Hypertonicity-induced aquaporin-1 (AQP1) expression is mediated by the activation of MAPK pathways and hypertonicity-responsive element in the AQP1 gene.

Aquaporin-1 (AQP1) is a water channel that is induced by hypertonicity. The present study was undertaken to clarify the osmoregulation mechanism of AQP1 in renal medullary cells. In cultured mouse medullary (mIMCD-3) cells, AQP1 expression was significantly induced by hypertonic treatment with impermeable solutes, whereas urea had no effect on AQP1 expression. This result indicates the requirement of a hypertonic gradient. Hypertonicity activated ERK, p38 kinase, and JNK in mIMCD-3 cells. Furthermore, all three MAPKs were phosphorylated by the upstream activation of MEK1/2, MKK3/6, and MKK4, respectively. The treatments with MEK inhibitor U0126, p38 kinase inhibitor SB203580, and JNK inhibitor SP600125 significantly attenuated hypertonicity-induced AQP1 expression in mIMCD-3 cells. In addition, hypertonicity-induced AQP1 expression was significantly reduced by both the dominant-negative mutants of JNK1- and JNK2-expressing mIMCD-3 cells. NaCl-inducible activity of AQP1 promoter, which contains a hypertonicity response element, was attenuated in the presence of U0126, SB203580, and SP600125 in a dose-dependent manner and was also significantly reduced by the dominant-negative mutants of JNK1 and JNK2. These data demonstrate that the activation of ERK, p38 kinase, and JNK pathways and the hypertonicity response element in the AQP1 promoter are involved in hypertonicity-induced AQP1 expression in mIMCD-3 cells.

Aquaporins (AQPs), 1 a family of water channels, function as a water-selective transporting protein in cell membranes (1). Aquaporin-1 (AQP1) was first discovered in human erythrocytes as a water channel for high osmotic water permeability (2,3). In addition to erythrocytes, AQP1 is abundantly present in the epithelium of kidney proximal tubules and descending thin limbs and endothelium of the descending vasa recta (4). AQP1 has been suggested to be important in constitutive water reabsorption, especially in the epithelial cells of the renal medulla.
The AQP1-expressing vasa recta of the renal medulla are critical in generating and maintaining an axial osmotic gradient through the medulla (5). Sodium chloride (NaCl), urea, and water transporters in the inner medullary collecting duct (IMCD) all play an important role in the regulation of solutefree water excretion in the kidney. Although most cells in mammals are not normally stressed by hypertonicity, epithelial cells of the renal medulla are constantly subjected to a hypertonic condition. Specifically, as a consequence of the urinary concentrating mechanism, cells in the renal inner medulla are normally exposed to a variety of high concentrations of NaCl and urea. Hypertonicity, which results from a high concentration of salt and urea, provides a mechanical stress to shrink medullary cells. However, medullary cells adapt to hypertonicity by a variety of responses through an acute influx of NaCl and water (6), chronic accumulation of organic osmolytes (7), and acute activation of immediate early and heat shock genes (8,9). Although the amounts of total RNA transcription, DNA synthesis, and protein synthesis are significantly decreased by hypertonicity, a limited number of genes are upregulated by hypertonicity-induced transcription (10,11).
The cellular and molecular mechanisms of aquaporin-2 (AQP2) regulation via vasopressin in the collecting duct of the kidney are well understood (12)(13)(14). However, the AQP1 regulation in the kidney is still unknown. Studies performed in the human primary proximal tubule epithelial cell (15) and mouse medullary cell line (mIMCD-3) (16) demonstrate that AQP1 mRNA and protein expressions were up-regulated by hypertonicity. It was also reported that the AQP1 transcript was upregulated by a hypertonicity response element (HRE) in the promoter region of AQP1 gene that was located at Ϫ54 to Ϫ46, which responded to hypertonic stimulation (17). This HRE sequence is different from previously reported tonicity-responsive enhancer (TonE) consensus sequence, which is also responsible for hypertonicity. Additional studies in mIMCD-3 cells showed that all three MAPKs (ERK, p38 kinase, and JNK) were activated by hypertonicity (18,19). Thus, it was possible that the osmoregulation of AQP1 may be mediated by MAPK pathways, which are induced by hypertonicity. However, the role of MAPKs in hypertonicity-induced AQP1 expression remains unclear.
The objective of the present study was to elicit the cellular and molecular mechanisms of hypertonicity-induced AQP1 expression in renal medullary cells. We show here that hypertonicity-induced AQP1 expression is mediated by the activation of the ERK, p38 kinase, and JNK pathways in mIMCD-3 cells.
