(6es) Engineering a Non-Enzymatic Analog of the Glycolysis Pathway

Engineering a Non-Enzymatic Analog of the Glycolysis Pathway

Marat Orazov* and Mark E. Davis

Chemical Engineering, California Institute of Technology

 Pasadena, CA 91125

Lactic acid and its esters have been shown to have a high potential as a renewable platform chemical for the production of polymers, green solvents, and fine chemicals. Currently, fermentation is the prevailing route for making L-lactic acid. There is a need for D-lactic acid, so catalytic routes to racemic lactic acid could, after separation, provide economic pathways to either enantiomer of lactic acid.

Glycolysis is the metabolic pathway by which cells convert glucose to pyruvate, capturing the energy associated with the process for further metabolic needs. Phosphorylation reactions that alter the stability of intermediates and minimize the potential of side reactions play a crucial role in natural glycolysis. The other essential reactions involved in the glycolysis pathway are the isomerization of glucose 6-phosphate to fructose 6-phosphate by isomerases, the retro-aldol cleavage of fructose 1,6- bisphosphate to glyceraldehyde 3-phosphate and dihydroxyacetone phosphate by aldolases, and the reaction cascade that, through isomerization, oxidation, and phosphate transfer takes these phosphorylated trioses to the pyruvate, producing energy in the process. The net result of glycolysis is the conversion of glucose to pyruvate, phosphorylation of ADP to ATP, and reduction of NAD⁺to NADH. For the process to keep going, the cells must re-oxidize NADH. Lactic acid fermentation is a strategy certain cells use to sustain glycolysis for energy production in the absence of oxygen by coupling NADH oxidation with pyruvate reduction to lactate.

Synthetic glycolysis is a catalytic strategy that mimics the natural glycolysis and lactic acid fermentation pathways for the conversion of glucose to lactate, without involving phosphorylation. The key steps for a synthetic glycolysis route are: isomerization of glucose to fructose, retro-aldol cleavage of fructose into trioses, glyceraldehyde (GLA) and dihydroxyacetone (DHA), and conversion of trioses into lactates via sequential dehydration to pyruvaldehyde, rehydration, and hydride-shift isomerization. In 2010, the Davis group pioneered the use of Lewis acidic molecular sieve Sn-Beta for the isomerization of glucose to fructose via a 1, 2 intramolecular hydride shift.¹ Since then, a number of groups have extended the applications of the 1, 2 intramolecular hydride shift using similar Lewis acidic zeotypes to other sugars. Notably, the conversion of trioses to alkyl lactates has been shown to be nearly-quantitative when catalyzed by Sn-Beta at moderate temperatures (ca. 100 °C).² However, a bottleneck in the overall process still exists, as retro-aldol reactions of hexoses are not thermodynamically favored at low temperatures, and have relatively high activation energies, even with the best reported catalysts. As a result, most investigations of the retro-aldol-based production of lactates have been carried out at elevated temperatures (160 °C and higher), with long reaction times. Under these conditions, ketohexoses like fructose are prone to side-reactions that can lead to poor yields of lactates.

Recently, in our characterization of the active site of Sn-Beta, we showed that alkali-exchanged Sn-Beta can catalyze glucose epimerization to mannose via 1, 2 intramolecular carbon shift^3,4, known as the Bilik reaction. Using such catalysts with fructose as a reagent at moderate temperatures (ca. 100 °C) yielded a complex reaction mixture of sugars, including the branched sugar, hamamelose, ketohexoses (sorbose, psicose, and tagatose), and retro-aldol products (DHA and GLA). Investigation of fructose reactions catalyzed by the traditional Bilik catalysts, i.e. molybdates, yielded similar results. We found that a variety of molybdenum (VI) species (MoO₃, (NH₄)₆Mo₇O₂₄·4H₂O, H₃PMo₁₂O₄₀, and Na₂MoO₄) catalyze retro-aldol reactions of ketohexoses at moderate temperatures in both aqueous and alcoholic solvents. However, under these conditions, the equilibrium of the retro-aldol reaction favors hexoses, and aldol condensation and carbon shift products (ketohexoses and branched hexoses, respectively) begin to form after relatively low concentrations of DHA and GLA are achieved. To overcome this limitation we developed co-catalyst systems where molybdates catalyze retro-aldol reactions of ketohexoses, and microporous stannosilicates promote the formation of lactic acid or alkyl lactates from the established pool of DHA and GLA. With such catalytic systems, we have achieved high yields of lactates (ca. 70% for both methyl- and ethyl- lactate) from fructose at 100 °C.

Overall, our resulting catalytic systems, when used together to convert glucose into lactates, mirror the biological glycolysis pathway. Because we have developed a strategy to decouple the retro-aldol reaction from the active site of the 1, 2 hydride shift utilized in the isomerization reactions, we can easily adjust the ratio of active sites for each reaction, much like a cell can adjust the expression of a particular enzyme in a complicated reaction pathway to meet its metabolic demand.

References:

[1] Moliner, M.; Román-Leshkov, Y.; Davis, M. E. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 6164−6168

[2] Osmundsen, C. M.; Holm, M. S.; Dahl, S.; Taarning, E Proc. R Soc. A, 2012, 468, 2000–2016

[3] Bermejo-Deval, R.; Orazov, M.; Gounder, R.; Hwang, S.; Davis, M. E. ACS Catal. 2014, 4, 2288–2297

[4] Orazov, M.; Bermejo-Deval, R.; Gounder, R.; Hwang, S.; Davis, M. E. 2014 AIChE Annual Meeting, 2014, Atlanta, Georgia, Nov 16-21

*Corresponding author. Email: morazov@caltech.edu