Synthesis of carbasugar-containing non-glycosidically linked pseudodisaccharides and higher pseudooligosaccharides

Carbohydrate Research 344 (2009) 2285–2310

Contents lists available at ScienceDirect

Carbohydrate Research

journal homepage: w ww.elsevier.com/locate/carres

Minireview

Synthesis of carbasugar-containing non-glycosidically linked

pseudodisaccharides and higher pseudooligosaccharides

*

Ian Cumpstey

Department of Organic Chemistry, Arrhenius Laboratory, Stockholm University, 106 91 Stockholm, Sweden

a r t i c l e i n f o

a b s t r a c t

Article history:

Received 26 June 2009

Received in revised form 2 September 2009

Accepted 9 September 2009

Available online 16 September 2009

This minireview covers synthetic methods towards carbasugar-containing non-glycosidically linked

pseudodisaccharides or higher pseudooligosaccharides. Carbocyclic pyranose mimetics (saturated or

unsaturated between C-5 and C-5a) are linked by ether, thioether or amine bridges to carbohydrates

or other carbasugars.

Keywords:

Carbasugars

Pseudodisaccharides

Epoxides

Triﬂates

Coupling

Amines

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2286

2. N-Linked . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2289

2.1. Epoxide and aziridine opening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2289

2.2. Reductive amination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2297

2.3. Alkylation: halide or sulfonate displacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2298

2.4. Palladium-catalysed allylic coupling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2300

2.5. Intramolecular reaction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2300

3. O-Linked . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2300

3.1. Triflate displacement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2300

3.2. Epoxide opening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2301

3.3. Other methods for ether-linked carbadisaccharide synthesis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2303

4. S-Linked . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2304

4.1. Epoxide opening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2304

4.2. Triflate displacement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2304

5. Summary, conclusions and outlook. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2304

5.1. Carbasugar electrophiles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2304

5.1.1. 5a-Carba-

5.1.2. 5a-Carba-

5.1.3. 5a-Carba-

5.1.4. 5a-Carba-

a

-mannopyranosides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2304

-glucopyranosides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2304

-xylo-hex(5,5a)enopyranosides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2304

-xylo-hex(5,5a)enopyranosides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2306

5.1.5. 5a-Carba-b-mannopyranosides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2306

5.1.6. 5a-Carba-b-galactopyranosides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2306

5.1.7. 5a-Carba-b-glucopyranosides. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2307

* Tel.: 0046 8 16 2481.

E-mail address: cumpstey@organ.su.se

doi:10.1016/j.carres.2009.09.008

2286

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

5.2. Carbasugar nucleophiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2307

5.2.1. 4-Substituted glucose. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2307

5.2.2. 3-Substituted glucose. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2309

5.2.3. 2-Substituted glucose. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2309

5.2.4. 6-Substituted glucose. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2309

5.2.5. 4-Substituted mannose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2309

5.2.6. 4-Substituted xylose. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2309

5.3. Conclusions and outlook. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2309

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2310

1. Introduction

Several N-linked pseudooligosaccharides containing carbocyclic

valienamine, valiolamine or validamine moieties occur naturally,

and some of these are shown in Figure 1.¹Some motivation for

their synthesis was found in the desire to prove the structures pro-

posed following isolation. The biological activity, in particular glu-

cosidase inhibitory activity, of some of these natural structures has

prompted the synthesis of analogues to help understand their bio-

logical mode of action or to try to alter or improve speciﬁcity and

afﬁnity for a glucosidase or other glycosidase of interest.^2,3Carba-

sugar pseudooligosaccharides have also been shown to interact

with other carbohydrate-binding proteins; some ether-linked car-

basugar pseudodisaccharides have been found to act as substrates

for glycosyltransferases;²some have been found to bind to lectins.⁴

The assessment of biological activity of the pseudodisaccharide

derivatives is not covered here, and neither are deprotection reac-

tions; the interested reader is referred to the original articles or to

review articles aimed more at these aspects.^2,3,5The synthesis of

This minireview concerns the synthesis of carbasugar-contain-

ing non-glycosidically linked pseudodisaccharides (or higher pseu-

dooligosaccharides) in which a carbocyclic pyranose mimetic,

either saturated or unsaturated between C-5@C-5a, is linked at

C-1 to a carbohydrate (or other carbasugar) by a one-atom nitro-

gen, oxygen or sulfur bridge. Formal replacement of a ring-oxygen

by a methylene group to give saturated carbadisaccharides is con-

ceptually the simplest modiﬁcation that leads to a hydrolytically

stable pseudodisaccharide of this type (Fig. 1a). Exchange of the in-

ter-glycosidic oxygen for nitrogen or sulfur and introduction of

unsaturation between C-5 and C-5a are further modiﬁcations that

have been studied.

HO

5a

2

O

OH

O

OH

O

1

6

5

4

O

OH

O

OR

HO

3

RO

HO

OR

Disconnection 1:

Allylic electrophile

α-maltoside

5a'-carba-α-maltoside

RO

X

Disconnection 2:

Carbohydrate electrophile

RO

HO

5a

6

5

1

RO

OR

HO

4

OH

N

2

H₃C

OH

H

O

OR

HO

3

O

N

HO

OH

H

HO

Salbostatin

O

OR

HO

OR

RO

HO

R = Glc₃: oligostatin C

R = Me : methyl oligobiosaminide

Disconnection 1:

H₃C

HO

RO

Carbasugar electrophile

O

N

OH

X

H

HO

Disconnection 2:

Carbohydrate electrophile

RO

HO

OR

HO

X¹

OH

R = Me : methyl acarviosin

R = Glc₂: acarbose

R = Glc : amylostatin (XG)

RO

OR

N

H

OH

OR

OH

X²

HO

Figure 2. Retrosynthetic disconnection of carbasugar pseudodisaccharides.

X¹= H, X²= H, ValidoxylamineA

X¹= OH, X²= H, ValidoxylamineB

X¹= H, X²= OH, Validoxylamine G

HO

OH

O

N

OH

H

HO

RO

OR

RO

O

RO

R = Glc : adiposin-1

R = Glc₂: adiposin-2

O

RO

OR

O

HO

saturated β

1,2-epoxide

saturated α

1,2-epoxide

saturated

1,5a-epoxide

HO

OH

H₂N

NH₂

OH

RO

HO

RO

HO

valiolamine

O

RO

valienamine

validamine

O

Figure 1. (a)

A

disaccharide and its O-linked carbasugar analogue; (b) some

unsaturated 'up'

1,2-epoxide

unsaturated 'down'

1,2-epoxide

naturally occurring carbasugar-containing pseudosaccharide structures; and (c)

carbocyclic amine substructures. This paper uses standard carbohydrate numbering

for carbasugars reﬂecting their homomorphic relationships to glycosides.

Figure 3. Main classes of carbasugar epoxides.

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

2287

the carbasugar building blocks has been reviewed elsewhere,^5,6so

is not included here. The focus of this minireview is on the chem-

istry leading to such pseudodisaccharide structures; usually by

coupling of a carbasugar to a carbohydrate, and where necessary

followed by further modiﬁcation.

ity of the allylic electrophiles in S_Nreactions⁷may make carbasug-

ar electrophiles (disconnection 1 in Fig. 2a) a better choice than

non-carbasugar electrophiles (disconnection 2). However, such

an allylic carbasugar electrophile could also be prone to a favour-

able elimination reaction leading to a conjugated diene.

The chemistry used to link a carbasugar to a carbohydrate by

ether, thioether or secondary amine bridges is quite different from

that used to synthesise a glycosidic bond. S_Nreactions have been

used for the synthesis of O- and S-linked pseudodisaccharides,

while both S_Nreactions and reductive amination have been used

for the synthesis of N-linked pseudodisaccharides. The bridging

bond may be disconnected retrosynthetically to give a carbasugar

C-1 electrophile and a carbohydrate nucleophile or to give a carba-

sugar C-1 nucleophile and a carbohydrate electrophile (Fig. 2).

The two disconnections are similar, but even discounting the

vagaries of real systems, they are non-equivalent. The C-1 position

is allylic in C-5@C-5a unsaturated systems. The enhanced reactiv-

S_Nreactions on carbohydrates (at carbons other than the ano-

meric carbon) are difﬁcult: the many electron-withdrawing groups

on the ring destabilise potential carbocations making S_N1 reactions

very unusual away from the anomeric centre. S_N2 reactions are dis-

favoured by the presence of electron-withdrawing groups in the b-

position to a leaving group (cf. disconnections 2 in Fig. 2a and b).⁸

C-1 of a carbasugar is ﬂanked by an alien methylene group (C-5a)

that does not bear an electron-withdrawing oxygen, which means

that it may be more electrophilic in S_N2 reactions than a carbon in

a carbohydrate ﬂanked by two carbons each bearing an electron-

withdrawing oxygen functionality. Stereoelectronic factors are also

important, with, for example, the Fürst–Plattner favoured trans-

Table 1

Synthesis of 1,2-trans a-manno imino-linked pseudodisaccharides and pseudotrisaccharides by opening of saturated b-manno 1,2-anhydro carbasugars

Entry

Epoxide

Amine

Conditions/t

Product

Yield

70%

Ref.

10

BnO

OH

NH

HO

NH₂

HO

BnO

N

O

OH

O

H

HO

BnO

1

A/5 days

HO

N

OBn

OH

O

3

H

1 (1.1 equiv.)

BnO

MeO

OBn

10

MeO

N₃

HO

OH

N₃

HO

H₂N

N

O

OH

O

H

2

1 (1.5 equiv)

A/6 days

61%

10

BnO

OBn

OMe

11

4

MeO

HO

OH

HO

H₂N

OH

N

H

O

OH

O

BnO

3

4

5

1 (1.5 equiv)

1 (2.2 equiv)

A/6 days

A/5 days

95%

72%

59%

10

11

OBn

OMe

12

5

MeO

H₂N

HO

H₂N

OH

O

10

6

AcO

OAc

HO

H₂N

HO

OH

HO

O

AcO

OAc

i) B/24 h

ii) Acetylate

O

HN

AcO

OMe

13

2

7 (1.3 equiv.)

OMe

AcO

OAc

H₃COH

AcO

O

H₂N

HO

AcO

H₃COAc

6

7

2

i) B/24 h

ii) Acetylate

56%

68%

11

O

HN

OMe

AcO

14

8 (1.5 equiv.)

O

OAc

O

OH

AcO

O

2

i) B/3 days

ii) Acetylate

AcO

OH

OAc

N

H

H₂N

AcO

15

9 (1.5 equiv.)

Reaction conditions: (A) propan-2-ol, sealed tube, 120 °C; (B) propan-2-ol/DMF, sealed tube, 100–120 °C.

2288

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

Table 2

Synthesis of 1,2-trans b-gluco and b-galacto N-linked pseudodisaccharides by opening of saturated

a-gluco or -galacto 1,2-anhydro carbasugars

Entry

Epoxide

Amine

t

Product

BnO

Yield

40%

Ref.

12

BnO OBn

OH

OBn

HN

HO

O

BnO

H₂N

HO

1

BnO

HO

OMe

HO

OMe

O

21

16

28

HO

Ph

O

OH

O

BnO

H

N

ca 30%

12

HO

Ph

O

HO

OMe

O

2

21

29

BnO

HO

O

OH

Ph

17

O

BnO

H

N

HO

O

OH

HO

OMe

30

O

OR

O

OR

OAc

OH

H

N

AcO

O

H₂N

HO

O

OMe

3

i) 2 Weeks

ii) Acetylate

31/66%

32/51%

12

15

OMe

OAc

O

OAc

OH

31, R = Bn

22 (1.3 equiv.)

18, R = Bn

32, R = TBDMS

19, R = TBDMS

OAc

OH

O

OBn

O

AcO

HO

4

5

18 (1.5 equiv)

i) 3 Weeks

ii) Acetylate

24%

38%

OMe

NHAc

OMe

NHAc

N

H₂N

H

OAc

23

33

O

OH

OBn

OAc

O

H

O

H₂N

HO

N

18 (10 equiv)

18 (2 equiv)

18 (1.1 equiv)

i) 3 Weeks

ii) Acetylate

14

O(CH₂)₇CH₃

AcO

NHAc

OAc

OBn

NHAc

24

34

OAc

O

AcO

38%

24%

N

H

O

NHAc

OH

OAc

O

35

HO

6

i) 3 Weeks

ii) Acetylate

OAc

O(CH₂)₇CH₃

NHAc

H₂N

OBn

AcO

HN

O

25

O(CH₂)₇CH₃

AcHN

O

36

OAc

O

OH

OBn

OAc

H

N

O

H₂N

7

8

i) 2 Weeks

ii) Acetylate

25%

40%

13

16

OMe

HO

OMe

AcO

OAc

NHAc

26

37

O

H₃C

HO

H₃C

NH₂

OH

H₃C

HO

O

H₃C

N

H

OH

27

OH

HO

O

38

20 (1.5 equiv.)

Reaction conditions: propan-2-ol, sealed tube, 120 °C.

diaxial products tending to dominate in epoxide-opening reac-

tions.⁹Alternative disconnections where the bridging C–X bonds

are formed before formation of the carbasugar ring have also been

put into practice.

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

2289

OBn

OH

OBn

O

N

H

O

propan-2-ol,

sealed tube,

Ph

O

120 ºC, 4 days

22

44 / 72%

O

+

O

BnO

O

OBn

Ph

NH₂

HO

39

42 (1 equiv.)

OBn

O

N

H

O

Ph

45 / 24%

BnO

OBn

OH

OBn

N

OBn

H

BnO

propan-2-ol,

sealed tube,

120 ºC, 110 h

OBn

Ph

O

18–21

O

BnO

46 / 75%

O

+

BnO

O

OBn

BnO

Ph

NH₂

42 (1.1 equiv.)

