Gyrate atrophy (GA) is a rare recessive disorder characterized by progressive blindness, chorioretinal degeneration and systemic hyperornithinemia. GA is caused by point mutations in the gene encoding ornithine δ-aminotransferase (OAT), a tetrameric pyridoxal 5′-phosphate-dependent enzyme catalysing the transamination of l-ornithine and α-ketoglutarate to glutamic–γ-semialdehyde and l-glutamate in mitochondria. More than 50 OAT variants have been identified, but their molecular and cellular properties are mostly unknown. A subset of patients is responsive to pyridoxine administration, although the mechanisms underlying responsiveness have not been clarified. Herein, we studied the effects of the V332M mutation identified in pyridoxine-responsive patients. The Val332-to-Met substitution does not significantly affect the spectroscopic and kinetic properties of OAT, but during catalysis it makes the protein prone to convert into the apo-form, which undergoes unfolding and aggregation under physiological conditions. By using the CRISPR/Cas9 technology we generated a new cellular model of GA based on HEK293 cells knock-out for the OAT gene (HEK-OAT_KO). When overexpressed in HEK-OAT_KO cells, the V332M variant is present in an inactive apodimeric form, but partly shifts to the catalytically-competent holotetrameric form in the presence of exogenous PLP, thus explaining the responsiveness of these patients to pyridoxine administration. Overall, our data represent the first integrated molecular and cellular analysis of the effects of a pathogenic mutation in OAT. In addition, we validated a novel cellular model for the disease that could prove instrumental to define the molecular defect of other GA-causing variants, as well as their responsiveness to pyridoxine and other putative drugs.