Advertisement
JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M301360200 on April 16, 2003

J. Biol. Chem., Vol. 278, Issue 26, 23786-23796, June 27, 2003
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
278/26/23786    most recent
M301360200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Meyer, E. L.
Right arrow Articles by Rogers, S. W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Meyer, E. L.
Right arrow Articles by Rogers, S. W.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Glutamate Receptor Subunit 3 Is Modified by Site-specific Limited Proteolysis Including Cleavage by {gamma}-Secretase*

Erin L. Meyer {ddagger} §, Nathalie Strutz ¶, Lorise C. Gahring {ddagger} || and Scott W. Rogers {ddagger} § **

From the {ddagger}Salt Lake City Veteran's Affairs Geriatrics Research, Education and Clinical Center and the Departments of §Neurobiology and Anatomy, ||Medicine, and Biology, University of Utah, Salt Lake City, Utah 84132

Received for publication, February 7, 2003 , and in revised form, March 18, 2003.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Ionotropic glutamate receptor (GluR) expression and function is regulated through multiple pre- and post-translational mechanisms. We find that limited proteolytic cleavage of GluR3 at two distinct sites generates stable GluR3 short forms that are glycosylated and found in association with other full-length GluRs in the mouse brain and cultured primary neurons. A combination of mutagenesis and transfection into HEK293 cells revealed cleavage by a {gamma}-secretase-like activity within the membrane-localized re-entry loop at or near the leucine-glycine pair (amino acids 585–586, GluR3s{beta}) and a second site within a proline-rich PEST-like sequence in the first cytoplasmic loop (Asp570-Pro571, GluR3s{alpha}). Generation of the prominent GluR3s{alpha} form was effectively abolished in the mutant, GluR3D570A, but inhibitors of lysosomes, the proteasome, caspases, or calpains had no effect. The possible impact of cleavage on receptor function was suggested when the co-expression of the GluR3P571Stop mutant (creating GluR3s{alpha}) co-assembled with other GluR subunits and decreased receptor function in Xenopus oocytes. In transiently transfected HEK293 cells, co-expression of GluR3s{alpha} alters the relative association between GluR1 and GluR3 during assembly, and the presence of the novel C-terminal proline-rich domain of GluR3s{alpha} imparts lateral membrane mobility to GluR complexes. These results suggest that limited proteolysis is another post-translational mechanism through which functional diversity and specialization between closely related GluR subunits is accomplished.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Neurons continually modify the relative expression, function, and subcellular localization of neurotransmitter receptors to maintain and fine-tune neurotransmission. Among the excitatory receptor systems modified are members of the {alpha}-amino-3-hydroxy-5-methylisoxazolepropionate (AMPA)1 family of the ionotropic glutamate receptor (GluR) that include subunits GluR1 thru GluR4 (1) where pre- and post-translational modifications range from RNA editing and alternative splicing to varied glycosylation (2) and phosphorylation (3). In addition, contained within the sequences of these subunits are amino acid motifs that can impart conditional functions including association with cellular proteins that govern appropriate sub-neuronal transport and localization (4, 5).

Proteolysis is another cellular mechanism for adjusting protein concentration and function. In particular, limited proteolysis through cleavage of the polypeptide at unique amino acid sequences affords a mechanism to impart distinctive functional differences between otherwise closely related proteins (6, 7). This mechanism appears operational on GluR members. For example, GluR1 is susceptible to activity-dependent limited proteolysis by a caspase 8-like protease in the C-terminal domain at sequence VSQD (residues 862–865, Ref. 8) that removes from the subunit sequences important for binding to cell substructure and subcellular localization (e.g. Ref. 9). In addition, GluR3 harbors a sequence in the first extracellular domain that exhibits glycosylation-sensitive susceptibility to cleavage by the serine protease, granzyme B (10).

Numerous sequence motifs that signal either direct sensitivity to proteolysis, or entry into degradation pathways have been identified (11). One of these, termed a PEST sequence (defined as polypeptide regions enriched for proline (P), glutamic acid (E, also aspartic acid, D), serine (S), and threonine (T) that are usually flanked by basic residues) is correlated with targeting proteins for rapid and often highly conditional site-specific cleavage or complete destruction (11, 12). Nontraditional PEST sequences may also occur at the N or C terminus of proteins, or possibly at or near boundaries of the polypeptide with membranes where these cytoplasmic domains are initiated or terminated (11). Here, we report that GluR3 is the substrate for limited cleavage by two distinct and independent proteolytic activities, the principal cleavage occurring at an aspartic acid-proline pair within a cytoplasmic-localized proline-rich PEST-like sequence. A second cleavage by a {gamma}-secretase/presenilin 1-related activity at or near a leucine-glycine pair occurs within the membrane re-entry loop that is proposed to construct the pore-forming domain.

Both proteolytic activities generate GluR3 short forms that are glycosylated and in stable association with other GluR subunits throughout the murine brain and in primary cultured cortical neurons. Notably, cleavage appears to be an intrinsic protein feature because the introduction of GluR3 cDNA into HEK293 cells by transient transfection or cRNA injection into Xenopus oocytes results in the generation of both GluR3 short protein forms observed in animal or cultured cell systems. Blocking the generation of the principal GluR3 form through modification of the proteolytic requirement of the aspartic acid to an alanine (GluR3D570A) to inhibit cleavage within the PEST-like sequence corresponds with enhanced amplitude of response to kainic acid relative to wild-type GluR3 when expressed in Xenopus oocytes. From these data we propose that GluRs contain multiple intrinsic signals for conditional modification by limited proteolysis, and these events contribute to subunit-specific modification of GluR function and expression.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Animal Tissues and Cell Culture—C57BL/6 mice were obtained from Jackson Laboratories (Bar Harbor, ME). Hippocampus, cortex (with some underlying basal ganglia), and cerebellum were dissected and solubilized for protein detection via Western blot analysis. Mixed neuronal/glial primary cortical cultures from mouse were generated and maintained as described elsewhere (13). Human embryonic kidney 293 cells (HEK293) were maintained in Dulbecco's Modified Eagle's Medium (Cellgro) containing 10% fetal bovine serum (Hyclone), Pen/Strep (Cellgro), and sodium pyruvate, and grown in a humidified incubator at 37 °C with 5% CO2 (see Ref. 10).

Transient transfection of HEK293 cells was done using the CalPhos Mammalian Transfection kit instructions (BD Clontech Laboratories, Inc) as described previously (8, 10). The mammalian expression vector used was pcDNA1/AMP (Invitrogen). In some cases, stably transfected colonies were selected using G418 (geneticin, Invitrogen). Limited dilutions were made to ensure the stable cell lines were clonal. Stable expression was confirmed by Western blotting using antibodies to both the N and C terminus of the protein. For experiments using varied temperature, HEK293 cells were transfected and maintained at 37 °C for ~12 h prior to moving them to a cell culture incubator kept at a different temperature at 5% CO2. Arrhenius plots were calculated as described elsewhere (8, 14).

Drugs (dissolved in Me2SO or cell growth media) at concentrations of 100–1000x were applied directly to the growth media at least 12-h post-transfection or 24-h post-plating if cells were not transfected. Cells were treated for 24–48 h. Protease inhibitors included; lysosomotropic agents 10 µM chloroquine and 4 mM ammonium chloride; caspase (Csp) inhibitors included the general caspase inhibitor Boc-D-FMK and Csp2: z-VDVAD-FMK, Csp3/6: z-DQMD-FMK, Csp6: z-VEID-FMK, Csp8: z-IWTD-FMK, all at 100 µM; calpain inhibitors, µ-Val-Hph-FMK, 100 µM and PD-150–606, 10–100 µM; the proteasome inhibitor lactacystin, 100 µM; and the {gamma}-secretase inhibitor, 100 µM. Deglycosylation of cultured cell lysates or murine hippocampal crude membranes (10, 15) was done after washing cells with PBS, solubilization in buffer containing 50 mM Tris, 150 mM NaCl, 1% Nonidet P-40, and incubating with N-glycosidase F (10 units/ml; 25,000 units/mg; RBI/Sigma Chemical Co.) for 2 h at 37 °C as before (10).

