Structure and Expression of Mouse α1-Acid Glycoprotein Gene-3 (AGP-3)

11, 4,

MaryAnnLiebert, Inc., Publishers

Pp.315-320

Structure and

_Expression

_of

Mouse

_cxi-Acid

_Glycoprotein

Gene-3

_(AGP-3)

CHING-JIN

_CHANG,

MING-YANG

_LAI,

DING-SHINN

_CHEN,

and SHENG-CHUNG LEE

ABSTRACT

The genome ofMus

domesticus

has

_multiple

genes of the

a,-acid

glycoprotein

(AGP).

Two_{cDNA clones}

wereidentified

_{corresponding}

to AGP-1 and AGP-2.

_Moreover,

two alíelesof AGP-1 exist in inbred mice.

The

_genomic

DNA _ofthe AGP-2gene has been cloned

and

studied. Herewe

_report

the

genomic

organiza-tion ofthree M.

domesticus AGP

_genes, the sequence

analysis

oftheAGP-3

_genomic

_DNA,

andthe expres-sion of the _{AGP-3 gene. The}

_major

structural differences between AGP-2 and AGP-3 genesare located in introns 1 _{and 5.} The _{low level} of

AGP-3

mRNA can be detected

by

the

_polymerase

chain reaction

(PCR).

The molecular basis of the low level

_expression

of

AGP-3

and the

_possible

classification of AGP-3 as a

pseudogene

arediscussed.

INTRODUCTION

Cci'AciD

glycoprotein

_(AGP)

is a

_single-chain

poly-peptide

withamolecular

_weight

of

_{approximately}

_44,000,

containing

about45%

carbohydrate.

This

_plasma

protein

is an

_acute-phase

reactant

synthesized

in the liver ofMus domesticus. Plasma concentrations of AGP in the mouse

increase several fold

_{during infection, inflammation,}

or are

induced

_following

subcutaneous

_injection

of

turpentine

or

lipopolysaccharide. Expression

of the AGP gene is regu-lated

_by

interleukin-1

_(IL-1),

tumor necrosis factor-a

(TNF-a),

interleukin-6

_(IL-6),

and

_{glucocorticoids}

(Bau-mann and

_Maquat,

1986;

Klein et

_{ai, 1987;}

Prowse and

Baumann,

1988;

Wonand

_{Baumann, 1990).}

The

_biological

function of AGP is

unknown,

but there are indications

that it may suppress the immune _response

_(Bennett

and

Schmid, 1980).

Some

_experiments

suggest that AGP is a

nonspecific

antiinfection agent

(Friedmann, 1983),

and

that it _possesses nerve

_{growth-promoting activity}

_(Liu

et

ai, 1988).

Multiple

forms ofmouse AGP canbedemonstrated

_by

two-dimensional

_gel

_{electrophoresis}

_(Baumann

_et

_ai,

1984),

anda

high degree

of

_{heterogeneity}

intheaminoacid

composition

of humanAGP has been described

(Schmid,

1975).

Denteet al.

(1987)

cloned and

_{sequenced genomic}

DNA segments

coding

for three humanAGPgenes. The

one

_designated

AGP-A,

_{is transcribed} to agreater extent

than AGP-B orAGP-C

(Dente

et

ai,

1987). Previously,

we

reported

the existence of at least two AGP _genes, AGP-1 and

_AGP-2,

in M. domesticus

(Lee

et

_ai,

_1989).

Both c/s-elements and

_{rra/is-acting}

factors

_reportedly

in-volved in the

_regulation

of AGP

_expression

havebeen

ana-lyzed

extensively

(Lee

et

_ai,

_{unpublished observations).}

One of

these,

AGP/EBP

(a

liver-enriched

_{transcription}

factor),

has been

_analyzed

and is

_key

to

_regulating

AGP

expression (Chang

et

_{ai, 1990).}

To

_study

the

genomic

orga-nization,

the molecular basis ofthedifferential

_expression,

and the

regulation

ofAGP

_expression,

we

analyzed

theM. domesticus AGP genes. We nowshow thattherearethree

AGP _{genes in} theM. domesticus _genome; we determined

the

_complete

_{sequence of}the

newly

identified_gene,

_AGP-3,

andshowthat AGP-3mRNA is

_expressed

at_verylowlevels.

MATERIALS

AND METHODS Isolation

_of

AGP

_genomic

clones

A

_{genomic library}

fromT

_lymphocytes

ofC57BL/6 M. domesticus andE. coliK803

(r",

m",

gal",

met")

were

gifts

of Dr. L. Mori

_(University

of

Milan,

_Italy).

The

_library

was _screened with an AGP-1 cDNA

probe (Lee

et

ai,

1989).

Of the

_{approximately}

4 x 105

_plaques

screened,

four

_{positive plaques}

were recovered and

_{purified by}

sec-ondary screening. Sequencing

wascarriedout

_by

the

dide-oxynucleotide

termination method

_(Sanger

et

_{ai, 1977).}

Institute ofBiological Chemistry, AcademiaSinicaand Institute of ClinicalMedicine,National TaiwanUniversity, Taipei,Taiwan.

316 _{CHANG ET AL.} RNase

_protection

assay

The RNase

_protection

_assaywas

performed

as described

by

Meltonetai

_(1984).

