Oxyfunctionalization of pyridine derivatives using whole cells of Burkholderia sp. MAK1

Pyridinols and pyridinamines are important intermediates with many applications in chemical industry. The pyridine derivatives are in great demand as synthons for pharmaceutical products. Moreover, pyridines are used either as biologically active substances or as building blocks for polymers with unique physical properties. Application of enzymes or whole cells is an attractive strategy for preparation of hydroxylated pyridines since the methods for chemical synthesis of pyridinols, particularly aminopyridinols, are usually limited or inefficient. Burkholderia sp. MAK1 (DSM102049), capable of using pyridin-2-ol as the sole carbon and energy source, was isolated from soil. Whole cells of Burkholderia sp. MAK1 were confirmed to possess a good ability to convert different pyridin-2-amines and pyridin-2-ones into their 5-hydroxy derivatives. Moreover, several methylpyridines as well as methylated pyrazines were converted to appropriate N-oxides. In conclusion, regioselective oxyfunctionalization of pyridine derivatives using whole cells of Burkholderia sp. MAK1 is a promising method for the preparation of various pyridin-5-ols and pyridin-N-oxides.

Several bacteria belonging to the genera Arthrobacter, Achromobacter, Rhodococcus, and Nocardia are able to grow on pyridin-2-ol [32][33][34][35] . The first common steps in the microbial metabolism of pyridin-2-ol involve the hydroxylation of the ring yielding di-or trihydroxypyridine intermediates 32,35,36 that are promising synthons for the preparation of substituted pyridines.
In this study, the oxyfunctionalization of the pyridine ring by whole bacterial cells was investigated. The pyridin-2-ol-degrading Burkholderia sp. MAK1 was found to be an efficient biocatalyst for the hydroxylation of various pyridin-2-ols and pyridin-2-amines. Moreover, Burkholderia sp. MAK1 was capable of oxidising several N-heterocyclic ring systems to corresponding N-oxides.
Selection of pyridine derivatives as substrates for hydroxylation with Burkholderia sp. As we found out that Burkholderia sp. MAK1 consumes pyridine-2-ol via pyridine-2,5-diol by supposedly pyridine-2-ol inducible pyridin-2-ol 5-monooxygenase we wanted to test whether Burkholderia sp. MAK1 is capable of hydroxylating other pyridine derivatives. In this study, more than 100 of pyridine, pyrimidine, and pyrazine derivatives were screened for the hydroxylation using Burkholderia sp. MAK1 as a whole-cell biocatalyst (Supplementary information Table S-2). The pyridin-2-ol-induced Burkholderia sp. MAK1 cells were incubated with a potential substrate as described in the Methods section. The progress of the reaction was followed by HPLC-MS. The efficiency of conversion of several compounds by whole cells of Burkholderia sp. MAK1 is presented as Supplementary information Table S-3.
It is worth mentioning that induction of Burkholderia sp. MAK1 hydroxylation activity was observed only in the presence of pyridin-2-ol. Several other tested compounds (pyridine, pyridine-2,5-diol, pyridin-2-amine) were not able to trigger the induction. Also no hydroxylation occurred when cells were cultivated with other sole carbon source (glucose or succinate) instead of pyridin-2-ol.
Optimization of cultivation and reaction conditions. Burkholderia sp. MAK1 grew poorly in rich nutrient medium, but the growth was observed in mineral medium (EFA or Koser) with pyridin-2-ol as a sole carbon source. The growth reached its peak after 40 h of incubation in EFA medium (OD 600 = 0.4). The optimal temperature for cultivation of Burkholderia sp. MAK1 appeared to be 30 °C. At higher tested temperature (37 °C), Burkholderia sp. MAK1 cells were not able to grow. Although bacterial growth was observed at 25 °C it was rather slow compared to 30 °C. The effect of temperature on Burkholderia sp. MAK1-mediated synthesis of hydroxylated pyridine derivatives was also investigated (Fig. 1). For this experiment 4-chloropyridin-2-amine was selected due to its great conversion percentage and definite product (Table 1). During the first hour of the experiment, the bioconversion of 4-chloropyridin-2-amine was most rapid at 30 °C and 35 °C with 6-amino-4-chloro-pyridin-3-ol production rate of 7 mg (g biomass)-1 h-1 and 7.4 mg (g biomass)-1 h-1, respectively. Higher temperatures (40-45 °C) were found to be unfavorable for the synthesis, probably because of the inactivation of the biocatalyst. The conversion reached near completion (~97%) after six hours at 30 °C.

