programmed cell death protein 1 (PD-1)/ programmed cell death ligand 1 (PD-L1) axis regulation has presented impressive therapeutic efficacy against multiple types of cancers by enhancing T cell infiltration and revitalizing exhausted cytotoxic T cells, which finally shifted the paradigm of cancer management. [1] However, although the current clinically approved and used anti-PD-L1 monoclonal antibodies are effective in the treatment of many cancer types, their widespread usage is still limited and risky owing to immune-related adverse effects, insufficient tumor specificity, and high cost. [2] Moreover, most anti-PD-L1 monoclonal antibodies could only disrupt the PD-1/ PD-L1 axis recognition on the surfaces of T cells and cancer cells, without taking into account the immune-regulation capacity of the cytoplasm or the nucleus distributed PD-L1 protein, which also seriously limited the efficacy of radioimmunotherapy and chemoimmunotherapy when combined with anti-PD-L1 monoclonal antibodies since anti-PD-L1 monoclonal antibodies could not influence the cytoplasm or the nucleus distributed PD-L1 protein and its mediated increased DNA damage repair. [3] Thus, there is still a need to better understand the mechanisms of regulating the PD-L1/PD-1 axis to augment approaches that target this pathway.Currently, the role of the lysosome, endoplasmic reticulum, or dictyosome in the transcription and translation of programmed cell death ligand 1 (PD-L1) is well revealed, but the role and function of mitochondria in the PD-L1 expression in tumors is still not fully researched, making it hard to offer a novel PD-L1 regulation strategy. In this research, it is newly revealed that mitochondria oxidative phosphorylation (OXPHOS) depression can be used as an effective PD-L1 down-regulation method. To offer an ideal and higheffective tumor mitochondria-targeted OXPHOS depression nanosystem, IR-LND is prepared by conjugating mitochondria-targeted heptamethine cyanine dye IR-68 with mitochondrial complexes I and II depression agent lonidamine (LND), which then further self-assembled with albumin (Alb) to form IR-LND@Alb nanoparticles. By doing this, PD-L1 expression in tumors is selectively and effectively depressed by IR-LND@Alb nanoparticles. As expected, the anti-tumor efficacy of such a PD-L1 depression strategy is superior to conventional anti-PD-L1 monoclonal antibodies. Interestingly, IR-LND can also be served as a novel ideal promising photodynamic therapy (PDT) drug with self-oxygen and self-PD-L1 regulation capacity. All in all, this tumorselective metabolic reprogramming platform to reactivate immunotherapy and sensitize for PDT effect, would open a new window for mitochondrial immunotherapy for cancer patients.