Functional analysis of the AQP1 promoter containing the HRE sequence also demonstrated that the AQP1 promoter activity was stimulated by hypertonicity and regulated by MAPK pathways.

EXPERIMENTAL PROCEDURES
Cell Line and Culture Condition-mIMCD-3 cells used in this study were obtained from the American Type Culture Collection. Cells were cultured at 37°C and 5% CO 2 in Dulbecco's modified Eagle's/F-12 medium supplemented with 10% fetal bovine serum.
Immunoblot Analysis-mIMCD-3 cells grown on 6-cm dishes were washed with ice-cold phosphate-buffered saline and suspended with 10 mM Tris-HCl, pH 7.5, containing 200 mM sucrose and homogenized by 10 passages through a 27-gauge needle. The homogenate was centrifuged at 3000 ϫ g for 10 min at 4°C. The supernatant was collected, and protein concentration was measured using the Bradford protein assay method (Bio-Rad protein assay kit). Protein (10 g) from the cell extract was resolved on a 12% SDS-polyacrylamide gel and transferred to a polyvinylidene fluoride membrane. The membrane was incubated with a rabbit polyclonal anti-human AQP1 antibody for 1 h. As a protein loading control, the membrane was also incubated with anti-␤-actin antibody (Sigma). For the analysis of MAPK signaling pathways, cells were lysed with ice-cold lysis buffer (50 mM ␤-glycerophosphate, pH 7.2, 0.5% Triton X-100, 0.1 mM sodium vanadate, 2 mM MgCl 2 , 1 mM EGTA, 1 mM dithiothreitol). The lysates were resolved on a 12% SDS-polyacrylamide gel and transferred to a polyvinylidene fluoride membrane. The membranes were incubated with antibodies to phospho-ERK, ERK, phospho-p38, p38, phospho-JNK, JNK, phospho-mitogen-activated ERK kinase (MEK) 1/2, MEK1/2, phospho-MAPK kinase (MKK) 3/6, MKK3, phospho-MKK4 (Cell Signaling Technology Inc., Beverly, MA), or MKK4 (StressGen, Victoria, British Columbia, Canada). After washing, the membrane was incubated with anti-rabbit IgG horseradish peroxidase secondary antibody (Amersham Biosciences). The immunoreactive bands were visualized by enhanced chemiluminescence method (PerkinElmer Life Sciences).
Plasmid Construction for Reporter Assay-The plasmid pCAT-basic (Promega, Madison, WI) was used to examine the promoter activity of the 5Ј-flanking region in the human AQP1 gene (17,20,21). The AQP1 promoter construct (Ϫ54/ϩ23; CP-54), which contains the HRE sequence, was generated by PCR and ligated into Hind III and XbaI sites of pCAT-basic vector. The sense primer corresponded to nucleotides Ϫ54 to Ϫ33 and contained an engineered Hind III restriction site. The antisense primer for amplification corresponded to nucleotides ϩ23 to ϩ4 and contained an engineered XbaI restriction site. The CAT construct was confirmed by sequence analysis.
Transient Transfections and CAT Assay-The transient transfection experiment was performed by the modification of previous reports (17, 20 -22). mIMCD-3 cells were resuspended at 2 ϫ 10 7 cells/ml in Dulbecco's modified Eagle's/F-12 medium without serum and transfected by electroporation under 300 V at 500 microfarads in a 0.4-mm cuvette. Each transfection was performed with 10 g of the AQP1 promoter-CAT construct (CP-54) and 5 g of the plasmid pSV-␤-galactosidase (Promega) on a total volume of 300 l. After overnight incubation, cells were incubated with or without 100 mM NaCl. After a 48-h incubation, the cells were harvested, and cell extracts were isolated using the reporter lysis buffer (Promega). ␤-Galactosidase activity was measured in every experiment for normalization. CAT activity was measured using D-threo[dichloroacetyl-1- 14 C]chloramphenicol (Amersham Biosciences) and the phase extraction method.

