RETRACTED ARTICLE: Chemical synthesis and antigenic activity of a phosphatidylinositol mannoside epitope from: Mycobacterium tuberculosis

Article abstract of DOI:10.1039/d0cc05573e

Phosphatidylinositol mannosides (PIMs) have been investigated as lipidic antigens for a new subunit tuberculosis vaccine. A non-natural diacylated phosphatidylinositol mannoside (Ac2PIM2) was designed and synthesized by mimicking the natural PIM6 processi

Full text of DOI:10.1039/d0cc05573e

ChemComm
View Article Online
View Journal
COMMUNICATION
Chemical synthesis and antigenic activity
of a phosphatidylinositol mannoside epitope
Cite this:DOI: 10.1039/d0cc05573e
from Mycobacterium tuberculosis†
Received 16th August 2020,
Accepted 1st October 2020
Shi-Yuan Zhao,‡a Na Li,‡a Wan-Yue Luo,‡a Nan-Nan Zhang,a Rong-Ye Zhou,a
abc
Chen-Yu Lia and Jin Wang
*
DOI: 10.1039/d0cc05573e
rsc.li/chemcomm
Phosphatidylinositol mannosides (PIMs) have been investigated as lipoarabinomannans (LAMs), are involved in the modulation of
lipidic antigens for a new subunit tuberculosis vaccine. A non- host immune responses and play important roles in the patho-
natural diacylated phosphatidylinositol mannoside (Ac2PIM2) was genesis of Mtb.5 In particular, PIMs can regulate cytokines and
designed and synthesized by mimicking the natural PIM6 processing stimulate early endosomal fusion by acting as ligands to receptors
procedure in dentritic cells. This synthetic Ac2PIM2 was achieved (MR, DC-SIGN and CR-3).6 In addition, a PIM was identified as a
from a-methyl D-glucopyranoside 1 in 17 steps in 2.5% overall yield. natural antigen for CD1d-restricted T cells,7 and synthetic PIMs
A key feature of the strategy was extending the use of the chiral and their analogues can also modulate cell-mediated immunity,8
myo-inositol building block A to the O-2 and O-6 positions of the indicating their potential as vaccine candidates.9
inositol unit to allow for introducing the mannose building blocks
Structurally, PIMs are each composed of a myo-inositol unit
B1 and B2, and to the O-1 position for the phosphoglycerol building with a diacylated glycerophospholipid moiety at position O-1 and
block C. Building block A, being a flexible core unit, may facilitate a-mannosylation sites at O-2 and O-6. The O-6 mannose unit of
future access to other higher-order PIM analogues. A preliminary the myo-inositol unit may be further substituted at position O-6
antigenic study showed that the synthetic PIM epitope (Ac2PIM2) by four mannosyl units, leading to PIM6. Additional lipid chains
was significantly more active than natural Ac2PIM2, which indicated may be linked at the O-6 position of the 2-O-mannosyl unit and
that the synthetic Ac2PIM2 can be strongly immunoactive and may the O-3 position of myo-inositol to form tetraacylated PIMs
be developed as a potential vaccine.
(Ac4PIM6),10 as shown in Fig. 1. Complex native antigen PIM6
requires processing within antigen-presenting cells to PIM2 to
Tuberculosis (TB) is a mycobacterial disease caused by Mycobac- become immunogenic.11 Lipid antigen PIM6 is recognized by
terium tuberculosis (Mtb), which ranks as the second-leading surface receptors (MR, DC-SIGN and CR3) and in this way can
infectious cause of mortality, after the human immunodeficiency enter antigen-presenting cells. The oligomannosidic moiety of
1
´
virus (HIV). Despite the discovery of the Bacillus Calmette-Guerin PIM6 was shown to be processed by a-mannosidase in the
(BCG) vaccine in the early 20th century2 and the development of a
new drug treatment for TB, this disease is still not under control.3
The envelope of Mtb is the causative agent of human tuberculosis,
and contains a variety of lipids with unique structures that can
serve as antigens for the immune system.4 Among the vital cell
envelope components, phosphatidylinositol mannosides (PIMs)
and their multiglycosylated counterparts, lipomannans (LMs) and
a School of Pharmacy, Yancheng Teachers University, Hope Avenue South Road No.
