Access the full text.
Sign up today, get DeepDyve free for 14 days.
Loading next page...
/lp/unpaywall/nrf-1-an-activator-involved-in-nuclear-mitochondrial-interactions-T6Ad9KpSP4
References
References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.
- Publisher
- Unpaywall
- ISSN
- 0890-9369
- DOI
- 10.1101/gad.7.12a.2431
- Publisher site
- See Article on Publisher Site
Abstract
Downloaded from genesdev.cshlp.org on November 17, 2021 - Published by Cold Spring Harbor Laboratory Press NRF-1, an activator involved in nuclear- mitochondrial interactions, utilizes a new DNA-binding domain conserved in a family of developmental regulators Ching-man A. Virbasius, Joseph V. Virbasius, and Richard C. Scarpulla 1 Department of Cell, Molecular, and Structural Biology, Northwestern University Medical School, Chicago, Illinois 60611 USA Nuclear respiratory factor 1 (NRF-1) was first discovered as an activator of the cytochrome c gene and was subsequently found to play a broader role in nuclear-mitochondrial interactions. We have now cloned a HeLa cDNA encoding NRF-1 using degenerate oligomers derived from tryptic peptide sequences for PCR amplification. The cDNA-encoded protein was indistinguishable from the authentic HeLa cell factor on denaturing gels, displayed the expected NRF-1 DNA-binding specificity, and made the same guanine nucleotide contacts as HeLa NRF-1 on binding known NRF-1 recognition sites. Antiserum raised against the highly purified recombinant protein recognized the identical DNA-protein complex formed using either a crude nuclear fraction or nearly homogeneous HeLa NRF-1. Recombinant NRF-1 also activated transcription through specific sites from several NRF-l-responsive promoters, confirming both the transcriptional activity and specificity of the cDNA product. Portions of NRF-1 are closely related to sea urchin P3A2 and the erect wing (EWG) protein of Drosophila. Both are recently identified developmental regulatory factors. The region of highest sequence identity with P3A2 and EWG was in the amino-terminal half of the molecule, which was found by deletion mapping to contain the DNA-binding domain, whereas the carboxy-terminal half of NRF-1 was highly divergent from both proteins. The DNA-binding domain in these molecules is unrelated to motifs found commonly in DNA-binding proteins; thus, NRF-1, P3A2, and EWG represent the founding members of a new class of highly conserved sequence-specific regulatory factors. [Key Words: Oxidative phosphorylation; nuclear respiratory factors; mitochondria; transcription] Received July 27, 1993; revised version accepted October 1, 1993. The vertebrate mitochondrion contains its own genome, a ribonucleoprotein enzyme that is thought to cleave along with the machinery required for its autonomous light-strand transcripts to form primers for heavy-strand transcription, translation, and replication (Attardi and DNA replication, have been cloned (Chang and Clayton Schatz 1988; Clayton 1991; Wallace 1992). Mitochon- 1989; Topper and Clayton 1990). A second nuclear gene drial DNA, however, has a coding capacity limited to product linking transcription and replication is mito- only 13 respiratory chain polypeptides, and 2 ribosomal chondrial transcription factor 1 (mtTF1), now called mt- and 22 transfer RNAs. Most respiratory proteins and all TFA (Xu and Clayton 1992). This factor recognizes the of those necessary for maintenance and expression of the divergent heavy- and light-strand promoters to stimulate mitochondrial genome are encoded in the nuclear DNA. transcription initiation by mitochondrial RNA polymer- Thus, understanding the genetic control of mitochon- ase in vitro (Fisher et al. 1987). Because light-strand tran- drial function largely becomes a problem of identifying scripts cleaved by mitochondrial RNA processing (MRP) the nuclear genes involved and investigating potential endonuclease prime heavy-strand replication, mtTFA mechanisms of regulated expression. has the potential to modulate both transcription and rep- Investigations of nuclear-mitochondrial interactions lication of mitochondrial DNA (Clayton 1991). The im- in mammalian cells have led to the cloning of nuclear portance of mtTFA to mitochondrial function in vivo is gene products required for mitochondrial DNA tran- supported by the observation that a null mutation in the scription and replication (Clayton 1991). The human and yeast counterpart to mtTFA (ABF2) results in the loss of mouse genes for the RNA subunit of MRP endonuclease, mitochondrial DNA (Diffley and Stillman 1991). Inter- estingly, this phenotype can be rescued by expression of human mtTFA in yeast (Parisi et al. 1993). 1Corresponding author. Regulatory elements common to nuclear genes with GENES & DEVELOPMENT 7:2431-2445 © 1993 by Cold Spring Harbor Laboratory Press ISSN 0890-9369/93 $5.00 2431 Downloaded from genesdev.cshlp.org on November 17, 2021 - Published by Cold Spring Harbor Laboratory Press Virbasius et al. from HPLC chromatography were chosen for microse- products that function in the mitochondria have also been described (Nagley 1991; Wallace 1993). Analysis of quencing. NRF-1(72) yielded 26 residues (SMILEDLE- cytochrome c and cytochrome oxidase promoters has led SALAEHAPAPQEVNSELP) from a single homogenous to the identification of transcriptional activators desig- peptide, whereas NRF-l(38) was a mixture of two pep- tides with a novel 8-residue sequence (VSWTQALR) de- nated as nuclear respiratory factors (NRF)-I (Evans and rived from the secondary product. The major NRF-l(38) Scarpulla 1989, 1990; Chau et a1.1992) and -2 (Virbasius peptide was from Ku antigen, a known contaminant of and Scarpulla 1991; Virbasius et al. 1993). Functional DNA-binding proteins (Kadonaga 1991). A series of de- NRF-1 sites have been found in genes encoding cy- generate oligonucleotides was designed for PCR ampli- tochrome c and at least one subunit each of respiratory fication of total eDNA prepared from HeLa poly(A) + complexes III, W, and V (Evans and Scarpulla 1989, 1990; RNA. Two of these probes yielded a 269-nucleotide PCR Chau et al. 1992}, suggesting a role for the factor in the product containing an open reading frame that was sub- coordinate expression of respiratory chain subunits. sequently used as a probe to obtain a 3-kb HeLa cDNA NRF-1 may also participate in mitochondrial gene ex- clone. The putative NRF-1 cDNA had a 503-amino-acid pression through its sequence-specific activation of open reading flame containing the complete sequence genes encoding both the MRP RNA (Evans and Scarpulla of the PCR product flanked by the two NRF-1 peptides 1990) and mtTFA (this paper; Virbasius and Scarpulla 1994). Activity of the proximal mtTFA promoter is (Fig. 1}. The cDNA product was overexpressed in Escherichia highly dependent on NRF-1 in both transfected cells and coli using an inducible T7 expression system (Studier et in in vitro transcription assays (Virbasius and Scarpulla 1994). Similarly, expression of the gene encoding 5-ami- al. 1990). A transformant containing the NRF-l-coding region transcribed from an inducible T7 promoter gave a nolevulinate (5-ALA} synthase, the rate-limiting enzyme major 68-kD protein on induction (Fig. 2, lanes 2,3). The in the biosynthesis of heme for respiratory cytochromes, 68-kD mass of the induced protein is identical to that requires two NRF-1 recognition sites within its pro- moter region (Braidotti et al. 1993). These findings are determined previously for highly purified HeLa NRF-1 consistent with an integrative role for NRF-1 in control- (Chau et al. 1992) but is greater than the 54-kD mass ling nuclear-mitochondrial interactions in higher organ- predicted by the NRF-1 open reading frame. As only the isms. In addition, functional NRF-l-binding sites are open reading frame fragment was cloned into the expres- sion vector, the mass of NRF-1 appears to be overesti- present in the genes for tyrosine aminotransferase and mated on denaturing gels. The induced 68-kD protein the translation initiation factor eIF-2c~ (Chau et al. 1992). was purified to >95% purity from sonified extracts by Like 5-ALA synthase, these two proteins participate in ammonium sulfate precipitation (lane 4) and heparin- the rate-limiting steps of their respective pathways of agarose fractionation (lane 5). tyrosine catabolism and protein synthesis. Thus, NRF-1 To determine whether the eDNA-encoded recombi- may integrate a number of metabolic processes by regu- nant protein was the same as that present in DNA-pro- lating the genes encoding key enzymes. tein complexes formed using HeLa cell NRF-1, goat an- As a prelude to molecular cloning, NRF-l-binding ac- tiserum was raised against the purified recombinant pro- tivity was purified over 30,000-fold to near homogeneity tein and tested for its ability to "supershift" NRF-1- and was found to reside in a single polypeptide of 68 kD DNA complexes in a gel-retardation assay. DNA- (Chau et al. 1992). Here, we use the sequences of tryptic protein complexes of identical migration were formed peptides to obtain a cDNA clone that encodes a protein with a labeled rat cytochrome c NRF-1 oligomer (RC4 with the DNA-binding and transcriptional specificities - 172/- 147) using crude nuclear extract (Fig. 3, lane 1), expected for NRF-1. Deletion mapping of NRF-1 estab- affinity-purified NRF-1 (lane 4), or recombinant NRF-1 lishes that its DNA-binding domain coincides with a (lane 7). In each case, the complex was supershifted with region of high sequence similarity with P3A2 (Calzone et antiserum raised against recombinant NRF-1 (lanes al. 1991; Hoog et al. 1991) and erect wing (EWG)(Desi- 3,6,9) under conditions where preimmune serum had no mone and White 1993), two recently identified develop- effect (lanes 2,5,8). Variations in the antibody complexes mental regulatory factors. Thus, these proteins define a formed with affinity-purified and recombinant proteins new family of sequence-specific