GENOMICS 46, 93–102 (1997)ARTICLE NO. GE974980

Genomic Structure and Chromosomal Mapping ofthe Nuclear Orphan Receptor RORg (RORC) Gene

Alexander Medvedev,* Anna Chistokhina,* Takahisa Hirose,† and Anton M. Jetten*,1

*Cell Biology Section, Laboratory of Pulmonary Pathobiology, National Institute of Environmental Health Sciences,National Institutes of Health, Research Triangle Park, North Carolina 27709;

and †Department of Medicine III, Osaka University Medical School,Yamadaoka, Suita, Osaka 565, Japan

Received March 28, 1997; accepted August 15, 1997

as well as a large number of receptors, referred to asThe nuclear orphan receptor subfamily ROR/RZR orphan receptors, for which no ligand has been identi-

is part of the steroid and thyroid hormone/retinoid fied (Evans, 1988; Keller and Wahli, 1993; Wilson etreceptor superfamily and consists of three different al., 1993; Qui et al., 1994; Hirose et al., 1994a, 1995).genes, a, b, and g. In this study, we determined the Based on their sequence aligments the nuclear recep-genomic structure of mouse RORg and the chromo- tors have been divided in several subfamilies (Laudetsomal localization of both mouse RORg and human et al., 1992; Escriva et al., 1997). Most nuclear receptorsRORg (HGMW-approved symbol RORC). The genomic contain five major domains: an A/B domain at thestructure of the mouse RORg gene was derived from amino-terminus, a highly conserved DNA-binding (C)the analysis of P1 vector clones containing large geno- domain containing two ‘‘zinc-finger motifs,’’ a hinge re-mic fragments encoding RORg. These results revealed gion (D domain), a ligand-binding (E) domain, and anthat the mRORg gene has a complex structure con- F domain at the carboxyl-terminus. The A/B and Esisting of 11 exons separated by 10 introns spanning

domains, which are poorly conserved, contain trans-more than 21 kb of genomic DNA. The DNA-bindingactivation functions. Nuclear receptors bind as mono-domain is contained in two exons, 3 and 4, each encod-mers or as homo- or heterodimers to response elementsing one zinc-finger. The splice site between exon 3 andcomposed of a single core motif containing the consen-exon 4 is identical to that found in RAR and TR3 recep-sus sequence PuGGTCA or to direct, palindromic, ortors. RORg is expressed as two mRNAs, 2.3 and 3.0inverted repeats of the core motif spaced by one or morekb in size, that are derived by the use of alternativenucleotides (Forman and Samuels, 1990; Truss andpolyadenylation signals. We show by fluorescence inBeato, 1993; Glass, 1994). Important roles for nuclearsitu hybridization that the mouse RORg gene is lo-receptors have been demonstrated in the control of em-cated on chromosome 3, in a region that corresponds

to band 3F2.1–2.2. The human RORg was mapped to bryonic development, cell proliferation, and differentia-chromosome region 1q21. The results demonstrate tion (De Luca, 1991; Tontonoz et al., 1994; Kastner etthat the RORg genes are located in chromosomal re- al., 1995). In addition, genetic alterations in the expres-gions that are syntenic between mouse and human. sion and structure of these receptors have been impli-q 1997 Academic Press cated in various disease processes (Hughes et al., 1988;

De Luca, 1991; Chomienne et al., 1990).The superfamily of nuclear receptors can be divided

INTRODUCTION into many subfamilies, each containing several closelyrelated genes (Laudet et al., 1992; Escriva et al., 1997).

The nuclear receptor superfamily constitutes a di- Recently, our laboratory cloned mouse and humanverse group of ligand-activated transcriptional factors RORg2 (Hirose et al., 1994b; Medvedev et al., 1996),that share a common modular structure (Evans, 1988; which constitutes the third member of the ROR (alsoLaudet et al., 1992; Kastner et al., 1995; Beato et al., named RZR) subfamily (Becker-Andre et al., 1993;1995). This superfamily includes receptors for steroid Carlberg et al., 1994; Giguere et al., 1994). The aminohormones, thyroid hormone, retinoids, and vitamin D acid sequence of RORg is highly conserved between

species; the amino acid sequences between mouse andSequence data from this article have been deposited with the

human RORg exhibit 88% identity (Hirose et al.,EMBL/GenBank Data Libraries under Accession Nos. AF019655–AF019660.

1 To whom correspondence should be addressed. Telephone (919) 2 The HGMW-approved symbol for the gene described in this paperis RORC.541-2768. Fax: (919) 541-4133. E-mail: jetten@niehs.nih.gov.

930888-7543/97 $25.00

Copyright q 1997 by Academic PressAll rights of reproduction in any form reserved.

AID GENO 4980 / 6r4e$$$301 10-31-97 15:15:57 gnmas

MEDVEDEV ET AL.94

FIG. 1. Genomic structure of the mouse RORg gene. Uppercase letters indicate exon sequences, and lowercase letters indicate intronsequences or upstream and downstream flanking sequences. Exons and introns are indicated on the right. The coding regions are markedby single-letter amino acid code. The numbers on the left indicate the amino acid in RORg. The start codon (ATG), the stop codon (TGA),and the polyadenylation signals are underlined. The start of the sequence of the RORg cDNA is marked by the solid arrow. The end of theprimer extension is marked by the open arrow.