Biotransformation of various pyridin-2-ols by Burkholderia sp. MAK1 cells. The study of
N-alkylpyridine transformation revealed that 1-methyl-, 1-ethyl-and 1-propylpyridin-2-ol were transformed to the final dihydroxy products by Burkholderia sp. MAK1 cells. In the chromatogram of 1-ethylpyridin-2-ol bioconversion, two dominant peaks A and B were detected (Supplementary information Fig. S-II) corresponding to the newly formed compound and the residual substrate, respectively. The absorption maximum of the product, compared with that of the substrate, shifted to longer wavelengths (~30 nm), which is characteristic of compounds with additional hydroxy group. Also, the mass of the molecular ion of the product was 16 Da higher than that of the parent compound, supporting the hydroxylation of 1-ethylpyridin-2-ol. Similar results were obtained with 1-methyl-and 1-propylpyridin-2-ol. In all cases, the formation of a single product was observed indicating the position-specific hydroxylation. Moreover, the apparent equivalence with pyridin-2-ol transformation suggested that 1-alkylpyridin-2-ols were hydroxylated at the 5-position. Of all the compounds tested, only 1-butylpyridin-2-ol remained unchanged, which is most likely due to its bulkiness. In summary, pyridin-2-ols containing small 1-alkyl substituent are hydroxylated regioselectively, but further pyridine ring opening reaction does not occur. Thus, Burkholderia sp. MAK1 is capable of producing 1-alkylpyridine-2,5-diols.
Another group of potential Burkholderia sp. MAK1 substrates comprised pyridin-2-ols substituted at position 3 ( Fig. 2). HPLC-MS analysis revealed that compounds containing hydroxyl, methyl, bromo, chloro, or fluoro functional groups were completely catabolized by Burkholderia sp. MAK1 cells since no significant peaks corresponding to any hydroxylated products were detected. The latter suggests that the hydroxylated metabolites were likely further metabolized to aliphatic products. However, 3-(trifluoromethyl)pyridin-2-ol was slowly converted into a detectable new compound whose molecular mass was 16 Da higher than that of the substrate. Burkholderia sp. MAK1 cells were not able to hydroxylate pyridin-2-ols containing carboxyl or methoxy groups at position 3.
Pyridin-2-ols carrying substituents at positions 3 and 6 were also examined. The pyridin-2-ol-induced cells were able to metabolize 2-hydroxy-6-methyl-pyridine-3-carbonitrile: substrate concentration decreased over time, and no new products were detectable by HPLC-MS. After incubation of Burkholderia sp. MAK1 with  3-amino-6-methyl-pyridin-2-ol, a new compound with a molecular mass of 278 Da accumulated in the reaction mixture. Since the molecular mass of the expected 3-amino-6-methyl-pyridin-2-ol hydroxylation product is 140 Da, it is likely that the oxidation of the substrate is followed by the spontaneous dimerization. When Burkholderia sp. MAK1 cells were incubated with 3-bromo-6-methyl-pyridin-2-ol, neither hydroxylation, nor any other transformation occurred suggesting that 3-bromo functional group disrupted the proper orientation of the substrate. Pyridin-2-ols substituted at positions 4 and/or 6 were also used as substrates in this study (Fig. 3). Pyridine-2,4-diol was completely oxidized by Burkholderia sp. MAK1 cells after 20 hours of incubation. However, the intermediate product accumulating in the reaction mixture was detected by HPLC-MS and its absorption spectra as well as molecular mass ( Pyridine-2,6-diol was transformed by Burkholderia sp. MAK1 to a blue pigment. Previously, Holmes with colleagues described dimerization of pyridine-2,3,6-triol, which led to the formation of a blue pigment 38 . Following this observation, the hydroxylation of the symmetric pyridine-2,6-diol by Burkholderia sp. MAK1 cells likely occurred at position 3 of the pyridine ring and the resulting pyridine-2,3,6-triol spontaneously dimerized to a blue compound. Moreover, if the sixth position of pyridin-2-ol was occupied by a small and uncharged functional group, the pyridine ring cleavage probably followed the hydroxylation event. Summarizing experiments with substituted pyridin-2-ols we can make the statement that most of the substrates were consumed without detectable products. Although we were unable to provide any data about structures of the detectible product there were strong evidences suggesting regioselective hydroxylation at 5-position ( Table 2).

