Reprogramming the Rossmann Fold Signature Motif Creates Orthogonal Redox Biocatalysts

Biological reducing power is carried by nicotinamide adenine dinucleotide (phosphate) (NAD(P)/H), which supports cellular functions and cannot be specifically directed to engineered metabolic pathways. Nicotinamide mononucleotide (NMN(H)) has emerged as an orthogonal redox cofactor to address this challenge, yet strategies remain elusive for creating NMN(H)-specific enzymes that no longer interact with the cellular NAD(P)/H pools. Previous designs avoided perturbing the most ancient and conserved GxGxxG motif in Rossmann fold enzymes, the root cause of persistent NAD(P)/H recognition. Herein, we demonstrated that this motif, though long considered essential, is in fact mutable to yield active NMN ⁺ -specific enzymes. This is implemented on two unrelated model enzymes chosen to garner orthogonal reducing power from either the cheap electron source phosphite or glycolysis. On phosphite dehydrogenase (PTDH), variants NRC-01 and NRC-02 eliminated electron leaking to numerous NAD(P)H-dependent side reactions in whole cells and crude cell lysates while driving NMNH-dependent biotransformation with ∼240-fold higher productivity than existing catalysts. On glyceraldehyde-3-phosphate dehydrogenase (GapA), variant RSQ featured a ∼2.9×10 ⁴ -fold cofactor specificity switch to NMN ⁺ from NAD ⁺ . Combined Rosetta modeling, systematic structural alignment, and experimental results revealed that Rossmann fold reprogramming, paired with engineered structural reinforcement, may offer a general route to orthogonal redox biocatalysts.