Angewandte
Chemie
alanine residues. These NCL global desulfurization proto-
cols[7] have since been employed in the synthesis of a number
of cysteine-free proteins with some success. In the course of
our own efforts to synthesize large, complex glycoprotein
targets we examined the feasibility of such methods for the
reduction of cysteine to alanine. In particular, we sought to
determine whether such metal-based protocols are compat-
ible with the extensive functionality present in our glycopep-
tide substrate systems. In this context, Pentelute and Kent
recently reported that the Raney nickel method can effec-
tively accommodate both methionine and the acetamido-
methyl (Acm) functionality, which is commonly employed as
a protecting group for cysteine.[8] However, a significant
drawback of this method is that it requires large excesses of
nickel.[9] Furthermore, the use of Raney nickel can cause
epimerization of secondary alcohols[10] and the reduction of
thiols, thioethers, and thioesters.[11] Meanwhile, although
Wong and co-workers have found that another metal-based
protocol, which utilizes Pd/Al2O3, is also able to accommo-
date both methionine and the Acm functionality,[12] we
observed that the thiazolidine (Thz) moiety, which serves as
an ideal masking group for N-terminal cysteine residues, is
not stable under these conditions.[13] Thus, given these
limitations, and given the complexity and level of function-
alization of our glycopeptide constructs, we sought to develop
a mild, nonmetal-based reduction method for cysteine. We
would require that such a method be tolerant of a range of
functional groups, including carbohydrate sectors, various
amino acids (particularly methionine), and a range of sulfur-
containing groups, such as Cys(Acm), biotin, Thz, and
thioesters.
to effect the reduction of model peptide substrates were met
with very limited success. In addition to small amounts of the
desired reduced adduct, we typically observed extensive side
products. We suspected that the problems observed might be
attributable to issues of phosphite solubility in aqueous
solution.
Accordingly, we came to consider the possibility of
utilizing trialkylphosphines to effect the desired cysteine
reduction in peptide settings. In particular, we were drawn to
tris(2-carboxyethyl)phosphine (TCEP), which has enjoyed
wide use as a disulfide reducing agent in peptide and
glycopeptide settings.[17,18] Many considerations served to
identify TCEP as the phosphine source. Most importantly, its
ability to tolerate a range of glycopeptide functionality is well-
established. Furthermore, it is easily manipulated in air and
reacts readily in aqueous solution over a wide pH range. We
selected, as the radical initiator, the water-soluble 2,2’-
azobis[2-(2-imidazolin-2-yl)propane]dihydrochloride (VA-
044), which has a very low temperature of decomposition.[19]
In the event, peptide 1 (Fmoc-RYKDSGCAHPRG-OH)
was exposed to TCEP, tBuSH, water, and VA-044 at room
temperature. We were pleased to observe nearly quantitative
conversion of the cysteine residue into alanine within 10 h, as
determined by LCMS (Figure 2). Following purification,
We began by taking note of a disclosure by Hoffmann
et al. in 1956 which described a desulfurization reaction
between mercaptan and trialkylphosphite derivatives under
both thermal and photochemical conditions.[14] Soon there-
after, in a key advance, Walling and Rabinowitz put forth a
proposed mechanistic sequence, wherein an alkylthiyl radical
adds reversibly to phosphite, thereby generating an inter-
mediate phosphoranyl radical.[15] Subsequent b scission was
envisioned to provide an alkyl radical, and rapid hydrogen
abstraction from the parent thiol would furnish the product
alkane, thereby serving to propogate the chain (Scheme 1). In
addition to this contribution to the mechanistic understanding
of the reduction, Walling et al. extended the reaction to
trialkylphosphines.[15b]
Figure 2. Model study for the selective free-radical desulfurization of
Cys to Ala in water at RT. LCMS chromatograms of: A) cysteinyl
peptide 1; B) crude products observed 3 h after treatment with TCEP,
tBuSH, VA-044 at RT; C) crude alanyl peptide 2 after 10 h. VA-
044=2,2’-azobis-[2-(2-imidazolin-2-yl)propane]dihydrochloride.
Fmoc=9-fluorenylmethoxycarbonyl.
On the basis of these findings, Valencia and co-workers
have developed a method by which cysteine can be reduced to
alanine through the action of triethylphosphite and a borane
radical initiator.[16] Our own efforts to apply such conditions
peptide 2 was isolated in 82% yield. We next prepared and
evaluated a series of peptide substrates, incorporating a range
of relevant functional groups (Table 1). Importantly, our
reaction conditions were able to efficiently accommodate a
variety of important functionalities, including methionine
(entry 4), Cys(Acm) (entry 5), Thz (entry 6), and biotin
(entry 7). To our knowledge, this TCEP-based method
Scheme 1. Proposed mechanism of radical desulfurization reaction.
Angew. Chem. Int. Ed. 2007, 46, 9248 –9252
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
&
&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
Take advantage of blue reference links
&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
Wan, Qian
