We are interested synthesizing molecules of the form RC(O)CH2SCH2CH(NH2)CO2H, especially where R = methyl or R = a catechol. Baker and coworkers (Goodman et al., J Org Chem 1958, pp. 1251-1258) synthesized an amino acid where R = phenyl, but they noted that it decomposed quickly in hot water, the recrystallizing solvent. The evidence for possible cyclization to an imine was a little ambiguous. Their ninhydrin test gave a pale yellow color, but by IR they saw an absorption corresponding to an aromatic ketone. They used chloroacetophenone or bromoacetophenone as their electrophile. They also attempted to use chloroacetone, but they saw highly colored, nonhomogeneous solids and nothing seemed to be the desired product (R = methyl).
Apparently, alpha-thioketones can react in the presence of thiols to produce ketones. A friend of mine wrote, “Alpha-carbonyl sulfides are reduced to the parent carbonyl
compounds by thiols (Ōki, M., Funakoshi, W. and Nakamura, A. (1971) Reaction
of α-carbonyl sulfides with bases. I. Reaction between α-carbonyl sulfides
with thiolates. Bull. Chem. Soc. Japan 44, 828-832).” A quick reading of this paper suggests two things: One, when one mole of alpha-thioketone reacts with one mole of thiolate, a ketone is produced. Two, when a thiolate reacts with an alpha-chloroketone, both the substitution product (alpha-thioketone) and the reduction product (ketone) may be observed. Unfortunately, the authors did only a limited study of the effects of changing the reaction stoichiometry. This study used simple aryl and alkyl groups, not amino acid derivatives. I wonder in the case where only one mole of thiolate is use, whether or not the order of addition (thiol to chloroketone vs. chloroketone to thiol) makes any difference.
There are a few of papers that use chloroacetone to inactivate enzymes, including at least one that has a nucleophilic cysteine residue. One of the studies used 2-mercaptoethanol as a model for a nucleophilic cysteine residue at the active site of an enzyme. They followed the reaction between chloroacetone and 2-mercaptoethanol by loss of reactivity to Ellman’s reagent (DTNB), but they did not characterize the product, apparently. Holloway et al., Biochem. J. (1980) 191, 811-826.
A couple of years ago we synthesized a compound with R = 3-pyridinyl and also the homocysteine version of the same molecule (unpublished). We did not look into the issue of possible cyclization, but we did take C-13 data. We did not do a formal analysis of the H-1 NMR splitting. I seem to recall seeing a ketone carbon in the C-13 NMR spectrum, but I wasn’t thinking about imines at the time. I plan to double-check the C-13 data.
We have attempted to alkylate cysteine with chloroacetone using a variety of bases and solvents. We see a loss of reactivity to DTNB. We have seen a bright orange-red compound or compounds, but we have not tried to characterize the colored material. We are just starting to try N-acetylcysteine, to see whether or not it produces the desired compound.
My tentative conclusion is that some alpha-thioketones are accessible, but I don’t yet know why some are difficult to make and some are not so difficult. Any thoughts or suggestions on what might be happening in the chloroacetone reactions or what we should try next to obtain our desired products would be appreciated.