Total RNA

_{(20 ^g)}

was dissolved in 20

_fd

of

_{hybridization}

buffer

(60% formamide,

40 mM

PIPES

pH 6.7,

400 mM

NaCl,

1 mM

EDTA)

containing

the

_riboprobe

(2-5

x 105

cpm).

The

hybridization

mixture

washeated at 85°C for 5

min,

followed

by

incubation at

45°C foraminimumof 4 hr. Then300

/tl

of RNase

diges-tion buffer

(10

mMTris«HCl

pH

7.5, 5

mMEDTA,

300 mM

_{NaCl) containing}

40

_fig/ml

ofRNase Aand50 U/ml

of RNaseTl. The reaction wascontinued at 30°C for 30

min,

then

_{stopped by}

additionof20

_fd

of10%

NaDodS04

and50_figof

proteinase

K. After

_incubating

at37°C for15

min,

the mixture was extracted with

phenol/chloroform

and

_precipitated

withethanol

(with

20 fig carrier tRNA

in-cluded).

The sequences oftheRNA

samples

werethen

de-termined and

_analyzed.

Polymerase

chain reaction

Each

polymerase

chainreaction

_(PCR)

included_{10 ng of} cloned DNA

_fragments

or _{100 ng of}

_{single-stranded}

cDNA,

1.25 units of of

Taq polymerase,

100

fiM

deoxy-ribonucleotides,

2.5 nMeach

_{oligonucleotide primer,}

1.5 mM

_MgCl2,

50 mM

KC1,

and 10 mMTris-HCl

_pH

8.3 ina

final volume of 100

_¡d.

The reaction mixture was overlaid

with mineral

_oil,

andrun

_through

30

cycles

of 1 mineach

at

_94°C,

1 minat

_57°C,

and30sec at72°C. The

product

was

_analyzed

_by

_native20%

_{polyacrylamide gel.}

proximately

16

kb,

while the

XAgp-25

plaques

contained

an insert of

approximately

18 kb. Three distinct AGP

geneswereidentified

by

restriction

mapping

and

partial

se-quencing

ofthese five clones

_(Fig.

_1).

cDNAs

correspond-ing

to twoof

them,

AGP-1 and

AGP-2,

hadbeen

charac-terized

_previously

_(Cooper

and

_{Papaconstantinou,}

_1986;

Leeet

ai,

1989),

andathird_gene

designated

AGP-3was

identified. AGP-1 is contained in X clone

Agp-1

while

AGP-2and AGP-3 are in

_XAgp-25

clones.

Only

two _types of cDNAs and the

_{corresponding}

pro-teins of AGP have been identified in the M. domesticus

(Baumann, 1984;

Leeet

ai,

1989).

To address the issue of whether or not AGP-3 is

expressed,

the

_genomic

DNA

containing

the

_coding

and

_noncoding

_sequencesofAGP-3

was

_sequenced.

This _{gene spans} 4 kb from 548

bp

up-streamfrom the

_{putative transcription}

initiation siteto 130

bp

downstream ofthe

polyadenylation signal,

AATAAA

(Fig.

2).

This _gene has 6 exonsand 5

introns,

a structure

analogous

tothe rat

_(Liao

et

_{ai, 1985;}

Reinkeand

Feigel-son,

1985),

human

(Dente

et

_ai,

1985; Dente et

_{ai, 1987),}

and mouse AGP-2 _genes

(Cooper

et

ai, 1987;

the

Papa-constantinou_group

_designates

it

AGP-1,

while our

desig-nation is

_AGP-2).

Theconsensus _{sequence for}

_splice

junc-tions is

preserved

in both AGP-2 and AGP-3 genes.

The mouse AGP-3 and AGP-2 _gene

(Cooper

et

ai,

1987)

sequences are very

similar; however,

there are two

major

differences between them: AGP-3 lacks 86

nucleo-tidesin intron 1

_(between

+331 and

+332),

whileAGP-2

has an additional

(GT)28

tract in intron 5

(+2815),

not

found in AGP-3.

RESULTS

Isolation and

characterization

of

M.

domesticus

AGP

genomic

clones

Several

_genomic

clones were obtained

by

_screening

the

EMBL-3

_{library (C57BL/6)}

with an AGP cDNA

_probe.

Four

_{positive plaques}

of

XAgp-1

containedaninsertof

ap-Differential

expression of

AGPgenes

It would be

_interesting

to know the relative level of mRNA

_expression

among

AGP-1, -2,

and -3 genes in M domesticus liver under normal conditions and

_during

the

acute-phase

reaction. Becauseofthe

_high

degree

of

simi-larity

amongthesethreegenes, anRNase

protection

analy-sis was

performed

to discriminate _among these mRNAs.

.lkb.

xAgp-1

B EHB

kX-xV^v\

X^ SH

AAgp-25

^\S\S\Vsl

35

BS A9P-2

^5

NNNXNXNXWI B BS BE 0.5kb 12 3 4 5 6

FIG. 1. Partial restriction_{map of«,-acid}

_glycoprotein

_{genomic clones,}

_XAgpl

_and

_XAgp25.

_{Shaded boxes represent}

AGP-1,

-2,and -3 _genes.Exonsare

_represented

by

solid boxes. Abbreviations: B,Bam

HI; E,

EcoRI;

H,

Hind III;

S,

Sail.