regulators that share a result from differences in the antigen-antibody ratios conserved DNA-binding domain. On the basis of func- present in each reaction. The multiple complexes ob- tional analysis of over a dozen NRF-1 sites, we predict served previously with binding of NRF-2 to its recogni- that the NRF-1 DNA-binding domain participates in the tion site in the cytochrome oxidase subunit Vb (MCO5b expression of >50 mammalian genes of known se- + 13/+33)gene (Virbasius et al. 1993)were unaffected quence. by the addition of anti-NRF-1 or preimmune serum to binding reactions (lanes 10-12). These results provide a direct link between the NRF-1 cDNA product and the Results NRF-l-binding activity present in HeLa cells. Molecular cloning and overexpression of recombinant NRF-1 DNA- bin cling specificity of recom bin an t NRF-1 Tryptic peptides were derived from -50 pmoles of puri- The recombinant protein should display the same DNA- fied HeLa NRF-1 (Chau et al. 19921. Two peptide peaks 2432 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on November 17, 2021 - Published by Cold Spring Harbor Laboratory Press Cloning of transcription factor NRF-1 AGCGCTGCCGCTTCTGAGT 20 TCCTGGTCAGAACTTTACACAGAAAAGGTGCTCAAAGGATATTTGTTTAATGAATGTGGTATGCTGACATTTAAAcAGGACAAAATTGTAGAACTTC i0 Met Glu Glu His Gly Val Thr Gln Thr Glu His Met Ala Thr Ile GIu Ala His AIa Val Ala Gin Gln Val Gln 119 ATG GAG GAA CAt GGA GTG ACC CAA ACC GAA CAT ATG GCT ACC ATA GAA GCA CAT GCA GTG GCC CAG CAA GTG CAG Gin Val His Val AIs Thr Tyr Thr Glu His Set Met Leu Set Ala Asp Glu Asp Set Pro Ser Set Pro Glu Asp 194 CAG GTC CAT GTG GCT ACT TAC ACC GAG CAT AGT ATG CTG AGT GCT GAT GAA GAC TCG CCT TCT TCG CCC GAG GAC 268 Thr Set Tyr Asp Asp Ser Asp Ile Leu ASh Set Thr Ala Ala Asp Glu Val Thr Ala His Leu Ala Ala Ala Gly 269 ACC TCT TAC GAT GAC TCA GAT ATA CTC AAC TCC ACA GCA GCT GAT GAG GTG ACA GCT CAT CTG GCA GCT GCA GGT 343 Pro Val Gly Met Ala AIs Ala Ala Ala Val Ala Thr Gly Lys Lys Arg LyB Arg Pro His Val Phe Glu Set ADn 344 CCT GTG GGA ATG GCC GCT GCT GCT GCT GTG GCA ACA GGA AAG AAA CGG AAA CGG CCT CAT GTA TTT GAG TCT PAT 418 Pro Set Ile Arg Lys Arg Gln Gln Thr Arg Leu Leu Arg Lys Leu Arg Ala Thr Leu Asp Glu Tyr Thr Thr Arg 419 CCA TCT ATC CGG AAG AGG CAA CAA ACA CGT TTG CTT CGG AAA CTT CGA GCC ACG TTA GAT GAA TAT ACT ACT CGT 493 Val Gly Gln Gln Ala Ile Val Leu Cys Ile Set Pro Ser Ly8 Pro Ash Pro Val Phe Lys Val Phe Gly Ala Ala 494 GTG GGA CAG CAA GCT ATT GTC CTC TGT ATC TCA CCC TCCAAA CCT AAC CCT GTC TTT AAA GTG TTT GGT GCA GCA 568 Pro Leu Glu Ash Val Val Arg Lys Tyr Lys Ser Met Ile Leu G~u ~s~ LeU Glu Set Kla Leu Ala Glu His A1s 569 CCT TTG GAG AAT GTG GTG CGT AAG TAC AAG AGC ATG ATC CTG GAA GAC CTG GAG TCT GCT CTG ~ 643 Pro Ala Pro Gln GI~ Val ASh Set Glu Leu Pro Pro Heu Thr Ile Asp Gly Ile Pro Val Ser Val Asp Lys Met 644 CCT GCG CCA CAG GAG GTT A~C T~A GAA CTG CCG 9CT CTC ACC ATC GAC GGA ATT CCA GTC TCT GTG GAC AAA ATe 718 Thr Gln Ala Gln Leu Arg Ala Phe Ile Pro Glu Met Leu Lys Tyr Ser Thr Gly Arg Gly Lys Pro Gly Trp Gly 719 ACC CAG GCC CAG CTT CGG GCA TTT ATC CCA GAG ATG CTC AAG TAC TCT ACA GGT CGG GGA AAA CCA GGC TGG GGr 793 Lys Glu Set Cys Lys Pro Ile Trp Trp Pro Glu Asp Ile Pro Trp Ala Asn Val Arg Ser Asp Val Arg Thr Glu 794 AAA GAA AGC TGC AAG CCC ATC TGG TGG CCT GAA GAT ATC CCC TGG GCA AAT GTC CGG AGT GAT GTC qqq hq A GAA 868 Glu Gln Lys Gln Arg Val Ser TrD Thr Gln Ala Le u Ara Thr Ile Val LyB ASh Cys Tyr LyB Gln Hil Gly Arg 869 GAG CAA RAG CAG AGG GTT ~C~ ~GG ACC CAG GCA CTA CGG ACC ATA GTT AAA AAC TGT TAT AAA CAG CAT GGG CGG 943 Glu Asp Leu Leu Tyr AIa Phe Glu Asp Gln Gin Thr Gln Thr Gln Ala Thr Ala Thr His Set Ile Ala His Leu 944 GRA GAC CTT TTG TAT GCC TTT GAA GAT CAG CAA ACG CAA ACA CAG GCC ACA GCC ACA CAT AGT ATA GCT CAT CTT 1018 Val Pro Set Gin Thr Val Val Gln Thr Phe Ser ASh Pro Asp Gly Thr Val Set Leu Ile Gln Val Gly Thr Gly Figure 1. Nucleotide and predicted 1019 GTA CCA TCA CAG ACT GTA GTC CAG ACT TTT AGT AAC CCT GAT GGC ACT GTC TCA CTT ATC CAG GTT GGT ACG GGG 1093 amino acid sequence of the HeLa NRF-1 Ala Thr Val Ala Thr Leu Ala Asp Ala Set Glu Leu Pro Thr Thr Val Thr Val Ala Gln Val Ash Tyr Set Ala 1094 GCA ACA GTA GCC ACA TTG GCT GAT GCT TCA GAA TTG CCA ACC ACG GTC ACC GTT GCC CAA GTG AAT TAT TCT GCC 1168 cDNA 5'-untranslated and coding region. Val Ala Asp Gly GIu Val Glu Gln Ash Trp Ala Thr Leu Gin Gly Gly Glu Met Thr Ile Gln Thr Thr Gin Ala Amino acid sequences matching those of 1169 GTG GCT GAT GGA GAG GTG GAA CAA AAT TGG GCC ACG TTA CAG GGA GGT GAG ATG ACC ATC CAG ACG ACG CAA GCA 1243 tryptic peptides from the purified NRF-1 Set Glu Ala Thr Gln Ala Val Ala Set Leu Ala Glu Ala Ala Val Ala Ala Ser Gin Glu Met Gln Gln Gly Ala 1244 TCA GAG GCC RCC CAG GCG GTG GCA TCG TTG GCA GAG GCC GCA GTG GCA GCT TCT CAG GAG ATG CAG CAG GGA GCT 1318 protein are underlined. The sequence co- Thr Val Thr Met Ala Leu ASh $er Glu Ala Ala Ala His Ala Val Ala Thr Leu Ala Glu Ala Thr Leu Gln Gly inciding with the PCR product amplified 1319 ACAGTC ACT ATGGCG CTT AAC AGC GAA GCT GCC GCC CAT GCT GTC GCC ACC CTG GCT GAG GCC ACC TTA CAA GGT 1393 from HeLa cDNA using primers derived Gly Gly Gln Ile Val Leu Set Gly Glu Thr Ala Ala Ala Val Gly Ala Leu Thr Gly Val Gln Asp Ala Ash Gly 1394 GGG GGA CAG ATC GTC TTG TCT GGG GAA ACC GCA GCA GCC GTC GGA GCA CTT ACT GGA GTC CAA GAT GCT AAT GGC 1468 from the NRF-1 peptide sequences is indi- Leu Val Gin Ile Pro Val Ser Met Tyr Gln Thr Val Val Thr Set Leu Ala Gln Gly Ash Gly Pro Val Gln Val 1469 CTG GTC CAG ATC CCT GTG AGC ATG TAC CAG ACT GTG GTG ACC cated with a bold underline. The complete AGC CTC GCC CAG GGC AAC GGA CCA GTG CAG GTG 1543 Ala Met Ala Pro Val ThE Thr Arg Ile Ser Asp Set Ala Val Thr Met Asp Gly Gin Ala Val GIu Val Val Thr 2970-nucleotide NRF-1 cDNA sequence 1544 GCC ATG GCC CCT GTG ACC ACC AGG ATA TCA GAC AGC GCA GTC ACC ATG GAC GGC CAA GCT GTG GAG GTG GTG ACA 1618 has been submitted to GenBank under ac- Leu Glu Gln End 1619 TTG GAA CAG TGA cession number L22454. binding properties ascribed previously to NRF-1 (Evans tested for their ability to stimulate the activity of a trun- and Scarpulla 1989, 1990; Chau et al. 1992). Recombi- cated cytochrome c promoter in transfected cells. The COXVb -109/-87 and mtTFA -73/-46 oligomers nant NRF-1 was thus used for competition DNase I foot- printing of the rat cytochrome c promoter region (Fig. 4). stimulated promoter activity 12.3-+3.4-fold and 6.0---1.9- In the absence of competitor, the recombinant protein yielded several intense enhanced cleavages at the 5' and 3' ends of the footprint, with the absence of cleavages ,-,= "E ® .E == throughout the intervening protected region (Fig. 4, lane O'~' ,P G 3). This pattern is identical to that observed previously STD - ~" E~ 0~o~ kD using preparations of HeLa NRF-1 (Evans and Scarpulla 1990). The footprint was eliminated by the inclusion of 200 an excess of unlabeled oligonucleotides of previously characterized NRF-1 sites from nuclear genes with prod- ucts that function in the mitochondria. These include cytochrome c (RC4, lane 4), cytochrome oxidase subunit 43 VIc (COXVIc, lane 5), and mouse MRP RNA (mMRP, lane 6). Moreover, a sequence from the cytochrome cl gene with two mismatches from the NRF-1 consensus did not compete (hCC1, lane 9), but a mutated derivative, 1 2 3 4 5 active in both NRF- 1 binding and transcriptional activity Figure 2. Expression and purification of recombinant NRF-1. {Evans and Scarpulla 1990), did (hCC1UP, lane 10). The Coomassie-blue stained SDS-PAGE of molecular mass stan- unrelated cytochrome c ATF/CREB site (RC4 -281/ dards (lane I} and 50 txl of log-phase culture of E. coli strain -256, lane 11) served as a negative control. BL21(DE3) transformed with the NRF-l-coding region in the In addition to the known NRF-1 sites, we observed pET3d expression vector uninduced (lane 2) or induced by the strong similarities to the NRF-1 consensus in recently addition of 0.4 mM IPTG (lane 3). A lysate of an induced culture isolated genes encoding cytochrome oxidase subunit Vb was purified by ammonium sulfate precipitation (lane 4) and (COXVb) (Basu and Avadhani 1991) and mtTFA (Tomi- fractionation of the pellet on a heparin-agarose column (lane 5}. naga et al. 1992). Oligomers of each of these sites were Lanes 4 and 5 contain equal amounts (4 ~g) of total protein. GENES & DEVELOPMENT 2433 Downloaded from genesdev.cshlp.org on November 17, 2021 - Published by Cold Spring Harbor Laboratory Press Virbasius et al. MCO5b ers are identical to those observed with the HeLa protein, RC4 -1721-147 +13/+33 providing further support for the conclusion that the iso- II I nuclear affinity recombinant nuclear lated cDNA encodes NRF-1. extract purified protein extract I ! ! I I a I . . + - + - . + - preimmune - + immune - - + - . + - + - . + Transcriptional activity and specificity of recombinant NRF-1 The binding of NRF-1 to its recognition site has been correlated with site-specific effects on promoter activity in transfected cells (Evans and Scarpulla 1989, 1990; Chau et al. 1992). The cloning of NRF-1 now affords an opportunity for a direct demonstration of its transcrip- tional activity and specificity. Promoter activation by NRF-1 was thus tested in an in vitro transcription assay using both a wild-type promoter from the rat cy- tochrome c gene (RC4CAT/-326} and the same tem- plate containing an insertional disruption of the NRF-1 :: ?;s site (RC4CAT/-326; LI- 162/- 159) that diminishes both NRF-1 binding and the activity of the transfected 1 2 3 4 5 6 7 8 9 10 11 12 Figure 3. Recognition of HeLa NRF-1 by antiserum directed against the recombinant protein. Binding reactions contained competitor 14 ~g of HeLa nuclear extract (lanes 1-3, 10-12), 12 ng of affin- ity-purified HeLa NRF-1 {lanes 4-6), or 20 ng of bacterial NRF-1 r heparin-agarose fraction (lanes 7-9). Labeled oligonucleotides .J.. , -.L' contained either an NRF- 1-binding site from the rat cytochrome c gene (RC4 - 172/- 147, lanes 1-9) or an NRF-2-binding