Rapid acidification of endocytic vesicles containing α2-macroglobulin

Cell, Vol. 28. 643-651,

March

1982,

Copyright

0 1982 by MIT

Rapid Acidification of Endocytic Containing cu,-Macroglobulin Benjamin Tycko and Frederick FL Maxfield Department of Pharmacology New York University Medical New York, New York 10016

Center

Summary We have used fluorescein-labeled armacroglobulin (F-&A) to measure pH changes in the microenvironment of internalized ligands following receptor-ma diated endocytosis. Fluorescence intensities of single BALB/c 3T3 mouse fibroblasts were measured by using a microscope spectrofluorometer with narrow bandpass excitation filters. The pH was determined from the ratio of fluorescein fluorescence intensities with 450 nm and 490 nm excitation. A standard pH curve was obtalned by incubating cells wlth F-a&l for 30 min at 37°C followed by fixation and lncubatlon in buffers of varying pH. To measure the pH of endocytic vesicles, cells were incubated wlth F-&f for 15 mln at 37%. Fluorescence intensities were measured on living cells within 5 min of rinsing. Under these conditions, the pH of the F-erM microenvironment was 5.0 f 0.2. Using colloidal gold+M for electron microscopic localization8 we have verified that, under these conditions, a&f is predominantly in uncoated vesicles that are negative for acid phosphatase activity. With further incubation for l/2 hr, we obtained a pH of 5.0 f 0.2 for the F-a&l. Using fluoresceln dextran, we obtained a iysosomai pH of 4.6 f 0.2. These results indicate that endocytlc vesicles become acidic prior to fusion with lysosomes. introduction Certain proteins, hormones, viruses and toxins that are known to bind to specific receptors on the cell surface have been shown to enter cells through a common pathway of receptor-mediated endocytosis (Goldstein et al., 1979, Pastan and Willingham. 1961). This pathway involves clustering of ligand-receptor complexes over clathrin-coated pits followed by internalization of the ligand in endocytic vesicles. An incomplete list of ligands that have been shown to enter cells through this pathway includes epidermal growth factor (Maxfield et al., 1978; Schlessinger et al., 1978, Haigler et al., 19791, insulin (Maxfield et al., 1978; Schlessinger et al., 19781, cm-macroglobulin (Maxfield et al., 1978; Willingham et al., 19791, low density lipoproteins (Brown and Goldstein, 1979; Goldstein et al., 19791, 3.3’,5-triiodo-L-thyronine (Cheng et al., 1980; Maxfield et al., 1981), immunoglobulins (Rodewald, 1980; Salisbury et al., 1980). asialoglycoproteins (Wall et al., 19801, viruses (Helenius et al., 1960; Dickson et al., 1981; Wolf et al., 1981) diphtheria toxin (Keen et al., 1982) and Pseudomonas exotoxin (FitzGerald et al., 1980).

Vesicles

The serum protein cun-macroglobulin &M) is internalized by fibroblasts (Maxfield et al., 1976; Van Leuven et al., 1978) and macrophages (Kaplan and Nielsen, 1979) through a receptor-mediated pathway. The large size of this protein, which is a tetramer with a total molecular weight of 780,000 daltons (780 kd; Frenoy et al., 19771, permits extensive labeling of the molecules with fluorescent chromophores. Studies of the internalization of fluoresceinor rhodamine-iabeled &I (Maxfield et al., 1978) and electron microscope studies with aZM labeled by peroxidase or ferritin conjugation or by adsorption to colloidal gold (Willingham et al., 1979; Dickson et al., 1981) have elucidated some of the ultrastructural details of both the internalization process and the pathway of ligand transport within the ceil subsequent to internalization. These studies have shown that (YAM is cleared from the cell surface of 3T3 mouse fibroblasts with a t1,2 of less than 10 min at 37°C. Initially, the internalized (Y*M appears in electron-lucent vesicles, which have been called “receptosomes” (Willingham and Pastan, 1980). Between 30 and 60 min after internalization the labeled a2M appears in small lysosomes. Similar pathways have been observed for other ligands (Goldstein et al., 1979; Wall et al., 1980). Changes in pH have profound effects on the interaction of many ligands with the cell surface. Several ligands, including epidermal growth factor (Haigler et al., 19801, insulin (Posner et al., 19771, lysosomal enzymes (Gonzalez-Noriega et al., 1980) and asialoglycoproteins (Ashwell and Morell, 1974) are released from their receptors when the pH of the medium is dropped below pH 5.5. The entry of diphtheria toxin (Draper and Simon, i980; Sandvig and Olsnes, 19801, Semliki Forest virus (Helenius et al., 19801, influenza virus (White et al., 19811, and vesicular stomatitis virus (White et al., 1981) into the cytoplasm is facilitated at a low pH. We have used fluorescein-labeled azM (F-asM) to measure the pH of the microenvironment of ligands undergoing receptor-mediated endocytosis. The excitation spectrum of fluorescein is strongly pH-dependent between pH 4 and pH 7.4, and fluorescently labeled compounds have been used previously as probes for the pH of lysosomes (Ohkuma and Poole, 1978; Poole and Ohkuma, 19811, the sperm acrosome (Working and Meizei, 19811, phagocytic vacuoles (Geisow et al., 1981; Segal et al., 1981) and the cytoplasm (Thomas et al., 1979; Heiple and Taylor, 1980). in this paper we present evidence for rapid acidification of primary endocytic vesicles prior to fusion of these vesicles with lysosomes.

