Proc. Natl. Acad. Sci. USAVol. 92, pp. 10604-10608, November 1995Developmental Biology

Isolation, regulation, and DNA-binding properties of threeDrosophila nuclear hormone receptor superfamily members

(20-hydroxyecdysone/Drosophila metamorphosis/gene regulation)

GREGORY J. FISK AND CARL S. THUMMELHoward Hughes Medical Institute and Department of Human Genetics, 5200 Eccles Institute of Human Genetics, University of Utah, Salt Lake City, UT 84112

Communicated by Mario R. Capecchi, University of Utah, Salt Lake City, UT, August 2, 1995 (received for review June 16, 1995)

ABSTRACT We have designed a rapid cloning and screen-ing strategy to identify new members of the nuclear hormonereceptor superfamily that are expressed during the onset ofDrosophila metamorphosis. Using this approach, we isolatedthree Drosophila genes, designated DHR38, DHR78, andDHR96. All three genes are expressed throughout third-instarlarval and prepupal development. DHR38 is the Drosophilahomolog of NGFI-B and binds specifically to an NGFI-Bresponse element. DHR78 and DHR96 are orphan receptorgenes. DHR78 is induced by 20-hydroxyecdysone (20E) incultured larval organs, and its encoded protein binds to twoAGGTCA half-sites arranged as either direct or palindromicrepeats. DHR96 is also 20E-inducible, and its encoded proteinbinds selectively to the hsp27 20E response element. The 20Ereceptor can bind to each of the sequences recognized byDHR78 and DHR96, indicating that these proteins may com-pete with the receptor for binding to a common set of targetsequences.

Retinoid, vitamin D, steroid, and thyroid hormones are smallhydrophobic ligands that initiate a diverse array of develop-mental and metabolic responses. The receptors that mediatethese responses form the basis of the nuclear hormone recep-tor superfamily (see ref. 1 for a recent review). This family isdefined by a characteristic protein domain structure includinga conserved DNA-binding domain and a ligand-binding/dimerization domain. Members of this superfamily can bedivided into three classes on the basis of their ligand-bindingand DNA-binding properties. Steroid receptors, including theestrogen and glucocorticoid receptors, form homodimers thatbind to an inverted repeat of 6-bp consensus half-sites (1, 2).The second class includes the retinoic acid and retinoid Xreceptors (RAR and RXR) as well as receptors for thyroidhormone and vitamin D. These receptors can bind to directrepeats of AGGTCA half-sites as homodimers or het-erodimers (3). Members of the third and largest class arereferred to as orphan receptors because their potential ligandsare unknown. At least some of these receptors, includingRev-Erb and NGFI-B, can bind to a single AGGTCA half-site(4, 5). Although extensive studies have provided significantinsights into the mechanisms by which nuclear hormonereceptors regulate the transcription of target genes, we stillknow little about how these changes in gene expression resultin specific and diverse developmental responses.

In an effort to understand the molecular mechanisms ofsteroid hormone action, we have focused on the role of20-hydroxyecdysone (20E) in directing the metamorphosis ofthe fruit fly Drosophila melanogaster. A high-titer 20E pulse atthe end of third-instar larval development triggers pupariumformation, followed 10 hr later by a 20E pulse that triggershead eversion and the onset of pupal development (6, 7). The20E receptor is encoded by two members of the nuclear

hormone receptor superfamily, EcR (8) and usp (9-11). TheUsP protein is most closely related to the vertebrate RXRfamily and can heterodimerize with vertebrate thyroid andvitamin D receptors, as well as with EcR (12-15). The abilityof RXRs to function as promiscuous heterodimerization part-ners, combined with the sequence similarity of many receptorbinding sites, suggests that other members of the superfamilymay function in transducing 20E signals, either by interactingdirectly with EcR and/or Usp or by competing for receptorbinding sites (16).At present, 13 members of the nuclear hormone receptor

superfamily have been identified in Drosophila. In addition toEcR and usp, these include kni (17), knrl (18, 19), egon (19), tll(20), svp (21), dHNF4 (22), E75 (23), E78 (24), FTZ-Fl (25),DHR3 (26), and DHR39 (27, 28). Seven of these genes appearto contribute to the 20E regulatory hierarchies that direct theonset of metamorphosis-E75, E78, PFTZ-FJ, DHR3, DHR39,EcR, and usp (16, 29, 30). Our goal in this study was tocomplete our collection of Drosophila nuclear hormone re-ceptor superfamily members that are expressed during theearly stages of metamorphosis. With these genes in hand, wecan undertake a comprehensive screen for protein-proteininteractions and initiate genetic studies aimed at defining thefunctions of these family members during Drosophila devel-opment. *

