Reactions of phosphate and phosphorothiolate diesters with nucleophiles: Comparison of transition state structures

Article abstract of DOI:10.1039/b707205h

A series of methyl aryl phosphorothiolate esters (SP) were synthesized and their reactions with pyridine derivatives were compared to those for methyl aryl phosphate esters (OP). Results show that SP esters react with pyridine nucleophiles via a concerted S_N2(P) mechanism. Bronsted analysis suggests that reactions of both SP and OP esters proceed via transition states with dissociative character. The overall similarity of the transition state structures supports the use of phosphorothiolates as substrate analogues to probe mechanisms of enzyme-catalyzed phosphoryl transfer reactions. This journal is The Royal Society of Chemistry.

Full text of DOI:10.1039/b707205h

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Reactions of phosphate and phosphorothiolate diesters with nucleophiles:
comparison of transition state structures
Jing-Dong Ye,†a Christofer D. Barth,†b Potluri S. R. Anjaneyulu,c Thomas Tuschld and Joseph A. Piccirilli*a
Received 14th May 2007, Accepted 8th June 2007
First published as an Advance Article on the web 29th June 2007
DOI: 10.1039/b707205h
A series of methyl aryl phosphorothiolate esters (SP) were synthesized and their reactions with pyridine
derivatives were compared to those for methyl aryl phosphate esters (OP). Results show that SP esters
react with pyridine nucleophiles via a concerted SN 2(P) mechanism. Brønsted analysis suggests that
reactions of both SP and OP esters proceed via transition states with dissociative character. The overall
similarity of the transition state structures supports the use of phosphorothiolates as substrate
analogues to probe mechanisms of enzyme-catalyzed phosphoryl transfer reactions.
Introduction
Results and discussion
Biological systems rely heavily on phosphoester transfer reactions,
devoting significant fractions of their genomes to encode protein
and RNA enzymes that catalyze these reactions.1–3 Phosphoester
transferases serve as integral components of pathways involved
in cellular replication and differentiation, playing central roles in
gene replication, recombination, and expression. Understanding
these important cellular processes at a chemical level requires
detailed knowledge of the catalytic mechanisms exploited by the
enzymes that mediate these processes. We now know that enzymes
frequently employ metal ions4–14 and/or general acids and bases
to catalyze phosphoester transfer reactions.15–21 Nevertheless,
defining these mechanisms within specific protein and RNA
enzymes remains challenging.
Phosphorothiolates, in which a sulfur atom replaces a bridging
oxygen, have provided effective probes for investigating the
mechanistic basis of phosphoryl transferases. This modification
is frequently used to study RNA enzymes including splicing
machineries (Group I, Group II, and the spliceosome),4–6,22–25 the
hepatitis delta virus (HDV)17 and hammerhead ribozymes,26–28
RNase P,29 as well as protein enzymes.30 These probes have unveiled
intricate and fundamental features of catalysis, including metal
ion coordination, hydrogen bond donation, and proton transfer
to the leaving group. These studies have relied on the underlying
assumption that phosphorothiolate and phosphate diesters react
similarly with nucleophiles and share a common pathway and
transition state. However, this assumption has never been tested.
Here, we synthesize a series of methyl aryl phosphorothiolate esters
and compare their non-enzymatic reactions directly with those of
an analogous series of methyl aryl phosphate esters. We obtain an
initial physical organic description of the transition state structure.
Synthesis of phosphates and phosphothiolate diesters
We chose to investigate reactions of methyl aryl phosphorothiolate
diester monoanions with pyridine nucleophiles for two reasons:
(1) aryl groups allow systematic perturbation of the nucleophile
and leaving group pKa values with minimal steric perturbation,
and (2) Kirby and Younas previously investigated reactions of OP
diesters with pyridine nucleophiles.31 We synthesized a series of
S-aryl methyl phosphorothiolate (SP) diesters (Scheme 1, 3a–d)
and aryl methyl phosphate (OP) diesters containing substituted
thiophenol and phenol leaving groups, respectively (Scheme 1).
OP diesters were synthesized following the procedure of Ba-
Saif et al.32 To synthesize SP diesters, we first prepared S-aryl
O,O-dimethyl phosphorothiolates 2a–d using a modified version
of the method described by Murdock and Hopkins for the
preparation of S-(4-chlorophenyl) O,O-dimethyl phosphorthioic
acid ester 2a.33 Rather than using bromotrichloromethane to
facilitate the radical-chain-transfer reaction, we reacted aryl
sulfenyl chlorides (generated in situ for 4-methylbenzenesulfenyl
chloride and 4-chlorobenzenesulfenyl chloride by reducing the
corresponding disulfides with chlorine gas) directly with trimethyl
phosphite to generate 2a–d. Dimethyl phosphorothiolates 2a–d
were demethylated with LiCl in dry acetone to give S-aryl methyl
phosphorothiolates 3a–d.31
Kinetic measurements
We used 1H and 31P NMR to monitor the reactions of the
monoanions of the OP and SP diesters with pyridine nucleophiles
(23 ◦C, 1.7 M ionic strength, 30 mM K2CO3, pH 10) (Fig. 1A and
1B). In these reactions, the initial products, pyridine methyl phos-
phoramidates, undergo rapid hydrolysis to pyridine and methyl
phosphate, the only phosphorus-containing product observed.31
Pseudo-first order rate constants (kobs, Fig. 1C) were obtained by
single exponential fits of the reactions with nucleophiles present in
large excess (10–50 eq.). Least-squares analysis of the plots of kobs
versus nucleophile concentration gave good linear fits (Fig. 1D).
