- Email: [email protected]
Identification and characterization of conserved cis-regulatory elements in the human keratocan gene promoter1
Biochimica et Biophysica Acta 1492 (2000) 452^459
www.elsevier.com/locate/bba
Promoter paper
Identi¢cation and characterization of conserved cis-regulatory elements in the human keratocan gene promoter1 Elena S. Tasheva *, Abigail H. Conrad, Gary W. Conrad Division of Biology, Ackert Hall, Kansas State University, Manhattan, KS 66506-4901, USA Received 20 March 2000; accepted 10 May 2000
Abstract Keratocan, along with lumican and mimecan, represent the keratan sulfate-containing proteoglycans of the vertebrate cornea that play a key role in development and maintenance of corneal transparency. In this study, we cloned 4.1 kb of the human Kera 5P-flanking region and characterized the promoter structure. Using primer extension and ribonuclease protection assay, we identify two major transcriptional start sites in the first exon. Using luciferase reporter gene transfection analysis of 5P-deletion and mutation constructs, we demonstrate positive and negative regulatory elements within a 1.3 kb upstream sequence. Comparison of human and bovine 5P-flanking sequences reveals three highly conserved regions: a 450 bp region in the first exon, a 92 bp promoter proximal conserved regulatory region identified as an enhancer in the natural context, and a 223 bp promoter distal conserved regulatory region identified as a silencer both in the natural context and in a heterologous promoter system. In addition, a conserved CArG-box residing 851 bp upstream of the first transcription start site also can lead to the repression of Kera expression in cultured corneal keratocytes. DNaseI footprinting and electrophoretic mobility shift assay demonstrate that cell type-specific factors bind to regulatory elements located in the conserved regions. Competition experiments indicate that the CTC factor and a protein that binds to the CAGA motif are likely to be among the multiple factors involved in the transcriptional regulation of the human Kera gene. ß 2000 Elsevier Science B.V. All rights reserved. Keywords : Promoter region ; Keratan sulfate proteoglycan ; Gene regulation; Corneal keratocyte
1. Introduction Keratocan, lumican and mimecan are keratan sulfate proteoglycans (KSPGs) expressed at high levels in the cornea, where the proteins are synthesized covalently associated with large, sulfated keratan sulfate chains [1]. Kera is also abundant in sclera and detectable in skin, ligament and skeletal muscle. However, similarly to lumican and mimecan, in these non-corneal tissues keratocan is present as a non-sulfated glycoprotein [2,3]. The high level of expression of KSPGs in cornea, combined with their unique Abbreviations : CPRR, conserved positive regulatory region; CNRR, conserved negative regulatory region; EMSA, electrophoretic mobility shift assay; GSP, gene-speci¢c primer; Inr, initiator element; Kera, gene encoding keratocan; KSPG, keratan sulfate proteoglycan ; PCR, polymerase chain reaction ; RACE, rapid ampli¢cation of cDNA ends; RPA, ribonuclease protection assay; tsp, transcription start point * Corresponding author. Fax: +1-785-532-6653; E-mail : [email protected] 1 The nucleotide sequence data reported here will appear in the GenBank nucleotide sequence database under the accession number: AF169962.
glycosylation, suggests that these molecules are important for the development and maintenance of corneal transparency. This conclusion is supported by several lines of evidence. Lack of sulfation of corneal KSPGs has been reported in macular corneal dystrophy type I and in opaque corneal scars [4,5]. The genomic organization and the protein structure are highly conserved between distant species, such as chicken, mouse and man [6^8]. Disruption of the lumican gene in knockout mice causes bilateral corneal opacity [9]. Despite considerable progress made in understanding the genetics and biology of KSPGs in the past decade, the mechanisms that control their tissue-speci¢c gene expression remain poorly understood. One possible explanation is that in vitro studies have been hampered by the ¢nding that stromal cells (keratocytes), producing KSPGs in the cornea, generally cease their production in vitro [10,11]. Recently, it has been reported that the native biosynthetic phenotype of keratocytes can be maintained for a few days in low serum and serum-free media [11]. These ¢ndings suggest a unique and highly regulated transcriptional mechanism. At present, there are no data on the
0167-4781 / 00 / $ ^ see front matter ß 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 7 8 1 ( 0 0 ) 0 0 1 2 9 - 9
BBAEXP 91424 3-7-00
E.S. Tasheva et al. / Biochimica et Biophysica Acta 1492 (2000) 452^459
453
Fig. 1. Nucleotide sequence of the proximal promoter region of the human Kera gene. The numbers to the right indicate the nucleotide position. Evolutionarily conserved homologous regions in the bovine Kera promoter are underlined and mismatches are indicated. The tsps are indicated by arrowheads and an asterisk. Putative binding sites for transcription factors discussed in the text are overlined and speci¢ed. The positions of restriction enzymes and primers used for generation of promoter/reporter constructs and primer extensions are overlined and shown in italics.
molecular mechanisms regulating the tissue-speci¢c expression of the Kera gene. In this study, we isolated the human Kera 5P-£anking region and characterized the promoter structure. Using a set of reporter gene constructs and transient transfection into corneal keratocytes, we identi¢ed positive and negative regulatory sequences that are conserved in the promoters of human and bovine Kera genes. Using DNaseI footprinting and electrophoretic mobility shift assay (EMSA), we showed that multiple cell type-speci¢c protein factors bind to these sequences. Given the importance of
KSPGs in providing corneal transparency, the results of this work provide novel information about Kera transcriptional regulation and thereby might suggest possible targets for therapeutic upregulation of this gene in pathological situations. 2. Comparison of the human and bovine Kera 5PP-£anking sequences reveals evolutionarily conserved regions Previously, we isolated genomic clones and determined
BBAEXP 91424 3-7-00
454
E.S. Tasheva et al. / Biochimica et Biophysica Acta 1492 (2000) 452^459
Kera gene. Part of this sequence (1.4 kb) is shown in Fig. 1. Using the Blast 2 Sequences Against Each Other program, we compared 2 kb of human and bovine Kera 5P-£anking sequences and identi¢ed three regions of signi¢cant evolutionary conservation : a 450 bp sequence of the ¢rst exon, the region between 3343 and 3249, and the region between 31244 and 31061 (Fig. 1). The order and spacing of the conserved sequences are also maintained. Analysis of these sequences reveals a number of potential binding sites for transcription factors known to enhance or suppress transcription in a cell type-speci¢c manner in the promoters of other genes. As shown in Fig. 1, the proximal conserved region contains mainly enhancer motifs, such as the SV40 enhancer core element and the T-cell receptor alpha enhancer [13]. The distal conserved region contains consensus sequences for silencers, such as the CAGA motif, essential for repression of the mouse alpha 2(I) collagen promoter [14], Pit1, a negative regulator of prolactin gene expression in non-pituitary cells [15], and a CTC factor binding motif [16]. A CArG-box at position 3851 is also conserved between the two species. 3. Identi¢cation of human Kera transcriptional start sites
Fig. 2. Analysis of transcription start sites of human Kera mRNA by primer extension and RPA assays. An outline of primer extension and RPA experiments is shown at the top; a: RPA using a 870 bp radiolabeled antisense RNA probe and total RNA from freshly isolated human corneal keratocytes. The RNA probe for RPA was synthesized from pGEM-T clones containing the 870 bp Kera DNA fragment that was generated by PCR. Primer extension and RPA were performed as described [18]. b: RNA was synthesized in in vitro transcription (IVT) reactions using a p(3770/+496) promoter^reporter construct and the promoterless pGL3-Basic plasmid (for negative control) as templates, and nuclear extract from primary bovine keratocytes prepared according to the method of Dignam [19]. IVT was carried out in a 25 Wl volume of reaction mixture containing 25 mM HEPES (pH 7.9), 10 mM KCl, 6 mM MgCl2 , 1 mM dithiothreitol (DTT), 10% (v/v) glycerol, 600 WM of each rNTP, 0.5^1 Wg DNA template and 10 Wl corneal nuclear extract. After a 45 min incubation at 30³C, 2.5 Wl RQ1 RNase-free DNaseI was added, incubation was continued for an additional 30 min at 37³C and RNA was isolated. Primer extension was performed with [Q-32 P]ATP end-labeled GSP(3385). Arrowheads indicate the protected bands that correspond to the major tsps. Asterisk indicates a possible tsp band(s) discussed in the text. M, size markers.
