Communication Vol. 269, No. 47, Issue of November 25, pp, 29379-29381, 1994 THE JOURNAL OF B I O ~ I C A L CHEMISTRY

0 1994 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.

The Tonicity-sensitive Element That Mediates Increased Transcription of the Betaine Transporter Gene in Response to Hypertonic Stress*

(Received for publication, September 26, 1994)

Masaru Takenaka, Agnes S. Preston, H. Moo Kwon, and Joseph S. Handler$ From the Division of Nephrology, Department of Medicine, The Johns Hopkins University School of Medicine, Baltimore, Maryland 21205

BGT1, the Na+- and CP-coupled betaine tranqporter, is responsible for the accumulation of high concentrations of the non-perturbing osmolyte betaine in hypertonic Madin-Darby canine kidney (MDCK) cells and presum- ably in the hypertonic renal medulla. In MDCK cells, the increase in activity of the betaine transporter is pre- ceded by an increase in transcription of BGTl and in the abundance of BGTl mRNA. To investigate the molecular mechanism of transcriptional regulation by tonicity, we have characterized the 5”flanking region of the gene. Transient transfection assays in MDCK cells cultured in isotonic or hypertonic medium using luciferase reporter constructs containing various fragments of the 5”flank- ing region revealed that the region spanning base pairs -69 to -50 5’ to the transcription initiation site (-691-50) has hypertonicity-responsive enhancer activity. A dou- ble-stranded -691-50 concatemer cloned 5’ to an SV40 basal promoter and luciferase reporter gene in hyper- tonic cells exhibited more than 11-fold the activity in isotonic cells. Expression assays and electrophoretic mobility shift assays of mutants of -69/-50 identified a smaller region that is required for hypertonicity to in- duce increased expression and a slowly migrating band on mobility shift assays.

In response to exposure to a hypertonic environment, pro- karyotes, plants, and animal cells accumulate small organic solutes that protect them from the adverse effects of hyperto- nicity (1). In contrast to high concentrations of intracellular electrolytes, the small organic solutes do not perturb the func- tion of macromolecules. Glycine betaine (betaine) is one of the major non-perturbing small organic solutes (osmolytes) (1) ac- cumulated by prokaryotes (21, plants (11, and animal (3) cells exposed to hypertonicity. In prokaryotes, hypertonicity induces a betaine/proline transport system by relieving repression of the proU operon (4). In animals, the renal medulla is the only tissue that normally, as par t of the urinary concentrating mech- anism, becomes hypertonic. In Madin-Darby canine kidney (MDCK)’ cells, betaine is accumulated to concentrations 1000

* This study was supported in part by National Institutes of Health Grant lPOl DK-44484. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ To whom reprint requests should be addressed. ‘The abbreviations used are: MDCK, Madin-Darby canine kidney;

times the extracellular concentration (5) as the result of an increase in the V,, of a sodium- and chloride- coupled trans- porter whose cDNA has been named BGT1. The cotransporter is a member of the sodium- and chloride-coupled neurotrans- mitter transporter gene family (6). The increase in BGTl trans- porter activity in response to hypertonicity is preceded by an increase in transcription of BGTl and in the abundance of BGTl mRNA (7). Hypertonicity also leads to the accumulation of sorbitol and myo-inositol in kidney cells by increasing the transcription of the genes for aldose reductase (8) and for the sodium-coupled myo-inositol transporter (9-111, respectively. The molecular mechanism of regulation of transcription of these genes in response to hypertonicity is not understood. We have characterized the 5”flanking region of the BGTl gene and have found a hypertonic stress-responsive element. Hyperto- nicity induces the formation of two complexes in MDCK cells that bind to the element.

EXPERIMENTAL PROCEDURES Cell Culture, Tkansient Dansfection, Luciferase Assays, and Chlor-

amphenicol Acetyltransferase Assays-MDCK cells were maintained in a defined medium (12). The day before transfection, MDCK cells were seeded onto 60-mm tissue culture dishes (approximately 5 x lo5 cells/ dish). They were transfected using DEAE-dextran (13) with plasmid DNA (10 pg of the luciferase reporter gene constructs containing vari- ous fragments of the 5”flanking region of the BGTl gene or double- stranded oligonucleotides, and, for assessing transfection efficiency, 5 pg of CMV-CAT, a plasmid containing the chloramphenicol acetyltrans- ferase gene under the control of the cytomegalovirus promoter). Trans- fected cells were maintained in isotonic medium for 24 h, then switched to hypertonic medium or maintained in isotonic medium for another 24 h. Forty-eight hours after the transfection, the cells were harvested and assayed for luciferase (14) and chloramphenicol acetyltransferase (15) activity. Each transfection was performed at least three times. Medium was made hypertonic by adding raffinose to 500 mosmkg H’O.

