Chinese Hamster Ovary Cell Mutants Deficient in an Anion Exchanger

Vol. 263,No. 35,Issue of December 15,pp. 18607-18613,1988 Printed in U.S. A.

THEJOURNALOF BIOLOGICAL CHEMISTRY

0 1988by The American Society for Biochemistry and Molecular Biology, Inc.

Chinese Hamster Ovary Cell Mutants Deficient in anAnion Exchanger 3” Functionally Similar to the Erythroid Band (Received for publication, April 13, 1988)

Ada ElgavishSO,Jeffrey D. Eskoll, and Alexander Knurr$)I From the Departments of $Pharmacologyand llBiochemistry, Schoolsof Medicine and Dentistry, University of Alabama, Birmingham, Alabama 35294

Studies in Chinese hamster ovary cells demonstrate isolated from the renal cortical epithelium (Schneider et al., the presence of an anion exchanger, which has some of 1984), Ehrlich ascites tumor cells (Levinson and Villereal, the properties of the band 3 transporter in erythro1974; Hoffman, 1982),Vero cells (Olsnes et al., 1987; Tonnescytes. 1) Extracellular chloride isa competitive inhib- sen et al., 1987),human lung fibroblasts (Elgavish et al., 1985), itor of sulfate influx and stimulates sulfate efflux, sug- Chinese hamster ovary (CHO)’ cells (Esko, 1986), tracheal gesting that the mechanism of uptake is SOf-/Cl- ex- epithelium (Elgavish et al., 1987) and skin fibroblasts (Elgavchange. 2) The anion exchange inhibitor 4,4‘-diiso- ish and Meezan, 1988). The substrate specificity of this carthiocyanostilbene-2,2‘-disulfonicacid inhibits sulfate rier-mediated transport is similar in erythroid and non-erythuptake in a dose-dependent manner. Half-maximal in- roid systems exhibiting a high affinity for chloride, bicarbonhibition is achieved at 0.06 k~ 4,4’-diisothiocyanostil- ate, and sulfate. This anion transportoccurs in erythroid and bene-2,2’-disulfonic acid. 3) Low extracellular pH non-erythroid cells via an electroneutral system (Knauf, 1979; markedly stimulates sulfate uptake.A 6-fold decrease K,,,is observed at pH,,,,, 5.5 as compared Lucke et al., 1981; Passow, 1986; Schneider et al., 1984). in the apparent to pK,, 7.5. However, studies carried out over a broad Several mechanisms by which electroneutrality maybe range of extracellular SOf- concentrations indicate the achieved have been proposed H+:SO;-/Cl- in erythrocytes presence of three components of this transport activity (Gunn, 1973), Na’/NaSO; or Na+/H+/SO:- in brush border in Chinese hamster ovary cells: two high affinity low membrane vesicles isolated from the ileum (Langridge-Smith < et al., 1983) and the renal cortex (Schneider et al., 1984), and capacity systems, one in the range0.5 k M < [sot-],,,, in brush 5 0 p M and one in the range50 p M < [so$-],,,, < 150 k M , a dual exchange mechanism Na‘/H’:SO;-/20Hand alow affinity, high capacity system (at [SOf-lOut> border membrane vesicles isolated from the ileum (Schron et 150 PM). These properties have not been previously al., 1985). It is possible that these differences are the result of 3 transporter. The reported for the erythroid band structural diversity in a basically similar aniontransport availability of mutants deficient in these activities has membrane protein leading to variations in the affinity for H+ enabled us to carry out studies which suggest that the or Na+ (K+). high affinity systems are functionally independent of The anion transportfunction of the erythroid band resides 3 the low affinity system, but that all systems are de- in a membrane-associated domain (60 kDa) (Cabantchik, pendent on the same anion exchange protein. Studies 1983). Several approaches have been used to correlate regions in a mutant which lacks allcomponents of the transport within this domain with distinct functional properties of the activity indicates that the anion exchanger may be carrier. One approach was to use a combination of labeling instrumental in the regulation of the intracellular pH and cleaving agents of defined chemical and permeation propin Chinese hamster ovary cells. erties. Data on labeling and cleaving of the band 3 protein by a variety of agents, in conjunction with the inhibitory effects of these agents on the transport activity, were taken to indiIn erythrocytes, anion transport occurs via an anion ex- cate that at least two transmembrane segments of band 3 changer, also referred to as band 3, which is a major integral cooperate in the formation of the putativetransportsite membrane protein (Knauf,1979; Passow, 1986). Certain func- (Cabantchik, 1983). A second powerful approach recently tional similarities between the band 3 anion exchanger in became available when a full-length cDNA clone encoding the exchange protein was isolated (Kopito erythrocytes and non-erythroid anion transportsystems may erythroid band-3 anion exist. As in the erythrocyte, anion transport has been shown and Lodish, 1985).The availability of the cDNA clones should to occur via a carrier-mediated system in hepatocytes (Von allow one to dissect the relationship of the primary sequence Dippe and Levy, 1982), the ileal epithelium (Knickelbein et of band 3to various transport properties. A third possible approach to determine how distinct funcal. 1985; Lucke et al., 1981) brush border membrane vesicles tional properties of the carrier are related to specific regions * This study was supported by a grant from the American Lung within the carrier protein is to use mutants deficient in various Association of Alabama/American Lung Association and the Cystic aspects of the transport activity. In this paper we show that Fibrosis Foundation (to A. E.) and by Grant GM33063 from the National Institutes of Health (to J. D. E.). The costs of publication an anion exchange activity functionally similar to the erythof this article were defrayed in part by the payment of page charges. roid band 3 exchanger is present in CHO cells. We compare This article must therefore be hereby marked “advertisement” in nine mutants deficient in anion transport activity and show accordance with 18 U.S.C. Section 1734 solelyto indicate this fact. § T o whom correspondence and reprint requests should be addressed Dept. of Pharmacology, University of Alabama at Birmingham, Birmingham, AL 35294. 11 Recipient of a medical student summer research fellowship (National Institutes of Health training grant).

