Transcriptional and neuroprotective effects of hexokinase-2 inhibitors administered after stroke

The delayed inflammatory response induced by stroke can cause secondary injury to peri-infarct tissue. Microglia and other immune cells that mediate this injury require increased glycolytic flux for pro-inflammatory activation. These cells, unlike neurons, astrocytes, and most other cell types, utilize hexokinase-2 (HK2) rather than hexokinase-1 for glycolysis. Accordingly, HK2 inhibitors may selectively target immune cells to suppress post-stroke inflammation. Here we compared the effects of the non-selective hexokinase inhibitor 2-deoxyglucose to the HK2-selective inhibitors lonidamine and 3-bromopyruvate on gene expression changes and secondary injury after stroke.

Stroke was induced in mouse motor cortex by photothrombosis, and the hexokinase inhibitors were administered intraperitoneally beginning three hours after stroke. Gene expression in microglia and neurons was evaluated in the adjacent and more distant peri-infarct cortex using in situ spatially defined cell type-specific whole transcriptomic profiling. Tissue injury was quantified using immunohistochemical methods, and mouse functional impairment was assessed over two weeks after stroke by the cylinder, rotating beam, and skilled reaching tasks.

Peri-infarct microglia exhibited an upregulation in pro-inflammatory gene expression, while neighboring neurons showed an upregulation in cell stress/death pathways. These changes were spatially dependent, being greatest adjacent to the infarct edge. The microglial gene expression changes correlated with both morphological activation and increased CD11b expression. The three hexokinase inhibitors had similar effects on gene expression, consistent with a shared mechanism of action. They suppressed pro-inflammatory gene upregulation in peri-infarct microglia, attenuated the cell stress responses in neighboring neurons, and had minimal effect on gene expression in uninjured cortex, with the HK2-selective inhibitors having greater anti-inflammatory effects than 2-deoxyglucose. The HK2-selective inhibitors were also more effective in suppressing microglial morphology changes and CD11b expression, and in suppressing oxidative stress and neurite loss in peri-infarct neurons. Mice treated with the HK2 inhibitor 3-bromopyruvate for four days after stroke showed long-term improvement in functional outcomes.

These findings confirm an essential role for glycolysis in post-stroke, pro-inflammatory microglial activation. Selective HK2 inhibitors provide a clinically applicable approach for suppressing microglial activation and thereby improving outcomes after stroke.