Sulfurization of dinucleoside phosphite triesters with chiral disulfides

Article abstract of DOI:10.1080/15257770.2011.597366

Sixteen chiral analogues of phenylacetyl disulfide (PADS) and 5-methyl-3H-1,2,4-dithiazol-3-one (MEDITH) were used to sulfurize five dithymidine phosphite triesters, each incorporating a β-cyanoethoxy or siloxy group. Each mixture of S_P:R_P phosphite triester diastereomers was combined with approximately one fourth of an equivalent of each of the sulfurizing reagents, and the R_PS:S_PS diastereomer ratios of the resulting phosphite sulfides or phosphorothioates were determined by reverse-phase HPLC. Diastereoselectivities and corresponding diastereomeric excess (de) values were calculated by correcting for the starting triester diastereomer ratios. The highest de values for R_PS and S_PS phosphorothioates were 14.7% and 7.9%, respectively, both using MEDITH analogues. Copyright Taylor and Francis Group, LLC.

Full text of DOI:10.1080/15257770.2011.597366

This article was downloaded by: [Cranfield University]
On: 23 April 2014, At: 09:46
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Nucleosides, Nucleotides and Nucleic
Acids
Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/lncn20
Sulfurization of Dinucleoside Phosphite
Triesters with Chiral Disulfides
Joshua A. Mukhlall a & William H. Hersh a
a Department of Chemistry and Biochemistry , Queens College and
the Graduate Center of the City University of New York , Flushing,
New York, USA
Published online: 08 Sep 2011.
To cite this article: Joshua A. Mukhlall & William H. Hersh (2011) Sulfurization of Dinucleoside
Phosphite Triesters with Chiral Disulfides, Nucleosides, Nucleotides and Nucleic Acids, 30:9, 706-725,
DOI: 10.1080/15257770.2011.597366
To link to this article: http://dx.doi.org/10.1080/15257770.2011.597366
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the
“Content”) contained in the publications on our platform. However, Taylor & Francis,
our agents, and our licensors make no representations or warranties whatsoever as to
the accuracy, completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content
should not be relied upon and should be independently verified with primary sources
of information. Taylor and Francis shall not be liable for any losses, actions, claims,
proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or
howsoever caused arising directly or indirectly in connection with, in relation to or arising
out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Conditions of access and use can be found at http://www.tandfonline.com/page/terms-
and-conditions

Nucleosides, Nucleotides and Nucleic Acids, 30:706–725, 2011
Copyright ꢀC Taylor and Francis Group, LLC
ISSN: 1525-7770 print / 1532-2335 online
DOI: 10.1080/15257770.2011.597366
SULFURIZATION OF DINUCLEOSIDE PHOSPHITE TRIESTERS
WITH CHIRAL DISULFIDES
Joshua A. Mukhlall and William H. Hersh
Department of Chemistry and Biochemistry, Queens College and the Graduate Center of the
City University of New York, Flushing, New York, USA
Sixteen chiral analogues of phenylacetyl disulfide (PADS) and 5-methyl-3H-1,2,4-dithiazol-3-
2
one (MEDITH) were used to sulfurize five dithymidine phosphite triesters, each incorporating a
β-cyanoethoxy or siloxy group. Each mixture of SP:RP phosphite triester diastereomers was combined
with approximately one fourth of an equivalent of each of the sulfurizing reagents, and the RPS:SPS
diastereomer ratios of the resulting phosphite sulfides or phosphorothioates were determined by reverse-
phase HPLC. Diastereoselectivities and corresponding diastereomeric excess (de) values were calcu-
lated by correcting for the starting triester diastereomer ratios. The highest de values for RPS and SPS
phosphorothioates were 14.7% and 7.9%, respectively, both using MEDITH analogues.
Keywords Sulfurization; chiral disulfides; dinucleoside phosphite triesters; antisense;
phosphorothioate oligonucleotides
INTRODUCTION
A variety of soluble sulfurization reagents[1,2] have been developed for
the synthesis of phosphorothioate oligonucleotides, materials that have at-
tracted intense interest as DNA analogues for antisense applications.[3,4]
While the substitution on the phosphate backbone of a terminal oxygen
with a sulfur atom gives an oligonucleotide that is more biologically stable
yet still possesses biological activity, this substitution also gives a new stereo-
genic center at every phosphorus atom on the oligonucleotide. Standard
DNA synthesis methods give rise to a random mixture of configurations at
phosphorus before sulfurization occurs, and following stereoretentive sul-
furization, a mixture of RPS and SPS diastereomers forms. These mixtures
are still acceptable for pharmaceutical use, but efforts to develop efficient
Received 5 January 2011; accepted 10 June 2011.
We thank William Berkowitz for helpful discussions of calculations of diastereomer selectivities,
Queens College for the purchase of HPLC equipment, and the City University of New York PSC-CUNY
Research Award Program for financial support.
Address correspondence to William H. Hersh, Department of Chemistry and Biochemistry, Queens
College and the Graduate Center of the City University of New York, Flushing, NY 11367-1597, USA.
E-mail: william.hersh@qc.cuny.edu
706

Sulfurization of Dinucleoside Phosphite Triesters with Chiral Disulfides
707
strategies for stereoselective phosphorylation to control stereochemistry at
phosphorus have been reported for years.[5–10] However, each reported
method evidently has flaws that prevent large-scale development, either in
the synthesis of the chiral phosphorylating reagents or in the coupling effi-
ciency.[11]
As part of a program to develop a new method for the synthesis of P-
stereogenic phosphorothioates, we have begun to examine the sulfurization
reaction. A recent report by Mikolajczyk’s group described the use of chiral
disulfides for the kinetic resolution of phosphines,[12] so a similar approach
involving the synthesis of chiral analogues of reagents used for phospho-
rothioate synthesis seemed warranted. In principle, an achiral reagent could
be used for oligonucleotide phosphite sulfurization, since the carbohydrate
rings attached to the phosphorus atoms generate P-epimers that are diastere-
omers, rather than the enantiomers examined by Mikolajczyk. Starting with
a mixture of phosphite epimers, the achiral sulfurization reagent could selec-
tively give one diastereomer at early reaction, but a chiral sulfurizing reagent
might do so more readily by double stereodifferentiation.[13]
We recently reported the synthesis of two classes of chiral disulfides, one
based on phenylacetyl disulfide[14,15] [PADS, (1a)–(1d)] and the other on
5-methyl-3H-1,2,4-dithiazol-3-one[16] [MEDITH, (2a)–(2e); Figure 1.[2] We
note that the usual name quoted in the literature, 3-methyl-1,2,4-dithiazolin-
5-one, appears to us to be in error; the parent ring system is fully unsaturated
(hence the “ol” ring termination) and the position of the unsaturation is in-
dicated by giving the H the lowest available number.[17] Since the P-epimers
do not interconvert at room temperature,[18] selective sulfurization alone
will not suffice to give P-stereogenic phosphorothioates. Nevertheless, in this
study, we report the results of screening 16 chiral disulfides with five dinucle-
oside phosphite triesters, including the synthesis of dinucleosides with differ-
ent degrees of steric hindrance at the phosphite triester, analytical methods
for measuring diastereomer ratios of the resultant phosphite sulfides, and
the derivation of a means of calculation of the sulfurization selectivity.
FIGURE 1 Chiral sulfurization reagents.

