Distinct Sensor Pathways in the Hierarchical Control of SNAT2, a Putative Amino Acid Transceptor, by Amino Acid Availability
2007; Elsevier BV; Volume: 282; Issue: 27 Linguagem: Inglês
10.1074/jbc.m611520200
ISSN1083-351X
AutoresRussell Hyde, Emma Cwiklinski, Katrina MacAulay, Peter M. Taylor, Harinder S. Hundal,
Tópico(s)Protein Kinase Regulation and GTPase Signaling
ResumoMammalian nutrient sensors are novel targets for therapeutic intervention in disease states such as insulin resistance and muscle wasting; however, the proteins responsible for this important task are largely uncharacterized. To address this issue we have dissected an amino acid (AA) sensor/effector regulon that controls the expression of the System A amino acid transporter SNAT2 in mammalian cells, a paradigm nutrient-responsive process, and found evidence for the convergence of at least two sensor/effector pathways. During AA withdrawal, JNK is activated and induces the expression of SNAT2 in L6 myotubes by stimulating an intronic nutrient-sensitive domain. A sensor for large neutral AA (e.g. Tyr, Gln) inhibits JNK activation and SNAT2 up-regulation. Additionally, shRNA and transporter chimeras demonstrate that SNAT2 provides a repressive signal for gene transcription during AA sufficiency, thus echoing AA sensing by transceptor (transporter-receptor) orthologues in yeast (Gap1/Ssy1) and Drosophila (PATH). Furthermore, the SNAT2 protein is stabilized during AA withdrawal. Mammalian nutrient sensors are novel targets for therapeutic intervention in disease states such as insulin resistance and muscle wasting; however, the proteins responsible for this important task are largely uncharacterized. To address this issue we have dissected an amino acid (AA) sensor/effector regulon that controls the expression of the System A amino acid transporter SNAT2 in mammalian cells, a paradigm nutrient-responsive process, and found evidence for the convergence of at least two sensor/effector pathways. During AA withdrawal, JNK is activated and induces the expression of SNAT2 in L6 myotubes by stimulating an intronic nutrient-sensitive domain. A sensor for large neutral AA (e.g. Tyr, Gln) inhibits JNK activation and SNAT2 up-regulation. Additionally, shRNA and transporter chimeras demonstrate that SNAT2 provides a repressive signal for gene transcription during AA sufficiency, thus echoing AA sensing by transceptor (transporter-receptor) orthologues in yeast (Gap1/Ssy1) and Drosophila (PATH). Furthermore, the SNAT2 protein is stabilized during AA withdrawal. Amino acids (AAs) 2The abbreviations used are: AA, amino acid; AS, asparagine synthetase; CaR, extracellular Ca2+ receptor; CMV, cytomegalovirus (promoter); ERK, extracellular signal-regulated kinase; IRES, internal ribosome entry segment; JNK, Jun N-terminal kinase; MAPK, mitogen-activated protein kinase; Me-AIB, α-methylaminoisobutyric acid; PI3K, phosphinositide-3-kinase; SNAT2, sodium-dependent neutral amino acid transporter 2; S6K, ribosomal protein S6 kinase; UTR, untranslated region; PBS, phosphate-buffered saline. 2The abbreviations used are: AA, amino acid; AS, asparagine synthetase; CaR, extracellular Ca2+ receptor; CMV, cytomegalovirus (promoter); ERK, extracellular signal-regulated kinase; IRES, internal ribosome entry segment; JNK, Jun N-terminal kinase; MAPK, mitogen-activated protein kinase; Me-AIB, α-methylaminoisobutyric acid; PI3K, phosphinositide-3-kinase; SNAT2, sodium-dependent neutral amino acid transporter 2; S6K, ribosomal protein S6 kinase; UTR, untranslated region; PBS, phosphate-buffered saline. act through the coordination of several signaling pathways to modulate distinct, albeit inter-related, processes (1Hyde R. Taylor P.M. Hundal H.S. Biochem. J. 2003; 373: 1-18Crossref PubMed Scopus (282) Google Scholar, 2Kilberg M.S. Pan Y.X. Chen H. Leung-Pineda V. Annu. Rev. Nutr. 2005; 25: 59-85Crossref PubMed Scopus (217) Google Scholar, 3Holsbeeks I. Lagatie O. Van Nuland A. d. Van V. Thevelein J.M. Trends Biochem. Sci. 2004; 29: 556-564Abstract Full Text Full Text PDF PubMed Scopus (149) Google Scholar), including protein synthesis and turnover, hormone action and release, and the synthesis, transport, and metabolism of AAs. The nutrient sensors that influence these processes are poorly defined in higher eukaryotes, so, in an attempt to clarify AA sensing in mammalian cells, the molecular regulation of a paradigm AA-sensitive process (the System A amino acid transporter) has been investigated. Sodium-dependent neutral amino acid transporter 2 (SNAT2, encoded by the gene SLC38A2) exhibits functional and regulatory properties of the classically defined System A transporter (4Mackenzie B. Erickson J.D. Pflugers Arch. 2004; 447: 784-795Crossref PubMed Scopus (404) Google Scholar). Hence, stimulation of System A by AA deprivation (aka adaptive regulation) or insulin occurs concurrent with increased SNAT2 expression (5Hyde R. Christie G.R. Litherland G.J. Hajduch E. Taylor P.M. Hundal H.S. Biochem. J. 2001; 355: 563-568Crossref PubMed Scopus (72) Google Scholar) or the plasma membrane recruitment of SNAT2 (6Hyde R. Peyrollier K. Hundal H.S. J. Biol. Chem. 2002; 277: 13628-13634Abstract Full Text Full Text PDF PubMed Scopus (89) Google Scholar), respectively. Palii et al. (7Palii S.S. Chen H. Kilberg M.S. J. Biol. Chem. 2004; 279: 3463-3471Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar) isolated a tripartite AA responsive domain in the first intron of SLC38A2 that is required for increasing SLC38A2 mRNA during nutrient stress. Pharmacological and genetic interventions have implicated the classical extracellular signal-regulated kinase (ERK) and stress-activated Jun N-terminal kinase (JNK) mitogen-activated protein kinases (MAPK) in the adaptive regulation of System A (8Franchi-Gazzola R. Visigalli R. Bussolati O. Dall'Asta V. Gazzola G.C. J. Biol. Chem. 1999; 274: 28922-28928Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar, 9Lopez-Fontanals M. Rodriguez-Mulero S. Casado F.J. Derijard B. Pastor-Anglada M. J. Gen. Physiol. 2003; 122: 5-16Crossref PubMed Scopus (30) Google Scholar) in certain cell types, although nutrient signaling loci upstream of MAPKs have not been identified to date. System A substrates, including the synthetic AA analogue α-methylaminoisobutyric acid (Me-AIB), suppress the increase in SNAT2 expression that occurs in AA-starved cells (9Lopez-Fontanals M. Rodriguez-Mulero S. Casado F.J. Derijard B. Pastor-Anglada M. J. Gen. Physiol. 2003; 122: 5-16Crossref PubMed Scopus (30) Google Scholar, 10Ling R. Bridges C.C. Sugawara M. Fujita T. Leibach F.H. Prasad P.D. Ganapathy V. Biochim. Biophys. Acta. 2001; 1512: 15-21Crossref PubMed Scopus (89) Google Scholar, 11Gazzola R.F. Sala R. Bussolati O. Visigalli R. Dall'Asta V. Ganapathy V. Gazzola G.C. FEBS Lett. 2001; 490: 11-14Crossref PubMed Scopus (73) Google Scholar). Hence, the concept that SNAT2 activity may regulate transporter expression has been developed and both direct and indirect mechanisms of transport-dependent sensing have been proposed (1Hyde R. Taylor P.M. Hundal H.S. Biochem. J. 2003; 373: 1-18Crossref PubMed Scopus (282) Google Scholar). SNAT2 may function as a hybrid transporter-receptor (transceptor) whereby structural changes during the transport cycle may be transduced to signaling pathways; such AA transceptors are well documented in Saccharomyces cerevisiae (12Donaton M.C. Holsbeeks I. Lagatie O. Van Zeebroeck G. Crauwels M. Winderickx J. Thevelein J.M. Mol. Microbiol. 