ATP-dependent Acidification of Intact and Disrupted Lysosomes EVIDENCE FOR AN ATP-DRIVEN PROTON PUMP*

Intralysosomal acidification, necessary for hydrolase activities that permit reutilization of the component parts of proteins and other macromolecules, was stud- ied by measuring the uptake of [14C]methylamine in purified and disrupted rat liver lysosomes. When MgATP was added to intact lysosomes, acidification was stimulated by 1 pH unit within 20 min at 25 “C. The acidification process involved nucleotide specificity for ATP; GTP stimulated less than ATP, whereas the nonhydrolyzable analog, adenosine fi’-(P,y-irnido)tri- phosphate, did not stimulate.

Intralysosomal acidification, necessary for hydrolase activities that permit reutilization of the component parts of proteins and other macromolecules, was studied by measuring the uptake of [14C]methylamine in purified and disrupted rat liver lysosomes. When MgATP was added to intact lysosomes, acidification was stimulated by 1 pH unit within 20 min a t 25 "C. The acidification process involved nucleotide specificity for ATP; GTP stimulated less than ATP, whereas the nonhydrolyzable analog, adenosine fi'-(P,y-irnido)triphosphate, did not stimulate. Acidification of lysosomes was inhibited by the proton ionophore, carbonyl cyanide m-chlorophenylhydrazone. ATP had little effect on lysosomal membrane potential, and the potassium ionophore, valinomycin, which has effects on membrane potential, had little effect on ATP-dependent acidification of lysosomes. In addition, an inhibitor of anion transport, diisothiocyanostilbene disulfonic acid, inhibited ATP-dependent acidification. Therefore, it is concluded that ATP-dependent acidification of lysosomes occurs electroneutrally.
ATP-dependent acidification and the membrane ATPase of lysosomes have several common characteristics. Both activities are stimulated by preincubation of lysosomes in salts that elevate intralysosomal pH; both are inhibited by dicyclohexylcarbodiimide; and both have pH optima slightly above pH 7. These characteristics further suggest that acidification and ATPase activities may be coordinated by the same lysosomal membrane protein.
Lysosomal preparations disrupted by freezing were shown to be capable of ATP-dependent acidification despite the loss of soluble lysosomal enzymes that act as Donnan effectors (93-100% loss). Furthermore, acidification occurred only when proteinase inhibitors were present during freezing. These results provide evidence for an ATP-driven pump that operates electroneutrally and enzymatically and results in acidification of lysosomes.
Liver lysosomes are known to be necessary for degradation of extracellular proteins such as asialoceruloplasmin (1-31, lipoproteins (4, 5) and altered proteins (6). In addition, lysosomes are probably necessary for degradation of intracellular protein (7-11). Degradation is required for reutilization of component parts of proteins and other macromolecules and, when impaired, results in serious disorders such as lysosomal storage diseases (12). The proper functioning of lysosomes depends on an acid pH inside; drugs such as chloroquine (13) and inhibitors of energy metabolism (14) elevate intralysosomal pH and inhibit protein degradation (4,(15)(16)(17).
Although intralysosomal pH is thought to be about pH 4.5 in cells (14,18,19), lysosomes purified from livers of rats treated with Triton WR-1339 have an internal pH of about 6 (20-22). The latter pH has been attributed to a Donnan equilibrium dependent on polyanions inside lysosomes and a membrane selectively permeable to protons (22). The polyanions acting as Donnan effectors are presumably soluble lysosomal enzymes, most of which are glycoproteins with acidic isoelectric points. However, it is questionable whether such a Donnan equilibrium can maintain the proper intralysosomal pH under in vivo conditions: these involve ( a ) a high concentration of cytoplasmic potassium ion with a probable leak rate into lysosomes of tl,z = 20 min (23), and ( b ) a considerable flow into lysosomes of protein buffered at neutral pH, probably 10% of plasma protein daily (24).
Evidence in support of ATP-dependent acidification of lysosomes has been reported. First,  isolated intact lysosomes containing radiolabeled albumin and showed that proteolysis of the albumin was stimulated by addition of ATP when the lysosomes were incubated in medium at neutral pH and not when incubated at acid pH. The effect of ATP was diminished by the proton ionophore dinitrophenol (26). Second, we and others have measured intralysosomal pH and observed acidification after addition of ATP (27)(28)(29)(30). In a preliminary report, the acidification activity was correlated with membrane ATPase activity in terms of effects of different salts, gramicidin, and uncoupler (29). Here, these studies are extended to show the effects of valinomycin, of different nucleotides, and of the ATPase inhibitors oligomycin, diamide, and dicyclohexylcarbodiimide. Also, the time course of acidification is examined, and the effects of ATP on the membrane potential and the internal volume of lysosomes are considered. Furthermore, it is shown that disrupted lysosomes, in which the soluble enzymes that have been postulated to be Donnan effectors are no longer latent, are active in ATP-dependent acidification.

