Overall Retention of Methyl Stereochemistry during B12-Dependent Radical SAM Methyl Transfer in Fosfomycin Biosynthesis

Methylcobalamin-dependent radical S-adenosylmethionine (SAM) enzymes methylate non-nucleophilic atoms in a range of substrates. The mechanism of the methyl transfer from cobalt to the receiving atom is still mostly unresolved. Here we determine the stereochemical course of this process at the methyl group during the biosynthesis of the clinically used antibiotic fosfomycin. In vitro reaction of the methyltransferase Fom3 using SAM labeled with 1H, 2H, and 3H in a stereochemically defined manner, followed by chemoenzymatic conversion of the Fom3 product to acetate and subsequent stereochemical analysis, shows that the overall reaction occurs with retention of configuration. This outcome is consistent with a double-inversion process, first in the SN2 reaction of cob(I)alamin with SAM to form methylcobalamin and again in a radical transfer of the methyl group from methylcobalamin to the substrate. The methods developed during this study allow high-yield in situ generation of labeled SAM and recombinant expression and purification of the malate synthase needed for chiral methyl analysis. These methods facilitate the broader use of in vitro chiral methyl analysis techniques to investigate the mechanisms of other novel enzymes.

. Plasmid map of btu-pBAD1030C-2.    Table S1. Radiochemical data for malates I and II isolated from chiral acetate samples.
All other chemicals and solvents were of reagent grade or higher.
Protein was concentrated aerobically in a sealed centrifugal filter device (EMD Millipore) to 2.5 mL, exchanged into gel filtration buffer (50 mM HEPES pH 7.5, 300 mM KCl, 15% (v/v) glycerol, 5 mM DTT) using a PD-10 desalting column (GE Healthcare), and concentrated again to 750 μL. The concentrated protein was flash frozen and stored in liquid nitrogen.
Determination of His6-SUMO-Fom3 cobalamin content. His6-SUMO-Fom3 was diluted to 40 μM in 50 mM NaOH, mixed with an equal volume of 0.2 M KCN in 10 mM NaOH, and incubated at 95 °C for 30 min. The resulting dicyanocobalamin was quantified by UV spectrophotometry (ε367 = 30,800 M -1 cm -1 ) 4 in comparison to a standard curve of HOCbl treated in the same manner.
Expression and purification of yeast malate synthase (ScMLS1). coScMLS1-pET28a was used to transform E. coli BL21(DE3) cells and 6 × 10 mL of starter culture was inoculated into 6 × 1 L of LB S5 medium + kanamycin (50 µg/mL). The cultures were shaken at 210 rpm and 37 °C; at an OD600 of 0.6, protein expression was induced with 50 μM IPTG and cultures were incubated at 20 °C and 210 rpm for 14 h. Cells (32 g) were harvested at 8,000 × g and resuspended in 200 mL of ice-cold lysis buffer (50 mM Tris pH 8.0, 20 mM KCl, 4 mM imidazole, 5 mM MgCl2, 1 mM TCEP) containing 0.5 mg/mL lysozyme, 1 mM PMSF, and 100 U/mL DNase I. Cells were lysed by sonication for 8 × 1 min and the lysate was centrifuged at 38,000 × g and 4 °C for 75 min. Clarified lysate was loaded onto a 2.5-cm diameter column containing 16 mL of cobalt TALON resin equilibrated in lysis buffer. The column was washed with 100 mL of ice-cold lysis buffer and protein was eluted with 60 mL of ice-cold elution buffer (50 mM Tris pH 8.0, 20 mM KCl, 250 mM imidazole, 5 mM MgCl2). Approx. 980 mg of His6-ScMLS1 was concentrated to 30 mL and dialyzed against 4 L of thrombin cleavage buffer (50 mM Tris pH 8.0, 10 mM CaCl2) at 4 °C overnight. The His6-tag was then cleaved from the protein by adding 4,000 U of thrombin (Sigma-Aldrich). The thrombin cleavage reaction was incubated at 23 °C for 25 h. Thrombin was removed by running the reaction solution through 5 mL of p-aminobenzamidine agarose (Sigma-Aldrich) at a flow rate of 1 mL/min; 2 mM imidazole was added to the flow-through and the cleaved His6-tag was removed by applying this solution to 16 mL of cobalt TALON resin equilibrated in thrombin cleavage buffer with 2 mM imidazole. The resin was washed with 20 mL of the same buffer, and the combined flow-through and wash fractions were concentrated to 30 mL. The resulting protein was divided into two 15-mL portions: the first portion was dialyzed against 1 L of malate synthase buffer (5 mM Tris-HCl pH 8.0, 10 mM MgCl2) at 4 °C overnight, and the second portion was dialyzed in the same manner against 1 L of malate synthase buffer containing 15% (v/v) glycerol. The dialysis was repeated for both portions using fresh buffer for 4 h. Initial activity assays were performed with both portions; the enzyme in glycerol-free buffer contained 17.0 mg/mL protein with 490 U/mL of malate synthase activity (29 U/mg protein), and the enzyme in glycerol-containing buffer contained 19.9 mg/mL protein with 490 U/mL of malate synthase S6 activity (25 U/mg protein). The protein in glycerol-free buffer (approx. 220 mg in 14 mL) was supplemented with 34 mg/mL sucrose as a lyoprotectant, divided into 0.5-mL aliquots, frozen overnight at −20 °C, cooled to -80 °C for 1 h, cooled in liquid nitrogen, and then lyophilized. The protein in glycerolcontaining buffer (approx. 300 mg in 15 mL) was flash frozen in 0.5-mL aliquots and stored at −80 °C.