2003; 50: 911-929Crossref PubMed Scopus (119) Google Scholar, 13Wu B. Ottow K. Poulsen P. Gaber R.F. Albers E. Kielland-Brandt M.C. J. Cell Biol. 2006; 173: 327-331Crossref PubMed Scopus (58) Google Scholar) but have yet to be verified in mammalian systems. However, this transceptor model cannot completely account for the control of System A/SNAT2, because AAs that are not substrates for this transporter (Trp, Tyr, Phe) also repress System A (14Heaton J.H. Gelehrter T.D. J. Biol. Chem. 1977; 252: 2900-2907Abstract Full Text PDF PubMed Google Scholar). Here, AA deprivation is shown to stimulate functional expression of SNAT2 in rat L6 myotubes and HeLa cells through transcriptional induction and protein stabilization. The AA-dependent control of the latter process is conferred by a region in the N terminus of SNAT2. Inhibitors of JNK impair adaptive regulation and the AAs shown to regulate SNAT2 expression are either capable of restraining JNK phosphorylation and/or interacting with SNAT2. By using a shRNA approach it is demonstrated that SNAT2 functions as a mammalian AA transceptor, which acts in an autoregulatory gene expression pathway. Materials—Culture media (α-minimal essential medium, α-MEM, and Dulbeccos MEM), fetal calf serum and antibiotic/antimycotic solution were from Invitrogen (Paisley, UK). Radiochemicals were from PerkinElmer Life Sciences (Cambridge, UK). AAs and reagent grade chemicals were from Sigma-Aldrich (Poole, UK). Inhibitors were from Tocris (Bristol, UK) and CN Biosciences (Nottingham, UK). Molecular Biology—Rat SNAT2 (15Yao D. Mackenzie B. Ming H. Varoqui H. Zhu H. Hediger M.A. Erickson J.D. J. Biol. Chem. 2000; 275: 22790-22797Abstract Full Text Full Text PDF PubMed Scopus (198) Google Scholar) and SNAT5 (16Nakanishi T. Kekuda R. Fei Y.J. Hatanaka T. Sugawara M. Martindale R.G. Leibach F.H. Prasad P.D. Ganapathy V. Am. J. Physiol. Cell Physiol. 2001; 281: C1757-C1768Crossref PubMed Google Scholar) cDNAs were placed in pcDNA6 (Invitrogen) following PCR mutagenesis to introduce a C-terminal V5-His6 tag. The SNAT2-SNAT5 NcoI chimera was generated in pCR2.1 and subcloned into pcDNA6. The rat SNAT2 promoter was cloned from L6 DNA and subcloned into pC3luc (see supplementary data). SNAT2 shRNAs were expressed from psiStrike (Promega) (Sequences: shSNAT2, ACCGATGAAC GTGTCCAAGA TTAAGTTCTC TAATCTTGGA CACGTTCATC TTTTTC; shSCRAMBLE, ACCGCTGGTT CAAACGTTAA GAAAGTTCTC TTCTTAACGT TTGAACCAGC TTTTTC). Cell Culture—L6 cells were grown and cultured to the stage of myotubes as described (6Hyde R. Peyrollier K. Hundal H.S. J. Biol. Chem. 2002; 277: 13628-13634Abstract Full Text Full Text PDF PubMed Scopus (89) Google Scholar). HEK 293 and HeLa cells were grown in Dulbecco's modified Eagle's medium/10% fetal calf serum/1% antibiotics. For experiments, cells were incubated in Earls' Balanced Salt Solution (EBSS) supplemented with an AA mix at 1× physiological concentration (5Hyde R. Christie G.R. Litherland G.J. Hajduch E. Taylor P.M. Hundal H.S. Biochem. J. 2001; 355: 563-568Crossref PubMed Scopus (72) Google Scholar) for 1 h, rinsed, and incubated in the appropriate buffer for cell stimulation. Monolayers were rinsed prior to lysis or cell membrane isolation (17Hajduch E. Alessi D.R. Hemmings B.A. Hundal H.S. Diabetes. 1998; 47: 1006-1013Crossref PubMed Scopus (295) Google Scholar). L6/HeLa cells transfected with pC3luc constructs were incubated in growth medium supplemented with 0.8 mg/ml G418 (Melford, Suffolk, UK). HEK 293 and HeLa cells were transiently transfected with pCDNA6 or psiStrike constructs (10 μg DNA/10-cm plate) using calcium phosphate. Solute Transport—Transport measurements were performed as previously described (17Hajduch E. Alessi D.R. Hemmings B.A. Hundal H.S. Diabetes. 