Expression of the Npt2 gene, encoding the type II sodium-dependent phosphate cotransporter, is restricted to renal proximal tubule epithelium. We have isolated a 4,740-bp fragment of the 5′-flanking sequence of the ratNpt2 gene, identified the transcription initiation site, and demonstrated that this 5′-flanking sequence drives luciferase-reporter gene expression, following transfection in the proximal tubule cell-derived opossum kidney (OK) cell line but not in unrelated cell lines. Analysis of the promoter sequence revealed the presence of 10 consensus binding motifs for the AP2 transcription factor. Transient transfection assays revealed an important effect of the number of tandemly repeated AP2 sites in enhancing promoter activity. The promoter sequence also revealed a pair of inverted repeats enclosing 1,324 bp of intervening sequence and containing 8 of the total 10 AP2 consensus sites in the promoter sequence. Deletion or reversal of orientation of the distal inverted repeat resulted in marked enhancement of promoter activity. Electrophoretic mobility shift analysis revealed a distinct pattern of transcription factor binding to oligonucleotides containing AP2 sites, using nuclear extracts from OK cells, compared with unrelated cell lines. Taken together, these results suggest an important role for AP2 consensus binding sites in regulating Npt2 gene expression and suggest a mechanism of regulation mediated by the interaction of inverted repeats enclosing these sites.
- inverted repeats
- proximal tubule
- transcription factor
the kidney represents an attractive model system for investigating mechanisms for the regulation of cell-specific gene expression. Recently, several genes whose expression is restricted primarily to the kidney or to specific cell types within the nephron have been identified (6, 11, 15, 16, 32, 35). One such gene (Npt2) encodes the type II sodium-dependent phosphate cotransporter, responsible for the sodium-dependent uptake of phosphate across the apical membrane of the proximal tubule (3, 26, 38). Handling of inorganic phosphate by the kidney represents an important determinant in maintaining overall body phosphate homeostasis and in maintaining the concentration of phosphate in the extracellular fluid within a narrow range. The apical sodium-dependent phosphate cotransporter of the proximal tubule appears to be a rate-limiting site for physiological regulation of overall phosphate reabsorption in response to numerous influences, such as parathyroid and other hormones and varying dietary phosphate intake, among others (27). Two major families of sodium-dependent phosphate cotransporters (type I and type II), have been identified, with little overall homology between members of the two families (38). Both type I and type II transporters are involved in proximal tubule phosphate reabsorption. However, only type II transporters are highly specific for sodium-dependent phosphate cotransport in a manner that is physiologically regulated in response to hormones and dietary intake (18, 19, 28, 39). Furthermore, immunohistochemical localization studies have revealed expression of type I transporters in multiple tissues and cell types, whereas type II transporter expression is restricted to the kidney proximal tubule. This expression is limited mainly to the S1 segment, with gradually decreasing expression toward the S3 segment, in agreement with the described pattern of intranephron heterogeneity for transport activity (2, 3, 8, 9, 36). Accordingly, analysis of the promoter for theNpt2 gene encoding this transporter might provide important insights with respect to its cell-specific expression and more generally with respect to differentiation of nephron cell types.
Previous studies of the murine 5′-flanking sequence (484 nt) and the human 5′-flanking sequence (2,400 nt), have focused primarily on the regulation of promoter activity by hormonal and dietary influences (12-14, 33, 34). In the current study, we have analyzed the 4,740 bp of the 5′-flanking region of the rat Npt2 promoter, and have focused on regions which might potentially mediate cell-type-specific promoter activity using opossum kidney (OK) cell lines, which express the endogenous gene, in comparison with other cell lines that do not express the gene.
MATERIALS AND METHODS
Cell Lines and Cell Culture
The opossum kidney OK cell line, the mouse fibroblast NIH3T3 cell line, and the human embryonic kidney HEK293 cell line were grown in DMEM containing 2 mM glutamine, 50 U/ml penicillin, 50 μg/ml streptomycin, and 10% FCS. The cells were kept at 37°C in a humidified atmosphere of 5% CO2 in air. Cells were subcultured weekly using Ca2+- and Mg2+-free PBS and 0.25% trypsin, and the medium was exchanged twice a week.