Site-directed Mutagenesis—A cassette portion of GluR3 between BamH1 (residue 1793) and SalI (residue 2360) was subcloned into the mutagenesis vector, pSP72 (Promega). The QuikChangeTM site-directed mutagenesis kit (Stratagene) was used according to kit instructions to produce mutations in the BamH1 to SalI fragment of GluR3 in pSP72. The mutated cassette was confirmed by automated sequencing (Sequencing Core facility, University of Utah) and then returned to GluR3 wild-type from which the corresponding BamH1-SalI fragment was removed. GluR subunits were subcloned into the mammalian expression vectors pcDNA1/AMP or pCDNA3.1 (both from Invitrogen) or into the RNA expression vector psGEM (a generous gift from Michael Hollmann). Qiagen Mini and Maxi kits were used to isolate plasmid DNA according to kit instructions. For translation in vitro, the T7 promoter system and rabbit reticulocyte lysate kit (Promega) supplemented with canine microsomes as per the manufacturer's directions was used.

Immunoprecipitation—Antibodies used include the mouse monoclonal antibody (mAb) to GluR3, mAb2F5 (10), mouse mAb 3A11, to GluR2 (Chemicon), rabbit anti-GluR1 polyclonal antibody from Oncogene, mouse and goat anti-presenilin1 (Santa Cruz Biotechnology) and anti-{beta}-amyloid (Zymed Laboratories Inc.). All secondary antibodies were from Jackson ImmunoResearch.

Transfected HEK293 cells were washed with PBS and dissolved with the aid of a glass Dounce homogenizer in RIPA buffer (50 mM Tris, 150 mM NaCl, 1% Nonidet P-40, 0.5% deoxycholate, 0.1% SDS, 0.2% Triton X-100, pH 7.5) containing protease inhibitors (phenylmethylsulfonyl fluoride (4 mM), iodoacetamide (10 mM), benzamidine (10 mM), and EDTA (10 mM), all freshly prepared). Solubilized cells were transferred to a microcentrifuge tube, cleared by centrifugation, and the supernatant dispensed to tubes with the appropriate antibody. The samples were rocked with antibody overnight at 4 °C, and then an additional hour at room temperature the following day. Protein-G-Sepharose beads were then added to each sample, and the tubes were rocked for an additional hour. The bead samples were then sedimented at high speed in a microcentrifuge, washed twice with RIPA buffer not containing Triton X-100, and placed in gel-loading buffer. These samples were boiled for 15 min, subjected to centrifugation, and the supernatant loaded onto an SDS-polyacrylamide gel. For immunoprecipitation of presenilin 1, RIPA was replaced with a buffer containing 10 mM Tris, 150 mM NaCl, 0.2% Triton X-100, 0.25% Nonidet P-40, 2 mM EDTA, 1% bovine serum albumin, pH 7.5 (16). The Triton X-100 and bovine serum albumin were omitted during the bead-washing step. A 1:1 mixture of protein A and protein G beads was used to precipitate presenilin 1.

Western Blot Analysis and Immunocytochemistry—Western blots were performed as described previously (8, 10). Briefly, transfected cells were harvested in immunoprecipitation buffer and mixed with 2x gelloading buffer containing dithiothreitol before boiling for 10–15 min followed by SDS-PAGE fractionation and transfer to nitrocellulose (10). Blots were blocked at room temperature for at least 1 h in PBS containing 5% dry milk and 0.05% Tween 20 (PBS-T). The blots were incubated overnight at 4 °C with slow agitation in primary antibody added to blocking solution. Blots were washed in successive changes of PBS-T and then incubated for 1 h in blocking solution containing peroxidase-conjugated secondary antibody. The blots were again washed with PBS-T and the bands detected on film after developing with the enhanced chemiluminescence kit (Amersham Biosciences). Gels were scanned and overlaid to compare the band sizes. Immunocytochemistry was done as in Ref. 8.

Electrophysiology—The full-length rat GluR3 cDNA and sequences modified as noted above were subcloned into the RNA expression vector, psGEM (from M. Hollmann). Xenopus oocytes were surgically removed and injected with 5 or 10 ng of cRNA that was synthesized using the Ambion kit for transcription in vitro. Yield and quantitation of injected RNA was measured using the RiboGreen RNA quantitation kit (Molecular Probes). Two-electrode voltage clamp recordings were performed by superfusion with kainic acid (300 µM) prepared in amphibian Ringer's solution. Oocytes were held at –70 mV, and the agonist was applied for 10–20 s at a flow rate of 10–14 ml/min.

Photobleaching Recovery Experiments—For photobleaching, transfected HEK293 cells grown on glass coverslips treated with polylysine were transferred to a live-cell chamber (30 °C) in Hanks' with 10 mM HEPES (pH 7.2) and no phenol red and then visualized with a Zeiss Axiovert 200 and Attoarc mercury lamp. A target cell was photographed and the mercury lamp light path was narrowed to a target beam of 2-µM diameter and power increased to 100 watts for ~1 min to quench the CFP. The power was returned to 25 watts, the iris opened, and photographs were taken at 30-s intervals for 7–10 min thereafter.

Reagents—{gamma}-Secretase inhibitor was obtained from Alexis Biochemicals. Caspase and calpain inhibitors were obtained from Calbiochem, Alexis Biochemicals, or Enzyme Systems Products. The metalloprotease inhibitor, KB8301, was from BD PharMingen and the inhibitors of ADAM proteases from Alexis Biochemicals. DNA-modifying enzymes were from New England BioLabs, Invitrogen, Promega, or Fermentas. Protein-G-Sepharose beads were from Amersham Biosciences and protein A beads from BioRad. All other reagents/drugs were from RBI/Sigma unless otherwise noted. Antibodies used include the mouse monoclonal antibody (mAb) to GluR3, mAb2F5 (10), mouse mAb3A11, to GluR2 (Chemicon), rabbit anti-GluR3 polyclonals 295, and 5209 (Carlson et al., Ref. 19), rabbit anti-GluR1 polyclonal antibody from Oncogene, mouse and goat anti-presenilin1 (Santa Cruz Biotechnology) and anti-{beta}-amyloid (Zymed Laboratories Inc.). All secondary antibodies were from Jackson ImmunoResearch.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
Western blot analysis of mouse brain tissue using antibodies prepared to AMPA family GluR subunits revealed two major reactive glycosylated species including the full-length GluR3 (~110 kDa) and a more rapidly migrating species at approximately ~72 kDa termed the GluR3 short form (Fig. 1). Similar fragments were not observed for GluR1 or GluR2 (Fig. 1A). The GluR3s nomenclature was selected to distinguish this R3 short form from a previously reported splice-variant found in cochlear cells (termed short-GluR3') that deletes 33 amino acids near the C terminus within the ligand-binding region of the S2 extracellular domain (17). To determine if GluR3s forms associate with other GluR subunits, detergent-solubilized membranes prepared from different brain regions of the C57BL/6 mouse were subjected to immunoprecipitation with either antibodies to GluR1 (Chemicon) or GluR2 (18, 19). Western blot analysis of the immunoprecipitate using a monoclonal antibody specific for GluR3 (mAb2F5, which binds to the extracellular region near the S1 domain, Ref. 10) revealed that both full-length GluR3 and GluR3s co-precipitated with either GluR1 or GluR2 in preparations from throughout the mouse brain (Fig. 1B). Also apparent on Western blots of immunoprecipitates was the clear distinction of two closely migrating GluR3s forms that differ in mobility by 1.8 kDa (termed GluR3s {alpha} or {beta} as in Fig. 1B).



View larger version (39K):
[in this window]
[in a new window]
  		FIG. 1.
GluR3 short forms are in association with other GluR subunits in the mouse brain and are produced in transfected cells. Panel A, Western blot analysis of whole protein from mouse (C57BL/6) cortex revealed full-length GluR1, GluR2, and GluR3 (~110 kDa, respectively) using antibodies specific to each subunit (see "Experimental Procedures"). The antibody used to detect GluR3 is a mouse monoclonal antibody (mAb2F5) that was prepared to an epitope in the first extracellular domain of GluR3, termed GluR3B (10). As shown, this antibody to GluR3 detects, in addition to the full-length receptor subunit, a smaller band at ~72 kDa (asterisk). This band is referred to as GluR3 short or GluR3s. Panel B, GluR3 short forms in the mouse brain are in association with other GluR subunits. Total protein from mouse cortex (Ctx), hippocampus (Hip), and cerebellum (Cb) was solubilized in RIPA buffer (see "Experimental Procedures") and enriched by immunoprecipitation with antibodies specific to GluR1 or GluR2. The immunoprecipitates were then subjected to SDS-PAGE and Western blot analysis using anti-GluR3 (mAb2F5). In all regions of the brain examined, GluR3 full-length (R3, bold arrow) and two distinct GluR3 short forms (R3s, termed {alpha} or {beta}; small arrows) co-precipitated in association with GluR1 or GluR2, respectively. Panel C, GluR3 short forms generated in HEK293 cells transfected with GluR3 cDNAs are glycosylated. Higher resolution Western blot analysis of protein from mouse hippocampus or HEK293 cells transiently transfected with cDNA encoding GluR3 revealed that the short GluR3 fragment is actually two stable products {alpha} (71 kDa) or {beta} (72.8 kDa) as indicated by arrows. Protein samples from these tissues were solubilized in deglycosylation buffer (see "Experimental Procedures") and incubated overnight with or without N-glycosidase-F (nGlycF). Migration of both the full-length GluR3 and GluR3s forms were reduced proportionately upon deglycosylation.

