If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
* This work was supported in part by a grant-in-aid for Scientific Research (B) from the Japan Society for the Promotion of Science, the Uehara Memorial Foundation, and the Danone Institute of Japan (to M. K.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Autophagy, a major bulk proteolytic pathway, contributes to intracellular protein turnover, together with protein synthesis. Both are subject to dynamic control by amino acids and insulin. The mechanisms of signaling and cross-talk of their physiological anabolic effects remain elusive. Recent studies established that amino acids and insulin induce p70 S6 kinase (p70S6k) phosphorylation by mTOR, involved in translational control of protein synthesis. Here, the signaling mechanisms of amino acids and insulin in macroautophagy in relation to mTOR were investigated. In isolated rat hepatocytes, both regulatory amino acids (RegAA) and insulin coordinately activated p70S6k phosphorylation, which was completely blocked by rapamycin, an mTOR inhibitor. However, rapamycin blocked proteolytic suppression by insulin, but did not block inhibition by RegAA. These contrasting results suggest that insulin controls autophagy through the mTOR pathway, but amino acids do not. Furthermore, micropermeabilization with Saccharomyces aureus α-toxin completely deprived hepatocytes of proteolytic responsiveness to RegAA and insulin, but still maintained p70S6k phosphorylation by RegAA. In contrast, Leu8-MAP, a non-transportable leucine analogue, did not mimic the effect of leucine on p70S6k phosphorylation, but maintained the activity on proteolysis. Finally, BCH, a System L-specific amino acid, did not affect proteolytic suppression or mTOR activation by leucine. All the results indicate that mTOR is not common to the signaling mechanisms of amino acids and insulin in autophagy, and that the amino acid signaling starts extracellularly with their “receptor(s),” probably other than transporters, and is mediated through a novel route distinct from the mTOR pathway employed by insulin.
Autophagy is a major mechanism for intracellular bulk degradation, and contributes to protein turnover and cell growth, together with protein synthesis (
). Macroautophagy, a dominant form in mammalian cells that is subject to nutritional and hormonal controls, consists of multiple steps starting with the formation of an autophagosome, followed by its acidification and fusion with the lysosome to become an autolysosome, and finally complete digestion of the sequestered organelles and macromolecules. Recent advances using molecular genetics have revealed that a variety of protein complexes, e.g. Apg12-Apg5, Apg1-Apg13 conjugation, and LC3 modification systems, participate in the initial autophagosome formation step in quite a complicated manner (
). Concerning proteolysis, among a number of proteolytic systems in the body, autophagic proteolysis is the only mechanism known to be subject to nutritional regulation. Because amino acids, together with hormones such as insulin and glucagon, are the principal physiological regulators of mammalian autophagy, the mechanism of amino acid signaling and its relationship with hormones are major targets to be elucidated. The hypothesis that amino acids have their own signal transduction mechanisms originated from the following evidence. In perfused liver, the extracellular concentrations of phenylalanine and leucine individually controlled autophagic proteolysis (
). These findings represented indirect evidence, but strongly suggested the possibility that, for autophagic regulation, amino acids have their own recognition sites at the cell surface and must have a signal transduction mechanism to the site of autophagosome formation inside the cell. A possible candidate for the leucine “receptor/sensor” protein has been reported (
) proposed the hypothesis that amino acids can control autophagy by inhibiting extracellular signal-regulated kinase 1/2-dependent GAIP phosphorylation. Another possibility is that the eukaryotic initiation factor 2α kinase signaling pathway is involved in amino acid starvation-induced autophagy in murine embryonic fibroblasts (
kinase is the central component of nutrient- and hormone-sensitive signaling pathways that regulate cell growth, including initiation of mRNA translation, transcription of enzymes for metabolic pathways, amino acid transport, and autophagy (
) pioneered their hypothesis that the inhibition of autophagy by amino acids is mediated by a signal transduction pathway via phosphorylation of ribosomal protein S6 in isolated hepatocytes. S6 phosphorylation is activated by p70 S6 kinase (p70S6k), which is a direct target of mTOR (
). However, among these inconsistent findings, the kinds of physiological effectors employed to regulate autophagy have not been considered seriously.