2, Yancheng, 224007, Jiangsu Province, P. R. China. E-mail: wangj01@yctu.edu.cn
b
´
´
Universite de Toulouse, Universite Toulouse III – Paul Sabatier,
118 route de Narbonne, 31062 Toulouse Cedex 9, France
c CNRS, IPBS (Institut de Pharmacologie et de Biologie Structurale),
205 route de Narbonne, 31077 Toulouse, France
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
d0cc05573e
Fig. 1 Structures of natural Ac4PIM6 and the non-natural Ac2PIM2.
‡ These authors contributed equally to this paper.
Chem. Commun.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Communication
ChemComm
presence of CD1e,12 and the acyl chains from the glycerol unit of
PIM6 were artificially cleaved by recombinant lipases (LPLA2 and
PLRP2).4 The resulting non-natural diacylated PIM2 (Ac2PIM2,
Fig. 1) that contains only two a-D-mannose units is the antigenic
form, which is presented by the CD1b protein to stimulate CD1b-
restricted T cells through the T-cell receptor (TCR).
The number of mannosyl residues and number of the acyl
chains present in the PIM molecules determine their antigenic
activities.13 As a result, elegant synthetic strategies have been
developed for native PIMs (PIM1, PIM2, PIM4, PIM5, PIM6)14 and
PIM analogues.15 However, most of these works have focused on
syntheses of higher PIMs, which often display poor immuno-
genicity, although the structure–activity relationships of PIMs
with different acylated forms have yet to be reported. The precise
acylation state of the PIM molecule appears to be crucial in the
immune responses.16 The relatively small synthetic PIM2 was
demonstrated by Painter et al. to be more active than natural
PIM2.17 Recently, Gilleron et al. reported that the lipase-cut PIM2
showed better antigenic activity than did natural PIM2,11 which
Scheme 1 Synthesis of inositol building block A. Reagents and condi-
tions: (a) TMSCl, Py, 0 1C; (b) (i) PhCHO, FeCl3Á6H2O; (ii) CH2Cl2/CH3CN,
0 1C to rt; (c) (i) Et3SiH; (ii) TBAF, 65%, 3 steps; (d) TBDMSCl, DMF, 45 1C,
quant.; (e) BH3ÁTHF, CoCl2, 0 1C, quant.; (f) SO3ÁPy, DIPEA, DMSO, 0 1C;
(g) K2CO3, Ac2O, 62%, 2 steps; (h) Hg(AcO)2, 64%; (i) NaBH(AcO)3, 93%.
contains two fatty acids at the sn-2 positions of the glycerol unit. ensure the a-glycosylation. This method proved advantageous
In the current work, we mimicked the natural PIM6 process- when compared to previous PIM syntheses. Examining the
ing procedure, and designed and synthesized an artificial PIM structure of Ac2PIM2 indicated the key step to be the synthesis
epitope Ac2PIM2 (Fig. 1). The preliminary immunomodulatory of myo-inositol A, a structural unit with different groups
activity of the synthetic Ac2PIM2 was also evaluated.
substituted at four of its positions (1, 2, 3, 6). The flexibility
Synthetic strategy: The developed retrosynthetic analysis of of the chiral inositol A core unit is expected to facilitate future
Ac2PIM2 is shown in Fig. 2. If this analysis were to be followed, access to other higher-order PIM analogues.
Ac2PIM2 would be assembled from the building blocks A, B1, B2
Synthesis of myo-inositol building block A: The generation of
and C. Specifically, Ac2PIM2 would be obtained via sequential optically pure myo-inositol derivatives with appropriate protect-
regioselective glycosylations, first of the inositol block A at the ing groups is a challenging work. Inspired by the work of Bender
O-6 position with mannose block B1, and then at the O-2 et al.18 on the conversion of methyl a-D-glucopyranoside to an
position with B2. Subsequent introduction of the phospho- enantiomerically pure myo-inositol using a Ferrier reaction, we
glycerol moiety to the O-1 position of the inositol unit would developed a new method for the synthesis of chiral myo-inositol
be carried out using C by applying the H-phosphonate method. building block A, as shown in Scheme 1. First, a-methyl
Further palmitoylations would be carried out, at the O-3 position D-glucopyranoside 1 was converted to tetra-O-silylated glucopyr-
of A and the O-6 position of B2. Final deprotection by catalytic anoside 2, which was treated with benzaldehyde and FeCl3Á6H2O
hydrogenolysis would afford Ac2PIM2. The stereoselectivity of to give the bis-benzylidene acetal 3 as a single diastereomer.19
each glycosydic bond formed would be ensured by neighboring This one-pot tandem reaction proceeded well at room tempera-
C-2 acyl participating groups.