site o -. i ' from the mouse COXVb gene (Virbasius et al. 1993) {lanes 10- .oo oo ~ ,=,== 12). Following the binding reaction, 1 ~1 of preimmune serum or goat antiserum was added to bacterially produced NRF-1, and the complexes were subsequently resolved on a native poly- acrylamide gel. fold, respectively, when cloned in cis, results similar to the value of 10.1 +3.1-fold obtained with the cytochrome c NRF-1 site (RC4 - 171/- 147). Both also formed spe- cific complexes with affinity-purified HeLa NRF-1 (not shown). The COXVb and mtTFA oligomers were also found to be specific competitors in the DNase I foot- printing assay using recombinant NRF-1 (Fig. 4, lanes 7,8). These results further substantiate the binding spec- ificity of the recombinant protein and indicate that COXVb and mtTFA genes are likely to have NRF-I-re- sponsive promoters. ww ~W :~ If the recombinant protein is NRF-1, it should contact DNA through specific guanine nucleotides spanning one tum of the DNA helix, as demonstrated using prepara- tions of the HeLa protein (Evans and Scarpulla 1990; Chau et al. 1992). Recombinant NRF-1 was thus used for 1 2 3 4 5 6 7 8 91011 methylation interference footprinting of known sites Figure 4. Binding of recombinant NRF-1 to the rat cytochrome from RC4, COXVIc, and MRP RNA genes. The pattem of c promoter region. An end-labeled RC4 promoter fragment con- guanine nucleotide contacts in each case was indistin- taining the NRF-l-binding site was subjected to DNase I diges- guishable from that obtained using HeLa NRF- 1 and con- tion following incubation in a mixture without added protein forms to the consensus derived previously (Fig. 5). The (lane 2) or with the addition of 20 ng of NRF-1 heparin-agarose RC4 site was known to deviate from the others by mak- fraction (lanes 2-11}. Competitor oligonucleotides indicated ing additional downstream contacts (Evans and Scarpulla above lanes 4-11 were added at a 200-fold molar excess before 1990). The same pattem is observed here with recombi- the addition of the labeled fragment. The extent of the NRF-1 nant NRF- 1. Therefore, the binding interactions between footprint is indicated by the vertical bar at right. {G) G reaction recombinant NRF-1 and cognate sites in several promot- of the labeled fragment. 2434 GENES & DEVELOPMENT ............. Downloaded from genesdev.cshlp.org on November 17, 2021 - Published by Cold Spring Harbor Laboratory Press Cloning of transcription factor NRF-1 mMRP COXVlc tivation of transcription by the recombinant protein (Fig. RC4 -3111-292 +20/+46 6A, lanes 2-4). Significant stimulation was observed us- -1721-147 ing 0.1 ~g of NRF-1, whereas 0.4 ~g was inhibitory. This ~ o. ~ ca. o inhibition likely results from the competitive displace- D. C3. 0 -~ ment of transcription complexes from the promoter tem- FB FB FB FB FB FB plate at high NRF-1 concentrations. In contrast, the linker insertion mutation resulted in a reduced level of transcription and completely eliminated activated ex- pression (lanes 5-7). The transcripts were initiated at the same position observed for the in vivo cytochrome c transcripts in liver RNA (lane 1), indicating that they accurately reflect promoter activation through the nor- mal initiation complex. The cytochrome c promoter has multiple cis-acting elements and therefore does not show complete depen- ~ o dence on NRF-1 for its activity (Evans and Scarpulla 1989). To enhance the NRF-l-dependent signal, four tan- dem sites from the cytochrome c (4XRC4) or the MRP < RC4CATI-326 tr RC4CAT/-326 LI -162/-159 =. • 0 0.1 0.4 0 0.1 0.4 pg NRF-1 RC4 -172 -147 I • • o o t TGCTAGCC CGCATGCGCGCGCACC TT ACGATCGG GCGTACGCGCGCGTGGAA 0 • • • • • COXVIc +46 +20 I • • I CTAGCAGCACGCATGCGCAGGAGC CGA GATC GTCGTGCGTACG CGTCCTCG GCT • • 0 • 1 2 3 4 5 6 7 mMRP RNA RC4CAT/-66 -311 -292 ,¢ z No 4XRC4 4XmMRP I • o o • I TAGTGCGCACGCGCAGGAG ¢: insert -172/-147 -3111-292 ATCACGCGTGCG CGTC CTC O • • 0 0.1 0.4 0 0.1 0.4 0 0.1 0.4 I.IgNRF-1 T T A Consensus: cGCGCAcGCGCG Figure 5. Recognition of NRF-l-binding sites by recombinant NRF-1 through characteristic guanine contacts. Fragments con- taining representative NRF-l-binding sites from the indicated promoters were labeled on upper or lower strands, partially methylated, and subjected to preparative scale mobility retar- 1 2 3 4 5 6 7 8 9 10 dation using recombinant NRF-1. Free DNA IF) and DNA iso- lated from bound complexes (B) were cleaved with piperidine, Figure 6. Transcriptional activation by recombinant NRF-1 and the products were analyzed on denaturing gels. (O) Guano- through specific NRF-1 recognition sites. (A) In vitro transcrip- sine bases that {when methylated) strongly inhibit NRF-1 bind- tion reactions with HeLa nuclear extract were gamed out with ing; (0) partial interference. Summarized below are the DNA 500 ng of a plasmid containing the RC4 promoter (lanes 2-4) or sequences of each site and the positions of guanine nucleotide a promoter with a linker insertion disrupting the NRF- 1-binding contacts compared with the consensus sequence and contacts site (lanes 5-7). Heparin-agarose-purified bacterial NRF-1 was derived from analysis of binding of HeLa NRF-1 to 10 known added as indicated (lanes 3,4,6,7). Transcription products were binding sites (Evans and Scarpulla 1990; Chau et al. 1992). analyzed by primer extension and compared with the primer extension product of 20 ~g of rat liver RNA {lane 1). (B) Products of in vitro transcription reactions using a truncated RC4 pro- moter construction RC4CAT/-66 (lanes 2-4) or the same pro- promoter (Evans and Scarpulla 1989). The results dem- moter with four tandem copies of the RC4 (lanes 5-7) or mMRP onstrate that a functional NRF-1 site is required for ac- (lanes 8-I0) NRF-1-binding sites cloned upstream. GENES & DEVELOPMENT 2435 Downloaded from genesdev.cshlp.org on November 17, 2021 - Published by Cold Spring Harbor Laboratory Press Virbasius et al. RNA (4XmMRP) promoters were cloned into an RC4 melanogaster participates in both nervous system and vector deleted of sequences upstream from - 66, and the flight muscle development (Desimone and White 1993). resulting constructs were used for in vitro transcription. Binding to DNA has been demonstrated for P3A2 but not Compared with the vector with no insert (Fig. 6B, lanes EWG, and neither has yet been shown to function di- 2--4), those with the RC4 (lanes 5-7) or mMRP (lanes rectly in transcriptional activation. 8-10) NRF-1 sites displayed a strong, dose-dependent in- Alignment of NRF-1 with P3A2 and EWG reveals a crease in transcription in response to added recombinant stretch of striking sequence conservation among all NRF-1. In this case, no inhibition was observed at 0.4 lag three proteins between NRF-1 residues 65 and 284 (Fig. of NRF-1 because of the increased binding capacity of the 7). This region coincides with the novel DNA-binding promoter template for NRF-1. As with the intact cy- domain identified previously for P3A2 (P3A2 residues tochrome c promoter, initiation occurred at the same 25-258) (Hoog et al. 1991) and corresponding here to site used by the liver initiation complex in the synthesis NRF-1 residues 61-290. In contrast, the three proteins of cytochrome c mRNA. These results establish that share little similarity in their carboxy-terminal halves or NRF-1 is a transcriptional activator that can function in an amino-terminal extension present in NRF-1 and both in the proper promoter context and in a minimal EWG. promoter to direct the synthesis of high levels of accu- To determine whether the highly conserved region co- rately initiated transcripts. incided with the NRF-1 DNA-binding domain, a dele- tion series of truncated NRF-1 molecules (summarized diagrammatically in Fig. 8A) was expressed by in vitro NRF-1 has a new DNA-binding domain conserved transcription and translation, and the products were as- in developmental regulatory factors sayed for binding to radiolabeled RC4 -172/- 147. As It was of interest to determine whether NRF-1 shares shown in Figure 8B, lane A, the intact eDNA yielded a structural features with other proteins. A computer translation product migrating at 68 kD. This protein was search revealed a region of extensive sequence similarity unaltered by deletion of the 3'-untranslated region to a with two recently described developmental regulatory position just downstream from the predicted NRF-1 factors (Fig. 7). The first, P3A2, has been implicated in translational terminator (lane B), confirming that the the correct expression of a cytoskeletal actin gene during translation product is derived from the NRF-1 open read- sea urchin development (Calzone et al. 1991; Hoog et al. ing frame. To demonstrate that the 68-kD translation 1991). The second, the EWG gene product of Drosophila product had the correct binding specificity, it was tested NRF-1 ........................................................... MEEHGVTQTEHMATI 15 P3A2 .......................................................................... EWG ATTSYRLWAPAGSQRSSTGNVVVTTTSSGSHSSNGANGGTGGTSAGSSTLGSGLNVTTITATSGGQLQSAGNT 75 NRF-1 EAHAVAQQVQQVHVATYTEHSMLSADEDSPSSPEDTSYDDSDILNSTAADEVTAHLAAAGPVGMAAAAAVATGKK 90 P3A2 .................... MMISEDISEPSSP.DTPFDDSDLLNSSMTDDVSAQLAASGPIGVRAAAAIATGKK 54 EWG SQSNGTTYKIEMLEEDIQSLGSDDDDEDLISSDGSLYEG..DLGSMPVNDDVAHQLAAAGPVGVAAAAAIASSKK 148 I II I0 IOIO IIIOIIOIO IIIIOIO II NRF-1 RKRPHVFESNPSIRKRQQTRLLRKLRATLDEYTTRVGQQAIVLCISPSKPNPVFKVFGAAPLENVVRKYKSMILE 165 P3A2 RKRPHSFETNPSIRRRQQTRLIRKLKATIDEYATRVGQQAVVLTCTPGKHDGNFKVFGAAPLENIMRNLKGIVLQ 129 EWG RKRPHCFETNPSVRKRQQNRLLRNVRAIIYEFTGRVGKQAVVLVATPGKPNTSYKVFGAKPLEDVLRNLKNIVMD 223 IIIII IIOIIIOIDIII IIOI 001 O IO0 III IIOOI OI I OIIIII III 0 I I ODD NRF-I DLESALAEHAPAPQEVNS...ELPPLTIDGIPVSVDKMTQAQLRAFIPEMLKYSTGRGKPGWGKESCKPIWWPED 237 P3A2 DLDNALAQRAPQPSNENSDLYELPPLVIDGIPTSVNKMTQAQLRAFIPLMLKYSTGRGKPGWGKESCRPVWWPSD 204 EWG ELDNALAQQAPPPPQDDPSLFELPGLVIDGIPTPVEKMTQAQLRAFIPLMLKYSTGRGKPGWGRESTRPPWWPKE 298 010 III II I I III IOIIIII I IIIIIIIIIIII IIIIIIIIIIIIIIOII OI IIl 0 NRF-I IPWANVRSDVRTEEQKQRVSWTQALRTIVKNCYKQHGREDLLYAF.EDQQTQTQATATHS ............. IA 298 P3A2 LPWANVRSDVRSEDEKRKVSWTHALVTIVINCYKHHGRDDLLPEFIEDKCKEIEASQNQ ......... VASLPTA 270 EWG LPWANVRMDARSEDDKQKISWTHALRKIVINCYKYHGREDLLPTFADDED.KVNALISQSGDEDEDMELSNPPTI 372 OIIIIII I IOIO I OOIII II II IIII IIIOIII I OI I HLVPSQTWQTFSNPDGTVSLIQVGTGATVATLADASELPTTVTVAQ ............................ 