Lithium bromide-DBU mediated synthesis of chlorophosphoramidate-activated morpholino nucleoside subunits

Article abstract of DOI:10.1016/j.tetlet.2012.09.126

An efficient and rapid protocol for the synthesis of chlorophosphoramidate- activated morpholino subunits has been developed. A unique combination of lithium bromide-DBU has been found to activate the 7′-hydroxyl toward the reaction. All the four morpholi

Full text of DOI:10.1016/j.tetlet.2012.09.126

Tetrahedron Letters 53 (2012) 6714–6717
Contents lists available at SciVerse ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Lithium bromide–DBU mediated synthesis of
chlorophosphoramidate-activated morpholino nucleoside subunits
⇑
Sankha Pattanayak, Surajit Sinha
Department of Organic Chemistry, Indian Association for the Cultivation of Science, Jadavpur, Kolkata 700 032, India
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 25 July 2012
Revised 25 September 2012
Accepted 27 September 2012
Available online 3 October 2012
An efficient and rapid protocol for the synthesis of chlorophosphoramidate-activated morpholino sub-
units has been developed. A unique combination of lithium bromide–DBU has been found to activate
the 70-hydroxyl toward the reaction. All the four morpholino units of adenosine, guanosine, cytosine,
and thymine responded smoothly to the developed protocol.
Ó 2012 Elsevier Ltd. All rights reserved.
Keywords:
Morpholino subunit
Chlorophosphoramidate
Nucleosides
Lithium bromide
DBU
Advancement of molecular genetics in recent years has recog-
nized antisense compounds as potential therapeutic agents and
powerful reagents for the exploration of gene functions.1,2 Devel-
oped by Summerton, phosphorodiamidate morpholino oligonucle-
otides (PMOs) are one of the highly favorable third-generation
antisense compounds, owing to its predictable in vivo activity in
blocking mRNA translation.3–9
starting materials remains unreacted, and with LiHMDS, longer
reaction time often leads to the formation of the phosphoric acid
as the major product. Notably, synthesis of the chlorophosphoram-
idate intermediates for medicinally-important prodrugs also re-
quires strong bases like t-BuOK14 and LiHMDS.15 These extremely
moisture-sensitive strong bases often need hazardous preparation
steps and additional precautionary measures. Because only a lim-
ited number of methods have been developed for synthesizing
these base-sensitive chlorophosphoramidate-activated morpho-
lino subunits, there is a further need for new mild and efficient
synthetic procedures. In connection with our studies on modified
morpholinos,16,17 herein, we report lithium bromide-diazabicyclo-
Morpholinos are synthetic DNA analogues (ꢀ25 bases long)
having morpholine ring in lieu of the ribose sugar moiety and con-
tain neutral phosphorodiamidate backbone.6 PMOs are prepared
on a solid support using an iterative two-step process, which com-
prises detritylation and subsequent assembling of the chloro-
phosphoramidate-activated subunits (Fig. 1). These activated
subunits were initially synthesized by reacting the corresponding
70-hydroxy morpholino units of adenosine (A), 5-methyl uridine
(T), guanosine (G), and cytidine (C) with N,N-dimethylphosphor-
amidodichloride in the presence of N-ethylmorpholine as a base
and N-methylimidazole as an additive.10 The conversion was later
improved by changing the base to 2,6-lutidine, and around 50%
yields were reported in the case of A, C, and T.11 However, longer
reaction time (6–16 h) often poses a problem as decomposition
of the product starts, and for G-morpholino, double base-protec-
tion is required to circumvent the very poor (5%) yield.12 An alter-
nate synthetic procedure was reported by Chen and co-workers
using a strong base such as LiHMDS which gave moderate to good
yields.13 In both of the above cases, a considerable amount of the
Cl
P O
N
BasePg
O
7'
N
BasePg
O
O
6'
5'
O
2'
3'
N
CPh3
CPh3
activated subunit
detritylation
O
BasePg
O
BasePg
O
O
activated subunit
coupling
N
N
BasePg
O P O
N
H
O
N
oligo elongation and
final removal from resin
CPh3
⇑
Corresponding author. Tel.: +91 33 2473 4971; fax: +91 33 2473 2805.
E-mail address: ocss5@iacs.res.in (S. Sinha).
Figure 1. General structure and oligomerization of activated subunits.