Isolation and Sequencing of Genomic Clones
A P1 clone was obtained using primers NAP3 5′-tgatgtcctacagcgagagattg-3′and NAP4 antisense 5′-ctcagactggagatggcataggt-3′ based on the published rat Npt2 cDNA sequence. Restriction digest and Southern blot analysis using NAP3 oligonucleotide as a probe, yielded a 5.5-kb Hind III genomic fragment that was cloned into Hind III site of the pBluescript (Stratagene) vector. To identify the genomic sequences upstream of the published cDNA, PCR analysis was performed, using two sets of primer pairs.Set 1 was NAP3 antisense 5′-caatctcttcgctgtaggacatca-3′ and T3 from the pBluescript vector; set 2 was NAP3 antisense and T7 from the pBluescript vector. This analysis revealed that the 5.5-kb genomic fragment contains 4,740 nt upstream of the published cDNA. Subfragments of this genomic fragment were subjected to manual chain-termination DNA sequencing, yielding the complete sequence for the entire 5.5-kb genomic fragment. The sequence was subjected to computer analysis [Wisconsin Package Version 8.0, Genetics Computer Group (GCG), Madison, WI] for identification of potential transcription factor binding sites.
5′-RACE analysis was performed following the instructions contained in the GIBCO-BRL 5′-RACE kit for rapid amplification of cDNA ends. One microgram total RNA from rat kidney cortex was reverse transcribed using primer NAP4 antisense. The anchor-ligated product was PCR amplified using the anchor primer and internal primers NAP3 antisense and NAP11 5′-catgcaccatgtgtctccctc-3′. PCR amplified products were cloned into pUC57 vector using the TA-cloning kit (MBI Fermentas) and manually sequenced (Sequenase version 2.0, US Biochemicals). The transcription initiation (TI) site, as determined by the 5′-RACE result, was assigned nucleotide position +1.
Total RNAs (23 μg/lane; Tri-Reagent, Molecular Research Center), extracted from OK, NIH3T3, and HEK cells were separated on 1.2% agarose gel and transferred to nylon membranes. Npt4 transcript was detected by hybridization with a 32P-labeled, 1.7-kbPst I Npt2 cDNA fragment (Npt2 cDNA was kindly provided by H. Murer).
Promoter-luciferase constructs. The proximal 5′-flanking region of the rat Npt2 promoter from −4035 down to, but not including, the ATG at +706 was cloned into the Hind III site of the polylinker of pGL3-basic reporter plasmid (Promega, Madison, WI). A series of progressively increasing 5′-flanking deletions (5′ ends at position −3458, −2326, −1514, −954, and −570, respectively) and all extending to +706 were constructed using native restriction endonuclease sites (BamH I, PflM I, Hpa I, Acc65 I, andHinc II, respectively) and cloned into the multiple cloning site of the vector.
Thymidine-kinase constructs. Double-stranded oligonucleotide agaaaga-3′ containing the Bgl II endonuclease restriction site was phosphorylated with T4 polynucleotide kinase and ligated to the BamH I site of pBLCAT2 reporter plasmid. Constructs containing three, five, and six copies of the insert were generated. In addition, a native rat Npt2 promoter DNA fragment from −3316 to −2757 that includes four AP2 consensus sites was filled in and ligated into BamH I site of pBLCAT2. The resultant construct contains eight copies of AP2 consensus sites.
Transient Transfections and Reporter Assays
NIH3T3 cells (3.75 × 104), OK cells (5 × 104), and HEK293 cells (5 × 104) were seeded 24 h before transfection in 24-well plates, in DMEM supplemented with 10% FCS. Cotransfections were performed using calcium phosphate precipitation with 0.1 μg promoter-luciferase reporter construct and 0.3 μg β-galactosidase expression vector pCH110 (Pharmacia) for calibration of transfection efficiencies. Transfection mixtures were adjusted with carrier DNA (pBluescript) to a total of 1 μg DNA per well. At 24 h after transfection the medium was changed and cells were incubated for an additional 48 h. Cell extracts were prepared and luciferase, chloramphenicol acetyl transferase (CAT), or β-galactosidase activities were measured.