Hypertonicty-induced AQP1 Expression in mIMCD-3
Cells-A previous study demonstrated that AQP1 expression was induced by hypertonic medium supplemented with NaCl in mouse renal medullary (mIMCD-3) cells (16). To examine the hypertonic induction of AQP1 in mIMCD-3 cells, cells were incubated in medium supplemented with hyperosmotic agents (NaCl and urea) and impermeable osmolytes (raffinose, glucose, sucrose, and sorbitol). After the treatment with hyperosmolar medium, cell lysates were immunoblotted by using anti-AQP1 antibody. AQP1 protein expression was increased by NaCl, raffinose, glucose, sucrose, and sorbitol but not urea (Fig.  1A). These data indicate that AQP1 induction by hypertonicity requires a hypertonic gradient. Next, to investigate the time course for hypertonic induction of AQP1 protein, mIMCD-3 cells were incubated in hypertonic medium by the addition of 100 mM NaCl. Samples were harvested at specific times for immunoblot analysis. As shown in Fig. 1B, AQP1 protein significantly increased by 12 h after exposure to hypertonic medium and peaked at 16 -20 h.
To test the AQP1 induction in medium osmolality, cells were incubated in medium by the addition of NaCl, raffinose, or glucose with different hyperosmolality. A minimum addition of 50 mM NaCl, 100 mM raffinose, or 100 mM glucose to the medium was needed to increase AQP1 protein expression in mIMCD-3 cells, and stronger induction of AQP1 was obtained by the exposure to higher medium osmolality (Fig. 2). However, the addition of 200 mM raffinose to the medium showed much less AQP1 expression than that of 100 and 150 mM raffinose, because the cell viability significantly decreased at that condition. Taken together, these data indicate that AQP1 expression in mIMCD-3 cells is significantly increased in hypertonic condition in a time-and dose-dependent fashion.

Transcriptional and Post-transcriptional Regulations of Hypertonicity-induced AQP1 Expression in mIMCD-3 Cells-
Hypertonicity-induced AQP1 expression may be caused by mRNA induction or stability or protein stability. To distinguish the mechanism, cells were pretreated with the RNA synthesis inhibitor actinomycin D or protein synthesis inhibitor cycloheximide and then incubated under isotonic or hypertonic condition. The expression levels of AQP1 mRNA and protein were examined by reverse transcription-PCR and immunoblot analysis, respectively. As shown in Fig. 3, pretreatment with actinomycin D or cycloheximide completely inhibited NaCl-induced AQP1 mRNA and protein expressions. These results suggest that both transcriptional and post-transcriptional regulations are required for AQP1 induction by hypertonicity.
ERK, p38 Kinase, and JNK Activation by Hypertonicity in mIMCD-3 Cells-Previous works show that all three MAPKs (ERK, p38 kinase, and JNK) can be activated by hypertonicity  (18,19). To determine whether MAPK-mediating signaling was involved in hypertonic induction of AQP1 in mIMCD-3 cells, antibodies specific for either the phosphorylated or total form of each of three MAPKs were used. Cells were treated with hyperosmotic condition containing 100 mM NaCl or 200 mM urea for 10 min, and then the phosphorylation of all three MAPKs were investigated by immunoblot analysis. NaCl significantly activated the ERK, p38 kinase, and JNK pathways in mIMCD-3 cells, whereas urea activated only the ERK pathway (Fig. 4). These data confirmed the results reported previously (18,23).
Next, activation of MAPKs is mediated by MEK or MKKs. ERK is activated by MEK1 and MEK2, whereas p38 kinase is activated by MKK3 and MKK6, and JNK is activated by MKK4 (24 -28). To further investigate the differential activation of MAPKs in mIMCD-3 cells in response to hypertonicity, the phosphorylation for the upstream activation of MAPKs was assessed using antibodies specific for the phosphorylated or total form of MEK1/2, MKK3/6, or MKK4. As shown in Fig. 4, NaCl activated the MEK1/2, MKK3/6, and MKK4 pathways, whereas urea activated only the MEK1/2 pathway. Thus, NaCl significantly stimulated the phosphorylation of all three subfamilies of MAPKs in mIMCD-3 cells, but urea only the ERK phosphorylation.
Inhibition by MAPK Inhibitors on Hypertonicity-induced AQP1 Expression-To determine whether the activation of ERK, p38 kinase, and JNK was involved in the regulation of hypertonicity-induced AQP1 expression, a MEK inhibitor, U0126, a p38 kinase inhibitor, SB203580, and a novel specific JNK inhibitor, SP600125 (29,30), were used. To investigate the specificity of each MAPK inhibitor, mIMCD-3 cells were pretreated with the different concentrations of U0126, SB203580, or SP600125, and then NaCl-induced ERK, p38 kinase, and JNK phosphorylations were analyzed by immunoblot. As shown in Fig. 5, all three MAPK inhibitors revealed the specific selectivity for other MAPKs within the range of a given concentration.
