Biocatalytic asymmetric phosphorylation of mevalonate

Abstract

The excellent selectivity of the mevalonate kinase-catalyzed phosphorylation of mevalonate simplifies lengthy multi-step routes to (R)-mevalonate-5-phosphate to a one-step biocatalytic reaction, because the phosphate group can be transferred directly and without any additional reaction steps involving introduction and removal of protecting groups. By adjusting the required reaction time for complete conversion, the kinetic resolution of racemic mevalolactone can be easily used for the preparation of (S)-mevalonate and (R)-mevalonate-5-phosphate. A new recombinant mevalonate kinase has been prepared by the expression of the mevalonate kinase gene from the hyperthermophilic archaeon Thermococcus kodakarensis in Escherichia coli and by the subsequent purification. Direct quantitative ³¹P-NMR kinetic analysis has been utilized to characterize the enantioselectivity of the mevalonate kinase. This method is useful for determining the biocatalyst's utility for the synthesis of enantiomerically pure (R)-mevalonate-5-phosphate as well as for biocatalytic process development.

---

Introduction

The availability of pure metabolites in sufficient quantities continues to be essential for robust functional bioanalyses, stability investigations, the elucidation of biochemical pathways and the discovery of novel biological functions. As the preparation of highly functionalized, water-soluble and often chiral metabolites can be challenging and involve numerous reaction steps with protection–deprotection schemes, the development of selective synthetic tools remains important.¹ Phosphorylated low molecular weight compounds represent a diverse and large group of metabolites, which are highly important for the molecular understanding of many central biochemical pathways, their regulation and molecular biology. As part of our modular approach towards the synthesis and analysis of phosphorylated biomolecules, we have focussed in the present study on phosphorylated carboxylic acids.

Pioneering work from the laboratories of Bloch,² Lynen³ and Popjak⁴ on the isolation of 5-phosphomevalonic acid have led to its establishment as an intermediate in the biosynthetic pathway to terpenes.⁵ In the early days no satisfactory synthesis of this metabolite was available and the work was focussed on the preparation of the racemic DL-5-phosphomevalonic acid,⁶ which already has been a major effort due to challenges in selective phosphorylation methods and subsequent product purification. However, since the key metabolite of the mevalonate pathway⁷ occurs in nature in only one enantiomeric form, the (R)-mevalonate-5-phosphate (Fig. 1), its straightforward synthesis and analysis has been a topic for more than 50 years and has continued to be highly relevant. The formation of a phosphorylated derivative of mevalonic acid was first demonstrated by Tchen using ¹⁴C-labelled mevalonic acid, ATP and soluble yeast extract fraction A.⁸

Fig. 1 Mevalonate pathway(s).

The synthetic routes for phosphorylated metabolites like mevalonate-5-phosphate require either quite a number of reaction steps with laborious protection–deprotection sequences or extensive purification steps for the removal of low molecular weight impurities (Fig. 2).^6,9 In order to develop a viable short route to enantiomerically pure (R)-mevalonate-5-phosphate without late-stage resolution of racemic mixtures and to avoid waste which is scaling stoichiometrically with increasing batch sizes, the selection of catalytic and asymmetric synthesis methods is highly desirable.

Fig. 2 Synthetic routes to mevalonate-5-phosphate.

Synthetic catalytic asymmetric phosphorylation reactions have obtained increased attention in recent years.¹⁰ As an example, D-myo-inositol-6-phosphate has recently been prepared from 1,3,5-tri-O-PMB-myo-inositol by catalytic P(III) phosphoramidite transfer using a tetrazole-functionalized peptide catalyst in the solvent chloroform and 10 Å molecular sieves.^10f The overall procedure includes subsequent oxidation with 30% hydrogen peroxide and deprotection with sodium metal and liquid ammonia to give the disodium salt of the D-myo-inositol-6-phosphate.