345 NRF-1 P3A2 TLLPSHAVVHTINNPDGTVSLIQVDTGATVATLAD ........................................ 305 EWG HTVTTMTPPTGNSNQPQQVNVVKINSAGTVITTHTAQSNTPAPTIIQSTNNQHVTTTATLPASTKIEICQAPAQN 447 O 0 0 I I O0 0 O II I Figure 7. Alignment of NRF- 1 protein se- NRF-1 .............. VNYSAVADGEVEQNWATLQGGEMT...IQTTQAS...EATQAVA ......... SLAEAAVA 391 quence with those of developmental regu- P3A2 ................... VTQVQQLTNLQTLQQVRLQPLQIQHALGNQQAEATQAVQ ......... TLAEVAAA 352 latory factors P3A2 and EWG. Sequences EWG QQHHQHHQTHLPNAVHIQPVAGGQPQTIQLTTASGTATATAVQTTAAA..VSAAQAHAHSQSQAHSQSSANQTVT 520 I0 I Ol 0 I011 0 1 O 0 of human NRF-1, sea urchin P3A2, and ASQ ...................... EMQQGATVTMALNSEAAAHAVATLAEATLQ..GGGQIVLSGETAAAVGAL 442 Drosophila EWG were aligned using the Nat-1 P3A2 QGG .................... DGELTEGQTVT ............. TLPEGT ....... QLVLASD ..... GSL 382 GAP program of the Genetics Computer EWG AQQIANAQVCIEPITLSDVDYTTQTVLSQNADGTVSLIQVDPNNPIITLPDGTTAQVQGVATLHQGEGGATIQTV 595 0 I II O I 0 [] DO Group (program manual, v.7, 1991J. (I} Residues identical in all three proteins~ ([31 Nay-1 TGVQDANG ................................ LVQIPVSMYQTVVTSLA..QGNGPVQVAMAPVTTR 483 QAINDGTAQG ............................... IVIPASVYQTVVAG ..... DGQPIQIANVNIAQQ 421 positions where all three proteins contain van2 r.WG QSLTDVNGHENMTVDLTETQDGQIYITTEDGQGYPVSVSNVISVPVSMYQSVMANVQQIQTNSDGTVCLAPMQVE 670 [] I [] DI IDIIDIDD [] [] similar (conservative} amino acid substitu- tions grouped as follows: (A S T); (D E); (N Nat-1 ISD ................. SAVTMDGQAV..EWTLEQ ........................... 503 SGG .......... GTTMAAIKNAVMQSQPIPSQVATLVVNAASHDQHT .................. 459 Q); (R K); (I L M V); (F Y W). Dots denote P3a2 EWG NGDQLETITMSPGMHQMMIQGGPGQEPQLV..QVVSLKDATLLSKAMEAINSGNVKSEDTIIMEQ 733 gaps introduced for optimum alignment. I D I~DI 2436 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on November 17, 2021 - Published by Cold Spring Harbor Laboratory Press Cloning of transcription |actor NRF-1 DNA NRF.1 503AA Binding [ w/////////////////////~ = "I ~ i RC4 RC4 COXVIc hQP 5O3 =. -172/-147 -172/-147 +20/+46 -66/-41 5o3 g &476-503 RC4 -172/-147 " " " + " " + " " + " " + " &419-503 + RC4 -281/-256 .... + " " + " " + " " + A331-503 &305-503 A264-503 . &238-503 41-77 J a1-109 K + &1-144 L kD A B C D E F G H kD B I J K L 218-- 218 100-- 1 2 3 4 5 6 7 8 1011121314 9 72-- 100 43-- 72 TAT-1090 oIF-2o yATPS COXVb 29-- /-1069 -421-21 +1/+20 -1091-87 RC4-172/-147 - + " " + " " + " " + " C ABCD E FGH B I J K L RC4-281/-256 - - + - - + " " + " " + Figure 8. Deletion mapping of the NRF-1 DNA-binding do- 1516 17 18 19 2021 22 23242526 main. (A) Schematic representation of in vitro-translated pro- teins tested for DNA binding. The shaded box represents the mtTFA mMRP hCC1UP hCC1 RC4 outer limits of the region required for DNA-binding activity. -781-49 -311/-292 -4541-431-4541-431-281/-256 Construct A includes the complete cDNA sequence. In con- RC4-1721-147 - + " " + " " + " " + " " + " struct B all but 35 nucleotides of the 3'-untranslated region is RC4 -2811-256 - - + - " + - - + - - + - - + deleted. C-H represent carboxy-terminal deletions of the resi- dues indicated and were generated by either cleavage at a re- striction endonuclease site in the native sequence (C,D,E,H) or insertion of a synthetic translation terminator (F,G). In I the entire 5'-untranslated region, which includes several potential upstream initiators, was deleted, and J,K and L represent amino- terminal deletions of the indicated residues and restoration of the initiator ATG by the addition of an NcoI linker. Activity of the proteins in DNA-binding assays is summarized by + or - (right). (B) SDS-PAGE showing [3SS]methionine-labeled transla- tion products corresponding to the constructions diagrammed 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 in A. Positions of molecular mass standards are at the left of each panel. (C) Electrophoretic mobility shift assay for binding Figure 9. Specific binding of in vitro-transcribed and -trans- activities of deletion mutants. Translation extracts containing lated NRF-1 to known NRF-1 recognition sites. Binding reac- the proteins diagrammed in A were incubated with end-labeled tions contained 5 ng of affinity-purified HeLa NRF-1 (lanes 3-5) RC4 -172/- 147 oligonucleotide, and the products were re- or 2 ~1 of wheat germ lysate without added RNA (lane 1 ), with solved on native acrylamide gels. RNA transcribed in vitro from an antisense NRF-1 template in pSG5 (lane 2), or with RNA transcribed from the NRF-1 se- quence in the sense orientation (lanes 6-41 ). End-labeled oligo- nucleotides (10 fmole/lane) contained the NRF-l-binding sites for binding to known NRF-1-binding sites from nine dif- designated above each panel. Lanes 1 and 2 also contained the ferent genes (Fig. 9). For each site, the major DNA-pro- labeled RC4 -172/-147 oligonucleotide. Binding reactions tein complex comigrated with the complex formed with were carried out in the presence (+) or absence (-) of a Z00-fold affinity-purified NRF-1 (lanes 3-5). Minor discrepancies excess of unlabeled specific competitor (RC4 - 172/- 147) or a result from differences in lengths of the labeled oligo- negative control oligonucleotide containing the rat cytochrome mers. All NRF-1 complexes were competitively dis- c ATF / CREB site (RC4 - 281 / - 256). DNA-protein complexes placed by an excess of unlabeled RC4 - 172/- 147 but were resolved on 5% acrylamide native gels. GENES & DEVELOPMENT 2437 Downloaded from genesdev.cshlp.org on November 17, 2021 - Published by Cold Spring Harbor Laboratory Press Virbasius et al. not by an excess of RC4 -281/- 256 ATF/CREB oligo- or amino termini, and the mass of HeLa NRF-1 was es- mer. The faster migrating complex appears to result from timated at -52 kD by glycerol gradient centrifugation in a truncated product formed during in vitro synthesis, both the presence and absence of its binding site (C. Vir- as indicated by the elimination of the lower complex basius, unpubl.). Finally, in vitro transcription experi- by carboxy-terminal deletion to residue 305 (Fig. 8C, ments unequivocally establish that the recombinant pro- lane F). tein has the transcriptional activity and specificity ex- Having confirmed the correct binding specificity, the pected for NRF-1. Taken together, the results presented carboxy-terminal deletions (Fig. 8A, constructs A-H) here allow us to conclude with reasonable certainty that were expressed (Fig. 8B, lanes A-H) and assayed for the cloned cDNA encodes NRF-1. However, given the DNA-binding activity (Fig. 8C, lanes A-H). DNA bind- existence of families of related transcription factors, it ing was unaffected until residues between 305 and 264 still remains a formal possibility that the true biological were removed (lanes F,G). Likewise, when the amino- activity results from a protein that has escaped our de- terminal deletions (Fig. 8A, constructs I-L) were ex- tection. pressed {Fig. 8B, lanes I-L) and assayed (Fig. 8C, lanes I-L), binding was lost on removal of residues 109-144 Conservation of the NRF-1 DNA-binding domain (lanes K,L). The precise deletion of the 5'-untranslated in P3A2 and E WG region in construct I removes several potential initiation codons without affecting the translated product, further The striking conservation of the NRF-1 DNA-binding confirming the identity of the NRF-1 reading frame. The domain in P3A2 and EWG suggests that these proteins carboxy-terminal boundary of the DNA-binding domain, constitute a new family of regulatory factors with di- determined here between NRF-1 residues 264 and 305, verse functions in eukaryotic development. The P3A2 compares favorably with that determined previously for DNA-binding domain coincides with that defined here P3A2 between NRF-1 residues 255 and 290. The amino- for NRF-1 and with the region of highest sequence con- terminal boundary determined here between residues servation among all three factors. Although NRF-1 and 109 and 144 is somewhat downstream from the P3A2 P3A2 are acidic proteins with predicted isoelectric boundary between NRF-1 residues 61 and 126 but is points of 4.71 and 5.49, respectively, nearly all of the overlapping in the region between residues 109 and 126. lysine and arginine residues [33/34 for NRF-1 and 34/37 These results establish that the major region of sequence for P3A2) are clustered into two sequence blocks within similarity among these proteins resides in their DNA- the most highly conserved regions of the DNA-binding binding domains. Thus, NRF-1, P3A2, and EWG define a domain. The sequence between NRF-1 residues 89 and new family of regulatory factors that share a highly con- 160 {Fig. 7) is 25% lysine plus arginine and has 85% served DNA-binding motif. sequence conservation {identical plus similar residues} with P3A2. Likewise, the NRF-1 sequence between res- idues 199 and 274 is 20% lysine plus arginine and has Discussion 91% sequence conservation with P3A2. In keeping with this structural conservation, the P3A2 Identification of the cDNA-encoded product as NRF-1 recognition sites strongly resemble those for NRF-1. Purification and molecular cloning of NRF-1 were un- Both proteins make major groove contacts through alter- dertaken as a requisite for further investigating its struc- nating GC base pairs, and high-affinity binding occurs tural characteristics and biological functions. Previ- through a tandem repeat of the T/CGCGCA motif ously, we had purified NRF-1 >30,O00-fold from HeLa {Evans and