Lysosomes
These were purified from rat liver using Triton WR-1339 treatment. Rats were injected intraperitoneally with Triton WR-1339 and 3% days later liver lysosomes were isolated by differential centrifugation and flotation through 34.5% sucrose, w/w, as described (31), except that, by not fasting the rats the night before killing, a higher enrichment was obtained (32). Isolation of lysosomes from rats fed ad libitum by this Triton technique provides highly purified preparations that have no detectable catalase nor cytochrome oxidase activity and are, thus, essentially free of peroxisomal and mitochondrial contamination. Glucose-6-phosphatase measurements indicate less than 10% contamination by endoplasmic reticulum. Purity was assessed by measuring N-acetylglucosaminidase and protein (32) and found to be at least 60-fold enriched over homogenate. As isolated at the 14-34R Sucrose (w/w) interface of the sucrose gradient, the lysosomes are intact and display 80-9056 latency for N-acetylglucosaminidase activity. In some experiments, lysosomal membranes were used; these were prepared free of the soluble hydrolases by washing with 0.2 M NaCl at pH 8 as described previously (33).

ATPase Assay
Lysosomes were mixed with 200 mM sucrose, 75 mM salt as indicated, 20 mM MOPS,' pH 7.0 with Tris, 3 mM MgC12, 1.5 mM ATP, 0.5 mM EGTA, and 0.5 mg of bovine serum albumidml. The assay at 37 "C was initiated by addition of lysosomes and terminated after 10 min by addition of trichloroacetic acid to 3% final concentration. After centrifugation to remove denatured protein, released phosphate was determined colorimetrically with ammonium molybdate-ascorbate (34,35).