Results Measurement of Lysosomal pH in BALB/c 3T3 Fibroblasts The excitation spectrum of fluorescein in solution is extremely sensitive to changes in pH (Ohkuma and

Cdl 644

Poole, 1978). In particular, the 450/490 excitation ratio, defined as the ratio of fluorescence intensity, measured at an emission wavelength of 520 nm, with excitation at 450 nm to that measured with excitation at 490 nm, is found to be highly pH-dependent between pH 4.0 and pH 7.4 (Figure 1). We have used fluorescein-conjugated dextran (F-dextran) as an indicator of the pH of BALB/c 3T3 fibroblast lysosomes. As is shown in Figure 1, a change in pH from pH 7.4 to pH 5.0 corresponds to a greater than threefold increase in the 450/490 excitation ratio of fluorescein fluorescence. This change in the excitation ratio is accompanied by a 45% reduction in the fluorescence excited at 450 nm and an 82% reduction in the 490 nm fluorescence between pH 7.4 and pH 5.0. Cells were allowed to internalize F-dextran by pinocytosis continuously for 18 hr at 37X, followed by rinsing in Dulbecco’s modified Eagle’s medium containing HEPES (DMEM/HEPES) and a 30 min chase in serum-free medium at 37X. At this time fluorescence is localized (within the cell) entirely in round

I 4

5

6

7

8

PH Figure

1. The pH-Dependence

of F-&f

Fluorescence

Cells were incubated with F-&A (60 pg/ml) in DMEM/HEPES for 30 min at 37’C, followed by rinsing and fixation in 2% formaldehyde and PBS. Fixed cells were then equilibrated for 1 hr in a series of buffers: 0.1 M sodium acetate (pH 4.0, pH 5.0). 0.1 M sodium phosphate (pH 6.0, pH 7.4) and 0.1 M Tris (pH 6.0). Fluorescence intensities at 450 nm and 490 nm excitation wavelengths were measured by microfluorometry. as described in Experimental Procedures. Data represent the mean of measurements from ten individual cells at each pH. corrected for cellular autofluorescence. The pH dependence of the IrJo/lrp~ ratio of F-&A and F-dextran in solution was also measured and did not differ significantly from that of cell-associated F-a2M (data not shown).

phase-contrast dense structures, which can be shown by cytochemical methods to correspond to lysosomes (Willingham and Yamada, 1978). Since dextrans are not degraded by lysosomal enzymes, the 450/490 excitation ratio provides a measure of lysosomal pH (Ohkuma and Poole, 1978). After incubation with F-dextran, cells were prepared for quantitative fluorescence microscopy as described in Experimental Procedures. A series of three experiments, each consisting of ten whole-cell measurements, gave a value for lysosomal pH of 4.6 f 0.2 (SE). The standard error represents the differences among the average pH values obtained in the three experiments. This pH value agrees well with values obtained by others for lysosomal pH in fibroblasts (Hollemans et al., 1981) and macrophages (Ohkuma and Poole, 1978). As a control, cells were fixed in 2% formaldehyde and phosphate-buffered saline (PBS; pH 7.4) followed by addition of chloroquine (500 $vt) to dissipate any pH gradients. After equilibration for 30 min at 23°C the measured pH of the F-dextran microenvironment had returned to pH 7.4. Measurement of the pH of the Microenvironment of internalized Fluorescein-a2-Macroglobulin. Fluorescein-conjugated an-macroglobulin was used as a probe for the pH of the microenvironment of internalized (YAM. The pH dependence of the 450/490 excitation ratio of cell-associated F-&l is shown in Figure 1. Measurements of the 450/490 excitation ratio of FlrPM in living cells were carried out under three different incubation conditions (Table 1). In experiment I, cell monolayers were exposed to F-(Y~M (80 pg/ml) in PBS containing CaC12 (1 mfvt) at 4°C for 1 hr. Cells were then rinsed three times in cold PBS and Ca2+ and immediately fixed in 2% formaldehyde and PBS. The 450/490 excitation ratio was measured from ten individual cells and corrected for cellular autofluorescence as described in Experimental Procedures. In experiment II, cells were incubated with F-a&t (80 pg/ml) in DMEM/HEPES at 37’C for 15 min. Cells were then rinsed three times with DMEM/ HEPES (37’C), incubated further for a chase period of 4 min at 37’C and rinsed twice in PBS, Ca2+ and glucose. In each replicate of this experiment, the 450/ 490 excitation ratio was measured from six individual live cells and corrected for cellular autofluorescence, as described in Experimental Procedures. The total chase time prior to observation was between 5 and 6 min. In experiment Ill, cells were treated as in experiment II. except that the chase time was extended to 30 min. In all these experiments, the ratio of F-cr2M fluorescence to background cellular autofluorescence and scattered light varied between 0.8 and 1 .O. As shown in Table 1, the microenvironment of F(YAM is acidified quite rapidly after entry of the ligand

Acidification 645

Table

of Endocytic

Vesicles

1. The pH of the Microenvironment

Experiment

Incubation

I

F-a&f

4°C. 60 min; rinsed

II

F-c&J

37’C.

of Internalized

Flourescein-Labeled

Condition9

15 min; rinsed;

un-Macroglobulin Observation

37°C

pH of F-a& Microenvironment’

Fixed

7.4 (n = 1)

5 min

Live cells

5.0 f 0.2 (n = 4) 5.0 f 0.2 (n = 4) 7.2 f 0.2 (n = 3)

4’ DMEM

Condition@

Ill

F-&t

37’C.