Screening of pyridin-2-amines as potential substrates for regioselective hydroxylation by
Burkholderia sp. MAK1 cells. The ability of Burkholderia sp. MAK1 to transform various pyridin-2-ols encouraged us to study pyridin-2-amines as another group of potential substrates. During the initial experiments, the cells were incubated with pyridin-2-amine for 20 hours. HPLC-MS analysis revealed that pyridin-2-amine  was completely consumed, and the new peak in the chromatogram belonged to the expected product. The molecular mass of the product, which was 16 Da higher than that of pyridin-2-amine, confirmed the notion that hydroxylation of the substrate occurred. The UV-Vis spectrum of the product was compared with spectra of commercially available reference standards (pyridin-2-amine hydroxylated at position 3, 4, or 6), yet none of these spectra matched that of the product (Supplementary information Fig. S-IV). From this we presume that in the case of Burkholderia sp. MAK1, pyridin-2-amine undergoes hydroxylation at position 5. Next, pyrazin-2-amine, a homolog of pyridin-2-amine containing two nitrogen atoms in the aromatic ring, was chosen as a substrate for the bioconversion. HPLC-MS analysis showed that the molecular mass of the biotransformation product was 16 Da higher than that of pyrazin-2-amine, suggesting that Burkholderia sp. MAK1 cells are also capable of pyrazin-2-amine hydroxylation.
Next, the ability of Burkholderia sp. MAK1 cells to transform pyridin-2-amines substituted at position 4 was investigated. Compounds with methyl, chloro, bromo, or fluoro substituents were hydroxylated. In all cases, the molecular mass of reaction products, as estimated by HPLC-MS, was 16 Da higher than that of parent compounds indicating that oxidation of substrates had occurred.

Oxyfunctionalization of pyridine, pyrazine and their derivatives using whole-cell biocatalyst. The study on pyridin-2-amine and pyridin-2-ol bioconversion by Burkholderia sp. MAK1 cells showed
that the pyridin-2-ol-inducible pyridin-2-ol 5-monooxygenase has broad substrate specificity and strict regiospecificity since it catalyzes hydroxylation at position 5 on the aromatic ring. With very few exceptions, microbial hydroxylation of pyridine-2-amines has been scarcely studied. One such exception is the study on the biotransformation of 4-methyl-3-nitro-pyridin-2-amine using whole-cells of fungus Cunninghamella elegans ATCC 26269. During this biotransformation, a mixture of three products, 6-amino-4-methyl-5-nitropyridin-3-ol, 2-amino-4-hydroxymethyl-3-nitropyridine, and 2-amino-4-methyl-3-nitropyridine-1-oxide was obtained suggesting that both aromatic and aliphatic positions as well as the heterocyclic nitrogen atom undergo oxidation 41 . In the case of Burkholderia sp. MAK1 cells, oxidation of the heterocyclic nitrogen atom was not observed when pyridin-2-ols were used as substrates. To determine if these bacteria were capable of producing N-oxides, various pyridine and pyrazine compounds without amino or hydroxy group at position 2 were tested as substrates for pyridin-2-ol-induced Burkholderia sp. MAK1 cells. HPLC-MS analysis showed that pyridine was transformed into a single product whose molecular mass was 16 Da higher than that of the parent compound. The UV spectrum of the product was very similar to that of pyridine yet did not match with the spectra of 2-, 3-, or 4-hydroxy-substituted pyridines at position suggesting that the product of pyridine biotransformation is pyridine-1-oxide (pyridine-N-oxide). The retention time, UV spectrum and ionisation profile of the bioconversion product matched those of analytical standard, pyridine-N-oxide, suggesting that Burkholderia sp. MAK1 catalyzes pyridine oxidation at position 1. Induction of cells with pyridin-2-ol was necessary for the oxidation of pyridine as well as for pyridin-2-ol and pyridin-2-amine transformation indicating that the same enzyme of Burkholderia sp. MAK1 is responsible for all these biotransformations.
A group of pyridines and pyrazines containing a methyl group attached to the aromatic ring at different positions (Fig. 5) was studied as potential substrates for Burkholderia sp. MAK1. The test revealed that the whole cells of Burkholderia sp. MAK1 catalyzed the transformation of 2-methyl-, 3-methyl-, and 4-methylpyridine into corresponding N-oxides whose structures were confirmed by HPLC-MS using analytical standards (Table 3). Burkholderia sp. MAK1 was also capable of transforming di-and trimethyl pyridines, except those in which both positions adjacent to nitrogen were occupied.
Based on HPLC-MS analysis, the biotransformation of pyrazine resulted in the formation of two products with molecular masses that were 16 Da and 32 Da higher than that of the parent compound. 1 H and 13 C NMR analysis allowed identification of these products as pyrazine Our research revealed that Burkholderia sp. MAK1 has also the ability to oxidize various methylpyrazines. For the oxidation of methylated pyrazines the single free position adjacent to either one of nitrogen atoms was a sufficient condition, e. g. the cells could oxidize 2,3,5-trimethylpyrazine, but not 2,3,5,6-tetramethylpyrazine.
To date, only a few reports regarding the microbial N-hydroxylation of pyridines have been published. The formation of pyridine N-oxides has been observed in fungi Cunninghamella elegans ATCC 26269 41 , Verticillium sp. GF39 31 , and other fungi 42 as well as in bacteria Methylococcus capsulatus 29 and Diaphorobacter sp. J5-51 43 . Also, the purified aromatic peroxygenase from fungus Agrocybe aegerita has been found to be active towards pyridine and its derivatives 30 . In this context, the results of this study not only broaden our understanding of microbial transformation but also provide a versatile tool that can be used in a regioselective oxyfunctionalization of various pyridine derivatives.

