Rat basophilic leukemia cells (RBL-2H3, #CRL-2256) were purchased directly from the American Type Culture Collection (ATCC, Manassas, VA, USA). Upon receipt, the cells were cultured as previously described [19], expanded, and cryopreserved in Bambanker freezing medium (NIPPON Genetics EUROPE GmbH). Experiments were performed using cells recovered from these early stocks and maintained for no more than 20 passages. Cells were routinely tested negative for mycoplasma contamination. Culture quality and phenotypic consistency were confirmed by morphological feature analysis and the expected $\beta$-hexosaminidase secretion response.

To generate RBL-2H3 cells expressing the wild-type MRGPRX2 receptor, we used the pReceiver-M06 plasmid encoding hemagglutinin-tagged human MRGPRX2 (#EX-T3417-M06, GeneCopoeia, Rockville, MD, USA). MRGPRX2 variants were generated using the QuikChange II Site-Directed Mutagenesis Kit (#200523, Agilent Technologies, Inc., Santa Clara, CA, USA) according to the manufacturer’s instructions. Briefly, mutagenic primers were designed using the Agilent web-based primer design tool [20]. The forward and reverse primers used to generate each variant were as follows:

N62S: Forward: 5^′-CAGAGAAGGCGCTCCTGCGCATGCGGAA-3^′

Reverse: 5^′-TTCCGCATGCGCAGGAGCGCCTTCTCTG-3^′

E164R: Forward: 5^′-AAGCCACAGAACTTCCCTCGCAAGATGCTCAGCAGTAG-3^′

Reverse: 5^′-CTACTGCTGAGCATCTTGCGAGGGAAGTTCTGTGGCTT-3^′

W248L: Forward: 5^′-GAATCCTTCCAGATCAATAATATTAGGAACCACTGAATGCCA-3^′

Reverse: 5^′-TGGCATTCAGTGGTTCCTAATATTATTGATCTGGAAGGATTC-3^′

S313R: Forward: 5^′-GACGGAAGCATCCTTCCCTGTGATCCACCTCAG-3^′

Reverse: 5^′-CTGAGGTGGATCACAGGGAAGGATGCTTCCGTC-3^′

S325L: Forward: 5^′-CCAGACTGCTTCTCAACATCTCCGGGGTG-3^′

Reverse: 5^′-CACCCCGGAGATGTTGAGAAGCAGTCTGG-3^′

Polymerase Chain Reactions (PCR) were performed using 25 ng of dsDNA template with the extension conditions described above (16 cycles, 8 min per cycle). The remaining PCR parameters were applied according to the manufacturer’s instructions. Following amplification, the parental methylated plasmid DNA was digested with DpnI, and the reaction mixture was transformed into chemically competent E. coli cells (XL1-Blue Supercompetent Cells). Bacterial colonies were selected on LB agar (#L3022, Sigma-Aldrich, St. Louis, MO, USA; #AGR001, BioShop, Burlington, ON, Canada) supplemented with ampicillin (#AMP201, BioShop, Burlington, ON, Canada). Positive clones were expanded, and plasmid DNA was purified using the Plasmid Midi AX Endotoxin-Free kit (#092EF-10, A&A Biotechnology, Gdańsk, Poland). Prior to transfection, the correct nucleotide sequences of all constructs were confirmed by Sanger sequencing (Genomed, Warsaw, Poland).

RBL-2H3 cells (0.8 $\times$ 10⁶) were transfected with 4 µg of the pReceiver-M06 plasmid encoding the wild-type MRGPRX2 receptor or its variants using the SF Cell Line 4D-Nucleofector X Kit (#V4XC-2012, Lonza, Gaithersburg, MD, USA), as previously described [19]. Cells were used within 24 h after transfection. Before each Ca²⁺ mobilization assay, the surface expression of MRGPRX2 or its variants on transfected RBL-2H3 cells was routinely assessed by flow cytometry [19] and analyzed using FlowJo software 10.10.0 (Waters Biosciences, Ashland, OR, USA).