A single aliquot of each portion was then thawed, the lyophilized aliquot was reconstituted with 0.5 mL water, and both protein concentration and activity were measured for both aliquots. The lyophilized aliquot contained 15.1 mg/mL protein and 380 U/mL of malate synthase activity (25 U/mg protein), and the frozen aliquot contained 19.9 mg/mL protein and 540 U/mL of malate synthase activity (27 U/mg protein).

Conversion of L-(methyl-13 C)methionine to (2-13 C)acetate.
Considering the low expected yield of acetate (0.3 mg) from each of the chiral methyl-labeled 2-HPP samples and the ubiquitous presence of acetic acid and acetate salts in common laboratory chemicals such as formic acid, a control sample of 13 Clabeled 2-HPP was oxidized to determine the yield of labeled acetate as well as any dilution with unlabeled acetate during the procedure. The (S)-(3-13 C)-2-HPP product of the MetK/Fom3/FomD replica reaction with L-(methyl-13 C)methionine was dissolved in water (3 mL) and (S)-( 13 C)-2-HPP was isolated by anion exchange chromatography. The residue from the formic acid fraction was dissolved in deionized water (5 mL) and Kuhn-Roth oxidized under optimized conditions (using KOH for neutralization of acetic acid) to furnish potassium acetate (8 mg, slightly yellowish powder) which was analyzed by NMR spectroscopy. 1

Conversion of potassium (2-2 H1)[2-3 H1]acetates 5 to malates I
Conversion of acetate 5a to malate Ia and isolation of malate Iaexperiment Ia-1. Potassium acetate 5a (2 µmol, 352 µL of stock solution, 12,179 Bq of 3 H) was diluted in carbonate buffer and water and reagents were added as described in Materials and Methods. The resulting reaction mixture was spiked with sodium [2-14 C]acetate (44 µL of stock solution, equivalent to 3,344 Bq of 14 C) and malate synthase (10 U), phosphotransacetylase (18 U) and acetate kinase (7 U) were added. After 2 h of reaction, unlabeled malic acid and perchloric acid were added, the mixture was filtered onto Dowex 1×8 resin, and the resin was washed as described in Materials and Methods. The fractions eluted with 0.8 M formic acid (2 × 25 mL, numbered 1 and 2) and 1.0 M formic acid (6 × 25 mL, numbered 3 to 8) were each spotted 7 times on cellulose TLC plates. They were developed with Et2O/HCO2H (99%)/H2O = 75:15:10 and dried in a vacuum desiccator at <1 mbar for 10 min. Then a solution of glucose (2 g) and aniline (2 mL) in a mixture of 1-butanol/ethanol/water 60:20:20 (100 mL) was applied. The plates were moved back and forth for 1 min to evaporate the liquid film from the surface. Finally, the plates were dried in a vacuum desiccator at <1 mbar for 15 min and heated in an oven at 140 °C for 10-15 min. Brown spots on a brownish background appeared; Rf = 0.60 for malic acid and 0.90 for fumaric acid; the latter stayed bound to the anion exchange resin. For stronger spots, small amounts (0.5 mL) of the aqueous fractions were put into vials (2 mL), cooled to 2-4 °C, then placed into a vacuum desiccator (15 mbar) over KOH and the water evaporated overnight. The residues were dissolved in water (50 µL, sonication for 30 s), and applied 5 times to a cellulose TLC plate as above. Malate Ia was detected