The TonE/TonEBP pathway mediates tonicity-responsive regulation of UT-A urea transporter expression.

The rat renal urea transporter UT-A includes four isoforms. UT-A1, UT-A3, and UT-A4 are transcribed from a single initiation site at the 5'-end of the gene; a distinct internal initiation site is used for UT-A2 transcription. We cloned 1.3 kilobases (kb) of the 5'-flanking region upstream of the transcription start site of UT-A1, UT-A3, and UT-A4. This region contains three CCAAT sequences but lacks a TATA motif. A tonicity-responsive enhancer (TonE) was identified at -377bp. The 1.3-kb full fragment subcloned into pGL3 vector induced luciferase activity in Madin-Darby canine kidney cells and in mouse inner medullary collecting duct cells in isotonic medium. Luciferase activity was increased significantly in hypertonic medium, whereas deletion or mutation of the TonE sequence abolished this response. Electrophoretic mobility shift assay using the 5' UT-A TonE sequence as DNA probe showed formation of a specific DNA-protein complex with nuclear extracts from cells exposed to hypertonic medium and was weakly detectable in isotonic controls. A supershift in the mobility of the DNA-protein complex was observed with antiserum targeted to the TonE-binding protein (TonEBP). Co-transfection with dominant-negative TonEBP abolished the luciferase activity induced by the UT-A 1.3-kb construct under hypertonic and isotonic conditions. These data suggest that the TonE/TonEBP pathway mediates tonicity-responsive transcriptional regulation of UT-A1, UT-A3, and UT-A4 expression.

Excretion of highly concentrated urine allows mammals to eliminate waste products while preserving body water to survive on dry land. Urea, the major end product of protein metabolism, and sodium chloride contribute the most to maintaining high osmolarity in the renal inner medulla. In this region of the kidney, high levels of urea are present in both intra-and extracellular compartments. Whereas urea can be viewed as an osmolyte, in high concentrations this solute is known to desta-bilize essential enzymatic activities (1,2). In this respect, urea differs substantially from non-perturbing osmolytes such as polyols, methylamines, and amino acids, which protect renal medullary cells during hypertonic stress (3)(4)(5)(6). Information about the adaptation of renal medullary cells to urea has begun to emerge (7,8). Epithelial cells of the inner medullary collecting duct (IMCD) 1 and descending thin limbs of Henle's loop are capable of rapidly transporting urea by a facilitated transport process that is mediated by the renal urea transporter UT-A. The mechanisms involved in regulating urea transport in these nephron segments are for the most part still unclear.
Four rat UT-A isoforms encoded by a single gene have been identified (9 -11). UT-A1, UT-A3, and UT-A4 are transcribed from a transcription site at the 5Ј-end of the gene, whereas UT-A2 is transcribed from a distinct internal transcription start site (12). UT-A1, UT-A3, and UT-A4 are expressed in the highly hypertonic renal inner medulla whereas UT-A2 is normally expressed in the outer medulla. Additional variants of UT-A1, UT-A2, and UT-A3, referred to as UT-A1b, UT-A2b, and UT-A3b, have been identified that include alternative 3Јuntranslated sequences but show a distribution similar to the parent transcript (12).
A consensus sequence for the TonE has been described (13), and the trans-activating factor for the TonE cis-element, TonEBP, was recently cloned (14). This pathway mediates increased transcription of genes involved in the accumulation of compatible osmolytes by renal cells in response to hypertonicity, (15)(16)(17)(18). The role of this tonicity-regulated pathway in other aspects of inner medullary cell physiology is still unclear.
The purpose of this study is to characterize part of the 5Ј-flanking region of the rat UT-A gene, to search for promoter activity and to identify factors that may be important in the transcriptional regulation of UT-A1, UT-A3, and UT-A4 expression in the renal medulla.
