Efficient mining of natural NADH-utilizing dehydrogenases enables systematic cofactor engineering of lysine synthesis...

Increasing the availability of NADPH is commonly used to improve lysine production by Corynebacterium glutamicum since 4 mol of NADPH are required for the synthesis of 1 mol of lysine. Alternatively, engineering of enzymes in lysine synthesis pathway to utilize NADH directly can also be explored for cofactor balance during lysine overproduction. To achieve such a goal, enzyme mining was used in this study to quickly identify a full set of NADH-utilizing dehydrogenases, namely aspartate dehydrogenase from Pseudomonas aeruginosa (PaASPDH), aspartate-semialdehyde dehydrogenase from Tistrella mobilis (TmASADH), dihydrodipicolinate reductase from Escherichia coli (EcDHDPR), and diaminopimelate dehydrogenase from Pseudothermotoga thermarum (PtDAPDH). This allowed us to systematically perturb cofactor utilization of lysine synthesis pathway of C. glutamicum for the first time. Individual overexpression of PaASPDH, TmASADH, EcDHDPR, and PtDAPDH in C. glutamicum LC298, a basic lysine producer, increased the production of lysine by 30.7%, 32.4%, 17.4%, and 36.8%, respectively. Combinatorial replacement of NADPH-dependent dehydrogenases in C. glutamicum ATCC 21543, a lysine hyperproducer, also resulted in significantly improved lysine production. The highest increase of lysine production (30.7%) was observed for a triple-mutant strain (27.7 g/L, 0.35 g/g glucose) expressing PaASPDH, TmASADH, and EcDHDPR. A quadruple-mutant strain expressing all of the four NADH-utilizing enzymes allowed high lysine production (24.1 g/L, 0.30 g/g glucose) almost independent of the oxidative pentose phosphate pathway. Collectively, our results demonstrated that a combination of enzyme mining and cofactor engineering was a highly efficient approach to improve lysine production. Similar strategies can be applied for the production of other amino acids or their derivatives.