The human DNA-binding protein, PO-GA, is homologous to the large subunit of mouse replication factor C: regulation by alternate 3′ processing of mRNA

Gene, 145 (1994) 261-265 0 1994 Elsevier Science B.V. All rights reserved.

261

0378-l 119/94/$07.00

GENE 07969

The human DNA-binding protein, PO-GA, is homologous to the large subunit of mouse replication factor C: regulation by alternate 3’ processing of mRNA (Recombinant DNA; poly(A)-addition

site; Drosophila; transcription factor)

Yang Lu” and Anna Tate RiegelaTb ‘Department of Pharmacology, and bVincent TLombardi Cancer Center, Georgetown University, Washington, DC 20007, USA Received by R. Padmanabhan:

27 December

1993; Accepted:

11 February

1994; Received at publishers:

17 March

1994

SUMMARY

We have previously cloned a human gene encoding a 128-kDa protein which we termed PO-GA [ Lu et al., Biochem. Biophys. Res. Commun. 193 (1993) 779-7861. In the present report, we compared PO-GA to recent DNA database entries and determined that PO-GA was 80% identical, at the amino-acid level, to the large subunit of replication factor C (activator 1) cloned from mouse [Burbelo et al., Proc. Natl. Acad. Sci. USA 91 (1994) in press]. This indicates that PO-GA probably represents the corresponding subunit of human replication factor C. In addition, PO-GA has high homology to a putative Drosophila transcription factor. All three proteins contain a nuclear translocation signal and an ATP/ADP-binding motif, and are highly conserved in regions with homology to Escherichia coli and yeast DNA ligases. We determined that PO-GA mRNA species of 5.3 and 4.5 kb can be detected in most human tissues, but levels are especially high in ovary. Analysis of the sequence of a new PO-GA cDNA clone that we obtained reveals a previously undetected 650-bp 3’-UTR extension. This region contains several A + U-rich regions potentially involved in regulation of mRNA stability. This fragment only hybridizes to the larger 5.3-kb mRNA. Comparison of cDNA sequences revealed that the two mRNA species arise as a result of alternate use of poly(A)-addition sites. Since the ratio of the two mRNA species is variable in different tissues, we speculate that alternative 3’ processing of the PO-GA mRNA is utilized as a mechanism of regulating cellular levels of mRNA.

INTRODUCTION

The PO-GA cDNA clone was initially detected (Lu et al., 1993) by expression screening of a HeLa cDNA Correspondence to: Dr. A.T. Riegel, Department of Pharmacology, Georgetown University Medical School, Washington, DC 20007, USA. Tel. (l-202) 687-1479; Fax (l-202) 687-6437. acid(s); ARE, A + U-rich response Abbreviations: aa, amino element(s); bp, base pair(s); GCG, Genetics Computer Group (Madison, WI, USA); kb, kilobase or 1000 bp; nt, nucleotide(s); oligo, oligodeoxyribonucleotide; ORF, open reading frame; PCR, polymerase chain reaction; PO-GA, DNA-binding protein; PO-GA, gene (DNA, RNA) encoding PO-GA; RFC, replication factor C; SDS, sodium dodecyl sulfate; UTR, untranslated region(s). SSDI 0378-I

119(94)00180-2

library with the cognate binding site of the PO-B transcription factor we have previously described (Wellstein et al., 1991). Although the properties of PO-GA indicated that it was not PO-B several features of this 128-kDa protein suggested it had a function in DNA regulatory mechanisms, i.e., regions of homology to DNA ligases and DNA repair proteins plus a strong preference to bind to G + A-rich DNA fragments (Lu et al., 1993). We speculated that this protein was likely to have a function in transcription, replication or repair processes (Lu et al., 1993). We now report that PO-GA has 80% identity to the large subunit of mouse replication factor C (RFC) (Burbelo et al., 1994). This RFC complex is also known as Activator 1 (Lee and Hurwitz, 1990; Lee et al., 1991). Purified RFC is composed of a number of proteins of

