Kidney aldose reductase gene transcription is osmotically regulated FRED
L. SMARDO,
JR., MAURICE
B. BURG,
AND
ARLYN
GARCIA-PEREZ
Laboratory of Kidney and Electrolyte Metabolism, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland 20892 Smardo, Garcia-Perez.
Fred
L.,
Jr.,
Maurice
B. Burg,
and
Arlyn
Kidney aldosereductasegenetranscription is osmotically regulated. Am. J. Physiol. 262 (Cell Physiol. 31): C776-C782, 1992.-Cells generally adapt to long-term hypertonic stress by accumulating organic osmolytes. PAP-HT25 renal medullary cellsin hypertonic mediumaccumulatesorbitol through a reaction catalyzed by aldosereductase and betaine through osmotically regulated transport. Hypertonicity increasesaldosereductaseprotein synthesisrate by elevating its mRNA abundance.To test whether the rise in aldosereductase mRNA is due to enhanced transcription, PAP-HT25 cells adapted to isotonic medium were switched to hypertonic medium, and transcription rate was measuredby nuclear run-on. Aldose reductasetranscription rate peaked at l7-fold the isotonic level after 12 h of hypertonicity. Then, transcription fell as sorbitol and betaine accumulated. Transcription stabilized at fivefold the isotonic level within days. Aldose reductase mRNA stability was not significantly different between the hypertonic and isotonic steady states. Thus aldosereductase mRNA is osmotically regulated through changesin its transcription. The osmotically induced rise in aldose reductase transcription is blunted by the accumulation of intracellular betaine and is exaggerated and prolonged by preventing the accumulation of both sorbitol (by aldosereductaseinhibition) and betaine (by removal from the medium). This study presents the first description of osmoregulatedgene transcription in animal cells. osmoticstress;osmolytes;sorbitol; betaine; stressproteins
THE COURSE of evolution, prokaryotic and eukaryotic plant and animal cells have adapted to longterm osmotic stress by acquiring the ability to accumulate intracellular osmolytes. The selection of osmotically active, compatible organic solutes, in contrast to inorganic salts, is thought to have occurred because such solutes tend not to perturb other intracellular macromolecules (23). These processes hold true for the mammalian kidney, which generates high concentrations of NaCl in the interstitium of the inner medulla during antidiuresis (11). Renal medullary cells achieve osmotic balance by accumulating four major compatible organic osmolytes, namely sorbitol, inositol, glycerophosphorycholine, and betaine (1, 9). PAP-HT25 is a line of rabbit inner medullary cells (22) that can accumulate large quantities of sorbitol (2), as well as several other renal organic osmolytes, under hyperosmotic conditions (14). Sorbitol is synthesized from glucose in a reaction catalyzed by the enzyme aldose reductase (EC 1.1.1.21) (2). In cells exposed to medium made hyperosmotic (500 mosmol/kgH20) by the addition of NaCI, sorbitol accumulates due to an elevation in the levels of aldose reductase mRNA (lo), aldose reductase protein synthesis (l3), and aldose reductase enzymatic activity (2,2l). The elevation in aldose reductase mRNA is selective; the abundance of some other mRNAs (e.g., THROUGHOUT
prostaglandin synthase) is not increased by hypertonic medium (10). In addition, the level of aldose reductase mRNA falls to the isotonic baseline level within 1 to 2 days after PAP-HT25 cells, adapted to hypertonic medium, are switched to isotonic (300 mosmol/kgHzO) medium. However, the catalytic activity of aldose reductase remains elevated in these cells because the aldose reductase protein has a half-life on the order of 1 wk (13). Although organic osmolytes have been found in many cell types, the molecular regulatory mechanisms involved in their accumulation have been best defined for a betaine transport system (proU) in bacteria. In Escherichia coli and S. typhimurium, transcription of the proU operon has been shown to be osmotically regulated (7, 17). The objective of this study was to determine how medullary cells regulate aldose reductase mRNA abundance. Therefore, we tested PAP-HT25 cells for changes in aldose reductase gene transcription and mRNA degradation under varying osmotic conditions, utilizing nuclear run-on transcription assays and mRNA stability analyses. We conclude that transcription is the major, if not the only, mechanism that regulates aldose reductase mRNA abundance. In addition, transcriptional experiments that featured the controlled accumulation of sorbitol and betaine were conducted in PAP-HT25 cells. The results from the nuclear run-on transcriptional assays proved consistent with the previously proposed hypothesis (13, 21) that a rise in the intracellular sodium plus potassium concentration (or the intracellular ionic strength) signals for an increase in aldose reductase protein. Our findings constitute the first demonstration of osmotically regulated mammalian gene transcription. EXPERIMENTAL
PROCEDURES
