[4156-16-5]  · C11H11NO4S  · N-Ethyl-5-phenylisoxazolium-3-sulfonate  · (MW 253.30)

(coupling reagent for peptide synthesis;2 protein modification;3 reduction of carboxylic acids into alcohols4)

Alternate Names: Woodward's reagent K; NEPIS.

Physical Data: mp 206-208 °C (dec); UV lmax (ε) 283 nm (22 500) in 0.1 N HCl.

Solubility: insol MeCN, MeNO2, DMF, CH2Cl2; sol H2O.

Form Supplied in: crystalline, anhydrous zwitterion; widely available as 95% pure that may contain some of the para sulfonate derivative.

Analysis of Reagent Purity: material that is sufficiently pure has mp 206-208 °C.

Purification: dissolve in aqueous 1 N HCl and reprecipitate by the slow addition of acetone. After filtration and drying the reagent is a fluffy solid.

Handling, Storage, and Precautions: stable to light and temperature and nonhygroscopic. No extra precautions need to be observed.

Peptide Coupling Reagent.

Woodward's reagent K reacts with an N-protected amino acid, under mild conditions, in the presence of a tertiary amine to form an enol ester intermediate, e.g. (1). This enol ester acylates an amino acid ester or peptide to afford the elongated peptide derivative (eq 1). The isoxazolium salt is not soluble in organic solvents, but as it reacts with a carboxyl group in the presence of a tertiary amine the reaction mixture clears.

Examples of di- and tripeptide couplings abound with yields typically >80%; minimal purification steps are required since the byproducts are water soluble (eq 2).2

The recommended coupling procedure is to activate the carboxyl group in MeNO2 or MeCN with vigorous stirring for 10 min at rt or 1 h at 0 °C. Then the amino ester, dissolved in MeNO2 or MeCN, is added and the coupling allowed to go at rt for several hours.

Good yields are obtained in coupling asparagine and glutamine. With some coupling reagents the carboxamide of these residues can undergo an intramolecular dehydration when the a-carboxylate is activated to give a nitrile residue (2) in the peptide.5 However, this byproduct has not been observed with reagent K.

An additional advantage of reagent K is that when serine, threonine, etc., are activated their hydroxyl groups do not require protection.6 For instance, the hydroxyproline tripeptide in eq 3 was obtained in good yield after crystallization.2

Several studies7 on the extent of racemization with peptide coupling reagents have been published. This generally is a problem when performing a fragment condensation, but normally not a problem when the amino group is protected as a carbamate. For best control of the extent of racemization, accurate measurement of the tertiary amine is advised.7b The extent of racemization observed with reagent K is just a few percent for acylated amino acids,7c which is no worse than most coupling reagents except for the azide coupling procedure. There are some disadvantages of this reagent in peptide couplings. The most serious one is the limitation of solvents. Acetonitrile and nitromethane are the best for high yield, high optical purity products. However, many peptides are not readily soluble in these solvents. The next disadvantage is that the enol ester intermediate can rearrange to the imide side-product (3) which is difficult to remove.1 Another problem with reagent K is its incompatibility with solid-phase peptide synthesis because of its poor solubility in organic solvents and relatively slow activation of carboxyl groups.8

A good comparison of reagent K with other reagents is in the synthesis of a protected derivative of oxytocin (eq 4).9 This product was further elaborated to give oxytocin. The results are compared with those for 1,3-Dicyclohexylcarbodiimide and N,N-Carbonyldiimidazole in Table 1.

An additional utility for reagent K is peptide cyclization.10 The activation and cyclization steps are performed separately; therefore each can be carried out under the appropriate dilution conditions. This is advantageous when the carboxyl group activates slowly since it is in concentrated solution.

The isoxazolium salt has been used in the synthesis of other difficult peptide structures. A peptidoglycan has been obtained in good yield by this procedure (eq 5).11

Additionally, this reagent has been used repetitively in the synthesis of a steroid-peptide adduct with a free hydroxyl group (eq 6).12 Deprotection and another coupling with Cbz-Arg(NG-nitro) gave the dipeptide-steroid adduct in good yield. In both couplings the steroid hydroxyl group was not acylated.

Protein Modification.

