Synthesis of SL0101 carbasugar analogues: Carbasugars via Pd-catalyzed cyclitolization and post-cyclitolization transformations

Article abstract of DOI:10.1021/ol101009q

A general approach to the stereoselective synthesis of 5a-carbasugars has been developed. The route mimics our palladium-catalyzed glycosylation/ postglycosylation approach to carbohydrates in that it also utilizes a highly regio- and stereospecific palladium-catalyzed allylation and postglycosylation reaction sequence for the installation of either d- or l-cyclitols. This cyclitolization/postcyclitolization sequence was used for the enantioselective synthesis of a cyclitol analogue of SL0101, its d-sugar enantiomer, as well as several acetylation pattern analogues.

Full text of DOI:10.1021/ol101009q

ORGANIC
LETTERS
2010
Vol. 12, No. 13
2986-2989
Synthesis of SL0101 Carbasugar
Analogues: Carbasugars via
Pd-Catalyzed Cyclitolization and
Post-Cyclitolization Transformations
Mingde Shan and George A. O’Doherty*
Department of Chemistry, West Virginia UniVersity, Morgantown, West Virginia 26506
george.odoherty@mail.wVu.edu
Received May 3, 2010
ABSTRACT
A general approach to the stereoselective synthesis of 5a-carbasugars has been developed. The route mimics our palladium-catalyzed
glycosylation/postglycosylation approach to carbohydrates in that it also utilizes a highly regio- and stereospecific palladium-catalyzed allylation
and postglycosylation reaction sequence for the installation of either D- or L-cyclitols. This cyclitolization/postcyclitolization sequence was
used for the enantioselective synthesis of a cyclitol analogue of SL0101, its D-sugar enantiomer, as well as several acetylation pattern analogues.
The p90 ribosomal S6 kinase (RSK) is a family of serine/
threonine kinases that have been identified as a promising target
for anticancer study.1 Among the four RSK isoforms (RSK1-4),
RSK1 and RSK2 are the most closely linked to cancer cell
growth.1 Therefore, inhibitors of RSK1-2 have the potential
to be chemotherapies of human cancers. So far, several small
molecules have been recognized to inhibit RSK1-2.1 Among
them, SL0101 and its analogues with different acetylation
patterns were initially discovered by Smith, Hecht, Lannigan,
et al. to be the first specific inhibitors of RSK1-2 (Figure 1).2
Not long after the discovery, Hecht reported a total synthesis
of SL0101.3 And later, additional analogues were synthesized
and screened for RSK2 inhibitory activities.4 As part of a larger
effort aimed at understanding the structural details behind
SL0101’s risk inhibition, we also reported a synthetic approach
to SL0101 and its analogues (1-3) via a Pd-catalyzed glyco-
sylation5 and subsequent postglycosylation transformations
(Scheme 1).6 In contrast to the other routes to SL0101, our route
also provided access to the D-sugar enantiomer of SL0101,
allowing us to gauge the role of the sugar in its structure-activity
relationship (SAR).
Figure 1. RSK2 inhibitor SL0101 and its analogues.
In this same vein, we turned our attention to the synthesis of
carbasugar analogues of SL0101 as part of a continuing effort
to search for more potent analogues of SL0101 (Figure 1).
(1) Nguyen, T. L. Anticancer Agents Med. Chem. 2008, 8, 710–716.
(2) Smith, J. A.; Poteet-Smith, C. E.; Xu, Y.; Errington, T. M.; Hecht,
S. M.; Lannigan, D. A. Cancer Res. 2005, 65, 1027–1034.
(3) Maloney, D. J.; Hecht, S. M. Org. Lett. 2005, 7, 1097–1099.
(4) (a) Smith, J. A.; Maloney, D. J.; Clark, D. E.; Xu, Y.; Hecht, S. M.;
Lannigan, D. A. Bioorg. Med. Chem. 2006, 14, 6034–6042. (b) Smith, J. A.;
Maloney, D. J.; Hecht, S. M.; Lannigan, D. A. Bioorg. Med. Chem. 2007,
15, 5018–5034.
(5) Shan, M.; O’Doherty, G. A. Org. Lett. 2006, 8, 5149–5152.
(6) (a) Babu, R. S.; O’Doherty, G. A. J. Am. Chem. Soc. 2003, 125,
12406–12407. (b) Babu, R. S.; Zhou, M.; O’Doherty, G. A. J. Am. Chem.
Soc. 2004, 126, 3428–3429.
10.1021/ol101009q  2010 American Chemical Society
Published on Web 06/02/2010