Electrophoretic Mobility Shift Assay (EMSA)
Nuclear extracts were prepared from OK and NIH3T3 cells as described previously (30). In vitro transcribed proteins were prepared using expression vectors of human AP2-α and murine AP2-β and AP2-γ (vectors kindly provided by R. Buettner) with the TNT kit (Promega). The human recombinant AP2-α protein was purchased from Promega. Fifteen micrograms of protein (or 300 ng recombinant AP2-α) was incubated with 300 ng poly(dI ⋅ dC) for 10 min at 4°C in a 20 μl reaction volume containing 10 mM HEPES, pH 7.9, 50 mM NaCl, 0.1 mM EDTA, 5% glycerol, 10 μg/ml BSA, and 1 mM DTT without or with 25-, 50-, and 200-fold molar excess of unlabeled competitor DNA. After incubation, 1 × 105 cpm of32P-labeled probe was added, and the reaction was incubated for an additional 20 min at 4°C. The DNA-protein complexes were separated by electrophoresis on a 5.3% polyacrylamide gel and visualized by autoradiography. For supershift assays, samples were preincubated with 5 μg antiserum against AP2-α (Santa Cruz Biotechnology, Santa Cruz, CA) for 1 h at 4°C. Double-stranded oligonucleotides containing the AP2 consensus binding sites located at nt −3362 and −3333 in the rat Npt2 promoter sequence were used as probes. The sequence of the oligonucleotide used was 5′-ccgggtttttaat- agaaagac-3′, which contains two versions of AP2 consensus sequence, cccagccc and cccagggg, with three overlapping nucleotides as indicated.
A 5.5-kb genomic fragment was isolated from a P1 clone that included 4,740 bp of 5′-flanking sequence upstream to the published rat Npt2 translation initiation ATG. This fragment was mapped by restriction digest and Southern blot analysis (data not shown), and the 5.5-kb sequence was determined (GenBank accession no. AF156188); 329 nucleotides at the 3′ end of the fragment correspond to the published rat Npt2 DNA sequence.
The TI site was mapped using 5′-RACE as described inmaterials and methods. The two products of 132 bp and 185 bp obtained from the 5′-RACE PCR reaction, as expected from the location of the two internal primers, were subcloned and sequenced (Fig. 1 A). The TI site was mapped 12 nt upstream of the published 5′ end of the rat Npt2 cDNA sequence, at the first A within the sequence GACT, corresponding to the TI consensus pyApypy for eukaryotic genes. The TI site was designated +1, and the rest of the sequence was marked accordingly. Typical TATA box and CAAT box motifs were identified at −31 nt and −137 nt, respectively. Computer analysis (Wisconsin GCG software) identified putative consensus transcription factor binding sites located in the promoter as described in Table 1. A CA repeat microsatellite (22 pairs) was identified at position −3047.
Interspecies Npt2 Promoter Sequence Comparison
A section of 165 bp of the rat Npt2 promoter in the proximity of the TI site was aligned with the published murine (12), human (12, 13, 34), and opossum (14) Npt2 promoters. High levels of similarity were observed when comparing the rat promoter region with the murine (89% identity), human (76% identity). and opossum promoters (62% identity) as shown in Fig. 1 B. Mutual putative TATA box and CAAT box motifs were identified in all three species at approximately the same distances. In comparison to the murine TI site, the rat transcript initiates one nucleotide further upstream.
