INTRODUCTION

The elevation of renal medullary extracellular NaCl and urea during antidiuresis has been known to be adjusted intracellularly by accumulation of osmolytes such as sorbitol, myoinositol, betaine, and glycerophosphorylcholine. Aldose reductase (AR),¹ an enzyme which catalyzes the reduction of various sugars to alcohols, including glucose to sorbitol, is highly expressed in kidney inner medulla (1, 2), and is believed to play a role in normal renal osmoregulation. A similar role for AR has been reported in plants that accumulate sorbitol during seed development (3). Contrary to the role of AR in osmoregulation, AR-mediated accumulation of sorbitol from excess glucose in diabetic patients has been suggested to be one of the triggering events in diabetic complications, including cataract, retinopathy, neuropathy, and nephropathy (4-6). The inhibition of this enzyme in diabetic animal models, such as galactose-fed rats has been successful in retarding or preventing such complications (7).

Elevation of AR enzymatic activity and protein synthesis in cultured rabbit renal inner medulla cells (PAP-HT25) exposed to hypertonic medium was first reported by Moriyama et al. (2) followed by additional findings that AR induction by hypertonic stress also occurs in other cell types and species by Kaneko et al. (8). Wang et al. (9) analyzed the human AR promoter up to 609 bp from the transcriptional start site using HepG2 cells (9). DNA protection in a CGGAA(A/G) motif (186 to 146), and a GC-rich region (87 to 31) was observed by in vitro DNase I footprinting, but no response to osmotic stress was observed with this promoter region in transfection studies. Ferraris et al. (10) observed a 40-fold increase in promoter activity with a 3.6 kilobase to +34-bp rabbit AR construct under osmotic stress (10). Three recent reports describe the involvement of a tonE-like osmotic response element of the human AR promoter at position 1,235 bp (11), the rabbit AR promoter at 1,108 bp (12), and the mouse AR promoter at 1,053 bp (13). A signal transduction study revealed that the p38 or SAPK/JNK pathway are not necessary for the transcriptional regulation of the AR promoter through osmotic response element (14).

In this paper, we describe the identification of a novel cis-element necessary for the constitutive activity and the osmotic response activity of the rat AR promoter which is located in the proximity of the previously described tonE-like element. This novel AEE element is shown to be occupied in vivo in both transiently transfected templates and in the endogenous promoter, suggesting that a stable association with transcription factors at this site is required for effective rat AR promoter activity.

EXPERIMENTAL PROCEDURES

The rat AR promoter was isolated from a phage rat genomic library (CLONTECH) (15). The 5-flanking sequence of the gene was sequenced up to 3.6 kb (LARK Technologies, Houston, TX). The promoter fragments were amplified by PCR and subcloned into the pGL3 luciferase reporter vector (Promega, Madison, WI) and sequenced. Five additional reporter constructs were made by ligation of 5-KpnI-GAGGGGTGTTGGAAGAGTGCCAAATTTCCGCCATT-KpnI-3 sequence in both sense and antisense orientation, 5-KpnI-AGGGGTGTTGGAAGAGTGCCAAATTTCC-KpnI-3, 5-KpnI-GGGTGTTGGAAGAGTGCCAAATTT-KpnI-3, 5-KpnI-GTGTTGGAAGAGTGCCAAAT-KpnI-3 to the 5 end of the thymidine kinase (TK) promoter (105 to 52 bp) in the pGL3basic luciferase reporter vector (16).

A normal rat liver cell line (Clone 9, ATCC, Rockville, MD) which expresses AR and responds to osmotic stress (data not shown) was cultured in 35-mm diameter culture wells (Falcon 3046) in Ham's F-12K medium (Life Technologies, Inc.), 10% fetal bovine serum (Life Technologies, Inc.), and 50 µg/ml gentamicin (Life Technologies, Inc.). When cells were 60% confluent, 1.7 µg of luciferase construct plasmid and 0.7 µg of pSV--galactosidase vector were transfected by the CaCl₂ precipitation method (Profection mammalian transfection system, Promega) without osmotic shock. After 48 h, six of the wells received normal medium and the other six received hypertonic medium (medium supplemented with 150 mM NaCl; 600 mosm/kg H₂O). Cells were harvested after 18 h, followed by luciferase assay and -galactosidase assay according to the manufacturer's instructions (Dual-Light, Tropix, Bedford, MA).