262 varying

molecular

DNA-dependent and Stillman,

activity

large subunit

proteins

of human

preference

indicates

for certain

those

associated

(Tsurimoto

and Stillman,

unusual with

the DNA binding

domain

sequence

cession may

Drosophila

No. L17340). perform

similar

transcription

identity

structures 3’

ends

of RFC of

(PO-GA)

of aa identity

factor (GenBank

It is possible functions

is the

and the large subunit

RFC have a high percentage

that in

to ac-

these proteins

DNA

region ( Fig. 1, aa 400-456) tween the human

specificity

of its largest subunit

(Lu et al., 1993) is not clear. PO-GA a putative

DNA recessed

and S. /~>/nhe ligascs (Lu et al.. 1993). Illtercstingly,

of the

to have DNA

1990). How the binding

to DNA is related to the apparent

and

in Melendy

that PO-GA

RFC. RFC appears

especially

of mouse

DNA binding

(reviewed

1992). The high degree of relatedness

mouse and human binding

mass and exhibits ATPase

regulatory

mechanisms. In our previous study we described two different POGA mRNA forms, 5.3 and 4.5 kb. In the present report we demonstrate that the larger form arises from a 6SO-bp extension of the 3’-UTR which we have now sequenced. The shorter 4.5 kb form arises from utilization of an earlier poly(A)-addition site. The preponderance of the 5.3 or 4.5kb forms of the PO-GA mRNA is variable, depending on the human tissue examined and suggests that differential processing of the PO-GA mRNA 3’ end may play a role in regulation of PO-GA (large subunit RFC) cellular transcripts.

(a) Homology

comparison and structural aspects of the

PO-GA gene Comparison of the original PO-GA sequence with recent entries in the GenBank and EMBL databases revealed that PO-GA had 80% aa identity with the large subunit of RFC from the mouse, (GenBank accession No. U01222; Burbelo et al., 1994). Interestingly, significant homology between the PO-GA sequence and a putative Drosophilu transcription factor (accession No. L17340) was also revealed by a database search (highlighted aa in Fig. 1). The identical aa which occur more than once between the mouse, human and Drosaplziiu sequences are shown as a consensus sequence in Fig. 1. The homology is especially apparent between the 3’ C-terminal sequences of PO-GA (distal to aa 350), the large subunit of mouse RFC and the Drosophila protein (Fig. I). Furthermore, in all three proteins, the N-terminal domain is Lys-rich and contains a conserved nuclear transport consensus sequences (Kate et al., 1992, and Fig. 1, Box I). We had previously observed that the PO-GA protein contained several domains with high homology to E. co/i

identical

proteins

this

(98%) be-

and has 67’/;, aa

with the D~~~~~~)~~j~u protein.

This domain

lies

protein (aa I90 - 52X)

of the human

that binds to DNA (Lu et al.,l993).

However,

the 3’ por-

tion of this domain,

a possible

helix-turn-

helix structure

which contains

in PO-GA,

is the least conserved

between all three proteins. influence

Sequences

the DNA-binding

Another situated

in the N-terminal

identical

between

of each protein.

region is between aa 653 and motif (Box II in Fig. I ) is portion

of this stretch

the three proteins. between

precisely

conserved

residues

may play a role in the tertiary

protein

through

disulfide

binding

structure

formation

of the

or an even

such as the zinc-finger

from the comparison

quence of the large subunit large

The Cys

structure

DNA-

motif.

We conclude, human

of which is

the three proteins. bridge

and is

Distal to this region

is an area rich in His and Cys. the spacing

more complicated

region

in this region might

specificity

highly conserved

796. An ATP~ADP-binding

PO-GA subunit

sequences, of RFC.

between

the aa se-

of RFC of the mouse and the that

PO-GA

In addition

is the human

it appears

that

a

D~~sop~j~~l protein exists that has high homology to PO-GA and it is tempting to speculate that this is a protein DNA

AND DISCUSSION

and mouse

within a larger portion

which plays a role in the Drosophila

tion process.

EXPERIMENTAL

is almost

Whether

tr~lnscription

these proteins

and repair

DNA replica-

also play a role in

processes

is a matter

of

conjecture.

