Total synthesis of mycobacterial arabinogalactan containing 92 monosaccharide units

Abstract

Carbohydrates are diverse bio-macromolecules with highly complex structures that are involved in numerous biological processes. Well-defined carbohydrates obtained by chemical synthesis are essential to the understanding of their functions. However, synthesis of carbohydrates is greatly hampered by its insufficient efficiency. So far, assembly of long carbohydrate chains remains one of the most challenging tasks for synthetic chemists. Here we describe a highly efficient assembly of a 92-mer polysaccharide by the preactivation-based one-pot glycosylation protocol. Several linear and branched oligosaccharide/polysaccharide fragments ranging from 5-mer to 31-mer in length have been rapidly constructed in one-pot manner, which enables the first total synthesis of a biologically important mycobacterial arabinogalactan through a highly convergent [31+31+30] coupling reaction. Our results show that the preactivation-based one-pot glycosylation protocol may provide access to the construction of long and complicated carbohydrate chains.

ARTICLE

Received 14 Sep 2016 | Accepted 7 Feb 2017 | Published 16 Mar 2017

Carbohydrates are diverse bio-macromolecules with highly complex structures that are involved in numerous biological processes. Well-dened carbohydrates obtained by chemical synthesis are essential to the understanding of their functions. However, synthesis of carbohydrates is greatly hampered by its insufcient efciency. So far, assembly of long carbohydrate chains remains one of the most challenging tasks for synthetic chemists. Here we describe a highly efcient assembly of a 92-mer polysaccharide by the preactivation-based one-pot glycosylation protocol. Several linear and branched oligosaccharide/ polysaccharide fragments ranging from 5-mer to 31-mer in length have been rapidly constructed in one-pot manner, which enables the rst total synthesis of a biologically important mycobacterial arabinogalactan through a highly convergent [31 31 30] coupling

reaction. Our results show that the preactivation-based one-pot glycosylation protocol may provide access to the construction of long and complicated carbohydrate chains.

DOI: 10.1038/ncomms14851 OPEN

Total synthesis of mycobacterial arabinogalactan containing 92 monosaccharide units

Yong Wu1, De-Cai Xiong1, Si-Cong Chen1, Yong-Shi Wang1 & Xin-Shan Ye1

1 State Key Laboratory of Natural and Biomimetic Drugs, School of Pharmaceutical Sciences, Peking University, Xue Yuan Road No. 38, Beijing 100191, China. Correspondence and requests for materials should be addressed to X.-S.Y. (email: mailto:[email protected]

Web End [email protected] ).

NATURE COMMUNICATIONS | 8:14851 | DOI: 10.1038/ncomms14851 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 1

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14851

Carbohydrates are involved in many key biological processes, such as cell signaling, cell proliferation and differentiation and viral and bacterial infections, as well as

immunoresponse13. Naturally occurring carbohydrates and glycoconjugates usually exist in microheterogeneous forms, making the isolation of pure carbohydrates and glycoconjugates from natural sources difcult or even impossible in most cases. Therefore, chemical synthesis becomes the main approach to obtain well-dened carbohydrates47. However, unlike peptides and oligonucleotides, which can be routinely prepared by automated solid-phase synthesizers, the oligosaccharide synthesis is much more difcult. The major challenge for oligosaccharide preparation is the regio- and stereochemistry8,9 issues in each glycosidic bond formation, making oligosaccharide synthesis a tedious and time-consuming process. Therefore, oligosaccharide synthesis becomes a daunting task, especially when polysaccharides are chosen as the target molecules. Indeed, only a few examples of the synthesis of oligosaccharide sequences containing 420 units have been reported over the past few decades1018. These syntheses are challenging because multiple steps of protective group manipulation and intermediate purication are required in most cases.

Arabinogalactan is an essential structural constituent of mycobacterial cell wall, which plays critical roles in the infectivity and pathogenicity of Mycobacterium tuberculosis19. Based on experiments and analyses2022, the primary structure of arabinogalactan has been established as a linear galactan composed of about 30 alternating b-(1-5)-linked and b-(1-6)-linked D-galactofuranose (Galf) residues, to which up to two22 highly branched arabinan chains (each containing 31 D-arabinofuranose (Araf) residues) are attached. The arabinogalactan motifs are useful probes for investigating the biosynthesis of mycobacterial cell wall, especially for characterization of the enzymes that process this polysaccharide, and those enzymes are attractive targets for the development of new antituberculosis drugs23. To this end, some solution phase14,15,2427 and automated solid-phase28 strategies have been developed for the assembly of motifs up to 22 residues, among which the Lowary group14 and Ito group15 have elegantly synthesized the docosasaccharide arabinan motif via the convergent [5 5 12]

and [7 7 8] coupling strategies, respectively. While almost all

these syntheses relied upon the stepwise synthesis of oligosaccharide fragments, the object of this study is to achieve the total synthesis of the whole complex polysaccharide rather than the truncated fragments in an efcient way.

In the preactivation-based one-pot glycosylation strategy, several glycosyl donors are allowed to react sequentially in the same vessel regardless of the anomeric reactivities, generating a single oligosaccharide as the main product, which can signicantly simplify the synthetic process and increase the overall efciency29. Herein, by utilizing the preactivation-based one-pot glycosylation protocol, we report the rst total synthesis of a biologically important mycobacterial arabinogalactan composed of 92 monosaccharide units. Our synthetic strategy involves: (1) several scalable one-pot coupling reactions to generate the linear and branched oligosaccharide fragments, (2) the stereoselective b-arabinofuranosylation by preactivation protocol, (3) the further one-pot coupling reactions of oligosaccharide fragments for the rapid assembly of polysaccharides up to 31-mer, and (4) the convergent [31 31 30] coupling reaction for the nal

construction of the target polysaccharide.