Second order rate constants (k2) were obtained from the slopes
of these plots. For reactions of pyridines with OP diesters at
39 ◦C investigated previously, k2 plots showed curvature due to
aDepartment of Biochemistry and Molecular Biology, Department of Chem-
istry, Howard Hughes Medical Institute, The University of Chicago, 929 E
57th St., CIS-W408A, Chicago, Illinois 60637, USA
bGeneral Anesthesiology, Cleveland Clinic, 9500 Euclid Avenue, Cleveland,
OH 44195, USA
c5329 W. Greenwood, Skokie, IL 60077, USA
dLaboratory of RNA Molecular Biology, Rockefeller University, 1230 York
Avenue, Box 186, New York, NY 10021, USA
† These authors contributed equally to this work.
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 2491–2497 | 2491
©

View Article Online
Scheme 1 Reagents and conditions: (i) Cl2, CH2Cl2, rt, 0.5 ∼ 1 h; (ii) trimethyl phosphite, 0 ∼ 40 ◦C, 0.5 ∼ 2 h, 65 ∼ 97%; (iii) LiCl, acetone, rt, 4 ∼ 24 h,
64 ∼ 90%.
Fig. 2 Plot of k2 versus pKa of the nucleophile for the reactions of
S-(2,4-dinitrophenyl) O-methyl phosphorothiolate (closed circles) and
2,4-dinitrophenyl methyl phosphate (open circles) with pyridine nucle-
ophiles. Second order rate constants (k2) were obtained as described
in Fig. 1. The pKa values43 of the pyridine nucleophiles are: 3-CN,
1.45; 3-COCH3, 3.18; 3-CONH2, 3.40; 3-H, 5.20; 3-methyl, 5.70; and
4-methyl, 6.02. The solid square symbol represents k2 for the reaction
of S-(2,4-dinitrophenyl) O-methyl phosphorothiolate with 2-picoline.
even though 2-picoline and 3-picoline have similar pKa values, the
Fig. 1 Monitoring the reaction of S-(4-nitrophenyl) O-methyl phospho-
sterically hindered 2-picoline reacts (k2 = 0.069 M−1 h−1) with S-
rothiolate with 3-picoline. 1H (A) and 31P (B) NMR spectra of the reaction
mixture at different times. The 1H NMR spectrum shows the methyl peak
as a doublet in the starting ester (S) and the methyl phosphate product
(P1). The 31P NMR spectrum shows the phosphorus peak in the starting
ester (S) and the phosphate monoester product (P2). (C) Data from A and
B are plotted against time and the observed rate constant (kobs) is obtained
(2,4-dinitrophenyl) methyl phosphorothiolate more than 100-fold
slower than does 3-picoline (k2 = 13 M−1 h−1). Reactions of S-(4-
nitrophenyl) methyl phosphorothiolate with peroxide exhibit an a
effect:35 peroxide reacts more than 100-fold faster than hydroxide,
even though peroxide has a pKa value almost four pH units
lower than hydroxide. This sensitivity of the reaction to steric
and electronic features of the nucleophile further suggests that
the pyridines act directly as nucleophiles rather than as general
acid/base catalysts. We conclude that SP diesters, like OP diesters,
react with nucleophiles through an SN 2(P) mechanism.31
from the slope of the plot. (D) Dependence of observed rate constant (kobs
)
on the concentration of 3-picoline. The slope of this linear dependence
gives the second order rate constant k2.
pyridine self-association.31,34 The reactions studied here◦showed
no curvature, possibly due to the lower temperature (23 C) and
lower concentration of pyridines used.
The linear dependence of kobs on the nucleophile concentration
and of k2 on the pKa of the nucleophile (Fig. 2) suggest that the
SP diesters react with pyridines by an SN 2 mechanism. Moreover,
Reactions of 3-picoline with a series of SP and OP diesters
show that k2 for these esters strongly depends on the leaving group
pKa (Fig. 3), exhibiting Brønsted blg values of −1.1 and −1.3 for
SP and OP diesters, respectively. blg values provide the change
of the effective charge that occurs on the leaving group as the
2492 | Org. Biomol. Chem., 2007, 5, 2491–2497
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
diesters is comparable to the published31 value of 0.44 measured at
39 ◦C and 1 M ionic strength.‡ These apparent bnuc values for the
phosphoryl transfer to pyridine likely deviate from the intrinsic bnuc
by the well-documented solvation effect.44,45 Using the equation
corr
b
b
= (bnuc − bd)/1 − bd) with bd = −0.2,46 we obtained corrected
nuc
corr
nuc
values of 0.42 and 0.56 for SP and OP diesters, respectively.
