5 pages

Purified Hepatocyte Nuclear Factor 1 Interacts with a Family of Hepatocyte-Specific Promoters

Please download to get full document.

View again

of 5
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Purified Hepatocyte Nuclear Factor 1 Interacts with a Family of Hepatocyte-Specific Promoters
  Proc. Natl. Acad. Sci. USA Vol. 85, pp. 7937-7941, November 1988 Biochemistry Purified hepatocyte nuclear factor 1 interacts withafamily ofhepatocyte-specific promoters (liver-specific expression/DNA-binding proteins/affinity purification/hepatitis B virus) GILLESCOURTOIS, SUSANNEBAUMHUETER, AND GERALD R. CRABTREE Department of Pathology, Stanford University School of Medicine, Stanford, CA 94305 Communicated by James E. Darnell, Jr., July 26, 1988 ABSTRACT Duringdevelopment cell types arise through the activation or repressionof classes of specific genes. One hypothesis is that this phenomenon is realized by tissue-specific factors playing a roleat the transcription level. Recently we have described a liver-specific nuclear protein, hepatocytenuclear factor 1, that appears to be involved in thetranscrip- tion of the fibrinogen and a1-antitrypsin genes. In this report we describe the purification of hepatocyte nuclear factor 1 and demonstrate that it interacts with essential promoter regionsof many liver-specific genes, including albumin, a-fetoprotein, and transthyretin. This finding suggests that hepatocyte nu- clearfactor 1 couldbeone factor necessary for establishingthe liver phenotype. We also show that this protein binds to the promoter of the surface-antigen gene of the hepatitis B virus, a virus characterized by a highdegreeofhepatotropism. Transcription in eukaryotes involves the specific interaction of nuclear proteins with discrete DNA elements located in the promoter or enhancer region ofgenes (1, 2 . Promoter regions cover, generally, the 100-200 base pairs(bp) upstream of the cap site, whereasenhancers can befound at large distances in the 5' or 3' regions of the genesthey regulate or even within them (1,2). In general, the full transcription activity of a gene implies the interaction between its promoterand enhancer(s) components through the various proteinsthey bind. Some of the knownDNA-binding proteins are restricted to a cell lineage (3-9). They interact with DNA sequences necessary for tissue-specificactivation or repression of genes and constitute at least a part of the machinery eukaryotes use to synthesizeproteins in a particular cell type. Transcrip- tional analysis of several hepatic genes (albumin, a-fetopro- tein, and transthyretin) has revealed the presenceof liver- specific promoters and/or enhancers (8-16). Inseveral in- stances proteinsbinding to these regions have been described (8-10, 12,17, 18). Analyzing the transcription of the fibrin- ogen genes we have localized in the 13chain promoter a functional sequence that binds a liver-specific nuclear pro- tein, hepatocyte nuclear factor 1 (HNF1)  8 . Because HNF1 also interacts with the gene for the a chainof fibrinogen and another liver-specific gene, a1-antitrypsin, we suggested that HNF1 couldbeinvolved in developmentally regulated gene expression in the liver  8 . The data presented here, which demonstrate that purified HNF1 interacts with a large family of hepatic genes, support this hypothesis. MATERIALS AND METHODS UV Cross Linking. The UV cross-linking experiments were done essentially according to Chodosh etal. (19). Seventynanograms of a 28-bpdouble-strand oligomer, corresponding to the nucleotides - 102 to - 75 of the rat fibrinogen ,-chain genepromoter, were hybridizedwith 3 Ag of the comple-mentary sequence 5'-TTTCCCTG-3'. By useof this primer the