Synthesis and biological activity of N-sulfonylphosphoramidates: Probing the electrostatic preferences of alkaline phosphatase

Article abstract of DOI:10.1021/jo010495q

N-Sulfonylphosphoramidates have been synthesized to investigate the electrostatic requirements for binding to alkaline phosphatase. Alkyl- and aryl N-benzylated sulfonamides were phosphorylated with bromophosphates or synthesized via phosphoramidite chemistry in moderate yields (44-77%.) The resulting tribenzylated N-sulfonylphosphoramidates may be deprotected in one step to give the free acids in quantitative yields. Physical data of N-sulfonylphosphoramidates, including pK_a's and stability toward hydrolysis, were determined. Inhibition data suggests that AP does not bind trianionic N-sulfonylphosphoramidates better than dianionic N-sulfonylphosphoramidates, although N-sulfonylphosphoramidates are bound tighter than N-phenylphosphoramidate. k_cat for the hydrolysis of N-sulfonylphosphoramidates by bovine and E. coli alkaline phosphatases is 10-60% that of p-nitrophenyl phosphate.

Full text of DOI:10.1021/jo010495q

VOLUME 66, NUMBER 23
NOVEMBER 16, 2001
© Copyright 2001 by the American Chemical Society
Articles
Syn th esis a n d Biologica l Activity of N-Su lfon ylp h osp h or a m id a tes:
P r obin g th e Electr osta tic P r efer en ces of Alk a lin e P h osp h a ta se
Benjamin T. Burlingham and Theodore S. Widlanski*
Department of Chemistry, Indiana University, Bloomington, Indiana 47405
twidlans@indiana.edu
Received May 15, 2001
N-Sulfonylphosphoramidates have been synthesized to investigate the electrostatic requirements
for binding to alkaline phosphatase. Alkyl- and aryl N-benzylated sulfonamides were phosphorylated
with bromophosphates or synthesized via phosphoramidite chemistry in moderate yields (44-77%.)
The resulting tribenzylated N-sulfonylphosphoramidates may be deprotected in one step to give
the free acids in quantitative yields. Physical data of N-sulfonylphosphoramidates, including pKa’s
and stability toward hydrolysis, were determined. Inhibition data suggests that AP does not bind
trianionic N-sulfonylphosphoramidates better than dianionic N-sulfonylphosphoramidates, although
N-sulfonylphosphoramidates are bound tighter than N-phenylphosphoramidate. kcat for the
hydrolysis of N-sulfonylphosphoramidates by bovine and E. coli alkaline phosphatases is 10-60%
that of p-nitrophenyl phosphate.
In tr od u ction
Two families of enzymes, kinases and phosphatases,
play important roles in maintaining the phosphorylation
state of proteins and small molecules vital for a wide
range of cellular processes.1 About half of all phos-
phatases are metallophosphatases. Alkaline phosphatase
F igu r e 1. Schematic representation of phosphate monoester
binding in the active site of alkaline phosphatase. The spheres
represent active site zinc atoms.
(AP), a well-studied member of this group, has served as
a model for the study of metallophosphatases. AP has a
shallow, positively charged active site containing an
arginine reside, two zinc ions, and a magnesium ion.2
Along with the arginine residue, the metal ions serve the
dual purpose of facilitating the binding of the negatively
charged phosphate moiety and activating the substrate
for attack by a nucleophilic serine residue (Figure 1).
Because the active site is shallow, the enzyme binds only
the phosphate moiety of the substrate, and recognition
depends little on the alcohol portion of the substrate.3
This fact makes it difficult to design inhibitors for this
enzyme. Because electrostatic interactions of the phos-
phate moiety with the enzyme are the main binding
elements in the active site of AP, a more detailed picture
* Corresponding author: Fax: 812-855-8300. Phone: 812-855-6863.
(1) Posada, J .; Cooper, J . A. Mol. Biol. Cell. 1992, 3, 583-592.
(2) Sowadski, J . M.; Handschumacher, M. D.; Krishna Murthy, H.
M.; Foster, B. A.; Wyckoff, H. W. J . Mol. Biol. 1985, 186, 417-433.
(3) Kim, E. E.; Wyckoff, H. W. J . Mol. Biol. 1991, 218, 449-464.
10.1021/jo010495q CCC: $20.00 © 2001 American Chemical Society
Published on Web 08/21/2001

7562 J . Org. Chem., Vol. 66, No. 23, 2001
Burlingham and Widlanski
F igu r e 4. N-Sulfonylphosphoramidate monoesters.
F igu r e 2. Schematic representation of inorganic phosphate
binding in the active site of alkaline phosphatase. The spheres
represent active site zinc atoms.
F igu r e 5. N-Sulfonylphosphoramidates.
pKa on the order of 7.5-10, depending on the nature of
the appended substituent. This range of pKa’s suggests
that a series of N-sulfonylphosphoramidates could be
constructed, some being primarily dianionic and others
being primarily trianionic at pH 9, the pH optimum of
AP.
F igu r e 3. A potential motif for alkaline phosphatase inhibi-
tors involves a bridging atom (X) capable of bearing a negative
charge. N-Sulfonylphosphoramidates fit this motif.
Sometime ago, N-sulfonylphosphoramidate monoesters
(Figure 4) were tested as substrates of AP, but function
only as weak inhibitors.8 In light of more recent struc-
tural knowledge of the active site of AP, it appears that
N-sulfonylphosphoramidate monoesters may be sterically
precluded from binding productively to the shallow active
site. On the other hand, N-sulfonylphosphoramidate
diacids, which more closely resemble phosphate mo-
noesters (the natural substrate), may not have this
difficulty. To test the binding of N-sulfonylphosphora-
midates to AP, three target compounds were chosen
(Figure 5). N-(Methylsulfonyl)phosphoramidate (3a ) is
the smallest possible N-sulfonylphosphoramidate, while
N-(phenylsulfonyl)phosphoramidate (3b) is a representa-
tive aryl N-sulfonylphosporamidate. The additional
strongly electron-withdrawing group of N-(trifluoro-
methylsulfonyl)phosphoramidate (3c) ensures that this
compound will have significantly different ionization
properties than either an alkyl- or aryl N-sulfonylphos-
phoramidate and will allow us to test the binding of a
trianionic molecule to AP.
of these interactions is critical if we are to develop a full
understanding of the enzyme and improve our ability to
design inhibitors of it.