708
J. A. Mukhlall and W. H. Hersh
RESULTS AND DISCUSSION
Phosphite Triesters
Initial screening was planned using the readily available dithymidine
phosphite triester (3), chosen since the β-cyanoethoxy group would nor-
mally be present using the phosphoramidite method of oligonucleotide syn-
thesis (Scheme 1). However, stereodifferentiation at phosphorus could be
difficult, since two of the alkoxy groups are primary and one is secondary. In
order to circumvent this, we included the tertiary 1,1-dimethyl-2-cyanoethoxy
protecting group in (4).[19,20] In addition, we tested the siloxy phosphite tri-
esters (5)–(7), which are readily available via the H-phosphonate (8) method
of oligonucleotide synthesis. In this way, the phosphite triester would consist
of a 1◦ alkoxy group due to the 5ꢁ-thymidine, a 2◦ alkoxy group due to the
3ꢁ-thymidine, and a 3◦-like alkoxy group due to the siloxy group, and thereby
provide a higher degree of steric bulk around the phosphorus, which might
allow greater selectivity for sulfurization.
SCHEME 1 Phosphite triesters for screening, sulfurization, and phosphorothioate formation for analysis.
Phosphite triester (3) was synthesized as previously described[18] and (4)
was readily prepared in a comparable manner starting from 1,1-dimethyl-
2-cyanoethanol[21] and (i-Pr2N)2PCl. Conversion of (8) and close ana-
logues to trimethylsiloxy phosphite triesters analogous to (5) using N ,O-
bis(trimethylsilyl)acetamide (BSA)[22–24] and other silylating reagents[23,25]
is known; procedures for the synthesis of phosphite triesters (6) and (7) have
not been reported, but the desired conversions were expected to proceed us-
ing tert-butyldimethylsilyl chloride (TBDMSCl) and Et3N and Ph3SiCl/Et3N,
respectively.[23,25,26] H-phosphonate (8) was prepared following literature
procedures reported for the 5ꢁ-O-TBDMS analogue.[26] Initial experiments
to find the best silylating conditions for the synthesis of phosphite triesters
(5) and (6) (Scheme 1) were carried out by simultaneously treating 5 mg
each of H-phosphonate (8) with 3 equivalents of BSA (the amount reported
by Shaw[23]) and TBDMSCl/Et3N in anhydrous acetonitrile-d3 at room tem-
perature. Since H-phosphonate (8) is a mixture of diastereomers, only two
peaks corresponding to the RP and SP diastereomers of phosphite triesters
(5) and (6) were expected after complete silylation. However, after ∼30

Sulfurization of Dinucleoside Phosphite Triesters with Chiral Disulfides
709
FIGURE 2 (A) 31P NMR spectrum of the reaction mixture of (8) and 3 equivalents of BSA after 1 day,
in the attempted synthesis of phosphite triester (5). (B) 31P NMR spectrum of the reaction mixture of
(8) and 5 equivalents of BSA after 20 minutes.
minutes reaction with BSA, the 31P NMR spectrum of the reaction solution
exhibited six peaks in the phosphite region near 127 ppm; no change was
observed after 1 day (Figure 2A). Similarly, the 31P NMR spectrum of the
reaction with TBDMSCl/Et3N exhibited several peaks in the phosphite tri-
ester region as well as in the H-phosphonate region (∼8 ppm) after ∼2
hours and was unchanged after 1 day. Surprisingly, an addition of 5 rather
than 3 equivalents of BSA to (8) gave two peaks in the 31P NMR spectrum
after 20 minutes (Figure 2B) and similarly 5 equivalents of TBDMSCl or
Ph3SiCl with 10 equivalents of NEt3 gave two peaks in the 31P NMR spectra
after 2 hours, each in virtually the same 56:44 RP:SP ratio as that seen for (8).
We suggest that the source of the additional peaks in the 31P NMR spectra
is simply the partial silylation of the carbonyl moieties on the thymine rings;
that is, silylation could occur at the 3ꢁ- and/or 5ꢁ-phosphorylated RP and
SP nucleosides. However, once a large excess of silylating reagent is added,
complete silylation can give only the RP and SP fully silylated P-epimers.
Analytical Methods for Determination of the Diastereomer Ratios
Each of the phosphite triesters (3)–(7) was converted to the corre-
sponding phosphite sulfides (9)–(13) by treatment with excess sulfur [(9)
and (11)–(13)] or MEDITH (10) at room temperature (Scheme 1).
Sulfur is known to react in a stereoretentive manner with chiral
phosphite triesters,[24,27] although the labels change according to the
Cahn–Ingold–Prelog rules and the RP epimer gives rise to the SPS epimer
upon sulfurization. Diastereomer assignments of the sulfides were made
by comparisons of the observed ratios to those of the phosphite triesters
from the 31P NMR spectra, although the ratios for (3) and (9) are so close
to 1:1 that the assignment must be viewed with caution; NMR and HPLC
(vide infra) data may be found in the Experimental section. Because the 31
P

710
J. A. Mukhlall and W. H. Hersh
NMR sulfide peak separations were quite small, however, it was clear that
they could not be integrated with sufficient precision to allow the detection
of small sulfurization selectivities. We therefore sought to use HPLC[28] for
the measurement of diastereomer ratios.
The diastereomers of (9) and (10) were examined first, and the di-
astereomers of (9) gave baseline separation by elution on a C18 column
with 65:35 acetonitrile/water with an integrated ratio similar to that seen by
NMR. However, we were unable to find any conditions for the separation
of the diastereomers of (10). We did not attempt any HPLC analysis of the
oxygen-sensitive phosphite triesters (3)–(7) or the water-sensitive siloxy es-
ter sulfides (11)–(13), but for these sulfides as well as (10), conversion to
phosphorothioate (14) was a potential alternative.
We expected that mild base treatment of (10) would give rise to stereo-
specific elimination of dimethylacrylonitrile,[29,30] and in fact, addition of 3.5
equivalents of diazabicycloundecene (DBU) followed by treatment with 2
M aqueous triethylammonium bicarbonate (TEAB) gave phosphorothioate
(14) (Scheme 1). The diastereomer ratios of (4) and (10) measured by 31
P
NMR were the same (45.5:54.5), and the diastereomers of (14) gave baseline
separation on a C8 column with 55:45 acetonitrile:0.1 M triethylammonium
acetate (TEAA), giving a diastereomer ratio of 45.66:54.34. In a similar man-
ner, the siloxy ester sulfides (11)–(13) were treated with TEAB[25] to give
the diastereomers of phosphorothioate (14) (Scheme 1), a reaction previ-
ously shown to be stereospecific,[27] in essentially the same ratio as that of
the H-phosphonate (8) precursor; the HPLC ratio of the diastereomers of
(14) in the final mixture was 43.44:56.56. Identification of (14) was con-
firmed by isolation from the reaction mixture and chromatography, giving
material that matched the beautifully detailed NMR data in CDCl3 reported
by Stawinski and coworkers.[31] The HPLC ratios of (14) from (12) and
(13) were the same within experimental error, as expected since these too
were derived from the same sample of (8). Assignments of configurations
of (4)–(7) and (10)–(14) were all determined on the basis of the reported
assignments of (8),[27] and since we obtained the diastereomers of (8) in
a ∼45:55 ratio, it was easy to assign the products formed from (8) [that is,
(11)–(14)] and the precursors of (11)–(14) [that is, (4)–(7) and (10)] on
the basis of that 45:55 ratio. It should be noted that (4) and (10) were not
derived from (8), but once the diastereomers of (14) were assigned, the
corresponding diastereomers of (4) and (10) that gave rise to (14) could be
assigned.
Reaction of Sulfurizing Reagents with Phosphite Triesters
Procedure
The phosphite triesters consist of a mixture of RP and SP diastereomers,
which do not interconvert on the reaction time scale.[18] Addition of less