0040-4039/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tetlet.2012.09.126

S. Pattanayak, S. Sinha / Tetrahedron Letters 53 (2012) 6714–6717
6715
undecene (DBU) as an effective combination to synthesize the acti-
vated subunits.
morpholino oligomerization.18 However, the effect of lithium salts
on the synthesis of chlorophosphoramidate-activated subunits has
never been studied. The high electropositive character and small
size of lithium cation make it a preferred additive in numerous
organic reactions. The LiBr–DBU combination is also known to
promote many organic transformations, for example, transesterifi-
cation,19 Michael addition,20 Baylis–Hillman reaction,21 cycloaddi-
tion reactions22,23 etc. In our model reaction, stoichiometric
amounts of LiBr in combination with 2.2 equiv of NMM, or 2,6-lu-
tidine or DBU in acetonitrile resulted in unsatisfactory conversions
(Table 1, entries 13–15). Among the non-nucleophilic bases
screened, DBU has the highest pKa value24 in non-aqueous solvents
and seemed promising from the initial results. To our delight,
when the amount of DBU and LiBr was increased to 3.2 equiv, at
room temperature, the reaction was completed within 20 min
and 46% yield was achieved (Table 1, entry 17). To minimize the
suspected decomposition of the product, the reaction was repeated
at 0 °C and better yield was obtained (Table 1, entry 18). Use of THF
as solvent was also found to yield comparative result (Table 1, en-
try 19). The nucleoside was soluble in DCM, whereas LiBr was sol-
uble in CH3CN. Thus a 50:50 mixture of DCM and CH3CN was used
as solvent and a satisfactory 61% yield for 1a was obtained (Table 1,
entry 20; best condition). To find out a better additive, other lith-
ium halides were also screened. Interestingly, none of LiF, LiCl,
and LiI exhibited satisfactory results (Table 1, entries 21–23). Lith-
The 70-hydroxy morpholino units of A, T, G, and C were first syn-
thesized following our established protocol.17 To find an efficient
way to carry out the transformation, we screened a number of con-
ditions using morpholino-T unit 1 as a model substrate (Table 1).
In the absence of any additive only a trace conversion was ob-
served when DIPEA, DBU, and DABCO (Table 1, entries 1–3) were
used as bases. Use of tetrabutyl ammonium hydroxide (n-Bu4NOH)
failed to effect any conversion (Table 1, entry 4). A powerful com-
bination of potassium carbonate-18-crown-6 was also proven to be
ineffective (Table 1, entry 5). When sodium hydride was employed,
at room temperature the products got decomposed (Table 1, entry
6). Careful optimization of the reaction temperature led to a pleas-
ing 50% yield of the activated morpholino-T unit (Table 1, entry 7)
but the purine morpholinos did not respond well to the NaH con-
dition. Expectedly, another strong base t-BuOK was quite effective
to promote the conversion (Table 1, entry 8), but was avoided be-
cause of its marked sensitivity to moisture to generate KOH which
hydrolyzed the product partially to give phosphoric acid. While DI-
PEA and DBU were ineffective alone, in combination with N-methyl
imidazole (30 mol %) as an additive, 19% and 26% yields of the
products were obtained, respectively (Table 1, entries 9 and 10).
Longer reaction time did not help the reactions to go to comple-
tion. In the presence of a catalytic amount of DMAP, DBU yielded
31% product (Table 1, entry 11) whereas N-methyl morpholine
(NMM) failed to promote any transformation (Table 1, entry 12).
Recently, the use of inorganic salts has been demonstrated to
accelerate the rate of phosphorodiamidate bond formation during
ium perchlorate (LiClO4) afforded comparable yields, with
a
slightly lower reaction rate (Table 1, entry 24). Bromides of other
representative alkali and alkaline earth metals could not promote
any conversion (Table 1, entries 25–27). The unique behavior of
Table 1
Screening of reaction conditionsa for the preparation of activated subunit 1a
O
P
N
HO
Thy
O
P
N
Cl
O
Base
Thy
O
O
Cl
Cl
Additive
solvent