1994b; Medvedev et al., 1996). The amino acid se- a different tissue distribution, suggesting that thesereceptors have distinct biological roles. While RORbquences of hRORg and mRORg are 50–51% identical

to those of the RORa and b receptors. The A/B, C, D, appears to be uniquely expressed in brain and retina,RORa and g were found to be highly expressed in sev-and E domains between mRORg and mRORa1 exhibit

50, 90, 34, and 57% homology, respectively (Giguere et eral tissues (Becker-Andre et al., 1993; Carlberg et al.,1994; Hirose et al., 1994b; Medvedev et al., 1996).al., 1994; Medvedev et al., 1996). These receptors do

not have an F domain. RORa and g have been demonstrated to be inducedduring adipocyte differentiation in several preadipo-The RORs have been shown to bind as monomers

to single core motifs preceded by an AT-rich sequence cyte cell lines (Adachi et al., 1996; Austin et al., 1997).A deletion in the RORa gene was found to be linked to(Giguere et al., 1994; Medvedev et al., 1996; Greiner

et al., 1996). Recently, several putative target genes, defects in staggerer mice with severe cerebellar ataxia(Hamilton et al., 1996; Matysiak-Scholtze and Nehls,including gF-crystallin, p21WAF1/CIP1, laminin B1, and

5-lipoxygenase, have been shown to contain response 1997). These studies support important regulatoryroles for RORs in different developmental and differen-elements in their promoter region that are able to bind

RORa (Tini et al., 1995; Schrader et al., 1996; Matsui, tiation processes.In this study, we describe the isolation of P1 vector1996). Each member of the ROR subfamily exhibits

AID GENO 4980 / 6r4e$$$301 10-31-97 15:15:57 gnmas

NUCLEAR ORPHAN RECEPTOR RORg (RORC) GENE 95

FIG. 1—Continued

bp. A new forward primer FP2 (5*-CACTTGGCTGTTTCAGCGGTC)clones containing genomic fragments of the mouse anddesigned from the intron sequence and a reverse primer RP2 (5*-human RORg genes. With the help of these clones,ATCGGTTGCGGCTGGTTCGGT) designed from the exon sequencewe characterized the genomic structure of the mouse were synthesized and amplified a 313-bp product from genomic

RORg gene and determined the chromosomal localiza- mouse DNA. This specific primer pair was used to screen by PCR ation of the mouse and human RORg genes. The mouse library of P1 vector clones containing 75- to 150-kb inserts of mouse

genomic DNA (Pierce et al., 1992; Genome Systems, St. Louis, MO).RORg gene was mapped to chromosome 3 and the hu-A similar strategy was employed for the isolation of P1 vector clonesman gene to chromosome 1. In addition, we demon-containing the human genomic RORg gene. The following amplifica-strate that RORg is expressed as two transcripts that tion cycles were employed: 30 cycles of 947C, 607C, and 727C for 1

are generated by the use of two different polyadenyla- min each, plus extension at 727C for 7 min. Three positive clonestion signals. were obtained. P1 vector DNA isolated by the standard NaOH-SDS

lysis method was purified by phenol/chloroform extraction and pre-cipitated with ethanol. DNA was digested with BamHI and frag-MATERIALS AND METHODSments were analyzed by Southern analysis using fragments fromdifferent regions of the mRORg cDNA as a probe. One of two P1Isolation of P1 vector clones. The plasmids mRORg-BSK andvector clones (P1-mRORg) that contained the full coding region ofhRORg-BSK containing the full coding region of the mouse and hu-RORg was used for further investigation. Fragments hybridizing toman RORg, respectively, were described previously (Hirose et al.,the RORg probe were cut from the gel, purified with microcon 301994b; Medvedev et al., 1996). From the cDNA sequence several(Amicon), and subcloned into Bluescript SK II. The size and orderprimer pairs were selected and used in PCRs with mouse genomicof BamHI fragments were subsequently confirmed by additional re-DNA or mRORg-BSK as template. The primer set (FP1, 5*-GTTATC-striction analysis, PCR, and sequence analysis.ACCTGTGAGGGGTG and RP2, 5*-GACATGCCCAGAGCCAGGCA)

DNA sequencing. Plasmids were purified using Wizard miniprepamplified a product of the expected size of 158 bp from mRORg-BSKor midiprep kits (Promega). Manual sequencing was performed usingDNA and amplified a 400-bp fragment from mouse genomic DNA.the dideoxynucleotide chain-termination method and the SequenaseThe sequence of the 400-bp fragment demonstrated that it contains

in addition to the 158-bp RORg coding sequence, an intron of 255 Quick-denature plasmid sequencing kit (Amersham). Automatic se-

AID GENO 4980 / 6r4e$$$301 10-31-97 15:15:57 gnmas

MEDVEDEV ET AL.96

TABLE 1

Summary of the Size of the Exons and Introns and the DNA Sequence of the Exon/Intron Junctionsin the Mouse RORg Gene

Exon Exon size 5*-Exon/intron junction Intron size 3*-Exon/intron junction aa interrupteda