-548 GGATCCTTTCCTGCTGTAGATACTGGGAGCTTTGCTGAACTAGATGTTCAAGTCA -4 93 GAATCAACCCTTTTTGGGCATTTGGATGCCTCTAGGCTGGGAAGGGGTCCTCAGG -438 AACATCACACTCCTTTGGAAACTAATCCATCTTTGTCCTTGGGCCTTAACTTGAG -383 CCCCTAAGTGTCTTCTATGTTCACTATGAACTTGACCTGGGACCCCTTCTTATCA -328 TGCTTGGGGGCGGGTTGATGTATGTGTAGGTTTCACTCCTGCTAGGCAGCTTCAT -273 GGGATAAGAGAGGGTGGGGACCACTGTCTGGGACCTAAGTATCATCAGGCTACCC -218 TGTACCCACCTTGACCATGAATCAGCCACTCTGGTGTAGGGCAGGAGCCTGTGTC -163 ACAGCCAGCTGGCTGAGAGAGCTGCACAAAGCTGGCTTGAGGGAACATTTTGCAC -108 AAGACATTTATCAAGTGCTGGTGAGTTTGTGGCACTGCTCTACAGCCCCTGGCTG +1 -53

CAGTCCCATGCCCTCCCCACATCCTGTtTATAÄJAAGCCACTGCACCCTCCAGCCAC

Exon 1 3 CAGTTATCTCTTCCAAGCTCTGGTGCCTCTGAGTATCCTCAGCATGGAACTACAC MetGluLeuHis 58 ACGGTTCTCATCATGTTGAGCCTCCTGCCACTGTTGGAAGCTCAGAACCCAGAGC ThrValLeuIleMetLeuSerLeuLeuProLeuLeuGluAlaGlnAsnProGluH 113 ATGCCATCAACATAGGCGACCCTATCACCAATGAGACCCTGAGCTGGGTAAGTGT isAlalleAsnlleGlyAspProIleThrAsnGluThrLeuSerTrp 168 CTGCCCGGGGCCTGGACCTGTTCACTGTAGGATTCTACTCTTTCCTCTGGGCTTT 2 2 3 CCCTTCCCTGGTGTCTGTGTTCAGCTCTGGGCTCCTGGTACTGCCCTTCCCACTT 2 78 GTGTATACCCTGGTGGCATCCCCCTGTCTCCAAAGGCAGAATCATCACTCTGAGC 333 CATAGCTTGCCTGCCCCTCATCGTGGATGAATGCCAAGGTCCTCACTACAAGGCC Exon 2 388 TGCTCATCTGTGTGCCTGCTTCTCCCCAGCTATCTGGCAAATGGTTTCTCATTGC LeuLeuGlyLysTrpPheLeuIleAl 44 3 TGTGGCTGACTCAGACCCTGATTATAGGCAGGAAATTCAAAAGGTACAGACTATA aValAlaAspSerAspProAspTyrArgGlnGluIleGlnLysValGlnThrlle 4 98 TTTTTTTACCTTACCCTAAACAAGATAAATGACACGATGGAGCTTCGAGAGTATC PhePheTyrLeuThrLeuAsnLeulleAsnAspThrMetGluLeuArgGluTyrH 553 ACACCAAGTGAGTCCTTGTAACAGCCAGCCCACCCTGGCCCTGGCTTCCACTCCC isThrLy 608 AGATTCCTAGAGACCTGAGCAAACTGGCTCTGCCTGGCCTCCCCACCCACTTTCA 663 GAAATGGGGACAGCTGTCTTGCCTCTTGCCCCCTTCTACCCTGGGCTAGTCAGAT Exon3 718 CACCTCTCCATCAGTTGTCCTCTCTCTTTGCTTTTTAGAGATGACCACTGTGTCT sAspAspHisCysValT 77 3 ATAACTCCAACCTTCTGGGATTCCAGAGAGAGAATGGGACCCTCTTCAAGTATGG yrAsnSerAsnLruLeuGlyPheLeuArgGluAsnGlyThrLeuPheLysTyrG 82 8 TGAAGGTGTGAACTCACTCCTTTCTGGGAGGGTATTGCCAGTTCTGAGGGGACAG 88 3 CAGAACAGGGCAGTTTGGTCTGTCAAGTCACTCTCTGGGGCTTGTAAGTGGACGT 93 8 AGATTTCAAACTGGAGTCCAGCTGCCAGGCCTGACTTGTTTTGATCAGGCTTTAT 993 GACCCTTCGTGACCTCAGAGTGGAAAGCCATGGGTTGGGAAACCAGTGACTTTAG 104 8 CCACACCCCCAGGTCACCAGGGACAGCATGGAGGAGAACAACCCAATTGCTGGTG 110 3 GGCCAGATCAGACACTGGTTAGCTTTTAATTGTCCTAAGCAGATGTTTTGATGTT 1158 TAAGGGGAAGTGTTAACATTATTATTGCCCCACCACCACACCACCAAGGGCCCTT 1213 TCCAAGGCCCCAGCTCTTCCATACCTATGAAGAATGAGAAGTGAGGCTTGCATCC 12 68 AGCATGAGGCTGAGCACATGGCAGCCTCAGGGAGCCCCAGGCACTTGCCATAGCT 1323 ATGTGCTTCCTTCCCTTGGGGATGTGAACCACATCATCATTCTAGTGACTCACAA Exon* 1378 AGACCTTCCTCCAACAGAAGGAGAAGTAGAAAACCCTTCTCACCTGAGAGTGCTA luGlyGluValGluAsnProSerHisLeuArgValLeu 