in fractions 3, 4 and sometimes also 5 as judged by the intensity of the spots. Thus fractions 2 -5 or sometimes even 2 -6 were pooled, concentrated, and dried as described in Materials and Methods to yield crystalline malate Ia (23 mg); total activity calculated from 3 H: 7,440 Bq, from 14 C: 2,330 Bq, RCY: 70% (based on 14 C), ratio 3 H/ 14 C = 3.19 (experiment Ia-1, Table S1). '-ATGGTGAAAGTGAGCCTGGATAACGTTAAACTGCTGGTGGACGTGGATAAAGAACCGTTCTTTAAGCCG  AGCAGCACCACTGTGGGTGACATTCTGACCAAAGATGCCCTGGAATTTATTGTTCTCCTGCATCGTACC  TTTAATAACAAACGCAAACAGCTGCTCGAAAATCGTCAAGTGGTTCAGAAGAAACTGGATAGCGGTAGC  TACCATCTGGACTTCCTGCCGGAGACGGCAAACATTCGTAATGATCCGACCTGGCAAGGTCCGATTCTG  GCACCTGGCCTCATCAATCGTAGCACGGAAATTACCGGCCCTCCGCTGCGTAACATGCTGATCAACGCC  CTGAATGCACCGGTTAATACCTACATGACGGATTTTGAAGACAGCGCCAGCCCGACCTGGAACAACATG  GTTTATGGTCAGGTGAACCTGTACGATGCGATTCGTAATCAGATTGACTTTGACACTCCTCGTAAGAGC  TACAAACTGAACGGTAATGTCGCAAATCTGCCGACCATTATCGTGCGCCCGCGTGGTTGGCACATGGTT  GAAAAACATCTGTATGTGGATGACGAACCGATCTCTGCCTCAATTTTCGATTTTGGCCTGTACTTTTAT  CACAATGCGAAAGAACTGATTAAACTGGGTAAAGGTCCGTACTTCTATCTGCCGAAAATGGAACATCAC  CTGGAAGCAAAACTGTGGAATGATGTGTTTTGCGTCGCGCAGGACTATATTGGCATTCCGCGCGGCACC  ATTCGCGCGACCGTCCTGATTGAAACCCTGCCGGCCGCATTTCAGATGGAAGAAATTATCTATCAGCTG  CGTCAGCACAGCTCAGGCCTGAATTGCGGTCGCTGGGATTATATTTTTAGCACGATTAAACGCCTGCGC  AACGACCCGAATCATATTCTGCCGAACCGTAACCAGGTGACGATGACCAGCCCGTTTATGGACGCGTAT  GTGAAGCGTCTGATTAACACCTGCCACCGTCGCGGTGTTCATGCCATGGGTGGCATGGCCGCGCAGATC  CCTATTAAAGATGATCCGGCCGCAAATGAAAAGGCGATGACTAAGGTGCGCAATGATAAAATTCGTGAA  CTGACCAATGGTCATGATGGTTCATGGGTTGCGCATCCGGCACTGGCGCCGATTTGCAACGAAGTTTTT  ATTAACATGGGCACCCCTAACCAGATTTACTTTATTCCGGAAAACGTTGTGACGGCCGCAAACCTCCTG  GAAACCAAAATTCCGAACGGCGAAATTACCACGGAAGGCATCGTGCAGAATCTGGATATTGGCCTGCAG   S14   TATATGGAGGCGTGGCTGCGTGGTAGCGGCTGCGTTCCGATCAACAATCTGATGGAAGATGCAGCGACC  GCCGAAGTCAGCCGTTGCCAGCTGTATCAGTGGGTGAAACACGGCGTGACCCTGAAAGATACGGGTGAG  AAAGTTACCCCGGAACTGACTGAGAAAATCCTGAAAGAGCAGGTTGAACGCCTGTCAAAAGCGAGCCCT  CTGGGCGATAAGAACAAATTCGCACTGGCGGCCAAATACTTCCTGCCGGAAATTCGTGGTGAAAAGTTT  TCAGAGTTCCTGACGACCCTTCTGTATGATGAAATTGTGTCAACCAAAGCGACGCCGACCGATCTGAGC AAACTGTAA-3' Figure S1. Plasmid map of btu-pBAD1030C-2 encoding the E. coli B12 uptake genes btuCEDFB, used for coexpression with Fom3. Corrected from ref. 1 (see "Sequence of btu-pBAD1030C-2" in Materials and Methods). S16 Figure S2. 1 H-decoupled 31 P NMR spectrum (242 MHz, D2O) of the product mixture from a MetK/Fom3/FomD reaction performed with (methyl-13 C)methionine. The only visible signal in the phosphonate range (>5 ppm) of the 31 P NMR spectrum is that of (3-13 C)-2-HPP, which appears as a doublet due to coupling with 13 C at C3. Under these conditions, the signal for 2-HEP would appear at 17-18 ppm and the phosphonate signals for 2-HEP-CMP and 2-HPP-CMP would appear at 14-15 ppm.   Figure 4. The finite intramolecular 2 H kinetic isotope effect on malate synthase (kH/kD = 3.8) 7 yields malate I as a mixture of species with opposite stereochemistry at C2. In turn, after full equilibration with fumarase, (R)-(2-2 H1) [2-3 H1] Mean F-value (%) 24.5 80.7 a acetate kinase and phosphotransacetylase were replaced by acetyl-CoA synthetase; b radiochemical yields (RCY) are based on 14 C.