An industrial process for synthesizing lodenosine (FddA)

Article abstract of DOI:10.1081/NCN-120021951

Two industrial synthetic approaches to Lodenosine (1, FddA, 9-(2,3-dideoxy-2-fluoro-β-D-threo-pentofuranosyl) adenine) via a purine riboside or a purine 3′-deoxyriboside are described. Several novel applications of deoxygenation and fluorination methods are compared considering reaction yields, economy, safety and environmental concerns.

Full text of DOI:10.1081/NCN-120021951

This article was downloaded by: [University of Calgary]
On: 09 September 2013, At: 04:16
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
An Industrial Process for Synthesizing Lodenosine
(FddA)
Kunisuke Izawa a c , Satoshi Takamatsu a , Satoshi Katayama a , Naoko Hirose a , Shigetaka
Kozai b & Tokumi Maruyama b
a AminoScience Laboratories, Ajinomoto Co., Inc., Kawasaki-ku, Kawasaki, Japan
b Faculty of Pharmaceutical Sciences, Tokushima Bunri University, Yamashiro-cho,
Tokushima, Japan
c AminoScience Laboratories, Ajinomoto Co., Inc., 1-1 Suzuki-cho, Kawasaki-ku, Kawasaki,
210-8681, Japan
Published online: 31 Aug 2006.
To cite this article: Kunisuke Izawa , Satoshi Takamatsu , Satoshi Katayama , Naoko Hirose , Shigetaka Kozai & Tokumi
Maruyama (2003) An Industrial Process for Synthesizing Lodenosine (FddA), Nucleosides, Nucleotides and Nucleic Acids,
22:5-8, 507-517, DOI: 10.1081/NCN-120021951
To link to this article: http://dx.doi.org/10.1081/NCN-120021951
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 & NUCLEIC ACIDS
Vol. 22, Nos. 5–8, pp. 507–517, 2003
An Industrial Process for Synthesizing Lodenosine (FddA)
1
1
Kunisuke Izawa,1, Satoshi Takamatsu, Satoshi Katayama,
*
Naoko Hirose,1 Shigetaka Kozai,2 and Tokumi Maruyama2
1AminoScience Laboratories, Ajinomoto Co., Inc.,
Kawasaki-ku, Kawasaki, Japan
2Faculty of Pharmaceutical Sciences, Tokushima Bunri University,
Yamashiro-cho, Tokushima, Japan
ABSTRACT
Two industrial synthetic approaches to Lodenosine (1, FddA, 9-(2,3-dideoxy-
2-fluoro-b-d-threo-pentofuranosyl) adenine) via a purine riboside or a purine
30-deoxyriboside are described. Several novel applications of deoxygenation and
fluorination methods are compared considering reaction yields, economy, safety
and environmental concerns.
Key Words: Anti-HIV; Nucleosides; Lodenosine; Fluorination; Radical
reduction.
INTRODUCTION
Lodenosine [9-(2,3-dideoxy-2-fluoro-b-d-threo-pentofuranosyl) adenine, FddA, 1]
is an acid-stable purine nucleoside analog with activity against human immuno-
deficiency virus (HIV).[1] It acts as an inhibitor of HIV reverse transcriptase (RT),
like other nucleosides derivatives such as zidovudine (AZT), 20,30-dideoxyadenosine
(2, ddA), and didanosine (3, ddI) (Fig. 1). FddA was evaluated as one of the most
*Correspondence: Kunisuke Izawa, AminoScience Laboratories, Ajinomoto Co., Inc., 1-1
Suzuki-cho, Kawasaki-ku, Kawasaki210-8681, Japan; E-mail: kunisuke_izawa@ajinomoto.com.
507
DOI: 10.1081/NCN-120021951
Copyright # 2003 by Marcel Dekker, Inc.
1525-7770 (Print); 1532-2335 (Online)
www.dekker.com

