Control of phosphoryl migratory transesterifications allows regioselecive access to sugar phosphates

Article abstract of DOI:10.1021/ol303271z

Phosphate esters in polyhydroxylated systems are normally blighted by uncontrolled migration under a variety of reaction conditions. Cesium fluoride is demonstrated as a reagent to control migration of primary phosphates during transesterifications. This allows easy exchange of phosphoryl protecting groups enabling enhanced synthetic strategic flexibility and regioselective phosphate installation. Mechanistic analysis suggests that a fluoride-induced extended solvent sphere modulates steric bulk at phosphorus to favor the primary position.

Full text of DOI:10.1021/ol303271z

ORGANIC
LETTERS
2013
Vol. 15, No. 2
346–349
Control of Phosphoryl Migratory
Transesterifications Allows Regioselecive
Access to Sugar Phosphates
Mitul K. Patel and Benjamin G. Davis*
Department of Chemistry, Chemistry Research Laboratory, University of Oxford,
Mansfield Road, Oxford, OX1 3TA, U.K.
Ben.Davis@chem.ox.ac.uk
Received November 28, 2012
ABSTRACT
Phosphate esters in polyhydroxylated systems are normally blighted by uncontrolled migration under a variety of reaction conditions. Cesium
fluoride is demonstrated as a reagent to control migration of primary phosphates during transesterifications. This allows easy exchange of
phosphoryl protecting groups enabling enhanced synthetic strategic flexibility and regioselective phosphate installation. Mechanistic analysis
suggests that a fluoride-induced extended solvent sphere modulates steric bulk at phosphorus to favor the primary position.
Phosphorylated carbohydrates are prevalent in nature.1
Their central role in metabolism and signaling make them
an important target for the synthetic chemist. However,
carbohydrates (and inositols) have an inherent structural
complexity that makes their selective phosphorylation
difficult.2 Installation of phosphoryl groups usually re-
quires multistep sequences that use protecting groups,
which results in poor overall yields. Since the high polarity
of unprotected phosphate groups can make conventional
purification difficult,3 they are typically also protected to
simplify handling. However, this strategy can cause further
problems with orthogonality to other functional groups.2,4
In particular, the migration and cyclization of the
phosphate group is a common problem requiring special
attention.2 Neighboring hydroxyls, especially those in a
vicinal position, are well placed to attack the phosphorus
center resulting in various side reactions. While protecting
groups can lessen this problem, later deprotections may
still expose neighboring hydroxyls to conditions favoring
phosphate cyclization and migration, resulting in mixtures
of polar compounds that can be extremely difficult to
purify. For some compounds, catalysts5 can offer elegant,
selective phosphorylations. However, for the majority of
carbohydrates and inositols, phosphate migrations remain
a major problem that can only be alleviated by careful
planning of the synthesis.
We sought conditions to access cyclic phosphate inter-
mediates, the ‘opening’ of which could parallel the well-
exploited 4,6-O-benzylidene ring openings where the choice
of reagent allows excellent regioselective control.6 Under
acid catalysis, electronic control dictates protonation leading
(1) Bowler, M. W.; Cliff, M. J.; Waltho, J. P.; Blackburn, G. M. New
J. Chem. 2010, 34, 784–794.
(5) (a) Nakajima, H.; Kondo, H.; Tsurutani, R.; Dombou, M.;
Tomioka, I.; Tomita, K. Enzymes in Carbohydrate Synthesis; American
Chemical Society: Washington, DC, 1991; pp 111À120. (b) Chenault,
H. K.; Mandes, R. F.; Hornberger, K. R. J. Org. Chem. 1997, 62,
331–336. (c) Sculimbrene, B. R.; Miller, S. J. J. Am. Chem. Soc. 2001,
123, 10125–10126.
(6) (a) Preparative Carbohydrate Chemistry; Hanessian, S., Ed.; Marcel
Dekker: New York, 1997; pp 53À68. (b) Greene, T. W.; Wuts, P. G. M.
Protective Groups in Organic Synthesis; 3rd ed.; John Wiley & Sons: New
York, 1999; pp 217À224. (c) Shie, C.-R.; Tzeng, Z.-H.; Kulkarni, S. S.;
Uang, B.-J.; Hsu, C.-Y.; Hung, S.-C. Angew. Chem., Int. Ed. 2005, 44,
1665–1668.
(2) (a) Billington, D. C. Chem. Soc. Rev. 1989, 18, 83–122.