A BAC rat genomic library was also screened. Seven BAC clone membranes were obtained from the BAC-PAC RESOURCES (Department of Cancer Genetics, Roswell Park Cancer Institute, Buffalo, NY). Membranes were screened with 32 P-radiolabeled gene specific probes spanning the region from intron 1 to exon 2. The DNA template, including Ϫ63 to ϩ287 of UT-A1 cDNA, was radiolabeled using Random Prime DNA labeling kit (Roche Molecular Biochemicals). Hybridization screening was performed according to the protocol recommended by the manufacturer. After overnight hybridization at 65°C, membranes were washed with 2ϫ SSC, 0.1% SDS at room temperature for 20 min, followed by 0.1ϫ SSC, 0.1% SDS at 65°C, twice, and exposed overnight at Ϫ70°C on Hyper film TM MP (Amersham Pharmacia Biotech). Two positive clones were identified and characterized by PCR, restriction analysis, Southern hybridization, and sequencing. Sequences of the 5Ј-region upstream of exon 1 were compared among two different BAC clones, the genomic clone identified with the Genome Walker kit, and rat genomic DNA (CLONTECH, Palo Alto, CA), to verify identity. Analysis of DNA sequences was performed using the Wisconsin Sequence Analysis Package and Lasergene software packages. Potential binding sites for transcription factors were searched using the GeneQuest program.
Reporter Plasmid Construction-To analyze promoter activity, deletions from the UT-A 1.3-kb 5Ј-flanking region identified from the genomic clones were made by PCR, with sense primers corresponding to Ϫ1258, Ϫ943, Ϫ616, Ϫ423, and Ϫ244 nt of the 1.3 kb sequence and an antisense primer corresponding to nt 61-85 of UT-A1 cDNA. PCRamplified products were subcloned into pGL3 basic luciferase reporter vector (Promega).
For site-directed mutagenesis, the wild-type TonE sequence 5Ј-TG-GAAAACTCC-3Ј in the pGL3 constructs was mutated by PCR using the sense primer 5Ј-gaattccaatGGAGTCCCTCAccctgaa-3Ј (the TonE sequence is in capital letters, and the mutated nucleotides are underlined) and the mutated antisense primer 5Ј-cagggtgAGGGACTCCattggaattc-3Ј. The resulting constructs carried the mutated TonE sequence 5Ј-TGAGGGACTCC-3Ј. These constructs were sequenced for verification.
Cell Culture, Transfection, and Determination of Promoter Activity-Madin-Darby canine kidney (MDCK) cells (American Type Culture Collection, Manassas, VA) were grown in minimum essential medium supplemented with 100 IU/ml penicillin, 100 IU/ml of streptomycin, 1 mM of sodium pyruvate, 1ϫ essential amino acid, and 10% fetal bovine serum (Mediatech, Herndon, VA). Mouse IMCD3 cells (American Type Culture Collection) were maintained in Dulbecco's modified essential medium/F-12 medium, supplemented with 100 IU/ml penicillin, 100 IU/ml of streptomycin, and 10% fetal bovine serum. NaCl was added to isotonic medium to raise the osmolarity up to 600 mosmol/kg H 2 O. Osmolarity was measured using a vapor pressure osmometer (Model 5500, Wescor, Inc., Logan, UT).
For transfection studies, cells were seeded 24 h before transfection in 12-well clusters (Corning, Marietta, GA). Cells in the control group were maintained in isotonic medium (300 mosmol/kg H 2 O) after the transfection. Cells in the hypertonic group were switched to a 600 mosmol/kg H 2 O hyperosmotic medium 24 h after transfection for 48 h. pGL3 constructs were transfected into the cells using Fugene 6 Transfection Reagent (Roche Molecular Biochemicals) according to the protocol suggested by the manufacturer. Transfection efficiency was determined by co-transfecting cells with an enhanced green fluorescent protein vector pEGFP and by counting the cells expressing fluorescent protein under fluorescent light 24 h after transfection. Transfection efficiency was estimated to be about 20% with this method.
For determination of promoter activity, 0.5 pmol (0.6 -1.2 g) pGL3 reporter construct and 10 ng of pRL-SV40, to control for transfection efficiency, were co-transfected into 50% confluent cells. After 72 h, the cell lysate was extracted with passive lysis buffer (Promega). Firefly luciferase activity from pGL3 reporter vector and Renilla luciferase activity from pRL-SV40 were measured by the Dual Luciferase assay system (Promega) on a TD-20/20 Luminometer (TURNER DESIGN, Sunnyvale, CA). Promoter activity was reported as the ratio between Firefly and Renilla luciferase activities in each sample. Differences between experimental and control cell groups were analyzed by 2-tailed, non-paired t test, with p Ͻ 0.05 indicative of statistical significance.
