Identification of astrocytic factors that could play a role in myelination

Multiple sclerosis (MS) is generally considered to be an autoimmune disease; it results in areas of focal demyelination known as plaques, where astrocytes are the main constituent and appear hypertrophic and stellate with increased expression of glial fibrillary acidic protein (GFAP). Although these reactive astrocytes can be beneficial for encapsulating areas of tissue necrosis within the glial scar, the pro-inflammatory cytokines and extracellular molecules they secrete are thought to be inhibitors of remyelination. Conversely, astrocytes can secrete anti-inflammatory cytokines in their less activated state and thereby can be beneficial for remyelination; although the mechanisms involved are not clearly understood. We have developed myelinating cultures of embryonic rat mixed spinal cord cells that are plated on neurosphere-derived astrocyte monolayers, which allow us to manipulate their reactive state and determine any effects of these astrocytes on myelination. To identify any differences in gene expression profiles of several purified astrocyte phenotypes we carried out a microarray analysis using myelinating cultures plated on astrocytes on poly-L-lysine (PLL-astrocytes) or on tenascin C (TnC-astrocytes), untreated or treated with ciliary neurotrophic factor (CNTF) since it was previously shown that CNTF could stimulate myelination in these myelinating cultures and that TnC-astrocytes supported myelination less than PLL-astrocytes. The microarray analysis suggested CCL2, CCL7 and SERPINB2 as potential myelination regulators, which were validated by qRT-PCR but failed to affect axonal ensheathment in myelinating cultures. A metabolomics analysis was also carried out using untreated or CNTF-treated PLL- and TnC-astrocytes. A list of metabolites were associated with different astrocyte phenotypes and highlighted the importance of lipid metabolism, which might be crucial for stimulating myelination. These in vitro astrocyte phenotypes were compared with previously identified in vivo astrocyte reactivity markers, which suggested that PLL- and TnC-astrocytes could reflect in vivo astrocytes to some extent and that TnC-astrocytes might not be as “quiescent” as currently considered. It is possible that TnC-astrocytes suppress myelination by creating an extracellular matrix (ECM) rich environment, while CNTF dissolves the ECM and thus creates a more regenerative environment. Despite similar stimulatory effects of CNTF on mixed neural cells on PLL and TnC, CNTF appears to affect them via different mechanisms showing that distinct reactive astrocytes could react to the same stimulus in different ways as also presented in in vivo studies.