BIOCHEMICAL
Vol. 171, No. 3, 1990 September 28, 1990
AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages 920-925
Purification of a stilbene sensitive chloride channel and reconstitution of chloride conductivity into phospholipid vesicles Harry C. Blair”
and Paul H. Schlesinger2
‘Department of Pathology, Jewish Hospital of St. Louis and 2Department of Biomedical Sciences, School of Dental Medicine; Washington University Medical Center, St. Louis, Missouri 63110 Received
August
7, 1990
A protein conferring passive chloride permeability was isolated from a N-octylglucoside solubilized extract of partially purified H+-transporting osteoclast cell membranes. Purification was achieved by binding of solubilized protein to an arninelinked 4,4’-diisothiocyanatostilbene-2,2’-disulfonate @IDS) Sepharose 4B column and elution with 50 mM KCl. A major protein, with MR = 60 kD on 10% SDS-PAGE, was obtained, which was further purified to homogeneity by HPLC gel filtration. This protein introduced 36C1-permeability when reconstituted in phospholipid membranes by equilibrium dialysis. The Cl- transport recovered in reconstituted membranes retained sensitivity to DIDS confirming the identity of the isolated protein as a stilbene-sensitive 01990Academic Press, Inc. chloride channel.
Cation concentrations (eg. K+ , Na+ , Ca+ + , and H+) across cell membranes are often held far from equilibrium by active transport. Conversely, the extracellular anion present in largest concentrations, Cl-, is generally distributed near its GibbsDonnan equilibrium concentration via a number of passive anion exchangers and channels (1). However, anion conductivity, while generally passive, may be a limiting process in important circumstances including the disease cystic fibrosis (2), in vacuolar acidification (3), renal acid secretion (4), and bone resorption (5). This suggests that the proteins mediating Cl- transport may be of considerable interest. We have recently determined that the massive proton flux used by the osteoclast to dissolve bone is provided by an ATP-dependent proton pump (5). Furthermore, we found that this H+ flux is dependent on [Cl-], and that 10 pM DIDS inhibits the process in a manner that is competitively reversed by increasing [ATP] (6). Therefore, we reasoned that the competitive DIDS binding could be exploited to isolate a Clchannel from partially purified osteoclast membranes. We herein describe a method by which the Cl- transporting membrane protein of osteoclasts was isolated with retention of its DIDS sensitive Cl- channel activity. *To whom reprint
requests
should
be addressed.
Abbreviations used: DIDS-4,4’-diisothiocyanatostilbene-2,2’disulfonate. 0006-291X&O $1.50 Copyright 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.
920
Vol.
171,
No.
3, 1990
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
Methods Membrane Preparation: Osteoclasts, isolated from long bones of calcium deprived laying hens as described (7), were fragmented via nitrogen cavitation in 250 mM sucrose, 1 mM EGTA, 1 mM D’IT, 10 mM Tris, pH = 7.0 (lysis buffer). Cells (2 x ld7) in 20 mls of lysis buffer were pressurized to 40 atm N,, at 4’C, in a stainless steel bomb, and allowed to equilibrate for 30 minutes, followed by decompression over 2 seconds through a 5 mm orifice. Membranes were fractionated by differential centrifugation, retaining the supernatants after 1000 x g for 5 minutes and then 4,700 x g for 10 minutes. The membranes were collected by centrifugation at 48,000 x g for 40 minutes after the method of Gluck and Al Awqati (8). The pellet (2.4 mg of protein) was primarily osteoclast ruffled membranes. Affinity substrate: Dry CNBr activated Sepharose 4B (Sigma Chenical Co. St. Louis, MO) was swollen in 50 mM carbonate, pH 9, and reacted with 1,5diaminopentane, 10 mM, for 36 h at 4’C. The beads were washed several times with water and suspended in 50 mM carbonate, pH 9, with 3 mM DIDS at 4’C for 48 h. After packing in a disposable glass column the DIDS-Sepharose was washed with > 50 volumes of 50 mM Na-cyclamate, 10 mM HEPES @H 7.0), 1.4% (w/v) Noctylglucoside (column buffer) at 4’C. Membrane solubiliition: Membrane