Synthesis and structural characterization of three unique helicobacter pylori a-cholesteryl phosphatidyl glucosides

Article abstract of DOI:10.1002/anie.201406529

Steryl glycosides produced by bacteria play important biological roles in the evasion and modulation of host immunity. Step-economical syntheses of three cholesteryl-6-Ophosphatidyl-α-d-glucopyranosides (αCPG) unique to Helicobacter pylori have been achie

Full text of DOI:10.1002/anie.201406529

.
Angewandte
Communications
DOI: 10.1002/anie.201406529
Natural Products
Synthesis and Structural Characterization of Three Unique
Helicobacter pylori a-Cholesteryl Phosphatidyl Glucosides**
Huy Q. Nguyen, Ryan A. Davis, and Jacquelyn Gervay-Hague*
Abstract: Steryl glycosides produced by bacteria play impor-
tant biological roles in the evasion and modulation of host
immunity. Step-economical syntheses of three cholesteryl-6-O-
phosphatidyl-a-d-glucopyranosides (aCPG) unique to Heli-
cobacter pylori have been achieved. The approach relies upon
regioselective deprotection of per-O-trimethylsilyl-a-d-choles-
terylglucoside at C6 followed by phosphoramidite coupling.
Figure 1. Structures of H. pylori cholesteryl glucosides. Lipid (length of
Global TMS ether deprotection in the presence of oxygen and
carbon chain/number of unsaturated bonds) C14:0, C16:0, C18:1,
subsequent deprotection of the cyano ethyl phosphoester
C19:0 (cyclopropyl).
afforded the target compounds in 16–21% overall yield
starting from d-glucose. The structures of these natural
products were determined using a combination of 2D NMR
methods and mass spectrometry. These robust synthesis and
characterization protocols provide analogues to facilitate
glycolipidomic profiling and biological studies.
approximately 25% of the total lipid content of H. pylori, and
can make up as much as 33% of the total lipid content in
other Helicobacter species.[2a,b,3]
To date, biological studies involving a-cholesteryl gluco-
sides have mostly relied upon mixtures from natural sources,
making it difficult to determine the independent roles of each
constituent.[2e,f] Moreover, biological studies have mainly
utilized TLC Rf values and/or mass spectrometry to character-
ize the components of the H. pylori glycolipid profile.[2a,b,f,3–5]
Recently, aCG (1) and aCAG (2) were synthesized and fully
characterized by NMR spectroscopy and mass spectrometry,
thus making pure samples of these compounds available to
the biological community for the first time.[6] The exact
structures of naturally occurring aCPG analogues (3) have
yet to be defined, as neither TLC nor MS can readily
distinguish the diversity of isomers resulting from esterifica-
tion of the phosphatidyl moiety.[2a,b,3]
Given the growing importance of understanding the
biological significance of H. pylori and related bacterial
immunomodulators, a synthetic campaign focused on devel-
oping step-economical syntheses of aCPG analogues (3a–c,
Figure 2) was initiated. The target compounds are composed
of three structural units, including a cholesteryl aglycon,
a sugar core, and a phosphate ester at C6 with variable
glycerol ester side chains. These structural units present
several synthetic challenges, including: 1) regioselective phos-
phorylation of the C6-hydroxy group,[7] 2) avoidance of
phosphite acetal formation as seen in previous research,[8]
3) installment of the phosphatidyl group while avoiding acyl
migration on the glycerol unit, and 4) regioselective oxidation
I
t has been 30 years since Marshall and Warren isolated and
characterized Helicobacter pylori bacteria from patients with
chronic gastritis.[1] Although the gram-negative bacterium
infects half of the world population, most people are
asymptomatic. However, patients who do show symptoms
are left with a gambit of illnesses ranging from peptic ulcers to
gastric carcinomas.[2] Extensive scientific efforts have con-
tributed to understanding the role H. pylori plays in these
illnesses. One area of research is the synthesis and chemical
characterization of the biomolecules produced by this patho-
gen.