ture and gave 4 in 65% yield over three steps. Silylation of 4 with
In this study, we employed an Fmoc group as a selectively tert-butyldimethylsilyl chloride (TBDMSCl) gave the 4,6-O-
removable protecting group for the mannose building blocks B. benzylidene hexopyranoside 5 in quantitative yield. Regioselec-
We chose the Fmoc group—and not an acetyl group—for B, tive reductive ring-opening of 5 was performed by using CoCl2
because the inositol building block A already has an acetyl and BH3ÁTHF, and gave 6 in quantitative yield.20 Swern oxidation
group at its O-1 position and this acetyl group was needed for of 6 provided the air-sensitive aldehyde 7. Without any purifica-
later introduction of the phosphoglycerol moiety. Thus, the O-2 tion, 7 was immediately treated with acetic anhydride and K2CO3
acetate of the mannose building blocks B had to be replaced to afford enol acetate 8 (62% yield, over two steps) as a mixture of
with another participating group, i.e., the Fmoc group, to Z- and E-isomers. It was not necessary to isolate the two isomers
because the geometry of the enol acetate 8 would not affect the
stereochemical outcome of the Ferrier reaction.21 The Ferrier
rearrangement of 8 gave 9 in 64% yield as a single product.
Subsequent reduction of 9 with NaBH(AcO)3 provided building
block A in 93% yield with a good diastereoselectivity and its
diastereomer 10 in 7% yield.
Preparation of the mannose building blocks B1 and B2: The
procedure for preparing the building blocks B1 and B2 is
shown in Scheme 2. First, 11 was produced from D-mannose
Fig. 2 Retrosynthetic analysis of Ac2PIM2.
by following the methodology developed by Iadonisi.22
Chem. Commun.
This journal is © The Royal Society of Chemistry 2020

View Article Online
ChemComm
Communication
inositol building block A at the O-6 position with building block
B1 in the presence of N-iodosuccinimide (NIS) and trimethylsi-
lyl trifluoromethanesulfonate (TMSOTf) gave pure a-linked
disaccharide 20 in 64% yield.29 Subsequent glycosylation of
the O-2 position of 20 with building block B2 afforded the pure
a-linked trisaccharide 21 in 82% yield. The Fmoc group needed
to be replaced with another protecting group for the later
reactions. The benzyloxy-methyl (BOM) group was selected to
replace the Fmoc group because the BOM group can be
introduced onto the OH group under mild conditions utilizing
DIPEA as a basic catalyst and hence without affecting the O-1
acetyl group of the inositol unit. The Fmoc group was removed
from 21 by treating 21 with Et3N to give 22 in 72% yield. The
resulting two OH groups of 22 were protected with the BOM
group to give 23 in 55% yield. Removal of the O-1 acetyl group
of 23 under Zemplen conditions gave 24 in quantitative yield.
Employing the H-phosphonate method,29 compound 24 was
coupled with building block C in the presence of pivaloyl
chloride (PivCl) and then oxidized with a solution of I2 in
aqueous pyridine to successfully produce the desired 25 in
85% yield (Scheme 4). Removal of the two TBS groups of 25 gave
26 in 85% yield, and palmitylations of the two free OH groups
of 26 with palmitic acid (C15H31COOH) afforded 27 in 89%
yield. Finally, global deprotection of 27 achieved by subjecting
it to catalytic (Pd(OH)2/C) hydrogenolysis led to Ac2PIM2 in 80%
yield (Scheme 4). The total yield of Ac2PIM2 based on building
block A was 11%.
