Portal de Periódicos da CAPES

A Mechanism Regulating Proteolysis of Specific Proteins during Renal Tubular Cell Growth

2001; Elsevier BV; Volume: 276; Issue: 22 Linguagem: Inglês

10.1074/jbc.m101777200

1083-351X

Harold A. Franch, Sira Sooparb, Jie Du, Nikia S. Brown,

Nuclear Structure and Function

Growth factors suppress the degradation of cellular proteins in lysosomes in renal epithelial cells. Whether this process also involves specific classes of proteins that influence growth processes is unknown. We investigated chaperone-mediated autophagy, a lysosomal import pathway that depends on the 73-kDa heat shock cognate protein and allows the degradation of proteins containing a specific lysosomal import consensus sequence (KFERQ motif). Epidermal growth factor (EGF) or ammonia, but not transforming growth factor β1, suppresses total protein breakdown in cultured NRK-52E renal epithelial cells. EGF or ammonia prolonged the half-life of glyceraldehyde-3-phosphate dehydrogenase, a classic substrate for chaperone-mediated autophagy, by more than 90%, whereas transforming growth factor β1 did not. EGF caused a similar increase in the half-life of the KFERQ-containing paired box-related transcription factor, Pax2. The increase in half-life was accompanied by an increased accumulation of proteins with a KFERQ motif including glyceraldehyde-3-phosphate dehydrogenase and Pax2. Ammonia also increased the level of the Pax2 protein. Lysosomal import of KFERQ proteins depends on the abundance of the 96-kDa lysosomal glycoprotein protein (lgp96), and we found that EGF caused a significant decrease in lgp96 in cellular homogenates and associated with lysosomes. We conclude that EGF in cultured renal cells regulates the breakdown of proteins targeted for destruction by chaperone-mediated autophagy. Because suppression of this pathway results in an increase in Pax2, these results suggest a novel mechanism for the regulation of cell growth. Growth factors suppress the degradation of cellular proteins in lysosomes in renal epithelial cells. Whether this process also involves specific classes of proteins that influence growth processes is unknown. We investigated chaperone-mediated autophagy, a lysosomal import pathway that depends on the 73-kDa heat shock cognate protein and allows the degradation of proteins containing a specific lysosomal import consensus sequence (KFERQ motif). Epidermal growth factor (EGF) or ammonia, but not transforming growth factor β1, suppresses total protein breakdown in cultured NRK-52E renal epithelial cells. EGF or ammonia prolonged the half-life of glyceraldehyde-3-phosphate dehydrogenase, a classic substrate for chaperone-mediated autophagy, by more than 90%, whereas transforming growth factor β1 did not. EGF caused a similar increase in the half-life of the KFERQ-containing paired box-related transcription factor, Pax2. The increase in half-life was accompanied by an increased accumulation of proteins with a KFERQ motif including glyceraldehyde-3-phosphate dehydrogenase and Pax2. Ammonia also increased the level of the Pax2 protein. Lysosomal import of KFERQ proteins depends on the abundance of the 96-kDa lysosomal glycoprotein protein (lgp96), and we found that EGF caused a significant decrease in lgp96 in cellular homogenates and associated with lysosomes. We conclude that EGF in cultured renal cells regulates the breakdown of proteins targeted for destruction by chaperone-mediated autophagy. Because suppression of this pathway results in an increase in Pax2, these results suggest a novel mechanism for the regulation of cell growth. epidermal growth factor 73-kDa heat shock cognate protein 96-kDa lysosomal glycoprotein transforming growth factor β1 glyceraldehyde-3-phosphate dehydrogenase phosphate-buffered saline A major response of cells to growth factors is a generalized increase in protein synthesis including the synthesis of specific classes of proteins (1Proud C.G. Nature. 1995; 371: 747-748Crossref Scopus (49) Google Scholar). In addition to controlling synthesis, growth factors can suppress the bulk degradation of proteins (2Ballard F.J. Knowles S.E. Wong S.S.C. Bodner J.B. Wood C.M. Gunn J.M. FEBS Lett. 1980; 114: 209-212Crossref PubMed Scopus (51) Google Scholar). For example, in renal tubular epithelial cells we found that EGF1 suppresses the breakdown of the mass of intracellular proteins (3Franch H.A. Curtis P.V. Mitch W.E. Am. J. Physiol. 