Cell culture. PAP-HT25 cells (passages 67 to 77) were grown in isotonic (300 mosmol/kgHzO) mediumor by adding NaCl in hypertonic (500 mosmol/kgHzO) medium, as describedpreviously (21, 22). Betaine was absent from the medium except in the experimentswhere it is otherwiseindicated. Cellswereused 3 days after becoming confluent. The culture medium was refreshed-14 h prior to eachexperiment. Isolation of nudei. Cell nuclei were isolated for nuclear runon assaysusing modified protocols of Sasaki et al. (18) and Petersonet al. (16). Culture plates (100mm diameter,Corning) with PAP-HT25 cells were rinsed at 4°C with cold phosphatebuffered salinethat wasbalancedto the tonicity of the growth medium.The following stepswere performed on one plate at a time, maintaining the others at 4°C. The saline solution was removed, and 2 ml of cold nuclear homogenization buffer (NHB) [composition: 10 mM IV-2-hydroxyethylpiperazine-IV’2-ethanesulfonic acid (HEPES), pH 8.0, 10 mM MgCL,, 250 mM sucrose,0.1% Triton X-100, 2.0 mM dithiothreitol] were added. The cells were scraped,homogenizedwith 15 strokes using a Teflon pestle, and centrifuged at 500g for 5 min at 4°C
C776 Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 25, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
OSMOREGULATED
ALDOSE
REDUCTASE
in a l&ml polypropylene culture tube. The supernatant was decanted into 4 ml of 5.5 M GTC solution (composition: 5.5 M guanidine thiocyanate, 10 mM EDTA, pH 8.0, 50 mM sodium citrate, pH 6.0, 120 mM 2-mercaptoethanol) to later isolate the cytoplasmic RNA according to Chirgwin et al. (5). The pelleted nuclei were dispersed in 5 ml of cold NHB and kept at 4°C until the remaining plates were processed. The nuclei were then centrifuged at 500 g for 5 min at 4OC, resuspended in 5 ml NHB, and an equal volume of 1.0 M sucrose in NHB was then layered underneath. Nuclei were centrifuged through this sucrose cushion at 500 g for 5 min. Nuclear pellets were resuspended in 75 ~1 of a solution containing the following: 50 mM HEPES, pH 8.0, 5 mM MgCl,, 0.1 mM EDTA, 25% glycerol, and 2 mM dithiothreitol. To normalize the number of nuclei among samples, DNA concentrations were measured in aliquots solubilized in 1.0% sodium dodecyl sulfate (SDS). Concentrations were adjusted to equal 1.0 mg DNA/ml by using the factor 20 pg DNA/ml per A 26()unit [instead of the classical 50 pg DNA/ml per A 260unit (12) to take into consideration the absorbance by nuclear protein, average A280/A260being 0.61. Purified nuclei were then used in nuclear run-on transcription assays. Nuclear run-on transcription assays.Duplicate run-on transcription assayswere performed on the nuclei isolated from each samplecell plate. For each assay reaction, 25 ~1 of ~4 transcription buffer [composition: 360 mM NH&l, 20 mM MgC12,2 mM MnC12,8 mM dithiothreitol, 2 mM ATP, 2 mM guanosine5’-triphosphate (GTP), 2 mM cytidine 5’-triphosphate (CTP)], 24 ~1 of 25 mg/ml acetylated bovine serum albumin, 2 ~1 (110 units) of human placenta ribonuclease (RNase) inhibitor (Boehringer-Mannheim), and 40 ~1 (40 pg DNA) of suspendednuclei were addedto a locking (Eppendorf) microcentrifugetube. After mixing, 10 ~1of 10 mCi/ml [cY-~~P] uridine 5’-triphosphate (UTP) (3,000 Ci/mmol; Amersham) were added. The reaction was incubated for 30 min at 26°C. The DNA was then digestedfor 10 min at 37°C using 10 ~1of 1 U/cl1RQl DNase (RNase Free; Promega)and 6 ~1of 20 mM CaC12.The reaction was incubated for 30 min at 37°C after addition of the following: 16 ~1 of X10 EST [composition: 50 mM EDTA, 5% SDS, 100 mM tris(hydroxymethyl)aminomethane (Tris) l HCl, pH 7.51, 2 ~1of 10 mg/ml proteinase K, and 5 ~1 of 10 mg/ml yeast tRNA. The digestion was then stoppedwith 350 ~1of 4 M GTC solution (a dilution of the 5.5 M GTC solution defined above). The final volume was 510 ~1. Isolation of newly transcribed RNA from nuclei. RNA was isolated by the method of Chomczynski and Sacchi (6) and Celanoet al. (4). After adding x0.1 vol of 2.0 M sodiumacetate, pH 4.0, each nuclear run-on reaction (510 ~1) was extracted with x 1.2 vol of water saturatedphenol-CHCls-isoamylalcohol (1.0:0.2:0.002).Sampleswere set on ice for 15 min and then centrifuged at 4°C for 15 min at 12,000g. The aqueousphase was precipitated with xl.0 vol of isopropanol and centrifuged at 12,000g for 15 min. The RNA wasdissolvedin 300 ~1of 4.0 M GTC, precipitated again with isopropanol,and washedwith 70% ethanol. To improve hybridization specificity, RNA samples were partially hydrolyzed in 100 ~1 of 50 mM NaOH at 25°C for 10 min. The reaction was stoppedby adding 100 ~1of a solution containing 50 mM HCl and 10 mM HEPES, pH 7.4. The RNA was precipitated as before and dissolvedin 10 mM Tris, pH 7.5, 1 mM EDTA, and 0.1% SDS (TES), and the nucleotide incorporation wasdetermined by precipitation with 10% trichloroacetic acid. Equal counts per minute (cpm) of the labelednuclear RNA werehybridized to aldosereductasecDNA slot blots asdescribedbelow. pARlO plasmid and antisense RNA probe preparation. The construction, isolation, and growth of the pARlO clone, a Bluescript plasmid (Stratagene) containing rabbit kidney aldose reductase cDNA, as well as the synthesis of antisense
GENE
TRANSCRIPTION
c777
RNA probes for Northern blot analyseshave been described previously (10). Hybridization selection of aldose reductase-specific transcripts using the pARlO cDNA insert. The aldose reductase cDNA