fractions in lysis buffer were dialyzed 24 h at 4’C in 12 - 14 kD cutoff cellulose membranes against 1000 volumes of the column buffer without N-octylglucoside, to which 0.5 mls of the acetate form of Dowex 1 had been added. This was repeated 3 times after which the dialyzed membranes were solubilized by the addition of 1.4 % (w/v) N-octylglucoside. DIDS-Sepharose chromatography of solubilized osteoclast membranes: 2.4 mg of solubili:zed membrane protein (in 1.0 ml) was passed over a 3 ml DIDS-Sepharose column equilibrated with the column buffer. The column was washed with a further 20 volume:s of the same buffer until A,,, had returned to a stable baseline, and then eluted with 50 mM KC1 + 10 mM HEPES @H 7.0) + 1.4% (w/v) N-octylglucoside. Pooled fractions were frozen at -80°C until used. Gel filtration of the fractions bound by DIDS-Sepharose: A Bio Sil Set 125 column (300 x ‘7.8 mm, BioRad, Richman CA) was equilibrated with 50 mM KCl, 10 mM HEPES (pH 7.0) and 1.4 % N-octylglucoside. The pooled DIDS-Sepharose peak was added in 0.5 ml of the same buffer and pumped through the column at 1 ml/minute. The column effluent was monitored at 280 nm and fractions of 800 ~1 were collected. The A,,, peaks were pooled separately and stored at -8O’C until used. Membrane reconstitution: Purified protein fractions were reconstituted into phospholipid vesicles by mixing 200 ~1 (l-3 pg of protein) of the pooled column fractions with 10 mg asolectin (Sigma Type 11s)and 9 mg of N-octylglucoside in 0.4 ml total volume. This mixture was dialized in a 12-14 kD cutoff cellulose dialysis membrane against 3 changes of 500 mls of 100 mM KCl, 100 mM sucrose, 10 mM HEPES, pH 7.0 (dialysis buffer) over a period of 18 h after the method of Kagawa and Racker (9) and Landry et al (10). 36Cl- Uptake: Reconstituted vesicles (30 ~1 and 0.05 - 0.15 pg of protein) were added to 70 ~1 of buffer (100 mM KC1 + 10 mM HEPES, pH 7.0) containing 0.05 &‘i of s6C1-, with or with out DIDS or valinomycin at the indicated concentrations. The reaction mixture was incubated at 25’C for 60 s and then run over a 1 ml Dowex 1 (acetate form) column which had been pre-washed with 1 ml of 25 mg/ml of BSA in 921
Vol. 171, No. 3, 1990
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
200 mM sucrose + 10 mM HEPES (pH 7.0). Immediately the column was further washed with 1 ml of 200 mM sucrose + 10 mM HEPES (pH 7.0) and the eluant pooled and counted for 36C1- (11). Concentrations and statistical methods: Protein concentrations were determined as described (12). Where a range is given it indicates mean f standard deviation.
Results From the nitrogen cavitation lysis of 2 x lo7 osteoclasts we obtained 250 mg of protein, after differential centrifugation the cell membranes contained 2.4 mg protein. These membranes consisted largely of ruffled membranes from the osteoclast bone attachment site. When these membranes were solubilized in 1.4 % N-octylglucoside and passed over an amine-linked DIDS-Sepharose affinity column 0.07&0.03% (n=3) of the protein was bound and eluted by 50 mM KC1 (Figure 1~). The chloride eluant contained one major protein and several minor bands by SDS-PAGE (Figure 1~). This mixture was resolved into Peak I, of approximately 60 kD, and Peak II, a low molecular weight “salt” peak, (Figure 2A) by HPLC gel filtration. SDS-PAGE of these two peaks on 10% gels revealed that Peak I contained one protein of approximately 60 kD (Figure 2B) and that Peak II contained several low molecular weight species (data not shown). Chloride transport activity was monitored by reconstitution of fractions into phospholipid vesicles using equilibrium dialysis as described. The vesicles formed by this procedure are shown in Figure 3. After the DIDS-Sepharose column we observed 36C1- uptake by reconstituted vesicles incorporating the KC1 eluted protein. The large unbound protein peak from the DIDS column was inactive (Table I). The HPLC gel filtration yielded a single major protein peak, Peak I, and a “salt” peak, Peak II. After reconstitution into lipid vesicles Peak I displayed 36C1- uptake greater than that of the active fractions from DIDS-Sepharose column (Table I). A.