The abbreviations used are: CHO, Chinese hamster ovary; Hepes, 4’-(2’-hydroxyethyl)-l-piperazineethanesulfonic acid; MES, 4-morpholinepropanesulfonic acid; DIDS, 4,4’-diisothiocyanostilbene-2,2’disulfonic acid SITS, 4-acetamido-4-isothiocyanostilbene-2,2’-disulfonic acid.

18607

18608

Anion Exchange Deficient CHO Mutants

that they define mutant alleles which in turn define functional domains involved in high affinity sulfate transport,low affinity sulfate transport, and regulation of cytoplasmic pH. MATERIALS AND METHODS

Cell Culture-Chinese hamster ovary cells (CHO-KI) were obtained from the American Type Culture Collection (CCL-61), Rockville,MD. The mutants described in this paper were isolated by replica plating and 35SO:-colony autoradiography, as described previously (Esko et al., 1985). Allof the mutants were recloned from cultures containing only deficient colonies in order to ensure their purity (Esko, 1986). In general, cells were maintained in Ham's F-12 (Ham, 1965)medium supplementedwith 10% (v/v) fetal bovine serum (Hyclone, Salt Lake City, UT), 100 pgof streptomycin sulfate/ml, and 100 units of penicillin G/ml. They were grown at 37 k 0.2 "C under a 5% COZ atmosphere a t 100% relative humidity, and subcultured every 3-4 days with 0.25% trypsin. Fresh cells were revived after 15-20 passages from stocks stored in liquid nitrogen in order to maintain their purity. Cells were grown to almost complete confluence, in 35-mm dishes, containing 3 ml of medium. Some experiments required the use of specially formulated growth medium. F-12 medium lacking sulfate was prepared from individual components (Ham, 1965), substituting chloride salts for sulfate salts andomitting streptomycin sulfate. Defined mediumwas supplemented with 10% (v/v) fetal bovine serum dialyzed 106-foldagainst phosphate-buffered saline (Dulbecco and Vogt, 1954). Transport Studies-Transport was measured essentially as described (Elgavish et al., 1985; Esko et al., 1986). Prior to the assay, cells were washed three timeswith an isosmotic medium. Uptake was initiated by adding 1ml of medium containing various concentrations of [36S]Na2S04(0.5 p M < [SO:-] < 10 mM). The composition of the medium varied among different experiments and is given in detail in the figure and table legends. Uptake was stopped by rapidly washing the cells six times with 3 ml of cold medium (4"C) containing 4 mM hemimagnesium gluconate, 100 mM sucrose, 100 mM NaN03, and10 Tris, pH 5.5). mM Tris-Hepes, pH 7.5 (or 25 mM MES, 4.6mM Finally, cells were extracted with 1 mlof0.1 NaOH for 30 min at room temperature, and the amount of radioactivity in 0.5-ml sample of extract was measured by liquid scintillation counting. Results were expressed as picomoles of sulfate taken up per milligram of cellular protein. Protein concentrations were determined by the method of Lowry et al. (1951). P'C]BenzoicAcid Uptake-Experiments were carried out as described (L'Allemain et al., 1985).For pHi, measurements, ["Clbenzoic acid was added at 1 pCi/ml. After incubation for various times, the dishes werewashed rapidly four times at 4 "C using the washing solution for sulfate transport experiments described above. Cells were solubilized in 0.1 N NaOH and radioactivity was assayed by liquid scintillation spectrometry. Extracellular Space-To estimate the radioactivity trapped in the extracellular space, the method described by L'Allemain et al. (1985) was used. [14C]Benzoic acid uptake was measured in the following way: the cells were first incubated with 20 mM NH4Cl in glucose/ saline (130 mM NaCl, 5 mM KCl, 2 mM CaClz, 1 mM MgCL, 5 mM glucose, and 20 mM Tris-Hepes, pH 7.5) for 60 min a t 37 "C. The medium was then aspirated, and thecells were quickly washed twice with sodium free saline at pH 8.0 (glucose/saline medium containing 130 mM choline chloride instead of NaCl). Cells were incubated 30 s in this medium and then incubated in the same medium containing [14C]benzoicacid (4"C). The dishes were then immediately washed four times with phosphate-buffered saline. Under these conditions, the intracellular pH is much more acidic than the external pH so that the intracellular accumulation of benzoic acid measured for a very short time (2 s) becomes negligible a t 2 "C. Intracellular Space-The intracellular space was calculated from the equilibrium uptake of ['4C]3-O-methyl-D-glucose.The cells were incubated with 5 pCi/ml of this nonmetabolizable hexose for 60 min at 37 "C at which time steady state had been reached. The intracellular volume accessible to 3-O-methyl-~-glucose was found to be 7.5 pl/mg protein. Calculation of pH,,-The counts measured after ['4C]benzoicacid uptake were corrected for radioactivity trapped in the extracellular space and loss of intracellular radioactivity during the washing procedure, as determined in a preliminary efflux experiment. Then, the intracellular concentration of benzoic acid Bi (cpm/pl) was calculated, assuming that there is 7.5 p1 of intracellular volume/mg of cell protein. The intracellular pH was calculated from: pHi, = pH,, + log (Bi/Bd,

