Activation of a nuclear DNA-binding protein recognized by a transcriptional element, bcn-1, from the laminin B2 chain gene promoter.

Treatment of mesangial cells with either phorbol 12-myristate 13-acetate (PMA) or interleukin-1β induces an increase in laminin B2 chain mRNA levels. In other systems, activation of gene expression by these agents is transcriptionally mediated. To identify transcription factors that control expression of laminin B2 chain gene, we employed a strategy consisting of a computer-based analysis of murine and human gene promoter sequences and gel shift assays. Although overall the laminin B2 chain promoters from the two species have low sequence similarity, the mouse promoter contained sequences that were also contained in one motif, 5′-CCCCGCCCACCTCGCGCGC-3′, designated bcn-1, from the human promoter. Treatment of mesangial cells with either PMA or interleukin-1β induced a transient increase in nuclear DNA binding activity, designated BCN-1, recognized the bcn-1 motif in a gel shift assay. A single nucleotide replacement in the bcn-1 motif abolished DNA binding, indicating that bcn-1·BCN-1 complex formation is highly specific. In transient transfections, the ability of PMA to induce the laminin B2 chain promoter was abolished by mutating the bcn-1 motif. These results suggest that the bcn-1 element and its cognate inducible BCN-1 protein regulate laminin B2 chain gene transcription.

Laminin is a major component of the glomerular basement membrane (1). It is a large non-collagenous glycoprotein composed of three chains connected by an ␣-helical coiled-coil domain (2). At least five distinct chains of laminin are known to exist. The basement membrane contains only a unique set of laminin chain heterotrimers; they all contain B2 chain in a combination with either A chain or merosin and either B1 or S chain (3)(4)(5). In the kidney, the expression of the variant laminin chains depends on the tissue type (6), stage of development (7)(8)(9), and disease (10). The level and distribution of laminin in the glomerular basement membrane also varies with types of glomerular disease (11,12). Although it is known that an abnormal accumulation of laminin chains contributes to the progression of glomerular scarring, the molecular mechanisms responsible for the increased synthesis of laminin are very poorly understood.
In the kidney, laminin is thought to be synthesized by both glomerular epithelial and mesangial cells (3,13). We have demonstrated previously that treatment of glomerular epithelial cells with IL-1␤ 1 induced a transient increase in laminin B2 chain mRNA levels (14). This increase was associated with the activation of NF-B DNA binding activity recognized by a Blike motif contained in the murine laminin B2 chain gene promoter (14). This association suggests that production of laminin chains in glomerular cells could, in part, be transcriptionally mediated.
Promoters of a number of the laminin chain genes share similarities, suggesting that transcription of laminin chain genes is regulated by similar or overlapping mechanisms. Human and murine B1 and B2 chain promoters have no TATA or CAAT boxes, contain a number of GC boxes, B consensus sequences, and potential Sp1 binding sites (GGGCGG) (15)(16)(17)(18)(19). TATA-less promoters typically contain a number of Sp1 DNAbinding sites, and the Sp1 transcription factor is thought to play a particularly key role in the initiation of transcription of genes that contain no TATA boxes in their promoters. The human B2 promoter contains a potential AP-2 binding site (CCCCAGGC) (15,17), suggesting that this or related factor along with NF-B and Sp1 might play a role in the transcriptional control of expression of laminin genes. To gain more insight into the transcriptional regulation of laminin genes, we used a computer-based analysis to compared the sequences of the murine and the human laminin B2 chain gene promoters (17,18). Here, we demonstrate that in mesangial cells IL-1␤-, PMA-, and TGF-␤induced transient increase in laminin B2 chain mRNA levels is preceded by the induction of nuclear DNA binding activity recognized by a highly conserved enhancer element, designated bcn-1, contained within both the rodent and the human laminin B2 chain gene promoter.

MATERIALS AND METHODS
Mesangial Cell Culture-The rat mesangial cell line was established from collagenase-treated glomeruli, and cells were characterized as described previously (20). Cells were grown in RPMI 1640 supplemented with 10% fetal bovine serum (Life Technologies, Inc.), 100 IU/ml penicillin, 100 g/ml streptomycin, and 200 g/ml L-glutamine. Cells were maintained at 37°C in 5% CO 2 in air and were passaged every 5-7 days by scraping.
