An improved synthesis of (R)-2,3-dihydroxypropyl 5-deoxy-5-dimethylarsinyl-β-D-riboside, a common marine arsenical

Article abstract of DOI:10.1071/CH01036

An efficient synthesis of (R)-2,3-dihydroxypropyl 5-deoxy-5-dimethylarsinyl-β-D-riboside, a common marine arsenical, from D-ribose, (S)-2,3-dibenzyloxypropanol and iododimethylarsine is described.

Full text of DOI:10.1071/CH01036

P u b l i s h i n g
publishing research papers from all fields of chemical
Volume 54, 2001
© CSIRO 2001
science, including synthesis, structure, new materials,
macromolecules, supramolecular chemistry, biological
chemistry, nanotechnology, surface chemistry, and
analytical techniques.
All enquiries and manuscripts should be directed to:
Dr Alison Green
Australian Journal of Chemistry–
an International Journal for Chemical Science
CSIRO
PUBLISHING
PO Box 1139 (150 Oxford St)
Collingwood, Vic. 3066, Australia
Telephone: +61 3 9662 7630
Fax: +61 3 9662 7611
E-mail: ajc@publish.csiro.au
CSIRO
Published by
PUBLISHING
for CSIRO and the Australian Academy of Science
w w w. p u b l i s h . c s i r o . a u / j o u r n a l s / a j c

Aust. J. Chem. 2001, 54, 181–183
An Improved Synthesis of (R)-2,3-Dihydroxypropyl
5-Deoxy-5-dimethylarsinyl-β-D-riboside, a Common Marine Arsenical
Robert V. Stick,A,B Keith A. StubbsA and D. Matthew G. TilbrookA
A Department of Chemistry, The University of Western Australia, 35 Stirling Highway,
Crawley WA 6009, Australia.
B Author to whom correspondence should be addressed (e-mail: rvs@chem.uwa.edu.au).
An efficient synthesis of (R)-2,3-dihydroxypropyl 5-deoxy-5-dimethylarsinyl-β-D-riboside, a common marine
arsenical, from D-ribose, (S)-2,3-dibenzyloxypropanol and iododimethylarsine is described.
C
H01036
AnI.Vmp.St rovick
d
b
isosafandCoD. m
M
l
Manuscript received: 17 April 2001.
Final Version: 24 May 2001.
Arsenic is an element that is widely distributed in both the
marine and terrestrial environments.[1] In the main, the
natural arsenic-containing chemicals are benign in their
action and pose little threat to mankind. However, recent
anthropogenic activity has reversed this situation—for
example, mining activity has resulted in huge stockpiles of
the extremely toxic arsenic(III) oxide, and deep wells in arid
areas have become contaminated with life-threatening levels
of arsenic.
–
Some two decades ago, arsenobetaine (Me3As+CH2CO2 )
was identified as a major and widespread marine arsenical;[2]
a little later, a range of glyceryl β-D-ribofuranosides,
containing arsenic at C 5 of the D-ribose, were found to be
widespread in the marine environment.[1] One of these
compounds, (R)-2,3-dihydroxypropyl 5-deoxy-5-dimethyl-
arsinyl-β-D-riboside (1), has proven to be a constant
component of most marine flora studied to date. So common
is compound (1) that it is now offered commercially and
there have been calls for its supply in gram amounts.
In order to establish the structure of (1), particularly that
of the stereochemistry of the glyceryl moiety, we set about its
synthesis some years ago.[3] This successful piece of work
was followed by a somewhat related route, described by
Irgolic some years later.[4] Neither synthesis provides good
quantities of compound (1). Here, we describe a hybrid
synthesis starting with D-ribose that allows for the synthesis
of gram amounts of (1) in a period of about three weeks.