Scarpulla 1990; Calzone et al. 1991). An ap- nuclear extracts and demonstrated that a single 68-kD parent difference is that P3A2 can bind a monomer of polypeptide accounted for specific binding to the known this sequence at reduced affinity, whereas stable binding NRF-1 sites (Chau et al. 1992). The tryptic peptide se- of NRF-1 requires a tandem direct repeat of this half-site quences described here were derived from -50 pmoles of {Table 1). No NRF-1 binding was detected to sequences the protein purified from >200 liters of HeLa cells. from the cytochrome Cl (hCC1 -454/-431) and COX- The evidence presented here is consistent with the VIc (COXVIc - 46/- 20) genes containing perfect NRF-1 isolated cDNA encoding NRF-1. Both peptides derived half-sites {Evans and Scarpulla 1990]. Two nucleotide from the purified protein were encoded in the cDNA, the changes in the hCC~ site that restore the direct repeat expressed product of which migrated identically to HeLa {hCC~UP - 454/- 431) also restore binding by NRF-1. NRF-1 on denaturing gels. Interestingly, the masses of This is confirmed here for these hCCI sites using the NRF-1, P3A2, and EWG were all overestimated on dena- recombinant protein (Figs. 4 and 9), making it unlikely turing gels by 30-50%, possibly reflecting shared struc- that NRF-1 would bind with high affinity to several of tural features. Recombinant NRF-1 also binds specifi- the P3A2 target sites. Also, in the highest affinity P3A2- cally to the known NRF-1 sites through characteristic binding sites, the half-site motifs are separated by inter- guanine contacts encompassing a single helical turn. Al- vening nucleotides and, in one case, are rotated by one- though the binding site is palindromic, the protein ap- half helical turn {Calzone et al. 1991). These features pears to bind as a monomer. Heterodimeric DNA-pro- have not been observed in the known NRF-1 recognition tein complexes were not detected when intact NRF-1 sites. It should be noted that P3A2 has been proposed to was mixed with derivatives truncated at either carboxyl be a negative regulator of transcription through its dis- 2438 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on November 17, 2021 - Published by Cold Spring Harbor Laboratory Press Cloning of transcription factor NRF-1 Table 1. NRF-l-binding sites Gene name a Sequence Location b I. Functional binding sites - 166/- 155 rat cytochrome c ~ CCCGCATGCGCG - 169/- 158 CCAGCATGCGCG human cytochrome c 2 + 39/+ 28 rat cytochrome c oxidase subunit VIc 2 CACGCATGCGCA - 92/- 103 CGCACATGCGCA mouse cytochrome c oxidase subunit Vb 3 - 11/-22 TGCGCAAGCGCC bovine cytochrome c oxidase subunit VIIa 4 TGCGCAGGCGCA - 53/- 64 human ubiquinone binding protein 2 CGCGCACGCGCG +4/+ 15 bovine ATP synthase 7-subunit s TGCGCACGCGCA - 308/- 297 mouse MRP RNA 2 CGCGCACGCGCA - 293/- 282 human MRP RNA 2 GGCGCAGGCGCG - 59/- 7O human mitochondrial transcription factor A - 77/- 88 rat 5-aminolevulinate synthase 6 CGCGCACGCGCA - 59/- 48 AGCGCATGCGCA - 1085/- 1074 rat tyrosine aminotransferase s TGCACATGCGCA - 37/- 26 human eukaryotic initiation factor 2 TCCGCATGCGCG a-subunit s Consensus PyGCGCAPyGCGCPu II. Potential binding sites identified in GenBank/EMBL data base c A. Metabolic enzymes -71/-61 human arylsulfatase A 7 CGAGCACGCGCA - 149/- 138 TCCGCATGCGCA human branched chain c~-keto acid dehydrogenase 8 human carbonyl reductase 9 CGCGCAGGCGCA - 46/- 57 human protein disulfide isomerase/prolyl 4-hydroxylase Bto CGCGCACGCGCC - 79/- 90 human a-enolase 1 CGGGCAGGCGCA - 734/- 723 CGAGCATGCGCA - 227/- 216 rat glutamate dehydrogenase ~2 TGCGCACGCGCG - 269/- 280 human steroid 5-a-reductase ~3 TGCGCACGCGCA - 266/- 277" mouse ornithine decarboxylase 14 CGCGCACGCGCA - 333/- 344" human ornithine decarboxylase ~s CGCGCACGCGCA - 268/- 279" rat omithine decarboxylase 16 CGCGCAAGCGCG -610/-621 rat fatty acid synthase *z CGCGCACGCGCG - 562/- 549" rat Na+/K + ATPase e~-I subunit 18 - 18/-7 human calcium-activated neutral protease ~9 TGCGCATGCGCA - 175/- 186 human cathepsin D 2° GGCGCACGCGCA B. Signal transduction - 981/- 992 mouse GM-CSF 2t CACGCACGCGCG - 1014/- 1003" mouse hepatocyte growth factor-like protein 22 CGCGCACGCGCA - 1066/- 1077 rat dopamine D 1 receptor 23 CACGCACGCGCA human insulin receptor 24 GGCGCACGCGCG - 1036/- 1025 human insulin-like growth factor receptor 2s CGCGCACGCGCC - 29/- 40 human interferon receptor 26 CGCGCACGCGCC - 85/- 96 TGCGCACGCGCT +i/-ii human cyclophilin 27 CGCGCACGCGCG - 173/- 162" human lipoprotein receptor-like protein 28 GGCGCATGCGCA - 151/- 140 rat calmodulin II129 CGCGCACGCGCG - 209/- 220" CGCGCACGCGCA - 197/- 208" GGCGCAGGCGCA - 53/- 42 human calretinin 3° human Go-or 31 AGCGCACGCGCG - 835/- 846 GGCGCAGGCGCA -215/-204 human ADP ribosylation factor 13~ CGCGCAGGCGCA - 532/- 543 human protein phosphatase 2A 0~ 33 TGCGCACGCGCC - 79/- 90 mouse cyclic nucleotide phosphodiesterase 34 Chromosome maintenance and nucleic acid metabolism human DNA polymerase ~3s TGCGCAAGCGCA - 235/- 246 - 19/-8 human topoisomerase 136 CGCGCAGGCGCA human H1 RNA 37 GGCGCACGCGCG - 145/- 156 - 558/- 569" human hnRNP core protein A138 TGCGCAGGCGCA -521/-510 TACGCATGCGCA TGCGCAGGCGCA -398/-401" TGCGCAGGCGCA - 35/- 46" mouse S 16 ribosomal protein 39 CGAGCACGCGCG -210/199 mouse histone H2a.24° CGCACACGCGCA -40/- 51 mouse histone H3 4~ (Table 1 continued on following page.} GENES & DEVELOPMENT 2439 Downloaded from genesdev.cshlp.org on November 17, 2021 - Published by Cold Spring Harbor Laboratory Press Virbasius et al. Table 1. NRF-I-binding sites (Continuted) Gene name a Sequence Location b human RCC142 TGCGCACGCGCA - 78/- 67" human cdc2 43 CCAGCATGCGCA - 717/- 706 D. Other human bcl-244 CACGCACGCGCA - 889/- 900 human GADD153 growth/DNA damage CGCGCACGCGCA - 175/- 186" inducible gene 4s hamster GADD 153 46 CGCGCACGCGCA -365/-354* human 86-kD heat shock protein 47 CGCGCAGGCGCA - 286/- 257 human synapsin p8 TGCGCACGCGCC -241/-252 rat synapsin 148 TGCGCACGCGCC - 240/- 251 mouse myb proto-oncogene 49 GGCGCACGCGCC - 295/- 306 aThe search was confined to primate and rodent genes. For part I, references are given for the demonstration of NRF-1 binding. For part II, publications of the gene sequence are cited. bCoordinates are given relative to the first transcription start site, if known. Otherwise the authors' numbering system is followed. In some cases, the reverse complement of the published sequence is given to conform to the consensus. Sites identical to those of known function are indicated by an asterisk (*). CRodent and primate sequences in the GenBank (release 76) and EMBL data bases (release 34) were searched with the Findpatterns program (Genetics Computer Group Manual, 1991), allowing one mismatch to the consensus given. Only mismatches found in the known sites in part I were allowed in a further screen of the identified sequences. Furthermore, only sites in upstream regions of published genomic sequences were included in the table. References: l{Evans and Scarpulla 1989), 2(Evans and Scarpulla 1990), 3(this work), 4(unpubl.), 5(Chau et al. 1992), 6(Braidotti et al. 1993), Z(Kreysing et al. 1990), 8(Mitsubuchi et al. 1991), 9(Forrest et al. 1991), t°(Tasanen et al. 1992), ~(Giallongo et al. 1990), ~2(Das et al. 1993), la(Jenkins et al. 1991), 14(Katz and Kahana 1988), lS(Moshier et al. 1992), 16(Wen et al. 1989), lZ(Amy et al. 1990), 18(Yagawa et al. 1990), W(Miyake et al. 1986), 2°(Cavailles et al. 1993), 21(Miyatake et al. 1985), 22(Degen et al. 1991), 2a{Zhou et al. 1992), ~4(Tewari et al. 1989), 2S(Mamula and Goldfine 1992), 26(Lutfalla et al. 1992), 27(Haendler and Hofer 1990), 28{Kutt et al. 1989), 29(Nojima 1989), S°(Parmentier and Lefort 1991), al(Tsukamoto et al. 1991), a2(Lee et al. 1992), a3(Khew-Goodall et al. 1991), 34(Kurihara et al. 1990), aS(Pearson et al. 1991), 36{Kunze et al. 1991), 37(Baer et al. 1990), aa{Biamonti et al. 1989), 39(Wagner and Perry 1985), 4°{Hurt et al. 1989), 4~(Sittman et al. 1983), 42(Furuno et al. 1991), 43(Ku et al. 1993), 44{Adachi and Tsujimoto 1990), 45(Park et al. 1992), 46(Luethy et al. 1990), 47(Walter et al. 1989), 48(Sauerwald et al. 1990), 49(Bender and Kuehl 1986). placement of a zinc finger protein that binds the same transport and oxidative phosphorylation is unique in sequence (Hoog et al. 1991). In contrast, NRF-1 clearly that both nuclear and mitochondrial genomes contribute functions as a positive activator of transcription. Thus, it protein subunits (Attardi and Schatz 1988; Clayton remains to be determined whether the structural conser- 1991; Wallace 1992). The sole purpose of the mitochon- vation between P3A2 and NRF-1 in their DNA-binding drial genetic system is to complement the contribution domains will be precisely reflected in their binding and of nuclear genes in maintaining respiratory function. transcriptional specificities. Such interplay between the two genomes might involve The EWG protein is required for viability of Droso- novel pathways of intracellular communication. phila embryos and for the proper development of the One possibility is that NRF-1 may help to coordinate embryonic nervous system (Desimone and White 1993). the expression of respiratory chain subunits with com- Its molecular mass of 116 kD on denaturing gels is ponents of the mitochondrial transcription and replica- greater than that observed for P3A2 (62 kD) and NRF-1 tion machinery. Such a model is consistent with the (68 kD) and exceeds the mass predicted by its amino acid finding of functional NRF-1 recognition sites in genes sequence (77 kD]. Like P3A2, the sequence conservation encoding respiratory subunits, the MRP RNA [Evans and with EWG is largely confined to the NRF-1 DNA-bind- Scarpulla 1990; Chau et al. 19921 and mtTFA. The latter ing domain. Although its nuclear localization and struc- two have the capability of communicating changes in tural conservation with P3A2 and NRF-1 are consistent nuclear gene expression to the mitochondria through with a function