Evaluation of vinylsulfamides as sulfhydryl selective alkylation reagents in protein modification

Article abstract of DOI:10.1016/S0960-894X(02)00950-2

The preparation of several model vinylsulfamides is described. Their excellent selective reactivity towards sulfhydryl group with regards to amino group has been demonstrated by the kinetics study between a model vinylsulfamide and cysteine or lysine at different pHs.

Full text of DOI:10.1016/S0960-894X(02)00950-2

Bioorganic & Medicinal Chemistry Letters 13 (2003) 383–386
Evaluation of Vinylsulfamides as Sulfhydryl Selective
Alkylation Reagents in Protein Modification
Min Li,* Robert S. Wu, Jane S. C. Tsai and Salvatore J. Salamone
Roche Diagnostics Corporation, 9115 Hague Road, Indianapolis, IN 46250, USA
Received 11 August 2002;revised 29 October 2002;accepted 6 November 2002
Abstract—The preparation of several model vinylsulfamides is described. Their excellent selective reactivity towards sulfhydryl
group with regards to amino group has been demonstrated by the kinetics study between a model vinylsulfamide and cysteine or
lysine at different pHs.
# 2002 Elsevier Science Ltd. All rights reserved.
Michael acceptors such as vinylogous amino acid deri-
vatives1 and vinylsulfone-containing molecular enti-
ties2,3 have been utilized in the design of novel protease
inhibitors. In particular, the vinylsulfone moiety has
recently been incorporated into several specific peptide
sequences which resulted in potent inhibitors for pro-
teasomes.4,5 The latter have been implicated in a num-
ber of diseases;growing evidence has suggested that
inhibitors of proteasomes may represent potentially new
and unique treatments for inflammatory diseases6 and
cancer.7 In another area, vinylsulfones were employed
to functionalize polyethylene glycol (PEG)8 and one of
the potential applications for such functionalized PEG
is to attach themselves to certain therapeutic proteins
through alkylation. The PEGylated proteins are usually
more stable and have longer in vivo half lives, which result
in improved therapeutic profiles for these proteins. On the
other hand, vinylsulfamides, which are structurally similar
to vinylsulfones and may be easier to prepare, have not
received attention in the aforementioned applications. In
this report, we describe the synthesis of several model
vinylsulfamides and the determination of their intrinsic
reactivity that favors sulfhydryl over amino group.
presence of triethylamine (TEA). The role of TEA was
not only to neutralize the first mole of HCl, resulting
from the reaction between the sulfonyl chloride and the
amine, but also to effect the elimination of the second
mole of HCl, which led to the isolation of the desired
vinylsulfamide in a single work up step. Because of this
reason, two equivalents (or three equivalents in the case
where the amine was used in its salt form) of TEA were
required in the reaction. The initial attempt to prepare
compound 2b was made through the hydrolysis of 2a in
a mixture of 2 N NaOH/MeOH (3/10, v/v) at 45 ꢀC for
3 h;nevertheless, the vinyl group was attacked by the
hydroxyl anion to give compound 4 under this condi-
tion (Scheme 1).9 The free acid, 4-(aminomethyl)benzoic
acid, was therefore allowed to react directly with
2-chloroethanesulfonyl chloride (Scheme 2). The desired
compound, 2b, was formed along with a relative 40% of
the corresponding dimer (5) as the by-product. The
reaction mixture containing 2b and 5 were then con-
verted to their NHS esters and separated via column
chromatography.10 Based upon its anticipated different
reactivity towards sulfhydryl (by the vinylsulfamide
moiety) and amino (by the NHS ester moiety) groups,
the purified compound (2c) appears to be a useful het-
erobifunctional cross-linking agent for proteins.11 On
the other hand, compound 3 could deliver the fluor-
escent label12 to a specific site within proteins that con-
tain a free cysteine residue. These expectations rely on
the presumption that the vinylsulfamides really possess
the intrinsic preferred reactivity towards sulfhydryl
group, which has been proven experimentally and is
described in the next paragraph.
The synthesis of all model sulfamides (Table 1) except
compound 2c is illustrated in Table 1, in which
2-chloroethanesulfonyl chloride was allowed to react
with an equal molar amount of a particular amine in the
*Corresponding author. Tel.: +1-908-284-2283, fax: +1-908-237-1009;
e-mail: minli88@att.net.
0960-894X/03/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00950-2