The binding of inorganic phosphate by alkaline phopha-
tase provides a model for electrostatic active site inter-
actions. Inorganic phosphate is one of the best inhibitors
of AP, with a Ki of about 5 µM for E. coli AP.4 In fact,
inorganic phosphate is bound tighter than a phosphate
monoester substrate under conditions of high pH. Figure
2 depicts a putative model for inorganic phosphate
binding to AP. It is unclear, however, whether inorganic
phosphate is bound as the di- or trianion. A recent study
suggests that an active site arginine present in AP may
be deprotonated and thus capable of accepting a hydrogen
bond from an H-bond donor.5 Accordingly, the enhanced
binding of inorganic phosphate, as compared to a phos-
phate monoester, may be due to a hydrogen bond between
the neutral arginine and a protonated phosphate oxygen.
On the other hand, it is possible that inorganic phosphate
binds as the trianion and thus takes maximum advan-
tage of electrostatic interactions with positively charged
groups in the active site. In an effort to gain further
information about the effect of substrate/inhibitor charge
on binding to AP, we sought to design synthetic molecules
that would predominately exist either in the dianionic
or trianionic form as a function of substitution pattern.
A phosphate monoester mimic with the potential to be
trianionic may be obtained by replacing the bridging
oxygen of the ester with a nitrogen atom (Figure 3).
Phosphoramidates have been tested as substrates for E.
coli AP, and previous reports suggest that they are
hydrolyzed with a kcat value 30-60% that of phosphate
monoesters.6 However, at biologically relevant pH’s, alkyl
or aryl phosphoramidates are not sufficiently acidic to
exist predominately as the trianion. Attaching a strongly
electron-withdrawing substituent to the nitrogen lowers
the third pKa of the phosphoramidate. N-Sulfonylphos-
phoramidates (Figure 3) have recently been tested as
inhibitors of carbonic anhydrase.7 They possess a third
Resu lts
A number of methods for the synthesis of N-sulfonyl-
phosphoramidate esters have been developed.9 Early
attempts at the synthesis of N-sulfonylphosphoramidic
acids and esters involved the reaction of sulfonamides
with PCl5, followed by hydrolysis or alcoholysis.10 This
hydrolysis may be affected by careful addition of formic
acid, although the product is susceptible to acid cleavage
of the P-N bond, releasing inorganic phosphate. This
four-step procedure was recently used to synthesize
inhibitors of carbonic anhydrase, though no yields were
reported.11 Alternate methods for the synthesis of N-
sulfonylphosphoramidate esters include a derivative of
the Arbuzov reaction, reaction of phosphite with N-
sulfonyl azides, and a variation of the Atherton-Todd
(7) Fenesan, I.; Popescu, R.; Scozzafava, A.; Crucin, V.; Mateiciuc,
E.; Bauer, R.; Ilies, M. A.; Supuran, C. T. J . Enzyme Inhib. 2000, 15,
297-310.
(8) Goldstein, J . A. J . Org. Chem. 1977, 42, 2466-2469.
(9) Lukanov, L. K.; Venkov, A. P.; Mollov, N. M. Synth. Commun.
1986, 16, 767-773.
(4) Widlanski, T. S.; Taylor, W. In Comprehensive Natural Product
Chemistry; Barton, D. B.; Nakanishi, K., Eds.; Elsevier: Oxford. 2000;
Vol. 4, Ch. 7.
(5) Holtz, K. M.; Stec, B.; Myers, J . K.; Antonelli, S. M.; Widlanski,
T. S.; Kantrowitz, E. R. Protein Sci. 2000, 9, 907-915.
(6) Snyder, S. L.; Wilson, I. B. Biochemistry 1972, 11, 1616-1623.
(10) Kirsanov, A. V.; Abrazhanova, A. Zh. Obshch. Khim. 1953, 23,
1048-1058.
(11) Fenesan, I.; Popescu, R.; Scozzafava, A.; Crucin, V.; Mateiciuc,
E.; Bauer, R.; Ilies, M. A.’ Supuran, C. T. J . Enzyme Inhib. 2000, 15,
297-310.

Synthesis and Bioactivity of N-Sulfonylphosphoramidates
J . Org. Chem., Vol. 66, No. 23, 2001 7563
Sch em e 1
Sch em e 2
Sch em e 3
approach.12 A series of N-(fluoroalkyl)sulfonylphosphor-
amidatic acids were recently constructed by reaction of
sulfonamide anions with chlorophosphates, though this
multistep scheme produced low overall yields.13
We reasoned that the synthesis of N-sulfonylphos-
phoramidates might be greatly simplified by adoption of
the proper protecting group strategy. The use of the
benzyl (Bn) protecting group offers a number of advan-
tages. A benzyl ester may be removed via hydrogenolysis
to yield a free acid in quantitative yield with no need for
further purification. Furthermore, because benzyl esters
are removed under mild, neutral conditions, the P-N
bond scission in N-sulfonylphosphoramidates, observed
under the acidic conditions employed in other syntheses,
may be avoided.
Ta ble 1. p Ka Va lu es of N-Su lfon ylp h osp h or a m id a tes
pKa
pKa
3
2
3a
3b
3c
5.1
5.2
∼3
10.3
9.4
∼7
Scheme 1 depicts the synthesis of tribenzyl N-sulfonyl-
phosphoramidates via phosphorylation of a sulfonamide,
adapted from a similar procedure developed by us and
others for the synthesis of phosphate esters.14 Oxidation
of tribenzyl phosphite with bromine gives dibenzyl
bromophosphate in situ. Reaction of N-benzylsulfona-
mides (1a -c) with dibenzyl bromophosphate gives the
fully benzylated N-sulfonylphosphoramidates (2a -c) in
reasonable yields (44-77%.) The protected N-sulfonylphos-
phoramidates (2a -c) may be deprotected by hydro-
genolysis to give N-sulfonlyphosphoramidates (3a -c) as
either the free acids or triethylammonium salts. This
facile synthesis is accomplished in just two steps from
the sulfonamide.