B. FT
Cl-
-30 0
20
40
50
Fraction
50 -
dye
Figure 1. A] Protein elution from DIDS-Sepharose. Protein in eluant fractions was monitored at 280 nm. 2.4 mg of osteoclast membrane protein in 50 mM Nacyclamate, 10 mM HEPES (pH=7.0) and 1.4% N-octylglucoside was placed on a column equilibrated with the same buffer. The column was washed until the A2so returned to the baseline. Then 50 mM KCl, 10 mM HEPES (pH 7.0) and 1.4% N-octylglucoside was used to further wash the column. A small peak that eluted at fractions 25-28 was pooled and stored at -8O’C. B] Polyacrylamide gel electrophoresis of the flow through (FT) and KCl-eluted peak (Cl-) under reducing conditions. The position of molecular weight markers and the dye front are indicated.
922
Vol.
171, No. 3, 1990
BIOCHEMICAL
A.
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
B. GF PeakI
Peak II
0.2
02
k 0
3
Fraction 9 9
12
-30 IS -
-dye
03
mtre 2. A] Elution profile on a GFE gel filtration column run in 50 mM KC1 + 10 mM HEPES (pH=7.0) and 1.4% N-octylglucoside. 20-30 pg of protein from the KC1 eluted peak of the DIDS-Sepharose column was placed on this gel filtration column and the eluant monitored at 280 nm. The first peak (Peak I) was pooled and stored at -80°C until used. The second peak (Peak II) was at the salt elution front of this column and contained no detectable high molecular weight proteins. B] Polyacrylamide gel electrophoresis in SDS of Peak I from the gel filtration. Approximately 1 pg of protein from Peak I was placed on the gel after reduction. The position of molecular weight markers and the dye front are indicated. Figure 3. Electromicroscopy of reconstituted vesicles. Vesicles reconstituted with Peak I were produced as described in methods, fixed with 0~01, pelleted by centrifugation, and processed for electron microscopy (magnification 31, ZOO).
DIDS (200 PM) inhibited uptake of 36C1- by vesicles reconstituted with both the DIDS-Sepharose peak and Peak I from the HPLC gel filtration (Table I). The addition of 20 PM valinomycin slightly increased uptake by vesicles reconstituted with protein from both active fractions.
Discussion It is established that anion conductivities are important in a number of physiological processes, including vacuolar acidification (l-4). However, identification of a specific Cl- conducting protein has proved elusive. In recent reports (8,13), indanyloxacetic acid derivatives were used as affinity substrates to purify 4 Clconducting proteins, but reconstitution of channels with DIDS sensitivity has not been reported. In the present paper, we report a novel application of DIDS competitive inhibition of chloride transport to isolate a homogenous Cl- conducting protein. The resulting 60 kD protein (Figure 2) demonstrates 36Cl- conducting activity, which when reconstituted into lipid vesicles, retains stilbene sensitivity (Table II). The sensitivity to DIDS is similar to that in osteoclast membranes where 200 PM DIDS completely inhibited the Cl- dependent acidification (6). These results suggest that a single protein mediates osteoclast H+-ATPase associated chloride transport, although with these data 923
BIOCHEMICAL
Vol. 171, No. 3, 1990
AND BIOPHYSICAL
TABLE
RESEARCH COMMUNICATIONS
I
36ClATransport by ReconstitutedVesicles’
Preparation
Lipid Vesicles2 DIDS-Sepharose DIDS-Sepharose PassThrough
Control
+ 2OpM 36C1--Uptake Valinomycin3
cpm
cpm
cpm
42+15 n=2
N.D.