(b) Billington, D. C. The Inositol Phosphates: Chemical Synthesis and
Biological Significance; VCH: 1993. (c) Sureshan, K. M.; Shashidhar, M. S.;
Praveen, T.; Das, T. Chem. Rev. 2003, 103, 4477–4504. (d) Chenault,
H. K.; Mandes, R. F.; Hornberger, K. R. J. Org. Chem. 1997, 62, 331–
336. (e) Ahmadibeni, Y.; Parang, K. Org. Lett. 2005, 7, 1955–1958.
(f) Ruhela, D.; Vishwakarma, R. A. J. Org. Chem. 2003, 68, 4446–4456.
(3) Feurle, J.; Jomaa, H.; Wilhelm, M.; Gutsche, B.; Herderich, M.
J. Chromatgr., A 1998, 803, 111–119.
(4) Greene, T. W.; Wuts, P. G. M. Protective Groups in Organic
Synthesis, 3rd ed.; John Wiley & Sons: New York, USA, 1999; pp 660À700.
r
10.1021/ol303271z
Published on Web 01/03/2013
2013 American Chemical Society

to the 6-ether, but bulky Lewis acids furnish the 4-ether, via
sterically more favorable 6-O activation. For phosphates,
the steric requirements of the implicated pentacoordinate
intermediates might allow similarly good electronic or sol-
vent-mediated control.
Table 1. Attempted Direct ÀOPh f ÀOH Deprotectiona
Scheme 1. Strategies for Regioselective Sugar Phosphorylation
entry
Rb
conditions
time
3/%
4/%
5/%
1
2
3
a
a
a
A
A
B
10 min
14 h
88
93
49
0
9
0
0
0
12 h
31
a Conditions: (A) NH4OH, H2O, tBuOH; (B) 10 equiv of CsF, H2O,
80 °C. b For the structures of sugar structures used in this paper please
refer to Scheme 1.
Scheme 2. ArO f OH Exchange during Hydrolysis Leads to
Irreversible Entry into a Charged Cyclic Phosphate Manifold
This suggested a phosphorylation strategy with a
bis-protected phosphoryl unit followed by cyclization
and controlled opening (Scheme 1, path a). If com-
bined with transesterification this could allow ready
manipulation of phosphoryl substituents. Such con-
trol of migration would also allow potential for re-
gioconvergence from an initially poorly regioselective
phosphorylation.
Trehalose, the R,R-1,1-nonreducing dimer of D-glucose,
has a complex polyhydroxylated structure that makes it a
particularly excellent test substrate. Its 6-O-phosphate has
been previously synthesized using chemical (POCl3,7 so-
dium phosphate8) or enzymatic methods.9 We have pre-
viously shown that diphenyl chlorophosphate can be used
to regioselectively install phenyl-protected phosphate
groups onto trehalose.10
Direct hydrolytic PhO f OH exchange in 2a was tested
first (Table 1), initially with basic hydrolysis.11 On expo-
sure of 2a to ammonium hydroxide, cyclic phosphate 3a
was produced in 88% yield within 10 min (entry 1); with
short reaction times, 9% of the partially deprotected, but
migrated, 4-O-phosphate 5a was also isolated. None of the
directly deprotected 6-O-phosphate was observed. Over
longer reaction times, 5 was not isolated and 3a was the
sole product (entry 2). Importantly, use of CsF12 (entry 3)
hinted at the potential for migration that might be
exploited later. While phosphate cyclization remained the
predominant pathway (3 in 49% yield), these conditions
also gave 4a, which crucially had retained the phosphoryl
group at the O-6-position.
These results may be rationalized via an entropically
favored intramolecular cyclization giving cyclic ester 6
(Scheme 2).13 Hydroxyl ion attack gives a pentacoordinate
intermediate (PI2) leading to three possible products
(7) Rønnow, E. C. L.; Meldal, M.; Bock, K. Carbohydr. Res. 1994,
260, 323–328.
(8) Tarelli, E.; Wheeler, S. F. Carbohydr. Res. 1994, 261, 25–36.
(9) (a) Doi, J.; Yokoigawa, K.; Isobe, Y.; Kawai, H. Biosci., Bio-
technol., Biochem. 1998, 62, 735–739. (b) Errey, J. C.; Lee, S. S.; Gibson,
R. P.; Fleites, C. M.; Barry, C. S.; Jung, P. M. J.; Sullivan, A. C. O.;
Davis, B. G.; Davies, G. J. Angew. Chem., Int. Ed. 2010, 49, 1234–1237.