40

HO

O

OBn

N

OBn

H

BnO

O

Ph

OBn

47 / 15%

OBn

N

OBn

H

BnO

OAc

OBn

O

i) propan-2-ol,

sealed tube,

120 ºC, 125 h

OBn

OAc

OBn

48 / 47%

OBn

OAc

AcO

18,19

O

BnO

+

O

OAc

O

OAc

O

AcO

ii) acetylate

OAc

OBn

BnO

NH₂

OAc

AcO

40

43

OAc

(1 equiv.)

N

OBn

H

BnO

OAc

OBn

AcO

O

AcO

49 / 30%

OAc

AcO

i) propan-2-ol,

sealed tube,

120 ºC, 92 h

Ph

O

AcO

17

NHR

O

55

RNH₂

BnO

+

ii) hydrogenolyse

iii) acetylate

OBn

AcO

42

NHR

OAc

AcO

RNH₂:

CH₃

56

OH

O

OH

O

OH

O

CH₃

O

HO

H₂N

HO

H₂N

HO

H₂N

HO

H₂N

HO

53

54

OMe

21

51

52

50

OMe

1.2 equiv. amine

diaxial, 34%

diequatorial, 11%

1.2 equiv. amine

diaxial, 43%

diequatorial, 9%

1.4 equiv. amine

diaxial, 51%

diequatorial, 8%

1.4 equiv. amine

diaxial, 44%

diequatorial, 5%

1.2 equiv. amine

diaxial, 57%

diequatorial, 10%

1.2 equiv. amine

diaxial, 48%

diequatorial, 5%

Scheme 1. Synthesis of pseudodisaccharides and pseudotrisaccharides from 1,5a-epoxides.

synthesis of N-linked pseudodisaccharide or higher pseudooligo-

saccharide structures. Carbasugar epoxides that have been opened

by amines to form a new C–N-bond at the carbasugar C-1 may be

2. N-Linked

2.1. Epoxide and aziridine opening

divided into four types: saturated 1,2-epoxides with

a and b con-

ﬁgurations, saturated 1,5a-epoxides, and ﬁnally C-5@C-5a unsatu-

rated 1,2-epoxide derivatives (Fig. 3).

The coupling of amines and epoxides is a synthetic strategy that

has been used very widely by the Ogawa group and others for the

The coupling reactions of saturated epoxides take place under

rather forcing reaction conditions: elevated temperatures, typically

120 °C in alcoholic solvents, in a sealed tube for several days. b-

manno-Conﬁgured 5a-carba-1,2-epoxides (1 and 2) were opened

It is noteworthy that nitrogen atoms bridging two secondary carbons in N-linked

pseudodisaccharide structures tend to be resistant towards acylation.

2290

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

stereospeciﬁcally and with excellent regioselectivity by carbohy-

drate amines (3–9) to give good yields of -manno-conﬁgured imi-

sically electronically favoured attack at C-1 (both due to C-1 being

ﬂanked by the small and relatively electron-rich methylene group)

and the energetically demanding stereoelectronic requirement to

reach a far-from-ground-state conformation at the transition state

to enable trans-diaxial opening before reverting to a more favoured

diequatorial ground state again for the product.

a

no-linked pseudodisaccharides and pseudotrisaccharides (10–15;

Table 1).^10,11Tribenzylated (1) and unprotected (2) epoxides have

both been used. The reaction was efﬁcient enough for a branched

pseudotrisaccharide (10) to be formed (Table 1, entry 1).¹⁰Indeed,

a monomeric substrate with two amino groups (6) reacted with

two equivalents of epoxide (1) to give the same pseudotrisaccha-

ride in good yield (Table 1, entry 4).¹⁰

The formation of regioisomeric products from b-manno 1,2-car-

basugar epoxides has not been reported. This may be due to the

following factors: Attack at C-1 (to give a-manno-conﬁgured carba-

sugar pseudodisaccharides) results in trans-diaxial opening as fa-

voured by the Fürst–Plattner guidelines, so is stereoelectronically

favourable; C-1 is ﬂanked by an electron-rich methylene group

whereas C-2 is ﬂanked by a carbon bearing an electron-withdraw-

ing group, so S_N2 is electronically favoured at C-1 over C-2; the

groups ﬂanking C-1 are smaller than those ﬂanking C-2 so attack

at C-1 may be favoured also from a steric point of view.

A carbasugar 1,5a-epoxide (42) has been opened regioselec-

tively by various amines (21, 39, 40, 50–54) with attack at C-1 to

give the trans-diaxial opening products (44, 46, 55) as major prod-

ucts along with the C-5a-substituted trans-diequatorial opening

products (45, 47, 56) as minor products (Scheme 1).^17–22A slightly

lower regioselectivity was seen when the pseudodisaccharide

epoxide 43, which is not conformationally locked by the presence

of a 4,6-benzylidene acetal, was opened with amine 40 to give

pseudotrisaccharides 48 and 49.^18,19A racemic 1,5a-epoxide (rac-

57) was opened by enantiopure amines (21 and 50) to give cou-

pling products as regioisomeric mixtures with the diaxially opened

compounds (59) as major components and diequatorial com-

pounds (60) as minor products (Scheme 2).^23–27When racemic

mixtures of both epoxide and amine components (rac-39 and

rac-58) were used, all eight possible products were formed (two

diastereomeric pairs of enantiomeric products arising from each

of diaxial and diequatorial opening), the difﬁcult separation of

which may account for the lower product yields.

The lower regioselectivity for diaxial opening for these 1,5a-

epoxides than was seen for the 1,2-carbasugar epoxides may pos-

sibly be explained as follows: In the 1,5a-epoxides, the ﬂanking

atom C-2 bears an electron-withdrawing oxygen atom, whereas

C-5 does not. This may favour S_N2 attack at C-5a over C-1 from

an electronic point of view. Attack at C-1 leading to the trans-diax-

ial product is favoured stereoelectronically, so these opposing

arguments may result in the observed mixtures of products. When

the C-5-oxygenated 1,5a-epoxide 62 was opened by amine 61, only

one product (63) was reported (Scheme 3).²²The increased steric

bulk at C-5 and the addition of an electron-withdrawing group at

C-5 would both tend to favour attack at C-1 over C-5a.

The reactions of a-gluco or a-galacto carbasugar epoxides (16–

20) with amines (21–27) to give epoxide-opened coupling prod-

ucts (28–38; Table 2)^12–16are reported to be generally slower than

those for the carbamannosides, and the yields tend to be lower.

Regioselectivity is an issue in such reactions and regioisomeric

mixtures of products were formed in some cases (Table 2, entries

2 and 6).^12,14A benzylated

a-galacto epoxide (16) ring-opened with

the attack only at C-2 (trans-diaxial product (28); Table 2, entry 1),

a benzylidene-protected gluco 1,2-epoxide (17) gave a mixture of

the products of C-2 attack (29) and C-1 attack (30) (Table 2, entry

2), while 3,4-isopropylidene-protected galacto derivatives (18 and

19) gave mostly or exclusively C-1 attack (Table 2, entries 3–7).

A 3,4-cyclohexylidene 6-deoxygalacto epoxide (20) also gave

exclusive C-1 attack (Table 2, entry 8), leading to the formation

of a 1,1-linked di-

The -gluco/galacto type epoxides shown in Table 2 must open

L

-fucoside (38).¹⁶

a

to give trans-diequatorial products if a carbadisaccharide is to re-

sult from attack at C-1. Fürst–Plattner favoured trans-diaxial open-

ing would arise from a C-2 attack, which would not lead to a

carbadisaccharide product. The generally slower reactions and

The axial OH-5a has been removed to give a methylene group

by two routes as shown in Scheme 4 (forming 65–67). First, the

C-5a hydroxyl group in 46 was converted into a good leaving

group, which then suffered intramolecular attack by the N-1 to

give the aziridine. This aziridine was ring-opened by a sulfur

lower yields in this series than in the

a-manno-forming reactions

can be put down to the conﬂict between the sterically and intrin-

OH

Ph

O

AcO

Ph

AcO

O

NHR

+ diastereomers / 59

O

propan-2-ol, sealed tube,

RNH₂

+

AcO

120 °C, 5–9 days

OAc

Ph

O

rac-57

NHR

OH

O

(1–1.2 equiv.)

AcO

+ diastereomers / 60

RNH₂:

OH

O

H₃C

O

H₂N

HO

H₂N

HO

O

HO

O

OMe

NH₂

OMe

24,26

23,26

25,27

rac-39

rac-58

21

50

diaxial, 44% (4 diastereomers)

diequatorial, 16% (4 diastereomers)

diaxial, 16% + 18%

diaxial, 15% + 16%

diaxial, 18% + 11% (both rac)

diequatorial, 2% + 9% (both rac)

diequatorial, 7% + 7%

diequatorial, 9% + 9%

(products were acetylated) (products were acetylated)

Using enantiopure starting materials:²²

diaxial, 32%

diequatorial, 12%

Scheme 2. Synthesis of pseudodisaccharides from racemic 1,5a-epoxides. (Mixtures of stereoisomers were formed. Only one diastereomer of each regioisomer is shown).

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

2291

O

4-amino-4,6-dideoxyglucose 50 (Table 3, entry 1),²⁸while the

enantiomerically pure carbasugar (69) coupled with various carbo-

hydrate amines to give the pseudodisaccharides in good yield (Ta-

ble 3, entries 2–5).^11,29A branched pseudotrisaccharide (78) was

formed in a single step from monomeric components, a diamine

(6) and two equivalents of epoxide (69), although here a mixture

of products was seen, possibly due to acyl migration (Table 3, entry

6).¹⁰Much lower reaction temperatures (50–70 °C) were used for

opening C-5@C-5a-unsaturated epoxides than for the correspond-

ing saturated derivatives (where 120 °C was typically used), pre-

sumably due the intrinsically higher reactivity of an allylic epoxide.

O

propan-2-ol,

sealed tube,

120 ºC, 25 h

BnO

O

NH₂

OBn

OH

OBn

61 (1.1 equiv.)

N

H

O

Bn

O

+

O

BnO

63 / 78%

OBn

O

62

The racemic

a-xylo-conﬁgured unsaturated carbasugar donor

(79) reacted with amines with the attack only at the allylic centre

(i.e., C-1) to give, after acetylation, the 1,2-trans iminocarbadisac-

charide coupling products (82–85) as diastereomeric mixtures

(Scheme 5).^28,30Again, lower reaction temperatures (45–55 °C)

were used for the coupling reactions of this allylic epoxide.

Ogawa and Tsunoda approached carbasugar analogues of

aminosugar disaccharides by aziridine opening.³¹They used 2,4-

Scheme 3. Opening a 1,5a-epoxide with oxygen functionality at C-5.²²

nucleophile to give the trans-diaxial-opened thioether. Reductive

cleavage of the thio group with Raney nickel gave the C-5a meth-

ylene group (65).^20,21Alternatively, conversion of OH-5a (in 44 and

63) into dithiocarbonates, followed by radical reduction with tribu-

tyltin hydride, gave the C-5a-methylene carbasugars (66 and 67).²²

The axial OH-5a group has been used as a handle for the introduc-

tion of C-5@C-5a unsaturation (Scheme 4). Chlorination of the

alcohol (64) with sulfuryl chloride or thionyl chloride gave the ax-

ial chloride, along with aziridine by-products. On treatment with

DBU, HCl was eliminated across C-5@C-5a to give the valienamine

derivative (68), along with aziridine by-products.^22,24,26

dinitrophenyl carbasugar aziridines with

a-gluco (86) or a-xylo

(87) conﬁguration. A fully protected aziridine derivative (the

3,4,6-triacetate of 86) failed to react to any great extent, while

the reaction between the unprotected carbasugar aziridine (86)

and amine 21 showed no regioselectivity, and the products of at-

tack at both C-1 (89) and C-2 (88) were seen, along with the prod-

uct of intramolecular attack by OH-6 (90) (Scheme 6).

A (racemic) 5,6-unsaturated derivative (87) reacted withamine 21

with better regioselectivity for C-1 attack. The resulting pseudodisac-

charide 91a was refunctionalised at C-5a–C-6 to give either the vali-

olamine-like 93 or valienamine-like 94 pseudodisaccharides (Fig. 4).

The complementary strategy for the synthesis of N-linked

An unsaturated epoxide with the b-lyxo conﬁguration (69) has

been opened regioselectively at the allylic position (C-1) to give

the 1,2-trans products with

a-lyxo conﬁguration (72–78; Table

3).^10,11,28,29Using a racemic carbasugar (rac-69) as starting mate-

rial resulted in diastereomeric mixtures of products, i.e., 72a,b

from the 4-amino-4-deoxyglucose 21 and 73a,b from the

pseudodisaccharides by epoxide opening is to condense

a

i) NaH, di-

imidazolylsulfone,

DMF, 50 ºC

OH

O

Ph

O

20,21

Ph

O

BnO

HN

BnO

HN

OBn

ii) MeC₆H₄SH,

propan-2-ol,

80 ºC

65 / 65%

46

BnO

iii) Raney-Ni

O

i) methylthio-

thiocarbonylation

OBn

OH

OBn

N

H

O

22

H

O

Bn

O

ii) reduction

Bu₃SnH

OBn

O

63

66 / 74%

O

OBn

OH

OBn

O

N

H

N

H

O

i) methylthio-

thiocarbonylation

O

22

OBn

ii) reduction

Bu₃SnH

67 / 50%

44

O

Ph

OAc

OH

OAc

O

N

O

H

O

H

i) SO₂Cl₂, 33%

O

22

OAc

ii) DBU, toluene,

100 ºC, 36%

64

68

O

Ph

Scheme 4. Modiﬁcation of pseudodisaccharides: deoxygenation of C-5a to achieve valienamines and saturated carbasugars.

2292

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

Table 3

Formation of pseudodisaccharide and pseudotrisaccharide from ‘up’ type unsaturated donors

Entry

Epoxide

Amine

Conditions

Product

Yield

Ref.

28

BzO

AcO

OAc

HN

72a, 46%

73a, 10%

CH₂R

O

CH₂R

O

AcO

72a, R = H

73a

H₂N

HO

BzO

AcO

OMe

, R = OAc

O

AcO

HO

1

i) 2-Propanol, 50 °C, 5–7 days

ii) Acetylate

OMe

OAc

AcO

(1–1.3 equiv.)