ResultsRetrosynthetic analysis. The target polysaccharide arabinogalactan 1 was disconnected into two sizeable fragments, that is, the

linear Galf30 acceptor 2 and the branched Araf31 donor 3 (Fig. 1). It was conceived that Galf30 acceptor 2 would be rapidly assembled via a ve-component one-pot coupling of several oligosaccharide fragments 47. As for the synthesis of Araf31 donor 3, oligosaccharide fragments 1012 were designed to carry out a four-component one-pot glycosylation reaction. For the preparation of heptasaccharide 10, thioglycoside donors 13ac and thioglycoside acceptor 14 were planned for the construction of the challenging b-arabinofuranosyl linkages.

Finally, it was expected that all the oligosaccharide fragments (8, 9, 1517) would be accessible by the preactivation-based one-pot oligosaccharide synthesis starting from various monosaccharide building blocks. Overall, it was anticipated that the major challenge of our plan towards the total synthesis of arabinogalactan 1 would rely on the efciency of one-pot glycosylation reactions, especially when large oligosaccharide fragments were attempted as the components in one-pot coupling reactions.

Synthesis of Galf30 acceptor 2. To test our one-pot strategy for oligosaccharide synthesis, three monosaccharide building blocks 1820 were designed and synthesized (Supplementary Fig. 1). Using these building blocks, the assembly of hexasaccharide 8 in a six-component one-pot manner (18 19 20 19 20 19) by preactivation protocol was

tried, which should be rather challenging as up to ve glycosidic linkages must be correctly constructed. To our delight, when promoted by stoichiometric p-toluenesulfenyl chloride/silver triate (p-TolSCl/AgOTf)29, all glycosylation steps underwent smoothly and none of the side products interfered with the reaction. After optimization of the reaction conditions, hexasaccharide 8 was obtained in 63% overall yield and on a perfect scale (1.07 g) within several hours (Fig. 2a). The desilylation of 8 provided 5 (85% yield), which was re-protected with benzoyl group to give 4 in 96% yield. Subsequently, the coupling reaction of 8 with 1-octanol afforded 21 (91% yield), which was followed by desilylation to provide 6 in 87% yield.

With hexasaccharides 46 in hand, the further iterative one-pot glycosylation was performed. The ve-component one-pot coupling of these oligosaccharides (4 5 5 5 6)

was realized successfully, producing the 30-mer polysaccharide 22 in 68% overall yield (Fig. 2b). It was noteworthy that some deletion sequences were difcult to be removed by column chromatography on silica gel. Gratifyingly, given the difference in molecular weight between the deletion sequences and desired product, size exclusion chromatography was then used to obtain the pure 30-mer polysaccharide 22 (Supplementary Fig. 2). The identity of 22 was conrmed by its nuclear magnetic resonance (NMR) and matrix-assisted laser desorption/ ionizationtime of ight (MALDI-TOF) mass spectra (see Supplementary Information for details). Finally, the global deprotection of 22 via successive debenzoylation and debenzylation provided the 30-mer galactan 23 ([M Na] m/z calcd.

for 5017.4, found: 5018.1).

Having established a highly efcient one-pot approach to the synthesis of polysaccharide up to 30-mer, we turned to accomplish the construction of Galf30 acceptor 2. As shown in

Fig. 1, an additional hexasaccharide 9 equipped with two levulinoyl groups was needed. Initially, monosaccharide 24a (Supplementary Fig. 3) was designed as a building block for a six-component one-pot assembly of 9. However, the efforts failed due to the migration of levulinoyl group from the O-5 to O-6 position. As an alternative route, disaccharide 24b was synthesized (Supplementary Fig. 3). Therefore, a four-component one-pot coupling reaction (18 24b 24b 19) nally gave the

2 NATURE COMMUNICATIONS | 8:14851 | DOI: 10.1038/ncomms14851 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14851 ARTICLE

OH

HO

O

HO

HO O

O

OH

O OH

OH

O OH O

OH

O

O

OH

O

HO O

OH

O

O

O

OH

O

OH

OH

OOH

O

HO HO

OH

OH

HO

O

Arabinan

O

OH

O

OH

OH

O OH

O

HO

O

HO O

O

HO

O

OC H

O

13

O

OH

HO

O

O

HOO

HO

OH

O

OH O

OH

O OH

HOO HO

OH

O

HO

O

HO O

OH

O

HO O

OH

O

O

O

OH

O

OH

O OH O

OH

O

OH

O OH

HOO

HO

O OH

OH

O

HOO

O

OH

HO HO

O

O

3

O

HO HO

OH

O

HO O

OH

O

O

OH

HOO

HO

O OH

O OH

OH

OOH

OH

O OH

OH

O OH

O

OH

OH

O

O

OH

HO

O

Arabinan

HO

O

O

HOO

O

13

O

HO

Galactan

HO O

O

OH

O

O

OH

OH

O OH

HOO

HO

O OH

O

O

HOO HO

O

OH

O

OH

OH

O

HO O

OH

O

O

OH

O OH

HOO HO

HO HO

O

O

HOO

HO

OH

9

HO OH

Arabinogalactan 1

OBz

BnO

O

BzO

[31+31+30]