Northern Blot Analyses-Total RNA was extracted essentially as described previously (21) with some modifications. Briefly, after treatment of mesangial cells (approximately 1.0 ϫ 10 7 cells) with inducing agents, cells were directly lysed and denatured in 2.0 ml of RNAzol B™ (Cinna Scientific Inc., Friendswood, TX) to isolate RNA. RNA was analyzed as described previously (22). A total of 15 g of RNA was electrophoresed through a 1.0% agarose gel containing 2.2 M formaldehyde and 0.2 M MOPS, pH 7.0. The gels were run for 2 h at 100 V, and RNA was transferred overnight to a nylon membrane (Hybond-N nylon membrane; Amersham Corp.) in 10 ϫ standard saline citrate (1 ϫ SSC, 150 mM NaCl, 15 mM sodium citrate) by a rapid downward transfer system (Turboblotter; Schleicher & Schuell). RNA was fixed to the membrane by shortwave UV cross-linking (120,000 J/cm 2 ). The murine laminin B2 chain cDNA (provided by Dr. Y. Yamada, National Institutes of Health, Bethesda, MD) (4) was labeled with 50 Ci of [ 32 P]dCTP using the Klenow fragment of Escherichia coli DNA polymerase I. Hybridization was conducted at 68°C for 24 h using 10 ml of Quick Hyb Solution (Stratagene, Menasha, WI)/blot, 100 g/ml salmon sperm DNA, and 2.0 ϫ 10 6 cpm/ml labeled cDNA probe. After hybridization, the membrane was washed three times for 5 min in 2 ϫ standard sodium phosphate-EDTA (1 ϫ SSPE, 150 mM NaCl, 10 mM NaHPO 4 , 1 mM EDTA, pH 7.4) with 0.1% SDS at room temperature and for at least 30 min until the background disappeared in 0.1 ϫ SSPE, 0.1% SDS at 60°C. The membrane was autoradiographed for 3-7 days at Ϫ70°C with intensifying screens. The membranes were stripped and reprobed under identical conditions with a [ 32 P]dCTP-labeled 28 S probe (22) (10 4 cpm/ml) to normalize amounts of RNA loaded per lane.
Electrophoretic Mobility Shift Assay-Electrophoretic mobility gel shift assay was performed as described previously (24) with some modifications. The double-stranded oligonucleotide probe used in gel shift assay was end-labeled with total 1.0 ϫ 10 6 cpm/sample [␥-32 P]ATP, and 0.2 unit/l T4 polynucleotide kinase (Life Technologies, Inc.). Binding reactions were carried out in a total volume of 20 l with 30 g of nuclear protein extract from mesangial cells, 8 -16 ng of oligonucleotides, and 4 g of poly(dI-dC) at room temperature for 30 min. The 4% polyacrylamide gel (19:1, acrylamide:bisacrylamide) electrophoresis was performed at 180 V for 2 h in 0.5 ϫ TBE (45 mM Tris borate and 1 mM EDTA, pH 8.0).
Plasmid Constructs-Synthetic bcn-1 oligonucleotide was used to construct reporter plasmid containing a dimer of wild-type and mutated bcn-1 oligonucleotides. The bcn-1 dimer inserts were subcloned in the BglII and MluI sites of pGL3 control vector (Promega, Madison, WI), and the nucleotide sequence of the inserts was confirmed by dideoxy sequencing with Sequenase (U. S. Biochemical Corp.) (28).
For site-directed mutagenesis of the bcn-1 motif, the rat laminin B2 chain gene promoter (Ϫ948 to Ϫ79 relative to the transcription initiation site) 2 was subcloned into a luciferase reporter gene, pGL3 enhancer (Promega, Madison, WI). Mutagenesis of the five base pairs (see bold below) required for protein binding was performed using Ouick-Change™ site-directed mutagenesis kit (Stratagene, LaJolla, CA) with oligonuceotides mut-bcn-1⅐BAMH1 sense 5Ј-CCACCGCCCCTGGATC-CTCGCGCCCTTCCC-3 and antisense 5Ј-GGGAAGGGCGCGAGGAT-CCAGGGGCGGTGG-3Ј. The mutation was verified by restriction analysis (a new BamHI site was created by the mutagenesis) and by direct dideoxy nucleotide sequencing with Sequenase (U. S. Biochemical Corp.).