Scheme 2 Syntheses of mannose building blocks B1 and B2. Reagents
and conditions: (a) (i) Ac2O, Py; (ii) I2, Et3SiH, CH2Cl2; (iii) 2,6-lutidine,
MeOH; (b) NaOMe, MeOH; (c) NaH, BnBr, 70%, 4 steps; (d) HgBr2, EtSH,
60%; (e) (i) NaOMe, MeOH; (ii) Fmoc-Cl, Py, 75%, 2 steps; (f) (i) TBDMSCl;
(ii) BnBr, NaH, 46%, 2 steps; (g) HgBr2, EtSH, 72%; (h) (i) NaOMe, MeOH,
quant.; (ii) Fmoc-Cl, Py, 86%.
Deacetylation of 11 under Zemplen conditions23 gave 12, which
was subsequently benzylated with benzyl bromide to provide
orthoester 13 in 70% yield over four steps. HgBr2-mediated
ring-opening of the orthoester 13 with ethanethiol gave 14 in
60% yield as a mixture of a and b anomers (with the major one
being a).24 Removal of the O-2 acetyl group of 14 was carried out
and the resulting product was treated with 9-fluorenylmethyl
chloroformate (Fmoc-Cl) to give B1 in 75% yield over two
steps.25 Similarly, regioselective silylation of the O-6 position
of 1226 and subsequent benzylation with benzyl bromide gave
15 in 46% yield over two steps. HgBr2-mediated ring-opening of
15 afforded 16 in 72% yield as a single a-anomer. Deacetylation
of 16 and protection of the resulting product with the Fmoc
group gave B2 in 86% yield.
Synthesis of building block C: At first, we set out to use either the
phoshphoramidite 19 or the H-phosphonate C as the building
block for the introduction of the phosphoglycerol moiety, because
both of them have been used successfully in previous syntheses of
PIM compounds.27 However, in our case, treatment of 17 with
2-cyanoethyl tetraisopropyl phosphorodiamidite 18 in the same
conditions as reported in the ref. 27 failed to yield the phosphor-
amidite 19. We suspected that the nearby benzyl groups in 17
sterically hindered the OH group and prevented this group from
reacting with the coupling reagent. Fortunately, the building block
Evaluation of antigenic activity: The IFN-g-producing activity of
synthetic Ac2PIM2 compared with those of the natural PIMs (PIM2,
PIM6, Ac4PIM2, Ac3PIM2, Ac2PIM2, Fig. 3A) are shown in Fig. 3B.
We achieved the chemical synthesis of a non-natural
phosphatidylinositol mannoside (PIM) epitope capable of bind-
ing the CD1b protein to activate T cells and release the IFN-g
C
was readily prepared according to literature procedures
(Scheme 3).28 The yield of building block C was 77%.
Assembly of Ac2PIM2: The synthesis of Ac2PIM2 is shown in
Scheme 4. Due to the equatorial OH at the 6 position of
myo-inositol building block A being more reactive than the
axial OH at the 2 position, selective mannnosylation of the
Scheme 4 Assembly of Ac2PIM2. Reagents and conditions: (a) NIS,
TMSOTf, À20 1C, 64%; (b) NIS, TMSOTf, À20 1C, 82%; (c) Et3N, THF, rt,
72%; (d) BOMCl, TBAI, DIPEA, rt, 55%; (e) NaOMe, MeOH, quant.; (f) (i) PivCl,
Py; (ii) I2, Py/H2O, 85%; (g) TBAF/THF, 40 1C, 85%; (h) C15H31COOH, DCC,
DMAP, toluene,100 1C, 89%; (i) Pd(OH)2/C, H2, rt, 80%.
Scheme 3 Synthesis of building block C. Reagents and conditions:
(a) 1H-tetrazole; (b) (i) PCl3, imidazole; (ii) Et3N, toluene, 0 1C to rt, 77%.
Chem. Commun.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Communication
ChemComm
4 P. Sander, K. Becker and M. D. Molin, Cell Chem. Biol., 2016, 23,
1044–1046.
