Gene, 98 (1991) 209-216

209

Elsevier

GENE

03901

Primary structure of the rat gene encoding an inhibitor of the insulin receptor tyrosine kinase (Recombinant

DNA;

chromosome

mapping;

promoter)

Laurence Falquerho a, Gilles Patey a, Laurent Paquereau a, Vakie Szpirer ‘, G&an Levand and Alphonse Le Cam a

Rossi%, Olivier Lahuna b, Josiane

Szpirer ‘, Claude

” Centre CNRS-INSERM de Pharmacologic-Endocrinologie, Montpellier (France); ’ Unit6 99 de I’INSERM, H6pital Henri Mondor, CrCteil (France) Tel. (33)49813544: ’ DPpartement de Biologie Mokculaire, UniversitC Libre de Bruxelles, Rhode-St-Genkse (Belgium) Tel. (32102 358 35 30; and ‘I Department of Genetics, University of GBteborg, Giiteborg (Sweden) Tel. (46)3185 32 90 Received by J.-P. Lecocq: 19 July 1990 Revised: 21 September 1990 Accepted: 26 October 1990

SUMMARY

The gene (PP63) encoding the inhibitor (PP63) of the insulin receptor tyrosine kinase was isolated from a rat genomic library. The intron/exon organization was deduced from Southern-blot analysis and sequence data (i.e., the exons + the boundaries). The PP63 gene, which maps to chromosome 11, spans approx. 8 kb and contains seven exons separated by six introns of different sizes. All of the boundaries match the consensus GT/AG sequence for donor and acceptor splice sites. Primer extension and Sl mapping experiments were used to locate the transcription start point (tsp) 73 nt upstream from the translational initiator. Both in vitro transcription assays and transcription of a chimeric gene in intact hepatoma cells indicated that the sequence located immediately upstream from the tsp contained a promoter. Several putative &-regulatory elements, including a TATA box and a C/EBP-binding site were found within the 250 bp preceding the tsp.

INTRODLJCTION

A large body of evidence has led to the hypothesis that protein phosphorylation of Tyr residues plays a major role in the control of cell growth and proliferation (Hunter and Cooper, 1985). Many oncogene products and growth factor Correv/~or2c/~,lcp to: Dr. A. Le Cam. Centre CNRS-INSERM cologie-Endocrinologie.

rut de la Cardonille.

Tel. (33)6754428X;

Abbreviations:

aa; amino

pair(s);

CAT,

encoding

CAT;

cDNA,

acid(s);

or 1000 bp;

AHSG,

chloramphenicol DNA

CCAAT/enhancer-binding otide(s);

complementary

protein; MIS,

NF-1, nuclear

HNF-I,

Mtillerian

factor

receptor

NaCI/O.O15 M Na, tion start point(s):

TyK);

human

a2-HS-glycoprotein;

acetyltransferase; hepatocyte

inhibiting

car, gene

to mRNA;

C/EBP,

NF-1; kb, kilo-

substance;

nt, nucle-

1; oligo, oligodeoxyribonucleotide;

open reading frame; PP63, phosphoprotein insulin

(France)

Fax (33)67542432.

bp; base

base(s)

de Pharma-

34094 Montpellier

PP63,

gene

of63 kDa (the inhibitor encoding

PP63;

SSC,

ORF, ofthe 0.15 M

citrate pH 7.6; SV40, simian virus 40; fsp, transcripTyK, tyrosine

kinase.

0378-I 119,‘911$03.50 0 1991 Elsewer Science Publishers B.V. (Biomedical Division)

receptors exhibit a TyK activity which is essential for their biological function (for a review, see Heldin et al., 1987). Although many laboratories have attempted to elucidate the mechanisms that regulate this enzyme activity, little is known, more particularly with respect to the existence and the role of putative natural effecters (i.e., activators and inhibitors). Only two examples of biological molecules which are capable of inhibiting TyK activity have so far been described. The MIS was found to inhibit the epidermal growth factor receptor TyK (Coughlin et al., 1987). PP63, a phosphoprotein secreted by normal rat hepatocytes (Le Cam et al., 1985) was shown to block the insulin receptor TyK and, concomitantly, inhibit the growthstimulating activity of insulin in rat hepatoma cells (Auberger et al., 1989). Gene cloning, which has already been achieved for MIS (Cate et al., 1986) represents an important step towards both the elucidation of the mechanism underlying the regulation of the production of these molecules and the understanding of their mode of action.

