Ontogeny Response

MOLECULARANDCELLULARBIOLOGY, Mar. 1993, p. 1933-1942 0270-7306/93/031933-10$02.00/0

Copyright © 1993,American Society for Microbiology

Kid-l, a Putative Renal Transcription Factor: Regulation during Ontogeny and in Response to Ischemia and

Toxic Injury

RALPHWITZGALL,"12EILEENO'LEARY,' REINHARD GESSNER,"12ANDRE J. OUELLETTE,34,5

ANDJOSEPH V. BONVENTRE'2*

Medical' * and Surgical3 Services, Massachusetts GeneralHospital, Departments of Medicine2 and Surgery,4 Harvard MedicalSchool, and Shriners Burns

Institute,'

Boston, Massachusetts 02114

Received 2 November 1992/Accepted 4 December 1992

We have identified a new putative transcription factorfrom the rat kidney, termedKid-i (for kidney, ischemia and developmentally regulatedgene1).Kid-]belongstothe C2H2 class of zinc fingergenes.Its mRNA accumulateswithageinpostnatalrenaldevelopmentand is detectedpredominantlyin thekidney. Kid-] mRNA levels declineafter renalinjurysecondarytoischemiaorfolic acidadministration,twoinsults which result in epithelialcelldedifferentiation,followedby regenerative hyperplasiaand differentiation.The lowexpression of Kid-i earlyin postnatal development, andwhenrenaltissue isrecovering after injury,suggeststhat thegene

product is involved in establishment of a differentiated phenotype and/or regulation ofthe proliferative

response.The deducedproteincontains 13C2H2zincfingersattheCOOH end ingroupsof 4and9separated bya32-amino-acidspacer. Thereare consensussites forphosphorylationintheNH2terminusnon-zinc finger region^as wellas in the spacerregion between zincfingers 4 and5. Aregion of the deduced protein shares extensive homology with a catalytic region of Raf kinases, ^a feature shared only with TFIIE among

transcriptionfactors. To determine whether Kid-i canmodulatetranscription,achimericconstructencoding theKid-1non-zincfinger region (senseorantisense) and theDNA-binding regionof GAL4wastransfected into COSand

LLC-PK,

cellstogetherwithachloramphenicolacetyltransferase(CAT)reporterplasmidcontaining GALA binding sites, drivenbyeithera minimal promoterora simian virus 40 enhancer. CAT activitywas

markedlyinhibited in cells transfected with thesenseconstructcomparedwiththeactivityincells transfected with the antisenseconstruct. Toourknowledge, this pattern ofdevelopmental regulation, kidney expression, andregulationoftranscriptionisuniqueamongtheC2H2class of zincfinger-containing DNA-binding proteins.

The kidney is a very heterogeneous organ with various differenttypesof cellsexpressingdifferentphenotypicchar- acteristics. Kidney developmentand cellular differentiation

arelikely regulated atthe level oftranscription and involve tissue-specificgeneexpression (16, 32, 34). One of theways to achievekidney-specific gene expression is to selectively

express trans-acting factors, kidney-specific transcription factors,whichmayplay importantroles in determination of the differentiated phenotypes of kidney cells and renal development.Theseproteinsmaybeimportant notonlyfor renal development and differentiation but also for the pro- cesses involved in repairof thekidneyafteran injury. The

process of kidney injury and repair recapitulates many

aspects of development since it involves dedifferentiation and regeneration of epithelial cells followed by differentia- tion (4, 35, 56).Theabilityto restoredifferentiated function and regenerateepithelialstructureafteranischemicortoxic insult is an importantproperty of the kidney,an abilitynot possessed bythe heart and thebrain,in whichmyocytesand

neurons are notreplaced.

One goal of the present study was to identify genes of potential importance in kidney development and repair whichare preferentially expressedinthekidney.Wereport the cloning of a novel cDNA which encodes a protein containing 13 zinc fingers in groups of 4 and 9. The gene, whichwe callKid-1 (for kidney, ischemia anddevelopmen- tally regulatedgene 1),isexpressed primarily in thekidney

* Correspondingauthor.

and is a single-copy gene. Kid-i mRNA levels accumulate with increasing postnatal age. mRNA levels are reduced after ischemia and reperfusion orfolic acid administration.

Thereareproteinsequence consensusmotifs in the non-zinc finger regionand intheregion separating the fourth and fifth zincfingersofthe deducedprotein, which suggeststhat the proteinmayberegulated by phosphorylation andmayhave kinase activity.

To determine whetherKid-1canmodulate transcription,a

chimericconstructencodingtheKid-1non-zincfinger region (sense or antisense) and the DNA-binding regionofGAL4

was transfected into COS cells, a large-T-antigen-trans- formedgreenmonkey kidneycellline,andLLC-PK,cells,a

highly differentiated porcine epithelial kidney cell line, to-

getherwith a chloramphenicol acetyltransferase (CAT) re-

porter plasmid containing GALA binding sites, driven by either a minimal promoter or a simian virus 40 (SV40) enhancer. CAT activity ^was markedly inhibited in cells transfected with thesense constructcomparedwith that in cells transfected with the antisenseconstruct.

MATERIALS ANDMETHODS

Construction and screening of cDNA libraries. cDNA li- braries wereconstructed from normal and postischemicrat

kidneys.MaleSprague-Dawleyrats(CharlesRiverBreeding Labs, Wilmington, Mass.) weighing 250 to 300 g were

anesthetized withanintraperitoneal injectionofpentobarbi- tal (6.5 mg/100^gofbody weight).The blood supplytoboth 1933

Vol. 13, No. 3

kidneys was interrupted by placement of a microaneurysm clamp (Roboz Surgical Instrument Co., Washington, D.C.) on the renal pedicles for 30 min. The clamps were then removed, and blood flow was reestablished for 1 h. Total RNAwasprepared from these kidneys and normal kidneys from other rats by previously reported techniques (39).