(b) Sequence of the PO-GA 3’4JTR extension Initial sequencing of overlapping PO-GA vealed

a cDNA

of approx.

clones

re-

4500 bp encompassing

a

S-UTR of 393 bp and a 3’-UTR of 666 bp (Lu et al., 1993) (depicted in Fig. 2A). A poly(A)-addition site (AATAAA)

was observed

an A, string conjectured

in the PC20 that

these

poly(A) polymerase.

approximately clone

A residues

Consistent

had been

and

the sequence

added

with this notion,

PC21 (Fig. 2B), the string of A residues not detected

140 bp before

(Lu et al., 1993). We by

in clone

after nt 4360 was

continued

for another

6.50 bp until an A,, string (Fig. 2B). This poly(A)

stretch

was preceded approx. 30 bp earlier by a putative poly(A) processing site (underlined in Fig. 28). We now report the

sequence

of the

650-bp

extension

in this

clone

(GenBank accession No. L24783). This sequence is compared in Fig. 2B to the 3’ sequence of clone PC20. Since the difference in size of these two clones was 650 bp, we speculated that this potential differential processing of the mRNA

could result in the two different

RNA species

263

H ..MDIRKFPG M ..MDIRKF?G D mqrgSdsFFk c --NDTru*FG

“IPSGKKl”s “IsSGKKpm

rip.......

ETvkKNEKTK ET”.KNEKTK

GdEeTlKaKK aSEgTvKgKK

.._....

“IPSGmt-v-

EN-KNEKTK

--E-T-K-KK

GIKElaVNSS G”KEAK”N”S . . ..AxakSa G-KKAKvHss

60 rKBDdfKqKQ gKBDasKPKQ eaBngetPsk -xED--KPI[Q

of 5.3 and 4.5 kb that we had previously Northern

blot analysis

observed

using

(Lu et al., 1993).

120

SmESmnlQ SDSESERTVQ ,S.SDeDRv"s SDSESEETVQ

!iIlz

"KNAKRppEK LpvSsKPGXI "XNAXKksBK LslSyKPGKV ppetl(l(rkas ktaS...... VxnAKL--EK L--S-KPGK-

SrqDPVTYIS SqMPmS SedDvYaatp s--rem-s

ETDEEDDFmC ETDEDDDFvC &PxakkaRng BTDE-DDF-C

(c) Northern blot analysis of human tissues detected with different PO-GA cDNA probes

H M 0 C

KWSKSKEN KKAASKSKEN qXpA. KKu.SKSKW

GrSTNShLGT G"STmyLGT

KLTPTSVLDY KLTPTSVLDY ..El KLTPTSVLLDY

180 w;TgS"QRSn POTES"QRSg FGgIz....t PCTBSVQRS-

from differential

G-STNS-LGT

SNMRKNEE~ SNVKKNEEN" SkLKrhvDpT SI-IImEm

H M D C

KKMVaSXRXE KKMVtSXRXE KrvIvpXpX. KKHV-SKRKE

1SQNTDESgL sSQNTEDSrL

NDEAIAKQLQ NDEAIAKQLQ

LH*DBEFART LHBDBEFART .fBnBDidRs LXED~EPART

240 LAMLDEEPKt LULDEEPKI LmevDlDesI IA-LDEEPKI

H

M D C

KKKR IYD SKKKRIIYD Rrrk Ii KKKPT YD

. . . . . . ...1

-sQNT--S-L

KTKNXPLSPI KTKNKPLSPI KTKNKPLSPI

WEDZxELBRQ LDEDRELERQ ..tknvLB.. NDEA1AKpL.p LDEDAELBRQ

300 H KKARKDtEaG

.ErPSS”Qan

M KKARKDsEeG D Xelapskkv. C mARmU)-E-G

eEsFSSVQdd

LSKAEKhKyP LSKAEKqKsP

hkmtaqvsd n..KaelfSt

-E-PSSYQ--

LSKAEK-K-P

---K----S-

H M D C

QhsKSSADkl ge"SSPXASs KLA=MKrKEE QpcK SAhrk eacSSPKASa KLAlMKaKEE raknSSpE.. Q--X%%.--- ---SSPIuLS- KLA-m-Km

