Synthesis of 7-Mercaptoheptano-ylthreonine Phosphate and Its Activity in the Methylcoenzyme M Methylreductase System*

The structure of component B of the methylcoenzyme M methylreductase of Methanobacterium thermoau-totrophicum was recently assigned as ‘7-mercaptohep- tanoylthreonine phosphate (HS-HTP) (Noll, K. M., Rinehart, K. L., Jr., Tanner, R. S., and Wolfe, R. S. (1986) (Proc. Natl. Acad. Sei. U. S. A. 83,4238-4242). We report here the chemical synthesis and biochemical activity of this compound. Thiourea and 7-bromohep-tanoic acid were used to synthesize 7,7’-dithiodihep- tanoic acid. This disulfide was then condensed with DL-threonine phosphate using N-hydroxysuccinimide and dicyclohexylcarbodiimide. The product was reduced with dithiothreitol to give HS-HTP. It could be oxi- dized in air in the presence of 2-mercaptoethanol to give the compound as it was isolated from cell extracts. The resulting product was identical to the authentic compound by ‘H NMR spectroscopy, mass spectrom- etry, and coelution using high performance liquid chromatography. The synthetic compound is active in the in vitro methanogenic assay at concentrations compa-rable to the authentic compound. This confirms the structure of component B as HS-HTP and provides a means to synthesize quantities sufficient for studies of the methylreductase system. Coupling of the N-Hydroxysuccinimide Ester and Threonine Phos- phate-A solution of DL-threonine phosphate (78.0 mg, 392 pmol) in 2.5 ml of water containing 108 pl (775 pmol) of triethylamine was added with stirring to a solution of the N-hydroxysuccinimide ester (100.9 mg, 196 pmol) in 5 ml of tetrahydrofuran; 1 ml of acetonitrile was added to the mixture to achieve a single phase. After 29 h at room temperature, the solvent was removed under vacuum. Inside an anoxic chamber (Coy Scientific Products, Ann Arbor, MI) the con-tents of the flask were dissolved in 10 ml of deaerated 50 mM ammonium bicarbonate containing 100 mM dithiothreitol. After standing in the chamber for 30 min, this mixture was applied to a 6- ml column of Accell QMA Anion-exchange Medium (Waters Associates, Milford, MA) pre-equilibrated with deaerated 50 mM ammonium bicarbonate. The column was washed with 15 ml of this buffer to remove the dithiothreitol, and HS-HTP was gradient-eluted with successive applications of seven 5-ml aliquots of deaerated buffer, each aliquot containing a 100 mM ammonium bicarbonate increment. Fractions of 5 ml were collected and assayed for HS-HTP. The active fractions were pooled and lyophilized repeatedly to remove the salt. The product, 52.5 mg (37%) of white crystals, was judged as pure by

3). The enzyme fractions consist of components Al, A2, A3, and C. The cofactor components are ATP, FAD, CH~-S-COM, and a heat-stable, low molecular weight cofactor present only in extracts of methanogenic bacteria, component B (2). Methane formation by extracts was shown to be completely dependent upon the presence of this cofactor. Until recently, this had been the only unidentified cofactor involved in methane formation. The purification of this cofactor was recently accomplished and the structure was proposed to be 7-mercaptoheptanoylthreonine phosphate (HS-HTP) (4). This structure was determined by the use of high resolution mass spectrometry, 'H NMR, 13C NMR, IR spectroscopy, and chemical analysis. Only 9 mg of pure cofactor could be obtained from 1 kg of wet cells by a laborious purification scheme. Thus, the amount available for its use in studies of methanogenesis was limited. We report here the chemical synthesis of this compound and demonstrate its activity in the in vitro methanogenic assay.

