ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS

Vol. 296, No. 2, August 1, pp. 539-546, 1992

Coupled Expression of Ca2+ Transport ATPase and a Dihydrofolate Reductase Selectable Marker in a Mammalian Cell System

Arif Hussain,*Tt David Lewis,* Carlota Sumbilla,* Li-Ching Lai,* Peter W. Melera,*,T$ and Giuseppe Inesi* *Department of Biological Chemistry, the TCancer Center, and the $Graduate Program in Molecular and Cell Biology, University of Maryland School of Medicine, Baltimore, Maryland 21201

Received January 10, 1992, and in revised form April 6, 1992

Stable expression of a full-length cDNA encoding chicken fast muscle Ca’+ transport ATPase was obtained in a Chinese hamster lung cell line (DC-3F), using a dual- promoter expression vector (pH@FCaA3) in which the ATPase was cloned downstream of a human &actin gene promoter, and a mutant dihydrofolate reductase cDNA (A3/DHFR) was cloned downstream of an SV40 pro- moter-enhancer. Owing to its essentially normal cata- lytic activity and modest (20-fold) resistance to the an- tifolate methotrexate (MTX), the A3DHFR mutant enzyme served as an efficient dominant selection marker in transfected cell populations challenged with MTX and, within a broad range of drug concentrations, allowed subsequent amplification and overexpression of vector sequences. In stable transfectants, the expressed ATPase was targeted to intracellular membranes, and the micro- somal fractions from those cells exhibited high rates of Ca2+ transport. In comparative experiments using tran- sient expression in COSl cells, the level of ATPase per transfected cell was greater, but less than 5% of the transfected population exhibited ATPase expression. Furthermore, as opposed to the stable lines, the tran- siently expressing cells could not be propagated. Overall, the yield of ATPase was 12-16 and 4-6 rg per milligram of microsomal protein in the stable and the transient e npression systems, respectively. The advantages of the stably transfected cell lines therefore lie in the homo- geneity of ATPase expression and its distribution in cells and microsomes, in the large yield of microsomes ob- tained by continuous cell propagation, and in the repro- ducible functional characteristics of the microsomes. Moreover, the microsomes derived from stably trans- fected cell lines provide a convenient system for studies of Ca2’ transport and ATPase partial reactions, elimi-

0003.9861/92 $5.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.

nating the need to conduct repetitive transient transfec- tions to obtain sufficient amounts of enzyme for functional studies. o 1992 Academic press, IIN.

The intracellular Ca’+-ATPases constitute a family of homologous enzymes that regulate the cytoplasmic Ca2+ concentration by active transport and sequestration into membrane-bound compartments. The physiological im- portance of these enzymes is related to their prominent role in control of contractile activation and other Ca2+- dependent functions. Operation of the Ca2+ pump by the isolated enzymes constitutes an important system for molecular studies of the mechanism by which ATP uti- lization is dependent on Ca2+ bound to enzyme, and cou- pled to active transport of bound Ca2+. These studies have been greatly advanced by the cloning and sequencing of cDNAs encoding these enzymes (1). Subsequent transient expression of these cDNAs in COSl cells allowed recovery of microsomal fractions exhibiting ATP-dependent Ca2’ transport from the transfected cells (2). This opened the way to site-directed mutagenesis and studies on relation- ships of structure and function (2).

Here we describe a system for the stable overexpression of Ca2+ transport ATPase in the Chinese hamster lung (CHL)l fibroblast cell line DC-3F. These cells were trans- fected with an expression vector containing the ATPase cDNA as well as a cDNA encoding a novel, mutant form

r Abbreviations used: CHL, Chinese hamster lung; DHFR, dihydro- folate reductase; MTX, methotrexate; FCS, fetal calf serum; dFCS, di- alyzed fetal calf serum; DMEM, Dulbecco’s modified Eagle’s medium; PBS, phosphate-buffered saline; SDS, sodium dodecyl sulfate; SSC, standard saline citrate.