second strand was synthesizedwith the Klenow enzyme in the presence of40 ACi of [a-32P]dATP (1 Ci = 37 GBq), 500 AM of bromodeoxyuridine,and 500 ILM of dCTP and dGTP.The full-length probe was purified on a 12% poly- acrylamide gel, and 0.5 ng (50,000 cpm) wereincubated with either 15 ,ug of a rat hepatoma cell line (Faza), liver nuclear extract, or 2 ,tg of heparin-Sepharose-purified liverextract for 45 min at room temperature.After30 min of UV irradiation at room temperaturewith a Fotodyne UV trans- illuminator (A emission = 310nm), the mixture was analyzedon an 8 sodium dodecyl sulfate (SDS)/polyacrylamide gel after reductionwith100 mM dithiothreitol. After fixation the gel was stained with Coomassie blue tovisualize the molec- ular massmarkers and autoradiographed. Purification of HNF1. Crude rat liver nuclear extracts wereprepared essentially as described (8). Starting with 50 rats we usually obtained 2.5-3.5 g of nuclear proteins in 150 ml of buffer C100 [50 mM Hepes, pH 7.8/100 mM KCl/0.1 mM EDTA/1 mM dithiothreitol/1 mM phenylmethylsulfonyl fluoride/10 (vol/vol) glycerol]. The extract was applied to a 250-ml heparin-Sepharose column pre-equilibrated with buffer C100. The columnwaswashed with 500 ml of buffer C100, and then the proteins were eluted with an800-ml linear KCl gradient (100 ml/hr). Fractionsof 4 mlwere collected and analyzed for HNF1 by a mobility-shift assay (see below). The HNF1-containing fractions werepooled (volume, 30-40 ml; 200-300 mg), dialyzed againstbuffer C100, andchroma-tographedon an affinity column. The column was prepared using CNBr-activatedSepharose CL-4B and multimers of the - 102- to - 75-bpsequenceof rat fibrinogen (3chain pro- moter (20). A coupling of -25 ,ug of DNA per ml of resin was achieved. Aliquots of 125 mg of heparin-Sepharose-purified HNF1 wereincubated 10 min at 4°C with poly(dI-dC) (10 ,ug/ml) and then loaded on a 5-ml affinity column previously equilibrated with buffer C100. Thecolumnwaswashed extensively with buffer C100 and eluted sucessively with20 ml of 0.2, 0.4, 0.6,0.8, and 1 M KCl in buffer C (buffer C100 without KCl). Fractions containing HNF1 (0.4 M KCl) were diluted to 0.1 M KCl, incubated at 4°C for 10 min with poly(dI-dC) (2 ,ug/ml), and reapplied to the same affinity column. Elutionof HNF1 from the column wasdone as described forthe first pass. Analysisof protein was con- ductedaccording to Laemmli (21) withuseofa silver-staining kit (Bio-Rad). Denaturation-Renaturation of HNF1.The protocol de- scribed byHager andBurgess (22) was followed with only minor modifications: 100 ,ug ofbovineserum albumin was added as protein carrier and, after denaturation with guani- dine hydrochloride, the denaturing agent was removed by dialysis against buffer C100. Abbreviations: HNF1, hepatocyte nuclear factor 1; SDS, sodiumdodecyl sulfate; ,B 28, 28-bpdouble-stranded probe containing the HNFl-binding site of the rat gene encoding the chainof fibrinogen. 7937 The publication costs of thisarticle were defrayed in part by page charge payment. This article must therefore be hereby marked  advertisement in accordance with 18 U.S.C. §1734 solely to indicate this fact.  7938 Biochemistry: Courtois et al. Mobility-Shift Assay and DNase I Footprinting. Thesetwo techniques wereconducted as described (8). When affinity- purifiedfractions were analyzed, 10 jug or25 pg of bovine serumalbumin wereadded as protein carrier for the mobility- shift assay and the DNase footprinting, respectively. The DNase footprinting probeswere prepared asfollows: The base-pair sequence from position - 177- to + 32 of the rat promoter of the gene forthe a chain of fibrinogen,insertedinto a chloramphenicol acetyltransferase-derived