DNA encoding the N-terminal 415 residues of the human thyrotrophin receptor (predicted to code for the large extracellular region) was introduced into Chinese hamster ovary (CHO) cells using the glutamine synthetase/cytomegalovirus amplifiable expression system, and into E. coli using the pGEX-3X expression vector. Substantial quantities of insoluble fusion protein product resulted from bacterial expression; by Western blot analysis, this was shown to be reactive with anti-receptor antibodies raised against a peptide corresponding to residues 313–330. Immunoreactivity was not retained by the solubilized protein. In eukaryotic expression, several successful CHO transfectants were observed and one (ExG2) was characterized thoroughly. Using agarose-bound Concanavalin A, a glycoprotein with an M r of approximately 60 000 was detected in a detergent extract of metabolically labelled ExG2 cells, agreeing with the predicted molecular size of 45 000, plus carbohydrate. The same protein could also be detected by immunoprecipitation using the experimental anti-peptide antisera and also sera from patients with Graves' disease. The protein was immunoreactive in Western blot analyses of ExG2 cells using the experimental antisera but not the pathological sera, supporting the view that linear sequences are not sufficient for autoantibody binding. These are the first studies in which visualization of eukaryotically expressed recombinant receptor by such immunological techniques has been possible, presumably because of the higher expression of the glutamine synthetase system.
Surprisingly, the recombinant protein was retained within the cells rather than being secreted. The recombinant protein was very effective at absorbing the adenylate cyclase-stimulating activity of the sera from patients with Graves' disease, but not that of thyrotrophin. This suggests that the large N-terminal extracellular region contains epitopes for stimulatory autoantibodies, but that high affinity thyrotrophin binding requires additional components.