insert waspurified from pARlO asfollows. The pARlO plasmid was cleaved with EcoR I and the 1,287-bp insert was size fractionated by electrophoresis through a preparative 1.0% agarosegel. The pARlO cDNA insert band was excised,electroeluted, extracted, and recovered by ethanol precipitations. With the useof a Hybri-slot manifold (GIBCO BRL), 1.5 pg of the aldosereductasecDNA insert or Bluescript plasmid DNA were adsorbedper slot on BA85/0.45 pm nitrocellulose paper (Schleicher and Schuell; Ref. 8). The nitrocellulose filter was cut so that each squarecontained one slot of aldosereductase insert and one slot of Bluescript plasmid.Each filter wasplaced into a 20-ml glassscintillation vial and incubated with 1 ml of prehybridization solution (composition:450 mM NaCl, 90 mM Tris HCl, 6 mM EDTA, x10 Denhardt’s solution, 0.1% SDS, 250 pg/ml yeast tRNA, 10 mM uridine) at 66°C for 2 h. The solution wasreplacedwith 300~1of hybridization buffer (composition: 450 mM NaCl, 90 mM Tris HCl, 6 mM EDTA, 20 mM sodiumphosphate, x10 Denhardt’s solution, 10% dextran sulfate, 0.1% SDS, 250 pg/ml yeast tRNA) containing 1 x lo7 cpm of newly transcribed 32P-labeledRNA and then covered with 1 ml of mineral oil to prevent evaporation. Hybridization was carried out at 66°C for 48 h. After hybridization, the vials were cooled to room temperature and the filters were batch rinsed with 50 ml of Xl SSC (composition: 150 mM NaCl, 15 mM sodium citrate, pH 7.0), 0.1% SDS, and then washedat 66°C for 30 min in another 50 ml of the samesolution. After three washesof 2 h eachat 66°C in x0.2 SSC, 0.1% SDS, the slot-blot filters were autoradiographed for l-10 days. The resultant bands were quantified using an LKB Ultroscan XL laser scanning densitometer.To normalize aldosereductasetranscription, all densitometry valuesin an experiment were divided by an internal control, which was the average densitometry value measuredfor PAP-HT25 cells grown in isotonic medium. Therefore, this arbitrarily set the transcription rate for the isotonic steady state equalto 1.0. Validation of nuclear run-on assays. In a representative experiment, four plates of PAP-HT25 cells grown in isotonic mediumwere switchedto hypertonic mediumfor 0, 12, 24, and 48 h prior to the start of the nuclear run-on assays.One plate of PAP-HT25 cells grown and maintained in hypertonic medium served as a hypertonic steady-state control. The average DNA content per plate was 180 t 30 pg, and the average incorporation of [CY-““P]UTPinto nuclear RNA was 10.3 * 1.0 million cpm (4.6 t 0.4% incorporation) per sample.Thus all nuclei incorporated approximately the sameamount of radioactivity regardlessof medium tonicity. In addition, by taking an equal number of cpm from each nuclear run-on reaction for hybridization selection to nonsaturable amounts of aldosereductase cDNA, the samples were normalized. Under these conditions, an increase greater than or equal to twofold in relative densitometry values representsa significant increase in the rate of transcriptional initiation for the aldosereductase gene. Northern blot analyses. Cytoplasmic and total RNA were first denatured in 2.2 M formaldehyde, 50% formamide for 5 min at 6O”C, and then size fractionated by electrophoresis through 1.5% agarosedenaturing gels in 2.2 M formaldehyde, 20 mM phosphate buffer, pH 7.8. RNA was transferred to nitrocellulose paper (Schlecher and Schuell) according to Thomas (20). Determination of aldose reductase mRNA half-life. With the use of hybridization selection, the half-life of aldosereductase mRNA was measuredas a function of the lossof 32P-labeled aldosereductasemRNA over time. First, confluent plates (100
Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 25, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
C778
OSMOREGULATED
ALDOSE
REDUCTASE
mm diameter, Corning) of PAP-HT25 cells adapted to isotonic or hypertonic medium were incubated for 24 h with 3 ml of otherwise identical medium containing 50 &i/ml [a-““P]UTP (3,000 Ci/mmol, Amersham). The cells were then rinsed twice with cold phosphate-buffered saline, and then 15 ml of fresh medium containing 10 mM uridine was added per plate to chase the label. At the given times of chase, the culture medium was removed and the RNA was isolated by a combined method of Cathala et al. (3) and Chomczynski and Sacchi (6), as follows. GTC (1.6 ml of 5.5 M) was added to each plate, and the lysate was vortexed to shear the DNA. After addition of ~5 volumes of 4 M LiCl, the mixture was stored for at least 4 h at 4°C. The precipitated RNA was centrifuged at 14,000 g for 20 min, and the pellet was dissolved in 500 ~1 of 4 M GTC. To this solution was added x0.1 volume of 2 M sodium acetate, pH 4.0. The RNA was extracted twice with ~1.2 volume of water-saturated phenol-CHC13-isoamyl alcohol (1.0:0.2:0.002) as above, precipitated twice with ~1 volume of isopropanol, and then partially hydrolyzed as described above for the nuclear run-on assays. The RNA isolated from each plate was hybridized to aldose reductase cDNA as described above for newly labeled transcripts; however, washes were performed only with xl.0 SSC, 0.1% SDS. To correct for losses due to RNA recovery, each slot-blot densitometry value was divided by the number of cpm in the corresponding hybridization sample. To measure the half-life of aldose reductase mRNA in PAPHT25 cells grown under conditions of varying medium tonicity, the densitometry values above were plotted against time, and the decay curves were drawn using linear regression analysis. Because the slope of each log-linear decay curve is equal to a decay rate and because the decay rate is proportional to halflife ( tIj2), the linear regression lines (half-lives) were compared by analysis of covariance to ascertain if their slopes were significantly different from one another (19). Any two regressions were regarded as significantly different if a value of P < 0.05 was obtained with the two-tailed F test. Addition of Tolrestat and betaine to PAP-HT25 cells. PAPHT25 cells were grown in isotonic mediumand then switched to hypertonic medium containing either Tolrestat (10 PM) or betaine (5 mM), or Tolrestat (10 PM) plus betaine (5 mM). After incubation for 21 or 45 h, Northern and nuclear run-on assayswere performed to measurealdose reductase mRNA abundanceand transcription rates. RESULTS Hypertonicity increases aldose reductase mRNA abundance and its transcription rate, both to approximately the same extent. Previous experiments revealed aldose
