- Email: [email protected]
Acidic Vesicles in Cultured Rat Hepatocytes
GASTROENTEROLOGY 1987;92:1251-61
Acidic Vesicles in Cultured Rat Hepatocytes Identification and Characterization of Their Relationship to Lysosomes and Other Storage Vesicles JOHN R. LAKE, REBECCA W. VAN DYKE, and BRUCE F. SCHARSCHMIDT The Department of Medicine and Liver Center, University of California, San Francisco, California
We and others recently have demonstrated adenosine triphosphate-dependent acidification in a variety of pre1ysosoma1 organelles isolated from liver including clathrin-coated vesicles, multivesicu1ar bodies, and Go1gi. Little is known, however, regarding the number or distribution of acidic compartments in intact hepatocytes. We therefore have utilized acridine orange, a fluorescent weak base, to study the number and distribution of acidic vesicles of rat hepatocytes in primary culture and compared these with the number and distribution of 1ysosomes and other storage vesicles. Hepatocytes were found to contain about 170 acidic compartments per cell Received August 20, 1986. Accepted September 26, 1986. Address requests for reprints to: John R. Lake, M.D., Gastrointestinal Research Unit, HSW 1120, University of California, San Francisco, California 94143. This work was supported in part by grants AM26270, AM26743, AM07453, and AM01254 from the National Institutes of Health, and grants from the Giannini Foundation, the American Liver Foundation, and the Walter C. Pew Fund for Gastrointestinal Research. This work was presented at the plenary session of the meeting of the American Association for the Study of Liver Disease, Chicago, Illinois, November 10-11,1984; and has been published previously in abstract form (Hepatology 1984;4:1070). The authors thank the University of California San Francisco Liver Center-supported (AM26743) Core Facilities, including the Cell Culture Facility (Dr. D. Montgomery Bissell, director) and the Electron Microscopy Facility (Dr. Albert 1. Jones, director). The authors also thank Jocelyn Matsumoto-Pan, Michael Wong, and Paul George for their technical support; Mark Malamud for his contribution in preparing the cultured hepatocytes, and Dr. Mary Barker for her assistance with the histochemical stain for acid phosphatase and electron microscopy. In addition, the authors thank Diana Fedorchak and Michael Karasik for their expert editorial assistance. © 1987 by the American Gastroenterological Association 0016-5085/871$3.50
by fluorescence microscopy. These vesicles were diffusely distributed throughout the cell cytoplasm, with about 50% in the perinuclear area by modified morphometry. The acridine orange staining of these vesicles was reversibly dissipated by monensin, NH4 C1, chloroquine, and primaquine, indicating these vesicles exhibit an acidic interior established by active proton transport. In addition, the cho1estatic agent chlorpromazine reversibly inhibited, in a dose-dependent fashion, the redevelopment of a pH gradient in the acidic vesicles after dissipation by monensin. The number and distribution of these acidic vesicles were not significantly different from the number and distribution of vesicles involved in the storage (up to 6 h after internalization) of the fluid phase marker fluorescein-dextran. By contrast, histochemically identifiable 1ysosomes were fewer in number and significantly more restricted in their distribution to the perinuclear area (89%) than either dextran-storing or acidic vesicles: Electron microscopic studies confirmed that endocytosed dextran as well as another fluid phase marker, colloidal gold, were found predominantly in acid phosphatase- and ary1su1fatase-negative vesicles for up to 6 h after internalization. These studies indicate that hepatoGytes contain numerous intracellular vesicles acidifed by an active H+ transport mechanism. Based on their comparative number and distribution, acidic vesicles probably include vesicles involved in fluid-phase endocytosis but only a minority are 1ysosomes. The findings also indicate that fluid-phase markers are stored predominantly in Abbreviations used in this paper: CPZ, chlorpromazine; FITCdextran, fluorescein isothiocyanate dextran.
1252
LAKE ET AL.
vesicles other than histochemically identifiable 1ysosomes for up to 6 h after internalization. Finally, this technique also affords the opportunity for studying the movement of such vesicles in a vital preparation. Recent studies by ourselves and others have demonstrated that a variety of intracellular organelles including clathrin-coated vesicles, multivesicular bodies, Golgi, and lysosomes are actively acidified via a proton-trans locating adenosine triphosphatase (1-4). The acidification of endocytic vesicles has been shown to be essential for the normal internalization and processing of ligands such as asialoglycoproteins (5), low-density lipoproteins (6), insulin (7), and epidermal growth factor (8), which are taken up into hepatocytes via receptor-mediated endocytosis. Agents such as protonophores and weak bases, which dissipate proton gradients responsible for vesicle acidification (1,9), prevent the normal dissociation of ligands from receptors (10), block the delivery of ligands to lysosomes (11), and interrupt the normal recycling of receptors back to the plasma membrane (12). In addition, vesicle acidification is important for the dissociation of iron from transferrin after endocytosis of the iron-transferrin complex (13), as well as the translocation of viruses (14) and diphtheria toxin from the vesicle interior to the cell cytosol (15). Whereas studies to date have focused primarily on the identification and characterization of the acidification mechanism in isolated organelles (1-3), less attention has been paid to the study of acidic compartments in situ. We therefore have developed a method for identifying acidic compartments via light microscopy in living cultured hepatocytes and have utilized this technique to (a) determine the number of intracellular acidic vesicles and the nature of the acidification mechanism and (b) compare the number and distribution of these acidic vesicles with the number and distribution of lysosomes and vesicles involved in the storage of fluid-phase markers.
Materials and Methods Reagents Fluorescein isothiocyanate dextran (FITC-dextran) (70,000 average mol wt), chloroquine (diphosphate salt), and primaquine (diphosphate salt) were obtained from Sigma Chemical Co., St. Louis, Mo. 6-Carboxyfluorescein diacetate was obtained from Molecular Probes, Inc., Junction City, Ore. Acridine orange was obtained from Eastman Kodak Chemicals, Rochester, N.Y., and monensin from Calbiochem-Behring, La Jolla, Calif. Cloroauric acid was purchased from ICN, Plainview, N.Y. Chlorpromazine was a gift from Dr. Carl Kaiser (Smith, Kline and French Laboratories, Philadelphia, Pa.).
GASTROENTEROLOGY Vol. 92, No.5. Part 1
Cultured Rat Hepatocytes Rat hepatocytes were prepared by collagenase perfusion of nonregenerating rat liver as previously described (16) and were plated on collagen-coated 35-mm plastic culture dishes at a density of 0.5-1.0 mg cell protein/dish. Cells were cultured at 37°C in the presence of 3% CO 2 for 48 h before use in media containing Hank's buffered salts and modified 199 amino acids supplemented with 1% rat serum, BME vitamins (10 mllL, Grand Island Biological Co., Grand Island, N.Y.), penicillin (100 U/ml), insulin (4 u/L), cortisosterone (1 pM), vitamin C (50 J.Lglml), and glucose (5.6 mM). Cell viability was tested by trypan blue exclusion as well as by release of lactate dehydrogenase into the incubation medium (17).
Acridine Orange Staining Two-day-old cultures of rat hepatocytes were incubated for 10 min in fresh media containing acridine orange at a concentration of 50 J.LM. Acridine orange-containing media was then removed and the cells were washed twice with 1 ml of a balanced electrolyte solution. After the second wash, a 25-mm circular glass coverslip was placed over the cells and the cells were visualized immediately under oil by transmission or fluorescence illumination using an Olympus fluorescence microscope (Olympus Corporation of America, New Hyde Park, N.Y.) equipped with a X40 oil immersion objective. Fluorescence visualization was accomplished using two Olympus BG-12 short-pass excitation filters, one neutral density excitation filter, and a single 0-530 long-pass emission filter. Cells were photographed during both transmission and fluorescence microscopy using an Olympus C-35 camera and 400 ASA Ektachrome film (Eastman Kodak). To minimize photo bleaching and photodamage, fluorescence photographs were taken after only a brief «5 s) visualization for identification of field and focusing.
Effects of Protonophores and Weak Bases on Acridine Orange Staining The effect of the proton-cation exchanging ionophore, monensin, was determined by exposing cells for 2 min to media containing 5 J.LM monensin followed by a 10-min incubation in media containing 50 J.LM acridine orange and 5 J.LM monensin. Cells were then visualized and photographed according to the protocol described previously. Reversibility of the monensin effect was assessed by incubating monensin-exposed cells in fresh, monensin-free media for 10 min followed by visualization after a second 10-min exposure to 50 J.LM acridine orange. The effects of the weak bases ammonium chloride (50 mM), chloroquine (0.5-2.0 mM), and primaquine (100-500 J.LM), as well as the reversibility of these effects, were determined in a similar fashion, the only difference being that the cells were exposed to the weak bases for 10 min before acridine orange exposure and visualization.
Effects of Chlorpromazine on Acridine Orange Staining The ability of chlorpromazine (CPZ) to inhibit development of a proton gradient across vesicle mem-
May 1987
branes was determined by incubating cells in media containing 10-100 pM CPZ for 5 min, followed by a 2-min exposure to monensin. Cells washed free of monensin were then visualized by fluorescence microscopy after a 10-min exposure to 50 pM acridine orange in CPZcontaining media and photographed. Reversibility of the CPZ effect was determined by incubating the CPZ-exposed cells for an additional 20 min in media free of CPZ and monensin, after which they were again exposed to 50 /LM acridine orange and visualized as described above.
Fluorescein Isothiocyanate-Dextran Uptake FITC-dextran (70,000 mol wt) was prepared before use by passage over a Sephadex G-75 (Pharmacia, Piscataway, N.J.) column (0.9 cm x 25 cm) to separate 70,000 mol wt FITC-dextran from free fluorescein and other lowmolecular-weight fluoresceinated contaminants. The void volume fraction was dialyzed against distilled water overnight at 4°C, lyophilized, and stored dessicated at 4°C until use. Cultured hepatocytes were incubated for 16 h (beginning 32 h after plating) in media containing 20 mg/ml FITC-dextran. Before visualization, hepatocytes were washed three times with buffer at 4°C, and reincubated in fresh FITC-dextran-free media at 37°C for 1, 2, 4, or 6 h. At the end of the chase period, cells were visualized and photographed using the same technique and filters used fbr acridine orange fluorescence.
Colloidal Gold Uptake Colloidal gold particles were prepared according to the method of Geoghegan and Ackerman (18). Cloroauric acid (0.05 g) was added to 500 ml of distilled H 2 0 and brought to a boil. The boiling solution was stirred vigorously while 12.5 ml of a 1% aqueous trisodium citrate solution was rapidly added. After a color change to orangered, the solution was cooled and 1.5 ml of a 1% polyethylene glycol (10,000 mol wt) solution was added to stabilize the suspension during storage at 4°C. This procedure reliably yields colloidal particles that have a crosssectional diameter by electron microscopy of 18-20 nm.
