Isolation and reconstitution of the dicyclohexylcarbodiimide-sensitive proton pore of the clathrin-coated vesicle proton translocating complex.

The clathrin-coated vesicle proton translocating complex is composed of a maximum of eight polypeptides. The function of the components of this system have not been defined. Proton pumping catalyzed by the reconstituted, 200-fold purified proton translocating complex of clathrin-coated vesicles is inhibited 50% at a dicyclohexylcarbodiimide (DCCD)/protein ratio of 0.66 mumol of DCCD/mg of protein. At an identical DCCD/protein ratio, the 17-kDa component of the proton pump is labeled by [14C]DCCD. Through toluene extraction, the 17-kDa subunit has been isolated from the holoenzyme. The 17-kDa polypeptide diminished proteoliposome acidification when coreconstituted with either bacteriorhodopsin or the intact clathrin-coated vesicle proton translocating ATPase. In both instances, treatment of the 17-kDa polypeptide with DCCD restored proteoliposome acidification. Moreover, the proton-conducting activity of the 17-kDa polypeptide is abolished by trypsin digestion. These results demonstrate that the 17-kDa polypeptide present in the isolated proton ATPase of clathrin-coated vesicles is a subunit which functions as a transmembranous proton pore.

has recently been purified 200-fold and reconstituted into liposomes prepared from both crude (11) and defined (12) lipids. At the current stage of purification, the enzyme preparation contains eight major polypeptides, and the holoenzyme has a molecular mass of 530 kDa (11). There is a similarity in the polypeptide composition of the coated vesicle proton pump to those of several ATPases which have been isolated from acidifying organelles prepared from fungal, plant, and mammalian sources (13)(14)(15)(16)(17). Notable are apparent conservations of 67-80-, 57-64-, and 15-19.5-kDa polypeptides amongst these preparations. Speculations as to the functions of these putative subunits have been based exclusively on labeling experiments. Controversy exists regarding the catalytic center of these ATPases in that there is not uniformity in labeling with ATP or ATP analogues (13,15,16). In contrast, most investigators have identified the low molecular weight species present within these ATPase isolates as the site of DCCD binding (13)(14)(15)(16)(17). Based upon an analogy to the FO component of mitochondrial ATPase (18), it has been proposed that this low molecular weight component is a DCCD-inhibitable proton pore, which serves to lower the activation energy for the movement of protons across the lipid bilayer.
Beyond such labeling experiments, there is little direct evidence for the functional properties of the putative subunits of this new class of proton pumps. Indeed, the actual subunit composition of these pumps is not known and the polypeptide components of the isolates range from three for plant (13,16) and fungal (17) ATPases to eight for the clathrin-coated vesicle complex (11). It is possible that subunits critical for proton pumping are missing from preparations with apparently fewer components.
In order to determine the minimal polypeptide requirements for proton pumping, as well as to define subunit function, we have undertaken the dissociation of the isolated clathrin-coated vesicle proton translocating complex. In this report, we describe the isolation and reconstitution of the 17-kDa subunit, which we demonstrate to serve as a DCCDsensitive, transmembranous proton pore.
Preparations-Clathrin-coated vesicles were isolated from batches of 30 bovine brains as described (19). The proton translocating complex was solubilized with the nonionic detergent polyoxyethylene 9-lauryl ether (C,,E,) and was purified as we have reported (11).
ATPase activities of these preparations range from 12-16 pmol Of pi. mg of protein". rnin".
Zsolatwn of the Proton Pore from the Intact Proton Translocating Complex of Clathrin-coated Vesicles-Twenty ml (150 pg protein/ml) of final glycerol gradient fractions containing the purified clathrincoated vesicle proton pump (11) were pooled, and the ATPase was concentrated by adding powdered (NH&S04 to a final saturation of 50% and centrifuging the mixture of 100,000 X g for 30 min. The pellets were dissolved in 2 ml of Buffer A, consisting of 0.5 mM dithiothreitol, 0.1% CIzEg, and 10 mM Tris brought to pH 7.0 with MES. Twenty ml of ice-cold toluene was added and the mixture was stirred for 17 h at 4 "C. (Extractions performed at room temperature did not yield a functional proton pore.) The toluene phase was collected and was dried under Nz. The residue was either dissolved in 1.5 ml of Buffer A for coreconstitution with the clathrin-coatedvesicle proton translocating complex or with 1.5 ml of 7 mM Tricine at pH 8.0 and 1.25% octylglucopyranoside for coreconstitution with bacteriorhodopsin. The isolated proton pore was stored at 4 "C and was stable for 2 weeks.
