tion when the microtubule-disrupting compound vinblastine was used as a stimulus. Similarly, others have found that cycloheximide pretreatment decreased ...
Journal of Cell Science 105,473-480 (1993) Printed in Great Britain © The Company of Biologists Limited 1993
473
Inhibition of protein synthesis separates autophagic sequestration from the delivery of lysosomal enzymes B. Paige Lawrence* and William J. Brown † Department of Biochemistry, Molecular and Cell Biology, Cornell University, Ithaca, NY 14853, USA *Present address: Environmental Health Sciences Center, and College of Veterinary Medicine, Oregon State University, Corvallis, OR 97339, USA †Author for correspondence
SUMMARY To investigate the role of newly synthesized proteins during autophagic sequestration and degradation, the effects of protein synthesis inhibition on autophagic vacuole (AV) formation and degradation were analyzed. The inhibition of protein synthesis was found to separate autophagic sequestration from the delivery of lysosomal enzymes to (AVs). Pretreatment with cycloheximide for ≥ 3 h caused a drastic inhibition of autophagy-induced degradation. Surprisingly, morphological analyses showed that the inhibition of protein synthesis for up to 12 h did not block the formation of nascent AVs; however, it did prevent their conversion into degradative AVs. Using immunoperoxidase cyto-
chemistry with an antibody against cathepsin D and labeling of lysosomes with endocytosed colloidal gold, we found that the nascent AVs that formed during prolonged cycloheximide pretreatment had not received lysosomal markers. The inhibition of autophagic degradation and lysosomal enzyme delivery were rapidly reversed following the removal of cycloheximide. These results suggest that there is a fairly rapid turnover of protein(s) that are necessary for lysosomal fusion, but that the initial formation of AVs is independent of new protein synthesis for a long period of time.
INTRODUCTION
whether or not autophagy requires protein synthesis is contradictory. On one hand, cycloheximide has been found to have no inhibitory effect on autophagy. For example, several groups have reported that pretreatment with cycloheximide did not inhibit the formation of glucagon-induced AVs or their conversion into a degradative compartment (Arstila and Trump, 1968; Shelburne et al. 1973; Wong and Woo, 1987). On the other hand, Amenta and Brocher (1981) reported that cycloheximide failed to inhibit autophagy stimulated by glucagon, but rapidly abrogated AV formation when the microtubule-disrupting compound vinblastine was used as a stimulus. Similarly, others have found that cycloheximide pretreatment decreased autophagic degradation or inhibited the formation of AVs in liver and pancreas of vinblastine-treated rats (Marzella and Glaumann, 1980a; Oliva et al., 1992). Because of the different experimental methods employed, it is very difficult to draw a final conclusion about the necessity of protein synthesis for autophagy. Two critical variables among these studies are the methods used to induce autophagy, and the timing of exposure to cycloheximide, including the length of treatment prior to the stimulation of autophagy. Studies using vinblastine to stimulate autophagy are particularly difficult to interpret due to its effect on microtubules. It has been proposed by some investigators that microtubule disruption inhibits the ability of
Classical or macroautophagy is a normal degradative pathway that exists in all eukaryotic cells (for reviews see Holtman, 1989; Seglen and Bohley, 1992). This degradative mechanism involves the sequestration of intracellular material, including organelles, by membranes of the ER (Marzella and Glaumann, 1987; Dunn, 1990a; Furuno et al., 1990; Ueno et al., 1991), to form a unique nascent autphagic vacuole (AV). The formation of nascent AVs is followed very rapidly by the delivery of lysosomal enzymes from pre-existing lysosomes to form degradative AVs (Ericsson, 1969; Lee et al., 1989; Lawrence and Brown, 1992). This autophagic process is the primary mechanism responsible for the turnover of unnecessary or dysfunctional organelles as well as for a random degradation of cytoplasmic proteins (Ahlberg et al., 1985; Ballard, 1987; Holtzman, 1989; Kopitz et al., 1990). Autophagy also plays a key role in embryonic programmed cell death, such as occurs during normal animal development (Holtzman, 1989; Clarke, 1990). Autophagy can be stimulated above this basal level by a variety of environmental stresses that enhance the need for cells to utilize their normal autophagic mechanism to survive. The complexity of this multi-step process suggests that it must be tightly regulated. The literature regarding