1998; 47: 1006-1013Crossref PubMed Scopus (295) Google Scholar), with slight alterations. Following cellular stimulations, L6 cells were rinsed with Hepes-buffered saline (HBS), exposed to uptake solutions (typically 10 μm [14C]Me-AIB (1 μCi/ml) in HBS) for 10 min at room temperature, then rinsed 3× with 0.9% NaCl (4 °C) and lysed in 50 mm NaOH. Nonspecific radiotracer binding was assessed with 10 mm unlabeled Me-AIB (or by brief tracer exposure in kinetics experiments). For cis-inhibition experiments, the uptake solution was supplemented with unlabeled AAs as noted. For trans-inhibition experiments (described under supplementary data), L6 myotubes were preincubated in AA-free EBSS for 4 h and then exposed to 2 mm unlabeled AA for 20 min (at 37 °C). Me-AIB uptake was subsequently measured as above. Uptake was standardized against protein recovery, or is presented relative to a control value. Immunoblotting—20-50 μg of cell protein was separated by SDS-PAGE as previously described (17Hajduch E. Alessi D.R. Hemmings B.A. Hundal H.S. Diabetes. 1998; 47: 1006-1013Crossref PubMed Scopus (295) Google Scholar). Polyvinylidene difluoride membranes were probed with signaling antibodies (New England Biolabs, Hitchin, UK), monoclonal antibodies against the Na/K-ATPase α1 subunit (α6F; 1/8000; DSHB, University of Iowa), the V5 epitope (1/5000; Invitrogen) or chicken anti-SNAT2 IgY (1/200, ∼0.5 μg/ml; Antigen, FLLESNLGKYET). Luciferase Activity—Luciferase assay was performed in 20 μg of cell lysate using the Luciferase Assay System (Promega, Southampton, UK; according to the manufacturer's protocol) in a TD20/20 luminometer (Turner BioSystems, Sunnyvale, CA). Luciferase activity in experimental samples was standardized relative to serum-starved, AA-supplemented cells. Immunofluorescence—HEK 293 cells were grown on coverslips and transfected with pcDNA6-V5His-Version A (Empty Vector, Invitrogen), pcDNA6-LacZ-V5His or the pcDNA6-SNAT2-V5His construct. Cells were rinsed with PBS and fixed with paraformaldehyde (4%, w/v) in PBS for 30 min. Cells were rinsed twice with PBS and rehydrated in PBS/2% fish skin gelatin (w/v) for 30 min. Cells were then incubated in PBS in the presence or absence of 0.2% Triton X-100 for 2 min. Detergent was removed with four changes of PBS. Coverslips were incubated overnight with anti-V5 monoclonal antibodies (1/250, v/v) in PBS/2% fish skin gelatin, were subsequently rinsed with PBS, and were incubated with fluorescein-conjugated anti-mouse antibody in PBS/gelatin. The rinsed and mounted coverslips were visualized by fluorescence microscopy. Statistical Analysis—One-way analysis of variance and non-linear regression were performed using GraphPad Prism 4 software and considered statistically significant at p < 0.05. For the non-linear regression analyses in Fig. 6, Equations 1 (inhibition) and 2 (repression) were used, y=ymax-((ymax·x)/(Kd+x)) y=ymax-((x·(ymax-ymin))/(Kd+x)) where the parameters ymax, ymin, and Kd are derived from the best-fit curve to the experimental data (y being transport rate/relative luciferase activity and x being inhibitor/repressor concentration) using GraphPad Prism 4. Online Supplemental Material—Supplementary methods containing extended protocols (cloning, transfection, buffers, IgY generation, and purification) and additional data showing: (i) the time-dependent increase in System A activity during AA deprivation; (ii) the repression of System A activity by individual AAs; (iii) an alignment of the N termini of vertebrate SNAT2 orthologues; (iv) the lack of effect of CaR modulators on adaptive regulation of System A; (v) the effect of altered extracellular pH on System A transport and adaptive regulation; and (vi) a comparison of the ability of individual AAs to cis-inhibit and trans-inhibit System A are available online. Amino Acid Specificity of System A