Transient Transfection of Mesangial Cells and Luciferase Reporter Gene Assay-Mesangial cells were transiently transfected using a lipofection method (29). Briefly, dioleoyl-L-␣-phosphatidylethanolamine (Sigma) and dimethyldioctadecylammonium bromide (Sigma) were mixed in absolute alcohol in a 0.5:1 molar ratio. Liposomes were prepared by injecting 50 l of this lipid solution into 1.0 ml of sterile water while vortexing (1 nmol of lipid/l of alcohol). A liposome complex was then formed with 5 g DNA (18:1 nmol of lipid:g of DNA) and applied directly to cells in serum-free medium for 4 h. Cells were then incubated in complete medium for 24 h. After treatment with or without an inducing agent, cells were pelleted and resuspended in 150 l of reporter lysis buffer (Promega, Madison, WI). Luciferase assay was carried out as described (30). Luciferase activity was quantitated for 30 s using a bioluminometer (LB9502; Wallac Inc., Gaithersburg, MD), and light units were adjusted for protein content.
Statistical Analysis-The means were compared by analysis of variance (ANOVA) using Fisher's test (31).

Northern Blot Analysis of the Laminin B2 Chain mRNA Levels in Mesangial Cells
Treated with Cytokines-In the glomerulus, laminin is a major constituent of the lamina rara externa of the glomerular basement membrane (13). Although the mechanisms responsible for the regulation of laminin chain synthesis in the glomerulus are poorly understood, laminins are thought to be produced by both the mesangial and glomerular epithelial cells. To elucidate the potential mechanisms that are responsible for the regulation of laminin production in glomerular cells, we examined the expression of laminin B2 chain mRNA levels in response to treatment of mesangial cells with a number of known cytokines and mitogens. Fig. 1 illustrates an autoradiograph of a Northern blot of total RNA isolated from untreated (Control, lane 1) and treated (lanes 2-10) mesangial cells probed with either 32 P-labeled murine laminin B2 chain cDNA (4)  In contrast to the laminin B2 chain message, the levels of 28 S RNA did not increase (lanes 1-10 in the lower panel) in response to treatment with these agents. These re-sults are analogous to the IL-1␤ effects seen in glomerular epithelial cells, where IL-1␤ also induced a transient increase in laminin B2 chain mRNA levels but the peak effect was seen after 2 h of treatment (14).

A Motif from the Laminin B2 Chain Gene Promoter Recognizes an Inducible DNA Binding Activity in Nuclear Extracts
from Mesangial Cells-In many cell systems PMA, IL-1␤, and TGF-␤ increase message levels through transcriptional mechanisms (32)(33)(34). Hence the transient increase in laminin B2 chain mRNA levels in mesangial ( Fig. 1) and glomerular epithelial cells (14) might, likewise, be transcriptionally mediated. To identify transcriptional elements contained within the laminin B2 chain gene promoter, we compared the human (17) and mouse (18) sequences 5Ј to the first exon using the GCG program (Genetics Computer Group, Madison, WI). Although the overall promoter sequences in the two species are significantly divergent, there was a region in the human promoter (Ϫ550 to Ϫ525 relative to the first codon) that contained sequences identical or nearly identical to short regions (9 -17 nucleotides long) present in the murine laminin B2 chain gene promoter ( Fig. 2). We reasoned that one or more of these similar promoter motifs might have been evolutionarily conserved in order to recognize the same or related transcription factors in rodents and humans. To test this hypothesis, we synthesized a doublestranded oligonucleotides that represented the human laminin B2 chain gene promoter sequence. To take into the account the two shorter murine fragments that were disparate from the human sequence at a single point, the synthetic oligonucleotide was programmed degenerate to contain either G or T at position Ϫ546, and either A or T at position Ϫ543 (human laminin B2 chain gene promoter) (17). Fig. 3. illustrates an autoradiograph of a gel shift assay performed on nuclear extracts from untreated and treated rat mesangial cells using as a probe the At given time points, total RNA was isolated, electrophoresed on a formaldehyde-agarose gel (15 g total RNA/lane), and blotted to a nitrocellulose membrane. The blot was hybridized with a 32 P-labeled murine laminin B2 chain cDNA (4) probe and was autoradiographed (top panel). The same membrane was then stripped, rehybridized with a 32 P-labeled 28 S probe as a loading control, and was autoradiographed (bottom panel).

