Cyclopropenium cation promoted dehydrative glycosylations using 2-deoxy- and 2,6-dideoxy-sugar donors

Article abstract of DOI:10.1021/ol200726v

Dehydrative glycosylation reactions using 2-deoxy- and 2,6-dideoxy-sugar donors promoted by a combination of 3,3-dichloro-1,2-diphenylcyclopropene and tetrabutylammonium iodide (TBAI) are described. The reactions are α-selective and proceed under mild con

Full text of DOI:10.1021/ol200726v

ORGANIC
LETTERS
2011
Vol. 13, No. 11
2814–2817
Cyclopropenium Cation Promoted
Dehydrative Glycosylations Using
2-Deoxy- and 2,6-Dideoxy-Sugar Donors
Jason M. Nogueira, Son Hong Nguyen, and Clay S. Bennett*
Department of Chemistry, Tufts University, Medford, Massachusetts 02155,
United States
clay.bennett@tufts.edu
Received March 17, 2011
ABSTRACT
Dehydrative glycosylation reactions using 2-deoxy- and 2,6-dideoxy-sugar donors promoted by a combination of 3,3-dichloro-1,2-
diphenylcyclopropene and tetrabutylammonium iodide (TBAI) are described. The reactions are R-selective and proceed under mild conditions
at room temperature without the need for special dehydrating agents. The reaction is shown to be effective with a number of glycosyl acceptors,
including those possessing acid and base sensitive functionality.
In many natural products the presence of deoxy-sugars
is often essential for biological activity.1 Additionally,
changing these sugars through glycorandomization can
dramatically alter the biological profile of a natural
product.2 As a consequence, there has been a significant
amount of effort directed at developing efficient methods
to synthesize deoxy-sugar containing oligosaccharides
over the past few decades.3,4 The most conceptually
straightforward of these approaches rely on so-called
“direct” glycosylation reactions which utilize fully functio-
nalized glycosyl donors. While a number of elegant direct
glycosylation reactions using deoxy-sugar donors have
been developed, most require the use of unstable activated
deoxy-sugar donors, rendering them very technically chal-
lenging for nonspecialists.5 In this communication, we
describe an operationally simple method for dehydrative
glycosylation reactions using deoxy-sugar donors that is
tolerant of acid and base sensitive functional groups.
Dehydrative glycosylation reactions use promoters to
activate lactols in situ. This approach has advantages over
other methods in that it does not require the synthesis and
isolation of highly unstable activated sugar donors.6 De-
spite the methods advantages, this approach has found
limited application in the synthesis of deoxy-sugars,7 in
part because homocoupling reactions can be problematic.8
While Mitsunobu glycosylations have been reported using
deoxy-sugar donors,9 they are only effective using acidic
(1) Kirschning, A.; Bechthold, A. F.-W. Top. Curr. Chem. 1997,
188, 1.
(5) For the use of allyl glycosides as donors, see: Park, J.; Boltje, T. J.;
Boons, G.-J. Org. Lett. 2008, 10, 4367.
(6) Nguyen, H. M.; Chen, Y.; Duron, S. G.; Gin, D. Y. J. Am. Chem.
€
(2) (a) Dong, S. D.; Oberthur, M.; Losey, H. C.; Anderson, J. W.;
Eggert, U. S.; Peczuh, M. W.; Walsh, C. T.; Kahne, D. J. Am. Chem. Soc.
ꢀ
~
2002, 124, 9064. (b) Menendez, N.; Nur-E-Alam, M.; Brana, A.; Rohr,
Soc. 2001, 123, 8766.
ꢀ
J.; Salas, J. A.; Mendez, C. Mol. Microbiol. 2004, 53, 903. (c) Langenhan,
J. M.; Peters, N. R.; Guzei, I. A.; Hoffman, M.; Thorson, J. S. Proc. Natl.
Acad. Sci. U.S.A. 2005, 102, 12305. (d) Iyer, A. K. V.; Zhou, M.; Azad,
N.; Elbaz, H.; Wang, L.; Rogalsky, D. K.; Rojanasakul, Y.; O’Doherty,
G. A.; Langenhan, J. M. ACS Med. Chem. Lett. 2010, 1, 326.
(3) Hou, D.; Lowary, T. L. Carbohydr. Res. 2009, 344, 1911.
(4) For indirect synthesis, see: (a) Roush, W. R.; Gung, B. W.;
Bennett, C. E. Org. Lett. 1999, 1, 891. (b) McDonald, F. E.; Reddy,
K. S.; Diaz, Y. J. Am. Chem. Soc. 2000, 122, 4304. (c) Haukaas, M. H.;
O’Doherty, G. A. Org. Lett. 2002, 4, 1771. (d) Yu, B.; Wang, P. Org.
Lett. 2002, 4, 1919.
(7) (a) Takeuchi, K.; Higuchi, S.; Mukaiyama, T. Chem. Lett. 1997,
26, 969. (b) Toshima, K.; Nagai, H.; Matsumura, S. Synlett 1999, 9,
1420. (c) Haberman, J. M.; Gin, D. Y. Org. Lett. 2003, 5, 2539. (d) Ye,
D.; Liu, W.; Zhang, D.; Feng, E.; Jiang, H.; Liu, H. J. Org. Chem. 2009,
74, 1733.
(8) Rodrı
ꢀ
´
guez, M. A.; Boutureira, O.; Matheu, M. I.; Dı
´
az, Y.;
Castillon, S.; Seeberger, P. H. J. Org. Chem. 2007, 72, 8998.
(9) (a) Smith, A. B., III; Hale, K. J.; Rivero, R. A. Tetrahedron Lett.
1986, 27, 5813. (b) Roush, W. R.; Lin, X.-F. J. Org. Chem. 1991, 56,
5740.
r
10.1021/ol200726v
Published on Web 05/06/2011
2011 American Chemical Society