MATERIALS AND METHODSConstruction of Receptor DNA-Binding-Domain Libraries.

cDNA templates were prepared from late larval and prepupalpoly(A)+RNA or from a high-titer lysate of the AHOR3 cDNAlibrary (identical to ANS reported in ref. 31). The primernucleotide sequences (5' to 3') were as follows: F3, ICITGYG-ARGGITGYAA; F4, CIGGIWWICAYTWYGG; F5, TGCII-IGTITGYGGIGA; R4, RCAYTTITKIRRICGRCA; R5,CATICCIACIKCIAIRCAYTT; R6, CATICCIACIKIIAIR-CAYTTIYKIARICGRCA; R8, YTCYTGIACIGCYTCIC-KYTTCATICC. All PCR amplifications were carried out in10-Al reaction mixtures containing 5 ,tg of bovine serumalbumin, 0.5% Ficoll, 10 mM tartrazine, 3 mM MgCl2, 0.5 mMeach dNTP, 0.5 ,uM each phosphorylated primer, 1 ng oftemplate, and 0.5 unit of Taq DNA polymerase (Promega) inan Idaho Technology (Idaho Falls) air thermocycler for 30cycles of 94°C for 0 sec, 40°C for 0 sec, and 74°C for 2 sec. Atleast 10 independent amplification reactions were performedwith each primer pair. The pooled PCR products were purifiedby agarose gel electrophoresis and inserted into theEcoRV siteof pBluescript (Stratagene).

Screening by Colony-Lift Hybridization. Twelve oligonucle-otides were synthesized, each derived from the central unique

Abbreviations: RAR, retinoic acid receptor; RXR, retinoid X recep-tor; 20E, 20-hydroxyecdysone; EcRE, ecdysone response element;NBRE, NGFI-B response element.*The sequences reported in this paper have been deposited in theGenBank data base (accession nos. U36762, U36791, and U36792).

10604

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement" inaccordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

June

10,

202

0

Proc. Natl. Acad. Sci. USA 92 (1995) 10605

region of the DNA-binding domains of the known Drosophilasuperfamily members (Fig. 1). Only dHNF4 was not repre-sented by an oligonucleotide probe. One oligonucleotide wasalso prepared from the sequence spanning the EcoRV site inthe pBluescript polylinker. The sequences of the oligonucle-otides were as follows: DHR3, TCAACTACCAGTGTCCGC;DHR39, AAAATCGCAAGAACTACGTG; E75, AAGATC-CAGTATCGCCCG; E78, GCAAATCGAATATCGCT-GTT; EcR, AGCGCCGTCTACTGCTG; egon, ATAGCG-GCGATTGCAGGA; FTZ-Fl, CAAGAAGGTCTACAC-CTGT; kni, ATCAGCACCATCAGCGAG; knrl, CTGTCG-TCGATATCCGAC; svp, AATCTAACTTACTCTTGCCG;tll, CCCGGCAGTATGTGTG; usp, GATCTCACATACG-CTTGCA; pBluescript, GGAATTCGATATCAAGCTTAT.The following oligonucleotides were prepared from DHR38,DHR78, and DHR96, after they were isolated: DHR38,CCAAGTATGTCTGCCTAG; DHR78, CTGGGCTAC-CAGTGTCG; DHR96, GCCAAGAAGCAGTTCACC. Col-onies containing fragments from genes of the nuclear hormonereceptor superfamily were lifted onto nitrocellulose. Thefilters were incubated in 5 x standard saline/phosphate/EDTA with 0.1% SDS and salmon sperm DNA at 100 ,ug/mlfor 1 hr at 51°C; then 15 ,uCi (555 kBq) of end-labeledoligonucleotide probe was added for an additional 18-hrincubation at 51°C. The filters were washed in 6x standardsaline/citrate twice for 15 min at 25°C, once for 3 min at 51°C,and once for 30 min at 25°C. DNA isolated from colonies thatfailed to hybridize to the mixed oligonucleotide probes wassequenced on an Applied Biosystems 373A automated se-quencing machine using a primer adjacent to the EcoRV siteof pBluescript.