Key words: autophagy, organelle fusion, cycloheximide
474
B. P. Lawrence and W. J. Brown
lysosomes to fuse with nascent AVs (Kovacs et al., 1982; Gordon and Seglen, 1988). Consequently, the increase in AVs may be due to the accumulation of new AVs that cannot be converted into degradative AVs. Consistent with this model is the observation that vinblastine-treated livers also have a lower level of proteolysis, suggesting that lysosomal degradation is somehow inhibited by vinblastine (Kovacs et al., 1982; Gordon and Seglen, 1988). However, other groups have reported contrary evidence, indicating that vinblastine treatment neither blocks lysosome fusion with nascent AVs nor inhibits lysosomal proteolysis (Arstila et al., 1974; Marzella and Glaumann, 1980a,b; Punnonen and Reunanen, 1990; Rez et al., 1990). Thus, it is not clear whether vinblastine actually increases the basal level of autophagy, or whether it prevents the delivery of lysosomal enzymes, leading to an accumulation of nascent AVs. Confusion about the actual effects of vinblastine on autophagic sequestration and degradation have direct bearing on understanding the contradictory nature of the effects of protein synthesis inhibition on autophagy. Many experiments to determine the effect of protein synthesis inhibition on autophagy have involved measuring proteolytic levels using vinblastine as the stimulus (Marzella and Glaumann, 1980a,b; Amenta and Brocher, 1981; Kovacs, 1982; Punnonen and Reunanen, 1990). Since it is unclear whether one secondary effect of vinblastine treatment may be an inhibition of lysosomal enzyme delivery by affecting microtubule function, it is difficult to interpret the data on the effects of cycloheximide on autophagy. Using a cultured cell system, we have shown that amino acid starvation alone or in combination with glucagon administration induces a 5-fold increase in autophagic sequestration and degradation, as measured both morphometrically and biochemically (Lawrence and Brown, 1992). Here we use this system to resolve whether or not newly synthesized proteins are required for autophagy by analyzing both autophagic sequestration and degradation as a function of cycloheximide pretreatment. MATERIALS AND METHODS Materials L-[2,3,4,5-3H]leucine at approx. 110 Ci/mmol and Ecolume scintillation counting solution were purchased from ICN Radiochemicals, Inc. (Costa Mesa, CA). Fab fragments of sheep anti-rabbit IgG conjugated with horseradish peroxidase (HRP) were from BioSys S.A. (Compiegne, France). Tetrachloroauric (III) acid, diaminobenzidine hydrochloride (type II), saponin, glucagon and cell culture media were from Sigma Chemical Co. (St. Louis, MO). Leupeptin (hydrogen sulfate) was from Boehringer Mannheim (Indianapolis, IN). Fetal bovine serum and MEM vitamin supplement were from Gibco Laboratories (Grand Island, NY). Spurr’s resin and other reagents for electron microscopy were obtained from Electron Microscopy Sciences (Fort Washington, PA).
Cell culture Fu5C8 cells were grown in minimal essential medium (MEM) supplemented with 5% fetal bovine serum (FBS), at 37°C in an atmosphere of 95% air and 5%CO2. All incubations were con-
ducted under these conditions. To induce autophagy by amino acid starvation, the cells were washed three times with Earle’s balanced salts solution with vitamins (EBSS), and then incubated in EBSS containing 86 nM glucagon for various lengths of time.
Electron microscopy Following various experimental conditions, cultured hepatocytes were fixed with 1% paraformaldehyde, 3% glutaraldehyde, 0.1 M sodium cacodylate, pH 7.4, 2 mM CaCl2, for 1.5 h at room temperature. After washing with 0.1 M sodium cacodylate, pH 7.4, the cells were post-fixed with 1% OsO4 in 0.1 M sodium cacodylate, pH 7.4, for 2 h, at 4°C. Osmicated samples were rinsed with 0.1 M sodium maleate buffer, pH 5.15, and then en bloc stained with 1% uranyl acetate, in 0.1 M sodium maleate buffer, pH 6.0, for 1 h at 25°C. After rinsing with 0.1 M sodium maleate buffer, pH 5.15, the samples were dehydrated in graded ethanol solutions, removed from plastic culture dishes with propylene oxide, and embeded in Spurr’s resin. The samples were thin sectioned, stained with lead citrate and uranyl acetate, and examined using a Philips 301 electron microscope. Immunoperoxidase cytochemistry was carried out as described by Lawrence and Brown (1992) using a polyclonal antibody against the lysosomal enzyme cathepsin D. Colloidal gold (8 nm) was prepared by the method outlined by Handley (1989), and stabilized with both BSA in 0.005 M NaCl, pH 7.4 (with K2CO3), and 0.2% PEG (8000).