vector (JYM-CAT) between Sph I and HindIII  8 , was labeled at the HindIII site. The sequence -291 to +1 of the rat /- fibrinogen promoter, insertedinto JYM-CAT plasmid be- tween Sph I and HindIII (8), was alsolabeled at the HindIII site. The Dde I fragment from - 117 to + 6 of the mouse a1-antitrypsin promoter, insertedintothe BamHI site of pGEM-1, was labeled using the Xba I site in pGEM-1. This corresponds to thelabeling of the coding strand of thea1-antitrypsin promoter. A Xba I (-430) to BstEII (+48)fragment derived from the rat albumin gene was labeled at BstEII. A BamHI (- 202) to Xba I (- 70) fragment derived from the mouse transthyretin gene was labeled at Xba I. A Ssp I (- 133) to BstEII (+ 11) fragment derived from the pre- Si promoter (23) of thehepatitis B virus was labeled at BstEIII. A HindIII (- 323) to Msp I (+ 8) fragment derived from the rat a-fetoprotein promoter (24) was labeled at Msp I. In each case, labeling was done with the Klenow enzyme. Materials. Synthetic oligomers were prepared by R. Bele- gage (EliLilly). A syntheticdouble-strand 23-mer corre- sponding to the base pairs from - 63 to - 41 of the rat albumin promoter wasfrom M. Yaniv (Pasteur Institute, Paris). Mouse a1-antitrypsin, rat albumin,and mouse prealbumin promoterswere from R. Costa, D. Grayson, and J. Darnell (Rockefeller University). Rat a-fetoprotein genomic clone wasfrom T. Sargent (National Institutes of Health), and a plasmid containing thehepatitis B virus genome was from A. Siddiqui (University of Colorado, Denver). Klenow enzyme was from Boehringer Mannheim; [a-32P]dATP was from Amersham; heparin-Sepharose CL-6B, CNBr-activated Sepharose CL-4B, and poly(DIdC) were fromPharmacia; guanidine hydrochloride and molecular weight markers were from Sigma; silver-staining kit was from Bio-Rad; bovine serum albumin was from Bethesda Research Laboratories; and restriction enzymes came from New England Biolabs. RESULTS UV Cross Linking of HNF1 to Its Binding Site. A cross- linking protocol was first devised to estimate the molecularmass of theliver-specific nuclear factor HNF1 directly from a crude nuclear extract. A bromodeoxyuridine-substituted probe, corresponding to the - 102- to - 75-bp region of the rat gene encoding the /3 chain for fibrinogen, was incubated with a nuclear extract prepared from Faza cells and irradiated under UV light to generate covalent links between the DNA and HNF1. The mixture was analyzed on a denaturing SDS/polyacrylamide gel and, as shown in Fig. 1A, a labeled band corresponding to a molecularmassof '110 kDa was detected (lane 1 . In the presenceof a large excess of unlabeled 28-mer this band disappeared, demonstrating that it represented thespecific interaction between HNF1 and the probe (Fig. la, lane 2 . In contrast, a smear covering the 40- to 60-kDa region was not competed with and most likely represented nonspecific interactions. A similar experiment was conducted with a nuclear extract derived from a rat liver, and the same 100-kDa protein was detected (Fig.lb,lanes 1 and 2 . When this extract was chromatographed through a heparin-Sepharose column (see below) and a fraction con- taining HNF1 was mixed with the probe, only the 110 kDa band was seen, further demonstrating that the 40-60 kDa a Faza kDa 97 - b Liver - + Hep *r* 4 110 66 - 36 - 1 2 go. I 1 23 FIG. 1. UV cross-linking of HNF1 to its binding site. (a) Detectionof HNF1 in a crudenuclear extract derived from the rat hepatoma cell line Faza without (lane 1) orwith (lane 2) a 100-fold excess ofunlabeled 3 28fragment. (b) Same analysis using acrude liver extract(lanes 1 and 2) ora heparin-Sepharose purifiedextract (Hep; lane 3). Arrow, HNF1. smear corresponded