the nucleotide sequence of the entire human and bovine Kera genes [8,12]. In the present study, we used a polymerase chain reaction (PCR)-based method to obtain and sequence 4.1 kb of the 5P-£anking region of the human
We ¢rst mapped the transcription start point(s) (tsp(s)) by ribonuclease protection assay (RPA) assay. Total RNA prepared from freshly isolated human corneal keratocytes hybridized to the 870 bp RNA probe depicted at the top of Fig. 2 produced two major and several minor protected fragments (Fig. 2a). The protected band at position +220 corresponds to the previously determined Kera cDNA end in human and bovine cells, using the 5P-rapid ampli¢cation of cDNA ends (RACE) technique [8,12], whereas the band at +1 represents a newly identi¢ed human Kera tsp. The minor band(s) indicated with an asterisk in Fig. 2a(position +316 in the human sequence) corresponds in size to the ¢rst exon of the mouse Kera gene as determined by RACE [6]. Repeated experiments using di¡erent RNA probes gave the same results. These data indicate that multiple tsps exist. To con¢rm this conclusion, we performed primer extension using gene-speci¢c primer (GSP)(3385) and in vitro synthesized human Kera RNA (Fig. 2b). The DNA fragment spanning nucleotides 3770 to +496 of the human Kera gene, cloned into the pGL3Basic vector (Promega), served as template for this synthesis. Two major tsps, at positions corresponding to +1 and +220, were also obtained in these experiments (Fig. 2b), indicating that these two tsps are utilized in an in vitro system as well. These two major tsps were subsequently con¢rmed by functional assays using reporter gene constructs and transient transfection (see below). Inspection of the nucleotide sequence reveals the following: (i) functionally optimal initiator element (Inr) consensus sequences, recently identi¢ed by saturation mutagenesis as CA+1(G/T)T [17], occur at +1 and +220 tsps (Fig. 1,
BBAEXP 91424 3-7-00
E.S. Tasheva et al. / Biochimica et Biophysica Acta 1492 (2000) 452^459
455
Fig. 3. E¡ect of upstream deletions, truncations and mutation on the promoter activity of the human Kera gene. A: Schematic representation of the 5P£anking region of the human Kera gene, with positions of restriction enzymes, GSPs and regulatory motifs shown at the top. Constructs used to test the functional activity of the Kera promoter in transient transfection assays are shown to the left. Luciferase expression assays are shown to the right. Primary bovine corneal keratocytes were isolated as described [20] and incubated at 2U104 cells/cm2 in 6-well cluster plates (Costar Corporation) in Dulbecco's modi¢ed Eagle's medium/nutrient mixture F-12 HAM (Sigma, catalog no. D-2906, without phenol red), supplemented with 1% (v/v) platelet-poor horse serum (Sigma) and antibiotics (100 Wg/ml each penicillin and streptomycin, 50 Wg/ml gentamicin and 2.5 Wg/ml amphotericin B). After 24 h, the cells were transiently transfected using TransFast transfection reagent (Promega) according to the standard protocol (9 Wl reagent per 3 Wg DNA). Transfections with the pGL3-Basic plasmid were used for background determination, and transfections with the pGL3-Control plasmid were used as positive controls. Co-transfections with 1 Wg pSV LGal plasmid (Clontech) were performed in all experiments to correct for transfection e¤ciency. The Steady Glo luciferase assay system (Promega) for ¢re£y luciferase activity and Galacto-Star (Tropix) for L-galactosidase activity were used according to the manufacturer's protocol. Luciferase activity was assayed using a Packard Top Count.NXT microplate scintillation and luminescence counter. The results are reported as the means þ S.E.M. from four separate transfections performed in duplicate. B: The e¡ect of the CNRR on a heterologous SV40 promoter. Results are means þ S.E.M. of three independent transfections.
arrows); (ii) a TATA-like motif (CATAAA in human and TATA-box in bovine) also can be found 30 bp upstream of the +220 tsp; and (iii) only the Inr's canonical CA+1 start site sequence is present at the third tsp at position +316.
4. Reporter gene analysis of human Kera promoter deletion constructs identi¢es positive and negative regulatory regions The pGL3 series of luciferase reporter gene plasmids were used for human Kera promoter analysis. The 4.1 kb Kera DNA fragment, containing exon 1 and 3.5 kb of the 5P-£anking region, was ampli¢ed by PCR from
BBAEXP 91424 3-7-00
456
E.S. Tasheva et al. / Biochimica et Biophysica Acta 1492 (2000) 452^459
the pGEM-T genomic clone using primers GSP(3496) and GSP3690 (5P-TAAAAGCTTGTGATATGCATATAGGG-3P). The PCR product was ligated into a SmaI site of the pGL3-Basic vector. Orientation of the cloned fragment
was determined by restriction endonuclease mapping and sequence analysis using RVprimer3 (Promega). The resulting p(33690/+496) clone was used to make deletions of the inserted DNA fragment using the restriction enzymes
BBAEXP 91424 3-7-00
E.S. Tasheva et al. / Biochimica et Biophysica Acta 1492 (2000) 452^459
457
Fig. 4. DNaseI footprinting and EMSA analyses of CPRR (A) and CNRR (B) of the human Kera promoter. For DNaseI footprinting, the DNA fragments spanning nucleotides ^374 to +238 and ^1276 to ^945 were obtained by PCR. Labeled DNA fragments were incubated with HeLa (Promega) or bovine corneal keratocyte nuclear extracts. DNaseI footprinting reactions were carried out using the core footprinting system (Promega) according to the manufacturer's protocol, except that the amount of diluted solution of DNaseI was increased to 15 Wl per reaction. The DNA samples were resolved on a 6% polyacrylamide gel containing 8 M urea. Oligonucleotides for EMSA were synthesized by Integrated DNA Technologies Inc. and Gibco BRL Life Technologies, and annealed to generate double-stranded DNA probes. These are as follows : for the CPRR probe pF1/F2 encompassing nucleotides 3294 to 3272 and probe F3 encompassing nucleotides 3271 to 3248; for the CNRR probe dF1/F2 (nucleotides 31187 to 31153), probe dF3a (nucleotides -1131 to -1103), probe dF3b (nucleotides 31105 to 31071), probe dF4 (nucleotides 31071 to 31040) and probe dF5 (nucleotides 31047 to 31021). Double-stranded DNA probes were 5P-end-labeled with [Q-32 P]ATP and T4 polynucleotide kinase. Nuclear extract (5 Wg) was incubated with 25 000 cpm of labeled probe in a 20 Wl volume of reaction bu¡er containing 50 mM NaCl, 10 mM Tris^HCl (pH 7.5), 1 mM MgCl2 , 0.5 mM EDTA, 0.5 mM DTT, 10% (v/v) glycerol and 2 Wg poly(dI-dC)-poly(dI-dC) (Amersham Pharmacia Biotech). Binding reactions were performed at room temperature for 25 min and resolved on a 4% non-denaturing acrylamide gel. The sequences of oligonucleotides used in competition experiments were: CAGA oligo, 5P-ATGCAGATCATCTGTTG-3P, containing the two CAGA motifs as read in their natural position without the rest of the sequence of the probe dF3b; and CTC oligo, 5P-TCCCTCCTGAGGGTTAG-3P, containing two CTCF motifs. AP1 and NF-kB oligonucleotides were obtained from Promega. A.1: footprint from the CPRR; F, free DNA; M, DNA size markers indicated on the left. The footprinted areas (F1^F3) are indicated on the right. A.2: Radiolabeled oligonucleotides corresponding to the footprinted sites from CPRR are indicated at the top. For lanes 2, 3, 7 and 8, incubations were performed with corneal (C) extract. For lanes 