Histochemical Staining for Lysosomes Light microscopy. Lysosomes were identified at the light microscopic level in 48-h-old cultures using a naphthol AS-TR phosphate-hexazonium pararosanilin histochemical stain for acid phosphatase (19). Cell mono layers were first fixed by incubation for 30 min at room temperature in a solution containing 1% CaCl 2 and 4% formaldehyde, pH 7.0, followed by a 2-h incubation at 37°C in 140 mM sodium cacodylate, pH 7.3. Cells were then rinsed in a large volume of veronal acetate buffer [prepared by dissolving 9.71 g sodium acetate· 3H 2 0 and 14.71 g sodium barbiturate in 500 ml of CO 2 -free distilled water and further diluting this solution 1:3 (vol/vol) with distilled water) and incubated in the naphthol AS-TR phosphate-hexazonium pararosanilin solution for 90 min at 37°C. The naphthol AS-TR phosphatehexazonium pararosanilin solution was prepared by dilut-
ACIDIC VESICLES IN CULTURED HEPATOCYTES
1253
ing 5 ml of the veronal acetate buffer in 12 ml of distilled water and adding 1 ml of a 10-mg/ml solution of naphthol AS-TR in N,N-dimethyl-formamide to which was added 0.8 ml of the pararosanilin solution (1 g pararosanilin hydrochloride in 20 ml of distilled water and 5 ml of concentrated HCI). The cells were then thoroughly washed in large volumes of the sodium cacodylate buffer and air-dried for 10-20 min before visualization under oil by transmission illumination. In preliminary studies, the effects of variable fixation times (2-16 h) and staining times (0.5-3 h) were studied and the protocol (above) yielding the maximum number and definition of individual lysosomes was utilized. Electron microscopy. Histochemical staining for acid phosphatase was performed in cultured cells incubated overnight (16 h) in media containing either dextran (20 mg/ml) or colloidal gold (10.4 mg/ml), washed, and placed in fresh, marker-free media as described above. After 1, 2, 4, or 6 h, the medium was aspirated and the cells were washed and fixed in a solution of 2.5% glutaraldehyde and 0.8% paraformaldehyde in 0.13 M sodium bicarbonate buffer (pH 7.4) for 10 min at room temperature followed by gentle mixing for 1 h at 37°C in RobinsonKarnofsky media (0.1 M acetate buffer, pH 5.0, containing 1 mM (3-glycerophosphate, and 2 mM CeCl 3 and 5% sucrose) as previously described (20). After the cytochemical reaction, the cells were washed twice in the above media and further incubated in the glutaraldehyde-paraformaldehyde fixation buffer for 1 h at 37°C, washed overnight in bicarbonate buffer at 4°C, osmicated, dehydrated in ethanol, and embedded in Epon. Sections were stained with lead citrate and examined in a Philips 300 electron microscope (Philips Electronic Instruments, Inc., Mahwah, N.J.). Histochemical staining for aryl sulfatase (21) was performed on cells fixed for 30 min at room teinperature in 1.5% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.4) containing 1% sucrose and washed with 0.1 M sodium cacodylate buffer containing 5% sucrose. Fixed cells were then incubated for 2-3 h at room temperature in medium prepared by dissolving 25 mg p-nitrocatchol sulfate (Sigma Chemical Co.) in 5 ml of acetate-veronal buffer containing 0.16 ml of 24% Pb(N0 3 Jz and adjusted to pH 5.5 with HCl. Immediately before incubation, 5% sucrose was added. After the incubation period, the cells were washed in the acetate-veronal buffer (pH 4.5) with 7% sucrose until the wash was clear. The reaction product was converted by incubation for 10 min in 2% ammonium sulfide in acetate-veronal buffer. After washing, the cells were osmicated, stained by incubation for 30 min with 0.5% uranyl acetate in acetate-veronal buffer (pH 6) and embedded in Epon.
Vesicle Counting To determine the number of acridine orange vesicles (Figure 1), FITC-dextran vesicles, and lysosomes per cell, as well as to assess their intracellular distribution, photomicrographs of ceils were projected onto a sheet of paper so that each cell was -5-10 in. in diameter. The outlines of the cell, the cell nucleus, and individually
1254
LAKE ET AL.
GASTROENTEROLOGY Vol. 92, No.5, Part 1
- - I - - - H - - • perinuclear'
cytoplasm
.. periphera I" cytoplasm
+--\,+-- 'vesicular structures
Figure 1. Morphometric technique used to localize vesicular structures to the "peripheral" as opposed to the "perinuclear" cytoplasm.
distinguishable intracellular vesicles were traced and the total number of vesicles as well as the percentage of vesicles in the perinuclear cytoplasm was determined. The perinuclear cytoplasm was defined by a line equidistant between the nuclear and plasma membranes (Figure 1). All counting of acridine orange vesicles was done without knowledge of the experimental protocol and cells for analysis were selected randomly from each photographic field. For each experimental manipulation, observations were recorded from 3 to 8 batches of cultured cells. Within each batch, several cells were counted from each of several .photographic fields for each experimental manipulation. In prelimihary studies, interobserver variation in the number and distribution of vesicles was determined to be ~7%. Hepatocytes cultured on a collagen matrix average 2 /-Lm in thickness Oones A, Barker M, unpublished observations), and the focal plane of the microscope used in these studies is 1-2 /-Lm. We confirmed that all FITC-dextrancontaining or acridine orange-staining intracellular vesicles were indeed seen within a single focal plane by visualizing cells obtained from several different batches of hepatocytes in several different focal planes and comparing the number of vesicles in each focal plane.
Statistics Student's t-test was used to evaluate significant differences in vesicle numbers and distribution. Differences with p < 0.05 were considered significant.
Results Acidic Vesicles Acidic vesicles were identified in 48-h-old hepatocyte cultures using acridine orange (Figures 2A-2C, Table 1). Acridine orange is a fluorescent weak base that emits a green fluorescence at tnicromolar concentrations, shifting to a red-orange
fluorescence when trapped and concentrated in acidic compartments (22). Thus, when acridine orange is present in the incubation medium in micromolar concentrations, acidic vesicles fluoresce red-orange against a green, comparatively alkaline, cytoplasmic background. Figure 2A shows the appearance of hepatocytes in monolayer culture by transmission illumination. As shown in Figure 2B, hepatocytes exposed to 50 J.LM acridine orange and visualized by fluorescence microscopy demonstrated numerous (169 ± 64/cell) red-orange vesicles that were diffusely distributed throughout the cell cytoplasm (47% ± 7% within the defined perinuclear area) (Table 1). To confirm that the vesicular red-orange fluorescence was due to a pH gradient and not simply to binding of dye to the vesicle membrane, we studied the effects of monensin, a cation-proton exchanging ionophore, which dissipates pH gradients, and the weak bases chloroquine, primaquine, and ammonium chloride, which directly alkalinize acidic compartments (9,12). As shown in Figure 2C, a 2-min incubation in 5 J.LM monensin completely abolished the vesicular fluorescence. This monensin effect was reversible inasmuch as orange vesicles reappeared in the same number and distribution after monensin washout. In a similar fashion, the red-orange acidic vesicles were reversibly abolished after a 10-min intubation in the weak bases NH 4 Cl (50 mM), chloroquine (0.5-2.0 mM), and primaquine (100-500 J.LM). The effect of monensin and the fully reversible nature of the effect of all these agents further indicate that the acidic interior of these vesicles is attributable to ongoing active H+ transport, and hot to either a Donnan effect or to a preformed pH gradient.
Effects of Chlorpromazine on Acidic Vesicles We have previously demonstrated that CPZ inhibits acidification of isolated clathrin-coated vesicles (23). In preliminary studies, a 10-min exposure to CPZ in concentrations of up to 100 J.LM had little effect on the acidic vesicles identified by acridine orange. Chlorpromazine might act, at least in part, by inhibiting active H+ transport rather than by discharging a preformed H+ gradient. Therefore, we examined the effect of CPZ on the redevelopment of acidic vesicle interiors after relaxation of the pH gradient by transient exposure to monensin. Cells incubated with CPZ (100 J.LM) failed to reestablish red-orange acidic vesicles after washout of monensin. After further incubation for 20 min in media free of CPZ and monensin, red-orange vesicles reappeared in approximately the same number and distribution as before incubation with CPZ and monensin.
May 1987
ACIDIC VESICLES IN CULTURED HEPATOCYTES
1255
Figure 2. Light microscopic (magnification x1500) appearance of cultured hepatocytes before or after the various experimental protocols. A. Appearance of hepatocytes in monolayer culture by transmission illumination. B. Appearance of cultured hepatocytes after a 10-min, exposure to acridine orange. Acidic vesicles fluoresce brightly against the relatively alkaline cytoplasmic background. C. Appearance of hepatocytes exposed to monensin (5 JLM) for 2 min, followed by a 10-min incubation in media containing 5 JLM monensin and 50 JLM acridine orange. D. Appearance of hepatocytes cultured overnight in media containing FITC-dextran (70,000 mol wt; 20 mg/ml). Fluorescence is confined to discrete vesiclelike structures distributed throughout the cell. E. Hepatocytes incubated for 20 min in media containing 6-carboxyfluorescein diacetate (35 JLM), a fluorescein derivative that distributes diffusely throughout the cell cytoplasm. F. Appearance of hepatocytes visualized by transmission illumination after histochemical staining for acid phosphatase. Lysosomes appear as dark structures which are concentrated perinuclearly.
This effect of CPZ on the redevelopment of acidic vesicles after exposure to monensin was examined over a concentration range of 10-100 JLM. Figure 3 shows the total number of "reappearing" acidic vesicles in cells exposed to varying concentrations of CPZ expressed as a percent of control. The concentration of CPZ that produces a 50% reduction in the number of vesicles is about 25 JLM.
Fluorescein Isothiocyanate-DextranContaining Vesicles Because most ligands taken up by receptormediated endocytosis are rapidly degraded in lysosomes, we chose to study endocytic vesicles containing dextran, a nondegradable fluid-phase marker (Figures 2D-2E, Table 1) (5,24,25). After an over-
1256
LAKE ET AL.
GASTROENTEROLOGY Vol. 92, No.5, Part 1
Table 1. Number and Distribution of Different Vesicle Populations
Vesicle type Acidic vesicles (acridine orange) Storage compartments (FITC-dextran) Lysosomes (acid phosphatase)
Number (mean ± SEM)
Percentage perinuclear (mean ± SD)
169 ± 64°
47 ± 7°
110 ± 25 b
42 ± 8°
38 ± 14
89 ± 11
FITC-dextran, fluorescein isothiocyanate dextran. ° p < 0.001 compared with lysosomes. b p < 0.05 compared with lysosomes.