BnO O

O

O OBz O

OBn

O OBz

OBn

OBz

O

O

OC H

O

O

OBn

O OBz

OBn

OOBz

O

BnOO BnO

BnOO

BnO

OBn

O

BnO O

OBz

O

O

OBz

BzO BnO

OBn

O OBz

BnOO

BnO

O OBz

OBn

O OBz

O

OBz

O

O

OBz

BnO

O

O

STol

BnOO HO

O

OBz

O

OBn

3

BzO

13

BnO O

O

BnOO BnO

O

BnOO

BnO

O OBz

OBz

OBn

O

O

OBz

O OBn

O OBz

O

O OBz

2

OBn

O

BnO O

OBz

O

O

OBz

O

O

OBn

BnOO BnO

BnOO

BnO

OBz

BzO BnO

O

BzO OBz

9

Araf31 donor 3

Galf30 acceptor 2

7: R1 = H, R2 = Lev, R3 = STol9: R1 = TBS, R2 = Lev, R3 = STol

one pot[7 + 7 + 5 + 6 + 6]

one pot[6 + 6 + 6 + 6 +6]

OBz

BnO

O

R

BzO

RO

O

BnOO R O

BnOO

BnO

BnO O

OOBz

OBn

OOBz

OBz

O

OBz

OBn

O OBz

OBn

O OBz O

OBz

OBn

O OBz O

OBn

O OBz

O STol

OBz

RO

OBz

O

BnOO

BnO

O

OBz

O

O

O

O

OBn

O OBz

O

O

OBz

O

STol

O

STol

BnO

BnOO

R O

O

OBz

OBn

O

BnO O

OBz

O

O

OBn

O

OBn

OBn

OBn

2

4

RO

BzO BnO

10 11: R = H

16: R = TBS

12: R = H

17: R = TBS

Arabinofuranosylation

4: R1 = Bz, R2 = Bn, R3 = STol

5: R1 = H, R2 = Bn, R3 = STol

6: R1 = H, R2 = Bn, R3 = OC8H17

8: R1 = TBS, R2 = Bn, R3 = STol

R O

BnO OR

OBn

O

BnO OR

O

O

OR

STol

O OBz O

O

13a: R1 = R2 = R3 = Bn13b: R1, R2 = -TIPDS-, R3 = Bn 13c: R1, R2 = -TIPDS-, R3 = TIPS

OBn

O OBz

O STol

14: R = H

15: R = Lev

OBn

OBz

OR

O

O

OBn

Figure 1 | The structure of mycobacterial arabinogalactan 1 and its retrosynthetic analysis. Bn, benzyl; Bz, benzoyl; Lev, levulinoyl; TBS, tert-butyl-dimethylsilyl; TIPDS, tetraisopropyldisiloxanylidene; TIPS, triisopropylsilyl; Tol, p-tolyl.

NATURE COMMUNICATIONS | 8:14851 | DOI: 10.1038/ncomms14851 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 3

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14851

a

19

c

BnO

BnOO

HO

STol

OBz

BnOO LevO

STol

24b

OBz

19

HO

BnOO

BnO

O

OBz

24b

19

HO

BnOO

BnO

STol

OBz

19

R

(12)

O

BnO

BnOO

TBSO

STol

OBz

BnOO BnO

BnOO

BnO

18

OBz

18

STol

OBz

O

BnOO

BnO

O

OBz

BnOO LevO

BnOO

BnO

OBz

OBz

BnO

BnOO

R O

O

OBz

2

O

OBz

BnO

BnOO

RO

O

OBz

BnOO

BnO

2

85% (4)

96% (5)

87% (4)

8: R1 = TBS, R2 = STol 5: R1 = H, R2 = STol4: R1 = Bz, R2 = STol 21: R1 = TBS, R2 = OC8H17

6: R1 = H, R2 = OC8H17

(6) 91%

90% (4)

85% (15)

9: R = TBS 7: R = H

5

b

5

d

5

6

5

(7)

OC H

4

O

(7)

BnOO BnO

BnOO

BnO

6

OBz

R OO

R O

OC H

BnOO

BnO

O OBz

R OO R O

O

BnOO RO

O

OBz

OR

3

O

OR

BnOO

BnO

O OBz

OR

BnOO BnO

O

R OO

R O

O

OR

R OO

R O

OBz

R O

14

22: R1 = Bz, R2 = Bn 23: R1 = R2 = H

O OBz

2

O

BnOO BnO

BnOO

BnO

OBz

(10), (11)

78%, 2 steps

25: R = Lev 2: R = H

BzO OBz

9

Figure 2 | Synthesis of Galf30 acceptor 2. (a) Synthesis of oligosaccharides 46; (b) Synthesis of 30-mer galactan 23; (c) Synthesis of hexasaccharide 7; (d) Synthesis of Galf30 acceptor 2. Reagents and conditions: (1) TTBP, 4 MS, CH2Cl2, p-TolSCl, AgOTf, then 19, 78 C to room temperature;

(2) p-TolSCl, AgOTf, then 20, 78 C to room temperature; (3) p-TolSCl, AgOTf, then 19, 78 C to room temperature; (4) HF-pyridine, THF/H2O (10:1),

35 C; (5) Bz2O, DMAP, pyridine, CH2Cl2, reux; (6) p-TolSCl, AgOTf, TTBP, 1-octanol, 4 MS, CH2Cl2, 78 C; (7) TTBP, 4 MS, CH2Cl2, p-TolSCl,