 

To assure the fidelity of co-precipitation, two experiments were done. First, cDNAs encoding GluR1 and GluR3, respectively, were translated either individually or together in vitro using rabbit reticulocytes supplemented with canine microsomes. These lysates were solubilized in RIPA buffer and immunoprecipitation performed as for membranes from intact tissues (see "Experimental Procedures"). Anti-GluR1 failed to co-immunoprecipitate GluR3 as determined by subsequent analysis on Western blots (not shown). Further, only full-length GluR3 was observed in lysate preparations suggesting this post-translational processing to generate GluR3s forms does not occur in reticulocyte lysates (not shown). The second method to assure co-precipitation fidelity was to transiently transfect HEK293 cells with cDNAs encoding either GluR1 or GluR3, and then mixing these independently transfected cells before preparing cell lysates by solubilization in RIPA buffer and subsequent immunoprecipitation and Western blot analyses. Similar to the results from translation in vitro, despite the presence of readily detectable GluR1 and GluR3 (full-length and short forms), no GluR3 was detected to co-precipitate with GluR1 (not shown). These experiments indicate that association between these respective subunits was not an artifact of immunoprecipitation conditions.

To determine if GluR3s forms were glycosylated, immunoprecipitates of GluR3 from hippocampal membranes were subjected to deglycosylation using N-glycosidase-F, an enzyme that removes all asparagine-linked glycosylations. As shown in Fig. 1C, there was an increase in mobility of full-length GluR3 and both GluR3s forms proportional to complete deglycosylation (see Ref. 10). The same fragments were identified on Western blots prepared from HEK293 cells transiently transfected with GluR3 cDNA (Fig. 1C) confirming the origin of these species to be from the GluR3 cDNA. Further, deglycosylation of this preparation resulted in changed migration of all forms of GluR3 proportional to equivalent glycosylation (Fig. 1C). The gels shown were selected for clarity of the appearance of both GluR3s forms. However, regardless of the protein source (transfected cells, brain tissue, or cultured neurons (not shown)), it is most typical for the GluR3s{alpha} to be prevalent and in many preparations GluR3s{beta} can be difficult to detect (see below). Inclusion of multiple protease inhibitors during all aspects of sample preparation had no effect on the incidence of GluR3s forms (not shown). Collectively, these results show that GluR3 short forms are present in preparations taken from various mouse brain regions and that these glycosylated forms are present in association with GluR1 and GluR2. Further, the generation of GluR3s forms do not occur when cDNA encoding GluR3 is translated in cell-free systems; however, upon introduction by transfection into cultured HEK293 cells, in addition to full-length GluR3 both GluR3s forms are observed, and they are glycosylated equivalently to the same species recovered from mouse brain suggesting their source is the result of a post-translational cell process.

Limited Proteolysis of GluR3 Produces Short Forms—Based upon the predicted molecular weight of the GluR3s forms, the likely site of GluR3s termination(s) was within the cytoplasmic domain between transmembrane 1 and the pore-forming re-entry loop (Fig. 2; see Ref. 1). In this region the GluR3 sequence is rich in prolines and acidic amino acids (Fig. 2A) that collectively resemble a PEST sequence (residues 564–575). Because GluR2 lacks detectable short fragments and the GluR2 sequence differs from GluR3 in the first cytoplasmic loop region, chimeras between these homologous regions were generated and expressed transiently in HEK293 cells (Fig. 2A). In all cases chimeras that disrupted GluR3Asp570 failed to generate GluR3s{alpha}; however, GluR3s{beta} was unaffected by the chimeras tested suggesting that GluR3s{alpha} and GluR3s{beta} are produced by distinct mechanisms and at different sites. Alanine mutagenesis confirmed the importance of GluR3Asp570 for generating GluR3s{alpha} since its production was effectively abolished when GluR3D570A was expressed in transfected cells (Fig. 2B). However, in some experiments (below), a light smear migrating at approximately this molecular weight was evident suggesting that additional minor fragments are revealed when GluR3s{alpha} is absent or that weak proteolysis of GluR3 in the same vicinity of GluR3D570A persists.



View larger version (85K):
[in this window]
[in a new window]
  		FIG. 2.
Disruption of GluR3Asp570 eliminates GluR3s{alpha}, and an inhibitor of {gamma}-secretase eliminates GluR3s{beta} Panel A, based upon the GluR3s molecular weight, the approximate number of amino acids in the GluR3s fragments placed the likely site of cleavage to generate GluR3s forms in the vicinity of the first cytoplasmic domain and re-entry loop (RL). Since there are no detectable short forms of GluR2, chimeras were generated containing the homologous region of this subunit in the GluR3 background between GluR3His554 and Phe575, arrows and amino acids in gray) and Western blot analysis done following transfection into HEK293 cells. In all cases, when GluR3Asp570 was altered, GluR3s{alpha} was absent. None of the chimeras generated influenced the formation of GluR3s{beta}. Amino acids are abbreviated with the standard one-letter code. Panel B, the importance of GluR3Asp570 to generation of GluR3s{alpha} was confirmed by site-directed mutagenesis (see "Experimental Procedures"). Conversion of this aspartic acid to an alanine (R3D570A) effectively eliminated the production of the GluR3s{alpha} fragment, but had no effect on the GluR3s{beta} proteolytic product. Panel C, in an attempt to identify the proteolytic activity responsible for cleavage at GluR3D570A, HEK293 cells were transfected with GluR3 wild-type (R3WT) and maintained at 37 °C for 24–48 h alone or in the presence of protease inhibitors. GluR3 was then subjected to immunoprecipitation using rabbit anti-GluR3 (polyclonal serum 295, Ref. 19) followed by detection on Western blots using anti-GluR3 mAb2F5 (see "Experimental Procedures"). A sampling of results is shown for GluR3 from cells placed in the indicated protease inhibitor (see "Experimental Procedures"). For caspases, which require an aspartic acid for substrate recognition, the Pan inhibitor exhibits a broad specificity toward caspase inhibition, whereas more specific caspase inhibitors included those that block 8, 6, 3/6, and 2. Neither the Pan inhibitor nor more specific inhibitors altered the generation of GluRs{alpha} or GluRs{beta} forms, respectively. For calpains, Mu (Mu-Val-HPh-FMK) is a nonspecific calpain inhibitor and Pd (PD150–606) is a specific calpain inhibitor. Lysosomotropic agents included ammonium chloride (NH4) and chloroquine (CQ) and a control (C). The proteasome/cathepsin A inhibitor, lactacystin (Lc, Ref. 45), is shown next to a control transfection (C). No protease inhibitor affected the GluR3s pattern with the notable exception of the transition state-specific inhibitor of {gamma}-secretase ({gamma}S, Ref. 21), which inhibited formation of the GluR3s{beta} form.

 

GluR3 Is a Substrate of {gamma}-Secretase and Associates with Presenilin 1—To identify the protease that cleaves at GluR3Asp570, attempts to inhibit this activity with a broad range of protease inhibitors toward caspases (based upon the requirement for GluR3Asp570, see also Refs. 8 and 20, calpains, lysosomal proteases, and the proteasome/cathepsin A (see Fig. 2C) as well as inhibitors of proline-endopeptidases, metalloproteases, ADAMS, and modulators of SREBP proteolytic activities, were tested without effect (not shown). However, it was observed that a state-dependent inhibitor of {gamma}-secretase (21) effectively and specifically abolished the formation of GluR3s{beta} (Fig. 2C). The inability to inhibit the formation of GluR3s{alpha} could have several explanations. For example, inhibitors of the responsible protease may not reach sufficient intracellular concentration to inhibit the cleavage. This is particularly true for peptide inhibitors that often cross the membrane relatively poorly and could themselves be substrates for proteolysis resulting in reduced efficacy. Of course, a proteolytic activity other than those tested for in these assays is certainly possible since it is likely that a multitude of cellular proteases, whose identities and novel subcellular localizations and function, remain to be determined.