Intralysosomal p H Measurements
Microfuge Assay-These measurements were carried out as described by others (20-22). Duplicate, l-ml samples of lysosomes containing either ['4C]methylamine and [:'H]water or ['4C]sucrose and ["]water were incubated a t 25 "C for the times indicated, usually 20 rnin. Then lysosomes were collected by centrifugation at 15,000 X g for 2 min in a Brinkmann Model 3200 microfuge. Supernatants were removed by using a Pasteur pipette drawn out to a fine tip, and the pellets were analyzed for radioactivity and protein. The amounts of radioactivity used were such that ["HI was about 10,OOO cpm/pl and I4C was about 500 cpm. Methylamine was 7 PM throughout, i.e. unlabeled methylamine was added to samples containing ['4C]sucrose. The internal volumes of lysosomes were found to be about 4 pl/mg of protein, using ['4C]sucrose to measure the pellet volume outside of lysosomes and ["]water to measure the total pellet volume, outside and inside. These volumes were not significantly changed in the presence of ATP, except in the absence of salt. The averages of 16 determinations in potassium chloride were 4.25 f 1.03 and 4.01 k 0.54 $/mg of protein in the absence and presence of ATP, respectively. The effect of ATP on internal volumes is so small that the observed changes in methylamine uptake cannot possibly be attributed to changes in internal volume alone; even if one assumes an error of 50% in determination of internal volume, then the IO-fold proton gradient generated by addition of ATP (Fig. l b ) would be reduced by half to 5-fold, a ApH of 0.70. The internal pH was calculated by equating the methylamine gradient to the proton gradient. This calculation has a theoretical basis (22,36). In addition, measurements of the internal pH of chromaffin granules gave similar results whether methylamine or NMR techniques were employed (37)(38)(39). It is important to point out that methylamine distributes across membranes more rapidly than can be detected using a 2-min centrifugation for separation of lysosomes. Although it is convenient to have methylamine present throughout the duration of these incubations, when methylamine was added just 1 min prior to centrifugation, results were unchanged.
Rapid Gel Filtration Assay-After incubation of lysosome preparations with methylamine, a 150.~1 aliquot of each mixture was applied to a column of Sephadex G-50 in a 1-ml Tuberculin syringe and equilibrated with 2 mM MgSO?, 1 mM ATP, and with MOPS, KCI, sucrose, and EGTA at concentrations equal to those in the incubation mixtures. These columns are pretreated by centrifugation a t 200 X g for 3 min and eluted by a second, identical Centrifugation. Previous results have shown that rapid gel filtration of lysosomes through Sephadex G-50 during 3 min of centrifugation at 200 X g affords a suitable separation, with 99.998% of the methylamine being retained by the Sephadex in the absence of vesicles (27). In the case of rapid gel filtration, ApH was calculated by dividing the amount of methylamine uptake per mg of protein observed in the presence of ATP by the amount of uptake in the absence of ATP, assuming no ATP-dependent volume changes. Although previous data indicate that this assumption is reasonable, it has not been verified directly using ["Hlwater to label the intralysosomal space because water diffuses very rapidly and labeled water is completely retained by the Sephadex. Regarding reproducibility of results, combined errors in measuring radioactivity and either internal volume or protein within an experiment are about 10% which would correspond to a ApH of 0.04. Potassium [I4C]thiocyanate uptake by lysosomes was used to measure membrane potential. The technique is based on the high permeability of membranes to thiocyanate anion; net uptake values permit calculation of membrane potentials from the Nernst equation (40). Measurements were made by mixing lysosomes in the same buffered salt media used for methylamine uptake except that potassium thiocyanate was 7 PM (60 mCi/mmol).

Intralysosomal Acidification
Measurement of IntralysosomalpH-The pH of lysosomes isolated in unbuffered sucrose media was measured using the technique of radioactive methylamine uptake and was found to be acid ( Fig. la). Since the pH outside of the lysosomes was 7.0, the ApH across the lysosomal membrane was about 0.6. This ApH was diminished to about 0.1 pH unit when lysosomes were preincubated at 0 "C for 1 h in 75 mM KC1 ( Fig. lb). In each case, addition of external MgATP promoted acidification to about pH 5.6 inside. After preincubation, lysosomes in ATP were 1.15 pH units more acidic inside than lysosomes incubated in EDTA and 1.35 pH units more acidic inside than out. ATP-dependent acidification occurred on a time scale of minutes and was complete by 20 min. In these acidification experiments, EDTA was added to lysosomes that were incubated without ATP as a control. The rationale is that both ATP and EDTA have chelating properties and similar negative charge. Although EDTA did not promote acidification, it was desirable to investigate further the requirement for ATP.
Nucleotide Specificity-Acidification showed nucleotide 'I were isolated from livers of Triton WR-1339-treated rats as described under "Materials and Methods." The lysosomes were mixed with 200 mM sucrose, 75 mM KCI, 20 mM MOPS (pH 7), with Tris, 0.5 mM EDTA, 3 mM MgCI,, 7 PM methylamine end either 1.5 mM EDTA (0) or 1.5 mM ATP (0) a t 25 "C; acidification was measured as uptake of radioactive methylamine by centrifugation at the indicated times. a, lysosomes used as isolated; b, preincubated at 0 "C for 1 h with 75 mM KC1 and 20 mM Tris-S04, pH 8. Fig. la. After incubation at 25 "C for 20 min, the lysosomes were isolated and ATP-dependent ApH was determined as in Table 11. ATPase activity was measured colorimetrically with ammonium molybdate-ascorbate as described under "Materials and Methods."

