ouabain-resistant transfectants of the murine ouabain resistance

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 267, No. 24, Issue of August 25, pp. 172’71-17278.1992 0 1992 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U. S. A.

Ouabain-resistant Transfectants of the Murine Ouabain Resistance Gene Contain Mutations in the a-Subunit of the Na,K-ATPase*

(Received for publication, February 14, 1992)

Lloyd G . CantleyS, Xiao-Mai Zhoug, Mary Jo Cunha, Jonathan Epsteinll, and Lewis C. Cantleyg From the Beth Israel Hospital, Haruard Medical School, Brigham and Women’s Hospital, THarvard Medical School and $Tufts New England Medical Center Boston, Massachusetts 02215

A 6.5-kilobase murine genomic DNA fragment isolated by Levenson et al. (Levenson, R., Racaniello, V., Albritton, L., and Housman, D. (1984) Proc. Natl. Acad. Sci. U. S. A. 81, 1489-1493) (called the ouabain resistance gene) has been shown to produce ouabain resistance in primate cells. Preliminary sequence information has revealed no homology with the coding sequence of the Na,K-ATPase. We have introduced this murine sequence into monkey and murine cells in an attempt to characterize its mechanism of action. In our experiments, transfection of this DNA fragment is as- sociated with the low frequency (1 in 8 X lo5 cells) appearance of ouabain-resistant clones of CV1, COS, and NIH 3T3 cells, an event not seen in control transfections.

Characterization of a new clone of ouabain-resistant CV1 cells (called OR8 cells) revealed a 5-fold increase in the IC50 for ouabain inhibition of rubidium uptake and a 10-fold increase in cell survival on ouabain. Although the murine sequence was detectable in South- ern blots of ouabain-resistant cells soon after transfection, this exogenous DNA was rapidly lost despite continued exposure to ouabain. Furthermore, we were unable to detect message expression by this genomic sequence in any of the three cell types tested. Instead, we found that all three ouabain-resistant cell lines exhibited point mutations in a domain of the a-subunit that has been implicated in ouabain sensitivity (Hl- H2). One of these mutations (Asp’21-Asn121 in OR8 cells) has been previously reported to cause ouabain resistance (Price, E. M., Rice, D. A., and Lingrel, J. B. (1989) J. Biol. Chern. 264,21902-21906). Other novel mutations in the HZ transmembrane domain were also detected. We postulate that the “ouabain resistance gene” is important in the early selection process on ouabain but that the permanent ouabain-resistant phenotype is due to a stable mutation in one allele of the e-subunit of the Na,K-ATPase.

A murine genomic DNA clone capable of causing ouabain resistance in primate cells was described in 1984 by Levenson et al. (1). The gene was isolated by transfection of murine genomic DNA into primate cells and selecting for resistance to the Na,K-ATPase inhibitor, ouabain. Human fibrosarcoma (HT1080) cells were transfected with chromosomal DNA from

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ To whom correspondence should be addressed Div. of Nephrol- ogy, Beth Israel Hospital, 330 Brookline Ave., Boston, MA 02215. Tel.: 617-735-2147; Fax: 617-735-5276.

mutagenized, ouabain-resistant mouse fibroblasts (LWT4) (3) and exposed to a lethal dose of ouabain. While most cells died, several clones survived and their DNA was subsequently transfected into monkey fibroblasts (CV1 cells). Again ouabain selection resulted in resistant colonies. Repeated sub- cloning of the DNA from resistant transfectants into fresh CV1 cells resulted in the concurrence of the ouabain-resistant phenotype with a single murine EcoRI-digested DNA fragment 6.5 kb’ in length. This 6.5-kb fragment of DNA (origi- nally named the ouabain resistance gene, designated in this paper as MOR6.‘) was then cloned into the vector pAT153 (a derivative of pBR322) and subsequently retransfected into CV1 cells, again with the development of ouabain-resistant colonies. Initial characterization of this DNA has shown it to have several mouse B1 and B2 repeats and a 300-base pair GC-rich region, but no significant homology with the coding sequence of either the a- or ,&subunit of the Na,K-ATPase or other known genes.’

Ouabain sensitivity of the Na,K-ATPase is at least partially determined by the sequence of the a-subunit. Although the majority of this protein is highly conserved across species, ouabain affinity can vary over 1000-fold. Furthermore, in certain species, there are at least three distinct forms of the a-subunit (designated al, a’, and as) which differ primarily in their affinity for ouabain (4, 5 ) . Correlation of ouabain binding with the amino acid sequence of these otherwise very similar proteins has lead to the discovery of a region of the a- subunit where the substitution of only two amino acids can alter the ouabain affinity over lOOO-fold, without any other apparent change in the pump (6, 7).

This region, designated by hydropathy plots as the H1-H2 extracellular domain (8), is approximately 40% conserved across species at the amino acid level. The principal changes are at the border residues between the membrane and the extracellular space, where substitution between charged and uncharged amino acids dramatically influences ouabain affinity. Thus, isoenzymes with charged residues at these two border positions (Arg”’ and Asp”’ in rat a1 for example) exhibit a very low affinity for ouabain, whereas those with uncharged amino acids at these sites (Gln”‘ and Asn”’ in human a1 and rat a3) have much higher affinities for ouabain. The evolutionary conservation of ouabain binding has lead to the postulation of the existence of an endogenous ligand which might act to regulate the pump. One possible explanation for the mechanism of action of MOR6.‘ has been that this murine sequence encodes a protein capable of binding at this site and

The abbreviations used are: kh, kilobase(s); MOR6.5, 6.5-kb murine ouabain resistance DNA fragment; MEM, modified Eagle’s media; FCS, fetal calf serum; DME, Dulbecco’s modified Eagle’s media; Hepes, 4-(2-Hydroxyethyl)-l-piperazineethanesulfonic acid PCR, polymerase chain reaction.

X.”. Zhou, manuscript in preparation.