1997; 273: C843-C851Crossref PubMed Google Scholar). The suppression of proteolysis in response to growth factors involved decreased lysosomal degradation rather than decreased proteasomal or calcium-sensitive proteases (3Franch H.A. Curtis P.V. Mitch W.E. Am. J. Physiol. 1997; 273: C843-C851Crossref PubMed Google Scholar). Despite reports that proteolysis is regulated, no one has determined if specific classes of proteins are being regulated by growth factors. Lysosomes degrade extracellular proteins (via endocytosis), membrane proteins, and organelles (via autophagy) and can degrade cytosolic proteins via direct import through the lysosomal membrane (4Lee H.K. Marzella L. Int. Rev. Exp. Pathol. 1994; 35: 39-147PubMed Google Scholar,5Klionsky D.J. Ohsumi Y. Annu. Rev. Cell Dev. Biol. 1999; 15: 1-32Crossref PubMed Scopus (389) Google Scholar). Dice and Terlecky (6Dice J.F. Terlecky S.R. Ciechanover A.J. Schwartz A.L. Cellular Proteolytic Systems. Wiley-Liss, New York1994: 134-159Google Scholar) showed that there is a specific import pathway involving the 73-kDa heat shock cognate protein (hsc73) called chaperone-mediated autophagy. Hsc73 binds to a penta-peptide motif (consensus sequence, KFERQ) on the target protein and, acting as a chaperone, unfolds the target protein (7Terlecky S.R. Dice J.F. J. Biol. Chem. 1993; 268: 23490-23495Abstract Full Text PDF PubMed Google Scholar). Hsc73 bound to the substrate protein then interacts with an intrinsic lysosomal membrane protein, the 96-kDa lysosomal glycoprotein (lgp96, also called lysosomal membrane protein 2a) (8Cuervo A.M. Dice J.F. Science. 1996; 273: 501-503Crossref PubMed Scopus (700) Google Scholar). After recruiting other accessory proteins, the target protein is transported through the lysosomal membrane and degraded (9Agarraberes F.A. Terlecky S.R. Dice J.F. J. Cell Biol. 1997; 137: 825-834Crossref PubMed Scopus (250) Google Scholar). Dice and co-workers (10Wing S.S. Chiang H.L. Goldberg A.L. Dice J.F. Biochem. J. 1991; 275: 165-169Crossref PubMed Scopus (115) Google Scholar) also showed that chaperone-mediated autophagy can be regulated by calorie deprivation, which accelerates the proteolysis of proteins with KFERQ motifs in the lysosomes from liver. In kidney and liver, up to 30% of proteins contains the KFERQ motif, including many of the proteins involved in glycolysis. Because most glycolytic proteins have long half-lives, an increase in degradation could function to down-regulate their abundance. Because we found that growth factors suppress lysosomal proteolysis in renal cells, we wanted to determine whether growth factors regulate the half-life of proteins that are substrates for chaperone-mediated autophagy. In pursuing this question, we uncovered a novel mechanism that leads to the accumulation of specific proteins involved in the regulation of cellular growth. All chemicals or reagents were purchased from Sigma, except Dulbecco's modified Eagle's medium, newborn calf serum, Trypsin-EDTA, and penicillin-streptomycin, which were obtained from Life Technologies, Inc. Recombinant human TGF-β1 and EGF were obtained from R&D Systems (Minneapolis, MN), andl-[U-14C]phenylalanine was obtained from PerkinElmer Life Sciences. Anti-hsc73 antiserum was purchased from Maine Biotechnology (Portland, ME), anti-M2 pyruvate kinase was purchased from Scebo-Tech, A.G. (Wettenburg, Germany), and anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was purchased from Biodesign International (Kennebunk, ME). Anti-Pax2 serum was purchased from Zymed Laboratories Inc.. Affinity-purified anti-sera to the penta-peptide KFERQ and to lgp96 were a generous gift of J. F. Dice (Tufts University). The anti-hexokinase serum was a gift of E. Knecht (Universidad de Vallencia, Spain). NRK-52E cells (a rat kidney epithelial cell line (11De Larco J.E. Todaro G.J. J. Cell. Physiol. 