Morphological definitions For quantitation of morphological observations, strict morphological criteria were used to define each compartment (Lawrence and Brown, 1992). Specifically, nascent AVs were defined as organelles bounded by more than one limiting membrane and which contained parts of the cell. Those organelles that contained recognizable parts of the cell but were bounded by only one membrane were counted as degradative AVs. The lysosomal compartment included all other single membrane-bounded degradative organelles.
Biochemical assay for autophagic protein degradation Increased protein degradation induced by stimulating autophagy was measured as described by Lawrence and Brown (1992). For this assay, the cells were labeled to isotopic steady-state (≥16 h) with 1.0 µCi/ml L-[2,3,4,5-3H]leucine in MEM, containing 5% FBS prior to incubation in either MEM containing 5% FBS and a 10-fold excess of unlabeled leucine (10× Leu), or in EBSS containing 10× Leu and 86 nM glucagon. The culture medium was collected and clarified by centrifugation, to remove any dead cells or debris. An equal volume of ice-cold 10% TCA was added to half of the medium. After a ≥1 h incubation at 4°C, the samples were microcentrifuged for 10 min to pellet all of the TCA-insoluble material. Following removal of the supernatant, the pellets were dissolved in 40 µl of 1 M NaOH. Both the TCA-soluble and TCA-insoluble radioactivity were determined using a scintillation counter. These data were expressed as the average TCA-soluble/TCA-insoluble radioactivity for each time point.
RESULTS Pretreatment with cycloheximide inhibits autophagy-stimulated degradation To investigate the requirement for new protein synthesis during autophagy, cells were pretreated for various periods of time with the protein synthesis inhibitor cycloheximide
Cycloheximide inhibits autophagic degradation
Fig. 1. Pretreatment with cycloheximide inhibits autophagic degradation. (A) [3H]leucine-labeled cells were pretreated for various periods of time (up to 12 h) with 2 µg/ml cycloheximide prior to the induction of autophagy. After these pretreatments, autophagy was induced by incubating cells in EBSS containing glucagon for 1 h (in the continuous presence cycloheximide). Degradation was monitored by the release of TCA-soluble radioactivity into the media. For the washout (arrow), autophagy was induced in the presence of cycloheximide for 1 h (following a 12 h pretreatment), then the cells were incubated for an additional 30 or 60 min in EBSS with glucagon, but without cycloheximide. Identical control (fed) samples were otherwise treated the same. The values shown are averages of triplicate samples. Amino acidstarved and glucagon-treated cells (open circles); Fed cells (filled squares). (B) [3H]leucine-labeled cells were treated for 6 h with or without 2 µg/ml cycloheximide prior to the induction of autophagy. After this pretreatment, cells were washed and autophagy was induced as usual (in either the absence or the continuous presence cycloheximide to match the pretreatment condition) for various periods of time as shown on the x-axis. 0 h represents the time at which autophagy was stimulated. Degradation was monitored by the release of TCA-soluble radioactivity into the medium. The values shown are averages of triplicate samples, and the error bars represent 1 s.d. Amino acidstarved cells (filled circles); amino acid-starved cells with cycloheximide treatment (open circles); Fed cells (filled squares); Fed cells with cycloheximide treatment (open squares).
(2 µg/ml) prior to induction of autophagy. Autophagy was stimulated by amino acid starvation and glucagon treatment (while still in the presence of cycloheximide) for 1 h, after which the medium was assayed for the release of degraded proteins. We found that the amount of autophagy-stimulated proteolysis decreased as a function of pretreatment time with cycloheximide (Fig. 1A). Increasingly longer incubations with cycloheximide led to a fairly rapid
475
decrease in the level of autophagic protein degradation. After 3 h of cycloheximide pretreatment, autophagic proteolysis was completely inhibited; however, cycloheximide pretreatment did not appear to affect the level of protein degradation in fed cells. This inhibition of autophagy-stimulated protein degradation was rapidly reversed even after long-term incubations in cycloheximide. After cells were treated with cycloheximide for 12 h, autophagy was induced for 1 h in the continuous presence of cycloheximide, then the cycloheximide was removed and the cells were incubated in EBSS containing glucagon for another hour. This treatment, indicated by the arrow in Fig. 1A, demonstrated that upon the reinitiation of protein synthesis, there was a very rapid increase in proteolysis in amino acid-starved, glucagontreated cells. After