Synthesis of D-glycero-D-manno-heptose 1,7-bisphosphate (HBP) featuring a β-stereoselective bis-phosphorylation

Article abstract of DOI:10.1016/j.tetlet.2017.08.005

D-glycero-D-manno-Heptose 1,7-bisphosphate (HBP) plays a unique role in bacteriology. We describe in this study a very efficient synthesis of HBP, featuring a highly 6-D-selective construction of the heptose scaffold as well as a double phosphorylation st

Full text of DOI:10.1016/j.tetlet.2017.08.005

Tetrahedron Letters xxx (2017) xxx–xxx
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Synthesis of D-glycero-D-manno-heptose 1,7-bisphosphate (HBP)
featuring a b-stereoselective bis-phosphorylation
⇑
Lina Liang, Stéphane P. Vincent
University of Namur (UNamur), Rue de Bruxelles 61, 5000 Namur, Belgium
a r t i c l e i n f o
a b s t r a c t
Article history:
D-glycero-D-manno-Heptose 1,7-bisphosphate (HBP) plays a unique role in bacteriology. We describe in
Received 22 June 2017
Revised 31 July 2017
Accepted 1 August 2017
Available online xxxx
this study a very efficient synthesis of HBP, featuring a highly 6-D-selective construction of the heptose
scaffold as well as a double phosphorylation step installing, in a single operation and in a b-stereoselec-
tive manner, the 1- and the 7-phosphates.
Ó 2017 Elsevier Ltd. All rights reserved.
Keywords:
Carbohydrates
Phosphorylation
Glycosidation
Bacterial saccharides
The preparation of ‘‘higher-carbon sugars” such as heptoses has
been investigated for more than a century and has witnessed all
the major (r)evolutions of organic chemistry that occurred during
the 20th century. The discovery of natural heptoses stimulated
the synthetic chemists to develop robust methodologies to synthe-
size them and demonstrate their structures, including in natural
products such as the spicamycin,1 miharamycins,2,3 and desfer-
6, the donor substrate of heptosyltransferases (WaaC, WaaF and
WaaQ) that will further construct the LPS molecule.14,15
A similar pathway has been evidenced for the D,D-heptosylated
oligosaccharides and the 6-deoxy-heptosides found in some bacte-
rial O-antigens as well as in Gram-positive bacteria. The kinase
HddA phosphorylates the same D,D-heptose 2 to give the a-anomer
7 with opposite stereoselectivity than HldE. This intermediate 7-
phosphate is then dephosphorylated by the phosphatase GmhB
and transformed into GDP-heptose 9 that can be further processed.
Our group has explored the synthesis of natural and non-natu-
risalmycin.4 Furthermore, the discovery that the glycero-
D-
manno-heptoses are common subunits of lipopolysaccharides
(LPS) and capsular saccharides of many pathogenic bacteria, inten-
sified the research by offering new applications: synthetic hepto-
sides could be designed, for instance, as antimicrobial agents or
synthetic vaccines.5–7 The main synthetic methods to construct
heptosides have been reviewed by Oscarson,8 Kosma,9 and more
recently by our group.10
ral heptosides as inhibitors of three enzymes of the LPS
D,D-heptose
pathway: GmhA, HldE and WaaC.16–19
Very recently, the group of Gray-Owen discovered that the
intermediate -glycero- -manno-heptose 1,7-bisphosphate 3
D
D
(HBP) plays a very unique role in bacteriology, distinct from the
LPS biosynthesis. Indeed, it was shown that HBP, once released
from pathogenic bacteria such as Nesseiria menengitidis, can trigger
an innate and adaptive immune response in mammals.20 These
remarkable and novel biological properties along with our contin-
uous efforts in generating glycomimetics of heptose-phosphates
pushed us to envisage the synthesis of HBP. This molecule can be
used as substrate of GmhB to develop enzymatic assays to discover
The biosynthesis of L,D-heptosides has attracted a strong atten-
tion because of their occurrence in the LPS of most gram-negative
bacteria, including in major human pathogens.11 Sedoheptulose-7-
phosphate 1, a product of the central metabolism, is the biosyn-
thetic precursor of the bacterial heptoses (Fig. 1).12,13 The first
enzyme GmhA isomerizes 1 into
D-glycero-D-mannoheptose 7-
phosphate 2 which is then phosphorylated by the kinase HldE
yielding heptose bisphosphate 3 (HBP), the object of the present
study. After hydrolysis of the terminal phosphate of 3, intermedi-
ate 4 is transformed into nucleotide-sugar 5. A regioselective
novel inhibitors of the D,D-heptose biosynthetic pathway, but it can
also be exploited to study the bacterial immunology.