Steryl glycosides are an intriguing class of bacteria-
derived immune modulators. Bacteria do not produce steroids
and the mechanisms they use to confiscate cholesterol from
the host are not well understood. Currently, the three known
cholesteryl glucosides isolated from H. pylori are aCG (1),
aCAG (2), an analogue of aCG that is acylated with
tetradecanoic acid at C6, and aCPG (3), a group of C6-
phosphorylated derivatives (Figure 1).[2a,b,3] The lipid portion
on the phosphatidyl glycerol unit varies in composition as
identified in lyso-CPG analogues isolated from H. pylo-
ri.[2a,b,3,4] Together, these a-cholesteryl glucosides make up
[*] H. Q. Nguyen, R. A. Davis, Prof. J. Gervay-Hague
Department of Chemistry, University of California, Davis
One Shields Ave, Davis, CA 95616 (USA)
E-mail: jgervayhague@ucdavis.edu
[**] This work was supported by the NIH R01GM090262, NSF CHE-
0196482, NSF CRIF Program (CHE-9808183), and NSF OSTI 97-
24412. Grants for 600 and 800 MHz instruments include NSF DBIO
722538 and NIH PR1973. J.G.H. gratefully acknowledges support
from the National Science Foundation for independent research and
development during her tenure as Division Director of Chemistry.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201406529.
Figure 2. Structures of H. pylori a-cholesteryl phosphatidyl glucosides.
13400
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 13400 –13403

Angewandte
Chemie
attempt was made to independently characterize the cyclo-
propyl isomers. The benzyl ether in compound 11 was
removed by hydrogenation to afford the desired glycerol 12,
which was characterized by NMR spectroscopy.
A DEPT NMR experiment was utilized to identify the sn-
1
2
13C (d 70.6 ppm), and the corresponding H NMR shift for
the sn-2 C-H (d 5.17 ppm) was assigned using HSQC NMR
experiments. Afterwards, COSY NMR experiments revealed
the 1H NMR shift for both the sn-1 (d 4.36 and 4.17 ppm) and
sn-3 (d 3.51 ppm) methylene protons. Likewise, HMBC NMR
experiments showed sn-1 CH2 correlation with the fatty ester
carbonyl 13C (d 173.2 ppm), thus distinguishing sn-1 from sn-3
(Figure 4).
Figure 3. Retrosynthesis strategy for the preparation H. pylori aCPGs.
of the phosphorus atom without oxidizing the double bond of
cholesterol.[9] We envisioned starting the synthetic route with
per-O-TMS aCG (4, Figure 3)[10] because of its ready avail-
ability in two steps from glucose and its compatibility with the
required criteria. Regioselective deprotection of the C6 ether
followed by condensation with an appropriate phosphorami-
dite would achieve the desired synthesis in a convergent
manner.
Three different fatty acids are required to address the
syntheses of glycerol esters. Although 1,2-dimyristoyl-sn-
glycerol (5) and 1-palmitoyl-2-oleoyl-sn-glycerol (6) are
commercially available, we wanted to establish a generalized
synthesis of various diacyl glycerols for future library devel-
opment. Thus we began our studies with the esterification of
1-O-benzyl-sn-glycerol (7) using myristic acid in the presence
of DCC and DMAP to afford 8 in 70% yield along with 23%
of the double-addition product (9).[11] Compound 8 was then
condensed a second time with oleic acid to form 10.
Diacylglycerol 11 was synthesized by cyclopropanation of 10
via a Simmons–Smith carbene utilizing Zn/Cu and diiodo-
methane (Scheme 1). Presumably, this reaction gave a mixture
of cyclopropyl diastereoisomers, which were not distinguish-
able by chromatography or NMR spectroscopy, and thus no
Figure 4. NMR experiments utilized to determine ester locations.
The next step in the synthesis was the conversion of
commercially available 1-palmitoyl-2-oleoyl-sn-glycerol to
the phosphoramidite needed to couple to cholesteryl glucose.
