THE JOURNAL OF BIOLOGICAL CHEMISTRY c 1991 by The American Society for Biochemistry and Molecular Biology. Inc

Vol. 266, No. 26. Issue of September 1.5, PP. 17707-17712, 1991 Printed in U. S. A.

Bafilomycin AI, a Specific Inhibitor of Vacuolar-type H+-ATPase, Inhibits Acidification and Protein Degradation in Lysosomes of Cultured Cells*

(Received for publication, April 15, 1991)

Tamotsu Yoshimori, Akitsugu Yamamoto, Yoshinori MoriyamaS, Masamitsu FutaiS, and Yutaka TashiroQ From the Department of Physiology, Kansai Medical University, Fumizono-cho, Moriguchi, Osaka 570, Japan and the $Department of Organic Chemistry and Biochemistry, Institute of Scientific and Industrial Research, Osaka Uniuersity, Ibaraki, Osaka 567, Japan

Bafilomycin AI is known as a strong inhibitor of the vacuolar type H+-ATPase in vitro, whereas other type ATPases, e.g. F1,Fo-ATPase, are not affected by this antibiotic (Bowman, E. M., Siebers, A., and Altendorf, K. (1988) Proc. Natl. Acad. Sci. U. S. A. 85, 7972- 7976). Effects of this inhibitor on lysosomes of living cultured cells were tested. The acidification of lyso- somes revealed by the incubation with acridine orange was completely inhibited when BNL CL.2 and A431 cells were treated with 0.1-1 PM bafilomycin A1. The effect was reversed by washing the cells. Both studies using 3-(2,4-dinitroanilino)-3’-amino-N-methyldipro- pylamine and fluorescein isothiocyanate-dextran showed that the intralysosomal pH of A431 cells in- creased from about 5.1-5.5 to about 6.3 in the presence of 1 PM bafilomycin A1. The pH increased gradually in about 50 min. In the presence of 1 PM bafilomycin A1, ‘“SI-labeled epidermal growth factor (EGF) bound to the cell surface at 4 “C was internalized normally into the cells at 37 “C but was not degraded at all, in marked contrast to the rapid degradation of lZ5I-EGF in the control cells without the drug. Immunogold electron microscopy showed that EGF was transported into ly- sosomes irrespective of the addition of bafilomycin A1. These results suggest that the vacuolar type H+-ATP- ase plays a pivotal role in acidification and protein degradation in the lysosomes in uiuo.

The central vacuolar system consists of the membrane organelles that belong to the exo- and endocytic pathways: the endoplasmic reticulum, the Golgi apparatus including the trans-Golgi network, secretory granules, secretory vesicles, endosomes, lysosomes, and the plasma membrane (Klausner, 1989). These multiple compartments communicate with each other by shuttling transport vesicles and are involved in biosynthesis, processing, sorting, transport, and degradation of proteins and other macromolecules. The internal acidifi- cation of some compartments in the central vacuolar system is suggested to play an important role in such processes (for review, see Mellman et al. (1986)).

* This work was supported in part by a grant-in-aid for Scientific Research from the Ministry of Education, Science, and Culture of Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

I To whom correspondence should be addressed: Dept. of Physi- ology, Kansai Medical University, Fumizono-cho 1, Moriguchi, Osaka 570, Japan.

Recently, proton-translocating ATPases have been identi- fied as a new class of H’-ATPase and referred to as the vacuolar type H’-ATPase (V-ATPases)’ (Nelson and Taiz, 1989). V-ATPases have been identified and purified from organelles belonging to the central vacuolar system, such as coated vesicles (Xie et al., 1986; Arai et al., 1987), chromaffin granules (Moriyama and Nelson, 1987), the Golgi complex (Moriyama and Nelson, 1989a), lysosomes (Moriyama and Nelson, 1989b), neurosecretory granules (Moriyama and Fu- tai, 1990), and fungal vacuoles (Uchida et al., 1985; Bowman and Bowman, 1986) and have been characterized extensively at the molecular level in recent several years. At present, V- ATPases are the most probable candidate for the generator of the acidity in the central vacuolar system. . V-ATPases show distinct sensitivities against several in-

hibitors when compared to F,,Fo-ATPase in the mitochondria and El,E2-ATPases in the plasma membranes. The V-AT- Pases are sensitive to N-ethylmaleimide but insensitive to oligomycin and azide, which are inhibitors of F,,Fo-ATPase, and to vanadate, which is an inhibitor of El,E2-ATPases. Moreover, recently, bafilomycin A1, a macrolide antibiotic isolated from Streptomyces sp. (Werner et al., 1984), was found to be a highly specific inhibitor of the V-ATPase (Bowman et al., 1988). This inhibitor is effective at nanomolar concentra- tions in vitro and inhibits neither Fl,Fo- nor El,E2-ATPases (Bowman et al., 1988; Hanada et al., 1990).