539

540 HUSSAIN ET AL.

E!coR I

FIG. 1. FIG. 1. Construction of pHBFCaA3. The cDNA encoding a mutant form of DHFR (originally isolated from the MTX-resistant CHL cell line Construction of pHBFCaA3. The cDNA encoding a mutant form of DHFR (originally isolated from the MTX-resistant CHL cell line DC-3F/A3) was inserted into the BglII site downstream of the SV40 promoter in pH@Apr-1-neo. NeoR sequences between the NaeI and EglII DC-3F/A3) was inserted into the BglII site downstream of the SV40 promoter in pH@Apr-1-neo. NeoR sequences between the NaeI and EglII sites were then removed to generate the lO.O-kb plasmid pH@A3-dhfr. The 3.2-kb avian Ca’+-ATPase cDNA FCa was then cloned between Sal1 sites were then removed to generate the lO.O-kb plasmid pH@A3-dhfr. The 3.2-kb avian Ca’+-ATPase cDNA FCa was then cloned between Sal1 and Hind111 sites and the final 13.2-kb construct, designated pHBFCaA3, used for transfection studies in DC-3F cells. IVSI represents intron I and Hind111 sites and the final 13.2-kb construct, designated pHBFCaA3, used for transfection studies in DC-3F cells. IVSI represents intron I of the human &actin gene and has enhancer-like activitv (7). while 5’ UTR corresponds to a 78-bp untranslated leader sequence. Restriction sites of the human &actin gene and has enhancer-like activitv (7). while 5’ UTR corresponds to a 78-bp untranslated leader sequence. Restriction sites _ _, in parentheses were destroyed during the cloning steps.

of dihydrofolate reductase (DHFR), originally isolated from the methotrexate (MTX)-resistant CHL cell line DC-3F/ A3 (3). This cDNA encodes a DHFR enzyme that has un- dergone a single amino acid change at position 22 (i.e., leu to phe), rendering it 20-fold resistant to MTX inhibition, while allowing it to maintain an essentially normal catalytic activity. These properties allow the mutant form of DHFR to act as an amplifiable dominant selectable marker (4).

Using this dual expression vector, we were able to develop CHL cell lines that stably integrate and express Ca2+- ATPase sequences in parallel with DHFR sequences. Se- lection with increasing concentrations of MTX allowed the isolation of cell populations that amplified and overex- pressed both sequences, yielding microsomal fractions ex- hibiting ATP-dependent Ca2+ transport. This system pro- vides a useful approach for obtaining large amounts of active microsomes, while avoiding the variability of, and the large effort required by, transient expression systems.

MATERIALS AND METHODS

Plusmidconstruct The full-length 3.2-kb avian Caz+-ATPase cDNA, designated FCa, was removed from pUC19FCa (5) by first linear&ii the plasmid at the 5’ end of the cDNA by partial cute with EcoRI (there is an internal EcoRI site within the coding sequence of FCa and also one at the 3’ end of the cDNA), followed by digestion at the 3’ end with I@nI. This fragment was then inserted between the EcoRI and I@nI sites of the 3.7-

kb plasmid pcDL-S&296 (6) to generate pcDL-SRolFCa. The ATPase cDNA in pcDL-S&FCa is driven by a promoter system which is composed of the SV40 early promoter, the R segment, and part of the U5 sequence of the long terminal repeat of human T cell leukemia virus type I (6). This construct was used to transfect COSl cells.

A second construct was built by inserting a cDNA encoding a mutant form of DHFR containing a leuzL + phe mutation (3) downstream of the SV-40 promoter in the dual-promoter expression vector pHflApr-1-neo (7) (Fig. 1). This involved removing the 713~bp mutant DHFR cDNA, designated A3, from pUCA3 by BumHI and BglII, and inserting it into the BglII site of pHflApr-l-neo. The neoR gene sequence between the BgZII and the A&I sites was then removed and the plasmid religated to itself after filling in the 5’ overhang generated by BglII with Klenow (Fig. 1). The final nature of this lo-kb construct, designatedpH@A3dhfr, was confinncd by sequencing across all cloning junctions. This vector allows selection of transfected cells by using MTX. Subsequently, the Ca*+-ATPase cDNA FCa was inserted downstream of the P-actin promoter in pHflA&dhfr to generate the 13.2- kb construct, pHmCaA3. In this configuration, the DHFR cDNA lies downstream of an SV40 promoter-enhancer, while the ATPase cDNA is driven by a human $actin promoter (Fig. 1). pH@CaA3 was subsequently used to transfect DC-3F cells. A single mutation (GAA-977 to CAG) in the ATPase cDNA was obtained with the aid of the Promega Mutagenesis Kit and confirmed by sequence analysis.

Cell culture and trunsfections. Monkey kidney COSl cells were maintained in DMEM media (GIBCO) without pyruvate and supple- mented with 4.5 g/liter glucose and 10% fetal calf serum (FCS), while the CHL cell line DC-3F was maintained in (Y-MEM/lO% dialyzed FCS (dFCS; GIBCO).