21, R = OH

50, R = H

AcO

rac-69

CH₂R

72b, 31%

73b, 7%

O

BzO

72b

HN

AcO

, R = H

OMe

73b, R = OAc

HO

OH

NH₂

O

HO

BzO

AcO

HO

O

2

i) 2-Propanol, 60 °C, 48 h

ii) Deprotect

78%

29

NH

HO

OMe

O

HO

69

70 (2.8 equiv.)

74

HO

OMe

BzO

AcO

OAc

HO

H₂N

HO

O

OAc

AcO

3

4

69

i) 2-Propanol, 60 °C, 48 h

ii) Acetylate

74%

68%

11

O

HN

OMe

AcO

75

7 (1.3 equiv.)

OMe

BzO

OAc

HO

H₃C

O

AcO

H₂N

HO

OAc

O

H₃C

69

i) 2-Propanol, 50 °C, 3 days

ii) Acetylate

HN

OMe

AcO

76

8 (1.5 equiv.)

OMe

O

OH

O

OAc

O

AcO

O

BzO

AcO

OH

5

6

i) 2-Propanol, 70 °C, 3 days

ii) Acetylate

64%

34%

11

10

OAc

N

H

H₂N

AcO

77

9 (1.5 equiv.)

BzO

AcO

OAc

H₂N

HO

H₂N

OH

O

AcO

69 (2.3 equiv)

i) 2-Propanol, 50 °C, 1 week

ii) Acetylate

NAc

AcO

N

H

OMe

6

OAc

O

BzO

AcO

OAc

78

MeO

carbasugar C-1 amine with an appropriate epoxide (disconnection

2 in Fig. 2). The majority of reactions published in this area have

been aiming for a (1?4)-linkage to a carbohydrate with gluco ste-

reochemistry, no doubt because this is the conﬁguration in many

naturally occurring pseudooligosaccharides (cf. Fig. 1). The epox-

ides that have been studied that lead to such an outcome may be

split into two types; both have galacto stereochemistry and both

rely on regioselective attack at C-4 to give the 4-amino-4-deoxy-

glucose derivatives. In the ﬁrst type, the carbohydrate ring-confor-

mation is not locked by a 1,6-anhydro bridge; the epoxide-opened

product will adopt a ⁴C₁conformation and hence trans-diequatorial

ring-opening with attack at C-4 will give the gluco conﬁguration. In

the second type, 1,6-anhydrosugars are locked in a ‘ring-ﬂipped’

conformation and attack at C-4 gives the gluco epoxide-opened

product with a trans-diaxial conﬁguration and ¹C₄conformation.

Conformationally unlocked galacto 3,4-epoxides: Opening the 3,4-

epoxides (97–99) with enantiopure carbasugar C-1 amines (58, 95

96) gave the 4-substituted glucose (104–107) and 3-substituted

gulose (112–115) derivatives with varying regioselectivity (Table

4, entries 1–4).^32–35Ogawa et al. noted that a slightly better regi-

oselectivity for the desired diequatorial (i.e., gluco) product was

seen when a C-6 hydroxyl group was present in the epoxide than

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

2293

BzO

CH₂X

H

N

AcO

O

OAc

OR

OAc

BzO

82a X = OAc, R = βR¹, 7% ³⁰

83a X = H, R = βR¹, 7% ³⁰

84a X = H, R = αMe, 26% ²⁸

CH₂X

O

H₂N

HO

AcO

+

HO

OR

i) 2-propanol, DMF,

45–55 ºC, 3–6 days

O

85a X = OAc, R = αMe, 25% ²⁸

(1–1.3 equiv.)

80, X = OH, R = βR¹

81, X = H, R = βR¹

50, X = H, R = αMe

21, X = OH, R = αMe

rac-79

OAc

AcO

BzO

ii) acetylate

CH₂X

OAcH

N

O

AcO

OY

O

OAc

OR

R¹

=

82b X = OAc, R = βR¹, 7% ³⁰

83b X = H, R = βR¹, 7% ³⁰

84b X = H, R = αMe, 31% ²⁸

85b X = OAc, R = αMe, 15% ²⁸

OY

Y = H (80,81)

Y = Ac (82,83)

Scheme 5. Formation of pseudodisaccharides and pseudotrisaccharides from ‘down’ type unsaturated donors. The products may be C-6 acetates or C-6 benzoates depending

on the work-up before separation of diastereomers.

OAc

O

OAc

AcO

OMe

AcO

HN

AcO

88 / 19%

RHN

i) DMF/2-propanol (1:1)

sealed tube, 120 ºC, 3 days

OAc

AcO

HO

OH

H

N

O

H₂N

HO

AcO

NHR

ii) acetylate

+

AcO

OMe

N

R

86

89 / 20%

HO

OMe

21 (1.5 equiv.)

R = 2,4-dinitrophenyl

O

OAc

The reaction was also

reported with rac-86

O

AcO

NHR

90 / 24%

OAc

H

N

AcO

O

NHR

AcO

OMe

i) DMF/2-propanol (1:1)

sealed tube, 80 ºC, 20 h

91a / 22%

OH

O

HO

+ diequatorial diastereomer, 21%

HO

H₂N

HO

AcO

OAc

+

ii) acetylate

N

rac-87

R

AcO

HO

OMe

21 (1.1 equiv.)

O

H

N

R = 2,4-dinitrophenyl

AcO

OMe

AcO

RHN

92a

+ diaxial diastereomer,

combined yield, 8%

Scheme 6. Aziridine opening approach to carbasugar pseudodisaccharides.³¹

when such a group was absent.^33,34,38However, the regioselectivity

of these reactions was usually quite poor and signiﬁcant amounts

of both regioisomers were formed. The total yield of epoxide-

opened products can be high, however.

OH

H

N

HO

O

HO

NHAc

Similar coupling reactions between racemic carbasugar amine

nucleophiles (rac-39 and rac-58) and galacto 3,4-epoxides (100–

103) gave the coupling products (108–111, 116–119) (Table 4,

entries 5 and 6).^36–38Each reaction gave a mixture of four prod-

ucts; the two regioisomers were both formed as diastereomeric

mixtures. Again the regioselectivity was poor, but better regiose-

lectivity was seen with a free OH-6 in the epoxide than when this

was absent. A number of related deoxygenated derivatives were

investigated (not shown).³⁹

HO

OMe

93

OH

H

N

HO

O

NHAc

HO

OMe

94

Figure 4. Valiolamine 93 and valienamine 94 derivatives formed from 91a.³¹

2294

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

Table 4

Formation of N-linked pseudodisaccharides and pseudooligosaccharides by reaction of carbasugar amines with galacto 3,4-epoxides

Entry

Amine

Epoxide

Conditions

Product/yield

Ref.

AcO

CH

O

AcO

O

3

O

CH

³O

H₃C

AcO

O

N

H

1

i) Propan-2-ol,

DMF (1:1)

120 °C, 70 h

ii) H⁺

32,33

OAc

OR²

OR

2

AcO

O

NH

OH

97

2

OAc

AcO

OAc

NH₂

OR

AcO

104 / 19%

OAc

112 / 30%

58 (1.2 equiv.)

iii) Acetylate

AcO

OH

O

OAc

O

AcO

OAc

O

AcO

N

2

58 (1.2 equiv)

i) Propan-2-ol,

DMF (1:1)

120 °C, 67 h

ii) H⁺

32,33

OAc

OR²

AcO

H

AcO

NH

OAc

OH

98

OAc

2

AcO

OR

105 / 33%

OAc

113 / 21%

iii) Acetylate

OH

O

AcO

OAc

O

OAc

O

H

N

O

3

4

i) Propan-2-ol,

DMF (1:1)

120 °C, 67 h

ii) Acetylate

34

35

O

NH₂

AcO

OMe

NH

OMe

AcO

O

OH

O

95

AcO

O

OMe

99 (1.2 equiv.)

114 / 39%

106 / 51%

AcO

OAc

O

AcO

HO

OAc

O

AcO

N

99 (1 equiv)

i) Propan-2-ol,

120 °C,

6 days,

sealed tube

ii) Acetylate

OAc

H

AcO

OH

NH₂

OMe

NH

OAc

115 / 36%

OAc

AcO

OMe

AcO

96

107 / 37%

OAc

AcO

OAc

AcO

OAc

O

OH

O

AcO

O

N

O

OAc

H

1

AcO

5

i) Propan-2-ol,

DMF 120 °C,

67 h

36

OR

NH

OAc

O

1

AcO

OR¹

OR

O

OH

NH₂

OAc

108a / 22%

+ diastereomer/25%

rac-39

100 (1 equiv.)

116a / 26%

+ diastereomer/26%

ii) H⁺

iii) Acetylate

AcO

OAc

AcO

OAc

O

AcO

CH X

O

N

H

2

OAc

AcO

OR

O

AcO

NH

OAc

AcO

OR

AcO

OAc

OR

OH

6

rac-58

i) Propan-2-ol,

DMF 120 °C,

40–60 h

36

109a / R = Me, X = OAc/17%

117a / R = Me, X = OAc/20%

99 R = Me, X = OH

102 R = R , X = OH

103 R = R , X = H

(1–1.2 equiv.)

1

+ diastereomer/20%

+ diastereomer/18%

118a / R = R , X = OAc/21%

36,37

1

ii) H⁺

36,37

38

1

110a / R = R , X = OAc/13%

iii) Acetylate

+ diastereomer/14%

111a / R = R , X = H/6%

+ diastereomer/16%

38

1

119a / R = R , X = H/24%

+ diastereomer/6%

+ diastereomer/32%

Note: when a racemic starting material was used, each regioisomer was formed as a mixture of diastereomers; only one diastereomer is shown.

O

OY

O

OY

O

YO

R¹

YO

R²

Y = H (100, 102, 103)

Y = Ac (108, 110, 111, 116, 118, 119)

O

Y = H (97, 98)

Y = Ac (104, 105, 112, 113)

OY

YO

Conformationally locked 1,6-anhydro galacto 3,4-epoxides: 1,6-

Anhydrosugar epoxides (126–133) were opened by carbasugar

C-1 amines (27, 28, 95, 120–123) with excellent regioselectivity

to give the trans-diaxial products, i.e., the 4-substituted gluco-

conﬁgured compounds (135–146), in good yields (55–93%; Table

5, entries 1–11).^16,40–46The reactions seem to be sensitive to ste-

ric considerations: A more hindered amine (121) reacted more

slowly than a less hindered amine (120) (cf. Table 5, entries 1

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

2295

Table 5

Formation of N-linked pseudodisaccharides by reaction of carbasugar amines with 1,6-anhydrosugar galacto 3,4-epoxides

Entry

Amine

Epoxide

Conditions

Product

Yield

76%

Ref.

O

OH

O

HO

O

HO

1

n-BuOH, 110 °C, 20 h

40

41

42

HO

HO N

H

NH₂

OH

OBn

120

126 (2.5 equiv.)

126 (3.1 equiv)

135

HO

*BnO

MeO

*BnO

O

OH

O

*BnO

2

3

4

n-BuOH, 110 °C, 30 h

72%

55%

70%

MeO

N

H

Bn*O

*BnO

NH₂

OBn

121

136

O

HO

OH

O

BnO

126 (4.6 equiv)

1-Propanol, 90 °C, 30 h

BnO

N

BnO

BnOBnO

OBn

H

NH₂

BnO BnO

122

137

O

OAc

O

i) Coupling

ii) Acetylate

NH₂

NH

AcO

OAc

O

127

138

95

OBn

O

OBn

OH

O

BnO

5

6

126 (2.3 equiv)

i) BuOH, 110 °C, 72 h

67%

43,44

NH₂

NH

BnO

123

OBn

BnO

OBn

139

O

X

OH

O

BnO

OH

NH

123

i) BuOH, 100 °C, 2.5 days

ii) BzCl, py

140, 30%

141, 41%

45

OBz

BnO

128, X = S

129, X = Se

(2 equiv.)

140, X = S

141, X = Se

O

OH

O

NH₂

OH

O

H₃C

HO

7

8

i) Propan-2-ol, 120 °C, 2 weeks

93%

51%

16

46

NH

OH

H₃C

OH

N₃

27

OH

HO

130 (1.6 equiv.)

142

O

OAc

O

AcO

O

i) Propan-2-ol, 120 °C, 10 days

OH

OAc

N

H

AcO

ii) H⁺

O

AcO

NH₂

131 (1.2 equiv.)

iii) Acetylate

143

58

O

OAc

O

AcO

9

58

130 (1.2 equiv)

i) Propan-2-ol, 120 °C, 4 days

ii) H⁺

61%

46

N

H

AcO

N₃

AcO

iii) Acetylate

144

O

OAc

O

AcO

10

58

i) Propan-2-ol, 120 °C, 7 days

ii) H⁺

iii) Acetylate

14%

46

N

H

AcO

F

AcO

132 (3 equiv.)

145

(continued on next page)

2296

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

Table 5 (continued)

Entry

Amine

Epoxide

Conditions

Product

Yield

71%

Ref.

46

O

OAc

O

AcO

11

12

58

i) Propan-2-ol, 120 °C, 3 days

N

H

AcO

ii) H⁺

STol

TolS

iii) Acetylate

133 (1.6 equiv.)

146

AcO

OAc

O

AcO

rac-58

127

i) Propan-2-ol, 120 °C

ii) H⁺

iii) Acetylate

48%

42^,a

47^,a

48^,a

N

H

+diastereomer, 48%

AcO

147a

O

OAc

O

NH₂

OH

H₃C

HO

HN

OAc

13

14

130 (1.6 equiv)

i) Propan-2-ol, 120 °C, 1 week

44%

N₃

ii) H⁺

+diastereomer, 48%

OH

H₃C

iii) Acetylate

rac-124

148a

OAc

AcO

O

OAc

O

i) Propan-2-ol, DMF (1:1) 70 °C, 7 days

ii) Acetylate

20%

N

O

H

OTs

134

+diastereomer, 19%

TsO

NH₂

149a

rac-125

Bn* = CD₂Ph.

a

Racemic starting material was used, so two diastereomeric products were formed from trans-diaxial opening after attack at the epoxide C-4 (only one is shown).