41+17 n=2
2878 n=l
1125-l
n=3 13+71
N.D.
6+1
2163+1015
n=2 HPLC Peak I
+2ooa DIDS4
n=3
n=2
3598&1709
n=3
4638+2085
n=3
750+ 147
n=3
‘Uptakevaluesare asmeancpm k standarddeviationandn is the numberof preparations in whicheachvalue wasdetermined.All uptakevaluesarenormalizedto 1 Pg of reconstituted protein. 2Lipidvesicleswerepreparedby substituting200~1of dialysisbuffer for the protein containingsamplefractionsusedelsewhere. 3Valinomycin(20 PM final concentration)wasincludedin the reactionmixture. Buffer compositionandtimeof incubationwasasdescribed in Methods.
4DIDS (200 PM final concentration) wasincluded in the reaction mixture. Buffer composition and time of incubation wasas describedin Methods. it is not possible to assesswhether multiple identical subunits are involved in transport by reconstituted vesicles.
The uptake of 36C1- did not require a vesicle membrane potential since the ionic composition was the same inside and outside the vesicles. In addition, 20 PM valinomycin slightly increased uptake under conditions where it would be expected to neutralize any membrane potential (I$=&= 100 mM). Uptake per pg of purified protein increased after HPLC gel chromatography consistent with the observed removal several contaminating proteins. It is possible that the 60 kD protein reported here is analogous to the 64 kD protein previously isolated from bovine kidney and trachea (8,13). They have similar molecular weights and the DllX insensitivity of the 64 kD bovine protein may reflect an artifact of the isolation technique used. However, there are most likely several distinct Cl- channels (l), and the two proteins, though close in molecular weight, may be structurally dissimilar and subserve different functions. 924
Vol.
171, No. 3, 1990
BIOCHEMICAL
AND BIOPHYSICAL RESEARCH COMMUNICATIONS
References 1. Franciolini, F., and Petris, A. (1990) B&him, Biolhys. Acta 1031, 247-259. 2. Li, M. McCann, J.D. Liedtke, C,M, Naim, A.C. Greengard, P. Welsh, M.J. (1988) Nature 331, 358-360. 3. Fuchs, R., Male, P. and Mellman, I. (1989) J. Biol. Chem. 264, 2212-2220. 4. Schmid, A., Burckhardt, G., and Gogelein, H. (1989) J. Membr. Biol. 111, 265275. 5. Blair, H.C., Teitelbaum, S.L., Ghiselli, R. and Gluck, S. (1989) Science 245, 855-857. 6. Manuscript submitted. 7. Blair, H.C., A.J. Kahn, E.C. Crouch, J.J. Jeffrey, and Teitelbaum, S.L. (1986) J. Cell Biol. 102, 1164-1172. 8. Gluck, S. and Al-Awqati, Q. (1984) J. Clin. Invest. 73, 1704-1710. 9. Kagawa, Y. and Racker, E. (1971) J. Biol. Chem. 246, 5477-5487. 10. Landry, D. W., Akabas, M.H., Readhead, C., Edelman, A., Cragoe, E.J. and AlAwqati, Q. (1989) Science 244, 1469-1472. 11. Breuer, W. (1989) J. Membr. Biol. 107, 35-42. 12. Groves, W.E., Davis, F.C. and Sells, B.H. (1968) Anal. Bioch. 22, 195-210. 13. Akabas, M., Edelman, A., Redhead, C., Land, D., and Al-Awqati, Q. (1990) Biophysical J. 57, 320a. 14. Blair, H., Koziol, C., Mead, R., Gluck, S., Teitelbaum, S., and Schlesinger, P. (1989) J. Cell Biol. 109, 137a.
