Flow chemistry kinetic studies reveal reaction conditions for ready access to unsymmetrical trehalose analogues

Article abstract of DOI:10.1039/c0ob00226g

Monofunctionalization of trehalose, a widely-found symmetric plant disaccharide, was studied in a microreactor to give valuable kinetic insights that have allowed improvements in desymmetrization yields and the development of a reaction sequence for large scale monofunctionalizations that allow access to probes of trehalose's biological function.

Full text of DOI:10.1039/c0ob00226g

View Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/obc | Organic & Biomolecular Chemistry
Flow chemistry kinetic studies reveal reaction conditions for ready access to
unsymmetrical trehalose analogues†
Mitul K. Patel and Benjamin G. Davis*
Received 7th June 2010, Accepted 1st July 2010
DOI: 10.1039/c0ob00226g
Monofunctionalization of trehalose, a widely-found symmet-
ric plant disaccharide, was studied in a microreactor to give
valuable kinetic insights that have allowed improvements in
desymmetrization yields and the development of a reaction
sequence for large scale monofunctionalizations that allow
access to probes of trehalose’s biological function.
In all the above applications, non-symmetrical trehalose ana-
logues are particularly desirable. The glycosylation of two glucose
units would provide a suitable method, but needs the simultaneous
formation of two a anomeric linkages. Methodologies to accom-
plish this have been reported, such as dehydrative glycosylations,7
and methyl ketoside glycosylations,8 but are limited to the
synthesis of manno and C-1 methyl analogues respectively. In-
tramolecular aglycone delivery9 is more general, but requires
more extensive protecting group manipulations. Alternatively,
direct modification of trehalose, a cheap, commercially available
starting material avoids difficult glycosylations. A drawback to this
strategy is lack of differentiation between the two glucose rings of
symmetric trehalose leading to low yields. Nonetheless, even via
such low yielding routes, several non-symmetrical analogues have
been synthesized demonstrating the viability of this strategy.5,10
In principle, the mono-functionalization of a symmetric polyol
such as 1 can be partitioned into two strategic levels of selectivity:
(a) regioselective modification of (primary) hydroxyls within a
symmetry unit and (b) simultaneous differentiation of the sym-
metry units (here glucose rings). Mode (a) selectivity is, of course,
well established for carbohydrates, where primary hydroxyls can be
regioselectively accessed in the presence of secondary alcohols.11
Efficient monofunctionalizations of symmetric diols (mode (b)
selectivity) have been developed and are now widely used.12,13
Analogous methods for the desymmetrization of trehalose might
usefully combine these two modes. The kinetics of mode (b) can
be modelled as two consecutive first order reactions (eqn (1)–(4)
and Scheme 1). The corresponding rate equations can be solved
to calculate the product distributions over time.
Trehalose (1) (Fig. 1), the a,a-1,1 disaccharide of glucose, is
an important biological molecule. It is synthesized by plants,
bacteria and all invertebrates1 as a protective agent against
stress conditions2 and can serve as a major carbon source for
metabolism. Analogues of trehalose are potential fungicides or
antibiotics by virtue of their ability to inhibit trehalase, an
enzyme critical to the metabolism of this sugar.3 Furthermore,
the human pathogen Mycobacterium tuberculosis incorporates
trehalose into the cell wall as an assortment of glycolipids,
such as mycolic acid esters, sulfolipids, polyacetylated species
and complex lipooligosaccharides.4 This lipid envelope of M.
tuberculosis plays important roles in bacterium-host interactions
and prevents uptake of potential drugs.4 Trehalose analogues that
disrupt cell wall synthesis may therefore act as novel antitubercular
agents.5
Fig. 1 Trehalose (1) and trehalose-6-phosphate (2).
In plants, bacteria and yeast, trehalose-6-phosphate (2) acts as a
signalling molecule controlling carbohydrate metabolism.1,6 While
for bacteria and yeast the mechanism of action is well understood,6
for plants very little is known.1 Understanding the role of 2 in
plants is crucial since manipulation of this pathway may lead to
considerable improvements in crop biomass. To realize this aim,
easy access to 2 and other variants is required.
Scheme 1 Conditions: (a) diphenyl chlorophosphate, py, 18 h; (b)
tert-butyldiphenylchlorosilane, imidazole, DMF, 26 h; (c) HN3, PPh3,
diisopropyl azodicarboxylate, 1,4-dioxane, 66 h.
Department of Chemistry, Chemistry Research Laboratory, Univer-
sity of Oxford, Mansfield Road, Oxford, UK OX1 3TA. E-mail:
ben.davis@chem.ox.ac.uk; Fax: + 44 (0)1865 285002; Tel: +44 (0)1865
285001
† Electronic supplementary information (ESI) available: Full experimental
procedures, microreactor reaction details, NMR spectra and compound
characterization data. See DOI: 10.1039/c0ob00226g
k
k
2
1
Tre + R ⎯⎯→Mono + R ⎯⎯→Di
d[Tre]/dt = -k1[Tre][R]
(1)
4232 | Org. Biomol. Chem., 2010, 8, 4232–4235
This journal is
The Royal Society of Chemistry 2010
©