Mobility-Shift Assays. The entire open reading frame ofeach gene was inserted into pET23a (Novagen) and eachprotein was purified to 50-80% homogeneity by means of itsC-terminal His6 tag. The sequences of the DR-3, DR-4, DR-5,TREpal, hsp27 ecdysone response element (EcRE), Rev-Erb,and FTZ-F1 oligonucleotides have been described (29). TheNGFI-B response element (NBRE) sequence is GGTTA-AAAGGTCACC (32). Approximately 100 ng of each proteinwas incubated on ice for 20 min in 20-,ul reaction mixturescontaining 20 mM Hepes, 100 mM KCl, 7.5% (vol/vol)glycerol, 2 mM dithiothreitol, 0.1% (vol/vol) Nonidet P-40,and 2 ,tg of poly(dI-dC) (Pharmacia). Each oligonucleotidewas end-labeled to a specific activity of 100 Ci/mmol, and 100fmol was added to each reaction mixture. Protein/DNA com-plexes were resolved by electrophoresis in 5% polyacrylamidegels run in 90 mM Tris phosphate/2 mM EDTA, pH 8.0, at40C.

UspSvp

dHNF4DHR3

DHR39E75E78EcR

FTZ-F1Tll

DHR38DHR78DHR96

FIG. 1. Alignment of Drosophila nuclear hormone receptor DNA-binding domains. An alignment of the DNA-binding domains of knownDrosophila nuclear hormone receptor superfamily members reveals two regions of conserved amino acids flanking a central unique region. Theconserved amino acids were used to design PCR primers for amplifying fragments of Drosophila receptors: F3, F4, F5, R4, R5, R6, and R8. Theunique region was used to design gene-specific oligonucleotide probes to eliminate previously identified family members from further study.

RESULTS AND DISCUSSIONA Rapid Screen to Identify New Members of the Nuclear

Hormone Receptor Superfamily. An alignment of the Dro-sophila nuclear hormone receptor DNA-binding domains re-veals a unique central region of 8-9 aa flanked by highlyconserved regions that each contain a Cys2Cys2 zinc finger(Fig. 1). We designed a set of degenerate PCR primers fromeach of the conserved flanking regions and used these inpairwise combinations to synthesize libraries of nuclear recep-tor gene fragments. The three forward PCR primers-F3, F4,and F5-each span DNA sequence encoding conserved aminoacids in the first zinc finger, and three of the four reverseprimers-R4, R5, and R6-span DNA sequence encodingconserved amino acids in the second zinc finger (Fig. 1). TheR8 reverse primer was designed to identify Drosophila RXRhomologs. This primer spans sequence immediately down-stream from the DNA-binding domain that is conservedamong members of the RXR family (33). Each oligonucleotideprimer contains three to nine deoxyinosine residues in orderto reduce degeneracy (no more than 128-fold), maximize therange of superfamily members that can be detected, andminimize the effects of alternative codon usage. Six pairwisecombinations of primers were selected to optimize the diver-sity of superfamily members that could be amplified: F3/R5,F3/R8, F4/R4, F4/R5, F5/R5, and F5/R6. cDNA preparedfrom staged late-third-instar larvae or prepupae was used as atemplate for PCR amplification, restricting our screen tosuperfamily members expressed during the onset of metamor-phosis. The PCR products were purified by gel electrophoresisand inserted into a plasmid vector for further analysis.A key step in our cloning strategy was to rapidly identify

sequences corresponding to known superfamily members andeliminate those fragments from further analysis. To achievethis goal, 12 gene-specific oligonucleotide probes were de-signed from the coding sequences for the central unique regionof each Drosophila DNA-binding domain (Fig. 1). Colony liftscontaining the cloned PCR fragments were hybridized with a

mixture of oligonucleotide probes, and those colonies thatfailed to hybridize were analyzed by DNA sequencing. A totalof 10,000 clones were screened. At least 10 independent PCRamplifications were performed with each of the six primerpairs, and between 1000 and 2000 clones derived from eachprimer pair were screened. Plasmid DNA was sequenced from423 clones that failed to hybridize to the gene-specific probes.The deduced amino acid sequences fell into five classes thatcontained key conserved amino acids characteristic of nuclearhormone receptors, as well as several classes that contained nosequence homology. Of the five classes that appeared tocorrespond to nuclear receptors, two were identical to DHR3