The inhibition of {gamma}-secretase had no effect on the generation of GluR3s{alpha} in cells transfected with GluR3 (Figs. 2C and 3A). Notably, when cells were transfected with the GluR3D570A construct in the presence of the {gamma}-secretase inhibitor, both GluR3s forms were effectively abolished (Fig. 3A). The efficacy of the {gamma}-secretase inhibitor was confirmed by demonstrating inhibition of {beta}-amyloid processing (Fig. 3A). Again, these data are consistent with the independent origin of the respective GluRs forms. Attempts to treat cultured neurons with this inhibitor were not successful due to apparent toxicity (not shown). Because {gamma}-secretase activity is related to presenilins (22), we determined if presenilin1 (PS1) associates with GluR3. To test this, HEK293 cells stably expressing GluR3 were subjected to immunoprecipitation with a PS1 monoclonal antibody and the immunoprecipitate probed on Western blots with anti-GluR3 (Fig. 3B). In this assay a unique band corresponding with full-length GluR3 (GluR3s forms were not detected in association with PS1) was revealed suggesting a relatively stable co-association between these respective proteins. Similar experiments on cells transfected with GluR1 indicated little or no co-precipitated GluR1 signal (not shown), nor did GluR3 co-precipitate when antibodies to presenilin2 were used (not shown). Double-label immunocytochemistry of cultured primary neurons using goat anti-PS1 and rabbit anti-GluR3 showed these proteins co-localized in the soma, particularly in perinuclear regions consistent with endoplasmic reticulum and in dispersed structures similar to elements of the Golgi apparatus (Fig. 3C). PS1 immunostaining in neuronal processes was weak and at this level of detection does not necessarily colocalize with GluR3 in the dendrite (Fig. 3C). Further, as shown in the neuron in Fig. 3C, the generally good agreement between PS1 and GluR3 immunostaining, particularly in the more dispersed structures in the soma is consistent with the localization of PS1 reported by others (23). Confirmation that inhibition of {gamma}-secretase in cultured neurons eliminated GluR3s{beta} formation was not successful due to the toxicity of the inhibitor under the multiple conditions attempted.



View larger version (65K):
[in this window]
[in a new window]
  		FIG. 3.
The generation of GluR3s{beta} is consistent with cleavage of GluR3 by {gamma}-secretase. Panel A, to confirm {gamma}-secretase cleavage of GluR3 was specific to GluR3{beta} formation, HEK293 cells were transfected with GluR3 or GluR3D570A and maintained at 28 °C (which increases the amount of GluR3{beta} form visible on blots) for 24–48 h with or without addition of the {gamma}-secretase inhibitor (see "Experimental Procedures"). GluR3 was measured by immunoprecipitation (rabbit polyclonal 295, Ref. 19) followed by Western blot analysis and detection of products with mAb2F5. As shown previously, GluR3s{beta} dominates the GluR3s forms generated from GluR3D570A although a smear of weaker bands was detected in this experiment in the vicinity of the absent GluR3s{alpha}. Also, cells transfected with GluR3 and treated with the {gamma}-secretase inhibitor failed to generate the GluR3s{beta} fragment. Combining the use of the {gamma}-secretase inhibitor on GluR3D570A-transfected cells resulted in almost complete elimination of GluR3s forms. In the blot to the right, the efficacy and specificity of the {gamma}-secretase inhibitor was confirmed by the inhibition of {beta}-amyloid cleavage (asterisk). Panel B, antibodies to presenilin1 (PS1) co-precipitate GluR3. Because the {gamma}-secretase activity is believed to be associated with presenilins, we determined if PS1 could be found in association with GluR3. To do this HEK293 cells were stably transfected with full-length GluR3 and then subjected to immunoprecipitation with a monoclonal antibody prepared to PS1 (see "Experimental Procedures") followed by Western blot analysis with mAb2F5 to detect GluR3 or secondary antibody (20) alone. The arrowhead indicates the presence of GluR3 immunoreactivity in anti-PS1 immunoprecipitate consistent with full-length receptor protein. GluR3s forms were not detected in these experiments suggesting that they are not in stable association with PS1. In similar experiments that substituted antibodies to presenilin 2 (PS2), no GluR3 was detected (not shown). Panel C, co-localization of GluR3 and PS1 in primary cultured neurons is shown in this double-labeled neuron. GluR3 was localized to perinuclear staining (presumably endoplasmic reticulum (ER) and Golgi) and dendritic processes (white arrows) whereas presenilin 1 (PS1) staining was prominently in ER and Golgi compartments. Upon merging the images, the co-localized signal (yellow) was particularly apparent in the soma. The weak signal for PS1 in processes rarely co-localized with GluR3 suggesting that co-localization of these proteins may be strongest in post-ER compartments and prior to transport into processes. The nucleus is noted by an asterisk. Panel D, to determine if membrane fusion is required for GluR3 cleavage by the {gamma}-secretase-like activity, the generation of GluR3s{beta} from HEK293 cells transiently transfected with wild-type GluR3 and maintained at the temperatures indicated for 30 h was measured by Western blot band intensities and the Arrhenius plot shown was derived. The plot is best fit by two lines (solid lines) with a break (dashed line), between 23 and 18 °C (indicated by an arrow). The Arrhenius break indicates that, consistent with vesicular transport, membrane fusion is required for generating GluR3s{beta}. Panel E, to identify further the sites of GluR3s cleavage, GluR3 constructs where codons encoding a translational stop were substituted in the GluR3 cDNA and transiently expressed in HEK293 cells. Western blots of total protein from these cells show that two of these constructs, GluR3P571Stop and GluR3G586Stop co-migrated with GluR3s forms {alpha} and {beta}, respectively. Accompanying diagrams show the relative location of GluR3s{alpha} and GluR3s{beta} and the proposed sites of restricted proteolysis indicated by scissors that correspond to GluR3P571Stop or GluR3G586Stop, respectively. Notably, the GluR3G586Stop is inthe membrane region of the re-entry loop, consistent with the site of cleavage for other substrates of {gamma}-secretase. Panel F, GluR3 and GluR2 (which has not been detected to be a {gamma}-secretase substrate), differ in sequence in the re-entry loop at position GluR3Q590 that in GluR2 is an arginine due to RNA editing that converts the genomic encoded glutamine (Q) codon to one encoding an arginine (R). To determine if this amino acid difference altered susceptibility to {gamma}-secretase the point mutant GluR3Q590R (as in GluR2) was generated and expressed in HEK293 cells transiently. Subsequent Western blot analysis revealed that this mutation had no effect on the generation of GluR3s forms. Other transfectants were GluR3 wild-type (WT) in the absence of {gamma}-secretase inhibitor or in its presence (WT {gamma}-sec) and GluR3D570A. The blot was probed with mAb2F5.

 