View Online
d[Mono]/dt = k1[Tre][R]-k2[Mono][R]
d[Di]/dt = k2[Mono][R]
(2)
(3)
(4)
d[R]/dt = -k1[Tre][R]-k2[Mono][R]
For a symmetric diol where modification of one hydroxyl does
not perturb the reactivity of the other, k1 = 2k2 giving a 50%
theoretical yield for the monofunctional product for 1 equivalent
of reagent R at the reaction endpoint. We attempted a range of
reactions on 1 in batch, but found substantially lower yields for
the desired product (Scheme 1), implying that k1 π 2k2 in all
cases. Phosphorylation, silylation and Mitsunobu reaction with
hydrazoic acid gave yields of 24%, 10% and 0% respectively
for the desired monofunctionalized product. Previous literature
accounts of trehalose modifications have reported similar low
yields. For example, halogenation of the primary hydroxyl with N-
halosuccinimides gave 13%, 21% and 37% yields for chloro, iodo
and bromo respectively,14 while tosylation afforded only the doubly
functionalized adduct.15 Together these preliminary results both
highlighted and confirmed existing inefficiencies in polyol systems
such as 1; although formally simple, such monofunctionalizations
remain an unsolved problem.
Fig. 2 Product distribution at various residence times for the phospho-
rylation of 1 in a microreactor chip.
We hypothesized that these non statistical yields were due to the
lower solubility of trehalose in organic solvents giving a low k1. In
such a scenario, after the first reaction, the monofunctionalized
trehalose adduct is rendered more soluble, and as a result k2 >
k1 leading to the observed low yields.‡ With this in mind, we
devised strategies to improve yields and to test the validity of this
hypothesis.
In this, we exploited the poor solubility of deprotected trehalose
analogues to achieve more efficient desymmetrizations. Rather
than attempting a single modification on 1 based on simultaneous
mode (a) and mode (b) selectivity, a reverse approach was used
applying first mode (a) then mode (b) selectivity (Scheme 2).
Removal of a silyl group from 6 gives 5, which is far less soluble.
Thus the subsequent deprotection to 1 occurs at a much slower
rate leading to higher yields for 5.
Flow chemistry is an emerging technology that allows precise
control of reaction conditions and is particularly applicable to
desymmetrizations.16 Direct analysis of product distribution by
HPLC against residence time can yield valuable kinetic data.
High reproducibility allows easy scale-up, a particular problem
for this type of regioselective reaction where the optimum end
point can be difficult to judge, often requiring a visual estimation
of product distribution by TLC. Phosphorylation of 1 in a
microreactor to optimize the yield of 3 was performed with
excess diphenyl chlorophosphate (4 equiv.) over different residence
times. Product distribution was monitored by HPLC (Fig. 2,
see ESI†).17 A maximum yield of 44% was obtained for a 2
min residence time. The reaction was performed on a 2 gram
scale with 39% isolation of 3. Non-linear regression analysis of
the data using eqn (1)–(4) (see ESI†) revealed that k1 = 0.157
mol-1dm3s-1 and k2 = 0.133 mol-1dm3s-1, better reflecting the ex-
pected theoretical yields and ratio of k1 to k2 compared to the batch
reaction.
While the improved yield of phosphorylation is beneficial, the
use of flow methods here is not a synthetic solution in its own
right since the technique is very solvent intensive (the necessity
for solubilisation of trehalose required a dilute 25 mM solution);
methods minimizing solvent usage are preferable.18 Furthermore,
for instalment of other functionality, conditions would have to
be individually re-optimized. Therefore, we chose to use the
same principles to develop more general methodology for the
desymmetrization of trehalose to give a versatile starting material
that could be further modified to other products.
Scheme 2 Reverse modification strategy for desymmetrizations. Desily-
lation reactions performed with TBAF on a microreactor.
The kinetics of this reverse approach were probed using a
microreactor chip. Two solvent conditions were explored: (i) high
solubility conditions (methanol:pyridine 1 : 1) where 6 and 5 show
similar solubility and (ii) low solubility conditions (THF:petrol
1 : 1) where the solubility of 5 is significantly reduced (Fig. 3).
Product distributions were again analyzed using non-linear regres-
sion based on eqn (1)–(4) (see ESI†) to calculate rate constants.
For the high solubility case k1 = 21.9 ¥ 10-3 mol-1dm3s-1 and
k2 = 17.9 ¥ 10-3 mol-1dm3s-1. For the low solubility case k1 =
21.6 ¥ 10-3 mol-1dm3s-1, effectively unaltered by the solvent.
On the other hand k2 = 7.67 ¥ 10-3 mol-1dm3s-1, substantially
reduced from the high solubility case. Thus, by altering the
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 4232–4235 | 4233
©