Since this reagent is water soluble and is reactive under aqueous conditions below pH 4.75, it can specifically modify carboxyl groups in proteins.3,13 This has proven useful in determining essential carboxyl groups in various enzymes. The enol ester is stable enough to isolate the modified protein and tag with a nucleophile for eventual identification of the modified amino acid. Other reagents such as the water-soluble carbodiimides are also used extensively in these mechanistic studies. However, an advantage with reagent K is that the enol ester formed has a UV absorbance at 340 nm (E340 = 7000 M-1 cm-1);14 thus the number of carboxyl groups modified can be quantified.

Reduction of Carboxylic Acids to Alcohols.

The conversion of carboxylic acids to alcohols occurs in a two-step process under mild conditions using reagent K with yields for simple alcohols ranging from 50-100%.4 The reduction of a dipeptide (eq 7) suggests a potential utility in modifying carboxyl groups in proteins.

Another method15 to obtain alcohols is to reduce the mixed carbonic acid anhydride prepared with Ethyl Chloroformate with an excess of Sodium Borohydride. This method uses less costly reagents; thus it is more useful for synthetic transformations.

Miscellaneous Reactions.

Pantothenic acid has been coupled to a long chain amine with reagent K (eq 8).16 This is another example of the nonreactivity of hydroxyl groups with the enol ester intermediate. Attempts to prepare this amide through the mixed anhydride procedure give low yields.

Reagent K may have utility in sequencing peptides and proteins from the C-terminus (eq 9).17 The 2-thiohydantoin produced is used to analyze a peptide sequence upon cleavage of the thiohydantoin from the peptide. This mild procedure gives good yields with amino acids that have sensitive side chains such as serine and lysine. The original Acetic Anhydride procedure18 for C-terminal sequencing does not work well with these amino acids; therefore this procedure is an improvement. Another activation reagent, hydroxysuccinimide esters of Fmoc-amino acids, fails to give the thiohydantoin.17

Related Reagents.

Benzotriazol-1-yloxytris(dimethylamino)phosphonium Hexafluorophosphate; 1-Cyclohexyl-3-(2-morpholinoethyl)carbodiimide; 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide Hydrochloride; Isobutyl Chloroformate; p-Nitrophenol; Pentafluorophenol; 1,1-Thionylimidazole.

1. Woodward, R. B.; Olofson, R. A.; Mayer, H. T 1969, 22(S8), 321.
2. Woodward, R. B.; Olofson, R. A. OS 1977, 56, 88.
3. Bodlaender, P.; Feinstein, G.; Shaw, E. B 1969, 8, 4941.
4. Hall, P. L.; Perfetti, R. B. JOC 1974, 39, 111.
5. Gish, D. T.; Katsoyannis, P. G.; Hess, G. P.; Stedman, R. J. JACS 1956, 78, 5954.
6. Klausner, Y. S.; Bodanszky, M. S 1972, 453.
7. (a) Kemp, D. S. The Peptides: Analysis, Synthesis and Biology; Academic: New York, 1979; Vol. 1, pp 315-383. (b) Woodward, R. B.; Woodman, D. J. JOC 1969, 34, 2742. (c) Kemp, D. S.; Wang, S. W.; Busby, G., III; Hugel, G. JACS 1970, 92, 1043.
8. Hudson, D. JOC 1988, 53, 617.
9. Fosker, A. P.; Law, H. D. JCS 1965, 4922.
10. Blaha, K.; Rudinger, J. CCC 1965, 30, 3325.
11. Merser, C.; Sinay, P.; Adam, A. Biochem. Biophys. Res. Commun. 1975, 66, 1316.
12. (a) Pettit, G. R.; Gupta, A. K. D.; Smith, R. L. CJC 1966, 44, 2023. (b) Pettit, G. R.; Smith, R. L.; Klinger, H. JOC 1967, 10, 145.
13. Baker, A. J.; Weber, B. H. JBC 1974, 249, 5452.
14. Sinha, U.; Brewer, J. M. Anal. Biochem. 1985, 151, 327.
15. Ramsamy, K.; Olsen, R. K.; Emery, T. S 1982, 42.
16. Wagner, A. P.; Retey, J. Eur. J. Biochem. 1991, 195, 699.
17. Boyd, V. L.; Hawke, D. H.; Geiser, T. G. TL 1990, 31, 3849.
18. Cromwell, L. D.; Stark, G. R. B 1969, 8, 4735.

Richard S. Pottorf

Marion Merrell Dow Research Institute, Cincinnati, OH, USA

Copyright 1995-2000 by John Wiley & Sons, Ltd. All rights reserved.