210 The aim of the present study was to determine the structure and organization of PP63, to analyze in detail the 5’-flanking sequence, and determine its chromosome location.

RESULTS

AND

INTRONS'

tt’t Bg P

DISCUSSION

(a) Isolation of genomic clones: organization of PP63 In a previous Southern-blot analysis, a PP63 cDNA probe which covered more than 75”, of the cDNA was shown to hybridize to a single, large sequence, (> 20 kb), BumHI rat genomic fragment (Auberger et al., 1989). This suggested that the complete gene might be included in such a fragment. Based on this observation, a rat genomic library, containing size-selected Bun?HI fragments was constructed and screened using a PP63 cDNA probe (Auberger et al., 1989). A clone bearing an insert of approx. 18 kb was isolated, purified and analyzed by restriction enzyme mapping using standard techniques (Maniatis et al., 1982). Restriction fragments, varying from roughly 2.5-0.5 kb and covering the entire sequence of the clone, were isolated and analyzed using the Southern-blot method, with various cDNA probes. These fragments were also subcloned in M 13 vectors and either totally or partially sequenced. All of the exons and the boundaries were sequenced on both strands. Thus, the number of exons and introns in PP63, as well as the exact sizes of the exons and the approximate sizes of the introns were determined. The organization of PP63 is represented in Fig. 1A. The gene spans approx. 8 kb and comprises seven exons and six introns. Comparison of the introniexon boundaries (Fig. 1B) shows that each corresponds to a canonical splice consensus sequence (Breathnach and Chambon, 1981). The exon size displays rather uncommon variability, ranging from 84-687 nt. As found for MIS (Cate et al., 1986), the two extreme exons (1 and 7) are the largest, whereas the central exons are much smaller (Fig. 1). The first exon contains the translation initiation site preceded by a short (73 nt) untranslated sequence, whereas exon 7 contains the terminator UAA (nt 1148-1150 in the cDNA sequence). This delineates an ORF encoding a 358-aa polypeptide, a size smaller (17 aa) than that of PP63 precursor which we originally reported (Auberger et al., 1989). This difference is accounted for by the fact that the first 45 nt of the 5’-end region of the gene (Fig. 5) differ from those reported in the original PP63 cDNA (Auberger et al., 1989). This 45-nt sequence, which is present both in the genomic clone and in a new cDNA clone which we have isolated and sequenced (data not shown), contains the tsp (see section b, below) and therefore belongs to the gene. However, although we have found agenomic clone bearing the 5’-end sequence (i.e., the first 140 nt) which we originally described as part of PP63

I~:\r,ll

A

Bg P

Bg E I

I

0

2

-4

\(’

a

\(i

Fig. I. Organization

of a genomic

library constructed

used to propagate

the library

phages

were screened

cDNA

probe,

performed

(specific

(i.e., for fragments

bearing

deletions,

diagrams, putative

activity

the diagram.

Bg,BgIII;

were generated

boxes arc cxons

cxons are indicated

below the horizontal

alignments

and homology

data bases were performed (C.I.T.I.2,

Paris).

PP63 cDNA

sequence

sequencing;

rcspectivell,

was

(Amers-

et al.. 1977). Ml3 clones, by Dale et al.

system).

and 0 corresponds

In the to the

site\ are noted belo\\

(I-VII).

data

The SIXS of mtrons

brackets.

searches

Computer

sequence

library

of

analysis programs

accession

GT and AG sequences

see A), at the donor

and

analysis

in protein and nucleic acid number

IS M29758. (Lower part) Intron-cxon

of PP63. The highly conserved genomic

(Sanger

as described

using BISANCE

The GenBank

(Pharmacia,

P, Purl; S, Snc,I. Open boxeh are introns

(l-6) and hatched sequences,

PP63

Sequencing

(IBI rapid dclctions

of the main restriction

E.EcoRI;

host

et al., 1982).