acceptors such as phenols or carboxylic acids. Attempts to
extend this reaction to aliphatic acceptors requires the use
of toxic mercury salts to drive the reaction.10 Thus, there is
still a need for a general promoter system for dehydrative
glycosylations using deoxy-sugar donors. Here we report
our initial efforts toward achieving this aim using armed
2-deoxy-sugar donors.
Table 1. Reaction Optimizationa
2
TBAI DIPEA
%
entry (equiv) (equiv) (equiv)
t
solvent additive yield R/β
1
1.5
1.5
1.5
1.5
0.76
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1
0
2
2
2
2
2
0
2
2
2
2
2
0
rt
rt
rt
rt
rt
rt
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
À
À
À
À
À
À
À
72 4:1
13 4:1
82 4.3:1
72 4:1
64 4:1
46 6:1
23 2.2:1
13 ND
2
Figure 1. Cyclopropenium/TBAI promoted dehydrative glyco-
sylation with deoxy-sugars.
3
5
4
10
5
5
6
5
7
5
40 °C CH2Cl2
8
5
rt
rt
rt
rt
rt
CH2Cl2 3AMS
We envisioned that a method that would promote the in
situ conversion of a lactol to a highly reactive species, such
as a glycosyl iodide,11 would present an attractive solution
to this problem. For this approach to be practical, the
conversion of the lactol to the halide would have to be
rapid in order to suppress homocoupling. Recently there
has beena report ofa novel system for the rapid conversion
of alcohols into alkyl chlorides based on cyclopropenium
cation activation.12 The operational simplicity of this
method made it appealing to us as a potential promoter
for dehydrative glycosylation reactions. While glycosyl
chlorides are not efficient glycosyl donors in the absence
of heavy metal promoters, it is well established that an
excess of iodide salt can transform the chloride into the
much more reactive glycosyl iodide.13 Thus, conducting
the reaction in the presence of a soluble source of iodide,
such as tetrabutylammonium iodide (TBAI) could poten-
tially lead to the formation of a glycosyl iodide such as 3
(Figure 1). This species could be trapped in situ with an
appropriate acceptor to form a new glycosidic linkage.
Importantly, in the presence of excess halide, glycosyl
iodides exist in equilibrium between the thermodynami-
cally favored R-iodide and the much more reactive β-
iodide.14 Under these conditions, the β-iodide reacts pre-
ferentially leading to formation of R-glycosides via SN2
displacement.15
9
5
CH2Cl2 AW300MS
CH2Cl2 AgOTf
À
ND
55 1.6:1
ND
80 4:1
10
11
12
5
5
MeCN
À
À
5
CH2Cl2 TTBP
a TBAI = tetrabutylammonium iodide, DIPEA = N,N-diisopropy-
lethylamine, TTBP = 2,4,6-tri-tert-butylpyrimidine, AW300MS = acid
washed 3 A molecular sieves, ND = not determined.
Preliminary reaction screening used cholesterol (5) as a
model small molecule acceptor (Table 1). As anticipated
activation of lactol 1 with 2 in the presence of TBAI and N,
N-diisopropylethylamine (DIPEA), followed by in situ
trapping with 5, led to clean formation of glycoconjugate
6 in good yield (72%) and 4:1 R:β selectivity (Table 1, entry
1). A control reaction, run in the absence of TBAI, led to
the formation of the desired product in a much lower yield
(13%, Table 1, entry 2). Investigations into the reaction of
1 with 2 in the absence of an acceptor and TBAI revealed
that the corresponding glycosyl chloride was being formed
in quantitative yields in less than 15 min. As expected, this
species was not reacting efficiently with 5. Increasing the
amount of TBAI in the reaction to 5 equiv led to a further
increase in yield to 82% (Table 1, entry 3), while additional
equivalents of TBAI did not have an impact on the course
of the reaction (Table 1, entry 4). Further attempts to
improve the reaction through varying stoichiometry, ad-
ditives, the use of acetonitrile,16 or heating all had a
detrimental effect on the course of the reaction (Table 1,
entries 5À11). Finally, although substitution of DIPEA
with the hindered base tri-tert-butylpyrimidine (TTBP) led
to the suppression of elimination byproducts, it did not
have additional effects on the reaction (Table 1, entry 12).
(10) Szarek, W. A.; Jarrell, H. C.; Jones, J. K. N. Carbohydr. Res.
1977, 57, C13.
(11) Gervay, J.; Nguyen, T. N.; Hadd, M. J. Carbohydr. Res. 1997,
300, 119.
(12) (a) Kelly, B. D.; Lambert, T. H. J. Am. Chem. Soc. 2009, 131,
13930. (b) Hardee, D. J.; Kovalchuke, L.; Lambert, T. L. J. Am. Chem.
Soc. 2010, 132, 5002. (c) Kelly, B. D.; Lambert, T. H. Org. Lett. 2011, 13,
740.
(13) Kronzer, F. J.; Schuerch, C. Carbohydr. Res. 1974, 34, 71.
(14) Lemieux, R. U.; Hendriks, K. B.; Stick, R. V.; James, K. J. Am.
Chem. Soc. 1975, 97, 4056.
(16) CH2Cl2 and acetonitrile have been established to be the optimal
solvents for conversion of alcohols into halides using 2; see ref 12a.
(15) Hadd, M. J.; Gervay, J. Carbohydr. Res. 1999, 320, 61.
Org. Lett., Vol. 13, No. 11, 2011
2815