Electrophoretic Mobility Shift Assays-Nuclei were isolated from MDCKcells cultured in isotonic or hypertonic medium (161, and nuclear extracts prepared (17). Nuclear extracts (4 pg) were incubated with a 3’P-end-labeled oligonucleotide corresponding to -691-35 of the BGTl gene (see Fig. 11, in 30 pl of binding reaction mixture (20 mM HEPES (pH 7.91, 100 mM KCl, 0.1 mM EDTA, 10% glycerol, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 1.5 pg of poly(dA-dT), 5 mM MgCl). After a 30-min incubation at room temperature, the mixture was elec- trophoresed on a 4% polyacrylamide gel (79:1, acrylamide: bisacrylamide) in TAE buffer (40 mM Tris acetate, pH 8.5, 2 mM EDTA) at 4 “C. The oligonucleotides used were as follows.

-69/-35: 5“TACTTGGTGGAAAAGTCCAGCTGGTGATTCACCCT ATGAACCACCTTTTCAGGTCGACCACTAAGTGGGA-5’

-69/-50: 5“TCGATACTTGGTGGAAAAGTCCAG ATGAACCACCTTTTCAGGTCAGCT-5’

-80/-56: 5“GATCCAGAGGGCAGGTACTTGGTGGAAAA GTCTCCCGTCCATGAACCACCTTTTCTAG-5’

AP1 consensus oligonucleotide was obtained from Promega.

RESULTS The BGTl gene has three independent transcription initia-

tion sites under control of three independent promoters.’ Only the 5’-flanking region of the most 5’ exon, lA, displays hyper-

BGT1, betaine-y-aminobutyric acid transporter; TonE, micity-respon- sive element; bp, base pair(s); M A P , mitogen-activated protein.

A., Kwon, H. M., and Handler, J. S. (1994) Proc. Natl. Acad. Sci. U. S. A,, ’ Takenaka, M., Bagnasco, S. M., Preston, A. S., Uchida, S.,Yamauchi,

in press.

29379

29380 Hypertonicity-responsive Element in the Canine BGTl Gene T,wx I

H,yprrtonir .stress induces luciferasr actioity in transfrrted MDCK crlls

Luciferase activity of the indicated plasmid constructs was assayed in transiently transfected MDCK cells. A CMV-CAT construct was cotrans- fected into MDCK cells to correct for transfection efficiency. Twenty-four hours after transfection in isotonic medium, cells were shifted to either isotonic (ISO) or hypertonic (HYP) medium, and maintained for another 24 h. Then the cells were harvested for assay of luciferase and CAT activities. Numbers in parentheses for each construct indicate positions of nucleotides a t both ends of the DNA fragment from the RGTl gene. Nucleotides are numbered from the first nucleotide (+1) of exon 1A. Nucleotides 3‘ to this are numbered positive and nucleotides 5’ to this a re negative. The DNA fragments were placed 5‘ to the luciferase gene alone (+ Luc) or the luciferase gene with an SV40 basal promoter (SV40Prom + Luc). Activity in cells transfected with the luciferase gene driven by a p-actin promoter (p-actin + Luc) was designated as 100 and results for each construct in that experiment are expressed relative to that value. Values given are means for isotonic and hypertonic media. HA is the mean ratio of activity in hypertonic medium divided by ac- tivity in isotonic medium in each experiment ? standard error. n is the number of independent transfections.

Constructs IS0 HYP HA n

Luc 1.3 0.8 0.8 2 0.12 12 (-501/+53) + Luc 8.2 80.2 10.9 2 2.16 6 (-373/+53) + Luc 6.2 65.5 11.4 ? 2.05 4 (-185/+53) + LUC 5.2 79.6 17.5 2 2.23 12 (-69/+53) + Luc 1.5 15.1 11.1 ? 2.65 4 (-49/+53) + Luc 2.8 2.6 0.8 0.09 12 SV40Prom + Luc 13.2 13.3 1.0 2 0.10 10 (-185l-50) + SV40Prom + Luc 16.3 203.8 13.2 ? 1.26 7 (-185/-70) + SV40Prom + Luc 12.8 22.1 1.9 2 0.30 5 p-Actin + Luc 100 100

tonicity-responsive enhancer activity,3 so we further character- ized only this region. We inserted various lengths of the 5’- flanking region of exon 1A of the BGTl gene’ 5‘ to a luciferase reporter gene and assayed luciferase expression in isotonic and hypertonic cells (Table I). All constructs demonstrated pro- moter activity as seen previously.2 Initial studies found that there was no difference in luciferase activity in response to hypertonicity in cells transfected with a constmct containing the 2,400 bp upstream of the transcription initiation site com- pared to cells transfected with a construct containing 500 bp (construct -501/+53 of Table I; data not shown). Further dele- tions from -501 to -373, to -185, to -69, manifested essentially the same enhancer activity. In contrast, a construct beginning 49 bp 5‘ to the start site of transcription was not responsive to hypertonicity (Table I), suggesting that the segment -69 to -50 contains the hypertonicity-responsive element. To confirm the location of the hypertonicity-responsive region, we tested -1851 -50 and -1851-70 for enhancer activity with a heterologous promoter. The former was very active and the latter inactive (Table I), establishing the presence of a tonicity-responsive el- ement (TonE) between -69 and -50. These results also elimi- nate the possibility that there is another independently acting TonE between -185 and -70.

