHEN, P., HERZOG, A., and

BOLLEN, A. (1985). Total DNA synthesis andcloning in E. coli ofagenecodingfor the human Growth Hormone

Releas-ing Factor. Biochimie67, 829-834.

DEVARE, S.G., SHATZMAN, A., ROBBINS, K.C.,

ROSEN-BERG, M.,andAARONSON,S.(1984).Expressionof

PDGF-relatedtransforming proteinof simiansarcomavirus in E. coli. Cell36,43-49.

FIELDING, C.J., SHORE, V.G.,andFIELDING,P.D.(1972). Aproteincofactor of lecithin: Cholesterolacyltransferase. Bio-chem. Biophys. Res. Commun. 46, 1493-1498.

FONTANA, A. (1972).Modification oftryptophanwith BNPS-skatol. MethodsEnzymol.25, 419-423.

FUJIMOTO, S., NAKATSU, M., KATO, K.,andKITANO,K.

(1988).Factorsaffectingthesynthesisof the N-terminal methi-onine-free molecule of recombinant human interleukin-2 by Escherichia coli. J. Ferment. Technol. 66, 181-185.

GLOMSET, J.A. (1968). The plasma lecithin: Cholesterol acyl-transferase reaction. J. Lipid Res.9, 155-167.

GORDON, J.I., SIMS, H., LENTZ, S.R., EDELSTEIN, C, SCANU, A.M., andSTRAUSS,A.W.(1983). Proteolytic

pro-cessing of human preproapolipoprotein A-I. J. Biol. Chem.

258,4037-4044.

HOLMQUIST, L., andCARLSON, K.(1977). Selective

extrac-tion of humanserumverylowdensityapolipoproteinswith

or-ganicsolvents. Biochim. Biophys. Acta493,400-409.

HOLMQUIST, L.,andBROSTROM, H. (1979).Asimple

Beck-manDUaccessoryfor location ofproteinsseparatedby isoelec-tricfocusingingranulated gel layers:Isolation of humanserum

verylow densityapolipoproteins C-II and C-III. J. Biochem.

Biophys. Methods 1, 117-127.

HOLMQUIST, L.(1987).Purification of humanplasmalecithin:

cholesterol acyltransferase by covalent chromatography. J. Biochem. Biophys. Methods 14,323-333.

JACOBS,P., CRAVADOR, A., LORIAU, R., BROCKLY, F., COLAU, B., CHUCHANA,P.,VANELSEN, A.,HERZOG, A., and BOLLEN, A. (1985). Molecularcloning, sequencing

and expressionin Escherichia coli of human preprourokinase cDNA. DNA 4, 139-146.

LAEMMLI, U.K.(1970). Cleavageof structural proteins during the assembly of the head of bacteriophage T4. Nature 227,

680-685.

LAWN, R.M., ADELMAN, J., BOCK, S.C., FRANKE, A.E., HOUCH, CM., NAJARÍAN, R.O., SEEBURG, P.H., and

WION, K.L. (1981). The sequence of human serum albumin cDNA and itsexpressioninE. coli. Nucleic Acids Res.9, 6103-6114.

LORENZETTI, R., SIDOLI, R., PALOMBA, R., MONACO, L., MARTINEAU, D., LAPPI, D.A.,andSORIA, M.(1986).

Expressionof the humanapolipoproteinA-I genefusedtothe E. coligenefor/3-galactosidase. FEBS Lett. 194, 343-346.

MALLORY, J.B., KUSHNER, P.J., PROTTER, A.A., COFER, CL., APPLEBY, V.L., LAU, K., SCHILLING, J.W., and VIGNE, J.-L. (1987). Expressionand characteriza-tion of human apolipoprotein A-I in chínese hamster ovary

cells. J. Biol. Chem. 262, 4241-4247.

MATTEUCCI, M.D.,andCARUTHERS, M.H.(1981).

Synthe-sis of deoxyoligonucleotides on a polymer support. J. Am.

Chem. Soc. 103,3185-3191.

MAXAM, A.M.,andGILBERT, W. (1977).Anewmethod for

sequencingDNA. Proc. Nati. Acad. Sei. USA 74, 560-564.