508
Izawa et al.
Figure 1. 20,30-Dideoxy purinenucleosides.
selective inhibitors in a series of 20,30-dideoxyadenosines with either an azido,
fluorine, or hydroxy group substituted in the ‘‘up’’ or ‘‘down’’ positions of C-20 and
C-30 of the ribose moiety, although it was less active than the parent compound, ddA,
2.[2] In addition, it showed the best selectivity index (activity versus cellular toxi-
city) in the series.[2] An advantage of FddA is its stability under acidic conditions
which decompose ddA and ddI.[1a,lb] Other promising aspects are that FddA does
not exhibit cross-resistance to AZT, zalcitabine (ddC) and ddI,[3] and even has
synergistic activity with AZT.[4] FddA also acts in synergy with other anti-HIV
agents including HIV protease inhibitors like Ritonavir, and thus provides new
possible combination therapy protocols for HIV.[5] Hence, FddA was developed
as a novel anti-HIV agent exhibiting significant advantages in terms of higher sta-
bility, high oral bioavailability[6] and effectiveness against strains resistant to exist-
ing dideoxynucleosides.[7]
The first synthesis of FddA, 1, was reported in 1987 by Herdewijn et al.[2] 30-
Deoxyadenosine (cordycepin) was used as a starting material, and is easily prepared
from adenosine using the method reported by Hansske and Robins.[8] 50-O-Trityl-
cordycepin was treated with diethylaminosulfur trifluoride (DAST) followed by detri-
tylation to give FddA, 1, in a 10% overall yield after purification. Several years
later, we examined a similar reaction with the 50-O-acetyl compound, but the yield
was also very poor.[9] At almost the same time as the first synthesis, Marquez and
co-workers extensively studied the synthesis of FddA and reported several
methods.[10] Some of their approaches involved the coupling of a fluorinated sugar
to a purine derivative. Although these reaction sequences provided a better yield of
FddA on a laboratory scale, they might not be appropriate for an industrial pro-
cess, since the reaction sequences required for the synthesis of the fluorinated sugar
were rather long and the coupling reaction was of low selectivity. Marquez also
reported the synthesis of FddA from ara-A with a shorter reaction sequence.[11]
However, since ara-A is a rather expensive starting material, the process is unlikely
to be economic. Very recently, an efficient synthesis of FddA was reported
by Jin et al.[12] They prepared a fluorinated sugar intermediate very efficiently
and also improved the yield for the coupling reaction. They reported an over-
all yield of 14.9% with a 9-step reaction sequence. Although the process requires
chromatography, this route might be one of the most economical processes
developed so far.

Synthesis of Lodenosine
509
Scheme 1.
In 1992, Pankiewicz, Watanabe and their co-workers reported a fluorination reac-
0
0
tion of O3 , O5 , N6-tritrityladenosine 4 with DAST.[13] Remarkably, they found that
30% of the desired product 5 was obtained along with an accompanying formation of
the purine base rearrangement product 6 in a 51% yield (Sch. 1). We speculate that this
same rearrangement might be the reason for the low yield in the Herdewijn synthesis.
SYNTHETIC STRATEGY
In the course of studies on the synthesis and activities of 6-chloropurine nucleo-
side analogs, Maruyama et al. found that this type of base rearrangement could be
completely suppressed by introducing a chlorine atom at the 6-position of the purine
base.[14] Thus, the reaction of the 30, 50-ditrityl derivative 8a of the 6-chloropurine
riboside 7 with DAST gave an 87% fluorinated product 9 (Sch. 2). Compound 9
was then successfully converted into F-ara-A 11.
Scheme 2.

