Scalable synthesis of substituted 2,7-dimethyl-9-phenylxanthen-9-ol (DMPx-OH): Useful for the preparation of crystalline 5′-O-DMPx-protected nucleosides

Article abstract of DOI:10.1016/j.tetlet.2012.06.081

A convenient single-step synthesis of several 2,7-dimethyl-9-phenylxanthen- 9-ol (DMPx-OH) analogs has been accomplished using a Friedel-Crafts reaction. Treatment of various DMPx-OH with unprotected 2′-O-methoxyethyl- ribonucleosides (MOE) in the presence of B(C₆F₅) _3, as a Lewis Acid catalyst, furnished 5′-O-protected derivatives of 2′-MOE-ribonucleosides in good yields. The deprotection of the DMPx groups was accomplished by acid hydrolysis under very mild conditions. Among the five DMPx analogs synthesized, the 2,7-dimethyl-9-(4-nitrophenyl) xanthene-9-yl group furnished crystalline products enabling non-chromatographic isolation of 5′-O-protetced nucleosides.

Full text of DOI:10.1016/j.tetlet.2012.06.081

Tetrahedron Letters 53 (2012) 4669–4672
Contents lists available at SciVerse ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Scalable synthesis of substituted 2,7-dimethyl-9-phenylxanthen-9-ol
(DMPx-OH): useful for the preparation of crystalline 50-O-DMPx-protected
nucleosides
Shyamapada Banerjee a, P. Srishylam a, S. Rajendra Prasad a, Michael T. Migawa b, Eric E. Swayze b,
Yogesh S. Sanghvi c,
⇑
a Sapala Organics Private Ltd, Plot No. 147, Phase II, IDA Mallapur, Hyderabad 500076, India
b Isis Pharmaceuticals, 2855 Gazelle Court, Carlsbad, CA 92010, USA
c Rasayan Inc., 2802 Crystal Ridge Road, Encinitas, CA 92024, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A convenient single-step synthesis of several 2,7-dimethyl-9-phenylxanthen-9-ol (DMPx-OH) analogs has
been accomplished using a Friedel–Crafts reaction. Treatment of various DMPx-OH with unprotected 20-O-
methoxyethyl-ribonucleosides (MOE) in the presence of B(C6F5)3, as a Lewis Acid catalyst, furnished 50-O-
protected derivatives of 20-MOE-ribonucleosides in good yields. The deprotection of the DMPx groups was
accomplished by acid hydrolysis under very mild conditions. Among the five DMPx analogs synthesized,
the 2,7-dimethyl-9-(4-nitrophenyl)xanthene-9-yl group furnished crystalline products enabling non-
chromatographic isolation of 50-O-protetced nucleosides.
Received 3 April 2012
Revised 13 June 2012
Accepted 15 June 2012
Available online 21 June 2012
Keywords:
Nucleosides
Protecting groups
DMT group
Ó 2012 Elsevier Ltd. All rights reserved.
Pixyl group
Dimethylpixyl group
Tris(pentafluorophenyl)borane
The selective protection of a hydroxyl group in a multifunc-
tional molecule continues to challenge chemists to design better
protecting groups.1 In this pursuit we have developed a variety
of protecting groups that are useful for the protection of nucleo-
sides.2 The recent surge in the large number of therapeutic oligo-
nucleotides progressing through advanced stages of clinical trials
has triggered a need for identifying process improvements that
will better enable commercial-scale synthesis of these molecules.3
In this vein, we believe that the ability to synthesize crystalline
products is of significant value. The 4,40-dimethoxytrityl (DMT)
protected nucleosides are key building-blocks for the synthesis of
therapeutic oligonucleotides; however, DMT-protected nucleo-
sides are often non-crystalline compounds requiring extensive
purification and isolation via column chromatography increasing
the cost-of-goods.
These criteria led us to the selection of the Px group which pos-
sesses a highly rigid and planar structure promoting crystalline
products while increasing the stability of the pixyl carbocation
over other similar groups, such as DMT. The use of a Px group as
a crystalline alternative to the DMT group for oligonucleotide syn-
thesis was first proposed by Reese and Yan.4 Subsequently, our
investigation indicated that use of Px group resulted in crystalline
products but at the cost of the formation of significant amounts of
the undesired 30,50-bis-protected nucleoside.5 Furthermore, the in-
creased lability of the Px group compared to DMT resulted in for-
mation of n+1 products during oligonucleotide synthesis.6 To fine
tune the lability of the Px group, the 2,7-dimethylpixyl (DMPx)
group was conceived as a desirable analog.7 However, the DMPx
group turned out to be more labile than the Px group during oligo-
nucleotide synthesis.8 As a result, we elected to modulate the acid-
lability of the DMPx group by the introduction of a p-substituent in
the 9-phenyl residue (R in Fig. 1). Herein, we describe the role of
the addition of an electron-withdrawing substituent on the DMPx
group to increase the stability of the original DMPx group while
maintaining the crystallinity of the 50-hydroxyl protected nucleo-
side as a whole.