Interestingly, when we started this study, there was only an
enzymatic synthesis of HBP but no chemical synthesis.21 Very
recently, we developed a methodology for the b-stereoselective
epimerization is then promoted by HldD, yielding ADP-L-heptose
phosphorylation of
D-mannosides, including D
-heptosides.22,23
⇑
Corresponding author.
E-mail address: stephane.vincent@unamur.be (S.P. Vincent).
http://dx.doi.org/10.1016/j.tetlet.2017.08.005
0040-4039/Ó 2017 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Liang L., Vincent S.P. Tetrahedron Lett. (2017), http://dx.doi.org/10.1016/j.tetlet.2017.08.005

2
L. Liang, S.P. Vincent / Tetrahedron Letters xxx (2017) xxx–xxx
OH
O
O
OH
OH
P
HO
O
GmhA
OH
OH
OH
HO
OH
O
HO
HO
OH
O
2
HO P O
OH
1
HldE
HddA
O
OH
OH
O
P
OH
P
O
HBP
OH
O
HO
OH
O
HO
HO
O
OH
HO
O
O
HO
O P OH
OH
HO
O P OH
3
7
OH
GmhB
GmhB
OH
OH
OH
O
HO
OH
HO
O
HO
HO
O P OH
OH
O
HO
HO
O
P OH
OH
4
O
8
HldE
HddC
OH
O
O
HO
HO
OH
OH
OH
O
O P O P OAde
OH OH
HO
HO
HO
O
5
O
O
HO
HldD
O P O P OGua
OH OH
9
OH
OH
D-glycero-D-manno-heptose-1 a-GDP
O
O
HO
HO
O
O P O P OAde
OH OH
HO
HddD
6
L-glycero-D-manno-heptose-1 b-ADP
S-layer glycoprotein
WaaC, WaaF, WaaQ
Lipid A
Fig. 1.
D-glycero-
D-manno-Heptose 1a-GDP biosynthetic pathway and
L
-glycero-
D-manno-heptose 1b-ADP pathway. Ade = adenosine, Gua = guanosine.
Our next challenge was thus the synthesis of the biologically rele-
vant HBP molecule.
We have to mention that, while we were writing this manu-
script, a chemical synthesis of HBP was disclosed by the group of
Fujimoto.24 Here, we describe our own approach and emphasize
the differences in the synthetic strategies.
Results
As mentioned above, many strategies have been exploited to
generate heptoses from hexoses. We selected the pathway
depicted in Scheme 1 because of its high stereoselectivity in favor
of the desired 6-D epimer 12.16,18,25
In our synthesis of fluorinated analogues of heptosides,16 we
had also explored the opening of an epoxide derived from an
alkene similar to 11, but we had found that this strategy is slightly
L-selective, as recently exploited by Fujimoto et al.24 Therefore, we
privileged a dihydroxylation of 11 that eventually lead to a high
9/1 stereoselectivity in favor of the D isomer.16,18
Scheme 1. Reagents and conditions: a) K2OsO4(H2O)2, K2CO3, K3Fe(CN)6, water/t-
BuOH/toluene, 0 °C-R.T, 89%, D/L = 9:1; b) TIPSCl, imidazole, THF, R.T, 93%, D/
L = 9:1; c) NaH, BnBr, DMF, R.T, 84%, the D/L = 9:1 epimers were separated at that
stage.