925
Vol. 171, No. 3, 1990 September 28, 1990
AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages 920-925
Purification of a stilbene sensitive chloride channel and reconstitution of chloride conductivity into phospholipid vesicles Harry C. Blair”
and Paul H. Schlesinger2
‘Department of Pathology, Jewish Hospital of St. Louis and 2Department of Biomedical Sciences, School of Dental Medicine; Washington University Medical Center, St. Louis, Missouri 63110 Received
August
7, 1990
A protein conferring passive chloride permeability was isolated from a N-octylglucoside solubilized extract of partially purified H+-transporting osteoclast cell membranes. Purification was achieved by binding of solubilized protein to an arninelinked 4,4’-diisothiocyanatostilbene-2,2’-disulfonate @IDS) Sepharose 4B column and elution with 50 mM KCl. A major protein, with MR = 60 kD on 10% SDS-PAGE, was obtained, which was further purified to homogeneity by HPLC gel filtration. This protein introduced 36C1-permeability when reconstituted in phospholipid membranes by equilibrium dialysis. The Cl- transport recovered in reconstituted membranes retained sensitivity to DIDS confirming the identity of the isolated protein as a stilbene-sensitive 01990Academic Press, Inc. chloride channel.
Cation concentrations (eg. K+ , Na+ , Ca+ + , and H+) across cell membranes are often held far from equilibrium by active transport. Conversely, the extracellular anion present in largest concentrations, Cl-, is generally distributed near its GibbsDonnan equilibrium concentration via a number of passive anion exchangers and channels (1). However, anion conductivity, while generally passive, may be a limiting process in important circumstances including the disease cystic fibrosis (2), in vacuolar acidification (3), renal acid secretion (4), and bone resorption (5). This suggests that the proteins mediating Cl- transport may be of considerable interest. We have recently determined that the massive proton flux used by the osteoclast to dissolve bone is provided by an ATP-dependent proton pump (5). Furthermore, we found that this H+ flux is dependent on [Cl-], and that 10 pM DIDS inhibits the process in a manner that is competitively reversed by increasing [ATP] (6). Therefore, we reasoned that the competitive DIDS binding could be exploited to isolate a Clchannel from partially purified osteoclast membranes. We herein describe a method by which the Cl- transporting membrane protein of osteoclasts was isolated with retention of its DIDS sensitive Cl- channel activity. *To whom reprint
requests
should
be addressed.
Abbreviations used: DIDS-4,4’-diisothiocyanatostilbene-2,2’disulfonate. 0006-291X&O $1.50 Copyright 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.
920
Vol.
171,
No.
3, 1990
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
Methods Membrane Preparation: Osteoclasts, isolated from long bones of calcium deprived laying hens as described (7), were fragmented via nitrogen cavitation in 250 mM sucrose, 1 mM EGTA, 1 mM D’IT, 10 mM Tris, pH = 7.0 (lysis buffer). Cells (2 x ld7) in 20 mls of lysis buffer were pressurized to 40 atm N,, at 4’C, in a stainless steel bomb, and allowed to equilibrate for 30 minutes, followed by decompression over 2 seconds through a 5 mm orifice. Membranes were fractionated by differential centrifugation, retaining the supernatants after 1000 x g for 5 minutes and then 4,700 x g for 10 minutes. The membranes were collected by centrifugation at 48,000 x g for 40 minutes after the method of Gluck and Al Awqati (8). The pellet (2.4 mg of protein) was primarily osteoclast ruffled membranes. Affinity substrate: Dry CNBr activated Sepharose 4B (Sigma Chenical Co. St. Louis, MO) was swollen in 50 mM carbonate, pH 9, and reacted with 1,5diaminopentane, 10 mM, for 36 h at 4’C. The beads were washed several times with water and suspended in 50 mM carbonate, pH 9, with 3 mM DIDS at 4’C for 48 h. After packing in a disposable glass column the DIDS-Sepharose was washed with > 50 volumes of 50 mM Na-cyclamate, 10 mM HEPES @H 7.0), 1.4% (w/v) Noctylglucoside (column buffer) at 4’C. Membrane solubiliition: Membrane fractions