to nuclearprotein interacting nonspe- cifically with the probe (Fig. lb, lane 3). Purification of HNF1. The purification of HNF1 was accomplished through two chromatographic steps. First, proteins ofa crude extract from ratliver were loaded on a heparin-Sepharose column and eluted with a linear KCl gradient. To determine the fractions containingthe factor HNF1 a mobility-shift assay was used with a 28-bp double-stranded probe  13 28) containing the HNF1-binding site of the rat geneencoding the /3 chainof fibrinogen. HNF1 activity was detected in fractionseluting between 0.35 and 0.4 M KCl (Fig. 2a). At this step an 8- to10-foldpurification was achieved, and the amount of protein recoveredallowed us to use directly an affinity-chromatography procedure as a sec- ond step. This chromatography was conducted essentially according to Kadonaga and Tjian (20). The column was prepared by covalent couplingof multimers of 8 28 to a CNBr-activatedSepharose CL-4B resin. The use of poly (dI1dC) as competitor for nonspecific interactions allowed purification to near homogeneity after two passes. As shown in Fig. 2b, the protein eluted from the affinity column with 0.4 M KC1,whereasmost of the loaded protein (Fig. 2c, lane 1) appeared in the flow-through (data not shown). The 0.4 M KCl fraction was analyzed on a denaturing SDS/acrylamide gelafter silver straining (Fig. 2c, lane 3). A major band migrating at -88 kDa and two minor ones migrating at 55 and 45 kDa were seen in the preparation represented; anotherpreparation gave usonly the 88-kDa species(data notshown). The nature of the weaker bands at 55 and 45 kDa remained to be determined,but we had noticed in variouspreparations that a proteolytic degradation of HNF1 generated a lower-molecular mass form of -45 kDa, still retaining its binding activity (data not shown). To definitely identify the protein in the purified HNF1 preparation, we carried out a denaturation-renaturation ex- periment (22). The proteinsmigrating aroundandbetween 90,55, and 45 kDa were eluted from an SDS/acrylamide gel, precipitated with acetone, fully denatured with6 M guanidine hydrochloride, and allowed to renature after removing the guanidine hydrochloride with dialysis. As shown in Fig. 2d the binding activity of HNF1 was detected mostly in the 90-kDa fraction (lane 3) but also in the 55-(lane 5) and45-kDa (lane 7) fractions. The specificity of the interaction was confirmed using as competition either an excessofunlabeled Proc. Natl. Acad Sci. USA 85 (1988)  Proc. Natl. Acad. Sci. USA 85 (1988) b HNFI , F1 0 . 4 p h ft w .   , i b RMF- ~~ == N km4j 88- 55- 1 2 3 4 5 6 7 8 FIG. 2. Purification of HNF1. (a) Heparin-Sepharosechromatography. The fractions were analyzed for HNF1 using a mobility-shift assay and the probe 28. p, Probe alone; ne, analysis of 5 Ag ofcrude liver extract; ft. flow-through; HNF1, fractions containing HNF1. (b) Affinity chromatography. p, Probe (328 alone; h, pooledheparin-Sepharose HNF1 fractions; ft, flow-through; w, wash; 0.2,0.4,0.6, and 0.8, fractions eluted at 0.2 M, 0.4 M, 0.6 M, and 0.8 M KCl, respectively. (c) Analysisof the purification using silverstaining. M, molecular mass markers; H, pooledheparin-Sepharose HNF1 fractions (=2 ,gg of protein); A, affinity chromatography. Approximately 70 ng of the 0.4 M KCI-eluted fraction (second pass) was analyzed. (d) Denaturation-renaturation experiment. Approximately 600 ng of affinity-purified HNF1 was loaded on a 7 SDS/polyacrylamide gel. Proteins migrating above 90 kDa (lane 2), in the 90-kDa region (lane 3) between 90 and 60 kDa (lane 4), in the 55-kDa region (lane 5), between 55 and 45 kDa (lane 6), in the 45-kDa region (lane 7), andbelow 45 kDa (lane 