1 146 bp TCTCGGG146gtaaga Ç2.0 kb ttacagA147GCTGCT Glu14

2 31 bp CACACCT176gtgagt Ç8.5 kb ttccagC177ACAAAT Ser24

3 82 bp GTGCAAG262gtgagt 255 bp tcccagG263GCTTCT Lys52/Gly53

4 140 bp CGAGATG404gtgaga 1153 bp cttcagC405TGTCAA Ala100

5 507 bp GACATAG911gtgagc 217 bp ccccagA912GTACCT Glu269

6 121 bp GAGGAAG1033gtaagg Ç1.4 kb ccgcagT1034CAATGT Lys309/Ser310

7 132 bp ACAGCAG1166gtgcac 167 bp atccagG1167AGCAAT Gly354

8 107 bp GCCTTGG1275gtgagg 157 bp cggcagG1276CTGCAG Gly390

9 110 bp AATGCCA1385gtgagt Ç1.3 kb cttcagA1386CCGTCC Asn427

10 109 bp AGCCAAG1495gtagga Ç4 kb tcctcagC1496TGCCAC Lys463/Leu464

11 1398 bp

Note. Exon sequences are shown in uppercase letters, and intron sequences are shown in lowercase letters.a The nucleotide numbers indicate the positive within the mouse RORg mRNA.

quencing was carried out using a Dye Terminator Cycle Sequencing 5*-RACE and nested primer PCR. Three reverse, mRORg-specificprimers NP4-6 (5*-GTGAGGCGCCCAGTGGTAGCT), NP5 (5*-GAC-Ready Reaction kit (Perkin–Elmer) and an ABI Prism 377 automatic

sequencer. DNA and deduced protein sequences were analyzed by ATGCCCAGAGCCAGGCA), and NP6 (5*-CAAGGGATCACTTCA-ATTTG) were used in 5*-RACE as well as in nested primer PCR.the UWGCG (Devereux et al., 1984) and MacVector (Kodak) sequence

analysis software packages. These primers were designed to be specific for regions in exon 5, 4,and 3, respectively. 5*-RACE was performed with a 5*/3 * RACE kitFluorescence in situ hybridization (FISH). The regional chromo-(Boehringer Mannheim) according to the manufacturer’s protocol.somal localization was determined by FISH using fragments of geno-The first strain cDNA was synthesized with AMV reverse tran-mic DNA containing either the human or the mouse RORg genescriptase and the primer NP4. The cDNA was then purified and acloned into P1 vectors as a probe (Stokke et al., 1995). FISH waspoly(A) tail added to its 3 *-end by terminal transferase reaction.carried out by Genome Systems. Two different methods were em-After purification, this modified cDNA was used as a template inployed. In the first method, DNA was labeled with digoxigenin–two subsequent rounds of PCR with oligo(dT) and the two remainingdUTP by nick-translation. Labeled probe was combined with shearedmRORg-specific primers NP5 or NP6. The PCR fragments were aga-mouse DNA and hybridized to normal metaphase chromosomes de-rose purified, subcloned, and sequenced. The inserts of all clonesrived from embryonic stem cells from male mice. Specific hybridiza-examined contained sequence identical to that of the mRORg cDNA.tion signals were detected by incubating the hybridized slides with

Nested primer PCR was performed in three rounds (25 cycles/fluoresceinated anti-digoxigenin antibodies followed by counter-round) with an Expand High Fidelity PCR System (Boehringerstaining with propidium iodide. These results indicated that theMannheim) and three different sets of nested primers consisting ofmouse RORg gene was located on chromosome 3. This was confirmedthree forward vector primers (NP1-3) and three reverse, mRORg-by cohybridizing the labeled P1 vector DNA with a biotin-labeledspecific primers (NP4-6). Mouse embryonic (17th day) and liverprobe specific for the centromeric region of chromosome 3 and coun-cDNA libraries were used as a template in the first round of PCRterstaining with 4*,6-diamidino-2-phenylindole (DAPI). The latter re-(25 cycles) with primers NP1 and NP4. The PCR products obtainedsulted in the specific labeling of the centromeric region of chromo-in each round were used in the subsequent rounds with NP2 andsome 3 after incubation with Texas red-labeled avidin. A total of 80NP5 or with NP3 and NP6. The final PCR products were then agarosemetaphase cells were analyzed, 71 of which exhibited specific label-purified, subcloned, and sequenced. The inserts of all clones exam-ing. Measurements were done on 10 specifically hybridized chromo-ined contained sequence identical to that of the mRORg cDNA.somes 3.

To localize the human RORg gene, digoxigenin-labeled DNA was Primer extension. The oligonucleotide primer PE-1 (5*-GTGGGT-CTTCTTTGCAGCCA) was end-labeled with [g-32P]ATP by T4 poly-hybridized to normal metaphase chromosomes derived from PHA-

stimulated human peripheral blood lymphocytes from a male donor. nucleotide kinase. After 10 min of labeling at 377C, the enzyme wasinactivated by incubation of the reaction mixture for 2 min at 907C.After the initial indication that the human RORg gene was localized

on chromosome 1, the labeled P1 vector DNA was cohybridized with The radiolabeled oligonucleotide was purified over NAP-5 columns(Pharmacia). For primer extension analysis 15 mg of total RNA froma biotin-labeled probe specific for the heterochromatic region of chro-

mosome 1. A total of 80 metaphase cells were analyzed, with 73 muscle or liver was annealed to the radiolabeled PE-1 and thenreverse transcribed using an AMV Reverse Transcriptase Primerexhibiting specific labeling. Measurements were performed on 10

specifically hybridized chromosomes 1. Extension System (Promega) according to the manufacturer’s proto-col. Primer extension products were analyzed on a denaturing, 6%Northern blot analysis. Total RNA from murine D1 adipocytespolyacrylamide sequencing gel together with a sequence ladder ob-was isolated as described previously (Adachi et al., 1996). Total RNAtained by dideoxy-sequencing of unrelated plasmid DNA, which(30 mg) was electrophoresed through formaldehyde 1.2% agarose gel,served as a size marker.blotted to Nytran membrane (Schleicher & Schuell), and UV-cross-

linked. The blots were hybridized with either a 32P-labeled, 2.1-kb5*-fragment of RORg cDNA (Medvedev et al., 1996) or the 3*-end of