15 min; rinsed;

DMEM

37°C

30 min

Live cells

IV

F-wM

37%

30 min; rinsed;

DMEM

37%.

5 min

Fixed, equilibrated in pH 7.4 buffer with 0.5 mM chloroquine

’ All incubations were at 60 &ml F-WM. In Experiment I the incubation buffer was PBS containing 1 mM CaCl* and 1 mg/ml BSA. In experiments II-IV, incubations were in DMEM/HEPES. ’ All observations were made at 37°C. as described in Experimental Procedures. ’ The values in parentheses are the number of replicate experiments. Measurements were made on six to ten cells in each experiment. The pH values were obtained from 1,,0/11, using Figure 1.

into the ceil. After 15 min of continuous uptake of FanM at 37X followed by a 5 min chase at 37’C, the pH of the F+M microenvironment is 5.0 f 0.2 (SE), quite close to the measured lysosomal pH of 4.6 f 0.2 (SE). Furthermore, there is no measurable decrease in pH when the chase time is extended to 30 min. Phase-contrast and fluorescence images of fibroblasts incubated under the conditions of experiment II are shown in Figure 2. Figure 2a shows the phasecontrast image of a cell that has been incubated under the conditions of experiment II and then fixed in 2% formaldehyde and PBS and equilibrated to pH 7.4. Figures 2b and 2c are fluorescence images of the same cell with excitation at 450 nm and 490 nm, respectively. Figures 2d, 2e and 2f show the corresponding images of a living cell. Note that, in addition to the marked decrease in the 450/490 excitation ratio on fixation at pH 7.4, there is an absolute increase in the fluorescence intensity at the 490 nm excitation wavelength. Time Course of Degradation of ‘251-a2Macroglobulin by BALB/c 3T3 Fibroblasts The finding that the microenvironment of F-(Y*M is rapidly acidified after endocytosis suggested the possibility that the interior of primary endocytic vesicles is acidified independently of fusion of these vesicles with lysosomes. It was therefore of interest to determine the time course of degradation of ‘251-~2M. The time course of degradation of internalized ‘?anM is shown in Figure 3. Cell monolayers were incubated with 12?-a2M for 15 min at 37°C. Cells were then rinsed twice and incubated further for 5 min. In Figure 3. the zero time point represents the end of this initial 5 min chase. Trichloroacetic acid- (TCA)-precipitable material was determined after additional chase times. Data points represent the mean of three experiments. BALB/c 3T3 fibroblasts degrade ‘251-a2M slowly: more than 80% of the material remains TCA-precipitable after a 1 hr chase. These results are consistent

with a nonlysosomal location for F-cu2M under the conditions of experiment II. The kinetics of degradation of ‘251-a2M were reproducibly nonlogarithmic. In a typical experiment, 25% of the radioactive counts remained TCA-precipitable after a 6 hr chase at 37’C (data not shown). The intracellular localization of this a2fvl is not known. A similar finding has been reported for degradation of rhodamine-a2M by Swiss 3T3 fibroblasts, but rhodamine-EY2M is rapidly degraded by normal rat kidney (NRK) cells (Maxfield et al., 1981). The possibility that (YAM might enter the lysosomal compartment under the conditions of experiment II but be relatively resistant to degradation by lysosomal proteases was ruled out by light and electron microscope studies as described below. Although these degradation experiments do not, by themselves, rule out exposure to lysosomal hydrolases, they do indicate that the F-(u~M molecules we are observing are intact. This would not be the case if NRK cells were used (Maxfield et al., 1981 a). Localization of Rhodamine-a2-Macroglobulin in Fibroblasts Stained for Acid Phosphatase BALB/c 3T3 fibroblasts were incubated with rhodamine-a& for 15 min at 37’C, followed by a 5 min chase at 37°C (the conditions of experiment II, Table 1). The cells were fixed and stained for acid phosphatase activity with the use of cytidine monophosphate as a substrate. Cells were viewed by bright-field and fluorescence microscopy (Figure 4). This staining procedure reveals a population of large lysosomes concentrated about the nucleus. In Figure 4, and in every other cell examined, the pattern of rhodamine-a?M fluorescence does not overlap with the pattern of large lysosomes. In no case was a bright dot seen by fluorescence found to correspond to an acid phosphatase-positive structure. These results are consistent with an extralysosomal location for a2M under the conditions of experiment II, Table 1. However, the possibility that a2M was contained in small (200 to 300 nm) lysosomes could not be excluded by this procedure.