This latter base was utilized for studying the scope of the
reaction in anticipation of using base sensitive acceptors.
Of particular note is the loss of selectivity observed when
silver(I) triflate was used to promote the reaction (1.6:1,
Table 1, entry 10).17 The silver salt is expected to promote gly-
cosylation through formation of an oxocarbenium cation.
The change in selectivity between our conditions (Table 1,
entries 3 and 12) and those in entry 10 indicates that the
reactions are proceeding through different mechanisms. This
observation, coupled with the low yields obtained in the
absence of TBAI, lends further support to our proposal that
the reaction proceeds through the formation of a glycosyl
iodide, followed by direct displacement by the nucleophile.
the reaction was more sluggish than other examples, the
product was formed as a single diastereomer. The increase
in selectivity observed with acceptors 11 and 13 could be a
result of the lower reactivity of these species leading them
to react preferentially (11), or exclusively (13) with the
more electrophilic β-anomer of 3 (Figure 1).15
Table 3. Scope of the Reaction with L-2,6-Dideoxy Donor 15
Table 2. Scope of the Reaction with D-2-Deoxy Donor 1
a 3 equiv donor used.
2,6-Dideoxy-sugars, which are commonly found in nat-
ural products, are more reactive than 2-deoxy-sugars
because they contain one less stabilizing oxygen moiety.
This increased reactivity can occasionally lead to problems
with glycosylation reactions.18 This proved not to be a
concern under our conditions, as 2,6-dideoxy donor 15
reacted smoothly with a number of substrates. In all cases
examined the reaction with donor 15 proceeded in higher
yield than donor 1 (Table 3). Notably, the R-product was
still favored despite changing the configuration of the
donor from a D-sugar to an L-sugar.
The increased yield is most notable in the case of the
hindered acceptor 13. Whereas the reaction of 13 with 1
proceededin moderate yield (50% Table2, entry 4), the use
of donor 15 in the reaction provided the desired product in
83% yield (Table 3, entry 4), albeit with slightly lower
selectivity. The reduction in selectivity in this reaction is
presumably a consequence of the higher reactivity of the
a 3 equiv donor used.
Having established conditions for the reaction, we ex-
amined its scope with different acceptors (Table 2). The
reaction using 2-deoxy-sugar 1 proceeded smoothly with
benzylic and primary alcohols (Table 2, entries 1 and 2).
The use of amino acid 11 as an acceptor led to a slight
decrease in the yield; however, the product was formed
with higher R-selectivity (Table 2, entry 3). Finally, the
more hindered acceptor 13 wasless efficient in the reaction,
initially providing the desired product in low (27%À34%)
yield. The yield could be improved by using a larger excess
of donor (3 equiv, Table 2, entry 4). Pleasingly, although
(18) Shan, M.; Sharif, E. U.; O’Doherty, G. A. Angew. Chem., Int.
Ed. 2010, 49, 9492.
(17) Lam, S. N.; Gervay-Hague, J. J. Org. Chem. 2005, 70, 8772.
2816
Org. Lett., Vol. 13, No. 11, 2011