If bafilomycin A, can permeate into the organelles of living cells through the plasma membranes without killing the cells, one could directly examine in vivo relationships between the V-ATPase, the acidification of the organelles, and the various functions of the central vacuolar system. In this study, we examined the effect of bafilomycin A, on living cultured cells and found that bafilomycin AI can effectively inhibit in viuo acidification of lysosomes and the degradation of an endocy- tosed protein.

EXPERIMENTAL PROCEDURES

Materials-The following materials were used: bafilomycin A,, kindly given by Prof. K. Altendorf (Universitat Osnabruck, Ger- many); acridine orange (Chroma, Germany); 3-(2,4-dinitroanilino)- 3’-amino-N-methyldipropylamine (DAMP) (Oxford Biochemical Re- search, Oxford, MI); fluorescein isothiocyanate (F1TC)-conjugated dextran (average molecular weight, 71,200; Sigma); serum-free com-

The abbreviations used are: V-ATPase, vacuolar type H+-ATP- ase; DAMP, 3-(2,4-dinitroanilino)-3’-amino-N-methyldipropyla- mine; FITC, fluorescein isothiocyanate; EGF, epidermal growth fac- tor; BSA, bovine serum albumin; MEM, minimal essential medium; DMEM, Dulbecco’s modified Eagle’s medium; FCS, fetal calf serum; Me,SO, dimethyl sulfoxide; PBS, phosphate-buffered saline.

17707

17708 Effects of Bafilomycin on Lysosomal Functions plete medium (COSMEDIUM, Cosmo-Bio, Japan); mouse epidermal growth factor (EGF) (Wako, Japan); '2'II-labeled EGF (Amersham, Japan); rabbit antiserum against dinitrophenol-conjugated bovine serum albumin (BSA) (Seikagaku Kogyo, Japan); mouse monoclonal antibody Ab-1 against human EGF (Oncogene Science, Inc., Man- hasset, NY); sheep antibody against human cathepsin B (Binding Site, Ltd., Birmingham, England); rabbit antibody against mouse IgG (Dakopatts, Denmark); rabbit antibody against sheep IgG (Cappel, West Chester, PA).

Cell Culture-A431, human epidermoid carcinoma cell line, and BNL CL.2, mouse normal embryonic liver cell line, obtained from the American Type Culture Collection (Rockville, MD), were cultured in FCS/DMEM (Dulbecco's modified Eagle medium (DMEM) con- taining 12% fetal calf serum (FCS)).

Vital Fluorescence Microscopy-The living cultured cells were stained with acridine orange following the method previously de- scribed (Geisow et al., 1980). The cells grown on cover slips were incubated a t 37 "C for 1 h with FCS/DMEM containing 1/100 volume of bafilomycin A, dissolved in dimethyl sulfoxide (Me,SO). For con- trol experiments, only Me2S0 was added to the medium instead of the bafilomycin A, solution. Then the cells were incubated a t 37 "C for 10 min with 5 pg/ml acridine orange in Hanks' solution. After four washings with Hanks' solution, the coverslips were examined by a fluorescence light microscopy (Olympus BH-2). Fluorescence mi- crographs were taken with the same shutter speed for each experiment by using Kodak Ektachrome film.

Measurement of Zntralysosomal pH-The intralysosomal pH of A431 cells was measured by the two methods. The first is a DAMP method according to Orci et al. (1986). Briefly, A431 cells grown in 96-well culture plates were incubated at 37 "C for 1 h with 1 p~ bafilomycin A, in FCS/DMEM or with only MezSO in the same medium. Then, DAMP was added a t a final concentration of 50 p ~ , and the cells were incubated at 37 "C for 1 more h. After three washings with phosphate-buffered saline (PBS), the cells were fixed, and the localization and quantity of DAMP were determined by immunogold electron microscopy as described below. We estimated pH according to the formula of Orci et al. (1986); pH = 7.0 - log I/ N , where N is the density of DAMP-specific gold particles in a pH 7.0 compartment (the nucleus) and L is the density of DAMP-specific gold particles in lysosomes.