Biocatalytic enantiospecific phosphorylations in aqueous solutions have been of great importance for biochemical analysis and are complementary synthetic tools suitable for the synthesis of water-soluble target molecules with high step economy and avoiding the use of protecting groups.¹¹ Numerous well-known biocatalytic phosphorylation reactions are key to central metabolic pathways and a variety of enzymes are available for catalyzing these phosphate transfer reactions.¹² The enzyme mevalonate kinase (ATP: mevalonate 5-phosphotransferase EC 2.7.1.36) has been found in a wide variety of biological systems.¹³ Classical biochemical experiments on the elucidation of the mevalonate pathway (Fig. 1), using radioactive labelling with carbon-14, synthetic experiments with phosphomevalonate barium salt and optical rotation measurements after barium salt removal, have shown that the assumptions that mevalonate kinase reacts stereospecifically are correct.¹⁴ Only the natural (R)-enantiomer of mevalonate in a racemic mixture is phosphorylated by the mevalonate kinase-catalyzed reaction to produce (R)-mevalonate-5-phosphate using ATP as phosphoryl group donor.¹⁴

The crystal structure of mevalonate kinase from Methanocaldococcus jannaschii reveals a mevalonate binding pocket that constrains the orientation of the 5-OH group of the two mevalonate enantiomers.¹⁵ When racemic mevalonate is used as starting material for the enzymatic phosphorylation, both (R)-mevalonate-5-phosphate and (S)-mevalonate can be obtained simultaneously (Fig. 3), with (R)-mevalonate as starting material only (R)-mevalonate-5-phosphate is formed after completion of the reaction.

Fig. 3 Mevalonate kinase-catalyzed stereospecific phosphorylation of racemic mevalonate in aqueous buffer at pH 8.

Extensive work-up procedures and in-process analysis, which previously involved indirect coupled spectrophotometric analyses and radioisotopic assays, can be avoided with the approach described here. Those preparations leading to a mevalonate-5-phosphate solution were laborious, time-consuming and with no indication of the enantiopurity.¹⁶

Results and discussion

A simple one-step biocatalytic asymmetric phosphorylation reaction with the maximum conversion of 50%, which is achievable in the kinetic resolution of racemic mevalolactone, is described using a cloned archaeal mevalonate kinase and its mevalonate substrate as starting material, which could also be the corresponding lactone form due to the equilibrium in aqueous solution (Fig. 3).

In order to accelerate the development of a robust synthetic procedure, a continuous and quantitative ³¹P-NMR method has been developed for the analysis of (R)-mevalonate-5-phosphate formation as described in the experimental part. The accurate product quantitation was performed with a calibration curve of different mevalonate-5-phosphate concentrations using the absolute concentration determined by quantitative ¹H-NMR. The comparison of the time course of the mevalonate kinase-catalyzed phosphorylation of (R)-mevalolactone with racemic mevalolactone (Fig. 4) demonstrates that the enantiomerically pure (R)-mevalolactone is phosphorylated to (R)-mevalonate-5-phosphate with a degree of conversion of 100%, compared to only 50% starting with racemic mevalolactone. No phosphorylation at all could be observed with (S)-mevalolactone as substrate.

Fig. 4 Time course of (R)-mevalonate-5-phosphate concentration [MVAP] in the enzymatic phosphorylation reaction by ³¹P-NMR. [MVAP] from (R)-mevalolactone, [MVAP] from racemic (R,S)-mevalolactone, [MVAP] from (S)-mevalolactone.

For preparative purposes, the long reaction time at room temperature can easily be shortened by raising the reaction temperature and increasing the volume activity of the mevalonate kinase used. The ease of purification for the recombinant mevalonate kinase and the opportunity to run the preparative reaction at 55 °C are major advantages and have been the starting point for selecting the cloning, expression and purification of the mevalonate kinase from Thermococcus kodakarensis.

Experimental

Materials and methods

Racemic and (R)-mevalolactone and other commercially available reagents were obtained from Sigma-Aldrich, where not mentioned otherwise. (S)-Mevalolactone has been prepared by enzymatic resolution of racemic mevalolactone as described in this paper.