N
N
CPh3
1
1a
CPh3
Entry
Base (equiv)
Additive (equiv)
Solvent
Temp (°C)
Time
Yieldb (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
DIPEA (2.2)
DBU (2.2)
None
None
None
None
18-Crown-6 (0.1)
None
DCM
DCM
DCM
THF
THF
THF
THF
THF
DCM
DCM
DCM
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
rt
rt
rt
rt
rt
rt/0
ꢁ10
0
rt
rt
rt
rt
rt
rt
rt
rt
rt
0
0
0
0
0
0
0
0
0
12 h
12 h
12 h
12 h
Trace
Trace
Trace
Trace
Trace
Decomp
50
63
19
26
31
Trace
Trace
11
16
21
46
55
47
61
DABCO (2.2)
n-Bu4NOH (2.2)
K2CO3 (2.2)
NaH (2.2)c
NaH (2.5)d
t-BuOK (2.2)
DIPEA (2.2)
DBU (2.5)
12 h
30 min
3.5 h
1 h
3 h
4 h
4 h
12 h
12 h
12 h
None
None
NMI (0.3)
NMI (0.3)
DMAP (0.1)
DMAP (0.1)
LiBr (1.1)
LiBr (1.1)
LiBr (1.1)
LiBr (2.2)
LiBr (3.2)
LiBr (3.2)
LiBr (3.2)
LiBr (3.2)
LiF (3.2)
LiCI (3.2)
LiI (3.2)
LiCIO4 (3.2)
NaBr (3.2)
KBr (3.2)
MgBr2 (3.2)
DBU (2.5)
NMM (2.2)
NMM (2.2)
2,6-Lutidine (2.2)
DBU (2.2)
DBU (3.2)
DBU (3.2)
12 h
12 h
CH3CN
CH3CN
THF
20 min
30 min
30 min
30 min
30 min
30 min
30 min
30 min
30 min
30 min
30 min
DBU (3.2)
DBU (3.2)
DBU (3.2)e
DBU (3.2)
CH3CN–DCM
CH3CN–DCM
CH3CN–DCM
CH3CN–DCM
CH3CN–DCM
CH3CN–DCM
CH3CN–DCM
CH3CN–DCM
6
5
11
54
Trace
Trace
Trace
DBU (3.2)
DBU (3.2)
DBU (3.2)
DBU (3.2)
DBU (3.2)
DBU (3.2)
0
a
b
c
0.05 mmol scale.
Isolated yields.
NMe2POCl2 added at 0 °C.
NMe2POCl2 added at ꢁ30 °C.
Best conditions.
d
e

6716
S. Pattanayak, S. Sinha / Tetrahedron Letters 53 (2012) 6714–6717
LiBr can be explained by assuming an optimum balance between
lattice energy and covalent character, so that it dissociates effi-
ciently while maintaining the coordination property of Li+ in or-
ganic solvents. Effectiveness of LiClO4 further proves the
assumption as it also has a loosely bound Li+ and low lattice
energy.
OH
BasePg
ring and
OH
N - inversion
O
N
Ph3C
BasePg
Ph3C
N
O
Cl
P
Cl
N
N
Having found the optimal conditions,25 the methodology was
then extended to the other 70-hydroxy morpholino subunits in a
similar manner (Table 2). The exocyclic amine protected cyti-
dine-morpholino subunit 2 was converted into the chlorophosph-
oramidate activated subunit 2a26 in 74% yield (Table 2, entry 2).
The 70-activated subunits of adenosine 3a and guanosine 4a were
also obtained in high yield within a short period of time (Table 2,
entries 3 and 4). It is worth mentioning here that such a clean
and high yielding method has not been reported earlier. After hav-
ing achieved a satisfactory way to synthesize regular activated
morpholino units, we were interested to see how efficiently this
methodology can be applied to other morpholino and nucleoside
derivatives. A novel morpholino derivative 5,17b integrating an
amino acid segment in the morpholine ring responded well to
furnish the corresponding product in 51% yield (entry 5). Interest-
ingly, when we proceeded to apply the method on acetonide-
protected 5-methyl uridine 6, a large amount of starting material
remained unreacted, and with 5 equiv of LiBr, only 21% yields of
6a were obtained. The reaction also remained unsuccessful on
20,30-dideoxythymidine.
N
O
H
O
LiBr
LiBr
Cl
Li
N
BasePg Li
O
DBU
Li
Ph3C
N
P
O
Cl
O
Cl
P
N
O
O
Ph3C
BasePg
N
Scheme 1. Plausible mechanism for the LiBr–DBU mediated activation.
The exact mechanism of the reaction needs to be further
investigated; however, a plausible mechanism (Scheme 1) can be
suggested from the clues obtained. In the examples of 6 and
20,30-dideoxythymidine the reactions were unsatisfactory. Notably,
in these, there are no ring nitrogen atoms like morpholinos which
clearly indicate the role of morpholine ring-N in coordination with
Li+. When the ring-N was protected with electron withdrawing
benzoyl group, only 9% yield of the activated unit was obtained
which further proves the involvement of the ring-N. The role of
Li+ was unambiguously established by repeating the model reac-
tion of 1 in the presence of lithium-coordinating solvent like
HMPA, as only a marginal conversion was observed. It seems that
3.2 equiv of the lithium salt is required to get the optimum results.