8) were eluted from the gel, renatured, andanalyzed with the probe 28 (lane 1) by use of a mobility-shift assay. f3 28 or of an unrelated oligomer (data not shown). We therefore assume that the 55- and45-kDa proteins represent degradationproductsof HNF1, which is a polypeptideof 88 kDa. The difference in size obtained from the UV crosslink and purification experiments could be attributed to the 28-bp DNA component that runs with the protein in the cross- linked sample. We favor this hypothesis because no differ- ence in mobility-shift assay was seen between the heparin aFg _w 0 ~ -_- -64   lm ow -- -35 _ MP f3Fg aAt Alb am- U.. a-   .- and affinity-purifiedfractions, therefore indicatingthat the 88-kDa band was not a degradation product of a 110-kDa form. Purified HNF1 Interacts with a Family of HepaticGenes. The purified protein HNF1 was used to footprint the pro- moter of the a and   chain of fibrinogen and a1-antitrypsin genes (Fig. 3). The synthetic oligomer (328, derived from the promoter for 3-chain fibrinogen, hadbeen used for HNF1 purification(see above), but we had previously shown that Tt aFp(d) aCFp(p) HBV _ _ -   -137 _ -86 -102 -W _ -58 - -37 -109   7 -87 44 -61 w -130   -62 -102 . to - -37 Z _ _~~~~~~t _ ~~~V  Ip ftf f a : 4 WIp I HNFI: -+ FIG. 3. Purified HNF1 interacts with a family of hepatic genes. The interaction of HNF1 with the fibrinogen a chain (aFg), fibrinogen chain ((BFg), a1-antitrypsin(aAt), albumin (Alb), transthyretin (Tt), a-fetoprotein (aFp) and presequence 1 surface-antigenpromoter of the hepatitis B virus (HBV) was analyzed with a DNase I footprinting assay. In each case -5fmol of labeled coding strands were incubatedwithout (-) or with (+) 10 ng (aFg, (3Fg, aAt, Alb, HBV) or 20 ng (Tt, aFp) of affinity-purified HNF1. aFp(d) and aFp(p) represent the distal and proximal HNF1-binding sites, respectively. a p n, f w C H M kDa -205 A M d -115 M.- - -97 _... - 45 -   -45 12 3 4 ,mAM6' .A- Al.._A. -,m6A -A AAMM ,- MR. I'VIN lqw   Biochemistry: Courtois et al. 7939  _  7940 Biochemistry: Courtois et al. this factor, or a similarprotein,also interacts with the a chainof fibrinogen and a1-antitrypsin promoters (8). Footprints similar to those seen with a crude liver extract were seen in each case. Because we had noticed that the albumin, a- fetoprotein, and transthyretin promoters had sequence sim- ilarities with the HNFl-binding site, we also analyzed their interaction with our purified HNF1 preparation. Footprint analysis of the albumin and transthyretin promoters clearly demonstrated that they contain an HNFl-binding site local- ized between - 2 and - 7 bp andbetween - 37 and -109 bp, respectively. In contrast, analysis of the a-fetoprotein promoter revealedtheexistence of two HNFl-binding sites. Their affinity for HNF1 appeared different-the distal site  - 130 to - 02bp)being stronger than the proximal (-61 to -37 bp) site. An HNFl-Binding Site Is Present in the Promoter for the Surface Antigen ofthe Hepatitis B Virus. Because hepatitis B virus infects the liver, we analyzed its genome to determine whether an HNFl-binding site was present. A good putative site was identified upstreamof the gene coding for the surface-antigen protein (pre-Sl promoter). We therefore attempted to footprint this region, either with a crude liver extract or withthe purified HNF1 protein. Analysis with the crude extract indicated protection of the putative HNF1 site, and this protection was specifically antagonized by,8 28 (data not shown). A similarresult was also seen with pure HNF1 (Fig. 3). The protectedregionextends from 62 to 87 bp upstreamof the cap site forthe pre-Sl mRNA. DISCUSSION We purified the liver-specific factor HNF1 