Second is a FITC-conjugated dextrans method devised by Ohkuma and Poole (1978). A431 cells grown on cover slips were exposed to 1 mg/ml FITC-conjugated dextran (previously dialyzed against PBS) for 24 h at 37 "C in the culture medium, washed with FCS/DMEM twice and incubated a t 37 "C for 1 h with 1 p~ bafilomycin A, in the same medium or for 10 min with other drugs. After washing the cells with PBS, fluorescence emission intensity, a t 510 nm, of the cells in PBS was measured with excitation a t 440 and 490 nm using a microspectrofluorometer (IMT-2-OSP, Olympus), the detailed com- position of which was described elsewhere (Tsunoda et al., 1989). The p H of any subcellular regions could be analyzed by using a pinhole unit (5 pm in diameter), a photomultiplier, and an Olympus UVFL X 100 objective lens. Backgrounds of cellular autofluorescence at each excitation wavelength (Zb440 and Zb490) measured in the cells contain- ing no FITC-dextrans were subtracted from the data (Z440 and 14w). The intralysosomal pH was calculated from a standard curve relating the ratio (Z440 - Zb440)/(Z490 - Zb490) with pH. The standard curve was generated using 220 pg/ml FITC-dextrans in 0.1 M Tris-HC1 or sodium acetate buffers a t various pH values. Fluorescence micro- graphs were taken by an Olympus BH-2. For the time course experi- ment, A431 cells were incubated with FITC-dextrans, and the change in the intralysosomal pH was investigated in serum-free complete medium COSMEDIUM by a Hitachi F-2000 fluorospectrophotometer according to the procedure of Ohkuma and Poole (1978) (fluorescence emission a t 519 nm with excitation a t 450 and 495 nm).

Measurement of Intracellular Degradation of 125Z-EGF--Intracel- lular degradation of '2'II-labeled EGF was measured by essentially the same method as described previously by Glenney et al. (1988). BNL CL.2 and A431 cells grown in 6-cm diameter dishes were washed twice with minimal essential medium containing 0.1% BSA (BSA/ MEM). After preincubation in BSA/MEM at 37 "C for 20 min, cells were incubated in the same medium containing 1 pCi (for BNL CL.2 cells) or 0.5 pCi (for A431 cells) of "'I-EGF a t 4 "C for 2 h. After three washings with BSA/MEM, cells were treated with 1 p~ bafi- lomycin A, in COSMEDIUM containing 0.1% bovine serum albumin or with only Me2S0 in the same medium at 37 "C. Aliquots of medium collected from dishes at the indicated times were precipitated with trichloroacetic acid a t a final concentration of 8.3%, placed in ice for

20 min, and then centrifuged to separate acid-soluble and -insoluble materials. Cells were solubilized with 1% Triton X-100, centrifuged to remove insoluble material, and then precipitated with trichloroa- cetic acid. Radioactivities of each sample were measured by a Packard MINAXI y auto-y counter. The values were indicated as percent of total counts (the sum of values from trichloroacetic acid-soluble and -insoluble fractions of cells and medium at the end of the incubation and Triton-insoluble fractions of cells; about 80,000 cpm for BNL CL.2 cells and about 400,000 cpm for A431 cells). The trichloroacetic acid-soluble counts of medium were regarded as degraded '*'I-EGF after correction by subtraction of the time 0 value (0.4-0.5%).

Measurement of Internalization of '2'Z-EGF-BNL CL.2 cells grown in 3.5-cm plastic dishes were treated with 0.25 pCi of "'I-EGF as in the degradation experiment described above. Internalization of cell surface-bound "'I-EGF into the cells was measured in the pres- ence of 1 PM bafilomycin A, or the absence (adding only Me,SO), as previously described (Haigler et al., 1980; Glenney et al., 1988); "'I- EGF that was bound to the cell surface was removed with 5 mM acetic acid containing 0.15 M NaCl, and the cells were then lysed with 1 M NaOH.

Zmmuno-gold Electron Microscopy-The DAMP-treated cells were fixed for 2 h with 1% glutaraldehyde in 0.1 M sodium phosphate buffer (pH 7.4). The fixed cells were embedded in Lowicryl K4M (Roth et al., 1981) together with the plastic dishes (96 wells), and the plastic dishes were removed after hardening the resin. Ultrathin sections of the cells were cut parallel to the bottom of the dishes. The ultrastructural localization and quantity of DAMP in these cells was investigated by a postembedding protein A-gold technique, as de- scribed elsewhere (Yoshimori et al., 1990), using rabbit anti-dinitro- phenol conjugated to BSA antiserum (diluted 1:50, v/v).

