Figure 1. (A) GABA_B(1b)ECD sequence, (B) sequence alignment, (C) model of the GABA_B(1)ECD and (D) designed constructs.

The natural signal sequence of GABA_B(1b) from R. norvegicus spans amino acids 1–29. Proline residues belonging to the N-terminal domain characteristic for the variant 1b appear in bold italic. The sequence encoded by exons 6–14 is underlined. The five potential N-glycosylation sites are indicated in boldface. (B) The amino acid sequence of the mGluR1 (1ewt) is derived from the PDB number 1EWT; GBR1b_ED is the sequence of the GABA_B(1b)ECD. Secondary structural elements observed (in the crystal structure 1EWT) or predicted (for GABA_B(1)ECD) are indicated with A for α-helices and B for β-strands. Identical residues in the two sequences appear in boldface. The following regions were omitted from the alignment used to generate the model: in mGluR1, residues 132–153 (not present in the PDB number 1EWT) and 348–405; in the GABA_B(1)ECD sequence, residues 316–347. (C) The model was ramp-coloured from blue to green and yellow to red, going from the N- to the C-terminus. Arrows indicate the generated GABA_B(1b) receptor deletion mutant proteins. The panel shows the plot obtained from the Verify3D algorithm using the co-ordinates of the modelled GABA_B(1)ECD. The x-axis indicates the residue numbers and y-axis gives the averaged 3D–1D scores. (D) The top part of the panel shows the N-terminal deletion mutants, all containing the natural signal sequence of the GABA_B(1b); the middle part of the panel shows the C-terminal deletions of the ECD. Details on construct generation are given in the Materials and methods section. The bottom part of the panel shows the sequence of the N-glycosylation mutants. The haemagglutinin tag is underlined. Asparagine residues, which can potentially be glycosylated in vivo, appear in boldface. All constructs are fused C-terminally through a short linker to the EGFP.