Transient transfections in COSl cells were carried out using the DEAE-dextran method (8). Briefly, exponentially growing COSl cells

Ca’+-ATPase EXPRESSION 541

were plated at a density of 5 X lo6 cells/l50mm dish 24 h pretransfection. Ethanol-precipitated plasmid DNA (at 4 gg DNA/plate) was resuspended in phosphate-buffered saline (1X PBS) and mixed with the diluted DEAE-dextran solution supplied by Stratagene (the final concentration of DEAE-dextran in media being 400 rig/ml) under conditions recom- mended by the manufacturer. On the day of transfection, media were removed, the plates were washed twice with 1X PBS, and 660 pl of the DNA/DEAE-dextran mixture was added dropwise to the plates. After incubating at 25°C for 15 min without CO*, the DNA solution was removed, and the plates were rinsed in 1X PBS and treated with 100 pM chloroquine (Sigma) in DMEM/2% FCS for 3 h at 37°C. Subse- quently, the media were aspirated and, after rinsing the plates in 1X PBS, the cells were fed with complete media (DMEM/lO% FCS). Forty- eight hours later, the cells were harvested for subsequent studies.

Stable transfections were carried out using calcium phosphate-DNA coprecipitates as previously described (9). Briefly, 1 X 10r’ cells were plated per lOO-mm dish the day prior to transfection, which was followed by 4 h exposure to calcium phosphate-plasmid (10 pg/plate) coprecip- itates the next day. The cells were glycerol-shocked (10% glycerol) for 1 min, washed twice in 1X PBS, and incubated in nonselective media supplemented with regular FCS overnight. Sixteen hours later, the media were changed to selective media (i.e., a-MEM/lO% dFCS plus 0.03 ng MTX (Lederle)/ml). Media were changed every 3-4 days and the colonies picked and clonally expanded after lo-14 days.

Zmmurwdetection studies. Immunofluorescent visualization of the expressed ATPase in the transfected cells was obtained as described by Karin et al. (5). Western blots were performed by the method of Towbin et al. (10) using the monoclonal antibody CaF3-IgG (5) which is specific for the avian sarcoplasmic reticulum ATPase. Quantitation of the ex- pressed ATPase was obtained by densitometry of films derived from Western blots, standardized with determinations of the phosphorylated ATPase intermediate by radioactive tracer.

Functional aasays. Preparation of microsomes from transfected cells,

5%. Furthermore, this high level of expression interferes with survival and reproduction of the transfected cells, thereby limiting further yield of protein.

We then attempted to obtain stable expression by trans- fecting DC-3F cells with pHpFCaA3 using MTX for selec- tion. The initial clonal populations obtained upon selection with MTX demonstrated minimal ATPase expression, as judged by the fluorescence signals in these cells (not shown). However, when DC-3F cells transfected with pH/3FCaA3 and initially selected in 0.03 pg MTX/ml were grown se- quentially in increasing concentrations of MTX, their expression of ATPase increased at each step of selection. Figure 2B represents a subclone obtained from a mass pop- ulation of transfected DC-3F cells that were grown stepwise in increasing MTX concentrations and finally maintained in 50 pg MTX/mI. In contrast to the transiently transfected COSl cells (Fig. 2A), this clonal population demonstrates a homogeneous distribution of ATPase-specific fluorescence in all the cells. The relatively high level of expression of ATPase in the stably transfected mass populations and in the clonal sublmes derived from them (see below) does not interfere with survival and reproduction of the transfected cells. Indeed, the level of ATPase expression in these lines is maintained over the course of multiple cell passages, with approximately a 50-60% decrease in expression occurring gradually over the course of 6 months of continuous maintenance in culture.

and studies of calcium transport and phosphoenzyme formation with [Y-~~P]ATP in the presence of Ca*+ were carried out as described by Clarke et al. (11).

Detection of Expressed ATPase in Microsomes Prepared

Dot blots. High molecular weight cellular DNA was prepared from from the Transfected Cells

various cell lines by proteinase K-SDS digestion followed by phenol- chloroform extraction (12). DNA was removed with RNaseA followed

Owing to the intracellular membrane targeting of the

by phenol-chloroform extraction and ethanol precipitation. Four-mi- Ca2+-ATPase, it is possible to recover the enzyme in mi-

crogram samples of DNA were then spotted onto nitrocellulose filters crosomal fractions obtained by homogenization and differ-

as previously described (13). Ca’+-ATPase and A3 DHFR cDNAs were ential centrifugation of transfected cells (2). Thus, micro- removed from pUC19FCa and pSVA3, respectively, and the inserts were oligolabeled (14) with [ol-32P]dCTP (3000 Ci/mmol; NEN) to specific

somes were prepared from transfected cells, their protein

activities > 5 X 10’ cpm/pg DNA and hybridized to the filters at 65°C components separated by gel electrophoresis, and Western

for 20-24 h in 4X SSC, 8 mM EDTA, and 300 pg/ml denatured salmon blots, using an anti-chicken Ca2+-ATPase monoclonal an-

sperm DNA. Subsequent washings of the filters were carried out at 65’C tibody, carried out. Microsomal preparations obtained from in 0.1X SSC and 0.1% SDS. a transiently transfected COSl cell culture, and a stably