Preparation of Nuclear Extracts, Electrophoretic Mobility Shift Assay (EMSA), and Supershift Assay-MDCK or mIMCD3 cells were seeded in 10-cm dishes and when 50% confluent were switched to hypertonic medium (600 mosmol/kg H 2 O). Cells were maintained in hypertonic medium for 48 h before harvesting; control cells were maintained in isotonic medium unless otherwise indicated. After washing with chilled phosphate-buffered saline, cells were scraped in 1 ml of cold phosphatebuffered saline on ice and transferred into a pre-chilled 1.5-ml Eppendorf tube. Cells were pelleted by 3,500 rpm centrifugation at 4°C, and the supernatant was discarded. Cell pellets were washed one time in 500 l of buffer (10 mM HEPES buffer, pH 7.9, 1.5 mM MgCl 2 , 10 mM phenylmethylsulfonyl fluoride; and 1 mM dithiothreitol), centrifuged at 3500 rpm, and resuspended in 80 l of the same buffer with 0.1% (v/v) Triton X-100 (Sigma) after the removal of the supernatant. After cell membrane disruption by 8 strokes with a Dounce glass homogenizer, nuclei were pelleted by centrifugation, washed, and resuspended in a buffer containing 20 mM HEPES, pH 7.9, 25% glycerol (v/v), 420 mM NaCl, 1.5 mM MgCl 2 , 0.2 mM EDTA, 1 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, 1.8 g/ml aprotinin, and 5 g/ml leupeptin. After a 30-min incubation on ice, samples were centrifuged at 14,000 rpm for 10 min at 4°C, and the supernatant was saved as nuclear extract at Ϫ80°C until usage.
For EMSA, each 5 g of nuclear extract was incubated for 30 min at room temperature with 32 P-radiolabeled oligonucleotide probes: wildtype TonE probe, 5Ј-gaattccaatGGAGTTTTCCAccctg-3Ј (TonE is in capital letters) and mutated TonE probe, 5Ј-gaattccaatGGAGTCCCTCAccctg-3Ј (mutated nucleotides are underlined), in the reaction buffer containing 12 mM HEPES, pH 7.9, 4 mM Tris-Cl, pH 7.5, 1 mM EDTA, pH 8.0, 1 mM dithiothreitol, 60 mM KCl, 12% glycerol (v/v), and 1 mM phenylmethylsulfonyl fluoride. The same reaction was also performed in the presence of 10-to 500-fold excess of wild-type or mutated nonradiolabeled probe to test specific binding. Electrophoresis was performed with a 5% polyacrylamide gel followed by autoradiography.
Supershift assay using specific anti-TonE-binding protein antiserum was performed to confirm the specific binding between the TonE DNA sequence and the TonE-binding protein, TonEBP, as described previously (14). Antiserum for TonEBP and a dominant-negative form of TonEBP, subcloned into the pCDNA vector, were kindly provided by Dr. H. Moo Kwon (Johns Hopkins University). The pGRE-tk promoter construct (19) was kindly provided by Dr. Jie Du (Emory University) and was used as a TonE-independent control promoter.

Analysis and Characterization of the 1.3-kb Segment from the 5Ј-Flanking Region of the UT-A Gene-Previously
, we identified a transcription start site for UT-A1, UT-A3, and UT-A4, localized 21-bp upstream of the beginning of the 5Ј-untranslated region of UT-A1, distinct from the one for UT-A2 (12). We characterized about 1.3 kb from the 5Ј-flanking region of the rat UT-A gene. This sequence (GenBank TM /EBI accession number AF214483) was identified in 3 independent genomic clones (BAC no. 409 L-16, BAC no. 578 J-7, and GW-1 Genome Walker Kit clone) upstream of the exon encoding the first 169 bp of UT-A1, UT-A3, and UT-A4. BAC no. 409 L-16 and BAC no. 578 J-7 include the two exons encoding the first 300 bp of these three isoforms (Fig. 1). This 1.3-kb segment of genomic DNA was fully sequenced from both strands and found to be 100% identical among the different genomic clones (Fig. 2). Sequence analysis of this region reveals three CCAAT sequences, two consensus sequences for AP-1, one for AP-2, three for the glucocorticoid response element GRE, and other putative cis-elements but no TATA motif. Interestingly, we identified a consensus sequence for TonE at Ϫ377-bp upstream from the transcription start site.