where BO= amount of [I4C]benzoicacid (cpm/pl) in the external medium. Analysis of Data-All transport experiments and assays were performed in duplicate or triplicate samples for each condition and the results are given as themean k standard deviation. The apparent K,,, and V., values obtained in kinetic studies were calculated using a computer program. The program calculates apparent K,,, and Vma, values by linear regression using Lineweaver-Burk, Woolf and Hofstee plots. Residual errors were minimized and were randomly distributed. Kinetic constants are generally the mean k standard deviation of values calculated from several experiments carried out in separate batches of cells. The number of experiments (n)is given in each case. Significant differences were tested using Student's t test. RESULTS

Characteristics of the Anion Transporter in Chinese Hamster Ovary Cells-In the present study we have confirmed our earlier finding (Esko et al., 1986) that in CHO cells sulfate transport occurs via a carrier-mediated system. In addition, we show that thiscarrier-mediated system which we probe by monitoring sulfate transport, resembles the band 3 anion exchanger of the erythroid plasma membrane. We base this contention on the following several lines of evidence: ( a ) substrate specificity, ( b ) specific inhibition, and (c) pH sensitivity. The time course of sulfate uptake into CHO cells is given in Fig. 1. In wild type cells, uptake is linear for about 10 min and then plateaus. In subsequent kinetic studies, initial rates were measured for times not exceeding 1 min. In thesekinetic studies sulfate uptake in Chinese hamster ovary cells was shown to be a saturable process (Fig. 2 A ) with an apparent K , = 2.9 & 0.6 mM and Vmax= 900.0 f 200 pmol/mg protein/ min ( n = 3).' Detailed kinetic studies carried out at sulfate concentrations lower than 150 p~ demonstrated the presence of two additional saturable components of sulfate uptake, one in the range of extracellular sulfate concentrations 50 pM < [so$-], < 150 pM and one in the range 0.5 pM < [so:-] < 50 p~ (Fig. 2B). Under physiological conditions (serum concentration of sulfate approximately 1 mM, Krijgsheld et al., 1980), the low affinity, high capacity saturable component of sulfate uptake is probably the main transport activity functioning (Fig. 2 A ) . The high affinity, low capacity components could represent SO:- binding to cell components. However, they probably constitute an authentic transport activity since CHO cells grown in medium containing low sulfate ( e 6 p M ) incorporate [35S]SOq-into proteoglycans efficiently (Esko et al., 1986). The substrate specificity of the carrier-mediated sulfate transport system was tested in cis-inhibition influx and transstimulation efflux studies. Extracellular thiosulfate, molybdate, and bicarbonate inhibited sulfate influx by about 50%, whereas identical concentrations of phosphate (5mM) had no significant effect (Table I). Sulfateefflux was markedly stimulated by extracellular chloride, whereas similar levels (5 mM) of phosphate or arsenate had no stimulatory effect (Table I). The effect of increasing extracellular chloride concentrations on sulfate influx into the cells is given in Fig. 3. Chloride inhibited sulfate influx in a dose-dependent manner. An increase in the extracellular chloride concentration increased the apparentK , of sulfate uptake very markedly, but had no significant effect on the Vmax(inset, Fig. 3). Thus, chloride appears to be a competitive inhibitor of sulfate uptake into the CHO cells. These results, taken together with the fact that extracellular chloride stimulates SO:- efflux, are consistent with the possibility that sulfate transport in the CHO cells occurs via a SO:-/Cl- exchange mechanism.

* Results are the mean k standard deviation of values obtained in three separate experiments.