the 1 h washout period, no significant change was observed in the proteolytic level among the fed cells, which were otherwise treated in the same way. To investigate whether the inhibition of protein degradation was due to a blockage or simply a slowing of the entire autophagic process, cells were incubated with or without cycloheximide for 6 h, followed by stimulation of autophagy (± cycloheximide) for various periods of time (Fig. 1B). Untreated cells exhibited a typical large increase in autophagy-stimulated degradation; however, cycloheximide-treated cells had only a modest increase that reached a plateau after 2 h. Significantly, degradation in cycloheximde-treated cells did not continue to increase with time (ultimately reaching control levels), indicating that cycloheximide treatment blocked degradation and did not merely slow the autophagic process down. Cycloheximide pretreatment inhibits the conversion of nascent AVs into degradative AVs Cycloheximide could be inhibiting autophagy either by preventing the initial sequestration by ER membranes to form nascent AVs or by blocking a subsequent degradative event. To determine whether the inhibition of protein synthesis blocked the initial formation of AVs, the effects of cycloheximide treatment and removal were investigated at the morphological level. In the absence of cycloheximide pretreatment, both nascent and degradative AVs were observed throughout the cytoplasm after stimulating autophagy for 1 h in the presence of cycloheximide (Fig. 2A). Following cycloheximide pretreatment of ≥3 h, many nascent AVs were found; however, degradative AVs were conspicuously absent (Fig. 2B). As expected from the degradation assay, upon removal of cycloheximide numerous degradative AVs were now observed throughout the cytoplasm (Fig. 2C). These observations were quantified by counting the numbers of nascent and degradative AVs observed after 1 h of amino acid starvation and glucagon treatment following increased lengths of pretreatment with cycloheximide. These values were compared with the relative numbers of nascent and degradative AVs when autophagy was induced in the absence of protein synthesis inhibition (Fig. 3). When autophagy was stimulated without cycloheximide, approx. 50% of all AVs were nascent, and the other half were degradative. Simultaneous induction of autophagy and treatment with cycloheximide for 1 h did not effect the relative numbers of nascent or degradative AVs that formed.
476
B. P. Lawrence and W. J. Brown
Fig. 3. Protein synthesis inhibition blocks the conversion of nascent AVs into degradative AVs. The number of nascent and degradative AVs present after various periods of cycloheximide treatment and removal were quantified. After treatment with cycloheximide, autophagy was induced for 1 h by amino acid starvation and glucagon, in the presence of cycloheximide. The data are expressed as the percentage of all AVs (nascent + degradative) that were nascent ones (AVi). The washout experiment was conducted as described for Fig. 1 (1 h). The open circle represents the % AVis after inducing autophagy for 1 h in the absence of cycloheximide. Approximately 100 AVs from random fields were counted for each time point. ×28,000.
However, increasingly longer exposures to cycloheximide prior to the induction of autophagy (for 1 h) resulted in up to 85% of all of the AVs being nascent ones, indicating that the cells were able to sequester material for autophagy but were prevented from degrading it. This inhibition was rapidly reversed upon removal of cycloheximide, as demonstrated by the dramatic increase in degradative AVs (approx. 75% of all AVs being degradative ones) after removal of cycloheximide and subsequent incubation in EBSS, containing glucagon for just 1 h. These results suggest that nascent AVs accumulated due to the inhibition of protein synthesis, and were rapidly converted into degradative AVs upon the re-initiation of protein synthesis.
Fig. 2. Cycloheximide pretreatment results in the reversible accumulation of nascent AVs. (A) No pretreatment. Following the induction of autophagy by amino acid starvation and glucagon treatment for 1 h in the presence of 2 µg/ml cycloheximide, numerous nascent (AVi) and degradative (AVd) AVs were observed throughout the cytoplasm. (B) 12 h pretreatment with cycloheximide. When autophagy was induced following >3 h pretreatment in cycloheximide, numerous AVis were found; however, AVds were rarely observed. (C) Washout. Following a 12 h pretreatment with cycloheximide, autophagy was induced in the presence of cycloheximide for 1 h, then the cells were incubated for an additional 60 min in EBSS with glucagon but without cycloheximide. Following the removal of cycloheximide, abundant AVds, at various stages of degradation, were observed. Bars: 0.50 µm (A, B); 1.0 µm (C).