17271

17272 Ouabain-resistant Cells Have Mutations in the Na,K-ATPase

antagonizing the action of ouabain. To address this question, we have transfected this murine

sequence into both primate and murine cells in an at tempt to understand its mechanism of action. We chose transfection of pRL (M0R6.'j in pAT153) into CV1 cells as the primary model since this was a duplication of the original experiment by Levenson et al. ( l ) , but also examined transfection utilizing a retroviral vector (pLJ) into COS and murine 3T3 cells.

We have confirmed previous findings of the appearance of ouabain-resistance in those cells transfected with MOR6.5, with no occurrence of ouabain resistance in control transfections (1). However, we have found no evidence that there is message expressed from this sequence, but rather have found that while the transfected DNA is present in abundance at the initial time of ouabain-resistant colony isolation, it is rapidly lost over the next 10-20 cell generations and is no longer detectable in the stable ouabain-resistant colonies. Instead, these colonies have mutations of the a-subunit of the Na,K-ATPase in the Hl-H2 domain. Interestingly, the mutations found to date are not in the 2 border residues described above, rather one is in a highly conserved residue adjacent to Asnlzz (Asp,,,) and the others are within the H 2 trans-membrane domain.

MATERIALS AND METHODS

Transfection of CV1 and COS Monkey Cells-CV1 cells were grown on MEM-a media containing 10% fetal calf serum (FCS). Transfec- tions were carried out when cells were 50% confluent, using Lipofectin (Bethesda Research Laboratories) as the transfection agent (9). Cells on 90-mm plates were initially washed three times with serum-free MEM-a and then fed with 5 ml of serum-free MEM-a. 100 pg of lipofectin mixed with 5 pg of either vector alone (pJ3r), vector containing 441 base pair of (bases 1133-1266 joined with bases 2574-2882) (pJB{-COR), or vector containing the entire MOR'.' sequence (pRL) was then added to the cells. After 5 h of incubation, the cells were fed MEM + 20% FCS. The plates were allowed to grow to confluence and passed on day 3. Selection was carried out by transiently (1-2 days) exposing the cells to 1 p~ ouabain (extended exposure resulted in the death of all cells). Over the next 2 weeks, cells were treated intermittently with 1 p~ ouabain as dictated by their appearance microscopically. Control (vector alone) transfected plates were maintained exactly as pJ3{-COR- and pRL-transfected plates.

COS cells were grown on MEM-a + 10% FCS and transfected on 60-mm plates when 80% confluent. Transfections were carried out with a retroviral vector containing an SV40 promoter and the neomycin resistance gene (pW). Cells were washed three times with opti- MEMR reduced serum medium from GIBCO. They were then transfected with 40 pg of lipofectin + 1 pg of DNA (either pLJ alone or pLJ containing MOR6.5 (pW-RR-)). After 6 h, 20% fetal calf serum was added, and the cells were allowed to incubate for 48 h without selection. Cells were then split into either 1 p~ ouabain or 1 mg/ml G418 (a neomycin analog). Again ouabain selection was carried out with intermittent exposure to 1 p~ ouabain.

Transfection of Murine Cells-NIH 3T3 cells were also transfected using the retroviral vector pLJ, but with only the first 4578 bases of the ouabain resistance gene (pW-BB-). For these transfections, the retrovirus was used to transfect a packaging cell line, and then intact virus from that cell supernatant was used to infect 3T3 cells. This technique has been successfully used in other instances to splice genomic DNA inserts and recover cDNA (Hirt DNA) (10) from those transfectants carrying the correct phenotype. First, qcm cells (a viral packaging cell line) were transfected in a manner identical to that described for COS cells. 48 h after transfection, the cell supernatant (containing intact virus with the BB- insert) was harvested and mixed with polybrene (8 pg/ml) and filtered. 3 ml of the filtered supernatant were then added to 80% confluent 3T3 cells, and infec- tion was allowed to take place overnight at 37 "C. Cells were then split to two plates and selected on either G418 (1 mg/ml) or ouabain (3 mM). For Hirt DNA isolation, the ouabain-resistant 3T3 cells were fused with COS cells using polyethyleneglycol to amplify the SV40 origin in the p U vector.

Ouabain Survival Experiments-The degree of ouabain resistance

of the CV1 transfected cell lines was determined by growing cells to 50% confluence. Ouabain, in concentrations ranging from 10 nM to 5 mM, was then added to duplicate plates. The culture media was changed every 2 days, and cell confluence and appearance were determined daily. Percent confluence was judged as the percent of the culture dish covered by cells in the field of view, averaged over at least 10 low power fields.

Rubidium Uptake-CV1 and OR8 cells were grown to confluence in DME + 10% FCS media (+1 p~ strophanthidin in OR8-induced cells). Cells were then washed with phosphate-buffered saline, trypsinized, and diluted 1:3 in DME + 10% FCS to inactivate the trypsin. Cells were pelleted at 1500 rpm for 6 min and resuspended to a concentration of approximately 1 X lo6 cells/ml in DME + 5 mM Hepes (pH 7.4, [K+] = 5.4 mM). Cell counts were determined in quadruplicate using a hemocytometer. Cells were maintained on ice until used for the experiment (less than 1 h).

Previous work has shown that Na,K-ATPase-dependent rubidium uptake into CV1 cells remains linear for the first 20 min (11). We examined rubidium uptake at 10 min in CV1 and OR8 cells in varying concentrations of ouabain. The cells were diluted 1:l with DME- Hepes containing twice the ouabain concentration to be tested and then incubated for 30 min at room temperature to allow equilibrium binding to the sodium pump. They were then prewarmed to 37 'C for 2 min prior to the addition of [=Rb]Cl. At time 0,200 pl of cells were added to 4 pl of [=Rb]Cl (neutralized with equimolar NaOH immediately prior to the experiment), and uptake was allowed to proceed for the next 10 min (final concentration of rubidium = 3 p ~ ) . Cells were then spun for 30 s through 300 pl of oil (1:l silicone/dioctyl phthalate), supernatants were aspirated, and the tubes were inverted. The tube bottoms (containing the cell pellets) were removed and counted for 10 min in scintillation fluid using a p counter. All experiments were done in quadruplicate. (In addition, four samples were spun through oil immediately after the addition of [=Rb]Cl to determine background counts (due to trapping of media between the cells and interaction of the rubidium with cell surface proteins). This value was then subtracted from the 10-min rubidium uptakes and the standard errors added.)