Recognizingthe limitationsof DMT group, wesoughtto developa
new hydroxyl protecting group with the following attributes: (i) it
should be easily synthesized on large-scale from readily available
chemicals; (ii) it should offer the desired regioselectivity for the pri-
mary hydroxyl group in the nucleosides; (iii) it should rapidly cleave
under acidic conditions, and (iv) it should offer crystalline products.
The conventional method for the introduction of the 9-phenylx-
anthene-9-yl (Px) group requires the use of pixyl chloride (Px-Cl)
in the presence of pyridine. The synthesis of Px-Cl was accom-
⇑
Corresponding author. Tel.: +1 760 944 1541; fax: +1 760 944 1543.
E-mail address: rasayan@sbcglobal.net (Y.S. Sanghvi).
0040-4039/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tetlet.2012.06.081

4670
S. Banerjee et al. / Tetrahedron Letters 53 (2012) 4669–4672
Figure 1. Various 50-O-protecting groups for nucleosides.
Table 1
plished via treatment of 9-phenylxanthen-9-ol (Px-OH) with acetyl
chloride,4 while the synthesis of Px-OH was accomplished by a
Grignard reaction of xanthone with phenyl magnesium bromide.
The use of hazardous reagents for the synthesis of Px-Cl motivated
us to develop a safer and scalable protocol for these molecules.
Large-scale synthesis of DMT-Cl is well established using Friedel–
Crafts reaction;9 therefore, we envisioned a synthetic route utiliz-
ing a diaryl ether reacting with benzoic acid in the presence of a
Lewis acid catalyst. Because ortho- and para-positions of the aryl
ether are reactive sites, we chose to employ di-p-tolyl ether (1)
to minimize side reactions. Replacement of benzoic acid with p-
substituted benzoic acids in this reaction would offer a straightfor-
ward protocol for synthesis of p-substituted DMPx-OH analogs in a
facile manner needed for the current study.
The general route for synthesis of DMPx-OH derivatives is
shown in Scheme 1. Key starting material 1 was synthesized from
p-cresol and 4-iodotoluene, following a published protocol in >80%
yield.10 Under the optimized conditions, reaction of di-p-toluyl
ether (1) and benzoic acid (R = H; 2a) in the presence of zinc chlo-
ride and phosphorus oxychloride at 95 °C furnished 2,7-dimethyl-
9-phenylxanthen-9-alcohol (DMPx-OH; R = H, 3a) in 66% yield.
This protocol was efficiently transferred to the synthesis of four
p-substituted analogs 3b–e in consistently good yield (Table 1).11
The scalability of this route was demonstrated by synthesis of
3a and 3e on large-scale to provide crystalline products without
the need for any chromatography. The procedure described herein
for the synthesis of DMPx-OH analogs is significantly simpler and
safer than the reported protocol4 for the original synthesis of Px-
OH.
Yield and analytical data for DMPx-OH
Compound (3)
R
Yield (%)
Mp (°C)
a
b
c
d
e
H
Br
Ph
t-Bu
NO2
66
59
71
70
86
110–115
151–153
202–204
115–120
125–130
Although such transformation is not unexpected, we attributed
this to the trace amount of acid triggering the formation of
DMPx-OMe. Further treatment of DMPx-OMe with AcOH (aq) fur-
nished DMPx-OH in quantitative yield.