HS-HTP Assays-Cells of
Methambacterium thermoautotrophicum strain AH were grown and extracts were prepared as described previously (4). Assays for HS-HTP were conducted using a cell-free extract stripped of soluble cofactors by passage twice through a column of Sephadex G-25 Superfine resin (Pharmacia P-L Biochemicals). The assay mixture contained all necessary components except HS-HTP. Activity was observed by the measurement of methane formation from the demethylation of CH3-S-CoM as previously described (4). Synthesis of 7,''-Dithiodiheptanoic Acid-7-Mercaptoheptanoic acid was synthesized by dissolving 3.24 g (42.65 mmol) of thiourea in a stirred solution of 7-bromoheptanoic acid (1.78 g, 8.53 mmol) in 20 ml of ethanol. This mixture was refluxed at 90 "C for 17 h, cooled to room temperature, and 5 ml of a 60% aqueous solution (w/v) of sodium hydroxide was added. The mixture was refluxed for an additional hour, cooled to room temperature, acidified with hydrochloric acid, and extracted with chloroform. The chloroform phase was extracted with a 1 M aqueous solution of sodium bicarbonate. The aqueous extract was acidified and extracted with chloroform. The thiol was oxidized to a disulfide by mixing the chloroform phase with an aqueous solution of 10% (w/v) iodine and 20% (w/v) potassium iodide until the brown color persisted. The aqueous phase was removed and the chloroform phase was washed twice with water, dried over anhydrous magnesium sulfate, and concentrated under vacuum. The product was crystallized twice from benzene-pentane to give 588 mg of white crystals (43%). Synthesis of the N-Hydroxysuccinimie Ester-The disulfide product was activated by synthesis of its N-hydroxysuccinimide ester using dicyclohexylcarbodiimide ( 5 ) . 7,7'-Dithiodiheptanoic acid (200.9 mg, 624 pmol) was dissolved in 6 ml of 1,4-dioxane at room temperature and the solution was stirred while 149.6 mg (1.26 mmol) of N-hydroxysuccinimide was added when this dissolved, 261.6 mg (1.27 mmol) of dicyclohexylcarbodiimide was added and the solution was stirred 20 h at room temperature. The precipitated dicyclohexylurea was removed by filtration and the filtrate was dried to a clear oil under vacuum. A small amount of 2-propanol was added to the oil, and this solution was dried to a white solid under vacuum. The product was recrystallized twice from 2-propanol to give 201.6 mg (63%) of white crystals. Coupling of the N-Hydroxysuccinimide Ester and Threonine Phosphate-A solution of DL-threonine phosphate (78.0 mg, 392 pmol) in 2.5 ml of water containing 108 pl (775 pmol) of triethylamine was added with stirring to a solution of the N-hydroxysuccinimide ester (100.9 mg, 196 pmol) in 5 ml of tetrahydrofuran; 1 ml of acetonitrile was added to the mixture to achieve a single phase. After 29 h at room temperature, the solvent was removed under vacuum. Inside an anoxic chamber (Coy Scientific Products, Ann Arbor, MI) the contents of the flask were dissolved in 10 ml of deaerated 50 mM ammonium bicarbonate containing 100 mM dithiothreitol. After standing in the chamber for 30 min, this mixture was applied to a 6ml column of Accell QMA Anion-exchange Medium (Waters Associates, Milford, MA) pre-equilibrated with deaerated 50 mM ammonium bicarbonate. The column was washed with 15 ml of this buffer to remove the dithiothreitol, and HS-HTP was gradient-eluted with successive applications of seven 5-ml aliquots of deaerated buffer, each aliquot containing a 100 mM ammonium bicarbonate increment.   HPLC and 'H NMR spectroscopy. The overall yield was 10%.