Harvesting RNA-Cells were allowed to grow to confluence on six 90-mm plates and then washed twice with ice-cold PBS (138 mM sodium chloride, 5 mM sodium phosphate, 3 mM potassium chloride, 2 mM potassium phosphate, pH 7.4) and scraped in 1.5 ml of GIT buffer (4 M guanidine isothiocyanate, 0.5% sodium N-laurylsarcosine, 25 mM sodium citrate, 0.1 M @-mercaptoethanol, pH 7.0). The resultant suspension was vigorously vortexed for 30 s and then loaded onto 3.5-ml CsCl cushions and centrifuged for 21 h at 174,000 X g. The RNA at the bottom of the gradient was resuspended in diethyl pyrocarbonate-treated water and then precipitated in 0.3 M sodium acetate, 2.5 volumes of ethanol. The resultant pellet was washed with 80% ethanol and resuspended in diethyl pyrocarbonate-treated water. Concentration and purity were determined by A ~ w / A z ~ .

Northern Blots-Approximately 1 mg of total RNA was poly(A) selected on 500-pl oligo(dT) columns (Pharmacia LKB Biotechnolo- gies Inc.) and 10 pg of both total and poly(A)+ RNA were run on 1% agarose/formaldehyde gels. RNA was then transferred overnight by capillary action onto GeneScreen membranes (Du Pont-New England Nuclear) and baked for 3 h in a vacuum oven at 80 "C. Equality of loading was determined by hybridization with a-tubulin. a-1 and p-1 probes for rat Na,K-ATPase were generously supplied by Dr. Robert Levenson from Yale University. These were radiolabeled with [32P] dCTP using random labeling with the Klenow fragment (Pharmacia).

PCR and Sequencing of H1 -H2 Domain-Total RNA was reverse transcribed with either random primers or oligo(dT) primers. This single strand cDNA then served as template for the polymerase chain reaction (PCR) using 21-mers constructed to amplify the H1-H2 membrane-spanning domain of the a-subunit of the Na,K-ATPase. This region has been previously shown by several investigators (6, 7) to be critical in modulating the ouabain affinity of the pump. Primer MA1-A (30'ATGTTACTGTGGATCCGAGC) contained two point mutations (in bold) to create a BamHI site, and primer MA1-B (486TTTGAAGAATTCCATGATCTT) contained one point mutation to create an EcoRI site. The predicted product of these primers crossed one intron and resulted in amplification of a band of 182 nucleotides in length. Parent CV1 or 3T3 cells were amplified as controls.

PCR was performed for 35 cycles with 30 s at 94 "C, 1 min at 50 "C, and 3 min at 72 "C using a Thermal Cycler from MJ Research. Control reactions with no template were run in parallel to rule out the possibility of contaminants from previous experiments. The PCR

Ouabain-resistant Cells Have Mutations in the Na,K-ATPase 17273

products were then electrophoresed through 2% agarose gels and transferred onto nylon membranes via overnight capillary transfer. These blots were then probed with a random-labeled cDNA probe of full-length mouse Na,K-ATPase a-subunit (generously provided by Sharon Graw and See-Ying Tam from Dr. David Housman’s labora- tory at the Whitehead Institute, Massachusetts Institute of Technol- 00).

Those PCR samples with a single band of the correct size were then cloned into the TA3 sequencingvector (Invitrogen) and resultant colonies screened with the same a-subunit probe noted above. Positive colonies were further amplified and plasmid minipreps prepared, followed by a restriction digest to confirm the correct insert. The original miniprep was then used for double-stranded sequencing by the dideoxynucleotide technique of Sanger et al. (12).

Southern Blots-Genomic DNA was isolated on cesium chloride gradients and digested with BglII prior to running 10 pgllane on 0.7% agarose gels. DNA was transferred for 24 h via capillary action onto Genescreen membranes. The resultant blots were probed at 60 “C overnight in a hybridization solution containing 0.5 M sodium acetate, 1 mM EDTA, 1% bovine serum albumin, 7% sodium dodecyl sulfate, pH 7.2, and washed for 30 min at 55 “C in a wash solution of 40 mM sodium phosphate, 1 mM EDTA, 1% sodium dodecyl sulfate. The probes utilized were made from the entire gel-purified MOR6.5 (or smaller restriction fragments when so noted) radiolabeled with [”PI dCTP using a random labeling kit (Pharmacia).

Cell Lysis and Western Blotting-Cells were grown to confluence on 90-mm dishes, washed with ice-cold phosphate-buffered saline, and scraped in 0.5 ml of RIPA cell-lysis buffer (150 mM NaCl, 1% Nonidet P-40,0.5% deoxycholate, 1% sodium dodecyl sulfate, 50 mM Tris, pH 7.4). The mixture was then spun for 10 min on a clinical centrifuge, and supernatants (containing crude cell membranes) were assayed for protein content. 65 pg of protein was then run per lane on sodium dodecyl sulfate, 10% polyacrylamide gels and transferred onto nitrocellulose membranes by electroblotting.

The blot was then blocked for 1.5 h in CTT solution (10% dried milk in Tris-buffered saline, 0.1% Tween-20, pH 7.6), and washed briefly two times in Tris-buffered saline (TBS), 0.1% Tween-20. The blot was then incubated with a rabbit polyclonal antibody against the rat a-subunit of the Na,K-ATPase (a generous gift of Dr. Robert Levenson) in TBS, 0.1% Tween-20 at 4 “C overnight.

Following hybridization of the antibody, the blots were washed twice and incubated for 20 min with a horseradish peroxidase-strep- tavidin-labeled, anti-rabbit antibody from Amersham Corp. in TBS, 0.1% Tween-20 at room temperature. Following three washes in TBS, 0.3% Tween-20 and three washes in TBS, 0.1% Tween-20, the blot was incubated for 1 min with 10 ml of a 1:l mixture of ECL detection reagents 1 and 2, and then exposed to Hyperfilm ECL for 5 s (reagents and film from Amersham Corp.).