The two epimers 12D and 12L could not be separated after the
dihydroxylation step. This epimeric mixture was thus engaged in
the two subsequent steps yielding pure 14 as a single D stereoiso-
mer. It is important to note that the two epimers could only be sep-
arated at the stage of molecule 14, but not 12 and 13. The
intermediate 14 features now a differentiated 7-position opening
the way for phosphorylation studies.
simultaneously. To do so, we transformed intermediate 14 into
16 by an acetolysis yielding intermediate bis-acetate 15 that was
de-acetylated to provide 16 in high yield (Scheme 2).
The b-stereoselective phosphorylation of manno-configured
sugars is a real synthetic challenge. We have recently realized a
methodological study to synthesize b-1-phosphates of a broad
range of carbohydrates, including mannosides and heptosides.22,23
Applied to intermediate 16, our own methodology only provided
bis-phosphate in low yield and poor stereoselectivity. This result
The key step of our synthetic strategy was to perform a double
phosphorylation of the heptose scaffold at the 7- and 1-positions,
Please cite this article in press as: Liang L., Vincent S.P. Tetrahedron Lett. (2017), http://dx.doi.org/10.1016/j.tetlet.2017.08.005

L. Liang, S.P. Vincent / Tetrahedron Letters xxx (2017) xxx–xxx
3
using a Mitsunobu reaction was not successful. This result is typi-
cal of the strong influence of protective groups on the carbohydrate
reactivity.
The subsequent deprotection of 17 could be classically achieved
by a global hydrogenolysis of all benzyl groups. The analytical data
of the final molecule corresponded well with literature data.21,24
In conclusion, we disclosed in this study a very efficient synthe-
OH
OTIPS
OBn
OAc
OBn
BnO
BnO
BnO
OBn
BnO
BnO
BnO
BnO
a)
b)
O
O
O
BnO
BnO
OH
OCH3
OAc
16
15
14
c)
sis of HBP, featuring a highly 6-D-selective construction of the hep-
O
O
Na+
O
O
tose scaffold as well as a double phosphorylation step installing, in
a single operation and in a b-stereoselective manner, the 1- and the
7-phosphates.
P
P
OBn
Na+
O
O
OBn
d)
O
O
BnO
BnO
BnO
OBn
HO
OH
O
O
P O Na+
HO
O
P OBn
OBn
O
HO
O
Na+
Acknowledgments
17
3
The authors thank the China Scholarship Council (CSC, PhD
grant n° 201506670006 to L.L.).
Scheme 2. Reagents and conditions: a) conc. H2SO4, AcOH/Ac2O, 0 °C, 67%; b)
MeONa, MeOH, R.T, 90%; c) dibenzyl phosphate, tris(4-chlorophenyl)phosphine,
Et3N, DEAD, THF,
(Na+).
a:b = 1:2, 53% (b); d) H2, Pd(OH)2/C, 1,4-dioxane/H2O, Dowex 50
A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.tetlet.2017.08.
005.
corroborates the literature data on the challenging b-mannosyla-
tion reactions whose stereochemical outcome highly depends on
the functional group at the 6-position, in a remote fashion. More-
over, this bisphosphorylation step was complicated by the fact that
the resulting bis-phosphate showed some unstability, especially
during the silica gel purification step.
References
1. Suzuki T, Chida N. Chem Lett. 2003;32:190–191.
2. Marcelo F, Abou-Jneid R, Sollogoub M, Marrot J, Rauter AP, Blériot Y. Synlett.
2009;1269–1272.
3. Marcelo F, Jimenez-Barbero J, Marrot J, Rauter AP, Sinay P, Bleriot Y. Chem A Eur
J. 2008;14:10066–10073.
In 2011, we had synthesized a series of D,D-heptose-7-phos-
phate analogues and we had found that a Mitsunobu phosphoryla-
tion at the position 7 was particularly efficient.18 Mitsunobu type
phosphorylations at the anomeric position of carbohydrates are
not so usual, but we decided to explore this strategy. To our great
pleasure, we could obtain bisphosphate 17 in good yield (60%) and
4. Dong L, Roosenberg JM, Miller MJ. J Am Chem Soc. 2002;124:15001–15005.
5. Yang Y, Martin CE, Seeberger PH. Chem Sci. 2012;3:896–899.
6. Crich D, Banerjee A. J Am Chem Soc. 2006;128:8078–8086.
7. Kong L, Vijayakrishnan B, Kowarik M, et al. Nat Chem. 2016;8:242–249.
8. Hansson J, Oscarson S. Curr Org Chem. 2000;4:535–564.
9. Kosma P. Curr Org Chem. 2008;12:1021–1039.
10. Tikad A, Vincent SP. Synthetic Methodologies Towards Aldoheptoses and Their
Applications to the Synthesis of Biochemical Probes and LPS Fragments. Wiley-
VCH; 2014. 29–65.