Because {gamma}-secretase/PS1 proteolytic activity is membrane-associated and co-localized with GluR3 in compartments throughout the soma, it was necessary to determine whether membrane fluidity and/or vesicular transport was required for GluR3s{beta} cleavage. To do this HEK293 cells were transfected with GluR3 cDNA, maintained at 37 °C for ~12 h, and then cultures were placed at six temperatures ranging from 7 to 37 °C thereafter before immunoprecipitating GluR3 protein for analysis by Western blot to generate an Arrhenius plot for this activity (Fig. 3D). As shown, the formation of GluR3s{beta} was greatly diminished below 18 °C indicating a break in the Arrhenius plot consistent with a requirement for a membrane fusion step in the formation of GluR3s{beta} that is unaffected by lysosomotropic agents (Fig. 2C). This result, in combination with the immunocytochemistry results, suggests that cleavage of GluR3s{beta} by the {gamma}-secretase/PS1 proteolytic activity is likely to require vesicular transport, possibly from the endoplasmic reticulum to Golgi structures, or membrane fluidity is required for protease interaction and cleavage. The formation of GluR3s{alpha} exhibited no Arrhenius plot break (not shown); however, the kinetics of formation of this species was complex, especially at lower temperatures where its formation was in some experiments actually increased when vesicular trafficking was inhibited. Proteolysis of substrate proteins such as Notch or {beta}-amyloid by {gamma}-secretase is at a section within the membrane-spanning domain whose consensus sequence is not defined (7, 22). To determine the most likely site of {gamma}-secretase cleavage of GluR3s{beta}, GluR3 was scanned using stop mutagenesis (introduction of stop codons within the amino acid coding region of the cDNA and analyzing the resulting products on Western blots of transfected HEK293 cells). Two of these constructs upon transfection generated stable and glycosylated GluR3s forms that align in mobility with native GluR3s forms (Fig. 3E). The first was GluR3P571Stop that generates a protein with mobility of GluR3s{alpha} as would be expected by cleavage at residue GluR3Asp570. The construct GluR3G586Stop produced a product consistent with the mobility of GluR3s{beta}. Since this sequence is proposed to be located within the membrane, this suggests that {gamma}-secretase cleaves at or near the GluR3Leu585-Gly586 residue pair is consistent with the substrate location of this activity (22, 24). Collectively, these data suggest that GluR3 is a substrate for {gamma}-secretase/presenilin proteolysis, and these respective proteins associate during cellular synthesis and/or transport to generate GluRs{beta}. Notably, GluR2, which lacks detectable short forms (Fig. 1A), differs from GluR3 in the vicinity of the putative GluR3s{beta} cleavage site only at GluR3Gln590 that in GluR2 is an arginine from modification of the codon by RNA editing (1). However, generation of a GluR3Q590R mutant to create the GluR2 re-entry loop sequence in the GluR3 background failed to alter the generation of either GluR3s form relative to wild-type GluR3 (Fig. 3F). Therefore, conversion of this amino acid to the edited sequence in the GluR2 homologue is not sufficient by itself to impart susceptibility or resistance to {gamma}-secretase cleavage. For that reason, the substrate-specific cleavage determinants on GluR3 recognized by this activity must reside elsewhere in the protein.

GluR3 Short Forms Associate with other GluR Subunits and Modify Receptor Function—The influence of GluR3s on GluR function, subunit association, and subcellular mobility was investigated further. Previous studies have demonstrated that short forms of GluRs generated by mutation can be used effectively to examine their influence on receptor function (17) and to determine features pertaining to determinants of GluR subunit assembly (25). Using a similar strategy, we determined if GluR3s{alpha} (GluR3P571Stop) or GluR3s{beta} (GluR3G586Stop) associates with other GluR subunits in cells co-transfected with cDNAs encoding GluR1, and/or GluR3WT. As will be shown in greater detail below (also see Fig. 5), immunoprecipitation from these cells using anti-GluR1-specific antibodies and subsequent Western analysis with anti-GluR3 revealed that both GluR3s form stable associations with GluR1, GluR3WT, GluR3D570A or mixtures of these subunits. Similar to the results from experiments noted above, no association between GluR1 and GluR3WT or GluR1 and GluR3P571Stop was observed when these proteins were co-translated in vitro (not shown) using rabbit reticulocytes and canine microsomes, nor were stable associations formed between subunits in mixtures of solubilized cells transfected independently with wild-type or short forms of the above subunits (not shown).



View larger version (90K):
[in this window]
[in a new window]
  		FIG. 5.
GluR3s{alpha} alters relative GluR1 and GluR3 subunit association and harbors sequences that modify membrane lateral mobility. Panel A, Western blots of protein from HEK293 cells transiently co-transfected with a fixed amount of GluR1 cDNA (1 µg/ml) but increasing amounts of wild-type GluR3 (R3WT), GluR3D570A, or GluR3P571Stop as indicated above the gels in (µg/ml) for the respective cDNA are shown. In the lower two gels, cells were transfected with constant GluR1 and GluR3D570A cDNA (1 µg/ml each) and increasing R3P571Stop or GluR3E561Stop (µg/ml as indicated). A model indicating the relative site of stop mutations and the proline-rich domain that is removed is shown. Receptor protein was immunoprecipitated and probed to reveal GluR1-associated GluR3 and/or GluR3s forms. Beneath each blot is shown ramps/bars to illustrate the relative change in signal of each transfected species (R1, GluR1; R3, full-length GluR3: WT or GluR3D570A; R3s, GluR3 short forms as indicated). Panel B, enhanced cyan fluorescent protein (eCFP) N-terminal fusion constructs of GluR3WT, GluR3D570A, and GluR3P571Stop were co-transfected with R1 as matched to the receptor complexes shown in panel A. Each photo set shows a representative cell before bleaching (time = 0), immediately after photobleaching (PhB, white arrowhead) and after 6 min of recovery at 30 °C. Cells co-transfected with GluR1+eCFP-GluR3WT, GluR1+eCFP-GluR3P571Stop, or GluR1+eCFP-GluR3D570A+GluR3P571Stop recovered whereas GluR1+eCFPGluR3D570A did not suggesting that the R3s{alpha} harbors a unique region that imparts lateral membrane mobility to the complex. Removal of the proline-rich region C terminus from GluR3s{alpha} (GluR1+eCFP-GluR3D570A+GluR3E561Stop) resulted in complexes that failed to recover after bleaching indicating that the proline-rich C-terminal domain is important for imparting the lateral mobility.

 

Since GluR3s{alpha} lacks the re-entry loop and part of the ligand-binding domain, S2 (see Fig. 2), but associates with full-length GluR subunits, how does this impact receptor function? To address this question, Xenopus oocytes were injected with RNA prepared from plasmids encoding either GluR3WT or GluR3P571Stop, respectively, or both. As in Fig. 4A, injection of GluR3WT produced receptors with a robust response to kainic acid, but when co-injected with GluR3P571Stop, which exhibits no response to KA alone, the total current was markedly decreased in all experiments and at all RNA concentrations tried (Fig. 4, A and B and data not shown). Similar results were obtained when GluR1 was expressed alone or in the presence of R3P571Stop (not shown). Notably, these results are consistent with studies previously reported for naturally occurring GluR3 short forms that are generated in vivo by alternative splicing (17). If the co-expression and association of GluR3s{alpha} with full-length GluR3 or GluR1 acts to decrease overall receptor function, then the absence of GluR3s forms should correspond with enhanced function. This expectation was confirmed when the expression of GluR3D570A alone or with GluR1 resulted in significantly enhanced current amplitudes relative to oocytes injected with either GluR3WT or GluR1 alone (Fig. 4C). These results support the conclusion that inclusion of GluR3s{alpha} early in receptor assembly could impact upon overall receptor function either through reducing receptor function directly or possibly through decreasing productive subunit associations.



View larger version (31K):
[in this window]
[in a new window]
  		FIG. 4.
Functional consequences of GluR3 limited proteolysis. Panel A, RNA encoding GluR3P571Stop (GluR3s{alpha}) was co-injected into Xenopus oocytes with GluR3WT, or alone as indicated and the response to kainic acid (KA) measured. Tracings are shown for GluR3WT (top) and for GluR3P571Stop (middle), which exhibited no KA response. The drawings show proposed structural models of GluR3 that would be encoded by the RNAs injected. Co-injection of GluR3P571Stop with GluR3WT diminishes the current response relative to GluR3WT alone. Western blot analysis of GluR3WT from oocytes showed that the GluR3s{alpha} form is generated by oocytes (not shown). Panel B, a summary of the data in Panel A collected from multiple trials is shown. The ordinate is the mean current amplitude in nA ± S.E., and the oocytes tested are indicated above each bar. In all cases the presence of GluR3P571Stop decreased the total current measured significantly (p < 0.01). Panel C, when the amount of GluR3 short forms is reduced or absent through injection of RNA encoding GluRD570A, the GluR current amplitude is enhanced significantly (p < 0.01) relative to GluR3WT alone or when injected in combination with GluR1. In all cases the amount of RNA injected was quantitated (see "Experimental Procedures") and equalized for each construct.