D-Ribose was easily converted into the methyl β-D-
ribofuranoside (2) (Scheme 1).[5] Although Irgolic describes
the conversion of (2) into the bromide (3) with sodium
bromide,[4] we found it much better to use tetrabutyl-
ammonium bromide [in dimethylformamide (DMF)]. The
transacetalization of (3) with the alcohol (4) (Scheme2), as a
neatmeltaccordingtoIrgolic,[4]certainlyproducedthedesired
glyceryl β-D-ribofuranoside (5), but never in the quantitative
Scheme 1. a) (i) MeOH, acetone, HCl (ii) p-toluenesulfonyl
chloride, pyr; b) Bu4NBr, DMF; c) (S)-2,3-dibenxyloxypropanol (4),
camphorsulfonic acid; d) (CH3)2AsI, Na, THF; e) Na, THF, NH3(l); f)
H2O2/H2O THF; g) TFA/H2O (9 : 1)
yield quoted. We made many futile attempts to conduct this
transformation in solution (toluene, xylene, DMF).
For the introduction of an arsenic atom at C 5 of (5) we
balked at the use of the difficult to prepare, toxic and
pyrophoric cacodyl (Me2AsAsMe2).[4] Instead, we relied on
the use of our tried and tested iododimethylarsine–sodium
couple,[3] and this produced the surprisingly stable (to air
oxidation) arsine (6). A by-product in the formation of (6)
was the original alcohol (4), probably the result of a Vasella-
© CSIRO 2001
10.1071/CH01036
0004-9425/01/030181

182
R. V. Stick et al.
(R)-2,3-Dibenzyloxypropyl 5-Bromo-5-deoxy-2,3-O-isopropylidene-β-
D-riboside (5)
The β-D-riboside (5) was prepared following the procedure of Irgolic,[4]
with the exceptions that camphorsulfonic acid was used in preference
to p-toluenesulfonic acid, and the reaction was subjected to usual
workup (EtOAc). Flash chromatography (EtOAc/petrol, 8 : 92) gave (5)
1
as a colourless oil (43%), [α]D –52.7° (lit.[4] –53.2°). H NMR (200
MHz, CDCl3) δ 1.37, 1.55, 2 × s, CMe2; 3.22–3.97, m, 7H, H 5,
aglycon; 4.37–4.43, m, H 4; 4.56–4.79, m, 6H, H 2,3,CH2Ph; 5.15, s,
H 1; 7.37, 10H, Ph.
(R)-2,3-Dibenzyloxypropyl 5-Chloro-5-deoxy-2,3-O-isopropylidene-β-
D-riboside (7)
Camphorsulfonic acid (170 mg) was added to a mixture of methyl
5-chloro-5-deoxy-2,3-O-isopropylidene-β-D-riboside (190 mg, 0.85
mmol) and the dibenzylglycerol (4) (230 mg, 0.85 mmol), and the
mixture was stirred at 80°C and 25 torr (20 h). Usual workup (EtOAc)
and flash chromatography (EtOAc/petrol, 3 : 22) yielded the chloride
(7) as a colourless oil (150 mg, 40%), [α]D –55.3° (Found: C, 64.9; H,
type reductive elimination[6,7] of the bromide (5) (Scheme 2).
The use of the chloride (7) did little to improve matters and
we accepted the reasonable yield of the arsine (6).
Any attempt to remove the benzyl protecting groups from
(6) by hydrogenolysis failed, probably owing to poisoning of
the palladium catalyst by the arsine residue in the substrate.
Nevertheless, treatment of (6) with sodium in liquid
ammonia gave the diol (8).
1
6.7%. C25H31ClO6 requires C, 64.9; H, 6.8%). H NMR (300 MHz,
CDCl3) δ 1.39, 1.53, 2 × s, CMe2; 3.41–3.60, m, 5H, aglycon; 3.70–
3.82, m, 2H, AB part of ABX pattern, H 5; 4.31–4.38, m, H 4; 4.56–
4.75, m, 6H, H 2,3,CH2Ph; 5.11, s, H 1; 7.27–7.42, 10H, Ph. 13C NMR
(75.5 MHz, CDCl3) δ 24.9, 26.4, CMe2; 44.2, C 5; 67.3, 69.7, 2C,
aglycon CH2; 72.3, 73.5, 2C, CH2Ph; 76.7, aglycon CH; 82.2, 85.1,
86.7, C 2,3,4; 108.4, C 1; 112.6, CMe2; 127.7, 127.8, 128.4, 138.1,
138.4, Ph.