FIG. 2.
Identification of a common motif in the murine and human laminin B2 chain gene promoter. The murine and human laminin B2 chain gene promoter sequences were aligned (Lasergene, DNASTAR). One motif designated, bcn-1, stretching in the human promoter from the nucleotide Ϫ550 to Ϫ525 (17) shared high sequence similarities or identity with short regions from the mouse promoter. Numbering of the nucleotides is shown relative to the first codon in the two species.  (Fig. 3C), but the magnitude was not as pronounced as that seen following treatment of these cells with either IL-1␤ or PMA (Fig. 3, compare panel C to panels A and B). These results demonstrate that in mesangial cells PMA-, IL-1␤-, or TGF-␤-induced transient increase in laminin B2 chain mRNA levels ( Fig. 1) is preceded by a transient increase in a nuclear DNA binding activity recognized by a conserved DNA motif present in the human and mouse laminin B2 chain gene promoter. This DNA-binding protein(s) does not bind to either sense or antisense strand of the bcn-1 motif (data not shown).
The Interaction of the bcn-1 Motif with Its Cognate Protein(s) Is DNA Sequence-specific-To examine whether binding of the BCN-1 protein to the bcn-1 probe is DNA sequence-specific, nuclear proteins extracted from PMA-treated mesangial cells were incubated with the 32 P-labeled bcn-1 probe in the absence (control, lane 1) or presence of increasing (lanes 2-10) concentrations of unlabeled competitor prior to analysis of DNA binding in a gel shift assay. The autoradiograph of the gel (Fig. 4A) demonstrates that binding to the 32 P-labeled bcn-1 motif was effectively competed by an excess unlabeled bcn-1 DNA (compare lanes 6 and 7 to lane 1) but not by a comparable excess of oligonucleotide bearing either the wild-type or mutated B motif (compare lanes 4 and 10 to lane 1). As a control for these experiments, we used the same nuclear extracts from PMAtreated mesangial cells but instead tested DNA binding to 32 P-labeled synthetic oligonucleotide containing the wild-type B (Fig. 4B) (24), alone or in the presence of an increasing excess of oligonucleotides containing either the wild-type or mutated B (24), or bcn-1 motif. Treatment of mesangial cells with PMA activates NF-B (Fig. 8). An autoradiograph (Fig.  4B) from these competition experiments shows that the binding to the 32 P-labeled NF-B probe was effectively competed by an excess of an unlabeled oligonucleotide containing the wild-type B motif (compare lanes 3 and 4 to lane 1), but not by a comparable amount of oligonucleotide containing either the mutated B or the bcn-1 motif (compare lanes 6,7,9, and 10 to lane 1). This series of experiments demonstrated that the bcn-1 motif recognizes its cognate nuclear protein (or protein complex) in a DNA sequence-specific manner and that the BCN-1 DNA-binding protein is not the transcription factor NF-B.

The (A/T)CCT Nucleotide Region Contained within the bcn-1 Motif Is Required for Recognition of the Cognate Protein-To
define the nucleotide bases contained within the bcn-1 motif that mediate protein binding, we designed "mutated" bcn-1 oligonucleotides and tested their ability to recognize nuclear proteins from PMA-treated mesangial cells. Prior to synthesis and their use in gel shift assays, the designed mutated bcn-1 sequences were rechecked against sequences in the transcription data base in GCG (Genetics Computer Group, Madison, WI) to ensure that a known transcription factor DNA recognition site was not created. An autoradiograph of a gel shift assay of nuclear proteins extracted from PMA-treated mesangial cells using these 32 P-labeled double-stranded oligonucleotides and their nucleotide sequences are shown in Fig. 5. Mutant bcn-1A (5Ј-CCCC-3Ј motif in the coding strand (sense) located in the 3Ј half of bcn-1 was converted to 5Ј-ATTA-3Ј motif) recognized PMA-inducible nuclear DNA binding activity with kinetics and electrophoretic mobility indistinguishable from the nuclear DNA binding activity recognized by the wild-type bcn-1 motif (compare lanes 4 -6 to lanes 1-3). In sharp contrast, a 32 Plabeled mutant bcn-1B (5Ј-(A/T)CCT-3Ј was converted to 5Ј-CTTC-3Ј) did not recognize any inducible nuclear DNA binding activity (compare lanes 7-9 to lanes 1-3). These gel shift assays indicate that the (A/T)CCT nucleotide stretch in the middle of the bcn-1 motif is required for the recognition of the nuclear PMA-inducible protein (or protein complex).