To evaluate the -691-50 region further, we cloned that 20-bp oligonucleotide upstream of a n SV40 promoter and the lucifer- ase reporter gene in the correct orientation and in the reverse orientation. Hypertonicity induced luciferase activity inde- pendent of the orientation (Table 11, part A). A concatemer of the 20-bp region was dramatically responsive. These results establish that the 20-bp region contains a TonE.

To test for interaction of TonE with transcription factors, we performed electrophoretic mobility shift assays using a double- stranded oligonucleotide corresponding to -691-35 as a probe. Nuclear extracts from isotonic MDCK cells contained at least one binding factor that resulted in a broad band (Fig. 1, lane 2 ).

’ M. Takenaka, A. S. Preston, H. M. Kwon. and J. S. Handler, unpuh- lished results.

TAIII.K I1 Charnrfrriznfion of TnnE nnrl I / . < nlrrtonts

Twenty base pair douhle-strandrd oligonuclcwtidc.s contninlng TonE and its mutants (Fig. 2 A 1 werr cloned upstrcvlm of an SV40 haw1 promoter (Prom) and luciferase CLUCI rrportrr grnr . Tr:tnsfrction. lucif- erase and CAT assays, and the summary of rrsults arc’ as in Tahlv I .

~~

Cnnstructs IS0 IIYI’ 11’1 I1 - ~~- ~

A. (-69/-50) + SV40Prom + I,uc 16.3 45.2 2.9 t 0.26 6 (-50/-691 + SV40Prom + Luc 17.4 41.9 2.5 2 0 . 4 6 4 (-69/-50) x2 + SV40Prom + IAIC 22.7 260.9 11.6 2 2.27 3 SV40Prom + Luc 13.2 I.‘M 1 . 0 ~ 0.10 10