Tribenzyl N-sulfonylphosphoramidates may also be
synthesized using phophoramidate chemistry (Scheme 2).
Treatment of a sulfonamide with dibenzyl diisopropyl-
phosphoramidite, followed by oxidation with m-CPBA
gives 2a ,b in yields comparable to those obtained via the
bromophosphate approach. Using this method, the alkyl
and aryl N-sulfonlyphosphoramidates (2a ,b) were ob-
tained. However, synthesis of the trifluoromethyl deriva-
tive (2c) failed.
conveniently and in high yield (Scheme 3). Aniline and
n-butylamine were phosphorylated with dibenzyl bromo-
phosphate to give the N-phenyl- and N-butylphosphor-
amidates (4a ,b) respectively. Because of the acid lability
of phosphoramidic acids, direct hydrogenolysis of the
benzyl-protected substituents could not be affected in
protic or aprotic solvents without significant product
decomposition. When carried out in ethanol, this reaction
provided a side product whose NMR spectra is consistent
with ethyl phosphate, presumably arising from the
ethanolysis of the phosphoramidates. Hydrogenolysis
could be achieved, however, by addition of triethylamine
to the reaction mixture prior to hydrogenolysis (Scheme
3). Under these conditions, 4b was cleanly deprotected
to give N-phenylphosphoramidic acid (5b), though minor
impurities were observed when (4a ) was subjected to
hydrogenolysis.
Ch em ica l Rea ctivity N-Sulfonylphosphoramidic ac-
ids are strong triacids, with a first pKa on the range of
2-3 in water at 25 °C.7 Second and third pKa values of
N-sulfonylphosphoramidates 3a -c were determined by
31P NMR titration.15 Approximately 2 mg of compounds
3a -c were dissolved in buffers with pH’s ranging from
2 to 10. The phosphorus chemical shift of each sample
was plotted as a function of buffer pH, and the pKa was
determined by logistic regression (SigmaPlot) as the point
of inflection of the sigmoidal curves.15 Table 1 lists the
2nd and 3rd pKa’s of the N-sulfonylphosphoramidates
(3a-c). N-(Trifluoromethyl)sulfonylphosphoramidate (3c)
showed only a gradual change in phosphorus chemical
Using the tribenzyl phosphite approach, N-alkyl- and
N-arylphosphoramidate esters may also be synthesized
(12) (a) Zhu, S.-Z. J . Chem. Soc., Perkin Trans. 1 1994, 15, 2077-
2081. (b) Znu, S.; Zhang, J .; Xu, B.; Qin, C. Org. Prep. Proced. Int.
1997, 29, 352-355.
(13) Zhu, S.; Xu, G.; Qin, C.; Xu, Y.; Chu, Q. Phosphorus, Sulfur
Silicon Relat. Elem. 1998, 140, 53-61.
(14) (a) Stowell, J . K.; Widlanski, T. S. Tetrahedron Lett. 1995, 36,
1825-1826. (b) Watanabe, Y.; Inada, E.; J inno, M.; Ozaki, S. Tetra-
hedron Lett. 1993, 34, 497.
(15) Cynthia J . Anderson, Ph.D. Thesis, Indiana University, 1996.

7564 J . Org. Chem., Vol. 66, No. 23, 2001
Burlingham and Widlanski
Ta ble 2. Rela tive Ra tes of Hyd r olysis of
N-Su lfon ylp h osp h or a m id a tes a n d
N-P h en ylp h osp h or a m id a te
Ta ble 3. IC50’s of N-Su lfon ylp h osp h or a m id a tes for F ou r
P h osp h a ta ses
IC50 (µM)
a
a
substrate
Vo (B AP)b
Vo (EC AP)c
substrate
B APa
EC APb
PuAPc
Stp1d
pNPP
3a
3b
3c
5b
100
28
12
39
21
100
57
16
37
19
3a
3b
3c
5b
pNPP
60
9
16
52
10
73
1800
630
770
10000
2500
.10000
370
Km ) 40
770
Km ) 10
Km ) 700
Km ) 200
Relative to pNPP. Bovine alkaline phosphatase. c E. coli
alkaline phosphatase.
a
b
a
b
Bovine alkaline phosphatase. E. coli alkaline phosphatase.
c Kidney bean purple acid phosphatase. Small tyrosine phso-
d
shift at varying pH, and the point of inflection of the
curves could not be mathematically fit. Approximate
pKa’s were therefore determined graphically.
phatase 1.
were investigated as inhibitors of AP. Under the assay
conditions, two of the N-sulfonylphosphoramidates (3a ,b)
exist primarily as dianions while the third (3c) is
trianionic. The inhibitory constants of 3a -c are all within
the same order of magnitude. These data suggests that
AP does not discriminate between dianionic and tri-
anionic inhibitors. If the enzyme were to bind trianions
significantly tighter than dianions, N-(trifluoromethyl-
sulfonyl)phosphoramidate (3c) would be expected to be
a better inhibitor of AP than 3a ,b. On the other hand, if
dianions bound more tightly to AP, 3c would not be
expected to inhibit the enzyme as effectively as dianionic
compounds such as 3a ,b.
Both bovine AP and E. coli AP recognize N-sulfonyl-
phosphoramidates as substrates and catalyze their hy-
drolysis at a rate 2-10 times slower than that observed
for pNPP. The hydrolysis of 3a ,b was not unexpected.
Hydrolysis of dianionic N-sulfonylphosphoramidate re-
quires the expulsion of a sulfonamide anion. These anions
are less basic than the alkoxide ions that serve as leaving
groups in the hydrolysis of phosphate esters. In contrast,
the ability of trianionic N-(trifluoromethyl)sulfonylphos-
phoramidate to serve as a substrate was surprising,
because this hydrolysis requires the expulsion of a
sulfonamide dianion.
One explanation for the observation that 3c acts as a
substrate for AP is that the dianionic form of 3c is acting
as the substrate in this reaction. Even though 3c exists
in primarily the trianionic state, a small portion of the
dianion will be present at pH 9. If 3c were to bind as the
dianion, it would be hydrolyzed in the same way as 3a ,b.