Poly(A) RNAwas selected by oligo(dT)-cellulose chroma- tography. First-strand cDNA was synthesized by using Moloney murine leukemia virus reverse transcriptase.

cDNAwas ligated into XgtlO with EcoRI linkers. Enzymes and componentsforlibraryconstructionwereobtained from GIBCO BRL, Gaithersburg, Md. In addition, kidney AgtlO and Xgt11 librarieswere purchased from Clontech Labora- tories, PaloAlto, Calif. For initial screeningof the libraries we constructed, approximately 125,000 phage were plated on Luria-Bertani agar plates. Subsequent screenings of libraries involved plating of106 phage. Tworeplicas of each plate were made with GeneScreen (New England Nuclear [NEN], Boston, Mass.) or Hybond (Amersham, Arlington Heights, Ill.)nylon filters. Replicafilters werethen hybrid- ized with oligonucleotides or cDNA by standard protocols (48).

DNA sequencing. Inserts from purified phage were sub- cloned intopBluescript (Stratagene, La Jolla, Calif.). Dou- ble-stranded DNAwas sequenced by thechain termination method(49).Eachclonewassequencedonboth strands. T3 and T7 Bluescript sequencing primers and oligonucleotide primers corresponding to the cDNA were synthesized by Oligosetc., Inc.,Wilsonville, Oreg. Products ofsequencing reactionswere electrophoresed on Hydrolink Long Ranger (AT Biochem, Malvern, Pa.)modifiedacrylamide gels.Com- puter analysis was carried out with the sequence analysis software package of the Genetics Computer Group at the Universityof

Wisconsin,

Madison

(17).

The

predicted

amino acid sequence of the Kid-1 protein was compared with sequences in the SwissProt, National Biomedical Research Foundation Protein Identification Resource, and GenBank databases, usingtheBLAST and FASTA algorithms.

Animal protocols. In experiments designed to examine Kid-i mRNA levels with ontogeny, the kidneys of young Sprague-Dawleyratswhichwere0.5, 10, 15, 20,or40to45 days old were rapidly frozen in liquid nitrogen and then homogenized in 6 Mguanidiumisothiocyanate for prepara- tion of total RNA(48). Kidneysof threeorfour animals <20 daysoldwerepooled. Forolderanimals,RNAwasprepared fromeachkidney separately.

For the ischemia-reperfusion experiments, a left-flank incision wasmade in 150- to 250-gmale rats and the renal pedicle was clamped with a

microaneurysm clamp.

The contralateral, nonischemickidney served as acontrol. The incisionwasclosed

temporarily

until 30 min

later,

when the clampwasremoved. The incisionwasthen sutured

closed,

and, after the stated timeofreperfusion,bothkidneyswere removed for extraction of RNA without

freezing.

Inother adult male rats, kidney injurywas produced by intraperitoneal administration of 250or350 mg of folic acid per kg of body weight. Kidneys were harvested at 3, 8 (350-mg-treatedanimals), and 24(250-mg-treated animals)h after the injection. Controls were injected with the carrier only (0.3 mmol of NaHCO3). The surgical wounds were

sutured, and the animalswere returnedto their cages until removal of bothkidneys underpentobarbital anesthesia.

Northern (RNA)blot

analysis.

TotalRNA wasisolatedby standard protocols (48), with the exceptionthat 6 Mguani- dinium isothiocyanate was used.

Twenty-five micrograms

(unless otherwise noted)of total RNAwas

lyophilized

and

dissolved subsequently in 16 ,u of buffer A [20 mM 3-(N- morpholino)propanesulfonic acid (MOPS), 8 mM Na ace- tate, 1 mM EDTA (pH 8.0), 6.2% formaldehyde, 50%

formamide]-4 ,ulof buffer B(10mM EDTA[pH 8,0], 0.25%

bromophenol blue, 50% glycerol, 0.1 mg of ethidium bro- mide perml) (31). Sampleswereheated 10 minat65°Cprior to being loaded onto a formaldehyde-agarose gel. After electrophoresis, the gel was treated for 30 min in 50 mM NaOH,15 minin 2x SSC(lxSSCis 0.15 M NaClplus0.015 M sodium

citrate),

and 15 min in lOx SSC before

being

blottedontoGeneScreen Plus(NEN).Blotswerehybridized with a356-bpEcoRI-XmnInon-zincfinger fragmentofZ5.9 (Fig. 1), Z5.9zf-, labelled by random priming 219, 20).

Hybridizationwasperformedat55°Cwith 2.5 x 10 cpm/ml in 2x SDE (2x SDE is 100 mM NaCl, 50 mM sodium phosphate [pH 7.0], and 5 mMEDTA)-5% sodium dodecyl sulfate

(SDS)-100

,ug each of yeast tRNA and denatured salmon sperm DNA per ml

(55).

Thesulfated

glycoprotein

2

(SGP-2)

rat cDNA

(1,364-bp)

probewas obtained from R.

Buttyan,Columbia, University. Afull-lengthhumanglycer-

aldehyde-3-phosphate dehydrogenase (GAPDH)

cDNA probe was provided by M.

Alexander-Bridges,

Massachu- setts General

Hospital. High-stringency

washeswere done twice for 10 minat650Cwith 1x SSC

(Z5.9zf-),

twice for 5 min at65°Cwith 0.1x SSC

(SGP-2),

ortwice for 10 min at

650Cwith 2x SSC

(GAPDH).

Membraneswere

exposed

at

-700CtoKodak X-OMAT film with

intensifying

screensfor 3 to 14

days.

Southern blot

analysis.

Genomic DNAwas

prepared

from human leukocytes by the

protocol

of John et al.

(25)

and fromratandmouseliver

according

tostandard

protocols (3).

Ten

micrograms

of

genomic

DNA was

digested

with the indicated restriction enzymes

(rat DNA)

orwith EcoRI

only (mouse

and human

DNA). Samples

were

separated by

agarosegel

electrophoresis

and transferredontoGeneScreen Plus membranes

(NEN). Hybridization

withZ5.9zf-

(2.5

x

105

cpm/ml)

was

performed

at

55°C

in 2x SDE-5% SDS-100 ,ugeach of yeast tRNA and denatured salmon sperm DNA perml

(55). High-stringency

washesweredone twicefor 10 minat650Cwith 1x SSC

(rat DNA)

or4x SSC

(human

and mouse

DNAs).

Membranes were

exposed

at -700C to KodakX-OMATfilm with

intensifying

^screens.

AmplificationofRNAbyPCR.A

515-bp fragment

of

Kid-i

(nucleotides 441 to 955 in the

cDNA)

was

amplified

from 1 ,ug of total RNA from various organs

by using primers 5'-AATlTlCTCTCCACTATGG-3' (antisense)

and 5'-CTGG AGAATTACAGCAACC-3'

(sense)

and a

GeneAmp

kit

(Perkin-Elmer, Norwalk, Conn.).

