Rational Protein Engineering of Bacterial N-demethylases to Create Biocatalysts for the Production of Methylxanthines

Abstract

Methylxanthines have a rich history as therapeutics and pharmaceuticals. However, natural dimethyl- and monomethylxanthines are difficult to produce synthetically, which has limited further exploration of these compounds in medicinal applications. A biosynthetic method for production of methylxanthines from whole cell biocatalysts is an attractive alternative. The bacterium Pseudomonas putida CBB5 contains a set of five enzymes, NdmABCDE, which are responsible for methylxanthine metabolism via N-demethylation to xanthine. The recent elucidation of the crystal structures of NdmA and NdmB, which remove the N1- and N3- methyl groups of caffeine, respectively, has opened new avenues to create biocatalysts for methylxanthine production. We have created a set of fifteen N-demethylase mutants and expressed them in E. coli BL21(DE3) as whole cell biocatalysts. The activity of each mutant was characterized for their affinity towards caffeine, theobromine, and theophylline. Two mutant enzymes in particular, labeled NdmA3 and NdmA4, both exhibited selectivity towards the N3-methyl group instead of the N1-methyl group. We also discovered that specific point mutations in NdmD resulted in the ability to tune the rate of the N-demethylase reaction. These new enzymes provide the capability of producing high-value methylxanthines, such as paraxanthine and 1-methylxanthine, through a biocatalytic route.

Highly Efficient Whole-Cell Biocatalysis for the Biosynthesis of 7-Methylxanthine and Other Xanthine Derivatives

“…However, the ability of NdmB to demethylate the N-1 methyl group of methylxanthines was weak, and the yield of 1-MX was not ideal. Since NdmB is virtually incapable of binding and demethylating caffeine, there are few suitable substrates for the synthesis of PX, but previous studies have shown that modification of NdmB allows it to recognize and demethylate caffeine to produce PX by assembling part of the active region of NdmB into NdmA, which is named NdmA4 . We tried to use BWN0673 strains to produce PX from caffeine, but the yield was not ideal, as it was only 0.13 mM.…”

“…Since NdmB is virtually incapable of binding and demethylating caffeine, there are few suitable substrates for the synthesis of PX, but previous studies have shown that modification of NdmB allows it to recognize and demethylate caffeine to produce PX by assembling part of the active region of NdmB into NdmA, which is named NdmA4. 52 We tried to use BWN0673 strains to produce PX from caffeine, but the yield was not ideal, as it was only 0.13 mM. Perhaps further modification of NdmB could improve its ability to produce PX (Figure 5B).…”

Caffeine degradation by food microorganisms

“…Mazzafera et al (1996) reported that bacterium Serratia marcescens could to degrade caffeine into paraxanthine and theobromine. Mills et al (2021) reported there are two N-demethylase enzymes found in bacterium Pseudomonas putida CBB5 i.e., N-demethylase A and N-demethylase B, that removes methyl group at N-1 and N-3. Since L. casei, L. mesenteroides, S. cerevisiae and R. oryzae were able to degrade caffeine into paraxanthine, therefore, they might be producing caffeine demethylase (N-demethylase B) to remove methyl group at N-3 of caffeine and produce paraxanthine.…”

Biocatalytic production of 7‐methylxanthine by a caffeine‐degrading Escherichia coli strain

“…This engineered bacterium harbors a mutant N ‐demethylase gene, ndmA4 (Kim et al, 2019; Mills et al, 2021), originally designed for the sole purpose of shifting the primary N ‐demethylation product of caffeine from theobromine to paraxanthine (Mills et al, 2021). NdmA4 carries three mutations around the binding pocket of the N 1 ‐demethylase NdmA (N282Q, F286L, and a swapped loop region) to mimic the N 3 ‐demethylase NdmB, which exhibits minimal activity toward caffeine due to steric hinderances (Kim et al, 2019; Mills et al, 2021). Strain MBM019 also overexpresses the ndmDP1 gene as well as two formaldehyde‐degrading genes, frmA and frmB (Supporting Information: Figure S1).…”

“…We have recently constructed E. coli strain MBM019 to N-demethylate caffeine primarily to paraxanthine (1,7dimethylxanthine), a caffeine metabolic intermediate (Mock et al, 2022) with the potential to treat and prevent Parkinson's Disease (Janitschke et al, 2021;Victorino et al, 2021). This engineered bacterium harbors a mutant N-demethylase gene, ndmA4 (Kim et al, 2019;Mills et al, 2021), originally designed for the sole purpose of shifting the primary N-demethylation product of caffeine from theobromine to paraxanthine (Mills et al, 2021). NdmA4 carries three mutations around the binding pocket of the N 1 -demethylase NdmA (N282Q, F286L, and a swapped loop region) to mimic the N 3demethylase NdmB, which exhibits minimal activity toward caffeine due to steric hinderances (Kim et al, 2019;Mills et al, 2021).…”