(10) Patel, M. K.; Davis, B. G. Org. Biomol. Chem. 2010, 8, 4232–
4235.
(12) Ogilvie, K. K.; Beaucage, S. L. Nucleic Acids Res. 1979, 7, 805–
823.
(13) (a) Fordham, W. D.; Wang, J. H. J. Am. Chem. Soc. 1967, 89,
4197–4203. (b) Buchwald, S. L.; Pliura, D. H.; Knowles, J. R. J. Am.
Chem. Soc. 1984, 108, 4916–4922.
(11) (a) Plourde, R.; d’Alarco, M. Tetrahedron Lett. 1990, 31, 2693–
2696. (b) Nanteuil, G. D.; Benoist, A.; Remond, G.; Descombes, J.;
Barou, V.; Verbeuren, T. J. Tetrahedron Lett. 1995, 36, 1435–1438.
(14) (a) Westheimer, F. H. Science 1987, 235, 1173–1178.
(b) Schroeder, G. K.; Lad, C.; Wyman, P.; Williams, N. H.; Wolfenden,
R. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 4052–4055.
Org. Lett., Vol. 15, No. 2, 2013
347

transesterification without concomitant deprotection al-
beit with substantial migration from O-6 to O-4 (63% r.r.
O-4 9a). In the presenceof CsF (entries 2, 3), O-4 migration
was reduced in a manner dependent on the amount of CsF
used, up to 10 equiv, which gave 65% r.r. of O-6 8a.
Importantly, these results demonstrated entry into a mani-
fold primed for migration and suggested that regioselec-
tivity could be modulated.
Table 2. Conditions for Controlled Transesterificationa
A striking feature of these reactions was the reduced
propensity for O-4 migration in the presence of CsF. The
reasons for these altered migratory properties were investi-
gated (Table 2). First, other fluoride salts with increasing
organic solubility (entries 4f3f5) reduced migration to O-
4. Similarly, increasing solvent proticity15 (entries
6f7f3f8) also decreased migration. Third, the reaction
was found to be essentially insensitive to pH (entries 9 and
10) and temperature (entry 11). Finally, with the different
counteranion ClÀ (entries 12, 13) 3a was formed exclusively,
presumably through Krapcho-type demethylation.16 Nota-
bly, fluorophosphates (potentially isolable compounds17)
were not detected in any reactions, ruling out direct fluoride
participation (see Supporting Information (SI) for control
reactions). It is worth noting that, over longer reaction times
(entry 14), further migration to O-2 and O-3 was observed.
These studies revealed that we were able to tune regio-
selective ratios from 68% in favor of one regioisomer, 9, to
100% of the other, 8, using appropriate conditions. To test
the scope of this strategy, we applied CsF-mediated trans-
esterification to other biologically relevant carbohydrates
(Table 3). The required starting diphenylphosphates were
synthesized in an analogous manner to that used for 2a.10
Pleasingly, subsequent transesterifications for 2bÀd all
proved successful, indeed more so than for 2a. Thus,
glucoside 2b gave the O-6-phosphate 8b as the major
product in 72% yield. Similarly, mannoside 2c furnished
the desired O-6-phosphate 8c in an excellent 84% yield.
For galactoside 2d, the initial cyclization was slower
(presumably due to the more hindered axial position of
OH-4) and the reaction required 40 h for completion.
Migratory ring opening exclusively yielded the O-6-phos-
phate 8d in 96% yield. Riboside 2e was also successful (8e,
96%); this proceeded via a mixed rather than cyclic ester
intermediate (see SI for details). Deprotections (Table 3) of
these dimethylphosphates 8aÀe were achieved with
TMSI18 in excellent yields of greater than 85% for all.