Based on this observation and earlier reports in the literature12
describing the use of Lewis acid for tritylation of alcohols, we were
inspired to develop a protocol that will directly install the DMPx-
OH onto the 50-hydroxyl group of nucleosides. Furthermore, the
additional cost of the conversion of DMPx-OH to the corresponding
chloride and the issues with the storage and stability of these
moisture sensitive reagents were yet another reason for us to ex-
plore a Lewis acid catalyzed protocol. Four Lewis acids were eval-
uated as potential candidates for the conversion of nucleoside 4a to
5a using DMPx-OH 3a in dichloromethane as a solvent. The results
are summarized in Table 2. We found that use of tris(pentafluor-
ophenyl)borane (TPB) furnished the best yield and regioselectivity
for the 50-O-protection of the MOE-T nucleoside, 4a. This finding is
not surprising in light of a recent report demonstrating chemose-
lective tritylation of alcohols with TPB.13
Because our focus of this study was to develop crystalline deriv-
atives of 20-MOE-nucleosides,14 we further explored the use of TPB
for direct installation of DMPx group on these nucleosides (Scheme
2). For our initial screening, we chose MOE-T 4a as a model com-
pound. Among the various solvents tested (DCM, ACN, DMF, DMSO,
and NMP), both dichloromethane and acetonitrile were found to
give the best results. The protection of MOE-T was regioselective
for all four substituted DMPx-OH derivatives (3b–e: R = Br, Ph, t-
Bu, and NO2); the yield and melting-point data for protected
MOE-T 5a–e are summarized in Table 3.15 The yield did not change
by extending the reaction time or changing the solvent to higher
boiling dichloroethane.
Next, we turned our attention to preparing 50-O-DMPx nucleo-
side analogs. Treatment of DMPx-OH (3a) with excess of acetyl
chloride in toluene furnished DMPx-Cl in quantitative yield. Inter-
estingly, protection of 20-O-methoxyethyl-thymidine nucleoside 4a
(MOE-T) with DMPx-Cl furnished a significant amount of bis-DMPx
protected nucleoside analogous to our previous study.5 Based on
our experience, the use of freshly crystallized reagent worked bet-
ter during the protection step; therefore, freshly crystallized
DMPx-Cl was used for the protection of nucleoside 4a. During at-
tempted crystallization of DMPx-Cl from MeOH, we were able to
transform 3e into corresponding DMPx-OMe analog quantitatively.
Table 2
Screening of Lewis acids for 50-O-DMPx-protection of MOE-T 4a in DCM at reflux for
4–24 h
Lewis Acid
Mol (%)
Yield (%)
ZnCl2
I2
PTSA
B(C6F6)3 (TPB)
10
10
10
3
50
10
40
60
Scheme 1. Synthesis of five DMPx-OH analogs.

S. Banerjee et al. / Tetrahedron Letters 53 (2012) 4669–4672
4671
Table 4
Regioselective
Protection
Deprotection half-lives of substituted DMPx groups
Pg
B
B
O
O
Compound
T1/2 (s)
3a e
-
O
Protecting Group
HO
DMT-MOE-T
90
9
40
5
15
63
O
O
3 Mol % Lewis acid
DCM/Reflux/3-4 h
HO
O
HO
O
5a
5b
5c
5d
5e
4a-d
5a-h
B = 5-Me-U, 5-Me-CBz, ABz, GiBu
Scheme 2.
Table 5
Yield for 50-O-protection of four MOE-nucleosides with 3e
Table 3
Various DMPx-groups 3a–e tested for the 50-O-protection of MOE-T 4a (B = 5-Me-U)
in DCM reflux for 4–24 h
Starting material (4)
Product (5)
Base (B)
Yield (%)
Compound (5)
R
Yield (%)
Mp (°C)
a
b
c
e
f
g
h
5-Me-U
5-Me-CBz
ABz
60
30
a
b
c
d
e
H
Br
Ph
t-Bu
NO2
68
31
35
59
60
89–92
60
102–105
122–126
115–120
125–130
d
GiBu
1016
(5d) had virtually, no effect, while the phenyl substituent (5c)
slightly increased the rate of cleavage. It is noteworthy that selec-
tion and placement of a single p-substituent in the phenyl residue
could be utilized to modulate the cleavage rate of the DMPx analogs
in a controlled manner.
Based on the data summarized above, it was clear that the p-ni-
tro derivative 3e had ideal characteristics as a potential alternative
protecting group in place of DMT group. The desirable attributes
include: (i) ease of synthesis in high yield as a crystalline product;
(ii) regioselectivity during protection with TBP as catalyst; (iii) and
ability to isolate protected nucleoside as crystalline product with-
out chromatography. Given the importance of the 20-MOE-nucleo-
sides,14 we utilized the p-nitro derivative 3e for the protection of
all four building-blocks. The results are summarized in Table 5.
The general protocol worked well for the protection of MOE-T 4a
and MOE-ABz 4c, both in 60% isolated yield.15 Whereas, the MOE-
5-Me-CBz 4b and MOE-GiBu 4d furnished 30% and 10% yield, respec-
tively. Replacement of DCM with acetonitrile as a solvent did not
help improve the yield. We attributed the lower yield for 5f and
5h due to the poor solubility in the solvents utilized.16 Neverthe-
less, the unprotected nucleosides 4a–d could be recovered from
the aqueous layer after extraction of the protected nucleosides 5.