Cycloheximide pretreatment inhibits the delivery of lysosomal markers to nascent AVs One possible explanation for the inhibitory effect of cycloheximide on the conversion of nascent AVs into degradative AVs is that a regulatory protein with a short half-life is involved either in the fusion of lysosomes with nascent AVs, or in the ability of lysosomal enzymes to carry out hydrolysis (e.g. acidification mechanism). To determine whether or not the nascent AVs that accumulated during cycloheximide pretreatment had received lysosomal enzymes, the above experiment was repeated and the intracellular distribution of the lysosomal enzyme cathepsin D (catD) was analyzed by immunoperoxidase cytochemistry. In the absence of cycloheximide treatment, antibodies against cathepsin D (catD) stained the rough ER, lysosomes and degradative AVs; however, nascent AVs rarely contained lysosomal enzymes (Dunn, 1990b; Lawrence and Brown, 1992). Occasionally, catD staining of nascent AV membranes was observed, presumably because the ER that formed it contained newly synthesized enzymes. Both lysosomes and degradative AVs still contained catD following
Cycloheximide inhibits autophagic degradation incubation with cycloheximide concomitantly with the induction of autophagy for 1 h. Pretreatment with cycloheximide for 3 h resulted in fewer degradative AVs, some of which still contained catD labeling (data not shown). Interestingly, after 12 h in cycloheximide, the accumulated nascent AVs did not contain any catD reaction product, whereas lysosomes stained positively for catD (Fig. 4A). Upon removal of cycloheximide, the degradative AVs that rapidly formed within 1 h were rich in catD (Fig. 4B). These results were quantified and compared with the distribution of catD in amino acid-starved glucagon-treated cells, and clearly show that the percentage of catD-positive AVs was greatly reduced (to approx. 10%) with increasing time of cycloheximide pretreatment (Fig. 5). In fact, after 12 h of pretreatment, no nascent AVs were found to be catD positive. The overall numbers of degradative AVs were greatly reduced after long-term protein synthesis inhibition and, of these, very few contained immunocytochemically detectable enzyme. Although very few catD-positive AVs were observed under these conditions, numerous catDpositive lysosomes were observed, ruling out a significant loss of enzyme due to turnover. In contrast, washout experiments showed that by 1 h after the removal of cycloheximide, 79% of the degradative AVs stained positively for catD (60% of the total AVs). To establish further that cycloheximide pretreatemnt was inhibiting the delivery of enzymes from pre-existing lysosomes, cells were allowed to endocytose 8 nm colloidal gold particles (Au8), prior to the induction of autophagy, under conditions that result in >95% of the gold being found in lysosomes (Lawrence and Brown, 1992). Following a 3 h pretreatment with cycloheximide to inhibit degradation and enzyme delivery, autophagy was induced for 1 h in the presence of cycloheximide. After this treatment, the cycloheximide was removed and the cells were incubated in EBSS containing glucagon for an additional 15, 30 and 60 min, before fixation and processing for electron microscopy. As expected, prior to the removal of cycloheximide, numerous lysosomes were found to contain Au8 particles but nascent AVs did not, demonstrating that little, if any, lysosomal Au8 was delivered to the AVs that formed during cycloheximide pretreatment. More importantly, following the removal of cycloheximide Au8 particles were rapidly delivered to AVs (Fig. 4C, D). Numerous degradative AVs that contained Au8 were observed following 1 h of cycloheximide washout. Actually, although less prevalent than after 1 h, degradative AVs were observed within 15 min after the cycloheximide was washed out, and some contained Au8. This observation further supports the role of pre-existing lysosomes as the primary source of hydrolytic enzymes during autophagy, because after 15 min it is unlikely that enzymes would be coming from the biosynthetic pathway. These and the previous experiments clearly indicate that blocking protein synthesis for ≥3 h results in the inability of pre-existing lysosomes to fuse with nascent AVs. DISCUSSION We have demonstrated that AVs form in the absence of
477
continued protein synthesis. The biochemical data alone would suggest that inhibiting protein synthesis for > 3 h completely inhibits autophagy. Surprisingly, however, the morphological and immunocytochemical data presented here demonstrate that cycloheximide pretreatment blocks a later stage in AV processing, but not their initial formation. Through the use of lysosomal markers, we found that pretreatment with cycloheximide for > 3 h prevented the delivery of lysosomal contents to nascent AVs. Thus, pretreatment with cycloheximide for increasing lengths of time effectively uncouples the process of AV formation from their subsquent fusion with lysosomes. The rapid appearance of two lysosomal markers (catD and Au8) following cycloheximide removal indicates that AVs are able to fuse with lysosomes shortly after the re-initiation of protein synthesis. It seems quite