Sulfurization of Dinucleoside Phosphite Triesters with Chiral Disulfides
711
than one equivalent of a sulfurizing reagent to the mixture would then
allow any selectivity for one P-epimer to be detected. The use of smaller
amounts of sulfurizing reagent would allow a more accurate determination
of selectivity, but since a reasonable amount of product was needed, approx-
imately one fourth of an equivalent seemed to represent a good compro-
mise. In order to increase the possibility of greater selectivity, the reactions
were carried out at −32◦C, necessitating a reaction time of 20 hours. After
sulfurization, anhydrous tert-butylhydroperoxide (TBHP)[32] was added to
quench the reaction in order to convert the remaining phosphite triester
to the phosphite oxide.[26,33] The SPS:RPS diastereomer ratio determined
by HPLC of the phosphite sulfide (9) after chiral sulfurization or of the
phosphorothioate (14) after sulfurization and elimination (10) or hydroly-
sis [(11)–(13)] can then be compared with the RP:SP ratio of the starting
phosphite triester—determined in the same way by complete, nonselective
sulfurization with excess sulfur or MEDITH—to determine the selectivity of
the reaction (vide infra).
Sulfurization Yields
While each screening reaction was carried out with ∼25 mol% of the
sulfurizing reagent with respect to the amount of phosphite triester used,
analysis of each reaction by 31P NMR and/or HPLC showed that more than
30% of the phosphite triester had been sulfurized in ∼30% of the reactions.
We ascribe this to weighing errors in the glove box, since there was no pattern
that might suggest (for instance) sulfurization by more than one sulfur atom
of the disulfides. A plot of the extent of sulfurization for each sulfurizing
reagent for each triester shows that the extent of sulfurization for most of the
reactions ranged from 20% to 40% (Figure 3). It can be seen that phosphite
triester (3) is less reactive toward PADS, especially its chiral analogues, and
phosphite triester (5) is generally more reactive to many sulfurizing reagents
compared with phosphite triesters (6) and (7).
RPS:SPS Diastereomer Ratios by HPLC
The RPS:SPS diastereomer ratios determined by HPLC analysis of phos-
phite sulfide (9) and phosphorothioate (14) derived from phosphite triesters
(10)–(13) are shown in Table 1. The first row displays data for reactions with
excess sulfur [or for (4), excess MEDITH instead], and so gives the initial
triester diastereomer ratio. The remaining data points are those for partial
sulfurization, that is, addition of ∼25 mol% of the sulfurizing reagent. The
data in Table 1 might in principle be corrected by taking into account the
fact that the sulfurizing reagents were not all 100% enantiomerically pure,[2]
but the correction was not necessary since the diastereomer ratio would only
change by ∼0.15% at most.[34]

712
J. A. Mukhlall and W. H. Hersh
100
80
60
40
20
0
(3)
(4)
(5)
(6)
(7)
Sulfurizing reagents
FIGURE 3 Extent of sulfurization of phosphite triesters (3)–(7) by ∼25 mol% of sulfurizing reagent.
Derivation and Calculation of Selectivity of the Chiral Sulfurization Reactions
For a reaction in which a single isomer reacts to give two stereoisomers,
the isomer ratio can be expressed as the enantiomeric excess (ee) if they are
enantiomers, or diastereomeric excess (de) if they are diastereomers. Unlike
a reaction that has a single isomer as the starting substrate, calculation
of the selectivity of a reaction such as phosphite triester sulfurization, in
which the starting substrate is a mixture of diastereomers, is neither standard
nor intuitive. Since sulfurization of the two diastereomers of a phosphite
triester occurs in a stereoretentive manner,[24,27] the RP configuration almost
certainly goes to the SPS phosphorothioate and the SP configuration goes
to the RPS phosphorothioate. The diastereoselectivities[35,36] for the R and S
configurations are expressed, respectively, as dsR(PS) and dsS(PS). That is, if a
sulfurizing reagent was 100% selective for the formation of SPS from RP, then
in that case, dsS(PS) = 100% and dsR(PS) = 0%, and we would only expect RP
to react leaving SP untouched. This leads to Equation (1), and for the case of
100% selectivity, the equation can always be inverted to prevent a zero in the
(SPS/RPS) = (RP/SP) × (dsS(PS)/dsR(PS)
dsS(PS) = 1 − dsR(PS)
)
(1)
(2)
(3)
(4)
dsR(PS) = RPS × RP/[(SP × SPS) + (RPS × RP)]
de = dsS(PS) − dsR(PS)

713

714
J. A. Mukhlall and W. H. Hersh
denominator. Rearranging Equation (1) to solve for dsS(PS)/dsR(PS) and sub-
stitution of dsS(PS) using Equation (2) give an expression for the diastereo-
selectivity dsR(PS) [Equation (3)], where SP and RP are the initial phosphite
percentages and RPS and SPS are the observed phosphite sulfide percentages
from SP and RP, respectively. Although it is not necessary, we found that
the de of the reaction [Equation (4)] allowed the results to be more easily
visualized; we have chosen that selectivity for the SPS phosphite sulfide or
phosphorothioate, derived from the more abundant RP isomer of (4) and
(8), although the less abundant (by a small amount) RP isomer of (3), will
give a positive de.
Using the RPS:SPS diastereomer ratios in Table 1 along with the initial
SP and RP phosphite triester values taken from the excess sulfur reactions
[or for (4), excess MEDITH], the diastereoselectivity ratios (dsR(PS):dsS(PS)
were calculated using Equation (3) for dsR(PS) and Equation (2) for dsS(PS)
)
.
The diastereoselectivities for (2a) and (2c) are the average values from the
repeated reactions. The resultant de values are plotted in Figures 4–6.
Achiral Sulfurization Selectivity
The use of excess sulfur for the sulfurization of (3)–(7) gives by definition
no selectivity. However, as seen in Figure 4, even the achiral sulfur, PADS,
and MEDITH reagents, when not present in excess, gave small selectivities,
particularly for (3), (4), and (7). In most cases, the selectivities favored the
RPS isomers and so gave negative de values—that is, the major isomer for
(3) but the minor isomer for (4)–(7) were favored. Somewhat surprisingly,
for instance, the least hindered phosphite triester (3) gave de values of
approximately 3% for the three achiral sulfurization reagents, while the
most hindered triesters (4) and (7) less surprisingly gave the highest de
values of −4.7% to −7.6% and −3.2% to −5.3%, respectively.
FIGURE 4 Diastereoselectivity excess (SPS positive) for sulfurization of phosphite triesters (3)–(7) by
achiral sulfurizing reagents.

Sulfurization of Dinucleoside Phosphite Triesters with Chiral Disulfides
715
FIGURE 5 Diastereoselectivity excess (SPS positive) for sulfurization of phosphite triesters (3)–(7) by
chiral PADS analogues (1).
Chiral Sulfurization Selectivity
Separate plots of de for the PADS and MEDITH analogues (Figures
5 and 6) exhibit a range of de values up to +7.5% and −14.7%. For the
PADS analogue sulfurizations, no discernable pattern of diastereoselectivity
was observed except for the reactions of (4), where in most cases a large
negative de was seen. The effect for (4) seems more due to the triester than
to the sulfurizing reagents, since even achiral PADS gave a de of −7.6%,
but an interesting effect was seen for isopropyl-substituted PADS analogue
(1c): the (R,R) isomer gave a small de of +1.2% while the (S,S) isomer gave
a large de of −12.2%. Otherwise, each of the other triesters gave bigger
FIGURE 6 Diastereoselectivity excess (SPS positive) for sulfurization of phosphite triesters (3)–(7) by
chiral MEDITH analogues (2).