The initial denaturation step involved incubation for 5 min at

95°C.

This was fol- lowed

by

40

cycles

of 1 minat

95°C,

1minat

450C,

and 1 min at

72°C,

with finalextensionfor 10 minat720C.No RNAwas

added to the

negative

control.

Equal aliquots

of the

poly-

merasechain reaction

(PCR) products

were

separated

on an

agarose

gel, blotted,

and

hybridized

as described under

"Southern blot

analysis."

Theblotwas

hybridized

with the

random-primed 515-bp fragment prepared by

PCR from the Z5.9/16

Kid-i

cDNAclone. Aftera1-h exposure, the areas on the membrane

corresponding

to the

signal

on the film were cut out and the

radioactivity

was

quantitated

in a

scintillation counter. To ensure the

quality

of the

RNA, amplification

wasalsocarriedoutwith

primers

for GAPDH.

Transfection protocols. LLC-PK1 and COS

cells,

growing in Dulbecco's minimal essential medium

supplemented

with 10% fetal calf serum, were

plated

at a

density

of

approxi-

mately2.5 ^x 105cells per100-mm-diameter dish2

days prior

to transfection. For

transfection,

cells were

exposed

to a

Kid-i, A PUTATIVE RENAL TRANSCRIPTION FACTOR 1935 total of 20 ,ug of DNA (including carrier DNA) in S ml of

Dulbecco's minimal essential medium-10% NuSerum (Col- laborativeBiomedical Products, Bedford,

Mass.)-400-Vg/ml

DEAE-dextran(Sigma,St.Louis,Mo.)-0.1mMchloroquine (Sigma).Four hoursafter the addition ofDNA,themedium was removed and cells were shocked for 2 min at room temperature with 10%dimethylsulfoxide in lx phosphate- buffered saline (PBS). After the shock treatment, the cells were washed oncewith PBS and new medium was added.

Cells were harvested 48 h after transfection for CAT and luciferase assays.

CATandluciferase assays.Transfected cellswerewashed twice with PBS, scraped with a rubber policeman into a microcentrifuge tube, and spun down. The cell pellet was resuspended in 200

VI

of 0.25 M Tris Cl, pH 7.8, and subsequently broken up by

freeze-thawing

three times by alternatingadry ice-ethanolbath anda37°Cwaterbath. The supernatant was assayed for CAT activity

by

standard protocols (48). Thin-layer

chromatography

with silica gel plates (Silica gel IB; J. T. Baker,

Phillipsburg,

N.J.) was used to separate acetylated from

nonacetylated

forms of chloramphenicol. Spots

corresponding

to monoacetylated and

nonacetylated

[

4C]chloramphenicol

were

scraped

from the thin-layer

chromatography

plate and quantitated in a liquid scintillationcounter.CATactivityisexpressedasthe ratio of monoacetylated

[14C]chloramphenicol

to total

[14C]chloramphenicol

and normalized to luciferase

activity

derived from a cotransfected luciferase reporter plasmid.

One microgramofa

luciferase-expressing plasmid, poLucSV/

Ti (53), was included in all cotransfection experiments to enablenormalizationfor transfectionefficiencies. Cellextracts were assayedfor luciferase

activity

with theluciferase assay systemfromPromega

(Madison, Wis.) according

totheman- ufacturer'sinstructionsusingamodel 20eLuminometer(Turn- erDesigns, Mountain

View, Calif.).

Nucleotide sequence accession number. The GenBank ac-

cessionnumber for the Kid-I sequence is M96548.

RESULTS

Isolation of cDNA clones. cDNAlibraries constructed from normal and postischemic kidneys were screened with a

degenerate oligonucleotide

[(A/G)TANGG(C/T)TT(C/T)TC

NCCNGT(A/G)TG] encoding the conserved peptide, HT GEKPY, which lies between the individual zinc

fingers

of the C2H2class of

transcription

factors

(52).

More than 150 hybridizing plaques were found in each

library.

Nineteen were furtherpurified, andfive of themweresubcloned into pBluescript.Partial sequencingof these clones revealed the presenceofC2H2 zinc

finger motifs, confirming

the success of thepursued strategy.

Oneof thesecDNAs,whichwecallZ5.9,derived from the postischemic

library,

contains 1,381

bp

and

hybridizes

to a 2.8-kbtranscriptwhich isexpressedprimarilyin the

kidney

and accumulates with age in the courseof

postnatal

devel- opment of the

kidney. Furthermore,

the levels of the tran-

script encoded by this gene, which we call

Kid-1 (kidney,

ischemia and developmentally

regulated),

decrease after ischemia-reperfusion and folic acid administration. Thisre- duction in mRNA levels

requires

morethan 1 h of

reperfu-

siontobeobserved,

accounting

forour

ability

toisolate the cDNA fromthe 1-h

postischemia kidney library.

Toobtain cDNA clones whichoverlap Z5.9and extend 5' and 3',

XgtlO

and Agtll rat

kidney

cDNA libraries from Clontech were screened with Z5.9. Positive

plaques

were

2563

552 1932

Kid- I

Z5.9 1932

Z5.9/16

950 2563

Z5.9/9C

FIG. 1. Relationship between clones analyzed. Z5.9 was the initial clone isolatedfrom theXgtlOkidney cDNA library. Clones Z5.9/16 and Z5.9/9Cwere selected by using PCR with 5' and 3' primers fromZ5.9 and lambda primers. The relationship of these clones to the full-length Kid-i cDNA is presented. Base pair numbers areindicated.

picked and subjected to PCR with internal Z5.9-specific primers and external lambda-specific primers. Clones ex- tendingfurther 5' or 3' were plaque purified, subcloned, and sequenced. Two overlapping cDNA clones were isolated, corresponding to atotalof2,563 bp (Fig. 1).