Exonuclease III (ExoIII)-mediated footprinting was carried out as described by Archer et al. with slight modification (17-19). Cells were cultured in 10-cm diameter dishes up to 60% confluency and were then transfected with 10 µg of the luciferase construct 4, which contains 1,094 bp of upstream AR promoter sequence. After 6 h, the cells were washed with medium and grown an additional 48 h. Cells were scraped and washed with phosphate-buffered saline followed by suspension in 5 ml of homogenization buffer (10 mM Tris-HCl, pH 7.4, 15 mM NaCl, 60 mM KCl, 1 mM EDTA, 0.1 mM EGTA, 0.1% Nonidet P-40, 0.15 mM spermine, 0.5 mM spermidine, 5% sucrose). The suspension was homogenized in a Dounce homogenizer and 1 ml of sucrose solution (10 mM Tris-HCl, pH 7.4, 15 mM NaCl, 60 mM KCl, 0.15 mM spermine, 0.5 mM spermidine, 10% sucrose) was added. After centrifugation at 1,400 × g for 20 min the pelleted nuclei were washed once with wash buffer (10 mM Tris-HCl, pH 7.4, 15 mM NaCl, 60 mM KCl, 0.15 mM spermine, 0.5 mM spermidine). Isolated nuclei (4.5 × 10⁵) were resuspended in 200 µl of enzyme digestion buffer (10 mM Tris-HCl, pH 7.4, 15 mM NaCl, 60 mM KCl, 0.1 mM EDTA, 5 mM MgCl₂, 5% glycerol, 1 mM dithiothreitol) and treated at 30 °C for 15 min with 200 units of Asp718 or HindIII restriction enzyme (Boehringer Mannheim) and 650 units of ExoIII (Boehringer Mannheim). One ml of proteinase K buffer (10 mM Tris-HCl, pH 7.6, 10 mM EDTA, 0.5% SDS, 0.2 mg/ml proteinase K) was added and incubated at 37 °C for 2 h to digest the nuclei. DNA was purified by phenol/chloroform extraction followed by ethanol precipitation. The DNA was resuspended and treated at 30 °C for 30 min with 50 units of mung bean nuclease (Boehringer Mannheim) in mung bean buffer (50 mM NaOAc, pH 5, 30 mM NaCl, 1 mM ZnSO₄). Twenty micrograms of isolated DNA was used for linear PCR. Primers used for linear PCR were 5-AAGCATGACCCAGCAGAAGGAGA-3 (974 bp to 952 bp, primer 1) for the coding strand and 5-AGTTGCCCCAAGAACAATGGCGGAA-3 (877 bp to 901 bp, primer 2) for the noncoding strand. For control experiments construct 4 was digested with HindIII and linear PCR was performed with primer 1 (Fig. 2, lane 1) or the construct 4 was digested with Asp718 and linear PCR was performed with primer 2 (Fig. 2, lane 3). Linear PCR product were resolved in a 6% sequencing gel (Sequagel-6, National Diagnostics, Atlanta, GA).

Cells were treated with 0.1% dimethyl sulfate for 90 s in both normal and hypertonic medium. DNA was extracted and cleaved by piperidine. Ligation-mediated (LM)-PCR was performed as described previously (20, 21). Primers used for analysis were primer 2 (shown above), 5-TGAACAGGCAGAATCCCATA-3 (747 to 766 bp, primer 3), and 5-CTGAAATAATCGGAGTTGCCCCA-3 (864 to 886 bp, primer 4). After LM-PCR, labeled products were resolved in a 6% sequencing gel (Sequagel-6, National Diagnostics, Atlanta, GA). At least two independent DNA preparations were used.