AgOTf, then 5, 78 C to room temperature; (8) p-TolSCl, AgOTf, then 5, 78 C to room temperature; (9) p-TolSCl, AgOTf, then 6, 78 C to

room temperature; (10) NaOCH3, CH3OH/CH2Cl2 (2:1); (11) Pd/C, H2, EtOAc/THF/1-PrOH/H2O (2:1:1:1); (12) TTBP, 4 MS, CH2Cl2, p-TolSCl, AgOTf, then 24b, 78 C to room temperature; (13) p-TolSCl, AgOTf, then 24b, 78 C to room temperature; (14) p-TolSCl, AgOTf, then 7, 78 C to room

temperature; (15) H2NNH2-AcOH, THF/CH3OH (10:1). DMAP, 4,4-dimethylaminopyridine; MS, molecular sieves; TTBP, 2,4,6-tri-tert-butylpyrimidine.

desired hexasaccharide 9 in 79% yield (Fig. 2c), which was subjected to desilylation to afford 7 (90% yield). Subsequently, polysaccharide 25 was assembled in a ve-component one-pot manner as described in the construction of 22 by coupling the oligosaccharide fragments (4 5 5 7 6) in

64% overall yield (Fig. 2d). Exposure of 25 to hydrazine acetate successfully fullled the preparation of the desired Galf30 acceptor 2 (85% yield).

Synthesis of Araf31 donor 3. The assembly of the branched Araf31 donor 3 required three oligosaccharide intermediates, that is, b-Araf-containing heptasaccharide 10, branched pentasaccharide 11 and linear hexasaccharide 12. For this purpose, a set of arabinofuranosyl building blocks (13ac, 2629) were designed and synthesized (Supplementary Fig. 4). A six-component iterative one-pot glycosylation of monosaccharides 26 and 27 (26 27 27 27 27 27) afforded

hexasaccharide 17 in excellent yield (73%) and on gram scale(1.20 g) (Fig. 3a). The desilylation of 17 resulted in the desired hexasaccharide 12 (92% yield). Likewise, the one-pot coupling reaction of building blocks 26, 28 (ref. 30) and 27 provided a branched pentasaccharide 16 very smoothly (78% yield), which was further converted into diol 11 by desilylation in 95% yield (Fig. 3b). For the preparation of heptasaccharide 10, another diol

14 was required. Initially, the glycosyl donor 29a (ref. 31) with chloroacetyl group at the O-2 position was chosen for the one-pot construction of pentasaccharide 15a (Supplementary Table 1), but the overall yield was moderate (43%). Fortunately, when the donor 29 equipped with a levulinoyl group was employed for the one-pot glycosylation reaction, pentasaccharide 15 was rapidly assembled in 76% overall yield (Fig. 3b). Ultimately, deacylation of 15 gave diol 14 (94% yield).

Our attention was then turned to the synthesis of heptasaccharide 10, which involved the stereocontrolled installation of two challenging b-arabinofuranosyl linkages. Among a number of innovative glycosyl donors developed by several groups25,3137, perbenzyl-protected thioglycoside 13a (ref. 25) and 3,5-O-tetraiso propyldisiloxanylidene-protected thioglycosides 13b,c (ref. 31) were synthesized for the current purpose (Supplementary Table 2). Although these donors proved useful for direct b-arabinofuranosylation, whether they could be subjected to thioglycoside acceptor 14 under the donor-preactivation conditions38 remained to be explored. After some optimization, the best result arised when 4.0 equiv. of 13b was preactivated by p-TolSCl/AgOTf and subsequently glycosylated with 1.0 equiv. of diol 14, delivering heptasaccharide 30b with good stereoselectivity (b,b-isomer/other isomers 9/1). Removal of

the silyl groups in 30b afforded 31 (74% over two steps, Fig. 3c), in which the newly formed b-arabinofuranosyl linkages were

4 NATURE COMMUNICATIONS | 8:14851 | DOI: 10.1038/ncomms14851 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14851 ARTICLE

a

c

27 27

OR

BnO

O

R O

BnO O

O

HO

OBz

STol

O

27

27

(7)

OBn

O OBz

O

Si O STol

Si

14

OBn

OBn

O OBz O

OBz

O

O

OBn

OBn

O

BnO O

OR

O

O

STol

O

O

O

OBn

TBSO

OBz

STol

O

26

R O

OBz

O

OBz

OBn

O OBz

O

O

13b

R O BnO

OBn

OBn

O

OBn

R

4

17: R1 = TBS, R2 = STol

32: R1 = H, R2 = OC8H17

12: R1 = H, R2 = STol

30b: R1, R2 = -TIPDS-

92% (3)

(4), (3)

74% (8)

86%2 steps (9) 97%

2 steps

31: R1= R2 = H 10: R1= R2 = Bz

b

d

11

HO

OOBz

STol

OR

R O

O

HO

10

12

12 or 32

HO

OOBz

STol

OBn

R O

28 27

R O O

O

OR

O OR

OR

O OR O

OR

TBSO

BnO

O

O

OOBz

STol

or

OOLev

STol

(5)

OR

O

R O O

OR

O

O

27

O

OR

OBn

OBn

O

26

OR

OR

OOR

29

R O R O

O

OR

O OR

O

OR

R O

OR

O OR

O

O

OOR

OBn

OOR

OR

R O

O

O

R

O

OR

OBn

O OBz O

OBn

O OBz

O STol

OBz

O

13

O

R O

R O O

O

O

OBn

OR

O

O

OR

O OR

O OR

O

O

OR

R O

16: R1 = TBS, R2 = Bz 11: R1 = H, R2 = Bz

O

OR

95% (3)