ATP-deDendent ADH
ATPase activitv specificity in that ATP was more effective than GTP (Table  I). Under these conditions intralysosomal pH in the absence of nucleotide was 6.35; thus adenosine 5'-(P,y-imido)triphosphate, a nonhydrolyzable analog of ATP, resulted in no acidification. On the other hand, GTP did support significant acidification inasmuch as a ApH of 0.25 corresponds to a 1.8 X proton gradient. In view of the nucleotide specificity, it appeared that ATP may promote acidification by driving ion movements across the lysosomal membrane. Accordingly, the effects of various salts on intralysosomal pH were studied. Salt Effects-Preincubation of lysosomes in various salts and subsequent measurement of internal pH showed that all were nearly equally effective at raising the intralysosomal pH (Table 11). Furthermore, in each case, addition of ATP caused acidification by at least 0.75 pH units. In view of the lack of ion specificity, it is not possible to say what ions are being induced to move across the lysosomal membrane when ATP is added; however, ATP-dependent acidifkation of lysosomes was inhibited nearly completely by 60 p~ diisothiocyanostilbene disulfonic acid and to a slight, reproducible extent by 1 p~ (Table 111). Since 1 p~ diisothiocyanostilbene disulfonic acid is known to inhibit anion transport in erythrocytes (41), ATP-dependent movement of protons and anions appears possible. Inasmuch as paired movement of protons and anions would be electroneutral, whereas proton movement alone would be electrogenic, the effect of ATP on lysosomal membrane potential was investigated.

TABLE I Nucleotide specificity for acidification of lysosomes and for
phosphate release by lysosomal membranes Lysosomes were freshly prepared and incubated in sucrose and potassium chloride; intralysosomal pH was measured 15 min after addition of 1.5 mM nucleotide as described in Fig. 1. The pH values were rounded off to the nearest 0.05 unit. The ApH values are relative to intralysosomal pH of lysosomes incubated similarly but with EDTA in place of nucleotide. Lysosomal membranes were incubated in a similar manner, and phosphate release was determined after 10 min as described under "ATPase Assay" (see "Materials and Methods").

TABLE I1
Effects of salts and ATP on intralysosomal pH Freshly prepared lysosomes were incubated for 1 h at 4 "C either as isolated in about 0.75 M sucrose or diluted 20% to give 20 m~ Tris-SO4 (pH 8), 75 mM salt as indicated, and about 0.6 M sucrose. "None" denotes lysosomes that were not preincubated in buffered salt and were incubated, as described in the legend to Fig. la, without KC1. ATPase activity was measured colorimetrically with ammonium molybdate-ascorbate as described under "Materials and Methods." ATPdependent ApH is defied as the intralysosomal pH value obtained for lysosomes in EDTA minus that value obtained for the same lysosomes in ATP. Effect of ATP on Membrane Potential-Using ['*C]thiocyanate distribution in lysosomes as a measure of membrane potential, ATP-dependent changes in membrane potential were found to be small (Fig. 2a) and can presumably be attributed to magnesium ion, because addition of MgEDTA gave similar changes. It is worth noting that the lack of a membrane potential is not due to membrane leakiness. When lysosomes at 25 "C were suspended in medium containing 150 mM KC1 and 1 pg/ml of valinomycin, a potassium ion-specific ionophore, the distribution of ['4C]thiocyanate indicated a membrane potential of +40 mV. A value of +40 mV is predicted by the diffusion potential if one assumes that the intralysosomal potassium concentration is 33 m~. In the absence or presence of ATP, the membrane potential of +40 mV was not diminished after 30 min of incubation at 25 "C. Thus, the lysosomal membrane is not freely permeable to ions, and the absence of a membrane potential during ATP-dependent acidification of lysosomes is, then, indicative of an electroneutral mechanism. If ATP-dependent acidification of lysosomes is electroneutral, then it should not be inhibited by the membrane potential imposed by a potassium chloride gradient and valinomycin. On the other hand, if the addition of ATP to lysosomes were to cause a membrane permeability change that would allow a Donnan equilibrium to establish itself, then a potassium chloride gradient and valinomycin should be inhibitory. These possibilities were considered next.