 

GluR3 Short and the Proline-rich First Cytoplasmic Domain Alter Relative GluR Subunit Association—In other studies, examination of subunit assembly using truncated GluR subunits in transfected cells has been successfully employed to reveal assembly determinants in the protein structure of GluRs (25). Although it is not yet known when GluR3 short forms are generated, if they are created during the assembly of GluRs (e.g. endoplasmic reticulum and Golgi as suggested by Fig. 3C), could they impact upon relative subunit association through disruption or modification of receptor assembly? To examine this, cells were transfected with a constant amount of GluR1 cDNA while increasing the input amount of cDNA encoding either GluR3 full-length or GluR3P571Stop, and then the relative amount of GluR3 in association with GluR1 was measured through immunoprecipitation with anti-GluR1 and the subsequent measurement of the respective associated subunits on Western blots. As anticipated, the relative ratios of GluR1+GluR3full-length+GluR3s (including both short forms generated from GluR3WT or GluR3s{alpha} in cells transfected with GluR3P571Stop) increased proportionally with input DNA (Fig. 5A). However, when GluR1 and GluR3D570A input cDNAs are held constant (see Fig. 5A) and GluR3P571Stop input cDNA is increased; there is a proportional increase in GluR1+GluR3P571Stop associations that appear to occur at the expense of associations with full-length GluR3D570A. Given that GluR1+GluR3WT exhibits a proportional increase in the incorporation of full-length and short forms, yet GluR3P571Stop apparently decreases full-length GluR3 associations with GluR1, this suggests that if GluR3 short forms are generated early in assembly, they could impact upon the subunit composition of the final receptor complex (Fig. 5A).

As noted above, one consequence of cleavage at GluR3D570A is to introduce a new C-terminal proline-rich cytoplasmic domain into GluR3s that is rich in the motif, PXXP, a characteristic of SH3-binding domains (26). In other proteins, these domains have been related to the regulation of receptor transport and lateral mobility, which is also implicated in controlling synaptic numbers (27–29). To determine if this domain influences relative subunit association, the proline-rich domain was deleted through introduction of a stop codon into the GluR3s (GluR3E561Stop) construct, and the above co-transfection experiments were repeated. GluR3E561Stop associated with both GluR3 full-length and GluR1 (Fig. 5A and data not shown), but in almost complete contrast to GluR3P571Stop, co-transfection of GluR1 and GluR3D570A with increasing GluR3E561Stop cDNA markedly decreased association with GluR1 but increased GluR3D570A. Therefore the inclusion of the proline-rich region within the GluR3s construct appears to harbor determinants that if present during assembly could in part govern the relative subunit composition of mature receptor complexes (Fig. 5A).

The Novel C-terminal Proline-rich Domain in GluR3s{alpha} Alters Receptor Lateral Membrane Mobility—As noted above, proline-rich domains containing SH3-like binding motifs in the C termini of many proteins are related to modulating subcellular receptor mobility (4, 5). Therefore, one possible function of the proline-rich C-terminal domain generated in GluR3s{alpha} could be to modify GluR cellular mobility. To test this hypothesis, we measured the relative lateral receptor mobility of different GluR3 subunit combinations transfected into 293 cells. To visualize GluR3 subunit mobility, a variation of the method of Shi et al. (30) was used who demonstrated that green fluorescent protein (GFP) fused to the N terminus of GluR1 was a reliable reporter for measuring the mobility of these receptors in the mouse hippocampus. Instead of GFP, the relatively easily quenched variant-enhanced cyan fluorescent protein (eCFP) was substituted for GFP and introduced into GluR3 (see "Experimental Procedures"). For each experiment, transfected cells were bleached in a small region (Fig. 5B), and the recovery of eCFP determined. Cells transfected with GluR1+eCFPGluR3WT exhibited strong cytoplasmic and perinuclear staining that recovered from bleaching within 3 min. Control cells were transfected with eCFP-KDEL (Clontech). The product of this construct accumulated in the ER but did not aggregate, and its recovery to bleaching occurred within seconds (not shown). If cells were co-transfected with GluR1+eCFPGluR3D570A, recovery to bleaching was not observed (Fig. 5B), but co-transfection of GluR1+eCFP-GluR3P571Stop exhibited effectively complete recovery of the bleached region suggesting that inclusion of GluR3s determines lateral mobility. This was confirmed by co-transfection with GluR1+eCFP-Glu-R3D570A+GluR3P571Stop where recovery of the bleached region was reconstituted suggesting that the association of GluR3s (most likely GluR3s{alpha}) is dominant in promoting AMPA receptor lateral diffusion/movement within the cell. To test the idea that the proline-rich region could provide a structure for binding to other cellular proteins to favor lateral mobility, GluR3E561Stop was substituted for GluR3P571Stop. As shown in Fig. 5B, when this proline-rich region is absent, lateral mobility to the GluR1+eCFP-GluR3D570A co-transfected GluR complexes is not restored. This finding suggests that limited proteolysis reveals in GluR3s{alpha} a cryptic signal in the proline-rich C-terminal domain that, at least in transfected cells, contributes to regulating the lateral movement of the receptor complex.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
From these findings we propose that limited proteolysis contributes to the highly specific regulation of expression and function observed for otherwise closely related GluR subunits. In particular, limited proteolysis could reveal cryptic protein domains that both alter receptor function and modulate receptor mobility within the cell. If cleavage occurs early in assembly, the possibility that novel C-terminal domains in the GluR3s forms alter relative subunit assembly is also suggested by these findings. Consequently, limited proteolysis at sites intrinsic to the GluR subunit may be a more general mechanism to regulate these receptors. Further, the proteolytic cleavage within the re-entry loop by {gamma}-secretase to generate a GluR3s form has added implications since the fidelity of this protease is linked to proper amyloid protein processing (7, 22, 31). Therefore the failure of this proteolytic system, in addition to generating toxic {beta}-amyloid fragments, could also directly impact upon GluR function, which is intriguing in light of the contribution made by dysregulated GluRs to neuronal death through excitotoxic mechanisms (32).

Particularly interesting is that GluR3 short forms do not necessarily target the receptor complex for degradation. This is supported by the constant presence of clipped forms in the mouse brain, cultured neurons or in transfected cells, and the stable inclusion of GluR3 stop mutants into complexes that can be analyzed by immunoprecipitation and demonstrate altered membrane mobility. Further, we have examined the relative stability of long versus short forms of GluR3 and GluR3D570A in the presence or absence of cycloheximide and find no detectable difference in the relative stability of any of these species (not shown). If GluR3 short forms targeted degradation, it would be expected that either GluR3D570A receptors would be more stable or that lysosomotropic agents or inhibitors of the proteasome would have preferentially decreased the degradation rate of receptors harboring the short form. As shown in Fig. 2 and in repeated experiments not shown, these inhibitors had no effect on the relative ratio of long to short form or the relative overall degradation rate, and at present we have found no evidence for a difference in the relative degradation rate of full-length GluR3 relative to GluR3s{alpha} (not shown). Therefore, it would appear that the limited cleavage of GluR3 does not necessarily lead to clearing of the substrate receptors, but rather suggests that their occurrence could contribute to the relative location and function of the resulting receptor complex.

Exactly when and where cleavage within the cell occurs has not been determined. The generation of GluR3s{beta} through {gamma}-secretase dependent cleavage of GluR3 appears to be mostly in an early compartment such as the endoplasmic reticulum or Golgi, consistent with other reports for this proteolytic activity (23). As noted above, if this occurs prior to, or possibly coincident with, receptor assembly into mature complexes, then the results in Figs. 3, 4, 5 would predict that GluR3 short forms could impact upon the relative inclusion of heterologous GluR subunits into the mature receptor assembly. The result in Fig. 5 shows that the GluR3 stop mutants indeed contain the minimum structural elements for receptor subunit association (as previously reported, see Ref. 25) to promote GluR subunit association and receptor assembly. Also, when co-expressed in cells transfected with GluR1, GluR2, or GluR3 (including GluR3D570A), the GluR3 short forms as simulated by the stop constructs form detergent-stable assemblies (not shown), and these modify function in a predictable way that is consistent with earlier reports in the literature for naturally occurring GluR short forms (17).

Of course, cleavage could also occur in the mature receptor complex, and this could be important to the successful activity-dependent remodeling or re-distribution of receptors in compartments such as the dendrite and spine. For example, cleavage(s) of GluR3 in the mature receptor would generate novel C-terminal regions (see Fig. 6), and these could impart novel functions to the receptor complex, especially altering the mobility of the receptor particularly through revealing cryptic SH3-like domains for adherence to cell substructure. In this context, however, limited cleavage within subunits of the mature receptor pool would be expected to also generate a large fragment containing a portion of the re-entry loop, the S2-domain, and final transmembrane domain that should be detected in immunoprecipitations of the GluR complexes (see Fig. 2). However, numerous attempts to identify this predicted C-terminal receptor fragment failed (not shown). This could reflect that additional cleavage of this region results in fragments too small to detect, that lack the epitopes required for detection by the antibodies available to us (mouse mAb2D8 toward the S2 region, Ref. 18, or rabbit anti-GluR2/3 to the C-terminal domain (Chemicon)), or this fragment dissociates from the receptor complex upon detergent solubilization. Another possibility is that cleavage occurs early in receptor assembly, and only the GluR3s fragment(s) are included in subunit association and receptor assembly. It is of interest that current models of receptor structure (1) based upon elegant electrophysiological studies support a tetrameric subunit configuration for GluRs (33). However, protein-based studies have indicated the possibility of stable GluR subunit associations consistent with a pentameric structure (e.g. Ref. 34). Perhaps the existence of GluR3s forms could explain this discrepancy since their occurrence could easily confuse this issue, especially in protein-based assays whose sensitivity to detection on sucrose gradients and non-denaturing gels rely upon molecular weights to infer complex subunit composition, which is inherently difficult to interpret for multimeric complexes.