reductase mRNA levels to be sixfold higher in PAPHT25 cells passaged in hypertonic medium as compared with isotonic medium (10). The present study confirms those findings. Aldose reductase mRNA abundance is seven times higher for cells maintained in the hypertonic steady state (Table 1). To discover if the rise in aldose reductase mRNA is due to an increase in the rate of transcripion, nuclear run-on experiments were conducted. For cells grown in hypertonic medium, the transcription rate is five times higher than that of cells grown in isotonic medium (Table 1). Thus an increase in the rate of transcription can account for most, if not all, of the rise in aldose reductase mRNA. The rate of aldose reductase transcription rises slowly after extracellular tonicity increases. The transcription
rate for aldose reductase was evaluated during a Z-day period after PAP-HT25 cells, adapted to isotonic me-
GENE
TRANSCRIPTION
Table
1. Effect of hypertonicity on aldose reductase mRNA abundance and transcription rate Medium Osmolality, mosmol/kgH20 300 500
Relative mRNA Abundance l.OkO.3 6.8t1.8
(7) (11)
Relative Transcription l.OkO.3 5.1t1.3
Rate (9) (7)
Values are means & SE. Number of cultures tested is included in parentheses. Aldose reductase mRNA abundance and transcription rates were measured by Northern and nuclear run-on transcription analyses on cells grown for several passages at indicated osmolality (22). Laser scanning densitometry was used to quantify specific hybrids shown by autoradiography. Osmolality was increased in hypertonic medium (500 mosmol/kgHzO) by adding NaCl. Isotonic medium is 300 mosmol/kgHzO.
dium (300 mosmol/kgH20), had been switched to hypertonic medium (500 mosmol/kgHzO). The transcription rate of aldose reductase in PAP-HT25 cells maintained in hypertonic medium served as the hypertonic steadystate control. No change in the transcription rate of aldose reductase was observed during the first 30 min after medium tonicity was increased (Fig. 1A). After 1 h, the transcription rate doubles, and then it increases steadily for 12 h (Fig. 1, A and B), at which time it is approximately U-fold greater than in isotonic medium. After 12 h, transcription falls due to the accumulation of intracellular sorbitol (13) and eventually reaches a steady-state level that is fivefold higher than the rate measured in isotonic medium. The rate of aldose reductase transcription falls rapidly after extracellular tonicity decreases. The rate of aldose
reductase transcription in PAP-HT25 cells adapted to hypertonic medium (500 mosmol/kgH20) was examined after switching these cells to isotonic medium (300 mosmol/kgH,O). The transcription rate of aldose reductase in PAP-HT25 cells maintained in isotonic medium served as the isotonic steady-state control. Within 30 min of switching the cells to isotonic medium, aldose reductase transcription rate falls to ~50% of its original level in hypertonic medium (Fig. 2). By 2 h, it is lowered to the isotonic steady-state level. Thus transfer from hypertonic to isotonic medium results in a rapid decrease in transcription. The half-life of aldose reductase mRNA does not significantly differ between isotonic and hypertonic steady states. As already stated, most, if not all, of the rise in
aldose reductase mRNA under hypertonic conditions can be accounted for by the increase in the rate of transcrip. tion. Therefore, any change in the stability of aldose reductase mRNA should be minor. Nevertheless, the following experiments were conducted to test directly whether aldose reductase mRNA stability is affected by medium tonicity. Transcription appears to contribute relatively little to the total level of aldose reductase mRNA over a 16-h period after switching the cells from hypertonic to isotonic medium. This is based upon the following observations. The level of aldose reductase mRNA falls from a hypertonic steady-state level of sixfold to about twofold the isotonic level in 24 h after switching from hypertonic to isotonic medium (10). In contrast, under the same conditions, aldose reductase transcription falls from a
Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 25, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
OSMOREGULATED
ALDOSE
REDUCTASE
c779
GENE TRANSCRIPTION 8
0 0.5
2
6
16 f-a” r
+
0”
I
0
1
”
”
2 Hours After
*
I
3 Raising
”
4
5
6
0
7
1
Hours
Medium Osmolality To 500 mosmol/kg With NaCl
B
20 ,
I 0
12
24
48
CQ
2 3 4 5 After Lowering Medium Osmolality To 300 mosmol/kg With NaCl
6
7
Fig. 2. Relative aldose reductase transcription rate in PAP-HT25 cells adapted to 500 mosmol/kgH20. All cells grown in hypertonic medium were switched, except for zero hour control, to isotonic medium, and nuclear run-on assays were performed as described in legend of Fig. 1. Slot blots shown in photographic inset contain aldose reductase cDNA insert or Bluescript plasmid (arrow). Data are normalized to a steadystate level of 1.0, shown by dotted line.
HypertonicSteadyState
0’
’ 12
* 24 Hours
’ 36
’ 48
.