remarkable that cells are able to form nascent AVs even after 12 h of cycloheximide pretreatment. This finding indicates that the proteins required for the initial sequestration of cytoplasm by the ER are quite longlived. It is also formally possible that cycloheximide pretreatment inhibits the closure of ER membranes as they engulf cytoplasm. Although we cannot definitively rule out this possibility, our numerous observations suggest that nascent AVs are usually fully enclosed organelles. Presumably, if pretreatment with cycloheximide is long enough, then even the initial sequestration (i.e. formation of nascent AVs) would be inhibited. However, we have not been able to determine this interval; and in any case, interpretation of the results would become increasingly difficult after such a long time in cycloheximide. In contrast to nascent AV formation, autophagy-stimulated protein degradation and delivery of lysosomal enzymes to AVs was completely inhibited by 3 h of cycloheximide pretreatment. These results suggest that at least one protein that is required for lysosomal enzyme delivery has a relatively short half-life (approx. 1.5 h). This suggestion is supported by finding that autophagic protein degradation could return to approx. 50% of control levels by 1 h after recovery from cycloheximide treatment. We have also found that after treatment with cycloheximide for > 3 h, only approx. 10% of the total AVs (nascent and degradative) contained the lysosomal enzyme catD. However, after 1 h of recovery, approx. 60% of the total AVs stained positively for catD. In addition, colloidal gold that was previously loaded into lysosomes was rapidly (within 2 h to be fully processed and delivered to lysosomes (Hasilik and Neufeld, 1980a,b; Park et al., 1991). Therefore, it is highly unlikely that enough active enzyme could reach autophagic vacuoles in order to achieve a 50% recovery in autophagystimulated protein degradation by 1 h after the removal of cycloheximide. We were prompted to conduct these studies because apparently no definitive conclusion had been reached about
478
B. P. Lawrence and W. J. Brown
Fig. 4. Localization of two lysosomal markers, cathepsin D (catD) and Au8 after cycloheximide pretreatment and removal. (A, B) Immunoperoxidase Fu5C8 cells were treated with cycloheximide for 12 h prior to the stimulation of autophagy for 1 h with EBSS, containing glucagon. Cells were fixed and processed for immunoperoxidase cytochemistry. (A) 12 h pretreatment, 1 h stimulation of autophagy. Numerous nascent AVs (AVi) were observed; however, they were not labeled by the antibody against catD. Many lysosomes (Lys) with reaction product were found throughout the cytoplasm. (B) Reversal. Following a 12 h pretreatment with cycloheximide, and stimulation of autophagy for 1 h (in the presence of cycloheximide), cycloheximide was washed out and the cells were incubated in EBSS, containing glucagon for an additional hour. Numerous catD-positive degradative AVs (AVd) were observed. catD-positive AVis were rarely observed, and when they were the staining was limited to the enveloping membranes. (C, D) Colloidal gold-labeled lysosomes. Lysosomes were labeled by a 16 h pulse and a 2 h chase with 8 nm colloidal gold prior to a 3 h incubation in MEM, containing 5% FBS and 2 µg/ml cycloheximide. (C) Autophagy was induced for 1 h in the continuous presence of cycloheximide. Nascent AVs were observed, but few, if any, contained Au8. Numerous gold-labeled lysosomes (Lys) were found throughout the cytoplasm. (D) Reversal. Cycloheximide was washed out and autophagy was stimulated for another hour as in B. During this time degradative AVs (AVd) appeared throughout the cytoplasm, and many now contained Au8. Bars: 0.5 µm (A, B); 1.0 µm (C, D).
Cycloheximide inhibits autophagic degradation
479
ever, cycloheximide pretreatment effectively blocked the processing of nascent AVs into degradative ones by inhibiting the delivery of enzymes from pre-existing lysosomes. By studying the effects of cycloheximide on autophagy in this manner, autophagic sequestration and degradation were separated into two distinct processes, presumably with very different metabolic regulators. This ability to prevent AV maturation will, hopefully, provide a valuable tool for investigating the unique characteristics of nascent AV membranes and for probing the molecular details of lysosomal enzyme delivery and the conversion to a degradative organelle. Fig. 5. Cycloheximide pretreatment inhibits the delivery of lysosomal enzymes from pre-existing lysosomes to nascent AVs. The distribution of catD in AVs after various lengths of cycloheximide treatment and removal were quantified. The data are expressed as the percentage of all AVs (both nascent and degradative) that were labeled by the antibody against catD. Autophagy was induced for 1 h by amino acid starvation and glucagon, in the absence (0 h bar) or presence (1 h bar) of cycloheximide. Cells were pretreated with cycloheximide for 12 h prior to the induction of autophagy (12 h bar). The washout experiment was conducted as described for Fig. 1 (reversal). Approximately 60 AVs from random fields were counted for each time point. ×28,000.