Sequence

analysis

ofKid-i. The nucleotide sequence of Kid-i containsalong open reading frame in the cDNA, from which a predicted protein product of approximately 66 kDa was deduced (Fig. 2). This size estimate is based on the usage of the 5'-proximal ATG codon at position 312. An in-frame ATG at position 468 could be an alternative start codonsince, exceptfor the presence of a pyrimidine rather than a purine at position -3, it lies in an almost perfect consensuscontext tooptimize translationinitiation (30). At thecarboxy end of the predicted proteinare13zinc fingers, which are clustered in groups of 4 and 9. Zinc fingers 4 and 5 are separatedby a32-amino-acid spacer. Upon alignment of the zinc fingers (Fig. 3), the invariant cysteines and histidines are clearly distinguished. In addition, there is significant conservationof manyamino acid residues within the fingers themselves as wellas in the H-Clink. There is moreidentityamongtheaminoacids of theindividual fingers inthe groupof nine zincfingers than in the group of four.

The amino-terminal non-zinc finger region ofKid-i con- tainsaKruppel-associatedbox(KRAB)domain (5) which is predictedtoformana-helix and may have a role in protein- protein interactions. If the alternativemethionine codon at position 468 is used, the KRAB-A box would not be ex- pressed,possiblyresultinginalteredinteractionof the Kid-1 protein with other proteins. Kid-1 also contains a nuclear translocationsignal,aspreviously described for nucleoplas- min(44).

Comprehensive sequence data base searches for genes encodingsimilarproteinsequencesrevealedregionsofiden- tityin themousegenes

Zfr-35

(55%) (14)andKr2(58%)and the human gene Hf.10 (51%). In each case, however, the identity was found in the zinc fingerregion of the protein sequence and there was no similarity seen in the non-zinc finger region. Thus, Kid-I encodes a novel zinc finger protein. Even though there is considerable homologywith other genes in the zinc finger region, Northern analysis of totalkidneyRNA,usingas aprobeZ5.9, which includes 11 of the 13 zinc fingers, revealed only one identifiable band under conditions of

high-stringency

washes (data not shown). Therefore,whilewe cannotexclude the possibility ofmorethanonetranscriptofsimilarsize,these datafurther suggest that there is no significant cross-hybridization of Kid-i cloneswith transcriptsof other genes.

Kid-i mRNA has two possible polyadenylation signals (57), one located 33 nucleotides from the poly(A) tail and another within 69 bases of the stop codon. There are structural featuresof theKid-i mRNAand deducedprotein sequence which suggest that both may be short-lived (Fig.

VOL. 13, 1993

1

1

A

1 COGAACAGTCTAAGGGTGTTCACTTTGGG&ACTCAGTGAGTCCCCCTCTGTTATOGAGTACAGTGTCTCTT 72 GAATCTTACTTCTGGSAATCCACGGACAGAGACTGGTGRCCTTGATTTGTGTGCCCACCCTCTAGAGGCA 148GACTGOCTCCTCCTCAOAOCCAGGcGCCAATOOGACGCTCGGAGGCTCGTCTCAGGTOCGA 224AGCAACATGTGCAACTCCTTTCTTCTTCCTCCCCTCAGATCCTCCTCCATCTTCACCCATCTTTAGAGGAAGAGCC MetAla Pro GluGln ArgGluGlyAla Ser Gln Val SerVal TSr Phe 300 TACAGAAAGAAGATGGCT CCT GAGCAA AGAG&AGGGGCCTCTCAGGTGTCAGTGACA TTT Glu AspValAla Val LeuPh. ThrArgAspGluTrpLysLysLeuAspLouSerGln 360 GAL GATGTG GCT GTGCTCTTTACTCOG GACGAGTGGAAGAAGCTGGATCTG TCTCAG

ArgSer LeuTyr ArgGluValMetLou Glu AenTyrSer Aan LouAla SlrMetAla 417 AGAAGC CTGTACCGTGAG GTG ATG CTGGAGAATTACAGCAACCTG GCC TCC ATG GCA

-

Gly Phe Lou PheThrLysProLysValIle Ser Lou LeuGln Gln Gly GluAsp Pro 474 GGA TTC CTG TTT ACCAAACCAAAG GTG ATCTCCCTGTTG CAGCAAGGAGAGGAT CCC

-I

TrpGln Val Glu Lys GluGly ProArgTyrPheSer Lou GlyLeuLys CysSerHis 531 TGG CAGGTG GAG AAA GAG GGCCCC AGA TAC TTC TCT CTCGGA TTGAAGTGC AGTCAT

put Cxzx

ArgThr Thr Lys SerThrGlnThr GlnAspSerSerPheGln Glu Leu IleVal Arg 588 AGA ACC ACTAAG TCA ACT CAA ACACAAGAC TCT TCATTTCAGGAACTG ATT GTAAGA

raf

Lys Ser Lys Arg Thr Phe Ala Phe GluProLeuAsnMet Lye SerGlu Asn LeuPhe 645 AAA TCTAAAAGA ACCTTTGCCTTCGAACCTTTG AACATG AAGTCAGAA AATCTT TTC Ile His Glu Gly Lys LeuGlu Glu Lye Trp Asp Lye Asn Thr LeuThr ValGluArg 702 ATA CATGAAGGC AAA TTA GAG GAA AAGTGG GAT AAGAATACT TTG ACTGTA GAA AGA

CRxx

Ser His Lye AsnAsnGlu Phe Ser Pro Lye Slr HisArgGlu Lye ArgSer SerGlu 759 AGC CAT AAA AAC AAT GAA TTT AGC CCA AAG TCC CAT AGA GAA AAA CGG TCC TCA GAA

lhc'law

Cys Lye Lye Gln Ile Ser TyrLeuSer Asn Pro Pro GlyIleThrProAspLye Arg 816 TGTAAAAAG CAGATATCTTATTTA TCT AACCCA CCAGGA ATC ACACCGGATAAACGC

CRx

Tyr LysCys Slr Met CysGlu Lye ThrPhe IleAsn ThrSer Slr LouArgLye His 873 TAT AAATGT AGC ATG TGE GAG AAA ACC TTC ATT AAC ACC TCA TCT CTT CGC AAA CAT Glu Lys Asn His SerGlyGlu LysLouPhe Lye Cys LysGluCys Slr Lye AlaPha

930 MAAG ^AGTGGAGAGAAATTA TTTAAATGTAAAGAATOTTCAAMAGCC^TTC

SerGln Ser SerAlaLouIle Gln HisGln Iie ThrHis ThrGlyGlu LysPro Tyr

987 ACA AGT TCAGCCCTTATTCAA CTAATAACT CACACTGGAGAGAAGCCTTAC

Val Cys Lys Glu CysGlyLysAlaPhe Thr LauSer ThrSlr Leu ryr Lys HisLou 1044 GTA TOTA"AGAA TGT AAG GCC TTC ACTCTC AGTACG TCC CTGTAT AAACAT CTC

