Journal of Cell Science 113, 4441-4450 (2000) 4441 Printed in Great Britain © The Company of Biologists Limited 2000 JCS1804

Unique properties of lamp2a compared to other isoforms

A. M. Cuervo* and J. F. Dice Department of Physiology, Tufts University School of Medicine, Boston, MA, USA *Author for correspondence (e-mail: [email protected])

Accepted 28 September; published on WWW 16 November 2000

SUMMARY

Lamp2a acts as a receptor in the lysosomal membrane for other lamp2s. Four positively-charged amino acids substrate of chaperone-mediated autophagy. uniquely present in the cytosolic tail of lamp2a are required Using antibodies specific for the cytosolic tail of lamp2a and for the binding of substrate proteins. Lamp2a also others recognizing all lamp2 isoforms, we found that in rat distributes to an unique subpopulation of perinuclear liver lamp2a represents 25% of lamp2s in the . We in cultured fibroblasts in response to serum show that lamp2a levels in the lysosomal membrane in withdrawal, and lamp2a, more than other lamp2s, tends to rat liver and fibroblasts in culture directly correlate with multimerize. These characteristics may be important for rates of chaperone-mediated autophagy in a variety lamp2a to act as a receptor for chaperone-mediated of physiological and pathological conditions. The autophagy. concentration of other lamp2s in the lysosomal membrane show no correlation under the same conditions. Key words: Lysosome, degradation, Membrane receptor, Furthermore, substrate proteins bind to lamp2a but not to Chaperone, Rat liver

INTRODUCTION Part of this matrix form of lamp2 reversibly aggregates with other lysosomal enzymes in a pH-dependent manner (Jadot et Lysosome-associated membrane proteins (lamps) are a al., 1997). The origin of the lumenal form of lamp2a is still group of lysosomal proteins with very similar structural unclear, but a direct deinsertion from the lysosomal membrane characteristics but of mostly unknown function (Fukuda, 1991; after a conformational change, as well as a release by Peters and von Figura, 1994). All of the lamps are type I proteolytic cleavage of the short transmembrane and cytosolic integral glycosylated proteins with a large lumenal domain, a tail have been proposed (Jadot et al., 1996). That kind of single transmembrane region of about 20 amino acids, and a deinsertion has been described for other type I membrane short (10-12 amino acid) carboxyl terminus tail at the cytosolic proteins (Nishiyama et al., 1999), and cleavage has been side of the lysosomal membrane (Akasaki and Tsuji, 1998). described for other lysosomal membrane proteins such as acid Lamps are mainly localized to lysosomes but can also be phosphatase (Gottschalk et al., 1989) and (Meikle et al., detected in lower amounts in endosomes and at the plasma 1999). The high level of of lamp2, in which the membrane (Furuno et al., 1989; Akasaki et al., 1993). Two protein core accounts for only 40 kDa of the final glycosylated different classes of lamps, lamp1 and lamp2, encoded by two product of 96 kDa, and the small size of the transmembrane different but evolutionarily-related have been described. and cytosolic tail, make it difficult to identify the matrix The lamp2 undergoes alternative splicing resulting in at forms of lamp2 as intact or truncated using conventional least three different mRNAs encoding different isoforms of electrophoretic methods. lamp2 (Gough et al., 1995; Hatem et al., 1995; Konecki et al., The function of most lamps remains unclear. They have been 1995). The three lamp2 isoforms (a, b and c) identified so far hypothesized to play a role in protecting the lysosomal show high amino acid sequence identity in their lumenal membrane from its associated hydrolases (Fukuda, 1991). region, but different transmembrane and cytosolic regions However, it has been shown recently that the complete (Gough et al., 1995). The lamp2 splice variants are expressed elimination of lamp1 that constitutes almost 40% of the at different levels in different tissues (Konecki et al., 1995; lysosomal membrane protein does not modify lysosomal Furuta et al., 1999) and have different distributions between the stability (Andrejewski et al., 1999). A role for lamp2 in cell- plasma membrane and lysosomes (Gough and Fambrough, cell or cell- adhesion has been proposed 1997). Lamp2 is concentrated in tissues undergoing apoptosis (Lippincott-Schwartz and Fambrough, 1986; Carlsson et al., during development, and the expression pattern for each lamp2 1988; Saitoh et al., 1992; Licheter-Konecki et al., 1999). A role isoform becomes more tissue and cell-type specific as for lamp2 in maturation of autophagic vacuoles has also differentiation progresses (Licheter-Konecki et al., 1999). been proposed (Tanaka et al., 2000). By analogy with other Lamp2s are present in the lysosomal lumen as well as the alternatively spliced proteins (Ravetch and Perussia, 1989), the lysosomal membrane (Jadot et al., 1996; Jadot et al., 1997). tissue-dependent expression of the different forms of lamp2 4442 A. M. Cuervo and J. F. Dice