A Single Nucleotide Mutation in the bcn-1 Motif Disrupts Protein Binding-To maximize the likelihood of identification of an inducible nuclear factor, we have initially used a preparation of degenerate synthetic oligonucleotides containing sequences that were degenerate in two positions; one position was programed to contain either G or T, and another position contained either A or T. Thus, this mixture was made up of four different sequences differing from each other by one or two nucleotide bases. To determine which sequence(s) in the degenerate mixture is responsible for the binding of the BCN-1 protein, we separately synthesized double-stranded oligonucleotides representing one of the four possible bcn-1 sequences. A gel shift assay (Fig. 6) demonstrated that the oligonucleotide representing exactly the human laminin B2 promoter sequences (Ϫ550 to Ϫ525) (Fig. 2) (17) (lanes 1-3) recognized a very strong PMA-inducible DNA binding activity. In a sharp contrast, the other three oligonucleotides either did not bind it at all (compare lanes 1-3 to lanes 4 -9) or the binding was very weak (lanes 10 -12). The observation that a single nucleotide base pair replacement in this promoter region was sufficient to block binding of the inducible factor indicates that the recognition of the bcn-1 nucleotide motif by the BCN-1 protein is remarkably specific. This and the previous set of experiments allow us to approximately define the 5Ј-CCCGCCCAC-CTCGCGC-3Ј motif as the relevant protein-binding nucleotide region and identify both G and A (shown in bold) as obligatory for protein binding. An identical sequence is also present in the mouse (Fig. 2) (18) and the rat laminin B2 chain gene promoter. 2 Search of the GCG transcription factor data base revealed that this motif contains a transcriptionally active element from the apoE B1 gene promoter, 5Ј-(G/C)CCCACCTCG-3Ј (35).
The bcn-1 Motif Does Not Recognize the Transcription Factor Sp1-To determine whether or not the bcn-1 motif contains a transcriptional element that recognizes a known transcriptional factor, we compared the bcn-1 motif with DNA elements in the transcriptional factor data base (GCG). The data base search revealed that the bcn-1 sequence is most similar to the motif that binds the Sp1 transcription factor. Sp1 is a ubiquitous transcription factor that plays an important role in transcriptional regulation of expression of many housekeeping genes (19). It drives gene expression by binding to GC-rich promoter nucleotide sequences (36). Because the bcn-1 motif is also GC-rich (Fig. 2), we next tested whether or not the bcn-1binding protein is in fact Sp1. To do that, nuclear extracts from PMA-treated cells were preincubated with either anti-Sp1 antibody or an excess of unlabeled oligonucleotide containing the Sp1-binding site prior to binding to the 32 P-labeled bcn-1 probe. The DNA-protein complexes were then resolved by gel shift assay as in Fig. 7. The autoradiograph from this experiment (Fig. 7A) demonstrates that the anti-Sp1 antibody did not supershift or alter in any other way the electrophoretic mobility of the bcn1-protein complex (compare lanes 2 and 3). This suggests that the protein recognized by the bcn-1 motif is not Sp1. An excess of unlabeled Sp1-binding DNA did not have an effect on the intensity of the bcn-1⅐BCN-1 complex, further showing that the bcn-1 motif does not recognize the Sp1 transcription factor (compare lanes 2 and 3). As a positive control for these experiments, we concurrently performed gel shift assay on the same nuclear extracts using a 32 P-labeled synthetic oligonucleotide containing the Sp1-binding site as the probe. Although not as dramatic as with the BCN-1 response, treatment of mesangial cells with PMA increased nuclear levels of Sp1 (Fig.  7, compare lanes 1 and 2 in panel A to those in panel B). The fact that the DNA-protein complexes shown here (Fig. 7B) contain Sp1 is demonstrated by the supershift seen when the nuclear extracts were preincubated with the anti-Sp1 antibody (compare lanes 2 and 3), and by the observation that unlabeled synthetic oligonucleotide containing the Sp1-binding site markedly diminished the intensity of the 32 P-labeled DNAprotein complex (compare lanes 2 and 4). Collectively, these results demonstrate that, although bcn-1 is a GC-rich motif, it does not recognize the transcriptional factor, Sp1.