This was achieved by first activating 2-cyanoethyl N,N,N,’N’-
tetraisopropylphosphordiamidite with tetrazole and its dis-
placement with 1-palmitoyl-2-oleoyl-sn-glycerol 6 to afford
the phosphoramidite 14 in 92% yield as a 1:1 mixture of
diastereomers (Scheme 2).[12] The same strategy was utilized
to prepare phosphoramidite 13 from 5[11a] and 15 from 12
(Scheme 2).
Scheme 1. Synthesis of diacylglycerols. Reaction conditions: a) myristic
acid, DCC, DMAP, CH2Cl2, 08C!RT, 15 h, 8: 70%, 9: 23%; b) oleic
acid, DCC, DMAP, CH2Cl2, RT, 15 h, 67%; c) Zn/Cu, CH2I2, Et2O,
reflux, 15 h, 62%; d) Pd(OH)2/C, H2 (1 atm), CH2Cl2, MeOH, RT, 2 h,
90%.
Scheme 2. Synthesis of phosphoramidites 13–15.
Angew. Chem. Int. Ed. 2014, 53, 13400 –13403
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 13401

.
Angewandte
Communications
Table 1: Selected 13C NMR resonances (in ppm) of compounds 16–19
and 3a–3c.
Regioselective phosphatidylation at the C6 position was
initiated with the synthesis of compound 4, which was
achieved using our glycosyl iodide one-pot glycosylation
protocol starting from per-O-TMS glucose.[10] In this manner,
we could generate per-O-silylated aCG (4) and then regio-
selectively desilylate the primary ether at C6 using ammoni-
um acetate in dichloromethane and methanol to afford the
free alcohol (16, Scheme 3).[13] The synthetic route to aCPG
through compound 16 avoids the formation of a phosphite
acetal between the hydroxy groups at C4 and C6 of
unprotected aCG(1) and any acyl migration event that
could occur.[8] Freshly prepared phosphoramidites 13–15
were then coupled to 16 using tetrazole as the promoter
(Scheme 3). Utilizing three molar equivalents of tetrazole was
key to the success of the coupling reaction, as smaller
amounts of tetrazole led to diminished yields (< 10%).
Even under these conditions, the yields were lower than
desired, presumably because of steric congestion around the
phosphoramidite.[14] Subsequent introduction of O2 and
DOWEX H+ resin resulted in the oxidation of the phosphorus
atom with concomitant deprotection of the TMS ethers.
Finally, the cyanoethyl protecting group was removed with
1,8-diazabicyclo[5.4.0]undec-7-ene (DBU). After purifica-
tion, aCPGs 3a–3c were obtained in 16–21% yield starting
from d-glucose (Scheme 3).
Extensive NMR analysis of the aCPG compounds
revealed the phosphatidyl group to be attached to the C6’
positions. In the 13C NMR spectrum of 16, C6’ appears as
a singlet at d 61.9 ppm (Table 1). After phosphatidylation
with 13–15, the C6’ was shifted downfield and became
a doublet because of the coupling with the phosphorus atom
(J = 3.5–6.8 Hz). Also of note, C4’, C3’, and C2’ remained as
singlets, indicating that the phosphate is not attached at these
positions. After coupling with the phosphoramidites, a mix-
ture of diastereomers (17–19) originating from the phospho-
rus center is formed. The diastereomeric mixture is observed
in the 31P and 13C NMR spectra with C3’, C5’, C6’, sn-1-C, sn-
2-C, and sn-3-C carbon resonances appearing as duplicates.
