Two forms of CX3CL1 display differential activity and rescue cognitive deficits in CX3CL1 knockout mice.

ABSTRACT

BACKGROUND: Fractalkine (CX3CL1; FKN) is a chemokine expressed by neurons that mediates communication between neurons and microglia. By regulating microglial activity, CX3CL1 can mitigate the damaging effects of chronic microglial inflammation within the brain, a state that plays a major role in aging and neurodegeneration. CX3CL1 is present in two forms, a full-length membrane-bound form and a soluble cleaved form (sFKN), generated by a disintegrin and metalloproteinase (ADAM) 10 or 17. Levels of sFKN decrease with aging, which could lead to enhanced inflammation, deficits in synaptic remodeling, and subsequent declines in cognition. Recently, the idea that these two forms of CX3CL1 may display differential activities within the CNS has garnered increased attention, but remains unresolved. METHODS: Here, we assessed the consequences of CX3CL1 knockout (CX3CL1-/-) on cognitive behavior as well as the functional rescue with the two different forms of CX3CL1 in mice. CX3CL1-/- mice were treated with adeno-associated virus (AAV) expressing either green fluorescent protein (GFP), sFKN, or an obligate membrane-bound form of CX3CL1 (mFKN) and then subjected to behavioral testing to assess cognition and motor function. Following behavioral analysis, brains were collected and analyzed for markers of neurogenesis, or prepared for electrophysiology to measure long-term potentiation (LTP) in hippocampal slices. RESULTS: CX3CL1-/- mice showed significant deficits in cognitive tasks for long-term memory and spatial learning and memory in addition to demonstrating enhanced basal motor performance. These alterations correlated with deficits in both hippocampal neurogenesis and LTP. Treatment of CX3CL1-/- mice with AAV-sFKN partially corrected changes in both cognitive and motor function and restored neurogenesis and LTP to levels similar to wild-type animals. Treatment with AAV-mFKN partially restored spatial learning and memory in CX3CL1-/- mice, but did not rescue long-term memory, or neurogenesis. CONCLUSIONS: These results are the first to demonstrate that CX3CL1 knockout causes significant cognitive deficits that can be rescued by treatment with sFKN and only partially rescued with mFKN. This suggests that treatments that restore signaling of soluble forms of CX3CL1 may be a viable therapeutic option for aging and disease.