In order to localize the endocytosed EGF, the A431 cells were preincubated as described above, treated with BSA/MEM containing 164 nM EGF at 4 "C for 2 h, and then incubated a t 37 "C for 5 or 30 min in the absence (only Me2SO) or the presence of 1 p~ bafilomycin A,. The cells were fixed for 10 min with 4% formaldehyde and 1% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4) at the end of the incubation. Embedding and ultrathin sectioning was carried out as in the DAMP experiment. The ultrastructural localization of endo- cytosed EGF in these ultrathin sections was investigated using mouse monoclonal anti-EGF antibody (20 pg/ml) and rabbit anti-mouse IgG (50 mg/ml). For double staining, after sequential reaction with the anti-EGF antibody, the anti-mouse IgG, and protein A-gold (4-nm gold particles), the remaining free binding sites of protein A were blocked with 10 pg/ml protein A in PBS containing 0.5% BSA a t room temperature for 10 min. Then the ultrathin sections were incubated with sheep anti-human cathepsin B antibody (50 pg/ml), rabbit anti-sheep IgG (50 pg/ml), and protein A-gold (8-nm gold particles), successively.

RESULTS

Bafilomycin A1 Inhibits Acidification of Lysosomes-Effects of bafilomycin A1 on acidification of lysosomes were tested by vital staining with acridine orange. This dye is an "acido- tropic" weak base, which is taken up by living cells and accumulates in acidified compartments such as lysosomes (Allison and Young, 1969; Holtzman, 1989). Fluorescence of acridine orange is green at low concentrations, whereas at high concentrations the fluorescence changes to orange (Ai- lison and Young, 1969). When A431 and BNL CL.2 cells were stained with acridine orange, nuclei, particularly the nucleoli and the cytoplasms, showed green fluorescence, whereas or- ange fluorescence was observed in a granular pattern in the cytoplasm (Fig. 1, panels I and 3 ) . This distribution pattern suggests that the orange fluorescence is due to acidified lyso- somes. Treatment of both cells with 1 PM bafilomycin AI at 37 "C for 1 h before acridine orange-staining caused complete disappearance of the orange fluorescence, whereas the green fluorescence remained as such as shown in Fig. 1, panel 2 (A431 cells). (A corresponding micrograph for BNL CL.2 cells was not shown.) Bafilomycin A, a t lower concentrations was also effective. 10 nM bafilomycin A1 markedly decreased or- ange fluorescence in BNL CL.2 cells (panel 4 ) , and at a final concentration of 100 nM, the fluorescence in these cells com-

FIG. 1. Acridine orange staining of bafilomycin Al-treated cells. Liv- ing A431 (panels 1 and 2) and BNL CL.2 (panels 3-6) cells were stained with ac- ridine orange after treatment of MeZSO (control, panels I and 3) or bafilomycin A, in MezSO at final concentrations of 1 p M ( p a n e l 2), 10 nM (panel 4) , or 100 nM (panel 5) for 1 h. Panel 6 shows the staining pattern after treatment with 100 nM bafilomycin A, for 1 h followed by washing and incubation without baf- ilomycin AI for 30 min. Bur, 30 pm.

Effects of Bafilomycin on Lysosomal Functions 17709

P . T"$ H :i'

pletely disappeared (panel 5 ) as at 1 pM (panel 2). After washing, the acidification of lysosomes in BNL CL.2 cells recovered completely, as shown in panel 6; the cells were treated with 100 nh4 bafilomycin AI for 1 h, washed, and then incubated for 30 min without the drug. These results suggested that bafilomycin AI can permeate through the plasma mem- branes and inhibit the acidification of lysosomes reversibly by interacting with vacuolar type H+-ATPase in the lysosomal membranes.

Increase in Intralysosomal pH by Bafilomycin A,-The in- crease in intralysosomal pH by 1 p~ bafiiomycin A, in A431 cells was also estimated by the DAMP method. DAMP is, like acridine orange, a weak base and accumulates in acidic com- partments. Moreover, this molecule is retained at the site of accumulation after aldehyde fixation and is detectable by immuno-gold electron microscopy because it contains a dini- trophenol group for easy detection with an anti-dinitrophenol antibody (for review, see Anderson and Orci (1988)). Accu- mulation of DAMP-specific gold particles was observed in lysosome-like multivesicular structures of DAMP-treated A431 cells (Fig. 2, left). Treatment of the cells with 1 PM