C2’
C3’
C4’
C5’
C6’
16[a] 74.2 75.1
72.4 73.0
61.0
17[b] 71.8 73.9, 69.3 70.3,
67.5 (d, J=5.4 Hz),
67.4 (d, J=5.4 Hz)
69.3 (d, J=6.8 Hz)
73.9
70.2
72.1 72.8,
72.7
70.3 71.1,
71.0
18[c] 75.9 74.4
19[b] 72.5 74.4
68.5 (d, J=5.3 Hz),
68.3 (d, J=3.5 Hz)
3a[d] 72.4 73.4
3b[e] 72.7 73.9
3c[f] 72.8 73.9
69.2 71.6 (d, J=3.2 Hz) 64.4 (d, J=5.6 Hz)
68.9 72.1 (d, J=2.6 Hz) 64.2 (d, J=6.8 Hz)
69.9 71.9 (d, J=5.3 Hz) 64.9 (d, J=6.4 Hz)
NMR solvents: [a] C6D6. [b] CDCl3/MeOD=5:1. [c] C5D5N/MeOD=5:1.
[d] CDCl3/MeOD/TEA=5:1.5:0.5 (0.1m). [e] CDCl3/DBU/
CD3COOD=5:0.8:0.2. [f] CDCl3/MeOD 1:1.
Different deuterated solvents were required to solubilize each
sample for NMR investigations, which resulted in chemical
shift fluctuations. Nevertheless, the 13C NMR shifts for 3a–3c
were all quite similar. Furthermore, HMBC NMR experi-
ments indicated that the sn-2-CH and sn-1-CH2 from aCPG
compounds 17–19 and 3a–3c were correlated with the
carbonyl carbon atom of the corresponding fatty acids, thus
providing evidence that acyl migration did not occur and the
phosphorus atom remained attached to the sn-3-CH2.
A synthetic protocol has been developed for the prepa-
ration of three different aCPG analogues (3a–c) associated
with H. pylori immune modulation. Glycosylation of per-O-
silylated glucose proceeds efficiently and with high a-
selectivity because of the armed nature of the per-O-silyl
donors.[10] Selective deprotection of the primary ether and
subsequent condensation with a highly functionalized phos-
phoramidite followed by concomitant oxidation and depro-
tection afforded the desired analogues. Importantly, we have
established a modular approach to preparing these natural
products that is amenable to library development. The
glycosyl iodide glycosylation is versatile, allow-
ing various cholesterol analogues as well as
other lipids to be attached to different carbo-
hydrate cores.[15] Likewise, we have demon-
strated that the phosphoramidite chemistry is
compatible with biologically relevant func-
tional groups such as olefins and cyclopropanes.
In light of the recent discovery that H. pylori
enzymes are promiscuous and readily incorpo-
rate a variety of cholesterol analogues,[16] this
modular platform offers accessibility to various
phosphatidyl glycolipids to study their biolog-
ical properties and also to aid in the discovery
of new analogues.
Received: June 25, 2014
Published online: September 4, 2014
Scheme 3. Synthesis of aCPG (3a–c). Reaction conditions: a) NH4OAc, CH2Cl2, MeOH,
RT, 20 h, 93%; b) 13, 14, or 15, tetrazole, CH2Cl2, CH3CN, RT, 40 h; c) O2 (1 atm),
MeOH, DOWEX H+, RT, 2 h; d) DBU, CH2Cl2, RT, 3 min; e) HOAc. [a] Yields based on
recovered aCG. [b] Total yields from glucose.
Keywords: glycosyl iodide · immunomodulator ·
phosphorylation · steroids · total synthesis
.
13402 www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 13400 –13403

Angewandte
Chemie
Kobayashi, X. F. Bao, J. Bugaytsova, T. Boren, R. S. Jin, Y. N.
Zong, P. H. Seeberger, J. Nakayama, M. Kronenberg, M.
Fukuda, Plos One 2013, 8, e78191.
[1] B. Marshall, J. R. Warren, Lancet 1984, 323, 1311 – 1315.
[2] a) Y. Hirai, M. Haque, T. Yoshida, K. Yokota, T. Yasuda, K.
Oguma, J. Bacteriol. 1995, 177, 5327 – 5333; b) M. Haque, Y.