Table 6

Formation of N-linked pseudodisaccharides by reaction of carbasugar amines with manno 2,3-epoxides

Entry

Amine

Epoxide

Conditions

Product

Yield

89%

Ref.

50

AcO

O

OAc

Ph

O

AcO

O

1

i) Propan-2-ol, 120 °C, 8 days

ii) H⁺

iii) Acetylate

N

AcO

H

O

NH₂

150 (1.3 equiv.)

153

58

O

HO

O

OH

49,50

58%

25%

OH

HN

O

HO

O

154

2

58

Propan-2-ol, 120 °C, 4 days

HO

OH

151 (1.3 equiv.)

O

N

H

O

OH

O

155

O

HO

O

OBn

OH

HN

OBn

6%

51

O

156

BnO

3

58 (1.2 equiv)

Propan-2-ol, 120 °C, 5 days

OMe

BnO

O

OMe

152

O

N

O

H

O

66%

MeO

157

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

2297

OAc

O

BnO

OMe

157a

HN

AcO

BnO

OH

O

AcO

O

BnO

i) Mesylation

ii) DBU

74% /

i) NaOAc,

AcOH, H₂O,

90 ºC, 3 days

AcO

OMe

N

156a:157a,

HN

O

5:4.

O

BnO

AcO

95%

O

157

O

ii) acetylate

O

158

O

HN

OMe

AcO

156a

AcO

O

OH^–resin,

MeOH

i) KOH,

EtOH

NaOAc,

AcOH

OAc

O

OAc

O

OAc

159 / 83%

OAc

N

ii) acetylate

OTs

RHN

R

160 / 87%

149a

O

= R

Scheme 7. Modiﬁcations after pseudodisaccharide formation.^48,51

O

and 2),⁴⁰and OH-5a-protected analogues of the nucleophile 122

failed to give any reaction with epoxide 126 (cf. Table 5, entry

3).⁴¹

Some similar coupling reactions between racemic starting

materials (rac-58, rac-124 and rac-125) and epoxides (127, 130,

134) have also been reported. The reactions proceeded with

excellent diastereoselectivity, but pairs of diastereomeric products

were formed (Table 5, entries 12–14).^42,47,48

manno 2,3-Epoxides: The discovery of the antibiotic Salbostatin,

consisting of valienamine (1?2)-linked to glucose, has led to the

synthesis of this compound and some analogues.^49,50A 4,6-benzyl-

idene-protected manno-conﬁgured 2,3-anhydrosugar (150) was

ring-opened with a protected valienamine (58), resulting in exclu-

sive formation of the undesired diaxial 3-amino-3-deoxyaltrose

(153) (Table 6, entry 1). The regioselectivity could be switched to

some extent by removing the benzylidene acetal. The unprotected

epoxide (151) reacted with the amine nucleophile (58) to give a

regioisomeric mixture of pseudodisaccharides with selectivity for

the diequatorial 2-amino-2-deoxyglucose compound (154) over

the diaxial 3-amino-3-deoxyaltrose (155). The unprotected epox-

ide was also found to be more reactive, so reducing the reaction

time (Table 6, entry 2). The major product (154) of this reaction

was converted into salbostatin.

the aziridine (Scheme 7).⁵¹Heating this aziridine (158) with so-

dium acetate in acetic acid gave a mixture of the two regioisomeric

products (156a and 157a). The regioselectivity was very low, but

the desired 2-amino-2-deoxyglucose (156a) was formed in slight

excess over the undesired 3-amino-3-deoxyaltrose. A related

modiﬁcation transformed a conformationally locked 3-amino-3-

deoxyglucose pseudodisaccharide (149a) into a 3-amino-3-deoxy-

mannose (160) (Scheme 7).⁴⁸Epoxide formation, rearrangement to

the aziridine and opening with sodium acetate in acetic acid, which

in this case proceeded with excellent regioselectivity for the diaxial

product, resulted in overall inversion at C-2.

2.2. Reductive amination

Reductive amination is the textbook method for secondary

amine formation, but it has not been widely used for the synthesis

of imino-linked pseudodisaccharides. An obvious conceptual dis-

advantage of this method over S_N2-type reactions such as epoxide

opening is that the reaction will be stereoselective to a greater or

lesser degree: two diastereomeric products can be formed, which

is not the case with stereospeciﬁc epoxide opening. On the other

hand, regioselectivity should not be a problem for reductive

amination.

A benzylated 2,3-epoxide (152) was ring-opened with the pro-

tected valienamine 58 with the aim of synthesising a pseudodisac-

charide consisting of valienamine (1?2)-linked to glucose as a

Kuzuhara and co-workers reported a reductive amination

between a carbasugar C-1 ketone (161) and a 4-amino-4,6-

dideoxyglucose derivative (164) (Table 7, entry 1).^52,53The stere-

potential inhibitor of

a

-glucosidase I,⁵¹but the required 2-substi-

oselectivity in the reduction step was in favour of the a product

tuted glucose derivative (156) was formed in very poor yield, the

major product being the 3-amino-3-deoxyaltrose (157) (Table 6,

entry 3). Analogues of the epoxide (152) with different protect-

ing-group patterns, with 4,6-benzylidene protection or with OH-

4 and OH-6 free, gave exclusive formation of the undesired 3-ami-

no-3-deoxyaltrose by attack at C-3. Hence, a strategy to isomerise

the 3-amino-3-deoxyaltrose product (157) into a 2-amino-2-

deoxyglucose derivative was developed. The free OH-2 in 157

was converted into a sulfonate, and treatment with a base gave

(166), which was formed in 30% yield, along with the b product

(167, 4%). Direct reduction products (carbasugar C-1 alcohols)

were obtained in 50% combined yield. The attempted reverse cou-

pling between a carbasugar C-1 amine and a trisaccharide C-4 ke-

tone failed, giving the galacto-conﬁgured product in 1% yield as

sole pseudotetrasaccharide, along with unidentiﬁed by-products

(not shown). Haines and Carvalho used reductive amination to link

a 3-amino-3-deoxyglucose (165) to an inositol-derived ketone

(162) in an attempt to synthesise an inhibitor of

a-glucosidase II

2298

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

Table 7

Reductive amination with carbasugar C-1 ketones

Entry

Ketone

Amine

Conditions

Product

Yield

30%

Ref.

OTr

H₃C

O

52,53

BnO

OR

N

H

BnO

O

Bn

TrO

BnO

H₃C

166

O

H₂N

BnO

1

NaBH₃CN, pH 6.2, 4 Å

O

BnO

164

BnO

OTr

OR

161 (2 equiv.)

H₃C

BnO

H

BnO

O

N

4%

OBn

BnO

OR

167

BnO

H₂N

OBn

BnO

O

BnO

2

i) Benzene, 4 Å, rt, 70 h

ii) NaBH₃CN, DMF, rt, 70 h, 100 °C, 1 h

61%

54,55

N

BnO

H

O

BnO

OMe

BnO

OMe

O

OBn

162

165

168

OBn

BzO

BnO

HO

H

N

BnO

OBn

BzO

23%

25%

BnO

HO

O

HO

OMe

H₂N

HO

12

169

3

NaBH₃CN, MeOH, reﬂux, 12 h

HO

O

OBn

OH

OMe

O

21

BnO

163

HOOMe

N

H

BzO

HO

170

O

OBn

O

R =

O

BnO

O

OBn

(Table 7, entry 2).^54,55The optimised conditions required pre-for-

mation of the imine and cyanoborohydride reduction in a second

step and gave the desired pseudodisaccharide (168) in 61% yield.

Attempted reductive amination between the same inositol ketone

(162) and a 2-amino-2-deoxyglucose failed, however. Imine for-

mation was very sluggish and on heating, an aromatic elimination

product was obtained. Ogawa et al. described a reductive amina-

tion for the formation of an N-linked lactose analogue (169) by

the reaction of a 4-amino-4-deoxyglucose (21) with a carbasugar

C-1 ketone (163) (Table 7, entry 3), but the diastereoselectivity of

try 1). The yield of the coupled products (177b and 178b) was

much lower when the 6-hydroxylated analogue (173b) was used.

Using a carbasugar C-1 amine (171) and carbasugar C-4 ketone

(174), the yield of pseudodisaccharides became very low and no

selectivity between the gluco (179) and galacto (180) products

was seen (Table 8, entry 2). Ogawa et al. used an unprotected b-ga-

lacto C-1 amine (172) as coupling partner with conformationally

locked carbasugar C-4 ketones (175 and 176a–c).⁵⁸These reactions

gave the pseudodisaccharides (181 and 182a–c) in respectable

yields and with good diastereoselectivities (Table 8, entries 3 and

4).

the reduction was low, with the a-linked compound (170) formed

in approximately the same amount as the desired b product.¹²

Stick and co-workers attempted a reductive amination between

2.3. Alkylation: halide or sulfonate displacement

an

a,b-unsaturated carbasugar C-1 ketone and a carbohydrate C-

4 amine, but no coupling product was obtained.⁵⁶

Direct S_N2 alkylation is not usually a recommended method

for the synthesis of secondary amines. However, in the systems

described below, overalkylation has not been described and it is

unlikely to be a problem, due to the low nucleophilicity of the

secondary amine nitrogen for both steric and electronic reasons.

This approach has been used in a few cases to access imino-

linked pseudodisaccharides. In fact, some of the ﬁrst syntheses

of N-linked carbasugar pseudodisaccharides were achieved by

alkylation of amines with carbasugar allyl bromides (Scheme 8).

Examples of successful couplings between carbasugar C-1

amines and carbohydrate ketones are known. Horii et al. reported

the synthesis of some carbasugar (1?4)-linked pseudodisaccha-

rides by this strategy with unprotected carbasugar C-1 amines.⁵⁷

A xylo-hexoside-4-ulose derivative (173a) was reductively con-

densed with a carbasugar C-1 amine (171) to give pseudodisaccha-

ride products (177a and 178a) in a reasonable yield and with good

diastereoselectivity for the gluco conﬁguration (177a) (Table 8, en-

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

2299

Table 8

Reductive amination with carbasugar C-1 amines

Entry Amine

Ketone

Conditions

Product

Yield

Ref.

CH₂R

O

HO

N

H

HO

CH₂R

177a, 43% 57

177b, 12%

HO

O

HO

OMe

HO

177a, R = H

O

177b, R = OH

O

OMe

1

i) NaBH₃CN, HCl, DMF, 65 °C, 18 h

HO

ii) Deprotect

NH₂

HO

173a, R = H

NH

HO

171

CH₂R

O

173b, R = OAc

HO

(1.7–2 equiv.)

HO

178a, 7%

178b, 4%

HO

OMe

178a, R =H

178b, R = OH

H₃C

HO

N

H

HO

H₃C

O

HO

8%

57

NH₂

179

O

2

3

4

171

i) NaBH₃CN, HCl, DMF, 65 °C, 18 h

ii) Deprotect

NHCbz

HO

NH

HO

CH₃

HO

174 (1.3 equiv.)

HO

7%

HO

180

HO

NH₂

HO

O

OH

O

AcO

OAc

i) HCl, MeOH, H₂O, NaBH₃CN, MgSO₄, reﬂux,

19 h

ii) Acetylate

47%

58

NH₂

O

NH

OH

172

OAc

175 (2 equiv.)

181

X

O

AcO

OAc

O

OBn

NH

OBn

172

i) HCl, MeOH, H₂O, NaBH₃CN, MgSO₄, reﬂux,

19 h

ii) Acetylate

182a, 60% 58

182b, 41%

182c, 50%

OAc

182a, X = OMe

182b, X = O(CH ) CH

176a, X = OMe

176b, X = O(CH ) CH

2 7

3

2 7

3

176c, X = N

3

182c, X = N

3

The allylic nature of the bromides presumably enhances their

reactivity to the extent that an alkylation reaction is possible;

S_N2 displacement of carbohydrate halides is not normally

straightforward.

Ogawa et al. coupled together two racemic carbasugar deriva-

tives, giving four stereoisomeric products (two diastereomeric

pairs of enantiomers). A primary allylic bromide (rac-183) reacted

with a carbasugar C-1 amine (rac-39) to give the coupled products

(rac-187) in good combined yield, but separation resulted in some

losses (the isolated yields of the separated diastereomers were 19%

and 27%).^59,60Note that the nitrogen in the products was not

fonamide or triﬂuoroacetamide nucleophiles failed to give

coupling products with these allylic bromides.

Kuzuhara and co-workers coupled an enantiopure carbasugar

C-1 bromide (185) with a 4-amino-4-deoxyglucose disaccharide

(186).⁶¹They found that the C-1 conformation of the bromide

was not stable under the reaction conditions, and therefore used

an

a,b-mixture of bromides, which gave an a,b-mixture of prod-

ucts (189). The reaction conditions used here involved the addition

of sodium iodide; the conditions used by Ogawa et al. (ⁱPr₂NH (sic),

DMF) reportedly failed to give the coupling product.