¨
5 (a) G. Kallenius, M. Correia-Neves, H. Buteme, B. Hamasur and
S. B. Svenson, Tuberculosis, 2016, 96, 120–130; (b) Z. Palcekova,
M. Gilleron, S. Angala, J. M. Belardinelli and M. C. Jackson, ACS
Infect. Dis., 2020, 6, 2235–2248.
6 J. B. Torrelles, A. K. Azad and L. S. Schlesinger, J. Immunol., 2006,
177, 1805–1816.
7 K. Fischer, E. Scotet, M. Niemeyer, H. Koebernick, J. Zerrahn, S. Maillet,
R. Hurwitz, M. Kursar, M. Bonneville, S. H. E. Kaufmann and
U. E. Schaible, Proc. Natl. Acad. Sci. U. S. A., 2004, 101, 10685–10690.
8 N. A. Parlane, B. J. Compton, C. M. Hayman, G. F. Painter,
R. J. Basaraba, A. Heiser and B. M. Buddle, Vaccine, 2012, 30,
580–588.
9 G. Larrouy-Maumus, E. Layre, S. Clark, J. Prandi, E. Rayner,
M. Lepore, G. de Libero, A. Williams, G. Puzo and M. Gilleron,
Vaccine, 2017, 35, 1395–1402.
10 P. S. Patil, T.-J. R. Cheng, M. M. L. Zulueta, S.-T. Yang, L. S. Lico and
S.-C. Hung, Nat. Commun., 2015, 6, 7239.
11 M. Gilleron, M. Lepore, E. Layre, D. Cala-De Paepe, N. Mebarek,
`
J. A. Shayman, S. Canaan, L. Mori, F. Carriere, G. Puzo and G. De
Libero, Cell Chem. Biol., 2016, 23, 1147–1156.
12 H. de la Salle, S. Mariotti, C. Angenieux, M. Gilleron, L.-F. Garcia-
Fig. 3 Use of AcnPIMs loaded onto CD1b to activate T cells. (A) Structures
of natural PIM2 (left) and PIM6 (right). The acylation sites are indicated by
R1–R4; (B) stimulation of CD1b-restricted T cells with different acyl forms
ˆ
Alles, D. Malm, T. Berg, S. Paoletti, B. Maıtre and L. Mourey, Science,
2005, 310, 1321–1324.
of AcnPIMs, Ag = antigen. ***P o 0.001 compared to natural Ac3PIM2 and 13 M. Gilleron, C. Ronet, M. Mempel, B. Monsarrat, G. Gachelin and
G. Puzo, J. Biol. Chem., 2001, 276, 34896.
###P o 0.001 compared to natural Ac2PIM2.
14 (a) G. D. Ainge, B. J. Compton, C. M. Hayman, M. William John,
S. M. Toms, D. S. Larsen, J. L. Harper and G. F. Painter, J. Org. Chem.,
2011, 76(12), 4941–4951; (b) L. Xinyu, B. L. Stocker and P. H.
factor. This synthetic Ac2PIM2 was achieved from 1 in 17 steps
Seeberger, J. Am. Chem. Soc., 2006, 128(11), 3638; (c) G. M. Rankin,
in a 2.5% overall yield. A series of efficient synthetic transfor-
mations incorporated three building blocks (A, B1, B2) into the
pseudo-trisaccharide 24. The overall efficiency of the assembly
process benefited from the use of shared mannose building
blocks (B1, B2) and the carefully chosen mannoside protecting
groups (Fmoc, BOM). The key intermediate building block A
allowed for efficient stereoselective glycosylations of the
mannoside buliding blocks B and introduction of the phos-
phoglycerol building block C. The synthetic PIM epitope
Ac2PIM2 was significantly more active than was natural Ac2PIM2
in inducing the production of IFN-g, and hence could be
developed as a potential vaccine against tuberculosis.
We thank the National Natural Science Foundation of China
(No. 81903427), the Jiangsu Key Research and Development
Plan (Social Development No. BE2020672) project, the Natural
Science Foundation of Jiangsu Province (BK20160443), and the
Six Talent Peaks Project in Jiangsu Province (SWYY-094) for
financial support.