I’
L. SMARDO,
JR., MAURICE
B. BURG,
AND
ARLYN
GARCIA-PEREZ
Laboratory of Kidney and Electrolyte Metabolism, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland 20892 Smardo, Garcia-Perez.
Fred
L.,
Jr.,
Maurice
B. Burg,
and
Arlyn
Kidney aldosereductasegenetranscription is osmotically regulated. Am. J. Physiol. 262 (Cell Physiol. 31): C776-C782, 1992.-Cells generally adapt to long-term hypertonic stress by accumulating organic osmolytes. PAP-HT25 renal medullary cellsin hypertonic mediumaccumulatesorbitol through a reaction catalyzed by aldosereductase and betaine through osmotically regulated transport. Hypertonicity increasesaldosereductaseprotein synthesisrate by elevating its mRNA abundance.To test whether the rise in aldosereductase mRNA is due to enhanced transcription, PAP-HT25 cells adapted to isotonic medium were switched to hypertonic medium, and transcription rate was measuredby nuclear run-on. Aldose reductasetranscription rate peaked at l7-fold the isotonic level after 12 h of hypertonicity. Then, transcription fell as sorbitol and betaine accumulated. Transcription stabilized at fivefold the isotonic level within days. Aldose reductase mRNA stability was not significantly different between the hypertonic and isotonic steady states. Thus aldosereductase mRNA is osmotically regulated through changesin its transcription. The osmotically induced rise in aldose reductase transcription is blunted by the accumulation of intracellular betaine and is exaggerated and prolonged by preventing the accumulation of both sorbitol (by aldosereductaseinhibition) and betaine (by removal from the medium). This study presents the first description of osmoregulatedgene transcription in animal cells. osmoticstress;osmolytes;sorbitol; betaine; stressproteins
THE COURSE of evolution, prokaryotic and eukaryotic plant and animal cells have adapted to longterm osmotic stress by acquiring the ability to accumulate intracellular osmolytes. The selection of osmotically active, compatible organic solutes, in contrast to inorganic salts, is thought to have occurred because such solutes tend not to perturb other intracellular macromolecules (23). These processes hold true for the mammalian kidney, which generates high concentrations of NaCl in the interstitium of the inner medulla during antidiuresis (11). Renal medullary cells achieve osmotic balance by accumulating four major compatible organic osmolytes, namely sorbitol, inositol, glycerophosphorycholine, and betaine (1, 9). PAP-HT25 is a line of rabbit inner medullary cells (22) that can accumulate large quantities of sorbitol (2), as well as several other renal organic osmolytes, under hyperosmotic conditions (14). Sorbitol is synthesized from glucose in a reaction catalyzed by the enzyme aldose reductase (EC 1.1.1.21) (2). In cells exposed to medium made hyperosmotic (500 mosmol/kgH20) by the addition of NaCI, sorbitol accumulates due to an elevation in the levels of aldose reductase mRNA (lo), aldose reductase protein synthesis (l3), and aldose reductase enzymatic activity (2,2l). The elevation in aldose reductase mRNA is selective; the abundance of some other mRNAs (e.g., THROUGHOUT
prostaglandin synthase) is not increased by hypertonic medium (10). In addition, the level of aldose reductase mRNA falls to the isotonic baseline level within 1 to 2 days after PAP-HT25 cells, adapted to hypertonic medium, are switched to isotonic (300 mosmol/kgHzO) medium. However, the catalytic activity of aldose reductase remains elevated in these cells because the aldose reductase protein has a half-life on the order of 1 wk (13). Although organic osmolytes have been found in many cell types, the molecular regulatory mechanisms involved in their accumulation have been best defined for a betaine transport system (proU) in bacteria. In Escherichia coli and S. typhimurium, transcription of the proU operon has been shown to be osmotically regulated (7, 17). The objective of this study was to determine how medullary cells regulate aldose reductase mRNA abundance. Therefore, we tested PAP-HT25 cells for changes in aldose reductase gene transcription and mRNA degradation under varying osmotic conditions, utilizing nuclear run-on transcription assays and mRNA stability analyses. We conclude that transcription is the major, if not the only, mechanism that regulates aldose reductase mRNA abundance. In addition, transcriptional experiments that featured the controlled accumulation of sorbitol and betaine were conducted in PAP-HT25 cells. The results from the nuclear run-on transcriptional assays proved consistent with the previously proposed hypothesis (13, 21) that a rise in the intracellular sodium plus potassium concentration (or the intracellular ionic strength) signals for an increase in aldose reductase protein. Our findings constitute the first demonstration of osmotically regulated mammalian gene transcription. EXPERIMENTAL
PROCEDURES