.00 R8'-- R5

F5N

F4 F3 ---- R4R6

CSICGDRASGKHYGVYSCEGCKGFFKRTVRKDLTYA-CRENR--NCIIDKRQRNRCQYCRYQKCLTCGMKREAVQEERQCVVCGDKSSGKHYGQFTCEGCKSFFKRSVRRNLTYS-CRGSR--NCPIDQHHRNQCQYCRLKKCLKMGMRREAVQRGRVCAICGDRATGKHYGASSCDGCKGFFRRSVRKNHQYT-CRFAR--NCVVDKDKRNQCRYCRLRKCFKAGMCKVCGDKSSGVHYGVITCEGCKGFFRRSQSSVVNYQ-CPRNK--QCVVDRVNRNRCQYCRLQKCLKLGMCPVCGDKVSGYHYGLLTCESCKGFFKRTVQNKKVYT-CVAER--SCHIDKTQRKRCPYCRFQKCLEVGMCRVCGDKASGFHYGVHSCEGCKGFFRRSIQQKIQYRPCTKNQ--QCSILRINRNRCQYCRLKKCIAVGMCKVCGDKASGYHYGVTSCEGCKGFFRRSIQKQIEYR-CLRDG--KCLVIRLNRNRCQYCRFKKCLSAGMCLVCGDRASGYHYNALTCEGCKGFFRRSVTKSAVYC-CKFGR--ACEMDMYMRRKCQECRLKKCLAVGMCPVCGDKVSGYHYGLLTCESCKGFFKRTVQNKKVYT-CVAER--SCHIDKTQRKRCPYCRFQKCLEVGMCKVCRDHSSGKHYGIYACDGCAGFFKRSIRRSRQYV-CKSQKQGLCVVDKTHRNQCRACRLRKCFEVGMCAVCGDTAACQHYGVRTCEGCKGFFKRTVQKGSKYV-CLADK--NCPVDKRRRNRCQFCRFQKCLVVGMCLVCGDRASGRHYGAISCEGCKGFFKRSIRKQLGYQ-CRGAM--NCEVTKHHRNRCQFCRLQKCLASGMCAVCGDKALGYNFNAVTCESCKAFFRRNALAKKQFT-CPFNQ--NCDITVVTRRFCQKCRLRKCLDIGM

unique region

Developmental Biology: Fisk and Thummel

Dow

nloa

ded

by g

uest

on

June

10,

202

0

10606 Developmental Biology: Fisk and Thummel

and E75, with the exception of a single-base-pair change in thecentral region where the gene-specific probes would hybridize.These changes most likely arose during PCR amplification andindicate that our hybridization conditions are sensitive enoughto discriminate sequences with single-base-pair resolution. Theother three classes corresponded to novel members of thenuclear hormone receptor superfamily, described in moredetail below. Because no new RXR family members wereidentified with the R8 primer (after screening 1831 colonies),it appears that usp is the only RXR gene expressed during theearly stages of metamorphosis.

Isolation, Cytogenetic Mapping, and DNA Sequence Anal-ysis of DHR38, DHR78, and DHR96 cDNAs. PCR fragmentscorresponding to each new superfamily member were used asprobes to isolate cDNA clones. Each gene was named on thebasis of its cytogenetic location in the polytene chromosomes:DHR38 maps to 38E1-2, DHR78 maps to 78D4-5, and DHR96maps to 96B12-14 (data not shown). Interestingly, none ofthese regions correspond to puffs in the salivary gland polytenechromosomes, suggesting that the superfamily members thatare represented by puff loci have been identified.The longest cDNA for DHR38 and full-length cDNAs for

DHR78 and DHR96 were sequenced on both strands [Gen-Bank accession nos. U36762 (DHR38), U36791 (DHR78), andU36792 (DHR96)]. Database searches revealed that DHR38shares extensive sequence homology with the rat proteinNGFI-B, an orphan nuclear hormone receptor, as well as itshuman homolog, NAK1 (Fig. 2A). The overall identity be-tween DHR38 and NGFI-B is 48%, while there is 64% identityin the 330-aa region extending from the DNA-binding domainthrough the C terminus of the DHR38 protein. This highdegree of sequence identity suggests that DHR38 is theDrosophila homolog of NGFI-B.

