Role of the signal peptide and its cleavage in the expression and trafficking of the glucagon-like peptide 1 receptor

Yan Huang1, Graeme Wilkinson2, Gary Willars1

1Dept. of Cell Physiology and Pharmacology, University of Leicester, Leicester, UK,
2Biological Chemistry, Discovery Enabling Capabilities and Sciences, AstraZeneca, Alderley Park, Macclesfield, UK

The glucagon-like peptide-1 receptor (GLP-1R) is a Family B G protein-coupled receptor (GPCR), which is a target for the treatment of type 2 diabetes. As integral membrane proteins, GPCRs contain a signal sequence but whether this is an N-terminal signal peptide (SP) that is potentially cleaved or a non-cleavable signal anchor (SA) sequence within the receptor has largely been determined by predictive algorithms. Although most Family B GPCRs have a predicted N-terminal SP, this is not cleaved in at least one member. Further, the role of these sequences in receptor synthesis and trafficking is unclear. Here we have examined the cleavage and role of the SP in the synthesis and trafficking of the GLP-1R.

Human GLP-1R constructs were made containing: an N- or C-terminal HA-tag; a C-terminal EGFP tag or; an N-terminal HA-tag and C-terminal EGFP-tag. In each, the SP sequence was either present or absent. After transient expression in HEK-293 cells, expression was determined by immunoblotting and subcellular localisation of EGFP determined by confocol microscopy. Ligand affinity was determined by homologous binding ([125I]-GLP-1 7-36 amide) and cAMP generation determined by radioreceptor assay.

In constructs with a SP and C-terminal HA- or EGFP-tag (GLP-1R-HA, GLP-1R-EGFP), immunoblotting revealed two bands (47 & 64kDa and 66 & 93kDa respectively). Treatment with endoglycosidase H or peptide-N-glycosidase F indicated that lower bands were the receptor at early stages of glycosylation in the ER and upper bands were the mature, fully-glycosylated form. Biotinylation of cell surface proteins showed only the latter to be present in the cell membrane. In constructs with the SP and an N-terminal HA-tag (HA-GLP-1R, HA-GLP-1R-EGFP), expression was not detectable using an anti-HA antibody although EGFP immunoreactivity was detected in the HA-GLP-1R-EGFP construct (66 & 93kDa). EGFP-fluorescence of both GLP-1R-EGFP and HA-GLP-1R-EGFP constructs was predominately at the plasma membrane. For these, ligand affinity and the potency and magnitude of GLP-1 7-36 amide-mediated cAMP production were similar to the untagged GLP-1R. Thus, the SP is cleaved during synthesis of the GLP-1R. Removal of the SP sequence from those constructs that had shown expression (GLP-1R-HA; GLP-1R-EGFP; HA-GLP-1R-EGFP) abolished the immunoreactivity of their C-terminal epitope-tags, EGFP-fluorescence, ligand binding and receptor-mediated cAMP generation. Thus, the SP is required for synthesis of the GLP-1R. In an additional EGFP-GLP-1R-HA construct, a single band (66kDa) was detected using either anti-HA or anti-GFP antibodies indicating synthesis of a full-length receptor but a lack of cleavage of the SP. However, ligand binding and cAMP generation were absent and the single immunoreactive band was consistent with that of an immature, ER glycosylated receptor. Thus, although a full-length receptor can be synthesised without cleavage of the SP, cleavage is critical for the appropriate subsequent processing of the GLP-1R.

In summary, the SP sequence is an absolute requirement for the synthesis of the GLP-1R, whilst cleavage of the SP is critical for processing to a fully glycosylated, mature form of the receptor that can be trafficked and inserted into the plasma membrane.