In summary, 2,7-dimethyl-9-(4-nitro)phenyl xanthen-9-ol
(R = NO2: 3e) has emerged as a protecting group of choice furnish-
ing crystalline products. It is noteworthy that we have achieved
most of our objectives, namely: (i) ease of protecting group synthe-
sis; (ii) regioselective introduction onto the 50-hydroxyl group, and
(iii) efficient cleavage under mild conditions. To the best of our
knowledge, this is the first example of protection of 50-hydroxyl
group in ribonucleosides using a Lewis acid catalyst and an aryl
alcohol with ability to ‘fine-tune’ the cleavage rate. This approach
eliminates the use of moisture sensitive reagents traditionally used
for the protection of nucleosides. The simplicity and ease of syn-
thesis of 3e is expected to open new possibilities of its use as a pro-
tecting group in a wider arena of organic synthesis.17 Furthermore,
the Px group has been investigated as a potential photocleavable
group for primary alcohols.18 Similarly, the DMPx analogs 3a-e
synthesized in this study may find application as photolabile pro-
tecting groups. The utility of 3e as a protecting group during oligo-
nucleotide synthesis is currently under evaluation.
We next studied the rate of cleavage of our newly synthesized p-
substituted DMPx protected nucleosides. Therefore, we carried
out a comparative unblocking study of the protecting groups from
compounds 5a–e, along with 50-O-DMT-MOE-T as a control. These
experiments were carried out following a method similar to that
used by Reese and Yan.4 Each of the protected MOE-T p-R-DMPx
analogs 5a-e were treated with dichloroacetic acid in dichlorometh-
ane in the presence of pyrrole, using tri-O-acetyl uridine as an inter-
nal standard. The solutions were then quenched at various time
points with 0.7 M methanolic triethylamine. The percent cleaved
was determined by LCMS, and then fitted using one phase decay
model in GraphPad Prism (Chart 1). The resulting half-lives are
shown in Table 4. All fits had an R square of 0.99 or better, except
5e (R2 = 0.898). It can be clearly seen from the chart, that increasing
the extent of electron withdrawing capabilities at the p-substituent
in the 9-phenyl residue of DMPx group greatly attenuates the rates
of cleavage. For example, the bromo-substituent (5b) resulted in a
rate nearly half that of the DMT group (50-O-DMT-MOE-T), versus
10 times less for the parent DMPx group (5a). The nitro substituent
(5e) had a rate that was even closer to that of DMT group, with an
initial rate that appears to be also very close. The t-butyl substituent
Supplementary data
Supplementary data (1H NMR and selected MS) associated with
this article can be found, in the online version, at http://dx.doi.org/
10.1016/j.tetlet.2012.06.081.
Chart 1. Cleavage of substituted DMPx groups.

4672
S. Banerjee et al. / Tetrahedron Letters 53 (2012) 4669–4672
afford 4-nitro-DMPx 3e as white powder. The crude product was dissolved in
References and notes
EtOAc and washed with NaHCO3 to remove the excess of 4–nitro benzoic acid.
The EtOAc layer was evaporated under reduced pressure to furnish the desired
product as fine powder 7.5 g (86%).
1. Kocienski, P. J. Protecting Groups; Thieme: Stuttgart, New York, 2004. pp 187–
364; Seliger, H. Current Protocols in Nucleic Acid Chemistry; John Wiley & Sons,
2000. pp. 2.3.1-2.3.34.
2. Oka, N.; Sanghvi, Y. S.; Theodorakis, E. A. Bioorg. Med. Chem. Lett. 2004, 14,
3241–3244; Oka, N.; Sanghvi, Y. S.; Theodorakis, E. A. Synlett 2004, 823–826;
Wen, K.; Chow, S.; Sanghvi, Y. S.; Theodorakis, E. A. J. Org. Chem. 2002, 67,
7887–7889.
3. Sanghvi, Y. S. Current Protocols in Nucleic Acid Chemistry; John Wiley & Sons,
2011. p 4.1.1-4.1.22, September.
4. Reese, C. B.; Yan, H. Tetrahedron Lett. 2004, 45, 2567–2570; Chattophadhyaya, J.
B.; Reese, C. B. J.C.S. Chem. Comm. 1978, 639–640.