RESULTS AND DISCUSSIONthe 3 *-UTR of RORg Hybridizations were performed for 1–2 h at687C using Quickhyb reagent (Stratagene), and blots were washedtwice with 21 SSC, 0.05% SDS at room temperature for 15 min Genomic structure of the mouse RORg gene. P1 vec-and subsequently with 0.11 SSC, 0.1% SDS at 607C for 30 min.

tor clones containing large fragments of mouse genomicAutoradiography was carried out with Hyperfilm-MP (Amersham)at 0707C. DNA encoding the RORg gene were isolated by screen-

AID GENO 4980 / 6r4e$$$301 10-31-97 15:15:57 gnmas

NUCLEAR ORPHAN RECEPTOR RORg (RORC) GENE 97

FIG. 2. Schematic presentation of the genomic structure of the mouse RORg gene. (A) The positions of the 11 exons corresponding tothe RORg mRNA are shown. The sizes of the RORg transcripts, 2.3 and 2.9 kb, are indicated. The translation start site, the stop codon,and the two major polyadenylation signals (PAS1,2) are marked. Open boxes in top row indicate the DNA-binding domain (DBD) and putativeligand-binding domain (LDB). Shaded area represents open reading frame. (B) Genomic structure of the RORg gene locus. Boxes indicateexons, black boxes represent open reading frames. BamHI (B), KpnI (K), EcoRI (E), HindIII (H), XhoI (X), and XbaI (Xb) restriction sitesare marked. (C) Sequence of DNA-binding domain of RORg in relation to genomic structure. Splice sites for the RORg, thyroid hormonereceptor (T3R), the retinoic acid receptor (RAR), the orphan receptor Nur77/NGFIB, and the steroid hormone receptors (SH) are indicated.NR indicates the splice site conserved among all nuclear receptors.

ing a P1 vector library by PCR using two specific prim- quently subcloned into pBluescript and then sequencedusing either internal primers derived from the cDNAers, FP2 and RP2. These primers yielded under appro-

priate conditions a PCR product of about 300 bp with sequence or vector primers. The order and orientationof the cloned BamHI fragments were established bymouse genomic DNA as template. Three P1 vector

clones containing the RORg gene were obtained; one PCR using P1-mRORg DNA as template and primersets that anneal to the ends of two different BamHIof two clones P1-mRORg containing the whole coding

region was selected for further analysis. The P1 vector fragments. This PCR positioned all cloned BamHI frag-ments and identified an additional fragment of 2.4 kb ofwas cut by BamHI and subjected to Southern analysis

using a 32P-labeled cDNA fragment containing the en- the mRORg gene that did not hybridize to the mRORgcDNA probe in Southern analysis of BamHI-digestedtire cDNA of RORg (Medvedev et al., 1996) as a probe.

Southern hybridization showed that BamHI fragments P1-mRORg DNA. This fragment was cloned and foundto consist of two BamHI fragments of 0.9 and 1.5 kb.of 8, 6.5, 4.7, 4.4, and 2 kb were recognized by the

labeled cDNA probe; these fragments were subse- The restriction map of the mRORg gene was con-

AID GENO 4980 / 6r4e$$$301 10-31-97 15:15:57 gnmas

MEDVEDEV ET AL.98

first zinc-finger and different from that of the orphanreceptors COUP-TF and Nur77/NGFI-B, which haveeither no splice site or a splice site at the fifth aminoacid, respectively (Watson and Milbrandt, 1989; Rit-chie et al., 1990) (Fig. 2C). The position of the nextsplice site (intron 4) at the C-terminus of the DNA-binding domain of RORg (Ala100) is highly conservedamong all nuclear receptors (Ponglikitmongkol et al.,1988; Ritchie et al., 1990; Lehmann et al., 1991). Basedon the positioning of these splice sites, the nuclearreceptors have been divided into different groups that

FIG. 3. The 2.3- and 3.0-kb RORg transcripts are generated by diverged from each other early during the evolutionarythe use of alternative polyadenylation signals. Total RNA (30 mg)isolated from D1 adipocytes was examined by Northern analysis us-ing 32P-labeled probes encoding a 2.1-kb 5*-fragment of the RORgcDNA (lane 1) or a 0.8-kb genomic fragment encoding a sequencejust downstream of the first polyadenylation signal (lane 2). Theformer fragment hybridizes to both transcripts while the latter hy-bridizes only to the 3.0-kb transcript.

structed based on the analysis of all cloned fragmentswith different restriction enzymes, including EcoRI,HindIII, XhoI, XbaI, and KpnI, and analyzed by South-ern blotting with radioactive probes specific for differ-ent regions of the cDNA. The exon–intron structure ofthe mRORg gene was confirmed by PCR analysis usingP1-mRORg DNA as a template and various primer setsthat amplify overlapping genomic DNA fragments. Asummary of the PCR, sequencing, and restriction map-ping analyses is presented in Fig. 1. These analysesrevealed that the mouse RORg gene spans more than21 kb and is composed of 11 exons separated by 10introns. A summary of the various sizes of the exonsand introns and the sequence of the exon/intron junc-tions is shown in Table 1. The sequences of these junc-tions are consistent with the consensus (A62G77/g100t100a60a74g88 for the 5* donor and y87ny97a100g100/G55

for the 3 * acceptor site) for known splice sites withineukaryotic genes (Mount, 1982). The DNA-binding do-main (from Cys31 to Cys91) of RORg is contained withintwo exons, 3 and 4 (Fig. 2). Exon 5 encodes the hingedomain while exons 6–10 encode the putative ligand-binding domain. A schematic view of the genomic struc-ture and the relationship between the various exonsand the DNA- and putative ligand-binding domains ispresented in Figs. 2A–2B.