18609

Anion Exchange DeficientCHO Mutants

h

/

I

I

. g

10oot

I

/

2

3

4

5

6

7

8

8

1

0

CSO.'-l mM

Time (mid

FIG.1. The time course of sulfate uptake in Chinese hamster ovary cells. Uptake was measured in a medium containing 4 mM hemimagnesium gluconate, 10 mM Tris-Hepes, pH 7.5, 150 mM NaCl, 10 p~ [35S]Na2S04in wild type cells (O),mutant 604 (U),and mutant 605 (A). Derivatives of stilbene disulfonic acid (DIDS, SITS) are potentinhibitors of anion exchange in several membrane systems (Cabantchik, 1978; Elgavish et al., 1985, 1987). We have previously shown that SITS inhibits sulfate transport in CHO cells (Esko et al., 1986). DIDS inhibited sulfate uptake into the Chinese hamster ovary cells in a dose-dependent manner (Fig. 4). The concentration of DIDS necessary for half-maximal inhibition was 0.06 p ~ which , indicates a high affinity of the anion exchanger for this inhibitor in the CHO cell. Studies in erythrocytes (Gunn, 1973; Milanick and Gunn, 1984) have led to the proposal of the titratable anion transporter model. Accordingto thismodel, protons activate sulfate fluxes by converting the anion exchanger from a monovalent 1 to a divalent anion transporter. InCHO cells, like the erythf o 50 100 = roid cells (Milanick and Gunn, 1984), low extracellular pH CSO,z-I IIM m markedly stimulates sulfate transport (Fig. 5 A ) . Sulfate upFIG.2. Kinetic studies of sulfate uptake in the wild type take in the cells was maximal after about 15 min. The maximum extent of sulfate uptake observed in the wild type cells and Chinese hamster ovary cell mutants deficient in sulfate at pH 5.5 was about 10-fold higher than the value measured transport. Afterremoving the growth medium, cells were rinsed times with sulfate-free growthmedium. Uptake assays were at the same time at pH 7.5. Kinetic studies showed that three initiated by adding 1 ml of growth medium supplemented with dilowering the extracellular pH caused a 6-fold decrease in the alyzed fetal bovine serum, [35S]NazS04,and various concentrations apparent K, of the transportsystem for sulfate (Table 11)but of Na2S04; A , results obtained at 0.15 mM C [SO:-],, C 10 mM; B , had no significant effect on Vmax.The decrease in the apparent results obtained at 0.5 pM < [SO:-],,t < 150 pM; inset, enlarged K , of sulfate transport at low extracellular pH suggests that display of results obtained at 0.5 pM C [SO:-], < 40 pM; c, results protons may be binding to the anion exchanger causing an obtained at 10 pM C SO^-],,, < 0.1 mM; 0, wild type; A, 13, 0, A, typical results obtained in a mutant from groups I, 11, 111, and IV, increase in the affinity of the substrate-binding site for sul- respectively. fate. Altered Sulfate Transport in CHO Mutants-In a previous study, we described two CHO mutants deficient in sulfate type and in mutants 604 and 605. Sulfate uptake measuretransport (Esko et al., 1986). Mutant 605 was shown to lack ments carried out for short times, i.e. before any significant the SITS-sensitive SO:-/CI- exchanger characterized in this "SOOg- incorporation can occur in macromolecular fractions report. This deficiency led to a dramatic inhibition of incor- (Esko et al., 1986), enabled us to measure sulfate transport poration of inorganic sulfateinto proteoglycans. Another per se. As shown in Fig. 1, the initial rate of sulfate transport mutant, strain 604, was shown to belong to the same genetic was 3-fold lower in mutant 604 and €$-fold lower in mutant complementation group as mutant 605 by cell hybridization. 605 than in the wild type. The defect was specific to the In contrast to mutant605, mutant 604 was only about 3-fold sulfate carrier, rather than a nonspecific defect in the memdefective in theincorporation of inorganic sulfate into proteo- brane as evident from the fact that L-cysteine transport was glycans. Using the same autoradiographic methods that led not affected in the mutants(Fig. 6). Kinetic studies were carried out in the wild type and nine to the isolation of mutants 604 and 605, we have since obdefective in thesulfate carrier. mutants and intriguing tained 36 other strains that are differences between the mutantswere The sulfate transport defects characterized in nine of these observed (Table 111, Fig. 2, A and C ) . In mutants belonging mutants are summarized in Table 111. to group I, the capacity of the low affinity, high capacity The time course of sulfate uptake was followed in the wild sulfate transport system was comparable to that in the wild

Anion Exchange Deficient CHO Mutants

18610

TABLEI The effect of various anions on the initial rate of sulfate influx and efflux For uptake, cells were incubated for 1min in a medium containing 0.1 mM [35S]Na2S04, 300 mM sucrose, 10 mM Tris-Hepes, pH 7.5, in the presence or absence of 5 mMof the various anions. For efflux measurements, cells were first incubated for 1h at 37 “C in a medium containing 0.1 mM [35S]Na2S04,1 mM CaC12, 5 mM KCl, 1 mM NaHtP04, 1 mMMgC12, 26 mM NaHC03, 5 mM D-glucose, 116 mM NaCl, pH 7.5. The medium was then removed and the amount of counts/minute accumulated in the cells (cpmo)was measured. Similarly labeled cells were washed and resuspended for 5 min in amedium containing 300 mM sucrose, 0.1 mM Na2S04,10 mM Tris-Hepes, pH 7.5, with or without 5 mM of various anions. Results are expressed as a percent of the [36S]sulfatecpm that have left the cells at time t (cpmo-cpmt),of the initial amount of counts/minute present in the cells when efflux started (cpmo). All results are the mean f standard deviation of triplicate samples. anion Added