R. Mut676.7 + SV40Prom + 1,uc 1 x 3 6X.5 .X7 = 0.4.3 4 Mut60/56 + SV40Prom + Luc lX .O 17.1 1.1 2 0.19 4 Mut55/51 + SV40l’rom + 1,uc l X . O 21.2 1.2 2 0.05 1 [+Actin + Luc 1 0 0 100

~~~ ~

competitor - - -691-35 APl -6g/-50 -801.56 n n n n n nuclear extract - I H I H I H I H I

, (1

.AP1 -Ilke

’ I)

1 2 3 4 5 6 7 8 9 1 0 1 1

-69 - 5 0 1.451.391.35

-69

AP1

-yi-E .69/-35 -50

-691-50

-801-56 - 8 0 - 5 6

FIG. 1. Electrophoretic mobility shift ansay of TonE. Radiola- beled douhle-strandrd oligonuclrotlde cnrrc~~ponrlln~ to -GV-35 in the 5”flanking region of the IKT1 gene was incuhatrd with 4 !IC: of nuclwr extract of “DCK cells cultured in isotonic or hypertonic mcvlium in thr. presence and ahsence of a 200-fold molar ~ X C P S S of various unlahrlrvl douhle-stranded oligonuclcwtidrs. I m c w p n n d shows n schvmntic rvp- resentatinn of the probe and cold comprtitors. AP-1-likr hlndlng sitv (18) is indicated hy the hntchrd box ITGATTCAI.

A similar band was also formed by nuclear rxtracts from hy- pertonic cells (lane 3 ). Binding to this band is markedly com- peted by a n excess of A P 1 sequencr (lnncs 6 and 7 ). Thr hand that remains in the presence of excess API srquencr may rep- resent incomplete competition or binding to anothrr srquence. There is an APl sequence (18) a t -451-39 in the probr. Extracts from hypertonic cells led to two additional hands, a broad band migrating faster than the AP1-like band that is designated p, and a narrow band migrating more slowly, designated (1. Thr 20-bp TonE element competed for hinding with both of t h r hypertonicity-induced bands tlnncs R and 9 ) . whereas the se- quence -80/-56 competed for p but not for (r ( Innrs 10 and l l ), suggesting that bands a and fl are independent and rrprr- sent binding at the 3’ and 5’ portions of the TonE rlrmrnt, respectively.

To characterize the element further, wr synthrsizrd thrre sets of oligonucleotides that contained contiguous 5 bp muta- tions in the sequence spanning -69/-FjO (Fig. 2 A ) and used them as cold competitors in the electrophoretic shift assays (Fig. 2R ). Mut67/63 competed for complex n but not complrx [3 (lanes 4 and .5) and, in expression assays increasrd lucifrrasr activity in hypertonic cells (Table 11, part R). In contrast. Mut60/56 did not compete for complexes a or [3 (Fig. 2R. lnnr 7 1

and, did not lead to increased lucifrrasr activity in tbr rxprrs-

Hypertonicity-responsive Element in the Canine BGTl Gene 29381

A Enhancer activity

WT 1-69) T A C l T G G a A A G S A G 1 . 5 0 ) + Mu167163 TA-TGGAAAAGTCCAG + Mu160/56 T A C T f G G T G m G T C C A G I

Mu155/51 T A C l T G G T G G A A A A m G I

B

1 2 3 4 5 6 7 8 9 FIG. 2. Electrophoretic mohil i ty shift assay of TonE and i t s

mutants.A. sequences of‘“TonFY and its various mutants. The inverted repeat in TonE is indicated by nrrou’s. B, nuclear extracts from MDCK cells cultured in isotonic or hypertonic medium were incubated with radiolabeled double-stranded oligonucleotide -69/-35 in the absence ( - )

or in the presence (+) of a 200-fold molar excess of the indicated unla- beled oligonucleotides (competitor).

sion assays (Table 11, par t B). Confirming the competition of -go/-56 for complex p in Fig. 1, Mut55/51 also competes for complex p (Fig. 2R, lane 9 ). However, complex a remained (lane 9) . Mut55/51 did not increase luciferase activity in hypertonic MDCK cells (Table 11, par t B). These results indicate that the sequence spanning -6O/-51 is important for tonicity-responsive enhancer activity and that complex a, but not complex p, is involved in the response. It is appropriate to note an inverted repeat sequence in the region -62/-51 (Fig. 2.4 ).

DISCUSSION

The transcriptional regulation of osmolyte transporter genes by tonicity is of great interest but is not understood. More information is available for transcriptional regulation of the bacterial osmolyte transporter gene, proU, which encodes a betaine/proline transport system in Escherichia coli and Sal- monella typhimurium. The abundant DNA-binding protein, H-NS, has been shown to interact with a relatively large por- tion of DNA surrounding the proU promoter of the proU operon and represses transcription initiation of proU in isotonic con- ditions (4). Transcription from the proU promoter is induced in response to increased extracellular osmolarity (19, 20), pre- sumably because the inhibitory effect of H-NS is relieved by high concentrations of intracellular K (4). which are a direct consequence of extracellular hypertonicity (21). Although in- creased intracellular cation concentration may also play a role in transcriptional activation of the aldose reductase gene in renal medullary cells (221, the relevance to animal cells of the findings in bacteria is not clear. In yeast, hypertonicity induces the osmoprotective accumulation of glycerol due to a transcrip- tional activation of enzymes involved in glycerol synthesis (23). Yeast mutants defective in either HOG1, a MAP kinase homo- log, or PBS2, a M A P kinase kinase homolog, do not accumulate glycerol and grow poorly in hypertonic medium (24), indicating

that the MAP kinase pathway is involved in regulation of the activation of transcription of the glycerol synthetic pathway through a two component signal transduction pathway (25 J. Protein kinase C-dependent activation of the MAP kinase cas- cade by hypertonicity has been reported in MDCK cells (26,. Since osmolyte transporter mRNA accumulation is stimulated by hypertonicity in protein kinase C-depleted MDCK cells,’ it is unclear whether MAP kinase is involved in regulation of os- molyte transporter genes in higher eukaryotes. Mammalian cells use three mechanisms for organic osmolyte accumulation: decreased depadation (glycerophosphorylcholine). induction of increased synthesis (aldose reductase), and increased uptake by induction of specific cotransporters (betaine, mvo-inositol. and taurine) ( 5 , 27). Thus, a number of mechanisms appear to be involved in this important function.

This is the first TonE identified. The data we present indicate that hypertonic stress induced the formation of complexes bound to the TonE in MDCK cells. It is not clear, rlt this point. whether other tonicity-regulated genes, namely aldose reduc- tase (8) and the my-inositol cotransporter 111), use common cis- and transacting elements responsive to hypertonicity. Fur- ther characterization of trans-acting factors that interact with TonE, a s well as identification of cis- and trans-acting elements responsive to hypertonic stress in the other tonicity-regulated genes, are needed.

Acknoudcdgrnentn-We thank Drs. Tamio Noplchi and Kazuya Yamadn for helpful advice..

7 .

8.

9.

In.

11.

12.

13. 14.

15. 1 fi. 17.

1% 19. 20.

21.

22.

23. 24.

25. 2fi.

27.

REFERENCES