Whole cell extracts were prepared as described previously (22). Rat liver cells were cultured in three 10-cm diameter dishes to 90% confluency. The cells were then grown in normal medium or hypertonic medium for 30 min or 3 h. After stress, cells were pelleted, washed with phosphate-buffered saline, and resuspended in 2 packed cell volumes of Buffer C (20 mM Hepes, pH 7.9, containing 0.42 M NaCl, 1.5 mM MgCl₂, 0.2 mM EDTA, 0.5 mM dithiothreitol, 25% (v/v) glycerol, 2 mM proteinase inhibitor 4-(2-aminoethyl)-benzenesulfonyl fluoride hydrochloride). The cells were snap frozen in ethanol/dry ice. Frozen cells were quickly thawed for homogenization and centrifuged for 1 h at 34,000 × g. The supernatant was used for EMSA without dialysis.

Whole cell extracts (2-6 µg of protein) were incubated with 0.1 pmol of ³²P-labeled oligonucleotide (10⁵ cpm/reaction mixture) in EMSA binding buffer containing 20 mM Tris-HCl, pH 7.5, 1 mM MgCl₂, 0.2 mM EDTA, 0.5 mM dithiothreitol, 5% (v/v) glycerol, and 2 µg of poly(dI-dC) for 30 min at 4 °C. The oligomers used were: probe A, 5-GAGGGGTGTTGGAAGAGTGCCAAATTTCCGCCATT-3, probe A8, 5-GAGGGGTGTTGAGGAGACACCAAATTTCCGCCATT-3. For the competition experiments, extracts were preincubated at 4 °C for 30 min with a × 100 excess of unlabeled oligomers prior to the addition of labeled probe.

RESULTS

Various luciferase reporter constructs of the rat AR promoter were made and transfected into rat liver cells to measure the promoter activity under normal and hypertonic conditions (Fig. 1). Rat AR constitutive promoter activity was constant for constructs 1 (3,104 to +23 bp) and 2 (2,371 to +23 bp), while further deletions increased the activity up to construct 5 (1,052 to +23 bp) which gave the highest constitutive activity. Deletion to construct 6 (895 to +23 bp) had a significant effect on the reduction of constitutive promoter activity.

[View Larger Version of this Image (21K GIF file)]

Induction of promoter activity by osmotic response of construct 1 was 4.4-fold while deletion to construct 2 resulted in an increase of 11.8-fold (Fig. 1). Further deletion to construct 4 (1,094 to +23 bp) gave the highest absolute promoter activity by osmotic response (100.00 under hypertonic conditions). Construct 5 (1,052 to +23 bp) which lacks the tonE element still maintained an osmotic response of 2.0-fold.

Construct 8, which lacks 8 bp of sequence between the two guanosines protected in vivo (915 to 908 bp) abolished the constitutive promoter activity. Construct 9, which contains all of the sequence of construct 1 except for a 177-bp deletion between 1,071 and 895 bp, abolished the constitutive activity and the osmotic response capability. These results indicated the presence of several regions involved in modulating the AR promoter activity and suggest the presence of an element(s) crucial for the constitutive activity and osmotic response located between 1,071 and 895 bp.

In Vivo Factor Binding to the Region 926 to 895 bp

Based on the luciferase gene expression experiments with construct 4, we tested transcription factor binding between 1,094 and 649 bp by using the ExoIII-mediated in vivo DNA footprinting method. This technique has been used to detect constitutive and hormone-induced factor binding to the mouse mammary tumor virus long terminal repeat (19), and is based on the fact that transcription factors upon binding, protect DNA from digestion with ExoIII nuclease. Under normal osmotic conditions, transient templates (construct 4) in isolated nuclei were digested with ExoIII and Asp718 (to supply an entry point for ExoIII) in the sense-directed digestion or with ExoIII and HindIII (to supply an entry point for ExoIII) in the antisense-directed digestion. A strong block to ExoIII digestion was detected at 926 bp for the sense strand-directed digestion and at 895 bp for the antisense strand-directed digestion (Fig. 2).