night (16 h) incubation in FITC-dextran-containing medium, cells were incubated for 1-6 h in dextranfree medium before visualization (see Materials and Methods). As shown in Figure 2D, the FITC-dextran within these cells appeared as punctate, fluorescence, consistent with its presence in vesicles. The vesicles were numerous (110 ± 25/cell) and diffusely distributed throughout the cell cytoplasm (42% ± 8% within defined perinuclear area). This vesicular appearance contrasts sharply with the appearance of cultured cells exposed to 6-carboxyfluorescein diacetate (Figure 2E), a low-molecular-weight fluorescein derivative that freely permeates cell membranes. When inside the hepatocytes, the acetate groups are hydrolyzed, releasing the fluorescent 6carboxyfluorescein species that distributes diffusely throughout the cytoplasm (26). In other cell types, fluid-phase markers are reportedly cleared with a half-time of 5-10 min from those vesicles that mediate initial uptake (27,28). Indeed, prolonged incubation in dextran has been used by previous workers in other cell types to identify lysosomes (5,25), a point of particular interest in light of our own findings (see below). We therefore extended our studies to examine the location of fluid-phase markers after increasingly longer chase periods. The number of FITC-dextran vesicles after chase periods of 1,2,4, and 6 h were 110 ± 25,117 ± 27, 132 ± 33, and 130 ± 24, respectively, with 42% ± 8%, 44% ± 6%, 43% ± 4%, and 43% ± 5% present, respectively, in the perinuclear area. The total number of vesicles as well as the percent located in the perinuclear area at 2, 4, and 6 h were not significantly different from that at 1 h. Fluorescein isothiocyanate is a pH-sensitive fluorophore that is partially quenched at an acidic pH. To assess whether quenching of FITC-dextran in acidic vesicles might cause an underestimation of the FITC-dextran-containing vesicles, we also determined the number of vesicles in FITC-dextran-
loaded cells incubated for 10 min with or without the addition of 1 mM chloroquine which, as shown above, alkalinizes the vesicle interior. No significant differences were seen in the number (130 ± 22 vs. 110 ± 25) or distribution (38% ± 4% perinuclear vs. 42% ± 8%) in cells incubated with or without chloroquine, respectively. Histochemical Staining for Lysosomes Lysosomes isolated from hepatocytes have been demonstrated to possess interiors acidifed by an active proton-transport mechanism (3). Indeed, acridine orange has been used as a "specific" stain for lysosomes (29-31); even though the specificity of acridine orange staining for lysosomes has not been critically tested. We therefore compared and contrasted the number and distribution of lysosomes, identified by histochemical staining for the lysosomal enzyme acid phosphatase, with that of the acidic vesicles, identified by acridine orange. Lysosomes were stained for acid phosphatase using napthol AS-TR phosphate as substrate and hexazonium pararosanilin for the coupled azo-dye reaction. Using this technique, lysosomes appeared as dark-staining structures within cultured hepatocytes. As compared with both the FITC-dextrancontaining or acidic vesicles, lysosomes (Figure 2F) appeared far fewer in number (38 ± 14/cell) and more restricted to the perinuclear area (89% ± 11%
e ~
'l5
80
~
IIu
60
'iii
?t
., '"c: c5., c:
40
:2
~ 0
10
25
50
100
Chlorpromazine concentration (}1M)
Figure 3, The concentration-dependent inhibition of vesicle acidification by CPZ. Hepatocytes were exposed to 10-100 J.LM CPZ for 5 min followed by a 2-min exposure to monensin. Cells, washed free of monensin, were then visualized by fluorescence microscopy after a 10-min exposure to 50 J.LM acridine orange in media containing the same concentration of CPZ as was present during the preincubation. The number of acridine orange-staining vesicles reappearing after monensin washout is plotted as a percent of control vs. CPZ concentration. Each concentration point represents the mean ± SD of 3-5 cells randomly selected from each of 4-6 plates of cells studied at each CPZ concentration.
ACIDIC VESICLES IN CULTURED HEPATOCYTES
May 1987
1257
I
I Figure 4. Electron micrograph of hepatocytes in monolayer culture exposed for 16 h to media containing dextran (70,000 mol wt; '20 mg/ml), chased for 1 h and then histochemically stained for acid phosphatase. Dextran, by electron microscopy, appears as a dark-staining granular material. It is seen in acid phosphatase-negative structures (solid arrows) next to acid phosphatasepositive lysosomes (open arrows) and was absent from cells incubated in dextran-free medium (magnification x48,000).
within a defined perinuclear area). Although we are unaware of previous estimates of lysosomal number based on histochemical identification at the light microscopic level, recent estimates by morphometry based on electron microscopic studies of intact liver yield greater numbers of lysosomes per cell (32). Lysosomes identified at the. electron-microscopic level, however, include very small «100 nm) primary lysosomes, which likely would be below light microscopic resolution. Electron Microscopy To confirm the observations at the light microscopic level that. dextran was present largely in acid phosphatase-negative storage compartments, cultured hepatocytes were studied by electron microscopy after an overnight exposure to media containing dextran (20 mg/ml) or another electron-dense fluid-phase marker (colloidal gold) (33) and stained for lysosomal enzyme (acid phosphatase and arylsulfatase) as described in Materials and Methods. Dextran can be identified in vesicles as a finely
granular, electron-dense material (34) and, as shown in Figure 4, was found in hepatocytes in numerous intracellular compartments including large electrolucent acid phosphatase-negative vesicles and multivesicular bodies. This granular, electron-dense material was not seen in identically processed control cells not exposed to dextran. The presence of dextran in acid phosphatase-negative vesicles in close proximity to acid phosphatase-positive vesicles (Le., lysosomes) makes it unlikely that the absence of acid phosphatase staining in these vesicles was attributable to technical problems such as variable penetration of fixative or stain. As compared with dextran, colloidal gold has a more characteristic appearance by electron microscopy (35) and can be seen even in organelles stained histochemically for lysosomal enzymes. As seen in Figure 5, colloidal gold, like dextran, is present predominantly in acid phosphatase-negative vesicles but can be seen in acid phosphatase-positive structures as well. Finally, to confirm the absence of lysosomal enzymes in at least some fluid-phase marker storage vesicles, we studied the subcellular distribution of dextran and colloidal gold in cells stained histochemically for a different
1258 LAKE ET AL.
GASTROENTEROLOGY Vol. 92, No.5, Part 1
Figure 5. Electron micrograph of hepatocytes in monolayer culture exposed for 16 h to media containing colloidal gold (18-20-nm particles), then washed and chased for varying periods (panel a, 1 h; panel b, 4 h; panel c, 6 h) in colloidal gold-free media. Colloidal gold particles are seen predominantly in acid phosphatase-negative (panel a and b) or arylsulfatase-negative (panel c) vesicles (solid arrows). Acid phosphatase- or arylsulfatase-positive vesicles (lysosomes) (open arrows) are present as well (magnification x35,OOO).
lysosomal enzyme, arylsulfatase, after the identical pulse-chase periods. Qualitatively similar results were obtained. These results confirm the observation made at the light microscopic level that histochemically identifiable lysosomes represent only a minority of the vesicles involved in the storage of dextran endocytosed 1-6 h earlier.
Discussion In these studies, we describe a method for visualizing acidic vesicles in cultured hepatocytes using the fluorescent weak base acridine orange and we have explored the identity of these acidic vesicles by comparing their number and intracellular distribution with the number and distribution of histochemically identifiable lysosomes as well as vesicles participating in the storage of FITC-dextran and colloidal gold, compounds that are taken up into hepatocytes by fluid-phase endocytosis. Before considering these findings, it is important to discuss the
validation and limitations of the methods used to identify the various vesicle populations. The use of acridine orange to identify acidic vesicles is based on its ability as a membrane-p81;meant weak base to concentrate in acidic compartments (22). Because acridine orange does not exhibit ideal weak-base behavior, the effects of protonophores and weak bases in this study are important in that they indicate that acridine orange staining of intracellular vesicles indeed reflects the presence of a proton gradient rather than nonspecific vesicle binding. Moreover, the effect of protonophores, which specifically enhance the proton conductance of biologic membranes, and the reversible nature of the effect of protonophores, as well as weak bases, indicate that the acidic interior represents ongoing active H+ transport rather than a preformed pH gradient or a Donnan effect (5,36). Dextran has been widely used as a marker of fluid-phase endocytosis by ourselves and others because it is known not to bind to specific cell receptors, exhibits cellular uptake that is linear with
May 1987
respect to extracellular concentration, and is not degraded in lysosomes (5,26,37,38). Moreover, we have recently demonstrated that dextran moves from perfusate to bile by a trans cellular vesicular mechanism in the isolated perfused rat liver in a concentration-independent fashion (24). Using the protocol employed in the present studies, dextran would be expected to be localized in storage compartments rather than early recycling compartments, an assumption supported by our observation that there were no changes in vesicle number or distribution with varying chase periods from 1 to 6 h (26,38). Indeed, our protocol is similar to that previously used to identify lysosomes (5,25,39). This has interesting implications that will be discussed below. Lysosomes in this study were identified at the light microscopic level by histochemical staining for acid phosphatase. Whereas it is conventional to define lysosomes based on their histochemicai properties (25), this approach may underestimate the actual number of lysosomal structures. For example, it requires the presence of acid phosphatase activity as well as sufficient resistance of the enzyme to fixation so as to permit histochemical identification. Our study design does not permit these assumptions to be directly tested; however, several staining protocols were tested so as to maximize the number of individual structures that could be visualized by light microscopy and complementary histochemical studies were performed at the electron microscopic level using two different histochemical stains for lysosomes as well. In addition, it is important to emphasize that the quantitation of vesicles in these studies is potentially limited by the microscopy technique used. For example, clathrin-coated vesicles from rat liver, which have been previously shown by us to possess an active acidification mechanism (1), are small (50-100 nm in diameter) (,*0) and would not be expected to be readily resolvable at the light microscopic level. The same considerations potentially apply to vesicles involved in fluid-phase endocytosis as well as to primary lysosomes. Finally, lightinduced excitation of fluorescent probes including fluorescein and acridine orange produces photobleaching of fluorescence and possible photo damage to cellular structures. Recognizing these potential problems, all measurements of vesicle number and distribution were made from photographs taken after only brief «5 s) preliminary inspection of the culture plate. This type of approach has also been used successfully by previous investigators working in other cell systems (37,41). With these considerations in mind, the present study demonstrates that living cultured hepatocytes contain an average of ~170 acidic intracellular ves-
ACIDIC VESICLES IN CULTURED HEPATOCYTES
1259
icles resolvable at the light microscopic level. These vesicles are distributed diffusely throughout the cell, with about one-half being localized to the perinuclear cytoplasm as assessed by two-dimensional morphometry. The acidified interior is attributable to ongoing active H+ transport, as would be anticipated based on work with isolated organelies, and cannot be explained by either a Donnan equilibrium or a preformed H+ gradient. As previously shown by us for isolated clathrin-coated vesicles (23), CPZ inhibited acidification of these vesicles in a concentration-dependent and reversible fashion. This is interesting in light of the known cholestatic effect of CPZ (42) and the possible importance of active H+ or HC0 3 - transport, or both, in bile formation (43). The present studies, however, permit no direct conclusions regarding the relationship of these observations to the effect of CPZ on bile formation. The present study also suggests that lysosomes account for only a minority of acidic intracellulpr compartments in hepatocytes. Whereas the considerations outlined above raise the possibility that histochemical staining may underestimate the number of lysosomes, we are not aware of any reason to believe that peripheral as opposed to perinuclear lysosomes would be selectively missed. On balance, it appears more likely that lysosomes are fewer.in number and more restricted in their distribution to the perinuclear cytoplasm than are acidic compartments. The findings thus indicate that acridine orange staining cannot be used as a specific marker of lysosomes. Our observations also suggest that acidic vesicles and vesicles involved in storage of fluid-phase markers represent overlapping populations and, indeed, that such storage compartments may account for a large proportion of nonlysosomal acidic compartments. this assertion is supported by the following observations: (a) isolated clathrin-coated vesicles and multivesicular bodies from rat liver, as well as endosomes from other tissues, all of which are likely involved in fluid-phase endocytosis, are actively acidified organelles (1,2,23,37,39,41,44); (b) vesicles involved in fluid-phase uptake by other epithelia have been shown to be acidic (45); and (c) the present studies indicate that the number and distribution of dextran-storing vesicles are similar to those of acidic vesicles. It is also likely that organelles other than those on the endocytic pathway, including some elements of the Golgi and endoplasmic reticulum, may account for some intracellular acidic vesicles in hepatocytes. These conclusions regarding the identity of nonlysosomal acidic organelles in hepatocytes must be considered tentative. It is of interest, however, that very similar findings have
1260 LAKE ET AL.
been reported for cultured human fibroblasts and a human hepatoma cell line using techniques that permit ultrastructural identification of acidified intracellular compartments (46,47). The discrepancy between the number and distribution of dextran-storing vesicles and the number of histochemically identifiable lysosomes is also of interest. Whereas the assumption that only lysosomes are involved in the storage of endocytosed material has been previously challenged and preliminary evidence that vesicles other than lysosomes participate in the storage of endocytosed material has recently been reported for other cell types (48,49), dextran-containing vesicles have frequently been defined as lysosomes and their properties, for example active acidification, attributed exclusively to lysosomes (5,25). In the present study, the majority of vesicles participating in the storage of internalized fluid-phase marker were found to be acid phosphatase-negative and aryl sulfatase-negative by light or electron microscopy, or both. Moreover, in separate studies (data not shown) the number of acid phosphatase-positive structures was not significantly different between hepatocytes that had or had not been exposed for 16 h to dextran. Thus, it does not appear that "induction" of lysosomes by dextran can account for the findings summarized in Table 1. On balance, it appears most reasonable to conclude that many vesicles, apart from lysosomes, participate in the storage of dextran. In any case, the present observations suggest that the use of dextran storage for functional identification of lysosomes merits reevaluation and verification. In addition to the implications of the findings already outlined, development of the technique used here sets the stage for the study of acidic vesicle trafficking in hepatocytes and the factors that regulate it. Our preliminary observations indicate that the same culture plate, mounted in a temperaturecontrolled flow-through chamber, can be repeatedly or continuously visualized by light or fluorescence microscopy while the perfusate is modified with respect to composition or pH, or both (50). The insertion into or removal from the plasma membrane of H+ -trans locating adenosine triphosphatase pump units via a vesicular shuttle mechanism involving vesicles with acidic interiors is believed to play a role in processes such as regulation of intracellular pH and transepithelial HC0 3 -IH+ transport in other tissues (37). Because the liver is an organ actively involved in endocytosis that is here demonstrated to have a large number of acidic intracellular compartments, dynamic studies of vesicle movement in living hepatocytes will be of considerable interest.