Transcriptional Activity of the Rat Npt2 Promoter
To examine whether the 4,740-bp 5′-flanking genomic region of theNpt2 gene can activate cell-specific transcription, transient transfection assays were performed. Luciferase reporter plasmids (pGL3-basic) containing the 4,740-bp genomic region upstream of the ATG were constructed and transfected into OK proximal tubule cells, which express the endogenous Npt2 gene, as well as in NIH3T3 and HEK293 cells, which do not express the gene as shown by Northern blot analysis (Fig. 2 A). Following transfection, luciferase activity was measured in cell lysates (Fig. 2,B and C). The 4,740-bp genomic fragment gave ∼120-fold increased activity in OK cells compared with the promoterless reporter pGL3-basic plasmid. Of note, this activity is even higher than that previously reported for a similar size fragment off the opossum Npt2 promoter (14). This may be attributed to differences in DNA binding motifs found 5′ to the region of higher similarity shown in Fig. 1 B. In contrast, there was no evident effect of this 5′-flanking promoter sequence in NIH3T3 and HEK293 cells. These results suggest that Npt2 transcription is highly activated in OK cells, which express the endogenous gene, but not in NIH3T3 and HEK293 cells, which do not express the endogenous gene.
To define the promoter region essential for transcriptional activation of Npt2 in OK cells, deletion analysis of the promoter was performed. Promoter fragments decreasing in size from the 5′ end of the 4,740-bp genomic fragment and ending at the TI of the gene were fused to the luciferase reporter gene in pGL3-basic vector as illustrated in Fig. 3. Transient transfection assays were performed in OK and NIH3T3 cells. As shown, the full-length promoter fragment (−4035 nt) confers ∼120-fold transcriptional activity compared with pGL3-basic. However, elimination of 574 bp from the 5′ region construct (−3458) further increased transcriptional activity to 270-fold. Subsequent deletions to −1514 nt did not reduce this highest level of transcriptional activation. Further deletions to −954 reduced activity to ∼120-fold, and deletions to −570 further reduced activity to ∼40-fold. These findings motivated the subsequent experiments to identify a potential mechanism whereby the distal 574-bp fragment could inhibit promoter activity.
Mechanism for Inhibition of Promoter Activity by the Distal Fragment
Sequence analysis of the Npt2 promoter revealed the presence of a pair of inverted repeat sequences of 137 bp and 127 bp at positions −3683 to −3547 and at −2222 to −2096, respectively, enclosing an intervening 1324-bp genomic region (Fig. 4 A). The inverted repeats are 95% identical, and include long stretches of A or T (Fig. 4 B). Eight putative consensus recognition sequences for the AP2 transcription factor are contained within the intervening sequence. To examine the significance of the inverted repeat in inhibiting promoter activity, a series of constructs were generated and tested. The distal repeat of −3683 to −3547 was digested and religated into the promoter in reverse orientation. In addition, a series of promoter fragments from the distal part of the promoter were ligated to an SV40 minimal promoter in pGL3 vector as shown in Fig. 5 A and examined for promoter activity. As shown, reversal of the most distal repeat resulted in 200-fold activity, compared with the 100-fold activity of the promoter with the native orientation of the two repeats. In addition, each of the smaller fragments in which one or both of the inverted repeats and intervening sequence had been deleted also yielded greater promoter activity compared with the full-length promoter (Fig. 5 B). The fragment from −4035 to −2420 nt, the fragment from −2420 to −1430 nt, and the fragment from −1430 to −954 nt yielded 22-, 16-, and 6-fold enhancement of promoter activity, respectively, compared with SV40 promoter alone. The results implicated interaction of the inverted repeats in the inhibition of promoter activity.