Therefore, in the present study, we started by confirming the effect of rapamycin on the amino acid regulation of autophagy. Then, by comparing the effectors, amino acids and insulin, to simulate the above conditions, we discovered that these effectors have distinct signaling pathways for controlling autophagic proteolysis in hepatocytes. The results clearly solve apparent discrepancies among the data reported above: insulin, and probably serum, regulate autophagy through an mTOR-dependent pathway, whereas amino acids exert their effect through an mTOR-independent pathway. Thus, in autophagic regulation, unlike protein synthesis, nutritional and hormonal signals do not converge at mTOR, which makes the story more complicated. In addition, we examined the effects of micropermeabilization by α-toxin, Leu8-MAP, and BCH, to further characterize how amino acid signals regulate autophagy in hepatocytes. All the results demonstrated that the signaling of amino acids in autophagic proteolysis starts with their “receptor(s),” probably other than transporters, at the plasma membrane, and has a novel route clearly distinct from the mTOR signaling pathway employed by insulin.
Materials—Insulin and glucagon were obtained from Sigma; rapamycin was from Wako Chemicals; l-[U-14C]valine (283 mCi/mmol) and l-[U-14C]leucine (292 mCi/mmol) were from American Radiolabeled Chemicals. The rabbit anti-rat p70 S6 kinase antibody was purchased from Santa Cruz Biotechnology Inc. The ECL Western blotting detection kit, horseradish peroxidase-conjugated donkey anti-rabbit IgG, and cAMP enzyme immunoassay system were purchased from Amersham Biosciences. All other chemicals were of analytical grade.
Preparation of Isolated Hepatocytes—Male Wistar/ST rats weighing about 250 g, meal-fed between 14:00 and 18:00 with a 35% casein diet, were used as hepatocyte donors. Liver parenchymal cells were isolated by the collagenase method of Seglen (
). Freshly isolated hepatocytes were suspended at a density of 2 × 106 cells/ml in Krebs-Ringer bicarbonate buffer (pH 7.4) containing 6 mm glucose and 0.5% bovine serum albumin oxygenated with O2:CO2 (95:5, v/v) gas. The cells were incubated in 3-ml suspensions in 10-ml flasks at 37 °C. The amino acids employed were those that have been shown to exert direct effects on autophagic proteolysis, called regulatory amino acids, and were added as multiples of the normal plasma concentrations as shown in the figure legends. The normal concentrations (1×) of the regulatory amino acids including coregulatory Ala (RegAA) were as follows (μm): Leu, 204; Tyr, 98; Pro, 437; Met, 60; His, 92; Trp, 93; Ala, 475 (
). After fresh hepatocytes were resuspended at a density of 2.5 × 106 cells/ml in HEPES-cytosol buffer (20 mm HEPES, 20 mm NaCl, 100 mm KCl, 5 mm MgSO4, 0.96 mm NaH2PO4, 0.49 mm CaCl2, 1 mm EGTA, and 0.5% bovine serum albumin, pH 7.3) oxygenated with O2 gas, they were treated with the toxin (20.3 rabbit hemolytic unit/ml) for 5 min at 37 °C. Then, the treated cells were centrifuged into a pellet and resuspended in HEPES-cytosol buffer to remove the excess toxin. To supply an energy source in the α-toxin-treated cells, an ATP regenerating system (1.5 mm ATP, 5 mm phosphocreatine, 5 units/ml creatine phosphokinase) was added (
), autophagic proteolysis was measured by Val release from hepatocytes in the presence of cycloheximide (20 μm). After incubation of hepatocytes with various additions for 30 min, cycloheximide was added. Samples for Val analysis were taken at 37 and 47 min. The proteolytic rate was calculated from the difference for this 10-min interval, which was designated “the short exposure method” (see Fig. 1). Samples were deproteinized in ice-cold perchloric acid (PCA, 6% final concentration). The supernatants were stored at -20 °C until analysis. Val was derivatized with dansyl chloride, using l-norvaline as an internal standard, and separated by reverse phase-high performance liquid chromatography using a Supelcosil LC-18 column (4.6 × 150 mm) as described previously (
Measurement of Protein Synthesis—The rate of protein synthesis was determined by the amount of [14C]Val incorporation into protein. After starting the hepatocyte incubation, [14C]Val and 5 mm unlabeled Val (specific activity: 0.02 μCi/μmol) were added at 60 min. Samples were taken at 90 min, and precipitated in an equal volume of 20% trichloroacetic acid. Protein was isolated from the precipitates as described previously, and the radioactivity was counted (