RESULTS

transfected DC-3F cell culture, are shown in Fig. 3. The exclusive immunoreactivity of a single band of comparable

Transient and Stable Expression of ATPase

In order to evaluate the efficiency of transfection under various conditions, we first carried out a series of exper- iments using a transient expression system by transfecting COSl cells with pcDL-SRaFCa. Detection of ATPase in the transfected cells, using specific anti-chicken Ca2+- ATPase monoclonal and secondary fluorescent antibodies, demonstrated maximal levels of expression 3 days follow- ing transfection. It is shown in Fig. 2A, for example, that extremely high levels of ATPase are expressed and extend throughout the entire cytoplasm. However, the percentage of effectively transfected cells, as revealed by the ratio of fluorescent to nonfluorescent cells, is limited to only 4-

intensity appears in the two preparations (Fig. 3A) and cor- responds to the electrophoretic mobility of ATPase in such gels. The specificity of the immunodetection is demonstrated by the lack of reaction in nontransfected samples (Al and A4). As lower amounts of microsomal protein were used in samples 5 and 6, it is apparent that the concentration of expressed ATPase is approximately three times higher in the microsomes derived from the stable DC-3F lines com- pared to that of the microsomes derived from the COSl cells. A somewhat lower expression was obtained with the mutated cDNA in both COSl and DC-3F lines.

Another method for the electrophoretic detection of expressed ATPase is related to the ability of this enzyme to form a phosphorylated intermediate by utilization of

542 HUSSAIN ET AL.

FIG. 2. Immunofluorescent visualization of avian Ca*+-ATPase expression. (A) High magnification image of COSl cells transfected with the pcDL-SRaFCa vector. The cells were lightly fixed in formaldehyde, permeabilized with saponin, exposed to CaF3-IgG, and visualized with rhodamine-conjugated goat anti-mouse IgG (5). Approximately 5% of the cells exhibit fluorescent staining in this preparation. (B) DC-3F cells transfected with pHPFCaA3 and selected with MTX. This was a clonal line obtained from a mass population sequentially selected in increasing MTX concentrations and finally maintained in 50 pg MTX/ml. All cells in this preparation exhibit homogeneous fluorescent staining.

the ATP terminal phosphate in the presence of Ca2+ (2). In fact, we were able to show a signal corresponding to the ATPase, following exposure of the microsomes to [y- 32P]ATP (Fig. 3B). In this case, a very faint band is noted in the nontransfected samples (Bl and B4), due to the presence of some native protein. On the other hand, a much stronger signal is noted in the samples transfected with wild-type cDNA (B2, B5, and B6), in proportion to the ATPase content revealed by the Western blots (A). A much lower signal is noted in the mutated samples (B3 and B7), comparable to or only slightly h~igher than that in the controls. It was previously reported that this single mutation interferes with Ca2+ activation of the enzyme and, therefore, with utilization of ATP (11).

Ca2+ Transport Activity of Microsomes Prepared from Transfected Cells

The most reassuring and quantifiable feature of the ex- pressed ATPase is its ability to sustain Ca2+ transport in the presence of ATP (2). The calcium transport activities

of microsomes prepared from COSl or DC-3F cells exhib- iting optimal levels of expression are shown in Figs. 4A and 4B, respectively. Both preparations sustain good and pro- longed activity when oxalate is included in the reaction mixture in order to limit inhibition of transport by an ex- cessive rise of the Ca2+ concentration in the lumen of the microsomal vesicles (15). It is apparent in Fig. 4 that the initial rate of transport is similar in experiments carried out with microsomes prepared from the stably transfected DC-SF cells or the transiently transfected COSl cells. How- ever, the activity of the microsomes obtained from tran- siently transfected cells tends to decline after several minutes (Fig. 4A), while the activity of the microsomes obtained from stably transfected cells does not exhibit any decline within the 30 min of experimental observation (Fig. 4B). As discussed below, this difference may be due to a more homogeneous distribution of enzyme in a greater proportion of microsomal vesicles in the stable lines.

The initial rates of Ca2+ transport provide a convenient means for evaluating expression of functional ATPase fol- lowing selection with MTX. The pHPFCaA3 construct re-

Ca’+-ATPase EXPRESSION

tains the coding sequences of the mutant A3 DHFR enzyme which serves as an efficient, dominant, and readily ampli- fiable selectable marker (4). Therefore, further selecting ini- tial transfectants in increasing concentrations of MTX al- lows for the isolation of cells that amplify exogenous DHFR sequences and coamplify additional sequences that are part of the same construct. Hence stepwise selection in MTX results in amplification and overexpression of the transfected ATPase sequences in DC-3F cells.