RESULTS

Transfection of Monkey and Mouse Cells with the Ouabain Resistance Gene-Consistent with the original results of Lev- enson et al. ( l) , transfection of CV1 cells with MOR6.5 resulted in a small fraction of cells resistant to 10 pM ouabain. By day 23, following 13 days of intermittent ouabain selection and then 10 days of continuous 1 p~ ouabain, all CV1 cells from the control transfections and the pJ3l“COR transfections were dead, but approximately five colonies/plate of dividing cells were present in those transfected with MOR6.5 (this equates to approximately 1 ouabain-resistant ce11/8 X lo6 cells transfected). These cells were trypsinized, pooled, and grown to confluence and were then subjected to 10 p~ ouabain for 4 days. This resulted in the death of many of the remaining cells, but the rapid appearance of normally dividing cells which reached confluence on this ouabain concentration. Al- though these cells (called ORs) represent a mixture of the original five ouabain-resistant colonies, the death of many cells when exposed to 10 UM ouabain would suggest that only one or two of the original clones survived. The cells appear morphologically identical to the parent CV1 cells, and the ouabain-resistant phenotype has remained stable for more than 12 months, even in cultures where ouabain selection was removed.

The ability of the murine genomic sequence to confer ouabain resistance has also been investigated using a retroviral vector to introduce the DNA into COS cells. These transfections were performed with the vector pLJ which in- cludes an SV40 promoter as well as the neomycin resistance gene. Again, no ouabain-resistant colonies were found in the control transfections selected on 1 p~ ouabain. In those COS cells transfected with MOR6.5, approximately 1 ouabain-resistant colony per 2 X IO5 cells was observed. Thus, the presence of the retroviral promoter did not significantly change the fraction of cells which ultimately became resistant to ouabain when compared to pRL transfections into CV1 cells. (The efficiency of transfection was approximately 15- 20% as judged by the number of neomycin-resistant clones.) Furthermore, if neomycin selection was performed first, and then neomycin-resistant colonies were secondarily selected with 1 WM ouabain, no ouabain-resistant colonies were observed (see “Discussion”). The stable ouabain-resistant (COS- RR-) transfectants were capable of dividing on 10 pM ouabain and were morphologically indistinguishable from parent COS cells.

Similarly, MOR6.5 conferred ouabain resistance to murine NIH 3T3 cells when introduced via a retroviral vector. In these experiments a fragment of the sequence containing the first 4578 bases (from the EcoRI to the BamHI site) was introduced using the pLJ vector as discussed under “Materials and Methods.” This fragment has been previously shown to be capable of conferring ouabain resistance in CV1 cells.3 The selection criteria for 3T3 cells was quite different than for COS and CV1 cells since the parent cell line is resistant to

M ouabain. Therefore, we selected these cells using 3 mM ouabain, a concentration which is eventually fatal to wild- type murine cells. In this case we saw multiple ouabain- resistant colonies in the pLJ-BB--transfected groups, but none in the mock-transfected controls. The ouabain-resistant transfectants continued to grow and divide normally even on 5 mM ouabain, a concentration which is rapidly fatal to wild- type 3T3. These cells, called 3T3-BB-, again represent a mixture of several, independent ouabain-resistant colonies.

Characterization of the Effect of Ouabain on the Transfected Cells-The ouabain-resistant transfectants were compared to the parent cell lines for ability to grow in the presence of ouabain and for [“Rb] uptake (a measure of the Na,K-ATPase activity). CV1 and OR8 cells were grown on varying concentrations of ouabain as described under “Materials and Meth- ods.” As seen in Fig. 1, at the higher concentrations of ouabain (>lo0 WM) all cells were dead by day 5. At intermediate concentrations of ouabain (1-10 pM), the OR8 cells continued to divide whereas CV1 cells all died. At 100 nM ouabain, OR8 cells were fully confluent by day 3, whereas CV1 cells were static on this concentration, neither dividing nor dying.

COS-RR- cells were slightly more resistant to ouabain than OR8 cells as judged by their ability to divide on 10 WM ouabain, but also were all killed on 100 PM ouabain (data not shown).

Rubidium uptake experiments indicated that the Na,K- ATPase of OR8 cells was more resistant to ouabain than that of the parent CV1 cells. Prior to preparing cells for uptake experiments, CV1 and OR8-uninduced cells were grown for greater than five passages in the absence of ouabain whereas OR8-induced cells were grown for 48 h in the presence of 1 p~ strophanthidin (see below). The whole cell rubidium uptake in the presence of 1 mM ouabain was subtracted from all values to yield Na,K-ATPase-dependent rubidium uptake. (The uptake in the presence of 1 mM ouabain was between 9-

L. G. Cantley, X.”. Zhou, R. Levinson, J. Epstein, and L. C. Cantley, unpublished results.


I2O E

A

0 I I r -: I

1 o - ~ 10" 1 0 . ~

FIG. 1. Survival of transfected primate cells on ouabain. CV1 and OR8 cells were grown to 50% confluence in MEM-a media and then exposed to ouabain at the indicated concentrations for the next 5 days. Cell confluence (% of plate covered by cells) was determined daily by examining multiple fields of duplicate plates. Fresh medium was applied every 2 days. e, CV1 cells; 0, OR8 cells.

Ouabain Concentration

300 r"""". .

Ouabain Concentration(M) FIG. 2. Rubidium uptake into transfected CV1 cells. Rubid-

ium uptake was allowed to proceed for 10 min at the concentration of ouabain shown. Uptake in the presence of 1 mM ouabain was subtracted from the other values to correct for non-Na,K-ATPase routes of entry for rubidium. Points are the mean of four measure- ments with bars indicating standard error of the mean. The solid line is a best fit to the CV1 data assuming a single ouabain-binding site/ pump. The two dashed lines are best fits to the OR8 and OR8-induced data, assuming two independent ouabain-binding sites/pump in each fit (see "Discussion"). All fits were performed with the data weighted using the standard error of the mean for each point.