GASTROENTEROLOGY Vol. 92, No.5, Part 1
References 1. Van Dyke RW, Steer q, Scharschmidt BF. Clathrin-coated vesicles from rat liver: enzymatic profile and characterization of ATP-dependent proton transport. Proc Natl Acad Sci USA 1984;81:3108-12. 2. Van Dyke RW, Hornick CA, Belcher J, Scharschmidt BF, Havel RJ. Identification and characterization of ATPdependent proton transport by rat liver multivesicular bodies. J Bioi Chern 1985;260:11021-6. 3. Schneider DL. ATP-dependent acidification of membrane vesicles isolated from purified rat liver lysosomes. Acidification activity requires phosphate. J Bioi Chern 1983;258: 1833-8. 4. Zhang F, Schneider DL. The bioenergetics of Golgi apparatus function: evidence for an ATP-dependent proton pump. Biochem Biophys Res Commun 1983;114:620-5. 5. Tycko B, Keith CH, Maxfield FR. Rapid acidification of endocytic vesicles containing asialoglycoprotein in cells of a human hepatoma line. J Cell Bioi 1983;97:1762-76. 6. Hornick CA, Jones AL, Renaud G, Hradek G, Havel RJ. Effect of chloroquine on low-density lipoprotein catabolic pathway in rat hepatocytes. Am J Physiol 1984;246:G187-94. 7. Posner BI, Khan MN, Bergeron HM. Effect of chloroquine on the internalization of p25-insulin into subcellular fractions of rat liver. J Bioi Chern 1982;257:5789-99. 8. Dunn wA, Hubbard AL. Receptor-mediated endocytosis of epidermal growth factor (EGF) by hepatocytes in the perfused rat liver: ligand and receptor dynamics. J Cell Bioi 1984; 98:2148-59. 9. Maxfield FR. Weak bases and ionophores rapidly and reversibly raise the pH of endocytic vesicles in cultured mouse fibroblasts. J Cell Bioi 1982;95:676-81. 10. Harford J, Wolkoff AW, Ashwell G, Klausner RD. Monensin inhibits intracellular dissociation of asialoglycoproteins from their receptor. J Cell Bioi 1983;96:1824-8. 11. Tartakoff AM. Perturbation of vesicular traffic with the carboxylic ionophore monensin. Cell 1983;32:1026-8. 12. Schwartz AL, Bolognesi A, Fridovich SE. Recycling of the as ialogl yeoprotein receptor and the effect of lysosomotropic amines in hepatoma cells. J Cell Bioi 1984;98:732-8. 13. Octave IN, Schneider YJ, Trovet A, Crichton RR. Iron uptake and utilization by mammalian cells. I. Cellular uptake of transferrin and iron. Trends Biochem Sci 1983;8:217-20. 14. Marsh M, Bolzau E, Helenius A. Penetration of Semliki Forest Virus from acid prelysosomal vesicles. Cell 1983;32:931-40. 15. Helenius A, Mellman I, Wall D, Hubbard A. Endosomes. Trends Biochem Sci 1983;8:245-50. 16. Bissell DM, Hammaker LE, Meyer UA. Parenchymal cells from adult rat liver in non-proliferating monolayer culture. I. Functional studies. J Cell Bioi 1973;59:722-34. 17. Scharschmidt BF, Stephens JE. Sodium, chloride and taurocholate transport by cultured rat hepatocytes. Proc Nat! Acad Sci USA 1981;78:986-90. 18. Geoghegan WD, Ackerman GA. Adsorption of horseradish peroxidase, ovomucoid and anti-immunoglobulin to colloidal gold for the indirect detection of concanavalin A, wheat germ agglutinin and goat anti-human immunoglobulin G on cell surfaces at the electron microscopic level: a new method, theory and application. J Histochem Cytochem 1977;25: 1187-200. 19. Barka T, Anderson PJ. Histochemical methods for acid phosphatase using hexazonium pararosanilin as coupler. J Histochem Cytochem 1962;10:741-53. 20. Robinson JM, Karnovsky MJ. Ultrastructural localization of several phosphatases with cerium. J Histochem Cytochem 1983;31:1197-208.
May 1987
21. Bentfeld-Barker ME, Bainton DF. Cytochemical localization of arylsulfatase B in rat basophils and mast cells. J Histochem Cytochem 1980;28:1055-61. 22. Lee HC, Forte JG. A study of H+ transport in gastric microsomal vesicles using fluorescent probes. Biochim Biophys Acta 1978;508:339-45. 23. Van Dyke RW, Scharschmidt BF, Steer q. ATP-dependent proton transport by isolated brain clathrin-coated vesicles. Role of clathrin and other determinants of acidification. Biochim Biophys Acta 1985;812:423-36. 24. Lake JR, Licko V, Van Dyke RW, Scharschmidt BF. Biliary secretion of fluid phase markers by the isolated perfused rat liver: the role of transcellular vesicular transport. J Clin Invest 1985;76:676-84. 25. Ohkuma S, Poole B. Fluorescein probe measurement of the intralysosomal pH in living cells and the perturbation of pH by various agents. Proc Natl Acad Sci USA 1978;75:3327-31. 26. Thomas JA, Buchsbaum RN, Zimniak A, Racker E. Intracellular pH measurements in Ehrlich ascites tumor cells utilizing spectroscopic probes generated in situ. Biochemistry 1979;19:2210-8. 27. Besterman JM, Airhart JA, Woodworth Re, Low RB. Exocytosis of pinocytosed fluid in cultured cells: kinetic evidence for rapid turnover and compartinentation. J Cell BioI 1981;91:716-27. 28. Storrie B, Pool RR, Sachdeva M, Maurey KM, Oliver C. Evidence for both pre-lysosomal and lysosomal intermediates in endocytic pathways. J Cell BioI 1984;98:108-15. 29. Herman B, Albertini DF. Ligand-induced rapid distribution of Iysosomes is temporally distinct from endosome translocation. Nature 1983;304:738-40. 30. Herman B, Albertini DF. A time-lapse video image intensification analysis of cytoplasmic organelle movements during endosome translocation. J Cell BioI 1984;98:565-76. 31. Robbins E, Marcus PI, Gonatas NK. Dynamics of acridine orange-cell interaction. II. Dye-induced ultrastructural changes in multivesicular bodies (acridine orange particles). J Cell BioI 1964;21:49-62. 32. Sewell RB, Dillon C, Grinpukel S, Yeomans ND, Smallwood RA. Peri canalicular location of hepatocyte Iysosomes and effects of fasting: a morphometric analysis. Hepatology 1986; 6:305-11. 33. Davies P, Allison AC, Haswell AD. The quantitative estimation of pinocytosis using radioactive colloidal gold. Biochem Biophys Res Commun 1973;52:627-34. 34. Ainsworth SK. An ultrastructural method for the use of polyvinylpyrrolidone and dextrans as electron opaque tracers. J Histochem Cytochem 1977;25:1254-9.
ACIDIC VESICLES IN CULTURED HEPATOCYTES
1261
35. de Bruijn WC, Schellens JPM, van Buitenen JMH, van der Meulen J. X-ray microanalysis of colloidal-gold-labelled Iysosomes in rat liver sinusoidal cells after incubation for acid phosphatase activity. Histochemistry 1980;66:137-48. 36. Yamashiro DJ, Fluss SR, Maxfield FR. Acidification of endocytic vesicles by an ATP-dependent proton pump. J Cell BioI 1983;97:929-34. 37. Gluck S, Cannon C, AI-Awqati Q. Exocytosis regulates urinary acidification in turt!e bladder by rapid insertion of H+ pumps into the luminal membrane. Proc Nat! Acad Sci USA 1982;79:4327-31. 38. Scharschmidt BF, Lake JR, Renner EL, Licko V, Van Dyke RW. Fluid phase endocytosis by cultured rat hepatocytes and perfused rat liver. Proc Nat! Acad Sci USA 1986:83:9488-92. 39. Ohkuma S, Moriyama Y, Takano T. Identification and characterization of a proton pump on Iysosomes by fluorescein isothiocyanate-dextran fluorescence. Proc Nat! Acad Sci USA 1982;79:2758-62. 40. Steer q, Klausner RD. Clathrin-coated pits and coated vesicles: functional and structural studies. Hepatology 1983;3: 437-54. 41. Tycko B, Maxfield FR. Rapid acidification of endocytic vesicles containing o2-macroglobulin. Cell 1982;28:643-51. 42. Bartholomew LG, Cain JC, Frazier SH, et al. Effect of chlorpromazine on the liver. Gastroenterology 1958;34:1096-107.' 43. Scharschmidt BF, Van Dyke RW. Mechanisms of hepatic electrolyte transport. Gastroenterology 1983;85:1199-214. 44. Galloway CJ, Dean GE, Marsh M, Rudnick G, Mellman I. Acidification of macrophage and fibroblast endocytic vesicles in vitro. Proc Nat! Acad Sci USA 1983;80:3334-8. 45. Gluck S, Kelly S, AI-Awqati Q. The proton trans locating ATPase responsible for urinary acidification. J BioI Chern 1982;257:9230-3. 46. Anderson RGW, Falck JR, Goldstein JL, Brown MS. Visualization of acidic organelles in intact cells by electron microscopy. Proc Nat! Acad Sci USA 1984;81:4838-42. 47. Schwartz AL, Strous GJAM, Slot JW, Gevze HJ. Immunoelectron microscopic localization of acidic intracellular compartments in hepatoma cells. EMBO J 1985;4:899-904. 48. McKinley DN, Wiley HS. The pathway and kinetics of fluid phase endocytosis in fibroblasts (abstr). J Cell BioI 1984; 99:376a. 49. Hohman TC, Bowers B. Phagosome-lysosome fusion is a function of phagocytic load (abstr). J Cell BioI 1984;99:363a. 50. Lake JR, Van Dyke RW, Scharschmidt BF. Visualization of vesicles in cultured hepatocytes: description of a technique and preliminary findings (abstr). Clin Res 1984;32:76A.