Cloning, expression and purification

Genomic DNA was derived from Thermococcus kodakarensis KOD1. E. coli DH5α and plasmid pUC19 were used for DNA manipulation. E. coli BL21-CodonPlus (DE3)-RIL (Stratagene, La Jolla, CA, USA) and pET21a (+) (Novagen, Wisconsin, WI, USA) were used for protein production. E. coli strains were cultivated in Luria–Bertani medium with 100 μg ml⁻¹ ampicillin at 37 °C. The mk_Tk gene (TK1474) was PCR amplified from the genomic DNA of T. kodakarensis using the primer set MK-5 and MK-3 (MK-5,5′-GGAGGGTCACATATGAGCGTTAAAAGGGTTCTTGC-3′; MK-3,5′-GAAGACGCTACCGCCGGAATTCATGAGTATCACTC-3′). The amplified DNA fragment was inserted into pET21a (+) at the NdeI and EcoRI sites after its sequence was confirmed. The plasmids were introduced into E. coli BL21-CodonPlus (DE3)-RIL cells for gene expression. E. coli cells harbouring the expression plasmid for the MK_Tk protein were grown in 700 ml LB medium. Gene expression was induced with 0.1 mM isopropyl-β-D-thiogalactopyranoside at the mid-exponential growth phase. After induction, cells were incubated at 20 °C for 12 h, then collected and washed with 50 mM Tris–HCl (pH 8.0) and resuspended in 20 ml of the same buffer. Cells were sonicated on ice and after centrifugation (20 000 × g, 30 min at 4 °C) the supernatant was incubated at 85 °C for 15 min, then immediately cooled on ice and centrifuged (20 000 × g, 30 min at 4 °C). The soluble protein in the supernatant was loaded on a Q-Sepharose anion exchange column (GE Healthcare, Little Chalfont, UK) that was equilibrated with 50 mM Tris–HCl (pH 8.0), and proteins were eluted with a linear gradient of NaCl (0.0 to 1.0 M). This was followed by gel filtration chromatography on a Superdex 200 HR 10/30 column (GE Healthcare) that was equilibrated with 50 mM Tris–HCl (pH 8.0) and 0.15 M NaCl. A mevalonate kinase solution with an activity of 43.6 U ml⁻¹ was obtained.

NMR spectroscopy

Quantitative ³¹P-NMR (qNMR) spectra were recorded at 242.98 MHz on a Bruker Avance III 600 spectrometer equipped with a 5 mm broadband z-gradient probe. Temperature stability was controlled by a BVT 3200 unit. A number of ³¹P-qNMR experiments were carried out with an interval of at least 60 minutes between each acquisition for analysis of mevalonate-5-phosphate formation. For reference, deuterated phosphoric acid was used to calculate the concentration of mevalonate-5-phosphate. Inverse gate proton decoupling was applied during data acquisition without signal enhancement due to the Nuclear Overhauser Effect (NOE) with a flip angle of 30° and a relaxation delay of 10 s. Each spectrum was acquired with 64 scans at a temperature of 298.2 K. After processing, phase and baseline correction of the ³¹P-qNMR spectra integration was subsequently executed with identical limits within the series of spectra. The integral of the standard D₃PO₄ was set to 100 and the resulting integral of mevalonate-5-phosphate used for calculation of the concentration.

Experimental approach with stem coaxial inserts. For the qNMR experiments specific requirements have to be fulfilled. The sample should be homogenous and when adding the reference standard it is crucial that the solution must be mixed to complete homogeneity. Another requirement when performing quantitative NMR analysis is that each compound needs to have at least one completely resolved signal appearing in the spectrum. When conducting proton qNMR on proton-rich samples this may be problematic, but for phosphorous qNMR with a chemical shift range of more than 400 ppm and proton decoupling, single, well resolved peaks will often be obtained. Quantitative NMR is a primary ratio method due to the fact that the peak areas are proportional to the number of corresponding nuclei giving rise to the signal, therefore it is important that sufficient time is allowed to pass between each scan and recovery of each nucleus to its equilibrium is completed.¹⁷

For the quantitative ³¹P NMR experiment we used 85% H₃PO₄ in D₂O (Sigma-Aldrich) as reference standard with a single signal at 0 ppm. But since the kinase is very sensitive to changes of the pH, the standard was placed into a stem coaxial insert (Wilmad) and this was then combined with the regular NMR tube which contained the reaction mixture. In this case the requirement of same volume for sample and reference was violated and for the calculation of the concentration of mevalonate-5-phosphate the difference between the two volumes was taken into account by carrying out additional calibration measurements with known concentrations of mevalonate-5-phosphate which led to the volume correlation factor F that gives the relation of the two volumes.

where I_MEP is the integral of the mevalonate-5-phosphate (MEP), P_MEP is the number of phosphorus nuclei generating the signal, c_PA is the concentration of phosphoric acid in D₂O, F is the volume correlation factor.