in lysis buffer were dialyzed 24 h at 4’C in 12 - 14 kD cutoff cellulose membranes against 1000 volumes of the column buffer without N-octylglucoside, to which 0.5 mls of the acetate form of Dowex 1 had been added. This was repeated 3 times after which the dialyzed membranes were solubilized by the addition of 1.4 % (w/v) N-octylglucoside. DIDS-Sepharose chromatography of solubilized osteoclast membranes: 2.4 mg of solubili:zed membrane protein (in 1.0 ml) was passed over a 3 ml DIDS-Sepharose column equilibrated with the column buffer. The column was washed with a further 20 volume:s of the same buffer until A,,, had returned to a stable baseline, and then eluted with 50 mM KC1 + 10 mM HEPES @H 7.0) + 1.4% (w/v) N-octylglucoside. Pooled fractions were frozen at -80°C until used. Gel filtration of the fractions bound by DIDS-Sepharose: A Bio Sil Set 125 column (300 x ‘7.8 mm, BioRad, Richman CA) was equilibrated with 50 mM KCl, 10 mM HEPES (pH 7.0) and 1.4 % N-octylglucoside. The pooled DIDS-Sepharose peak was added in 0.5 ml of the same buffer and pumped through the column at 1 ml/minute. The column effluent was monitored at 280 nm and fractions of 800 ~1 were collected. The A,,, peaks were pooled separately and stored at -8O’C until used. Membrane reconstitution: Purified protein fractions were reconstituted into phospholipid vesicles by mixing 200 ~1 (l-3 pg of protein) of the pooled column fractions with 10 mg asolectin (Sigma Type 11s)and 9 mg of N-octylglucoside in 0.4 ml total volume. This mixture was dialized in a 12-14 kD cutoff cellulose dialysis membrane against 3 changes of 500 mls of 100 mM KCl, 100 mM sucrose, 10 mM HEPES, pH 7.0 (dialysis buffer) over a period of 18 h after the method of Kagawa and Racker (9) and Landry et al (10). 36Cl- Uptake: Reconstituted vesicles (30 ~1 and 0.05 - 0.15 pg of protein) were added to 70 ~1 of buffer (100 mM KC1 + 10 mM HEPES, pH 7.0) containing 0.05 &‘i of s6C1-, with or with out DIDS or valinomycin at the indicated concentrations. The reaction mixture was incubated at 25’C for 60 s and then run over a 1 ml Dowex 1 (acetate form) column which had been pre-washed with 1 ml of 25 mg/ml of BSA in 921
Vol. 171, No. 3, 1990
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
200 mM sucrose + 10 mM HEPES (pH 7.0). Immediately the column was further washed with 1 ml of 200 mM sucrose + 10 mM HEPES (pH 7.0) and the eluant pooled and counted for 36C1- (11). Concentrations and statistical methods: Protein concentrations were determined as described (12). Where a range is given it indicates mean f standard deviation.
Results From the nitrogen cavitation lysis of 2 x lo7 osteoclasts we obtained 250 mg of protein, after differential centrifugation the cell membranes contained 2.4 mg protein. These membranes consisted largely of ruffled membranes from the osteoclast bone attachment site. When these membranes were solubilized in 1.4 % N-octylglucoside and passed over an amine-linked DIDS-Sepharose affinity column 0.07&0.03% (n=3) of the protein was bound and eluted by 50 mM KC1 (Figure 1~). The chloride eluant contained one major protein and several minor bands by SDS-PAGE (Figure 1~). This mixture was resolved into Peak I, of approximately 60 kD, and Peak II, a low molecular weight “salt” peak, (Figure 2A) by HPLC gel filtration. SDS-PAGE of these two peaks on 10% gels revealed that Peak I contained one protein of approximately 60 kD (Figure 2B) and that Peak II contained several low molecular weight species (data not shown). Chloride transport activity was monitored by reconstitution of fractions into phospholipid vesicles using equilibrium dialysis as described. The vesicles formed by this procedure are shown in Figure 3. After the DIDS-Sepharose column we observed 36C1- uptake by reconstituted vesicles incorporating the KC1 eluted protein. The large unbound protein peak from the DIDS column was inactive (Table I). The HPLC gel filtration yielded a single major protein peak, Peak I, and a “salt” peak, Peak II. After reconstitution into lipid vesicles Peak I displayed 36C1- uptake greater than that of the active fractions from DIDS-Sepharose column (Table I). A.