Sulfate influx

5 mM

Sulfate efflux

pmol/mgproteinlmin[fcpmo-cpmJ/cpmJ

None Na2S203 Na2Mo04 NaHC03 NaH2P04 NaHAsOl NaCl

245.8 k63.4 24.0 149.3 +- 22.1 130.7 f 28.4 136.7 f 16.3 55.1 3.6 228.1 f

I50

0.01 0.1

1

10

100 1000

I

log CDIDSI pM

FIG. 4. DIDS inhibits the initial rate of sulfate uptake in a dose-dependent manner. Uptake (3 “C)was measured for 30 s in a medium containing4 mM hemimagnesium gluconate, 150 mM sodium gluconate, 10 PM [35S]Na2S04,and 10 mM Tris-Hepes, pH 7.5.

95.5

I

1147.1

f

146.9

0.5

1125.7

f

120.5

!

0

I

D‘ote,”?.,.)

1.3 f 0.2

3.2

I

f 0.9 53.0 f 2.9 94.2 f 2.0

1149.9

f

100

w

f 2.6

,0mo:er/vg

0.4

X

h

0.05

-

6

0

10

20

30 0

10

20

30 0

10

20

30

Time (mid

I 0

J 150 50

100

ca-I m~ FIG. 3. Extracellular chloride inhibits the initialrate of sulfate uptake in a dose-dependent manner. Uptake was measured for 1 min in 1 ml of medium containing 4 mM hemimagnesium gluconate, 10 PM [35S]NazS04,and various ratios of NaCl/sodium gluconate (up toa total of 150 mM), 10 mM Tris-Hepes, pH 7.5. Inset, sulfate uptake was measured for up to 1 min in an identical medium but containing various concentrations of Na2S04. Apparent K , and Vmaxvalues were calculated as described under “Materials and Methods.” Significant differences (*) were tested using Student’s t test ( p c 0.01).

FIG. 5. Low extracellular pH stimulates sulfate uptake in the wild type and mutant 604 but not in mutant 606. Uptake was measured in a medium containing 4 mM hemimagnesium gluconate, 150 mM NaCl, 10 p M [35S]Na2S04,and either 10 mM TrisHepes, pH 7.5 (A),or 25 mM MES, 4.6 mM Tris, pH 5.5 (0).A , wild type; B , mutant 604; C, mutant 605.

TABLEI1 The effect of the extracellular pH on sulfate uptake Sulfate uptake was measured for up to 1min in a medium containing [36S]Na2S04, various concentrations of Na2S04,150 mM NaCl, 4 mM hemimagnesium gluconate, and either 10 mM Tris-Hepes, pH 7.5, or 25 mM MES, 4.6 mM Tris, pH 5.5. Apparent K,,, and Vmax values were calculated as described under “Materials and Methods.” Results are given as mean f standard deviation of values obtained in two to three separate experiments. Significant differences (*) were tested usine Student’s t test (P c 0.01). DH Auuarent K, V,.

type. However, the affinity of the system for sulfate was mhf proteinlmin pmollmg markedly lower. In mutants belonging to group 11, both the 876.0 f 210.5 2.97.5 f 0.60 affinity of this transport system for sulfate and its capacity 812.2 f 146.5 0.55.5 f 0.04, were lower than in the wild type. When the high affinity, low capacity transport system operating at 50 p M C [SO~-],,t C 150 p~ was probed in mutants belonging to groups I and 11, 2, A and C) was not a saturable function of the extracellular both the affinity and thecapacity of this transportcomponent sulfate concentration over the entire range of concentrations were found to be markedly lower than in the wild type (Table tested, and as described previously (Esko et al., 1986), this 111; Fig. 2C). The capacity of this high affinity transport mutant lacks carrier activity entirely. Low extracellular pH stimulates sulfate uptake in mutant system in mutants belonging to group I and I1 was also found to be lower than that in mutantsbelonging to group I11 (Fig. 604 (Fig. 5B) but the extentof the stimulation is lower than 2 C ) . Mutants belonging to group I11 displayed uptake identi- in the wild type suggesting that a proton-binding sitemay be cal to thewild type at 0.5 p~ c [SO:-],ut c 50 p ~ but , uptake altered in this mutant. No stimulation bylow extracellular was lower than that in the wild type at higher extracellular pH can be demonstrated in mutant 605 (Fig. 5C), but this is lacks carrier SO:- concentrations (Fig. 2C). The initialrate of sulfate notsurprising considering thatthismutant uptake in the mutant belonging to group IV (Table 111, Fig. activity altogether (Fig. 2, A and C).