Role of AP2 Consensus Binding Sites
Ten putative AP2 consensus binding sites are found in the 5′-flanking region of the rat Npt2 promoter at positions −3635, −3586, −3351, −3346, −3191, −3142, −2923, −2873, −1356, and −428. Eight of ten are located within the DNA sequence enclosed by the inverted repeats. To examine the role of the AP2 consensus binding sites in the regulation of Npt2 gene expression, EMSA experiments were performed, using nuclear extracts from OK and NIH3T3 cells and a double-stranded oligonucleotide probe corresponding to the sequence located at −3362 to −3333 nt of the rat Npt2 promoter. This sequence includes two overlapping putative AP2 consensus binding motifs. Fig. 6 A shows the electrophoretic mobility shift of a prominent band using OK but not NIH3T3 nuclear extracts. This complex formation was completely competed by addition of 200-fold molar excess of unlabeled oligonucleotide. Several other nonspecific bands of lower molecular weight were observed using NIH3T3 nuclear extract. No supershift of the band was observed by addition of antibodies raised against human AP2-α and AP2-β (data not shown) that cross-react with mouse and rat AP2-α and AP2-β proteins. Recombinant human AP2-α protein and in vitro synthesized AP2-α, AP2-β, and AP2-γ used in these experiments with the same probe yielded DNA-protein complexes with a different electrophoretic mobility, compared with the complex generated by OK nuclear extract and the labeled oligonucleotide. The complexes generated by the recombinant and in vitro synthesized proteins did supershift when the same antibodies were added (Fig. 6 B).
To examine the role of the putative AP2 consensus binding motifs in enhancing transcription activation in vivo, a set of promoter-reporter constructs were prepared and tested following cellular transfection. Three, five, and six copies of the oligonucleotide used for the EMSA experiments described above were inserted in front of TK minimal promoter in the pBLCAT2 vector. An additional reporter construct was generated by inserting two copies of a 560-bp native promoter fragment (−3316 to −2757) in tandem that included eight AP2 sites altogether, as indicated in Fig. 6 C. Computer analysis of this sequence did not reveal any other transcription factor binding sites beside the AP2 consensus motifs. All constructs were examined in transient transfection assays of OK and NIH3T3 cells. CAT activity measurements showed that six copies of the AP2 oligonucleotide gave 2.7-fold enhancement compared with the minimal promoter in OK cells but not in NIH3T3 cells. In comparison, the two copies of the native promoter fragment that includes eight copies of AP2 consensus binding motifs conferred much greater (8-fold) enhancement of promoter activity. Taken together with the EMSA results, these transfection studies indicate an important role for AP2 consensus binding motifs in cell-specific promoter activation.
Expression of the Npt2 gene is largely restricted to the proximal convoluted tubule of the kidney with little if any expression in other cell types (8). The OK opossum kidney cell line is a continuous cell line in culture that serves as a useful model for studying the expression of certain genes, which are characteristic of epithelial cells of the proximal convoluted tubule (21). Npt2and its encoded product have been among the most extensively studied in this regard. However, relatively little is known about the factors that regulate the proximal tubule cell-specific expression of these and other genes.
In the current study, we have isolated a 4,740-bp fragment of 5′-flanking sequence of the rat Npt2 gene, and demonstrate that this 5′-flanking sequence powerfully drives reporter gene expression following transfection in the OK cell line, but not in other cell lines (NIH3T3 and HEK293). Deletion analysis yielded maximum cell-specific promoter activity with a 5′-flanking fragment beginning at −3458 nt, with a progressive decline in promoter activity concomitant with progressive 5′-flanking deletions. Examination of the entire 5′-flanking sequence revealed multiple potential transcription factor binding sites (see Table 1). Among others, these include, binding motifs for tissue restricted transcription factors expressed in the kidney [e.g. AP2 (25), CEBP (23), E2A (29)] and for signal transduction-mediated regulation [e.g., PPAR-α (4)]. Among the known putative consensus binding motifs, AP2 sites were most abundant, and the initial deletion analysis results led us to focus our attention on these AP2 consensus sites.
Promoter construct −570 containing one AP2 consensus site yielded ∼30-fold promoter activity, while construct −954 containing one additional AP2 site, significantly further augmented promoter activity to ∼120-fold. Maximum promoter activity of native 5′-flankingNpt2 sequence was observed with construct −3458, which includes eight AP2 consensus sites.