Identification of Tonicity-sensitive Promoter Activity-To test the promoter activity of this region we used MDCK and mIMCD3 cells. Although mIMCD3 cells can express UT-A2 after prolonged exposure to hypertonic medium (20), neither mIMCD3 nor MDCK cells express UT-A1, UT-A3, or UT-A4 transporters. Both cell lines provide an established in vitro model to study the epithelium of the distal nephron, and MDCK cells are known to express tonicity-responsive genes, like the betaine/␥-aminobutyric acid (GABA) cotransporter (BGT1) (21). The entire 1.3-kb 5Ј-flanking region, as well as several deletions, subcloned into pGL3-Basic vector, were transfected into MDCK cells and/or mIMCD3 cells to measure the luciferase activity as a function of promoter activity. The 1.3-kb construct transfected into MDCK cells induced basal promoter activity in isotonic medium (0.93 Ϯ 0.03 F/R luciferase activity Ϯ S.D., n ϭ 3), which increased to 2.78 Ϯ 0.18 after exposure to hypertonic medium for 48 h, resulting in a 2.8-fold increase (Fig. 3A). Similar results were observed in mIMCD3 cells, where in isotonic medium the 1.3-kb construct induced detectable promoter activity (0.32 Ϯ 0.01), which increased to 0.91 Ϯ 0.02 after exposure to hypertonic medium for 48 h resulting in a 2.9-fold increase. The effect of progressive deletions was tested in MDCK cells (Fig. 3A). The largest deletion, a 300-bp insert devoid of TonE sequence, induced negligible luciferase activity in hypertonic medium and in isotonic control medium. However, after 48 h in hypertonic medium, each construct carrying the TonE sequence showed significantly higher luciferase activity compared with isotonic control. Mutation of the TonE sequence in the 1.3-kb pGL3 construct completely abolished the luciferase activity increase observed in hypertonic medium, but did not affect luciferase activity in isotonic medium (Fig. 3B). These observations indicate that the TonE motif found at Ϫ377 in the 5Ј-flanking region of the UT-A gene is active and may be involved in regulating UT-A promoter activity in response to hypertonicity.
Detection of TonE/TonEBP Binding-By gel shift assay, a DNA-protein complex was demonstrated in mIMCD3 cells and in MDCK cells grown in hypertonic medium but was barely detectable in isotonic medium (Fig. 4). No binding was observed with a mutated TonE probe. Competition with excess unlabeled probe effectively prevented binding. Higher concentrations of excess unlabeled oligonucleotides in which the TonE sequence was mutated did not prevent binding. These results demonstrate specific binding between TonE and a nuclear protein that is enhanced in hypertonic conditions. To test whether TonEBP could be identified in the DNA-protein complex described above, we performed a supershift assay, using specific antiserum toward TonEBP (Fig. 5). With this assay, we detected a significant upward shift in the mobility of the DNAprotein complex when antiserum was added into the reaction mixture containing nuclear extract from hypertonic cells. (Adding the anti-serum before adding the probe to the reaction mixture produced the same results.) These observations support specific binding between this TonE element in the 5Ј- flanking region of the UT-A gene and TonEBP.
Effect of Dominant-Negative TonEBP-To further examine the role of TonEBP in 5Ј UT-A promoter activity and its response to hypertonicity, we tested the effect of co-transfecting MDCK cells with a dominant-negative form of TonEBP (14) subcloned into pCDNA3 in the presence of the 1.3-kb pGL3 construct (pUT-A) or with the tonicity-independent control promoter construct pGRE-tk (Fig. 6). These experiments were performed in MDCK cells maintained in isotonic medium and in MDCK cells exposed to hypertonicity. Co-transfection with the dominant-negative TonEBP effectively decreased the 5Ј UT-A promoter activity in hypertonic as well as in isotonic condtions, supporting a role of TonE and TonEBP in regulating the 5ЈUT-A promoter activity. Minimal residual UT-A promoter activity was still detected in the presence of TonEDN, probably reflecting incomplete inhibition at the concentration used, as previously observed by Myiakawa et al. (14) with TonEdriven luciferase constructs. TonEDN did not significantly inhibit the activity of the pGRE-tk control promoter. DISCUSSION In this study, we demonstrate that the TonE/TonEBP pathway participates in regulating expression of the rat urea transporter UT-A by stimulating transcription from the 5Ј-region of the UT-A gene in hypertonic conditions. Our findings provide novel evidence that tonicity-responsive regulation mediated by the TonE/TonEBP pathway may affect the expression of genes that are not directly involved in osmoprotection in the renal inner medulla.