1978; 94: 335-342Crossref PubMed Scopus (232) Google Scholar), passage 15) were obtained from ATCC (Manassas, VA), subcultured, and grown in high glucose Dulbecco's modified Eagle's medium supplemented with 25 mm HEPES, 25 mm glutamine, and 5% calf serum. Studies were performed on cells from passages 19–29. Cells in 6-well plates were grown to confluence and rendered quiescent by serum removal 48 h prior to experimental treatment. The cell culture medium was refreshed every 24 h to maintain a constant pH; it did not differ between control and treatment groups. Recombinant human TGF-β1 was reconstituted in 4 mm HCl containing 0.1% heat-treated bovine serum albumin. Recombinant human EGF was reconstituted in PBS containing 0.1% heat-treated bovine serum albumin. In all studies, concentrations of 10−10m (TGF-β1), 10−8m (EGF), and 10 mm (NH4Cl) were used (12Franch H.A. Shay J.W. Alpern R.J. Preisig P.A. J. Cell Biol. 1995; 129: 245-254Crossref PubMed Scopus (88) Google Scholar); the appropriate vehicle was added to control cells. After exposure to an experimental variable, cells were washed with PBS, incubated with 0.05% trypsin/0.5 mm EDTA for 5 min, centrifuged at 1500 × g for 5 min, and washed with PBS. The final pellet was resuspended in 1 ml of 50 mmNa2PO4 (pH 7.4) and lysed on ice by repeated passage though a 27-gauge needle. The lysate was divided and stored at −70 °C for protein and DNA determination as described (12Franch H.A. Shay J.W. Alpern R.J. Preisig P.A. J. Cell Biol. 1995; 129: 245-254Crossref PubMed Scopus (88) Google Scholar). Protein degradation was measured as the release ofl-[U-14C]phenylalanine from cells prelabeled as described (3Franch H.A. Curtis P.V. Mitch W.E. Am. J. Physiol. 1997; 273: C843-C851Crossref PubMed Google Scholar, 4Lee H.K. Marzella L. Int. Rev. Exp. Pathol. 1994; 35: 39-147PubMed Google Scholar, 5Klionsky D.J. Ohsumi Y. Annu. Rev. Cell Dev. Biol. 1999; 15: 1-32Crossref PubMed Scopus (389) Google Scholar, 6Dice J.F. Terlecky S.R. Ciechanover A.J. Schwartz A.L. Cellular Proteolytic Systems. Wiley-Liss, New York1994: 134-159Google Scholar, 7Terlecky S.R. Dice J.F. J. Biol. Chem. 1993; 268: 23490-23495Abstract Full Text PDF PubMed Google Scholar, 8Cuervo A.M. Dice J.F. Science. 1996; 273: 501-503Crossref PubMed Scopus (700) Google Scholar, 9Agarraberes F.A. Terlecky S.R. Dice J.F. J. Cell Biol. 1997; 137: 825-834Crossref PubMed Scopus (250) Google Scholar, 10Wing S.S. Chiang H.L. Goldberg A.L. Dice J.F. Biochem. J. 1991; 275: 165-169Crossref PubMed Scopus (115) Google Scholar, 11De Larco J.E. Todaro G.J. J. Cell. Physiol. 1978; 94: 335-342Crossref PubMed Scopus (232) Google Scholar, 12Franch H.A. Shay J.W. Alpern R.J. Preisig P.A. J. Cell Biol. 1995; 129: 245-254Crossref PubMed Scopus (88) Google Scholar, 13Jurkovitz C.T. England B.K. Ebb R.G. Mitch W.E. Kidney Int. 1992; 42: 595-601Abstract Full Text PDF PubMed Scopus (40) Google Scholar). Briefly, 5 mm unlabeled phenylalanine was added to the medium to minimize reuse of the phenylalanine released by protein breakdown, and an initial 4-h washout period was used to eliminate short lived proteins and unincorporatedl-[14C]phenylalanine. Aliquots of the medium were removed at intervals and treated with trichloroacetic acid to remove protein, and the radioactivity was determined. At the end of the experiment, cell protein was solubilized in 1 ml/well of 1% SDS, and the remaining radioactivity was measured. The protein degradation rate was calculated as the slope of the logarithm of the [14C]phenylalanine remaining in cell proteinversus time. Confluent cells in 100-mm dishes were incubated with 100 μCi ofl-[35S]cystine/methionine (ICN, Costa Mesa, CA). For GAPDH, the labeling was performed in serum-free Dulbecco's modified Eagle's medium with cold cystine/methionine present for 72 h. For Pax2, cells were treated with EGF in cystine/methionine-free medium for 20 h to increase the labeling of Pax2 because its abundance is very low in quiescent cells. After two washes in serum-free medium, a 4-h washout in serum- and growth factor-free cystine/methionine-containing medium was performed prior to the addition of growth factors. Subsequently, cells were washed twice with serum-free medium before adding the experimental variable in the medium containing an