Another possibility is that the trianion may be reactive
enough to undergo enzyme-catalyzed hydrolysis. For this
hydrolysis to occur, the leaving sulfonamide would have
to form a relatively stable dianion. The first pKa of
trifluoromethanesulfonamide is 6.3, and evidence sug-
gests that trifluoromethylsulfonamide may be doubly
deprotonated in aqueous solution at high pH.19 It there-
fore seems possible that the dianion of trifluoromethyl-
sulfonamide may function as a leaving group in this
reaction, and that trianionic N-(trifluoromethylsulfonyl)-
phosphoramidate may serve as a substrate for AP.
While the greater ionization ability of N-sulfonylphos-
phoramidates does not seem to enhance the inhibitory
capacity of these molecules, the sulfonyl group of N-
sulfonylphosphoramidate does seem to affect binding.
N-phenylsulfonylphosphoramidate has a significantly
Under acidic conditions, N-sulfonylphosphoramidates
are known to suffer hydrolysis of the P-N bond. To
determine the stability of 3a -c toward hydrolysis at
higher pH’s, degradation of the N-sulfonylphosphorami-
dates was followed over time by 31P NMR. In buffers of
pH 5 and 6, 3a and 3b have a half-life of approximately
24 h (data not shown). At pH 9, however, no hydrolysis
product was observed. In contrast, no hydrolysis was
detected for 3c at pH 5, 6, or 9. These data suggests that
N-sulfonylphosphoramidates are not susceptible to hy-
drolysis at pH’s significantly above their pKa2.
Alk a lin e P h osp h a ta se Recogn ition of N-Su lfon yl-
p h osp h or a m id a tes. To determine how well alkaline
phosphatase utilizes N-sulfonylphosphoramidates as sub-
strates, relative rates of hydrolysis were determined. The
rate of phosphate release was measured by the Lanzetta
assay16 for the AP-catalyzed hydrolysis of N-sulfonylphos-
phoramidates, and the rate was normalized relative to
the catalyzed hydrolysis of p-nitrophenyl phosphate
(pNPP.) Table 2 lists the relative initial rates of hydroly-
sis. At pH 9, N-sulfonylphosphoramidates are hydrolyzed
by bovine AP at 12-39% the rate of phosphate mono-
ester. The rate of catalyzed hydrolysis of 5b was also
measured to be 19% that of pNPP. Literature values for
the relative rate of hydrolysis of N-(phenyl)phosphor-
amidate are on the range of 29-37%.17
The inhibitory abilities of 3a -c were measured for four
phosphatases: bovine AP, E. coli AP, purple acid phos-
phatase (PuAP), and small tyrosine phosphatase 1 (Stp1).
IC50’s for N-sulfonylphosphoramidate were determined
with enzymes using pNPP as a substrate18 (Table 3).
N-Sulfonylphosphoramidates inhibit AP at significantly
lower concentrations than they inhibit PuAP or Stp1.
Furthermore, the three N-sulfonylphosphoramidates tested
possess significantly lower IC50 values than phenylphos-
phoramidate.
Discu ssion
To investigate the effects of charge on substrate
binding to AP, a series of N-sulfonylphosphoramidates
(16) Lanzetta, P. A.; Alvarez, L. J .; Reinach, P. S.; Candia, O. A.
Anal. Biochem. 1979, 100, 95-97.
(17) (a) Snyder, S. L.; Wilson, I. B. Biochemistry, 1972, 11, 1616-
1623. (b) Williams, A.; Naylor, R. A. J . Chem. Soc. (B) 1971, 1973-
1979.
(18) Because N-sulfonylphosphoramidates are sustrates of phos-
phatases, their binding ability could, in principle, be determined as
Km values. In practice, however, Km values for these substrates cannot
be determined accurately because of the difficulty in measuring the
concentration of inorganic phosphate released in assays performed at
low substrate concentration. A measurement of inhibition, such as an
IC50 value, also provides the pertinent binding information without
the technical difficulties associated with the direct measurement of
Km.
(19) 19F NMR spectra were obtained for aqueous samples of tri-
fluoromethanesulfonamide at various pHs. When dissolved in aqueous
buffers below its first pKa, 6.33, trifluoromethanesulfonamide has a
chemical shift of -80.3, and above this pH, it has a chemical shift of
-80.6. When dissolved in 10 M NaOH, a new shift (-79.3 ppm) occurs.
The trifluoromethanesulfonamide dianion is a species consistent with
this significant chemical shift at high pH.

Synthesis and Bioactivity of N-Sulfonylphosphoramidates
J . Org. Chem., Vol. 66, No. 23, 2001 7565
Chemical shifts are reported relative to residual solvent peaks.
Phosphorus-31 nuclear magnetic resonance (31P NMR) spectra
were recorded at 121.5 MHz. Chemical shifts are reported
relative to an external neat H3PO4 sample (0 ppm). Fluorine-
19 nuclear magnetic resonance (19F NMR) spectra were
recorded at 282 MHz. Chemical shifts are reported relative to
an external neat CF3COOH standard (-78.5 ppm). Infrared
(IR) spectra (thin film, KBr pellet, or single beam) were
recorded by FT-IR. Mass spectra were determined by high
resolution (HRMS) or high-resolution fast-atom bombardment
mass spectra (HRFABMS). Melting points were determined
in open capillary tubes and are uncorrected.
In h ibition of Bovin e Alk a lin e P h osp h a ta se. Assays
were conducted in 50 mM Tris HCl buffer at pH 9.0 with 1
mM MgCl2. The initial rate (∼5%) of product formation was
measured for the following ranges of inhibitor concentration:
3a , 0-100 µM; 3b, 0-35 µM; 3c, 0-67 µM; 5b, 0-2.2 mM. A
solution of substrate and inhibitor (590 µL of 175 µM pNPP)
was incubated in a cuvette in a temperature-controlled jacket
at 25 °C for 2 min. Enzyme solution was added (10 µL, 395
ng/mL), and the cuvette was inverted until the solution was
properly mixed. The absorbance was measured continuously
at 405 nm for one minute, and the initial rate was obtained
as the slope of the resulting plot of absorbance change versus
time. Percent activity was calculated by comparing the rates
of product release at various concentrations of inhibitor to the
rate of product release in the absence of inhibitor. The inverse
of the initial rate of reaction was plotted as a function of
inhibitor concentration. In this plot, IC50 is equal to the
opposite of the x-intercept.