4, 5, 9 and 10, incubations were performed with HepG2 liver (L) extract (Geneka). Arrows indicate speci¢c DNA^protein complexes. Unlabeled oligonucleotides added to the reactions as competitors are indicated at the bottom with (+) and con¢rm the tissue speci¢city of the formed complexes. B.1: Footprint from the CNRR ; F, free DNA; M, DNA size markers indicated on the left. The footprinted areas (F1^F5) are indicated on the right. B.2: Radiolabeled oligonucleotides corresponding to the footprinted sites from CNRR (indicated at the top) were incubated with nuclear extracts from bovine cornea (C) (lanes 2, 2a, 6, 7, 8, 13, 14 and 15), HeLa (H) cells (lanes 3, 17 and 18) or human HepG2 (L) cells (lanes 4, 9, 10, 11 and 16). Lanes 1, 5 and 12 were not supplemented with nuclear extracts. Unlabeled double-stranded oligonucleotides added to the reactions in lanes 7, 8, 10, 11, 14, 15 and 18 as competitors are indicated at the bottom. Arrows mark speci¢c DNA^protein complexes. Lane 2a is a longer exposure of lane 2. B.3: Radiolabeled oligonucleotides corresponding to the footprinted sites from CNRR and radiolabeled NF-kB and AP1 oligonucleotides used as controls are indicated on the top. Unlabeled oligonucleotides added to the reactions as competitors are indicated at the bottom with (+). For lanes 1, 2, 3, 4, 13, 14 and 15, incubations were performed with corneal (C) extract. For lanes 5, 6, 7, 8, 16, 17, 18 and 19, incubations were performed with HepG2 (L) extract. For lanes 9, 10, 11, 20 and 21, incubations were performed with HeLa (H) extract. Arrows indicate speci¢c DNA^protein complexes and asterisk indicates the non-speci¢c complex.
6
and GSPs shown in Figs. 1 and 3A, top. Intact, deleted, truncated, mutated and reversed orientation DNA fragments were constructed as shown in Fig. 3A, middle, left. These constructs were transfected into primary bovine keratocytes cultured in 1% (v/v) platelet-poor horse serum (Sigma), conditions in which KSPG production can be maintained for several days ([11] and our unpublished data). To test functionally the two major tsps, we constructed the reporter plasmids p(+163/+496) that contained only the tsps at +220 and at +316, p(331/+163) that contained only the tsp at +1, and p(331/+496) that contained all three tsps. As shown in Fig. 3A, middle, left, the construct with only the newly identi¢ed +1 tsp showed promoter activity, indicating that this site is utilized in in vitro assays. The construct containing all three tsps showed the highest promoter activity. These data are consistent with the ¢ndings in the initial primer extensions and RPA experiments. Further extension of the promoter sequence revealed that the regions between nucleotides 3446 to 3208 and 3770 to 3556 enhance the reporter activity, while the regions between nucleotides 3891 to 3770 and 31890 to 3891 had a negative e¡ect on reporter activity. Three of these DNA fragments were subjected to further analysis, since they contained the conserved sequences (labeled as conserved positive regulatory region (CPRR) and conserved negative regulatory region (CNRR) in Fig. 3A) and the conserved CArG-box. The functions of CPRR and CNRR in corneal keratocytes were tested by generation of spliced constructs
p(CPRR331/+496) and p(CNRR331/+496), in which the CPRR and CNRR sequences were directly adjoined to the reporter construct p(331/+496). Transfection experiments showed that the CPRR exhibits enhancer activity whereas the CNRR exhibits silencer activity compared to the p(331/+496) construct alone in cultured keratocytes. The importance of the CArG-box was demonstrated by mutating its sequence from CCAAATAAGG to GGAAACAAGG. As shown by transient expression with p(3891/+496CArGmut), this mutation caused about a 3-fold increase of reporter activity compared to the p(331/+496) construct. These results indicate that the CArG-box is important for the suppression of Kera gene expression in cultured keratocytes. In addition, the CNRR was cloned also into a pGL3-Promoter vector (Promega) that contains the heterologous SV40 promoter (Fig. 3B). When this construct was transfected into corneal keratocytes and four other non-corneally derived cell lines, the human Kera CNRR had a negative regulatory e¡ect on the SV40 promoter in all cells. 5. DNaseI footprinting analyses and EMSA identify cell type-speci¢c factors that bind to the conserved regions CPRR and CNRR were further analyzed by DNaseI footprinting and EMSA. The results of these experiments are shown in Fig. 4. Three footprints were obtained with the CPRRs (Fig. 4A.1, F1^F3) that appear identical be-
BBAEXP 91424 3-7-00
458
E.S. Tasheva et al. / Biochimica et Biophysica Acta 1492 (2000) 452^459
tween corneal and HeLa extracts. However, when the footprints were analyzed by EMSA, di¡erent speci¢c complexes were formed in corneal compared to liver nuclear extracts (Fig. 4A.2, lanes 2 compared to 4, and 7 compared to 9). These results demonstrate that tissue-speci¢c factors bind to the footprinted regions. Five footprints were obtained with the CNRR (Fig. 4B.1, F1^F5). The F1 region contains the HLA-DQbeta gene octamer motif; F2 contains a binding site for Pit1 ; F3 contains a binding site for CTC factor and two CAGA motifs ; F4 contains two CTC factor binding motifs and overlapping PEA3/ NF-kB binding sites; F5 contains an AP1 transcription factors binding site (see also the region between nucleotides 31241 and 31011 in Fig. 1). EMSA analyses of these footprints demonstrate many tissue di¡erences. (a) Cornea-, liver- and HeLa-speci¢c nuclear complexes form with dF1/F2 (Fig. 4B.2, lanes 1^4 and 2a), dF3 (Fig. 4B.2, lanes 6, 9, 13, 16 and 17), dF4 (Fig. 4B.3, lanes 2, 5 and 9) and dF5 (Fig. 4B.3, lanes 13, 16 and 21). (b) Proteins found in cornea and liver nuclear extracts bind to GAGGG (in the sense strand), the binding site of CTC factor, as judged by the competition experiments when labeled dF3a (Fig. 4B.2, lanes 5^11) and labeled dF4 (Fig. 4B.3, lanes 1^4) oligonucleotides were competed with unlabeled CTC oligo. (c) A cornea-speci¢c protein binds to the CAGA motifs, as judged by the competition experiments using labeled dF3b oligonucleotide probe and unlabeled CAGA oligo as competitor (Fig. 4B.2, lanes 12^ 15). (d) Non-corneal protein, most likely a member of the NF-kB/Rel family, participates in complex formation at the dF4 site (Fig. 4B.3, lanes 5^11). As shown, when NF-kB oligonucleotide at 50-fold excess was added, the intensity of the fast mobility complex in liver extract diminished and a new slow migrating complex was formed (lane 6). Increasing the amount of competitor further to 100-fold excess increased the intensity of this new complex and diminished the intensity of the fast complex (lane 7). These results can be explained if two or multiple proteins were involved in the complex formation. The presence of overlapping binding sites for three DNA binding proteins in this region supports this conclusion. (e) Liver-speci¢c proteins bind to the AP1 motif in dF5 (Fig. 4B.3, lanes 15 and 17). Acknowledgements We thank Dr. Ron Walkenbach and Tony Bavuso for the human eyes, members of the Dr. Terry C. Johnson laboratory for supplying the bovine eyes, Heideh Fattaey for kindly providing HB and NIH3T3 cell lines, and Keith Woods for the HCT-8 cell line. This work was supported by NIH Grant EY00952 to G.W.C.