We characterized part of the 5Ј-flanking region upstream from the transcription start site for the rat UT-A1, UT-A3, and UT-A4. Our observations are consistent with a TATA-less promoter controlling the transcription from the 5Ј-flanking region of the UT-A gene. Analysis of luciferase expression reveals a modest level of promoter activity induced by transfecting this segment of genomic DNA into MDCK cells and mIMCD3 cells in isotonic conditions. However, the promoter activity is significantly increased in response to hypertonicity in both cell types. Of particular interest in this respect is the identification of a consensus sequence for TonE within approximately 400 bp from the 5Ј-transcription start site. With progressive deletion and mutation analysis, we show that this TonE motif is active in inducing the tonicity-sensitive increase in promoter activity. Our results suggest that this sequence by itself may influence UT-A transcriptional activity. However, analysis of a larger segment of genomic DNA further upstream from this region may reveal other TonE sequences that may additionally stimulate promoter activity as has been observed for other genes (15).
TonE-mediated stimulation of promoter activity involves binding to the trans-activating factor TonEBP, which has been characterized previously (14,22). We identified a protein-DNA complex consistent with TonE/TonEBP binding in nuclear extracts from hypertonic cells, indicating that the 5Ј UT-A TonE was activated through this pathway. The negative effect of co-transfecting dominant-negative TonEBP on the promoter activity of MDCK cells also supports this hypothesis.
Tonicity-sensitive gene transcription, activated through the TonE/TonEBP pathway, has been proposed as a major regulatory mechanism for the expression of genes involved in osmolyte accumulation. This study expands the role of tonicityresponsive transcription mediated by TonE to include regulation of UT-A transporter expression, a gene involved in the transport of the perturbing solute urea that is essential for renal function.
Regulation of transcription by the TonE/TonEBP pathway has mostly been investigated in the context of transient, adaptive responses to a shift from isotonic to hypertonic conditions. A high expression of UT-A1, UT-A3, and of the UT-A3 variant UT-A3b mRNA transcript is detected physiologically in the inner medulla (12). The findings described in this study, together with the above in vivo observations, support the hypothesis that normally a basal level of transcriptional activity is needed to maintain a steadily high expression of UT-A1, UT-A3, UT-A3b, and to a lesser extent of UT-A1b, and UT-A4. A certain degree of constant activity for the TonE/TonEBP pathway could ensure adequate abundance of the UT-A urea transporters and possibly other genes in the hypertonic inner medulla. Interestingly, TonEBP abundance in the nucleus of inner medullary cells increases in water-deprived rats, presumably resulting in activation of tonicity-responsive transcription above basal levels (23).
We recently demonstrated significantly increased expression of UT-A3 and UT-A3b mRNA and negligible increases in UT-A1 and UT-A1b expression in the inner medulla of waterdeprived rats compared with control animals (12). The reason behind the scarce effect of water deprivation on UT-A1 and UT-A1b mRNA abundance is not immediately apparent because both are transcribed from the same region as UT-A3 and UT-A3b, and their transcription should be similarly regulated. One would expect that during water deprivation, a higher tonicity in the inner medulla could activate the TonE/TonEBP pathway and stimulate transcription of UT-A1, UT-A1b, UT-A3, UT-A3b, and UT-A4 to an equivalent extent. Thus, the differential effect of water deprivation on the expression of UT-A1 and UT-A3 observed in rat inner medulla cannot be explained at the transcriptional level. This discrepancy may possibly be caused by factors affecting mRNA stability or splicing events that are currently unknown.
Further studies on the mechanisms regulating UT-A expression will be needed and may also explain why expression of UT-A1, UT-A3, and UT-A4 does not occur under normal culture conditions in any renal epithelial cell line, similar to Aquaporin 2 (24). Lack of expression of an unidentified trans-activating factor, activation of negative cis-elements, or other unknown mechanisms may prevent generation of these UT-A mRNA transcripts in cultured cells. The role of post-transcriptional events, which may influence expression of individual UT-A transporters in vitro and in vivo, remains to be elucidated.