SSY-E-E--A

skRKEnAlK1 arRKEsATeP .ppKpksTKs --RKB-ATK-

SSYkE=Ep"A SSYnEtEllA

H M D C

RKBLXERYGG AKSLZERYGG AeSvIkeYOG AKSLIERYGG

KVTGNVSKKT KmNVSXXT KVmtvVgXK1 KVTGNVSKKT

NYLVMGRDSG NYI.mmRDSG kYLVvCeEaG HYLVMGRDSG

QSKSDKAAAL QSKSDKA?aL pkKlavAeeL *SISDKuAL

480 GTKI~~GL GTKI.LDBrmL nipILsBDGL GTKILDEDGL

MIU(B.SKLER MRKEkSKLER gXKBvktsrR MKKB-SKLER

TPQKNvQGKR TPQKNoQGKR ssdKkEkeat TPQm-*GKR

540 XlSPsKXESE RISPaKKESE KlkygeKh d KISP-KKESE

XEqvaeetsg KS........ Kkepssqkeh KE________

dskaRnladD ..thgn ppspR..taD _-_-R-_--D

600 sSeNK"EnLL rSsNKeEcLL lktld""gMa _S_~"E_LL

SKKsrpTskr SKKcKlTllk laXhKvkeeh SKK_X_T___

dSlaKTIKKE tD"fwksLDf nSPmKa"XXE astcprgZ,D" tSPkeTkdR1 nD"pa"tLk" _S*_I[T_KI[E _D"____LDV

660 H WVDXYXPTSL M WMXYgPaSL D WVDXhKPTSl C WVDKYKPTSL

KtIIQQQGDQ SD.NKLLlWL RNWqKSSsED KMIGQQGDQ SCANKLLrWL RNWhKSSpEE XeTVlipaGaa SnvtKLMnLn skWyvnhdgn K-IIGC4GDQ SCANltIL-WL BNW-KSS-E-

SLYC SLVC tLVv SLYC

RHH.sKFGKF KKHaaXFGKL XX.pqrpnpW KXR--KPGK-

sgKDDnSSFK AsKDDGSSFK AknDDGSfyK A-~~DDGSSPK

QPLGYSYVRL QBLGYSWSL kELGFdaVBP *ELGYSYvEL

720 NrSDTKSKsS LKAIVAESLN NTSIKGPYsn tW3tX'WRnS Ll‘?+WAESLN NrsIKGFVTs NASBT'Rt!,Wl Wlde'VstlLs NkSIsGYFT. IPlrSDTBSX-S L.KP.-VAESLNNTSTIICPYT-

SIAFKEGLRI SIAPREGLKI SIcPRELkvRI SIAPI[EGLKT

PPPAMNRIIL PPPAMNWIL sPakve&IIa PPPAMNXIIL

840

H M D C

YCFDLRBQRB YO?X,RFQRP YCMfldlppRp YCPDLRPQRP

RvtppIXGAMM RVBQlXsAML RlBQXKGkiM RvEQItG-

GANQDIRQVL KNLSMWCArS G.WQWRQ"L "NLSMWCAqS atNnDIRQsi nhiaLlsRke G~NQDrrtpvr. BNLSMWCA-S

human

adult

QENYiHVKPV

QENYLHVKP" QqNyLqVlP. QEKnaVRPV

AAGGDMKKHL MLLSRAADSI AAGGDMKKHL MLLSRRADSI ..qGnlcKd"L akvaatADa1 AAGGDHXRRL MLLSRJADSI

CMjDLVDsQI CWDLVDnQI slGDL"EkrS CDGDLM-QI

RSKQNWSLLP RSKQNWSLLP RansaWSLLP RSRQNWLLP

960 aQAIYASVLP TQAIYASVLP TQAfFsSYLP TQAIYAS"LP

H M D C

LVqPLTSQG" LVRPLTSQG" l"RPLakdGq LVRPLTSQGV

DG"Qd""aLM EGaQhVIkLM EGVpaaldvM EGVQ-"--La

DTYYLMKEDP ENIMBISSWG DTYYLMYdtDF HNIMBVSSWG kdYhL,,rl(DLDslvBltSWP D TrnxKEDP ENmE-BSRG