(Konecki et al., 1995) suggests that they might have different of serum, plates were extensively washed with Hanks’ balanced salts cellular functions. solution (Life Technologies, Gaithersburg, MD) and medium without We identified the lamp type 2a (lamp2a) at the lysosomal serum was added. membrane as a receptor for substrates of chaperone-mediated Chemicals autophagy (Cuervo and Dice, 1996). In this pathway, specific cytosolic proteins are directly transported through the Sources of chemicals and antibodies were as described previously (Terlecky and Dice, 1993; Cuervo et al., 1994; Cuervo et al., 1995; lysosomal membrane into the lysosomal matrix where they are Cuervo and Dice, 1996). The antibody against the cytosolic tail of rat degraded (Cuervo and Dice, 1998; Dice, 2000). Substrate lamp2a was raised in our laboratory (Cuervo and Dice, 1996). The proteins bind first to a constitutively expressed heat shock monoclonal antibodies against the lumenal side of rat lamp2a, rat protein of 73 kDa (hsc73) in the cytosol that targets them to lamp1, human influenza hemaglutinin protein (HA) and cathepsin A lysosomes (Chiang et al., 1989). A second chaperone located were gifts from Dr Michael Jadot (Facultes Universitaires Notre- in the lumen of the lysosomes, the lysosomal hsc73, is required Dame de la Paix, Namur, Belgium), Dr Ira Mellman (Yale University for the complete transport of substrate proteins into lysosomes School of Medicine, New Heaven, CT), Dr Anjana Rao (Department (Agarraberes et al., 1997; Cuervo et al., 1997). We of Pathology, Harvard Medical School, Boston, MA), and Dr demonstrated that substrate proteins bind to the cytosolic tail Alessandra D’Azzo, respectively. The monoclonal antibodies against of lamp2a at the lysosomal membrane before their transport the matrix region of human, mouse, and hamster lamp2s, and human lamp1 were obtained from the Developmental Studies Hybridoma and degradation in the lysosomal matrix (Cuervo and Dice, Bank (Iowa City, IA). Aminolink and Sulfolink gels and 1996). Inhibition of that binding with specific antibodies crosslinking agents were from Pierce (Rockford, IL). against the cytosolic tail of lamp2a or by competition with a synthetic of the same amino acid sequence as the tail, Isolation of subcellular fractions results in blocking of substrate uptake and degradation in Rat liver lysosomes were isolated from a mitochondrial- lysosomes (Cuervo and Dice, 1996). Interestingly, the lysosomal fraction in a discontinuous metrizamide density gradient overexpression of only lamp2a in cultured cells increased the (Wattiaux et al., 1978) by the shorter method reported previously rates of chaperone-mediated autophagy, suggesting that (Aniento et al., 1993). After isolation lysosomes were resuspended binding of the substrates to lamp2a at the lysosomal membrane in MOPS buffer (0.3 M sucrose/10 mM 3-(N-morpholino) might be a rate-limiting step in this pathway (Cuervo and Dice, propanesulfonic acid (MOPS), pH 7.2). In some experiments, two separate lysosomal fractions with different levels of chaperone- 1996). mediated autophagy were isolated as described (Cuervo et al., 1997). In the present study, using an antibody specific for lamp2a Contamination of the lysosomal fraction with mitochondria (based and another that recognizes all lamp2s we compare on the activity of ornithine transcarbamoylase and succinate concentrations, biochemical properties, subcellular location, dehydrogenase) or cytosol (based on the activity of lactate and lysosomal distribution of lamp2s in rat liver and fibroblasts dehydrogenase and GAPDH) could account for less than 0.5% of the in culture, the two systems in which chaperone-mediated lysosomal fraction. Lysosomes from cultured cells were isolated as autophagy has been well-characterized. We demonstrate that described (Storrie and Madden, 1990). Integrity of the lysosomal binding of substrate proteins to lamp2a is a rate-limiting membrane after isolation was measured by β-hexosaminidase latency step in chaperone-mediated autophagy under a variety of as previously described (Terlecky and Dice, 1993). Only preparations physiological and pathological conditions, while the other with more than 95% intact lysosomes were used. Lysosomal matrices and membranes were obtained as described by Oshumi (Ohsumi et lysosomal forms of lamp2a do not affect chaperone-mediated al., 1983). autophagy. Finally, we identify unique properties of lamp2a, such as the presence of a group of positive residues in its Purification of lamp2a from rat liver lysosomes cytosolic tail essential for substrate binding/uptake. Lysosomal membranes were solubilized in 50 mM Tris-HCl, pH 8, 150 mM NaCl, 1% Nonidet-P40, 0.5% sodium deoxycholate, 0.1% SDS for 2 hours at 0°C. Lysates were cleared by centrifugation at 100,000 g for 30 minutes, and solubilized membrane proteins were MATERIALS AND METHODS recovered in the supernatant. Lysosomal membrane and matrix proteins were separately loaded in a column containing the antibody Animals and cells against the cytosolic tail of lamp2a immobilized in Aminolink Plus Adult male Wistar rats weighing 200-250 g and fasted for 20 hours gel previously equilibrated in lysis buffer. After extensive washing before sacrifice were used. Where indicated fasting was extended for immobilized proteins were eluted with 100 mM glycine, pH 2.3, and longer times with free access to water. An age-controlled rat strain collected in neutralizing Trizma base. (Fischer 344) was used for the study of age-related changes, and 3- and 22-month old rats were compared. For some studies, 2,2,4- Uptake and degradation of substrate proteins by isolated trimethylpentane, dissolved 1:1 in corn oil, was administered to the rat liver lysosomes animals by gavage (1 g/kg of body weight) during 7 consecutive days. Substrate proteins were incubated with chymostatin-treated Human lung fibroblasts (IMR-90) were from the Coriell Cell lysosomes as previously described (Aniento et al., 1993; Cuervo et Repositories (Camden, NJ), and Chinese Hamster ovary cells (CHO), al., 1994). Transport was measured after proteinase K treatment of the human embryo kidney cells (HEK293), and rat lung fibroblasts (RFL- samples, SDS-PAGE and immunoblot, as the amount of substrate 6) were from the American Type Culture Collection (Manassas, VA). resistant to the . Degradation of glyceraldehyde-3-phosphate Mouse skin fibroblasts were generously provided by Dr Alessandra dehydrogenase (GAPDH) by isolated intact lysosomes was measured D’Azzo (St Jude Children’s Research Hospital, Memphis, TN). Cells as described (Terlecky and Dice, 1993). Lysosomes (25 µg protein) were maintained in Dubelcco’s modified Eagle’s medium (Sigma, St were incubated in 10 mM MOPS, pH 7.2, 300 mM sucrose, 1 mM Louis, MO) in the presence of 10% newborn calf serum (NCS), except dithiothreitol, 5.4 µM cysteine for 30 minutes at 37°C with 260 nM for CHO cells that were grown in F-12 medium (Life Technologies, GAPDH radioactively labeled with [14C] by reductive methylation Gaithersburg, MD) with the same amount of serum. To deprive cells (Jentoft and Dearborn, 1983). Reactions were stopped by the addition Lamp2s and chaperone-mediated autophagy 4443 of trichloroacetic acid to a final concentration of 10%. Acid-soluble analyzed and colocalization determined using MetaMorph (Universal material (amino acids and small ) was collected by filtration Imaging). All digital microscopic images were prepared using Adobe through a Millipore Multiscreen Assay System (Millipore, Bedford, Photoshop 5.0 software (Adobe Systems Inc., Mountain View, CA). MA) using a 0.45 µm pore filter, and the acid-precipitable material (protein and larger peptides) was collected on the filter. Radioactivity Centrifugation in sucrose gradients in the samples was converted to disintegrations per minute in a Rat liver lysosomes were incubated with 1 mM dithiobis(suc- P2100TR Packard liquid scintillation analyzer by correcting for cinimidylpropionate), a membrane- permeable reversible crosslinker, quenching using an external standard (Packard Instruments, Meriden, in MOPS buffer for 30 minutes at 25°C. The reaction was quenched CT). Proteolysis was expressed as a percentage of the initial acid- with 50 mM Tris-HCl, pH 7.5, and lysosomes were recovered by insoluble radioactivity converted to acid-soluble radioactivity at the centifugation and separated into membrane and matrices as described end of the incubation. above. Lysosomal membranes were solubilized in 50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.2% octyl-glucoside. Solubilized lysosomal Expression of lamp2a and HA-lamp2a membranes were loaded on the top of a continuous gradient of sucrose The cDNAs for native (Carlsson et al., 1988; Fukuda, 1991) and (20% to 60%) (5 ml total) in solubilization buffer. Samples were mutated human lamp2a were subcloned in the pCR3 mammalian centrifuged at 140,000 g for 24 hours and aliquots of 200 µl were expression vector (Invitrogen, San Diego, CA), and CHO cells were separately collected from the top to the bottom of the gradient. Thirty transfected with these constructs by the calcium phosphate method µl of each aliquot were used for SDS-PAGE and immunoblot with (Maniatis et al., 1982). After Geneticin (Life Technologies) selection antibodies. A parallel sucrose gradient with molecular mass markers 5-10 different clones for each construct were isolated and assayed for (Sigma) was included in each experiment. their protein expression levels by immunoblot. Clones expressing similar levels of lamp2a were used for the protein degradation assays. General methods For purification purposes three consecutive HA epitope-tags were Protein was determined by the Lowry method (Lowry et al., 1951) inserted in the lumenal hinge region of the cDNAs for native and using bovine serum albumin as a standard. Standard procedures were mutated human lamp2a. Human embryo kidney cells were transfected used for the determination of enzymatic activities as reported with those constructs by the same method as above. After 48 hours previously (Terlecky and Dice, 1993; Cuervo et al., 1994). CHO cells cells were harvested and solubilized in 1% octyl-glucoside in 50 mM were stably transfected