The enethiolate anion reaction products of EpiD. Pka value of the enethiol side chain is lower than that of the thiol side chain of peptides.

One of the steps involved in the biosynthesis of the lantibiotic epidermin is the oxidative decarboxylation reaction of peptides catalyzed by the flavoenzyme EpiD. EpiD catalyzes the formation of a (Z)-enethiol derivative from the C-terminal cysteine residue of the precursor peptide of epidermin and related peptides. The UV-visible spectra of the reaction products of EpiD are pH-dependent, indicating that the enethiol side chain is converted to an enethiolate anion. The pKa value of the enethiol group was determined to be 6.0 and is substantially lower than the pKa value of the thiol side chain of cysteine residues. The increased acid strength of the enethiol side chain compared with that of the thiol group is attributed to the resonance stabilization of the negative charge of the anion.

Several new posttranslational modification reactions, such as dehydration of serine and threonine residues, thioether formation, and formation of D-alanine residues from L-serine residues, are involved in the biosynthesis of the ribosomally synthesized lantibiotics (reviewed in Refs. 1 and 2). The lantibiotic epidermin contains the ␣,␤-unsaturated amino acid didehydroaminobutyric acid, S-[(Z)-2-aminovinyl]-D-cysteine, 3-methyl-lanthionine, and meso-lanthionine (3). The formation of epidermin from the precursor peptide EpiA (4) includes the oxidative decarboxylation of the C-terminal cysteine residue of EpiA to a (Z)-enethiol catalyzed by the FMN-containing enzyme EpiD (5,6). Two reducing equivalents from the C-terminal cysteine residue are removed, a double bond is formed, and the coenzyme FMN is reduced to FMNH 2 . The decarboxylation occurs spontaneously or is also catalyzed by EpiD. The (Z)enethiol derivative is the intermediate in the formation of the C-terminal S-[(Z)-2-aminovinyl]-D-cysteine residue of epidermin (6). The unusual enethiol structure has been confirmed by using mass spectrometry, tandem mass spectrometry, UV-visible spectroscopy, two-dimensional NMR spectroscopy, and conversion of the enethiol to a mixed disulfide with 5,5Ј-dithiobis(2-nitrobenzoic acid) (Ellman's reagent (7,8)). EpiD has a low substrate specificity, and most of the peptides with the sequence [V/I/L/(M)/F/Y/W]-[A/S/V/T/C/(I/L)]-C at the C terminus are substrates of EpiD, as elucidated by analysis of the reaction of EpiD with single peptides and peptide libraries, respectively (7).
Here we report the further characterization of the reaction products of EpiD. We demonstrate that at physiological pH, the dominant form of the reaction product is the enethiolate anion and that the pK a value of the enethiol group is 6.0. We assume that the enethiolate anion is strongly nucleophilic and is the substrate for the Michael addition-like formation of S-[(Z)-2aminovinyl]-D-cysteine. Furthermore, we discuss the possibility that enzymes exist in nature that catalyze the formation of enethiol groups from non-C-terminal cysteine residues.

Synthesis and Purification of SFNSYV-NH-CHϭCH-SH and SFN-
SYT-NH-CHϭCH-SH-SFNSYVC and SFNSYTC were synthesized as described previously (7). For generation of the enethiol derivative of SFNSYVC, approximately 25 mg of peptide were dissolved in 50 ml of 20 mM Tris/HCl (pH 8.0) containing 4 mM DTT 1 and 1 mg of MBP-EpiD (5). For generation of the enethiol derivative of SFNSYTC, approximately 12 mg of peptide and 500 g of MBP-EpiD were used. After 25 min of incubation at 37°C, 500 l of trifluoroacetic acid was added. The reaction mixture was then separated by reversed-phase chromatography using a ProRPC HR10/10 column (Pharmacia Biotech Inc.). Peptides were eluted with a linear gradient of 0 -80% acetonitrile, 0.1% trifluoroacetic acid in 150 ml at a flow rate of 2 ml/min. Peptides with a high absorbance at 260 nm were collected and diluted with H 2 O, 0.1% trifluoroacetic acid. DTT was added to 3 mM, and the samples were subjected to a second reversed-phase chromatography using a RPC C2/C18 SC 2.1/10 column (Pharmacia). Peptides were eluted with a linear gradient of 0 -50% acetonitrile, 0.1% trifluoroacetic acid in 5.8 ml at a flow rate of 200 l/min. The absorbance was measured simultaneously at 214, 260, and 280 nm. The modified peptide was identified by the high A 260 /A 214 ratio and collected by the peak fractionation method using the SMART TM system (Pharmacia, Freiburg, FRG). Samples were dried with a vacuum concentrator (yields: approximately 4.4 mg of SFNSYV-NH-CHϭCH-SH and 1 mg of SFNSYT-NH-CHϭCH-SH, respectively) and stored at Ϫ70°C. The modified peptides were analyzed by electrospray mass spectrometry as described recently (7).