Acidic Vesicles in Cultured Rat Hepatocytes Identification and Characterization of Their Relationship to Lysosomes and Other Storage Vesicles JOHN R. LAKE, REBECCA W. VAN DYKE, and BRUCE F. SCHARSCHMIDT The Department of Medicine and Liver Center, University of California, San Francisco, California
We and others recently have demonstrated adenosine triphosphate-dependent acidification in a variety of pre1ysosoma1 organelles isolated from liver including clathrin-coated vesicles, multivesicu1ar bodies, and Go1gi. Little is known, however, regarding the number or distribution of acidic compartments in intact hepatocytes. We therefore have utilized acridine orange, a fluorescent weak base, to study the number and distribution of acidic vesicles of rat hepatocytes in primary culture and compared these with the number and distribution of 1ysosomes and other storage vesicles. Hepatocytes were found to contain about 170 acidic compartments per cell Received August 20, 1986. Accepted September 26, 1986. Address requests for reprints to: John R. Lake, M.D., Gastrointestinal Research Unit, HSW 1120, University of California, San Francisco, California 94143. This work was supported in part by grants AM26270, AM26743, AM07453, and AM01254 from the National Institutes of Health, and grants from the Giannini Foundation, the American Liver Foundation, and the Walter C. Pew Fund for Gastrointestinal Research. This work was presented at the plenary session of the meeting of the American Association for the Study of Liver Disease, Chicago, Illinois, November 10-11,1984; and has been published previously in abstract form (Hepatology 1984;4:1070). The authors thank the University of California San Francisco Liver Center-supported (AM26743) Core Facilities, including the Cell Culture Facility (Dr. D. Montgomery Bissell, director) and the Electron Microscopy Facility (Dr. Albert 1. Jones, director). The authors also thank Jocelyn Matsumoto-Pan, Michael Wong, and Paul George for their technical support; Mark Malamud for his contribution in preparing the cultured hepatocytes, and Dr. Mary Barker for her assistance with the histochemical stain for acid phosphatase and electron microscopy. In addition, the authors thank Diana Fedorchak and Michael Karasik for their expert editorial assistance. © 1987 by the American Gastroenterological Association 0016-5085/871$3.50
by fluorescence microscopy. These vesicles were diffusely distributed throughout the cell cytoplasm, with about 50% in the perinuclear area by modified morphometry. The acridine orange staining of these vesicles was reversibly dissipated by monensin, NH4 C1, chloroquine, and primaquine, indicating these vesicles exhibit an acidic interior established by active proton transport. In addition, the cho1estatic agent chlorpromazine reversibly inhibited, in a dose-dependent fashion, the redevelopment of a pH gradient in the acidic vesicles after dissipation by monensin. The number and distribution of these acidic vesicles were not significantly different from the number and distribution of vesicles involved in the storage (up to 6 h after internalization) of the fluid phase marker fluorescein-dextran. By contrast, histochemically identifiable 1ysosomes were fewer in number and significantly more restricted in their distribution to the perinuclear area (89%) than either dextran-storing or acidic vesicles: Electron microscopic studies confirmed that endocytosed dextran as well as another fluid phase marker, colloidal gold, were found predominantly in acid phosphatase- and ary1su1fatase-negative vesicles for up to 6 h after internalization. These studies indicate that hepatoGytes contain numerous intracellular vesicles acidifed by an active H+ transport mechanism. Based on their comparative number and distribution, acidic vesicles probably include vesicles involved in fluid-phase endocytosis but only a minority are 1ysosomes. The findings also indicate that fluid-phase markers are stored predominantly in Abbreviations used in this paper: CPZ, chlorpromazine; FITCdextran, fluorescein isothiocyanate dextran.
1252
LAKE ET AL.
vesicles other than histochemically identifiable 1ysosomes for up to 6 h after internalization. Finally, this technique also affords the opportunity for studying the movement of such vesicles in a vital preparation. Recent studies by ourselves and others have demonstrated that a variety of intracellular organelles including clathrin-coated vesicles, multivesicular bodies, Golgi, and lysosomes are actively acidified via a proton-trans locating adenosine triphosphatase (1-4). The acidification of endocytic vesicles has been shown to be essential for the normal internalization and processing of ligands such as asialoglycoproteins (5), low-density lipoproteins (6), insulin (7), and epidermal growth factor (8), which are taken up into hepatocytes via receptor-mediated endocytosis. Agents such as protonophores and weak bases, which dissipate proton gradients responsible for vesicle acidification (1,9), prevent the normal dissociation of ligands from receptors (10), block the delivery of ligands to lysosomes (11), and interrupt the normal recycling of receptors back to the plasma membrane (12). In addition, vesicle acidification is important for the dissociation of iron from transferrin after endocytosis of the iron-transferrin complex (13), as well as the translocation of viruses (14) and diphtheria toxin from the vesicle interior to the cell cytosol (15). Whereas studies to date have focused primarily on the identification and characterization of the acidification mechanism in isolated organelles (1-3), less attention has been paid to the study of acidic compartments in situ. We therefore have developed a method for identifying acidic compartments via light microscopy in living cultured hepatocytes and have utilized this technique to (a) determine the number of intracellular acidic vesicles and the nature of the acidification mechanism and (b) compare the number and distribution of these acidic vesicles with the number and distribution of lysosomes and vesicles involved in the storage of fluid-phase markers.
Materials and Methods Reagents Fluorescein isothiocyanate dextran (FITC-dextran) (70,000 average mol wt), chloroquine (diphosphate salt), and primaquine (diphosphate salt) were obtained from Sigma Chemical Co., St. Louis, Mo. 6-Carboxyfluorescein diacetate was obtained from Molecular Probes, Inc., Junction City, Ore. Acridine orange was obtained from Eastman Kodak Chemicals, Rochester, N.Y., and monensin from Calbiochem-Behring, La Jolla, Calif. Cloroauric acid was purchased from ICN, Plainview, N.Y. Chlorpromazine was a gift from Dr. Carl Kaiser (Smith, Kline and French Laboratories, Philadelphia, Pa.).
GASTROENTEROLOGY Vol. 92, No.5. Part 1
Cultured Rat Hepatocytes Rat hepatocytes were prepared by collagenase perfusion of nonregenerating rat liver as previously described (16) and were plated on collagen-coated 35-mm plastic culture dishes at a density of 0.5-1.0 mg cell protein/dish. Cells were cultured at 37°C in the presence of 3% CO 2 for 48 h before use in media containing Hank's buffered salts and modified 199 amino acids supplemented with 1% rat serum, BME vitamins (10 mllL, Grand Island Biological Co., Grand Island, N.Y.), penicillin (100 U/ml), insulin (4 u/L), cortisosterone (1 pM), vitamin C (50 J.Lglml), and glucose (5.6 mM). Cell viability was tested by trypan blue exclusion as well as by release of lactate dehydrogenase into the incubation medium (17).
Acridine Orange Staining Two-day-old cultures of rat hepatocytes were incubated for 10 min in fresh media containing acridine orange at a concentration of 50 J.LM. Acridine orange-containing media was then removed and the cells were washed twice with 1 ml of a balanced electrolyte solution. After the second wash, a 25-mm circular glass coverslip was placed over the cells and the cells were visualized immediately under oil by transmission or fluorescence illumination using an Olympus fluorescence microscope (Olympus Corporation of America, New Hyde Park, N.Y.) equipped with a X40 oil immersion objective. Fluorescence visualization was accomplished using two Olympus BG-12 short-pass excitation filters, one neutral density excitation filter, and a single 0-530 long-pass emission filter. Cells were photographed during both transmission and fluorescence microscopy using an Olympus C-35 camera and 400 ASA Ektachrome film (Eastman Kodak). To minimize photo bleaching and photodamage, fluorescence photographs were taken after only a brief «5 s) visualization for identification of field and focusing.
Effects of Protonophores and Weak Bases on Acridine Orange Staining The effect of the proton-cation exchanging ionophore, monensin, was determined by exposing cells for 2 min to media containing 5 J.LM monensin followed by a 10-min incubation in media containing 50 J.LM acridine orange and 5 J.LM monensin. Cells were then visualized and photographed according to the protocol described previously. Reversibility of the monensin effect was assessed by incubating monensin-exposed cells in fresh, monensin-free media for 10 min followed by visualization after a second 10-min exposure to 50 J.LM acridine orange. The effects of the weak bases ammonium chloride (50 mM), chloroquine (0.5-2.0 mM), and primaquine (100-500 J.LM), as well as the reversibility of these effects, were determined in a similar fashion, the only difference being that the cells were exposed to the weak bases for 10 min before acridine orange exposure and visualization.
Effects of Chlorpromazine on Acridine Orange Staining The ability of chlorpromazine (CPZ) to inhibit development of a proton gradient across vesicle mem-
May 1987
branes was determined by incubating cells in media containing 10-100 pM CPZ for 5 min, followed by a 2-min exposure to monensin. Cells washed free of monensin were then visualized by fluorescence microscopy after a 10-min exposure to 50 pM acridine orange in CPZcontaining media and photographed. Reversibility of the CPZ effect was determined by incubating the CPZ-exposed cells for an additional 20 min in media free of CPZ and monensin, after which they were again exposed to 50 /LM acridine orange and visualized as described above.
Fluorescein Isothiocyanate-Dextran Uptake FITC-dextran (70,000 mol wt) was prepared before use by passage over a Sephadex G-75 (Pharmacia, Piscataway, N.J.) column (0.9 cm x 25 cm) to separate 70,000 mol wt FITC-dextran from free fluorescein and other lowmolecular-weight fluoresceinated contaminants. The void volume fraction was dialyzed against distilled water overnight at 4°C, lyophilized, and stored dessicated at 4°C until use. Cultured hepatocytes were incubated for 16 h (beginning 32 h after plating) in media containing 20 mg/ml FITC-dextran. Before visualization, hepatocytes were washed three times with buffer at 4°C, and reincubated in fresh FITC-dextran-free media at 37°C for 1, 2, 4, or 6 h. At the end of the chase period, cells were visualized and photographed using the same technique and filters used fbr acridine orange fluorescence.
Colloidal Gold Uptake Colloidal gold particles were prepared according to the method of Geoghegan and Ackerman (18). Cloroauric acid (0.05 g) was added to 500 ml of distilled H 2 0 and brought to a boil. The boiling solution was stirred vigorously while 12.5 ml of a 1% aqueous trisodium citrate solution was rapidly added. After a color change to orangered, the solution was cooled and 1.5 ml of a 1% polyethylene glycol (10,000 mol wt) solution was added to stabilize the suspension during storage at 4°C. This procedure reliably yields colloidal particles that have a crosssectional diameter by electron microscopy of 18-20 nm.