Arg ThrHis !hr Vai GluLye Slr ?yrArgCyJLye GluCYJGiyLyJSar Phe Gly 1101 AmI.CACACTGTGGAGAAATCC TACAGA TOTAAGGAL TOT GT TCC TTT GOC Gln Arg S-rGly Lou Phe lie Hi GlnLye Ii.lesAla ArgGlu Asn ProHisArg 1158CAAAGOTCA GOTCTT ACC CGA GAAAAC CCT CATAGA TyrAsn Pro GlyArgLysAlaS-rAlaS-rLouSerGlyCysGlnArg AlaHis Ser 1215 TATAAC CCAGGAAGG AAG GCA TCC GCT TCC CTC TCT OGATGCCAGAGA GCT CAT TCC ArgLys LysThrTyrLeu CysAen GluCy&GlyAen Thr Ph. Lye Slr SlrSer Ser 1272 AGG AAG AAG ACC TAC TTGTOTAAT GTATOT GGCAAC ACC SSCALO SCT LOC SCCTCC LoU Arg 'yr Hil Gln ArgXI1 liarhr Gly GluLyeProPh.ArgCye SarGluCyJ 1329 CTC CGTTACCAT CAG AGA ATC CAC ACCGGA GAGAAA CCTTTCAGATGTAGC GAA TGC GlyArgAiaPh? lSrGln SerAlaSlr Lau Ii1 Ginlil GluArgZle ils ThrGly

1386 CA AtO QC ACGGGA

GluLyePro ryrArg CysGly Glu Cys GlyLye Gly PherhrSlrI1-SlrArg Lou 1443 GAAAAG CCC TACAGG TOT 000GGGa GGCTMCACTTCT ATC TCAAGA CTC AsnArg Hi ArgZle IZe HiSrhr Gly GluLye Louryr Asn Cys AsnGMuCysGly 1500 AAT AGACACCGGATA ATT CAT ACA GAGAGAAA TTG TAT AAT GC AATGAGTGT GGC LysAlaLouSerSlrHisSerrhrLauZ1a ZaeHiSGluArgz1a HisThr Gly Glu 1557 AAAGCCTTaAGT SCCCACTCAACT CTT ATT ATTCAC_AACOA ASC QCA ACT GGA GAG Lye ProCys Lys Cys LysVai Cya GlyLye AlaPheArgGlnSlr SlrAla Lou Ile 1614 AAACCG TGT AAA TGT AAAGTTTGT 00 AAAGCC TTC AGACAGAGTTCA GCTCTGATC GlnHisGlnArg MetHis rhrGly GluArg Pro!'yrLysCysAsnGlu CysGlyLye 1671 CAG ACTGGG GA AGACCC TAT AAATOT ALTGAGTOC 000AAA ThrPhaArgCys AsnSlr SlrLouSarAsnHisGlnArgIalHis rhr Gly Glu Lys 1728 ACATTCAGGTGT AACTCA TCC CTAAOT AAC QCCAGCGAATCCACACTGGA GAG AAA Pro TyrGln CysZleGluCysGlyMetSerPha Gly GlnSarSerAlaLauIleGln 1785 CCT TAT CA TOTATAGAATOT000ASGTCG TSTGGA CAA OTTCTGCT CTT ATT CAG His Arg ArgZaeHisrhrGlyGluLys ProPheLysCysAsn ThrCys GlyLys Thr 1842 CAC CGA AGG ATT CAC ACGGGA GAGAARCCG TTT AAATGTAAC ACA TGCGGA AAGACC PhaArgGln Ser Ser SerArgZla Ala HisGlnArgZ1aHisThrGlyGluLySPro 1899 TTTAGO CAGAGC TCC TQ COTATAGCACAT CAG CGAAT?CAT ACT GGG GAG AAA CCC TyrGlu Cys AsnThrCys Gly Lys LouPhe AsnTyrArgSerSlrLouThr AsnHis 1956 TACGAA TOTAACACGTGTGGGAILACTT TTCAACTAT AGG TCATCCCTTACCAAT CAT

2013

ryr Lys lie HisVal Asp Glu AspProTER

,TATAAA AC CAT GTG GAT GAG GAC CCT TAAAAAGTAAATTTGCATGTAAAAAAGCCTTAAACCAA 2078AGCTCTTCAGAGAATGGGCTTGAGAGCTGTMZAaCATACTATGGCGTCCCTGTAGTCAIAGTTACTGTCAAC 2154 TAAATTCCATGGATAAATCCTTAACAA TCATGCCT CACTTTAAAAAAATOC 2230 TTGAAATCCGGTGGTTATGTGTAGTCCTGOCTACCGOGGACCCTGAGGATTGCTTGACTCCACAAGTTTGAAGAGC 2306AAATGTTATGGAGCTTTCCATCTCAGAAGATTARCTTCTGGACTTCA 2382GGTATTTGGAAAATGGTGCTTTTCTCTCTOCAACAGTGTGTOCTAGATACAATTGTAATCATAGGCATCTAAGTGA 2458ACAGAGGAGTOGrATGGACTTAGCAAAACACAAAACTCTACTGTTGTGATACTTTTATACTATTTTAAZ&ATATCC 2534AAAAGGrATCAACACAAATCAAAAAAAkA

cDNA

ATG AlG

Vo00

Protein

,200 bp

4Zincfingers

*P "

125 bp

Met KRAB-A KRAB-B PEST Raf Nuleus 4Zincfingers

t?I CK CK CK

FIG. 2. Structure of Kid-i. (A) Sequenceof the Kid-I cDNA and thepredicted protein product. Nucleotide sequenceiswritten in the 5'-to-3'direction. The deducedamino acid sequence is shown from theputativestart codon atposition312. The zincfingersare underlined, and the proteinsequence isin italics. KRAB-A and -B, thePEST sequence, theraf homology domain, thenuclear translocation signal (Nucleus), and casein kinase IIconsensussites(CKII)areindicated. Putativepolyadenylationsequences(AATAAAandAATAAT)andan

instabilitymotif(ATITA)inthe3' untranslated regionareunderlined. (B) Relationshipsof theKid-i cDNA andstructuraldomains of the Kid-1protein. CK,casein kinase.