>4Oll Ci:mmol)

technique

the scale (in kb) is at the bottom IS/J. The positions

(Maniatis

>0.5 kb), overlapping

(1985), using the ‘Cyclone’ system

into the

LE392. 4 x 10’ recombinant

and sequencing.

ham, UK), usmg the chain-termination various

( IS-20 kb) were

into pUC or M 13 vectors

When required

from the

from a rat genomic

WI). The bacterial

methods

for amplification

\t;

using the B,$II 878.bp

to standard

with [a-“S]dCTP

wah deduced

and inserted

Madison,

was strain

were subcloned

Sweden)

( ( ( (

DNA fragments

by hybridization

according

DNA fragments Uppsala,

I (Promega,

iGEM-

8kb

map of gene PP63. (Upper

(104J”,)

gradient

I

I (i(‘\(;

was isolated

as follows. BnmHl

f

E

E

6

restriction

clone which

t

/

shown in the diagram

on a sucrose

bacteriophage

I

\(‘....................

and partial

part) The gene organization

size-selected

\

PBg

Bg p

I

(1.... \(’

structure

t t f/K.

A P s

of the

boundaries

(determined

and acceptor

bq

splice sites.

arc underlined.

cDNA, the latter did not overlap the genomic clone described above. In addition, this 140-nt sequence does not hybridize to the mRNA species recognized by the rest of the cDNA, suggesting that it might represent a cloning artifact. Besides this discrepancy, two other differences were found between the sequence of the exons determined in this study and that ofthe previously published cDNA: GC substituted for CG at nt positions 901-902 (Arg”” + Ala) and an additional stretch of 26 nt was found at the 3’-end of the seventh exon. Again. these differences were found in the sequence

211 ofthe new cDNA clone. These modifications were therefore taken into account when the sequence was entered into the data bank. Comparison of PP63 cDNA with sequences found in data banks reveals a high degree of identity (approx. 747,) with AHSG cDNA (Chong-Chou et al., 1987). The aa sequences of precursor polypeptides are also identical by of 60”,, ; more importantly, there is a perfect conservation the Cys residues, shown to be crucial in determining the secondary structure of AHSG (Kellerman et al., 1989), throughout the sequences. In addition, PP63 (Le Cam and Le Cam, 1985) and AHSG (Lebreton et al., 1979) are both negative acute-phase reactants secreted by the liver. (b) Determination

of

the

tsp and identification

of

a

functional promoter in PP63

To identify the 5’ boundary ofPP63 mRNA, a [ “‘P]oligo complementary to the 5’ -end region of the coding strand of the cDNA (nt + 77 to + 101; Fig. 5) was hybridized to total hepatocytc mRNA. The primer was then extended with avian myeloblastosis virus reverse transcriptase (Pharmacia, Uppsala, Sweden) and the sizes of the resulting products were determined by gel electrophoresis under denaturing conditions, followed by autoradiography. The major cxtcnsion species has a length of 101 nt; a minor species, which is 2 nt shorter, is also detected (Fig. 2). This suggests that tl-~~nscriptjon may start at a T residue, 73 nt upstream from the start codon AUG. The tsp was confirmed by S 1 nuclease mapping. When a uniformly labelled, single-stranded M 13 anti-sense probe, corresponding to a P.crI-LlgllI (nt -282 to + 246; Fig. 5) fragment ofthe 5’-end region of PI>63 was hybridized to total hepatocyte mRNA and digested with S 1 nuclease, a major protected fragment of approx. 250 nt was detected (Fig. 3, lane 4). This result agrees \?~ll with that of the primer-extension experiment. To determine whether the cloned genomic DNA containing fP63 S’-flanking sequence actively initiates and supports tr~~nscription, we performed an in vitro transcription assay using liver nuclear extracts and, as a template, a plasmid bearing the BglII fragment than encompasses the t.sp (nt -692 to 246; Fig. 5). The transcription products wcrc hybridized to the same probe as the one reported above (S 1 Inapping), digested with nuclease S 1 and analyzed on a polyacrylamide gel under denaturing conditions. A major protected fragment of approx. 270 nt was detected (Fig. 3, lane 3). This fragment is slightly larger than the protected fragment obtained after mRNA hybridization. This difference is accounted for by the presence of the same vector sequence (i.e., a fragment of the polylinker containing the SUCI, f(pflI, Smal and BarnHI restriction sites) in both the M 13-derived probe and the pEMBL-derived template used for transcription. Thus, it seems likely that transcription was initiated at the correct position, in vitro.