in gene regulation, there is as yet no their effects on mitochondrial DNA replication and tran- evidence for DNA-binding or transcriptional effects, nor scription. In keeping with this hypothesis, we have re- have potential target genes been identified. cently established that the proximal promoter for the mtTFA gene is almost completely dependent on a NRF-1 recognition site for its activity both in transfected cells and in an in vitro transcription assay using recombinant NRF-1 and the nuclear control of mitochondrial NRF-1 (Virbasius and Scarpulla 1994). NRF-1 control function over mitochondrial function is also supported by the re- The mitochondrion serves to compartmentalize diverse cent observation that the activity of 5-ALA synthase cellular metabolic systems largely regulated by enzymes gene promoter is highly dependent on tandem NRF-1 encoded in the nuclear DNA. The apparatus for electron recognition sites [Braidotti et al. 1993). Thus, NRF-1 con- 2440 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on November 17, 2021 - Published by Cold Spring Harbor Laboratory Press Cloning of transcription factor NRF-1 trol of this nuclear gene may serve to regulate the levels sis and branched-chain amino acid catabolism, respec- of the heine cofactor required by the respiratory cy- tivelyl. This is also true of tyrosine aminotransferase, tochromes encoded by both genomes. These observa- 5-ALA synthase, and eIF-2a, supporting a role for NRF-1 tions constitute a compelling case for an important in- in integrating a variety of metabolic pathways by mod- tegrative function for NRF-1 in communication between ulating the expression of a key activity. Putative NRF-1 sites in genes encoding a number of receptors and com- nuclear and mitochondrial genetic compartments. ponents of signal transduction networks may indicate a role in the establishment or maintenance of regulatory Other regulatory functions for NRF-1 in coordinate systems that influence these or other cellular functions. gene expression Also prominent in the list are genes involved in chromo- some maintenance and nucleic acid metabolism. This Our interest in nucleus-encoded mitochondrial func- may reflect a requirement for coordinating the expres- tions, along with the remarkable consistency in the ap- sion of the replication, transcription, and translation ma- pearance of NRF-1 sites in the majority of genes in this chinery with organelle biogenesis under certain condi- category, has led to a direct functional characterization of these sites (Table 1). However, the identification of tions. Similarly, NRF-1 sites are found in genes that may NRF-1 sites in the tyrosine aminotransferase and eIF-2oL be directly involved in cell cycle regulation (cdc2, RCC1) or are regulated by cell growth (omithine decarboxylase, genes suggested a broader integrative function for NRF-1 (Chau et al. 1992) and prompted a systematic search for DNA polymerase-a, and GADD153). Maintenance of potential binding sites in published gene sequences (Ta- mitochondria might be expected to require sensitivity to ble 1). The sites listed have only a single mismatch with proliferative signals, and it is tempting to speculate that the consensus derived from the 14 tested binding sites; NRF-1 may function in transducing such signals. Thus, and in each case, the mismatch is known to be allowed although the best defined biological role for NRF-1 is in at that position in the functional sites. These imposed the nuclear control of mitochondrial function, the constraints in selecting potential binding sites make it NRF-1 protein or related proteins having the NRF-1 DNA-binding domain may have the potential for inte- likely that the genes containing these sites are targets for grating diverse functions required for cell maintenance, NRF-1. In fact, 14 of the 48 putative target genes in Table growth, and proliferation. 1 have NRF-1 sites identical to those of known function. It should be emphasized, however, that the effects of NRF-1 on basal promoter activity are influenced by pro- moter context. For example, mutation of the NRF-1 sites Materials and methods in the cytochrome c and COXVIc genes (Evans and Scar- Purification and amino acid sequencing of NRF-1 pulla 1989, 1990) results in a more modest effect on pro- DEAE-agarose and heparin-agarose fractionation of HeLa nu- moter activity than mutation of the NRF-1 sites in the clear extracts have been described (Virbasius et al. 1993), except mtTFA (Virbasius and Scarpulla 1994) or 5-ALA syn- NRF-1 fractions were eluted with HEPES-D, 0.45 M KC1, diluted thase genes (Braidotti et al. 1993). Also, the conservation to 0.1 M KC1 with HEPES-D, 0.1% NP-40, and loaded onto a of the NRF-1 DNA-binding domain in P3A2 and EWG NRF-l-specific affinity column as described (Chau et al. 1992). suggests that this domain may be conserved in a mam- Affinity-purified NRF-1 was isolated by SDS-PAGE, transferred malian family of related factors that mediate different to nitrocellulose, and the NRF-1 band was identified by Pon- biological functions through similar recognition sites. ceau-S staining (Aebersold et al. 1987). In situ tryptic digestion Thus, a rigorous analysis ultimately requires an evalua- and peptide sequencing was performed by William S. Lane (Har- tion of the NRF-1 sites within the proper promoter con- vard Microchemistry Facility, Cambridge MA). text and the identification of the cognate activator pro- tein. With these caveats in mind, some interesting observa- Amplification of NRF-1 sequence and cDNA library screening tions emerge from Table 1. In addition to the cy- One sense primer and two sets of antisense primers were used to tochrome c and MRP RNA genes, there are several genes amplify the NRF-l-coding sequence. One set of primers, 5'- (omithine decarboxylase, GADD153, and synapsin I) GCIGAICATGCICCIGCICCICAIGAIGTIAACTC-3' derived where the NRF-1 site is conserved in a similar location from the NRF-l(72) peptide and 5'-GCYTGNGTCCANGA- in different species. A majority of the genes are ubiqui- NAC-3' derived from the NRF-l(38) peptide, yielded a PCR product. Briefly, cDNA was synthesized using AMV reverse tously expressed, consistent with the wide distribution transcriptase {Promega) with 2 p.g of oligo(dT)-primed HeLa of the NRF-1-binding activity. In cases where ubiquitous poly(A) RNA in a total volume of 20 Izl. The product (2 ~1) was and tissue-specific members of a gene family exist mixed with two different pairwise combinations of sense and (cytochrome c, 5-ALA synthase, enolase, and the Na/ antisense primers and amplified with AmpliTaq DNA polymer- K ATPase), NRF-1 sites are detected only in the widely ase (Perkin-Elmer Cetus) for 50 cycles (94°C for 1 rain, 50°C for expressed gene, suggesting involvement of NRF-1 in 2 min, 72°C for 2 rain). The products were ligated to M13mpl8 general, rather than tissue-specific cellular functions. for sequencing. A 269-bp PCR product, encoding portions of the Among a variety of metabolic enzymes encoded by two NRF-1 peptides, was subcloned into pGEM4 Blue. The in- these genes, several (omithine decarboxylase and the sert was labeled by nick translation for screening a HeLa cDNA branched-chain a-keto acid dehydrogenase) catalyze the library in KZAPII (a gift of Dr. R. Morimoto, Northwestern Uni- versity, Evanston, IL). The 3-kb insert of one of two positive rate-limiting step of their pathways (polyamine synthe- GENES & DEVELOPMENT 2441 Downloaded from genesdev.cshlp.org on November 17, 2021 - Published by Cold Spring Harbor Laboratory Press Virbasius et al. phages was subcloned into pGEM7zf(+) (Promega} using the COXVb-109/-87 flanking XhoI and XbaI sites in the phage to generate pGEM7zf- GATCCAGAACTGCGCATGTGCGGCGTCA NRF1. Subclones in M13 were sequenced on both strands using GTCTTGACGCGTACACGCCGCTGATCGA Sequenase (U.S. Biochemical). mtTFA-78/-49 CGCTCTCCCGCGCCTGCGCCAATT GGGCGCGGACGCGGTTAAGGCGGG Plasmid constructions The full-length NRF-1 cDNA was put into the pSG5 expression For antiserum supershift experiments, 1 ~1 of anti-NRF-1 or vector (Green et al. 1988} in the sense orientation by ligating the preimmune serum was added to binding reactions that had been upstream EcoRI-PstI fragment and downstream PstI-BamHI incubated for 15 rain as described above. After an additional 15 fragment of pGEM7zf-NRF1 into the EcoRI and BamHI sites of rain of incubation, the reaction was fractionated by electropho- pSG5 to produce pSG5NRF1/1-2970. The antisense construc- resis on 4% (58 : 1) acrylamide/bisacrylamide gels. tion was generated by ligating the XhoI-BamHI fragment into the same sites of a pSG5 vector modified by insertion of a XhoI linker at the EcoRI site. pSGSNRF1/1-1662 (construct B in Fig. Expression and purification of the recombinant NRF-1 8A) was generated by addition of a BamHI linker at the AccI site and antiserum preparation 35 bp downstream of the termination codon, removing the 3'- NRF-1 was expressed using the T7 expression system {Studier et untranslated region, pSGSNRF 1 / 1-1030 and pSGSNRF 1 / 1-908 al. 1990). An NcoI site was introduced at the NRF-1 initiation (constructs F and G in Fig. 8A) were generated by exonuclease III codon by PCR using a sense primer, GAACTCCATGGAG- digestion from the AccI site, followed by addition of an Asp718I GAACAC, and the same antisense primer as above. The PCR linker and cloning into a pGEM7zf(+) containing a synthetic product was digested with NcoI and PstI to give a 221-bp frag- universal translation terminator (Pharmacia) in its Sinai site. ment, ligated with a PstI-BamHI fragment containing the rest Deleted fragments were recloned to pSG5 using flanking PstI of NRF-1-coding region to the NcoI and BamHI sites of pET3d and BamHI sites. Amino-terminal truncations were generated and used to transform E. coli BL21(DE3). Partial purification of either by PCR cloning or restriction enzyme cleavages. Briefly, the overexpressed protein was as described (Pognonec