CJ&O&
Analytical Methods-High performance liquid chromatography was performed using a semipreparative (300 X 7.8 mm) pBondapak Cla column (Waters Associates). The column was equilibrated with 40% methanol in 20 mM ammonium acetate (pH 5). The sample was eluted by application of a 15-min linear gradient to 50% methanol in this buffer at a flow rate of 1 ml/min, beginning immediately after injection of the sample. The effluent was monitored by UV absorbance using a Varian Instruments Model 2050 variable wavelength UV detector. Waters Associates Model 45 pumps and a Model 660 solvent programmer were used.
'H NMR spectroscopy was performed using a Nicolet 360-MHz Fourier transform spectrometer. Mass spectrometry was performed using a VG Analytical ZAB SE mass spectrometer. Conditions for these analyses were described previously (4).
Chemicals-All chemicals and solvents were of reagent grade. 7-Bromoheptanoic acid was purchased from ICN Pharmaceuticals, Inc., Plainview, NY. DL-Threonine phosphate and thiourea were from Sigma. N-Hydroxysuccinimide and dicyclohexylcarbodiimide were from Aldrich.

RESULTS AND DISCUSSION
The synthesis described here yields a product that is active in the methanogenic assay. The activity of the thiol form of HS-HTP had been noted previously (4), but all chemical characterizations had been performed using the mixed disulfide of HS-HTP and 2-mercaptoethanol, the form that was isolated from cell extracts. This disulfide can be readily obtained by adding an excess of 2-mercaptoethanol to a solution of HS-HTP and drying the solution under vacuum. To compare the authentic and synthetic cofactors, the same form of the coenzyme is required. For the analyses described here, this mixed disulfide was made and the product purified by the HPLC method described under "Experimental Procedures." The mixed disulfide of the synthetic product was found to coelute with the mixed disulfide of the authentic cofactor (Fig.  1). When mixed together, the two compounds eluted as a single, uniform peak, thus supporting their identity.
The positive ion of the synthetic mixed disulfide had a mass of 366.0755, as determined by high resolution, fast atom bombardment mass spectrometry while the positive ion of the monosodium salt of HS-HTP has a calculated mass of 366.0755. This confirms the structural formula of C,,Hz2NO7PSNa for this ion of the synthetic compound. A fragment ion of mass 246.1169 was also observed from the synthetic compound. This may be assigned to a fragment lacking a phosphate (CllH2003NS, calculated mass 246.1164).
The 'H NMR spectrum of the synthetic compound was also identical to the authentic compound as shown in Table I. In the thiol form, the triplets at 2.89 and 3.87 ppm are absent and the triplet at 2.78 ppm shifts to 2.56 ppm. This is consistent with the loss of 2-mercaptoethanol and the theoretical chemical shift of a methylene adjacent to a free thiol (6). The multiplets at 1.65 and 1.71 ppm also change and simplify to an apparent sextet (pair of triplets). This is due to the shift upfield of the 1.71-ppm multiplet so that it nearly overlaps the 1.65-ppm multiplet. This corresponds to the expected shift of the methylene protons that are p to the thiol.
The identity of the synthetic product was further confirmed by its biochemical activity. Authentic HS-HTP exhibited onehalf the maximal rate of methane formation at a concentration of 3 b~ (4). The synthetic mixed disulfide was also active as shown in Fig. 2. One-half maximal activity was found at a concentration of 6 PM. This higher K,,, may be due to the fact that 0-phospho-DL-threonine was used in the synthesis. A racemic mixture of HS-HTP, therefore, was synthesized. The configuration of the authentic compound is unknown.
The effect of added HS-HTP on the length of time before the onset of a linear rate of methane formation (the lag time) was also similar. The addition of increasing amounts of HS-HTP shortened the lag time (Fig. 2). This lag was found to decrease from 18 min (no HS-HTP added) to 7 min (22.5 nmol of HS-HTP added). (Please note that in Ref. 4 The present work outlines a method for the synthesis of HS-HTP using readily available chemicals in a simple synthetic scheme. A more complex synthesis was recently presented (7) starting from diethyl pimelate and using the method of Walton et al. (8) for synthesis of lipoic acid. That synthesis involved seven steps and gave a final yield of 0.4%. The method presented here is a simplification of that method and provides a much better yield. Most importantly, the chemical synthesis of HS-HTP confirms the previously proposed structure. Now, all the known cofactors involved in methanogenesis have been identified. With this synthesis, it will be possible to obtain quantities of this cofactor sufficient for the elucidation of its role in methane formation.