Retinoid resistance and multifaceted impairment of retinoic acid synthesis in glioblastoma

Objectives Measuring concentrations of the differentiation‐promoting hormone retinoic acid (RA) in glioblastoma tissues would help to understand the reason why RA treatment has been inefficient in clinical trials involving brain tumor patients. Here, we apply a recently established extraction and measurement protocol to screen glioblastoma tissues for the levels of the RA precursor retinol and biologically active RA. Combining this approach with mRNA analyses of 26 tumors and 8 normal brains, we identify a multifaceted disturbance of RA synthesis in glioblastoma, involving multiple aldehyde dehydrogenase 1 family and retinol dehydrogenase enzymes. Through database studies and methylation analyses, we narrow down chromosomal deletions and aberrant promoter hypermethylation as potential mechanisms accounting for these alterations. Employing chromatin immunoprecipitation analyses and cell‐culture studies, we further show that chromatin at RA target genes is poised to RA substitution, but most glioblastoma cell cultures are completely resistant to RA treatment. This paradoxical RA response is unrelated to alternative RA signaling through the fatty acid‐binding protein 5/peroxisome proliferator‐activated receptor delta axis. Our data suggest a multifaceted disturbance of RA synthesis in glioblastoma and contribute to reconsider current RA treatment strategies. GLIA 2015;63:1850–1859 Main Points Retinoic acid synthesis is impaired in glioblastoma. Most glioblastoma cultures are resistant to retinoic acid. Current retinoic acid treatment strategies in glioblastoma need to account for retinoid resistance in tumor cells.