After correction of the observed bnuc values, the difference
between OP and SP diesters decreases modestly but remains
significant. The interaction coefficient pxy could account for part
of this difference. pxy defines the dependence of bnuc on the pKa
of the leaving group or the dependence of blg on the pKa of
the nucleophile (pxy = ∂bnuc/∂pK1ag = ∂blg/∂pKnauc).47 The pKa
values of 2,4-dinitrothiophenol36 and 2,4-dinitrophenol32 are 3.2
and 4.1, respectively. Using an interaction coefficient pxy of 0.013
for pyridine attack on OP diesters,31,47 we obtain bnuc = 0.56 −
0.013 × (4.1 − 3.2) = 0.55 for pyridine nucleophiles reacting
with a hypothetical aryl methyl phosphate diester containing a
leaving group with pKa = 3.2. Thus, the pKa difference of 2,4-
dinitrothiophenol and 2,4-dinitrophenol is not sufficient to explain
the difference in bnuc between the OP and SP diesters.
Fig. 3 Plot of k2 versus pKa of the leaving group for the reactions of S-aryl
O-methyl phosphorothiolates (closed circles) and aryl methyl phosphates
(open circles) with 3-picoline. The pKa values36,37 of the leaving groups for
S-aryl O-methyl phosphorothiolates are: 2,4-dinitro, 3.2; 4-nitro, 5.11;
4-chloride, 7.06; and 4-methyl, 8.07. The pKa values32 of the leaving
groups for aryl methyl phosphates are: 2,4-dinitro, 4.1; 2,5-dinitro, 5.22;
4-Cl-2-nitro, 6.46; and 4-nitro, 7.14.
The beq for phosphoryl group transfer to pyridines is indepen-
corr
nuc
dent of the leaving group, allowing direct comparison of b
corr
values. The b values indicate that SP diesters have less N-P
nuc
reaction progresses from ground state to transition state, relative to
the effective charge change defined by the ionization equilibrium:
RXH ꢀ RX− + H+ (X = O for OP and S for SP diesters). The
effective charges on X for RXH and RX− are defined as 0 and −1,
respectively.
bonding in the transition state than do OP diesters. Given that
the P–O and P–S bonds undergo similar extents of cleavage in
the transition state, these results suggest that SP diesters require
less bonding to the nitrogen nucleophile than do OP diesters to
achieve the same degree of bond cleavage to the leaving group. The
Hammond effect may account for this difference,35,48 reflecting the
weaker thermodynamic stability of the P–S bond compared to the
P–O bond.49,50
To facilitate comparison of the OP and SP diester reactions, we
may obtain a crude estimate of the beq for pyridine-PO2(OCH3)−
transfer equilibria from the pyridine-acyl transfer data (where
beq = +1.6),42 as described above for thiophenol-PO2(OCH3)−
transfer equilibria. This enables construction of the effective
charge map (Scheme 2) and representation of the transition state
in a More O’Ferrall–Jencks diagram (Fig. 4).51,52 The data suggest
that OP and SP diesters react with nucleophiles via a concerted
mechanism§ involving a dissociative transition state.53¶ Previous
studies on chemical models of phosphate diester monoanions,32,54
ab initio calculations of phosphorothiolate diesters,55 and Kirby’s
bnuc values31‡corroborate the OP diester transition state model that
To obtain the effective charge and bonding extent for the leaving
group in the transition state, blg values require calibration against
beq, the substituent effect on the reaction equilibrium.38–40 beq
defines the effective charge change on the leaving group for the
overall reaction equilibrium, Keq. For reactions of aryl phosphate
monoanion, beq = −1.74,41 meaning that the oxygen leaving group
bears an effective charge of +0.74 in the starting ground state and
an effective charge of −1 (by definition) in the phenolate product
state. The value of blg = −1.3 for OP diesters suggests that the
phenolate oxygen bears an effective charge of −0.6 (+0.74 − 1.3)
in the transition state. beq for cleavage of S-aryl phosphorothiolate
esters has not been determined and is challenging to measure.
However, acyl and -PO3H− groups exhibit similar electropositive
character in phenol transfer equilibria (with beq = 1.70 and 1.74,
respectively42). Assuming that the acyl and -PO3H− groups exhibit
similar electropositivity in thiophenol transfer, we may use beq
=
1.38 for thiophenol-acyl transfer equilibria42 as a crude estimate of
beq for thiophenol-PO3H− and thiophenol-PO2(OCH3)− transfer
equilibria. The value of blg = −1.1 for SP diesters suggests that
the sulfur leaving group has an effective charge of −0.7 (+0.38 −
1.1) in the transition state. OP and SP diesters therefore appear
to undergo similar extents of bond breaking (a) in their respective
reactions with pyridine: aS = bS1g/beSq = 0.79; aO = b1Og/beOq = 0.75.
To probe the degree of bond making between the nucleophile
and OP and SP diesters, we studied reactions of OP and SP
diesters with a series of pyridine derivatives. The second order
rate constants for reaction of pyridine nucleophiles with S-(2,4-
dinitrophenyl) methyl phosphorothiolate and 2,4-dinitrophenyl
methyl phosphate show linear dependencies on nucleophile pKa
(Fig. 2). The slopes give the Brønsted coefficient, bnuc, as 0.3 and
0.47 for SP and OP diesters respectively. The bnuc of 0.47 for OP
‡ The text of the published report31 states that bnuc = 0.31; the plot in
Fig. 2 in that report has a slope of 0.44. We plotted the data from Table 2
in that report and obtained bnuc = 0.44, in excellent agreement with our
measurement.