MONACO, L., BOND, H.M., HOWELL, K.E.,andCÓRTESE,

R.(1987). Arecombinant apoA-I-proteinAhybrid reproduces the binding parameters of HDL to its receptor. EMBO J. 6,

3253-3260.

MOTT, J.E., GRANT, R., HO, Y.Z., and PLATT, T. (1985).

Maximizing gene expression from plasmid vectors containing

the

PL

promoter: Strategies for overproducing transcription

termination factor P. Proc. Nati. Acad. Sei. USA82, 88-92.

PIRAN, V.,andMORIN,R.J.(1979).Arapid radioassay proce-dure forplasmalecithimcholesterolacyltransferasebycovalent

chromatography. J. Lipid Res.20, 1040-1043.

ROSS, S.E., and CASRSON, S.D. (1985). Rapid

Chromato-graphie purificationofapolipoproteinsA-IandA-II from

hu-manplasma.Anal. Biochem. 149,166-168.

SANGER, F., NICKLEN, S., and COULSON, A.R. (1977). DNAsequencingwith chainterminatinginhibitors. Proc. Nati. Acad. Sei. USA74,5463-5467.

SEILHAMER, J.J., PROTTER, A.A., FROSSARD, P., and

LEVY-WILSON, B. (1984). Isolation and DNA sequence of

full-lengthcDNA and of the entiregenefor human

apolipopro-teinA-I. Discoveryofa newgenetic polymorphismin theapo A-Igene. DNA3,309-317.

SHAPIRO, D., BALLENTYNE, F.C., and SHEPHERD, J.

(1980).Comparisonofimmunonephelometryand electroimmu-noassay for estimation of plasma apolipoprotein A-I. Clin. Chim. Acta 103,7-13.

WU, A.L.,andWINDMUELLER, H.G.(1979). Relative contri-butionsbyliver and intestinetoindividual plasma

apolipopro-teins in therat. J. Biol. Chem. 254,7316-7322.

ZANNIS, V.l., KARATHANASIS, S.K., KEUTMANN, T., GOLDBERGER, G., and BRESLOW, J.L. (1983). Intracellu-lar and extracelluIntracellu-lar processingof human apolipoprotein A-I: SecretedapolipoproteinA-Iisoprotein2 isapropeptide.Proc.

NatiAcad. Sei.USA80,2574-2578.

Address

reprint

requests to:

Dr. A. Bollen Service de

Génétique

Appliquée

U.L.B.

ruede l'Industrie24 B-1400

Nivelles,

Belgium

Received forpublicationDecember27, 1988,and in revised form

Imagem

FIG. 1. Construction of plasmid pNIV1617. A 744-bp
FIG. 3. Immunodetection of proapo A-I on Western blots. Protein extracts from induced and noninduced bacteria carrying plasmid pNIV1617 were prepared for electrophoresis on NaDodS04-polyacrylamide gels
FIG. 5. Phenyl-sepharose chromatography. Peak III of
Table 2. Stimulation of Lecithin ¡Cholesterol Acyltransferase Activity by

Referências

Documentos relacionados

“adquirir, tratar, tornar acessíveis e disseminar os recursos de informação, mas também enriquecer, conservar e preservar as colecções bibliográficas que integram o

Após a realização deste trabalho é possível retirar algumas conclusões sobre a degradação de poluentes emergentes, nomeadamente o IBP, através do processo de

Functional expression and purification of bovine enterokinase light chain in recombinant Escherichia coli. Prep

The prevalence of virulence genes expressing fimbriae, production of hemolysin, colicin and aerobactin, was determined in Escherichia coli isolates from healthy cow’s genital

This section seeks to present an employed notation (3.1) and a proposed formulation (3.2), the Facility Location Problem and the application and measurement of a heuristic model

Lignin modifying enzymes from selected white rot fungi: production and role in lignin degradation. Purification and characterization of peroxidases from the dye decolorizing

Como pontos críticos desses programas foram observados: 1) falta de consulta aos professores sobre suas necessidades, bem como a impossibilidade de eles modifi- carem um curso

O nível de satisfação dos consumidores de jazz presentes nos eventos “Dia Internacional do Jazz” e “Uma Sexta de Jazz” também foram medidos através da escala de Likert com