384
M. Li et al. / Bioorg. Med. Chem. Lett. 13 (2003) 383–386
Table 1. Synthesis and structures of model vinylsulfamides1
It was demonstrated previously that the vinylsulfones
showed good selectivity towards sulfhydryl group with
regards to amino group.8 Here, we evaluated the reac-
tivity of the vinylsulfamides towards sulfhydryl and
amino groups by measuring the half-lives of the model
vinylsulfamide (1),13 under different pHs at room tem-
perature (21Æ1 ꢀC). The measurement experiments were
performed in phosphate buffers where five times molar
excess of cysteine or lysine was added versus 1 in order
to achieve pseudo-first order reaction condition (experi-
ment conditions listed in the Note under Table 2). The
concentrations (A, as represented by peak Area) of the
vinylsulfamide (1) was followed by HPLC and the Ln
(A/A0) was plotted against time (an example is shown in
Fig. 1). The half-life was then calculated from the slope
and the results are summarized in Table 2. Since the
goal of the present study was to obtain qualitative
kinetics, the reactions were performed in a thermostat-
controlled laboratory rather than a preciously con-
trolled temperature bath.14 Nevertheless, the pseudo-
first order plot displayed excellent linearity with a
correlation coeffeciency (R2) of 0.9967 (Fig. 1). At pH
8.1, vinylsulfamide 1 showed, respectively, a half-life of
27 min in the presence of five times more cysteine and a
half-life of 7.4 days (or 177.4 h) in the presence of five
times more lysine. In another word, the selectivity of the
vinylsulfamide (1) is approximately 390 times in favor of
sulfhydryl over amino group. Around neutrality (pH
7.2), 1 displayed a half-life of 77 min when cysteine was
present;at the same pH;the half-life of 1 was too slow
to measure in the presence of lysine within a limited
time frame. On the other hand, the half-life of 1
decreased to 29.6 h at pH 9.2 when lysine was present,
whereas in the case of cysteine at the same pH, the half-
Compd
R
Isolated
Yield (%)
1
47
46
2a
2b
Note 2
2c
3
23
63
Notes: (1) General reaction procedure: to an ice–water cooled solution of
2-chloroethanesulfonyl chloride in methylene chloride (0.2 M) was added one
equivalent of the amine (or its HCl salt) and one (or two) equivalent(s) of TEA,
respectively. After the resulting mixture was stirred at 0 ꢀC for 1 h, a second
equivalent of TEA was added and the reaction was continued for another 1 h.
The reaction mixture was then poured onto a mixture of saturated NaHCO3/ice,
followed by addition of more methylene chloride. The organic phase was sepa-
rated and the aqueous phase extracted with methylene chloride (3Â). All organic
solutions were combined and washed with saturated NaHCO3/ice–water (1/4, v/v,
1Â), chilled 0.5 M HCl (2Â), and brine (1Â). The washed solution was then
dried (Na2SO4) and evaporated to give the vinylsulfamide. (2) The compound
was not isolated at this stage. It was directly converted into its NHS ester and
then isolated through column chromatography.
Scheme 1.
Scheme 2.