The EDw of DC-3F cells for MTX is 0.02 pg/ml in CV- MEM, dFCS. Multiple colonies resulted upon transfection of DC-3F with pH@FCaA3 when initial selection was carried out in 0.03 fig MTX/ml. Low levels of ATPase were ex- pressed in many of these clonal populations, as demon- strated by fluorescence (not shown) and Ca’+ uptake (Fig. 5) studies. Several of these initial clones were then grown sequentially in increasing concentrations of MTX. Figure 5 demonstrates the initial rates of Ca2+ uptake by the mi- crosomal fractions obtained from cell populations selected in increasing MTX concentrations. Cells selected in 0.03 pg MTX/ml are clonal populations, while subsequent se- quential selection of such clones in 0.3, 1.0, 2.0, 5.0, 25.0, 50.0, and 100.0 pg MTX/ml results in generation of mass populations at each step of selection. Thus, Ca2’ uptake studies at the different selection concentrations reflect mi- crosomal ATPase activity of the mass populations at those

concentrations. As seen in Fig. 5, the transport activity of microsomes prepared from DC-3F cells maintained in MTX concentrations greater than 25 pg/ml is approximately 20- fold higher than that of microsomal fractions obtained from the initial clonal populations that were selected at 0.03 pg MTX/ml. Essentially no further increase in transport ac- tivity was seen in the mass populations by increasing the selective MTX concentration above 25 pg/ml.

It should be pointed out that if MTX is removed from the maintenance media, ATPase activity of microsomal fractions from the stable lines decreases as a function of time. Thus, activity decreased to 75% of maximal 1 week off drug, to 50% 2 weeks off drug and was unde- tectable 4 weeks off drug. This demonstrates the ne- cessity of maintaining selective pressure in order to re- tain exogenous DHFR (and hence ATPase) sequences in transfected cells. Interestingly, however, ATPase expression decreased by 50-60% over a period of 6 months of continuous cell passage even though the mass populations were maintained in the appropriate con- centrations of methotrexate. While these results are not surprising in light of the fact that MTX provides selective pressure on DHFR and not on ATPase, they do suggest that, with time, cells expressing relatively high levels of ATPase might be selected against. It is likely that the relative instability of amplified DNA (16,

544 HUSSAIN ET AL.

12 3 4567

FIG. 3. Electrophoretic characterization of expressed ATPase. West- ern blots (A) and autoradiograms of the ATPase-phosphorylated inter- mediate (B) were obtained with microsomes isolated from COSl cells transfected with pcDL-SRoFCa (l-3) and from DC-3F cells transfected with pHPFCaA3 and selected with MTX (4-7). The numbers correspond to microsomes obtained from control cells (1 and 4), cells transfected with wild-type cDNA (2, 5, and 6), and cells transfected with cDNA containing the single mutation GAA-977 to CAG, resulting in change of Glu309 to Gln in the expressed protein (3 and 7). In order to evaluate the specific content of ATPase in the microsomes, the following amounts of microsomal protein were placed in the electrophoretic wells: 20 pg in wells 1, 2, 3, 4, and 7; 5 kg in well 5; and 10 pg in well 6.

17) contributes in part to the loss of ATPase activity in the long-term cell cultures.

Evaluation of Gene Copy Numbers

The transport activity of the microsomal fractions at each level of MTX selection represents, with the exception of the clonal lines that were selected at 0.03 pg MTX/ml, the average activity of all the cells within the heteroge- neous mass populations. The dot blot in Fig. 6A dem- onstrates Ca’+-ATPase-specific sequences in a clonal line selected at 0.03 1.18 MTX/ml and in other sublines derived from it by further selecting in sequentially increasing concentrations of MTX. As can be seen in Table I, se- lecting transfected cells in increasing MTX concentra- tions results in amplification of the introduced Ca’+- ATPase sequences. However, the gene copy number does not necessarily increase at each step of selection.

Cloning individual cells from mass populations enabled us to obtain cell lines that were homogeneous with respect to ATPase distribution. Several subclones were obtained from a mass population that was resistant to 50 fig MTX/ ml, and Fig. 2B represents an example of one such clonal line. It should be pointed out that this parental mass pop- ulation (designated DC-3F/50.0) had been maintained in 50 /*g MTX/ml continuously for several months and there- fore, at the time of selecting clonal sublines from it, the transport activity of its microsomal fraction had decreased from 30 to 11 nmol/mgprotein/min (Table II). As expected, the three subclones had different gene copy numbers and rates of Cazf uptake from each other and from the parental mass population, reflecting the heterogeneity of the cell