Reconstitutions-For reconstitution of both bacteriorhodopsin and the clathrin-coated vesicle proton translocating complex, liposome composition was exactly as described (12). Briefly, lipid mixtures (125 mg/ml) were prepared with phosphatidylcholine, phosphatidylethanolamine, phosphatidylserine, and cholesterol at a weight ratio of 40:26.5:7.5:26. The mixture was sonicated for 3 10-min cycles under Nz and stored at 4 "C.
The clathrin-coated vesicle proton pump was reconstituted by the freeze-thaw, cholate-dilution method as described (12). Designated amounts of purified holoenzyme, with or without the isolated 17-kDa subunit, were added to 7.5 pl of the lipid mixture. Glycerol, sodium cholate, KCl, and MgCl, were added to the protein/lipid mixture at final concentrations of 20% (v/v), 1%, 0.1 M, and 2.5 mM, respectively. After vortexing, the mixture was frozen at -70 "C for 15 min, the proteoliposomes were thawed on ice and then were added to 1.5 ml of assay buffer (described below) for measurement of ATP-generated acridine orange quenching. dilution method (19). Fifteen pl of bacteriorhodopsin (3 mg protein/ Bacteriorhodopsin was reconstituted by the octylglucopyranoside ml in 150 mM KCl, 1.25% octylglucopyranoside) was mixed with 5 pl of the pure lipid mixture, with or without addition of the 17-kDa subunit. The mixture was incubated at 25 'C for 5 min and was subsequently added to 1.0 ml of 150 mM KC1 for measurement of light-energized proteoliposome acidification, as described below.
Measurement of Proton Pumping and ATPase Activities and Performance of SDS-Polyacrylamide Gel Electrophoresis and Autoradwgraphy-In experiments designed to determine the effects of coreconstitution of the proton pore with the intact proton pump on proteoliposome acidification, proton pumping was assessed in an SLM DWPC dual wavelength spectrophotometer, using the pH-sensitive dye, acridine orange. Measurement of ATP-generated decreases in the absorbance of acridine orange was performed as L L~~~~.~~ and quantified as the initial slope of ATP-induced decrease in absorbance, as reported (1,8). The reaction mixture (1.5 ml) consisted of 100 mM KCl, 20 mM Tricine at pH 7.0, 5 mM MgCIZ, 3 mM NaN3, 6.7 PM acridine orange, and designated amounts of proteoliposomes. Reactions were initiated by adding 10 pl of 0.2 M NaATP at pH 7.0 and 1.5 p1 of 2 mM valinomycin.
Measurement of proteoliposome acidification catalyzed by bacteriorhodopsin was performed essentially as described (20). Illumination of a water-jacketed cuvette fitted with a stirrer was performed with a Bessler lantern slide projector equipped with a Sylvania 300watt tungsten halogen bulb situated at a constant distance of 7 cm from the cuvette. Alkalinization of the proteoliposome media was measured with a Fisher calomel "Microprobe" pH electrode attached to a Corning model 135 pH meter with integral potentiometer. Recording of light-induced changes in media pH was obtained with a Lindi variable chart speed recorder equipped with a variable rheostat. The reaction solution (1 ml) consisted of 150 mM KC1 and 3 p~ valinomycin and calibration was performed after experiments by injecting known quantities of HCl into the cuvette, ATPase activity was measured as the liberation of 32Pi from [y-32P]ATP. The assay solution consisted of 40 mM KC], 40 mM Tris-MES, pH 7.0, 4 mM MgClz, and 4 mM [y-32P]ATP (400 cpm/nmol). Samples were preincubated with 2.5 pg of phosphatidylserine for 5 min at 25 'C, and the reaction was initiated by the addition of 200 pl of assay solution and after a 10-min incubation at 37 "C, was terminated by the addition of 1.0 ml of 1.25 N perchloric acid. Liberated 3zPi was assayed as described (21). the isolated proton pore with ["CIDCCD were performed as described Labelings of the intact clathrin-coated vesicle proton pump and in the legends to figures. SDS-polyacrylamide gel electrophoresis was performed as described (22). Autoradiography was performed with Kodak XAR-5 film which was exposed to Enlightning-treated gels overnight at -70 "C in the presence of an enhancer screen. Protein was determined by the Amido-Schwartz method (23).

