Thanks to its very high genome-editing efficiency, CRISPR-Cas9 technology could be a promising anti-cancer weapon. Clinical trials using CRISPR-Cas9 nuclease to ex vivo edit and alter immune cells are ongoing. However, to date, this strategy is not still applied in clinical practice to directly target cancer cells. Targeting a canonical metabolic pathway essential to the good functioning of the cells without potential escape would represent an attractive strategy. Here, the inventors propose to mimic a genetic metabolic disorder in cancer cells in order to weaken cancer cells, independently of their genomic abnormalities. Mutations affecting the heme biosynthesis pathway are responsible for porphyria, and most of them are characterized by an accumulation of toxic photoreactive porphyrins. This study aimed to mimic porphyria by using CRISPR-Cas9 to invalidate UROS, leading to porphyrin accumulation in a prostate cancer model. Prostate cancer is currently the leading cancer in men and has a high mortality rate despite therapeutic progress, with a primary tumor accessible to light. By combining light with gene therapy, the inventors obtained high efficiency in vitro and in vivo, with considerable improvement in the survival of mice. Finally, the inventors achieved the preclinical proof-of- principle of performing cancer CRISPR gene therapy.