24% of the total uptake by the cells and was not substantially different from the uptake at 100 p~ ouabain in either cell type. Since previous work has shown that the primate Na,K- ATPase is fully inhibited by 1 mM ouabain, and both CV1 and OR8 cells all died in the presence of 100 p~ ouabain, we believe that this assumption is valid.) Fig. 2 shows the results of these experiments. As can be seen, the ICs0 for ouabain is increased approximately 5-fold in the OR8 cells compared to the parent CV1 cell line.

Since prior data have shown that a different ouabain- resistant cell line (OR6 cells) required 24 h of growth on ouabain or strophanthidin for the appearance of ouabain- resistant rubidium uptake (11, 13, 14), we also chose to test whether preincubating one group of OR8 cells for 48 h on 1 p~ strophanthidin (designated ORB-induced) resulted in a difference in the ouabain effect on cell rubidium uptake. Strophanthidin was used for induction since this ouabain analog more readily dissociates from the sodium pump and can therefore be washed off prior to the rubidium uptake experiments (11). As can be seen in Fig. 2, there was no

significant difference in rubidium uptake in the two OR8 groups, suggesting that pregrowth in ouabain is not required for the resistant phenotype to appear in these cells.

Northern and Southern Analysis of Ouabain-resistant Cells-In contrast to results reported in ouabain-resistant OR6 cells (see "Discussion") (13), multiple attempts to show that MOR6.s encoded a messenger RNA in the OR8 or COS- RR- cells were unsuccessful. These included Northern analysis of poly(A)+ RNA probed with multiple regions of the murine sequence and PCR of potential coding sequences using reverse-transcribed mRNA as template. Similarly, attempts to determine potential splicing sites using a retroviral packaging system in 3T3 transfectants, as well as screening cDNA libraries from mouse liver and mouse fibroblast with restriction fragments of MOR6.', also failed.'

Furthermore, Southern analysis of genomic DNA from OR8 cells performed approximately 36 cell generations after the original transfection failed to show the presence of MOR',' (Fig. 3, lane 7). (The autoradiograph shown was obtained after probing the blots with a 1.4-kb BglII fragment of MOR6.5. When the full-length mouse genomic fragment was used as a probe, there was weak homology with a single band at approximately 4 kb which was found in both untransfected CV1 cells and OR8 cells and represents homology with an endogenous sequence in these cells.)' In contrast, early isolation of DNA from the ouabain-resistant COS transfectants (COS- RR-) revealed a large number of copies of MOR6.5 (lane 3) . (A lane of native mouse (NIH 3T3) DNA is included ( l a n e 1 ). Comparison of the signal intensity between the mouse DNA (presumably with two copies of the gene/cell) and the COS- RR- DNA suggests approximately 50 copies of MOR6.'//cell in these transfectants.) However, despite the continued selection pressure of 1 PM ouabain, these cells also lost the murine genomic DNA, with no detectable copies present after 40 more cell generations (lane 6). COS cells from this same

r

636 1--

2322- 2027-

1371-

5 6 4 -

1 2 3 4 5 6 7 FIG. 3. Genomic Southern analysis of MOR'." incorporation

into ouabain-resistant transfectants. Southern analysis of 10 pg of BglII-digested genomic DNA from cells transfected with MOR6.'. Four individual blots were probed with a 1371-base pair genomic BglII fragment of MOR6.'. Lane I is from untransfected 3T3 mouse cells, showing the size (1371 base pairs) and amount of the native genomic BglII fragment of MOR6.5. Lane 2 is untransfected COS monkey cells. Lane 3 is from ouabain-resistant colonies isolated 4 weeks after MOR6.s transfection into COS cells. Lane 4 is from the same transfection but selected from the beginning with neomycin instead of ouabain (DNA isolated 2 weeks after transfection). Lanes 5 and 6 are the same ouabain-resistant cells shown in lane 3, maintained on 1 p~ ouabain for a further 4 and 20 weeks, respectively. Lane 7 is DNA from ouabain-resistant CVl cells (OR8) isolated approximately 28 weeks following transfection. DNA marker sizes in kilobases are shown on the left.

Ouabain-resistant Cells Have Mutations in the Na,K-ATPase

transfection which were first selected for neomycin resistance revealed no detectable copies of MOR6.5 ( l a n e 4 ) . As discussed above, these cells all died when subsequently cultured in the presence of 1 pM ouabain.

Thus, although the appearance of ouabain resistance coin- cides with incorporation of at least part of MOR6.5 into the genome of COS cells, long term survival on ouabain was not dependent on the continued presence of this exogenous DNA. This suggests that while the murine sequence is important for the initial appearance of ouabain resistance, a second event might account for the stable phenotype.

Characterization of the Na,K-ATPase in Ouabain-resistant Cells-The disappearance of the murine DNA from CV1 and COS transfectants despite the stability of the ouabain-resistant phenotype, coupled with our inability to prove that this sequence resulted in mRNA production, led us to examine other potential mechanisms of ouabain resistance in these cells. Since ouabain binds to the a-subunit of the Na,K- ATPase, we investigated message levels, membrane protein levels and specific sequence information for this protein.

Northern blots of the ouabain-resistant OR8 and COS-RR- cells failed to show a difference in the size or amount of Na,K- ATPase a1 mRNA compared to that in parent CV1 and COS cells (Fig. 4). Rehybridization of these same blots with a Na,K-ATPase p1 subunit probe also showed comparable levels of p1 message in the two cell types (data not shown). Hybrid- ization with @* and H,K-ATPase @ subunit probes failed to show these messages. Furthermore, Western analysis of membrane proteins from CV1, OR8, COS, and COS-RR- cells showed that the amount of a-subunit was virtually identical in all four cell types (Fig. 5). These results are consistent with the comparable levels of Na,K-ATPase-dependent [=Rb]Cl uptake seen in CV1 and OR8 cells (Fig. 2), and do not support the theory that copy number of the sodium pump is involved in the stable phenotype of ouabain resistance in these cells.