the effect of cycloheximide on autophagy. On one hand, several studies have found that short-term exposures (< 90 min) to cycloheximide prior to stimulating autophagy did not result in any inhibition (Arstila and Trump, 1968; Shelburne et al., 1973; Amenta and Brocher, 1981; Wong and Woo, 1987). Other investigators, most of whom used vinblastine to stimulate autophagy, found just the opposite result: 30-90 min incubations with cycloheximide inhibited autophagy (Marzella and Glaumann, 1980a; Amenta and Brocher, 1981; Kovacs et al., 1982; Oliva et al., 1992). To confuse the issue further, Glaumann and colleagues (Glaumann et al., 1981; Marzella and Glaumann, 1980c) suggested that vinblastine may also inhibit non-lysosomal proteolysis. Therefore, it would be very difficult to distinguish between the inhibitory effects produced by cycloheximide and by vinblastine. This may also explain some of the contradictory results between experiments using vinblastine versus glucagon as an autophagic stimulus. Our studies do not address the question of what factors might have turned over in the 3 h of cycloheximide pretreatment, leading to the inhibition of AV-lysosome fusion. However, obvious candidates would be any one of the several components of the NSF-fusion particle that appears to be required for fusion between various membranous organelles (for review see Wilson et al., 1992), although the requirement for an N-ethylmaleimide-sensitive protein has not yet been demonstrated for autophagic degradation. In any case, since we have been able to reversibly separate the process of nascent AV formation from fusion with lysosomes (and consequent conversion into a degradative AV), it should be possible to exploit these findings in order to identify factors that are specifically involved with each step. In this report, we have demonstrated that pretreatment of cells for up to 12 h with the protein synthesis inhibitor cycloheximide does not block the formation of AVs. How-
We thank Marian Strang for producing the thin sections for EM and the members of the lab. for comments on this manuscript. This work was supported by a grant from the National Institutes of Health (DK 37249) to W. Brown.
REFERENCES Ahlberg, J, Berkenstam, Henell, F. and Glaumann, H. (1985). Degradation of short and long lived proteins in isolated rat liver lysosomes: effects of pH, temperature and proteolytic inhibitors. J. Biol. Chem. 260, 5847-5854. Amenta, J. S. and Brocher, S. C. (1981). Mechanisms of protein turnover in cultured cells. Life Sci. 28, 1195-1208. Arstila, A. U., Nuuja, I. J. M. and Trump, B. F. (1974). Studies of cellular autophagocytosis: vinblastine-induced autophagy in the rat liver. Exp. Cell Res. 87, 249-252. Arstila, A. U. and Trump, B. F. (1968). Studies on autophagocytosis: The formation of autophagic vacuoles in the liver following glucagon administration. Am. J. Pathol. 53, 687-693. Ballard, F. J. (1987). Regulation of intracellular protein breakdown with special reference to cultured cells. In Lysosomes: Their Role in Protein Breakdown (ed. H. Glaumann and F.J. Ballard), pp. 285-318. Academic Press, Inc., New York, London. Clarke, P. G. H. (1990). Developmental cell death: morphological diversity and multiple mechanisms. Anat. Embryol. 181, 195-213. Dunn, W. A. Jr (1990a). Studies on the mechanisms of autophagy: formation of the autophagic vacuole. J. Cell Biol. 110, 1923-1933. Dunn, W. A. Jr (1990b). Studies on the mechanisms of autophagy: maturation of the autophagic vacuole. J. Cell Biol. 110, 1935-1945. Ericsson, J. L. E. (1969). Studies on the mechanisms of autophagy: maturation of the autophagic vacuole. Exp. Cell Res. 55, 95-106. Furuno, K., Ishikawa, T., Akasaki, K., Lee, S., Nishimura, Y., Tsuji, H., Himeno, M. and Kato, K. (1990). Immunocytochemical study of the surrounding envelope of autophagic vacuoles in cultured rat hepatocytes. Exp. Cell Res. 189, 261-268. Glaumann, H., Ericsson, J. L. E. and Marzella, L. (1981). Mechanisms of intralysosomal degradation with special reference to autophagocytosis and heterophagocytosis of cell organelles. Int. Rev. Cytol. 73, 149-182. Gordon, P. B. and Seglen, P. O. (1988). Prelysosomal convergence of autophagic and endocytic pathways. Biochem. Biophys. Res. Commun. 