excess of cold cystine and methionine. The culture medium was always changed daily with an additional wash. At time 0 and at various times up to 72 h for GAPDH and up to 24 h for Pax2, cells were lysed in a 1% Nonidet P-40 lysis buffer containing 100 μg/ml phenylmethylsulfonyl fluoride, 2 mmsodium EDTA, 4 μg/ml aprotinin, 2 μg/ml leupeptin, and 2 μg/ml pepstatin. One μg/ml anti-GAPDH or Pax2 anti-serum was added to equal amounts of cellular protein that was precipitated with protein G-Sepharose beads. After three washes with lysis buffer, the immunoprecipitate was separated by SDS-polyacrylamide gel electrophoresis, underwent autoradiography, and was quantitated by use of the Signmagel program. The protein half-life was calculated from the slope of the logarithmic transformation of the densitometry data plotted against time. We documented the completeness of recovery by performing Western blots on the supernatants after immunoprecipitation (data not shown). Cells in 60-mm tissue culture dishes were washed twice in ice-cold PBS and lysed in a buffer containing 100 μg/ml phenylmethylsulfonyl fluoride, 2 mmsodium EDTA, 2 μg/ml aprotinin, 2 μg/ml leupeptin, and 2 μg/ml pepstatin. After centrifugation, the proteins in the supernatant were determined and boiled in buffer containing 1% SDS and 0.5% β-mercaptoethanol, separated by SDS-polyacrylamide gel electrophoresis, and transferred to nitrocellulose filters, and 5% fat-free milk protein or 3% bovine serum albumin was used as a blocking reagent. Antibodies were detected using the ECL system (Amersham Pharmacia Biotech) and Kodak BCL film. Lysosomes were isolated as described by Cuervo et al. (14Cuervo A.M. Hildebrand H. Bomhard E.M. Dice J.F. Kidney Int. 1999; 55: 529-545Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar). Briefly, cells in two 25-cm2plates/group were washed in ice-cold PBS and then homogenized after scraping in ice-cold buffer (2.5 mm Tris (pH 7.2), 0.25m sucrose) by 20 strokes of a Teflon Polytron homogenizer at 4 °C. Then 1 g of protein/7 ml of 0.25 m sucrose was centrifuged at 2500 × g for 10 min and the post-nuclear supernatant was placed on a discontinuous gradient of 35 and 17% metrizamide (pH 7.0) and 6% Percoll and centrifuged at 6800 × g for 25 min. The lysosome/mitochondrial fraction at the metrizamide/Percoll interface was resuspended to a final concentration of 57% metrizamide. On top of this fraction, there was a discontinuous metrizamide gradient of equal volumes of 35, 17, and 5%, metrizamide with a final 0.25% sucrose layer. This gradient was centrifuged for 1 h at 95,000 × g. The lysosomes sediment to the interface of 5–17% metrizamide and mitochondria at the 35–57% interface. Purity of the lysosomal or mitochondrial fractions was determined by the activity of β-N-hexosaminidase (lysosome) and mitochondrial succinic dehydrogenase and by the presence of lgp96 (15Aniento F. Roche E. Cuervo A.M. Knecht E. J. Biol. Chem. 1993; 268: 10463-10470Abstract Full Text PDF PubMed Google Scholar). Results are expressed as mean ± S.E. Because there was experiment-to-experiment variation in the magnitude of responses, results are presented as a percentage of the control value determined simultaneously. The differences between two groups were analyzed by the Student's t test, but multiple comparisons were analyzed by analysis of variance. Comparisons of slopes of lines representing the release ofl-[U-14C]phenylalanine were done by analysis of co-variance. We treated NRK-52E cells with growth factors and found different growth properties (Fig. 1): EGF causes hyperplasia (increased DNA content) and increases (∼30%) the half-life of long lived proteins. TGF-β1 increases in the protein/DNA ratio (hypertrophy) but does not significantly suppress proteolysis. The combination of EGF plus TGF-β1 causes hypertrophy with the suppression of proteolysis, whereas ammonia causes less hypertrophy even though there is an even greater suppression of proteolysis. We have shown previously that EGF, TGF-β1, and EGF plus TGF-β1 increase protein synthesis, whereas ammonia does not affect synthesis (12Franch H.A. Shay J.W. Alpern R.J. Preisig P.A. J. Cell Biol. 1995; 129: 245-254Crossref PubMed Scopus (88) Google Scholar, 13Jurkovitz C.T. England B.K. Ebb R.G. Mitch W.E. Kidney Int. 1992; 42: 595-601Abstract Full Text PDF PubMed Scopus (40) Google Scholar, 16Franch H.A. Preisig P.A. Am. J. Physiol. 