Stick et al. formed carbasugar pseudodisaccharides by coupling

of unsaturated carbasugar C-1 amines (123, 190 and 191) with ax-

ial C-4 or C-3 carbohydrate triﬂates (192–197). Straightforward

displacement in N,N⁰-dimethyl-2-imidazolidinone (DMI) as solvent

gave the xylo- or gluco-conﬁgured all-equatorial products (199–

204) stereospeciﬁcally with inversion of conﬁguration (Table

9).^56,62–65It was important to have a neighbouring benzoate

protecting group and not a benzyl ether for good yields to be ob-

tained. The 2,3-dibenzylated triﬂate (analogous to 193) gave essen-

tially only the elimination products with amine 123.⁶⁶In most cases

acetylated. An

a C-1 allylic bromide (rac-184) coupled with the

carbasugar C-1 amine (rac-39) to give the b products (rac-188, as

a mixture of two diastereomers, both racemic, combined yield

40%), i.e., with inversion of conﬁguration at C-1. Heating the reac-

tion mixture resulted in degradation, so the reaction was con-

ducted at rt. But the b-bromide (C-1 epimer of rac-184) also gave

the same b-conﬁgured amine products (rac-188); neighbouring

group participation or epimerisation of the bromide to the (pre-

sumably) more reactive

a-isomer are possible explanations. Sul-

2300

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

ates for syntheses of ‘pseudoacarviosin’⁶⁸(a carbasugar analogue

O

of acarviosin) and validoxylamine G⁶⁹were obtained by this meth-

od (Table 10, entries 4 and 5).

O

59,60

NH₂

i) DIPEA, CHCl₃,

reflux, 3 days

O

rac-39

HN

AcO

rac-187a

2.5. Intramolecular reaction

+

ii) acetylate

Br

AcO

Knapp et al. described intramolecular strategies for N-substi-

tuted C-5@C-5a unsaturated carbasugar synthesis.⁷⁰Only one

example of the formation of a pseudodisaccharide by such a

method has been described, however. A carbasugar precursor

allylic alcohol (221) was coupled to a carbohydrate isothiocyanate

(222) to give a carbonimidothioate (223) (Scheme 9). On heating,

this compound underwent a 3,3-sigmatropic shift to form the N–

C-1 bond. The reaction product was a thiocarbamate, which was

deprotected under dissolving metal conditions to give the free

amine (224). While this strategy worked well to form a (1?6)-

linked derivative, an analogous reaction to form a (1?4)-linked

glucose pseudodisaccharide failed, possibly owing to the in-

creased steric hindrance around the secondary carbon of the

amine.

AcO

OAc

+ rac-diastereomer

combined yield, 80%

HO

OAc

rac-183

O

59,60

O

NH₂

DIPEA, DMF

RT, 20 days

O

rac-39

AcO

NH

+

AcO

rac-188a

+ rac-diastereomer

combined yield, 40%

OAc

AcO

OAc

AcO

Br

3. O-Linked

rac-184

3.1. Triﬂate displacement

BzO

Ether-linked pseudodisaccharides have been synthesised by S_N2

reaction between carbasugar C-1 alcohols and carbohydrate tri-

ﬂates. The alkoxide nucleophile is a strong base, so potential com-

peting elimination reactions must be minimised.

Paulsen et al. identiﬁed triﬂates that can be used to synthesise

6-substituted glucose (236–238) and 4-substituted glucose (239–

243) pseudodisaccharides.⁷¹The displacement of the primary gluco

triﬂate (233) by alkoxides derived from various carbasugar C-1

alcohols (225–227) took place in THF at rt, giving the ether-linked

(1?6)Glc pseudodisaccharides (236–238) in excellent yield (Table

11, entries 1–3).⁷¹The alkoxides were generated from the alcohols

and NaH, but KOtBu could also be used in combination with 18-

crown-6.

The synthesis of pseudodisaccharides containing 4-substituted

glucose was more difﬁcult; the reaction products are sec–sec

ethers, and S_N2 is expected to be more difﬁcult at a secondary than

a primary carbon, while competing elimination does not suffer the

same steric constraint. A galactose 4-triﬂate in the ⁴C₁conforma-

tion (245; Fig. 5) gave only the product of elimination. Therefore,

the conformationally locked (¹C₄) 1,6-anhydrogalactose derivative

(234) with an equatorial C-4 triﬂate was used instead; here the

antiperiplanar relationship of H and OTf is avoided, thus disfavour-

ing a competing elimination pathway. Yields of pseudodisaccha-

ride coupling products in THF were low, but adding DMF or

HMPA as co-solvent gave better yields, and some ether-linked

pseudodisaccharides (239–243) could be synthesised from the cor-

responding alcohols (225 and 228–231) by this method (Table 11,

entries 4–8).^71,41In my laboratory, we synthesised a pseudodisac-

charide based on 3-substituted glucose (244) by coupling of a car-

basugar C-1 alcohol (232) with an allo 3-triﬂate (235) (Table 11,

entry 9).⁷²

The yields for the synthesis of (1?6)- and (1?4)-linked com-

pounds, at least, are high, and apparently independent of the struc-

ture of the nucleophile, from which we may infer that this

approach may be used to synthesise other carbasugar pseudodisac-

charides linked (1?4) or (1?6) to glucose. This approach to

pseudodisaccharides requires triﬂates that will not easily undergo

1,2-elimination even under strongly basic conditions, so if further

such carbohydrate triﬂates are discovered, the method may be ex-

tended to other linkages.

BnO

BzO

Br

61

OBn

BnO

185

CH₃

NaI, DMSO

3 days

O

NH

+

OBn

BnO

CH₃

OR

O

189 / 13%,

α/β ca 1:1

H₂N

BnO

OR

O

186 (6 equiv.)

OBn

O

R =

OBn

Scheme 8. Alkylation of amines using carbasugar allylic bromides.^59–61(187 and

188 were formed as mixtures of two diastereomers, both of which were racemic.

Only one diastereomer is shown.)

where a neighbouring benzoate was present, the coupled products

(199–203) were obtained in reasonable yield (43–56%, Table 9, en-

tries 1–5),^56,62–64but a pseudotetrasaccharide (205) was only

formed in very low yield by attempted coupling of a carbasugar

amine (123) with a trisaccharide triﬂate (128) (Table 9, entry 6).⁶⁵

2.4. Palladium-catalysed allylic coupling

The allylic nature of the C-1 position in C-5@C-5a unsaturated

carbasugars opens the way for

allylic amination reaction for the synthesis of N-linked

pseudodisaccharides.

a transition-metal-catalysed

Shing et al. developed this idea, using C-5@C-5a unsaturated

carbasugars bearing chloride at C-1 (209–212) as the electrophilic

coupling partner and other C-1 or C-4 amino carbasugars (58, 125

and 206–208) as nucleophiles (Table 10).^67–69Pseudodisaccharides

(214–218) were formed with retention of conﬁguration at the

allylic carbasugar C-1 and with minimal elimination by-products

(219 and 220) under the optimised conditions (i.e., with a palla-

dium catalyst and a phosphonate ligand). The yield of elimination

by-products was minimised by using acetal protecting groups

rather than benzyl ethers on the electrophiles (cf. Table 10, entries

2 and 3).⁶⁷Outstanding yields of pseudodisaccharide intermedi-

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

2301

Table 9

Formation of N-linked pseudodisaccharides and pseudotetrasaccharides by reaction of carbasugar C-1 amines with carbohydrate triﬂates

Entry

Amine

Triﬂate

Conditions

Product

Yield

56%

Ref.

TfO

H

N

BnO

O

NH₂

BzO

1

BzO

DMI, 60 °C, 20 h

62

63

56

64

BnO

OMe

BnO

OBn

OBz

192

190 (2.5 equiv.)

199

OBn

TfO

CH₃

H

N

CH₃

O

BnO

HO

O

2

3

4

5

BzO

DMI, rt, 4 days

DMI, rt, 5 days

DMI, rt

36%

43%

41%

55%

NH₂

BzO

BnO

HO

OMe

OBn

OBz

193

BnO

OBz

200

191 (2.5 equiv.)

OBn

TfO

CH

³O

H

N

CH₃

BnO

O

NH₂

BzO

OMe

BzO

BnO

OBn

194

BnO

OBn

201

123 (2 equiv.)

TfO OBz

O

BzO

OBz

O

H

N

123 (2.1 equiv)

BzO

OMe

BnO

OBn

OBz

195

202

CH₃

OBn

BzO

O

BzO

H

N

OMe

123 (2.5 equiv)

DMI, rt, overnight

OBz

BnO

OTf

OBn

196

203

TfO

H

CH₃

CH

³O

N

O

BzO

X

BnO

X

OBn

6

DMI, 50–60 °C, 20 h

OBz

204, 20%

205, 4%

65

OBz

197, X = H

198, X = X

204, X = H

205, X = X

1

OBn

O

OBn

O

X¹

=

O

BnO

O

BnO

OBn

OMe

OBn

3.2. Epoxide opening

ride product (253) and none of the pseudotrisaccharide (256; Table

12, entry 2).¹⁰

Ogawa et al. have shown that epoxide opening can be used to

form ether-linked pseudodisaccharides and pseudotrisaccharides.

Lewis acid catalysis promoted a coupling reaction between a b-

manno-conﬁgured carbasugar 1,2-epoxide (1) and a primary carbo-

hydrate alcohol (246) to form an ether-linked carbasugar pseudo-

The products of these epoxide-opening reactions all have the

a

-manno conﬁguration. Attempts to obtain b-gluco- or b-galacto-

conﬁgured carbasugar pseudodisaccharides by analogous trans-

diequatorial opening of the corresponding -carbasugar epoxides

a

with alcohol nucleophiles failed, which contrasts with the results

with amines, where imino-linked pseudodisaccharides could be

formed from such epoxides (vide supra). Only complex mixtures

disaccharide (252) (Table 12, entry 1), but

a secondary

carbohydrate alcohol (249) failed to open the epoxide under these

conditions.⁷³Efﬁcient coupling could, though, be achieved under

basic conditions; heating the epoxide (1) with the alcohol (249)

in DMF with excess base gave the sec–sec ether (255) in 35% yield.

Running the reaction in the presence of a crown-ether gave a higher

yield of the pseudodisaccharide (Table 12, entry 4).^10,73Two b-man-

no-conﬁgured carbasugar 1,2-epoxides (1 and 258) were used, and

these coupled with primary or secondary carbohydrate alcohols,

(247–251 and 259–268) to give the corresponding pseudodisaccha-

rides (253–255, 257 and 269a–k) or pseudotrisaccharide (256) in

good yields and with excellent regioselectivity for the trans-diaxi-

al-opening products (Table 12, entries 12–17).^{10,12–14,35,73–79}

A branched pseudotrisaccharide (256) was formed by this

method using a pseudodisaccharide alcohol (250) as a nucleophile

(Table 12, entry 5),^10,73but when the diol 247 was used as a nucle-

ophile, only the primary alcohol reacted, giving a pseudodisaccha-

of elimination products were seen when

a-gluco and a-galacto

1,2-epoxides (16–18) were heated with alcohols and base in

DMF. Even a simple model alcohol, octanol, gave the coupling

product in only 5% yield.^12,13A procedure was developed to ac-

cess ether-linked carbasugar pseudodisaccharides with conﬁgu-

rations other than

a-manno, based on epimerisations of either

C-1 or C-2, or both (Scheme 10).^{12–14,75–79}Oxidation of the free

carbasugar OH-2 (269) gave the C-2 ketone (270), which could

be epimerised at C-1 under basic conditions to give a mixture

of

rated, and which usually contained the b isomer (271) as the

major component. Reduction of the -conﬁgured C-2 ketone

(270) gave then mixtures of gluco (272) and manno (269)-conﬁg-

ured -linked carbasugar pseudodisaccharides, while reduction

of the b-conﬁgured C-2 ketone (271) gave mixtures of the gluco

a- and b-carbahexuloses (270 and 271), which were sepa-

a

2302

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

Table 10

Formation of N-linked pseudodisaccharides by palladium catalysis

Entry Amine

Electrophile

Conditions

Product

Yield

Ref.

67

O

1

Pd(dba)₂, TMPP. MeCN, Et₃N,

rt

78% +219/15%

O

N

H

O

NH₂

Cl

214

209

58 (4.4 equiv.)

OR

O

BnO

HO

BnO

OR

OH

BnO

OBn

BnO

2

3

Pd(dba)₂, TMPP. MeCN, 50 °C

a, R = Ac; 42% +220/38%

b, R = CONMe₂; 50% +220/

30%

67

OBn

N

H

BnO

Cl

NH₂

206

210a−c

215a−c

c, R = MOM; 57% +220/24%

O

Pd(dba)₂, TMPP. MeCN, Et₃N,

rt

94%

84%

99%

O

N

H

O

NH₂

Cl

211

216

125 (2.4 equiv.)

O

OMe

O

MeO

O

MeO

O

NH₂

CH₃

O

H₃C

4

5

Pd(dba)₂, TMPP. MeCN, Et₃N,

rt

68

69

O

OMe

O

HN

MeO

TBDMSO

MeO

O

MeO

O

207 (3 equiv.)

OTBDMS

212

217 MeO

OMe

O

OTMS

MeO

212

Pd(dba)₂, TMPP. MeCN, Et₃N,

rt

O

OMe

O

MeO

N

H

NH₂

O

TMSO

208 (3.8 equiv.)

MeO

218

BnO

O

Et

BnO

O

TMPP (trimethylolpropane phosphite):

Elimination products:

O

OR

220a–c

O

219

P

BnO

(273) and manno (274) b-linked carbasugar pseudodisaccharides;

different reduction conditions often gave different diastereose-

lectivities, as shown in Scheme 10. Further protecting group

manipulation and inversion of conﬁguration at C-4 by S_N2 reac-

tion or oxidation–reduction gave galacto-conﬁgured carbasugars

(not shown).^{12–14,75,78}The C-2 epimerisation (manno 254?gluco

275) has also been demonstrated with a different protecting-

group pattern (Scheme 11). Carbaglucosamine derivatives were

accessible from sulfonation of OH-2 and S_N2 displacement with

nitrogen nucleophiles. This substitution worked better for b-

BnO

OBn

HO

221

OBn

OMe

O

BnS

i) NaH, THF

+

N

ii) BnBr

O

BnO

NCS

O

BnO

223

BnO

OMe

222

manno-conﬁgured starting material than

a-manno, with elimina-

i) Toluene, reflux, 48 h

ii) debenzylation

tion dominating in the latter case.⁷⁷Introduction of C-5@C-5a

unsaturation into an ether-linked carbasugar was also achieved

at the pseudodisaccharide level (not shown).³⁵

HO

HN

HO

O

Ogawa has suggested that the b-talo-conﬁgured 1,2-epoxide

276 (Fig. 6) could be used to avoid some of the extensive post-pro-

HO

224 / 52%

OMe

cessing necessary to achieve

a-galacto-conﬁgured ether-linked

carbasugar pseudodisaccharides, but no detailed results from cou-

Scheme 9. Intramolecular amino delivery reaction for the synthesis of an N-linked

pseudodisaccharide.⁷⁰

pling with this epoxide are given.⁷⁶

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

2303

Table 11

Ether-linked pseudodisaccharide synthesis by triﬂate displacement

Entry

Alcohol

Triﬂate

Conditions

Product

Yield

84%

Ref.