14 3 3 GAGAAACATGGGGCCATCATGCTTTTCTTTGACCTGAAGGATGAGAAGAAACGGG ArgLysHisGlyAlalleMetLeuPhePheAspLeuLysAspGluLysLysArgG 1488 GACTGTCCCTGAGCGGTAGGGTCCTCTCATCCCTGGTGCGCCCAGCTCAACTGGC lyLeuSerLeuSerA 154 3 CTTACTCTTGGTCACCCACCCACATCCCACATCCCTACCTGGCCTCCCATCTACT Exon 5 15 98 GGACCCATAGCCAGCATAACTTTGGATCCCTTCTCCCATGCAGCTAGAAGGCCAG laArgArgProA 1653 ATATCCCCCCGGAGCTGCGGGAAGTATTTCAGAAGGCTGTCACACACGTGGGCAT spIleProProGluLeuArgGluValPheGlnLysAlaValThrHisValGlyMe 17 08 GGATGAATCAGAAATCATATTTGTCGACTGGAAAAAGGTAAATGAAGGAGGCTGT tAspGluSerGluIlellePheValAspTrpLysLys 17 63 ACGATACCACCCCAGCAGTGCGCCCAGTGTCAGTGACCCTAGAGGCTCAGAGAGG 1818 GCAAGCTTCTGGTTAAGGCAGCTCAGCGAGGCAGGTATCTTGTTAACTCTCCTGC 187 3 CTCCTCCTCATCAGGAGATCACAGAGACCCTAGATGGGCAGTGAGCCTCAGGGAG 192 8 GTGAAGTTAAGTAGGAGGTCCTGGAAAGCTTGTGGAGGATAAGAGGAAGATCAGG 198 3 AGGGTCACTTAGGGAACAGCCAGTGCCAGGGTGCCAGGTTTCTTCCTGTCCTTCA 2 0 3 8 TATTACTACCTTTTCAAGCAGGAGTTTTGATTGACATCTTCCATGTCACCCCAAC 2 0 93 TCCAGCAAGCCCGATGGCTTTGATAGGCAGGGTTGACCACACTGAGACTCTTGAT 214 8 GTCCGGTCTACACATTGTGCAGAGGGAGAGGCAGCATCAGTTTTGTTTTTCACCC 22 03 ATGGCGAATGCATGGGATCAAACAGTCACCTTGCATGTAGTTTAAGATACTCAAT 22 58 AGCTTTTGTAACTTCATTCTCTGGTCACCTGAGCCTTTCCTGGCATCATCACCAG 2313 CCCCAGGATTCCCGGGAGAGGTGCCTGCACACAGACACTGCCATTCACAGCATGA 2 368 CTTCCACCCACACCAGTGGGCCAGTAGACTCATCCTGCACCTGTGGACAGAAGTG 2423 TTAGATAATGCCTGCCCTTTGGGGATTCTGCTCACAATCAATGGGTGAATAAGCC 2 478 GGAGCTCAGAGATGAGGGACAACTTACCCTAGACTAGTGGTTCTCTAGGATGGGT 2 5 33 CTCAACCTGTATGTCAGATGTCCTGCACGTCAGGTATTTATTGATTTATAATAGT 2588 AGCATATAATTACAGTTATTGAAGTAGCAATGAAATCATGGTAGGTGATCACCAC 2 64 3 AACACAAGAAACTGTGTTAAAGTTTTGCATTATTAGGAAGGCTGAGAACCACTGT 2 6 98 CCTGCCACTGCAGGGAGCCATGGCAGATCTAAGACACATCTGGTTGACACTACCG 2753 GGCCATTTTGACCAACAACAGTACTCCCCCCAACCCACCTCACAATAGGTGTATT 2808 _{CATAGCTAG_TGTG_T_G;rGC_A_TGT_GTGA_C_TGTG_TGGGTACACAAGCATGCTATAAGA} 2 863 CATGTGTGGATGTCAGAGGACACCTGTGGGCTATGTCCTCTTCTACCATTCTCTC 2918 CTGGGCTCTGGTTAAGGCTGGGTTGGCTTCAAGCTGCCCCTCAGGCTTACCTACC 2 973 TTGCCATTTTTTTTTGTTGTTCTGTCCTGTTTTTTCTGTTTTGTTTTGTTTTTGT 3028 ATTTAATCTTGCAGCCCAGGCTACTCTACTGCAACTCATAGCAATCCTCTTGCCT 30 83 CAGTATTCATCAACCCTGGTGTGTGCCACCAGCCCTGGCTTACTCACTCTGCTCT Exon6 3138 CCTCCCTGATATCTTCCAGGACAGGTGCAGTGAACAGGAAAAGAAGCATCTTGAG AspArgCysSerGluGlnGluLysLysHisLeuGlu 3193 TTGGAGAAGGAGACCAAGAAAGATCCTGAGGAAAGCCAGGCATGAACTCAGCTCT LeuGluLysGluThrLysLysAspProGluGluSerGlnAla 3243 CTGGTCTCCTTGGGCTGTCCCCATGTGTACCACACCCTACCCCATCCTGGTCACT 3303 TTGATTCTGTCTCTGTAAdAATAÄÄfeGTTTGCTGACACTGTCAATATCATTTCTT 3358 TGCTCCCTTCCTTTTCCTCCCTCCCTCCCTCCCTTCGTGGAGAGTCTTGAGTGGA 3413 GCTAGCTAAGTCAATAACCCTGCCAGGAATTCGAAAGGCTCT