RESULTS
As we have previously reported, proton pumping and ["Pi] ATP exchange activities of intact clathrin-coated vesicles are inhibited by DCCD (1). Shown in Fig. 1 are the effects of DCCD on ATPase activity catalyzed by the solubilized, 200fold purified proton ATPase and on proton pumping catalyzed by the proton pump when reconstituted into liposomes composed of pure lipids. Under described conditions (i.e. 5-min preincubation with DCCD at 25 "C), ATP hydrolytic activity is inhibited 50% by 130 p~ DCCD (0.69 pmol of DCCD/mg of protein) and proton pumping, as assessed by the initial rate of ATP-generated acridine orange quenching, is inhibited 50% by Isolation of the 17-kDa polypeptide from the 200-fold purified holoenzyme was accomplished by toluene extraction. Approximately 60 rg of protein was extracted from the 3 mg of starting material. Shown in Fig. 2, panel 11 are the Coomassie-stained extract and the ['4C]DCCD-labeled extract. Loss of a sharply focused band on SDS-polyacrylamide gel electrophoresis was a consistent (and insurmountable) feature and likely was caused by the co-extraction of lipids present within our ATPase preparation. Nonetheless, the predominant constituent within the extract is a protein with an apparent molecular mass of 15-20 kDa.
To determine if the isolated 17-kDa polypeptide could function as a proton pore, coreconstitution of the toluene extracted protein was performed using two different proton gradient generating systems: the light-driven proton pump, In separate experiments, we have found that this amount of DCCD does not affect light-energized proteoliposome acidification catalyzed by bacteriorhodopsin (data not shown).
It should be noted that a very high DCCD/protein ratio was required to inhibit the proton conductance induced by the 17-kDa polypeptide in the experiments of Fig. 3. This is likely due to partitioning of DCCD into the lipids present in the mixture. In order to address this issue, experiments were performed (Fig. 4) in which the 17-kDa polypeptide was treated with DCCD prior to coreconstitution with bacteriorhodopsin. Under these conditions, incubation of the 17-kDa polypeptide with DCCD a t a ratio of 0.35 pmol of DCCD/mg For all experiments, bacteriorhodopsin (45 p g ) was reconstituted with 625 pg of lipids and assayed as described under "Experimental Procedures." Trace A is the control experiment performed with bacteriorhodopsin proteoliposomes. Trace B is the effect of coreconstituting 0.8 pg of the 17-kDa subunit with bacteriorhodopsin. Trace C, 0.8 p g of the 17-kDa subunit was preincubated with DCCD at a DCCD/ protein ratio of 0.7 pmol of DCCD/mg of protein for 20 min prior to coreconstitution with bacteriorhodopsin. For all experiments, 1.5 pgof the purified clathrincoated vesicle proton pump was reconstituted with 300 pg of lipids, and proteoliposome acidification was assessed by ATP-generated acridine orange quenching, as described under "Experimental Procedures." Circles (-17 kDa) indicate the effect of preincubating the holoenzyme for 5 min prior to reconstitution with designated concentrations of DCCD. Squares (+17 kDa), indicate the effect of coreconstituting 0.4 pg of the 17-kDa subunit with holoenzyme. The 17-kDa subunit was preincubated for 5 min with designated concentrations of DCCD, as indicated. In all instances, (squares and circles) the amount of DCCD in the final reaction mixture was identical for each preincubation concentration.
of protein results in restoration of bacteriorhodopsin catalyzed medium alkalinization. Taken together, these results indicate that the uncoupling effect of the 17-kDa polypeptide on proteoliposome acidification is DCCD-sensitive and thus demonstrate that the 17-kDa subunit acts as a proton pore.
Similar results were obtained when the 17-kDa subunit was coreconstituted with the clathrin-coated vesicle proton translocating complex, as shown in Fig. 5. As can be seen, addition of 0.4 pg of the isolated 17-kDa polypeptide to the reconstitution mixture, in the absence of DCCD, reduces the initial slope of ATP-generated acridine orange quenching from a control rate of 4.6 to 2.1. With pretreatment of the 17-kDa subunit with DCCD (25 p~ or 0.35 pmol of DCCD/mg of protein), restoration of the initial rate of proton pumping is achieved to a rate (4.1) which is indistinguishable from that occurring with the holoenzyme reconstituted with an equal amount of DCCD, but without the 17-kDa component. At higher concentrations of DCCD (40 p~ or 0.56 pmol of DCCD/mg of protein), the inhibitory effect of DCCD on the

TABLE I Effect of trypsin on the inhibition by the 17-kDa polypeptide of bacteriorhodopsin-catalyzed proteoliposome acidification
Bacteriorhodopsin (45 pg) was reconstituted exactly as described in the legend to Fig. 3. For the control, 2 pl of soy trypsin inhibitor (3 mg/ml) and 20 units of trypsin were added prior to reconstitution. The 17-kDa subunit (0.8 pg) was incubated at 25 "C for 14 h in the absence or presence of 2 p1 (20 units) of trypsin, and 2 pl (3 mg/ml) of soy trypsin inhibitor was added prior to coreconstitution with bacteriorhodopsin. Proton pumping was assessed as described under "Experimental Procedures." holoenzyme is predominent; i.e. the effects of DCCD on the 17-kDa component are obscured by direct inhibition of the proton gradient generating system itself, the proton ATPase. Thus, in a second proton gradient generating system, the 17-kDa polypeptide preparation can be shown to function as a DCCD-inhibitable proton pore.