Regulation—AA starvation of L6 myotubes induces a time-dependent increase in System A activity ((5Hyde R. Christie G.R. Litherland G.J. Hajduch E. Taylor P.M. Hundal H.S. Biochem. J. 2001; 355: 563-568Crossref PubMed Scopus (72) Google Scholar), Fig. 1A, and supplemental Fig. S1) through a process dependent upon new protein synthesis (net induction in nutrient-deprived cells: 92.8 ± 32.3 pmol/min/mg; net induction in the presence of the translation inhibitor cycloheximide: 16.1 ± 3.8 pmol/min/mg; p < 0.05). The AAs that regulate System A activity were determined by providing individual AAs (at 2 mm) to otherwise AA-starved myotubes (Fig. 1A and supplemental Fig. S2). Consistent with the idea that the System A transporter might generate a repressive signal during the transport cycle, AAs shown to be the best substrates for SNAT2 (18Reimer R.J. Chaudhry F.A. Gray A.T. Edwards R.H. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 7715-7720Crossref PubMed Scopus (162) Google Scholar) were generally the compounds with the greatest repressive effect, although aromatic AAs were also found to significantly repress System A induction (p < 0.05). Cis-inhibition experiments were used to demonstrate that although 2 mm Tyr impaired System A up-regulation, Tyr did not significantly interact with System A at this concentration (Fig. 1B). Nevertheless, (l)-Gln, Ala, and Me-AIB all interact with SNAT2 and the non-repressive AAs did not (Fig. 1B). SNAT2 abundance increases in AA-deprived L6 myotubes (5Hyde R. Christie G.R. Litherland G.J. Hajduch E. Taylor P.M. Hundal H.S. Biochem. J. 2001; 355: 563-568Crossref PubMed Scopus (72) Google Scholar). Hence, the effects of the above AAs on SNAT2 protein expression were assessed using chicken anti-SNAT2 antibodies (Fig. 1C). These antibodies were designed against a region in the cytoplasmic N-terminal domain of SNAT2 (FLLESNLGKYET) and bind to two broad bands (≈35 kDa and 55 kDa) that increase in response to AA deprivation. Given the co-regulation of these proteins we propose that the higher band is full-length SNAT2, and the lower band is a cleaved form of SNAT2. A cross-reactive band (≈75 kDa) is also observed; however, this protein is restricted to intracellular membranes (data not shown). Tyr, Ala, and (l)-Gln prevented the induction of SNAT2 protein and Me-AIB had a slight repressive effect after 4 h (Fig. 1C), which was almost complete after 6 h of incubation with Me-AIB. Amino acids that were non-repressive following4hof stimulation (Leu, Tyr, Arg) remained non-repressive even when incubated with cells for 6 h. Because Me-AIB had a modest repressive effect on SNAT2 expression, but a large effect on System A transport, the overall repressive effect of Me-AIB on System A activity may largely be due to trans-inhibition of plasma membrane SNAT2 by cytoplasmic Me-AIB (19Bracy D.S. Handlogten M.E. Barber E.F. Han H.P. Kilberg M.S. J. Biol. Chem. 1986; 261: 1514-1520Abstract Full Text PDF PubMed Google Scholar). To test whether Me-AIB exerted a repressive effect strictly on SNAT2 gene expression, analysis of the SNAT2/SLC38A2 promoter was performed. Amino Acids Control SNAT2 Transcription and Protein Stability—The SLC38A2 AA responsive domain isolated by Palii et al. contains three elements, a CAAT-box, an amino acid response element (AARE in Fig. 1D) and a purine-rich element (PuR in Fig. 1D), each of which contribute to SNAT2 expression during AA limitation (7Palii S.S. Chen H. Kilberg M.S. J. Biol. Chem. 2004; 279: 3463-3471Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar). The rat SNAT2 promoter was cloned and two versions of this promoter, one (-1 + 1) containing the first intron of the gene and one (-1 + 0.3), which lacks the first intron and its associated nutrient responsive domain, were