References [1] J.L. Funderburgh, M.L. Funderburgh, M.M. Mann, G.W. Conrad, Physical and biological properties of keratan sulphate proteoglycan, Biochem. Soc. Trans. 19 (1991) 871^876. [2] L.M. Corpuz, J.L. Funderburgh, M.L. Funderburgh, G.S. Bottomley, S. Prakash, G.W. Conrad, Molecular cloning and tissue distribution of keratocan, bovine corneal keratan sulfate proteoglycan 37A, J. Biol. Chem. 271 (1996) 9759^9763. [3] J.L. Funderburgh, L.M. Corpuz, M.R. Roth, M.L. Funderburgh, E.S. Tasheva, G.W. Conrad, Mimecan, the 25 kDa corneal keratan sulfate proteoglycan, is a product of the gene producing osteoglycin, J. Biol. Chem. 272 (1997) 28089^28095. [4] J.R. Hassell, C. Cintron, C. Kublin, D.A. Newsome, Proteoglycan changes during restoration of transparency in corneal scars, Arch. Biochem. Biophys. 222 (1983) 362^369. [5] K. Klintworth, C.F. Smith, Abnormalities of proteoglycans and glycoproteins synthesized by corneal organ cultures derived from patients with macular corneal dystrophy, Lab. Invest. 438 (1983) 603^ 612. [6] C.Y. Liu, A. Shiraishi, C.W. Kao, R.L. Converse, J.L. Funderburgh, L.M. Corpuz, G.W. Conrad, W.W. Kao, The cloning of mouse keratocan cDNA and genomic DNA and the characterization of its expression during eye development, J. Biol. Chem. 273 (1998) 22584. [7] J.R. Dunlevy, P.J. Neame, J.-P. Vergnes, J.R. Hassell, Identi¢cation of the N-linked oligosaccharide sites in chick corneal lumican and keratocan that receive keratan sulfate, J. Biol. Chem. 273 (1998) 9615^9621. [8] E.S. Tasheva, J.L. Funderburgh, M.L. Funderburgh, L.M. Corpuz, G.W. Conrad, Structure and sequence of the gene encoding human keratocan, DNA Seq. 10 (1999) 67^74. [9] S. Chakravarti, T. Magnuson, J.H. Lass, K.J. Jepsen, C. LaMantia, H. Carroll, Lumican regulates collagen ¢bril assembly : Skin fragility and corneal opacity in the absence of lumican, J. Cell Biol. 141 (1998) 1277^1286. [10] G.W. Conrad, A. Dorfman, Synthesis of sulfated mucopolysaccharides by chick corneal ¢broblasts in vitro, Exp. Eye Res. 18 (1974) 421^433. [11] M.P. Beales, J.L. Funderburgh, J.V. Jester, J.R. Hassell, Proteoglycan synthesis by bovine keratocytes and corneal ¢broblasts: maintenance of the keratocyte phenotype in culture, Invest. Ophthalmol. Vis. Sci. 40 (1999) 1658. [12] E.S. Tasheva, J.L. Funderburgh, L.M. Corpuz, G.W. Conrad, Cloning, characterization and tissue-speci¢c expression of the gene encoding bovine keratocan, a corneal keratan sulfate proteoglycan, Gene 218 (1998) 63^68. [13] A. Travis, A. Amsterdam, C. Belanger, R. Grosschedl, LEF-1, a gene encoding a lymphoid-speci¢c protein with an HMG domain, regulates T-cell receptor alpha enhancer function, Genes Dev. 5 (1991) 880^894. [14] G. Karsenty, P. Golumbek, B. de Crombrugghe, Point mutations and small substitution mutations in three di¡erent upstream elements inhibit the activity of the mouse alpha 2(I) collagen promoter, J. Biol. Chem. 263 (1988) 13909^13915. [15] Z.X. Zhang, V. Kumar, R.T. Rivera, J. Chisholm, D.K. Biswas, cisacting negative regulatory element of prolactin gene, J. Biol. Chem. 265 (1990) 4785^4788. [16] M. Burcin, R. Amold, M. Lutz, B. Kaiser, D. Runge, F. Lottspeich, G.N. Filippova, V.V. Lobanenkov, R. Renkawitz, Negative protein 1, which is required for function of the chicken lysozyme gene silencer in conjunction with hormone receptors, is identical to the multivalent zinc ¢nger repressor CTCF, Mol. Cell Biol. 17 (1997) 1281^1288. [17] R.J. Kraus, E.E. Murray, S.R. Wiley, N.M. Zink, K. Loritz, G.W.
BBAEXP 91424 3-7-00
E.S. Tasheva et al. / Biochimica et Biophysica Acta 1492 (2000) 452^459 Gelembiuk, J.E. Mertz, Experimentally determined weight matrix de¢nitions of the initiator and TBP binding site elements of promoters, Nucleic Acids Res. 24 (1996) 1531^1539. [18] E.S. Tasheva, M.L. Funderburgh, J. McReynolds, J.L. Funderburgh, G.W. Conrad, The bovine mimecan gene. Molecular cloning and characterization of two major RNA transcripts generated by alternative use of two splice acceptor sites in the third exon, J. Biol. Chem. 274 (1999) 18693^18701.
459
[19] J.D. Dignam, R.M. Lebovitz, R.G. Roeder, Accurate transcription initiation by RNA polymerase II in a soluble extract from isolated mammalian nuclei, Nucleic Acids Res. 11 (1983) 1475^1489. [20] J.L. Funderburgh, M.L. Funderburgh, M.M. Mann, S. Prakash, G.W. Conrad, Synthesis of corneal keratan sulfate proteoglycans by bovine keratocytes in vitro, J. Biol. Chem. 271 (1996) 31431^ 31436.