Immunoblotting of p70 S6 kinase—As an indicator of mTOR kinase activity, the rapamycin-sensitive phosphorylation of p70S6k was measured. Fresh hepatocytes were incubated at 37 °C. Five-ml samples were taken at 30 min, and added to 10 ml of ice-cold homogenization buffer (20 mm HEPES, 100 mm KCl, 2 mm EGTA, 50 mm sodium fluoride, 0.2 mm EDTA, 50 mm β-glycerophosphate, 0.1 mm dithiothreitol, 0.1 mm phenylmethanesulfonyl fluoride, 1 mm benzamidine, 0.5 mm sodium vanadate, pH 7.4). The cells were centrifuged into a pellet, resuspended in 4 ml of ice-cold homogenization buffer, and then homogenized with a tight-fitting Dounce homogenizer (100 strokes, Wheaton Science Products) on ice. The homogenates were centrifuged at 10,000 × g for 10 min at 4 °C. The supernatants were boiled for 5 min in 10 mm Tris-HCl (pH 6.8), 12.5% glycerol, 1.25% SDS, and 1.25% β-mercaptoethanol. The p70S6k in the supernatants was separated by SDS-PAGE according to Laemmli (
), and then transferred to a polyvinylidene difluoride membrane. After blocking with 5% skim milk in phosphate-buffered saline/Tween 20, the membrane was incubated with the rabbit anti-rat p70S6k antibody for 1 h at room temperature, followed by incubation with the horseradish peroxidase-conjugated donkey anti-rabbit IgG. An ECL Western blot detection kit was used as the substrate for detection of the horseradish peroxidase-conjugated secondary antibody and the membrane was exposed to Hyperfilm ECL (Amersham Biosciences). In the case of the effect of individual amino acids, because of their small changes, the electrophoretic forms with different mobilities, designated as α-, β-, γ-, and δ-form, were densitometrically quantified, expressed as the percentage of total p70S6k in hyperphosphorylated forms (β + γ + δ/total), and statistically estimated.
Leu Transport Assay—Leu transport activity was measured using the uptake of [14C]Leu into the cells. Hepatocytes were exposed to 4× Leu containing [14C]Leu (specific activity: 94 μCi/mmol). After [14C]Leu addition, cell suspensions were taken at 0.5, 1, 2, 5, and 10 min, layered on top of a separation medium in centrifuge tubes, and then quickly centrifuged at 10,000 × g for 1 min. The separation medium was prepared by stacking a hydrocarbon oil mixture (1-bromodecane, n-dodecane = 10:1, d = 1.037) on 10% PCA (
). Intact cells were harvested in the PCA layer after this centrifugation. Leu uptake into hepatocytes was measured by the radioactivity in the PCA layer and the values were expressed as nanomole of Leu/106 cells.
Analytical Procedures—Protein was measured by the Bradford method (
) using bovine IgG as the standard. Radioactivity was determined with Aloka liquid scintillation analyzer LSC-1000. All the data were expressed as mean ± S.E. Statistical significance was evaluated by Student's t test.
Lack of Effect of Rapamycin on Proteolytic Suppression by Amino Acids—Because Blommaart et al. (
) reported that ribosomal protein S6 phosphorylation via mTOR was partially involved in the control of autophagy by amino acids in isolated hepatocytes, we started by verifying this possibility. First, the effect of rapamycin, a specific inhibitor of mTOR, on amino acid-sensitive proteolysis was tested using our regular short exposure method (Fig. 1). The physiologically maximal level of RegAA (4 times the normal plasma levels, 4×) suppressed proteolysis to about 55%. Rapamycin (100 nm) neither increased proteolysis significantly by itself, nor affected that suppressed by RegAA. No effects were seen even at the much higher dose of 2.5 μm. These unexpected results were inconsistent with those of Blommaart et al. (
). Because there were several differences in methodology, we repeated their method, i.e. Val release in the presence of cycloheximide throughout a 90-min incubation (a long exposure method) and other additions. As shown in Fig. 2, the Val release suppressed by RegAA was partially restored by rapamycin, which was essentially the same result as theirs. Several possibilities for this discrepancy were considered, e.g. the time course, combination of reagents, etc. Then, because rapamycin was not effective at 30 min, it was tested whether a longer exposure was required for its effectiveness in our study. Hepatocytes were incubated with rapamycin and RegAA for 60 or 90 min and then our short exposure method was employed for proteolytic measurement. However, rapamycin was not effective (data not shown). Next, the possibility of disturbance by cycloheximide was suspected, because it was reported that long exposure to cycloheximide not only inhibits protein synthesis but also perturbs proteolysis (
). Therefore, to investigate this possibility, an indirect method without cycloheximide was employed. As shown in Table I, protein synthesis and net Val release were determined simultaneously, and then proteolysis was calculated by the sum of synthesis and net release, a method that has been validated previously (
). The results were exactly consistent with those in Fig. 1. Therefore, it was proved that the long exposure to cycloheximide was the cause of the discrepancy, and was not appropriate for the measurement of proteolysis.