716
J. A. Mukhlall and W. H. Hersh
differences between the enantiomeric PADS analogues and so would seem
to be more effective than (4) as screening choices. The highest de other than
for (4) of −11.1% was observed for sulfurization of trimethylsiloxyphosphite
(5) with the hindered isopropyl PADS (R,R)-(1c). The negative de indicates
selectivity for the minor RPS isomer, so the observed 49.0:51.0 RPS:SPS ratio of
(14) from (11) must be referenced to the starting 43.4:56.6 SP:RP ratio of (5)
to generate the (relatively) high de. The largest absolute difference in the
observed diastereomers was the 40.2:59.8 RPS:SPS ratio seen for sulfurization
of (7) by (S,S)-(1d), which translates into a de of +6.4%.
In contrast to the PADS data, the MEDITH analogues clearly exhibited
a pattern for the pairs of enantiomeric disulfides (2a) and (2c)—that is,
each pair gave a similar pattern of selectivities with the five phosphites, al-
though the patterns were different for the different MEDITH analogues. For
(2b) and (2e), the similarities were not strong, but the results nevertheless
suggested that the chiral MEDITH analogues were racemizing under the
reaction conditions, a concern since these compounds had been observed
to racemize upon silica gel chromatography.[2] MEDITH analogue (2a) was
tested for racemization, by running sulfurization reactions of (5) with ex-
cess (R) or (S)-(2a), and the unreacted (2a) was then analyzed by HPLC as
previously described[2] to test for enantiopurity. Complete racemization was
observed.
Because of the racemization of the MEDITH analogues, the sulfurization
selectivities could arise from faster reaction with one enantiomer, or simply
arise as a mixture of two selectivities. Interestingly, higher selectivities were
observed than for the nonracemizing PADS analogues [which were tested
in the same manner, using excess (R,R) and (S,S)-(1a) with (4) and (5), with
the result that no racemization was found]. For instance, (S)-(2d) gave three
of the highest de values of −14.7%, −8.9%, and −12.6% with (5)–(7), and
(S)-(2e) [but not (R)-(2e)] gave a de of −12.5% for the least hindered siloxy
phosphite (5). The most hindered siloxyphosphite, (7), gave the highest de
values of −8.9% to −11.4% with (2a)–(2b), while (S)-(2c) gave the highest
positive de of 7.9% 0.3% with the trimethylsiloxyphosphite (5) [that for
(R)-(2c) was 5.7% 0.9%], and hence the highest absolute selectivity ratio of
39.6:60.4 RPS:SPS. No significant selectivity was seen for any of the MEDITH
sulfurizations of β-cyanoethoxy phosphite (3).
EXPERIMENTAL SECTION
General
NMR spectra were recorded on a 400 MHz Bruker spectrometer with
tetramethylsilane and 85% H3PO4 in CD3CN, CDCl3, and toluene-d8 as ex-
ternal standards. High-resolution mass spectrometry was performed on an
Agilent G6520A Q-TOF instrument. Air- and water-sensitive compounds were

Sulfurization of Dinucleoside Phosphite Triesters with Chiral Disulfides
717
handled in an inert atmosphere glove box under nitrogen. While oligonu-
cleoside reagents do not require glove box use, we have found it to be
extraordinarily convenient in maintaining nonoxidizing conditions for the
more air-sensitive reagents, and in maintaining dry solvents for all reactions
and NMR analyses. Amines, acetonitrile, 1,4-dioxane, dimethyformamide
(DMF), and pyridine were distilled from calcium hydride under nitrogen,
and diethyl ether and tetrahydrofuran (THF) were distilled from sodium
benzophenone under nitrogen. Flash column chromatography was carried
out on 230–400 mesh silica gel, and thin layer chromatography was carried
out on aluminum-backed silica gel F (200 microns). Water (HPLC grade,
submicrofiltered), dichloromethane, hexanes, and acetic acid were used
1
as received. For the peak assignments in the H NMR spectra of (3) and
(4) that follow, the 3ꢁ-phosphorylated thymidine and the 5ꢁ-phosphorylated
thymidine are labeled T1 and T2, respectively; see ref. [18] for details on
the methods of assignment. Details of the syntheses of dinucleosides (3),
(4), and (8) are given below because these procedures do not appear to be
readily available.
Synthesis of 2-Cyanoethyl
5ꢀ-O-(4,4ꢀ-Dimethoxytrityl)Thymidin-3ꢀ-yl
3ꢀ-O-(Tertbutyldimethylsilyl) Thymidin-5ꢀ-yl Phosphite (3)
The synthesis of (3) was carried out on a 0.51 g scale of phosphoramidite
as previously described,[18] but instead of adding Me2SBH3 to the crude
mixture to effect separation of diastereomers, the mixture was taken into
the glove box and purified by using a short column of silica gel (∼20 g
of silica gel in a 60 mL coarse frit) and solvent mixture of THF/hexane
(2:1). All fractions with Rf of 0.55 (THF/hexane 2:1) were evaporated under
reduced pressure to afford (3) as a white foam (0.60 g, 94% yield). 1H NMR
(400 MHz, CD3CN, two diastereomers): δ 9.37 (br s, NH, 4H), 7.48–7.26
(m, includes 4 × H-6, 22H), 6.91–6.88 (m, 8H), 6.30 (m, 2 × T1H-1ꢁ, 2H),
6.19 (m, 2 × T2H-1ꢁ, 2H), 5.03 (m, 2 × T1H-3ꢁ, 2H), 4.41 (m, 2 × T2H-3ꢁ, 2H),
4.12 (m, 2 × T1H-4ꢁ, 2H), 4.08–3.91 (m, 2 × T2H-4ꢁ, 2 × T2H-5ꢁ, 2 × T2H-5ꢁꢁ,
2 × CH2OP, 10H), 3.79 (s, 2 × CH3O, 6H), 3.78 (s, 2 × CH3O, 6H), 3.35
(m, 2 × T1H-5ꢁ, 2 × T1H-5ꢁꢁ, 4H), 2.65 (br t, J = 5.9 Hz, 2 × CH2CN, 4H), 2.44
(m, 2 × T1H-2ꢁ, 2 × T1H-2ꢁꢁ, 4H), 2.19 (m, 2 × T2H-2ꢁ, 2 × T2H-2ꢁꢁ, 4H), 1.83
(m, 2 × CH3C-5, 6H), 1.52 (d, J = 1.1 Hz, CH3C-5, 3H), 1.50 (d, J = 1.1 Hz,
CH3C-5, 3H), 0.91 (s, 2 × (CH3)3CSi, 18H), 0.12 (s, (CH3)2Si, 6H), 0.10
(s, (CH3)2Si, 6H). 31P NMR (161 Mz, CD3CN, two diastereomers): δ 140.41,
140.38; (toluene-d8): δ 139.7 (SP), 139.3 (RP), 50.7:49.3; assigned from the
NMR spectra of the isolated diastereomers.[18] HRMS (ESI): Calculated for
C50H61N5O13PSi [M-H]−: 998.3773, found 998.3777.