Some of the intron/exon junctions are conserved be-tween members of the nuclear receptor superfamily.The location of the second intron (at Ser24) in the RORggene is shared with a splice site in the RORa gene(Giguere et al., 1994). The splice site (intron 3) betweenthe two zinc-fingers of RORg is located at the second FIG. 4. Determination of the transcriptional initiation site of theamino acid from the last cysteine of the first zinc-finger mouse RORg gene. For primer extension analysis, 15 mg of total(Fig. 2C); this position is shared with that of the thy- RNA from muscle or liver was annealed to the end-labeled oligonucle-

otide PE-1 and then reverse-transcribed as described under Materi-roid hormone and retinoic acid nuclear receptors, in-als and Methods. Primer extension products were analyzed on acluding the retinoic acid receptor RARg (Ritchie et al.,denaturing, polyacrylamide gel together with a sequence ladder1990; Lehmann et al., 1991). This splice site is different (lanes A, C, G, and T) obtained by dideoxy sequencing of unrelated

from that of the steroid hormone receptor that is lo- plasmid DNA that served as a size marker. The arrow indicates themajor primer extension product of 171 bp.cated 10 amino acids 3 * from the last cysteine of the

AID GENO 4980 / 6r4e$$$301 10-31-97 15:15:57 gnmas

NUCLEAR ORPHAN RECEPTOR RORg (RORC) GENE 99

process. In this respect, ROR fits in the T3R/RAR sub-group.

Alternative polyadenylation signals. Northern blotanalysis with RNA from various tissues demonstratedthat a probe consisting of a 2.1-kb 5*-fragment of theRORg cDNA hybridizes to two transcripts about 2.3and 3.0 kb (Hirose et al., 1994b; Medvedev et al., 1996)(Fig. 3). These transcripts could be generated by alter-native splicing, use of alternative promoters and tran-scription start sites, and/or the selection of alternativepolyadenylation signals. The RORa gene has been re-ported to generate through alternative splicing threeisoforms that differ at the amino terminus (Giguereet al., 1994). This alternative splicing involves exonsupstream from the splice site that is conserved betweenRORa and RORg (junction of exons 2 and 3 of mRORg).To examine the possibility of alternative splicing of FIG. 5. Regional mapping of the RORg gene by fluorescence in

situ hybridization to mouse chromosome 3. (A) The location of theRORg transcripts, we performed 5*-RACE using RNARORg gene was identified on metaphase chromosomes from murinefrom liver and muscle as a template and three nestedmale embryonic stem cells using a digoxigenin dUTP-labeled, geno-reverse primers annealing to a region in exon 5, 4, or mic fragment containing the RORg gene. Chromosome 3 was identi-

3. The sequences of the inserts from seven different fied by cohybridization with a biotinylated probe specific for the cen-tromere of chromosome 3. Only the staining for RORg is shown. (B)clones were found to be all identical to that reportedThe idiogram indicates that mouse RORg maps to band 3F2.1–2.2.for mRORg and extended into the 5*-UTR. In a secondThe arrowhead indicates the interval within which the hybridizationapproach, we performed nested PCR using a mousesignal was detected on a sample of 71 chromosomes. No specific

embryonic and liver cDNA library. The inserts of the signal was detected on other chromosomes.19 clones analyzed varied in length; however, their se-quence was identical to that of the 5*-end of mRORg.

in the control of the stability or rate of translation ofThese results of these two strategies did not providethe RORg mRNA.any evidence for alternative splicing of mRORg mRNA

Transcription initiation site. To determine thein the 5* region.mRORg transcription initiation site(s), primer extensionWe next examined the possibility of the use of alter-was performed using RNA from mouse liver and musclenative polyadenylation signals. The 3 *-end sequence ofas a template and the reverse primer PE-1 designedthe RORg cDNAs described previously (Medvedev etfrom the sequence of the second exon (Fig. 4). Reactionsal., 1996) contained a poly(A)/ tail eight nucleotideswith both templates identified a major extension productfrom the sequence AATAAGAATATA containing twoof 171 bp. A weak, 86-bp extension product was alsoconsecutive polyadenylation signals (Wickens, 1990).observed. These results suggest that the transcriptionTo determine whether the larger transcript was de-start site of RORg mRNA is 106 bases from the startrived by the use of an alternative polyadenylation sig-codon. Location of the transcription initiation site wasnal, we amplified a 0.8-kb region just downstream fromconfirmed by RT-PCR using mouse liver RNA as a tem-the first polyadenylation signal with genomic DNA asplate. Oligonucleotides just upstream and downstreama template and used it as a probe in Northern blotfrom the putative transcription initiation site were usedanalysis. As shown in Fig. 3, in contrast to the 2.1-as forward primers and PE-1 was used as a reversekb 5*-probe, this probe hybridized only to the 3.0-kbprimer. Only the RT-PCR with the downstream primertranscript. Examination of the sequence of this frag-yielded a PCR product of the predicted size while PCRment identified two consecutive polyadenylation sig-with the upstream primer did not give a PCR productnals, AGTAAA and CATAAA (Wickens, 1990), 0.7 kb(data not shown). These results support the location ofdownstream from the first site. A labeled probe derivedthe transcriptional initiation site identified by primerfrom the genomic fragment just downstream fromextension.these two polyadenylation signals did not hybridize to