2). An AUUUA sequence at residues 2137 to 2141 may contribute to a short half-life of the message, since this pentanucleotidesequenceconfersmRNAinstabilityandhas been found in the 3' untranslated regions of several rapidly turning over mRNAs, including those of cytokines, lym- phokines, andoncogenes(7). One could alsospeculate that there exists another form of mRNA in which the polyade- nylation signal further 5' was used and which, because it doesnotcarrythe AUUUAsequence,would bepredictedto

have alonger half-life.Atpresent,wedo notknow whether this mRNAspeciesexistsand,if itdoes,whether it exerts a

distinctbiologicalfunction. There also isaconsensusregion (PEST sequence)in theNH2terminus of thededucedKid-1 proteinwhichispresentinproteinswithshort half-lives(45).

This peptide sequence, rich in proline, acidic, serine, and threonineresidues, hasbeenproposedto be the correlateof

theAUUUAsequence ontheproteinlevel.

Anotherimportantfeature of thepredicted Kid-1 polypep-

MTAM MTMT

9 Zincfingers TM ATTrA

_I_(A)n

6- S, RS<>

9 Zinc fingers Stop

_Ml-

10

0o'O

A

Kid-i, A PUTATIVE RENAL TRANSCRIPTION FACTOR 1937

190 211 239 267 327 355 383 411 439 467 495 523 551

S GE KL F K

TGE KP Y V

TV EKS Y R

T GE KP F R

TG EKP Y R

T GE KL Y N

TG EKP C K

TG ERP Y K

TGE KP Y Q

TGE KP F K

T G E K P Y E C C C C C

SM KE KE KE NE SE GE NE KV NE IE NT NT

E K T F I N T S K A F S Q S G K A F T L S G K S F G Q R G N T F K S S G R A F S Q S G K G F T S I G K A L S S H G K A F R Q S G K T F R C N G M S F G Q S G K T F RQ S G K L F N Y R

S SLR K S ALI Q T SLY K S GLF I S SLR Y A SLI Q S RLN R S TLI I S ALI Q S SL S N S AL I Q S SRI A S SLT N H H H H H H H H H H H H N

E K N Q I T L R T Q K I Q R I E R I R I I E R I Q R M Q R I R R I Q R I Y K II

consensus T G E K P F/y- C C G K F S/T - L - - H - - - H

FIG. 3. Zincfingerdomains ofKid-1. Fingerrepeats arealigned with consensus amino acids indicated at the bottom. Invariant cysteines andhistidines areboxed. There isa32-amino-acidspacer between fingers4and 9. Amino acidnumbers are indicated tothe left of the repeats.

tide is the presence ofa 12-amino-acid motif shared by all members of the Raffamily of serine/threonine kinases (22) (Fig. 4). This motif lies in the catalytic subdomain VI of protein kinases(22, 28, 29), suggesting that Kid-1 may have kinase activity. The glutamic acid residue in the catalytic domain, togetherwith theasparticacid residue 7 amino acids downstream, may support ATPbinding (9). Aregion some- what homologousto kinase subdomain VI has been identi- fied in the deducedproteinsequenceofoneof the subunits of TFIIE, a general transcription factor(40).

Thereare structural aspects of thededuced Kid-1protein which suggest that it may beasubstrate forphosphorylation.

There are three potential casein kinase II phosphorylation sites (41), twoin the non-zinc finger region and one in the first zinc finger. In addition, the 32-amino-acid spacer be- tween thefourth and fifth zincfingers contains four serines andonethreonine. Two of the serines and the threonineare preceded byan arginine at position -3, a consensus motif forcyclic AMP (cAMP)-dependent protein kinase and pro- teinkinase C (27).

Southernanalysis. SouthernanalysisofratgenomicDNA cut with BamHI, EcoRI, HindIII, or PstI and hybridized with Z5.9zf-yieldedonlyoneband in eachcase(Fig. SA).

Thus,Kid-i is asingle-copygene.Whenmouse and human genomicDNAswere cutwithEcoRI andratZ5.9zf- cDNA was used as aprobe, therewas a single bandseen in each lane, indicatingthatKid-i is conserved among rodents and primates (Fig. 5B).

Developmentalandtissueexpression ofKid-l. Kid-I mRNA levels change with kidney ontogeny. Kid-I mRNA is only marginally detectablebyNorthernanalysisinkidneystaken

%c0

e

^I.

- 23.1 kbp

- 9.4

- 6.6 - 4.4

23.1 - 94- 66- 4.4-

2.3-2.0 -

1 .4 - ~..

FIG. 5. Southern analysisofKid-i. (A) Rat genomicDNA hy- bridized withKid-i cDNA. When 10 pLgofratgenomic DNA cut with each of four different restriction enzymes wasprobed with Z5.9zf-,the non-zincfinger region of Z5.9, onlyoneband could be detected in each lane.(B)Mouseand humangenonicDNAcutwith EcoRIand hybridizedwithratZ5.9zf-.Asingle band can be seen in mouse (leftlane)and human(right lane) genomicDNAs.

at the timeof birth but becomes easilyseen at 15 days and accumulates togreaterlevels in the adultrat(Fig. 6).

Kid-i is expressedpredominantly in thekidney.When 50 ,ug of total RNA from a number of different organs was hybridized with Z5.9zf-, a 2.8-kb band could be detected only in the kidney (Fig. 7A). To further increase the sensi- tivity of detection ofKid-i mRNA in different organs, we also used the PCRtoamplifyKid-i mRNA sequences,using reverse transcriptase and primers selected to amplify a 515-bp fragmentof theKid-i cDNA sequence(seeMaterials and Methods). A probewas made from Z5.9/16, amplified with thesameprimers, and labelledby randompriming.The probewas hybridizedtoproductsof the PCR in aSouthern blot.Hybridizationwasquantitated by cuttingoutthe bands.

