Is the C3a receptor antagonist SB290157 a useful pharmacological tool?

To the Editor,

We refer to the recent article by Ahmad et al. (2020) entitled “C3a receptor antagonist therapy is protective with or without thrombolysis in murine thromboembolic stroke” published in the British Journal of Pharmacology.

The complement C3a receptor is a G‐protein coupled receptor that exerts pro‐inflammatory but also immunomodulatory roles based on the disease and cell types examined (Coulthard & Woodruff, 2015). The study by Ahmad et al. demonstrates heightened expression of plasma C3a and brain C3a receptors following experimental thromboembolic stroke in mice, suggesting that it could play a role in ischaemic stroke pathophysiology. Ahmad et al. conclude, through the use of a commercially available C3a receptor antagonist, SB290157, that inhibition of C3a receptors reduces infarct volume and improves neurological outcomes in the model.

Although this manuscript proposes a promising avenue for future clinical therapies for stroke patients, we believe it is valuable to point out to the readers that these conclusions were obtained with the use of a C3a receptor inhibitor with a dubious pharmacological record. Others (Mathieu et al., 2005) and ourselves (Proctor et al., 2004; Woodruff & Tenner, 2015) have previously detailed concerns for conclusions obtained when solely using this compound to interrogate C3a receptor function. The IUPHAR/BPS Guide to PHARMACOLOGY “Complement peptide receptors” database also comments on the issues with this compound. We acknowledge that the authors refer in the final sentence of their discussion the potential for off‐target effects of SB290157. However, we believe it is important to clearly highlight the limitations with this compound, particularly when using it as the sole tool to define C3a receptor biology.

Below we summarise the concerns with this inhibitor as a pharmacological tool:

• 1 Off‐target activity

The current study by Ahmad et al. utilised SB290157 at a relatively high dose of 20 mg kg⁻¹, which is 20‐fold higher than most other studies in mice using this compound. High doses of drugs administered in vitro or in vivo, invariably have unintended side effects which can obscure the interpretation of the experimental results. The authors used a single high dose of the C3a receptor antagonist, without explanation why this dose was chosen, and without analysis of plasma levels in their report, which would have markedly assisted interpretation of the data. In our earlier study in rats, published in the British Journal of Pharmacology (Proctor et al., 2004), we performed in vivo dose–response curves (0.1, 0.3, 1.0 mg kg⁻¹ i.v.), measured plasma pharmacokinetics, and performed receptor binding of SB290157, to clarify precisely these issues, ultimately determining that the in vivo therapeutic activity of SB290157 was NOT due to C3a receptor inhibition. This was further replicated in a myocardial ischaemia and reperfusion injury model in mice by an independent group, confirming off‐target in vivo activity for SB290157 (Busche & Stahl, 2010). Our basic query is therefore this: how do the authors know their results are solely, or even partly due to antagonism of C3a receptors?

• 2 Is SB290157 really a C3a receptor antagonist?

Despite being originally reported as a full antagonist, SB290157 was later discovered to display full “agonist” activity at C3a receptors in many cell models. Activation of human and mouse C3a receptors is seen at doses as low as 30 nM (Mathieu et al., 2005), and at higher doses, SB290157 can also partly activate C5a₂ receptors (Li, Lee, Kumar, & Woodruff, 2020). Thus, any results obtained using SB290157 must be considered as potentially resulting from either full antagonist activities, full agonist activities, or the pharmacological spectrum in between.

The study by Ahmad et al. is otherwise performed very well indeed, and we commend the authors for the detailed experiments and subject matter which are of high interest. We do feel however that, as SB290157 was the only method used to confirm a possible role for C3a receptors in the pathogenicity of thromboembolic stroke, the conclusions will need to be confirmed in further studies. We pharmacologists understand that there is no such thing as a purely selective or absolutely specific drug and that unwanted “off‐target” effects are a reality and are highly context‐dependent, especially, as noted above, with increasingly high doses. Given this, it is extremely challenging to control for every potential variable when utilising drugs. However, at least for SB290157, there are clear reasons for not relying solely on its pharmacological use.

There are a few ways out of this dilemma. Correlating the various pharmacodynamic and pharmacokinetic parameters is one way (although, admittedly such correlations will provide only circumstantial evidence). Using better probes is another way, although, unfortunately the reason why SB290157 usage persists is because there are few available C3a receptor ligands. The third way, of using C3a receptor knockout models to buttress the use of SB290157, is probably the best approach available today. SB290157 is not a “clean” probe. We and others have emphasised this over the years, and we would not like uncritical dogma to prevail, when the evidence is otherwise. In conclusion, to answer our posed title question, we believe SB290157 is a pharmacological tool of limited usefulness to study C3a receptors.