In h ibition of E. coli Alk a lin e P h osp h a ta se. Assays were
conducted in 50 mM Tris HCl buffer at pH 9.0 with 1 mM
MgCl2. The initial rate (∼5%) of product formation was
measured for the following ranges of inhibitor concentration:
3a , 0-133 µM; 3b, 0-33 µM; 3c, 0-133 µM; 5b, 0-2.0 mM. A
solution of substrate and inhibitor (590 µL of 50 µM pNPP)
was incubated in a cuvette in a temperature-controlled jacket
at 25 °C for 2 min. Enzyme solution was added (10 µL, 1.88
U/mL), and the cuvette was inverted until the solution was
properly mixed. The absorbance was measured continuously
at 405 nm for one minute, and the initial rate was obtained
as the slope of the resulting plot of absorbance change versus
time. Percent activity was calculated by comparing the rates
of product release at various concentrations of inhibitor to the
rate of product release in the absence of inhibitor. The inverse
of the initial rate of reaction was plotted as a function of
inhibitor concentration. In this plot, IC50 is equal to the
opposite of the x-intercept.
In h ibit ion of P u r p le Acid P h osp h a t a se. Assays were
conducted in 100 mM NaOAc buffer at pH 5.0 containing 50
mM NaCl. Product formation (∼5%) was measured at the
following ranges of inhibitor concentration: 3a , 0-4.0 mM;
3b, 0-2.8 mM; 3c, 0-1.4 mM. To a solution of substrate and
inhibitor (590 µL of 3.0 mM pNPP) was added enzyme solution
(10 µL of 4.6 µg/mL enzyme in buffer contain 50 µg BSA/mL.)
After 5 min, the reaction was quenched by the addition of 100
µL of 1.0 M NaOH. The absorbance of the resulting solution
was measured at 405 nm. Percent activity was calculated by
comparing the amount of product release at various concentra-
tions of inhibitor to the rate of product release in the absence
of inhibitor. The inverse of the initial rate of reaction was
plotted as a function of inhibitor concentration. In this plot,
IC50 is equal to the opposite of the x-intercept.
lower IC50 (50-55-fold) than N-phenylphosphoramidate.
At the same time, N-phenylsulfonylphosphoramidate is
hydrolyzed by E. coli AP at a slower rate than is
N-phenylphosphoramidate. These data suggest that the
sulfonyl group is playing some role in interacting with
the protein active site. Ligation of the sulfonyl group to
the active site metal might be one possible explanation.
Such a phenomenon has been employed in the design of
other inhibitors of AP,20 and crystal structures indicate
that such interactions may be significant in determining
the mode of binding.5 The sulfonyl group may also confer
some selectivity among phosphatase targets. N-Sulfonyl-
phosphoramidates selectively inhibit alkaline phosphatas-
es as compared to an acid phosphatase (PuAP) and Stp1,
a nonmetallophosphatase (Table 3.)
Con clu sion s
While phosphorylation of alcohols with bromo- and
iodophosphates has been previously recognized as a
useful means to form phosphate triesters, this methodol-
ogy can be extended to the formation of phosphoramidate
and N-sulfonylphosphoramidate esters as well. N-Sulfo-
nylphosphoramidates were shown to be slow substrates
and moderate inhibitors of AP. Triply anionic N-sulfo-
nylphosphoramidates do not appear to bind tightly in the
alkaline phosphatase active site, suggesting AP does not
prefer to bind trianionic molecules in preference to
dianions, in spite of the plethora of positive charges in
the active site of AP.
Exp er im en ta l Section
Diethyl ether (Et2O) and tetrahydrofuran (THF) were
distilled under nitrogen from sodium-benzophenone ketyl.
Methylene chloride (CH2Cl2) was distilled from calcium hy-
dride. Pyridine (pyr.) was distilled from barium oxide. Butyl-
amine was distilled out of magnesium sulfate. Aniline was
distilled from crushed potassium hydroxide under reduced
pressure. Reactions were carried out under nitrogen atmo-
sphere. Biochemical assays were carried out in doubly distilled
and deionized waster. Bovine alkaline phosphatase, E. coli
alkaline phosphatase, and p-nitrophenyl phosphate were
purchased from Sigma. Kidney bean purple acid phosphatase
stock was from a previous study.20 Stp-1 was prepared accord-
ing to a literature procedure.21
Analytical thin-layer chromatography (TLC) was performed
on precoated (0.25 mm thickness) glass plates (E. Merck, silica
gel 60 F-254). Components were visualized by UV light (254
nm) if possible and by staining plates with either an acidic
solution of (NH4)6Mo7O24 monohydrate and ceric sulfate or an
acidic solution of p-anisaldehyde, followed by heating. Column
chromatography was performed using Kieselgel-60 230-400
mesh silica gel (E. Merck.) Bulk solvent removal was carried
out on a rotary evaporator at water aspirator pressure. All
nonvolatile compounds were routinely dried under high vacuum.
Proton nuclear magnetic resonance spectra (1H NMR) were
recorded on a 400 MHz spectrometer. The chemical shifts are
reported in parts per million (δ, ppm) relative to an internal
tetramethylsilane standard. Coupling constants are reported
in hertz (Hz), and the data are presented in the following
form: chemical shift (multiplicity, coupling constants, number
of protons.) Multiplicities are recorded by the following ab-
breviations: s, singlet; d, doublet; t, triplet; q, quartet; sx,
sextet; J , coupling constant (Hz). Carbon-13 nuclear magnetic
resonance (13C NMR) spectra were recorded at 101 MHz.