These results suggest a progressive increase in promoter activity with progressively available AP2 consensus binding sites. The potential importance of AP2 consensus motifs to OK cell-specific promoter activity was further highlighted by the analysis of additional 5′-flanking sequence. In particular, we noted a nearly perfect 137-nt inverted repeat located at position −3683 to −3547 with 95% inverted homology to a corresponding 127-nt sequence at position −2222 to −2096, with 1,324 bp of intervening sequence. The inverted repeats and intervening sequence contain eight of the total ten AP2 consensus sites in the total promoter sequence subjected to analysis. Furthermore, we found that promoter activity declined in the −4,035-nt construct that includes an additional two AP2 consensus motifs, but which also includes the distal inverted repeat. Moreover, greatest promoter activity was observed using a modified construct including all 10 AP2 consensus motifs, but in which the distal inverted repeat was reversed in orientation. This modified construct in which the inverted repeats were altered to direct repeats yielded 200-fold promoter activity compared with 100-fold activity of the corresponding native promoter construct containing the intact inverted repeats. This direct repeat construct containing all 10 AP2 consensus motifs yielded greater promoter activity than that observed with simple deletion of sequence containing the distal repeat, which itself also includes two AP2 consensus sites. These findings suggest a potential functional role for the inverted repeats in the regulation of promoter activity, through the formation of a possible stem-loop structure resulting from the apposition of corresponding complementary strands of each of the inverted repeats (Fig.7). Such a structure could limit access to transcription factors that bind to sites within the intervening sequence.
Inverted repeats have been described in various eukaryotic DNA sequences, and have been shown to form hairpin-like stem-loop structures, which could modulate sequence accessibility to binding proteins (1, 20, 37, 41). Poly(dA) nucleotide stretches that have been described within some inverted repeats have been shown to modify transcriptional activity in vivo by modulating DNA accessibility to specific transcription factors (17, 31). In the human globin genes, the inverted repeats are separated by 700–800 nt (7), whereas the inverted repeats in the rat Npt25′-flanking region are separated by 1,325 nt.
The possibility that the deleted distal 577-nucleotide sequence in itself contains a silencing element whose removal enhanced promoter activity was excluded by studies in which this fragment was fused upstream to the −570 rat Npt2 promoter construct, and this was tested in transient transfection of OK cells. Under these circumstances, no inhibition of promoter activity was observed (data not shown). Taken together, these results show a progressive contribution to promoter activity of an increasing number of potentially accessible AP2 consensus motifs. This pattern was disrupted for the AP2 sites surrounded by the inverted repeats and restored when the potential effects of the inverted repeats were negated by reversal of orientation or deletion.
Since AP2 consensus binding motifs comprise the most abundant and striking putative binding sites within the intervening sequence and throughout the promoter, we evaluated their functional importance in a separate series of experiments, using two experimental approaches. First, we used EMSA to determine whether there was a distinct pattern of potential transcription factor binding to AP2 site-containing oligonucleotides using nuclear extracts from OK cells and NIH3T3 cells. Indeed, the electrophoretic mobility shift in OK cells revealed a specific band that migrated more slowly than the corresponding bands observed using NIH3T3 cell nuclear extracts. Purified human AP2-α, in vitro transcribed human AP2-α, and murine AP2-β and AP2-γ were all able to bind specifically to the same rat Npt2 promoter-derived AP2 binding consensus site. However, the electrophoretic mobility of the complexes formed using the purified or synthesized AP2 proteins is not the same as was observed using the OK cell nuclear extract. Furthermore, supershift assays using commercial anti-AP2-α binding protein antibodies did not induce a supershift when tested using OK cell nuclear extracts. Taken together, these results indicate the presence in OK cell nuclei of DNA binding protein(s), which specifically recognized the rat AP2 binding motif. However, these results are not conclusive with respect to the identity of these binding proteins and cannot tell us whether they do or do not contain native opossum AP2 binding proteins. The difference in the pattern of the electrophoretic mobility shift could be due to the occurrence of a multimeric transcription factor complex, antigenic differences in opossum AP2 binding proteins compared with those of other species, or other as yet unknown variables. It should be noted that studies of AP2 consensus sites in other promoters have also shown evidence for regulation by binding of as yet unidentified transcription factors, distinct from AP2-α, AP2-β, and AP2-γ (22). Further studies using species-specific reagents and/or nuclear extracts from other species, as well as more direct protein purification or one-hybrid cloning strategies, will be required to clarify the nature of the DNA binding complexes observed in the current study. The importance of identifying the DNA binding proteins involved is motivated by the evidence for the importance of these sites in overall promoter activity.