placed upstream of firefly luciferase (Fig. 1D). SNAT2-luciferase vectors were stably expressed in L6 cells. The inclusion of 1 kb of upstream sequence in both SNAT2-luciferase promoters produced a substantial basal level of transcription relative to CMV minimal promoter (Fig. 1D), as also seen with upstream regions of the mouse SNAT2 promoter (7Palii S.S. Chen H. Kilberg M.S. J. Biol. Chem. 2004; 279: 3463-3471Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar). The SNAT2-1+1-luciferase transgene was significantly induced following7hofAA deprivation (Fig. 1D). SNAT2-luciferase induction was dependent upon the presence of intron 1, because SNAT2-1+0.3-luciferase activity was not stimulated by AA deprivation. The relative activation produced by the nutrient responsive domain of rat SNAT2 intron 1 in the absence or presence of AA is around 2.2 (Fig. 1D), which, given the short duration of AA starvation we used (7 h versus 10 or 12 h in other studies), is comparable to results reported elsewhere (∼1.8-2.5 for mouse SNAT2 intron 1 (7Palii S.S. Chen H. Kilberg M.S. J. Biol. Chem. 2004; 279: 3463-3471Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar); ∼3 for human SNAT2 intron 1, (20Gaccioli F. Huang C.C. Wang C. Bevilacqua E. Franchi-Gazzola R. Gazzola G.C. Bussolati O. Snider M.D. Hatzoglou M. J. Biol. Chem. 2006; 281: 17929-17940Abstract Full Text Full Text PDF PubMed Scopus (87) Google Scholar). As a further negative control, a luciferase construct driven by the cytomegalovirus (CMV) promoter was also unaffected by nutrient availability. In the presence of the repressive AAs ((l)-Gln, Ala, Tyr, Me-AIB; 2 mm), SNAT2-1+1-luciferase activity was reduced to control (AA+) levels, whereas Arg, a non-repressive AA, did not affect expression of this construct. Given that nutrient-responsive transcriptional controls appear to impart only a modest stimulus for SNAT2 induction, we examined the possibility that additional levels of control contributed to transporter activation. Hatzoglou's group (20Gaccioli F. Huang C.C. Wang C. Bevilacqua E. Franchi-Gazzola R. Gazzola G.C. Bussolati O. Snider M.D. Hatzoglou M. J. Biol. Chem. 2006; 281: 17929-17940Abstract Full Text Full Text PDF PubMed Scopus (87) Google Scholar) has recently found an internal ribosome entry sequence (IRES) in the 5′-untranslated region (UTR) of the SLC38A2 gene; however, unlike in certain other transcripts where an IRES has been documented (e.g. CAT1 (21Fernandez J. Yaman I. Sarnow P. Snider M.D. Hatzoglou M. J. Biol. Chem. 2002; 277: 19198-19205Abstract Full Text Full Text PDF PubMed Scopus (166) Google Scholar)), the SNAT2 IRES does not stimulate translation but permits uninterrupted transporter expression during nutrient stress, a condition where global translation initiation may be restricted. We proposed that SNAT2 protein may be stabilized during nutrient stress and to test this we expressed a CMV-driven SNAT2 construct (containing a C-terminal V5-His6 epitope tag) in HeLa cells, since these are more readily transfected than L6 cells and in separate experiments (not shown) were found to also exhibit System A adaptation in a JNK-dependent manner. V5 immunoreactivity was observed in lysates of SNAT2V5-transfected HeLa cells with a banding pattern appropriate for the full-length SNAT2 glycoprotein (55-60 kDa). The SNAT2V5 signal was absent from cells which had been cotransfected with anti-SNAT2 shRNAs (but not from cells transfected with scrambled shRNA controls), thus validating that the protein bands observed are a specific consequence of SNAT2V5 expression. AA deprivation of the SNAT2V5-transfected HeLa cells for 8 h led to a marked increase in V5 immunoreactivity, implying an increased abundance of SNAT2V5 protein in these cells (Fig. 2A). Our