718
J. A. Mukhlall and W. H. Hersh
Synthesis of 1,1-Dimethyl-2-Cyanoethyl
5ꢀ-O-(4,4ꢀ-Dimethoxytrityl)Thymidin-3ꢀ-yl
3ꢀ-O-(Tertbutyldimethylsilyl) Thymidin-5ꢀ-yl Phosphite (4)
The starting phosphoramidite, 1,1-dimethyl-2-cyanoethyl 5ꢁ-O-(4,4ꢁ-
dimethoxytrityl)thymidin-3ꢁ-yl N ,N -diisopropylphosphoramidite, was syn-
thesized as follows: under a nitrogen atmosphere, a solution of 5ꢁ-O-(4,4ꢁ-
dimethoxytrityl)thymidine[37] (0.6008 g, 1.103 mmol) in 4 mL CH3CN
was added dropwise over 2–3 minutes to a magnetically stirred solution
of (i-Pr2N)2POC(CH3)2CH2CN (0.7837 g, 2.379 mmol, prepared in 90%
yield using the reported procedure[38] from 1,1-dimethyl-2-cyanoethanol[21]
and (i-Pr2N)2PCl) and N -methylimidazolium trifluoromethanesulfonate[39]
(0.1578 g, 0.6796 mmol) in 10 mL CH3CN. The reaction solution was stirred
for 24 hours at room temperature, after which the solvent was evaporated
on a vacuum line to give a gum. The gum was dissolved in 50 mL of CH2Cl2,
washed with 15 mL saturated Na2CO3 solution, the organic layer was dried
with anhydrous MgSO4 and the solvent was evaporated on a vacuum line
to give a gum. This material was eluted through a pad of silica gel (20 g)
with ethyl acetate, and the UV-active material with Rf = 0.9 (EtOAc) was
collected and the solvent evaporated on a vacuum line to give 0.85 g (100%
yield) of product as a slightly yellow foam, which was used in the next
step.
The synthesis of (4) was then carried out in the same manner as described
for the synthesis of (3), starting from 0.793 g (1.03 mmol) of the above phos-
phoramidite. The crude compound was taken out of the glove box and
purified by column chromatography (∼28 g of silica gel in a 30 mm diame-
ter column) using a mixture of ethyl acetate/hexane (3:1). All fractions with
Rf = 0.38 (ethyl acetate/hexane 3:1) were evaporated under reduced pres-
1
sure to afford (4) as a white foam (0.60 g, 60% yield). H NMR [400 MHz,
CDCl3, two diastereomers; T1 and T2 assigned by comparison to (3)]: δ 8.34
(br s, NH, 4H), 7.59–7.24 (m, includes 4 × H-6, 22H), 6.85–6.83 (m, 8H),
6.41 (m, 2 × T1H-1ꢁ, 2H), 6.22 (m, 2 × T2H-1ꢁ, 2H), 4.97 (m, 2 × T1H-3ꢁ, 2H),
4.34 (m, 2 × T2H-3ꢁ, 2H), 4.18 (m, 2 × T1H-4ꢁ, 2H), 4.07–3.92 (m, 2 × T2H-4ꢁ,
2 × T2H-5ꢁ, 2 × T2H-5ꢁꢁ, 6H), 3.79 (s, 4 × CH3O, 12H), 3.42 (m, 2 × T1H-
5ꢁ, 2 × T1H-5ꢁꢁ, 4H), 2.54 (m, 2 × CH2CN, 4H), 2.55–2.31 (m, 2 × T1H-2ꢁ,
2 × T1H-2ꢁꢁ, 4H), 2.29–2.08 (m, 2 × T2H-2ꢁ, 2 × T2H-2ꢁꢁ, 4H), 1.89 (d, J =
1.0 Hz, CH3C-5, 3H), 1.88 (d, J = 1.0 Hz, CH3C-5, 3H), 1.510, 1.505, 1.49,
1.47 (4s, C(CH3)2OP,12H), 1.45 (d, J = 1.0 Hz, CH3C-5, 3H), 1.44 (d,
J = 1.0 Hz, CH3C-5, 3H), 0.89 (s, (CH3)3CSi, 9H), 0.88 (s, (CH3)3CSi, 9H),
0.075, 0.070 (2s, (CH3)2Si, 6H), 0.062, 0.058 (2s, (CH3)2Si, 6H). 31P NMR
(161 MHz, CDCl3, two diastereomers): δ 135.15 (SP), 134.75 (RP), 45.5:54.5;
assigned by conversion to (14).

Sulfurization of Dinucleoside Phosphite Triesters with Chiral Disulfides
719
Synthesis of 5ꢀ-O-(4,4ꢀ-Dimethoxytrityl)Thymidin-3ꢀ-yl
3ꢀ-O-(Tertbutyldimethylsilyl) Thymidin-5ꢀ-yl-(H-Phosphonate) (8)
Compound (8) was synthesized by following the literature proce-
dure used for the synthesis of the 5ꢁ-O-(tert-butyldimethylsilyl) ana-
logue of compound (8),[26] starting first with the preparation of [5ꢁ-
O-(4,4ꢁ-dimethoxytrityl)thymidine 3ꢁ-(H-phosphonate)]− (CH3CH2)3NH+:
under
a
nitrogen atmosphere,
a
solution of 2-chloro-4H-1,3,2-
benzodioxaphosphorin-4-one (1.0326 g, 5.0985 mmol) dissolved in 5 mL
of 1,4-dioxane was added dropwise over ∼3 minutes to a rapidly stirred so-
lution of 5ꢁ-O-(4,4ꢁ-dimethoxytrityl)thymidine[37] (2.0107 g, 3.6921 mmol)
and pyridine (0.8080 g, 10.21 mmol) in 40 mL of 1,4-dioxane. The flask
was rinsed with 5 mL of 1,4-dioxane and transferred to the reaction solu-
tion. The solution was stirred at room temperature for ∼20 minutes, then
0.5 mL of water was added, and the mixture was diluted with 60 mL of
CH2Cl2/Et3N (99/1 v/v). The solution was washed with 25 mL of 2 M TEAB
(vide infra). The organic layer was separated and the aqueous layer was back
extracted with 2 × 25 mL CH2Cl2. The combined organic layers were dried
with anhydrous MgSO4, filtered, and the solvent was removed on a rotary
evaporator to give an orange gum. The product was purified by silica gel
chromatography using a 0%–9% gradient of CH3OH in 0.5% Et3N/CH2Cl2
(50 g silica gel in a 50 mm diameter column). All fractions with Rf = 0.24
(10% MeOH/CH2Cl2) were combined and evaporated on a vacuum line
1
to afford the product as a white foam (2.01 g, 77% yield). H, 13C, and 31P
NMR spectra matched those taken on an authentic sample purchased from
ChemGene.
Under a nitrogen atmosphere, a sample of the above H-phosphonate
(0.8339 g, 1.175 mmol) and 3ꢁ-O-(tert-butyldimethylsilyl)thymidine[40]
(0.4114 g, 1.154 mmol) was co-evaporated with 12 mL of pyridine to re-
move water and then redissolved in 10 mL of pyridine. With stirring, pivaloyl
chloride (0.3290 g, 2.729 mmol) was added dropwise over ∼1 minute. The
reaction solution was stirred at room temperature for 1 hour, and then evap-
orated on a vacuum line to give a white foam. The foam was dissolved in
THF and then eluted with THF through a short column of silica gel (∼10 g
silica gel in a 60 mL coarse frit). The UV-active material that came through
with the solvent front was collected and evaporated on a vacuum line to give
the crude product as a white foam. The compound was purified by silica gel
chromatography using a 0%–6% gradient of CH3OH in CH2Cl2 (60 g silica
gel in a 50 mm diameter column). The appropriate fractions with Rf = 0.6
(10% MeOH/CH2Cl2) were combined, and the solvent was evaporated on a
vacuum line to afford H-phosphonate (8) as a white foam (0.76 g, 70%). 1H
and 31P NMR spectra matched those reported in the literature[27]
:
31P NMR
(161 MHz, DMSO-d6, two diastereomers): δ 9.44 (SP), 8.76 (RP), 44.4:55.6.