Fig. 2. Determination

ofthe tsp. Total hepatocyte

from normal adult rats as described

by Chirgwin

ized with a 2%nt oligo (see Fig. 5). labelled nucleotide

kinase

(Boehringer,

Mannheim.

Primer-extended

products

polyacrylamide)

and sized by comparison

ladder

run alongside

indicated includes

RNA

( 10pg),prepared

et al. (1979). was hybridat the 5’-end F.R.G.)

were run on a sequencing

and

with poly[Y-~‘P]ATP.

gel (8 M urea/6%

with a dideoxy

sequencing

(the first four lanes from the left). The main band,

by an arrow, corresponds

to a lOl-nt extended

fragment.

which

the 25-nt primer.

The ability of the cloned putative PP63 promoter to support transcription in intact cells was also demonstrated using a transient transfection assay. For this purpose we transfected a plasmid bearing theBg/II-BstNI, PP63 5’-end fragment (nt --692 to + 52) inserted in a CAT vector, into a human hepatoma cell line (HepG2, Knowles et al., 1980). Based on CAT assays, a significant promoter activity which reaches about 25 ‘;, of that of the strong SV40 viral promoter, was observed for the chimeric PP63 construct (Fig. 4). (cl Analysis of 5’-flanking putative cis-regulatory

PP63 sequence; identi~cation

of

elements

The sequence of the region (approx. 800 bp) located upstream from the tsp was determined (Fig. 5) and a search for potential cis-acting regulatory elements was carried out.

212

+982

+3-ac +I -ac

~270 4-250 -OR

1

2

Fig. 4. PP63 expression

3

as measured

by the transfection

mids into intact cells. The following chromatography

on QUIAGEN

were used: pEMBLCAT, same plasmid (-692

1234 Fig. 3. In vitro transcription

ofPP63.

Assays were performed

by Gorski et al. (1986), using nuclear from 3-month-old (1 ~g per assay)

(Dente et al., 1983)containing, genomic

proteins

male rats (Dignam was a plasmid

fragment

and

(45 pg per assay) prepared

derived

from

a pEMBLCAT

at

55’C

using a nuciease for

pH 7.4)/N&J stranded Davies

16 h

the PP63 zsp and includes

(1 mM), to a uniformIy

(0.4 M);!EDTA

M 13 probe

scribed

above.

F.R.G.)

was carried

enzyme,

in a medium

products

(50000 cpm per assay) containing

Digestion

with

prepared

the PstI-EstNI Sl

nuclease

labelied

containing

fragment

(Boehringer

by de-

Mannheim, were then

to Fig. 2), and their

sizes were

determined by comparison with a dideoxy sequencing ladder run alongside [the first four lanes on the left). Lanes: 1, migration of undigested probe; 2, probe hybridized the in vitro transcription denatured protected

to 25 pg yeast tRNA; products;

total hepatocyte fragments

RNA.

