Drug repurposing is used to propose new therapeutic perspectives. Here, we introduce “Drug Upgrade”, that is, characterizing the mode of action of an old drug to generate new chemical entities and new therapeutics. We proposed a novel methodology covering target identification to pharmacology validation. As an old drug, we chose hydroxychloroquine (HCQ) for its well-documented clinical efficacy in lupus and its side effect, retinal toxicity. Using the Nematic Protein Organization Technique (NPOT®) followed by liquid chromatography-tandem mass spectrometry analyses, we identified myeloperoxidase (MPO) and alpha-crystallin β chain (CRYAB) as primary and secondary targets to HCQ from lupus patients' peripheral blood mononuclear cells (PBMCs) and isolated human retinas. Surface plasmon resonance (SPR) and enzymatic assays confirmed the interaction of HCQ with MPO and CRYAB. We synthesized INS-072 a novel analog of HCQ that increased affinity for MPO and decreased binding to CRYAB compared to HCQ. INS-072 delayed cutaneous eruption significantly compared to HCQ in the murine MRL/lpr model of spontaneous lupus and prevents immune complex vasculitis in mice. In addition, long-term HCQ treatment caused retinal toxicity in mice, unlike INS-072. Our study illustrates a method of drug development, where new applications or improvements can be explored by fully characterizing the drug's mode of action.