with the human lamp2a cDNA and mutant Tris-HCl, pH 8, 150 mM NaCl. Point mutations were performed with lamp2a cDNAs as described previously (Cuervo and Dice, 1996). the QuikChangeTM Site-Directed Mutagenesis Kit (Stratagene, La Intracellular protein degradation in control and transfected CHO cells Jolla, CA) following the manufactures instructions and were verified was measured as described (Auteri et al., 1983). Hsc73 was purified by DNA sequencing. from rat liver cytosol by ATP-agarose affinity chromatography (Welch and Feramisco, 1985). Cellular fractions and isolated proteins were Binding of proteins to lamp2a subjected to SDS-PAGE in slab gels (Laemmli, 1970). Gels were Specific binding of cytosolic proteins to lamp2a was analyzed as electrotransferred to nitrocellulose membranes using a Mini-Trans described (Cuervo and Dice, 1996). Briefly, lysosomal membranes Blot SD wet transfer cell (Bio-Rad, Richmond, VA), and were subjected to SDS-PAGE and electrotransferred to a immunoblotting was performed following standard procedures nitrocellulose membrane. After blockage with 5% nonfat dry milk in (Towbin et al., 1979). Membranes were developed by 20 mM Tris-HCl, pH 7.2, 150 mM NaCl and 0.1% Tween-20, chemiluminescence methods (Renaissacence, NEN-Life Science membranes were incubated with 14C-GAPDH (250 nM) in a Products, Boston, MA). Densitometric quantification of the renaturation buffer (50 mM Tris-HCl, 100 mM potassium acetate, 150 immunoblotted membranes was performed in Kodak, Scientific mM NaCl, 1 mM dithiothreitol, 5 mM MgCl2, 1 mM EDTA, and 0.3% Imagin Film, using an Image Analyzer System (Inotech S-100, Tween-20) for 12 hours at 4°C. Bound protein was detected after Sunnyvale, CA). Statistical analyses were carried out using the extensive washing by exposure to a PhosphorImager screen. For the Student’s t-test. binding of GAPDH to wild type and mutant recombinant HA-lamp2as the solubilized fraction from transiently transfected HEK293 cells was passed through an anti-HA Sulfolink matrix and extensively washed with 50 volumes of 50 mM Tris-HCl, pH 8, 500 mM NaCl and 0.2% RESULTS octyl-glucoside. The beads were resuspended in 10 volumes of 50 mM Tris-HCl, pH 8, 150 mM NaCl, 0.2% octyl-glucoside. For each Characterization of the membrane and matrix forms binding assay 15 µl of the resuspended beads were incubated with rat of lamp2 in rat liver lysosomes µ µ liver cytosol (30 g protein) in 200 l of the above buffer for 45 The presence in rat liver of three different mRNA transcripts minutes at 37°C. After two washes with 10 volumes of buffer, the beads were collected by centrifugation and subjected to SDS-PAGE for lamp2 suggest that several forms of lamp2 proteins might and immunoblot for GAPDH. The amount of GAPDH bound to beads coexist in that tissue (Konecki et al., 1995). We have previously containing the anti-HA antibody but without lamp2a was subtracted developed an antibody against the twelve amino acids of the from each experimental value. GAPDH binding was normalized to the cytosolic region of rat lamp2a (Cuervo and Dice, 1996). This amount of HA-lamp2a in each sample. antibody does not recognize the cytosolic tail of the human lamp2a that only differs from the rat in two out of twelve amino Confocal microscopy acids (GLKHHHAGYEQF (human) vs GLKRHHTGYEQF Immunofluorescence staining of cultured cells was performed (rat)). The cytosolic regions of the other two predicted splice following conventional procedures. Cells were grown on coverslips variants of lamp2 in rodent, lamp2b and lamp2c, differ from until confluent and kept in the presence or absence of serum for 20 the a isoform in 8 and 9 out of 12 amino acids, respectively hours before fixing with a 3% formaldehyde solution. Coverslips were blocked and incubated with the primary and corresponding fluorescein (Fig. 1A, shadowed residues). These large differences between isothiocyanate (FITC)- or Texas Red (TR)-conjugated secondary the cytosolic regions of the predicted spliced variants of rodent antibodies. Images were acquired as single scans on a Bio-Rad MRC- lamp2 make it very unlikely that the antibody to the cytosolic 1024 confocal microscope (Bio-Rad Laboratories, Hercules, tail of lamp2a also recognizes any of the other two variants. California, USA) equipped with an argon laser. Images were further To verify the specificity of the antibody, as well as to 4444 A. M. Cuervo and J. F. Dice determine the contribution of lamp2a to the total content of lamp2s in rat liver lysosomes, we subjected lysosomal membranes and matrices to affinity chromatography through a matrix containing the antibody raised against the cytosolic tail of lamp2a. We then analyzed the retained fraction and the flow through with the same antibody (Fig. 1B, left), and with an antibody raised against the lumenal region common for all lamp2s (Fig. 1B, right). Fig. 1B, right, shows the presence of lamp2 isoforms that did not bind the antibody (lanes 3 and 4), and Fig. 1B, left, shows that the antibody against the cytosolic tail of lamp2a did not recognize these isoforms (lanes 3 and 4). We found similar electrophoretic mobility for the lamp2 in the retained and unretained membrane fractions after completely removing the groups with different glycosidases (data not shown). These results suggest that lamp2 not retained by the antibody against the lamp2a cytosolic tail corresponded to a complete form of the protein with a different cytosolic region and support the specificity of our antibody for lamp2a. We were able to isolate lamp2a from both the membrane and, in lower amount, the matrix of lysosomes (Fig. 1B left, Fig. 1. Distribution of lamp2s in lysosomes. (A) Amino acid lanes 1 and 2). The quantification of the purification of lamp2a sequence of the cytosolic region of the different mouse lamp2 by affinity chromatography revealed that lamp2a accounts for isoforms (Cough et al., 1995). (B) Solubilized lysosomal membranes approximately 25% of the total lamp2s at the lysosomal (MB) and matrices (Mtx) were loaded onto a column containing membrane, and 10% of the total lamp2s in the lysosomal immobilized antibody against the cytosolic region of lamp2a. matrix. In both fractions some of the lamp2 was not recognized Aliquots of the flow through and the eluted fraction were subjected to by the cytosolic tail antibody so did not contain the cytosolic SDS-PAGE and immunoblotted with the antibody against the region (truncated form) or contained a different one not cytosolic region of lamp2a (left) or against a lumenal region common recognized by the antibody (Fig. 1B, right, lanes 3-4). Other for all lamp2s (right). (C) Lysosomes from rat liver (100 µg protein) authors have suggested that the lamp2s detected in the were isolated and their corresponding membranes (MB) and matrices lysosomal matrix might be truncated forms lacking the (Mtx) were subjected to SDS-PAGE and immunoblotted for lamp2a (anti-cytosolic), all lamp2s (anti-matrix), lamp1, and cathepsin A, as cytosolic and transmembrane regions (Jadot et al., 1996; labeled. Akasaki and Tsuji, 1998). However, at least for lamp2a, most of the protein detected in the matrix still contains the cytosolic of the substrates to lysosomes (Cuervo and Dice, 1996). In tail. In addition, we found similar electrophoretic mobility addition, overexpression of human lamp2a in CHO cells by for the membrane and matrix lamp2a after complete itself increased the activity of chaperone-mediated autophagy deglycosylation (data not shown). These results suggest that (Cuervo and Dice, 1996). most of the lamp2a in the matrix corresponds to the intact We have used different approaches to analyze the effect that protein. changes in the lysosomal content and distribution of lamp2a For the purification studies of lamp2a (Fig. 1B) we started have on the activity of the chaperone-mediated autophagic with the same amount of protein from lysosomal membranes pathway. Working with individual clones of CHO cells stably and matrices (300 µg), but the matrix normally contributes transfected with human lamp2a we found a very high 70% of the total lysosomal protein. When we analyzed the correlation between the levels of human lamp2a expressed in distribution of lamp2a in isolated liver lysosomes we found that each clone and rates of chaperone-mediated autophagy in those 70-80% of lamp2a was located in the lysosomal membrane and clones (Fig. 2A). In the clones overexpressing lamp2a the rates 30-20% in the matrix (Fig. 1C, lanes 1 and 2). In contrast, the of proteolysis in the presence of serum were only slightly distribution of all the lamp2s considered together was 55% in higher than in the control CHO cells. However, the increase in membrane and 45% in matrix (Fig. 1C, lanes 3 and 4). The protein degradation after serum removal was significantly presence of another lysosomal membrane protein, lamp1, only higher in clones 2 and 3 of the lamp2a-expressing cells. Fig. in the lysosomal membrane fraction (Fig. 1C, lanes 5 and 6), 2A shows the increase in the rate of proteolysis induced and the predominantly matrix protease, cathepsin A, only in by serum deprivation for each clone corrected for their the matrix (Fig. 1C, lanes 7 and 8) demonstrate the purity of corresponding values in the presence of serum. The response our membrane and matrix preparations. to serum removal proportionally increased with the increase in the lysosomal content of human lamp2a. These results suggest Changes in levels of lamp2a at the lysosomal that levels of lamp2a at the lysosomal membrane can be a rate- membrane correlate with changes in the rate of limiting step in the transport and degradation of substrate chaperone-mediated autophagy proteins in lysosomes. We have previous evidence suggesting that lamp2a is the form To determine if the levels of lamp2a at the lysosomal of lamp2 contributing most to substrate binding. Thus, using a membrane also correlate with the activity of the chaperone- synthetic peptide with identical amino acid sequence as the mediated autophagic pathway under physiological conditions cytosolic tail of lamp2a we blocked almost 80% of the binding we isolated lysosomes from rats subjected to different periods Lamp2s and chaperone-mediated autophagy 4445