There was avery strong signal for the kidney and signals near background level for other organs (Fig. 7B). On the Southern blot itself, very faint but clearly present bands

Kid-1 Human c-raf Rat c-raf Xenopus c-raf Human A-raf Rat A-raf Mouse A-raf

Human putative ser/thr kinase Human B-raf

Mouse B-raf Drosophila raf

NMKSENLF IHEG D---N-I-L--- D---N-I-L--- D---N-I-L--- DL--N-I-L--- DL--N-I-L--- DL--N-I-L--- DL--N-I-L--- DL--N-I-L--D DL--N-I -L--D DL--N-I-L--D

FIG. 4. Homology ofKid-1 with members of theraf familyof serine/threoninekinases. Acatalytic regionof the Rafproteinfamily ishighlyconserved inKid-1. Hyphensdenote identitywithKid-1.

Thesingle-letteramino acid code isemployed.

Age (days)

1 1 0 10 15 15 2 0 4 5-5 0 28 S -

18 S -

:x

-__

__li

18 ^{S -}

@@ @W@ 0

Kid- 1

GAPDH FIG. 6. Northern analysis of total RNA collected from rats at various stages ofpostnatal development.Blotswerehybridizedwith Z5.9zf- and GAPDH cDNAprobes.

VOL. 13, 1993

A $

Q7 4S.

A I \q % e es \e \e °

28 S-

18 S- 28 S-

18 s-

B ORGANS

Control Sm. Bowel Lg.Bowel Testis Brain Spleen

Liver Kidney

'U

, Kid-i

1 5 48 96 hours

IC C I C I C

28 S-

18=. :p...

s s:

18 S- k

28 S-

18 S- 18 S -

,.) Kid-1

mma SGP-2

GAPDH FIG. 8. Northern analysis of kidney RNA obtained after isch- emia and various periods of reperfusion. Northern blots of total RNAtaken from kidneys which underwent 30 minof ischemia and various times of reperfusion (lanes I) or from the contralateral control kidneys (lanes C) were hybridized with either Z5.9zf-, SGP-2,orGAPDHcDNAprobes.

r g. U

a 2500 5000 7500 10000 12500 15000

PCR amplified Kid-i (cpm)

FIG. 7. ExpressionofKid-linvarious tissues.(A) Northern blot analysis of 50 ,ugoftotalRNAisolated from each of the indicatedrat organs. Among the large number of rat organs tested, Kid-i is expressed onlyin thekidney. Neither brain, lung, heart, liver, testis, spleen, large bowel,norskeletalmuscle tissue showsanydetectable expressionof Kid-i mRNA. ThelowerpartofpanelAshows the ethidium bromidestainingof thegel, indicating approximately the

same amountsof RNA in each lane, as reflected by the relative intensities of the 28S and18S bands. (B) Analysis of expression of Kid-i in varioustissuesby PCR.cDNAwaspreparedfrom different tissue RNAs withreversetranscriptase. By using specific primers forKid-i,afragment of the Kid-i cDNAwasamplified by PCR and samplesof theamplified productwerethensubjectedtoSouthern blot analysis with Z5.9zf- labelled with 32P by random priming.

Bandswerecut outandquantitated by liquid scintillation counting.

were seeninsamples from the spleen, brain, liver, and testis (data notshown). When PCRwas performed with oligonu- cleotides complementary to GAPDH, approximately the

same amount of amplified DNAwas obtained from each

organ. Thus, Kid-1 is expressed predominantly in the kid-

ney.

mRNA levelsafter ischemia- andfolic acid-induced tubular necrosis. Sincekidney injury is likely associated with dedif- ferentiation of surviving cells (6, 35, 56), and since repair follows the same general pattern as nephrogenesis (4), the effects of ischemia and reperfusion, as well as folic acid administration, on expression of Kid-i were evaluated.

Kid-i mRNA levels are reduced after renal ischemia and reperfusion, a stimulus which results in dedifferentation, mitogenesis, and tissue repair. Ascanbeseen inFig. 8, the levels of Kid-i mRNA decline after 30 min of unilateral ischemia and 5 h ofreperfusion. There isamarked decrease in the steady-state levels of Kid-i mRNAup to 96 h after

reperfusion. Normal mRNAlevels are restored after 7 days of reperfusion(datanotshown).Toprove that thedecline in mRNA levels was not a nonspecific response to tissue damage or energy depletion, which might have a general effectontranscription,wehybridizedthe same blot with the cDNAforsulfatedglycoprotein 2 (SGP-2), a gene develop- mentally

regqlated

inthe kidneywhoseexpression is down- regulated inepithelia that are terminally differentiated (23).

Thisgene has also beenimplicatedinprostate cell death(10) and has previously been shown to be induced with renal ischemiaandreperfusion (46).There was a marked increase in SGP-2 mRNA levels in thepostischemic kidneyat48and 96 h ofreperfusion,at atime when Kid-i mRNA levels were reduced, comparedwith levels in the contralateralkidney.

To evaluate whether a decrease in Kid-i mRNA levels could bereproduced byanother stimulus forkidney epithe- lialcellproliferation,wemeasuredKid-i mRNAlevelsafter rats were administered folic acid. High concentrations of folic acid induce acute tubular necrosis with subsequent regeneration of epithelial cells derived from the surviving cells(2, 21).

Kid-i

mRNAlevelswere suppressed3 h after folic acid administration and remained suppressed for at least 24 h (Fig. 9). The same blots were hybridized with a cDNAfor SGP-2. SGP-2mRNAaccumulatedtohighlevels at8 and 24 h after folic acidadministration, atthe sametimes thatKid-I mRNAlevels were reduced.

Effect of theN-terminal, non-zincfinger region of

Kid-i

on transcriptionalactivity.TheKid-icDNA wasmodifiedin the following way. A 5' fragmentwas

amplified

by PCR with primers spanning position 312 (5'-CCAACA'iTI7AAGCT TCTAGACTGCAGCTCGAGGCCACCATGGCTCCT GAGCAAAG-3' [the ATG startingat position312 is under- lined and in boldface type]) to position 541 (5'-TCCACC

TGCCAGGGATCCTCIT-3').

