Coenzyme nicotinamide adenine dinucleotide (NAD, including the oxidized form-- NAD+ and reduced form--NADH) and the phosphorylated form--nicotinamide adenine dinucleotide phosphate (NADP, including NADP+ and NADPH) are two of the most important biological electron carriers. Most NAD(P) dependent redox enzymes show a preference of either NADP or NAD as an electron acceptor or donor depending on their unique metabolic roles. In biocatalysis, the low enzymatic activities with unnatural coenzymes have made it difficult to replace costly NADP with economically advantageous NAD or other biomimetic coenzyme for catalysis. This is a significant challenge that must be addressed should in vitro biocatalysis be a viable option for the practical production of low-value biocommodities (i.e., biohydrogen). There is a significant need to first address the coenzyme selectivity of the NADP-dependent dehydrogenases and evolve mutated enzymes that accept biomimetic coenzymes. This is a major focus of this dissertation. Establishment of efficient screening methods to identify beneficial mutants from an enzymatic library is the most challenging task of coenzyme engineering of dehydrogenases. To fine tune the coenzyme preference of dehydrogenases to allow economical hydrogen production, we developed a double-layer Petri-dish based screening method to identify positive mutant of the Moorella thermoacetica 6PGDH (Moth6PGDH) with a more than 4,278-fold reversal of coenzyme selectivity from NADP+ to NAD+. This method was also used to screen the thermostable mutant of a highly active glucose 6-phosphate dehydrogenase from the mesophilic host Zymomonas mobilis. The resulting best mutant Mut 4-1 showed a more than 124-fold improvement of half-life times at 60oC without compromising the specific activity. The screening method was further upgraded for the coenzyme engineering of Thermotaga maritima 6PGDH (Tm6PGDH) on the biomimetic coenzyme NMN+. Through six-rounds of directed evolution and screening, the best mutant showed a more than 50-fold improvement in catalytic efficiency on NMN+ and a more than 6-fold increased hydrogen productivity rate from 6-phosphogluconate and NMN+ compared to those of wild-type enzyme. Together, these results demonstrated the effectiveness of screening methods developed in this research for coenzyme engineering of NAD(P) dependent dehydrogenase and efficient use of the less costly coenzyme in ivSB based hydrogen production. / Ph. D. / NADP and NAD are two of the most important electron carriers in cellular metabolism, and they play distinctive roles in anabolism and catabolism, respectively. Most NAD(P)-dependent dehydrogenases exhibit a strong preference for either NADP or NAD. This coenzyme preference, however, make it nearly impossible to replace the costly NADP with less costly NAD or biomimetic coenzymes in the biocatalysis application. How to engineer dehydrogenases through directed evolution and effective screening method to accept NAD or biomimetic coenzymes, is critical and the focus of this dissertation. The use of in vitro synthetic biosystem (ivSB) to produce hydrogen form starch, is one of the most important in vitro synthetic biology projects, and it depends on NADP coenzyme. With other issues in this system solved, the efficient use of dehydrogenases along with low cost and stable coenzyme is the last obstacle to hydrogen production through industrial biomanufacturing. However, the 6-phosphogluconate dehydrogenase (6PGDH), one of the rate-limiting enzymes in this biosystem, exhibits a strong coenzyme preference for NADP⁺ . For producing low-cost hydrogen, the coenzyme engineering of this dehydrogenase is urgently required. Its activity with less costly NAD or biomimetic coenzymes must be improved. The establishment of an effective screening method is the most challenging task for coenzyme engineering of dehydrogenases. In this research, we developed a Petri-dish double-layer based screening method for coenzyme engineering of thermophilic 6PGDH for activity for NAD⁺ . This screening method was also used to improve the thermostability of a highly active glucose 6-phosphate dehydrogenase from a mesophilic host, where the evolved mutant had a greatly improved thermostability without losing activity. The screening method was further upgraded to develop for coenzyme engineering on biomimetic coenzyme NMN⁺ . The engineered mutant showing a more than 50-fold increase in catalytic efficiency on NMN⁺ was used to develop the first biomimetic coenzyme dependent electron transfer chain for hydrogen production. This screening method is suitable to change the coenzyme selectivity of series of NAD(P)-dependent redox enzymes and show great potential in improving other properties, such as thermostability, substrate scope and optimal pH, of different dehydrogenases. With this method developed, we can efficiently use the low cost stable coenzyme in the biocatalysis, and break the last obstacle to industrial biomanufacturing of hydrogen production.