In contrast, DHR78 and DHR96 appear to encode novelorphan receptors. The DNA-binding domain of DHR78 is74% identical to the TR2 orphan receptor and 67% identicalto Usp (Fig. 2B). The DHR78 ligand-binding/dimerizationdomain (aa 397-601) is most similar to that of the vitamin Dreceptor, with 28% identity. The DNA-binding domain ofDHR96 is 64% identical to the human vitamin D receptor and52% identical to EcR (Fig. 2C). The DHR96 ligand-bindingdomain (aa 501-723) is most similar to that of thyroid hormonereceptor, with 23% identity.

ADHR38NGFI-BNAK1hRXRa

UspFTZ-F1xRARca

BDHR78TR2Usp

hRXRahCOUP-TF

EcRhRARa

Temporal Profiles of DHR38, DHR78, and DHR96 Tran-scription During the Onset of Drosophila Metamorphosis. Asa first step toward understanding the functions of DHR38,DHR78, and DHR96 during Drosophila development, we de-termined their temporal patterns of transcription during theonset of metamorphosis. Northern blots containing total RNAisolated from staged third-instar larvae and prepupae wereprobed to detect transcripts from each of the three genes (Fig.3).DHR38 encodes a predominant transcript of 1.9 kb, al-

though 3.6-, 4.0-, and 6.2-kb transcripts can be detected uponlonger exposure (Figs. 3 and 4 and data not shown). The 1.9-kbDHR38 mRNA is expressed throughout third-instar larval andprepupal development, with a reduction in levels in mid-third-instar larvae (Fig. 3). DHR78 encodes a single 2.3-kb transcriptand DHR96 encodes two transcripts, 2.8 and 0.6 kb in length.Both genes are expressed throughout third-instar larval andprepupal development, with distinct increases in abundance at106 hr after egg laying (Fig. 3). A subset of 20E-regulated genesare coordinately induced at these times (34), suggesting thatDHR78 and DHR96 may be 20E-inducible. Curiously, the600-bp DHR96 transcript, which can also be detected with aprobe derived from the DNA-binding domain (data notshown), is present at the beginning of the third larval instar andis downregulated 94 hr after egg laying. This response parallelsthe repression of theAdh larval promoter (34), which has beenproposed to be a 20E-regulated event (36).The temporal patterns of DHR38, DHR78, and DHR96

transcription most closely resemble those of the genes encod-ing the 20E receptor. EcR and usp mRNAs can be detectedthroughout third-instar larval and prepupal development (34,37). Temporal specificity is dictated by the pulses of 20E thatare required for receptor activation. The temporal profiles ofDHR38, DHR78, and DHR96 transcription are consistent withthe possibility that these genes, like EcR and usp, may encodeligand-dependent receptors.DHR78 and DHR96 Are Induced by 20E in Cultured Larval

Organs. To determine whether DHR38, DHR78, and DHR96are 20E-inducible, we analyzed RNA extracted from mass-isolated larval organs cultured for 0-8 hr with the hormone.Equal amounts of RNA from each time point were analyzedby Northern blot hybridization (Fig. 4). Hybridization to detectrp49 mRNA confirmed that equal amounts of RNA had been

10 20 30 40 50 60Ident ty88%88%64%64%61%59%

Ident i ty74%67%65%62%62%62%

10 2,0 30

DHR96 IdentityhVDR G R H*'MKRIAL- D R IKDN' HS*K 64%EcR L R Y L VTKSAV K GRA EMDMYM~ KS 52%

FTZ-Fl HYLLK VQNEIV--Y ESHgDKTQK PYF V 49%

hRARa F Ni 5 HYGV'S GG 'SIQKNMV--Y HRDK INK 'NR S FEV 49%hCOUP-TF VEii5 nKHYGQF ffG 3S ;'SVRRNLT-Y3A PDQHHN BY K KVE 49%

hT3Ra V HYRCI IG IQKNLHPTYS KYDSC VDK 'NQ L'FKF IAV 49%

FIG. 2. Alignments of DNA-binding-domain sequences. The DNA-binding-domain sequence of each gene was used to search the PIR/Swiss-Prot/GenBank databases. An alignment of each sequence with representative matches from the databases is presented. Shaded boxes indicateidentity with the new protein sequence, and the percent identity is shown to the right of each sequence. x, Xenopus; h, human; VDR, vitamin Dreceptor; T3R, triiodothyronine receptor.