FIG. 2. Ultrastructural localization of DAMP in baf'ilomy- cin AI-treated cells. A431 cells were incubated at 37 "C for 1 h in the absence (only Me,SO) (left panel) or the presence of 1 p~ bafilomycin A, (right panel). 50 pM DAMP was then added to the cells, which were incubated at 37 "C for 1 more h. The cells were fixed, embedded in Lowicryl K4M, and processed for protein A-gold localization of DAMP. Accumulation of DAMP-specific gold particles (8-nm diameter) that were numerously observed in the lysosome-like multivesicular bodies in the control cells (left) was inhibited by the treatment with bafilomycin A, (right). Bar, 0.3 pm. Magnification, X29,400.

bafilomycin A, for 1 h decreased markedly the number of gold particles in these compartments (Fig. 2, right). Table I shows that the density of DAMP-specific gold particles in lysosomes

17710 Effects of Rafilomycin on Lysosomal Functions

ntonhwlprn' Control

Lysosomes 4 1 .?l 5 5 Nucleus 1.9

Lvsosomes 8.4 6.3 Nucleus 1.7

Ratilomycin-treated

"Density of gold particles in the lysosomes and nucleus was cor- rected for nonspecific binding measured on the portion o f the u l t r a - thin sections without cell profile.

F I ~ ; . :\. Fluorescen t pa t t e rns of pinocytosed FI ' I 'C-dextrans. A431 cells were exposeed \ v i t h 1 mg/ml I'I'lY-dextrans a t :K "C for 24 h. Fluorescent micrographs after incuhation of these cells at 37 "C for 1 h in the absence (only Me&O) (lrft pnnel) or the presence of 1 p~ hafilomycin A, (right p o n d ) show a punctate lysosome-like pat- tern. Hnr. 40 pm.

Addition Rat io" oH

No addition 1.49 f 0.4Bh (67)" 5.14 Hafilonlycin AI ( 1 p M ) 4.71 -C 0.9.5 (66) 6.30 Nigerycin ($5 pg/ml) 4.75 -e 0.96 (101) 6.33 Ammonium chloride (10 mM) 3 5 2 k 0.76 (79) 5.92 Chloroquine (100 pM) 4.58 k 0.95 (77) 6.26 Methylamine (10 mM) 3.84 c 10.6 (87) 5.98

j ' (I,'!,, - I h , ~ ~ ~ l ~ / ~ I . l , ~ ~ - I h d . " Mean f S.D. ' Numher of cells measured.

of the bafilomycin-treated cells was about one-fifth that in lysosomes of control cells, whereas the density in neutral pH compartments such as the nucleus was not affected by the drug treatment. From the DAMP-specific gold particle den- sity, we estimated that the intralysosomal pH of A431 cells increased from 5.5 to 6.3 by 1 p~ bafilomycin AI (for the formula, see "Experimental Procedures").

This result was confirmed by measuring the fluorescence emission of fluid phase pinocytosed FITC-dextrans according to the procedure of Ohkuma and Poole (1978). Incubation of A431 cells with FITC-dextrans for 24 h resulted in accumu- lation of this fluorescence probe within the cells as a punctate lysosome-like pattern (Fig. 3, l e f t ) , and their fluorescence intensity was increased by the treatment with 1 p~ bafilo- mycin A, for 1 h (Fig. 3, right). It has been reported that the fluorescence intensity of FITC increases according to the rise in surrounding pH (Ohkuma and Poole, 1978). Next, the ratios in fluorescence emission intensities a t 510 nm following excitation at 490 and 440 nm were measured by microspectro- fluorometry (Table 11). Intralysosomal pH was calculated as 5.14 in control A431 cells and 6.30 in 1 pM bafilomycin- treated cells from the standard curve. The degree of the increase in pH by 1 p~ bafilomycin AI was similar to that by the other various agents, such as acidic ionophore and weak bases, which are known to increase in intralysosomal pH.

The time course of increase in intralysosomal pH by 1 p~ bafilomycin Ai in A431 cells labeled with FITC-dextrans was monitored by the spectrofluorometer. The pH gradually in- creased and reached maximum value (about 6.0) after 50 min (Fig. 4). The intralysosomal pH of BNL CL.2 cells also increased in a similar way after the addition of 1 pM bafilo- mycin Ai (data not shown).

Rafilomycin A i Abolishes Cellular Degradation of "."I-EGF- Epidermal growth factor is internalized into cells through receptor-mediated endocytosis, then degraded into low molec- ular weight products in lysosomes (Carpenter, 1987). It is interesting to see the effect of bafilomycin A,, because acidic pH of lysosomes may be important for the processes of deg- radation of EGF. Fig. 5 shows the effects of bafilomycin AI on degradation of ""I-EGF in the BNL CL.2 and A431 cells. After incubation at 4 "C with ""I-EGF, both cells degraded about 80% of the cell surface-bound '"'I-EGF a t 37 "C and

_.. .

