The "C3aR Antagonist" SB290157 is a Partial C5aR2 Agonist

Abstract

Innate immune complement activation generates the C3 and C5 protein cleavage products C3a and C5a, defined classically as anaphylatoxins. C3a activates C3aR, while C5a activates two receptors (C5aR1 and C5aR2) to exert their immunomodulatory activities. The non-peptide compound, SB290157, was originally reported in 2001 as the first C3aR antagonist. In 2005, the first report on the non-selective nature of SB290157 was published, where the compound exerted clear agonistic, not antagonistic, activity in variety of cells. Other studies also documented the non-selective activities of this drug in vivo. These findings severely hamper data interpretation regarding C3aR when using this compound. Unfortunately, given the dearth of C3aR inhibitors, SB290157 still remains widely used to explore C3aR biology (>70 publications to date). Given these issues, in the present study we aimed to further explore SB290157's pharmacological selectivity by screening the drug against three human anaphylatoxin receptors, C3aR, C5aR1 and C5aR2, using cell models. We identified that SB290157 exerts partial agonist activity at C5aR2 by mediating β-arrestin recruitment at higher compound doses. This translated to a functional outcome in both human and mouse primary macrophages, where SB290157 significantly dampened C5a-induced ERK signaling. We also confirmed that SB290157 acts as a potent agonist at human C3aR in transfected cells, but as an antagonist in primary human macrophages. Our results therefore provide even more caution against using SB290157 as a research tool to explore C3aR function. Given the reported immunomodulatory and anti-inflammatory activities of C5aR2 agonism, any function observed with SB290157 could be due to these off-target activities.

Keywords: C3a anaphylatoxin; C3aR; C5a anaphylatoxin; C5aR1; C5aR2; SB290157; complement; complement C3a.

FIGURE 1

SB290157 potently activates C3aR-mediated ERK signaling in transfected CHO cells. (A) Agonism testing of SB290157 on CHO-C3aR cells. Serum-starved CHO-C3aR cells (50,000/well) were stimulated with respective concentrations of SB290157 or purified human C3a for 10 min before being lysed. (B) Agonism testing of SB290157 on CHO-C5aR1 cells. Serum-starved CHO-C5aR1 cells (50,000/well) were stimulated with the respective ligands at the indicated concentrations for 10 min before being lysed. (C) Antagonism testing of SB290157 on CHO-C5aR1 cells. Serum-starved CHO-C5aR1 cells (50,000/well) were pre-treated with SB290157 (10 µM) or vehicle (0.1% DMSO) for 30 min before being stimulated with 0.3 nM of C5a for 10 min and then lysed. The phospho-ERK1/2 content in the cell lysate was measured and normalised to the maximum C3a-induced (for A) or 0.3 nM C5a-induced (for B, C) levels before being combined. Data represent mean ± S.E.M. of triplicate measurements from 3 to 6 independent experiments (n = 3–6).

FIGURE 2

SB290157 induces C5aR2-mediated β-arrestin 2 recruitment in transfected HEK293 cells. HEK293 cells were transiently transfected using C5aR2-Venus and β-arrestin 2-Rluc8 BRET pairs for 24 h and seeded (100,000/well) overnight. Filtered light emissions between 460–485 nm (Rluc8) and 520–545 nm (Venus) were continually monitored for 90 min with SB290157, P32, C5a or vehicle added at the 0 min time point. Data represent (A) the time course of ligand-induced BRET ratios (Venus/Rluc8 emission ratio) caused by the respective ligands at the indicated concentrations, (B) the corresponding concentration-response curves for SB290157 and P32 at 40 min post ligand addition, and (C) the ligand-induced BRET ratios (efficacy) of SB290157, P32 or C5a at 40 min post ligand addition. Data represent the mean $\pm$ S.E.M. of triplicate measurements from 3 to 5 independent experiments (n = 3–5).

FIGURE 3

SB290157 potently inhibits C3aR-mediated ERK signaling in human monocyte-derived macrophages. (A) Antagonism testing of SB290157 on HMDMs. Serum-starved HMDMs (50,000/well) were pre-treated with various doses of SB290157 for 30 min before being stimulated with purified human C3a (5 nM) for 10 min and then lysed. (B) Agonism testing of SB290157 on HMDMs. HMDMs were stimulated with respective doses SB290157 for 10 min and then lysed. The phospho-ERK1/2 content in the cell lysate was measured and normalised to the C3a-induced levels before being combined. Data represent mean ± S.E.M. of triplicate measurements using cells from 2-3 independent donors (n = 2–3).

FIGURE 4

C5aR2 activation dampens C5a-induced ERK signaling in human monocyte-derived macrophages and mouse bone marrow-derive macrophages. HMDMs (A) or BMDMs (B,C) were serum-starved overnight and pre-treated with SB290157 (50 µM) or P32 (100 µM) and the corresponding solvent-only control (Vehicle) for 30 min, prior to stimulation with recombinant human C5a (10 min, for A) or recombinant mouse C5a (5 min, for B and C). The phospho-ERK1/2 content in the cell lysate was measured and normalised to the maximum vehicle-treated C5a-induced levels before being combined. Data represent mean ± S.E.M. of triplicate measurements using cells from 3-6 independent donors (n = 3–6) for HMDMs, or 4 mice (n = 4) for BMDMs. Two-way ANOVA with Sidak's post hoc test. *p < 0.05, **p < 0.01, ****p < 0.0001, P32 or SB290157 pre-treated vs. control-treated cells stimulated by respective concentrations of C5a.