et al. pSG5NRF1/348-1662 (construct J in Fig. 8A) was generated by 1991). The ammonium sulfate fraction was diluted 10-fold with PCR using a sense primer, CCCATGGGAATGGCCGC, and an HEPES-D and applied to a 1-ml heparin-agarose column in antisense primer, CCACGGCAGAATAATTC, matching se- HEPES-D, 0.1 M KC1. NRF-1 was eluted in a 0.1-1 M KC1 gra- quence downstream of the natural EcoRV site. The PCR product dient. Goat anti-NRF-1 serum was raised against the heparin- was cut with NcoI and EcoRV, and the 480-bp fragment was agarose peak fraction (East Acres Biologicals, Southbridge, MA). cloned into pGEM5Zf( + ) (Promega). An EcoRI linker was added at an adjacent EagI site. EcoRI and HincII digestion of this vec- Methylation interference and footprinting tor released a fragment that was then ligated with the HinclI/ BamHI fragment of pSGSNRF1/1-1662 into the EcoRI and Methylation interference and DNase I footprinting were de- BamHI sites of pSG5. pSG5NRF1/444-1662 and pSG5NRF1/ scribed previously (Evans and Scarpulla 1990). A 130-ng recom- 551-1662 (constructs K and L in Fig. 8A) were generated by binant NRF-1 ammonium sulfate pellet was used in the prepar- digestion of pSGSNRF1/1-1662 with AflIII and DraI, respec- ative shift of methylated fragments, and a 20-ng NRF-1 heparin- tively, followed by the addition of NcoI linkers. Recloning into agarose fraction dialyzed to 0.1 M KC1 was used in footprinting. pSG5 was the same as described for pSG5NRF1/349-1662. When indicated, a 200-fold excess of NRF-1-specific or -nonspe- pSG5NRF1/119-1662 (construct I in Fig. 8A) was generated by cific competitors was added before the addition of labeled frag- ligating the 221-bp NcoI-PstI fragment from pET3dNRF1 (see ment in footprinting. below) into pGEM5zf and cloning back into pSG5. In vitro transcription In vitro transcription and analysis of transcripts were done as In vitro transcription, translation, and mobility shift assay described (Virbasius et al. 1993), except 54 ~g of HeLa nuclear The pSG5 vectors described above include a T7 promoter up- extract and 0.5 ~g DNA template were used. The recombinant stream of the cloning site. To generate runoff transcription tem- NRF-1 used was the dialyzed heparin-agarose peak fraction. plates, pSGSNRFI/1-2970 was linearized with BamHI to gen- Constructions used as templates have been described (Evans erate the full-length (construct A in Fig. 8A), NcoI (C, A476- and Scarpulla 1989, 1990). 503), EcoNI (D, A419-503), BglI {E, d1331-503), or EcoRV (H, A238-503). The other carboxy-terminal and all amino-terminal Transient transfection deletions were linearized with BamHI. In vitro transcription was performed by using T7 polymerase (Promega) and RNA For transfection, 3x 106-4X 106 COS cells were resuspended in translated in wheat-germ extract (Promega). Reactions con- 0.8 ml of ZAP buffer (20 mM HEPES, 137 mM NaC1, 0.5 mM KC1, tained unlabeled methionine for use in mobility-shift assays or 0.7 mM Na2HPO4, 6 mM dextrose adjusted to pH 7.05) and [aSS]methionine for analysis of the protein products on SDS-- mixed with 5 ~g of reporter plasmid and 15 ~g of pGEM4blue polyacrylamide gels. Mobility shift assays were done as de- carrier. The cells were then subjected to a single pulse {270 V, scribed (Evans and Scarpulla 1990). Binding reactions contained 960 ~F} using a Bio-Rad Gene Pulser. The cells were harvested 1 ~g of sonicated calf thymus DNA and 2 ~g of BSA in HEPES- after 48 hr, and extracts were analyzed for chloramphenicol D, 100 mM KC1. When indicated, a 200-fold excess of NRF-1- acetyltransferase (CAT) activity and CAT-coding DNA in Hirt specific or -nonspecific competitors was added. In addition to supematants as described previously (Evans and Scarpulla oligonucleotides described previously (Evans and Scarpulla 1988). The reporter plasmids used were either RC4CATBA/ 1990; Chau et al. 1992), the following oligonucleotides were - 66BA or RC4CATBA/- 66BA with NRF-! oligonucleotides employed in binding assays: from MCOSb, mtTFA, or RC4 cloned upstream as described 2442 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on November 17, 2021 - Published by Cold Spring Harbor Laboratory Press Cloning ot transcription [actor NRF-1 previously (Evans and Scarpulla 1990). Values were the average piratory factor 1 activation sites in genes encoding the of six separate determinations -+ S.D. gamma-subunit of ATP synthase, eukaryotic initiation fac- tor 2a, and tyrosine aminotransferase. Specific interaction of purified NRF-1 with multiple target genes• J. Biol. Chem. Acknowledgments 267: 6999-7006. Clayton, D.A. 1991. Replication and transcription of vertebrate We thank William S. Lane of the Harvard Microchemistry Fa- mitochondrial DNA. Annu. Rev. Cell Biol. 7: 453-478. cility for his invaluable guidance and expertise in performing Das, A.T., A.C. Amberg, H. Malingre, P. Moerer, R. Charles, the peptide mapping and amino acid sequencing of NRF-1. This A.F. Moorman, and W.H. Lamers. 1993. Isolation and char- work was supported by U.S. Public Health Service grant acterization of the rat gene encoding glutamate dehydroge- GM32525-10 from the National Institutes of Health. R.C.S. is nase. Eur. J. Biochem. 211: 795-803. the recipient of Faculty Research Award FRA-361 from the Degen, S.J., L.A. Stuart, S. Hart, and C.S. Jamison. 1991. Char- American Cancer Society• acterization of the mouse eDNA and gene coding for a he- The publication costs of this article were defrayed in part by patocyte growth factor-like protein: Expression during de- payment of page charges• This article must therefore be hereby velopment. Biochemistry 30:9781-9791. marked "advertisement" in accordance with 18 USC section Desimone, S.M. and K. White. 1993. The Drosophila erect wing 1734 solely to indicate this fact. gene, which is important for both neuronal and muscle de- velopment, encodes a protein which is similar to the sea urchin P3A2 DNA binding protein• Mol. Cell. Biol. 13: References (6) 3641-3949. Adachi, M. and Y. Tsujimoto. 1990. Potential Z-DNA elements Diffley, J.F. and B. Stillman. 1991. A close relative of the nu- surround the breakpoints of chromosome translocation clear, chromosomal high-mobility group protein HMG1 in within the 5' flanking region of bcl-2 gene. Oncogene yeast mitochondria. Proc. Natl. Acad. Sci. 88: 7864-7868. 5: 1653-1657. Evans, M.J. and R.C. Scarpulla. 1988. Both upstream and intron Aebersold, R.H., J. Leavitt, R.A. Saavedra, L.E. Hood, and S.B.H. sequence elements are required for elevated expression of Kent. 1987. Internal amino acid sequence analysis of pro- the rat somatic cytochrome c gene in COS-1 cells. Mol. Cell. teins separated by one- or two-dimensional gel electropho- Biol. 8: 35--41. resis after in situ protease digestion on nitrocellulose. Proc. • 1989. Interaction of nuclear factors with multiple sites Natl. Acad. Sci. 84: 6970-6974. in the somatic cytochrome c promoter. Characterization of Amy, C.M., B. Williams-Ahlf, J. Naggert, and S. Smith. 1990. upstream NRF-1, ATF and intron Spl recognition sites. ]. Molecular cloning of the mammalian fatty acid synthase Biol. Chem. 264: 14361-14368. gene and identification of the promoter region. Biochemistry • 1990. NRF-I: A trans-activator of nuclear-encoded res- 271: 675-679• piratory genes in animal cells. Genes & Dev. 4: 1023-1034. Attardi, G. and G. Schatz. 1988. Biogenesis of mitochondria. Fisher, R.P., J.N. Topper, and D.A. Clayton. 1987. Promoter Annu. Rev. Cell Biol. 4: 289-333. selection in human mitochondria involves binding of a tran- Baer, M., T.W. Nilsen, C. Costigan, and S. Altman. 1990. Struc- scription factor to orientation-independent upstream regula- ture and transcription of a human gene for H1 RNA, the tory elements. Cell 50: 247-258. RNA component of human RNase P. Nucleic Acids Res. Forrest, G.L., S. Akman, J. Doroshow, H. Rivera, and W.D. Ka- 18: 97-103. plan. 1991. Genomic sequence and expression of a cloned Basu, A. and N.G. Avadhani• 1991. Structural organization of human carbonyl reductase gene with daunorubicin reduc- nuclear gene for subunit Vb of mouse mitochondrial cy- tase activity• Mol. Pharmacol. 40: 502-507. tochrome c oxidase. J• Biol. Chem. 266: 15450-15456• Furuno, N., K. Nakagawa, U. Eguchi, M. Ohtsubo, T. Nishim- Bender, T.P. and W.M. Kuehl. 1986. Murine myb protooncogene oto, E. Soeda, and M. Ohtubo. 1991. Complete nucleotide mRNA: eDNA sequence and evidence for 5' heterogeneity. sequence of the human RCC1 gene involved in coupling Proc. Natl. Acad. Sci. 83: 3204-3208. between DNA replication and mitosis. Genomics 11: 459- Biamonti, G., M. Buvoli, M.T. Bassi, C. Morandi, F. Cobianchi, and S. Riva. 1989. Isolation of an active gene encoding hu- Giallongo, A., D. Oliva, L. Cali, G. Barba, G. Barbieri, and S. Feo. man hnRNP protein A1. Evidence for alternative splicing. J. 1990. Structure of the human gene for alpha-enolase. Eur. ]. Mol. Biol. 207: 491-503. Biochem. 190: 567-573. Braidotti, G., I.A. Borthwick, and B.K. May. 1993. Identification Green, S., I. Isseman, and E. Sheer. 1988. A versatile in vivo and of regulatory sequences in the gene for 5-aminolevulinate in vitro eukaryotic expression vector for protein engineering. synthase from rat. J. Biol. Chem. 268:1109-1117. Nucleic Acids Res. 16: 369. Calzone, F.J., C. Hoog, D.B. Teplow, A.E. Cutting, R.W. Zeller, Haendler, B. and E. Hofer. 1990. Characterization of the human R.J. Britten, and E.H. Davidson. 1991. Gene regulatory fac- cyclophilin gene and of related processed pseudogenes. Eur. tors of the sea urchin embryo. I. Purification by affinity chro- J. Biochem. 190: 477-482. matography and cloning of P3A2, a novel DNA binding pro- Hoog, C., F.J. Calzone, A.E. Cutting, R.J. Britten, and E.H. Dav- tein. Development 112: 335-350• idson. 1991. Gene regulatory factors of the sea urchin em- Cavailles, V., P. Augereau, and H. Rochefort. 