Cdl 646

20 I

1

TIME Figure Figure

2. F-a2M in Endocytic

Vesicles

Cells were incubated with F-a&f for 15 min at 37OC followed by a 5 min chase, as in experiment II of Table 1, and were observed after fixation in 2% formaldehyde and equilibration to pH 7.4 in PBS with 0.5 mM chloroquine (a-c) or live in PBS with Ca*+ and glucose (d-e). (a) Phase-contrast image of a fixed cell. (b and c) Fluorescein fluorescence images of the same cell with excitation at 450 nm or 490 nm. respectively. (d-f) Same as (a-c) but on a living cell. In the fixed cell equilibrated to pH 7.4, the F-a&I fluorescence with 490 nm excitation (c) is much brighter than with 450 nm excitation (b). In the living cell, the fluorescence with 490 nm excitation (0 has been reduced to approximately the same level as the 450 nm excitation (e) as expected at pH values near 5. All fluorescence images were obtained with the television camera operating at the same gain. Images were recorded on videotape and photographed from the monitor during playback. Bar = 10 PM.

Subcellular Localization of wMacroglobulin Adsorbed to Colloidal Gold To establish with more confidence that, under the conditions of experiment II (Table l), F-(w&l is present mainly in primary endocytic vesicles, we have incubated BALB/c 3T3 fibroblasts with (Y&I adsorbed to colloidal gold &M-gold) and examined thin sections by electron microscopy. The application of the colloidal gold technique to studies of the subcellular localization of internalized apM has been described by Dickson et al. (1981). Briefly, anM is adsorbed to colloidal gold particles with a size distribution of 10 to 20 nm. These cy,M-gold particles have been shown to bind specifically to the (Y*M receptor of the plasma membrane, to cluster over coated pits and to be internalized in electron-lucent

3. Degradation

3

2

of ‘25la&t

(hours)

by BALE/c

3T3 Fibroblasts

Cell monolayers were incubated in DMEM/HEPES (0.1% BSA) with “‘I-a2M (0.1 Pg/ml) for 15 min at 37°C. rinsed and chased for an additional 5 min at 37°C. as in experiment II, Table 1. The zero time point represents the end of this initial pulse-chase. Cells were rinsed again and incubated further for the indicated times at 37’C in DMEM/ HEPES. Trichloroacetic acid-precipitable radioactive counts were determined at the indicated times. All measurements were made in duplicate. Data shown represent the mean of three experiments. In a typical experiment, cells had internalized 2 ng of ‘*‘I-a2M per mg of cell protein by the end of the initial pulse-chase.

endocytic vesicles (Willingham and Pastan, 1980). The time course of internalization has been shown to be identical to that of aaM-ferritin and unlabeled ~llpM localized by the immunoperoxidase method (Pastan and Willingham, 1981 I. When BALB/c 3T3 fibroblasts are incubated in the presence of cynM-gold (30 pg anM/ml) under conditions identical to those of experiment II (Table l), greater than 90% of the internalizal gold particles are seen in electron-lucent vesicles, while less than 10% of the particles are seen in lysosomes. Sections of cells stained for acid phosphatase activity (Figure 5) show no detectable activity in endocytic vesicles. This finding strongly suggests that, under .the conditions of experiment II (Table l), the internalized F-a2M is present largely in endocytic vesicles that have not yet fused with lysosomes. Discussion A great variety of molecules, including certain serum proteins, hormones, viruses and toxins that can bind

Acidification 647

of Endocytic

Figure 4. Locakation Phosphatase

Vesicles

of Rhodamine-&t

in a Cell Stained

for Acid

Cells were incubated with rhodamintizM (80 Fg/ml) for 15 min at 37OC. followed by a 5 min chase at 37% (the conditions of experiment II, Table 1). Cells were then Rxed and stained for acid phosphatase activity with the use of @dine monophosphate as a substrate. (a) Bright-field Image of a flbrobfast stained for acid phosphatase. (b) Fluorescence image of the same cell. Lysosomes are seen clustered about the nucleus, while endocytic vesicles containing rhodaminec&t are seen scattered throughout the cell. The box is placed in the same position in both fields. No overlap between the pattern of rhodamine-a~M fluorescence and the pattern of acid phosphatase activity was seen in any of more than 20 cells examined. Bar = 6 FM.

with high affinity to specific receptor sites on the cell membrane are internalized by cells through a shared pathway, which has been termed receptor-mediated endocytosis (Goldstein et al., 1979; Helenius et al., 1980; Wall et al., 1980; Pastan and Willingham, 1981; Wolf et al., 1981). This process is characterized by an initial binding of the ligand to its receptor, clustering of ligand-receptor complexes over clathrin-coated pits and internalization of the ligand in primary endocytic vesicles (Dickson et al., 1981; Pastan and Willingham, 1981). These vesicles, which have a characteristic appearance when observed by electron microscopy (Dickson et al., 1981), have been shown to eventually fuse with small lysosomes (Wall et al., 1980; Willingham and Pastan, 1980; Dickson et al., 19811,