Sequence Analysis of the Na,K-ATPase al-Subunit from Ouabain-resistant Cells-Based on the published evidence that the a-subunit H1-H2 domain is an important binding site for ouabain (2,6, 7); we investigated the sequence of this domain in several of our transfectants to determine if a spontaneous mutation may account for the stable phenotype of ouabain resistance.

Using the PCR technique, cDNA corresponding to the coding region of the extracellular loop between the H1 and H2 domains was amplified, cloned, and sequenced as described under “Materials and Methods.” The cDNA sequence from the parent CV1 (monkey) cells differed from human sequence at positions 381 (C>T) and 420 ( A X ) . These differences

28s -

18s-

T A T A T A T A

cv1 OR8 cos COS-RR- FIG. 4. Northern analysis of ouabain-resistant cells. North-

ern blots of total and poly(A)’ RNA from CVl,OR8, COS, and COS- RR- cells. 28 and 18s are labeled as size markers. The blots were probed with a murine Na,K-ATPase-al probe (expected message size is 3.7 kb).

205 -

116-

66-

45“

17275

1 2 3 4 FIG. 5. Protein gel electrophoresis of ouabain-resistant cell

membranes. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of crude cell membranes isolated from CV1 (lane I ) , OR8 (lane 2) , COS (lane 3 ) , and COS-RR- (lane 4 ) cells. There is no observed difference in the amount of a-subunit between either of the transfectants and their parent cell type (CV1 versus OR8 and COS versus COS-RR-). 65 pg of protein was loaded per lane and equality of loading was further assessed by Coomassie Blue staining of duplicate lanes. The blot was hybridized with a polyclonal antibody against rat Na,K-ATPase-a1 (molecular mass = 100 kDa) and exposed for 5 s. Molecular mass markers are indicated in kDa.

resulted in no change in the amino acid sequence from that of human (see Fig. 6). However, in the ouabain-resistant OR8 cells, two different sequences were found. One was identical to the CV1 sequence and is presumably the native monkey sequence. The second sequence, confirmed by independent reverse transcription and PCR, differed from the monkey sequence a t nucleotide 382 (G is changed to A) and resulted in a single amino acid substitution a t position 121 (aspartic acid is replaced by asparagine). This substitution of an uncharged residue (at a position which in all native forms of the pump sequenced so far is invariably aspartic acid (15)) has been previously shown by Price et al. (2) to confer ouabain resistance to primate cells. Thus, the decrease in ouabain sensitivity of rubidium uptake observed in Fig. 2 can be explained by this mutation.

Sequence analysis of the Hl-H2 domain of ouabain-resistant COS-RR- and 3T3-BB- cells has also revealed point mutations. Unlike the mutation in OR8 cells, these two cell lines have mutations in the predicted membrane spanning region, H2, rather than the extracellular domain (see Fig. 7). In COS-RR- cells, nucleotide 424 is changed from A to G with the resultant amino acid change from isoleucine 135 to valine 135. Two mutations have been found in 3T3-BB- cells. On separate clones of the Hl-H2 domain from the ouabain- resistant 3T3 transfectants, we have found a mutation from A-G at position 392 resulting in tyrosine 124-cysteine 124 and C-G a t position 440 resulting in serine 140-cysteine 140. In both COS-RR- and 3T3-BB- cells, the parent al-subunit sequence was also found, implying that only one allele is mutated in these cells. The finding of two separate mutations in 3T3-BB- cells suggests that this cell line is a mixture of


4201 382 CTGTACCTGGGTGTGGTGCTATCAGCCGTTGTAATCAT~CTGGTTGCTTCTCCTACTAT~G~~~G~~~~.~.

CTGTACCTGGGTGTGGTGCTATCAGCC(;TTGTCATCATAACTGGTTGCTTCTCC~ACTAT~.~G~~G~?.~~.,.~.

CTGTACCTGGGTGTGGTGCTA~CAGCCGTTGTCA?CA~~CTG~?TGCTTCTC~TACTATC~.G~~G~~--:-.

l 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 I l l l l l l I I I I l I I l I I I I l I I I I I I I I l l / I I / I l I / I I l I / I l / l ’ : l

1 l 1 1 / 1 1 1 1 1 l 1 l I l l l l l l l l l I I l / / l I l I / l / 1 1 1 l / / I l l I I I / I l / 1 l l l l l I ; l 1 0 1 l l / I

Human -

Monkey -

OR8 -

121‘

FIG. 6. Partial nucleotide and amino acid sequences of the Na,K-ATPase of ouabain-resistant OR8 cells. Results of PCR amplification of the H1-H2 domain of the Na,K-ATPase a-subunit. Nucleotide and amino acid sequences are shown. Predicted extracellular domain is outlined by the box. The monkey (CV1) sequence differs from the human sequence at positions 381 and 420. These changes fall at the third base of the codon with no amino acid change. The G>A mutation at position 382 in OR8 cells results in the substitution of Asp for Asn.

O 1 ’ ’\&: T y r l 2 4 - C y s l 2 4 ( 3 T 3 - B B - )

A s p 1 2 1 - A m 1 2 1 (ORE)

Intracellular B J,

NH2

l l e l 3 5 - V a l - 1 3 5 (COS-RR-)

S e r l 4 0 - C y s l 4 0 ( 3 T 3 - B B - )

11 COOH

FIG. 7. Sites of mutations discovered in ouabain-resistant cells. Model of the Hl-HZ domain of the Na,K-ATPase (adapted from Lingrel et al. (22)) with sites of mutations discovered to date in the ouabain-resistant cell transfectants highlighted.

several different clones which independently mutated during selection with ouabain, a finding consistent with the observed result on the culture plate. Of note, not all of these mutations have been confirmed from independent reverse transcription reactions (see “Discussion”). In preliminary results, no mutations have been found in the H3-H4 or H5-H6 regions of the a-subunit.