720
J. A. Mukhlall and W. H. Hersh
Synthesis of [5ꢀ-O-(4,4ꢀ-Dimethoxytrityl)Thymidin-3ꢀ-yl
3ꢀ-O-(Tertbutyldimethylsilyl)
Thymidin-5ꢀ-yl-Phosphorothioate]−[(CH3CH2)3NH]+ (14)
Under a nitrogen atmosphere, BSA (20 µL, 0.082 mmol) was added to
a solution of (8)(15 mg, 0.016 mmol) in 0.3 mL CH3CN. After stirring for
20 minutes, sulfur (6.1 mg, 0.19 mmol S or 0.024 mmol S8) was added. After
stirring the suspension (with the addition of 1 mL more CH3CN) for 2 hours,
2 M TEAB (vide infra) was added (50 µL, 0.10 mmol), and the volatiles were
removed using a vacuum pump. The white and brown solid was taken up
in chloroform and filtered, and then chromatographed on a 0.5 × 4 cm
silica column using 10% CH3OH in ethyl acetate with ∼0.5% Et3N added
and then with 1:1 CH3OH/ethyl acetate once the product began to elute.
Fractions with Rf ≈ 0.15 (4:1 ethyl acetate:CH3OH with ∼0.5% Et3N) were
1
combined. While the 31P NMR only exhibited peaks due to (14), the H
NMR exhibited impurities near 0 ppm in particular, so the chromatography
was repeated. The product was treated with 50 µL TEAB in CH3CN, and
the solvent was removed to give (14) as a white, crystalline solid (10.9 mg,
1
63% yield). The 31P and H NMR spectra matched those reported[31]
;
31P
NMR data follow for the chromatography sample, where some fractionation
occurred, so the ratios cannot be compared with other samples of (14), and
the chemical shifts were found to be sensitive to purity: (CD3CN) δ 56.555
(SPS), 56.519 (RPS), ∼61:39, HPLC 55.13:44.87 SPS:RPS; (CDCl3) δ 57.41
(RPS), 57.28 (SPS), 44.5:55.5.
Preparation of 2 M Aqueous TEAB
The following was adapted from a literature method using CO2 gas[26]: 56
mL of Et3N was added to 144 mL of deionized H2O, resulting in a two-layer
solution. With vigorous stirring, pieces of dry ice (CO2) were added to the
solution until a homogenous solution of pH ∼ 8 was attained, as determined
with Alkacid test paper. The slightly yellow solution obtained was stored at
room temperature.
Preparation of 0.1 M TEAA Buffer for HPLC Analysis
At room temperature, 13.9 mL of Et3N was added to a vigorously stirred
solution of 950 mL HPLC grade H2O and 5.6 mL of acetic acid. After ∼20
minutes, acetic acid was added dropwise to the solution until the pH was 7.0
as determined using an Accumet Basic pH meter (Fisher Scientific). The
solution was stored in a sealed bottle and kept at room temperature and
protected from ambient light.

Sulfurization of Dinucleoside Phosphite Triesters with Chiral Disulfides
721
Chiral Sulfurization Procedure
Typically up to eight reactions were run at once, in sets of four phos-
phite triesters and two enantiomers of a sulfurizing reagent, all in the inert
atmosphere glove box under nitrogen. For each reaction, a ∼0.05 M stock
solution of (3), (4), or (8) in CH3CN was prepared, and 100 µL of this solu-
tion [giving 0.0054–0.0060 mmol, ∼5 mg each of (3), (4), or (8)] was used
for each sulfurization. For (5), BSA was added to the solution of (8) directly
(7 µL ≈ 5.3 equivalents), while for (6) and (7), CH3CN solutions of TBDM-
SCl (100 µL of a 0.27 M solution, ∼5.1 equivalents) or Ph3SiCl (200 µL of a
0.14 M solution, ∼5.3 equivalents) were added along with Et3N (8 µL, ∼10.7
equivalents). Each solution was then diluted with either 100 or 200 µL of
CH3CN as appropriate to give ∼300 µL of solution (giving concentrations
ranging from ∼0.017 to 0.019 M in phosphite). After allowing the silylations
to proceed at room temperature for 2 hours, the reactions were cooled in
the glove box freezer at −32◦C for ∼20 minutes, and then the sulfurizing
reagents (0.25 equivalents) were added as CH3CN solutions [volumes ranged
from 16 to 65 µL for 0.02–0.09 M solutions, ∼0.0013–0.0015 mmol, ∼0.3–
0.8 mg of (1) or (2)]. The solutions were magnetically stirred in the glove
box freezer by placing a stirrer below the freezer, for 1 hour, and kept at
−32◦C for a total of 20 hours. The solutions were then quenched by the
addition of 10 µL of a ∼3.3 M solution of anhydrous TBHP in toluene[32] in
order to oxidize the unreacted phosphite.[26,33] For (4), deprotection with
DBU (3.5 equivalents) over 45 minutes followed by the addition of 20 µL
of 2 M TEAB gave the triethylammonium phosphorothioate salt (14). For
(5)–(7), hydrolysis of the siloxy sulfide and oxide to phosphorothioate (14)
and the phosphodiester was carried out by the addition of 10 µL of 2 M
TEAB.
Analysis of samples was carried out by HPLC by the injection of 3 µL of
solution into a Dionex Acclaim PolarAdvantage II C18 column (5 µm parti-
˚
cle size, 120 A pore size, 4.6 × 150 mm) for (9) and into a Dionex Acclaim
˚
120 C8 column (5 µm, 120 A, 4.6 × 250 mm) for (14). In all cases, reactions
were analyzed at 30◦C with a flow rate of 1.0 mL/minute and monitored at
254 nm, using 65:35 CH3CN:H2O for (9) and 55:45 or 52:48 CH3CN:TEAA
for (14). Elution times for (9) were 15.75 (RPS) and 16.79 (SPS) minutes,
and for the phosphate triester diastereomers, 7.87 and 8.16 minutes; elution
times and integrations gave virtually identical results for repeat injections.
Elution times for (14) were 16.12 0.09 (RPS) and 20.28 0.13 (SPS) min-
utes, and 9.86 minutes for the phosphate diester for 55:45 CH3CN:TEAA,
and 23.98 0.18 (RPS) and 31.08 0.28 (SPS) minutes, and 13.81 minutes
for the phosphate diester for 52:48 CH3CN:TEAA, and integration repro-
ducibility for these repeat injections was 0.2%; for repeated experiments
(a total of 31 for 14 experiments), integration reproducibility was 0.9%.
Peak homogeneity was assessed using a diode array detector capable of