Histochemical Staining for Lysosomes Light microscopy. Lysosomes were identified at the light microscopic level in 48-h-old cultures using a naphthol AS-TR phosphate-hexazonium pararosanilin histochemical stain for acid phosphatase (19). Cell mono layers were first fixed by incubation for 30 min at room temperature in a solution containing 1% CaCl 2 and 4% formaldehyde, pH 7.0, followed by a 2-h incubation at 37°C in 140 mM sodium cacodylate, pH 7.3. Cells were then rinsed in a large volume of veronal acetate buffer [prepared by dissolving 9.71 g sodium acetate· 3H 2 0 and 14.71 g sodium barbiturate in 500 ml of CO 2 -free distilled water and further diluting this solution 1:3 (vol/vol) with distilled water) and incubated in the naphthol AS-TR phosphate-hexazonium pararosanilin solution for 90 min at 37°C. The naphthol AS-TR phosphatehexazonium pararosanilin solution was prepared by dilut-
ACIDIC VESICLES IN CULTURED HEPATOCYTES
1253
ing 5 ml of the veronal acetate buffer in 12 ml of distilled water and adding 1 ml of a 10-mg/ml solution of naphthol AS-TR in N,N-dimethyl-formamide to which was added 0.8 ml of the pararosanilin solution (1 g pararosanilin hydrochloride in 20 ml of distilled water and 5 ml of concentrated HCI). The cells were then thoroughly washed in large volumes of the sodium cacodylate buffer and air-dried for 10-20 min before visualization under oil by transmission illumination. In preliminary studies, the effects of variable fixation times (2-16 h) and staining times (0.5-3 h) were studied and the protocol (above) yielding the maximum number and definition of individual lysosomes was utilized. Electron microscopy. Histochemical staining for acid phosphatase was performed in cultured cells incubated overnight (16 h) in media containing either dextran (20 mg/ml) or colloidal gold (10.4 mg/ml), washed, and placed in fresh, marker-free media as described above. After 1, 2, 4, or 6 h, the medium was aspirated and the cells were washed and fixed in a solution of 2.5% glutaraldehyde and 0.8% paraformaldehyde in 0.13 M sodium bicarbonate buffer (pH 7.4) for 10 min at room temperature followed by gentle mixing for 1 h at 37°C in RobinsonKarnofsky media (0.1 M acetate buffer, pH 5.0, containing 1 mM (3-glycerophosphate, and 2 mM CeCl 3 and 5% sucrose) as previously described (20). After the cytochemical reaction, the cells were washed twice in the above media and further incubated in the glutaraldehyde-paraformaldehyde fixation buffer for 1 h at 37°C, washed overnight in bicarbonate buffer at 4°C, osmicated, dehydrated in ethanol, and embedded in Epon. Sections were stained with lead citrate and examined in a Philips 300 electron microscope (Philips Electronic Instruments, Inc., Mahwah, N.J.). Histochemical staining for aryl sulfatase (21) was performed on cells fixed for 30 min at room teinperature in 1.5% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.4) containing 1% sucrose and washed with 0.1 M sodium cacodylate buffer containing 5% sucrose. Fixed cells were then incubated for 2-3 h at room temperature in medium prepared by dissolving 25 mg p-nitrocatchol sulfate (Sigma Chemical Co.) in 5 ml of acetate-veronal buffer containing 0.16 ml of 24% Pb(N0 3 Jz and adjusted to pH 5.5 with HCl. Immediately before incubation, 5% sucrose was added. After the incubation period, the cells were washed in the acetate-veronal buffer (pH 4.5) with 7% sucrose until the wash was clear. The reaction product was converted by incubation for 10 min in 2% ammonium sulfide in acetate-veronal buffer. After washing, the cells were osmicated, stained by incubation for 30 min with 0.5% uranyl acetate in acetate-veronal buffer (pH 6) and embedded in Epon.
Vesicle Counting To determine the number of acridine orange vesicles (Figure 1), FITC-dextran vesicles, and lysosomes per cell, as well as to assess their intracellular distribution, photomicrographs of ceils were projected onto a sheet of paper so that each cell was -5-10 in. in diameter. The outlines of the cell, the cell nucleus, and individually
1254
LAKE ET AL.
GASTROENTEROLOGY Vol. 92, No.5, Part 1
- - I - - - H - - • perinuclear'
cytoplasm
.. periphera I" cytoplasm
+--\,+-- 'vesicular structures
Figure 1. Morphometric technique used to localize vesicular structures to the "peripheral" as opposed to the "perinuclear" cytoplasm.
distinguishable intracellular vesicles were traced and the total number of vesicles as well as the percentage of vesicles in the perinuclear cytoplasm was determined. The perinuclear cytoplasm was defined by a line equidistant between the nuclear and plasma membranes (Figure 1). All counting of acridine orange vesicles was done without knowledge of the experimental protocol and cells for analysis were selected randomly from each photographic field. For each experimental manipulation, observations were recorded from 3 to 8 batches of cultured cells. Within each batch, several cells were counted from each of several .photographic fields for each experimental manipulation. In prelimihary studies, interobserver variation in the number and distribution of vesicles was determined to be ~7%. Hepatocytes cultured on a collagen matrix average 2 /-Lm in thickness Oones A, Barker M, unpublished observations), and the focal plane of the microscope used in these studies is 1-2 /-Lm. We confirmed that all FITC-dextrancontaining or acridine orange-staining intracellular vesicles were indeed seen within a single focal plane by visualizing cells obtained from several different batches of hepatocytes in several different focal planes and comparing the number of vesicles in each focal plane.
Statistics Student's t-test was used to evaluate significant differences in vesicle numbers and distribution. Differences with p < 0.05 were considered significant.
Results Acidic Vesicles Acidic vesicles were identified in 48-h-old hepatocyte cultures using acridine orange (Figures 2A-2C, Table 1). Acridine orange is a fluorescent weak base that emits a green fluorescence at tnicromolar concentrations, shifting to a red-orange
fluorescence when trapped and concentrated in acidic compartments (22). Thus, when acridine orange is present in the incubation medium in micromolar concentrations, acidic vesicles fluoresce red-orange against a green, comparatively alkaline, cytoplasmic background. Figure 2A shows the appearance of hepatocytes in monolayer culture by transmission illumination. As shown in Figure 2B, hepatocytes exposed to 50 J.LM acridine orange and visualized by fluorescence microscopy demonstrated numerous (169 ± 64/cell) red-orange vesicles that were diffusely distributed throughout the cell cytoplasm (47% ± 7% within the defined perinuclear area) (Table 1). To confirm that the vesicular red-orange fluorescence was due to a pH gradient and not simply to binding of dye to the vesicle membrane, we studied the effects of monensin, a cation-proton exchanging ionophore, which dissipates pH gradients, and the weak bases chloroquine, primaquine, and ammonium chloride, which directly alkalinize acidic compartments (9,12). As shown in Figure 2C, a 2-min incubation in 5 J.LM monensin completely abolished the vesicular fluorescence. This monensin effect was reversible inasmuch as orange vesicles reappeared in the same number and distribution after monensin washout. In a similar fashion, the red-orange acidic vesicles were reversibly abolished after a 10-min intubation in the weak bases NH 4 Cl (50 mM), chloroquine (0.5-2.0 mM), and primaquine (100-500 J.LM). The effect of monensin and the fully reversible nature of the effect of all these agents further indicate that the acidic interior of these vesicles is attributable to ongoing active H+ transport, and hot to either a Donnan effect or to a preformed pH gradient.
Effects of Chlorpromazine on Acidic Vesicles We have previously demonstrated that CPZ inhibits acidification of isolated clathrin-coated vesicles (23). In preliminary studies, a 10-min exposure to CPZ in concentrations of up to 100 J.LM had little effect on the acidic vesicles identified by acridine orange. Chlorpromazine might act, at least in part, by inhibiting active H+ transport rather than by discharging a preformed H+ gradient. Therefore, we examined the effect of CPZ on the redevelopment of acidic vesicle interiors after relaxation of the pH gradient by transient exposure to monensin. Cells incubated with CPZ (100 J.LM) failed to reestablish red-orange acidic vesicles after washout of monensin. After further incubation for 20 min in media free of CPZ and monensin, red-orange vesicles reappeared in approximately the same number and distribution as before incubation with CPZ and monensin.
May 1987
ACIDIC VESICLES IN CULTURED HEPATOCYTES
1255
Figure 2. Light microscopic (magnification x1500) appearance of cultured hepatocytes before or after the various experimental protocols. A. Appearance of hepatocytes in monolayer culture by transmission illumination. B. Appearance of cultured hepatocytes after a 10-min, exposure to acridine orange. Acidic vesicles fluoresce brightly against the relatively alkaline cytoplasmic background. C. Appearance of hepatocytes exposed to monensin (5 JLM) for 2 min, followed by a 10-min incubation in media containing 5 JLM monensin and 50 JLM acridine orange. D. Appearance of hepatocytes cultured overnight in media containing FITC-dextran (70,000 mol wt; 20 mg/ml). Fluorescence is confined to discrete vesiclelike structures distributed throughout the cell. E. Hepatocytes incubated for 20 min in media containing 6-carboxyfluorescein diacetate (35 JLM), a fluorescein derivative that distributes diffusely throughout the cell cytoplasm. F. Appearance of hepatocytes visualized by transmission illumination after histochemical staining for acid phosphatase. Lysosomes appear as dark structures which are concentrated perinuclearly.
This effect of CPZ on the redevelopment of acidic vesicles after exposure to monensin was examined over a concentration range of 10-100 JLM. Figure 3 shows the total number of "reappearing" acidic vesicles in cells exposed to varying concentrations of CPZ expressed as a percent of control. The concentration of CPZ that produces a 50% reduction in the number of vesicles is about 25 JLM.
Fluorescein Isothiocyanate-DextranContaining Vesicles Because most ligands taken up by receptormediated endocytosis are rapidly degraded in lysosomes, we chose to study endocytic vesicles containing dextran, a nondegradable fluid-phase marker (Figures 2D-2E, Table 1) (5,24,25). After an over-
1256
LAKE ET AL.
GASTROENTEROLOGY Vol. 92, No.5, Part 1
Table 1. Number and Distribution of Different Vesicle Populations
Vesicle type Acidic vesicles (acridine orange) Storage compartments (FITC-dextran) Lysosomes (acid phosphatase)
Number (mean ± SEM)
Percentage perinuclear (mean ± SD)
169 ± 64°
47 ± 7°
110 ± 25 b
42 ± 8°
38 ± 14
89 ± 11
FITC-dextran, fluorescein isothiocyanate dextran. ° p < 0.001 compared with lysosomes. b p < 0.05 compared with lysosomes.