these two RNAs (not shown). These results indicate Chromosomal localization of the mouse RORg gene.The regional chromosomal localization was determinedthat the generation of the 2.3- and 3.0-kb RORg tran-

scripts is related to the use of alternative polyadenyla- by FISH using fragments of genomic DNA of about 100kb containing either the mouse or the human RORgtion signals. It is interesting to note that the ratio

between the two transcripts can vary dramatically be- gene cloned into P1 vectors as a probe. These P1 vectorclones were obtained by screening a P1 vector librarytween different tissues (Medvedev et al., 1996), sug-

gesting that it is regulated in a tissue-specific manner. by PCR using two specific primers that yielded a PCRproduct of about 300 bp with genomic DNA as template.The different lengths of the 3 *-UTR may be important

AID GENO 4980 / 6r4e$$$301 10-31-97 15:15:57 gnmas

MEDVEDEV ET AL.100

FIG. 6. Regional mapping of the RORg gene by fluorescence in situ hybridization to human chromosome 1. (A) The RORg gene waslocalized on normal metaphase chromosomes derived from PHA-stimulated human peripheral blood lymphocytes from a male donor usinga digoxigenin dUTP-labeled, genomic fragment containing the RORg gene. Chromosome 1 was identified by cohybridization with a biotinyl-ated probe specific for the centromere of chromosome 1. Only the staining for RORg is shown. (B) The idiogram indicates that RORg mapsto band 1q21. The arrowhead indicates the interval within which the hybridization signal was detected on a sample of 73 chromosomes.No specific signal was detected on other chromosomes.

DNA was labeled with digoxigenin–dUTP by nick- procedure mapped the RORg gene to human chromo-some 1. This assignment is in agreement with thattranslation. Labeled probe was combined with sheared

DNA and hybridized to normal metaphase chromo- obtained from Southern analysis using a BIOS Blotsomes derived from male embryonic stem cells. Specific Somatic Cell Hybrid Panel (New Haven, CT) consistinghybridization signals were detected by incubating the of TaqI-digested genomic DNA isolated from a panelhybridized slides in fluoresceinated anti-digoxigenin of human/hamster and human/mouse hybrid cell linesantibodies followed by counterstaining with propidium (Hirose et al., 1994b). In the second experiment, a bio-iodide. This resulted in the specific labeling of the fourth tin-labeled probe specific for the heterochromatic re-largest chromosome, suggesting that the RORg gene gion of chromosome 1 was cohybridized with the probewas contained in chromosome 3 (not shown). This was specific for hRORg. The band assignment was deter-confirmed by a second experiment in which the labeled mined by measuring the fractional chromosome lengthP1 vector DNA was cohybridized with a biotin-labeled and by analyzing the banding pattern. Measurementsprobe specific for chromosome 3. The latter resulted in of 10 specifically hybridized chromosomes 1 demon-the specific labeling of the centromeric region of chromo- strated that RORg is located at a position that is 7%some 3 after incubation with Texas red avidin. The re- of the distance from the heterochromatic–euchromaticgional assignment of the RORg probe was determined boundary to the telomere of chromosome arm 1q, anby the analysis of the fractional chromosome length; area that corresponds to band 1q21. This is shown sche-10 specifically hybridized chromosomes were measured. matically in Fig. 6.These calculations indicated that the RORg gene The biological functions of RORg have not been pre-mapped to a position that is 54% of the distance from cisely established. RORg has been shown to be ex-the heterochromatic–euchromatic boundary to the te- pressed in a tissue-specific manner; it is highly ex-lomere of chromosome 3, an area that corresponds to pressed in liver, kidney, and thymus. Recent studiesband 3F2.1–2.2. This is schematically shown in Fig. 5. have indicated that members of the ROR/RZR family

play important roles in the regulation of development,Chromosomal localization of the human RORg gene.Similar procedures were used to map the chromosomal differentiation, and cell growth. A deletion in the RORa

gene identified in staggerer mice (Hamilton et al., 1996;localization of hRORg. In this case, normal metaphasechromosomes were derived from PHA-stimulated pe- Matysiak-Scholtze and Nehls, 1997) supports this con-

clusion. RORg has been demonstrated to be inducedripheral blood lymphocytes from a male donor. The first

AID GENO 4980 / 6r4e$$$301 10-31-97 15:15:57 gnmas

NUCLEAR ORPHAN RECEPTOR RORg (RORC) GENE 101

Greiner, E. F., Kirfel, J., Greschik, H., Dorflinger, U., Becker, P.,during fat cell differentiation in 3T3-L1 and D1 preadi-Mercep, A., and Schule, R. (1996). Functional analysis of retinoidpocytes (Adachi et al., 1996; Austin et al., 1997), sug-Z receptor b, a brain-specific nuclear orphan receptor. Proc. Natl.gesting a role in the control of adipocyte function. In Acad. Sci. USA 93: 10105–10110.