1996; 270: C932-C938Crossref PubMed Google Scholar). GAPDH has a KFERQ consensus sequence (17Dice J.F. Chiang H.L. Biochem. Soc. Symp. 1989; 55: 45-55PubMed Google Scholar) and is a classic substrate for chaperone-mediated autophagy (15Aniento F. Roche E. Cuervo A.M. Knecht E. J. Biol. Chem. 1993; 268: 10463-10470Abstract Full Text PDF PubMed Google Scholar,18Cuervo A.M. Knecht E. Terlecky S.R. Dice J.F. Am. J. Physiol. 1995; 269: C1200-C1208Crossref PubMed Google Scholar). As shown in Fig. 2 and Table I, the half-life of GAPDH measured in a pulse-chase experiment increased by ∼ 90% in cells treated with either EGF or ammonia but did not significantly change in cells treated with TGF-β1 alone.Table IGrowth factors change GAPDH protein half-lifeHalf-lifePercentage of Control Half-lifeh%Control38.1 ± 5.3100EGF (10−8 M)74.9 ± 3.1 1-ap < 0.05versus control, n = 4–7.196TGF (10−10 M)48.6 ± 3.8128Ammonia 10 mM73.7 ± 10.81-ap < 0.05versus control, n = 4–7.193NRK-52E cells were radiolabeled when quiescent growth factors were added during a cold-chase period, and then GAPDH was immunoprecipitated at different times. Protein half-life was calculated from the slope of the logarithmic transformation of the densitometry of autoradiograms plotted against time.1-a p < 0.05versus control, n = 4–7. Open table in a new tab NRK-52E cells were radiolabeled when quiescent growth factors were added during a cold-chase period, and then GAPDH was immunoprecipitated at different times. Protein half-life was calculated from the slope of the logarithmic transformation of the densitometry of autoradiograms plotted against time. We also tested whether these agents change the abundance of specific proteins with the KFERQ lysosomal import sequence. GAPDH abundance increased with treatments that suppress proteolysis, but TGF-β1, which does not affect proteolysis, did not increase GAPDH. We also examined the M2 isoform of pyruvate kinase because it is a glycolytic enzyme with KFERQ consensus sequences (19Noguchi T. Inoue H. Tanaka T. J. Biol. Chem. 1986; 261: 13807-13812Abstract Full Text PDF PubMed Google Scholar), it binds hsc73, and it is imported into lysosomes (6Dice J.F. Terlecky S.R. Ciechanover A.J. Schwartz A.L. Cellular Proteolytic Systems. Wiley-Liss, New York1994: 134-159Google Scholar). In contrast, hexokinase is a glycolytic enzyme that lacks a KFERQ sequence (15Aniento F. Roche E. Cuervo A.M. Knecht E. J. Biol. Chem. 1993; 268: 10463-10470Abstract Full Text PDF PubMed Google Scholar). Conditions that stimulate cell growth also increase the abundance of the M2 isoform of pyruvate kinase as well as other proteins recognized by anti-KFERQ affinity-purified sera (Fig. 3). Hexokinase abundance did not increase with any of the growth factors. Although the pattern of changes was similar between proteins recognized by anti-KFERQ serum and the M2 isoform of pyruvate kinase, the magnitude of increase in the M2 isoform of pyruvate kinase was much greater, suggesting that its synthesis is also stimulated (Fig. 3, B andC). Glycolytic enzymes are not the only proteins containing KFERQ sequences, which may be important intermediates influencing renal cell growth. For example, the renal paired box-related transcription factor, Pax2, contains a conserved KFERQ sequence at amino acids 38–42 (20Eccles M.R. Wallis J.L. Fidler A.E. Spurr N.K. Goodfellow P.J. Reeve A.E. Cell Growth Differ. 1992; 3: 279-289PubMed Google Scholar). We found that the half-life of Pax2 (Fig.4) also increased with EGF treatment and that the abundance of Pax2 was increased by growth factors; the smallest rise occurred with TGF-β1. Interestingly, the Pax2 abundance also increased when lysosomal proteolysis was inhibited with NH4Cl (Fig. 3 D). Thus different stimuli causing growth also increase Pax2 abundance. To determine the mechanisms that control suppression of proteolysis, we examined if the regulatory proteins of chaperone-mediated autophagy change in response to growth factors. Hsc73 did not change in abundance (Fig. 3 A). We also examined the lysosomal membrane receptor for protein translocation, lgp96, in lysates and in association with isolated lysosomes using sera directed against the 12-amino acid cytoplasmic portion of lgp96 that binds to hsc73 (8Cuervo A.M. Dice J.F. Science. 