Table IProteolysis calculated by protein synthesis and net balance in hepatocytes
Effect of Rapamycin on the Control of p70S6k Phosphorylation by Amino Acids—The lack of effect of rapamycin raised doubts as to the usefulness of this reagent in our study. Next, we compared the effect of rapamycin on autophagic proteolysis and mTOR activity in relation to amino acid control. The mTOR activity was estimated by the phosphorylation state of p70S6k, a direct target of mTOR kinase (
). Phosphorylation of p70S6k on several sites leads to at least 4 bands with decreased mobility in SDS-PAGE gels, as indicated by an upward shift in the banding pattern (Fig. 3B). RegAA treatment caused an upshift of the p70S6k bands, which indicated stimulation of mTOR kinase activity. Leu, Tyr, and Gln, which are all regulatory amino acids, exhibited significant 22, 32, and 26% suppressions of proteolysis, respectively (Fig. 3A). However, the phosphorylation of p70S6k by Leu, Tyr, or Gln was increased by 18 ± 6(p < 0.05), 9 ± 10, 0 ± 4% compared with the control, respectively, and only Leu had a significant effect on the phosphorylation (Fig. 3B). Rapamycin completely blocked the phosphorylation of p70S6k induced by RegAA, but had no effect on the proteolytic effect of RegAA (Fig. 3A). Nonregulatory amino acids, consisting of 12 amino acids without proteolytic effects, had no effect on proteolysis or p70S6k phosphorylation. These results clearly demonstrate that the mTOR-mediated pathway is not involved in the control of autophagic proteolysis by amino acids, and the effects of amino acids on these two processes seem to be coincidental rather than a causal relationship.
p70S6k Phosphorylation and Proteolytic Suppression by Insulin—Because it has been established that the mTOR signaling pathway is activated by growth factors (
), the contribution of this pathway to autophagic regulation by insulin was then tested. As shown in Fig. 4B, insulin mildly induced phosphorylation of p70S6k, and this was blocked by rapamycin. At the same time, the suppression of autophagic proteolysis by insulin was completely blocked by rapamycin, and the proteolytic rate was restored to the control level (Fig. 4A). Moreover, when hepatocytes were incubated with insulin and RegAA together, proteolysis was further suppressed and p70S6k phosphorylation was stimulated additively. This combined effect of insulin and RegAA on the phosphorylation was completely blocked by rapamycin, but the effect on proteolysis was only partially blocked by the reagent (Fig. 4B). These results strongly indicate that insulin uses the mTOR-dependent signaling pathway for autophagic regulation, whereas amino acids use another unknown pathway.