None
Effects of Ionophores and Inhibitors on ATP-dependent Acidification- Table IV shows that ATP-dependent acidification was not inhibited by valinomycin plus potassium ion. In addition, the proton ionophore, C-CCP did inhibit the ATP effect which shows that ATP leads to formation of a net proton gradient. If ATP promotes acidification of lysosomes by driving ion pump(& then acidification might be sensitive to inhibition by specific compounds. ATP-dependent acidification of lysosomes was inhibited by 10 PM dicyclohexylcarbodiimide and 50 p~ diamide; oligomycin (0.5 pg/ml) did not inhibit acidification. The relatively small amounts of the inhibitors employed, 10 nmol of dicyclohexylcarbodiimide and 50 nmol of diamide, would not be expected to react with all of the soluble lysosomal enzymes that are possible Donnan effectors but rather with specific proteins. Therefore, the presence of a lysosomal protein that requires ATP to drive catalytically proton translocation across the lysosomal membrane seems more probable. An ATP-driven proton pump would be expected to have a pH optimum, a point at which the ApH across the lysosomal membrane is maximal; in contrast, a Donnan equilibrium should maintain a constant ApH as the outside pH is varied.
Effect of External pH on ATP-dependent Acidification-Intact lysosomes were incubated in the absence and presence of external MgATP at various pH values, and intralysosomal

Effects of ionophores on ATP-dependent acidification of lysosomes
Lysosomes were prepared and incubated for 20 min as in Fig. Ib, except that salt was either 100 mM KC1 or 100 mM NaCl, as indicated, and that ionophores were added. Valinomycin was used at a concentration of I pg/ml, and C-CCP at IO p~; stock solutions were 1 mg/ml and IO mM. resrtectivelv. in 95% ethanol. pH, determined by distribution of [I4C]methylamine, was found to be greatest when the external pH was slightly alkaline (Table V). The observed maximum in ATP-dependent ApH at slightly alkaline external pH is what one might expect for a catalyzed process and is difficult to reconcile with a Donnan equilibrium. Therefore, the possibility that a lysosomal membrane protein with ATPase activity may act as a proton pump is to be considered.

ATPase Activity Associated with ATP-dependent Acidification
Lysosomal membranes are known to have ATPase activity (35). Table I compares the phosphohydrolase activity of membranes acting on various nucleotides with the ability of these nucleotides to promote acidification of intact lysosomes. ADP is hydrolyzed even though it does not promote acidification; this activity corresponds to the previously reported nucleotide diphosphatase activity (49) and has been recently reported to co-sediment with ATPase activity when Triton X-100-solubilized lysosomal membranes are centrifuged into sucrose gradients (50). Although G T P was a better substrate than ATP at 1.5 mM, previous results reveal that ATP is the preferred substrate at 0.2 m~ and below (35). ATP, above 2 mM, has an inhibitory effect on ATPase activity which is probably related to the fact that 1.5 m~ G T P is a superior substrate. The observations that neither ADP and GTP promote acidification as well as ATP may indicate that ATP is the normal substrate for the ATPase, but further comment is required. It is possible that both ADP and GTP uncouple phosphohydrolase activity from proton pump activity. Alternatively, it is possible that there are multiple nucleoside phosphatases in the lysosomal membrane and that ADP and GTP phosphohydrolases are different activities than the ATPase.
The ATPase activity of intact lysosomes was stimulated by preincubation in buffered salt solution in parallel with ATPdependent acidification (Table 11). The increase in activity may be due to ( a ) an increase in intralysosomal pH, ( b ) a dependency on ions, or (e) lysis of the lysosomal membrane. An increase in intralysosomal pH is known to occur (Fig. 1). Preincubation with salt led to an increase in intralysosomal pH when the lysosomes were buffered at pH 8 (Table 11) or a t pH 7 (Fig. 2). There was an apparent lack of ion specificity (Table 11), and the latency of N-acetylglucosaminidase was only slightly affected by preincubation (27). By the process of elimination, the major factor determining the increase in ATPase activity may be the increase in intralysosomal pH.
The inhibitors of ATP-dependent acidification, diisothiocyanostilbene disulfonic acid, dicyclohexylcarbodiimide, and diamide were also found to inhibit lysosomal membrane ATPase activity and thus suggest a link between ATPase and acidification activities. Dicyclohexylcarbodiimide completely inhibited ATP-dependent acidification, while it inhibited ATPase activity 60%. However, diisothiocyanostilbene disulfonic acid inhibited ATP-dependent acidification with little effect on ATPase activity (Table 111). One possibility is that proton pump activity is more sensitive to inhibitors and inactivation than ATPase activity. An illustration of the apparent lability of the pump activity is the observation that it usually cannot be measured in disrupted lysosome preparations even though the ATPase activity is unaffected.