addition, RORg is highly expressed in T-lymphocytes Hamilton, B. A., Frankel, W. N., Kerrebrock, A. W., Hawkins, T. L.,but is not expressed in B-lymphocytes (Hirose et al., FitzHugh, W., Kusumi, K., Russell, L. B., Mueller, K. L., van Ber-

kel, V., Birren, B. W., Kruglyak, L., and Lander, E. S. (1996). Dis-1994b; Medvedev et al., 1996; unpublished observa-ruption of the nuclear hormone receptor RORa in staggerer mice.tions). The latter finding suggests a role for RORg inNature 379: 736–739.regulating specific T-cell functions. Genetic alterations

Hirose, T., Fujimoto, W., Yamaai, T., Kim, K. H., Matsuura, H., andinvolving the human 1q21 region where the RORg geneJetten, A. M. (1994a). TAK1: Molecular cloning and characteriza-

is located have been implicated in a variety of malig- tion of a new member of the nuclear receptor family. Mol. Endocri-nancies including lymphomas and renal carcinomas nol. 8: 1667–1680.(Whang-Peng et al., 1995; Weterman et al., 1996). Fu- Hirose, T., Smith, R. J., and Jetten, A. M. (1994b). RORg: The third

member of the ROR/RZR orphan receptor subfamily that is highlyture studies have to determine whether genetic alter-expressed in skeletal muscle. Biochem. Biophys. Res. Commun.ations in the RORg gene are involved in these disease205: 1976–1983.processes.

Hirose, T., O’Brien, D. A., and Jetten, A. M. (1995). RTR: A newmember of the nuclear receptor superfamily that is highly ex-pressed in murine testis. Gene 152: 247–251.REFERENCES

Hughes, M. R., Malloy, P. J., Kieback, D. G., Kesterson, R. A., Pike,J. W., Feldman, D., and O’Malley, B. W. (1988). Point mutations inAdachi, H., Dawson, M. I., and Jetten, A. M. (1996). Suppression bythe human vitamin D receptor gene associated with hypocalcemicretinoids of the induction of the CAAT/enhancer-binding proteinrickets. Science 242: 1702–1705.a and the nuclear receptors PPARg and RORg during adipocyte

differentiation of 3T3-L1 cells. Mol. Cell. Differ. 4: 365–381. Kastner, P., Mark, M., and Chambon, P. (1995). Nonsteroid nuclearreceptors: What are the genetic studies telling us about their roleAustin, S., Medvedev, A., Yan, Z-H., Adachi, H., Hirose, T., and Jet-in real life? Cell 83: 859–869.ten, A. M. (1997). Induction of the nuclear orphan receptor RORg

during adipocyte differentiation of D1 and 3T3-L1 cells. Cross-talk Keller, H., and Wahli, W. (1993). Peroxisome proliferator-activatedbetween nuclear receptor pathways. Cell Growth Differ. Submit- receptors. A link between endocrinology and nutrition? Trends En-ted. docrinol. Metab. 4: 291–296.

Beato, M., Herrlich, P., and Schutz, G. (1995). Steroid hormone recep- Laudet, V., Hanni, C., Coll, J., Catzeflis, and Stehelin, D. (1992).tors: Many actors in search of a plot. Cell 83: 851–857. Evolution of the nuclear receptor gene superfamily. EMBO J. 11:

1003–1013.Becker-Andre, M., Andre, E., and DeLamarter, J. F. (1993). Identifi-cation of nuclear receptor mRNAs by RT-PCR amplification of con- Lehmann, J. M., Hoffmann, B., and Pfahl, M. (1991). Genomic orga-served zinc-finger motif sequences. Biochem. Biophys. Res. Com- nization of the retinoic acid receptor gamma gene. Nucleic Acidsmun. 194: 1371–1379. Res. 19: 573–578.

Carlberg, C., van Huijsduijnen, R. H., Staple, J. K., Delamarter, J. F., Matsui, T. (1996). Differential activation of the murine laminin B1and Becker-Andre, M. (1994). RZRs, a new family of retinoid-re- gene promoter by RAR alpha, ROR alpha, and AP1. Biochem. Bio-lated orphan receptors that function as both monomers and homo- phys. Res. Commun. 220: 405–410.dimers. Mol. Endocrinol. 8: 757–770. Matysiak-Scholtze, U., and Nehls, M. (1997). The structural integrity

of RORa isoforms is mutated in staggerer mice: Cerebellar coex-Chomienne, C., Ballerini, P., Balitrand, N., Huang, M. E., Krawice,pression of RORa1 and RORa4. Genomics 43: 78–84.I., Castaigne, S., Fenaux, P., Tiollais, P., Dejean, A., Degos, L., and

de The, H. (1990). The retinoic acid receptor a gene is rearranged in Medvedev, A., Yan, Z-H., Hirose, T., Giguere, V., and Jetten, A. M.retinoic acid-sensitive promyelocytic leukemias. Leukemia 4: 802– (1996). Cloning of a cDNA encoding the murine orphan receptor807. RORg and characterization of its response element. Gene 181:

199–206.De Luca, L. M. (1991). Retinoids and their receptors in differentia-tion, embryogenesis and neoplasia. FASEB J. 5: 2924–2933. Missbach, M., Jagher, B., Sigg, I., Nayeri, S., Carlberg, C., and Wie-

senberg, I. (1996). Thiazolidine diones, specific ligands of the nu-Devereux, J., Haeberli, P., and Smithies, O. (1984). A comprehensiveclear receptor retinoid Z receptor/retinoid acid receptor-related or-set of sequence analysis programs for the VAX. Nucleic Acids Res.phan receptor a with potent antiarthritic activity. J. Biol. Chem.12: 387–395.271: 13515–13522.Escriva, H., Safi, R., Hanni, C., Langlois, M., Saumitou-Laprade, P.,

Mount, S. M. (1982). A catalogue of splice junction sequences. NucleicStehelin, D., Capron, A., Pierce, R., and Laudet, V. (1997). LigandAcids Res. 10: 459–472.binding was acquired during evolution of nuclear receptors. Proc.