M. Li et al. / Bioorg. Med. Chem. Lett. 13 (2003) 383–386
385
Table 2. Summary of half-lives of vinylsulfamide 1 in the presence of
five times molar excess of cysteine or lysine
free sulfhydryl groups were found to be alkylated within
3 h at room temperature.17 When the molar ratio was
raised to 200:1, all of the sulfhydryl groups were reacted
under the same condition.
pH
7.2
8.1
9.2
Cysteine
Lysine
77 min
ND2
27 min
7.4 days
ND1
29.6 h
In summary, we have shown that the vinylsulfamides
are readily prepared in most cases and their excellent
selectivity towards sulfhydryl group relative to amino
group has been demonstrated by the kinetics study
between the model vinylsulfamide and cysteine or
lysine. Since amino-containing compounds are widely
available or readily accessible, their conversion into the
corresponding vinylsulfamides via reaction with
2-chloroethanesulfonyl chloride should bring about a
variety of molecular entities that combine a particular
functionality (as in the cases of compounds 2c and 3)
with the sulfhydryl-selective vinylsulfamide moiety.
Such resulting compounds will be useful for use as site-
specific alkylation agents in protein/peptide modifi-
cation and may also find their utility in the inhibitor
design of proteasomes and certain proteases.
Note: The measurement of the half-lives was performed in 50 mM
phosphate buffers containing 10% EtOH and 5 mM EDTA at room
temperature (21Æ1 ꢀC). The reaction solutions were monitored by an
HP1100 HPLC system that was equipped with a Hypersil C18 column
(4.6Â150 mm, 5 mm) and a diode-array detector with detection set at
280 nm. The elution was effected by a gradient of 10–40% acetonitrile
relative to the aqueous component consisting of 0.1% trifluoroacetic
acid in water over a period of 30 min. Initial concentration of 1 was
1.33 mM. ND1: reaction rate was too quick to determine. ND2: reac-
tion rate was too slow to measure within the time frame of the current
study.
life was too quick to measure accurately with the
methodology used in this study.
Morpurgo et al. measured the half-lives of several com-
pounds bearing the sulfhydryl group in the presence of
ten time molar excess of vinylsulfone PEG.8 At pH 8.0,
N-acetylcysteine had a half-life of 34 min, which is very
comparable to the value of 27 min recorded in our case
at pH 8.1. In their study, Morpurgo et al. was unable to
determine the half-life of the amine-containing com-
pound (Na-acetyllysine methyl ester) at pH 8.0, since no
reaction could be detected after 24 h. This may be
attributable to the amine assay method used in their
study (TNBS assay15) which is less sensitive in detecting
small changes than the HPLC method employed in the
current study.
References and Notes
1. Hanzlik, R. P.;Thompson, S. A. J. Med. Chem. 1984, 27,
711.
2. Palmer, J. T.;Rasnick, D.;Klaus, J. L.;Bromme, D.
J. Med. Chem. 1995, 38, 3193.
3. Bromme, D.;Klaus, J. L.;Okamoto, K.;Rasnick, D.;Pal-
mer, J. T. Biochem. J. 1996, 315, 85.
4. Bogyo, M.;McMaster, J. S.;Gaczynska, M.;Tortorella,
D.;Goldberg, A. L.;Ploegh, H. L.
U.S.A. 1997, 94, 6629.
Proc. Natl. Acad. Sci.
5. Bogyo, M.;Shin, S.;McMaster, J. S.;Ploegh, H. L.
J. Agric. Food Chem. 1990, 38, 990.
6. Iqbal, M. A.;Ator, M. A. In High Throughput Screening
for Novel Anti-inflammatories;Khan, M., Ed.;Birkhauser
Verlag: Basel, 2000;pp 19–34.
To test the alkylating ability of the vinylsulfamides
towards the cysteine residue in a protein, compound 1
was treated with bovine serum albumin (BSA) which
contains one free sulfhydryl group at Cys-34.16 At a
molar ratio of 20:1 (4 vs. BSA) at pH 7.0, 23% of the
7. Teicher, B. A.;Ara, G.;Herbst, R.;Palombella, V. J.;
Adams, J. Clin. Cancer Res. 1999, 5, 2638.
8. Morpurgo, M.;Veronese, F. M.;Kachensky, D.;Harris,
J. M. Bioconj. Chem. 1996, 7, 363.
9. 1H NMR of the sodium salt of 4 (DMSO-d6) d 7.85 (d, 2H,
J=8.1 Hz), 7.70 (br, 1H), 7.25 (d, 2H, J=8.1 Hz), 4.13 (s,
2H), 3.62 (t, 2H, J=6.4 Hz), 3.21 (t, 2H, J=6.4 Hz).
10. Procedures of the reactions and separation by column
chromatography: to a solution of 2-choloroethanesulfonyl
chloride (345 mL, 3.3 mmoL) in 31 mL of DMF that was
cooled in an ice–water bath, was added 2-(amino-
methyl)benzoic acid (500 mg, 3.3 mmoL) and TEA (878 mL,
6.3 mmoL), respectively. The resulting mixture was stirred at
0 ꢀC for ꢁ1 h and a second batch of TEA (502 mL, 3.6 mmoL)
was added. The reaction mixture was allowed to warm to
room temperature overnight and then filtered through a sin-
tered glass funnel. The filtrate was evaporated in vacuo and
the residue, after re-dissolved in a small amount of DMF, was
poured into an ice-chilled 1 N HCl solution. The mixture was
extracted with methylene chloride (5Â) and the organic extract
1
yielded 350 mg residue after removal of the solvent. H NMR
Figure 1. Linearity plot of the pseudo first order reaction between
cysteine and vinylsulfamide 1 at pH 8.1. Note: The x-axis is the reac-
tion time in min. The y-axis is Ln [A/A0] where A is the concentration
of vinylsulfamide 1 and A0 is the concentration at time 0. The pseudo
first order reaction constant, k1, equals to the absolute value of the
slope. The half-life, t1=2, can therefore be calculated as t1=2=0.693/
k1=27 min.
analysis of the residue indicated that it contained 2b and the
dimer (5).10a The mixture was converted to their NHS esters
according to the published procedure,10b and the reaction
mixture was loaded onto a flash silica gel column which was
eluted with ethyl acetate. The fast-eluted fractions gave the
desired NHS ester, 2c. 1H NMR (CDCl3) d 8.10 (d, 2H, J=8.3