Fig. 2. Lysosomal levels and distribution of lamp2a under conditions that result in changes in the rate of chaperone-mediated autophagy. (A) CHO cells were stably transfected with human lamp2a cDNA, and individual clones were grown separately. The rate of total protein degradation in the presence or absence of serum was measured for three different clones as described in Materials and Methods. The best exponential decay curve was calculated by linear regression analysis. At each time point the percentage of total protein degraded in the presence of serum was subtracted from the percentage of protein degraded in the absence of serum for each individual clone. The graphic shows the mean values of those differences for each clone in 4 different experiments. Standard errors were less than 10% of the mean values. Inset shows the levels of human lamp2a in homogenates (H) and lysosomes (L) from the selected clones and nontransfected CHO cells detected by immunoblot. For reference a homogenate from human fibroblasts (IMR-90) is shown. (B) Lysosomes were isolated from livers of rats starved for different periods of time as labeled. Levels of lamp2a (top) and lamp1 (bottom) in 100 µg of protein from lysosomal membranes were analyzed by SDS-PAGE and immunoblot. (C) Lysosomes were isolated from rat fibroblasts cultured in the presence (s +) or absence of serum (s −) for 20 hours. Total lysosome (Lysosom), lysosomal membranes (L.MB) and matrices (L.Mtx) (10 µg of protein) were processed as in B. (D) Two lysosomal populations with a high (h +) and low (h −) activity of chaperone-mediated autophagy were isolated as described under Materials and Methods. Intact lysosomes (100 µg of protein) and their corresponding membranes and matrices were processed as in B. of starvation. The lysosomal membrane levels of lamp2a slightly higher in the more active population compared to the increased with the starvation time through the entire less active group. After separating lysosomal membranes and starvation period analyzed (88 hours) (Fig. 2B, top), as did matrices we found that 78% of the lamp2a was present at the also the rate of chaperone-mediated autophagy (Cuervo et al., lysosomal membrane and 22% at the matrix in the active 1995). We found an increase in the lysosomal levels of lamp1 population while the lamp2a distribution was 57% membrane during the first 20 hours of starvation, but after that time and 43% matrix in the less active population (Fig. 2D top, lanes lamp1 levels remained steady (Fig. 2B, bottom). Levels of 3-6). These results support the idea that changes in the lamp2a in the lysosomal matrix did not significantly change distribution of lysosomal lamp2a between the membrane and until the most prolonged periods of starvation, when they the matrix could contribute to the regulation of chaperone decreased (data not shown). We also found an increase in the mediated-autophagy (Cuervo and Dice, 2000a). For unknown lysosomal membrane content of lamp2a in rat fibroblasts in reasons lamp1 was enriched in the membrane of the less active culture in response to serum deprivation, but no changes in group of lysosomes (Fig. 2D, bottom). the lysosomal membrane content of lamp1 (Fig. 2C, lanes 3- Other conditions in which we have recently identified 4). Levels of lamp2a in the lysosomal matrix remained increased chaperone-mediated autophagy in liver and kidney unchanged after serum removal (Fig. 2B, lanes 5-6). Thus, include a chemically-induced nephropathy (Cuervo et al., rather than total lysosomal levels of lamp2a, levels of the 1999). Chaperone-mediated autophagy is activated after protein at the lysosomal membrane correlate with the activity exposure to gasoline additives and results in increased uptake of the lysosomal pathway. of one substrate, alpha-2-microglobulin. In these studies we We then analyzed the lamp2a distribution in lysosomal found higher levels of lamp2a in the membranes of lysosomes populations with different rates of chaperone-mediated isolated from intoxicated animals than from the controls. We autophagy. Rat liver lysosomes can be separated based on found no significant differences in matrix content of lamp2a or density into two groups with a different content of hsc73 in lysosomal levels of hsc73 (Cuervo et al., 1999). their matrices and different activity for the direct transport of A decrease in levels of lamp2a at the lysosomal membrane substrate proteins (Cuervo et al., 1997). As shown in Fig. 2D also occurs when the activity of the pathway decreases. Thus, top (lanes 1 and 2) total lysosomal levels of lamp2a are only the activity of the chaperone-mediated autophagic pathway 4446 A. M. Cuervo and J. F. Dice gN µ 2 α ++lamp2a SR(88h) SR(64h) HSC- FR/S+ OR(22m) +lamp2a SR(20h)/YR A HSC + S - 5 r2 = 0.978 4 y = 1.09 x p < 0.001 3