studies in L6 cells had shown that the CMV promoter used here was not sensitive to AA-deprivation (Fig. 1D) and to verify the specificity of this AA-dependent increase in SNAT2V5 protein, rat SNAT5 (a structurally related transporter from the SLC38 gene family) was V5-tagged and expressed in HeLa cells. Although analogous to the SNAT2V5 construct, an equivalent induction of SNAT5V5 under nutrient deprivation was not observed (Fig. 2A). Additionally, given that the two constructs were expressed from an identical transcriptional promoter (CMV), SNAT5V5 appears to be considerably more stable than SNAT2V5 in HeLa cells. Our interpretation of these data is that during AA-sufficiency SNAT2 protein is considerably less stable than during AA-restriction. A corollary of the proposal that SNAT2 is stabilized during nutrient deprivation is that a particular region of the SNAT2 sequence may confer AA-regulated stability effects. It should therefore be possible to identify such regions through mutagenesis. The cytoplasmic domains of SNAT2 are the best candidate regions for associating with the proteolytic machinery; however, the structure of SNAT2 within the cell membrane is poorly characterized. Two topologies have been proposed which differ in the number of transmembrane domains (15Yao D. Mackenzie B. Ming H. Varoqui H. Zhu H. Hediger M.A. Erickson J.D. J. Biol. Chem. 2000; 275: 22790-22797Abstract Full Text Full Text PDF PubMed Scopus (198) Google Scholar, 18Reimer R.J. Chaudhry F.A. Gray A.T. Edwards R.H. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 7715-7720Crossref PubMed Scopus (162) Google Scholar, 22Sugawara M. Nakanishi T. Fei Y.J. Huang W. Ganapathy M.E. Leibach F.H. Ganapathy V. J. Biol. Chem. 2000; 275: 16473-16477Abstract Full Text Full Text PDF PubMed Scopus (231) Google Scholar), as illustrated in Fig. 2B. Both models predict a large intracellular N terminus, but the 12-pass model proposed by Sugawara et al. (22Sugawara M. Nakanishi T. Fei Y.J. Huang W. Ganapathy M.E. Leibach F.H. Ganapathy V. J. Biol. Chem. 2000; 275: 16473-16477Abstract Full Text Full Text PDF PubMed Scopus (231) Google Scholar) differs from the 11-pass model of Yao et al. (15Yao D. Mackenzie B. Ming H. Varoqui H. Zhu H. Hediger M.A. Erickson J.D. J. Biol. Chem. 2000; 275: 22790-22797Abstract Full Text Full Text PDF PubMed Scopus (198) Google Scholar) and Reimer et al. (18Reimer R.J. Chaudhry F.A. Gray A.T. Edwards R.H. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 7715-7720Crossref PubMed Scopus (162) Google Scholar) in placing the C terminus at an intracellular location. To help distinguish between the two models, SNAT2V5 was expressed in HEK 293 cells and the cells were subject to analysis by immunofluorescence (Fig. 2C). Unlike the cytoplasmically localized LacZV5 control, HEK293 cells transfected with SNAT2V5 did not require permeabilization with Triton X-100 to allow an appreciable signal to be observed in anti-V5 immunofluorescence. Hence, in SNAT2V5, the C-terminal V5 epitope is accessible from the extracellular domain, supporting the 11-pass topology of SNAT2. Sequence identity between SNAT proteins is highest within the transmembrane domains and lowest within the hydrophilic N terminus (see supplemental Fig. S3) and the large extracellular loop and consequently it is predicted that isoform-specific regulatory domains are present within the hydrophilic regions. To test whether the N terminus of SNAT2 might control AA-dependent stability, a CMV-driven chimera was generated between SNAT2 and SNAT5V5 using a conserved NcoI site that maps to the first extracellular loop of the two proteins (as illustrated, Fig. 2B). Only 3 nucleotides from the SNAT2 5′-UTR were present in this construct (to minimize the potential influence of the 5′-UTR