The future of biomanufacturing is dependent on rewiring biological systems to establish an alternative approach to our current chemical industries. However, a key limitation in biomanufacturing is that desired processes must rely on the same two redox cofactors as natural metabolism, nicotinamide adenine dinucleotide (phosphate) NAD(P)+, to shuttle electrons energy. Thus, competition of resources with natural reactions within host cells is nearly unavoidable. One strategy to overcome redox cofactor resource competition is the implementation of a third, noncanonical redox cofactor, such as nicotinamide mononucleotide (NMN+), which supports specific electron delivery to desired reactions. Here, we redesign the Escherichia coli pyruvate dehydrogenase multienzyme complex (PDHc) to specially utilize NMN+ by engineering its E3 subunit (Lpd). Through rational design, we discover a cofactor promiscuous variant Lpd Penta (G182R-I186T-M206E-E205W-I271L) with an ∼2500-fold improvement in NMN+ apparent turnover number. We tailor the enzyme to exclusively use NMN+ through computational design and construct Lpd Ortho (Penta-R292E-Q317L) with a 2.4 × 105-fold cofactor specificity improvement toward NMN+ compared to the wild type. Molecular simulation allowed tracking of the cofactor's alternative binding poses that emerge as the enzyme evolves, which was crucial to precisely guide engineering. We demonstrate that the engineered NMN+-specific PDHc functions in E. coli cells to sustain the life-essential pyruvate metabolism, in an NMN+-dependent manner. These results expand the available NMN+ toolkit to include the high flux and nearly irreversible reaction of PDHc as an insulated electron source.