In h ibition of Stp 1. Assays were conducted in 50 mM
succinate buffer at pH 6.0 containing 150 mM NaCl and 1 mM
EDTA. Product formation (∼5%) was measured for the follow-
ing ranges of inhibitor concentration: 3a , 0-4.0 mM; 3b, 0-2.7
mM; 3c, 0-2.7 mM. To a solution of substrate and inhibitor
(595 µL of 800 µM pNPP) was added 5 µL of enzyme solution
(1 µg/mL). After 10 min, the reaction was quenched by the
addition of 100 µL of 1.0 M NaOH. The absorbance of the
resulting solution was measured at 405 nm. Percent activity
was calculated by comparing the amount of product release
at various concentrations of inhibitor to the rate of product
release in the absence of inhibitor. The inverse of the initial
(20) Myers, J . K.; Antonelli, S. M. Widlanski, T. S. J . Am. Chem.
Soc. 1997, 119, 3163-3164.
(21) Zhang, Z. Y.; Zhou, G.; Wu, L.; Tang, X.; Monderert, O.; Russell,
P.; Butch, E.; Guan, K. L. Biochemistry 1995, 34, 10560-10568.

7566 J . Org. Chem., Vol. 66, No. 23, 2001
Burlingham and Widlanski
rate of reaction was plotted as a function of inhibitor concen-
tration. In this plot, IC50 is equal to the opposite of the
x-intercept.
mL). The ether layer was collected and washed with 30 mL of
10% aq sodium hydroxide followed by 30 mL of brine. The
ether layer was dried over magnesium sulfate, and the ether
was removed in vacuo. The crude material was purified by
silica gel column chromatography (2a : 1:1 Hex:EtOAc; 2b: 2:1
Hex:EtOAc).
Diben zyl N-Ben zyl-N-(m eth ylsu lfon yl)p h osp h or a m i-
d a te (2a ). Yield: Method A: 61% isolated, Method B: 64%
colorless oil: 1H NMR (400 MHz, CDCl3): δ 2.76 (s, 3H), 4.68
(d, J ) 10.8 Hz, 2H), 5.04 (m, 4H), 7.2-7.35 (m, 15H) ppm;
13C NMR (100 MHz, CDCl3): δ 42.21, 51.54, 69.75 (d, J ) 5.3
Hz), 128.04, 128.16, 128.39, 128.49, 128.61, 128.89, 135.03 (d,
6.9 Hz), 136.21 ppm; 31P NMR (121 MHz, CDCl3): -0.14 ppm;
IR (neat): 1165, 1275, 1355 cm-1; HRFABMS (M + H) calcd
for C22H25NO5PS 446.1191, found 446.1194.
Diben zyl N-Ben zyl-N-(p h en ylsu lfon yl)p h osp h or a m i-
d a te (2b). Yield: Method A: 77%, Method B: 61% white
solid: mp 77-78 °C; 1H NMR (400 MHz, CDCl3): δ 4.67 (d, J
) 11.6 Hz, 2H), 4.93 (m, 4H), 7.20-7.76 (m, 15H) ppm; 13C
NMR (100 MHz, CDCl3): δ 52.07, 69.60 (d, J ) 5.3 Hz) 127.77,
127.85, 128.18, 128.28, 128.51, 128.57, 128.62, 128,75, 133.06,
135.21 (d, J ) 7.6 Hz), 136.34, 139.92 ppm; 31P NMR (121
MHz, CDCl3): 0.23 ppm; IR (thin film): 1010, 1173, 1276,
1367, 3066 cm-1; HRMS (M + H) calcd for C27H27NO5PS
508.1348, found 508.1368.
Diben zyl N-Ben zyl-N-(tr iflu or om eth ylsu lfon yl)p h os-
p h or a m id a te (2c). Yield: Method A: 44%, Method B: 0%
colorless oil: 1H NMR (400 MHz, CDCl3): δ 4.77 (d, J ) 11.6
Hz, 4H), 4.90 (t, J ) 10.4 Hz, 2H), 7.18-7.53 (m, 15H) ppm;
13C NMR (100 MHz, CDCl3): δ 54.10, 70.59 (d, J ) 5.3 Hz),
119.55 (q, J ) 323 Hz), 128.12, 128.56, 128.61, 128.80, 129.35,
134.55 (d, J ) 7.9 Hz), 135.20 ppm; 31P NMR (121 MHz,
CDCl3): -2.75 ppm; 19F NMR (282 MHz, CDCl3): δ -74.56
ppm; IR (neat): 1015, 1139, 1230, 1288, 1403 cm-1; HRMS
(M - H) calcd for C22H20F3NO5PS 498.0752, found 498.0750.
Gen er a l P r oced u r e for th e P r ep a r a tion of N-Su lfon yl-
p h osp h or a m id ic Acid s (3). Dibenzyl N-benzyl-N-sulfonyl-
phosphoramidate (2) (0.5 mmol) was dissolved in 2 mL of
ethanol. Approximately 50 mg of 10% activated palladium on
carbon was added to the reaction. The flask was sealed and
evacuated in vacuo, followed by introduction of an H2 atmo-
sphere. After 2 h, the reaction flask was purged with nitrogen
and the solution was filtered through Celite. The ethanol was
evaporated in vacuo to yield 3a ,b quantitatively, with no need
for further purification. 3c was not stable as the free acid, but
could be isolated as the triethylammonium salt by dissolving
the crude product in triethylamine/bicarbonate buffer and
evaporating the buffer in vacuo.
N-(Meth ylsu lfon yl)p h osp h or a m id ic Acid (3a ). White,
highly hygroscopic solid: 1H NMR (400 MHz, CDCl3): δ 3.13
(s) ppm; 13C NMR (100 MHz, CDCl3): δ 42.90 ppm; 31P NMR
(121 MHz, CDCl3): -4.62 ppm; IR (neat): cm-1; HRFABMS
(M + H) calcd for CH7NO5PS 175.9781, found 175.9783.
N-(P h en ylsu lfon yl)p h osp h or a m id a ic Acid (3b). White
solid: mp147-148 °C (148-149 °C);24 1H NMR (400 MHz,
CDCl3): δ 7.51-7.98 (m) ppm; 13C NMR (100 MHz, CDCl3): δ
128.20, 129.99, 133.87, 143.44 ppm; 31P NMR (121 MHz,
CDCl3): -5.15 ppm; IR (KBr): 1182, 1215, 1332, 2363, 2413,
3229 cm-1; HRMS (M + H) calcd for C6H9NO5PS 237.9939,
found 237.9950.