Cell culture. PAP-HT25 cells (passages 67 to 77) were grown in isotonic (300 mosmol/kgHzO) mediumor by adding NaCl in hypertonic (500 mosmol/kgHzO) medium, as describedpreviously (21, 22). Betaine was absent from the medium except in the experimentswhere it is otherwiseindicated. Cellswereused 3 days after becoming confluent. The culture medium was refreshed-14 h prior to eachexperiment. Isolation of nudei. Cell nuclei were isolated for nuclear runon assaysusing modified protocols of Sasaki et al. (18) and Petersonet al. (16). Culture plates (100mm diameter,Corning) with PAP-HT25 cells were rinsed at 4°C with cold phosphatebuffered salinethat wasbalancedto the tonicity of the growth medium.The following stepswere performed on one plate at a time, maintaining the others at 4°C. The saline solution was removed, and 2 ml of cold nuclear homogenization buffer (NHB) [composition: 10 mM IV-2-hydroxyethylpiperazine-IV’2-ethanesulfonic acid (HEPES), pH 8.0, 10 mM MgCL,, 250 mM sucrose,0.1% Triton X-100, 2.0 mM dithiothreitol] were added. The cells were scraped,homogenizedwith 15 strokes using a Teflon pestle, and centrifuged at 500g for 5 min at 4°C
C776 Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 25, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
OSMOREGULATED
ALDOSE
REDUCTASE
in a l&ml polypropylene culture tube. The supernatant was decanted into 4 ml of 5.5 M GTC solution (composition: 5.5 M guanidine thiocyanate, 10 mM EDTA, pH 8.0, 50 mM sodium citrate, pH 6.0, 120 mM 2-mercaptoethanol) to later isolate the cytoplasmic RNA according to Chirgwin et al. (5). The pelleted nuclei were dispersed in 5 ml of cold NHB and kept at 4°C until the remaining plates were processed. The nuclei were then centrifuged at 500 g for 5 min at 4OC, resuspended in 5 ml NHB, and an equal volume of 1.0 M sucrose in NHB was then layered underneath. Nuclei were centrifuged through this sucrose cushion at 500 g for 5 min. Nuclear pellets were resuspended in 75 ~1 of a solution containing the following: 50 mM HEPES, pH 8.0, 5 mM MgCl,, 0.1 mM EDTA, 25% glycerol, and 2 mM dithiothreitol. To normalize the number of nuclei among samples, DNA concentrations were measured in aliquots solubilized in 1.0% sodium dodecyl sulfate (SDS). Concentrations were adjusted to equal 1.0 mg DNA/ml by using the factor 20 pg DNA/ml per A 26()unit [instead of the classical 50 pg DNA/ml per A 260unit (12) to take into consideration the absorbance by nuclear protein, average A280/A260being 0.61. Purified nuclei were then used in nuclear run-on transcription assays. Nuclear run-on transcription assays.Duplicate run-on transcription assayswere performed on the nuclei isolated from each samplecell plate. For each assay reaction, 25 ~1 of ~4 transcription buffer [composition: 360 mM NH&l, 20 mM MgC12,2 mM MnC12,8 mM dithiothreitol, 2 mM ATP, 2 mM guanosine5’-triphosphate (GTP), 2 mM cytidine 5’-triphosphate (CTP)], 24 ~1 of 25 mg/ml acetylated bovine serum albumin, 2 ~1 (110 units) of human placenta ribonuclease (RNase) inhibitor (Boehringer-Mannheim), and 40 ~1 (40 pg DNA) of suspendednuclei were addedto a locking (Eppendorf) microcentrifugetube. After mixing, 10 ~1of 10 mCi/ml [cY-~~P] uridine 5’-triphosphate (UTP) (3,000 Ci/mmol; Amersham) were added. The reaction was incubated for 30 min at 26°C. The DNA was then digestedfor 10 min at 37°C using 10 ~1of 1 U/cl1RQl DNase (RNase Free; Promega)and 6 ~1of 20 mM CaC12.The reaction was incubated for 30 min at 37°C after addition of the following: 16 ~1 of X10 EST [composition: 50 mM EDTA, 5% SDS, 100 mM tris(hydroxymethyl)aminomethane (Tris) l HCl, pH 7.51, 2 ~1of 10 mg/ml proteinase K, and 5 ~1 of 10 mg/ml yeast tRNA. The digestion was then stoppedwith 350 ~1of 4 M GTC solution (a dilution of the 5.5 M GTC solution defined above). The final volume was 510 ~1. Isolation of newly transcribed RNA from nuclei. RNA was isolated by the method of Chomczynski and Sacchi (6) and Celanoet al. (4). After adding x0.1 vol of 2.0 M sodiumacetate, pH 4.0, each nuclear run-on reaction (510 ~1) was extracted with x 1.2 vol of water saturatedphenol-CHCls-isoamylalcohol (1.0:0.2:0.002).Sampleswere set on ice for 15 min and then centrifuged at 4°C for 15 min at 12,000g. The aqueousphase was precipitated with xl.0 vol of isopropanol and centrifuged at 12,000g for 15 min. The RNA wasdissolvedin 300 ~1of 4.0 M GTC, precipitated again with isopropanol,and washedwith 70% ethanol. To improve hybridization specificity, RNA samples were partially hydrolyzed in 100 ~1 of 50 mM NaOH at 25°C for 10 min. The reaction was stoppedby adding 100 ~1of a solution containing 50 mM HCl and 10 mM HEPES, pH 7.4. The RNA was precipitated as before and dissolvedin 10 mM Tris, pH 7.5, 1 mM EDTA, and 0.1% SDS (TES), and the nucleotide incorporation wasdetermined by precipitation with 10% trichloroacetic acid. Equal counts per minute (cpm) of the labelednuclear RNA werehybridized to aldosereductasecDNA slot blots asdescribedbelow. pARlO plasmid and antisense RNA probe preparation. The construction, isolation, and growth of the pARlO clone, a Bluescript plasmid (Stratagene) containing rabbit kidney aldose reductase cDNA, as well as the synthesis of antisense
GENE
TRANSCRIPTION
c777
RNA probes for Northern blot analyseshave been described previously (10). Hybridization selection of aldose reductase-specific transcripts using the pARlO cDNA insert. The aldose reductase cDNA