5.75 -

5.50 -

4.500 1111111..I 10 20 30 40 50 60 70 80 90

Time ( m i d

FIG. 4. T i m e c o u r s e of increase i n i n t r a l y s o s o m a l pH b y baf i lomycin A,. Change in the fluorescence emission intensity at 519 nm of the A431 cells containing FITC-dextrans as in Fig. 3 was monitored with excit.ation a t 450 and 495 nm. After 5 min, bafilo- mvcin A, was added at a final concentration of 1 p M (nrrorcr). Values were corrected for autoluminescence of the cells measured in the cells not exposed to FITC-dextran, and pH was calculated from the stand- ard curve.

2 * o p " p 7 1 60 .

A A 1 - 0 2 3 4

Time (hr)

FIG. :i. I n h i b i t i o n o f the i n t r a c e l l u l a r d e g r a d a t i o n o f "'1- EGF w i t h b a f i l o m y c i n Ai. After preincuhation in the serum-free medium, HNI, CL.2 (pnnrl A ) and A431 (pond H ) cells were incu- hated with ""I-EGF at 4 "C for 2 h. Then the unhound ligand was removed by several washings, and the cells were incuhated at :l7 "C for the indicated time in the ahsence (only Me,SO) (0) or the presence of 1 p~ bafilomvcin A, (0). Degradation of '"'I-EGF internalized into the cells was measured as described under "Experimental Proce- dures." Values were corrected by suhtracting the values at time 0 .

Effects of Bafilomycin on Lysosomal Functions 17711

released acid-soluble low molecular weight products to the medium within 4 h (Fig. 5, A and B, open circles). When 1 p~ bafilomycin A1 was added to the incubation medium, degra- dation of '"I-EGF was completely abolished (Fig. 5, A and B, closed circles). A431 cells are known to overexpress EGF receptor (Carpenter, 1987). In fact, A431 cells degraded about 10 times as much '*'I-EGF as BNL CL.2 cells. The inhibition by bafilomycin A1 was complete even in A431 cells, and almost 100% of the radioactivities associated with both cells were in acid-insoluble forms even after incubation for 4 h. Thus the inhibition of the degradation by bafilomycin A1 was not due to an inhibition of a release of degraded Iz5I-EGF.

Bafilomycin A, Does Not Inhibit Endocytosis of lZ5I-EGF- In order to know which step in the process of endocytosis and degradation of EGF was inhibited by bafilomycin Al, first, internalization of lZ5I-EGF into BNL CL.2 cells was chased under the presence of 1 p~ bafilomycin A1. As shown in Fig. 6, the cell surface-bound T - E G F was rapidly internalized into the cells independently whether bafilomycin A1 was added or not. The decrease in lZ5I-EGF within the cells, which appeared at 45 min in the absence of bafilomycin A,, may reflect the release of degraded lZ51-EGF.

Next, delivery of internalized EGF to lysosomes was inves- tigated by immuno-gold electron microscopy using mono- clonal anti-EGF antibody. When EGF was bound to the surface of A431 cells a t 4 "C, then internalized a t 37 "C for 30 min, gold particles were detected in lysosome-like multive- sicular bodies (Fig. 7A), whereas after incubation for 5 min at 37 "C, gold particles were distributed on the plasma mem- brane (data not shown). Similar results were obtained when 1 MM bafilomycin A, was added to the medium at the 37 "C incubation (Fig. 7B). This compartment in which EGF was localized was determined to be lysosomes by the double- staining method using anti-EGF and anti-cathepsin B anti- bodies (Fig. 7, C and D). Irrespective of the addition of 1 p~ bafilomycin A,, the endocytosed EGF was co-localized with cathepsin B, which is a representative lysosomal marker enzyme. Morphology of lysosomes was not altered by the bafilomycin A, treatment.

Thus, bafilomycin A, does not appear to affect the endocytic pathway of EGF to lysosomes. These results suggest strongly that the arrest of degradation of lZ51-EGF by bafilomycin A1 is closely related to the increase in the intralysosomal pH.