The upstream primer con- tained a perfect Kozak box for ATG-312 and

recognition

sites for HindIII,

XbaI,

PstI, and XhoI. The downstream primerincluded the internal BamHI site at

position

524, so that the resulting PCR fragment could be cut with BamHI and ligated to the BamHI-BsaHI 1.6-kbp

fragment

which 4

g

I

15

I I-V

I_I

i c

mg* Oa

m a" a

Kid-1, A PUTATIVE RENAL TRANSCRIPTION FACTOR 1939

3 8 24 hours

FA C FA C FA C

18 S-

28 S-

18 S-

18 S-

0

^SGP-2

,,,*i4, '1j GAPDH

i.f *

FIG. 9. Northern analysis of kidney RNA collected 3, 8,or24 h afterfolic acidadministration. Northernblotswerehybridizedwith Z5.9zf-, SGP-2, and GAPDH cDNAs. Lanes FA, RNA from a

kidneyfromananimal treated with folic acid; lanes C, RNA froma

kidneyfrom ^avehicle-treatedcontrol animal.

was isolated from ^a phage clone containing the full-length Kid-i cDNA. Sequencing confirmed the absence of any

mutations. The non-zinc finger domain of Kid-i was subse- quently excised and ligated to the DNA-binding region of GAL4(encoding amino acids 1to147) in thevectorpBXG1, both in sense and antisense directions, yielding pBXG1/

Kid-iN sense orpBXG1/Kid-lN antisense (Fig. 1OA). The reporter plasmid contained five GAL4 binding sites and a

CATgenedownstreamfromaminimalpromoter(pG5EC)or

SV40 enhancer (pG5SV-BCAT). pBXG1, pG5EC, and pG5SV-BCATwere generous gifts ofM. Ptashne (26, 47).

Ten micrograms of either pBXG1/Kid-1N ^{sense or} pBXG1/Kid-lN antisensewascotransfectedwith 3,ugofthe pG5ECreporterplasmidtoevaluate whethertheN-terminal, zinc finger-free domain of Kid-1 ^was able to modulate transcriptional activity when coupled to the DNA-binding region of GAL4. We chose COS cells, a large-T-antigen- transformed green monkey kidney cell line, and LLC-PK1 cells, a highly differentiated porcine epithelial kidney cell line, to assayfor transcriptional regulation in two different eukaryotic environments. There was low-level constitutive expression from the pG5EC reporter plasmid, which con-

tainsonlyaweakpromoter(47). In neithercell line couldwe

detect anypositive influenceof theGAL4-Kid-lN chimeric proteinontranscription. In fact, the fusionprotein exerteda

negative effect^on transcription in both cell types (Fig. lOB andC).

To furtherevaluate thepossibilitythat theGAL4-Kid-lN chimeric protein might alter transcription, ^we performed experiments with pG5SV-BCAT, a reporter plasmid in which theCATgeneis drivenbythestrong SV40enhancer (33). This construct showed strong CAT activity when cotransfected with pBXG1/Kid-lN antisense; however, CAT activities dropped markedly when pBXG1/Kid-lN

sense was the expression plasmid (Fig. lOB and C). In a

subsequent experiment, we evaluated how the suppression oftranscription varied with the amountof expression plas- midtransfected. A50% reduction in CAT activity^{was seen} when 1 ,ug ofexpression plasmidwas cotransfected with3 ,ugofreporterplasmid, withgreaterreductionsobservedas

largeramountsoftheexpression plasmidwerecotransfected (Fig. lOD).

DISCUSSION

The kidney is a complex organ, consisting of cells with highly varied differentiated phenotypes. This phenotypic complexity is likely determined in part by differential expres- sion of various genes, some of which may be expressed primarily in the kidney. Kidney-specific expression may be modulated in part by kidney-specific transcription factors.

Our studies reveal a novel Kid-i cDNA which hybridizes to a 2.8-kb mRNA transcript which is expressed primarily in the kidney and accumulates in the course of postnatal development of the kidney. It is possible that Kid-i plays a role in renal development. Whereas in humans renal devel- opment is complete at the time of birth, in rats a substantial amount of development takes place after birth (50). S-shaped bodies can still be found 4 to 5 days after birth (12), and thymidine incorporation does not decrease to background levels until 15 to 35 days after birth (12, 54). In the newborn rat kidney, Kid-I is barely detectable. Kid-i is much more strongly expressed in the adult. Its mRNA levels increase between days 15 and 20 after birth, at a time when the thymidine incorporation begins to decrease to baseline (12, 54). Thus Kid-i mRNA levels are correlated with epithelial cell differentiation in postnatal development.

The potential importance ofKid-I is also underscored by thefact that its expression is predominantly kidney specific, a feature comparable to that of the helix-loop-helix protein MyoD in muscle (15). Some of the other known zinc finger genes are preferentially expressed in particular cells, and some are developmentally regulated. MZF-I, a human zinc finger gene, is preferentially expressed in myeloid cells (24).

ZFX and ZFY are zinc finger genes on the Y chromosome which may have a role in testis development (51). ZFX and ZFY are transcribed, however, in many human tissues. The mouse gene

Zfp-35,

which shares homology withKid-iin the zinc finger region only, contains 18 zinc fingers and is expressed predominantly in the testis (14). The Wilms' tumor gene has four C2H2 zinc fingers and is expressed primarily in the kidney and the spleen (11). To our knowl- edge, Kid-i is the first example of a developmentally regu- lated putative zinc finger transcription factor expressed primarily in the kidney.

Kid-i encodes the same number of zinc fingers (13) as MZF-I, ZFX, and ZEY. MZF-1 has a 24-amino-acid spacer between the fourth and fifth zinc fingers, a pattern similar to that ofKid-1,in which the spacer consists of 32 amino acids.

The amino acids of the spacer regions of the two proteins are different, however. In MZF-1, the spacer is rich in glycines and prolines. In the deducedKid-1 protein, the spacer region contains four serines and one threonine. Two of the serines and the threonine are preceded by an arginine at position -3, a consensus motif for cAMP-dependent protein kinase and protein kinase C (27). The zinc fingers in ZFX and ZFY are notdivided by a spacer.

Decreased Kid-i mRNA levels are seen when epithelial cells are dedifferentiated and proliferate: early in postnatal development, during ischemia and reperfusion, and after folic acidadministration. It is possible that the gene product may exert a positive effect on kidney cell differentiation and/or a negative effect on growth. We have shown that the non-zinc finger NH2 terminus ofKid-1 (Kid-iN) can serve as astrongtranscriptional suppressor when fused to the DNA- binding domain of GAL4. This suppressor effect is not species dependent, since theGAL4-Kid-1construct inhibits CAT activity from a reporter plasmid containing GAL4 binding sites in both COS cells, derived from the green VOL. 13, 1993