View larger version (39K):
[in this window]
[in a new window]
  		FIG. 6.
Summary of GluR3 residues from Glu560 to Asp594 with key cleavage sites and proposed subdomains indicated. Shown are the GluR3s{alpha} site of cleavage and the proposed GluR3s{beta} site of cleavage by {gamma}-secretase. Key sequence motifs including the PEST-like sequence, and two other sequence motifs specifically the Src-homology3 (SH3)-domain (minimal motif: PXXP, Ref. 26), and the Homer binding domain (minimal motif: PPXXF, Ref. 28) are indicated. Cleavage at these sites would introduce a novel C terminus and N-terminal region to an intact GluR3 subunit. Especially notable is that limited proteolysis at GluR3D570 is between the SH3-like domain (that harbors sequences imparting lateral mobility when R3s{alpha} is included in transfected cells, see Fig. 5 and text) and the Homer domain motif suggesting the possibility that these sites could be differentially exposed or modified upon limited proteolysis.

 

The consequences of limited proteolysis of GluR3 could be numerous and not necessarily restricted to just this subunit. For example, GluRs as targets of calpains is well known (6), and GluR1 harbors a cytoplasmic C-terminal sequence that imparts susceptibility to a caspase-like protease (8). In GluR1, cleavage at this C-terminal site removes from the subunit sequences implicated in modifying subcellular localization and receptor mobility (4, 5). Limited proteolysis may also occur at sequence-defined sites in the extracellular domain as demonstrated for GluR3, which contains a sequence that if not glycosylated is cleaved by the serine protease, granzyme B (10). The possible role of a PEST-like sequence in altering GluR3 susceptibility to limited cleavage (Fig. 6) would not necessarily be surprising since limited proteolysis at other receptor proteins within PEST or PEST-like sequences is well established (e.g. Ref. 11). Further, in some proteins, such as NOTCH, limited proteolysis at the PEST sequence results in the release from the membrane-bound complex of stable protein intermediaries important for imparting the signaling role of this molecule (e.g. Ref. 35). Numerous neuronal proteins also harbor PEST sequences ranging from neuronal cytoskeleton proteins such as MAP2 (36) to the cytoplasmic domains of neuronal nicotinic receptor subunits alpha2, alpha3, and alpha4 that despite often low sequence identity are well conserved among species (not shown). Although the role of these sequences in regulating receptor function through limited proteolysis is not yet fully resolved, the possibility that they impart susceptibility to limited proteolysis in response to rapidly changing conditions is suggested. Further, in neurons where GluRs are often concentrated within highly compartmentalized subcellular environments, such as the dendritic spine, limited proteolytic activity could provide a post-translational mechanism that is particularly responsive to rapidly changing conditions such as those described for GluR redistribution and functional modification that coincides with treatments favoring the establishment of long-term potentiation or long-term depression (e.g. Refs. 30 and 37–39).

In addition to modifying receptor function, the location of the identified sites of limited proteolytic cleavage near two domains implicated in receptor anchoring to cell substructure and mobility within the neurons is remarkably coincidental (Fig. 6). For example, a HOMER-binding domain motif (PPXXF, see Ref. 28) is present (PPNEF, GluR3 residues 571–575) immediately C-terminal to the GluR3s{alpha} cleavage site at GluR3Asp570. This sequence also occurs in GluR4 (PPNEF, human GluR4 sequence accession number P48058 [GenBank] ) at the site homologous to GluR3, but in the absence of the preceding proline-rich region. In other proteins, such as group 1 metabotropic glutamate receptors (40–42), the HOMER domain has been implicated in tethering the receptor to the region immediately adjacent to the postsynaptic density (40, 43, 44). Similarly, cleavage at GluR3Asp570 also reveals a novel C terminus in the first cytoplasmic domain of GluR3 (Fig. 6) that is rich in the motif, PXXP, a characteristic of SH3-binding domains (26). These domains at the protein C terminus, are associated with regulation of protein interaction with the cytoskeleton and subcellular transport related to controlling synaptic numbers and localization (27–29). It is tempting to speculate that successive (or independent) cleavage by the {gamma}-secretase-like activity would reveal the HOMER domain at the C terminus of GluR3s{beta} while the cleavage at GluR3Asp570 would remove the HOMER domain but reveal the SH3-like domain. Notably, cleavage to generate the novel C terminus was required to reveal the function of this domain in membrane lateral mobility since CFP-GluR3D570A failed to exhibit lateral mobility as did complexes where this domain was removed from the short form stop construct (i.e. those containing GluR1:CFP-GluR3D570A: GluR3E561Stop). Only constructs containing GluR3P571Stop (GluR3s{alpha}) exhibited the lateral mobility characteristic of wild-type GluR3 receptors. In the context of the present study, the presence (or absence) of this proline-rich domain alters receptor mobility and possibly relative subunit assembly (Figs. 5 and 6), both functions that in the neuron would be likely to contribute to the highly regulated expression of the AMPA-class GluRs and introduce additional structural complexity to explain how the diversity and regional specificity is further customized by the inclusion or exclusion during assembly of receptors from otherwise closely related subunits.


   FOOTNOTES
 
* This work was supported by National Institutes of Health Grant NS35181, the Val A. Browning Foundation and Deutsche Forschungsgemeinschaft (N. S.). 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. Back

** To whom correspondence should be addressed: University of Utah School of Medicine, Neurobiology and Anatomy, MREB 403, 50 North Medical Dr., Salt Lake City, UT 84132. Tel.: 801-585-6339; Fax: 801-585-3884; E-mail: Scott.Rogers{at}HSC.Utah.Edu.

1 The abbreviations used are: AMPA, {alpha}-amino-3-hydroxy-5-methylisoxazolepropionate; z, benzyloxycarbonyl; FMK, fluoromethylketone; PBS, phosphate-buffered saline; RIPA, radioimmune precipitation assay buffer; mAb, monoclonal antibody; GFP, green fluorescent protein; GluR, glutamate receptor; GluR3s, glutamate receptor short form; WT, wild type; HEK, human embryonic kidney; SH3, Src homology domain 3. Back