11. Raetz CR, Whitfield C. Annu Rev Biochem. 2002;71:635–700.
12. Kneidinger B, Marolda C, Graninger M, et al. J Bacteriol. 2002;184:363–369.
13. Kneidinger B, Graninger M, Puchberger M, Kosma P, Messner P. J Biol Chem.
2001;276:20935–20944.
satisfactory
a/b selectivity (a/b = 1:3), using the standard PPh3/
DEAD Mitsunobu conditions. However, we must specify that the
separation of the desired b-anomer by standard silica gel chro-
matography was extremely difficult. To optimize the yield, we
eventually found that using tris(4-chlorophenyl)phosphine in
place of PPh3 facilitated the purification. Under these conditions,
bisphosphate 17 was obtained in 53% yield with a
a/b selectivity
of 1/2, a yield and a stereoselectivity that have to be considered
very satisfactory based on literature data.8–10 The b configuration
was ascertained by NMR techniques as in our previous studies.23
14. Czyzyk DJ, Liu C, Taylor EA. Biochemistry. 2011;50:10570–10572.
15. Gronow S, Brabetz W, Brade H. Eur J Biochem. 2000;267:6602–6611.
16. Dohi H, Perion R, Durka M, et al. Chem A Eur J. 2008;14:9530–9539.
17. Durka M, Buffet K, Iehl J, Holler M, Nierengarten JF, Vincent SP. Chem A Eur J.
2012;18:641–651.
In particular, we found that the 1J(C1-H1) of the
a anomer was
18. Durka M, Tikad A, Perion R, et al. Chem A Eur J. 2011;17:11305–11313.
19. Tikad A, Fu H, Sevrain C, Laurent S, Nierengarten J-F, Vincent SP. Chem A Eur J.
2016;22:13147–13155.
20. Gaudet RG, Sintsova A, Buckwalter CM, et al. Science. 2015;348:1251–1255.
21. Wang L, Huang H, Nguyen HH, Allen KN, Mariano PS, Dunaway-Mariano D.
Biochemistry. 2010;49:1072–1081.
176.4 Hz and 163.9 Hz for the b anomer. Moreover, through-space
NOE correlations between H1 and H5 of the b anomer confirmed
this assignment.
We have to mention that Fujimoto et al. also developed a Mit-
sunobu phosphorylation at the anomeric position.24 However their
synthetic scheme involves two distinct and subsequent phospho-
rylation techniques and a protective group manipulation between
the two phosphorylations. While this manuscript was in the
review/revision process, Zamyatina et al. also disclosed an efficient
synthesis of HBP 3.26 Here again, two subsequent phosphorylation
steps were required. Interestingly, in their approach, acetates were
selected as protective groups and the anomeric phosphorylation
22. Li T, Tikad A, Pan W, Brissonnet Y, Vincent SP. Stereocontrolled b and
a
phosphorylations of -mannose. In: Vidal S, Roy R, editors. Carbohydrate
D
Chemistry: Proven Synthetic Methods, Vol. 3. CRC Press; 2015:133–140.
23. Li T, Tikad A, Pan W, Vincent SP. Org Lett. 2014;5628–5631.
24. Inuki S, Aiba T, Kawakami S, Akiyama T, Inoue JI, Fujimoto Y. Org Lett.
2017;19:3079–3082.
25. Li T, Tikad A, Durka M, Pan W, Vincent SP. Carbohydr Res. 2016;432:71–75.
26. Borio A, Hofinger A, Kosma P, Zamyatina A. Tetrahedron Lett.
2017;58:2826–2829.
Please cite this article in press as: Liang L., Vincent S.P. Tetrahedron Lett. (2017), http://dx.doi.org/10.1016/j.tetlet.2017.08.005