§ Formally, the reactions may occur via a stepwise addition-elimination
(associative) mechanism in which the nucleophile first adds to the phos-
phorous center to form a dianionic phosphorane intermediate followed by
expulsion of the leaving group. Our results argue against the associative
mechanism. If the addition step limited the reaction rate, we would
expect small values of bnuc and blg (relative to the corresponding beq).
If the elimination step limited the reaction rate, we would expect more
substantial values for both bnuc and blg. However, we observe relatively
large blg values and relatively small bnuc values, consistent with neither
rate-limiting addition nor rate-limiting elimination.
¶ A dissociative transition state, defined by bnuc/beq + (1 − blg/beq) <
1, indicates that bond fission exceeds bond formation. An associative
transition state, defined by bnuc/beq + (1 − blg/beq) > 1, indicates that
bond formation exceeds bond fission).
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 2491–2497 | 2493
©

View Article Online
Scheme 2 Effective charges of the nucleophiles and leaving groups in the reactions of pyridine derivatives with OP and SP esters. aValues are estimated.
leaving groups become more pronounced. For example, the pKas
of 2,4-dinitrophenol and 2,4-dinitrothiophenol differ by 0.9 units,
whereas the pKas of ethanol and ethanethiol differ by ∼5 units.
For OP esters, the pxy value predicts that a 5-unit pKa decrease
in the leaving group would decrease bnuc by more than 10%,
thereby increasing the dissociative character of the transition
state. Therefore the pKa difference between S and O leaving
groups in nucleic acids would be expected to induce an even
greater increase in the dissociative character of the transition
state than observed here for sulfur substitution in methyl aryl
phosphodiesters. Together with the increases in volume and
bond lengths that accompany sulfur substitution, the more open
transition state structure may explain in part why enzymes catalyze
reactions of phosphorothiolates less efficiently than those of their
natural substrates.
Fig. 4 Reaction maps for transfer of the methyl phosphoryl group from
phenols and thiophenols to pyridine nucleophiles. The horizontal and
Conclusions
vertical scales are calibrated in terms of the overall effective charge change
on the leaving and entering atoms, respectively using values reported in
In summary, we have obtained an initial model for the transition
state structure of nonenzymatic reactions of phosphodiester
and phosphorothiolate diesters with pyridine nucleophiles. Both
oxygen and sulfur leaving groups bear similar effective charges
and undergo similar degrees of bond cleavage in the transition
state. Although we can only estimate bonding extents due to
lack of knowledge about beq, this work provides a useful direct
comparison of the reactions of SP and OP diesters with pyridine
nucleophiles. Both types of esters react with pyridine nucleophiles
via a concerted, dissociative mechanism, but the SP diesters react
via a slightly more open transition state than do the OP diesters.
We conclude that sulfur substitution of the leaving group does
not grossly perturb the reaction pathway, supporting the use of
Scheme 2. The position of the transition state of the reaction of S-aryl
O-methyl phosphorothiolates with pyridines is shown as a closed square
and that of the reaction of aryl methyl phosphates with pyridines is shown
as an open square.
emerges from the current work. Comparison of our data for SP
diesters to these OP data suggests that nucleophiles react with the
SP diesters via a similar mechanism and transition state structure.
The analysis above provides a foundation for studying the
phosphoester transfer in biological systems. In the case of nucleic
acid substrates, in which the leaving oxygen atoms are bonded to
alkyl groups rather than to aryl groups like the model compounds
used here, the pKa differences between the sulfur and oxygen
2494 | Org. Biomol. Chem., 2007, 5, 2491–2497
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
phosphorothiolates as substrate analogues to probe mechanisms
of enzyme-catalyzed phosphoryl transfer reactions.
J 6.6), 121.43, 127.88, 131.89 (d, J 4.9), 148.43, 151.87 (d, J 4.4);
dP 21.14.
Lithium S-(4-chlorophenyl) O-methyl phosphorothiolate (3a).
A solution of 2a (23 g, 90 mmol) in 50 ml acetone was stirred
with LiCl (3.8 g, 90 mmol) for 8 hours at 20 ◦C; filtration yielded a
white precipitate 3a (15 g, 69%); mp >250 ◦C; (Found: C, 34.11; H,
2.95; P, 12.73; S, 13.99. C7H7ClLiO3PS requires C, 34.38; H, 2.89;
P, 12.66; S, 13.11%); kmax (MeOH)/nm 226 (e/dm3 mol−1 cm−1
81 000); mmax(KBr)/cm−1 3429, 2988, 2946, 2844, 2359, 2343, 1637,
1476, 1390, 1209, 1084, 1013, 820, 783, 744; dH 3.56 (3 H, d, J 12.9),
7.29–7.32 (2 H, m), 7.44–7.47 (2 H, m); dC 54.12 (d, J 6.6), 128.77
(d, J 6.6), 130.06 (d, J 1.6), 134.87 (d, J 3.3), 136.36 (d, J 4.9);
dP 19.26; HRMS (FAB−) calculated for C7H7ClO3PS: 236.9542,
found: 236.9536.
Experimental
Synthesis
S-(4-Chlorophenyl) O,O-dimethyl phosphorothiolate (2a). An
ice cooled solution of bis(4-chlorophenyl) disulfide 1a (20 g,
0.068 mol) in dichloromethane (100 ml) was saturated with
chlorine gas by bubbling it through the solution for 1 hour then
excess gas and solvent were removed by vacuum evaporation.