94% (6)

R O R O

O

OR

O

R O O

OR

O

3: R1 = Bz, R2 = Bn, R3 = STol

15: R1 = Bn, R2 = Lev 14: R1 = Bn, R2 = H

33: R1 = Bz, R2 = Bn, R3 = OC8H17 34: R1 = R2 = H, R3 = OC8H17

(13), (14)

83%, 2 steps

Figure 3 | Synthesis of Araf31 donor 3. (a) Synthesis of oligosaccharides 12 and 32; (b) Synthesis of diols 11 and 14; (c) Synthesis of heptasaccharide 10; (d) Synthesis of Araf31 donor 3 and 31-mer arabinan 34. Reagents and conditions: (1) TTBP, 4 MS, CH2Cl2, p-TolSCl, AgOTf, then 27, 78 C to room

temperature; (2) p-TolSCl, AgOTf, then 27, 78 C to room temperature; (3) TBAF, AcOH, THF; (4) p-TolSCl, AgOTf, TTBP, 1-octanol, 4 MS, CH2Cl2, 78 C; (5) TTBP, 4 MS, CH2Cl2, p-TolSCl, AgOTf, then 28, 78 C to room temperature; (6) H2NNH2-AcOH, THF/CH3OH (10:1); (7) p-TolSCl, AgOTf,

then 14, 78 C; (8) TBAF, THF; (9) Bz2O, DMAP, pyridine, CH2Cl2, reux; (10) TTBP, 4 MS, CH2Cl2, p-TolSCl, AgOTf, then 11, 78 C to room

temperature; (11) p-TolSCl, AgOTf, then 12, 78 C to room temperature; (12) p-TolSCl, AgOTf, then 12 or 32, 78 C to room temperature; (13) NaOCH3,

CH3OH/CH2Cl2 (2:1); (14) Pd/C, H2, EtOAc/THF/1-PrOH/H2O (2:1:1:1). TBAF, tetra-n-butylammonium uoride.

conrmed by the 13C NMR spectrum (appearance at 99.6 and99.1 p.p.m.)39. Finally, the re-protection of 31 with benzoyl groups yielded the desired heptasaccharide 10 (97% yield).

With oligosaccharide building blocks 1012 in hand, the assembly of Araf31 donor 3 by preactivation-based one-pot glycosylation protocol was attempted. This one-pot reaction was expected to be more challenging due to the steric hindrance in the double glycosylation of Araf5 acceptor 11 using

Araf7 donor 10. Surprisingly, the reaction proceeded smoothly when 2.3 equiv. of 10 was reacted with 1.0 equiv. of 11, delivering an Araf19 intermediate, which was sequentially coupled with two Araf6 acceptors 12 in a single ask without any intermediate isolation to afford the Araf31 donor 3 in 70% overall yield (Fig. 3d). To further conrm the identity of this 31-mer polysaccharide, an Araf6 acceptor 32 bearing an alkyl group at the reducing end was synthesized (Fig. 3a). Thus a four-component one-pot coupling reaction of oligosaccharides 1012 and 32 gave a similar 31-mer polysaccharide 33 in 65% yield (Fig. 3d), which was fully deprotected via deacylation and hydrogenolysis to afford the arabinan 34 ([M Na]

monoisotopic m/z calcd. for 4246.4, found: 4246.3; [M K]

monoisotopic m/z calcd. for 4262.4, found: 4262.2). Gratifyingly, the 1H and 13C NMR data of 34 were found to be identical

with previous reports14,40 except for the differences in some repeating units.

Assembly of arabinogalactan 1. Our nal task was the glycosylation of Galf30 acceptor 2 with Araf31 donor 3 to nish the assembly of target polysaccharide. To the best of our knowledge, no glycosylation reactions between polysaccharide sequences composed of 420 units were reported to date. For the planned [31 31 30] coupling reaction, it was anticipated that the

biggest challenge would come from the steric hindrance by the bulky size of both the donor and acceptor, especially when a double glycosylation was required. Indeed, when a wide variety of promoter systems such as p-TolSCl/AgOTf29, NIS/AgOTf41, NIS/TfOH42, N-(p-methylphenylthio)-e-caprolactam/Tf2O43, TBPA44, Ph3Bi(OTf)2 (ref. 45), BSP/Tf2O46 and Ph2SO/Tf2O47 were examined (Supplementary Table 3), no double glycosylation product or only some monoglycosylation product was observed before the donor decomposed, prompting us to further screen the reaction conditions. Encouragingly, it was found that benzenesulnyl morpholine/triic anhydride (BSM/ Tf2O)48 developed by our group is the most effective promoter.

And indeed, when promoted by BSM/Tf2O, this double

NATURE COMMUNICATIONS | 8:14851 | DOI: 10.1038/ncomms14851 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 5

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14851

OR1

R2O

O

Araf31 donor 3 + Galf30 acceptor 2

(1)

R1O

R2O O

O

84%

OR2

O OR1

OR2

O OR

1 O

OR1

O

O

OR2

O

R2O O

OR1

O

O

OR2

OC8H17

O

O

R2OO

R2O

O

OR1

OR2

OOR1

R2OOR2O OR

1

R1O R

2O

O

OR2

O OR1

O

OR1

R2OO

R2O

O OR1

O

OR2

O OR1

O

OR1

R2O

O

O

OR2

O

O

R2OO

O

OR1

3

R1O

13

R2O O

OR1

O

OR2

O

O

OR1

O OR2

O OR1

R2OO

R2O

O

R2OO R2O

O

O

OR1

OR2

O

R2O O

OR1

O

O

OR1

O

2

OR2

O OR1

R2OO

R1O R

2O

O

O

R2OO

R2O

(2), (3)

75%, 2 steps

Figure 4 | Assembly of arabinogalactan 1. Reagents and conditions: (1) BSM, Tf2O, 4 MS, CH2Cl2, 40 C; (2) NaOCH3, CH3OH/THF (2:1);

(3) Pd/C, H2, EtOAc/THF/1-PrOH/H2O (2:1:1:1). BSM, benzenesulnyl morpholine.

