Knockout controls and the specificity of cannabinoid CB₂ receptor antibodies

In Atwood and Mackie's (2010) excellent review of the status of the cannabinoid CB₂ receptor, evidence for CB₂ receptor expression in the cerebellum obtained with immunohistochemistry is discussed. This includes a discussion of work published by my colleagues and myself (Ashton et al., 2006). In addition to the studies reviewed, Suarez et al. (2008; 2009;) have also published results for cannabinoid receptor immunolabelling in the rodent brain, including the cerebellum. The second of these studies (Suarez et al., 2009) is notable because it uses cannabinoid receptor knockout mice to test for antibody specificity. These studies are particularly relevant to the issues discussed by Atwood and Mackie, who critically discussed the use of controls in immunohistochemistry, including the use of knockout mice. Suarez et al. (2008) reported widespread neuronal CB₂ receptor immunolabelling in the cerebellum, in contrast to our results (Ashton et al., 2006). Suarez et al. (2009) then reported widespread neuronal CB₂ receptor immunolabelling in the hippocampus, using the same antibody. In this later paper, knockout controls for both CB₁ and CB₂ receptor antibodies were used. However, close investigation of the published images reveals that the CB₂ receptor antibody gives a very similar – but fainter – staining pattern in the knockout mouse as in the wild-type mouse. In contrast, staining patterns are very different between knockout and wild-type strains for CB₁ receptor immunolabelling. Therefore, these studies illustrate a rule that convincing knockout controls on antibody specificity should show the redistribution rather than the reduction of immunolabelling. These investigations therefore provide an informative addition to Atwood and Mackie's comprehensive review.