thereby initiating degradation of the ligand (Wall et al., 1980). The large serum protein ewe-macroglobulin, a protease inhibitor whose exact biological function is unknown, is internalized by fibroblasts (Van Leuven et al., 1978; Pastan and Willingham, 1981) and macrophages (Kaplan and Nielsen, 1979) through a receptor-mediated pathway. In this study, we have used fluorescein-conjugated an-macroglobulin as a probe for the pH of the microenvironment of internalized a&l. We have found that, in BALB/c 3T3 fibroblasts, the microenvironment of F-(YM is acidified quite rapidly after entry of the labeled protein into the cell. When cells are allowed to internalize F-&l continuously for 15 min at 37’C, followed by a 5 min chase at 37°C has a measured pH of the F-&l microenvironment 5.0 f 0.2 (SE). This value does not decrease after an additional 25 min chase and is quite close to the measured lysosomal pH of 4.6 f 0.2 (SE). Although it would be extremely useful to obtain pH measurements at earlier times, we were unable to do this reliably. The intensity of fluorescein fluorescence drops as the pH is lowered, and a 15 min continuous uptake was required to obtain a signal that would yield valid pH measurements. Also, the 4 min chase before starting measurements was necessary to reduce nonspecific binding. Despite these limitations, our results demonstrate that the vast majority of endocytic vesicles are acidic. We can estimate an upper limit for the percentage of vesicles at neutral pH by calculating the value of the 450/490 excitation ratio for various mixtures of F-cmM at pH 4.6 (lysosomal PHI and pH 7.4 (extracellular pH). If 10% of the F-(Y*M were in a pH 7.4 environment with the remainder at pH 4.6, the measured pH for the mixture would be 5.3; if 20% of the F+M were at pH 7.4, the measured pH would be 6.0. Since we measure a pH of 5.0 after a 15 min incubation and 5 min chase, we estimate that less than 10% of the F-(Y~M is still at pH 7.4. This means that even the F-(u~M bound to the cell surface at the end of the incubation period must be mostly acidified by the time the measurements are made. Our studies of the degradation of ‘251-~2M suggest that, after a 15 min pulse and a 5 min chase, internalized ligand is not being degraded at a significant rate. We cannot rule out the possibility that CX~M is degraded slowly even after entering lysosomes. However, light microscopic studies and electron microscopic ultrastructural studies using clr2M adsorbed to colloidal gold as an electron-dense tracer have clearly demonstrated that, under these same conditions, the internalized a2M is predominantly localized in endocytic vesicles. These vesicles do not contain detectable levels of acid phosphatase activity. We conclude that primary endocytic vesicles are acidified prior to and independent of fusion of these vesicles with lysosomes.

Cell 640

Figure

5. Endocytic

Vesicle

Containing

alM Adsorbed

to Colloidal

Gold. Stained

for Acid Phosphatase

Cell monolayers were incubated with a&l-gold at an anM concentration of 30 pg/ml under the conditions of experiment II. Table 1: a 15 min/b min pulse-chase at 37OC. Cells were fixed in 2% glutaraldehyde and stained for acid phosphatase activity with the use of cytidine monophosphate as a substrate. Cells were postfixed in osmium tetroxide and viewed without further staining. Endocytic vesicles containing cr*M-gold (arrows) were consistently negative for acid phosphatase. By comparison, lysosomes included in the field stain positively for acid phosphatase. (A) Several lysosomes in the same field as an endocytic vesicle. The arrowhead shows two gold particles that have entered a lysosome. (6) A higher magnification image of a single endocytic vesicle in close proximity to a lysosome. Bars = 0.2 pM.

Acidification 649

of Endocytic

Vesicles

The interaction of many ligands with their receptors or with the plasma membrane is markedly influenced by changes in pH. The hormones insulin (Posner et al., 1977) and epidermal growth factor (Haigler et al., 1979) lysosomal enzymes (Gonzalez-Noriega et al., 1980) and asialoglycoproteins (Ashwell and Morell, 1974) all undergo rapid dissociation from their receptors if the pH is dropped below pH 5.5. Our results suggest that acidification of primary endocytic vesicles permits ligand-receptor dissociation prior to entry of the ligand into the degradative compartment of the ceil. This process may be intimately coupled with recycling of functional receptors back to the plasma membrane or to some other compartment of the cell. A similar model has been proposed to explain the effects of lysosomotropic amines, which dissipate the lysosomal pH gradient, on the secretion and endocytosis of lysosomal enzymes (Gonzalez-Noriega et al., 1980; Stahl et al., 1980). The fusion of the membrane of Semliki Forest virus with the plasma membrane and entry of the viral nucleocapsid into the cytoplasm can be induced by dropping the pH of the external medium below pH 6.0 (White et al., 1981). In addition, the fusion of vesicular stomatitis virus and influenza virus with the plasma membrane is facilitated at pH 5.0 (White et al., 1981). A similar pH dependence has been described for entry of the A subunit of diphtheria toxin into cells (Draper and Simon, 1980; Sandvig and Olsnes, 1980; Sandvig and Olsnes, 1981). These viruses (Dales and Choppin, 1962; Patterson et al., 1979; Helenius et al., 1980; Dickson et al., 1981) and diphtheria toxin (Keen et al., 1982) have all been shown to enter cells through the receptor-mediated endocytic pathway. In light of the results presented here, it is likely that the biologically effective viruses and toxin chains gain access to the cytoplasm directly from acidified primary endocytic vesicles, thus circumventing the degradative machinery of the cell. Experimental