18611

Anion Exchange Deficient CHO Mutants TABLE I11 Mutants of Chinese hamster ovary cells deficient in sulfate transport The symbols used are: +, not altered; -, altered. Low affinity, high capacity sulfate transport system

High affinity, low capacity sulfate transport system Strain A”

c“

B”

Capacity Affinity Capacity Affinity

+ +

+

Wild type CHO-K1 Mutants Group I (813,632,668) Group I1 (647, 613,604) Group I11 (625,679) inactive Group IV (605) a A, 0.5 pM < [SO:-],t < 50 pM; B, 50 pM < [so!-]w ND, not determined.

+

Affinity

Capacity

+

+

+

+

-

-

-

+-

-

NDb inactive

inactive

150 pM c , 0.15 mM < [so!-]mt < 10 mM.

. m A

b

5

10 0

5

10

1

Time (mid

0

10

20

30

40

50

60

Time (mid

.,

FIG.6. L-Cysteine uptake is not affected in the sulfate transport-deficient mutants of Chinese hamster ovary cells. Uptake was measured in a medium containing 4mM hemimagnesium gluconate, 150 mM NaC1, 10 mM Tris-Hepes, pH 7.5, and 10 p M I%] L-cysteine. a, wild type; A, mutant 604; mutant 605.

Mutant 605 (group IV) has been very useful in demonstrating that the anion exchanger present in CHO cells may be involved in pH regulation. In the experiment depicted in Fig. 7, wild type and mutant 605 (group IV) CHO cells were made acidotic using a method previously described by L’Allemain et al. (1985). Cells were first preincubated for 60 min in a medium containing 20 mM NH4C1.For pHinestimates, uptake of [14C]benzoic acid was then followed in an identical NHZfree medium. Under these conditions, the rapid diffusion of the uncharged species NH3 out of the cells preincubated with NH4C1 would beexpected to cause intracellular acidification (L’Allemain et al., 1985). In the presence of extracellular Na+, the acidification occurred, as expected, followedby almost complete recovery of the intracellular pH with time, both in the wild type and in mutant 605. However, in the absence of extracellular Na+, the wild type was able to recover from acidosis, whereas mutant 605, which lacks anion exchange activity, lagged very markedly in its recovery. These results suggest that in CHO cells, the anion exchanger may play a role in intracellular pH regulation. DISCUSSION

Our studies in wild type CHO cells indicate the presence of an anion exchanger (SOq-/Cl-) in the plasma membrane of

FIG. 7. Recovery from an acid load in wild typeand mutant (606) Chinese hamster ovary cells. CHO cells were preincubated with a glucose/saline buffer (130 mM NaC1, 5 mM KCl, 2 mM CaC12, 1 mM MgC12, 5 r n glucose, ~ 10 mM Tris-Hepes, pH 7.5) containing 20mM NH4Cl and 1 mM NazS04for 30 min at 37 “C. At time 0, extracellular NH: was eliminated, and cultures were incubated for various times with [“Clbenzoic acid in a medium containing 150 mM NaCl ( A )or 150 mM N-methyl glucamine C1 ( B ) ,10 mM Tris-Hepes, pH 7.5, 1 mM Na2S04, 10 mM NaHC03, 4 mM hemimagnesium gluconate. Intracellular pH, was calculated as described under “Materials and Methods” and is given as a mean f standard deviation of triplicate samples for the wild type (0)and mutant 605 (a).For determination ofpHi, a t time 0, before NH: removal, [“Clbenzoic acid was added to thepreincubation medium for the last 1 min.

these cells. By criteria of substrate specificity, specific inhibition by DIDS andSITS andstimulation at low extracellular pH, the anion exchanger in CHO cells appears similar to the well-described band-3 anion exchanger in erythrocytes (Cabantchik et al., 1978; Knauf, 1979; Passow, 1986).Anion exchangers with similar properties have been previously demonstrated in non-erythroidcells such as Ehrlich ascites tumor cells (Levinson and Villereal, 1974; Hoffman, 1982), lung fibroblasts (Elgavish et al., 1985; Jiang et al., 1986) and skin fibroblasts (Elgavish and Meezan, 1988).More recently, anion exchange activity was measured in epithelial plasma membranes such as renal brush border membranes (Pritchard, 1987) and tracheal apical membranes (Elgavish et al., 1987). Structural evidence that erythroid and non-erythroidplasma membranes contain similar anion exchangers has emerged from studies in which polypeptides immunologically related to the band-3 anion exchanger were found in several nonerythroid cell systems (Cox et al., 1980; Kay et al., 1983). A distinguishing feature of the anion exchanger in theCHO cell is the presence of high affinity components (Fig. 2B). The erythroid anionexchanger has not been probed at extra, may cellular sulfate concentrations lower than 100 p ~which