DISCUSSION

We have investigated effects of a specific V-ATPase inhib- itor, bafilomycin A1, on the acidification and digestive func- tion of lysosomes in uiuo. This drug strongly inhibited the

Time (mid

FIG. 6. Internalization of lZ5I-EGF into the cells. "'I-EGF was bound to the cell surface of BNL CL.2 cells at 4 "C and internal- ized into the cells at 37 "C as described in Fig. 5. "'I-EGF bound to the cell surface (circles) and "'I-EGF internalized into the cells (triangles) were measured in the absence (only Me,SO) (open symbols) or the presence (closed symbols) of 1 FM bafilomycin AI (see "Exper- imental Procedures").

acidification of lysosomes and increased the intralysosomal pH in the two cell lines, BNL CL.2 and A431 cells. Further- more, it abolished lysosomal degradation of endocytosed EGF completely, whereas the reagent did not affect internalization of EGF and its transport into the lysosomes. These results suggest strongly that proton-translocating activity of V-AT- Pases is actually essential for the lysosomal function, such as EGF digestions, through maintaining the luminal acidic en- vironment of the organelles.

The specificity and useful properties of bafilomycin A1 have been demonstrated mainly by i n vitro experiments (Bowman et al., 1988; Moriyama and Nelson, 1989a; Hanada et al., 1990; Moriyama and Futai, 1990). There have been only a few reports using the drug against living cells. Klionsky and Emr (1989) showed partial missorting of a vacuolar protein oc- curred after treatment with a high concentration (10 p M ) of bafilomycin AI. Bafilomycin Al obviously inhibited bone re- sorption by osteoclasts, although the effective site was not identified (Sundquist et al., 1990). Cell-killing action of diph- theria toxin on monkey kidney cells was inhibited by bafilo- mycin A1 at the step of toxin penetration from endosomes to cytosol (Umata et al., 1990). I t is now confirmed by the present results that bafilomycin A, is a useful inhibitor for the inves- tigation of V-ATPases in intracellular organelles in living cells.

Abnormal morphology of the cells and the organelles was not observed when bafilomycin A1 was used at a concentration of 1 p ~ . Intracellular protein transport, including the endo- cytotic pathway, appeared not to be inhibited by the bafilo- mycin A, treatment (Figs. 4 and 5). Furthermore, bafilomycin AI did not perturb bulk secretions of proteins from the two cell lines derived from rat hepatoma and pituitary, respec- tively.' These results gave credence to use of the reagent in living cells to inhibit specifically V-ATPases.

I t is notable that the increase in intralysosomal pH by bafilomycin A1 is limited to about 6.3. Addition of amines or ionophores also increases the pH to about 6.0-6.5 (Ohkuma and Poole (1978) and our results shown in Table 11). On the other hand, the degradation of EGF was completely inhibited by bafilomycin A,, in spite of the fact that several proteolytic enzymes in lysosomes have been reported to have optimal pH near 6 i n vitro (Barrett, 1972). One possibility is that EGF is a specific substrate for a particular hydrolase(s), which is inactive at pH 6.0. However, this possibility is unlikely be- cause overall degradation of endogenous proteins was also markedly inhibited by bafilomycin A,.' Perhaps the optimum pH for the enzymes in active lysosomes within living cells may differ from those assayed i n vitro as the isolated form, or there may be some regulation systems for the function of the lysosomal enzymes in lysosomes i n vivo, which are sensitive to small changes in intralysosomal pH.

The importance of acidification in the functions of lyso- somes and other organelles in the central vacuolar system has been suggested mainly based upon indirect evidence (for re- view, see Mellman et al. (1986)), using acidotropic weak bases (Holtzman, 1989), and no direct approaches to physiological roles of V-ATPases have been reported. We showed that the effect of bafilomycin A1 may be limited strictly to the inhibi- tion of V-ATPase activity. The complete abolishment of the degradation of EGF by the drug suggests that the acidification of the lysosomes is essential for the activation of lysosomal hydrolases. Now we have established that bafilomycin Al could be a very strong tool for studying the roles of V-ATPases in lysosomes within living cells.

'T. Yoshimori, A. Yamamoto, Y. Moriyama, M. Futai, and Y. Tashiro, manuscript in preparation.

17712 Effects of Bafilomycin on Lysosomal Functions

FIG. 7. Ultrastructural localiza- tion of ECF internalized into the cells. EGF was bound to and internal- ized into A431 cells as described in Fig. 5. After incubation at 37 “C for 30 min in the absence (only Me2SO) (panel A ) or the presence of 1 p~ bafilomycin A, (panel E ) , the cells were fixed and as- sayed by postembedding protein A-im- muno-gold electron microscopy using monoclonal anti-EGF antibody. Gold particles (8-nm diameter, indicated by arrowheads) representing the presence of EGF are localized on the multivesi- cular bodies in both panels A and B. Bar, 0.5 pm. Magnification, X25,500. Panels C and D show the results of the double- labeling experiments using the anti-EGF antibody and anti-cathepsin B antibody in the cells treated similarly to A and B, respectively. EGF-specific gold particles (4-nm diameter, indicated by arrow- heads) are co-localized with cathepsin B- specific gold particles (8-nm diameter, indicated by arrows) on the multivesi- cular bodies in both panels C and D. Bar, 0.3 pm. Magnification, X40,200.