The day after transfection, the cells were plated at a density of 0.05 $\times$ 10⁶ cells per well in Corning® 96-well flat clear-bottom black polystyrene TC-treated microplates (#3603, Corning, Glendale, AZ, USA) in culture medium. The cells were then incubated for 3.5 h at 37 °C in 5% CO₂. The cells were subsequently washed twice with HBSS buffer (#14025-092, Gibco, Thermo Fisher Scientific, Waltham, MA, USA), after which 100 µL/well of Fluo-4 NW dye-loading solution (#F36206, Fluo-4 NW Calcium Assay Kit, Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA) was added, and the cells were incubated at 37 °C for 30 min. The plate was then placed in a CLARIOstar Plus plate reader (BMG LabTech, Ortenberg, Germany) preheated to 37 °C and equilibrated for an additional 15 min.

Fluorescence was measured at an excitation wavelength of 485 nm and an emission wavelength of 530 nm. The measurements were recorded for 60 s to obtain baseline values (F₀; 0–59 s) and then for 200 s after stimulus addition (F; 60–260 s), with readings collected every 4 s. Data were normalized as $\mathrm{\Delta }$F/F₀ = (F – F₀)/F₀ using the mean of the first eight baseline readings as F₀, and the area under the curve (AUC) of $\mathrm{\Delta }$F/F₀ was calculated for 60–260 s ($AU{C}_{60-260}={\int }_{60}^{260}\frac{\mathrm{\Delta }F}{F_0}𝑑t$).

Cells were stimulated with the following reagents: SP (#1156, Tocris Bioscience, Bio-Techne, Bristol, UK; used as a positive control for MRGPRX2-dependent stimulation), vancomycin (Vancomycin-MIP, MiP Pharma, Gdańsk, Poland), atracurium (atracurium besilate, Tracrium, Aspen Pharma, Durban, South Africa), and ciprofloxacin (Ciprofloxacin Kabi, Fresenius Kabi, Warsaw, Poland). Immediately before stimulation, all reagents were prepared in assay buffer (from the Fluo-4 NW Calcium Assay Kit), which was also used as a negative control. The tested ligands were used at the following final concentrations: SP, 25 µM; vancomycin, 1 mg/mL; atracurium, 500 µg/mL; and ciprofloxacin, 250 µg/mL. The concentrations were selected based on previous publications, choosing those reported to elicit the maximal response [16, 19]. The numbers of biological replicates for each condition were as follows: atracurium (n = 16, 7, 10, 10, 7, 10 for WT, N62S, E164R, W248L, S313R, and S325L, respectively); ciprofloxacin (n = 22, 9, 10, 10, 7, 12, respectively); vancomycin (n = 19, 7, 8, 11, 10, 8, respectively); negative control (n = 16, 7, 4, 4, 7, 5, respectively); and substance P (n = 18, 6, 6, 5, 6, 7, respectively).

Data were analyzed using Prism software 8.0.1 (GraphPad Software, Inc., La Jolla, CA, USA). Data distribution was assessed with the Shapiro–Wilk test. Data are presented as medians and interquartile ranges (IQRs). Multiple comparisons were conducted using the Kruskal–Wallis test followed by Dunn’s post hoc test. Differences were considered statistically significant at p 0.05.

Our results show that the S313R variant displays a response profile largely comparable to WT across all tested ligands, with a tendency toward enhanced activation upon SP stimulation (Fig. 2). This variant was first described by Quan et al. [16] in a patient with vancomycin-induced immediate hypersensitivity. In that study, functional analyses were performed in HEK293LTV cells using Ca²⁺ mobilization and inositol 1-phosphate production assays as readouts. Consistent with our findings, S313R-transfected cells showed stronger responses to SP and weaker responses to ciprofloxacin compared with WT. The authors also reported enhanced activation by vancomycin in this variant [16]. To date, however, no further experimental validation of drug responses in S313R-expressing cells has been reported by other research groups.