Tr ieth yla m m on iu m N-(Tr iflu or om eth ylsu lfon yl)p h os-
p h or a m id a te (3c). White, hygroscopic solid: mp 81-83 °C;
1H NMR (400 MHz, CDCl3): δ 1.27 (t, J ) 6.4 Hz, 9H), 3.08
(q, J ) 6.4 Hz, 6H) ppm; 13C NMR (100 MHz, CDCl3): δ 8.03,
45.77, 121.41 (q, J ) 322 Hz) ppm; 31P NMR (121 MHz, CDCl3):
-2.33 ppm; 19F NMR (282 MHz, CDCl3): δ -79.87 ppm IR
(KBr): 1165, 1206, 1266, 1476, 2491, 2677, 2976 cm-1; HR-
FABMS (M + H) calcd for C13H34F3N3O5PS 432.1909, found
432.1900.
N-Su lfon ylp h osp h or a m id a tes a s Su bstr a tes of Alk a -
lin e P h osp h a ta se. The relative rates of hydrolysis of pNPP
and N-sulfonylphosphoramidates 3a -c were followed by de-
tection of inorganic phosphate release, detected by the method
of Lanzetta et al.15 To 1 mL of 500 µM substrate was added
enzyme (20 µL of 395 nM/mL bovine AP, or 15 µL of 1.71 U/mL
E. coli AP). At time points (0-60 min), 100 µL aliquots of the
reaction solution were quenched into 800 µL of molybdenate
color indicator (ꢀ ) 78 000 Abs/M). After 60 s, the color
indicator was quenched with 100 µL of 34% sodium citrate
solution. After 1 h, the absorbances of these solutions were
measured at 660 nm. Blanks contained 100 µL of 500 µM
substrate with no enzyme mixed with 800 µL of color indicator
and quenched as above. Initial rates (∼10%) were determined
from the slope of the line of a plot of the concentration of
inorganic phosphate released as a function of time.
N-P h en ylp h osp h or a m id a te (5b) a s
a Su bstr a te of
Alk a lin e P h osp h a ta se. The relative rates of hydrolysis of
pNPP and N-phenylphosphoramidates 5b was followed by
detection of inorganic phosphate release, detected by a varia-
tion of the method of Lanzetta, et al.15 To 1 mL of 500 µM
substrate was added enzyme (20 µL of 395 nM/mL bovine AP,
or 15 µL of 1.71 U/mL E. coli AP.) At time points (0-60 min),
100 µL aliquots of the reaction solution were quenched into
800 µL of molybdenate color indicator (ꢀ ) 70 000 Abs/M). After
6 s, the color indicator was quenched with 100 µL of 34%
sodium citrate solution. (Because the substrate is not stable
under the acidic assay conditions, the assay was quenched
quickly, resulting in only 10% loss of sensitivity.) After 1 h,
the absorbances of these solutions were measured at 660 nm.
Blanks contained 100 µL of 500 µM substrate with no enzyme
mixed with 800 µL of color indicator and quenched as above.
Initial rates (∼10%) were determined from the slope of the
line of a plot of the concentration of inorganic phosphate
released as a function of time.
Gen er a l P r oced u r e for th e P r ep a r a tion of Diben zyl
N-Ben zyl-N-su lfon ylp h osp h or a m id a t es (2). Met h od A:
Br om op h osp h a te. N-Benzylsulfonamide (1a -c) (1 mmol)
was dissolved in 2 mL of dry tetrahydrofuran and added
dropwise to sodium hydride (1.1 mmol) that had been pre-
washed in 5 mL of pentane. After stirring for 45 min, no gas
evolution was visible. Tribenzyl phosphite22 (1.5 mmol) was
dissolved in 2 mL of dry methylene chloride at 0 °C. Bromine
(1.45 mmol), dissolved in 3 mL of methylene chloride, was
added to the phosphite with stirring over the course of 5 min.
After 10 min, the solution was warmed to room temperature
and stirred an additional 15 min. To the deprotonated sulfon-
amide solution at -30 °C was added the phosphorylating
reagent, dropwise over 10 min. The reaction was warmed to
room temperature and allowed to stir for 60 min. The solvent
was removed in vacuo, and the residue was partitioned
between 25 mL of ether and 25 mL of 1 M HCl. The ether
layer was collected and washed with brine (25 mL) and then
dried over magnesium sulfate. The ether was removed in
vacuo, and the crude material was purified by silica gel
chromatography (2a : 1:1 Hex:EtOAc; 2b: 2:1 Hex:EtOAc;
2c: 2:1 Hex:Et2O). Meth od B: P h osp h or a m id a te. N-Benzyl-
sulfonamide(1) (1 mmol) and 1-H-tetrazole (1 mmol) were
dissolved/suspended in 2 mL of dry methylene chloride.
Dibenzyl diisopropylphosphoramidite23 (1 mmol) was added to
the reaction dropwise. Upon stirring for 1 h, the solution grew
clear and then cloudy. The solution was cooled to -40 °C, and
m-CPBA (2 mmol), in 3 mL of methylene chloride, was added
dropwise. After stirring the reaction for 1 h, the reaction was
allowed to warm to room temperature. The solvent was
evaporated in vacuo, and the crude material was partitioned
between ether (30 mL) and 10% aq sodium bicarbonate (30
Gen er a l P r oced u r e for th e P r ep a r a tion of Diben zyl
P h osp h or a m id a te (4). Tribenzyl phosphite22 (2 mmol) was
(22) Saady, M.; Lebeau, L.; Mioskowski, C. Helv. Chim. Acta 1995,
78, 670-678.
(23) J enkins, D. J .; Riley, A. M. Potter, B. V. L. J . Org. Chem. 1996,
61, 7719-7726.
(24) Houmbold, W.; Becke-Goering, M. Z. Anorg. Allg. Chem. 1967,
352, 113-121.