[View Larger Version of this Image (75K GIF file)]

To observe whether the ExoIII protected region (926 to 895 bp) was also protected in the endogeneous gene, dimethyl sulfate in vivo footprinting was performed (20). Liver cells were grown in normal or hypertonic medium for 30 min or 3 h. The cells were then treated for 90 s with 0.1% dimethyl sulfate. As a control, DNA extracted from cells and treated with 0.1% dimethyl sulfate in vitro for 90 s was also analyzed by ligation-mediated PCR (LM-PCR) (Fig. 3). When the intensity of the bands between the in vitro lane and the in vivo lanes were compared, two guanosines, at positions 915 and 908 bp, were strongly and reproducibly protected from methylation in cells grown in both normal or hypertonic medium as shown in comparison between lanes 4 and lanes 1-3. The relative intensity of the two guanosine bands was similar for normal and osmotic stress. This result confirms that a factor(s) binding to the region 926 to 895 bp occurs in vivo, and shows that occupancy of this site is constitutive in the endogeneous rat AR promoter.

[View Larger Version of this Image (34K GIF file)]

DNA probes were used to further evaluate putative transcription factor binding at 926 and 895 bp in the rat AR promoter. Probe A, which encompasses the 32-bp of sequence protected by ExoIII-mediated footprinting (Fig. 2), gave several complexes in EMSA experiments (Fig. 4, lanes 2-4). These complexes were similar in normal and hypertonically-treated cells. In competition experiments, cold probe A caused inhibition of band complex formation (Fig. 4, lanes 5-7). Probe A8, which contains eight mutations, including the two guanosine nucleotides protected from methylation by dimethyl sulfate in vivo footprinting, did not show significant binding of the same factors (Fig. 4, lanes 8-10). This result was confirmed by a competition assay with cold mutated probe A8 which did not compete with wild type probe A (Fig. 4, lanes 11-13).

[View Larger Version of this Image (62K GIF file)]

The effect on constitutive activity and osmotic response of the thymidine kinase (TK) promoter (105 to +52 bp) was studied by fusing a portion of 35-bp ExoIII-protected region to the 5-end of the TK promoter (Fig. 5). Construct pAEE35-TK, which contains the full 35-bp sequence of AEE, gave a 4-fold increase in basal activity as well as a 5.3-fold increase with osmotic stress. Construct pAEE28-TK (28 bp) which lacks 7 nucleotides of AEE gave the highest basal activity (7.7-fold) and the highest absolute activity under osmotic stress (100.00). Construct pAEE24-TK, which lacks 11 bp, was found to be the minimum required sequence for the functional induction of basal promoter activity and for the osmotic response. The constructs, pAEE20-TK and pTK, gave similar basal activity and no osmotic response.

[View Larger Version of this Image (20K GIF file)]

DISCUSSION

The transfection studies of the AR promoter region between 1,071 and 895 indicate important elements necessary for the increase of AR constitutive promoter activity and osmotic response (Fig. 6). Within this region a tonE-like element (1,071 to 1,059 bp), originally found as an osmotic response enhancer of the dog betaine transporter gene (Table I), was found (23). The sequence and the location of the element matches other tonE-like elements recently reported in the rabbit AR promoter between 1,108 and 1,092 bp (12), and in the mouse AR promoter between 1,053 and 1,040 bp (13). The human tonE-like element has been reported at two positions, between 1,230 to 1,220 bp and 1,157 to 1,148 bp (11).

