The PHD3-FOXO3 axis modulates the interferon type I response in microglia aggravating Alzheimer's disease progression
The PHD3-FOXO3 axis modulates the interferon type I response in microglia aggravating Alzheimer's disease progression
Sanchez-Garcia, M. A.; Lara-Urena, N.; March-Diaz, R.; Ortega-de San Luis, C.; Quinones-Canete, S.; Barba-Reyes, J. M.; Cabello-Rivera, D.; Munoz-Cabello, A. M.; Mora-Romero, B.; Romero-Molina, C.; Heras-Garvin, A.; Navarro, V.; Lopez-Barneo, J.; Vizuete, M.; Vitorica, J.; Munoz-Machado, A. B.; Cockman, M.; Rosales-Nieves, A. E.; Pascual, A.
AbstractMicroglia respond to Alzheimer\'s disease (AD) with a variety of transcriptional responses. However, the regulation of specific transcriptional signatures and the contribution of each individual response to disease progression is only starting to be characterized. We have previously shown that hypoxia via hypoxia inducible factor 1 (HIF1) is a strong regulator of A{beta} plaque-associated microglia (A{beta}AM). Here, we characterize the role of HIF1-mediated transcription of Egln3, encoding for PHD3, in A {beta}AM. We show that oligomeric A {beta} treatment (oA{beta}) in vitro induces the expression of Hif1a and Egln3 in microglia, which correlates with the transcriptional activation of genes involved in the interferon type I signature (IFNS) in a PHD3-dependent manner. Mechanistically, we demonstrate FOXO3 to be an important repressor of IFNS in microglia, whose abundance decreases upon A{beta} presence, and, correspondingly, both in human single-nucleus (sn) and mouse A{beta}AM transcriptomics, FOXO3 DNA binding sites define the IFNS. FOXO3 repression of the IFNS is dependent on PHD3, with our results suggesting a physical interaction between both proteins in vitro. In vivo, loss of PHD3 correlates with abrogation of the IFNS and activation of the disease-associated microglia signature (DAM) in A{beta}AM. Transcriptional changes in microglia associate with increased microglia proximity to A{beta} plaques, augmented phagocytosis of A{beta} by microglia, reduced parenchymal levels of A{beta}, and an increase in small-sized plaques. PHD3 deficiency also reduced the A{beta} plaque-associated neuropathology and rescued behavioural deficits of an AD mouse model. Finally, we also demonstrate that microglial PHD3 overexpression during development in the absence of A{beta} pathology is sufficient to induce the IFNS and behavioural alterations. Altogether, our data strongly indicate that the PHD3-FOXO3 axis controls the microglial IFNS in a cell autonomous manner, contributing to the progression of AD.