Synthesis and Bioactivity of N-Sulfonylphosphoramidates
J . Org. Chem., Vol. 66, No. 23, 2001 7567
dissolved in 3 mL of methylene chloride at 0 °C. Bromine (1.95
mmol) dissolved in 2 mL of methylene chloride was added to
the phosphite dropwise. The reaction was stirred for 20 min
and then warmed to room temperature and stirred for an
additional 20 min. In a separate flask, the amine (3 mmol)
was dissolved in 5 mL of methylene chloride at -30 °C. The
phosphorylating reagent was added to the amine dropwise over
10 min. The reaction was warmed to room temperature and
stirred for 2 h. The solvent was evaporated in vacuo, and the
residue was partitioned between 25 mL of ether and 25 mL of
2 M HCl. The ether layer was separated from the organic layer
and washed again with 25 mL of 2 M HCl and then 25 mL of
brine. The organic layer was collected and dried with magne-
sium sulfate, followed by evaporation of the solvent in vacuo.
The crude product was purified by silica gel chromatography
(1:1 Hex:EtOAc for both 4a and 4b.)
activated palladium on carbon was added to the reaction. The
flask was sealed and evacuated, followed by introduction of
an H2 atmosphere. After 0.5 h, the reaction flask was purged
with nitrogen, and the solution was filtered through Celite.
The ethanol was evaporated in vacuo to yield 5a ,b.
Tr ieth yla m m on iu m N-(n -Bu tyl)p h osp h or a m id a te (5a ).
White solid, mp 74-76 °C; 1H NMR (400 MHz, CD3OD): δ
0.97 (t, J ) 9.6 Hz, 3H), 1.32 (t, 9.6 Hz, 4.5H), 1.42 (sx, J )
10.0 Hz, 2H), 1.67 (p, J ) 10 Hz, 2H), 3.04 (m, 2H), 3.19 (q, J
) 9.6 Hz, 3H) ppm; 13C NMR (100 MHz, CD3OD): δ 9.26,
14.16, 21.22, 31.50 (d, J ) 7.6 Hz), 44.15, 47.56 ppm; 31P NMR
(121 MHz, CD3OD): 0.54 ppm; IR (single beam): 976, 1081,
1173, 1235, 1464, 2959 cm-1; HRFABMS (M + H) calcd for
C
10H28N2O3P 255.1838, found: 255.1829.
Tr iet h yla m m on iu m P h en ylp h osp h or a m id a t e (5b ).
White solid: mp 102-104 °C; 1H NMR (400 MHz, CD3OD): δ
1.25 (t, J ) 7.2 Hz, 9H), 3.07 (q, J ) 7.2 Hz, 6H), 6.74-7.13
(m, 5H) ppm; 13C NMR (100 MHz, CD3OD): δ 9.16, 47.40,
118.01 (d, J ) 6.9 Hz), 120.07, 129.82, 145.13 ppm; 31P NMR
(121 MHz, CD3OD): 0.84 ppm; IR (KBr): 1082, 1154, 1313,
1501, 1604, 2359, 2677, 2979, 3218 cm-1; HRFABMS (M + H)
calcd for C12H24N2O3P: 275.1525, found: 275.1534.
Diben zyl N-(n -Bu tyl)p h osp h or a m id a te (4a ). Yield: 80%
white solid: mp 46-47 °C; 1H NMR (400 MHz, CDCl3): δ 0.84
(t, J ) 7.2 Hz, 3H), 1.26 (sx, J ) 7.2 Hz, 2H), 1.39 (p, J ) 7.2
Hz, 2H), 2.84 (d q, J ) 9.6 Hz, 7.2 Hz, 2H), 3.06 (m, 1H), 5.04
(d, J ) 5.6 Hz), 7.20-7.40 (m, 10H) ppm; 13C NMR (100 MHz,
CDCl3): δ 13.54, 19.57, 33.55 (d, J ) 6.8 Hz), 40.97, 67.63 (d,
4.5 Hz), 127.57, 128.03, 128.34, 136.47 (d, 7.6 Hz) ppm; 31P
NMR (121 MHz, CDCl3): 10.36 ppm; IR (thin film): 1000,
1456, 1548, 2959, 3218 cm-1; HRFABMS (M + H) calcd for
Ack n ow led gm en t. Funding was provided by the
National Institute of Health (Grant CA 71736-03.)
Thanks to Professor Z. Y. Zhang (Albert Einstein) for
the generous gift of Stp-1 plasmid. B.T.B. thanks J ohn
Strelow for the isolation and purification of Stp-1,
Stephen Antonelli for helpful discussions, and Procter
& Gamble for a research fellowship.
C
18H25NO3P 134.1572, found: 134.1557.
Diben zyl N-P h en ylp h osp h or a m id a te (4b).25 Yield: 97%
white solid: mp 88-89 °C; 1H NMR (400 MHz, CDCl3): δ 5.09
(m, 4H), 6.94-7.30 (m, 15H) ppm; 13C NMR (100 MHz,
CDCl3): δ 68.44 (d, 4.6 Hz), 117.60 (d, J ) 7.6 Hz), 121.60,
127.95, 128.31, 128.41, 129.16, 135.69 (d, J ) 8.3 Hz), 139.51
ppm; 31P NMR (121 MHz, CDCl3): 3.11 ppm; IR (thin film):
1001, 1227, 1422, 1500, 1605, 2912, 3093, 3169 cm-1; HR-
FABMS (M + Na) calcd for C20H20NO3PNa 376.1079, found:
376.1072.
Su p p or tin g In for m a tion Ava ila ble: 1H NMR spectra of
all compounds lacking elemental analysis (2a , 2b, 2c, 4a , 5a ,
5b). 13C NMR spectra of 3a . 31P and 19F NMR spectra of 3c.
19F NMR spectra of trifluoromethanesulfonamide in pH 4
buffer, pH 9 buffer, and 10 M NaOH. This material is available
free of charge via the Internet at http://pubs.acs.org.
Gen er a l P r oced u r e for th e P r ep a r a tion of Tr ieth yl-
a m m on iu m P h osp h or a m id a tes (5). Dibenzyl phosphora-
midate (4) (0.3 mmol) and triethylamine (0.75 mmol) were
dissolved in 2 mL of ethanol. Approximately 40 mg of 10%
(25) Zwierzak, A. Synthesis 1975, 8, 507-509.
J O010495Q