UV-visible Spectroscopy-UV-visible spectra were obtained with a Beckman DU 7500 spectrophotometer. Spectral data were recorded in the range of 200 -600 nm at 21°C at various pH values in quartz cuvettes (1-cm light path; the reference cuvette contained the used buffers and on occasion 4 mM DTT). For determination of the pK a , 0.6-ml aliquots of a stock solution of the oxidatively decarboxylated peptide SFNSYV-NH-CHϭCH-SH were mixed with 0.15 ml of 50 mM sodium phosphate (pK a ϭ 2.15, 7.20, and 12.38), 50 mM sodium citrate buffer (pK a ϭ 3.13, 4.76, and 6.40) (final buffer concentration was 10 mM each for phosphate and citrate). The pH of the buffers was adjusted by adding HCl. pH values of the 10 mM phosphate/citrate buffers and of the buffered peptide solutions were determined using a Knick pH meter (761 Calimatic).

RESULTS
pH Dependence of UV-visible Spectra of Oxidatively Decarboxylated Peptides-The reaction products of the flavoenzyme EpiD are oxidatively decarboxylated peptides with a C-terminal enethiol side chain (6 -8). Recently, we demonstrated that the reaction products of EpiD have an absorption maximum at 260 nm when measured in 0.1% trifluoroacetic acid, 30% ace-* This work was supported by grants from the Deutsche Forschungsgemeinschaft (SFB 323). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
tonitrile, 70% H 2 O (6, 7). In order to determine the acid strength of the -SH group of enethiol-containing peptides, we investigated whether the UV-visible spectra of SFNSYV-NH-CHϭCH-SH and SFNSYT-NH-CHϭCH-SH are pH-dependent. It is well established that the thiolate anion of cysteine can be detected by its absorbance at 240 nm (9,10) and that the pK a value of the -SH group of cysteine-containing peptides is in the range of 8.5-9.5.
The UV-visible spectrum of the peptide SFNSYV-NH-CHϭCH-SH purified by reversed-phase chromatography was determined in the pH range 2.5-8.0. To ensure that spectral changes were not due to oxidation of the peptide to the dimer, the UV spectra were recorded in the presence of 4 mM DTT. Interestingly, the UV spectrum was pH-dependent ( Fig. 1) with an absorption maximum of 259 nm under acidic conditions (pH Յ 4.2) and of 283 nm at pH 7.0. Due to ionization of the tyrosine residue, the absorbance increased significantly below 260 nm at pH 8.0; however, there was no further shift of the absorption maximum to wavelengths greater than 283 nm (data not shown). Similar pH-dependent spectral changes were also observed for the peptide SFNSYT-NH-CHϭCH-SH (data not shown but see Fig. 2).
The pH dependence of the UV spectrum of the oxidatively decarboxylated peptide indicates the ionization of the enethiol form to the enethiolate anion form (Fig. 1, A and B). From the UV spectra, the apparent pK a value of the enethiol group was determined to be pH 6.0 (Fig. 1B). The molar extinction coefficient (⑀) of SFNSYV-NH-CHϭCH-SH at 259 nm and of SFNSYV-NH-CHϭCH-S Ϫ at 283 nm was determined to be at least ⑀ Ϸ 6,800 and ⑀ Ϸ 7,200 M Ϫ1 cm Ϫ1 , respectively. However, this value may even be higher due to error in weighing out the small amounts of available peptide and due to the presence of sodium chloride and water in the peptide sample. Assuming that the peptides SFNSYVC and SFNSYV-NH-CHϭCH-SH have the same ⑀ 214 value, ⑀ 259 of SFNSYV-NH-CHϭCH-SH and ⑀ 283 of SFNSYV-NH-CHϭCH-S Ϫ were calculated to be approximately 10,000 M Ϫ1 cm Ϫ1 .
Reversed-phase Chromatography of Enethiol-containing Peptides under Acidic and Neutral Conditions-In order to confirm the pH dependence of the UV spectrum of enethiol-containing peptides, the reaction mixture containing the substrate peptide SFNSYTC (821 m/z), the reaction product, and MBP-EpiD was analyzed by reversed-phase chromatography under acidic conditions (0.1% trifluoroacetic acid, approximately pH 2.0) and under neutral conditions (25 mM TEAA, pH 7.0, Fig. 2). The reaction product (775 m/z) was identified by high resolution electrospray mass spectrometry as a monomer (data not shown). Degradation products (741 and 759 m/z) of the reaction product were also present. The A 260 /A 280 ratio of the reaction product was 2.3 in 0.1% trifluoroacetic acid/acetonitrile/ H 2 O and 0.6 in 25 mM TEAA/acetonitrile/H 2 O (compare Fig. 2,  A2 and B2). Comparable results were obtained for the reaction products of SFNSYVC and SFNSYCC (data not shown). DISCUSSION The determined pK a value of 6.0 for the enethiol group of SFNSYV-NH-CHϭCH-SH is significantly (3 pH units) lower than the reported pK a value for the thiol group of cysteine residues in small peptides. Peptides with a C-terminal cysteine residue in which the negative charge of the carboxyl group is removed (cysteine amides, a negative charge in the vicinity of the thiol group decreases the acidity of the thiol group (11)), as is also the case for the investigated oxidatively decarboxylated peptides SFNSYV-NH-CHϭCH-SH and SFNSYT-NH-CHϭCH-SH, are reported to have a pK a value in the range of pH 8.8 (12). The increased acid strength of the enethiol group compared with the acid strength of the thiol group is explained by a resonance effect in which the negative charge on the sulfur anion is delocalized in the enethiolate. Delocalization of the negative charge is the reason for stabilization of enolate ions so that the acidity of enols is increased compared with alcohols (13).