B. FT
Cl-
-30 0
20
40
50
Fraction
50 -
dye
Figure 1. A] Protein elution from DIDS-Sepharose. Protein in eluant fractions was monitored at 280 nm. 2.4 mg of osteoclast membrane protein in 50 mM Nacyclamate, 10 mM HEPES (pH=7.0) and 1.4% N-octylglucoside was placed on a column equilibrated with the same buffer. The column was washed until the A2so returned to the baseline. Then 50 mM KCl, 10 mM HEPES (pH 7.0) and 1.4% N-octylglucoside was used to further wash the column. A small peak that eluted at fractions 25-28 was pooled and stored at -8O’C. B] Polyacrylamide gel electrophoresis of the flow through (FT) and KCl-eluted peak (Cl-) under reducing conditions. The position of molecular weight markers and the dye front are indicated.
922
Vol.
171, No. 3, 1990
BIOCHEMICAL
A.
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
B. GF PeakI
Peak II
0.2
02
k 0
3
Fraction 9 9
12
-30 IS -
-dye
03
mtre 2. A] Elution profile on a GFE gel filtration column run in 50 mM KC1 + 10 mM HEPES (pH=7.0) and 1.4% N-octylglucoside. 20-30 pg of protein from the KC1 eluted peak of the DIDS-Sepharose column was placed on this gel filtration column and the eluant monitored at 280 nm. The first peak (Peak I) was pooled and stored at -80°C until used. The second peak (Peak II) was at the salt elution front of this column and contained no detectable high molecular weight proteins. B] Polyacrylamide gel electrophoresis in SDS of Peak I from the gel filtration. Approximately 1 pg of protein from Peak I was placed on the gel after reduction. The position of molecular weight markers and the dye front are indicated. Figure 3. Electromicroscopy of reconstituted vesicles. Vesicles reconstituted with Peak I were produced as described in methods, fixed with 0~01, pelleted by centrifugation, and processed for electron microscopy (magnification 31, ZOO).
DIDS (200 PM) inhibited uptake of 36C1- by vesicles reconstituted with both the DIDS-Sepharose peak and Peak I from the HPLC gel filtration (Table I). The addition of 20 PM valinomycin slightly increased uptake by vesicles reconstituted with protein from both active fractions.
Discussion It is established that anion conductivities are important in a number of physiological processes, including vacuolar acidification (l-4). However, identification of a specific Cl- conducting protein has proved elusive. In recent reports (8,13), indanyloxacetic acid derivatives were used as affinity substrates to purify 4 Clconducting proteins, but reconstitution of channels with DIDS sensitivity has not been reported. In the present paper, we report a novel application of DIDS competitive inhibition of chloride transport to isolate a homogenous Cl- conducting protein. The resulting 60 kD protein (Figure 2) demonstrates 36Cl- conducting activity, which when reconstituted into lipid vesicles, retains stilbene sensitivity (Table II). The sensitivity to DIDS is similar to that in osteoclast membranes where 200 PM DIDS completely inhibited the Cl- dependent acidification (6). These results suggest that a single protein mediates osteoclast H+-ATPase associated chloride transport, although with these data 923
BIOCHEMICAL
Vol. 171, No. 3, 1990
AND BIOPHYSICAL
TABLE
RESEARCH COMMUNICATIONS
I
36ClATransport by ReconstitutedVesicles’
Preparation
Lipid Vesicles2 DIDS-Sepharose DIDS-Sepharose PassThrough
Control
+ 2OpM 36C1--Uptake Valinomycin3
cpm
cpm
cpm
42+15 n=2
N.D.