ATP-dependent Acidification of Disrupted Lysosomal Preparations
Lysosomal preparations were disrupted by freezing and storing overnight at -20 "C. Disruption was complete as evidenced by a lack of latent N-acetylglucosaminidase activity. Freshly prepared lysosomes were 80-90% latent; frozenstored ones were, 0 to at most 7%, latent. The membranes of disrupted lysosomes apparently reseal spontaneously because an amount of ['4C]s~~rose that corresponds to less than 1 pl/ mg of protein was included in the membranes after gel fiitra-

Effect of external pH on ATP-dependent acidification of lysosomes
Freshly prepared lysosomes were incubated for 30 min in the absence and presence of ATP as in Fig. la In Vitro Acidification tion. The lack of latency indicates release of most of the soluble enzymes from the lysosomes. Since the soluble enzymes are postulated to be the polyanions that create a Donnan effect, the disrupted preparations should not be capable of acidification if the only mechanism of acidification is Donnan equilibrium.
To study acidification of disrupted lysosome preparations, it was necessary to use rapid gel filtration for separation of lysosomal vesicles from their external medium because centrifugation at 15,000 X g for 2 min did not result in complete pelleting. Table VI shows results obtained with intact lysosomes analyzed by the two methods. ATP-dependent acidification was observed in both cases; the apparent ApH across the lysosomal membrane was 0.65 and 0.70 pH units by the centrifugation and rapid gel filtration method, respectively. This good agreement between the two methods supports the validity of the rapid fitration technique.
When disrupted lysosome preparations were incubated with ['4C]methylamine in the absence of ATP, considerably less methylamine uptake was seen than with intact lysosomes (Table VII). Less uptake is to be expected because the internal volumes of membrane vesicles are nearly 10 times smaller than those of intact lysosomes. However, ATP-dependent acidification of the vesicles occurred provided that the preparations were stored with proteinase inhibitor (Table VII). Furthermore, the magnitudes of the proton gradients in the disrupted lysosomal preparations were equal to those of intact preparations. Acidification occurred in disrupted preparations despite the possibility of increased leakiness to protons and the possibility of outside-in, as well as outside-out, orientation