2

1 lamp2a (times control value) (times control membrane Lysosomal levels lamp2a levels Lysosomal 0 0 1 2 3 4 B 4 2 r = 0.008 lamp1 y = - 0.06 x membrane 3 p = NS

2

1 (times control value) (times control Lysosomal levels lamp1 levels Lysosomal 0 0 1 2 3 4 Lysosomal Activity Fig. 3. Correlation between lysosomal membrane levels of lamp2a and activity of chaperone-mediated autophagy. Lysosomal Fig. 4. Lamp2 isoforms other than lamp2a and chaperone-mediated membranes from rat liver or cultured rat fibroblasts were prepared as autophagy. (A) Lysosomal membranes from rat liver or cultured described in Materials and Methods. Levels of lamp2a (A) and lamp1 fibroblasts as described in Fig. 3 were subjected to SDS-PAGE and (B) in those fractions were determined by densitometric analysis immunoblotted for all lamp2s. Total lamp2s values and chaperone- after immunoblotting. Lysosomal activity was measured as the ability mediated activity were calculated as described in Fig. 3. Values are of isolated lysosomes to degrade a radiolabeled substrate protein, means + s.e.m. of 5 to 10 different experiments for each of the 14C-GAPDH. The lysosomal activity for each condition analyzed is conditions analyzed. (B) Rat liver lysosomal membranes (L.MB) expressed relative to the activity of lysosomes isolated from fed rats were immunoprecipitated with the specific antibody against lamp2a. or from cells maintained in the presence of serum that were given the Lysosomal membranes, precipitates (IP) and supernatants (Post-IP) arbitrary value of 1. Values are means + s.e.m. of 5 to 10 different were subjected to SDS-PAGE and immunoblotted for lamp2a (top), experiments. Lysosomes were isolated from: fed (FR) or starved rats all lamp2s (middle) or assayed for 14C-GAPDH binding (bottom) as (SR), 3-months old (YR) or 22-months old rats (OR), 2,2,4- described in Materials and Methods. Bound protein was detected trimethylpentane-treated rats to induce α2µglobulin nephropathy using a phosphorImager screen. (α2µgN), fibroblasts maintained in the presence (S+) or absence (S-) of serum, and CHO cells stably transfected with human lamp2a and expressing different levels (+, ++ lamp2a). Two groups of section is shown in Fig. 3A, top. In contrast, we did not find a lysosomes with high (HSC+) and low (HSC-) activity for chaperone- significant correlation between lysosomal membrane levels of mediated autophagy were isolated from rats starved for 20 hours. lamp1 (Fig. 3B) or matrix levels of lamp2a (data not shown) Best straight lines were calculated by linear regression. and rates of chaperone-mediated autophagy. decreases in senescent fibroblasts in culture (Dice, 1993), and Contribution of the different forms of lamp2 to the also in lysosomes isolated from livers of old rats (Cuervo and binding of substrates for chaperone-mediated Dice, 2000b). Levels of lamp2a at the lysosomal membrane are autophagy significantly lower in old rats when compared with young rats. We also measured the lysosomal membrane levels of all forms The content of lamp2a in the lysosomal matrix does not change of lamp2 considered together in the different conditions or only decreases slightly with age. described for lamp2a. We did not find a significant correlation The linear correlation between the lysosomal membrane between the levels of all lamp2s at the lysosomal membrane levels of lamp2a and the rates of chaperone-mediated and chaperone-mediated autophagy activity (Fig. 4A). This autophagy under the different conditions described in this lack of correlation suggests that if other forms of lamp2 are Lamp2s and chaperone-mediated autophagy 4447

Fig. 5. Effect of changes in the amino acid sequence of the cytosolic tail of lamp2a on chaperone-mediated autophagy. (A) Amino acid sequence of the cytosolic region of lamp2a (L2) and the mutations performed (L2/x). Replaced amino acids are shadowed. (B) CHO cells were stably transfected with native (L2) and mutated (L2/x) human lamp2a cDNAs and individual clones were grown separately. For stably transfected clones expressing similar levels of the respective form of human lamp2a the rate of total protein degradation in the absence of serum was measured as described in Materials and Methods. Values were calculated as in Fig. 2A and are means of 6 different experiments. Standard error were less than 10% of mean values. (C) The rate of 14C-GAPDH degradation by intact lysosomes isolated from CHO cells and the different lamp2a clones described in A was measured as described in Materials and Methods. Values are means + s.e.m. of 7 different experiments. Differences from control values are significant to *P<0.01 and **P<0.002. (D) Native (L2) and mutant (L2/x) HA-lamp2a were immobilized in Sulfolink containing anti-HA antibody and their ability to bind GAPDH present in rat liver cytosol was analyzed as described in Materials and Methods. Values are means + s.e.m. of the densitometric quantification of 5 different experiments similar to the one shown in the inset. Differences from control values are significant to *P<0.01. involved in chaperone-mediated autophagy they are at least not 4 positive charges in the cytosolic tail of lamp2a by alanines rate-limiting components in that process. the rates of protein degradation during serum deprivation were To directly analyze the ability of the other forms of lamp2 indistinguishable from control cells (Fig. 5B) indicating that to bind substrates we immunoprecipitated lamp2a from the lamp2a/4A mutation was not an active receptor. solubilized lysosomal membranes and then measured the Isolated lysosomes containing the lamp2a lacking the four binding of 14C-GAPDH to the remaining lamp2s. GAPDH positive residues showed lower ability for GAPDH binding, only bound to the fractions containing lamp2a but not to uptake, and degradation than the ones containing native or the fractions containing the remaining forms of lamp2s (Fig. 4B, other mutated forms of lamp2a (Fig. 5C). Using two different lanes 1-3). These results confirm our previous hypothesis that substrate binding assays to immobilized lysosomal membranes lamp2a is the only isoform of lamp2 involved in binding of (data not shown) and to purified recombinant HA-lamp2a (Fig. substrate proteins for chaperone-mediated autophagy. 5D), we demonstrated that the decreased GAPDH degradation was a consequence of lower binding of substrate proteins to Amino acid residues of lamp2a required for the lamp2a lacking the four positive residues. In this binding substrate binding assay we found that elimination of a single positive residue or We have previously demonstrated that substrate proteins changes in residues distal to the GY dipeptide, including the directly bind to the cytosolic region of lamp2a (Cuervo and hydrophobic terminal residue, did not modify lamp2a substrate Dice, 1996). To analyze which of the 12 amino acid residues binding ability (Fig. 5D). of the cytosolic tail of lamp2a, if not all, were involved in substrate binding we performed site-directed mutagenesis of Unique characteristics of lamp2a compared to other that area as shown in the diagram in Fig. 5A. We did not lamp2s modify the GY dipeptide because it is conserved in all forms Besides the above described differences in lysosomal content of lamp2s and it is required for targeting of lamp2a to and distribution between lysosomal membrane and matrix (Fig. lysosomes (Williams and Fukuda, 1990). The last hydrophobic 1), the membrane-associated form of lamp2a also differed from amino acid (F) is also important for lamp2a targeting other forms of lamp2 in its sedimentation rates after cross- (Guarnieri et al., 1993), but since it is not conserved in the other linking. We found that after centrifugation of detergent- forms of lamp2a we mutated it and analyzed its binding solubilized lysosomal membranes through a continuous 20%- properties in vitro. Except for that mutation all the other 60% sucrose gradient lamp2a could be detected in four lamp2a mutants were properly targeted to lysosomes (data not different regions of the gradient corresponding approximately shown). As we have previously shown for wild type lamp2a, to 100, 200, 400 kDa and 800 kDa (Fig. 6A). After reversing cells overexpressing lamp2a/EQ or lamp2a/A mutants had the crosslinking and analyzing the different fractions by SDS- higher rates of protein degradation after serum removal than PAGE and silver staining, we found that lamp2a was the only control CHO cells (Fig. 5B). However, after replacement of the protein component of the 400 kDa complex (data not shown). 4448 A. M. Cuervo and J. F. Dice

