Effects of neuron-specific depletion of cholesterol biosynthesis on dendritic spines

Abstract

Cholesterol is an important constituent of plasma membranes, including dendritic spines, where it helps to maintain their shape and size and influences transmembrane proteins. Spines undergo remodeling in response to stimulation, which is important during spine maturation and synaptic plasticity. Cholesterol in the brain is mainly synthesized by glial cells and trafficked to neurons. Most cholesterol is excreted from the brain through hydroxylation to the membrane-permeable 24-hydroxycholesterol (24-OHC), which diffuses from the brain into the bloodstream. This conversion is catalyzed by the neuron-specific enzyme cholesterol 24-hydroxylase (CYP46A1). Recent studies have postulated that CYP46A1 is activated after neuron stimulation resulting in cholesterol loss from synaptic membranes. Cholesterol synthesis requires considerable investment of cellular energy, and it is assumed that mature neurons obtain most of their cholesterol from glia cells. On the other hand, even mature neurons have cholesterol biosynthetic enzymes, and a distinct cholesterol precursor profile compared to glia, suggesting some level of cholesterol synthesis in neurons. We hypothesize that neuronal cholesterol synthesis serves a role that cannot be fully compensated for by cholesterol uptake from glia. To address this hypothesis, late-stage cholesterol biosynthetic enzymes DHCR24 and LSS were depleted specifically in neurons of hippocampal mixed neuron-glia cultures through neuron-specific RNA interference. Neurons depleted of DHCR24 showed a higher spine density and appeared more susceptible to cell death following NMDA receptor activation with glycine and chemical induction of long-term potentiation. Although we were unable to directly measure cholesterol levels in spines, my findings support a role for neuronal cholesterol synthesis in synaptic function. I propose a working model whereby neuronal cholesterol synthesis can specifically replenish cholesterol lost during synaptic activation of CYP46A1. Further studies are required to test this model.