insert waspurified from pARlO asfollows. The pARlO plasmid was cleaved with EcoR I and the 1,287-bp insert was size fractionated by electrophoresis through a preparative 1.0% agarosegel. The pARlO cDNA insert band was excised,electroeluted, extracted, and recovered by ethanol precipitations. With the useof a Hybri-slot manifold (GIBCO BRL), 1.5 pg of the aldosereductasecDNA insert or Bluescript plasmid DNA were adsorbedper slot on BA85/0.45 pm nitrocellulose paper (Schleicher and Schuell; Ref. 8). The nitrocellulose filter was cut so that each squarecontained one slot of aldosereductase insert and one slot of Bluescript plasmid.Each filter wasplaced into a 20-ml glassscintillation vial and incubated with 1 ml of prehybridization solution (composition:450 mM NaCl, 90 mM Tris HCl, 6 mM EDTA, x10 Denhardt’s solution, 0.1% SDS, 250 pg/ml yeast tRNA, 10 mM uridine) at 66°C for 2 h. The solution wasreplacedwith 300~1of hybridization buffer (composition: 450 mM NaCl, 90 mM Tris HCl, 6 mM EDTA, 20 mM sodiumphosphate, x10 Denhardt’s solution, 10% dextran sulfate, 0.1% SDS, 250 pg/ml yeast tRNA) containing 1 x lo7 cpm of newly transcribed 32P-labeledRNA and then covered with 1 ml of mineral oil to prevent evaporation. Hybridization was carried out at 66°C for 48 h. After hybridization, the vials were cooled to room temperature and the filters were batch rinsed with 50 ml of Xl SSC (composition: 150 mM NaCl, 15 mM sodium citrate, pH 7.0), 0.1% SDS, and then washedat 66°C for 30 min in another 50 ml of the samesolution. After three washesof 2 h eachat 66°C in x0.2 SSC, 0.1% SDS, the slot-blot filters were autoradiographed for l-10 days. The resultant bands were quantified using an LKB Ultroscan XL laser scanning densitometer.To normalize aldosereductasetranscription, all densitometry valuesin an experiment were divided by an internal control, which was the average densitometry value measuredfor PAP-HT25 cells grown in isotonic medium. Therefore, this arbitrarily set the transcription rate for the isotonic steady state equalto 1.0. Validation of nuclear run-on assays. In a representative experiment, four plates of PAP-HT25 cells grown in isotonic mediumwere switchedto hypertonic mediumfor 0, 12, 24, and 48 h prior to the start of the nuclear run-on assays.One plate of PAP-HT25 cells grown and maintained in hypertonic medium served as a hypertonic steady-state control. The average DNA content per plate was 180 t 30 pg, and the average incorporation of [CY-““P]UTPinto nuclear RNA was 10.3 * 1.0 million cpm (4.6 t 0.4% incorporation) per sample.Thus all nuclei incorporated approximately the sameamount of radioactivity regardlessof medium tonicity. In addition, by taking an equal number of cpm from each nuclear run-on reaction for hybridization selection to nonsaturable amounts of aldosereductase cDNA, the samples were normalized. Under these conditions, an increase greater than or equal to twofold in relative densitometry values representsa significant increase in the rate of transcriptional initiation for the aldosereductase gene. Northern blot analyses. Cytoplasmic and total RNA were first denatured in 2.2 M formaldehyde, 50% formamide for 5 min at 6O”C, and then size fractionated by electrophoresis through 1.5% agarosedenaturing gels in 2.2 M formaldehyde, 20 mM phosphate buffer, pH 7.8. RNA was transferred to nitrocellulose paper (Schlecher and Schuell) according to Thomas (20). Determination of aldose reductase mRNA half-life. With the use of hybridization selection, the half-life of aldosereductase mRNA was measuredas a function of the lossof 32P-labeled aldosereductasemRNA over time. First, confluent plates (100
Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 25, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
C778
OSMOREGULATED
ALDOSE
REDUCTASE
mm diameter, Corning) of PAP-HT25 cells adapted to isotonic or hypertonic medium were incubated for 24 h with 3 ml of otherwise identical medium containing 50 &i/ml [a-““P]UTP (3,000 Ci/mmol, Amersham). The cells were then rinsed twice with cold phosphate-buffered saline, and then 15 ml of fresh medium containing 10 mM uridine was added per plate to chase the label. At the given times of chase, the culture medium was removed and the RNA was isolated by a combined method of Cathala et al. (3) and Chomczynski and Sacchi (6), as follows. GTC (1.6 ml of 5.5 M) was added to each plate, and the lysate was vortexed to shear the DNA. After addition of ~5 volumes of 4 M LiCl, the mixture was stored for at least 4 h at 4°C. The precipitated RNA was centrifuged at 14,000 g for 20 min, and the pellet was dissolved in 500 ~1 of 4 M GTC. To this solution was added x0.1 volume of 2 M sodium acetate, pH 4.0. The RNA was extracted twice with ~1.2 volume of water-saturated phenol-CHC13-isoamyl alcohol (1.0:0.2:0.002) as above, precipitated twice with ~1 volume of isopropanol, and then partially hydrolyzed as described above for the nuclear run-on assays. The RNA isolated from each plate was hybridized to aldose reductase cDNA as described above for newly labeled transcripts; however, washes were performed only with xl.0 SSC, 0.1% SDS. To correct for losses due to RNA recovery, each slot-blot densitometry value was divided by the number of cpm in the corresponding hybridization sample. To measure the half-life of aldose reductase mRNA in PAPHT25 cells grown under conditions of varying medium tonicity, the densitometry values above were plotted against time, and the decay curves were drawn using linear regression analysis. Because the slope of each log-linear decay curve is equal to a decay rate and because the decay rate is proportional to halflife ( tIj2), the linear regression lines (half-lives) were compared by analysis of covariance to ascertain if their slopes were significantly different from one another (19). Any two regressions were regarded as significantly different if a value of P < 0.05 was obtained with the two-tailed F test. Addition of Tolrestat and betaine to PAP-HT25 cells. PAPHT25 cells were grown in isotonic mediumand then switched to hypertonic medium containing either Tolrestat (10 PM) or betaine (5 mM), or Tolrestat (10 PM) plus betaine (5 mM). After incubation for 21 or 45 h, Northern and nuclear run-on assayswere performed to measurealdose reductase mRNA abundanceand transcription rates. RESULTS Hypertonicity increases aldose reductase mRNA abundance and its transcription rate, both to approximately the same extent. Previous experiments revealed aldose