Medium alkalinization
The results of experiments listed in Table I demonstrate that the uncoupling effect of the 17-kDa polypeptide is trypsin-sensitive and hence due to the protein moiety of the subunit, as opposed to contaminating lipids.

DISCUSSION
Taken together, these experiments demonstrate that the 17"kDa polypeptide of the clathrin-coated vesicle proton translocating complex is a DCCD-inhibitable, transmembranous proton pore. Resemblance to the Fo component of FIFo ATPases (18) is striking and supports the popular thesis that endomembrane pumps in general, and the clathrin-coated vesicle proton pump in particular, are F,Fo-type pumps and share a distant ancestry with the mitochondrial pump in the proton ATPase of anaerobic bacteria (25). Indeed, other investigators have used the coreconstitution of mitochondrial Fo (18,26) and chloroplast Fo (27) with bacteriorhodopsin to demonstrate the proton-conducting activity of these proteolipids. Furthermore, the finding that the coreconstitution of the 17-kDa subunit with the clathrin-coated vesicle proton pump yields a DCCD-inhibitable proton conductance is highly reminiscent of experiments conducted with submitochondrial particles which were partially depleted of F1 (28).
Because of the close analogy to mitochondrial Fo, it is important to note that the 17-kDa subunit differs from mitochondrial Fo in several respects, and thus we have isolated the proton-conducting unit of the clathrin-coated vesicle proton pump. First, the 8-10-kDa mitochondrial Fo is substantially smaller (18) than the 17-kDa proton pore of the clathrincoated vesicle proton pump. Second, 100-fold greater DCCD/ protein ratios are required to inhibit the clathrin-coated vesicle proton pump and the 17-kDa subunit, as compared to the mitochondrial FIFo (1). Third, in side-by-side comparison (11) we have shown that our final clathrin-coated vesicle ATPase preparation shares no polypeptides in common with purified mitochondrial F,. As our final step in purification is a glycerol gradient centrifugation, it is highly unlikely that the 530-kDa clathrin-coated vesicle preparation is contaminated with dissociated mitochondrial Fo, a complex of a molecular mass of 8 kDa.
Further experiments are required to determine the compo-sition of the clathrin-coated vesicle proton pore. By analogy to Fo, it is likely that the 17-kDa subunit of the clathrincoated vesicle proton pump is a proteolipid. We cannot at present discern whether the proton pore of coated vesicles is a single polypeptide or is a composite of several polypeptides and proteolipids (as is Fo); loss of resolution after extraction precludes such definition and at present we claim isolation, rather than purification, of the proton pore.
A key issue in the field of vacuolar H' ATPases, of which the clathrin-coated vesicle proton pump is an example, is the possible relationship to FIFO type pumps. Although the presence of a dissociable proton pore (and likely proteolipid) in the coated vesicle ATPase suggests a relationship, a more important differentiating feature is the mechanism of the catalytic sector. The vanadate insensitivity of the coated vesicle ATPase (1) suggests the absence of a phosphoaspartyl intermediate which participates in the catalytic cycle of E1E2type ATPases such as Na,K-ATPase (29). However, sensitivity to vanadate can be a latent feature of known E,E,-type enzymes (30), and at present, the lack of vanadate inhibition of our system can be viewed as only indirect evidence that the coated vesicle pump does not have a phosphoenzyme intermediate. It is our view that if the clathrin-coated vesicle proton pump is an FIFo-type ATPase, then the divergence from the mitochondrial ATPase is extreme. In order to gain insight into this issue, resolution and functional definition of the catalytic cycle of the clathrin-coated vesicle proton translocating ATPase is necessary.