Hirai, K. Yokota, N. Mori, I. Jahan, H. Ito, H. Hotta, I. Yano, Y.
Kanemasa, K. Oguma, J. Bacteriol. 1996, 178, 2065 – 2070;
c) R. M. Peek, Jr., M. J. Blaser, Nat. Rev. Cancer 2002, 2, 28 – 37;
d) H. M. Algood, T. L. Cover, Clin. Microbiol. Rev. 2006, 19,
597 – 613; e) C. Wunder, Y. Churin, F. Winau, D. Warnecke, M.
Vieth, B. Lindner, U. Zahringer, H. J. Mollenkopf, E. Heinz, T. F.
Meyer, Nat. Med. 2006, 12, 1030 – 1038; f) M. Beigier-Bompadre,
V. Moos, E. Belogolova, K. Allers, T. Schneider, Y. Churin, R.
Ignatius, T. F. Meyer, T. Aebischer, J. Infect. Dis. 2011, 204,
1339 – 1348; g) J. Park, D. Forman, E. R. Greenberg, R. Herrero,
Curr. Oncol. Rep. 2013, 15, 517 – 525.
[6] R. A. Davis, C.-H. Lin, J. Gervay-Hague, Chem. Commun. 2012,
48, 9083 – 9085.
[7] A. Abragam Joseph, C.-W. Chang, C.-C. Wang, Chem. Commun.
2013, 49, 11497 – 11499.
[8] M. K. Patel, B. G. Davis, Org. Lett. 2013, 15, 346 – 349.
[9] a) T. Yamada, K. Imagawa, T. Mukaiyama, Chem. Lett. 1992,
2109 – 2112; b) J. A. R. Salvador, S. M. Silvestre, V. M. Moreira,
Curr. Org. Chem. 2006, 10, 2227 – 2257.
[10] R. A. Davis, J. C. Fettinger, J. Gervay-Hague, J. Org. Chem.
2014, DOI: 10.1021/jo501371h.
[11] a) P. V. Ioannou, G. H. Dodd, B. T. Golding, Synthesis 1979,
939 – 941; b) P. Greimel, Y. Ito, Tetrahedron Lett. 2008, 49, 3562 –
3566.
[12] H. J. Lin, A. K. Adak, L. V. R. Reddy, S. H. Wu, C. C. Lin,
Chem. Eur. J. 2013, 19, 7989 – 7998.
[3] M. Haque, Y. Hirai, K. Yokota, K. Oguma, J. Bacteriol. 1995,
177, 5334 – 5337.
[4] A.-H. Lebrun, C. Wunder, J. Hildebrand, Y. Churin, U.
Zꢀhringer, B. Lindner, T. F. Meyer, E. Heinz, D. Warnecke, J.
Biol. Chem. 2006, 281, 27765 – 27772.
[13] Y. Cui, Z. Cheng, J. Mao, Y. Yu, Tetrahedron Lett. 2013, 54,
3831 – 3833.
[5] a) H. Shimomura, S. Hayashi, K. Yokota, K. Oguma, Y. Hirai,
FEMS Microbiol. Lett. 2004, 237, 407 – 413; b) H.-J. Wang, W.-C.
Cheng, H.-H. Cheng, C.-H. Lai, W.-C. Wang, Mol. Microbiol.
2012, 83, 67 – 84; c) Y. Ito, J. L. Vela, F. Matsumura, H. Hoshino,
A. Tyznik, H. Lee, E. Girardi, D. M. Zajonc, R. Liddington, M.
[14] X. Wei, Tetrahedron 2013, 69, 3615 – 3637.
[15] S. S. Kulkarni, J. Gervay-Hague, Org. Lett. 2008, 10, 4739 – 4742.
[16] H. Shimomura, K. Hosoda, D. J. McGee, S. Hayashi, K. Yokota,
Y. Hirai, J. Bacteriol. 2013, 195, 359 – 367.
Angew. Chem. Int. Ed. 2014, 53, 13400 –13403
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 13403