Carbasugars have been known as sugar mimics, in which the
ring oxygen of the sugar moiety was replaced with a methylene
group.7 This substitution imparts stability to acid and enzymatic
hydrolysis and thus provides substantially improved biostability.
While the carbasugar motif has been used by both man8 and
nature,9 a unified strategy for their synthesis was lacking.
Scheme 2. Synthesis of Two R-Boc-enones from Quinic Acid
Scheme 1. De Novo Approach to Normal Sugar and Carbasugar
prepared in 12 and 11 steps from quinic acid 10 (Scheme 2).10
Although the route was a little long, it provided ample quantities
of the two enantiomeric D- and L-Boc-enones for our method-
ological and medicinal chemistry studies (vide infra).
Scheme 3. Pd-Cyclitolization for R-Boc-enones with BnOH
In this regard, we were particularly interested in developing
a practical and general approach to carbasugar synthesis, which
mimics our Pd-catalyzed glycosylations (7a to 8a) and postg-
lycosylation approach (8a to 9a) (Scheme 1). Because our
palladium glycosylation reaction uses the double bond to
stabilize the carbocation intermediate (via a ꢀ-Pd-π-allyl
intermediate), the ring oxygen is not needed. For our proposed
cyclitolization reaction variant to work as well, the electron-
withdrawing C-4 ketone must direct the incoming nucleophile
to the C-1 sugar position.
Our carbasugar studies began with our investigations of the
palladium(0)-catalyzed cyclitolization. In practice, R-D-Boc-
enone (ent-7b) was treated with BnOH in CH2Cl2 in the
presence of 10 mol % of Pd(PPh3)2 at 0 °C for 12 h. As a result,
the reaction afforded glycosylated enone 11 in a reasonable 60%
yield (Scheme 3).
The desire for this transformation preforced the use of a Boc-
enone 7b instead of a Boc-pyranone 7a in an analogous Pd-
catalyzed cyclitolization (7b to 8b), which would install the
carbasugar glycosidic bond in 8b. In turn, suitable postcycli-
tolization transformations (8b to 9b) could be used to install
the remaining carbasugar functionality. Herein we describe our
successful efforts to expand our de novo approach to carbo-
hydrates to include the synthesis of carbasugars. In addition,
we demonstrate its utility in the synthesis of novel SL0101
carbasugar analogues and their enantiomers.
Scheme 4. Post-Cyclitolization Transformation
Recently, we developed a stereodivergent synthesis of either
enantiomer of the required Boc-enones from D-quinic acid.
Thus, both R-L-Boc-enone 7b and R-D-Boc-enone ent-7b were
In order to construct the sugar functionality, we next turned
our attention to the postcyclitolization transformation. In
particular, we hope to develop a practical way to the manno/
rhamno stereochemistry since the sugar moiety in SL0101 and
its analogues are rhamno-sugars. In analogy to our pyranone
chemistry, we first explored the Luche-type 1,2-reduction of
the R,ꢀ-unsaturated ketone. Unfortunately, these conditions
(NaBH4/CeCl3 at -78 °C) gave the allylic alcohol with only a
1.5:1 diastereoselectivity, which is poor in comparison to the
pyranone chemistry (dr >20:1). After screening a variety of
reducing agents, we found LiAlH4 reduction at -78 °C resulted
in a reasonable diastereoselectivity of 11:1 to afford allylic
alcohol 12 with 85% yield. The minor diastereomer could be
removed by silica gel chromatography. To install the cis-diol,
allylic alcohol 12 was then dihydroxylated at 0 °C upon Upjohn
(7) Carbohydrate Mimics. Concepts and Methods; Chapleur, Y., Ed.;
Wiley-VCH: Weinheim, 1998.
(8) For examples of cyclitol synthesis, see ref 10 and: (a) Ortiz, J. C.;
Ozores, L.; Cagide-Fagin, F.; Alonso, R. Chem. Commun. 2006, 4239–
4241. (b) Chiara, J. L.; Garcia, A.; Sesmilo, E.; Vacas, T. Org. Lett. 2006,
8, 3935–3938. (c) Chakraborty, C.; Vyavahare, V. P.; Dhavale, D. D.
Tetrahedron 2007, 63, 11984–11990. (d) Lysek, R.; Schutz, C.; Favre, S.;
O’Sullivan, A. C.; Pillonel, C.; Krulle, T.; Jung, P. M. J.; Clotet-Codina,
I.; Este, J.; Vogel, R. Bioorg. Med. Chem. 2006, 14, 6355–6282. (e) Corsaro,
A.; Pistara, V.; Catelani, G.; D’Andrea, F.; Adamo, R.; Chiacchio, M. A.
Tetrahedron Lett. 2006, 47, 6591–6594. (f) Frigell, J.; Cumpstey, I.
Tetrahedron Lett. 2007, 48, 9073–9076. (g) Sardinha, J.; Guieu, S.; Deleuze,
A.; Fernandez-Alonso, M. C.; Rauter, A. P.; Sinay, P.; Marrot, J.; Jimenez-
Babero, J.; Sollogoub, M. Carbohydr. Res 2007, 342, 1689–1703. (h)
Plumet, J.; Gomez, A. M.; Lopez, J. C. Mini-ReV. Org. Chem. 2007, 4,
201–216.
(9) For a review of cyclitol-containing natural products, see: Flatt, R.;
Mahmud, T. Nat. Prod. Rep. 2007, 24, 358–392.
(10) (a) Shan, M.; O’Doherty, G. A. Org. Lett. 2008, 10, 3381–3384.
(b) Shan, M.; O’Doherty, G. A. Synthesis 2008, 3171–3179.
Org. Lett., Vol. 12, No. 13, 2010
2987