More extensive data are available for the N62S variant. It has been reported in patients with perioperative anaphylaxis to NMBAs and in individuals with immediate DHRs to fluoroquinolones, although it also occurs at a comparable frequency in the general population [9, 16]. In our experiments, N62S-transfected cells displayed responses similar to or slightly numerically lower than those observed in WT-transfected cells, but not reaching statistical significance. Yang et al. [15], using HEK293-G$\alpha$15 cells, likewise did not observe increased responsiveness to SP associated with this variant. Quan et al. [16] reported a trend toward enhanced activation of N62S-expressing HEK293LTV cells following stimulation with SP, vancomycin, and ciprofloxacin. The variant has also been linked to increased severity of chronic spontaneous urticaria, with RBL-2H3 cells expressing N62S showing enhanced responses to SP and compound 48/80 in assays measuring degranulation and calcium mobilization [23]. The opposite, protective clinical role has been attributed to the N62S variant in ulcerative colitis with the authors suggesting that this effect may result from increased $\beta$-arrestin recruitment, leading to enhanced receptor desensitization and reduced susceptibility of intestinal mast cells to agonist-induced activation [24].

Functional characterization of MRGPRX2 variants has typically relied on classical agonists, such as SP or compound 48/80. Drug ligands were evaluated only in a limited number of studies, although, as noted above, the same variant may exhibit ligand-specific responses. Therefore, we extended our analysis to variants previously characterized as gain-of-function (S325L) [17] and loss-of-function (E164R) [18].

S325L is a missense mutation located in the C-terminal region of MRGPRX2, which is involved in downstream signaling processes, including G-protein phosphorylation and $\beta$-arrestin recruitment [17]. This variant has been reported to enhance SP-induced activation of RBL-2H3 cells [17]. In our study, S325L showed a modest tendency toward enhanced activation following SP stimulation compared with WT, whereas responses to the remaining ligands were comparable to those of WT. The E164R variant was previously shown by Reddy et al. [18] to impair MRGPRX2 activation in HEK-293 cells stimulated with SP or compound 48/80. In line with these findings, E164R-transfected cells in our experiments displayed markedly reduced responses to all tested ligands.

W248 is a pocket residue located on the extracellular segment of transmembrane helix 6 (TM6) of MRGPRX2. In our previous computational study [19], we demonstrated that alanine substitution at this position leads to reduced binding free energies of the receptor for relevant drug molecules—specifically, fluoroquinolones and laudanosine (a metabolite of atracurium)—relative to the wild-type receptor. This diminished binding affinity was attributed to the loss of $\pi$-stacking interactions between the aromatic structures of these ligands and the indole motif of W248.

Although no direct contact is observed between W248 and the neuropeptide SP in the MRGPRX2 complex (PDB ID: 7VDM) [25], our experimental data demonstrate that the W248L substitution renders the receptor completely unresponsive to activation by this ligand, which served as a positive control in our assays. We hypothesize that this loss of function does not result from disruption of a direct protein–ligand interaction, but rather from a role of W248 in maintaining the receptor’s tertiary helix bundle structure. Inspection of available MRGPRX2 structural data [25] reveals a strong hydrogen bond between the N–H group of W248 and the carboxylate moiety of D184 (Fig. 3), a transmembrane helix 5 (TM5) residue previously shown to be critical for ligand binding through electrostatic interactions with the positively charged nitrogen atom of MRGPRX2 ligands [19, 25]. Notably, this hydrogen bond also constitutes the sole polar contact between TM5 and TM6 in the entire protein, highlighting its potential importance in stabilizing the relative orientation of these two helices, preserving the native conformation of the binding pocket, and thereby maintaining the functional integrity of the receptor.

Fig. 3.

Detailed view of the interaction between the MRGPRX2 agonist (R)-ZINC-3573 and residue W248 within the receptor binding pocket (PDB ID: 7S8N). In addition to direct contact with the ligand, W248 contributes to stabilization of the binding pocket by forming a hydrogen bond with D184.