Contributions of Angiomotin-Like-1 on Astrocytic Morphology: Potential Roles in Regulating Connexin-43-Based Astrocytic Gap Junctions, Remodeling the Actin Cytoskeleton and Influencing Cellular Polarity

Abstract

Glioblastoma is a lethal cancer that arises from support cells in the nervous system and kills around 20,000 people in the United States each year. While much is known about the highly malignant primary glioblastoma, the natural history of lower grade glioma (LGG) is less understood. While the majority of LGGs are initiated by a mutation in isocitrate dehydrogenase, the events leading to their malignant progression into a grade IV tumor are not known. Analysis of primary tumor sample data has revealed that low transcript levels of Angiomotin-like-1 (AmotL1) strongly associate with poor outcomes of patients with these cancers. Follow-up RNA-sequencing of human embryonic astrocytes with AmotL1 silencing revealed the downregulation of many transcripts that encode proteins mediating gap junctions (GJ) between astrocytes, especially connexin-43 (Cx43). Cx43 protein oligomerizes to form functional channels comprising the astrocytic GJ. AmotL1 knockdown through RNA interference decreases Cx43 transcript and protein levels while increasing its distribution to GJs. This suggests increased GJ formation and intercellular communication, as similar localization patterns are observed in differentiated astrocytes. Astrocytes with AmotL1 knockdown also display a pronounced pancake-like morphology, suggesting that the actin cytoskeleton is affected. Imaging reveals that cells with reduced AmotL1 have characteristic losses in both stress fibers and focal actin under the cell body but notable increases in cortical F-actin. Consistent with previous studies, AmotL1 may promote increases in the number and thickness of F-actin fibers. Because actin binding to related angiomotins is inhibited by phosphorylation from the LATs kinases, I define the effects of expressing wildtype AmotL1 versus mutants that mimic or prevent phosphorylation by LATs1/2. Interestingly, expression of AmotL1 S262D in combination with NEDD4-1, a ubiquitin ligase, results in a profound loss of actin stress fibers. Dependence on NEDD4-1 suggests that this phenotype is due to the induced degradation of proteins that promote F-actin, e.g. RhoA. These results directly support a model in which phosphorylated AmotL1 specifically inhibits F-actin formation as opposed to unphosphorylated AmotL1 which is known to promote stress fiber formation. Thus, in addition to regulating polarity and YAP/TAZ transcriptional co-activators, AmotL1 plays major functions in dictating cellular F-actin dynamics.