1996; 273: 501-503Crossref PubMed Scopus (700) Google Scholar). The quality of lysosomes isolated did not vary between the control and EGF-treated cells as assayed by hexosaminidase activity (TableII). Isolated lysosomes exhibited immunostaining for lgp96 (Fig.5 A); the level was 7-fold higher than in whole cell lysates (Fig. 5 B). Lgp96 was not detected in the mitochondrial fractions. In whole cell lysates, lgp96 abundance decreased by 30–40% after 24 or 96 h of treatment with stimuli that suppress proteolysis (Fig. 5, C andE). In contrast, TGF-β1, which did not affect proteolysis, did not affect lgp96 levels. Because the lysosomal-associated lgp96 correlates more closely with the activity of chaperone-mediated autophagy than total cellular lgp96 (21Cuervo A.M. Dice J.F. J. Cell Sci. 2000; 113: 4441-4450Crossref PubMed Google Scholar), we examined lysosomal-associated lgp96 with EGF treatment and found a 48 ± 10% decrease (p < 0.05, n = 3) compared with lysosomes isolated from control cells (Fig. 5,D and E).Table IIHexosaminidase activity in isolated lysosomesHexosaminidase activityProteinSpecific activity%%Control (n = 3)Homogenate1001001Post-nuclear supernatant28.0 ± 8.614.4 ± 1.11.95 ± 0.61Mitochondria0.40 ± 0.150.49 ± 0.260.92 ± 0.19Lysosomes2.55 ± 0.380.80 ± 0.013.19 ± 0.48 2-ap < 0.05versus homogenate.EGF (n = 3)Homogenate1001001Post-nuclear supernatant16.4 ± 4.012.3 ± 0.71.36 ± 0.41Mitochondria0.87 ± 0.420.54 ± 0.161.52 ± 0.34Lysosomes1.60 ± 0.250.58 ± 0.182.93 ± 0.502-ap < 0.05versus homogenate.NRK-52E cells were grown as in Fig. 1, and lysosomes and mitochondria were isolated by metrizamide density gradient centrifugation. There are no significant differences between the control and EGF.2-a p < 0.05versus homogenate. Open table in a new tab NRK-52E cells were grown as in Fig. 1, and lysosomes and mitochondria were isolated by metrizamide density gradient centrifugation. There are no significant differences between the control and EGF. In the early 1980s, it was recognized that specific growth factors and activated oncogenes could suppress protein degradation in certain cell types including epithelial cells (22Ballard F.J. Wong S.S.C. Knowles S.E. Partridge N.C. Martin T.J. Wood C.M. Gunn J.M. J. Cell. Physiol. 1980; 105: 335-346Crossref PubMed Scopus (47) Google Scholar). It was not known, however, which classes of proteins develop longer half-lives during growth or how this response was regulated. We found that EGF suppresses the breakdown of the bulk of proteins in NRK-52E cells by a mechanism that involved the suppression of lysosomal function but not proteolysis by proteasomal or calcium-activated proteases (3Franch H.A. Curtis P.V. Mitch W.E. Am. J. Physiol. 1997; 273: C843-C851Crossref PubMed Google Scholar). Physiologic conditions can regulate specific pathways of lysosomal proteolysis. For example, calorie deprivation increases the degradation of proteins with a KFERQ motif in liver and kidney lysosomes (10Wing S.S. Chiang H.L. Goldberg A.L. Dice J.F. Biochem. J. 1991; 275: 165-169Crossref PubMed Scopus (115) Google Scholar). How does this finding bear on growth factor-induced renal cell growth? Conditions stimulating renal cell growth increase glycolysis, and many glycolytic enzymes contain KFERQ motifs (17Dice J.F. Chiang H.L. Biochem. Soc. Symp. 1989; 55: 45-55PubMed Google Scholar, 23Tang M.J. Suresh K.E. Tannen R.L. Am. J. Physiol. 