The protocol we describe allows control of regiochem-
istry during phosphoryl migrations, which occur via a
cyclic intermediate13 with the outcome dictated by a
yieldb/
%
rrc
8:9/% products
other
R
conditions
1
a NEt3 (1 equiv), MeOH, 3 h
54 37 63
7a
(22%)
À
2
3
4
5
a CsF (2.5 equiv), MeOH
a CsF (10 equiv), MeOH
a KF (10 equiv), MeOH
a TBAF (10 equiv), MeOH
91 58 42
94 65 35
93 61 39
51 85 15
À
À
3a
(24%)
3a
6
7
a CsF (10 equiv), DMF/MeOH
59 55 45
62 59 41
31 100d 0d
(1:1)
(22%)
3a
a CsF (10 equiv), tBuOH/MeOH
(1:1)
8d a CsF (10 equiv), H2O/tBuOHd
(1:3)
(11%)
3a
(49%)
À
9
a CsF, NH3, MeOH, pH 13
92 64 36
93 68 34
90 63 37
10 a CsF, HCl, MeOH, pH 2
11 a CsF (10 equiv), MeOH, 50 °C
12 a CsCl (10 equiv), MeOH
À
À
0
0
0
3a
(82%)
3a
13 a MgCl2 (10 equiv), MeOH
0
0
0
(85%)
e
14 a Na2CO3 (2.5 equiv), DCM/MeOH 34 32 68
(4:1), 80 h
15f a Ionic liquid/MeOH (1:1), 18 h,
NEt3 (4 equiv)f
57 100
0
À
a Reactions were performed under reflux for 16 h unless otherwise
noted. b Combined yield. c Regioselectiviy ratio normalized as %, de-
termined by 1H NMR of crude mixture and HPLC. d From Table 1
giving 4 and 5. e Minor O-2,3 migration products. f Conducted in
dimethylimidazolium dimethylphosphate and per-O-acetylated to aid
purification.
(3, 4, or 5). Of these, the diester 3 exists in deprotonated
form, and the repulsive effects of this charge are the source
of extreme hydrolytic stability.1,14 Thus, nucleophilic attack
on 3 is disfavored and its formation is irreversible, while 4
and 5 are still inclined to cyclization and so are formed
reversibly; over time the cyclic phosphate 3 is the major
product.
We reasoned that OPh f OR transesterification would
produce a cyclic phosphate intermediate that would not be
charged and could therefore be formed reversibly. Hence,
we tested reactions using MeOH as solvent (Table 2).
Initial investigations under strongly basic conditions re-
sulted in BAl2/SN2-type solvolysis of exchanged intermedi-
ates leading to monodeprotection and entry into the
unwanted charged cyclic phosphate (3) manifold. Under
milder conditions methyl loss did not occur; we were
pleased to see that uncharged cyclic phosphate 7a
was quickly formed and, over time, reacted further to
give 8a and 9a. Thus, triethylamine (entry 1) allowed
(15) (a) Kamlet, M. J.; Taft, R. W. J. Am. Chem. Soc. 1976, 98, 377–
383. (b) Taft, R. W.; Kamlet, M. J. J. Am. Chem. Soc. 1976, 98, 2886–
2894.
(16) Krapcho, A. P.; Glynn, G. A.; Grenon, B. J. Tetrahedron Lett.
1967, 8, 215–217.
(17) (a) Gupta, A. K.; Acharya, J. D.; Pardasani, D.; Dubey, D. K.
Tetrahedron Lett. 2008, 49, 2232–2235. (b) Furegati, S.; Ganci, W.;
€
Gorla, F.; Ringeisen, U.; Ruedi, P. Helv. Chim. Acta 2004, 87, 2629–
2661 and references therein for preparation, isolation, and even reaction
in alcoholic solvents of fluorophosphates.
(18) (a) Blackburn, G. M.; Ingleson, D. J. Chem. Soc., Chem.
Commun. 1978, 870–871. (b) Olah, G. A.; Narang, S. C. Tetrahedron
1982, 38, 2225–2277.
348
Org. Lett., Vol. 15, No. 2, 2013

Table 3. Transformations to 8, 9, and 10a
Scheme 3
starting
time/
h
yield
9/%b
yield
8/%b
yield
10/%c
entry
2
1
2
3
4
5
a
b
c
16
16
16
40
16
33
18
5
61
72
84
96
96
87
88
86
92
87d
pronounced in D-galactosides and is consistent with the
trend seen for 2df8d (Table 3, entry 4). A similar unusual
role for CsF in reaction selectivity has also been noted
previously in Michael additions.29 Based on this analysis,
a reaction was performed in an ionic liquid (Table 2,
entry 15) that possesses a highly ordered long-range
structure30 and therefore similarly large solvation spheres.
We chose a dimethylphosphate derived ionic liquid to
prevent interference from the anion.31 Under reflux in a
1:1 ratio with methanol and 4 equiv of NEt3 as base, the
reaction gave a highly regioselective ratio of >98% for 8a,
consistent with the solvent sphere hypothesis.