An attractive conclusion now was to treat the diol (8) with
iodine in methanol, thereby removing the isopropylidene
group[8] and converting the arsine, after hydrolytic workup,
into the arsine oxide.[9] Although the desired natural product
(1) was formed, there was apparently some leakage to the
methyl D-ribofuranoside (9). Therefore, the diol (8) was
easily oxidized to the arsine oxide (10) and a subsequent and
careful treatment with aqueous trifluoroacetic acid gave the
natural product (1).[3]
(R)-2,3-Dibenzyloxypropyl 5-Deoxy-5-dimethylarsino-2,3-O-isopro-
pylidene-β-D-riboside (6)
(i) Iododimethylarsine (370 mg, 1.6 mmol) was added to sodium
(75 mg, 3.0 mmol) in dry THF (7 mL) with stirring under argon.[3]
After the addition, the yellow mixture was stirred (30 min, 20°C) and
then allowed to stand under argon (1 h). The supernatant liquid was
added to the bromide (5) (200 mg, 0.4 mmol) and the mixture was
stirred (1 h). The reaction was quenched (H2O) and, after usual workup
(EtOAc) and flash chromatography (EtOAc/petrol/triethylamine,
6 : 93 : 1), the arsine (6) was obtained as a colourless oil (87 mg, 41%),
[α]D –24.8° (Found: C, 61.0; H, 7.1%. C27H37AsO6 requires C, 60.9; H,
Experimental
1
7.0%). H NMR (300 MHz, CDCl3) δ 0.98, 0.99, 2 × s, AsMe2; 1.38,
1.47, 2 × s, CMe2; 1.59–1.89, m, 2H, AB part of ABX pattern, H 5;
3.51–3.98, m, 5H, aglycon; 4.28–4.33, m, H 4; 4.51–4.78, m, 6H, H 2,3,
CH2Ph; 5.09, s, H 1; 7.33–7.39, m, 10H, Ph. 13C NMR (75.5 MHz,
CDCl3) δ 9.17, 9.23, AsMe2; 25.0, 26.5, CMe2; 34.7, C 5; 67.2, 70.3,
2C, aglycon CH2; 72.2, 73.4, 2C, CH2Ph; 76.9, aglycon CH; 85.0, 85.7,
85.8, C 2,3,4; 108.8, C 1; 112.2, CMe2; 127.6, 127.7, 128.3, 128.4,
138.2, 138.5, Ph.
(ii) Iododimethylarsine (250 mg, 1.1 mmol) and sodium (51 mg,
2.2 mmol) in dry THF (5 mL) were used, as described above, to react
with the chloride (7) (125 mg, 0.3 mmol), and the mixture was stirred
(3 h). The reaction was quenched (H2O) and, after usual workup
(EtOAc) and flash chromatography (EtOAc/petrol, 1 : 10), the arsine (6)
was obtained as a colourless oil (70 mg, 49%). The 1H NMR (300 MHz)
spectrum for the compound was in good agreement with that obtained
from the bromide (5).
Experimental details have been given previously.[10]
Methyl 2,3-O-Isopropylidene-5-O-p-toluenesulfonyl-β-D-riboside (2)
The tosylate (2) was prepared following a two-step procedure as
described by Lerner,[5] with the exception that the reaction was
quenched with saturated NaHCO3, not water as reported, to give (2) as
colourless crystals (40%), [α]D –49.6° (lit.[11] –50.0°), m.p. 81–82°C
(EtOH: lit.[5] 83–84°C).