1993. Cathepsin D bryo. II. Two dissimilar proteins, P3A1 and P3A2, bind to the gene is controlled by a mixed promoter, and estrogens stim- same target sites that are required for early territorial gene ulate only TATA-dependent transcription in breast cancer expression• Development 112:351-364. cells. Proc. Natl. Acad. Sci. 90: 203-207. Hurt, M.M., N. Chodchoy, and W.F. Marzluff. 1989. The mouse Chang, D.D. and D.A. Clayton. 1989. Mouse RNase MRP RNA histone H2a.2 gene from chromosome 3. Nucleic Acids Res. is encoded by a nuclear gene and contains a decamer se- 17: 8876. quence complementary to a conserved region of mitochon- Jenkins, E.P., C.L. Hsieh, A. Milatovich, K. Normington, D.M. drial RNA substrate. Cell 56: 131-139. Berman, U. Francke, and D.W. Russell. 1991. Characteriza- Chau, C.A., M.J. Evans, and R.C. Scarpulla. 1992. Nuclear res- tion and chromosomal mapping of a human steroid 5 alpha- GENES & DEVELOPMENT 2443 Downloaded from genesdev.cshlp.org on November 17, 2021 - Published by Cold Spring Harbor Laboratory Press Virbasius et al. reductase gene and pseudogene and mapping of the mouse vitro. Mol. Cell. Biol. 13: 1951-1961. homologue. Genornics 11:1102-1112. Park, J.S., J.D. Luethy, M.G. Wang, J. Fargnoli, A.J. Fornace, Jr., Kadonaga, J.T. 1991. Purification of sequence-specific binding O.W. McBride, and N.J. Holbrook. 1992. Isolation, charac- proteins by DNA affinity chromatography. Methods Enzy- terization and chromosomal localization of the human rnol. 208: 10-12. GADD153 gene. Gene 116: 259-267. Katz, A. and C. Kahana. 1988. Isolation and characterization of Parmentier, M. and A. Lefort. 1991. Structure of the human the mouse ornithine decarboxylase gene. J. Biol. Chem. brain calcium-binding protein calretinin and its expression 263: 7604-7609. in bacteria. Eur. J. Biochern. 196: 79-85. Khew-Goodall, Y., R.E. Mayer, F. Maurer, S.R. Stone, and B.A. Pearson, B.E., H.P. Nasheuer, and T.S. Wang. 1991. Human Hemmings. 1991. Structure and transcriptional regulation of DNA polymerase alpha gene: Sequences controlling expres- protein phosphatase 2A catalytic subunit genes. Biochemis- sion in cycling and serum-stimulated cells. Mol. Cell. Biol. try 30: 89-97. 11: 2081-2095. Kreysing, J., K. von Figura, and V. Gieselmann,. 1990. Structure Pognonec, P., H. Kato, H. Sumimoto, M. Kretzschmar, and R.G. of the arylsulfatase A gene. Eur. L Biochern. 191: 627-631. Roeder. 1991. A quick procedure for purification of func- Ku, D.H., S.C. Wen, A. Engelhard, N.C. Nicolaides, K.E. Lipson, tional recombinant proteins over-expressed in E. coli. Nu- T.A. Marino, and B. Calabretta. 1993. c-myb transactivates cleic Acids Res. 19: 6650. cdc2 expression via Myb binding sites in the 5'-flanking re- Sauerwald, A., C. Hoesche, R. Oschwald, and M.W. Kilimann. gion of the human cdc2 gene. J. Biol. Chem. 268: 2255-2259. 1990. The 5'-flanking region of the synapsin I gene. A Kunze, N., G.C. Yang, M. Dolberg, R. Sundarp, R. Knippers, and G + C-rich, TATA- and CAAT-less, phylogenetically con- A. Richter. 1991. Structure of the human type I DNA topoi- served sequence with cell type-specific promoter function. J. somerase gene. J. Biol. Chem. 266: 9610-9616. Biol. Chem. 265: 14932-14937. Kurihara, T., K. Monoh, K. Sakimura, and Y. Takahashi. 1990. Sittman, D.B., R.A. Graves, and W.F. Marzluff. 1983. Structure Alternative splicing of mouse brain 2',3'-cyclic-nucleotide of a cluster of mouse histone genes. Nucleic Acids Res. 3'-phosphodiesterase mRNA. Biochern. Biophys. Res. Corn- 11: 6679--6697. rnun. 170: 1074-1081. Studier, F.W., A.H. Rosenberg, J.J. Dunn, and J.W. Dubendorff. Kutt, H., J. Herz, and K.K. Stanley. 1989. Structure of the low- 1990. Use of T7 RNA polymerase to direct expression of density lipoprotein receptor-related protein (LRP) promoter. cloned genes. Methods Enzyrnol. 185: 60-89. Biochirn. Biophys. Acta 1009: 229-236. Tasanen, K., J. Oikarinen, K.I. Kivirikko, and T. Pihlajaniemi. Lee, C.-M., R.S. Haun, S.-C. Tsai, J. Moss, and M. Vaughan. 1992. Promoter of the gene for the multifunctional protein 1992. Characterization of the human gene encoding ADP- disulfide isomerase polypeptide. Functional significance of ribosylation factor 1, a guanine nucleotide-binding activator the six CCAAT boxes and other promoter elements. J. Biol. of cholera toxin. J. Biol. Chem. 267: 9028-9034. Chem. 267: 11513-11519. Luethy, J.D., J. Fargnoli, J.S. Park, A.J. Fornace, Jr., and N.J. Tewari, D.S., D.M. Cook, and R. Taub. 1989. Characterization Holbrook. 1990. Isolation and characterization of the ham- of the promoter region and 3' end of the human insulin re- ster gadd 153 gene. Activation of promoter activity by agents ceptor gene. J. Biol. Chem. 264: 16238-16245. that damage DNA. J. Biol. Chem. 265: 16521-16526. Tominaga, K., S. Akiyama, Y. Kagawa, and S. Ohta. 1992. Up- Lutfalla, G., K. Gardiner, D. Proudhon, E. Vielh, and G. Uze. stream region of a genomic gene for human mitochondrial 1992. The structure of the human interferon alpha/beta re- transcription factor 1. Biochim. Biophys. Acta 1131: 217- ceptor gene. J. Biol. Chem. 267: 2802-2809. Mamula, P.W. and I.D. Goldfine. 1992. Cloning and character- Topper, J.N. and D.A. Clayton. 1990. Characterization of hu- ization of the human insulin-like growth factor-I receptor man MRP/Th RNA and its nuclear gene: Full length MRP/ gene 5'-flanking region. DNA Cell Biol. 11: 43-50. Th RNA is an active endoribonuclease when assembled as Mitsubuchi, H., Y. Nobukuni, F. Endo, and I. Matsuda. 1991. an RNP. Nucleic Acids Res. 18: 793-799. Structural organization and chromosomal localization of the Tsukamoto, T., R. Toyama, H. Itoh, T. Kozasa, M. Matsuoka, gene for the E 1 beta subunit of human branched chain alpha- and Y. Kaziro. 1991. Structure of the human gene and two rat keto acid dehydrogenase. J. Biol. Chem. 266: 14686-14691. cDNAs encoding the alpha chain of GTP-binding regulatory Miyake, S., Y. Emori, and K. Suzuki. 1986. Gene organization of protein Go: Two different mRNAs are generated by alterna- the small subunit of the human calcium-activated neutral tive splicing. Proc. Natl. Acad. Sci. 88: 2974-2978. protease. Nucleic Acids Res. 14: 8805-8807. Virbasius, J.V. and R.C. Scarpulla. 1991. Transcriptional activa- Miyatake, S., T. Otsuka, T. Yokota, F. Lee, and K. Arai. 1985. tion through ETS domain binding sites in the cytochrome c Structure of the chromosomal gene for granulocyte-macro- oxidase subunit IV gene. Mol. Cell. Biol. 11: 5631-5638. phage colony stimulating factor: Comparison of the mouse ~. 1994. Activation of the human mitochondrial transcrip- and human genes. EMBO J. 4: 2561-2568. tion factor A gene by nuclear respiratory factors: A potential Moshier, J.A., D.L. Osborne, M. Skunca, J. Dosescu, J.D. Gilbert, regulatory link between nuclear and mitochondrial gene ex- M.C. Fitzgerald, G. Polidori, R.L. Wagner, R. Friezner, S.J. pression in organelle biogenesis. Proc. Natl. Acad. Sci. {in Degen, and G.D. Luk. 1992. Multiple promoter elements pressl. govern expression of the human omithine decarboxylase Virbasius, J.V., C.A. Virbasius, and R.C. Scarpulla. 1993. Iden- gene in colon carcinoma cells. Nucleic Acids Res. 20: 2581- tity of GABP with NRF-2, a multisubunit activator of cy- tochrome oxidase expression, reveals a cellular role for an Nagley, P. 1991. Coordination of gene expression in the forma- ETS domain activator of viral promoters. Genes & Dev. tion of mammalian mitochondria. Trends Genet. 7: 1-4. 7: 380-392. Nojima, H. 1989. Structural organization of multiple rat cal- Wagner, M. and R.P. Perry. 1985. Characterization of the multi- modulin genes. J. Mol. Biol. 208: 269-282. gene family encoding the mouse $16 ribosomal protein: Parisi, M.A., B. Xu, and D.A. Clayton. 1993. A human mito- Strategy for distinguishing an expressed gene from its pro- chondrial transcriptional activator can functionally replace a cessed pseudogene counterparts by an analysis of total geno- yeast mitochondrial HMG-box protein both in vivo and in mic DNA. Mol. Cell. Biol. 5: 3560-3576. 2444 GENES & DEVELOPMENT Downloaded from genesdev.cshlp.org on November 17, 2021 - Published by Cold Spring Harbor Laboratory Press Cloning of transcription factor NRF-1 Wallace, D.C. 1992. Diseases of the mitochondrial DNA. Annu. Rev. Biochem. 61: 1175-1212. ~. 1993. Mitochondrial diseases: Genotype versus pheno- type. Trends Genet. 9: 128-133. Walter, T., B. Drabent, H. Krebs, M. Tomalak, S. Heiss, and B.J. Benecke. 1989. Cloning and analysis of a human 86-kDa heat-shock-protein-encoding gene. Gene 83:105-115. Wen, L., J.K. Huang, and P.I. Blackshear. 1989. Rat ornithine decarboxylase gene. Nucleotide sequence, potential regula- tory elements, and comparison to the mouse gene. J. Biol. Chem. 264: 9016-9021. Xu, B. and D.A. Clayton. 1992. Assignment of a yeast protein necessary for mitochondrial transcription initiation. Nucleic Acids Res. 20: 1053-1059. Yagawa, Y., K. Kawakami, and K. Nagano. 1990. Cloning and analysis of the 5'-flanking region of rat Na +/K( + ]-ATPase alpha 1 subunit gene. Biochim. Biophys. Acta 1049: 286- Zhou, Q.Y., C. Li, and O. Civelli. 1992. Characterization of gene organization and promoter region of the rat dopamine D1 receptor gene. J. Neurochem. 59: 1875-1883. GENES & DEVELOPMENT 2445 Downloaded from genesdev.cshlp.org on November 17, 2021 - Published by Cold Spring Harbor Laboratory Press NRF-1, an activator involved in nuclear-mitochondrial interactions, utilizes a new DNA-binding domain conserved in a family of developmental regulators. C A Virbasius, J V Virbasius and R C Scarpulla Genes Dev. 1993, 7: Access the most recent version at doi:10.1101/gad.7.12a.2431 This article cites 71 articles, 31 of which can be accessed free at: References http://genesdev.cshlp.org/content/7/12a/2431.full.html#ref-list-1 License Receive free email alerts when new articles cite this article - sign up in the box at the top Email Alerting right corner of the article or click here. Service Copyright © Cold Spring Harbor Laboratory Press
Journal
Genes & Development – Unpaywall
Published: Dec 1, 1993