to near homoge- neity with the protocolrecently described by Kadonaga and Tjian (20). The remarkable efficiency of this procedure is illustrated by the fact that after only onepass a pattern of protein similar to the one represented on Fig. 2d was already seen, suggesting an importantdegree of purification(data not shown).Considering that thethree bands obtained after the second affinity pass correspond to HNF1 (see below), we estimate theoverall purification of HNF1 from total nuclear proteins to be 85,000 fold. The purified HNF1 preparationrevealed a major band of 88 kDa and in some preparations two additional bands at 55 and 45 kDa (Fig.2c). A similar major band (92 kDa) was seen using nonreduced samples (data not shown). The definitive demonstration that the 88-kDa polypeptidecontained thebinding activity of HNF1 and that the 55- and45-kDa polypeptides likely representeddegradationproductsof the same protein came from a denaturation-renaturation exper- iment, which indicated that in each casea binding activityspecific for the probe 28 was recovered. Because a similar mobility shift was seen in a crude extract and after renatur- ation of the 88-kDa polypeptide (data not shown), we assume that HNF1 is composed ofonly a single type of chain, but at this point we cannot exclude that it interacts with its binding site as a noncovalently linked homodimer (see below). Because we have not yet developed an assay to measure in vitro the activity of HNF1, we are unable to determinewhether our purified HNF1 preparations still have an effect on fibrinogen p-chain transcription or whether we have purified only the bindingsubunitof HNF1. If our preparation is active, it will be interesting to determine, forinstance, whether HNF1 can restore transcription activity to the promoter for fibrinogen p chain in a nonhepatic extract. Such an effect has been seen with a pituitary-specific factor for the human growth hormone gene in a HeLa extract (25). Purified HNF1 protein appears to interact with many hepatic gene promoters. Nevertheless, the definitive dem- onstration that exactly the same protein functionally interacts with all these promoters in vivo remains to be established. Preliminary experiments using probes containing the a- fibrinogen,a1-antitrypsin, andalbumin putative HNF1 site, in a mobility-shift assay, allowed thedetection in each case of a liver-specific band that comigrated with thefactor HNF1 (data not shown). In almost everycase where functional data are available, the HNFl-binding site is localized in an essential promoter region. We have already shown that the fibrinogen a and p-chain HNF1 sites appear necessary for promoter activityfor those genes in hepatoma cell lines (8). Recently De Simone et al. (10) reported that clustered pointmutations in the human a1-antitrypsin promoter between positions -68 Table 1. HNF1-binding sequences Promoter Species Sequence Fibrinogen a chain* Rat G G T G A T G A T T AA C -47 Fibrinogen a chain* Rat G T C A AA T A TT A A C -84 Fibrinogen f3 chain Human A TT AA A T A TT AA C -77 Albumin Mouse G T TAA T GA T C T A C -52 Albumin* Rat G T T AA T G A T C T A C -53 Albumin Human G TT AA T A A T C T A C -51 a-Fetoprotein Mouse G T T AC T A G TT A AC -50 a-Fetoprotein Mouse G TT A A T T A TT G G C -116 a-Fetoprotein* Rat G T T AC T A G TT A A C -49 a-Fetoprotein* Rat G TT AA T T A TT GG C -115 a-Fetoprotein Human G T T AC T A G TT A A C -47 a-Fetoprotein Human G TT A A T T A TT G G C -118 Transthyretin* Mouse G T T AC TT A TT C T C -118 Transthyretin Rat G TT A C TT A TT C T C -116 Transthyretin Human G TT AC T T A TT C TC -116 a1-Antitrypsin* Mouse G T T A A T - A T T C A T -63 al-Antitrypsin* Human G T T A A T - A TT C A C -63 Hepatitis B virus (pre S)* Human G T T A A T C A TT A C T -75 HNF1 consensus site: G17 T17T17 A17 A12T16 N A15 T18 T15 A7 All C16(18 sequences) A common DNA motif is recognized by HNF1. Only the 13-bp sequence that appears to be shared by the promoters analyzed is represented. The asterisksindicate promoters whose interaction with HNF1 is documented (this report and ref. 8 . Nucleotides matching the HNF1 consensussequence are in boldface. Numbers at right indicate thelocation of the HNF1 sites referenced to the last cytosine (or thymine) as reference. Proc. Natl. Acad Sci. USA 85 (1988)  Proc. Natl. Acad. Sci. USA 85 (1988) 7941 and -64 or between -77 and -72 drastically affects the transcription of this gene. The albumin-binding site also appears to participate in transcription of the albumin gene(M. Yaniv,personal communication). Similarly a 5'-deletion analysis of the mouse a-fetoprotein gene hasrevealed that the region between residues - 85 and - 52, which destroys the proximal HNF1 binding site (see Fig. 3), results in a large drop of transcriptional activity (13). In contrast, removal of the distal site does not appear to significantly affect tran- scription of the a-fetoprotein gene. It is noteworthy that the affinity of HNF1 for the proximal site is fairly low compared with the other sites tested (see the legend forFig. 3 . This fact couldsupport the observationof Jones et al. (26)for the binding of CTF/nuclear factor 1 to the f3globin promoter, which suggests that thelocalization and the context of a site, and not only its intrinsicaffinity, might also be importantparameters for its role in transcription. Functional informa- tion is limited forthe promoter region of the transthyretin gene. Costa et al. (11) have demonstrated that the region frombp - 151 to - 108, which contains the HNFl-binding site (Fig. 3), is involved in the tissue-specific expression of this gene,but a more precise analysis is necessary. Finally, no functional information is available concerning the HNF1 site in thehepatitis B virus pre-Sl promoter.Nevertheless, its position, -55 bp upstream of a TATA box, strongly suggests that it mightbe involved in the transcription of thesurface- antigen gene. A compilation of the DNA sequences protectedwith HNF1 reveals a common structural motif (Table 1 . Moreover, examinationof these sequences in different species reveals that they are generally extremely conserved suggesting their physiologic importance. We propose as a consensus site for HNF1 the sequence GTTAATNATTAAC (where N = A, C, T, G, or no nucleotide). This site contains an inverted repeat, indicating that HNF1 might bind as a homodimer generated by the noncovalent association of two88-kDa polypeptide chains. A quick surveyof other liver-specific genes suggests that an HNF1-binding site may exist in several ofthem, such as factor VIic (27), C-reactive protein (28), L-type pyruvatekinase (29) and play a role in their transcription. The discovery that HNF1 interacts with many hepatic genes supports the hypothesis that tissue-specific expression of families ofgenes is achieved through the use of a limited number of DNA-binding proteins. An obvious advantage of this system would be thecontrol of many genes solely by action on their common regulator(s). HNF1 could be one of these regulators in the differentiated hepatocyte. In support of such an idea, is our recent observationof a striking correlation between the presenceof HNF1 and the differen- tiation state of the hepatocyte (30). Upon dedifferentiation, HNF1 disappears and is replaced by a protein of lowermolecularmass withvery similar sequence specificity; the relationship of this smaller protein