BBAEXP 91424 3-7-00
www.elsevier.com/locate/bba
Promoter paper
Identi¢cation and characterization of conserved cis-regulatory elements in the human keratocan gene promoter1 Elena S. Tasheva *, Abigail H. Conrad, Gary W. Conrad Division of Biology, Ackert Hall, Kansas State University, Manhattan, KS 66506-4901, USA Received 20 March 2000; accepted 10 May 2000
Abstract Keratocan, along with lumican and mimecan, represent the keratan sulfate-containing proteoglycans of the vertebrate cornea that play a key role in development and maintenance of corneal transparency. In this study, we cloned 4.1 kb of the human Kera 5P-flanking region and characterized the promoter structure. Using primer extension and ribonuclease protection assay, we identify two major transcriptional start sites in the first exon. Using luciferase reporter gene transfection analysis of 5P-deletion and mutation constructs, we demonstrate positive and negative regulatory elements within a 1.3 kb upstream sequence. Comparison of human and bovine 5P-flanking sequences reveals three highly conserved regions: a 450 bp region in the first exon, a 92 bp promoter proximal conserved regulatory region identified as an enhancer in the natural context, and a 223 bp promoter distal conserved regulatory region identified as a silencer both in the natural context and in a heterologous promoter system. In addition, a conserved CArG-box residing 851 bp upstream of the first transcription start site also can lead to the repression of Kera expression in cultured corneal keratocytes. DNaseI footprinting and electrophoretic mobility shift assay demonstrate that cell type-specific factors bind to regulatory elements located in the conserved regions. Competition experiments indicate that the CTC factor and a protein that binds to the CAGA motif are likely to be among the multiple factors involved in the transcriptional regulation of the human Kera gene. ß 2000 Elsevier Science B.V. All rights reserved. Keywords : Promoter region ; Keratan sulfate proteoglycan ; Gene regulation; Corneal keratocyte
1. Introduction Keratocan, lumican and mimecan are keratan sulfate proteoglycans (KSPGs) expressed at high levels in the cornea, where the proteins are synthesized covalently associated with large, sulfated keratan sulfate chains [1]. Kera is also abundant in sclera and detectable in skin, ligament and skeletal muscle. However, similarly to lumican and mimecan, in these non-corneal tissues keratocan is present as a non-sulfated glycoprotein [2,3]. The high level of expression of KSPGs in cornea, combined with their unique Abbreviations : CPRR, conserved positive regulatory region; CNRR, conserved negative regulatory region; EMSA, electrophoretic mobility shift assay; GSP, gene-speci¢c primer; Inr, initiator element; Kera, gene encoding keratocan; KSPG, keratan sulfate proteoglycan ; PCR, polymerase chain reaction ; RACE, rapid ampli¢cation of cDNA ends; RPA, ribonuclease protection assay; tsp, transcription start point * Corresponding author. Fax: +1-785-532-6653; E-mail : [email protected] 1 The nucleotide sequence data reported here will appear in the GenBank nucleotide sequence database under the accession number: AF169962.
glycosylation, suggests that these molecules are important for the development and maintenance of corneal transparency. This conclusion is supported by several lines of evidence. Lack of sulfation of corneal KSPGs has been reported in macular corneal dystrophy type I and in opaque corneal scars [4,5]. The genomic organization and the protein structure are highly conserved between distant species, such as chicken, mouse and man [6^8]. Disruption of the lumican gene in knockout mice causes bilateral corneal opacity [9]. Despite considerable progress made in understanding the genetics and biology of KSPGs in the past decade, the mechanisms that control their tissue-speci¢c gene expression remain poorly understood. One possible explanation is that in vitro studies have been hampered by the ¢nding that stromal cells (keratocytes), producing KSPGs in the cornea, generally cease their production in vitro [10,11]. Recently, it has been reported that the native biosynthetic phenotype of keratocytes can be maintained for a few days in low serum and serum-free media [11]. These ¢ndings suggest a unique and highly regulated transcriptional mechanism. At present, there are no data on the
0167-4781 / 00 / $ ^ see front matter ß 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 7 8 1 ( 0 0 ) 0 0 1 2 9 - 9
BBAEXP 91424 3-7-00
E.S. Tasheva et al. / Biochimica et Biophysica Acta 1492 (2000) 452^459
453
Fig. 1. Nucleotide sequence of the proximal promoter region of the human Kera gene. The numbers to the right indicate the nucleotide position. Evolutionarily conserved homologous regions in the bovine Kera promoter are underlined and mismatches are indicated. The tsps are indicated by arrowheads and an asterisk. Putative binding sites for transcription factors discussed in the text are overlined and speci¢ed. The positions of restriction enzymes and primers used for generation of promoter/reporter constructs and primer extensions are overlined and shown in italics.
molecular mechanisms regulating the tissue-speci¢c expression of the Kera gene. In this study, we isolated the human Kera 5P-£anking region and characterized the promoter structure. Using a set of reporter gene constructs and transient transfection into corneal keratocytes, we identi¢ed positive and negative regulatory sequences that are conserved in the promoters of human and bovine Kera genes. Using DNaseI footprinting and electrophoretic mobility shift assay (EMSA), we showed that multiple cell type-speci¢c protein factors bind to these sequences. Given the importance of
KSPGs in providing corneal transparency, the results of this work provide novel information about Kera transcriptional regulation and thereby might suggest possible targets for therapeutic upregulation of this gene in pathological situations. 2. Comparison of the human and bovine Kera 5PP-£anking sequences reveals evolutionarily conserved regions Previously, we isolated genomic clones and determined
BBAEXP 91424 3-7-00
454
E.S. Tasheva et al. / Biochimica et Biophysica Acta 1492 (2000) 452^459
Kera gene. Part of this sequence (1.4 kb) is shown in Fig. 1. Using the Blast 2 Sequences Against Each Other program, we compared 2 kb of human and bovine Kera 5P-£anking sequences and identi¢ed three regions of signi¢cant evolutionary conservation : a 450 bp sequence of the ¢rst exon, the region between 3343 and 3249, and the region between 31244 and 31061 (Fig. 1). The order and spacing of the conserved sequences are also maintained. Analysis of these sequences reveals a number of potential binding sites for transcription factors known to enhance or suppress transcription in a cell type-speci¢c manner in the promoters of other genes. As shown in Fig. 1, the proximal conserved region contains mainly enhancer motifs, such as the SV40 enhancer core element and the T-cell receptor alpha enhancer [13]. The distal conserved region contains consensus sequences for silencers, such as the CAGA motif, essential for repression of the mouse alpha 2(I) collagen promoter [14], Pit1, a negative regulator of prolactin gene expression in non-pituitary cells [15], and a CTC factor binding motif [16]. A CArG-box at position 3851 is also conserved between the two species. 3. Identi¢cation of human Kera transcriptional start sites
Fig. 2. Analysis of transcription start sites of human Kera mRNA by primer extension and RPA assays. An outline of primer extension and RPA experiments is shown at the top; a: RPA using a 870 bp radiolabeled antisense RNA probe and total RNA from freshly isolated human corneal keratocytes. The RNA probe for RPA was synthesized from pGEM-T clones containing the 870 bp Kera DNA fragment that was generated by PCR. Primer extension and RPA were performed as described [18]. b: RNA was synthesized in in vitro transcription (IVT) reactions using a p(3770/+496) promoter^reporter construct and the promoterless pGL3-Basic plasmid (for negative control) as templates, and nuclear extract from primary bovine keratocytes prepared according to the method of Dignam [19]. IVT was carried out in a 25 Wl volume of reaction mixture containing 25 mM HEPES (pH 7.9), 10 mM KCl, 6 mM MgCl2 , 1 mM dithiothreitol (DTT), 10% (v/v) glycerol, 600 WM of each rNTP, 0.5^1 Wg DNA template and 10 Wl corneal nuclear extract. After a 45 min incubation at 30³C, 2.5 Wl RQ1 RNase-free DNaseI was added, incubation was continued for an additional 30 min at 37³C and RNA was isolated. Primer extension was performed with [Q-32 P]ATP end-labeled GSP(3385). Arrowheads indicate the protected bands that correspond to the major tsps. Asterisk indicates a possible tsp band(s) discussed in the text. M, size markers.