Trimethyl phosphite (18 ml, 0.15 mol) was added dropwise
under argon over 15 minutes to the stirred solution of 4-
chlorobenzenesulfenyl chloride intermediate. The solution was
stirred for an additional 30 minutes, solvent was then removed
by evaporation under vacuum, and excess trimethyl phosphite
was distilled off at 70 ◦C/0.01 mmHg to give 2a as a colorless oil
(27 g, 79%); mmax(film)/cm−1 3116, 3087, 2966, 2864, 2362, 2345,
1603, 1546, 1459, 1444, 1345, 1263, 1186, 1032, 905, 843, 799, 733,
559 (lit.,33 3070, 1263, 1180); dH(acetone-d6) 3.83 (6 H, d, J 12.7),
7.40–7.44 (2 H, m), 7.58–7.62 (2 H, m); dC 54.57 (d, J 6.6), 126.07
(d, J 7.1), 130.28 (d, J 2.2), 135.78 (d, J 3.3), 136.89 (d, J 5.6); dP
24.53.
Lithium S-(4-methylphenyl) O-methyl phosphorothiolate (3b).
The general synthetic procedure described for the synthesis of
3a was followed. After 24 hours of stirring the reaction mixture
containing 2b, a precipitate of 3b was isolated by filtration (15 g,
64%); mp >250 ◦C; (Found: C, 42.08; H, 4.69; P, 14.00; S, 13.92.
C8H10LiO3PS requires C, 42.87; H, 4.50; P, 13.82; S, 14.30%);
kmax (MeOH)/nm 226 (e/dm3 mol−1 cm−1 67 000); mmax/cm−1 3422,
3020, 2980, 2946, 2844, 2360, 2343, 1636, 1493, 1224, 1082, 810,
782; dH 2.22 (3 H, s), 3.54 (3 H, d, J 12.7), 7.09–7.13 (2 H, m), 7.34–
7.38 (2 H, m); dC 21.06, 54.12 (d, J 6.0), 126.29 (d, J 6.6), 130.87
(d, J 1.6), 135.10 (d, J 4.4), 139.90 (d, J 3.0); dP 20.06; HRMS
(FAB−) calculated for C8H10O3PS: 217.0137, found: 217.0077.
S-(4-methylphenyl) O,O-dimethyl phosphorothiolate (2b).
Compound 2b was prepared from p-tolyldisulfide (15 g,
0.061 mol) using the procedure described for the synthesis of
2a with the following modifications: chlorine gas was bubbled
for 30 minutes, trimethyl phosphite (16 ml, 0.14 mol) was added
dropwise over 10 minutes under argon at 0 ◦C. The resulting
solution of p-tolylsulfenyl chloride was stirred for 10 minutes.
Vacuum distillation yielded 2b as a yellow oil (28 g, 97%);
mmax/cm−1 (KBr) 3022, 2953, 2850, 2358, 1902, 1800, 1596, 1492,
1457, 1261, 1182, 1017, 811, 788, 765, 600, 567; dH(acetone-d6)
2.32 (3 H, d, J 1.7), 3.77 (6 H, d, J 12.5), 7.19–7.23 (2 H, m),
7.46–7.50 (2 H, m); dC 21.04, 54.35 (d, J 6.0), 123.53 (d, J 7.1),
130.99 (d, J 2.2), 135.45 (d, J 4.9), 140.21 (d, J 3.3); dP 25.56.
Lithium S-(4-nitrophenyl) O-methyl phosphorothiolate (3c).
The general procedure described for the synthesis of 3a was
followed except that the reaction mixture containing 2c was stirred
for 15 hours. Crystals were isolated by filtration, and dried under
vacuum to yield 6.2 g (90%) of 3c containing 0.77 equivalents of
1
acetone (by H NMR). Removal of solvent from crystal lithium
salt proved difficult, recrystallization in chloroform–methanol
(12 : 1) gave 3c with 0.87 equivalents of crystal methanol, mp
◦
>250 C; (Found: C, 30.84; H, 3.22; N, 5.13; P, 10.79; S, 11.93.
C7H7LiNO5PS·0.87 CH3OH requires C, 33.40; H, 3.73; N, 4.95;
P, 10.94; S, 11.33%.); kmax (MeOH)/nm 306 (e/dm3 mol−1 cm−1
68 000); mmax/cm−1 (KBr) 3380, 3102, 2984, 2845, 2366, 2345, 1638,
1598, 1578, 1514, 1344, 1236, 1082, 1045, 854, 787, 744, 685, 580;
dH 3.61 (3 H, d, J 12.9), 7.69–7.72 (2 H, m), 8.11–8.14 (2 H, m); dC
54.21, 124.82, 134.21 (d, J 5.5), 140.86 (d, J 6.0); dP 17.23; HRMS
(FAB−) calculated for C7H7NO5PS: 247.9783, found: 247.9796.
