Site-directed mutagenesis of N-linked glycosylation sites on the gamma-aminobutyric acid type A receptor alpha 1 subunit.

Oligonucleotide-directed mutagenesis was used to mutate the two potential sites for N-linked glycosylation on the rat gamma-aminobutyric acid (GABA)A receptor alpha 1 subunit. Wild-type (WT) or mutant alpha 1 subunits [asparagine to glutamine substitutions at position 10 (alpha 1Q10), 110 (alpha 1Q110), or both 10 and 110 (alpha 1Q10/110)] were coexpressed with beta 1 and gamma 2 subunits in Xenopus oocytes. Removal of either one or both potential sites for N-linked glycosylation resulted in expression, in Xenopus oocytes, of functional GABAA receptors with pharmacological properties similar to those observed for the WT receptor. WT and mutant alpha 1 subunits were co-transfected with beta 1 and gamma 2 subunits in human embryonic kidney 293 cells. WT and mutant alpha 1 subunits expressed in 293 cells were photoaffinity labeled with [3H]flunitrazepam. Co-transfection of alpha 1WT, alpha 1Q10, or alpha 1Q110 subunits in combination with beta 1 and gamma 2 GABAA receptor subunits resulted in the labeling of single bands, with approximate molecular masses of 54, 49, and 50 kDa, respectively. The decrease in molecular mass for both the alpha 1Q10 and alpha 1Q110 mutants suggests that both consensus sequences for N-linked glycosylation are used in 293 cells. Low levels of [3H]flunitrazepam binding prevented visualization of the alpha 1Q10/110 double mutant. The 293 cells transfected with either the alpha 1Q10 or alpha 1Q110 mutant in combination with beta 1 and gamma 2 subunits expressed significantly lower levels of [3H]Ro15-1788 binding, relative to WT levels. In addition, [3H]Ro15-1788 binding was undetectable in 293 cells expressing the alpha 1Q10/110 double mutant. When transfected 293 cells were grown at 30 zero, [3H]Ro15-1788 binding to alpha 1Q10 and alpha 1Q110 GABAA receptors was restored to levels comparable to that for WT receptors. [3H]Ro15-1788 binding to alpha 1Q10/110 was not reliably detected at 30 zero. Similar results were observed using [3H]muscimol. These data suggest that intracellular processing and transport of the glycosylation-deficient GABAA receptor alpha 1 subunit is temperature sensitive. Furthermore, the observed differences between the two expression systems may be accounted for by the typically lower temperature used for maintaining microinjected Xenopus oocytes. Thus, although glycosylation is not an absolute requirement for GABAA receptor expression, it has a profound effect on the processing of at least the alpha 1 receptor and its subsequent assembly into a mature receptor.