Many artificial enzymes that catalyze redox reactions have important energy, environmental, and medical applications. Native metalloenzymes use a set of redox-active amino acids and cofactors as redox centers, with a potential range between -700 and +800 mV versus standard hydrogen electrode (SHE, all reduction potentials are versus SHE). The redox potentials and the orientation of redox centers in native metalloproteins are optimal for their redox chemistry. However, the limited number and potential range of native redox centers challenge the design and optimization of novel redox chemistry in metalloenzymes. Artificial metalloenzymes use non-native redox centers and could go far beyond the natural range of redox potentials for novel redox chemistry. In addition to designing protein monomers, strategies for increasing the electron transfer rate in self-assembled protein complexes and protein-electrode or -nanomaterial interfaces will be discussed. Redox reactions in proteins occur on redox active amino acid residues (Tyr, Trp, Met, Cys, etc.) and cofactors (iron sulfur clusters, flavin, heme, etc.). The redox potential of these redox centers cover a ∼1.5 V range and is optimized for their specific functions. Despite recent progress, tuning the redox potential for amino acid residues or cofactors remains challenging. Many redox-active unnatural amino acids (UAAs) can be incorporated into protein via genetic codon expansion. Their redox potentials extend the range of physiologically relevant potentials. Indeed, installing new redox cofactors with fined-tuned redox potentials is essential for designing novel redox enzymes. By combining UAA and redox cofactor incorporation, we harnessed light energy to reduce CO2 in a fluorescent protein, mimicking photosynthetic apparatus in nature. Manipulating the position and reduction potential of redox centers inside proteins is important for optimizing the electron transfer rate and the activity of artificial enzymes. Learning from the native electron transfer complex, protein-protein interactions can be enhanced by increasing the electrostatic interaction between proteins. An artificial oxidase showed close to native enzyme activity with optimized interaction with electron transfer partner and increased electron transfer efficiency. In addition to the de novo design of protein-protein interaction, protein self-assembly methods using scaffolds, such as proliferating cell nuclear antigen, to efficiently anchor enzymes and their redox partners. The self-assembly process enhances electron transfer efficiency and enzyme activity by bringing redox centers into close proximity of each other. In addition to protein self-assembly, protein-electrode or protein-nanomaterial self-assembly can also promote efficient electron transfer from inorganic materials to enzyme active sites. Such hybrid systems combine the efficiency of enzyme reactions and the robustness of electrodes or nanomaterials, often with advantageous catalytic activities. By combining these strategies, we can not only mimic some of nature's most fascinating reactions, such as photosynthesis and aerobic respiration, but also transcend nature toward environmental, energy, and health applications.