722
J. A. Mukhlall and W. H. Hersh
monitoring the UV spectrum at 190–400 nm at any point on a selected
peak, and periodic checks showed the integrated peaks to have identical UV
spectra near the beginning, middle, and end of each peak.
Analysis of samples by 31P NMR was carried out by preparation as de-
scribed above for (5)–(7) and (9)–(13), but for (9)–(13), excess sulfur
[(MEDITH for (10)] was used; HPLC ratios for (10)–(13) are reported for
the hydrolysis product (14): (5): δ 127.76 (RP), 127.67 (SP), 55.9:44.1; (6): δ
127.87 (RP), 127.31 (SP), 56.3:43.8; (7): δ 128.34 (RP), 126.89 (SP), 55.7:44.3;
(9): δ 67.04 (RPS), 66.98 (SPS), 50.3:49.7, HPLC 50.59:49.41 RPS:SPS; (10): δ
58.93 (RPS), 58.58 (SPS), 45.5:54.5, HPLC 45.66:54.34 RPS:SPS; (11): δ 55.38
(RPS), 55.11 (SPS), 44.4:55.6, HPLC 43.44:56.56 RPS:SPS; (12): δ 55.55 (SPS),
55.49 (RPS), 56.8:43.2, HPLC 56.62:43.38 SPS:RPS; (13): δ 55.16 (SPS), 55.05
(RPS), 55.4:44.6, HPLC 56.64:43.36 SPS:RPS.
CONCLUSION
For screening the chiral analogues of PADS and MEDITH, in addition
to the dithymidine phosphite triesters prepared by the phosphoramidite
method, phosphite triesters with different degrees of steric hindrance
around the phosphorus were successfully prepared via the H-phosphonate
method, using BSA, TBDMSCl/Et3N, and Ph3SiCl/Et3N. Because silylation
occurred at the H–P=O group and on the thymine bases, it was necessary
to use ≥5 equivalents of the silylating reagents to give completely silylated
phosphite triesters. After sulfurization, analysis of the RPS:SPS diastereomeric
ratios of the resulting phosphite sulfides or phosphorothioates was success-
fully determined by reverse-phase HPLC; 31P NMR did not give sufficient
resolution of peaks or accuracy of peak integration to be useful.
A numerical procedure was developed to express the diastereoselectivity
(dsR(PS) and dsS(PS)) of the sulfurization reactions, and the de (de = dsS(PS)
– dsR(PS)) was used to compare selectivities of the reactions. The analogues
of MEDITH were found to be generally more selective than those of PADS,
despite the fact that in many cases, racemization of the MEDITH analogues
occurred. A possible reason for the poor selectivity of the PADS analogues
could be the known decomposition of PADS itself into a more active sulfu-
rizing reagent that might consist of sulfur oligomers, when it is “aged” in
3-picoline solution.[41] If sulfur oligomers were the primary active reagent,
then the chirality of the intact disulfides would have no effect. Fresh solutions
of PADS are in fact known to sulfurize phosphites in a 1:1 stoichiometry,[41]
so our use of fresh acetonitrile solutions of the PADS analogues and the
observation of any selectivity for the chiral analogues suggest that we are in
fact observing direct sulfurization by the intact disulfides.
Since the best actual selectivities were achieved with MEDITH analogues
(S)-(2d) (the naproxen derivative of MEDITH) and (S)-(2e) (an ibuprofen

Sulfurization of Dinucleoside Phosphite Triesters with Chiral Disulfides
723
derivative of MEDITH), and the best difference in the diastereomer ratio
was achieved with MEDITH analogue (S)-(2c), which contained an isopropyl
group, a direction for future work will be chiral (and achiral) disulfides with
larger groups adjacent to the disulfide. The approach of MEDITH toward
the phosphite triester will necessarily be different from that of PADS, and
these results suggest that it will be a better choice for selectivity. Development
of a practical method for syntheses of P-stereogenic phosphorothioates will
still require a means to epimerize the triester, but the use of the five triesters
in this study still provides a viable route to screening of new disulfides for
diastereoselectivity.
REFERENCES
1. See the following and references therein: Guzaev, A.P. Reactivity of 3H-1,2,4-dithiazole-3-thiones and
3H-1,2-dithiole-3-thiones as sulfurizing agents for oligonucleotide synthesis. Tetrahedron Lett. 2011,
52, 434–437.
2. Mukhlall, J.A.; Noll, B.C.; Hersh, W.H. Synthesis of chiral disulfides: potential reagents for enantio-
selective sulfurization. J. Sulfur Chem. 2011, 32, 199–212.
3. Stein, C.A.; Cheng, Y.-C. Antisense oligonucleotides as therapeutic agents—is the bullet really magi-
cal? Science 1993, 261, 1004–1012.
4. Rayburn, E.R.; Zhang, R. Antisense, RNAi, and gene silencing strategies for therapy: mission possible
or impossible? Drug Discov. Today 2008, 13, 513–521.
5. Stec, W.J.; Grajkowski, A.; Kobylanska, A.; Karwowski, B.; Koziolkiewicz, M.; Misiura, K.;
Okruszek, A.; Wilk, A.; Guga, P.; Boczkowska, M. Diastereomers of nucleoside 3ꢁ-O-(2-thio-1,3,2-
oxathia(selena)phospholanes): building blocks for stereocontrolled synthesis of oligo(nucleoside
phosphorothioate)s. J. Am. Chem. Soc. 1995, 117, 12019–12029.
6. Stec, W.J.; Karwowski, B.; Boczkowska, M.; Guga, P.; Koziolkiewicz, M.; Sochacki, M.; Wieczorek,
M.W.; Blaszczyk, J. Deoxyribonucleoside 3ꢁ-O-(2-thio- and 2-oxo- “spiro” -4,4-pentamethylene-1,3,2-
oxathiaphospholane)s: monomers for stereocontrolled synthesis of oligo(deoxyribonucleoside phos-
phorothioate)s and chimeric PS/PO oligonucleotides. J. Am. Chem. Soc. 1998, 120, 7156–7167.
7. Wang, J.-C.; Just, G. Indolyl-oxazaphosphorine precursors for stereoselective synthesis of phosphite
triesters and dithymidinyl phosphorothioates. J. Org. Chem. 1999, 64, 8090–8097.
8. Wilk, A.; Grajkowski, A.; Phillips, L.R.; Beaucage, S.L. Deoxyribonucleoside cyclic N -
acylphosphoramidites as
a new class of monomers for the stereocontrolled synthesis of
oligothymidylyl- and oligodeoxycytidylyl-phosphorothioates. J. Am. Chem. Soc. 2000, 122, 2149–2156.
9. Yu, D.; Kandimalla, E.R.; Roskey, A.; Zhao, Q.; Chen, L.; Chen, J.; Agrawal, S. Stereo-enriched
phosphorothioate oligodeoxynucleotides: synthesis, biophysical and biological properties. Bioorg.
Med. Chem. 2000, 8, 275–284.
10. Oka, N.; Yamamoto, M.; Sato, T.; Wada, T. Solid-phase synthesis of stereoregular oligodeoxyribonu-
cleoside phosphorothioates using bicyclic oxazaphospholidine derivatives as monomer units. J. Am.
Chem. Soc. 2008, 130, 16031–16037.
11. Almer, H.; Szabo, T.; Stawinski, J. A new approach to stereospecific synthesis of P-chiral phosphoro-
thioates. Preparation of diastereomeric dithymidyl-(3ꢁ-5ꢁ) phosphorothioates. Chem. Commun. 2004,
290–291.
12. Perlikowska, W.; Gouygou, M.; Mikolajczyk, M.; Daran, J.-C. Enantiomerically pure disulfides: key
compounds in the kinetic resolution of chiral PIII-derivatives with stereogenic phosphorus. Tetrahedron
Asymmetry 2004, 15, 3519–3529.
13. Eliel, E.L.; Wilen, S.H.; Mander, L.N. Stereochemistry of Organic Compounds, Wiley-Interscience, New
York, NY, 1994, p. 965.
14. Kamer, P.C.J.; Roelen, H.C.P.F.; van den Elst, H.; van der Marel, G.A.; van Boom, J.H. An efficient
approach toward the synthesis of phosphorothioate diesters via the Scho¨nberg reaction. Tetrahedron
Lett. 1989, 30, 6757–6760.