OR1

35: R1 = Bz, R2 = Bn

1: R1 = R2 = H

R2O

9

1

R1O OR

glycosylation was extremely clean and complete (indicated by thin-layer chromatography analysis), delivering the fully protected arabinogalactan 35 in 84% yield (Fig. 4). Although signals of the anomeric protons in 1H NMR spectrum were obscured due to the extensive overlapping, the anomeric carbons were distinctive in 13C NMR spectrum (all anomeric carbons of a-Araf residues and b-Galf residues were between 105 and 107 p.p.m., and anomeric carbons of b-Araf residues appeared at 101.0 and 100.6 p.p.m.). The identity of 35 was further supported by its MALDI-TOF mass spectrum ([M Na] m/z calcd. for

33885.4, found: 33884.7). Finally, the global deprotection of 35 by the successive deacylation and hydrogenolysis was conducted, affording the target polysaccharide arabinogalactan 1 successfully.

DiscussionWe have developed a concise and highly efcient strategy for the rst total synthesis of 92-mer mycobacterial arabinogalactan 1. This work not only represents the longest well-dened carbohydrate chain synthesis up to date, but also provides useful compounds as probes for further investigations on mycobacterial cell wall-related biological events. Our synthetic strategy highlights a series of efcient preactivation-based one-pot glycosylation reactions to minimize the synthetic steps, the stereoselective b-arabinofuranosylation by preactivation protocol and the convergent [31 31 30] double glycosylation reaction,

thus offering a straightforward access to the target polysaccharide. Our work may open an avenue to the synthesis of complex polysaccharides with biological importance that are either difcult or impossible to access through isolation or semisynthesis.

Methods

General. The complete experimental details and compound characterization data can be found in Supplementary Methods. For the NMR, HPLC and MALDI-TOF mass spectra of the compounds in this article, see Supplementary Figs 5126.

General procedure for preactivation-based one-pot glycosylation reaction.

A mixture of glycosyl donor, TTBP and freshly activated 4 molecular sieves in anhydrous CH2Cl2 under argon atmosphere was stirred for 20 min at room

temperature and cooled to 78 C. After 5 min, stoichiometric amount of

p-TolSCl was added to the mixture, followed by the addition of AgOTf. After another 5 min, a solution of glycosyl acceptor in anhydrous CH2Cl2 was slowly added. The resulting mixture was slowly warmed to room temperature within 2 h, stirred for another 20 min and cooled back to 78 C. The glycosylation operation

mentioned above was repeated until the generation of the desired product.

Data availability. The authors declare that the data supporting the ndingsof this study are available within the article and its Supplementary Information les. And all data are available from the authors upon reasonable request.

References

1. Varki, A. Biological roles of oligosaccharides: all of the theories are correct. Glycobiology 3, 97130 (1993).

2. Bertozzi, C. R. & Kiessling, L. L. Chemical glycobiology. Science 291, 23572364 (2001).

3. Boltje, T. J., Buskas, T. & Boons, G.-J. Opportunities and challenges in synthetic oligosaccharide and glycoconjugate research. Nat. Chem. 1, 611622 (2009).

4. Danishefsky, S. J., McClure, K. F., Randolph, J. T. & Ruggeri, R. B. A strategy for the solid-phase synthesis of oligosaccharides. Science 260, 13071309 (1993).

5. Yamada, H., Harada, T., Miyazaki, H. & Takahashi, T. One-pot sequential glycosylation: a new method for the synthesis of oligosaccharides. Tetrahedron Lett. 35, 39793982 (1994).

6. Zhang, Z. et al. Programmable one-pot oligosaccharide synthesis. J. Am. Chem. Soc. 121, 734753 (1999).

7. Plante, O. J., Palmacci, E. R. & Seeberger, P. H. Automated solid-phase synthesis of oligosaccharides. Science 291, 15231527 (2001).

8. Nigudkar, S. S. & Demchenko, A. V. Stereocontrolled 1,2-cis glycosylation as the driving force of progress in synthetic carbohydrate chemistry. Chem. Sci. 6, 26872704 (2015).

9. Wang, C.-C. et al. Regioselective one-pot protection of carbohydrates. Nature 446, 896899 (2007).

10. Matsuzaki, Y., Ito, Y., Nakahara, Y. & Ogawa, T. Synthesis of branched poly-N-acetyl-lactosamine type pentaantennary pentacosasaccharide: glycan part of a glycosyl ceramide from rabbit erythrocyte membrane. Tetrahedron Lett. 34, 10611064 (1993).

11. Pozsgay, V. A new strategy in oligosaccharide synthesis using lipophilic protecting groups: synthesis of a tetracosasaccharide. Tetrahedron: Asymmetry 11, 151172 (2000).