the nucleotide sequence of the entire human and bovine Kera genes [8,12]. In the present study, we used a polymerase chain reaction (PCR)-based method to obtain and sequence 4.1 kb of the 5P-£anking region of the human
We ¢rst mapped the transcription start point(s) (tsp(s)) by ribonuclease protection assay (RPA) assay. Total RNA prepared from freshly isolated human corneal keratocytes hybridized to the 870 bp RNA probe depicted at the top of Fig. 2 produced two major and several minor protected fragments (Fig. 2a). The protected band at position +220 corresponds to the previously determined Kera cDNA end in human and bovine cells, using the 5P-rapid ampli¢cation of cDNA ends (RACE) technique [8,12], whereas the band at +1 represents a newly identi¢ed human Kera tsp. The minor band(s) indicated with an asterisk in Fig. 2a(position +316 in the human sequence) corresponds in size to the ¢rst exon of the mouse Kera gene as determined by RACE [6]. Repeated experiments using di¡erent RNA probes gave the same results. These data indicate that multiple tsps exist. To con¢rm this conclusion, we performed primer extension using gene-speci¢c primer (GSP)(3385) and in vitro synthesized human Kera RNA (Fig. 2b). The DNA fragment spanning nucleotides 3770 to +496 of the human Kera gene, cloned into the pGL3Basic vector (Promega), served as template for this synthesis. Two major tsps, at positions corresponding to +1 and +220, were also obtained in these experiments (Fig. 2b), indicating that these two tsps are utilized in an in vitro system as well. These two major tsps were subsequently con¢rmed by functional assays using reporter gene constructs and transient transfection (see below). Inspection of the nucleotide sequence reveals the following: (i) functionally optimal initiator element (Inr) consensus sequences, recently identi¢ed by saturation mutagenesis as CA+1(G/T)T [17], occur at +1 and +220 tsps (Fig. 1,
BBAEXP 91424 3-7-00
E.S. Tasheva et al. / Biochimica et Biophysica Acta 1492 (2000) 452^459
455
Fig. 3. E¡ect of upstream deletions, truncations and mutation on the promoter activity of the human Kera gene. A: Schematic representation of the 5P£anking region of the human Kera gene, with positions of restriction enzymes, GSPs and regulatory motifs shown at the top. Constructs used to test the functional activity of the Kera promoter in transient transfection assays are shown to the left. Luciferase expression assays are shown to the right. Primary bovine corneal keratocytes were isolated as described [20] and incubated at 2U104 cells/cm2 in 6-well cluster plates (Costar Corporation) in Dulbecco's modi¢ed Eagle's medium/nutrient mixture F-12 HAM (Sigma, catalog no. D-2906, without phenol red), supplemented with 1% (v/v) platelet-poor horse serum (Sigma) and antibiotics (100 Wg/ml each penicillin and streptomycin, 50 Wg/ml gentamicin and 2.5 Wg/ml amphotericin B). After 24 h, the cells were transiently transfected using TransFast transfection reagent (Promega) according to the standard protocol (9 Wl reagent per 3 Wg DNA). Transfections with the pGL3-Basic plasmid were used for background determination, and transfections with the pGL3-Control plasmid were used as positive controls. Co-transfections with 1 Wg pSV LGal plasmid (Clontech) were performed in all experiments to correct for transfection e¤ciency. The Steady Glo luciferase assay system (Promega) for ¢re£y luciferase activity and Galacto-Star (Tropix) for L-galactosidase activity were used according to the manufacturer's protocol. Luciferase activity was assayed using a Packard Top Count.NXT microplate scintillation and luminescence counter. The results are reported as the means þ S.E.M. from four separate transfections performed in duplicate. B: The e¡ect of the CNRR on a heterologous SV40 promoter. Results are means þ S.E.M. of three independent transfections.
arrows); (ii) a TATA-like motif (CATAAA in human and TATA-box in bovine) also can be found 30 bp upstream of the +220 tsp; and (iii) only the Inr's canonical CA+1 start site sequence is present at the third tsp at position +316.
4. Reporter gene analysis of human Kera promoter deletion constructs identi¢es positive and negative regulatory regions The pGL3 series of luciferase reporter gene plasmids were used for human Kera promoter analysis. The 4.1 kb Kera DNA fragment, containing exon 1 and 3.5 kb of the 5P-£anking region, was ampli¢ed by PCR from
BBAEXP 91424 3-7-00
456
E.S. Tasheva et al. / Biochimica et Biophysica Acta 1492 (2000) 452^459
the pGEM-T genomic clone using primers GSP(3496) and GSP3690 (5P-TAAAAGCTTGTGATATGCATATAGGG-3P). The PCR product was ligated into a SmaI site of the pGL3-Basic vector. Orientation of the cloned fragment
was determined by restriction endonuclease mapping and sequence analysis using RVprimer3 (Promega). The resulting p(33690/+496) clone was used to make deletions of the inserted DNA fragment using the restriction enzymes
BBAEXP 91424 3-7-00
E.S. Tasheva et al. / Biochimica et Biophysica Acta 1492 (2000) 452^459
457
Fig. 4. DNaseI footprinting and EMSA analyses of CPRR (A) and CNRR (B) of the human Kera promoter. For DNaseI footprinting, the DNA fragments spanning nucleotides ^374 to +238 and ^1276 to ^945 were obtained by PCR. Labeled DNA fragments were incubated with HeLa (Promega) or bovine corneal keratocyte nuclear extracts. DNaseI footprinting reactions were carried out using the core footprinting system (Promega) according to the manufacturer's protocol, except that the amount of diluted solution of DNaseI was increased to 15 Wl per reaction. The DNA samples were resolved on a 6% polyacrylamide gel containing 8 M urea. Oligonucleotides for EMSA were synthesized by Integrated DNA Technologies Inc. and Gibco BRL Life Technologies, and annealed to generate double-stranded DNA probes. These are as follows : for the CPRR probe pF1/F2 encompassing nucleotides 3294 to 3272 and probe F3 encompassing nucleotides 3271 to 3248; for the CNRR probe dF1/F2 (nucleotides 31187 to 31153), probe dF3a (nucleotides -1131 to -1103), probe dF3b (nucleotides 31105 to 31071), probe dF4 (nucleotides 31071 to 31040) and probe dF5 (nucleotides 31047 to 31021). Double-stranded DNA probes were 5P-end-labeled with [Q-32 P]ATP and T4 polynucleotide kinase. Nuclear extract (5 Wg) was incubated with 25 000 cpm of labeled probe in a 20 Wl volume of reaction bu¡er containing 50 mM NaCl, 10 mM Tris^HCl (pH 7.5), 1 mM MgCl2 , 0.5 mM EDTA, 0.5 mM DTT, 10% (v/v) glycerol and 2 Wg poly(dI-dC)-poly(dI-dC) (Amersham Pharmacia Biotech). Binding reactions were performed at room temperature for 25 min and resolved on a 4% non-denaturing acrylamide gel. The sequences of oligonucleotides used in competition experiments were: CAGA oligo, 5P-ATGCAGATCATCTGTTG-3P, containing the two CAGA motifs as read in their natural position without the rest of the sequence of the probe dF3b; and CTC oligo, 5P-TCCCTCCTGAGGGTTAG-3P, containing two CTCF motifs. AP1 and NF-kB oligonucleotides were obtained from Promega. A.1: footprint from the CPRR; F, free DNA; M, DNA size markers indicated on the left. The footprinted areas (F1^F3) are indicated on the right. A.2: Radiolabeled oligonucleotides corresponding to the footprinted sites from CPRR are indicated at the top. For lanes 2, 3, 7 and 8, incubations were performed with corneal (C) extract. For lanes 