Procedures

Cell Culture BALB/c 3T3 cells were obtained from Ira Pastan and grown in Dulbecco-Vogt’s medium containing 10% (v/v) calf serum. Experimental cultures were plated in 35 mm Falcon tissue culture dishes in 2 ml of medium at a density of 1.5 X 1 O5 cells per dish. Cells for fluorescence microscopy were plated on 22 mm glass coverslips. Cultures were used two to three days after plating. Confluent cell monolayers were used in studies of ‘2Jl-a2M binding and degradation; cultures for fluorescence microscopy were approximately two thirds confluent at the time of observation. Preparation of Fluorescetn-Conjugated nrMacroglobulin c&facroglobulin was prepared from whole human plasma by a slight modification (Willingham et al., 1979) of the procedure of Wickerhauser and Hao (1972). Fluorescein-conjugated OLM was prepared as described by Maxfield et al. (I 976). Preparation of ‘*2-a~M azM was iodinated by the chloramine-T procedure (Masher et al., was 10’ cpm/Ag. Radioactive 1977). Specific activity of “?a2M counts were greater than 90% TCA-precipitable after separation from

free Na”‘l

by chromatography

over a Sephadex

G25 column.

Binding and Degradation of ‘2’l-a2M Confluent cell monolayers were incubated at 37°C with ‘25l-azM (0.1 pge/ml) in DMEM containing 20 mM HEPES (DMEM/HEPES) and 0.1% bovine serum albumin (BSA). At the specified times, cells were rinsed twice with DMEM/HEPES and incubated further for the specified chase times. Cells were again rinsed twice with DMEM/HEPES. Trichloroacetic acid (10% v/v. one ml) was added to the dish and allowed to stand for 5 min. TCA was then drawn off and the precipitated material was taken up in 1 N sodium hydroxide. Radioactive counts were measured in a Beckman 4000 gamma counter. All measurements were made in duplicate. Quantitative Fluorescence Microscopy Fluorescence intensities of single cells were measured with the use of a Leitz MPV Compact microscope spectrofluorometer mounted on a Leitz Orthoplan fluorescence microscope. The microscope was equipped with a 75W xenon lamp and with exchangeable 450 nm and 490 nm narrow bandpass filters (Edmund Scientific) for incident light and a 520 nm bandpassfilter for emitted light. A 63x (N. A. 1.3) objective was used, and a circle of approximately 20 pM diameter was illuminated. During the period of illumination (0.25 set) the loss of fluorescence intensity due to photobleaching was not more than 10% of the total. Before observation, coverslip cultures were washed in PBS containing CaCb (1 mM) and glucose (2 mg/ml). Coverslips were then inverted over depression slides filled with PBS, Ca2+ and glucose. The microscope stage was maintained at 37’C by a Sage Air-Curtain Incubator. The background of cellular autofluorescence was determined in each experiment by averaging intensity measurements from ten individual cells that had not been exposed to F-a2M. Competition with a 60-fold excess of unlabeled uzM indicated that nonspecific binding of F-e&t was less than 10% of the total. Lysosomal pH was measured by incubating cells for 16 hr in the presence of fluorescein dextran (Sigma molecular weight 40,000, 1 mg/ml). Coverslips were then rinsed twice in DMEM/HEPES and incubated for an additional 30 min prior to observation. Fluorescence intensities of single cells were then measured as described above. Fluorescence measurements on solutions of F-oI~M or F-dextran were made by placing the solutions in a cell counting chamber and using the microscope spectrofluorometer as described by Maxfield et al. (196la). Video Intendtlcation Microecopy The microscope used for microfluorometry was also equipped with an image intensifier television camera (Dage/MTl , model 65) with selectable manual or automatic gain control. Video images were stored on a videotape recorder (Panasonic NV 6030) and photographs were made of the video monitor during playback. Acid Phosphatase Staining for Light Microscopy Cells were stained for acid phosphatase activity and observed by light microscopy as described by Willingham and Yamada (I 976) except that the illumination was by bright field instead of phase contrast. Fluorescence and bright-field images were recorded on videotape by using the video intensification microscope system. Preparation of a&Colloidal Gold Complexes for Electron Microscopy Colloidal gold particles with a diameter of between 10 and 20 nm were prepared according to the method of Dickson et al. (1961). a& was adsorbed to gold particles as described by Dickson et al. (1961). except that anM was added to colloidal gold at a protein concentration that was in 10% excess of the amount required to stabilize the colloidal gold against flocculation in 5% sodium chloride. Generally, a small amount of ‘251-a2M was added as a tracer. Particles of a*Mgold were separated from free anM by centrifugation (20.000 x g. 1 hr) and resuspended in PBS containing 0.02% polyethylene glycol. In one preparation, ‘25l-a2M was conjugated to colloidal gold. Under the conditions used for electron microscopy, 90% of the “‘I-a2M: colloidal gold uptake could be blocked by 2 mg/ml of a&f.