GKPSPFSKLD GKPSaFSKLD GKkSPldavD GKPSPPSKLD

1080 PKVKAAFTRA PXVKaaFTRA gr"KiW.LTRs PKVKAAFTRA

H M D C

WLTPY YNRsRHLTPY YMCEvmaysY YWKBABLTPY

aLQAIKaSRh SLQvVKtSRl SaQAgikkkk SLQA-K-SR-

STsPsLDSEY STgPaLDSEY SeaagaDdDV ST-P-LDSEY

. . . ..SQSDE . . . ..tQS.E agghlSseED -----SQS-E

KDQDAIETDA KEQDAVETDA eDkDnlElDS KDQDA-ETDA

1140

"EELnEDD.. sEBFqEDD.. lDEgpgBEdg -ES--ED-1176

H MIKKK.....

M Mnau. D LIKaKkrttt C KmM-----

TKSSKPSKpE kDDKEprRGI[GRssKK' TrSSKPSItsE REKEskRGKG EnwKR SKaSggSKka RssrasKsKa Ka.KK TKSSI[PSI[-E R-R.E--KGRG r--m

tissues

PO-GA

mRNA

Northern both

blot, containing species

(Fig. 3A). Interestingly,

species was especially

a different extended in most the 5.3-kb

preponderant

in

human ovary when compared with other tissues (Fig. 3A and B, lane 4). Although both the 5.3 and 4.5-kb species were detectable in most tissues it is clear that the 4.5-kb species is less abundant and the ratio of the two species is highly tissue dependent (e.g., Fig. 3A, compare ovary to testis, lane 4 vs. 5). After stripping the blot of PC1 probe we re-hybridized with the 190-bp 3’-UTR probe PRl shown in Fig. 2A. This probe only hybridized to the 5.3-kb species and no 4.5-kb species was detected (Fig. 3B) even upon a twoweek exposure of the blot (data not shown). This result fits the prediction that the 650-bp fragment is not present in the 4.5-kb species because the earlier poly(A)-addition site at 4360 is used to generate the lower molecular mass species. Since the ratio of the two species is organdependent it appears that this processing site is underutilized in some tissues compared with the later site at 5067 (Fig. 2A). However, the half-life of the two species may not be equivalent. In this regard it should be noted that the 650-bp 3’-UTR extension contains four signals similar to the ARE (A + U-rich response element) (Shaw and Kamen, (highlighted

H M D C

species in many human

are present

M X"SPtKrES" D KpaadlesSV C K-SP-K-ES"

H M D C

mRNA

the observation

that

PC1 and

(Lu et al., 1993). Hybridization

360 KGETkTPKKT KGEktTPKKT KatrprvRKe KGET-TPKI[T

CLEGLxFVIT CLEGLTWlT CtsGLTFVVT CLEGLTWIT

SKYHiAvEtE SXYRmAaEaE pKkghssEek SKYE-A-B-E

to both PO-GA

tissues but not placenta

previously,

SKBIPKGAEN SKBIIPKGAEN SKUPKfispd SKEIPKGAEN

. . . . . . ..Ck . ..Xr &vkeekXs --------I-

bridizes

blot with two probes

a

PC1 is in the PO-GA ORF and hy-

set of tissues from those examined

MREGPKALG LE(REGPKALG kNRssclnpG LmREGPmLG

LnLIRmPG LDLIRTMPG FDLIReksGr LDLIRTMPG-

tissue Northern

(Fig. 3). Probe

to a different human

RTNYQAYRSY RTNVQAVRSY RasavlYqkY RTIWQAYRSY

H M D C

PRl

species resulted

of the 3’end, we hybridized

SKyeSSkEsq gKgRaSeDaK SitRSSpspK SX-MS---X

SPEDS..EKK SPEDS..$KX 1tDEerhBrK SPED?.--EXK

GVLESIBRDE GVLESIBRDB GVLESmEREE GVLESIERDE

human

PO-GA mRNA

processing

eRKsYSPrKq aRKtYSPaKH . . . . . . . ..H -RR-YSP-W

420 H KsSPaKkESV

To test if the different

1986) which has a loose consensus AUUUA in Fig. 2B) and has been reviewed by Sachs