B. J. Compton, K. A. Johnston, C. M. Hayman, G. F. Painter and
D. S. Larsen, J. Org. Chem., 2012, 77(16), 6743–6759; (d) D. Wang,
D.-C. Xiong and X.-S. Ye, Chin. Chem. Lett., 2018, 29(9), 1340–1342;
(e) A. Yohei, T. Shota, M. Takanori, Y. Sho and F. Yukari, Org. Biomol.
Chem., 2020, 18, 3659–3663.
15 (a) B. Siwarutt, L. Xinyu, M. Mario, L. Bernd and P. H. Seeberger,
J. Am. Chem. Soc., 2008, 130(49), 16791; (b) G. D. Ainge, M. William
John, B. J. Compton, C. M. Hayman, D. S. Larsen, Y. Sung-Il,
I. A. Wilson, J. L. Harper and G. F. Painter, J. Med. Chem., 2011,
54(20), 7268–7279.
16 B. S. Dyer, J. D. Jones, G. D. Ainge, D. Michel, D. S. Larsen and
G. F. Painter, J. Org. Chem., 2007, 72, 3282–3288.
¨
17 G. D. Ainge, N. A. Parlane, M. Denis, B. S. Dyer, A. Harer,
C. M. Hayman, D. S. Larsen and G. F. Painter, J. Org. Chem., 2007,
72, 5291–5296.
18 S. L. Bender and R. J. Budhu, J. Am. Chem. Soc., 1991, 113,
9883–9885.
19 Y. Bourdreux, A. Lemetais, D. Urban and J.-M. Beau, Chem. Com-
mun., 2011, 47, 2146–2148.
20 S. Tani, S. Sawadi, M. Kojima, S. Akai and K.-i. Sato, Tetrahedron
Lett., 2007, 48, 3103–3104.
21 H. Takahashi, H. Kittaka and S. Ikegami, J. Org. Chem., 2001, 66,
2705–2716.
`
22 M. Adinolfi, A. Iadonisi, A. Ravida and M. Schiattarella, Tetrahedron
Lett., 2003, 44(43), 7863–7866.
´
23 G. Zemplen and E. Pacsu, Ber. Dtsch. Chem. Ges. (A and B Ser.), 1929,
Conflicts of interest
62, 1613–1614.
24 Z.-G. Wang, J. D. Warren, V. Y. Dudkin, X. Zhang, U. Iserloh,
M. Visser, M. Eckhardt, P. H. Seeberger and S. J. Danishefsky,
Tetrahedron, 2006, 62(20), 4954–4978.
There are no conflicts to declare.
25 J. Kandasamy, M. Hurevich and P. H. Seeberger, Chem. Commun.,
2013, 49, 4453–4455.
Notes and references
´
26 A. Lubineau and D. Bonnaffe, Eur. J. Org. Chem., 1999, 2523–2532.
1 M. Pai, M. A. Behr, D. Dowdy, K. Dheda, M. Divangahi, C. C.
Boehme, A. Ginsberg, S. Swaminathan, M. Spigelman, H. Getahun,
D. Menzies and M. Raviglione, Nat. Rev. Dis. Primers, 2016, 2, 16076.
2 J. Guiard, A. Collmann, M. Gilleron, L. Mori, G. De Libero, J. Prandi
and G. Puzo, Angew. Chem., Int. Ed., 2008, 120, 9880–9884.
3 J. D. Simmons, C. M. Stein, C. Seshadri, M. Campo, G. Alter, S. Fortune,
E. Schurr, R. S. Wallis, G. Churchyard, H. Mayanja-Kizza, W. H. Boom
and T. R. Hawn, Nat. Rev. Immunol., 2018, 18, 575–589.
27 G. M. Rankin, B. J. Compton, K. A. Johnston, C. M. Hayman,
G. F. Painter and D. S. Larsen, J. Org. Chem., 2012, 77(16),
6743–6759.
28 H.-H. Tan, A. Makino, K. Sudesh, P. Greimel and T. Kobayashi,
Angew. Chem., Int. Ed., 2012, 51(2), 533–535.
29 G. D. Ainge, J. Hudson, D. S. Larsen, G. F. Painter, G. S. Gill and
J. L. Harper, Bioorg. Med. Chem., 2006, 14(16), 5632–5642.
Chem. Commun.
This journal is © The Royal Society of Chemistry 2020