night (16 h) incubation in FITC-dextran-containing medium, cells were incubated for 1-6 h in dextranfree medium before visualization (see Materials and Methods). As shown in Figure 2D, the FITC-dextran within these cells appeared as punctate, fluorescence, consistent with its presence in vesicles. The vesicles were numerous (110 ± 25/cell) and diffusely distributed throughout the cell cytoplasm (42% ± 8% within defined perinuclear area). This vesicular appearance contrasts sharply with the appearance of cultured cells exposed to 6-carboxyfluorescein diacetate (Figure 2E), a low-molecular-weight fluorescein derivative that freely permeates cell membranes. When inside the hepatocytes, the acetate groups are hydrolyzed, releasing the fluorescent 6carboxyfluorescein species that distributes diffusely throughout the cytoplasm (26). In other cell types, fluid-phase markers are reportedly cleared with a half-time of 5-10 min from those vesicles that mediate initial uptake (27,28). Indeed, prolonged incubation in dextran has been used by previous workers in other cell types to identify lysosomes (5,25), a point of particular interest in light of our own findings (see below). We therefore extended our studies to examine the location of fluid-phase markers after increasingly longer chase periods. The number of FITC-dextran vesicles after chase periods of 1,2,4, and 6 h were 110 ± 25,117 ± 27, 132 ± 33, and 130 ± 24, respectively, with 42% ± 8%, 44% ± 6%, 43% ± 4%, and 43% ± 5% present, respectively, in the perinuclear area. The total number of vesicles as well as the percent located in the perinuclear area at 2, 4, and 6 h were not significantly different from that at 1 h. Fluorescein isothiocyanate is a pH-sensitive fluorophore that is partially quenched at an acidic pH. To assess whether quenching of FITC-dextran in acidic vesicles might cause an underestimation of the FITC-dextran-containing vesicles, we also determined the number of vesicles in FITC-dextran-
loaded cells incubated for 10 min with or without the addition of 1 mM chloroquine which, as shown above, alkalinizes the vesicle interior. No significant differences were seen in the number (130 ± 22 vs. 110 ± 25) or distribution (38% ± 4% perinuclear vs. 42% ± 8%) in cells incubated with or without chloroquine, respectively. Histochemical Staining for Lysosomes Lysosomes isolated from hepatocytes have been demonstrated to possess interiors acidifed by an active proton-transport mechanism (3). Indeed, acridine orange has been used as a "specific" stain for lysosomes (29-31); even though the specificity of acridine orange staining for lysosomes has not been critically tested. We therefore compared and contrasted the number and distribution of lysosomes, identified by histochemical staining for the lysosomal enzyme acid phosphatase, with that of the acidic vesicles, identified by acridine orange. Lysosomes were stained for acid phosphatase using napthol AS-TR phosphate as substrate and hexazonium pararosanilin for the coupled azo-dye reaction. Using this technique, lysosomes appeared as dark-staining structures within cultured hepatocytes. As compared with both the FITC-dextrancontaining or acidic vesicles, lysosomes (Figure 2F) appeared far fewer in number (38 ± 14/cell) and more restricted to the perinuclear area (89% ± 11%
e ~
'l5
80
~
IIu
60
'iii
?t
., '"c: c5., c:
40
:2
~ 0
10
25
50
100
Chlorpromazine concentration (}1M)
Figure 3, The concentration-dependent inhibition of vesicle acidification by CPZ. Hepatocytes were exposed to 10-100 J.LM CPZ for 5 min followed by a 2-min exposure to monensin. Cells, washed free of monensin, were then visualized by fluorescence microscopy after a 10-min exposure to 50 J.LM acridine orange in media containing the same concentration of CPZ as was present during the preincubation. The number of acridine orange-staining vesicles reappearing after monensin washout is plotted as a percent of control vs. CPZ concentration. Each concentration point represents the mean ± SD of 3-5 cells randomly selected from each of 4-6 plates of cells studied at each CPZ concentration.
ACIDIC VESICLES IN CULTURED HEPATOCYTES
May 1987
1257
I
I Figure 4. Electron micrograph of hepatocytes in monolayer culture exposed for 16 h to media containing dextran (70,000 mol wt; '20 mg/ml), chased for 1 h and then histochemically stained for acid phosphatase. Dextran, by electron microscopy, appears as a dark-staining granular material. It is seen in acid phosphatase-negative structures (solid arrows) next to acid phosphatasepositive lysosomes (open arrows) and was absent from cells incubated in dextran-free medium (magnification x48,000).
within a defined perinuclear area). Although we are unaware of previous estimates of lysosomal number based on histochemical identification at the light microscopic level, recent estimates by morphometry based on electron microscopic studies of intact liver yield greater numbers of lysosomes per cell (32). Lysosomes identified at the. electron-microscopic level, however, include very small «100 nm) primary lysosomes, which likely would be below light microscopic resolution. Electron Microscopy To confirm the observations at the light microscopic level that. dextran was present largely in acid phosphatase-negative storage compartments, cultured hepatocytes were studied by electron microscopy after an overnight exposure to media containing dextran (20 mg/ml) or another electron-dense fluid-phase marker (colloidal gold) (33) and stained for lysosomal enzyme (acid phosphatase and arylsulfatase) as described in Materials and Methods. Dextran can be identified in vesicles as a finely
granular, electron-dense material (34) and, as shown in Figure 4, was found in hepatocytes in numerous intracellular compartments including large electrolucent acid phosphatase-negative vesicles and multivesicular bodies. This granular, electron-dense material was not seen in identically processed control cells not exposed to dextran. The presence of dextran in acid phosphatase-negative vesicles in close proximity to acid phosphatase-positive vesicles (Le., lysosomes) makes it unlikely that the absence of acid phosphatase staining in these vesicles was attributable to technical problems such as variable penetration of fixative or stain. As compared with dextran, colloidal gold has a more characteristic appearance by electron microscopy (35) and can be seen even in organelles stained histochemically for lysosomal enzymes. As seen in Figure 5, colloidal gold, like dextran, is present predominantly in acid phosphatase-negative vesicles but can be seen in acid phosphatase-positive structures as well. Finally, to confirm the absence of lysosomal enzymes in at least some fluid-phase marker storage vesicles, we studied the subcellular distribution of dextran and colloidal gold in cells stained histochemically for a different
1258 LAKE ET AL.
GASTROENTEROLOGY Vol. 92, No.5, Part 1
Figure 5. Electron micrograph of hepatocytes in monolayer culture exposed for 16 h to media containing colloidal gold (18-20-nm particles), then washed and chased for varying periods (panel a, 1 h; panel b, 4 h; panel c, 6 h) in colloidal gold-free media. Colloidal gold particles are seen predominantly in acid phosphatase-negative (panel a and b) or arylsulfatase-negative (panel c) vesicles (solid arrows). Acid phosphatase- or arylsulfatase-positive vesicles (lysosomes) (open arrows) are present as well (magnification x35,OOO).
lysosomal enzyme, arylsulfatase, after the identical pulse-chase periods. Qualitatively similar results were obtained. These results confirm the observation made at the light microscopic level that histochemically identifiable lysosomes represent only a minority of the vesicles involved in the storage of dextran endocytosed 1-6 h earlier.
Discussion In these studies, we describe a method for visualizing acidic vesicles in cultured hepatocytes using the fluorescent weak base acridine orange and we have explored the identity of these acidic vesicles by comparing their number and intracellular distribution with the number and distribution of histochemically identifiable lysosomes as well as vesicles participating in the storage of FITC-dextran and colloidal gold, compounds that are taken up into hepatocytes by fluid-phase endocytosis. Before considering these findings, it is important to discuss the
validation and limitations of the methods used to identify the various vesicle populations. The use of acridine orange to identify acidic vesicles is based on its ability as a membrane-p81;meant weak base to concentrate in acidic compartments (22). Because acridine orange does not exhibit ideal weak-base behavior, the effects of protonophores and weak bases in this study are important in that they indicate that acridine orange staining of intracellular vesicles indeed reflects the presence of a proton gradient rather than nonspecific vesicle binding. Moreover, the effect of protonophores, which specifically enhance the proton conductance of biologic membranes, and the reversible nature of the effect of protonophores, as well as weak bases, indicate that the acidic interior represents ongoing active H+ transport rather than a preformed pH gradient or a Donnan effect (5,36). Dextran has been widely used as a marker of fluid-phase endocytosis by ourselves and others because it is known not to bind to specific cell receptors, exhibits cellular uptake that is linear with
May 1987
respect to extracellular concentration, and is not degraded in lysosomes (5,26,37,38). Moreover, we have recently demonstrated that dextran moves from perfusate to bile by a trans cellular vesicular mechanism in the isolated perfused rat liver in a concentration-independent fashion (24). Using the protocol employed in the present studies, dextran would be expected to be localized in storage compartments rather than early recycling compartments, an assumption supported by our observation that there were no changes in vesicle number or distribution with varying chase periods from 1 to 6 h (26,38). Indeed, our protocol is similar to that previously used to identify lysosomes (5,25,39). This has interesting implications that will be discussed below. Lysosomes in this study were identified at the light microscopic level by histochemical staining for acid phosphatase. Whereas it is conventional to define lysosomes based on their histochemicai properties (25), this approach may underestimate the actual number of lysosomal structures. For example, it requires the presence of acid phosphatase activity as well as sufficient resistance of the enzyme to fixation so as to permit histochemical identification. Our study design does not permit these assumptions to be directly tested; however, several staining protocols were tested so as to maximize the number of individual structures that could be visualized by light microscopy and complementary histochemical studies were performed at the electron microscopic level using two different histochemical stains for lysosomes as well. In addition, it is important to emphasize that the quantitation of vesicles in these studies is potentially limited by the microscopy technique used. For example, clathrin-coated vesicles from rat liver, which have been previously shown by us to possess an active acidification mechanism (1), are small (50-100 nm in diameter) (,*0) and would not be expected to be readily resolvable at the light microscopic level. The same considerations potentially apply to vesicles involved in fluid-phase endocytosis as well as to primary lysosomes. Finally, lightinduced excitation of fluorescent probes including fluorescein and acridine orange produces photobleaching of fluorescence and possible photo damage to cellular structures. Recognizing these potential problems, all measurements of vesicle number and distribution were made from photographs taken after only brief «5 s) preliminary inspection of the culture plate. This type of approach has also been used successfully by previous investigators working in other cell systems (37,41). With these considerations in mind, the present study demonstrates that living cultured hepatocytes contain an average of ~170 acidic intracellular ves-
ACIDIC VESICLES IN CULTURED HEPATOCYTES
1259
icles resolvable at the light microscopic level. These vesicles are distributed diffusely throughout the cell, with about one-half being localized to the perinuclear cytoplasm as assessed by two-dimensional morphometry. The acidified interior is attributable to ongoing active H+ transport, as would be anticipated based on work with isolated organelies, and cannot be explained by either a Donnan equilibrium or a preformed H+ gradient. As previously shown by us for isolated clathrin-coated vesicles (23), CPZ inhibited acidification of these vesicles in a concentration-dependent and reversible fashion. This is interesting in light of the known cholestatic effect of CPZ (42) and the possible importance of active H+ or HC0 3 - transport, or both, in bile formation (43). The present studies, however, permit no direct conclusions regarding the relationship of these observations to the effect of CPZ on bile formation. The present study also suggests that lysosomes account for only a minority of acidic intracellulpr compartments in hepatocytes. Whereas the considerations outlined above raise the possibility that histochemical staining may underestimate the number of lysosomes, we are not aware of any reason to believe that peripheral as opposed to perinuclear lysosomes would be selectively missed. On balance, it appears more likely that lysosomes are fewer.in number and more restricted in their distribution to the perinuclear cytoplasm than are acidic compartments. The findings thus indicate that acridine orange staining cannot be used as a specific marker of lysosomes. Our observations also suggest that acidic vesicles and vesicles involved in storage of fluid-phase markers represent overlapping populations and, indeed, that such storage compartments may account for a large proportion of nonlysosomal acidic compartments. this assertion is supported by the following observations: (a) isolated clathrin-coated vesicles and multivesicular bodies from rat liver, as well as endosomes from other tissues, all of which are likely involved in fluid-phase endocytosis, are actively acidified organelles (1,2,23,37,39,41,44); (b) vesicles involved in fluid-phase uptake by other epithelia have been shown to be acidic (45); and (c) the present studies indicate that the number and distribution of dextran-storing vesicles are similar to those of acidic vesicles. It is also likely that organelles other than those on the endocytic pathway, including some elements of the Golgi and endoplasmic reticulum, may account for some intracellular acidic vesicles in hepatocytes. These conclusions regarding the identity of nonlysosomal acidic organelles in hepatocytes must be considered tentative. It is of interest, however, that very similar findings have
1260 LAKE ET AL.
been reported for cultured human fibroblasts and a human hepatoma cell line using techniques that permit ultrastructural identification of acidified intracellular compartments (46,47). The discrepancy between the number and distribution of dextran-storing vesicles and the number of histochemically identifiable lysosomes is also of interest. Whereas the assumption that only lysosomes are involved in the storage of endocytosed material has been previously challenged and preliminary evidence that vesicles other than lysosomes participate in the storage of endocytosed material has recently been reported for other cell types (48,49), dextran-containing vesicles have frequently been defined as lysosomes and their properties, for example active acidification, attributed exclusively to lysosomes (5,25). In the present study, the majority of vesicles participating in the storage of internalized fluid-phase marker were found to be acid phosphatase-negative and aryl sulfatase-negative by light or electron microscopy, or both. Moreover, in separate studies (data not shown) the number of acid phosphatase-positive structures was not significantly different between hepatocytes that had or had not been exposed for 16 h to dextran. Thus, it does not appear that "induction" of lysosomes by dextran can account for the findings summarized in Table 1. On balance, it appears most reasonable to conclude that many vesicles, apart from lysosomes, participate in the storage of dextran. In any case, the present observations suggest that the use of dextran storage for functional identification of lysosomes merits reevaluation and verification. In addition to the implications of the findings already outlined, development of the technique used here sets the stage for the study of acidic vesicle trafficking in hepatocytes and the factors that regulate it. Our preliminary observations indicate that the same culture plate, mounted in a temperaturecontrolled flow-through chamber, can be repeatedly or continuously visualized by light or fluorescence microscopy while the perfusate is modified with respect to composition or pH, or both (50). The insertion into or removal from the plasma membrane of H+ -trans locating adenosine triphosphatase pump units via a vesicular shuttle mechanism involving vesicles with acidic interiors is believed to play a role in processes such as regulation of intracellular pH and transepithelial HC0 3 -IH+ transport in other tissues (37). Because the liver is an organ actively involved in endocytosis that is here demonstrated to have a large number of acidic intracellular compartments, dynamic studies of vesicle movement in living hepatocytes will be of considerable interest.