throughout the . Chaperone-mediated autophagy is A Sucrose 60% activated in these cells after serum removal (Dice et al., 1986). 20% Under those conditions lamp2a was preferentially found in lysosomes surrounding the nucleus (Fig. 7A and B,). A similar 120 pattern of distribution has been previously observed for the lysosomes more active for chaperone-mediated autophagy 800 KDa 200 KDa 100 100 KDa 400 KDa based on their hsc73 content (F. Agarraberes and J. F. Dice, unpublished results). We did not find differences in the 80 distribution of the other forms of lamp2 in the presence or absence of serum (Fig. 7A). The right panel shows the merged 60 image of the staining with both antibodies (notice that the antibody against lamp2s recognizes lamp2a but with low

lamp2a (a.d.u.) 40 affinity (Fig. 1A right, lane 1). Though to a lesser extent than for lamp2a, the staining for lamp1 also increased in the 20 perinuclear region after removal of serum (Fig. 7B, bottom).

0 0 2 4 6 8 10 12 DISCUSSION B Volume (ml) Chaperone-mediated autophagy is responsible for the selective Sucrose 60% degradation of cytosolic proteins in lysosomes during stress 20% conditions (Cuervo and Dice, 1998; Dice, 2000). Here we 120 show that the activity of chaperone-mediated autophagy is modulated under different physiological and pathological 100 conditions by changes in the levels of lamp2a at the lysosomal membrane. Binding of substrate proteins to lamp2a is a rate- 80 limiting step for their degradation (Fig. 2A), and there is a direct correlation between the levels of lamp2a at the lysosomal membrane and the rate of chaperone-mediated autophagy (Figs 60 2 and 3). Interestingly, in spite of the high similarity between different lamp2s, other isoforms of lamp2 do not participate in 40 lamp2s (a.d.u.) substrate binding (Fig. 4). The presence of a group of positive residues in the cytosolic tail of lamp2a, absent in the other 20 forms of lamp2, is required for its ability to selectively bind cytosolic protein substrates for chaperone-mediated autophagy 0 0 2 4 6 8 10 12 (Fig. 5). In addition, the unique ability of the lamp2a at the lysosomal membrane to multimerize may be of relevance for Volume (ml) substrate transport into lysosomes (Fig. 6). Fig. 6. Multimerization of lamp2a in the lysosomal membrane. Rat Different cellular functions for different forms of lamp2 liver lysosomes were reversibly crosslinked and separated into have been proposed based on their tissue-specific expression membranes and matrices as described in Materials and Methods. (Konecki et al., 1995). However, a specific function for one of Solubilized lysosomal membranes were subjected to sedimentation them but not for the others has not been described until now. through a continuous sucrose gradient (20-60%) as described in We show here that only lamp2a is involved in the selective Materials and Methods. Aliquots of each of the fractions collected lysosomal binding of substrates for chaperone-mediated from the top of the gradient were subjected to SDS-PAGE and autophagy in rat liver and fibroblasts in culture (Fig. 4). We immunoblotted for lamp2a (A) or for all lamp2s (B). The cannot discard the possibility that the antibody raised against densitometric quantification of those immunoblots and the calculated the cytosolic region of lamp2a might also recognize other molecular mass in kDa are shown. lamp2s with minor modifications in that region but certainly not the spliced forms described so far. In addition, it is clear These results suggest that, though lamp2a could also associate that at least some of the other lysosomal forms of lamp2 are with other proteins, at least part of the lamp2a at the lysosomal not recognized by this antibody (Fig. 1B, lanes 3 and 4, and membrane is in the form of homomultimers. The other lamp2 Fig. 4B, lane 3). Further studies with specific antibodies for forms were mainly located in the 100 to 200 kDa region (Fig. each of the described cytosolic tails of lamp2 isoforms might 6B). Therefore, the ability to form high molecular mass lead to a better understanding of the roles of the other isoforms. complexes at the lysosomal membrane seems more of a Several types of evidence support our hypothesis that characteristic of lamp2a than the other isoforms. lamp2a is responsible for most of the binding of substrates to When we analyzed the intracellular distribution of lamp2a the lysosomal membrane in all the conditions that we have compared to other isoforms of lamp2 (Fig. 7A, top) or to lamp1 analyzed: (1) the overexpresssion of lamp2a alone increases (Fig. 7B, top) in mouse skin fibroblasts cultured under normal rates of chaperone-mediated autophagy (Fig. 2A; Cuervo and conditions we did not find significant differences. The three Dice, 1996), (2) the increase or decrease of lysosomal levels antibodies displayed a typical vesicular pattern distributed of lamp2a results in increase or decrease in binding of substrate Lamp2s and chaperone-mediated autophagy 4449