3, probe hybridized

4, probe hybridized

(QUIAGEN,

Studio City, CA),

devoid of a promoter

(lane 3); the

fragment

a plasmid

of PP63 (putative

bearing

the strong

promoter)

viral

(lane l),

SV40 promoter

plastic

dish)

in Dulbecco

HepG2

serum150 119 gentamycin described

phate precipitate

by Chen and Okayama

was left overnight

[dichloroa~etyi-l-‘~Cl species,

luciferase

2pg

gene driven

(1987). The

DNA-Ca phos-

48 h later (Gorman

et al., 1982), using r>-threo-

(spec.

act.

Saclay,

(ch) and the acetylated were resolved by thin-layer

by autoradiography. efficiency,

fetal calf

B per ml, using the

with the cells, and the yield

g L’Energie Atomique,

chloramphenicol

et al., 1980)

in contact

chloramphenicol

ated; 3-ac: 3-acetylated) detected

cells (Knowles Eagle’s medlum/lO”,,

per ml/l0 pg amphotericin

ofthe CAT was measured (Commissariat

hepatoma

and Vogt-modified

To account

of a plasmid

forms

“C-1abelled

(I-ac: l-ucetyl-

chromatography

for the variations

(pRSVLUC),

by the Rous sarcoma

SO-60 mCi~mmol~

France).

bearing

virus promoter,

and

in transthe firefly were co-

transfected with the test plasmid. Luciferase activity, present in detergent lysed cells, was determined as described by Brasier et al. (1989). The amounts

of cell extracts

containing

were used for the CAT assays.

the same number

of luciferase

units

OR, origin.

to

to 20 big of heat-

The sizes of the probe

(in nt) are indicated

grown

fection

of IO units of

0.25 M NaCI. The products

gel (see legend

single-

as described

genomic

out at 45°C for 1 h, in the presence

run on a sequencing

were

of CAT plasby ion-exchange

(lane 2). used as a reference. Highly purified supercoiled plasmid DNA (6 pg) was transfected into exponentially growing (5-7 x 10’ cells/60 mm

procedure

part of

protection assay. They were first hybridized, formam~de~Tris HCl buffer (20 mM 807,

in

et al. (1986),

vector

in front ofthe CAT, the 335-bpP.TzI-BstNI

which encompasses

to + 52) genomic

diameter

et al., 1983). The DNA template

the first exon (nt -282 to i 52) (see Fig. 5). Transcription analyzed

as reported

columns

a construct

purified

in front of the cat, the 5’-end 744-bp ffglII-EstNl

bearing

pSVZCAT,

plasmids,

and of the

on the right margin.

An oligo that could represent a TATA box is found at the correct position, namely 25-30 nt before the potential tsp. However, this A + T-rich sequence, 5’-AAATAAATT (nt -3 1 to -23), does not closely match those considered as optimal for strong TATA boxes (Efstratiadis et al., 1980). The gene appears to lack a typical CCAAT box, the sequence originally found to bind the nuclear factor CjEBP which was first isolated from rat liver nuclei (Graves et al.,

1986; Johnson et al., 1987). However, the sequence 5’-TTACGCAAT, found at nt positions -59 to -5 1 in the PP63 promoter, contains the pentanucleotide 5’-GCAAT which was recently shown to be optimal for C/EBP-binding sites (Vinson et al., 1989). This sequence also matches the consensus sequence of c®ulatory elements reported to define a new class of liver-specific genes which includes the hemopexin gene (Poli et al., 1989). These genes, like PP63, do not contain a binding site for HNF-1, the nuclear factor shown to confer liver-specific expression to a large group of genes (Courtois et al., 1987). Considering the prominent