Effects of RegAA and Insulin on mTOR Activity in α-Toxin-treated Hepatocytes—While investigating the signaling mechanism of amino acids in autophagic regulation, we observed that micropermeabilization of hepatocytes by Staphylococcal α-toxin completely eliminated their sensitivity to amino acids (
). Because the α-toxin treatment was demonstrated to maintain the autophagic activity of hepatocytes, these semi-intact cells make it possible to examine the responses of proteolysis and mTOR signaling to amino acids and insulin. After investigating various sources of the toxin, a recombinant α-toxin showed quite stable and reproducible activity. First, the micropermeabilization was evaluated by the release of small molecules from the cell. After glucagon addition, cAMP was rapidly produced and accumulated within intact hepatocytes (Fig. 5A, left panel). The α-toxin treatment of hepatocytes led to a rapid release of cAMP into the medium, although the cells still produced cAMP in response to glucagon (Fig. 5A, right panel). Next, as illustrated in Fig. 5B, α-toxin-treated cells maintained autophagic/lysosomal proteolysis, as demonstrated by the retention of chloroquine-inhibitable proteolysis, but the sensitivities to amino acids and insulin were completely lost. Thus, it was thought that the loss of these physiological controls was because of disruption of the signaling pathways of amino acids and insulin. To determine whether the mTOR signaling pathway was involved in these alterations of the responses, p70S6k phosphorylation in the α-toxin-treated cells was measured. The induction of p70S6k phosphorylation by RegAA was maintained in the α-toxin-treated cells (Fig. 5C), although at a lower level than in intact cells (Fig. 3). On the other hand, the phosphorylation by insulin was completely lost. In addition, insulin did not synergize p70S6k in the presence of RegAA (compare with Fig. 4). Thus, the amino acid-induced mTOR signaling pathway was maintained in α-toxin-treated cells, despite the loss of proteolytic control. On the other hand, insulin-induced signaling was completely disrupted. These results support our hypothesis that mTOR is independent of the autophagy signaling pathway by amino acids, but is involved in the signaling pathway by insulin.
Effects of a Nontransportable Leucine Analogue and BCH—Leu8-MAP is formed as a branched isopeptide with eight residues of Leu attached to its N-terminal and has a molecular weight of about 1,900 (
). Fig. 6A shows that the inhibitory effect of Leu8-MAP on proteolysis was equal to that of Leu. In contrast, neither Ile nor Ile8-MAP inhibited proteolysis. These MAP derivatives have been demonstrated not to be transported into cells or degraded to produce free amino acids during a short incubation such as the 45 min in the present study (
Therefore, it is conceivable that the effect of Leu8-MAP is sensed by a Leu recognition site at the plasma membrane, and mediated by an intracellular signaling pathway. To detect the effect of a single amino acid effectively, we next examined a time course of the stimulation of p70S6k phosphorylation. After addition of Leu, phosphorylation was quickly induced at 10 min and then attenuated at 20 min (Fig. 6B). When hepatocytes were incubated with Leu8-MAP, Ile, or Ile8-MAP for 10 min, phosphorylation was not stimulated by these reagents at all (Fig. 6C), which implied that Leu8-MAP did not mimic the effect of Leu on p70S6k phosphorylation. In other words, it was concluded that the proteolytic effect of Leu, mimicked by Leu8-MAP, does not require the mTOR signaling pathway.
The above results raised the question of whether an amino acid transport step at the plasma membrane was involved in control of autophagic proteolysis, because the only Leu-specific binding protein in mammalian cells known to date is the System L transporter. Thus, to investigate the possibility of System L involvement, an experiment using BCH, a competitive inhibitor of the System L transporter in rat liver (
), was carried out. In hepatocytes, 40× (10-fold) BCH against 4× Leu inhibited about 60% of Leu transport (Fig. 7A). As shown in Fig. 7B, BCH itself did not exhibit any proteolytic effects up to the 40× level. Even when the cells were incubated with 4× Leu in the presence of equal (4×) or 10-fold (40×) BCH, the inhibitory effect of Leu was not blocked by BCH at all. Next, we tested whether BCH affected the Leu-induced mTOR signaling pathway. BCH itself tended to slightly increase p70S6k phosphorylation. However, it had no effect on the phosphorylation by Leu (Fig. 7C). These results imply that the effects of Leu on autophagy and the mTOR signaling pathway do not occur in relation to BCH or the System L transporter. Thus, it is highly conceivable that Leu controls autophagic proteolysis via an unknown Leu-recognition site(s) in the plasma membrane, other than the System L transporter.
In the present study, we found that rapamycin had no effect on autophagic proteolysis suppressed by amino acids in isolated hepatocytes. However, the results of Blommaart et al. (
) that, when hepatocytes were incubated with cycloheximide during 90 min, proteolytic rates were partially improved by rapamycin were also confirmed. The inconsistency between these results seemed to be because of the method of proteolysis. It has been reported that long exposure to cycloheximide disturbs the proteolytic rate in perfused rat liver (
). Thus, because all the evidence indicated that cycloheximide may have disturbed these measurements, we examined the effect of rapamycin in the absence of cycloheximide. As demonstrated in Table I, this was proved to be the case, and it is necessary to take care with the use of cycloheximide.
