with M-CSF alone, while the activation marker Msr1 was less expressed (Supplementary Fig. 2c). Interestingly, also Cd83 expression was considerably increased by the addition of TGF-β, thus resembling a homeostatic, less activated phenotype. To assess the effect of cellular activation, we treated microglial cultures with IFN-γ, TNF-α, or LPS, which are all associated with a pro-inflammatory profile, as well as IL-4, a prototypic cytokine-inducing alternative activation. While IFN-γ had no influence on Cd83 expression, both TNF-α and LPS significantly decreased Cd83 mRNA levels (Supplementary Fig. 2d). Interestingly and in sharp contrast, treatment with IL-4 strongly increased Cd83 expression levels. When we compared the expression pattern of Cd83 with that of Tmem119 or Msr1 in adult microglia cultures, we noticed that pro-inflammatory stimuli down-modulated Cd83 and Tmem119 but had the contrary effect on Msr1 (Supplementary Fig. 2e). Intriguingly, IL-4, which increases Cd83 expression, reduces microglial Tmem119 expression, while enhancing Msr1. Thus, regulation of Cd83 expression in microglia is similar to homeostatic gene transcripts but is also responsive to alternative activation. The opposing effects of TNF-α and IL-4 on Cd83 expression were also evident in cultures of neonatal microglia, which strongly suggests a context-dependent regulation of Cd83 expression in murine microglia (Supplementary
