D2.335 - Single-nucleotide polymorphisms in the Mas-related G protein-coupled receptor X2 alter in vitro cellular responses to specific drugs

It has been demonstrated that MRGPRX2 is activated by certain drug ligands, including neuromuscular blocking agents, fluoroquinolones, and vancomycin. However, little is still known about how this translates into drug hypersensitivity reactions (DHRs) observed in some patients exposed to these medications. One hypothesis proposes that single-nucleotide polymorphisms (SNPs) in MRGPRX2 alter receptor reactivity and thereby influence drug tolerance. Here, we investigated whether selected SNPs—previously described in patients with DHR (N62S, S313R), known but not previously tested with drugs (S325L, E164R), and predicted in silico (W248L)—affect in vitro responses to representative drug ligands.

The rat basophilic leukemia (RBL-2H3) cell line served as a functional model. Site-directed mutagenesis was used to generate MRGPRX2 variants in plasmids. Each mutant construct underwent DNA sequencing prior to transfection to confirm sequence integrity. Cells were subsequently transfected with the corresponding plasmids, and the expression of MRGPRX2 and its variants was confirmed by flow cytometry. Transfected RBL-2H3 cells were exposed to substance P, three MRGPRX2 drug agonists (ciprofloxacin, atracurium, vancomycin), and reactive buffer as a negative control. Calcium mobilization was used as the read-out.

RBL-2H3 cells showed equivalent levels of cell surface expression across all tested variants. Non-transfected cells did not respond to any of the tested ligands. Cells transfected with the variants did not respond in the presence of reactive buffer (negative control). As shown in the figure, the N62S variant exhibited a slightly weaker response to all tested agonists compared with the wild-type (WT) variant. The S313R and S325L variants displayed responses comparable to WT, with a tendency toward enhanced responses to substance P and vancomycin, respectively. The E164R variant demonstrated clear loss-of-function (LoF) characteristics, similar to the in silico–predicted W248L variant.

The findings for (1) N62S are consistent with our previous clinical observations that this variant does not explain the occurrence of DHR in affected patients; (2) the results for W248L indicate that in silico modeling may assist in assessing the pathogenicity of specific MRGPRX2 variants; and (3) the remaining variants show only partial concordance with related published data, which may reflect differences in experimental settings and underscores the need for replication studies.