DISCUSSION

The 6.5-kb murine genomic DNA isolated by Levenson et al. (1) enhances the rate of appearance of ouabain-resistant clones when transfected into monkey cells or when introduced into mouse cells via a retrovirus. The Na,K-ATPase of the resistant cells appears to have a lower affinity for ouabain as judged by ouabain sensitivity of rubidium uptake. At a time when the cells had been exposed to ouabain for several months, no significant change in the expression of the Na,K- ATPase was found in any of the cell lines characterized. Rather, DNA sequence analysis of the resistant cell lines revealed point mutations leading to amino acid substitutions in the Hl-H2 domain. One of these amino acid changes (Asp121-Asnlzl in OR8 cells) was previously examined by Price

et al. (2) in the sheep a1-subunit. In their work, HeLa cells transfected with the mutated sheep gene acquired a level of ouabain resistance similar to that seen in OR8 cells.

Since our sequence information for OR8 cells shows that both the native monkey a-subunit sequence as well as the sequence containing the Asp121-Asn121 point mutation are present, we fit the rubidium uptake data to a two-component model of ouabain binding. The high affinity site was assigned the same Ki as the parent CV1 cells (255 nM), with the low affinity component being allowed to vary. Using the Mac- Intosh Kaleidagraph curve fitting program, the Ki for the low affinity site (reflecting the mutated a-subunit) was determined to be 3.5 PM, or approximately 10-fold less avid for ouabain. The lines in Fig. 2 were drawn using these constants.

The other mutations which we have found (in COS-RR- and 3T3-BB- transfectants) are confined to membrane spanning regions and have not been previously investigated. How- ever, a recent abstract by Horisberger et al. (16) has shown wabain resistance due to a change in a residue of the H1 membrane-spanning domain (cysteinel13-tyrosinel13). They suggest that the highly hydrophobic nature of ouabain may result in its partitioning into the membrane during binding,

Ouabain-resistant Cells Have Mutations in the Na,K-ATPase 17277

with resultant changes in affinity when mutations occur at intramembrane residues. Although not all of the mutations we report have been confirmed from independent reverse transcription reactions, several factors lead us to believe that they are correct and may play a role in the ouabain-resistant phenotype. These factors include the absence of mutations in the parent cell lines, the absence of mutations in the cytoplasmic domains of the enzyme, and the observation that no mutations have been found which did not lead to an amino acid change. Our results to date would suggest that ouabain affinity, while greatly affected by changes in the amino acids at the first and second membrane-cytoplasmic border, may depend to a lesser degree on several amino acids within the membrane itself. However, reproduction of our newly found H2 point mutations and reintroduction into primate cells will be necessary to determine if they are sufficient to account for the ouabain-resistant phenotype in COS-RR- and 3T3-BB- cells.

The lack of evidence for stable incorporation or expression of MOR6.5 in transfectants and the unexpected finding of point mutations in the ouabain-binding domain of the a- subunit of the Na,K-ATPase imply that the murine gene does not play a role in the long term phenotype of ouabain resistance. Yet when cells were exposed to lethal doses of ouabain, only those cells transfected with the mouse DNA developed ouabain-resistant clones. Further, when the transfection vector contained both neomycin resistance and MOR6.5, the murine sequence was lost during the 2 weeks of selection on neomycin, but still present in large copy numbers at 4 weeks when the cells were initially ouabain-selected.

These results suggest a role for MOR6.5 during the initial selection on ouabain. Several previous reports have also dem- onstrated the occurrence of stable ouabain-resistance in mouse L, HeLa, and MDCK cells (3, 17, 18). Although sequence information is not available from these cells, the stable nature of the resistant cells, the low rate of occurrence of the phenotype (approximately 1 colony/108 cells), and the presence of two populations of Na,K-ATPase in cell membrane preparations suggest a mutation in one allele of the Na,K- ATPase gene as the mechanism of ouabain resistance in these cells. Interestingly, the selection criteria used in those experiments involved concentrations of ouabain resulting only in inhibition of cell division, not cell death (3, 18). However, when lethal concentrations of the drug were used in MDCK cells (17) no occurrences of spontaneously ouabain-resistant cells were reported in over lo9 cells. In this experiment, the mutagen EMS again caused the occurrence of ouabain-resistant MDCK cells even with selection on lethal concentrations of ouabain. In agreement with this, we have also found that low doses of ouabain (100 nM) did result in the spontaneous appearance of ouabain-resistant clones in untransfected MDBK cells (data not presented).

The preceding information can be interpreted to mean that, when exposed to ouabain, spontaneous mutations of the Na,K-ATPase a-subunit which diminish ouabain affinity are selected for because they stabilize the intracellular milieu and allow for continued cell growth and division. However, if the selection criteria are too stringent (ie. the ouabain dose kills the cells too rapidly), the cells are unable to undergo these changes. How then does the murine sequence, MOR6.5, alter this balance and allow for the development of ouabain-resistant cells even in the presence of lethal concentrations of the drug? Obviously, many possibilities exist, including alteration of potassium or sodium leak rates to decrease Na,K-ATPase work load, upregulation of Na,K-ATPase copy numbers in

the membrane, or a direct mutagenizing effect of the transfected sequence.

Although a definitive answer is not yet known, we have found that a 300 base GC-rich region of MOR6.5 cross-hybrid- izes with the 5‘ end of the a-subunit of the Na,K-ATPase.’ When this region was more closely examined with homology search techniques, multiple short (7-12 base) stretches of >90% homology with MOR6.5 were found in the -750 to -100 region of the murine a-subunit. Interestingly, when this same 300-base region of MOR6.5 was used to search GenBank, similar homologies were found, again primarily in the 5’- noncoding regions of mammalian genes, including the @- subunit of murine Na,K-ATPase.’ While as yet unproven, the homology between this segment of MOR6.‘ and the 5’ promoter regions of the a- and @-subunit of the Na,K-ATPase could lead to the transfected murine DNA acting either in cis or trans to transiently upregulate the sodium pump. This could allow the cells to survive long enough for a mutation to occur in the ouabain-binding region of the a-subunit and produce the stable, ouabain-resistant phenotype. Once this mutation has occurred there is no more selection pressure to incorporate MOR6.5, and this exogenous DNA is lost with no further upregulation of pump numbers.