(1993). This element, either alone or in multiple copies, has previously been shown to destabilize some cellular

Fig. 1. Comparison of the human PO-GA aa sequence (H) (Lu et al., 1993) with the sequence of the mouse large subunit of RFC (M) GenBank Accession No. U01222 (Burbelo et al., 1994) and with a putative Drosophila transcription factor (D) GenBank Accession No. L17340. The dashes represent areas of non-identity of aa between the three proteins. The dots represent gaps which improve the homology alignment. The consensus sequence (C) represents aa shared by two or three proteins (bold). Identical aa between the H, M and D proteins are highlighted. Box I represents an area with homology to a nuclear localization consensus sequence (Kato et al., 1992) and Box II represents an area containing the ATP/ADP binding site consensus. Methods: Sequence alignments were performed using the GCG program.

264 A.

PC1

PRl

A*lt

PC20

PO-GA

cDNA clones

t

PCZI

PI tAA

t

7.5-

B

_is’../

4

4.4-

*

,.

,

Fig. 3. Northern analysis of a multiple human Palo Alto. CA, USA) with two different probes (A) The blot was first hybridized

tissue blot (Clontcch. from PO-GA cDNAs.

with the 982-bp

PC1 probe which is

contained

within the ORF of the PO-GA cDNA (Fig. 2A). The two species of PO-GA mRNA detected at 5.3 and 4.5 kb are indicated by arrows.

PBL-peripheral

blood leukocytes.

(B) The blot was then rehy-

bridized with probe PRl representing nt 203 to 393 of the 3’-LTTR sequence of clone PC21 shown in Fig. 2B. Only a 5.3-kb band hybridized as indicated Fig. 2. Alternative

poly(A)-addition

sites of the 3’-UTR

described

of PO-GA

by arrow.

Methods: The PC1 probe was generated

for colony hybridization

as

in the legend to Fig. 2. Hybridization

mRNA. (A) Two PO-GA cDNA clones PC20 and PC21 were purified from an UniZAP HeLa cDNA library by hybridization with the 5’ PO-

and prehybridization solution were composed of 5 x SSPE (1 x SSPE is 0.8 M NaCIIO.O1 M NaHLPO,/l mM EDTA pH 7.4): IO x Denhardt’s

GA cDNA

solution/50% formamide. 2”/;I SDS100 ug per ml of freshly denatured. sheared salmon sperm DNA. The DNA probes. PCI and PRI. were

clones

sequence

(PCl)

and their position

(Lu et al., 1993). The overlap relative

shown. The heavy line represents tion ofprobable poly(A)-addition represents’s

DNA fragment

to the full length

between

PO-GA

the

cDNA

is

the ORF of PO-GA cDNA. The posisignals are indicated (AATAAA). PRl

used for hybridization

of Northern

blots

(see Fig. 3). (B) The nt sequence of these clones is identical until nt 4498 at which point a poly(A)-tail is added in clone PC20. PC21 contains

labeled

by nick translatton

buffers at a concentration

and mixed thoroughly of 2 x IO6 cpmiml.

in the hybridization

The blot was hybridized

at 42°C for 24 h in a shaking water bath. The blot was washed at room temperature for IO min in 2 x SSC ( I x SSC is 0.15 M NaCl,/O.O15 M Na,citrate

pH

7.6)iO.5%

SDS.

twice.

and

then

washed

with

no poly (A) sequence at this position but extends an extra 649 bp (GenBank accession no L24783), at which point a string of A bases is

0.1 x SSC/O.5% autoradiography,

observed preceded by a poly(A)-addition signal, AATAAA (underlined). The extended 3’-liTR in clone PC21 contains several ARE (Shaw and

The blot was re-exposed for autoradiography to check if all the probe had been removed. The clean blot was rehybridized with a PRI probe,

Kamen.

which was generated as a PCR product from PC21 template with a pair of primers. T3j4: 5’-GAC CAG TAG GAA CCA CCC; and T3X:

1986) which are highlighted.

