Implementation of the synthetic arabinose 5-phosphate dependent glycolaldehyde assimilation (SAGA) pathway in Pseudomonas putida.

Ethylene glycol (EG) is a promising substrate for biochemical product syntheses as it can be sourced sustainably from CO2 or plastic waste. The synthetic, arabinose 5-phosphate dependent glycolaldehyde assimilation (SAGA) pathway has been recently developed to improve carbon efficiency for the conversion of EG to acetyl-CoA. This pathway performs poorly in Escherichia coli, because the initial NAD-dependent EG oxidation is thermodynamically unfavorable (ΔrG'° = +23.7 kJ mol-1), resulting in slow and incomplete consumption of the substrate. Transferring the pathway to Pseudomonas putida KT2440 overcomes this barrier, as its native PQQ-dependent dehydrogenases PedE and PedH oxidize EG irreversibly. Here, the SAGA pathway was extended to enable the formation of mevalonate (MVA). Carbon tracing experiments were used to quantify the contribution of glucose metabolism and synthetic or native EG-assimilating pathways to the formation of acetyl-CoA. Deletion of the two known GA dehydrogenases (aldB-I and aldB-II) resulted in detectable SAGA pathway activity which accounted for 2.5% of acetyl-CoA production but did not prevent its formation through EG assimilation via the native pathway (10.0%). To improve GA assimilation, we identified all GA reductases in KT2440, which when deleted, eliminated GA to EG conversion but did not further improve the performance of the SAGA pathway. Limited regeneration of the GA-acceptor glyceraldehyde-3-phosphate (GA3P) was addressed by supplying glucose as a co-substrate. Deletion of the PQQ-dependent glucose dehydrogenase gcd alleviated competition of Gcd and PedE for the cofactor, permitting simultaneous uptake of EG and glucose. Optimized expression of all SAGA pathway genes raised the contribution of the synthetic pathway to 3.0% of total acetyl-CoA production. Additional removal of the major GA3P dehydrogenase gapA increased the SAGA pathway's contribution by 91% to 5.7%. When glucose was provided at reduced rates using feed beads, the share of EG-derived acetyl-CoA increased to 17.3% out of which 9.3% were channeled through the SAGA pathway. However, under these conditions the total MVA accumulation after 24 h dropped from 2.9 mM observed for the high-glucose condition to 0.3 mM.