Further evidence for the importance of the AP2 consensus sites as regulatory elements came from transient transfection studies using constructs containing different numbers of synthetic AP2 binding sites or Npt2 promoter fragments containing only AP2 consensus sites and devoid of other known potential regulatory elements. Two properties of interest were highlighted by these studies. First was the finding of an important effect of the number of tandemly repeated AP2 sites in enhancing promoter activity, similar to the pattern of a progressive contribution of increasing number of AP2 sites to native 5′-flanking promoter activity. Second was the finding that AP2 sites within the context of the native promoter yielded a much greater enhancement of minimal TK promoter activity, compared with the effect of a synthetic set of six tandemly repeated AP2 sites not in the context of the native Npt2 promoter. Taken together, these results highlight the functional role for the AP2 consensus motif and in particular the importance of both the number of repeats of this consensus site and the native promoter context in yielding optimal promoter and transcriptional activity.
The finding of an important role for AP2 consensus sites in proximal tubule gene expression is consistent with previous studies, which reported the occurrence, and in some cases the functional role, for AP2 sites in the promoters of other genes expressed in the proximal tubule, such as γ-glutamyl transpeptidase type II, the plasminogen activator inhibitor type I, and the aquaporin-1 gene. AP2-α and AP2-β are members of the AP2 family of transcription factors, which are expressed in a specific temporally and spatially restricted pattern during embryonic development of the kidney (25). AP2-α levels are highest during embryonic development, and levels are reported to decrease following birth, at which time AP2-β expression increases. Mutual inhibitory roles for AP2-α and AP2-β have been described in which an alternatively spliced transcript of the AP2-α gene was shown to block sequence-specific DNA binding (5). A number of studies have provided evidence for fine regulation of AP2 transcription factor stoichiometry in the regulation of gene expression for a number of genes containing AP2 consensus sites in their promoters. The findings in the current study reinforce the probable role of the AP2 binding sites in proximal tubule cell-specific gene regulation and suggest an important role for stoichiometric effects (24, 40). However, it is possible that related transcription factors other than AP2 may be involved. Furthermore, the results in the current study suggest a potential novel mechanism for secondary structure in modifying the accessibility of AP2 sites to transcription factors in the final determination of promoter activity. Further studies will be required to determine whether the accessibility of AP2 binding sites is subject to physiological regulation in response to hormonal, dietary, and other factors that modify Npt2 gene expression in vivo. Further studies will also be needed to identify the nature of the protein factors that bind to the AP2 binding consensus sites in the Npt2 promoter and promoters for other genes expressed in the proximal tubule. For further analysis of promoter regulation within the same species, it will be important to utilize primary cultures of rat kidney cortex or extracts from rat proximal tubule suspensions.
Taken together, the findings in the current study point to an important role for AP2 consensus binding sites in mediating Npt2 promoter activity. The occurrence and functional effect of a set of inverted repeats flanking multiple intervening AP2 consensus binding sites suggests a potential novel mechanism for regulating promoter activity under varying physiological conditions. The significance of such a mechanism and the involvement of AP2 binding sites in cell-specific gene expression in vivo will require future studies utilizing transgenic experimental models.
We are indebted to Dr. H. Murer for providing the rat Npt2 cDNA and to Dr. R. Buettner for providing AP2-α, AP2-β, AP2-γ cDNAs. We thank Orr Sharpe, Yael Ravel, and Nira Leider for skillful technical assistance.
Address for reprint requests and other correspondence: K. L. Skorecki, Bruce Rappaport Faculty of Medicine and Research Institute, Technion, Israel Institute of Technology, Dept. of Molecular Medicine. PO Box 9649, Bat Galim, Haifa 31096, Israel (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
- Copyright © 2000 the American Physiological Society