12. Fraser-Reid, B., Lu, J., Jayaprakash, K. N. & Lpez, J. C. Synthesis of a 28-mer oligosaccharide core of Mycobacterial lipoarabinomannan (LAM) requires only two n-pentenyl orthoester progenitors. Tetrahedron: Asymmetry 17, 24492463 (2006).

6 NATURE COMMUNICATIONS | 8:14851 | DOI: 10.1038/ncomms14851 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14851 ARTICLE

13. Cheon, H.-S., Lian, Y. & Kishi, Y. Highly stereoselective and iterative synthesis of a-(1-4)-linked polysaccharides composed of 3-O-methyl-D-mannose. Org.

Lett. 9, 33233326 (2007).14. Joe, M., Bai, Y., Nacario, R. C. & Lowary, T. L. Synthesis of the docosanasaccharide arabinan domain of mycobacterial arabinogalactanand a proposed octadecasaccharide biosynthetic precursor. J. Am. Chem. Soc. 129, 98859901 (2007).

15. Ishiwata, A. & Ito, Y. Synthesis of docosasaccharide arabinan motif of mycobacterial cell wall. J. Am. Chem. Soc. 133, 22752291 (2011).

16. Calin, O., Eller, S. & Seeberger, P. H. Automated polysaccharide synthesis: assembly of a 30mer mannoside. Angew. Chem. Int. Ed. 52, 58625865 (2013).

17. Hansen, S. U., Miller, G. J., Cliff, M. J., Jayson, G. C. & Gardiner, J. M. Making the longest sugars: a chemical synthesis of heparin-related [4]n oligosaccharides from 16-mer to 40-mer. Chem. Sci. 6, 61586164 (2015).

18. Islam, M., Shinde, G. P. & Hotha, S. Expedient synthesis of the heneicosasaccharyl mannose capped arabinomannan of the Mycobacterium tuberculosis cellular envelope by glycosyl carbonate donors. Chem. Sci. doi:10.1039/C6SC04866H (2017).

19. Brennan, P. J. Structure, function, and biogenesis of the cell wall of Mycobacterium tuberculosis. Tuberculosis 83, 9197 (2003).

20. Alderwick, L. J. et al. Deletion of Cg-emb in corynebacterianeae leads to a novel truncated cell wall arabinogalactan, whereas inactivation of Cg-ubiA results in an arabinan-decient mutant with a cell wall galactan core. J. Biol. Chem. 280, 3236232371 (2005).

21. Bhamidi, S. et al. The identication and location of succinyl residues and the characterization of the interior arabinan region allow for a model of the complete primary structure of Mycobacterium tuberculosis mycolyl arabinogalactan. J. Biol. Chem. 283, 1299213000 (2008).

22. Bhamidi, S. et al. Detailed structural and quantitative analysis reveals the spatial organization of the cell walls of in Vivo grown Mycobacterium leprae and in vitro grown Mycobacterium tuberculosis. J. Biol. Chem. 286, 2316823177 (2011).

23. Lowary, T. L. Twenty years of mycobacterial glycans: furanosides and beyond. Acc. Chem. Res. 49, 13791388 (2016).

24. Mereyala, H. B., Hotha, S. & Gurjar, M. K. Synthesis of pentaarabinofuranosyl structure motif A of Mycobacterium tuberculosis. J. Chem. Soc. Chem. Commun. 685686 (1998).

25. DSouza, F. W. & Lowary, T. L. The rst total synthesis of a highly branched arabinofuranosyl hexasaccharide found at the nonreducing termini of mycobacterial arabinogalactan and lipoarabinomannan. Org. Lett. 2, 14931495 (2000).

26. Completo, G. C. & Lowary, T. L. Synthesis of galactofuranose-containing acceptor substrates for mycobacterial galactofuranosyltransferases. J. Org. Chem. 73, 45134525 (2008).

27. Wang, S., Meng, X., Huang, W. & Yang, J.-S. Inuence of silyl protections on the anomeric reactivity of galactofuranosyl thioglycosides and application of the silylated thiogalactofuranosides to one-pot synthesis of diverse b-D-oligogalactofuranosides. J. Org. Chem. 79, 1020310217 (2014).

28. Kandasamy, J., Hurevich, M. & Seeberger, P. H. Automated solid phase synthesis of oligoarabinofuranosides. Chem. Commun. 49, 44534455 (2013).

29. Huang, X., Huang, L., Wang, H. & Ye, X.-S. Iterative one-pot synthesis of oligosaccharides. Angew. Chem. Int. Ed. 43, 52215224 (2004).

30. Liang, X.-Y., Deng, L.-M., Liu, X. & Yang, J.-S. Efcient one-pot syntheses of a-D-arabinofuranosyl tri- and tetrasaccharides present in cell wall polysaccharide of Mycobacterium tuberculosis. Tetrahedron 66, 8793

2010:

31. Ishiwata, A., Akao, H. & Ito, Y. Stereoselective synthesis of a fragment of mycobacterial arabinan. Org. Lett. 8, 55255528 (2006).

32. Gadikota, R. R., Callam, C. S., Wagner, T., Fraino, B. D. & Lowary, T. L. 2,3-Anhydro sugars in glycoside bond synthesis. Highly stereoselective syntheses of oligosaccharides containing a- and b-arabinofuranosyl linkages.

J. Am. Chem. Soc. 125, 41554165 (2003).33. Lee, Y. J., Lee, K., Jung, E. H., Jeon, H. B. & Kim, K. S. Acceptor-dependent stereoselective glycosylation: 20-CB glycoside-mediated direct b-D-arabinofuranosylation and efcient synthesis of the octaarabinofuranoside in mycobacterial cell wall. Org. Lett. 7, 32633266 (2005).

