Identification and targeting oxidative phosphorylation/glycolysis to overcome anti-CSF-1R therapy resistance in glioblastoma

The standard care of glioblastomas (GBM) confers limited survival benefit for patients due to the rapid tumor recurrence. Targeting tumor-associated macrophages/microglia via colony-stimulating factor 1 receptor (CSF-1R) inhibition is potentially effective in suppressing GBM recurrence. However, clinical trials of CSF-1R inhibitors failed to achieve their goal due to GBM resistance to anti-CSF-1R therapy. Here, we identified and verified key resistance mechanisms of anti-CSF-1R therapy by translatome profiling-combined analyses. To solve above problem, we have established a highly stable and refractory mouse G422^TN-GBM model, in which temozolomide (TMZ) is the most effective monotherapy but can only slightly extend animal survival. To identify effective resistance mechanism of anti-CSF1R therapy in GBM, we first apply the Translating ribosome affinity purification (TRAP) RNA-sequencing techniques in GBM tissues, which have previously used in neuroscience. TRAP-seq identified oxidative phosphorylation/glycolysis as anti-CSF1R therapy resistance mechanism, and it's combined with Cancer Therapeutics Response Portal (CTRP) identified piperlongumine (PL) or vorinostat (SAHA) as targeting drugs. PL or SAHA enhanced PLX3397 efficacy by reversing oxidative phosphorylation/glycolysis dysregulation in vitro and in vivo. The triple combination of PLX3397, TMZ, and PL/SAHA significantly improved survival in G422^TN-GBM mice. In conclusion, targeting oxidative phosphorylation/glycolysis by PL or SAHA prominently improves therapeutic efficacy of PLX3397 + TMZ in GBM, which deserves priority for clinical trials. Our study also reveals that translatome profiling is efficient for uncovering drug-resistant targets.