To further pursue the possibility of a contribution by the mTOR pathway to amino acid signaling on autophagic proteolysis, we tried other approaches using different techniques. Micropermeabilization with α-toxin completely abolished the proteolytic responsiveness to amino acids, although p70S6k was still phosphorylated (Fig. 5). The fact that easier access of amino acids to intracellular sites because of permeabilization was neither beneficial nor harmful strongly suggests that, for autophagic regulation by amino acids, the maintenance of plasma membrane integrity is more important compared with their role in mTOR activation. These results support our hypothesis that the signaling route of amino acids for autophagic suppression is independent of the mTOR pathway. The possibility that some steps downstream of mTOR were impaired by α-toxin treatment cannot be ruled out, but it seems negligible considering the above results with rapamycin.
On the other hand, autophagic suppression and induction of p70S6k phosphorylation by insulin were completely blocked by rapamycin in intact hepatocytes. Furthermore, when hepatocytes were exposed to insulin and RegAA together, their combined effect was only partially blocked by rapamycin. Therefore, these results demonstrate that the mTOR signaling pathway in macroautophagy is sensitive to insulin, but insensitive to amino acids. In contrast, in the control of mRNA translation, this pathway was stimulated by both amino acids and insulin (
). Therefore, we concluded with our observation that both phosphorylation of p70S6k and suppression of proteolysis were regulated by amino acids was coincidental rather than a causal relationship. In the route of insulin signaling through the mTOR pathway in autophagic proteolysis, downstream targets of mTOR are of interest, but still unknown, because there is no evidence that p70S6k itself controls autophagy. Indeed, Tap42, which is a downstream target of TOR signaling and responsible for protein translation control in yeast, was tested, and turned out not to be involved in the regulation by TOR of the autophagic machinery (
). Thus, from the above evidence, it may also be reasonable to consider the control by mTOR separately from p70S6k in the mammalian autophagy. The recognition in this way is a critical point not to confuse the interpretation of the data.
When amino acids were tested individually, leucine, tyrosine, and glutamine had the same extent of suppressive activity on autophagy (Fig. 3). However, they exhibited a slight and various degree of increases in p70S6k phosphorylation. In a variety of cultured cells, investigators have reported that leucine, in particular, enhanced the phosphorylation of p70S6k and/or 4E-BP1, in parallel downstream of mTOR, e.g. in FAO hepatocytes (
). Glutamine has not been observed to stimulate the mTOR signaling pathway. Thus, the inconsistencies among the group of effective amino acids between autophagic control and p70S6k phosphorylation may be more evidence against the hypothesis that amino acids affect autophagy through the mTOR pathway.
In addition, a more specific result was obtained in that Leu8-MAP inhibited proteolysis without stimulating p70S6k by mTOR, the latter of which was supported by a study of 4E-BP1 in adipocytes (
), clearly indicating that the mTOR signaling pathway is not involved in proteolytic inhibition by leucine. On the contrary, it can be presumed that p70S6k phosphorylation by leucine may depend on the increase in its intracellular level. Furthermore, the stimulation of p70S6k phosphorylation by amino acids was conserved in α-toxin-treated hepatocytes, supporting the same conclusion. In accordance with this, the results by Christie et al. (
) are quite interesting, namely that, in Xenopus laevis oocytes, rises in intracellular amino acid concentrations, by microinjection of amino acids and overexpression of the System L transporter, induce p70S6k phosphorylation. However, when we examined this using BCH, inhibition of the leucine transport by excess BCH did not affect p70S6k phosphorylation. Our data did not demonstrate the possible involvement of System L in the leucine-stimulated mTOR pathway. A recent study has shown that BCH had no effect on the phosphorylation of 4E-BP1 by leucine in adipocytes (
). We investigated the inhibitory activity of BCH on leucine transport into rat hepatocytes, and its effect was about 50% inhibition. Because BCH inhibits leucine transport by System L competitively, the concentration ratio of BCH to leucine is critical. Although we examined this using a 10-fold concentration of BCH to leucine, we should investigate the effect of much higher concentration ratios of BCH.