OTf

MeO

BnO

O

OMe

OBn

MeO

BnO

O

BnO

O

OH

1

NaH, THF, 0 °C to rt, 2 h

71

BnO

OMe

225

OBn

BnO

236

233 (1.2 equiv.)

OTBDPS

BnO

O

OMe

OBn

2

233 (1.3 equiv)

NaH, THF, 0 °C to rt, 12 h

88%

BnO

O

BnO

OH

226

237

OBn

BnO

OTBDPS

MeO

BnO

MeO

BnO

O

OMe

OBn

BnO

O

3

4

5

233 (1.5 equiv)

NaH, THF, 0 °C to rt, 12 h

86%

90%

88%

71

BnO

OH

227

238

OBn

BnO

O

OBn

O

OBn

O

MeO

BnO

TfO

225

NaH, HMPA, 0 °C to rt, 12 h

O

BnO

OBn

BnO

234 (1.3 equiv.)

239

OMEM

BnO

MEMO

O

OBn

O

MeO

BnO

234 (2 equiv)

NaH, THF, HMPA, 0 °C to rt, 24 h

MeO

BnO

O

BnO

OH

228

240

OMEM

OH

O

OBn

O

BnO

6

7

234 (2 equiv)

i) NaH, HMPA, 0 °C to rt, 12 h

ii) HCl, MeOH

88%

87%

71

OH

O

BnO

229

BnO

241

OBn

O

OBn

O

OBn

MeO

BnO

NaH, THF, HMPA, 0 °C to rt, 12 h

NaH, THF, HMPA, 0 °C to rt, 17 h

NaH, DMF, rt

O

BnO

MeO

BnO

230

BnO

OH

BnO

242

OBn

O

BnO

OBn

O

BnO

8

9

234 (2.7 equiv)

89%

34%

41

72

O

BnO

231

BnO

OBn

OH

BnO

243

O

OBn

O

OBn

O

BnO

O

OH

TfO

OBn

O

OBn

232

O

235 (2 equiv.)

244

3.3. Other methods for ether-linked carbadisaccharide

synthesis

Sinaÿ has developed modiﬁcations of the Ferrier II rearrange-

ment where the glycosidic substituent is retained. The reaction

has been applied to the synthesis of pseudodisaccharides, and

the carbocycles elaborated to give carbadisaccharides. Hence,

disaccharides were modiﬁed to give exo-glycals (277), which rear-

ranged upon treatment with TIBAL to give pseudodisaccharides

(278) (Scheme 12).⁸⁰

Two approaches to carbasugar pseudodisaccharides where the

carbasugar is assembled after attachment to a carbohydrate com-

ponent are described in this section. These approaches are quite

different to the other work described in this minireview in that

they completely remove the need to form an ether bond between

two sterically encumbered and electronically unreactive partners

(a carbasugar and a carbohydrate); the linkage is put in place at

an earlier stage, either as a glycoside or as an ester.

The reaction worked well for a range of (1?6)- or (1?4)-linked

compounds, but debenzylation was a competing reaction when the

anomeric substituent was an

a-methyl glucoside (278a,b). The

reaction was diastereoselective and the products with an axial

2304

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

OBn

to give an axial chloride with overall retention of conﬁguration)

followed by basic elimination.

TfO

BnO

O

BnO

OMe

245

4.2. Triﬂate displacement

Figure 5. An ineffective carbohydrate triﬂate.

Triﬂate displacement strategies have also been used to synthe-

sise thioether-linked carbasugar pseudodisaccharides.

C-1 orientation were formed. The reaction also worked in a bis

manner to give a pseudodisaccharide (not shown)⁸¹and even in

a tris manner to give a pseudotrisaccharide (279?280) in 33%

yield (Scheme 12).⁸²Modiﬁcation at C-5 of a pseudodisaccharide

(281) obtained by this method gave a carbaidopyranoside pseudo-

disaccharide (284) (Scheme 13).⁸³

C-6-Deprotected carbaidopyranosides have been converted to

carbaglucopyranosides by an oxidation–epimerisation–reduction

sequence, although this has only been shown on a glycosidically

linked pseudodisaccharide (285?286) as yet (Scheme 14).⁸⁴

Treatment of the exo-glycal rearrangement precursors with tri-

methylaluminium instead of TIBAL gave a rearrangement to cyclo-

hexane products with no debenzylation observed.⁸⁵Here though, a

methyl group was added from the reagent (rather than a hydride

from TIBAL), giving a carbahexopyranose (287), again with stere-

oselectivity for the C-1-axial product (Scheme 15).

Stick et al. coupled the carbasugar C-1 thiol of a C-5@C-5a

unsaturated carbapentose derivative (299) with a pentose C-4 tri-

ﬂate (192), which gave the thioether carbasugar pseudodisaccha-

ride (300) with inversion of conﬁguration at C-4 (Scheme 18).⁶²

My group used the opposite strategy, with displacement of a b-glu-

co-conﬁgured carbasugar C-1 triﬂate (301) by C-3 thiols with gluco

(302) or manno (303) conﬁguration giving, with inversion of con-

ﬁguration at C-1, the a-gluco thioether carbasugar pseudodisaccha-

rides (304 and 305) (Scheme 19).⁸⁷

5. Summary, conclusions and outlook

In this ﬁnal section, different useful routes to each of the syn-

thesised linkages are compared. It is assumed that the required

nucleophiles/electrophiles are accessible. This identiﬁes those

areas where methodology exists for the formation of a given ste-

reochemical linkage as well as those areas where more work is

required.

Secondly, Mootoo and co-workers have also developed

a

route based on post-conjugation carbasugar synthesis, and dem-

onstrated this principle for the synthesis of a carbagalactoside

(Scheme 16).⁸⁶The acyclic compound 288 is a carbasugar pre-

cursor. This was coupled to a carbohydrate alcohol by esteriﬁca-

tion. Tebbe methylenation gave the enol ether (289), which,

upon activation with methyl triﬂate, underwent a cyclisation

reaction to give the carbocyclic enol ether (290). This was then

hydrated to give the carbasugar pseudodisaccharide (291). The

method has only been demonstrated for a single example of

an ether-linked carbadisaccharide, but a glycosidically linked

compound was also made in the same way starting from a hemi-

acetal (not shown).

5.1. Carbasugar electrophiles

This section summarises those reactions in which the new C–X

bond is the carbasugar C-1–X bond, corresponding to disconnec-

tion 1 in Scheme 2.

5.1.1. 5a-Carba-a-mannopyranosides

N, O and S-linked pseudodisaccharides are all accessible by con-

densation of a carbasugar 1,2-epoxide with a nucleophile: amines

(yields are good, 56–95%; Table 1) for N-linked pseudodisaccha-

rides; alcohols (yields are reasonably high, 36–82%; Table 12) for

O-linked pseudodisaccharides or thiols (only one example known,

Scheme 17) for S-linked pseudodisaccharides. The synthesis is efﬁ-

cient in that the a-mannosides are obtained directly in the conden-

sation, and the trans-diaxial ring-opening attack at C-1 is

favourable, so good regioselectivity is always seen.

4. S-Linked

Reports of thioether-linked carbasugar pseudodisaccharides are

scarcer than those of their amine- or ether-linked counterparts.

The higher nucleophilicity and lower basicity of sulfur compared

to oxygen mean that thioetheriﬁcation reactions are less likely to

suffer from elimination than are the corresponding etheriﬁcations.

However, due to the very limited results in this area, it is difﬁcult

to draw extensive conclusions.

5.1.2. 5a-Carba-

Carbasugar 1,5a-epoxides are opened by amines to give 5a-hy-

droxy-5a-carba- -glucosides with good regioselectivity for diaxial

a-glucopyranosides

a

opening but rather low yields (34–57%) of coupling product. Com-

peting formation of diequatorial product accounts for the lower

yields in these reactions (Scheme 1). Such compounds have been

4.1. Epoxide opening

deoxygenated at C-5a to give N-linked carba-a-gluco pseudodisac-

Ogawa and co-workers synthesised thioether-linked carbasugar

pseudodisaccharides by epoxide opening.³⁵Opening a 1,2-epoxide

(292) with a thiol generated and activated in situ by treatment of

the thioacetate (293) with 2-aminoethanethiol and 1,4-dithio-

charides (Scheme 4). Analogous epoxide opening has also been

demonstrated for a thiol nucleophile (Scheme 17), but not for alco-

hol nucleophiles.

O-Linked carba-

mannosides by

achieve epimerisation at C-2 (Schemes 10 and 11).

S-Linked carba- -glucosides have been synthesised in two

a

-glucosides are available from the carba-

a-

erythritol gave the

opening product, in excellent yield and with excellent regioselec-

tivity (Scheme 17). The -manno-conﬁgured carbasugar was con-

verted to the required -gluco compound by oxidation of the free

a-manno derivative (294), i.e., the trans-diaxial

a

two-step oxidation–reduction sequence to

a

ways: either by coupling of a carba-b-glucose 1-triﬂate and carbo-

hydrate C-3 thiols (Scheme 19); or by C-2 epimerisation of the

OH-2, followed by reduction of the resulting ketone to give a thio-

ether-linked carbamaltose (295) (Scheme 17). Opening a carbasug-

ar 1,5a-epoxide (296) with a glucose C-4 thiol nucleophile (from

ﬁrst-formed carba-a-mannoside by an oxidation–reduction se-

quence (Scheme 17).

293) gave the 5a-hydroxy

a-gluco-conﬁgured carbasugar pseudo-

disaccharide (297) resulting from attack at C-1 of the epoxide

5.1.3. 5a-Carba-

b-lyxo-Type unsaturated donors are opened by amine

attack exclusively at C-1 to give good yields of N-linked -lyxo

a-xylo-hex(5,5a)enopyranosides

and trans-diaxial opening. This compound was converted into the

required

a-xylo C-5@C-5a unsaturated carbasugar pseudodisac-

a

charide (298) by chlorination of the free hydroxyl group (OH-5a,

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

2305

Table 12

Formation of pseudodisaccharides and pseudotrisaccharides by opening of 1,2-epoxides

Entry Epoxide

Alcohol

Conditions

Product

Yield

Ref.

OH

BnO

HO

OBn

O

BnO

O

BnO

AllO

O

1

A

37% (58% alcohol

recovered)

73

252

OBn

O

OMe

BnO

AllO

1 (1 equiv.)

246

OMe

HO

BnO

OBn

O

HO

BnO

HO

OBn

O

2

3

1 (2 equiv)

1 (1 equiv)

B, 70 °C, 2 h

65%

10

O

BnO

OMe

HO

247

253

OMe

OH

BnO

OBn

O

HO

BnO

B, 70 °C, 4 h

45% (46% alcohol

recovered)

35,73

O

254

BnO

OMe

248

BnO

OMe

Ph

O

OBn

O

HO

O

Ph

O

OBn

O

4

1 (2.8 equiv)

B, 70 °C, 2 h

64% (27% alcohol

recovered)

10,73

HO

BnO

OMe

249

OBn

255

MeO

OH

BnO

O

OH

BnO

OBn

O

BnO

O

5

1 (3 equiv)

BnO

B, 70 °C, 2 days

44%

10,73

BnO

OBn

250

OBn

O

BnO

MeO

OBn 256

MeO

OH

BnO

OH

O

BnO

O

6

7

8

9

1 (2.4 equiv)

B, 70 °C, 2 h

82%

74

257

O(CH₂)₇CH₃

O

OBn

BnO

251

O(CH₂)₇CH₃

OBn

Ph

OH

O

OBn

Ph

O

OBn

O

HO

BnO

B, 70 °C, 3–26 h

66%

14,75

76

O(CH₂)₇CH₃

NHAc

BnO

O

O(CH₂)₇CH₃

NHAc

269a

BnO

258 (3 equiv.)

259

OBn

BnO

O

HO

O

BnO

HO

O

258 (3 equiv)

B, 80 °C, 3 days

36–48%

O(CH₂)₇CH₃

O

NHAc

O(CH₂)₇CH₃

OBn

AcHN

Ph

269b

260

OAc

O

OBn

BnO

258 (2 equiv)

248

i) B, 50 °C,

overnight

ii) Acetylate

60%

O

BnO

269c

BnO

OMe

(continued on next page)

2306

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

Table 12 (continued)

Entry Epoxide

Alcohol

Conditions

Product

Yield

52%

Ref.

77

Ph

OH

O

BnO

O

10

258

B, 80 °C, 6 h

O

OH

261

269d

O

Ph

OH

O

OBn

O

OBn

O

BnO

HO

BnO

11

12

13

14

258 (2 equiv)

258 (2.5 equiv)

258 (2.3 equiv)

258 (1.5 equiv)

B, 50 °C, overnight

84%

70%

68%

44%

78

O(CH₂)₁₁CH₃

O

OBn

O(CH₂)₁₁CH₃

BnO

269e

262

263

264

OBn

Ph

OH

O

OBn

O

HO

BnO

B, 70 °C, 3 h

12,13

OMe

BnO

O

OBn

OMe

BnO

269f

OBn

OH

O

OBn

O

OBn

O

HO

BnO

B, 70 °C, 25 h

13

OMe

O

NHAc

OMe

BnO

269g

NHAc

Ph

OAc

OBn

O

OBn

HO

BnO

i) B, 60 °C, 27 h

ii) Acetylate

BnO

79

O

265

BnO

269h

OH

O

BnO

OH

O

BnO

15

16

258 (1.5 equiv)

B, 60 °C

56%

45%

79

O

269i

O

BnO

266

BnO

OBn

Ph

O

HO

O

HO

B, 60 °C, 16 h

O

267

O

Ph

269j

OH

OBn

O

OBn

HO

BnO

17

258 (1.5 equiv)

B, 60 °C

17%

79

BnO

O

268

269k

BnO

Reaction conditions: (A) BF₃ꢀOEt₂in CH₂Cl₂, ꢁ150 °C; and (B) NaH, DMF, 15-crown-5, 50–80 °C.

unsaturated pseudodisaccharides (2-epi-valienamine derivatives)

5.1.5. 5a-Carba-b-mannopyranosides

with yields between 64% and 78% (Table 3).