On the basis of the result in Fig. 5, we proceeded to assess the effect of MAPK inhibitors on NaCl-induced AQP1 protein expression. Cells were pretreated with each MAPK inhibitor and stimulated with NaCl. Cell lysates were analyzed by immunoblot. Results were normalized to expression of ␤-actin by densitometry of immunoblot (Fig. 6). NaCl-induced AQP1 expression was significantly blocked by U0126 (0.5-2.5 M), SB203580 (0.5-2.5 M), and SP600125 (2-10 M). The treatment of a higher concentration with each MAPK inhibitor almost completely abolished NaCl-induced AQP1 expression (data not shown). Therefore, this result provides that AQP1 induction by hypertonicity requires the activation of ERK, p38 kinase, and JNK pathways.
Inhibition by Dominant-negative Mutants of JNK Isoforms on Hypertonicity-induced AQP1 Expression-As an alternative approach to the involvement of JNK pathway on hypertonicity- were incubated in hypertonic medium containing 100 mM NaCl for 10 min. Cells were harvested, and cell lysates were prepared using lysis buffer. Immunoblot analysis was performed by using antibodies to phospho-ERK, ERK, phospho-p38 kinase, p38 kinase, phospho-JNK, or JNK.

FIG. 2. Dose-dependent induction on hypertonicity-induced AQP1 expression in mIMCD-3 cells.
Cells were incubated in medium with different hyperosmolality for 16 h. Cells were harvested, and total protein was analyzed by immunoblot.

FIG. 3. Effects of RNA and protein synthesis inhibitors on hypertonicity-induced AQP1 expression in mIMCD-3 cells. Cells
were incubated without or with 100 mM NaCl for 16 h in the presence or absence of actinomycin D (5 g/ml) or cycloheximide (2 g/ml). A, reverse transcription-PCR analysis was performed for the detection of AQP1 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNAs. B, immunoblot analysis was performed for the detection of AQP1 protein.

FIG. 4. Activation of MAPK and MAPK kinase pathways by NaCl and urea in mIMCD-3 cells.
For the MAPK immunoblotting, cells were incubated in isotonic or hypertonic medium containing 100 mM NaCl or 200 mM urea for 10 min. Cells were harvested, and cell lysates were prepared using lysis buffer. Immunoblot analysis was performed by using antibodies to phospho-ERK, ERK, phospho-p38 kinase, p38 kinase, phospho-JNK, JNK, phospho-MEK1/2, MEK1/2, phospho-MKK3/6, MKK3, phospho-MKK4, or MKK4. induced AQP1 expression, mIMCD-3 cells were transiently cotransfected with the dominant-negative mutants of JNK1 (JNK1-DN) or JNK2 (JNK2-DN) and pSV-␤-galactosidase plasmids. After 16 h of incubation without or with 100 mM NaCl, AQP1 protein expression was examined by immunoblot analysis. ␤-Galactosidase activity was measured in each experiment for correction of transfection efficiency. Results were normalized to expression of ␤-actin by densitometry of immunoblot. As shown in Fig. 7, NaCl-induced AQP1 expression was significantly reduced by both the JNK1-DN-and JNK2-DN-transfected mIMCD-3 cells. Taken together, the activation of ERK, p38 kinase, and JNK pathways is necessary for AQP1 upregulation in mIMCD-3 cells that is responsible for hypertonicity.
Effect of Potent ERK Stimulators on AQP1 Expression-Our results demonstrated that both NaCl and urea activated the ERK pathway. To determine whether the ERK activation is sufficient to induce AQP1 expression, arginine vasopressin, phorbol-12-myristate 13-acetate, or calcium ionophore (ionomycin), which are known as potent stimulators of ERK pathway, were tested. As shown in Fig. 8A, all three potent stimulators indeed activated the ERK pathway. However, they did not induce AQP1 protein expression under isotonic or hypertonic condition (Fig. 8B). Therefore, the ERK activation is necessary but not sufficient to induce AQP1 expression by hypertonicity.
Inhibition of AQP1 Promoter Activity by U0126, SB203580, SP600125, or Dominant-negative Mutants of JNK1 or JNK2-Our previous study demonstrates that a novel HRE is present in the AQP1 promoter region, which is located in the region from Ϫ54 to Ϫ46 (17). This critical element responds to hypertonicity and is required for basal and hypertonicity-induced expression of the AQP1 gene. To investigate the involvement of ERK, p38 kinase, and JNK pathways in hypertonicity-induced AQP1 expression at the transcriptional level, transient transfection experiments were performed using the AQP1 promoterchloramphenicol acetyltransferase construct CP-54 (Ϫ54/ϩ23), which contains the HRE sequence. These results are showed in Fig. 9. When cells were transfected with the CP-54, the treatment with NaCl led to an ϳ3.4-fold increase in the promoter activity, whereas the treatment with urea did not. NaCl-induced promoter activity was attenuated in the presence of U0126, SB203580, and SP600125 in a dose-dependent manner.