2,6-dideoxy-sugars. Since both the R- and β-iodides gen-
erated from 15 are anticipated to be more electrophilic
than the corresponding iodides generated from 1, it is
possible that the R-iodide can react with 13 to a limited
extent, leading to some β-product.
presence of TBAI this species exists in equilibrium with
>10% of another species, possibly iodide 3r or 3β (see
Supporting Information). These latter two species are
presumably the reactive donors, as chloride 23 is not a
good donor under these conditions (Table 1, entry 2).
Excess iodide in the reaction will promote an equilibrium
between the more stable 3r and the more reactive 3β. The
observed R-selectivity of the products arises through pre-
ferential reaction of the acceptors with 3β via an SN2-like
pathway (path A).21 As discussed above, stereochemical
outcome of the reaction will depend on the reactivity of the
acceptor, with less reactive acceptors affording the higher
levels of selectivity.
In conclusion we have developed a new promoter system
for dehydrative glycosylation reactions. The system toler-
ates acid and base sensitive substrates, and reactions work
especially well with armed 2,6-dideoxy donor species.22
The reaction compares favorably with previously reported
dehydrative glycosylation reactions in that homocoupling
of the donor sugar is not observed. Additionally, unlike
previously reported Mitsunobu coupling approaches to
deoxy-sugars, the system permits glycosylations with ali-
phatic acceptors. Moreover, to the best of our knowledge
this is the first example where cyclopropenium cation
activation has been used to promote intermolecular cou-
pling between complex molecules. Finally, since the reac-
tion uses stable lactols as donors, special precautions are
not necessary to purify donors or carry out the reaction.
Attempts to further improve the scope (including extend-
ing the methodology to disarmed donors) and selectivity of
the reaction are currently under investigation in our
laboratory.
Acknowledgment. Support for this research was pro-
vided in part by a Tufts University Faculty Research Fund
Award. J.M.N. is supported by a GAANN fellowship.
We thank Professor Tristan H. Lambert (Columbia
University) for helpful discussions.
Figure 2. Proposed reaction mechanism.
Note Added after ASAP Publication. Errors in Table 1
were corrected in the version reposted May 27, 2011.
A mechanistic proposal for the reaction is shown in
Figure 2. Prior to addition of the lactol, cyclopropene 2 is
generated in situ from diphenylcyclopropenone. This spe-
cies is in equilibrium with 21, which rapidly converts lactol
1 into R-chloride 23, possibly through the intermediacy of
22.19,20 NMR experiments demonstrate that in the
Supporting Information Available. Experimental pro-
cedures and characterization data. This material is avail-
able free of charge via the Internet at http://pubs.acs.org.
(21) The fact that both D- and L-donors afford R-products is evidence
against the formation of oxocarbenium cation intermediates, which
would be expected to lead to “matched” and “mismatched” products.
See: Spijker, N. M.; van Boeckel, C. A. A. Angew. Chem., Int. Ed. Engl.
1991, 30, 180.
(19) NMR experiments have shown that 1 is converted into 23 in less
than 15 min; see Supporting Information.
(20) One referee raised the possibility that oxalyl chloride/TBAI
could be an effective promoter in the absence of the cyclopropenone.
Preliminary data indicate that this may be the case; however the reaction
is not as efficient as when the dichlorocyclopropene is used as the
promoter.
(22) At present, the reaction does not work well with disarmed
donors, including fully substituted pyranoses.
Org. Lett., Vol. 13, No. 11, 2011
2817