Abstract
Depression is associated with peripheral inflammation, but its link with brain microglial activity remains unclear. In seven healthy males, we used repeated translocator protein-Positron Emission Tomography (TSPO-PET) dynamic scans with [(11)C]PBR28 to image brain microglial activation before and 24 h after the immune challenge interferon (IFN)-α. We also investigated the association between changes in peripheral inflammation, changes in microglial activity, and changes in mood. IFN-α administration decreased [(11)C]PBR28 PET tissue volume of distribution (Vt) across the brain (-20 ± 4%; t(6) = 4.1, p = 0.01), but after correction for radioligand free-plasma fraction there were no longer any changes (+23 ± 31%; t = 0.1, p = 0.91). IFN-α increased serum IL-6 (1826 ± 513%, t(6) = -7.5, p < 0.001), IL-7 (39 ± 12%, t(6) = -3.6, p = 0.01), IL-10 (328 ± 48%, t(6) = -12.8, p < 0.001), and IFN-γ (272 ± 64%, t(6) = -7.0, p < 0.001) at 4-6 h, and increased serum TNF-α (49 ± 7.6%, t(6) = -7.5, p < 0.001), IL-8 (39 ± 12%, t(6) = -3.5, p = 0.013), and C-reactive protein (1320 ± 459%, t(6) = -7.2, p < 0.001) at 24 h. IFN-α induced temporary mood changes and sickness symptoms after 4-6 h, measured as an increase in POMS-2 total mood score, confusion and fatigue, and a decrease in vigor and friendliness (all p ≤ 0.04). No association was found between changes in peripheral inflammation and changes in PET or mood measures. Our work suggests that brain TSPO-PET signal is highly dependent of inflammation-induced changes in ligand binding to plasma proteins. This limits its usefulness as a sensitive marker of neuroinflammation and consequently, data interpretation. Thus, our results can be interpreted as showing either that [(11)C]PBR28 is not sensitive enough under these conditions, or that there is simply no microglial activation in this model.