Conclusions
In summary, whole cells of Burkholderia sp. MAK1 have high activity towards pyridin-2-amines and pyridin-2-ols, and are applicable for the synthesis of pyridin-5-ols from the corresponding substrates. Moreover, unsubstituted pyridine and pyrazine as well as their methylated derivatives can be converted into the corresponding N-oxides using pyridin-2-ol-induced Burkholderia sp. MAK1 (Fig. 6). The approach presented here offers a promising alternative to chemical synthesis of hydroxylated pyridines.

Bioconversion of pyridines, pyrimidines and pyrazines using the cells of
MAK1 cells was resuspended in 100 ml of 10 mM potassium phosphate buffer, pH 7.2 supplemented with 15 mM of glucose and 0.25 mM of corresponding substrate and incubated at 30 °C. After bioconversion the cells of Burkholderia sp. MAK1 were separated by centrifugation. The supernatant liquid was vaporized to dryness under reduced pressure. The residue was dissolved in 5 ml of deionized water and purification of the product was carried out using reverse phase chromatography (12 g C-18 cartridge). Prior the purification the column was equilibrated with water. A mobile phase that consisted of water and methanol delivered in the gradient 10:0 → 10:5 elution mode. The collected fractions were analyzed by HPLC-MS. The fractions containing pure product were joined, and the solvent was removed under reduced pressure. 1 H NMR spectra were recorded in DMSO-d 6 or CDCl 3 on Bruker Ascend 400, 400 MHz, and 13 C NMR were recorded on Bruker Ascend 400, 100 MHz. Chemical shifts are reported in parts per million relative to the solvent resonance signal as an internal standard.

HPLC-MS analysis.
Before the analysis the cells were separated from the reaction mixture by centrifugation.
The resultant supernatant was mixed with an equal part of acetonitrile, centrifuged and analyzed using a high performance liquid chromatography system (CBM-20A controller, two LC-2020AD pumps, SIL-30AC auto sampler and CTO-20AC column oven; Shimadzu, Japan) equipped with a photo diode array (PDA) detector (SPD-M20A Prominence diode array detector; Shimadzu, Japan) and a mass spectrometer (LCMS-2020, Shimadzu, Japan) equipped with an ESI source. The chromatographic separation was conducted using a YMC Pack Pro column, 3 × 150 mm (YMC, Japan) at 40 °C and a mobile phase that consisted of 0.1% formic acid water solution (solvent A), and acetonitrile (solvent B) delivered in the 0 → 60% gradient elution mode. Mass scans were measured from m/z 10 up to m/z 700, at 350 °C interface temperature, 250 °C DL temperature, ± 4,500 V interface voltage, neutral DL/Qarray, using N 2 as nebulizing and drying gas. Mass spectrometry data was acquired in both the positive and negative ionization mode. The data was analyzed using the LabSolutions LCMS software.
Activity assay of pyridine-2,5-diol 5,6-dioxygenase from Burkholderia sp. MAK1. Burkholderia sp. MAK1 was grown at 30 °C for 20 hours in two 150 ml flasks, one containing 25 ml EFA medium (pyridin-2-ol induced cells), other containing 25 ml EFA medium where pyridin-2-ol is substituted for succinate (uninduced cells, negative control). The cells were harvested by centrifugation, washed twice with 10 mM potassium phosphate buffer (pH 7.2), suspended in 5 ml of the same buffer and sonicated. In 1.5 ml tubes three separate reaction mixtures were combined: internal control (990 μ l 10 mM potassium phosphate buffer, pH 7.2 and 10 μ l 2 mg/ml pyridine-2,5-diol solution), negative control (890 μ l 10 mM potassium phosphate buffer, pH 7.2, 10 μ l 2 mg/ml pyridine-2,5-diol solution and 100 μ l cell-free extract of uninduced cells) and sample (890 μ l 10 mM potassium phosphate buffer, pH 7.2, 10 μ l 2 mg/ml pyridine-2,5-diol solution and 100 μ l cell-free extract of induced cells). 100 μ l of each reaction mixture was transferred to a 96 well plate and change in absorbance (λ max 320 nm) per 30 minutes was measured. Overall change in absorbance was evaluated by subtracting noise data (internal and negative controls) from sample data. We were able to achieve 200-250 mU per 1 l medium, where 1 enzyme unit (U) is an amount of enzyme that catalyzes depletion of 1 μ mol pyridine-2,5-diol per minute. The measured molar extinction coefficient of pyridine-2,5-diol in ethanol was 9800 M −1 •cm −1 .