Time (minutes) oe-7i7r- 30

Time (minutes)

FIG. 4. Calcium transport activity of microsomes obtained from cells transfected with vectors containing the inserts (+), or vectors which did not contain the ATPase insert (A). Microsomal vesicles were obtained from COSl cells transfected with pcDL-SRoFCa and exhibiting transient expression (left panel) or from DC-3F cells transfected with pHPFCaA3 and maintained in MTX selection (right panel). Calcium transport was obtained at 25°C in a reaction mixture containing 20 mM Mops, pH 7.0,60 mM KCl, 5 mM MgC12,0.2 mM EGTA, 0.2 mM [%a]CaClB, 10 c(g protein/ml, 5 mM potassium oxalate, and 3 mM ATP. Fractional samples were filtered serially through 0.45-pm Millipore filters. The filters were washed with 3 mM LaC13 and processed for scintillation counting.

population making up the parental line (Fig. 6B, row a, and Table II). The Ca2+-ATPase gene copy number in DC-SF/ 50.0 is the average of all the Ca2+-ATPase sequences present in individual cells making up this mass population. As seen

6oa

a

0 l

0

l

l

l

01 ’ 9 . . . . . ., . . . . . . , . . . . . . , . . . . . . . , 0.01 0.1 1 10 100

@g/ml) Methotrexate FIG. 5. Initial rates of Ca’+ transport as a function of increasing MTX selection. DC-3F cells transfected with pHBFCaA3 were initially selected in 0.03 gg MTX/ml and sequentially grown in increasing con- centrations of MTX. Microsomal preparations were obtained at the different selection concentrations, and transport studies were performed as described in the legend to Fig. 4. The rates were derived from the initial slopes of time curves as shown in Fig. 4.

Ca*+-ATPase EXPRESSION 545

12 3 4 5 6 1 2 3 4 5

a a

Panel A PanelB

FIG. 6. DNA dot blots. (A) (24 h exposure) Genomic DNA samples from a DC-3F cell line transfected with pHbFCaA3 and selected at 0.03 pg MTX/ml and other sublines derived from it (by further selecting in sequentially higher concentrations of MTX) were spotted onto nitro- cellulose filters (4 kg DNA per dot) and hybridized to a 3.2-kb avian Ca’+-ATPase cDNA insert probe (FCa) (a). A duplicate blot was hy- bridized to a 1.4-kb PstI fragment of the hamsterp-glycoprotein cDNA (19) as an internal control (b). Column 1, nontransfected DC-3F; column 2, DC-3F/0.03; column 3, DC-3F/0.3; column 4, DC-3F/2.0; column 5, DC-3F/25.0; column 6, DC-3F/50.0. (B) (9 h exposure) Genomic DNA samples from a mass population of DC-SF cells originally transfected with pHfiFCaA3 and maintained in 50 gg MTX/mI (designated DC- 3F/50.0), along with DNA samples from three clonal sublines derived from it, were spotted onto nitrocellulose filters. Each dot contains 4 pg of DNA. Column 1, nontransfected DC3F; column 2, DC-3F/50.0; col- umn 3, clone 1; column 4, clone 2; column 5, clone 3. Filter hybridized with the Ca’+-ATPase cDNA FCa (a). Filter hybridized with the A3/ DHFR cDNA (b). Both filters were stripped and hybridized to a 1.4-kb PstI fragment of the hamster p-glycoprotein cDNA (19) as an internal control (c).

in Table II, individual clonal sublines could be obtained from this mass population, in which Ca’+-ATPase sequences were present in much higher numbers than in the parental line. Analysis of the same dot blots with DHFR-specific insert probes (Fig. 6B, row b) revealed that the ATPase gene copy number essentially changed in parallel with the DHFR gene copy number (Table II). This coupling is con- sistent with the fact that both the ATPase and the DHFR cDNA sequences lie on the same transfecting plasmid pHPFCaA3 (Fig. 1).

As is apparent from Tables I and II, increasing gene dos- age may or may not be associated with a commensurate increase in Ca’+-ATPase expression (i.e., Ca’+ transport activity). The fact that the gene copy number of transfected sequences does not necessarily correlate one-to-one with its expression level (which may be higher or lower than the respective gene copy number) is consistent with the random nature of integration of exogenous sequences into the host genome during transfection and with the occurrence of gene rearrangements during the amplification process. That is to say, sites of integration of exogenous sequences and their possible rearrangements during amplification influence the efficiency with which they are subsequently transcribed.

