Microbial α-transaminases are useful as industrial biocatalysts to prepare non
proteinogenic L-amino acids from a-keto acids and an amino donor. In this study the use of the branched chain aminotransferase, ilvE, from E. coli for the industrial production of the non proteinogenic acid L-Tertiary Butyl Glycine,L-TBG 4 from Trimethylpyruvic acid, TMP 1 was investigated. L-TBG 4 is an important amino acid target particularly in the pharmaceutical industry where it is used in the synthesis of the antiretroviral drug Atazanavir, marketed under the trade name Reyataz.
It was discovered that when using equimolar concentrations of TMP 1 and L-glutamate 2 as substrates a 31% conversion of TMP 1 to L-TBG 4 is achieved at equilibrium. This low yield and corresponding low product purity was not feasible for large-scale production and prompted efforts to overcome this limitation of an otherwise attractive biocatalytic process. Methods were sought to increase the reaction yield by removal of the α-ketoglutarate 3 by-product.
Two potential equilibrium shifted processes were investigated and compared during this work. Both of these processes involved the addition of a second enzyme which removed the by-product a-ketoglutarate 3 resulting in a shift in the equilibrium towards L-TBG 4. The first of these is a process developed by Excelsyn (now AMRI) and uses aspartate 5 transaminase, aspC, from E. coli to remove the α-ketoglutarate 3. This enzyme converts α-ketoglutarate
3 and aspartate 5 to L-glutamate 2 and oxaloacetate 6. The oxaloacetate 6
decarboxylates to form pyruvate 7, which is then converted to acetolactate 8 by a third enzyme, acetolactate synthase from B. Subtilis. The acetolactate 8 then decarboxylates to form acetoin 9. At substrate concentrations of TMP 1 (0.4 M), L-glutamate 2 (0.05 M) and aspartate 5 (0.65 M) this process was found to achieve a 75 % conversion of TMP 1 to L-TBG 4 generating a L-TBG 4 yield of 40 g/L.
The second process used α-ketoglutarate dehydrogenase to remove the α-ketoglutarate 3. α- Ketoglutarate dehydrogenase is part of the E. coli Krebs cycle and converts the α-ketoglutarate 3, produced by the transamination of TMP 1 to L-TBG 4 with aspartate, to succinyl-CoA 67 using the cofactors; NAD⁺ and CoASH (Scheme 2). The products from this are then funnelled in to the E.coli Krebs cycle.
Scheme 2:The use of whole cell biocatalysts containing the chromosomal levels of ketoglutarate dehydrogenase provided proof of principle for the clean conversion of TMP 1 to L-TBG 4. The best process described to date was run at a substrate concentration of 0.1 M at pH and a cell loading of 50 g/L. After 96 hours a 90 % conversion of TMP 1 to L-TBG 4 is observed, resulting in an L-TBG 4 yield of 11 g/L. ¹³C NMR analysis of the resulting broth showed that it contained almost pure L-TBG 4, with only small quantities of L-glutamate 2 (10 mM) and acetate 11. Such a clean biotransformation is highly attractive for a potential industrial process for the production of L-TBG 4. It was found that this
biotransformation yield could be improved to 72 % at a substrate concentration of 0.3 M by increasing the levels of ketoglutarate dehydrogenase within the biotransformation. The use of acetate 11, succinate 12 and malate 13 as carbon sources for the growth of a biocatalyst overexpressing ilvE were shown to increase the levels of ketoglutarate dehydrogenase within the resulting cell. Attempts to overexpress the three genes that encode the ketoglutarate dehydrogenase in P. putida in an expression vector have so far proved unsuccessful.