Proc. Natl. Acad. Sci. USA 92 (1995)

r-

Dow

nloa

ded

by g

uest

on

June

10,

202

0

Proc. Natl. Acad. Sci. USA 92 (1995) 10607

MoltPuparium

Wandering formation Head eversion

Third larval instar + Prepupal i Pupal

11 1100c tCD 0 (OOtD0001

c"'a 0Mo N Ito Mo N tCsom g 8 ° -D CD N o D+ Ct'-t - N 0 M -4 -o -4 -i -O -4 -4 -4 -4 4 -4 co on -. - _. __

DHR38

DHR78

DHR96

FIG. 3. Temporal profiles of DHR38, DHR78, and DHR96 transcription during the onset of metamorphosis. Northern blots containing RNAsamples isolated from staged third-instar larvae and prepupae collected at 2-hr intervals were probed to detect DHR38, DHR78, and DHR96mRNAs. These blots have been used previously for detailed studies of 20E-regulated gene transcription (34). One set of blots was sequentiallystripped and hybridized with probes from each gene, in order to allow direct comparison of transcription patterns. The blots were also hybridizedto detect rp49 mRNA, as a control for equal loading (data not shown). Developmental times are shown at the top as hours after egg laying, forthird-instar larval development, and as hours after puparium formation, for prepupal and pupal development. Landmark 20E-triggereddevelopmental transitions are shown at the top.

loaded and transferred. Transcripts from all three receptorgenes can be detected at the 0-hr time point, consistent withtheir expression in late-third-instar larvae (Fig. 3). Two tran-scripts of 1.9 kb and 4.0 kb, present at low levels, are detectedby a DHR38 antisense RNA probe (Fig. 4). The 4.0-kbtranscript may be weakly induced by 20E, showing a peak at4-8 hr after 20E addition, although the levels of the 1.9-kbtranscript remain unchanged throughout the time course. Incontrast, DHR78 transcripts begin to accumulate to higherlevels after several hours of culture, peaking at 3-8 hr after20E addition, indicating that this gene is induced by 20E. Asexpected, the 600-bp DHR96 transcript is not detectable incultured late larval organs (Fig. 3). The overall increase in

Hours after 20E addition

0 .25 .5 1 1.5 2 3 4 6 8

ME.S.g_ ... .... ..

%: '~ l...

i~~~~~~~~~~~~~~~~~~~. ... ...ex

4.0 kb_

1.9 kb_'

2.3 kb 0

2.8 kb _-

DHR38

DHR78

DHR96

e~~~~~~~~~~~~~ r 49

FIG. 4. Time course of DHR38, DHR78, and DHR96 transcriptionin cultured larval organs treated with 20E. Mass-isolated late-third-instar larval organs were treated with 0.5 ,uM 20E for the times shown,as described (35). Equal amounts of total RNA isolated from each timepoint were fractionated by formaldehyde/agarose gel electrophoresis,transferred to a nylon membrane, and hybridized with probes to detectDHR38, DHR78, DHR96, and rp49 mRNAs. One Northern blot wassequentially stripped and hybridized with a probe from each gene, inorder to allow direct comparison of transcription patterns. Detectionof DHR38 transcripts required the use of an antisense RNA probe.

DHR96 transcript levels indicates that this gene is also 20E-inducible.DNA-Binding Properties of DHR38, DHR78, and DHR96

Proteins. In an initial effort to characterize the DNA-bindingactivities of the DHR38, DHR78, and DHR96 proteins, wetested their abilities to bind to eight oligonucleotides carryingdifferent arrangements of single or double half-sites (Fig. 5).DHR38 binds specifically to the NBRE, as expected for aDrosophila homolog of NGFI-B (Fig. 5). Mutagenesis of theNGFI-B DNA-binding domain identified amino acids down-stream from the zinc fingers that are required for specificbinding-site recognition (38). These residues are conserved inDHR38, consistent with its ability to bind to an NBRE. Theselectivity of this interaction argues that the two adenosinesupstream from the AGGTCA half-site are critical for DHR38binding, as has been shown for NGFI-B (32). Furthermore, theability of DHR38 to bind to a single AGGTCA half-sitesuggests that, like NGFI-B (5), this protein can bind DNA asa monomer.DHR78 binds to the DR-3, DR-4, DR-5, and TREpal