ATP-dependent acidification of lysosome preparations measured
by rapid gel filtration L-fraction was prepared from normal rat liver without use of Triton WR-1339 by differential centrifugation as described (48) and was 15fold enriched in N-acetyl-P-D-glucosaminidase relative to homogenate. Lysosomes containing Triton WR-1339 were isolated as described under "Materials and Methods." Preparations were frozen and stored either in sucrose as isolated or with Tris, free base, and aprotinin added to give 2 mM and 1.7 p~, respectively. of Lysosomes by ATP of the presumed proton pump. An absolute requirement for proteinase inhibitor was observed. Aprotinin (shown), soybean trypsin inhibitor, and p-aminobenzamidine were all effective; phenylmethylsulfonyl fluoride and N-ethylmaleimide were not (not shown). It is interesting to note that disrupted lysosomes in the L-fraction preparation displayed a greater capacity for methylamine uptake than those in the Triton WR-1339 preparation. First, this greater uptake may indicate that Triton can cause membrane damage and leakiness to protons. Second, and perhaps more significant, the contaminants in the L-fraction preparation, mitochondria, peroxisomes, and endoplasmic reticulum, may act as substrates for the proteinases in the preparation and thereby spare the lysosomal membrane proteins. These data are the first demonstration of acidification of disrupted lysosomes and may have significant implications for the relative roles of membrane pumps and Donnan effectors, as will be discussed below.