View Online
of pyridine. 6 was suspended in diethyl ether and methanol
to give a saturated solution, which was deprotected with HCl
generated in situ from acetyl chloride.22 Regular monitoring of
the reaction allowed isolation of 5 in 55% yield over 2 steps.
Compound 5 was phosphorylated regioselectively affording 9 in
88%. Subsequent deprotection of the silyl group under acidic
conditions proceeded smoothly in 94% yield. Hydrogenation of
3 allowed near quantitative phosphate deprotection to furnish 2
in a 45% yield over five steps. Thus, we obtained phosphorylation
yields similar to those discovered through flow chemistry using
a convenient and scalable batch process that offers a significant
improvement in yield over the conventional reaction (Scheme 1a).
In conclusion, we have used flow chemistry to study the
desymmetrization of trehalose. This has produced kinetic data
that has given vital insights into the mechanism of these seem-
ingly simple reactions. This data can explain previous literature
reports of low yielding monofunctionalizations on deprotected
trehalose. Strategies to improve dissolution of trehalose, such as
dilute reaction mixtures, or protection of secondary hydroxyls
are possible workarounds. Alternatively, a reverse modification
route can exploit phase effects to give greater than statistical
monofunctionalization yields. The facile, scalable synthesis of
5 can be readily incorporated into synthetic routes for other
non-symmetrical trehalose analogues and, together with the
reverse modification approach in general, could lead to significant
improvements in yield. The potential also exists for extension
of the principle, for example, to pseudosymmetric disaccharide
polyols allowing access to modifications at one primary hydroxyl
in the presence of another.
Fig. 3 Yield of 5 versus time for different solvents.
solvent to exploit solubility differences of trehalose adducts, yields
of monofunctionalization were improved through manipulation
of k2.‡
Such “phase control” of reactions has been utilized in enzymatic
synthesis to shift reaction equilibria towards the product by judi-
cious choice of solvent.19 Understanding the physics of attrition
grinding deracemizations (another “phase controlled” process)20
has allowed for the enantiopure synthesis of the non-steroidal anti-
inflammatory drug Naproxen.21 Analogously, by understanding
the kinetics of trehalose desymmetrizations, we have optimized
conditions for the synthesis of 2, an important natural signalling
molecule.
To avoid the potential for precipitation under such reduced
solubility conditions (with subsequent blockage of the microreac-
tor) and to create a more generally applicable route for increased
scale, we used the kinetic data and conditions discovered using
the flow methods to develop a batch method. In this way,
flow discovery allowed batch optimization. Thus under these
newly discovered conditions, silylation of 1 smoothly gave 6
(Scheme 3). The crude product was used directly after removal
Acknowledgements
We are grateful to the International AIDS Vaccine Initiative for
financial support. BGD is a Royal Society Wolfson Research Merit
Award recipient and supported by an EPSRC LSI Platform grant.
Notes and references
‡ The rate constants k1 and k2 refer to the apparent rate constants. Due to
the poor solubility of the trehalose analogues, the actual concentration of
these compounds in solution cannot be determined. Thus, while the actual
rate constants cannot be measured, it is mathematically convenient in the
context of these discussions to express the reductions in rate as apparent
rate constants.
1 M. J. Paul, L. F. Primavesi, D. Jhurreea and Y. Zhang, Annu. Rev. Plant
Biol., 2008, 59, 417.
2 J. H. Crowe, J. F. Carpenter and L. M. Crowe, Annu. Rev. Physiol.,
1998, 60, 73; F. Albertorio, V. A. Chapa, X. Chen, A. J. Diaz and P. S.
Cremer, J. Am. Chem. Soc., 2007, 129, 10567.
´
´
3 J. L. Chiara, I. S. de Gracia, A. Garc´ıa, A. Bastida, S. Bobo and M. D.
Mart´ın-Ortega, ChemBioChem, 2005, 6, 186; Y. Hui and C. T. Chang,
Org. Lett., 2002, 4, 2245.
4 P. J. Brennan and H. Nikaido, Annu. Rev. Biochem., 1995, 64, 29; D. E.
Minnikin, L. Kremer, L. G. Dover and G. S. Besra, Chem. Biol., 2002,
9, 545.
5 J. D. Rose, J. A. Maddry, R. N. Comber, W. J. Suling, L. N. Wilson and
R. C. Reynolds, Carbohydr. Res., 2002, 337, 105; J. Wang, B. Elchert,
Y. Hui, J. Y. Takemoto, M. Bensaci, J. Wennergren, H. Chang, R. Rai
and C. T. Chang, Bioorg. Med. Chem., 2004, 12, 6397.
6 M. Paul, Curr. Opin. Plant Biol., 2007, 10, 303; I. Pe´rez-Victoria, S.
Kemper, M. K. Patel, J. M. Edwards, J. C. Errey, L. F. Primavesi, M. J.
Paul, T. D. W. Claridge and B. G. Davis, Chem. Commun., 2009, 5862.
7 M. A. Rodr´ıguez, O. Boutureira, M. I. Matheu, Y. D´ıaz, S. Castillo´n
and P. H. Seeberger, J. Org. Chem., 2007, 72, 8998.
Scheme 3 Conditions: (i) tert-butyldiphenylchlorosilane, py, 14 h; (ii)
AcCl, MeOH–Et2O 1 : 3 18 h; (iii) diphenyl chlorophosphate, py, 15 h; (iv)
AcCl, MeOH, 18 h; (v) H2, PtO2, AcOH, EtOH, 14 h.
4234 | Org. Biomol. Chem., 2010, 8, 4232–4235
This journal is
The Royal Society of Chemistry 2010
©