41+17 n=2
2878 n=l
1125-l
n=3 13+71
N.D.
6+1
2163+1015
n=2 HPLC Peak I
+2ooa DIDS4
n=3
n=2
3598&1709
n=3
4638+2085
n=3
750+ 147
n=3
‘Uptakevaluesare asmeancpm k standarddeviationandn is the numberof preparations in whicheachvalue wasdetermined.All uptakevaluesarenormalizedto 1 Pg of reconstituted protein. 2Lipidvesicleswerepreparedby substituting200~1of dialysisbuffer for the protein containingsamplefractionsusedelsewhere. 3Valinomycin(20 PM final concentration)wasincludedin the reactionmixture. Buffer compositionandtimeof incubationwasasdescribed in Methods.
4DIDS (200 PM final concentration) wasincluded in the reaction mixture. Buffer composition and time of incubation wasas describedin Methods. it is not possible to assesswhether multiple identical subunits are involved in transport by reconstituted vesicles.
The uptake of 36C1- did not require a vesicle membrane potential since the ionic composition was the same inside and outside the vesicles. In addition, 20 PM valinomycin slightly increased uptake under conditions where it would be expected to neutralize any membrane potential (I$=&= 100 mM). Uptake per pg of purified protein increased after HPLC gel chromatography consistent with the observed removal several contaminating proteins. It is possible that the 60 kD protein reported here is analogous to the 64 kD protein previously isolated from bovine kidney and trachea (8,13). They have similar molecular weights and the DllX insensitivity of the 64 kD bovine protein may reflect an artifact of the isolation technique used. However, there are most likely several distinct Cl- channels (l), and the two proteins, though close in molecular weight, may be structurally dissimilar and subserve different functions. 924
Vol.
171, No. 3, 1990
BIOCHEMICAL
AND BIOPHYSICAL RESEARCH COMMUNICATIONS
References 1. Franciolini, F., and Petris, A. (1990) B&him, Biolhys. Acta 1031, 247-259. 2. Li, M. McCann, J.D. Liedtke, C,M, Naim, A.C. Greengard, P. Welsh, M.J. (1988) Nature 331, 358-360. 3. Fuchs, R., Male, P. and Mellman, I. (1989) J. Biol. Chem. 264, 2212-2220. 4. Schmid, A., Burckhardt, G., and Gogelein, H. (1989) J. Membr. Biol. 111, 265275. 5. Blair, H.C., Teitelbaum, S.L., Ghiselli, R. and Gluck, S. (1989) Science 245, 855-857. 6. Manuscript submitted. 7. Blair, H.C., A.J. Kahn, E.C. Crouch, J.J. Jeffrey, and Teitelbaum, S.L. (1986) J. Cell Biol. 102, 1164-1172. 8. Gluck, S. and Al-Awqati, Q. (1984) J. Clin. Invest. 73, 1704-1710. 9. Kagawa, Y. and Racker, E. (1971) J. Biol. Chem. 246, 5477-5487. 10. Landry, D. W., Akabas, M.H., Readhead, C., Edelman, A., Cragoe, E.J. and AlAwqati, Q. (1989) Science 244, 1469-1472. 11. Breuer, W. (1989) J. Membr. Biol. 107, 35-42. 12. Groves, W.E., Davis, F.C. and Sells, B.H. (1968) Anal. Bioch. 22, 195-210. 13. Akabas, M., Edelman, A., Redhead, C., Land, D., and Al-Awqati, Q. (1990) Biophysical J. 57, 320a. 14. Blair, H., Koziol, C., Mead, R., Gluck, S., Teitelbaum, S., and Schlesinger, P. (1989) J. Cell Biol. 109, 137a.
925