to HNF1 remains to be elucidated. In contrast, a revertant cell line that has recov- ered hepatic functions contains a factor apparently identical to HNF1. Uncovering the functionalrole of HNF1 and its behavior during the differentiation process will, neverthe- less, necessitate the cloning of this protein. Its purification, described in this report, represents a first step inthis direction. We thank R.Costa, D.Grayson, J. Darnell, T. Sargent, A. Siddiqui, R. Belegage, and M. Yaniv for generous gifts ofplasmids or synthetic oligonucleotides. We also thank J. and R. Conaway for help during the preparation of the ratliver nuclear extracts, J. P. Shaw for review of the manuscript, and N.Leger for editorial assistance. This work was supported by National Institutes ofHealthGrants CA 39612 and HL 33942 to G.R.C. 1. Dynan, W. S.   Tjian, R. (1985) Nature (London) 316, 774- 778. 2. Maniatis, T., Goodbourn, S.   Fisher, J. A. (1987) Science 236, 1237-1245. 3. Landolfi, N. F., Capra, J. D.   Tucker, P. W. (1986) Nature (London) 323, 548-551. 4. Staudt, L. M., Singh, H.,Sen, R., Wirth, T., Sharp, P. A.   Baltimore, D. (1986) Nature(London) 323, 640-643. 5. Scheidereit,C., Heguy, A.   Roeder, R. G. (1987) Cell 51, 783- 793. 6. West, B. L., Catanzaro,D. F., Mellon, S. H., Cattini,P. A., Baxter, J. D.   Reudelhuber, T. L. (1987) Mol. Cell.Biol. 7, 1193-1197. 7. Lefevre, C., Inagawa,M., Dana, S., Grindlay, J., Bodner, M.   Karin, M. (1987) EMBO J.6, 971-981. 8. Courtois,G., Morgan, J. G., Campbell, L. A.,Fourel, G.   Crabtree, G. R. (1987) Science 238 688-692. 9. Grayson, D. R., Costa, R.H., Xanthopoulos, K.G.   Darnell, J. E. (1988) Science 239, 786-788. 10. De Simone, V., Ciliberto, G., Hardon, E., Paonessa, G., Palla, F., Lundberg, L.   Cortese, R. (1987) EMBO J. 6, 2759-2766. 11. Costa, R. H., Lai, E.   Darnell, J. E. (1986) Mol. Cell. Biol. 6, 4697-4708. 12. Costa, R. H., Lai, E., Grayson, D. R.   Darnell, J. E. (1988) Mol. Cell. Biol. 8, 81-90. 13. Godbout, R., Ingram, R.   Tilghman, S. M. (1986) Mol. Cell.Biol. 6, 477-478. 14. Ott, M.-O., Sperling, L., Herbomel, P., Yaniv, M.   Weiss, M. C. (1984) EMBO J.3, 2505-2510. 15. Gorski, K., Carneiro, M.   Schibler, U. (1986) Cell 47, 767- 776. 16. Pinkert, C. A., Ornitz, D.M., Brinster, R.L.   Palmiter, R. D. (1987) Gene Dev. 1, 268-276. 17. Cereghini, S., Raymondjean, M., Carranca, A. G., Herbomel, P.   Yaniv, M. (1987) Cell 50, 627-638. 18. Lichtsteiner, S., Wuarin, J.   Schibler, U. (1987) Cell 51, 963- 973. 19. Chodosh, L. A., Carthew, R. W.   Sharp, P. A. (1986) Mol. Cell. Biol. 6, 4723-4733. 20. Kadonaga, J. T.   Tjian, R. (1986) Proc. Natl. Acad. Sci. USA 83, 5889-5893. 21. Laemmli, U. K. (1970) Nature(London) 227, 680-685. 22. Hager, D. A.   Burgess, R.R. (1980) Anal. Biochem. 109, 76- 86.23. Ono, Y., Onda, H., Sasada, R., Igarashi, K., Sugino, Y.   Nishioka, K. (1983) NucleicAcids Res. 11, 1747-1757. 24. Nahon, J. L., Danan, J. L.,Poiret, M., Tratner, I., Jose Estanyol, M.   Sala-Trepat, J. M. (1987) J. Biol. Chem. 762, 12479-12487. 25. Bodner, M.   Karin, M. (1987)Cell 50, 267-275. 26. Jones, K. A., Kadonaga, J. T., Rosenfeld, P. J., Kelly, T. J.   Tjian, R. (1987) Cell48, 79-89. 27. Gitschier, J., Wood, W. I., Goralka, T. M., Wion, K. L., Chen, E. Y., Eaton, D. H., Vehar, G. A., Capon, D. J.   Lawn, R. M. (1984) Nature(London) 312, 326-330. 28. Woo, P., Korenberg, J. R.   Whitehead, A. S. (1985) J. Biol. Chem. 260, 13384-13388. 29. Cognet,M., Lone, Y. C., Vaulont, S., Kahn, A.   Marie, J. (1987) J. Mol. Biol. 196, 11-25. 30. Baumhueter, S., Courtois, G.   Crabtree, G. R. (1988) EMBO J.7, 2485-2493. Biochemistry: Courtois et al.
Related Documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!