oligonucleotides, the only oligonucleotides containing twoAGGTCA half-sites (Fig. 5). This requirement for two half-sites, as well as the presence of at least three size classes ofprotein/DNA complexes, suggests that DHR78 binds DNA asa multimer. Interestingly, the hsp27 EcRE is the only oligo-nucleotide bound by DHR96 (Fig. 5). The EcRE consists of apalindromic arrangement of the imperfect half-sites AGtgCAand gGtTCA. DHR78 and DHR96 thus recognize distinctsequences that can also be bound by the EcR/Usp heterodimer(29). These distinct binding specificities are consistent with theP-box sequences of the DHR78 and DHR96 proteins. TheDHR78 P-box, EGCKG, like that of DHR38, directs bindingto an AGGTCA half-site sequence (1). In contrast, DHR96contains a unique P-box sequence not seen in any superfamilymember described to date-ESCKA. Determination of apreferred DHR96 binding site should allow us to determinethe significance of its ability to bind the hsp27 EcRE.

All of the tested oligonucleotides that are specifically boundby the EcR/Usp heterodimer are also bound by either DHR78or DHR96 (29). This may allow these proteins to modulate 20Eresponses by competing with the 20E receptor for binding sitesin target gene promoters. It will be of interest to determinewhether DHR78 or DHR96 can heterodimerize with EcR,

m M ..

ow"41".4"OI P.I.

Developmental Biology: Fisk and Thummel

Dow

nloa

ded

by g

uest

on

June

10,

202

0

10608 Developmental Biology: Fisk and Thummel

IC

4

DHR96

FIG. 5. DNA-binding specificities ofDHR38, DHR78, and DHR96proteins. Each protein was overproduced in Escherichia coli, purified,and tested for its ability to bind to eight oligonucleotides in electro-phoretic mobility-shift assays. The names of the oligonucleotides areshown at the top. In all cases, binding could be blocked by the additionof an excess of the appropriate unlabeled oligonucleotide as compet-itor (data not shown).

Usp, or any of the Drosophila orphan receptors. The multipleforms of DHR78 protein/DNA complexes detected by mobil-ity-shift assay suggest that this protein is capable of forminghigher-order binding complexes. The requirement for twohalf-sites for DHR78 and DHR96 binding is also consistentwith these proteins binding as dimers (Fig. 5). Biochemical andgenetic studies ofDHR38, DHR78, and DHR96 should providefurther insights into their possible regulatory functions andinteractions with other Drosophila superfamily members.These studies should also lead to a clearer understanding ofhow nuclear hormone receptor superfamily members mightfunction together to control the diverse developmental path-ways associated with insect metamorphosis.

We thank Tatyana Kozlova for providing information prior topublication, Igor Zhimulev for his assistance with the cytology of the38E region, Pat Hurban for the gift of the XHOR3 cDNA library, andmembers of the Thummel laboratory for critical comments on themanuscript. C.S.T. is an Associate Investigator of the Howard HughesMedical Institute.

1. Tsai, M.-J. & O'Malley, B. W. (1994) Annu. Rev. Biochem. 63,451-486.

2. Gronemeyer, H. (1992) FASEB J. 6, 2524-2529.3. Stunnenberg, H. G. (1993) BioEssays 15, 309-315.4. Harding, H. P. & Lazar, M. A. (1993) Mol. Cell. Biol. 13, 3113-

3121.5. Wilson, T. E., Fahrner, T. J. & Milbrandt, J. (1993) Mol. Cell.

Biol. 13, 5794-5804.6. Pak, M. D. & Gilbert, L. I. (1987) J. Liq. Chromatogr. 10,

2591-2611.7. Richards, G. (1981) Mol. Cell. Endocrinol. 21, 181-197.8. Koelle, M. R., Talbot, W. S., Segraves, W. A., Bender, M. T.,

Cherbas, P. & Hogness, D. S. (1991) Cell 67, 59-77.9. Henrich, V. C., Sliter, T. J., Lubahn, D. B., MacIntyre, A. &