View Online
8 R. Namme, T. Mitsugi, H. Takahashi and S. Ikegami, Eur. J. Org.
Chem., 2007, 3758.
16 B. P. Mason, K. E. Price, J. L. Steinbacher, A. R. Bogdan and D. T.
McQuade, Chem. Rev., 2007, 107, 2300; K. Ja¨hnisch, V. Hessel, H.
Lo¨we and M. Baerns, Angew. Chem., Int. Ed., 2004, 43, 406.
17 N. Vervoort, D. Daemen and D. To¨ro¨rk, J. Chromatogr., A, 2008, 1189,
92.
9 M. R. Pratt, C. D. Leigh and C. R. Bertozzi, Org. Lett., 2003, 5, 3185.
10 F. L. Lin, H. van Halbeek and C. R. Bertozzi, Carbohydr. Res., 2007,
´
342, 2014; D. Rodr´ıguez-Lucena, J. M. Benito, E. Alvarez, C. Jaime,
18 C. Capello, U. Fischer and K. Hungerbu¨hler, Green Chem., 2007, 9,
J. Perez-Miron, C. O. Mellet and J. M. G. Fernandez, J. Org. Chem.,
2008, 73, 2967; D. Rodr´ıguez-Lucena, C. O. Mellet, C. Jaime, K. K.
Burusco, J. M. G. Ferna´ndez and J. M. Benito, J. Org. Chem., 2009, 74,
2997; S. Shin and T. V. RajanBabu, Org. Lett., 1999, 1, 1229.
11 T. Tsuchiya, Glycoscience: Chemistry and Biology (ed., B. Fraser-Reid,
K. Tatsuta and J. Thiem), 1, 2001, Springer-Verlag, Berlin, pp117-194.
12 P. G. McDougal, J. G Rico, Y. I. Oh and B. D. Condon, J. Org. Chem.,
1986, 51, 3388.
927.
19 R. V. Ulijn and P. J. Halling, Green Chem., 2004, 6, 488; B. Cho, J. Seo,
T. Kang and B. Kim, Biotechnol. Bioeng., 2003, 83, 226; V. N. Barai, S.
V. Kvach, A. I. Zinchenko and I. A. Mikhailopulo, Biotechnol. Lett.,
2004, 26, 1847.
20 D. G. Blackmond, Chem.–Eur. J., 2007, 13, 3290; Y. Saito and H.
Hyuga, J. Phys. Soc. Jpn., 2009, 78, 104001; C. Viedma, Phys. Rev.
Lett., 2005, 94, 065504.
21 W. L. Noorduin, B. Kaptein, H. Meekes, W. J. P. van Enckevort, R. M.
Kellogg and E. Vlieg, Angew. Chem., Int. Ed., 2009, 48, 4581.
22 E. M. Nashed and C. P. J. Glaudemans, J. Org. Chem., 1987, 52, 5255.
13 A. Bouzide and G. Sauve´, Tetrahedron Lett., 1997, 38, 5945; A. Bouzide
and G. Sauve´, Org. Lett., 2002, 4, 2329.
14 S. Hanessian and P. Lavalle´e, Carbohydr. Res., 1973, 28, 303.
15 G. Birch and A. C. Richardson, Carbohydr. Res., 1968, 8, 411.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 4232–4235 | 4235
©