Another possibility is that this sequence can act directly to enhance the rate of mutation in the transfected cells (19,201. In support of this is the finding that the rate of appearance of ouabain-resistant cells following selection on ouabain is higher than the spontaneous rate reported by Baker et al. (3) in 3T3 cells, but is similar to the rate that they and others report following ethyl methanesulfonate mutagenesis (17).

Of note in our results is the lack of similarity between OR8 cells and the previously reported ouabain-resistant OR6 cells (11, 13, 14). These discrepancies include the lower level of ouabain resistance of OR8 cells compared to OR6 cells, the constitutive nature of ouabain resistance in OR8 cells, and the inability to find messenger RNA from MOR6.5 in OR8 cells. Recent investigation of the Na,K-ATPase in OR6 cells has revealed the unexpected result that these primate cells contain the message for the murine al-subunit of the sodium pump, indicating that they were derived by transfection of murine genomic DNA rather than MOR6.5. Furthermore, these cells are more resistant to ouabain than native murine cells, suggesting that some mechanism other than simply the addition of the murine al-subunit is involved. We are pres- ently attempting to characterize the mechanism of “super ouabain resistance” in OR6 cells and believe that direct com- parisons with our CV1 transfectants are not applicable at this juncture? Rather, we feel that the ouabain-resistant OR8 cells reported here are more similar to pRLouaR-6 cells (also derived by transfecting MOR6.5 into CV1 cells) as reported by Pressley and Edelman (21). Both cell types display similar ouabain sensitivity and a lack of inducibility with pregrowth on ouabain.

In conclusion, in our experiments, selection of cultured primate or murine cells on lethal doses of ouabain has never resulted in the appearance of ouabain-resistant clones of cells. However, the 6.5-kb murine genomic DNA described by Levenson et al. (1) is first introduced into these cells, ouabain- resistant clones do appear, but at a very low frequency. Homology between a GC-rich segment of MOR6.5 and the 5’ regulatory region of the Na,K-ATPase a-subunit and other genes may be an important factor in transiently rescuing the cells from the lethal effects of ouabain. However, the stable phenotype of ouabain resistance is not related to the long

X.”. Zhou, L. G. Cantley, M. J. Cunha, J. Epstein, and L. C. Cantley, manuscript in preparation.


term incorporation of the murine DNA into these cells, but rather is due, at least in the case of OR8 cells, to a point mutation in the ouabain-binding domain of the Na,K-ATPase a-subunit.

Acknowledgments-We thank Dr. Robert Levenson for many helpful discussions as well as the gift of several of the cDNA probes and antibodies used in these experiments. We also thank Sharon Graw, See-Ying Tam, and Dr. David Housman for helpful thoughts and the gift of the mouse a-subunit cDNA that we used.

REFERENCES 1. Levenson, R., Racaniello, V., Albritton, L., and Housman, D. (1984) Proc.

2. Price, E. M., Rice, D. A,, and Lingrel, J. B. (1989) J. Biol. Chem. 264,

3. Baker, R. M., Brunette, D. M., Mankovitz, R., Thompson, L. H., Whitmore,

4. Linerel. J. B.. Orlowskl. J., Price, E. M., and Pathak, B. G. (1991) SOC. Gen.

Natl. Acad. Sci. U. S. A. 81, 1489-1493

21902-21906

G. F., Siminovitch, L., and Till, J. E. (1974) Cell 1.9-21

P%ysiol. Ser. 46, 1-16 5. Sweadner, K. J. (1991) Soc. Gen. Physiol. Ser. 46,63-76

6. Price, E. M., and Lingrel, J. B. (1988) Biochemistry 27,8400-8408 7. Emanuel, J. R., Graw, S., Housman, D., and Levenson, R. (1989) Mol. Cell

8. Shull, G. E., Schwartz, A,, and Lingrel, J. B. (1985) Nature 316,691-695 9. Felgner, P. L., Gadek, T. R., Holm, M., Roman, R., Chan, H. W., Wenz,

Acad. Sa . (I. S. A. 84, 7413-7417 M., Northrop, J. P., Ringold, G. M., and Danielsen, M. (1987) Proc. Natl.

11. Schulz, J. T., and Cantley, L. C. (1988) J. Biol. Chem. 263,624-632 10. Hirt, B. (1967) J. Mol. Biol. 26,365-369

12. Sanger, F., Nicklen, S., and Coulson, A. R. (1977) Proc. Natl. Acad. Sci.

13. English, L., Epstein, J., Cantley, L., Housman, D., and Levenson, R. (1985)

14. Epstein, J. A,, and Lechene, C. (1988) Am. J. Physiol. 264, C847-C854 15. Takeyasu, K., Lemas, V., and Fambrough, D. M. (1990) Am. J. Physiol.

16. Horisberger, J. D., and Rossier, B. C. (1991) J. Cell Biol. 116, 201 (abstr.) 17. Soderberg, K., Rossi, B., Lazdunski, M., and Louvard, D. (1983) J. Biol.

18. Robbins, A. R., and Baker, R. M. (1977) Biochemistry 16,5163-5168 19. Bedryshev, G. D., and Masiuk, A. I. (1974) Tsitol Genet 8,452-464 20. Charikova, E. V., Lukash, L. L., and Zalmanzon, E. S. (1983) Biol. Nauki

Bio. 9,3744-3749

U. S. A. 74,5463-5467

J. Biol. Chem. 260,1114-1119

269, C619-C630

Chem. 268,12300-12307

21. Pressley, T. A., and Edelman, I. S. (1986) J. Biol. Chem. 261,9779-9786 22. Lingrel, J. B:, Orlowski, J., Shull, M. M., and Price, E. M. (1990) Prog.

4,86-88

Nuckzc Acld Res. Mol. Bwl. 38,37-89

ouabain-resistant transfectants of the murine ouabain resistance

Documents