In cotransfection experiments, NaCl-induced promoter activity was significantly inhibited by the dominant-negative mutants of JNK1 (JNK1-DN) or JNK2 (JNK2-DN). These results indicate that the HRE in AQP1 promoter is essential for the hypertonicity-induced regulatory effects by ERK, p38 kinase, and JNK pathways on AQP1 expression. DISCUSSION In the book From Fish to Philosopher, Smith suggests that the capacity to conserve solute-free water excretion by activating a urinary concentrating mechanism was an important factor in the evolution from fresh water fishes to land-dwelling  DN) and pSV-␤-galactosidase plasmids followed by treatment without or with 100 mM NaCl for 16 h. Cells were harvested, and total protein was analyzed by immunoblot. The membranes were also incubated with anti-␤-actin antibody as a protein loading control in immunoblot. Transfection efficiency to each experiment was determined by ␤-galactosidase activity. Each protein blot was analyzed by densitometry. AQP1 expression in 100 mM NaCl was assigned as 100%. Values represent the mean Ϯ S.E. of three independent sets of experiments.

FIG. 8. Effect of potent ERK stimulators on AQP1 expression in mIMCD-3 cells.
A, ERK activation. Cells were incubated with arginine vasopressin (AVP; 10 or 100 nM), phorbol-12-myristate 13acetate (PMA) (1 or 10 nM), or ionomycin (10 or 100 nM) for 10 min. Cells were harvested, and cell lysates were prepared using lysis buffer. Immunoblot analysis was performed by using antibody specific for the phospho-ERK or ERK. B, AQP1 expression. Cells were incubated without or with 100 mM NaCl for 16 h after pretreated with arginine vasopressin (10 or 100 nM), phorbol-12-myristate 13-acetate (1 or 10 nM), or ionomycin (10 or 100 nM) for 30 min. Cells were harvested, and total protein was analyzed by immunoblot. The data are representative of three independent experiments. Similar results were obtained in the other two experiments.
Aquaporin-1 Induction by Hypertonicity mammals (31). Much has been learned about the mammalian urinary-concentrating mechanism, including the osmotic and non-osmotic release of the antidiuretic hormone, vasopressin (32), the countercurrent concentrating and exchange mechanisms (33), the cloning of the collecting duct vasopressin V2 receptor (34), and most recently, the cloning of the renal water channels, termed aquaporins (35). Although the up-regulation and trafficking to the collecting apical membrane of AQP2 is a critical factor in the urinary-concentrating mechanism (36), there is recent evidence implicating an important role of AQP1. AQP1 resides in the proximal tubule, descending thin limb of Henle's loop and endothelial cells of the vasa recta (4). Mutations of the AQP1 gene both in knockout mice (37) and humans (38) have been shown to exhibit a defect in maximal urinary concentrating capacity. A recent study shows that a hypertonic NaCl environment in mouse medullary cells is associated with an up-regulation of AQP1 (16). The mechanism for this effect has, however, not been elucidated. Such an up-regulation of AQP1 protein expression would likely enhance the urinary concentrating mechanism, since mutations of AQP1 impair urinary concentration. The present study was undertaken to examine the cellular and molecular mechanisms on the effect of hypertonic stress to up-regulate AQP1 protein expression.
The initial study was performed with mouse medullary (mIMCD-3) cells incubated in medium supplemented with hyperosmotic NaCl or urea as well as with medium containing impermeable osmolytes including raffinose, glucose, sucrose, and sorbitol. AQP1 protein expression was increased in medium containing all of these substrates except urea. Because urea crosses cell membrane freely and, thus, does not create an osmotic gradient, these results implicate a hypertonic gradient in this induction of AQP1 protein expression. Further studies demonstrated the time-and dose-dependent effects of the hypertonic gradient using NaCl, raffinose, or glucose to induce AQP1 protein expression. The actinomycin D and cyclohexi-mide treatment experiments suggested the respective involvement of transcriptional and post-transcriptional regulations in hypertonicity-induced AQP1 expression. A recent study demonstrated that AQP1 protein stability was crucial in AQP1 induction by hypertonicity in mouse BALB/c fibroblasts, which was mediated by the proteasome degradation pathway (39). Therefore, we conclude that AQP1 induction by hypertonicity occurs at both the transcriptional and posttranscriptional levels.