The Protein Synthesis Inhibitor Cycloheximide Blocks the Induction of the BCN-1 DNA Binding Activity-To determine whether new protein synthesis is needed for the activation of the BCN-1 DNA binding activity, mesangial cells were treated for 1 h with 25 g/ml cycloheximide, a protein synthesis inhibitor (37), prior to treatment with PMA. At given time points cells were harvested, nuclei were isolated, and nuclear protein extracts were assayed in a gel shift assay using either 32 Plabeled bcn-1 (Fig. 8A) or B (Fig. 8B) motif. As before (Fig. 3), treatment of mesangial cells with PMA potently activated the BCN-1 DNA binding activity (Fig. 8A, lanes 1-3). In sharp contrast, pretreatment of cells with cycloheximide completely blocked the constitutive and the inducible nuclear BCN-1 DNA binding activity (lanes 4 -6). As a positive control, gel shift analysis was done with the same extracts using instead the B motif, which recognizes the transcription factor, NF-B. The NF-B transcription factor is sequestered in the cytoplasm by the inhibitor IB. Treatment of cells with cycloheximide blocks the synthesis of IB, permitting NF-B both to translocate to the nucleus and to bind the B motif (38). In agreement with previous reports in other cell systems, treatment of mesangial cells with either PMA (Fig. 8B, lanes 1-3) or cycloheximide (Fig. 8B, lanes 4 -6) activated nuclear NF-B DNA binding activity. Pretreatment of mesangial cells with several cell-permeable protease inhibitors (10 g/ml E-64, 10 g/ml antipain dihydrochloride, and 10 g/ml leupeptin) (39) prior to treatment with cycloheximide did not restore the PMA ability to activate BCN-1 DNA binding activity (data not shown). These series of experiments, therefore, suggest that activation of nuclear BCN-1 DNA binding activity requires new protein synthesis.
The bcn-1 Motif Exhibits Inducible Transcriptional Activity Either within the Context of the Laminin B2 Chain Promoter or When Cloned Upstream of Heterologous Promoter-To assess the transcriptional activity of the bcn-1 motif, we cloned a synthetic oligonucleotide containing either a wild-type or a mutant type bcn-1 dimer motif into a pGL3-control vector bearing the luciferase reporter gene (Promega, Madison, WI). Luciferase activity of the pGL3-control vector containing the wildtype bcn-1 dimer transiently expressed in untreated mesangial cells was 97.4 Ϯ 13.6% (Fig. 9A, lane 1) in untreated cells and increased to 174.4 Ϯ 12.9% (lane 2) after 24 h of treatment of cells with PMA, (n ϭ 6, p Ͻ 0.005). Relative luciferase activity of the pGL3-control vector containing the mutant-type bcn-1 dimer in untreated cells was 34.2 Ϯ 11.3% (Fig. 9, lane 3) and 38.2 Ϯ 2.9% (lane 4) after 24 h of PMA stimulation, a 4% difference that was not statistically significant (n ϭ 6, p ϭ 0.73).