510
Izawa et al.
Scheme 3.
However, this process had drawbacks: the yield of 30,50-ditritylation was very
low and the product 8a had to be purified by column chromatography. To apply this
methodology to synthesize FddA, 1, the selective protection of the 30- and 50-posi-
tions by a different group is necessary. Another important factor for industrial syn-
thesis is the fluorination reagent. We found that DAST is not a suitable agent
because of its availability, safety and cost on a large scale.
Based on this consideration, we attempted to develop two different routes
to FddA from inosine (Sch. 3). Method A involves fluorination, followed by
deoxygenation of the 30-OH group in the later stages of the synthesis. For this
purpose, it was necessary to achieve selective protection of the 30 and 50 positions of
the 6-chloropurine riboside. On the other hand, Method B involves the synthesis
of 6-chloropurine 30-deoxyriboside with the fluorination step in the later part of
the synthesis.
A Synthesis of FddA via 6-Chloropurine Riboside (Method A-1)[15]
We first used the 20,30-O-di-n-butylstannylene complex of 6-chloropurine riboside
to introduce only one benzoyl group at the 20-OH and=or 30-OH. Thus, 6-chloro-
purine riboside 7 was successively treated with di-n-butyltin oxide and excess
benzoyl chloride (BzCl) in the presence of triethylamine (Et3N) in MeOH. After
the usual work-up and purification by silica gel chromatography, an inseparable
mixture of 30-O-benzoate 17 and 20-isomer was subjected to crystallization from
MeOH to afford 96.1% pure 30-O-benzoate 17 in 65% yield in crystalline form.
Compound 17 was reacted with trityl chloride (TrCl) in the presence of Et3N and
4-dimethylaminopyridine (DMAP) in N,N-dimethylformamide (DMF). Thin-layer
chromatography analysis indicated a small amount of migration of the benzoyl

Synthesis of Lodenosine
511
Scheme 4.
group from the 30-position to the 20-position under these reaction conditions. For-
tunately, however, the desired 30-O-benzoate 13a was obtained in a 71% yield with
98% purity after crystallization from MeOH (Sch. 4). Since the fluorination reac-
tion with DAST should be carried out in the presence of amine to protect the
50-O-trityl group, the stability of 30-O-benzoate 13a was checked before the reac-
tion. Thus, the compound was heated under reflux with 6 equiv. of pyridine in
CH2Cl2 for 5 h. To our delight, only 3.1% of the 20-O-benzoate 13b was observed
in the mixture. We then examined the reaction of 13a with DAST using the same
conditions and obtained the product 14 in a 78% yield.
The 30-O-benzoyl-50-O-trityl protecting system has two advantages: it is easy to
introduce the protecting group compared with the di-O-trityl group and the selective
removal of either the 30- or 50-O-protecting group proceeds smoothly. Thus, deprotec-
tion of the 30-O-benzoyl group and displacement of the 6-chloro group of 14 could be
accomplished by treatment with ammonia in MeOH to afford 18 in a 73% yield.
Deoxygenation of the 30-hydroxy group was achieved by the conventional method.
Compound 18 was treated with phenyl chlorothionoformate to give 19 in an 80%
yield. The product was then treated with tris(trimethylsilyl)silane in the presence of
2,20-azobis(isobutyronitrile) (AIBN) in toluene under N2 atmosphere to give 20 as
white crystals in a 73% yield. Acid treatment of 20 gave the desired 9-(2,3-dideoxy-
2-fluoro-b-d-threo-pentofuranosyl)adenine (FddA, 1) as prisms in an 88% yield. Ana-
lytical data for this sample were identical in all respects to the published data.
An Improved Synthesis of FddA via 6-Chloropurine Riboside
(Method A-2)[16]
The more efficient protection methodology we developed is shown in Sch. 5. The
selective 50-O-tritylation of 6-chloropurine riboside (79%) was performed with 2

512
Izawa et al.
Scheme 5.
equiv. of trityl chloride in the presence of diisopropylamine (iPr2NH) in DMF at 40ꢀC
(Sch. 6). Selective benzoylation of the 30-position was carried out using excess benzoyl
chloride in the presence of pyridine in toluene followed by treatment with a catalytic
amount of triethylamine in the same solvent to give 13a in a 70% yield. Before fluor-
ination, the 20-hydroxy group of 13a was converted to imidazolesulfonate 22a or tri-
fluoromethanesulfonate 22b. Compound 22a was treated with Et3Nꢁ3HF in toluene
at 70ꢀC to give 14 in a 78% yield, including imidazolesulfonation. The trifluoro-
methanesulfonate 22b, which was quantitatively obtained from 13a, was reacted with
6 equiv. of Et3Nꢁ3HF and 3 equiv. of Et3N in ethyl acetate (EtOAc) at reflux tem-
perature. Fluorination proceeded very smoothly to give 14 in an 88% yield. To the
best of our knowledge, this is the highest reported yield in fluorination at the 20-posi-
tion of a purine riboside. Next, we treated compound 14 with ammonia to give 18 by
simultaneous 6-amination and 30-benzoyl deprotection in almost quantitative yield.
To complete the synthesis of F-ddA, deoxygenation of the 30-hydroxy group in
18 was carried out by the Barton-McCombie method[17] with silane, thus avoiding
the use of toxic tin hydride. Compound 18 was converted with phenyl chlorothiono-
formate to the 30-O-thiocarbonyl derivative 19 in a 92% yield. The product was then
subjected to deoxygenation with various silanes in the presence of AIBN. The results
are summarized in Table 1.
Among these, diphenylsilane was the best of choice to produce a reduction pro-
duct 20 in an 81% yield, which in turn was treated with 80% acetic acid to give FddA,
1, of an excellent yield and quality.
Although we successfully synthesized FddA in good yield, several points should
be improved. When we examined the cost in further detail, the silane used in the
deoxygenation step was found to be relatively expensive. Therefore, we examined
other radical reductions and reduction of the xanthate 23 with hypophosphorous
acid proved to be the most efficient. Table 2 summarizes the results.[18]