S-(4-nitrophenyl) O,O-dimethyl phosphorothiolate (2c). Under
argon, trimethyl phosphite (8.8 ml, 0.074 mol) was added dropwise
over 10 minutes to a solution of 4-nitrobenzenesulfenyl chloride
(13 g, 0.068 mol) in dichloromethane (55 ml); the solution was
brought to reflux and then stirred for an additional hour at room
temperature. Dichloromethane was evaporated, and the residue
was recrystallized in 15 ml methanol to yield 2c (10 g, 65%); mp
56–58 ◦C; mmax/cm−1 (KBr) 3104, 2955, 2852, 2366, 2345, 1578,
1513, 1444, 1349, 1265, 1179, 1012, 854, 835, 789, 765, 746, 555;
dH (acetone-d6) 3.85 (6 H, d, J 12.7), 7.86–7.90 (2 H, m), 8.22–
8.26; dC 54.97 (d, J 6.0), 124.99 (d, J 1.6), 135.51 (d, J 5.5), 136.54
(d, J 6.6), 148.96; dP 14.32.
Lithium S-(2,4-dinitrophenyl) O-methyl phosphorothiolate (3d).
The general procedure described for the synthesis of 3a was
followed. A reaction mixture containing starting material 2d
became yellow, and crystals formed after 4 hours, crystals were
isolated and dried under vacuum. The yield of 3d (16 g) was
quantitative, however acetone was present. Recrystallization in
chloroform–methanol (4 : 1) was accompanied by decomposition,
and yellow crystals of 3d were recovered containing 0.7 equivalents
of methanol, mp >250 ◦C; Found: C, 30.84; H, 3.22; N, 5.13;
P, 10.79; S, 11.93. C7H7LiNO5PS·0.7 CH3OH requires C, 33.40;
H, 3.73; N, 4.95; P, 10.94; S, 11.33%.); kmax (MeOH)/nm 312
(e/dm3 mol−1 cm−1 67 000); mmax/cm−1 (KBr) 3418, 3110, 2952,
2848, 2364, 2344, 1595, 1529, 1458, 1344, 1267, 1093, 1044, 917,
843, 786, 746, 735, 669; dH 3.59 (3 H, d, J 12.9), 8.14 (1 H, d, J
9.0), 8.31 (1 H, dd, J 9.0), 8.72 (1 H, d, J 2.2); dC 54.45 (d, J 6.6),
S-(2,4-dinitrophenyl) O,O-dimethyl phosphorothiolate (2d).
The general procedure as described for the synthesis of 2b was fol-
lowed. The reaction mixture containing 2,4-dinitrobezenesulfenyl
chloride starting material was stirred for 2 hours at 20 ◦C after
addition of trimethyl phosphite. Recrystallization in methanol
gave 2d (15 g, 73%); mp 69–70 C; mmax/cm−1 (KBr) 3116, 3087,
2966, 2864, 2362, 2345, 1603, 1546, 1345, 1263, 1186, 1023, 905,
843, 799, 734, 559; dH 3.87 (6 H, d, J 12.9), 8.33 (1 H, dd, J 1.7 and
8.8), 8.56 (1 H, dd, J 2.7 and 8.8), 8.84 (1 H, d, J 2.4); dC 55.60 (d,
◦
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 2491–2497 | 2495
©

View Article Online
121.64, 127.75, 135.78 (d, J 4.4), 138.02 (d, J 4.4), 146.61, 150.00
(d, J 2.7); dP 14.74; HRMS (FAB−) calculated for C7H7N2O7PS:
292.9633, found: 292.9622.
Kinetics of reactions of oxyanion nucleophiles with
S-(4-nitrophenyl) methyl phosphorothiolate
In measuring k2 for the reaction of hydroxide nucleophile with
S-(4-nitrophenyl) methyl phosphorothiolate, potassium hydroxide
concentrations of 0.3 to 1.7 M were used. In addition, to determine
the rate by following the formation of methyl phosphate, the kobs
value was calculated from conversion of substrate peak to the
product thiophenolate peak. The latter measurement gave slightly
lower k2. This may be due to the formation of a by-product, as
suggested by new resonances in the aryl region (according to the
1H NMR spectra). We did not identify the structure of the by-
product. The k2 value was not corrected for carbon attack because
we didn’t detect any products arising from attack at carbon.
In reactions of peroxide nucleophile with S-(4-nitrophenyl)
methyl phosphorothiolate, peroxide nucleophile stock solution
was prepared fresh with one equivalent of potassium hydroxide.
Nucleophile concentration ranged from 0.1 to 0.3 M. Values of
kobs were determined by following the reaction over the first two
half lives.
Measurement of kobs and k2
All reaction solutions contained 10 mM arylmethylphosphate
(OP) or phosphorothiolate esters (SP) in deuterium oxide and
were maintained at 1.7 M ionic strength using potassium chloride.
Reactions were carried out in NMR tubes at room temperature
and kept in the dark. A Bruker Avance DMX 500 (500 MHz) and
1
Avance DRX 400 (400 MHz) were used to record H NMR and
31P NMR, 1H NMR was referenced to deuterium oxide (dH 4.63)
and 31P NMR was referenced externally to 85% H3PO4 (dP 0.00).
Generally, to measure kobs, we monitored the conversion of
substrate into methyl phosphate by 1H NMR, reactions were
followed for at least seven time points and, unless otherwise
noted, rates were measured for more than 3.5 half lives. Values of
kobs were measured for five different nucleophile concentrations.