reductase mRNA levels to be sixfold higher in PAPHT25 cells passaged in hypertonic medium as compared with isotonic medium (10). The present study confirms those findings. Aldose reductase mRNA abundance is seven times higher for cells maintained in the hypertonic steady state (Table 1). To discover if the rise in aldose reductase mRNA is due to an increase in the rate of transcripion, nuclear run-on experiments were conducted. For cells grown in hypertonic medium, the transcription rate is five times higher than that of cells grown in isotonic medium (Table 1). Thus an increase in the rate of transcription can account for most, if not all, of the rise in aldose reductase mRNA. The rate of aldose reductase transcription rises slowly after extracellular tonicity increases. The transcription
rate for aldose reductase was evaluated during a Z-day period after PAP-HT25 cells, adapted to isotonic me-
GENE
TRANSCRIPTION
Table
1. Effect of hypertonicity on aldose reductase mRNA abundance and transcription rate Medium Osmolality, mosmol/kgH20 300 500
Relative mRNA Abundance l.OkO.3 6.8t1.8
(7) (11)
Relative Transcription l.OkO.3 5.1t1.3
Rate (9) (7)
Values are means & SE. Number of cultures tested is included in parentheses. Aldose reductase mRNA abundance and transcription rates were measured by Northern and nuclear run-on transcription analyses on cells grown for several passages at indicated osmolality (22). Laser scanning densitometry was used to quantify specific hybrids shown by autoradiography. Osmolality was increased in hypertonic medium (500 mosmol/kgHzO) by adding NaCl. Isotonic medium is 300 mosmol/kgHzO.
dium (300 mosmol/kgH20), had been switched to hypertonic medium (500 mosmol/kgHzO). The transcription rate of aldose reductase in PAP-HT25 cells maintained in hypertonic medium served as the hypertonic steadystate control. No change in the transcription rate of aldose reductase was observed during the first 30 min after medium tonicity was increased (Fig. 1A). After 1 h, the transcription rate doubles, and then it increases steadily for 12 h (Fig. 1, A and B), at which time it is approximately U-fold greater than in isotonic medium. After 12 h, transcription falls due to the accumulation of intracellular sorbitol (13) and eventually reaches a steady-state level that is fivefold higher than the rate measured in isotonic medium. The rate of aldose reductase transcription falls rapidly after extracellular tonicity decreases. The rate of aldose
reductase transcription in PAP-HT25 cells adapted to hypertonic medium (500 mosmol/kgH20) was examined after switching these cells to isotonic medium (300 mosmol/kgH,O). The transcription rate of aldose reductase in PAP-HT25 cells maintained in isotonic medium served as the isotonic steady-state control. Within 30 min of switching the cells to isotonic medium, aldose reductase transcription rate falls to ~50% of its original level in hypertonic medium (Fig. 2). By 2 h, it is lowered to the isotonic steady-state level. Thus transfer from hypertonic to isotonic medium results in a rapid decrease in transcription. The half-life of aldose reductase mRNA does not significantly differ between isotonic and hypertonic steady states. As already stated, most, if not all, of the rise in
aldose reductase mRNA under hypertonic conditions can be accounted for by the increase in the rate of transcrip. tion. Therefore, any change in the stability of aldose reductase mRNA should be minor. Nevertheless, the following experiments were conducted to test directly whether aldose reductase mRNA stability is affected by medium tonicity. Transcription appears to contribute relatively little to the total level of aldose reductase mRNA over a 16-h period after switching the cells from hypertonic to isotonic medium. This is based upon the following observations. The level of aldose reductase mRNA falls from a hypertonic steady-state level of sixfold to about twofold the isotonic level in 24 h after switching from hypertonic to isotonic medium (10). In contrast, under the same conditions, aldose reductase transcription falls from a
Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 25, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
OSMOREGULATED
ALDOSE
REDUCTASE
c779
GENE TRANSCRIPTION 8
0 0.5
2
6
16 f-a” r
+
0”
I
0
1
”
”
2 Hours After
*
I
3 Raising
”
4
5
6
0
7
1
Hours
Medium Osmolality To 500 mosmol/kg With NaCl
B
20 ,
I 0
12
24
48
CQ
2 3 4 5 After Lowering Medium Osmolality To 300 mosmol/kg With NaCl
6
7
Fig. 2. Relative aldose reductase transcription rate in PAP-HT25 cells adapted to 500 mosmol/kgH20. All cells grown in hypertonic medium were switched, except for zero hour control, to isotonic medium, and nuclear run-on assays were performed as described in legend of Fig. 1. Slot blots shown in photographic inset contain aldose reductase cDNA insert or Bluescript plasmid (arrow). Data are normalized to a steadystate level of 1.0, shown by dotted line.
HypertonicSteadyState
0’
’ 12
* 24 Hours
’ 36
’ 48
.
I’