also worth mentioning that the Pd catalyst loading could be
lowered to 5 mol % with the reaction times remaining in the
0.5-2 h range.
We next turned our attention to the synthesis of SL0101
carbasugar glycoside analogues by examining the Pd-cyclito-
lization reaction with a suitably protected SL0101 aglycon 15
(Scheme 5). In practice, reaction between flavonol 15 and R-D-
Boc-enone 7b went smoothly in CH2Cl2 in the presence of 5
mol % of Pd catalyst at 0 °C in 30 min to afford desired
glycosylated enone 16 in 84% yield.
Table 1. Pd-Cyclitolization of Phenols and Enol with
R-D-Boc-enone
Scheme 5
.
Pd-Cyclitolization for Kaempferol with
R-L-Boc-enone
Scheme 6. Synthesis of SL0101 5a-Carbasugar R-L-glycoside 4
conditions (OsO4/NMO)11 which afforded triol 13 in 90% yield
with complete stereocontrol (Scheme 4).
With this proof of concept established, we decided to next
investigate the Pd-cyclitolization with phenolic and enolic
nucleophiles because the aglycon of SL0101 is an enolic
nucleophile. For this purpose, a number of phenol/enol nucleo-
philes with different substitution patterns were chosen for testing
(Table 1). It turned out that the Pd-cyclitolization worked very
well with reasonable to excellent yields even for those phenols/
enols either with sterically demanding or highly electron
deficient substituents. With the phenol/enol nucleophiles, there
is the added issue of O- vs C-allylation and/or accompanied
Claisen rearrangements.12 Thus, we were delighted to see that
for this cyclitolization, this appeared not to be a problem. It is
To install the remaining SL0101 functionalities in the
carbasugar, we explored the postcyclitolization transformation.
Although we were concerned about the ketone functionality in
the aglycon, this turned out not to be a problem. Thus, reduction
of the glycosylated enone 16 with LiAlH4 at -78 °C afforded
allylic alcohol (89%, dr 11:1), which was acylated to give allylic
acetate 17 in 93% yield (Scheme 6). Dihydroxylation of the
olefin furnished a diol 18 as a single diastereomer (73%), which
was then debenzylated via hydrogenolysis with Pd/C to afford
one of our desired carbasugar glycoside analogues of SL0101,
the C-4 monoacetate 6, in 68% yield.
In order to achieve the diacetate carbasugar glycoside
analogues of SL0101, we thought a selective C-2 acylation
could be carried out on diol 18 via orthoacetate formation and
kinetic hydrolysis. This was successfully used in our synthesis
of SL0101 (Scheme 7). To our surprise, when diol 20 was
reacted with trimethyl orthoacetate at 0 °C in the presence of a
(11) VanRheenen, V.; Kelly, R. C.; Cha, D. Y. Tetrahedron Lett. 1976,
17, 1973–1976.
(12) (a) Trost, B. M.; Toste, F. D. J. Am. Chem. Soc. 1998, 120, 815–
816. (b) Satoh, T.; Ikeda, M.; Miura, M.; Nomura, M. J. Org. Chem. 1997,
62, 4877–4879. (c) Nay, B.; Peyrat, J.-F.; Vercauteren, J. Eur. J. Org. Chem.
1999, 2231–2234.
2988
Org. Lett., Vol. 12, No. 13, 2010