I

Acknowledgment-We express our thanks to Dr. S. Yodozawa for assistance and advice with microspectrofluorometry.

REFERENCES

Allison, A. C., and Young, M. R. (1969) in Lysosomes in Biology and Pathology (Dingle, J. T., and Fell, H. B., eds) Vol. 2, pp. 600-628, North-Holland Publishing Company, Amsterdam

Anderson, R. G. W., and Orci, L. (1988) J. Cell Biol. 106,539-543 Arai, H., Berne, M., Terres, G., Terres, H., Puopolo, K., and Forgac,

M. (1987) Biochemistry 26,6632-6638 Barrett, A. J. (1972) in Lysosomes, A Laboratory Handbook (Dingle,

J. T., ed) pp. 46-135, North-Holland Publishing Company, Am- sterdam

Bowman, B. J., and Bowman, E. J. (1986) J . Membr. Biol. 94,83-97 Bowman, E. M., Siebers, A., and Altendorf, K. (1988) Proc. Natl.

Carpenter, G . (1987) Annu. Rev. Biochem. 56,881-914 Geisow, M. J., Beaven, G. H., D’Arcy Hart, P., Young, M. R. (1980)

Exp. Cell Res. 126, 159-165 Glenney, J. R., Jr., Chen, W. S., Lazar, C. S., Walton, G. M., Zokas,

L. M., Rosenfeld, M.G., and Gill, G. N. (1988) Cell 52,675-684 Haigler, H. T., Maxfield, F. R., Willingham, M. C., and Pastan, I.

(1980) J. Biol. Chem. 255, 1239-1241 Hanada, H., Moriyama, Y., Maeda, M., and Futai, M. (1990) Biochem.

Biophys. Res. Commun. 170,873-878 Holtzman, E. (1989) Lysosomes, pp. 95-100, Plenum Press, New York Klausner, R. D. (1989) Cell 57,703-706 Klionsky, D. J., and Emr, S. D. (1989) EMBO J. 8,2241-2250

Acad. Sci. U. S. A. 85,7972-7976

Mellman, I., Fuchs, R., and Helenius, A. (1986) Annu. Reo. Biochem.

Moriyama, Y., and Futai, M. (1990) J. Biol. Chem. 265,9165-9169 Moriyama, Y., and Nelson, N. (1987) J. Biol. Chem. 262,9175-9180 Moriyama, Y., and Nelson, N. (1989a) J. Biol. Chem. 264, 18445-

Moriyama, Y., and Nelson, N. (1989b) Biochim. Biophys. Acta 980,

Nelson, N., and Taiz, L. (1989) Trends Biochem. Sci. 14,113-116 Ohkuma, S., and Poole, B. (1978) Proc. Natl. Acud. Sci. U. S. A. 75,

Orci, L., Ravazzola, M., Amherdt, M., Madsen, O., Perrelet, A., Vassalli, J.-D., and Anderson, R. G. W. (1986) J. Biol. Chem. 103,

Roth, J., Bendavan, M., Carlemalm, E., Villiger, W., and Caravito,

Sundquist, K., Lakkakorpi, P., Wallmark, B., and Vaananen, K.

Tsunoda, Y., Yodozawa, S., and Tashiro, Y. (1989) J. Histochem.

Uchida, E., Ohsumi, Y., and Anraku, Y. (1985) J. Biol. Chem. 260,

Umata, T., Moriyama, Y., Futai, M., and Mekada, E. (1990) J. Biol.

Xie, X.-S., and Stone, D. K. (1986) J. Biol. Chem. 261,2492-2495 Werner, G., Hagenmaier, H., Drautz, H., Baumgartner, A., and Zah-

Yoshimori, T., Semba, T., Takemoto, H., Akagi, S., Yamamoto, A.,

55,663-700

18450

241-247

3327-3331

2273-2281

M. (1981) J. Histochem. Cytochem. 29,663-671

(1990) Biochem. Biophys. Res. Commun. 168,309-313

Cytochem. 37,999-1005

1090-1095

Chem. 265,21940-21945

ner, H. (1984) J. Antibiot. (Tokyo) 37, 110-117

and Tashiro, Y. (1990) J. Biol. Chem. 265,15984-15990