in AA-dependent control). Upon transfection of HeLa cells, SNAT2-5V5 protein expression was comparable to that of SNAT2V5 and was increased upon AA deprivation (Fig. 2A). Hence, by grafting the SNAT2 N terminus onto SNAT5 we have rendered the stability of the SNAT5 protein sensitive to AA availability. Signaling Pathways Mediating System A Adaptation—AA availability controls a variety of signaling pathways. To identify those pathways relevant to System A regulation, L6 myotubes were incubated with signal transduction inhibitors during our adaptation protocol prior to measuring Me-AIB transport (Fig. 3A). Inhibitors of the ERK pathway (PD-098059, 50 μm), the p38 MAPK pathway (SB-203580, 10 μm), or mTOR (rapamycin, 100 nm), neither mimicked (data not shown) nor impaired System A adaptation. In contrast, inhibitors of PI3K (wortmannin, 100 nm, and LY-294002, 50 μm) and JNK (SP-600125, 30 μm) significantly inhibited System A adaptation. 3-methyladenine (10 mm) inhibits class III PI3K isoforms (23Blommaart E.F. Krause U. Schellens J.P. Vreeling-Sindelarova H. Meijer A.J. Eur. J. Biochem. 1997; 243: 240-246Crossref PubMed Scopus (723) Google Scholar) and has been shown to inhibit autophagic proteolysis in AA-starved mammalian cells. This compound did not affect System A adaptation, which argues against a role for class III PI3K in the nutritional regulation of System A. AA withdrawal is shown to increase the phosphorylation (activation) of both ERK and JNK in L6 cells (Fig. 3B). Compared with previous studies (8Franchi-Gazzola R. Visigalli R. Bussolati O. Dall'Asta V. Gazzola G.C. J. Biol. Chem. 1999; 274: 28922-28928Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar, 9Lopez-Fontanals M. Rodriguez-Mulero S. Casado F.J. Derijard B. Pastor-Anglada M. J. Gen. Physiol. 2003; 122: 5-16Crossref PubMed Scopus (30) Google Scholar), wherein JNK and ERK signaling peaks within 30 min, MAPK activation was both delayed and prolonged in L6 cells, being apparent between 2 and 6 h of nutrient restriction. During AA deprivation, SEK1 (a MAPK kinase lying directly upstream of JNK) was rapidly phosphorylated, whereas the mTOR target p70-S6 kinase (S6K) was dephosphorylated and glycogen synthase kinase 3 (a downstream effector of class I PI3K signaling) was unaffected. The protein kinase GCN2 has been shown to phosphorylate the translation factor eIF2α during amino acid restriction (24Jousse C. Averous J. Bruhat A. Carraro V. Mordier S. Fafournoux P. Biochem. Biophys. Res. Commun. 2004; 313: 447-452Crossref PubMed Scopus (92) Google Scholar); however, although eIF2α phosphorylation increased with prolonged incubation time, AA withdrawal did not affect this in L6 myotubes. Tunicamycin, which affects eIF2α independently of GCN2, did stimulate eIF2α phosphorylation in L6 myotubes (21Fernandez J. Yaman I. Sarnow P. Snider M.D. Hatzoglou M. J. Biol. Chem. 2002; 277: 19198-19205Abstract Full Text Full Text PDF PubMed Scopus (166) Google Scholar). Ligand Sensitivity of JNK Signaling—If a single AA sensor was responsible for restraining SNAT2 adaptation, then that same sensor should also restrain signaling through the pathways that induce SNAT2 expression. Data from ourselves (this report) and others (9Lopez-Fontanals M. Rodriguez-Mulero S. Casado F.J. Derijard B. Pastor-Anglada M. J. Gen. Physiol. 2003; 122: 5-16Crossref PubMed Scopus (30) Google Scholar) implicate JNK in this process. However, relative to the GCN2 and mTOR pathways, AA signaling via JNK is poorly documented. To address this, we assessed the ability of specific AAs to inhibit the JNK activation typically observed in AA-starved cells (Fig. 4). Of the AAs tested, (l)-Gln and Tyr (2 mm) substantially reduced JNK phosphorylation, while Ala (also Ser and Trp) had more modest
Referência(s)