FIG. 2. Nucleotide _{sequence of}mouse AGP3 gene. The

putative

TATA box and

_poly(A)

addition

_signal

sequence

(AATAAA)

areboxed. Sixexonsandencoded amino acid

sequencesareshown. The siteof initiationof

transcription

is

_{depicted by}

+1.

The

_{specific riboprobes}

of AGP-1 and AGP-2

_(Fig.

3a,

upper

panel)

was

designed

from their cDNA sequences

(Lee

et

ai,

1989).

The

_riboprobe

of AGP-3 was derived

from the

_genomic

segments that contains the

putative

exon

4

_(Fig.

3a).

The result of RNase

_{protection analysis}

is shownin

_Fig.

3a. When

using

RNAfrom normalliver and

from

_{lipopolysaccharide-stimulated}

liver,

protected

bands

weredetected

_{corresponding}

to 247

bp

ofAGP-1 and 318

bp

ofAGP-2

_(Fig.

3a, lower

panel,

lanes _1, 2orN-Land

LpS-L).

However,

there were no

signals

for the

probe

de-rived from AGP-3. The

_expression

level of AGP-1 isabout fivefold

_higher

than that of AGP-2 and both _genes

re-spond

tothe

acute-phase reaction,

thusAGP _appearsas a

liver-specific

gene, because no

_signal

wasdetectedin RNA

from the

spleen.

Because the level of

_expression

of AGP-3 is much less than AGP-1 and

_AGP-2,

a more sensitive method

(e.g.,

PCR)

wasusedtodetectit. PrimersforAGP-2 andAGP-3

were

synthesized

for PCR

experiments (Fig.

3b,

upper

panel).

Using

genomic

DNAofAGP-2orAGP-3_genesas

templates,

we showed that the

primers

derived from

AGP-2 and AGP-3 were

_specific

for the

_{corresponding}

templates

(Fig.

3b,

lower

_panel,

lanes

_{1, 2, 4,}

5). However,

when these

primers

were used for reverse

_{transcriptase}

(RT)-PCR

using

RNAderived from M. domesticus

liver,

an

AGP-2,

but not anAGP-3

fragment

canbeseeninthe agarose

gel by

ethidium bromide. This does not exclude

the

_possibility

thatthelevel of

_expression

ofAGP-3

_might

318 CHANG ET AL. riboprobe protectedlength RI Anp-2 437

|

^-(390ni) (318ni) riboprobe protected length

^g

WiW (570ol) riboprobe

(108) protectedlength _{Agp2 primer R: 5'CCATGACAAGAATCATGTGC3'}

+67 +52 N-L LPSL LPSS Probe 123123123123 L: 5'ATCTCTTCCAAGCCCTG3' +8 +24

Agp3 primer6F H R: 5' GAACCGTGTGTAGTT 3'

+64 +50

L: 5'ATCTCTTCCCAGCTCT 3'

+8 +24

1 2 3 4 5 6 7 8

FIG.3. a.

Upper panel.

RNase

_protection

assay.

Specific riboprobes

andtheir

protected length

for

AGP-1, -2,

and-3 are

represented.

The

riboprobes

of AGP-1 and -2are derived from their cDNAs in

pGEM3

vector.The

fragments

are

numberedrelative to the mRNA

_{transcription}

initiation site indicated

_by

+1. The

riboprobe

of AGP3 containsexon4

andintronsegmentsin

pGEM

4vector. Lower

_panel.

RNase

_{protection experiment}

onRNA

prepared

frommouseliver and

_spleen.

1, 2,

and 3 indicate the

riboprobes

derived from

AGP1, 2,

and 3.

N-L,

LPS-L, and LPS-S indicate the

sources of RNA from normal

liver,

LPS-stimulated

liver,

and LPS-stimulated

spleen, respectively. Specific protection

bands are

represented by

arrows, b. PCR

_{amplification}

of AGP-2 andAGP-3. The

_right

(R)

and left

(L) primers

of

AGP-2 and AGP-3 are shown above, and the PCR

_{products analyzed}

on 20%

polyacrylamide

gel

are shown below. Lanes 1-6, Ethidium bromide

_staining

patterns; lanes7 and

8,

autoradiographic

patterns. Lanes

1-3,

7, AGP2

_primer;

lanes4-6, 8, AGP3

_{primer. Templates}

used: lanes 1 and4,AGP2

_genomic

DNA; lanes 2 and5, AGP3

_genomic

DNA; lanes 3, 6, 7, 8,

single-stranded

cDNAderived from mouseliver.