) proposed that a leucine recognition site(s) at/near the plasma membrane may generate a signal(s) in response to extracellular leucine in hepatocytes. Our data confirmed that Leu8-MAP has a leucine-mimicking effect on autophagic proteolysis. Moreover, we investigated the possible involvement of System L as a possible recognition site of leucine, but this possibility was not feasible. Although a putative leucine-binding protein(s) has not yet been identified, a candidate was detected using a photoreactive derivative of Leu8-MAP (
). Our findings and those of others strongly support the hypothesis that macroautophagy responds to the extracellular level of amino acids, whereas mRNA translation responds to their intracellular level. Employing a discussion by Lynch (
), there may be at least two leucine recognition sites in hepatocytes, an extracellular site for autophagy regulation and an intracellular site for mTOR regulation. The results with α-toxin-permeabilized hepatocytes raise many possibilities for the amino acid signaling pathway in macroautophagy. One straightforward hypothesis is the possible existence of a small molecule(s) that mediates an amino acid effect to the intracellular autophagic machinery, but easily leaks out through the α-toxin pores. Indeed, we obtained preliminary results that a soluble fraction from liver stimulated by RegAA specifically suppressed proteolysis in α-toxin-treated cells.
R. Akaishi, T. Kanazawa, and M. Kadowaki, manuscript in preparation.
Of course, the possibility that α-toxin treatment impairs an as yet unknown signaling pathway(s) of amino acids cannot be ruled out.
Another hypothesis for the regulatory mechanism of amino acids on autophagic proteolysis has been proposed. The cell hydration hypothesis is applicable to amino acids, which are transported by sodium-dependent transporters, e.g. glutamine. It was reported that autophagic inhibition by glutamine occurs by cell swelling via p38MAPK, a stress-responsive MAP kinase (
). This is worth verification, although it may not be applicable to amino acids transported in a sodium-independent manner, such as leucine. Whether various amino acids have a common signaling mechanism or different ones remains to be clarified. Another possible route proposed was that the extracellular signal-regulated kinase 1/2 kinase pathway is involved in the amino acid signaling pathway in human colon cancer HT-29 cells (
). In isolated hepatocytes, the intracellular signaling pathway is still unclear. However, LC3, an autophagosomal membrane protein, is a strong candidate for a downstream target of amino acid signaling.
T. Kanazawa and M. Kadowaki, manuscript in preparation.
The hypothesis that protein synthesis and degradation are coordinately controlled by a common signaling pathway is quite attractive. Indeed, in the case of insulin, the mTOR pathway shares its mechanism with both processes. It is tempting to extend this hypothesis to nutrient regulation. However, in the case of amino acids, the results of the present study suggest that the mechanisms are much more complicated. The effective amino acids differ between both processes, except for leucine. In proteolysis, the effect of each amino acid is demonstrable by adding singly, whereas, in protein synthesis, the presence of all the other amino acids are necessary to exhibit the effectiveness of a specific amino acid, e.g. leucine. Furthermore, the signaling mechanism of amino acids seems to be different, as demonstrated in the present study. Recently, even in protein synthesis, the possibility has arisen that amino acids induce translation of TOP mRNA without requiring activation of p70S6k and S6 phosphorylation (
). Altogether, the hypothetical signaling pathways of amino acids and insulin for autophagic proteolysis are illustrated in Fig. 8.
In conclusion, it has been regarded that the signaling mechanism by amino acids in macroautophagy is mediated via the mTOR pathway, which is also involved in translational control. However, our data demonstrate that the mechanism of amino acid signaling in macroautophagy is distinct from that for mRNA translation and mediated by a novel rapamycin-insensitive pathway. On the other hand, the signaling control by insulin is mediated by the mTOR pathway as its translational control.
We are grateful to Dr. Glenn E. Mortimore for encouragement and valuable discussions throughout this study. We also thank Dr. Giovanni Miotto (University of Padova, Italy) for providing the Leu8-MAP and Ile8-MAP, Dr. Eric A. Lehoux (Oklahoma State University, OK) for helpful discussions regarding α-toxin permeabilization, and Dr. Yoshiyuki Kamio (Tohoku University, Japan) for providing the recombinant α-toxin preparation.
Lysosomal Pathways of Protein Degradation. EUREKAH. COM/Landes Bioscience,