For several enzymes, such as the protein thiol-disulfide oxidoreductase DsbA from Escherichia coli, human glutathione transferase, and cysteine peptidases, it has been shown that the catalytic active cysteine thiol groups have a pK a value near 4.0 (14 -16). This low pK a value is partially explained by stabilization of the thiolate anion by interaction with histidine or lysine side chains in the tertiary structure of the proteins. It is very unlikely that the peptides SFNSYV-NH-CHϭCH-SH and SFNSYT-NH-CHϭCH-SH have a defined three-dimensional structure and that this structure is responsible for stabilization of the enethiolate anion.
There is a strong correlation between the low pK a value of the thiol group of Cys 30 of DsbA and the redox potential of DsbA (17). The two cysteine residues of the active site Cys 30 -Pro 31 -His 32 -Cys 33 form a disulfide bridge, and disulfide-bonded DsbA is a potent oxidant that generates protein disulfide bonds (18,19). Because of the correlation between the pK a value of the thiol group and the redox potential, we are interested in determining the redox potential of peptides containing both a cysteine residue and a C-terminal enethiol group (e.g. oxidatively decarboxylated precursor peptide EpiA), and we are interested in investigating the reaction between protein thioldisulfide oxidoreductases and such substrates.
As known for thiols (20), we assume that enethiols are far more reactive as the enethiolate anions. Therefore, reasonable chemical intuition suggests that the intermediate in the Michael addition-like formation of the S-[(Z)-2-aminovinyl]-D-cysteine residue of epidermin is the enethiolate and not the enethiol form of posttranslationally modified precursor peptide EpiA. Furthermore, we postulate that in the case of lanthionine formation the thiolate anions of the involved cysteine residues are stabilized by interaction with positive charges within the precursor peptide or by interaction with the enzyme responsible for the thioether formation (EpiC and/or EpiB). Glutathione transferase catalyzes a similar reaction, the nucleophilic addition of glutathione to xenobiotic compounds containing an electrophilic center. It has been discussed that the pK a of the thiol group of glutathione is shifted from 9.0 in aqueous solution to 6.6 in the active site of glutathione transferase so that the predominant species in the active site is the thiolate anion of glutathione (21).
A fascinating question is whether enzymes exist in nature that catalyze the formation of enethiols/enethiolates from cysteine residues that are not located at the C terminus. Principally, such a reaction can be involved in the formation of the thiazole backbone modification of the gyrase inhibitor microcin B17, whose structure has been elucidated recently (22,23) (Fig.  3). Enzymes involved in microcin B17 modification have not been characterized, and there is no experimental evidence for the existence of the proposed enethiol-forming activity in microcin B17 modification. By converting the thiol side chain of cysteine residues in peptides and proteins to enethiol/enethiolates the (pH dependence of the) activity and stability of proteins may be influenced to a large extent. However, nothing is known about the stability of these enethiol side chains. Peptides with a C-terminal enethiolate group are unstable, but they are stabilized in the presence of zinc ions (6 -8). In addition to the flavoprotein EpiD, another enzyme catalyzing an ␣,␤-dehydrogenation reaction of peptides has already been characterized; heme-containing L-tryptophan 2Ј,3Ј-oxidase from Chromobacterium violaceum is responsible for the synthesis of ␣,␤-dehydrotryptophan derivatives (24,25).
Acknowledgments-We thank Christoph Kempter for assisting in electrospray mass spectrometry and Karen A. Brune for critically reading the manuscript. The recently postulated mechanism for the posttranslational modification of -Gly-Cys-Gly-segments of the microcin precursor peptide to the thiazole backbone modification of microcin B17 involved the attack of the thiol group of cysteine on the peptide carbonyl of the preceding glycine residue, the elimination of water, and the oxidation (removal of H 2 (22,23)). Here we describe that, in principle, the oxidation can be the first reaction and that an enethiol is formed as the first intermediate. Due to the increased acidity of the enethiol, the reactive nucleophile is the enethiolate anion. A similar reaction is the equilibrium between thiamine (vitamin B 1 ) and its enethiolate form (26,27).