Hypertonicity activated all three MAPK signaling pathways (ERK, p38 kinase, and JNK) in mIMCD-3 cells. These signaling pathways were shown to mediate hypertonicity-induced AQP1 expression. Specifically, it was demonstrated that the blockade of any one of ERK, p38 kinase, or JNK signaling pathway by a specific inhibitor significantly reduced hypertonicity-induced AQP1 expression. U0126, SB203580, or SP600125 selectively inhibited the ERK, p38 kinase, or JNK pathway in a dose-dependent fashion, respectively. Interestingly, the partial ERK inhibition by SB203580 was observed at 5 M but not 2.5 M (data not shown). The maximal dose used in the present study was 2.5 M. The similar observation was previously reported by Yang et al. (23) and suggested a potential interrelationship between ERK and p38 kinase activation. Although SP600125, which was recently identified as a specific JNK inhibitor, blocked the JNK activation in a dose-dependent fashion, it has been demonstrated that this inhibitor also had a partial p38 inhibition at a concentration higher than the maximal dose used in the present study (29). It should be noted that mIMCD-3 cells are highly sensitive to pharmacological inhibition of MAPK signaling pathway. Furthermore, we showed that the blockade of JNK pathway by dominant-negative mutants of JNK1 and JNK2 as an alternative approach significantly reduced hypertonicity-induced AQP1 expression. Taken together, our findings suggest that the activation of all three MAPK is indispensable for AQP1 induction by hypertonicity.
In mIMCD-3 cells, although NaCl activated the ERK, p38 kinase, and JNK, urea activated only the ERK. Although both NaCl and urea activated the ERK, only NaCl induced AQP1 expression. Furthermore, other ERK stimulators such as arginine vasopressin, phorbol-12-myristate 13-acetate, and ionomycin did not induce AQP1 expression under isotonic or hypertonic condition. These results indicate that the ERK activation is necessary but not sufficient to induce AQP1 expression by hypertonicity. A similar study also demonstrates that hypertonicity-induced AQP5 expression in mouse lung epithelial cells was mediated by the ERK-dependent pathway, although the ERK activation was not sufficient for AQP5 induction by hypertonicity (40). The detailed regulatory mechanism via MAPK pathways in hypertonicity-induced AQP1 expression is still unclear. Our data suggest that the three MAPKs may be necessary to phosphorylate transcriptional factors that bind to hypertonicity-responsive elements in the AQP1 gene.
The induction of AQP1 gene by hypertonicity results from a sequential MAPK-signaling pathway that directly affects the promoter region of AQP1 gene. Previous studies demonstrate that the genes of the sodium/myo-inositol cotransporter (41), sodium/chloride/betaine cotransporter (42), and aldose reductase (43) are up-regulated in response to hypertonicity. The tonicity responsive enhancer (TonE) consensus sequence present in these genes mediates the transcriptional stimulation. It seems that AQP1 induction by hypertonicity is mediated by an osmotic response element such as TonE. However, no TonE consensus sequence is present in a 1.8-kilobase AQP1 promoter region (20). We recently characterized a novel HRE in the AQP1 gene that is different from TonE consensus sequence (17). Using the AQP1 promoter-chloramphenicol acetyltrans- ferase construct that contains the HRE sequence, we showed that any of three MAPK specific inhibitors and dominantnegative mutants of JNK1 or JNK2 significantly reduced NaClinduced AQP1 promoter activity in a dose-response manner. In contrast, urea did not induce the promoter activity. In addition, the promoter activity was not affected by U0126, SB203580, or SP600125 under isotonic conditions. 2 Thus, the novel HRE, which only responds to hypertonicity, is essential for the transcriptional regulation of hypertonicity-induced AQP1 expression. Identification of a trans-acting HRE-binding protein that binds to the HRE should further elucidate the transcriptional regulation of hypertonicity-induced AQP1 expression.
In summary, the present results demonstrate that hypertonicity-induced AQP1 expression is regulated by ERK, p38 kinase, and JNK activation and the HRE in the AQP1 promoter. Inhibition of each MAPK pathway results in significant reduction of hypertonicity-induced AQP1 expression, indicating that all three MAPK signaling pathways are indispensable for AQP1 induction.