To overcome

_this,

5'-end labeled

_primers

were

employed

for RT-PCR. As demonstrated

by

this

analysis,

the

_signal

for AGP-3 isatleasttwoorders of

magnitude

weakerthan AGP2

_(Fig.

3b,

compare lower

panel

lanes 7 and

8).

DISCUSSION

We have isolated the

_genomic

clones for the entire

se-quenceof three ofthe M. domesticus AGP genes, AGP-1,

-2,

and -3. The

_newly

identified gene,

AGP-3,

is located

approximately

6 kbupstreamfrom AGP-2.

By

the restric-tionmapof the X clones of

Agp-1

and

Agp-25

and the

evi-dence that AGP-1is

_proximal

tothecentromere

(Baumann

et

ai,

1984),

we

_predict

the

_{genomic organization}

of these three _genes is:

AGP-1, AGP-3,

and

AGP-2, arrayed

in

tandem. AGP-3 and AGP-2 are

_closely

linked while there

is some distance

_{(not determined)}

between AGP-2 and AGP-3.

AGP-1 AGP-2 AGP-3 AGP-1 AGP-2 AGP-3 AGP-1 AGP-2 AGP-3 AGP-1 AGP-2 AGP-3 AGP-1 AGP-2 AGP-3 AGP-1 AGP-2 AGP-3 AGP-1 AGP-2 AGP-3 AGP-1 AGP-2 AGP-3 CGGCAGGAGTCTGTGTCAGGACCAGT GGGCAGGAGTCTGTGTCAGGGCC-GG GGGCAGGAGCCTGTGTCACAGCCAGC -ieo -170 _16° AGGTTGAGGGAGCTGCATAAAGCTGG -CTGCGAGGGAGCTGCACAAAGCTGG TGGCTGAGAGAGCTGCACAAAGCTGG -ISO -140 C T T G A G G S A A íi A T T ! T G C G C A A G A C A CTTGAGAGAACATTTTGCGCAAGACA CTTGAGGGAACAT TTTG CACAAGA C A -130 -120 -110 TTTCCCAAGTGCTGGTGAGATTGTG- TTTCCCAAGTGCTGGGGAGAATGTG-TTTATCAAG TGCTGGTGAG T T T G T G G -100 -90 --C-CACAGCTCTACTGTCCCTGGCT --C-CAGGGCTCTAGAGGCCCTGGCT C A C TCACTGCTCTACAGCCCCTGGCT TCAGTCCCATGCCCTCCCCACATCG-GCAGTCCCATGCCCTCGCCACATC-T GCAGTCCCATGCCCTCCCCACATCCT -40 -30 A T A A A T A A A T A A ATGTTGCTGCACCATCC AAGCCACTGTACCCTC-AAGCCACTGCACCCTCC -20 -ÏÔ AACACCAATGATCTTTTCCCAGGCCT --CACCAGTTATCTCTTCCAAGCCCT GCCACCAGTTATCTCTTCCAAGCTCT +1 10 +20

FIG.4.

Comparison

of_sequencesin thepromoter

region

ofmouseAGP genes. The TATAboxisboxed. The

num-bers are indicated

_according

to the AGP3 _sequence. The

AGP/EBP

_binding

motifsareindicated

by

solid

underlin-ing,

GREs are indicated

by overlining,

and theconsensus

acute-phase

sequence is indicated

by

dashed

underlining.

AGP-3 mRNA

_by

RT-PCR. AGP-3 mRNAwas detected

in this

experiment,

but the

signal

was weaker than thatof

AGP-2.

Therefore,

the AGP-3_geneis transcribedata_very lowlevel. The

_expression

of AGP-3

_protein

isyettobe

de-termined.

We do not know

why

AGP-3 is so

poorly expressed.

One

_possibility

isthat structuralaberrationsarepresent in the introns.

_Comparing

AGP-3and

_AGP-2,

the

_major

dif-ferences are the lack of 86

_bp

of intron 1 and no

(GT)28

tract in intron 5. The 86

_bp

of intron _{1 may} contain the

branch site of

splicing;

the deletion _{of this sequence may}

cause AGP-3 RNA to be

spliced inefficiently.

Another

possible

explanation

is that

alternating purine-pyrimidine

sequencescanformZ-DNA,which_mayintroducea poten-tial sitefor_gene

regulation.

_{Alternatively,}

the low-level

ex-pression

may be due to the

positional

effect. Proudfoot

(1986)

has shown that

_{transcription}

ofthe first _gene in a genecluster interfereswith the

transcription

ofthe

follow-ing

gene. It is

possible

that the

transcription

of AGP-1 gene interferes with the

transcription

of AGP-3 gene.

Two AGP

proteins

have been identified

_{corresponding}

to the

products

of AGP-1 and

AGP-2;

a

protein

product

ofAGP-3 has not been identified.

_By

_comparing

the de-duced amino acid _sequenceofAGP-3and those of AGP-1 and AGP-2

_(Fig.

5),

AGP-3 would contain 206 amino

AGP-1 MALHTVLIILSLLPMLEAQNPEHAN 25

AGP-2 _{MALHMILVMVSLLPLLEAQNPEHVÑ}

AGP-3

MELHTVLIMLSLLPLLEAQNPEHA-upstream from the

putative

transcription

initiation siteto

130

_bp

3' ofthe

polyadenylation

signal.