This can be explained as Li+ also interacts with the amide function-
ality of acylated nucleobases and NMe2POCl2. In the presence of
lithium cation, evidence for amide–imidic acid tautomerism was
found in FTIR spectrum (suppression of peak at 1690 cmꢁ1 and
emergence of sharp peak at 3566 cmꢁ1), when the model reaction
of 1 was performed without the addition of NMe2POCl2. In another
test reaction of NMe2POCl2, DBU, and LiBr, without the addition of
1, the 31P NMR signal shifts to downfield (from d 19.85 to 20.00),
possibly due to the coordination of Li+ with NMe2POCl2. Thus
mechanistically (Scheme 1), on intramolecular chelation with 70-
OH and morpholine ring-N,27 the lithium cation forms a six-mem-
bered ring through ring and nitrogen inversion. This chelation in
turn increases the acidity of the 70-OH proton, making it easier
for the base to abstract it. The alkoxide thus generated then reacts
with dimethylphosphoramidodichloride to form the desired chlo-
rophosphoramidate-activated subunit. The Li+ chelation step is be-
lieved to be the rate determining step as a thermodynamically
unfavorable inversion28,29 equilibrium proceeds to the forward
direction slowly.
Table 2
Scope of the LiBr–DBU mediated activated subunit synthesis
Entry Starting material
Product
Time
(min)
Yielda
(%)
HO
O
P
Thy
O
Cl
O
Thy
O
N
CPh3
N
1
30
25
30
20
61
74
69
56
1
N
CPh3
1a
O
CytBz
HO
HO
O
CytBz
Cl P O
N
O
N
CPh3
2
3
4
2
N
CPh3
2a
AdeBz
O
P
N
O
AdeBz
Cl
Cl
O
O
O
N
CPh3
3
N
CPh3
3a
Guai-Bu
HO
O
P
N
O
Guai-Bu
O
N
CPh3
4
N
CPh3
4a
HO
O
Thy
O
N
It is noteworthy to mention here that, the chlorophosphorami-
date products are not stable in solution at room temperature for
longer time, especially, in the presence of any base. In all the prep-
aration methods (including previously reported) for activated sub-
units, the isolated yields were often less than the conversion
observed (TLC). It is therefore advisable to purify the products
immediately after the reaction to obtain good yields.
In conclusion, we have developed a mild and high yielding
method for the preparation of chlorophosphoramidate-activated
morpholino subunits. An unprecedented application of the LiBr–
DBU reagent combination has also been uncovered. Owing to the
utilities of chlorophosphoramidate intermediates in the synthesis
Cl P O
N
Thy
O
5
6
40
60
51
21
Ph
O
N
Ph
O
O
5
5a
O
HO
O
Cl P O
N
Thy
O
Thy
O
O
O
6
O O
6a
a
Isolated yields.

S. Pattanayak, S. Sinha / Tetrahedron Letters 53 (2012) 6714–6717
6717
16. Paul, S.; Nandi, B.; Pattanayak, S.; Sinha, S. Tetrahedron Lett. 2012, 53, 4179–4183.
17. (a) Pattanayak, S.; Paul, S.; Nandi, B.; Sinha, S. Nucleos. Nucleot. Nucl. 2012, in
press.; (b) Sinha, S.; Pattanayak, S.; Paul, S.; Nandi, B. WO2011018798A2; Inter.
Pat: Appl, 2011.
of morpholino oligomers and medicinally important prodrugs, the
present method may find potential applications.
18. Harakawa, T.; Tsunoda, H.; Ohkubo, A.; Seio, K.; Sekine, M. Bioorg. Med. Chem.
Lett. 2012, 22, 1445–1447.
Acknowledgments
19. Seebach, D.; Thaler, A.; Blaser, D.; Ko, S. Y. Helv. Chim. Acta 1991, 74, 1102–
1118.
S.S. thanks the Department of Biotechnology, Government of
India, for financial support by a Grant no [BT/PR13426/BRB/10/
761/2009]. S.P. is grateful to CSIR for his fellowship.