DISCUSSION

TABLE I

Gene Copy Number of Ca’+-ATPase Sequences in the Stably Transfected DC-3F Cells Shown in Fig. 6A

Fold amplification Ca”+ uptake

(nmol/mg protein/min)

DCBF 1.0 0.0 DC-3F/0.03 1.3 2.0 DC-3F/0.3 5.4 18.0 DC-3F/2.0 5.0 20.0 DC-3F/25.0 12.0 47.0 DC-3F/50.0 15.0 48.0

Note. DC-3F/0.03 represents a clone derived by selecting DC-3F in 0.03 pg MTX/ml post-transfection with pHBFCaA3; DC-3F/0.3 is de- rived from DC-3F/0.03 by further selecting the latter in 0.3 Kg MTX/ ml; and DC-3F/2.0 is derived from DC-3F/0.3 by selecting in 2.0 pg MTX/ml. The DC-3F/25.0 cell line was derived from DC-3F/2.0 in several steps (i.e., DC-3F/2.0 was grown sequentially in 5.0, 10.0, and then 25.0 pg MTX/ml). DC-3F/50.0 was derived from DC-3F/25.0 in a single step by selecting the latter in 50.0 fig MTX/ml. A Betascope 603 Blot Analyzer was used to quantitate the signal intensity of each dot and normalized to that of DC-3F. The corresponding rates of Ca” transport for each line are also shown.

quently be carried out. By using a novel mutant form of DHFR as a selectable marker, we were able to develop cell lines whose microsomal fractions were active with respect to ATPase transport activity. DHFR is an essential enzyme for reduction of dihydrofolate to tetrahydrofolate and is in- hibited by the folate analog MTX. The mutant A3/DHFR cDNA used in our experiments was originally isolated from the MTX-resistant CHL cell line DC-3F/A3 (3). The A3 mutant enzyme has undergone a leu to phe change at amino acid position 22 of the coding sequence, rendering the en-

TABLE II

Gene Copy Number of Ca”-ATPase and DHFR Sequences in DC-3F/50.0 and Three Sublines Derived from It

DC-SF DC-3F/50.0 Clone 1 Clone 2 Clone 3

Fold amplification

Ca”-ATPase DHFR

1.0 1.0 4.0 5.0 8.0 6.0

15.0 13.0 20.0 19.0

Ca2+ uptake (nmol/mg protein/min)

0.0 11.0 14.0 16.0 29.0

Note. DC-3F/50.0 was obtained from DC-3F/0.03 by growing the latter in sequentially increasing concentrations of MTX. Originally, the Ca*+ transport activity of DC-3F/50.0 was 30 nmol/mg protein/min, but dropped to 11 nmol/mg protein/min at the time subclones were derived from it. Clones 1, 2, and 3 were derived from DC-3F/50.0 by dilutional

The experiments reported here were motivated by a spe- cloning. Quantitation of the above sequences was done as in Table I. It

cific interest in expressing relatively large amounts of Ca2+ should be noted that DC-3F/50.0 in this table represents a different

transport ATPase so that functional studies could subse- mass population from DC-3F/50.0 of Table I; hence, the Ca*’ transport activities of the two lines are different.

546 HUSSAIN ET AL.

zyme 20-fold resistant to MTX inhibition, while allowing it to maintain the catalytic activity of the wild-type enzyme (18). Resistance to MTX inhibition in conjunction with an essentially normal catalytic activity allows the mutant DHFR to act as an efficient dominant selectable marker. Indeed, initial selection at low levels of MTX (i.e., 0.03 pg/ ml) post-transfection with pH@FCaA3 resulted in numerous colonies, many of which expressed Ca’+--ATPase activity, albeit at low levels (Fig. 5). Increasing the selective concen- tration of drug stepwise to 25 pg MTX/ml generated anti- folate-resistant mass populations whose ATPase activity increased at each step of selection, resulting in an approx- imately 20-fold elevation in expression of Ca2+-ATPase ac- tivity over the initial clones.

As opposed to the homogeneous and stable expression in CHL cells transfected with pHBFCaA3 and selected with MTX, only a small percentage of COSl cells transfected with pcDL-SRaFCa express ATPase (compare Figs. 2A to 2B). However, expression in each of the transiently trans- fected COSl cells is exceedingly high, which may in part be due to the efficiency of the promoter system of the original pcDL-SRa296 plasmid (6). The transiently transfected COSl cells express ATPase for only several days, and therefore, to obtain sufficient amounts of enzyme for studies of transport function, repetitive transfections have to be carried out. Expression in stable lines established by MTX selective pressure involves the entire cell population, but the lower expression per cell in these lines results in an overall yield of ATPase per milligram microsomal protein that is approximately three times greater than that in the transient system. The stable systems, however, allow con- tinuous propagation of the line and unlimited ATPase yield.