This_genecontains

6 exons and 56 introns. Its

coding

sequences are normal

and contain no frameshift or nonsense mutations. The

exon/intron

_splicing

_sites,

andthe

_5'-flanking

and

3'-flank-ing

sequences are

conserved,

as

compared

to the AGP-2

gene. The

5'-fianking

regions

of three AGP_genesare

simi-lar inthefirst 180

bp

upstream from thecap

site,

asshown in

_Fig.

4. AGP-3aswellasAGP-1 andAGP-2contain the.

potential glucocorticoid-responsive

element

_(-125

to

-113)

and threeAGP/EBP

_binding

sites locatedat -119

to

-110,

-107 to

_-98,

and -87 to -78

_(Chang

et

ai,

1990).

A sequence of 38

bp

(-15

to

₊₂₃₎

issimilarto se-quences observed in the three human

acute-phase proteins

(Dente

et

_{ai, 1985)}

located inAGP-3_gene.Part ofthe

reg-ulatory

region (-180

to

+60)

of these three _genes were

separately

fusedto the

_{chloramphenicol}

_acetyl

transferase

(CAT) (Gorman

et

_ai,

1982)

reportergeneand then

trans-fected intoaBHKcell line. The

expression

was monitored

by assaying

the

_activity

ofthe CAT enzyme. The

_activity

of AGP-3 promoter is similar to those of AGP-1 and

AGP-2

(data

not

_shown).

Using

RNase

protection

to_{assay the}

_endogenous

expres-sion ofthethree AGP_genes, weshowedthat theamount

ofAGP-1 wasabout fivefold

_higher

thanAGP-2 ina

nor-malM. domesticus liver and

_during

acute-phase

reaction.

However,

the

expression

ofAGP-3 could not bedetected

by

the RNase

protection

assay. Todetermineif the AGP-3 gene is a bona

fide pseudogene

or

_just

_expressed

at very low

level,

we

synthesized specific

_primers

to

_amplify

AGP-1 FTIGEPITNETLSWLSDKWFFMGAA 50 AGP-2 ITIGDPITÑETLSWLSDKWFFIGAA AGP-3 INIGDPITNETLSWLLGKWFLIAVA AGP-1 FRKLETRQAIQTMQSEFFYLTTNLI 75 AGP-2 VLNPDYRQEIQKTQMVFFNLTPNLI AGP-3 DSDPDYRQEIQKVQTIFFYLTLNLI AGP-1 _{NDTIELRESQTIGDQCVYNSTHLGF 100} AGP-2 NDTMELREYHTIDDHCVYNSTHLGI AGP-3 NDTMELREYHTKDDHCVYNSNLLGF AGP-1 QRENGTFSKYEGGVETFAHLIVLRK 125 AGP-2 _{QREÑGTLSKYVGGVKIFADLIVLKM} AGP-3 LPEÑGTLFKYEGEVENPSHLRVLRK AGP-1 HGAFMLAFDLKDEKKRGLSLYAKRP 150 AGP-2 HGAFMLAFDLKDEKKRGLSLNAKRP AGP-3 HGAIMLFFDLKDEKKRGLSLSARRP AGP-1 DITPDLRDVFQKAVTHVGMDESEII 175 AGP-2 DITPDLRDVFQKAVTHVGMDESEII AGP-3 DIPPDLRDVFQKAVTHVGMDESEII AGP-1 _{FV3WKKDRCGQQEKKQLELGKETKK 200} AGP-2 FVDWKKDRCSQQEKQQLELEKETKK AGP-3 FVÛWKKDRCSEQEKKHLELEKETKK AGP-1 D P L £ G Q A AGP-2 D P E E G Q A AGP-3 D P E E S Q A

FIG. 5. The amino acid _{sequence of} mouse

_AGP-1,

-2,

and the

putative

AGP-3. Dots above the

_{single-letter}

amino acid mark the

_position

of

_{putative glycosylation}

sites

(sequence

_{Asn-X-Ser/Thr).}

320 CHANG ET AL.

acids,

including

the 18-residue

_{putative signal peptide.}

Thereare45amino acid substitutionsbetween the encoded AGP-3

_protein

and that ofAGP-1 or AGP-2. AGP is a

highly

glycosylated protein; therefore,

itwasofinterestto

localize

_potential

carbohydrate

attachment sites indicated

by

the_sequenceof Asn-X-Thr/Ser. Five

potential

sitescan

be foundin the AGP-1 and six

potential

sitesin

AGP-2;

however,

only

three

_{potential glycosylation}

sites existedin

the

_putative

AGP-3

_polypeptide.

If AGP-3 _expresses a

functional

_protein,

theaminoacidsubstitutions would af-fectitsfunction. The existenceand the

potential

functional

implications

of AGP-3

_protein

remained to be

investi-gated.

ACKNOWLEDGMENTS

We thank Dr.

George

Bolton for

_editing

and Ms. Joanne Kahrmannand Ru-Ju Chen for

_typing

this

manu-script.

This researchwas

supported by

Grant NSC80-0412-B002-09 from the NationalScience Council.

REFERENCES

BAUMANN, H., and _MAQUAT, L.E. _(1986). Localization of DNA _sequence involved in dexamethasone-dependent expres-sion of the rat_{al-acid glycoprotein gene.} Mol. Cell. Biol. _6, 2551-2561.