Synthesis of Lodenosine
513
Table 1. Deoxygenation of 19 with silanes.
Run
R
Silane
Solvent
Yield %
1
2
3
4
Ph
Ph
Ph
Ph
[(Me)3Si]3SiH
Ph2SiH2
Ph2SiH2
Toluene
Dioxane
Toluene
Dioxane
78
81
74
69
Ph2MeSiH
Scheme 6.
The best yield, 93%, was obtained when we used 10 equiv. of hypophosphorous
acid and 11 equiv. of Et3N in DME=H2O (Sch. 7). In summary, we achieved the
synthesis of FddA from 6-chloropurine riboside in a 32.8% overall yield and 8 steps.
A Synthesis via 6-Chloropurine 30-Deoxyriboside (Method B)[19]
Although the overall yield for the reaction sequence in Method A-2 was the best
thus far, we further investigated other routes in an attempt to lower the production
costs. Thus, we attempted the synthesis of 6-chloropurine 30-deoxyriboside 27 from a
readily available inosine, 12. The synthesis of 27 was achieved as shown in Sch. 8.
The acetoxybromination technology used is that which we had already developed
Table 2. Deoxygenation of xanthate 23.
Run Solvent
M-H
P-H=eq. Et3N=eq. AIBN=eq. Temp.=ꢀC Time=hrs Yield %a
1
2
3
4
5
6
7
DME H3PO2
DME H3PO2
DME H3PO2
DME H3PO2
DME H3PO2
5
10
10
5
2.5
10
10
10
20
11
5.5
2.8
0.3
0.6
0.6
0.1
0.2
0.6
0.6
Reflux
Reflux
Reflux
Reflux
Reflux
Reflux
Reflux
1
84
86
93
90
90
84
82
1.8
1.8
1.2
1.3
2
DME (MeO)2P(O)H
DME (EtO)2P(O)H
–
–
2
aCalculated from the results of HPLC analysis.

514
Izawa et al.
Scheme 7.
for the large scale synthesis of ddA.[20] The reaction proceeded well to afford 24 in an
80% overall yield from 12. We then carried out the radical debromination of 24 using
hypophosphorous acid as a reducing agent in the presence of a water-soluble radical
initiator such as V-50 or VA-044 (Fig. 2 to give 25 in almost quantitative yield.[18]
When we used AIBN as a radical initiator, we observed a lower isolated yield due
to a loss of product during the purification to remove residual AIBN.
6-Chlorination and subsequent deacetylation were carried out in a conventional
manner, and the product 27 was tritylated to give 15 for the next fluorination step in
85% yield.
Another high-cost step is fluorination. Accordingly, we examined the use of a
perfluoroalkanesulfonyl fluoride such as perfluoro-1-butanesulfonyl fluoride (non-
afluoro-1-butanesulfonyl fluoride, C4F9SO2F, NfF) or perfluoro-1-octanesulfonyl
fluoride (C8F17SO2F, OctF) for fluorination, and compared the yields with those
using other fluorination agents. Vorbruggen and co-workers reported that the com-
¨
bination of NfF with 1,8-diazabicyclo[5,4,0]undec-7-ene (DBU) in non-polar sol-
vents efficiently converts primary and secondary alcohols into their corresponding
Scheme 8.