   ACKNOWLEDGMENTS
 
Emily Days and Karina Persiyanov are thanked for excellent technical assistance. Dr. Gisi Seebohm is thanked for RNA quantitation.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Dingledine, R., Borges, K., Bowie, D., and Traynelis, S. F. (1999) Pharmacol. Rev. 51, 7–61[Abstract/Free Full Text]
  2. Standley, S., and Baudry, M. (2000) Cell Mol. Life Sci. 57, 1508–1516[CrossRef][Medline] [Order article via Infotrieve]
  3. Swope, S. L., Moss, S. I., Raymond, L. A., and Huganir, R. L. (1999) Adv. Second Messenger Phosphoprotein Res. 33, 49–78[Medline] [Order article via Infotrieve]
  4. O'Brien, R. J., Lau, L. F., and Huganir, R. L. (1998) Curr. Opin. Neurobiol. 8, 364–369[CrossRef][Medline] [Order article via Infotrieve]
  5. Tomita, S., Nicoll, R. A., and Bredt, D. S. (2001) J. Cell Biol. 153, F19–24[CrossRef][Medline] [Order article via Infotrieve]
  6. Villa, P., Kaufmann, S. H., and Earnshaw, W. C. (1997) Trends Biochem. Sci. 22, 388–393[CrossRef][Medline] [Order article via Infotrieve]
  7. Nunan, J., and Small, D. H. (2000) FEBS Lett. 483, 6–10[CrossRef][Medline] [Order article via Infotrieve]
  8. Meyer, E. L., Gahring, L. C., and Rogers, S. W. (2002) J. Biol. Chem. 277, 10869–10875[Abstract/Free Full Text]
  9. Leonard, A. S., Davare, M. A., Horne, M. C., Garner, C. C., and Hell, J. W. (1998) J. Biol. Chem. 273, 19518–19524[Abstract/Free Full Text]
  10. Gahring, L., Carlson, N. G., Meyer, E. L., and Rogers, S. W. (2001) J. Immunol. 166, 1433–1438[Abstract/Free Full Text]
  11. Rechsteiner, M., and Rogers, S. W. (1996) Trends Biochem. Sci. 21, 267–271[CrossRef][Medline] [Order article via Infotrieve]
  12. Rogers, S., Wells, R., and Rechsteiner, M. (1986) Science 234, 364–368[Abstract/Free Full Text]
  13. Carlson, N. G., Bacchi, A., Rogers, S. W., and Gahring, L. C. (1998) J. Neurobiol. 35, 29–36[CrossRef][Medline] [Order article via Infotrieve]
  14. Hough, R., Pratt, G., and Rechsteiner, M. (1986) J. Biol. Chem. 261, 2400–2408[Abstract/Free Full Text]
  15. Rogers, S. W., Hughes, T. E., Hollmann, M., Gasic, G. P., Deneris, E. S., and Heinemann, S. (1991) J. Neurosci. 11, 2713–2724[Abstract]
  16. Pack-Chung, E., Meyers, M. B., Pettingell, W. P., Moir, R. D., Brownawell, A. M., Cheng, I., Tanzi, R. E., and Kim, T. W. (2000) J. Biol. Chem. 275, 14440–14445[Abstract/Free Full Text]
  17. Sekiguchi, M., Doi, K., Zhu, W. S., Watase, K., Yokotani, N., Wada, K., and Wenthold, R. J. (1994) J. Biol. Chem. 269, 14559–14565[Abstract/Free Full Text]
  18. Siegel, S. J., Janssen, W. G., Tullai, J. W., Rogers, S. W., Moran, T., Heinemann, S. F., and Morrison, J. H. (1995) J. Neurosci. 15, 2707–2719[Abstract]
  19. Carlson, N. G., Gahring, L. C., Twyman, R. E., and Rogers, S. W. (1997) J. Biol. Chem. 272, 11295–11301[Abstract/Free Full Text]
  20. Stennicke, H. R., Renatus, M., Meldal, M., and Salvesen, G. S. (2000) Biochem. J. 350, 563–568[CrossRef][Medline] [Order article via Infotrieve]
  21. Moore, C. L., Diehl, T. S., Selkoe, D. J., and Wolfe, M. S. (2000) Ann. N. Y. Acad. Sci. 920, 197–205[Medline] [Order article via Infotrieve]
  22. Sisodia, S. S., Annaert, W., Kim, S. H., and De Strooper, B. (2001) Trends Neurosci. 24, Suppl. 11, S2–6[CrossRef][Medline] [Order article via Infotrieve]
  23. Lah, J. J., Heilman, C. J., Nash, N. R., Rees, H. D., Yi, H., Counts, S. E., and Levey, A. I. (1997) J. Neurosci. 17, 1971–1980[Abstract/Free Full Text]
  24. Steiner, H., and Haass, C. (2000) Nat. Rev. Mol. Cell. Biol. 1, 217–224[CrossRef][Medline] [Order article via Infotrieve]
  25. Leuschner, W. D., and Hoch, W. (1999) J. Biol. Chem. 274, 16907–16916[Abstract/Free Full Text]
  26. Feller, S. M., Ren, R., Hanafusa, H., and Baltimore, D. (1994) Trends Biochem. Sci 19, 453–458[CrossRef][Medline] [Order article via Infotrieve]
  27. Arnold, D. B., and Clapham, D. E. (1999) Neuron 23, 149–157[CrossRef][Medline] [Order article via Infotrieve]
  28. Xiao, B., Tu, J. C., and Worley, P. F. (2000) Curr. Opin. Neurobiol. 10, 370–374[CrossRef][Medline] [Order article via Infotrieve]
  29. Ango, F., Pin, J. P., Tu, J. C., Xiao, B., Worley, P. F., Bockaert, J., and Fagni, L. (2000) J. Neurosci. 20, 8710–8716[Abstract/Free Full Text]
  30. Shi, S. H., Hayashi, Y., Petralia, R. S., Zaman, S. H., Wenthold, R. J., Svoboda, K., and Malinow, R. (1999) Science 284, 1811–1816[Abstract/Free Full Text]
  31. Murphy, M. P., Hickman, L. J., Eckman, C. B., Uljon, S. N., Wang, R., and Golde, T. E. (1999) J. Biol. Chem. 274, 11914–11923[Abstract/Free Full Text]
  32. Rothstein, J. D. (1996) Neurology 47, Suppl. 2, S19–S25; discussion S26[Medline] [Order article via Infotrieve]
  33. Rosenmund, C., Stern-Bach, Y., and Stevens, C. F. (1998) Science 280, 1596–1599[Abstract/Free Full Text]
  34. Wenthold, R. J., Yokotani, N., Doi, K., and Wada, K. (1992) J. Biol. Chem. 267, 501–507[Abstract/Free Full Text]
  35. Sriuranpong, V., Borges, M. W., Strock, C. L., Nakakura, E. K., Watkins, D. N., Blaumueller, C. M., Nelkin, B. D., and Ball, D. W. (2002) Mol. Cell. Biol. 22, 3129–3139[Abstract/Free Full Text]
  36. Friedrich, P., and Aszodi, A. (1991) FEBS Lett. 295, 5–9[CrossRef][Medline] [Order article via Infotrieve]
  37. Carroll, R. C., Lissin, D. V., von Zastrow, M., Nicoll, R. A., and Malenka, R. C. (1999) Nat. Neurosci. 2, 454–460[CrossRef][Medline] [Order article via Infotrieve]
  38. van Rossum, D., and Hanisch, U. K. (1999) Trends Neurosci. 22, 290–295[CrossRef][Medline] [Order article via Infotrieve]
  39. Luscher, C., Nicoll, R. A., Malenka, R. C., and Muller, D. (2000) Nat. Neurosci. 3, 545–550[CrossRef][Medline] [Order article via Infotrieve]
  40. Xiao, B., Tu, J. C., Petralia, R. S., Yuan, J. P., Doan, A., Breder, C. D., Ruggiero, A., Lanahan, A. A., Wenthold, R. J., and Worley, P. F. (1998) Neuron 21, 707–716[CrossRef][Medline] [Order article via Infotrieve]
  41. Tu, J. C., Xiao, B., Yuan, J. P., Lanahan, A. A., Leoffert, K., Li, M., Linden, D. J., and Worley, P. F. (1998) Neuron 21, 717–726[CrossRef][Medline] [Order article via Infotrieve]
  42. Tu, J. C., Xiao, B., Naisbitt, S., Yuan, J. P., Petralia, R. S., Brakeman, P., Doan, A., Aakalu, V. K., Lanahan, A. A., Sheng, M., and Worley, P. F. (1999) Neuron 23, 583–592[CrossRef][Medline] [Order article via Infotrieve]
  43. Kennedy, M. B. (1998) Brain Res. Brain Res. Rev. 26, 243–257[CrossRef][Medline] [Order article via Infotrieve]
  44. Kennedy, M. B. (1997) Trends Neurosci. 20, 264–268[CrossRef][Medline] [Order article via Infotrieve]
  45. Cuervo, A. M., Mann, L., Bonten, E. J., d'Azzo, A., and Dice, J. F. (2003) EMBO J. 22, 47–59[CrossRef][Medline] [Order article via Infotrieve]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 J. Biol. Chem.Home page
Y. Araki, N. Miyagi, N. Kato, T. Yoshida, S. Wada, M. Nishimura, H. Komano, T. Yamamoto, B. De Strooper, K. Yamamoto, et al.
Coordinated Metabolism of Alcadein and Amyloid {beta}-Protein Precursor Regulates FE65-dependent Gene Transactivation
J. Biol. Chem., June 4, 2004; 279(23): 24343 - 24354.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
278/26/23786    most recent
M301360200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Meyer, E. L.
Right arrow Articles by Rogers, S. W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Meyer, E. L.
Right arrow Articles by Rogers, S. W.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals   Molecular and Cellular Proteomics 
 Journal of Lipid Research   ASBMB Today 
Copyright © 2003 by the American Society for Biochemistry and Molecular Biology.
Advertisement
spacer
Advertisement
Advertisement