Methyl 5-Bromo-5-deoxy-2,3-O-isopropylidene-β-D-riboside (3)
A mixture of the tosylate (2) (8.7 g, 24 mmol) and tetrabutyl-
ammonium bromide (23 g, 72 mmol) in DMF was heated under reflux
(24 h). Concentration of the mixture and usual workup (Et2O) gave a
pale yellow oil. Rapid suction-filtration chromatography (EtOAc/
petrol, 17 : 83) yielded the bromide (3) as a colourless oil (5.6 g, 89%),
[α]D –79° (lit.[4] –79.6°). 1H NMR (200 MHz, CDCl3) δ 1.31, 1.48, 2 ×
s, CMe2; 3.25–3.42, m, 2H, H 5; 3.32, s, OCH3; 4.25–4.35, m, H 4;
4.53–4.58, m, H 3; 4.65–4.70, m, H 2; 4.99, s, H 1.
In both the above preparations of the arsine (6), the presence of the
alcohol (4) in the crude reaction mixture was obvious (TLC).
(R)-2,3-Dihydroxypropyl 5-Deoxy-5-dimethylarsino-2,3-O-isopropyl-
idene-β-D-riboside (8)
Methyl 5-Chloro-5-deoxy-2,3-O-isopropylidene-β-D-riboside
Sodium (115 mg, 5.0 mmol) was added to a solution of the arsine
(6) (370 mg, 0.70 mmol) in THF and liquid ammonia (11 mL, 3 : 8) at
–78°C. The mixture was left (1 h) and then quenched with solid NH4Cl.
The ammonia was left to evaporate and, after the usual workup
(CH2Cl2) and flash chromatography (EtOAc/petrol, 1 : 1), the diol (8)
was obtained as a yellow oil (172 mg, 71%), [α]D –20.1° (Found: C,
44.2; H, 7.2%. C13H25AsO6 requires C, 44.1; H, 7.0%). 1H NMR (300
MHz, CDCl3) δ 0.98, 0.99, 2 × s, AsMe2; 1.38, 1.47, 2 × s, CMe2; 1.55–
1.89, m, 2H, AB part of an ABX pattern, H 5; 3.03, 2H, OH; 3.50–3.91,
A mixture of the tosylate (2) (7.3 g, 20 mmol) and tetrabutylammonium
chloride (16 g, 59 mmol) in DMF was heated at reflux (36 h).
Concentration of the mixture and usual workup (Et2O) gave a pale
yellow oil. Rapid-suction filtration column chromatography (EtOAc/
petrol, 17 : 83) yielded the title chloride as a pale yellow oil (3.1 g,
1
71%), [α]D –92° (lit.[12] –93°). H NMR (200 MHz, CDCl3) δ 1.28,
1.43, 2 × s, CMe2; 3.29, s, OCH3; 3.38–3.59, m, 2H, H 5; 4.25–4.29, m,
H 4; 4.55, d, J2,3 6.2 Hz, H 3; 4.64, d, H 2; 4.99, s, H 1.

An Improved Synthesis of a Common Marine Arsenical
183
m, 5H, aglycon; 4.23–4.26, m, H 4; 4.55–4.69, m, H 2,3; 5.06, s, H 1.
13C NMR (75.5 MHz, CDCl3) δ 9.63, 9.77, AsMe2; 25.5, 27.0, CMe2;
35.1, C 5; 64.3, 70.9, 2C, aglycon CH2; 71.2, aglycon CH; 85.5, 86.1,
86.3, C 2,3,4; 110.0, C 1; 113.0, CMe2.
m, aglycon CH2; 3.44, 3.58, AB part of ABX pattern, aglycon CH2;
3.72–3.76, m, aglycon CH; 3.97, d, J2,3 3.8 Hz, H 2; 4.08–4.14, m,
H 3,4; 4.85, s, H 1. 13C NMR (75.5 MHz, D2O) δ 16.4, 16.8, AsMe2;
38.4, C 5; 64.7, 71.2, 2C, aglycon CH2; 72.6, aglycon CH; 76.5, 78.1,
79.2, C 2,3,4; 109.7, C 1.