18612

Anion Exchange DeficientCHO Mutants

be the reason for the fact that highaffinity, low capacity cells have the ability to obtain all the inorganic sulfate necsystems have not been previously reported in erythrocytes essary for glycosaminoglycan sulfation by catabolizing L-cys(Knauf, 1979; Passow, 1986). This is not surprising, since the teine (Eskoet al., 1986). Under the conditions inwhich those concentration of inorganic sulfate in the human serum is studies were carried out, ([SOf],t < 6 PM) sulfate transport normally about 300 ktM (Pillion et al., 1986). However, serum via the anion exchangerdoes not limit glycosaminoglycan levels have been shown to dropby as much as &fold when an sulfation. animal is maintained ona low protein diet (Krijgsheld et al., A CHO mutant which lacks anion exchange activity alto1982). Thus, the high affinity, low capacity components of gether (mutant 605), enabled us to reveal a possible function cells. Since HCO; is the anion exchange system may be useful under these condi- the anion exchanger has in the CHO tions. Several lines of evidence indicate that thehigh affinity, known to be a substrate for the anion exchanger in erythrolow capacity components of the sulfate uptake system repre- cytes (Knauf, 1979; Olsnes et al., 1987; Passow, 1986) and sent transport into the cells. In previous studies (Esko et al., non-erythroid cell systems(Olsnes et al., 1987), andour 1986), we have demonstrated that["SJSOP was incorporated results suggested that it is also a substrate for the exchanger efficiently into cellular proteoglycans at an extracellular sul- in CHO cells (Table I),we postulated that the anion exchanfate concentration of 6 k ~ a ,finding which suggested that ger in the CHOcells may play a role in the regulation of the intracellularpH. Totestthishypothesis, recovery from SO:- was taken up under these conditions. In the present study, we showed that sulfate uptake at low extracellular NH4C1--induced acidosis was followed in thewild type and in SO:- concentration (10 PM) displayed features characteristic mutant 605 (Fig. 7). In the presence of Na+, both the wild to the carrier-mediated anion exchanger (Figs. 3 and 5). Thus, type and mutant 605 recovered from acidosiswith similar sulfate uptake at SO^-],,, = 10 pM was inhibited in a dose- kinetics. We assumed that this recovery was made possible dependent manner by chloride (Fig. 3), and it was stimulated by the presence of a Na+/H+ exchanger which is uniquitous in plasma membranes of mammalian cells and which, we severalfold at low extracellular pH (Fig. 5). The anion exassumed, is mostprobably not altered in the mutant. Our change-deficient CHO mutants were very useful in demonassumption appears to be justified by results in Fig. 7. In the strating that the three saturable components of the anion which exchange system, the two high affinity, low capacity systems absence of extracellularNa+,anextremecondition mimics an altered Na+/H+ exchanger (L'Allemain et al., 1985), (Fig. 2, B and C) and the low affinity, high capacity system the recovery from acidosis of the mutant lacking the anion (Fig. 2 A ) were independent functional activities, but appeared exchanger was very markedly delayed as compared to the wild to depend on the gene product defined by the mutant. We type. These results suggest that the anion exchanger may play cannot exclude the possibility that the gene product defined cells. Such a role by the mutants formscomplexes with itself or other proteins a regulatory role in pH homeostasis in CHO for an anion exchanger has also been suggested from studies which manifest three kineticmodes. Thefunction of the high affinityandthe low affinity of Vero cells (Olsnes et al., 1987) and a CHO cell mutant systems appeared to be independent of each other. Whereas deficient in Na+/H+ antiport(L'Allemain et at., 1985). both the affinity and the capacity of the high affinity transport in mutantsbelonging to groupI (Table 111, Fig. 2B) were low, only the affinity of the low affinity transport was altered in these mutants (Fig. 2 A ) . In mutants belonging to group 111, the high affinity transport system operating a t 0.5 I . ~ M< [SOf],,, < 50 PM displayed wild type phenotype, but the high at 50 pM < < affinity transport system operating 150 PM was clearly deficient (Table 111, Fig. 2C). Thus, the high and low affinity transport systems appear to be independent functionally; they may be properties of the same membrane protein as suggested by the fact that all of the mutants, regardless of kinetic differences were members of the same genetic complementation group (Esko et al., 1986). Thus, the various mutants define different mutant alleles of the same gene. Saturable, high affinity andlow affinity components appear to be well-conserved properties of anion transporters since theyare found in cells which synthesize relativelylarge amounts of extracellularmatrix,suchaslungfibroblasts (Elgavish and (Elgavish et al., 1985) andskinfibroblasts Meezan, 1988), and cells, like CHO, (Figs. 2, A and B ) which do not synthesize large amounts of matrix (Eskoet al., 1986). In cell systems which synthesize large amounts of sulfated molecules, it is possible that under conditionsof sulfate deprivation or when the low affinity high capacity transport activity is altered, sulfate transport at the plasma membrane, which can now proceed only via the low capacity transport system, becomes a limiting factor for sulfation. This may in part explain why aortic endothelial cells (Humphries et al., 1986), EHS tumor cells (Tyree et al., 1986), and epiphyseal cartilage in culture (It0 et al., 1982) synthesize undersulfated glycosaminoglycans when the extracellular SO?- concentration drops below 100 PM. However, we have shown that CHO