4, 5, 9 and 10, incubations were performed with HepG2 liver (L) extract (Geneka). Arrows indicate speci¢c DNA^protein complexes. Unlabeled oligonucleotides added to the reactions as competitors are indicated at the bottom with (+) and con¢rm the tissue speci¢city of the formed complexes. B.1: Footprint from the CNRR ; F, free DNA; M, DNA size markers indicated on the left. The footprinted areas (F1^F5) are indicated on the right. B.2: Radiolabeled oligonucleotides corresponding to the footprinted sites from CNRR (indicated at the top) were incubated with nuclear extracts from bovine cornea (C) (lanes 2, 2a, 6, 7, 8, 13, 14 and 15), HeLa (H) cells (lanes 3, 17 and 18) or human HepG2 (L) cells (lanes 4, 9, 10, 11 and 16). Lanes 1, 5 and 12 were not supplemented with nuclear extracts. Unlabeled double-stranded oligonucleotides added to the reactions in lanes 7, 8, 10, 11, 14, 15 and 18 as competitors are indicated at the bottom. Arrows mark speci¢c DNA^protein complexes. Lane 2a is a longer exposure of lane 2. B.3: Radiolabeled oligonucleotides corresponding to the footprinted sites from CNRR and radiolabeled NF-kB and AP1 oligonucleotides used as controls are indicated on the top. Unlabeled oligonucleotides added to the reactions as competitors are indicated at the bottom with (+). For lanes 1, 2, 3, 4, 13, 14 and 15, incubations were performed with corneal (C) extract. For lanes 5, 6, 7, 8, 16, 17, 18 and 19, incubations were performed with HepG2 (L) extract. For lanes 9, 10, 11, 20 and 21, incubations were performed with HeLa (H) extract. Arrows indicate speci¢c DNA^protein complexes and asterisk indicates the non-speci¢c complex.
6
and GSPs shown in Figs. 1 and 3A, top. Intact, deleted, truncated, mutated and reversed orientation DNA fragments were constructed as shown in Fig. 3A, middle, left. These constructs were transfected into primary bovine keratocytes cultured in 1% (v/v) platelet-poor horse serum (Sigma), conditions in which KSPG production can be maintained for several days ([11] and our unpublished data). To test functionally the two major tsps, we constructed the reporter plasmids p(+163/+496) that contained only the tsps at +220 and at +316, p(331/+163) that contained only the tsp at +1, and p(331/+496) that contained all three tsps. As shown in Fig. 3A, middle, left, the construct with only the newly identi¢ed +1 tsp showed promoter activity, indicating that this site is utilized in in vitro assays. The construct containing all three tsps showed the highest promoter activity. These data are consistent with the ¢ndings in the initial primer extensions and RPA experiments. Further extension of the promoter sequence revealed that the regions between nucleotides 3446 to 3208 and 3770 to 3556 enhance the reporter activity, while the regions between nucleotides 3891 to 3770 and 31890 to 3891 had a negative e¡ect on reporter activity. Three of these DNA fragments were subjected to further analysis, since they contained the conserved sequences (labeled as conserved positive regulatory region (CPRR) and conserved negative regulatory region (CNRR) in Fig. 3A) and the conserved CArG-box. The functions of CPRR and CNRR in corneal keratocytes were tested by generation of spliced constructs
p(CPRR331/+496) and p(CNRR331/+496), in which the CPRR and CNRR sequences were directly adjoined to the reporter construct p(331/+496). Transfection experiments showed that the CPRR exhibits enhancer activity whereas the CNRR exhibits silencer activity compared to the p(331/+496) construct alone in cultured keratocytes. The importance of the CArG-box was demonstrated by mutating its sequence from CCAAATAAGG to GGAAACAAGG. As shown by transient expression with p(3891/+496CArGmut), this mutation caused about a 3-fold increase of reporter activity compared to the p(331/+496) construct. These results indicate that the CArG-box is important for the suppression of Kera gene expression in cultured keratocytes. In addition, the CNRR was cloned also into a pGL3-Promoter vector (Promega) that contains the heterologous SV40 promoter (Fig. 3B). When this construct was transfected into corneal keratocytes and four other non-corneally derived cell lines, the human Kera CNRR had a negative regulatory e¡ect on the SV40 promoter in all cells. 5. DNaseI footprinting analyses and EMSA identify cell type-speci¢c factors that bind to the conserved regions CPRR and CNRR were further analyzed by DNaseI footprinting and EMSA. The results of these experiments are shown in Fig. 4. Three footprints were obtained with the CPRRs (Fig. 4A.1, F1^F3) that appear identical be-
BBAEXP 91424 3-7-00
458
E.S. Tasheva et al. / Biochimica et Biophysica Acta 1492 (2000) 452^459
tween corneal and HeLa extracts. However, when the footprints were analyzed by EMSA, di¡erent speci¢c complexes were formed in corneal compared to liver nuclear extracts (Fig. 4A.2, lanes 2 compared to 4, and 7 compared to 9). These results demonstrate that tissue-speci¢c factors bind to the footprinted regions. Five footprints were obtained with the CNRR (Fig. 4B.1, F1^F5). The F1 region contains the HLA-DQbeta gene octamer motif; F2 contains a binding site for Pit1 ; F3 contains a binding site for CTC factor and two CAGA motifs ; F4 contains two CTC factor binding motifs and overlapping PEA3/ NF-kB binding sites; F5 contains an AP1 transcription factors binding site (see also the region between nucleotides 31241 and 31011 in Fig. 1). EMSA analyses of these footprints demonstrate many tissue di¡erences. (a) Cornea-, liver- and HeLa-speci¢c nuclear complexes form with dF1/F2 (Fig. 4B.2, lanes 1^4 and 2a), dF3 (Fig. 4B.2, lanes 6, 9, 13, 16 and 17), dF4 (Fig. 4B.3, lanes 2, 5 and 9) and dF5 (Fig. 4B.3, lanes 13, 16 and 21). (b) Proteins found in cornea and liver nuclear extracts bind to GAGGG (in the sense strand), the binding site of CTC factor, as judged by the competition experiments when labeled dF3a (Fig. 4B.2, lanes 5^11) and labeled dF4 (Fig. 4B.3, lanes 1^4) oligonucleotides were competed with unlabeled CTC oligo. (c) A cornea-speci¢c protein binds to the CAGA motifs, as judged by the competition experiments using labeled dF3b oligonucleotide probe and unlabeled CAGA oligo as competitor (Fig. 4B.2, lanes 12^ 15). (d) Non-corneal protein, most likely a member of the NF-kB/Rel family, participates in complex formation at the dF4 site (Fig. 4B.3, lanes 5^11). As shown, when NF-kB oligonucleotide at 50-fold excess was added, the intensity of the fast mobility complex in liver extract diminished and a new slow migrating complex was formed (lane 6). Increasing the amount of competitor further to 100-fold excess increased the intensity of this new complex and diminished the intensity of the fast complex (lane 7). These results can be explained if two or multiple proteins were involved in the complex formation. The presence of overlapping binding sites for three DNA binding proteins in this region supports this conclusion. (e) Liver-speci¢c proteins bind to the AP1 motif in dF5 (Fig. 4B.3, lanes 15 and 17). Acknowledgements We thank Dr. Ron Walkenbach and Tony Bavuso for the human eyes, members of the Dr. Terry C. Johnson laboratory for supplying the bovine eyes, Heideh Fattaey for kindly providing HB and NIH3T3 cell lines, and Keith Woods for the HCT-8 cell line. This work was supported by NIH Grant EY00952 to G.W.C.