Kinetic analysis of mevalonate phosphorylation by ³¹P-NMR

Enantiomerically pure (R)- and (S)-mevalonate respectively. A solution of 149 mg (0.27 mmol) ATP disodium salt hydrate, 18 mg MgCl₂ hexahydrate and 336 mg Tris base in 3 ml Tris buffer 1 M pH 8.0 was added to 0.27 mmol of the enantiomerically pure mevalonate or mevalolactone respectively, which results in a mevalonate concentration of 80 mM.

To this solution 2 μl of mevalonate kinase (MVK) as described above were added and, after mixing, 0.5 ml thereof were transferred to an NMR tube to which was further added an insert containing 2.3 M D₃PO₄ in D₂O.

Racemic mevalonate. A solution of 194 mg (0.35 mmol) ATP disodium salt hydrate, 22.3 mg MgCl₂ hexahydrate and 429 mg Tris base in 4 ml Tris buffer (1 M, pH 8.0) was added to 45.8 mg (0.35 mmol) of racemic mevalolactone, giving a mevalonate concentration of 78 mM.

To this solution 2 μl of mevalonate kinase (MVK) solution as described above were added and, after mixing, 0.5 ml thereof were transferred to an NMR tube that contained an insert with 2.3 M D₃PO₄ in D₂O.

Synthetic application of mevalonate kinase-catalyzed phosphorylation of (R)-mevalolactone

A mixture of 7.73 g (14.0 mmol) adenosine 5′-triphosphate disodium salt hydrate (ATP), 1.54 g (12.6 mmol) R-mevalolactone and 820 mg magnesiumchloride hexahydrate in 21 ml H₂O was adjusted to pH 8 with 30 ml 1 M LiOH and then 65.4 ml 0.1 M LiOH. At 55 °C and with stirring 0.1 ml mevalonate kinase solution were added. During 25 hours the pH was kept constant at 8 by adding 0.1 M LiOH with a autoburette. A turnover of ∼90% was calculated based on the consumption of LiOH and thin layer chromatography (TLC) showed only a trace of remaining mevalolactone.

The reaction mixture was concentrated on a rotary evaporator at 40 °C, the residue dissolved in 44 ml H₂O and then 88 ml methanol were added dropwise with stirring. The resulting suspension was filtered and the absence of phosphomevalonate in the solid residue checked by TLC. To the filtrate 300 ml acetone were added with stirring and the resulting suspension was again filtered. The residue, which was washed with some acetone and which contained no starting material and only minor amounts of ATP/ADP as judged by TLC, was dissolved in a minimum amount of H₂O, adjusted to pH 2.5 with formic acid and chromatographed on silica (propanol–H₂O–formic acid = 12 : 3 : 1). The fractions containing the desired product were pooled, evaporated and the residue taken up in a minimum amount of H₂O and passed through a cation exchanger column (Dowex 50WX8, H+ form). The product containing fraction was concentrated, the residue mixed several times with H₂O and concentrated to remove formic acid, then adjusted to pH 8 with 1 M LiOH and lyophilized which gave 1.92 g (65%) (R)-5-phospho-mevalonate lithium salt. The white powder had >98% purity by TLC. Elemental analysis: C 26.09%, H 4.57%, Li 8.43%, water content 11.0%, [α]²⁰_D = −2.6° (c = 0.1 in 0.1 N HCl), ¹H-NMR and ³¹P-NMR spectra were conform with structure.

The enzymatic ATP-regeneration using the phosphoenolpyruvate/pyruvate kinase system has also been established and enables easy scalability and the realization of the full potential of this process also for large scale.