213 AGATCTTCAGGGAGGAGGGTGAGGTCTTGAAGGAGGAGGGGTGAGGTC -750

TGCATGGAGGAGGGTGAGGTCTTCAGGGAGGAAGGTGAGGTCTGCATGGA -

-700

GGAAGGTGAGATCTTCAGGGAGGAGGGTAAGTTCTTTAGGGAGGAGGGTG

-650

AGGTCTTCAGGGAATAGGGTGAGGTCTGCAGGGAGGAGGGTGAGGTCTGC

-600 -550 -500

CATTCTTCTGAAGTCTTTGGCATCTCCCATGACTCACTGTATTTTTGCTA

-450

CATGTGTGGACCTTATTCTCCTTCTAGGAGAATATCTTGAAGGAGATCAA

-400

TCAACAACATTAGAAGAGTCAGTAAGAGCATGTTATTGCCCTAGCATGCT

-350

CACATTCAGGGTGCCATGGCATGATCATATTCAGGTAGAGGACCCCCTGT

-300

CCCCACTCCCTGCTGCCACTGCAGCTACGCTCTGTGCAGCTCGTGTCCAT

-250 -200 I--Cl CGGGGATAGATGATGTCCTAACTTATTATTTGCTTTCCCAGAGCTGTTGTTTG

-150

-b-

-50

CAAGGATGATTTGGAACCAGAACAAAAATCAGCCCACGCCTTTACGCAAT -aTCCTTCAGTGGGCTCTGTCAAATMATTGAGCCCTCTGCCCCTCTATCGG

1

TCTAGCTCTCCAAGCTGATTATCTGGGCTGCCCCTGACATTTGTCCATCT

-100

CCCAGGGCCTCTCTGGAGCAGCCATGAAGTCCCTGGTCCCTGGTCCTGCTCCTTTGT __--__---------

i-51

TTTGCTCAGCTCTG~GCTGCC~TCAGCTC~C~GGTGCAGGGCTGGG

+101 +151 +201

TAGCCGTGCACTACCTCAATAAACATCTTCTTCAGGGATTCAGGCAGATC

+251

T

Fig. 5.Sequence ofthe 5’.flanking M 13mp19 and sequenced located

downstream

region ofPP63.

on both strands

(see

Two BgUI genomic

fragments

legend toFig. 1). Numbering

from the rsp are those which differ from the original

( -798 to -692 and -692 to + 246) were subcloned

of nt, noted on the left margin, PP63 cDNA

sequence

(Auberger

starts

into M13mp18

from the putative

and

tsp.The first 45 nt

et al., 1989). The following

putative

DNA

regulatory elements were tentatively identified in the sequence: a, A + T-rich region or TATA box: b, C/EBP-binding site; c, cctamer binding site; d, NF- 1 binding site. The 18-m repeated sequence is underlined with a continuous line. The sequence selected to make the oligo used in the primer extension experiment

is underlined

with a dashed

line. The GenBank

data library

role of C/EBP in regmating gene expression in differentiated cells, in particular hepatocytes and adipocytes (for a review, see McKnight et al., 1989) and the fact that PP63 is specifically expressed in liver (Auberger et al., 1989) this DNA element is likely to be involved in regulating PP63 promoter activity. Besides these elements, an oligo: 5’-TGGCAA, that perfectly matches the potential DNA-binding site for the ubiquitous transcription factor NF-1 (Jones et al., 1987) was also found at nt positions -162 to -157. Furthermore, the ohgo 5’-ATTTGCTT, found at positions -126 to

accession

number

of PP63 5’-flanking

sequence

is M36547.

-119, corresponds, with a single mismatch, to the inverted complementary sequence of the octamer motif 5’-ATGCAAAT which binds both tissue specific (Ott-2A and Ott-2B) and ubiquitous (Ott- 1) transcription factors (Kemler et al., 1989). These DNA elements, which are capable of binding ubiquitous factors, might contribute to the strength of PP63 promoter. Lastly, it is worth noting that the G-rich 18-nt oligo: S’-GGAGGAGG~TGAGGTCT~ was repeated ten times, with a conserved 3-nt spacer, between nt -787 and -581.

214

+++I

+++‘I+

i+il,++,,+

t~++l+++i

+*

+

-+

+ + + + + + /

+++t,

+ i

+

I

fi

+

++

I

+

-+

+

c

i-

i

.cf

‘r,

0

a