References [1] J.L. Funderburgh, M.L. Funderburgh, M.M. Mann, G.W. Conrad, Physical and biological properties of keratan sulphate proteoglycan, Biochem. Soc. Trans. 19 (1991) 871^876. [2] L.M. Corpuz, J.L. Funderburgh, M.L. Funderburgh, G.S. Bottomley, S. Prakash, G.W. Conrad, Molecular cloning and tissue distribution of keratocan, bovine corneal keratan sulfate proteoglycan 37A, J. Biol. Chem. 271 (1996) 9759^9763. [3] J.L. Funderburgh, L.M. Corpuz, M.R. Roth, M.L. Funderburgh, E.S. Tasheva, G.W. Conrad, Mimecan, the 25 kDa corneal keratan sulfate proteoglycan, is a product of the gene producing osteoglycin, J. Biol. Chem. 272 (1997) 28089^28095. [4] J.R. Hassell, C. Cintron, C. Kublin, D.A. Newsome, Proteoglycan changes during restoration of transparency in corneal scars, Arch. Biochem. Biophys. 222 (1983) 362^369. [5] K. Klintworth, C.F. Smith, Abnormalities of proteoglycans and glycoproteins synthesized by corneal organ cultures derived from patients with macular corneal dystrophy, Lab. Invest. 438 (1983) 603^ 612. [6] C.Y. Liu, A. Shiraishi, C.W. Kao, R.L. Converse, J.L. Funderburgh, L.M. Corpuz, G.W. Conrad, W.W. Kao, The cloning of mouse keratocan cDNA and genomic DNA and the characterization of its expression during eye development, J. Biol. Chem. 273 (1998) 22584. [7] J.R. Dunlevy, P.J. Neame, J.-P. Vergnes, J.R. Hassell, Identi¢cation of the N-linked oligosaccharide sites in chick corneal lumican and keratocan that receive keratan sulfate, J. Biol. Chem. 273 (1998) 9615^9621. [8] E.S. Tasheva, J.L. Funderburgh, M.L. Funderburgh, L.M. Corpuz, G.W. Conrad, Structure and sequence of the gene encoding human keratocan, DNA Seq. 10 (1999) 67^74. [9] S. Chakravarti, T. Magnuson, J.H. Lass, K.J. Jepsen, C. LaMantia, H. Carroll, Lumican regulates collagen ¢bril assembly : Skin fragility and corneal opacity in the absence of lumican, J. Cell Biol. 141 (1998) 1277^1286. [10] G.W. Conrad, A. Dorfman, Synthesis of sulfated mucopolysaccharides by chick corneal ¢broblasts in vitro, Exp. Eye Res. 18 (1974) 421^433. [11] M.P. Beales, J.L. Funderburgh, J.V. Jester, J.R. Hassell, Proteoglycan synthesis by bovine keratocytes and corneal ¢broblasts: maintenance of the keratocyte phenotype in culture, Invest. Ophthalmol. Vis. Sci. 40 (1999) 1658. [12] E.S. Tasheva, J.L. Funderburgh, L.M. Corpuz, G.W. Conrad, Cloning, characterization and tissue-speci¢c expression of the gene encoding bovine keratocan, a corneal keratan sulfate proteoglycan, Gene 218 (1998) 63^68. [13] A. Travis, A. Amsterdam, C. Belanger, R. Grosschedl, LEF-1, a gene encoding a lymphoid-speci¢c protein with an HMG domain, regulates T-cell receptor alpha enhancer function, Genes Dev. 5 (1991) 880^894. [14] G. Karsenty, P. Golumbek, B. de Crombrugghe, Point mutations and small substitution mutations in three di¡erent upstream elements inhibit the activity of the mouse alpha 2(I) collagen promoter, J. Biol. Chem. 263 (1988) 13909^13915. [15] Z.X. Zhang, V. Kumar, R.T. Rivera, J. Chisholm, D.K. Biswas, cisacting negative regulatory element of prolactin gene, J. Biol. Chem. 265 (1990) 4785^4788. [16] M. Burcin, R. Amold, M. Lutz, B. Kaiser, D. Runge, F. Lottspeich, G.N. Filippova, V.V. Lobanenkov, R. Renkawitz, Negative protein 1, which is required for function of the chicken lysozyme gene silencer in conjunction with hormone receptors, is identical to the multivalent zinc ¢nger repressor CTCF, Mol. Cell Biol. 17 (1997) 1281^1288. [17] R.J. Kraus, E.E. Murray, S.R. Wiley, N.M. Zink, K. Loritz, G.W.
BBAEXP 91424 3-7-00
E.S. Tasheva et al. / Biochimica et Biophysica Acta 1492 (2000) 452^459 Gelembiuk, J.E. Mertz, Experimentally determined weight matrix de¢nitions of the initiator and TBP binding site elements of promoters, Nucleic Acids Res. 24 (1996) 1531^1539. [18] E.S. Tasheva, M.L. Funderburgh, J. McReynolds, J.L. Funderburgh, G.W. Conrad, The bovine mimecan gene. Molecular cloning and characterization of two major RNA transcripts generated by alternative use of two splice acceptor sites in the third exon, J. Biol. Chem. 274 (1999) 18693^18701.
459
[19] J.D. Dignam, R.M. Lebovitz, R.G. Roeder, Accurate transcription initiation by RNA polymerase II in a soluble extract from isolated mammalian nuclei, Nucleic Acids Res. 11 (1983) 1475^1489. [20] J.L. Funderburgh, M.L. Funderburgh, M.M. Mann, S. Prakash, G.W. Conrad, Synthesis of corneal keratan sulfate proteoglycans by bovine keratocytes in vitro, J. Biol. Chem. 271 (1996) 31431^ 31436.
BBAEXP 91424 3-7-00