Natl. Acad. Sci. USA 94: 6803–6808. Pierce, J. C., Sternberg, N., and Sauer, B. (1992). A mouse genomiclibrary in the bacteriophage P1 cloning system: Organization andEvans, R. M. (1988). The steroid and thyroid receptor superfamily.characterization. Mamm. Genome 3: 550–559.Science 240: 889–895.

Ponglikitmongkol, M., Green, S., and Chambon, P. (1988). GenomicForman, B. M., and Samuels, H. H. (1990). Interaction among a sub-organization of the human oestrogen receptor gene. EMBO J. 7:family of nuclear hormone receptors: The regulatory zipper model.3385–3388.Mol. Endocrinol. 4: 1293–1301.

Qui, Y., Tsai, S. Y., and Tsai, M-J. (1994). COUP-TF an orphan mem-Giguere, V., Tini, M., Flock, G., Ong, E., Evans, R. M., and Otulakow-ber of the steroid/thyroid hormone receptor superfamily. Trendsski, G. (1994). Isoform-specific amino-terminal domains dictateEndocrinol. Metab. 5: 234–239.DNA-binding properties of RORa, a novel family of orphan nuclear

receptors. Genes Dev. 8: 538–553. Ritchie, H. H., Wang, L-H., Tsai, S., O’Malley, and Tsai, M-J. (1990).COUP-TF gene: A structure unique for the steroid/thyroid hor-Glass, C. K. (1994). Differential recognition of target genes by nu-mone receptor superfamily. Nucleic Acids Res. 18: 6857–6862.clear receptor monomers, dimers and heterodimers. Endocr. Rev.

15: 391–407. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989). ‘‘Molecular

AID GENO 4980 / 6r4e$$$301 10-31-97 15:15:57 gnmas

MEDVEDEV ET AL.102

Cloning: A Laboratory Manual,’’ Cold Spring Harbor Laboratory tions with deoxyribonucleic acid and transcriptional factors. Endo-crine Rev. 14: 459–479.Press, Cold Spring Harbor, NY.

Watson, M. A., and Milbrandt, J. (1989). The NGFI-B gene, a tran-Schrader, M., Danielsson, C., Wiesenberg, I., and Carlberg, C. (1996).scriptionally inducible member of the steroid receptor gene super-Identification of natural response elements of the nuclear receptorfamily: Genomic structure and expression in rat brain after seizureROR. They also bind COUP homodimers. J. Biol. Chem. 271:induction. Mol. Cell. Biol. 9: 4213–4219.19732–19736.

Weterman, M. A., Wilbrink, M., Dijkhuizen, T., van den Berg, E.,Stokke, T., Collins, C., Kuo, W. L., Kowbel, D., Shadravan, F., Tan- and Geurts van Kessel, A. (1996). Fine mapping of the 1q21

ner, M., Kallioniemi, A., Kallioniemi, O. P., Pinkel, D., and Deaven, breakpoint of the papillary renal cell carcinoma-associated (X;1)L. (1995). A physical map of chromosome 20 established using translocation. Hum. Genet. 98: 16–21.fluorescence in situ hybridization and digital analysis. Genomics Whang-Peng, J., Knutsen, T., Jaffe, E. S., Steinberg, S. M., Raffeld,26: 134–137. M., Zhao, W. P., Duffey, P., Condron, K., Yano, T., and Longo, D. L.

(1995). Sequential analysis of 43 patients with non-Hodgkins’sTini, M., Fraser, R. A., and Giguere, V. (1995). Functional interac-lymphoma: Clinical correlations with cytogenetic, histologic, im-tions between retinoic acid receptor-related orphan receptor (RORmunophenotyping and molecular studies. Blood 85: 203–216.alpha) and the retinoic acid receptors in the regulation of the

Wickens, M. (1990). How the messenger got its tail: Addition ofgamma F-crystallin promoter. J. Biol. Chem. 270: 20156–20161.poly(A) in the nucleus. Trends Bio. 15: 277–281.

Tontonoz P., Hu, E., and Spiegelman, B. M. (1994). Stimulation of Wilson, T. E., Fahrner, T. J., and Millbrandt, J. (1993). The orphanadipogenesis in fibroblasts by PPARg2, a lipid-activated transcrip- receptors NGFI-B and steroidogenic factor 1 establish monomertion factor. Cell 79: 1147–1156. binding as a third paradigm of nuclear receptor–DNA interaction.

Mol. Cell. Biol. 13: 5794–5804.Truss, M., and Beato, M. (1993). Steroid hormone receptors: Interac-

AID GENO 4980 / 6r4e$$$301 10-31-97 15:15:57 gnmas