In this paper, we have demonstrated methodology that
can offer regiocontrol of phosphate migrations. This now
enables phosphate transesterification and substituent ex-
change with all of the associated synthetic strategic ad-
vantages. Crucially, exchange of protecting groups allows
the introduction of phosphates through a bulky phosphor-
ylating agent prior to alteration to something with ortho-
gonality that allows removal under conditions compatible
with other functionality. We believe that this approach will
prove useful for installation of the key sugar modification
of phosphorylation and may prove applicable to other
vital modifications such as sulfation.
d
e
0
0
a Conditions: (i) CsF (10 equiv), MeOH, reflux; (ii) TMSI, dioxane,
15 min, rt. b Isolated from step (i). c Isolated from step (ii). d C-1 epimer
seen.
pentacoordinate P center (Scheme 3).19 A kinetic prefer-
enceformigration toOH-4(cf. Table2, entry14), observed
previously,20,21 can be explained by apicophilicity of the
most electronegative, primary OH-6 in 11 (see SI for
discussion);20,22,23 extrusion of the longer PÀO(6) bond
would be more facile giving 9. However, sterics may over-
ride electronics.22 FÀ ions interact strongly with hydroxyls24
and create H-bonding networks within solvent25 (described
as “structure markers” that increase solvent ordering26)
particularly in methanolic solutions of CsF.27,28 This sug-
gests that F-mediated H-bonding may create a large solva-
tion sphere around the charged phosphoryl group (12,
Scheme 3).
The increased bulk at OH-4 would then overcome the
apicophilicity of OH-6. This effect would be more
(19) (a) Westheimer, F. H. Acc. Chem. Res. 1968, 1, 70–78.
(b) Dennis, E. A.; Westheimer, F. H. J. Am. Chem. Soc. 1966, 88,
3432–3433. (c) Gillespie, P.; Ramirez, F.; Ugi, I.; Marquarding, D.
Angew. Chem., Int. Ed. 1973, i, 91–119.
(20) Hall, C. R.; Inch, T. D. Tetrahedron 1980, 36, 2059–2095.
(21) (a) Cooper, D. B.; Harrison, J. M.; Inch, T. D.; Lewis, G. J.
J. Chem. Soc., Perkin Trans. 1 1974, 1058–1068. (b) Szabo, P.; Szabo, L.
J. Chem. Soc. 1960, 3758–3762. (c) Inch, T. D.; Lewis, G. J. J. Chem.
Soc., Chem. Commun. 1973, 310–311.
(22) Trippett, S. Phosphorus Sulfur 1976, 1, 89–98.
(23) (a) Mikhailov, S. N.; Oivanen, M.; Oksman, P.; Lonnberg, H.
J. Org. Chem. 1992, 57, 4122–4126. (b) Oivanen, M.; Schnell, R.;
Pfeiderer, W.; Lonnberg, H. J. Org. Chem. 1991, 56, 3623–3628.
(c) Jarvinen, P.; Oivanen, M.; Lonnberg, H. J. Org. Chem. 1991, 56,
5396–5404.
Supporting Information Available. Experimental de-
tails. This material is available free of charge via the
Internet at http://pubs.acs.org.
Acknowledgment. We thank the International AIDS
Vaccine Initiative for funding. B.G.D. is a Royal Society
Wolfson Research Merit Award recipient.
€
€
€
(29) Ishikawa, T.; Oku, Y.; Kotake, K.; Ishii, H. J. Org. Chem. 1996,
61, 6484–6485.
ꢀ
ꢀ
(24) Gonzalez, R.; Ortiz, P.; Reguera, E.; Bertran, J. F. J. Fluorine
(30) Hardacre, C.; Holbrey, J.; Nieuwenhuyzen, M.; Youngs,
T. G. A. Acc. Chem. Res. 2007, 40, 1146–1155.
(31) If substitution were to occur, 8a would be formed (rather than
Chem. 2001, 110, 5–10.
(25) (a) Hefter, G. T. Pure Appl. Chem. 1991, 63, 1749–1758.
(b) Hefter, G. T.; McLay, P. J. J. Sol. Chem. 1988, 17, 535–546.
(26) Frank, H. S.; Wen, W. Y. Disc. Faraday Soc. 1957, 24, 133–140.
(27) Marcus, Y.; Hefter, G. Chem. Rev. 2006, 106, 4585–4621.
(28) Hefter, G. T.; Salomon, M. J. Sol. Chem. 1996, 25, 541–553.
side products that would result with other ionic liquids).
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 2, 2013
349