To evaluate the transcriptional activity of the bcn-1 motif within the context of the laminin B2 promoter, the Ϫ948 to Ϫ78 (relative to the transcription initiation site) fragment of the rat promoter containing the bcn-1 motif was cloned upstream of a luciferase reporter gene. Mutagenesis of the five bcn-1 bases that are required for protein binding was performed using sense and antisense synthetic oligonucleotides that were designed to contain mutated bcn-1 motif and a BamHI site (see "Materials and Methods"). Relative luciferase activity of the pGL3-enhancer vector containing the wild-type laminin B2 chain promoter transiently expressed in mesangial cells was 107.5 Ϯ 24.2% (Fig. 9B, lane 1) 4, p ϭ 0.25). These results demonstrate that the bcn-1 motif from the laminin B2 chain gene promoter is transcriptionally active and PMA-responsive. The wild-type laminin B2 promoter was only weakly responsive to IL-1␤ treatment, and the small response was not blocked by mutating the bcn-1 motif (data not shown). While the activation of the BCN-1 DNA binding activity by PMA in mesangial cells has consistently been very robust, the level of BCN-1 activation by IL-1␤ has not been as large (Fig. 3, compare panel  4). DNAprotein complexes were resolved by gel electrophoresis and gels were autoradiographed as before (Fig. 3). A to panel B). Therefore, the discrepancy between the measured IL-1␤-induced BCN-1 DNA binding and the reporter gene activity in transient transfections may be accounted for by the fact that the gel shift analysis is more sensitive. Also of note is the observation that at 2 h of IL-1␤ the BCN-1 DNA binding activity is nearly gone, while the laminin B2 chain mRNA does not peak until 4 h. These discrepancies may reflect involvement of multiple factors in the IL-1␤-induced gene expression.

DISCUSSION
In vitro studies have demonstrated that glomerular mesangial and epithelial cells are the likely source of laminin chains in the glomerulus (13). Both of these glomerular cell types have been successfully grown in culture (22,40), providing useful model systems to examine mechanisms that regulate laminin chain synthesis. A number of growth factors activate glomerular mesangial and epithelial cells, including IL-1 and TGF-␤ (41), and some of these agents have previously been reported to stimulate laminin production in these cells (42,43). We have shown previously that IL-1␤ increases laminin B2 chain mRNA levels in cultured glomerular epithelial cells (14). In the present study, we have extended these observations by demonstrating that treatment of mesangial cells with either PMA, IL-1␤, or TGF-␤ also results in a transient rise in laminin B2 chain mRNA levels (Fig. 1). The growth factor-induced increase in laminin B2 chain gene expression in glomerular mesangial (Fig. 1) and epithelial (14) cells is likely to be transcriptionally mediated.
A number of methods have been used to identify transcrip-tional elements. The traditional method consists of deletion analysis of gene promoters using reporter genes such as chloramphenicol acetyltransferase (44) or luciferase (30) as a readout for transcriptional activity. This approach resulted in the identification of most of the known transcription factors. The present study illustrates application of an alternative strategy to search for promoter motifs that recognize inducible proteins. This strategy is based on the fact that even in related species the non-coding DNA regions have low overall sequence homology, but in order to recognize homologous transcription factors, the regulatory DNA elements contained in the untranslated regions are very likely to be highly conserved across related species. The conserved regulatory motifs are simply identified by computer-driven alignment of the promoter sequences from different species. The promoter sequences of regions that are conserved are then analyzed against transcription data base to determine (i) whether or not other gene promoters contain these motifs, (ii) whether or not these motifs are known to be transcriptionally active, and (iii) whether or not these or homologous nucleotide motifs are known to recognize transcription factor (s). Based on the computer data base search, one can then choose which of the conserved promoter motifs ought to be used to design synthetic oligonucleotides to test if they recognize inducible DNA binding activities in gel shift assays. The advantage of this strategy is its simplicity because it is based on computer analysis and gel shift assays. This strategy is particularly suitable for analysis of those gene promoters whose nucleotide sequences in related species have an overall FIG. 9. Assessment of bcn-1 transcriptional activity using the luciferase reporter gene. Liposomes (dimethyldioctadecylammonium bromide:dioleoyl-L-␣-phosphatidylethanolamine in a 0.5:1 molar ratio) were prepared as described (29). Liposome-DNA complexes were formed with a lipid:DNA (nmol:g) ratio of 18:1. After incubating the cells with these complexes, the cells were incubated in complete medium for 24 h, and were subsequently treated with or without 10 Ϫ7 M PMA for 24 h. After treatment, the cells were pelleted and lysed in reporter lysis buffer. Luciferase assay was carried out as described (30) using a bioluminometer. Light units were adjusted for total protein content in the samples. Panel A, mesangial cells (approximately 5.0 ϫ 10 6 ) were transiently transfected with 5 g of pGL luciferase control plasmid (Promega, Madison, WI) bearing either the wild-type (lanes 1 and 2) or mutated (lanes 3 and 4) bcn-1-dimer insert. Data are shown as mean Ϯ S.E. of relative luciferase activity (n ϭ 6). Panel B, mesangial cells were transiently transfected with 5 g of pGL luciferase plasmid containing the Ϫ948 to Ϫ79 (relative to the transcription initiation site) fragment of the rat laminin B2 chain promoter bearing either the wild-type (lanes 1 and 2) or a mutated (lanes 3 and 4) bcn-1 motif. Data are shown as mean Ϯ S.E. of relative luciferase activity (n ϭ 4). Mutated nucleotides are shown in bold. low sequence homology but contain short (20 -30 nucleotides) regions that are conserved.