Conclusions

The results clearly show the excellent enantioselectivity and complete conversion of the biocatalytic asymmetric phosphorylation of mevalonate catalyzed by the recombinant Thermo-coccus kodakaraensis mevalonate kinase, which is superior to any classical chemical synthetic route. Lengthy multi-step routes to (R)-mevalonate-5-phosphate involving the selective introduction and removal of protecting groups or extensive purification steps have been simplified to a one-step biocatalytic asymmetric phosphorylation reaction. This represents a major improvement according to many criteria of green chemistry. The required reaction time for achieving complete conversion in the kinetic resolution of racemic mevalolactone can be easily adjusted for the preparation of (S)-mevalonate and (R)-mevalonate-5-phosphate, which can of course also be prepared by starting from (R)-mevalolactone (Fig. 3). This continuous and direct kinetic ³¹P-NMR methodology can now not only be used to characterize the enantioselectivity of the various mevalonate kinases available and determine their utility for the synthesis of both enantiomerically pure as well as racemic mevalonate-5-phosphates, stable isotope-labelled versions and other substrate analogues, but also for the exploration of a variety of other kinases.

A key element for robust and successful kinetic ³¹P-NMR experiments was the use of 85% phosphoric acid in D₂O in a stem coaxial insert. Consistent ³¹P-NMR spectra of the biocatalytic phosphorylation reaction can then be accumulated as a function of time in a straightforward way. The growth of the ³¹P-NMR signals due to the mevalonate-5-phosphate formation as well as the concomitant conversion of ATP to ADP can then be monitored and quantified, as shown in Fig. 5 by a few selected ³¹P-NMR spectra during the course of the phosphorylation reaction.

Fig. 5 Direct ³¹P-NMR spectra at the start (bottom) and at selected time points (middle and top) during the course of the mevalonate kinase-catalyzed phosphorylation reaction of (R)-mevalolactone. The peak at ~4 ppm originates from mevalonate-5-phosphate, the remaining peaks from ATP and ADP respectively.

The direct and quantitative ³¹P-NMR kinetic analysis of product formation in enzymatic phosphorylation reactions is equally suitable for other phosphate transfer reactions and is of broad interest in biology, chemistry, biotechnology and biomedical sciences.