Synthesis of Lodenosine
515
Figure 2.
fluorides.[21] In our first experiment, we noticed that NfF reacted slowly with 15 in
the presence of Et3N, even at 50ꢀC. Using pyridine as a base, the reaction did not
proceed at all. However, immediately after DBU was added to this mixture, 15
was converted into the fluorinated product 16 and the elimination product 28. We
next sought to improve the ratio of fluorination to elimination. The base must be
a good proton acceptor with a moderate basicity to avoid elimination and must also
have low nucleophilicity. Among the bases we studied, N-ethylpiperidine (NEP) gave
better results with regard to both the reaction rate and selectivity. At 50ꢀC, with 4
equiv. of NfF and 2 equiv. of NEP as a base, the reaction went to completion after
44 h with 64 area% in HPLC. N,N-Dimethylcyclohexylamine (DMCHA) also seems
to be a good substitute for DBU. With 2 equiv. of NfF and 2 equiv. of DMCHA as a
base, HPLC analysis indicated a 62.4% yield after aqueous work-up. The yield of 16
from 15 with NfF was much greater than that using DAST and almost the same as
the previous 2-step method involving trifluoromethanesulfonylation and fluorination
with Et3Nꢁ3HF (Sch. 9).[22]
We also successfully removed the elimination side-product 28 simply by treating
the reaction mixture with 80% acetic acid in toluene at 10ꢀC for 3 h. The desired
product was isolated by crystallization without any loss. The second synthesis of
F-ddA (Method B) from inosine is outlined in Sch. 10.
6-Chloropurine 30-deoxyriboside was efficiently synthesized using radical debro-
mination. Fluorination was successfully carried out using perfluoroalkanesulfonyl
Scheme 9.

516
Izawa et al.
Scheme 10.
fluoride. Thus, we synthesized FddA in a 36% overall yield and a total of 9 steps.
This process is far superior to the previous method.
Unfortunately, the development of FddA was suspended[23] for reasons of
toxicity just after the pilot plant production of compound 27 on a 3000-L scale.
However, the methodology developed in this study might be very useful for the
synthesis of related nucleosides.
ACKNOWLEDGMENTS
We are grateful to Mr. M. Naito, Dr. E. De Cock, Mr. G. Schelkens, Dr. M.
Demillequand, Dr. J. Brepoels, and Dr. Y. Sato for their contributions to this study.
We also wish to acknowledge the valuable suggestions of Dr. K. Watanabe.
REFERENCES
1. (a) Marquez, V.E.; Tseng, C.K.-H.; Kelly, J.A.; Mitsuya, H.; Broder, S.; Roth,
J.S.; Driscoll, J.S. Biochem. Pharmacol. 1987, 36, 2719–2722; (b) Marquez,
V.E.; Tseng, C.K.-H.; Mitsuya, H.; Aoki, S.; Kelly, J.A.; Ford, H., Jr.; Roth,
J.S.; Broder, S.; Johns, D.G.; Driscoll, J.S. J. Med. Chem. 1990, 33, 978–985;
(c) Graul, A.; Silvestre, J.; Castaner, J. Drugs Future 1998, 23, 1176–1189
and references cited therein.
2. Herdewijn, P.; Pauwels, R.; Baba, M.; Balzarini, J.; De Clercq, E. J. Med.
Chem. 1987, 30, 2131–2137.
3. Driscoll, J.S.; Mayers, D.L.; Bader, J.P.; Weislow, O.S.; Johns, D.G.; Buckheit,
R.W. Antiviral Chem. Chemother. 1997, 8, 107–111.
4. (a) US Bioscience, Inc.: West Conshohocken, PA, Press Release, Jan. 06, 1998;
(b) Billich, A. Current Opinion Anti-Infective Drugs 1999, 1, 179–185.