GASTROENTEROLOGY Vol. 92, No.5, Part 1
References 1. Van Dyke RW, Steer q, Scharschmidt BF. Clathrin-coated vesicles from rat liver: enzymatic profile and characterization of ATP-dependent proton transport. Proc Natl Acad Sci USA 1984;81:3108-12. 2. Van Dyke RW, Hornick CA, Belcher J, Scharschmidt BF, Havel RJ. Identification and characterization of ATPdependent proton transport by rat liver multivesicular bodies. J Bioi Chern 1985;260:11021-6. 3. Schneider DL. ATP-dependent acidification of membrane vesicles isolated from purified rat liver lysosomes. Acidification activity requires phosphate. J Bioi Chern 1983;258: 1833-8. 4. Zhang F, Schneider DL. The bioenergetics of Golgi apparatus function: evidence for an ATP-dependent proton pump. Biochem Biophys Res Commun 1983;114:620-5. 5. Tycko B, Keith CH, Maxfield FR. Rapid acidification of endocytic vesicles containing asialoglycoprotein in cells of a human hepatoma line. J Cell Bioi 1983;97:1762-76. 6. Hornick CA, Jones AL, Renaud G, Hradek G, Havel RJ. Effect of chloroquine on low-density lipoprotein catabolic pathway in rat hepatocytes. Am J Physiol 1984;246:G187-94. 7. Posner BI, Khan MN, Bergeron HM. Effect of chloroquine on the internalization of p25-insulin into subcellular fractions of rat liver. J Bioi Chern 1982;257:5789-99. 8. Dunn wA, Hubbard AL. Receptor-mediated endocytosis of epidermal growth factor (EGF) by hepatocytes in the perfused rat liver: ligand and receptor dynamics. J Cell Bioi 1984; 98:2148-59. 9. Maxfield FR. Weak bases and ionophores rapidly and reversibly raise the pH of endocytic vesicles in cultured mouse fibroblasts. J Cell Bioi 1982;95:676-81. 10. Harford J, Wolkoff AW, Ashwell G, Klausner RD. Monensin inhibits intracellular dissociation of asialoglycoproteins from their receptor. J Cell Bioi 1983;96:1824-8. 11. Tartakoff AM. Perturbation of vesicular traffic with the carboxylic ionophore monensin. Cell 1983;32:1026-8. 12. Schwartz AL, Bolognesi A, Fridovich SE. Recycling of the as ialogl yeoprotein receptor and the effect of lysosomotropic amines in hepatoma cells. J Cell Bioi 1984;98:732-8. 13. Octave IN, Schneider YJ, Trovet A, Crichton RR. Iron uptake and utilization by mammalian cells. I. Cellular uptake of transferrin and iron. Trends Biochem Sci 1983;8:217-20. 14. Marsh M, Bolzau E, Helenius A. Penetration of Semliki Forest Virus from acid prelysosomal vesicles. Cell 1983;32:931-40. 15. Helenius A, Mellman I, Wall D, Hubbard A. Endosomes. Trends Biochem Sci 1983;8:245-50. 16. Bissell DM, Hammaker LE, Meyer UA. Parenchymal cells from adult rat liver in non-proliferating monolayer culture. I. Functional studies. J Cell Bioi 1973;59:722-34. 17. Scharschmidt BF, Stephens JE. Sodium, chloride and taurocholate transport by cultured rat hepatocytes. Proc Nat! Acad Sci USA 1981;78:986-90. 18. Geoghegan WD, Ackerman GA. Adsorption of horseradish peroxidase, ovomucoid and anti-immunoglobulin to colloidal gold for the indirect detection of concanavalin A, wheat germ agglutinin and goat anti-human immunoglobulin G on cell surfaces at the electron microscopic level: a new method, theory and application. J Histochem Cytochem 1977;25: 1187-200. 19. Barka T, Anderson PJ. Histochemical methods for acid phosphatase using hexazonium pararosanilin as coupler. J Histochem Cytochem 1962;10:741-53. 20. Robinson JM, Karnovsky MJ. Ultrastructural localization of several phosphatases with cerium. J Histochem Cytochem 1983;31:1197-208.
May 1987
21. Bentfeld-Barker ME, Bainton DF. Cytochemical localization of arylsulfatase B in rat basophils and mast cells. J Histochem Cytochem 1980;28:1055-61. 22. Lee HC, Forte JG. A study of H+ transport in gastric microsomal vesicles using fluorescent probes. Biochim Biophys Acta 1978;508:339-45. 23. Van Dyke RW, Scharschmidt BF, Steer q. ATP-dependent proton transport by isolated brain clathrin-coated vesicles. Role of clathrin and other determinants of acidification. Biochim Biophys Acta 1985;812:423-36. 24. Lake JR, Licko V, Van Dyke RW, Scharschmidt BF. Biliary secretion of fluid phase markers by the isolated perfused rat liver: the role of transcellular vesicular transport. J Clin Invest 1985;76:676-84. 25. Ohkuma S, Poole B. Fluorescein probe measurement of the intralysosomal pH in living cells and the perturbation of pH by various agents. Proc Natl Acad Sci USA 1978;75:3327-31. 26. Thomas JA, Buchsbaum RN, Zimniak A, Racker E. Intracellular pH measurements in Ehrlich ascites tumor cells utilizing spectroscopic probes generated in situ. Biochemistry 1979;19:2210-8. 27. Besterman JM, Airhart JA, Woodworth Re, Low RB. Exocytosis of pinocytosed fluid in cultured cells: kinetic evidence for rapid turnover and compartinentation. J Cell BioI 1981;91:716-27. 28. Storrie B, Pool RR, Sachdeva M, Maurey KM, Oliver C. Evidence for both pre-lysosomal and lysosomal intermediates in endocytic pathways. J Cell BioI 1984;98:108-15. 29. Herman B, Albertini DF. Ligand-induced rapid distribution of Iysosomes is temporally distinct from endosome translocation. Nature 1983;304:738-40. 30. Herman B, Albertini DF. A time-lapse video image intensification analysis of cytoplasmic organelle movements during endosome translocation. J Cell BioI 1984;98:565-76. 31. Robbins E, Marcus PI, Gonatas NK. Dynamics of acridine orange-cell interaction. II. Dye-induced ultrastructural changes in multivesicular bodies (acridine orange particles). J Cell BioI 1964;21:49-62. 32. Sewell RB, Dillon C, Grinpukel S, Yeomans ND, Smallwood RA. Peri canalicular location of hepatocyte Iysosomes and effects of fasting: a morphometric analysis. Hepatology 1986; 6:305-11. 33. Davies P, Allison AC, Haswell AD. The quantitative estimation of pinocytosis using radioactive colloidal gold. Biochem Biophys Res Commun 1973;52:627-34. 34. Ainsworth SK. An ultrastructural method for the use of polyvinylpyrrolidone and dextrans as electron opaque tracers. J Histochem Cytochem 1977;25:1254-9.
ACIDIC VESICLES IN CULTURED HEPATOCYTES
1261
35. de Bruijn WC, Schellens JPM, van Buitenen JMH, van der Meulen J. X-ray microanalysis of colloidal-gold-labelled Iysosomes in rat liver sinusoidal cells after incubation for acid phosphatase activity. Histochemistry 1980;66:137-48. 36. Yamashiro DJ, Fluss SR, Maxfield FR. Acidification of endocytic vesicles by an ATP-dependent proton pump. J Cell BioI 1983;97:929-34. 37. Gluck S, Cannon C, AI-Awqati Q. Exocytosis regulates urinary acidification in turt!e bladder by rapid insertion of H+ pumps into the luminal membrane. Proc Nat! Acad Sci USA 1982;79:4327-31. 38. Scharschmidt BF, Lake JR, Renner EL, Licko V, Van Dyke RW. Fluid phase endocytosis by cultured rat hepatocytes and perfused rat liver. Proc Nat! Acad Sci USA 1986:83:9488-92. 39. Ohkuma S, Moriyama Y, Takano T. Identification and characterization of a proton pump on Iysosomes by fluorescein isothiocyanate-dextran fluorescence. Proc Nat! Acad Sci USA 1982;79:2758-62. 40. Steer q, Klausner RD. Clathrin-coated pits and coated vesicles: functional and structural studies. Hepatology 1983;3: 437-54. 41. Tycko B, Maxfield FR. Rapid acidification of endocytic vesicles containing o2-macroglobulin. Cell 1982;28:643-51. 42. Bartholomew LG, Cain JC, Frazier SH, et al. Effect of chlorpromazine on the liver. Gastroenterology 1958;34:1096-107.' 43. Scharschmidt BF, Van Dyke RW. Mechanisms of hepatic electrolyte transport. Gastroenterology 1983;85:1199-214. 44. Galloway CJ, Dean GE, Marsh M, Rudnick G, Mellman I. Acidification of macrophage and fibroblast endocytic vesicles in vitro. Proc Nat! Acad Sci USA 1983;80:3334-8. 45. Gluck S, Kelly S, AI-Awqati Q. The proton trans locating ATPase responsible for urinary acidification. J BioI Chern 1982;257:9230-3. 46. Anderson RGW, Falck JR, Goldstein JL, Brown MS. Visualization of acidic organelles in intact cells by electron microscopy. Proc Nat! Acad Sci USA 1984;81:4838-42. 47. Schwartz AL, Strous GJAM, Slot JW, Gevze HJ. Immunoelectron microscopic localization of acidic intracellular compartments in hepatoma cells. EMBO J 1985;4:899-904. 48. McKinley DN, Wiley HS. The pathway and kinetics of fluid phase endocytosis in fibroblasts (abstr). J Cell BioI 1984; 99:376a. 49. Hohman TC, Bowers B. Phagosome-lysosome fusion is a function of phagocytic load (abstr). J Cell BioI 1984;99:363a. 50. Lake JR, Van Dyke RW, Scharschmidt BF. Visualization of vesicles in cultured hepatocytes: description of a technique and preliminary findings (abstr). Clin Res 1984;32:76A.