BAUMANN, H., HELD, W.A.,andBERGER,F.G._(1984).The

acute _{phase response of} mouse liver. Genetic analysis of the majoracute _phasereactants. J. Biol. Chem. 259,566-573.

BENNETT, M., and SCHMID, K. _{(1980). Immunosuppression} byhumanplasmaod-acidglycoprotein:Importanceof the

car-bohydrate moiety. Proc. Nati. Acad. Sei. USA77,6109-6113.

CHANG, C.J., CHEN, T.T., LEI, H.Y., CHEN, D.S., and

LEE, S.C. (1990). Molecular_cloningofa_{transcription}factor,

AGP/EBP, that _belongs to members of the C/EBP family. Mol. Cell. Biol. 10, 6642-6653.

COOPER, R., and PAPACONSTANTINOU, J. (1986). Evi-dence for the existence ofmultipleai-acidglycoproteingenesin the mouse. J. Biol. Chem.261, 1849-1853.

COOPER, R., _ECKLEY, D.M.,and PAPACONSTANTINOU, J. (1987). Nucleotidesequenceof themousea,-acid

glycopro-teingene 1. Biochemistry26, 5244-5250.

DENTE, L., GILBERTO, G., andCÓRTESE, R. (1985).

Struc-tureof the humanai-acidglycoproteingene:Sequence homol-ogywith other humanacute_{phaseproteingenes. Nucleic Acids}

Res. 13, 3941-3952.

DENTE, L., PIZZA, M.G., METSPALU, A., and CÓRTESE,

R._{(1987).Structure andexpressionof thegenescoding}for

hu-mana.-acid _{glycoprotein.}EMBO J. _{6, 2289-2296.}

FRIEDMAN, M.L. _{(1983). Control of malaria virulenceby}

a¡-acid _{glycoprotein (orosomucoid),} an acute phase (inflamma-tory)reactant. Proc. Nati. Acad. Sei. USA80, 5421-5424.

GORMAN, CM., MOFFAT, L.F.,andHOWARD,B.H.(1982).

Recombinant _genomes which express chloramphenicol acetyl-transferase in mammalian cells. Mol. Cell. Biol. 2, 1044-1051.

KLEIN, E.S., REINKE, R., FEIGELSON, P., and RINGOLD,

G.M. (1987). Glucocorticoid-regulated expressionfrom the 5'-flanking region of therat aracid glycoproteingene. J. Biol. Chem. 262, 520-523.

LEE, S.C, CHANG, C.J., LEE, Y.M., LEI, H.Y., LAI, M.Y.,

and CHEN, D.S. (1989). Molecular cloning of cDNAs

corre-spondingto two_{genes ofaracidglycoprotein}and characteriza-tion oftwo _{alíeles of AGP-1 in the}mouse. DNA8, 245-251.

LIAO, Y.C, TAYLOR, J.M., VANNICE, J.L., CLAWSON, G.A., and SMUCKLER, E.A. (1985). Structure of therat

a,-acid glycoproteingene. Mol. Cell. Biol. 5, 3634-3639.

LIU, H.M., TAKAGAKI, K.,andSCHMID, K. (1988).In vitro nerve-growth-promoting activityofhuman_plasmaa,-acid gly-coprotein. J. Neurosci. Res. 20,64-72.

MELTON, D.A., KRIEG, P.A., REBAGLIATI, M.R.,

MANI-ATIS, T., ZINN, K., and GREEN, M.R. (1984). Efficient in vitro_synthesisof_biologicallyactive RNA and RNA

hybridiza-tionprobesfromplasmids containingabacteriophageSP6

pro-moter. Nucleic AcidsRes. 12, 7035-7056.

PROUDFOOT, N.J.(1986).Transcriptionalinterference and

ter-mination between_{duplicated a-globin gene}constructs _suggests

anovel mechanism forgeneregulation. Nature322, 562-566.

PROWSE, K.R.,andBAUMANN,H.(1988). Hepatocyte-stimu-lating factor, a¡-interferon, and interleukin-1 enhance expres-sion of therat <x,-acidglycoproteingeneviaadistal upstream

regulatory region. Mol. Cell. Biol. 8,42-51.

REINKE, R., and FEIGELSON, P. _(1985). Rat a,-acid glyco-protein. Gene sequence and regulation by glucocorticoids in transfected L-cell. J. Biol. Chem. 260,4397-4403.

SANGER, F., NICKLEN, S., and COULSON, A.R. (1977).

DNA_sequencingwith_{chain-terminating}inhibitors.Proc.Nati. Acad. Sei. USA74,5463-5467.

SCHMID,K.(1975). InThe PlasmaProteins,vol. I. F.Putnam,

ed. (Academic Press,New _York) pp. 184-228.

WON, K.-A.,andBAUMANN,H.(1990).The_{cytokineresponse} element of the rat _{a,-acid glycoprotein gene is} a complex of

several _{interacting regulatory sequence.} Mol. Cell. Biol. 10,

3965-3978.

Address

_reprint

_requests to:

Dr.

Sheng-Chung

Lee Institute

_of

Biological

Chemistry

Academia Sínica

Taipei,

Taiwan

Received for publicationOctober _{21, 1991,} and in revised form December 3, 1991.