Analysis of the potential role of GluA4 carboxyl-terminus in PDZ interactions

Abstract

Background: Specific delivery to synapses of alpha-amino-3-hydroxy-5-methylisoxazole-4-propionate (AMPA) receptors with long-tailed subunits is believed to be a key event in many forms of activity-dependent changes in synaptic strength. GluA1, the best characterized long-tailed AMPA receptor subunit, contains a C-terminal class I PDZ binding motif, which mediates its interaction with scaffold and trafficking proteins, including synapse-associated protein 97 (SAP97). In GluA4, another long-tailed subunit implicated in synaptic plasticity, the PDZ motif is blocked by a single proline residue. This feature is highly conserved in vertebrates, whereas the closest invertebrate homologs of GluA4 have a canonical class I PDZ binding motif. In this work, we have examined the role of GluA4 in PDZ interactions.

Methodology/principal findings: Deletion of the carboxy-terminal proline residue of recombinant GluA4 conferred avid binding to SAP97 in cultured cells as shown by coimmunoprecipitation, whereas wild-type GluA4 did not associate with SAP97. Native GluA4 and SAP97 coimmunoprecipitated from mouse brain independently of the GluA1 subunit, supporting the possibility of in vivo PDZ interaction. To obtain evidence for or against the exposure of the PDZ motif by carboxyterminal processing of native GluA4 receptors, we generated an antibody reagent specific for proline-deleted GluA4 C-terminus. Immunoprecipitation and mass spectrometric analyses indicated that the carboxyl-terminus of native GluA4 AMPA receptors is intact and that the postulated single-residue cleavage does not occur to any significant extent.

Conclusion/significance: We conclude that native GluA4 receptors are not capable of canonical PDZ interactions and that their association with SAP97 is likely to be indirect.

Figure 1. Sequence characteristics of long-tailed AMPA receptor subunits.

(A) Alignment of the unique carboxyterminal extensions of rat long-tailed GluA1, GluA2_L and GluA4 subunits. The accession codes (SwissProt/TrEMBL) for the sequences are: GluA1, P19490; GluA4, P19493; GluA2_L, P23819-3. (B) Conservation of GluA4 C-terminal sequence in vertebrate evolution. The indicated GluA4 orthologs represent diverse vertebrate lineages: mammals (Rattus norvegicus, rat, P19490), birds (Gallus gallus, chicken, Q90858), bony fishes (Danio rerio, zebra fish, Q71E58) and amphibians (Xenopus tropicalis, western clawed frog; the sequence represents a virtual translation of Genbank EST CX366243). (C) Alignment of mammalian GluA4 carboxyterminal sequence with its closest invertebrate homologs: Mammal (rat); Beetle, (Tribolium castaneum, Red flour beetle, XP 968786); Louse, (Pediculus humanus corporis, human body louse, XP 002430327); Snail (Lymnaea stagnalis, great pond snail, CAA42683); Aplysia (Aplysia californica, California sea har, ABB03888). In all alignments, the residues conforming to the class I PDZ motif (-Thr/Ser-X-Φ; Φ denoting an amino acid residue with large aliphatic side chain, X standing for any amino acid) are highlighted in yellow. Asterisks indicate identical residues, whereas strong and weak similarities (according to Gonnet Pam250 matrix [49]) are indicated by colons and dots, respectively.

Figure 2. Deletion of proline-902 exposes a functional PDZ motif in GluA4 and confers binding to SAP97.

(A) Expression of wild-type or mutant GluA4, with or without co-expressed myc-tagged SAP97 in HEK293 cells. Upper panels show expression of all proteins; lower panels show co-immunoprecipitation of GluA4ΔP, but not full-length GluA4 with SAP97. Immunoblotting antibodies are indicated on right. (B) Transiently expressed GluA4ΔP can co- immunoprecipitate with endogenous SAP97 from HEK293 cells. Upper panel shows similar expression levels of transfected GFP-tagged constructs. Lower panel show immunoprecipitation with anti-SAP97 specific antibody. Both blots were probed with anti-GFP IgG. The extreme carboxyterminal sequences of the expressed proteins are shown below.

Figure 3. Native GluA4 AMPA receptors interact with SAP97.

Whole brain detergent extracts prepared from (A) wild-type (WT, GluA1^+/+) and (B) GluA1 knockout mice (GluA1^−/−) were subjected to immunoprecipitation. Immunoprecipitating antibodies are indicated on top; whereas antibodies used for detection of the immunocomplexes are shown on the left.

Figure 4. Analysis of AMPA receptors with an antibody specific for the exposed PDZ motif in GluA4ΔP.

(A) HEK293 cells expressing flag-tagged AMPA receptor subunits with all potential wild-type CTDs and the mutant GluA4ΔP (indicated above) were immunoblotted with the antibodies indicated to the left. Short-tailed isoforms of GluA2 and GluA4 are indicated by SH. The initial antiserum, anti-BD_L detects both A2 and A4 long tails isoforms and GluA4ΔP. After the depletion procedure and purification, the anti-ΔP IgG only recognises GluA4ΔP (lower panel). (B) Anti-ΔP IgG labels a single 100 kD band in rat cerebellar tissue; this is specifically blocked by preincubation with 13mer peptide (upper panels). Similarly anti-GluR4 IgG also labels a 100 kDa band. This labelling is blocked by pre-incubation with 14 mer peptide (lower panels). (C) Immunoprecipitation from rat cerebellar extract using independent AMPA receptor antibodies fails to bring down anti-DΔP immunoreactivity (upper panel). An alternative antibody shows GluA4 levels were highly enriched in the immunoprecipitates (lower panel).