Fig. 7. Serum-dependent changes in the subcellular distribution of lamp2a. Mouse skin fibroblasts grown in the presence (top) or absence (bottom) of serum were processed for indirect immunofluorescence and labeled with the antibody for lamp2a and all lamp2s (A) or for lamp2a and lamp1 (B) as labeled. Right panels show merging images. ×800. proteins to lysosomes (Figs 2 and 3), (3) the group of positive amino acids that participate in substrate binding are absent in the cytosolic tail of the other forms of lamp2 (Figs 1A, 5), (4) the multimerization of lamp2 at the lysosomal membrane may be important for the selective uptake of substrates in lysosomes (Cuervo and Dice, unpublished results), and is mainly detected for lamp2a (Fig. 6), and (5) the serum-related changes in the subcellular location of lamp2a are similar to the changes in cellular distribution of active lysosomes for chaperone- mediated autophagy (Fig. 7; F. Agarraberes and J. F. Dice, unpublished results). The cytosolic protein substrates for chaperone- mediated autophagy all contain a pentapeptide motif biochemicaly related to KFERQ that targets them for lysosomal degradation (Chiang and Dice, 1988; Dice, 1990). That motif is not involved in the binding to the receptor at the lysosomal membrane (A. M. Cuervo and J. F. Dice, unpublished results). The requirement for the four positive residues in the cytosolic tail of lamp2a for substrate binding, described here (Fig. 5), suggests a charge-mediated substrate/receptor interaction. Different explanations have been reasoned about the origin of the lamp2 detected in the lysosomal matrix such as its cleavage from the lysosomal membrane or its direct transport by vesicular fusion from the endosomal compartment lysosomal distribution of different forms of lamp2 (Fig. 2D) (Jadot et al., 1996; Jadot et al., 1997). Since glycosylated and also serum-dependent changes in the intracellular proteins have hydrodynamic volumes that differ per unit of distribution of lamp2a (Fig. 7). Similar perinuclear molecular mass from those of globular proteins used as distribution to the one shown here for lamp2a has been molecular mass standards in our studies, we can not rule out described for the lysosomes most active for chaperone- the possibility that some of the matrix lamp2s not recognized mediated autophagy using other markers such as hsc73 (F. by the antibody against the cytosolic tail correspond to Agarraberes and J. F. Dice, unpublished results). In addition, truncated forms of lamp2a (Fig. 1A, right lane 4). Other after prolonged starvation in rats, and based on the lysosomal authors have described formation of tetramers of truncated hsc73 content, we have proposed the recruitment for lamp2s, lacking a portion of its carboxy-terminal region chaperone-mediated autophagy of lysosomes normally less (Akasaki and Tsuji, 1998). However, we demonstrate here that active for this pathway (Cuervo et al., 1997). It is thus at least a part of the lamp2a present in the lysosomal matrix possible that under extreme conditions, when chaperone- corresponds to the intact protein (Figs 1, 2, 4B). Thus, cleavage mediated autophagy reaches its maximum activity, a specific might explain the origin of some of the truncated forms of group of lysosomes receives lamp2a by fusion with other lamp2 in the matrix, but there must be other mechanisms by vesicles (endosomes or lysosomes) containing lamp2a. which intact forms of lamp2a appear in the lysosomal lumen Levels of lamp2a not only in the lysosomal membrane but (Cuervo and Dice, 2000a). also in the matrix are dynamic and independent of the changes We now have evidence supporting the recruitment of part of of other lysosomal membrane proteins (Figs 2, 3 and 4; Cuervo the matrix lamp2a toward the lysosomal membrane under and Dice, 2000a). Those changes in levels of lamp2a at the specific conditions that activate chaperone-mediated lysosomal membrane under different conditions have a direct autophagy (Cuervo and Dice, 2000a). Interestingly, although effect on the rates of chaperone-mediated autophagy (Fig. 3). matrix levels of lamp2a do not correlate in general with the rate Further studies on the mechanisms that control levels of of substrate degradation (Fig. 3B), in conditions such as lamp2a at the lysosomal membrane might help us to prolonged starvation, we found an opposite change in understand how chaperone-mediated autophagy itself is membrane and matrix levels of lamp2a (Fig. 3A and B). regulated. In addition to the transfer from the lysosomal membrane, it is still possible that part of the matrix lamp2a might This work was supported by the National Institutes of Health grants originate from other vesicular structures after fusion with AG06116 (J.F.D.) and AG00829 (A.M.C.), and by a Research Grant lysosomes. In this work we show heterogeneity in the from the American Federation of Aging Research (A.M.C.). 4450 A. M. Cuervo and J. F. Dice