A total of five to six runs were performed for each substrate–
nucleophile combination. The second order rate constant, k2,
was obtained from the slopes of the linear plots of kobs versus
nucleophile concentration. We used 31P NMR to confirm the
assignment of the product peak as methyl phosphate for each
substrate–nucleophile combination. Precipitate developed in some
reactions with phosphorothiolate diesters but not with phosphate
diesters. The formation of precipitate did not appear to affect
Kinetics of reactions of phosphorothiolate and phosphate diesters
with pyridine nucleophiles
Reactions of phosphorothiolate and phosphate diesters with
pyridine nucleophiles were buffered with 30 mM potassium
bicarbonate buffer (pH 10). The concentration ranged from 0.1
to 0.3 M for the picoline nucleophiles and from 0.1 to 0.5 M
for all other pyridine nucleophiles. Concentrations of bases were
not adjusted for association, as we observed, under our reaction
conditions, no obvious curvature in the plots of kobs vs. pyridine
nucleophile concentration (pyridine nucleophile concentration ≤
0.5 M, 23 ◦C; 1.7 M ionic strength).
measurements of kobs
.
Measurement of phosphorothiolate diester and methyl phosphate
relaxation times
To acquire accurate integration values by 31P NMR for phospho-
rothiolate diesters and methyl phosphate, we must allow the 31P nu-
clei to relax fully following excitation. Therefore, we measured the
relaxation times of the phosphorous nucleus for each compound
by a standard inversion recovery test. Relaxation times (Table 1)
were 5.303 s, 5.339 s, 4.906 s, and 3.390 s for phosphorothiolates
3a-d, respectively. Variation of phosphorothiolate concentration
does not appear to influence relaxation time measurements except
for 3a. Relaxation times of 3a were measured to be 5.123 s at
50 mM and 5.483 s at 200 mM. The average value was used.
Methyl phosphate, synthesized by hydrolysis of phosphorothiolate
diesters with 1.0 M potassium hydroxide, had an average relaxation
time of 11.003 s (literature,56 12 s).
In the reaction of 3-picoline with 4c and 5c–d, kobs was
determined by following reactions for more than 3.5 half lives.
For 4a–b and 5a–b, kobs was determined from initial rates. For
phosphate diester 5a, ten runs were performed and kobs was
averaged due to the slow rate of this reaction. NMR tubes were
flame closed to prevent the evaporation of 3-picoline.
In the reaction of 3-cyanopyridine with phosphate diester 5d,
a substituted pyridine by-product was observed. Although this
by-product does not appear to affect kobs over the first 1.5 half
lives, it does affect it increasingly thereafter. Consequently, kobs
was determined from the initial rate for this reaction (seven time
points). The by-product was isolated by TLC (ethyl acetate–
1
hexane, 6 : 1) and identified by H NMR and mass spectra as
nicotinamide.
Table 1 Relaxation times of the phosphorous nucleus for SP diesters
Acknowledgements
Relaxation times/sa
Chemical shift (dP)
Jing-Dong Ye is a research associate, Christofer D. Barth is a
Medical Student Fellow and Joseph A. Piccirilli is an Investigator
of the Howard Hughes Medical Institute.
3a
5.123
5.483 (50 mM)
5.339
4.906
3.39
10.823 (50 mM)
11.183
11.184
17.54
17.77
20.00
18.98
15.0
5.15
5.30
3b
3c
3d
References
Methyl phosphate
1 M. J. Fedor and J. R. Williamson, Nat. Rev. Mol. Cell Biol., 2005, 6,
399–412.
2 E. A. Galburt and B. L. Stoddard, Biochemistry, 2002, 41, 13851–13860.
3 F. H. Westheimer, Science, 1987, 235, 1173–1178.
5.23
a Concentrations are at 200 mM unless otherwise noted.
2496 | Org. Biomol. Chem., 2007, 5, 2491–2497
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
4 J. A. Piccirilli, J. S. Vyle, M. H. Caruthers and T. R. Cech, Nature, 1993,
361, 85–88.
5 S. Shan, A. Yoshida, S. Sun, J. A. Piccirilli and D. Herschlag, Proc.
Natl. Acad. Sci. U. S. A., 1999, 96, 12299–12304.
6 E. J. Sontheimer, P. M. Gordon and J. A. Piccirilli, Genes Dev., 1999,
13, 1729–1741.
7 R. X. Xu, A. M. Hassell, D. Vanderwall, M. H. Lambert, W. D. Holmes,
M. A. Luther, W. J. Rocque, M. V. Milburn, Y. Zhao, H. Ke and R. T.
Nolte, Science, 2000, 288, 1822–1825.
8 N. C. Horton and J. J. Perona, Nat. Struct. Biol., 2001, 8, 290–293.
9 M. Nowotny and W. Yang, EMBO J., 2006, 25, 1924–1933.
10 D. E. Wilcox, Chem. Rev., 1996, 96, 2435–2458.
11 W. Richter, Proteins, 2002, 46, 278–286.
29 J. M. Warnecke, E. J. Sontheimer, J. A. Piccirilli and R. K. Hartmann,
Nucleic Acids Res., 2000, 28, 720–727.
30 E. A. Galburt, B. Chevalier, W. Tang, M. S. Jurica, K. E. Flick, R. J.
Monnat, Jr and B. L. Stoddard, Nat. Struct. Biol., 1999, 6, 1096–1099.
31 A. J. Kirby and M. Younas, J. Chem. Soc. B, 1970, 1165–1172.
32 S. A. Ba-Saif, A. M. Davis and A. Williams, J. Org. Chem., 1989, 54,
5483–5486.