Acknowledgments-The excellent technical assistance of Sandra Vatani and Julie Weinke is gratefully acknowledged. We would like to thank Diane Lee for typing this manuscript. REFERENCES Cabantchik, Z. I. (1983) in Structure and Function of Membrane Proteins (Quagliarello, E., and Palmieri, F.,eds), pp. 271-283, Elsevier Scientific Publishing Co., New York Cabantchik, Z. I., Knauf, P. A., and Rothstein, A. (1978) Biochim. Biophys. Acta 515, 239-302 Cox, J . V., Moon, R. T., and Lazarides, E. (1980) J. Cell Biol. 100, 1548-1557 Dulbecco, R., and Vogt, M. (1954) J. Exp. Med. 99, 167-182 Elgavish, A., Dibona, D. R., Norton, P., and Meezan, E. (1987) Am. J. Physiol. 253, C416-C425 Elgavish, A., Smith, J . B., Pillion, D. J., and Meezan, E. (1985) J . Cell. Physiol. 1 2 5 , 243-250 Elgavish, A,, and Meezan, E. (1988) Biochem. Biophys.Res. Commun. 1 5 2 , 99-106 Esko, J. D. (1986) Methods Enzymol. 129,237-253 Esko, J. D., Elgavish, A., Prasthofer, T., Taylor, W. H., and Weinke, J. L. (1986) J. Biol. Chem. 261,15725-15733 Esko, J. D., Stewart, T. E., and Taylor, W. H. (1985) Proc. Natl. Acad. Sci. U. S. A. 82,3197-3201 Gunn, R. B. (1973) in Erythrocytes, Thrombocytes, Leucocytes (Gerlach, E., Moser, K., Deutsch, E., and Wilmans, W., eds) pp. 77-79 Thieme, Stuttgart Ham, R. G. (1965) Proc. Natl. Acad. Sci. U. S. A. 53,288-293 Hoffman, E. K. (1982) Phil. Trans. R. SOC.Lond. 299,519-535 Humphries, D. E., Silbert, C. K., and Silbert, J. E. (1986) J . Biol. Chem. 261,9122-9127 Ito, K., Kimata, K., Sobue, M., and Suzuki, S. (1982) J. Biol. Chem. 257,917-923 Jiang, X., Meezan, E., and Elgavish, A. (1986) Fed. Proc. 45, 1006 Kay, M.M.B., Tracey, C. M., Goodman, J. R., Cone, J. C., and Bassel, P. S. (1983) Proc. Natl. Acad. Sci. U. S. A. 80,6882-6886 Knauf, P. A. (1979) in Current Topics Membr. Transport 12, 251363

Anion Exchange Deficient CHO Mutants Knickelbein, R. G., Aronson, P. S., and Dobbins, J. W. (1985) J. Membr. Biol. 88, 199-204 Kopito, R. R., and Lodish, H. F. (1985) J. Cell. Biochern. 29, 1-17 Krijgsheld, K.R., Scholtens, E., and Mulder, G. J. (1982) Biochem. Pharmacol. 31, 3997-4000 Krijgsheld, K. R., Scholtens, E., and Mulder, G. J. (1980) Camp Biochem. Physiol. 67,683-686 (abstr.) L'A1lemain? G ' ? Paris, '., and Pouyssewr7 J' (1985) J' Bioi. 260,4877-4883 Langridge-Smith, J. E., Sellin, J. H., and Field, M. (1983) J. Membr. Biol. 72, 131-139 Levinson, c.9 and Villereal, M. L. (1974) J . Cell. Phwiol. 85, 1-14 Lowry, 0. H., Rosebrough, N. J., Farr, A.L., andRandall, R. J. (1951) J . Biol. Chem. 193, 265-275 Lucke, H., Stange, G., and Murer, H.(1981) Gastroenterology 80,2230

18613

Milanick, M.A., and Gunn, R. B. (1984)Am. J. Physiol. 247, C247C259 Olsnes, S., Ludt, J., Tonnessen, T. I., and Sandvig, K. (1987) J. Cell. PhYS. 132, 192-202 Passow, H. (1986) Reu. Physiol. Biochem. Pharrnacol. 103,62-186 Pillion, D. J., Neumeier, T. T., and Meezan, E. (1986) C h . Chim. Acta 142,241-247 Pritchard, J. B. (1987) Am. J. Physiol. 252, F346-F356 Schneider, E. G., Durham, J , C., andSacktor, B. (1984) J , Bioi, Chern, 259,14591-14599 Schron, C. M,, ~ ~ i ~ k ~ R. lG., b A~ i ~ ,~p. s., ~ puca, ~ J., ~ and Dobbins, J. W. (1985) Am. J. Physiol. 249, (36074613 Tonnessen, T. I., Ludt, J., Sandvig, K., and Olsnes, S. (1987) J . Cell. physiol, 132, 183-191 Tyree, B., Hassell, J. R., and Hascall, V. C. (1986) Arch. Biochem. Biophys. 250,202-210 Von Dippe, P., and Levy, D. (1982) J. Biol. Chem. 257,4381-4385

~