DISCUSSION
The results clearly show an ATP-dependent lowering of the intralysosomal pH by 1 pH unit (Fig. lb) and thus substantiate preliminary communications (27-30). Furthermore, the kinetics of this in vitro acidification are in good agreement with those observed in leukocyte lysosomes after phagocytosis of dye-impregnated yeasts (18, 19). It is important to point out that methylamine uptake by lysosomes is a measure of their overall proton gradient and reflects both the rate of proton pumping and leaking. The pH lowering to about 5.5 is not as great as one would expect on the basis of measurements in living cells which indicate an intralysosomal pH of 4.5 (14,18). This probably can be attributed to leakiness because Triton WR-1339 treatment, which was used in these experiments, is known to result in an increased permeability of lysosomal membranes to protons (42). ATP-dependent acidification of L-fraction lysosomes was greater than that of Triton lysosomes (Table VII); in fact, since L-fraction preparations are only about 20% pure, the actual lysosomal ApH might be signifkantly greater and correspond to an intralysosomal pH of 4.5.
ATP-dependent acidification suggests the presence of a proton pump in the lysosomal membrane which may be related to its known ATPase activity (35). Consistent with this suggestion are the similar and coordinate effects of salt on both ATPase and acidification activities (cf Table 11). The possibility that these activities are due to the same protein is further suggested from experiments in which DCCD completely inhibited ATP-dependent acidification while it inhibited ATPase activity 60%. In mitochondria, proton pump activity is also more sensitive to inhibition by DCCD than is ATPase activity: 1 nmol of DCCD/mg of submitochondrial particle protein completely inhibited energy-linked (pump) activity (51) but inhibited ATPase activity about 25% (52). The purported ATP-driven proton pump in lysosomes appears to operate electroneutrally, perhaps in conjunction with anion transport, because acidification is inhibited by the inhibitor of anion transport, diisothiocyanostilbene disulfonic acid (Table 111). Also, measurement of ["Clthiocyanate uptake (Fig. 2) shows that ATP does not significantly affect the membrane potential of lysosomes during the course of acidification. An electroneutral proton pump could be either due to a proton-cation exchange process or due to a proton-anion cotransport process. Although it is not yet possible to say with certainty which mechanism is operating, proton-anion cOtransport would agree well with inhibition by diisothiocyanostilbene disulfonic acid. It is uncertain whether the lysosomal membrane contains only one nucleoside triphosphatase (50) or multiple nucleoside triphosphatases, only one of which is a proton pump. Multiple phosphatases are certainly present in the lysosomal membrane because 5'-nucleotidase activity (43) and ATPase activity (30) are both known to be present. Purification of the ATPase(s) would provide useful information in this regard and is being pursued.
Inhibition of intralysosomal acidification by ATP in the presence of valinomycin plus potassium chloride would be expected if membrane potential were of critical importance in regulating intralysosomal pH; the slight stimulation observed indicates that membrane potential is not regulatory (Table  IV). Experiments with resealed lysosomal membranes indicate, on the other hand, that transmembrane pH is important in regulating ATPase activity (30) and is maximal when the external pH is slightly alkaline ( Table V). The inhibition of acidification by C-CCP (Table IV) shows that ATP generates a proton gradient and is notable: the effect of C-CCP is not consistent with a Donnan equilibrium acting as the proton motive force, because a proton ionophore should facilitate, rather than inhibit, a Donnan equilibrium. Also, it is interesting to note that C-CCP stimulates ATPase activity (29).
It is necessary to point out that the determination of intralysosomal pH is dependent on measuring both internal volume and methylamine uptake. Although this represents a potential source of error, the effects of ATP on the internal volumes of lysosomes are interestingly slight, lysosomes suspended in sucrose alone being an exception. It seems quite possible that one difference between the results presented here and those reported by Hollemans and co-workers (44) is the determination of internal volume. Although ['4C]sucrose and ["HIwater were used in both cases, the absolute sucrose concentrations were very different, as these results were obtained with at least 200 mM sucrose throughout and Hollemans and co-workers (44) used submicromolar concentrations. Their results are possibly compromised by absorptive binding of sucrose. We have observed that lysosomes in 0.16 WM sucrose do bind sucrose and that the binding is diminished by addition of ATP. Another reason for the discrepancy between the results may be the quality of the lysosome preparation.
At present, one suspects that the major difference between the ATP-dependent acidification reported here and the lack of acidification observed by others (44) is due to proteolytic damage that occurred when their preparations were centrifuged to form a pellet, as a means of concentrating the lysosomes after flotation gradient centrifugation. ATP stimulation of methylamine uptake by lysosomes does decrease as the isolated lysosomes are stored on ice. Our measurements of acidification by intact lysosomes were made within 4 h after isolation. In addition, the lysosomes were used as isolated in hyperosmotic sucrose (14.3 to 34.5% sucrose interface, about 0.75 M) because centrifugation to form a pellet of concentrated lysosomes and resuspension in isoosmotic medium significantly affects membrane integrity as indicated by changes in the latency of N-acetylglucosaminidase. Lysosomes disrupted by freezing and storing overnight at -20 "C were active in ATP-dependent acidification, but only if proteinase inhibitors were employed (Table VII). Inasmuch as acidification of disrupted preparations was as great as that of intact preparations, disruption by freeze-thawing apparently causes neither greatly increased leakiness to protons nor a significant number of outside-in vesicles. ATP-dependent acidification of disrupted lysosomes is significant; if the Donnan effectors were released, then acidification must be due to ATP-driven ion movement across the lysosomal membrane. Presumably the Donnan effectors are the soluble lysosomal enzymes, and enzyme latency should be an accurate measure oftheir release. Disrupted preparations showed 100-93% loss oflatency, which is indicative of nearly complete release. Experiments on acid-ification of lysosome ghosts separated from the soluble enzymes are in progress and will provide additional information on the role of ATP-driven ion movements that result in intralysosomal acidification.
The evidence reported here for the possibility of an ATPdriven proton pump in lysosomes agrees with a number of observations. Lysosomal proteinases are known to require an acid pH for activity (45), and external ATP stimulates the degradation of intralysosomal albumin in vitro (25). Protein degradation is energy-dependent (15-17), and inhibitors of energy metabolism raise the intralysosomal pH of living cells (14). Permeant weak bases like the antimalarial drug chloroquine are accumulated in lysosomes of cells to such an extent that active processes seem likely to be involved (13,36). Also, the acidification properties of lysosomes are similar to those of chromaffin granules in which a proton pump is generally accepted (37,39,40,46). Acidification of the chromaffin granules requires about 20 min, which is in good agreement with the data reported here for lysosomes. However, the systems differ in that the chromaffin granule proton pump appears to be electrogenic (47), whereas the lysosomal proton pump appears to be electroneutral.
Although future experiments are required to evaluate more quantitatively the relative roles of ATP-driven proton pump and Donnan effectors in acidification of lysosomes, at present it seems very reasonable to conclude from these experiments on intact and disrupted lysosomes that a proton pump mechanism may be operable.