33 L. L. Murdock and T. L. Hopkins, J. Org. Chem., 1968, 33, 907–908.
34 A. J. Kirby and W. P. Jencks, J. Am. Chem. Soc., 1965, 87, 3209–3216.
35 D. Herschlag and W. P. Jencks, J. Am. Chem. Soc., 1990, 112, 1951–
1956.
36 K. Pilgram and F. Korte, Tetrahedron, 1965, 21, 1999–2013.
37 E. A. Castro and C. Ureta, J. Chem. Soc., Perkin Trans. 2, 1991, 63–68.
38 J. E. Leffler, Science, 1953, 117, 340.
12 C. Walsh, in Enzymatic Reaction Mechanisms, W. H. Freeman, San
Francisco, 1979, pp. 184–207.
13 H. Ke and H. Wang, Curr. Top. Med. Chem., 2007, 7, 391–403.
14 R. B. Silverman, in The Organic Chemistry of Enzyme-Catalyzed
Reactions, Academic Press, San Diego, 2000.
39 A. M. Davis, A. D. Hall and A. Williams, J. Am. Chem. Soc., 1988, 110,
5105–5108.
40 A. Williams, in Advances in physical organic chemistry, ed. D. Bethell,
Academic Press Limited,1992, vol. 27, pp. 1–55.
15 A. T. Perrotta, I. H. Shih and M. D. Been, Science, 1999, 286, 123–126.
16 S. Nakano, D. M. Chadalavada and P. C. Bevilacqua, Science, 2000,
287, 1493–1497.
17 S. R. Das and J. A. Piccirilli, Nat. Chem. Biol., 2005, 1, 45–52.
18 R. Pinard, K. J. Hampel, J. E. Heckman, D. Lambert, P. A. Chan, F.
Major and J. M. Burke, EMBO J., 2001, 20, 6434–6442.
19 Y. I. Kuzmin, C. P. Da Costa, J. W. Cottrell and M. J. Fedor, J. Mol.
Biol., 2005, 349, 989–1010.
41 N. Bourne and A. Williams, J. Org. Chem., 1984, 49, 1200–1204.
42 D. J. Hupe and W. P. Jencks, J. Am. Chem. Soc., 1977, 99, 451–464.
43 A. J. Kirby and Varvogli Ag, J. Chem. Soc. B, 1968, 135–141.
44 D. A. Jencks and W. P. Jencks, J. Am. Chem. Soc., 1977, 99, 7948–7960.
45 W. P. Jencks, S. R. Brant, J. R. Gandler, G. Fendrich and C. Nakamura,
J. Am. Chem. Soc., 1982, 104, 7045–7051.
46 W. P. Jencks, M. T. Haber, D. Herschlag and K. L. Nazaretian, J. Am.
Chem. Soc., 1986, 108, 479–483.
20 S. Loverix, A. Winqvist, R. Stromberg and J. Steyaert, Chem. Biol.,
2000, 7, 651–658.
21 R. T. Raines, Chem. Rev., 1998, 98, 1045–1066.
22 E. J. Sontheimer, S. G. Sun and J. A. Piccirilli, Nature, 1997, 388, 801–
805.
47 D. Herschlag and W. P. Jencks, J. Am. Chem. Soc., 1989, 111, 7579–
7586.
48 D. Herschlag and W. P. Jencks, J. Am. Chem. Soc., 1989, 111, 7587–
7596.
49 J. B. Pedley and E. M. Marshall, J. Phys. Chem. Ref. Data, 1983, 12,
967–1031.
50 J. Drowart, C. E. Myers, R. Szwarc, A. Vanderau and O. M. Uy, High
Temp. Sci., 1973, 5, 482–488.
51 R. A. More O’Ferrall, J. Chem. Soc. B, 1970, 274.
52 W. P. Jencks, Chem. Rev., 1972, 72, 705–718.
23 P. M. Gordon, E. J. Sontheimer and J. A. Piccirilli, RNA, 2000, 6,
199–205.
24 P. M. Gordon and J. A. Piccirilli, Nat. Struct. Biol., 2001, 8, 893–898.
25 A. Yoshida, S. Shan, D. Herschlag and J. A. Piccirilli, Chem. Biol.,
2000, 7, 85–96.
26 W. B. Derrick, C. H. Greef, M. H. Caruthers and O. C. Uhlenbeck,
Biochemistry, 2000, 39, 4947–4954.
27 S. Wang, K. Karbstein, A. Peracchi, L. Beigelman and D. Herschlag,
Biochemistry, 1999, 38, 14363–14378.
28 K. Suzumura, M. Warashina, K. Yoshinari, Y. Tanaka, T. Kuwabara,
M. Orita and K. Taira, FEBS Lett., 2000, 473, 106–112.
53 P. J. O’Brien and D. Herschlag, Biochemistry, 2002, 41, 3207–3225.
54 A. C. Hengge, A. E. Tobin and W. W. Cleland, J. Am. Chem. Soc., 1995,
117, 5919–5926.
55 D. M. Zhou and K. Taira, Chem. Rev., 1998, 98, 991–1026.
56 D. Herschlag, J. A. Piccirilli and T. R. Cech, Biochemistry, 1991, 30,
4844–4854.
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 2491–2497 | 2497
©