Materials: PC1 is a 982-bp

frag-

ment representing nt 953 to 1934 of the full length PO-GA cDNA clone (Lu et al., 1993). Methods: PC1 was labelled for library hybridization using nick translation. described previously 5 x IO4 recombinants

The screening

of the HeLa cDNA

libraries

was

(Lai et al.. 1992). Briefly, plaques were plated at per plate and transferred to nitrocellulose filters.

After denaturation and renaturation then hybridized with the radiolabeled

the filters were prehybridized and PC1 probe. Initial positive clones

were subjected to second and third rounds of screening to obtain pure positive clones. All positive clones were subcloned into pBluescript phagemid vector by excision. The sucloned inserts were sequenced on both strands using synthetic oligos and the dideoxy chaintermination method.

mRNAs (Shaw and Kamen, 1986). Reiterations of this sequence are capable of conferring instability on a previously stable mRNA (Shaw and Kamen, 1986). The mechanism seems to be through direct interactions of trans-acting cellular proteins with the ARE sequence (Kruys et al., 1989; Malter, 1989; Gillis and Malter, 1991; Bohjanen et al., 1991; 1992; Vakalopoulou et al., 1991).

5’-CTA GGT

SDS once at 50 C and once at 65’ C for 20 min. After the PC1 probe was removed from the blot by boiling.

CCA TAG CCT CAC (Fig. 2B), using 35 cycles of 94‘C

for denaturing, 56C for annealing and 72’.C for extension. The PCR product was purified with the Stratagene DNA fragment purrhcation kit (LA Jolla, CA, USA) and was analyzed

by

I O/O agarose gel electro-

phoresis. The same washing condition were used as described for the PC1 probe. After washing the blot was exposed for autoradiography.

Conversely, in some contexts this sequence may also be involved in prolonging the half-life of some cellular mRNAs (Schuler and Cole, 1988; Lindsten et al., 1989). Clearly the precise role of these potential ARE in the regulation of cellular levels of PO-GA will have to be determined by mutational analysis. However it should be noted that similar signals are not present in the 3’-UTR of the shorter 4..5-kb mRNA. (d) Conclusions PO-GA is the human large subunit of RFC sharing 80% identity with the mouse homologue of RFC and

265 significant homology with a putative Drosophila transcription factor. The two PO-GA mRNA species detected in many human tissues arise from different utilization of poly(A)-addition sites in the primary transcript. Since the ratio of the two mRNA species varies quite substantially between different human tissues we speculate that utilization of these sites may be a factor in determining the overall amount of a particular species of PO-GA cellular mRNA present in human tissues.

Kruys,

V., Marinx,

Translational quences.

O., Shaw,

blockade

G.,

Deschamps,

imposed

J. and

by cytokine-derived

Huez,

G.:

UA-rich

se-

Science 245 (1989) 852-855.

Lai, S., Czubayko, human

F., Riegel, A.T. and Wellstein,

heparin-binding

Biophys.

growth

Res. Commun.

factor

A.: Structure

gene pleiotrophin.

of the

Biochem.

187 (1992) 1113-l 122.

Lee, S.H. and Hurwitz,

J.: Mechanism

of elongation

by DNA polymerase

d, proliferating

cell nuclear

of primed antigen

DNA

and activa-

tor I. Proc. Nat]. Acad. Sci. USA 87 (1990) 5672-5676. Lee, S.H., Kwong, A.D., Pan, Z.Q. and Hurwitz, J.: Studies activator

I protein

nuclear

complex

an accessory

antigen-dependent

on the

factor for proliferating

DNA polymerase

cell

d. J. Biol. Chem. 266

(1991) 5944602. Lindsten,

T., June, C.H., Ledbetter,

Regulation

ACKNOWLEDGEMENTS

of lymphokine

T cell activation

We thank Dr. Anton Wellstein for advice during this project and helpful review of the manuscript. This work was supported, in part, by grants from the National Institutes of Health MH49105 and DK43127 (A.T.R.) and a grant from American Cancer Society (Y.L.), IRG-193. A.T.R. is the recipient of a Research Career Development Award (DK02141).

pathway.