We applied this strategy to begin identification of transcription factors that potentially control laminin B2 chain gene expression. While the coding nucleotide sequences of the mouse and human laminin B2 chain genes share high homology, the region 5Ј to the first exons are overall divergent (4,17,18,45,46). We identified highly conserved nucleotide regions and showed that one of the conserved motifs, termed bcn-1 (Fig. 2), did recognize an inducible nuclear protein(s), BCN-1 (Fig. 3). We demonstrated that a single nucleotide base pair replacement was sufficient to block protein binding (Fig. 5), indicating that this interaction is highly specific. The BCN-1 DNA binding activity is activated by the same growth factors (Fig. 3) that increase laminin B2 chain mRNA levels (Fig. 1). This association suggests that the BCN-1 factor might contribute to the enhanced laminin B2 chain gene expression following treatment of mesangial cells with these growth factors. This postulate is supported by the observation that the bcn-1 motif is transcriptionally active and PMA-responsive (Fig. 9).
Analysis of the bcn-1 nucleotide sequence revealed that it contains a potential GC-rich Sp1-binding site. Although there are nucleotide sequence similarities between the bcn-1 motif and the Sp1 DNA recognition site, neither anti-Sp1 antibody nor an excess of cold Sp1-binding oligonucleotide altered either the electrophoretic mobility or the intensity of the bcn-1⅐BCN-1 complex in a gel shift assay (Fig. 7A). These results demonstrated that the Sp1 transcription factor is not recognized by the bcn-1 motif. Similar competition studies and gel shift analyses were done using an excess of synthetic oligonucleotides containing binding sites for NF-B (33), AP-1 (26), and AP-2 (27). Like BCN-1, NF-B and AP-1 are factors known to respond to treatment of cells with either PMA or IL-1 (26,47,48). As with Sp1-binding DNA (Fig. 8A), neither NF-B- (Fig. 4A), AP-1-, nor AP-2-binding oligonucleotides altered the intensity of the 32 P-labeled bcn-1-protein complex (data not shown). Recently, a PMA-inducible factor, denoted PDGF-A-BP-1, was identified in mesangial cells that is recognized by 5Ј-GGCCCG-GAATCCGGGGGAGGC-3Ј motif from the PDGF-A chain gene promoter (49). Because neither the sense nor antisense bcn-1 transcriptional element resembles the sequence from the PDGF-A chain promoter, the BCN-1 and PDGF-A-BP-1 factors are probably different. These results and considerations suggest that the nuclear factor recognized by the bcn-1 motif is a novel protein complex. This contention is consistent with the fact that search of the transcription data base (GCG, Madison, WI) did not reveal a known factor that can recognize the core bcn-1 nucleotide sequence 5Ј-CCCCGCCCACCTCGCG-3Ј, a motif that contains nucleotides required to bind the BCN-1 nuclear protein (Figs. 6 and 7). Although the factor that binds the 5Ј-CCCCGCCCACCTCGCG-3Ј motif remains to be identified, this nucleotide region contains a transcriptionally active element from the apoE B1 gene promoter, 5Ј-(G/C)CCCCACCT-3Ј (35).
In summary, a simple strategy was applied to identify an inducible nuclear DNA binding activity recognized by a conserved motif from the laminin B2 chain gene promoter. In the context of the laminin B2 chain gene promoter cloned upstream of a reporter gene, the bcn-1 motif is transcriptionally active and PMA-inducible, suggesting that it plays a role in the induction of the laminin B2 chain gene transcription in mesangial cells. Work is in progress to clone the cDNA encoding this protein.