Notes and references

1. R. Wohlgemuth, Biotechnol. J., 2009, 4, 1253–1269.
2. M. Lindberg, C. Yuan, A. DeWaard and K. Bloch, Biochemistry, 1962, 1, 182–188.
3. B. W. Agranoff, H. Eggerer, U. Henning and F. Lynen, J. Biol. Chem., 1960, 235, 326–332.
4. (a) G. Popjak, Annu. Rev. Biochem., 1958, 27, 533–560; (b) G. Popjak, Methods Enzymol., 1969, 15, 410–413.
5. (a) J. W. Cornforth, Chem. Soc. Rev., 1973, 2, 1–20; (b) G. Popjak and J. W. Cornforth, Adv. Enzymol. Relat. Subj. Biochem., 1960, 22, 281–335.
6. (a) F. Lynen, H. Eggerer, U. Henning and I. Kessel, Angew. Chem., 1958, 70, 738–742; (b) H. Hellig and G. Popjak, Biochem. J., 1961, 80, 47P; (c) K. Kirschner and F. Lynen, Chem. Ber., 1961, 94, 1595–1606; (d) H. Machleidt, E. Cohnen and R. Tschesche, Justus Liebigs Ann. Chem., 1962, 655, 70–80; (e) C. D. Foote and F. Wold, Biochemistry, 1963, 2, 1254–1258.
7. (a) J. L. Goldstein and M. S. Brown, Nature, 1990, 343, 425–430; (b) H. Nishimura, Y. Azami, M. Miyagawa, C. Hashimoto, T. Yoshimura and H. Hemmi, Biochem. J., 2013, 153, 415–420; (c) M. Thurnher, G. Gruen-bacher and O. Nussbaumer, Biochim. Biophys. Acta, Mol. Cell Biol. Lipids, 2013, 1831, 1009–1015; (d) P. P. Peralta-Yahya, F. Zhang, S. B. Del Cardayre and J. D. Keasling, Nature, 2012, 488, 320–328.
8. T. T. Tchen, J. Am. Chem. Soc., 1957, 79, 6344–6345.
9. (a) C. Z. Wang and H. M. Miziorko, Anal. Biochem., 2003, 321, 272–275; (b) C. S. Lee and W. J. O'Sullivan, Biochim. Biophys. Acta, 1985, 839, 83–89; (c) D. Pilloff, K. Dabovic, M. J. Romanowski, J. B. Bonanno, M. Doherty, S. K. Burley and T. S. Leyh, J. Biol. Chem., 2003, 278, 4510–4515.
10. (a) B. R. Sculimbrene and S. J. Miller, J. Am. Chem. Soc., 2001, 123, 10125–10126; (b) B. R. Sculimbrene, A. J. Morgan and S. J. Miller, J. Am. Chem. Soc., 2002, 124, 11653–11656; (c) A. J. Morgan, S. Komiya, Y. J. Xu and S. J. Miller, J. Org. Chem., 2006, 71, 6923–6931; (d) B. R. Sculimbrene, Y. J. Xu and S. J. Miller, J. Am. Chem. Soc., 2004, 126, 13182–13183; (e) C. M. Longo, Y. Wei, M. F. Roberts and S. J. Miller, Angew. Chem., Int. Ed., 2009, 48, 1–5; (f) P. A. Jordan, K. J. Kayser-Bricker and S. J. Miller, Proc. Natl. Acad. Sci. U. S. A., 2010, 107, 20620–20624.
11. (a) D. C. Crans and G. M. Whitesides, J. Am. Chem. Soc., 1985, 107, 7008–7018; (b) Y. G. Kwon, J. Srinivasan, M. Mendelow, S. Pluskey and D. S. Lawrence, J. Am. Chem. Soc., 1993, 115, 7527–7528; (c) H. K. Chenault, R. F. Mandes and K. R. Hornberger, J. Org. Chem., 1992, 62, 331–336; (d) V. M. Rios-Mercadillo and G. M. Whitesides, J. Am. Chem. Soc., 1979, 101, 5828–5829; (e) T. van Herk, A. F. Hertog, A. M. van den Burg and R. Wever, Adv. Synth. Catal., 2005, 347, 1155–1162.
12. W. W. Clelland and A. C. Hengge, Chem. Rev., 2006, 106, 3252–3278.
13. (a) J. W. Porter, Methods Enzymol., 1985, 110, 71–78; (b) R. Matsumi, H. Atomi, A. J. Driessen and J. van der Oost, Res. Microbiol., 2011, 162, 39–52.
14. (a) F. Lynen and M. Grassl, Hoppe-Seyler's Z. Physiol. Chem., 1958, 313, 291; (b) R. H. Cornforth, J. W. Cornforth and G. Popjàk, Tetrahedron, 1962, 18, 1351; (c) T. T. Tchen, J. Biol. Chem., 1958, 233, 1100; (d) H. R. Levy and G. Popjak, Biochem. J., 1959, 72, 35P; (e) A. DeWaard and G. Popjak, Biochem. J., 1959, 73, 410–415; (f) R. H. Cornforth, K. Fletcher, H. Hellig and G. Popjak, Nature, 1960, 185, 923–924; (g) G. Popjak and J. W. Cornforth, Biochem. J., 1966, 101, 553–568.
15. D. Yang, L. W. Shipman, C. A. Roesner, A. I. Scott and J. C. Sacchettini, J. Biol. Chem., 2002, 277, 9462–9467.
16. (a) S. Bazaes, E. Beytia, A. M. Jabalquinto, F. Solis de Ovando, I. Gomez and J. Eyzaguirre, Biochemistry, 1980, 19, 2300; (b) H. R. Levy and G. Popjàk, Biochem. J., 1960, 75, 417; (c) J. Eyzaguirre and S. Bazaes, Methods Enzymol., 1985, 110, 80–82.
17. (a) F. Malz and H. Jancke, J. Pharm. Biomed. Anal., 2005, 38, 813–823; (b) T. J. Henderson, Quantitative NMR spectroscopy using coaxial inserts containing a reference standard: Purity determinations for military nerve agents, Anal. Chem., 2002, 74, 191–198.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra01299b

---