J.A., Stella, G. and Thompson,

mRNA

stability

Science 244 (1989) 339-343.

Lu, Y., Zeft, AS. and Riegel, A.T.: Cloning human

DNA

Commun.

binding

protein,

Malter, J.S.: Identification

tively to AUUUA

multimers

in the 3’untranslated

region of lympho-

of a AUUUA-specific

Mol. Biol. 6 (1992) 129-158. Sachs, A.B.: Messenger RNA

messenger

degradation

RNA bind-

Nucleic

in eukaryotes.

Acids

Cell

74

(1993) 4133421. regulated

and oncogene

in trans

mRNA stabiht-

in a mouse

monocytic

Cell 55 (1988) 115551122.

ficities and affinities. J. Biol. Chem. 267 (1992) 6302-6309. Burbelo, P.D., Utani, A., Pan, Z.Q. and Yamada, Y.: Cloning of activator

1 (replication

of the

factor C) reveals homology

DNA ligases. Proc. Nat]. Acad. Sci. USA 91 (1994)

Gillis, P. and Malter, J.S.: The adenosine-uridine

binding

nizes the AU-rich elements of cytokine, lymphokine mRNAs. J. Biol. Chem. 266 (1991) 317223177. G.J., Lee, W.F., Chen, L. and Dang,

mains and interaction

degradation.

factor recogand oncogene

C.V.: Max: functional

with c-Myc. Genes Dev. 6 (1992) 81-92.

do-

Cell 46 (1986) 6599667.

Tsurimoto, T. and Stillman, B.: Purification of a cellular replication factor RF-C that is required for co-ordinated synthesis of leading and

Mol. Cell. Biol. 11 (1991) 32883295.

Bohjanen, P.R., Petryniak, B., June, C.H., Thompson, C.B. and Lindsten, T.: AU RNA binding factors differ in their binding speci-

Kato,

Res.

Shaw, G. and Kamen, R.: A conserved AU sequence from the 3’untranslated region of GM-CSF mRNA mediates selective mRNA

Bohjanen, P.R., Petryniak, B., June, C.H., Thompson, C.B. and Lindsten, T.: An inducible cytoplasmic factor (AU-B) binds selec-

with bacterial in press.

of a novel

Biophys.

ing protein. Science 246 (1989) 664-666. Melendy, T. and Stillman, B.: SV40 DNA replication.

tumor.

large subunit

Biochem.

193 (1993) 779-786.

ies are independently

kine mRNA.

and expression

PO-GA.

Schuler, G.D. and Cole, M.D.: GM-CSF

REFERENCES

C.B.:

by a surface-mediated

lagging

strands

during

SV40 replication.

Mol. Cell. Biol. 9

(1989) 609-619. Tsurimoto, T. and Stillman, B.: Functions of replication proliferating cell nuclear antigen: functional similarity merase accessory

proteins

from human

factor C and of DNA poly-

cells and bacteriophage

Proc. Natl. Acad. Sci. USA 87 (1990) 1023-1027. Vakalopoulou, E., Schaack, J. and Shenk, T.: A 32-kilodalton

protein

binds to the AU-rich domains in the 3’untranslated regions rapidly degraded mRNAs. Mol. Cell. Biol. 11 (1991) 3355-3364. Wellstein,

A, Dobrenski,

A.F.,

Radonovich,

M.N.,

Brady,

T4.

J.F.

of and

Riegel, A.T.: Purification of PO-B, a protein that has increased affinity for the pro-opiomelanocortin gene promoter after dephosphorylation.

J. Biol. Chem. 266 (1991) 12234-12241.