1989; 256: C532-C539Crossref PubMed Google Scholar, 24Fukuhara Y. Yamamoto S. Yano F. Orita Y. Fujiwara Y. Uneda N. Kamada T. Noguchi T. Tanaka T. Koide H. Endou H. Kurokawa K. Cellular and Molecular Biology of the Kidney. Karger, Basel, Switzerland1991: 222-228Google Scholar, 25Brink U. Eigenbrodt E. Oehmke M. Mazurek S. Fischer G. Virchows Arch. 1994; 424: 177-185PubMed Google Scholar). Thus, by acting in the opposite fashion as calorie deprivation, growth factors could suppress the degradation of glycolytic enzymes and contribute to the increase in glycolysis that accompanies renal growth. Our results confirm that EGF acts to prolong the half-life of the classic substrate for chaperone-mediated autophagy, GAPDH, and increase the abundance of KFERQ-containing proteins. Our results provide additional insights into the relationship among growth factors, cell growth, and lysosomal protein degradation. First, only specific growth factors influence lysosomal function. For example, EGF clearly stimulates cell growth and suppresses total proteolysis and the proteolysis of substrates of chaperone-mediated autophagy. In contrast, TGF-β1 caused the smallest increase in growth and has almost no effect on proteolysis. We do not conclude that suppression of proteolysis is the sole mechanism causing KFERQ-containing protein accumulation, because the accumulation of KFERQ proteins that occurred with TGF-β treatment almost certainly reflects increased synthesis (Figs. 3, B and C). Second, our results show that regulation of this lysosomal pathway by growth factors leads to prolongation of the half-life of Pax2, which has been implicated in renal cell growth in development, cyst formation, and renal cell carcinoma (20Eccles M.R. Wallis J.L. Fidler A.E. Spurr N.K. Goodfellow P.J. Reeve A.E. Cell Growth Differ. 1992; 3: 279-289PubMed Google Scholar, 26Winyard P.J. Risdon R.A. Sams V.R. Dressler G.R. Woolf A.S. J. Clin. Invest. 1996; 98: 451-459Crossref PubMed Scopus (167) Google Scholar). Because there is also an increase in the abundance of Pax2 in cells treated with EGF and because EGF causes only trivial increases in Pax2 mRNA in renal tubular cells (27Liu S. Cieslinski D.A. Funke A.J. Humes H.D. Exp. Nephrol. 1997; 5: 295-300PubMed Google Scholar), the increase in half-life we found could be physiologically relevant. Because Pax2 acts as a transcription factor, these responses suggest a new mechanism by which growth factors regulate cell growth; not only do they suppress the degradation of the bulk of cytoplasmic proteins (3Franch H.A. Curtis P.V. Mitch W.E. Am. J. Physiol. 1997; 273: C843-C851Crossref PubMed Google Scholar), but they increase the availability of at least one critical transcription factor. Finally, our results provide unexpected information about a potential mechanism by which ammonia could increase cell growth. The growth of renal cells characteristically found in response to metabolic acidosis is attributed to ammonia, which can reach concentrations as high as 5 mm in the cortex of the kidney (28Halperin M.L. Kamel K.S. Ethier J.H. Stinebaugh B.J. Jungas R.L. Seldin D.W. Giebisch G. The Kidney: Physiology and Pathophysiology. Raven Press, Ltd., New York1992: 2645-2680Google Scholar). Ammonia had been thought to act only by changing lysosomal pH and nonspecifically suppressing lysosomal proteolysis leading to the accumulation of cytosolic proteins (16Franch H.A. Preisig P.A. Am. J. Physiol. 1996; 270: C932-C938Crossref PubMed Google Scholar, 29Golchini K. Norman J. Bohman R. Kurtz I. J. Clin. Invest. 1989; 84: 1767-1779Crossref PubMed Scopus (61) Google Scholar). However, our results suggest that ammonia also acts by suppressing the degradation of specific signaling proteins such as Pax2. The up-regulation of transcription factors could allow the expression of particular proteins important for growth without an increase in global protein synthesis. Regarding the mechanism involved in changing lysosomal degradation, we and others find that the abundance of hsc73 does not change even when activity of this pathway changes (9Agarraberes F.A. Terlecky S.R. Dice J.F. J. Cell Biol. 1997; 137: 825-834Crossref PubMed Scopus (250) Google Scholar, 14Cuervo A.M. Hildebrand H. Bomhard E.M. Dice J.F. Kidney Int. 1999; 55: 529-545Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar). Curiously, hsc73 contains KFERQ sequences but is resistant to degradation within hepatic lysosomes responding to starvation (9Agarraberes F.A. Terlecky S.R. Dice J.F. J. Cell Biol. 1997; 137: 825-834Crossref PubMed Scopus (250) Google Scholar). On the other hand, we did observe a decrease in the abundance of lgp96 including a sharp decrease in the amount of lgp96 