Transcription of the sodium/myo-inositol cotransporter gene is regulated by multiple tonicity-responsive enhancers spread over 50 kilobase pairs in the 5'-flanking region

Abstract

The sodium/myo-inositol cotransporter is a plasma membrane protein responsible for concentrative cellular accumulation of myo-inositol in a variety of tissues. When cells in kidney and brain are exposed to a hyperosmolar salt condition (hypertonicity) due to the operation of urinary concentration mechanism and pathological conditions, respectively, they survive the stress of hypertonicity by raising the cellular concentration of myo-inositol. Transcription of the sodium/myo-inositol cotransporter gene is markedly stimulated in response to hypertonicity, leading to an increase in the activity of the cotransporter, which in turn drives the osmoprotective accumulation of myo-inositol. To understand the molecular mechanisms by which hypertonicity stimulates transcription, we analyzed the 5'-flanking region of the cotransporter gene for cis-acting regulatory sequences. We identified five tonicity-responsive enhancers that are scattered over 50 kilobase pairs. All the enhancers are variations of the same type of enhancer interacting with the transcription factor named tonicity-responsive enhancer binding protein. In vivo methylation experiments demonstrated that exposure of cells to hypertonicity increases the binding of tonicity-responsive enhancer binding protein to the enhancer sites, indicating that all of these enhancers are involved in the transcriptional stimulation. We conclude that the sodium/myo-inositol cotransporter gene is regulated by a large region (approximately 50 kilobase pairs) upstream of the gene.

Fig. 1. Map of the SMIT gene and TonE sequences

The restriction map shown at the bottom is constructed from two P1 clones 3283 and 3284 using pulse-field gel electrophoresis and Southern blot analysis: M, MluI; C, ClaI; N represents two NotI sites at −248 and −129 relative nucleotide positions relative to the SMIT gene. The first exon of the SMIT gene is indicated by a vertical bar and the second exon by a box within which the entire protein coding region (shaded box) is located. Ovals labeled S14 and S31 indicate SacI fragments in which TonE sequences are identified (see Figs. 2 and 3 and Table II). The narrow oval P denotes position of a TonE sequence (TonEp) at −331 relative nucleotide position (Table II).

Fig. 2. Localization of tonicity-responsive enhancer activity

SacI restriction fragments S14 (top) and S31 (bottom) were subcloned from the P1 clone 3283 (Fig. 1). S14 and S31 and their smaller restriction derivatives shown were individually cloned in front of the SMIT promoter and the Photinus luciferase gene. Each construct was transfected into MDCK cells, and the transfected cells were cultured in isotonic or hypertonic medium. Activity of luciferase was measured from cell lysates, and -fold induction of luciferase by hypertonicity was calculated by dividing the activity in hypertonic condition by the activity in isotonic condition (mean values are shown; n = 2–5) as described under “Experimental Procedures.” Luciferase expression driven by the SMIT promoter alone was stimulated 11% (1.11-fold induction) by hypertonicity (n = 6). S, SacI; P, PstI; H, HindIII; K, KpnI; X, XbaI; N, NheI.

Fig. 3. Characterization of the SMIT TonEs

DNA fragments S14-c (top), S14-g (middle), and S31-d (bottom) are shown as thick lines with positions of TonE sequences indicated (not to scale) as boxes. S31-ds was derived from S31-d using PCR. TonEA in S14-c and TonEB(2/3) in S14-g were inactivated (as indicated by X marks) by mutating key nucleotide residues as described under “Experimental Procedures” to generate S14-c/mTonEA and S14-g/mTonE(2/3), respectively. Likewise, TonEC1 and TonEC2 in S31-ds were mutated to generate S31-ds/mTonEC1, S31-ds/mTonEC2, and S31-ds/mTonEC1/C2. TonEB1(2/3) (middle) and TonEC123 (bottom) are described in Table I. In other DNAs, TonE sequences marked by X were mutated by changing the 2nd G to T, the 5th A to T, and the 9th pyrimidine to G. Each of the wild-type and mutant DNA fragments was tested for its tonicity-responsive enhancer activity as described in Fig. 2. -Fold induction of luciferase in response to hypertonicity in cells transfected with each construct is shown: mean ± S.E., n = 4–9.

Fig. 4. Binding of TonE sequences to nuclear extracts from MDCK and HeLa cells

A, autoradiogram of an EMSA gel using ³²P-labeled hTonE (right) or TonEA (left). Each lane was loaded with a binding reaction containing 5 μg of nuclear extract prepared from MDCK or HeLa cells cultured in isotonic (I) or hypertonic (H) medium for 24 h. Slowly migrating bands representing TonEBP and free probes are marked on the right. B, autoradiogram of an EMSA gel showing only the TonEBP bands. Each lane was loaded with a binding reaction containing 0.5 nm ³²P-labeled hTonE, 5 μg of nuclear extract prepared from MDCK cells cultured in hypertonic medium, and 10 nm of a competitor TonE sequence indicated on the top. Percent competition (or reduction) of ³²P-labeled hTonE binding to TonEBP by a TonE sequence was calculated from the radioactivity of the slow bands in the lane without (first lane) and in the lane with that particular TonE competitor. Means of triplicate determinations are shown at the bottom.

Fig. 5. Changes in the occupancy of SMIT TonE sites in response to hypertonicity

Site occupancy was determined by in vivo footprinting assay in which the efficiency of methylation of G residues by dimethyl sulfate in vivo was measured. HeLa cells cultured in isotonic (I) or hypertonic medium (H) for 24 h were treated with dimethyl sulfate before isolation of DNA. The DNA was cleaved with piperidine, and quantitative PCR was performed to detect methylated G residues. ³²P-Labeled primers were used in the last two rounds of PCR to visualize amplified products in sequencing gels as shown. Each panel shows PCR products of a primer, and the region corresponding to its TonE site is demarcated along with the sequence. Bands representing the 2nd and 3rd G residues in TonEA, the 2nd G residue in TonEB2 and TonEC2, and the G residues complementary to the C residues in the 10th and 11th position of TonEp are indicated by filled arrowheads and bold letters. The radioactivity of these bands in each lane was divided by the radioactivity of the reference bands marked by open triangles to calculate the corrected intensity. Percent decrease in the corrected intensity of these bands in hypertonic cells relative to that in isotonic cells are shown at the bottom (mean and S.E., n = 4).