Cell 650

Electron Microscopy Confluent cell monolayers. grown on Thermanox plastic coverslips (Lux Scientific), were incubated in DMEM at 37’C with &l-gold at an a&I concentration of 30 pgg/ml for 15 min, then rinsed and incubated for an additional 5 min at 37°C. as in experiment II (Table 1). Cells were then fixed in 2% glutaraldehyde and stained for acid phosphatase activity by using a cytidine monophosphate substrate, as described by Willingham and Yamada (I 978). Cells were posfflxed in osmium tetroxide (2%) and embedded for thin sectioning in “Dircupan” (Fluka). Thin sections were viewed at 50 kV with a JEOL 100s electron microscope. To determine the distribution of IX& colloidal gold, approximately BOO gold particles were counted and observed to be either in acid-phosphatase-negative endocytic vesicles or in acid-phosphatase-positive lysosomes. Acknowledgments We are grateful to Sharon Fluss for expert technical assistance. This work was supported by the Irma T. Hirsch1 Charitable Trust and by a grant from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Received

October

6. 1961;

revised

November

30. 1981

Ashwell. G. and Morell, A. G. (I 974). The role of surface carbohydrates in the hepatic recognition and transport of circulating glycoproteins. Adv. Enzymol. 4 1, 99-I 26.

Cheng. S.-Y., Maxfield. F. Ft., Robbins, J.. Willingham, M. C. and Pastan, I. (1960). Receptor-mediated uptake of 3.3’,5-triiodo-L-thyronine by cultured fibroblasts. Proc. Nat. Acad. Sci. USA 77, 34253429. and penetration

of

Dickson, R. B., Willingham. M. C. and Pastan, I. (1981). u2-macroglobulin adsorbed to colloidal gold: a new probe in the study of receptor-mediated endocytosis. J. Cell Biol. 89, 29-34. Draper, R. K. and Simon, M. I. (1980). The entry of diphtheria toxin into the mammalian cell cytoplasm: evidence for lysosomal involvement. J. Cell Biol. 87, 849-854. FitzGerald, D., Morris, mediated internalization Cell 2 1, 867-873.

R. E. and Saelinger, C. 6. (1960). Receptorof Pseudomonas toxin by mouse fibroblasts.

J. J., D’Arcy Hart, P. and Young, R. M. (1981). Temporal of lysosome and phagosome pH during phagolysosome in macrophages: studies by fluorescence spectroscopy. J. 89. 645-652.

Gonzalez-Noriega, A., Grubb. J. H., Talkad. V. and Sly, W. S. (1980). Chloroguine inhibits lysosomal enzyme pinocytosis and enhances lysosomal enzyme secretion by imparing receptor recycling. J. Cell Biol. 85, 839-852. Haigler. H. T.. McKanna. J. A. and Cohen, S. (1979). Direct visualization of the binding and internalization of a ferritin conjugate of epidermal growth factor in human carcinoma cells A-431. J. Cell Biol. 81, 382-395. H. T., Maxfield,

F. R., Willingham,

M. C. and

Pastan,

pH in single motile

Helenius. A., Kartenbeck, J., Simon& K. and Fries, E. (1980). On the entry of Semliki Forest virus into BHK-21 cells. J. Cell Biol. 84, 404420. Hollemans. M., Elferink. R. 0.. De Groot, P. G., Strijland, A. and Tayer, J. M. (1981). Accumulation of weak bases in relation to intralysosomal pH in cultured human fibroblasts. Biochim. Biophys. Acta 643, 140-l 51. Kaplan, J. and Nielsen, M. L. (1979). Analysis receptors. J. Biol. Chem. 254, 7323-7328.

of macrophage

surface

Keen, J. H., Maxfield. F. R.. Hardegree. M. C. and Habig. W. H. (1962). Proc. Nat. Acad. Sci. USA, in press. Receptor-mediated endocytosis of diphtheria toxin by cells in culture. Maxfield. F. R., Schlessinger. J.. Shechter. Y., Pastan. I. and Willingham, M. C. (1978). Collection of insulin, EGF and a?-macroglobulin in the same patches on the surface of cultured fibroblasts and common internalization. Cell 74, 805-810. Maxfield, F. A., Willingham, M. C.. Haigler, H. T.. Dragsten. P. and Pastan. I. H. (1981a). Binding, surface mobility, internalization and degradation of rhodamine-labeled an-macroglobulin. Biochemistry 20,5353-5358.

Ohkuma, S. and Poole, B. (I 978). Fluorescence probe measurement of the intralysosomal pH in living cells and the perturbation of pH by various agents. Proc. Nat. Acad. Sci. USA 75. 3327-3331. Pastan. I. and Willingham. M. C. (1981). Receptor-mediated endocytosis of hormones in cultured cells. Ann. Rev. Physiol. 43, 239250. Patterson, S.. Oxford, J. S. and Dourmashkin. R. R. (1979). Studies on the mechanism of influenza virus entry into cells. J. Gen. Virol. 43, 223-229. Poole, S. and Ohkuma, S. (1981). intralysosomal pH in mouse peritoneal 665-669.

Effect of weak bases on the macrophages. J. Cell Biol. 90,

Rodewald, R. (1960). Distribution of immunoglobulin receptors small intestine of the young rat. J. Cell Biol. 85, 18-32.