(R)-2,3-Dihydroxypropyl 5-Deoxy-5-dimethylarsinyl-2,3-O-isopropyl-
idene-β-D-riboside (10)
References
A solution of the arsine (8) (57 mg, 0.16 mmol) in THF (3 mL) was
treated with 30% hydrogen peroxide (0.06 mL, 0.5 mmol) and the
solution stirred (30 min). Evaporation and flash chromatography of the
residue (MeOH/CHCl3, 2 : 3) yielded the arsine oxide (10) as a
colourless oil (49 mg, 82%), [α]D –0.18° (lit.[13] –0.2°). 1H NMR
(300 MHz, CDCl3) δ 1.26, 1.43, 2 × s, CMe2; 1.66, 1.78, 2 × s, AsMe2;
2.21, dd, J5,5 14.1, J4,5 3.9 Hz, H 5; 2.82, dd, J4,5 12.9 Hz, H 5; 3.39–
3.49, m, aglycon CH2; 3.58, 3.91, AB part of ABX pattern, aglycon
CH2; 3.74–3.76, m, aglycon CH; 4.40–4.72, m, H 2,3,4; 5.06, s, H 1.
13C NMR (75.5 MHz, CDCl3) δ 15.03, 15.07, AsMe2; 24.8, 26.3,
CMe2; 38.0, C 5; 64.3, 69.3, 2C, aglycon CH2; 71.0, aglycon CH; 82.0,
85.0, 85.2, C 2,3,4; 108.9, C 1; 112.7, CMe2.
[1] J. S. Edmonds, K. A. Francesconi, R. V. Stick, Nat. Prod. Rep.
1993, 10, 421.
[2] J. R. Cannon, J. S. Edmonds, K. A. Francesconi, C. L. Raston,
J. B. Saunders, B. W. Skelton, A. H. White, Aust. J. Chem. 1981,
34, 787.
[3] D. P. McAdam, A. M. A. Perera, R. V. Stick, Aust. J. Chem. 1987,
40, 1901.
[4] K. J. Irgolic, J. Liu, D. H. O’Brien, Appl. Organomet. Chem.
1996, 10, 1.
[5] L. M. Lerner, Carbohydr. Res. 1977, 53, 177.
[6] B. Bernet, A. Vasella, Helv. Chim. Acta 1984, 67, 1328.
[7] J. K. Gallos, T. V. Koftis, A. E. Koumbis, J. Chem. Soc., Perkin
Trans. 1 1994, 611.
[8] W. A. Szarek, A. Zamojski, K. N. Tiwari, E. R. Ison, Tetrahedron
Lett. 1986, 27, 3827.
[9] H. Bauer, J. Am. Chem. Soc. 1945, 67, 591.
[10] J. C. McAuliffe, R. V. Stick, Aust. J. Chem. 1997, 50, 193.
[11] F. Sarabia-Garcia, F. Lopez-Herrera, Tetrahedron 1996, 52, 4757.
[12] S. Hanessian, M. M. Ponpipom, P. Lavallee, Carbohydr. Res.
1972, 24, 45.
(R)-2,3-Dihydroxypropyl 5-Deoxy-5-dimethylarsinyl-β-D-riboside (1)
The arsine oxide (10) (50 mg) was dissolved in trifluoroacetic acid/H2O
(9 : 1, 1 mL) and the solution stirred (10 min). Rapid concentration of
the solution under vacuum at room temperature, followed by
evaporation after the addition of water (2 mL) and pyridine (2 mL), gave
a colorless syrup. Passage through a column of resin (Amberlite IRA
400, OH–) gave the arsenical (1) (33 mg, 80%) as a syrup, [α]D –2.5°
(MeOH; lit.[3] –2.6°). 1H NMR (300 MHz, D2O) δ 1.68, 1.70, AsMe2;
2.32, dd, J5,5 13.9, J4,5 9.6 Hz, H 5; 2.48, dd, J4,5 2.9 Hz, H 5; 3.39–3.49,
[13] K. A. Francesconi, J. S. Edmonds, R. V. Stick, J. Chem. Soc.,
Perkin Trans. 1 1992, 1349.