151, 40-47. Handley, D. A. (1989). In Colloidal Gold: Principles, Methods, and Applications (ed. M. A. Hayat), pp. 13-32. Academic Press, Inc., New York, London. Hasilik, A. and Neufeld, E. F. (1980a). Biosynthesis of lysosomal enzymes in fibroblasts: synthesis as precursors of higher molecular weight. J. Biol. Chem. 255, 4937-4945. Hasilik, A. and Neufeld, E. F. (1980b). Biosynthesis of lysosomal enzymes in fibroblasts: phosphorylation of mannose residues. J. Biol. Chem. 255, 4946-4950. Holtzman, E. (1989). Lysosomes, pp. 439. Plenum Publishing Corp., New York. Kopitz, J., Kisen, G. Ø., Gordon, P. B., Bohley, P. and Seglen, P. O. (1990). Nonselective autophagy of cytosolic enzymes in isolated rat hepatocytes. J. Cell Biol. 111, 941-953. Kovacs, A. L., Reith, A. and Seglen, P. O. (1982). Accumulation of
480
B. P. Lawrence and W. J. Brown
autophagosomes after inhibition of hepatocytic protein degradation by vinblastine, leupeptin, or a lysosomotropic amine. Exp. Cell Res. 137, 191-201. Lawrence, B. P. and Brown, W. J. (1992). Autophagic vacuoles rapidly fuse with pre-existing lysosomes in cultured hepatocytes. J. Cell Sci. 102, 515-526. Lee, H.-K., Myers, R. A. and Marzella, L. (1989). Stimulation of autophagic protein degradation by nutrient deprivation in a differentiated murine teratocarcinoma (F9 12-1a) cell line. Exp. Mol. Pathol. 50, 139146. Marzella, L. and Glaumann, H. (1980a). Increased degradation in rat liver induced by vinblastine. I. Biochemical characterization. Lab. Invest. 42, 8-17. Marzella, L. and Glaumann, H. (1980b). Increased degradation in rat liver induced by vinblastine. II. Morphologic characterization. Lab. Invest. 42, 18-27. Marzella, L. and Glaumann, H. (1980c). Inhibitory effects of vinblastine on protein degradation. Virchows Arch. B Cell Pathol. 34, 111-122. Marzella, L. and Glaumann, H.(1987). Autophagy, microautophagy, and crinophagy as mechanisms for protein degradation. In Lysosomes: Their Role in Protein Breakdown (ed. H. Glaumann and F.J. Ballard), pp. 319367. Academic Press, Inc., New York, London. Oliva, O., Rez, G., Palfia, A. and Fellinger, E. (1992). Dynamics of vinblastine-induced autophagocytosis in murine pancreatic acinar cells: influence of cycloheximide post-treatments. Exp. Mol. Pathol. 56, 76-86. Park, J. E., Draper, R. K. and Brown, W. J. (1991). Biosynthesis of lysosomal enzymes in cells of the End 3 complementation group conditionally defective in endosomal acidification. Somat. Cell Mol. Genet. 17, 137-150.
Plomp, P. J. A. M., Gordon, P. B., Meijer, A. J., Høyvik, H. and Seglen, P. O. (1989). Energy dependence of different steps in the autophagiclysosomal pathway. J. Biol. Chem. 264, 6699-6704. Punnonen, E.-L. and Reunanen, H. (1990). Effects of vinblastine, leucine, histidine, and 3-methyladenine on autophagy in Ehrlich ascites cells. Exp. Mol. Pathol. 52, 87-97. Rez, G., Laszlo, L., Fellinger, E., Kovacs, A. L. and Kovacs, J. (1990). Time course of the quantitative changes in the autophagic-lysosomal and secretory granule compartments of murine liver cells under the influence of vinblastine. Virchows Arch. B Cell Pathol. 58, 189-197. Seglen, P. O. and Bohley, P. (1992). Autophagy and other vacuolar protein degradation mechanisms. Experientia 48, 158-172. Shelburne, J. D., Arstila, A. U. and Trump, B. F. (1973). Autophagocytosis: the relationship of autophagocytosis to protein synthesis and to energy metabolism in rat liver and flounder kidney tubules in vitro. Am. J. Pathol. 73, 641-662. Ueno, T., Muno, D. and Kominami, E. (1991). Membrane markers of endoplasmic reticulum preserved in autophagic vacuole membranes isolated from leupeptin-administered rat liver. J. Biol. Chem. 266, 1899518999. Wilson, D. W., Whiteheart, S. W., Wiedmann, M., Brunner, M. and Rothman, J. E. (1992). A multisubunit particle implicated in membrane fusion. J. Cell Biol. 117, 531-538. Wong, S. S. C. and Woo, P. T. L. (1987). Stimulation by glucagon and adenosine-3′,5′-cyclic monophosphate of protein degradation in Reuber H35 hepatoma monolayers. J. Biochem. 19, 443-447. (Received 24 November 1992 - Accepted, in revised form, 22 March 1993)