Pd-catalyzed cyclitolization of kaempferol 15 with Boc-enone
ent-7b afforded 85% yield of enone ent-16 under the same
conditions as before, and after the same sequence of postcy-
clitolization transformation, carbasugar glycoside analogue of
SL0101, C-4 monoacetate ent-6 was obtained (Scheme 8).
Scheme 7
.
Synthesis of SL0101 5a-Carbasugar R-L-Glycoside 5
and 6
Scheme 9. Synthesis of ent-5 and ent-4
catalytic amount of p-toluenesulfonic acid followed by the
hydrolysis by using 90% aqueous acetic acid furnished a
mixture of 2,4-diacetate 19 and 3,4-diacetate 20 in a 1.5:1 ratio.
This result stood in contrast to the completely regioselective
acylation of the axial hydroxy group in the synthesis of
SL0101,5 indicating a significantly less rigid chair conformation
for this rhamno-carbasugar versus the rhamno-sugar. This is
presumably due to the loss of anomeric effect. After separation
of these two regioisomers on silica gel chromatography, 19 and
20 were per-debenzylated with H2 upon Pd/C producing the
other two desired SL0101 carbasugar analogues, the C-2,C-4
diacetate 5 (63%) and the C-3,C-4 diacetate 4 (75%), respec-
tively.
By a similar orthoacetate formation and kinetic hydrolysis
with 90% acetic acid, the diol ent-18 was converted to a mixture
of diacetate ent-19 and ent-20 in a ∼1:1 ratio. Global deben-
zylation of these two precursors by hydrogenation afforded the
enantiomeric carbasugar glycoside of SL0101, C-2,C-4 diacetate
ent-5 and C-3,C-4 diacetate ent-4 (Scheme 9).
Scheme 8. Synthesis of ent-6
In conclusion, six SL0101 carbasugar glycoside analogues
in either enantiomeric form have been synthesized successfully.
The formation of the key glycosidic bond features a highly
regio- and stereospecific Pd-catalyzed cyclitolization. The
functionalities on the sugar moieties have been established via
corresponding postcyclitolization transformations. Further ap-
plications of this new methodology along with the associated
medicinal chemistry studies will be reported in due course.
Acknowledgment. We are grateful to NIH (GM088839)
and NSF (CHE- 0749451) for the support of our research
program and NSF-EPSCoR (0314742) for a 600 MHz NMR
at WVU. A dissertation fellowship to M.S. from WVU’s Office
of Graduate Study and Life is highly appreciated.
Supporting Information Available: Complete experimen-
tal procedures and spectral data for all new compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
The synthesis of the enantiomeric carbasugar glycoside
analogues (ent-4, ent-5, and ent-6) was accomplished by simply
switching the R-L-Boc-enone 7b to R-D-Boc-enone ent-7b. A
OL101009Q
Org. Lett., Vol. 12, No. 13, 2010
2989