O-Linked compounds are available from the

a-mannosides by

Very good results for

formation have been obtained by palladium-catalysed nucleophilic

substitution of allylic C-1 -chlorides with retention of conﬁgura-

a

-lyxo unsaturated pseudodisaccharide

multistep epimerisation at C-1 (Scheme 10). N-Linked and S-linked

examples are not known.

a

tion. Acetal-protecting groups are necessary for excellent yields

5.1.6. 5a-Carba-b-galactopyranosides

(Table 10).

N-Linked compounds are available directly by attack of a carbo-

hydrate amine at C-1 of a carbasugar 1,2-epoxide, but regioselec-

tivity can be problematic (Table 2). Product yields are lower than

5.1.4. 5a-Carba-

a-xylo-hex(5,5a)enopyranosides

Very good results for the formation of N-linked carba-

pseudodisaccharides (valienamine derivatives) have been ob-

tained by palladium-catalysed nucleophilic substitution of allylic

a

-xylo

those for the formation of carba-

24% and 66%. O-Linked compounds have only been synthesised

by extensive modiﬁcation of carba- -mannosides at the pseudodi-

a-mannosides, being between

a

C-1

a

-chlorides with retention of conﬁguration. Again, acetal

saccharide level, that is, epimerisation of the carbamannose at C-1,

C-2 and C-4;^{12–14,75,78}or by Mootoo’s post-conjugation carbasugar

assembly (Scheme 16).

protecting groups are necessary for the obtention of excellent

yields (Table 10).

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

2307

oxidation

PCC, CH₂Cl₂

or

epimerisation

DBU, toluene

or

Ph

OH

O

Ph

^tBuOK, ^tBuOH

O

DMSO, Ac₂O

O

BnO

OR

BnO

a–k

71–99%

a–k

44–93%

OR

O

269

270

271

Ph

O

Ph

O

Ph

OH

O

OR

BnO

OR

BnO

HO

OH

272

OR

274

273

f : A, manno : gluco 1:2

f : B, gluco : 77%

g : A, gluco : 66%; manno : 22%

h : A : gluco, 43%; manno, 28%

j : A : gluco, 64%; manno, 28%

k : A : gluco, 53%; manno, 31%

a : C, gluco, 78%; manno, 15%

a : D, gluco, 47%; manno, 50%

b : D, manno : gluco, 1:1

c : C, gluco, 69%; manno, 15%

d : C, gluco, 61%

e : C, gluco, 52%; manno, 9%

f : A, manno, 42%; gluco, 25%

f : B, manno, 67%

Reduction conditions

A : NaBH₄, CH₂Cl₂, MeOH, 0 ºC

B : L-selectride, THF

C : borane, THF, 0 ºC

D : NaBH₄, MeOH, CeCl₃

g : A, gluco, 43%; manno, 44%

h : D : gluco, 73%; manno, 22%

i : A : gluco, 39%; manno, 56%

j : D: gluco, 50%; manno, 43%

k : D : gluco, 47%; manno, 38%

OBn

O

R =

BnO

R =

O(CH₂)₇CH₃

BnO

OBn

NHAc

79

14,75

h

a

O

R =

O(CH₂)₁₁CH₃

BnO

e

OBn

78

O

BnO

R =

O(CH₂)₇CH₃

NHAc

R =

OBn

14,75

79

b

i

O

R =

OMe

BnO

OBn

Ph

O

12,13

f

O

R =

BnO

OBn

O

76

OMe

79

c

j

R =

NHAc

OBn

13

g

O

R =

BnO

77

O

d

79

k

Scheme 10. Epimerisation of C-1 and/or C-2 after pseudodisaccharide formation.

5.1.7. 5a-Carba-b-glucopyranosides

pseudodisaccharides are available from the carba-

a

-mannosides

For the synthesis of N-linked pseudodisaccharides by opening

of 1,2-carbasugar epoxides by amines, the regioselectivity is not

by the well-worked-out sequence of double epimerisation at C-1

and C-2 (Scheme 10).

as good as that for the synthesis of carba-a-mannosides, but attack

at C-1 has been observed (only one example; Table 2). O-Linked

5.2. Carbasugar nucleophiles

i) PCC,

CH₂Cl₂BnO

BnO

ii) NaBH₄,

MeOH,

CeCl₃

This section summarises those reactions in which the new C–X

bond is not the carbasugar C-1–X bond, corresponding to discon-

nection 2 in Scheme 2.

BnO

OH

BnO

OBn

HO

O

BnO

254

BnO

275 / 45%

(+ 254 / 38%)

BnO

OMe

5.2.1. 4-Substituted glucose

For the formation of N-linked pseudodisaccharides,

conformationally unlocked galacto 3,4-epoxides give mixtures of

BnO

OMe

Scheme 11. Epimerisation of C-2 after pseudodisaccharide formation.^35,73

2308

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

BzO

OBz

O

BzO

276

Figure 6. A b-talo epoxide proposed by Ogawa.

O

BnO

HO

O

TIBAL, PhMe,

50 ºC, 4 h

BnO

O

BnO

OR

O

OR

277a–e

SPh

279

278a–e

BnO

OBn

O

BnO

O

BnO

i) TIBAL, PhMe,

50 ºC, 4 h

ii) Ac₂O, DMSO

33%

OMe

BnO

b / 11%

(+ debenzylated 48%)

BnO

OBn

c / 65%

OMe

a

/ 63%

(+ debenzylated 26%)

O

OBn

BnO

O

BnO

SPh

OMe

BnO

O

BnO

OBn

e / 60%

OBn

d / 65%

BnO

O

BnO

O

SPh

BnO

280

OBn

Scheme 12. Glycoside to pseudodisaccharide and pseudotrisaccharide rearrangements.

HO

O

BnO

AlMe₃, PhMe,

60 ºC, 2 h

BnO

PCC,

CH₂Cl₂

OBn

BnO

HO

O

OR

O

OR

277a

BnO

282 BnO

287 / 75%

281

BnO

OMe

OBn

O

R =

Tebbe reagent

56%

BnO

OMe

Scheme 15. Glycoside to pseudodisaccharide rearrangement.

H₂C

i) BH₃. THF

BnO

HO BnO

BnO

OBn

ii) NaOH,

H₂O₂

saccharides have also been satisfactorily formed by the displace-

ment of galacto C-4 triﬂates with vicinal benzoate protection:

45–55% yields have been obtained with b-valienamine N-1 nucle-

ophiles, forming (1?4)Glc linkages with inversion of conﬁguration

at C-4 (Table 9).

O

BnO

O

BnO

284

48%

BnO

283

OMe

BnO

OMe

Scheme 13. Conversion of pseudodisaccharides into carbaidose-containing

carbadisaccharides.

i) ROH, esterification,

80%

O

OBn

O

OBn

regioisomers when attacked by carbasugar C-1 amines (Table 4).

The yields of the diequatorial gluco-conﬁgured coupled products

are between 19% and 37%. Conformationally locked (1,6-anhydro)

galacto 3,4-epoxides give much better regioselectivity when at-

tacked by carbasugar C-1 amines (Table 5). The yields of the diaxial

gluco-coupled product are typically 47–76%. N-Linked pseudodi-

O

ii) Tebbe reagent,

86%

OH

OR

SPh

289

MeOTf,

DTBMP,

CH₂Cl₂

75%

288

O

OBn

O

i) BH₃.SMe₂

OBn

OH

OR

i) Swern oxidation

ii) pyridine, MeOH,

50 ºC, 32 h

BnO

ii) H₂O₂, 78%

O

BnO

OR

OH

291

BnO

HO

290

O

BnO

O

iii) NaBH₄, THF, H₂O

OBn

BnO

O

BnO

O

BnO

O

R =

OBn

54% over 3 steps

BnO

OBn

BnO

285

OMe

286

Scheme 14. C-5 (ido?gluco) Epimerisation of

a

glycosidically linked

Scheme 16. Formation of a b-galacto carbasugar pseudodisaccharide by post-

carbadisaccharide.

conjugation carbocylisation.

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

2309

BzO

O

292

OH

BzO

1,4-dithioerythritol,

2-aminoethanethiol,

i) Ac₂O, DMSO

+

OBz

HO

OBz

O

ii) NaBH₄, CeCl₃,

MeOH

S

HMPA, RT, 2 h

O

BzO

AcS

BzO

294 / 89%

BzO

295 / 52%

OMe

BzO

OMe

293 (1.5 equiv.)

Ph

O

BzO

O

OBz

OH

296

+

O

BzO

O

Ph

1,4-dithioerythritol,

2-aminoethanethiol,

O

i) SOCl₂, pyridine

ii) DBU, PhMe

OBz

O

BzO

OBz

BzO

O

S

BzO

S

HMPA, RT, 24 h

O

AcS

BzO

298 / 40%

BzO

297 / 37%

BzO

OMe

BzO

OMe

293 (1.7 equiv.)

Scheme 17. Epoxide opening to form the carbasugar C-1–S bond, and modiﬁcation of carbasugar pseudodisaccharides.³⁵

with carbasugar C-1 alcohol nucleophiles (Table 11). HMPA is

needed as solvent for good yields.

BnO

SH

BnO

299 (1.1 equiv)

5.2.2. 3-Substituted glucose

S

O

DBU, toluene,

RT, 6 h

BzO

An electrophile, an allofuranose C-3 triﬂate (235), has been

identiﬁed that leads to an O-linked (1?3)Glc derivative with a car-

basugar C-1 alcohol nucleophile (only one example; Table 11). An

N(1?3)-linked glucose derivative is formed by S_N2 reaction of a

benzoate-protected allo triﬂate (196) and a carbasugar N-1 amine

nucleophile (only one example; Table 9).

BnO

OMe

+

OBn

OBz

TfO

BzO

300/ 72%

O

OMe

OBz

192

Scheme 18. C-1–S bond formation by S_N2 reaction between carbasugar C-1 thiol

and carbohydrate C-4 triﬂate.⁶²

5.2.3. 2-Substituted glucose

Conformationally unlocked manno-conﬁgured 2,3-epoxides

have been used to synthesise N(1?2)-linked glucose derivatives

(Table 6). The regioselectivity for the diequatorial 2-amino-2-

deoxyglucose derivative over the diaxial 3-amino-3-deoxyaltrose

is better when the epoxide is not conformationally locked by a

4,6-benzylidene acetal. A 3-amino-3-deoxyaltrose pseudodisac-

charide was converted into the 2-amino-2-deoxyglucose isomer

after coupling (Scheme 7).

An electrophile, a 1,6-anhydrosugar galacto C-4 triﬂate (234),

has been identiﬁed that leads to O-linked (1?4)Glc derivatives

OBn

BnO

OTf

BnO

Ph

BnO

301 (1.6 equiv.)

O

BnO

i)NaH, DMF,

50 ^ºC

ii) acetylate

5.2.4. 6-Substituted glucose

S

OBn

+

A glucose C-6 triﬂate (233) has been identiﬁed that leads to O-

linked (1?6)Glc derivatives with carbohydrate C-1 alcohol nucle-

ophiles (Table 11).

O

BnO

OBn

304 / 47%

O

Ph

O

AcO

MeO

HS

HO

OMe

302

5.2.5. 4-Substituted mannose

An N-linked derivative was formed from a 4-substituted glu-

cose pseudodisaccharide by multistep epimerisation (only one

example, Scheme 7).

OBn

BnO

Ph

O O

OTf

BnO

5.2.6. 4-Substituted xylose

BnO

301

NaH, DMF,

50 ºC

S

An S-linked xylose derivative was formed by direct S_N2 dis-

placement of a C-4 triﬂate with a thiol nucleophile (only one exam-

ple, Scheme 18).

OBn

+

OBn

O

BnO

OBn

Ph

O

305 / 54%

MeO

HS

5.3. Conclusions and outlook

OMe

303 (1.6 equiv.)

Many synthetic routes to non-glycosidically linked carbasugar

pseudodisaccharides and pseudooligosaccharides have been

Scheme 19. C-1–S bond formation by S_N2 reaction between carbasugar C-1 triﬂate

and carbohydrate C-3 thiols.⁸⁷

2310

I. Cumpstey / Carbohydrate Research 344 (2009) 2285–2310

28. Ogawa, S.; Yasuda, K.; Takagaki, T.; Iwasawa, Y.; Suami, T. Carbohydr. Res. 1985,

141, 329–334.

explored. The chemistry summarised here that leads to this group

of structures is diverse, and includes different coupling reactions

associated with each of nitrogen, oxygen and sulfur nucleophiles.

The presence of C-5@C-5a unsaturation in the ring can be impor-

tant in enhancing the reactivity of an electrophile (e.g., in epoxide

opening) or introducing new reactivity (e.g., in palladium cou-

pling). The reactions often appear to suffer from sensitivity to the

sterically crowded and electron-poor nature of both nucleophile

and electrophile, so that chemistry that may work well on simple

systems can fail here.