20. Kanemasa, S.; Uchida, O.; Wada, E. J. Org. Chem. 1990, 55, 4411–4417.
21. Mamaghani, M.; Radmogadam, K.; Badrian, A. Asian J. Chem. 2006, 18, 840–844.
22. Dondas, H. A.; Grigg, R.; Kilner, C. Tetrahedron 2003, 59, 8481–8487.
23. Hanessian, S.; Bayrakdarian, M. Tetrahedron Lett. 2002, 43, 9441–9444.
24. Kaljurand, I.; Kütt, A.; Sooväli, L.; Rodima, T.; Mäemets, V.; Leito, I.; Koppel, I. A.
J. Org. Chem. 2005, 70, 1019–1028.
Supplementary data
25. Typical procedure for the synthesis of chlorophosphoramidate activated
morpholino subunits. To a solution of the nucleoside (0.5 mmol) in dry DCM–
CH3CN (6 ml, 50:50) was added LiBr (1.6 mmol) at 0 °C under Ar atmosphere
and stirred for 2 min. Then DBU (1.6 mmol) in dry CH3CN (1 mL) was added
dropwise at 0 °C to form a white precipitation. After stirring for another 2 min,
dimethylphosphoramidodichloride (0.8 mmol) in CH3CN (1 mL) was added
dropwise and the solution became almost clear within 15 min. The stirring was
continued for the required time, the mixture was filtered through celite and
the solvent was reduced under vacuum. The residue was immediately purified
by silica gel flash column chromatography (230–400 mess, 7–12% acetone–
DCM) to afford the activated subunit.
Supplementary data (characterization data for 1a–6a and copies
of spectra for the new compound 5a) associated with this article
can be found, in the online version, at http://dx.doi.org/10.1016/
j.tetlet.2012.09.126.
References and notes
1. Kramer, R.; Cohen, D. Nat. Rev. Drug Disc. 2004, 3, 965–972.
2. Chan, J. H. P.; Lim, S.; Wong, W. Clin. Exp. Pharmacol. Physiol. 2006, 33, 533–540.
3. Entire issue of Genesis 2001, 30, July 30.
26. Characterization data for compound 2a. White solid, mixture of diastereomers.
IR (KBr): mmax 1666, 1485 cmꢁ1 1H NMR (500 MHz, CDCl3) d 1.28 (t, 1H,
.
4. Eisen, J. S.; Smith, J. C. Development 2008, 135, 1735–1743.
5. Heasman, J. Dev. Biol. 2002, 243, 209–214.
J = 11.5 Hz), 1.53 (2d, 1H, J = 11.0 Hz), 2.62 (s, 3H), 2.65 (s, 3H), 3.16 (d, 1H,
J = 12.0 Hz), 3.59 (d, 1H, J = 11.5 Hz), 4.10–4.18 (m, 2H), 4.43–4.44 (m, 1H), 6.27
(d, 1H, J = 9.0 Hz), 7.18 (t, 3H, J = 7.0 Hz), 7.29 (t, 6H, J = 7.5 Hz), 7.36–7.53 (m,
9H), 7.57 (t, 1H, J = 7.5 Hz), 7.72 (t, 1H, J = 8.3 Hz), 7.88 (d, 2H, J = 7.5 Hz), 8.90
(br s, 1H); 13C NMR (125 MHz, CDCl3) d 36.7, 48.9, 52.7, 67.3, 74.8, 77.2, 82.2,
97.0, 126.7, 127.2, 127.5, 127.8, 128.0, 128.3, 128.6, 129.0, 129.3, 133.1, 133.2,
144.5, 147.1, 162.4; HRMS (ESI) m/z [M+Na]+, Calculated for C37H37ClN5O5PNa
720.2119. Found 720.2119.
6. Summerton, J.; Weller, D. Antisense Nucleic Acid Drug Dev. 1997, 7, 187–195.
7. Corey, D. R.; Abrams, J. M. Genome Biol. 2001, 2, 1015.1–1015.3.
8. Summerton, J. Biochim. Biophys. Acta 1999, 1489, 141–158.
9. Du, L.; Gatti, R. A. J. Immunol. Methods 2011, 365, 1–7.
10. Summerton, J.; Weller, D. U.S. Patent 5,185,444, 1993.
11. Weller, D.; Hassinger J. N. U.S. Patent 20090088562A1, 2009.
12. Reeves, M. D.; Weller, D. U.S. Patent 20090131624A1, 2009.
13. Shestopalov, I. A.; Sinha, S.; Chen, J. K. Nat. Chem. Biol. 2007, 3, 650–651.
14. Wang, P.; Mondal, M.; Wang, Y. Eur. J. Org. Chem. 2009, 2055–2058.
15. Meyers, C. L. F.; Hong, L.; Joswig, C.; Borch, R. F. J. Med. Chem. 2000, 43, 4313–
4318.
27. Tatsukawa, A.; Kawatake, K.; Kanemasa, S.; Rudzinski, J. M. J. Chem. Soc., Perkin
Trans. 2 1994, 2525–2530.
28. Lunazzi, L.; Casarini, D.; Cremonini, M. A. Tetrahedron 1991, 47, 7465–7470.
29. Jones, A. J.; Beeman, C. P.; Hasan, M. U.; Casy, A. F.; Hassan, M. A. A. Can. J. Chem.
1976, 54, 126–135.