specifically associated with lysosomes (Fig. 5). This finding is consistent with the close correlation between the lgp96 associated with lysosomes and the activity of chaperone-mediated autophagy (14Cuervo A.M. Hildebrand H. Bomhard E.M. Dice J.F. Kidney Int. 1999; 55: 529-545Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar, 21Cuervo A.M. Dice J.F. J. Cell Sci. 2000; 113: 4441-4450Crossref PubMed Google Scholar, 30Cuervo A.M. Hu W. Lim B. Dice J.F. Mol. Biol. Cell. 1998; 9: 1995-2010Crossref PubMed Scopus (132) Google Scholar). Although there are similarities between the effects of EGF and ammonia on the proteolysis of KFERQ-containing proteins and lgp96 levels, there are differences in their actions on lysosomes. We found that pharmacologic agents that specifically inhibit lysosomal proteolysis (ammonia, methylamine, bafilomycin A1, or leupeptin plus the protease inhibitor, E64) convert the cellular proliferation in response to EGF into hypertrophy (31Franch H.A. J. Am. Soc. Nephrol. 2000; 11: 1631-1638PubMed Google Scholar). The change in lgp96 abundance may be a common pathway increasing growth-promoting proteins such as glycolytic enzymes and Pax2, and an additional influence of lysosomal inhibitors may account for the conversion of hyperplasia to hypertrophy. Besides the regulation of chaperone-mediated autophagy, EGF could affect the function of other pathways of lysosomal proteolysis. Autophagy may also be regulated by growth factors (32Blommaart E.F.C. Luiken J.J.F.P. Meijer A.J. Histochem. J. 1997; 29: 365-385Crossref PubMed Scopus (216) Google Scholar), leading to slower degradation of organelles and membranes. EGF acts through phosphoinositide 3-kinase as it suppresses proteolysis in renal cells, 2H. A. Franch and J. Du, unpublished observation. and phosphoinositide 3-kinase has been reported to regulate autophagy in cultured liver cells (33Petiot A. Ogier-Denis E. Blommaart E.F.C. Meijer A.J. Codogno P. J. Biol. Chem. 2000; 275: 992-998Abstract Full Text Full Text PDF PubMed Scopus (1031) Google Scholar). One practical prediction of these results is that a KFERQ sequence may be used to identify proteins that are up-regulated during renal cell growth. Besides glycolytic enzymes, there are a large number of proteins in the National Center for Biotechnology Information data base that contain conserved KFERQ sequences and are important for renal tubule cell growth. These proteins include enzymes involved in phospholipid metabolism (choline kinase (GenBankTMaccession number 139962) and phosphorylcholine transferase (34Kalamar G.B. Kay R.J. Lachance A. Aebersold R. Cornell R.B. Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 6029-6033Crossref PubMed Scopus (129) Google Scholar)), ion transporters (β subunit of the sodium/potassium ATPase (35Mercer R.W. Schneider J.W. Savitz A. Emmanuel J. Benz E.J. Levenson R. Mol. Cell. Biol. 1986; 6: 3884-3890Crossref PubMed Scopus (88) Google Scholar)), and signaling molecules such as Pax2 (20Eccles M.R. Wallis J.L. Fidler A.E. Spurr N.K. Goodfellow P.J. Reeve A.E. Cell Growth Differ. 1992; 3: 279-289PubMed Google Scholar). The KFERQ sequence is present in the Pax isoforms expressed in the urinary tract (Pax2, -5, and -8) but not in other Pax isoforms, suggesting that the link between Pax proteins and this proteolytic pathway may be specific to the urinary tract (36Poleev A. Frickenscher H. Mundlos S. Winterpacht A. Zabel B. Fidler A.E. Gruss P. Pleiman C.M. Development. 1993; 116: 611-623Google Scholar, 37Adams B. Doerfler P. Aguzzi A. Kozmik Z. Urbanek P. Maurer-Fogy I. Busslinger M. Genes Dev. 1992; 6: 1589-1607Crossref PubMed Scopus (467) Google Scholar). Finally, the signaling proteins MARKS and IκB have been shown to have their abundance regulated by this pathway (30Cuervo A.M. Hu W. Lim B. Dice J.F. Mol. Biol. Cell. 1998; 9: 1995-2010Crossref PubMed Scopus (132) Google Scholar,38Spizz G. Blackshear P.J. J. Biol. Chem. 1997; 272: 23833-23842Abstract Full Text Full Text PDF PubMed Scopus (36) Google Scholar). We thank Drs. J. Fred Dice and Ana Maria Cuervo for help with reagents and techniques and Drs. William Mitch and Russ Price for advice and critical reading of the manuscript.