An important difference between the microsomes derived from transient and stable expression experiments pertains to the distribution of expressed ATPase in the microsomes, and the ATPase density within the microsomal membrane. Using specific anti-chicken Ca2+-ATPase monoclonal and gold-labeled secondary antibodies, we found that the ATPase was highly concentrated in a very small fraction (2-3s) of the microsomes derived from transient expression experiments, while a lower density of ATPase was found in approximately 50% of the microsomes derived from stable lines generated by using MTX selection. Consequently, the greater microsomal volume available per ATPase unit in stably transfected cells allows Ca” transport to proceed at constant velocity for a long period of time without inhibition by the rising concentration of lumenal Ca2’, in contrast to what is seen with microsomes derived from transient expression experiments (Fig. 4).

In conclusion we demonstrated that, owing to its novel enzymological characteristics, the mutant A3/DHFR can be used as a dominant selectable marker, allowing am- plification and overexpression of Ca2+-ATPase cDNA se- quences that are placed in the same vector under control

of an independent promoter. The resulting CHL cell line carries out homogeneous and stable expression of ATPase targeted to intracellular membranes and yields micro- somes well suited to studies of active transport and in- termediate ATPase reactions. Although stable expression of avian Ca2+-ATPase has been previously described in nonamplifiable expression systems (5), functionally intact microsomal fractions cannot be successfully obtained from such lines. Use of a novel, amplifiable dominant selectable marker has allowed us to develop stable lines which, in principle, can yield an unlimited supply of microsomes that maintain Ca2+-ATPase transport activity.

ACKNOWLEDGMENTS The authors are indebted to Drs. Malcolm Campbell, Douglas Fam-

brough, and Norman Karin for supplying the cDNA encoding the Ca2+- ATPase and for help with the fluorescence studies. The authors also thank Jan Price for technical assistance. This work was supported in part by Grant PO 1 HL27867 from the National Heart, Lung, and Blood Institute (G.I.), NIH Grant CA-44678 (P.W.M.), NC1 Clinical Inves- tigator Award CA-01298 (A.H.), the American Heart Association (G.I.), and SAPEC S.A.

REFERENCES

1.

2.

3.

4.

5.

6.

I.

8.

9.

10.

11.

12.

13.

14. 15.

16.

17.

18.

19.

MacLennan, D. H., Brandl, C. J., Korczak, B., and Green, N. M. (1985) Nature 316, 696.

Maruyama, K., and MacLennan, D. H. (1988) Proc. Natl. Acad. Sci. USA 86,3314.

Melera, P. W., Davide, J. P., Hession, C. A., and Scotto, K. W. (1984) Mol. Cell. Biol. 4, 38.

Hussain, A., Lewis, D., Yu, M., and Melera, P. W. (1992) Gene. 112,179.

Karin, N. J., Kaprielian, Z., and Fambrough, D. M. (1989) Mol. Cell. Biol. 9, 1978.

Takebe, Y., Seiki, M., Fujisawa, J. I., Hoy, P., Yokota, K., Arai, K. I., Yoshida, M., and Arai, N. (1988) Mol. Cell. Bial. 8, 466.

Gunning, P., Leavitt, J., Muscat, G., Ng, S. Y., and Kedes, L. (1987) Proc. Natl. Acad. Sci. USA 84, 4831.

Lopata, M. A., Cleveland, D. W., and Sollner-Webb, B. (1984) Nucleic Acids Res. 12, 5707.

Wigler, M., Pellicer, A., Silverstein, S., and Axel, R. (1978) Cell 14, 725.

Towbin, H., Staehlin, T., and Gordon, J. (1979) Proc. Natl. Acad. Sci. USA 76,435O.

Clarke, D. M., Maruyama, K., Loo, T. W., Leberer, E., Inesi, G., and MacLennan, D. H. (1989) J. Biol. Chem. 264,11,246.

Gross-Bellard, M., Oudet, P., and Chambon, P. (1973) Eur. J. Biochem. 36,32.

Kafatos, F. C., Weldon-Jones, C., and Egstratiodis, A. (1979) Nucleic Acids Res. 7, 1541.

Feinberg, A. P., and Vogelstein, B. (1983) Anal. Biachem. 132, 6. Hasselbach, W. (1964) Prog. Biophys. Biophys. Chem. 14, 169.

Kaufman, R. J., and Schimke, R. T. (1981) Mol. Cell. Biol. 1,1069.

Federspiel, N. A., Beverly, S. M., Schilling, J. W., and Schimke, R. T. (1984) J. Biol. Chem. 269,9127.

Albrecht, A. M., Biedler, J. L., and Hutchinson, D. J. (1972) Cancer Res. 32, 1539.

Devine, S. D., Hussain, A., Davide, J. P., and Melera, P. W. (1991) J. Biol. Chem. 266, 4545.