Synthesis of Lodenosine
517
5. Palmer, S.; Russell, J.D.; Pallansch, L.A.; Driscoll, J.S.; Buckheit, R.W. Anti-
viral Res. 1999, 41, LB4.
6. (a) Ruxrungtham, K.; Boone, E.; Ford, Jr. H.; Driscoll, J.S.; Davey, R.T., Jr.;
Lane, H.C. Antimicrob. Agents Chemother. 1996, 40, 2369–2374 and refer-
ences cited therein; (b) US Bioscience, Inc.: West Conshohocken, PA, Press
Release, July 02, 1998.
7. Ueno, T.; Mitsuya, H. Biochemistry 1997, 36, 1092–1099.
8. Hansske, F.; Robins, M.J. Tetrahedron Lett. 1985, 26, 4295–4298.
9. Shiragami, H.; Tanaka, Y.; Uchida, Y.; Iwagami, H.; Izawa, K.; Yukawa, T.
Nucleosides & Nucleoticles 1992, 11, 391–400.
10. (a) Wysocki, R.J.; Siddiqui, M.A.; Barchi, J.J., Jr.; Driscoll, J.S.; Marquez,
V.E. Synthesis 1991, 1005–1008; (b) Siddiqui, M.A.; Marquez, V.E.; Driscoll,
J.S.; Barchi, J.J., Jr.; Tetrahedron Lett. 1994, 35, 3263–3266; (c) Siddiqui,
M.A.; Driscoll, J.S.; Abushanab, E.; Kelly, J.A.; Barchi, J.J., Jr.; Marquez,
V.E. Nucleosides, Nucleotides & Nucleic Acids 2000, 19, 1–12.
11. Siddiqui, M.A.; Driscoll, J.S.; Marquez, V.E. Tetrahedron Lett. 1998, 39,
1657–1660.
12. Jin, F.; Wang, D.; Confalone, P.N.; Pierce, M.E.; Wang, Z.; Choudhury, A.;
Nguyen, D. Tetrahedron Lett. 2001, 42, 4787–4789.
13. Pankiewicz, K.W.; Krzeminski, J.; Ciszewski, L.A.; Ren, W.-Y.; Watanabe,
K.A. J. Org. Chem. 1992, 57, 553–559.
14. Maruyama, T.; Sato, Y.; Oto, Y.; Takahashi, Y.; Snoeck, R.; Andrei, G.;
Witvrouw, M.; De Clercq, E. Chem. Pharm. Bull. 1996, 44, 2331–2334.
15. Maruyama, T.; Takamatsu, S.; Kozai, S.; Sato, Y.; Izawa, K. Chem. Pharm.
Bull. 1999, 47, 966–970.
16. Takamatsu, S.; Maruyama, T.; Katayama, S.; Hirose, N.; Izawa, K. Tetra-
hedron Lett. 2001, 42, 2321–2324.
17. Chatgilialoglu, C.; Ferreri, C. Res. Chem. Intermed. 1993, 19, 755–775.
18. Takamatsu, S.; Katayama, S.; Hirose, N.; Naito, M.; Izawa, K. Tetrahedron
Lett. 2001, 42, 7605–7608.
19. (a) Takamatsu, S.; Maruyama, T.; Katayama, S.; Hirose, N.; Naito, M.; Izawa,
K. Tetrahedron Lett. 2001, 42, 2325–2328; (b) Takamatsu, S.; Maruyama, T.;
Katayama, S.; Hirose, N.; Naito, M.; Izawa, K. J. Org. Chem. 2001, 66, 7469–
7477.
20. Shiragami, H.; Amino, Y.; Honda, Y.; Arai, M.; Tanaka, Y.; Iwagami, H.;
Yukawa, T.; Izawa, K. Nucleosides & Nucleotides 1996, 15, 31–45.
21. (a) Bennua-Skalmowski, B.; Krolikiewicz, K.; Vorbruggen, H. Bull. Soc. Chim.
¨
¨
Belg. 1994, 103, 453–461; (b) Bennua-Skalmowski, B.; Vorbruggen, H. Tetra-
hedron Lett. 1995, 36, 2611–2614.
22. Takamatsu, S.; Katayama, S.; Hirose, N.; De Cock, E.; Schelkens, G.;
Demillequand, M.; Brepoels, J.; Izawa, K. Nucleosides, Nucleotides &
Nuclei Acids 2002, 21 (11&12), 849–861.
23. US Bioscience, Inc.: West Conshohocken, PA, Press Release, Oct. 14, 1999.