12. Adinolfi, M.; Barone, G.; Iadonisi, A.; Schiattarella, M. Tetrahedron Lett. 2003, 44,
3733–3735; Salehi, P.; Iranpoor, N.; Behbahani, F. K. Tetrahedron 1998, 54,
943–948.
13. Reddy, C. R.; Rajesh, G.; Balaji, S. V.; Chethan, N. Tetrahedron Lett. 2008, 49,
970–973.
14. Currently there are >25 active clinical trials in progress with antisense
oligonucleotides containing 20-MOE-modification (www.isispharm.com).
15. General Protocol for Nucleoside Protection: 4-(NO2) DMPx-protected (2-
methoxyethyl)-5-methyluridine:
A
mixture of 20-O-methoxyethyl-thymidine
(MOE-T, 5 g, 15.8 mmol) and B(C6F5)3 (242 mg, 3 mol %) in CH2Cl2 (100 mL)
was stirred at room temperature for 30 min. Next, 4-(NO2)-DMPx-OH (3e,
5.48 g, 15.80 mmol) was dissolved in CH2Cl2 (50 mL) and added to the above
reaction mixture drop wise under nitrogen atmosphere. The resulting reaction
mixture was stirred at room temperature for 3 h. The progress of the reaction
was monitored by TLC. After maximum conversion of the reaction, reaction
mixture was washed with saturated NaHCO3 (50 mL) solution and brine
solution (25 mL). The organic layer was dried over anhydrous Na2SO4 and
evaporated under reduced pressure. Crude product upon crystallization from
EtOH furnished off white powder 6.19 g (60%).
5. Day, R. T.; Williams, D.; Soriano, P.; Sanghvi, Y. S. Nucleosides, Nucleotides
Nucleic Acids 2005, 24, 1135–1138.
6. Unpublished results. The n+1 mer formation is relatively higher (2–5%) when Px
group is used for solid-phase oligonucleotide synthesis compared to the use of
standard DMT-protecting group for same sequence.
7. Song, Q.; Khammungkhune, S.; Ross, B.S.; Griffey, R.H. 2007, U.S. Patent
Application No.: US 2007/0276139 A1.
8. Tram, K.; Sanghvi, Y. S.; Yan, H. Nucleosides, Nucleotides Nucleic Acids 2011, 30,
12–19; Millar, S.; Yan, H.; Sanghvi, Y. S.; Pon, R. T. Nucleic Acids Symp. Ser. No.
2008, 52, 311–312.
9. Rathore, M.; Rani, P.; Mathur, N. K.; Narang, C. K. Indian J. Chem. 1995, 34B,
634–635.
10. Chang, J. W.; Chee, S.; Mak, S.; Buranaprasertsuk, P.; Chavasiri, W.; Chan, P. W.
Tetrahedron Lett. 2008, 49, 2018–2022.
11. Representative Experimental Protocol: 2,7-Dimethyl-9-(4-nitro)phenyl xanthen
-9-ol (R = NO2: 3e): Di-p-tolyl ether (1, 5 g, 25.5 mmol), 4-nitro benzoic acid
(5.13 g, 30.7 mmol) and ZnCl2 (10 g, 73.4 mmol) are taken into a 100 ml double
neck RB flask and charged with POCl3 (7.5 mL) and heated to 95 °C for 4 h. The
reaction mixture turned into a brown viscous mass that was difficult to stir.
The reaction mixture was then cooled to room temperature and crushed ice
was added slowly while keeping the low temperature. To the above mixture
was added H2O (75 mL) and stirred for 12 h. The above reaction mixture was
filtered and the residue was washed with H2O (20 mL) and hexane (20 mL) to
16. We were able to increase the overall isolated yield to 40% for 50-O-protected 5h
by using a mixed solvent system (DCM/DMF; 1:1 v/v) where solubility of the
starting material 4d was superior compared to the use of DCM alone.
Therefore, we believe that good solubility of the starting material is essential
for improved yields.
17. We were able to selectively protect the primary hydroxyl group in: (i) 2-
hydroxy-propanol (60% yield); (ii) thymidine (58% yield); and (iii) 20-O-
methyl-uridine (60% yield). The NMR data for these three 4-(NO2)DMPx-
protected compounds are included in the Supplementary data. These examples
further supports the possible use of 4-(NO2)DMPx-OH as an acid labile
protecting group that may furnish crystalline products.
18. Misetic, A.; Boyd, M. K. Tetrahedron Lett. 1998, 39, 1653–1656; Coleman, M. P.;
Boyd, M. K. J. Org. Chem. 2002, 67, 7641–7648.