REFERENCES Gough, N. R. and Fambrough, D. M. (1997). Different steady state subcellular distribution of the three splice variants of lysosome-associated membrane Agarraberes, F., Terlecky, S. R. and Dice, J. F. (1997). An intralysosomal protein LAMP-2 are determined largely by the COOH-terminal amino acid hsp70 is required for a selective pathway of lysosomal protein degradation. J. residue. J. Cell Biol. 137, 1161-116779. Cell Biol. 137, 825-834. Guarnieri, F. G., Arterburn, L. M., Penno, M. B., Cha, Y. and August, J. T. Akasaki, K., Fukuzawa, H., Kinoshita, H., Furuno, K. and Tsuji, H. (1993). (1993). The motif tyr-X-X-hydrophobic residue mediates lysosomal Cycling of two endogenous lysosomal membrane proteins, lamp-2 and acid membrane targeting of lysosome-associated membrane protein 1. J. Biol. phosphatase, between the cell surface and lysosomes in cultured rat Chem. 268, 1941-1946. hepatocytes. J. Biochem. 114, 598-604. Hatem, C. L., Gough, N. R. and Fambrough, D. M. (1995). Multiple mRNAs Akasaki, K. and Tsuji, H. (1998). Purification and characterization of a soluble encode the avian lysosomal membrane protein LAMP-2 resulting in form of lysosome-associated membrane -2 (lamp-2) from rat liver alternative transmembrane and cytoplasmic domains. J. Cell Sci. 108, 2093- lysosomal contents. Biochem. Mol. Biol. Int. 46, 197-206. 2100. Andrejewski, N., Punnonen, E. L., Guhde, G., Tanaka, Y., Lullmann-Rauch, Jadot, M., Wattiaux, R., Mainferme, F., Dubois, F., Claessens, A. and R., Hartmann, D., von Figura, K. and Saftig, P. (1999). Normal lysosomal Wattiaux-De Coninck, S. (1996). Soluble form of Lamp II in purified rat morphology and function in LAMP-1-deficient mice. J. Biol. Chem. 274, liver lysosomes. Biochem. Biophys. Res. Commun. 223, 353-359. 12692-12701. Jadot, M., Dubois, F., Wattiaux-DeConinck, S. and Wattiaux, R. (1997). Aniento, F., Roche, E., Cuervo, A. M. and Knecht, E. (1993). Uptake and Supramolecular assemblies from lysosomal matrix proteins and complex degradation of glyceraldehyde-3-phosphate dehydrogenase by rat liver lipids. Eur. J. Biochem. 249, 862-869. lysosomes. J. Biol. Chem. 268, 10463-10470. Jentoft, N. and Dearborn, D. G. (1983). Protein labeling by reductive Auteri, J. S., Okada, A., Bochaki, V. and Dice, J. F. (1983). Regulation of alkylation. Meth. Enzymol. 91, 570-579. intracellular protein degradation in IMR- 90 human diploid fibroblasts. J. Cell. Konecki, D. S., Foetisch, K., Zimmer, K. P., Schlotter, M. and Lichter- Physiol. 115, 159-166. Konecki, U. (1995). An alternatively spliced form of the human lysosome- Carlsson, S. R., Roth, J., Piller, F. and Fukuda, M. (1988). Isolation and associated membrane protein-2 gene is expressed in a tissue-specific manner. characterization of human lysosomal membrane , h-lamp-1 and Biochem. Biophys. Res. Commun. 215, 757-767. h-lamp-2. J. Biol. Chem. 263, 18911-18919. Laemmli, U. (1970). Cleavage of structural proteins during the assembly of the Chiang, H. L. and Dice, J. F. (1988). Peptide sequences that target proteins for head of the bacteriophage T4. Nature 227, 680-685. enhanced degradation during serum withdrawal. J. Biol. Chem. 262, 6797- Licheter-Konecki, U., Moter, S. E., Krawisz, B. R., Schlotter, M., Hipke, C. 6805. and Konecki, D. S. (1999). Expression patterns of murine lysosome- Chiang, H. L., Terlecky, S. R., Plant, C. P. and Dice, J. F. (1989). A role for associated membrane protein 2 (Lamp-2) transcripts during morphogenesis. a 70-kilodalton heat shock protein in lysosomal degradation of intracellular Differentiation 65, 43-58. proteins. Science 246, 382-385. Lippincott-Schwartz, J. and Fambrough, D. M. (1986). Lysosomal membrane Cuervo, A. M., Terlecky, S. R., Dice, J. F. and Knecht, E. (1994). Selective dynamics: structure and interorganellar movement of a major lysosomal binding and uptake of ribonuclease A and glyceraldehyde-3-phosphate membrane glycoprotein. J. Cell Biol. 102, 1593-1605. dehydrogenase by rat liver lysosomes. J. Biol. Chem. 269, 26374-26380. Lowry, O. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J. (1951). Cuervo, A. M., Knecht, E., Terlecky, S. R. and Dice, J. F. (1995). Activation Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 265- of a selective pathway of lysosomal proteolysis in rat liver by prolonged 275. starvation. Am. J. Physiol. 269, C1200-C1208. Maniatis, T., Fritsch, W. F. and Sambrook, J. (1982). Molecular Cloning. A Cuervo, A. M. and Dice, J. F. (1996). A receptor for the selective uptake and Labroratory Manual. Cold Spring Harbor: Cold Spring Harbor Laboratory degradation of proteins by lysosomes. Science 273, 501-503. Press. Cuervo, A. M., Dice, J. F. and Knecht, E. (1997). A population of rat liver Meikle, P. H., Yan, M., Ravenscroft, E. M., Isaac, E. L., Hopwood, J. J. and lysosomes responsible for the selective uptake and degradation of cytosolic Brooks, D. A. (1999). Altered trafficking and turnover of LAMP-1 in Pompe proteins. J. Biol. Chem. 272, 5606- 5615. disease-affected cells. Mol. Genet. Metab. 66, 179-188. Cuervo, A. M., Hildebrand, H., Bomhard, E. M. and Dice, J. F. (1999). Direct Nishiyama, K., Fukuda, A., Morita, K. and Tokuda, H. (1999). Membrane lysosomal uptake of alpha2-microglobulin contributes to chemically induced deinsertion of SecA underlying proton motive force-dependent stimulation of nephropathy. Kidney Int. 55, 529-545. protein translocation. EMBO J. 18, 1049-1058. Cuervo, A. M. and Dice, J. F. (1998). Lysosomes, a meeting point of proteins, Ohsumi, Y., Ishikawa, T. and Kato, K. (1983). A rapid and simplified method chaperones, and . J. Mol. Med. 76, 6-12. for the preparation of lysosomal membranes from rat liver. J. Biochem. 93, Cuervo, A. M. and Dice, J. F. (2000a). Regulation of lamp2a levels in the 547-556. lysosomal membrane. Traffic 1, 570-583. Peters, C. and von Figura, K. (1994). Biogenesis of lysosomal membranes. Cuervo, A. M. and Dice, J. F. (2000b). Age-related decline in chaperone- FEBS Lett. 346, 108-114. mediated autophagy. J. Biol. Chem. 275, 31505-31513. Ravetch, J. V. and Perussia, B. (1989). Alternative membrane forms of Fc Dice, J. F., Chiang, H.-L., Spencer, E. P. and Backer, J. M. (1986). Regulation gamma RIII(CD16) on human natural killer cells and neutrophils. Cell type- of catabolism of microinjected ribonuclease A: Identification of residues 7-11 specific expression of two genes that differ in single nucleotide substitutions. as the essential pentapeptide. J. Biol. Chem. 262, 6853-6859. J. Exp. Med. 170, 481-497. Dice, J. F. (1990). Peptide sequences that target cytosolic proteins for lysosomal Saitoh, O., Wang, W. C., Lotan, R. and Fukuda, M. (1992). Differential proteolysis. Trends Biochem. Sci. 15, 305-309. glycosylation and cells surface expression of lysosomal membrane Dice, J. F. (1993). Cellular and molecular mechanisms of aging. Physiol. Rev. glycoproteins in sublines of a human colon exhibiting distinct 73, 149-159. metastatic potentials. J. Biol. Chem. 267, 5700-5711. Dice, J. F. (2000). Lysosomal Pathways of Protein Degradation. Austin, TX, Storrie, B. and Madden, E. A. (1990). Isolation of subcellular organelles. Meth. Landes Bioscience. Enzymol. 182, 203-225. Fukuda, M. (1991). Lysosomal membrane glycoproteins. Structure, Tanaka, Y., et al. (2000). Accumulation of autophagic vacuoles and biosynthesis, and intracellular trafficking. J. Biol. Chem. 266, 21327-21330. cardiomyopathy in Lamp-2-deficient mice. Nature 406, 902-906. Furuno, K., Yano, S., Akasaki, K., Tanaka, Y., Yamaguchi, Y., Tsuji, H., Terlecky, S. R. and Dice, J. F. (1993). Polypeptide import and degradation by Himeno, M. and Kato, K. (1989). Biochemical analysis of the movement of isolated lysosomes. J. Biol. Chem. 268, 23490-23495. a major lysosomal membrane glycoprotein in the endocytic membrane system. Towbin, H., Staehelin, T. and Gordon, J. (1979). Electrophoretic transfer of J. Biochem. 106, 717-722. proteins from polyacrylamide gels to nitrocellulose sheets: procedures and Furuta, K., Yang, X. L., Chen, J. S., Hamilton, S. R. and August, J. T. (1999). some applications. Proc. Nat. Acad. Sci. USA 76, 4350-4353. Differential expression of the lysosome-associated membrane proteins in Wattiaux, R., Wattiaux-De Coninck, S., Ronveaux-Dupal, M. F. and Dubois, normal human tissues. Arch. Biochem. Biophys. 365, 75-82. F. (1978). Isolation of rat liver lysosomes by isopycnic centrifugation in a Gottschalk, S., Waheed, A., Schmidt, B., Laidler, P. and von Figura, K. metrizamide gradient. J. Cell Biol. 78, 349-368. (1989). Sequential processing of lysosomal acid phosphatase by a cytoplasmic Welch, W. J. and Feramisco, J. R. (1985). Rapid purification of mammalian thiol proteinase and a lysosomal aspartyl proteinase. EMBO J. 8, 3215-3219. 70,000-dalton stress proteins: affinity of the proteins for nucleotides. Mol. Cell Gough, N. R., Hatem, C. L. and Fambrough, D. M. (1995). The family of Biol. 5, 1229-1237. LAMP-2 proteins arises by alternative splicing from a single gene: Williams, M. A. and Fukuda, M. (1990). Accumulation of membrane characterization of the avian LAMP-2 gene and identification of mammalian glycoproteins in lysosomes requires a tyrosine residue at a particular position homologs of LAMP-2b and LAMP-2c. DNA Cell Biol. 14, 863-867. in the cytoplasmic tail. J. Cell Biol. 111, 955-966.