Gilbert, L. I. (1990) Nucleic Acids Res. 18, 4143-4148.10. Shea, M. J., King, D. L., Conboy, M. J., Mariani, B. D. & Kafatos,

F. C. (1990) Genes Dev. 4, 1128-1140.11. Oro, A. E., McKeown, M. & Evans, R. M. (1990) Nature (Lon-

don) 347, 298-301.12. Yao, T., Segraves, W. A., Oro, A. E., McKeown, M. & Evans,

R. M. (1992) Cell 71, 63-72.13. Thomas, H. E., Stunnenberg, H. G. & Stewart, A. F. (1993)

Nature (London) 362, 471-475.14. Yao, T., Forman, B. M., Jiang, Z., Cherbas, L., Chen, J. D.,

McKeown, M., Cherbas, P. & Evans, R. M. (1993) Nature (Lon-don) 366, 476-479.

15. Koelle, M. R. (1992) Ph.D. thesis (Stanford Univ., Stanford,CA).

16. Richards, G. (1992) Curr. Bio. 2, 657-659.17. Nauber, U., Pankratz, M. J., Lienlin, A., Seifert, E., Klemm, U.

& Jackle, H. (1988) Nature (London) 336, 489-492.18. Oro, A. E., Ong, E. S., Margolis, J. S., Posakony, J. W., McKe-

own, M. & Evans, R. M. (1988) Nature (London) 336, 493-496.19. Rothe, M., Nauber, U. & Jackle, H. (1989) EMBO J. 8, 3087-

3094.20. Pignoni, F., Baldarelli, R. M., Steingrimsson, E., Diaz, R. J.,

Patapoutian, A., Merriam, J. R. & Lengyel, J. A. (1990) Cell 62,151-163.

21. Mlodzik, M., Hiromi, Y., Weber, U., Goodman, C. S. & Rubin,G. M. (1990) Cell 60, 211-224.

22. Zhong, W., Sladek, F. M. & Darnell, J. E., Jr. (1993)EMBO J. 12,537-544.

23. Segraves, W. A. & Hogness, D. S. (1990) Genes Dev. 4, 204-219.24. Stone, B. L. & Thummel, C. S. (1993) Cell 75, 307-320.25. Lavorgna, G., Ueda, H., Clos, J. & Wu, C. (1991) Science 252,

848-851.26. Koelle, M. R., Segraves, W. A. & Hogness, D. S. (1992) Proc.

Natl. Acad. Sci. USA 89, 6167-6171.27. Ohno, C. K. & Petkovich, M. (1992) Mech. Dev. 40, 13-24.28. Ayer, S., Walker, N., Mosammaparast, M., Nelson, J. P., Shilo, B.

& Benyajati, C. (1993) Nucleic Acids Res. 21, 1619-1627.29. Horner, M., Chen, T. & Thummel, C. S. (1995) Dev. Biol. 168,

490-502.30. Woodard, C. T., Baehrecke, E. H. & Thummel, C. S. (1994) Cell

79, 607-615.31. Hurban, P. & Thummel, C. S. (1993) Mol. Cell. Biol. 13, 7101-

7111.32. Wilson, T. E., Fahrner, T. J., Johnston, M. & Milbrant, J. (1991)

Science 252, 1296-1300.33. Lee, M. S., Kliewer, S. A., Provencal, J., Wright, P. E. & Evans,

R. M. (1993) Science 260, 1117-1121.34. Andres, A. J., Fletcher, J. C., Karim, F. D. & Thummel, C. S.

(1993) Dev. Biol. 160, 388-404.35. Thummel, C. S., Burtis, K. C. & Hogness, D. S. (1990) Cell 61,

101-111.36. Savakis, C., Ashburner, M. & Willis, J. H. (1986) Dev. Biol. 114,

194-207.37. Henrich, V. C., Szekely, A. A., Kim, S. J., Brown, N. E., An-

toniewski, C., Hayden, M. A., Lepesant, J.-A. & Gilbert, L. I.(1994) Dev. Biol. 165, 38-52.

38. Wilson, T. E., Paulsen, R. E., Padgett, K. A. & Milbrandt, J.(1992) Science 256, 107-110.

Proc. Natl. Acad. Sci. USA 92 (1995)

Dow

nloa

ded

by g

uest

on

June

10,

202

0