34. Zhu, X., Kawatkar, S., Rao, Y. & Boons, G.-J. Practical approach for the stereoselective introduction of b-arabinofuranosides. J. Am. Chem. Soc. 128, 1194811957 (2006).

35. Imamura, A. & Lowary, T. L. b-Selective arabinofuranosylation using a 2,3-O-xylylene-protected donor. Org. Lett. 12, 36863689 (2010).

36. Thadke, S. A., Mishra, B. & Hotha, S. Facile synthesis of b- and a-arabinofuranosides and application to cell wall motifs of M. tuberculosis. Org. Lett. 15, 24662469 (2013).

37. Liu, Q.-W., Bin, H.-C. & Yang, J.-S. b-Arabinofuranosylation using 5-O-(2-quinolinecarbonyl) substituted ethyl thioglycoside donors. Org. Lett. 15, 39743977 (2013).

38. Crich, D., Pedersen, C. M., Bowers, A. A. & Wink, D. J. On the use of 3,5-O-benzylidene and 3,5-O-(di-tert-butylsilylene)-2-O-benzylarabinothiofuranosides and their sulfoxides as glycosyl donors for the synthesis of b-arabinofuranosides: importance of the activation method. J. Org. Chem. 72, 15531565 (2007).

39. Mizutani, K., Kasai, R., Nakamura, M. & Tanaka, O. N.M.R. spectral study of a- and b-L-arabinofuranosides. Carbohydr. Res. 185, 2738 (1989).

40. Lee, R. E. B., Li, W., Chatterjee, D. & Lee, R. E. Rapid structure characterization of the arabinogalactan and lipoarabinomannan in live mycobacterial cells using 2D and 3D HR-MAS NMR: structural changes in the arabinan due to ethambutol treatment and gene mutation are observed. Glycobiology 15, 139151 (2005).

41. Konradsson, P., Udodong, U. E. & Fraser-Reid, B. Iodonium promoted reactions of disarmed thioglycosides. Tetrahedron Lett. 31, 43134316 (1990).

42. Vwneman, G. H., van Leeuwen, S. H. & van Boom, J. H. Iodonium ion promoted reactions at the anomeric centre. II. An efcient thioglycoside mediated approach toward the formation of 1,2-trans linked glycosides and glycosidic esters. Tetrahedron Lett. 31, 13311334 (1990).

43. Maity, S. K., Basu, N. & Ghosh, R. Efcient activation of thioglycosides with N-(p-methylphenylthio)-e-caprolactam-TMSOTf. Carbohydr. Res. 354, 4048 (2012).

44. Marra, A., Mallet, J.-M., Amatore, C. & Sina, P. Glycosylation usinga one-electron-transfer homogeneous reagent: a novel and efcient synthesis of b-linked disaccharides. Synlett 572574 (1990).

45. Goswami, M., Ellern, A. & Pohl, N. L. B. Bismuth(V)-mediated thioglycoside activation. Angew. Chem. Int. Ed. 52, 84418445 (2013).

46. Crich, D. & Smith, M. 1-Benzenesulnyl piperidine/triuoromethanesulfonic anhydride: a potent combination of shelf-stable reagents for the low-temperature conversion of thioglycosides to glycosyl triates andfor the formation of diverse glycosidic linkages. J. Am. Chem. Soc. 123, 90159020 (2001).

47. Code, J. D. C. et al. Ph2SO/Tf2O: a powerful promotor system in chemoselective glycosylations using thioglycosides. Org. Lett. 5, 15191522 (2003).

48. Wang, C., Wang, H., Huang, X., Zhang, L.-H. & Ye, X.-S. Benzenesulnyl morpholine: a new promoter for one-pot oligosaccharide synthesis using thioglycosides by pre-activation strategy. Synlett 28462850 (2006).

Acknowledgements

This work was nancially supported by the grants from the National Natural Science Foundation of China (21232002) and the Ministry of Science and Technology of China (2013CB910700, 2012CB822100). We thank Professor Qin Li and Professor Lijun Zhong at Peking University Health Science Center for their helpful assistance in analysis of glycan structures.

Author contributions

D.-C.X. and X.-S.Y. conceived the research. Y.W., D.-C.X. and X.-S.Y. designed the experiments. Y.W. performed all the experiments. S.-C.C. and Y.-S.W. synthesized some building blocks. Y.W., D.-C.X. and X.-S.Y. analyzed the data. Y.W. and X.-S.Y. wrote the manuscript. X.-S.Y. supervised the project.

Additional information

Supplementary Information accompanies this paper at http://www.nature.com/naturecommunications

Web End =http://www.nature.com/ http://www.nature.com/naturecommunications

Web End =naturecommunications

Competing interests: The authors declare no competing nancial interests.

Reprints and permission information is available online at http://npg.nature.com/reprintsandpermissions/

Web End =http://npg.nature.com/ http://npg.nature.com/reprintsandpermissions/

Web End =reprintsandpermissions/

How to cite this article: Wu, Y. et al. Total synthesis of mycobacterial arabinogalactan containing 92 monosaccharide units. Nat. Commun. 8, 14851 doi: 10.1038/ncomms14851 (2017).

Publishers note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional afliations.

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the articles Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

Web End =http://creativecommons.org/licenses/by/4.0/

r The Author(s) 2017

NATURE COMMUNICATIONS | 8:14851 | DOI: 10.1038/ncomms14851 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 7