Goldstein, J. L.. Anderson, R. G. W. and Brown, M. S. (1979). Coated pits, coated vesicles and receptor-mediated endocytosis. Nature 2 79, 679-685.

Haigler,

Intracellular

Posner, B. I., Josefsbert. 2. and Bergeron, J. J. M. (1977). Characterization of insulin binding sites in Golgi fractions from the liver of female rats. J. Biol. Chem. 253, 4067-4073.

Frenoy, J., Bourrillon. R., Lippoldt. R. and Edelhoch, H. (1977). Stability and subunit structure of human az-macroglobulin. J. Biol. Chem. 252, 1129-l 133. Geisow. changes formation Cell Biol.

Heiple. J. M. and Taylor, D. L. (I 980). cells. J. Cell Biol., 88, 885-890.

Mosher. D. F., Saksela. 0. and Vaher. A. (1977). Synthesis and secretion of alpha P-macroglobulin by cultured adherent lung cells. J. Clin. Invest. 60. 1036-l 045.

Brown, M. S. and Goldstein, J. L. (1979). Receptor-mediated endocytisis: insights from the lipoprotein receptor system. Proc. Nat. Acad. Sci. USA 76, 3330-3337.

Attachment

Dansylcadaverine inhibits internalization of ‘261-epidermal factor in Balb 3T3 cells. J. Biol. Chem. 255, 1239-l 241.

Maxfield, F. R., Willingham, M. C.. Pastan. I.. Dragsten, P. and Cheng, S.-Y. (1981 b). Binding and mobility of the cell surface receptors for 3.3’,5-triiodo-L-thyronine. Science 27 7, 63-65.

References

Dales, S. and Choppin. P. W. (1962). influenza virus. Virology 18, 489-493.

(1960). growth

I.

in the

Salisbury. J. L., Condeelis, J. S. and Satir, P. (1980). The role of coated vesicles, microfilaments and calmodulin in receptor-mediated endocytosis by cultured B lymphoblastoid cells. J. Cell Biol. 87, 132141. Sandvig. facilitated

K. and Olsnes, S. (1980). Diphtheria at low pH. J. Cell Biol. 87, 828-832.

toxin entry into cells is

Sandvig, K. and Olsnes. S. (1981). Rapid entry of nicked toxin into cells at low pH. J. Biol. Chem. 256, 9068-9076.

diphtheria

Schlessinger, J., Shechter. Y., Willingham. M. C. and Pastan. I. (1978). Direct visualization of the binding, aggregation, and internalization of insulin and epidermal growth factor on fibroblastic cells. Proc. Nat. Acad. Sci. USA 75, 2659-2663. Segal. W. A., Geisow. M.. Garcia, R., Harper, A. and Miller, R. (1981). The respiratory burst of phagocytic cells is associated with a rise in vacuolar pH. Nature 290, 406-409. Stahl, P.. Schlesinger, P. H.. Sigardson. E.. Rodman. J. S. and Lee, Y. C. (1980). Receptor-mediated pinocytosis of mannose glycoconjugates by macrophages: characterization and evidence for receptor recycling. Cell 19, 207-215.

Acidification 651

of Endocytic

Vesicles

Thomas, J. A., Buchsbaum, Ft. N., Zimniak, A. and Racker, E. (1979). Intracellular pH measurements in Ehrlich ascites tumor cells utilizing spectroscopic probes generated in situ. Biochemistry 18, 22102216. Van Leuven. F., Cassiman. J. J. and Van den Berghe, H. (1976). Uptake and degradation of a*-macroglobulin-protease complexes in human cells in culture. Exp. Cell Res. 17 7, 273-262. Wall, D. A., Wilson, G. and Hubbard, A. (1960). The galactosespecific recognition system of mammalian liver: the route of ligand internalization in rat hepatocytes. Cell 21. 79-93. White, J.. Matkin. Forest, influenza, 674-679. Wickerhauser. macroglobulins. Willingham. destruction 480-487.

K. and Helenius. A. (1981). Cell fusion by Semllki and vesicular stomatitls viruses. J. Cell Biol. 89,

hf. and Hao, Y. L. (1972). VOX Sanguinis 23, 119-l

Large 25.

scale preparation

of

M. C. and Yamada. S. S. (1976). A mechanism for the of pinosomes in cultured fibroblasts. J. Cell. Biol. 78,

Willingham. M. C. and Pastan. I. (1980). The receptosome: an intermediate organelle of receptor-mediated endocytosis in cultured fibroblasts. Cell 27, 67-77. Willingham. M. C., Maxfield. F. R. and Pastan. I. (1979). a,-macroglobulin binding to the plasma membrane of cultured fibroblasts. J. Cell Biol. 82, 614-625. Wolf, J. L.. Rubin. D. H.. Fuiberg. R., Kauffman, R. S.. Sharpe. A. H.. Trier, J. S. and Fields, B. N. (1961). Intestinal M Cells: a pathway for entry of reovirus into the host. Science 212, 471-472. Working, functions acrosome.

P. K. and Meizel, S. (1961). Evidence that an ATPase in the maintenance of the acidic pH of the hamster sperm J. Biol. Chem. 256. 4706-4711.