724
J. A. Mukhlall and W. H. Hersh
15. Cheruvallath, Z.S.; Carty, R.L.; Moore, M.N.; Capaldi, D.C.; Krotz, A.H.; Wheeler, P.D.; Turney, B.J.;
Craig, S.R.; Gaus, H.J.; Scozzari, A.N.; Cole, D.L.; Ravikumar, V.T. Synthesis of antisense oligonu-
cleotides: replacement of 3H-1,2-benzodithiol-3-one 1,1-dioxide (Beaucage reagent) with pheny-
lacetyl disulfide (PADS) as efficient sulfurization reagent: from bench to bulk manufacture of active
pharmaceutical ingredient. Org. Process Res. Dev. 2000, 4, 199–204.
16. Zhang, Z.; Nichols, A.; Tang, J.X.; Han, Y.; Tang, J.Y. Solid phase synthesis of oligonucleotide phos-
phorothioate analogues using 3-methyl-1,2,4-dithiazolin-5-one (MEDITH) as a new sulfur-transfer
reagent. Tetrahedron Lett. 1999, 40, 2095–2098.
17. Powell, W.H. Revision of the extended Hantzsch-Widman system of nomenclature for heteromono-
cycles. Pure Appl. Chem. 1983, 55, 409–416.
18. Mukhlall, J.A.; Hersh, W.H. Measurement of the barrier to inversion of configuration in acyclic
phosphite triesters. Inorg. Chim. Acta 2011, 369, 62–70.
19. Marugg, J.E.; Nielsen, J.; Dahl, O.; Burik, A.; Van der Marel, G.A.; Van Boom, J.H. (2-cyano-1,1-
dimethylethoxy)bis(diethylamino)phosphine: a convenient reagent for the synthesis of DNA frag-
ments. Recl. Trav. Chim. Pays-Bas 1987, 106, 72–76.
20. Nielsen, J.; Caruthers, M.H. Directed Arbuzov-type reactions of 2-cyano-1,1-dimethylethyl deoxynu-
cleoside phosphites. J. Am. Chem. Soc. 1988, 110, 6275–6276.
21. Ando, K.; Kawai, M.; Masuda, T.; Omura, H. Benzimidazolone Derivatives as CB2 Receptor Ligands
and Their Preparation, Pharmaceutical Composition, and Use for Treatment of Various Types of Pain,
WO/2006097808, Pfizer Japan, Inc., Tokyo, 2006.
22. Kume, A.; Fujii, M.; Sekine, M.; Hata, T. Acylphosphonates. 4. Synthesis of dithymidine phosphonate:
a new method for generation of phosphonate function via aroylphosphonate intermediates. J. Org.
Chem. 1984, 49, 2139–2143.
23. Sergueev, D.S.; Shaw, B.R. H-phosphonate approach for solid-phase synthesis of oligodeoxyribonu-
cleoside boranophosphates and their characterization. J. Am. Chem. Soc. 1998, 120, 9417–9427.
24. Sergueeva, Z.A.; Sergueev, D.S.; Shaw, B.R. Synthesis of dithymidine boranophosphates via stere-
ospecific boronation of H-phosphonate diesters and assignment of their configuration. Tetrahedron
Lett. 1999, 40, 2041–2044.
25. Fujii, M.; Ozaki, K.; Sekine, M.; Hata, T. Acylphosphonates. 7. A new method for stereospecific and
stereoselective generation of dideoxyribonucleoside phosphorothioates via the acylphosphonate
intermediates. Tetrahedron 1987, 43, 3395–3407.
26. deVroom, E.; Dreef, C.E.; van den Elst, H.; van der Marel, G.A.; van Boom, J.H. Preparation of
internucleotide phosphate analogues via the corresponding H-phosphonate diester. Recl. Trav. Chim.
Pays-Bas 1988, 107, 592–595.
27. Seela, F.; Kretschmer, U. Diastereomerically pure RP and SP dinucleoside H-phosphonates—the
stereochemical course of their conversion into P-methylphosphonates, phosphorothioates, and [18O]
chiral phosphates. J. Org. Chem. 1991, 56, 3861–3869.
28. Stec, W.J.; Zon, G.; Uznan`ski, B. Reversed-phase high-performance liquid chromatographic sep-
aration of diastereomeric phosphorothioate analogues of oligodeoxyribonucleotides and other
backbone-modified congeners of DNA. J. Chromatogr. 1985, 326, 263–280.
29. Marugg, J.E.; Dreef, C.E.; Van der Marel, G.A.; Van Boom, J.H. Use of 2-cyano-1,1-dimethylethyl as
a protecting group in the synthesis of DNA via phosphite intermediates. Recl. Trav. Chim. Pays-Bas
1984, 103, 97–98.
30. Sekine, M.; Tsuruoka, H.; Iimura, S.; Kusuoku, H.; Wada, T.; Furusawa, K. Studies on steric and
electronic control of 2ꢁ-3ꢁ phosphoryl migration in 2ꢁ-phosphorylated uridine derivatives and its
application to the synthesis of 2ꢁ-phosphorylated oligouridylates. J. Org. Chem. 1996, 61, 4087–
4100.
31. Bartoszewicz, A.; Kalek, M.; Stawinski, J. The case for the intermediacy of monomeric metaphos-
phate analogues during oxidation of H-phosphonothioate, H-phosphonodithioate, and H-
phosphonoselenoate monoesters: mechanistic and synthetic studies. J. Org. Chem. 2008, 73,
5029–5038.
32. Hill, J.G.; Rossiter, B.E.; Sharpless, K.B. Anhydrous tert-butyl hydroperoxide in toluene: the preferred
reagent for applications requiring dry TBHP. J. Org. Chem. 1983, 48, 3607–3608.
33. Hayakawa, Y.; Uchiyama, M.; Noyori, R. Nonaqueous oxidation of nucleoside phosphites to the
phosphates. Tetrahedron Lett. 1986, 27, 4191–4194.

Sulfurization of Dinucleoside Phosphite Triesters with Chiral Disulfides
725
34. Mukhlall, J.A. Chiral sulfurization for synthesis of antisense oligonucleotides, Ph.D. thesis, City
University of New York, New York, 2010.
35. Eliel, E.L.; Wilen, S.H.; Mander, L.N. Stereochemistry of Organic Compounds, Wiley-Interscience, New
York, NY, 1994, p. 860.
36. Gawley, R.E. Do the terms “% ee” and “% de” make sense as expressions of stereoisomer composition
or stereoselectivity? J. Org. Chem. 2006, 71, 2411–2416.
37. Schaller, H.; Weimann, G.; Lerch, B.; Khorana, H.G. Studies on polynucleotides. XXIV. The stepwise
synthesis of specific deoxyribopolynucleotides (4). Protected derivatives of deoxyribonucleosides and
new syntheses of deoxyribonucleoside-3ꢁ phosphates. J. Am. Chem. Soc. 1963, 85, 3821–3827.
38. Nielsen, J.; Marugg, J.E.; van Boom, J.H.; Honnens, J.; Taagaard, M.; Dahl, O. Thermal instability of
some alkyl phosphorodiamidites. J. Chem. Res. (S) 1986, 26–27.
39. Xie, C.; Staszak, M.A.; Quatroche, J.T.; Sturgill, C.D.; Khau, V.V.; Martinelli, M.J. Nucleosidic phos-
phoramidite synthesis via phosphitylation: activator selection and process development. Org. Process
Res. Dev. 2005, 9, 730–737.
40. Ogilvie, K.K.; Thompson, E.A.; Quilliam, M.A.; Westmore, J.B. Selective protection of hydroxyl groups
in deoxynucleosides using alkylsilyl reagents. Tetrahedron Lett. 1974, 15(33), 2865–2868.
41. Krotz, A.H.; Gorman, D.; Mataruse, P.; Foster, C.; Godbout, J.D.; Coffin, C.C.; Scozzari, A.N. Phos-
phorothioate oligonucleotides with low phosphate diester content: greater than 99.9% sulfurization
efficiency with “aged” solutions of phenylacetyl disulfide (PADS). Org. Process Res. Dev. 2004, 8,
852–858.