The coupling reactions can be classiﬁed as indirect routes,

which rely on post-coupling modiﬁcation (e.g., epimerisation or

reduction) to get the carbasugar or carbohydrate into its ﬁnal form,

and direct routes, where this is not necessary. Direct routes sim-

plify already complex reaction sequences, and so such methods

would be preferable for all linkages. Such direct routes exist in

29. Cottaz, S.; Brimacombe, J. S.; Ferguson, M. Carbohydr. Res. 1993, 247, 341–345.

30. Ogawa, S.; Sugizaki, H. Carbohydr. Res. 1986, 156, 264–272.

31. Ogawa, S.; Tsunoda, H. Liebigs Ann. Chem. 1993, 755–769.

32. Ogawa, S.; Shibata, Y. J. Chem. Soc., Chem. Commun. 1988, 605–606.

33. Shibata, Y.; Ogawa, S. Carbohydr. Res. 1989, 189, 309–322.

34. Ogawa, S.; Uchida, C.; Shibata, Y. Carbohydr. Res. 1992, 223, 279–286.

35. Tsunoda, H.; Sasaki, S.-I.; Furuya, T.; Ogawa, S. Liebigs Ann. 1996, 159–165.

36. Ogawa, S.; Iwasawa, Y.; Toyokuni, T.; Suami, T. Carbohydr. Res. 1985, 141, 29–

40.

37. Ogawa, S.; Iwasawa, Y.; Toyokuni, T.; Suami, T. Chem. Lett. 1983, 337–340.

38. Ogawa, S.; Sugizaki, H.; Iwasawa, Y.; Suami, T. Carbohydr. Res. 1985, 140, 325–

331.

39. Shibata, Y.; Kosuge, Y.; Mizukoshi, T.; Ogawa, S. Carbohydr. Res. 1992, 228, 377–

398.

40. Paulsen, H.; Röben, W. Liebigs Ann. Chem. 1985, 974–994.

41. Paulsen, H.; Mielke, B. Liebigs Ann. Chem. 1990, 169–180.

42. Ogawa, S.; Shibata, Y.; Kosuge, Y.; Mizukoshi, T.; Uchida, C. J. Chem. Soc., Chem.

Commun. 1990, 1387–1388.

43. McAuliffe, J. C.; Stick, R. V.; Stone, B. A. Tetrahedron Lett. 1996, 37, 2479–2482.

44. McAuliffe, J. C.; Stick, R. V. Aust. J. Chem. 1997, 50, 219–224.

45. Stick, R. V.; Tilbrook, D. M. G.; Williams, S. J. Aust. J. Chem. 1999, 52, 885–894.

46. Ogawa, S.; Aso, D. Carbohydr. Res. 1993, 250, 177–184.

47. Ogawa, S.; Sekura, R.; Maruyama, A.; Odagiri, T.; Yuasa, H.; Hashimoto, H.

Carbohydr. Lett. 2000, 4, 13–20.

some cases. For example, the synthesis of a-manno-conﬁgured car-

basugars with N-, S- or O-linkages by a one-step 1,2-epoxide open-

ing seems apparently generally applicable.

Some other direct routes, such as the palladium-catalysed cou-

48. Ogawa, S.; Sugizaki, H. Chem. Lett. 1986, 1977–1980.

49. Yamagishi, T.; Uchida, C.; Ogawa, S. Bioorg. Med. Chem. Lett. 1995, 5, 487–490.

50. Yamagishi, T.; Uchida, C.; Ogawa, S. Chem. Eur. J. 1995, 1, 634–636.

51. Ogawa, S.; Ashiura, M.; Uchida, C. Carbohydr. Res. 1998, 370, 83–95.

52. Hayashida, M.; Sakairi, N.; Kuzuhara, H. Carbohydr. Res. 1986, 158, C5–C8.

53. Hayashida, M.; Sakairi, N.; Kuzuhara, H. Carbohydr. Res. 1989, 194, 233–246.

54. Haines, A. H.; Carvalho, I. Chem. Commun. 1998, 817–818.

55. Carvalho, I.; Haines, A. H. J. Chem. Soc., Perkin Trans. 1 1999, 1795–1800.

56. McAuliffe, J. C.; Stick, R. V.; Tilbrook, D. M. G.; Watts, A. G. Aust. J. Chem. 1998,

51, 91–95.

pling routes to N-linked

a-gluco carbasugar pseudodisaccharides

and the formation of 4-O-substituted glucose derivatives by triﬂate

displacement, are also apparently very efﬁcient but maybe some-

what less explored. It would be desirable to see the development

of direct coupling routes for those cases where none exists already.

For example, new shorter and general routes to O-linked

a- and b-

galacto or N-linked -galacto derivatives would be welcome. The

a

biological activity of at least O- and N-linked carbasugar pseudodi-

saccharides justiﬁes the further exploration of this type of struc-

ture, and hopefully in the future, efﬁcient coupling methods for

the synthesis of all the relevant linkages can be found.

57. Horii, S.; Fukase, H.; Matsuo, T.; Kameda, Y.; Asano, N.; Matsui, K. J. Med. Chem.

1986, 29, 1038–1046.

58. Ogawa, S.; Funayama, S.; Okazaki, K.; Ishizuka, F.; Sakata, Y.; Doi, F. Bioorg. Med.

Chem. Lett. 2004, 14, 5183–5188.

59. Ogawa, S.; Toyokuni, T.; Suami, T. Chem. Lett. 1981, 947–950.

60. Toyokuni, T.; Ogawa, S.; Suami, T. Bull. Chem. Soc. Jpn. 1983, 2999–3004.

61. Sakairi, N.; Kuzuhara, H. Tetrahedron Lett. 1982, 23, 5327–5330.

62. McDonough, M. J.; Stick, R. V.; Tilbrook, D. M. G. Aust. J. Chem. 1999, 52, 143–

147.

63. Fairweather, J. K.; McDonough, M. J.; Stick, R. V.; Tilbrook, D. M. G. Aust. J. Chem.

2004, 57, 197–205.

64. McDonough, M. J.; Stick, R. V.; Tilbrook, D. M. G.; Watts, A. G. Aust. J. Chem.

2004, 57, 233–241.

References

1. Chen, X.; Fan, Y.; Zheng, Y.; Shen, Y. Chem. Rev. 2003, 103, 1955–1977.

2. Ogawa, S. Trends Glycosci. Glyc. 2004, 16, 39–53.

3. Ogawa, S.; Kanto, M.; Suzuki, Y. Mini-Rev. Med. Chem. 2007, 7, 679–691.

4. Fernández-Alonso, M. C.; Cañada, F. J.; Solís, D.; Cheng, X.; Kumaran, G.; André,

S.; Siebert, H.-C.; Mootoo, D. R.; Gabius, H.-J.; Jiménez-Barbero, J. Eur. J. Org.

Chem. 2004, 1604–1613.

5. Suami, T.; Ogawa, S. Adv. Carbohydr. Chem. Biochem. 1990, 48, 21–90.

6. Arjona, O.; Gómez, A. M.; López, J. C.; Plumet, J. Chem. Rev. 2007, 107, 1919–

2036.

7. Streitweiser, A.; Jayasree, E. G.; Leung, S. S.-H.; Choy, G. S.-C. J. Org. Chem. 2005,

70, 8486–8491.

65. Stick, R. V.; Tilbrook, D. M. G.; Williams, S. J. Aust. J. Chem. 1999, 52, 895–904.

66. McAuliffe, J. C.; Stick, R. V. Aust. J. Chem. 1997, 50, 203–207.

67. Shing, T. K. M.; Kwong, C. S. K.; Cheung, A. W. C.; Kok, S. H. L.; Yu, Z.; Li, J.;

Cheng, C. H. K. J. Am. Chem. Soc. 2004, 126, 15990–15992.

68. Shing, T. K. M.; Cheng, H. M.; Wong, W. F.; Kwong, C. S. K.; Li, J.; Lau, C. B. S.;

Leung, P. S.; Cheng, C. H. K. Org. Lett. 2008, 10, 3145–3148.

69. Shing, T. K. M.; Cheng, H. M. Org. Lett. 2008, 10, 4137–4139.

70. Knapp, S.; Naughton, A. B. J.; Dhar, T. G. M. Tetrahedron Lett. 1992, 33, 1025–

1028.

71. Paulsen, H.; von Deyn, W. Liebigs Ann. Chem. 1987, 141–152.

72. Akhtar, T.; Cumpstey, I. Tetrahedron Lett. 2007, 48, 8673–8677.

73. Ogawa, S.; Sasaki, S.-I.; Tsunoda, H. Chem. Lett. 1993, 1587–1590.

74. Ogawa, S.; Furuya, T.; Tsunoda, H.; Hindsgaul, O.; Stangier, K.; Palcic, M. M.

Carbohydr. Res. 1995, 271, 197–205.

75. Ogawa, S.; Matsunaga, N.; Palcic, M. Carbohydr. Lett. 1997, 2, 299–306.

76. Ogawa, S.; Ohmura, M.; Hisamatsu, S. Synthesis 2001, 312–316.

77. Ogawa, S.; Hirai, K.; Ohno, M.; Furuya, T.; Sasaki, S.-I.; Tsunoda, H. Liebigs Ann.

1996, 673–677.

78. Ogawa, S.; Aoyama, H.; Sato, T. Carbohydr. Res. 2002, 337, 1979–1992.

79. Ogawa, S.; Senba, Y. J. Carbohydr. Chem. 2006, 25, 69–93.

80. Pearce, A. J.; Sollogoub, M.; Mallet, J.-M.; Sinaÿ, P. Eur. J. Org. Chem. 1999, 11,

2103–2117.

81. Pearce, A. J.; Mallet, J. M.; Sinaÿ, P. Heterocycles 2000, 52, 819–826.

82. Pearce, A. J.; Chevalier, R.; Mallet, J. M.; Sinaÿ, P. Eur. J. Org. Chem. 2000, 2203–

2206.

83. Sollogoub, M.; Pearce, A. J.; Hérault, A.; Sinaÿ, P. Tetrahedron: Asymmetry 2000,

11, 283–294.

84. López-Méndez, B.; Jia, C.; Zhang, Y.; Zhang, L.-H.; Sinaÿ, P.; Jiménez-Barbero, J.;

Sollogoub, M. Chem. Asian J. 2008, 3, 51–58.

85. Jia, C.; Pearce, A. J.; Blériot, Y.; Zhang, Y.; Zhang, L.-H.; Sollogoub, M.; Sinaÿ, P.

Tetrahedron: Asymmetry 2004, 15, 699–703.

8. Shaik, S. J. Am. Chem. Soc. 1983, 105, 4359–4367.

9. Fürst, A.; Plattner, P. A. Helv. Chim. Acta 1949, 32, 275–283.

10. Ogawa, S.; Sasaki, S.-I.; Tsunoda, H. Carbohydr. Res. 1995, 274, 183–196.

11. Ogawa, S.; Nakamura, Y. Carbohydr. Res. 1992, 226, 79–89.

12. Ogawa, S.; Hirai, K.; Yamazaki, T.; Nakajima, A.; Matsunaga, N. Carbohydr. Lett.

1996, 2, 183–188.

13. Ogawa, S.; Hirai, K.; Odagiri, T.; Matsunaga, N.; Yamazaki, T.; Nakajima, A. Eur.

J. Org. Chem. 1998, 1099–1109.

14. Ogawa, S.; Matsunaga, N.; Li, H.; Palcic, M. Eur. J. Org. Chem. 1999, 631–642.

15. Okazaki, K.; Nishigaki, S.; Ishizuka, F.; Kajihara, Y.; Ogawa, S. Org. Biomol. Chem.

2003, 1, 2229–2230.

16. Ogawa, S.; Fujieda, S.; Sakata, Y.; Ishizaki, M.; Hisamatsu, S.; Okazaki, K.; Ooki,

Y.; Mori, M.; Itoh, M.; Korenaga, T. Bioorg. Med. Chem. 2004, 12, 6569–6579.

17. Shibata, Y.; Kosuge, Y.; Ogawa, S. Carbohydr. Res. 1990, 199, 37–54.

18. Ogawa, S.; Miyamoto, Y. J. Chem. Soc., Chem. Commun. 1987, 1843–1844.

19. Ogawa, S.; Miyamoto, Y.; Nose, T. J. Chem. Soc., Perkin Trans. 1 1988, 2675–2680.

20. Miyamoto, Y.; Ogawa, S. J. Chem. Soc., Perkin Trans. 1 1989, 1013–1018.

21. Miyamoto, Y.; Ogawa, S. Chem. Lett. 1988, 889–890.

22. Ogawa, S.; Sato, K.; Miyamoto, Y. J. Chem. Soc., Perkin Trans. 1 1993, 691–696.

23. Ogawa, S.; Toyokuni, T.; Iwasawa, Y.; Abe, Y.; Suami, T. Chem. Lett. 1982, 279–

282.

24. Ogawa, S.; Ogawa, T.; Chida, N.; Toyokuni, T.; Suami, T. Chem. Lett. 1982, 749–

752.

25. Ogawa, S.; Iwasawa, Y.; Toyokuni, T.; Suami, T. Chem. Lett. 1982, 1729–1732.

26. Ogawa, S.; Ogawa, T.; Iwasawa, Y.; Toyokuni, T.; Chida, N.; Suami, T. J. Org.

Chem. 1984, 49, 2494–2599.

86. Cheng, X.; Khan, N.; Kumaran, G.; Mootoo, D. R. Org. Lett. 2001, 3, 1323–1325.

87. Cumpstey, I.; Alonzi, D. S.; Butters, T. D. Carbohydr. Res. 2009, 344, 454–459.

27. Ogawa, S.; Iwasawa, Y.; Toyokuni, T.; Suami, T. Carbohydr. Res. 1985, 144, 155–

162.

Article Doi

Synthesis of carbasugar-containing non-glycosidically linked pseudodisaccharides and higher pseudooligosaccharides

DOI: 10.1016/j.carres.2009.09.008

Source and publish data:

Authors:

Article abstract of DOI:10.1016/j.carres.2009.09.008

Full text of DOI:10.1016/j.carres.2009.09.008

Products guided by the article

R&D Labs maybe for 198202-24-3

Relevant to this article

Hot Product