[View Larger Version of this Image (17K GIF file)]

Table I. Comparison of the TonE-like sequences in AR promoters Dog TonE TACTTGGTGGAAAAGTCCAG Rat TonE-like sequence TGGAAAATCA Mouse TonE-like sequence TGGAAAATCA Rabbit TonE-like sequence CGGAAAATCA Human TonE-like sequence (1) TGGAAAATAT TonE-like sequence (2) TGGAAAATCA

Surprisingly in the rat promoter, the deletion of the tonE-like element decreased, but did not abolish the osmotic response. Also, the lack of the tonE-like element had no effect on the high constitutive promoter activity. As the 5-end of the promoter was deleted from construct 1 to construct 5, constitutive activity increased 7.0-fold suggesting that several negative regulatory elements may lie within this region. These results indicate that the most proximal cis-element required for the osmotic response and the constitutive promoter activity lie downstream of the tonE-like element.

The ExoIII-mediated DNA footprinting method was used to scan the region between 1,094 and 649 bp for any DNA protection under normal osmolarity. ExoIII digestion was halted at positions 926 bp in the coding direction and at 895 bp in the noncoding direction, indicating that a 32-bp region was protected under normal osmotic conditions. This putative transcription factor binding sequence, indicated by the ExoIII method, was further studied by dimethyl sulfate in vivo DNA footprinting to determine whether the factor was bound to the endogeneous AR promoter and how it would be affected by osmotic stress. Two guanosines located at 915 and 908 bp within the 32-bp protected region were found to be protected from dimethyl sulfate methylation under both normal and hypertonic conditions. No change in the intensity of the DNA protection was observed under normal culture conditions, 30 min osmotic stress, or 3 h osmotic stress. No other sequence surrounding this ExoIII-protected region was observed protected from methylation by these experiments.

Further confirmation of transcription factor binding and delineation of the critical nucleotides involved in binding were determined by EMSA. A probe containing the 35-bp protected sequence gave several intense complexes under normal and osmotic stress, while a probe containing eight nucleotide mutations, between and including the two guanosines protected in the dimethyl sulfate in vivo experiment, significantly reduced the transcription factor binding. Competition assays also confirmed that the sequences between the two guanosines are crucial for the transcription factor(s) to bind. No match was found to this sequence in the transcription factor data base. This novel cis-element was named aldose reductase enhancer element (AEE).

The regulatory properties of this novel cis-element were further demonstrated when various lengths of AEE were fused to the 5-end of the thymidine kinase promoter (105 to +52 bp). Surprisingly, construct pAEE28-TK gave a significant basal activity increase of 7.7-fold and an osmotic response of 5.7-fold, which is a higher increase than construct pAEE35-TK (Fig. 5). The minimum functional sequence for AEE was determined as 5-GGGTGTTGGAAGAGTGCCAAATTT-3 by construct pAEE24-TK and pAEE20-TK.

Finally, luciferase assays were performed with constructs 8 and 9 to determine the effect of inhibition of AEE binding or inhibition of AEE plus tonE binding on promoter activity. Construct 8 was designed to only inhibit AEE binding by deleting 8 bp (5-GAAGAGTG-3) of the critical sequence determined by the EMSA experiment (Fig. 4). Both constructs 8 and 9 completely abolished the constitutive activity of the rat AR promoter, while construct 8 which includes tonE retained an osmotic response (2.5-fold). This unexpected finding suggests to us that the AEE element is critical for maintenance of constitutive activity. Deletion of AEE, which resulted in continued osmotic response, indicates the presence of other positive or negative elements.

In conclusion, we have identified a new cis-element AEE, located downstream of the tonE-like element, which has a strong effect on the constitutive activity and osmotic response of the rat aldose reductase promoter. This enhancer is capable of inducing the basal activity of a heterologous promoter by 7.7-fold and osmotic response up to 6.7-fold. From our data on AEE and considering the fact that tonE-like elements have been reported in different locations in the AR promoter (11), it is likely that several elements might be involved in the control of the osmotic stress response. Identification of the relative role of these respective elements and their connection to the signal transduction pathway will be important in understanding the process of cellular osmoregulation. The in vivo occupancy of the AEE in the endogeneous AR promoter strongly suggests that this element plays a physiological role in the regulation of AR expression.