386
M. Li et al. / Bioorg. Med. Chem. Lett. 13 (2003) 383–386
Hz), 7.47 (d, 2H, J=8.3 Hz), 6.49 (dd, 1H, J1=16.6 Hz,
J2=9.2 Hz), 6.27 (d, 1H, J=16.6 Hz), 5.95 (d, 1H, J=9.2 Hz),
4.82 (t, 1H, J=6.1 Hz), 4.32 (d, 2H, J=6.1 Hz), 2.92 (s, 4H).
(a) Part of the 2-chloroethanesulfonyl chloride molecules
apparently acted as the activation reagent by reacting with 2b
to form a ‘mixed anhydride’ intermediate. Attack of this
intermediate by a second molecule of 2b yielded the dimer 5.
(b). Li, M.;Tsai, S.-F.;Rosen, S. M.;Wu, R. S.;Reddy, K. B.;
DiCesare, J.;Salamone, S. J. . J. Agric. Food Chem. 2001, 49,
1287.
where specified) (CDCl3) d 8.18 (d, 2H, J=8.8 Hz), 7.51 (d,
2H, J=8.8 Hz), 6.51 (dd, 1H, J1=9.6 Hz, J2=16.4 Hz), 6.25
(d, 1H, J=16.4 Hz), 5.96 (d, 1H, J=9.6 Hz), 5.11 (t, 1H,
J=6.4 Hz), 4.30 (d, 2H, J=6.4 Hz). ESI-MS: m/z 243 (M +
H+). (2) The adduct (6) formed between cysteine and 1 was
1
isolated and characterized by H NMR: (D2O/DCl) d 7.41 (d,
2H, J=8.7 Hz), 6.79 (d, 2H, J=8.7 Hz), 3.62 (s, 2H), 3.56 (m,
1H), 2.65 (m, 2H), 2.49–2.27 (m, 2H), 2.15 (m, 2H). (3) The
yield for the adduct (7) derived from lysine and 1 was quite
low isolation of the product was not attempted.
11. (a) For a comprehensive review of heterobifunctional
cross-linking reagents, see: (a)Wong, S. S. Chemistry of Pro-
tein Conjugation and Cross-Linking;CRC: Boca Raton, FL,
1991. (b) Hermanson, G. T. Bioconjugate Techniques;Aca-
demic Press: San Diego, CA, 1996.
1
12. Characterization of 3: H NMR (400 MHz, DMSO-d6) d
8.05 (t, 1H, J=5.9 Hz), 7.75 (d, 1H, J=8.8 Hz), 7.03 (d, 1H,
J=2.5 Hz), 6.98 (dd, 1H, J1=8.8 Hz, J2=2.5 Hz), 6.86 (dd,
1H, J1=16.4 Hz, J2=9.9 Hz), 6.31 (s, 1H), 6.12 (d, 1H,
J=16.4 Hz), 6.04 (d, J=9.9 Hz), 4.31 (d, 2H, J=5.9 Hz), 3.87
(s, 3H). 13C NMR (100 MHz, proton-decoupled, DMSO-d6) d
162.45, 160.11, 154.89, 152.39, 136.65, 126.03, 125.78, 112.21,
110.84, 109.76, 100.90, 55.94, 42.33. FAB-MS: m/z 296.2
(M+H+). IR: 3240, 3094, 1720, 1610, 1139 cmÀ1. UV
(MeOH) 217 (e 19,117), 322 nm (e 13,773). The corresponding
fluorescent 4-(aminomethyl)-7-methoxycoumarin was synthe-
sized according to the procedure of Li and White: Li, M.;
White, E. H. Bioconjugate Chem. 1994, 5, 454.
14. Such a practice has been adopted in the kinetics study of
Morpurgo et al.8
15. Snyder, S. L.;Sobocinsky, P. Z. Anal. Biochem. 1975, 64,
284.
16. Brown, J. R.;Shockley, P. In Serum Albumin: Structure
and Characterization of Its Ligand Binding Sites in Lipid Pro-
tein Interactions;Jost, P. C., Griffith, O. H., Eds.;John Wiley
& Sons: New York, 1982;Chapter 2, pp 25–68.
17. The content of the free sulfhydryl groups were assayed
using the Ellman’s reagent: Riddles, P. W.;Blakeley, R. L.;
Zerner, B. Methods Enzymol. 1983, 91, 49.
13. (1) Characterization of 1: 1H NMR (200 MHz except